Iron catalysed selective reduction of esters to alcohols

Article abstract of DOI:10.1039/c8ob02661k

The reaction of (dppBIAN)FeCl₂ with 3 equivalents of n-BuLi affords a catalytically active anionic Fe complex; the nature of the anionic complex was probed using EPR and IR experiments and is proposed to involve a dearomatized, radical, ligand scaffold. This complex is an active catalyst for the hydrosilylation of esters to afford alcohols; loadings as low as 1 mol% were employed.

Full text of DOI:10.1039/c8ob02661k

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
Iron catalysed selective reduction of esters to
alcohols†
Cite this: DOI: 10.1039/c8ob02661k
Sem Raj Tamang, Anthony F. Cozzolino
and Michael Findlater
*
Received 26th October 2018,
Accepted 18th December 2018
DOI: 10.1039/c8ob02661k
rsc.li/obc
The reaction of (dppBIAN)FeCl2 with 3 equivalents of n-BuLi in catalysis have shown that reactions that have competing
affords a catalytically active anionic Fe complex; the nature of the single electron transfer process can now readily and easily cat-
anionic complex was probed using EPR and IR experiments and is alyze reactions that typically undergo 2e− redox process.7
proposed to involve a dearomatized, radical, ligand scaffold. This
The reduction of carbonyls and their derivatives into alco-
complex is an active catalyst for the hydrosilylation of esters to hols is an important transformation in organic synthesis and
afford alcohols; loadings as low as 1 mol% were employed.
has found widespread use in the chemical industry.8 While, the
reduction of aldehydes and ketones to their respective alcohols
have been broadly studied and can be achieved under facile and
mild conditions, the reduction of esters to alcohols with a
broad substrate scope and employing non-precious metals
Introduction
remains
a challenge. Typical methods employed in the
Transformation of organic molecules into compounds of
increased value, as synthetic intermediates or as final products,
is often achieved by employing transition metals as catalysts.1
In recent years the ability of main group elements,2 and organo-
catalyts3 to effect similar catalytic transformations has also
been explored. Although there are many important aspects of
catalyst design, recent advances in the rational application of
metal–ligand cooperativity have proven highly successful. For
example, as a metal center typically undergoes oxidation state
changes during catalysis, redox-active ligands assume an
increasingly significant role in stabilizing the metal at variety of
oxidation states to prevent the degradation of the catalyst.4
Transition metal catalysis now encompasses a huge range
of metals and ligand-types which have been successfully
applied in catalysis.5,4b One such class is the ‘redox non-inno-
cent’ ligand; whereby the ligands themselves participate in
mediating the catalysis via action as either electron donor or
acceptor in concert or independently from the metal. This non-
innocence is useful as oxidation states, which are not common
or unconventional, are avoided. Redox non-innocent ligands
derived from diimine backbones and complexation with tran-
sition metals have been extensively studied and are well docu-
mented in the literature.6 The applications of these complexes
reduction of esters include hydrogenation,9 use of strong redu-
cing agents (e.g., LiAlH4, NaBH4)10 or hydrofunctionalization
(i.e., hydroboration,11 and hydrosilylation12). From the perspec-
tive of atom-economy, hydrogenation is the most attractive
option, but reaction conditions typically employ high pressure
and temperature. Metal-catalyzed hydrofunctionalization pro-
vides an alternative, and synthetically viable, method to effec-
tively and efficiently reduce esters into alcohols. There are
several reports of metal catalyzed hydrosilylation of esters13 but
opportunities exist to improve upon recent progress, especially
in the application of earth abundant transition metals as cata-
lysts in this transformation. Moreover, reports of metal com-
plexes involving redox active ligands in the reduction of esters
via hydrosilyation remain limited,14 especially in comparison to
other transformations where their use is extensive.15a,7c,15b–d
Our group has been actively pursuing applications of earth
abundant transition metals (M = Fe,16 Ni,17 Co18) in catalytic
hydrofunctionalization chemistry. We have previously reported
the syntheses of two dppBIANFe complexes: dppBIANFeCl2 and
dppBIANFe(C7H8) (BIAN
=
1,2-((bis-2,6-diisopropylphenyl)
imino)acenaphthene)16a in which the acenaphthene backbone
provides a rigid scaffold for the Fe complex (Scheme 1). The
latter was found to be active in the hydrosilylation of aldehydes
and ketones under solvent and additive free conditions. In
2016, Hoyt and coworkers reported dppBIANFeBr2 which is,
after activation with alkali metal salts, capable of the hydro-
silylation of 1-hexene (Scheme 1).19 In 2017, von Wangelin
Department of Chemistry & Biochemistry, Texas Tech University, Lubbock, TX 79409,
USA. E-mail: michael.findlater@ttu.edu
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8ob02661k
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
Scheme 1 Examples of dppBIAN[Fe] in hydrofunctionalization catalysis.
reported that the addition of 3 equivalents of n-BuLi to a solu-
tion of dppBIANFeCl2 in toluene generated an anionic (but
otherwise unknown) [Fe] species which was catalytically active
in the hydrogenation of alkenes (Scheme 1).20 Intrigued by
this report, we wondered if a similar approach could be used
in the reduction of esters.
Scheme 2 Ligand effects on catalytic hydrosilylation of methyl benzo-
ate. Yields determined by GC-MS.
Initially, we chose methyl benzoate as a model substrate to
explore hydrogenation of esters. Thus, we exposed a reaction
mixture of methyl benzoate, 10 mol% dppBianFeCl2 and
30 mol% n-BuLi in toluene to H2 at 20 bar pressure; the reac-
tion mixture was heated at 100 °C. Analysis of the reaction
mixture by GC-MS after 24 h showed little to no reaction had
occurred. Increasing the applied H2 pressure to 100 bar simi-
larly afforded no reaction. Next, we began to explore hydrosilyl-
ation as a reduction method; initially a variety of silanes were
screened for activity. Initial screening of the catalytic trans-
formation of methyl benzoate in the presence of silane was
performed using 10 mol% dippBIANFeCl2, 30 mol% n-BuLi,
and 3 mmol polymethylhydrosiloxane (PMHS); 100% conver-
sion of methyl benzoate (2a) was observed when the reaction
mixture was analyzed by GC-MS after 20 h. Lowering the cata-
lytic loading to 1 mol% did not substantially affect catalytic
performance as 97% of benzyl alcohol was formed under
otherwise identical reaction conditions.21
Table 1 Optimization with various metals
Entry
[M]
Conversiona (%)
1
2
3
4
5
6
7
8
9
dppBIANFeCl2
dppBIANCoCl2
dppBIANMnCl2
dppBIANCuCl2
dppBIANPdCl2
Fe(acac)3
97
44
95
0
0
44
0
Fe(OTf)3
C7H5FeIO2
46
29
dppBIANFe(η6-C7H8)b
a Yields determined by GC-MS employing mesitylene as internal stan-
dard. b No alkyllithium present.
Subsequently, we investigated the influence of different
types of redox active (and redox inactive) ligands (Scheme 2).
No significant difference was observed employing either PDI
(L2) or mesDAB (L3) as the supporting ligand. The reactions per- salts such as Fe(acac)3, Fe(OTf)3 and cyclopentadienyliron
formed using either FeCl2/L2 or FeCl2/L3 afford excellent yields dicarboxyl iodide (C7H5FeIO2) showed little to moderate
of benzyl alcohol under our optimized reaction conditions. activity, and dppBIANFe(η6-C7H8) which was previously reported
When FeCl2/L4 (bipyridine) was used as the catalyst, only 62% by our group to be highly active in hydrosilylation of carbo-
of benzyl alcohol was formed (Scheme 2). Little to no reaction nyls16a yielded only 29% of benzyl alcohol in the absence of
was observed when redox innocent NHC-based ligands (L5 and n-BuLi.
L6) were employed. This implies that the electronic nature
The use of a range of sterically and electronically diverse
(redox innocent vs. non-innocent) of the ligand may play an silylating reagents was examined (Table 2). GC-MS analysis of
important role in facilitating this catalytic transformation. the reaction mixtures showed excellent conversion of methyl
With these results in hand and our prior work with the BIAN benzoate when phenylsilane and triethoxysilane were used.
(L1) framework16a we decided to pursue further studies Moderate activity was observed when 1,1,3,3-tetramethyl-
employing MCl2/L1.
disiloxane (TMDS) and diphenylsilane were used as the silylat-
The effect of changing the metal on catalysis was also ing reagents.
tested. Different metal salts, M2+ (M = Cu, Mn, Co, and Pd),
With optimized conditions in hand, we screened a wide
were complexed with L1 (Table 1, entries 1–5). Mn showed variety of esters bearing electron donating, electron withdraw-
good activity as the reaction yielded 95% of benzyl alcohol. ing, and halide substituents, N-heteroarenes, alicylic, and ali-
Interestingly, little to no reaction was observed when com- phatic esters were also explored in an effort to show the
plexes of Pd and Cu were used. Commercially available iron general applicability of our method (Scheme 3). In most cases,
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Communication
radical transfer led to the deactivation of the catalyst.14b In an
effort to elucidate the mechanistic details in our system, FTIR
and EPR studies were conducted. In particular, the role of
n-BuLi was one we wished to address.
Table 2 Screening of various silylating reagents
The direct reaction of dppBIAN with t-BuLi to generate a
radical (dearomatized) BIAN backbone has been reported in
the literature.22 Treatment of dppBIAN with one equivalent of
n-BuLi in hexanes or diethyl ether has been reported to give
Entry
[Si]
Yielda (%)
1
2
3
4
5
PMHS
97
29
80
85
54
the
crystallographically
characterized
dppBIAN(n-Bu)Li
Ph2SiH2
PhSiH3
(EtO)3SiH
TMDS
(Scheme 4).23 An additional equivalent of n-BuLi gives
[{dppBIAN(n-Bu)Li}n-BuLi]2, which exists as a dimer in the solid
state. EPR analysis of a solution of dppBIAN in toluene after
addition of 1 equivalent of n-BuLi revealed an EPR signal with
hyperfine coupling (AN-14 = 4.31 G, ALi-6/7 = 3.84/3.87 G) and
g-factor of 2.003 which implies the presence of an organic
radical (Fig. 1) that is perhaps more consistent with a dppBIAN
radical anion.24 The paramagnetic nature of this product
seems initially at odds with the results of Fedushkin et al.23
The major difference between the two experiments is in the
choice of solvent and the concentration. Here, toluene was
used, and the concentrations were much lower than those
used in the preparation and isolation of dppBIAN(n-Bu)Li. If
the reaction begins with a single electron transfer to the
dppBIAN ligand, the resulting butyl radical can either react
with the dppBIAN radical anion, or abstract hydrogen from a
solvent molecule. Here, the lower concentration allows for the
observation of the presumed dppBIANLi radical as shown in
Scheme 4. Addition of subsequent aliquots of n-BuLi resulted
a Based on GC-MS analysis employing mesitylene as an internal
standard.
Scheme 4 Known and proposed reaction pathways of dppBIAN with
n-BuLi.
Scheme 3 Substrate scope for hydrosilylation of esters. Isolated yields;
a values in parentheses represent GC-MS conversion employing mesity-
lene as an internal standard.
good to excellent isolated yields were obtained except for 2h,
where no reaction was observed. N-Heteroarenes bearing ester
substrates showed limited reactivity (2l, 7% and 2m, 10%).
Saturation of the α,β double bond was observed for trans
methyl cinnamate (2n); and mixtures of alcohols from hydro-
silylation were also observed.
Trovitch and coworkers have recently proposed a mecha-
nism in which a radical transfer and abstraction of H atom
occurs in the hydrosilylation of esters in the presence of a
manganese catalyst supported by a redox active ligand (PDI);
the final silyl ether product arises from the subsequent hydro-
silylation of the aldehyde generated in situ after β-alkoxide
elimination.14a In subsequent work, these authors employed
Fig. 1 X-band EPR spectra of dppBIAN with 1 eq. of n-BuLi in toluene at
detailed experimental and computational analysis to show that 298 K. Top: Simulated. Bottom: Experimental.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
in a significant decrease in the intensity of the EPR signal of a new peak at 1653 cm−1 suggesting an interaction of the
(Fig. S1†). Similarly, titration of n-BuLi was performed on (now) active catalyst with the ester. This coordination could be
dppBIANFeCl2 and EPR spectra recorded after each addition of presumed to be an initial step prior to the cleavage of the C–O
n-BuLi. dppBIANFeCl2 was found to be EPR silent. A weak bond. A comprehensive study is currently undergoing in our
signal with hyperfine coupling of 4.2 G (g factor = 2.003) to laboratory to isolate intermediates, and understand the role of
two equivalent 14N nuclei was observed when 1 equiv. of the redox active ligand, dppBIAN, in the hydrosilyation reac-
n-BuLi was added to the reaction mixture. An additional equi- tions of esters.
valent of n-BuLi led to a signal of higher intensity with a broad
In conclusion, we have shown that hydrosilylation of esters
coupling pattern (Fig. 2). This spectrum can be simulated with can be achieved upon activation of dppBIANFeCl2 with 3 equiv.
the convolution of two EPR active species with g-factors of of n-BuLi. Electron donating, electron withdrawing, aliphatic,
2.003. The first contains hyperfine coupling to a single 14N and alicyclic esters were tolerated and good to excellent reac-
(A = 4.2 G) and 6/7Li (A = 3.7 G). The second is a broad EPR tion yields were observed. Preliminary work suggests the possi-
signal. The signal increases with the addition of the third equi- bility of the involvement of radical transfer in catalysis.
valent of n-BuLi (Fig. S2†). The hyperfine coupling remains the Further investigation employing both computational and
same, but the contribution from the species with the broad experimental studies, designed to more fully understand the
EPR features increases. A possible model for this could be mechanism, are currently ongoing in our laboratory.
competing ligand metathesis and complex reduction at each
step (Scheme 5).
FTIR data of the sample which was taken upon addition of
a stoichiometric amount (relative to the catalyst) of methyl
Conflicts of interest
benzoate to the solution after activation of the catalyst with 3 There are no conflicts to declare.
equiv. of n-BuLi revealed a change in the stretching mode of
the νc–o band of the ester (Fig. S3 and S4†), and the appearance
Acknowledgements
The financial support of the Robert A. Welch Foundation is
gratefully acknowledged (Grant No. D1807). Dr Anne-Marie
Dechert-Schmitt is thanked for helpful discussions and com-
ments on an earlier draft of this manuscript.
Notes and references
1 (a) J. Maes and B. U. W. Maes, in Adv. Heterocycl. Chem, ed.
E. F. V. Scriven and C. A. Ramsden, Academic Press, 2016,
vol. 120, pp. 137–194; (b) J. E. Zweig, D. E. Kim and
T. R. Newhouse, Chem. Rev., 2017, 117, 11680–11752;
(c) J. D. Hayler, D. K. Leahy and E. M. Simmons,
Organometallics,
organomet.8b00566.
2018,
DOI:
10.1021/acs.
Fig. 2 X-band EPR spectra of dppBIANFeCl2 with 3 eq. of n-BuLi in
toluene at 298 K. Top: Simulated. Bottom: Experimental.
2 M. S. Hill, D. J. Liptrot and C. Weetman, Chem. Soc. Rev.,
2016, 45, 972–988.
3 J. Alemán and S. Cabrera, Chem. Soc. Rev., 2013, 42, 774–
793.
4 (a) H. Grützmacher, Angew. Chem., Int. Ed., 2008, 47, 1814–
1818; (b) J. R. Khusnutdinova and D. Milstein, Angew.
Chem., Int. Ed., 2015, 54, 12236–12273.
5 S. Díez-González, N. Marion and S. P. Nolan, Chem. Rev.,
2009, 109, 3612–3676.
6 (a) P. J. Chirik, Inorg. Chem., 2011, 50, 9737–9740;
(b) D. Zhu, I. Thapa, I. Korobkov, S. Gambarotta and
P. H. M. Budzelaar, Inorg. Chem., 2011, 50, 9879–9887.
7 (a) V. Lyaskovskyy and B. de Bruin, ACS Catal., 2012, 2,
270–279; (b) J. I. van der Vlugt, Eur. J. Inorg. Chem., 2012,
2012, 363–375; (c) O. R. Luca and R. H. Crabtree, Chem.
Soc. Rev., 2013, 42, 1440–1459.
Scheme 5 Proposed reaction pathways of dppBIANFeCl2 with n-BuLi.
Black reaction arrows depict metathesis steps, blue reaction arrows
depict complex reduction.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Communication
8 J. Magano and J. R. Dunetz, Org. Process Res. Dev., 2012, 16, 15 (a) S. Blanchard, E. Derat, M. Desage-El Murr,
1156–1184.
L. Fensterbank, M. Malacria and V. Mouriès-Mansuy,
Eur. J. Inorg. Chem., 2012, 2012, 376–389; (b) A. Quintard
and J. Rodriguez, Angew. Chem., Int. Ed., 2014, 53, 4044–
4055; (c) P. J. Chirik, Acc. Chem. Res., 2015, 48, 1687–1695;
(d) I. J. S. Fairlamb, Angew. Chem., Int. Ed., 2015, 54, 10415–
10427.
9 (a) J. Zhang, G. Leitus, Y. Ben-David and D. Milstein,
Angew. Chem., Int. Ed., 2006, 118, 1131–1133;
(b)
S. Chakraborty,
H. Dai, P. Bhattacharya,
N. T. Fairweather, M. S. Gibson, J. A. Krause and H. Guan,
J. Am. Chem. Soc., 2014, 136, 7869–7872; (c) S. Werkmeister,
K. Junge and M. Beller, Org. Process Res. Dev., 2014, 18, 16 (a) F. S. Wekesa, R. Arias-Ugarte, L. Kong, Z. Sumner,
289–302; (d) S. Werkmeister, K. Junge, B. Wendt,
E. Alberico, H. Jiao, W. Baumann, H. Junge, F. Gallou and
M. Beller, Angew. Chem., Int. Ed., 2014, 53, 8722–8726;
(e) S. Elangovan, M. Garbe, H. Jiao, A. Spannenberg,
K. Junge and M. Beller, Angew. Chem., Int. Ed., 2016, 55,
G. P. McGovern and M. Findlater, Organometallics, 2015,
34, 5051–5056; (b) A. D. Smith, A. Saini, L. M. Singer,
N. Phadke and M. Findlater, Polyhedron, 2016, 114, 286–
291; (c) S. R. Tamang and M. Findlater, J. Org. Chem., 2017,
82, 12857–12862.
15364–15368;
(f)
J.
Yuwen,
S.
Chakraborty, 17 S. R. Tamang, A. Singh, D. K. Unruh and M. Findlater, ACS
W. W. Brennessel and W. D. Jones, ACS Catal., 2017, 7,
3735–3740.
10 J. Seyden-Penne, Reductions by the Alumino– and
Catal., 2018, 8, 6186–6191.
18 S. R. Tamang and M. Findlater, Dalton Trans., 2018, 47,
8199–8203.
Borohydrides in Organic Synthesis, Wiley-VCH, Weinheim, 19 M. J. Supej, A. Volkov, L. Darko, R. A. West, J. M. Darmon,
Germany, 2nd edn, 1997.
11 C. C. Chong and R. Kinjo, ACS Catal., 2015, 5, 3238–3259.
C. E. Schulz, K. A. Wheeler and H. M. Hoyt, Polyhedron,
2016, 114, 403–414.
12 S. Díez-González and S. P. Nolan, Org. Prep. Proced. Int., 20 M. Villa, D. Miesel, A. Hildebrandt, F. Ragaini,
2007, 39, 523–559.
D. Schaarschmidt and A. Jacobi von Wangelin,
13 (a) S. C. Berk, K. A. Kreutzer and S. L. Buchwald, J. Am.
ChemCatChem, 2017, 9, 3203–3209.
Chem. Soc., 1991, 113, 5093–5095; (b) S. C. Berk and 21 Initial optimization revealed a combination of 1 mol%
S. L. Buchwald, J. Org. Chem., 1992, 57, 3751–3753;
(c) Z. Mao, B. T. Gregg and A. R. Cutler, J. Am. Chem. Soc.,
1995, 117, 10139–10140; (d) M. Igarashi, R. Mizuno and
T. Fuchikami, Tetrahedron Lett., 2001, 42, 2149–2151;
catalyst and 2 mol% n-BuLi to be effective in the hydrosilyl-
ation of 2a; however, extension of methodology to various
substrates bearing aliphatic, alicylic, and electron donating
esters revealed poor to moderate activity (Scheme S1†).
(e) S. Das, K. Möller, K. Junge and M. Beller, Chem. – Eur. 22 (a) D. A. Evans and A. H. Cowley, J. Am. Chem. Soc., 2012,
J., 2011, 17, 7414–7417.
14 (a) T. K. Mukhopadhyay, M. Flores, T. L. Groy and
134, 15672–15675; (b) D. A. Evans, I. Vargas-Baca and
A. H. Cowley, J. Am. Chem. Soc., 2013, 135, 13939–13946.
R. J. Trovitch, J. Am. Chem. Soc., 2014, 136, 882–885; 23 I. L. Fedushkin, M. Hummert and H. Schumann,
(b) T. K. Mukhopadhyay, C. L. Rock, M. Hong, D. C. Ashley, Eur. J. Inorg. Chem., 2006, 2006, 3266–3273.
T. L. Groy, M.-H. Baik and R. J. Trovitch, J. Am. Chem. Soc., 24 I. L. Fedushkin, A. A. Skatova, V. A. Chudakova and
2017, 139, 4901–4915.
G. K. Fukin, Angew. Chem., Int. Ed., 2003, 42, 3294–3298.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.