Hydrogen-deuterium exchange of methane on nickel and potassium promoted nickel prepared by the reduction of nickel oxide

Article abstract of DOI:10.1016/j.molcata.2010.01.005

The hydrogen-deuterium exchange reactions of methane in a deuterium stream were studied by a pulse experiment over a reduced nickel (Ni1373 prepared from nickel oxide calcined at 1373 K, and Ni773 prepared from nickel oxide calcined at 773 K) and K₂O promoted reduced nickel (K-Ni1373 and K-Ni773). Ni1373 had a higher exchange activity than Ni773. The effects of the addition of K₂O resulted in the decrease in the exchange activity due to the inhibition of step defect sites and/or the promotion of the change of nickel crystal structure. High activity of Ni1373 could be due to that the Ni1373 had the surface of a higher Ni(1 0 0)/Ni(1 1 1) ratio.

Full text of DOI:10.1016/j.molcata.2010.01.005

Journal of Molecular Catalysis A: Chemical 320 (2010) 68–71
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Hydrogen–deuterium exchange of methane on nickel and potassium promoted
nickel prepared by the reduction of nickel oxide
∗
Tsutomu Osawa , Takashi Futakuchi, Tomoharu Imahori, I-Yin Sandy Lee
Graduate School of Science and Engineering for Research, University of Toyama, Gofuku, Toyama 930-8555, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The hydrogen–deuterium exchange reactions of methane in a deuterium stream were studied by a pulse
experiment over a reduced nickel (Ni1373 prepared from nickel oxide calcined at 1373 K, and Ni773
prepared from nickel oxide calcined at 773 K) and K2O promoted reduced nickel (K-Ni1373 and K-Ni773).
Ni1373 had a higher exchange activity than Ni773. The effects of the addition of K2O resulted in the
decrease in the exchange activity due to the inhibition of step defect sites and/or the promotion of the
change of nickel crystal structure. High activity of Ni1373 could be due to that the Ni1373 had the surface
of a higher Ni(1 0 0)/Ni(1 1 1) ratio.
Received 10 September 2009
Received in revised form 6 January 2010
Accepted 6 January 2010
Available online 15 January 2010
Keywords:
Nickel oxide
Hydrogen–deuterium exchange
Methane
©
2010 Elsevier B.V. All rights reserved.
Nickel
K2O promoted nickel
1
. Introduction
nickel oxide calcined at 773 K (NiO773) [7]. This could be due to
the difference between the surface structure of Ni1373 and that of
Ni773.
A tartaric acid-modified nickel catalyst is one of the most
successful enantio-selective heterogeneous catalysts. For the
hydrogenation of prochiral ketones, high enantioselectivity was
attained for ␤-ketoesters (up to 98% e.e.) and 2-alkanones (up to
The hydrogen–deuterium (H/D) exchange reaction of CH4 with
D2 over various metals [8–10] and metal oxides [11–13] has been
studied for many years. As for the nickel catalysts reported in the
literature, they were prepared by the direct reduction of nickel salt
precursors at 673 K to 923 K [12,14,15] or prepared through the
8
5% e.e.) [1–3]. The catalyst is simply prepared by immersing an
activated nickel catalyst (base nickel catalyst) in a solution of tar-
taric acid [4]. The enantioselectivity is obtained by the adsorption
of the tartaric acid on the nickel surface. For the base nickel cata-
lyst, various types of nickel were used, for example, Raney nickel,
supported nickel, reduced nickel (nickel prepared by the reduc-
tion of nickel oxide by hydrogen gas), and nickel fine powder.
From these studies, it has been revealed that the condition of the
nickel surface is one of the important factors for attaining a high
enantioselectivity, i.e. the enantio-selective hydrogenation over a
tartaric acid-modified nickel catalyst is a structure-sensitive reac-
tion [5,6]. It was reported that the modified catalyst with greater
than a 40-nm nickel crystallite size gave a high enantioselectivity
reduction of nickel oxide, which was calcined at 673 K, on SiO [16].
2
As the H/D exchange of methane would also be a structure-sensitive
reaction (the activity over Ni(1 0 0) is higher than Ni(1 1 1)) [17,18],
the exchange activity reveals which the surface is predominant for
the catalytic reaction. In this work, we have drawn attentions on the
H/D exchange reactions over Ni773 and Ni1373 in order to reveal
the differences of these surfaces. In the previous paper, Ni773 and
Ni1373 were prepared at a constant reduction temperature (reduc-
tion of nickel oxide to nickel was carried out at 623 K) [18]. We have
found the features of the exchange reaction of Ni1373 were much
different from those of Ni773 and those reported Ni catalyst in the
literature. That is, the preparation of the nickel catalyst through the
nickel oxide calcined at high temperature would result in a differ-
ent nickel surface, when compared to the preparation not through
the nickel oxide.
[
6]. It was also reported that the enantioselectivity of a modified
reduced nickel catalyst depended on the preparation temperature
of the nickel oxide. The reduced nickel (Ni1373) prepared from
the nickel oxide calcined at 1373 K (NiO1373) had a higher enan-
tioselectivity than the reduced nickel (Ni773) prepared from the
In the present study, the hydrogen–deuterium (H/D) exchange
reaction of CH4 with D2 was carried out over Ni773 and Ni1373.
The effects of the reduction temperature of nickel oxide to nickel
and furthermore, the addition of potassium on both the exchange
activity and the exchanged species were examined.
∗ Corresponding author.
E-mail address: osawa@sci.u-toyama.ac.jp (T. Osawa).
1
381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.01.005

T. Osawa et al. / Journal of Molecular Catalysis A: Chemical 320 (2010) 68–71
69
2
. Experimental
which however, overlap each other. Therefore, to obtain the pure
molecular ion peaks of the methane isotopologues, corrections
have to be made for the observed ion peaks. In the present study,
to determine each of the isotopologues in the pulse experiment,
the pulse areas of the ion current were calculated by our newly
proposed method. The detailed correction method is described
elsewhere [19].
2.1. Materials
CH4 (99.99%), D2 (99.995%), and He (99.995%) were supplied
from the Takachiho Trading Co., Ltd., Japan.
2.2. Preparation of nickel oxide
The composition of the isotopologue P(CH D4-i) (i = 4–0) after the
i
exchange reaction was calculated using the peak area S(CH D4-i) of
i
Nickel oxides were prepared by decomposing and calcining
its corrected intrinsic parent ion (Eq. (1)).
nickel hydroxide (Wako Pure Chemical Ind., Ltd., Lot ACJ6158) in
3
−1
3
−1
P(CH4) = 100 × S(CH4)/Ssum
P(CH D) = 100 × S(CH D)/S
a mixture of N2 (40 cm min ) and O2 (10 cm min ) for 3 h.
The nickel oxides prepared at 773 K and 1373 K were designated
NiO773 and NiO1373, respectively.
3
3
sum
P(CH D ) = 100 × S(CH D )/Ssum
2
2
2
2
(1)
2
.3. Preparation of K O promoted nickel oxide
P(CHD3) = 100 × S(CHD3)/Ssum
P(CD ) = 100 × S(CD )/Ssum
2
4
4
Potassium carbonate (Wako Pure Chemical Ind., Ltd., Lot LTE
645, 0.00367 g) was dissolved in distilled water (40 cm ). Nickel
3
Ssum = S(CH ) + S(CH D) + S(CH D ) + S(CHD ) + S(CD )
1
4
3
2
2
3
4
hydroxide (2.16 g) was added to the potassium carbonate solution
and stirred at 353 K for 2 h and dried. The resulting solid was cal-
3. Results and discussion
3
−1
3
−1
cined in a mixture of N2 (40 cm min ) and O2 (10 cm min ) at
73 K or 1373 K for 3 h. The K O promoted nickel oxides prepared
7
3.1. The H/D exchange activity over the reduced nickel catalyst
and the K2O promoted reduced nickel catalyst
2
at 773 K and 1373 K were designated K-NiO773 and K-NiO1373,
respectively.
Table 1 shows the H/D exchange activities of Ni773 and Ni1373
prepared at reduction temperatures from 573 K to 723 K. The
exchange activity of Ni1373 has turned out to be much higher than
that of Ni773 at all the reduction temperatures. Taking into account
that the exchange activity of Ni(1 0 0) would be higher than Ni(1 1 1)
[17,18], the surface of Ni1373 could have a higher Ni(1 0 0)/Ni(1 1 1)
ratio. This idea would also be supported by the following reasons.
(i) It was reported that NiO(1 0 0) can be reduced to Ni(1 0 0) by
hydrogen at 623 K [20]. (ii) As nickel oxide is a crystal like NaCl,
the NiO(1 0 0) face has the lowest surface energy density (ꢀ). It is
reasonable to infer the equilibrium form of the crystal is a hex-
2
.4. Catalytic test
A continuous gas flow fixed bed glass reactor (14 mm id) was
used for the catalytic test. The measurement without an exchange
reaction was first carried out by a CH pulse (0.6 cm ) without
a catalyst in a He (50 cm min ) stream. Consequently, NiO773,
NiO1373, K-NiO773, or K-NiO1373 (0.127 g) was placed in the reac-
tor and treated with a D stream (mixture of D2 (3 cm min
3
4
3
−1
3
−1
)
2
3
−1
and He (50 cm min )) at an atmospheric pressure for 1 h. The
reduction of nickel oxide was confirmed by XRD (Shimadzu XD-
3
A) experiments. The resulting reduced nickel catalysts were
ahedron, which mainly consists of (1 0 0) faces. When ꢀ(1 1 1) is
√
designated Ni773, Ni1373, K-Ni773, and K-Ni1373 respectively.
Reduction was carried out at 573, 623, 673, or 723 K. The result-
ing catalysts were designated (K-)Ni773-573, (K-)Ni773-623 and so
on. The H/D exchange reaction was carried out in the same reactor
by a CH4 pulse (0.6 cm ) under a continuous gas flow of a mix-
ture of D (3 cm min ) and He (50 cm min ) at 573 K. The outlet
from the reactor was connected to a Q-mass spectrometer (PFEIF-
FER Vacuum Prisma QMS 200) by a silica capillary tube (0.05 mm
id × 2.5 m). The data set of the ion current (m/z = 12–20) was col-
lected every 0.7 s. Background corrections to the observed spectra
were made (H O at m/z = 18). The second exchange reaction was
carried out at a 45-min interval to check the stability of the cata-
lyst. From the beginning of the first run to the end of the second
smaller than 3ꢀ(1 0 0), however (1 1 1) may appear on the surface
of the crystal [21]. As the NiO1373 was calcined at high tempera-
ture, it is reasonable to assume that the crystal would be relatively
stable and have a high NiO(1 0 0)/NiO(1 1 1) ratio. (iii) As the crys-
tallinity of Ni1373 was higher than that of Ni773 (nickel crystallite
size, Ni1373: 68 nm, Ni773: 28 nm) [22], the Ni1373 would have
a larger terrace without disorder than Ni773. Table 1 also indi-
cates the exchange activity of Ni773 and Ni1373 decreased with
the increase in the reduction temperature of nickel oxide. As the
increase of the reduction temperature caused the sintering and
decrease in the BET surface area, the specific exchange activity is
also shown in Table 1. The specific activity over Ni773 decreased
with the increase in the reduction temperature, while that over
Ni1373 increased with the increase in the reduction temperature.
3
3
−1
3
−1
2
2
run, the gas flow of D and He was maintained. As the results of the
2
first run and the second run were almost the same, the catalysts
were considered stable during the experiment and the results of
the first run are shown in this paper.
Table 1
Exchange activity of Ni773 and Ni1373 prepared at various reduction temperatures.
Catalysta
Exchange
activity/%
BET surface
Specific exchange
activity /%
2.5. Determination of BET surface area
2
−1
b
area/m
g
Ni773-573
Ni773-623
Ni773-673
Ni773-723
29
7.0
4.1
2.4
2.1
4.1
1.4
0.63
0.62
The BET surface area of the reduced nickel catalysts was mea-
sured using a Micromeritics Gemini 2375 by N2 adsorption at 77 K.
5.6
1.5
1.3
2
.6. Determination of the exchanged species by the correction of
Ni1373-573
Ni1373-623
Ni1373-673
Ni1373-723
48
43
41
37
4.2
3.3
2.4
1.9
12
13
18
20
the methane isotopologues
The determination of the isotopologues of methane was carried
out by quadrupole mass spectroscopy. The methane isotopologues
CH D (i = 4–0) have shown characteristic fragmentation patterns,
a
Ni773-***, Ni1373-***: *** indicates the reduction temperature of nickel oxide.
b
Exchange activity/BET surface area.
i
4-i

7
0
T. Osawa et al. / Journal of Molecular Catalysis A: Chemical 320 (2010) 68–71
As the crystallinity of NiO773 was low, the increase in the reduc-
tion temperature resulted in the change of the crystal structure,
i.e. the structure with moderately high Ni(1 0 0) ratio to that with
high Ni(1 1 1) ratio that is the stable face of fcc crystal. The ratio
of even higher index facets would also decrease by the sintering,
resulting in low exchange activities. This idea is supported by DFT
studies that decomposition of methane on nickel was higher on
Ni(2 1 1) than Ni(1 1 1) [23]. On the other hand, as the crystallinity
of Ni1373 was high, the crystal structure was not changed easily
and the decrease of the proportion of Ni(100) would be small. The
increase of the specific activity with the increase in the reduction
temperature could be attributed to the decrease of the high-index
facets by the sintering. The H/D exchange activity could be higher
on Ni(1 0 0) than on the high-index facets.
In order to investigate the activity on the high-index facets of
the crystal, K O promoted nickel was used for the H/D exchange
Fig. 1. Percentage of the exchanged species over Ni773 at various reduction tem-
peratures.
2
reaction. It is reported that when nickel was promoted by K O, K O
2
2
was localized on the step defect sites and it inhibited these sites
24]. It is also expected that the addition of K O to the nickel crys-
tal would result in the decrease of the crystallinity and accelerate
the change of crystal structure. Table 2 shows the H/D exchange
[
3.2. The exchanged species over the reduced nickel catalyst and
the K2O promoted reduced nickel catalyst
2
activity, BET surface area, and the specific activity of K O promoted
Figs. 1 and 2 show the percentages of all the exchanged species
over Ni773 and K-Ni773, respectively. For the reactions over Ni773,
CH2D2 was observed for the Ni773-573 that had relatively high
exchange activity. However, the percentage of CH2D2 decreased
with increases in the reduction temperatures and little CH2D2 was
observed over Ni773-623, -673, and -723. As for K-Ni773, low per-
centage of CH2D2 was observed for all the reduction temperatures.
Two mechanisms for the H/D exchange reaction of CH4 were
proposed, that is, a stepwise exchange and a multiple exchange
[12]. For the stepwise exchange, one hydrogen atom in methane
is exchanged at each residence on the surface. For the multi-
ple exchange, more than one hydrogen atoms in methane are
exchanged in one residence. It was reported that the activation
energy of CH2(adsorbed) to CH(adsorbed) in the multiple exchange
was lower than that of the other dissociation steps on the surface
and that further exchanges rather than the desorption to the gas
phase of CH2D2 occurred [9]. In this study, on the surfaces of Ni773-
623, -673, -723, and all the K-Ni773, CH2D2 was not observed. As
mentioned in 3.1, the H/D exchange reaction would occur mainly on
Ni(1 1 1) for these catalysts. The Ni(1 1 1) had low exchange activ-
ity and the multiple exchange would have occurred on this surface.
As Ni773-573 would have much Ni(1 0 0) and/or high-index facets
than Ni773-623, -673, -723, stepwise exchange would occur and
CH2D2 was observed.
2
nickel (K-Ni773 and K-Ni1373) prepared at reduction temperatures
from 573 K to 723 K. Comparing the specific activity of K-Ni773 and
that of Ni773 (Table 1), the activity of K-Ni773-573 and K-Ni773-
6
23 were lower than that of the corresponding Ni773. This could be
due to that Ni773-573 and Ni773-623 had the step defect sites that
were inhibited by K O and/or the addition of potassium facilitated
2
the structural change of nickel crystal. The inhibition of exchange
reaction by potassium could be explained by the increase of the dis-
sociation barrier of methane. DTF study indicated that increase of
the barrier was given by the interaction between the induced dipole
moment in the transition state of the dissociating methane and
the electrostatic field induced by the potassium adatoms [25]. On
the other hand, the activity of K-Ni773-673 and K-Ni773-723 were
as low as that of Ni773-673 and Ni773-723. It could be when the
reduction temperature was as high as 673 K and 723 K, the change
to Ni(1 1 1) from Ni(1 0 0) takes place irrespective of the presence of
potassium and the predominance of Ni(1 1 1) lowered the activity.
As for K-Ni1373 and Ni1373, the specific activity of K-Ni1373-573
and Ni1373-573 were almost the same. Taking into account that
they had similar BET surface area, the results would indicate that
Ni1373 did not have sites that were inhibited by K O and that the
2
promotion of structural change by the addition of potassium did
not observed at the reduction of 573 K. However, for the nickel
prepared by the reduction over 623 K, the specific activity signifi-
cantly decreased by the addition of potassium, probably because of
the decrease in the proportion of Ni(1 0 0) by the structural change
promoted by potassium.
Figs. 3 and 4 show the percentage of each exchanged species
over Ni1373 and K-Ni1373, respectively. Over Ni1373, CH2D2 was
observed at similar percentage to the other isotopologues at all the
Table 2
Exchange activity of K-Ni773 and K-Ni1373 prepared at various reduction
temperatures.
Catalysta
Exchange
activity/%
BET surface
Specific exchange
activity /%
area/m2
g
−1
b
K-Ni773-573
K-Ni773-623
K-Ni773-673
K-Ni773-723
6.9
2.4
1.8
1.7
8.3
5.3
2.8
2.5
0.84
0.45
0.65
0.69
K-Ni1373-573
K-Ni1373-623
K-Ni1373-673
K-Ni1373-723
45
30
17
9.1
4.0
4.1
2.5
1.7
11
7.1
6.9
5.5
a
K-Ni773-***, K-Ni1373-***: *** indicates the reduction temperature of nickel
oxide.
Fig. 2. Percentage of the exchanged species over K-Ni773 at various reduction tem-
peratures.
b
Exchange activity/BET surface area.

T. Osawa et al. / Journal of Molecular Catalysis A: Chemical 320 (2010) 68–71
71
of the step defect sites by potassium and/or the promotion of the
change of the nickel crystal structure. The present results support
the idea that Ni1373 with higher exchange activity than Ni773
could be attributed to a higher Ni(100)/Ni(1 1 1) ratio on the surface
for the former.
This study analyzed the exchanged species by the pulse of
methane in a D2 stream. Although the pulse technique is not a
steady-state reaction, the results of this study could be similar to
the results obtained in the early stage of the steady-state reaction,
because the analyses of each isotopologue in this study were car-
ried out by the integration of each peak. This would be supported by
the fact that the results obtained by Ni773-673 (the ratio of CH D2
2
was low and those of CH D and CD4 were high) were similar to
3
those reported using a conventional closed gas circulation system
[
12,16]. However, the pulse experiment would be a good technique
for analyzing the structure of a fresh catalyst surface, because it can
observe the fresh surface of the catalyst not adsorbed and covered
by the reactant.
Fig. 3. Percentage of the exchanged species over Ni1373 at various reduction tem-
peratures.
References
[
[
[
[
1] D.Yu. Murzin, P. Mäki-Arvela, E. Toukoniity, T. Salmi, Catal. Rev. 47 (2005) 175,
and references therein.
2] T. Osawa, T. Harada, O. Takayasu, Curr. Org. Chem. 10 (2006) 1513, and refer-
ences therein.
3] T. Sugimura, in: K. Ding, Y. Uozumi (Eds.), Handbook of Asymmetric Heteroge-
neous Catalysis, Wiley Verlag, 2008, p. 357, and references therein.
4] T. Harada, M. Yamamoto, S. Onaka, M. Imaida, H. Ozaki, A. Tai, Y. Izumi, Bull.
Chem. Soc. Jpn. 54 (1981) 2323.
[
[
5] Y. Nitta, F. Sekine, T. Imanaka, S. Teranishi, Bull. Chem. Soc. Jpn. 54 (1981) 980.
6] T. Osawa, S. Mita, A. Iwai, O. Takayasu, H. Hashiba, S. Hashimoto, T. Harada, I.
Matsuura, J. Mol. Catal. A: Chem. 157 (2000) 207.
[
7] T. Osawa, K. Koganei, N. Ito, M. Kitano, T. Harada, O. Takayasu, Catal. Lett. 100
(
2005) 189.
[
[
8] C. Kemball, Adv. Catal. 11 (1959) 223, and references therein.
9] C. Kemball, Catal. Rev. 5 (1972) 33, and references therein.
[
10] A. Ozaki, Isotopic Studies of Heterogeneous Catalysis, Kodansha Ltd. and Aca-
demic Press, Inc., 1977, and references therein.
[
[
[
11] J.G. Larson, W.K. Hall, J. Phys. Chem. 69 (1965) 3080.
12] J.M. Cece, R.D. Gonzalez, J. Catal. 28 (1973) 260.
13] J.S.J. Hargreaves, G.J. Hutchings, R.W. Joyner, S.H. Taylor, Appl. Catal. A: Gen.
Fig. 4. Percentage of the exchanged species over K-Ni1373 at various reduction
temperatures.
227 (2002) 191.
[
[
[
14] H.F. Leach, C. Mirodatos, D.A. Whan, J. Catal. 63 (1980) 138.
15] J.A. Dalmon, C. Mirodatos, J. Mol. Catal. 258 (1984) 161.
16] K. Otsuka, S. Kobayashi, S. Takenaka, J. Catal. 200 (2001) 4.
examined reduction temperatures. This could be also due to that
Ni1373 had more Ni(1 0 0) sites where stepwise exchange occurred.
On the other hand, as for the reaction over K-Ni1373, the percentage
[17] T.P. Beebe Jr., D.W. Goodman, B.D. Kay, J.T. Yates Jr., J. Chem. Phys. 87 (1987)
305.
18] T. Osawa, Y. Hamano, S. Saga, O. Takayasu, J. Mol. Catal. A: Chem. 298 (2009)
11.
[19] T. Osawa, K. Sugimoto, M. Miyamoto, O. Takayasu, React. Kinet. Catal. Lett. 90
2007) 93.
2
of CH D decreased with increases in the reduction temperatures,
2
2
[
because addition of potassium would promote the change of crystal
structure to decrease the proportion of Ni(1 0 0).
1
(
[
[
[
20] R.P. Furstenau, G. McDougall, M.A. Langell, Surf. Sci. 150 (1985) 55.
21] R. Lacmann, N. J. Miner. Abh. 122 (1974) 36.
22] T. Osawa, Y. Amaya, T. Harada, O. Takayasu, J. Mol. Catal. A: Chem. 211 (2004)
93.
4
. Conclusion
H/D exchange reactions of methane in a deuterium stream in a
[
[
23] H.S. Bengaard, J.K. Nørskov, J. Sehested, B.S. Clausen, L.P. Nielsen, A.M. Molen-
broek, J.R. Rostrup-Nielsen, J. Catal. 209 (2002) 365.
24] J.R. Rostrup-Nielsen, J. Sehested, Adv. Catal. 47 (2002) 65.
fixed bed flow reactor were studied over the reduced nickel cata-
lyst. We found Ni1373 had a higher exchange activity than Ni773.
And the promotion by potassium resulted in a significant decrease
in the exchange activity. This could be caused by the inhibition
[25] H.S. Bengaard, Ib Alstrup, Ib Chorkendorff, S. Ullmann, J.R. Rostrup-Nielsen, J.K.
Nørskov, J. Catal. 187 (1999) 238.