Solvent-free ketone hydrogenations catalyzed by molybdenum complexes

Article abstract of DOI:10.1039/b401760a

Et₂C=O is hydrogenated under solvent-free conditions using a catalyst prepared by hydride abstraction from HMo-(CO)₂] η⁵:η¹-C₅H₄(CH ₂)₂PCy₂]; the catalyst functions at low catalyst loadings (< 0.4 mol%).

Full text of DOI:10.1039/b401760a

Solvent-free ketone hydrogenations catalyzed by molybdenum
complexes†
Barbara F. M. Kimmich,a Paul J. Fagan,b Elisabeth Hauptmanb and R. Morris Bullock*a
a Chemistry Department, Brookhaven National Laboratory, Upton, New York 11973, USA.
E-mail: bullock@bnl.gov; Fax: (+1) 631 344-5815
b Central Research and Development Department, The DuPont Company, Experimental Station, P.O. Box
80328, Wilmington, Delaware 19880, USA
Received (in West Lafayette, IN, USA) 4th February 2004, Accepted 23rd February 2004
First published as an Advance Article on the web 22nd March 2004
Et2CNO is hydrogenated under solvent-free conditions using a
catalyst prepared by hydride abstraction from HMo-
(CO)2[h5:h1-C5H4(CH2)2PCy2]; the catalyst functions at low
catalyst loadings ( < 0.4 mol%).
Environmental concerns provide a compelling incentive to develop
chemical reactions that minimize waste production. The develop-
ment of solvent-free (or ‘solventless’) reactions1,2 is an especially
attractive goal in the context of Green Chemistry,3 since solvents
are used on a large scale (over 15 billion kilograms of organic
solvents produced worldwide annually!).1 Catalytic hydrogen-
ations of ketones and other organic substrates are traditionally
carried out using platinum-group metals like rhodium and ruthe-
nium,4 though there are a few examples of catalysts not requiring
precious metals. Darensbourg and co-workers reported the catalytic
hydrogenation of ketones at 125 °C using
5
mol%
[W(CO)5(OAc)]2 as the catalyst precursor.5 Our research on the
reactivity of metal hydrides has led to a class of homogeneous
hydrogenation catalysts using the base metals molybdenum and
tungsten.6 Research on these ‘Cheap Metals for Noble Tasks’ offers
the potential for development of new types of catalysts that may
provide cost savings and potentially other advantages over
conventional homogeneous catalysts. One goal of our research is to
develop catalysts based on inexpensive metals that may be useful in
industrial processes to replace stoichiometric reagents like LiAlH4
or NaBH4. Our efforts thus far have concentrated on using the
mechanistic understanding gained from studies of proton transfer
and hydride transfer reactions of metal hydrides to guide the
rational design of catalysts. A series of ketone complexes,
Scheme 1
5
1
to formation of the molybdenum hydrides HMo(CO)2[h :h -
C5H4(CH2)2PR2]‡ and expulsion of one CO as the phosphine
[CpM(CO)2(PR3)(h -Et2CNO)]+BArA42 [R = Ph or Me; M = Mo
1
or W; ArA = 3,5-bis(trifluoromethyl)phenyl], were shown to
function as catalyst precursors for the hydrogenation of ketones
under mild conditions (room temperature, 4 atm H2).6 The
proposed mechanism for these ionic hydrogenations is shown in
Scheme 1.
coordinates to the metal. The metal hydrides exhibit broad 31P
NMR resonances at d 74.4 (R = Ph), 280.6 (R = Cy) and 103.8
t
The turnover rates are relatively slow, however, and the catalysts
suffer decomposition. We report here a new generation of Mo
catalysts that exhibits substantially improved lifetimes, and which
are shown to be effective for the hydrogenation of ketones under
solvent-free conditions and low catalyst loadings.
(R = Bu), consistent with coordination of the phosphine to the
5
1
metal. The 31P NMR resonance of HMo(CO)2[h :h -
C5H4(CH2)2PCy2] at 289 °C de-coalesces into two shaper
resonances (w1/2 ≈ 7 Hz) at d 78.7 (98%) and 89.3 (2%). The 1H
5
1
NMR spectrum of HMo(CO)2[h :h -C5H4(CH2)2PCy2] recorded at
289 °C also indicates 98% trans isomer. These spectra indicate that
the hydrides exist as an interconverting mixture of cis and trans
isomers. Isomerization between the trans and cis isomers was
reported previously for closely related complexes with un-
substituted Cp ligands such as CpMo(CO)2(PPh3)H,10 CpMo-
(CO)2(PMe3)H,11 and CpMo(CO)2(PCy3)H.12 Whereas such iso-
merizations are well-known, it is notable that the trans isomer
The appearance of phosphonium cations, HPR3+, in catalytic
1
reactions using [CpM(CO)2(PR3)(h -Et2CNO)]+ suggested that the
decomposition pathway involved dissociation of a phosphine,
which is protonated under the reaction conditions. We sought to
prepare catalysts with improved lifetimes by suppression of
phosphine dissociation through chelation of the phosphine to the
cyclopentadienyl ring by a two-carbon bridge. Reaction of
(CO)3Mo(diglyme)7 with Li+[C5H4(CH2)2PR2]2 8 produced Li+-
5
1
strongly predominates in HMo(CO)2[h :h -C5H4(CH2)2PCy2], in
5
{Mo(CO)3[h -C5H4(CH2)2PR2]}2. IR spectra of the n(CO) bands
contrast to HMo(CO)2(PCy3) where the cis isomer is favored by a
were similar to those previously analyzed in detail for Li+Mo-
5
1
ratio of 9:1.12 The structure of HMo(CO)2[h :h -C5H4(CH2)2PCy2]
was determined by single crystal X-ray diffraction and indicates the
trans isomer in the solid state; details of the structure will be given
in a full paper.
(CO)3Cp2,9 and 31P NMR peaks at d 213.2 (R = Ph), 21.8 (R =
Cy) and 31.8 (R
=
tBu) were indicative of uncoordinated
phosphines. Protonation of the anionic complexes with HOAc led
5
1
The metal hydrides, HMo(CO)2[h :h -C5H4(CH2)2PR2], react
cleanly with HOTf (OTf = OSO2CF3) at room temperature to form
† Electronic supplementary information (ESI) available: additional spectro-
scopic data and description of the synthetic procedures. See http://
www.rsc.org/suppdata/cc/b4/b401760a/
5
1
cis- and trans-Mo(CO)2[h :h -C5H4(CH2)2PR2](OTf), presumably
through unobserved dihydride or dihydrogen complexes that
1014
C h e m . C o m m u n . , 2 0 0 4 , 1 0 1 4 – 1 0 1 5
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

5
1
Table 1 Hydrogenation of Et2CNO by HMo(CO)2[h :h -C5H4(CH2)2PR2]
2
5
1
+ Ph3C+BArA4 or Mo(CO)2[h :h -C5H4(CH2)2PR2](OTf). 4 atm H2,
except where noted
TON(total)
t = 10 daysa
R
A2
T/°C
Mo/mol%
2
2
2
2
Ph
BArA4
BArA4
BArA4
BArA4
BF4
OTf
50
23
50
50
50
50
75
50
0.35
0.35
0.35
0.35
0.35
0.35
0.086
0.35
62 [1]
23 [ < 1]
132 [10]
283 [34]b
99 [6]
120 [4]
462 [16]
82 [6]
Cy
Cy
Cy
Cy
Cy
Cy
tBu
eliminate H2. Clean hydride transfer reactivity from HMo-
2
5
1
2
(CO)2[h :h -C5H4(CH2)2PR2] to Ph3C+BF4 in CD3CN is ob-
served, producing complexes with acetonitrile ligands, trans- and
OTf
cis-Mo(CO)2[h :h -C5H4(CH2)2PR2(NCCD3)]+BF42. The cis and
trans isomers of these complexes are readily distinguished by their
1H NMR spectra. The trans isomers have a plane of symmetry and
exhibit two sets of resonances for the C5H4 protons. In contrast, the
cis isomer has no plane of symmetry, so its cyclopentadienyl
protons appear as four separate multiplets.
5
1
2
BArA4
a TON (turnover number) after 10 days. The number in brackets indicates
the TON for the ether condensation product, (Et2CH)2O. b H2 pressure of 55
atm; hydrogenation complete at t = 8 days.
In our earlier studies, the ketone complexes [CpM(CO)2-
PR3(Et2CNO)]+BArA42 (M = Mo, W; R = Ph, Me) were isolated
and used as the catalyst precursors. Catalysis from CpM(CO)2-
(PCy3)H was initated by carrying out the hydride transfer in situ,
without isolation of the ketone complex. Similar reactions of
HMo(CO)2[h :h -C5H4(CH2)2PPh2] with Ph3C+BArA4 with 1–2
equivalents of Et2CNO gave spectral evidence for the formation of
the ketone complexes, but pure ketone complexes were not isolated.
Thus the catalytic reactions were carried out by hydride transfer
from the metal hydride to Ph3C+BArA42 in situ, as was done for the
PCy3 complexes previously.
This new generation of Mo catalysts was rationally designed
from mechanistic principles to disfavor a decomposition pathway
identified for previously studied complexes. Several advantages
were found with these new complexes, including low catalyst
loadings ( < 0.4 mol%), higher thermal stability, substantially
longer lifetimes (hundreds of turnovers), and utility under solvent-
free conditions. Further improvements are being sought to produce
catalysts that might provide attractive replacements for stoichio-
metric hydrogenation reagents like LiAlH4.
We thank the U.S. Department of Energy, Division of Chemical
Sciences, Office of Basic Energy Sciences, and the Laboratory
Technology Research Program for support. This research was
carried out at Brookhaven National Laboratory under contract DE-
AC02-98CH10886 with the U.S. Department of Energy.
5
1
2
Initial conditions for the catalytic hydrogenations were identical
to those used earlier in the non-bridged systems (30 mM metal
hydride + 30 mM Ph3C+BArA42 in CD2Cl2 under 4 atm H2 at 23 °C,
Notes and references
5
1
‡ Spectral data for HMo(CO)2[h :h -C5H4(CH2)2PCy2]: 1H NMR (400
2
MHz, CD2Cl2): d 25.81 (d, JPH = 26.8, 1H, MoH), 1.24–2.00 (m, 22H,
10 equivalents of Et2CNO (initially 300 mM). Under these
conditions, the C2-PCy2 catalyst produced 2.8 turnovers in 6 h,
indicating that it is less reactive than the catalyst formed from
CpMo(CO)2(PCy3)H, which produced about 8.6 turnovers in 6 h.
Although the new C2-bridged catalyst is initially less active than the
related complex without the C2 bridge, an advantage of the C2-
bridged complex is its much longer lifetime compared with the
unbridged catalyst precursor. While these conditions provided a
direct comparison of the bridged vs. unbridged catalysts and
afforded the convenience of direct NMR tube monitoring of the
reaction, they were not ideal since it was necessary to periodically
replenish the reaction tube with more ketone and H2. Solvent-free
hydrogenations were carried out neat Et2CNO, and the results are
summarized in Table 1.
Cy), 2.38 (m, 2H, PCH2), 2.47 (m, 2H, C5H4CH2), 4.96 (br, 2H, a-C5H4),
5.09 (m, 2H, b-C5H4); 31P{1H} NMR (162.0 MHz, CD2Cl2): d 80.6 (br, w1/2
= 200 Hz); 13C{1H} NMR (100.6 MHz, CD2Cl2): d 25.8 (br, PCH2), 26.7
(s, m- or p-Cy), 27.6 (d, 2JCP = 13 Hz, o-Cy), 29.0 (s, m- or p-Cy), 37.0 (d,
2JCP = 23 Hz, ipso-Cy), 37.3 (v br, C5H4CH2), 81.8 (s, o-C5H4), 87.4 (s, m-
C5H4), 125.1 (d, 3JCP = 6 Hz, ipso-C5H4), 234 (v br, CO); IR (THF): n(CO)
= 1927 (s), 1848 (vs) cm21; Anal. calc. for C21H31O2PMo (442.4): C,
57.02; H, 7.06. Found: C, 56.92, H, 6.99%. Additional spectroscopic data
and a description of the synthetic procedures are provided in the ESI.†
1 J. M. DeSimone, Science, 2002, 297, 799–803.
2 G. W. V. Cave, C. L. Raston and J. L. Scott, Chem. Commun., 2001,
2159–2169.
3 P. T. Anastas and M. M. Kirchhoff, Acc. Chem. Res., 2002, 35,
686–694.
4 P. A. Chaloner, M. A. Esteruelas, F. Joó and L. A. Oro, Homogeneous
Hydrogenation, Kluwer Academic Publishers, Dordrecht/Boston/Lon-
don, 1994.
5 P. A. Tooley, C. Ovalles, S. C. Kao, D. J. Darensbourg and M. Y.
Darensbourg, J. Am. Chem. Soc., 1986, 108, 5465–5470.
6 R. M. Bullock and M. H. Voges, J. Am. Chem. Soc., 2000, 122,
12 594–12 595; M. H. Voges and R. M. Bullock, J. Chem. Soc., Dalton
Trans., 2002, 759–770.
The PCy2-bridged systems give higher turnovers than analogs
with either a PPh2 or PtBu2 bridge. The C2-bridge that chelates the
cyclopentadienyl to the phosphine clearly confers substantial
additional stability to the complex, as shown by the experiments at
50 and 75 °C. In contrast, the unchelated systems reported earlier
showed some decomposition even at room temperature. The
2
BArA42 anion gives superior performance compared with the BF4
anion, presumably due to more facile decmposition pathways
available to the BF4 complexes. Another improvement over the
7 M. S. Kralik, A. L. Rheingold, J. P. Hutchinson, J. W. Freeman and R.
D. Ernst, Organometallics, 1996, 15, 551–561.
2
8 R. T. Kettenbach, W. Bonrath and H. Butenschön, Chem. Ber., 1993,
126, 1657–1669; H. Butenschön, Chem. Rev., 2000, 100, 1527–1564.
9 M. Y. Darensbourg, Prog. Inorg. Chem., 1985, 33, 221–274.
10 J. W. Faller and A. S. Anderson, J. Am. Chem. Soc., 1970, 92,
5852–5860.
11 P. Kalck, R. Pince, R. Poilblanc and J. Roussel, J. Organomet. Chem.,
1970, 24, 445–452.
12 T.-Y. Cheng, B. S. Brunschwig and R. M. Bullock, J. Am. Chem. Soc.,
1998, 120, 13 121–13 137.
unchelated earlier systems is that complexes with a bound triflate
ligand give catalysis, though as shown in the table for reactions
5
1
2
with HMo(CO)2[h :h -C5H4(CH2)2PCy2], the BArA4 counterion
provides a better performance than the complex with a bound
triflate. Although most experiments were carried out at 4 atm H2, a
higher pressure can be beneficial. Of particular note is the complete
hydrogenation of neat Et2CNO catalyzed by 0.35 mol% Mo
complex in 8 days at 50 °C at 55 atm H2.
C h e m . C o m m u n . , 2 0 0 4 , 1 0 1 4 – 1 0 1 5
1015