Evidence for two competing mechanisms for n-butane oxidation catalyzed by vanadium phosphates [12]

Full text of DOI:10.1021/ja9929180

11024
J. Am. Chem. Soc. 1999, 121, 11024-11025
Evidence for Two Competing Mechanisms for
n-Butane Oxidation Catalyzed by Vanadium
Phosphates
Bin Chen and Eric J. Munson*
Department of Chemistry, UniVersity of Minnesota
207 Pleasant St. SE, Minneapolis, Minnesota 55455
ReceiVed August 12, 1999
The conversion of n-butane to maleic anhydride by vanadium
phosphorus oxide (VPO) catalysts is recognized as one of the
most complex selective oxidation reactions used in industry today,
as it involves the abstraction of eight hydrogen atoms and insertion
of three oxygen atoms.1-3 The reaction is proposed to proceed
by either an alkoxide or an olefinic intermediate route. Under
standard operating conditions no intermediates have been observed
in the reaction products, so the proposed mechanisms for this
reaction are based upon studies of the kinetics of the reaction
and the observation of anticipated intermediates using nonstandard
conditions.3,4 In the olefinic route, the reaction is proposed to
proceed as shown:
This route is based on the observation of butenes, butadiene, and
furan at very low concentrations (<2%) in the reaction products
using low oxygen and high n-butane concentrations on catalysts
with average vanadium oxidation states ∼4 or below.2,5,6 For this
reason the role of the proposed intermediates in the reaction
mechanism has been questioned.2,3 However, all of the intermedi-
ates react on VPO to produce maleic anhydride significantly faster
than does n-butane. This has been used to explain why it is
difficult to observe these intermediates in the reaction products.
In this paper we present the first direct evidence from isotopic
labeling studies that the mechanism for n-butane oxidation
proceeds via two different routes.
Figure 1. 13C NMR spectra of the reaction products of n-butane or
butadiene on VPO 4.56. (a) 1,4-13C n-butane reacting at 380 °C and (b)
same as part a, except the ampule was evacuated prior to sealing. Arrows
indicate 13C satellites from 2,3 carbons in 1,4-13C maleic acid. (c)
Uniformly 13C-labeled n-butane reacting at 380 °C. (d) 2-13C butadiene
reacting at 330 °C. In all spectra dashed and solid lines indicate gas and
liquid peaks, respectively.
We have investigated the conversion of n-butane to maleic
anhydride by reacting selectively 13C-labeled n-butane on VPO
catalysts.7 The products were collected, sealed in a glass ampule,
and analyzed using 13C NMR spectroscopy.8 Figure 1a shows the
13C NMR spectrum of the reaction products collected after flowing
1,4-13C-labeled n-butane over VPO catalyst at 380 °C that has
an average vanadium oxidation state of +4.56 (denoted VPO
4.56). No molecular oxygen was present in this experiment. The
spectrum contains large peaks for unreacted gas and liquid phase
n-butane, as well as CO, CO2, ethylene, maleic acid, fumaric acid,
and methanediol.9
All of the proposed reaction mechanisms indicate that 1,4-13C
n-butane will produce 1,4-13C maleic acid (peak at 170 ppm).
While most of the maleic acid is labeled in the 1,4 positions,
there is a peak corresponding to maleic acid in which the labeled
carbons are in the 2,3 positions (132 ppm). The assignment of
the 132 ppm peak was confirmed by reacting 1,4-13C n-butane
on VPO 4.56 at 380 °C and evacuating the ampule prior to
analysis (Figure 1b). All gaseous and volatile species such as CO,
CO2, and unreacted n-butane were removed, leaving only peaks
for maleic acid, fumaric acid, and methanediol in the spectrum.
* To whom correspondence should be addressed.
(1) Centi, G. Catal. Today 1993, 16, 1-153.
(2) Centi, G.; Trifiro, F.; Ebner, J. R.; Franchetti, V. M. Chem. ReV. 1988,
88, 55-80.
(3) Cavani, F.; Trifiro, F. Catalysis 1994, 11, 246-317.
(4) Zang-lin, Y.; Forissier, M.; Sneeden, R. P.; Vedrine, J. C.; Volta, J. C.
J. Catal. 1994, 145, 256-266.
(5) Kubias, B.; Rodemerck, U.; Zanthoff, H. W.; Meisel, M. Catal. Today
1996, 32, 243-253.
(8) The VPO catalyst (∼0.2 g) was placed in the center of a glass reactor.
The reactor was attached to a high-vacuum line and evacuated to less than 5
× 10-3 Torr. The reactants (butane or butadiene, ∼50 µmol) were condensed
in one end of the reactor with liquid N2. After the catalyst was preheated in
a furnace to the desired temperature, liquid N2 was shifted to the other end of
the reactor. After the products were collected, the glass ampule containing
the reaction products and the remaining reactant was flame sealed for NMR
analysis. 13C NMR spectra were acquired on a home-built spectrometer
operating at 50.197 MHz. Single-pulse 13C excitation (Bloch decay) with proton
decoupling (pulse delay ) 1-3 s, 10000-100000 transients) was used to
obtain all of the spectra shown.
(9) The assignment of the peaks at 84, 132, and 133 ppm to methanediol,
maleic acid, and fumaric acid, respectively, was made by comparison to 13C
chemical shifts and coupling constants in solution and by determining the
number of attached protons from the proton coupled spectrum. From this
information the peaks for both fumaric and maleic acid could be assigned
unambiguously.
(6) Rodemerck, U.; Kubias, B.; Zanthoff, H. W.; Wolf, G. U.; Baerns, M.
Appl. Catal. A-Gen. 1997, 153, 217-231.
(7) Three VPO catalysts used in this study were prepared from two different
precursors. VPO 3.92 was prepared by mixing a stoichiometric amount of
V2O5 and H3PO4 (85%) in ethanol and refluxing for 16 h. The resulting
precursor was blue and identified by XRD as VOHPO4‚0.5H2O. The precursor
was then calcined in nitrogen at 550 °C for 3 h. The XRD pattern for the
resulting catalyst was consistent with (VO)2P2O7. VPO 4.95 was prepared by
calcining the precursor at 450 °C for 24 h in air. For VPO 4.56, the procedure
described by Cavani et al. (Cavani, F.; Centi, G.; Trifiro`, F. Appl. Catal. 1984,
9, 191-202) was used. The precursor was then calcined at 380 °C for 3 h in
air. The average valence state of vanadium in the VPO catalysts was
determined using the potentiometric method described by Niwa and Murakami
(Niwa, M.; Murakami, Y. J. Catal. 1982, 76, 9-16). 1,4- and fully 13C-labeled
n-butane were obtained from Isotec. 2-13C-labeled butadiene was obtained
from Cambridge Isotope Laboratories.
10.1021/ja9929180 CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/12/1999

Communications to the Editor
J. Am. Chem. Soc., Vol. 121, No. 47, 1999 11025
Figure 3. Effect of the mole ratio of oxygen:n-butane on the percentage
of label scrambling for n-butane and oxygen reacting on VPO 3.92 at
380 °C.
13C n-butane using VPO catalyst with an average vanadium
oxidation state of 3.92 (denoted VPO 3.92) under anaerobic
conditions at 540 °C. Figure 2c shows the expanded region of
Figure 2a. All of the proposed intermediates in the n-butane
oxidation were observed, and no maleic acid was produced. This
result is consistent with previous studies using TAP (temporal
analysis of products) reactors and the requirement that V5+ is
necessary for the oxidation to maleic anhydride.1,5,6,10,11 The most
interesting feature is that the peaks for 2,3-13C butadiene (137-
138 ppm) are almost as large as the peaks for 1,4-13C-labeled
butadiene (114-117 ppm). Label scrambling does not occur under
identical conditions for 2-13C butadiene (Figure 2b,d), indicating
that it must have occurred prior to the formation of butadiene.
The product distribution for reaction temperatures from 450 to
600 °C was identical, indicating that the mechanism for butadiene
formation from n-butane on VPO 3.92 is independent of reaction
temperature.
All of the experiments presented so far were performed using
anaerobic conditions. We simulated industrial reaction conditions
by running the reaction in the presence of oxygen using VPO
3.92 at 380 °C.12 The products (CO, CO2, maleic acid, fumaric
acid, methanediol, and ethylene) were the same as observed for
n-butane reacting on VPO 4.56 (Figure 1a). None of the proposed
intermediates were observed. The percent of label exchange vs
oxygen:n-butane ratio is shown in Figure 3. Label scrambling
was highest at low oxygen:n-butane ratios, decreased until a
minimum was reached at about 10:1, and increased slightly at
higher oxygen:n-butane ratios. This result is consistent with the
greatest percentage of label scrambling occurring for low average
vanadium oxidation states.
Figure 2. 13C NMR spectra of the reaction products of 1,4-13C n-butane
(a) or 2-13C-butadiene (b) on VPO 3.92 at 540 °C. Parts c and d are
expansions of parts a and b, respectively. In all spectra dashed and solid
lines indicate gas and liquid peaks, respectively.
The peak assigned to 2,3-13C maleic acid is bracketed by two
smaller peaks assigned to the 2,3 carbons of maleic acid at natural
abundance. The resonance of the natural abundance carbons is
split into a doublet by 13C-13C couplings to the 13C labels in the
1 and 4 positions of maleic acid. Uniformly 13C labeled n-butane
run under identical conditions (Figure 1c) produced essentially
the same product distribution, except that the peaks for ethylene
and acetylene are much larger than observed for 1,4-13C n-butane.
It also shows that the parameters used to acquire the spectrum
resulted in the carbonyl resonance of maleic acid being ∼6.5 times
smaller than the olefinic resonance, although the ratio should be
1:1. Spectra acquired with longer repetition times between
acquisitions showed that the intensities of all of the other peaks
were quantitative. From the integrated intensity ratios of 1:5 for
the carbonyl and olefinic resonances, respectively, in Figure 1b
and using a scaling factor of 6.5, we estimate that the amount of
label exchange for this sample was ∼3%. This is consistent with
the intensities of the natural abundance carbons (1.1%) observed
in Figure 1b.
Our results show that the label in butadiene, produced from
1,4-13C n-butane, is completely scrambled, but in maleic acid,
also produced from 1,4-13C n-butane, the label is largely
unscrambled. This makes it unlikely that maleic acid is formed
predominantly by a butadiene intermediate. However, the small
amount of label scrambling in maleic acid suggests that both
reaction pathways may be occurring simultaneously to produce
maleic acid. We are currently investigating the influence of
oxidation state and phosphorus-to-vanadium ratios on the percent
of label exchange.
We have analyzed over 100 samples of 1,4-13C n-butane
reacting on VPO catalysts in average vanadium oxidation states
between 4.3 and 5 and at reaction temperatures between 330 and
500 °C and found that the amount of label scrambling varied
between 0 and 7%, depending upon both oxidation state and
reaction temperature. Label scrambling was greatest at oxidation
states <4.6, although scrambling was observed at higher oxidation
states as well. We explored the possibility that label scrambling
occurred when butadiene reacted on VPO. Figure 1d shows the
reaction of 2-13C butadiene on VPO 4.56 at 330 °C. No evidence
of label scrambling was observed. This suggests that the label
scrambling occurred prior to butadiene formation.
Acknowledgment. We thank Dr. John Bemis for preparing one of
the catalysts, Professor Lannie Schmidt for helpful discussions, and
Professor Peter Carr for critical reading of the manuscript. This work
was supported in part by the Petroleum Research Fund, administered by
the American Chemical Society, PRF 30584-G5, and by NSF grant
CHE-9624161.
JA9929180
(10) Coulston, G. W.; Bare, S. R.; Kung, H.; Birkeland, K.; Bethke, G.
K.; Harlow, R.; Herron, N.; Lee, P. L. Science 1997, 275, 191-193.
(11) Rodemerck, U.; Kubias, B.; Zanthoff, H. W.; Baerns, M. Appl. Catal.
A-Gen. 1997, 153, 203-216.
(12) Herron, N.; Thorn, D. L.; Harlow, R. L.; Coulston, G. W. J. Am. Chem.
Soc. 1997, 119, 7149-7150.
Figure 2a shows the 13C NMR spectrum of the reaction of 1,4-