ABS acylonitrile

7/27/2019 ABS acylonitrile

1/14

FIRE AND MATERIALS, VOL. 10,93-105 (1986)

Acrylonitrile-Butadiene-Styrene Copolymers (ABS):Pyrolysis and Combustion Products and theirToxicity-A Review of the LiteratureJoseph V. Rutkowski and Barbara C. Levintus Department ofCommerce, National Bureau of Standards, National Engineering Laboratory, Center for Fire Research,Gaithersburg, MD 20899, USA

A review of the literature was undertaken to ascertain the current knowledge of the nature of the thermal decomposit ionproducts generated from ABS and the toxicity of these evolved products in toto. The literature review encompassesEnglish language publications available through June 1984. This literature surveyed showed that the principal ABSthermooxidative degradation products of toxicologic importance are carbon monoxide and hydrogen cyanide. Theexperimental generation of these and other volatile products is principally dependent upon the combustion conditionsand the formulation of the plastic. The toxicity of ABS thermal degradation products has been evaluated by fivemethods. The LCso (30 min exposure + 14 day post-exposure period) values for flaming combustion ranged from15.0 mg 1- 1 to 28.5 mg 1-1. In the non-flaming mode of combustion, the LCso values ranged from 19.3 mg 1- 1 to64.0 mg 1- 1. Therefore, no apparent toxicological difference exists between the flaming mode and the non-flamingmode. The toxicity of ABS degradation products was found to be comparable with the toxicity of the thermaldecomposition products of other common polymeric materials.Keywords: ABS plastics; carbon monoxide; combustion products; hydrogen cyanide; li terature reviews;thermal decomposition; toxicity.

Figure 1. Chemicalstructure of ASS monomers.

StyreneO~=C-HIH(m.w. = 104.14)

predominant atomic species (atomic weight/molecularweight). The only source of nitrogen in the plastic isacrylonitrile, which contains approximately 26% nitrogen. The weight percentage of hydrogen is relatively small.Acrylonitrile is a colorless liquid characterized by astrong odor. The boiling point of acrylonitrile is 77.3 Cand its freezing point is approximately - 83.55 0c. Acrylonitrile is soluble in most solvents, both polar and nonpolar. Water solubility increases with temperature. 2 Acrylonitrile is capable of undergoing many chemical reactions, the majority of which occur at the carbon-carbondouble bond. It isat this site that the homopolymerizationof the acrylonitrile monomer occurs. This reaction takesplace readily in the presence of initiators through either afree radical mechanism or an anionic mechanism ofpolymerization.3 The resultant polymer, polyacrylonitrile, is a crystalline-type compound, which isnot soluble in

INTRODUCTION

Although the hazard posed by toxic gases and vaporsgenerated in a fire environment has long been recognized, 1 information regarding the toxic agent(s) generated from ABS which are responsible for fire deaths notattributable to burns is relatively sparse. The growth ofthe science of combustion toxicology over the last tenyears has resulted in the generation of a large quantity ofinformation regarding the thermal decomposition products of many natural and synthetic materials, and therelative toxicity of those materials' decomposition products in toto. This paper is a review of the Englishliterature, available through June 1984, on the thermaldecomposition products from acrylonitrile-butadienestyrene (ABS) terpolymers and an evaluation of thetoxicity data from animals exposed to those products.The ABS plastics are of particular interest since they arefound in numerous household items, such as telephones,appliances, luggage and piping, as well as in automobileand aircraft interior structures.

CHEMICAL STRUCTURE ANDTHERMOPHYSICAL PROPERTIESABS polymers are synthesized from three monomericchemicals: acrylonitrile, butadiene, and styrene (Fig. 1);from which the acronym is derived. The atomic components of these three chemical compounds are, exclusively,carbon, nitrogen and hydrogen, with carbon being thetAuthor to whom correspondence should be addressed.

AcrylonitrileHIC=C-C!!NI IH H(m.w. = 53.03)

ButadieneH H H1 I IC=C-C=CI I IH H H(m.w = 54.09)

0308-0501j86j030093-13$07.50 U.S. National Bureau of StandardsReceived 20 February 1 98 6

Accepted 24 July 1986


2/14

94 1. V. RUTKOWSKI AND B. C. LEVIN

most common solvents, e.g. acids, greases, oils, water andhypochlorite solutions. It will, however, degrade whenexposed to concentrated cold or hot dilute alkalinesolutions, as well as warm 50% sulfuric acid. Although theacrylonitrile monomer is significantly toxic by eitheringestion of the liquid or inhalation of the vapor, thehomopolymer is considered essentially non-toxic.3

1,3-Butadiene is a colorless gas. The boiling point ofbutadiene is - 4.5 e and its freezing point is - 108.96 dc.lt is soluble in organic solvents.4 Due to the presence oftwo double bonds, butadiene is a very reactive compound,allowing a variety of polymer structures to be formed. Allpolymerizations require either an initiator or catalyticfactor and occur by a free-radical, an ionic or a coordinatemechanism. The end result, polybutadiene, is a rubberpolymer formed by addition polymerization. S

Styrene is a colorless to yellowish, oily liquid which hasa penetrating odor. The boiling point range of styrene is145 to 146C and the freezing point is - 30.6C. Styreneis only slightly soluble in water; however, it is soluble inalcohol, ether, methanol, acetone and carbon disulfide.Styrene is not considered very toxic upon ingestion but isa significant irritant when applied to the skin, eyes andmucous membranes. The rate of polymerization of thestyrene monomer is slow at room temperature andincreases with heat. Since the polymerization is anexothermic process, the reaction can become selfaccelerating. Styrene homo polymerization can occureither by free-radical or ionic reactions. The resultantpolymer, polystyrene, is a clear, crystalline-typecompound.6-sABS polymers (Fig. 2) can be produced by two general

methods, each of which generates a different type ofplastic. Type A ABS polymers are produced by a mechanical blending of a styrene-acrylonitrile copolymer resinwith a butadiene-based elastomer (butadieneacrylonitrile rubber). The production of Type B ABSinvolves a grafting of styrene and acrylonitrile ontopolybutadiene. The Type B polymer also contains astyrene-acrylonitrile copolymer and ungrafted polybutadiene. The types of plastics are varied further bydifferences in the relative proportions of the monomercomponents. The styrene-acrylonitrile copolymer, foreither type of production procedure, can be composed of65-76% styrene and 24-35% acrylonitrile.9lo Since theformulations of ABS plastics are rarely the same andpoorly defined in most of the studies examined for thisreview, it is difficult to generalize or compare ABSsamples.

The temperatures at which ABS processing occursrange from 325-550C, depending upon the type andgrade of plastic being produced as well as the techniquebeing used. The melting point range for extrusion gradesof ABS is 88-120 C and for the injection molding gradesof ABS, 110-125CY The autoignition temperature of

tHH Htt H H HnHtI I I I I I Ic-c-c c=c-c-c CI I I I I I ACNH H H H 0y zFigure 2. Representative chemical structure of an ASSpolymer.

the finished products, generally termed ABS, has beenreported as 466 ell and between 500 and 575 0c.l3 ABSpolymers burn readily with the generation of densesmoke, which is characteristic of polymers containingaromatic rings.l4

In general, the ABS plastics exhibit good durability andresilience, as well as mechanical strength. In addition,these thermoplastic polymers demonstrate a favorablecombination of thermal, electrical and mechanical properties. The different properties may be selectively enhanced for specific applications by modification of thecomposition and/or relative proportions of the components. Such selective production is usually at the expenseof one or more of the other properties.9lO,lS

THERMAL DECOMPOSITIONThe type and quantity of the thermal degradationproducts from ABS plastics depends upon the combustion chamber utilized and the existing combustion conditions, as. well as the composition of the plastic. Themajority of the studies of the chemical analysis of thedecomposition products of ABS has employed a combustion tube with a ring furnace to decompose the material.This type of combustion system is similar to that used inthe DIN 53 436 toxicity test method (see below). Thisapparatus has the advantages of being easy to set up andoperate and requiring only a small sample for testing. Thecombustion atmosphere is easily changed, allowing forpyrolysis studies under inert conditions as well as thermoxidative degradation experiments. This type of furnacemay be programmed either to provide a constant temperature zone or to heat at a fixed rate. The disadvantageswithin this type of system also must be considered. One ofthe most important limitations is the large internal surfacearea on which combustion products may adsorb or react.In addition, the small sample size, listed as an advantageabove, may also be considered a disadvantage since itdoes not provide enough material to be conducive to aflaming test mode. During the actual combustion, airand/or other gases are fed into the system in a mannerwhich is contrary to the natural convective flow of airobserved under natural combustion conditions.16

Other furnace systems have been used less frequently incombustion product generation and analysis. Therefore,their pertinent advantages and limitations will be addressed when the experiments conducted in these particular systems are initially discussed.

In the following sections, the results of the chemicalanalyses of the combustion atmospheres of ABS areseparated according to the prevailing experimental conditions; i.e. inert or oxidative atmospheres. Although realfires occur normally under oxidative conditions, veryoften (especially in large fires) little or no O2 reaches thepolymer surface. Therefore, the study of the volatileproducts generated under inert atmospheres providesinformation on the molecular mechanisms of thermaldegradation under controlled experimental conditions.Also, information may be obtained on potential decomposition products that could be generated under vitiatedatmospheres or at the polymer surface in real fires.A summary of all the decomposition products of ABS

,


3/14

ACRYLONITRILE-BUTADIENE-STYRENE COPOLYMERS (ABS) 95

Table 1. Degradation products of ADS

'I = Inert pyrolysis atmosphere.0= Oxidative combustion atmosphere.identified under all the described experimental conditionsis presented in Table 1, in which the combustion atmosphere (inert or oxidative) is indicated as are thestudies which identified each product. The data indicatethat the majority ofthe products generated are characteristic of both inert and oxidative conditions.

ProductAcetophenoneAcids (-COOH)AcroleinAcrylonitrileAldehydes (total ) (HCHO)BenzaldehydeCarbon dioxideCarbon monoxideCresolDimethylbenzeneEthanalEthylbenzeneEthylmethylbenzeneFormaldehydeHydrogen cyanideIsopropyl benzeneC(- MethylstyreneP -MethylstyreneNitric oxideNitrogen dioxidePhenolPhenyl cyclohexane2- Phenyl-2-propanol3-Phenyl-1-propenen- Propyl benzeneStyrene4- Vinyl-1 -cyclohexene

Atmosphere' ReferencesI. 0 19o 24I, 0 19,23,24I, 0 19,23,24o 241,0 19,23o 13,17o 13,17,18,21,241,0 19I, 0 191,0 191,0 191,0 19o 241,0 13,17,21,24I, 0 19I. 0 191,0 19o 21o 21I, 0 19I, 0 19I, 0 191,0 19I, 0 19I, 0 19,23,24I, 0 19

was not detectable when this ABS sample was decomposed under the described inert pyrolysis conditions,

Hoff et al.19 analyzed by gas chromatography-massspectrometry the volatile products generated from twodifferent ABS samples decomposed under an inert atmosphere. The two ABS samples were identified asUgikral RA 10331 (Ugine Kulman; referred to as ABS-l)and Cycolac AM 25139 (Borg-Warner; referred to asABS-2). ABS-l and ABS-2 were pyrolyzed at 210C in aPyrex glass degradation tube with a nitrogen gas flow of20-30mlmin-1. The two ABS polymers demonstrateddifferent properties upon thermal degradation. ABS-2, theless thermostable of the two, decomposed considerablyfaster and formed nearly three times as much styrene asABS-l (Table 2). ABS-2 also released higher amounts ofother volatiles common to both plastics as degradationproducts, such as 2-phenyl-2-propanol, acetophenone, !Y.methylstyrene and isopropylbenzene. ABS-l did notgenerate any volatile products other than styrene inquantities greater than 100 flg g - 1 of ABS sample pyroIyzed. Some of the degradation products from the ABSsamples pyrolyzed under inert conditions contain oxygen,indicating the presence of oxygen-containing additives.The extensive list of volatile products identified by Hoffet at. 19 (Table 2) may be a result of the experimentalmethodology. The decomposition products evolved during the thermal degradation of the polymer were removedfrom the furnace/combustion site by the nitrogen gas flow(20-30mlmin-1), potentially reducing their further degradation. Also, the temperature used is typical ofthose inindustrial processing and, at most, may represent thetemperature conditions during the early phase of a fire.The different amounts of styrene evolved from the twoABS plastics may be related to a difference in theacrylonitrile content of the plastic, as it has been shownthat an increase in acrylonitrile content of styreneacrylonitrile polymers favors a higher styrene yield duringdegradation.2o

Table 2. Thermal degradation products of two types of ABSdecomposed in nitrogen 19Degradation product ASS.1 a ABS-2b

'Ugikral RA 10331, Ugine Kulman.bCycolac AM 25139, Borg-Warner.cValues are micrograms of product evolved per gram of materialpyrolyzed.dOxygen-containing products.

Thermal degradation in inert atmospheresThis section will cover experiments performed in atmospheres devoid of oxygen. This experimental atmosphere, in effect, limits the type of degradation productsformed, as oxidative decomposition is prevented.Sumi and Tsuchiya 17 measured the amount of hydrogen cyanide (HCN) generated when chopped ABS pipewas exposed to a temperature of 800C in a combustiontube furnace with a nitrogen flow of II min -1. Theevolved HCN was collected in a 0.4% sodium hydroxide(NaOH) trap and quantified with a cyanide ion (CN -)specific electrode. The maximum concentration of HCNgenerated under these experimental conditions was0.034 g HCN g-l ABS sample. The generation of HCNunder nitrogen did not differ quantitatively from thatgenerated under oxidative conditions (see below).The amount of carbon monoxide (CO) generated bypyrolysis of an ABS sample in a helium atmosphere wasdetermined by Michal.18 A horizontal combustion tubewas used with a moving ring oven set at 550C. Heliumgas was fed into the combustion chamber at a flow rate of30 ml min -1.The ABS used was identified as Forsan 752E(Kaucuk, Kralupy, Czechoslovakia): The results indicated that CO (as measured by gas chromatography)

Styrene2- phenyl-2 -propanoldAcetophenonedC(- MethylstyreneIso-propylbenzene4-vinyl-1-cyclohexaneBenzaldehydedEthylbenzeneEthylmethylbenzenen- Propyl benzenefJ- MethylstyreneAcrylonitrileCresoldPhenoldPhenylcyclohexaneDi-methylbenzene3-Phenyl-1-propene

2400C202312421333231118


4/14

96 J. V. RUTKOWSKI AND B. C. LEVIN

Thermooxidative degradationWhen sufficient oxygen (02) is present in the atmosphere,the primary gases formed during the combustion of r.lostmaterials, including ABS, are CO and carbon die xide(C02), Hydrogen cyanide generation from nitrogencontaining materials occurs during thermo oxidative degradation as well as during thermal degradation underinert atmospheres. Other thermo oxidative products fromABS are listed in Table 1.Table 3 summarizes the quantitative data in the litera

ture on the generation of CO, CO2 and HCN from ABSdecomposed under oxidative conditions. Although variations in test methods and ABS samples do not allow fordirect, quantitative comparisons between studies, generally more CO and CO2 were generated in the flamingmode than in the non-flaming one. The production ofHCN is approximately the same in both modes. Thefollowing sections will discuss the experimental studies oneach of these primary gases separately.Carbon monoxide. Gross et al.21 measured the amount ofCO generated during the combustion of two genericsamples of ABS. The composition ratio of the monomersin the plastics was the same for both samples; however, thesamples did differ in their physical characteristics andappearance. ABS sample 1 was 0.045 inches thick,weighed 1.33 kg m - 2 and was a semi-rigid sheet with adark gray matte finish. The second ABS sample was 0.08inches thick, weighed 2.35 kg m - 2 and was a rigid sheetwith a green polished surface. Each ABS sample wasplaced in an 18f t3 chamber and was subjected to thermalirradiation at a level of 25kWm-2 to the normal-useexterior (or front) surface. Gas analysis during the courseof experimentation was performed using colorimetricindicator tubes.b The maximum value from two separatedeterminations was reported; if these values differed by afactor of two, additional measurements were performed.

A ten-fold increase in the maximum generation of CO wasdemonstrated under flaming conditions as compared withthe CO levels generated in the non-flaming or smolderingconditions (Table 3).Similar findings are evident in the data from the

interlaboratory evaluation (ILE) of the National Bureauof Standards (NBS) Toxicity Test Method (Table 3).13Four laboratories used the same ABS material andessentially the same combustion conditions. However,variations in the results can be attributed to individuallaboratory modifications of the designed experimentalprotocol. An ABS copolymer in pellet form was thermooxidatively degraded in a crucible (or cup) furnace.The amount of air flow around the test sample in thefurnace is unknown and not controlled. This mayaffect the formation of secondary combustion productsand the completeness of combustion of evolved products.In the ILE of this test system, toxic gas analysis wasperformed by continuous flow, non-dispersive infra-redanalysis or gas chromatography for CO and CO2 andeither gas chromatography or specific ion electrodes wereused for HCN measurement. The observed CO levelsgenerated in the flaming mode generally exceeded levelsmeasured in the non-flaming mode of combustion of theABS; however, the differences were not as great as thosefound by Gross et al. (i .e. ten-fold).21 One of the four ILEparticipating laboratories who examined ABS reportedno significant difference in mean CO levels (laboratory#3). The reason for the differences between the data ofGross et al.21 and of Levin et al.13 is unknown.Sumi and Tsuchiya 17 used gas chromatography to

measure the amount of CO generated from chopped ABSpipe subjected to a series of increasing temperatures in acombustion tube furnace (Table 3). An initial increase inCO production was observed when the temperature wasraised from 500-600 0c. Further elevation of temperaturein 100 increments up to a maximum of 800C resulted ina decrease in CO levels. However, when the O2 con-

Table 3. Primary toxicants from experimental ABS thermooxidative degradationCombustion

Temperature orpparatustmosphereadiant f luxModeO'CN'co:zaadiant heatingir5kWm-2Sample 1b 21.8.0.629.8.2.7ombustion tubeir00C 48923 55172 36554 27697 800C 52458 550Cd 62 26'Air F100342 NF7 140.514144126 752154 396 NF60 184154 75676 1492 160 NF52-70 86686 18390 104 NF5 30.506 27bFlaming combustion.Non-flaming combustion.The numbers respond to the laboratory code of those laboratories par ticipating in the interlaboratory evaluation.Values are the mean standard deviation.


5/14

ACRYLONITRILE-BUT ADIENE-STYRENE COPOLYMERS (ABS) 97

centration was decreased by 50% by nitrogen, the COlevel increased two-fold over that found at 800C in airalone. The authors provide no indication of whether ornot the specimens flamed during their tests. However, asnoted above, the quartz tube furnace is not conducive toflaming combustion.Hydrogen cyanide.The generation of HCN, like that of CO,is dependent upon the combustion conditions. The datafrom the previously described work of Gross et aFl andLevin et al,u (laboratory #6 of the ILE) demonstrate nodifference between the evolution of HCN gas duringflaming combustion and during non-flaming thermaldegradation (Table 3).However, some data (laboratories 1and 3 ofthe ILE) indicate a somewhat greater productionof HCN in the flaming mode of combustion than in thenon-flaming one (Table 3).13 Sumi and Tsuchiya 17 demonstrated an increase in HCN production as thetemperature increased from 500 to 800C. The incidenceof flaming combustion, however, was not noted. For the300 C increase in temperature, the generation of HCNincreased approximately two-fold. At 800C, a 50%decrease in the O2 content of the combustion atmosphereincreased HCN production by approximately 20%.Carbon dioxide. Sumi and Tsuchiya 17 examined the effectoftemperature on the production of CO2, Carbon dioxideproduction was approximately five- to ten-fold greaterthan CO production (Table 3).The CO2 production at alltemperatures was approximately the same, ranging from223 to 297 mg g-l sample. The occurrence of flamingcombustion was not reported. In a combustion atmosphere of air and nitrogen (50/50; 10% O2), the yields of COand CO2 were of the same order of magnitude (52 and58 mg g - 1 sample, respectively).The data reported by Levin et al.13 also demonstrated a

preponderance of CO2 over CO generated in the combustion of ABS (Table 3). Except for the results reported byone of the participating laboratories (laboratory #3), CO2production was greater in the flaming mode than in thenon-flaming one. The mean CO2 values differed betweencombustion modes by a factor of two to three.

Organic volatiles. The classes of organic volatiles and thesummed quantities of the individual agents generatedduring the experimental thermo oxidative degradation ofvarious samples of ABS are presented in Table 4 anddiscussed below. Some of these compounds, such asacrolein and styrene, have known irritant properties andmay be present in high enough concentrations to compound the hazard of the fire environment by affecting themucous membranes and respiratory tracts of exposedindi viduals.Zitting and Heinonen23 measured the levels of styrene,

benzaldehyde, acrolein and acrylonitrile when CycolacT 10000 (Borg Warner) was decomposed in air at 250, 300and 350C in a combustion tube furnace with an air flowof 0.51 min - I. Styrene and benzaldehyde were measuredby gas chromatography, acrolein by high pressure-liquidchromatography (HPLC) of its 2,4-dinitrophenylhydrazone derivative and acrylonitrile by colorimetric indicatortubes.b The aromatic hydrocarbon yield (in this case,styrene) increased with each 50C increase in temperature,as did the aldehydes (benzaldehyde and acrolein) andacrylonitrile (Table 4). The amount of benzaldehydepresent exceeded the measured amount of acrolein onehundred-fold.Zitting and Savolainen24 determined the quantity ofvarious organic volatiles when Cycolac T 1000 was ther

mooxidatively degraded at temperatures of 300 and330C. Gas chromatographic analyses of styrene (listed asan aromatic hydrocarbon in Table 4) indicated that theconcentration doubled with the 30C increase in degradation temperature. The same relative increase was seenwith acrylonitrile and the organic acids which weremeasured by back titration with NaOH. The aldehydeconcentrations (acrolein, formaldehyde and total aldehydes) also increased with the increase in temperature;however, this increase was less than two-fold. Acroleinwas measured by the method of Cohen and Altshuller. 25Formaldehyde concentration was determined by achromatropic method26 and total aldehydes by sodiumbisulfite titration.Hoff et al.19 thermo oxidatively decomposed two ABS

plastics, Ugikral RA 10331 (ABS-l) and Cycolac AM

Table 4. Organic volatiles from experimental ABS thermooxidative degradationCombustion Temp. Aromatic

Reference Apparatus atmosphere (Oe) Hydrocarbons hydrocarbons Alcohols Ketones Aldehydes Acids Acrylonitrile19 Combustion tube Air080 I1g9 -, 2445 I1gg-'721199-'51199-'9 I1gg-' 191199-' 197821199-' 825611gg-'700l1g 9 -, 450l1g g-'1811gg-'1Ol1gg-'ombustion tubeir50 35mgm-3'-3mgm-3b1mgm-3 87mgm-36mgm-3mgm-320mgm-3 -3mg m-3mgm-3Combustion tubeir001.8ppm'- 8.4ppm'2.1 ppmdppm4.5ppm -1.5ppm3.9ppmppm

Styrene.bSenzaldehyde with smal l amount of acrolein.'Acrolein. formaldehyde and total aldehydes.dOrganic acids (as -COOH).


6/14


Table 5. Thermooxidative degradation products of two types ofABS19Degradation product ABS-"BS-Z' 2300c800


7/14

ACRYLONITRILE-BUTADIENE-STYRENE COPOLYMERS (ABS) 99

LC",(30min)(mgl-') LC",(30min+ 14d)(mgl-')F NFF40'C17.4 22.05.09.30.0(13.9-21.9) (17.6-27.5)12.3-18.3)13.9-26.9)26.5-34.0)15.6>38.05.638.438.0

(13.2-18.4 ) (13.2-18.4)20.833.00.83.337.6

(18.9-22.9) (22.8-47.8)15.9-27.2)23.1-47.9)22.1>32.59.30.9.D.9

(20.1-24.4 )(16.7-22.3)21.2-45.0)28.5 64.0(23.6-34.5)

(54.2-75.5)440'C~20.2

[15.1-25.2]9.0(4.7-17.3)

46.544002.18006032.57001609.8400 29.0400bAverage CO calculated for the 30 min exposure at the LCso (30 min).NBS large furnace.

The concentrations of CO and HCN in the combustionatmosphere are measured according to the protocol of theNBS test method and the results are presented in Table 7.More CO was generated in the flaming than the nonflaming exposure atmospheres and correspondinglyhigher COHb levels were noted in the exposed animals;however, the amount of CO was not sufficient to accountfor the lethality of the animals.29 On the other hand, themeasured HCN concentrations in the exposure atmospheres were sufficient by themselves or with the additionalCO to account for the deaths that were observed inlaboratories 1,3 and 6 in the flaming mode. In the nonflaming mode, laboratory 1 reported HCN and CO levels

that either individually or together were too low toaccount for the observed within-exposure deaths.Whereas, laboratories 3 and 6 reported HCN levelssufficiently high enough to have caused within-exposuredeaths, in actuality, no deaths were observed (Table 7).Under flaming conditions, the LCso values for ABS

were lower than those found for Douglas fir by approximately a factor of two (Table 8). This difference may notbe regarded as toxicologically significant. There is nosignificant difference in toxicity between ABS and Douglas fir when combusted in the non-flaming mode, eventhough the wood generates a significantly greater amountof CO under both combustion conditions. These results


8/14


support the earlier implication of the involvement ofanother toxicant (presumably HCN) in the toxic action ofABS combustion products.

University of Michigan test methodsTwo test systems, one static and one dynamic, each withdistinct characteristics, were developed by Cornish andcoworkers.30-32 The static system consists of a 15001stainless steel exposure chamber in which mice areexposed in a whole-body fashion to the combustionproducts for a period of 4 h. Combustion products aregenerated by placing the test material in a Vycor tubewrapped with resistance wire which produces a combustion temperature of 700-800 DC in 1-2 min. The LC50 isdetermined after a 7-day observation period. Additionalanimals may be tested and sacrificed following exposurefor histological/pathological examination.The dynamic University of Michigan test systemconsists of a flow-through Pyrex glass tube chamber with

side ports for the head-only exposure of rats. Thecombustion products are generated from materials placedin ceramic boat in a combustion furnace. The furnacetemperature is programmed to increase 3 to 5 DC perminute to a maximum temperature of700-800 0c. Carrierair is passed through the combustion apparatus at a rateof 11min - 1 and an additional diluting air stream of21 min -1 is added before the combustion products enterthe exposure chamber. The exposure period parallels thetime course of combustion product generation during thedegradation of the material sample.The amount of material consumed which kills 50% of

the exposed animals (LA 50)C from a number of studies onABS decomposed in both the static and the dynamic testsystems are listed in Table 9. The specimens were notidentified sufficiently to determine if the ABS materialtested was identical in all three studies. The data providesome information on the relative toxicity of fire retardant(unspecified) treated ABS (ABS-FR) and non-fire retardedABS samples and other materials used in these studies.Because of the great differences in design and operation ofthe static and dynamic exposure systems the specifictoxicity of anyone material in either system should not becompared. However, the relative toxicities of differentmaterials studied within one system should be comparable with that found in others. This does not alwayshold.32In the static exposure system red oak is a factor of twomore toxic than ABS and ABS-FR (Table 9). However,

Table 9. Toxicity data-University of Michigan test methodLAso values'taticynamiceference 18.70.81.22 20.20.32 9.62 28.92

"Amount (grams) of material consumed necessary to cause the deathof 50% of the exposed animals.bFire-retarded treated sample.

the results of the experiments with the dynamic exposuresystem rank both ABS and ABS-FR of approximatelyequal toxicity with red oak (Table 9). When one considersthe relative toxicities of ABS and polypropylene in thestatic exposure system, polypropylene is equal in toxicityto ABS; whereas in the dynamic exposure system it ismore toxic (Table 9).32Therefore, these data indicate thata change in combustion systems will affect differentmaterials differently.The LAso values from the experiments using both thedynamic and the static systems30.31 do not show muchdifference between the apparent toxicities of the combustion products of ABS and those of ABS-FR (Table 9).However, the specific mortality results, taken in a separatestudy,32 do indicate a difference. For example, when ratswere exposed in the static system to the combustionproducts of a 20 g sample of ABS- FR, 14 out of 15 animals(93%) died. Three of four rabbits also died when subjectedto the same exposure conditions.32 This mortality issomewhat greater than anticipated, since a 20 g sample isessentially equivalent to the reported LA 50 value.30.32 Incontrast, no rats or rabbits died from the exposure when a20 g sample of non-fire retarded ABS was decomposed,32although the reported LAso value is also approximately20 g30.31 or slightly greater (33 g).32 These inconsistencieswhich are evident between the mortality data32 and theLAso values30-32 for ABS and ABS-FR are notexplained.

University of Pittsburgh (PITT) methodThis dynamic test system was developed by Alarie andAnderson.33 A Lindberg furnace programmed to heat at20C min -1, starting at room temperature, is responsiblefor the generation of combustion products. The testmaterial starts in the non-flaming mode and will flamewhen the material's autoignition temperature (underthese combustion conditions) is exceeded. Animal exposure begins in the non-flaming mode when 0.2% of thematerial has decomposed. An air flow of IIImin - 1 ismaintained through the furnace during the combustion ofthe material. Total air flow entering the exposure chamberamounts to 201 min -1 following dilution of the combustion gases with cooling air. The 2.31 glass exposurechamber accommodates four mice in the head-onlyexposure mode. The biological endpoints/parameterswhich are determined include the RDso (i.e. the concentration of smoke that produces a 50% decrease inrespiratory rate), a sensory irritation stress index (SISI),asphyxiant effects, lethality and histopathologicalchanges. Analyses of the combustion atmosphere forprimary toxicants and O2 depletion are also performed.A very limited amount of ABS toxicity informationderived from this test system is available. Alarie andAnderson34 determined an LCso of 6.3g (1O.5mgl-1)dwhere the LCso is the amount of material loaded in thefurnace that is required to kill 50% of the animals withinan exposure period of 30 min and a post-exposure periodof 10min. The ABS sample used was only described as'standard acrylonitrile-butadiene-styrene'. The L Tso, thetime at which 50% of the animals die at the sample weightkilling 50% of the animals (or the sample weight nearestabove it), was recorded as 9 min. According to the

,

./


9/14


10/14


30-min exposure was 120000 ppm. min (equivalent to a30-min exposure to 4000 ppm). A CO2 index of 300%. min(equivalent to a 30-min exposure to 100000 ppm) willresult in the inhibition of respiration. It is evident from thedata (Table 11) that the CO was not present in amountsthat can solely account for the observed mortality. Thesedata suggest, as has the work of other authors, that CO isnot the only toxicant involved.

University of San Francisco (USF) methodThe apparatus for this method consists of a Lindberghorizontal tube furnace which is either programmed toheat at a rate of 40C min -1 from 200C to a maximum of800 C or set at a fixed temperature. The toxic combustionproducts which are generated are carried into the exposure chamber with or without a forced air flow. Theexposure chamber is a 4.21 hemisphere constructed ofpolymethylmethacrylate and provides for whole-bodyexposure offour mice. The biological endpoints which areused in this system are time intervals to various observedsigns of intoxication, such as incapacitation, staggering,convulsions, collapse or death. The mortality is recorded.Analysis of CO and CO2 in the chamber atmosphere isperformed at the conclusion of the experiment. Oxygenlevels, however, are monitored periodically during theexposure.The principal disadvantage to this system is that allendpoints are based on the subjective measurements ofthe one individual conducting the test. In addition, theseendpoints provide no insight into the cause of theobserved response. Since CO and CO2 are measured atthe end of the exposure period, limited inference can bemade concerning the role of these gases in the observedtoxic effects. Moreover, time to effect is not a truetoxicologic parameter and, in this case, is related to boththe time necessary for material degradation and thatrequired for the toxic insult to take effect. Toxicity is basedon the quantity of a substance required to elicit a certainmeasurable effect. The 'time to effect' is more appropriately considered important in the overall hazard assessment of a material in a particular use rather than being anindex of toxicity.Hilado and co-workers have amassed a sizeableamount of data on the toxicity of materials, including

Table 12. ADS toxicity data-USF method38

ABS. Only selected representative work by Hiladoet a1.3839 will be presented here in an effort to avoidredundancy.The data for ABS toxicity are presented in Table 12.The endpoints used by Hilado are the following:(1) Time to incapacitation (the first observation of stag

gering, prostration, collapse or convulsions);(2) Time to staggering (the initial observation of loss ofequilibrium);(3) Time to convulsions (the first observation of

convulsive-type movement);(4) Time to collapse (the first observation of muscular

support loss); and(5) Time to death (the time at which cessation of move-

ment and respiration are observed).The mean values and standard deviations of times tospecified animal responses presented in the studies byHilado et at. are derived from the means of two or morereplicate tests and reflect between-experiment variationrather than individual animal variability. The ABSmaterial used by Hilado et al. was identified as Dylel 702,natural, lot A54000 from Sinclair-Koppers, Co., Kobuta,PA. The data in Table 12 compare various experimentalconditions and their influence on the toxicity of ABS. Theamount of ABS pyrolyzed under each set of experimentalconditions was approximately 1.0g . From these results itis evident that the determination of the relative toxicity ofABS depends largely upon the test conditions used. Withthis test system the most toxic exposure condition is afixed furnace temperature of 800C with forced air flow, atwhich animals respond most rapidly. This short animalresponse time is most likely due to the more rapidgeneration of toxic decomposition products at this temperature. The addition of a forced air flow through thecombustion system into the exposure chamber alsoaugments the degree of toxicity, as reflected in the shortertime to response at 600 and 800 dc. The apparent COconcentration (ppm) in the animal chamber at the end ofthe exposure period did not correlate with the time toresponse. In fact, at 800C with forced air flow, theamount of CO was lower at the shorter response times, anobservation that was not evident at a furnace temperatureof 600C.In another study of ABS by Hilado and Huttlinger39 amaterial sample size-dependent phenomenon was de-

,

Experimentalconditions "ortalitylow Time to Time toime toime toime toumberOImin-')ncapaci ta tion (min)taggering (min)onvulsions (min)o llapse (min)eath (min)# dead/# exposed)ppm)3.52 3.46b (5)C6.48 6.44 (2)5.53 4.09 (5)8.09 (1)7.624.77 (5)0/20.832.10 (2).50 2.83 (2).39 1.90 (2).71 3.68 (2)8.52 5.08 (2)/8975 (2).84 0.06 (2) 3.39 0.41 (2).41 0.30 (2).53 0.16 (2)/8405 (2).30 0.04 (2) 2.93 0.21 (2).96 2.61 (2)/8700 (1).09 1.57 (2).46 1.99 (4)501.25 (4).02 0.72 (4).51 4.09 (4)6/16600 (1).69 0.05 (2) 1.72 0.09 (2).72 (1).80 1.53 (2)/8435 (2).47 0.49 (2)1.12 (1).76 0.26 (2).03 0.08 (2).47 0.39 (2)/8695 (2)

'Ramped heating mode.bMeans S. D.cNumber in parenthesesrepresentsthe number of experiments from which the meanswere used to calculate the means S. D.dFixedheating mode.


11/14

ACR YLO NITRILE-BUT ADIENE-STYRENE CO PO LYMERS (ABS) 103

Table 14. Relative toxicity of ABS compared with variouswood41

Table 13. Time to death dependency on sample size- USF testmethod 39

'Dylel 702 Natural lot A54000, Sinclair-Koppers Co., Kobuta, PA.bMean standard deviation.CLustran, Monsanto Co., St Louis, MO.

'Concentration-time products (ppm. min) for CO.bSamples were pyrolyzed with a ramped furnace temperature at 40Cmin -, to a maximum of 800C with no forced air flow through thesystem.

SUMMARYThe processes of pyrolysis and combustion under variousatmospheric conditions result in the production of manydecomposition products, some which are known toxicants and others whose toxicities have not been investigated. In addition, there may be toxic products generated that have not been identified. The conditions underwhich decomposition occurs are important in determining which degradation products will be generated and atwhat concentrations.

In the case of ABS, twenty-seven chemicalcompounds have been identified as combustionproducts.13,17-19,21,23,24 It is obvious that more com-bustion products are generated, but have not beenidentified or investigated to date. The degradation products which appear to be of primary toxicologic concernare CO and HCN.13,36 These two decomposition products have been detected under a wide variety of experimental conditions and are likely to be generated underreal fire situations. Although each toxicant alone may notbe present in adequate concentrations to pose a healththreat, recent studies at NBS on the interaction of CO andHCN indicate that these two toxicants may act in anadditive fashion.29 The data presented in this review havebeen evaluated to determine the extent to which CO andHCN together could account for the combustion producttoxicity. It appears that the concomitant generation ofCO and HCN is responsible for the toxic/lethal effects ofthe combustion products of ABS. However, the involvement of other toxicants cannot be precluded prior tocomprehensive analyses of the degradation products.In general, the degradation products of ABS from theflaming mode produced LC 50values that were lower thanthose of wood (either Douglas fir or red oak)(Fig. 3)/3,32,34 however, some findings to the contraryhave been noted,32 The magnitude of the differences inLCso values ranges from a two-fold difference in a staticexposure system13 to a ten-fold difference in a dynamictest system.34 The ABS LCso values for flaming combustion ranged from 10,5mg 1-134 to 22.1 mg 1- 113 for a30-min exposure period and 15.0 mg 1- 113 to28.5 mgl-128 for the 30-min exposure and 14-day observation period. In the non-flaming mode, both ABS andDouglas fir combustion products are equally toxic.13 Inthe non-flaming mode of combustion, the ABS LCsovalues ranged from 22.0 mg 1- 113for a 30-min exposureto 64.0 mg I-lor greater13,28 for the 30-min exposure and14-day observation period.

of CO was calculated to be 31 020 ppm. min, indicatingthat the amount of CO generated by the thermal decomposition of ABS was barely in the critical range (3000045000 ppm. min for a 30-min exposure period is characteristic of materials whose principal toxicant is C041).(Caution should be used, however, in extrapolatingconcentration-time products to other exposure timessince this value is not a constant.) While CO was found tobe an important contributor to ABS combustion producttoxicity,40 Hilado and Huttlinger39 did not consider it tobe the sole toxicant.

Weight o f samp leNumber o f

Chargedyrolyzedime to (g)eath (min).0500.05006.66 1.87b.00283.984138.524.812.00393 10.02 0.36.00170.87166.41 025.10384.103375.28 1.75.00080.987040.58 0.012.00271.001540.880.44.00262.89112.96 0.08.00011.99374.51 1.09.00355.99133.58013.00227.95491.27 0.0.99974.99648.101.12.00144.76249.17 0.66.00481.97572.65 0.11

ABS-2 fixed

ABS-l fixed

ABS-2C ramped

monstrated under fixed and ramped temperature conditions in the time to death of the animals (Table 13).Twosamples of ABS, Dylel 702 natural Lot A54000, SinclairKoppers Co., Kobuta, PA (ABS-1), and Lustran, Monsanto Co., St Louis, MO (ABS-2), were tested. As the massofthe pyrolyzed sample increased, the mean time to deathof the experimental animals decreased, a result which isnot surprising since the amount of toxic gases produced isdirectly correlated with the amount of sample decomposed. The decrease is directly correlated with the amountof sample decomposed. The decrease in time of deathobserved with both samples under ramped and fixedtemperature conditions was comparable.

Hilado and Huttlinger compared the relative toxicity ofABS to various woods under ramped temperature conditions and no air flOW.41The results are presented inTable 14. The data presented by Hilado41 show that theCO concentration generated by the ABS sample at thetime of death is less than that generated by any of the ninewood samples tested under similar conditions. Also, theconcentration-time product for ABS is lower than that ofany wood sample tested. The concentration-time product

Sample TemperatureABS-1' ramped

T ime todeath COone. t ime, (min)ppm)ppm. min) 18.527001020 14.0650002720 13.8214009390 15.0920005270 14.5002006980 14.911003920 14.761003580 14.3735008500 15.4203009710 15.562005790


12/14

104 1. V. RUTKOWSKI AND B. C. LEVINConcentration - response

Much more : More toxic : As toxictoxic than : thon wood ; as woodwood' .0.1 1000.0

I ~

IClass B

Closs C

r GramsUF cellUlose] tI~~125] ~rr[Q::;woodGM57 [Df"MO:"- ~PVC-A~t Closs A

GM 2' ......- PE [

GM 47 :u'\ IJ. PVC PE IPVC-CN ~.(G~49 GM 23I GMzl PEABS-3

TPCP-CN

'"Q)+::lC

...Q)'O+-0"'0~3...Q)"O1;;g 3.0.E3~cUo:>~

~ ~ 1

Q) ., .~ Co ,g 10.0a. +-IIIQ)...IQ)E~

r Io ~

g:J 30.01;;0"0..",03

to Closs 0Figure 3. Relative toxicity of ABS-Pitt method.33 Each point represents theamount of material (grams on the X-axis) which produced suf ficient smoke to kill50% of the animals (LC50) and the time (minutes on the Y-axis) required to kill 50%of the animals (L T50) using that amount of material. Reading the graph vertically,each material is classified in terms of potency while each material is classified interms of onset of action by reading horizontally. To combine both, parallelquadrants separate classes A-D. (Reproduced with the permission of the authors.Copyright 1981 Academic Press, Inc.)

CONCLUSIONSThe currently available literature on ABS combustionproduct toxicity provide a great deal of data; however,these results have only very limited usefulness. The fullassessment of the relative toxicity of the combustionproducts of ABS is complicated by a large number ofpotential variations in the chemical formulations of ABSand a variety of test systems and experimental protocols.However, the following generalizations are possible:(1) Many combustion products are generated during thedecomposition of ABS.(2) The principal gases which appear to be responsible forthe observed toxicity are HCN and CO.

(3) The thermal degradation products of ABS plasticsdemonstrate a measureable degree of toxicity. However, the toxicity of ABS combustion products iscomparable with the toxicity of the degradationproducts of other common polymeric materials.

AcknowledgementsThe authors gratefully acknowledge the help of Ms N. Jason, Ms M.Diephaus, Ms A. Durham and Ms C. S. Bailey who performed theliterature search and compiled the references. This experimental workwas performed under Contract CPSC-IAG-74-25, Task Order 84-8from the US Consumer Product Safety Commission, Bethesda, Maryland 20207, with Dr Rita Orzel as Project Officer. The opinionsexpressed herein are those of the authors and not those of CPSc.

NOTES

a Certain commercial equipment, instruments or materials areidentified in this paper in order to adequately specify the experimental procedure. In no case cioes such identification implyrecommendation or endorsement by the National Bureau ofStandards, nor does it imply that the equipment or materialidentif ied is necessarily the best available for the purpose.

b Indicator tubes, such as Draeger Tubes, do provide a convenientand inexpensive indication of the presence of measureableamounts of individual combustion products; however, i f interfering gases are present in the test atmosphere, a high degree ofprecision and accuracy is not attainable.

C The use of the convention, LC50 by Cornish et at. to describe avalue expressed in grams and not grams per unit volume isinappropriate. The convention LA", (lethal amount", ) is used as amore appropr iate subst itute in the discussion of these data.

d Alarie repor ts his toxicological results in grams of material. Forcomparison purposes the gram units have been converted tomg 1-1 units by the following equat ion:

wt (g)chamber air flow rate (I min-1) x exposure t ime (min)x1000mg/g


13/14

ACRYLONITRILE-BUTADIENE-STYRENE COPOLYMERS (ABS)

REFERENCES

105

1. G. E.Ferguson, Firegases. Nat. Fire ProtoAssoc. 27, 110 (1933).2. M. Windholz (ed.). The Merck Index, 10th edn, Merck and Co.,

Inc., Rahway, NJ, p.20 (1983).3. C. H. Bamford and G. C. Eastmond, Acrylonitrile polymers. InEncyclopedia of Polymer Science and Technology, H. F. Mark,N. G. Gaylord and N. M. Bikales (eds), Interscience, New York,

pp.374-425 (1964).4. M. Windholz (ed.). The Merck Index, 10th edn, Merc & Co., Inc.,

Rahway, NJ, p.209 (1983).5. W. M. Saltman, Butadiene polymers. In Encyclopedia of Polymer Science and Technology, H. F. Mark, N. G. Gaylord and

N. M. Bikales (eds), Interscience, New York, pp.678-754(1965).6. M. Windholz (ed.), The Merck Index, 10th edn, Merck & Co.,

Inc., Rahway, NJ, p. 1270 (1983).7. K. E. Coulter and H. Kehde, Styrene polymers (monomers). In

Encyclopedia of Polymer Science and Technology, H. F. Mark,N. G. Gaylord and N. M. Bikales (eds), Intersc ience, New York,pp.135-55 (1970).

8. A. E. Platt, Styrene polymers (polymerization). In Encyclopediaof Polymer Science and Technology, H. F. Mark, N. G. Gaylordand N. M. Bikales (eds), Interscience, New York, pp. 156-206(1970).9. C. H. Basdekis, ABS Plastics, Reinhold, New York (1964).10. A. Lebovits, Acrylonitrile-butadiene-styrene copolymers. InEncyclopedia of Polymer Science and Technology, H. F. Mark,

N. G. Gaylord and N. M. Bikales (eds), Intersc ience, New York,pp.436-44 (1964).11. Guide to Plastics, Property and Specification Charts by theEditors of Modern Plastics Encyclopedia, McGraw-Hi li, New

York, pp.4-5 (1983).12. A. H. Landrock, Handbook of Plastics Flammabi lity and Com

bustion Toxicology, Noyes Publications, Park Ridge, NJ(1983).13. B. C. Levin, M. Paabo and M. Birky, An interlaboratory evaluation of the 1980 version of the National Bureau of Standards

test method for assessing the acute inhalation toxicity ofcombustion products. Nat. Bur. Stand. (US) NBSIR 83-2378,Apri l 1983.

14. C. F. Cullis and M. M. Hirschler, The combustion of organicpolymers. In The International Series of Monographs onChemistry, J. E. Baldwin, J. B. Goodenough, J. Halpern andJ. S. Rowlinson (eds), The Clarendon Press, Oxford (1981).15. J. M. Lantz, ABS, Modern Plastics Encyclopedia 1984-85, J.Agranoff (ed.), McGraw-Hili, New York, pp.6-7 (1984).16. H. L. Kaplan, A. F. Grand and G. E. Hartzell, CombustionToxicology-Principles and TestMethods, Technomic Publishing Co., Lancaster, PA (1983).17. K. Sumi and Y. Tsuchiya, Toxic ity of decomposition productspolyacrylonitri le, nylon and ABS. Build ing Research Note No.111, Division of Bui lding Research, National Research Councilof Canada, April 1976.18. J. Michal, Determination of carbon monoxide in thermal degradation products of polymeric materials. Fire & MatIs 5,149-52 (1981).19. A. Hoff, S. Jacobsson, P. Pfaffi, A. Zitting and H. Frostling,Degradation products of plastics: Polyethylene and styrenecontain ing thermoplast ics-Analytical, occupational and toxicologic aspects. Scand. J. Work Environ. Health, Suppl. 2 a,9-60 (1982).20. N. Grassie and D. R. Bain, Thermal degradation ofcopolymers ofstyrene and acrylonitr ile: I . preliminary investigation of changesin molecular weight and the formation of volatile products. J.Polym. Sci. A-1 8,2653-64 (1970).21. D. Gross, J. J. Loftus, T. G. Lee and V. E.G ray, Smoke and gasesproduced by burning aircraft inter ior mater ials. Nat. Bur. Stand.,Washington, DC. Prepared for the Dept of Trans, FAA. ReportNo. NA-68-36 (DS-68-16) (June 1968).22. Standard Guide for Measurement of Gases Present or GeneratedDuring Fires, ASTM E 800-81, American Society for Testing

Mater ials, Phi ladelphia, PA (1984).23. A. Zitting and T. Heinonen, Decrease of reduced glutathione in

isolated rat hepatocytes cause by acrolein, acrylonitri le, and thethermal degradation products of s tyrene copolymers. Toxicology 17,333-41 (1980).24. A. Zitting and H. Savolainen, Effects of single and repeatedexposures to thermooxidative degradation products of poly(acrylonitrile-butadiene-styrene) (ABS) on rat lung, l iver, k idney, and brain. Arch. Toxicol. 46, 295-304 (1980).

25. J. R. Cohen and A. P. Altshuller, A new spectrophotometricmethod for the determination of acrolein in combustion gasesand in the atmosphere. Ana!. Chem. 33, 726-33 (1961).

26. NIOSH American Public Health Association, Washington,NIOSH Manual of Analytical Methods of Air Sampling andAnalysis, p.194 (1972).

27. B. C. Levin, A. J. Fowell, M. M. Birky, M. Paabo, A. Stolte andD. Malek, Further development of a test method for theassessment of the acute inhalation toxicity of combustionproducts. Nat. Bur. Stand. (US) NBSIR 82-2532, June 1982,133p.28. R. C. Anderson, P. A . Croce, F. G. Feeley and J. D. Sakura,Study to assess the feasibility of incorporating combustiontoxicity requirements into building material and furnishingcodes of New York State. Final Report to Department of State,Office of Fire Prevention and Control, Albany, NY, Reference88712, Arthur D. Little, Inc. (1983).29 B. C. Levin, M. Paabo, J. L. GUTman and S. E. Harris, Effects ofexposure to single or multiple combinations of the predominanttoxic gases and low oxygen atmospheres produced in fires.Fund. & Appl . Tox. (in press).30. H.Cornish, Toxicity ofthermal-degradation products of plast ics.Proceedings of Symposium on Products of Combustion of(Plastics) Building Materials, Lancaster, PA (March 1973).31. H. H. Cornish, K. J. Hahn and M. L. Barth, Experimental toxicology of pyrolysis and combustion hazards. Environ. HealthPerspect.11, 191-6 (1975).32. H. H. Cornish, M. L. Barth and L. H. Kolman, Comparativetoxicology of plastics during thermodecomposition. Utah University International Symposium (March 1976).

33. Y. C. Alarie and R. C. Ande'rson, Toxicologic and acute lethalhazard evaluat ion of thermal decomposition products of synthetic and natural polymers. Toxico!. App!. Pharmacal. 51,341-62(1979).34. Y. Alarie and R. C. Anderson, Toxicological classification ofthermal decomposi tion products of synthetic and natural polymers. Toxicol. Appl. Pharmacal. 57, 181-8 (1981).35. German Standards Insti tution, DIN 53436, Erzeugung Thermischer Zersetzungsprodukte von Werkstoffen unter Luftzufuhrund ihre Toxiko logische Prufung. Part 1, Zersetzungserat undBestimmung der Versuchstemperatur. Part 2, Verfahren zurThermischen Zersetzung (draft) (January 1979).36. M. G. Kimmerle, Aspects and methodology for the evaluation oftoxicological parameters during fire exposure. J. Fire &Flamm./Comb. Tax. 1,4-51 (1974).

37. C. Herpo!. Biologica l evaluation of toxic ity caused by combustion of build ing materials . Fire & MatIs 4, 127-43 (1980).38. C. J. Hilado, H. J. Cumming, A. M. Michado and J. E.Schneider. Comparison of animal responses to the combust ionproducts generated by two test procedures, the USF/NASAmethodology and the FAA/CAMI system. J. Comb. Tox. 4,325-59 (1977).39. C. J. Hi lado and N. V. Huttlinger, Concentration-response dataon toxicity of pyrolysis gases from six synthetic polymers. J.Comb. Tax. 5, 81-100 (1978).

40. C. J. Hilado and H. J. Cumming, Studies with the USF/NASAtoxicity screening test method: carbon monoxide and carbondioxide concentrations. J. Comb. Tax. 4, 376-84 (1977).41. C. J. Hilado, Carbon monoxide as the principal toxicant in thepyrolysis gases from materials. J. Comb. Tax. 6,177-84 (1979).


14/14

(

ABS acylonitrile

Documents

Transcript of ABS acylonitrile