Carbon efficiency and the surface chemistry of the actinides: Direct formation of furan from acetylene over β-UO3

Article abstract of DOI:10.1006/jcat.2001.3483

Uranium oxide is a very good catalyst for oxidation reactions, and bare U atoms are very active for carbon-carbon coupling reactions. The direct oxidative coupling of two molecules of acetylene to furan (C₄H₄O) over the surface of pure polycrystalline β-UO₃ was presented. For comparison, only traces of furan were formed over α-U₃O₈ and none on UO₂ surfaces. Comparison to the reactions of other C₂ compounds (ethanol, acetaldehyde, and ethylene) over β-UO₃ showed the presence of two routes for making furan from C₂ compounds, i.e., via aldolization and via oxidative coupling.

Full text of DOI:10.1006/jcat.2001.3483

Journal of Catalysis 206, 155–158 (2002)
doi:10.1006/jcat.2001.3483, available online at http://www.idealibrary.com on
Carbon Efficiency and the Surface Chemistry of the Actinides:
Direct Formation of Furan from Acetylene over β-UO3
H. Madhavaram and H. Idriss1
Materials Chemistry, Department of Chemistry, The University of Auckland, Private Bag 92019, Auckland, New Zealand
Received September 13, 2001; revised November 27, 2001; accepted November 27, 2001
might help in designing a system where the dynamic ori-
This work presents the direct oxidative coupling of two molecules ents toward a desired chemical product. A simple way to
of acetylene to furan (C4H4O) over the surface of pure polycrys- view such a surface dynamic is via the “so-called” Mars–van
talline β-UO3. For comparison, only traces of furan are formed over
α-U3O8 and none on UO2 surfaces. A comparison to the reactions
Krevelen mechanism (4), where an adsorbed molecule
(
such as CO) might be oxidized by lattice oxygen anions to
of other C2 compounds (ethanol (H. Madhavaram and H. Idriss,
J. Catal. 184, 553 (1999)), acetaldehyde (H. Madhavaram and H.
Idriss, Catal. Today 63, 309 (2000)), and ethylene (H. Madhavaram
and H. Idriss, J. Catal. 184, 553 (1999); H. Madhavaram and H.
Idriss, Stud. Surf. Sci. Catal. 110, 265 (1997)) over β-UO3 indicates
the presence of two routes for making furan from C2 compounds:
one via aldolization and the other via oxidative coupling. ꢀc 2002
Elsevier Science (USA)
CO2. As a result the lattice is depleted (due to the formation
of oxygen vacancies) and gas-phase oxygen molecules are
required to restore these surface point defects. To achieve
this, at least two requirements have to be satisfied: (i) high-
2
−
surfaceoxygenmobility(oxidescontainingO withasmall
Madelung potential (5, 6)) and (ii) the presence of sur-
face metal cations having a coordinative unsaturation. For
4
5
+
Key Words: uranium oxide; oxidative coupling; actinides; furan. example, Ru (5c stands for five-fold coordinated) cations
c
of the rutile RuO2(110) single crystal showed activity for
4
6
+
CO oxidation, while Ru cations (fully coordinated) did
c
INTRODUCTION
not (7). Moreover, the presence of double vacancies may
accommodate two adsorbed molecules over one surface
cation. This may result in coupling reactions. The Ziegler–
Natta polymerization reaction is a chief example (8). The
best known example of an oxide surface is the “so-called”
The surface chemistry of the uranium oxide system is
one of the richest and most complex to investigate. This
is due to several factors including the large range of oxi-
dation states and the presence of multiple thermodynami-
cally stable phases; for example, UO3 exhibits seven poly-
crystalline structures and one amorphous structure (1). In
addition, phase transformation from one structure to an-
other requires very small activation energies making it a
real challenge to produce a pure form of the desired phase
4
+
ketonization reaction (9, 10) over Ti of the {114}-faceted
4
c
TiO2(001) single crystal in which the dissociative adsorp-
tion of carboxylic acids on the cation–anion centers results
in carboxylate and hydroxide formation—in general any
oxide having one coordinative unsaturation may catalyze
this step. However, the presence of two coordinative unsat-
4
4
+
(
probably with the exception of α-U3O8).
Consequently, investigating the surface bulk dynamic of
urations on the Ti cations is poised to accommodate two
c
carboxylate species that may couple together yielding one
molecule of ketone and one molecule CO2 (11).
these actinide oxides is an essential task. Yet, the surface re-
action of pure uranium oxide phases has received very little
attention. This work is motivated by the important techno-
logical applications of uranium oxide because it presents an
extension to our basic understanding of the surface reactiv-
ity of earlier transition metal oxides and might show new
reactions not shared by other oxides; see, for example, H2
production form water over PuO2 (2) or the formation of
cis-endiolates over organometallic compounds of uranium
from CO (3). The rich range of stoichiometry and structures
This work proves that we succeeded in combining
the aforementioned points. In other words, to make furan
from acetylene the presence of two coordinative unsatura-
tions and a high-surface oxygen anion mobility is required,
and it appears that the surface of β-UO3 provides this
requirement.
Furan and furan derivates, important chemicals in many
industrial processes and products such as those in phar-
maceuticals, herbicides, stabilizers, and fine chemistry, are
commercially manufactured from the decarbonylation of
furfural (12–14) and can also be made by oxidation of buta-
1
To whom correspondence should be addressed. Fax: 64 9 373 7422.
E-mail: h.idriss@auckland.ac.nz.
diene (15, 16). All these routes are based on C reactants.
4
155
0
021-9517/02 $35.00
c 2002 Elsevier Science (USA)
All rights reserved.
ꢀ

1
56
MADHAVARAM AND IDRISS
Yet, it was recently indicated that it would be challenging dation states of U cations on the surfaces of UO2, α-U3O8,
to produce furan from these C4 molecules in a one-step and UO3 to quantitatively determine the atomic concen-
x+
process (17). Moreover, while furan might be a reaction trations of U cations on the surface and near surface
intermediate in the production of maleic anhydride from of these oxides and can be found in Ref. (21). A Spec-
C4 hydrocarbons over VPO catalysts (18), a good part of tra Vision quadrupole mass spectrometer (scanning up to
maleic anhydride is converted back to furan! Thus, it is 200 amu) connected to a high-vacuum chamber (base pres-
−
7
clear, that achieving the direct synthesis of furan from non- sure ca. 10 Torr, working pressure during temperature-
−
6
oxygen-containing C2 molecules (such as acetylene) would programmed desorption (TPD) = 2–3×10 Torr) and mul-
have a crucial impact on catalysis by oxide materials.
tiplexed to a PC was used. The ramping rate was kept at
−
1
0
.25 Ks . The cyclic rate was ca. 12 amu/3 s. Thus, spec-
EXPERIMENTAL
tral resolution was close to 1 K. β-UO3 powder, 50 mg,
was loaded into a fixed-bed quartz reactor and heated to
β-UO3 was prepared from a uranium nitrate solution in 523 K for 2 h prior to dosing with C H . Ultrapure C H
2
2
2
2
which U(OH)6 was precipitated with ammonia solution at was used at concentration of 10 ppm (Ar balance). The dos-
−
7
−6
pH 9. The precipitate was filtered and washed in deion- ing amounts varied from 3.5 × 10 to 4.2 × 10 mol (per
ized water, dried at 373 K overnight, and calcined at 773 K 50 mg). Quantitative reaction products were analyzed fol-
for 18 h. Preheating the furnace to 773 K was essential in lowing the method described previously (24). The reactor
−
3
preventing the formation of the more stable γ -UO3 and α- pressure during TPD was ca. 10 Torr at a pumping speed
3
U3O8 phases. An X-ray diffraction (XRD) spectrum and a of ca. 10 m /h.
scanning electron microscopy image are presented in Fig. 1.
Most of the lines in the XRD spectrum can be attributed to
the β-UO3 phase indicating a high purity of the oxide; the
relatively high signal-to-noise ratio shows a high crystalline-
RESULTS AND DISCUSSION
There are two main differences between β-UO3 and
to-amorphous ratio. Irregularly shaped particles of a size α-U3O8: they are structural and electronic. The unit lat-
6
+
less than 500 nm containing some nanopores (<10 nm) are tice of β-UO3 is composed of five U cations (Fig. 2): three
shown. UO2 was formed from reduction of α-U3O8 by H2 of them are sixfold coordinated to oxygen anions (i.e., con-
6
6
+
above 800 K for at least 4 h. X-ray photoelectron spec- taining two coordinative unsaturations, U ) and the re-
c
troscopy (XPS) was previously used to investigate the oxi- maining two are in a pentagonal bipyramid structure (i.e.,
2
FIG. 1. XRD of pure β-UO3. An SEM (Ep = 5 kV) image of a carbon-coated β-UO3 sample is inserted. BET surface area = 15 m /g.

FORMATION OF FURAN FROM ACETYLENE OVER β-UO3
157
FIG. 2. A schematic representation of the structure of β-UO3. Two
molecules of acetylene at the proximity of atom 2 (sixfold coordinated U
ion) are also shown. The arrows indicate a plausible adsorption site. The
structure is adapted from Ref. (19).
each U cation is sevenfold coordinated to oxygen anions)
(
19). On the other hand, all U cations of α-U3O8 are sev-
enfold coordinated to oxygen anions, also in a pentagonal
bipyramid structure. While U cations in α-U3O8 are in +6
and +4 oxidation states (in a theoretical ratio of 2 to 1–
although we have shown, by detailed analysis of the XPS
U4f region that the real surface of U3O8 may be composed
FIG. 3. Temperature-programmed desorption after acetylene adsor-
ption at room temperature over β-UO3. The insert presents the computed
furan desorption at a different initial dosing of acetylene. m/e 25 is mainly
an acetylene fragment.
of up to 83% of U6 cations (20)), the surface of β-UO3 is
+
composed exclusively of U6 cations (21).
+
Figure 3 presents TPD of acetylene (total dosing = 3.5 ×
−
7
1
0
mol) over β-UO3. Three desorption masses are shown:
We have previously observed the formation of furan
unreacted acetylene (C2H–, m/e 25), furan molecular from ethanol (20), acetaldehyde (22), and ethylene (20, 23).
weight (C4H4O, m/e 68), and furan main fragmentation Although the highest yield for furan (prior to this work)
(
C3H3, m/e 39). The very low desorption peak tempera- was observed from acetaldehyde, considerable forma-
ture at 390 K indicates that the reaction may have been tion of crotonaldehyde (formed by β-aldolization of two
completed during the adsorption process at room tem- molecules of acetaldehyde) was also observed. (Furan is
−
6
perature. Increasing the dosing amount to 2 × 10 and easily formed from crotonaldehyde.) Moreover, acetalde-
−6
4
.2 × 10 mol resulted in the desorption of a small sec- hyde was formed from ethylene as well as from ethanol
ond peak of furan at ca. 450 K. However, the amount of
unreacted acetylene remained negligible. These observa-
tions might indicate that either the surface is saturated or
large amounts of furan are formed at room temperature—
while dosing—this formation prevented further adsorption
of acetylene molecules. Similar reactions were carried out
over α-U3O8, but only traces of furan were observed. These
trace amounts are most likely formed over a discrete phase
of UO3 at the surface of U3O8 (not observed by XRD).
The average furan carbon yield, from four different runs at
TABLE 1
Carbon Selectivity of Furan during TPD of Four Different C2
a
Molecules over β-UO3
Furan
selectivity
(C%)
TPD, peak
temperature
(K)
Other
products
Reactant
References
23
CH2 == CH2
5–6
CH3CH2OH
CH3CHO
510
−
7
3
.5×10 mol/50 mg, was found to be equal to 83% which is
CH3CH2OH
CH3CHO
20–25
CH3CHO
510
420
20
22
−
6
−7
2
equal to ca. 3×10 mol/g (or 2×10 mol/m ; BET surface
30–40b
C4H6O,b
2
area ≈15 m /g of UO3). Catalyst regeneration could be ob-
crotonaldehyde
CHCH
≈100
—
390
This work
tained after heating the used oxide in 30 atm of O2 at 673 K
for at least 2 h. No other products were observed (with the
exception of CO2, mainly formed from bulk transformation
of UO3 to U3O8 at T > 600 K during TPD).
a
The activity to furan is in the following order: CH2CH2 <
CH3CH2OH < CH3CHO < CHCH.
b
Selectivity changed with coverage.

1
58
MADHAVARAM AND IDRISS
REFERENCES
1. Colmenares, C. A., Prog. Solid State Chem. 15, 257 (1984).
2. Haschke, J. M., Allen, T. H., and Morales, L. A., Science 287, 285
(
2000).
3
4
5
6
7
. Kahn, B. E., and Rieke, E. D., Chem. Rev. 88, 733 (1988).
. Mars, P., and van Krevelen, D. W., Chem. Eng. Sci. 3, 41 (1954).
. Idriss, H., and Barteau, M. A., Adv. Catal. 45, 261 (2000).
. Chong, S. V., Griffiths, T. R., and Idriss, H., Surf. Sci. 444, 187 (2000).
. Over, H., Kim, Y. D., Seitsonen, A. P., Wendt, S., Lundgren, E.,
Schmid, M., Varga, P., Morgante, A., and Ertl, G., Science 287, 1474
(
2000).
8
9
. Boor, J., Jr., “Ziegler–NattaCatalystsandPolymerizations.”Academic
Press, New York, 1979.
. Kim, K. S., and Barteau, M. A., J. Catal. 125, 353 (1990).
1
0. Idriss, H., Kim, K. S., and Barteau, M. A., Stud. Surf. Sci. Catal. 64,
SCHEME 1. TwodistinctroutesformakingfuranfromC2 compounds
327 (1991).
over β-UO3. Route 1. Coupling of two molecules of ethylene (or two
molecules of acetylene) followed by oxidation to furan. Route 2. Aldoliza-
tion of acetaldehyde, followed by C–O bond formation (and cyclization).
1
1
1. Barteau, M. A., Chem. Rev. 96, 1413 (1996).
2. Othmer, K., “Ecyclopedia of Chemical Technology,” third edition,
Vol. 11. Wiley, New York, 1982.
1
1
1
3. Banford, W. H., and Manes, M. M., U.S. Patent 2 846 449, 1958.
4. Manly, D. G., U.S. Patent 3 021 342, 1962.
5. Garnett, D. I., and Peterson, M. L., U.S. Patent 4 172 838, 1978.
adding a furthercomplication tothe scheme. The absence of
acetaldehyde and crotonaldehyde, during acetylene-TPD, 16. Farha, F. E., Jr., U.S. Patent 3 906 009, 1975.
clearly suggests the presence of a direct route for making 17. Wilderberg, M. D., Mallat, T., Gobel, U., and Baiker, A., Appl. Catal. A
6
8, 69 (1998).
furan from acetylene.
1
8. Taufiqyaq, Y. H., Sakakini, B. H., and Waugh, K. C., Catal. Lett. 46,
73 (1997).
9. Debets, P. C., Acta Crystallogr. 21, 589 (1966).
Table 1 compares data for furan formation from the
four different C2 molecules, while Scheme 1 shows the two
routes as evidenced from TPD studies.
2
1
20. Madhavaram, H., and Idriss, H., J. Catal. 184, 553 (1999).
1. Madhavaram, H., Buchanan, P., and Idriss, H., J. Vac. Sci. Technol. A
2
1
5, 1685 (1997).
CONCLUSION
2
2
2. Madhavaram, H., and Idriss, H., Catal. Today 63, 309 (2000).
3. Madhavaram, H., and Idriss, H., Stud. Surf. Sci. Catal. 110, 265
(1997).
4. Idriss, H., Kim, K. S., and Barteau, M. A., J. Catal. 139, 119 (1993).
5. Hutchings, G. J., Heneghan, C. S., Hudson, I. D., and Taylor, S. H.,
Nature 384, 341 (1996).
6. Nozaki, F., and Ohki, K., Bull. Chem. Soc. Jpn. 45, 3473 (1972).
7. Heinemann, C., Cornehl, H. H., and Schwartz, H., J. Organomet.
Chem. 501, 201 (1995).
Uranium oxide is considered to be a very good catalyst
for oxidation reactions (25, 26), and bare U atoms are very
active for carbon–carbon coupling reactions (27). This work
demonstratesthatbothroutescanbesuccessfullycombined
on β-UO3 as evidenced by the direct synthesis of furan
2
2
2
2
(
C4H4O) from acetylene (C2H2).