CuI/H2/NaOH-catalyzed cross-coupling of two different alcohols for carbon-carbon bond formation: "Borrowing hydrogen"?

Article abstract of DOI:10.1002/chem.201101752

Not borrowed: A new method was developed for the cross-coupling of two different alcohols (see scheme). This carbon-carbon bond formation was realized by using a small amount of CuBr (0.05-0.2 mol %) and NaOH (4-20 mol %) in the presence of H₂ (1 atm), and provides an alternative procedure for longer-chain alcohols. The catalytic cycle was different from those reported to date for the Guerbet reaction involving "borrowing hydrogen". Copyright

Full text of DOI:10.1002/chem.201101752

DOI: 10.1002/chem.201101752
CuI/H2/NaOH-Catalyzed Cross-Coupling of Two Different Alcohols for
Carbon–Carbon Bond Formation: “Borrowing Hydrogen”?
Takashi Miura,[a] Osamu Kose,[a] Feng Li,[a, c] Sun Kai,[a, d] and Susumu Saito*[a, b]
To minimize salt waste byproducts and to save energy and
resources during the manufacture of longer-chain (higher)
alcohols, transition-metal-catalyzed direct coupling between
alcohols has recently been reinvestigated.[1] Homocoupling
of the carbon atoms of primary alcohols was first demon-
strated as a Guerbet reaction[2] using excess alkaline metal
ductivity of higher alcohols using a hetereogeneous catalys-
t.[7a] We report herein a notable advancement in this field,
based on our demonstration that simple CuI[7]/H2/NaOH cat-
alysts are able to catalyze the cross-coupling between a
range of two different alcohols; the coupling showed a
wider substrate scope with the best practicality for the selec-
tive synthesis of higher alcohols (Scheme 1). Most impor-
alkoxides at
a
high pressure/temperature (ꢀ2208C).[2a]
Recent Guerbet reactions using transition-metal catalysts
were proposed to involve a “borrowing hydrogen”[1b] pro-
cess consisting of 1) transition metal-catalyzed oxidation
(transfer dehydrogenation) of both alcohols to form the cor-
responding carbonyl compounds; 2) base-catalyzed aldol
condensation of them to form a,b-unsaturated carbonyl
compounds, and 3) transition metal-catalyzed concomitant
reduction (transfer hydrogenation) of the C=C and C=O
bonds using the borrowed hydrogen atoms from both alco-
hols. Thus far, precious metals Ir[3,4] and Ru,[4,5] respectively,
have been the most widely used for this purpose and such
catalyst precursors have shown reasonable reactivity by
subtle modification of the original conditions. However, re-
action conditions used to date for Guerbet variants are so
diverse that the critical factors for promoting the reactions
have been veiled: in many cases, the reaction utilizes signifi-
cant amounts of base (1–3 equiv with respect to one of two
substrates),[3a,c,e,4a,b,5a,c–g,6b–d,h,7] and/or sacrificial hydrogen
(H2) acceptors,[3b,5a,f] with a few exceptions.[4c,5b,h,6a] In some
cases, the molar amount of one of the two different alcohols
used exceeded[3a,6b,7a] that of the other (e.g., 1/2=1:2[5a] or
2:1[6a] equiv). A H2 atmosphere was detrimental to the pro-
Scheme 1. General Scheme for coupling of two different alcohols.
tantly, the catalytic cycle that we found here was redox-free
and was different from those involving “borrowing hydro-
gen” that have been reported thus far for the modified
Guerbet reaction.
We first chose alcohol 1a as a model substrate (Table 1),
since a considerable amount of ketone 4aa is generated as a
side product when using a Ir[3a] or Ru[5a] catalyst system.[8]
Treatment of a secondary (28) alcohol 1a with a primary
(18) alcohol 2a in the presence of CuBr and NaOH (0.2/
4 mol%: [CuBr]0 =1.6ꢀ10ꢁ3 m) in p-xylene at 1358C under
H2 (1 atm) afforded, after the mixture was purified by
column chromatography on silica gel, the coupling product
alcohol 3aa in an yield of 92% (Table 1, entry 5).
No reaction was observed when the base NaOH was lack-
ing.[9] In the presence of dppe or dppp, the productivity de-
creased significantly, whereas moderate product alcohol/
ketone selectivity was retained (Table 1, entries 2 and 3). In
comparison, H2 (1 atm) was purged into the reaction solute
with an expectation that this treatment would eliminate the
formation of ketone 4aa.[10] In fact, conversion and product
alcohol selectivity were enhanced (Table 1, entries 4 and 5),
compared with the results obtained without H2 gas (en-
tries 1–3). By taking advantage of the H2 effects, the use of
CuBr (0.05 mol%) and NaOH (4 mol%) gave a turnover
number (TON) of>1300 at 1458C (Table 1, entry 6). The
[a] T. Miura, O. Kose, Dr. F. Li, Dr. S. Kai, Dr. S. Saito
Department of Chemistry, Graduate School of Science
& Research Center for Materials Science, Nagoya University
Chikusa, Nagoya 464-8602 (Japan)
Fax : (+81)52-789-5945
E-mail: saito.susumu@f.mbox.nagoya-u.ac.jp
[b] Dr. S. Saito
Institute for Advanced Research, Nagoya University
Chikusa, Nagoya 464-8601 (Japan)
[c] Dr. F. Li
Current Address: School of Chemical Engineering
The Nanjing University of Science and Technology (P.R. China)
[d] Dr. S. Kai
Current Address: School of Pharmacy, Jilin University (P.R. China)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101752.
11146
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11146 – 11151

COMMUNICATION
Table 1. Effects of different factors on the coupling of 1a with 2a.[a]
To elucidate the reaction mechanism of this efficient cata-
lyst system, several key experimental results should be
noted. 1) Those aryl substrates having a halogen, which re-
acted with Pd catalysts[6a] (e.g., 3bd:ꢀ0%), also underwent
cross-coupling under the present conditions. Even the aryl
iodide remained intact upon the reaction of 2e (Table 2,
entry 7). Hence, Cu0 is unlikely to play an essential role as
the active species.[13] 2) Neither O2 nor air was needed for
reoxidation of such low-valent Cu species, if in fact any
were present, even though apparent oxidation of alcohol
Entry
Ligand
H2
[atm]
Total yield[%][b]
(3aa/4aa)
TON[c]
N
1
2
3
4
None
dppp
dppe
dppp
None
None
–
–
–
1
1
1
74 (62:38)
50 (74:26)
30 (70:30)
88 (93:7)
>99 (92:8)
69 (97:3)
370
250
150
440
500
1380
was involved throughout the overall process. 3) Na/air[14a]
[14b]
5
and NaOH (or KOH)/O2
were recently reported to pro-
6[d]
mote the oxidation and the Guerbet reaction of benzylic-
type alcohols, respectively, without any transition metals. In
our case, catalytic Na or NaH, in place of NaOH under oth-
erwise identical reaction conditions (Table 2, entry 1),
showed a similar reactivity (e.g., 3ba: ꢀ95%) with rigorous
exclusion of air from the reaction mixture.
[a] CuBr/ligand/base/1a/2a=1:1ACTHNUTRGNE(UNG or 0):20:500:500. Unless otherwise
specified, the reaction was performed in p-xylene at 1358C for 48 h
under H2 or Ar. Diastereoselectivity of 3aa=ꢀ1:1.2. [b] Total yield of
3aa and 4aa. [c] Turnover number, calculated as (yield(%) of 3aa+
4aa)/(CuBr mol%). [d] CuBr/dppp/NaOH/1a/2a=1:0:80:2000:2000 at
1458C for 96 h.
Given these experimental findings, several control experi-
ments were carried out using 1b or 2a (Table 3). The role of
CuBr/NaOH/(H2 or dppp) was to start the oxidation of 1b,
but also more slowly, that of 2a (Table 3, entries 1–3).[15] No
more than 5% of 1b underwent oxidation even after 24 h.
CuBr2 showed a similar ineffectiveness (Table 3, entry 4).
These results are in contrast to CuI- or CuII-initiated oxida-
tion of alcohols in the presence of O2.[13c] Without CuBr, H2,
and air, NaOH alone barely promoted the oxidation of both
1b and 2a (Table 3, entry 5). Hence, a major pathway for
the oxidation of 2a finally started presumably through a
Meerwein/Ponndorf/Verley/Oppenauer (MPVO)-type hy-
drogen transfer from 2a to 5b.
Next, we intended to doubly check such a primary role of
CuBr, and to clarify the mechanism of the reaction steps fol-
lowing the induction period of the catalysis. The cross-cou-
pling between 18 and 28 alcohols 1b and 2a was conducted
using different reaction conditions, which were monitored in
the absence of CuBr and H2, and instead by adding likely
carbonyl compound(s) that might be generated in situ (5b,
6a, 5b+6a, or 7ba) (Table 4). We observed a similarity and
significant difference in the influence of each additive on
the yield of 3ba. Among the four different additives
(Table 4, entries 1–5), 5b+6a (0.5 or 1 mol% each) led to
an excellent yield of 3ba (ꢀ90%) with a marginal amount
of 4ba (ꢀ3%; Table 4, entries 3 and 4). It was speculated
that under basic conditions with a reaction temperature of
1358C, 5b+6a underwent rapid cross-aldol condensation
leading to enone 7ba. However, when 7ba (1 mol%) was
used separately as an additive, the reaction rate was slightly
inferior to that obtained using 5b+6a (Table 4, entry 4 vs.
5). In any event, we were able to avoid using CuBr and H2
by taking advantage of the notable influence of 5b+6a as
an initiator of the reaction.
solvent screening with CuBr (0.2 mol%)/NaOH (4 mol%)
([CuBr]0 =1.6ꢀ10ꢁ3 m) suggested that polar solvents, includ-
ing DMSO, DMF and 1,4-dioxane were not promising (3aa:
<3%).
Given the optimal conditions (1 (100 mol%),
2
(100 mol%), CuBr (0.2 mol%), H2 (1 atm), NaOH
(4 mol%), and p-xylene at 1358C), various substrates were
investigated and the results are presented in Table 2. Some
characteristic features of the CuBr/H2/NaOH catalyst
system are as follows. 1) The reaction showed substrate gen-
erality with respect to both 18 and 28 alcohol components.
Under the optimal conditions, benzylic-type (hetero)aromat-
ic 18 and 28 alcohols were the most promising substrates,
both of which were joined into the respective products 3 in
high to quantitative yields. 2) The ratio 3/4 (the correspond-
ing ketone) obtained using aliphatic alcohols was uniformly
excellent (>12:1ꢀ >99:1) under H2 (Table 2, entries 15–17
and 21–24).[11] The capability of a catalytic amount of base
as co-catalyst, thus far underestimated in the cross-coupling
of aliphatic alcohols, was enhanced significantly. These re-
sults are in contrast to preceding Guerbet variants utilizing
sacrificial H2 acceptors.[3b,5a,f] 3) The procedure using H2 was
also beneficial for obtaining b-branched 28 alcohols 3da and
3ea (Table 2, entries 19 and 20). The reactivity of acyclic 28
alcohols such as 1d and 1e having an elongated carbon
chain (R2 ¼H) was overlooked in preceding attempts to
modify the Guerbet conditions.[12] 4) In some cases, the addi-
tion of dppp without purging H2 was enough to ensure the
smoothest conversion (Table 2, entries 4, 9, and 12).
A molar concentration of the aliphatic aldehydes, generat-
ed upon oxidation of the corresponding alcohols 2k and 2l,
would be kept at a minimum, so that neither the aldehydes
nor their self-condensation was detected at least by TLC
analysis. The cross-aldol condensation between aldehydes
and ketones should proceed more rapidly. These points are
consistent with the unique facets of Guerbet conditions in-
cluding the present system.
Additional control reactions in the absence of CuBr were
carried out to probe a reducing agent for the prospective in-
termediate 7ba (Table 5). The 28 alcohol 1b showed reduc-
ing capability superior to the 18 alcohol 2a for the transfer
hydrogenation (TH) of a carbonyl group (Table 5, entry 1
Chem. Eur. J. 2011, 17, 11146 – 11151
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11147

S. Saito et al.
Table 2. Coupling between two different alcohols: substrate scope.[a]
of the olefin of allylic alcohol 8 also produced desir-
able 3ba (Table 5, entry 3); however, this process
may involve the transfer dehydrogenation (TDH:
Oppenauer oxidation) of 8, followed by 1,4-reduc-
tion of the resulting 7ba and subsequent TH of the
carbonyl group of 4ba (MPV reduction). In fact, an
equilibrium that shifts from 8 to the respective
ketone 7ba is possible through a self-induced TDH/
TH (MPVO) cycle, which indeed afforded 3ba and
4ba (entry 4). Here again, the apparent 1,4-reduc-
tion of 7ba prevails over 1,2-reduction (Table 5, en-
tries 3 and 4). Overall, in the reaction of 1b with 2a
(Table 2, entry 1), formation of 7ba should be fol-
lowed by a reaction sequence involving TH of the
a,b-olefin of 7ba (apparent 1,4-reduction)[16] and
subsequent TH of the carbonyl group of 4ba (1,2-
reduction), in which 1b and 2a can promote 1,4-
and 1,2-reduction.[17] Although a possibility of the
inverse sequence (first step: TH of the carbonyl
group of 7ba; second step: TH of the olefin)
cannot be fully excluded, recovery of 8 was unde-
Entry 28 Alcohol 18 Alcohol CuBr/dppp/NaOH/
Product
Yield
[%][b]
1
2
1/2
alcohol 3
1
2
3
4
5
6
7
8
9
1b
1b
1b
1b
1b
1b
1b
1b
1b
1b
2a
2a
2b
2b
2c
2d
2e
2 f
2 f
2g
1:0:20:500:500
1:0:40:2000:2000
1:0:20:500:500
1:1:20:500:500
1:0:20:500:500
1:0:100:500:500
1:0:100:500:500
1:0:20:500:500
1:1:20:500:500
1:0:20:500:500
3ba: Y=H
3ba
3bb: Y=F
3bb
3bc: Y=Cl
3bd: Y=Br
3be: Y=I
3bf: Y=OMe
3bf
95
98[c]
83
97[d]
97
98
98
91
96[d]
98
10
3bg: Y=SMe
11
12
1b
1b
2h
2h
1:0:20:500:500
1:1:20:500:500
3bh
3bh
71
87[d]
1
tectable by H NMR analysis in a crude mixture of
each reaction (Table 5, entries 1–4).
13
14
1b
1b
2i
2j
1:0:20:500:500
1:0:20:500:500
3bi: Z=O
3bj: Z=S
87
99
According to these control experiments, a likely
catalytic cycle viable for promoting a major path-
way can be depicted (Scheme 2). One of the most
important roles of CuBr/NaOH/(H2 or dppp) is to
initiate the oxidation of 28 alcohol 1 (and a very
small amount of 18 alcohol 2), giving a mixture of
ketone 5 and aldehyde 6.[18] Both reaction partners
readily form enolate IA through deprotonation of 5
with catalytically active species M-OR. Subsequent
cross-aldol condensation with 6 in IA leads to the
respective enone 7. Sequential hydrogen transfer
from 2 (or 1) to 7 and from 1 (or 2) to IB gives de-
sirable 3. Concomitant formation of IC, followed by
a rapid process involving intramolecular proton
transfer (deprotonation of 5) regenerates IA. The
sequence of multistep reactions corresponds to the
basic unit of overall catalytic cycles and can circu-
late iteratively even without a transition metal.
Thus a prospective species that catalyzes this self-
induced cycle (M-OR, Scheme 2) in a hydrocarbon
solvent may be a molecular cluster[19] consisting of
15
16
17
1b
1b
1b
2k
2k
2l
1:0:100:500:500
1:1:100:500:500
1:0:100:500:500
3bk: R3 =cHex
50[e]
80[e]
3bk
3bl: R3 =Me
N
18
1c
2a
1:0:100:500:500
3ca
99[e]
19
20
1d
1e
2a
2a
1:0:100:500:500
1:0:100:500:500
3da: R2 =Me
3ea: R2 =Ph
85[e,f]
82[e,f]
21
22
23
1 f
1 f
1 f
2a
2l
2l
1:0:100:500:500
1:0:100:500:500
1:1:100:500:500
3 fa: R3 =Ph
98[e]
86[e]
3 fl: R3 =Me
N
3 fl
24
1g
2a
1:0:100:500:500
3ga
68[e]
[a] Unless otherwise specified, CuBr/
scale was used and the ratios 3/4 (4 is the corresponding ketone, e.g., 4bl (ketone in-
stead of alcohol; R3 =Me
(CH2)10)) were consistently>93:<7. The reaction was per-
formed in p-xylene at 1358C for 48 h under H2 (1 atm). [b] Of those isolated purified
product alcohols 3 was determined by 1H NMR spectroscopy based on the internal
standard (1,1,2,2-tetrachloroethane). [c] At 1458C for 6 days. [d] Under Ar (1 atm).
[e] At 1458C for 48–144 h. [f] Diastereoselectivity was approximately 1:1–1.5.
ACHUTGTNRENNUG(dppp)/NaOH/1/2=0.02:ACHTUNGTERN(NUGN 0.02):0.4:10:10 mmol
(NaOR)mACHTUNGTRNEUNG
(CuOR)n (RO=alkoxide or enolate[20] de-
rivatives; m=an integer but not 0; n=0[21] or an in-
teger), which is thought to have the potential to
promote different sets of MPVO-type hydrogen
transfer.[6a,22]
AHCTUNGTRENNUNG
In summary, we have developed a highly efficient
and practical procedure for producing longer-chain
alcohols using a catalytic amount of base under
vs. 2). In contrast, the TH of the a,b-olefin of 7ba can take
place using either 1b or 2a (Table 5, entries 1 and 2). The
latter TH (1,4-reduction) showed an apparent reaction rate
faster than the former, judging from the product distribution
in favor of olefin reduction (Table 5; TH of carbonyl/TH of
olefin=1:1.8 (entry 1) and <1:12 (entry 2)). Apparent TH
anaerobic conditions. The carbon–carbon cross-coupling of
two different alcohols, which were mixed in a 1:1 molar
ratio, took place with high product alcohol selectivity using
H2. The substrate scope was broadened using inexpensive
metal sources CuBr and NaOH. We eventually found a tran-
11148
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11146 – 11151

CuI/H2/NaOH-Catalyzed Cross-Coupling
COMMUNICATION
Table 3. Control experiments: oxidation of 1b or 2a under different re-
action conditions.[a]
Entry Catalyst system Reaction 1b!5b
Reaction 2a!6a
ACHTUNGTRENNUNG
PhCOMe (5b) [%][b] PhCHO (6a) [%][b]
1
2
3
CuBr (0.2)
NaOH (4); Ar
CuBr (0.2)
NaOH (4); H2
CuBr (0.2)
NaOH (4)
5
5
4
ꢀ1
ꢀ1
ꢀ1
dppp (0.2); Ar
CuBr2 (0.2)
NaOH (4); Ar
NaOH (4); Ar
4
5
2
2
ꢀ1
ꢀ1
[a] Unless otherwise specified, the reaction was carried out using 1b or
2a in p-xylene at 1358C for 24 h using each catalyst system. [b] Yields de-
termined by NMR spectroscopy.
Table 4. Effects of additive(s) on reaction of 1b with 2a.[a]
Scheme 2. Plausible major catalytic cycle.
[10]
of H2 and the catalytically most important species partici-
Entry
Additive(s) [mol%]
3ba[%][b]
pating in this CuBr/H2/NaOH/alcohol/carbonyl compound(s)
system is now underway in our laboratory.
1
2
3
4
5b (1)
6a (1)
5b (1)+6a (1)
5b (0.5)+6a (0.5)
39
39
93 (3)[c]
82 (2)[c]
Experimental Section
General procedure: A degassed and subsequently argon-filled balloon
(1 L) was equipped with a Schlenk tube (100 mL). p-Xylene (10 mL), 1a
(10 mmol, 1482 mg), 2a (10 mmol, 1081 mg), CuBr (0.02 mmol, 2.87 mg),
and NaOH (0.4 mmol, 16.0 mg) was added to this vessel at 258C. A sus-
pension inside the Schlenk tube was subjected to the freeze (0.5–
1 mmHg, at ꢁ1968C)–thaw (filled with Ar, at 258C) cycles (ꢀ 3), filled
with H2 (1 atm, in the 1 L balloon) and was stirred at 1358C for 48 h,
during which time droplets of water were sticking around the neck and
the upper rim of the Schlenk tube. The mixture was cooled down to
258C and directly purified by column chromatography on silica gel
(EtOAc/n-hexane=from 1:100 to 1:30) to give a mixture of 3aa and 4aa
(2381 mg,>99% yield) in a ratio of 92:8 determined by internal standard
5
7ba (1)
72 (3)[c]
[a] Unless otherwise specified, the reaction was carried out using 1b
(10 mmol) and 2a (10 mmol) using NaOH (4 mol%) and additive(s) in
p-xylene at 1358C for 48 h under Ar; [NaOH]0 =ꢀ0.04m. [b] 1HNMR
yields. [c] NMR yields of 4ba.
Table 5. Control reactions using 7ba or 8.[a]
1
(1,1,2,2-tetrachloroethane) method using H NMR spectroscopy.
Entry
Starting
material
Additive
[mol%]
Yield[%][b]
N
(3ba/4ba/8/7ba)[c]
1
2
3
4
7ba
7ba
8
1b (500)
2a (500)
2a (100)
–
90 (49:41:<1:<1)
86 (<4:44:<1:38)
95 (32:63:<1:<1)
94 (13:81:<1:<1)
Acknowledgements
8
This work was partially supported by a Grant-in-Aid for Basic Research
(B) from JSPS and for Scientific Research on Priority Areas “Advanced
Molecular Transformations of Carbon Resource” from MEXT, and by
funds from Asahi Glass, Asahi Kasei Chemicals, Daiso, Kuraray, Mitsu-
bishi Chemical, Mitsui Chemicals, Nissan Chemical Ind., and Sumitomo
Chemical. The authors wish to thank GCOE in Chemistry, Nagoya Univ.
(NU) and Prof. R. Noyori (NU & RIKEN) for their fruitful discussions.
[a] Unless otherwise specified, the reaction was carried out using 7ba or
8 and additive with NaOH (4 mol%) in p-xylene at 1358C for 24 h under
Ar; [NaOH]0 =0.04m. [b] 1H NMR yields (3ba+4ba+8+7ba). [c] De-
1
termined by H NMR spectroscopy.
sition-metal-free[14b,23] version of this process in the presence
of not more than 1 mol% of carbonyl compound(s) under
conditions much milder than the original Guerbet condi-
tions. Thus this self-induced catalytic cycle, initiated by the
Keywords: alcohols · alkoxides · CꢁC coupling · cross-
coupling · copper · hydrogenation
CuBr/H2ACHTUNGTRENNUNG(or dppp)/NaOH, clearly demonstrated a mechanis-
tically distinct alternative to the “redox” or “borrowing hy-
drogen” mechanism in the previous Guerbet reaction that
has been believed over some decades to be facilitated by a
transition-metal catalyst.[24] A further elucidation of the role
[1] Recent reviews: a) G. Guillena, D. J. Ramꢂn, M. Yus, Angew. Chem.
2007, 119, 2410–2416; Angew. Chem. Int. Ed. 2007, 46, 2358–2364;
b) M. H. S. A. Hamid, P. A. Slatford, J. M. J. Williams, Adv. Synth.
Catal. 2007, 349, 1555–1575. Related reviews concerned about “bor-
Chem. Eur. J. 2011, 17, 11146 – 11151
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11149

S. Saito et al.
rowing hydrogen” or “hydrogen autotransfer”: c) T. D. Nixon, M. K.
Whittlesey, J. M. J. Williams, Dalton Trans. 2009, 753–762; d) G.
Guillena, D. J. Ramꢂn, M. Yus, Chem. Rev. 2010, 110, 1611–1641;
e) G. E. Dobereiner, R. H. Crabtree, Chem. Rev. 2010, 110, 681–
703.
[11] 28 alcohols of R2 ¼H are generally less reactive than 28 alcohols of
R2 =H. When we carried out separately the reaction of 3ba with 2a
at 135–1458C under otherwise identical reaction conditions, 95% of
3ba was recovered with ꢀ4% of 4ba, checked by 1H NMR spec-
troscopy. In fact, in situ benzylation of 3ba did not take place
(Table 2, entry 1) with prolonged heating at 1358C, as checked by
GC-MS.
[12] In the initial attempts, a higher loading of CuBr (2 mol%)/NaOH
(20–30 mol%) was assumed to ensure a smooth oxidation of aliphat-
ic 18 and 28 alcohols, since these substrates should be rather reluc-
tant to undergo oxidation, compared with benzylic alcohols. Howev-
er, the product alcohol selectivity was marginal: the reaction be-
tween 1b and 2l at 1358C and 1458C (CuBr/dppp/NaOH=
2:2:30 mol%) for 48 h under Ar gave alcohol 3bl and ketone 4bl in
ratios of 69:31 and 28:72 with total yields of 68 and 85%, respective-
ly.
[13] Dehydrogenation of 18 and 28 alcohols using Cu or CuO requires a
higher temperature (250–3008C): a) M. Hudlicky, Oxidations in Or-
ganic Chemistry, American Chemical Society, Washington DC, 1990,
pp. 114–149; CuI- or CuII-initiated oxidation of alcohols in the pres-
ence of O2 and/or TEMPO: b) R. A. Sheldon, I. W. C. E. Arends,
Adv. Synth. Catal. 2004, 346, 1051–1071; c) I. W. C. E. Arends, P.
Gamez, R. A. Sheldon, Adv. Inorg. Chem. 2006, 58, 235–279;
d) M. J. Schultz, M. S. Sigman, Tetrahedron 2006, 62, 8227–8241;
e) C. S. Cho, W. X. Ren, S. C. Shim, Tetrahedron Lett. 2006, 47,
6781–6785; f) C. S. Cho, W. X. Ren, N. S. Yoon, J. Mol. Catal. A:
Chem. 2009, 299, 117–120.
[2] a) M. C. R. Guerbet, Compt. Rend. Acad. Sci. 1909, 149, 129–132;
b) H. Machemer, Angew. Chem. 1952, 64, 213–220; c) E. F. Pratt,
D. G. Kubler, J. Am. Chem. Soc. 1954, 76, 52–56; d) G. Gregorio,
G. F. Pregaglia, R. Ugo, J. Organomet. Chem. 1972, 37, 385–387;
e) E. Klein, F. Thçmel, H. Winkler, Justus Liebigs Ann. Chem. 1973,
1004–1017; f) P. L. Burk, R. L. Pruett, K. S. Campo, J. Mol. Catal.
1985, 33, 1–14; g) P. L. Burk, R. L. Pruett, K. S. Campo, J. Mol.
Catal. 1985, 33, 15–21.
[3] Ir catalyst: a) K.-i. Fujita, C. Asai, T. Yamaguchi, F. Hanasaka, R.
Yamaguchi, Org. Lett. 2005, 7, 4017–4019; b) T. Matsu-ura, S. Saka-
guchi, Y. Obora, Y. Ishii, J. Org. Chem. 2006, 71, 8306–8308; c) A.
Pontes da Costa, M. Viciano, M. Sanaffl, S. Merino, J. Tejeda, E.
Peris, B. Royo, Organometallics 2008, 27, 1305–1309; ketone-alcohol
coupling: d) K. Taguchi, H. Nakagawa, T. Hirabayashi, S. Sakaguchi,
Y. Ishii, J. Am. Chem. Soc. 2004, 126, 72–73; e) A. P. da Costa, M.
Sanaffl, E. Peris, B. Royo, Dalton Trans. 2009, 6960–6966.
[4] Ir and Ru catalysts: a) D. Gnanamgari, C. H. Leung, N. D. Schley,
S. T. Hilton, R. H. Crabtree, Org. Biomol. Chem. 2008, 6, 4442–
4445; b) D. Gnanamgari, E. L. O. Sauer, N. D. Schley, C. Butler,
C. D. Incarvito, R. H. Crabtree, Organometallics 2009, 28, 321–325;
c) G. Onodera, Y. Nishibayashi, S. Uemura, Angew. Chem. 2006,
118, 3903–3906; Angew. Chem. Int. Ed. 2006, 45, 3819–3822.
[5] Ru catalyst: a) C. S. Cho, B. T. Kim, H.-S. Kim, T.-J. Kim, S. C.
Shim, Organometallics 2003, 22, 3608–3610; b) G. R. A. Adair,
J. M. J. Williams, Tetrahedron Lett. 2005, 46, 8233–8235; c) R. Martí-
nez, D. J. Ramꢂn, M. Yus, Tetrahedron 2006, 62, 8982–8987; d) M.
Viciano, M. Sanaffl, E. Peris, Organometallics 2007, 26, 6050–6054;
e) A. Prades, M. Viciano, M. Sanaffl, E. Peris, Organometallics 2008,
27, 4254–4259; ketone-alcohol coupling: f) C. S. Cho, B. T. Kim, T.-
J. Kim, S. C. Shim, Tetrahedron Lett. 2002, 43, 7987–7989; g) R.
Martínez, D. J. Ramꢂn, M. Yus, Tetrahedron 2006, 62, 8988–9001;
h) P. A. Slatford, M. K. Whittlesey, J. M. J. Williams, Tetrahedron
Lett. 2006, 47, 6787–6789.
[6] Pd catalyst: a) O. Kose, S. Saito, Org. Biomol. Chem. 2010, 8, 896–
900; b) C. Carlini, M. D. Girolamo, A. Macinai, M. Marchionna, M.
Noviello, A. M. R. Galletti, G. Sbrana, J. Mol. Catal. A: Chem. 2003,
204–205, 721–728; Pd or Ni for ketone–alcohol coupling: c) C. S.
Cho, J. Mol. Catal. A: Chem. 2005, 240, 55–60; d) M. S. Kwon, N.
Kim, S. H. Seo, I. S. Park, R. K. Cheedrala, J. Park, Angew. Chem.
2005, 117, 7073–7075; Angew. Chem. Int. Ed. 2005, 44, 6913–6915;
e) Y. M. A. Yamada, Y. Uozumi, Org. Lett. 2006, 8, 1375–1378;
f) Y. M. A. Yamada, Y. Uozumi, Tetrahedron 2007, 63, 8492–8498;
g) F. Alonso, P. Riente, M. Yus, Eur. J. Org. Chem. 2008, 4908–
4914; alcohol–alcohol coupling gave ketones using Ag: h) K. Shimi-
zu, R. Sato, A. Satsuma, Angew. Chem. 2009, 121, 4042–4046;
Angew. Chem. Int. Ed. 2009, 48, 3982–3986; Gold nanoparticle: i) S.
Kim, S. W. Bae, J. S. Lee, J. Park, Tetrahedron 2009, 65, 1461–1466.
[7] Heterogeneous Cu/Cr/Mn and Cu Raney catalysts with excess
NaOR: a) C. Carlini, M. D. Dirolamo, M. Marchionna, M. Noviello,
A. M. R. Galletti, G. Sbrana, J. Mol. Catal. A: Chem. 2002, 184,
273–280; b) C. Carlini, A. Macinai, A. M. R. Galletti, G. Sbrana, J.
Mol. Catal. A: Chem. 2004, 212, 65–70.
[14] a) L.-H. Zhou, X.-Q. Yu, L. Pu, Tetrahedron Lett. 2010, 51, 475–
477; aliphatic alcohols showed scant reactivity using O2: b) L. J.
Allen, R. H. Crabtree, Green Chem. 2010, 12, 1362–1364.
[15] Hence, a molar concentration of aldehyde should be consistently
much lower than that of ketone. This may be one of the reasons
why aliphatic aldehydes barely underwent their self-condensation.
[16] MPV-type 1,2-reduction of a,b-enone 7ba, followed by stepwise 1,3-
hydrogen shift (hydrogen transfer from the a- to the g-carbon of 8)
giving 4ba also could not be fully ruled out.
[17] Deuterium experiments also revealed that deuterium as Dꢁ shifted
from [D1]1b (PhCDOH(Me)) or from [D2]2a (PhCD2OH) to the g-
carbon (with respect to the OH group of 3ba: the b-carbon of a,b-
enone 7ba) and the a-carbon of 3ba (the carbonyl carbon of 7ba)
under otherwise identical conditions (Table 2, entry 1). In these re-
actions, the deuterium ratio (D%: thus hydrogen ratio (H%) is
100–D%) at the a- and g-carbon of 3ba was 40 and 20% using
[D1]1b and 2a, and 44% and 66% using 1b and [D2]2a. 1H NMR
indicated that the b-carbon of 3ba (originally, the a-carbon of 7ba)
was not labeled by deuterium in both cases, suggesting that the a-
carbon of 7ba is protonated by H+ from the OH group of alcohols.
This shows a typical nature of (apparent) MPV-type 1,4-reduction of
a,b-enones.
[18] The production of a metal hydride from a metal alkoxide indeed
confirms a proposed unique characteristic of copper: G. V. Goeden,
K. G. Caulton, J. Am. Chem. Soc. 1981, 103, 7354–7355.
[19] a) M. G. Stanton, M. R. GagnØ, J. Am. Chem. Soc. 1997, 119, 5075–
5076; b) M. G. Stanton, C. B. Allen, R. M. Kissling, A. L. Lincoln,
M. R. GagnØ, J. Am. Chem. Soc. 1998, 120, 5981–5989 and referen-
ces cited therein; c) R. M. Kissling, M. R. GagnØ, J. Org. Chem.
2001, 66, 9005–9010.
[8] For example, with Ir (Ref. [3a]) and Ru catalyst (Ref. [5a]) in the
presence of a stoichiometric or an excess amount of base (KOH or
[20] a) P. J. Pospisil, S. R. Wilson, E. N. Jacobsen, J. Am. Chem. Soc.
1992, 114, 7585–7587; b) J. M. Gruver, L. R. L. Liou, A. J. McNeil,
A. Ramirez, D. B. Collum, J. Org. Chem. 2008, 73, 7743–7747;
c) T. S. De Vries, A. Goswami, L. R. Liou, J. M. Gruver, E. Jayne,
D. B. Collum, J. Am. Chem. Soc. 2009, 131, 13142–13154 and refer-
ences cited therein.
Na
ACHTUNGTRENNUNG
[9] In the absence of NaOH (4 mol%), CuAHCUTNGTRENNUNG
unable to promote the coupling reaction. See also the Appendix of
Supporting Information (page S13–S14) for more detailed analysis
of reaction conditions.
[21] We always observed that black solids of Cu0, the amount of which
was yet unclear, precipitated out from the reaction mixture, sticking
on the wall of a Pyrex reaction vessel. The Inductively Coupled
Plasma Emission (ICPE) analysis of the solution phase of the reac-
tion as in entry 1, Table 2 indicated that the Cu content was less
than about 1/500 of the Cu mass initially used (the Cu mass changed
[10] The mechanism on how H2 improved the reaction rate and product
alcohol selectivity is still under scrutiny and needs profound investi-
gation. When D2 was used in place of H2 in the reaction of 1b with
2a (Table 2, entry 1), no appreciable incorporation of deuterium
1
into 3ba was observed using H NMR analysis.
11150
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11146 – 11151

CuI/H2/NaOH-Catalyzed Cross-Coupling
COMMUNICATION
from 116 to <0.62 ppm). However, we could not fully rule out a
possibility of the participation of some Cu(OR) species in catalysis.
[22] a) R. Martínez, D. J. Ramꢂn, M. Yus, J. Org. Chem. 2008, 73, 9778–
9780; b) H. V. Mierde, P. V. D. Voort, F. Verpoort, Tetrahedron Lett.
2008, 49, 6893–6895; c) H. V. Mierde, P. V. D. Voort, F. Verpoort,
Tetrahedron Lett. 2009, 50, 201–203.
ed that the rate of content of each heavy metal (Ag, Cr, Fe, Mo, Ni,
Zn) representative of any metal impurities was less than 1 ppm.
[24] We had already submitted our primary results related to the Cu/
(H2)NaOH system as a JP patent, dated on December 29th, 2009
(R. Noyori, S. Saito, H. Naka, O. Kose, T. Miuta (Nagoya Universi-
ty) 2009-299231, 2009) before submitting this paper.
[23] The ICPE analysis of NaOH pellet commercially available from Al-
drich (99.999% purity) and Kanto (99.98% purity) we used indicat-
Received: June 6, 2011
Published online: August 31, 2011
Chem. Eur. J. 2011, 17, 11146 – 11151
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11151