Time-resolved infrared spectral studies of photochemically induced oxidative addition of benzene to trans-RhCl(CO) (PMe3)2

Article abstract of DOI:10.1021/om970679o

Time-resolved infrared and time-resolved optical spectroscopy were used to examine the pathway-(s) by which 355 nm photolysis of the rhodium(I) species trans-RhCl(CO)(PMe₃)₂ (1) in benzene leads to the C-H oxidative-addition product (Ph)(H)RhCl(CO)(PMe₃)₂ (2). Two reaction pathways to the formation of 2 were found. One of these was the prompt formation (<150 ns) of 2, the apparent result of a direct reaction of the electronic excited state of the four-coordinate species 1 with the C₆H₆ solvent. The second route was more a convoluted stepwise process, involving CO photodissociation to give the incoordinate intermediate RhCl-(PMe₃)₂ followed by benzene oxidative addition then CO addition to give 2.

Full text of DOI:10.1021/om970679o

5592
Organometallics 1997, 16, 5592-5594
Tim e-Resolved In fr a r ed Sp ectr a l Stu d ies of
P h otoch em ica lly In d u ced Oxid a tive Ad d ition of Ben zen e
to tr a n s-Rh Cl(CO)(P Me3)2
J on S. Bridgewater,† Brian Lee,† Stefan Bernhard,‡ J on R. Schoonover,‡ and
Peter C. Ford*,†
Department of Chemistry, University of California, Santa Barbara, California 93106, and
Bioscience/ Biotechnology Group (CST-4), Chemical Science and Technology Division,
Los Alamos National Laboratory, Los Alamos, New Mexico 87545
Received August 4, 1997X
Summary: Time-resolved infrared and time-resolved
optical spectroscopy were used to examine the pathway-
(s) by which 355 nm photolysis of the rhodium(I) species
trans-RhCl(CO)(PMe3)2 (1) in benzene leads to the C-H
oxidative-addition product (Ph)(H)RhCl(CO)(PMe3)2
(2). Two reaction pathways to the formation of 2 were
found. One of these was the “prompt” formation (<150
ns) of 2, the apparent result of a direct reaction of the
electronic excited state of the four-coordinate species 1
with the C6H6 solvent. The second route was more a
convoluted stepwise process, involving CO photodisso-
ciation to give the “tricoordinate” intermediate RhCl-
(PMe3)2 followed by benzene oxidative addition then CO
addition to give 2.
C-H activation step in the photocatalytic carbonylation
of benzene by RhCl(CO)L2 (eq 1).
hν, RhCl(CO)L2
C6H6 + CO
8 C6H5CHO
(1)
Among the mechanisms proposed for this reaction,
two pathways for the C-H activation step appear to be
the most credible. One mechanism is the oxidative
addition of a benzene C-H bond directly to an electronic
excited state [RhCl(CO)L2]* giving the hexacoordinate
Rh(III) transient species in a single step (Scheme 1).
Migratory insertion of CO into the resulting Rh-Ph
bond followed by reductive elimination of the acyl
hydride would give benzaldehyde. Although such ex-
cited states have been shown to be short lived (<∼250
ps),5 this would not be a limitation to a bimolecular
reaction since the substrate is the solvent.
The alternative mechanism involves initial photodis-
sociation of CO to give the “tricoordinate” intermediate
“RhClL2” (3, Scheme 2). Evidence that such a species
is considerably activated toward oxidative-addition
reactions comes from the demonstration in this labora-
tory that H2 addition to RhCl(PPh3)2 occurs ∼5 orders
of magnitude more rapidly than the analogous H2
addition to RhCl(PPh3)3.6 Theoretical studies appear to
confirm this point.4
Hexacoordinate complexes (Ph)(H)RhCl(CO)(PMe3)2
(2) have been identified by NMR spectroscopy3 to be
stable intermediates formed by photolysis of low-tem-
perature tetrahydrofuran/benzene solutions of trans-
RhCl(CO)(PMe3)2 (1). However, it should be noted that
both the single-step excited-state oxidative-addition
pathway (Scheme 1) and the stepwise mechanism
(Scheme 2) lead to the same product. The issue of which
scheme is preeminent can be addressed by flash pho-
tolysis studies of sufficient temporal resolution.
Selective catalytic functionalization of hydrocarbons
has long been a chemical “holy grail” of substantial
economic importance.1 The key step in any such scheme
is the activation of C-H bonds which may be ac-
complished by free-radical hydrogen-atom abstraction
and electrophilic attack or oxidative addition of a low-
valent transition metal. Oxidative addition has the
potential to provide the highest selectivity, although
most examples of C-H activation by this route have not
led to catalytic hydrocarbon functionalization. As a
consequence, the several reports that Rh(I) complexes
of the type trans-RhCl(CO)L2 (where L ) a trialkyl or
triaryl phosphine R3P) may serve as precursors to
catalytic or photocatalytic species for the functionaliza-
tion of various hydrocarbons have drawn considerable
experimental2,3 and some theoretical interest.4 De-
scribed here are the results of laser flash photolysis
studies using both time-resolved infrared (TRIR) and
time-resolved optical (TRO) detection to probe the key
† University of California, Santa Barbara.
‡ Los Alamos National Laboratory.
X Abstract published in Advance ACS Abstracts, December 1, 1997.
(1) Armdtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H.
Acc. Chem. Res. 1995, 28, 154-162.
FTIR investigations of low temperature (-60 °C)
solutions of 1 in THF/C6H6 (5/1 v/v) under conditions
analogous to those used for the NMR studies described
above demonstrated that photolysis with 355 nm light
led to the disappearance of the νCO band of 1 at 1962
(2) (a) Kunin, A. J .; Eisenberg, R. J . Am. Chem. Soc. 1986, 108, 535-
536; Organometallics 1988, 7, 2124-2129. (b) Sakakura, T.; Tanaka,
M. J . Chem. Soc., Chem. Commun. 1987, 758-759. (c) Sakakura, T.;
Sodeyama, T.; Tanaka, M. New J . Chem. 1989, 13, 737-745. (d)
Sakakura, T.; Sodeyama, T.; Sasaki, K.; Wada, K.; Tanaka, M. J . Am.
Chem. Soc. 1990, 112, 7221-7229. (e) Iwamoto, A.; Itagaki, H.; Saito,
Y. J . Chem. Soc., Dalton Trans. 1991, 1093-1097. (f) Maguire, J . A.;
Boese, W. T.; Goldman, A. S. J . Am. Chem. Soc. 1989, 111, 7088-
7093. (g) Boese, W. T.; Goldman, A. S. J . Am. Chem. Soc. 1992, 114,
350-351.
(3) (a) Boyd, S. E.; Field, L. D.; Partridge, M. G. J . Am. Chem. Soc.
1994, 116, 9492-9497. (b) Rosini, G. P.; Boese, W. T.; Goldman, A. S.
J . Am. Chem. Soc. 1994, 116, 9498-9505.
(4) (a) Sakaki, S.; Ujino, Y.; Sugimoto, M. Bull. Chem. Soc. J pn.
1996, 69, 3047-3057. (b) Margl, P.; Ziegler, T.; Blo¨chl, P. E. J . Am.
Chem. Soc. 1995, 117, 12625-12634; 1996, 118, 5412-5419. (c) Koga,
N.; Morokuma, K. Chem. Rev. 1991, 91, 823-842.
cm-1 and appearance of a strong band at 2070 cm-1
,
which we attribute to the formation of 2. (Note: No
new absorptions were seen in the 2000-2150 cm-1
region of the spectrum when the low-T photolysis was
carried out in neat THF.7) The NMR trapping studies
(5) Ford, P. C.; Netzel, T. L.; Spillett, C. T.; Pourreau, D. B. Pure
Appl. Chem. 1990, 62, 1091-1094. (b) Netzel, T. L. unpublished studies
(private communication to P.C.F.).
(6) Wink, D. A.; Ford, P. C. J . Am. Chem. Soc. 1987, 109, 436-442.
S0276-7333(97)00679-1 CCC: $14.00 © 1997 American Chemical Society

Communications
Organometallics, Vol. 16, No. 26, 1997 5593
Sch em e 1
Sch em e 2
Sch em e 3
a prompt rise (within the 150 ns rise time of the system),
consistent with direct reaction with the short-lived
excited state as shown in Scheme 1 (Figure 1 inset).
However, these data also indicate another process, a
“delayed” rise of the absorbance attributed to 2, which
could be fit to a first-order rate law with kobs ) (8.3 (
F igu r e 1. TRIR difference spectrum (recorded by the
stepped-scan technique) in the νCO region of a C6D6 solution
of trans-RhCl(CO)(PMe3)2 subsequent (∼0-100 µs) to 355
nm flash photolysis. The negative peak represents deple-
tion of 1, while the positive peak represents formation of
species attributed to 2. Inset: Temporal IR absorbance
change at 2070 cm-1 subsequent to 355 nm flash photolysis
of a C6H6 solution of trans-RhCl(CO)(PMe3)2, showing the
prompt and delayed fomation of νCO absorbance attributed
to 2 in this medium (see text). The vertical scale on both
the large figure and the inset is ∆OD × 103.
1) × 104 s-1
. The analogous pathway in perprotio-
benzene gave kobs ) (1.0 ( 0.1) × 105 s-1, indicating, at
most, a small deuterium isotope effect (kh/kd ≈ 1.3 (
0.3) as is often the case with oxidative-addition pro-
cesses.11 Under these conditions the ratio of the signals
due to prompt and delayed formation of 2 was about
1.2.12 However, when the analogous experiment was
carried under 0.1 atm CO of (7 × 10-4 M), the temporal
response of the absorbance at 2050 cm-1 changed; the
prompt rise remained, but the slower, first-order rise
was quenched.
These observations can be interpreted in terms of
Scheme 3 which incorporates key features of both
Schemes 1 and 2. Prompt formation of 2 via direct
reaction with [RhCl(CO)L2]* should be independent of
[CO], while CO quenching of the slower stepwise
pathway can be attributed to formation of the CO-
dissociated intermediate 3. This intermediate reacts
competitively by benzene oxidative addition to give 2
and CO addition to regenerate 1. Under argon, the first-
order oxidative-addition pathway of 3 is quite competi-
indicated two isomers of 2; however, the FTIR spectra
displayed only one broad band, which may be due to
similar νCO frequencies for the isomers. Two other new
bands, approximately one-fourth the intensity of that
at 2070 cm-1, appear at 2130 and 2018 cm-1. These
bands are consistent with Rh-H stretching frequencies
reported for similar compounds.8 Consistent with this
assignment is the absence of these bands when the
photolysis was carried out in C6D6. The νCO band
appeared at lower frequency (2050 cm-1), perhaps owing
to the absence of Fermi coupling to another vibration
of comparable frequency.
Figure 1 illustrates the result of laser flash photolysis
(355 nm) of 1 in benzene-d6 solution under Ar using a
stepped-scan infrared spectrometer for TRIR detection.9
The transient difference spectrum displays a bleaching
at 1962 cm-1 representing the loss of 1 and a new, broad
absorbance at 2050 cm-1 corresponding to the formation
of an intermediate. When temporal absorbance changes
in the IR were examined using a tunable single-fre-
quency probe source,10 formation of the transient shows
(10) (a) DiBenedetto, J .; Ryba, D. W.; Ford, P. C. Inorg. Chem. 1989,
28, 3503-3507. (b) Ford, P. C.; DiBenedetto, J . A.; Ryba, D. W.; Belt,
S. T. SPIE Proceedings 1992, 1636, 9-16. (c) Boese, W. T.; Ford, P. C.
J . Am. Chem. Soc. 1995, 117, 8381-8391.
(11) (a) Tolman, C. A. In The Hydrogen Series; Muetterties, E. L.,
Ed.; Marcel Dekker: New York, 1971; Vol. 1, pp 271-312. (b)
Rosenberg, E. Polyhedron 1989, 8, 383-405. (c) Wink, D.; Ford, P. C.
J . Am. Chem. Soc. 1985, 107, 1794-1796. (d) Zhou, P.; Vitale, A. A.;
Filippo, J . S., J r.; Saunders, W. H., J r. J . Am. Chem. Soc. 1985, 107,
8049-8054.
(12) The first-order nature of the slow rise suggests that the rate-
limiting step for the formation of complex 2 is the oxidative-addition
step and not the second-order reaction of intermediate I with carbon
monoxide. The quantum yield for the loss of CO was reported to be
0.1 in cyclohexane. By comparing the amount of complex 2 formed via
the prompt process versus the stepwise process and assuming the same
quantum yield for CO loss in benzene, an upper limit of 0.12 can be
placed on the quantum yield for the production of complex 2 via the
prompt process. The mixed-order kinetics reinforces that the quantum
yield of 0.12 for the formation of complex 2 via the prompt process is
an upper limit.
(7) It should be noted that after very long term photolysis of 1 in
low temperature THF, previous workers did observe some photoprod-
ucts but the photolysis times used in those experiments were about 2
orders of magnitude longer than in the present experiments, suggesting
that these photoreaction pathways are quite inefficient in THF.
(8) (a) Kaesz, H. D.; Saillant, R. B. Chem. Rev. 1972, 72, 231-281.
(b) Intille, G. M. Inorg. Chem. 1972, 11, 695-702.
(9) Schoonover, J . R.; Strouse, G. F.; Omberg, K. M.; Dyer, R. B.
Comm. Inorg. Chem. 1996, 18, 165-188.

5594 Organometallics, Vol. 16, No. 26, 1997
Communications
tive with second-order recombination with the labilized
CO, but at excess CO concentrations as low as 7 × 10-4
M CO, the latter pathway apparently dominates. The
features of Scheme 3 are also evident when a benzene
solution of 1 (0.5 mM) was subjected to flash-photolysis
excitation (355 nm) using TRO detection. The difference
spectrum in the visible region shows a transient absorp-
tion from 400 to 550 nm with a λmax at 430 nm. Under
Ar, the decay of this peak could be fitted equally well
as a first-order decay or as a second-order decay, but
under CO, the decay could be fit only to a first-order
rate law. (Note: the spectrum did not fully decay to
the base line within the TRO detection limits (∼50 ms),
even under higher [CO]). Longer-lived intermediate(s)
consistent with the formation of 2 persisted, although
comparisons of spectra before and after photolysis using
a standard optical spectrometer indicated eventual
decay to the base line. The kobs values obtained for the
decay of the first transient were found to be first order
in [CO]; a plot of kobs vs [CO] is a straight line with a
slope (kCO) of (1.2 ( 0.1) × 109 M-1 s-1 and a non-zero
not mutually exclusive mechanisms. Under the ap-
propriate conditions both are effective, but under condi-
tions designed to favor hydrocarbon carbonylation (ex-
cess CO), the unsaturated species 3 would be trapped
by CO at a rate sufficient to minimize its participation
in C-H activation. Thus, another mechanism such as
direct activation of the hydrocarbon by the four-coor-
dinate excited state3b must be responsible for the low-
yield photocatalytic carbonylations of aromatic hydro-
carbons reported. Nonetheless, in the absence of CO
and other ligands with a high affinity for Rh(I), 3 is
likely to play an important role in the activation of
various C-H bonds. Studies describing the detailed
kinetics of various intermediates of this type will be
reported.14
Ack n ow led gm en t. This research was sponsored by
a grant (Grant No. DE-FG-03-85ER13317) to P.C.F.
from the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy, and by a
Collaborative UC/Los Alamos Research (CULAR) Initia-
tive grant from Los Alamos National Laboratory. TRO
experiments were carried out on systems constructed
with instrumentation grants from the U.S. Department
of Energy University Research Instrumentation Grant
Program (Grant No. DE-FG-05-91ER79039) and from
the National Science Foundation. We thank T. L. Netzel
of Georgia State University for communicating data
cited in ref 5b.
intercept of ∼1.5 × 105 s-1 13
.
The observation of a non-zero intercept for this plot
would be consistent with partitioning of intermediate
3 between the second-order trapping by CO to reform 1
and other CO-independent pathways, one being an
oxidative-addition pathway to give 2. However, the
TRO technique is not as effective as TRIR detection in
separating the various processes, owing to the similarity
and overlap of the electronic spectra of reactants,
products, and intermediates. As a consequence, the
intercept (which is prone to sizable experimental un-
certainty) is likely to reflect other pathways as well as
benzene oxidative addition to give 2.13
OM970679O
(13) This intercept was obtained for flash photolysis studies with a
low concentration of 1 (0.1 mM). Higher values of the intercept (but
the same value of the slope) were found when higher initial concentra-
tions of 1 were used. This observation suggests that, under such
conditions, trapping of the intermediate 3 by excess 1 to form dimers
is a significant contributor to the CO-independent decay pathways
sampled by the TRO experiment.
In summary, these data indicate that the alternative
C-H activation steps described by Schemes 1 and 2 are
(14) Bridgewater, J . S.; Ford, P. C. Manuscript in preparation.