I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
628
Vol. 18, No. 10
of Moore (IO). Aliquots of this extract were measured on the colorimeter and the c a r o t p e concentrations were determined by measurements at 4620 A. on a Beckman spectrophotometer, using the absorption coefficients as reported for pure carotenes in hexane by Zscheile et al. (15). The slit width was 0.03 mm. and 1-cm. cells were employed.
and for its carefully controlled storage, and eliminates the operation of weighing out the standard and the necessity for spectrophotometric study of the standard prior to each use.
These data, when plotted, give essentially the same calibration curve. This indicates that when a good preparation of carotene is used, the analytical results will be the same with either the direct or indirect method of calibration, Any discrepancy in a comparison of this type will probably,be due to deterioration of the crystalline carotene standard, so that the indirect calibration curve is considered the more reliable.
Benne, E. J., Rose, D. I., and Comar, C. L., J . Assoc. Oj’icial Agr. Chem.,27,517 (1944). Comar, C. L., dissertation, Purdue University, 1942. Comar, C. L., IND.ENG,CHEM.,ANAL.ED.,14,677 (1942). Comar, C. L., Benne, E. J., and Buteyn, E. R., Ibid., 15, 524
A D V A N T A G E S OF INDIRECT C A L I B R A T I O N
The fundamental advantage of the indirect method of calibration as outlined here is the reliability of the results obtained, which will lead to dependable interlaboratory agreements where this situation is now far from satisfactory in many cases. Another advantage lies in the simplicity of the indirect meth0.d. After the details have been R-orked out, the spectrophotometric determination on an aliquot of the extract is usually a simple matter of making measurements on a solution a t a few given wave lengths. I t can usually be arranged to have the solutions shipped to another laboratory for the spectrophotometric measurement, if necessary. This practice eliminates the need for ob-, taining the purified material either by purchase or preparation
LITERATURE CITED
(1943).
Comar, C. L., and Zscheile, F. P., Plant Physiol., 17,196 (1942). Hogness, T. R., Zscheile, F. P., Jr., and Sidwell, A. E., Jr., J . Phys. Chem., 41,379 (1937). Mellon, M. G., IND.ENG.CHEM.,ANAL.ED.,17,81 (1945). Mellon, M.G., Proc. Am. SOC.Testing Materials, 44,733 (1944). Miller, E. S., ”Quantitative Biological Spectroscopy”, Minneapolis, Burgess Printing Co., 1939. Moore, L. A , , IND. ENG.CHEM.,A x . 4 ~ED., . 12,726 (1940). Smith, J. H. C., J . Am. Chem. SOC.,58, 247 (1936). Vandenbelt, J. M., Forsyth, J., and Garrett, -4.,IND.ENG. CHEM.,ANAL.ED.,17,235 (1945). Zscheile, F. P., and Comar, C. L., Botan. Gaz., 102,463 (1941). Zscheile, F. P., Comar, C. L., and Mackinney, G., Plant PhysioZ., 17,666 (1942). Zscheile, F. P., White, J. W.,J r . , Beadle, B. W.,and Roach, , J. R., Ibid., 17,331 (1942). PRESENTED before the Florida Section of the AMERICAN CEEMICALSOCIETY, Tampa, Fla., &fay 12, 1945. Published with the permission of the director of the Florida Experiment Station.
Determination of Conjugated Diolefins with Chloromaleic A nhydride S. T. PUTNAM, M. L. MOSS,
AND
R. T. HALL, Hercules Experiment Station, Hercules Powder Company, Wilmington, Del.
A new method of determining conjugated, diolefins in hydrocarbon mixtures utilizes chloromaleic anhydride as the dienophilic reagent. This method is based on the quantitative addition of chloromaleic anhydride to conjugated dienes to form an adduct containing a highly reactive tertiary chlorine atom which can be determined by the Volhard method after refluxing with aqueous silver nitrate; the vinyl chlorine of the reagent is completely unreactive under the same conditions.
D
U R I S G the early phases of the synthetic rubber program, an investigation of the preparation of conjugated dienoid monomers by the cracking of terpenes was undertaken in this laboratory. I t soon developed that the determination of conjugated dienes in the pyrolysis products was one of the more important phases of the problem. Since no satisfactory procedure n-as available for the analysis of such mixtures, a number of physical and chemical methods were considered. It Tvas recognized that several physical methods including ultraviolet absorption, infrared absorption, and mass spectrometry give excellent results on mixtures of known constituents. Hovever, to obtain reliable results by the absorption methods it is generally necessary to know not only the functional groups but also the molecular species present. Moreover, a simple chemical method requiring no special equipment appeared desirable ior the routine determination of conjugated dienes in the crude products from cracking experiments. Consequently, a number of reagents were considered which included bromine (4, 7 ) , sulfur dioxide (10, l a ) , aromatic diazonium coumpounds ( I I ) , cuprous chloride (9), maleic anhydride (3, 12), and chloromaleic anhydride. All these reagents except the last two either proved nonspecific for conjugated dienes or did not react quantitatively. The first procedure developed involved fractionation of the samples and determination of the conjugated diene content of
the cuts by a refined gas volumetric procedure utilizing maleic anhydride as the diene absorbent. However, the need for a more rapid procedure-particularly one that would not involve preliminary fractionation of the sample-resulted in the development of a second method using the sRme reagent. I n this procedure the conjugated diene coiitcnt was ohtxined by reacting the sample with a known excess of maleic anhydride and determining the unreacted reagent polarographically. Both procedures were time-consuming and provided only an indirect determination of t,he conjugated dienes The belief that chloromaleic anhydride would also react quantitatively with conjugated dienes and, in addition, would provide a direct and more rapid determination led to the development of the present method.
l
Figure 1.
L
Sample Bottle Bulb,Cap Pressure Bottle, and
ANALYTICAL EDITION
October, 1946 Table
I.
Determination OF Chlorine in Isoprene-Chloromaleic Anhydride Adduct.
Conditions of Pretreatment Chlorine in Adduct Temp., Time, , Theoretical, Found, Compounds Added O C. hours % % 17.7 None ... 17.68 17.6 None ... ... 17.6 ... None 2 17.9 ... Chloromaleic anhydride, 1 gram 55 17.8 Chloromaleic anhydride, 1 gram 2.5 ... 50-55 17.7 2.5 50-55 Mixed Amylenes, 0.2 ml. ... 17.7 2 Chloromaleic anhydride, 1 gram 100 ... 17.7 2 100 ... 17.8 Inhibitor solution, 2 drops 1 ... 75-80 17.8 75-80 1 Sample weights varied from 0.2 to 0.5gram.
... ... ...
...
...
Q
DEVELOPMENT OF CHLOROMALEIC ANHYDRIDE METHOD
Chloromaleic anhydride reacts with conjugated diolefins in the same manner as maleic anhydride to form a Diels-Alder adduct. Thus, isoprene (I) reacts with chloromaleic anhydride (11) to give l-chloro-4-methyl-4-cyclohexene-l,2-dicarboxylic anhydride (111) and 2-chloro-4-methyl-4-cyclohexene-1,2-dicarboxylic anhydride (IV).
//
CH2
0
c1
CH-C'
+
-C
/\
CHI
CH1
I
'
629
able for adduct formation. The formation of chloromaleic acid can, however, be prevented by proper care in handling the anhydride. Even with extensive purification i t was difficult t o obtain chloromaleic anhydride that gave theoretical adduct formation. Consequently, it mas found desirable to use chloromaleic anhydride that gave results between 95 and lOO'j?? of theory and apply an empirical correction factor determined by analyzing a sample of known conjugated diene content. However, it is believed that if absolutely pure chloromaleic anhydride were available, it would be possible to eliminate the correction factor, since in one instance it has been reduced to 1.005. Low and variable results were also obtained in the presence of peroxides which catalyze the copolymerization of conjugated dienes with chloromaleic anhydride. Apparently the copolymer contains more than a 1 to 1 ratio of conjugated diene to chloromaleic anhydride, and furthermore, it is so insoluble in aqueous silver nitrate that the analysis is unsatisfactory. In view of this adverse effect of natural peroxides, a small amount of inhibitor, such as p-tert-butylcatechol, was added as a precaution in all subsequent analyses. Ai study of the optimum time and temperature for adduct formation showed that heating a t 55" C. for 2 hours was sufficient for 0 samples containing over 25% isoprene or for cyclopentadiene. Samples containing smaller amounts of ,()y'\o isoprene should be heated a t higher temperatures and/or for longer f l\ l' periods of time-for exarnpIe, EarnLl ples containing 5 t o 10% should be C' heated for 4 hours a t 75 ' C. Butadiene reacts more slowly than isoprene and it is necessary to heat the IV samples a t 55' C. for 12 hours to obtain satisfactory results. Isoprenebutadiene mixtures containing 5 to 10% butadiene require 6 to 8 hours a t 55" C. and mixtures containing 20 to '25% butadiene require 8 to 12 hours at 55' C. The method should be applicable to samples containing less than 5% diene, but the optimum reaction conditions have not been established experimentally. The accuracy in this range might be improved by increasing the sample weight.
''
/
I"
C1
XO
I1
I11
The difference in reactivity between the highly active tertiary chlorine of the adduct and the unreactive vinyl chlorine of the reagent permits a quantitative determination of the adduct in the presence of excess reagent. Refluxing with aqueous silver nitrate removes the chlorine completely from the adduct, leaving the chlorine of the chloromaleic anhydride untouched. It is believed that the utilization of the high reactivity of tertiary halogens as a basis of organic analysis is a principle that has considerable possibilities and has been exploited to only a limited extent. This same principle has been applied recently in the authors' laboratories to the determination of cupdimethylstyrene in styrene mixtures ( 5 ) . The practicality of the new method was demonstrated by substituting chloromaleic anhydride for maleic anhydride in the gasometric absorption procedure, and also by synthesizing pure isoprene-chloromaleic anhydride adduct (melting point 35-36 'C.) and showing that the chlorine of the adduct could be determined by the Volhard method after refluxing with aqueous silver nitrate. -4s the results in Table I show, the agreement with theoretical is equally good when the determination is made in the presence of compounds added t o simulate conditions of an actual analysis. Samples containing conjugated dienes were then heated with excess chloromaleic anhydride in specially designed pressure bottles (6) (Figure I ) , and the amount of adduct formed was determined as outlined above. I t was soon apparent, as is shown by the data in Table 11, that it was necessary to purify the commercial-grade chloromaleic anhydride to obtain quantitative adduct formation. Chloromaleic acid proved to be one of the most troublesome impurities, as it is readily formed by hydrolysis of the anhydride. The low and erratic results obtained in the presence of appreciable amounts of this acid, shown in Table 111, may be due to the acid catalyzing the dimerization and polymerization of conjugated dienes, thus rendering them unavail-
B
CHLOROMALEIC ANHYDRIDE REAGENT
Commercial chloromaleic anhydride is purified by filtering the crystals from the oily mother liquor and washing them with dry hexane. (Commercial chloromaleic anhydride may be pur-
Table
II. Effect of Purity OF Chloromaleic Anhydride on Determination of Conjugated Dienes
Chloromaleic Anhydride Sample purification Commercial One recrystallization Two recrystallizations Three recrystallizations Two recrystallizations, one distillation Theoretical
Table
Melting point, O C. 27 (ea.) 31.532.5 32-34 32.5-34
...
E.f i s )
Conjugated Diene, h a l y s i s Results. % ' of Theory 62.2 * 5 91.6 * 2 ,
93.1 * 1 95.4 * 1 99.0 * 1 100.0
Ill. Effect OF Chloromaleic A c i d on Determination OF Conjugated Dienes with Chloromaleic Anhydride
Chloromaleic Acid Added Weight % of Total Reagenks 0 1 2 3 10 50
Weight % of Isoprene Found (Ay.) 93.9 90.6 92.0 91.5 85.3 82.2
630 Table
INDUSTRIAL AND ENGINEERING CHEMISTRY IV. Analysis of Standard Conjugated Diene h n p l e s with
Chloromaleic Anhydride Conjugated Dienes, Weight yo Composition of Sample Present Found0 99.7 100.0 Isoprene 99.7 100.0 99.7 100.0 100.0 100.0 100.1 100.0 7 ; . ; 75.2 [soprene-an>lene mixture 75.1 75.D 60.6 G0.8 Rn R
A1 2
60.8 60.8
59.6 61.8 59.7 25.2 26.2 25.4 9.5 9.9 10.2 9.7 4.6 4.8 97.5 98.5 97.8 97.9
GO 8
z
25.7 28
Dei-iation -0.3 -0.3 -0.3 AO.0
*o. 1 -0.3 -0.4 -0
a
-0 , - 4
-1.2 +1.0 -1.1 -0.5 TO.5 -0.3 -0.4
25.I boprene-pentane mixture 9.9 9.9 fO.0 9.9 +0.3 9.9 -0.2 5.0 -0.4 5.0 -0.2 Cyclopentadiene 98.9 -1.4 98.9 -0.4 Butadieneb 98.4 -0.6 98.4 -0.5 5 Results corrected by a factor of 1.03. 5 Butadiene samples heated for 12 houra at 5.5' C.instead of 2 hocre. ~
~~
Table V. Precision of Typical Determinations on W i d e Boiling Range Isoprene Sampler b y Chloromaleic Anhydride M e t h o d Weight % ' of Conjugated Dienee Sample Designation Found, Calculated as CcHr 89.6.89.7 A 74.4,75.0 B 77.7,77.5 C 85.3,86.0 D 92.5,92.8 E 95.4,95.3 F ' 95.9,96.0 G 91.6,gl.O H 28.2,27.3 I 93.8,94.0 J
chased from the Kational Aniline Division, Allied Chemical and Dye Corporation, 40 Rector St., Kevi York, N. Y. By special request one sample of chloromaleic anhydride was obtained which was of satisfactory analytical grade as received. Should such a reagent become generally available, the purification procedure would be unnecessary.) The crystalline chloromaleic anhydride is dissolved in the minimum amount of dry benzene and then hexane is added to incipient, turbidity a t room temperature. The mixture is seeded and cooled in an ice bath or a refrigerated room (+3' C.). The recrystallized product is filtered as before and then distilled in vacuo (water pump with drying tube in the line) under dry carbon dioxide or nitrogen. The fraction boiling a t 110" 9. a t 14 mm. of mercury is collected and sealed in small glass ampoules under nitrogen or carbon dioxide until ready for use. The purified product melts a t 32 to 34" C. and is completely soluble in dry benzene (insoluble material is usually chloromaleic acid). The correction factor for the reagent is determined by analyzing a sample of isoprene of known purity. If it is great,er than 1.05, further purification is recommended. ANALYTICAL PROCEDURE
For the analysis of volatile samples veigh a small glass bulb (Figure 1) similar to those used in acid analysis to the nearest 0.1 .mg. Warm the weighed bulb slightly with a small flame and invert the stem into t)he sample, using a lead washer for support. Cool the bulb with a snugly fitting piece of dry ice until the desired amount of sample (0.1 to 0.2 gram) is drawn into I t and then quickly seal the stem in a hot flame. To prevent carbonization and fractionation during sealing keep the bulb well cooled by holding a small piece of dry ice just below the tip of the stem. Reweigh the bulb after it reaches room temperature and place it in a pressure bottle (Figure 1) with 1.0 * 0.1 gram of chloromaleic anhydride. [These small pressure bottles, for use with commercial bottle caps, are available from the Ace Glass Co., Vineland; New Jersey. Bottle caps with cellophane liners ("spots") are recommended. ] Add one drop (approximately 0.02 ml.) of a 10% solution of p-tert-butylcatechol in nitrobenzene, cap the pressure bottle using a household bottle capper, and break the sample bulb by striking the bottle sharply against a moderately hard object like t h e heel of one's shoe.
Vol. 18, No. 10
Heat for 2 hours a t 55" C., cool, open the bottle, and transfer the reaction products to a 250-mi. Erlenmeyer flask through a funnel with 10 ml. of acetone and distilled water (50 t o 75 ml.) (These conditions are for samples containing over 25% isoprene See the preceding discussion regarding the conditions to use for other samples.) Use 2 ml. of the acetone to rinse the bottle cap Add 20 ml. of 0.2 N aqueous silver nitrate solution and reflux the mixture for 1 hour. After cooling, filter, wash the precipitate thoroughly with distilled water, add 5 ml. of 1to 1nitric acid and 1 ml. of ferric alum indicator to the filtrate, and titrate the excess silver nitrate with 0.1 N potassium thiocyanate solution. Nonvolatile samples may be weighed directly into the pressure bottle from a small weight buret. From this point the procedure is the same. Calculate the results as follows (for unknown samples an approximate molecular weight may be assumed or the results ma! be expressed as diene numbers) :
% conjugated diene
AgSOa X N AgN03) - (ml. uf mol. wt. of diene X correction factor
= [(ml. of
1
DISCUSSION OF RESULTS
The conjugated diene content of a number of standard sampler of isoprene, isoprene-amylene mixtures, isoprene-pentane mixtures, cyclopentadiene, and butadiene was determined by the chloromaleic anhydride method. The results given in,Table IV indicate that the average error of the determination expressed aE: per cent conjugated diene is about 0.5 unit. Duplicate determinations on a number of crude isoprene samples are given in Table V and show the precision of the method to be of the same order of magnitude as the accuracy. Styrene interferes in the determination because of the formation of a copolymer vith chloromaleic anhydride, even in the absence of peroxide catalysts. The copolymer naturally contains tertiary chlorine which is determined along with the chlorinr of the conjugated diene-chloromaleic adduct. Although the method has been used most extensively for thr determination of isoprene, it has been applied successfully to the determination of butadiene, cyclopentadiene, and other conj ugated dienes. Preliminary experiments indicate that trans- ' piperylene and ?-methy1-1,3-pentadiene can be determined with chloromaleic anhydride, although 4methyl-l,3-pentadiene cannot. The behavior of the latter compound is not surprising. since it has been found (2) that it will not form a Diels-Alder adduct with maleic anhydride. I t is probable that the method could be applied to other conjugated dienes by varying the conditions of adduct formation to obtain a quantitative reaction. ACKNOWLEDGMENTS
The authors wish to express their gratitude to C. L. Dunn, R A. Krieger, and R. L. Shirley, whose assistance in the development of this method Ti-as invaluable. Thanks are due t o Lawrence H. Flett of the National Anilinr Division of Allied Chemical and Dye Corporation for his cooperation in making available several samples of purified chloromaleic anhydride. LITERATURE CITED
Auwers and Harres, Ber., 62B, 1678 (1929). Bachman and Goebel, J . A m . C'hem. Soc., 6 4 , 7 8 7 (1942). Basset and Killiams, J . Chem. Soc., 1932, 2324. Dobryanskii, Petroleum Domain, 1925, 574 (1926). Elliott and Cook, IND. ENG.CHEM.,ANAL.ED.,16, 20 (1944). Hershberg, Wolfe, and Fieser, J . B i d . Chem., 140, 218 (1941). Manning, King, and Sinnatt,, Fuel Research, Tech. Paper, 19, 19 (1928).
Perkin, J . Chem. Soc., 53, 703 (1888). Schultz, R. F., private commu ication. Sorokin and Puzitzkii, Sintet. Kauchuk, 6 , 12 (1933). Terent'ev and Demidova, J . Gen. Chem. (U.S.S.R.), 7 , 2464 (1937).
Tropsch and Mattox, IXD.END. CHEM., ANAL. ED., 6, 104 (1934).