Oxy-functionalization of nucleophilic rhenium(I) metal carbon bonds catalyzed by selenium(IV)

Full text of DOI:10.1021/ja806814c

Published on Web 01/22/2009
Oxy-Functionalization of Nucleophilic Rhenium(I) Metal Carbon Bonds
Catalyzed by Selenium(IV)
William J. Tenn III,† Brian L. Conley,† Claas H. Ho¨velmann,§ Ma˚rten Ahlquist,‡ Robert J. Nielsen,‡
Daniel H. Ess,‡,§ Jonas Oxgaard,‡ Steven M. Bischof,§ William A. Goddard III,*,‡ and Roy A. Periana*,§
Department of Chemistry, UniVersity of Southern California, Los Angeles, California 90089, DiVision of Chemistry
and Chemical Engineering, California Institute of Technology, Pasadena, California 91125, and Scripps Energy
Laboratories, The Scripps Research Institute, Jupiter, Florida 33458
Received September 10, 2008; E-mail: rperiana@scripps.edu; wag@wag.caltech.edu
Electrophilic C-H bond activation (CHA) with strong electro-
philes,1,2 such as PtII, PdII, AuIII, and HgII, followed by reductive
oxy-functionalization (OF) of the resulting positively polarized
metal alkyl (M-Rδ+) intermediate is a well-known strategy for
hydrocarbon hydroxylation.3 However, water and alcohol inhibition
of the CHA reaction renders this strategy commercially impractical.
We4 and others5 have shown that CHA with weakly electrophilic
cations, such as IrIII and RuII, are less prone to such inhibition.
However, the reductive OF reactions utilized with highly electro-
philic catalysts are generally not applicable to weakly electrophilic
systems because the reduced oxidation potential thermodynamically
disfavors reductive functionalization and/or the reduced electro-
philicity can result in M-Rδ- polarized bonds that exhibit high
barriers for nucleophilic attack at the carbon.6 Therefore we have
set out to design new approaches for functionalization of M-Rδ-
polarized bonds.
developed to addresses these types of undesired reactions. We chose
to study the OF reactions of the (CO)5ReIR and the more electron-
rich analogue (CO)3(bpy)ReIR motifs because our screens showed
these complexes were relatively soluble and sufficiently stable in
aqueous media to allow reactions to be examined without special
techniques (Table 1). Here we report conditions for the facile,
stochiometric as well as catalytic OF of Re-Rδ- complexes to the
corresponding alcohols via a TRF mechanism.
Table 1. TRF Reactions Studieda
We recently proposed a new strategy for facile, nonredox, OF
of electron-rich M-C bonds that involves insertion of O-electro-
philes, YO.6 A related strategy is to functionalize by alkyl transfer
to a more electrophilic, redox active, Z-X co-catalyst, followed
by oxidation and reductive OF from Z-R (Scheme 1). This trans-
alkylation reductive functionalization (TRF) sequence, when coupled
to the C-H bond activation and oxidant regeneration reactions
(dashed lines, Scheme 1), could lead to tandem catalytic systems
for alkane hydroxylation.
a General reaction conditions: 0.01 mmol Re complex (1 equiv) and
SeO2 (2 equiv) in a J-Young NMR tube are dissolved in 0.45 mL of
CD3CN/D2O (9:1). The reaction is heated to 100 °C for the
pentacarbonyl complexes and stirred at rt for the bpy complexes. The
yield of the seleninic acids is determined by H NMR comparison to an
external standard. NaIO4 (2 equiv) is added and the reaction is heated to
Scheme 1. Generalized Scheme for Coupling a TRF Catalytic
Cycle (Solid Lines) to CHA and Oxidant Regeneration Reactions
(Dashed Lines)
1
100 °C for 14 h for oxidation to the alcohols.
In contrast to earlier reports of high yield OF reactions of
CH3ReO3 that occur via O-atom insertion,6 reactions of
(CO)5ReCH3 with electrophilic O-donors, such as PhIO and NaIO4,
in a 9:1 acetonitrile/water solution at 100 °C gave low yields of
methanol (∼20-30%, see Supporting Information, SI), and PyO
yielded no methanol. Oxidants such as CuI/CuII and PdII salts in
the presence or absence of air did not yield methanol. We considered
reactions with selenium oxides since (1) similar to SO2, SeO2 is
known to react with M-R to generate R-SeIVspecies,7 (2)
R-SeIV(O)OH can be oxidized to R-SeVI(O)2OH that readily
converts to alkanols and SeIV,8 and (3) Se oxides are commercially
utilized as co-catalysts in hydrocarbon oxidation reactions.9,10
The reaction of (C-O)5ReCH3 with excess SeO2, which exists
as D2SeIVO3 in D2O,11 in a 9:1 v/v solution of CD3CN/D2O
produced methaneseleninic acid, CH3-SeIVO2D, in quantitative
We focused on designing TRF reactions with d6, octahedral, low
oxidation state M-Rδ- complexes that would be good models for
M-R intermediates generated from weakly electrophilic, water
insensitive CHA systems.4 Key to facilitating study of these M-R
complexes is identifying systems where undesired reactions, such
as metal centered oxidations and alkyl protonolysis, can be
minimized. In actual catalytic systems, strategies will have to be
1
yield as identified by comparison of the H, 13C NMR, and GC/
† University of Southern California.
‡ California Institute of Technology.
§ The Scripps Research Institute.
MS spectra to an authentic sample (Table 1). This reaction was
slow at room temperature, but heating to 50 °C for 2 h gave nearly
9
2466 J. AM. CHEM. SOC. 2009, 131, 2466–2468
10.1021/ja806814c CCC: $40.75
2009 American Chemical Society

C O M M U N I C A T I O N S
1
quantitative formation of CH3SeO2D as monitored by H NMR.
with SeO2 gives fast quantitative methyl transfer, but attempted
reaction of (CO)3(bpy)ReCH3 with IO4 and catalytic amounts of
H2SeO3 at 100 °C led only to formation of methane. We suspect
Control experiments using (CO)5Re13CH3 confirmed that the
CH3SeO2D was derived from the (CO)5ReCH3 and not from side
reactions with the solvent. The formation of CH3SeO2D followed
first-order kinetics with an experimental ∆H‡ ) 8.0 ( 1.7 kcal/
mol. Given the observed ∼100% methyl mass balance, the likely
reaction stoichiometry is shown in eqs 1 and 2.
-
that this is the result of a higher rate of protonolysis for
δ-
(CO)3(bpy)ReCH3 vs (CO)5ReCH3 (due to a more basic Re-CH3
bond) even though the transalkylation reactions with H2SeO3 are
also more efficient. In actual tandem catalytic systems, when
coupled to a reVersible CHA system, this M-R protonolysis will
not prevent product formation. The relative reactivity of
(CO)5ReICH3 and (CO)3(bpy)ReICH3 with SeO2 and oxidants is
currently being examined in detail.
2H2SeO3 + (CO)5ReCH3 f (CO)5Re(OSeO2H) +
CH3SeO2H + H2O (1)
(CO)5Re(OSeO2H) + H2O f (CO)5Re(OH) + H2SeO3 (2)
To examine the scope of these OF reactions, we also studied the
reaction of (CO)3(bpy)Re-R complexes where R ) methyl, ethyl,
and n-propyl (Table 1). Of particular interest was whether primary
rhenium alkyls would generate primary, rather than rearranged, seleno-
functionalized products upon treatment with SeO2. Significantly, as
in the case of R ) CH3, reaction of the ethyl and n-propyl ReI
complexes led to essentially quantitative yields of the corresponding
seleno-functionalized product, RSeIV(O)OH, where R is ethyl and
n-propyl, respectively. Since SeIV is a relatively strong oxidant and
used for oxidation of activated CH bonds in organic chemistry,9 it is
significant that no rearrangements or oxidation of the ethyl and n-propyl
groups take place during functionalization. Equally significant, treat-
ment of the resulting RSeIV(O)OH solutions with IO4- led to
regioselective formation of the corresponding alcohols, CH3CH2OH
and CH3CH2CH2OH, Table 1. These results are encouraging and show
the potential for SeIV based tandem CHA/TRF catalysis that can lead
to alcohols where the regioselectivity is set by the CH activation step.
This observation, coupled with the known higher selectivity for CHA
for primary over secondary or tertiary C-H bonds with weakly
electrophilic catalysts,1,4 could suggest methods for the selective C–H
functionalization of alkanes to primary alcohols.
The expected Re products, (CO)5Re(OSeO2H) or (CO)5ReOH,
could not be directly observed, likely due to known oligomerization
reactions12 and displacement of CO ligands by the coordinating
solvent. Significantly, no methane was observed in the headspace
or solution. This, along with the high reaction yield, suggests that
the transalkylation reaction with D2SeO3 is much faster than
protonolysis of the Re-CH3 bond or oxidation at the Re center by
D2SeO3. To directly obtain methanol from (CO)5ReICH3 we
examined whether it would be possible to in situ oxidize
CH3SeIV(O)OH to CH3SeVI(O)2OH which could generate methanol
and regenerate H2SeIVO3. Significantly, heating (CO)5ReCH3 to 100
°C for 12 h in an ∼8:2 v/v mixture of CD3CN/D2O with excess
IO4- along with a catalytic amount of D2SeO3 (10 mol %) gave an
∼80% yield of methanol. Controls carried out in the dark show
the reactions are not light induced. CH3SeIVO2H was observed as
1
a reaction intermediate by H NMR, indicating that oxidation of
CH3SeIVO2H to methanol, likely via CH3SeVI(O)2OH, is the rate
limiting step in the overall conversion to methanol. Control
experiments confirmed that treatment of CH3SeIV(O)OH with IO4
-
under the conditions of the catalytic reaction generates methanol
without observation of CH3SeVI(O)2OH (Scheme 2).8
To determine if the seleno-functionalization could be also
extended to aryl rhenium bonds, we examined the reaction of SeO2
with (CO)3(bpy)Re-R, R ) phenyl and para-tolyl. These complexes
also reacted quantitatively and rapidly to generate the corresponding
RSe(O)OH products. Significantly, R ) para-tolyl led to exclusive
formation of the para-tolyl seleninic acid, p-CH3C6H4Se(O)OH,
and shows that functionalization likely proceeds by attack at the
ipso carbon of the aryl ring. Treatment of (CO)3(bpy)Re-Ph with
catalytic amounts of SeO2 and excess IO4- did not generate phenol
and gave only PhSeIV(O)OH and PhSeVI(O)2OH. Consistent with
this observation, control studies show that, unlike CH3SeVI(O)2OH
(Vide infra), C6H5SeVI(O)2OH is a thermally stable species in
aqueous solution and can be readily synthesized by treatment of
Scheme 2. Proposed Mechanism of Methanol Formation
Seeking a more reactive system that could also generate a well-
defined Re product that could be characterized, we reasoned that
replacing CO groups with more electron-donating, less labile bipyridine
(bpy) should increase the nucleophilicity of the Re-CH3δ- group and
lead to cleaner, faster reactions with electrophiles such as SeIV.
Consequently, we examined the reaction of (CO)3(bpy)ReCH313 with
1 equiv of H2SeO3 in a 9:1 mixture of CH3CN/H2O. This system was
indeed found to be more reactive, and the reaction was very efficient
at rt and even to temperatures as low as -35 °C to generate
(CO)3(bpy)Re(OSe(O)CH3) in high yield on mixing. This species was
isolated as a fine yellow solid in 65% yield after precipitation by diethyl
14
PhSeIV(O)OH with oxidants such as MnO4 (or IO4-). Since
methods to convert these aryl metal complexes to phenol will be
important, efforts are underway to identify conditions to convert
(CO)5RePh, PhSeIV(O)OH, and/or PhSeVI(O)2OH to phenol.
-
The mechanism for the transalkylation step was also studied using
density functional theory.15 The most favorable reaction pathway
was found to be nucleophilic attack by Re-CH3δ- on the electro-
philic SeIV center (Figure 1). All reactions involving prior dissocia-
tion of CO are energetically disfavored. Consistent with this,
experiments run at 70 °C under 1 atm of CO gas showed no
decrease in rate relative to those without added CO. The initial
step in the reaction mechanism is the conversion of ground state
H2SeO3 to the more electrophilic SeO2, with the H2O weakly bound
to selenium (Se-OH2 2.6 Å). As expected, this reaction was
calculated to be endothermic with a ∆H of 5.5 kcal/mol. The
electrophilic Se center is then attacked by the negatively polarized
methyl group of (CO)5Re-CH3δ-. We found that keeping the water
molecule associated with the SeO2 moiety in the transition state
1
ether and characterized by H NMR and elemental analysis. Use of
excess H2SeO3 produced CH3SeO2H and (CO)3(bpy)Re(OSeO2H)
which was also isolated in high yield and characterized analogously.
It is significant that the rate of methyl transfer can be substantially
increased by the use of more electron donating ligands since fast
reactions will be required for the trapping of M-R intermediates
when coupling functionalizations to CHA reactions. Also relevant
to eventual coupling of these reactions to CHA is the lack of
competitive protonolysis or oxidation of the Re(I) center. However,
at low temperatures the stoichiometric reaction of (CO)3(bpy)ReCH3
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2467

C O M M U N I C A T I O N S
lowers the reaction barrier by 3.0 kcal/mol, most likely due to a
stabilization of the partial negative charge built up at the oxygen
atoms. We calculated the reaction to have an enthalpic barrier of
7.6 kcal/mol, giving an overall barrier of 13.2 kcal/mol relative to
selenious acid and (CO)5ReCH3. This is consistent with our
experimental ∆H‡ value of 8.0 ( 1.7 kcal/mol.16 For the overall
reaction from H2SeO3 and (CO)5ReCH3 we calculated an entropic
barrier of ∼4.4 kcal/mol (T∆S‡ ) -14.8 eu) (298 K), yielding a
∆G‡ of 17.6 kcal/mol at rt. This relatively large negative entropy
is consistent with the associative nature of the transition state.
In conclusion, we have identified a facile, catalytic OF route for
ReI-CH3 bonds with SeIV and iodine based O-atom donors. We
are exploring strategies to incorporate this reaction sequence into
catalytic cycles for the overall conversion of hydrocarbons to
alcohols utilizing low-valent, weakly electrophilic transition metals
for C-H bond activation.17
Acknowledgment. We thank Chevron Company and Scripps
Florida for financial support, Prof. William C. Kaska for helpful
suggestions, and Ariana Nussdorf for work on kinetic analysis. M.A.
acknowledges The Knut and Alice Wallenberg Foundation. This work
was supported by a fellowship within the Postdoc-Programme of the
German Academic Exchange Service (DAAD) supporting C.H.H.
Supporting Information Available: Synthetic procedures, experi-
mental and computational details. This material is available free of
charge via the Internet at http://pubs.acs.org.
References
(1) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H. Acc.
Chem. Res. 1995, 28, 154 and citations therein.
(2) Crabtree, R. H. J. Organomet. Chem. 2004, 689, 4083.
(3) (a) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879. (b) Periana,
R. A.; Mironov, O.; Taube, D.; Bhalla, G.; Jones, C. Science 2003, 30,
814. (c) Lin, M.; Hogan, T.; Sen, A. J. Am. Chem. Soc. 1997, 119, 6048.
(d) Muehlhofer, M.; Strassner, T.; Herrmann, W. A. Angew. Chem., Int.
Ed. 2002, 41, 1745.
(4) (a) Tenn, W. J., III.; Young, K. J. H.; Bhalla, G.; Oxgaard, J.; Goddard,
W. A., III; Periana, R. A. J. Am. Chem. Soc. 2005, 127, 14172. (b) Tenn,
W. J., III; Young, K. J. H.; Oxgaard, J.; Nielsen, R. J.; Goddard, W. A.,
III; Periana, R. A. Organometallics 2006, 25, 5173.
(5) (a) Klei, S. R.; Golden, J. T.; Tilley, D.; Bergman, R. G. J. Am. Chem.
Soc. 2002, 124, 2092. (b) Feng, Y.; Lail, M.; Barakat, K. A.; Cundari,
T. R.; Gunnoe, T. B.; Petersen, J. L. J. Am. Chem. Soc. 2002, 127, 14174.
(c) Kloek, S. M.; Heinekey, D. M.; Goldberg, K. I. Angew. Chem., Int.
Ed. 2007, 46, 4736.
Figure 1. Proposed mechanism for methyl transfer to H2SeO3.
A scan of the intrinsic reaction coordinate (IRC) shows that the
transition state collapses to a structure where one of the oxygen
atoms of the methylseleninic moiety coordinates to Re (Figure 2,
bottom right). The overall reaction of (CO)5ReCH3 with selenious
acid to form the O-methylseleninate complex is exothermic by 21.5
kcal/mol. Interestingly, we found that rearrangement to the S-methyl-
seleninate is unfavorable by 15.4 kcal/mol.
(6) (a) Conley, B. L.; Ganesh, S. K.; Gonzales, J. M.; Tenn, W. J., III; Young,
K. J. H.; Oxgaard, J.; Goddard, W. A., III; Periana, R. A. J. Am. Chem.
Soc. 2006, 128, 9018. (b) Conley, B.; Ganesh, S. K.; Gonzales, J. M.; Ess,
D. H.; Nielsen, R. J.; Ziatdinov, V. R.; Oxgaard, J.; Goddard, W. A., III;
Periana, R. A. Angew. Chem., Int. Ed. 2008, 41, 7849.
(7) (a) Bibler, J. P.; Wojcicki, A. J. Am. Chem. Soc. 1964, 86, 5051. (b)
Wojcicki, A. Acc. Chem. Res. 1971, 4, 344. Lorenz, I.-P. Angew. Chem.,
Int. Ed. Engl. 1978, 17, 53.
(8) Bird, L.; Challenger, F. J. Chem. Soc. 1942, 570.
(9) (a) Sharpless, K. B.; Gordon, K. M. J. Am. Chem. Soc. 1976, 98, 300. (b)
Corey, E. J.; Schaefer, J. P. J. Am. Chem. Soc. 1960, 82, 918. (c) Selenium
in Natural Products Synthesis, Nicolaou, K. C.; Petasis, N. A.; CIS;
Philadelphia, 1984. (d) Sharpless, K. B.; Lauer, R. F. J. Am. Chem. Soc.
1972, 94, 7154. (e) Umbreit, M. A.; Sharpless, K. B. J. Am. Chem. Soc.
1977, 99, 5526.
(10) (a) Patel, B. M.; Price, G. L. Ind. Eng. Chem. Res. 1990, 29, 730. (b) Mann,
R. S.; Lao, K. C. Ind. Eng. Chem. Res. 1967, 6, 263.
(11) Waitkins, G. R.; Clark, C. W. Chem. ReV. 1945, 36, 235.
(12) Beck, W.; Raab, K.; Nagel, U.; Steinmann, M. Angew. Chem., Int. Ed.
Engl. 1982, 21, 526.
Figure 2. Transition state (top left) and snapshot views from the IRC of
the subsequent decomposition to O-seleninate product.
(13) (a) Lucia, L. A.; Burton, R. D.; Schanze, K. S. Inorg. Chim. Acta 1993,
208, 103. (b) Worl, L. A.; Duesing, R.; Chen, P.; Della Ciana, L.; Meyer,
T. J. J. Chem. Soc., Dalton Trans. 1991, 849.
(14) Haas, A.; Schinkel, K. Chem. Ber. 1990, 123, 685.
(15) (a) All calculations were performed using the B3LYP hybrid density
functional as implemented in Jaguar 6.5 and 7.5. For all atoms except Se,
the LACV3P** basis set was used for geometry optimizations and solvation
energies, and LACV3P**++ for single-point gas phase energy corrections.
For Se, the augmented MSV all electron basis set was used. See Supporting
Information for details. (b) Jaguar, version 7.5; Schrodinger, LLC: New
York, 2008.
(16) We believe that part of the discrepancy between the experimental and
theoretical activation enthalpy is due to the theoretical solvation model.
This computes the free energy of solvation and therefore contains entropic
contributions from solvent rearrangement.
Figure 3. Solvent (MeCN) optimized transition states for O-atom transfer
and H2O attack.
(17) There have been several recent examples of Re mediated C-H activation.
For example: (a) Kuninobu, Y.; Nishina, Y.; Matsuki, T.; Takai, K. J. Am.
Chem. Soc. 2008, 130, 14062. (b) Kuninobu, Y.; Nishina, Y.; Okaguchi,
K.; Shouho, M.; Takai, K. Bull. Chem. Soc. Jpn. 2008, 81, 1393. (c)
Kuninobu, Y.; Nishina, Y.; Nakagawa, C.; Takai, K. J. Am. Chem. Soc.
2006, 128, 12376. (d) Kuninobu, Y.; Nishina, Y.; Shouho, M.; Takai, K.
Angew. Chem., Int. Ed. 2006, 45, 2766. (e) Kuninobu, Y.; Tokunaga, Y.;
Kawata, A.; Takai, K. J. Am. Chem. Soc. 2006, 128, 202. (f) Kuninobu,
Y.; Kawata, A.; Takai, K. J. Am. Chem. Soc. 2005, 127, 13498. (g)
Gonzales, J. M.; Oxgaard, J.; Periana, R. A.; Goddard, W. A., III.
Organometallics 2007, 26, 1505.
The direct O-atom transfer between CH3SeIV(O)OH and IO4- to
generate CH3SeVI(O)2OH has an activation free energy of 24 kcal/
mol, relative to the methyl SeIV intermediate, which is consistent with
the need to heat the reaction to 100 °C for several hours to give
methanol (Figure 3). Selenium oxidation is most likely followed by
SN2 attack of the CH3SeVI bond by water (∆G‡ ) 18 kcal/mol). An
alternative pathway without Se oxidation via a Baeyer-Villiger type
mechanism has a large activation barrier (∆G‡ ) 53 kcal/mol).
JA806814C
9
2468 J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009