+z9J~Prod uct a d Process A section devoted to information on the development of products 'and the processes for making them on any scale with industrial implications, and including economics and market development
ying Ion Exchange Resins as Catalysts L. M. REED', 1. A. WENZEL, AND J. B. O'HARA2 Deporfment of Chemicul Engineering, Lehigh Univerrify, Bethlehem, Po.
HE production of ethylene glycol rose from 226,673,000 pounds in 1947 to 630,261,000 pounds in 1953 (6). This increased demand for ethylene glycol in recent, years has stimulated research in quest of more efficient manufacturing processes. Ethylene glycol is manufactured in the United States by three processes. Thc most common and oldest of these is the chlorohydrin process, which n-as originally employed by the Carbide and Carbon Chemicals Corp. in 1922 ( 2 , 3). Ethylene chlorohydrin is produced by the reaction of chlorine and water on ethylene. This material may then he hydrolyzed directly to glycol or made t o react with calcium hydroxide (milk of lime) to produce ethylene oxide, which is subsequently hydrolyzed t o ethylene glycol. X more recent process of direct oxidation of ethylene with air to ethylene oxide, followed by hydration to ethylene glycol, is assuming major importance (8). Another process is based on the production of glycolic acid from formaldehyde and carbon monoxide (6, 7 ) . The acid is then esterified and reduced to form ethylene glycol. In both processes employing ethylene as a base material the final step may involve the hydration of ethylene oxide t o ethylene glycol. This hydration has usually been conducted in the liquid phase and catalyzed by sulfuric acid. The reaction proceeds readily and high yields are obtaiiicd; ho\i-ever, the sulfuric acid contaminates the product, and makes purification difficult. The uncatalyzed reaction proceeding a t high pressure and elevated temperatures has bcen studied (4),although commercial interest has not been evident. Basic catalysts, mainly calcium and sodium hydroxides, have albo proved effective. The use of a solid catalyst xi-ould simplify t,he purification problem significantly. Such was the aim of Cartniell and others ( 1 ) in t,heir studies of the vapor phase hydrat,ion of ethylene oxide over a variety of solid catalysts. Acid-type catalysts such as phosphoric acid on a!umina, silica gel, or thoria, gave no glycol, but silver oxide on an alumina. catalyst was found t o be effective. Conversions to glycols ranged from 20 to 30yowith a corresponding yield of about 80 and 40yoof ethylene glycol. Ion exchange materials appear ideally suited for this use, being established as effective ca,talysts in numerous organic reactions. Possible advantages of an ion exchange material over the conventional acids or hases are: The product8 are obtained free of contamination by the catalyet, the catalyst is readily retained for continued usage, and the catalyst is often more 1 2
Present address, Air Products, Inc., Allenton.n, Pa. Present address, Oh-Mathieson Chemical Corp., Niagara Falls, K . Y.
February 1956
?peelfir in obtaining the desiied product. This study was therefore conducted to explore the feasibility of the use of ion cxrhange materials in catalyiing the hydration of ethylrne oxide t o ethylene glycol. Thc experimental apparatus was designed to study the reaction in it continuous process with fixed-bed reactor. Separate systems n ere rmploj cd for introducing the water and ethylene oxide under ront rolled conditions. Apparatus was designed for continuous reaetion in fixed-bed reactor
Figure 1 is a schematic diagram of this apparatus. The reactants, ethylene oxide and water, are contained in receivers 16 iind 17, from where they pass through 8eparat.e rotameters, 5 and 15, arid heating coils, 6 and 14, and finally enter the t o p of the reactor, 8. The reactants pass down through the catalyst bed and are collected in a receiver immersed in an ice bath, 11. The vent from the receiver passes through a dry ice trap, 12, t o collect any uncondensed vapor of ethylene oxide. The analysis of the products was made by distillation. The reactor consisted of a 3/r-inch, schediile 40 stainless steel pipe, 24 inches in length, with convenient fittings for introducing the catalyst. Seven thermocouple wells were installed in the center of the pipe for measuring the temperature in the catalyst bed. .bound the pipe was coiled a 750-watt asbestos-covered Chromel-A wire. This heater was controlled with a variable voltage transformer. Procedure. The ethylene oxide and water feed systems were charged and pressurized. The catalyst bed was brought to the desired temperature and pressure by passing steam through it. Fo1lov;ing this ethylene oxide was introduced, and the system was allowed to reach equilibrium. Sufficient product was collected to ensure reasonable accuracy in analytical procedure. The time required t o collect a sample ranged from 30 minutes to 1.5 hours. The products were distilled a t two pressure levels. All components with boiling temperatures equal t o and below that' of water. were distilled at atmospheric pressure in a small column Rith packed height equivalent t o approximately 10 theoretical stages. The product remaining vias fractionated a t a pressure of 20 mm. of mercury in a 12-mm. Todd column (9). This column has a packed height equivalent t o 41 theoretical stages, and a holdup in the column of about 10 cc. Product components could be separated by this procedure with an accuracy within f l ml.
INDUSTRIAL AND ENGINEERING CHEMISTRY
205
PRODUCT AND PROCESS DEVELOPMENT
ION EXCHANGE MATERIALS make possible catalytic hydration of ethylene oxide
. . . without product contamination
. . . without consumption of catalyst As an average run resulted in the collection of 50 t o 75 ml. of converted products, this results in an analytical accuracy within f 2 % . Ethylene oxide could be separated more sharply, giving conversions accurate t o i1%. Ethylene oxide i s readily hydrated at moderate temperatures and pressures
range and the mole ratio was chosen t o give high yields. T h e temperatures plotted are bottom temperatures in the catalyst bed and the pressures correspond closely t o the saturated steam pressure for the temperature plotted. The yield declined slowly from 83% a t 180’ F. to 68% a t 300’ F., and fell rapidly t o 59% a t 330’ F. The percentage conversion increased rapidly from 17y0 a t 180” F. t o near complete conversion a t 250’ F. and above. Effect of Mole Ratio on Yield and Conversion. The effect of the mole ratio, moles of water per mole of ethylene oxide, on the percentage yield and percentage conversion is shown in Figure 3. The temperature was held constant within a few degrees for these runs. It is evident that a large excess of water is necessary to prevent the formation of higher glycols. The maximum yield is reached at mole ratios of approximately 15 to 1. The percentage conversion increases rapidly a t mole ratios above 12 t o 1. This cannot be explained on the basis of increased concentration of the water, and must be a result of altering the diffusional rates within the catalyst bed.
It was found that all strong acid and strong base ion exchange Satisfactory catalyst life i s expected materials in the hydrogen and hydroxyl forms would catalyze the reaction a t moderate temperatures and pressures. Only The studies made here allow no final conclusion t o be drawn commercially available ion exchange materials were employed, on the life t o be expected of the catalyst. Tests under and these are described in Table I. The major portion of the mixed phase conditions up t o 12 hours were made with no experimental work was done with catalyst A, a strong acid resin of the nuclear sulfonic type. Catalyst F, a phosphonic-type resin, functioned as well as the sulfonic types. Catalyst G, a strong base resin, catalyzed the reaction, though it produced more of the higher glycols. Catalyst H, a weakly acidic ion exchange material of the carboxylic type, failed t o speed the reaction. Tables I1 through I V summarize the results. The percentage conversion is the per cent of the ethylene oxide fed which is converted t o other products. The percentage yield is the percentage ratio of the moles of ethylene glycol formed per mole of ethylene oxide reacted. This term is sometimes called “selectivity.’ The “space velocity” term used in the table is a modified term, or mass velocity: mass flow of reactants per unit volume of catalyst bed per unit time. 15 The mass flow is the total flow of reactants, both water and ethylene oxide, regardless of the phase conditions. I t s use is convenient, since runs were made in the heterogeneous liquid vapor 3 phase region. Here the relative amounts of liquid and vapor present change during passage through the bed. Phase Conditions. Runs 14, 15, 4A, and 5A were made under conditions of temperature and pressure t o assure vapor phase throughout. 2 The yields were low in all vapor phase runs, ranging from 42 to 51%. The major portion of the work was therefore conducted under conditions of temperature and pressure which established a heterogeneous system of vapor Figure 1. Schematic diagram of experimental apparatus and liquid phases. Effect of Temperature and Pressure. The 1. Ethylene oxide cylinder 10. W a t e r cooler 1 1 . Receiver in ice bath combined effect of temperature and pressure on 2. Line valve 12. Dry ice trap 3. Needle control valve the percentage yield and percentage conversion 13. Vent 4. Nitrogen line is shown in Figure 2 . The data for these curves 14. W a t e r heater 5. Rotameter, ethylene oxide were obtained using catalyst A, with the mole 15. Rotameter, water 6. Heater, ethylene oxide 16. Pressurizing vessel, ethylratio and mass space velocity constant. The 7 . Pressure g a g e ene oxide space velocity was chosen to give a wide range of a. Reactor 17. Pressurizing vessel, water conversions over the available temperature 9. Thermocouples 18. Filling plug
-
$
206
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 48, No. 2
PRODUCT AND PROCESS DEVELOPMENT
Table I.
Properties of Ion Exchange Materials Used a s Catalysts
Catalyst
Trade Name
G e n e r a l Properties
A
Permutit Q ( P e r m u t i t Co.) Dowex 5 0 (Dow Chemical Co.) Zeo K a r b (Permutit Co.) Zeo R e x ( P e r m u t i t Co.) Permutit H ( P e r m u t i t Co.) D u o l i t e C-60 (Chemical Process Co.) Permutit S ( P e r m u t i t Co.)
C a t i o n exchange resin of sulfonated polystyrene, nuclear sulfonic exchange group, strong acid t y p e . -1640-mesh beads C a t i o n exchange resin of sulfonated polystyrene. nuclear sulfonic exchange group, strong acid t y p e , 9% cross-linked, - 2050-mesh b e a d s C a t i o n exchange m a t e r i a l of sulfonated coal; sulfonic, carboxylic, a n d hydroxyl exchange g r o u p s ; strong acid t y p e , -2050-mesh g r a n u l a r particles C a t i o n exchange resin of sulfonated phenolic t y p e . sulfuric a n d h y d r o x y l exchange groups, s t r o n g acid t y p e , - 2050-mesh g r a n u l a r particles C a t i o n exchange resin of carboxylic acid t y p e ; weak acid t y p e with carboxyl exchange group, - 1640-mesh g r a n u l a r particles C a t i o n exchange polystyrene resin of nuclear phosphonic t y p e , m e d i u m s t r e n g t h acid t y p e w i t h phosphonic exchange group, -2060-mesh beads Anion exchange resin of a m i n e t y p e , a m i n e exchange group. strong base t y p e , - 1640-mesh b e a d s
B C D
E F G
+
+
+
+
+
+
+
90
5ln 80 W
z > 70
I-
u W w 60
a n
: - 50 >
+
TE hi PER AT U R E , 'F.
Figure 2. 0. 0.
W
X
Combined effect of t e m p e r a t u r e a n d pressure on yield a n d conversion
40
w
n
70conversion of ethylene oxide 70yield of ethylene glycol
30
Mole ratio, H?O/C?H40, 12.8: 1 to 13.0: 1 Mass space velocity, grams per hour X ml., 5.2 to 5.7 Catalyst, Permutit Q, H form Pressures approximately those of saturated steam at corresponding
5
Figure 3.
temperatures
noticeable degradation of the catalyst. Throughout the mixed phase studies no attrltion of the catalyst was found. Thus, a satisfactory life is expected. I n runs made under vapor phase conditioiis an initial, highlv exothermic reaction was n o t e d . This reaction iesulted in uncontrollable reactor bed temperatures
T a b l e II.
10 15 20 MOL R A T I O , H z O * C 2 H 4 O
25
Effect of m o l e ratio o n yield a n d conversion
Catalyst, Permutit Q, H form Temperature range, 203' to 21 1
F.
and charring of the catalyst. I n runs made under heterogeneous phase conditions, this initial high rate of energy evolution is absorbed in vaporizing part of the water present and hence is
Experimental Data
Run S o .
12
13
Catalyst
A , H Form
14 C, H
Form
15 A, H
Form
16n Sand
1.1
29
3A
4.1
A , H Form
12 12 12 24 12 12 12 12 12 Bed depth, inches 100 100 100 100 100 200 100 100 100 Catalyst, ml. Temp., r. 238 213 230 235 215 207 207 230 206 TOP 284 216 240 208 219 226 215 249 271 Middle 244 320 225 225 226 221 268 271 206 Bottom 200 20.3 2 0 . 0 20.5 20.0 20.0 20.0 25.0 25.3 Reactor pressure, lb./sq. inch abs. Mass space velocity, g./hr. X inl. 5.8 5 7 6.0 9.3 5.7 3.1 3.1 6 0 5.5 Total 085 0.75 1 . 2 3 0.79 047 0.98 0.73 0.40 0.90 CzHaO 4 9 8.1 5.0 2.6 5.3 4.5 5,0 2.7 5.1 Hz0 15.4 14.0 1 7 . 3 16 0 17.1 13.4 16.4 1 4 . 1 11.6 Mole ratio H Z O / C ~ H I O 58 42 70 48 74 8 61 56 97 Conversion. 93 Yield, 70 82 74 60 50 48 83 81 46 Ethylene glycol 9 8 23 21 7 9 12 20 Diethylene glycol 18 15 1.3 9 32 8 10 42 Higher polymers , . . .. , , . . 1G Intermediates b 0 R u n 16 was not analyzed because of small conversion. b Analyses of runs 3A and before did not include products boiling between l o O D and 212' F., termed "inter mediates."
February 1956
INDUSTRIAL AND ENGINEERING CHEMISTRY
207
PRODUCT AND PROCESS DEVELOPMENT Table 111. Run No. Catalyst
5A
6A
7A
SA
SA
10A
11A
12A
13A
A, H Form
Bed depth, inches Catalyst, ml. Temp., ' F. Top Middle Bottom Reactor pressure, lb./sq. inch abs. Mass space velocity, g./hr. X ml. Total CzHaO
Hz0
Mole ratio HzO/C?HaO Conversion, % Yield % Etiylene glycol Diethylene glycol Higher polymers Intermediatesa Products boiling between
Experimental Data
12 100
12 100
12 100
12 100
12 100
290 291 311 20.0
198 228 238 25.0
213 225 232 25.0
212 242 247 25.0
217 209 238 234 248 270 30.0 45.0
12 100
12 100
12 100
177 214 204 178 220 205 180 221 210 16.0 10.0 2 0 . 0
5.4 0.62 4.8 18.7 42
5.3 0.71 4.6 15.5 94
4.4 1.37 3.0 5.4 58
2.6 0.83 1.8 5.3 97
5.6 0.85 4.8 13.8 98
5.6 0.89 4.7 12.8 99
5.3 0.83 4.5 13.2 40
5.2 0.83 4.4 13.0 17
5.2 0.84 4.4 12.9 63
61 20
82 13 5
68 16
52 22 20
75 14 7 4
70 14 12 4
78 9
10
82 10 8
76 13 9 2
19A G, H Form
20A I),H Form
22rlb 21A C , H A, Na Form Form
...
19
looo and
12 100
16
. . . . . .
6
...
3
212O F.
Table IV. 14A
R u n No.
Experimental Data 15Aa
A, H Form -___
Catalyst Bed depth, inches Catalys5 ml. Temp., F. TOP Middle Bottom Reactor pressure, Ib./sq. inch ahs. Mass space velocity, g./hr. X ml. Total CzHaO Hz0 Mole ratio HzO/C&rO Conversion, % Yield % Etiylene glycol Diethylene glycol Higher polymers Intermediates C
16Ab E, H Form
12 100
12 100
292 303 302 65.0
324 207 215 331 214 215 330 213 220 95.0 20.0 20.2
5.4 5.7 0.81 0.83 4.6 4.9 14.0 14.0 100 100
12 100
17A D, H Form
5.2 0.85 4.4 12.5 7
.,. . . . . . . . . . . . .
67 25 4 4
59
. . . . . .
12 100
18A F,H Form 12 100
12 100
12 100
12 100
12 100
194 206
182 194 194 196 206 194 202 198 212 192 211 202 17.0 18.0- 1 5 . 5 1 5 . 7 15.5 29.5
210
4.2 0.72 3.5 11.8 64
4.8 0.76 4.0 12.8 86
78 9 11 2
86
4.6 0.49 4.1 21.0
78 53 30 1
12
2
7
5.2 0.97 4.2 10.5 25
5.1 0.90 4.2 11.5 91
78 15
76 18 6
7
5.1 0.91 4.2 11.1 4
. . . . . . . . . . . . . . .
a Decomposition occurred during distillation after the ethylene glycol cut, which prevented a complete analysis of R u n 15A. b Runs 16A and 22A were not analyzed because of small conversions. C Products boiling between 100" and 212' F.
Table R u n No. Catalyst Bed depth, inches Catalyst, ml. Temp., F.
23A
12 100
V.
244 12 100
196 198 209 205 Middle 209 206 Bottom Reactor pressure, Ib. sq. inch abs. 15.5 15.5 Mass space velocity, g./hr. X ml. Total 5.2 3.7 CzHaO 0.47 0.52 HzO 4.7 3.2 Mole ratio HzO/CzHaO 24.4 15.0 Conversion, 82 57 Yield, % Ethylene glycol 84 81 Diethylene glycol 9 11 7 8 Higher polymers Intermediatesa . . . . . .
TOP
a
Experimentut Data 25A 12
100
26.4
27.4 28A A, H Form
12
12 100
100
298 12 100
12
100
30A 12 100
31.4 12
100
32.4
12 100
198 200 200 195 199 199 202 197 204 205 204 200 204 203 206 201 208 211 205 203 209 203 204 204 15.5 15.5 15.5 16.2 16.0 15.5 15.7 16.0 2.9 1.7 5.1 0.51 0.53 0.50 2.4 1.15 4 . 6 11.5 5.3 22.4 54 54 7a
10.7 0.97 9.7 24.1 49
8.0 5.3 1.0 1.04 7.0 4.3 1 6 . 5 10.1 33 31
3.2 2.9 1 . 0 4 0.94 2.1 2.0 5 1 5.2 35 30
81 11 8
84
85 8 7
65 15 20
63 20 17
82 9 9
8
8
79 11 10
69 11 20
h y d r o x y l f o r m . All the strong cation and anion eschange materials are found t o catalyze the reaction; however, the weakly acidic resins fail. The reaction is catalyzed in either the vapor phase or the heterogeneous system of vapor and liquid phases; o n l y t h e h e t e r o geneous system gives satisfactory yields. I n determining the yield the mole ratio of water t o ethylene oxide is the most significant factor. A mole ratio of 10 t o 1 is necessary t o obtain yields of 80y0 and an even higher ratio is required t o obtain a maximum of about 85%. The temperature and corresponding pressure greatly influence the reaction rate and t h e percentage yield. A t e m p e r a t u r e of 200' F. is necessary to obtain a substantial reaction rate, and the rate increases rapidly up t o 240OF. The percentage yield drops slowly wit,h increasing temperature in the range between 180" and 250" F., and above 300' F. drops rapidly with increasing temperatures. I o n e x c h a n g e materials have made available a catalyst which does not cont,aminate the products and is n o t c o n s u m e d i n t h e course of the reaction. These catalysts should find extcnsive use in the product,iori of ethylene glycol.
literature cited (1) Cartmell, R. R., Galloway,
J. R., Olsen, R. W., and Smith, J. M., ISD. EA-G. CHEM.40,389 (1948). (2) Curme, G. O., U. S. Patent 1,422,184 (1922). (3)
C . 0.. Ibid., 1,456,916
. . . . . . . . . . . . . . . . . . . . . . . .
Products boiling between 100' and 212' F.
Curme, G. O., and Young, (1922).
(4)
Davis, P. C., von Waaden, C. E., and Kurata, F , Chem. Eng. Prog. Symposium Ser. KO.4, 48,
not evident in the data. After this initial reaction takes place, the catalyst can be safely used for either vapor phase or heterogeneous phase operation. No further change in catalyst activity was observed.
91-7 (1952). (5) Hibben, J. H., IND. ENG.C I ~ E M 46, 1123-30 (1954). ( 6 ) Larson, A. T., U. S. Patent 2,152,064 (1939). (7) (8) (9)
D. J., Ibid.p 2t2859448 (1942).
McBee, E. T., H a w H. B.. a n d Wiseman, P. A., IND.ENG.C m x . 37, 432 (1945). Todd, F., IND. ENG.C H m r . , ANAL.ED.17, 175 (1945).
Summary a n d conclusions The between Oxide and water is catalyzed by ion exchange materials in either the hydrogen or
208
RECEIVED for review September 9, 1955. ACCEPTEDDecember 5, 1955. Research based on work for ri doctoral thesis submitted by L. M. Reed to Lehigh University.
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 48.No. 2