Catalytic carbonyl hydrosilylations: Via a titanocene borohydride-PMHS reagent system

Article abstract of DOI:10.1039/c7cy01088e

Reduction of a wide range of aldehydes and ketones with catalytic amounts of titanocene borohydride in concert with a stoichiometric poly(methylhydrosiloxane) (PMHS) reductant is reported. Preliminary mechanistic studies demonstrate that the reaction is mediated by a reactive titanocene(iii) complex, whose oxidation state remains constant throughout the reaction.

Full text of DOI:10.1039/c7cy01088e

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Catalytic carbonyl hydrosilylations via a
titanocene borohydride–PMHS reagent system†
Cite this: DOI: 10.1039/c7cy01088e
*
Godfred D. Fianu, Kyle C. Schipper and Robert A. Flowers II
Received 31st May 2017,
Accepted 1st July 2017
Reduction of a wide range of aldehydes and ketones with catalytic amounts of titanocene borohydride in
concert with a stoichiometric polyIJmethylhydrosiloxane) (PMHS) reductant is reported. Preliminary mecha-
nistic studies demonstrate that the reaction is mediated by a reactive titanoceneIJIII) complex, whose oxida-
tion state remains constant throughout the reaction.
DOI: 10.1039/c7cy01088e
rsc.li/catalysis
Carbonyl reductions to alcohols are an important and
fundamental reaction in organic chemistry. Synthetic organic
chemists have routinely applied this transformation to form
basic motifs in multi-step syntheses of complex organic mole-
cules, of biological and synthetic utility.1–5 In view of its signifi-
cance, a large number of methods have been developed for car-
bonyl reductions to their respective alcohols. Notable methods
include reductions mediated by hydride sources such as lith-
ium aluminium hydride (LAH) and sodium borohydride
(NaBH4),6 Meerwein–Ponndorf–Verley reductions,7,8 samarium-
mediated reductions,9,10 catalytic hydrogenations11–13 and low-
valent metal-catalyzed hydroborations and hydrosilylations.14–17
Work described herein is primarily focused on carbonyl
reductions mediated by catalytic titanocene borohydride. We
were inspired by seminal work of Schwartz and coworkers
whereby simple ketones were reduced to alcohols in aprotic
media by titanocene borohydride: a complex that is easily
generated in situ, by mixing titanocene dichloride and NaBH4
in dimethoxyethane (DME) (Scheme 1).18 Although this
method was simple, prepared from readily available reagents,
and performed under mild conditions, it required stoichio-
metric amounts of titanocene borohydride. To the best of our
knowledge, a catalytic approach to this process has not yet
been developed. Conversely, Buchwald and coworkers devel-
oped an elegant approach to carry out hydrosilylation of ke-
tones with subsequent workup to reduce alcohols with cata-
lytic low-valent titanocene hydride complexes and an
inexpensive siloxane polymer, polyIJmethylhydrosiloxane)
(PMHS) as the stoichiometric reductant. In this approach,
the active titanocene hydride species was either synthesized
from a premade titanocene precatalyst or generated in situ at
very low temperatures with reagents that required careful
handing (Scheme 1).19–22
Based on the aforementioned titanocene mediated reduc-
tions developed by Schwartz and Buchwald, we envisioned
designing a system whereby carbonyls could be reduced to
their respective alcohols with catalytic amounts of titanocene
borohydride. Ideally, this complex would be generated under
mild conditions, from inexpensive, readily available
titanocene dichloride and a hydride source as a stoichiomet-
ric reductant. Although catalytic approaches have been
reported for reducing carbonyls to alcohols in refluxing diiso-
propylether (DIPE) using catalytic ZnIJ2-ethylhexanoate)2 and
NaBH4 with PMHS as the stoichiometric reductant,23 a simi-
lar approach with commercially available titanocene
dichloride has not been reported. Herein, we present a proce-
durally simple, inexpensive, and mild approach for the hydro-
silylation of aldehydes and ketones and subsequent conver-
sion to their respective alcohols at room temperature with
catalytic amounts of titanocene borohydride and PMHS as
the stoichiometric reductant (Scheme 1).
To optimize the reaction conditions, a series of experi-
ments were performed and the results are summarized in
Table 1. As shown, 2-octanone (1) was successfully reduced to
Department of Chemistry, Lehigh University, Bethlehem, PA 18015, USA.
E-mail: rof2@lehigh.edu
† Electronic supplementary information (ESI) available: Experimental details, IR
and NMR spectra. See DOI: 10.1039/c7cy01088e
Scheme 1 Low-valent titanocene catalyzed carbonyl reductions.
This journal is © The Royal Society of Chemistry 2017
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
2-octanol (2) with 10 mol% catalyst and with NaBH4 as the
stoichiometric hydride source in DME (Table 1, entry 1). The
reduction of 1 to 2 was ineffective in DME solvent with either
only NaBH4 as reductant or with the addition of 5 mol% cata-
lyst (Table 1, entries 2 & 3). Complete reduction of 1 to 2 was
achieved with 5 mol% catalyst and PMHS as stoichiometric
reductant (Table 1, entry 4). Control experiments showed that
the reaction was unsuccessful without the active titanocene
catalyst (Table 1, entries 5 & 6).
Reductions were also performed in THF, which proceeded
smoothly but relied on a very low catalyst to solvent ratio
(6.25 mg (Cp2TiCl2) : 10 mL (THF) or 2.5 mM of catalyst). At
higher catalyst to solvent ratios (25 mg (Cp2TiCl2) : 10 mL
(THF) or 10 mM of catalyst), the active titanocene borohy-
dride complex was not formed until after three days due to
the poor solubility of NaBH4 in THF, corroborating observa-
tions made by Schwartz and coworkers.18 Attempts to form
the active titanocene catalyst directly by mixing titanocene
dichloride with PMHS were not successful. In addition to
NaBH4, attempts to form the active titanocene borohydride
intermediate with milder hydride sources including sodium
cyanoborohydride (NaBH3CN) and sodium triacetoxy-
borohydride (NaBHIJOAc)3), were not successful and provided
only recovered starting material (see ESI† for details). With
these control studies in hand, further reactions were
performed using DME as the solvent of choice because it pro-
vided the fastest preparation of the active catalyst.
To further investigate the breadth of this approach, we
probed the reduction of a variety of aldehydes and ketones
with different steric and electronic properties. Our findings
are summarized in Scheme 2. All of the aldehydes and ke-
tones examined were successfully reduced to alcohols
irrespective of their structural features. Under the reaction
conditions, the cis-isomer was preferred to the trans-isomer
for ketones 2-l & 2-n, suggesting that the hydride preferen-
tially attacks the carbonyl carbon at the least sterically hin-
dered position. Nardi and coworkers observed similar trends
in their ErIJOTf)3/NaBH4 catalyzed reductions of α,β-unsatu-
rated ketones.15 Like Luche reductions,24,25 α,β-unsaturated
ketones 2-l and 2-u favorably reduced via 1,2 addition under
the reaction conditions. This approach serves as a compli-
Scheme
2
Reduction of aldehydes and ketones Cp2TiCl2/NaBH4/
PMHS in DME. Conversion to alcohol based on NMR analysis & isolated
yields in parenthesis; reaction time was 1 hour; reaction time was 1–
3 hours; reaction stirred overnight.
1
2
3
ment to titanoceneIJIII) catalyzed reductions of α,β-unsaturated
ketones to ketones.26 1-(2-Phenylcyclopropyl) ethan-1-one (1-v)
was successfully reduced to 1-(2-phenylcyclopropyl) ethan-1-ol
(2-v) without any evidence of ring opening. This observation
suggests that this reaction proceeds via a hydride transfer as
opposed to a hydrogen atom transfer (HAT). Heterocycles
with pendant carbonyls (1-w & 1-x) were also effectively re-
duced to their respective alcohols (2-w & 2-x).
Table 1 Optimizing conditions and control experiments
From a mechanistic standpoint, we hypothesized that the
carbonyl was being reduced via a hydrosilylation instead of a
hydrogen atom transfer (vide supra). To further study the pro-
cess, the model reaction shown in Scheme 3 was performed
and monitored by ReactIR. The C–H wag of the Cp ligand is
known to be sensitive to the oxidation state of the titanocene
complex as well as the nature of the ligands coordinated to
the titanium metal.27,28 Additionally, in situ monitoring also
enabled us to follow carbonyl reduction and conversion of
the Si–H bond to an Si–O bond. Since PMHS has an absorp-
tion that interferes with the C–H wag of the Cp ligand of the
titanocene, we employed phenylsilane (PhSiH3) as there was
no interference from this terminal reductant.
Cp2TiCl2
Entry (mol%)
NaBH4
(mol%)
PMHS
(mol%)
Time
(h)
Yielda
(%)
1
2
3
4
5
6
10
—
5
5
—
200
400
150
10
—
10
—
—
—
150
150
150
4
96
4
1
24
24
95
41
57
100b
—
—
13
a
b
GC yields with biphenyl as internal standard. Only product
observed by GC.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Catalysis Science & Technology
Paper
The titanocene borohydride complex whose X-ray structure
is known,29 was generated by the addition of sodium borohy-
dride to titanoceneIJIV) dichloride. Reduction of TiIJIV) to TiIJIII)
and formation of the titanocene borohydride was demon-
strated by the shift of the C–H wag of the Cp ligand (Fig. 1a)
from 820 cm−1 to 809 cm−1 and a change in color from red to
violet was observed. Once this conversion was complete, ke-
tone 1 was added to the solution and the C–H absorption of
the Cp shifted from 809 cm−1 to 799 cm−1 (Fig. 1b) and a
change in color from violet to green was observed.
Fig. 2a contains a plot of the change in absorption over
time for the C–H wag for the Cp of titanocene borohydride
before and after addition of ketone 1. Upon addition of the
ketone, the absorption at 1719 cm−1 was observed. Concur-
rently there was a disappearance of the C–H wag absorption
for titanocene borohydride at 809 cm−1, and the subsequent
growth of another absorption at 799 cm−1 consistent with the
C–H wag of a Cp2Ti–O interaction.27,28 The initial growth of
ketone observed at 1719 cm−1 was followed by a minor drop
in the absorption consistent with initial reduction of 2-octan-
one by titanocene borohydride (Fig. 2a). Furthermore, the
C–H wag absorption at 799 cm−1 characteristic of Cp2TiIJIII)
complexes indicates that the titanocene catalyst remains in
the +3 oxidation state bound to the alkoxide.18,30,31
Once no further changes were observed, phenylsilane was
added to the solution to initiate the reaction. The ketone ab-
sorption at 1719 cm−1 decayed over the course of approxi-
mately 100 minutes until the reaction was complete (Fig. 2b).
The Si–H stretch of phenylsilane was observed at 705 cm−1.
The phenylsilane peak decreased with concomitant develop-
ment of another absorption at 835 cm−1 which is attributed
to the OSi–H stretch due to silylation of the TiIJIII)-bound alk-
oxide (see ESI† for details). Interestingly, the C–H wag absor-
bance for titanocene at 799 cm−1 remains after ketone reduc-
tion, indicating the presence of either TiIJIII)–O species or an
active titanocene hydride intermediate. At this point, the re-
action was exposed to air and worked up with 1 M NaOH,
and provided a nearly quantitative yield of the carbinol
product.
Fig. 1 a) Formation of titanocene borohydride monitored by ReactIR.
b) Shift in C–H wag after ketone addition (see ESI† for details).
To verify the intermediacy of a TiIJIII) hydride, a titanocene
hydride complex was generated via the approach described
by Buchwald20–22 and monitored by ReactIR. In this experi-
ment, titanocene difluoride was mixed with phenylsilane in
refluxing THF. The formation of a titanoceneIJIII) hydride spe-
cies was demonstrated by the shift in the C–H wag of the Cp
ligand of titanocene from 813 cm−1 to 799 cm−1 and a change
in color from yellow to green was observed (see ESI† for ex-
perimental details). The C–H wag absorbance at 799 cm−1
suggests that a TiIJIII) hydride complex is formed as an inter-
mediate in the reduction process with titanoceneIJIII) borohy-
Fig. 2 a) Trend observed on ReactIR for 2-octanone and C–H wag of
Cp for titanocene complex. b) Monitoring reaction after addition of
phenylsilane (see ESI† for experimental details).
dride and PMHS as the stoichiometric reductant. We also ob-
served the formation of PhSiF2–H through the presence of
absorbances at 876 and 858 cm−1 respectively.
Scheme 3 Reaction conditions for ReactIR studies.
This journal is © The Royal Society of Chemistry 2017
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
To determine if the titanocene hydride intermediate in
our approach is the same as the one generated by Buchwald,
the reduction of 2-octanone to 2-octanol was performed via
the method described by Buchwald and coworkers20–22 and
the results are summarized in Table 2. Reactions were moni-
tored after an hour and it was observed that reduction of
2-octanone only proceeded under refluxing conditions and
product yield was improved when the active catalyst was pre-
pared with 20 mol% PhSiH3 (Table 2, entries 4 & 5). With our
optimized reaction conditions however, full conversion to
product was realized within an hour and at room tempera-
ture (Table 1, entry 4). These results indicate that the
titanocene hydride generated in situ with our approach is not
identical to that generated in the system developed by
Buchwald and coworkers.
Fig. 3 TitanoceneIJIII) hydride complexes proposed by Harrod.32
Attempts to isolate and crystallize the intermediate gener-
ated using our approach has not been successful thus far.
Nonetheless, it is useful to examine other protocols devel-
oped in the literature for generating titanoceneIJIII) hydride
complexes. To this end, we looked at the studies on
titanoceneIJIII) hydride complexes done independently by
Harrod32 and Brintzinger.33,34 Brintzinger studied titanocene
hydride complexes generated with either hydrogen gas,
organolithiums or Grignard reagents.33,34 Harrod, however,
studied titanoceneIJIII) hydride complexes generated (Fig. 3) by
reacting dimethyltitanocene with phenylsilane.32 We there-
fore hypothesized that the titanocene hydride intermediate
generated in our system would be similar but not necessarily
identical to that generated by Harrod. Complex b was pre-
pared by reacting stoichiometric amounts of dimethyl-
titanocene (1 : 1 mole ratio) with phenylsilane while complex
c was prepared by reacting catalytic amounts of dimethyl-
Scheme 4 Proposed catalytic cycle for carbonyl reduction.
our titanocene borohydride–PMHS system and the approach
outlined by Buchwald. It should also be noted that the pres-
ence of a TiIJIII)-bound alkoxide cannot be ruled out as a possi-
ble intermediate since its Cp ligand has the same C–H wag
absorption at 799 cm−1.
titanocene with phenylsilane. They observed that
c
decomposed to b, which eventually decomposed to a. Inter-
mediate a was observed by ESR spectroscopy. All the interme-
diates were found to be reactive in polymerizing silanes with
concomitant evolution of H2 gas.32
It is therefore possible that for our system, the reduction
could be mediated by either one or more titanocene hydride
complexes generated in situ, which may have varying reactiv-
ity towards carbonyl reductions. This could explain the dis-
crepancies observed between the reductions performed via
Based on the results obtained from monitoring this pro-
cess via in situ IR spectroscopy, a catalytic cycle was proposed
as shown in Scheme 4. In the proposed cycle, the titanocene
borohydride, formed via the method described in the litera-
ture,18,35 reduces the carbonyl to form the titaniumIJIII) alkox-
ide.18,30,31 It is then posited that the titanium alkoxide reacts
with the silane, via a method similar to that proposed by
Buchwald and coworkers,20,21,36 leading to the formation of
silylated product and a titanoceneIJIII) hydride intermediate,
whose structural identity at the moment is unknown. This
intermediate presumably reacts with another carbonyl to
complete the cycle.
Table 2 Reduction of 1 to 2 with Cp2TiF2
In conclusion, a procedurally straightforward, inexpensive,
and mild approach to the reduction of aldehydes and ketones
with a titanocene borohydride–PMHS system has been demon-
strated. This method was used to successfully reduce a wide
range of aldehydes and ketones. Preliminary mechanistic stud-
ies suggest that the reaction is mediated by a reactive
titanocene hydride complex. Results from our studies also
demonstrate that the titanocene complex remains in the +3-oxi-
dation state throughout the reaction. Future endeavours will fo-
cus on asymmetric carbonyl reductions11,12,20,37 catalyzed by
titanocene borohydride. We are optimistic that this potentially
Entry
R3SiH
T (°C)
Yielda (%)
1
2
3
4
5
PhSiH3
PMHS
PMHS2
PMHS
PMHSb
r.t.
r.t.
r.t.
60
—
—
—
45
91
60
a
b
GC yields after 1 h with biphenyl as internal standard. Active
catalyst formed with 20 mol% PhSiH3 under reflux.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Catalysis Science & Technology
Paper
has direct application to the synthesis of complex organic com-
pounds. Work is also currently underway to extend this system
to reduce other related functional groups including esters, am-
ides, imines and nitriles. Initial results from these studies are
promising and will be reported in future publications.
14 M. Zhao, W. Xie and C. Cui, Chem. – Eur. J., 2014, 20,
9259–9262.
15 M. Nardi, G. Sindona, P. Costanzo, M. Oliverio and A.
Procopio, Tetrahedron, 2015, 71, 1132–1135.
16 F. S. Wekesa, R. Arias-Ugarte, L. Kong, Z. Sumner, G. P.
McGovern and M. Findlater, Organometallics, 2015, 34,
5051–5056.
17 A. Kaithal, B. Chatterjee and C. Gunanathan, Org. Lett.,
2015, 17, 4790–4793.
18 M. Barden and J. Schwartz, J. Org. Chem., 1995, 60,
5963–5965.
19 K. J. Barr, S. C. Berk and S. L. Buchwald, J. Org. Chem.,
1994, 59, 4323–4326.
Acknowledgements
We thank the U.S. National Science Foundation (CHE-
1123815) and a Faculty Innovation Grant from Lehigh Univer-
sity for the support of this work. We also thank Dr.
Sandeepan Maity for insightful discussions.
20 J. Yun and S. L. Buchwald, J. Am. Chem. Soc., 1999, 121,
5640–5644.
Notes and references
1 P. C. Fuenfschilling, P. Hoehn and J. P. Mutz, Org. Process
Res. Dev., 2007, 11, 13–18.
21 X. Verdaguer, M. C. Hansen, S. C. Berk and S. L. Buchwald,
J. Org. Chem., 1997, 62, 8522–8528.
2 C. Y. Chen, L. F. Frey, S. Shultz, D. J. Wallace, K.
Marcantonio, J. F. Payack, E. Vazquez, S. A. Springfield, G.
Zhou, P. Liu, G. R. Kieczykowski, A. M. Chen, B. D. Phenix,
U. Singh, J. Strine, B. Izzo and S. W. Krska, Org. Process Res.
Dev., 2007, 11, 616–623.
3 J. E. Cabaj, D. Kairys and T. R. Benson, Org. Process Res.
Dev., 2007, 11, 378–388.
4 T. Ikemoto, T. Ito, H. Hashimoto, T. Kawarasaki, A.
Nishiguchi, H. Mitsudera, M. Wakimasu and K. Tomimatsu,
Org. Process Res. Dev., 2000, 4, 520–525.
22 X. Verdaguer, U. E. W. Lange, M. T. Reding and S. L.
Buchwald, J. Am. Chem. Soc., 1996, 118, 6784–6785.
23 H. Mimoun, J. Org. Chem., 1999, 64, 2582–2589.
24 J.-L. Luche, J. Am. Chem. Soc., 1978, 100, 2226–2227.
25 J. L. Luche, L. Rodriguez-Hahn and P. Crabbe, J. Chem. Soc.,
Chem. Commun., 1978, 601–602.
26 A. D. Kosal and B. L. Ashfeld, Org. Lett., 2010, 12, 44–47.
27 A. Gansäuer, D. Von Laufenberg, C. Kube, T. Dahmen, A.
Michelmann, M. Behlendorf, R. Sure, M. Seddiqzai, S.
Grimme, D. V. Sadasivam, G. D. Fianu and R. A. Flowers,
Chem. – Eur. J., 2015, 21, 280–289.
5 R. Hett, Q. K. Fang, Y. Gao, S. A. Wald and C. H.
Senanayake, Org. Process Res. Dev., 1998, 2, 96–99.
6 J. Magano and J. R. Dunetz, Org. Process Res. Dev., 2012, 16,
1156–1184.
28 A. Gansäuer, S. Hildebrandt, A. Michelmann, T. Dahmen, D.
von Laufenberg, C. Kube, G. D. Fianu and R. A. Flowers,
Angew. Chem., Int. Ed., 2015, 54, 7003–7006.
7 P. K. Sharma, A. Kolchinski, A. Shea, J. J. Nair, Y. Gou, L. J.
Romanczyk, H. H. Schmitz, I. Company, H. Street and E.
Street, Org. Process Res. Dev., 2007, 11, 422–430.
8 J. S. Cha, Org. Process Res. Dev., 2006, 10, 1032–1053.
9 T. A. Davis, P. R. Chopade, G. Hilmersson and R. A. Flowers,
Org. Lett., 2005, 7, 119–122.
29 S. J. Lippard, D. Coucouvanis and K. M. Melmed, Inorg.
Chem., 1973, 12, 232–236.
30 K. Matsubara, S. Niibayashi and H. Nagashima,
Organometallics, 2003, 22, 1376–1382.
31 S. Niibayashi, K. Mitsui, Y. Motoyama and H. Nagashima,
J. Organomet. Chem., 2005, 690, 276–285.
10 T. V. Chciuk, W. R. Anderson and R. A. Flowers, J. Am.
Chem. Soc., 2016, 138, 8738–8741.
32 C. T. Aitken, J. F. Harrod and E. Samuel, J. Am. Chem. Soc.,
1986, 108, 4059–4066.
11 Y. Li, S. Yu, X. Wu, J. Xiao, W. Shen, Z. Dong and J. Gao,
J. Am. Chem. Soc., 2014, 136, 4031–4039.
12 W. Wu, S. Liu, M. Duan, X. Tan, C. Chen, Y. Xie, Y.
Lan, X. Q. Dong and X. Zhang, Org. Lett., 2016, 18,
2938–2941.
33 H. Brintzinger, J. Am. Chem. Soc., 1967, 89, 6871–6877.
34 J. E. Bercaw, R. H. Marvich, L. G. Bell and H. H. Brintzinger,
J. Am. Chem. Soc., 1972, 94, 1219–1238.
35 C. R. Lucas, Inorg. Synth., 1977, 17, 91–94.
36 M. T. Reding and S. L. Buchwald, J. Org. Chem., 1995, 60,
7884–7890.
13 T. J. Connolly, M. Matchett, P. Mcgarry, S. Sukhtankar and J.
Zhu, Org. Process Res. Dev., 2004, 8, 624–627.
37 S. Xin and J. Harrod, Can. J. Chem., 1995, 73, 999–1002.
This journal is © The Royal Society of Chemistry 2017
Catal. Sci. Technol.