CuO - Activated carbon catalysts for methanol decomposition to hydrogen and carbon monoxide

Article abstract of DOI:10.1139/v03-146

A comparison of the abilities of CuO - activated carbon catalysts, prepared by different copper precursors and preparation techniques, in the methanol decomposition reaction to carbon monoxide and hydrogen, was undertaken. Higher catalytic activity and stability are found for the catalysts obtained from an ammonia solution of copper carbonate. The nature of the catalytic active complex in the samples is also discussed.

Full text of DOI:10.1139/v03-146

1
096
CuO – activate d carbon catalys ts for me thanol
de compos ition to hydroge n and carbon monoxide
Ta nya Ts onc he va , Ra dos tin Nic kolov, Sve tos la va Va nkova , a nd
Dim ita r Me ha ndjie v
Abstract: A comparison of the abilities of CuO – activated carbon catalysts, prepared by different copper precursors
and preparation techniques, in the methanol decomposition reaction to carbon monoxide and hydrogen, was undertaken.
Higher catalytic activity and stability are found for the catalysts obtained from an ammonia solution of copper carbon-
ate. The nature of the catalytic active complex in the samples is also discussed.
Key words: methanol decomposition, CuO – activated carbon catalysts, catalytic active complex.
Résumé : On a réalisé une comparaison de la réaction de décomposition du méthanol en monoxyde de carbone et en
hydrogènes sous l’action de divers catalyseurs de CuO sur des charbons activés, préparés à partir différents précurseurs
de cuivre et par diverses techniques. On a observé que l’activité catalytique et la stabilité des catalyseurs étaient plus
élevées pour les catalyseurs obtenus à partir de solutions ammoniacales de carbonate de cuivre. On discute aussi de la
nature du complexe catalytique actif dans les échantillons.
Mots clés : décomposition du méthanol, catalyseurs de CuO sur du charbon activé, complexe catalytique actif.
[
Traduit par la Rédaction] Tsoncheva et al. 1100
Introduc tion
Sample Cu-AC-N
The active phase was deposited by four-fold spraying of
the support with an aqueous solution of Cu(NO ) .
In view of the application of methanol as an alternative
ecological fuel (1–4), its catalytic decomposition to CO and
hydrogen is a process of considerable interest. Some copper-
containing materials appear to be promising catalysts (5–7)
for this purpose. In our previous investigations we consid-
ered the favorable effect of activated carbon used as a sup-
port for the preparation of efficient catalysts for methanol
decomposition (8–13). The role of the copper precursor in
the catalytic activity and selectivity of these materials has
been also reported (10, 12, 13). In this aspect, we would like
to draw attention to some peculiarities in methanol decom-
position on copper-oxide-supported catalysts obtained by
different preparation techniques and to look for a correlation
with the nature of the catalytic active complex (CAC) in
them (14).
3
2
Sample Cu-AC-A1
The active phase was deposited by four-fold spraying of
the support with 2.4 wt% basic copper carbonate in ammo-
nia water and ammonium carbonate, pH = 9.9.
Sample Cu-AC-A2
The active phase was deposited by single spraying of the
support with 2.4 wt% basic copper carbonate in ammonia
water and ammonium carbonate, pH = 9.9.
After each deposition, the samples were kept for 24 h in a
closed volume and subjected to drying for 1 h under vacuum
at 298 K and then for 1 h in air at 313 K. The samples were
decomposed at 480 K in air for 1 h.
Expe rim e nta l
Methods and apparatus
Copper content in the samples was determined by atomic
absorption spectrometry (Pye Unicam SP 90B spectrometer).
X-ray diffraction powder patterns were obtained on a
Phylips APD-15 diffractometer equipped with a Cu Kα radi-
ation source and graphite monochromator. XP and Auger
spectra were registered using an ESCA LAB Mk II (VG Sci-
entific). The ESR spectra were obtained in an ERS-220
Materials
This investigation encompasses three copper-containing
catalysts supported on activated carbon, a commercial
product (15). The specific surface area of the support is
2
–1
2
–1
1
060 m g ; the pore volume is 0.91 cm g . The catalysts
were prepared by an incipient wetness method as follows:
Received 29 March 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 30 September 2003.
1
T. Tsoncheva. Institute of Organic Chemistry, Bulgarian Academy of Sciences, Sofia 1113, Bulgaria.
R. Nickolov and D. Mehandjiev. Institute of General and Inorganic Chemistry, Bulgarian Academy of Sciences, Sofia 1113,
Bulgaria.
S. Vankova. Department of Chemistry, “St. Kl. Ohridski”, Sofia 1000, Bulgaria.
1Corresponding author (e-mail: ormm@orgchm.bas.bg).
Can. J. Chem. 81: 1096–1100 (2003)
doi: 10.1139/V03-146
© 2003 NRC Canada

Tsoncheva et al.
1097
spectrometer (Germany) at 298 K. The magnetic measure-
ments were performed with a Faraday-type magnetic balance
in a temperature range of 298–473 K.
Table 1. Copper content, specific surface area, and magnetic sus-
ceptibility of the initial samples obtained by different preparation
techniques.
The catalytic experiments were carried out in a flow reac-
tor with partial pressure of methanol at 1.57 kPa and WHSV
Cu content
(wt%)
BET
(m g )
χg
2
–1
6
Samples
(g ions × 10 )
–
1
at 1.5 h in a thermo-programmed regime with a heating
–
1
AC
0.0
4.0
5.0
5.0
1060
960
958
657
–0.6
19.0
7.5
rate of 1 K min . A standard catalytic test with time on
stream (about 200 min) at selected temperatures was also
performed. The on-line gas chromatographic analysis was
performed.
Cu-AC-N
Cu-AC-A1
Cu-AC-A2
18.4
Fig. 1. Total conversion vs. temperature on various copper-
loaded samples.
Re s ults a nd dis c us s ion
In Table 1 data of the copper content and specific surface
area for the initial copper-oxide-supported materials are pre-
sented. The magnetic susceptibility of the active phase, cal-
culated as CuO at 298 K, is also given. An essential
decrease in the BET surface area for all copper-loaded sam-
ples in comparison with the parent activated carbon is ob-
served. Therefore, a highly dispersed copper species, located
mainly within the pores of the support, seems to be formed.
In Fig. 1 the temperature dependencies of methanol con-
version for the copper oxide – activated carbon samples, ob-
tained by different techniques, are presented. The main
registered products are CO and H . Methyl formate (MF) is
2
also found at lower temperatures. A well-defined tendency
for increased conversion with increasing temperatures is reg-
istered for Cu-AC-A1. In this case the conversion remains
unchanged (about 100%) above 530 K. On the contrary, a
maximum in the conversion curve at 510 K for Cu-AC-N is
observed. Further increase in methanol conversion above
netic susceptibilities, per g mol Cu2+ ions in the initial sam-
ples, are rather different (Table 1). The lower magnetic sus-
ceptibility for Cu-AC-A1 could be an indication of the
presence of copper species in a lower oxidative state in this
sample, since their effective magnetic susceptibility is zero
(17). Indeed, only in the case of the initial Cu-AC-A1 does
6
00 K is also found. The catalytic activity for Cu-AC-N is
lower compared with that for Cu-AC-A1 in the whole tem-
perature interval that was investigated. The conversion curve
for Cu-AC-A2 is close to that for Cu-AC-A1, up to 600 K,
but a sharp decrease above this temperature is registered. So,
a favourable effect of active phase deposition from ammonia
solution is concluded. However, the preparation technique
that includes a four-fold spraying of the support with copper
solution facilitates the formation of more stable catalysts.
As far as the fact that the activated carbon does not ex-
hibit a catalytic activity in methanol decomposition on its
own (not shown), some significant differences in the copper
state for the initial catalysts could be assumed. Further
changes in the copper state during the catalytic process
could also be expected.
1
+
XPS analysis show the presence of Cu ions (Table 2). So,
the participation of copper species in different oxidative
2
+
2+
2+
1+
states (Cu , Cu –Cu , and Cu ) in the CAC Cu-AC-A1
2
+
2+
2+
could be assumed. On the contrary, Cu and Cu –Cu
ions are exclusively present in the CAC of the initial
Cu-AC-N. In this relation, the observed higher activity for
Cu-AC-A1 (Fig. 1) could be ascribed to the presence of
1
+
Cu ions in the initial CAC. The favourable role of the am-
monia copper precursor for these types of CAC formations
in the process of catalyst preparation is considered. This as-
sumption is also proved by the higher turnover frequencies
(TOF), calculated on the basis of the initial conversion at
different temperatures, for Cu-AC-A1 in comparison with
Cu-AC-N (Fig. 4). However, the similarities of TOF values
for both samples above 530 K could be evidence for some
essential changes in CAC for Cu-AC-N during the reaction.
This conclusion is confirmed by the XPS spectra for Cu-AC-N,
To describe the peculiarities of the active-phase state on
the investigated catalysts, a complex study by different
methods was performed. The XRD spectra for the initial
samples are identical, and CuO is the only phase observed
(
Fig. 2, curve 1). However, two ESR signals are also found
in all investigated samples (Fig. 3), in spite of the fact that
CuO does not register on ESR spectra. This is also in accor-
dance with the results in ref. 16. A sharp ESR signal, typical
0
2+
1+
where Cu , together with Cu and Cu , is found (Table 2).
2
+
2+
0
of single Cu ions, is registered. The presence of Cu ions
Intensive signals, typical of Cu , are also observed in the
2
+
2+
interacting with each other (Cu –Cu ) and characterized
with a broader ESR signal are found as well. So, despite the
nature of the used copper precursor, the CAC in the initial
sample contains predominantly Cu2 ions, differing in coor-
dination and type of interaction. This conclusion is proved
by the magnetic measurements as well. The calculated mag-
XRD spectrum for this sample (Fig. 2, curve 2). On the con-
0
trary, no Cu is found for the used Cu-AC-A1. Hence, a
more stable CAC in Cu-AC-A1 with respect to the reaction
medium influence is most probably formed.
The higher stability of CAC in the ammonia-obtained
sample compared with that in Cu-AC-N could also be
+
©
2003 NRC Canada

1
098
Can. J. Chem. Vol. 81, 2003
Table 2. Binding energies of Cu 2p3/2 and O 1s and the kinetic energy of the Cu L M4,5M4,5 peak (eV) of the initial (I) samples and
3
those used in the catalytic test at 700 K (T).
Cu 2p3/2
Cu LMM Auger
O 1s
Samples
CuO
Cu O
—
933.6
932.5
Cu
CuO
Cu O
—
917.2
917.1
Cu
CuO
Cu O
—
530.3
530.9
Cu(OH)2
531.6
531.9
2
2
2
Cu-AC-N (I)
Cu-AC-A1 (I)
Cu-AC-N (T)
934.1
934.0
933.6
917.9
917.8
917.9
529.8
530.0
530.4
932.0
919.0
531.7
Fig. 2. XRD patterns of the initial Cu-AC-N (curve 1) and that used in the catalytic test at 700 K (curve 2).
Fig. 3. ESR spectra of activated carbon (curve 1) and the initial
Cu-AC-N (curve 2).
Fig. 4. Temperature dependences of TOF for Cu-AC-A1 and Cu-
AC-N.
©
2003 NRC Canada

Tsoncheva et al.
1099
Fig. 5. Evolution of the conversion and product distribution with time on stream on Cu-AC-N at 473 K (a) and 551 K (b).
Fig. 6. Evolution of the conversion and product distribution with time on stream on Cu-AC-A1 at 490 K (a) and 575 K (b).
proved by the catalytic experiments in isothermal conditions
Figs. 5 and 6). For Cu-AC-N, the conversion remains un-
temperatures are needed. In relation to this, further investi-
gations are in progress.
(
changed only at lower temperatures (Fig. 5a). However, it
gradually decreases at 550 K (Fig. 5b), and the catalyst prac-
tically loses its catalytic activity after about 5 h. On the con-
trary, no essential changes with time on stream are
registered for Cu-AC-A1 even at 575 K (Fig. 6).
Re fe re nc e s
1
. L. Pettersson and K. Sjöström. Combust. Sci. and Tech. 80,
265 (1991).
2
. À. Yildiz and K. Pekmez. In Hydrogen energy system. NATO
ASI Series. Ser. E. Vol. 295. Edited by Y. Yürüm. Kluwer Ac-
ademic Publshers, Netherlands. 1995. p. 195.
Conc lus ions
3
. W.H. Cheng and H.H. Kung (Editors). In Methanol production
and use. Marcel Dekker, New York. 1994. Chap 1.
. W.H. Cheng. Acc. Chem. Res. 32, 685 (1999).
. M. Clement, Y. Zhang, D.S. Brands, E.K. Poels, and A. Bliek.
Stud. Surf. Sci. Catal. 130, 2123 (2000).
In comparison with the nitrate-obtained, copper-loaded
activated carbon materials, the catalysts prepared from the
ammonia precursor exhibit higher catalytic activity during
methanol decomposition to CO and hydrogen. In this case,
the preparation technique that includes four-fold spraying of
the support with the precursor is favoured as producing the
catalyst with the best stability in this reaction. It results in
easier formation and better stabilization of the CAC, con-
4
5
6
7
. A. Fisher and A.T. Bell. J. Catal. 184, 357 (1999).
. D.B. Clarke, D.-K. Lee, M.J. Sandoval, and A.T. Bell. J. Catal.
1
50, 81 (1994).
8
9
. T. Tsoncheva, R. Nickolov, Ya. Neinska, Chr. Minchev, and D.
Mehandjiev. Bulg. Chem. Commun. 32, 218 (2000).
. R. Nickolov, T. Tsoncheva, and D. Mehandjiev. Fuel, 81, 203
(2002).
1
+
taining Cu ions on the support in the process of catalyst
preparation. Indeed, the nature of the CAC is only one as-
pect of the complex problem of optimization of catalysts.
Higher catalytic activity, selectivity, and stability at low
10. T. Tsoncheva, R. Nickolov, and D. Mehandjiev. React. Kinet.
Catal. Lett. 72, 383 (2001).
©
2003 NRC Canada

1
100
Can. J. Chem. Vol. 81, 2003
1
1. T. Tsoncheva, R. Nickolov, and D. Mehandjiev. In Proceedings
of the 9th International Symposium of Heterogeneous Cataly-
sis, Edited by L. Petrov, Ch. Bonev, and G. Kadinov. Bulgaria.
15. V. Minkova, M. Razvigorova, M. Goranova, L. Ljutzakanov,
and G. Angelova. Fuel, 70, 713 (1991).
16. N. Stankova, M. Khristova, and D. Mehandjiev. J. Colloid In-
terface Sci. 241, 439 (2001).
2
000. p. 755.
1
1
1
2. T. Tsoncheva. C. R. Acad. Bulg. Sci. 54, 59 (2001).
3. T. Tsoncheva. C. R. Acad. Bulg. Sci. 54, 65 (2001).
4. D. Mehandjiev. In Proceedings of the 9th International Sympo-
sium of Heterogeneous Catalysis. Edited by L. Petrov, Ch.
Bonev, and G. Kadinov. Bulgaria. 2000. p. 19.
17. D. Mehandjiev and S. Angelov. In Magnetochemistry of the
solid state. Nauka I Izkustvo, Sofia, Bulgaria. 1979. p. 134.
©
2003 NRC Canada