Arylglyoxylrhodium complexes, their thermolysis, and attempted generation by carbonylation of an aroylrhodium complex

Article abstract of DOI:10.1021/om101079e

Reactions of ArCOCOCl (Ar = p-ClC₆H₄, Ph) with Rh(acac)(CO)₂ proceeded readily to afford dimeric arylglyoxyl rhodium complexes [Rh(μ-Cl)(acac)(CO)(COCOAr)]₂. [Rh(μ-Cl)(acac)(CO) {COCO(p-ClC₆H₄)}]₂ was characterized by X-ray diffraction. Thermolysis of the products showed that [Rh(μ-Cl)(acac)(CO) {COCO(p-ClC₆H₄)}]₂ was more stable than [Rh(μ-Cl)(acac)(CO)(COCOPh)]₂. The reaction of p-CH ₃OC₆H₄COCOCl with Rh(acac)(CO)₂ did not form a similar dimeric complex as final product, but gave p-CH ₃OC₆H₄COCl as major product, showing thermal instability of corresponding arylglyoxyl and aroyl rhodium complexes. The reaction of C₆F₅COCOCl did not form a simlar dimeric complex either, but a mononuclear complex RhCl(acac)(COCOC₆F ₅)(CO)₂ was generated as a transient intermediate, which was readily transformed to a furanone arising from reductive elimination of the acac ligand and C₆F₅COCO moiety followed by cyclization. In the thermolysis of [Rh(μ-Cl)(acac)(CO)(COCOAr)]₂, only ArCOCl and Rh(acac)(CO)₂ were formed, and any ArCORh species could not be detected during the thermolysis process. However, the reaction of Rh(acac)(CO)₂ with PhCOCl formed RhCl(acac)(CO)₂(COPh), albeit only to a small extent, suggesting that the reaction is not thermodynamically favored. The reaction of RhCl(CO)(PMe₃)₂ with p-ClC₆H₄COCOCl also proceeded cleanly to furnish p-ClC₆H₄COCORhCl₂(CO)(PMe₃) ₂. Thermolysis of the complex formed p-ClC₆H ₄CORhCl₂(CO)(PMe₃)₂, indicating slow reductive elimination of ClC₆H₄COCl as compared with the ArCORh species, generated in the thermolysis of [Rh(μ-Cl)(acac)(CO)(COCOAr)] ₂. Treatment of ClC₆H₄CORhCl ₂(CO)(PMe₃)₂ with carbon monoxide generates p-ClC₆H₄COCORhCl₂(CO)(PMe₃) ₂, although the yield was low.

Full text of DOI:10.1021/om101079e

Organometallics 2011, 30, 239–244 239
DOI: 10.1021/om101079e
Arylglyoxylrhodium Complexes, Their Thermolysis, and Attempted
Generation by Carbonylation of an Aroylrhodium Complex
Taigo Kashiwabara and Masato Tanaka*
Chemical Resources Laboratory, Tokyo Institute of Technology, 4259-R1-13 Nagatsuta, Midori-ku,
Yokohama 226-8503, Japan
Received November 15, 2010
Reactions of ArCOCOCl (Ar=p-ClC6H4, Ph) with Rh(acac)(CO)2 proceeded readily to afford
dimeric arylglyoxyl rhodium complexes [Rh(μ-Cl)(acac)(CO)(COCOAr)]2. [Rh(μ-Cl)(acac)(CO)-
{COCO(p-ClC6H4)}]2 was characterized by X-ray diffraction. Thermolysis of the products showed
that [Rh(μ-Cl)(acac)(CO){COCO(p-ClC6H4)}]2 was more stable than [Rh(μ-Cl)(acac)(CO)-
(COCOPh)]2. The reaction of p-CH3OC6H4COCOCl with Rh(acac)(CO)2 did not form a similar
dimeric complex as final product, but gave p-CH3OC6H4COCl as major product, showing thermal
instability of corresponding arylglyoxyl and aroyl rhodium complexes. The reaction of C6F5CO-
COCl did not form a simlar dimeric complex either, but a mononuclear complex RhCl(acac)-
(COCOC6F5)(CO)2 was generated as a transient intermediate, which was readily transformed to a
furanone arising from reductive elimination of the acac ligand and C6F5COCO moiety followed by
cyclization. In the thermolysis of [Rh(μ-Cl)(acac)(CO)(COCOAr)]2, only ArCOCl and Rh(acac)-
(CO)2 were formed, and any ArCORh species could not be detected during the thermolysis process.
However, the reaction of Rh(acac)(CO)2 with PhCOCl formed RhCl(acac)(CO)2(COPh), albeit only
to a small extent, suggesting that the reaction is not thermodynamically favored. The reaction of
RhCl(CO)(PMe3)2 with p-ClC6H4COCOCl also proceeded cleanly to furnish p-ClC6H4COCORhCl2-
(CO)(PMe3)2. Thermolysis of the complex formed p-ClC6H4CORhCl2(CO)(PMe3)2, indicating slow
reductive elimination of ClC6H4COCl as compared with the ArCORh species, generated in the
thermolysis of [Rh(μ-Cl)(acac)(CO)(COCOAr)]2. Treatment of ClC6H4CORhCl2(CO)(PMe3)2 with
carbon monoxide generates p-ClC6H4COCORhCl2(CO)(PMe3)2, although the yield was low.
Introduction
acyl species, has also been reported.4 Here we wish to disclose
5,6
that Rh(I) complexes such as Rh(acac)(CO)2 and RhCl-
One of us and Yamamoto’s group reported palladium-
catalyzed carbonylation of aromatic halides forming R-keto
acid derivatives in 1982.1 Although a mechanism involving
ArCOCOPd species was one of the options, the possibility
was clearly excluded by the facile decarbonylation observed
by Sen and co-workers.2 Since then we have considered
tacitly that R-keto acyl complexes are thermodynamically
unfavorable, although Sen and co-workers reported that
R-keto acyl platinum complexes are more stable than palla-
dium congeners.3 Palladium-catalyzed carbonylation of imi-
doyl chlorides, possibly involving somewhat similar R-imino
(CO)(PMe3)2 react readily with ArCOCOCl to afford sig-
nificantly stable arylglyoxyl rhodium complexes (ArCOCORh)
in high yields. Thermolysis of the products and attempted
generation of arylglyoxyl rhodium species by carbonylation
of an aroyl rhodium complex are also reported.
Results and Discussion
Reaction of Arylglyoxyl Chlorides with Rh(acac)(CO)2
Forming Di- or Mononuclear Arylglyoxyl Complexes. When
a toluene solution of Rh(acac)(CO)2 (1) and p-ClC6H4CO-
COCl (2a, 2 equiv) was stirred for 6 h at 30 °C, a yellow
powder precipitated. Routine workup afforded an analyti-
cally pure sample of 3a in 93% yield (Scheme 1). NMR and
IR spectroscopy and elemental analysis were satisfactory,
and an X-ray diffraction study verified that 3a was a
*To whom correspondence should be addressed. E-mail: m.tanaka@
res.titech.ac.jp.
(1) (a) Ozawa, F.; Soyama, H.; Yamamoto, T.; Yamamoto, A.
Tetrahedron Lett. 1982, 23, 3383. (b) Ozawa, F.; Soyama, H.; Yanagihara,
H.; Aoyama, I.; Takino, H.; Izawa, K.; Yamamoto, T.; Yamamoto, A. J. Am.
Chem. Soc. 1985, 107, 3235. (c) Kobayashi, T.; Tanaka, M. J. Organomet.
Chem. 1982, 233, C64. (d) Sakakura, T.; Yamashita, H.; Kobayashi, T.;
Hayashi, T.; Tanaka, M. J. Org. Chem. 1987, 52, 5733, and references
therein. (e) Kobayashi, T.; Sakakura, T.; Tanaka, M. Tetrahedron Lett.
1987, 28, 272.
(2) Sen, A.; Chen, J.-T.; Vetter, W. M.; Whittle, R. R. J. Am. Chem.
Soc. 1987, 109, 148.
(3) For the chemistry of R-keto acyl complexes relevant to double
€
carbonylation and their generation, see: des Abbayes, H.; Salaun, J.-Y.
(4) Amii, H.; Kishikawa, Y.; Kageyama, K.; Uneyama, K. J. Org.
Chem. 2000, 65, 3404.
(5) Relevant catalysis using this complex will be reported elsewhere.
(6) To the best of our knowledge, there is only one publication
reporting oxidative addition of organic halides with Rh(acac)(CO)2.
See: Varshavsky, Yu. S.; Cherkasova, T. G.; Struchkov, Yu. T.; Batsanov,
A. S.; Bresler, L. S.; Marasanova, N. N. Koord. Khim. 1988, 14, 1105.
Chem. Abstr. 1989, 111, 39553t.
Dalton Trans. 2003, 1041.
r
2010 American Chemical Society
Published on Web 12/22/2010
pubs.acs.org/Organometallics

240 Organometallics, Vol. 30, No. 2, 2011
Kashiwabara and Tanaka
Table 1. Time Course of the Reaction of Rh(acac)(CO)2 with 2da
conversion (%)
Rh(acac)(CO)2
yield (%)b
2dc
4
5
b
time (h)
1
2
3
4
37
72
100
100
39
70
100
100
10
55
96
97
26
14
3
0
a Reaction conditions: 2d (0.053 mmol), Rh(acac)(CO)2 (0.053 mmol),
chloroform-d (0.5 mL), 30 °C, in an NMR tube. b Determined by 1H
NMR spectroscopy. c Determined by 19F NMR spectroscopy using
hexafluorobenzene as an internal standard.
Figure 1. Molecular structure of 3a. Thermal ellipsoids are
drawn at the 50% probability level. Hydrogen atoms were
omitted for clarity.
Scheme 2
Scheme 1
Scheme 3
chlorine-bridged dimeric arylglyoxyl complex (Figure 1).7
Its Rh-C separation in the arylglyoxyl rhodium moiety
(1.974 A) is somewhat short as compared with those of
˚
bound to the acyl functionality suppress decarbonylation
in rhodium-catalyzed reactions of acyl chlorides.8a,b,10
The reaction of C6F5COCOCl (2d) with 1 also formed an
arylglyoxyl complex (5) as transient species, but the complex
was mononuclear and was too labile in another direction to
isolate (Scheme 2). A preliminary experiment using 1 and 2d
in a 1:1 ratio at 30 °C for 6 h gave, besides [RhCl(CO)2]2,
compound 4 as the sole product.11,12 Time course study of
the reaction showed complex 5 being formed initially, but later,
the quantity of 5diminished to give 4at the expense of 5(Table 1).
The formation of 4 can be rationalized as shown in Scheme 3.
The structure of 5 could not be fully characterized due to
its lability. However, we are able to safely conclude that
compound 5 is a mononuclear arylglyoxyl complex on the
bases of (1) NMR spectral features being different from
those of 3a and 3b, (2) FAB-MS analysis displaying m/z 481
([M - Cl]þ) unlike 3a and 3b, which displayed m/z 399 and
433, respectively, both corresponding to [M/2 þ H]þ, (3)
final product 4 being formed at the expense of 5, and (4)
[RhCl(CO)2]2 being nearly the sole rhodium-containing
species in the final solution. Somewhat puzzling is that the
selective formation of 4 takes place with 2d, but not with 2a
and 2b. This may be associated with the high electrophilicity
of the C6F5COCO group,13 resulting in the facile C-acylation
with the rhodium acetylacetonate, relative to the CO dis-
sociation prerequisite for the formation of 3a and 3b.
acyl-rhodium complexes, although the deviation from the
normal range (2.058-1.992 A) is not significant8,9 and agrees
˚
with its unexpected thermal stability (mp ∼126 °C).
Plain C6H5COCOCl (2b) also reacted with 1 similarly to
afford 3b in 76% yield. However, p-MeOC6H4COCOCl (2c)
displayed totally different reaction behavior under identical
1
conditions. H NMR analysis of the resulting mixture in-
dicated that p-MeOC6H4COCl arising from decarbonyla-
tion was the major product (65%), and unreacted 2c (27%)
and 1 (89% based on 1 charged) were also found. Although
formation of 3c could not be confirmed, we can safely
presume, in view of decarbonylation and reductive elimina-
tion affording p-MeOC6H4COCl having taken place, that
oxidative addition took place to generate 3c as transient
intermediate. Thus, the reactivity of 2c differs very much
from those of 2a and 2b, but the reactivity trend agrees with
our previous observations that electronegative groups
(7) An iodine-bridged dimeric acetyl rhodium complex having nearly
the same configuration has been reported. See ref 6.
˚
(8) The Rh-C separation is shorter than 2.058-1.992 A reported
for RhCl2(CO)(COCH2Cl)(PMePh2)2, RhCl2(CO)(COCH2Cl)(PMe3)2,
Rh(COC6F5)Cl2(CO)(PMe3)2, and cis-RhCl2(COPh)(dppp), but a
˚
little longer than 1.953 A for cis-Rh(COC2H5)CI2(P(C6H5)3)2. See:
(a) Kashiwabara, T.; Kataoka, K.; Hua, R.; Shimada, S.; Tanaka, M.
Org. Lett. 2005, 7, 2241. (b) Kashiwabara, T.; Fuse, K.; Hua, R.; Tanaka, M.
Org. Lett. 2008, 10, 5469. (c) McGuiggan, M. F.; Doughty, D. H.; Pignolet,
L. H. J. Organomet. Chem. 1980, 185, 241. (d) Shie, J.-Y.; Lin, Y.-C.; Wang,
Y. J. Organomet. Chem. 1989, 371, 383.
Thermolysis of [Rh(μ-Cl)(acac)(CO)(COCOAr)]2 and Re-
action of Rh(acac)(CO)2 with PhCOCl Relevant to the Ther-
molysis. Although the foregoing experiments have already
provided a rough view of the substituent-dependent thermal
˚
(9) A very short Rh-C separation (1.85 A) was noted for Rh-
(COCH2CH2Ph)CI2(P(C6H5)3)2 in a trigonal-bipyramidal arrangement
in which the two phosphines were in apical positions. However this value
came from a preliminary X-ray structure determination, which was
stopped at R = 0.20 stage of the refinement. The final value does not
appear to have been published. See: Lau, K. S. Y.; Becker, Y.; Huang,
F.; Baenziger, N.; Stille, J. K. J. Am. Chem. Soc. 1977, 99, 5664.
(10) (a) Hua, R.; Shimada, S.; Tanaka, M. J. Am. Chem. Soc. 1998,
120, 12365. (b) Hua, R.; Onozawa, S.-y.; Tanaka, M. Chem. Eur. J. 2005, 11,
3621.
(11) Nonhebel, D. C.; Smith, J. J. Chem. Soc. (C) 1967, 1919.
(12) The structure was confirmed by X-ray analysis. See the Support-
ing Information.
(13) Kivinen, A. In The Chemistry of Acyl Halides; Patai, S., Ed.;
Interscience Publishers: London, 1972; Chapter 6.

Article
Organometallics, Vol. 30, No. 2, 2011 241
Scheme 4
Scheme 5
Figure 2. Time course of the thermolysis of complexes 3a and
3b. Thermolysis temperature: 30 °C (0-3 h), 50 °C (3-10.5 h),
60 °C (10.5-24 h).
stability of arylglyoxyl rhodium complexes, high stability of
3a as compared with 3b is more distinctly seen in the
thermolysis of these complexes. A benzene-d6 solution of
3a or 3b was heated in an NMR tube under nitrogen at 30 °C
for 3 h, then at 50 °C for an additional 7.5 h, and finally at
60 °C for an additional 13.5 h. The thermolysis proceeded
cleanly to form only Rh(acac)(CO)2 and corresponding
ArCOCl. In the thermolysis of 3a, 81% of 3a still remained
unchanged even after these three-step processes (Figure 2).
Complex 3b also survived after the second process at 50 °C,
albeit in a less significant quantity (75%), clearly suggesting
that the benzoyl group in the phenylglyoxyl group is elec-
tronegative enough to stabilize 3b. However, since 3b did not
have an extra electronegative group like chlorine in 3a, it
underwent decarbonylation more rapidly upon heating at 60 °C.
The rate constants of the thermolysis of 3a and 3b at 60 °C
were 4.2 ꢀ 10-11 and 3.0 ꢀ 10-10 mol/s, respectively.
Scheme 6
What is interesting in the thermolysis study is that 1H NMR
spectroscopy showed basically only the signals of complex 3,
aroyl chloride 7, and 1; any other species such as correspond-
ing aroyl rhodium complex 6 was not detected in an appreci-
able quantity. This indicates that the thermal decomposition
proceeds as shown in Scheme 4 and that 6, a possible inter-
mediate, is far more labile toward reductive elimination than 3.
Despite the foregoing statement, careful analysis of a 1:1
mixture of complex 1 and benzoyl chloride in benzene-d6 left
standing at room temperature for 1 h did show new species
(6-Ph), but the yield was only 3%. The yield did not change
over a period of an additional 2 h, indicating that the extent
of the reaction was thermodynamically controlled under the
conditions. The equilibrium constant under the conditions is
evaluated at 1.52ꢀ103 mol-1. Following this, benzoyl chloride
(2 equiv) was supplemented to the mixture and the reaction
was continued further for another 2 h to boost the yield to
12%. Another sequence of similar operations carried out at
60 °C also provided similar results (Scheme 5), but the yield
of 6-Ph was generally higher.14 The equilibrium constant on
the basis of the conversion after the initial 2 h is evaluated at
7.28 ꢀ 103 mol-1. Although species 6 was formed to a minor
extent under these conditions, the observations also con-
clude that the oxidative addition of 7 with 1, unlike the
reaction of 2 with 1, is not a thermodynamically favored
process and that a majority of species 6 possibly generated in
the thermolysis of 3 reductively eliminates 7 readily.
Formation of p-ClC6H4COCORhCl2(CO)(PMe3)2 by the
Reaction of p-Chlorophenylglyoxyl Chlorides with RhCl-
(CO)(PMe3)2 and Its Thermolysis. Finally, we took a brief
look at the reaction of RhCl(CO)(PMe3)2 with 2a (1 equiv) at
30 °C, which did furnish corresponding arylglyoxyl complex
8 in 61% yield (Scheme 6). We were unable to obtain a crystal
suitable for X-ray analysis. However, characterization by
NMR and IR spectroscopy combined with our previous
observations8a,b,10 supports that the structure is that illu-
strated in Scheme 6.
Thermolysis of complex 8 has revealed that the complex is
also fairly stable, in particular under carbon monoxide.
Unlike the lack of observable formation of 6 (Scheme 4) in
the thermolysis of 3a and 3b, thermolysis of 8 at 80 °C leads
to the formation of fairly stable p-chlorobenzoylrhodium
complex 9 besides other products arising from 8 and/or 9
(Table 2). The higher stability of 9 as compared with possible
intermediate 6 is associated with the trimethylphosphine
ligated to the rhodium center.
In practice of the thermolysis, a toluene-d8 solution of 8 in
an NMR tube was heated at 80 °C for 1 h under nitrogen.
Only 12% of 8 remained and 9 was found in 56% yield, as
analyzed by 1H NMR spectroscopy. After heating at 80 °C
for an additional 1 h, analysis of the reaction mixture by
31P{1H} NMR spectroscopy confirmed that 8 did not remain
in the final solution and that, in addition to 9 (47%, which
was quantified by 1H NMR spectroscopy), mer,trans-RhCl3-
(CO)(PMe3)2 (mer,trans-10) was also formed. More detailed
analysis by 1H NMR spectroscopy (with p-dimethoxybenzene
(14) As Scheme 5 indicates, upon cooling the mixture to room
temperature after the sequential treatment at 60 °C, the yield decreased.

242 Organometallics, Vol. 30, No. 2, 2011
Table 2. Thermolysis of p-ClC6H4COCORhCl2(CO)(PMe3)2 (8) at 80 °C
Kashiwabara and Tanaka
yield (%)b, Ar = p-ClC6H4
other conditionsa
recovery of 8 (%)b
9
fac,cis-10
Ar2CO
(ArCO)2
ArCOCl
Ar2
1 h, N2
2 h, N2
1 h, CO 80 atm
12
0
56
47
22
3
0
3
20
6
8
0
trace
trace
>72c
<20c
a Other conditions: 80 °C, toluene-d8 solvent. b Recovery of 8 and formation of 9 and mer,trans-10 were confirmed by 31P NMR spectroscopy of the
reaction mixture (toluene-d8 solution) by using authentic samples. Note that mer,trans-10, which was detected in the toluene-d8 solution, isomerized to
fac,cis-10 upon replacement of the solvent with CDCl3 due to the solvent-dependent isomerization. Recovery of 8 and the yields of the products listed
were determined by 1H NMR spectroscopy in CDCl3 after evaporation of toluene-d8 and dissolving the residue in CDCl3. Yields of Ar2CO, (ArCO)2,
ArCOCl, and Ar2 are referenced to the quantity of Ar group. c See ref 16.
as internal standard) using a CDCl3 solution of a residue
obtained by evaporation of the final mixture in toluene-d8
revealed the formation of dichlorobenzil, p-chlorobenzoyl
chloride, and dichlorobiphenyl. Dichlorobenzophenone was
not found at all. mer,trans-RhCl3(CO)(PMe3)2 (mer,trans-
10), which was detected before the replacement of the solvent
from toluene-d8 to CDCl3, was not found either, but instead,
fac,cis-RhCl3(CO)(PMe3)2 (fac,cis-10) was found in the
CDCl3 solution. In view of facile isomerization between
mer,trans-10 and fac,cis-10, depending on the polarity of
the solvent,15 mer,trans-10 found in the thermolysis mixture
(a toluene-d8 solution) appeared to have isomerized to fac,cis-
10 when the toluene-d8 solvent was replaced by CDCl3.
On the other hand, upon thermolysis under CO (80 atm) at
80 °C for 1 h, >72% of 8 survived and 9 was formed only in
<20%.16 Other products were also formed as summarized in
Table 2.
Attempts to Generate p-ClC6H4COCORhCl2(CO)(PMe3)2
by Carbonylation of p-ClC6H4CORhCl2(CO)(PMe3)2. Worth
noting, associated with the foregoing formation and the stability
of 8, is that treatment of p-ClC6H4CORhCl2(CO)(PMe3)2 (9)
with carbon monoxide generates species 8 in situ (Scheme 5),
albeit in a low yield due to side reactions of 9. When a solution
of 9 in CDCl3 was heated at 40 °C under carbon monoxide
(80 atm) for 6 h, only decomposition of 9 appeared to have
taken place to give dichlorobenzophenone, dichlorobenzil,
p-chlorobenzoyl chloride, and others, besides unreacted 9
(18%) and fac,cis-RhCl3(CO)(PMe3)2 (10, 46%) as major
rhodium-containing species (Table 3, Figure 3).17 However,
another reaction run at 80 °C in toluene-d8 for 6 h resulted in
emergence of 31P NMR signals characteristic of 8 (-3.44 ppm,
d, JP-Rh = 84.1 Hz in CDCl3). Its yield was estimated at 4% by
1H NMR spectroscopy, and other products from 8 and/or 9
were also formed. Although the yield of 8 is low, this observa-
tion is encouraging in view of application to rhodium-catalyzed
reactions of aroyl chlorides under CO pressure to afford
arylglyoxylated products.5
To summarize, arylglyoxyl rhodium complexes can be
readily synthesized by oxidative addition of arylglyoxyl
chloride. Unlike R-keto acyl palladium complexes, they are
fairly stable as opposed to our tacit understanding and can be
generated by CO insertion reaction with aroyl rhodium
complexes. Catalysis involving them as key intermediates
will be reported shortly.
Experimental Section
[Rh(μ-Cl)(acac)(CO)(COCOC6H4-p-Cl)]2 (3a). To a yellow
solution of Rh(acac)(CO)2 (97.1 mg, 0.376 mmol) in toluene
(5 mL) placed in a 20 mL Schlenk tube was added p-ClC6H4CO-
COCl (151.8 mg, 0.752 mmol) at 30 °C, and the mixture was
stirred for 6 h at the temperature. While the reaction was in
progress, yellow powders precipitated. Volatiles were removed
in vacuo to give a yellow material, which was rinsed with ether
(1 mL) and hexane (3 mL ꢀ 3) and was dried in vacuo to afford
analytically pure 3a (153.0 mg, 0.176 mmol, 93%): yellow needles:
mp 125.8-126.3 °C (under nitrogen). 1H NMR (300 MHz,
CDCl3): δ 7.78 (d, J=8.58 Hz, 2H, m-Ph), 7.44 (d, J=8.58 Hz,
2H, o-Ph), 5.62 (s, 1H, acac-CH), 2.19 (s, 3H, Me), 1.97 (s, 3H,
Me). 13C{1H} NMR (75 MHz, CDCl3):δ209.8 (br, Rh-CO), 186.9
(acac-CO), 186.7 (ArCOCORh), 181.4 (acac-CO), 176.9 (d,
1JRh-C=66.8 Hz, Rh-COCO), 141.6 (p-Ph), 131.5 (o-Ph), 129.4
(m-Ph), 128.0 (ipso-Ph), 100.5 (CH), 26.9 (Me), 26.4 (Me). IR
(KBr, cm-1): 2106 (Rh-CO), 1736, 1678 (RhCOCOAr), 1558,
(15) (a) Browning, J.; Goggin, P. L.; Goodfellow, R. J.; Norton,
M. G.; Rattray, A. J. M.; Taylor, B. F.; Mink, J. J. Chem. Soc., Dalton
Trans. 1977, 2061. (b) Kashiwabara, T.; Fuse, K.; Muramatsu, T.; Tanaka, M.
J. Org. Chem. 2009, 74, 9433.
(16) To evaluate the conversion from 8 to 9, we have to take into
account the fact that the conversion is much faster under nitrogen than
that under CO. In the thermolysis under CO, CO was introduced
approximately 3 min after the toluene-d8 solution of complex 8 had been
charged to the autoclave preheated at 80 °C. Then the mixture was stirred
for 1 h at 80 °C. In view of the foregoing thermolysis effected in the
absence of carbon monoxide, thermal decomposition of 8 is envisioned
to have taken place, during this 3 min time lag under nitrogen atmo-
sphere, to a small extent (∼5% conversion of 8; see Table S1 for detailed
conversion data obtained in the thermolysis under nitrogen to estimate
this conversion during the 3 min time lag). If we deduct the conversion of
8 made before introduction of CO, the net conversion of 8 after CO
introduction can be estimated at ∼23% (recovery of 8: ∼77%) and the
quantity of 9 formed at ∼17%. The same argument can be made for the
quantity of other products listed in Table 2.
(17) We are unable to rigorously exclude the possibility that the
species displaying a doublet at -7.11 ppm (JP-Rh = 117.2 Hz) in the
31P NMR spectrum (Figure 3, the second spectrum from the bottom) is
an intermediate toward the formation of 8.

Article
Organometallics, Vol. 30, No. 2, 2011 243
Table 3. Reaction of p-ClC6H4CORhCl2(CO)(PMe3)2 (9) with Carbon Monoxide
yield (%)b, Ar = p-ClC6H4
conditionsa
recovery of 9 (%)b
8
fac,cis-10
Ar2CO
(ArCO)2
ArCOCl
Ar2
CDCl3, 40 °C
toluene-d8, 80 °C
18
0
0
4
46
61
15
33
20
10
21
0
0
5
a Other conditions: CO 80 atm, 6 h. b Refer to footnote b of Table 2.
1725, 1675 (RhCOCOPh), 1560, 1522 (acac-CO). MS (FAB,
matrix=3-nitrobenzyl alcohol): m/z 399 ([M/2 þ H]þ). HRMS
(FAB, matrix=3-nitrobenzyl alcohol): calcd for C14H12ClO5Rh
([M/2 þ H]þ) m/z 398.9507, found 398.9510. Anal. Calcd for
C28H24Cl2O10Rh2: C, 42.19; H, 3.03. Found: C, 41.44; H, 3.35.
For NMR spectra of this compound, see Appendices 1 and 2 in
the Supporting Information.
Attempted Synthesis of [Rh(μ-Cl)(acac)(CO)(COCOC6H4-p-
OMe)]2 (3c) by the Reaction of Rh(acac)(CO)2 with p-MeOC6H4-
COCOCl (2c). To a yellow solution of Rh(acac)(CO)2 (8.2 mg,
0.0318 mmol) in C6D6 (0.4 mL) placed in an NMR tube was
added p-MeOC6H4COCOCl (9.0 mg, 0.0454 mmol) at room
temperature, and the mixture was stirred for 6 h at 30 °C.
1H NMR analysis of the final mixture using intentionally
added silicone grease as an internal standard revealed that
p-MeOC6H4COCl (65% relative to the quantity of 2c charged),
unreacted 2c (27%), and 1 (89% based on 1 charged) were
present. No signal assignable to 3c was observed.
Reaction of Rh(acac)(CO)2 with C6F5COCOCl (2d) Affording
4. To a yellow solution of Rh(acac)(CO)2 (30.2 mg, 0.233 mmol)
in toluene (5 mL) placed in a 20 mL Schlenk tube was added
C6F5COCOCl (60.3 mg, 0.0233 mmol) at room temperature,
and the mixture was stirred for 6 h at 30 °C. Volatiles were
removed in vacuo. A ocher solid material obtained was rinsed
with hexane (3 mL ꢀ 3) and ether (1 mL ꢀ 2) and was dried in
vacuo to afford analytically pure 4 (56.3 mg, 0.174 mmol, 75%).
4-Acetyl-2-hydroxy-5-methyl-2-pentafluorophenyl-3-furanone
(4): pale yellow solid, mp 152.9-153.8 °C. 1H NMR (300 MHz,
CDCl3): δ 5.39 (br, 1H, OH), 2.69 (s, 3H, Me), 2.48 (s, 3H, Me).
13C{1H} NMR (75 MHz, CDCl3): δ 196.8 (CH3CO), 193.6 (d,
JCF = 30.4 Hz, C6F5CCO), 147.4-136.3 (m, ArF), 112.8
(C6F5COCCH3), 109.4 (t, JCF = 10.8 Hz, COH), 100.6 (AcC),
29.9 (CH3CO), 18.4 (C6F5COCCH3). 19F NMR (282 MHz,
CDCl3): δ -139.1 (m, 2F, o-Ph), -150.1 (m, 1H, p-Ph),
-160.2 (m, 2F, m-Ph). IR (KBr, cm-1): 3102 (br, νOH), 1731
(νCO), 1657 (νCO), 1550 (νCdC). MS (FAB, matrix = 3-nitro-
benzyl alcohol): m/z 323 ([M þ H]þ). Anal. Calcd for
C13H7F5O4: C, 48.46; H, 2.19. Found: C, 48.10; H, 2.28.
Time-Course Study of the Reaction of Rh(acac)(CO)2 with
C6F5COCOCl (2d). To a yellow solution of Rh(acac)(CO)2
(13.7 mg, 0.0531 mmol) in CDCl3 (0.5 mL; a small quantity of
silicone grease was intentionally added as an internal standard
for 1H NMR spectroscopy) placed in an NMR tube were added
C6F5COCOCl (13.8 mg, 0.0531 mmol) and C6F6 (2.0 μL;
internal standard for 19F NMR spectroscopy) at room tempera-
Figure 3. 31P NMR spectra of final mixtures of carbonylation
of complex 9 (after replacement of the solvent with CDCl3 in the
case of the reaction run in toluene-d8) with reference complexes.
1519 (acac-CdO). MS (FAB, matrix=3-nitrobenzyl alcohol):
m/z 433 ([M/2 þ H]þ). Anal. Calcd for C28H22Cl4O10Rh2: C,
38.83; H, 2.56. Found: C, 38.44; H, 2.92.
[Rh(μ-Cl)(acac)(CO)(COCOC6H5)]2 (3b). To a yellow solu-
tion of Rh(acac)(CO)2 (74.5 mg, 0.289 mmol) in toluene (3 mL)
placed in a 20 mL Schlenk tube was added PhCOCOCl (50.0 mg,
0.296 mmol) at room temperature, and the mixture was stirred
for 6 h at that temperature. Volatiles were removed in vacuo. A
yellow oil obtained was rinsed with ether (1 mL ꢀ 2) and hexane
(1 mLꢀ2) to generate a yellow powder, which was dried in vacuo
to afford 3b (87.1 mg, 0.11 mmol, 76%): yellow crystals, mp 72.3-
73.9 °C (under nitrogen, dec). 1H NMR (300 MHz, CDCl3): δ
7.84 (d, J=8.13 Hz, 2H, Ph), 7.59 (t, J=7.59 Hz, 1H, p-Ph), 7.46
(m, 2H, Ph), 5.48 (s, 1H, acac-CH), 2.17 (s, 3H, acac-Me), 1.85
(s, 3H, acac-Me). 13C{1H} NMR (75 MHz, CDCl3): δ 210.2 (d,
1JRh-C=29.0 Hz, Rh-CO), 187.1 (acac-CO), 186.6 (PhCOCORh),
1
ture, and H and 19F NMR spectra were recorded. Then the
mixture was stirred for 4 h at 30 °C. The progress of the reaction
was monitored by 1H and 19F NMR analysis every 1 h (Table 1).
FAB-MS analysis using a 1 μL portion taken from the
reaction mixture was also performed after stirring for 2 h to
1
182.4 (acac-CO), 177.3 (d, JRh-C = 66.9 Hz, PhCOCORh),
134.8 (p-Ph), 130.4 (o-Ph), 129.0 (ipso-Ph), 128.2 (m-Ph), 100.5
(CH), 26.3 (Me), 25.7 (Me). IR (KBr, cm-1): 2108 (Rh-CO),

244 Organometallics, Vol. 30, No. 2, 2011
Kashiwabara and Tanaka
gain information of mononuclear R-keto acyl complex 5. Also,
after 4 h, the IR spectrum was recorded to show νCO bands at
2104, 2085, 2036, and 2002 cm-1, which agreed with literature
data for [RhCl(CO)2]2.18 Finally, PMePh2 (26.3 mg, 0.1314
mmol) was added to the mixture; bubbles coming out from
the mixture could be confirmed by sight, and RhCl(CO)-
(PMePh2)2 having been formed was also confirmed by 1H and
Rh-COCOAr). HRMS (FAB, matrix=3-nitrobenzyl alcohol):
calcd for C15H22Cl2O3P2Rh ([M - Cl]þ) m/z 484.9476, found
484.9467. For NMR spectra of this compound, see Appendices
3-5 in the Supporting Information.
RhCl2(CO)(COC6H4Cl)(PMe3)2 (9). To a yellow solution of
RhCl(CO)(PMe3)2 (100.1 mg, 0.314 mmol) in toluene (5 mL)
placed in a 20 mL Schlenk tube was added p-ClC6H4COCl(40.0μL,
0.315 mmol) at room temperature. The mixture was stirred at
30 °C for 3 h, while the color changed gradually to slightly dark
yellow. The resulting mixture was evaporated. The yellow solid
residue was rinsed with ether (2 mL ꢀ 2) and hexane (3 mL ꢀ 2)
and was dried in vacuo to afford analytically pure 9 (136.2 mg,
0.276 mmol, 88%): pale yellow powder, mp 100.2-102.8 °C
(under nitrogen, dec). 1H NMR (CDCl3, 400 MHz): δ 8.06 (d, J =
8.79 Hz, 2H, o-Ph), 7.50 (d, J=8.79 Hz, 2H, m-Ph), 1.71 (virtual
t, J=3.60 Hz, 18H, PMe3). 13C{1H} NMR (75 MHz, CDCl3): δ
132.4, 130.1, 128.2, 129.4, 14.6 (virtual t, JPC=16.7 Hz, Me).
Signals associated with carbons in the carbonyl groups were too
weak to be observable due presumably to multiple coupling with
P and Rh nuclei. 31P{1H} NMR (CDCl3, 162 MHz): δ -4.63 (d,
JP-Rh = 86.02 Hz). IR (KBr, cm-1): 2083 (RhCO), 1675
(RhCOAr). MS (FAB, matrix = 3-nitrobenzyl alcohol): m/z
457 ([M - Cl]þ). Anal. Calcd for C14H22Cl3O2P2Rh: C, 34.04;
H, 4.49. Found: C, 33.90, H, 4.13.
Treatment of RhCl2(CO)(COC6H4Cl)(PMe3)2 with CO (80
atm) at 80 °C in Toluene-d8. To RhCl(CO)(PMe3)2 (18.3 mg,
0.0370 mmol) placed in a 20 mL Schlenk tube were added
silicone grease (2.9 mg; internal standard for 1H NMR analysis)
and toluene-d8 (2.0 mL). The solution was transferred to a
20 mL autoclave (fitted with a glass insert). Carbon monoxide
was introduced at 80 atm, and the autoclave was heated at 80 °C
for 6 h. Carbon monoxide was vented, and the yellow solution
taken out from the autoclave was transferred to a Schlenk tube
and evaporated. p-Dimethoxybenzene (5.8 mg) as internal
standard for NMR spectroscopy and CDCl3 were added, and
the resulting solution was analyzed by NMR spectroscopy.
1H and 31P NMR analysis revealed that rhodium-containing
species, 8 and fac,cis-RhCl3(CO)(PMe3)2 (fac,cis-10), were in
this CDCl3 solution together with other organic compounds
shown in Table 3. Starting rhodium complex 9 did not remain at
all. The 31P NMR spectrum of this CDCl3 solution and that of a
similar reaction run at 40 °C for 6 h are shown in Figure 3, which
includes relevant species for comparison.
1
31P NMR spectroscopy (89% as quantified by H NMR spec-
troscopy).
RhCl(acac)(COCOC6F5)(CO)2 (5). This compound is a tran-
sient species found during the foregoing time-course study. Collec-
tion of full characterization data was hampered by the lability.
However, its identity is well supported as discussed in the foregoing
section. 1H NMR (300 MHz, CDCl3): δ 2.36 (s, 6H, Me), 5.51 (s,
1H, acac-CH). 19F NMR (282 MHz, CDCl3): δ -136.2 (m, 2F,
o-Ph), -144.1 (m, 1F, p-Ph), -159.6 (m, 2F, m-Ph). MS (FAB,
matrix=3-nitrobenzyl alcohol): m/z 481 ([M - Cl]þ).
Time-Course Study on Thermolysis of [Rh(μ-Cl)(acac)(CO)-
(COCOC6H4X)]2 (3a; X=Cl, 3b; X=H). Complex 3a (6.6 mg) or
3b (8.54 mg) was dissolved in C6D6 (0.5 mL, a small quantity of
silicone grease was added intentionally as internal standard for
NMR spectroscopy), and the solution was heated at 30 °C for 3
h, then at 50 °C for an additional 7.5 h, and finally at 60 °C for a
final 13.5 h (altogether 24 h). The progress of the thermolysis
was monitored by 1H NMR spectroscopy. The progress is
illustrated in Figure 2.
RhCl(acac)(CO)2(COPh) (6-Ph). This compound could not
be isolated due to instability, but the compound in the mixture of
the reaction of Rh(acac)(CO)2 (1) with PhCOCl (7-Ph) in benzene-
d6 could be characterized. 1H NMR (300 MHz, C6D6): δ 7.93 (m,
2H, o-Ph), 7.08 (t, J=7.38 Hz, p-Ph), 4.93 (s, 1H, CH), 1.58 (s,
6H, Me). The signal of m-Ph overlapped with free PhCOCl. IR
(benzene-d6 solution, cm-1): 2088, 2032 (Rh-CO), 1731 (Rh-
COPh), 1205, 1174 (acac-CO). HRMS (FAB, matrix=3-nitro-
benzyl alcohol): calcd for C14H13ClO5Rh ([M þ H]þ) m/z
398.9507, found 398.9519.
RhCl2(CO)(COCOC6H4Cl)(PMe3)2 (8). To a yellow solution
of RhCl(CO)(PMe3)2 (97.0 mg, 0.305 mmol) in toluene (5 mL)
placed in a 20 mL Schlenk tube was added p-ClC6H4COCOCl
(61.9 mg, 0.305 mmol) at room temperature. The color of the
mixture changed to reddish-yellow instantaneously. The mixture
was stirred at 30 °C for 30 min and was evaporated. The orange
solid residue was rinsed with ether (2 mL ꢀ 2) and hexane (3 mL ꢀ
2) and was dried in vacuo to afford 8 (97.14 mg, 0.186 mmol,
61%): pale yellow powder, mp 112.0-113.3 °C (under nitrogen).
1H NMR (CDCl3, 400 MHz): δ 7.98 (d, J=8.39 Hz, 2H, o-Ph),
7.46 (d, J=8.39 Hz, 2H, m-Ph), 1.70 (virtual t, J=4.00 Hz, 18 H,
Acknowledgment. This work was supported by a
Grant-in-Aid for Scientific Research on Priority Areas
(No. 18065008) from MEXT, Japan, and by a research
fellowship to T.K. from JSPS.
PMe3). 13C{1H} NMR (CDCl3, 75 MHz): δ 223.8 (dt, JC-P
5.81 Hz, JC-Rh=30.52 Hz, Rh-CO), 186.7 (dt, JC-P=2.18 Hz,
C-Rh = 4.36 Hz, RhCOCO), 182.7 (dt, JC-P = 10.17 Hz,
=
J
JC-Rh = 66.85 Hz, RhCOCO), 141.6 (p-Ph), 131.4 (o-Ph),
129.3 (m-Ph), 128.5 (ipso-Ph), 14.22 (virtual t, JC-P=18.17 Hz,
=
Supporting Information Available: Experimental details and
characterization of new compounds, NMR spectra of 3b and 8,
which did not furnish satisfactory elemental analysis, and
crystallographic data for 3a and 4 in CIF format. This material
is available free of charge via the Internet at http://pubs.
acs.org.
PMe3). 31P{1H} NMR (CDCl3, 162 MHz): δ -3.44 (d, JP-Rh
84.08 Hz). IR (KBr, cm-1): 2072 (Rh-CO), 1714, 1668 (br,
(18) Cramer, R. Inorg. Synth. 1974, 15, 17.