2834
INDUSTRIAL AND ENGINEERING CHEMISTRY
point of acetonyl acetone cannot be utilized, however, because of the acidity of the phosphorylated compounds. This acidity is sufficient to cyclicize the compound to 2,5-dimethylfuran and hence the reaction temperature must be kept a t 90” to 100” C. Table I1 shows result’s of treatment of phosphorylated materials with sodium iodide in acetonyl acetone. Sample 3 was prepared by treatment of hydroxyethylcellulose with phosphorus oxychloride in pyridine a t room temperature for 35 days. The product contains 9.7% phosphorus which is approximately 9 times that contained in a phosphorylated cellulose prepared in the same manner. This may be dne to the phosphorylat,ion occurring in the hydroxyl group of t’hehydroxyethyl chain, which should be the most easily accessible. Since all samples underwent a replacement of part of the phosphorus and part of the chlorine, it appears probable that some of each is attached in the primary position. CONCLUSIONS
The combined phosphorus in cellulose phosphate prepared by treatment of cellulose with urea phosphate is probably entirely in the form of a monosubstituted phosphate ester. There is no evidence of formation of other, more highly substituted products, as presumed by Coppick and Hall ( 5 ) . The structure of cellulose phosphate prepared by the treatment of cellulose with phosphorus oxychloride-pyridine mixture is similar, except, that in a typical sample approximately 23y0 of the phosphorus can be accounted for as a disubstituted phos-
Vol. 41, No. 12
phate ester. The chlorine also present in this product is probably attached directly to carbons of the glucose units. Nitrogen is found in typical samples, presumably due to pyridine. ACKNOWLEDGMENT
The authors are indebted to W. A. Pons, Mrs. V. 0.Cirino, and Miss E. R. McCall of the analytical section of this laboratory for the analytical data given. LITERATURE CITED
(1) Allison, J. B., and Hixon, R. M., J . Am. Chem. Soc., 48, 406 (1926). (2) Barham, H. N., Stickley, E. S., and Caldwell, M. J., Ibid.,68, 1018 (1946). (3) Buras, E. M., and Reid, J. D., IND.ENG.CHEM.,ANAL.K D . , 16, 591 (1944). (4) Ibid., 17, 120 (1945). (5) Coppick, S.,and Hall, TT. P., in “Flameproofing Textile Fahrics,” by R. W. Little, A.C.S. Monograph 104, pp. 182-3, New York, Reinhold Publishing Corp., 1947. (6) Gerrard, W., J. Chem. Soc.. 1946, 741. (7) Gerritz, H. I\’., J . Assoc. 0&. A y r . Chemists, 23,321 (1940). ( 8 ) Hess, K . , and Stcnzol, H., Ber., 68,981 (1935). (9) Kumler, W.D., and Eiler, J. J., J . Am. Cham. Soc., 65, 2355 (1943). (10) Nuessle, A. C., J . SOC.D y e i s Colourists, 64,342 (1948). I m . ENG.CHEM.,41, 2828 (11) Reid, J. D., and Mameno, L. W., (1949).
RECEIVED October 4, 1948. Presented before the Division of Sugar Chemistry 5nd Technology and the Division of Cellulose Chemistry a t the 114t,h Meeting of the . ~ M E R I C A X CHEMICIL S O C I ~ T PPortland, Y, Ore.
J
L. $1. WELCH, J. F. NELSON, AND H. L. WILSON Esso Laboratories, Standard Oil Development Company, Linden, N . J .
The physical properties of Butyl type vulcanizates are discussed, pointing out the variation between polymers containing a wide concentration range of isoprene, butadiene, piperylene, and dimethylbutadiene as diolefins.
A
LTHOUGH Butyl polymers can be made by copolymerizing isobutylene with a variety of diolefins ( 4 ) ,isoprene is the diolefin used in the commercial production of Butyl rubber (GR-I). The choice of isoprene as the diolefin was based on its availability, the polymer quality, and its applicability in the process. At the present time the major part of GR-I production goes into inner tubes, primarily because of its air-holding properties and excellent tear resistance. However, there is a growing market for Butyl in mechanical goods, curing bags, proofed goods, and farm tractor tires. McKinley ( 3 ) has reported that Butyl is a n outstanding polymer for use in wire insulation. This results from its good age resistance, low moisture retention, and good electrical properties. The other uses mentioned require good aging properties above all else. The well-known superior aging properties of Butyl are attributed to its low unsaturation, but low unsaturation results in an inherently slow cure rate which has, in some instances, limited its use.
Since the unsaturation in Butyl is a variable which can be adjusted a t viill, within rather wide limits, the manufacture of different grades of polymer has been a logical step in supplying the rubber industry with the polymer which suits its needs best Table I lists three of the grades of Butyl which are manufactured a t the prescnt time: GR-I, GR-1-18, GR-1-25, having unsaturations ( 1 ) of approximately 1.1, 1.6, and 2.2 mole 7 0 ,respectively. The specifications for GR-I, GR-1-15, and GR-1-25 illustrate the range of properties obtained in a 50-part easy processing carbon black (EPC) formulation. Of particular interest hero is the increasing cure rate as evidenced by the 300% moduli. The production (or consumption) trend established by the rubber industry as a whole is toward polymers of higher unsaturation. Although only 1.6% of the United States Butyl production was GR-1-15 in 1946, this increased to 32.0% in 1947 and to 75.0y0 in 1948. This trend results from the advantages of GR-I15 for inner tubes-namely, faster cure rate with attendant higher state of cure which isdesirable for better inner bube performance. The volume of GR-1-25 remains low because GR-1-25 compounds either tend to scorch or precure duringinner tube processing, or the over-all physical properties of a pncessable compound are unsatisfactory. From this it is evident that the proper balance between polymer unsaturation, acceleration, and physi-
INDUSTRIAL A N D E N G I N E E R I N G CHEMISTRY
December 1949
2835
cal properties must be mainTABLE TREND IN BUTYLPRODUCTION TO POLYMERS OF HIGHER UNSATURATION tained in order successfully t o utilize d i f f e r e n t B u t y l Recipe (Ciire at 307O F.) Parts polymers. Polymer 100 As mentioned above, a wide EPC 50 Zinc oxide 5 variety of diolefins have been Stearic acid 3 Tuads 1 copolymerized with isobutyCaptax 0.5 Sulfur 2 lene, but very few have given Tensile polymer quality equal to that Avera e Lb./S.q. I h h Elongation, 300% Moduli % of Total Mole $a Minimum, % Min., of isoprene. However, the Unsatu40-Minute 40-iqinute 20 Minutes 40 Minutes 60 Minutes U. S. Production Grade ration Cure Cure Min. Max. Min. Max. Min. Max. 1946 1947 1948 properties desired in a vulcani1.1 2500 650 GR-I 575 775 875 1125 1200 1500 9 8 . 4 6 6 . 4 22.2 zate are related very closely GR-1-15 1.6 2400 550 750 950 1125 1375 1475 1775 1 . 6 3 2 . 0 75.0a t o the use, and the concept of 2.2 2300 5 00 GR-1-25 900 1100 1325 1575 1750 2050 0.0 1.6 2.2 which properties of a vulo Includes GR-1-17 which has a Mooney viscosity of 60 to 70 and a cure rate equivalent to that of GR-1-15, canizate are important in a given use changes as experience is gained with a new product. I n the case of Butyl, experience in the inner tube field has shown the advantages of higher unsaturation. As Butyl is applied in other fields, the requirements will undoubtedly be 2300 d somewhat different and other types of Butyl will be found more suitable. In this study a comparison is made of some of the physical properties of Butyl A (butadiene), Butyl B (isoprene), -I 2100 i Butyl C (piperylene), and Butyl D (dimethylbutadiene) covering
r.
-
16
c
-
a
t; z
a fairly wide range of unsaturation for each type.
c"
1900
v)
TENSILE, ELONGATION, AND MODULUS
W
'-I
Isoprene. The vulcanizate properties of GR-I type polymers which are produced commercially were illustrated above in a 50part EPC recipe. For the following studies, however, a tube stock recipe is used in order more nearly to approximate the type of compounds and cure times employed by inner tube manufacturers. Table I1 shows t h e effect of unsaturation on the tensileelongation-modulus properties of isobutylene-isoprene copolymers over the range 1.2 t o 5.3 mole % isoprene in the copolymer. The molecular weight is held reasonably constant, and in this range the physical properties are essentially unaffected by small variations in molecular weight. In regard t o tensile strength, it is apparent t h a t overcure results both from excessive cure time, as shown by the 1.2 and 2.0 mole yo unsaturation polymers, and from too high a n unsaturation for the cure time and acceleration employed, as shown by the 3.1 and 5.3 mole yo unsaturation polymers for the 4- and &minute cures. Figure 1 illustrates this more clearly by plotting tensile strength as a function of time of cure and mole per cent unsaturation. The similarity between the curves indicates t h a t cure time and unsaturation function in a
u)
z
w
)TOO
I
I
1~001 0
I
I
I
I
2
I
I
I
t
4 MOLE % UNSATURATION
3
I
I
5
I
I
I
6
Figure 2 I
I
I
-
MODULI
I
l
l l l COPOLYMERS (TUBE STOCK FORMULATION I
e
I
i
l
l
l
l
1
1
OF ISQBUTYLENE-BUTADIENE
3 4 MOLE % UNSATURATION
R
Figure 3
I
2
3
MOLE T.
4
I
uNsAruwrioti
Figure 1
e
9
1200
I
I
4
I
I
I
I
I
I
I I I I 12 16 PO 24 TIME OF CURE,MINUTES AT 32O.F
a
Figure 4
I
I
28
I
I
32
I
Vol. 41, No. YZ
INDUSTRIAL A N D ENGINEERING CHEMISTRY
2836
PHYSICAL PROPERTIES OF DIMETHYLBUTADIENE AND PIPERYLENE TYPE BUTYLS (TUBE STOCK FORMULATION)
0.48
2
5
1.12
1.78
MOLE % OF DIMETHYLBUTADIENE UMSATURATION 3.05 0.48 1.12 1.78 3.05 0.48
2
2000
w, d 15'00
0
5
@c? w
(Ij
2 -m l rn
2
J
w
500
.#
w
i-
800
O
4816
4816
4816
4816
1.78
3.05
4816
4816
4 8
800
-'f
2 -600
600
06
W
1000
1.12
1000
1000
2500
z
400
q400
om 5200
200
*
0
n 4816
4816
4816
4816
'4816
TIME OF CURE, MINUTES AT 320"E
,MOLE % 1.3 2.2 3.8 4.8 5.6
OF PIPERYLENE UNSATURATION 1.3 2.2 3.8 4.8 5.6
i.3 2.2 3.8 4.8 5.6
i" L, F
g
22000
2
w -
* " g\ d J v)
z w t-
800
1500
9
1000
2 400
r-, W
60 0
0
500 8
8
8
8
8
J w 2 00 8 0
8
8
8
8
8
TIME OF CURE, MINUTES AT 320°F. Figure 5
similar manner. Increasing either cure time or unsaturation, of course, has the effect of increasing the number of cross links per unit volume. This in turn decreases the degree of orientation of the chains attainable on stretching. However, different diolefins impart different tensile-modulus-cure time rclittions t o Butyl polymers, so that the state of cure or number of cross links per unit volume is not the only variable. Butadiene. I n the early development of Butyl rubber, butadiene was used as the diolefin, but it mas later found that isoprene was more suitable in &he process. There was very little choice in so far as physical properties of polymers of low unsaturation (about 1.0%) were concerned. I n some properties the butadiene polymers appear superior, but in other propertics the isoprene polymers are superior, as will be disoussed later. Table I11 gives the properties of a series of isobutylenebutadiene copolymers ranging in unsaturation from 0.6 to 5.6 mole %. The tensile strength of the vulcanixates, as shown more clearly in Figure 2, decreases with increased unsaturation in much the same manner as the isoprene polymers. The somewhat higher tensilo
OF UXSATURATION OK I S O B ~ ~ T Y L E N E - ~ COPOLYMERS SOPRRM~ TABLE I1, EFFECT
Tube Stock Rwrpe (Cure at 320" F,) Parts
Tensile, Lb./Sq. Inch 4 8 16 Mole c/c ininminminIsoprene in Jteudinger Utes utes utei Mol. Wt. Copolymer" i j i o 2010 1710 37,000 1.2 igno 2030 1510 2.0 45,000 1670 18311 40,ono 3.1 1640 1490 35,000 5.3 a Iodine-mercuric acetate method ( 1 ) .
ElongationL &4 8 16 minminmin. utes utea rites goo 780 620 790 700 470 680 450 693 440
_______I_
O F UNS.4TURATION TABLZ T I L EFFECT
3007 \ I O C ~ U ~ U ~ Lh.O/&. Inrh 4 8 16 minmmminUtes utes Ute8 140 370 680 190 610 880 580 1070 050 080
ON I s O B U T Y L E ; v E - B U r A D r ~ ~COPOLYMERS e
Tube Stock Recipe (Cure a t 320° F.) Parts Polymer in0 SRF 50 Zinc oxide E: Sulfur 2 Tuads I Captax 0.6 Tensile, Lb./Sq. Inch 4 8 !6 Mole o/a minrninminButadiene in Staudinger utes Utes utes Copolymers Mol. Wt. 68,000 , , 2310 2400 0.6 , 2320 2400 1.0 47,000 37,000 igio 2140 2210 1.8 43,000 2080 2180 2.2 2650 zino .. 3.3 4n.onn 40,nnn 1870 1820 ,~ 3.7 1700 1640 .. 5.6 34,000 a Iodine-mercnrio acetate method ( 1 ) .
Elongation& 8 !6 minminutes utei Utes .. 740 700
4 rnin-
iio
Bio
510 420
740
mo
550 570 390 340
ea0 56c
470 .~
3007 Modolue,
-~ Lb.pSq. 4
minutes
.. 470
I n o h 8 1.8 minmmutes ut% 440 570
so0
iSn
790 io40
pilo
1400
1240
1450
530 970
BO61 I240
December 1949
INDUSTRIAL AND ENGINEERING CHEMISTRY
2837
TEAR RESISTANCE OF ISOPRENE AND BUTADIENE T Y P E BUTYLS (TUBE STOCK FORMULATION ) AGED (48 HRS. AT 250°F IN OVEN )
UNAGED
f
G R-l I .o
37,000
GR-1-25 2.1
I .5
.BUTYL A GR-I 3.1 MOLEXUNSAT 1.0
37,000
55,000
35,000 STAUD. MOL.WT
4 8 1632
4 8 1632
GR-1-25 2 .I
BUTYL A
1.5
3.1
*
4 8 1632 4 8 1632
4 8 1632
4 8 I632
4 8 I632
CURE T I M E , MINUTES AT 320°F
CURE T I M E , MtNUTES AT 320°F
,
4 8 I632
g 200 22 150
200 I50
4 8 I632
4 8 1632
4 8 1632
4 8 1632 4 8 I632
CURE T I M E , M I N I I T E S AT 320°F
4 8 I632
4 8 I632
4 8 I632
CURE T I M E , M I N U T E S AT 320°F Figure 6
XT
atrengths for the polymers of low unsaturation are attributed, in part a t least, t o the higher molecular weights of these polymers. The relation between unsaturation, modulus, and time of cure is illustrated in Figure 3. The rapid increase in modulus with unsaturation is t o be expected by analogy with the isoprene polymers, but the butadiene type polymers seem to be capable of obtaining a higher modulus while maintaining a somewhat better tensile strength. For example, in comparing the isoprene and butadiene type polymers of 5 + mole % unsaturation a t the 8minute cure in Tables I1 and I11 the butadiene type shows a tensile of 1640 pounds per square inch and a 300% modulus of 1450 pounds per square inch compared to 1490 pounds per square inch tensile and 980 pounds per square inch modulus for the isoprene type polymer. The retention of tensile strength with time of cure for isoprene and butadiene polymers of essentially the same unsaturation and molecular weight is shown in Figure 4. It is apparent that the butadiene type polymers have less tendency to overcure in the tube stock recipe. I n other formulations this tendency of isoprene type polymers to overcure is not so pronounced. Dimethylbutadiene and Piperylene. The vulcanizate properties of the dimethylbutadiene type Butyl polymers are similar
t o the corresponding isoprene polymers. Typical results for copolymers containing from 0.5 to 3.0 mole yodimethylbutadiene are shown graphically in the upper part of Figure 5. I n this series the molecular weight of the polymers decreased as the unsaturation increased, so that the tensiles and moduli for the two higher unsaturation polymers could normally be expected t o be somewhat higher at a higher molecular weight level. In the lower part of Figure 5, similar data are presented for a series of isobutylene-piperylene copolymers in which piperylene content ranges from 1.3 to 5.6 mole %. The molecular weights of the various polymers are reasonably constant. It is readily apparent that the tensile strengths are not a s high as those obtained with the other diolefins, but on the other hand the moduli for a given unsaturation are in about the same range as the other diolefins. TEAR RESISTANCE
It was shown by Turner et al. (6) that although carbon blacks do not reinforce Butyl so far as tensile strength is concerned, they do improve tear resistance and increase the modulus of Butyl compounds. Although the tear resistance of Butyl is much higher in channel black stocks, the data presented in Figure 6 on the tear resistance of isoprene and butadiene type polymers were
Vol. 41, No. 12
INDUSTRIAL AND ENGINEERING CHEMISTRY
2838
REBOUND AND GOODRICH FLEXOMETER DATA CARCASS STOCK
GR-l I .o
40
100
GR-1-25
TREAD STOCK
GR-l
BUTYL A MOLE %-
2.1
I .5
40 100
40 100
3 * 1- UNSAT.
40 100
I,o
40
100
GR-1-25 2.1
40
100
BUTYL A
3 .I
I .5
40
100
40
100
TEMPERATURE OF BALL, "C.
60 120
60 120
60 120
60 120
60
I20
60 120
60 120
60 120
TIME OF CURE, MINUTES AT 307°F. Figure 7
obtained using the standard tube stock recipe containing 50 parts of semireinforcing (SRF) black. The tear values are therefore somewhat lower than can be obtained by different compounding, but the comparative data of Figure 6 tend to show polymer differences. Of the four polymers tested, two were commercial plant products, GR-I and GR-1-25, and the other two mere pilot plant samples of the butadiene type. The lowei uiisaturation Butyl A (1.5) was of higher molecular weight: 55,000 compared to 35,000 to 37,000 Staudinger for the other three. The upper two sets of data give the room temperature tear for both uriaged and aged samples (48 hours a t 250" F.). Butyl A shous better tear resistance, both aged and unaged. The hot tear values given in the lower portion of Figure 6 are also somrwhat higher for Butyl A. Tear resistance decreases rapidly with increased cure times above 8 minutes, and somewhat lower tear values ai e shown for the polymers of higher unsaturation.
slightly higher rebound than Butyl -4a t both 40' and 100" C. in the following carcass and tread stocks:
REBOUND AND THE GOODRICH FLEXOMETEK
For many uses it is desirable for vulcanizates to maintain good flexibility and high retraction rate a t low temperatures. Figure 8 compares the retraction rate of Butyl A and Butyl B tube stork vulcanizates at -30" F. These data are obtained as follows:
Polymer Zinc oxide SRF black Stearic acid
EPC black Tuads Captax Sulfur Dibenzo ether
Carcass Stock, Parts lo?
Tread Stock, Parts 100
J
5 3
363
..
1 0.5 2.0 5.0
50 1
0.5 2.0 5.0
The temperature rise observed in the Goodrich flexometer test also shows a.n advantage for Butyl B. The polymers of higher unsaturation for both t,ypes give slightly less temperature rise. It, t,herefore appears that isoprene type polymers have somewhat. better hysteresis properties t)han have the butadiene type. RETRACTION RATE AT LOW TEMPERATURE
It was shown by Haworth and Baldwin (5') that Butyl polymers of higher unsaturation have better hysteresis properties as measured by rebound and Goodrich flexometer data. Figure 7 compares the isoprene and butadiene type of Butyls in these tests. The upper part of the graph shows that Butyl B gives
h standard 2-inch T-50 dumbbell of the vulcanisate is stretched t o 3.25 inches (62.570) and placed in an oven for 15 minutes a t
December 1949
INDUSTRIAL AND ENGINEERING CHEMISTRY
2839
c
GR!-(I.~ %ISOPRENE)
40
R E T R A C T I O N T I M E A T -30.F..
Figure 8.
SECONDS
Retraction Rate at Low Temperature 35.
e
0
150 F. After this relaxation time, the specimen is immersed in an alcohol bath at -30" F. for 2 minutes. The sample is then released and the time required for the sample to retract to 2.75 inches (or 37.5% extension) a t -30' F. is recorded.
nl
Isoprene polymers are superior to butadiene polymers in retraction rate a t any given modulus level. The rather wide range of values given for each type results from the variation in molecular weight, unsaturation, and cure temperature. However, these other factors were found to be of much less importance than modulus or state of cure. This is, of course, in line with the T-50 test for measuring the state of cure of natural rubber vulcanizates. At a low modulus level the retraction rate is low, but the retraction rate increases rather rapidly with increased moduli up to 700 to 900 pounds per square inch and then little further improvement is noted with increased modulus. These data, of course, apply only to the particular formulation given. Various blacks and dasticizers have marked effects on retraction rates, but a discussion of compounding variations is outside the scope of this paper. I n any formulation, however, a relatively high state of cure is essential for high retraction rate a t low temperatures.
6 8 AGING CYCLES
le
IO
'
(CYCLS CONSISTS OF 8 HRS.lN 100PSIG.STLAM--BHRS.80
[email protected] AT P6O.F)
Figure 9
types of Butyl appear t o be more resistant to the hydrochloric acid than to nitric acid. I n general, Butyl C and Butyl D are inferior to Butyl A and Butyl B in resistance to nitric and hydrochloric acids. This is more noticeable with the lower unsaturated stocks of each type. All four types of polymer are fairly resistant to attack by phosphoric acid, although Butyl C seems to be somewhat inferior t o the others. I n summary, these data indicate that Butyl C is the least acid resistant of the four types, closely followed by the dimethylbutadiene type Butyl. Little, if any, significant difference can be found between Butyl A and Butyl B in acid resistance.
TABLEIV.
RESISTANCE OF BCTYL POLYMERS TO ACIDS Recipe (Cure a t 320" F.) Parts Polymer 100 Wyex 20 Zinc oxide 5 Sulfur 2 TuadP 1 Captrtx 0.5 7 % Tensile Strength Retained5 Original Sulfuric Hydrochlorx Phosphoric Tensile, Acidb Acid Acid0 Acid C
ACID RESISTANCE a
(4,5 ) that Butyl rubber vulcanizates are much more resistant to acids than is natural rubber. Table IV shows that the type of diolefin in Butyl affects the acid resistance. The percentage of tensile strength remaining after 2 weeks of immersion in concentrated hydrochloric, nitric, and phosphoric acids is a good index t o the degree of attack by these acids. Sulfuric acid caused such severe disintegration of the polymer t h a t the test in this acid was discontinued after 7 days of aging. These data indicate that the isoprene and butadiene type polymers are somewhat better than the piperylene and dimethylbutadiene type polymers in sulfuric acid resistance. These four
4
It has been shown previously
Nitric
Polymer ButylA ButylA Butyl B Butyl B Butyl C Butyl C Butyl D Butyl D
Mole % Unsatd. 0.80 2.58 0.60 2.25 1.18 2.83 1.06 2.55
8 minutes 1900 2835 3310 2685 1875 2795 2055 2755
16 ininminute Utes cure 2695 13 2915 D 3265 D 1500 D 3145 D 3395 D D 1320 3170 D
minute cure 17
D D D D D D
22
mlnute cure 54 56 57 48 24 41 30 56
minute cure 62 61 67 66 31 44 40 53
minute cure 73 50 74 51 45 48 27 70
minute cure 64 40 73 58 57 60 37 77
min- minute ute cure cure 82 82 85 89 91 87 84 83 74 80 78 77 80 96 87 88
Standard Scott dumbbells were completely imrrersed i n t h e fluid during t h e aging, then
the. dumbbells were water washed and soaked, a n d finally dried before pulling t h e dumbbells. b C
Acid aging 7 days a t room temporatwe (75' to 88' F.); D = duiubbells disintegrated. Acid aging 14 days a t room temperature (75' to 88O F.).
2848
INDUSTRIAL AND ENGINEERING CHEMISTRY STEAM-AIR AGING
The rubber compound used in curing bags for tire production must be capable of withstanding high temperature cyclic steam and air aging. Natural rubber compounds tend to harden in this +rvice with failure resulting from brittleness and eventual cracking, GR-I, on the other hand, has a tendency to become soft and the inside of tho curing bags deteriorates to the point of plugging -he valve stems. Figure 9 shows the effect of diolefin type on thr *team-air aging of Butyl polymers. Each cycle in this accelerated Tzst consists of 8 hours in 100 pounds per square inch gage stram rnd 8 hours in 80 pounds per square inch gage air a t 260 F. Thp .hange in Durometer hardness is an index of the resistance of Butjl compounds t o this type of aging. Considering first the GR-I type, improved results are obtained with inrreasrd unYaturation. The GR-1-25 compound is shown to have a Durom+ter hardness of 45 after twelve cycles compared to three cycles ro reach the same hardness in the GR-I compound. Piperylene is in the same class as isoprene. On the other hand, butadiene is much superior. Very little loss in Durometer hardness is noted inr the 2.2y0hutadirne Dolvmer and an increase in hardnev i s
Vol. 41, No. 12
observed with polymers above 3.0% butadiene. IIigher iwprrne polymers, on the other hand, continue to soften. ACKNOWLEDGMENT
The authors wish to acknowledge the assistance of their colleagues in the rubber laboratory of the Standard Oil Development Company in the evaluation of these polymers. I n particular, the work of L. T. Eb>--,D. J. Buckley, and F. P. Ford on developing methods for testing IOK temperature properties made that: portion of the paper possible, For other evaluations the authors are indrbted to S. B. Robison, A, W.Hubbard, and W. F Fischer. LITERATURE CITED
(1) Gallo, 8. G., Wiese, H. K., and Nelson, J Fa,IND.ENC.C : m x . " 40, 1277-80 (1948). (2) Haworth, J. P., and Baldwin, F. P., Ibl:d., 34,13014 (1942). . N o . 25, 52 (Dec. 22, 1945). (3) McKinley, R. B., Electric W ~ r l d 124, !4) Thomas, R. M.,Lightbown, I. E., Sparks, W. J , , Frolich, P. K., and Murphree, E. V., IXD.ENG.CHEM.,32, 1253-92 (1940). (5) Turner, L. B . , Haworth, J. P., Smith, W. C., and %appg11,. L , Ibid., 35, 958-63 ( I 943). RECEIVED August 7. 1948. Presented before the meeting of the IXvision of Rubber Chemistry, A X E R I C A N CHEYICAJ.~ SOCIFTY,Lon Angeles, Calif., July 22: 1948.
Stabilization of Cyanogen Chloride 91. S. KHARASCH, AL4N R. STILES, ELWOOD V. AND
JEXSEN.
DANIEL W. LEWIS
University of Chicago, Chicago, I l l , Commercial cyanogen chloride, stored in steel containers, contains water and soluble iron compounds which catalyze its transformation into cyanuric chloride. Sodium pyrophosphate, 2 to 5 % by weight, inhibits the effect o f these impurities and is a very effective agent for stabilizing cyanogen chloride.
C
YAXOGEK chloride, when stored in steel containers, is gradually transformed into a nonvolatile solid consisting mainly of cyanuric chloride. This process is exothermic and is markedly accelerated with increasing temperature, When the material is stored in large containers so that the heat cannot be dissipated, the temperature and consequently the pressure may 4se rapidly; several explosions of containers filled with cyanogen ahloride have occurred during storage. i n an attempt to find a substanc? which would inhibit the oolymerization of cyanogen chloride, a study was made of the Factors which influence the rate of the polymerization. On the hasis of these findings, compounds were selected for testing as potential stabilizing agents for cyanogen chloride. One compound, sodium pyrophosphate, n a s found to be v x y effective in inhibiting the polymerization of cyanogen chloride a t temperalures between 25' and 125' C. FACTORS INFLUESCINC, THE POLYRIERIZATION O F CYANOGEN CHLORIDE
Commercial cyanogen chloride, prepared by the chlorination of hydrogen cyanide, is about 98% pure. The following impurities were found to be present, varying from batch to batch within the limits given: hydrogen cyanide, 0.37 to 2.350/00; hydrogen chloride, 0.06% ; arid water, 0.03 to 0.4%. After storage in steel containers t'lere is also present a small amount c.f a soluble iron compound, presumably formed by the action cf the container
either of cyanogen chloride or of the water and hydrogen chloride present as impurities. The presence of iron is probably the most important factor in promoting the polymerization of cyanogen chloride. The effect of iron is clearly shown by comparing the stability of cyanogen chloride in glass containers with that of material in similar containers to which have been added srnall amounts of steel turnings. The data recorded in Table I show the catalytic effect o f iron.
TABLE I.
EFFECT O F I R O N O S STABILITY OF PUR.E CYANOGEN
CHLORIDE .-.~~~ Days for Complete
Solidificationa Steel turnings added 65 66, 7 0 b (A-1). 22 (24-2) 100 40 7 128 20 1.5 a All tests v e r e carried o u t in duplicate; a different batch of cyanogen chloride was used a t each temperature. b Figures indicate t h a t duplicate samples solidified a t different times. C .&-I and A-2 refer t o two different samples of steel used. 'Temp.,
c.
Glass alone 148
The tendency of cyanogen chIoride to polymerize in the presence of iron increases with increasing water content. When 0.2 and O.5y0of water, respectively, were added t o samples of cyanogen chloride, the stability of the cyanogen chloride was markedly decreased, both in the presence and absence of stecl. Itloreover, the performance of the stabilizing agent was adversely affected by the presence of excessive quantities of water. The data are summarized in Table 11. Higher concentrations of hydrogen chloride decrease the stability of cyanogen chloride somewhat, but the effect seems to be less marked than that of water and iron. The amount of hydrogen chloride in commercial cyanogen chloride F a s consistently
