Post on 07-Jan-2022
AMPHOLYTE MICROGELS WITH CONTROLLED
DISTRIBUTION OF IONIZABLE GROUPS
VON DER FAKULTÄT FÜR MATHEMATIK, INFORMATIK UND NATURWISSENSCHAFTEN DER
RWTH AACHEN UNIVERSITY ZUR ERLANGUNG DES AKADEMISCHEN GRADES EINER DOKTORIN
DER NATURWISSENSCHAFTEN GENEHMIGTE DISSERTATION
VORGELEGT VON
DIPLOM-CHEMIKERIN
RICARDA SCHRÖDER
AUS DÜSSELDORF-HEERDT, DEUTSCHLAND
BERICHTER:
UNIVERSITÄTSPROFESSOR PROF. DR. ANDRIJ PICH
UNIVERSITÄTSPROFESSOR PROF. DR. WALTER RICHTERING
TAG DER MÜNDLICHEN PRÜFUNG: 17.02.2016
DIESE DISSERTATION IST AUF DEN INTERNETSEITEN DER UNIVERSITÄTSBIBLIOTHEK ONLINE
VERFÜGBAR.
II
III
Für meine Eltern
IV
Table of Contents
1 Introduction…………………………………………………………………………. 1
2 Thesis Outline……………………………………………………………………….. 3
3 State of the Art……………………………………………………………………… 5
3.1. Microgels 5
3.2. Poly-N-vinylcaprolactam and Poly-N-isopropylacrylamide 8
3.3. PVCL and PNIPAm microgels 11
3.4. Literature 16
4 Experimental Part………………………………………………………………….. 20
4.1. Chemicals 20
4.2. Analytical Instrumentation and Experimental Procedures 20
4.3. Microgel Synthesis 24
PART I: Synthesis and Characterization
5 Zwitterionic Microgels…………………………………………………...………… 25
5.1. Introduction 25
5.2. Synthesis Procedure 28
5.3. Precipitation Polymerization…………………………………...……... 30
5.3.1. Introduction 30
5.3.2. Synthesis Procedure 32
5.3.3. Results and Discussion 33
5.3.4. Summary 43
5.4. Emulsion Polymerization…………………………………...………….. 44
5.4.1. Introduction 44
5.4.2. Synthesis Procedure 47
5.4.3. Results and Discussion 49
5.4.4. Summary 59
5.5. Literature 60
6 Ampholyte Microgels with Statistical Distribution of Ionizable Groups………. 63
6.1. Introduction 63
6.2. Synthesis Procedure 66
6.3. Results and Discussion 67
6.4. Summary 76
6.5. Literature 77
V
7 Ampholyte Microgels with Core-Shell Structure……………………………….. . 79
7.1. Introduction 79
7.2. Synthesis Procedure 82
7.3. Results and Discussion 84
7.4. Summary 92
7.5. Literature 93
8 Ampholyte Janus-like Microgels………………………………………………….. 94
8.1. Introduction 94
8.2. Synthesis Procedure 97
8.3. Results and Discussion 99
8.4. Summary 105
8.5. Literature 105
PART II: Interactions with Proteins
9 Interactions with Proteins………………………………………………………….. 106
9.1. Introduction 106
9.2. Experimental Part 109
9.3. Results and Discussion 110
9.4. Summary 122
9.5. Literature 123
PART III: Microgel-Based Materials
10 Biomineralization / Biohybrid systems………………………………………...….. 124
10.1. Growth of CaCO3 in Zwitterionic Microgels………………………..... 124
10.1.1. Introduction 124
10.1.2. Synthesis Procedure 127
10.1.3. Results and Discussion 130
10.1.4. Summary 142
10.1.5. Literature 143
10.2. Microgels and β-TCP as Bone Substitute Material…………...……… 144
10.2.1. Introduction 144
10.2.2. Synthesis Procedure 147
10.2.3. Results and Discussion 148
10.2.4. Summary 155
10.2.5. Literature 156
VI
11 Coating of Surfaces: Protein-Repellant Surfaces with Zwitterionic Microgels… 158
11.1. Introduction 158
11.2. Experimental Part 159
11.3. Results and Discussion 160
11.4. Summary 166
11.5. Literature 167
12 Summary……………………………………………………………………...…….. 168
13 Nomenclature……………………………………………………………………….. 171
14 Appendix…………………………………………………………………………….. 173
Danksagung 178
Schlusserklärung 179
VII
1
1 Introduction
While manifold examples of the synthesis and characterization of ampholyte microgels can be
found in literature, only little attention has been paid to the spatial distribution of ionizable
groups within the three-dimensional polymer network of microgel particles and their
influence on the particles’ properties. Often, microgels are praised as smart drug delivery
systems. In this regard, the location of ionizable groups in the microgels is an essential
parameter to control and predict effective release of drugs and proteins.
T. Hoare and R. Pelton were the first to introduce the problem of different distributions of
ionizable groups. They synthesized microgels based on poly-N-isopropylacrylamide
(PNIPAm) with different acids as comonomers and showed that different microgel
morphologies were obtained depending on the type of comonomers used due to different
reactivity parameters of the monomers [1]. They summarized that a comonomer that reacts
slower than NIPAm is more localized on or close to the microgel surface.
A later work by the same authors analyzed the influence of the distribution of ionizable
groups in PNIPAm-based microgels on the uptake and release of various drugs [2]. Here, the
microgels were functionalized with different acids such as acrylic acid, methacrylic acid,
vinylacetic acid, and fumaric acid. Again, the different reactivity ratios between the respective
acid and NIPAm resulted in microgels with varying distribution of ionizable groups.
Microgels containing vinylacetic acid led to particles with a high surface-functionalization,
while methacrylic acid containing microgels led to particles with ionizable groups located in
the core of the particle. Uptake and release experiments revealed that a higher amount of
drugs could be loaded in the microgels when the charges were situated in the core. Charges on
the surface of the particles lead to a so-called “skin layer” on the surface of the particles that
prevents further uptake of drugs.
In 2006, T. Hoare and D. McLean attempted to develop a kinetic model in order to predict the
distribution of ionizable groups in microgel particles [3]. They state that predictions can be
made on copolymerization ratios, homo-polymerization constants, and initial monomer
concentrations alone. A comparison between the radial density diagrams based on kinetic
model predictions and TEM images (in which carboxylic groups are stained with U(Ac)3)
show a good correlation with regard to the distribution of carboxylic groups. The model could
also be used to predict the radial distribution of crosslinker density within the particle.
The work by T. Hoare and R. Pelton show that the distribution of ionizable groups influences
the swelling and electrophoretic properties of microgel particles and is crucial for many
applications such as drug delivery or chemical sensing.
2
1.2 Literature
[1] T. Hoare, R. Pelton, Langmuir 2004, 20, 2123–33.
[2] T. Hoare, R. Pelton, Langmuir 2008, 24, 1005–12.
[3] T. Hoare, D. McLean, J. Phys. Chem. B 2006, 110, 20327–20336.
3
2 Thesis Outline
The aim of this work is the understanding of how the distribution of functional groups (i.e.
basic and acid moieties added through copolymerization) within the polymer network of
environmentally responsive microgels influences the microgels’ properties. Applications of
microgels are manifold and range from encapsulation/controlled release of drugs and genes
[1][2] and chemical sensing and/or filtration in membranes [3][4] to coating for non-biofouling
and protein-repellent surfaces [5][6] and magnetic resonance imaging [7][8]. With regard to each
application, the microgel architecture plays a crucial role as discussed in the previous chapter.
This work is therefore divided into three parts:
(1) Synthesis and characterization of ampholyte microgel particles.
Microgels are based on poly(N-vinylcaprolactam) (PVCL) and poly(N-
isopropylacrylamide) (PNIPAm) with various distributions of basic and acid moieties
within the microgel particle. Here, zwitterionic as well as ampholyte microgels are
synthesized and the functional groups are incorporated either statistically, as core-shell
or Janus-like (see Figure 2.1). Depending on the amount of hydrophilic comonomers
used, either surfactant-free free-radical precipitation polymerization in water or free-
radical inverse mini-emulsion in heptane/water are used as polymerization techniques.
The synthesis of statistical microgels of both zwitterionic and ampholyte nature
requires a one-step synthesis, while core-shell particles are prepared by a two-step
synthesis. For the synthesis of Janus-like microgels, the particle formation in the
beginning of the synthesis is closely studied to predict the right moment for mixing
separate microgel dispersion at the right time so they form Janus-like particles.
Figure 2.1 Various distributions of ionizable groups aimed at in this work.
The influence of the distribution of ionizable groups as well as their amount in the
microgel on the microgels’ properties such as particle size distribution, hydrodynamic
radius, temperature- and pH-sensitivity, and softness is looked at. Another focus is on
the special behavior and the degradation potential of non-covalently crosslinked
zwitterionic microgels after the addition of salt.
4
(2) Interactions of ampholyte microgels with charged species.
The second part looks closely at the influence of different microgel architecture on the
uptake and release of charged species. Subsequently, the size of the charged species is
increased from simple salts to proteins. The microgels’ properties such as
hydrodynamic radius and swelling behavior are greatly influenced by the addition of
charged species. In case of the protein, where cytochrome c is used as a model protein,
not only the uptake is studied, but also the release using various triggers such as
temperature, the addition of salt and a change in pH. Here, the focus is on how
different distributions of ionizable groups lead to different uptake and release
mechanisms.
(3) Microgel-based materials.
Hybrid materials of the above discussed microgels are studied in this part of the work.
First, ionizable microgels are used as nano-containers for biomineralization of CaCO3
and for bone substitution with β-TCP. Furthermore, zwitterionic microgels are
deposited on surfaces to study the influence of the betaine amount on the surface’s
hydrophilicity, protein-repellency, and self-healing properties.
2.1 Literature
[1] X. Li, J. Yuan, H. Liu, L. Jiang, S. Sun, S. Cheng, J. Colloid Interface Sci. 2010, 348,
408–15.
[2] G. Liu, X. Li, S. Xiong, L. Li, P. K. Chu, S. Wu, Z. Xu, J. Fluor. Chem. 2012, 135, 75–
82.
[3] D. Menne, F. Pitsch, J. E. Wong, A. Pich, M. Wessling, Angew. Chemie - Int. Ed.
2014, 53, 5706–5710.
[4] L. Zhang, M. W. Spears, L. A. Lyon, Langmuir 2014, 30, 7628–7634.
[5] A. W. Bridges, N. Singh, K. L. Burns, J. E. Babensee, L. Andrew Lyon, A. J. García,
Biomaterials 2008, 29, 4605–15.
[6] C. M. Nolan, C. D. Reyes, J. D. Debord, A. J. García, L. A. Lyon, Biomacromolecules
2005, 6, 2032–9.
[7] L. Zhang, H. Xue, Z. Cao, A. Keefe, J. Wang, S. Jiang, Biomaterials 2011, 32, 4604–8.
[8] X. Zheng, J. Qian, F. Tang, Z. Wang, C. Cao, K. Zhong, ACS Macro Lett. 2015, 4,
431–435.
5
3 State of the Art
3.1 Microgels
Microgels are intermolecular crosslinked, three-dimensional colloidal polymer networks with
a porous structure and a high capability to absorb water, hence the name “gel”. The amount of
water that can be absorbed ranges from 10 – 90 wt-% and depends on the hydrophilicity,
presence of charges, and the crosslinking density within the microgel particle [1]. Their sizes
lie in the range of 50 nm to 5 µm [1][2]. Therefore, the term “microgel” is misleading since the
part “micro” is referring to the size of the particles. Though “nanogel” would be the more
appropriate term, in this work, the word “microgel” is still used in the following. The color of
microgel dispersions is highly dependent on the particles’ sizes as well as the degree of
swelling: Dispersions with small or collapsed particles are milky white, while dispersions
with larger or highly swollen particles are nearly transparent, since their refractive index is
close to that of water [3]. Though the swelling of microgels in water is a result of their high
hydrophilicity, they do not dissolve in water due to their internal crosslinking. This
crosslinking can be of covalent nature, but also – dependent on the kind of polymers used -
originate from hydrogen bonding, electrostatic interactions or entanglement of polymer chains
[4]. Microgels that are crosslinked covalently are also referred to as chemically crosslinked
particles, while the others are physically crosslinked. The latter ones are only stable under
certain conditions: changes in the environment such as salt or temperature will lead to the
dissolution of microgels into loose polymer chains. Covalently crosslinked microgels can be
obtained by either the copolymerization of a crosslinker such as N,N’-
Methylenebis(acrylamide) (BIS) [5][6] or a chemical reaction between functional groups in the
polymer chain [7][8]. The latter is often done in a post-polymerization and leads to microgels
that are only crosslinked on the surface.
Microgel particles obtain their colloidal stability by three means: steric hindrance,
electrostatic repulsion or a combination of both [9]. Steric hindrance is caused by the loose
ends of polymer chains on the particle’s surface or by post-modification, e.g. via grafting with
PEGMA [10][11] or PEO [9][12]. Electrostatic repulsion is based on the repulsion forces between
charges on the particle surface and is mostly gained by the use of a charged initiator during
the synthesis, e.g. 2,2'Azobis (2-methylpropionamidine) dihydrochloride (AMPA) or the use
of charged surfactants such as sodium dodecyl sulfate (SDS). The third method combines the
preceding two methods. The good colloidal stability goes so far that microgels are able to re-
disperse in water after being freeze-dried or centrifuged [1].
6
Microgels can be divided into non-responsive and stimulus-responsive gels. This chapter will
only focus on the latter. Stimulus-sensitive microgels undergo volumetric changes when
responding to small variations in the environment. This fascinating behavior has been
extensively studied experimentally and theoretically over the past decades. Most studies in
literature concerning sensitive microgels focus on temperature-sensitive microgels, i.e. they
swell or shrink in size dependent on the solution temperature. The temperature at which the
chemical and physical characteristics change dramatically is called Volume Phase Transition
Temperature (VPTT). Typical examples for temperature-sensitive polymers are poly-N-
isopropyl acrylamide (PNIPAm) and poly-N-vinylcaprolactam (PVCL)-based systems. Both
have a VPTT of around 32 °C [13][14]. This closeness to body temperature makes these
monomers therefore very attractive for biomedical applications.
By the use of comonomers further sensitivities for example to pH, ionic strength, light,
magnetic or electric field, can be incorporated. Because their behavior can be controlled and
tuned externally, they are often called “smart materials” [15][16]. In contrast to macroscopic
gels, micro- and nanogels respond very fast to external stimuli, usually within few seconds
[17][18] due to their high surface area-to-volume ratio [19]. The reaction of macroscopic gels
usually is within several hours because their reaction is dependent on the diffusion speed into
the network. Many authors have reported on the relationship between size and response time
[20][21]. The responsiveness of microgels to external stimuli makes them attractive for a wide
range of applications. They range from targeted drug delivery [22][23][24], catalysis [25][26], tissue
engineering [27], to chemical separation processes [28] and color-tunable colloidal crystals [29].
Microgels have a bulk phase behavior very much different than that of hard microspheres
such as polystyrene (PS) [30][31]. They are liquid at low volume fractions, and show crystal [32]
and glassy phases at high volume fractions. With regard to their softness, they are located
between molecular polymers and hard colloidal particles. As a result, they have a higher
packing density that is beyond the random close packing fraction of hard spheres due to their
ability to deform [33].
The softness of microgels can be changed by variation of the amount of crosslinker as well as
through its response to environmental changes like temperature, pH, and ionic strength. A
different approach was used by Z. Zhou et al. who prepared microgels containing a PS core
and a PNIPAm shell [33]. They could tune the softness of the particles by variation in
temperature and shell thickness. The softness of particles greatly influences their
deformability. Contrary to hard spheres, deformation and interpenetration of polymer chains
can occur for microgels. The softness and deformability of particles are an important factor
7
for the stabilization of Pickering emulsions. Classically, Pickering emulsions are emulsions
stabilized by solid particles. In the last years, thermosensitive microgels were used as
stabilizers due to their high deformability in order to obtain responsive emulsions [34]. In
contrast to hard spheres, microgels deform strongly at the oil-water interface and loose their
spherical shape. The extent of deformation is greatly influenced by the degree of crosslinking.
This effect was intensively studied by M. Destribats et al. [35]. They showed that a higher
deformability leads to kinetically well stabilized emulsions, while a higher crosslinking or an
increase in temperature (resulting in an increase in crosslinking density due to particle
shrinkage) reduces the stability. The degree of deformability also becomes important with
regard to substrate deposition. Deformable microgels can be deposited on surfaces that are not
accessible for rigid particles such as PS or silica [19].
8
3.2 Poly-N-vinylcaprolactam and Poly-N-isopropylacrylamide
N-Vinylcaprolactam (VCL) is a seven-membered vinylamide with a lactam ring that is
soluble in both polar and non-polar solvents. It is a white, crystalline solid with a melting
point of 35 - 38 °C [36]. It is amphiphilic due to its hydrophilic carboxylic and amide group
and its hydrophobic vinyl group. N-isopropylacrylamide (NIPAm) is an acrylamide with a
melting point of 60 – 63 °C [37]. Both monomers are commercially available. The structures of
both VCL and NIPAm can be seen in Figure 3.1.
Figure 3.1 Chemical structures of (left) N-Vinylcaprolactam (VCL); (right) N-
isopropylacrylamide (NIPAm).
In comparison to NIPAm, VCL is more biocompatible. When NIPAm hydrolyses, it produces
toxic amines of low molecular weight [38][39]. Hydrolysis of VCL can occur under strong
acidic conditions and results in the formation of caprolactam (see Figure 3.2) [40].
Most articles in literature discuss the polymerization of VCL via radical polymerization
[42][43][44]. M. Hurtgen et al. prepared thermo-responsive block copolymers of VCL and
vinylacetate in a cobalt-mediated radical polymerization [45]. The reported polydispersity of
Ð = 1.1 confirms the advantage of this method. M. Beija et al. report the polymerization of
PVCL via Reversible Addition-Fragementation Chain Transfer Polymerization (RAFT) and
Macromolecular Design via the Interchange of Xanthates (MADIX) [46]. They were able to
synthesize polymers with Mn of 150,000 g/mol and dispersities below Ð < 1.2. A different
approach was followed by S. C. Cheng et al. They prepared PVCL polymers via γ-radiation in
high yields. They demonstrated how the molecular weight could be tuned with the radiation
dose as well as the monomer concentration [47].
Synthesis procedures to obtain un-crosslinked PNIPAm polymers are rare in literature since
most authors focus on the polymer’s thermosensitivity for the preparation of PNIPAm gels. T.
H. Ho et al. obtained PNIPAm polymer chains with low Ð of 1.05 via RAFT polymerization
[48]. There are, however, many reports about using PNIPAm homo- or copolymers as brushes
or monolayer for the stabilization of nanoparticles. X. Jiang et al., for instance, prepared
poly(acrylic acid)-g-poly(N-isopropylacrylamide) graft copolymers to coat water-soluble iron
oxide nanoparticles [49]. The polymer was obtained in three-step synthesis and served as a
9
template for the preparation of well-defined Fe3O4 nanoparticles. It was shown that the length
of the graft polymer chain influenced the size of the nanoparticles. B.-Y. Zhang et al. prepared
a copolymer of PNIPAM and poly(ethylene glycol) via ATRP [50]. In a subsequent step, this
block copolymer was added to poly(N,N-dimethylamino-ethylmethacrylate) in a click
reaction. The final terpolymer was both temperature- and pH-sensitive.
Figure 3.2 Hydrolysis of N-vinylcaprolactam [41].
Both poly-N-vinylcaprolactam (PVCL) and poly-N-isopropylacrylamide (PNIPAm) are non-
ionic, water-soluble, and temperature-sensitive polymers with a lower critical solution
temperature (LCST) at around 32 °C in water [13][51][52]. The temperature-sensitivity is based
on the presence of both hydrophobic (alkyl chains or vinyl backbone) and hydrophilic groups
(amide or carbonyl groups). Below the LCST, non-crosslinked PVCL polymer chains form
random coils in water because its hydrophilic carbonyl group forms hydrogen bonds with the
surrounding water molecules (polymer-solvent interactions) giving a favorable contribution to
the free mixing energy (ΔHm < 0). Simultaneously, the hydrophobic alkyl ring forms
polymer-polymer interactions. Since the hydrogen bonds are stronger than the hydrophobic
interactions, PVCL is soluble in water. When the temperature reaches the LCST, PVCL
undergoes dramatic changes in its interactions with the surrounding water molecules. The
10
polymer-solvent bonds are disrupted and the polymer-polymer interactions become stronger.
PVCL turns from a random coil to a globular polymer that is insoluble in water. As formerly
ordered water molecules are released by this process, an increase in entropy results. This
process is similar for non-crosslinked PNIPAm polymer chains. The LCST can be tuned by
the ratio between hydrophobic and hydrophilic groups in the polymer. The higher the number
of hydrophobic segments, the lower the LCST becomes [21][53][54].
Berghman et al. distinguish three different types of LCST: type I, type II, and type III [55][56].
PVCL shows a type I LCST transition behavior, i.e. the position of the critical point can be
shifted towards lower concentrations by an increase in the polymer chain length [38].
Since PVCL has a type I transition behavior, its LCST is dependent on the molecular weight
of the polymer as well as the concentration. An increase in Mw and concentration leads to a
decrease of the LCST of PVCL [56][46]. M. Beija et al. could show how the cloud point of
PVCL microgels can easily be tuned [46]. By increasing the molecular weight from 18.000 to
150.000 g/mol, the LCST was shifted from 33 to 46 °C. The incorporation of comonomers is
another possibility to influence the LCST of polymers [14][57]. For instance, A. Kermagoret et
al. synthesized copolymers of PVCL with N-vinylamide and vinylester and were able to shift
the LCST from 32 °C to 81 °C and 19 °C for molar fractions of 0.57 % of N-vinylacetamide
and 0.45 % of vinylacetate, respectively [52].
PNIPAm exhibits an LCST in the same range as PVCL. In contrast to PVCL, PNIPAm is a
type II polymer, i.e. its critical point is almost independent of the polymer chain length.
Poly(vinyl methyl ether) (PVME) is a typical polymer having a type III transition behavior.
Type III polymers have two critical points: one for low and one for high polymer
concentrations, thus showing type I and type II behavior, i.e. the LCST of PVME is
depending both on the polymer molar mass and the concentration [58][59].
Polymers having an LCST are of great interest as their behavior resembles the denaturation
process of proteins and enzymes. As I. Bischotberger et al. explain, the globule-to-coil
transition of a polymer is similar to the denaturation of proteins. Above the LCST, a polymer
is in a compact, “folded” state but it unfolds to an expanded coil below the LCST [60] .
Polymers based on PVCL are used in a wide range of applications. The great potential of
PVCL for the preparation of thermosensitive microgels is discussed in the next chapter.
Copolymers of VCL and N-vinylpyrrolidone (VP) are sold by BASF under the trademark
LUVITEC© and are used as hot-melt adhesives, glue sticks and hydrogel adhesives [61][62].
Copolymers of VCL and 1-vinylimidazole (Vim) were used to prepare protein-like
11
copolymers with hydrophobic units in the core and polar units in the shell of a globule [63].
VCL copolymers are also used in hairsprays and hair-fixation resins [64].
12
3.3 PVCL and PNIPAm microgels
The first microgels were synthesized via dispersion polymerization by R. Pelton in 1986 [65].
They obtained PNIPAM microgels with sizes of about 1000 nm below the VPTT.
Generally, PNIPAm and PVCL microgels are obtained in a free radical polymerization at
synthesis temperatures above the LCST of the respective monomer, i.e. 60 – 80 °C. The main
requirement is that the polymer is insoluble at this temperature, otherwise a macrogel instead
of a microgel will form. At this temperature, the water-soluble, free-radical initiator
decomposes and starts the growth of a monomer radical to a polymer chain via radical
propagation (see Figure 3.3). Through the use of an ionic initiator such as AMPA or
potassium persulfate (KPS), the charge at the polymer chain end will contribute to the
resulting microgel’s colloidal stability. In contrast to the synthesis of PNIPAm microgels,
however, initiators such as KPS cannot be used to produce PVCL microgels due to the
sensitivity of the VCL ring to oxidation (see Figure 3.2) [41][66]. Beyond a critical chain length
of the collapsed polymers, phase separation occurs due to the polymer’s insolubility in water
and the aggregates form precursor particles that grow in size by further aggregation, by the
addition of monomer molecules or by the addition of growing oligo-radicals. The reaction
stops when the amount of monomers is exhausted. During the cooling down of the solution at
the end of the polymerization to T < LCST, the microgel network swells in water [1] [67] [26]
[68]. After synthesis, the microgel dispersion needs to be purified due to the presence of
unreacted monomers, surfactants and/or linear or slightly branched polymer chains. Common
techniques are centrifugation, filtration, decantation, and dialysis.
Figure 3.3 Microgel formation during free radical precipitation polymerization.
The synthesis with free radical polymerization technique is advantageous since it produces
monodisperse particles and particle properties such as particle size and charge are easily
controlled [68]. Still there are some limitations that make other polymerization techniques more
attractive. First, the incorporation of comonomers is very much dependent on their
hydrophilicity [67]. Very hydrophilic monomers may not precipitate at the critical chain length
to form precursor particles. Differences in polymerization parameters between different
13
comonomers as well as between monomer and crosslinker lead to the formation of
inhomogeneous particles. Usually, microgels have a higher crosslinking density in the core
than in the outer shell of the particles [69]. Furthermore, J. Gao and B. J. Frisken have shown
that monomers are able to self-crosslink [70]. They could prepare PNIPAM microgels without
the use of the crosslinker BIS and obtained particles with a narrow particle size distribution.
The authors explained this result by chain transfer reactions preferably at the hydrogen atom
on the tert-C of the main chain backbone. This leads to problems regarding the possible
application as a drug delivery system for microgels containing cleavable crosslinkers (with
regard to biocompatibility and removal of particles from the body) where a self-crosslinking
of monomers results only in partial degradation. Another drawback is the encapsulation of
temperature-sensitive drugs during the synthesis due to the high temperatures used during
precipitation polymerization (usually 60 °C [71] - 70 °C [72][73][74]). X. Hu et al. proposed a
different mechanism to obtain microgels via precipitation polymerization at low temperatures
[68]. They assumed that because the LCST of PNIPAm is at 32 °C, nucleation of particles
should be possible at temperatures far below 60 °C, but above 32 °C. Classical precipitation
polymerization is carried out at high temperatures since the number of radicals decreases with
decreasing temperature. X. Hu bypassed this using two different approaches: (a) the use of the
redox initiator system ammonium persulfate (APS)/ N,N,N’,N’-tetramethylethylenediamine
(TEMED) which allows initiation below the thermal decomposition of APS. As already
shown by X. D. Feng, the mechanism is based on a contact charge transfer complex and a
cyclic transition state between APS und TEMED [75]; (b) the decomposition of the initiator
2,2’-azobis isobutyronitrile (AIBN) was realized by UV radiation. Both variations allowed the
preparation of spherical PNIPAM microgels with a narrow particle size distribution.
The size of the particles can be controlled by various means. An increase in the synthesis
temperature leads to the formation of smaller particles [76][77]. Smaller particles can also be
obtained by the addition of an ionic surfactant [5]. D. Dupin et al. describe how the surfactant
absorb onto the PNIPAm surface and thus lead to stabilization at an early stage of particle
formation [78]. A higher concentration of surfactants leads to even smaller particles. Nonionic
surfactants are usually used for steric stabilization [79]. A. Lee et al. used poly(vinyl alcohol)
of various molecular weight and showed that PNIPAm microgels could be stabilized over a
wide ionic strength. If, in contrast, the size of particles is to be increased, electrolytes can be
added during the synthesis procedure. H. Shimizu et al. added sodium chloride to the
monomer solution to obtain particles with sizes between 0.55 and 1.60 µm [80]. An increase in
size, however, is accompanied with an increase in polydispersity. The particle size of
14
PNIPAM microgels prepared by T. Still et al. in a semi-batch emulsion polymerization could
be tuned between 0.8 – 4 µm depending on the charge concentration [5]. Since the particle size
increases linearly with time, the reaction can be stopped at specific times in order to obtain the
desired particle size.
Micro- or hydrogels based on polymers possessing a LCST such as NIPAm and VCL show a
volume phase transition temperature (VPTT) themselves. That means that particles are
swollen and hydrophilic below the VPTT, above the VPTT they collapse and expel water
from the polymer network (see Figure 3.4). Similar to their polymers, the VPTT of microgels
can be tuned by varying the ratio between hydrophobic and hydrophilic comonomers.
Figure 3.4 Hydrodynamic radius RH of PVCL microgels as a function of temperature.
This phenomena for non-ionic microgels was first shown for PNIPAm microgels by
Hirokawa et al. [81]. PNIPAm-based microgels exhibit a VPTT of 32-34 °C [14][81][82]. Below
the VPTT, the behavior of crosslinked PNIPAm is similar to that of non-crosslinked PNIPAm
as described above. Crosslinked microgels are surrounded by structured water molecules and
thus swollen. I. Bischotberger et al. have discussed the increased solubility of the hydrophobic
parts of PNIPAm at low temperatures resulting from the formation of a hydration shell around
the hydrophobic groups that involves stronger hydrogen bonds between the water molecules
than in the bulk [60]. With an increase in temperature, however, contact between the
hydrophobic groups becomes thermodynamically more favorable than contact between the
hydrophobic groups and water [83]. Subsequently, the polymer network undergoes a
conformation change so that hydrophobic groups are surrounded by more hydrophilic groups.
In contrast to non-crosslinked polymers, however, crosslinked polymer networks cannot
dissolve in water above the VPTT due to the crosslinks.
When NIPAm is modified with an additional hydrophobic methyl group, it shows a much
higher phase transition temperature. N-isopropylmethacrylamide (NIPMAm) has an LCST of
15
38 - 45 °C which is attributed to the steric hindrance of the methyl group that hinders
favorable association due to restraints of the chain conformational change [84][85].
Microgels based on PVCL retain its monomers’ reversible thermosensitivity and most authors
agree that PVCL microgels have a VPTT at around 32 °C in water [9][12][41], tough their
transition is not as sharp as the transition of PNIPAm microgels [86][87]. High-sensitivity
differential scanning calorimetric studies by L.M. Mikheeva et al., however, show that PVCL
microgels reveal two separate transition temperatures: at 31.5 °C and at 37.6 °C [83]. The
authors assign the higher transition to the gel volume collapse, while the formation of
hydrophobic domains within the microgel occurs at the lower temperature. These
hydrophobic domains are a result of the amphiphilic character of the PVCL polymer chain. It
consists of cyclic amide with a hydrophobic polymer backbone (see Figure 3.1). The
formation of the hydrophobic domains is thermodynamically more favored than the complete
collapse of the overall gel which occurs only at a higher temperature of 37.6 °C.
The VPTT is influenced by changing the polymer-solvent interactions. This can be realized
by a variation in crosslinking density [88], pH, solvent [89], the presence of salt [90] or
surfactants [13], and the amount of ionic [91], hydrophobic [92] or hydrophilic groups [93][94] in
the polymer. For instance, Mikheeva et al. studied the effect of NaCl and sodium dodecyl
sulfate (SDS) on the transition temperature of PVCL microgels. They showed that the
presence of NaCl results in increased hydrophobic interactions, hence both transition
temperatures are shifted to lower temperatures, i.e. room temperature. Amphiphilic SDS on
the other hand promotes the formation of hydrophobic domains, therefore pronouncing the
difference between the two separate transition temperatures while at the same time shifting
them to higher temperatures. At an SDS concentration of 5.2 mM, the transition temperatures
shift to 67.6 °C and 78.8 °C, respectively. A. Pich et al. showed that the amount of ethanol in
an aqueous PVCL microgel solution greatly influences the transition temperature [89].
Depending on the amount of ethanol, the VPTT decreases, increases or vanishes completely.
All these examples, however, are not suitable if it comes to applications under physiological
conditions. Variations in solvent or ionic strength, for instance, are hardly or not at all feasible
within the body. Also the use of ionic surfactants is not without difficulty since they cause the
denaturation of proteins [95]. Therefore, the most practicable approach is the incorporation of
hydrophobic or hydrophilic comonomers. Hydrophilic units change the VPTT to higher
temperatures, while hydrophobic comonomers shift it to lower temperatures. A. Pich et al., for
instance, prepared VCL-co-AAEM-co-Vim microgels and showed that a higher amount of
Vim leads to an increase in the VPTT. At pH = 4, the protonation of the Vim-segments leads
16
to a drastic shift of the VPTT to ~ 50 °C compared to ~ 28 °C for pure PVCL-co-AAEM
microgels [93].
While properties of PNIPAm microgels such as a VPTT close the human body temperature
are important for applications in the biomedical field, the problem of colloidal stability under
physiological conditions must be solved. H.-Y. Tsai report the aggregation of PNIPAm
microgels in buffer solution when the temperature rises over the transition temperature [96].
Steric stabilization or the incorporation of acid or basic moieties improves the stability.
Another drawback of PNIPAm microgels is the toxicity of the monomer NIPAm and its
carcinogenetic properties [97]. PNIPAm microgels therefore are neither biocompatible nor
biodegradable. VCL, on the other hand, is more suitable for bio-applications since it is stable
against hydrolysis [40]. Its biocompatibility [47][51], non-toxicity [98][99], solubility in both water
and organic solvents, inexpensiveness [46], and high absorption ability [100] makes its use
attractive for many medical and biochemical applications. S. Peng and C. Wu showed that
microgels of PVCL and sodium acrylate are able to build complexes with Hg2+, Cu2+, Ca2+,
and Na+ making it suitable for water treatment [101]. Y. Wang et al. prepared nanogels based
on PVCL with ketal linkages that are acid cleavable. They studied the drug release behavior
using the drug doxorubicin (DOX) and demonstrated the fast and efficient release at low pH.
17
3.4 Literature
[1] R. Pelton, T. Hoare, in Microgel Suspens. Fundam. Appl., 2011.
[2] R. Pelton, Adv. Colloid Interface Sci. 2000, 85, 1–33.
[3] Y. Zhang, Y. Guan, S. Zhou, Biomacromolecules 2007, 8, 3842–3847.
[4] T. Hirotsu, Gels Handbook, 2001.
[5] T. Still, K. Chen, A. M. Alsayed, K. B. Aptowicz, a. G. Yodh, J. Colloid Interface Sci.
2013, 405, 96–102.
[6] S. Nayak, L. A. Lyon, Angew. Chemie 2005, 117, 7862–7886.
[7] C. J. Coester, K. Langer, H. van Briesen, J. Kreuter, J. Microencapsul. 2000, 17, 187–
193.
[8] S. Azarmi, Y. Huang, H. Chen, S. McQuarrie, D. Abrams, W. Roa, W. H. Finlay, G. G.
Miller, R. Löbenberg, J. Pharm. Pharm. Sci. 2006, 9, 124–132.
[9] A. Laukkanen, S. K. Wiedmer, S. Varjo, M.-L. Riekkola, H. Tenhu, Colloid Polym.
Sci. 2002, 280, 65–70.
[10] B. S. Ho, B. H. Tan, J. P. K. Tan, K. C. Tam, Langmuir 2008, 24, 7698–703.
[11] H. Ahmad, D. Dupin, S. P. Armes, A. L. Lewis, Langmuir 2009, 25, 11442–9.
[12] A. Laukkanen, S. Hietala, S. L. Maunu, H. Tenhu, Macromolecules 2000, 33, 8703–
8708.
[13] E. E. Makhaeva, H. Tenhu, A. R. Khokhlov, Macromolecules 1998, 31, 6112–6118.
[14] H. Feil, Y. H. Bae, J. Feijen, S. W. Kim, Macromolecules 1993, 26, 2496–2500.
[15] N. Welsch, A. L. Becker, J. Dzubiella, M. Ballauff, Soft Matter 2012, 8, 1428.
[16] M. Muratalin, P. F. Luckham, J. Colloid Interface Sci. 2013, 396, 1–8.
[17] J. M. D. Heijl, F. E. Du Prez, Polymer (Guildf). 2004, 45, 6771–6778.
[18] L. Y. Chu, J. W. Kim, R. K. Shah, D. Weitz, Adv. Funct. Mater. 2007, 17, 3499–3504.
[19] H. Y. Tsai, K. Vats, M. Z. Yates, D. S. W. Benoit, Langmuir 2013, 29, 12183–12193.
[20] A. Bin Imran, T. Seki, Y. Takeoka, Polym. J. 2010, 42, 839–851.
[21] Y. Qiu, K. Park, Adv. Drug Deliv. Rev. 2001, 53, 321–39.
[22] H. Ichikawa, Y. Fukumori, J. Control. Release 2000, 63, 107–119.
[23] W. Leobandung, H. Ichikawa, Y. Fukumori, N. a. Peppas, J. Appl. Polym. Sci. 2002,
87, 1678–1684.
[24] W. H. Blackburn, L. A. Lyon, Colloid Polym. Sci. 2008, 286, 563–569.
[25] D. E. Bergbreiter, B. L. Case, Y. Liu, J. W. Caraway, Macromolecules 1998, 9297,
6053–6062.
[26] A. Pich, W. Richtering, Chemical Design of Responsive Microgels, Springer Berlin
Heidelberg, Berlin, Heidelberg, 2011.
[27] K. Y. Lee, D. J. Mooney, Chem. Rev. 2001, 101, 1869–1880.
[28] H. Kawaguchi, K. Fujimoto, Bioseparation 1999, 7, 253–258.
[29] C. D. Sorrell, M. C. D. Carter, M. J. Serpe, Adv. Funct. Mater. 2011, 21, 425–433.
18
[30] D. Paloli, P. S. Mohanty, J. J. Crassous, E. Zaccarelli, P. Schurtenberger, Soft Matter
2013, 9, 3000.
[31] A. S. J. Iyer, L. A. Lyon, Angew. Chemie - Int. Ed. 2009, 48, 4562–4566.
[32] P. S. Mohanty, W. Richtering, J. Phys. Chem. B 2008, 112, 14692–14697.
[33] Z. Zhou, J. V Hollingsworth, S. Hong, G. Wei, Y. Shi, X. Lu, H. Cheng, C. C. Han,
Soft Matter 2014, 10, 6286–6293.
[34] M. Destribats, F. Pinaud, E. Sellier, U. De Bordeaux, A. P. Berland, P. Cedex, C. De
Recherche, P. Pascal, A. A. Schweitzer, Langmuir 2013, 29, 12367–12374.
[35] M. Destribats, V. Lapeyre, M. Wolfs, E. Sellier, F. Leal-Calderon, V. Ravaine, V.
Schmitt, Soft Matter 2011, 7, 7689.
[36] Lookchem, “http://www.lookchem.com/N-Vinylcaprolactam/,” 29.06.2015
[37] Sigma-Aldrich,
“http://www.sigmaaldrich.com/catalog/product/aldrich/415324?lang=de®ion=DE,”
29.06.2015
[38] S. Sun, P. Wu, J. Phys. Chem. B 2011, 115, 11609–18.
[39] L. Hou, P. Wu, RSC Adv. 2014, 4, 39231.
[40] J. Ramos, A. Imaz, J. Forcada, Polym. Chem. 2012, 3, 852.
[41] A. Imaz, J. I. Miranda, J. Ramos, J. Forcada, Eur. Polym. J. 2008, 44, 4002–4011.
[42] O. F. Solomon, D. S. Vasilescu, V. Tǎrǎrescu, J. Appl. Polym. Sci. 1969, 13, 1–7.
[43] S. F. Medeiros, J. C. S. Barboza, R. Giudici, A. M. Santos, J. Macromol. Sci. Part A
2013, 50, 763–773.
[44] K. Nakabayashi, H. Mori, Eur. Polym. J. 2013, 49, 2808–2838.
[45] M. Hurtgen, J. Liu, A. Debuigne, C. Jerome, C. Detrembleur, J. Polym. Sci. Part A
Polym. Chem. 2012, 50, 400–408.
[46] M. Beija, J.-D. Marty, M. Destarac, Chem. Commun. (Camb). 2011, 47, 2826–8.
[47] S. C. Cheng, W. Feng, I. I. Pashikin, L. H. Yuan, H. C. Deng, Y. Zhou, Radiat. Phys.
Chem. 2002, 63, 517–519.
[48] T. H. Ho, M. Levere, J.-C. Soutif, V. Montembault, S. Pascual, L. Fontaine, Polym.
Chem. 2011, 2, 1258.
[49] X. Jiang, S. Zhai, X. Jiang, G. Lu, X. Huang, Polymer (Guildf). 2014, 55, 3703–3712.
[50] B. Y. Zhang, W. D. He, W. T. Li, L. Y. Li, K. R. Zhang, H. Zhang, Polymer (Guildf).
2010, 51, 3039–3046.
[51] V. Boyko, A. Pich, Y. Lu, S. Richter, K.-F. Arndt, H.-J. P. Adler, Polymer (Guildf).
2003, 44, 7821–7827.
[52] A. Kermagoret, C.-A. Fustin, M. Bourguignon, C. Detrembleur, C. Jérôme, A.
Debuigne, Polym. Chem. 2013, 4, 2575.
[53] M. A. Ward, T. K. Georgiou, Polymers (Basel). 2011, 3, 1215–1242.
[54] J. Yin, J. Hu, G. Zhang, S. Liu, Langmuir 2014, 30, 2551–2558.
[55] F. Afroze, E. Nies, H. Berghmans, J. Memb. Sci. 2000, 554, 55–68.
19
[56] F. Meeussen, E. Nies, H. Berghmans, S. Verbrugghe, E. Goethals, F. Du Prez, Polymer
(Guildf). 2000, 41, 8597–8602.
[57] L. D. Taylor, L. D. Cerankowski, J. Polym. Sci. 1975, 13, 2551–2570.
[58] V. Aseyev, S. Hietala, A. Laukkanen, M. Nuopponen, O. Confortini, F. E. Du Prez, H.
Tenhu, Polymer (Guildf). 2005, 46, 7118–7131.
[59] V. Aseyev, H. Tenhu, F. M. Winnik, Adv Polym Sci 2011, 242, 29–89.
[60] I. Bischotberger, D. C. E. Calzolari, P. De Los Rios, I. Jelezarov, V. Trappe, Sci. Rep.
2014, 4, 4377.
[61] BASF, http://www.micronal.de/portal/streamer?fid=436077, 29.06.2015
[62] BASF, http://www.dispersions-pigments.basf.com/portal/ba, 29.06.2015
[63] V. I. Lozinsky, I. a Simenel, V. K. Kulakova, E. a Kurskaya, T. a Babushkina, T. P.
Klimova, T. V Burova, A. S. Dubovik, I. Y. Galaev, B. Mattiasson, et al.,
Macromolecules 2003, 36, 7308–7323.
[64] E. D. Goddard, J. V. Gruber, Principles of Polymer Science and Technology in
Cosmetics and Personal Care, 2000.
[65] R. H. Pelton, P. Chibante, Colloids and Surfaces 1986, 20, 247–256.
[66] J. Ferguson, V. Sundar Rajan, Eur. Polym. J. 1979, 15, 627–630.
[67] D. Klinger, Light-Sensitive Polymeric Nanoparticles Based on Photo-Cleavable
Chromophores, 2013.
[68] X. Hu, Z. Tong, L. A. Lyon, Langmuir 2011, 27, 4142–4148.
[69] I. Varga, T. Gilányi, R. Mészáros, G. Filipcsei, M. Zrínyi, J. Phys. Chem. B 2001, 105,
9071–9076.
[70] J. Gao, B. J. Frisken, Langmuir 2003, 19, 5212–5216.
[71] M. H. Kwok, Z. Li, T. Ngai, Langmuir 2013, 29, 9581–9591.
[72] S. Schachschal, A. Balaceanu, C. Melian, D. E. Demco, T. Eckert, W. Richtering, A.
Pich, Macromolecules 2010, 43, 4331–4339.
[73] A. Balaceanu, Y. Verkh, D. E. Demco, M. Möller, A. Pich, Macromolecules 2013, 46,
4882–4891.
[74] C. D. Jones, L. A. Lyon, Macromolecules 2003, 36, 1988–1993.
[75] K. Y. Q. X D Feng, X Q Guo, Die Makromol. Chemie 1988, 189, 77–83.
[76] J. W. Goodwin, J. Hearnb, C. C. Ho, R. H. Ottewill, Br. Polym. J. 1973, 5, 347–362.
[77] J. W. Goodwin, R. H. Ottewill, R. Pelton, G. Vianello, D. E. Yates, Br. Polym. J. 1978,
10, 173–180.
[78] D. Dupin, S. Fujii, S. P. Armes, P. Reeve, S. M. Baxter, Langmuir 2006, 22, 3381–
3387.
[79] A. Lee, H. Y. Tsai, M. Z. Yates, Langmuir 2010, 26, 18055–18060.
[80] H. Shimizu, R. Wada, M. Okabe, Polym. J. 2009, 41, 771–777.
[81] Y. Hirokawa, T. Tanaka, J. Chem. Phys. 1984, 81, 6379.
[82] B. Özkahraman, I. Acar, S. Emik, Polym. Bull. 2010, 66, 551–570.
20
[83] L. M. Mikheeva, N. V Grinberg, A. Y. Mashkevich, V. Y. Grinberg, L. Thi, M. Thanh,
E. E. Makhaeva, A. R. Khokhlov, Macromolecules 1997, 9297, 2693–2699.
[84] Y. Tang, Y. Ding, G. Zhang, J. Phys. Chem. B 2008, 112, 8447–8451.
[85] K. Von Nessen, M. Karg, T. Hellweg, Polym. (United Kingdom) 2013, 54, 5499–5510.
[86] S. Peng, C. Wu, J. Phys. Chem. B 2001, 105, 2331–2335.
[87] V. Boyko, S. Richter, A. Pich, K.-F. Arndt, Colloid Polym. Sci. 2003, 282, 127–132.
[88] W. Su, K. Zhao, J. Wei, T. Ngai, Soft Matter 2014, 10, 8711–8723.
[89] A. Pich, Y. Lu, V. Boyko, S. Richter, K. F. Arndt, H. J. P. Adler, Polymer (Guildf).
2004, 45, 1079–1087.
[90] Z. H. Farooqi, H. U. Khan, S. M. Shah, M. Siddiq, Arab. J. Chem. 2013, DOI
10.1016/j.arabjc.2013.07.031.
[91] S. Hirotsu, Y. Hirokawa, T. Tanaka, J. Chem. Phys. 1987, 87, 1392.
[92] J. D. Debord, L. A. Lyon, Langmuir 2003, 19, 7662–7664.
[93] A. Pich, A. Tessier, V. Boyko, Y. Lu, H. P. Adler, Macromolecules 2006, 7701–7707.
[94] T. Hoare, R. Pelton, Langmuir 2004, 20, 2123–33.
[95] A. Lee, S. K. Y. Tang, C. R. MacE, G. M. Whitesides, Langmuir 2011, 27, 11560–
11574.
[96] H.-Y. Tsai, A. Lee, W. Peng, M. Z. Yates, Colloids Surfaces B Biointerfaces 2014,
114, 104–110.
[97] J. B. Thorne, G. J. Vine, M. J. Snowden, Colloid Polym. Sci. 2011, 289, 625–646.
[98] H. Vihola, A. Laukkanen, L. Valtola, H. Tenhu, J. Hirvonen, Biomaterials 2005, 26,
3055–3064.
[99] L. K. Kostanski, R. Huang, R. Ghosh, C. D. M. Filipe, J. Biomater. Sci. Polym. Ed.
2008, 19, 275–290.
[100] H. Vihola, A. Laukkanen, J. Hirvonen, H. Tenhu, Eur. J. Pharm. Sci. 2002, 16, 69–74.
[101] S. Peng, C. Wu, Polymer (Guildf). 2001, 42, 6871–6876.
21
4 Experimental Part
4.1. Chemicals
N-vinylcaprolactam (VCL, 98 %, destilled before use), [2-(methacryloyloxy)ethyl]-dimethyl-
(3-sulfopropyl)-ammonium hydroxide (sulfobetaine, SB, 97 %), N,N‘-
methylene(bis)acrylamide (BIS, 99 %), itaconic acid (IA, ≥ 99 %), 1-Vinylimidazole (Vim, ≥
99 %), N-isoproylacrylamide (NIPAm, 97 %), Span®80 (sorbitan monooleate), Tween®80
(polyethylene glycol sorbitan monooleate), 2,2‘-azobis(2-methylpropionamidine)
dihydrochloride (AMPA, 97 %), ammonium carbonate (≥ 30 % NH3 basis), phosphate
buffered saline (PBS, tablets), (3-aminopropyl)trimethoxysilane (APTMS, 97 %), 2-
(methacryloyloxy)ethyl acetoacetate (AAEM, 95 %), 2-(N,N’-dimethylamino)ethyl
methacrylate (DMAEM, 98 %), cytochrome c (from equine heart, ≥ 95 %), and acetone (for
HPLC, ≥ 99.8 %) were purchased from Sigma Aldrich and were used without further
purification. 2-(methacryloyloxy)ethyl 2-(trimethylammonio)ethyl phosphate (phospho-
betaine, PB, ≥ 98 %) was purchased from Santa Cruz Biotechnology. β-propiolactone (97 %)
was obtained from Alfa Aesar. Albumin from borine serum tetramethylrhodamine conjugate
(BSA) was purchased from Life technologiesTM. Toluene (min. 99.5 %) was obtained from
VWR Chemicals. Calcium chloride (95 %, granulated) was purchased from Riedel-de-Haёr.
Potassium chloride (pro analysi) and sodium chloride (for analysis) were obtained from
Merck Millipore. Water used in the experiments was purified using a Millipore water
purification system with a minimum resistivity of 18 MΩ*cm. For light scattering
experiments, sterile water from Roth was used.
4.2 Analytical Instrumentation
Dynamic Light Scattering (DLS) The hydrodynamic radius RH of the microgel particles was
measured using an ALV/CGS-3 goniometer with an ALV/LSE 7004 Tau Digital Correlator
and a JDS Uniphase laser operating at 633 nm. The measurements were taken at 25 °C at an
angle of 90 °C after equilibrating the samples for at least 10 min. Temperature trends were
measured in a temperature range of 15 °C to 50 °C in 3 °C-steps with at least five
measurements per temperature step. Before all measurements, the samples were filtered with
a 1.2 µm PTFE filter. Water was filtered with 5 µm, 1.2 µm, and 0.45 µm filters prior to use.
Another device for dynamic light scattering is the Zetasizer Nano ZS (Malvern Instruments,
UK) equipped with a 532 nm laser. Measurements were performed at a scattering angle of
22
173 °. Measurement conditions (equilibration time, measurement time, number of
measurements, temperature range) are identical to the measurements performed at the ALV.
Static Light Scattering (SLS) The molecular weight of the microgel particles was measured
with static light scattering using a Fica goniometer (SLS Systemtechnik, Germany) with a
laser of 543 nm. The samples were measured in an angle range of 25 ° to 145 ° with 5 °
intervals. For each sample, five dilutions were prepared with concentration from 0.1 mg/mL
to 2 mg/mL. Prior to SLS measurements, the change of the refractive index (dn/dc) was
determined using a refractometer (SLS Systemtechnik, Germany). Both measurements were
performed at 20 °C.
Electrophoretic Mobility The electrophoretic mobility was measured at a Zetasizer NanoZS
(Malvern Instruments, UK). Each sample contained 1 mM NaCl and was measured at 25 °C
after an equilibration of at least 10 min. The size of microgels at different pH were measured
using 0,1 M HCl and NaOH to adjust the pH. Measurements were taken between pH = 3 and
10 in 0.5 steps at 25 °C. Before all measurements, the samples were filtered with a 1.2 µm
PTFE filter.
Nuclear Magnetic Resonance (NMR) To determine the comonomer content in microgels,
1H-NMR spectra were taken with a Bruker DPC at a frequency of 300 MHz. 5 wt-% of
freeze-dried microgels were dissolved in D2O and measured at room temperature.
Fourier-transformed Infrared Spectroscopy (FTIR) FTIR measurements were performed on
freeze-dried microgel samples at a Bruker, Alpha-P apparatus at room temperature. The
samples were mixed with KBr powder and then pressed to form a transparent KBr pellet.
UV-Vis Spectroscopy (UV-Vis) UV-Vis spectra were taken at a UV-visible
spectrophotometer using a CARY 100 Bio (Agilent Technologies, USA). A given
concentration of microgel solution in water was measured at room temperature in the range of
900 – 200 nm using 1 cm path length quartz cuvettes.
For the determination of the UCST and LCST of a solution, the temperature was controlled
using a heating circulator and a cooler. The temperature was increased from 5 °C to 70 °C
with a heating rate of 0.4 K/min. The UCST or LCST was defined as the temperature at the
maximal slope for the absorbance against the temperature.
Field Emission Scanning Electron Microscopy (FE-SEM) FE-SEM images were taken at a
S4800 (Hitachi) equipped with a field emission cathode. The electron high tension was set to
1-2 keV. One drop of the microgel solution was put on aluminum foil and dried overnight to
prepare samples for FE-SEM investigation. To improve the electric conductivity, a gold layer
was sputtered for 100 s onto the samples prior to measurements.
23
Transmission Electron Microscopy (TEM) TEM images were taken using a Zeiss Libra 100
(Zeiss, Germany). For the sample preparation one drop of the microgel solution was deposited
at room temperature on a TEM copper grid (Formvar/ carbon film, 200 mesh, Plano GmbH,
Germany) the day before and dried overnight. The TEM was operated at an acceleration
voltage of 120 kV. The images were recorded using a CCD camera system (Ultra Scan 1000,
Gatan, Germany).
Atomic Force Microscopy (AFM) Atomic Force Microscopy images were taken at an
Asylum Research MFP-3D (Santa Barbara, CA) in AC mode. Cantilevers of silicon nitride
were purchased from NanoWorld (Neuchatel, Switzerland) with a force constant of 42 N/m.
Images (10.0 x 10.0 µm²) were acquired using the tapping mode. Force mapping was done in
contact mode. The root-mean-square (rms) roughness gives the height deviation from a mean
line and was determined to quantitatively describe the roughness of the surface:
𝑹𝒒 = [(𝟏
𝑳) ∫ 𝒁(𝒙𝟐)𝒅𝒙
𝑳
𝟎]𝟏𝟐⁄
(1)
with L = length; Z(x) = profile height function.
Thermogravimetric Analysis (TGA) Thermogravimetric analysis (TGA) measurements were
performed using a Netzsch TG 209C unit (Netzsch Gerätebau GmbH, Germany) operating
under nitrogen atmosphere with a flow rate of 10 mL/min. 20–30 mg of samples were placed
in standard Netzsch alumina 85 μL crucibles and heated at 20 K/min from 20 °C to 850 °C.
Differential Scanning Calorimetry (DSC) Measurements were performed at a Netzsch DSC
204 (Netzsch Gerätebau GmbH, Germany). Samples were solved in 1 – 2 drops of D2O in
aluminum crucibles and weighed. The crucibles were closed with an aluminum lid. As
reference, pure D2O was used. The scanning was done from 5 – 70 °C at a heating rate of
2 K/min under nitrogen flow. An isotherm was recorded for 2 min before the sample was
cooled down again to 5 °C at a heating rate of 2 K/min. The cycle was repeated once.
Isothermal Titration Calorimetry (ITC) The measurements were performed at a TAM III
from TA Instruments company, Germany. A given amount of microgel and protein were
solved in PBS buffer, respectively. A solution of pure PBS buffer was used as a reference. All
samples were degassed for 10 min at room temperature and 635 mmHg to remove CO2 from
the solution. 500 µL of microgel solution were filled in the ampoule of stainless steel. The
stirrer in the reaction ampoule was set to 200 rpm. 640 µL (i.e. 500 µL + half of volume of
titrated volume) of the reference solution were filled in the reference ampoule. The ampoules
are lowered stepwise into the calorimeter chamber. After being lowered completely, the
system was allowed to equilibrate for at least 60 min. 280 µL of aqueous CaCl2 (c = 1
24
mg/mL) solution was filled into a glass syringe. It is important that no gas bubbles are inside
the syringe. Measurements were performed at 25 °C and 37 °C. The temperature could be set
with an accuracy of ± 0.0001 °C. 28 injections of 10 µL each were done, each injection lasted
10 s. Between measurements, the signal was allowed to return to the baseline in 15 min. Each
measurement was repeated with pure buffer solution in the syringe to determine dilution heat.
Flow Field-Flow Fractionation (F-FFF) Measurements were done using an AF 2000 MT
separation system (Postnova Analytics, Germany). 20 mL of aqueous microgel solutions with
a concentration of 1 mg/mL were filled into the channel that had a thickness of 294 µm and a
length of 30 cm. The eluent was filtered deionized water. Particles were collected using a
regenerated cellulose membrane (30,000 MWCO). The eluting particles were detected online
by static light scattering.
Colloidal stability measurements The stability of the microgel particles was studied with a
LUMiFuge 1112-33 (L.U.M. GmbH, Germany). The centrifuge measures sedimentation
velocities of dispersions under centrifugal force. Microgel samples were measured in a 2 mm
LUM, rectangular Polycarbonate cell. Centrifugation was done at acceleration velocity
2000 rpm (corresponds to 800 g) at intervals of 10 s at T = 25 °C and 50 °C. The slope of
sedimentation curves was used to calculate the sedimentation velocity and to gain information
about colloidal stability of the samples.
IGASorp Moisture sorption-desorption isotherms were measured using an IGAsorp moisture
sorption analyzer (Hiden Isochema) at 37 °C. The sample was put in a ultra-sensitive balance
and the relative humidity (RH) was increased from 2 % RH to 94 % RH in steps of 10 %.
Plasma activation and spin-coating The substrates were cleaned prior to use with water and
ethanol and dried with nitrogen to remove impurities. Then they were put in an air plasma
oven (Plasma Activate Flecto 10 USB) and activated at 0.2 mbar for 60 s at room
temperature. The substrates were coated with 30 µL of microgel sample via spin-coating with
a rotational speed of 1000 rpm/s for 60 s at room temperature.
Contact angle measurements Measurements were conducted at a Contact Angle Measuring
System G2-Mk4 from Krüss GmbH, Germany. A water droplet of Millipore water of 5 µL
was dropped on a coated Si-wafer. 10 images of the droplet on the surface were immediately
taken and the contact angle was measured with the software.
Protein adsorption tests To determine protein repellent properties of a modified surface, the
wafers were incubated with 700 µL of a 5 µG/mL solution of BSA, BODIFY FL conjugate
for 20 min. The protein solution was removed and substituted with 700 µL of PBS buffer for
10 min. This step was repeated twice. At last, 700 µL of water were added for 10 min and
25
removed again. The measurements were conducted at a fluorescence microscope (Carl Zeiss
Axioplan 2) with 10-fold magnification. A halogen lamp XBO 75 was used and the exposure
time was set to 20 s. As a reference, uncoated wafers were used.
Circular dichroism (CD) spectroscopy CD spectra were recorded on a JASCO J-1100 CD
spectropolarimeter using a quartz cuvette with a 0.1 cm path-length at room temperature.
Protein solutions of cytochrome c with a concentration of 0.0005 g/mL were used in 10 mM
PBS buffer with pH = 3 and pH = 8, respectively. The reported CD profiles were recorded
from 190 to 300 nm to measure secondary protein structure and are an average of three
successive scans obtained at a 10 nm min−1 scan rate corrected by a baseline. The measured
spectra are converted to mean residue ellipticity (MRE) as follows:
[𝜽] = 𝜽
𝟏𝟎∙𝒏∙𝒄∙𝒍 (2)
with [θ] = mean residue ellipticity; θ = observed ellipticity (in mdeg); n = amino acid residue
(104 for cytochrome c [1]); c = protein concentration (in mol); l = pathlength of cuvette.
The content of α-helix, beta sheet, turn, and random coil conformations are determined via the
Structural Estimation Application software using Yang’s reference.
4.3 Microgel synthesis
The syntheses of the microgel particles and their purification are described in the respective
chapters.
26
5 Zwitterionic Microgels
5.1. Introduction
Zwitterionic microgels contain the same amount of cationic and anionic groups and are thus
electrically neutral. In contrast to ampholyte microgels, both charges are located in the same
monomer unit (see Figure 5.1). This ensures that always an equal amount of charges is present
in the particle. In general, the cation does not have an associated hydrogen atom.
A prominent group of zwitterions are betaines. Betaines consist of a quaternary ammonium or
pyridinium group [1] and a negative pendant group. In most cases, the latter is either a sulfo-,
carboxy- or phosphogroup. Hence, the correspondent betaines are called sulfobetaine (SB),
carboxybetaine (CB), and phosphobetaine (PB) (see Figure 5.1).
The main difference between zwitterionic and ampholyte microgels becomes apparent by
looking at their behavior in aqueous solution. In water, ampholyte microgels are swollen
through the electrostatic repulsion of the charged groups. The addition of salt or the change of
pH leads to a collapse of the ampholyte microgel structure because the charges are screened.
This phenomenon is called the polyelectrolyte effect. Zwitterionic microgels, on the other
hand, behave differently. Their structure expands after the addition of salt or the change of pH
which is called the anti-polyelectrolyte effect [2].
SB is similar to taurine, an amino acid found in the human body [3]. It is an essential part of
the muscles and the nervous system [4]. Contrary to other betaines, SB is electrically neutral
over a wide range of pH due to its high acid dissociation constant. CB is similar to glycine
betaine which is an essential part of the osmotic regulation in living organisms [3]. PB is
similar to phosphorylcholine [3] which is an essential part of biological membranes [1]. Thus
much effort has been done to study the behavior of phospholipid analogues. However, all
phosphobetaines described in the literature are not analogues of the described sulfo- and
carboxybetaines, i.e. while SB and CB have the anionic group at the tail of each monomer
unit, PB has the positively charged ammonia group at the end. Even in publications
comparing all three betaines, e.g. by N.Y. Kostina [5], no remark concerning the different
structures is given.
Betaines are well known for their ultra low-fouling properties [5][6][7]. These properties will be
discussed in more detail in chapter 11. In addition to their possible functionalization and their
anti-polyelectrolyte effect, polymers containing betaines are highly interesting for biomedical
applications such as drug delivery [6] and tissue engineering [8].
27
Figure 5.1 Betaine structures: (a) sulfobetaine; (b) carboxybetaine; (c) phosphobetaine.
General structure of (a) zwitterions; (b) ampholytes.
Y. Chang et al. used SB and PNIPAm to obtain zwitterionic copolymers [9]. They stress the
usefulness of the particles due to their biocompatibility and their temperature-sensitivity. The
copolymer exhibits both UCST and LCST depending on the copolymer composition, solution
concentration, solution polarity, and ionic strength. For instance, a copolymer composition of
29.0 mol-% polySB and 71.0 mol-% PNIPAm exhibits an UCST of 15 °C and an LCST of
41 °C in water at a polymer concentration of 5 wt-%. Furthermore, the authors could show
that the same copolymer exhibits extremely low protein adsorption on surfaces. Y. Su et al.
synthesized poly(acrylonitrile)-co-SB copolymers for the preparation of zwitterionic
membranes [10]. The on-off behavior for transport of proteins such as lysozyme and BSA can
be controlled by the electrolyte concentration. If no electrolytes are present, attractive
interactions between the proteins and the sulfobetaine exists, while those interactions are
screened at higher electrolyte concentrations.
The synthesis of zwitterionic hydrogels has been reported by many authors [11][12][13]. In order
to improve mechanical stability and incorporate further desired properties these hydrogels are
often copolymerized. One drawback is the decreased non-fouling properties when other
monomers are inserted. N.Y. Kostina et al. present brushes containing SB, CB or PB and 2-
hydroxyethyl methacrylate (HEMA) [5]. They showed that only CB-co-HEMA copolymers
exhibited a sufficient fouling resistance and thus used them to obtain hydrogels. Hydrogels
with a CB content of 10 mol-% retained the good mechanical stability characteristic for pure
PHEMA hydrogels while at the same time possessing a four-fold amount of water promoting
protein-repellent properties.
a b c d e
28
M. Das et al. were the first to synthesize zwitterionic microgels based on PNIPAm using
sulfobetaine as the zwitterionic comonomer [2]. They showed that only a limited amount of SB
can be incorporated into the microgel network using free radical precipitation polymerization.
With varying pH, no change in particle size could be observed due to the permanent negative
charge of the sulfo group. Interestingly, the authors could detect no anti-polyelectrolyte
behavior of the particles. Furthermore, the authors used the zwitterionic microgels for in-situ
synthesis of gold and silver nanoparticles.
The great potential of zwitterionic structures to synthesize, trap, encapsulate and/or release
nanoparticles is also shown in other literature [14]. L. Zhang et al. used polyCB nanogels of
~ 110 nm in size to encapsulate monodisperse Fe3O4 nanoparticles that can be used for
enhancing Magnetic Resonance Imaging (MRI) performances [6]. They showed that only 3 %
of the nanoparticles are released in PBS, but much faster (80 %) in the presence of the
reducing agent dithiothreitol representing an intracellular environment. The release of
nanoparticles occurred through the cleavage of the disulfide crosslinker. Furthermore, the
nanogels are able to degrade into small polyCB polymers that can be removed from the body.
29
5.2. Synthesis Procedure
Synthesis of Carboxybetaine Carboxybetaine was synthesized as described in my diploma
thesis following the synthesis procedure by Z. Zhang [15].
50 mL of dried acetone were ice-cooled in a flask. DMAEM (1.54 g, 9.9 mmol) and β-
propiolactone (0.99 g, 13.7 mmol) were slowly added with a syringe pump and the mixture
was allowed to stir for 5 h at 15 °C. The product was obtained as a white precipitate. The
solvent was removed and the product was washed three times with anhydrous ether and
acetone.
1H-NMR (300 MHz): δ = 6.04 (s, 1H, H-2), 5.65 (s, 1H, H-1), 4.53 (t, 2H, H-4), 3.69 (t, 2H,
H-7), 3.55 (t, 2H, H-5), 3.07 (s, 6H, H-6), 2.61 (t, 2H, H-8), 1.82 (s, 3H, H-3) ppm.
Figure 5.2 1H-NMR spectrum of carboxybetaine.
Synthesis of Phosphobetaine Commercially available phosphobetaine has its two ionizable
groups in reverse order compared to the other betaines used in this work, i.e. the quaternary
ammonium group is terminal. Therefore, phosphobetaine with a terminal phosphogroup was
synthesized and friendly provided by Rüdiger Bormann (IOC, RWTH Aachen).
Figure 5.3 Structure of phosphobetaine. (a) As commercially available. (b) As synthesized by
R. Bormann; numbers refer to the NMR spectrum in Figure 5.4.
30
Figure 5.4 1H-NMR spectrum of phosphobetaine. Numbers above the spectrum refer to the
peak position, numbers below the spectrum refer to the integral.
1H-NMR (300 MHz): δ = 6.03 (s, 1H, H-1), 5.64 (s, 1H, H-2), 4.51 (s, 2H, H-4), 3.91 (m, 2H,
H-8), 3.76 (m, 2H, H-7), 3.68 (m, 2H, H-5), 3.12 (s, 6H, H-6), 2.08 (t, 3H, H-3) ppm.
31
5.3. Precipitation Polymerization
5.3.1. Introduction
Precipitation polymerization is the main synthesis procedure for microgel synthesis.
Zwitterionic microgels with sulfo-, carboxy-, and phosphobetaines as comonomers were
already synthesized and characterized in my diploma thesis. In addition to crosslinked
zwitterionic microgels, samples without crosslinker BIS were analyzed regarding their
particle size.
In this work, however, non-covalently crosslinked (hence called non-crosslinked) samples
with varying betaine content were further studied with regard to their degradability upon salt
addition and the particle formation process in comparison to conventionally BIS-crosslinked
microgels.
It is postulated that zwitterionic compounds form internal ionic bridges of two different types
(see Figure 5.5)[16]: (a) The molecular structure is rigid and no internal bridges between the
ionic groups can be formed. (b) Intermolecular bridges between two different zwitterionic
groups are formed through antiparallel arrangement. (c) Intramolecular bridges between the
charged groups of one zwitterionic unit are formed; this type is dependent on the length of the
methylene units between the charged groups.
Figure 5.5 Types of ion pairing within zwitterionic structures: (a) rigid structure; (b)
intermolecular bridges; (c) intramolecular bridges.
M. Szafran described the factors that lead to bridges of either type (b) or (c): (1) flexibility of
the chain; (2) bulkiness and hydration of the charged groups; and (3) electrical properties of
both the charged group and the solvent [17]. S. Chen et al. suggested that the charged groups
arrange antiparallel as to minimize net dipole moments [18].
The formation of ionic bridges of type (b) lead to the suggestion that it is possible to
synthesize microgels without the use of conventional crosslinker such as BIS because
intermolecular bridges are expected to form sufficiently strong ion pairing. These bridges,
however, should easily be dissolved by the addition of saline solution.
In literature, degradation of microgels mostly occurs by the use of a degradable crosslinker
that degrades by the use of UV light [19][20][21], pH [22][23], and enzymes [24][25][26]. V. Bulmus for
a b c
32
instance synthesized microgels based on poly(hydroxyethyl methacrylate) (pHEMA) and used
several diacrylate- and dimethylacrylate-functionalized crosslinkers to obtain acid-cleavable
microgels [27]. They studied the hydrolysis time at two different pH values, neutral and acidic,
and showed that particles are stable at neutral pH, while they degrade in acidic medium in less
than 60 min. The authors explained this by the cleavage of the acetal crosslinks resulting in
free linear crosslinker chains. In this procedure, the precursor polymers are recovered and the
degradation of the microgel does not lead to toxic or harmful products which is an essential
criterion for the use as bio- and cytocompatible materials. A. M. Hawkins et al. specifically
focused on the toxicity of the degradation products of poly (β-amino ester) hydrogels and
analyzed the cellular response of mesenchymal cells [28]. M. Patenaude and T. Hoare show in
their work the possible application of NIPAm hydrogels that can be injected over a syringe to
act as a drug delivery system [29]. They point out two challenges: The microgels have to be
robust enough to withstand the shear stress in a syringe. On the other hand, it must be ensured
that the hydrogel degrades without the generation of toxic products and without accumulating
in the body. They could show that their hydrogels degrade in acid and form the formerly used
precursor copolymers. No considerable toxicity or only slight toxicity at high concentrations
was observed in cell tests.
There are only few examples in literature concerning the degradation of zwitterionic
polymers. L. Zhang et al. synthesized carboxybetaine nanogels using a disulfide crosslinker
that can be cleaved in the presence of dithiothreitol (DTT) [6]. Furthermore, they tested the
stability of their particles in PBS buffer for a period of six month. They showed that the
hydrodynamic radius of about 110 nm did not change during this time.
So far, no examples for the synthesis of zwitterionic microgels without the use of crosslinker
can be found in the literature. In contrast to L. Zhang’s work, these microgels are expected to
decompose in saline solution because the crosslinking ionic groups of the zwitterions are
screened and results in a disintegration of the microgel structure. The use of saline solution
offers a simple, non-toxic, and low-cost method for degradation.
33
5.3.2. Synthesis Procedure
Microgel Synthesis in Reactor VCL (1.877 g, 13.49 mmol), the respective betaine (see
Table 5.1), and AMPA (0.06 g, 0.221 mmol) are added in 150 mL of water in a double-walled
reactor. While stirring at 200 rpm, the mixture is purged with nitrogen at 70 °C. The reaction
is started by adding the initiator AMPA. The reaction is allowed to stir at 70 °C for 5 h.
Microgel solutions are cleaned via dialysis in a composite regenerated cellulose membrane
from Millipore (NMWCO 30,000) for 3 d against water.
Table 5.1 Amounts of betaines used for microgel synthesis.
m (Betaine) / g wt-% * n (SB) / mmol n (CB) / mmol n (PB) / mmol
0.02 1 0.072 0.083 0.071
0.04 2 0.143 0.165 0.142
0.06 3 0.215 0.248 0.213
0.08 4 0.286 0.330 0.285
0.10 5 0.358 0.413 0.356
* of betaines with respect to amount of VCL and BIS.
Microgel Synthesis in Calorimeter For calorimetric studies, the same amounts of monomers
were used. They were performed out in a reaction calorimeter RCle from Mettler Toledo with
a 500 mL 3-wall AP01-0.5-RTCal reactor equipped with a Hastelloy® stirrer, a baffle and a
Turbido™ turbidity probe from Solvias. The measurements were done in isothermal mode, in
which the desired reaction temperature (Tr) is set at a constant value and the jacket
temperature (Tj) changes automatically to maintain Tr at the desired value. Data were
evaluated with the software iControl RC1e™ 5.0. For the determination of the net reaction
time the end of the reaction was defined to be when the heat flow was back to a value close to
zero and constant.
For in-situ DLS studies, the amounts of monomers were down-scaled to 100 mL H2O.
Immediately after the addition of the initiator, the size of the particles was determined in 99
steps with 42 s intervals (30 s measurement + 12 s processing time).
34
5.3.3. Results and Discussion
PVCL-based microgels without crosslinker and varying betaines content (up to 5 wt-%) were
successfully synthesized via precipitation polymerization. No aggregation occurred during the
synthesis or after cooling down of the solution. FTIR measurements confirmed the successful
incorporation of SB into the microgel polymer network (graph see appendix).
In my diploma thesis, it was shown that the size of the particles changes in the same way as
microgels containing the crosslinker BIS: While carboxybetaine microgels increase in size
with increasing CB content, SB microgels do not change in size, and PB microgels decrease
in size. This very same behavior is also observed for non-crosslinked particles (see Figure
5.6).
Figure 5.6 Change of particle size RH with increasing amount of betaine for non-crosslinked
particles.
Both crosslinked and non-crosslinked sulfobetaine microgels were further analyzed with SLS
to obtain information about the radius of gyration (see Figure 5.7). Experiments were
performed below the VPTT (i.e. 20 °C).
Figure 5.7 SLS data for crosslinked and non-crosslinked zwitterionic microgels with 5 wt-%
SB.
35
SLS measurements provide the radius of gyration RG. It is 118 nm and 102 nm for crosslinked
and non-crosslinked samples, respectively. The ratio RG/RH gives information about the shape
parameter. It is 0.775 for an ideal, hard sphere [30][31]. The ratio RG/RH is 0.55 and 0.46 for
crosslinked and non-crosslinked microgels, respectively, and is thus much smaller than the
ratio for a hard sphere. V. Boyko gives the range of 0.3 - 0.6 as a typical range for microgels
[32]. He explained this behavior by the inhomogenous crosslinking density in a microgel
particle: it has a dense inner core with a less dense outer shell of dangling polymer chains.
Figure 5.7 shows that zwitterionic microgels crosslinked with BIS are closer to the value of a
hard sphere due to the fact that they are both chemically and physically crosslinked.
Zwitterionic microgels that were prepared without the use of the crosslinker BIS are only
physically crosslinked over intermolecular ionic bridges are thus less compact and have more
pending segments on the particle surface. M. Antonietti et al. already demonstrated that an
increase in crosslinker results in an increase in the ratio RG/RH [33]. They state that a microgel
with RG/RH < 0.6 is highly swollen both in the core and the shell.
AFM measurements give the particle size in dried state. Both crosslinked and non-crosslinked
samples with 1 and 5 wt-% SB were spin-coated on Si-wafer to see differences in the softness
of the particles (see Figure 5.8). These measurements were done solely for SB microgels since
a variation in betaine content does not lead to a change in the particle size (see Figure 5.6).
The following trends can be drawn from the AFM images: (1) Overview images show that
non-crosslinked microgels lead to a better and more even coating than crosslinked particles.
This might be caused by the fact that non-crosslinked samples are softer and can thus adapt
their structure to the surface in a better way than crosslinked particles. (2) The height profiles
reveal that both for 1 and 5 wt-% SB microgels, crosslinked particles have a bigger height
than non-crosslinked samples: 141 nm and 121 nm, respectively, for 1 wt-% SB, and 150 nm
and 144 nm, respectively, for 5 wt-% SB. Furthermore, non-crosslinked samples spread more
on the surface than crosslinked samples: 682 nm and 467 nm, respectively, for 1 wt-% SB,
and 387 nm and 366 nm, respectively, for 5 wt-% SB. The hydrodynamic radius as obtained
from DLS measurements for all samples is ~ 220 nm. This behavior arises from the fact that
crosslinked particles are better able to retain their shape when in contact with a solid surface
than non-crosslinked particles. The latter spread pancake-like and loose thus in height
compared to their crosslinked analogues (see Figure 5.8d). Furthermore, it is obvious that
non- crosslinked particles with a betaine content of 5 wt-% do not spread as pronounced as
microgels with 1 wt-% SB. The higher betaine content and thus the higher amount of internal
36
Figure 5.8 AFM images of 1 and 5 wt-% SB microgels, crosslinked and non-crosslinked.
From top to bottom: (a) Overview of Si-wafer (225 µm²); (b) Zoom-in with lines for height
profiles (25 µm²); (c) Height profile of marked particles in (b) with average sizes; (d)
Schematic draft of microgel shape on surface, in proportion to each other; (e) Zoom-in
(1 µm²); (f) Phase image of (d).
ionic bridges lead to a better stability and thus higher rigidity of the particles. (3) The phase
images of the zoomed-in images reveal that crosslinked particles have a “flat” surface, while
non-crosslinked particles have a “globular” sub-structure. This argues for a homogenous
distribution of the crosslinker BIS in the particles, while the sulfo groups lead to a
deformation and distortion of the network.
37
As shown in the AFM images above, the use of BIS as a crosslinker leads to different
properties of the microgels regarding their size and morphology. In the following,
calorimetric and kinetic studies were performed to understand differences during the synthesis
of crosslinked and non-crosslinked microgels. First, the synthesis was done in a calorimeter
measuring the polymerization heat and the change of turbidity during the synthesis. Second,
in-situ DLS measurements were performed to study the particle growth. Those experiments
were done for PVCL-SB (5 wt-% SB) microgels with and without BIS. Reference
measurements for several monomers and monomer combinations were done to study the
influence and contribution of each monomer to the particle formation process.
Figure 5.9 Heat profile qr (black), hydrodynamic radius RH (red), and turbidity (blue); (a)
non-crosslinked 5 wt-% SB microgels; (b) crosslinked 5 wt-% SB microgels. Measurements
were performed at 70 °C under nitrogen atmosphere.
Figure 5.9 gives the change of the heat profile qr, the hydrodynamic radius RH, and the
turbidity for the copolymerization of VCL/SB without the crosslinker BIS (a) and VCL/SB
with BIS (b). For the heat profile (black curve), two peaks can be identified, belonging to two
separate steps in the polymerization process: one flat peaks shortly after the start of the
reaction (0 to ~5 min and 5 to ~10 min), and one sharp peak (10 to ~ 15 min). The first peak is
not clearly distinguishable. Integration of the curve of the heat profile curves gives the
polymerization heat: it is 0.844 kJ and 0.735 kJ for crosslinked and non-crosslinked
microgels, respectively. Regarding the particle growth (red curve), three steps can be
observed which occur simultaneously with the two peaks seen for the reaction heat: first, the
particles grow for ~ 6 min (to 88 nm and 141 nm for non-crosslinked and crosslinked
particles, respectively), then the particle size remains stable until ~ 12 min, followed by
another growth in particle size (145 nm and 187 nm for non-crosslinked and crosslinked
particles, respectively). The main difference between non-crosslinked and crosslinked
38
particles can be observed during the last part: crosslinked particles grow more rapidly in size
than non-crosslinked particles. For non-crosslinked particles, the particle size remains stable
after ~ 24 min, for crosslinked particles already after ~ 12 min. The turbidity only shows two
separate phases: first a slow increase from ~ 10 % to 11 % after 10 min, then a faster increase
to 14.5 and 16 % for non-crosslinked and crosslinked particles, respectively.
In order to identify the two heat peaks, the monomers were synthesized separately with the
initiator AMPA (see Figure 5.10a). Here, it can be seen that SB does not react when
individually polymerized with AMPA. It is believed that AMPA is not the right initiator if SB
is polymerized alone. Probably, a non-ionic initiator such as azobisisobutyronitrile (AIBN)
would be better suited for this task. Contrary, VCL reacts immediately after the addition of
the initiator resulting in a narrow and sharp peak. The reaction is complete after ~ 10 min.
Integration of the peak gives a polymerization heat of 0.997 kJ.
Figure 5.10 Heat profile qr as measured with calorimetry (a) of monomers SB and VCL with
AMPA; (b) combinations of VCL with BIS and SB; (c) crosslinked and non-crosslinked 5 wt-
% SB microgels.
Furthermore, VCL was copolymerized with BIS alone (see Figure 5.10b). It can be seen that
the reaction is clearly delayed compared to the homopolymerization of VCL. Furthermore, the
former sharp peak is broadened. Its form resembles the broad second peak in Figure 5.10c.
The polymerization heat is 0.802 kJ. All polymerization heats are summarized in Table 5.2.
39
The following conclusions can be drawn from these observations for the identification of the
three peaks as observed in the copolymerization of VCL-SB with and without BIS: The first
peak is caused by SB. In a second step, VCL is incorporated into the microgel network.
Therefore, it can be assumed that the resulting microgel has a diffuse core-shell structure with
SB mostly located in the core of the particle, while VCL is located in the shell.
Table 5.2 Polymerization heats of copolymerization of VCL with SB, and/or BIS.
Monomers used for polymerization Polymerization heat / kJ
VCL-SB-BIS 0.844
VCL-SB 0.735
SB no reaction
VCL 0.997
VCL -BIS 0.802
As already explained above, the particles are crosslinked internally through ion pairing of the
quaternary ammonium group and the sulfo group (see Figure 5.5). This is indirectly proven by
the fact that DLS and AFM images show the formation of microgel particles that would be
loose polymer chains instead if no internal crosslinking would have taken place. Another
possibility to show the presence of those internal ion pairing is FTIR. Y. Su et al. showed for
their poly(acrylonitrile)-based zwitterionic membranes that FTIR is able to detect differences
between dissociated and undissociated sulfo groups [10]. Depending on the degree of
dissociation, the characteristic peak at 1042 cm-1 belonging to the symmetric stretch vibration
is split into two separate peaks.
Therefore, a crosslinked sample with 5 wt-% SB was analyzed with FTIR in three different
states: In dried state, in water, and in 1 M KCl solution that was stirred overnight. For this
experiment, a crosslinked sample was analyzed to ensure that the addition of KCl does not
lead to the degradation of the microgel particle which would influence not only the peaks of
interest but the whole spectrum.
40
Figure 5.11 FTIR spectra of 5 wt-% SB microgel sample in (black) solid state; (blue) in pure
water; (red) in 1 M KCl solution.
The peak of the sulfo group at 1040 cm-1 is most prominent in solid state (see Figure 5.11,
black curve). The spectra for the aqueous samples (blue curve) were subtracted with the
spectrum of pure water. After subtraction, the peak of the sulfo group is very weak due to the
small amount of sulfo groups in the sample. The positions of the peaks were identified by
determining the first deflection. It is 1038 cm-1 in pure water and 1045 cm-1 in 1 M KCl
solution. In agreement with the results shown by Y. Su et al., the peak at 1038 cm-1 is
associated with the undissociated sulfo group with the charged groups being oriented in an
anti-parallel way. The peak at 1045 cm-1 is induced by the dissociated sulfo groups, i.e. a
sulfo group being surrounded by counterions, in this case potassium. In contrast to Y. Su et
al., no splitting into two different peaks could be observed. This suggests that the dissociation
in 1 M KCl is complete and no transition states exist.
The uncrosslinked samples with the highest amount of betaine (i.e. 5 wt-%) were analyzed
concerning their change in size with increasing KCl content. It is expected that the
intramolecular ion bridges are dissolved by the addition of salt because attractive charge-
dipole and dipole-dipole interactions are screened leading to the degradation of the microgel
particles. Freeze-dried microgel samples were dissolved in saline solution and stirred for 3 d
to ensure complete disintegration of the polymer network.
It can be seen in Figure 5.12 that the particle size of all betaine microgels are unaffected at
low salt concentrations, i.e. up to an KCl concentration of 0.05 M KCl. Beginning at 0.1 M
KCl for all samples, a second, smaller peak at around 20 nm can be observed. This indicates
the disintegration of the microgel network to loose polymer chains.
41
Figure 5.12 Change of particle size distribution with varying salt concentration. (a)
sulfobetaine; (b) carboxybetaine; (c) phosphobetaine.
Images taken with FESEM support the data obtained with DLS. Figure 5.13 shows images of
particles with 5 wt-% SB (a) before the addition of salt and (b) after the addition of salt.
Figure 5.13a shows monodisperse particles with a diameter of 200 nm (the particles are in a
dried state and thus slightly smaller then they appear with DLS), while the particles in Figure
5.13b show aggregates and single particles of different sizes. Probably, the small particles of
~ 20 nm form aggregates during the drying process and are thus not seen in the image.
Figure 5.13 FE-SEM images of sulfobetaine microgels (5 wt-%) without crosslinker BIS. (a)
before the addition of salt; (b) after the addition of salt.
a b
c
198 nm
a b
2.00 µm 3.00 µm
42
This behavior can also be observed optically. Figure 5.14a shows non-crosslinked PVCL
microgels at 70 °C and 25 °C. At 70 °C, the microgels are in a collapsed state thus appearing
milky. At 25 °C, however, the swelling of the polymer network causes a rupture leading to
single polymer chains. Non-crosslinked PVCL-co-SB microgels, in contrast, are stable at both
70 °C and 25 °C supporting the assumption that zwitterionic bridges lead to an internal
crosslinking of the particles (Figure 5.14b). These internal ionic bridges are destroyed at room
temperature through the addition of salt indicated by the mostly colorless solution.
Figure 5.14 (a) Non-crosslinked PVCL microgels at 70 °C (left) and 25 °C (right); (b) non-
crosslinked PVCL-co-SB microgels at 70 °C (left), 25 °C (middle), and 25 °C in the presence
of salt (right).
Figure 5.15 (a) Sedimentation velocities of crosslinked (X) and non-crosslinked (non-X)
samples with 1 and 5 wt-% SB, respectively, with and without the addition of salt; (b)
Sedimentation front for sample 1 wt-% SB, non-crosslinked, with salt; drawn in are the
positions of air, water and dispersion for better clarification; (c) Sedimentation front for
sample 1 wt-% SB, non-crosslinked, without salt.
Since no apparent difference could be observed between SB, CB, and PB, the colloidal
stability of the particles was quantified with LUMiSizer measurements only for samples
containing 1 wt-% and 5 wt-% SB. Images of the samples after centrifugation show that there
43
is almost no difference between the crosslinked and the non-crosslinked samples if no salt is
present. In the presence of salt, the non-crosslinked samples are considerably less colloidally
stable than the respective crosslinked sample.
This behavior is also mirrored in the sedimentation fronts given in Figure 5.15b+c. The
sedimentation front is defined as the distance from the rotor center to the phase boundary
between sediment and supernatant [34]. After transforming these curves into transmission-time
curves (not shown), the sedimentation velocity can be calculated through the slope of the
curves. Figure 5.15b shows the sedimentation front for the non-crosslinked sample 1 wt-% SB
after the addition of salt. The sedimentation front moves from the left to the right side with
time indicating a fast precipitation of the particles. The same sample without the presence of
salt (Figure 5.15c) does not exhibit a moving sedimentation front thus showing a higher
colloidal stability.
The sedimentation velocities are summarized in Figure 5.15a. The following conclusions can
be drawn: (1) Samples with salt are less colloidally stable, i.e. they have a higher
sedimentation velocity, than samples that contain no added salt. For crosslinked samples, this
can be explained by the fact that the particles increase in size after the addition of salt due to
the antipolyelectrolyte effect [35][36][37]. The relation between particle size r and sedimentation
velocity ν is as follows [38]:
𝜈 = 2𝑟² ∆𝑝 ×𝑔
9𝜂 (1)
With Δp = density difference between the particle and the medium; g = centrifugal force; η =
viscosity of the medium.
Presuming that both the centrifugal force (for the sedimentation velocity of 2000 rpm and a
cuvette radius of 2 mm) and the viscosity are constant, it follows that the sedimentation
velocity is dependent on both the particle size and the density difference between the particle
and the medium. Assuming further that the change of density of the particle is negligible, the
sedimentation velocity is increasing with increasing particle size. (2) Non-crosslinked samples
are considerably less stable in the presence of salt than crosslinked samples: for the sample
1 wt-% SB, the sedimentation velocity is approximately 30 times higher than for the
respective crosslinked sample (115.7 µm/s and 4.324 µm/s, respectively). This is due to the
degradation of the microgel network. (3) The sample 1 wt-% SB is less stable in the presence
of salt than the sample 5 wt-% SB (115.7 µm/s and 66.67 µm/s, respectively). This is because
the sample with the higher amount of betaine is better able to preserve the internal ionic
bridges than the sample with a lower amount of betaine.
44
5.3.4. Summary
In this chapter, non-covalently crosslinked (i.e. without the use of the crosslinker BIS)
zwitterionic microgels were studied. Zwitterionic microgels can be synthesized without the
use of a crosslinker since the positively charged ammonium group and the negatively charged
sulfo groups are expected to build internal ionic groups that are sufficiently strong enough to
hold the microgel network together. The analysis focuses on two main aspects: first, the
polymerization process itself and second, the degradability of the particles in saline solution.
The polymerization process was studied with calorimetry and in-situ DLS. For both
crosslinked and non-crosslinked particles, two steps could be identified: first, SB was
polymerized, followed by VCL. This suggests that the resulting microgels have a diffuse core-
shell structure with SB located in the particle interior. Furthermore, it could be seen that the
presence of the crosslinker BIS fastens the polymerization process.
Differences in morphology and rigidity between crosslinked and non-crosslinked particles
could be seen with AFM. Particles without the crosslinker spread in a higher extent on a solid
surface than crosslinked particles. This is expected since crosslinked particles are more rigid
than non-crosslinked particles. Furthermore, differences could be observed between particles
with a different content of SB: particles with more SB are more rigid and do not spread in the
same extent than particles with a lower amount of SB.
Since the microgels are crosslinked via internal ionic bridges, the particles were expected to
degrade after the addition of salt that is able to screen the charges. First, the existence of these
internal bridges was verified with FTIR. It could be shown that the peak of the sulfo group at
1038 cm-1 in water is shifted to 1045 cm-1 in 1 M KCl. The first peak indicates the presence of
the undissocaiated sulfo groups with internal ionic bridges. The shift suggests the dissociation
of the sulfo group that is surrounded by potassium ions.
After the addition of KCl, DLS measurements showed the formation of smaller peaks of about
20 nm at higher KCl concentrations indicating the presence of loose polymer chains. Since no
apparent difference could be observed between SB, CB, and PB, colloidal stability
measurements were done solely for SB microgels. Again, it could be seen that non-
crosslinked microgels are considerably less stable in KCl solution that crosslinked samples,
the sedimentation velocity was up to 30 times higher. Eventually, FE-SEM images supported
the assumption of the degradation of non-crosslinked microgel particles in saline solution.
45
5.4. Inverse Mini-Emulsion Polymerization
5.4.1. Introduction
An important factor in precipitation polymerization is the hydrophobicity of the formed
polymers. It is difficult to incorporate larger amounts of hydrophilic comonomers in
microgels by precipitation polymerization [39]. This behavior was shown in chapter 5.3 where
microgels with up to only 5 wt-% of betaines could be synthesized via precipitation
polymerization. To increase the amount of the zwitterionic comonomer in the microgels, a
different technique, namely inverse mini-emulsion (or nano-emulsion) free radical
polymerization, is used and discussed in this chapter. This polymerization technique allows
the incorporation of betaine amounts of 30 mol-% and higher.
Mini-emulsions are aqueous dispersions of oil droplets that are stabilized by a surfactant [40].
They are formed mostly through shear forces implied by ultrasonication, microfluidizer or a
homogenizer that break the monomer droplets which results in the formation of droplets in a
size range of 20 – 500 nm [41]. In this work, ultrasonication is used to create small droplets.
With every wave, ultra sound produces a compression and rarefaction of molecules. During
the rarefaction, small cavitation bubbles form and grow that implode again in one of the
following compressions. The collapse of the cavitation bubbles forms powerful shock waves
in the surrounding liquid. The resulting liquid jets of high speed are the reason for larger
monomer droplets to disperse in the continuous phase. The longer the ultrasonication time and
the higher the ultrasound power, the smaller the droplet size becomes [42][43].
The system usually consists of a dispersed phase, a continuous phase, one or more surfactants
and a strong hydrophobe. The surfactant protects droplets from diffusional degradation
through steric hindrance. Due to the small droplet size, the resulting large droplet surface area
leads to the main amount of surfactant being absorbed on the droplet surface. Thus, only a
small amount of surfactant is able to form micelles. The surfactant greatly influences the
resulting particle size. The more surfactant is used, the more the particles are covered by the
surfactant and subsequently the smaller the particles become. Generally, a smaller size will
also lead to a broader particle size distribution [44]. Due to their size, dispersions appear
transparent or translucent. Syntheses found in literature use anionic, cationic, non-ionic or a
mixture of surfactants, though the first are the most common used. Sodium dodecyl sulfate
(SDS) is often use for the preparation of mini-emulsions [43][45][46]. Non-ionic surfactants are
rarely used in classical emulsions since they are often soluble in the monomer droplets. This
is different for mini-emulsions: Here, the larger surface area leads to a larger fraction of non-
46
ionic surfactant at the surface in a thermodynamic equilibrium. They are mostly used in
controlled living polymerization with atom transfer radical polymerization (ATRP) where
anionic surfactants cannot be used [47].
A co-stabilizer is mostly a low-molecular weight compound that is insoluble in water and
prevents diffusional degradation [48]. The hydrophobe is the key agent in a mini-emulsion. It is
monomer-soluble, but highly water-insoluble (less than 10-7 ml/mL [42]) and thus does not
diffuse between the different monomer droplets through the aqueous phase [44]. In a classical
emulsion polymerization, the diffusion of monomers from small droplets to larger droplets
(i.e. Ostwald ripening) reduces the total surface energy. In a mini-emulsion, however, the
monomer diffusion will increase the concentration of the hydrophobe and thus result in an
increase in free energy. An osmotic pressure is created inside the monomer droplet that
counterveils the Laplace pressure that is induced by the interface energy (i.e. the pressure
difference between inside and outside the droplet). This ideally stabilizes the monomer
droplets completely against Ostwald ripening. In reality, however, the hydrophobe only
retards the monomer diffusion, though the destabilization of droplets is in the range of months
[49]. It was shown that in contrast to the surfactant, the hydrophobe does not influence the
resulting droplet size [40][50]. Usually, high-molecular weight compounds such as long-chain
alkanes and long-chain alcohols are used as hydrophobes [51]. In some papers, the hydrophobe
is also called co-surfactant. This, however, is a misleading term, since the hydrophobe
stabilizes the monomer droplet not on the surface but in the bulk [48]. Due to the lack of
diffusion between droplets they act as “nanoreactors” [52]. As a result, the size as well as the
composition of the latex particles is identical with the composition of the monomer phase. In
contrast to a classical emulsion polymerization, the polymerized particles have the same size
as the initial monomer droplets as demonstrated by K. Landfester et al. with surface tension
measurements and conductometry (see Figure 5.16) [45].
Figure 5.16 Schematic representation of polymerization processes for mini-emulsion.
Water-in-oil mini-emulsions are called inverse mini-emulsions and are used for the formation
of hydrophilic, water soluble polymers such as polyacrylamide, polyacrylic acid or
polymethacrylic acid. Here, initiation can occur in the hydrophobic phase or within the water
47
droplet. Instead of a strong hydrophobe as required in direct mini-emulsions, a lipophobe such
as water or salt is used. Because oil is used as the continuous phase, non-ionic surfactants
with a low HLB (hydrophilic-lipophilic balance) are favorable [51].
Inverse mini-emulsions were first introduced by K. Landfester et al. in 2000 [40]. They
prepared hydroxyethyl methacrylate polymers in cyclohexane as the continuous phase. They
showed that the minimal concentration for the surfactant was 1.6 wt-% with respect to the
monomer. For the synthesis of polyacrylamide, an ionic salt was added as a lipophobe and
greatly increased the stability of the emulsion droplets.
Inverse mini-emulsions are used to synthesize a spectrum of polymers. Many examples in
literature deal with the synthesis of acrylamides by inverse mini-emulsion. S. Wiechers and
G. Schmidt-Naake, for instance, prepared pH-sensitive copolymers of 2-acrylamido-2-methyl-
1-propanesulfonic acid and 1-vinylimidazole with a particle size of 200 nm and demonstrated
that the conversion is dependent on the type of surfactant used during the reaction [53]. D.
Klinger et al. prepared enzymatically degradable microgels based on acrylamide and dextran
methacrylate in water and cyclohexane [54]. The use of two different initiators allowed the
synthesis at 70 °C and 37 °C to obtain monodisperse, spherical particles as confirmed with
SEM.
There are several examples for the preparation of microgels with various sensitivities via
mini-emulsion. To obtain microgels, they have to be crosslinked before they are transferred to
the continuous phase [55]. H. Dong, for instance, prepared magnetic microgels based on
di(ethylene glycol) methyl ether methacrylate via nano-emulsion [56]. The authors used
activators generated by electron transfer atom transfer radical polymerization (AGET ATRP)
as a polymerization technique to obtain degradable, temperature-sensitive microgels that were
used for the entrapment of iron oxide nanoparticles. D. Klinger and K. Landfester synthesized
photo-cleavable PMMA-based microgels via mini-emulsion by free radical polymerization
[52]. They obtained stable emulsion with particles with a size of 140 to 200 nm. Later, the
same authors prepared microgels based on 2-hydroxyethyl methacrylate and methacrylic acid
that were both sensitive to pH and UV light [39]. In contrast to the previous work, these
particles were formed in an inverse mini-emulsion.
48
5.4.2. Synthesis Procedure
VCL, the appropriate amount of sulfobetaine (Table 5.3), 3 wt-% BIS, and 5 wt-% NaCl as a
lipophobe were dissolved in 3.8 mL H2O. In a second flask, 0.60 g (1.400 mmol) Span 80 and
0.20 g (0.153 mmol) Tween 80 (ratio 3:1) were dissolved in 100 mL heptane and dispersed
with ultrasonic. The aqueous solution was added dropwise to the oil phase while
ultrasonicating for 5 min at level 4, 10 %, under ice cooling to obtain a stable emulsion.
Afterwards, polymerization was carried out within 5 min. The emulsion was heated up to
70 °C under nitrogen atmosphere and vigorous stirring. The initiator AMPA was solved in
0.2 mL H2O and added to the emulsion. The reaction was carried out for 2 h.
After cooling down, heptane was removed from the emulsion via centrifugation at 11,000 rpm
at room temperature. The precipitate was washed and centrifuged alternately with water and
heptane for three times each. Between centrifugation, the microgels were shaken in the
respective solvent for 30 min. Afterwards, the samples were dialyzed against water for 7 d.
Table 5.3 Amounts of betaines used for microgel synthesis.
Ratio
VCL:SB
n(VCL)
/ mmol
m(VCL)
/ g
n(betaine)
/ mmol
m(betaine)
/ g *2
2:1 2.443 0.340 1.222 0.341
1:1 2.443 0.340 2.443 0.682
1:2*1 1.222 0.170 2.443 0.683
*1 For this batch, half of the amount of solvents and monomers was used.
*2 Before synthesis, it was ensured that SB is not soluble in heptane in the concentration
range used for the syntheses.
Further experiments were conducted in the same way as described above except synthesis was
performed in a smaller batch using 50 mL of heptane and 2 mL of water (a) varying the
crosslinker amount (see Table 5.4) and (b) substituting VCL for N-vinylpyrrolidone (VPy)
and N-vinylpiperidone (VPi) (see Table 5.5).
Table 5.4 Variation in crosslinker concentration. Amounts of other monomers are identical to
amounts given in Table 5.3. The ratio VCL:SB was 1:1.
m (BIS) / g n (BIS) / mmol % BIS in microgel
0.002 0.015 0.5
0.005 0.031 1
0.008 0.053 2
0.013 0.083 3
49
Table 5.5 Amounts of N-vinylpyrrolidone (VPy) and N-vinylpiperidone (VPi). The ratio
vinylamid:SB was 1:1 in all cases. 3 wt-% of crosslinker was used.
n(VPi)
/ mmol
m(VPi)
/ g
n(VPy)
/ mmol
m(VPy)
/ g
n(betaine)
/ mmol
m(betaine)
/ g *2
1.173 0.147 - - 1.178 0.329
- - 1.19 0.131 1.183 0.330
Samples were purified in the same way as described above.
50
5.4.3. Results and Discussion
Inverse mini-emulsion was performed using Span 80 and Tween 80 as surfactants. They are
both non-ionic and biodegradable [57]. Tween 80 is more hydrophilic than Span 80 due to its
high number of PEG groups. In combination with Tween 80, Span 80 acts as a co-emulsifier.
Span 80 and Tween 80 have HLB (Hydrophile Lipophile Balance) values of 4.3 and 15.0,
respectively. In a ratio of Tween80/Span80 = 1:3, an overall HLB of ~ 7 is achieved
according to the following equation:
(2)
with X = overall HLB.
According to literature, the mixture of a non-ionic surfactant with a low HLB and a nonionic
one with a high HLB improves the stability of inverse mini-emulsions [58]. The overall HLB
should ideally be between 3 and 8 [59].
Figure 5.17 Schematic structure of surfactant orientation in water droplets.
AMPA was chosen as the initiator because of its solubility in water. It therefore starts the
reaction within the water droplets.
Before the synthesis of microgels via inverse mini-emulsion, various reactions were carried
out to analyze the stability of the emulsion droplets. Therefore, the oil and the water phase
were ultrasonicated and the emulsion measured with DLS for 60 min. As expected, the
stability of the droplet size could be influenced by the amount of surfactant used in the
reaction (graph see appendix).
51
An emulsion with a ratio between Tween80/Span80 = 0.1/0.3 breaks immediately. A fourfold
amount of surfactant, in contrast, is stable over the complete measurement time of 60 min.
Since a high amount of surfactant requires a long time of cleaning via dialysis, a third sample
was analyzed. Here, the amount of surfactant was 0.2/0.6. Measurements showed that the
droplets are only stable up to 20 min. This time, however, is thought to be sufficient to initiate
and formulate monodisperse microgel particles. Before breaking, the droplets have a
hydrodynamic radius of 257 ± 25 nm. Breaking of the mini-emulsion can be attributed to two
factors: (1) Droplets coalescence due to Brownian motion and van der Waals forces; (2)
Droplet degradation through monomer diffusion [48]. Coalescence occurs when the adhesion
energy between two droplets is equal or larger than the turbulent energy that keeps the
droplets separated.
The concentrations of surfactants used in the syntheses are above the cmc of both Span 80 and
Tween 80 (1.8 * 10-5 mol/L and 1.2 * 10-5 mol/L, respectively). A test measurement using
concentrations below the cmc’s of both surfactants showed that the monomer droplets are not
stable and aggregate quickly with a PDI between 0.7 and 1.0 (graph shown in the appendix).
Microgel samples were synthesized via inverse mini-emulsion polymerization. After
synthesis, they were cleaned via repeated centrifugation and dialysis. 1H-NMR measurements
(not shown) showed that removal of the surfactants took approximately 1 week. Re-dispersion
in water after freeze-drying resulted in stable microgel dispersions. Colloidal stabilization was
obtained through the use of the cationic initiator AMPA.
First, the incorporation of betaine in the microgel was measured with FTIR (for spectrum see
appendix). The peak of PVCL at 1672 cm-1 is set in proportion to the peak of SB at 1050 cm-1.
By using a calibration line (see appendix), the actual ratio between VCL and SB can be
determined. The data is in good agreement with the theoretical values (see Table 5.6).
Table 5.6 Amount of incorporated SB as determined with FTIR spectroscopy.
Sample name n(VCLtheo)
/ mol
n(SBtheo)
/ mol
n (VCLFTIR)
/ mol
n (SBFTIR)
/ mol
1:2 1 2 2 3*
1:1 1 1 1 1
2:1 2 1 1.4 1
* The correct ratio is 61/100 ~ 2/3.
The size of the particles was then determined with DLS (Table 5.7). Particles become slightly
smaller with increasing content of SB, while the PDI remains constant at ~ 0.04. According to
52
K. Landfester, the narrow particle size distribution is attributed to the fact that all droplets
nucleate nearly at the same time [44]. The microgel size obtained for the sample VCL:SB =
1:1, i.e. 227 nm, correlates closely with the size of the water droplets before the
polymerization.
Table 5.7 Hydrodynamic radius RH and polydispersity index (PDI) for samples of different
composition.
VCL:SB Size (TEM)
/ nm
R,dried (TEM)
/ nm
RH (DLS) /
nm
PDI
(DLS)
RH (DLS) /
nm (after 3
months)
PDI (DLS)
(after 3
months)
2:1 - - 242 ± 16 0.043 243 ± 7 0.052
1:1 with BIS 459 ± 18 230 ± 9 227 ± 19 0.044 221 ± 8 0.048
1:2 - - 216 ± 8 0.047 211 ± 11 0.041
The particle size was measured a second time after a time period of 3 months. The sizes of all
particles as well as their PDI remained stable.
In order to obtain an optical confirmation of this result, TEM images were taken of the sample
with a ratio VCL:SB = 1:1. Figure 5.18 shows that the particles have a spherical shape and are
monodisperse. In dried state, they have a radius of 230 ± 9 nm. This is very close to the result
obtained via DLS. One might expect the size in dried state is smaller than the hydrodynamic
radius in aqueous solution, but particles presumably spread pancake-like on the surface, thus
appearing bigger in size.
Next, the influence of varying temperature on the particles’ properties was investigated. Pure
PVCL microgels have a broad transition with a VPTT of 30.7 °C (Figure 5.18a). This
temperature is well reported in literature [60][61]. The VPTT is shifted towards higher
temperatures for a ratio of VCL:SB = 2:1 to 37.7 °C. Interestingly, a second VPTT appears at
low temperatures if the content of SB is further increasing. For a ratio of VCL:SB = 1:1,
transition temperatures can be seen at 9.7 °C and 42.9 °C. Both are shifted further to higher
temperatures for a ratio of VCL:SB = 1:2, 18.7 °C and 48.5 °C. Below the first VPTT and
above the second VPTT, particles are in a collapsed state, while they are swollen in the
temperature range in between. This behavior for linear VCL-co-SB copolymers was already
described by B. Yang et al [12]. While the transition temperatures are similar, a major
difference between linear copolymers and microgels is visible: While the microgel particles
described here are stable within the whole temperature range, linear copolymers aggregate
53
above and below the LCST and UCST, respectively. The latter’s hydrodynamic diameter
increases rapidly from ~ 40 nm to above 2 µm.
An increase of both VPTTs can be explained by the weakening of the hydrophobic effects of
VCL induced by the caprolactam ring with an increasing amount of zwitterionic, hydrophilic
SB in the polymer network. This effect is not yet seen for a VCL:SB ratio of VCL:SB = 2:1
indicating that a specific amount of SB is needed for the formation of a doubly
thermosensitive microgel.
Figure 5.18 Hydrodynamic radius RH of PVCL-co-SB microgels as a function of temperature
measured with DLS. (a) Pure PVCL microgel; (b) VCL:SB = 2:1; (c) VCL:SB = 1:1; (d)
VCL:SB = 1:2. The TEM image belongs to the sample VCL:SB = 1:1.
B. Yang et al. describe the arrangement of different polymer chain segment depending on the
temperature: Between the UCST and LCST, both VCL and SB are hydrophilic, so that long
linear chains can be observed [12]. If the temperature falls below the UCST, SB becomes
hydrophobic due to intra- and intermolecular bridges between the zwitterionic groups and
subsequently, aggregates with an SB core and a VCL shell are formed. Analogously, at
temperatures above the LCST, VCL becomes hydrophobic whereby hydrogen bonds between
VCL and water break and subsequently, aggregates with a VCL core and an SB shell are
formed.
This behavior can only partly be transferred to VCL-co-SB microgels. Since the polymer
network is chemically crosslinked a complete “inversion” of an SB/VCL core/shell is not
54
possible. Still, due to the change in particle size, a “shifting” of polymer segments to more
favorable positions in a less pronounced way as for linear polymer chains is expected.
Figure 5.19 Absorbance of VCL-co-SB microgels as a function of temperature. (a) VCL:SB
= 2:1; (b) VCL:SB = 1:1; (c) VCL:SB = 1:2.
The position of a phase transition of the microgels can also be observed in UV-Vis
measurements as a change of absorbance (Figure 5.19). At the transition point where microgel
particles go from a swollen to a collapsed state, the turbidity of the microgel dispersion
increases, thus leading to an increase in absorbance.
Similar to DLS measurements, two transition temperatures can be seen for the samples
VCL:SB = 1:1 and 1:2. These measurements confirm the results obtained with DLS.
Figure 5.20 Hydrodynamic radius RH (right axis, upper curves) and electrophoretic mobility
EM (left axis, lower curves) as a function of pH.
55
The sulfo group of sulfobetaine has a low pKa value of -3 whereby the it remains
deprotonated over the whole pH range [62]. Therefore, the electrophoretic mobility is negative
for all samples within the pH 3 to 10 and hardly pH-dependent (see Figure 5.20). The same
can be said about the particle size. Due to the formation of an internal salt within the microgel
network, the particles do not swell with varying pH.
The colloidal stability of the microgel solutions was examined measuring their sedimentation
velocities below and above the VPTT (i.e. T = 25 °C and 50 °C). The samples were
centrifuged while a laser detects the de-mixing of the samples over time. The higher the
sedimentation speed is, the lower is the colloidal stability of the solution.
Figure 5.21 (a) Sedimentation velocities indicating colloidal stability for mini-emulsion
samples with different monomer ratios. Measurements were conducted at T = 25 °C (black
bars) and 50 °C (red bars) and pH = 7. (b) Sedimentation diagram for VCL:SB = 1:1.
Figure 5.21a shows that all crosslinked samples have approximately the same sedimentation
velocities. Below the VPTT, the sedimentation velocities are ~ 20 µm/s. They increase by ca.
50 % at T = 50 °C. This is caused by a loss of steric repulsion and an increase of Van-der-
Waals attractions. Because swollen microgels have a Hamaker constant close to that of water
at room temperature, attractive Van-der-Waals attractions can be neglected [63]. Attractive
forces increase with increasing temperature, therefore leading to a decrease in colloidal
stability above the VPTT. A complete aggregation is, however, prevented by an increase in
electrostatic repulsion between charges on the surface originating from the initiator.
The polydispersity of non-crosslinked samples can also be seen in the sedimentation curves.
Vertical lines that move to the right side of the diagram as seen in Figure 5.21b are caused by
a slow sedimentation of monodisperse particles.
The swelling behavior of the particles can be tuned by varying the crosslinking content.
Samples discussed above contained 3 wt-% of BIS. The amount was decreased to 2 wt-%,
a
b
56
1 wt-% and 0.5 wt-%. L. A. Shah et al. claim that a small increase in crosslinker content does
not influence the position of the VPTT [64]. Beginning from 10 wt-% of crosslinker content,
the transition temperature is increased due to the solubility of BIS in the polymer network. A
change in the crosslinker content, however, changes the extent of swelling of particles. A
higher crosslinker content leads to a higher particle stiffness and thus a suppressed swelling
ability [65][66]. Figure 5.22 shows that the position of both transition temperatures is hardly
effected by a variation in crosslinker content. While all microgel dispersions have roughly the
same size of ~ 205 nm at 5 °C, their degree of swelling is considerably different in the range
of 14 °C to 35 °C. The higher the crosslinking content, the smaller are the particle size in this
range due to an increase in polymer network stiffness.
Figure 5.22 Thermosensitive behavior of VCL-co-SB microgels (VCL:SB = 1:1) with
varying crosslinker content. (a) Hydrodynamic radius RH (measured with DLS). (b)
Absorbance (measured with UV-Vis). Microgels containing only PVCL are given as a
reference.
AFM images (see Figure 5.23) confirm the variation in microgel stiffness. Height profiles
show that the particles spread more the lower the content of crosslinker is, thus appearing
smaller on the Si-wafer surface. While particles have an average inter-particle distance of ~
90 nm for crosslinker contents of 2 and 3 wt-%, there are no gaps between particles with the
lowest crosslinker content of 0.5 wt-%.
a b
57
Figure 5.23 Height profiles and AFM images of VCL:SB = 1:1 microgels with various
crosslinker content.
Both PVCL and PNIPAm polymers exhibit an LCST of ~ 31 - 37 °C [67]. For some
applications, a shift of this narrow temperature range to different temperatures is necessary. It
is well known in literature that the copolymerization with a hydrophobic comonomer lowers
the transition temperature, while the incorporation of hydrophilic comonomers shifts the
transition to higher temperatures. In the following, a different approach will be followed
where PVCL is substituted for its five- and six-ring homologues (see Figure 5.24).
Figure 5.24 Structures of N-vinylpyrrolidone (VPy), N-vinylpiperidone (VPi), and N-
vinylcaprolactam (VCL).
N-vinylpyrrolidone (VPy) contains a five carbon ring. Its polymer, poly-N-vinylpyrrolidone
(PVPy) is water-soluble, hydrophilic, non-ionic and exhibits good biocompatibility and low
cytotoxicity [68][69]. Its transition temperature in aqueous solution is above 100 °C. Adding salt
to the dispersion decreases the transition temperature considerably [70].
5 µm
5 µm
5 µm
5 µm
58
Compared to PVCL and PVPy, few reports have been published about the
(co-)polymerization of N-vinylpiperidone (VPi) and the properties of its polymer, poly(N-
vinylpiperidone) (PVPi). In 2010, V. N. Kizhnyaev et al. were the first to polymerize VCL
with VPi [71]. Ieong et al. described the synthesis of homo- and copolymers via RAFT
polymerization [72]. The homopolymer exhibits a sharp LCST between 87 – 68 °C depending
on the molecular weight.
Figure 5.25 1H-NMR spectra of monomers N-vinylpyrrolidone, N-vinylpiperidone, N-
vinylcaprolactam, and sulfobetaine. Graphs in inset are close-ups of the respective microgel
spectra (vinylamide:SB = 1:1) that show peaks of the respective vinylamide and sulfobetaine
that were used to calculate the ratio between the two monomers.
1H-NMR spectra confirm the successful copolymerization of the respective vinylamide and
SB. For the calculation of the ratio between the two monomers, the peaks of the vinylamide
and SB at 2.99 and 3.25 ppm, respectively, were used (see Figure 5.25). The calculated ratios
are in accordance with the feed ratios (see Table 5.8).
Table 5.8 Feed and calculated ratios between monomers as obtained from 1H-NMR.
Sample Vinylamidefeed VinylamideNMR SBfeed SBNMR
VCL:SB 1 1 1 0.94
VPi:SB 1 1 1 1.07
VPy:SB 1 1 1 1.18
59
In the following, the influence of the lactam-ring size on the temperature behavior is
examined. Microgels containing homologues of VCL with five- and six-ring are prepared in
the same way via inverse mini-emulsion. A ratio of vinylamide:SB of 1:1 with 3 wt-% BIS is
used.
First, the temperature sensitivity is studied with DLS and UV-Vis. DLS measurements reveal
that the LCST of the vinylamide is shifted to higher temperatures with decreasing ring size.
PVCL and PVPi have a sharp LCST of 42.9 °C and 73.0 °C, respectively, while no LCST can
be observed for PVPy. The UCST of SB remains constant at around 10 °C. It is noticeable
that additionally, the UCST becomes broader with decreasing ring size. UV-Vis
measurements confirm the results obtained via DLS (see Figure 5.26).
Figure 5.26 Thermosensitive behavior of vinylamide-co-SB microgels (vinylamide:SB = 1:1)
with varying ring-size content. (a) Hydrodynamic radius RH (measured with DLS). (b)
Absorbance (measured with UV-Vis).
60
5.4.4. Summary
In this chapter, zwitterionic microgels that were synthesized via inverse mini-emulsion
polymerization were discussed. It was shown that it is possible to incorporate a higher amount
of sulfobetaine (SB) than is possible with precipitation polymerization. Here, even ratios of
VCL:SB = 1:2 were accomplishable. The successful incorporation of high amounts of SB was
verified with FTIR. Samples with a ratio of VCL:SB = 1:1 and 1:2 do not show only an LCST
at 42.9 °C and 48.5 °C, respectively, as is typical for VCL-based microgels, but also an UCST
at 9.7 °C and 18.7 °C, respectively. Both UCST and LCST increase with increasing amount of
SB. This behavior was also confirmed by UV-Vis measurements since absorbance increases
for particles in a collapsed state. As expected for SB microgels, the particle size and their
electrophoretic mobility are not dependent on pH due to the low pKa value of the sulfo group.
Samples synthesized with varying amount of crosslinker were analyzed with DLS and AFM.
DLS measurements confirm that the degree of swelling is dependent on the amount of
crosslinker in the microgel particles. An increase in crosslinker leads to a suppressed swelling
while the positions of the phase transition temperatures remain constant. The decreasing
stiffness with decreasing content of crosslinker was also observed with AFM. Particles with
the lowest amount of crosslinker spread flat on the surface of a Si-wafer, while stiffer
particles were better able to retain their spherical shape.
Furthermore, the influence of the caprolactam-ring size on the temperature-sensitive behavior
was studied. Therefore, the VCL homologues N-vinylpyrrolidone (VPy) and N-
vinylpiperidone (VPi) were used as copolymers for the preparation of vinylamide-co-SB
microgels. 1H-NMR measurements confirmed the successful incorporation of both monomers
in ratios in accordance with the feed ratios. It was shown that the position of the UCST
originating from the SB moieties remains constant for all samples, while the LCST
originating from the vinylamide moieties are strongly dependent on the ring size. VCL- and
VPi-based samples have LCSTs of 42.9 °C and 73.0 °C, respectively, while VPy-based
microgels have no LCST below 92 °C.
61
5.5. Literature
[1] A. B. Lowe, C. L. McCormick, Chem. Rev. 2002, 102, 4177–89.
[2] M. Das, N. Sanson, E. Kumacheva, Chem Mater 2008, 20, 7157–7163.
[3] M.-C. Sin, S.-H. Chen, Y. Chang, Polym. J. 2014, 46, 436–443.
[4] J. Menzie, H. Prentice, J.-Y. Wu, Brain Sci. 2013, 3, 877–907.
[5] N. Y. Kostina, C. Rodriguez-Emmenegger, M. Houska, E. Brynda, J. Michálek,
Biomacromolecules 2012, 13, 4164–70.
[6] L. Zhang, H. Xue, Z. Cao, A. Keefe, J. Wang, S. Jiang, Biomaterials 2011, 32, 4604–8.
[7] L. Mi, M. T. Bernards, G. Cheng, Q. Yu, S. Jiang, Biomaterials 2010, 31, 2919–25.
[8] P. Liu, J. Song, Biomaterials 2013, 34, 2442–54.
[9] Y. Chang, W.-Y. Chen, W. Yandi, Y.-J. Shih, W.-L. Chu, Y.-L. Liu, C.-W. Chu, R.-C.
Ruaan, A. Higuchi, Biomacromolecules 2009, 10, 2092–100.
[10] Y. Su, L. Zheng, C. Li, Z. Jiang, J. Phys. Chem. B 2008, 112, 11923–8.
[11] Z. Zhang, T. Chao, S. Jiang, J. Phys. Chem. B 2008, 112, 5327–32.
[12] B. Yang, C. Wang, Y. Zhang, L. Ye, Y. Qian, Y. Shu, J. Wang, J. Li, F. Yao, Polym.
Chem. 2015, DOI 10.1039/C5PY00123D.
[13] J. Chang, Y. Tao, B. Wang, B. Guo, H. Xu, Y. Jiang, Y. Huang, J. Mater. Chem. B
2015, 3, 1097–1105.
[14] T. P. de Souza, H. Chaimovich, A. Fahr, B. Schweitzer, A. Agostinho Neto, I. M.
Cuccovia, J. Colloid Interface Sci. 2012, 371, 62–72.
[15] Z. Zhang, S. Chen, S. Jiang, Biomacromolecules 2006, 7, 3311–5.
[16] J. G. Weers, J. F. Rathman, F. U. Axe, C. a. Crichlow, L. D. Foland, D. R. Scheuing,
R. J. Wiersema, A. G. Zielske, Langmuir 1991, 7, 854–867.
[17] M. Szafran, Z. Dega-szafran, A. Katrusiak, G. Buczak, T. Głowiak, J. Sitkowski, L.
Stefaniak, X. A. Surfactants, J. Org. Chem. 1998, 63, 2898–2908.
[18] S. Chen, J. Zheng, L. Li, S. Jiang, J. Am. Chem. Soc. 2005, 127, 14473–8.
[19] S. Bian, J. Zheng, W. Yang, J. Polym. Sci. Part A Polym. Chem. 2014, 52, 1676–1685.
[20] D. Klinger, K. Landfester, Soft Matter 2011, 7, 1426.
[21] S. Seiffert, W. Oppermann, K. Saalwächter, Polymer (Guildf). 2007, 48, 5599–5611.
[22] G. Agrawal, J. Wang, B.Brüster, X.Zhu, M. Möller, A. Pich, Soft Matter 2013, 9, 5380.
[23] S. B. Jhaveri, K. R. Carter, Macromolecules 2007, 40, 7874–7877.
[24] R. Månsson, G. Frenning, M. Malmsten, Biomacromolecules 2013, 14, 2317–25.
[25] L. Bromberg, M. Temchenko, V. Alakhov, T. A. Hatton, Langmuir 2005, 21, 1590–8.
[26] S. V. Ghugare, E. Chiessi, B. Cerroni, M. T. F. Telling, V. G. Sakai, G. Paradossi, Soft
Matter 2012, 8, 2494.
[27] V. Bulmus, Y. Chan, Q. Nguyen, H. L. Tran, Macromol. Biosci. 2007, 7, 446–55.
[28] A. M. Hawkins, T.A. Milbrandt, D.A. Puleo, J. Z. Hilt, Acta Biomat. 2011, 7, 1956–64.
[29] M. Patenaude, T. Hoare, ACS Macro Lett. 2012, 1, 409–413.
62
[30] V. Boyko, S. Richter, W. Burchard, K.-F. Arndt, Langmuir 2007, 23, 776–84.
[31] I. Varga, T. Gilányi, R. Mészáros, G. Filipcsei, M. Zrínyi, J. Phys. Chem. B 2001, 105,
9071–9076.
[32] V. Boyko, N -Vinylcaprolactam Based Bulk and Microgels : Synthesis , Structural
Formation and Characterization by Dynamic Light Scattering, Dresden University of
Technology, 2004.
[33] M. Antonietti, W. Bremser, M. Schmidt, Am. Chem. Soc. 1990, 23, 3796–3805.
[34] S. Schachschal, A. Balaceanu, C. Melian, D. E. Demco, T. Eckert, W. Richtering, A.
Pich, Macromolecules 2010, 43, 4331–4339.
[35] P. K. Singh, V. K. Singh, M. Singh, e-Polymers 2007, 7, 1–34.
[36] W.-F. Lee, Y.-J. Chen, J. Appl. Polym. Sci. 2001, 81, 2888–2900.
[37] S. E. Kudaibergenov, N. Nuraje, V. V. Khutoryanskiy, Soft Matter 2012, 8, 9302.
[38] V. Boyko, A. Pich, Y. Lu, S. Richter, K.-F. Arndt, H.-J. P. Adler, Polymer (Guildf).
2003, 44, 7821–7827.
[39] D. Klinger, K. Landfester, Macromolecules 2011, 44, 9758–9772.
[40] K. Landfester, M. Willert, M. Antonietti, Macromolecules 2000, 33, 2370–2376.
[41] C. Solans, P. Izquierdo, J. Nolla, N. Azemar, M. Garciacelma, Curr. Opin. Colloid
Interface Sci. 2005, 10, 102–110.
[42] B. Abismaïl, J. P. Canselier, a. M. Wilhelm, H. Delmas, C. Gourdon, Ultrason.
Sonochem. 1999, 6, 75–83.
[43] K. Landfester, N. Bechthold, F. Tiarks, M. Antonietti, Macromolecules 1999, 32,
5222–5228.
[44] K. Landfester, F. J. Schork, V. a Kusuma, Comptes Rendus Chim. 2003, 6, 1337–1342.
[45] K. Landfester, N. Bechthold, S. Förster, M. Antonietti, Macromol. Rapid Commun.
1999, 20, 81–84.
[46] C. S. Chern, J. C. Sheu, Polymer (Guildf). 2001, 42, 2349–2357.
[47] M. Li, K. Matyjaszewski, Macromolecules 2003, 36, 6028–6035.
[48] J. M. Asua, Prog. Polym. Sci. 2002, 27, 1283–1346.
[49] H. Huang, H. Zhang, J. Li, S. Cheng, B.Tan, J. Appl. Polym. Sci. 1998, 68, 2029–2039.
[50] A. Pich, S. Datta, a. Musyanovych, H. J. P. Adler, L. Engelbrecht, Polymer (Guildf).
2005, 46, 1323–1330.
[51] K. Landfester, A. Musyanovych, Adv. Polym. Sci 2010, 234, 39–63.
[52] D. Klinger, K. Landfester, Soft Matter 2011, 7, 1426.
[53] S. Wiechers, G. Schmidt-Naake, Macromol. React. Eng. 2008, 2, 126–134.
[54] D.Klinger, E.M.Aschenbrenner, C.K.Weiss, K. Landfester, Polym. Chem. 2012, 3, 204.
[55] A. Pich, W. Richtering, Chemical Design of Responsive Microgels, Springer Berlin
Heidelberg, Berlin, Heidelberg, 2011.
[56] H. Dong, V. Mantha, K. Matyjaszewski, Chem. Mater. 2009, 21, 3965–3972.
63
[57] “Span and Tween,”
http://www.crodaoilandgas.com/home.aspx?view=dtl&d=content&s=155&r=378&p=2
404&prodID=169, 20.06.2015
[58] M. Barari, M. Abdollahi, M. Hemmati, Iran. Polym. J. 2011, 20, 65–76.
[59] A. Renken, D. Hunkeler, Polymer (Guildf). 1999, 40, 3545–3554.
[60] E. E. Makhaeva, H. Tenhu, A. R. Khokhlov, Macromolecules 1998, 31, 6112–6118.
[61] H. Feil, Y. H. Bae, J. Feijen, S. W. Kim, Macromolecules 1993, 26, 2496–2500.
[62] G. Jander, K. F. Jahr, G. Schulze, Maßanalyse, Walter De Gruyter, 2002.
[63] J. Wu, G. Huang, Z. Hu, Macromolecules 2003, 36, 440–448.
[64] L. A. Shah, J.Chem.Soc.Pak. 2013, 35, 1522–1529.
[65] S. Wellert, M. Richter, T. Hellweg, R. von Klitzing, Y. Hertle, Zeitschrift für Phys.
Chemie 2014, 0, DOI 10.1515/zpch-2014-0568.
[66] E. Mah, R. Ghosh, Processes 2013, 1, 238–262.
[67] N. A. Yanul, Y. E. Kirsh, E. V. Anufrieva, J. Therm. Anal. Calorim. 2000, 62, 7–14.
[68] F. Haaf, a. Sanner, F. Straub, Polym. J. 1985, 17, 143–152.
[69] S. Vijaykumar, S. Prasannkumar, B. S. Sherigara, N. B. Shelke, T. M. Aminabhavi, B.
S. R. Reddy, Macromol. Res. 2009, 17, 1003–1009.
[70] Y. Maeda, T. Nakamura, I. Ikeda, Macromolecules 2002, 35, 217–222.
[71] V. N. Kizhnyaev, E. A. Krakhotkina, T. L. Petrova, G. V. Ratovskii, O. V. Tyukalova,
F. A. Pokatilov, A. I. Smirnov, Polym. Sci. Ser. B 2010, 52, 480–486.
[72] N. S. Ieong, M. Redhead, C. Bosquillon, C. Alexander, M. Kelland, R. O’Reilly,
Macromolecules 2011, 44, 886-893.
64
6 Microgels with Statistically Distribution of Ionizable Groups
6.1. Introduction
Microgels that contain both acidic and basic functional groups are ampholytes [1]. They
exhibit different behavior than nonionic, anionic, cationic or zwitterionic microgels.
Polyelectrolyte microgels having only basic or acidic groups swell with a change to acidic or
basic pH, respectively, due to the ionization of the respective groups resulting in a repulsion
of like charges. Upon the addition of salt, they collapse due to the screening of electrostatic
repulsion forces leading to aggregation [1]. The particle size of ampholyte microgels with
anionic and cationic groups based on strong bases and acids, in contrast, is independent on the
pH [2]. The particles do not swell or de-swell with a change in pH. More common, however,
are ampholyte microgels with anionic and cationic groups based on weak bases and acids:
they swell at both high and low pH. In between, both charges are present creating a
zwitterionic microgel that results in a de-swelling of the particle due to attractive interactions
between positive and negative charges [3]. The presence of an isoelectric point (IEP) is a
characteristic feature of ampholyte microgels. The position of the IEP can be shifted by
varying the composition of the microgel [2]. Ampholyte microgels de-swell in a smaller extent
or even swell if salt is added because intramolecular ionic crosslinks that are present in the
absence of salt or at low salt concentrations are broken. This behavior is called “anti-
polyelectrolyte” effect.
Many reports can be found in literature on the synthesis of ampholyte microgels, they have,
however, mostly a core-shell structure [4][5][6]. If no core-shell structure is prepared, the
distribution of acidic and basic groups is hardly investigated or mentioned [7][8][9]. K.
Christodoulakis et al. prepared 2-(diethylamino)ethyl methacrylate-co-methacrylic acid
microgels (DEA-co-MAA) with DEA and MAA either located in a core-shell structure or
randomly distributed [4]. The latter structure was prepared by the simultaneous addition of
both monomers during the reaction. The ratio MAA:DEA was 56:44. SEM images reveal the
spherical shape of the monodisperse particles that have a size of ~ 350 nm. The particles were
stained with potassium hexachloroplatinate or cadmium nitrate. PtCl62- and Cd2+ selectively
stain the positively charged DEA and the negatively charged MAA groups, respectively. TEM
images show uniformly stained particles, however, a close look-up at a single particle for
better visualization is lacking.
In this chapter, ampholyte microgels with statistically, i.e. homogenously, distributed
ionizable groups are synthesized and characterized. N-Vinylcaprolactam (VCL) was used as a
main monomer to obtain temperature-sensitive microgels. N-isopropylacrylamide (NIPAm)
65
microgels were prepared for comparison. The particles are moreover sensitive to changes in
pH and ionic strength [3][7][4][10] influencing their properties such as particle size and
electrophoretic mobility. To obtain ampholyte microgels, itaconic acid (IA) and 1-
vinylimidazole (Vim) (see Figure 6.1) were used as comonomers. In this work, Vim and IA
were used because of their favorable pKa values. Itaconic acid has a pKa of 3.84 and 5.55
[11][12], 1-vinylimidazole has a pKa of 6.0 [13]. This means that microgel particles containing
both monomers are expected to be positively charged below pH 6.0 and negatively charged
above pH 5.5. The isoelectric point (IEP) is expected to be between pH 5.5 and 6.0,
depending on the amount of IA and Vim incorporated in the microgel particles.
Figure 6.1 Temperature, pH, and salt sensitivity of ampholyte microgels.
C. Erbil et al. have shown for copolymers of NIPAm and IA that copolymers with a statistical
distribution of the monomers in the copolymer are obtained [14]. A. Pich have demonstrated
for NIPAm and Vim that the latter is homogenously distributed in particles [15]. For VCL and
Vim, on the contrast, Vim is thought to be in the loosely crosslinked shell. The authors
explain this by the presence of AAEM. A. Pich and his group have shown that VCL and
AAEM form microgels with core-shell structure with AAEM in the center and VCL in the
shell due to different reactivity ratios [15][16][17]. This “template” forces Vim to accumulate in
the shell and not homogenously throughout the whole particle. Therefore, AAEM was not
used a comonomer because this structure is thought to hinder the formation of microgels
where IA and Vim are homogenously distributed in the particle.
Itaconic acid can easily be used as a comonomer [11] and increases the hydrophilicity of the
microgel due to the presence of two carboxylic groups in the side chain (see Figure 6.2) [18]. 1-
vinylimidazole is a weak base having a nitrogen atom that is protonated at low pH. In
literature, it is often used as comonomer to NIPAm-based microgels due to its metal-binding
properties [19][20].
66
Figure 6.2 Chemical structure of itaconic acid and 1-vinylimidazole.
While the synthesis and characterization of microgels based on NIPAm with either IA [21] or
even more so with Vim [15][22][23][24] have been studied in literature, only S. Schachschal et al.
have – to our knowledge – prepared microgels containing both IA and Vim [5]. They prepared
core-shell microgels based on VCL with IA in the core and Vim in the shell and studied the
effect of pH and temperature change on the swelling behavior. The properties of and
preparation techniques for core-shell microgels is more closely looked at in chapter 7.
Experimental data described in this chapter were published in Macromolecules, 2015 (DOI:
10.1021/acs.macromol.5b01305).
67
6.2. Synthesis Procedure
Microgels were synthesized via precipitation polymerization in water. Experiments showed
that for microgels with a higher amount of IA, the synthesis performed at pH 10, and for
microgels with a higher amount of Vim or equal amounts of IA and Vim, the synthesis done
at pH 3 lead to colloidal stable particles. Appropriate amounts (see Table 6.1 and 6.2) of
NIPAm/VCL, IA, Vim, and BIS were dissolved in 80 mL water and heated up to 70 °C while
purging with N2. After 1 h, the initiator AMPA was added and the reaction carried out for 5 h
under constant stirring. After synthesis, microgel solutions were cleaned via dialysis using a
composite regenerated cellulose membrane from Millipore (NMWCO 30,000) for 3 d against
water and subsequently lyophilized.
Table 6.1 Amounts of monomers used for PVCL microgel synthesis.
Molar ratio
Vim:IA
VCL / g BIS / g IA / g Vim / g AMPA / g
0:20 0.8 0.040 0.187 0 0.03
5:15 0.8 0.040 0.140 0.034 0.03
10:10 0.8 0.040 0.093 0.068 0.03
15:5 0.8 0.040 0.047 0.101 0.03
20:0 0.8 0.040 0 0.135 0.03
5:5 0.8 0.030 0.042 0.030 0.03
15:15 0.8 0.038 0.160 0.116 0.04
20:20 0.8 0.044 0.249 0.180 0.04
25:25 0.8 0.053 0.374 0.270 0.04
Table 6.2 Amounts of monomers used for PNIPAm microgel synthesis.
Molar
ratio
Vim:IA
NIPAm / g BIS / g IA / g Vim / g AMPA / g
0:20 0.8 0.041 0.230 0 0.03
5:15 0.8 0.042 0.173 0.042 0.03
10:10 0.8 0.041 0.115 0.083 0.03
15:5 0.8 0.041 0.058 0.125 0.03
20:0 0.8 0.041 0 0.166 0.03
68
6.3. Results and Discussion
In this part, ampholyte microgels based on either PVCL or PNIPAm were synthesized. In
both cases, the ratio between IA and Vim was varied (see Table 6.1 and Table 6.2).
Furthermore, for PVCL microgels, particles with equal amount of ionizable groups were
synthesized. Here, the molar amount of IA and Vim was increased from 5 to 25 mol-%.
PVCL and PNIPAm microgels with unequal amount of ionizable groups
The molar ratio between the two ionizable groups is thought to influence the properties of the
microgel particles. Therefore, FTIR measurements were done to ensure that the actual ratio is
in agreement with the theoretical value. The FTIR spectra of the monomers VCL, NIPAm, IA
and Vim are shown in Figure 6.3.
N-Vinylcaprolactam shows peaks at 2934 cm-1 (C-H stretching band), 1654 cm-1 (C=O
stretching), 1620 cm-1 (amide I band), and 1481 cm-1 (C-N stretching). The following bands
derive from the deformation of –CH2: 1433 cm-1, 1391 cm-1, 1320 cm-1, 1256 cm-1, 1183 cm-1,
967 cm-1, and 874 cm-1. These results are in good agreement with literature [25][26][27].
N-isopropylacrylamide has peaks at 2962 cm-1 (C-H stretching band), 1657 cm-1 (amide I),
1623 cm-1 (C=O stretching), 1548 cm-1 (amide II), and 1241 cm-1 (amide III). These results
are in accordance with literature [28].
Itaconic acid shows peaks at 3070 cm-1 (CH2 vibration), 1675 cm-1 (C=O stretching), 1372
cm-1 (C-O-H in-plane bending), 1212 cm-1 (C-O stretching), 1066 cm-1 (O-H out-of-plane
bending). These results can found in literature [29] [30] [31].
1-vinylimidazole shows peaks at 3112 cm-1 (N=CH, C=CH stretching), 3005 cm-1 (CH2
asymmetric stretching), 1649 cm-1 (vinyl off-ring), 1511 cm-1 (C=N, C=C stretching), 1496
cm-1 (CH2 bending and CH2 stretching), 1420 cm-1 (CH2 bending and ring stretching), 1372
cm-1 (CH2 und CH wagging), 1326 cm-1 (CH2 and CH wagging), 1281 cm-1 (ring vibration),
1228 cm-1 (ring vibration), 1106 cm-1 (CH in-plane bending), 1081 cm-1 (CH in-plane
bending), 960 cm-1 (CH out-of-plane bending), 820 cm-1 (CH out-of-plane bending), 735 cm-1
(CH2 rocking), 655 cm-1 (ring torsion), and 597 cm-1 (N=CH and C=CH wagging). These
results are in good agreement with spectra found in literature [32].
Peaks at ca. 3400 cm-1 in all spectra derive from OH from water.
Table 6.3 and 6.4 give the amounts of IA and Vim incorporated into the microgel particle as
obtained from FTIR measurements. The amount of VCL and PNIPAm is set to 80 mol%,
respectively. For PVCL and PNIPAm, the peaks of the carbonyl group at 1654 and 1657 cm-1,
69
Figure 6.3 FTIR spectra of (a) monomers VCL, NIPAm, IA, and Vim; (b) VCL:Vim:IA =
80:10:10 microgel as a representative sample. Measurements were conducted at 25 °C.
respectively, were used as references. In all cases, the amounts of IA and Vim as obtained
from FTIR are close to the theoretical values verifying that the incorporation of IA and Vim
in the desired amounts was successful.
Table 6.3 Amounts of ionizable groups in PVCL microgels as obtained with FTIR.
Ratio
VCL:Vim:IA
VCL %
(1654 cm-1
)
Vim %
theor.
Vim % FTIR
(1228 cm-1
)
IA % theor. IA % FTIR
(1675 cm-1
)
80:0:20 80 0 0 20 16
80:5:15 80 5 4 15 14
80:10:10 80 10 10 10 11
80:15:5 80 15 13 5 6
80:20:0 80 20 18 0 0
Table 6.4 Amounts of ionizable groups in PNIPAm microgels as obtained with FTIR.
Sample NIPAm %
(1657 cm-1
)
Vim %
theor.
Vim % FTIR
(1228 cm-1
)
IA % theor. IA % FTIR
(1675 cm-1
)
80:0:20 80 0 0 20 16
80:5:15 80 5 5 15 13
80:10:10 80 10 6 10 7
80:15:5 80 15 17 5 6
80:20:0 80 20 21 0 0
a b
70
Microgel particles with different amounts of Vim and IA are expected to behave differently
with varying pH with regard to their electrophoretic mobility and their size. Therefore, those
measurements were conducted between pH 3 and 10.
Figure 6.4 shows the dependence of the electrophoretic mobility on the variation of pH. The
black and red curves belong to the mono-ionic microgels with only IA and Vim, respectively.
Microgels with solely IA as comonomer (black curve) are negatively charged throughout the
whole pH range. No differences can be detected between PVCL and PNIPAm microgels.
Microgels with solely Vim (red curve) as comonomer are positively charged with varied pH.
Microgel particles with both IA and Vim are positively charged at low pH and negatively
charged at high pH, thus exhibiting an isoelectric point (IEP). The position of the IEP is
dependent on the amount of IA and Vim present. The more IA is incorporated the lower the
IEP. Thus, PVCL microgels with 5 mol% IA have an IEP of 6.5, 10 mol% of 6.2, and
15 mol% of 5.0. As for PNIPAm microgels, the respective IEPs are at 5.9, 4.8, and 3.9 and
thus slightly lower.
Figure 6.4 Electrophoretic mobility of PVCL (left side) and PNIPAm (right side) particles
with varying pH. The horizontal line at 0 µmcm/Vs is a visual help to better identify the
isoelectric point (IEP). Error bars are not included in order not to overload the graphic, but are
in average ± 0.04 µmcm/Vs.
The shifts as well as the height of the measured electrophoretic mobilities further support that
varying amount of ionizable groups were incorporated for different samples.
The presence of charges influences the size of the particles. Ampholyte microgels swell in
acidic and basic regions due to electrostatic repulsion. On the other hand, the particles
collapse at or close to the IEP.
71
Figure 6.5 shows the change in size with varying pH. To better foreground the dependence
between electrophoretic mobility and particle size, the electrophoretic mobility is shown for
each sample again.
Figure 6.5 Change of particle size (RH, black curve) and electrophoretic mobility (EM, red
curve) with varying pH. Left side: PVCL microgels. Right side: PNIPAm microgels. The
content of itaconic acid is decreasing from top (20 mol%) to bottom (0 mol%).
For samples containing both acidic and basic groups, a minimum in size (black curve) can be
observed that overlaps with the IEP (red curve). At the IEP, no charges are present in the
microgel leading to a collapse in the internal structure. PVCL microgels with a Vim:IA ratio
of 10:10 have an IEP of 6.2 and a hydrodynamic radius of 94 nm.
At pH 3 and pH 10, particles are swollen due to electrostatic repulsion between charged
groups. Vim is positively charged at low pH, while IA is negatively charged at high pH. For
samples with a Vim:IA ratio of 10:10, the change in size from pH 3 to pH 6.2 and from pH
6.2 to pH 10 should be equal if the acidic and basic moieties are distributed equally. An
unequal change in size would indicate a core-shell structure where the charged monomer that
leads to a more prominent change in size if situated in the shell. Figure 6.5 shows an equal
change in size for both PVCL and PNIPAm microgels indicating a homogenous distribution
of ionizable groups. From their minimum at the IEP, PVCL microgels swell 77 % at pH 3 and
90 % at pH 10. For PNIPAm, the swelling degree is 71 % at pH 3 and 79 % at pH 10.
VCL:Vim:IA =
72
In all cases, the change in size with varying pH is very broad. Particles do not change their
size abruptly but gradually over the whole pH range. This indicates a homogeneous
distribution of ionizable groups. Because of their distribution all over the particle, ionizable
groups do not interact with the environment simultaneously, but gradually.
These results show that it is possible to synthesize particles that are highly pH-sensitive and
that their response to changes in pH can be tuned and controlled by the amount of
incorporated ionizable groups.
Furthermore, it is tested whether the presence of charges has an influence on the temperature
sensitivity of the particles. These trends are measured at pH 3 because the particles are more
stable due to the presence of positive charges. Measurements at the IEP of each sample were
not possible because the particles aggregated when the temperature was increased.
Measurements at pH 10 are expected to mirror the results obtained at pH 3 and were thus not
performed.
As shown in various works by A. Pich [5][33], pure PVCL microgels have a Volume Phase
Transition Temperature (VPTT) of 28-35 °C at neutral pH. S. Schachschal [5] et al. have
shown that the incorporation of Vim lead to a shift of the VPTT at pH 4 to over 35 °C. A
charged microgel is more hydrophilic and thus better able to contain the surrounding water
molecules and shift the collapse to higher temperatures.
Figure 6.6 Change of RH with varying temperature at pH 3. (a) PVCL microgels; (b)
PNIPAm microgels.
Figure 6.6 shows that all samples exhibit broad temperature transitions between 30 and 35 °C
with no distinguishable trend. It was expected that the VPTT would differ dramatically
between samples that are fully charged (80:20:0) and samples that have no charges (80:0:20)
at pH 3. The sample containing only Vim was expected to have a significantly higher VPTT
because the protonated Vim groups show a better solvation in water and thus a higher
temperature is needed to disrupt hydrogen bridges.
a b
73
While the VPTT does not change significantly, the extent of de-swelling differs from pure
PVCL microgels containing no ionizable groups. As shown in previous works by A. Pich,
PVCL microgels are twice as big in the swollen as in the de-swollen state [34] (i.e. the ratio
between the size at 20 °C and 40 °C, R40°C/R20°C, is approximately 2.0). Similar behavior can
be observed for the samples 80:0:20, i.e. with IA as the only comonomer. At pH 3, no charges
are present and the microgel particle is thus neutral. As a consequence, it behaves like pure
PVCL microgels with R50°C/R20°C = 2.12 for PVCL and 2.05 for PNIPAm. The opposite can
be seen for the samples 80:20:0 where Vim as the only comonomer is positively charged. For
PVCL and PNIPAm, R50°C/R20°C is 1.54 and 1.57, respectively. The positive charges prevent a
complete collapse of the network. The same is true for the ampholyte microgels with both
Vim and IA. Here, the ratio R50°C/R20°C is between 1.56 and 1.70.
Figure 6.7 AFM in-situ images of microgels with a ratio of VCL:Vim:IA = 80:5:15. First
row: zoom-in images of particles; second row: force constants of particles shown in the first
row; third row: cross sections (see red line in first row). Force constant images are slightly
shifted due to adjustments between measurement modes. (a) pH 3; (b) pH ~ 6.5; (c) pH 10.
All measurements were conducted in water at 20 °C.
AFM in-situ measurements were conducted in water at pH 3, 6.5, and 10. Images in
Figure 6.7 confirm that the particles are spherical and monodisperse and swollen in water at
pH 3 and pH 10, while they are in collapsed state at pH close to the IEP. This agrees with the
data obtained via DLS. Additionally, the force constants were measured to evaluate
a b c
276 ± 50 nm 156 ± 9 nm 291 ± 36 nm
74
mechanical properties of the particles (see second row in Figure 6.7). Hereby, the darker a
spot in the image, the softer the spot is. Purely white spots are generated by the glass
substrate. All particles have two areas: a lightly colored, soft shell and a darker, stiffer core.
This does not display a “real” core-shell structure, but is caused by different thicknesses of
those two areas due to the spreading of the soft microgel particles: the cantilever “senses” the
thick core and the thinner shell by indentation. Determination of the force constants gives the
Young’s Modulus of the particle at each spot. The data is given in Figure 6.8. Since “core”
and “shell of the polymer particle differ considerably, they are analyzed separately.
Figure 6.8 (a) Young’s Modulus for the sample VCL:Vim:IA = 80:5:15. Error bars were
determined by evaluating a representable number of data points in Figure 6.7. (b) AFM
cantilever approaching the surface. The Young’s Modulus E is defined as E = F/z with F =
force and z = distance.
Figure 6.8a shows the Young’s moduli (YM) for VCL:Vim:IA = 80:5:15 microgels at various
pH. A low YM is attributed to a high stiffness indicating that particles are in a collapsed state.
While the YM for the shell does not change significantly, the YM of the core is strongly
dependent on the pH of the dispersion. At low and high pH, particles have a high YM above
1000 kPa. At pH = IEP, the stiffness increases due to attractive forces between opposite
charges that induces the collapse of the particle network. A decrease in stiffness correlates
with a swelling of the particles at low and high pH and confirm the previous DLS
measurements (see Figure 6.5). It can also be seen that particles have a higher YM at pH 10
than at pH 3. At pH 10, particles are swollen due to the deprotonation of the carboxylic
groups. The asymmetric ratio of Vim:IA = 5:15 leads to a higher swelling at pH 10 and thus a
lower YM.
a b
75
PVCL microgels with equal amount of ionizable groups
It was shown that microgel particles with a changing ratio between Vim and IA express
different behavior regarding their location of the IEP and their swelling behavior with varying
temperature and pH. Particles with an equal ratio between Vim and IA, but an increasing
amount of ionizable groups should show similar behavior regarding the location of the IEP,
but different extends in swelling behavior. In the following, samples with Vim:IA ratios of
5:5, 10:10, 15:15, 20:20, and 25:25 are discussed. Since no significant differences were
detected between PVCL and PNIPAm microgels for a Vim:IA ratio of 10:10 in the chapter
above, these particles were only prepared with VCL.
To ensure that the desired amount of ionizable groups is incorporated in the particles, FTIR
measurements were performed.
Table 6.5 Amounts of ionizable groups in PVCL microgels as obtained with FTIR.
Sample VCL %
(1619 cm-1)
Vim %
theor.
Vim % FTIR
(665 cm-1
) IA % theor.
IA % FTIR
(1454 cm-1
)
90:5:5 90 5 6 5 4
80:10:10 80 10 10 10 11
70:15:15 70 15 14 15 15
60:20:20 60 20 16 20 16
50:25:25 50 25 21 25 22
As seen in Table 6.5, FTIR data confirm that the desired amounts of IA and Vim are in
agreement with the actual amounts.
TEM images support the data obtained with FTIR. The sample based on PVCL with a Vim:IA
ratio of 20:20 is taken as an example to visualize the distribution of ionizable groups. The
sample is stained with uranyl acetate that binds to the negatively charged carboxylic groups.
The heavy metal ion U3+ enhances the contrast by increasing the electron density [35] and
black dots visualize its location and distribution in the microgel particle. This method is used
by many groups such as Jones and Lyon [36] and Xing [37].
Figure 6.9 shows the distribution of ionizable groups for the sample with a Vim:IA ratio of
20:20. The image shows slightly deformed spherical particles due to the removal of water
while drying. The images support the thesis of a random distribution of itaconic acid in the
microgel particle. Particles also appear to be rather polydisperse. According to Lyon [36], this
is an illusion caused by an differing spreading of the particles on the TEM grid. Information
about the polydispersity of a sample is better analyzed with DLS.
76
Figure 6.9 Transmission electron image of PVCL microgel particles with a ratio of
VCL:Vim:IA = 60:20:20. The image on the right side is a zoom-in.
Figure 6.10 Change of (a) electrophoretic mobility EM and (b) hydrodynamic radius RH of
PVCL microgels with equal amount of Vim:IA.
In the previous section, the behavior of PVCL microgels with a Vim:IA ratio of 10:10 was
discussed. Samples with an equal but increasing ratio should show very similar behavior
when the pH of the solution is changed. Figure 6.10 shows the change of the electrophoretic
mobility EM and the hydrodynamic radius RH as a function of pH. The graphs indicate that
the degree of swelling becomes more pronounced with a higher concentration of ionizable
groups. This shows that the electrostatic repulsion becomes stronger when more charges are
incorporated. Furthermore, it can be seen the size of particles in collapsed state decreases with
increasing amount of ionizable groups. Particles with a ratio of Vim:IA = 5:5 have a
hydrodynamic radius of ~ 100 nm at pH 6.2, while particles with a ratio of 25:25 have a
hydrodynamic radius of ~ 55 nm.
a b
77
6.4. Summary
This chapter showed that it is possible to synthesize ampholyte microgels with balanced as
well as unbalanced amounts of acidic and basic moieties with statistical distribution of
ionizable groups. Both PVCL and PNIPAm microgels were prepared with itaconic acid (IA)
and 1-vinylimidazol (Vim) as comonomers. The successful incorporation of both
comonomers in desired amounts was verified with FTIR spectroscopy. Also TEM images of
stained samples indicated the random distribution of ionizable groups. It was further shown
that the size of the particles can be tuned with the pH of the solution. For all ampholyte
microgels, a V-shaped curve could be obtained, this means that particles were swollen at high
and low pH and collapsed at the isoelectric point (IEP). These data correlated very well with
measurements for the determination of the electrophoretic mobility. The position of the IEP
and thus the minimum of the size curve could be shifted to higher or lower pH depending on
the amount of IA and Vim. Samples containing only Vim or IA as a comonomer were only
swollen in one pH regime. Microgels with a balanced ratio between IA and Vim, but different
amounts all showed the typical V-shaped curve for their sizes with varying pH with an IEP at
~ 6.2. It could be seen, however, that particles with the highest amount of comonomers
(VCL:Vim:IA = 50:25:25) were smallest in the collapsed state compared to samples with a
lower amount of ionizable groups (VCL:Vim:IA = 90:5:5). Furthermore, the degree of
swelling at pH 3 and pH 8, respectively, compared to pH = IEP was higher for samples with a
higher amount of ionizable groups. This was explained thereby that a higher amount of
charges leads to more electrostatic repulsion between the charged groups thus extending the
microgel network.
AFM images in buffer were taken of samples with an unbalanced amount of ionizable groups
and verified the production of spherical, monodisperse microgel particles. Force maps of the
same particles revealed that particles are stiffer, that means they have a higher Young’s
modulus, at pH = IEP than at pH 3 and pH 8. These results correlate with the results obtained
with DLS. At pH = IEP, the particles are in a collapsed state and are therefore stiffer than in
the swollen state.
78
6.5. Literature
[1] T. Hoare, R. Pelton, Biomacromolecules 2008, 9, 733–740.
[2] M. Bradley, B. Vincent, G. Burnett, Aust. J. Chem. 2007, 60, 646–650.
[3] S. E. Kudaibergenov, N. Nuraje, V. V. Khutoryanskiy, Soft Matter 2012, 8, 9302.
[4] K. E. Christodoulakis, M. Vamvakaki, Langmuir 2010, 26, 639–47.
[5] S. Schachschal, A. Balaceanu, C. Melian, D. E. Demco, T. Eckert, W. Richtering, A.
Pich, Macromolecules 2010, 43, 4331–4339.
[6] W. Richtering, A. Pich, Soft Matter 2012, 8, 11423.
[7] H. Ni, H. Kawaguchi, T. Endo, Macromolecules 2007, 6370–6376.
[8] M. Kashiwabara, K. Fujimoto, H. Kawaguchi, Colloid Polym Sci 1995, 273, 339–345.
[9] K. Ogawa, A. Nakayama, E. Kokufuta, Langmuir 2003, 19, 3178–3184.
[10] M. Bradley, B. Vincent, G. Burnett, Colloid Polym. Sci. 2008, 287, 345–350.
[11] T. Willke, K.-D. Vorlop, Appl. Microbiol. Biotechnol. 2001, 56, 289–295.
[12] J. Li, T. B. Brill, J. Phys. Chem. A 2001, 105, 10839–10845.
[13] T. Hakamatani, S. Asayama, H. Kawakami, Nucleic Acids Symp. Ser. (Oxf). 2008,
677–8.
[14] C. Erbil, B. Terlan, Ö. Akdemir, A. T. Gökçeören, Eur. Polym. J. 2009, 45, 1728–
1737.
[15] L. Shen, A. Pich, D. Fava, M. Wang, S. Kumar, C. Wu, G. D. Scholes, M. a. Winnik, J.
Mater. Chem. 2008, 18, 763.
[16] G. Agrawal, M. Schürings, X. Zhu, A. Pich, Polymer (Guildf). 2012, 53, 1189–1197.
[17] C. Cheng, X. Zhu, A. Pich, M. Möller, Langmuir 2010, 26, 4709–16.
[18] M. Babic, J. Jovasevic, J. Filipovic, S. Tomic, Hem. Ind. 2012, 66, 823–829.
[19] M. Muratalin, P. F. Luckham, J. Colloid Interface Sci. 2013, 396, 1–8.
[20] P. Chen, L. Cao, G. Wang, J. Wang, Polym. Adv. Technol. 2013, 24, 764–771.
[21] M. K. Krušić, J. Filipović, Polymer (Guildf). 2006, 47, 148–155.
[22] L. Li, C. Cheng, M. P. Schürings, X. Zhu, A. Pich, Polymer (Guildf). 2012, 53, 3117–
3123.
[23] S. Bhattacharya, F. Eckert, V. Boyko, A. Pich, Small 2007, 3, 650–7.
[24] A. Pich, A. Tessier, V. Boyko, Y. Lu, H. P. Adler, Macromolecules 2006, 7701–7707.
[25] Y. Maeda, T. Nakamura, I. Ikeda, Macromolecules 2002, 35, 217–222.
[26] J. Spěváček, J. Dybal, L. Starovoytova, A. Zhigunov, Z. Sedláková, Soft Matter 2012,
8, 6110.
[27] S. Sun, P. Wu, J. Phys. Chem. B 2011, 115, 11609–18.
[28] B. Özkahraman, I. Acar, S. Emik, Polym. Bull. 2010, 66, 551–570.
[29] N. B. Milosavljević, M. Đ. Ristić, A. a. Perić-Grujić, J. M. Filipović, S. B. Štrbac, Z. L.
Rakočević, M. T. K. Krušić, Colloids Surfaces A Physicochem. Eng. Asp. 2011, 388,
59–69.
79
[30] T. Betancourt, J. Pardo, K. Soo, N. A. Peppas, J. Biomed. Mater. Res. A 2010, 93, 175–
88.
[31] R. M. Silverstein, F. X. Webster, D. Kiemle, Spectrometric Identification of Organic
Compounds, John Wiley & Sons, 2005.
[32] J. E. Hansen, B. I. Rickett, J. H. Payer, H. Ishida, J. Polym. Sci. Part B Polym. Phys.
1996, 34, 611–621.
[33] V. Boyko, A. Pich, Y. Lu, S. Richter, K.-F. Arndt, H.-J. P. Adler, Polymer (Guildf).
2003, 44, 7821–7827.
[34] A. Balaceanu, D. E. Demco, M. Martin, A. Pich, Macromol. React. Eng. 2011, 44,
2161–2169.
[35] T. Hoare, R. Pelton, Curr. Opin. Colloid Interface Sci. 2008, 13, 413–428.
[36] C. Jones, L. Lyon, Macromolecules 2000, 33, 8301–8306.
[37] Z. Xing, C. Wang, J. Yan, L. Zhang, L. Li, L. Zha, Colloid Polym. Sci. 2010, 288,
1723–1729.
80
7 Microgels with Core-Shell Structure
7.1. Introduction
Due to the broad application range for microgels, microgel architectures tend to become more
and more complex in order to meet multiple requirements in one multifunctional microgel
polymer network. The next step from microgels with a statistical distribution of ionizable
groups that were discussed in the last chapter, with regard to advanced architecture are
microgels with a core-shell structure with ionizable groups spatially separated. These are the
subject of this chapter.
Core-shell microgels consist of two chemically crosslinked networks [1]. Particles can be
prepared either in a one-step synthesis using monomers of different reactivity ratios or in a
two-step synthesis by adding the monomers for the shell after the reaction of the core is
complete. Both steps are carried out at T > LCST. The latter method ensures a strict spatial
separation of core and shell and a well-defined surface between them [2]. A two-step synthesis
is possible because the particles of the first step act as seeds for the second polymerization
step where oligomers precipitating from the aqueous phase are connecting preferably to the
existing particle cores by chain transfer due to their relatively higher hydrophobicity [3][4]. The
total particle number remains constant during the addition of the shell [3].
Due to the soft nature of microgels, core and shell influence each other greatly [1]. Swelling or
collapsing of one network either causes or hinders the swelling or collapsing of the other. I.
Berndt et al. prepared core-shell microgels with PNIPAm in the core and poly(N-
isopropylmethacrylamide) (PNIPMAm) in the shell [5]. PNIPAm and PNIPMAm exhibit
LCSTs of 34 °C and 44 °C, respectively. The core thus collapses at 34 °C, while the shell
remains hydrophilic until a temperature of 44 °C is reached. In between, core and shell
counteract each other creating a field of stress within the particle. SANS measurements
revealed that the shell pulls the core outwards which is intensified with increasing shell
thickness.
Not only the synthetic protocol is important for the successful preparation of core-shell
particles, another important factor is the compatibility of the two monomers as well as the
interfacial tension between the two polymers [6].
In literature, a wide range of analysis techniques are used to gain information about the
internal structure of core-shell microgels. TEM and cryo-TEM measurements give an optical
conformation of the core-shell nature of particles and their morphology [4]. Staining with
phosphate tungsten acid [7], cadmium nitrate, potassium hexachloroplatinate [8] or uranyl
acetate [9] greatly enhances the contrast between core and shell. If nanoparticles are already
81
embedded in the microgel network, no further staining is necessary [10]. Cryo-TEM images
presented by M. Ballauff and L. Yu of poly(styrene) core and PNIPAm shell microgels
clearly distinguish between core and shell and show that the shell is slightly deformed due to
density fluctuations of the shell [4]. Their data confirm results for the same microgel system
that were obtained earlier in the same group by small angle X-ray scattering (SAXS) [11] and
that were predicted in theory by M. Shibayama [12]. These inhomogeneities are explained by
thermal fluctuations of the polymer chains that become stiff when crosslinked. Furthermore, it
was shown that the shell of those particles that were prepared in a two-step synthesis is not
continuously attached to the core and detaches slightly when the particles swell below the
VPTT [2].
Further methods for the characterization of core-shell microgels are SANS [13], NMR [14],
AFM [15][16] or titration [17][18].
Multi-responsive microgels react to more than one external stimulus, mostly pH and
temperature. T. Hu et al. synthesized core-shell particles with a PNIPAm-co-acrylic acid 2-
hydroxyethyl ester (HEA) core and a PNIPAm shell via free-radical RAFT living
polymerization [19]. They could show that core and shell exhibited different transition
temperatures. DLS measurements revealed that the copolymer core collapsed between 28 –
32 °C, while the collapse of PNIPAM shell is shifted to 32 – 25 °C. A dual temperature-
sensitive core-shell microgel was also prepared by A. Balaceanu et al [20]. They prepared
poly(vinylcaprolactam) (PVCL) core and poly(N-isopropylmethacrylamide) (PNIPMAm)
shell and reverse microgels. PVCL and PNIPMAm have volume transition temperatures at
32 °C and 45 °C, respectively. In PNIPMAm core/PVCL shell microgels, DLS measurements
giving the ratio of the core Rcore to the hydrodynamic radius RH indicate that the PVCL shell
collapses first upon an increase in temperature as can be seen from an increase in the ratio
Rcore/RH. A further increase in temperature leads to a collapse of the PNIPMAm core,
decreasing the Rcore/RH ratio.
So, far only few examples for ampholyte core-shell microgels can be found in literature. K. E.
Christodoulakis and M. Vamvakaki prepared 2-(diethyl-amino)ethyl methacrylate (DEA) or
tert-butyl methacrylate (t-BuMA) core and t-BuMA or DEA shell, respectively, in a two-step
emulsion polymerization [17]. Both core and shell could be protonated and deprotonated
separately. They compared the swelling/ de-swelling results to a microgel of the same
composition where t-BuMA and DEA are randomly distributed and found that in the latter the
basic and acid moieties are protonated/ deprotanated simultaneously. S. Schachschal et al.
synthesized ampholyte microgels with anionic core and a cationic shell based on VCL in a
82
one-step synthesis [14]. The core-shell structure arises from different reactivities and water
solubilities of the comonomers itaconic acid dimethyl ester (IADME) and vinylimidazole
(Vim). The ampholyte character of the particles was obtained after hydrolyzation of IADME
to itaconic acid (IA). Using proton high-resolution transverse magnetization relaxation under
magic angle sample spinning (MAS), the authors could show that IA is mostly located in the
core of the microgels. Both the VPTT and the IEP of the particles can be tuned by variation of
the IA and Vim content.
Based on the last work, core-shell microgels with NIPAm or VCL as main monomers in the
core or shell and IA and Vim as comonomers in the core or shell were prepared. Figure 7.1
shows the structure of the particles and the sample names used in this chapter.
In contrast to the work done by S. Schachschal, the core-shell microgels are prepared in a
two-step synthesis to ensure a strict spatial separation between core and shell. Particles with
basic moieties in the core and acid moieties in the shell as well as their reverse counterparts
were prepared to study the influence of distribution of ionizable groups on swelling
properties. The location of ionizable groups will further be important for uptake and release
studies in later chapters. Moreover, the location of VCL and NIPAM was varied. VCL is less
cytotoxic than NIPAm [21][22], its location in the shell might therefore enhance
biocompatibility of the whole core-shell particle. NIPAm-based microgels, however, have a
lower polydispersity and a sharper VPTT than PVCL-based microgels [22]. The
copolymerization of two temperature-sensitive monomers is a facile method for the
preparation of new smart materials.
Figure 7.1 Structure of core-shell microgels and the sample names used in this chapter
(c = core, s = shell).
83
7.2. Synthesis Procedure
Microgels were synthesized in a two-step precipitation polymerization in water. For the
synthesis of the core, appropriate amounts (see Table 7.1) of NIPAm/VCL, IA/Vim, and BIS
were dissolved in 150 mL water and heated up to 70 °C while purging with N2. After 1 h, the
initiator AMPA was added and the reaction carried out for 3 h under constant stirring. The
core-microgels were left in the reactor without cooling down or purification. For the addition
of the shell, the appropriate amounts (see Table 7.2) of VCL/NIPAm, Vim/IA, BIS and
AMPA were added to the aqueous solution without addition of further water. The reaction
was carried out for further 3 h under N2. After synthesis, microgel solutions were cleaned via
dialysis using a composite regenerated cellulose membrane from Millipore (NMWCO 30,000)
for 3 d against water and subsequently lyophilized for 3 d.
Table 7.1 Amounts of monomers used for synthesis of the core.
Monomer
in core
VCL/NIPAm / g IA / g Vim / g BIS / g AMPA / g
VCL 1.897 - 0.169 0.062 0.053
1.876 0.175 - 0.065 0.053
NIPAm 1.568 - 0.156 0.062 0.049
1.501 0.179 - 0.063 0.050
Table 7.2 Amounts of monomers used for synthesis of the shell (added to the core
dispersion).
Monomer
in shell
VCL/NIPAm / g IA / g Vim / g BIS / g AMPA / g Sample name
NIPAm 1.532 0.182 - 0.062 0.058 cVCL+/sNIPAm-
1.572 - 0.126 0.062 0.056 cVCL-/sNIPAm+
VCL 1.891 0.181 - 0.061 0.051 cNIPAm+/sVCL-
1.871 - 0.104 0.062 0.050 cNIPAm-/sVCL+
Further, core-shell microgels cVCL-/sNIPAm+ were prepared with an increasing shell
thickness. Core particles as prepared according to Table 7.1 were synthesized. The amounts of
the shell can be seen in Table 7.3. For the samples with a core:shell ratio of 1:2.5 and 1:5, the
monomers for the shell were dissolved in 190 mL and 375 mL of water (pre-heated to 70 °C),
respectively, and instantaneously added to the core dispersion.
84
Table 7.3 Amounts of monomers used for synthesis of the shell with increasing shell
thickness (added to the core dispersion).
Ratio
core:shell
NIPAm / g Vim /
g
BIS / g AMPA / g Sample name
1:1 1.532 0.104 0.062 0.050 cVCL-/sNIPAm+ (1:1)
1:2.5 3.802 0.250 0.149 0.132 cVCL-/sNIPAm+ (1:2.5)
1:5 7.581 0.526 0.312 0.260 cVCL-/sNIPAm+ (1:5)
85
7.3. Results and Discussion
The synthesis of core-shell microgels led to colloidally stable dispersions. Measurements of
the colloidal stability (see appendix) reveal that all particles have the same colloidal stability
regardless of the distribution of monomer and ionizable groups.
To ensure that the desired amounts of monomers are incorporated in the microgels, samples
were analyzed with FTIR (see Figure 7.2). The discussion of the monomer spectra and the
respective peak assignment can be found in chapter 6.
Figure 7.2 FTIR spectra for core-shell PVCL-co-Vim-co-IA microgels.
Table 7.4 Amounts of ionizable groups in core-shell microgels as obtained with FTIR.
Sample name VCL %
(1654 cm-1
)
Vim %
theor.
Vim % FTIR
(1284 cm-1
)
IA % theor. IA % FTIR
(1153 cm-1
)
cVCL+/sNIPAm- 80 10 10 10 8
cVCL-/sNIPAm+ 80 10 9 10 9
cNIPAm+/sVCL- 80 10 11 10 10
cNIPAm-/sVCL+ 80 10 9 10 11
In all cases, the amounts of IA and Vim as obtained from FTIR are close to the theoretical
values indicating that the incorporation of IA and Vim in the desired amounts was successful.
TEM images confirm the formation of core-shell microgels (see Figure 7.3). The particles
were stained with uranyl acetate, U(Ac)3, which binds to the carboxylic groups. Therefore,
regions with IA appear darker than regions with Vim. Figure 7.3a shows that IA is mostly
located in the core of the particles. A thin shell can be seen surrounding the particles. Figure
7.3c shows particles which are completely stained. IA is located in the shell and thus masks
the inner core.
86
Figure 7.3 TEM images of core-shell particles stained with U(Ac)3. (a) cVCL-/sNIPAm+ (b)
cVCL-/sNIPAm+, closer look; (c) cVCL+/sNIPAm-; (d) cVCL+/sNIPAm-, closer look.
Figure 7.4 (a) Hydrodynamic radius RH and (b) electrophoretic mobility EM of core
microgels at 20 °C with varying pH. The horizontal line at 0 µmcm/Vs is a visual help to
better identify the isoelectric point.
Particles were analyzed regarding their pH-sensitivity. First, the pure core particles were
examined (see Figure 7.4). Particles with basic moieties swell at low pH and collapse at high
pH from ~ 270 nm to ~ 100 nm (i.e. - 63 %). Particles with acid moieties swell at high pH due
to the deprotonation of the carboxylic groups and collapse at low pH from ~ 230 nm to ~
125 nm (i.e. - 46 %). The electrophoretic mobility reflects this behavior. While the EM of the
cores containing Vim decreases from 2.1 µmcm/Vs to 0 µmcm/Vs from pH 3 to pH 10, it
decreases less significantly for cores containing IA: from 0 µmcm/Vs at pH 3 to -
0.5 µmcm/Vs at pH 10. This asymmetric behavior is reflected in the FTIR data. Slightly less
IA is incorporated into the core polymer network than Vim. The swelling of both particle
a c
b d
a b
shell
core
87
types is very broad and constant indicating a homogenous distribution of the acidic and basic
moieties in the particles.
Pure IA and Vim have pKa and pKb values of 3.84/5.55 and 6.0, respectively. In the core
microgels, the swelling already starts at pH ~8 for Vim-core particles and at pH ~ 6 for IA-
core particles. The “premature” swelling between pH 6 and pH 8 for Vim-core microgels is
caused by the greater hydrophilicity of VCL compared to Vim due to the carbonyl group of
VCL. Hydrophilic VCL-co-Vim microgels will more readily swell in water than pure Vim. IA
on the other hand is already more hydrophilic than VCL due to its two carboxylic groups,
therefore, no “premature” swelling can be observed. This behavior is in accordance with
results obtained by M. Bradley et al [23].
Figure 7.4 also shows that there are no differences between PVCL- and PNIPAm-based core
microgels regarding their EM with varying pH. Looking at their RH, PNIPAm microgels are
slightly bigger in size than the respective PVCL microgel in the swollen state, while both
have the same size in collapsed state. Since FTIR data confirm that both samples have the
same amount of basic or acid moieties incorporated, the difference in swelling might be
attributed to different crosslinker densities. A higher crosslinker density in the core for VCL-
based microgels would suppress the swelling behavior.
Figure 7.5 shows the pH-responsive behavior of the respective core-shell particles that were
discussed in Figure 7.4. Figure 7.5b shows that core-shell particles show very different
behavior depending on the location of ionizable groups. For particles with Vim located in the
core with IA in the shell (black and blue curves), the particle size constantly decreases from
pH 3 to pH 10. For this system, the core is positively charged at low pH, while the shell is
negatively charged at high pH. Therefore, a V-shaped curve as observed for microgels with
statistical distribution of ionizable groups (see chapter 6) was expected. The collapse of the
shell at low pH exerts an inward pressure on the swollen core, still a strong increase in
particle size (compared to neutral state at pH ~ 6.2) can be observed indicating that the
collapse of the shell has only a small effect. The fact that particles constantly de-swell with
increasing pH indicates that at high pH, the negatively charged shell is not “strong” enough to
counteract the collapse of the neutral core. While the IA-shell supposably increases in size
due to swelling, the overall particle size decreases because the Vim-core pulls the shell
inwards (see Figure 7.6). Though the particle size decreases from low to high pH, two distinct
areas can be differentiated. A rapid decrease occurs from pH 3 to pH 6 and a slower decrease
from pH 6 to pH 10. This indicates that the swelling/ de-swelling transitions of the core and
the shell are independent from each other.
88
Figure 7.5 (a) Hydrodynamic radius RH and (b) electrophoretic mobility EM at 20 °C with
varying pH of core-shell microgels. The horizontal line at 0 µmcm/Vs is a visual help to
better identify the isoelectric point.
Figure 7.6 Schematic illustration of pH-dependent behavior of core-shell particles. (a) core:
Vim, shell: IA; (b) core: IA, shell: Vim. Arrows indicate direction of swelling (green) and de-
swelling (red). X and y denote the particle size at low (x) and high (y) pH.
A different situation arises if IA is located in the core and Vim in the shell (red and magenta
curves). Additionally, a difference can be seen between NIPAm and VCL. The system
cNIPAm-/sVCL+ shows a slight V-shaped curve indicating that particles swell at high and
low pH. The system cVCL-/sNIPAm+, in contrast, reaches a plateau in size at pH > 6.5. For
both systems, results indicate that the positively charged shell at low pH is able to
“counteract” the collapse of the neutral core by exerting a strong outward pressure on the
latter. This is in contrast to the “weak” IA-shell discussed above. The reason for this is likely
to be the nature of the core. The Vim-cores discussed above de-swells highly at high pH and
subsequently pulls the shell inwards. The IA-core, on the other hand, does not de-swell in the
same extent as the Vim-core when neutralized due to its higher hydrophilicity. This could
already be seen in Figure 7.4. Neutral IA-cores (i.e. at pH 3) have an RH of 125 nm, while
neutral Vim-cores (i.e. at pH 10) have a lower RH of 102 nm.
The swelling degrees q can be seen in Table 7.5. They are calculated as follows:
a b
89
𝑞𝑎𝑐𝑖𝑑𝑖𝑐 = (𝑅𝐻,𝑝𝐻3)³
(𝑅𝐻,𝑝𝐻6)³ (1)
𝑞𝑏𝑎𝑠𝑖𝑐 = (𝑅𝐻,𝑝𝐻10)³
(𝑅𝐻,𝑝𝐻6)³ (2)
The results show that particles with PNIPAm in the shell swell in a higher extent than
particles with PVCL in the shell. This is in accordance with the results obtained for the core-
particles where NIPAm-based particles swelled more than VCL-based microgels (see Figure
7.4). This is attributed to the different reaction parameters between BIS with VCL and
NIPAm, respectively, leading to different incorporation of the crosslinker in the polymer
network and thus a different swelling behavior. The table also emphasizes the difference
between particles having Vim or IA in the shell. Particles with IA in the shell have a ratio
below 0 indicating that the particles become smaller in volume.
Particles with Vim in the core and IA in the shell have an IEP of pH ~ 5.5, while particles
with a reverse distribution of ionizable groups have an IEP of pH 6.5. This discrepancy can be
explained by the fact that electrophoresis measures the amount of charges on the particle
surface. Particles with IA in the shell have therefore an IEP hat is closer to the pKa of IA than
to the pKa of Vim. Accordingly, particles with Vim in the shell have a pKa closer that to of
Vim than to the one of IA. For all samples, regardless of the charge location, the absolute
value for the electrophoretic mobility is below 2 µmcm/Vs. M. Semmler et al. state that the
typical value for hard spheres is ~ 4 µmcm/Vs [24]. This indicates that the particles are swollen
therefore showing no distinct, but a diffuse surface typical for microgels.
Table 7.5 Swelling ratio q of core-shell microgels at pH 3 and 10, respectively, compared to
the minimum in size at pH 6.2 (~ IEP). Swelling ratios were calculated as follows:
q1 = RH(pH3)³/RH(pH6.2)³ and q2 = RH(pH10)³/RH(pH6.2)³. The table also shows the swelling ratios for
selected microgels with statistical distribution of ionizable groups.
Sample qacidic qbasic
cNIPAm-/sVCL+ 3.70 1.57
cNIPAm+/sVCL- 3.99 0.44
cVCL-/sNIPAm+ 4.05 1.06
cVCL+/sNIPAm- 4.68 -0.40
VCL:Vim:IA = 80:10:10 statistical 5.50 6.91
VCL:Vim:IA = 80:0:20, statistical - 1.75
90
Comparing the swelling degrees at low and high pH of core-shell particles with particles with
a statistical distribution of ionizable groups (see Chapter 6 and Table 7.5), it can be observed
that microgels with a core-shell distribution swell less at pH 3 and considerably less at pH 10
than microgels with a statistical distribution of ionizable groups. According to T. Hoare and
R. Pelton, microgels with a homogeneous distribution throughout the particle swell in a less
extent [25]. The polyion concentration is lower due to a greater volume that is available. In
core-shell microgels, in contrast, charges are localized in a smaller volume thus repelling each
other in a higher degree resulting in a more pronounced swelling.
This, however, is not observed for the above discussed system. Assuming that the core has a
higher influence on the swelling/ de-swelling behavior, the swelling degrees of microgels with
a statistical distribution of ionizable groups is best compared to core-shell particles with a
swollen core and a neutral shell. For particles with Vim in the core, q is approximately in the
same range as for particles with statistical distribution of ionizable groups. The smaller value
can be explained by the stiff shell of neutral IA suppressing further swelling. For particles
with IA in the core, however, q is considerably lower. A comparison with the sample
VCL:Vim:IA = 80:0:20 with statistical distribution of ionizable groups (it should be noted,
however, that this sample has a twofold amount of IA incorporated), shows that this value is
comparable with the value for core-shell particles containing IA in the core.
Figure 7.7 Hydrodynamic radius RH as a function of temperature for (a) cVCL+/sNIPAm-
and the respective core, measured at pH 3; (b) cVCL-/sNIPAm+ and the respective core,
measured at pH 10. Pure PVCL microgels are given as a reference.
In the following, the hydrodynamic radius of the core-shell microgels is compared to the
respective core-microgels. Measurements were only conducted for particles with PVCL in the
91
core and PNIPAm in the shell and at a pH where the core is charged, i.e. at pH 3 for cores
containing Vim and at pH 10 for cores containing IA. Measurements at the isoelectric point
Table 7.6 Volume phase transition temperatures for core and core-shell microgels.
Sample Measured at pH VPTT Core / °C VPTT Core-Shell / °C
cVCL+/sNIPAm- 3 31.8 33.8
cVCL-/sNIPAm+ 10 31.8 33.0
cNIPAm+/sVCL- 3 31.8 34.4
cNIPAm-/sVCL+ 10 31.7 33.5
led to aggregation at higher temperatures due to the lack of charges. All samples show a broad
transition temperature compared to pure PVCL microgels (see Figure 7.7). Pure PVCL
microgels have a transition temperature of 31 °C. The incorporation of basic or acid moieties
into the core particles does not change the transition temperature significantly (see Table 7.6).
Also, at all pH, core and core-shell microgels are larger in size than pure PVCL microgels due
to charged moieties repelling each other.
In the following, the thickness of the shell in the system cVCL-/cNIPAm+ was increased to
study its effect on the particle’s properties. As shown in Figure 7.4, the size of this sample
decreases until a pH of 6.5, above this pH, the particle size remains constant. To study
whether the swelling behavior at high pH can be influenced, the same particles with an
increasing shell thickness were prepared. The ratio core:shell was increased from 1:1 (as
discussed above) to 1:2.5 and 1:5.
First, an increasing amount of 1-vinylimidazole was confirmed with FTIR (see appendix). It
can be seen that the peak corresponding to Vim at 1284 cm-1 increases with an increasing
shell thickness.
Figure 7.8 TEM images of core-shell microgels (cVCL-/sNIPAm+) with increasing shell
thickness. (a) Ratio core:shell = 1:1, (b) 1:2.5; (c) 1:5. Images with single particles are shown
to better visualize the increase in shell thickness.
TEM images confirm that the particles increase in size (see Figure 7.8). Dark spots in the
corona come from “entangled” U3+ aggregates in the shell polymer network. The thickness of
the shell increases from 24 nm to 50 nm and 71 nm.
c a b
92
Figure 7.9 RH and EM at 20 °C with varying pH of core-shell microgels with increasing shell
thickness.
Figure 7.9 gives the hydrodynamic radius and the electrophoretic mobility for samples with
increasing shell thickness as a function of pH. RH remains constant for a pH > 6.5 for samples
with a ratio core:shell = 1:1 as already discussed above. For a ratio core:shell = 1:2.5, the
particle size slightly increases for pH > 7.5. For a ratio core:shell = 1:5, the particle size
increases in a more pronounced way for a pH > 7. This shows that the inward force exerted by
the collapse of the neutral core on the negatively charged, swollen shell is diminished by an
increase in shell thickness. Still, the degree of swelling at high pH is considerably lower than
for samples with a statistical distribution of ionizable groups. The electrophoretic mobility is
shifted to higher pH for an increasing ratio core:shell while the absolute value increases at low
pH. This can be explained by the higher number of positively charged groups in the particle
shell.
The size of the particles was also measured varying the solution temperature. As for the core-
shell microgels discussed above, no significant changes in the VPTT can be detected. The
graphs are shown in the appendix.
93
7.4. Summary
This chapter discussed the preparation and analysis of ampholyte core-shell microgels. N-
vinylcaprolactam (VCL) and N-ispropylacrylamide (NIPAm) were used as main monomers in
either the core or the shell. Itaconic acid (IA) and 1-vinylimidazole (Vim) were used as
comonomers. Particles were prepared in a two-step synthesis to ensure a strict spatial
separation between the core and the shell. The core-shell structure was visualized with TEM
after selectively staining of the carboxylic groups with U(Ac)3. The successful incorporation
of desired amounts of IA and Vim were confirmed with FTIR spectroscopy. The pH-
sensitivity of the cores was analyzed with regard to the particles’ size and their electrophoretic
mobility. Particles containing IA swell at high pH due to repulsion of negatively charged
groups. Particles containing Vim, in contrast, swell at low pH due to the repulsion of
positively charged groups. Core-shell particles, on the other hand, do not show a V-shaped
curve for the particle size as expected for ampholyte microgels. It was shown that particles
swell greatly at low pH, while the particle size either decreases continuously with increasing
pH (for particles containing IA in the shell) or is constant in size above a pH ~ 7. Also the
degree of swelling greatly differs from particles containing a statistical distribution of
ionizable groups. It could be observed that a neutral shell hinders the complete swelling of the
core. Temperature-dependent size measurements showed that particles have a broadened
transition while particles are larger in size compared to neutral PVCL microgels due to the
repulsion of like charges. An increase of the shell thickness led to an improved degree of
swelling at high pH. Finally, the thermal decomposition of core-shell particles was studied
with TGA. A two-step decomposition was detected caused by different decomposition
temperatures of PVCL and PNIPAm.
94
7.5. Literature
[1] W. Richtering, A. Pich, Soft Matter 2012, 8, 11423.
[2] J. J. Crassous, M. Ballauff, M. Drechsler, U. Schmidt, Y. Talmon, Langmuir 2006, 22,
2403–2406.
[3] D. Gan, L. A. Lyon, Prog. Colloid Polym. Sci. 2006, 133, 1–8.
[4] M. Ballauff, Y. Lu, Polymer (Guildf). 2007, 48, 1815–1823.
[5] I. Berndt, J. S. Pedersen, P. Lindner, W. Richtering, Langmuir 2006, 22, 459–468.
[6] P. Marion, Macromolecules 1997, 30, 123.
[7] Z. Zhou, J. V. Hollingsworth, S. Hong, G. Wei, Y. Shi, X. Lu, H. Cheng, C. C. Han,
Soft Matter 2014, 10, 6286-6293.
[8] S. E. Kudaibergenov, N. Nuraje, V. V. Khutoryanskiy, Soft Matter 2012, 8, 9302.
[9] C. Jones, L. Lyon, Macromolecules 2000, 33, 8301–8306.
[10] Y. Lu, S. Proch, M. Schrinner, M. Drechsler, R. Kempe, M. Ballauff, J. Mater. Chem.
2009, 19, 3955.
[11] N. Dingenouts, C. Norhausen, M. Ballauff, Macromolecules 1998, 31, 8912–8917.
[12] M. Shibayama, Macromol. Chem. Phys. 1998, 199, 1–30.
[13] K. Kratz, T. Hellweg, W. Eimer, Polymer (Guildf). 2001, 42, 6631–6639.
[14] S. Schachschal, A. Balaceanu, C. Melian, D. E. Demco, T. Eckert, W. Richtering, A.
Pich, Macromolecules 2010, 43, 4331–4339.
[15] G. Li, P. D. Pandya, S. S. Seo, Int. J. Polym. Anal. Charact. 2009, 14, 351–363.
[16] A. Burmistrova, M. Richter, M. Eisele, C. Üzüm, R. von Klitzing, Polymers (Basel).
2011, 3, 1575–1590.
[17] K. E. Christodoulakis, M. Vamvakaki, Langmuir 2010, 26, 639–47.
[18] R. Saito, K. Ishizu, T. Fukutomi, J. Appl. Polym. Sci. 1991, 43, 1103–1109.
[19] T. Hu, Y. You, C. Pan, C. Wu, J. Phys. Chem. B 2002, 106, 6659–6662.
[20] A. Balaceanu, Y. Verkh, D. E. Demco, M. Möller, A. Pich, Macromolecules 2013, 46,
4882–4891.
[21] L. Hou, P. Wu, Soft Matter 2015, 11, 2771–2781.
[22] J. Ramos, A. Imaz, J. Forcada, Polym. Chem. 2012, 3, 852.
[23] M. Bradley, B. Vincent, G. Burnett, Langmuir 2007, 23, 9237–9241.
[24] M. Semmler, E. K. Mann, J. Ricka, M. Borkovec, Langmuir 1998, 14, 5127–5132.
[25] T. Hoare, R. Pelton, Langmuir 2004, 20, 2123–33.
95
8 Janus-like Microgels
8.1. Introduction
Janus is the name of the old Roman god of end and beginning, of entrance and exit, of old and
new. Not only is the month “January” named after him (January being the month linking the
old and the new year), but due to his two faces (that he used to look in both directions
simultaneously) his name is used in modern-day science to term things or phenomena that
have two “faces” or two aspects. The term is mostly used in chemistry. Here, Janus particles
are colloids with clearly distinguished hemispheres with different chemical or physical
properties such as form [1], material, chemical functionalization, or polarity [2].
Janus particles exhibit many different shapes and forms (see Figure 8.1).
Figure 8.1 Different forms of Janus particles: a Bicompartmental; b Dumbbell; c Half-
raspberry; d Acorn; e Snowman.
In nature, dissymmetry in a molecule offers specific properties. The best examples are
proteins that have an asymmetric distribution of ionizable groups and thus a large dipole
moment. For Janus particles, the dissymmetry offers an additional level of freedom: the
particles can distinguish between left and right, between up and down [1]. This behavior can be
utilized for arranging particles or for the self-assembly of complex structures. Tuning the
physical or chemical properties of each hemisphere, new materials with dual characteristics
are possible. Simulations predict that these highly complex structures are different from
traditional materials [3]. While in “old” materials, atoms and molecules are the building
blocks, anisotropic particles open the pathway to novel supermolecules with exactly the
desired features.
Different synthesis routes have developed addressing the desired properties of the respective
particles. The most common three approaches are discussed in the following.
Immobilization on a substrate. In this process, the particles are deposited on a surface and are
immobilized. Then, “functionalization” is induced from one side leading to an altered upper
hemisphere of the particles. The release of the particles from the surface results in Janus
particles. M. Bradley and J. Rowe used a similar technique to prepare Janus microgels [4].
Instead of a flat surface, they used negatively charged poly(2-vinylpyridine-co-styrene) latex
particles as a surface to mask one side of the positively charged poly(2-vinylpyridine)
microgels that aggregated onto the much larger latex particles. The exposed side of the
96
microgels was functionalized with PNIPAm. They obtained pH- and temperature-sensitive
microgel particles after the dissolution of the latex particles by a reduction of the pH.
Modification of colloids at liquid-liquid interfaces. The functionalization of the particles takes
place at the liquid-liquid interface. H. Kawaguchi et al. who were the first to prepare Janus-
like microgels. PNIPAm microgels were copolymerized with acrylic acid (AAc) and amino
groups were introduced via a carbodiimide coupling reaction on the side of the particles in
contact with water.[5]. To prevent the microgels from “wobbling” at the interface, Y. Umeda
et al. stabilized their PNIPAm-co-AAc microgels at the interface by cooling down the
emulsion to 4 °C [6].
Figure 8.2 Different swelling behavior of a microgel particle in different solvents.
However, the synthesis of Janus microgels via emulsion polymerization entails certain
problems, the main one being that microgels swell differently in different solvents, thus
leading to an inhomogeneous surface functionalization (see Figure 8.2).
Microfluidics The relatively new method of microfluidics mixes two non-miscible phases
with a photoinitiator to form droplets due to surface tension effects in a co-flowing aqueous
phase at the Y-junction of a microchannel. A UV source ensures the polymerization of the
desired particle morphology. Shape, size, and functionality are determined by microfluidic
set-up, i.e. the geometry and the flowing rate of the media. This approach, however, has two
main drawbacks: firstly, it only allows producing a restricted amount of particles at a time,
secondly, usually only spherical shapes or deformed spheres can be realized [7]. Microgels in
micrometer size using microfluidics were prepared by S. Seiffert et al [8]. They mixed pre-
modified and unmodified precursor PNIPAm polymers and crosslinked them via exposure to
UV light. Functionalization with red and green fluorescence labels enabled the verification of
the Janus-like structure with confocal scanning laser fluorescence microscopy.
Further, less frequent used techniques are self-assembly [9][10], phase separation [11][12], and
controlled surface modification [13][14].
Janus particles find applications in microfluidics, stabilization of emulsions, drug delivery,
electronics, and catalysis [2].
In this chapter, ampholyte Janus particles based on N-isopropylacrylamide (NIPAm) were
synthesized using a novel technique. Based on phase separation of the monomers in the first
moments after initiation, the precursor particles or microgel nuclei are mixed at the moment
97
when two different kinds of precursor particles are not able to completely mix with each other
(see Figure 8.3).
Figure 8.3 Synthesis procedure for Janus particles. In Reaction 1 microgel A reacts with
functional group A’, in reaction 2 microgel B reacts with functional group B’. tmix is the time
at which the mixing of dispersions lead to the formation of Janus particles (b). (a) The
dispersion are mixed before tmix leading to one particle with homogenous functional group
distribution. (c) The dispersions are mixed after tmix leading to separate particles A and B
bearing separate functional groups A’ and B’.
a b c
98
8.2. Synthesis Procedure
Calorimetric measurements were performed in a reaction calorimeter RCle from Mettler
Toledo with a 500 mL 3-wall AP01-0.5-RTCal reactor equipped with Hastelloy® stirrer, a
baffle, and a TurbidoTM turbidity probe from Solvias. The measurements were done in
isothermal mode, in which the desired reaction temperature (Tr) is set at a constant value and
the jacket temperature (Tj) changes automatically to maintain Tr at the desired value. The
amounts of monomer used can be seen in Table 8.1. 300 mL of water was heated up to 70 °C
under nitrogen atmosphere and constant stirring at 200 rpm. After 30 min, the respective
amounts of monomers were added. Simultaneously, a probe recorded the heat generated. If a
stable base line in the produced heat was reached (minimum waiting time was 30 min), the
initiator AMPA was added. The production of heat and the turbidity were recorded in-situ
during the reaction using the iControl RCLeTM 5.0 software. The reaction was allowed to
continue for 3 h and subsequently cooled down to room temperature. For the determination of
the net reaction time the end of the reaction was defined to be when the heat flow was back to
a value close to zero and constant. The fractional conversion can be obtained integrating the
reaction heat curve.
Size kinetics were measured in a 100 mL reactor equipped with a Nano-Flex probe (Particle
Matrix company, Germany) measuring in-situ dynamic light scattering. The aqueous
monomer solution was heated up to 70 °C and degassed under vigorous stirring to avoid
bubbles on the probe surface. After equilibrating under nitrogen atmosphere at 200 rpm for
30 min, the initiator was added. The stirring was stopped after 5 s to ensure accurate DLS
measurements in 33 s intervals. The reaction was allowed to continue for 99 measurements.
The data was analyzed using the Microtrac Felx (v. 11.0.0.4) software providing the
hydrodynamic radius and the polydispersity index.
For the synthesis of Janus particles, two separate reactions were carried out in two different
reactors (see Table 8.1).
Synthesis 1. NIPAm, Vim, and BIS were added in 100 mL H2O. After heating up to 70 °C, the
solution was stirred at 200 rpm under nitrogen atmosphere for 1 h.
Synthesis 2. NIPAm, IA, and BIS were added in 100 mL H2O. After heating up to 70 °C, the
solution was stirred at 200 rpm under nitrogen atmosphere for 1 h.
99
Table 8.1 Amounts of monomers used for the reaction.
Chemical Synthesis 1 Synthesis 2
Amount / g
NIPAm 1.003 1.004
Vim 0.158 -
IA - 0.293
BIS 0.049 0.049
AMPA 0.038 0.037
In both reactors, the initiator AMPA was added simultaneously. 0.9 mL of solution was
removed from each solution after 1 min, 2 min, 3 min, 4 min, 5 min, 6 min, 8 min, 10 min,
12 min, 14 min, 16 min, and 18 min. The two solutions were immediately combined in one
flask and allowed to stir for 3 h at 70 °C at 200 rpm.
To visualize the success of the reaction, transmission electron microscopy (TEM) was
performed at a Zeiss Libra TM 120 (Carl Zeiss, Oberkochen, Germany). 20 µL of each
sample was stained with 5 µL of uranyl acetate U(Ac)3 and one drop was put on a carbon
coated copper grid. Samples were then dried overnight.
100
8.3. Results and Discussion
Janus-like particles based on PNIPAm with itaconic acid (IA) and 1-vinylimidazole (Vim) as
comonomers are discussed in this chapter. An anisotropic distribution of acidic and basic
moieties is aimed at. In contrast to core-shell particles, where one kind of ionizable group is
located in the core and the other in the center of the microgel, Janus-like particles have their
ionizable groups on opposite “sites”.
The synthesis of Janus-like microgels is based on phase separation phase separation of the
precursor particles. The approach via substrate immobilization and emulsion polymerization
lead to particles with only surface-modified functionalization. The aim in this work, however,
is the synthesis of ampholyte microgels with ionizable groups located throughout the whole
microgel.
First preliminary experiments with either VCL or NIPAm as main monomer and the
subsequent analysis with TEM showed that the synthesis process works better for NIPAm
microgels (TEM images for NIPAm will be discussed later in this chapter). Therefore, the
following experiments were solely performed for NIPAm microgels.
Before the synthesis of the actual Janus-like particles, experiments to understand the
polymerization kinetics and particle formation were performed. Therefore, separate reactions
for NIPAM-IA and NIPAm-Vim as well as pure NIPAm microgels were carried out in a
calorimeter and the change of heat flow and particle size were measured starting with the
addition of the initiator (start of the reaction). Syntheses were performed at different pH
according to the type of comonomer (pH 3 for Vim and pH 8 for IA).
First, the change in reaction heat was measured. It was shown that it was not possible to
copolymerize NIPAm with IA in the calorimeter, contrary to the reactions that were
successfully carried out in the double-walled glass reactors. Instead of a stable microgel
dispersion, a sticky hydrogel-like material was obtained. The reason for this is unknown,
though a reaction between itaconic acid and the metal probes is likely. Both the curves for
pure PNIPAm and PNIPAm-co-Vim microgel look similar, though the reaction time is
slightly longer for PNIPAm-co-Vim microgels. Both reactions are finished after ~ 30 min.
It was, however, possible to measure in-situ time-dependent DLS for all three samples. Figure
8.4 shows that the reaction for pure PNIPAm microgels is the fastest and that particles have
reached a hydrodynamic radius of ~ 140 nm after already 4 min. PNIPAm-co-Vim microgels
need a slightly longer reaction time of ~ 8 min at which point they have obtained a
hydrodynamic radius of ~ 250 nm. At first glance, these results are contrary to the results
101
Figure 8.4 (a) Calorimetric measurements for pure PNIPAm microgels (black) and PNIPAm-
co-Vim microgels (red). The numbers give the reaction heat that was created. (b) In-situ time-
dependent DLS measurements for pure PNIPAm (black), PNIPAm-co-IA (blue), and
PNIPAm-co-Vim (red) microgels.
obtained with calorimetry which showed that both reactions continue until ~ 30 min. This
shows, however, that the particles obtain their final size after a very short time and that further
addition of monomers (generating more heat) does not contribute to an increase in particle
size. Microgels containing PNIPAm and IA need about 40 min to reach a hydrodynamic
radius of ~ 500 nm.
The synthesis of Janus particles was realized by mixing PNIPAm-co-IA and PNIPAm-co-Vim
microgel dispersions after a certain amount of time. The results in Figure 8.4 suggest that the
point of mixing time is a crucial factor that influences the form and properties of the resulting
microgel particles. Both reactions (1. NIPAm + IA; 2. NIPAm + Vim) were done
simultaneously in two different reactors. After the addition of the initiator, samples were
taken from both reactions and mixed immediately at various reaction times. The reactions
were allowed to continue for 4 h in a shaker at 200 rpm and 70 °C. Depending on the time of
mixture, colorless (early mixing time) or milky (late mixing time) dispersions were obtained.
Afterwards, samples were stained with U(Ac)3 and analyzed with transmission electron
microscopy (TEM) to confirm the success of the formation of Janus particles. U(Ac)3 binds to
the carboxylic groups of the itaconic acid and gives a clear distinction between parts with
itaconic acid and parts with 1-vinylimidazole.
The reaction times before mixing were changed:
a. NIPAm-co-IA and NIPAm-co-Vim were initiated simultaneously. 1 mL of
each sample was taken after 1, 2, 3, 4, and 5 min from each reactor and mixed
immediately.
a b
102
b. NIPAm-co-Vim was initiated 2.5 min after the initiation of NIPAm-co-IA.
1 mL of each sample was taken after 0.5, 1, 2, 3, and 4 after the initiation of
NIPAm-co-Vim min from each reactor and mixed immediately.
c. NIPAm-co-Vim was initiated 5 min after the initiation of NIPAm-co-IA. 1 mL
of each sample was taken after 0.5, 1, 2, 3, and 4 min after the initiation of
NIPAm-co-Vim from each reactor and mixed immediately.
Figure 8.5 shows TEM images for a batch of samples that were started simultaneously
through the addition of the initiator AMPA (case a). PNIPAm-co-Vim and PNIPAm-co-IA
solution were mixed after 1, 2, 3, 4, and 5 min. TEM images reveal that different mixing
times lead to the formation of microgels with different architectures. When mixed after 1 min,
particles with a homogenous distribution of both IA and Vim are obtained. Mixing after 2 min
leads to particles with one side clearly darker than the other. This suggests that the stained
carboxylic group of IA is located only on one side of the microgel leading to the formation of
Janus-like particles. Later mixing times lead to the formation of two separate particles that
exist side by side. Large aggregates are visible for PNIPAm-IA microgels (see Figure 8.5f).
Figure 8.5 TEM images of PNIPAm-co-IA-co-Vim microgels obtained following route a.
PNIPAm-co-IA and PNIPAm-co-Vim solutions were mixed after (a) 1 min; (b) 2 min; (c) 2
min – zoom-in; (d) 3 min; (e) 4 min; (f) 5 min.
The synthesis was repeated with different initiation times of the two reactions. As shown in
Figure 8.4b, the reaction NIPAm-co-IA is slower than the reaction NIPAm-co-Vim.
Therefore, the reaction NIPAm-co-Vim was started 2.5 min after the initiation of NIPAm-co-
IA (case b). Samples were mixed after 0.5, 1, 2, 3, and 4 min of the initiation of NIPAm-co-
Vim.
a b
d e f
c
103
Figure 8.6 (a+b) TEM images of PNIPAm-co-IA-co-Vim microgels obtained following route
b. Mixing time was 0.5 min. (c) NIPAm-co-Vim microgels; (d) NIPAm-co-IA microgels as
reference.
TEM images reveal that only the mixture of 0.5 min produced Janus-like particles (see Figure
8.6). TEM images for different mixing times are not shown as they resemble images shown in
Figure 8.5 and are of no further interest.
The synthesis following route c (NIPAm-co-Vim initiated 5 min after initiation of NIPAm-co-
IA) did not result in the formation of Janus-like particles independent of the mixing time.
(It was also tried to vary the amount of volume of NIPAm-co-Vim and NIPAm-co-IA,
respectively. Therefore, separate solutions were mixed after simultaneous initiation, but
mixed in different ratios, i.e. NIPAm-co-IA:NIPAm-co-Vim = 2:1 and 1:2. However, in
neither case TEM images revealed the formation of Janus-like particles.)
Particles obtained via the synthesis route b were further characterized with temperature-
dependent DLS to obtain the particle’s hydrodynamic radius (see Figure 8.7a). It can be seen
that the VPTT is ~ 29 °C. In comparison to ampholyte microgels with statistical or core-shell
distribution of ionizable groups, the VPTT is only slightly lower.
Furthermore, the particles’ pH sensitivity was examined. Figure 8.7b shows that particles are
swollen at high (~ 550 nm) and low (~ 650 nm) pH, while being in a collapsed state at pH 5
(430 nm). This and the fact that the absolute value of the electrophoretic mobility is higher at
pH 3 than at pH 10, indicates that slightly more Vim is incorporated than IA.
a b
c d
104
Figure 8.7 (a) RH with varying temperature at pH 6; (b) Electrophoretic mobility EM and RH
with varying pH at 25 °C. Janus-like particles were obtained via synthesis route b.
In the following, the pH-sensitivity of the Janus-like microgels is compared to the pH-
sensitivity of particles with statistical and core-shell distribution of ionizable groups (see
chapters 6 and 7). Figure 8.8 therefore shows the electrophoretic mobility and the
hydrodynamic radius as a function of pH for microgels with various distributions of ionizable
groups.
Figure 8.8 RH and EM as a function of pH for microgels with various distribution if ionizable
groups: Janus-like (black), core-shell (red), and statistical (blue). All samples have a monomer
ratio of VCL:Vim: IA = 80:10:10. Measurements were conducted at T = 25 °C. The
horizontal line (EM = 0 µmcm/Vs) helps to identify the isoelectric point.
All samples have a similar isoelectric point (IEP) of pH ~ 5 - 6. At lower pH, the microgels
are positively charged, while they are negatively charged at higher pH. A difference can be
seen between Janus-like and statistical on the one side and core-shell particles on the other
side. The latter has a more negative EM than the other two because here itaconic acid is
exclusively located in the shell.
a b
105
Furthermore, the hydrodynamic radius RH is within the same size range, i.e. between 100 and
300 nm. Differences can be seen in the swelling behavior. Samples with a statistical
distribution of ionizable groups show a V-shaped behavior, i.e. they are swollen at low and
high pH, while the particles collapse at the IEP. Since the sample has a symmetric ratio
between Vim and IA, the particles swell in the same extent at low and high pH. Core-shell
microgels, on the other hand, swell only at low pH, while a constant decrease in size can be
observed with increasing pH. The reasons are explained in chapter 7. The pH-dependency of
Janus-like microgels is V-shaped for a pH range of 3 – 6.5, while RH remains constant at pH >
6.5. This comparison shows that the distribution of ionizable groups within the microgel
particles is of great importance and highly influences the pH behavior.
Similarly, the colloidal stability of all three samples was studied and compared.
Measurements were conducted at pH 8 because the greatest differences between the above
discussed samples is expected to be seen at high pH for the following reason: All samples are
negatively charged at high pH. The core-shell sample has itaconic acid groups located
exclusively in the shell, Janus-like particles exclusively on one “side”, while particles with a
statistical distribution have itaconic acid located throughout the polymer network. Colloidal
stability obtained via electrostatic interactions is therefore likely to differ considerably. These
assumptions are confirmed in Figure 8.9. Core-shell microgels have the lowest sedimentation
velocity, i.e. they have the highest colloidal stability, while microgels with a statistical
distribution of ionizable groups have the highest sedimentation velocity. Note, however, that
overall, the sedimentation velocities are relatively low for all samples regardless of their
architecture indicating their good colloidal stability.
Figure 8.9 Sedimentation velocities for microgels with various distribution of ionizable
groups. Samples have a monomer ratio of VCL:Vim: IA = 80:10:10.
106
8.4. Summary
This chapter showed that it is possible to synthesize Janus-like particles by mixing two
separate microgels solutions shortly after initiation of each solution. For this, PNIPAm-co-IA
and PNIPAm-co-Vim microgels were synthesized in different reactors, initiated by the
addition of AMPA and subsequently mixed at a certain mixing time. TEM images confirmed
that Janus-like particles were formed at the right mixing time. The formation was possible if
both reactions were initiated simultaneously or shortly after each other. For simultaneous
initiation, Janus-like particles were obtained after 2 min after the addition of AMPA. In a
second experiment, PNIPAm-co-IA was initiated 2.5 min before PNIPAm-co-IA. Mixing
after 3 min after the initiation of PNIPAm-co-IA also led to the formation of Janus-particles.
Since it was shown that this synthesis procedure leads to a spatial separation of acidic and
basic moieties, further studies are needed so that various mixing times lead to variable ratios
between both functional groups resulting in structures such as bicompartmental, dumbbell or
snowman.
8.5. Literature
[1] A. Perro, S. Reculusa, S. Ravaine, E. Bourgeat-Lami, E. Duguet, J. Mater. Chem.
2005, 15, 3745.
[2] A. Walther, A. H. E. Müller, Soft Matter 2008, 4, 663.
[3] S. C. Glotzer, Science 2004, 306, 419–20.
[4] M. Bradley, J. Rowe, Soft Matter 2009, 5, 3114-3119.
[5] D. Suzuki, S. Tsuji, H. Kawaguchi, J. Am. Chem. Soc. 2007, 129, 8088–9.
[6] Y. Umeda, T. Kobayashi, T. Hirai, D. Suzuki, Colloid Polym. Sci. 2010, 289, 729–737.
[7] D. Dendukuri, D. C. Pregibon, J. Collins, T. A. Hatton, P. S. Doyle, Nat. Mater. 2006,
5, 365–9.
[8] S. Seiffert, M. B. Romanowsky, D. A. Weitz, Langmuir 2010, 26, 14842-14847.
[9] Y. Yin, Y. Lu, Y. Xia, J. Am. Chem. Soc. 2001, 123, 771–772.
[10] Y. Yin, Y. Lu, B. Gates, Y. Xia, J. Am. Chem. Soc. 2001, 123, 8718–8729.
[11] P. Mulvaney, M. Giersig, T. Ung, L. M. Liz-Marzán, Adv. Mater. 1997, 9, 570–575.
[12] H. Gu, R. Zheng, X. Zhang, B. Xu, J. Am. Chem. Soc. 2004, 126, 5664–5.
[13] S. Reculusa, C. Poncet-Legrand, S. Ravaine, C. Mingotaud, E. Duguet, E. Bourgeat-
Lami, Chem. Mater. 2002, 14, 2354–2359.
[14] H. Yu, M. Chen, P. M. Rice, S. X. Wang, R. L. White, S. Sun, Nano Lett. 2005, 5,
379–82.
107
9 Interactions with Proteins
9.1 Introduction
The low solubility of many drugs in water is a major problem modern, intelligent polymer
systems should address when utilized as drug carriers. Microgels can be employed due to the
great potential of precisely designing their properties. Functionalization allows tuning the
interactions between microgel and drug that determines the uptake and release of the latter.
Microgels can be tailored to possess crucial characteristics an intelligent drug carriers must
fulfill: hydrophilicity/ hydrophobicity, sensitivity to environmental changes, biocompatibility,
degradability, and specific size. Using microgels not only offers the possibility of using a
smart polymer network for controlled delivery, they also protect proteins against
environmental hazards such as drastic pH or changes in redox potential by providing a
hydrophilic environment that preserves the protein’s activity [1][2]. Additionally, microgels are
capable to store large amounts of proteins in a relatively small volume [3].
The interactions between microgels and proteins are complex. Proteins can undergo
conformational changes, while microgels can swell or collapse upon the sorption of proteins.
The uptake/ release of proteins can be controlled by the following: (a) diffusion; (b) osmotic
pressure; (c) degradability of the polymer; (d) responsiveness of the polymer to environmental
changes such as pH or enzymes [1]. Covalent binding between microgel and protein is avoided
because covalent attachment may change the drug’s properties such as activity.
Diffusion The driving force behind diffusion of protein molecules into oppositely charged
microgels is the gain in free energy. Initially, after mixing of protein and microgel, the strong
binding causes a strong increase in protein concentration at the interface between microgel
and surrounding solution. After time, with proceeding diffusion of protein into the microgel,
the concentration becomes more homogenous throughout the particle.
Swelling/ Osmotic pressure The uptake of protein by a charged microgel causes a release of
counterions. This lowers the osmotic swelling of the microgel’s osmotic swelling pressure. A
collapse of the charged microgel network can induce an increase in charge density, thus
promoting further protein uptake.
Degradability of microgel The use of a cleavable crosslinker enables the degradation of the
microgel with the help of e.g. enzymes or acids [4]. Ideally, the particle consists of a shell that
is impermeable to the protein. Degradation of the microgel particle leads to the release of the
protein.
108
Responsiveness of microgel pH-sensitive microgels can bind proteins electrostatically. The
increase of ionic strength leads to a screening of the electrostatic attractions between protein
and the oppositely charged groups in the microgel, resulting in a release of the protein from
the polymer network. Temperature-sensitive microgels can either release proteins upon
heating by “squeezing” them out or contrary retain the protein through entrapment or
increased charge density [5].
Usually, a combination of all factors affects the degree of uptake and release of proteins to
and from microgels.
V. A Kabanov et al. prepared poly(acrylic acid) (pAA) and N,N-dimethyl-N-ethylaminoethyl
methacrylate bromide) hydrogels and studied the uptake of different proteins [6]. They
described that the uptake of protein takes place primarily in the microgel shell. This leads to a
collapse of the shell while the particle core is still highly swollen. Eventually, the protein-
microgel complex propagates inward leading to further de-swelling of the microgel interior
until the complete polymer networks collapses. This was confirmed by C. Johansson et al.
who further investigated the diffusion of lysozyme into pAA microgels [3]. Microgels were 50
– 150 µm in size which enables visualization with confocal microscopy. They could see that a
shell of lysozyme initially forms rapidly preventing additional uptake of lysozyme. After ~
180 s, however, lysozyme diffuses deeper into the microgel core while no further de-swelling
of the polymer network can be observed. They state that lysozyme molecules diffusing to the
core do not originate from the lysozyme adsorbed in the shell, but from “new” lysozyme
molecules that diffuse through the already existing shell.
H. Bysell and M. Malmsten looked closer at the interaction between lysine and pAA
microgels as a function of ionic strength, pH, and protein concentration [7]. They state that the
de-swelling of the microgel network as described by V. A. Kabanov is greatly depending on
the mesh size of the polymer network and the protein size. A protein that is larger than the
mesh size is exclusively located in the microgel shell, while smaller proteins can penetrate
into the microgel core. They confirm that the formation of a protein-microgel complex in the
shell lead to the collapse of the shell. This happens faster than the protein diffusion, thereby
preventing further uptake into the microgel core. The mesh size in turn is dependent on the
crosslinking density within a microgel particle. The relation between crosslinking density and
protein uptake was investigated by G. E. Eichenbaum et al [8]. They determined the pore size
of the poly(methacrylic acid-co-acrylic acid) microgel via protein size exclusion. They found
out that the loading of proteins is highly dependent on the crosslinking density.
109
A both pH- and temperature-dependent uptake of proteins to microgels was studied by C. S.
O. Silva et al [9]. They used human immunoglobin G (IgG) and cationic PNIPAm microgels
that were functionalized with amine groups. They found out that the high degree of swelling
of the microgel network at low temperatures reduced the uptake of IgG, while ~ 90 % of the
initial IgG concentration could be adsorbed at temperatures above the VPTT. The authors
explained this by the higher charge density in the microgel at higher temperatures inducing
favorable electrostatic interactions.
The first step in testing a polymer’s ability as drug carrier is performing studies using a model
drug that is well described in literature. Many authors use cytochrome c (cyt-c) as a model
protein to study the interactions between proteins and microgels. Cyt-c is found e.g. in the
human body. It is for instance the main initiator of programmed cell death [10]. It has 104
amino acids and a heme group with the Fe(III) being six-coordinated (see Figure 9.1). It is
surrounded by four alpha helices and is covalently bound by two cysteine side chains [11]. The
properties of cyt-c are given in Table 9.1. It is expected that microgels and proteins interact
over electrostatic interactions between the negatively charged carboxylic groups in the
microgel network and the positively charged groups on the outside surface of cyt-c.
Figure 9.1 (a) Structure of cytochrome c; (b) Electrostatic surface properties of cytochrome c,
color-coded by electrostatic potential, blue: positive, red: negative. Images were generated
using the program Discoverstudio.
Table 9.1 Properties of cytochrome c [11][12][13].
Molecular weight 12.3 kDa
Isoelectric point pH 10.2
Radius of gyration 12.8 nm (native state)
Overall net charge at neutral pH +6.9 (at pH 10.2)
a b
110
9.2 Experimental Part
Specifications regarding DLS, SLS, and UV-Vis measurements can be found in chapter 4
(experimental part).
UV-Vis measurements were conducted in order to analyze the uptake and release kinetics of
cytochrome c. 0.005 g of microgel were dissolved in 3 mL of water at three different pH
values: 3, IEP (i.e. ~ 6), and 8. For each sample, the theoretical amount of carboxylic groups
was calculated and cytochrome c in a ratio of 1:1 was added assuming that each carboxylic
group is able to bind one protein molecule. For instance, for the microgel with statistical
distribution of ionizable groups and a monomer ratio VCL:Vim:IA = 60:20:20, 0.005 g
contain 7.26 * 10-6 mM of carboxylic groups. Therefore, 7.26 * 10-6 mM (0.089 g) of
cytochrome c was added.
For kinetic release studies, the respective amounts of microgel and protein in water at three
different pH values were allowed to stir overnight to reach equilibrium. After several hours of
stirring, UV-Vis measurements at 25 °C from 400 to 900 nm were carried out.
Afterwards, the samples were dialyzed in a composite regenerated cellulose membrane from
Millipore (NMWCO 30,000) in water with the respective pH. UV-Vis measurements were
repeated after specific times and the dialysis resumed (note: the dialysis water was not
changed during the dialysis process). UV-Vis measurements of the dialysis water were also
conducted under the same conditions to analyze the amount of cytochrome c released from
the microgel.
For controlled release studies, microgel samples and cytochrome c were dissolved in water of
pH = 8 with the same amounts explained above. After stirring overnight, UV-Vis
measurements was carried out. For dialysis, several parameters were changed:
(1) The pH of the dialysis water was changed from pH 8 to pH 3.
(2) The temperature was changed from T = 25 °C to T = 37 °C (i.e. above VPTT).
(3) NaCl was added.
Again, the samples were measured with UV-Vis after specific times.
To study the effect of ionic salt on the particle size and molecular weight, DLS and SLS
measurements were performed. Microgel and protein were mixed in buffer with pH 8 of the
according NaCl concentration (0, 30, 50, 75, 100, and 125 mM). The solution was stirred for
2 d to allow the sample to reach equilibrium. The samples were centrifuged to remove
unbound protein from the solution. For SLS measurements, the samples were lyophilized to
determine the exact concentration and re-dispersed in buffer with the according salt
concentration. SLS measurements were performed at 25 °C.
111
9.3 Results and Discussion
In this chapter, the interactions between microgels with statistical distribution of ionizable
groups as well as core-shell structure and the protein cytochrome c (cyt-c) are looked at.
Cyt-c is used for this study because it is positively charged over a broad pH range (see Figure
9.3). The microgels that are used, however, are positively charged at pH 3, neutral at pH ~ 6,
and negatively charged at pH = 8. Thus, interactions between microgels and protein can be
studied at these three different pH values. Furthermore, the red color of cyt-c makes it suitable
for UV-Vis measurements to analyze the binding kinetics between protein and microgels.
First, however, the conformational structure of cytochrome c at pH 3 and pH 8 is measured
via circular dichroism (CD) spectroscopy (see Figure 9.2).
Amount of conformation / %
pH 3 pH 8
α-helix 41.3 48.5
β-sheet 13.6 4.6
turn 20.1 26.3
random 25.1 20.6
Figure 9.2 CD spectra of cytochrome c in PBS buffer with different pH and structural
analysis using Yang’s reference.
Both spectra have two negative minima at around 222 and 208 nm and a positive maximum at
around 194 nm. The double minimum is typical for a helical structure [14][15]. A small
shoulder at around 217 nm can be attributed to pleated β-sheet conformation. The structures
were analyzed using Yang’s reference structural analysis (see Figure 9.2).
Yang’s reference predicts that the globular domain can support a mixture of all conformations
at both pH, though α-helix and random coil conformation are dominant. The amount of β-
sheets is decreased slightly at pH 8 compared to pH 3. The positions of the peaks do not vary
when the pH of the solution is changed indicating that cyt-c does not change its conformation
dramatically.
Figure 9.3 shows that cytochrome c is positively charged over a broad pH range. Its
isoelectric point (IEP) is above pH 8. Figure 9.3 gives also the electrophoretic mobility of the
ampholyte PVCL-co-Vim-co-IA microgel with a statistical distribution of ionizable groups
112
and a monomer ratio of 60:20:20. It is positively charged at low pH, while being negatively
charged at high pH. The IEP is at ~ 6.2.
Figure 9.3 Electrophoretic mobility of cytochrome c (black curve) and microgel with
statistical distribution of ionizable groups as an example (monomer ratio VCL:Vim:IA =
60:20:20, red curve). The measurements were conducted at T = 25 °C.
Next, UV-Vis spectra of pure cyt-c were measured at pH 3, 6, and 8 (spectrum shown in the
appendix). Cyt-c exhibits two prominent peaks at 402 and 416 nm (γ-band). These peaks are,
however, prone to change from one to two peaks depending on the oxidation state of iron: the
peak at 402 nm belongs to Fe(III), while the peak at 416 nm can be assigned to Fe(II) [16].
Therefore, the smaller peak at 530 nm (α-band) was used as reference peak. Pure microgel
samples show no peaks in this range (for spectrum see appendix).
Microgel samples with statistical distribution of ionizable groups and a monomer ratio of
VCL:Vim:IA = 60:20:20 were mixed with cyt-c in buffer of pH 3, 6 or 8 at room temperature.
After incubating for 24 h, the samples were dialyzed against buffer of the respective pH.
Depending on the pH, the protein was either bound by the microgel and could thus not be
washed out the dialysis bag or not bound by the microgel and could thus be washed out the
dialysis bag. The release of cyt-c was studied with UV-Vis spectroscopy after various time
intervals. Samples were taken both inside and outside the dialysis bag.
Figure 9.4 shows the change of the UV-Vis spectra over time. The absorbance of the peak at
530 nm before the dialysis is the same for all samples indicating that the concentration in all
samples is the same. Already after 30 min of dialysis, the absorbance for pH 3 is decreased
clearly, while at the same time, the same peak increases in the dialysis water. This indicates
that cytochrome c is not bound within the microgel particle and is able to penetrate through
the dialysis membrane. After 360 min of dialysis, no cyt-c is left in both the samples at pH 3
and pH 6.2. This is also confirmed optically: only the solution at pH 8 maintains its red color.
113
Figure 9.4 UV-Vis measurements of microgel sample with statistical distribution of ionizable
groups and monomer ratio VCL:Vim:IA = 60:20:20. Black curve: pH 3; red curve: pH 6.2;
blue curve: pH 8. (a) Before the start of dialysis; (b) After 30 min of dialysis; (c) After 180
min of dialysis; (d) After 360 min of dialysis. Inset graphs show UV-Vis measurements of
dialysis water. Inset images give optical confirmation of protein release.
The decrease of the peak at 530 nm is analyzed over time and shown in Figure 9.5 for the
sample with statistical distribution of ionizable groups. The graphs show the kinetics of the
dialysis process. An absorbance of 1 means that all carboxylic groups within the microgel
particle bind cyt-c (i.e. 100 %). The measurements were repeated for microgels with a
different monomer ratio as well as a core-shell structure (graphs see appendix). The amount of
cyt-c remaining in the microgel after 24 h of dialysis is analyzed with regard to microgels
with varying distributions of ionizable groups and various amounts of incorporated basic and
acidic moieties at different pH values (see Table 9.2). No binding (i.e. absorbance = 0) can be
seen at pH 3 and pH 6.2 for microgels with statistical distribution of ionizable groups as well
as for the core-shell particles with itaconic acid in the shell. The situation, however, is
different at pH 8 where all of them bind ~ 50 % of the theoretical amount of cyt-c. This
surprisingly low amount can be explained as follows: The first cyt-c molecules that bind to a
microgel particle will bind to the first carboxylic group it “meets”, i.e. on the particle surface.
c d
a b
114
Figure 9.5 Change of absorbance with dialysis time for ampholyte microgel with statistical
distribution of ionizable groups (VCL:Vim:IA = 60:20:20). Black curve: pH 3; red curve: pH
6.2; blue curve: pH 8.
As more and more protein is bound, they will eventually block the polymer network for
further cyt-c to enter. So, though more protein could be bound theoretically, the microgel
interior is probably relatively protein-free. This behavior was already seen by T. Hoare and R.
Pelton [17]. They analyzed the uptake of a cationic drug to PNIPAM microgels containing
carboxylic groups. The authors observed a local collapse of the polymer network in response
to the binding of the protein to the anionic groups. The decreased charge density reduces
repulsion forces between like charges and leads to a phase transition within the microgel.
Further diffusional uptake of protein into the microgel network is hindered or limited.
Table 9.2 Amount of cyt-c (in %) remaining after end of dialysis for microgel samples with
different distributions of ionizable groups. Measurements were conducted at T = 25 °C.
Statistical Core-Shell (IA:Vim = 10:10)
80:10:10 60:20:20 Core +
Shell -
Core -
Shell + (1:1)
Core -
Shell + (1:5)
pH 3 0 0 0 12 14
pH 6.2 0 0 0 28 40
pH 8 46 53 41 52 71
Core-shell particles with itaconic acid in the core, however, show a different behavior. Even
at pH 3, with the shell being positively charged and the core being neutral, ~ 13 % cyt-c is
bound by the microgel. At pH 6, both core and shell are neutral. 28 % and 40 % of cyt-c are
bound by core-shell particles with a ratio between core and shell of 1:1 and 1:5, respectively.
At pH 8, the sample with the thicker shell (1:5) is able to bind ~ 70 % of cyt-c, though the
neutral shell is not expected to bind cyt-c over electrostatic interactions. These latter results
show that cyt-c is bound by the microgels even in the absence of electrostatic
115
forces supporting binding (pH 6) and even so in the presence of electrostatic forces opposing
binding (pH 3). This leads to the assumption that cyt-c is bound to the microgel particle not
only over electrostatical interactions, but suggests the existence of another type of interaction,
hydrophobic interactions, between cyt-c and microgel. The idea of hydrophobic interactions
was first introduced by W. Kauzmann in 1959 [18]. He described them as the tendency of
nonpolar solutes to adhere to each other in aqueous solutions.
In order to analyze the influence of hydrophobic interactions on the binding of cyt-c, the
release kinetics between neutral core-shell microgels (i.e. containing no acid and basic
moieties) and cyt-c were studied with UV-Vis.
Figure 9.6 reveals that ~ 11 % of cyt-c remains in the microgel at pH 3 and 8 at T = 25 °C.
The measurements indicate clearly that interaction between neutral microgel and positively
charged protein takes place at both pH. Since the microgel has no charges at both pH, these
results suggest that hydrophobic, non-coulombic interactions play an important role in the
binding of cyt-c. Hydrophobic interactions between microgels and proteins were also
observed by other authors [19][20][21]. Hydrophobic interactions between cyt-c and the
ampholyte microgel take place between the hydrophobic amino acid residues of the N- and C-
terminal α-helices of cyt-c and the hydrophobic backbone of the polymer chain in the
microgel [22].
Figure 9.6 Change of absorbance with dialysis time for core-shell microgels without
ionizable groups. Black curve: pH 3; blue curve: pH 8.
The hydrophobicity of a microgel increases when the solution temperature is increased from
25 °C to 37 °C (i.e. above the VPTT) [23]. At 37 °C, microgels are in a collapsed state
(combined with a decrease in size of ~ 61 %) exhibiting more hydrophobic domains. No
specific driving force for the release of cyt-c is expected at this elevated temperature. Figure
9.6 reveals, however, that at higher temperatures significantly more cyt-c remains in the
116
microgel after 48 h. Compared to the amount of protein remaining at 25 °C, about 73 % more
protein is bound at 37 °C. This was already observed by other authors [23]. N. Shamim for
instance prepared thermosensitive Fe3O4 nanoparticles and studied the uptake of lysozyme
below and above the VPTT of the PNIPAm shell [24]. They confirmed that hydrophobic
interactions are the driving forces behind the uptake of the protein at elevated temperatures.
This strongly supports the assumption that the increase of hydrophobicity with rising
temperature increases the hydrophobic interactions and leads to more hydrogen bonds
between polymer network and protein [23]. Hydrophobic interactions that occur at high
temperatures are an entropically driven process, while electrostatic interactions at lower
temperatures at pH 8 are enthalpically driven processes [20].
As it was shown that protein uptake is possible at pH 8 due to the opposite charges of
microgel and cyt-c, the release from the microgel was studied using various triggers. This is
studied for microgels with a statistical distribution of ionizable groups (VCL:Vim:IA =
60:20:20) as well as for core-shell microgels with a thick shell (cVCL-/sNIPAm+, 1:5). All
microgels were loaded with cyt-c at pH 8 in the same way as explained above.
The first trigger studied is temperature. Due to the particles’ temperature sensitivity, they
shrink at temperatures above ~ 32 °C. It is a likely scenario that the particles “squeeze” out
the protein at higher temperatures. Therefore, the loaded microgel particles were heated up to
37 °C and dialyzed. If protein is released from the microgel, it will leave the dialysis tube.
Again, UV-Vis measurements were performed to monitor this process. Figure 9.7a shows that
most protein is staying inside the microgel and cannot be released by a change in temperature.
Even after three days of dialysis, almost 80 % of the initial protein amount remains. This
behavior is attributed to hydrophobic interactions between protein and microgel that are
pronounced at elevated temperatures as discussed above.
A different trigger is the addition of salt. Salt ions can shield the interactions between the
carboxylic groups and the protein. Figure 9.7b shows that almost 70 % of the previously
bound cyt-c is released within the first 4 hours. It indicates further that other than electrostatic
interactions between microgel and protein play a role and prevent the complete release of cyt-
c form the microgel. This non-electrostatic part is attributed to hydrophobic interactions as
already discussed above.
In a next experiment, the pH of the microgel dispersion was changed from pH 8 to pH 3
(Figure 9.7c). Carboxylic groups that are negatively charged in basic conditions, become
117
neutral at low pH. After almost 24 h, no protein remains in the microgel particle indicating
that a change in pH results in a full release of cyt-c.
Figure 9.7 Change of absorbance with dialysis time for ampholyte microgel with statistical
distribution of ionizable groups (VCL:Vim:IA = 60:20:20). (a) Change of temperature from
25 to 37 °C; (b) Addition of 0.1 M NaCl; (c) Change of pH 3 to 8; and (d) core-shell structure
cVCL-/sNIPAm+ (1:5), change of pH = 3 to 8.
These results lead to the following conclusions about the release of cyt-c by ampholyte
microgels with a statistical distribution of ionizable groups:
- Temperature is no effective trigger for the release of cyt-c from the microgel. This
indicates that the protein is mainly bound on the surface. At elevated temperatures,
hydrophobic interactions between microgel and protein increase preventing a complete
release.
- The addition of salt suppresses the interaction between cyt-c and microgel. As a
consequence, cyt-c is released from the polymer network. The release, however, is
only partial (~ 70 %) and further suggests the presence of non-coulombic,
hydrophobic interactions.
- For a complete release of protein, a change in pH from 8 to 3 proved to be most
a
c d
b
118
effective. In this case, hydrophobic interactions are overcome by electrostatic
repulsion.
The release of protein through a change in pH from core-shell particles was also studied. For
this, core-shell microgels with a negative core and a positive, thicker shell were taken (cVCL-
/sNIPAm, 1:5). The results can be seen in Figure 9.7d. A clear difference between statistical
and core-shell particles is discernible. While all protein is released from microgel with a
statistical distribution of ionizable groups, this is only about 50 % for core-shell particles.
Obviously, a different mechanism prevents the complete release of cyt-c.
At pH 8, the core is negatively charged, while the shell is neutral. Protein is bound in the core.
A change in pH to acidic conditions leads to a neutralization of the core, while the shell
becomes positively charged. The results indicate that the protein is trapped inside the particle
because the positively charged shell prevents release from the interior.
In the following, was studied how the uptake of cyt-c as well as the presence of salt influences
the particle size. The particle size was measured with increasing ionic strength (0 – 125 mM)
at pH 8. At pH 8.0, the microgel particles are negatively charged, while cyt-c is positively
charged, thus interaction between both is expected. With an increase in ionic strength,
however, the charges of both are screened and less interaction is anticipated. This behavior is
tested for four different samples, namely core-shell particles cVCL-/sNIPAm+ (1:1) and
cVCL-/sNIPAm+ (1:5) and microgels with a statistical distribution of ionizable groups and a
monomer ratio of VCL:Vim:IA = 80:10:10 and 60:20:20.
Figure 9.8 shows that unloaded microgels with statistical distribution of ionizable groups and
a monomer ratio of VCL:Vim:IA = 80:10:10 decrease in size from 230 nm at 0 mM NaCl to
200 nm at 125 mM NaCl correlating to a decrease in volume of 34 %. At pH ~ 8, the microgel
particles are negatively charged. The charges repel each other resulting in a higher particle
size. With increasing ionic strength these charges become more and more screened leading to
a contraction of the polymer network. For the sample 60:20:20, the decrease in volume is only
14 %. Loaded microgels, in contrast, increase in size with increasing ionic strength (34 vol-%
and 18 vol-% for 80:10:10 and 60:20:20, respectively, from 0 mM to 125 mM NaCl). With no
salts present, the negatively charged carboxylic groups and the positively charged cyt-c
interact. With increasing ionic strength, the electrostatic affinity between charges of the
carboxylic groups and the protein decreases. As a result, the microgel becomes more and
more negative and increases in size. For unloaded and loaded microgels, the change in
volume generated by the addition of salt is less pronounced for microgels with a higher
119
Figure 9.8 Influence of ionic strength on particle radius RH for microgel with statistical
distribution of ionizable groups with monomer ratio VCL:Vim:IA = (a) 80:10:10, (b)
60:20:20. Black curve: unloaded microgel; red curve: loaded microgel. All measurements
were conducted in buffer at pH 8.0. Microgel sketches in (a) illustrate the interaction between
salt and cytochrome c .
amount of ionizable groups. The less pronounced change in size for microgels with a high
amount of ionizable groups (60:20:20) further supports the assumption explained above. The
uptake of cyt-c is primarily taking place in the outer shell of a microgel particle, while the
core is mostly unaffected. A higher amount of “unaffected” charges in the core thus results in
a higher particle size of VCL:Vim:IA = 60:20:20 compared to 80:10:10. The release of cyt-c,
therefore, leads to a less pronounced size change for highly charged microgels.
Core-shell particles show a similar behavior (see Figure 9.9). The particle size decreases for
unloaded microgels with increasing ionic strength due to negative charges being screened at
pH 8. Particles loaded with cyt-c increase in size with increasing ionic strength which is
attributed to negative charges being generated as cyt-c is released from the polymer network.
It is noticeable that the degree of swelling of loaded microgels with increasing ionic strength
is considerably lower than for microgel with a statistical distribution of ionizable groups.
Figure 9.10. shows the curves for loaded microgels of all architectures discussed above.
- -
-
- -
- -
- -
-
- -
-
a b
120
Figure 9.9 Influence of ionic strength on particle radius RH for microgel with core-shell
distribution of ionizable groups with increasing shell thickness. Core:shell ratio: (a) 1:1, (b)
1:5. Black curve: unloaded microgel; red curve: loaded microgel. All measurements were
conducted in buffer at pH 8.0.
Instead of the hydrodynamic radius RH, the ratio between the radius at an ionic strength = 0
mM (R0) and the radius at the respective ionic strength (Rx) is given to better visualize the
particle swelling. Note that before each measurement, all samples were allowed to reach
equilibrium. Core-shell particles have a considerably lower slope than microgel with s
statistical distribution of ionizable groups. This suggests that particles with a statistical
distribution easily release cyt-c upon the addition of salt resulting in a swelling of the polymer
network. Core-shell particles, in contrast, release cyt-c in a less pronounced way suggesting
that the neutral shell hinders the complete release of protein. This behavior is more
pronounced for core-shell microgels with a higher shell thickness.
Figure 9.10 Influence of ionic strength on ratio R0/Rx (R0 = RH at ionic strength = 0 mM; Rx
= RH at respective ionic strength) for microgels loaded with cytochrome c. Graphs are shown
for particles with statistical as well as core-shell charge distribution. All measurements were
conducted in buffer at pH 8.0 and at T = 25°C.
a b
121
The release of cyt-c upon the increase of the ionic strength also influences the molecular
weight MW of the microgels. This was already studied by M. Smith and A. L. Lyon for the
interactions between cyt-c and negatively charged PNIPAm-co-AAc microgels [25]. They
showed that the molecular weight decreases with increasing ionic strength due to the release
of cyt-c.
Therefore, MW of microgels with various distributions of ionizable groups (statistical and
core-shell) and various monomer ratios (80:10:10 and 60:20:20 for statistical distribution) in
the absence of cyt-c (black dots), in the presence of cytochrome c at low ionic strength
(30 mM, red dots) and high ionic strength (125 mM, blue dots) was measured at pH 8 (see
Figure 9.11). Note that samples were centrifuged before each measurement to remove
unbound protein.
In the absence of cyt-c, MW of microgels with statistical distribution of ionizable groups and
monomer ratios of 80:10:10 and 60:20:20 does not differ significantly (~ 2x 108 g/mol, blue
dots). MW increases continuously for core-shell microgels with increasing shell thickness
(blue dots).
A loading of the respective microgel particles with cyt-c leads to an increase in MW (red dots).
The sample 80:10:10 increases ~ 1.6 times in MW indicating that a large amount of cyt-c was
incorporated. Samples with twice the amount of ionizable groups (60:20:20) increase only 1.9
times in MW. This discrepancy between theoretical capacity of protein uptake and measured
capacity was already observed discussed for release experiments and confirms the formation
of a protein layer on the microgel surface hindering further uptake. The amount of protein that
is bound by core-shell particles is strongly dependent on the shell thickness. An increase in
shell thickness leads to an increase in MW, though the addition of a thicker shell does not add
further acidic moieties (i.e. itaconic acid) that are able to bind an increasing amount of cyt-c.
This indicates that hydrophobic interactions between protein and neutral shell play an
important role. The difference between unloaded and loaded microgels allows to estimate the
amount of cyt-c (MW = 12.300 g/mol) bound by microgels. The results are given in Table 9.3.
An increase in ionic strength from 30 mM (red dots) to 125 mM (blue dots) at pH 8 leads to a
release of cyt-c due to screening of attractive forces accompanied with an increase in particle
size. For microgel with a statistical distribution of ionizable groups, only small differences
between unloaded microgels (black dots) and loaded microgels (blue dots) in 125 mM salt
solution can be seen suggesting an almost complete release of cyt-c in high salt solution. The
difference between unloaded and loaded microgels at 125 mM salt concentration is more
pronounced for core-shell microgels and rises significantly with increasing shell thickness.
122
This suggests that the protein is not released completely through the addition of salt
supporting the results shown in Figure 9.7 and Figure 9.9.
Table 9.3 Amount of cytochrome c bound by microgels at pH 8 and T = 25 °C.
Sample Number of bound
protein at 30 mM
VCL:Vim:IA = 80:10:10,
statistical
8.5 x 103
VCL:Vim:IA = 60:20:20,
statistical
3.9 x 104
cVCL-/sNIPAm+, c:s = 1:1 1.1 x 105
cVCL-/sNIPAm+, c:s = 1:2.5 9.5 x 105
cVCL-/sNIPAm+, c:s = 1:5 2.0 x 106
Figure 9.11 Relationship of molecular weight MW and microgel architecture in the absence
(black dots) and presence (red dots) of cytochrome c in 30 mM NaCl and 125 mM NaCl (blue
dots). The dashed line is a visual help to separate microgels with core-shell structure from
microgels with statistical distribution of ionizable groups.
123
9.4 Summary
This chapter investigated the use of ampholyte microgels as container for the uptake and
release of proteins. As a model protein, cytochrome c (cyt-c) was used due to its red color and
its positive charge over a wide pH range. The interactions between ampholyte microgels and
cyt-c were studied at pH 3, 6, and 8. UV-Vis measurements confirmed that no cyt-c was
bound at pH 3 for microgels with statistical distribution of ionizable groups and core-shell
microgels with IA in the shell. A small amount of protein, however, remained in the microgel
at this pH for core-shell microgels with Vim in the shell. Since both shell and protein are
positively charged at this pH, the interaction was attributed to hydrophobic interactions. This
was confirmed by UV-Vis studies performed with neutral core-shell microgels. About 50 %
of cyt-c was bound at pH 8.
After loading the microgels with cyt-c at pH = 8, various triggers for the release of cyt-c from
the microgel network were studied. A raise in temperature did not prove to be an effective
trigger to release cyt-c. About 80 % of the initial concentration remained in the particle. The
addition of salt released about 70 % of the bound protein. The non-complete release indicated
the presence of hydrophobic interactions that are not influenced by the addition of salt.
Finally, the pH of the solution was changed from pH 8 to pH 3. For microgels with a
statistical distribution of ionizable groups, this led to a complete release of protein. Core-shell
microgels were able to “trap” the protein inside the particle and thus hindering its release.
The addition of salt also influences the microgel size. Unloaded particles decrease in size with
increasing ionic strength. Since measurements were performed at pH 8, this reflects that
negative charges are shielded by the salt ions and thus leading to a collapse of the microgel
structure. Loaded particles in contrast increased in size with an increasing ionic strength. This
can be attributed to the release of cytochrome c from the microgel, generating more and more
negative charges that repel each other leading to an expansion of the microgel structure. The
change of molecular weight MW for loaded und unloaded microgel in the presence (125 mM)
and absence (30 mM) of salt supported these results. While MW is identical for unloaded and
loaded microgels with statistical distribution of ionizable groups in 125 mM salt solution
indicating a complete release of protein, MW for loaded core-shell microgels in 125 mM salt
solution is higher than for unloaded core-shell microgels. The difference was shown to
increase with increasing shell thickness suggesting that the shell hinders a complete release of
the protein.
The results discussed in this chapter illustrated the importance of different distribution of
ionizable groups on the uptake and release of proteins from the polymer network.
124
9.5 Literature
[1] W. B. Liechty, D. R. Kryscio, B. V. Slaughter, N. A. Peppas, Annu. Rev. … 2010, 1,
149–173.
[2] S. H. Gehrke, J. P. Fisher, M. Palasis, M. E. Lund, Ann. N. Y. Acad. Sci. 1997, 831,
179–207.
[3] C. Johansson, P. Hansson, M. Malmsten, J. Phys. Chem. B 2009, 113, 6183–6193.
[4] N. Murthy, M. Xu, S. Schuck, J. Kunisawa, N. Shastri, J. M. J. Fréchet, Proc. Natl.
Acad. Sci. U. S. A. 2003, 100, 4995–5000.
[5] H. Bysell, R. Månsson, P. Hansson, M. Malmsten, Adv. Drug Deliv. Rev. 2011, 63,
1172–1185.
[6] V. a Kabanov, V. B. Skobeleva, V. B. Rogacheva, a B. Zezin, J. Phys. Chem. B 2004,
108, 1485–1490.
[7] H. Bysell, M. Malmsten, Langmuir 2006, 22, 5476–5484.
[8] G. M. Eichenbaum, P. F. Kiser, D. Shah, S. a. Simon, D. Needham, Macromolecules
1999, 32, 8996–9006.
[9] C. S. O. Silva, R. P. Baptista, A. M. Santos, J. M. G. Martinho, J. M. S. Cabral, M. Â.
Taipa, Biotechnol. Lett. 2006, 28, 2019–2025.
[10] R. Sardari, R. Rezaee, Int. J. Electrochem. Sci. 2013, 8, 11578–11587.
[11] I.-J. Hsu, Y.-J. Shiu, U.-S. Jeng, T.-H. Chen, Y.-S. Huang, Y.-H. Lai, L.-N. Tsai, L.-Y.
Jang, J.-F. Lee, L.-J. Lin, et al., J. Phys. Chem. A 2007, 111, 9286–9290.
[12] O. Kimhi, H. Bianco-Peled, Langmuir 2007, 23, 6329–35.
[13] J. D. Carbeck, I. J. Colton, J. R. Anderson, J. M. Deutch, G. M. Whitesides, J. Am.
Chem. Soc. 1999, 121, 10671–10679.
[14] T. K. Das, S. Mazumdar, S. Mitra, Eur. J. Biochem. 1998, 254, 662–670.
[15] C. P. Lee, Current Topics in Bioenergetics, Elsevier, 2014.
[16] M. Hulko, I. Hospach, N. Krasteva, G. Nelles, Sensors (Basel). 2011, 11, 5968–80.
[17] T. Hoare, R. Pelton, Langmuir 2008, 24, 1005–12.
[18] W. Kauzmann, Adv. Protein Chem. 1959, 14, 1–63.
[19] X. Chen, S. Chen, J. Wang, Analyst 2010, 135, 1736–41.
[20] K. H. Kim, E. K. Lee, Biotechnol. Bioprocess Eng. 2007, 12, 366–371.
[21] Y. Liu, Y. Liu, R. Guo, J. Colloid Interface Sci. 2010, 351, 180–9.
[22] P. P. Parui, M. S. Deshpande, S. Nagao, H. Kamikubo, H. Komori, Y. Higuchi, M.
Kataoka, S. Hirota, Biochemistry 2013, 52, 8732–8744.
[23] N. Welsch, A. Wittemann, M. Ballauff, J. Phys. Chem. B 2009, 113, 16039–45.
[24] N. Shamim, H. Liang, K. Hidajat, M. S. Uddin, J. Colloid Interface Sci. 2008, 320, 15–
21.
[25] M. Smith, L. A. Lyon, Macromolecules 2011, 44, 8154-8160.
125
10. 1 Biomineralization of CaCO3 in Zwitterionic Microgels
10.1.1 Introduction
Nature is an excellent teacher when it comes to combining the mechanical strength and
toughness of inorganic material with the softness and formability of polymers to form
complex structures with unique properties and specific functions. The control of organic
material over inorganic material to form minerals is called biomineralization. Examples are
manifold – so far over 60 different minerals have been detected [1] - and vary from magnetic
crystals of magnetotactic bacteria, shells, and corals to tooth enamel and bones in humans.
Biomineralization is very complex and many factors such as the organic matrix or the ion
concentration influence the nucleation process and the mineral orientation, size, density, and
morphology [2].
In general, biomineralization is divided into two categories: Biologically induced
mineralization (BIM) and biologically controlled mineralization (BCM) [1]. Biologically
induced mineralization occurs extracellularly or epicellularly (i.e. on the cell surface) when
the mineralization is little controlled and an organism segregates a metabolic by-product that
interacts with the environment leading to the precipitation of the mineral [3]. The resulting
minerals are poorly crystallized and polydisperse in size, have no specific morphology, but
often impurities. Biologically controlled mineralization arises when the organism exerts a
high degree of control over the mineralization process. This happens for instance if a defined
space such as vesicles or macromolecules exists and leads to minerals that are monodisperse
in size with a well-ordered crystal morphology. BCM can occur intracellularly meaning that
the mineralization process is not as sensitive to environmental conditions such as pH,
temperature, pressure, and redox potential as BIM [3].
Most minerals in nature are based on calcium. This can be explained by the abundance of
calcium in the world’s oceans as well as the low water solubility of calcium salts [1]. Calcium
carbonate CaCO3 is used in nature to strengthen the skeletons of mussel shells, snails, or
corals. It is found in six different modifications: amorphous (i.e. unordered, isotropic) CaCO3
(ACC), the three anhydrous crystalline forms vaterite, aragonite, and calcite, and the two
hydrated crystalline forms calcium carbonate monohydrate and calcium carbonate
hexahydrate (ikaite) [1]. Though aragonite and calcite are the most stable forms, stabilization
of less stable forms can be achieved by the addition of other ions such as Mg2+ or Li+ or by
enclosing it in an organic matrix.
In literature, different types of polymers are used for mimicking the biomineralization
process: Peptides [4][5], dendrimers [6][7], synthetic polymers and monolayers. In the following,
126
two of them are described in more detail: Self-assembled monolayers (SAMs) and hydrogels.
Though none of the approaches are able to mimic the complex processes that enfold in nature,
they can be used to gain insights in the principles of biomineralization.
Hydrogels enable a stable transport via diffusion that is advantageous for nucleation. They,
however, consist of many pores where precursor ions are isolated. Nucleation is thus more
hindered than in solution because nucleation only occurs if a certain limit of supersaturation is
reached [2]. Still, this approach enables a better control of mineral growth than from
supersaturated solutions.
For instance, K. Furuichi et al. prepared a poly(acrylic acid) hydrogel containing phosphate
ions [8]. Through diffusion of calcium ions into the matrix, nanocrystalline hydroxylapatite
were formed. They observed that the soft hydrogel was transformed into a solid disk after
72 h of calcification. After heating up of the hybrid material to 700 °C, a porous
hydroxylapatite structure remained. O. Grassmann and P. Löbmann synthesized hydrogels
containing sulfonic acid to mineralize CaCO3 [9]. They showed that an increase in sulfonic
acid content leads to a change in morphology. While a low content of sulfonic acid leads to
the formation of pseudo-octahedral morphology, a higher content leads to the formation of
cuboctahedron morphology.
P. Liu and J. Song mention the possibility to use zwitterionic hydrogels for the growth of
hydroxyapatite [10]. They emphasize the anti-polyelectrolyte behavior and the ability of
zwitterionic polymers to bind both cationic and anionic precursor ions and predict a lower
activation energy for nucleation and growth. A comparison with other charged and uncharged
comonomers revealed that nucleation of hydroxylapatite is possible in all hydrogels, though
differences could be observed in the mineralization degree: in uncharged hydrogels,
mineralization occurred on the surface, while for anionic and cationic hydrogels, spherical
mineral nodules were formed. Sulfobetaine hydrogels, however, had a three times higher
mineral content than anionic or cationic hydrogels. The authors attribute this to two factors:
first, the anti-polyelectrolyte effect leads to a swelling of the hydrogels a facilitation in
electrolyte uptake, and secondly, the ability of zwitterions to bind and stabilize cationic and
anionic precursor ions. Furthermore, the authors could show the influence of the crosslinker
degree on the crystal size: the higher the amount of crosslinker, the smaller the formed
nodules. They assume that a higher crosslinking degree hinders a free diffusion of precursor
ions which leads to a gradient in the concentration of precursor ion in the hydrogel network
resulting in a lower mineral growth rate.
127
This chapter is inspired by the work done by P. Liu and J. Song. Here, phosphobetaine
microgels were used to study their influence on the mineralization process of CaCO3. CaCO3
was used as a standard model mineral because of its well studied properties and high
abundance in nature [11][12]. Phosphobetaine was used due to its known high affinity for
calcium ions [13][14]. The complexation of Ca2+ leads to a saturation of ions in a confined space
that is essential for the nucleation of CaCO3 crystals. The use of PVCL-based zwitterionic
phosphobetaine microgels offers other important advantages apart from their low cytotoxicity
and anti-polyelectrolyte effect that was not studied by P. Liu and J. Song: their temperature-
and pH-sensitivity as well as their confined volume. PVCL microgels shrink at temperatures
around 32 °C [15][16]. Phosphobetaine is pH-sensitive [17] and a change of the pH of the
surrounding medium leads to a change in particle size. Both effects are likely to influence the
nucleation and mineralization process of CaCO3.
128
10.1.2 Synthesis Procedure
For this chapter, zwitterionic microgels based on PVCL with phosphobetaine as a comonomer
were used. The microgel synthesis via precipitation polymerization is described in chapter 5.
For the synthesis of CaCO3, the pH of the solution is required to be above ~ 6. In acidic
medium, CaCO3 decomposes according to the following equation:
CaCO3 + 2 HCl → CaCl2 + H2O + CO2 (1)
Biomineralization is initiated in two different ways: through direct synthesis and via gas phase
diffusion (see Figure 10.1). In one batch prepared via gas diffusion, the samples were dialyzed
before being put in a desiccator.
Figure 10.1 Overview of the synthesis routes for mineralization of CaCO3. (1) Gas diffusion;
(2) Gas diffusion with prior dialysis; (3) Direct synthesis.
(1) Direct Synthesis
The respective microgel was dissolved in destilled water. CaCl2·2H2O was dissolved in 1 mL
of destilled water and added to the aqueous microgel solution. (NH4)2CO3 was dissolved in
5 mL of destilled water and added dropwise under vigorous stirring to the aqueous
microgel/CaCl2 solution. After 10 min of stirring the solution was allowed to stand for 24 h at
room temperature. The product was dialyzed against water for 3 d. The respective amounts of
educts can be seen in Table 10.1.
129
Table 10.1 Amounts of educts used for direct synthesis of CaCO3.
Sample name Theor. amount of
CaCO3 / wt-%
m(CaCl2·2H2O)/
g
m((NH4)2CO3)/
g
m(microgel)/g
D(2)13* 13.0 0.015 0.009 0.065
D(2)28 28.4 0.040 0.026 0.065
D(2)37 37.6 0.058 0.038 0.065
D(2)50 50.4 0.097 0.064 0.065
D(2)67 67.0 0.195 0.127 0.065
* Explanation of the sample name: “D” = “Direct synthesis”; “(2)” = Microgel contains 2 wt-
% PB; “13” = Theoretical amount of CaCO3 is 13 wt-%.
(2) Gas Phase Diffusion
The respective microgel was dissolved in destilled water. CaCl2·2H2O was dissolved in 1 mL
of destilled water and added to the aqueous microgel solution. The sample amounts
correspond to the ones given in Table 10.1 (the sample names are e.g. E(2)13: “E” =
Exsikkator = Desiccator). The aqueous microgel/CaCl2 solution is put in a sealed desiccator
with solid (NH4)2CO3 and allowed to stand for 24 h at room temperature. The product was
dialyzed against water for 3 d.
The following variations from the above described procedure were performed:
(a) Samples with a higher microgel concentration were prepared. Therefore, freeze-dried
microgel samples were dissolved in water with a ratio of 1:3. The viscous solution was
treated as described above. Here, the time in the desiccator was increased to 48 h. The
sample amounts are given in Table 10.2.
(b) After the addition of CaCl2, the samples were dialyzed to remove all Ca2+ that is not
bound by the anionic groups in the microgel. After dialyzing for 2 d, the samples were
put in a desiccator for 24 h and dialyzed anew. The sample names are e.g. E(2)13v:
“v” = vordialysiert = dialyzed before addition of (NH4)2CO3.
(c) The synthesis of pre-dialyzed samples was performed at 40 °C, i.e. above the VPTT of
the microgel. This was done for the sample E(10)37.
(d) The pH of the solution was varied. The pH of the original samples is ~ 8.5. It was
adjusted to pH 3 and pH 10 with 1 M NaOH and 37 % HCl, respectively. This was
done for the sample E(10)67.
130
Table 10.2 Amounts of educts used for gas diffusion synthesis of CaCO3.
Sample
name
Amount of PB /
wt-%
m(microgel) /
g
Theor. amount
of CaCO3 / wt-
%
m(CaCl2·2H2O)/g
E(2)13H* 2 0.287 13.0 0.064
E(4)13H 4 0.337 13.0 0.076
E(4)7H 4 0.416 7.0 0.050
E(4)5H 4 0.852 5.0 0.073
E(10)5H 10 0.476 5.0 0.041
* Explanation of the sample name: “E” = “Exsikkator = Desiccator”; “(2)” = Microgel
contains 2 wt-% PB; “13” = Theoretical amount of CaCO3 is 13 wt-%; “H” = Highly
concentrated.
131
10.1.3 Results and Discussion
In this chapter, the process of biomineralization of CaCO3 in zwitterionic microgels is
analyzed stepwise. Therefore, the behavior of the particles is analyzed as follows:
(a) Pure PB Microgels
(b) Microgels with CaCl2
(c) Microgels with CaCO3
(a) Pure PB Microgels
The properties of phosphobetaine microgels obtained via precipitation polymerization were
analyzed and discussed in my diploma thesis. It was shown that the particle size was strongly
dependent on the amount of phosphobetaine incorporated. The size of particles with 2 wt-%
of PB at 25 °C was 190 nm and decreased to 130 nm with 10 wt-% PB. AFM images show
the formation of spherical, monodisperse particles (see Figure 10.2). Due to the low amount
of crosslinker (3 wt-%), the particles spread widely on a solid surface.
Figure 10.2 (a) AFM image of microgel particles with 5 wt-% PB; (b) Cross-section of
particles marked in the left image.
The particle size was analyzed dependent on the temperature and the pH (see Figure 10.3).
PVCL-based microgels shrink with increasing temperature. The Volume Phase Transition
Temperature (VPTT) of the zwitterionic particles is slightly shifted to higher temperatures
(30.7 °C to 31.2 °C for 0 wt-% and 10 wt-% PB, respectively) due to the increased
hydrophilicity that retards the shrinkage of the of the zwitterionic particles. Furthermore, the
temperature range of the phase transition becomes broader with increasing PB content. While
pure PVCL-microgels show a sharp transition, microgels with 10 wt-% PB have a very broad
a
b
132
transition range. This further suggests that the shrinkage of the microgels is retarded by the
incorporation of the hydrophilic PB comonomer.
Figure 10.3 (1) Temperature- and (2) pH-sensitivity of microgels with different PB content.
A change in pH, however, does not lead to a change in particle size. It is expected that the
hydroxyl group of the PB is protonated at low pH while the quaternized ammonium group
leads to an overall positive charge of the molecule. Within the microgel network, these
charges are thought to repel each other leading to a swelling in particle size. At high pH, PB is
neutral leading to shrinkage in particle size. The absence of pH-sensitivity in phosphobetaine
microgels, however, is due to the low acid dissociation constant of diethyl phosphoric acid of
1.39 [18] meaning that PB is able to maintain its zwitterionic nature at a wide range of pH.
(b) Microgels with CaCl2
It was shown that microgels with 5 wt-% PB start to swell from 110 nm in water to 165 nm at
a concentration of 1 M CaCl2 (i.e. by 50 %). Complementary, the saturation limit of the
particles was determined gravimetrically. Therefore, a known amount of freeze-dried
microgel (1 and 5 wt-% PB) is solved in water to which a known amount of CaCl2 is added.
The samples are shaken overnight and centrifuged three times to remove unbound CaCl2.
Afterwards, the samples are freeze-dried and weighed.
Figure 10.4 shows that microgels with a PB content of 1 wt-% reach a saturation limit at a low
ratio of CaCl2:microgel, i.e. 2.6. Samples with 5 wt-% PB, however, reach the saturation limit
at a higher ratio of 15.5. This approximately corresponds to the fivefold amount of 1 wt-% PB
microgels.
133
Figure 10.4 Gravimetrically determined saturation limit of PB microgels with CaCl2.
The interactions between Ca2+ and microgels were analyzed in a more detailed way with ITC.
Here, CaCl2 solution of a known concentration is titrated to a microgel solution of known
concentration. The raw data, i.e. the generated heat over time, is given in Figure 10.5.
Figure 10.5 ITC raw data for titration of CaCl2 to microgel solution containing 1, 5 or 10 wt-
% PB. The concentration for CaCl2 was 1 g/L in all measurements. Titration was performed at
T = 25 °C.
The graphs show that microgel samples with increasing amount of PB need increasing amount
of CaCl2 to become saturated. The graph for 10 wt-% PB was fitted to calculate the saturation
limit.
With respect to the given concentration and volumes, the ratio between Ca2+ and microgel
gives the binding constant Kb. At the equilibrium
A + B ↔ AB (2)
Kb is defined as:
𝐾𝑏 = [𝐴𝐵]
[𝐴]∙[𝐵] (3)
134
with [A] = concentration of species A that can be bound; [B] = concentration of free binding
sites; [AB] = concentration of occupied binding sites.
At 25 °C, Kb increases with increasing amount of incorporated content of PB. The linear
relationship can be seen in Figure 10.6. Additionally, the measurements were repeated at
37 °C, i.e. above the VPTT of the microgels. It is noticeable that the affinity is about one
order of magnitude lower than at room temperature. This is attributed to the collapse of the
polymer network preventing the uptake of larger amounts of Ca2+.
Figure 10.6 Binding constants Kb of binding of Ca2+ to zwitterionic microgels with varying
PB content as determined from ITC analysis at different temperatures.
(c) Microgels with CaCO3
- Direct synthesis
The synthesis via direct synthesis leads to the precipitation of the microgels forming a white,
ductile disk (see Figure 10.7).
Figure 10.7 Mineralized microgel.
These plates retain the thermosensitivity that is typical for PVCL microgels. They shrink
when heated up to 50 °C and swell again when cooled to room temperature (see Figure 10.8).
The extent of swelling and shrinking is less pronounced when compared to pure microgels
due to the hybrid material’s macroscopic size.
The crystal structure of CaCO3 inside the particles was determined with XRD. For all
samples, independent on PB or CaCO3 content, calcite was found to be the only crystal
135
Figure 10.8 Thermosensitivity of mineralized microgels. (1) 25 °C; (2) 50 °C; (3) 25 °C.
structure according to standard JCPDS files. The reflections at 2Θ = 29.5 (104), 36.1 (110),
and 39.5 (113) confirm the formation of calcite and are in good agreement with the literature.
Figure 10.9 shows the XRD diffraction pattern for the sample D(4)50 as an example. The
assignments of the peaks can be found in the appendix. This result is not surprising as calcite
is the thermodynamically favored modification which exists at ambient temperature and
pressure [19][20][21]. The sharp peaks suggest that CaCO3 was well-crystallized.
Figure 10.9 XRD diffraction pattern of microgel with 4 wt-% PB and 50 wt-% CaCO3 via
direct synthesis. Black lines indicate the peak positions for calcite as found in literature
(JCPDS: 86-2334).
Another means to determine the crystal structure of CaCO3 is FTIR (see Figure 10.10).
Calcite, vaterite, and aragonite show different peaks in the FTIR spectrum. The peaks found
in Figure 10.10 agree with the data obtained with XRD, namely that only calcite is present in
the sample. The peaks belonging to calcite are highlighted in the spectrum above. The peaks
at 1420 cm-1, 875 cm-1, and 711 cm-1 correspond to the asymmetric stretching, out-of-plane
bending, and in-plane-bending of the carbonate group, respectively [22]. The microgel has a
characteristic peak at 1620 cm-1, induced by the carbonyl group of PVCL.
Apart from the crystal structure, an increase in the peak intensity of calcite can be observed.
Thus, it can be qualitatively, though not quantitatively, stated that the amount of CaCO3 is
increasing with increasing theoretical amount.
1 2 3
136
Furthermore, the form and morphology of the hybrid microgels was examined with
microscopy.
Figure 10.10 (a) FTIR spectrum of microgels with 10 wt-% PB and varying CaCO3 content,
synthesized via direct synthesis. (Red) Pure CaCO3 (calcite); (green) Unmineralized microgel
with 10 wt-% PB; (black) Microgel with 13 wt-% CaCO3; (blue) Microgel with 37 wt-%
CaCO3; (turquois) Microgel with 67 wt-% CaCO3. (b) FESEM image of 10 wt-% PB
microgels with 28 wt-% of CaCO3.
The images in Figure 10.10 show large, trigonal-rhombohedral crystals (ca. 4 µm) that are
covered completely or partly by a microgel film. The sample D(10)28 is characteristic for all
samples obtained with direct synthesis. They clearly indicate that the microgel structure has
no influence on the crystal growth. It can be assumed that the direct synthesis is too fast to
enable a controlled biomineralization process and CaCO3 crystals grow randomly with no size
control. Thus, no further characterization of these hybrid particles was performed.
Therefore, a different route to obtain CaCO3 crystals whose sizes are controlled by the
microgel network is followed. Since a fast reaction between CaCl2 and (NH4)2CO3 leads to
large crystals of about 3 – 5 µm, a slow reaction via gas diffusion might enable the microgel
network to influence the crystal formation process.
- Gas Phase Diffusion
According to N. Gehrke, the slow crystallization process via gas diffusion leads to well-
facetted calcite crystals [23]. In the desiccator, (NH4)2CO3 slowly disintegrates:
(NH4)2CO3 ↔ 2 NH3 + CO2 + H2O (4)
a b
137
The synthesis of CaCO3 in PB microgels via gas phase diffusion in a desiccator resulted in
white disk that retain the microgel’s thermosensitivity identical to the results shown in Figure
10.7 and Figure 10.8.
XRD and FTIR measurements reveal that calcite is the only crystal structure obtained with
gas diffusion. Changes in microgel concentration, temperature or pH lead to the same results.
All XRD diffraction patterns as well as the FTIR spectra can be found in the appendix.
FESEM images show the location, size and form of the formed calcite crystals. Figure 10.11
(1) shows 10 wt-% PB microgels with 67 wt-% CaCO3. The sample was dialyzed before it
was put in the desiccator to react with (NH4)2CO3 in order to remove all Ca2+ that is unbound
by the phosphogroup inside the microgel. The image reveals that small CaCO3 minerals were
formed within the microgel structure. These crystals have sizes between 20 and 60 nm. The
microgel structure remains spherical, but disrupted. (2) shows a sample that was not dialyzed
before being put in the desiccator. Here, the image resembles the images taken after the direct
synthesis. Large, trigonal-rhombohedral crystals were formed that exceed the size of
microgels (size (CaCO3) > 2 µm) so that it can be assumed that the microgel network had no
influence on the mineral growth. In the next step, the influence of PB amount on the crystal
growth was studied. (3) shows hybrid particles with less CaCO3 compared to image (1). In a
“broken” microgel particle, small crystals with a size of ~ 30 nm can be seen (red arrow). It
seems that a lower content of PB does not lead to a deformation of the particle in the same
extent as can be observed for microgels with 67 wt-% PB. Also, less crystals with a smaller
size are formed. EDX measurements confirm that less Ca is present in the sample E(10)37
compared to E(10)67 as indicated by the increased Ca/C ratio (Figure 10.12). (4) shows the
biomineralization in microgels without PB. Here, the crystals and completely covered in a
layer of microgels and thus resemble the crystals obtained in (2). The microgel obviously has
no influence on the mineral growth. This indicates that the presence of zwitterionic PB plays a
crucial role in the controlled growth of CaCO3. Finally, (5) and (6) show the influence of pH
on the biomineralization process. (6) shows the results if the microgel/CaCl2 solution was
acidified after the pre-dialysis (pH 3). It is obvious that no mineral was formed since only
spherical, monodisperse microgel particles can be seen. This is expected since CaCO3
decomposes in acidic medium as follows:
CaCO3 + 2 HCl → CaCl2 + H2O + CO2 (5)
(5) shows that crystals inside the microgel structure were also formed under basic conditions
(pH 10). Compared to (1), (pH 8.5), the crystals are smaller in size (~ 30 nm) and more
monodisperse.
138
Figure 10.11 FESEM images. (1) E(10)67, pre-dialyzed before placing into desiccator, pH
8.5; (2) Same sample as in (1), undialyzed; (3) E(10)13, pre-dialyzed; (4) E(0)67; (5)
E(10)67, pH 10; (6) E(10)67, pH 3.
To sum up, the best results for the formation of CaCO3 were obtained if the following
requisitions were met:
(1) The presence of PB is crucial for a controlled mineralization within the microgel
network. The polar phosphogroup effectively immobilizes Ca2+ leading to a
supersaturation of ions that is essential for mineralization. Furthermore, it can be
assumed that due to the anti-polyelectrolyte effect, diffusion of Ca2+ ions into the inner
microgel network is facilitated.
(2) The sample has to be dialyzed before being put in the desiccator to remove all
unbound Ca2+.
139
(3) The pH of the solution has to be pH > 8.5. At low pH, CaCO3 decomposes to CaCl2.
Figure 10.12 EDX analysis of (a) E(10)13 and (b) E(10)67. Traces of Al, P, and Cl derive
from the sample holder.
Figure 10.13 TEM images of (a) 0 wt-% PB; (b) 5 wt-% PB; (c) 10 wt-% PB. Each sample
contains 67 wt-% CaCO3. The images in the second row (‘) are a zoom-in of the respective
image in the first row.
The influence of the PB content on the crystal size and amount is further studied with TEM
(see Figure 10.13). The images show that the mineralization of CaCO3 in uncharged PVCL
microgels leads to a crystal growth on the surface of the particle. Contrary, zwitterionic
microgels induced a crystal growth within the microgel network. A higher content of PB leads
to an increased amount of crystals while the spherical shape of the microgel is still
maintained. The ability of the zwitterionic microgels to highly swell after the addition of Ca2+
supports the internal growth of CaCO3.
a b
200 nm
500 nm 500 nm 200 nm
200 nm 100 nm
a b c
b‘ a‘
1
c‘
140
A quantitative conclusion about the incorporated amount of CaCO3 can be drawn from
thermogravimetric analysis. Here, the dried sample is heated up to 900 °C. At temperatures
around 450 °C, the microgel is decomposed. CaCO3 is decomposed at around 750 °C
according to the following equation:
CaCO3 → CaO + CO2 (6)
Looking at the remaining mass after the microgel’s decomposition (600 °C, see black line in
Figure 10.14), one can read the amount of CaCO3 in the sample. The amounts obtained from
TGA measurements are in good agreement with theoretical values.
Further TGA measurements were performed for hybrid particles that were synthesized at
different temperatures as well as pH (see Figure 10.15). (1) shows that ~ 14 % less CaCO3 are
formed within the microgel particle at 40 °C compared to 25 °C. At 40 °C, the microgels are
collapsed, thus hindering a free diffusion of Ca2+ within the microgel network. These results
are in agreements with ITC data (see Figure 10.5).
Figure 10.14 Decomposition of 10 wt-% PB microgels with different CaCO3 content. The
black vertical line shows the point where the remaining amount of CaCO3 is determined
(600 °C).
Figure 10.15b demonstrates that the highest amount of CaCO3 crystals is gained at pH 8.5 At
pH 10, ~ 11 % less CaCO3 is formed, while at pH 3 no CaCO3 is formed at all. The remaining
4.9 % seen in the graph agree with the remains of pure PB microgels at 600 °C (see (a)).
141
f Figure 10.15 Decomposition of 10 wt-% PB microgels. (a) With 37 % CaCO3, synthesized in
desiccator at different temperatures. (b) With 67 % CaCO3, synthesized at different pH. The
black vertical line shows the point where the remaining amount of CaCO3 is determined
(600 °C).
Figure 10.7 suggest that the hybrid particles retain their temperature sensitivity even after
mineralization. This was studied in more detail with DLS. Figure 10.16 shows that pure PB
microgels have a VPTT of 30.9 °C. The respective crystallized sample as well shrinks with
increasing temperature. The constant decrease in size beginning at 15 °C makes it difficult to
give an exact VPTT for the hybrid particles. Interestingly, the hybrid particles have roughly
the same size in collapsed state than the unmineralized particles.
Figure 10.16 Temperature-sensitivity of pure 2 wt-% PB microgels (red line) and sample
E(2)13v (black line).
The obvious temperature-sensitivity of the hybrid microgels leads to the assumption that
temperature can be used as a trigger to release CaCO3 crystals “on demand”. Therefore, the
sample E(2)13v was heated up to 50 °C and analyzed. TEM measurements (Figure 10.17)
reveal small particles of a size of 59 ± 8 nm and larger particles of a size of 168 ± 14 nm.
Since the smaller particle did not appear in images taken at room temperature, they are
assumed to be CaCO3 crystals that are “squeezed” out of the microgel particles. The crystals
a b
142
are relatively monodisperse and spherical (a). The microgels as well retain their spherical
structure after the removal of CaCO3 (b).
Figure 10.17 TEM images of E(10)67 after heating up to 50 °C. (b) is a zoom-in of (a).
The sample was fractionated with F-FFF to separate particles with different sizes (Figure
10.18). Furthermore, this method gives the radius of gyration, RG.
Figure 10.18 Fractionation of sample E(10)67 at (a) 25 °C and (b) 50 °C, the arrow marks the
peaks generated by small particles. The left-most peak represents the void peak.
The measurement at 25 °C (Figure 10.18a) reveals no distinct, but a broad and undefined
peak. The presence of CaCO3 in the microgel particle leads to a disruption of the polymer
network and thus high polydispersity as was already seen in TEM (Figure 10.13). Contrary, at
50 °C, two separate peaks can be seen (Figure 10.18b). The first peak has its maximum after
9 min corresponding to an RG of ~ 60 nm. This size is identic with the value obtained from
TEM measurements. The main peak has its maximum at t = 23 min corresponding to an RG of
140 nm. This corresponds to the value obtained from DLS. The ratio between RH and RG
gives information about the hardness of microgels. W. Burchard specifies the value of
microgels with 0.3 to 0.6 [24]. The examined sample has a ratio of 0.74 that is close to the
value of a hard sphere (0.775 [25]). This surprisingly high value can be explained by the
collapsed state of the microgel particle at 50 °C which is intensified by the electrostatic
attraction of positive and negative charges that lead to a further “hardening” of the polymer
network.
a b
143
10.1.4 Summary
This chapter elaborated the important parameter necessary for successful biomineralization of
CaCO3 in zwitterionic microgels based on PVCL. The most convenient method was slow gas
diffusion. At pH above 8.5 and room temperarure, crystals with sizes of ~ 60 nm could be
detected inside the microgel particle. This required that the microgel sample in aqueous CaCl2
solution was dialyzed before being put in a desiccator in order to remove all unbound Ca2+.
The crystal structure was found to be calcite and confirmed with XRD and FTIR. TGA
measurements showed that the CaCO3 amount was found to be in accordance with the
theoretical values.
Since the hybrid particles retain their thermosensitivity, temperature was used as a trigger to
“squeeze” CaCO3 crystals out of the microgel particles. Both TEM and F-FFF confirmed that
it is possible to separate microgels and CaCO3 at temepratures above the VPTT. This offers
possible applications of this hybrid material as CaCO3 carrier where the CaCO3 can be
released on demand.
144
10.1.5 Literature
[1] F. C. Meldrum, Int. Mater. Rev. 2003, 48, 187–224.
[2] H. Li, L. a Estroff, J. Am. Chem. Soc. 2007, 129, 5480–3.
[3] R. B. Frankel, Rev. Mineral. Geochemistry 2003, 54, 95–114.
[4] D. B. Deoliveira, R. A. Laursen, J. Am. Chem. Soc. 1997, 119, 10627–10631.
[5] D. L. Masica, S. B. Schrier, E. A. Specht, J. J. Gray, J. Am. Chem. Soc. 2010, 132,
12252–12262.
[6] J. J. J. M. Donners, E. W. Meijer, R. J. M. Nolte, C. Roman, A. P. H. J. Schenning, N.
A. J. M. Sommerdijk, B. R. Heywood, Chem. Commun. 2000, 1937–1938.
[7] D.-K. Keum, K. Naka, Y. Chujo, Bull. Chem. Soc. Jpn. 2003, 76, 1687–1691.
[8] K. Furuichi, Y. Oaki, H. Ichimiya, J. Komotori, H. Imai, Sci. Technol. Adv. Mater.
2006, 7, 219–225.
[9] O. Grassmann, P. Löbmann, Chem. Eur. J. 2003, 9, 1310–1316.
[10] P. Liu, J. Song, Biomaterials 2013, 34, 2442–54.
[11] F. C. Meldrum, H. Cölfen, Chem. Rev. 2008, 108, 4332–432.
[12] A.-W. Xu, M. Antonietti, H. Cölfen, Y.-P. Fang, Adv. Funct. Mater. 2006, 16, 903–
908.
[13] V. Wintgens, F. Dalmas, B. Sébille, C. Amiel, Carbohydr. Polym. 2013, 98, 896–904.
[14] F. Degryse, B. Ajiboye, R. D. Armstrong, M. J. McLaughlin, Soil Sci. Soc. Am. J.
2013, 77, 2050.
[15] J. Ramos, A. Imaz, J. Forcada, Polym. Chem. 2012, 3, 852.
[16] A. Pich, A. Tessier, V. Boyko, Y. Lu, H. P. Adler, Macromolecules 2006, 7701–7707.
[17] L. Nong, C. Xiao, Z. Zhong, J. Surfactants Deterg. 2011, 14, 433–438.
[18] W. D. Kumler, J. J. Eiler, J. Am. Chem. Soc. 1943, 65, 2355–2361.
[19] X. Wang, H. Sun, Y. Xia, C. Chen, H. Xu, H. Shan, J. R. Lu, J. Colloid Interface Sci.
2009, 332, 96–103.
[20] X. Li, Q. Shen, Y. Su, F. Tian, Y. Zhao, D. Wang, Cryst. Growth Des. 2009, 9, 3–9.
[21] C. Cheng, Z. Shao, F. Vollrath, Adv. Funct. Mater. 2008, 18, 2172–2179.
[22] T. Z. Forbes, a. V. Radha, A. Navrotsky, Geochim. Cosmochim. Acta 2011, 75, 7893–
7905.
[23] N. Gehrke, H. Cölfen, N. Pinna, M. Antonietti, N. Nassif, Cryst. Growth Des. 2005, 5,
1317–1319.
[24] W. Burchard, Light Scattering from Polymers, Springer Berlin Heidelberg, Berlin,
Heidelberg, 1983.
[25] W. Burchard, K. Kajiwara, M. Chemie, A. Ludwigs, J. Polym. Sci. Part B Polym.
Phys. 1982, 20, 157–171.
145
10.2 Microgels and β-TCP as Bone Substitute Material
10.2.1 Introduction
Materials that are used for bone repair are substances that support, promote or stimulate the
healing process of bone either on their own or in combination with other materials. Ideally, it
will also involve the surrounding soft tissue to stimulate the differentiation of osteogenic stem
cells. For this, the material has to be as similar as possible to natural bone in its properties [1].
For the synthetic fabrication of such material the following parameters must be addressed:
biocompatibility, bioresorbability, surface texture (influencing the response of the cell’s, bone
and surrounding tissue), geometry, the presence of pores with defined size and volume, facile
flow and incorporation of oxygen, nutrients or drugs, and chemical purity [2][3].
Synthetic inorganic material
Calcium phosphates are promising materials for biomedical application and the fabrication of
biomaterials. For instance, in the field of bioceramics, hydroxyapatite (HA, (Ca10(PO4)6(OH)2,
Ca/P ratio 1.67) and tricalcium phosphate are used as possible materials for bone substitution,
fixation of implant, and tooth remineralization due to their similarity to human bone apatite
(Ca/P ratio ~ 1.6, depending on the type and age of the bone) [4][5]. β-tricalcium phosphate (β-
TCP, Ca3(PO4)2, Ca/P ratio 1.5) is a polymorph form of tricalcium phosphate and is
characterized by its high biocompatibility with surrounding living tissue, high solubility, and
bioactivity [3][6][7][8]. It is also osteoconductive. In contrast to HA, it decomposes in the human
body and releases a high concentration of Ca2+ and PO4- ions [2]. Ca2+ is essential blood
clotting [9], stabilization of cell membranes [10], and muscle contraction [11]. Chang et al. could
show that an increase in the local Ca2+ and PO4- concentration supports the mineralization
process [12]. Already in 1977, Cameron et al. used β-TCP as bone implantation material [8].
Depending on the temperature during the sintering process, both α- and β-TCP can be
produced working at 1370 °C and 1200 °C, respectively [13]. α- and β-TCP differ in
mechanical strength and their solubility in water. The latter is important for the decomposition
of the ceramic.
A big difference between biologically and synthetically fabricated β-TCP is that the latter
does not possess interconnecting pores that are essential the success of osteoconductive
treatment since they facilitate tissue infiltration. For those, the ceramic has to be submitted to
special treatments. F. Zhang et al. for instance produced macroporous β-TCP that they used
for bone regeneration [14]. They were able to tune the porosity from 55 to 70 % with pore sizes
between 300 and 500 µm using spark plasma sintering that allows the fast production of solid
146
nanostructured ceramics. In their review that illuminates the connection between porosity,
pore sizes and bone regeneration, V. Karageorgiou and D. Kaplan state that the ideal pore size
is considered to be ~ 300 µm, but concede that the minimum pore size is ~ 100 µm to
regenerate new bone [15].
Pure β-TCP as bone substitute material is not well suited due to its gradual, still fast and
unpredictable resorption [3]. A decomposition that is too fast is both detrimental to bone
formation [16] and may lead to inflammation due to high concentration of ceramic particles [2].
In acid environment, β-TCP degrades as follows:
Ca3(PO4)2 𝐻+
→ Ca2+ + 2 CaHPO4 𝐻+
→ 3 Ca2+ + 2 PO43- (1)
An acid environment is generated by H+ secreting macrophages that are the first cells that
cover a new ceramic implantation site. Calcium as the degradation product is either excreted
or incorporated in the body. K. Lin, however, showed that it is possible to slow down the
degradation process through control of the powder size. The authors used nano-sized β-TCP
instead of micro-sized powder [17]. A second drawback of pure β-TCP is its low mechanical
strength [1] and elasticity and its high brittleness [18], it is therefore often used in composites.
Composites
Due to the manifold combination possibilities between inorganic and organic material, a
classification of composites is difficult. Literature offers countless examples for composite
materials.
Pure inorganic materials such as HA and β-TCP have low mechanical strength and low
toughness, so composite materials, i.e. the blending of advantageous properties of both
inorganic and organic materials, are studied to develop new materials suitable for bone
substitution and regeneration. A number of researchers mix β-TCP with bioresorbable and
biocompatible polymers that act as a binder [19]. Crosslinkers are further added for two
reasons: to increase the persistence in tissue and to improve mechanical properties. C.-H. Yao
et al. for instance used gelatin as a binder and glutaraldehyde as a crosslinker. They found out
that their composite materials need at least four days of soaking to improve biocompatibility
[20]. In a later study, the authors substituted glutaraldehyde with genipin, a crosslinker that is
bound in gardenia and shows very low toxicity [19][21][22]. They obtained a formable filling
material that causes no inflammation and leads to and participates in the formation of new
bone. C. Zou et al. used the natural polymer collagen for the fabrication of porous β-
TCP/collagen composites [23]. They were able to achieve a porosity of 95 % with pore sizes
of 50-100 µm. The authors reported that the bone was fully regenerated after 12 weeks in
147
animal testing. Z.-J. Zhu et al. prepared β-TCP/ polylactide composites that showed promising
results [24]. The authors used this combination to prevent early degradation through the
neutralization of alkaline degradation products of β-TCP and the acidic degradation product
of polylactide. Kikuchi et al. added poly(L-lactide-co-glycolide-co-ε-caprolactone) to β-TCP
to stabilize the pH of 6-7 during the acidic degradation of the polyester through the
dissolution of β-TCP [25]. The pH of pure polymer solutions, in contrast, decreased to pH 3
after the degradation to glycolic and L-lactic acid.
So far, only few examples of composites that combine β-TCP with microgels or microspheres
exist in literature. S.M. Kuo prepared chitosan microspheres of roughly 200 – 400 µm in
diameter and used them to encapsulate β-TCP in the interior [26]. Chitosan is an antibacterial
natural polymer that is able to facilitate wound healing. Many studies have been performed to
examine its release properties, primarily of drugs [27][28][29]. M. Kucharska et al. pursued this
idea and demonstrated the potential of β-TCP/chitosan microspheres as possible bone
scaffolds [30]. They obtained a 3D structure through the acidic dissolution and consequent
agglomeration of the chitosan spheres. The material had pores with sizes of ~ 280 µm for the
highest β-TCP concentration (i.e. 10 wt-%), but the porosity of 20 – 25 % failed to copy the
typical porosity of bone (20 – 90 %). Y. Wang prepared microspheres of poly(lactid acid) and
added β-TCP to increase the biocompatibility [31]. The spheres had a higher pore volume
compared to pure poly(lactic acid) spheres, but did not improve the compressive modulus.
In this chapter, PVCL-based microgels were chosen as a binder for β-TCP because of their
high elasticity and water adsorption capacity. In water, they are thought to bind β-TCP
particles through intermolecular and interparticular hydrogen bondings. The size and the
content of microgels in the mixture are likely to influence the porosity of the composite.
Results discussed in this chapter were already published in Biomed. Tech. 2015, DOI:
10.1515/bmt-2014-0141, pii: /j/bmte.ahead-of-print/bmt-2014-0141/bmt-2014-0141.xml.
148
10.2.2 Synthesis Procedure
Microgel synthesis Microgels were prepared via precipitation polymerization in water.
Appropriate amounts of VCL, Vim, BIS, and AAEM (see Table 10.3) were mixed in 500 mL
of water and heated up to 70 °C while stirring at 200 rpm under nitrogen atmosphere. After
stirring for 60 min, the initiator AMPA (0.16 g, 0.59 mmol) was added. The reaction was
allowed to carry on for 4 h. After the synthesis, the samples were purified with dialysis for 3 d
with a composite regenerated cellulose membrane from Millipore (NMWCO 30,000). Four
different samples with varying amount of Vim were prepared.
Table 10.3 Amounts of monomers used for the microgel synthesis.
Amount of monomer / g
Sample VCL Vim BIS AAEM
0 % Vim 6.000 - 0.240 1.100
5 % Vim 6.000 0.367 0.240 1.100
Preparation of ß-TCP Phase pure ß-TCP was prepared by calcination using a calcium
deficient hydroxyapatite with 3 wt% calcium hydrogen phosphate (CHP) at 1000 °C for 2 h.
CHP was added to adjust the Ca/P ratio to exactly 1.5. Chemical and phase purity was
confirmed by X-ray fluorescence (XRF) spectroscopy and X-ray diffraction (XRD)
spectroscopy.
Preparation of microgel- ß-TCP composite material Various concentrations of microgel
dispersions were mixed with an appropriate amount of ß-TCP and either directly freeze-dried
or spray-dried (Mini Spray Dryer B-290, Büchi) at 220 °C to yield granulates with a microgel
content of 10 or 5 wt%. Before spray-drying, the suspensions were milled on a roller platform
with zirconia milling balls (in each case 1 kg of 1 cm and 2 cm in diameter) for 1 day. After
the addition of 1 to 2 drops CONTRASPUM CONC., the suspensions were sieved with a
125 µm sieve.
149
10.2.3 Results and Discussion
Microgels with 0 and 5 mol-% Vim (henceforth called PVCL-Vim) were synthesized. AFM
images show the formation of monodisperse, spherical particles (see Figure 10.19). The
microgel suspensions were mixed with ß-TCP to obtain composite materials with 5, 10, and
15 wt-% microgel content (henceforth called e.g. C10-5% meaning the composite containing
10 wt-% microgel in the suspension with a Vim content of 5 wt-%).
Figure 10.19 (a) AFM image of 5 wt-% Vim microgel; (b) FTIR spectrum of PVCL-Vim
microgels, pure ß-TCP, and the composite C10-5%.
The interaction between the microgel network and β-TCP was investigated with FTIR.
Therefore, FTIR spectra of PVCL-Vim microgels, pure β-TCP, and the composite C10-5%
were recorded (see Figure 10.19). Prominent peaks for β-TCP are at ~ 900 – 1200 cm-1 (PO43-
stretching mode) and 553 and 606 cm-1 (both PO43-
vibration mode). PVCL-Vim microgels show
characteristic peaks at 2921 cm-1 (C-Hring stretching mode), 2856 cm-1 (C-Hstretching), 1633 cm-1
(amide I), 1485 cm-1 (C-Nstretching), 1437 cm-1 and 1418 cm-1 (both C-Hdeformation bands), 1264
cm-1 (C-Hring in-plane bending and C-Nring stretching mode), 1230 cm-1 (C-Hring in-plane bending and C-Nring
stretching mode), and 975 cm-1 (ring deformation mode) [32]. For the amide I band of PVCL, a shift
from 1633 cm-1 to 1618 cm-1 can be detected in the composite spectrum. This indicates that
the C=O group of PVCL interacts with the water molecules on the surface of β-TCP. Further
possible interactions between Vim and β-TCP cannot be detected due to the low concentration
of Vim in the sample. It is likely, however, that the electron pair of nitrogen in Vim interacts
with Ca2+ (see Figure 10.20).
To confirm this data, electrophoretic measurements were performed (see Figure 10.21a).
Uncharged PVCL microgels show no pH dependency. A slight negative charge is caused by
the initiator and results in the colloidal stabilization of the particles. PVCL-Vim microgels
a b
150
with 5 % Vim exhibit a positive charge over a broad range of pH due to the protonation of
Vim. The particles become neutral at the isoelectric point of ~ 9.5. β-TCP is neutral at pH < 5
Figure 10.20 Interactions between PVCL-Vim microgel and β-TCP.
and becomes negative with increasing pH. Measurements below pH 5 were not possible due
to aggregation of β-TCP. A mixture of 5 wt-% microgel in the suspension containing 5 wt-%
Vim has a pH of ~ 7. Thus, at pH 7, a strong interaction between positively charged microgels
and negatively charged β-TCP can be assumed in the composite.
Figure 10.21 (a) Electrophoretic mobility with varying pH. Blue curve: 5 wt-% Vim
microgel; black curve: uncharged PVCL microgel (reference); green curve: pure β-TCP. (b)
XPS spectra of pure β-TCP (green curve), 5 wt-% Vim microgel (blue curve), and composite
material C10-5% (red curve).
XPS measurements reveal the chemical composition of the composite (Figure 10.21b). The
microgel sample with 5 wt-% Vim shows signals at 282 eV (C 1s), 396.5 eV (N 1s) and 528.5
eV (O 1s). Signals in the XPS spectrum of β-TCP at 130 eV and 344 eV were assigned to Ca
2p and P 2p, respectively. The spectrum of the composite C10-5% exhibits signals that are a
combination of both. This further supports the assumption of a successful incorporation of the
microgel into the composite.
a b
151
An important property of a composite that is a possible substitute for bone material is a high
and fast adsorption of water while at the same time it has to be ensured that the material stays
stable and does not decompose. Therefore, the water uptake through the increase in the
sample weight as a function of environmental humidity was investigated using a humidity-
controlled thermomechanical analyzer. The shape of the resulting adsorption isotherms gives
useful knowledge of the microstructure of materials [33].
Figure 10.22 Water sorption for (a) 0 wt-% PVCL-Vim microgel (red curves) compared to
composite C10-0% (blue curves); (b) 5 wt-% PVCL-Vim microgel (red curves) compared to
composite C10-0% (blue curves). Graphs show adsorption and desorption isotherms; (c)
water sorption in porous materials.
Figure 10.22 gives the adsorption and desorption isotherms for pure β-TCP, 0 and 5 wt-%
PVCL-Vim microgels, and their respective composites C10-0% and C10-5%. Measurements
were performed at 37 °C to mimic body temperature. Pure 0 and 5 wt-% PVCL-Vim
microgels show a high water uptake of 32 and 34 %, respectively. The amount is slightly
higher for microgels containing Vim due to the increased hydrophilicity. In the respective
composites, the water sorption is distinctively lower, it decreased to 4 % for both 0 and 5 wt-
a b
c
152
% PVCL-Vim microgels. The isotherms of pure β-TCP are given as a reference to illustrate
the poor water sorption of the material.
For many industrial materials, S-shaped (type II) isotherms are measured (see Figure 10.22c)
[33]. These curves give information about the type of bound water in the porous material. At
low pressures, water is tightly bound by the material. Between the first and second inflection
point, water is loosely bound, while at high pressures, the water molecules are able to move
freely.
In the graphs measured for the microgels and the composites, no sigmoidal curve can be
observed. Instead, only two areas can be distinguished that resemble the areas characteristic
for loosely bound and free water. This can be explained by the lyophilization of the microgel
before the measurement. Through handling at ambient conditions, the microgel adsorbs water
from the air, thus tightly bound water is already present prior to each measurement. If further
water is tightly bound in the microgel network at the beginning of each measurement, the
amount is assumed to be so small that the first inflection is not or only barely visible.
Furthermore, the BET surface area can be calculated from these measurements. The BET
surface area was developed by Brunauer, Emmett, and Teller in 1938 [34] and is an important
factor since the surface area of the pores influences for instance the rate of biodegradation and
chemical reactions [33]. The prerequisite for this is that at low pressures, i.e. p/p0 = 0.05 – 0.35
(with p = pressure at equilibrium; p0 = vapor pressure at saturation), the sorption is strictly
reversible meaning a free movement of the adsorbate on the surface of the adsorbent. Further,
it has to be ensured that no chemical reaction takes place.
The theory is based on the assumption that the adsorption of e.g. gas or liquids on a surface
takes place beginning with a single layer of adsorbate molecules advancing to multiple layers
of adsorbate at saturation [33]. The measurement of grams of absorbate molecules that form a
monolayer, called the monolayer capacity, allows the determination of the number of
molecules and thus the surface area. The actual value of the surface area is dependent on the
type of molecule used for the measurement. For instance, the water molecule is smaller in size
than a nitrogen molecule, therefore water is better able to penetrate into pores and the
resulting surface area is bigger.
The ratio p/p0 can be obtained by the division of RH (in %) by 100%. The change in weight
wt-% is converted into the adsorbed water volume V as follows:
𝑉 = 𝑡𝑜𝑡𝑎𝑙 𝑤𝑒𝑖𝑔ℎ𝑡−𝑑𝑟𝑦 𝑚𝑎𝑠𝑠
𝑑𝑟𝑦 𝑚𝑎𝑠𝑠 (1)
153
The total weight is determined at each point of RH, while the dry mass was determined after
lyophilization.
Plotting p/p0 against (p/p0) / V(1-p/p0), the intercept and the gradient of the resulting graphs
can be used to calculate the BET surface area (see Figure 10.23):
𝐵𝐸𝑇 =(
1
𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡+𝑔𝑟𝑎𝑑𝑖𝑒𝑛𝑡)
𝑀𝑊∗ 𝑁 ∗ 𝐴 (2)
with MW = MW of adsorbate (i.e. H2O, 18.01 g/mol); N = Avogadro’s number ( 6.022 * 1023);
A = molecular area adsorbate (10.78 Ų for water [35]). The BET surface areas are given in
Table 10.4.
Figure 10.23 Graphs for the calculation of the BET surface area.
Table 10.4 BET surface areas for microgel samples and composite material.
Sample 0 wt-% Vim µ-gel 5 wt-% Vim µ-gel C10-0% C10-5%
BET [m²/g] 163 204 18.5 18.2
Pure microgels have an approximately tenfold higher surface area than the composite
material. BET surface areas values for pure β-TCP are given in literature as 1-2 m²/g [6][36].
Therefore, the incorporation of microgels into the composite material greatly influences the
surface area values and changes the textural properties of β-TCP.
The water sorptions for the microgels and their respective composite materials as a function
of time are given in Figure 10.24. The sorption rates for all samples decreased with increasing
humidity, i.e. a longer time was needed to reach equilibrium. Microgel samples containing
5 wt-% Vim have a longer sorption time than uncharged microgels due to the higher
hydrophilicity of the positively charged microgels. Furthermore, in accordance with Figure
10.22, it can be seen that pure microgel samples take up higher amounts of water than their
respective composite material.
154
Figure 10.24 Sorption rates of pure β-TCP, microgel samples and their respective composite.
The thermochemical stability of the microgels as well as the composites was analyzed (see
Figure 10.25). The graph for pure β-TCP is shown as reference to demonstrate that β-TCP is
stable until temperatures to 850 °C. For pure microgels, the first weight loss occurs at around
100 °C resulting from the loss of water. In the range between 150 and 350 °C, acrylic groups
are decomposed. The third step at 441 °C is due to the decomposition of the polymer
backbone.
Figure 10.25 Thermogravimetrical analysis of 5 wt-% PVCL-Vim microgel, pure β-TCP and
composite material with various microgel content in suspension.
The composite materials show the same decomposition behavior. The polymer decomposition
is shifted slightly to 450 °C indicating an improved thermal stability. At 600 °C, 96.4 %,
90.7 %, and 87.8 % of remaining β-TCP were recorded for the samples C5-5%, C10-5%, and
C15-5%, respectively. This data further indicates that the desired amount of microgel was
assured in the suspension.
Further testing whether the composites are suitable material for bone tissue repairing, 3 D
powderbed printing is a convenient tool to create samples that can be tested regarding their
155
mechanical strength. To ensure that the composite material is applicable for 3 D powderbed
printing, knowledge about the morphology and particle size distribution is essential.
Figure 10.26 SEM images of composites (a) C10-5%; (b) C5-5%; and (c) pure β-TCP; all
after spray-drying.
Figure 10.26 shows that the composites consist of granulates (a,b). The granulate size of
composite C5-5% is bigger compared to C10-5%. The spherical form of granulates as well as
the grain size (< 1µm) makes it difficult to distinguish between β-TCP and PVCL-Vim
microgels. β-TCP on the other hand forms only few granulates, while the main material
consists of single grains with an average size of 1.46 µm (c). This supports the assumption
that microgels act as glue for the grains to form granulates. The printed sample C10-5% can
be seen in Figure 10.27.
Figure 10.27 Printed sample C10-5%.
156
10.2.4 Summary
This work aimed at the preparation and characterization of β-TCP/microgel composites. For
this, cationic PVCL/AAEM/Vim microgels were used because of the possible interactions
that arise from the carboxylic group of PVCL as well as the free electron pair of the nitrogen
of Vim and Ca2+ and PO43- in β-TCP. The microgels were further chosen as binding material
between β-TCP particles because of their high water content. This could be proven with FTIR
spectroscopy and TEM. TGA measurements confirmed that the sintering process pyrolyzes
the microgels and leaves pure β-TCP. It was shown, however, that the presence of microgels
facilitates the pressing of the composites and the consequent spray-drying process that was
not discussed in this chapter. Thermoanalytical measurements demonstrated that the
composite material absorbs a higher amount of water due to the excellent water absorption
capability of microgels. The BET surface, i.e. the surface area of pores inside the material, is
also increased by a factor of 10 in comparison to pure β-TCP. A poor stability of the material
in water, however, suggests further improvement of the material with regard to possible
application as bone scaffolds. In future, the microgels are also applicable to be loaded with
osteogenic or osteoinductive drugs to further enhance bone regeneration.
157
10.2.5 Literature
[1] C. Rentsch, B. Rentsch, D. Scharnweber, H. Zwipp, S. Rammelt, Unfallchirurg 2012,
115, 938–49.
[2] K. A. Hing, Philos. Trans. A. Math. Phys. Eng. Sci. 2004, 362, 2821–50.
[3] J. M. Rueger, Orthopade 1998, 27, 72–79.
[4] N. H. A. Camargo, S. A. de Lima, Enori Gemelli, Am. J. Biomed. Eng. 2012, 2, 41–47.
[5] S. M. Schnürer, U. Gopp, K.-D. Kühn, S. J. Breusch, Orthopade 2003, 32, 2–10.
[6] K. Adamska, M. Woźniak, A. Voelkel, Ann. UMCS, Chem. 2010, 65, DOI
10.2478/v10063-010-0004-4.
[7] K.-W. Wang, L.-Z. Zhou, Y. Sun, G.-J. Wu, H.-C. Gu, Y.-R. Duan, F. Chen, Y.-J. Zhu,
J. Mater. Chem. 2010, 20, 1161.
[8] H. U. Cameron, I. Macnab, R. M. Pilliar, J. Biomed. Mater. Res. 1977, 11, 179–186.
[9] E. Kaufmann, Klin. Wochenzeitschrift 1926, 11, 453–455.
[10] M. R. Hadi, N. Karimi, J. Plant Nutr. 2012, 35, 2037–2054.
[11] L. J. Clelland, B. M. Browne, S. M. Alvarez, A. S. Miner, P. H. Ratz, J. Muscle Res.
Cell Motil. 2011, 32, 77–88.
[12] Y. L. Chang, C. M. Stanford, J. C. Keller, J. Biomed. Mater. Res. 2000, 52, 270–8.
[13] B. Mehdikhani, B. Mirhadi, N. Askari, J. Ceram. Process. Res. 2012, 13, 486–490.
[14] F. Zhang, K. Lin, J. Chang, J. Lu, C. Ning, J. Eur. Ceram. Soc. 2008, 28, 539–545.
[15] V. Karageorgiou, D. Kaplan, Biomaterials 2005, 26, 5474–5491.
[16] A. Jamali, A. Hilpert, J. Debes, P. Afshar, S. Rahban, R. Holmes, Calcif. Tissue Int.
2002, 71, 172–8.
[17] K. Lin, J. Chang, J. Lu, W. Wu, Y. Zeng, Ceram. Int. 2007, 33, 979–985.
[18] M. Neumann, M. Epple, Eur. J. Trauma 2006, 32, 125–131.
[19] C. H. Yao, B. S. Liu, S. H. Hsu, Y. S. Chen, Biomaterials 2005, 26, 3065–3074.
[20] F. H. Lin, C. H. Yao, J. S. Sun, H. C. Liu, C. W. Huang, Biomaterials 1998, 19, 905–
917.
[21] C. Hoon Lee, S.-C. Kwak, J.-Y. Kim, H. Mee Oh, M. Chual Rho, K.-H. Yoon, W.-H.
Yoo, M. Su Lee, J. Oh, J. Pharmacol. Sci. 2014, 124, 344–353.
[22] J. R. Khurma, D. R. Rohindra, A. V. Nand, Polym. Bull. 2005, 54, 195–204.
[23] C. Zou, W. Weng, X. Deng, K. Cheng, X. Liu, P. Du, G. Shen, G. Han, Biomaterials
2005, 26, 5276–5284.
[24] Z. J. Zhu, H. Shen, Y. P. Wang, Y. Jiang, X. L. Zhang, G. Y. Yuan, Asian Pac. J. Trop.
Med. 2013, 6, 753–756.
[25] M. Kikuchi, Y. Koyama, T. Yamada, Y. Imamura, T. Okada, N. Shirahama, K. Akita,
K. Takakuda, J. Tanaka, Biomaterials 2004, 25, 5979–5986.
[26] S. M. Kuo, G. C. C. Niu, S. J. Chang, C. H. Kuo, M. S. Bair, J. Appl. Polym. Sci. 2004,
94, 2150–2157.
158
[27] S. Zhu, P. Lu, H. Liu, P. Chen, Y. Wu, Y. Wang, H. Sun, X. Zhang, Q. Xia, B. C.
Heng, et al., Ann. Rheum. Dis. 2013, 1, 1–11.
[28] A. G. Sullad, L. S. Manjeshwar, T. M. Aminabhavi, Polym. Bull. 2014, 72, 265–280.
[29] R. Wary, S. Sivaraj, R. Kumar, S. A. I. L. Mari, G. Dasararaju, G. Kannayiram, Int. J.
Pharm. Pharm. Sci. 2014, 6, 94–100.
[30] M. Kucharska, K. Walenko, M. Lewandowska-Szumieł, T. Brynk, J. Jaroszewicz, T.
Ciach, J. Mater. Sci. Mater. Med. 2015, 26, DOI 10.1007/s10856-015-5464-9.
[31] Y. J. Wang, J. H. Ge, Y. D. Zheng, X. H. Zhang, X. F. Chen, W. Ran, G. Wu, Key Eng.
Mater. 2005, 280-283, 1613–1618.
[32] N. Pekel, O. Güven, Polym. Int. 2008, 57, 637–643.
[33] P. A. Miller, N. L. Clesceri, Waste Sites as Biological Reactors: Characterization and
Modeling, CRC Press, 2010.
[34] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309–319.
[35] G. Wirzing, Anal. Chemie 1980, 302, 97–108.
[36] E. B. Montufar, C. Gil, T. Traykova, M. P. Ginebra, J. Planell, Key Eng. Mater. 2008,
363, 323–326.
159
11 Coating of Surfaces: Protein-repellant Surfaces with Zwitterionic Microgels
11.1 Introduction
A material that comes in contact with the human body triggers a biological response that can
lead to an immune reaction, thrombosis or embolism [1][2]. The design of so-called non-
fouling, hemocompatible materials is essential and required in the biomedical field, especially
for the development of contact lenses, implants, prostheses, drug delivery carriers, and
sensors [3][4]. In general, a non-fouling material should be hydrophilic, overall electrically
neutral and preferably a hydrogen bond acceptos [5]. Many authors agree that a hydrophilic
surface minimizes protein adhesion[2][1][3]. The surface should be electrically neutral so no
electrostatic interactions between the surface and the protein occur. The absence of hydrogen
bond donors might reduce the amount of hydrogen bonds [3].
The exact process of protein adhesion onto a surface is not yet fully understood. The
hydration state of a surface, however, has a great influence on the adhesion process of
proteins onto a surface. K. Ishihara et al. describe this phenomenon as follows [6]: The
adsorption of proteins onto a polymer surface in aqueous media results in the loss of bound
water by the protein and thus a conformational change. This causes irreversible adsorption of
the protein. If the water state on the surface, however, is similar to the surrounding aqueous
solution, no water needs to be released by the protein and no conformational change takes
place. Thus, proteins can reversibly come into contact with a surface.
Not only coating with proteins leads to fouling, but also the coating with bacteria. Here, too,
the exact adhesion process is unclear. G. Cheng describes the process of the creation of an
undesired biofilm. Bacteria deposit non-specifically and reversibly on the surface [7].
Subsequently, bacteria begin to produce exopolysaccarides (EPS) that enclose the bacteria
film. Due to the charged and neutral polysaccharide groups, the EPS is able to filter nutrients
from the surrounding liquid medium and further protect the bacteria film from the body’s
immune response.
Betaines are known for their ultra low-fouling properties. Zhang et al. used carboxybetaine
(CB) to polymerize nanogels that encapsulate dextran as a model drug and Fe3O4 as imaging
reagents for magnetic resonance imaging (MRI) [8]. They could show in their study that there
is a low uptake amount of macrophage cells indicating little interactions with the immune
system. CB is therefore a promising candidate for non-fouling materials and a possible
substitute for prevalent polyethylene glycol (PEG) and phosphorylcholine (PC). These
materials are widely used for e.g. marine coatings, but offer various disadvantages: PEG, for
instance, is prone to oxidation, and both PEG and PC are short of functional groups that might
160
immobilize ligands [4]. Even before the use of PEG and PC, neutral poly-2-hydroxyethyl
methacrylate (PHEMA) was used for the coating for non-fouling materials, but several
authors reported the considerably decreased anti-fouling properties when in contact with
human blood serum and plasma [9][10][11]. S. Chen et al. interpreted the high resistance of
zwitterions to protein adhesion as the result of the strong hydration through ionic solvation.
They prepared zwitterionic hydrogels and showed that the adhesion of both IgG and
fibrinogen is very low under physiological conditions [4]. Z. Zhang et al. confirm that the
tightly bound water around the negatively charged pendant group is the driving factor behind
the non-adhesion behavior [12]. Especially in zwitterionic hydrogels, the ratio between free and
bound water is higher as compared to other hydrophilic polymers. Y.-J. Shih and Y. Chang
found out that the non-fouling behavior of SB can be tuned by the molecular weight of the
polymer [3]. A polySB polymer with a molecular weight of 135 kDa exhibited both the
strongest hydration capability for binding water and subsequently the most effective non-
fouling behavior.
11.2 Experimental Part
For the coating of surfaces, zwitterionic microgels from my diploma thesis were used. They
have a zwitterionic content of 0, 2, and 5 wt-%.
The coating procedure is described in the Experimental Part in chapter 4.
161
11.3 Results and Discussion
Zwitterionic microgels with a varying amount of phosphobetaine (PB) as well as pure PVCL
microgels were coated on Si-wafers after plasma activation to study the interfacial properties
of the microgels. Si-wafers were used because of the facile deposition of a variety of
polymers, their clear chemical structure, and their wide applications as various biomaterials.
Figure 11.1 Schematic representation of zwitterionic microgels on hydrophilic plasma-treated
Si-wafer.
AFM images show the surface topography before and after the coating. The images reveal
that betaine-free microgels form a homogeneous coating layer with equal distances between
the particles and no gaps or voids (see Figure 11.2). Microgels with 2 wt-% PB also form a
monolayer with particles closer to each other. The smaller distance is due to the zwitterionic
nature of particles and charges that attract each other. Particles with 5 wt-% PB form a
thorough multilayer with very small interparticle distances.
The roughness of the surfaces was analyzed quantitatively by determining the rms roughness.
According to Y. Xu, the protein repellence properties of a surface are directly linked to its
roughness [13]. A surface is considered to be smooth when it has an rms below 0.5 µm [14]. The
native wafer (not shown) has a very low rms of 7 nm. The values for the surface roughness
decrease slightly from microgels with 0 wt-% to 2 wt-% and 5 wt-% PB and are 19.3 ±
5.8 nm, 14.0 ± 5.1 nm, and 10.5 ± 3.7 nm, respectively. A comparison of those values with
the PDIs of the microgels (0.048, 0.031, and 0.029, respectively) indicates that there is a
direct correlation between the polydispersity of the sample and the surface roughness.
Another explanation is the structure of the particles. All rms are still below the microgel
particle sizes by a factor of roughly 10 meaning that the surface looks smooth on a length
scale of the particle size. On a molecular level, the microgels have a “hairy” surface with
loose ends that form loops and dangling ends that might lead to a higher surface roughness.
An increased content of zwitterionic PB, however, leads to the formation of inter- and
intramolecular ionic bridges between the negatively charged quaternary ammonia group and
162
the positively charged phosphonium group, thus “gluing” the tangling loose ends to the
particle surface.
Figure 11.2 Coating of Si-wafer with (a) pure PVCL microgels; (b) 2 wt-% PB; (c) 5 wt-%
PB. Insets show the respective contact angle measured in air at room temperature.
Furthermore, the contact angle of the coatings was determined to investigate the wettability of
the microgel-coated surfaces. Hydrophilic surfaces have a contact angle of 0 – 6 ° and
hydrophobic surfaces of 50 – 70 ° [15]. The contact angle was measured immediately after the
deposition of the water droplet and again after 2 min. A disparity in the contact angle over
time can be attributed to two factors: swelling of the dried microgels and reorientation of
functional groups on the surface. Table 11.1 gives the values for all coated surfaces. The
contact angle of uncoated Si-wafer is given in literature as ~ 11 ° [14]. All coated surface have
a significant higher contact angle than uncoated Si-wafer due to the hydrophobic backbone of
both PB and PVCL. For the coated surfaces, it is apparent that the surface becomes more
hydrophilic with increasing PB content in the microgels due to the zwitterionic nature of PB.
Table 11.1 Contact angles of PB microgels with varying PB content on Si-wafer in air at
room temperature. The values are the average of three separate drops on different areas.
PB content / % Contact angle immediately
after deposition / °
Contact angle after 2 min
of deposition / °
0 56.4 ± 3.10 49.8 ± 0.35
2 50.1 ± 4.32 36.5 ± 4.95
5 31.2 ± 5.64 16.5 ± 2.24
PB microgels coated on Si-wafer were exposed to BSA BODIPY FL conjugate protein to test
the effect of zwitterionic coatings on protein adsorption. BSA was chosen due to its well-
known properties (IEP ~ 5) and its wide-ranged use as a model protein. Measurements were
carried out in PBS buffer (pH ~ 7.4) to mimic typical physiological condition. Figure 11.3
shows fluorescence microscopy images after the incubation with BSA BODIPY FL
a b c
163
conjugate. Betaine-free PVCL microgels are shown on the respective left side as a reference.
High amounts of protein adsorb to the surface as indicated by the light green coloring. BSA
will bind to the hydrophilic silicon oxide surface though BSA has almost no net charge at pH
~ 5 [16]. Still, the local positive charges of BSA are attracted by the slightly negatively charged
oxide [17].
Figure 11.3 Images of protein BSA adhesion on Si-wafers coated with (a) 2 wt-% PB; (b)
5 wt-% PB; and (c) 10 wt-% PB. The reference PVCL microgels containing no PB is given on
the respective left side. (d) Determination of cell adhesion with human dermal fibroblasts on
coated Si-wafer.
The coating with zwitterionic microgels greatly affects the adhesion of proteins. The light
green color vanishes increasingly with a growing content of PB in the microgel. Since
electrostatic attraction is one of two main factors for protein adhesion (besides hydrophobic
interactions), a decrease in protein adhesion on zwitterionic surfaces confirms the
electroneutrality of the surface layer. These results are in agreement with earlier observations
made by E. Hatakeyama et al. [18]. They compare quaternary phosphonium polymers with
PEG-based polymers and state that some analogues have the same or even better protein-
repellency than PEG-based polymers. These results are also in agreement with the results
Reference Reference
Reference
a b 2 wt-% PB 5 wt-% PB
c 10 wt-% PB d
164
given by E. F. Murphy et al. [17]. They found that the outer layer of 25 Å of a thin layer of
phosphoryl choline on a surface consisted of approximately 86 % water. Therefore, it can be
assumed that an increase of PB leads to an increasing steric hindrance that hinders protein
deposition. Because methacrylate groups increase the adhesion of proteins on a surface [17],
this suggests further that the zwitterionic compound in the microgel is located on the particle
surface.
E. Ostuni et al. describe how the adsorption of proteins entails cell and bacteria adhesion [19].
A surface that is capable to prevent protein adsorption should thus be able to suppress cell
adhesion. Therefore, cell tests with human dermal fibroblasts were performed. Figure 11.3d
presents the results after 24 h of incubation. Pure PVCL microgels as well as PVCL microgels
with a low amount of PB exhibit the same high cell adhesion behavior as the uncoated
surface. In contrast, the cell adhesion decreases significantly with increasing PB content.
After coating, 73% less cell adsorption was observed in comparison to the surface coated with
the betaine-free or low amount of PB coated surface.
Both the reduced cell and protein adhesion can be attributed to an enhanced hydration of the
surface through the zwitterionic PB.
In a next step, samples with a very high content of betaine are examined. The microgels were
synthesized via emulsion polymerization (described in chapter 5). A sulfobetaine sample with
a monomer ratio of VCL:SB = 1:1 as an exemplary sample was coated on a glass surface and
dried at air overnight. Similar to the results discussed above, the sample was exposed to BSA
tetramethylrhodamine conjugate protein and analyzed with a light microscope as well as
contact angle measuring device.
Figure 11.4a shows the improved protein repellency of microgels with a high betaine content
of ~ 50 mol-%. No protein can be seen on the coated surface, while the reference surface
coated with pure PVCL microgels shows a bright red indicating a complete protein coating.
The protein-repellency is greatly improved compared to surfaces coated with zwitterionic
microgels with a low betaine content (see Figure 11.3). The surfaces are further more
hydrophilic with a contact angle of 9 ± 2 °.
165
Figure 11.4 (a) Protein-repellency of zwitterionic microgels with a monomer ratio of
VCL:SB = 1:1. (b) Contact angle measurement.
It is obvious that for a possible application for protein-repellent catheters in the medical field,
it has to be ensured that zwitterionic microgels are biocompatible. Zwitterionic microgels
(VCL:SB = 1:1) as well as pure PVCL microgels as reference were tested with regard to their
cytotoxicity. A negative test, however, states only that a material contains no harmful
substances and causes no harmful effects, it can therefore only be understood as an indication
whether a material is biocompatible, not as a final proof.
In the following, the influence of coated glass disks on cell adhesion during osteogenic
differentiation was analyzed for three different types of cells: human mesenchymal stem cells
(hMSC), diploid fibroblast cells (dFb), and mouse fibroblast cells (L929). Fibroblasts and
mesenchymal stem cells are the natural precursor cells of bone tissue [20]. Fibroblasts are cells
of connective tissue that differentiated from mesenchymal stem cells and form collagen to
strengthen the extracellular matrix. Mesenchymal cells are cells from bone marrow that are
still undifferentiated and are able to differentiate in various cell types such as bone or
cartilage. Cytotoxicity experiments were conducted by cand. med. P. Kandt and S. Neuss-
Stein of the Institute of Pathology and at the Biointerface Group of Helmholtz Institute for
Biomedical Engineering (RWTH Aachen, University Hospital) in the framework of the ERS
Project “Antimicrobial and tissue regenerating nanogels for implant coating”. Representative
images are given in Figure 11.5 for the microgel sample VCL:SB = 1:1 prepared via inverse
mini-emulsion (see Chapter 5) and for pure PVCL microgels as a reference.
9 ± 2 ° b
166
Figure 11.5 Fluorescent (FDA/PI) live/dead staining after osteogenic differentiation of (a) +
(d) dFb; (b) + (e) hMSC; and (c) + (f) L929. (a) – (c) PVCL microgels; (d) – (f) Microgels
prepared via inverse mini-emulsion with a monomer ratio of VCL:SB = 1:1. The scale bars
represent 75 (a,c,d,f) and 100 (b,e) µm. Images were kindly provided by cand. med. Pierre
Kandt.
Cells show either typical flat and spindle-shaped (dFb and hMSC) or spherical morphology
(L929). The ratio between living (green) and dead (red) cells indicates whether the surface is
cytotoxic. According to ISO 10993, microgels are not considered cytotoxic if not more than
5 % of dead L929 cells are present. This is true for both microgel samples and both pure
PVCL and zwitterionic microgels are therefore not cytotoxic. Images show that the surfaces
coated with zwitterionic microgels display an improved cell response. Figure 11.5c reveals
that a surface coated with PVCL exhibits more dead cells than a surface coated with SB
(Figure 11.5f). Furthermore, SB coated surfaces are better coated than PVCL coated surfaces.
a b c
d e f
167
11.4 Summary
This chapter showed the possible application of zwitterionic microgels as coating material in
the biomedical field. Coating experiments on Si-wafer as a model surface resulted in
homogeneous layers without gaps and voids. Contact angle measurements showed that coated
surfaces are less hydrophilic than uncoated surfaces due to the hydrophobic polymer
backbone. Still, an increase in PB leads to a decrease in the contact angle. Amongst others, K.
Ishihara et al. demonstrated that a high uptake of water is a prerequisite for successful protein
repellence [6]. Fluorescence active BSA was used to test the protein-repellent properties. Data
show that an increase in PB content results in a decrease in protein adhesion. This is due to
the improved hydration of the surface. An improvement in lower cell adhesion further
demonstrates the possible potential of zwitterionic microgels for protein and cell repellent
surfaces. Experiments were repeated for microgels samples with a high SB content that were
prepared via inverse mini-emulsion. Their protein-repellent properties are increased compared
to PB microgels with a low betaine content.
168
11.5 Literature
[1] Y. Chang, W.-Y. Chen, W. Yandi, Y.-J. Shih, W.-L. Chu, Y.-L. Liu, C.-W. Chu, R.-C.
Ruaan, A. Higuchi, Biomacromolecules 2009, 10, 2092–100.
[2] Y. Chang, W. Chang, Y. Shih, T. Wei, G. Hsiue, Appl. Mater. Interfaces 2011, 3,
1228–1237.
[3] Y.-J. Shih, Y. Chang, Langmuir 2010, 26, 17286–94.
[4] S. Chen, S. Jiang, Adv. Mater. 2008, 20, 335–338.
[5] E. Ostuni, R. G. Chapman, R. E. Holmlin, S. Takayama, G. M. Whitesides, Langmuir
2001, 17, 5605–5620.
[6] K. Ishihara, H. Nomura, T. Mihara, K. Kurita, Y. Iwasaki, N. Nakabayashi, J. Biomed.
Mater. Res. 1998, 39, 323–30.
[7] G. Cheng, Z. Zhang, S. Chen, J. D. Bryers, S. Jiang, Biomaterials 2007, 28, 4192–9.
[8] L. Zhang, H. Xue, Z. Cao, A. Keefe, J. Wang, S. Jiang, Biomaterials 2011, 32, 4604–8.
[9] C. Zhao, L. Li, Q. Wang, Q. Yu, J. Zheng, Langmuir 2011, 27, 4906–13.
[10] B. Mrabet, M. N. Nguyen, A. Majbri, S. Mahouche, M. Turmine, A. Bakhrouf, M. M.
Chehimi, Surf. Sci. 2009, 603, 2422–2429.
[11] C. Yoshikawa, A. Goto, Y. Tsujii, T. Fukuda, T. Kimura, K. Yamamoto, A. Kishida,
Macromolecules 2006, 39, 2284–2290.
[12] Z. Zhang, T. Chao, L. Liu, G. Cheng, B. D. Ratner, S. Jiang, J. Biomater. Sci. Polym.
Ed. 2009, 20, 1845–59.
[13] Y. Xu, M. Takai, K. Ishihara, Biomaterials 2009, 30, 4930–8.
[14] R. S. Faibish, W. Yoshida, Y. Cohen, J. Colloid Interface Sci. 2002, 256, 341–350.
[15] G. Kissinger, W. Kissinger, Phys. Status Solidi 1991, 123, 185–192.
[16] Y. Mukai, E. Iritani, T. Murase, J. Memb. Sci. 1997, 137, 271–275.
[17] E. F. Murphy, J.R. Lu, J. Brewer, J. Russell, J. Penfold, Langmuir 1999, 15, 1313–
1322.
[18] E. S. Hatakeyama, H. Ju, C. J. Gabriel, J. L. Lohr, J. E. Bara, R. D. Noble, B. D.
Freeman, D. L. Gin, J. Memb. Sci. 2009, 330, 104–116.
[19] E. Ostuni, R. G. Chapman, M. N. Liang, G. Meluleni, G. Pier, D. E. Ingber, G. M.
Whitesides, Langmuir 2001, 17, 6336–6343.
[20] D. Duarte, C. Panfil, K. Raupach, S. Neuss, M. Weber, C. Otten, U. Reisgen, H.
Fischer, Biomed. Eng. / Biomed. Tech. n.d., 40.
169
12 Summary
The aim of this work was the synthesis of polyampholyte microgels and the understanding of
the influence of distribution of ionizable groups on the microgels’ properties, their
interactions with proteins and possible applications.
In the first part, both zwitterionic and ampholyte microgels based on poly(N-
vinylcaprolactam) (PVCL) and poly(N-isopropylacrylamide) (PNIPAm) were prepared and
the ionizable groups were incorporated either statistically, as core-shell or Janus-like.
Chapter 5 dealt with the synthesis and characterization of zwitterionic particles. In the first
part, particles with a low amount of betaine (i.e. up to 10 wt-%) were prepared with
surfactant-free free-radical precipitation polymerization in water. Particles containing
sulfobetaine (SB) have a hydrodynamic radius of ~ 200 nm independent on the amount of SB
incorporated. Since zwitterionic microgels are able to crosslink even without the presence of a
classical crosslinker such as N,N’-methylenebis(acrylamide) (BIS) due to the formation of
intra- and intermolecular ionic bridges between the sulfo group and the quaternary ammonium
group, microgels without the use of BIS (“non-crosslinked”) are prepared and compared to
classically crosslinked particles (“crosslinked”). The softness of the particles was investigated
with AFM. It could be shown that crosslinked particles are stiffer than non-crosslinked
particles. This was attributed to the increase in inter- and intramolecular ionic bridges. The
degradation of crosslinked and non-crosslinked particles after the addition of salt is studied
with FTIR, DLS, and FESEM spectroscopy. It could be seen that particles degrade completely
in 1 M of NaCl. DLS measurements show that the particle size decreases from ~ 200 nm to ~
20 nm. Microscopy images confirm that the previous monodisperse, spherical particles
degrade to non-defined polymer aggregates.
Since it is shown that the incorporation of amounts of betaines higher than 10 wt-% is not
possible using free-radical precipitation polymerization, particles with a higher amount of SB
are prepared using free-radical inverse mini-emulsion in the second part of chapter 5. The
focus of particle characterization was on the temperature-sensitivity of the particles due to an
effect not seen for microgels with a low betaine content: beginning from a monomer ratio of
VCL:SB = 1:1, microgels not only exhibited an LCST, but also a UCST. The first is
generated by the collapse of PVCL polymer chains, the latter by the collapse of polySB
chains. The position of the phase transition temperatures is strongly dependent on the amount
of SB incorporated. An increasing amount of SB shifted both UCST and LCST to higher
temperatures. This behavior was both monitored with DLS and UV-Vis measurements.
Chapter 6, 7, and 8 discussed the synthesis of ampholyte microgels with itaconic acid (IA)
170
and vinylimidazole (Vim) with different architectures: statistically, core-shell, and Janus-like,
respectively. The amounts of ionizable groups was varied and verified with FTIR
spectroscopy. The varying distributions of ionizable groups were confirmed with TEM after
staining of the carboxylic groups with U(Ac)3.
Microgels with a statistical distribution of ionizable groups were obtained by mixing all
monomers in water at 70 °C. DLS and electrophoresis showed that particles swell at high and
low pH, while they are in a collapsed state at the particles’ isoelectric point (IEP). It is
possible to tune the position of the IEP as well as the degree of swelling by the ratio between
ionizable groups. Particles with a symmetric ratio of ionizable groups have a typical V-shaped
curve displaying the hydrodynamic radius as a function of pH. In-situ AFM measurements in
water further affirmed that particles are stiffer at pH ~ IEP due to the collapse of the microgel
network compared to particles at low and high pH.
Core-shell microgels were prepared in a two-step synthesis wherein the shell was synthesized
after the complete synthesis of the core. DLS measurements showed that the location of the
negative charge in the core or in the shell greatly influences the swelling behavior of the
particles in different pH. Particles with IA in the core and Vim in the shell behaved like
microgels with a statistical distribution of ionizable groups, i.e. the hydrodynamic radius
exhibited a typical V-shaped curve with varying pH, though the degree of swelling is much
less pronounced at high pH than at low pH. An increase in shell thickness, however, led to an
intensified swelling of the shell at high pH. Particles with a reverse location of ionizable
groups, i.e. Vim in the core and IA in the shell, do not swell at high pH. This was attributed to
the fact that the shell is too “weak” to counteract the collapse of the particle core.
In chapter 8, the polymerization of NIPAm with either Vim or IA as comonomer was closely
investigated with calorimetry to obtain information about particle formation and monomer
reactivities. For the synthesis of ampholyte Janus-like particles, both microgel dispersions
(NIPAm-IA and NIPAM-Vim) were initiated separately and mixed after a defined amount of
time. TEM measurements confirmed the formation of Janus-like particles if the dispersions
were mixed after 1 min after both polymerizations were initiated.
The second part of this thesis looked closely at the influence of different microgel architecture
on the uptake and release of proteins. The uptake and release of cytochrome c (cyt-c) as a
model protein is investigated in chapter 9. Since cyt-c is positively charged over a broad range
of pH, while ampholyte microgels have an IEP of ~ 6.5, experiments were conducted at
different pH where microgel and protein have opposite or like charges. The kinetic of release
171
was monitored with UV-Vis spectroscopy as well as optically due to the red color of the
protein’s heme group. Different triggers such as pH, temperature and salt solution were tested
for a complete or partial release of protein. SLS and DLS measurements showed the influence
of protein uptake on the microgels’ hydrodynamic radius and molecular weight.
The third part studied possible applications of the microgels prepared in the previous chapters.
Chapter 10 dealt with the use of charged microgels in biohybrid systems. The first part of the
chapter focused on zwitterionic microgels based on PVCL and phosphobetaine (PB) as
possible reaction vessels for the controlled growth of CaCO3 crystals. Two different synthesis
routes were tested to ensure crystal growth within the microgel particle. Following, the
maximum capacity of CaCl2 uptake was investigated gravimetrically as well as with ITC. As
expected, there was a higher capability to bind Ca2+ with a higher amount of PB. The
properties of CaCO3 were studied with XRD and FTIR. It could be shown that the common
calcite is the only crystal structure formed. SEM and TEM measurements optically confirmed
the formation and location of crystals and demonstrated that only specific synthesis
parameters lead to a crystal growth within the microgel network. Heating of the microgel
dispersion released the CaCO3 crystals as was shown with TEM and F-FFF.
The second part looked at the application of positively charged microgels based on PVCL and
Vim as bone substitute material in combination with β-tricalcium phosphate (β-TCP). In
collaboration with the university hospital, this work focused on the chemical properties of the
microgels with or without the presence of β-TCP, so the best parameters for the formation of
green bodies were identified. Here, FTIR spectroscopy was used for the confirmation of the
successful incorporation of Vim, while electrophoresis, TGA, and XPS measurements
confirmed the incorporation of Vim-microgels in the hybrid material. Water sorption
measurements were conducted to study the amount as well as the kinetics of moisture uptake
as well as determine the BET surface of the composite. SEM completed the characterization
of the biohybrid material.
Chapter 11 had a closer look on the properties of zwitterionic microgel coatings on Si-wafer.
In a first step, the right parameters for a complete formation of mono- or thin poly-layer were
investigated via AFM. Contact angle measurements delivered information of the
hydrophilicity of the coated surfaces. An increase in the betaine amount leads to a decrease in
hydrophilicity. Protein-repellency was examined with the help of fluorescence microscopy.
Here, surfaces coated with zwitterionic microgels showed great improvement over surfaces
coated with non-zwitterionic microgels.
172
13 Nomenclature
AAEM 2-(Methacryloyloxy)ethyl acetoacetate
AFM Atomic Force Microscopy
AGET ATRP Activators generated by electron transfer atom transfer radical polymerisation
AIBN Azobisisobutyronitrile
AMPA 2,2-Azobis(2-methylpropionamidine) dihydrochloride
APS Ammonium persulfate
APTMS (3-Aminopropyl)trimethoxysilane
BET Brunauer-Emmett-Tellett
BIS N,N’-Methylenebis(acrylamide)
BSA Albumin from borine serum
β-TCP β-tricalcium phosphate
CB Carboxybetaine
δ Chemical Shift
Ð Dispersity
DLS Dynamic Light Scattering
DSC Differential Scanning Calorimetry
EM Electrophoretic mobility
EPS Exopolysaccharides
F-FFF Flow Field-Flow Fractionation
FTIR Fourier-Transformed Infrared Spectroscopy
HAp Hydroxylapatite
HEMA 2-hydroxyethyl methacrylate
IA Itaconic Acid
IEP Isoelectric point
ITC Isothermal Titration Calorimetry
KPS Potassium persulfate
λ Wavelength
LCST Lower Critical Solution Temperature
MADIX Macromolecular Design via the Interchange of Xanthates
MMA Methylmethacrylate
Mn Number average molar mass
MRI Magnetic Resonance Imaging
NIPAm N-isopropylacrylamide
NMR Nuclear Magnetic Resonance
PB Phosphobetaine
PBS Phosphate Buffered Saline
PC Phosphorylcholine
PDI Polydispersitätsindex
PDMS Polydimethylsiloxan
PEG Polyethylene glycol
PHEMA Poly-2-hydroxyethyl methacrylate
PMMA Polymethylmethacrylate
PS Polystyrene
173
PVCL Poly-N-Vinylcaprolactam
PNIPAm Poly-N-isopropylacrylamide
RAFT Reversible Addition-Fragementation Chain Transfer Polymerization
RH Hydrodynamic radius
rpm Rounds per minute
SAM Self-assembled monolayer
SB Sulfobetaine
SDS Sodium dodecyl sulfate
SEM Scanning Electron Microscope
SLS Static Light Scattering
Span 80 Sorbitan monooleate
T Temperature
TEM Transmission Electron Microscope
TEMED N,N,N’,N’-tetramethylethylenediamine
TGA Thermogravimetric Analysis
Tween 80 PEG 20-Sorbitan monooleate
UCST Upper Critical Solution Temperature
VCL N-Vinylcaprolactam
Vim 1-Vinylimidazole
VPTT Volume Phase Transition Temperature
wt-% Weight percent
XRD X-Ray Diffraction
174
14 Appendix
Chapter 5.1 – Non-crosslinked zwitterionic microgels via precipitation polymerization
Figure 14.1 FTIR spectra of non-crosslinked zwitterionic microgels prepared via precipitation
polymerization. Inset shows zoom-in of peak at 1050 cm-1 belonging to the SO3- group of SB.
For determination of SB content, the calibration curve in Figure 14.2 was used.
Chapter 5.2 – Zwitterionic microgels via emulsion polymerization
Figure 14.2 (1) Calibration curve for determination of sulfobetaine (SB) amount in PVCL-SB
microgels via FTIR spectra. (2) FTIR spectra of zwitterionic microgels with SB prepared with
emulsion polymerization. Inset shows peaks at 1051 cm-1 (SB) and 1635 cm-1 (PVCL) that
were used for generation of calibration curve.
175
Figure 14.3 Change of hydrodynamic radius RH of emulsion droplets over time with varying
amount of surfactant. Measurements were conducted at 70 °C. Inset shows the point of time
when the emulsion breaks.
Figure 14.4 Change of hydrodynamic radius RH of emulsion droplets over time with
surfactant concentrations below the cmc. Measurements were conducted at 70 °C. The PDI of
the droplets fluctuates between 0.7 and 1.0.
Chapter 7 - Core-Shell microgels
Figure 14.5 Colloidal stability of core-shell microgels at 20 °C, pH 7.
176
Figure 14.6 FTIR spectra for core-shell microgels with increasing shell thickness.
Figure 14.7 RH as a function of temperature for core and core-shell microgels and their
respective transition temperatures. Measurements were conducted in pH ~ 6.0 phosphate
buffer.
Chapter 9 - Interactions with proteins
jkkjhkjhkhjkj hjjj
Figure 14.8 UV-Vis spectra of (a) pure cytochrome C and (b) pure microgel sample
(VCL:Vim:IA = 80:10:10) at pH 3, 6.2, and 8. Measurements were conducted at T = 25 °C.
a b
177
Figure 14.9 Change of absorbance with dialysis time for ampholyte microgel with (a)
statistical distribution of ionizable groups (VCL:Vim:IA = 80:10:10); and core-shell
structures: (b) cVCL-/sNIPAm+ (1:1); (c) cVCL+/sNIPAm-; (d) cVCL-/sNIPAm+ (1:5).
Black curve: pH 3; red curve: pH 6.2; blue curve: pH 8.
Chapter 10.1 - Biomineralization
Figure 14.11 FTIR spectrum of microgels with 10 wt-% PB and varying CaCO3 content,
synthesized via gas diffusion. (Red) Pure CaCO3 (calcite); (green) Unmineralized microgel
with 10 wt-% PB; (black) Microgel with 13 wt-% CaCO3; (blue) Microgel with 28 wt-%
CaCO3; (turquois) Microgel with 67 wt-% CaCO3.
a b
c d
178
Table 14.1 Assignment of XRD peaks of calcite.
2Θ / ° Peak
29.4 (104)
36.1 (110)
39.3 (113)
43.7 (202)
47.6 (018)
48.4 (116)
57.6 (122)
60.8 (214)
64.8 (300)
Figure 14.10 XRD diffraction pattern of microgel with 2 wt-% PB and 13 wt-% CaCO3 via
gas diffusion: (a) with pre-dialysis; (b) without pre-dialysis; (c) high concentration of
microgel. Black lines indicate the peak positions for calcite as found in literature (JCPDS: 86-
2334).
b
c
a
179
Danksagung
Ich möchte mich herzlich bei denjenigen bedanken, die mir während meines Studiums und
meiner Dissertation beigestanden haben.
Zuerst möchte ich Prof. Dr. Andrij Pich für die interessante Themenstellung und für sein
fortwährendes Interesse an meiner Arbeit danken. Des Weiteren bedanke ich mich für die
Freiheit, meine Arbeit zu großen Teilen selbst gestalten zu können, sowie die Möglichkeit, an
nationalen und internationalen Konferenzen und Fortbildungen teilzunehmen.
Für die finanzielle Unterstützung danke ich dem Sonderforschungsbereich SFB 985.
Prof. Dr. Walter Richtering, Prof. Dr. Igor Protemkin, Andreas Schmid, Andrey Rudov und
Arjan Gellissen danke ich für die gute Zusammenarbeit und die vielen fruchtbaren Treffen im
Rahmen des SFB985.
Zusätzlich bin ich auch Pierre Kandt für die gute Zusammenarbeit im Rahmen des Projektes
„Antimicrobial and tissue regenerating nanogels for implant coating“ dankbar und für seine
geduldigen Erklärungen zum Thema Zellwachstum.
Weiterhin möchte ich Prof. Dr. Andrew Lyon dafür danken, dass ich 5 Monate in seinem
Arbeitskreis forschen durfte und für seine Zeit und Interesse, die er meiner Arbeit und meinen
Ergebnissen entgegengebracht hat.
Natürlich bin ich auch dankbar für die Hilfe der Studenten, die ihre Forschungs- und
Bachelorarbeit bei mir erstellt sowie als Hilfskraft gearbeitet haben. Dazu gehören Keti
Piradashvili, Alexander Töpel, Helmi Maria Ala-Mutka, Artur Gantarev, Julia Blöhbaum und
Miriam Al Enezy.
Des Weiteren danke ich meinen Kollegen und Freunden innerhalb des Arbeitskreises für die
wirklich tolle Atmosphäre und den sehr guten Zusammenhalt während der letzten Jahre.
Hierbei sei besonders Labor 3.04 und 3.05 hervorgehoben (Christian Willems, Dominic
Kehren, Andreea Balaceanu, Jason Zografou, Thorsten Palmer).
Zu guter Letzt bedanke ich mich bei meiner Familie, die mich während des gesamten
Studiums unterstützt hat.
180
Schlusserklärung
Hiermit erkläre ich, dass ich die vorliegende Arbeit selbstständig verfasst und keine anderen
als die hier angegebenen Quellen und Hilfsmitteln benutzt habe.
Ferner erkläre ich, dass ich nicht anderweitig mit oder ohne Erfolg versucht habe, eine
Dissertation einzureichen oder mich einer Doktorprüfung zu unterziehen.
Aachen, den 03.07.2015
………………………………………………………………………….
Ricarda Schröder