ZINC CHLORIDE AS CATALYST In the present i n v e s t i g a t i o n a study was made of the polymerization of propylene, using anhydrous zinc chloride as catalyst. Experiments were conducted in a closed-pressureapparatus, and various conditions of t e m p e r a t u r e , pressure, and time of contact were employed with the object of determining which set of conditions p r o d u c e d a m a x i m u m yield of liquid product of gasoline type. For each set of experimental conditions investigated a quantity of liquid product sufficient to enable close fractionation was prepared. Examination of the resultant fractions gave information concerning the nature of their constitutents.
FIQURE 1. PRESSURE BOMBAND RECOVERY SYSTFJM
PROPYLENE POLYMERIZATION
THE
Materials
recent activity in the field of polymerization of gaseous olefins is readily understood when we consider that there is available per year approximately 300 billion cubic feet * of olefin-containing gas from the petroleum cracking industry. Thus, for 1930 Dunstan, Hague, and Wheeler (8) estimated the volume of such gas produced as 275 billion cubic feet, from which the quantities of gaseous olefins obtainable, expressed in million of cubic feet, were about as follows: ethylene, 16,500; propylene, 22,000; butylenes, 11,000. A recent publication by Ipatieff and Egloff (9) describes the polymerization of such material to yield a highly antiknock gasoline. In 1933 Gayer ( 7 ) reported results on the polymerization of propylene a t atmospheric pressure and temperatures of 340 " to 350' C. Activated Floridin, synthetic aluminum silicate, and alumina-on-silica were found to be active catalysts; the last mentioned was most effective. The portion of the polymeric product boiling up to 150' C. consisted of a series of homologous olefins from Cg to C,, inclusive (the Cs fraction was the largest), and a small proportion of paraffin hydrocarbons. A comprehensive review of the decomposition and polymerization of the olefinic hydrocarbons, including propylene, by thermal, chemical, electrical, and alpha-particle action was given by Egloff, Schaad, and Lowry (4).
The propylene used in this experimental work was obtained from the Carbide and Carbon Chemicals Corporation. The gas analyzed 97 per cent propylene, the remainder being chiefly propane. Finely ground Baker's anhydrous c. P. zinc chloride was employed as catalyst for the experiments at 290" to 310°C.; for the experiments at 200" to 21OOC. and a t 150' to 160" C. the zinc chloride catalyst was distributed on the surface of pea-size pumice by heating these substances together, atmospheric moisture being excluded, until incipient fusion of the salt caused its adherence to the pumice surface.
Apparatus and Experimental Procedure The apparatus used in the polymerization experiments is shown in Figure 1: The steel pressure bomb, of approximately 1400 cc. capacity, was mounted in a horizontal, electrically heated furnace, and could be rotated at approximately 25 r. p. m. In each experiment the catalyst was distributed along the length of the pressure bomb to secure better contact of catalyst and reactant. The weight of propylene used was found from the differehce in weight of the transfer vessel when filled with liquid propylene from the supply cylinder and its weight after transfer of the ropylene to the pressure bomb. After introduction of the cataryst and the propylene, the pressure bomb was heated t o and maintained at
TABLEI. RESULTB OF TYPICAL POLYMERIZATION EXPERIMENTS Expt. No.
Weight of ZnCh Grams
Molar Ratio, ZnClz/ CaHs
Temp.
Time,at Working Temp.
c.
Min.
a
Yield of (Pressure Drop Lb./sq. in. gage
K g . / s q . cm.
Max. Pressure
P % $
SP. %* (:!) of Liquid Product
Lb./sq. in. gage K g . / s q . cm. Weight %
The same sample of catalyst was used in experimenta 1, 2, and 3. Distributed on 72 grams of pumioe. In these control experiments 72 grams of pumice were placed in the presaure bomb to correspond to the pumice carrier used in the experiments made under corresponding conditions. a b
0
554
MAY, 1936
INDUSTRIAL AND ENGINEERING CHEMISTRY
555
to those for experiment 9. At the end of this series a weight the desired temperature for a given time. Regular and frequent of liquid product equal to 6.5 times that of the catalyst was observations of time, temperature, and pressure were made. At the conclusion of each ex eriment the pressure bomb was obtained, the activity of the latter had dropped to 50 removed from the furnace anzcooled t o room temperature in a value. Per cent of its current of air. Liquid product was then collected in the liquid trap of the recovery system, any unreacted propylene being slowly Tables I1 to V record the data secured by fractionating vented through the spiral and e x a m i n i n g the fraccondenser and receiver cooled tions from the combined to -35" to -40" C. At this 0. L. BRANDES, W. A. GRUSE, AND ALEXANDER LOWY liquid Products of the 290temperature any highly vola310" C., 200-210" C., and tile product that might otherUniversity of Pittsburgh and wise have escaped from the the 150-160" C. e x p e r i Mellon Institute of Industrial Research, Pittsburgh, Pa. merits. The correspondliquid trap d u r i n g release of- the unreacted -propylene ing distillation curves are was recovered and added shown in Figures 2,3, and 4. to the main body of product collected in the liquid trap. The latter was then allowed to Table V shows that the C6 and C9fractions from the original stand for 8 to 10 hours at room temperature while connected t o fractionation of the 150-160 " C. product were refractionated a condenser and receiver at 0" to -10" C. in order to permit into closer boiling fractions. These latter were then examined. evolution of dissolved gases without loss of volatile product. While the composition of the gaseous hydrocarbons acAfter weighing, the liquid product was stored in a dark, cold room until the time of examination. companying the liquid polymers was not determined, it is of interest that liquefied samples of the gases from the 150Examination of Liquid Products 160" C. and the 200-210" C. experiments distilled off a t a temperature corresponding to that for liquid propylene The combined liquid products collected for each set of experi(-47" C.). This would tend to indicate that in these cases mental conditions investigated were carefully distilled through a fractionating column designed after that described by Guthrie the gaseous products consisted essentially of unreacted propyland Higgins (8), and equipped with a Kester and Andrews' ene. type of still head (1.4). Characteristics of the column were as The control experiments, in which no zinc chloride catalyst follows: length, 3 feet (91.4 cm.); internal diameter, 0.75 was present, showed practically no conversion of propylene to inches (1.9 om.) ; packing, approximately 0.5-inch (1.27-cm.) lengths of 4-mm. bore glass tubing. Adiabatic control was efliquid products, particularly at 200" to 210" C . and a t 150' fected by means of a dead air space and electrically heated jacket to 160" C. The 4.3 per cent by weight conversion in the conencircling the column. A reflux ratio of approximately 20 t o 1 trol experiment a t 290" to 310" C. (experiment 4, Table I) was maintained, and the distillation rate was such as to give may have resulted from the mild catalytic action of the steel 1 cc. of distillate per minute. At the time the distillations were made, the barometric pressure was 744 (*2) mm. Except where walls of the pressure apparatus, since Dunstan, Hague, and otherwise noted, fractions were collected for each 10" F. (5.6' C.) Wheeler (3) obtained an 80 per cent by weight conversion of rise in temperature, as read from an A. S. T. M. thermometer. propylene into liquid products by using a mild steel autoclave, During collection of the lower boiling fractions, suitable prea temperature of 400" to 404" C., a pressure of 71 kg. per cautions were taken t o guard against loss through volatilization. Each fraction was stored in the dark under an atmosphere of sq. cm., and a time period of 2.5 hours. carbon dioxide until ready for use. As an additional precaution Further, the data in Tables I1 to IV and the distillation the lower boiling fractions were kept in a refrigerator at 40" F. curves shown in Figures 2 to 4, inclusive, demonstrate that as (4.4" C.). The results of the distillations are expressed graphithe experimental temperature was lowered (from 290-310' C. cally by plotting the volume (in c ~ . of ) distillate per given fraction os. the average boiling temperature of the fraction. to 200-210" C., and then to 150-160" C.), a larger perSpecific gravities (io), refractive indices (TI%),and iodine numcentage by volume of the product distilled in the gasoline bers were determined for the various fractions. The iodine numrange (to 200" C.); likewise, a larger percentage by volume of bers were found by the Johansen method ( I $ , in which a correcthe product corresponded to the tripolymer of propylene. tion for iodine substitution is applied. Molecular weights were estimated by the method of Francis (6). From the molecular weights and iodine-addition values the percentage unsaturation of the fractions was calculated. ~~
Results of Polymerization The findings of typical polymerization experiments are given in Table I. The liquid products possessed the following characteristics: light straw color, slight fluorescence, neutral reaction to litmus, negative Beilstein halogen test response, and sweet pleasant odor characteristic of lower members of the olefin series. I n a separate experiment (not shown in Table I) a maximum yield of 81.5 per cent by weight of liquid product was obtained at 260" to 270" C., 3225 pounds per square inch pressure (226.7 kg. per sq. cm.), and a time period of 75 minutes. I n the 150-160" C. experiments it was necessary, because of the slow reaction rate, to employ a higher pressure and a longer time period to secure satisfactory yields of liquid products. The sample of catalyst described for experiment 7, Table I, was used for a series of eight experiments under conditions closely approximating those for experiment 7. An average yield of 35.6 per cent by weight of liquid product was obtained. At the end of the series the catalyst had produced an amount of liquid product equal to 7.7 times its own weight, and still retained 87 per cent of its initial activity. Similarly, the sample of catalyst described in experiment 9 was employed in a series of five experiments under conditions corresponding
This paper presents the results obtained in a study of the polymerization of propylene at elevated temperatures and pressures in the presence of zinc chloride as catalyst. Experiments were conducted i n a closed pressure apparatus under three different conditions of temperature, pressure, and time. Liquid polymer yields as high as 74.2, 43.5, and 64.4 per cent by weight were obtained under the three respective conditions. A total of 74.5 per cent by volume of the 290-310" C. product, 85.9 per cent of the 200-210" C. product, and 92 per cent of the 150-160° C. product distilled in the motor gasoline range (to 200"C.). The data obtained on fractionation of the polymeric products and examination of the resultant fractions are given, together with a discussion of their probable chemical composition.
INDUSTRIAL AND ENGINEERING CHEMISTRY
556
TABLE11. FRACTIONATION AND EXAMINATION OF LIQUID PRODUCT FROM 290-310 C. EXPERIMENTS O
Group
(Volume fractionated = 520 cc.; specific gravity ( Z O ) = 0.7557) Per Iodine Cent ReAddi- Iodine B. P. fractive tion Substiby of Average Sp. Gr Index Vol. Value, tution Fraction B. P. Volume Distd N Value (3 O
c.
30.5-40.5 40.5-50.0
Ca
50- 60 60- 70 70- 80
55 65 75
80- 90 90-100
85 95
100-110 110-120 120-128
105 115 124
128-130 130-139 139-140 140-148 148-150
129 134.5 139.5 144.0 149.0
C7
CS
co
35.5 45.2
ClO t o CII, ,in-
150-160 160-170 170-180 180-190 190-200 200-210 210-220
155 165 175 185 195 205 215
CIS
220-230 230-234
225 232
cluslve
c c.
c.
Cb
19 11
0.6431 0.6360
1.3765 1.3721
145 132
0.0 0.0
72 77
41 40
0.6531 0.6570 0.6706
1.3752 1.3865 1.3938
128 141 161
0.0 0.0
82 87 92.
41 48 58
0.6826 0.6845
1.3980 1.4060
123 125
1.2 5.1
96 101
47 50
0.7012 0.7075 0.7192
1.4061 1.4120 1,4162
122 134 115
8.1 7.8 11.4
107 112 116
51 59 52
0.7235 0.7281 0.7311 0.7367 0.7410
1.4190 1.4218 1.4240 1,4252 1.4290
0.7476 0.7601 0.7665 0.7714 0.7780 0,7923 0.8052
1,4311 1.4370 1.4404 1,4440 1.4481 1.4550 1.4613
0,8109 0.8190
1.4653 1.4681
__ 30 13.8 20.4 18.4
5.8
52.6 11.0 19.5
10.1
30.5 20.5 15.2 31.5
5.9
67.2 8.0 50.1 6.0 29.0 4.2
12.9
___ --
--_
--_
97.3 15.8 22.0 20.2 23.0 29.0 16.0 16.0
-142 0 13.0 11.5
__ 18.7
_27.3
_-
24.5 4.7 Residue = 61 cc. = 11.7 per cent; per cent by volume distilling to 200°
represented by the Ca compounds. Moreover, a t 290' to 310" C. the liquid product contained an appreciable amount of CS,C,, and C, hydrocarbons, as compared to the relatively small amounts of these compounds in the liquid products from the lower temperature experiments. Gayer ( 7 ) referred to the presence of these nonmultiple polymers in the liquid product resulting from the polymerization of propylene. It is probable that, by lowering the experimental temperature still more (below 150" C.), the polymerization reaction could be further controlled so as to yield a still higher percentage of the lower polymers, such as the dimer and trimer compounds; but slower reaction rates a t lower temperatures might prove objectionable. With reference to polymerization by means of chemical reagents, such as zinc chloride and anhydrous ferric and 11
122
167
TEMPERATURE "F 212. Z R 301.
341
Mol. Per Wt., Cent, A4 linsatn.
392.
0.25
VOL. 28, NO. 5
aluminum chlorides, Brooks (1) says that all such cases probably involve addition of the reagent to the double bond, followed b y dissociation or decomposition of the addition product to give hydrocarbon residues having a t the moment bivalent or trivalent carbon atoms. Combination of these r e s i d u e s results in polymer formation,
Probable Composition of Main Fractions of Liquid Products
The fractions of the liquid products of most interest were naturally those present in larg111 12.2 118 52 18.5 131 121 62 est amounts; these fractions 124 132 13.0 64 115 10.3 126 57 comprised the main peaks of the 10.3 124 128 62 distillation curves shown in Figures 2,3, and 4. By com112 13.0 132 58 parison of the properties of 33.8 136 90 48 34.6 141 86 48 these fractions with those for 34.4 85 146 49 41.4 70 known hydrocarbons1 of simi151 41 40.0 39 156 24 lar boiling points, the following 37.0 38 162 24 conclusions as to the approximate c o m p o s i t i o n of t h e 44.0 30 167 20 51.0 19 170 13 fractions were drawn: .polymer PRODUCT AT 290 O TO 310 O C. C. = 74.5 The data for the C6-C8,inclusive, fractions of this product (Table 11) indicated them to consist predominantly of olefinic and paraffinic hydrocarbons. The high specific gravities and refractive indices of the naphthenic, cycloolefinic, and aromatic hydrocarbons eliminate the possibility of their presence in the fractions, except in very small amounts. Moreover, treatment of the Cgfraction of 134.5' C. average boiling point with fuming nitric acid, as in the Garner method (6),and consequent renitration of the fuming nitric extract, failed to yield any crystallihe nitroaromatic derivatives of the xylenes (meta and para), which 1 A long list of hydrocarbons with their boiling points, specific gravities, refractive indices, and literature references is contained in the thesis of 0 L. Brandes from which this paper is taken. TEMPERATURE
'F
431
8 w
5
9
TEMPER~TVRE 'C.
FIQURE2. DISTILLATION CURVEOF POLYMER MADEAT 290-310" C.
5
TEMPERATURE
'c
175
UXI
225
FIGURE 3. DISTILLATION CURVEOF POLYMER MADEAT 200-210" c.
MAY, 1936
INDUSTRIAL AND ENGINEERING CHEMISTRY
557
have boiling points w i t h i n TABLE111. FRACTIONATION AND EXAMINATION OF THE LIQUIDPRODUCT t h e r a n g e of this fraction. FROM 200-210" C. EXPERIMENTS The specific gravity of known (Volume fractionated = 860 0 0 . ; speoifio gravity (.5 0 ) = 0.7523) olefins and paraffins and the Per relatively high iodine subCent Iodine by AddiIodine Per stitution values of the C9 AverVol. sP. t i ~ e tion SubstiMol. Cent age velDisGr. Index Value, tution Wt., Unfractions indicate that an I3. P of Fraction M satn. Group B. P. ume tilled (z') (a%?) N Value a p p r e c i a b l e c o n t e n t of F. c. c. cc. b r a n c h e d -e h a i n hydrocar100-134 44.4-56.7 50.5 0.6 0.07 c b r c 6 -__ bons, such as the m e t h y l 66.7-62.2 59.4 84 71 134-144 5.0 0.6868 1.3930 214.5 CS 0.0 octanes, methyl octenes, and 86 75 144-154 62.2-67.8 65.0 28.1 0.6846 1.3948 221.0 0.0 89 75 154-164 67.8-73.3 7 0 . 5 0.7036 1.4009 1 8 . 5 213.0 0 . 0 the dimethyl heptanes, was 73.3-78.9 76.1 8.2 92 76 164- 174 0.7045 1.4030 209.0 0.0 ____ present. 59R 7 0 In the case of the Clo-12 174- 184 78.9-84 4 81.6 4.0 Ci 184- 194 87.2 4.0 84.4-90.0 f r a c t i o n s from each of the 194-204 90.0-95.6 92.8 2.0 liquid products (Tables 11, 204-214 95.6-101.1 98.3 4.0 ____ 111, and V), the percentage 14.0 1.6 unsaturation is observed to be 214-224 101.1-106.7 103.9 2.0 cs 224234 106.7-112.2 109.4 3 . 6 lower than that of the lighter 234-244 112.2-117.8 115.0 6.0 244-254 117.8-123.3 120.5 6.0 fractions, particularly in the _____ 200-210' C. and the 150-160' 17.6 2.0 254-264 cs 123.3-128.9 1 2 6 . 1 26.1 0.7383 1,4209 141.0 3.9 117 65 C. products, while the iodine 264-274 128.9-134.4 131.6 144.0 0.7418 1.4235 136.0 5.2 123 120 67 64 substitution values are corre274-284 134.4-140.0 137.2 188.0 0.7436 1.4258 139.0 3.5 284-294 140.0-145.6 125 68 142.8 31.6 0.7514 1.4280 138.0 3.8 s p o n d i n g l y higher. These 294-304 145.6-151.1 148.3 14.0 ____ f a c t s w o u l d indicate that 403.7 47.0 larger amounts of compounds, 304-314 151.1-156.7 153.9 12.0 ClO t o 10.0 such as branched-chain oleC~z,,in- 314-324 156.7-162.2 159.4 162.2-167.8 165.0 9.0 clusive 324-334 fins and paraffins, which con334-344 167.8-173.3 170.5 17.0 344-354 173.3-178.9 176.1 11.0 tain secondary and tertiary 354-364 178.9-184.4 181.6 145 50 36.0 87.0 24.0 0.7696 1.4370 364-374 184.4-190.0 187.2 84.0 148 49 24.5 84.6 hydrogen atoms capable of 0.7766 1.4392 374-384 190.0-195.5 192.7 48.0 0.7764 1.4405 85.5 150 50 25.0 substitution by iodine, were 384-394 195.5-201.1 198.3 16.0 394-404 201.1-206.6 203.8 10.0 present in the heavier frac404-414 206.6-212.2 209.4 7.0 414-424 212.2-217.8 215.0 7.5 tions. Here also the high 424-434 217.8-223.3 220.5 8.2 specific gravities and refrac-275.7 32.0 tive indices of the aromatic 434-444 223.3-228.9 226.1 13.0 Cia,Cia hydrocarbons boiling in this 444-454 228.9-234.4 231.6 17.0 454-464 234.4-240.0 237.2 14.6 range led to the belief that they were not p r e s e n t i n 44.6 5.2 Residue = 40 cc. = 4.6 per cent; per cent hy volume distilling t o 200' C . = 85.9 significant amounts. Support for the p r e s e n c e of naphthenic hydrocarbons was going from 134' to 146' F. (56.7' to 63.3' C.). The same found, however, from the fact that the observed specific behavior is noted for the 200-210" C. product in going from gravities.of the fractions were higher than could be explained 134' to 154" F. (56.7' to 67.8' C.). This minimum was not on the basis of known olefins and p a r a f i s present to the observed in the C6fractions from the 290-310" C. product. As extent indicated by the degree of unsaturation. Thus, if we the specific gravities of the 56.7-62.2' C. fraction from the select the fraction of 180-190' C., boiling point from the 200-210' C. product, and the 56.7-59.4' C, and 59.4-62.2" C. 290-310' C. product (Table 11),and choose from the literafractions from the 150-160" C. product were higher than could ture representative hydrocarbons of similar boiling point, be explained on the basis of known olefins of similar boiling the calculated specific gravity of the fraction becomes: point, the presence of cyclic hydrocarbons in the saturated 1. Assuming only olefins and paraffins were present, portion of these fractions was inferred. For purposes of = 0.3739 0.49 (% unsatn.) X 0.7630 (1-undecene) 0.51 X 0.7655 (2,6-dimethyl-3-isopropylheptane)= 0.3904 illustration, let it be assumed that the saturated portion of the 59.4-62.2' C. fraction from the 150-160' C. product (Table 0.7643 V) was composed entirely of naphthene hydrocarbons. The Observd. sp. gr. = 0.7714 calculated gravity of this fraction would then be: 2. Assuming olefins, paraffins, and naphthenes were present 0.71 (% unsatn. X 0.6817 (2-methyl-1-pentene) = 0.4840 = 0.3739 0.49 X 0.7630 (1-undecene) = 0.2014 0.29 X 0.6946 ~1,2,3-trimethglcyclopropane) 0.30 X 0.7655 (2,6-dimethyl-3-isopropylheptane) = 0.2296 0 . 2 1 X 0.801 (pentamethylcyclohexane) = 0.1682 0.6854 Obsvd. sp. gr. = 0.6867 0.7717
fz;-
O
Here the assumption of 21 per cent of a naphthene in the saturated portion of the fraction gives a calculated specific gravity close to that observed. PRODUCTS AT 200-210" C. AND 150-160' C. The unsaturation of the Ce fractions from these products was from 19 to 36 per cent higher than for the corresponding fractions from the 290-310' C. product. Also, in the case of the 150-160' C. product, the specific gravity passes through a minimum in
Probably, then, the saturated portion of the fractions just cited contained appreciable amounts of naphthenes. The other CRfractions had properties indicating a high content of olefinic and paraf€inic hydrocarbons. An explanation for the absence of naphthenic hydrocarbons in the corresponding C g fractions of the 290-310" C. product is suggested by the research of Ipatieff and Rutala (11) and of Ipatieff and Huhn (IO). In the polymerization of ethylene by zinc chloride the
INDUSTRIAL AND ENGINEERING CHEMISTRY
558
VOL. 28, NO. 5
TABLEIV. FRACTIONATION OF LIQUID PRODUCT FROM 150-160 C. EXPERIMENTS (Volume fractionated = 1240 cc.: specific gravity
Group
cs
c8
c7
44.4-56.7
134-144 144-154 154-164 164-174
56.7-62.2 62.2-67.8 67.8-73.3 73.3-78.9
174-184 184-194 194-204 204-214
-
59.4 65.0 70.5 76.1
78.9-84.4 84 .4-90.O 90.&95.6 95.6-101.1
81,6 87.2 92 8 98.3
132.0 5.0 3.0 5.0 2.0
cs
214-224 224-234 234-244 244-254
101.1-106.7 106.7-112.2 112.2-117.8 117.8-123.3
103.9 109.4 115.0 120.5
e 9
254-264 264-274 274-284 284-294 294-304
123.3-128.9 128.9-134.4 134.4-140.0 140.0-145.6 145.6-151.1
126.1 131.6 137.2 142.8 148.3
21.0 28.0 378.0 240,O 31 .O 16.0
153.9 159.4 165.0 170.5 176.1 181.6 187.2 192.7 198.3 203.8 209.4 215.0 220.5
693,O 10.0 8.0 6.0 10.0 17.0 56.0 116.0 40.0 13.0 5.0 5.0 6.0 7.0
226.1 231.6 237.2 242.7
299.0 11.0 12.3 10.0 15.0
304-314 314-324 324-334 334-344 344-354 354-304 364-374 374-384 384-394 394-404 404-414 414-424 424-434
Cis,
434-444 444-454 454-404 464-474
clusive
cI4
223.3-228.9 228.9-234.4 234.4-240.0 240.0-245.5
0.3
24.0 92.0 13.0 3.0
15.0 4.0 4.0 6.0 7 .O
ClO t o CIZ,in-
by VOl.
Distd.
.
c.
110-134
0,7400) Per cent
Average B. P. Volume c. cc 50.5 3.5
B. P. of Fraction
F. Cr,
=
( ' 0 )
10.6 1.2
__
55.9
24.1
-
48.3 3.9 (Residue = 28 oc. = 2.3 per cent. per cent by volume diatilling to 2000 e.'= 92.0)
former authors observed that p o l y m e r i z a t i o n began a t 275" C. and that no lower n a p h t h e n e s were formed. They concluded that a t this temperature the zinc chloride probably produced an isomerization of lower naphthenes to olefins. The distillation curves of the liquid products under consideration (Figures 3 and 4) reveal very pronounced peaks corresponding to Cg compounds and representing theatrimer stage of polymerization. Similarly, Tables 111 andl IV show that, for the 200-210" C. and the 150160" C. products, the percentages by volume distilling in the COrange were 47 and 55.9, respectively. A comparison of the properties of the CDfractions of these products with those of known hydrocarbons of similar boiling point indicated that the frac-
TABLEV.
Group
110-134
C8
tions contained, in addition to olefinic and paraffinic hydrocarbons, approximately 15 per cent of naphthenes. It was concluded that aromatics were not present, except possibly in very small amounts: (1) because of their very high specific gravities and refractive indices, as compared to the corresponding values for the fractions; and ( 2 ) because of the low experimental temperature. Cycloolefins may have been present in small amounts (probably not exceeding 5 per cent), Similar considerations for the Clo-12 fractions that were examined (Tables 111 and V) led to the logical assumption of the presence of 30 to 45 per cent of naphthenes, in addition to olefins and paraffins, taking into account the observed specific gravities of the fractions, their percentage unsaturation, and the properties of known hydrocarbons of similar boiling point. The presence of more than small amounts of aromatic hydrocarbons was unlikely, for reasons previously indicated.
PROPERTIES OF FRACTIONS FROM LIQUIDPRODUCT OBTAINED IN 150-160' C. EXPERIMENTS
B . P. of Fraction F.
cs, Ce cs
FIGURE4. DISTILLATION CURVEOF POLYMER MADEAT 150-160" C .
1.7
Average B. P.
Iodine bddi- Iodine Refractive tion Substitution Mol. Sp. Gr. Index Value, Weight, Value N M (G?)
CO)
Per Cent
Un-
44.4- 56.7
50.5
Volume cc. 5.0
-
0,6960
1.3860
198
0.0
79
62
56.7- 59.4 59.4- 62.2 62.2- 63.5 63.3- 64.4 64.4- 65.6 65.6- 06.7 66.7- 67.8 67.8- 73.3 68.9- 73.3 73.3- 80.0
58.0 60.8 62.7 63.8 65.0 66.1 67.2 70.5 71.1 76.6
4.6 4.4 17.4 31 .O 26 .O 13.0 6.2 9.0 12.0 9.2
0.6915 0.6867 0.6838 0.6870 0.6883 0.6912 0.6946 0.6942 0.6944 0.7003
1.3879 1.3891 1.3921 1.3940 1,3956 1.3872 1.3985 1.3984 1.3995 1.4012
205 214 222 223 222 221 220 219 212 205
0.0
83 84 85 86 86 87 88 89 89 92
67 71 74 75 70 76 76 77 74 74
132.8 4.2 6.4 32.0 22.0 45.0 143.8 146.6 123.0 53.0 15.0 10.0 5.5 2.0
0.7347 0.7372 0.7384 0.7398 0.7416 0.7428 0,7441 0.7455 0.7466 0.7488 0.7492 0.7527 0.7564
c.
O
c.
123.3-126..l 126.1-128.9 128.9-130.0 130.0-131.1 131.1-132.2 132.2-133.3 133.3-134.4 134.4-135.6 135.6-136.7 136.7-137.8 137.8-140.0 140.0-145.6 145.6-151.1
-
-
0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0
satn.
63 62 65 65 65 66 67 67 67 66 66 64 65
1.0 4.7 3.4 5.0 3.0 2.2 1.1 2.7 3.5 3.3 0.5 0.5
ROU . 5
Clot0 364-364 Clz,,in- 364-374 clusive 374-384
178.9-184.4 181.6 184.4-190.0 187.2 190.0-195.6 192.7
56 116 40 212
0.7731 0.7763 0.7811
1.4388 1.4402 1.4420
80 74
SO
27.2 27.1 20.0
145 148 150
46 43 47
MAY, 1936
INDUSTRIAL AND ENGINEERING CHEMISTRY
It is interesting to recall a t this point that Ivanov (12) reported in 1933 that the condensate obtained by the polymerization of so-called ethylene concentrate in the presence of zinc chloride at 320-360' C. and superatmospheric pressure was made up of a mixture of paraffin, aromatic, naphthene, and olefin hydrocarbons. The amount of the olefins present was stated not to exceed 30 per cent.
Conclusions The following data were obtained from the fractionation of the liquid products under the various conditions shown in Table I:
Exptl. TemD.
c:
290-310 200-210 150-180
---Chief
cs 5.8 0.07 0.3
Product Distilling in the Motor Gasoline Range
Constituents Present in LiquidProduct cs ci CS c 8 Cia-12 (t0200' c.) Per cent b y volume , 7 4 . 5 ' 10.1 5.9 12.9 18.7 27.3 7.0 1.6 2.0 47.0 32.0 85.9 10.6 1.2 1.7 55.9 24.1 92.0
'
As the experimental temperature was lowered, a successively higher percentage of the liquid product corresponded to the tripolymer of propylene, represented by the CScompounds. Corresponding to the above data, the distillation curve of the 290-310" C. product showed prominent peaks pointing to CS,CS,and Clo-12 compounds, with somewhat lower peaks for the CS, Ca,and C7 compounds. The distillation curves of the 200-210' C. and 150-160" C. products had very pronounced peaks corresponding to CScompounds, and somewhat
559
lower peaks for Clo-u compounds. The C , C7, and CScompounds were present' only in small amounts. From a comparison of the properties of the fractions comprising the peaks of the distillation curves with those of known hydrocarbons of similar boiling point, conclusions were drawn regarding the probable composition of the fractions. No evidence was found of the presence of aromatic hydrocarbons. The data indicated the presence of appreciable amounts of naphthenic hydrocarbons in certain of the fractions, in addition to olefins and paraffins.
Literature Cited Brooks, J . Inst. Petroleum Tech., 14, 751 (1928). Dunstan, Hague, and Wheeler, IND. ENG. C H E M . , 26, 307 (1934).
Ibid., 26, 311 (1934). Egloff, Schaad, and Lowry, J . Phys. Chem., 35, 1825 (1931). Francis, IND. ENG.C H E M . ,18, 821 (1926). Garner, J . Inst. Petroleum Tech., 14, 695 (1928). Gayer, IND.ENQ.CHEM.,25, 1122 (1933). Guthrie and Higgins, Bur. Mines, Rept. Investigations 3159 (1932).
Ipatieff and Egloff, iVatl. Petroleum News, 27, 24G (May 15, 1935).
Ipatieff and Huhn, Ber., 36, 2014 (1903). Ipatieff and Rutala, Ibid., 46, 1748 (1913). Ivanov, J. Applied Chem. (U. S. R. R.), 6, 103 (1933). Johansen, IND. ENQ.CHEM.,14, 288 (1922). Kester and Andrews, IND.ENQ. C H E Y . , Anal. Ed.. 3, 373 (1931). RECEIVED December 23, 1935. Contribution 301 of the Department of Chemistry, University of Pittsburgh. This paper is an abstract of a thesis presented to the Graduate Sohool of the University of Pittsburgh b y 0. L. Brandes in partial fulfillment of the requirements for the Ph.D. degree.
4GHLOROACETOPHENONE Catalytic Oxidation in the Liquid Phase J. 3.STUBBS AND C. E. SENSEMAN Bureau of Chemistry and Soils, Washington, D. C.
oh
CETOPHENONE and its derivatives are easily oxidized by the usual oxidizing agents to the corresponding benzoic acids. Recently VanArendonk and Cupery (6) showed that 4chloroacetophenone and other derivatives can be oxidized in good yields with sodium hypochlorite. The writers (6) found that, although under carefully controlled conditions good yields of acetophenone could be obtained by the catalytic oxidation of ethylbenzene in the liquid phase, there was a tendency toward further oxidation of the ketone to benzoic acid; this behavior was especially evident a t slightly elevated temperatures. Flemming and Speer (1)recently extended the scope of their patented process for the liquid-phase oxidation of aliphatic ketones to include the oxidation of acetophenone. A yield of 25 grams of benzoic acid and 7 grams of formic acid is claimed when 50 grams of acetophenone in acetic acid is oxidized for 12 hours a t 105' C. in the presence of a small quantity of manganese acetate by means of air bubbled through a filter plate a t the rate of 5 liters per hour. The recent work of Groggins and Nagel (9)showing that both acyl groups of acid anhydrides can be made to react with aromatic compounds in the Friedel and Crafts con--
-
.
densation by using increased proportions of aluminum chloride indicates that acetophenones might well be used as a starting material for the production of certain substituted benzoic acids. Thus, Newton and Groggins (3) described the preparation of 4-chlorobenzoic acid by the oxidation of 4-chloroacetophenone with chromic anhydride, 95 per cent yields being obtained. The purpose of the present investigation was to determine the optimum conditions of temperature and of time for the production of 4-chlorobenzoic acid by the liquid-phase catalytic oxidation of 4-chloroacetophenone, using atmospheric oxygen as the oxidizing agent.
Oxidation Experiments The experiments in general consisted of passing air, dispersed by a sintered glass disk, through a solution of 10 grams of 4-chloroacetophenone in 100 ml. of glacial acetic acid in the presence of 0.5 gram of specially prepared manganese dioxide (4). Preliminary experiments had shown that, when
The oxidation of 4-chloroacetophenone in the liquid phase, using manganese dioxide as a catalyst and air as the oxidizing agent, is studied. An apparatus is described, and optimum conditions of temperature and time for the production of 4-chlorobenzoic acid are determined. Yields of the acid in excess of 90 per cent are obtained.
