Ind. Eng. Chem. Res. 1992,31,469-474 Bach, H. C. Polymerization of Primary Aromatic Diamines to Azopolymers by Oxidative Coupling. Adv. Chem. Ser. 1968, 91, 679-691. Barron, C. H.; OHern, H. A. Reaction Kinetics of Sodium Sulfite Oxidation by the Rapid-Mixing Method. Chem. Eng. Sci. 1966, 21, 397. Cha, J. A.; Berry, K. B.; Lim, P. K. Aerobic Coupling of Aqueous Phenols Catalyzed by Binuclear Copper: Ring Substituent Effect and the Kinetics of the Coupling of o-Methylphenol. AICHE J. 1986,32,477-485. Chao, Taina H.; March, Juan. A Study of Polypyrrole Synthesized with Oxidative Transition Metal Ions. J. Polym. Sci., Part A: Polym. Chem. 1988,26,743-753. Cornils, B.; Falbe, J. Hydroformylationwith Water-Soluble Rhodium Catalysis. In Fundamental Research in Homogeneous Catalysis; Shilov, A. E., Ed.; Gordon and Breach Science Publishers: New York, 1984; Vol. 1,pp 487-508. Dumas, T.; Bulani, W. Oxidation ofPetrochemicak Chemistry and Technology; Wiley: New York, 1974; pp 29, 39-46. Endres, G. F.; Kwiatek, J. Polymerization by Oxidative Coupling. 111. Mechanistic .Type in the Copper-Amine Catalyzed Polymerization of 2,6-Dimethylphenol. J. Polym. Sci. 1962,58, 593-609. Endres, G. F.; Hay, A. S.; Eustance, J. W. Polymerization by Oxidative Coupling. V. Catalytic Specificity in the Copper-AmineCatalyzed Oxidation of 2,6-Dimethylphenol. J. Org. Chem. 1963, 28,1300-1305. Fendler, J. H.; Fendler, E. J. Principles of Micellar Catalysis in Aqueous Solutions. Catalysis in Micellar and Macromolecular Systems; Academic Press: New York, 1975; pp 86-103. Fritz, J. J. Solubility of Cuprous Chloride in Various Soluble Aqueous Chlorides. J. Chem. Eng. Data 1982,27, 188-193. Ganesan, S.; Lloyd, P. B.; Gilleskie, G. L.; Lim, P. K. Segragation and Chemical Conversion at Oil-Water Interface. Chem. Eng. Sci. 1989,44, 171-177. Giles, D. W.; Cha, J. A.; Lim, P. K. The Aerobic and Peroxide-Induced Coupling of Aqueous Thiols. I. Kinetic Results and Engineering Significance. Chem. Eng. Sci. 1986,41, 3129-3140. Hay, A. S. Polymerization by Oxidative Coupling. 11. Oxidation of 2,6-Disubstituted Phenols. J. Polym. Sci. 1962, 58, 581-591. Hay, A. S.; Blanchard, H. S.; Endres, G. F.; Eustance, J. W. Polymerization by Oxidative Coupling. J. Am. Chem. SOC.1959, 81, 6335-6336. Hay, A. S.; Shenian, P.; Gowan, A. L.; Erhardt, P. F.; Haat, W. R.; Theberge, J. E. Phenols, Oxidative Polymerization. Encyclopedia of Polymer Science and Technology; Wiley: New York, 1969; Vol.
469
10, pp 92-111. Henglein, A. Catalysis of Photochemical Reactions by Colloidal Semiconductors. Pure Appl. Chem. 1984,56, 1215-1224. Keim, W. The Impact of Transition Metal-Based Homogeneous Catalysis in Industrial Processes. In Fundamental Research in Homogeneous Catalysis; Grasiani, M., Giongo, M., Us.; Plenum Press: New York, 1984, pp 131-144. Kobayashi, M.; Chen, J.; Chung, T. C.; Moraes, F.; Heeger, A. J.; Wudl, F. Synthesis and Properties of Chemically Coupled Poly. (Thiophene). Synth. Met. 1984,9, 77-86. Lilly, M. D.; Woodley, J. M. Biocatalytic Reactions Involving Water-Insoluble Organic Compounds. In Biocatalysts in Organic Synthesis; Tramper, J., van der Plas, H. C., Linko, P., Eds.; Elsevier: New York, 1985; pp 179-192. Lloyd, P. B.; Ganesan, S.; Lim, P. K. Segregation of Metals at the Oil-Water Interface: Results and Implications. Ind. Eng. Chem. Res. 1989,28,577-584. Sittig, M. Combine Hydrocarbons and Oxygen for Profits-1968, Noyes Development Corp.: Park Ridge, NJ, 1968; pp 75-79. Starks, C. M., Ed.; Phase-Transfer Catalysis: New Chemistry, Catalysts, and Applications; ACS Symposium Series 326, American Chemical Society: Washington, DC, 1987; pp 1-14,96-127. Starks, C. M.; Liotta, C. Introduction. Phase-Transfer Catalysis: Principles and Techniques;Academic Press: New York, 1978; pp 1-12. Talsi, E. P.; Shaikhutdinova, N. I.; Shubin, A. A.; Chinakov, V. D.; Khlebnikov, B. M.; Yudkin, B. I.; Nekipelov, V. M.; k a r a e v , K. I. Relation between Catalyst Structure and Selectivity of the Oxidative Coupling of 2,6-Dimethylphenol in the Presence of Copper Complexes: Studies Usina- EPR and NMR. J.Mol. Catal. 1990,57,326351. Tsuchida, Eishun; Kaneko, Masao; Nishide, Hiroyuki. The Kinetics of the Oxidative Polymerization of 2.6-Xvlenol with a Comer* * Amine Complex. Makromol. Chem. 1972; 151, 221-234. White, Dwain M. I3C NMR Spectra of Poly(2,6-Dimethyl-l,4Phenylene Ether)s and Related Compounds. Polym. Prepr. (Am. Chem. SOC.,Div. Polym. Chem.) 1972,13, 373-378. Zhong, Y.; Rodberg, J. A.; Lim, P. K. ESR Asymmetric Line Broadening and Intramolecular Electron Exchange Involving a Transition Metal-RadicalComplex. 1. Theoretical Considerations and Simulation and Experimental Evidence. J. Phys. Chem., 1992, submitted for publication.
Received for review July 2, 1991 Accepted October 17, 1991
Promotional Effects of Potassium on Pd/A1203Selective Hydrogenation Cata1ysts Yeung H. Park and Geoffrey L. Price* Department of Chemical Engineering, Louisiana State University, Baton Rouge, Louisiana 70803
The catalytic behavior of K-doped Pd/A1203 catalysts for selective acetylene hydrogenation was investigated using steady-state-reaction, temperature programmed reaction (TPR),and deuterium tracer techniques. Catalysts were prepared by introducing K both before and after Pd impregnation, and both types of catalysts displayed similar traita. K addition enhanced the partial hydrogenation of acetylene to ethylene but increased oligomer formation. An increase in the rate of acetylene conversion was also observed upon the addition of K. K addition shifts T P R products to lower temperatures, which indicates that desorption of the surface species is enhanced. Ethylene deuteration results showed that the probability of ethylene desorption from the catalyst increased with increasing levels of K. A decrease in the rate of ethylene deuteration with K addition was observed, which contrasts with the enhanced rate of acetylene hydrogenation. All of these effects may be explained by K-inducedreduction in hydrocarbon adsorption strength. The effect of K arises through metal-support interactions. from ethylene feedstocks. Catalysts with higher selectiv-
Introduction
* To whom correspondence should be addressed.
Palczewska et al., 1984) and alloying of Pd with group 1B transition metals including Ag (Bond et al., 1958), Cu
0888-588519212631-0469$03.00/00 1992 American Chemical Society
470 Ind. Eng. Chem. Res., Vol. 31, No. 2, 1992
(Schay et al., 1983; Leviness et al., 1984; Weiss et al., 1984), and Au (Visser et al., 1974) have been reported. An enhancement in the selectivity and reaction rate with addition of group 1B transition metals has been observed and is thought to originate from the donation of electrons from these metals to Pd (Boitiaux et al., 1985a). Alkali metal cations, also potential electron donors, might yield similar effects; they have been shown to favorably alter the activity and selectivity of transition metal catalysts for CO hydrogenation and ammonia synthesis (Mross, 1983). The effect of alkali metals on these reactions is known to be due to electronic factors. The adsorption of CO and nitrogen on the metal becomes stronger while the adsorption of hydrogen and ammonia becomes weaker (Moss, 1983; Dry et al., 1969; Frankenburg, 1955). In a previous paper (Park and Price,1991a), we reported global rate and selectivity effects of K on the selective hydrogenation of trace C2H2in C2Hk In this paper, we report experiments aimed at understanding the effects of the K promoter on the catalytic behavior of Pd/Al203. The techniques used to study K effects were temperature programmed reaction (TPR) and ethylene deuterium tracer experiments. Background work on the applicability of TPR for this system has been reported (Park and Price, 19914. The reaction of ethylene with deuterium, for which a quantitative assessment of product composition is possible by Kemball’s method (Kemball, 1956; Kemball and Wells, 1968), is known to effectivelymeasure the relative surface populations of adsorbed species (Briggs et al., 1980).
Experimental Section Materials. A 1%Pd/A1203 catalyst was prepared by impregnation of Pd(NH3)4(N03)3on y-alumina (Linde 503). The Pd salt was dissolved in deionized water in an amount equivalent to the pore volume of alumina and was added to the dried alumina while being stirred. The catalyst was then dried at 393 K overnight and reduced in H2 at 773 K for 1 h. This catalyst will be designated “Pd/Al2O3”. One series of doped catalysts was prepared by adding 1and 4 wt % K as potassium carbonate to this prereduced catalyst by impregnation with an aqueous solution using the same procedure as Pd loading. This series of catalysts was then dried at 393 K and will be designated “K-Pd/A1203”. Another series was prepared by pretreating alumina with 1,2, and 4 w t % K in a procedure similar to that used by Figoli and L’argeniere (1989). Dry alumina was impregnated with K&03 solution and then dried and calcined at 873 K for 7.5 h. Pd was loaded on this treated support by impregnation, and the catalyst was dried before storage. Reduction was subsequently accomplished as a pretreatment procedure before any of the catalytic testing procedures. These materials are deeignated “Pd/K-A1203”. The dispersion of all catalysts was determined by chemisorption of carbon monoxide in a static adsorption apparatus using the difference between strongly and weakly bound CO. The BET surface areas were measured by an Omnisorp 360C adsorption apparatus. Hydrogen (99.999%) and helium (99.999%) from the gas cylinders were passed through a Supelpure (Supelco) purifier to remove trace oxygen. Acetylene (99.6%) was purified by bulb-to-bulb distillation using liquid nitrogen. Ethylene and carbon monoxide gases (CP grade) were used directly from the lecture bottle. Apparatus and Procedures. The apparatus and procedures used for the present study have been described previously (Park and Price, 1991b,c). Many experiments on each type of catalyst have been run in an effort to determine conditions which give optimum and reprodu-
cible results. In a steady-state-reaction experiment, 100 mg of catalyst was loaded into a Pyrex U-tube reactor, reduced in an atmospheric H2 flow at 773 K for 2 h, and purged with atmospheric He for 30 rnin at the same temperature. In the batch recirculation reactor 5.33 kPa of acetylene and 5.33 kPa of H2 were premixed and reacted at 298 K. The reaction mixture was sampled periodically and analyzed by GC. The GC column (3-m Durapak) was immersed in an ice-salt bath (258 K) for separation of C2 products at the initial stage of analysis and then in a water bath (298 K) for C4 product analysis. The thermal conductivity detector was calibrated using standard samples of known composition. For deuterium-exchangereactions, 2.66 kPa of ethylene and 2.66 kPa of D2were reacted in the batch recirculation reactor at 223 K over 10 mg of catalyst using the same pretreatment procedures as described above. Samples of the reaction mixture were injected into a GC column (2-m Porapak NS) for separation followed by analysis on a UTI lOOC mass spectrometer. Mass spectra were obtained with an electron impact source operated at 70 eV. The deuterium distribution of the ethylene reactant and the ethane product were determined by comparing the raw mass spectral intensities with experimental fragmentation patterns for each of the isotopic species (Park and Price, 1991b) . In a typical TPR experiment, 50 mg of 200-325-mesh catalyst particles were packed in a quartz tube reactor and reduced at 773 K for 1h in the recirculation system filled with 53.2 kPa of H2 and 53.2 kPa of He with a liquid nitrogen trap in the circulation loop. After reduction, the catalyst was flushed with He at 773 K for 30 min to remove adsorbed H2, and it was then cooled to the adsorption temperature (223 K) in He flow. A mixture of 73.2 Pa of acetylene in 106.4 kPa of He was circulated over the catalyst at 223 K for 15 min, and then the apparatus was switched to TPR mode and the catalyst was flushed with He for 30 min to remove weakly bound acetylene and weakly bound self-hydrogenation products. The mass spectrometer could be used to monitor this flushing process if desired. The catalyst was then further cooled to 173 K, a flow of 20 cm3/min H2 and 180 cm3/min He was established, and a linear temperature program at 5 K/min was initiated. Products were sampled by mass spectrometry every 20 s. The mas8 spectra of TPR products, which consisted of a mixture of hydrocarbons, were converted into the individual component spectra by a calculational method which has been described in a previous report (Park and Price, 1991~). Ammonia and acetylene TPD of the catalyst support were also used as characterization tools. In these experiments, 50 mg of A1203 or K-treated A1203 was treated with He at 773 K for 1h and then cooled to 298 K. The sample was exposed for 15 min to a mixture of 0.67 kPa of ammonia or 1.33 kPa of acetylene in 106.4 kPa of He circulating in the batch system and flushed with He for 1h to remove weakly adsorbed species. The sample was then heated at 10 K/min, and the desorbing product was monitored by mass spectrometry.
Results Physical properties of the catalysts are given in Table I, and temperature programmed desorption (TPD) spectra of ammonia and acetylene are given in Figures 1 and 2, respectively. The BET surface area of the alumina decreased only slightly with K addition. Pd dispersion was difficult to control from batch to batch as can be seen in two separate preparations shown in Table I. These two preparations will be referred to separately by their degree
Ind. Eng. Chem. Res., Vol. 31, No. 2, 1992 471 2.50
1.25
- - 4%K-A1203
I
7\
- - 4%K-A1203
2%K-A1203
c 0
E 0.50
.& E
v
S I rd 0501
\
m 0.25
e:
0 00
!
0.W
300
400
500
600
100
800
250
Temp (K) Figure 1. TPD spectra of NH3 adsorbed on K-doped aluminas at 298 K. Table I. Physical ProDerties of the Catalysts surface NH3 dispersion," area,b capacity: catalyst % m2/g umollg 215.6 292.9 Pd/Alz03c 36.2 292.9 16.7 215.6 Pd/A1203d 215.6 292.9 1%K-Pd/Al203 33.5 215.6 292.9 4% K-Pd/AlZ03 26.1 257.9 189.8 P d / l % K-AlzO3 17.2 186.5 233.7 Pd/2% K-A1203 32.5 80.8 23.2 182.7 Pd/4% K-Alz03
CzHz capacity,b rmol/g 22.9 22.9 22.9 22.9 18.8 17.5 81.2
"Based upon CO/Pd = 1/1. bProperties before Pd impregnation. cFirst batch. dSecond batch.
of dispersion. The acidity of the support, measured by ammonia TPD, decreased as expected with K addition so that 72% of the total acid sites were removed at 4% K content, but as can be seen in Figure 1, the high-temperature NH3 shoulder (indicating stronger acid sites) is especially strongly attenuated. K had little effect on acetylene adsorption except at the 4% level where an increase in capacity of about 4 times was observed. The increased CzHzcapacity is also noted as a large yield of Cz species in the TPR experiments (described below). In the TPD experiments, the acetylene peak temperature did not shift significantly as K was added to A1203(Figure 2). The results for the acetylene hydrogenation reaction over K-doped Catalysts are given in Table 11. Reaction curves were consistent with c w e s reported by others (Bond and Wells, 1966) which begin with a slow zero-order region that extends until almost all the acetylene has been depleted.
Temp (K) Figure 2. TPD spectra of C2Hzadsorbed on K-doped aluminas at 298 K.
Product selectivitiesare virtually constant with conversion in this region, so we have elected to report product compositions at 70% acetylene conversion (which is still within the zero-order region in all cases) and we have computed a global rate from the slope of the zero order line which is also reported in Table 11. From the table, we immediately see that reversing the order of K doping has very little effect on the catalytic properties. However, global effects of K are apparent. First, K improves the rate of acetylene conversion, and second, the ratio of ethylene to ethane was generally improved for all levels of K content. These observations are consistent with a previous report which studied global rates and selectivities for K-promoted materials using trace CzHzin C,H, (Park and Price, 1991a). Yields of C4products also increased with K addition, but this increase was largely due to an increase in the yield of 1B-butadiene. Because the increase in C4yield offsets the improved selectivity of acetylene to ethylene, the overall ethylene selectivity is relatively constant. Differences in the Pd dispersion (as determined by several separate observations of the potassium-free materials with disparate levels of Pd dispersion) between the catalysts has only a slight effect compared to the effect of K. The TPR spectrum of acetylene over Pd/AlZO3is given in Figure 3, and peaks for ethylene, ethane, n-butane, n-hexane, benzene, and cyclohexane are evident. The spectrum was found to depend heavily upon the acetylene adsorption conditions, so conditions from a previous study (Park and Price, 1991~)were employed to facilitate interpretation of the spectra. In the previous study, we
Table 11. Effect of K on Acetylene Hydrogenation Product Distribution and Reaction Rate Pd/ 1% Pd/2% Pd/4% 1% catalyst Pd/A1203 K-A1203 K-Alz03 K-A1203 K-Pd/Alz03 36.2 16.7 17.2 32.5 23.2 33.5 dispersion, % 5.94 7.75 9.86 5.87 rate" 5.87 4.68 ~
ethane ethylene n-butane 1-butene trans-2-butene cis-2-butene butadiene
4.5
Product 10.7 78.4 0.5 3.7 1.7 1.3 3.7
CzH4/Czb CzH4 (global)" C4 (globalId
91.8 72.6 20.9
88.0 70.7 19.7
7.2 81.1 0.3 4.0 1.8 1.1
Yc,h/(Yc,
93.2 73.4 21.3
Selectivities 93.6 72.0 23.0
92.2 68.0 26.3
5.8 81.2 0.5 3.5 1.9 1.4 5.7
3.6 82.7 0.0 3.7 1.5 1.3 7.3
93.3 71.9 23.0
95.8 72.6 24.2
for the zero-order initial region. bDefied as 100 X Yc2h/(YQh + Yc ,where Y is yield. "Defined as 100 + ~Yc,), where Cz = C(al1 two-carbon products) and C4 = C(all four-carbon producd. dDefined as 2Yc,/(Yc2 + 2Yc,).
" Rate of reaction in pmol/min X
Distribution (mol % at 70% Conversion) 6.0 5.6 6.6 82.1 81.3 78.2 0.6 0.3 0.3 3.4 3.6 4.2 1.8 2.0 2.0 1.4 1.7 1.4 4.7 5.4 7.2
~
4% K-Pd/AlzOS 26.1 7.93
472 Ind. Eng. Chem. Res., Vol. 31, No. 2, 1992 Table 111. Product Yields" from Acetylene TPR (73.2 Pa of CzHz Adsorbed, 15-minExposure, 223 K, 50 mg of Catalyst) Pd/l% Pd/2% Pd/4% 1% 4% PdfA1203 K-Alz03 K-Al203 K-A1203 K-PdfA1203 K-PdfAlZOB rmol T, K wmol T, K rmol T, K rmol T,K rrmol T, K rrmol T,K umol T. K acetylene 35.49 238 25.04 221 0.34 211 ethylene 3.77 208 2.04 206 8.48 234 1.06 206 0.81 223 ethane 37.86 221 8.61 204 13.27 213 32.66 209 63.86 241 14.63 208 72.09 241 n-butane 7.24 234 5.90 226 4.99 225 6.06 235 15.35 243 6.39 231 13.77 246 n-hexane 2.75 2.24 286 285 2.20 275 1.81 271 3.10 275 2.41 279 2.40 273 benzene 2.35 326 1.83 327 7.79 309 9.00 316 7.08 321 6.37 318 3.94 323 1.04b 628 l.Ogb 631 1.03b 647 cyclohexane 8.76 342 8.76 341 4.79 354 2.51 360 0.35 319 4.36 345 0.23 315 dispersion, % 36.2 16.7 17.2 32.5 23.2 33.5 26.1 ~~
"Yields based on 1g of catalyst. bBroad peak not finished by 673 K.
X
-E 0
1.50
100-
e .-u e
I
j l K y n e ,
,
~ 0 . 5 0 ~
,
,
,
,
Acetylene 0 00
150
250
350 450 550 Temp (K)
650
150
3 00
250
350 450 550 Temp (K)
650
3.00 I
8 2504 3
X
1
B n-Butane/3
;I
n-Hexane
0.00,
,
,
,
,
,
~
,
,
,
,
350 450 550 650 Temp (K) Figure 3. TPR spectra of CzH2adsorbed on Pd/Al2O3 (dispersion = 36.2%) at 223 K.
350 450 550 650 Temp (K) Figure 4. TPR spectra of CzHzadsorbed on 4% K-Pd/A1203 at 223 K.
determined that C, products were largely the result of physisorbed acetylene, which desorbs first in the TPR experiment and is hydrogenated on Pd sites, while oligomer bands largely represent the species which were originally adsorbed on Pd sites. The TPR spectra of acetylene over Pd/A1,03 and Kdoped catalysts are summarized in Table 111, and a typical spectrum for a K-doped catalyst (4% K-Pd/AlzO,) is given in Figure 4. Large changes in the peak area and temperature were observed for catalysts which contain K compared to Pd/AlZO3. Changes in the peak temperatures for oligomers are especially evident. Hexane shifted 8-14 K lower over K-doped catalysts compared to Pd/Al2O3 while peak areas remained similar on all catalysts. Butane displayed a similar trend except for catalysts with 4 % K where the yield was roughly twice as large. The peak temperatures for benzene from K-containing catalysts shifted to lower temperatures while the peak area increased significantly, but the cyclohexane yield was attenuated.
Over Pd/A1203,most of the C6cyclic component appears as cyclohexane while benzene is favored over K-doped materials, but it appears at lower temperatures. Changes in the TPR spectra in the C2 region were also significant but are leas meaningful than the oligomerization results since Cz products originate from the physisorbed acetylene as discussed above. Yields of Cz products follow the trends we expect from the acetylene capacity measurements on the support (Table I), where we observed a large capacity increase for the 4 % K support. Ethylene deuteration results in the form of Kemball's parameters (Kemball, 1956; Kemball and Wells, 1968) are given in Table IV. Over Pd/AlZO3catalysts at 223 K ethane-d, was dominant, which agrees with the results by Bond et al. (1966). Conversely, ethane-d, was dominant over K-doped catalysts, showing that H/D exchange which occurs along with deuteration has slowed relative to the rate of deuteration. Deuteration experiments were repeated at 263 K for K-doped catalysts to check for deu-
150
250
150
250
Ind. Eng. Chem. Res., Vol. 31, No. 2, 1992 473 Table IV. Kemball’s Parameters from Ethylene Deuteration catalyst Pd/A1203 (16.7% dispersion) Pd/4% K-Al203 Pd/A1203 (36.2% dispersion) 4% K-Pd/A1203
T,K 223 223 263 223 223 263
conversion/reaction time 16%/4 rnin 13%/67 rnin 15%/13 rnin 15%/10 rnin 15%/75 rnin 17%/20 rnin
(100 - p*): % 11 38 35 13 40 41
(loo - r*),b % 8 25 24 8 27 27
q*, s+,c %
32,37 55, 56 62,62 26,32 62, 62 64,64
a Equivalent to the probability of ethylene desorption. *Equivalent to the probability of ethyl hydrogenation to ethane. CEquivalentto the probability of D incorporation upon hydrogenation.
terium distribution changes at higher overall reaction rates, but the dominance of ethane-d2 was unchanged. These qualitative features are reflected quantitatively in the values of Kernball’s parameters given in Table IV. We have also observed a much lower rate of ethylene deuteration for K-doped catalysts than for the undoped catalysts at 223 K.
Discussion The major changes in the acetylene hydrogenation reaction which we observed upon addition of K to Pd/A1203 were the enhancement of the partial hydrogenation of acetylene to ethylene, an increase in the rate of acetylene conversion, and an increase in the C4 product selectivity. We also observed that K addition caused a shift in the oligomerization products from acetylene TPR to lower temperatures and noted an increase in the probability of ethylene desorption upon K addition during the deuteration of ethylene. The rate of ethylene deuteration (hydrogenation) was also reduced upon K addition. Several possible explanations for these effects exist, but the results of the ethylene deuterium exchange reaction, which show that ethylene desorption is favored upon the addition of K, suggest that the strength of hydrocarbon adsorption decreases with K content. Oligomer bands from acetylene TPR appear a t lower temperatures for K-promoted materials compared to Pd/A1203 which is also consistent with this postulate. The enhancement in the probability of ethylene desorption with the addition of K to Pd/A1203could also be explained by a reduction in the availability of surface hydrogen. If hydrogen availability is limited by K, the probability of the hydrogenation of adsorbed ethyl radicals to ethane would be suppressed, resulting in a greater probability that ethylene desorbs rather than reacts. However, our results show that the probability of ethyl hydrogenation also increases with K content and, therefore, this postulate is unlikely. We might also attribute greater benzene/cyclohexane ratios from acetylene TPR for Kcontaining catalysts to a limited hydrogen supply, but this would be inconsistent with the ethylene deuteration results. Also, Rieck and Bell (1986) have reported that alkali metal promotion of Pd/Si02 does not significantly alter the amount of hydrogen adsorption. The higher benzene/cyclohexane ratios fall into the more likely postulate that the strength of hydrocarbon adsorption is reduced with K content. Benzene desorption would therefore be favored over hydrogenation to cyclohexane. The strong similarities of K-Pd/A1203 and Pd/K-A1203 suggest that the order of K addition has the same effect on the catalytic properties, and it appears that the effect is related to K-induced changes in the properties of the support which indirectly affects the Pd probably through metal-support interactions. Though the surface area of A1203 was reduced 10-15% by the addition of K, transport property effects cannot completely explain our TPR results. If oligomers were eluted from K-containing materials at lower temperatures because of transport effects, all
oligomers should shift to lower temperatures, and this was not generally the case for cyclohexane. The improved rate of acetylene hydrogenation and the contrasting decrease in the rate of ethylene hydrogenation caused by the addition of K may also be explained by a reduced strength of hydrocarbon adsorption. Acetylene dominates the Pd surface at the expense of hydrogen so that acetylene hydrogenation is slowed by the lack of hydrogen. If K reduces the strength of acetylene adsorption, hydrogen could compete more effectively for sites and a net increase in rate is expected. For the ethylene hydrogenation reaction, however, ethylene does not completely dominate the Pd surface so that a reduction in ethylene surface coverage should reduce the hydrogenation rate as we have observed. The reduced strength of adsorption of hydrocarbons may be caused by an increased electron density on Pd particles. Pd would then have less propensity to share electrons with adsorbate hydrocarbons. The addition of piperidine (an electron donor) to the reaction mixture during the hydrogenation of l-butyne over Pd/A1203resulted in an increase in both selectivity to l-butene and overall hydrogenation activity (Boitiaux et al., 1985b), which is similar to our observations on the effect of K on acetylene hydrogenation. Electronic effects on Pd can take place either through direct K-Pd interaction or through a modification of the support properties which induces a change in metal-support interactions. Our acetylene and NH3 TPD spectra of AZO3and K-A1203 show that support properties change from acidic to basic upon K addition as expected. Acetylene is known to be slightly acidic (Yates and Lucchesi, 1961; Morrison and Boyd, 1983) so that the dramatically higher capacity of 4% K-A120, for acetylene adsorption indicates a strong shift toward a basic support. Electron transfer between metals and acidic supports has been suggested by several researchers. Figoli and L’argeniere (1989) studied selective hydrogenation of styrene over Pd/AI2O3catalysts and found changes in the electronic state of Pd via X-ray photoelectron spectroscopy when sodium was added to the support. Free electrons on Pd were transferred to lower electron density sites on acidic A1203and electron transfer was attenuated with added sodium. Figueras et al. (1973) studied benzene hydrogenation over Pd supported on several different acidic oxides and found that the support influences the electronic properties of Pd. Similar theories have been reported by Gallezot et al. (1977) and Anderson (1975). In our case, we suggest by analogy that the more basic nature of Kdoped A1203 results in an electron transfer to Pd, resulting in a higher Pd electron density, and ultimately, a reduced strength of hydrocarbon adsorption. This leads to an increase in C2H4 selectivity from acetylene because ethylene desorbs more readily before it can be further hydrogenated to ethane and because the ethylene product is less likely to readsorb and become hydrogenated. One last point concerning the production of C4oligomers deserves discussion under this theory. Other researchers
474 Ind. Eng. Chem. Res., Vol. 31, No. 2, 1992
have suggested that acidic support sites contribute to oligomerization reactions (Trimm,1980; Chauda and Gosch, 1969; Kranich et al., 1985), but we have observed an increase in C4 oligomers when the support is made more basic via the addition of K, which we attribute to a lower strength of hydrocarbon adsorption on Pd/A1203modified by K. These theories are not necessarily mutually exclusive. If some oligomers are formed at acidic support sites and the addition of K reduces the number of these sites, the reduced strength of hydrocarbon adsorption on Pd could possibly overcompensate for this effect resulting in a net increase in oligomers. The observation that the activity and selectively of Pd/A1203for acetylene hydrogenation are improved via the addition of K suggests that these materials should be further tested as candidates for trace acetylene removal from ethylene streams. Activity, selectivity, and aging characteristics need further examination under more realistic industrial conditions. If our observations can be extrapolated to industrial conditions, the improved ethylene selectivity from acetylene related to K addition would need to offset the simultaneous K-induced increase in oligomerization products for these materials to be a significant improvement in current technology, and it is likely that this requirement can be realized. It is known that ethylene hydrogenation proceeds independent of acetylene hydrogenation in the industrial situation where trace acetylene in an ethylene feedstock is hydrogenated (Moses et al., 1984). K-modified Pd/A1203should reduce the rate of direct ethylene hydrogenation and improve the rate of acetylene removal while the expected increase in oligomers, which comes from the trace acetylene component in the feedstock, should be trivial. We can also postulate that K-modified Pd/A1203will be less susceptible to aging since desorption of low molecular weight oligomers may result in an attenuated accumulation of high molecular weight oligomers that foul the catalysts. Acknowledgment We appreciate the financial support of the Exxon Foundation. Literature Cited Anderson, J. R. Structure of Metallic Catalysts; Academic Press: London, 1975; p 278. Boitiaux. J. P.: Cosvns. J.: Derrien. M.: Leaar. G. Newest Hvdrogenation C a k y s k . Hydrocarbon Process. 1985a, March, 51. Boitiaux, J. P.; Cosyns, J.; Vasudevan, S. Hydrogenation of Highly Unsaturated Hydrocarbons over Highly Dispersed Palladium Catalyst. Part I 1 Ligand Effect of Piperidine. Appl. Catal. 1985b, 15 (2), 317. Bond, G. C.; Wells, P. B. The Hydrogenation of Acetylene. 11. The Reaction of Acetylene by Alumina-Supported Palladium. J. Catal. 1966, 5, 65. Bond, G. C.; Dowden, D. A.; Mackenzie, N. Selective Hydrogenation of Acetylene. Trans. Faraday SOC.1958,54, 1537. Bond, G. C.; Philipson, J. J.; Wells, P. B.; Winterbottom, J. M. Hydrogenation of Olefins Part 3. Reaction of Ethylene and Propylene with Deuterium over Alumina-Supported Palladium and Rhodium. Trans. Faraday SOC.1966,62,443. Briggs, D.; Dewing, J.; Burden, A. G.; Moyes, R. B.; Wells, P. B. Support Effects in Ethene Hydrogenation by Platinum. J.Catal. 1980, 65 (l),31. Chauda, M.; Gosch, S. S. Hydropolymerization of Acetylene over Palladium catalysts in Fixed Bed Part I Effect of Different Variables on the Pattern of Acetylene Conversion. J.Indian Inst. Sci. 1969, 51, 180. Dry, M. E.; Shingles, T.; Boschoff, L. J.; Oosthuizen Heat of Chemisorption on Promoted Iron Surfaces and the Role of Alkali in Fisher-Tropsch Synthesis. J. Catal. 1969, 15, 190.
Figoli, N. S.; L'argeniere, P. C. Selective Hydrogenation Catalysts. Influence of the Support on the Sulphur Resistance. Catal. Today 1989,5,403. Figueras, F.; Gomez, R.; Primet, M. Adsorption and Catalytic Properties of Palladium Supported by Silica, Alumina, Magnesia, and Armorphous and Crystalline Silica-Aluminas. Adu. Chem. Ser. 1973, 121, 480. Frankenburg, W. G. Catalytic Synthesis of Ammonia. In Catalysis; Emmett, P. H., Ed.; Reinhold New York, 1955; Vol. 111,Chapter 6, p 237. Gallezot, P.; Datka, J.; Massadir, J.; Primet, M.; Imelik, B. Unusual Catalytic Behavior of Very Small Platinum Particle Encaged in Y Zeolites. Proceedings of the Sixth International Congress on Catalysis, London (1976); Bond, G. C., et al., Eds.; Chemical Society: London, 1977; Vol. 2, p 696. Kemball, C. The Deuterium and Exchange of Ethylene on Evapo1956, ration Metal Catalysts at Low Temperature. J. Chem. SOC. 735. Kemball, C.; Wells, P. B. A Revised Theory of Calculating Distribution of Products from the Catalytic Reaction of Ethylene with Deuterium. J. Chem. SOC.( A ) 1968,444. Kranich, W. L.; Weiss, A. H.; Schay, Z.; Guczi, L. Acetylene Hydrogenation Using Palladium Zeolite Catalysts. Appl. Catal. 1985, 13, 257. Leviness, S.; Nair, V.; Weiss, A. H.; Schay, Z.; Guczi, L. Acetylene Hydrogenation Selectivity Control on Pd Cu/A1203Catalysts. J. Mol. Catal. 1984,25, 131. Lindlar, H.; Dubuis, R. Palladium Catalyst for Partial Reduction of Acetylenes. Org. Synth. 1966, 46, 89. Miller, S. A. Acetylene: Its Properties, Manufacture and Uses; Academic Press: New York, 1966; Vol. 2, p 1. Morrison, R. T.; Boyd, R. N. Organic Chemistry, 4th ed.; Allyn and Bacon: Boston, 1983. Moses, J. M.; Weiss, A. H.; Matusek, K.; Guczi, L. The Effect of Catalysts Treatment on the Selective Hydrogenation of Acetylene over Palladium/Alumina. J. Catal. 1984,86, 417. Mross, W. D. Alkali Doping in Heterogeneous Catalysis. Catal. Rev. Sci. Eng. 1983, 25 (4), 591. Palczewska, W.; Ratajczykowa, I.; Szymerska,I.; Krawczyk, M. Lead and Carbon Monoxide as Additives Modifying the Selectivity of Palladium Catalysts in Partial Hydrogenation of Acetylene. Proceedings of the Eigth International Congress on Catalysis; Verlag Chemie: Weinheim, 1984; Vol. IV, p 173. Park, Y. H.; Price, G. L. Postassium Promotor for Palladium on Alumina Selective Hydrogenation Catalysts. J. Chem. Soc., Chem. Commun. 1991a, 17, 1188. Park, Y. H.; Price, G. L. Deuterium Tracer Study on the Effect of Carbon Monoxide on the Selective Hydrogenation of Acetylene over Pd/Al2O8. Znd. Eng. Chem. Res. LSSLb, 30, 1693. Park, Y. H.; Price, G. L. Temperature Programmed Reaction Study on the Effect of Carbon Monoxide on the Acetylene Reaction over Pd/A1203. Ind. Eng. Chem. Res. 1991c, 30, 1700. Peterson, J. R. Hydrogenation Catalysts; Noyes-Data Corp.: Park Ridges, NJ, 1977; p 183. Rieck, J. S.; Bell, A. T. Studies of the Interactions of H2and CO with Pd/Si02 Promoted with Li, Na, K, Rb, and Cs. J. Catal. 1986, 100, 305. Schay, Z.; Sarkany, A.; Guczi, L.; Weiss, A. H.; Nair, V. Selective Hydrogenation of Trace Acetylene in Ethylene over PalladiumCopper/Alumina Catalyst. Proceeding, Fifth International Symposium on Heterogeneous Catalysis; Elsevier: New York, 1983; Vol 1, p 315. Trimm, D. L. Design of Industrial Catalysts;Elsevier: Amsterdam, 1980; p 229. Visser, C.; Zuidwijk, J. G. P.; Ponec, V. Reactions of Hydrocarbons on Palladium-Gold Alloys. J. Catal. 1974, 35, 407. Weiss, A. H.; Leviness, S.; Nair, V.; Guczi, L.; Sarkany, A.; Schay, Z. The Effect of Pd Dispersion in Acetylene Selective Hydrogenation. Proceedings of the Eigth International Congress on Catalysis; Verlag Chemie: Weinheim, 1984; p 541. Yates, D. J. C.; Lucchesi, P. J. Infrared Spectra of Acetylene Derivatives Adsorbed on Alumina and Silica. J. Chem. Phys. 1961, 35, 243. Received for review June 5, 1991 Revised manuscript received October 16, 1991 Accepted November 19, 1991
