Applied Catalysis, 58 (1990) 227-239 227

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

Activity and Selectivity of Pd -AI203 for Ethyne

Hydrogenation in a Large Exces§-of-Ethene and
Hydrogen

H.R. ADIJRIZ, P. BODNARIUK, M. DENNEHY and C.E. GIGOLA*

Planta Piloto de Ingenieria Quimica, Plapiqui, 12 de Octubre 1842, 8000 Bahia Blanca
(Argentina)
(Received 20 July 1989, revised manuscript received 18 October 1989)

ABSTRACT

The hydrogenation of ethyne in the presence of large amounts of hydrogen and ethene was
studied at 1,520 kPa over Pc~/o~-A1203catalysts of varying dispersion. Both TON and selectivity
for ethane/(Sc2Hs vmoles df ethane produced from both ethyne and ethene per mole of ethyne
converted) were found to decrease as the palladium dispersion increased. The former result was
consistent with a clear dependence of the rate order for ethyne on the particle size of the palladium:
it was 0 for large particles and - 0 . 5 for small particles. This kinetic behaviour was ascribed to a
stronger interaction of the ethyne on the small palladium particles. By increasing the temperature
at constant space velocity, SC2H~ Was measured over a wide conversion range. The amount of
ethane formed was high oi~largeOarticles hydrogen-to-palladium ratio ( < 0.2 ) and negligible on
s~all particles. Using an ethene-free mixture, it was shawrr that the selectivity for ethene
~Sc2H,) is close to 100% up to high conversion. Moreover~ Sc2m d~es not depend on the particle
ize of the palladium. Consequently, the higl~ Sc2m~-----o~serv~d.don]o~v-dispersion catalysts was due
p thee parallel hydrogenation of ethene. Thes~s.result)/were int~ipreted by assuming that the pres-
ence of multiple adsorbed species on large palladium particles prevents the strong bonding of
ethyne and therefore the simultaneous adsorption and reaction of ethene is permitted.

INTRODUCTION

The selective hydrogenation of ethyne in the presence of large amounts of

ethene is a process of considerable importance for the manufacture of polymer-
grade ethene. From an industrial point of view, the main objective is the re-
duction of the ethyne concentration to less than 10 ppm without ethene hy-
drogenation. Two different hydrogenation processes are used for this purpose:
the front-end and the tail-end configuration. The former involves the hydro-
genation of the cracked gas mixture (with large amount of ethene and hydro-
gen) after caustic scrubbing to remove sulphur contaminants. In the tail-end
process the hydrogenation is carried out after the addition of a stoichiometric
amount of hydrogen as required by the C2H2--, C2H4 reaction. Commercial pal-

0166-9834/90/$03.50 © 1990 Elsevier Science Publishers B.V.

228

ladium catalysts with low metal loading and low dispersion are now used for
both processes but carbon monoxide is continuously added in trace amounts
to inhibit the undesirable hydrogenation of the alkenes [1,2]. The present
investigation was performed to gain a better understanding of both the activity
and the selectivity of ethyne hydrogenation on Pd/a-A1203 using a front-end
reaction mixture without carbon monoxide. If the selectivity of palladium is
determined only by the thermodynamic factor, we may simply expect to ob-
serve the conversion of ethyne, with no formation of ethane from ethene until
most of the ethyne has been eliminated. However, literature results concerning
the selectivity of palladium catalysts for the semi-hydrogenation of alkynes
seem to vary depending on particle size and feed composition. According to
Borodzirlski et al. [3], large particles favour the direct conversion of ethyne
into ethane owing to the presence of the fl-hydride phase.
In line with this observation, Carturan et al. [4] demonstrated that small
palladium particles are more selective for the half-hydrogenation of phenyl-
acetylene and it was concluded that this selectivity arises from the strong in-
teraction between the alkyne and the palladium phase. In the presence of dis-
solved hydrogen the comparable strengths of the metal-alkyne and metal-al-
kene bonds lead to a lower selectivity. However, they also proposed, as suggested
by Borodziriski et al. [3 ], that the direct conversion of alkynes to alkanes may
occur in the presence of the fl-hydride phase. Both studies were conducted with
alkyne-hydrogen mixtures at atmospheric pressure.
Using isotopic labelling, McGown et al. [5 ] showed that the ethane produced
during the hydrogenation of ethyne-ethene rich mixtures on commercial pal-
ladium catalysts is mainly formed from ethene. To explain this result, it was
postulated that two types of sites exist on the metal surface. A similar tech-
nique was used by Weiss and co-workers [6-8] to demonstrate that on palla-
dium black and Pd/A1203, ethyne is hydrogenatedto both ethene and to ethane,
whereas ethene hydrogenation occurs at low ethyne concentration on the same
active sites. It was also suggested that ethane formation from ethyne involves
alkylidene species at intermediates. Using a simulated tail-end mixture [8] at
60% ethyne conversion the selectivity for ethane was in the range 2-10%.
The effect of palladium dispersion on these reactions has also been investi-
gated [9]. The selectivity for ethane formation from ethyne was observed to
decrease by a factor of four whereas the dispersion increased from 2% to 60%.
The effect of dispersion on the parallel hydrogenation of ethene was found to
be similar. Again it was proposed that the dissociative adsorption of ethyne
plus the presence of the fl-hydride phase on large particles leads to ethane
formation. According to the authors, this mode of chemisorption does not block
the simultaneous adsorption and hydrogenation of ethene, as expected on the
basis of thermodynamic arguments.
Although the influence of the fl-palladium hydride phase on selectivity has
been invoked in several studies, the low hydrogen partial pressure in tail-end
 229

hydrogenation mixtures (1% hydrogen + ethyne balanced by ethene) seems to

exclude that possibility. The thermodynamics of the Pd-H system for bulk
palladium indicates that a minimum pressure of 40 Torr (1 Tort = 133.322 Pa)
is needed to form the fl-palladium hydride phase at 300 K [10]. When ethyne
is removed by the front-end hydrogenation process (H2/C2H2=50; H2/
C2H4= 2 ), the high hydrogen pressure supports the formation of the fl-hydride
phase. Under this condition the specific activity for ethyne hydrogenation was
found to decrease with increasing dispersion [2]. In contrast, the results of
S~irk~iny et al. [9] for tail-end hydrogenation demonstrated just the opposite
effect. Consequently, it was considered interesting to compare the selectivity
for ethane formation of these different reaction mixtures as a function of pal-
ladium particle size.
This paper mainly reports the selective hydrogenation of a simulated front-
end ethane cracker gas mixture on palladium catalysts of varying dispersion.
A basic kinetic study was also performed. In contrast to previous work, the
investigationwas conducted at high pressure in the absence of carbon monoxide.

EXPERIMENTAL

Pd/AI203 catalysts containing 0.058% and 0.09% of palladium were pre-

pared by wet impregnation of a-alumina (Rh5ne Poulenc SCS9, 13 m 2 g-l)
using palladium acetylacetonate dissolved in benzene as precursor. After 72 h
the liquid was removed and the preparations were dried at 383 K followed by
calcination for 1 h in air at 573 K. A portion of the 0.058% palladium catalyst
was reduced in a flow of hydrogen by raising the temperature at 5 K min- 1up
to 573 K and maintaining it for 1 h. The remaining part was sintered in hydro-
gen at 1073 K for 16 h. The 0.09% palladium catalyst was reduced at 773 K for
2 h and a very low dispersion sample was obtained owing to the higher metal
loading. A commercial Pd/a-Al203 catalyst (IC138-1 ) was also available. The
metal content in the reduced catalysts was determined by atomic absorption
spectrometry. The palladium dispersion was calculated from measurements
performed in a low-dead-volumeglass apparatus, following the back-sorption
technique proposed by Boudart and Wang [11]. The samples were first re-
duced at 573 K and then evacuated at the same. temperature for 12 h. After
measuring the sorption isotherms at room temperature, the samples were evac-
uated for 0.5 h at the same temperature and a second measurement provided
the amount of absorbed and weakly bound hydrogen. The extent of hydrogen
chemisorption was obtained by subtracting these contributions from the total
uptake. A summary of the catalysts prepared and their characterization is given
in Table 1. Several compressed reaction mixtures, listed in Table 2, were pre-
pared by mixing CP-grade hydrocarbons (methane, ethyne and ethene from
Matheson) with ultra-high-purity hydrogen and nitrogen (from AGA) in steel
cylinders. They were rolled for several days before use. Mixtures 1-6 were used
230

TABLE 1

Composition and hydrogen adsorption data for Pd/a-A1203

Catalyst P d (%) H/Pd

L 1-1 0.058 0.58

L 1-2 0.058 0.10
L 2 0.090 0.05
C 1 (ICI 38-1) 0.040 0.18

TABLE 2

Inlet composition of reaction mixtures (mol-%)

Mixture CH 4 C2H 2 C2H 4 H2 N2

1 6.37 0.49 42.14 51.00 -

2 8.99 0.64 60.36 30.00 -
3 10.92 0.84 72.24 16.00 -
4 11.71 0.90 77.48 9.90 -
5 11.30 0.98 64.70 23.10 -
6 9.20 0.19 68.40 22.20 -
7 - 0.85 - 15.33 83.80
8 - 0.85 4.00 17.33 77.80

as feed gases to obtain rate data at low conversion. The effect of ethyne con-
version on selectivity was studied mainly with mixture 3. In order to compare
our results with those of previous studies, ethene was excluded or added in
small amounts in mixtures 7 and 8, respectively. Although ethane was excluded
from all mixtures to facilitate the measurement of small amounts formed dur-
ing reaction, it was present as an impurity at a level of 0.013%.
The apparatus for the activity and selectivity measurements was essentially
the same as that employed previously [2 ], except that a two-stage flow reactor
with intermediate cooling was used to minimize thermal effects. Two SS Swa-
gelok unions (1/8 in.) holding catalyst charges of 0.01 g were connected with
2 m of 1/16-in. stainless-steel tubing and immersed in a diethylene glycol bath.
The feed mixtures and the reaction products were analysed using on-line gas
chromatography (GC). In order to test the selectivity of unsupported palla-
dium a special reactor, as shown in Fig. 1, was constructed. The reaction mix-
tures were passed over the external surface of a palladium tube (14 cm X 0.3
cm), sealed to a stainless-steel shell by Swagelok unions, while the internal
surface was exposed to pure nitrogen. Fresh samples were pretreated under the
reaction mixture for 4 h at 288 K and 1520 kPa. Following this pretreatment,
conversion measurements at constant space velocity were carried out at in-
creasing temperatures up to the point where the hydrogenation of ethene was
 231

tl

t
it
t t
It
II

',! s

tl

Fig. 1. Schematic diagram of catalytic wall reactor. 1 = Reactor inlet; 2 = reactor outlet; 3 = P d
tube; 4 = N2 filling; 5 = Swagelok fitting.

the predominant reaction. In order to obtain kinetic data under differential

conditions, the flow-rate was adjusted at selected temperatures. Steady-state
operation was maintained for 1 h before changing the operating conditions to
the next selected values.

RESULTS

The rate of ethyne hydrogenation was measured at low conversion (15%)

between 273 and 313 K. Linear Arrhenius plots were obtained, as shown in Fig.
2. An apparent activation energy of 16 _+2 kcal tool- 1 was obtained for all the
catalysts tested. Using these data and the information in Table 1, the specific
activity at 288 K was calculated and is presented in Table 3 as a function of
dispersion. A clear trend of decreasing specific activity with increasing disper-
sion is observed. As reported previously [2 ], it decreases by one order of mag-
nitude as the dispersion increases from 10% to 60%. Kinetic data obtained with
mixtures 1-6 for samples L 1-1, L 2 and C 1 were fitted to a power-law rate
equation. Using mixtures 5 and 6 the dependence on ethyne partial pressure
was found to be zero and - 0 . 5 for catalysts C 1 and L 1-1, respectively. The
main results are summarized in Table 4. Despite the five-fold change in ethyne
concentration, the TON did not vary for the low-dispersion sample. In con-
trast, when sample L 1-1 was tested the rate increased as the concentration
decreased, leading to a negative value of - 0.5. Consequently, the dependence
232

0
I-.-

k-1-1

0J I I I i
312 3.3 314 315 3.6

T-l,, 103 ( K -1 )
Fig. 2. Arrhenius plots for hydrogenation of ethyne on palladium catalysts with mixture 3.

TABLE 3

Kinetic parameters for ethyne hydrogenation on Pd/t~-A1203 as a function of palladium dispersion

Feed gas: mixture 3.

Catalyst H/Pd E (kcal mo1-1) T O N (s -1) at 288 K ma

L 1-1 0.58 17 0.28 1.6

C 1 0.18 16 1.1 1.3
L 1-2 0.10 17 4.7 -
L 2 0.05 14 7.3 1.3

aH 2 order from Fig. 3.

of TON on dispersion was less pronounced at low ethyne concentrations. Re-

garding the dependence on hydrogen partial pressure, linear plots of
log (activity) vs. log(pHi) were obtained with mixtures 1-4, as can be seen in
Fig. 3. The rate order for hydrogen was found to be between 1.3 and 1.6 for all
the samples tested.
The effect of increasing temperature on ethyne conversion and ethane for-
mation, at constant flow-rate, is presented in Fig. 4. There is a clear distinction
between low- and high-dispersion samples. For catalyst L 1-2 (H/Pd = 0.10)
 233

TABLE 4

Dependence of reaction rate on ethyne concentration

Feed gas: mixtures 5 and 6.

Catalyst C2H2 (mol-%) H2 (mol-%) TON (s- 1) at 288 K n

L 1-1 0.98 23.0 0.58

-0.5
L 1-1 0.19 22.2 1.42
C 1 0.98 23.0 2.70
0.0
C 1 0.19 22.2 2.58

,0 [;;:0.0o,
A

'E

A
% [::.:

I"1-

C
z
O
I--

10
PH2( at m )
Fig. 3. Dependence of reaction rate on hydrogen partial pressure. Effect of palladium dispersion.
Feed compositions: mixtures 1-4; T = 288 K.

the amount of ethane formed increased continuously up to a value of 0.2% at

80% conversion. In contrast, for the high-dispersion catalyst L 1-1 ( H /
Pd = 0.58) the ethane content was very low, less than 0.02 % up to 50% ethyne
conversion. It should be mentioned that small changes in ethene concentration
could not be followed owing to the large excess of this component in mixture
3. Consequently, only the selectivity for ethane (Scan6) was calculated from
the GC data. This selectivity is plotted versus conversion in Fig. 5 for all the
samples tested. There is a clear trend of decreasing SC2H6 with increasing dis-
persion. Selectivity values equal to or greater than 1 were obtained with un-
supported palladium, although in this instance the ethyne conversion was lim-
ited to 10-15%. At higher conversion values there was a marked decrease in
ethene concentration owing to hydrogenation, and isothermal conditions could
234

80 0 193 )
Z 0.053
o_
(./3 60 olol
n"
ILl

Z
0 40 o.o59

£ 20
0029
0.024
V/oo:'
0.014
I 0.014 I I

283 293 303 313

T(K)

Fig. 4. Ethyne conversion ( % ) vs. T ( K ) at constant space velocity [ 29 c m 3 ( S T P ) g - c a t - 1 s - 1].

The numbers denote the ethane content ( m o l - % ) . Initial value: 0.013 m o l - % . Feed composition:
mixture 3.

100 unsupported /
o" Pd
Pd = 0 . 0 5 )

J L 1-2
/
( H/Pd= 0.10 )

20 ~ P d =0.,~)

20 z,O 60 80 t00
C2 H2 CONVERSION (%)
Fig. 5. Ethane selectivity (%) vs. ethyne conversion (%). Effect of palladium dispersion. Feed
composition: mixture 3.

not be maintained. Low-dispersion catalysts L 1-2 and L 2 gave higher selec-

tivities and ethane formation was detected even at low ethyne conversion. On
the other hand, with the high-dispersion sample L 1-1, ethene was the main
product of ethyne hydrogenation.
In an attempt to determine whether the ethane was formed from ethyne or
ethene, similar experiments were conducted with mixtures 7 and 8 over unsup-
 235

TABLE 5

Effect of ethyne conversion on ethane and ethene selectivities over unsupported palladium
Data obtained by increasing both temperature and flow-rate.

Mixture 7 Mixture 8

C2H2 conversion (%) Sc2H6 SC~H4 C2H2 conversion (%) SC2H8 SC2H4

6.8 0.0 1.00 10.0 0.36 0.79

10.2 0.0 1.00 12.5 0.45 0.65
15.6 0.05 1.00 20.4 0.39 0.59
27.3 0.11 0.88 32.2 0.22 0.80
55.8 0.12 0.87 54.1 0.73 0.33
75.6 0.17 0.82 72.3 0.87 0.16
85.5 0.18 0.81 90.0 0.97 0.06

ported palladium (see Table 5). First the reaction was performed in the ab-
sence of ethene (mixture 7) while keeping the concentration of ethyne con-
stant at the same values as for mixture 3. Then the concentration of ethene
was increased to 4% (mixture 8). In these experiments the ethene selectivity
was also calculated as moles of ethene formed or consumed per mole of ethyne
converted.
Table 5 shows that ethane was not formed in the absence of ethene; initially
the reaction exhibited a selectivity of 100% towards ethene. When the ethyne
conversion was higher than 15%, small amounts of ethane were formed, lead-
ing to a parallel decrease in Sc2m. Similar results were obtained with samples
C 1 and L 1-1. On addition of 4% ethene, pronounced changes in these selec-
tivities were observed. Within experimental error, the initial value of Sc2H6
was fairly high and it increased continuously over the whole conversion range.
An opposite tendency was observed for SC2H4 but a net gain of ethene was
maintained.
On the basis of these observations, it is clear that ethyne hydrogenation
proceeds exclusively in the direction of ethene formation. Consequently, ethane
formation is mainly due to ethene hydrogenation and this reaction occurs only
on large palladium particles.

DISCUSSION

In agreement with previous data [2] the specific activity was found to de-
crease with increasing dispersion. This behaviour has already been observed
in the hydrogenation of several alkynes and dienes such as 1-butyne, 1,3-bu-
tadiene, isoprene, valylene (2-methyl-l-buten-3-yne) and phenylacetylene
[4,12,13]. The former reaction has also been studied on Pt/A1203 and Rh/
236

A1203 catalysts [14,15] with similar results. Consequently, the change in TON
with particle size cannot be related to the formation of the fl-hydride phase.
Boitiaux et al. [ 14 ] attributed this sensitivity to metal dispersion to a "multi-
complexation" mechanism involvinga metal atom of low coordination number
(small particles) and two hydrocarbon molecules. This interpretation was
consistent with the observed variations in reaction order with particle size: for
Pt/A1203 the apparent order with respect to 1-butyne varies between 0 and - 1
as the dispersion increases from 3% to 96%. Our kinetic data seem to follow
this pattern. The rate order for ethyne decreased from 0 to - 0.5 as the disper-
sion increased from 18% to 58% (Table 4). Therefore, at a relatively low ethyne
concentration (high conversion) the difference in TON between small and
large particles would not be significant. At high ethyne concentration small
particles could be deactivated owing to multiple adsorption of this hydrocar-
bon, as postulated by Boitiaux et al. [ 14 ]. Recently, on the basis of an X-ray
photoelectron spectroscopic study, Hub et al. [ 16 ] argued that small "electron-
defficient" palladium particles are responsible for this behaviour.
Regarding the influence of the hydrogen partial pressure, our results are also
consistent with those of Boitiaux et al. in the sense that the correspondent
order does not change with dispersion. However, the value of 1.3-1.6 differs
from the their reported first-order behaviour. Orders in hydrogen greater than
1 have been quoted by Bond et al. [17] for ethyne hydrogenation on platinum
and palladium catalysts, under different experimental conditions. In addition,
the data in Fig. 2 show that the apparent activation energy remained constant
regardless of the metal dispersion. The value of 16 + 2 kcal mol-1 is in reason-
able agreement with those reported for ethyne and valylene hydrogenation on
supported palladium of 14 and 16 kcal mol-1, respectively [13,18]. These sim-
ilarities of kinetic parameters at least suggest that the same mechanism and
nature of the rate-limiting step are valid for different alkyne hydrogenation
reactions.
A clear effect of particle size on the overall selectivity for ethane formation
is shown by the data in Fig. 5. It should be noted, however, that sample L 1-1
(60 % dispersion ) behaves differently from the others. The selectivity for ethane
was less than 1% up to 50% conversion. Therefore, it seems convenient to
discuss first the implications of this result. The very low concentration of ethane
in the product stream indicates that ethyne hydrogenation on small palladium
particles proceeds almost exclusively in the direction of ethene formation. In
other words, the intrinsic selectivity for ethene is close to 100%. As generally
accepted, this behaviour arises from the large difference in the strength of
adsorption between ethyne and ethene. Supporting evidence for this explana-
tion has been provided by a recent theoretical study [19]. Ethane formation
becomes important at low ethyne concentration owing to the parallel hydro-
genation of ethene. On the basis of our data, the hydrogenation of ethene is
suppressed if the ethyne concentration remains above 0.3%. This result is con-
 237

sistent with those obtained in earlier studies under different experimental con-
ditions. S~irkEiny et al. [9], using atmospheric pressure, reported an overall
ethane selectivity of 15% at 60-70% conversion using a comparable catalyst,
an initial ethyne concentration of 0.3% and a larger excess of ethene. In this
instance the amount of ethane produced directly from ethyne was of the order
of 4%. Hub et al. [ 16 ] observed that small palladium particles give a constant
selectivity of 98% towards 1-butene during the hydrogenation of 1-butyne, up
to a limiting conversion of 70%. Consequently, our first conclusion is that the
selectivity of high-dispersion palladium catalyst for the hydrogenation of
ethyne-ethene mixtures at medium pressure is mainly governed by the relative
heats of adsorption of these hydrocarbons.
Returning to the curves in Fig. 5, it is seen that low-dispersion samples pro-
duce significant amounts of ethane at all conversion levels. An interesting fea-
ture of these results is the marked increase in ethane selectivity observed in
the 0-10% dispersion range. In addition, the selectivity was constant up to a
minimum value of conversion, and this limit decreased at low palladium dis-
persion. Both tendencies are in very good agreement with those studies cited
above. S~irkEinyet al. [9] found a high ethane selectivity of 67% (at 60-70%
conversion) in the low-dispersion range (7%). As the dispersion increased to
38% the selectivity decreased by a factor of 3. During the hydrogenation of 1-
butyne, Hub et al. [16] observed that the selectivity for 1-butene, for a 26%
dispersion catalyst, was very high and constant (98%) up to 50% conversion
whereas, as mentioned before, it was around 70% for high-dispersion samples.
According to the results in Fig. 5, the undesirable formation of ethane at low
conversion could be expected if the metal dispersion is less than 20%.
An important question to address is whether this ethane is produced by
ethyne or ethene hydrogenation. The experimental evidence in Table 4 strongly
suggests that ethane formation on large palladium particles is due only to eth-
ene hydrogenation. The fact that a similar activation energy was obtained for
all the catalysts does not support a change in reaction mechanism with particle
size. Consequently, we postulate that the selectivity for ethene in the C2H2+ H2
reaction is very high regardless of the metal dispersion. This implies that the
parallel hydrogenation of ethene occurs more easily as the particle size in-
creases, leading to a pronounced effect on the amount of ethane formed. One
argument that could be invoked to explain this behaviour is the presence of
/?-hydride phase which is due to the high hydrogen pressure used in our exper-
iments. However, it is noted that a marked change in selectivity occurs when
the dispersion falls below 20%, although the fl-hydride phase has been detected
even on samples of 50% dispersion [20]. The C2H4-~ C2H6 reaction could also
be ascribed to the presence of ethylidyne on the metal surface. Although sev-
eral adsorbed species derived from ethyne and ethene have been identified on
platinum, palladium and rhodium, it is difficult to prove that they play an
active role in the hydrogenation processes [21,22]. In our experiments, the
238

initial activity for ethene hydrogenation observed on low-dispersion samples

is consistent with the kinetic behaviour already discussed. The decrease in
TON and reaction order with increasing dispersion indicates a stronger ethyne
adsorption on the small palladium particles, which in turn precludes the hy-
drogenation of ethene. In addition, it has recently been suggested [22 ] that on
large palladium particles the presence of adsorbed species such as ethylidyne
or ethylidene inhibits the strong adsorption of ethyne to the metal surface.
Consequently, the adsorption and reaction of ethene could be facilitated.
Therefore, we can reconcile the effect of particle size with the activity and
selectivity results. On the basis of this study, it would be necessary to inhibit
the hydrogenation of ethene in order to improve the selectivity of low-disper-
sion Pd/a-A1203 catalysts. In commercial practice this is achieved by addition
of carbon monoxide. Another approach involves the use of bimetallic catalysts.
The beneficial effect of alloying palladium with lead has already been reported
[231 for ethyne hydrogenation.

CONCLUSIONS

The main result that emerges from this study is that the dependence of TON
on palladium dispersion, previously observed for ethyne hydrogenation in the
presence of a large excess of ethene and hydrogen, is less pronounced when the
reaction is carried out at low ethyne concentration (ca. 0.2% ). Our kinetic data
have demonstrated that the rate order for ethyne varies from 0 to - 0.5 as the
dispersion increases. Consequently, the TON for the removal of trace amounts
of ethyne from front-end mixtures does not depend on the palladium disper-
sion. In other words, the particle size is not an important variable for improv-
ing the hydrogenation activity of Pd/~-A1203. However, low-dispersion cata-
lysts are used in industrial applications. Their lower rates per unit mass result
in a more stable reactor operation as the occurrence of thermal effects is
minimized.
It has also been shown that the selectivity for ethene of Pd/o~-A1203 is very
high regardless of the palladium dispersion and the high ethene and hydrogen
partial pressures. Low-dispersion catalysts favour the undesirable formation
of ethane via the parallel hydrogenation of ethene. This reaction is almost
absent on high-dispersion samples owing to the strong and multiple adsorption
of ethyne [ 14 ]. In commercial practice the addition of carbon monoxide inhib-
its the ethene hydrogenation, which in turn improves the selectivity.

REFERENCES

1 M.L. Derrien, in L. Cerven~ (Ed.), Catalytic Hydrogenation (Studies in Surface Science and
Catalysis, Vol. 27) Elsevier, Amsterdam, 1986, p. 613.
 239

2 C.E. Gigola, H.R. Addriz and P. Bodnariuk, Appl. Catal., 27 (1986) 133.
3 A. Borodzirlski, R. Dug, R. Frak, A. Janko and W. Palczewska, in G.C. Bond, P.B. Wells and
F.C. Tompkins (Eds.), Proceedings of the 6th International Congress on Catalysis, London,
July 12-16, 1976, Chemical Society, London, 1977, p. 150.
4 G. Carturan, G. Facchin, G. Cocco, S. Enzo and G. Navazio, J. Catal., 76 (1982) 405.
5 W.T. McGown, C. Kemball, D.A. Whan and M.S. Scurrell, J. Chem. Soc., Faraday Trans. I,
73 (1977) 632.
6 L. Guczi, R.B. LaPierre, A.H. Weiss and E. Biron, J. Catal., 60 (1979) 83.
7 J. Margitfalvi, L. Guczi and A.H. Weiss, J. Catal., 72 ( 1981 ) 185.
8 A. S~irkEiny,L. Guczi and A.H. Weiss, Appl. Catal., 10 (1984) 369.
9 A. S~irkdny, A.H. Weiss and L. Guczi, J. Catal., 98 (1986) 550.
10 J.J.S. Scholten and J.A. Konvalinka, J. Catal., 5 (1966) 1.
11 M. Boudart and H.S. Wang, J. Catal., 39 (1975) 44.
12 J.P. Boitiaux, J. Cosyns and S. Vasudevan, Appl. Catal., 6 (1982) 41.
13 H.R. Adtlriz, P. Bodnariuk, B. Coq and F. Figueras, in preparation.
14 J.P. Boitiaux, J. Cosyns and E. Robert, Appl. Catal., 32 (1987) 145.
15 J.P. Boitiaux, J. Cosyns and E. Robert, Appl. Catal., 32 (1987) 169.
16 S. Hub, L. Hilaire and R. Touroude, Appl. Catal., 36 (1988) 307.
17 G.C.Bond, D.A. Dowden and N. Mackenzie, Trans. Faraday Soc., 54 (1958) 1537.
18 W.L. Kranich, A.H. Weiss, Z. Schay and L. Guczi, J. Catal., 13 (1985) 257.
19 H. Nakatsuji, M. Hada and T. Yonezawa, Surf. Sci., 185 (1987) 319.
20 R.K. Nandi, P. Georpopoulos, J.B. Cohen, J.B. Butt, R.L. Burwell and D.H. Bilderback, J.
Catal., 77 (1982) 421.
21 T.P. Beebe and J.T. Yates, J. Am. Chem. Soc., 108 (1986) 663.
22 N.R.M. Sassen, Ph.D. Thesis, University of Leiden, 1989.
23 W. Palczewska, A. Jablonski and Z. Kaszkur, J. Mol. Catal., 25 (1984) 307.