JOURNAL OF CATALYSIS 40, 94- 100 (1975)

Catalysis and Coordinative Unsaturation of

Active Sites on Sulfurated Nickel Catalyst1

I. Reversible Formation of Active Sites for

Partial Hydrogenation of Acetylene

ATSUSHI TAKEUCHI, KEN-ICHI TANAKA, ISAMU TOYOSHIMA

AND KOSHIRO MIYAHARA
Research Institute for Catalysis, Hokkaido University, Sapporo 060, Japan
Received March 5, 1975

Active sites are generated on sulfurated nickel surface by contacting with acetylene, on
which the hydrogenation of acetylene and ethylene as well as the Hz-D, equilibration reac-
tion proceed. These sites are wiped up by exhaustion of acetylene, and such reversible for-
mation and vanishment of the active sites controls the partial hydrogenation of acetylene to
ethylene on the sulfurated nickel catalyst. A nickel sulfide having an atomic ratio of
S/Ni = 0.62 had the activity for the partial hydrogenation of acetylene, while a nickel sul-
fide of S/Ni = 0.96 had no catalytic activity. The surface composition of the sulfurated
nickel estimated by Auger electron spectroscopy was a value of the ratio S/Ni = 0.69.
From these results it is concluded that partial hydrogenation is enabled by reversible for-
mation of active sites on sulfide with appropriate coordinative unsaturation of nickel atoms.

INTRODUCTION tions. This paper gives the evidences for

Modification of nickel catalyst with chal- reversible formation of active sites for hy-
cogen such as sulfur and selenium, and drogenation on sulfurated nickel by con-
with halogen enhances the catalytic selec- tact with acetylene (7,8). Surface composi-
tivity for partial hydrogenation of acety- tion of sulfurated nickel was estimated by
lene to ethylene (Z), and similar selective Auger electron spectroscopy (AES). And
partial hydrogenation of polyolefins over the catalytic activity and selectivity of sul-
sulfurated nickel and nickel sulfide have furated nickel are well understood by the
been reported; selective hydrogenation of coordinative unsaturation of surface nickel
isoprene to methyl-butenes (2) and cyclic atoms (8).
polyenes to corresponding cyclic mono-
olefins (3) over sulfurated nickel, partial EXPERIMENTAL METHODS
hydrogenation of 4-vinylcyclohexene on Reactions were carried out in a conven-
nickel subsulfide N&S, (4). Partial hy- tional, closed circulation system similar to
drogenation of polyenes to corresponding the previous one (9) with a volume of 600
monoolefins was found to take place on ml and the circulating velocity of about
nickel sulfide NiS (5), however, it was 1700 ml/min. Hydrogen or deuterium was
found to be Ni& under its working state purified by diffusing a palladium-silver
(6). thimble. High purity carbon monoxide and
It is interesting to throw light on the role oxygen (Takachiho Chem. Co.) were used
of sulfur in the selective partial hydrogena- without further purification and acetylene
1 Abstracted in part from the doctoral thesis of A. (Takachiho Chem. Co.) was purified by
Takeuchi, Hokkaido University. distillation in vacua.
94
Copyri.@t @ 1975 by Academic Press. Inc.
AU rights of reproduction in any form reserved.
 CATALYSIS ON SULFURATED NICKEL. 1 9.5

Analysis of reacting gases were carried Co.) at 500°C in air was reduced with
out by a gas chromatograph with a silica hydrogen at 350°C for 100 hr in a reactor,
gel column for methane, ethane, ethylene and nickel powder obtained was sulfurated
and acetylene, and with a VZ-7 column with hydrogen sulfide at 300°C. The rate
(Gaschro Industrial Co.) for propylene, of sulfuration became very slow at SlNi
isomers of butene, 1,3-butadiene and n-bu- approached to 1.0 and the amount of hy-
tane. Deuterated hydrocarbons and hy- drogen sulfide reacted with nickel powder
drogen were analyzed with a mass spec- was estimated from the amount of re-
trometer (Hitachi RMU-6). Ethylene was maining hydrogen sulfide condensed into
analyzed at 13 V of the ionization voltage, liquid nitrogen trap.
while hydrogen at 70 V. Acetylene and By using a nickel plate of 1 X 10 X 10
ethane were analyzed with gaschro-mass mm3 (Toshiba Elec. Co.), surface composi-
spectrometer with a silica gel column at 70 tion of the sulfurated nickel was examined
V of the ionization voltage. by Varian 98 1-1000 analytical Auger spec-
Nickel wires (Toshiba Electric Co.) of trometer, of which the primary electron
0.1 mm in diameter and of 2 10 m (Fig. l), was 1.5 keV. The base pressure of the
130 m (Fig. 3) or 160 m (Fig. 4) in length system was of the order of 10es mm Hg.
were used as catalyst. Nickel wire None of metals other than nickel was de-
mounted in a reactor was heated prelimi- tected by the AES analysis of the nickel
narily in 1 mm Hg oxygen at 500°C for 5 plate and wire used in this work.
min and reduced with hydrogen of 60 mm NiS powder of 99.9% purity (Yamanaka
Hg at 500°C for 1 hr. The wire was ox- chemical Industrial Co.) was disked by a
idized again in 1 mm Hg oxygen at 500°C pressure of 560 kg/cm2 and adopted as a
for 5 min and reduced with hydrogen of standard sample for AES analysis.
few tens mm Hg at 300°C for 10 hr, and In AES analysis, the sample was set in
finally evacuated for 2 hr at 300°C. the AES apparatus in air and after evacua-
As the catalytic properties of sulfurated tion at a room temperature differential en-
metals depend on the sulfur compound ergy distribution &V(E)/& was recorded
used, duration and temperature of sulfura- vs Auger electron energy E. Table 1
tion (2,3,5), sulfuration in this work was shows the peak positions of respective ele-
carried out by hydrogen sulfide to exclude ments, which are assigned by energy at the
deposition of carbon compounds on the high-energy minimum in differential energy
surface of catalyst. Hydrogen sulfide of distribution curve (20).
0.8-l. 1 mm Hg was completely decom- In order to estimate the penetration of
posed on nickel catalyst at 300°C for 2 sulfur in bulk, the plate sample was sput-
min. After evacuation of evolved hydrogen tered by argon ions (600 eV) at a pressure
for 4 min at 300°C the reactor was cooled of 5 X lo-’ mm Hg. The change of peak-
to reaction temperature. Amount of sulfur
TABLE 1
deposited on nickel is much more than
AUGER PEAK POSITIONS
monolayer and the atomic ratio (S/Ni) of
deposited sulfur to total nickel of wire is Element Energy (eV) Transition
about 1O-4.
Ni 783 LMM
In order to compare the catalytic prop- 61 MMM
erties of sulfurated nickel, nickel sulfides
having SlNi ratio of 0.02, 0.62 and 0.96
were prepared as follows. Nickel oxide of 0 510 KLL
about 0.5 g prepared by pyrolysis of nickel C 272 KLL
S 152 LMM
carbonate (special grade from Koso Chem.
96 TAKEUCHI ET AL.

peak amplitudes of differential curve was simultaneous dimerization reaction also

plotted as a function of sputtered depth es- has the same induction period and 1,3-bu-
timated by number of argon ions, Nion, tadiene formed is hydrogenated to butene
bomberded per unit surface area of sample during hydrogenation of acetylene. Con-
which is given by secutive hydrogenation of produced ethyl-
ene and butene are very slow even after
I* x t the exhaustion of acetylene in contrast
Nion = 1.6 x lo-Is’
with nickel catalyst, over which olefin is
where IA is a current density and t is a successively hydrogenated as soon as acet-
duration of argon ion bombardment. ylene is exhausted (9,13).
Number of sputtered atoms from the Even though the rate of acetylene hy-
sample, Natom,is given by Y X Nion where drogenation on sulfurated nickel is less
Y is sputtering yield. If one assumes 1015 than one tenth of that of nickel, distribu-
atoms/cm2 on sample surface, number of tion of products during acetylene hy-
sputtered layers is given by Nab,,,/ 1015,and drogenation is nearly the same on nickel
the sputtered depth (A) is given by a fol- and sulfurated nickel as illustrated in Fig.
lowing equation: 2. Selectivity Sz is a ratio of ethylene and
ethane (C,) to the total amount of con-
aXZ,XYXt
L= sumed acetylene, and selectivity SE is a
1.6 x 1O-4 ’
percentage of ethylene in Cz products (9).
where u is the lattice constant of sample. This result suggests that active sites for
In this case IA was 4.8 PA/cm2 and a and hydrogenation of acetylene and its dimeri-
Y are approximated with those of nickel, zation on nickel and sulfurated nickel are
a = 3.2 8, and Y = 1.2 (II). commonly nickel atoms and sulfur retards
the consecutive hydrogenation of pro-
RESULTS AND DISCUSSION duced monoolefins.
Typical time course of acetylene hy- After the first run of Fig. 1, run 2
drogenation on freshly prepared sulfurated (broken curve) was carried out on the sul-
nickel is shown in Fig. 1 with solid lines. furated nickel which was treated with 30
The freshly sulfurated surface becomes ac- mm Hg hydrogen at 300°C for 18 hr and
tive for hydrogenation of acetylene after evacuated. After run 2, catalyst was kept
an induction period of several hours. The at room temperature in contact with
reacted gas for more than 1 wk and run 3
was started after degassing at 119°C. In-
duction period is shortened and catalytic

100
a+@ 52
60 0
t---
lb) 26Ot , , , , ,

time (hrs) 0 20 40 60 60 ICXJ

FIG. 1. Hydrogenation of acetylene on sulfurated acetylene conversion (%I
nickel at 119°C. Initial pressure are CzHz = 19.8 mm FIG. 2. Dependence of selectivities upon acetylene
Hg and H, = 39.0 mm Hg. conversion on nickel (0) and sulfurated nickel (a).
 CATALYSIS ON SULFURATED NICKEL. 1 97

activity is increased by these treatments as

well as by repeated uses of the catalyst. In
this series of experiments, none of sulfur
compounds were detected in gas phase
with mass spectrometer. These facts
suggest that nickel is enriched on catalyst
surface by diffusion of sulfur into metal J.

bulk and this was proved by AES analysis. 0 5 IO

I
15
1
20 25
Hydrogenation of ethylene on freshly time (hrs)
sulfurated nickel was so slow as to give FIG. 4. Induction phenomena during acetylene hy-
only 2% of conversion in 45 hr at 120°C drogenation on sulfurated nickel at 119°C. Initial
while consecutive hydrogenation of pro- pressures are C,H, = 18.9 mm Hg and Hz = 74.3 mm
Hg. Arrows (l), (2) and (3) the addition of acetylene;
duced ethylene is obviously accelerated in (0) removing of acetylene by trapping; (---) slopes ex-
a brief time after the exhaustion of acety- pected at no induction period.
lene as shown in Fig. 1 and more clearly in
Fig. 3. The surface I of sulfurated nickel is also
These results and the presence of induc- inactive for the Hz-D, equilibration reac-
tion period infer that the freshly sulfurated tion except the abnormal isotopic mixing
surface (surface I) is changed during the at the initial few minutes. When acetylene
induction period to surface II being active is added to a mixture of H, and DI, hy-
for hydrogenation of acetylene as well as drogenation starts with an induction
ethylene, and that the surface II goes back period. If acetylene is removed from the
reversibly to surface I by exhaustion of reacting gas with liquid nitrogen trap, then
acetylene. Hz-D, equilibration reaction proceeds
Experiments given in Fig. 4 ensures this markedly for a brief time, as shown in Fig.
inference. The addition of acetylene during 5, similarly to the consecutive hydrogena-
its hydrogenation gave no activity change, tion of ethylene demonstrated in Fig. 3.
whereas obvious induction phenomena The surface II, but not the surface I, of
were observed after several hours of com- sulfurated nickel is thus proved to be ac-
plete consumption of acetylene or by trap- tive for the Hz-D, equilibration as well as
ping of acetylene with liquid nitrogen. hydrogenation.

0 IO 20 30
0 time (hrs)
0 IO 20 30 FIG. 5. Equilibration reaction of H, and D,
time (hrs)
(HZ/D2 = 2/3) on sulfurated nickel at 120°C. Acety-
FIG. 3. Hydrogenation of acetylene on sulfurated lene is added at 5.1 hr and trapped at 18.0 hr. (equ.)
nickel at 119°C (run 2 of Fig. 1). An equilibrium composition.
98 TAKEUCHI ET AL.

Deuterium distribution in acetylene, eth- the nickel sulfide of WNi = 0.62. Thus the
ylene and ethane formed by the reaction of partial hydrogenation of acetylene closely
acetylene with a mixture of Hz and DI relates to the presence of an induction
(Hz/D, = 2/3) is given in Table 2. Eth- period. The composition of the reaction
ylene-d,, -dl and -dz are the main species products over the sulfide of S/Ni = 0.62
of deuteroethylene, obviously differing was similar to those over the nickel (9) and
from a random distribution of D atoms. the sulfurated nickel. Hydrogenation of
Even though the measurement is less accu- acetylene over this sulfide was reprodu-
rate, relative abundances of deu- cible in contrast with the sulfurated nickel
teroethanes from -dl to -de are given in (S/Ni = 10m4), suggesting that a steady
Table 2. Deuterium is not found in acety- ratio of nickel to sulfur on the sulfide
lene at all, revealing that acetylene adsorbs makes a reproducible activity in contrast
irreversibly on sulfurated nickel. with the rising of activity on sulfurated
Hydrogenation of acetylene on nickel nickel. On the other hand, the nickel sul-
sulfides was studied out to compare with fide having ratio of S/Ni = 0.96 was inac-
sulfurated nickel. On nickel sulfide of tive for hydrogenation of acetylene, re-
S/Ni = 0.02, acetylene was hydrogenated vealing that nickel atoms of this sulfide are
to ethane without induction period as was coordinatively saturated by sulfur.
the case of nickel catalyst (9,22,13). In Figure 6 shows a typical result of AES
contrast with this, the selective partial hy- measurement of freshly sulfurated nickel,
drogenation of acetylene to ethylene and a and the changes of the peak-peak ampli-
marked induction period were observed on tude are plotted as a function of sputtered
depth L. Sulfur diffuses in nickel bulk as
TABLE 2
DEUTERKJM DISTRIBUTIONS (%) IN ACETYLENE,
deep as 500 A, showing that sulfur is con-
ETHYLENE AND ETHANE FORMED BY C,H, siderably mobile in nickel bulk. Peak-peak
HYDROGENATION WITH HI-D* MIXTURE amplitude of 152 eV sulfur peak decreases
IN FIG. 5 at the surface as shown in Fig. 6, which
may be masked by the adsorbed of oxygen
Time (hr): 12.9 17.9
and impurities came from air.
Ethylene-d, 4.3 3.6 (0.6)n Figure 7 shows the result of AES analy-
4 5.7 6.4 (6.3) sis of a NiS disk. Each peak-peak ampli-
4 19.0 20.4 (24.4) tude is constant in bulk phase, suggesting
4 35.0 36.6 (41.8)
do 36.0 33.0 (26.9)
Av D content (o/o) 27 28 wlfurated nickel

Ethane-d, - 0.8 (0.2)

4s - 2.8 (2.3)
4 9.9 9.2 (10.1)
4 21.1 19.3 (24.3)
4 43.2 33.9 (32.7)
4 25.8 34.0 (23.4)
do - - (7.0)
Av. D content (%) 36 36

Acetylene-d,, -d2 0 0 depth (%,

do 100 100 AES analysis of sulfurated nickel having

heen contacted with ethylene; Ni at 61 eV (O), Ni at
a Parentheses show the random distribution of 783 eV (0), S at 152 eV (a), C at 272 eV (A) and 0
deuterium. at 510 eV (0).
 CATALYSIS ON SULFURATED NICKEL. 1 99

2 A nickel catalyst poisoned by a trace of

5 I nickel sulfide (NISI
CO also enhances the partial hydrogena-
tion of acetylene and 1,3-butadiene (7).
Nickel atom in N&S, is surrounded by
four S atoms, while by six S atoms in NiS,
and hence Ni atoms on the surface of
Ni.& as well as sulfurated nickel may
have appropriate coordinative unsaturation
of nickel atoms in contrast with NiS. In
conclusion, the partial hydrogenation is
mainly controlkd by reversible formation
of coordinatively unsaturated nickel atoms
dspth 6,
being generated by replacing with acety-
FIG. 7. AES analysis of standard NiS sampIe. As-
signment of plots is the same as Fig. 6. lene.

ACKNOWLEDGMENTS
that the change of bulk canrposftion The arthors are grateful to Dr. Yoshitaka Hayashi
caused by tht! argon ion bomb&r&rent, if and Dr. Mitsuko Tomori of NEVA (Nippon Electric-
any, is negligible. Peak-peak amplitude of Varian Co.) for their valuable AES measurements.
nickel at 61 eV changes in parallel with
that of sulfur at 152 eV. On accotHlt of es- REFERENCES
cape depth of electron (14), nickel peak at 1. Kwan, T., Shokubai 1, 28 (1946). Matsuda, A.,
61 eV is more appropriate than that at 783 Shokubai 3, 1 (1948). Kwan, T., and Izu, T.,
eV to estimate the ratio S/Ni on the cata- Shokstbai 3, 6, 14 (1948). Kwan, T., Izu, T.,
lyst surface. Figure 7 shows that peak of Takasaki, Y., Koshino, K., and Murayama, Y.,
Shokubai 3, 2 1 (1948).
sulfur is three times as sensitive as that of 2. Bourne, K. H., Holmes, P. D., and Pitkethly, R.
nickel at 61 eV. Figure 8 is obtained from C., Proc. Int. Congr. Catal., 3rd, I964 2, 1400
Fig. 6 by the use of above ratio of sensitiv- (1965).
ity of sulfur to nickel. Atomic ratio S/Ni of 3. Zuech, E. A., U. S. Pat. 3, 493, 625 (1970).
the sulfurated nickel surface is estimated 4. Johnson, M. M., Nowack, G. P., and Tabler, D.
C., J. Catal. 27, 397 (1972).
as 0.69 which is nearly equal to 0.67 of 5. Greensfelder, B. S., and Peterson, W. H., U. S.
nickel subsulfide Ni& known as an effec- Pat. 2, 402, 493 (1946).
tive catalyst for selective partial hydrogen- 6. Kirkpatrick, W. J., in “Advances in Catalysis”
ation (6). (W. G. Frankenhurg, V. I. Komarewsky and
E. K. Rideal, Eds.), Vol. 3, p. 329. Academic
Press, New York, 1951.
I.0 --_-- ------ pJis 7. Takeuchi, A., Tanaka, K., and Miyahara, K.,
Chem. tett. (Chem. Sot. Jup.) 171 (1974).
0.8 8. Takeuchi,A., Tanaka, K., and Miyahara, K.,
-----------N,& Chem. Lett. (Chem. Sot. Jap.) 411 (1974).
0.6 9. Takeuchi, A., and Miyahara, K., J. Res. Inst.
S/Ni Catal. Hokkaido Univ. 21, 132 (1973).
0.4 10. Chang, C. C., Surface Sci. 25, 53 (1971); Palm-
i berg, P. W. et al., “Handbook of Auger Elec-

O.‘L
tron Spectroscopy.” Physical Electronics In-
dustries, Inc., Minnesota, 1972.
11. Kaminsky, M., “Atomic and Ionic Impact Phe-
0 00200300400500600
nomena on Metal Surfaces.” Springer-Verlag,
%ezfhI, l%S.
FIG. 8. Change of atomic ratio S/Ni in &&rated 1.2. Bond, G. C., “Catalysis by Metals.” Academic
nickel along the sputtered depth. (--) Vahcs for stoi- Press, New York, 1962.
chiometric NiS and Ni&. 13. Sheridan, J., J. Chem. Sot. London 373 (1944);
100 TAKEUCHI ET AL.

133, 301, 305 (1945); Bond, G. C.,J. Chem. 27 1 (1970); Mann, R. S., and Naik, S. C., Can.
Sot., London 2705, 4288 (1958); Bond, G. C., J. Chem. 45, 1023 (1967); Mann, R. S., and
and Mann, R. S., J. Chem. Sot., London 4738 Khulbe, K. C., J. Catal. 13, 25 (1969); Can. J.
(1958); Hirota, K., Yoshida, N., and Teratani, Chem. 45, 2755 (1967).
S., J. Catal. 15, 425 (1969); Teratani, S., and 14. Brundle, C. R., J. Vuc. Sci. Technol. 11, 2 12
Hirota, K., 2. Phys. Chem. Neue Folge 69, (1974); Bauer, E., Vacuum 22, 539 (1972).