COMPOSITION of VAPORS from BINARY SOLUTIONS Systems Used in Butadiene Manufacture from Butylene Glycol DONALD F. OTHMER, N A T H A N SHLECHTER, AND WALTER A. KOSZALKA POLYTECHNIC INSTITUTE OF BROOKLYN. N. Y .
Butylene glycol is studied as a souree of butadlene for synthetio rubber manufacture. After a speelal fermentation of m o l a w r or graln, extraotlon and dlstlllatlon procedures are necessary for Its subsequent recovery; then it Is esterifled t o give the diacetate which, in turn, is pyrolyzed t o give butadiene. In order t o design the evaporatlng and dlstlllation equipment required, vapor-llq uld eq ullibrl u m relations were determined for the systems water2,3butylene glycol, acetlc acld-2,3-butylene glyool dlaeetate, 2,3-butylene #lycol-2,3-butytene glycol diacetate, methyl vinyl carbinol acetate-2,3-butylene glycol, and methyl vinyl carbinol acetate-acetic acid. T h e e equlllbrlum data were determined a t atmospheric and various subatmospheric prerures; and the data are oorrelated by methods previously suggested t o make possible the evaluation of the several steps of the procress and the design of the equipment Involved.
by distillation as a means of recovering the diol from a Wered and limed ferment solution has shown possibilities in another investigat'on (6). This work showed that three solvents studied (n-butyl alcohol, butylene glycol diacetate, and methyl vinyl carbinol acetate) had desirable properties. In the esterification unit following the recovery steps, the water formed in the rmction, as well as any prment with the diol, is removed in an azeotropic distillation with a suitable entrainer in the manner already described (3). This work is part of a study of the recovery of 2,%butylene glycol from fermented mash liquors. Vapor-liquid equilibria of various binary systems encountered in the distillation processes following the proposed extractions are considered in order to obtain data necewry for the comparison of these three p r o c e m and for the design and operation of final plants. The system watelc2,&butylene glycol was studied under subatmospheric pressures and is of interest in the design of the preliminary evcrporator. This unit concentrates the fermented liquor up to a strength where diol in appreciable quantities a p pe.ars in the vapors or where the concentration of impurities in the liquor becomes high enough to interfere with the subsequent extraction. Since all the water in the ferment liquors is eventually removed, in order to recover the solids for fertilizer or cattle feed, no heat economy in the over-all procew is to be effected by running an extraction directly on the dilute ferment liquors, as would be the case where the liquors were wasted. However, considerable saving will be expected in the extraction and distillation operation if a concentrated feed is used. This will mean a smaller extracting column and, more important, the use of considerably less solvent and lower heat costs in the subsequent distillations. The system acetic acid-2,3-butylene glycol diacetate is of intereRt in the design of the euterification unit, sinee acetic acid
N 1930 Fulmer and Werkman (1) showed that 2,3-butylene glycol may be obtained by the fermentation of glucose. This diol has since been considered as the starting point for new plastics and has possibilities of being used as an antifreeze. In 1941 the diacetate of the diol became important as a possible source of butadiene for use in synthetic rubber manufacture; as compared with alcohol, in the pyrolysis of butadiene it gives comparatively small amounts of by-products which are readily separable. Various investigations were started on its production and conversion to butadiene. The major steps are: (a) fermentation of carbohydrates (grain under present war conditions) to give liquors containing 3 . 5 4 5 % of 2.3401; (a) recovery of the diol from fermentation liquors; (c) esterification of the diol to Z,%butylene diacetate; (d) pyrolysis of the diacetate to buta1.4500 , , , , , , , , , , , , , diene; (e) separation of the butadiene from the other products of pyrolysis. I n addition to the diol, the fermented liquors contain proteins, alcohols, polysaccharides, and other by-products of fermentation, as well as substantial amounts of suspended material resulting from dead bacteria and broken-down grain structures. Attempted recovery of the diol, by steam distillation of filtered fermented liquor, to which has been added an excess of lime, showed that an over-all recovery of the diol of only a little over 90% could be effected. If the ferment liquor is not made basic and filtered, i t is possible to obtain only a 50% recovery of the diol due to decomposition. These recoveries are not high enough ''2'0' ''.'O' ' ' d o ' "6'0' ''7I0,''@IO*I for a successful commercial prows. Operation is also difficult due to the precipitation of solid WEIGHT PERCENT O f 2.3WJTYLENE GLYCOL material, extreme foaming, and the very viscous Flgura 1. Rdnotive Indices for T h m Syatems bottoms obtained. Solvent extraction followed 805
I
I
"A'
"A'
1
,
I
,
,,
,
~
1 . 1 . 'SO
10'0
INDUSTRIAL AND ENGINEERING CHEMISTRY
896
Vol. 37, No. 9
TABLE I. REFRACTIVE INDEXDATA Water-meso-2,3Butylene Glycol Wt. % butylene Refractive glycol in index a t soh. 24' C. 0 1.3330 9.21 1.3435 17.01 1.3530 23.63 1.3610 29.12 1.3680 33.60 1.3740 38.26 1 3825 41.90 1.3850 45.01 1.3880 47.75 1 3915 49.86 1 3930 50.00 1 3930 51.97 1 3960 55.03 1 3990 57.90 1 4020 62.15 1 4070 64.90 1 4100 70.52 14111 76.10 14190 82.53 1 4250 91.13 14310 100 1 4366
e 760mm o50Omm 350 m m o250mm.
MOL PERCENT COMPONENT A IN LIQUID
Figure 2.
Vapor-Liquid Compositions for Water-maso-2,S-Butyl-
ene Glycol and levo-2,3-Butylene Giycol-meso-2,3 Butylene Glycol Diacetate Systems Component A i s water In t h e upper section and bwo-2,3 butyleno glycol in t h e lower system
and ester must be separated in the esterification still. This problem is complicated by the fact that butylene glycol diacetate tends to decoppose when heated in the presence of mineral acids used for catalysts in the esterification. Therefore, data must be available for vacuum and hence, low-temperature distillation. The system 2,a-butylene glycol-2,3butylene glycol diacet,ate is also of interest in the design of the esterification unit, It is important to be able to separate any unreackd glycol in the column. The butylepe glycol and the diacetate have close boiling points (180' and 192'C.).
The system methyl vinyl carbinol acetate-2,a-butylene glycol is of interest in the separation of these materials following extraction with methyl vinyl carbinol acetate as solvent; i t is believed that this separation could be accomplished satisfactorily at atmospheric pressure. The system methyl vinyl carbinol roetate-acetic acid is of interest in the separation of these materials in the event an excess of solvent is present in a mixture with acetic acid subsequent to the extraction step. MATERIALS A N D METHODS
2,3-BUTYLENE GLYCOL. Technicalgrade meso-2,3-butylene glycol (Schenley RePearcl.1 Institute) consisted of 92% 2,3diol and 8% impurities. The liquid was carefully rectified; and the dis.
-
-P
T
C.
172.0 167.2 158.6 149.6 132.0 110.6 106.0 104.6 102.4
-P
T
C.
191.0 180.5 175.2 166.7 158.1 145.8 127.9 121.4 120.0
,--P T o C. 193.0 189.8 187.0 184.0 181.5 180.4 178.6 177.6 177.6 177.7 179.0
Zevo-2.3-Butylene Glycol-meso-Butylene Glycol Diacetate Wt. % butylene Refractive glycol in index a t soln. 24O C. 0 1.4130 8.74 1.4140 17.05 1.4150 22.30 1.4155 28.25 1.4165 33.10 1.4170 37.50 1.4175 41.00 1.4185 44.20 1.4190 48.80 1.4195 49.45 1.4200 50.2 1,4195 53.1 1.4200 56.1 1.4210 59.9 1.4215 65.4 1.4230 71.2 1.4240 77.9 1.4255 84.5 1.4270 82.3 1.4290 100 1,4300
Methyl Vinyl Carbinol Acetate-levo-2 3Butylene Glyiol butylene Wt. % Refractive butylene wt. % Refractive glycol in index a t glycol in index at soln. 25' C. soln. 25O C. 0 1.3990 52.30 1.4155 9.92 1 ,4025 55.00 1.4160 18.19 1.4040 58.02 1.4168 24.96 1.4060 61.28 1.4178 30.77 1.4080 64.81 1.4190 35.61 1.4100 69. 08 1.4205 39.82 1.4115 73,71 1.4220 43.42 1.4125 76.76 1.4230 1.4139 46.74 84.79 1.4260 49.70 1.4150 92.32 1.4285 52.40 1.4160 100 1.4310
TABLE 11. VAPOR-LIQUID EQUILIBRIUM DATA
-
-
SYSTEM WATER-me80-2,3-BUTYLENE GLYCOL 760 M m . 7 7 - P 500 Mm.-P 350 Mm.7 - P = 200 Mm.Mole ,% water Mole .Yowater Mole *Yowater Mole .% water in: in: in: in: Liquid Vapor' T O C. Liquid Vapor T C. Liquid Vapor T C. Liquid Vapor 3.70 31.75 155.4 4 . 8 5 55.69 126.0 2.90 33.60 140.0 4 . 8 5 60.70 6.00 42.00 153.4 8.30 64.90 119.6 4 . 0 0 42.70 135.8 8.30 69.70 10.00 57.00 144.0 9.50 63.00 123.0 15.00 80.00 109.5 12.69 82.40 15.75 70.80 136.1 19.16 87.20 14.70 75.50 115.4 86.4 27.70 94.60 26.70 87.50 111.4 31.90 93.00 105.8 25.70 92.00 73.6 46.30 9 7 , 6 0 87.6 47.50 97.40 48.80 96.60 9 5 . 2 56.90 98.70 69.4 66.20 98.70 83.6 66.20 98.70 62.00 98.00 79.40 99.00 91.7 6 8 . 3 82.30 99.20 71.50 98.50 81.2 80.60 99.00 89.7 88.00 99.40 67.8 87.70 99.40 81.40 99.00 79.4 91.20 99.10 89.2 92.70 99.60 .67.2 94.50 99.50
SYBTEM ACETICACXD-mesO-2.3-BUTYLENE GLYCOL DIACETATE ,--P = 500 Mm.-P = 300 M m . 7 7 - P = 760 Mm.Mole % acid Mole % acid Mole.% acid in: in: in. Liquid Vapor T O C. Liquid Vapor T O C. Liquid Vapor T C. 9.03 157.5 2.79 16.48 137.8 1.15 2.16 14.12 177.6 5.61 30.96 136..l 2.67 21.26 153.5 7.66 43.40 173.6 11.28 55.43 168.4 7.16 37.89 145.0 10.15 55.32 129.2 14.32 65.77 119.3 13.80 59.85 139.6 13.99 68.60 159.8 23.03 76.87 138.5 29.85 86.89 119.5 35.25 89.86 106.4 98.8 113.3 45.50 93.59 34.13 89.58 120.9 54.85 96.69 63.65 97.10 85.5 69.19 97.49 112.6 70.49 98.52 103.9 94.3 82.56 99.08 81.3 84.96 99.20 109.3 79.80 98.80 89.9 96.95 99.52 75.2 90.86 99.70 106.8 8 4 . 9 8 99.15
-
-
-
150 Mm.? Mole, % acid in: Liquid Vapor 0.37 2.65 1.65 16.53 6.71 48.39 13.32 70.59 26.41 86.13 36.72 91.72 61.50 96.92 71.49 98.26 89.36 99.17
SYBTEM leVO-2,3-BUTYLENE GLYCOL-meSO-BUTYLBNE GLYCOL DXACBTATE = 760 Mm.,--P 500 Mm.F--P 350 M m . 7 -P = 250 M m . 7 Mole $G glyool Mole % glycol Mole glycol Mole % glycol in: in: in: in: Liquid Vapor T C. Liquid Vapor T e C. Liquid Vapor' T C. Liquid Vapor 8.0 17.0 153.8 1.0 6.5 8.5 24.0 165.2 2.6 4 . 5 172.0 8.5 19.0 16.0 27.0 150.6 22.2 28.5 158.9 7.4 168.5 4.8 148.0 16.2 30.0 54.0 163.9 48.0 37.0 167.2 28.0 9.6 14.0 49.5 65.0 145.8 28.5 64.0 61.0 64.0 153.0 19.0 2 6 . 0 164.7 39.5 48.0 6 8 . 8 144.4 70.0 71.0 153.0 71.0 34.2 4 1 . 5 164.6 58.5 61.0 7 2 . 5 143.5 75.5 75.0 153.1 77.0 4 8 . 8 164.6 41.0 71.5 66.5 75.0 143.5 80.5 79.0 153.2 83.0 61.0 164.8 55.0 8 0 . 5 72.5 90.0 83.0 143.8 88.0 84.5 153.7 71.3 73.0 165.1 84.0 75.0 87.5 144.0 93.3 8 7 . 5 154.0 92.5 80.8 7 9 . 5 165.2 91.5 82.5 9 3 . 0 144.5 194.5 98.0 9 0 . 5 165.4 95.0 95.5 93.8 100.0 100.0 100.0 100.0 165.8 100.0 100.0 154.8 100.0 100.0 145.0
-
September, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
a97
excess of acetic anh dride for 1-2 hours, washed with water, dried with calcium cgloride and finally purified by distillation. The purified methyl vinyi carbinol acetate distilled a t 112113' C. METHODS. The experimental apparatus and procedure were described previously (9,4 ) . .Because of. the high-boiling temperatures encountered, the still was h e a d y insulated to reduce radiation losses and errors. The three systems containing diol were analyzed in a refractometer. Refractive index data obtained for these systems from synthetic mixtures are shown in Table I and Figure 1. The two systems containing acetic acid were analyzed by titration of weighed samples for acetic acid. CORRELATION OF DATA
VAPOR PRESSURE OF WATER m m.Wg
Figure 3. Log Plot of Partial P~OSBU~OS of meso-2,3Butylene Glycol from Aqueous Solutions VI. Vapor Pressures of Water a t Same Temperatures
tillate coming over up to 183" and above 184' C. at 760 mm. mercury waa discarded. The fraction collected had a refractive index of 1.4366. This material was used in studyin the system water-diol and for makin the ester used in the otter two syeteras. The diol for the k r d and fourth systems was Zeuo-2,3butrlene gl col similarly purified with a boiling range of 176177 C. a n i s refractive index of 1.4130. The technical-grade me80 2,Q-BUTYLENEGLYCOL DIACETATE. ester (Schenley Research Institute) waa refined in the same manaa the diol. The distillate was collected over a ran e of 190ner 193' C. and had a refractive index of 1.4330 at 24' Them values are in agreement with accepted data. ACETICACID. Chemically pure glacial acetic acid w&s used. The trace of water present waa removsd by discarding f o r e m of the distillations, since it steam-distilled over to give B sli'ghtly milky condensate. METHYLVINYLCABBINOL ACETATB. This compound (M.C. V.A.) waa made from meth 1 vinyl carbinol obtruned from Shell Development Company. $he alcohol waa refluxed with a slight
8.
The data are given in Tables I1 and 111, and plots were made of the liquid and vapor relations with temperature (boiling point end dew point curves) and of the vapor composition as a function of liquid composition alone (2,y curves). Plots were also made of the logarithms of partial pressure, vapor composition, relative volatility, activity, and equilibrium constant against the logarithms of vapor pressure of water a t the same temperature and against the total pressure at the same temperature. These methods of correlation of P, T,2,y data were discussed elsewhere (6).,Only a few of these plots are included (Figures 1 to 8 ) to show the relative consistency of the experimental work. The moothed values of vapor atid liquid comuositions of Table IV were obtained by picking off values from logarithmic plots of all the experimental data. WATER- 2,%BUTYLENE GLYCOL.The data for this system (Figures 2, 3, and 4) exhibit the characteristics of a normal binary mixture of two components with widely differing boiling points. No constant-boiling mixture is formed a t any pressure; the relative volatility of water is always great and increases with decreasing pressure. This is important in the operation of the evaporator and a simple rectifying system which may be attached to it, since considerable water can be evaporated from the aqueous solutions without any loss of diol overhead. Figure 4 (vapor composition plotted on log paper against total pressure) is a convenient means of correlating data of this type. The slopes of the lines on the total pressure plots shown are nearly all the same, as expected, since the molar latent heats are relatively close for the two compounds (water 9700 Figure 4. Log Plot of Vapor calories, diol 1 1 , 7 0 0 Composition of System calories). Water-mew 2,3 Butylene ACETIC Acr~-2,3-BuG l y w l a t Constant Liquid TYLENE GLYCOLDIACECompositions vs. Total TAW. This system shows Pnuuns
- -
838
Vol. 37, No. 9
INDUSTRIAL AND ENGINEERING CHEMISTRY
some irregularities, which are expected since acetic acid is known to be associated. The wide variation of the boiling points also increased experimental difficulties and probably experimental errors. The separation, however, between the diester
and acetic acid is good (Figures 5 and 6 ) . Change of pressure seems to have little effect on the relative volatilities, so that rectification of acetic acid from the diester should be equally easy at m y pressure. When the s-y diagrams for the several pressures are superimposed rather than separated, as in Figure 5, the curves a t constant pressures intersect and cross one another. This behavior may result from the variation of the degree of associat>ionof acetic acid molecules with temperature. 2,3-BUTYLENE GLYCOL-2,3-BUTYLENE GLYCOLDIACETATE. This system exhibits a minimum constanhboiling mixture (CBM). On a log plot of total pressure ageinst vapor pressure of water, the lines cross as previously noted for binary mixtures having CBM. Reduction of the pressure increases the amount of the diester (less volatile) in the CBM (Figure 2 ) which would be expected by analogy with other systems. At 250 mm. mercury pressure, however, the composition of the CBM is still very high in glycol (63.2 mole %). The following shows the change in composition of the CBM with total pressure: Pressure, Mm. H g 760 500 350 250
yo Diol in CBM
w., c.
77.0 71.2
177.6
63.2
143.4
67.5
0
164.6 153.0
Figure 7 shows log vapor composition as a function of log total pressure; the change of the slopes of the lines with differentvalues of 5 is due to the presence of the constant-boiling mixture. The slopes of the line at higher concentrations of diester are positive. As the concentration decreases, the slope decreases until it passes through zero and the line is horizontal. A horizontal line on this plot represents a concentrabion where the vapor-liquid equilibrium relation is independent of pressure (and also of temperature). Figure 7 shows this at about 55% where the curves for various pressures cross a common point. In general, such a fixed point (where, a t constant x, both temperature and
a: 0 a
a >
z P 0 U
0 t-
w
u a I-
z
w
0
CT W
a -1
0
E
Figure 6.
Vapor-Liquid Compositions for System Acetic Acid-2,3-Butylene Glyool Diacetate
Vapor oomporltlons t o 60% 8wtlo 8ctd a n ind1oat.d on left and above So%, on right.
MOL PERCENT ACETIC ACID
Figure 6. Boiling Point and Dew Point Curves for System Acetic Acid-2,3-Butylene Glycol Disoetate
INDUSTRIAL AND ENGINEERING CHEMISTRY
September, 1945
899
pressure may vary without changing y) 80 MVCA-ZWCA,~-BUTYLENE GLYCOLAT ATMOSPHERIC may be expected if PRESSURE there is a variation 60 Mole % MVCA in: T:"~, Mole % MVCA in: in slope from plus Liquid Vapor Liqd Vapor to minus, or minus 0 0 179.0 0 0 10.0 4.6 0.8 3.8 176.6 to plus, in the log 40 2.4 6.9 8.0 170.4 24.8 plots of y against 4.7 9.1 11.9 166.0 64.8 22.2 6.9 18.2 160.4 63.6 total pressure or y 21.0 26.0 11.6 141.2 72.6 30 30.7 36.2 17.9 133.9 80.4 against vapor pres43.6 88.7 31.2 126.6 88.0 > sure of a reference 44.2 48.6 120.6 46.4 91.7 61.6 66.3 76.7 114.9 97.6 I Rubstance. 86.6 74.2 69.4 114.0 98.6 a eo 82.2 98.8 88.3 78.7 113.6 The presence of W 100 100 90.0 87.6 111.8 c t h i s CBM indi94.3 92.3 u) 100.0 100.0 IS cates the importance of oomplete esterification of the 0 glycol in the esteriIO fying column. It TABLEI IV. SMOOTHED DATAOF VAPOR-LIQUIDEQUILIBBIA shows that the di.B 8 Y I T m Y WATEU-mbsO-2,3-BUTYLElNE GLYCOL .7 acetate may be used P 760 Mm. P = 600 Mm. P 360 Mm. P 200 Mm. aa a solvent only Mole Mole Mole Mole if esterification in TOTAL PRESSURE IN mm I49 % % % % Mole p / . water water water water the presence of a Waterin T in T in T in T in great excess of the Figure 7. Log Plot of Vapor ComLiquid C. vapor C. vapor C. vapor * C. vapor positions of System 2,3-Butylene diester is feasible 0.0 142.0 0.0 0 182.5 0 . 0 168.0 0 . 0 168.6 Giycoi-2,3-Butyiene Glycol Diace6 168.6 38.6 152.6 4 3 . 6 141.2 64.20 126.3 60.2 or if some other 10 168.7 6 6 . 6 143.0 64.6 130.7 71.70 116.2 76.0 tate a t Constant Liquid Composimethod of separa20 79.2 127.0 83.8 114.0 87.60 97.6 90.1 142.3 tions vs. Total Pressures 80 89.8 113.7 91.6 101.2 93.70 84.8 96.2 127.8 tion than ordinary 96.7 96.60 77.6 97.2 92.1 40 116.7 94.6 103.6 rectification is used 98.0 97.6 72.7 98.6 60 110.3 96.7 8 7 . 0 97.90 99.1 for the diol and its diacetate. Such an additional method 70.4 60 106.6 97.7 94.7 98.3 8 4 . 6 98.60 69.0 99.3 70 104.6 98.6 92.6 9 8 . 8 82.8 99.00 would be, for example, the separation of the diol by a partial 99.6 80 102.8 9 9 . 0 91.0 99.2 81.3 99.40 68.1 pressure distillation of the heterogeneous CBM of the diol 101.2 99.4 89.7 99.6 80.1 99.70 67.3 99.8 100.6 99.7 89.3 99.8 79.6 99.86 6 7 . 0 99.9 and kerosene away from the diacetate. 100 100.0 100.0 88.7 100.0 79.3 100.00 66.6 100.0
TABLE111. VAPOR-LIQUIDEQUILIBRIUM DATAFOR
SYSTEM
':'8?
!4
*
5
-
-
-
O
O
O
!!
CONCLUSIONS &%TIM
Aoid in % Liquid 0
6
10 20 30 40 60 60
70 80
90 96 100
ACETIC AClD-meUO-2,3-BVTYL1NE GLYCOL DIACETATE P-760Mm. P-600Mm. P-300Mm. P-160Mm. Mole Mole Mole Mole apid % in * C. vapor * C. 193.7 178.7 0 171.3 29.3 186.6 164.3 6 0 . 0 177.6 160.5 76.0 162.1 149.7 138.6 86.4 141.0 130.0 92.0 123.6 96.0 136.6 118.0 97.4 131.3 113.0 98.4 127.0 123.2 99.2 108.7 99.2 120.0 99.7 106.7 99.7 99.8 118.6 99.8 104.7 117.8 100 104.4 100
T
acid "/c in vapor 0 30.1 60.2 76.2 86.8 92.2 94.7 96.5 98.1
P = 760 Mm. Mole Mole % G1 aolm
d&d 0 6 10 20 30 40 60 60 70 80 80
96 100
T
P
-
%
T
glycol in liquid
* C. 192.7 0 189.8 7.6 187.2 14.7 184.0 28.6 182.0 37.7 180.6 46.7 179.3 66.7 178.3 83.6 72.0 177.7 177.6 79.0 86.8 178.0 178.4 9 2 . 0 179.0 100
500 Mm.
1Mole
7
T
glycol in liquid
' C. 177.6 0 10.6 173.6 171.8 17.6 168.8 28.6 167.0 38.7 166.9 47.7 166.2 66.0 164.8 6 3 . 0 164.7 70.6 164.7 76.6 166.1 86.0 166.4 90.7 166.7 100
in
O C. vapor 161.0 0 163.7 31.2 146.8 60.6 134.0 76.4 86.1 123.7 116.3 9 1 . 9 110.6 96.1 106.2 96.7 100.1 98.2 96.7 99.1 99.7 91.7 90.4 99.8 89.6 100
P
-
%
T
apfd
350 Mm. Mole
%
T
glycol
ln
O'C. liquid 166.6 0 12.0 162.4 160.6 20.0 168.0 29.7 166.3 38.7 164.8 47.2 163.7 65.7 163.1 62.6 163.0 68.6 163.2 74.6 163.8 83.6 164.2 90.0 164.8 100
7
T
apfd
1. The evaporation of fermented liquors containing 2,3butylene glycol prior to extraction can be carried out without excessive loss of the diol. Any diol that may come over can be easily returned in a small rectifying column in the vapor line of
O C. vapor 138.3 0 131.6 40.0 124.7 60.0 112.6 79.9 103.3 89.0 96.3 93.6 90.7 96.8 86.1 97.3 82.0 98.3 78.3 99.0 74.7 99.7 73.0 99.8 71.6 100
p = 260 ~ m . Mole
%
glycol In O C. liquid 164.6 0 162.0 12.0 160.0 21.6 147.4 32.6 146.6 4 1 . 2 144.5 48.6 143.7 56.5 143.6 62.0 143.6 67.3 143.9 72.0 144.6 81.8 144.7 89.0 146.1 100
MOL PERCENT COMPONENT A IN LIQUID
Figure 8. Vapor-Liquid Compositions for MVCAIevo-2,3-Butyiene Glycol and MVCA-Aoetio Add Systems
INDUSTRIAL A N D ENGINEERING CHEMISTRY
900
the final effect of a multiple-effect evaporator. The uw of low pressures is advantageous for this separation. 2. To separate acetic acid from 2,Sbutylene glycol diacetate with a sufficiently low bottom temperature (120" C.) to minimize decomposition due to the presence of sulfuric acid as an esterification catalyst, very low pressures (100 mm.) are required. I t may be necessary to neutralize the sulfuric acid before r'emoving tkie acetic acid from the diester by distillation, 3. Separation of 2,Sbutylene glycol and 2,Sbutylene glycol diacetate by distillation is complicated even a t very low pressures, owing to the formation of a CBM. Therefore, the use of diester as a solvent for the diol is dependent upon a successful esterification of diol in the presence of excess ester or the use of another method of separation than straight rectification.
Vol. 37, No. 9
Appreciat,ion is expressed to Sclieiiley Research Institute aiici especially to the director, A. J. Liebmann, for the cooperntion which made thiR work possihle. LITERATURE CITED
(1)
Fulmer, E. I., and Werkman, C. H., Proc. Iowa Acad. Sn'., 36, 114 (1929).
(2) Gilmont,
R.,and Othmer, D. F., IND.ENQ.CXEM.,36, 1081
(1944).
(3) Othmer, D. F., Chem. & Met. EW., 48,91 (1941). Othmer, D. F., IND. ENG.CHEW,35, 614 (1943).
(4)
(5) Othmer, D. F., Bergen, W. S., Shlechter, N., and Bruins, P. I?.. Ibid.,37,890 (1945). ( 8 ) Othmer, D. F., and Gilmont, R.,Ibid., 36,858 (1944). PREBENTED before the Division of Indutrial and Engineering Chemistry et the 107th Meeting of the AMERICAN CHRMICAL SOCIETY in Cleveland, Ohio.
ESTERIFICATION of 2,3-BUTY LENE GLYCOL ACETIC ACID NATHAN SHLECHTER, DONALD F. OTHMER, AND SEYMOUR MARSHAK POLYTECHNIC INSTITUTP: OF BROOKLYN, N. Y .
T h e rates of esterification a t different temperatures of both glycol and the glycol monoester were determined, using sulfuric acid as catalyst. The reaction was also studied In an excess of glycol diacetate equivalent to the amount which would be present in an extract layer from a liquid extractton column, using butylene glycol diacetate as a solvent for the recovery of glycol. The reaction does not appear t o be a single, simple one of first, second, or third order. Equilibrium constants and heats of reactions are reported. An empirical p l o t was developed which correlated the reaction data a t several temperatures, and the slopes of the straight lines obtained were used as a measure of activation energy. A continuous esterification was operated with an entraining agent to withdraw the water formed.
STEBIFICATION of 2,3-butylene glycol to its diacetate is an important step in the conversion of the 2,3-diol to l,&butadiene. The kinetics of this reaction at different temperatures are important for the design and operation of an esterification unit (either batch or continuous). The stoichiometric equation for the esterification of diol with acetic acid and the reverse reaction is written &s follows:
E
CH3 I
CHOH + 2CHsCOOH I CHOH I
.Diol
ticetic acid
CH,
I
CHOOCCHa f 2Hz0
Diacetate
Water
Probably this reaction should be regarded as taking place in two steps, a monoacetate being formed. as an intermediate product. This mechanism would give two opposite secondorder reactions forming and decomposing the monoacetate, followed by two more opposing second-order reactions forming and decomposing the diacetate. The monoacetate was actually found
in the reaction vessel ( 8 ) ; it was prepared by treating the diol with the calculated amount of acetic anhydride to give a material having .the following properties ( 2 ): clear, colorless liquid with a mild pleasant odor; soluble in water, alcohol, ether, benzine, and acetone; sparingly soluble in aliphatic hydrocarbons; refractive index 1.4214 at 25" C.; boiling range 178-183" C. The industrial application of the reaction to give the diacetate would be complicated by the high boiling points of the alcohol (181O C.) and the ester (192O C.), A projected method is to urn a continuous esterification column fed with the stoichiometric quantities of glycol and acetic acid plus the required amount of sulfuric acid catalyst. An entrainer may be recycled in the upper or dehydration section to remove the water of reaction. The glycol and the diacetate form an azeotrope (6),and the reaction must proceed to complete utilization of the glycol. This can be accomplished by having excess acetic acid present as permanent holdup in the column below the feed plate. EXPERIMENTAL METHODS
To determine conditions at equilibrium, a galvanized iron tank, 12 X 9 X 14inches (31 X 23 X 36 cm.), was set up as a constanttem erature bath and lag ed with a 2-inch layer of magnesia. S.A.%. 30 motop oil was uw%in the bath and was heated by a 750watt immersion heater which was connected in series with a relay and a mercury temperature regulator. The regulator an: an agitator held the bath temperature constant within, *I C. The reaction took place in a one-liter, three-neck flarjk immersed in the oil bath. A thermometer, water-cooled condenser, and stirrer, inserted through a mercury seal, were installed in the three openings. The lower-boiling reactant, acetic acid, contained the required amount of sulfuric acid (optimum = 0.0162 mole per mole of glycol, 2 ) . It was heated in the reaction Bask, and the highboiling reactant butylene glycol, was heated in another flask. When both liquids were a t reaction temperature the glycol was rapidly introduced into the flask. At the same instant an electric timer was started. A t regular intervals samples were withdrawn by a I-ml. pipet. Samples, 0.50 to 0.85 cc. in volume, were transferred to Erlenmeyer flasks kept a t about -10'. C . by immersion in a mixture of ice water, and calcium chloride. The m p l e s froze instantly and thus arrested further reaction. Each sample waa later titrated for free acetic acid.
