Gravimetric Determination of Styrene as Styrene Nitrosite G. R. BOND, J R . , Special Projects Division, Houdry Process Corp. of Pennsylvania, .Marcus Hook, Pa. An improted method for the determination of styrene has been developed, based on the reaction of styrene with nitrogen trioxide to give a definite proportion of a crystalline nitrosite. Various olefinic impurities, which interfere in previous methods of analysis, either do not react or give products which are soluble in cold alcohol. Styrene nitrosite is of very limited solubility in cold alcohol.
N O K G the methods 11-hich have been utilized for the qunntiTable I.
tative determination of styrene are (a)the determination of purity by freezing point (a)which is applicable only to samples containing over 90% of styrene, and when there is only one major impurity present; ( b ) estimat,ion of mole per cent olefinic unsaturation from the bromine number, Fvhich is of little value when other unsaturated compounds are present; (c) determination of refractive index or refractivit,y intercept, applicable t o mixtures containing only xylenes or ethylbenzene as diluents; ( d ) selective polymerization ( 3 ) ,which may give erroneous results in the presence of conjugated diolefins or other materials which give rise to heavy polymers under the conditions of the test; and ( e ) determination by the use of the ultraviolet spectrophotometer ( I ) , which is subject to serious interference from impurities that show strong absorption in the near ultraviolet. A new method, which avoids the defects of these earlier procedures, has been developed for the quantitative determination of styrene, based on the reaction of styrene u-ith nitrogen trioxide (nitrous anhydride) to form a bulky crystalline nitrosite, prohably according to the following equation:
Conversion of Styrene to Styrene Nitrosite
Volume % Styrene (25-Ml. Sample) 1.0 2.0 3.0 5.0 7.0 10.0
Styrene WeightSitrosite, of Crystalline Grams 0,300 0.630 1.035 1.590 2.260 3,160
Apparatus and Reagents. Pyrex Tutwiler buret. The lower stopcock should be a t least of 4-mm. bore. Pyrex sintered-glass crucible of medium porosity. Saturated sodium nitrite solution prepared by dissolving 72 grams of sodium nitrite in 84 ml. of water, warming slightly if necessary. Dilute sulfuric acid prepared by diluting 1 volume of concentrated sulfuric acid with 4 volumes of water. Six milliliters of this acid are equivalent to 5 ml. of the nitrite solution. Ethyl alcohol, 95%. Petroleum ether, boiling point 35" to 60" C. Naphtha (diluent) treated with sulfuric acid to remove olefins and aromatics, boiling range approximately 120" to 150" C. Sample Size. Samples containing up to 10% styrene may be treated directly, using a 25-ml. portion. For higher concentrations, an accurately measured (75' F.) volume of saniple containing not over 10 ml. of actual styrene should be diluted with acid-treated naphtha to 100 ml. in a volumetric flask and a 25ml. aliquot taken for analysis. Likewise, because of the appreciable solubility of the styrene nitrosite in aromatics, the portion taken for analysis should, if necessary, be diluted with naphtha to reduce to beloiv 50% the percentage of any aromatics aside from styrene. Analytical Procedure. Partly evacuate the bulb of the Tutwiler buret and draw in through the upper section 10 ml. of the sodium nitrite solution. Rinse out the upper buret with water and pipet in 25 ml. of the sample a t 75' F. (diluted if necessary a8 stated above), drawing it completely into the lower bulb. Again evacuate the bulb to about 0.25 atmosphere and fill the upper buret with the dilute sulfuric acid. Permit the acid t o flow into the bulb in about 0 . 5 4 . increments, shaking vigorously between additions, thus generating nitrogen trioxide which reacts immediately with any styrene present. Continue the addition of acld as described, cooling under running water if excessive heat develops, until pressure commences to develop in the bulb, as indicated by gas being forced up through the acid upon opening the upper stopcock carefully, having first removed the stopper on the buret, (Avoid drawing air into the bulb a t any time to prevent oxidation of nitrogen trioxide to nitrogen tetroxide.) Re-evacuate the bulb and continue the careful addition of acid until 1 ml or less produces pressure, signifying the completion of the reaction. As a safety measure, a face mask should be worn a t all times in case excessive pressure should blow out the stopcock. Pour off any excess acid from the upper buret, draw off the contents of the bulb into a tared Pyrex sintered-glass crucible, and filter with mild suction. Kash the bulb four times with 5to 10-ml. portions of water to remove the bulk of any remaining unreacted hydrocarbon and precipitate onto the crucible and to dissolve inorganic salts. Now wash the bulb with three successive 5-ml. portions of petroleum ether, transferring these washings to the crucible to dissolve any remaining hydrocarbon. Again rinse the bulb with three 3-ml. portions of 95yo ethyl alcohol a t room temperature, and pass these washings also through the crucible to remove water and oily precipitate which forms in certain cases. (In case diisobutylene is present in considerable amounts, five alcohol washes may be necessary. If desired, the alcohol used in the second and subsequent washes may have
Exclusion of air during the reaction is important to prevent oxidation of nitrogen trioxide to nitrogen tetroxide which would lead to the formation of an oily nitrosate in place of the crystalline nitrosite. Although nitrous anhydride has long been used to identify various terpenes by means of the physical properties of their crystalline nitrosites (4),it is believed that this method has not been applied before to the identification or quantitative determination of styrene. DESCRIPTION OF METHOD
Scope. This method appears to be fairly specific for monomeric styrene in the presence of such other materials as would normally be encountered as impurities, including phenylacetylene or butadiene dimer. The normal amounts of polymerization inhibitors which may be present do not interfere. It is directly applicable to samples containing up to 10% styrene and to higher percentages on proper dilution. The method should preferably he applied to a narrow-boiling fraction in the styrene boiling rsnge to reduce possible contamination from such other olefinic materials as may react with nitrogen trioxide, although acceptable results have been obtained on cracked naphthas boiling from 40' to 210" C. Such other olefinic hydrocarbons as do react with riitrogen trioxide (notahly diolefins and a limited number of monoolefins) yield products which are generally oily and much more aoluble in cold aLcoliolthan is styrene nitrosite. The presence of considerable a.iu)'l i l t * of diisobutylene may lead to high results unless the prwrtiit,it)ns noted in the procedure are observed. Phenolic materiais r~iltyinterfere, but may readily be removed by caustic washing.
390
391
JUNE 1947 been previously saturated with nitrosite.) Careful agitation of the precipitate with each wash by means of a small stirring rod or wire assists the purification step. The precipitate should now be white or pale tan and crystalline. Transfer the crucible a t room temperature to a vacuum desiccator containing calcium chloride and evacuate to approximately 50 mm. of mercury. After about 3 hours a precipitate should have attained constant weight. If not, place again in the desiccator and reweigh a t 0.5-hour intervals until there is no further appreciable loss in weight. Any precipitate which still clings persistently to the inside of the Tutwiler may be dissolved in a small amount of acetone and allowed t o evaporate a t room temperature on a small tared watch glass. Its weight should be added to that of the main precipitate. Calculation Weight of precipitate x 77.5 = % styrene by volume volume of sample, ml.
Table 111. Effect of 4-Vinyl-I-cyclohexene on Styrene Analysis Styrene Present Vol. % Approximately 5 Approximately 5
4-Vinyl1-c clohexene %resent Vol. 5% 0 1.2
Styrene Found Vol. % 4.62 4.62
Table IV. Determination of Styrene in Presence of Phenylacetylene Sample
Styrene Found Styrene Theoretical Vol. % Vol. % Phenylacetylene (impure) 9.7 Cnknown, b u t estimated a t a b o u t 9.6 107, from other d a t a 9.6 99.9 Approximately 100 Styrene Phenylacetylene styrene 19.9 Approximately 19 2
+
The term 77.5 is derived from the equation 1.21 0.903 where
104 = molecular weight of styrene 180 = molecular weight of styrene nitrosite 0.903 = density of styrene a t 75" F.
(E) based on the percentage of 82.3 the crystalline form of nitrosite obtained, as described under development of the method.
and 1.21 is a correction factor
Purification Step. In the case of impure samples, such as cracked distillates containing highly reactive olefinic materials or phenols which are found to give an oily precipitate, a purification step prior to the nitrogen trioxide treatment will be found of great advantage. This step also removes already-polymerized styrene and does not affect the styrene monomer content. Pipet a sample of suitable size (measured a t 75" F.), depending on the approximate styrene content, into a 25- or 50-ml. glassstoppered graduate. Saturate it with sulfur dioxide, introducing the gas a t a fairly rapid rate through a tube drawn out to a slender capillary and cooling the sample in ice water if appreciable heat develops. Allow the sample to stand for approximately 2 hours; the impurities generally settle out as a black tar. Transfer the mixture completely to a 100-ml. separatory funnel with successive portions of cold water (any of the black tar adhering to the graduate may be left in the graduate, provided it is washed free of the supernatant sample), avoiding loss, and vash three times Tvith about 10 to 15 ml. of cold water. Then wash with two 10-ml. portions of 10% sodium hydroxide solution, once with water, and transfer completely t o a 100-ml. distillation flask. (The funnel may be rinsed with 2 to 5 ml. of petroleum ether if Jesired.) Rapidly steam-distill into a clean, chilled separatory funnel until only traces of oil come over, leaving polymerized or tarry impurities in the flask. Draw off nater from the distillate and transfer the remaining hydrocarbon completely to the main bulb of an evacuated Tutwiler, rinsing the funnel with successive -mall ortions of cold water. Draw off the rinse water from the Tutwirer as fully as possible, and precipitate the nitrosite according to the analytical procedure.
Table 11. Reaction of h-itrogen Trioxide with Various Unsaturated Hydrocarbons and Phenol Compound Ethylbenzene Toluene Xylene Phenol a-Methvlstvrene p-Methyl-& methylstyrene Dipentene Isoprene 2-Methvlpentidiene Di c y c 1o pentadiene Terpinolene
-
Phellandrene Diisobutylene Cyclohexene
Reaction Product None h-one None D a r k red oil D a r k red oil Yellow oil Little gummy oil D a r k oil D a r k oil G u m m y yellow p p t . Little brownish oil White crystalline ppt. a n d yellow oil Some crystallization on standing Slight ppt.
Remarks
Soluble in alcohol Soluble in alcohol Soluble in alcohol Soluble in alcohol
Moderately soluble in alcohol Not likely t o be encountered N o t likely to be encountered Soluble i n alcohol
DEVELOPMENT OF METHOD ' -4 series of known blends of styrene (Eastman Kodak Co., inhibited Tyith hydroquinone) in acid-treated naphtha was prepared and the nitrosite precipitated as described under procedure. The resulting weights of nitrosite and the corresponding percentages by volume of styrene in the blends are shown in Table I. When these values are plotted, they are found to lie almost on a straight line passing through the origin. On the basis of this curve, 100 ml. of 10% (volume) styrene solution, corresponding to 10 ml. or 9.03 grams of styrene, will give 12.88 grams of crystalline nitrosite. Since a theoretical weight of 15.65 grams should be obtained based on molecular weights of 104 for styrene and 180 for the nitrosite, the washings were investigated and it was found that a definite constant proportion (17.7%) of oily, alcoholsoluble nitrosite was formed a t the same time. From these data, a factor of 77.5/volume sample was derived for the calculation of the volume percentage of styrene in unknown samples from the weight of crystalline nitrosite. To determine possible interference from various impurities which might be present, the action of nitrogen trioxide on a number of aromatics, diolefins, and other materials was investigated (Table 11). Subsequent analyses of styrene mixtures containing many of these impurities have shown no appreciable interference. In most cases, any reaction products were of an oily nature and readily soluble in cold alcohol. No reaction was observed with aromatic hydrocarbons. Two of the impurities normally most troublesome, which are likely to be encountered in styrene mixtures, are phenylacetylene and 4-vinyl-1-cyclohexene (butadiene dimer), the latter occurring in the recovered styrene obtained from a synthetic rubber copolymerization process. Both these materials were found to give a slight gummy precipitate with nitrogen trioxide, which, however, was sufficiently soluble in cold alcohol to be removed by the washing operation. A 5% (volume) blend of slightly impure styrene in acid-treated naphtha n-as prepared and two 25-ml. aliquots were analyzed, to one of which was first added 1.2y0 (volume) of butadiene dimer. Results are shown in Table 111. Thus the presence of butadiene dimer equal to 247, of the styrene present had no adverse effect on the analysis, and purity of the styrene was established as 96.4%. At a cooperating laboratory (Texas Co., Beacon, K. Y.), a sample of phenylacetylene, which appeared to be contaminated with about 10% stryene and a blend of this phenylacetylene with additional styrene, were analyzed by this procedure (Table IV). Styrene may thus be determined satisfactorily in the presence of phenylacetylene. To determine any possible interference from general olefinic impurities of the type which might be encountered in cracked naphthas, a series of blends of styrene m s prepared in gasoline having a bromine number of 61 and in styrene-free reformer naphtha. Analyses by a routine operator are shown in Table V.
V O L U M E 19, NO. 6
392 Table V. Determination of Styrene in Olefinic Gasoline5 Base Stock
Styrene Added Vol. % 0 1.0 2.0 3.0 5.0 10.0 10.0
Cracked gasoline Cracked gasoline Cracked gasoline Cracked gasoline Cracked gasoline Cracked gasoline Reformer naphtha.
Styrene Found Vol. % 0 0.95 2.0 3.1 5.0, 4 . 8 10.2, 9 . 9 10.1, 9 . 7
Table VI. Effect of Contact Time w-ith Sulfur Dioxide on Styrene 2 5 24 S a O H wash only 1 4 53 4.48 4.42 4.55 4.29
Contact time, hours Styrene content, vol. %
There was some tendency for the nitrosite precipitates to be gummy, evidently due to reaction of certain olehic components of the base stock with the nitrogen trioxide. However, the alcohol wash satisfactorily removed these other compounds, as evidenced in Table V, leaving the styrene nitrosite as pale tan crystals. As a result of this tendency of certain impure samples to yield gummy or oily precipitates, it seemed advisable to develop some means of removing the bulk of the troublesome impurities without affecting the styrene content. In the course of such an investigation it was observed that certain olefinic gasoline samples, on saturation with sulfur dioxide gas a t room temperature, quickly darkened and deposited a greenish-black tar. Since these samples had also yielded an oily precipitate with nitrogen trioxide prior t o the sulfur dioxide treatment, it was thought this might provide a means for purifying cracked naphthas prior to a styrene determination. Steam distillation, following removal of excess sulfur dioxide, was chosen as the most expedient means of recovering completely the unreacted portions of the sample with a minimum chance of causing polymerization of any styrene as a result of prolonged or excessive heating. Samples treated in this manner and then analyzed showed a very marked improvement in the crystalline nature of the nitrosite. A blend of styrene in naphtha was prepared and saturated with sulfur dioxide, and samples were withdrawn a t the end of definite intervals of time for neutralization, distillation, and subsequent analysis as shown in Table VI. Since up to 5 hours’ contact time showed no appreciable effect on the styrene, a 2-hour period was adopted as satisfactory on the basis of convenience of operation and apparent completion of reaction with tar-forming constituents in this period. A series of blends of styrene plus various olefinic impurities in acid-treated naphtha was prepared and treated with sulfur dioxide for 2 hours as described in the purification step under procedure, then analyzed for styrene with results tabulated in Table VII. The first determination of the blend with diisobutylene was appreciably high. It was felt that this was due to coprecipitation of a little of the bis-nitrosite of diisobutylene which might be removed by two additional 3-ml. alcohol washes. The second determination showed this to be the case. Using this figure, the average of the determinations on the various blends was 7.29% with a maximum deviation from this average of 0.0570, which is 0.7% of‘the styrene content.
Table VII. Effect of Impurities and Sulfur Dioxide Purification on Styrene Determination Impurity Present
voz. %
None 2-Methylpentadiene Phenylacetylene Cyclobexene Diisobutylene Diisobutylene
.. 4 2
4 4 4
Styrene Present Vol. % About 7 . 3 About 7 . 3 About 7 . 3 About 7 . 3 About 7 . 3 About 7 . 3
Styrene Found Vol. % 7.24 7.28 7.33 7.32 7 . 5 6 ( 3 alcohol washes) 7 . 2 6 ( 5 alcohol washes)
As none of these added impurities showed any tar formation with the sulfur dioxide, the exact nature of the impurities which give rise to such tars is still unknown. Table VI1 indicates that sulfur dioxide has no adverse effect on the styrene determination; therefore the purification step is warranted where it results in an improvement in the crystalline nature of the styrene nitrosite. DISCUSSION
The crystalline styrene nitrosite is a fragrant, fluffy white powder, having a greenish or yellowish tint. It is of very limited solubility in 95% ethyl alcohol a t room temperature (9 ml. of alcohol dissolve 3 mg. of styrene nitrosite), thus facilitating the removal of the reaction products of any impurities which may be present, since such other products in general have been found to be soluble in cold alcohol. It is soluble in benzene a t room temperature to the extent of 1 to 2% and a determination of molecular weight by depression of the freezing point of benzene indicates it to be monomolecular in this solvent, corresponding to the formula CsHs.CHT\’O CH,N02. Analysis by the Dumas method of a sample of nitrosite prepared from pure styrene gave 15.5y0 nitrogen compared to the theoretical value of 15.6%. The nitrosite melts with vigorous decomposition a t about 93 O t o 94’ C. and is appreciably volatile a t lower temperatures; so heating should be avoided during the drying operation. Upon distillation of the nitrosite with dilute alkali, benzaldehyde appears in the distillate. The nature of the other products formed has not yet been fully investigated. The benzaldehyde thus produced may be oxidized with alkaline potassium permanganate to benzoic acid, which may be extracted with chloroform after acidification of the mixture and has been found to equal in weight approximately one third of the weight of the original nitrosite. It thus appears that 2 moles of the styrene nitrosite yield 1 mole of benzaldehyde under these conditions. This fact has been utilized in estimating the styrene nitrosite content of an occasional oily or extremely gummy precipitate, since the impurities present do not yield benzaldehyde by this treatment. This method offers certain distinct advantages over previous methods of analysis. It is not affected by the presence of most unsaturated hydrocarbons, including phenylacetylene, vinyl cyclohexene, conjugated diolefins, or other materials which give rise to heavy polymers on heating; it does not require a different correction factor for the lower percentages, as is required in a thermal polymerization method (3); it may be applied to such complex mixtures as highly cracked naphthas; and it is simple and rapid (about 0.5 hour aside from the drying operation, provided the purification step is not needed) and requires a minimum of equipment. I n the case of very impure samples, the sample may first be submitted to a purification step involving treatment with sulfur dioxide and steam distillation, thus frequently effecting separation of the impurities as a tarry residue without affecting the styrene content, and giving a crystalline nitrosite much easier to handle in the subsequent analytical step. Judging from the data presented in Table I, V, VI, and VI1 it is believed that monomeric styrene c+n be determined by this method well within an accuracy of +=370of the styrene content. ACKNOWLEDGMENT
The author wishes to thank Harry Levin of the Texas Company for his contribution on the determination of styrene in the presence of phenylacetylene. LITERATURE CITED
(1) M e e h a n , E. J., J . Polymer Sci., 1, 175-82 (1946). (2) S m o k e r a n d Burchfield, IND. EXG.CHEM.,ANAL.ED.,15, 128
(1943). (3) U n i t e d Gas I m p r o v e m e n t
Laboratories, P h i l a d e l p h i a , P a . , communication. (4) Wallach, O.,“ T e r p e n e u n d C a m p h e r , ” 2 n d ed., p p . 69 e t seq., Leipsig, T’eit & Co., 1914.