Furan decomposes on Pd(111) at 300 K to form H and CO plus C3H3, which can dimerize to benzene at 350 K

Full text of DOI:10.1021/ja9532042

J. Am. Chem. Soc. 1996, 118, 907-908
Furan Decomposes on Pd(111) at 300 K To Form H
907
and CO plus C
3 3
H , Which Can Dimerize to Benzene
at 350 K
Tracy E. Caldwell, Ihab M. Abdelrehim, and
Donald P. Land*
Department of Chemistry, UniVersity of California
DaVis, California 95616
ReceiVed September 19, 1995
Figure 1. Profile of the magnitudes for selected masses observed in
LITD/FT mass spectra, each obtained from a single laser shot at a
different spot, after warming of a 0.3 L exposure of furan on Pd(111)
to the temperatures indicated for 1 min. The plot shows a decay in
furan (m/z 68) followed by growth of CO (m/z 28) and then benzene
We report recent results on the products and mechanism of
the reaction of furan with a clean Pd(111) surface. The desire
to refine hitherto underutilized hydrocarbon sources requires
the removal of significant quantities of sulfur-, nitrogen-, and
(m/z 78). The benzene signal has been increased by a factor of 10.
1
oxygen-containing species. The same catalysts are used for
Several spectra are obtained at each temperature, and error bars are
displayed for the one data point with the largest scatter within each
set.
all three classes and typically consist of alumina-supported
sulfides of Mo or W.2 Oxygenated compounds are the most
prevalent hetero compounds in liquids derived from coal and
2,3
biomass, with furanic rings among the most important of these,
to yield CO. Direct formation of CO from furan has been
but removal of oxygen requires higher temperatures and
significantly more hydrogen, on a molar basis, than removal of
sulfur. The addition of a late transition metal, such as Co or
Ni, as a promoter has been reported to increase catalytic activity
13
observed on Mo(100) and Mo(110) surfaces, also.
For laser-induced thermal desorption (LITD), a Nd:YAG laser
is focused onto the surface. Adsorbed molecules are vaporized
from the surface as neutrals without substrate ablation. The
neutrals are then postionized by an electron beam and detected
by FTMS, yielding a complete mass spectrum of the postionized
species for each laser pulse. The desorbed species have been
2
,4
5
by an order of magnitude; however, little is known about the
surface chemistry of heterocycles on late transition metals. We
report here initial results on the chemistry of furan on Pd(111)
6
7
and relate the results to those for furan on Mo foils, Cu(110),
shown to be representative of the surface composition in the
8
9
10
Ag(110), and O/Ag(110), and for thiophene on Pd(111).
14,15
majority of cases.
Unlike conventional thermal desorption,
The experiments are performed in an ultrahigh-vacuum
chamber equipped for adsorption/desorption measurements
utilizing both conventional and laser-induced heating methods.
LITD rapidly achieves high surface temperatures and, under
these conditions, entropically favored processes (such as direct
16
desorption) can be favored.
1
1
Details of the apparatus have been published elsewhere. The
During the LITD experiments, the Pd sample is heated to a
selected temperature and allowed to equilibrate for 1 min before
the laser is fired. Figure 1 shows the signal intensities from
the LITD/FT mass spectra for furan (m/z 68), CO (m/z 28), and
benzene (m/z 78) as a function of sample temperature. Each
spectrum is obtained from a single laser shot at a different spot
on the surface. Several spectra are taken at each temperature,
and the signal magnitudes are averaged for Figure 1. The LITD/
FTMS signals are typically proportional to the surface concen-
+
Pd(111) crystal is cleaned by Ar bombardment and by heating
in oxygen. Furan (Aldrich Chemical Company 99+%) is
purified by freeze-pump-thaw cycles. The Pd sample is
cooled to 100 K and is then dosed with furan by backfilling
the chamber. Exposures have been corrected for ion gauge
sensitivities.
Temperature programed reaction (TPR) studies are performed
-
6
using a 0.3 L (L ) Langmuir ) 10 Torr‚s) exposure of furan
and a heating rate of 3 K/s. These studies use Fourier transform
mass spectrometry (FTMS); hence, a complete mass spectrum
11
trations for each species. These data clearly illustrate the loss
of furan accompanied by the growth of CO at 280-320 K
followed by benzene above 350 K. The peak at m/z 78 is
indicative of benzene and first appears at 350 K. No benzene
is observed in TPR, since, as previous studies show, low
(
m/z 10-650) of desorbing species is obtained every 2 s. The
only species observed are furan, H2, and CO (Tmax ) 265, 360,
and 450 K, respectively). To estimate the relative yield of
reversibly vs irreversibly adsorbed furan, the integrated areas
under the thermal desorption traces are calculated for CO (m/z
17
coverages of benzene decompose on Pd(111) before desorbing
and our LITD signals for benzene are indicative of very low
coverages. The benzene produced is only observed because
the laser heating favors direct desorption. Above 380 K, the
CO LITD signal decreases due to depletion by conventional
desorption.
By 320 K, the furan signal has almost completely disappeared
and CO formation is essentially complete. Presumably, by 320
K, C3H3 formation is also essentially complete; however, we
see no evidence for this species in LITD. This species is
probably too tightly bound or too labile to be desorbed intact,
even with rapid laser heating.
2
8) and furan (sum of m/z 39 and 68 areas). The furan total is
12
corrected using the ion gauge sensitivity (3.29) as an estimate
of the relative sensitivity by electron ionization in the mass
spectrometer. The ratio of the m/z 28 area relative to the
corrected total of m/z 28, 39, and 68 is approximately 0.4,
indicating that 40% of the initial coverage of furan decomposes
(
1) Speight, J. G. The Desulfurization of HeaVy Oils and Residua; Marcel
Dekker, Inc.: New York, 1981; vol. 4.
(
(
(
(
2) Furimsky, E. Catal. ReV. Sci. Eng. 1983, 25, 421.
3) Cadey, W. E.; Seelig, H. S. Ind. Eng. Chem. 1952, 44, 2636.
4) Rollmann, L. D. J. Catal. 1977, 46, 243.
5) Chary, K. V. R.; Rao, K. S. R.; Muralidhar, G.; Rao, P. K. Carbon
1
991, 29, 478.
(13) Tinseth, G. R.; Watson, P. R. J. Am. Chem. Soc. 1991, 113, 8549.
(14) Land, D. P.; Pettiette-Hall, C. L.; Hemminger, J. C.; McIver, R. T.,
Jr. Acc. Chem. Res. 1991, 24, 42.
(
2.
(
(
(
(
(
6) Gellman, A. J.; Neiman, D.; Somorjai, G. A. J. Catal. 1987, 107,
9
7) Sexton, B. A. Surf. Sci. 1985, 163, 99.
(15) Land, D. P.; Wang, D. T.-S.; Tai, T.-L.; Sherman, M. G.;
Hemminger, J. C.; McIver, R. T., Jr. In Lasers in Mass Spectrometry;
Lubman, D. M., Ed.; Oxford University Press: New York, 1990; pp 157.
(16) George, S. M. In InVestigations of Surfaces and InterfacessPart
A; 2nd ed.; Rossiter, B. W., Baetzold, R. C., Eds.; John Wiley & Sons:
New York, 1993; Vol. IXA, p 453.
8) Solomon, J. L.; Madix, R. J.; St o¨ hr, J. J. Chem. Phys. 1990, 94, 4012.
9) Crew, W. W.; Madix, R. J. J. Am. Chem. Soc. 1993, 115, 729.
10) Caldwell, T. E.; Abdelrehim, I. M.; Land, D. P. In preparation.
11) Land, D. P.; Abdelrehim, I. M.; Thornburg, N. A.; Sloan, J. T. Anal.
Chim. Acta 1995, 307, 321.
12) Bartmess, J. E.; Georgiadis, R. M. Vacuum 1983, 33, 149.
(
(17) Patterson, C. H.; Lambert, R. M. J. Phys. Chem. 1988, 92, 1266.
0
002-7863/96/1518-0907$12.00/0 © 1996 American Chemical Society

9
08 J. Am. Chem. Soc., Vol. 118, No. 4, 1996
Communications to the Editor
One might question whether laser heating drives some
Scheme 1
intermediate (such as a metallacycle with a C6H6 component)
to benzene before or during desorption. Because of its rapidity,
however, laser heating is likely to produce only subtle changes
in the desorbing species.18 Thus, changes in the LITD mass
spectra as surface temperature is varied can still be correlated
with changes in surface composition. Laser-driven recombina-
tion of C3H3 can be ruled out, since, if laser-driven recombina-
tion of two C3H3 species were an efficient process, benzene
desorption would have been observed at 320 K when most of
the C3H3 had formed. Clearly, something about the surface
composition changes between 320 and 350 K. Our data indicate
that C3H3 is stable for many minutes between 320 and 350 K
on Pd(111). At 350 K, these species are able to dimerize to
form a species similar, if not identical, to adsorbed benzene. It
is somewhat surprising that C3H3 would be stable and mobile
at these temperatures on group VIII metals, considering their
dehydrogenating nature. Typically, alkyl fragments are only
observed to recombine on coinage metals,19 although methyl
coupling to yield C2 species has been observed on other late
transition metal surfaces, including Pd, following the photodis-
sociation of methyl halides.2
performed on the same day, did show a strong signal for H2
from surface H using the same conditions described here.) The
results from TPR and LITD indicate that H2 evolution in TPR
of furan on Pd(111) is rate limited by R-CH bond breaking.
These experiments were carried out three times with identical
results, as illustrated in Scheme 1. In contrast to these Pd results,
7
other late transition metals, such as clean Cu(110) and Ag-
8
(
110), yield no decomposition products during TPR. It has
also been shown that Ni and Co are more effective than Zn in
2
5
the promotion of Mo and W HDO catalysts. It appears, then,
that the Co and Ni group metals should be expected to be more
effective promoters than the Cu and Zn group metals. However,
furan coadsorbed with oxygen on Ag(110) has been shown to
produce CO2 and H2O, both of which are evolved starting at
0-22
The data from LITD can be used to determine the relative
yields of furan, CO, and benzene by making assumptions about
the relative efficiency of laser-induced desorption for each
species. Assuming that the efficiency is the same for each, we
calculate that 60% of a 0.3 L exposure of furan decomposes to
give CO (plus H and, presumably, C3H3) and 2% of the furan
ends up as benzene. Thus, approximately 3% of the C3H3
species (assuming that CO and C3H3 are formed in equal
quantities) form benzene; the remainder decompose on the
surface. However, the benzene yields reported here are lower
limits, since furan and CO are likely to have higher efficiencies
for laser-induced thermal desorption than benzene. Both CO
and furan desorb at lower temperatures (430 and 280 K,
respectively, in TPR) than does benzene (520 K). In fact, as
mentioned above, low coverages of benzene decompose com-
pletely in TPR. Therefore, the efficiency of thermal desorption
in LITD is likely to be smallest for benzene.16 Furthermore,
the possibility of defect sites inducing the formation of benzene
in such low yields may be argued; however, it is difficult to
conceive how a defect would lead to benzene formation, since
defects are usually associated with higher binding energies and
enhanced decomposition.
9
about 300 K during TPR. At higher temperatures (520 K),
very small amounts of partial oxidation products, such as maleic
9
acid, benzene, and bifuran, are evolved. In these studies,
benzene formation was also attributed to the cycloaddition of
two C3H3 fragments, but only after oxidative activation of furan
decomposition. The Pd surface used in our study does not
require this highly oxidative environment.
FTMS monitors all masses from m/z 10 to 650 simulta-
neously, and all observed mass peaks can be attributed to furan,
CO, hydrogen, and benzene. No C4 species are observed. In
contrast, studies of furan HDO over supported Mo/C catalysts
5
show C4 species as the major products. HDS of thiophene on
6
Mo or W catalysts results in only C4 species, and our recent
results for thiophene decomposition on Pd(111) show exclu-
10
sively C4 products using LITD/FTMS and FT-TPR.
The
formation of furan from adsorbed oxygen and acetylene on Pd-
26
(
111) is also believed to occur via a C4H4 intermediate. The
exclusive production of CO, H, and the C3H3-derived benzene
from furan is, therefore, an unexpected result. However,
6
2
Gellman et al. and Furimsky both report that furan HDO on
Mo-based catalysts does produce some C3 product (primarily
propene); however, Furimsky’s data show that C4 species are
dominant by a factor of 3-5 over C3 products. Furthermore,
whereas HDO of furanic species over typical hydrotreating
catalysts requires temperatures in excess of 670 K, Pd catalyzes
rapid deoxygenation at 300 K. This may point to the funda-
mental action of late transition metal promoters in HDO
catalysts. Our results also show, however, that significant
amounts of carbonaceous material may be left behind on the
Pd surface. These species may act to poison further catalytic
reactivity unless removed, e.g., by hydrogen supplied to the feed.
+
H2 is monitored in a separate experiment performed with a
lower magnetic field strength. These experiments follow the
same procedures outlined above; however, a slightly higher
2
3
exposure (0.5 L) of furan is used (saturation requires ∼2 L ).
TPR (3 K/s) shows H2 evolution beginning at ∼300 K with the
desorption peak centered around 360 K. (Hydrogen on clean
Pd(111) desorbs with a peak centered at 310 K using a heating
2
4
rate of 25 K/s. ) Using LITD, no significant signal for H2 is
observed after heating a furan-dosed Pd(111) surface. (Experi-
ments on the dehydrogenation of cyclohexene on Pd(111),
Acknowledgment is made to the donors of the Petroleum Research
Fund, administered by the American Chemical Society for support of
this research. T.E.C. also acknowledges administrative aid from
Wellwood Elementary School and financial support from the Patricia
Roberts Harris Foundation. We thank Ted Picciotto and Alex Caldwell
for assistance in data collection.
(
18) Sherman, M. G.; Land, D. P.; Hemminger, J. C.; McIver, R. T., Jr.
Chem. Phys. Lett 1987. 137, 298.
(
(
19) Paul, A.; Bent, B. E. J. Catal. 1994, 147, 264.
20) Ali, A.-K. M.; Saleh, J. M.; Hikamat, N. A. J. Chem. Soc., Faraday
Trans. 1987, 83, 2391.
(
21) Zhou, Y.; Henderson, M. A.; Feng, W. M.; White, J. M. Surf. Sci.
1
989, 224, 386.
JA9532042
(
(
(
22) Solymosi, F.; Kiss, J.; Revesz, K. J. Phys. Chem. 1991, 94, 8510.
23) Ormerod, R. M. Ph.D. Dissertation, Cambridge, 1989.
(25) Patzer, J. F.; Montagna, A. A. Ind. Eng. Chem., Process Des. DeV.
1980, 19, 382.
(26) Ormerod, R. M.; Lambert, R. M. Catal. Lett. 1990, 6, 121.
24) Rucker, T. G.; Logan, M. A.; Gentle, T. M.; Muetterties, E. L.;
Somorjai, G. A. J. Phys. Chem. 1986, 90, 2703.