Laser-ablated volume and depth as a function ofpulse duration in aluminum targets

Boris Le Drogoff, François Vidal, Stéphane Laville, Mohamed Chaker, Tudor Johnston,Olivier Barthélemy, Joëlle Margot, and Mohamad Sabsabi

The ablated depth and volume per laser pulse from an aluminum target were measured for pulsedurations that ranged from 80 fs to 270 ps at an average fluence of �100 J�cm2 and a wavelength of 0.8�m. The ablated volume showed a flat maximum for subpicosecond pulses and a minimum for �6 ps. Thecrater diameters were rather constant up to pulse durations of �6 ps and increased for larger pulsedurations. As a result, the ablated depth also showed a plateau for subpicosecond pulses but decreasedmonotonically with pulse duration. A physical interpretation of these results and their consequences forlaser applications are discussed. © 2005 Optical Society of America

OCIS codes: 140.3390, 140.0140.

1. Introduction

Removal of matter by means of a laser beam, termedlaser ablation, has been accomplished in micromach-ining,1 thin-film deposition,2 and chemical analysis.3

For a given laser fluence (laser energy per unit area)and wavelength, the ablation dynamics and resultsvary considerably with laser pulse duration �.4 Al-though understanding the influence of pulse durationon ablation efficiency (i.e., on the volume of matterablated per laser pulse) for a given wavelength andfluence is fundamental for most laser applications,this issue has been discussed in only a few papers todate. One of the most significant investigations waspresented by Semerok et al.,5 who carried out mea-surements of a variety of metals and investigated thedependence of the ablation efficiency on pulse dura-tion for only three specific values of pulse duration(150 fs, 25 ps, and 4 ns). However, various fluences

and various laser wavelengths were compared foreach pulse duration. The main conclusion of thatstudy was that the ablated volume per pulse and perunit laser energy in several metals, displayed as afunction of the pulse duration, seems to show a min-imum in the picosecond regime. Additionally, thesame team presented ablation experiments with cop-per at a fluence of 21 J�cm2 with the pulse durationin the range 0.1–10 ps.6 The results clearly showed arough plateau in the ablated depth for subpicosecondpulses and a monotonic decrease for pulse durationslarger than �1 ps.

Even though these papers have, to some extent,addressed some features of the influence of the pulseduration on the ablation efficiency, a parametric anddetailed experimental determination of the ablationefficiency, with only the pulse duration varied, hadstill to be made to verify the general tendencies in-ferred by Semerok et al.5

The main purpose of the study reported here is topresent the results of laser-ablation efficiencies whena single laser wavelength, a single average fluence,and several pulse durations are used. A physical in-terpretation of the results is also discussed on thebasis of our simulation results.7

2. Experimental Setup

The experimental setup used in this study is shownin Fig. 1. The measurements were all performed in asingle-shot mode, in air at atmospheric pressure, byuse of a Ti:sapphire laser system �� � 0.8 �m� basedon the chirped-pulse amplification technique that al-lows the pulse duration to vary from 80 fs to 270 ps.

When this research was performed, B. Le Drogoff, F. Vidal(vidal@inrs-emt.uquebec.ca), S. Laville, M. Chaker, and T. John-son were with the Institut National de la Recherche Scientifique-Énergie, Matériaux et Télécommunications, 1650 BoulevardLionel-Boulet, Varennes, Quebec J3X 1S2, Canada. B. Le Drogoff,along with M. Sabsabi, is now with the Institut des MatériauxIndustriels, Conseil National de Recherches du Canada, 75 Bou-levard de Mortagne, Boucherville, Quebec J4B 6Y4, Canada. O.Barthélemy and J. Margot are with the Département de Physique,Université de Montréal, 2900 Boulevard Édouard-Montpetit, Mon-treal, Quebec H3C 3J7, Canada.

Received 18 June 2004; accepted 16 July 2004.0003-6935/05/020278-04$15.00/0© 2005 Optical Society of America

278 APPLIED OPTICS � Vol. 44, No. 2 � 10 January 2005

(We obtained the 270-ps laser pulses by bypassingthe compressor.) The incident Gaussian laser pulsewas spatially shaped by the knife-edge method with a5-mm-diameter diaphragm and focused with a35-mm objective lens to a diameter (at 1�e2) esti-mated as �5 �m at focus (using a knife-edge method),with a depth of 3.5 �m. The beam, directed along thenormal to the target surface, was focused at 25 �mbehind the incidence surface of the 100-�m-thick alu-minum target. (We verified that no air breakdownoccurred for the laser energy used, which was 60 �J.)The fluence at the focal spot was thus estimated as�300 J�cm2. As the pulse diameter is not constantover the target thickness, this value constitutes ahigh upper limit for the fluence used in the experi-ments. We estimate that all our ablated volume anddepth measurements were performed at an averagelaser fluence of roughly 100 J�cm2.

It is worth noting that we obtained reproduciblemeasurements only by reducing the size of the focalspot to at most a few tens of micrometers, for whichrelatively short focal lengths must be used. For largerfocal spots it is likely that the lack of reproducibilityis due to the presence of hot spots whose amplitudeand distribution can change from one shot to another.The pulse’s radial profile becomes more regular as thefocal spot diameter approaches the diffraction limit.

We define the average ablated depth per pulse as� � h�Nb, where h is the known aluminum-foil thick-ness �h � 100 �m� and Nb is the number of laser shotsof given energy necessary to bore through the alumi-num foil. Because optical microscope observationsshowed that the crater is approximately conical [Fig.1B], the ablated volume per pulse, V, is approxi-mately related to the ablated depth by means of thecrater diameter at the target surface, D, by the rela-tion V � ���12�D2�. (The crater diameter was deter-mined by optical microscopy.)

3. Results and Discussion

Figure 2 shows the measured ablated depth and vol-ume as a function of pulse duration. In this figure theright- and left-hand scales have been set such thatthe two curves coincide for the shortest pulse dura-tions. The error bars indicate the dispersion of 10

measurements. One can observe from Fig. 2 that thetwo curves can be made to coincide up to �6 ps.

For long pulses (�6 ps and more), Fig. 2 showssignificant differences in the variation of V and �.When the ablated volume starts to increase, the ab-lated depth continues to decrease with pulse dura-tion. The different behaviors of the ablated depth andthe volume are related to that of crater diameter D,which is shown in Fig. 3. The crater diameter can beseen to be equal to �24 �m below �6 ps and toincrease monotonically past this value. Thus one cansay that, whereas the crater diameter does not in-crease appreciably, the system is essentially one di-mensional (1-D), and that, in that regime, the ablateddepth and volume behave in the same way. For largerpulse durations, most of the ablated volume is due toan increase in the crater diameter, as the ablated

Fig. 3. Crater diameter as a function of pulse duration.

Fig. 1. A, Schematic of the ablation experiments. B, Conicalshape of the resultant crater.

Fig. 2. Measured ablated volume per pulse, V (filled squares), andablated depth per pulse �, (open circles), as functions of pulseduration. The scales were chosen such that the shortest pulseswould coincide.

10 January 2005 � Vol. 44, No. 2 � APPLIED OPTICS 279

depth decreases. A similarly decreasing ablateddepth for increasing fluence was noted for steel tar-gets8 for fluences smaller than �3000 J�cm2. How-ever, at higher fluences the ablated depth shows twomaxima8 in the same way as the volume shown inFig. 2.

The influence of the pulse duration on the ablationefficiency for aluminum was investigated theoreti-cally previously by means of a detailed 1-D fluidcode.7 This code includes laser energy absorption inthe cold solid as well as in the subsequent plasma,and a realistic equation of state for aluminum. Thecalculated ablated depth is shown in Fig. 4 as a func-tion of pulse duration for a laser wavelength of 0.8�m and for several values of fluences. One can ob-serve that, for a given fluence, (i) for pulses shorterthan �1 ps, the ablated depth is nearly independentof the pulse duration; (ii) for fluences higher than �10J�cm2, the ablated depth as a function of pulse dura-tion presents two maxima, the highest of which oc-curs for subpicosecond pulses, with a minimum thatoccurs near 10 ps, in qualitative agreement with Se-merok et al.5; and (iii) for longer pulse durations,ablation decreases as the corresponding ablationthreshold is approached.

The basic concepts can be obtained by examinationof the simulation snapshots in detail. For subpicosec-ond laser pulses, laser energy absorption occurs es-sentially close to the solid density, because mattercannot move much in such a short time. For longerpulses, significant heat conduction in the target takesplace while the pulse is on, resulting in less energyremaining to accomplish ablation. For these longerpulses, an important fraction of the incoming laserenergy is absorbed in the expanding plasma formedin front of the target surface, and this extra reservoirof energy assists ablation at late times.

The observed behavior of the measured ablatedvolume shown in Fig. 2 is in qualitative agreement

with the predictions of this model. It might seempuzzling at first that the measured ablated volume isalso in agreement with the ablated depth obtainedfrom the 1-D model and shown in Fig. 4, because theablated depth is seen to decrease in Fig. 2. The solu-tion to this apparent dilemma is that, in the 1-Dsimulations, it is implicitly assumed that no thermalconduction or plasma expansion is allowed in thelateral directions. As a result, in the 1-D model all theabsorbed laser energy is used to increase the depth ofthe crater instead of increasing its diameter. There-fore, although the crater diameter is assumed to beconstant in the 1-D model (with an undefined value),the corresponding ablated volume has the correct be-havior as a function of the pulse duration.

By analyzing the code details we found that theincrease in the ablated depth past the minimum ob-served in Fig. 4, near 10 ps, was essentially due to themore efficient transfer of laser energy to the plasmaformed with the longer pulses. This energy transferresulted in heating of the solid surface by thermalconduction, which naturally raised the amount of ab-lation. The increase of crater diameter for � � 6 psobserved in Fig. 3 is thus likely connected with thesize of the laser-heated plasma bowl formed in frontof the target surface. Lateral refraction of the laserlight by the expanding plasma may perhaps also be-gin to play a role in this increase in crater diameter atthe longer pulse durations.

4. Conclusions

In summary, the ablated volume per pulse measuredin an aluminum target for an average fluence of theorder of 100 J�cm2 shows a nearly flat plateau forsubpicosecond pulses, then a drop by a factor of 1.8with a minimum near � � 6 ps, and a modest increaseup to subnanosecond pulse durations. In contrast, theablated depth per pulse decreases monotonically as afunction of pulse duration for � � 1 ps. The craterdiameter is constant up to �6 ps and increases rap-idly beyond that value. From our 1-D code, the in-crease of ablated volume for longer pulse durations isinterpreted as being due to the absorption of laserenergy in the expanding plasma. A more detailedinterpretation of the experimental results, in partic-ular of the dependence of the crater’s diameter anddepth on the pulse duration, would require consider-ation of a two-dimensional model, at least for pulsedurations greater than �6 ps.

Our results clearly show that, if one wishes to per-form efficient ablation with lateral precision, pulsesof 1 ps or less are essential, whereas pulse durationslarger than �10 ps should be avoided. This advan-tage of subpicosecond laser pulses is now well estab-lished for micromachining of a large variety ofmaterials3,9 and is being investigated for medical ap-plications.10

The present investigation has clearly identified aboundary between short pulses and long pulses thatis near 6 ps for aluminum. Whereas a 1-D code, suchas the one discussed in this paper, cannot describethe observed two-dimensional features of the ablation

Fig. 4. Calculated ablated depth in aluminum as a function ofpulse duration for three fluences.

280 APPLIED OPTICS � Vol. 44, No. 2 � 10 January 2005

craters, it is likely that the 1-D code can still be usedto estimate this boundary with good precision. Al-though this study was performed with aluminum, webelieve that the results apply generally, except thatthe minimum ablated volume may well occur at othervalues than �6 ps for other materials.

This study was financed partly by the NationalScience and Engineering Research Council ofCanada.

References1. M. von Allmen and A. Blattner, Laser–Beam Interaction with

Materials: Physical Principles and Applications (Springer-Verlag, Berlin, 1995).

2. D. B. Chrisey and G. K. Hubler, eds., Pulse Laser Depositionof Thin Films (Wiley, New York, 1994).

3. J. D. Winefordner, I. B. Gornushkin, D. Pappas, O. I. Matveev,and B. W. Smith, “Novel uses of lasers in atomic spectroscopy,”J. Anal. Atom. Spectrom. 15, 1161–1189 (2000).

4. C. Momma, S. Nolte, B. N. Chichkov, F. von Alvensleben, andA. Tünnermann, “Precise laser ablation with ultrashortpulses,” Appl. Surf. Sci. 110, 15–19 (1997).

5. A. Semerok, C. Chaléard, V. Detalle, J.-L. Lacour, P. Mauch-ien, P. Meynadier, C. Nouvellon, B. Sallé, P. Palianov, M.Perdrix, and G. Petite, “Experimental investigations of laserablation efficiency of pure metals with femto, pico and nano-second pulses,” Appl. Surf. Sci. 139, 311–314 (1999).

6. B. Sallé, O. Gobert, P. Meynadier, M. Perdrix, G. Petite, and A.Semerok, “Femtosecond and picosecond laser microablation:ablation efficiency and laser microplasma expansion,” Appl.Phys. A 69, S381–S383 (1999).

7. S. Laville, F. Vidal, T. W. Johnston, O. Barthélemy, M. Chaker,B. Le Drogoff, J. Margot, and M. Sabsabi, “Fluid modeling ofthe laser ablation depth as a function of the pulse duration forconductors,” Phys. Rev. E Part 2 66, 066415 (2002).

8. C. Momma, B. N. Chichkov, S. Nolte, F. von Alvensleben, A.Tünnermann, H. Welling, and B. Wellegehausen, “Short-pulselaser ablation of solid targets,” Opt. Commun. 129, 134–142(1996).

9. B. N. Chichkov, C. Momma, S. Nolte, F. von Alvensleben,and A. Tünnermann, “Femtosecond, picosecond and nano-second laser ablation of solids,” Appl. Phys. A 63, 109–115(1996).

10. K. R. Sletten, K. G. Yen, S. Sayegh, F. Loesel, C. Eckhoff, C.Horvath, M. Meunier, T. Juhasz, and R. M. Kurtz, “An in vivomodel of femtosecond laser intrastromal refractive surgery,”Ophthalmic Surg. Lasers 30, 742–749 (1999).

10 January 2005 � Vol. 44, No. 2 � APPLIED OPTICS 281