title studies on explosive reaction of vinyl chloride mixed...

Title Studies on explosive reaction of vinyl chloride mixed withoxygen

Author(s) Suga, Masao

Citation The Review of Physical Chemistry of Japan (1960), 29(2): 73-80

Issue Date 1960-03-28

URL http://hdl.handle.net/2433/46779

Right

Type Departmental Bulletin Paper

Textversion publisher

Kyoto University

The Review of Physical Chemistry of Japan Vol. 29 No. 2 (1959)

TllE REa't E\c OF PHYSICAL C}IEat[sraY OP JAYAV, VoL. 29, Nu. 2, JIAR. 1960

STUDIES ON EXPLOSIVE REACTION OF VINYL CHLORIDE

MIXED WITH OXYGEN

By MASAU SucA

(Received February IS, 7960)

The explosion limits of vinyl chloride mixed with oxygen are determined a[ reduced pressures by the admission method. Isochors and isotherms of the explosion

limit and also the colors of ignition Homes ace shown. There is no essential difference between the behavior of vinyl chloride mined with oxygen and that of

ethylene mixed with oxygen except the generation of hydrogen chloride under the slow reaction. Patterns of these explosion limits are similar to each other. The

induction periods in the explosion peninsula of vinyl chloride mixed with oxygen have also been measured and i[ is found tlmt these periods are shorter than those

of ethylene mixed with oxygen. The apparent activation energy o{ the reaction is estimated from the explosion limits of vinyl chloride mixed with oxygen according [o Semeno6's theory and its value, 8, is about 34 kcal.

In recent years vinyl chloride has come to assume commercial importance as a monomer of

its polymer. There are a few datatl in the literature in regard [n °limi[ of inflammability",

but observations experienced under certain conditions can hardly 6e accepted as indicative of the

mnnomer~s safety range in other set of conditions. And with the increasing use in industry of

monomer vapor at elevated temperatures and pressures, it is important for information relevant

to these condiiions to be aa-ailable over sufficiently wide ranges.

In the previous papers?•a•+>, the explosion limit of ethylene mixed with oxygen and that with

the air were determined by the "admission method ". The present author has tontinvally studied

the explosion limit o[ vinyl chloride mixed with oxygen by the same method. for the purpose of

gaining the knowledge of the explosive reaction. These results are compared avith the results of

ethylene mixed with oxygen and briefly discussed. In this case, the results- obtained are unsatis-

factory from the practical point of view. However, it is interesting to mention [hat the explosion

peninsula is found also in the explosion limit of vinyl chloride mixed with oxygen, and it is made sure that the explosion peninsula involved in the explosion limit of ethylene mixed with oxygen

is not the sole nature oC ethylene itself as reported in the previous works),

Experimentals

I) Division Coating Technical Service Bulletin, I'inyl Chloride :Ifonorne., The Dow Chemical Co. plastics

2) h1. Suga. Dull. Chem. Sac. Japan, 31, 515 (1958) 3) N1. Suga, ibid., 32, 425 (1959)

4) \f. Suga, Thir Journal. 2g, 67 (1959)

i


74 T1. Saga

Materials The vinyl chloride employed in this experiment was prepared by dehydro-

chlorinating 1, 1-dichloroethane (b. p. 83S`C) through the anion of alkali in methanol solution

and then stored in a 'small cylinder. When necessary, it tuns fractionated, puri5ed by the action

of concentrated aqueous solution of sodium hydroxide, ammonical cuprous chloride solution and

dilute sulfuric acid. and then dried with soda lime and phosphorous pentoxide. The oxygen

employed was obtained from a commercial cylinder (purity: 99.4;.). after being passed through

a train of bottles wntaining a concentrated aqueous solution of sodium hydroxide, concentrated

sulfuric acid and phosphorous pentoxide.

Apparatus and procedure The apparatus used and the procedure adopted were of the

types described in the previous paperz> except [he removal of the mercury manometer previously

connected with the reaction vessel, taking into consideration the mercury vapor effects on the ex-

plosion limit. Next, since [he results were particularly influenced by the previous history of the reaction vessel and these effects not being reproducible, the following standard procedure was

adopted for the purpose of eliminating these effects.

Alter a previous run, the reaction vessel was heated [0 600°C and evacuated to 10-ammHg

for an hour.

Pressure change was measured with a movable membrane manometer containing a thin mica

or aluminium membrane, with the object of recording the pressure-time curve. Thereupon, the

displacement of membrane was recorded by an ink oscillograp6 and strain meter` cmmected with

a strain gauge pasted on the membrane. But on the general experimental runs of the determina-

tion of the occurrence of explosion. all manometers were removed in order to simplify the

operations.

Results and Considerakions

Observation The determination of the occurrence of explosion was made directly by

naked eyes. The results of observation are cla_ssi5ed thus:

I) After the ]apse of various induction periods (usually less than IOsec) luminous flames

(blue, green or yellow flames) were normally preceded by a glow which rapidly built up

into a bright Sash 511ing the vessel and accompanied by a sound.

1) After the lapse of relatively longer induction periods, pale blue flames were ohserved in

the dark room.

3) After the lapse of shorter induction periods, consecutive glows were observed for several

seconds in the dark room. This is not found in the experiments of ethylene mixed with

oxygen.

4) No change was observed.

I[ was noticed that the white products, presumably polymer, adhered sometimes on the inner

wall of the vessel after the reaction in case (3). Besides, after the explosion of case (1) black

products containing carbon were found on the inner wall of the tube connected with the reaction

• DS~P type mire resistance strain meter, Shinkoh Communication Industry Co., Ltd., Tokyo


Studies on Expbsive Reaction of Vinyl Chloride mixed with Ozygen 15

vessel. It is pointed nut that [his phenomenon differs from the case of ethylene mixed with

oxygen Fine point aside, cases (1), (2) and (3) are conveniently defined to be the explosions.

Recording of pressure variation with time The typical pressure-time curves which

were measured by means of the movable membrane manometer are schematically indicated in

Fig. 1. Curve (A) indicates a very rapid increase of pressure with luminous yellow, green or blue

lAl

~---- (C7

- ~

-..

~a)

(D) i~-- ~---'

r• is.

Time

1 Schematic pressure-lime curves i

!lames and this Corresponds to a normal ignition. Curve (B) shows the pressure change which

appears to be a two-stage ignition and occurs under conditions near the thermal explosion limit.

Dark blue flames which occured on the vessel wall grew immediately into luminous flames. It

is considered that curve (A) is essentially the same as curve (B). Curve (C) indicates a feebler

and milder increase of pressure due to slower reaction than in the case of curve (B) and is ac-

U

d

e E v F

700

650

fi00

550

500

0 iao zoo 300

Pressure,

400

mmHg

500 000

i

a z 3

Fig. 2 isoc6ars for ezplosion limit of vinyl ehloride•oxygen mixtures

Curve I-20g', 2-40$b, 3-60°0, 4-80% vin}•1 chloride

1

i

i i i

I

i

i


16 3I. Suga

companied with pale-blue (lames found under the explosipn peninsula. Curve (U) indicates no

increase of pressure with no flame. This corresponds to the so-called "slow reaction".

The explosion limit of vinyl chloride mixed with oxygen Isochors were determined

for the mixtures of the various compositions. 'Some of [he isochors are shown in Fig. 2. Only

the isotherm at 600°C is shown in Fig. 3. The curves shown in these figures exhibit the minima

i t

I 1

~ , ~~ i r r

r ~ , , r 500

i r r i

' 1 i ~

400 ~ I r I g , 'a'i i 1 I I T , 400

, t ~ , I c`7r o I m K '

, 1 S c 1 ' e'1 Er I

E ~ ~ ~ Green flames 1 Yellow (tames ' E CTI q c r ~ . 1 i e $ r o U j 3011 „~- 1 1 / G ~ / . /

1 1 I aV l E I

1 r I~ rI I r / L ~ 11 11 ~ / C '~)

\ r

1 / / l0U _ ~ ~ _ / 100

0 10 20 30 40 50 fi0 70 8D 90 1000 Volume .ti' of vinyl chloride in gas mixture with oxygen

Fig. 3 Isotherm for explosion limit of vinyl chloride-oxygen mixtures at 600'C

at the composition of about 50% vinyl chloride and the explosion limit situated about 3~95p

vinyl chloride (containing glow range}. The explosion peninsula is found a[ the definite tempera-

ture and in pressure condition is the mixture range 3--27 o vinyl chloride as indicated in the

lower closed curve in Fig. 2. The glow which was observed in the region of [he mixture of more

than 90 a vinyl chloride could not be recognized in the case of ethylene, and its presumed limit

is indicated with the broken line.

The explosion peninsula The explosion peninsula lies on the range bounded by the

first and second pressure limits of the explosions with which pale blue flames occur. Fig. 4 shows

isochor for the explosion limit o[ 20% vinyl chloride mixed with oxygen. Qn , Q and x indicate

respectively the cases Q), (2) and (4) of the observations. Also, Q indicates the case of the luminous flame propagating into the storage bulb through the induced capillary tube owing to

very short induction period. The broken line in Fig. 4 represents the explosion of 20% ethylene

mixed with oxygen.

AC this point, [he present author first considers that the explosion peninsula of ethylene mixed

with oxygen was not always related to the nature of ethylene but was caused by carbon monoxide

accumulated in the slow reac[ions•<). Then, the following mechanism is supposedr.):

5) C. N. Ilinshelwaod, The 6:inelicr of Chemical CGange, Oxford (1940), p. lG8

e Clue flmn


Studies an Explosive Reaction of Vinyl Chloride mined with Oxygen 77

700 ~d6

1tSa o$

o~ 3b1 0 ® e ~o

u 45'9 10.6 8.06 4.2 ~° m Fig. 4 Explosion limit of 10°S vinyl ~29 x x O O1'0 chloride mixed with orygen

174A G13 446 -- ~ .p (Induction periods in seconds are O O O O 6r x`~ x indicated by numbers beside $

SOLI 48b observation points .) E

,rte, 171.3126.1891 3,2 oa0 0 0 0 ,•x x x iti, o

1iD.J zsgo o ; ~ x 347.3 1956 ' 14

o O x x x ~

X•-+( x

a~00 50 l0U

Pressure, mmHg

CH;= HCOH HCOH -~ HCHO ~ HCOOH -~ H:CO,

CH.~ HCH ~ HCOH CO+H: CO+A: CO+H-O COo+H:O

In analogy to [his mechanism it is considered that in the case of vinyl chloride, carbon monoxide

is accumulated and is followed by the generation of hydrogen chloride in the process of slow

reaction.

CH, HCOH HCOH J HCHO --~ HCOOH ~ H.COz

CHCI ~ CIA ~ CI(i;OH ~ HCI+CO CO+H:O CO:+H:O

If [he essential difference be[a'eea the respective behaviors of ethylene and vinyl chloride mixed

700

U ~~ GO nZro

r 6(10

n550

Is~

~~

lox

Fig. 5

100 15(1 200 250 30U 350

Pressure, mmHg

(sochors foc explosion limits of vinyl chloride-oxygen mixtures

i

i


7s M. Suxa

with oxygen is the generation of hydrogen chloride, the discrepancy between the explosion limits

is well understood in conformity with the factsf that hydrogen chloride is an inhibitor of carbo¢

monoxide oxidation. Fig. 5 shows the various isochors far the explosion Iimil of amyl chloride

mixed with oxygen. which indicate the carious shapes of the explosion peninsula according to [he

composition of gas mixtures. That is to say, [he composition of gas mixtures considerably affects

the explosion limit, and particularly it is noticed that the explosion peninsula disappears at over

2i~ vinyl chloride. It is similar to the case of ethylene.

The explosion limit and induction period The results of the experiment with regard

to the induction period are shon•n by numbers beside the observation points in Fig. 4. The induction

period was measured by a stopwatch. Fortunately, the induction periods are fairly long in the

region of the explosion peninsula and are reproducible so far as precautions are taken in accord

with the standard procedure. Pig. 6 shows [he relationship between the explosion limit pressure

m x e a

v w

a

40

30

10

0

1~ •

`

. ~ ̀ \ ~T6e

Qb,

second pressure limit

The 5rst

~0C

pressure limit

o so loo Iso 200 ~0 3ao aso aoo asa

Induction period, sec

Fig. 6 Relationship between the explosion limit pressure and induction periods at the definite temperature for 2096 vinyl chloride mixed with oxygen

Table f The

for

induction periods

2096 ethylene or

(in sec) at the definite vinyl chloride mixed wi

pressures

th oxygen

and temperatures

10 30 50Press.mmFlg

Temp.'Ce[hylene

viaylchloride etkyleoe

viaylehloride ethylene

vinylchloride

590

600

620

640

617.i

368.7

128.9

33.1

260.0

207.4

IOLI

2t.9

2i1.0

1 u9.4

72.4

>os

124.8

89.1

48,6

Izs

x

x

3S.I

13.1

X

X

X

S.fi

x :non-explosion

6) D. E. Houre and A, D. Walsh, Trans. Faraday Soc., a0, 3i (1914)


Studies on Cxplosive Reaction of Vinyl Chloride mixed xah Oxygen 79

and Che induction period at various constant temperatures for 20 m vinyl chloride mined with

oxygen. In Table 1, the induction periods in the explosion peninsula of vinyl chloride mixed

with oxygen are compared with 2D;b ethylene mixed with oxygen.

It is interesting to.say that the explosion peninsula is more restricted but the induction periods

of the explosion are shorter in [he case of vinyl chloride than in those of ethylene. In other

words, it is considered that hydrogen chloride is an inbitor of carbon monoxide combustion but

will eventually be a promoter of the slow reaction of vinyl chloride.

The apparent activation energy of the reaction In Fig. 7 is shown the relationship

5.8

S.G

5.4

ti 5.2

4 on ^° 5 .0

4.8

A.fi

., '

Z

~3

Fig. 7 Logro P/T against 1/T curves for the ex-

plosion limit of vinyl chloride oxygen mixtures

Curve I=20%. 2=40%, 3=60%, 4=80% vinyl chloride

1.0 1.1 12 1.3 1.4 ]/Tx10'

between log„PJT and 1 JT for the third pressure limit of Fig. 2. These graphs are essentially linear

a[ a lower temperature than 600°C. This corresponds to the relationship IogsP/T=AJTtB. The

case is similar to that of ethylene mixed with oxygen. I[ the third pressure limit is thermal in

Table 2 Activation energies estimated from the explosion limits of vinyl chloride-oxygen mixtures

CrHxCI g; E-1.303x 2x 1.98ixA (kcal/mole)

80 3.70 33.8

fi0 3.74 34.2

40 3.71 33.9

20 3.i3 34.1

i

I,


80 \l. Suga

character, the gradient of these lines.:1, will be equal to E/2.303 x 2 x R. Thus. values for E,

the apparent activation energy. may be calculated as in Table 2.

By comparison with the values of ethylene mixed with oxygen?), it is found that the "E"

values of vinyl chloride mired with oxygen almost .analogous to those of ethylene. Of course, there will be no diretY meaning in the kinetic interpretation, but the explosion reaction of vinyl

chloride mixed with oxygen may apparently be considered as thermal in nature in this experi-

mental ranges as in the case of ethylene.

Summary

The explosion limit of vinyl chloride mixed with oxygen was determined at. reduced pressure

by the "admission method". Isnchors and iso[hertcs of the explosion limit and also the colors

of ignition flames are shown. it is considered that there is no esential difference between the

behaviours of tiny] chloride mixed with oxygen and ethylene mixed with oxygen except the

generation of hydrogen chloride under the s'Imv reaction. The induction periods in the explosion

peninsula of vinyl chloride mixed with oxygen have also been measured and it is found that

these values are shorter than those of ethylene mixed with oxygen. The apparent energy of the

reaction is estimated from the explosion limit of vinyl chloride mixed with oxygen according to

Scmenoff's theory and its mean value "L•" is about 34kcal.

The author wishes to thank Prof. R'asaburo Jono. Dr. ]iro Osugi and the late Prof. Ryo

Kiyama for their kind advised and encouragement throughout this work. He is grateful to D4r.

Hiromi[su -Naono and ~Ir. Junichi Rizuki for their assistances of the obsen•ations in part of the

present experiment.

Depart racnt of Chemistry Fatally of Scfeuce aad Literaftrrc

Ehime Uai•~~ersi+y .iJa(suymna, Japan

title studies on explosive reaction of vinyl chloride mixed...

Documents