OXYGEN-18 AND KINETICS OF BUTADIENE EPOXIDATION
4. CONCLUSIONS
367
Of course, this is an approximate expression since it does
not include the other kinetic parameters that have been
discussed earlier (1), namely the kinetic dependency of bu-
tadiene and the negative reaction orders in CO2 pressure.
The goal here is not to attempt to write a complete rate ex-
pression that satisfies all observed kinetic parameters, but
to reconcile the strong kinetic inhibition by EpB within the
framework of dissociation of molecular oxygen as a RDS
for EpB formation. Clearly, although the strongly adsorbed
state of EpB does not affect the RDS in a direct way, it can
still have a strong effect on the kinetics, as demonstrated by
the denominator of Eq. [5].
The O-18 results in this study directly support the con-
clusion that atomic oxygen (16, 30–32), and not a molecular
oxygen species, is the active form of oxygen that reacts with
olefins to form olefin epoxides. The very high selectivities
for EpB formation observed in this study (99% ) and in ear-
lier work (1, 2) (94–96% ) support the assignment of atomic
oxygen as the active oxygen species for butadiene epoxi-
dation. If we use the argument of Sachtler et al. (33), com-
monly known as the 6/7 rule for ethylene oxide formation
which states that the selectivity for ethylene oxide forma-
tion cannot exceed 6/7, or 85.7% , if molecular oxygen is the
active oxygen for epoxidation, then the rule for butadiene
epoxidation becomes 11/12 for molecular oxygen. Clearly,
the selectivity values cited above for EpB formation are
much higher than 11/12, or 91.7% , indicating that molec-
ular oxygen is not the active oxygen that electrophilically
addsto the C == C bond duringepoxide formation. Thisstudy
1
8
Kinetic isotope effect data using O2 for butadiene epox-
idation over Cs-promoted, supported Ag catalysts have
been experimentally determined to show that the rate-
determining step for butadiene epoxidation is dissociation
�
1
ofa molecular oxygen species(O2) on a vacant Agsurface
site. As confirmation, the experimentally obtained KIE val-
ues have also been compared to calculated KIE values for
other reaction steps (other than O–O dissociation) involv-
ing bond-making or bond-breaking steps in which oxygen
is involved. In all these instances the calculated KIE val-
ues are much lower than the KIE actually observed. To
the best of our knowledge, this is the first instance where
18
O2 has been used at steady-state olefin epoxidation con-
ditions to confirm the nature of the oxygen active in olefin
epoxidation.
The O-18 results in this study also directly support the
conclusion that atomic oxygen, and not a molecular oxygen
species, is the active form of oxygen that reacts with olefins
to form olefin epoxides. Finally, comparison of the kinet-
ics for butadiene epoxidation with the kinetics for ethy-
lene epoxidation shows that the rate-limiting steps for the
two reactions are different. For ethylene epoxidation, the
surface reaction between adsorbed ethylene and adsorbed
oxygen is considered to be the limiting step, while disso-
ciation of molecular oxygen is rate limiting for butadiene
epoxidation.
ACKNOWLEDGMENTS
marks the first time that direct evidence has been obtained
under steady-state conditions to confirm the nature of the
oxygen active in olefin epoxidation. In addition, compari-
One of us (J.R.M.) gratefully acknowledges Eastman Chemical Com-
pany for permission to publish this material and the efforts of Ms. Libby
son of the kinetics for butadiene epoxidation with the ki- Cradic in preparing this manuscript. J.W.M. and M.A.B. acknowledge the
netics for ethylene epoxidation shows that the rate-limiting financial support of the Department of Energy, Office of Science, Division
of Chemical Sciences (Grant FG02-84ER-13290). J.W.M. also acknowl-
edges support from the National Science Foundation in the form of a
graduate fellowship.
steps for the two reactions are different. For ethylene epox-
idation, the surface reaction between adsorbed ethylene
and adsorbed oxygen is considered to be the limiting step
(
32, 34), while we have shown that molecular oxygen dis-
REFERENCES
sociation is rate limiting for butadiene epoxidation. This
may largely reflect the significant differences in stability of
the oxametallacycle intermediates calculated for the two
epoxidation processes (6, 7) and the resulting differences
in the occupancy of surface sites by these intermediates un-
der reaction conditions. Furthermore, the similar promoter
requirements for styrene (13) and norbornene (14) epoxi-
dation and the existence of strong inhibition effects of both
epoxide products on these reactions also suggest that the
RDS for epoxidation of these olefins is very likely molec-
ular oxygen dissociation as well. In fact, it is possible that
of all the olefins capable of being selectively epoxidized by
O2 using silver catalysts, the epoxidation of ethylene is the
reaction that is different from that of other olefins. If so, this
is somewhat ironic, since almost all of our understanding to
date regarding olefin epoxidation using O2 and Ag catalysts
is based on epoxidation of ethylene.
1. Monnier, J. R., Stud. Surf. Sci. Catal. 110, 135 (1997).
2
3
4
. Monnier, J. R., and Muehlbauer, P. J., U.S. Patent 4,950,773 (1990) to
Eastman Chemical Company.
. Monnier, J. R., and Muehlbauer, P. J., U.S. Patent 5,138,077 (1992) to
Eastman Chemical Company.
. Stavinoha, J. L., Jr., Monnier, J. R., Hitch, D. M., Nolen, T. R., and
Oltean, G. L., U.S. Patent 5,362,890 (1994) to Eastman Chemical
Company.
5
6
. Barnicki, S. D., and Monnier, J. R., U.S. Patent 6,011,163 (2000) to
Eastman Chemical Company.
. Medlin, J. W., Mavrikakis, M., and Barteau, M. A., J. Phys. Chem. B
1
03, 11169 (1999).
7. Medlin, J. W., Barteau, M. A., and Vohs, J. M., J. Mol. Catal. 163, 129
2000).
. Cowell, J. J., Santra, A. K., and Lambert, R. M., J. Am. Chem. Soc.
22, 2381 (2000).
. Chem. Eng. News, p. 7, August 21, 1995.
(
8
9
1
1
0. Chem. Week, p. 28, April 22, 1998.
11. Monnier, J. R., and Hartley, G. W., J. Catal., in press.