Deuterium kinetic isotope effects in butadiene epoxidation over unpromoted and Cs-promoted silver catalysts

Article abstract of DOI:10.1006/jcat.2001.3381

The kinetic isotope effects (KIE) in the epoxidation of butadiene over both unpromoted and Cs-promoted catalysts were studied using selectively deuterium-labeled isotopomers of 1,3-butadiene (CD₂CDCDCD₂, CD₂CHCHCD₂, and CH₂CDCDCH₂) as feed reactants. On unpromoted silver catalysts, 3,4-epoxy-1-butene formation was characterized by an inverse KIE that results from a primary effect in total oxidation to CO₂. These KIE were only significant for butadiene isotopomers that are D-labeled in the 1- and 4-positions, indicating that combustion was initiated by cleavage of a terminal C-H bond. In analogous fashion to previous KIE studies for the epoxidation of ethylene, the observed KIE could be explained by a reaction scheme in which partial and total oxidation products were generated from a common C₄H₆(ads) surface intermediate. This scheme was also consistent with previous surface science experiments, suggesting that the C₄H₆O(ads) intermediate could be a surface oxametallacycle. The results of KIE experiments performed over Cs-promoted silver catalysts were in accord with the proposed reaction scheme and showed that the function of the Cs promoter could be to lower the activation energy for epoxy-butene formation from the C₄H₆O(ads) species.

Full text of DOI:10.1006/jcat.2001.3381

Journal of Catalysis 204, 71–76 (2001)
doi:10.1006/jcat.2001.3381, available online at http://www.idealibrary.com on
Deuterium Kinetic Isotope Effects in Butadiene Epoxidation
over Unpromoted and Cs-Promoted Silver Catalysts
�
�,1
J. Will Medlin, John R. Monnier,† and Mark A. Barteau
�
Center for Catalytic Science and Technology, Department of Chemical Engineering, University of Delaware, Newark, Delaware 19716;
and †Eastman Chemical Company, Research Laboratories, P.O. Box 1972, Kingsport, Tennessee 37662
Received February 15, 2001; revised August 1, 2001; accepted August 1, 2001
butadiene epoxidation, it may be possible to obtain a better
Selectively deuterium-labeled isotopomers of 1,3-butadiene understanding of the mechanism for silver-catalyzed olefin
CD CDCDCD , CD CHCHCD , and CH CDCDCH ) have been epoxidations.
(
2
2
2
2
2
2
used as feed reactants to study kinetic isotope effects (KIEs) in the
epoxidation of butadiene over both unpromoted and Cs-promoted
catalysts. On unpromoted catalysts, the rate of formation of the
epoxide product, 3,4-epoxy-1-butene (EpB), is increased relative to
that for unlabeled butadiene for the butadiene isotopomers labeled
on the terminal carbon atoms. This indicates that EpB formation is
characterized by an inverse KIE that is caused by a primary KIE in
the rate of the combustion reaction and also demonstrates that com-
bustion is initiated by cleavage of a terminal C–H bond. These KIE
One useful method for probing the reaction mechanism
is the measurement of kinetic isotope effects (KIEs). Cant
and Hall (3) previously studied the silver-catalyzed epoxi-
dation of unlabeled and deuterium-labeled ethylene. They
observed a sharp enhancement in the rate of ethylene ox-
ide (EO) production when C2D4 was used as the reactant.
This result seems somewhat surprising, since deuterium ki-
netic isotope effects are normally associated with a decrease
phenomena can be explained by a reaction mechanism in which in rate for D-labeled compounds. However, Cant and Hall
both partial and complete oxidation products are produced from were able to interpret their results in terms of a mechanism
a common C
H O(ads) surface intermediate. The measured KIEs in which both ethylene oxide and CO were generated from
4 6
2
for the reaction over the Cs-promoted catalyst are also consistent
with this mechanism. Comparisons between KIE measurements in
ethylene and butadiene epoxidation are discussed, and links with
surface science studies of C H O oxametallacycle reactions are es-
4 6
tablished. �c 2001 Academic Press
a common surface intermediate. Under this mechanism, the
rate of epoxide production was increased due to an increase
in the selectivity to EO. In the present work, we report re-
sults from the study of KIEs in the butadiene epoxidation
reaction for three deuterium-labeled butadiene reagents:
CH2CDCDCH2, CD2CHCHCD2, and C4D6. These results
are then compared to those obtained for the ethylene epox-
idation process in order to understand them in terms of a
common mechanism for olefin epoxidation.
Key Words: deuterium isotope effect; butadiene oxidation; olefin
epoxidation; epoxybutene; silver catalyst; epoxidation kinetics.
INTRODUCTION
The recent discovery of a selective catalyst for the epox-
idation of 1,3-butadiene to 3,4-epoxy-1-butene (EpB) (1)
EXPERIMENTAL
2
has raised some interesting questions about the mecha-
nism of epoxidation reactions. Although epoxidations of
both ethylene and butadiene are carried out over silver
catalysts, the catalyst for the latter incorporates a much
higher loading of the Cs promoter (2). Differences in
the measured kinetics within each of these processes
also suggest that the relative rates of various elementary
steps for these two processes may be different. In other
words, the rate-determining steps may be different in each
case. By performing comparative studies of ethylene and
Experiments were conducted in an atmospheric pressure
flow reactor system equipped with on-line sampling, similar
to systems described previously (4). Standard feed condi-
tions consisted of 9 mol% butadiene and 18% oxygen in
a helium diluent, with a total flow of 50–60 standard cubic
centimeters per minute (sccm). A temperature-controlled
furnace was used to maintain a steady reaction tempera-
ture, which was measured via a thermocouple inserted into
a thermal well in the catalyst bed. For the KIE experiments,
a switching valve (Valco Valves) system was used to switch
the flow between the unlabeled and labeled butadiene.
The isotopes were introduced into the system using
a microprocessor-controlled and well-calibrated syringe
pump (Harvard Apparatus, Model 44 Infusion Pump); all
1
To whom correspondence should be addressed. E-mail: barteau@
che.udel.edu. Fax: (302)-831-8201.
2
EpB is a registered trademark of Eastman Chemical Company.
71
0021-9517/01 $35.00
Copyright �c 2001 by Academic Press
All rights of reproduction in any form reserved.

7
2
MEDLIN, MONNIER, AND BARTEAU
TABLE 1
Kinetic Isotope Effects in Butadiene Epoxidation on an Unpromoted Catalyst
Selectivitya
(C4H6)
TOF, s
� 1
CH2CDCDCH2
CD2CHCHCD2
CD2CDCDCD2
Product
EpB
CO2
Furan
(C4H6)
(rD /rH )
(rD /rH )
(rD /rH )
rD /rH SDb
4
3
5
5
26.2%
68.6%
3.3%
6.50 � 10�
1.01
0.98
1.22
0.77
1.38
0.95
1.10
0.78
1.36
0.96
1.38
0.68
0.02
0.03
0.02
0.09
�
1.71 � 10
�
8.08 � 10
Crotonaldehyde
1.6%
3.87 � 10�
a
�
Reactor bed temperature = 270 C, butadiene conversion = 1.1% .
b
The standard deviation represents the maximum standard deviation of the 3 rD /rH columns.
other components were admitted through mass flow con- site blocking by a “strongly bound” EpB state (2) that will
trollers (Tylan Corporation, Model FC-260). Butadiene- be referred to here as C4H6O(ads). After approximately
2
,3-d2, butadiene-1,1,4,4-d4, and butadiene-d6 were obta- 2 h on stream, a low but steady-state activity was observed.
ined from CDN isotopes, and had isotopic purities >98% . In order to obtain a quantifiable product yield, the reaction
�
Unpromoted and Cs-promoted Ag/�-Al2O3 catalysts temperature was set at 270 C to give a steady-state butadi-
were prepared and characterized via previously described ene conversion of 1.1% . The first three columns in Table 1
methods (5). Cesium chloride was used as the promoter salt show the product distribution and rates of formation under
for the promoted catalyst. The silver loadings for the unpro- these reaction conditions. Carbon dioxide is the major reac-
moted and promoted catalysts were 13.4 and 12.0% , respec- tion product, followed by EpB, furan, and crotonaldehyde.
tively. Scanning electron microscopy (SEM) was used to Clearly, after a relatively short time onstream, the unpro-
determine the distribution of Ag particles on the catalysts. moted silver catalyst is neither highly active nor selective
Using automatic particle counting/measuring software to for butadiene epoxidation.
determine Ag crystallite diameters from the scanning elec-
In order to determine the influence of butadiene par-
tron micrographs, the average Ag particle diameters were tial pressure on the reaction kinetics, the butadiene feed
calculated to be 0.11 and 0.14 �m for the unpromoted and mole fraction was varied between 4.5 and 18.0% holding
promoted catalysts, respectively. Assuming hemispherical all other variables constant. Figure 1 depicts a logarithmic
Ag particles, the Ag surface site densities were thus calcu- plot of the rate of formation of each product as a function of
18
18
lated to be 8.05� 10 Ag sites/g catalyst and 5.37� 10 Ag the butadiene mole fraction. If one assumes power rate law
sites/g catalyst, respectively, for the unpromoted and pro- kinetics, the reaction order in butadiene can be determined
moted catalysts. Previous O2 chemisorption analyses (5) for from the slope of each data set on this plot. As shown in
a similar series of CsCl-promoted Ag catalysts showed that
normalized O2 uptake values for optimal CsCl-promoted
catalysts were approximately one-half of the O2 uptake for
the unpromoted catalyst at comparable Ag loadings. Thus,
we assume the surface Ag site concentration of the pro-
18
moted catalyst in this study to be approximately 2.7� 10 /g
catalyst. These values for Ag site densities were used for the
calculation of turnover frequencies (TOFs) for EpB and
CO2 formation. Both catalysts were sieved to 18/25 mesh
and loaded above a coarse frit in the Pyrex reactor. Catalyst
loadings of 2.00 and 1.00 g of catalyst were used for exper-
iments with unpromoted and promoted catalysts, respec-
tively. Catalysts were calcined in a flowing stream of 20%
�
O2 in He at 250 C for 2 h before being brought on line.
RESULTS AND DISCUSSION
�
After bringing the catalyst on stream at 200 C, the activ-
ity of the unpromoted catalyst for epoxidation of butadiene
dropped dramatically over a period of 1–2 h. This behavior
FIG. 1. Power rate law plot for the rate of product formation as a
has been observed previously, and has been attributed to function of butadiene partial pressure.

KINETIC ISOTOPE EFFECTS IN BUTADIENE EPOXIDATION
73
Fig. 1, the rates of formation of all the partial oxidation primarily via the reverse reaction (with rate constant k1)
products (EpB, furan, and crotonaldehyde) are nearly in- to regenerate EpB. Small amounts of 2,5-dihydrofuran
dependent of butadiene partial pressure; CO2 formation, (2,5-DHF) are also produced on the reduced silver surface
however, has an apparent order of � 0.5 with respect to during butadiene epoxidation (2). Under the oxidative con-
butadiene. Thus, increasing the mole fraction of butadiene ditions associated with the butadiene epoxidation reaction,
inhibits the combustion process without substantially af- 2,5-DHF rapidly undergoes allylic oxidation to form furan
fecting the rate of formation of other oxidation products. (2); therefore, we conclude that furan is also produced from
One possible explanation for this result is that increased the common reaction intermediate. Since previous studies
butadiene adsorption on the catalyst may decrease the cov- have shown that 2,5-DHF does not adsorb on model sil-
erage of available surface oxygen responsible for deep ox- ver surfaces to form a stable surface oxametallacycle (8),
idation. Further studies are necessary, however, to confirm the reaction described by rate constant k2 is taken to be
this hypothesis.
irreversible.
The results of experiments conducted with deuterium-
It is possible that crotonaldehyde is also produced from
labeled butadiene are summarized in Table 1, columns the common C4H6O(ads) intermediate, but the reaction
–7. The observed KIE values are reported as rate ratios pathway for crotonaldehyde production cannot be deduced
rD /rH ), i.e., the rate of formation of each product from from the studies carried out to date. There is no evidence
4
(
the labeled butadiene relative to the rate of formation of from our surface science studies that crotonaldehyde can
that product from butadiene-d0. The rates of formation of be produced from reactions of surface oxametallacycles.
EpB and furan are characterized by having inverse isotope Furthermore, the observed KIEs for crotonaldehyde for-
effects, whereas the rates of formation of CO2 and cro- mation are consistent with any mechanism that requires
tonaldehyde undergo the more usual decrease in rate with C–H scission in the rate-determining step. Thus, it is possi-
isotopic substitution. These results are similar to those ob- ble that crotonaldehyde is formed from the series reaction
served by Cant and Hall for ethylene epoxidation, in which of another product. Studies are underway to investigate
a large inverse KIE for ethylene oxide and a smaller nor- the reaction ofC4H6O oxametallacycleson oxygen-covered
mal effect for CO2 were observed (3). As demonstrated by Ag(110) in order to more fully evaluate the mechanism of
Cant and Hall, KIE values of this type can be explained butadiene epoxidation by determining whether furan, CO2,
by a mechanism in which multiple products arise from a and crotonaldehyde are reaction products on the oxidized
common reaction intermediate.
To explain the experimental data, we propose the buta-
surface.
The mechanism proposed in Scheme 1 can be used to
diene epoxidation mechanism shown in Scheme 1. The cen- qualitatively explain the essential features of the isotope
tral feature of this mechanism—namely, that EpB and CO2 effects reported in Table 1. Because further oxidation of
are produced from a common surface intermediate—is di- the C4H6O(ads) intermediate involves C–H bond cleavage,
rectly analogous to the scheme proposed by Cant and Hall the rate constants for oxidation (k3–6) should decrease for
for ethylene epoxidation. This mechanism is also consistent deuterium-labeled butadiene. The decrease in these rate
with studies of EpB adsorption on model silver surfaces. It constants will result in an increase in the concentration of
has been shown that EpB adsorbs on the Ag(110) surface C4H6O(ads), which will therefore increase the rate of EpB
(
in a step associated with rate constant k� 1 in Scheme 1) formation. The increase in the C4H6O(ads) concentration
to form a strongly bound C4H6O(ads) oxametallacycle in- can also explain the relatively small KIEs observed for CO2;
termediate (6, 7). This oxametallacycle intermediate reacts i.e., although deuterium labelingdecreasesthe rate constant
for combustion, k3, this decrease is nearly offset by the in-
crease in C4H6O(ads) concentration.
In their studies of ethylene and propylene epoxidation,
Cant and Hall (3) present a series of rate equations derived
from the pseudo-steady-state approximation (PSSA) for
the common intermediate. From these rate equations, the
authors determine KIEs in terms of rate constants for indi-
vidual mechanistic steps. Neglecting the rate of formation
of furan and crotonaldehyde for simplicity (low TOF values
in Table 1), a completely analogous treatment (the reader
is referred to Ref. (3) for the details) yields
r1 = k1[C4H6O(ads)] � k� 1 PEpB
[1]
for the rate of formation of EpB, where PEpBis the partial
SCHEME 1. Proposed mechanism for butadiene oxidation.
pressure of EpB. The concentration of the common surface

7
4
MEDLIN, MONNIER, AND BARTEAU
intermediate, [C4H6O(ads)], is evaluated using the PSSA rate is observed for C4D6 and CD2CHCHCD2 than for
to be
CH2CDCDCH2. This observation suggests that rate of fur-
ther oxidation of the C4H6O(ads) intermediate is largely
[2] governed by the rate of reaction at its terminal C–H bonds.
Oxidation at the 2- and 3-positions apparently has no mea-
surable effect on the rate of combustion.
k� 1 PEpB + kR PO PC H
,
2
4
6
[
C4H6O(ads)] =
k1 + k3
where PO2 and PC H are the partial pressures of oxygen
4
6
As shown in Table 1, KIEs have also been detected in the
rates of formation of the two minor reaction products, fu-
ran and crotonaldehyde. The use of the deuterium-labeled
isotopomers results in opposite KIEs for furan and cro-
tonaldehyde formation: the rates of furan formation are in-
creased, whereas the rates of crotonaldehyde formation are
decreased. As discussed in the following, this suggests that
and butadiene, respectively. In their treatment, Cant and
Hall (3) neglect the rate of reverse reaction for epoxide
formation (i.e., assume k� 1 is negligible); following this
procedure leads to the following expression for the inverse
KIE for EpB production at steady state:
r1D
r1H
k1 + k3H
k1 + k3D
=
.
[3]
the rate-determining step for crotonaldehyde formation
involves C–H bond cleavage, while the rate-determining
Here, rate constants expected to be affected by deuterium step for furan formation does not. The rate of furan for-
labeling are denoted with a subscript D or H. Based on mation is observed to increase for all three labeled buta-
the selectivity data in Table 1, we estimate that k3H = 2.6 k1 dienes, indicating that furan production is also enhanced
for unlabeled butadiene, and we also follow the common due to the accumulation of the C H O(ads) intermediate
4
6
assumption that k3H = 2k3D (10) for reactions limited by upon isotopic substitution. Interestingly, the rate increase
C–H bond cleavage. From Eq. [3], this yields a predicted is greatest for C D , intermediate for CH CDCDCH , and
4
6
2
2
maximum inverse KIE of approximately 1.6, significantly least for CD CHCHCD . Furan has been shown (2) to
2
2
above that reported for any of the isotopomers in Table 1. be formed from the rapid oxidative dehydrogenation of
However, the PSSA treatment is probably not as appro- 2,5-dihydrofuran (2,5-DHF). Because the formation of 2,5-
priate in this case (in comparison to the studies of Cant DHF from C H O(ads) is expected to be the rate-limiting
4
6
and Hall) for several reasons. First, butadiene epoxidation step in furan formation, deuterium substitution at the 1-
is a much more complex reaction, with a broader array of and 4-positions, by the previous arguments, would increase
pathways and products, than the epoxidation of ethylene the rate of formation of furan. However, the rate of fu-
or propylene. Thus, the estimation of a large number of ran production will also be controlled by the rate at which
individual rate constants is required for a full PSSA treat- furan is consumed by further oxidation to CO ; the obser-
2
ment. Furthermore, the previous analysis assumes a low vation that r /r is higher for CH CDCDCH than for
D
H
2
2
surface coverage of the C4H6O(ads) intermediate, and a CD CHCHCD suggests that the further oxidation of fu-
2
2
correspondingly low value for the rate constant k� 1. This ran at its 3- and 4-positions may have the strongest effect
simplification is probably not valid in the case of a strongly on the net rate of furan formation. The yields of croton-
adsorbed C4H6O(ads) intermediate, where the activa- aldehyde, yet another nonselective product, are too low to
tion energy for EpB ring opening has been measured evaluate trends in reaction rate as a function of the
on Ag(110) to be lower than the activation energy for deuterium-labeling position. It is clear, however, that deu-
C4H6O(ads) ring closure (6). In other words, it is not valid terium labeling of butadiene results in a decrease in the
to assume that EpB formation from C4H6O(ads) is an irre- rate of formation of crotonaldehyde. Crotonaldehyde is
versible step. Including the rate constant k� 1 in the previ- most likely formed via an isomerization of some other
ous analysis will result in a decrease in the predicted KIE, C H O species (EpB, 2,5-DHF, or the strongly adsorbed
4
6
possibly bringing it more in line with the values reported C H O(ads) intermediate) that requires hydrogen transfer
4
6
in Table 1. Finally, as has been pointed out recently (10), steps. In isomerization of these species to crotonaldehyde,
detailed mechanisms derived from macrokinetic data are a 1–4 hydrogen shift involving cleavage of a C–H bond at
often misleading. Thus, we instead focus on qualitative as- the 1- or 4-position of butadiene is expected. The obser-
pects of the proposed mechanism in order to demonstrate vation of a clear KIE for butadiene labeled at the 2- and
trends in the experimental results.
3-positions, however, suggests that crotonaldehyde may be
With the mechanism in Scheme 1 as a basis for inter- formed by a more complex series of steps, and possibly in-
preting the data, it is useful to examine the results in volves hydride transfer steps along the carbon backbone of
Table 1 in closer detail. For example, the rD /rH ratio for a C H O species.
4
6
EpB is seen to be a nearly identical, 1.36–1.38 for C4D6
We now briefly consider other mechanistic schemes that
and CD2CHCHCD2, whereas the rate is practically un- could potentially be used to describe the epoxidation of bu-
changed for CH2CDCDCH2. This effect is mirrored by tadiene. For example, one might suggest a scheme in which
the rate of CO2 formation, in which a larger decrease in EpB and CO are produced from parallel channels for the
2

KINETIC ISOTOPE EFFECTS IN BUTADIENE EPOXIDATION
75
TABLE 2
Kinetic Isotope Effects in Butadiene Epoxidation on a Cs-Promoted Catalyst
Selectivitya
(C4H6)
TOF, s
� 1
CH2CDCDCH2
CD2CHCHCD2
CD2CDCDCD2
Product
EpB
CO2
Furan
(C4H6)
(rD /rH )
(rD /rH )
(rD /rH )
rD /rH SDb
2
4
96.6%
3.4%
0.0%
0.0%
2.57 � 10�
1.01
0.99
N/A
N/A
1.01
0.84
N/A
N/A
0.99
0.85
N/A
N/A
0.02
0.02
N/A
N/A
�
9.05 � 10
0
0
Crotonaldehyde
a
�
Reactor bed temperature = 173 C, butadiene conversion = 8% .
b
The standard deviation represents the maximum standard deviation of the 3 rD /rH columns.
reaction of butadiene and oxygen. This pathway is rejected unpromoted case, the activity of the promoted catalyst dra-
because it cannot account for the significant KIEs observed matically increased over the first several hours on stream
for EpB formation. In a purely parallel mechanism, a de- until a high steady-state activity was achieved. This induc-
crease in the rate constant for combustion would lead to an tion period has been attributed to the enhancement of silver
increase in the partial pressure of the reactants, which could dispersion and/or the removal of excess chlorine from the
in turn lead to an increase in the rate of EpB production. catalyst (11). Due to the greatly improved catalytic activ-
�
However, at the butadiene conversion levels (1.1% ) of our ity, a much lower reaction temperature of 173 C was em-
experiments, any decrease in the combustion rate constant ployed. Even at these lower temperatures, the butadiene
would have a negligible effect on the partial pressure of the conversion was nearly 8% , with selectivity to EpB greater
reactant; i.e., the increase in reactant concentration would than 96% . Carbon dioxide was the only nonselective
be less than 1.1% even if the rate of combustion were fully product detected. Table 2 summarizes the observed KIE
suppressed. This effect is far too small to account for the data for the promoted catalyst. No effect on the rate of
measured 36% increase in the rate of EpB production. Thus EpB formation is discernible for any of the isotopically la-
a common surface intermediate is necessary to explain the beled butadienes, but a small decrease in the rate of CO2
isotope effects observed.
formation is observed. These results are consistent with the
Nevertheless, it should be pointed out that it is possible to basic mechanism shown in Scheme 1; because the rate of
qualitatively explain the observed KIE values for the major EpB formation is so much greater than the rate of CO2
products using other mechanistic schemes. For example, a formation, small changes in the rate constants for combus-
purely sequential scheme is possible, i.e.,
tion will have no measurable effect on the rate at which
EpB is produced. It has been suggested that the role of the
[4] Cs promoter is to increase the rate of EpB formation from
C4H6 → EpB ↔ C4H6O(ads) → CO2.
the strongly adsorbed C4H6O intermediate (2, 6) by low-
ering the desorption energy of strongly bound EpB. That
is, in terms of the mechanism of Scheme 1, the addition
of Cs results in a substantial increase in k1. The observed
KIE values for CO2 again suggest that combustion is ini-
tiated by C–H bond cleavage at the 1- and 4-positions of
butadiene. In summary, these studies indicate that there
are apparent similarities between the mechanisms of the
silver-catalyzed epoxidation of ethylene and butadiene, as
evidenced by similar trends in KIE values for unpromoted
catalysts. However, key differences in the product distri-
bution complexity, adsorption energy of the products, and
promoter levels used in industrial catalysts point to oppor-
tunities for elucidating the mechanism of olefin epoxidation
through comparative studies of these chemistries.
In such a scheme, using D-labeled butadiene would again be
expected to lower the rate constant for complete combus-
tion, thus increasing the concentration of the C4H6O(ads)
intermediate. An increase in the concentration of this sur-
face intermediate would again increase the rate of forma-
tion of EpB, consistent with the effects observed in Table 1.
We have favored the mechanism of Scheme 1 based on its
consistency with the mechanism for ethylene epoxidation
previously suggested by Cant and Hall (3). In the case of
the less complex ethylene epoxidation reaction, the authors
were able to show that this type of model was quantita-
tively consistent with observed KIEs. Moreover, whether
the mechanism suggested in Scheme 1 or Eq. [4] is more
representative of the “correct” epoxidation mechanism, the
central conclusion of this work is the same: namely, the ob-
served KIE values can be explained by the influence of the
strongly adsorbed C4H6O(ads) oxametallacycle intermedi-
ate on the overall epoxidation kinetics.
CONCLUSIONS
Finally, we consider isotope effects for butadiene epox-
Selectively deuterated 1,3-butadiene isotopomers have
idation over the Cs-promoted catalyst. In contrast to the been used to study KIEs in the silver-catalyzed epoxidation

7
6
MEDLIN, MONNIER, AND BARTEAU
of butadiene. On unpromoted silver catalysts, epoxybutene support of the Department of Energy, Office of Science, Division of Chem-
ical Sciences (Grant FG02-84ER13290). JWM also acknowledges support
from the National Science Foundation in the form of a graduate fellowship.
formation is characterized by an inverse KIE that results
from a primary effect in total oxidation to CO2. These KIEs
are only significant for butadiene isotopomers that are D-
labeled in the 1- and 4-positions, indicating that combustion
is initiated by cleavage of a terminal C–H bond. In analo-
REFERENCES
gous fashion to previous KIE studies for the epoxidation of
ethylene, the observed KIEs can be explained by a reaction
scheme in which both partial and total oxidation products
are generated from a common C4H6O(ads) surface inter-
mediate. This scheme is also consistent with previous sur-
face science experiments, suggesting that the C4H6O(ads)
intermediate may be a surface oxametallacycle. The results
of KIE experiments conducted over Cs-promoted silver
catalysts are also in accord with the proposed reaction
scheme and demonstrate that the function of the Cs pro-
moter may be to lower the activation energy for epoxy-
butene formation from the C4H6O(ads) species.
1. Denton, D., Falling, S., Monnier, J. R., Stavinolia, J., and Watkins, W.,
Chim. Oggi 5, 17 (1996).
2. Monnier, J. R., Stud. Surf. Sci. Catal. 110, 135 (1997).
3. Cant, N. W., and Hall, W. K., J. Catal. 52, 81 (1978).
4. Monnier, J. R., Medlin, J. W., and Kuo, Y. J., Appl. Catal. A 194, 463
(2000).
5. Monnier, J. R., and Muehlbauer, P. J., US Patent 4,897,498 to Eastman
Chemical Co.
6
. Medlin, J. W., Barteau, M. A., and Vohs, J. M., J. Mol. Catal. A 163,
29 (2000).
1
7. Medlin, J. W., Sherrill, A. B., Chen, J. G., and Barteau, M. A., J. Phys.
Chem. B 105, 3769 (2001).
8. Medlin, J. W., unpublished results.
9. Moore, J. W., and Pearson, R. G., “Kinetics and Mechanism.” Wiley,
New York, 1981.
10. Over, H., Kim, Y. D., Seitsonen, A. P., Wendt, S., Lundgren, E.,
ACKNOWLEDGMENTS
Schmid, M., Varga, P., Morgante, A., and Ertl, G., Science 287, 1474
(
2000).
JRM gratefully acknowledges Eastman Chemical Company for permis-
sion to publish this material. JWM and MAB acknowledge the financial
11. Monnier, J. R., and Muehlbauer, P. J., US Patent 4,950,773 to Eastman
Chemical Co.