Publisher’s version / Version de l'éditeur:
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
Research Paper (National Research Council of Canada. Division of Building
Research); no. DBR-RP-353, 1968-02-01
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
https://nrc-publications.canada.ca/eng/copyright
NRC Publications Archive Record / Notice des Archives des publications du CNRC :
https://nrc-publications.canada.ca/eng/view/object/?id=995d42b7-f173-477c-b29a-253e71a40234
https://publications-cnrc.canada.ca/fra/voir/objet/?id=995d42b7-f173-477c-b29a-253e71a40234
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.
https://doi.org/10.4224/40001535
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Thermal decomposition products of polyvinyl chloride
THERMAL DECOMPOSITION PRODUCTS OF
BOEYVINYE CHLORIDE
When plastics arc involved in n lire they may yicltl toxic dccomposition produc~s. Some cluantitalivc data on
thc decomposition prodclcts of plastics arc available in ~ h c lircrature, but i t is ditiicc~it to asscss the tlangcr from
the dilrcrcnt amo~!nts of variocls products ~ C C ~ L I S C of Ilic ;I[)SCI~CC of a suitable mcthod of cvaluarion. Tlic
authors have proposed a method of cv:tluntion busctl on pyl.olysis followcd by gas cliromatogr;tl,hic analysis
a n d have uscd it to asscss toxicity from various Ihcrmal dccomposition products of polyvinyl chloriclc.
liytlrogcn cliloridc was found to bc tlic main toxic tlccoml)osilio~i product.
Introduction
Tlic increasing use of plastics and other organic polyn~ers raises the possibility that when involved in a fire they niap yield toxic decomposition products in quantities sutxcient to produce a dangerous atmosphere. Some q ~ ~ a n t i t a t i v c results on the decomposition products of plastics are available in the literature, but it is difficult to compare data because of difl'erences in experimental methods and differences in presentation of experimental data.
I t is also difiicult to assess the danger from the different amounts of various decomposition products because of the abscncc of a suitable method of evaluation. I n view of thc above factors, the authors belicve that a nced exists for a systematic study to provide quantitative results on the voiatile deconiposition products of a wide variety of plastics in both inert and oxidising atmospheres. A necd also cxists for a method of assessing dangcr from tlie different quantities o f dccomposition products. This investigation was carried out to mcet thcse necds.
Polyvinyl chloride (PVC) is the plastics niatcrial that has probably rcccived the most attention froni firc authorities because of its relatively wide use and the possibility of its thermal decomposition leading to the formation of toxic gaseous products. This polymer was, therefore, selected for t-he present study.
Coleman & Thonias' determined the combustion products of PVC and other chlorinated plastics using a static system. The specinicns were decon~posed in a flask over a temperature range of 300 to 1000". The ratio of plastics to air was varied by using different weights of sample, and the conlbustion products werc analysed by conventional methods. Schries- heim2 employed a similar method over a temperature range of 250 to 550" and analysed the products of combustion by a niass spectrometer and chemical niethods. Stroniberg et
d."
studied the mechanism of thernial decomposition of PVC in vacuum. The volatile products were analysed by niass spectro- metry and found to consist almost entirely of hydrogen chloride and small proportions of bcnzene, toluenc and other hydrocarbons. Gilbert & Kipling4 investigated the carbonisa- tion of PVC and other vinyl polymcrs by decomposing the spcci~ncns, under vacuum, in a n inert alniosphcre and in air. Aflcr removal of hydrogcn chloridc thc rcsiduc was dc- composed and the products wcrc analysed by gas chromato- graphy.
In the present investigation the thcrnial deconiposition products of PVC wcre deterrnincd, firstly, in a n inert atnios- pherc and, secondly, in air. A flow system was adopted
instcad of thc static system ~ ~ s c d by other in\.cstig:itors.'
"
bccac~se the decomposition conditions during a n cupcrimcnt can bc more closely dcf ncd in a flow system than in a static system.Experimental Material
The PVC used in this s t ~ ~ d y was a co~i~mercially available, general purpose resin in powder form. I t did not contain any plasticiser.
Thermal decomposition
PVC was deconiposed in a furnacc through ~ h i c h inert gas or- air was passed and the gaseous products wcre collcctcd in
a flask. The sample was wcighed in a ceramic boat, to wli~ch a stainless steel wirc and a piece of iron werc attached, so that it could bc moved inside a 19-mm diamctcr t ~ ~ b c to tlic hot zone of the furnace by means of an extcrnal rn:Igllet. The furnace tcmpcraturc was maintained at isot1icrm;tl \.alucs of 350, 600 or 850'. Helium was i~scd to pro\.idc a n incrt atmosphere in onc scrics of expcrimcnts anti air was L I S C ~ I to provide an oxidising atniosphcrc in thc othcr scl.ics. After some preliminary s t ~ ~ d i e s a
as
flow ratc of 450 cm" pcr min was adopted. This corresponded to a nican linear vclocity of about 160 cni pcr min. The sample weights selccted for this study were 0.5 g in helium atmosphere and 0.5 and 0.25 g in air. The amount of air used during each experiment was slightly in excess of that required for coniplcte coni- bustion of 0.5 g of PVC to carbon dioxidc, water and liydro- gen chloride. The smaller weight of 0.255 was also uscd in a n effort to cxamine the influence of effectively increasing air supply o n the formation of deconiposition products.The higher boiling components of the decomposition products werc collected in a U-tube filled with glass beads and maintained at 0". The unreacted gas and the remainder of the volatile deconiposition products were collected in :I 3- litre flask that had been evacuated prior to the expcrimcnt. The sarnplc was kept in the furnace until tlie gases filled the flask. Each experiment lasted about 8 win.
Analysis
G a s chroniatog~~apliy was the main mcthod ilscd in tlic analysis of decomposition products. As thcsc were con1pt)scd of widely ditl'cring materials, thrcc clir-omatogr:~pIiic con- ditions had to bc employed (two with :I thermistor-typc
thcrmal conductivity ccll detector, with different colunins). First a ~nolccular sieve 5A (60- to 80-nicsli) colunin of length 2 rn and diamcter J in. (0.6cni) was uscd with the J. appl. Chem., 1967, Vol. 17, December
~ h c ~ m a l ccrn~i!~cci\ity dctcclur a1 room ccmpcrnt~~rc a n d a cilrricl. gas [hclillm) llow race of 60cm3/nlin for the ~ietcr- minalion of hydrogen and carbon monoxide. Second, n silica gel (60- to SO-mesh) column of the same length and diamctei as 1111: ~ i i o l c c ~ l a r sieve column was ~lszd a t 8 0 Sor the determination of carbon dioxide and chlorine. For quantitative analysis with the ther~nal conductivity dctector the response of the detector was calibrated against its res- ponse to pure gases.
A hydrogcn flamc ionisation detector was used for the third condition, with another silica gel (60- to 80-mesh) column of length 6 ft (1.8 m) and diameter
&
in. (0.3 cm) for determination of various hydrocarbons. A programmed temperature of 80" for 5 min., followed by a heating rate of IO0/min. up to 300", was used with a carrier gas (helium) llow rate of 30 cm3/min. Most of the peaks of the chromato- grams obtained under this condition were identified by comparison with retention times of pure materials. Identifica- tion of sonic of the peaks was further confirmed by trapping the ctllucnts, followed by analysis with a mass spectrometer. l o r q~~nntitativc analysis with the flame ionisation detector the response of tllc detector to n-butane was measured, and 'apprnsimatc substance specific cor~.ection factors' according to Kaiser's definition,' were used to calculate the response of other components.A calcium chloride trap was placed a t the discharge end of the furnace tube to remove water vapour evolved. Some hydrogcn chloride was absorbed by water vapour and was thcrcfore trapped; the rest of the hydrogen chloride was determined along with carbon dioxide by a gravin~etric nlethod in which ascarite was used. The amount of hydrogen chloride not absorbed in water vapour was determined by subtracting the amount of carbon dioxide previously deter- mined by gas chroniatography.
In a n effort t o determine chlorine, if present, a sensitive colorimetric detector tube (Kitagawa) was used. This de- tector is recon~n~ended for concentrations of I to 40 ppm.
Results and discussion
The results of the determination of tlie volatile thermal decomposition p~.oducts of PVC in a n inert atmosphere and in air arc given in Table I. They are reported as percentage by wcight of original polymer, and are based o n means of two 01' more rcpcat tcsts.
The main dccon~position product of PVC was hydrogen chloride. Almost all the chlorine in PVC was converted t o hydrogcn chlnl.ide. T l ~ c amount of hydrogen chloride pro- ct~~ccct in an incrt atmosphere approached the theoretical value
clpcctcd Crom I'VC. The ;1mount determined in an oxidising ncmosphcrc was sligl~tl!. Io\vcr. bccn~lsc wntcr vnpour pro- duced in a n a[m~)splicrc ol' ;lir ;~bso~.hcci some of chc hydrogen chloride. Tlic amount of hydrogcn cl~lo~.idc ; ~ b s o r h c ~ I i l l
\\':11er WilS il1tclltioll;l~~y l l e g l e ~ t ~ d ;111d tllc I o \ \ ~ I . v;lluc 1.c- ported bccausc tlie toxicity of hyclrogcn cliluri~ic p~.oduccrl at a fire is primarily due to the gas. l ' h c amount absorbcct in the water would condense on cool surfaces and play a comparatively minor role. If this amount had been included, the results for hydrogen chloride would have approached the theoretical value because no appreciable amounts o f other chlorine conlpounds were found in the decomposition products. Neither chlorine nor phosgene was detected. The minimum amount of chlorine that could have been determined in the present investigation was 0.003 wt.-od of the original polynler.
The results for the two different sample weights were very similar. At the lower sample/air ratio, slightly more carbon dioxide was produced a t the expense of hydrocarbons. The highest carbon n~onoxide concentration was found a t 600"; the amount was less a t 850" because of further oxidation to carbon dioxide.
Benzene was always found in the product in both a n inert atniospliere and air. The formation of benzene and other aromatics can be explained o n the basis of the removal of hydrogcn chloride from the polymer, producing a polyene system followed by the formation of stable six-membered r i n m 3
-
The similarity of the volatile deco~nposition products o f PVC found in two different atmospheres suggests that the decompos~tion mechanism is non-oxidative and essentially thermal. The only oxidation products found were carbon monoxide and carbon dioxide. Other oxidation products such as acids, aldehydes, alcol~ols, ketones, etc. were not found. This observation is in agreement with the author's experience with the deconlposition of polyethylene in air.
Efforts were made t o assess the danger that might arise from various quantities of thermal decon~position products o f PVC as determined in the present investigation. A literature search was undertaken to find data o n relative toxicity of different materials. The most detail available IS o n maximum allowable concentrations of gases and vapours at which n o adverse effect is e x p e ~ t e d . ~ - ~ Some information is available o n higher levels of concentrations such as those fatal to animals in a short time and the authors believe that such data would be more applicable t o fire situations than the maxi- mum allowable concentrations.
Difficulties were experienced in finding consistent data o n TABLE 1 Decomposition products of PVC Temp. of decomp., "C Wt of sample, g Atmosphere Decomposition HCI products. wt.-':.<; CO sample C o 2 H 2 CHa CIH, CIH, Benzene Toluene Residue 850 350 0.5 0.5 H e Air -- 57.9 47.2 - 1
.o
- 1.7 0.47 - 3.2 -- 0.32 - 2.5 - 5.9 5.4 0.87 5.1 40.2 600 0.5 Air 850 350 600 0.5 0.25 0.25Air Air Air
40.6 48.5 42.3 16.6 1.2 43.0 54.8 2.5 65.7 0.50 - 0.1 l 3.4 - 1.7 0.06 - 0.14 1.3 - 0.38 3.1 5.1 4.8 0.89 0.1 8 - 39.5 0.3 850 0.25 Air -24-3 18.5 99.7 0.37 2.4 0.03 0.41 1.3 0.31 - J. appl. Cllem., 1967, Vol. 17, December
Toxicity from dccontpositio~~ products of PVC
---
Tcmp. of dccomp., O c 350 600 8 5 0 350 600 S50 550 600 8.50
M't of samplc, g 0.5 0.5 0.5 0 . 5 0 . 5 0.5 0.25 0.25 0.25
.At niosphc'rc He Hc H c. Air Air Air Air :\ir Air
. . . -- -. - . . . . - . . - ~ . .- - .. . . - -
HCI 0.23 0.33 0.36 0.2') 0 0.25 0.30 0.26 11. I 5
1 '
co
-- - . . .- 0.002 0.07 0.03 0.002 0 .i)')Tolicit)., t = -
,.,
C O , 0.00003 0.00 1 0.00 1 0~00006 0.002 0.01 0,003C 6 l - f ~ 0.001 0.003 0.003 0.003 0.002 0.00 1 0.007 0,001 O.OOO(1
the relative toxicity of tlie main decomposition products or
PVC for very short exposures. Data on concentrations of gases and vapours considered fatal to man in a 30-niin ex- posure ( c I ) were assembled. Toxicity ( t ) due to a gaseous or volatile product was assumed to be proportional to its con- centration and to its relative toxicity (t,) i.e., t a c t , . Relative
1
toxicity was defined as: t , =
-.
CI
The toxicity of a deconiposition product, based o n both the nature of the niaterial and the quantity evolved, then becomes:
C
t a - . ('I
By comparing \,nlues of for different deconiposition protlucls cvol\cd undcr one set of co~ldltions, toxicity from each p ~ . ~ d i ~ c t co~lld LJC :ISSCSSC~. In order to present data o n toxicity in 11 consistent nianner the authors suggest using the equation:
wherc c, is the concentration of a volatile o r gaseous product evolved whcn one gram of original material is deco~ilposed and the decon~position products arc diffused in a volu~ne of
I m3, and c I is the concentration of gases lethal to man after 30 min.
Toxicity, based o n the above equation, was determined for the analytical results obtained in the current investigation and presented in Table 11.
The main decorllposition product of PVC was hydrogen chloride. The toxicity due to that product was found t o be three to ten times as great as that due to carbon nionoxide when PVC was decomposed in the presence of air. Even if it were possible to convert all the carbon of PVC to carbon monoxide the degree of toxicity (based o n the present method of evaluation) due to hydrogen chloride would be twice as great as that due to carbon monoxide.
The toxicity due to the dccomposition products of different polyn~ers can also be evaluated if, as a first approximation, synergism is neglected and it is assumed that the combined effect of toxicity 1s additive. The data for PVC arc also given in Tablc 11. Thc toxicity t l ~ ~ c to decomposition undcr different
experimental conditions c o ~ ~ l d also bc stutlictl i~sing this n ~ c t h o d of evaluation by varying temperature, samplc/nir ratio, etc.
The authors realise the limitations of the nic11iods of evaluation. Toxicities are not ~lecessarily aclditive, and tlic mechanisms of the toxicity for hydrogen chloride and carbon monoxicle are quite different. The most serious effects of hydrogen chloride are on the eyes o r respiratory tract. The toxic effect of carbon monoxide is due primarily to its affinity for haenloglobin, and this could lead to damage of tlie brain. Synergistic effects between different conditions, c.2.. I:cat 2nd carbon monoxide concentration, and oxygen depiction and carbon monoxide, were neglected because the present st:~tc o f knowledge o n this subject precludes consideration in a sinlple formula. 111 spitc of these limitations, the proposed
mcthods of evaluation should be of \:aluc. cspccially in the design of animal cxperinicnts to obtain data on toxicity.
This paper is a contribution from the Division of Building Rescnrch, National Kcsearch Council, Canada, and is pub- lished with the approval of tlie Director of thc Division.
Fire Kesearch Scct~on,
Division of Building Research, National Research Counal,
Ottawa, Canada
Received 22 February, 1967: a~ncnded manuscr~pt, 16 June, 1967 References
Colema~i, E. H., & Thornas, C . H., J. rrpy'. Clrerrr.. Lorrtl., 1953, 4.379
Sch;iesheim, A , , J. Res. rlnttr. Rltr. Strrtrd., 1956. 57, (4). 245
'
Strombcig, R. R., Strai~s, S. 6i Aclihn~n~ne;, B. G., J. Polvtrr.Sci., 1959, 35, 355
Gilbert, J. B., & Kipling, J. J., File/, Loud., 1962, 4 1 , 249 Kaiser, R., 'Gas Phase Chromatography, Vol. I l l . Tables for
Gas Chromatography', 1963, p. 101 (London: Buttcrworths)
'
Tliresl~old Limit Valucs for 1965, American Confcrcncc of Governn~cntal Industrial HygienistsFieldncr, A. C., Katz, S. H., & Kinncy, S. P., U . S . BIII.. A4itre.s, fcclr. Prtp. 248, 1921, 1,. 58
Jacobs, M. B., 'Analytical Chemistry of Intlustriill Poisons. Hazards and Solvents,' 1049, p.788 (Ncw York: Interscience)
