Rethinking the Claisen–Tishchenko Reaction

Article abstract of DOI:10.1002/anie.201611186

Pincer-type complexes [OsH₂(CO){PyCH₂NHCH₂CH₂NHPtBu₂}] and [OsH₂(CO){HN(CH₂CH₂PiPr₂)₂}] catalyze the disproportionation reaction of aldehydes via an outer-sphere bifunctional mechanism achieving turnover frequencies up to 14 000 h^?1. The N?H group of the catalysts is a key player in this process, elucidated with the help of DFT calculations.

Full text of DOI:10.1002/anie.201611186

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201611186
German Edition: DOI: 10.1002/ange.201611186
Pincer Ligands
Rethinking the Claisen–Tishchenko Reaction
Stacey A. Morris and Dmitry G. Gusev*
Abstract: Pincer-type complexes [OsH2(CO){PyCH2NH-
CH2CH2NHPtBu2}] and [OsH2(CO){HN(CH2CH2PiPr2)2}]
catalyze the disproportionation reaction of aldehydes via an
transfer step of the Meerwein–Pondorf–Verley (MPV) reac-
tion.[5]
The use of d-block metals for the ester synthesis of
Scheme 1 is relatively uncommon.[6] [RuH2(PPh3)4] was the
first complex discovered to catalyze the disproportionation of
aldehydes by Yamamoto in 1982.[6a] Other prominent d-metal
systems include the Shvo catalyst,[6d] [OsH6(PiPr3)2],[6e] and
[RhH2(NCMe)(PhB(CH2PPh2)3)].[6k] With the exception of
the Shvo catalyst, the d-metal compounds tested afforded
TON = 100–200 and thus little incentive for replacing the
main-group catalysts. This is somewhat surprising, consider-
ing how dramatically the mechanistically related MPV
reaction was transformed by the development of the bifunc-
tional ruthenium catalysts.[7]
outer-sphere bifunctional mechanism achieving turnover fre-
À1
À
quencies up to 14000 h . The N H group of the catalysts is
a key player in this process, elucidated with the help of DFT
calculations.
T
he Claisen–Tishchenko reaction (Scheme 1) is an atom-
efficient method of producing esters from aldehydes.[1] The
aldehyde disproportionation was first reported in 1887 by
Claisen who obtained benzyl benzoate by treating benzalde-
hyde with sodium ethoxide in ethanol.[2] Subsequently,
Tishchenko developed the general catalytic approach of
Scheme 1 using aluminum and magnesium alkoxides.[3]
Modern catalysts achieve up to 200 turnovers,[4] although
turnover numbers (TONs) under 100 and the use of up to
5 mol% catalyst are more common.[1]
Different reaction mechanisms have been considered with
the d-metals. The prevalent thought is that vacant coordina-
tion sites on the metal and substrate coordination are
À
important. Aldehyde insertion into a M H bond to give
a metal alkoxide is another key mechanistic event.
Yamamoto originally proposed that the coordinated
aldehyde undergoes b-H elimination to give a metal acyl
hydride intermediate, then a second molecule of the aldehyde
À
inserts into the Ru H bond to form an alkoxide. Ultimately,
Scheme 1. The Claisen–Tishchenko reaction.
À
the C O coupling of the acyl with the alkoxide affords the
ester.[6a] The mechanism of Tejel[6k] (Scheme 3) resembles the
The Claisen–Tishchenko reaction has been comprehen-
sively reviewed.[1] The generally accepted mechanism of the
aldehyde disproportionation is summarized in Scheme 2. The
À
key steps of the C O bond formation and the hydride transfer
occur in the coordination sphere of the metal. Step II of the
Claisen–Tishchenko reaction is analogous to the hydride
Scheme 3. A mechanism of the d-metal-catalyzed ester synthesis.
classical mechanism (Scheme 2), except that the final hydride
transfer is to the metal. In yet another mechanism, Shvo
proposed that the aldehyde is first hydrogenated to produce
the primary alcohol and a 16-electron metal intermediate.[6d]
The alcohol further reacts with the substrate in solution to
give the hemiacetal. This hemiacetal is dehydrogenated by the
16-electron catalyst intermediate to afford the final product.
Both the hydrogenation and dehydrogenation events with the
Shvo catalyst were envisaged to occur in an outer-sphere
fashion.
Scheme 2. Key intermediates of the Claisen–Tishchenko reaction
(M=metal from Main Groups I–III).
The experimental and computational work presented here
[*] S. A. Morris, Prof. Dr. D. G. Gusev
Department of Chemistry and Biochemistry
Wilfrid Laurier University
À
establish that the C O bond formation of the Claisen–
Tishchenko reaction can be an outer-sphere process. Thus, the
mechanism does not require a vacant coordination site, and
the disproportionation of aldehydes is most efficiently
promoted by thermally robust 18-electron bifunctional cata-
lysts affording turnover frequencies on the order of 104 hÀ1.
These findings provide insights into the unusual role of the
75 University Ave. W., Waterloo ON N2L 3C5 (Canada)
E-mail: dgoussev@wlu.ca
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under
http://dx.doi.org/10.1002/anie.201611186.
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
conversion of benzaldehyde in 7 h, and S/C = 1000 gave 98%
conversion of trans-cinnamaldehyde in 3 h. This behavior is
indicative of some catalyst deactivation that can be due to
a reaction with the substrate or due to a reaction with a trace
impurity.
The results of Scheme 4 present several salient points.
Both 1 and 2 are outstanding catalysts (TOF up to 14000 hÀ1),
one to two orders of magnitude more efficient than the known
catalysts developed in the past 100 years. Only the Shvo
catalyst, tested with heating at 658C for several hours,
afforded comparable initial turnover frequencies reaching
5000 hÀ1. We note that 1, 2, and the Shvo complex are
bifunctional carbonyl hydrides; the significance of this
observation should become clear when we discuss the
reaction mechanism. Catalysts 1 and 2 well tolerate the
sterically more hindered and extended substrates.
A special feature of 1 is the ability to catalyze the Claisen–
Tishchenko reaction of a,b-unsaturated aldehydes, which are
very challenging substrates with the existing catalysts. The
near quantitative formation of E6 and E7 from cinnamalde-
hyde and 2-hexenal, respectively (Scheme 4), is unprece-
=
dented. Generally, the C C bonds of the unsaturated
Scheme 4. Disproportionation products and turnover numbers of the
tested aldehydes (3m for 10-undecenal and 5m for the other sub-
strates in toluene) with 1 and 2 (for details see the Supporting
Information, Table S1, and Figures S1–S7).
substrates seem to be unaffected by 1 and are retained in
the products (E3 and E4, Scheme 4). Complex 2 proved to be
inactive toward E6 and E7.
Complex 1 and especially 2 are thermally robust species.
We propose that the reactions leading to the esters of
Scheme 4 involve 18-electron intermediates and outer-
sphere steps utilizing the bifunctional nature of 1 and 2. The
elucidation of the mechanism was guided by DFT (M06L-D3)
calculations, using the disproportionation of acetaldehyde to
ethyl acetate as the model reaction.
The proposed catalytic cycle (Scheme 5) starts by an
outer-sphere hydride transfer from 1 or 2 to the substrate,
followed by a rearrangement of the intermediate alkoxide to
give Int 2. This transformation is downhill with acetaldehyde:
DG = À8.8 and À6.4 kcalmolÀ1 with 1 and 2, respectively.
Int 2 is the ground-state species of the reaction mixture. A
second molecule of the aldehyde reacts with Int 2 in an outer-
À
N H group in the reactions promoted by the bifunctional
catalysts.
We tested complexes 1[8a] and 2[8b] (Scheme 4) with
representative substrates: enolizable (butyraldehyde, iso-
butyraldehyde, 10-undecenal, 3-cyclohexene-1-carboxalde-
hyde) and non-enolizable aldehydes (benzaldehyde, trans-
cinnamaldehyde, and trans-2-hexenal). The catalytic solutions
in toluene were stirred at room temperature for 10 min, then
the first NMR spectra of the reaction solutions were collected
1
in 5–7 minutes. The H and 13C NMR spectra of the product
esters E1–E7 are provided with the Supporting Information
(Figures S1–S7).
À
sphere fashion; the C O bond is fully formed in Int 3. Step III
The Claisen–Tishchenko reaction generates a considera-
ble amount of heat. The reactions of Scheme 4 toward E1–E5
were fast, with the temperature reaching 60–758C on the
surface, registered by an IR thermometer. The corresponding
turnovers of Scheme 4 are at 100% conversion, with the
reaction times being ꢀ 20 min. Formation of E6 and E7 was
less rapid and required 1–2 h to give > 99% conversion and
the turnovers of Scheme 4. The non-enolizable substrates
afforded the product esters E5, E6, and E7 selectively, while
the formation of E1–E4 was accompanied by a trace byprod-
uct tentatively identified as a primary alcohol (see Figures S1–
S7 for details). Using 2 produced slightly more of this
byproduct whose origin is uncertain.
The turnovers of Scheme 4 are the limiting numbers, and
increasing the S/C ratio resulted in product mixtures with the
unreacted substrates. Larger turnover numbers could be
achieved for E3, E5, and E6 with 1 by allowing longer
reaction times. For example, using S/C = 7000 gave 98%
conversion of 10-undecenal in 6 h, S/C = 5000 gave 88%
can be viewed as a nucleophilic addition of the alkoxide onto
=
À
the carbonyl, facilitated by the C O···H N hydrogen bond-
ing. A similar mechanistic event is part of the dehydrogen-
ative homocoupling of alcohols, while the reverse of step III
occurs in catalytic ester hydrogenation.[9]
For the product ester of Scheme 5 to form, Int 3 should
rearrange to TS3. It is likely that this involves Int 4, which is
the thermodynamic “sink” in step V. The relevant 1-ethoxy-
ethanolate complex was documented by ESI-MS in ethanol
dehydrogenation to ethyl acetate catalyzed by 1, and the
collision-induced dissociation experiments demonstrated for-
mation of ethyl acetate from this species in the gas phase.[10]
The calculations suggest that step II of Scheme 5 is
unexpectedly complex and involves formation of the alcohol
RCH2OH in the coordination sphere of the metal (Scheme 6).
À
From Int 1, formation of the O H bond in Int 1a is downhill.
The alcohol of Int 1a rearranges easily to give the 18-electron
À
complex Int 1b. Finally, the O H proton returns to the
nitrogen to give the alkoxide Int 2. All of the steps of
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
give the corresponding ethoxides is exoergic by DG = À4.0
and À12.9 kcalmolÀ1, respectively.[12] Thus, the catalytic
reaction of Scheme 5 proceeds with net retention of the
À
N H bond. Dub and Gordon recently proposed a revised
mechanism for the Noyori hydrogenation, suggesting reten-
[13]
À
tion of the N H bond of the catalyst in that reaction as well.
The use of a non-protic solvent is crucial for the ester
synthesis of Scheme 5. For example, the outcome of the
reaction changes when toluene is replaced by methanol.
When a 2m solution of butyraldehyde in MeOH was treated
with 1 or 2 (S/C = 900), the ensuing rapid reaction afforded
a mixture of butanol and methyl butyrate (Figure S8). This
observation is easily rationalized by formation of the meth-
oxide intermediate in step II of Scheme 7 and the outer-
sphere coupling of the methoxide with butyraldehyde. The
overall reaction is that of Scheme 8a. It is interesting whether
catalysts and/or reaction conditions can be developed to
circumvent the insertion step I of Scheme 7 by allowing rapid
H2 elimination from the hydride intermediate, to produce the
methyl ester selectively per reaction of Scheme 8b.
Scheme 5. d-Metal-catalyzed ester synthesis. The origin of the M06L-
D3 free energies for the formation of ethyl acetate (underlined values,
kcalmolÀ1, in toluene at 298.15 K) is OsH(OEt)(CO)[PyCH2NHCH2-
CH2NHPtBu2] (int 2) with CH3CHO. Mass balance is ensured through-
out.
Scheme 6. A path connecting Int 1 and Int 2. The M06L-D3 free
energies (kcalmolÀ1) are for the reactions of 1 and 2 in toluene
(R=Me, M=Os).
Scheme 6 leading to TS1c proceed on a relatively flat,
practically barrier-less free energy surface. We note that
Scheme 6 also explains formation of Ru-alkoxides as the
kinetic products observed by Bergens in the reactions of
ketones with the Noyori catalyst, trans-[Ru((R)-BINAP)-
(H)2((R,R)-dpen)].[11] When considering the reverse process
Scheme 7. d-Metal-catalyzed methyl ester synthesis in methanol.
À
of Scheme 6, it is clear that the N H group should be
instrumental in repositioning the alkoxide from Int 2 to Int 1
in the dehydrogenation reactions of alcohols catalyzed by
1.[8a,9d,10]
Scheme 8. Methyl ester formation in methanol.
16-Electron intermediates have become a mechanistic
staple in the catalytic cycles constructed based on the ideas of
metal–ligand cooperation (MLC). Considering Int 1a, it is
intriguing to see no free alcohol RCH2OH released in step II
along with formation of the 16-electron amido complexes
[OsH(CO){PyCH2NCH2CH2NHPtBu2}] (3, from 1), or [OsH-
(CO){N(CH2CH2PiPr2)2}] (4, from 2). Our DFT calculations
suggest that the alcohol elimination is unfavorable versus
Int 2 in toluene. For example, addition of EtOH to 3 and 4 to
One can reasonably expect that a reaction with water,
analogous to Scheme 7, should afford an osmium hydroxide
and ultimately a stable carboxylate complex[14] that can be
detrimental to the ester synthesis. The presence of water is
inhibitory to the classical Tishchenko reaction catalyzed by
aluminum alkoxides.[1d] Another practical point related to the
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
Chem. Int. Ed. 2011, 50, 11047 – 11049; Angew. Chem. 2011, 123,
11241 – 11243; d) A. M. P. Koskinen, A. O. Kataja, Org. React.
2015, 86, 105 – 409.
catalyst stability is the well-known tendency of aldehydes to
decarbonylation with d-metal complexes.[6e] This unwanted
side-reaction can be minimized by employing 4 or 5d-metal
carbonyls, such as 1, 2, and the Shvo catalyst, where the
formation of di- or polycarbonyls is relatively unfavorable.
In conclusion, this paper broadens the scope of the
[2] L. Claisen, Ber. Dtsch. Chem. Ges. 1887, 20, 646 – 650.
[3] a) V. E. Tishchenko, J. Russ. Phys. Chem. Soc. 1906, 38, 355 –
418; b) V. E. Tishchenko, J. Russ. Phys. Chem. Soc. 1906, 38,
482 – 540; For electronic access, see http://books.e-heritage.ru/
book/10088469 and http://books.e-heritage.ru/book/10088471,
respectively.
[4] For example: Z. Guo, S. Wang, X. Wei, J. Organomet. Chem.
2016, 818, 115 – 122.
[5] R. Cohen, C. R. Graves, S.-B. T. Nguyen, J. M. L. Martin, M. A.
Ratner, J. Am. Chem. Soc. 2004, 126, 14796 – 14803.
À
catalytic bifunctional outer-sphere reactions. Clearly, M X
À
=
nucleophiles other than X H can be usefully coupled with
À
the carbonyls (Scheme 9). The N H group of the bifunctional
À
catalyst is a key player in this process. The N H is directing
[6] a) T. Ito, H. Horino, Y. Koshiro, A. Yamamoto, Bull. Chem. Soc.
Jpn. 1982, 55, 504 – 512; b) S.-I. Murahashi, T. Naota, K. Ito, Y.
Maeda, H. Taki, J. Org. Chem. 1987, 52, 4319 – 4327; c) K. A.
Bernard, J. D. Atwood, Organometallics 1988, 7, 235 – 236; d) N.
Menashe, Y. Shvo, Organometallics 1991, 10, 3885 – 3891; e) P.
Barrio, M. A. Esteruelas, E. OÇate, Organometallics 2004, 23,
1340 – 1348; f) T. Suzuki, T. Yamada, T. Matsuo, K. Watanabe, T.
Katoh, Synlett 2005, 1450 – 1452; g) S. Omura, T. Fukuyama, Y.
Murakami, H. Okamoto, I. Ryu, Chem. Commun. 2009, 6741 –
6743; h) M.-O. Simon, S. Darses, Adv. Synth. Catal. 2010, 352,
305 – 308; i) S. Ogoshi, Y. Hoshimoto, M. Ohashi, Chem.
Commun. 2010, 46, 3354 – 3356; j) Y. Hoshimoto, M. Ohashi, S.
Ogoshi, J. Am. Chem. Soc. 2011, 133, 4668 – 4671; k) C. Tejel,
M. A. Ciriano, V. Passarelli, Chem. Eur. J. 2011, 17, 91 – 95.
[7] a) R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed. 2001, 40, 40 –
73; Angew. Chem. 2001, 113, 40 – 75; b) D. Wang, D. Astruc,
Chem. Rev. 2015, 115, 6621 – 6686.
Scheme 9. Outer-sphere bifunctional reactions of carbonyls.
and activating the carbonyl substrate toward the nucleophile
XÀ through the C O···H N hydrogen bonds. More intrigu-
=
À
À
ingly, the N H can “shuttle” an alkoxide in and out of the
metal coordination sphere, thus allowing facile heterolytic
À
C H activation in systems where b-H elimination is not
feasible. The latter removes the mechanistic need for opening
a coordination site through elimination of an ancillary ligand
and allows working with thermally stable, electronically and
coordinatively saturated d-metal catalysts.
[8] a) D. Spasyuk, C. Vicent, D. G. Gusev, J. Am. Chem. Soc. 2015,
137, 3743 – 3746; b) M. Bertoli, A. Choualeb, A. J. Lough, B.
Moore, D. Spasyuk, D. G. Gusev, Organometallics 2011, 30,
3479 – 3482.
The metal complexes tested in this work, particularly
complex 1, by a large margin outperform all of the known
catalysts for the Claisen–Tishchenko reaction developed in
the last 100 years. The selective and efficient ester formation
with 1 and 2 makes the homocoupling of aldehydes a useful
contribution to the library of “click” reactions.[15] The
chemistry of Scheme 7 is promising for the development of
a catalytic approach toward coupling aldehydes with alcohols
for the synthesis of the corresponding mixed esters.
[9] For details, see refs. [9a–d] and references therein: a) F. Hasa-
nayn, A. Baroudi, A. A. Bengali, A. S. Goldman, Organometal-
lics 2013, 32, 6969 – 6985; b) F. Hasanayn, A. Baroudi, Organo-
metallics 2013, 32, 2493 – 2496; c) F. Hasanayn, H. Harb, Inorg.
Chem. 2014, 53, 8334 – 8349; d) D. G. Gusev, ACS Catal. 2016, 6,
6967 – 6981.
[10] C. Vicent, D. G. Gusev, ACS Catal. 2016, 6, 3301 – 3309.
[11] a) P. A. Dub, T. Ikariya, J. Am. Chem. Soc. 2013, 135, 2604 –
2619; b) R. J. Hamilton, S. H. Bergens, J. Am. Chem. Soc. 2008,
130, 11979 – 11987; c) S. Takebayashi, N. Dabral, M. Miskolzie,
S. H. Bergens, J. Am. Chem. Soc. 2011, 133, 9666 – 9669; see also
d) W. Baratta, G. Chelucci, S. Gladiali, K. Siega, M. Toniutti, M.
Zanette, E. Zangrando, P. Rigo, Angew. Chem. Int. Ed. 2005, 44,
6214 – 6219; Angew. Chem. 2005, 117, 6370 – 6375.
Acknowledgements
[12] These energies should change when the alcohol is hydrogen-
bonded in solution. For example, the enthalpy of hydrogen-
bonded EtOH in toluene is lowered by ca. À6.5 kcalmolÀ1 (see
the Supporting Information for details) compared to the
enthalpy of a single EtOH molecule in toluene.
We are grateful to NSERC Canada, CFI LOF program,
SHARCNET, and Wilfrid Laurier University for support.
Conflict of interest
[13] a) P. A. Dub, N. J. Henson, R. L. Martin, J. C. Gordon, J. Am.
Chem. Soc. 2014, 136, 3505 – 3521; b) P. A. Dub, J. C. Gordon,
Dalton Trans. 2016, 45, 6756 – 6781.
The authors declare no conflict of interest.
[14] a) M. L. Buil, J. J. F. Cardo, M. A. Esteruelas, E. OÇate, Organo-
metallics 2016, 35, 2171 – 2173; b) F. Hasanayn, L. M. Al-Assi,
R. N. Moussawi, B. S. Omar, Inorg. Chem. 2016, 55, 7886 – 7902.
[15] a) The term “click chemistry” applies to reliable, high-yielding,
and regio-selective methods of joining building blocks to
produce new compounds through a modular approach; b) X.
Hou, C. Ke, J. F. Stoddart, Chem. Soc. Rev. 2016, 45, 3766 – 3780.
Keywords: aldehyde disproportionation · bifunctional catalysts ·
ester synthesis · metal–ligand cooperation · pincer complexes
[1] a) O. P. Tçrmꢀkangas, A. M. P. Koskinen, Recent Res. Dev. Org.
Chem. 2001, 5, 225 – 255; b) T. Seki, T. Nakajo, M. Onaka, Chem.
Lett. 2006, 35, 824 – 829; c) W. I. Dzik, L. J. Gooßen, Angew.
Manuscript received: November 15, 2016
Revised: December 13, 2016
Final Article published: && &&, &&&&
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Pincer Ligands
S. A. Morris, D. G. Gusev*
&&& — &&&
Rethinking the Claisen–Tishchenko
Reaction
Think again: Pincer-type complexes 1 and
2 catalyze the homocoupling of aldehydes
via an outer-sphere bifunctional mecha-
nism achieving turnover frequency of up
to 14000 hÀ1.
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!