Key words cross-coupling - alcohols - ketones - ruthenium catalysis - homogeneous catalysis
- green synthesis
1
Introduction
Ketones are important compounds that are widely used as solvents in the chemical industry
and also serve as basic building blocks in syntheses of many natural products.[1 ] In particular, α- and β-substituted ketones are structural motifs present in many
natural products and bioactive compounds. These α- and β-substituted ketones are usually
synthesized by the reaction of an alkyl halide with a ketone in the presence of a
strong base such as BuLi or LDA.[2 ] Such conventional alkylation methods suffer from various disadvantages, such as
the use of toxic alkyl halides, the need for cryogenic conditions, or the production
of stoichiometric amounts of metal salts as chemical waste, thereby decreasing the
atom economy of the alkylation reactions. Because of the versatile utility of ketones,
the development of new methods for their synthesis has received much attention. As
a result, atom-economical, practical, environmentally friendly, and efficient catalytic
methods have been devised to produce α- and β-substituted ketones. Sustainable development
from renewable raw materials is one of the contemporary challenges for organic synthesis.[3 ] Alcohols are good choices as alkylating partners for alkylation reactions because
they are abundant in nature, biorenewable, cheap, and easy to handle and store.[4 ]
Scheme 1 Borrowing-hydrogen methodology in cross-coupling of secondary alcohols
Subramanian Thiyagarajan (left) was born in Karur, Tamilnadu, India. He received his B.Sc. in 2012 and his
M.Sc. in 2014 from the St. Joseph’s College (Bharathidasan University), Tiruchirappalli.
In 2015 he joined the research group of Dr. C. Gunanathan as a junior research fellow
for his doctoral studies in the School of Chemical Sciences, National Institute of
Science Education and Research (NISER), Bhubaneswar, India. His current researches
focus on syntheses and green catalytic transformations in the presence of ruthenium
pincer complexes. Chidambaram Gunanathan (right) was born in Kothamangalam, Tamilnadu, India. He completed his B.Sc. in chemistry
at RKM Vivekananda College (1997), and his M.Sc. in organic chemistry at the Department
of Organic Chemistry, University of Madras, Chennai (1999). He obtained his Ph.D.
in chemistry from the Central Salt and Marine Chemicals Research Institute (CSMCRI,
2005), where he worked in the group of Professor S. Muthusamy. For a postdoctoral
stint, he joined the groups of Professors David Milstein and Hadassa Degani at the
Weizmann Institute of Science, Israel, where he was also a Dean of Faculty Postdoctoral
Fellow. He then spent two years as an Alexander von Humboldt Research Fellow in the
group of Professor Walter Leitner at RWTH Aachen University, Germany. In 2011, he
joined NISER, Bhubaneswar, as an assistant professor . He is a recipient of a Ramanujan
Fellowship from DST-SERB, New Delhi. He was promoted to reader in 2013 and to associate
professor in 2017. His research interests include the catalytic applications of pincer
complexes for sustainable development and the development of atom-economical hydroelementation
processes.
In recent years, the borrowing-hydrogen methodology (or hydrogen autotransfer process)
has gained importance for the construction of C–C and C–N bonds by using alcohols
(Scheme [1 ]).[4 ]
[5 ] Notably, H2 O is the only byproduct from this method. An alternative approach for the formation
of new C–C and C–N bonds through acceptorless dehydrogenation of alcohols has also
been developed.[6 ] Remarkably, these processes are green and sustainable as they do not use hydrogen
acceptors or oxidants, and they liberate hydrogen and water as the only byproducts.
Catalytic Self- or Cross–Coupling of Alcohols and Selectivity Challenges
2
Catalytic Self- or Cross–Coupling of Alcohols and Selectivity Challenges
By using borrowing-hydrogen or acceptorless-dehydrogenation concepts, environmentally
benign synthetic methodologies have been developed. Catalytic self-couplings and cross-couplings
of primary alcohols to provide esters have been reported.[7 ]
[8 ] Catalytic dimerization of primary alcohols to give β-alkylated alcohols have also
been reported in the literature.[9 ] Self-couplings of secondary alcohols to provide ketones have been developed (Scheme
[2a ]).[10 ] Recently, cross-couplings of secondary alcohols with primary alcohols to provide
substituted ketones or alcohols have also been reported (Scheme [2b ]).[8 ]
[11 ]
Scheme 3 Ruthenium-catalyzed direct cross-coupling of secondary alcohols and selectivity challenges.
(Reprinted with permission from reference 12. © 2019, American Chemical Society)
Scheme 2 Catalytic self-couplings and cross-couplings of alcohols. (Reprinted with permission
from reference 12. © 2019, American Chemical Society.)
Unlike other coupling reactions of alcohols, the catalytic cross-coupling of secondary
alcohols was unknown until our recent report (Scheme [3a ]).[12 ] Cross-coupling of secondary alcohols is a reaction in which two different secondary
alcohols react together with the help of a metal catalyst to form a new covalent bond.
In general, the major challenge in cross-coupling of secondary alcohols is to minimize
the self-coupling of ketones generated in situ from both secondary alcohols. In addition,
under basic conditions ketones can provide undesired aldol side reactions (Scheme
[3b ]). Controlling the C–C coupling of secondary alcohols to give β-disubstituted ketones
exclusively is highly challenging. Thus, catalytic methods have been developed to
synthesize β-disubstituted ketones.
Recent Developments in the Synthesis of β-Disubstituted Ketones
3
Recent Developments in the Synthesis of β-Disubstituted Ketones
A few methods have recently been reported for the synthesis of β-disubstituted ketones
(Scheme [4 ]). In 2015, Sun and co-workers reported a peroxide-promoted decarboxylative alkylation
of cinnamic acids with alkenes to provide β-disubstituted ketones.[13 ] More recently, an iron-catalyzed oxyalkylation of alkynes with peroxides to provide
various β-disubstituted ketones was demonstrated.[14 ] Reaction of chloro ketones with a Grignard reagents in the presence of zinc chloride
and copper acetate gave ketones in good yields.[15 ] Moreover, a iridium-catalyzed photoredox oxyalkylation of styrene derivatives and
a decarboxylative alkylation of silyl enol ethers have been developed to provide a
variety of β-disubstituted ketones.[16 ] These reported methods all suffer from various serious drawbacks, such as the use
of excess amounts of acids or peroxides as additives that produce copious amounts
of waste; the use of Grignard reagents and chloro ketones, which produce hazardous
organometallic waste; or the use of use of prefunctionalized N -(acyloxy)phthalimides as alkylating partners. Moreover, all these reactions necessitate
the use of expensive chemicals, excess amounts of oxidants or metallic halides, or
prefunctionalized starting materials, which are undesired as they do not provide environmentally
benign processes. Thus, it was important to develop a practical catalytic protocol
that does not require stoichiometric additives or prefunctionalization and that can
proceed without producing hazardous chemical waste.
Scheme 4 Recent reports in synthesis of β-disubstituted ketones
Catalytic dehydrogenative coupling of alcohols has been recognized as an alternative
protocol for the synthesis of substituted ketones. Recently, Donohoe and co-workers
reported an alkylation of ketones by using secondary alcohols to provide β-disubstituted
ketones through a borrowing-hydrogen methodology.[17 ] However, this alkylation protocol requires excess amounts of base, and its substrate
scope is limited to bulky ketones [e.g., 1-(pentamethylphenyl)ethanone]. The synthesis
of β-disubstituted ketones directly from two different secondary alcohols requires
four elementary steps: dehydrogenation, aldol condensation, hydrogenation of an α, β-unsaturated ketone, and oxidation of the resulting secondary alcohol functionality.
Inspired by recent green catalytic transformations that emanated from our group,[18 ] we have developed a direct synthesis of β-disubstituted ketones from two different
secondary alcohols.[12 ] Notably, all the fundamental transformations occur in a single-step and the catalytic
cross-coupling reaction does not require stoichiometric bases or oxidants; instead,
it requires only a catalyst and a catalytic amount of base. Water and liberated H2 are the only byproducts.
Scope of Ruthenium-Catalyzed Cross-Coupling of Secondary Alcohols
4
Scope of Ruthenium-Catalyzed Cross-Coupling of Secondary Alcohols
To verify our hypothesis, we screened various reaction conditions for the ruthenium-catalyzed
cross-coupling of secondary alcohols. Inspired by our recent studies, we chose Ru-MACHO
1 (1 mol%) as the catalyst with 1-phenylethanol (0.5 mmol) and cyclohexanol (0. 5 mmol)
as reactants in the presence of t -BuOK (2 mol%) in toluene solution at 135 °C as our model conditions; these provided
the complete conversion of both alcohols, and the expected cross-coupled product was
isolated in 69% yield (Table [1 ], entry 1). Traces of the self-coupled product and unreacted acetophenone were also
obtained. A high catalytic reactivity was observed when the base loading was increased
to 5 mol% and the temperature was reduced to 125 °C, which provided the product in
86% yield (entry 3). Further increases in the base loading, decreases in the catalyst
loading, lower temperatures, or replacement of t -BuOK with t -BuONa resulted in considerably lower yields (entries 4–8). Control experiments with
only the base and without the catalyst 1 or base confirmed that no product formation occurs without the catalyst (entries
9 and 10).
Table 1 Optimization of the Reaction Conditionsa
Entry
Catalyst (mol%)
Base (mol%)
Temp (°C)
Conv.b (%)
Yieldc (%)
1
1
2
135
>99
69 (74)
2
1
5
135
>99
85 (90)
3
1
5
125
>99
86 (91)
4
1
10
125
>99
70 (76)
5d
1
5
125
>99
87 (94)
6
1
5
115
94
63 (69)
7
0.5
5
125
>99
70 (72)
8e
1
5
125
>99
79 (83)
9f
–
5
125
5
–
10f
–
–
125
–
–
a Reaction conditions: 1-phenylethanol (0.5 mmol), cyclohexanol (0.5 mmol), toluene
(1.5 ml), catalyst 1 (1 mol%), t -BuOK (5 mol%), 125 °C, under flowing argon.
b Conversion of 1-phenylethanol as determined by GC analysis with benzene as internal
standard.
c Yields of isolated products after column chromatography; yields calculated by GC
analysis of the reaction mixtures are given in parentheses.
d Cyclohexanol (2 equiv) was used.
e 5 mol% of t -BuONa was used.
f The reaction was performed for 24 h.
With the optimized reaction condition in hand, we examined the scope of the secondary
alcohols in cross-coupling with cyclohexanol (Scheme [5 ]). 1-Phenylethanols containing electron-donating substitution on the aryl ring afforded
the corresponding cross-coupled products 2a–i in yields of 60–90% (Scheme [5 ]). Interestingly, aryl secondary alcohols having electron-withdrawing groups were
well tolerated: the reaction of 1-(4-chlorophenyl)ethanol with cyclohexanol provided
product 2j in 48% yield. Heterocyclic secondary alcohols reacted with cyclohexanol to give the
cross-coupled products 2k and 2l in good yields. Bicyclic aromatic secondary alcohols gave products 2m –o in very good yields.
Scheme 5 Scope of ruthenium-catalyzed cross-coupling of secondary alcohols with cyclohexanol
We next investigated whether the same strategy could be applied to other cyclic or
acyclic secondary alcohols. Surprisingly, both cyclic and acyclic secondary alcohols
directly coupled with benzylic secondary alcohols to give good to excellent yields
of the cross-coupled ketone products (Scheme [6 ]). Under the optimized reaction conditions, substituted cyclohexanols gave the corresponding
cross-coupled products as mixtures of diastereomers whose diastereomeric ratios were
determined by 1 H NMR analysis of the crude reaction mixtures. The reaction of various 1-arylethanol
derivatives with 4-methylcyclohexanol gave the cross-coupled ketone products 3a –c (dr = 80:20, 81:19, and 78:22, respectively) in good yields as mixture of diastereomers.
Similar diastereoselectivities were observed for 4-propyl-, 4-tert -butyl-, and 4-phenylcyclohexanol, which gave products 3d –g in good yields. 1-Cycloheptanol and 2-norborneol reacted with 1-phenylethanol derivatives
to give the corresponding ketone products 3h –k in good yields (Scheme [6 ]). 1,4-Dioxaspiro[4.5]decan-8-ol and diphenylmethanol reacted with 1-arylethanol
derivatives to give products 3l and 3m in moderate yields. Next, highly challenging and nonactivated linear-chain aliphatic
secondary alcohols were employed in this selective cross-coupling of secondary alcohols
in the presence of an increased catalyst loading of 1 (4 mol%) and of base (20 mol%). Ultimately, a variety of secondary alcohols such
as propan-2-ol, butan-2-ol, pentan-2-ol, pentan-3-ol, and heptan-4-ol were well tolerated
and were selectively converted into the corresponding β-disubstituted ketones 3n –t in good to excellent yields.
Mechanistic Studies and Proposed Mechanism
5
Mechanistic Studies and Proposed Mechanism
To gain mechanistic insights, the cross-coupling of 1-phenylethanol with cyclohexanol
was monitored under the optimized reaction conditions. GC analysis of the reaction
mixture at regular intervals indicated that the reaction follows first-order kinetics
with respect to the consumption of 1-phenylethanol. Interestingly, when 1-mesitylethanol
reacted with sterically hindered adamantan-2-ol, the olefin product 5a formed selectively and was isolated in 87% yield (Scheme [7a ]). Analysis of the crude reaction mixture by GC clearly indicated the absence of
alkylated product. Moreover, the reaction of 1-mesitylethanol with heptan-4-ol under
the optimized conditions provided alkylated product 3t and olefin product 5b in a 90:10 ratio (Scheme [7b ]). These results indicate that the borrowing-hydrogen pathway was interrupted due
to steric encumbrance of the ruthenium on catalyst 1 and that the reaction proceeded via α,β-unsaturated ketone intermediates. In addition,
deuterium-labeling experiments conducted by using deuterated alcohols suggested that
the liberated dideutrium/dihydrogen from cyclohexanol was predominantly reinstalled
on an α,β-unsaturated intermediate formed in situ, rather than 1-phenylethanol (Schemes
7c and 7d).
On the basis of our previous reports[18 ] and experimental evidence, a catalytic cycle for the cross-coupling of secondary
alcohols catalyzed by the Ru-MACHO catalyst 1 is proposed (Scheme [8 ]). Previously, facile O–H, O–D, N–H, and C(sp)–H bond-activation reactions by catalyst
1 have been established.[12 ]
[18 ] Catalyst 1 reacts with base to generate a coordinatively unsaturated reactive intermediate I , previously observed in mass spectrometric analyses.[18d,19 ] The reactive intermediate I reacts with both the secondary alcohols to provide the alkoxy-ligand-coordinated
ruthenium intermediates II and II′ upon O–H activation, as previously established by us.[18f ] The amide donor present in the unsaturated intermediate I accepts a proton upon activation of the O–H bond and become the amine donor in intermediates
II and II′ . In concert with the metal center, the ligand motif participates in bond formation
and bond breaking, and hence displays metal–ligand cooperation. Further, β-hydride
elimination from intermediates II and II′ might result in the formation of ketone intermediates A and B , and both dehydrogenation reactions converge to provide the same ruthenium dihydride
complex III . Although there is no evidence for it, the involvement of other mechanistic pathways
cannot be ruled out.[20 ] Furthermore, in the presence of base, a cross-aldol condensation reaction between
ketones A and B , formed in situ, generates the α,β-unsaturated carbonyl compound C . Further, selective hydrogenation by complex III provided the desired β-disubstituted ketone. The amine–amide metal–ligand cooperation
that is operative in these catalytic intermediates allows the regeneration of catalytic
active intermediate I upon hydrogenation, as well as the liberation of a H2 molecule from the ruthenium dihydride III .
Scheme 6 Scope of the ruthenium-catalyzed cross-coupling of secondary alcohols
Scheme 7 Mechanistic investigations. (Reprinted with permission from reference 12. © 2019,
American Chemical Society.)
Scheme 8 Proposed catalytic cycle for the cross-coupling of secondary alcohols. (Reprinted
with permission from reference 12. © 2019, American Chemical Society.)
6
Conclusion
We have demonstrated an unprecedented ruthenium-catalyzed selective and highly efficient
cross-coupling of secondary alcohols to provide β-disubstituted ketones. The catalytic
system exhibits a high activity and selectivity with a broad substrate scope. The
reaction proceeded through acceptorless dehydrogenation of secondary alcohols to provide
the corresponding ketones. Furthermore, under basic conditions, an aldol condensation
followed by selective hydrogenation of an α,β-unsaturated ketone intermediate, formed
in situ, provides β-disubstituted ketones.
Detailed kinetic and deuterium-labeling studies suggested that aliphatic secondary
alcohols oxidize faster than do benzylic secondary alcohols, which facilitates the
selective formation of cross-coupling products. Notably, hydrogen and water are the
only byproducts liberated by this green catalytic protocol. This new cross-coupling
reaction protocol will be an important development in the synthesis of β-disubstituted
ketones, and should be applicable to laboratory and industrial-scale chemical syntheses.
