Iridium-catalyzed direct dehydroxylation of alcohols

Article abstract of DOI:10.1002/ejoc.201301293

Iridium-catalyzed direct dehydroxylation of alcohols with hydrazine was developed through a combination of the oxidation of alcohols and the Wolff-Kishner reduction. This protocol is simple to perform and highly efficient for a series of primary, benzylic and allylic alcohols. Iridium-catalyzed direct dehydroxylation of alcohols with hydrazine is developed through a combination of the oxidation of alcohols and Wolff-Kishner reduction. This protocol is simple to perform and highly efficient for a series of primary alcohols, especially benzylic and allylic ones. Copyright

Full text of DOI:10.1002/ejoc.201301293

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201301293
Iridium-Catalyzed Direct Dehydroxylation of Alcohols
[a]
[a]
[a]
Jian-Lin Huang, Xi-Jie Dai, and Chao-Jun Li*
Keywords: Synthetic methods / Dehydroxylation / Reduction / Iridium / Hydrazines / Alcohols
Iridium-catalyzed direct dehydroxylation of alcohols with
hydrazine was developed through a combination of the oxi-
dation of alcohols and the Wolff–Kishner reduction. This pro-
tocol is simple to perform and highly efficient for a series of
primary, benzylic and allylic alcohols.
Introduction
The dehydroxylation of alcohols is arguably one of the
most fundamental transformations in organic chemistry
and continuously plays an insurmountable role in the total
synthesis of complex natural products bearing multifunc-
[
1]
tional groups. In addition, there is a great demand for
more efficient dehydroxylation procedures in biomass con-
[
2]
versions. The best known method for the dehydroxylation
of alcohols was established by Barton and McCombie in
1975 through a radical synthetic procedure, which requires
multistep conversions and proceeds with overall low atom
[
3]
efficiency. Many alternatives, including benzoyl-ester-me-
[
4]
[5]
diated and the phosphite-mediated variants, have since
been developed. However, these strategies still require the
conversion of an alcohol into the preactivated intermediate,
which is subsequently transformed into the corresponding
C–H bond by a radical process [Scheme 1, Eq. (1)].
The ever-increasing desire for efficient chemical transfor-
mations in the synthetic community calls for more direct
dehydroxylation methods. Recently, two prominent
attempts to achieve this challenging goal were reported. In
2
010, Barrero and co-workers described the groundbreak-
ing dehydroxylation of benzylic and allylic alcohols utilizing
Nugent’s reagent (Cp TiCl) to homolyze the C–O bond
Scheme 1. Conventional radical dehydroxylation methods versus
efforts toward direct dehydroxylation. Cp = cyclopentadienyl, ppy
= 2-phenylpyridinato, LED = light-emitting diode.
2
[
Scheme 1, Eq. (2)].[6] Very recently, Stephenson and co-
workers developed visible-light-photoredox-catalyzed one-
pot dehydroxylation through an organic iodide intermedi-
ate, with a combination of the Garegg–Samuelsson reaction
and flow chemistry [Scheme 1, Eq. (3)].[ Although these
two procedures represent major ongoing efforts toward the
direct dehydroxylation of alcohols, they still suffer from cer-
tain limitations: the substrate scope of the former is restric-
ted to reactive candidates such as benzylic and allylic
alcohols, and the latter is operationally complex, requiring
a photoreactor with continuous flow. Thus, the develop-
ment of a simple and versatile dehydroxylation approach is
highly desirable.
7]
The main challenge in the development of a one-step de-
hydroxylation reaction is the poor leaving group of the hy-
droxy group, owing to its large CϪO bond dissociation en-
–
ergy, as well as the strongly basic hydroxide ion (OH ),
which is generated. To overcome these intrinsic problems,
most known tactics [Scheme 1, Eqs. (1) and (3)], are de-
signed toward obtaining reactive intermediates with supe-
rior leaving groups from alcohols. Given that ways to gener-
ate such intermediates are generally compatible with many
follow-up transformations, the direct dehydroxylation reac-
[
a] Department of Chemistry and FQRNT Center for Green
Chemistry and Catalysis, McGill University,
Montreal, Quebec, H3A 0B8, Canada
E-mail: cj.li@mcgill.ca
http://cjli.mcgill.ca/cj page.htm
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201301293.
6496
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 6496–6500

Iridium-Catalyzed Direct Dehydroxylation of Alcohols
tion can thus be achieved. For this matter, we planned to Table 1. Optimization of reaction conditions.[a]
take advantage of the better reactivity of carbonyl com-
pounds, which can be generated in situ by oxidation of
alcohols using transition-metal catalysts. Compatible with
this process is the Wolff–Kishner reduction, a classical text-
book reaction, representing a powerful synthetic strategy to
[
8]
Entry Catalyst
Base
Solvent[b]
T
[
Yield
deoxygenate carbonyl groups in organic compounds. We
envisaged that by employing hydrazine as a nucleophilic re-
ducing reagent in addition a suitable transition-metal cata-
lyst, the oxidation of alcohols might be combined with the
Wolff–Kishner reduction to enable the direct dehydroxyl-
ation of alcohols. Herein, we report a simple and highly
efficient method for the direct dehydroxylation of alcohols
through an oxidation/Wolff–Kishner sequence [Scheme 1,
Eq. (4)]. Remarkably, this reaction can even be carried out
in water, albeit with a decreased yield.
°C] [%][c]
1
2
3
4
5
6
7
8
9
10
[Cp*IrCI
2
]
2
NaHCO
NaHCO
Cs CO
Et
Et
Et
Et
Et
Et
3
2
H O
120
120
120
120
120
120
120
120
120
120
120
160
160
Ͻ5
8
(PCP)IrH(Cl)
(PCP)IrH(Cl)
(PCP)IrH(Cl)
3
H
2
O
O
O
O
O
O
O
2
3
H
H
2
15
29
32
31
9
15
29
38
59
70
3
N
N
N
N
N
N
2
(PPh
[(C
(PPh
(PPh
(PPh
3
)
H
2
IrCl(CO)
12)Ir(OMe)]
Ru(CO)H
Ru(CO)(Cl)H
IrCl(CO)
3
H
H
2
8
2
3
2
3
)
3
2
3
H
H
2
3
)
3
3
2
3
)
2
3
CHCl
3
(PPh ) IrCl(CO)
3
2
Et N
3
MeOH
MeOH
MeOH
MeOH
[
d]
1
12
1
(PPh
(PPh
(PPh
3
)
2
IrCl(CO)
IrCl(CO)
IrCl(CO)
KOH
KOH
KOH
[
d]
3
)
2
[
d,e]
[f]
1
3
3
)
2
99
96[
95
50
0
g]
Results and Discussion
[
d–f,h]
1
4
(PPh
(PPh
none
(PPh
(PPh
3
)
2
IrCl(CO)
KOH
KOH
KOH
KOH
KOH
MeOH
neat
MeOH
MeOH
MeOH
160
160
160
160
160
[d]
To test our hypothesis, the pilot study was investigated 15
between piperonyl alcohol (1a) and hydrazine hydrate (2,
3 2
) IrCl(CO)
[d–f]
1
1
1
6
[
d–f,i]
7
3
)
2
IrCl(CO)
Ru(CO)(Cl)H
0
92
2
equiv.) in H O in the presence of [Cp*IrCI ] (1.0 mmol-
2 2 2
[d–f]
8
3
)
3
%
, Cp* = C H ) and NaHCO (1 equiv.) (Table 1, see the
5 5 3
[
9]
[a] Reaction conditions: 1a (15 mg, 0.1 mmol),
2
(10 mg,
Supporting Information for full optimization data). To
our delight, a trace amount of corresponding dehydroxyl-
0
.2 mmol), catalyst (1 mol-%), base (0.1 mmol), under an air atmo-
sphere. [b] Solvent (0.1 mL) was used (c = 1 m), unless otherwise
1
ated product 3a was detected after 12 h reaction at 120 °C noted. [c] Determined by H NMR spectroscopy by using nitro-
(
Table 1, entry 1). A slightly higher yield was obtained upon methane as an internal standard. [d] 0.2 mmol KOH was used.
[e] 10 μL MeOH was used (c = 10 m). [f] Reaction time: 3 h. [g] Re-
using the iridium pincer complex (PCP)IrH(Cl) {PCP =
action time: 2 h. [h] Under an atmosphere of Ar. [i] Without 2.
C H -2,6-(CH PtBu ) } as the catalyst (Table 1, entry 2).
6
3
2
2 2
Subsequently, a variety of bases were examined and Et N
3
was found to be the best (Table 1, entries 3 and 4; Support- this yield was unaffected by an argon atmosphere (Table 1,
ing Information, Table S1, entries 1–4). Screening of transi- entry 14). The product was obtained in only 50% yield un-
tion-metal catalysts revealed that iridium complexes dis- der solvent-free conditions even after 10 h (Table 1, en-
played higher catalytic activity than ruthenium complexes try 15). Furthermore, control experiments demonstrated
(
Table 1, entries 5–8; Table S1, entries 5–8), whereas neither that both the iridium catalyst and hydrazine hydrate were
rhodium nor iron complexes showed any catalytic activity indispensable for this dehydroxylation, as no reaction oc-
Table S1, entries 9 and 10). Among all the iridium catalysts curred in the absence of either reagent (Table 1, entries 16
tested, Vaska’s complex [(PPh ) IrCl(CO)] gave the best re- and 17). It is noteworthy that the use of (PPh ) Ru(CO)-
(
3
2
3 3
sult, and desired product 3a was detected in 32% yield (Cl)H also afforded the product, albeit in a slightly lower
(
Table 1, entry 5). Encouraged by these results, we next in- yield relative to that obtained with (PPh ) IrCl(CO)
3 2
vestigated the influence of solvents, with MeOH providing (Table 1, entry 18).
the highest yield (Table 1, entries 9 and 10; Table S1, en-
Under the optimized conditions, the substrate scope was
tries 11–22). A further increase in yield was obtained by explored using 2 (2 equiv.) as the reagent, (PPh ) -
3
2
replacing Et N (1 equiv.) with KOH (2 equiv.) as the base IrCl(CO) (1 mol-%) as the catalyst, KOH (2 equiv.) as the
3
(Table 1, entry 11); this was attempted in view of KOH be- base in MeOH (c = 10 m) at 160 °C under an air atmo-
ing commonly used in the Wolff–Kishner reduction. While sphere for 3 h (Table 2). In general, benzylic alcohols
optimizing the conditions, we discovered that this reaction showed excellent reactivity, including those with both elec-
was insensitive to air. Moreover, all of the signals in the tron-donating and electron-withdrawing substituents on the
1
H NMR spectrum of the crude reaction mixture could be benzene rings (Table 2, entries 1–12); nearly quantitative
ascribed to either 1a or 3a, which indicated that it was pos- yields were obtained for electron-rich benzylic alcohols in
sible to entirely consume 1a. On the basis of the above ob- almost all cases (Table 2, entries 1–3 and 7). Interestingly,
servations, our attention was turned to the reaction tem- the yield decreased significantly (99 to 58%) upon moving
perature. A higher reaction temperature was found to be the methoxy group from the para position to the ortho posi-
beneficial (59 and 70% at 120 and 160 °C, respectively; tion, whereas the meta isomer remained unaffected (Table 2,
Table 1, entry 11 vs. entry 12). Notably, an increased reac- entries 3–5). The decrease in the yield of the product ob-
tion concentration resulted in a shorter reaction time and tained through the use of the ortho isomer can possibly be
nearly quantitative yield of the product (Table 1, entry 13); attributed to the ready chelation of the ortho-methoxy and
Eur. J. Org. Chem. 2013, 6496–6500
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6497

J.-L. Huang, X.-J. Dai, C.-J. Li
SHORT COMMUNICATION
hydroxy groups to the iridium catalyst; this chelation occu- try 19), whereas the C=C bond remains intact in the pre-
III
[6]
pies the empty coordination site, which is required for acti- vious Ti -promoted strategy. However, only a moderate
vation of the β C–H bond. Moreover, this chelation effect yield was observed for secondary cyclic alcohol 1t (Table 2,
was not observed with the more sterically bulky benzoxy entry 20), possibly as a result of increased steric hindrance,
substituent (Table 2, entry 6). Next, primary alcohols other which offers potential regioselective dehydroxylation of mo-
than benzylic alcohols were tested and most of them gave lecules bearing multiple hydroxy groups.
satisfactory results (Table 2, entries 13–19). To our delight,
This new (PPh ) IrCl(CO)/NH NH ·H O dehydroxyl-
3 2 2 2 2
the reaction scope was extended to heteroaromatic, hetero- ation protocol exhibits many appealing practical features:
cyclic, and aliphatic alcohols and diols, as the direct dehy- (1) All reactions take place either in air or under inert gas
droxylation of these candidates has not been previously re- atmosphere. (2) Reactions can be simply performed by add-
ported (Table 2, entries 14–18).[
6,7]
Significantly, both the ing all chemicals into the sealed Ace pressure glass tube and
hydroxy group and the C=C bond can be efficiently reduced then heated. (3) The workup procedure is very straightfor-
by employing this dehydroxylation strategy (Table 2, en- ward and only requires filtration through a silica plug.
Table 2. Scope of the Ir-catalyzed direct dehydroxylation.[a]
[
1
a] Reaction conditions: 1 (0.3 mmol), 2 (30 mg, 0.6 mmol), (PPh
3
)
2
IrCl(CO) (2.3 mg, 1 mol-%), KOH (34 mg, 0.6 mmol), MeOH (30 μL),
60 °C, 3 h, under an air atmosphere. [b] Determined by H NMR spectroscopy by using nitromethane as an internal standard; yields of
the isolated products are given in parentheses. [c] MeOH (90 μL) was used (c = 3 m). [d] Reaction conditions: 1 (0.3 mmol), 2 (60 mg,
.2 mmol), (PPh IrCl(CO) (1 mol-%), KOH (68 mg, 1.2 mmol), MeOH (30 μL), 160 °C, 3 h, under an air atmosphere. [e] A trace amount
of mono-dehydroxylated product was detected. [f] Reaction time: 12 h.
1
1
3 2
)
6498
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 6496–6500

Iridium-Catalyzed Direct Dehydroxylation of Alcohols
To explore the mechanism of this dehydroxylation reac- following the Wolff–Kishner reduction. After ligand disso-
tion, some control experiments were conducted under the ciation of the carbonyl compound from C, iridium hydride
optimized conditions (Scheme 2). Tertiary 2-phenyl-2-prop- complex D is protonated by H O to regenerate the active
2
anol (4) did not react with hydrazine hydrate (2) due to its species A, with concomitant release of hydrogen gas. Inter-
lack of a β-H (Scheme 2, a). In addition, an intermolecular estingly, the presence of hydrogen acceptors in this catalytic
reaction between piperonyl alcohol (1a) and styrene (6) pro- system results in simultaneous reduction of their double
vided corresponding dehydroxylated product 3a and ethyl- bonds through an insertion and protonation sequence
benzene (7), both in quantitative yield. This implies that (Table 2, entry 9 and Scheme 2, b). However, the role of
hydrogen gas generated in situ can participate in the hydro- iridium complexes other than Vaska’s complex in the cata-
genation of the double bond catalyzed by Vaska’s complex lytic cycle is still unclear at this stage and needs further
(Scheme 2, b). Anticipating two key intermediates (the alde- investigation.
hyde and the hydrazone) in the overall dehydroxylation pro-
cess, we also carried out an NMR spectroscopy experiment
in deuterated benzene (Scheme 2, c). As expected, a small
1
amount of piperonyl aldehyde (8) was detected by H NMR
spectroscopy upon treatment of piperonyl alcohol (1a) with
(
PPh ) IrCl(CO) (1 equiv.) and KOH (1 equiv.) at 65 °C for
3 2
1
h. However, 8 was not observed in the absence of base
even if the reaction mixture was heated. The subsequent
addition of 2 (1 equiv.) to that reaction mixture led to the
formation of a trace amount of piperonyl hydrazone (9).
1
Moreover, a quartet was observed in H NMR spectrum
around 10 ppm (JP,H = 21.6 Hz), which can be assigned to
the Ir–H bond (see the Supporting Information for more
details).
Scheme 3. Tentative mechanism for the Ir-catalyzed direct de-
hydroxylation of alcohols.
Conclusions
In summary, we have reported herein an unprecedented
iridium-catalyzed direct alcohol dehydroxylation reaction to
form C–H bonds in a simple and efficient fashion. This
reaction is proposed to proceed through an oxidation/
Wolff–Kishner reduction sequence. The present approach
highlights a useful alternative to the classical multistep de-
hydroxylation strategy of alcohols, especially for benzylic
and allylic primary alcohols. Notably, even water can be
used as a solvent for this reaction. Further efforts to expand
the reaction scope, clarify the reaction mechanism, and ex-
plore the synthetic applications of this reaction are cur-
rently in progress in our laboratory.
Scheme 2. Investigation of the mechanism for the dehydroxylation.
On the basis of the obtained experimental results as well
as literature studies on the “borrowing hydrogen” strat-
egy[ and the Wolff–Kishner reduction, a tentative mecha-
nism for this iridium-catalyzed direct dehydroxylation of
alcohols is proposed in Scheme 3. Given that the activation
of the alcohol does not occur without involvement of a
base, we postulate an initial equilibrium between Vaska’s
complex and hydroxy-substituted derivative A in the pres-
ence of KOH. Then, alcohol 1 replaces the hydroxy group
in complex A to generate complex B, which is subsequently
converted into complex C through β-H elimination. The
10]
Experimental Section
General Procedure: An Ace pressure glass tube (4 cm) was charged
with (PPh
0
0
3
)
2
IrCl(CO) (2.3 mg, 0.003 mmol), KOH (34 mg,
.6 mmol), the alcohol (0.3 mmol), NH NH ·H (29 μL,
.6 mmol), and MeOH (30 μL) under an air atmosphere. The Ace
2
2
2
O
tube was sealed, and the reaction mixture was stirred at 160 °C for
h. After that, the reaction mixture was cooled to room tempera-
ture and filtered through a short column made of a silica plug by
flushing it with CH Cl (10 mL). The filtrate was dried with
3
2
2
carbonyl derivative coordinating to C dissociates from the Na SO and concentrated under reduced pressure to give a residue,
2
4
1
iridium metal center to give rise to corresponding alkane 3 which was first subjected to analysis by H NMR spectroscopy by
Eur. J. Org. Chem. 2013, 6496–6500
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6499

J.-L. Huang, X.-J. Dai, C.-J. Li
SHORT COMMUNICATION
using nitromethane (5.4 μL, 0.1 mmol) as an internal standard and
then further purified by preparative TLC or flash chromatography
on silica gel to afford the desired product.
McLaughlin, L. L. Adduci, J. J. Becker, M. R. Gagné, J. Am.
Chem. Soc. 2013, 135, 1225–1227.
3] a) D. H. R. Barton, S. W. McCombie, J. Chem. Soc. Perkin
Trans. 1 1975, 1574–1585; b) D. Crich, L. Quintero, Chem. Rev.
[
1
989, 89, 1413–1432; c) W. Hartwig, Tetrahedron 1983, 39,
Supporting Information (see footnote on the first page of this arti-
1
2609–2645.
cle): Full optimization data, synthesis, and copies of the H NMR
[
4] a) I. Saito, H. Ikehira, R. Kasatani, M. Watanabe, T. Matsu-
ura, J. Am. Chem. Soc. 1986, 108, 3115–3117; b) K. Lam, I. E.
Marko, Org. Lett. 2008, 10, 2773–2776; c) K. Lam, I. E.
Marko, Chem. Commun. 2009, 95–97; d) K. Lam, I. E. Marko,
Synlett 2012, 1235–1239; e) K. Lam, I. E. Marko, Org. Lett.
and 13C NMR spectra of all isolable compounds.
Acknowledgments
2011, 13, 406–409.
[
[
[
5] a) L. Zhang, M. Koreeda, J. Am. Chem. Soc. 2004, 126, 13190–
The authors are grateful to the Canada Research Chair Foundation
1
2
3191; b) P. A. Jordan, S. J. Miller, Angew. Chem. 2012, 124,
961–2965; Angew. Chem. Int. Ed. 2012, 51, 2907–2911.
(
grant to C. J. L.), Canada Foundation for Innovation (CFI),
Fonds Québécois de la Recherche sur la Nature et les Technologies
FQRNT) Center for Green Chemistry and Catalysis, FQRNT
6] H. R. Dieguez, A. Lopez, V. Domingo, J. F. Arteaga, J. A. Do-
bado, M. M. Herrador, J. F. Quilez del Moral, A. F. Barrero, J.
Am. Chem. Soc. 2010, 132, 254–259.
7] J. D. Nguyen, B. Reiss, C. Dai, C. R. Stephenson, Chem. Com-
mun. 2013, 49, 4352–4354.
8] D. Todd, Org. React. 1947, 4, 378–422.
9] a) K.-I. Fujita, K. Yamamoto, R. Yamaguchi, Org. Lett. 2002,
4, 2691–2694; b) O. Saidi, A. J. Blacker, M. M. Farah, S. P.
Marsden, J. M. J. Williams, Chem. Commun. 2010, 46, 1541–
1543; c) O. Saidi, A. J. Blacker, G. W. Lamb, S. P. Marsden,
J. E. Taylor, J. M. J. Williams, Org. Process Res. Dev. 2010, 14,
1046–1049.
(
Equipe (grants to D. G. and C. J. L.), Natural Sciences and Engi-
neering Research Council of Canada (NSERC), and McGill Uni-
versity for support of the research.
[
[
[
1] a) S. W. McCombie, Comprehensive Organic Synthesis (Eds.:
B. M. Trost, I. Fleming), Oxford, 1991, vol. 8, p. 811–833; b)
R. C. Larock, Comprehensive Organic Transformations, 2nd ed.,
Wiley, New York, 1999, p. 49–52; for examples in a complex
natural molecule setting, see: c) K. C. Nicolaou, T. V. Koftis,
S. Vyskocil, G. Petrovic, W. Tang, M. O. Frederick, D. Y. K. [10] a) M. H. S. A. Hamid, P. A. Slatford, J. M. J. Williams, Adv.
Chen, Y. Li, T. Ling, Y. M. A. Yamada, J. Am. Chem. Soc.
006, 128, 2859–2872; d) D. S. Palacios, T. M. Anderson, M. D.
Synth. Catal. 2007, 349, 1555–1575; b) T. D. Nixon, M. K.
Whittlesey, J. M. J. Williams, Dalton Trans. 2009, 753–762; c)
G. E. Dobereiner, R. H. Crabtree, Chem. Rev. 2010, 110, 681–
703; d) A. J. Watson, J. M. Williams, Science 2010, 329, 635–
636; e) M. G. Edwards, R. F. Jazzar, B. M. Paine, D. J. Shermer,
M. K. Whittlesey, J. M. Williams, D. D. Edney, Chem. Com-
mun. 2004, 90–91.
2
Burke, J. Am. Chem. Soc. 2007, 129, 13804–13805; e) G. E.
Veitch, E. Beckmann, B. J. Burke, A. Boyer, S. L. Maslen, S. V.
Ley, Angew. Chem. 2007, 119, 7773–7776; Angew. Chem. Int.
Ed. 2007, 46, 7629–7632.
[
2] a) J. ten Dam, U. Hanefeld, ChemSusChem 2011, 4, 1017–1034;
b) M. Shiramizu, F. D. Toste, Angew. Chem. 2012, 124, 8206–
Received: August 27, 2013
8210; Angew. Chem. Int. Ed. 2012, 51, 8082–8086; c) M. P.
Published Online: September 17, 2013
6500
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 6496–6500