Catalytic carbon-carbon and carbon-hydrogen bond cleavage in lower alkanes. Low-temperature hydroxylations and hydroxycarbonylations with dioxygen as the oxidant

Article abstract of DOI:10.1021/ja953670r

RhCl₃, in the presence of several equivalents of Cl^- and I^- ions, catalyzed the direct formation of methanol and acetic acid from methane, carbon monoxide, and dioxygen at 80-85°C in a 6:1 mixture of perfluorobutyric acid and water (approximate turnover rate: 2.9/h based on Rh). It was possible to selectively form either methanol or acetic acid by a simple change in the solvent system. As might be anticipated, ethane was more reactive than methane, and under similar reaction conditions formed methanol, ethanol, and acetic acid (approximate turnover rate: 7.5/h based on Rh). For both methane and ethane, the product alcohols were less reactive than the starting alkanes. Methyl iodide was also less reactive than methane. Most significantly, for ethane and higher alkanes products derived from C-C cleavage dominated over those derived from C-H cleavage on a per bond basis. Indeed, C-C cleavage products were virtually all that were observed with butane, isopentane, and 2,3-dimethylbutane. While the mechanism of the C-H and C-C cleavage steps remains to be elucidated, preliminary indications are that outer-sphere electron transfer or bond homolysis resulting in the formation of alkyl radicals did not occur.

Full text of DOI:10.1021/ja953670r

4574
J. Am. Chem. Soc. 1996, 118, 4574-4580
Catalytic Carbon-Carbon and Carbon-Hydrogen Bond
Cleavage in Lower Alkanes. Low-Temperature Hydroxylations
and Hydroxycarbonylations with Dioxygen as the Oxidant
Minren Lin, Terrence E. Hogan, and Ayusman Sen*
Contribution from the Department of Chemistry, The PennsylVania State UniVersity,
UniVersity Park, PennsylVania 16802
ReceiVed October 31, 1995X
Abstract: RhCl3, in the presence of several equivalents of Cl- and I- ions, catalyzed the direct formation of methanol
and acetic acid from methane, carbon monoxide, and dioxygen at 80-85 °C in a 6:1 mixture of perfluorobutyric
acid and water (approximate turnover rate: 2.9/h based on Rh). It was possible to selectively form either methanol
or acetic acid by a simple change in the solvent system. As might be anticipated, ethane was more reactive than
methane, and under similar reaction conditions formed methanol, ethanol, and acetic acid (approximate turnover
rate: 7.5/h based on Rh). For both methane and ethane, the product alcohols were less reactive than the starting
alkanes. Methyl iodide was also less reactive than methane. Most significantly, for ethane and higher alkanes
products deriVed from C-C cleaVage dominated oVer those deriVed from C-H cleaVage on a per bond basis.
Indeed, C-C cleavage products were Virtually all that were observed with butane, isopentane, and 2,3-dimethylbutane.
While the mechanism of the C-H and C-C cleavage steps remains to be elucidated, preliminary indications are
that outer-sphere electron transfer or bond homolysis resulting in the formation of alkyl radicals did not occur.
This report encompasses the highly catalytic, low temperature
two of the highest volume functionalized organics produced
commercially are methanol and acetic acid whose 1993 U.S.
hydroxylations and hydroxycarbonylations of C-H and C-C
bonds in lower alkanes with dioxygen as the oxidant.1 The
simultaneous cleavage of C-H and C-C bonds of alkanes, as
well as O2 activation, has been achieved. We are unaware of
any report of such a highly catalytic system for the direct
functionalization of methane and ethane by dioxygen at low to
moderate temperatures (e100 °C) although the use of other
oxidants has been reported under these conditions.2 Most
significantly, our observations constitute the first examples of
metal catalyzed oxidative functionalization of C-C bonds of
simple alkanes in solution although several instances of C-C
cleavage by metal complexes have been reported previously.3
The lower alkanes, such as methane and ethane, are the least
reactive and most abundant of the hydrocarbon family with
known reserves equal to that of petroleum.4 Thus, the selective
oxidative functionalization of these alkanes to more useful
chemical products is of great practical interest.5 For example,
productions were 10.5 × 109
and 3.7 × 109 lbs, respectively.6
The current technology for the conversion of alkanes to these
products involves multi-step processes: (a) the high-temperature-
steam reforming of alkanes to a mixture of H2 and CO,7 (b) the
high-temperature conversion of the mixture of H2 and CO to
methanol,7 and (c) the carbonylation of methanol to acetic acid,8
mainly through the “Monsanto process”.9 Clearly, the direct,
low temperature conversion of the lower alkanes to methanol
and acetic acid would be far more attractive from an economical
standpoint. Of particular interest would be the formation of
the same end product(s) from different starting alkanes, thus
obviating the need to separate the alkanes. For example, natural
gas is principally methane with 5-10% ethane. A system that
converts both methane and ethane to the same C1 product, such
as methanol, would not require the prior separation of the
alkanes. Of course, the formation of C1 products from ethane
and higher alkanes would require an unprecedented catalytic
cleavage and oxidation of C-C bonds which we have now
achieved. Additionally, catalytic C-C cleavage is the key step
X Abstract published in AdVance ACS Abstracts, May 1, 1996.
(1) Preliminary account of conversion of methane to acetic acid in
aqueous medium: Lin, M.; Sen, A. Nature 1994, 368, 613.
(2) (a) Nishiguchi, T., Nakata, K., Takaki, K.; Fujiwara, Y. Chem. Lett.
1992, 1141. (b) Goldstein, A. S.; Drago, R. S. J. Chem. Soc., Chem.
Commun. 1991, 21. (c) Kao, L.-C.; Hutson, A. C.; Sen, A. J. Am. Chem.
Soc. 1991, 113, 700. Examples at higher temperatures (∼200 °C): (d)
Vargaftik, M. N.; Stolarov, I. P.; Moiseev, I. I. J. Chem. Soc., Chem.
Commun. 1990, 1049. (e) Periana, R. A.; Taube, D. J.; Evitt, E. R.; Lo¨ffler,
D. G.; Wentrcek, P. R.; Voss, G.; Masuda, T. Science 1993, 259, 340.
Also: Snyder, J. C.; Grosse, A. V. U.S. Patent 2,493,038, 1950.
(3) (a) Perthuisot, C.; Jones, W. D. J. Am. Chem. Soc. 1994, 116, 3647.
(b) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10015.
(c) Chiang, H. J.; Liu, C.-S. J. Organomet. Chem. 1993, 458, 167. (d) Gozin,
M.; Aizenberg, M.; Liou, S.-Y.; Weisman, A.; Ben-David, Y.; Milstein, D.
Nature 1994, 370, 42. (e) Gozin, M.; Weisman, A.; Ben-David, Y.; Milstein,
D. Nature 1993, 364, 699. (f) Rondon, D.; Chaudret, B.; He, X.-D.; Labroue,
D. J. Am. Chem. Soc. 1991, 113, 5671. (g) Hartwig, J. F., Andersen, R. A.;
Bergman, R. G. J. Am. Chem. Soc. 1989, 111, 2717. (h) Bunel, E.; Burger,
B. J.; Bercaw, J. E. J. Am. Chem. Soc. 1988, 110, 976. (i) Crabtree, R. H.,
Dion, R. P., Gibboni, D. J., McGrath, D. V.; Holt, E. M. J. Am. Chem.
Soc. 1986, 108, 7222. (j) Suggs, J. W.; Jun, C.-H. J. Am. Chem. Soc. 1986,
108, 4679. (k) Watson, P. L.; Roe, D. C. J. Am. Chem. Soc. 1982, 104,
6471.
(4) (a) Axelrod, M. G.; Gaffney, A. M.; Pitchai, R.; Sofranko, J. A. In
Natural Gas ConVersion II; Curry-Hyde, H. E.; Howe, R. F., Eds.;
Elsevier: Amsterdam, 1994; p 93. (b) Masters, C. D.; Root, D. H.; Attanasi,
E. D. Science 1991, 253, 146. (c) Starr, C.; Searl, M. F.; Alpert, S. Science
1992, 256, 981.
(5) Recent reviews: (a) Olah, G. A.; Molna´r, A. Hydrocarbon Chemistry;
Wiley: New York, 1995. (b) Crabtree, R. H. Chem. ReV. 1995, 95, 987.
(c) Labinger, J. A. Fuel Process. Technol. 1995, 42, 325. (d) Fierro, J. L.
G. Catal. Lett. 1993, 22, 67. (e) Srivastava, R. D.; Zhou, P.; Stiegel, G. J.;
Rao, V. U. S.; Cinquegrane, G. Catalysis (London) 1992, 9, 183.
(6) Chem. Eng. News 1994, July 4, 31.
(7) (a) Methanol Production and Use; Cheng, W.-H.; Kung, H. H., Eds.;
Marcel Dekker: New York, 1994. (b) Wade, L. E.; Gengelbach, R. B.;
Trumbley, J. L.; Hallbauer, W. L. In Kirk-Othmer Encyclopedia of
Chemical Technology; Wiley: New York, 1981; Vol. 15, p 398.
(8) (a) Acetic Acid and Its DeriVatiVes; Agreda, V. H., Zoeller, J. R.,
Eds.; Marcel Dekker: New York, 1993. (b) Wagner, F. S. In Kirk-Othmer
Encyclopedia of Chemical Technology; Wiley: New York, 1978; Vol. 1, p
124.
(9) Review: Forster, D. AdV. Organomet. Chem. 1979, 17, 255.
S0002-7863(95)03670-5 CCC: $12.00 © 1996 American Chemical Society

Catalytic C-C and C-H Bond CleaVage in Lower Alkanes
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4575
back to the acid and methanol, it should be possible to design
a system where the acid is recycled and methanol is the end
product. With respect to product stability, control experiments
using methanol (initial concentration: 0.45 M) and acetic acid
(initial concentration 0.30 M) as substrates indicated that they
were quite stable under the reaction conditions (14% of methanol
reacted in 44 h; 5% of acetic acid reacted in 72 h). Furthermore,
when 12CH4 and 13CO were employed, an analysis of the gas
phase revealed only a trace of 12CO2 (13CO2:12CO2 ) 20). The
relative rate of methane versus methanol oxidation in the
presence of the former is given below.
A close examination of Figure 1 reveals that the methyl ester:
acetic acid ratio decreased with time. One explanation for this
trend would be that the methanol derivative was the initial
product and that this was being carbonylated to acetic acid (the
catalyst system consisting of a mixture of RhCl3 and I- ions
resembles the “Monsanto system” for the carbonylation of
methanol to acetic acid9). However, as in the aqueous system,1
we were able to rule out this possibility through the following
experiment. A 6:1 mixture (v/v) of perfluorobutyric acid and
water was made 0.50 M in 13CH3OH (90% of which was
converted to the methyl ester) and the methane functionalization
reaction was run under conditions identical to those given in
the caption of Figure 1. The products observed after 36 h were
C3F7CO212CH3 + 12CH3OH (0.13 M), 12CH3CO2H (0.22 M),
H12CO2H (0.02 M) and C3F7CO213CH3 + 13CH3OH (0.44 M),
H13CO2H (0.02 M). Clearly, methanol or its ester was not an
intermediate in the formation of acetic acid from methane since
no 13CH3CO2H was formed. Note that at 1000 psi of pressure,
the concentration of methane in pure water is 0.055 M. While
the solubility of methane in a 6:1 mixture (v/v) of perfluorobu-
tyric acid and water is expected to be higher, it is not likely to
significantly exceed 0.5 M, the concentration of added 13CH3-
OH. Finally, the complementary labeling experiment was
performed using 0.21 M 12CH3OH and 180 psi of 13CH4, 100
psi of 12CO, 50 psi of O2, and 170 psi of N2. After 18 h, the
organic products observed in solution were C3F7CO212CH3 +
12CH3OH (0.21 M), H12CO2H (0.05 M) and C3F7CO213CH3 +
13CH3OH (0.06 M), 13CH3CO2H (0.01 M), H13CO2H (0.04 M),
thus confirming (a) that the products, methanol and acetic acid,
were derived from methane and (b) that methane was signifi-
cantly more reactive than methanol or its ester.
Figure 1. Plots of products formed from methane versus time. Reaction
conditions: RhCl3‚3H2O (5 mg, 0.005 M), KI (10 mg, 0.017 M), NaCl
(15 mg, 0.073 M) in a 3.5 mL 6:1 (v/v) mixture of C3F7CO2H and
D2O, CH4 (1000 psi), CO (300 psi), O2 (100 psi), 80 °C. The reaction
was run in a 125-mL stainless steel bomb equipped with a glass liner.
in petroleum cracking and a “one-pot” system for both cracking
and subsequent oxidation of the light residues is of great interest
in the context of the need for oxygenates in clean-burning
gasoline.10
Results and Discussion
1. Catalyst. The catalyst system consists of RhCl3 along
with several equivalents of Cl- and I- ions dissolved in an acidic
solvent, and the alkane functionalizations were carried out in
the presence of O2 and CO at 80-85 °C. The use of this system
for the direct conversion of methane to acetic acid in aqueous
medium was previously reported in a brief communication.1
2. Functionalization of Methane. The principal problem
when water was used as the solvent was the low turnover rates
(approximately 0.1 turnover/h based on Rh).1 We now find
that it is possible to both significantly increase the reaction rate
and change the product specificity from acetic acid to methanol
by the choice of appropriate solvents. Our results for the
functionalization of methane in a 6:1 mixture (v/v) of perfluo-
robutyric acid and water as the solvent are summarized in Figure
1. Formic acid was the only significant byproduct observed.
Starting with pure water as solvent, the reaction rate was found
to increase steadily with increasing concentration of added
perfluorobutyric acid. However, water was essential for the
reaction since product formation was not observed in pure
perfluorobutyric acid. For methane, when a 6:1 mixture (v/v)
of perfluorobutyric acid and water was used as the solvent, the
total concentration of acetic acid and methanol (and its ester)
formed under reaction conditions similar to that used in pure
water exceeded 1.1 M after the reaction was run for 80 h at 80
°C (approximate turnover rate: 2.9/h based on Rh). Along with
an increase in rate, the product selectivity for methane func-
tionalization was found to change from virtually all acetic acid
to one in which the methanol derivative dominated. Since
methyl perfluorobutyrate is both volatile and easily hydrolyzed
A second interesting result came out of the first of the two
experiments described above that was run under standard
conditions. There was a switch in product selectivity upon the
addition of methanol to the solvent mixture: acetic acid rather
than the methanol derivative was now the principal product and
this was fully consistent with the gradual decrease in methyl
ester:acetic acid ratio with time that is seen in Figure 1.
Surmising that acetic acid became the favored product when
perfluorobutyric acid present in the solvent mixture was
converted to its ester, the methane functionalization was rerun
in a solvent mixture in which part of the perfluorobutyric acid
was replaced by 1,1,1,3,3,3-hexafluoro-2-propanol which would
be expected to form an ester with the remaining acid. Indeed,
acetic acid again became the predominant product. Thus, when
a 4:2:1 (v/v/v) mixture of perfluorobutyric acid, 1,1,1,3,3,3-
hexafluoro-2-propanol, and water was used as the solvent the
products obtained after 34 h under the same conditions were
C3F7CO2CH3 + CH3OH (0.09 M), CH3CO2H (0.22 M), and
HCO2H (0.02 M).
While the detailed mechanism remains to be elucidated, the
ratio of alcohol derivative to the corresponding higher acid may
be assumed to be a function of the relative rates of nucleophilic
attack versus carbon monoxide insertion into a common Rh-
(10) Unzelman, G. H. Oil Gas J. 1991, 89, 44, 62.

4576 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Lin et al.
Figure 3. 1H-NMR spectrum obtained after the following reaction
conditions: RhCl3‚3H2O (5 mg, 0.005 M), KI (10 mg, 0.017 M), NaCl
(15 mg, 0.073 M), 13CH3CH2OH (0.40 M) in a 3.5 mL 6:1 (v/v) mixture
of C3F7CO2H and D2O, C2H6 (550 psi), CO (150 psi), O2 (75 psi), 80
°C, 20 h. A capillary containing a mixture of Me2SO and D2O was
used as the internal standard.
trace of propionic acid was formed from ethane in a 6:1 mixture
(v/v) of perfluorobutyric acid and water. The generally lower
amount of hydroxycarbonylation observed for ethane may be
due to a higher rate of nucleophilic attack on the Rh-CH2CH3
species when compared to the corresponding Rh-CH3 species.
This observation finds parallel in the rate of hydrolysis of the
analogous Pt(IV) compounds [Cl5Pt-C2H5]2- and [Cl5Pt-
CH3]2-, which can be synthesized through the reactions of
Figure 2. Plots of products formed from ethane versus time. Reaction
conditions: RhCl3‚3H2O (5 mg, 0.005 M), KI (10 mg, 0.017 M), NaCl
(15 mg, 0.073 M) in a 3.5 mL 6:1 (v/v) mixture of C3F7CO2H and
D2O, C2H6 (550 psi), CO (150 psi), O2 (50 psi), 80 °C. The reaction
was run in a 125 mL stainless steel bomb equipped with a glass liner.
alkyl bond (i.e., kNu/kCO, see eq 1). While, to a first-order
Nu–
kNu
2-
PtCl4 with the corresponding alkyl iodide.11 In pure water
R
Nu
at ambient temperature, [Cl5Pt-C2H5]2- was hydrolyzed within
hours to ethanol whereas [Cl5Pt-CH3]2- was stable for days.
On a per bond basis, the ratio of products formed through
C-H versus C-C cleavage of ethane was 0.6. Control
experiments indicated that neither acetic acid nor ethanol was
the source for the C1 product, methanol. This is shown by the
1H NMR spectrum obtained when both 13CH312CH2OH (0.4 M)
and 12C2H6 (550 psi) were present in the reaction mixture (Figure
3). Clearly, the labeled group of ethanol was not the source
for either the methanol derivative or the acetic acid. Note that
at 550 psi of pressure, the concentration of ethane in pure water
is 0.03 M. While the solubility of ethane in a 6:1 mixture (v/
v) of perfluorobutyric acid and water is expected to be higher,
it is likely to be lower than 0.4 M, the concentration of added
13CH312CH2OH. In order to confirm that the observed methanol
derivative was not formed in some way by CO hydrogenation,
the ethane oxidation was repeated using 12C2H6 and 13CO. No
13C-containing organic products were observed except in acetic
acid which was an approximate 4:1 mixture of 12CH312CO2H
and 12CH313CO2H. Thus, the experiment also confirmed that
the majority of the acetic acid was derived from straightforward
ethane oxidation rather than through C-C cleavage and
carbonylation of a resultant Rh-CH3 species. Interestingly, the
results from the two labeling studies taken together showed that
ethanol was not even an intermediate in the conversion of ethane
to that fraction of acetic acid not formed via C-C cleavage.
Finally, in order to conclusively demonstrate that ethane was
indeed the source for methanol (and its ester), 13C2H6 was
employed as the substrate (gas pressures: 13C2H6, 100 psi; CO,
200 psi; O2, 100 psi; N2, 800 psi) resulting in the exclusive
formation of C3F7CO213CH3 + 13CH3OH.
LxRh
R
Nu–
CO
kCO
RCO Nu
LxRh COR
(Nu = OH, C3F7CO2)
(1)
approximation, kCO is likely to be independent of the solvent,
kNu will depend on the nature of the nucleophile derived from
the solvent. Presumably, the perfluorobutyrate ion is a better
nucleophile than water since more of the alcohol derivative was
formed in a perfluorobutyric acid-water mixture than in pure
water. This also explains why acetic acid was once again the
major product when the perfluorobutyrate ion was tied up as
-
the ester. In principle, nucleophiles stronger than the C3F7CO2
ion should be able to compete with the latter when added to
the reaction mixture. However, the addition of 7 equiv of
acetate ion strongly retarded the reaction, possibly due to
coordination to the metal center. In the same vein, the
substitution of acetic acid for perfluorobutyric acid in the usual
solvent mixture resulted in a drastic reduction in reaction rate
(<10% of the original rate). Finally, consistent with the
mechanistic scenario shown in eq 1 was also the observation
that the ratio of acetic acid to methanol derivative formed from
methane increased with increasing pressure of CO although the
overall reaction was sharply inhibited at high CO pressures.
3. Functionalization of Ethane. As might be anticipated,
ethane was more reactive than methane and a higher reaction
rate was observed for the conversion of ethane (Figure 2).
Under similar reaction conditions, the total concentration of
acetic acid, ethanol, and methanol (and their esters) exceeded
2.4 M after 66 h (approximate turnover rate: 7.5/h based on
Rh).
(11) (a) Sen, A.; Lin, M.; Kao, L.-C.; Hutson, A. C. J. Am. Chem. Soc.
1992, 114, 6385. (b) Luinstra, G. A.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 1993, 115, 3004. (c) Kusch, L. A.; Lavrushko, V. V.; Misharin,
Yu. S.; Moravsky, A. P.; Shilov, A. E. NouV. J. Chim. 1983, 7, 729.
Unlike methane, little hydroxycarbonylation was observed
for ethane and higher alkanes. For example, no more than a

Catalytic C-C and C-H Bond CleaVage in Lower Alkanes
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4577
cleavage. Two reasons are usually cited for the general lack
of C-C cleavage compared to corresponding C-H cleavage
in reactions mediated by metal complexes in solution.12 First,
C-C bonds are sterically less accessible to transition metal
centers surrounded by bulky ligands. Second, metal-carbon
bonds tend to be weaker than metal-hydrogen bonds, again
due to steric repulsions between the ligands surrounding the
metal and the alkyl group bound to it. C-C cleavage is,
however, commonly observed in the interaction of bare metal
cations with alkanes.13 In this case, there is no steric hindrance
and the metal-carbon and metal-hydrogen bond strengths are
comparable (approximately 60 kcal/mol).13a,b,14
Before a case is made for the cleavage of C-C and C-H
bonds mediated by the metal center in the present instance, it
is necessary to rule out alternative pathways. The possibility
that a variation of oxychlorination15 may be occurring in the
gas phase (eq. 2) can be ruled out by the following experiment.
The reaction involving methane was run in the absence of
dissolved Rh salt, but in the presence of O2, the promoters (KI,
NaCl), and a large excess of dissolved Cl2 (the solvent was
saturated with Cl2 prior to reaction). Such a reaction condition
would be particularly favorable for oxychlorination since all
the ingredients (Cl2, HCl, and O2) were present. No product
formation was observed. Also note that in view of the inertness
of CH3I (see below) it is most unlikely that CH3Cl, if formed,
would convert to the observed products. Finally, simple alkyl
radicals do not undergo C-C cleavage reactions.
Figure 4. 1H-NMR spectrum obtained after the following reaction
conditions: RhCl3‚3H2O (5 mg, 0.005 M), KI (10 mg, 0.017 M), NaCl
(15 mg, 0.073 M), C4H10 (5.6 mmol) in a 3.5 mL 6:1 (v/v) mixture of
C3F7CO2H and D2O, CO (300 psi), O2 (100 psi), N2 (800 psi), 90 °C,
66 h. A capillary containing a mixture of Me2SO and D2O was used as
the internal standard.
The C-C cleavage of ethane was also observed when pure
water was used as the solvent. For example, an approximate
3:1 mixture of 13CH313CO2H and 13CH312CO2H was also formed
in water starting with 13C2H6, 12CO, and O2 (gas pressures:
13C2H6, 100 psi; CO, 200 psi; O2, 100 psi; N2, 600 psi). The
complementary experiment with 12C2H6, 13CO, and O2 gave a
3:1 mixture of 12CH312CO2H and 12CH313CO2H.
4. Functionalization of Higher Alkanes. When propane
was used as the substrate, a mixture of n- and isopropyl esters
and acetone were formed with a 5:1 preference for attack on
the secondary C-H bond (on a per hydrogen basis). Thus, the
reaction rate increased with decreasing C-H bond dissociation
energy. In addition, significant quantities of acetic acid, ethanol,
and methanol derivatives were formed through C-C cleavage.
A particularly striking example of catalytic C-C function-
alization was obtained with n-butane. As shown in Figure 4,
Virtually all the products were derived through C-C cleavage!
Especially noteworthy was the formation of ethanol derivative.
Starting with 0.61 mmol of butane, the organic products
observed after 20 h were C3F7CO2CH2CH3 + CH3CH2OH (0.18
mmol), CH3CO2H (0.34 mmol), CH3CHO (0.04 mmol), C3F7-
CO2CH3 + CH3OH (0.17 mmol), HCO2H (0.63 mmol), along
with unreacted n-butane (0.05 mmol). The overall yield was
85.0% based on carbon content of n-butane. Note that the
butane concentration was much higher for the reaction shown
in Figure 4 and resulted in a slightly different product distribu-
tion. Finally, control experiments indicated that, like ethane,
neither 1- or 2-butanol was the source of ethanol in the above
reaction.
As with butane, products derived from C-C cleavage were
Virtually all that were observed with 2-methylbutane (isopen-
tane). Starting with 0.42 mmol of substrate, the organic products
observed after 20 h were C3F7CO2CH(CH3)2 + (CH3)2CHOH
(0.03 mmol), (CH3)2CO (0.07 mmol), C3F7CO2CH2CH3 + CH3-
CH2OH (0.07 mmol), CH3CO2H (0.20 mmol), C3F7CO2CH3 +
CH3OH (0.14 mmol), HCO2H (0.50 mmol), along with unre-
acted alkane (0.04 mmol). The overall yield was 87.2% based
on carbon content of 2-methylbutane.
2CH4 + Cl2 + 1/2O2 f 2CH3Cl + H2O
(2)
Another possibility is that a radical cation may form from
the alkane by outer-sphere electron transfer and this species, in
turn, would fragment some of the time by C-C cleavage (cf
eq 3). For example, the 1e- oxidation of ethane leads to a
–H+
–e–
•
•+
C2-products
C2H5
[C2H6
]
C2H6
•
CH3+ + CH3
C1-products
(3)
substantial weakening of the C-C bond (from 90 to 45-37
•
+ 16
kcal/mol) resulting in fragmentation to CH3 + CH3
.
The
direct experimental precedent for such a step is Olah’s observa-
tion of CH3NO2 as the principal product in the reaction of ethane
with NO2+PF6- in aprotic solvents (although the products may
also arise from the direct attack of the electrophile on the C-C
bond).17 Additionally, we have shown that in 98% sulfuric acid,
a variety of 1e- and 2e- oxidants will oxidize ethane to a
mixture of C1 and C2 derivatives through a similar electron
transfer pathway at 180°C.18
Several observations appear to rule out the involvement of
free radicals formed by simple bond homolysis that would
(12) (a) Halpern, J. Inorg. Chim. Acta 1985, 100, 41. (b) Halpern, J.
Acc. Chem. Res. 1982, 15, 238.
(13) Reviews: (a) Armentrout, P. B. In SelectiVe Hydrocarbon Oxidation
and Functionalization; Davies, J. A., Watson, P. L., Greenberg, A., Liebman,
J. F., Eds.; VCH: New York, 1990; p 467. (b) Armentrout, P. B.;
Beauchamp, J. L. Acc. Chem. Res. 1989, 22, 315. (c) Schwarz, H. Acc.
Chem. Res. 1989, 22, 282.
(14) Simoes, J. A. M.; Beauchamp, J. L. Chem. ReV. 1990, 90, 629.
(15) Review: Treger, Yu. A.; Rozanov, V. N. Russ. Chem. ReV. 1989,
58, 84.
(16) Lathan, W. A.; Curtiss, L. A.; Pople, J. A. Mol. Phys. 1971, 22,
1081.
Finally, the hindered C-C bond of 2,3-dimethylbutane was
also cleaved by the present system. Starting with 1.16 mmol
of substrate, the organic products observed after 40 h were C3F7-
CO2CH(CH3)2 + (CH3)2CHOH (0.12 mmol), (CH3)2CO (0.41
mmol), CH3CO2H (0.31 mmol), C3F7CO2CH3 + CH3OH (0.03
mmol), and HCO2H (0.48 mmol). The overall yield was 40.0%
based on carbon content of 2,3-dimethylbutane.
(17) Olah, G. A.; Lin, H. C. J. Am. Chem. Soc. 1971, 93, 1259.
(18) Sen, A.; Benvenuto, M. A.; Lin, M.; Hutson, A. C.; Basickes, N. J.
Am. Chem. Soc. 1994, 116, 998.
5. Mechanistic Considerations. The most significant aspect
of the work described above is the observation of alkane C-C

4578 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Lin et al.
2-
conditions, we examined the reaction of S2O8 with methane
and ethane. We had earlier demonstrated that in pure water,
2-
S2O8
functionalized these alkanes through a hydrogen-
abstraction step resulting in the formation of the corresponding
alkyl radical as the intermediate (eq 4).20 Additionally, we have
shown that at 180 °C in 98% sulfuric acid, the S2O82- ion will
oxidize ethane to a mixture of C1 and C2 derivatives through
an electron transfer pathway.18 However, under the reaction
conditions shown in Figure 1 except with K2S2O8 replacing the
Rh-based catalyst system, no product was formed from methane
after 20 h. When ethane was used as the substrate (gas
pressures: C2H6, 550 psi; CO, 150 psi; O2, 50 psi), the following
products were observed after 48 h: CH3CO2H (0.01 mmol),
CH3CHO (0.15 mmol), HOCH2CH2OH and its mono- and
diesters (0.13 mmol). Ethene, presumably formed by the
deprotonation of the ethyl cation (see eq 4), may be an
intermediate in the formation of the diol and its derivatives.
This was demonstrated by replacing ethane by ethene (550 psi)
in the above reaction, resulting in the formation of 0.71 mmol
of HOCH2CH2OH and its mono- and diesters in 20 h.
Figure 5. 1H-NMR spectrum obtained after the following reaction
conditions: RhCl3‚3H2O (5 mg, 0.005 M), KI (10 mg, 0.017 M), NaCl
(15 mg, 0.073 M), 13CH3I (0.22 M) in a 3.5 mL 6:1 (v/v) mixture of
C3F7CO2H and D2O, CH4 (1000 psi), CO (300 psi), O2 (100 psi), 80
°C, 20 h. A capillary containing a mixture of Me2SO and D2O was
used as the internal standard.
•-
S2O82- f SO4
•-
•-
SO4
SO4
+
•
C2H6
8 C2H6
8 C2H5+ y-H z
follow electron transfer from the alkane. First, when the
methane functionalization reaction was run under conditions
identical to that shown in Figure 1 except that the reaction
mixture was 0.22 M in 13CH3I, the products observed were C3F7-
CO212CH3 + 12CH3OH (13CH3I remained unchanged) (Figure
CH2dCH2 f C2-products (4)
A comparison of the reactivity pattern of the Rh-based catalyst
system with that of K2S2O8 further supports the proposition that
the former does not simply function as an 1e- oxidant or a
H-atom abstractor vis-Åa-vis alkanes. Clearly further studies are
needed to elucidate the nature of the C-H and C-C cleavage
steps. One should note that none of the experiments establish
that the metal center is directly responsible for C-C cleavage.
It is possible, for example, that the metal ion forms an alkyl
species through C-H activation and that the C-C cleavage
products are formed through a subsequent â-alkyl abstraction
step. Such a sequence of steps has been reported in the reactions
of bare metal cations with alkanes.21 Still another possibility
is that a metal alkoxide is formed initially from the alkane
through a nonradical pathway and that a subsequent bond
homolysis generates the alkoxy radical which then undergoes
known fragmentations through C-C cleavage;22 the fact that
alcohols are unreactive in our system (see above) does not
necessarily rule out this scenario.
The role of carbon monoxide in the rhodium-catalyzed alkane
functionalizations is curious. Without CO there was no reaction
even when the product was simply an alcohol derivative and,
therefore, CO was not required by stoichiometry. One role of
CO may simply be to serve as a stabilizing ligand, especially
for Rh(I). Indeed, Rh(I) carbonyl complexes are formed under
the reaction conditions.1 However, it is likely that there is a
second, more important, role of CO in the system. CO is also
an essential reagent in a metallic palladium-catalyzed system
for alkane functionalization that we reported earlier.23 Our
studies on that system showed that CO participates in the water-
gas shift reaction to generate H2 (together with CO2) which, in
turn, reacts with O2 to form H2O2, the real oxidant involved in
alkane functionalization. It is tempting to postulate a similar
5), i.e., the significantly stronger CH3-H bond (105 kcal/mol19a
)
was being broken in preference to the weak CH3-I bond (57
kcal/mol19a); the ionization potential of CH3I is also significantly
lower than that of methane (9.5 vs 11.5 eV19b). The unusual
reactivity pattern was further confirmed by the complementary
labeling experiment involving 12CH3I and 13CH4 which resulted
in the formation of C3F7CO213CH3 + 13CH3OH only. Second,
in the reaction run in the presence of 13CH3OH (vide supra),
0.21 mmol of 13CH3OH was oxidized at the same time as 1.23
mmol of 12CH4 was converted to C3F7CO212CH3 + 12CH3OH
+
12CH3CO2H. Thus, methane was at least 5 times more
reactiVe than methanol and its derivative (note that an even
higher selectivity for methane oxidation was observed in the
reaction involving 13CH4 and 12CH3OH) although the homolytic
C-H bond energy in the latter is 10 kcal/mol lower. However,
this does not take into account the increase in the C-H bond
energy when methanol is converted to the ester (the following
C-H bond-energy data illustrate the point: H-CH2OH, 94 kcal/
mol; H-CH2OCOC6H5, 100.2 kcal/mol19c). Finally, using a
1:1 mixture of C2H6 and C2D6 (250 psi each), a primary isotope
effect (kH/kD) of 2.65(5) was observed for the formation of ethyl
heptafluorobutyrate. This is inconsistent with the mechanism
shown in eq 3 since electron transfer, which is expected to be
the rate-limiting step, should exhibit only a very modest
secondary isotope effect. Thus direct C-H bond cleavage
within the coordination sphere of the metal is indicated. It
should be noted that, due to the presence of excess I- ions in
the system, it is possible to have neither a radical more reactive
than I• nor an oxidant stronger than I2. Since neither of these
species is capable of reacting with methane, this provides
additional support that the alkane functionalization is occurring
within the coordination sphere of the metal.
(20) Lin, M.; Sen, A. J. Chem. Soc., Chem. Commun. 1992, 892.
(21) Specific recent example: van Koppen, P. A. M.; Bowers, M. T.;
Fisher, E. R.; Armentrout, P. B. J. Am. Chem. Soc. 1994, 116, 3780.
(22) Review on fragmentation of alkoxy radicals: Kochi, J. K. In Free
Radicals; Wiley: New York, 1973; Vol. II, p 665.
In order to further delineate the reaction pathway followed
by an 1e- oxidant/H-atom abstractor under the present reaction
(19) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC
Press: Boca Raton, 1991; (a) p 9-121, (b) p 10-213, (c) p 9-114.
(23) Lin, M.; Sen, A. J. Am. Chem. Soc. 1992, 114, 7307.

Catalytic C-C and C-H Bond CleaVage in Lower Alkanes
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4579
role for CO in the present rhodium-based system. Indeed, Rh
salts in the presence of I- ion as promoter have been shown to
be active catalysts for the water-gas shift reaction.24 Moreover,
as with the metallic palladium based system, alkane function-
alization does not occur if water is absent from the solvent
mixture. Additionally, we observed the conversion CO to CO2
in parallel with the alkane functionalization (see section 2
above). However, H2O2, if formed subsequent to the water-
gas shift reaction, would be expected to react rapidly with the
added I- ions to generate I2, a species that is thermodynamically
not competent to react with methane or ethane. Nevertheless,
we examined the effect of replacing O2 in our system with a
number of O-atom donor oxidants, such as H2O2, m-ClC6H4-
COOOH, C6H5IO, and IO4-, both in the presence and absence
of CO. In no case was product formation observed from
methane. The role of CO must, therefore, await a more detailed
analysis of the system.
6. Comparison with the Monsanto System. Finally, it is
instructive to compare our reaction rates with that in the
“Monsanto system” for the carbonylation of methanolsthe
currently preferred commercial process for the manufacture of
acetic acid.8,9 In a 6:1 mixture (v/v) of perfluorobutyric acid
and water the rate of formation of acetic acid + methanol from
methane was 3.2 × 10-4 M/min at 80 °C. In the Monsanto
process, the rate of formation of acetic acid from methanol is
384 × 10-4 M/min at 180 °C.25 Thus, the rate is 120 times
faster in the latter system. However, the substrate concentration
in the Monsanto system is typically 5 M. On the other hand,
at the 1000 psi pressure that we employ, the aqueous concentra-
tion of methane is approximately 0.05 M. If the solution
concentrations (more precisely, the activities) of methane in a
6:1 mixture (v/v) of perfluorobutyric acid and water are similar,
then under similar substrate concentrations our rate of methane
conversion is comparable to the rate of methanol conversion
in the Monsanto system (this does not take into account the
much higher reaction temperature employed for the latter).
employed to generate a high-energy oxygen species, such as a
metal-oxo complex, from the second oxygen atom (eq 5).26
The “Gif” system of Barton is also designed on this premise.27
In at least the metallic Pd-based system,23 the coreductant,
carbon monoxide, is employed to generate dihydrogen (eq 6),
the latter being formally equivalent to 2H+ + 2e- that is
employed in the biological systems (cf eqs 5 and 7).
O2 + 2H+ + 2e- f H2O + [O]
CO + H2O f CO2 + H2
O2 + H2 f H2O + [O]
(5)
(6)
(7)
How general is this requirement for a coreductant (e.g., CO
or H2) in achieving “difficult” catalytic hydrocarbon oxidations
by dioxygen? Our work has provided two examples of catalytic
systems, one homogeneous (this report) and the other hetero-
geneous,23 that operate in this manner (i.e., as monooxygenase
analogs). There have been other recent publications on catalytic
systems for the oxidation of hydrocarbons, including olefins
and aromatics, that also call for either CO or H2 as the
coreductant.28 While, from a practical standpoint, it is more
desirable for both oxygen atoms of O2 to be used for substrate
oxidation, there appears to be no known catalytic system that
operates as an artificial “dioxygenase” under mild conditions
toward “difficult” substrates, such as those possessing unacti-
vated primary C-H bonds.
Experimental Section
General. The following chemicals were used as received: RhCl3‚
3H2O (Johnson Matthey); NaCl, KI, 13CH3I, isopentane, 2,3-dimeth-
ylbutane (Aldrich); K2S2O8 (Fisher Scientific Co); methane, ethane,
propane, n-butane, nitrogen, oxygen, and carbon monoxide (Matheson);
D2O, 13CO, 13C2H6, and C2D6 (Isotec, Inc.); and 13CH4 (Cambridge
Isotopes). Reactions under pressure were carried out in Parr general
purpose bombs using glass liners. 1H-NMR spectra were recorded on
a Brucker AM 300 FT-NMR spectrometer using solvent resonance at
the appropriate frequency or an external standard consisting of a
capillary tube containing 1 µL of DMSO in 60 µL of D2O for lock,
reference, and as an integration standard. GC-MS was performed using
a combination of Carls ERBA gas chromatograph equipped with a J
& W Scientific DBS column (60 m × 0.255 mm) and a Kratos MS-25
Conclusion
We have discovered the first catalytic system that simulta-
neously cleaves C-H and C-C bonds of simple alkanes, as
well as activates dioxygen, under mild conditions. For most
of the alkanes studied, products derived from C-C cleavage
dominated over those derived from C-H cleavage on a per bond
basis. While the mechanism of the C-H and C-C cleavage
steps remains to be elucidated, preliminary indications are that
outer-sphere electron transfer or bond homolysis resulting in
the formation of alkyl radicals does not occur.
A curious aspect of the previous metallic Pd23 and now the
homogeneous Rh-based systems is that, apart from their ability
to activate both dioxygen and the alkane, both require a
coreductant (carbon monoxide). Thus, there is a striking
resemblance with monooxygenases.26 In nature, while the
dioxygenases utilize the dioxygen molecule more efficiently, it
is the monooxygenases that carry out “difficult” oxidations, such
as alkane oxidations. In the latter, one of the two oxygen atoms
of dioxygen is reduced to water in a highly thermodynamically
favorable reaction and the free energy gained thereby is
1
mass spectrometer. In all cases, the products were identified by H-
NMR spectroscopy and confirmed by GC/MS and/or by comparison
with authentic samples. The points in Figures 1 and 2 were generated
by running a reaction for a certain time period, depressurizing,
examining the reaction mixture by NMR spectroscopy, adding the
reaction mixture back to the reactor, and resuming the reaction for the
next time period.
Caution: Appropriate precautions should be taken while working
with gases under high pressures. Particular attention should be paid
to flammability limits of gas mixtures.
1. Functionalization of Alkanes with a Rhodium-Based Catalyst
.
System. (a) Functionalization of Methane. RhCl3 3H2O (5.0 mg,
0.02 mmol), KI (10.0 mg, 0.06 mmol), and NaCl (15 mg, 0.26 mmol)
were dissolved in 0.5 mL of D2O in a glass container, 3 mL of C3F7-
CO2H was then added slowly, and a brownish mixture was formed.
The glass container was placed in a high-pressure bomb which was
then sealed. The bomb was purged and pressurized to 300 psi with
CO, to 1300 psi with CH4, then to 1400 psi with O2. The contents
were stirred at 80-85 °C for a certain time period. After venting, the
(24) Baker, E. C.; Hendriksen, D. E.; Eisenberg, R. J. Am. Chem. Soc.
1980, 102, 1020.
(25) Hjortkjaer, J.; Jensen, V. W. Ind. Eng. Chem., Prod. Res. DeV. 1976,
15, 46.
(26) Reviews: (a) Valentine, J. S. In Bioinorganic Chemistry; Bertini,
I.; Gray, H. B.; Lippard, S. J., Valentine, J. S., Eds.; University Science
Books: Mill Valley, CA, 1994; p 253. (b) Mansuy, D.; Battioni, P. In
Bioinorganic Catalysis; Reedijk, J., Ed.; Marcel Dekker: New York, 1993;
p 395.
1
products were examined by H-NMR spectroscopy.
(27) Barton, D. H. R.; Doller, D. Acc. Chem. Res. 1992, 25, 504.
(28) Representative examples: (a) Otake, M. Chemtech 1995, No. 9, 36.
(b) Miyake, T.; Hamada, M.; Sasaki, Y.; Oguri, M. Appl. Catal. A: General
1995, 131, 33. (c) Teranishi, T.; Toshima, N. J. Chem. Soc., Dalton Trans.
1995, 979. (d) Wang, Y.; Otsuka, K. J. Catal. 1995, 155, 256. (e) Benvenuto,
M. A.; Sen, A. J. Chem. Soc., Chem. Commun. 1993, 970.

4580 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Lin et al.
(b) Functionalization of Methane in the Presence of CH3I. The
reaction was conducted as described above, except that 48 µL of 13CH3-
I (0.75 mmol) was added to the reaction mixture. The products formed
were 12CH3OH and its ester (0.700 mmol), 12CH312CO2H (0.140 mmol),
H12CO2H(0.21 mmol), along with a trace of 13CH3CO2H.
and 100 psi of O2 at 85 °C for 20 h. The products formed were CH3-
CH2OH and ester (0.18 mmol), CH3CO2H (0.34 mmol), CH3CHO (0.04
mmol), CH3OH and ester (0.17 mmol), HCO2H (0.63 mmol), along
with unreacted n-butane (0.05 mmol). The yield was 85.0% based on
carbon content of butane.
(g) Functionalization of n-Butane under 13CO. The reaction was
conducted as described in procedure (f), except that 300 psi of 13CO
was used instead of 12CO. Only 12C-containing products were formed.
In the reverse labeling experiment, 9 µL of 12CH3I (0.145 mmol)
containing 12CH3OH (0.0026 mmol) was added to 1 mL of the reaction
mixture. The bomb was then purged and pressurized to 250 psi with
13CH4, to 350 psi with CO, to 450 psi with N2, then to 500 psi with O2.
At the end of 18 h, 0.004 mmol of 13CH3OH and its ester were the
only significant products observed; there was virtually no change in
the amount of 12CH3OH and its ester (0.0023 mmol) that was present
initially.
(c) Functionalization of Methane in the Presence of CH3OH. The
reaction was conducted as described in procedure (b), except that 73
µL of 13CH3OH (1.75 mmol) was added to the reaction mixture instead
of 13CH3I. The products observed were 12CH3OH and ester (0.46
mmol), 12CH3CO2H (0.77 mmol), H12CO2H (0.07 mmol), 13CH3OH and
ester (1.54 mmol), and H13CO2H (0.07 mmol).
(h) Functionalization of Isopentane. The reaction was conducted
as described in procedure (f), except that 0.420 mmol (0.120 M) of
isopentane was used instead of n-butane. The products formed were
2-propanol and ester (0.03 mmol), CH3COCH3 (0.07 mmol), CH3CH2-
OH and ester (0.07 mmol), CH3CO2H (0.20 mmol), CH3OH and ester
(0.14 mmol), HCO2H (0.50 mmol), along with unreacted isopentane
(0.039 mmol). The yield was 87.2% based on carbon content of
isopentane.
(i) Functionalization of 2,3-Dimethylbutane. The reaction was
conducted as described in procedure (f), except that 1.16 mmol (0.330
M) of 2,3-dimethylbutane was used instead of n-butane. The organic
products observed after 40 h were 2-propanol and ester (0.12 mmol),
(CH3)2CO (0.41 mmol), CH3CO2H (0.31 mmol), CH3OH and ester (0.03
mmol), and HCO2H (0.48 mmol). The overall yield was 40.0% based
on carbon content of 2,3-dimethylbutane.
In the reverse labeling experiment, 8.4 µL of 12CH3OH (0.21 mmol)
was added to 1 mL of the reaction mixture. The bomb was then purged
and pressurized to 180 psi with 13CH4, to 280 psi with CO, to 450 psi
with N2, then to 500 psi with O2. After 18 h, the organic products
observed in solution were 12CH3OH and ester (0.21 M), H12CO2H (0.05
M) and 13CH3OH and ester (0.06 M), 13CH3CO2H (0.01 M), and
H13CO2H (0.04 M).
(d) Functionalization of Ethane. The reaction was conducted as
described in procedure (a) except that 150 psi of CO, 550 psi of C2H6,
and 50 psi of O2 were added into the bomb.
2. Functionalization of Alkanes with K2S2O8. (a) Functional-
ization of methane. The reaction was conducted as described in
procedure (a) in section 1, except that 272 mg of K2S2O8 (1.0 mmol)
.
was used instead of RhCl3 3H2O and its promoters. No product
formation was observed after 20 h.
(e) Functionalization of Ethane in the Presence of 13CH3CH2OH.
RhCl3‚3H2O (5.0 mg, 0.02 mmol), KI (10.0 mg, 0.06 mmol), NaCl
(15 mg, 0.26 mmol), and 13CH3CH2OH (80 µL, 1.4 mmol) were
dissolved in 0.5 mL of D2O in a glass container, and 3 mL of C3F7-
CO2H was then added slowly to the mixture. The glass container was
placed in a high-pressure bomb which was then sealed. The bomb
was pressurized to 500 psi with CH3CH3, to 700 psi with CO, to 1200
psi with N2, to 1300 psi with O2. The contents were stirred at 80-85
°C for 24 h. The products formed were 12CH3CO2H (0.74 mmol),
12CH3CH2OH (0.64 mmol), 12CH3OH and ester (0.15 mmol), and
HCO2H (2 mmol).
(b) Functionalization of Ethane. The reaction was conducted as
described in procedure (a), except that 150 psi of CO, 550 psi of C2H6,
and 50 psi of O2 were added to the bomb and the reaction was carried
out at 85 °C for 48 h. The products formed were CH3CO2H (0.01
mmol), CH3CHO (0.15 mmol), and HOCH2CH2OH and its mono- and
diesters (0.13 mmol).
(c) Functionalization of Ethene. The reaction was conducted as
described in procedure (b), except that 550 psi of ethene was used
instead of ethane and the reaction was carried at 85 °C for 20 h. The
products formed were HOCH2CH2OH and its mono- and diesters (0.71
mmol).
(f) Functionalization of n-Butane. The reaction was conducted
as described in procedure (a), except that n-butane was bubbled through
the solution in the glass container for 10 min following which the
solution concentration of n-butane in the solution was measured by
1H-NMR spectroscopy (0.173 M, 0.61 mmol). After the bomb was
sealed, the contents were stirred under 300 psi of CO, 800 psi of N2,
Acknowledgment. This research was funded by the National
Science Foundation and in part by the CANMET Consortium
on the Conversion of Natural Gas.
JA953670R