ACS Catalysis
Research Article
According to the above mechanistic studies, catalytic cycle A
that includes the formation of a five-membered metallacyclic
intermediate and subsequent hydrolysis via attacking on the
phosphorus of the PR2OH ligand by H2O is strongly
supported for (P∼P)Pt(PR2OH)X(OTf)-catalyzed nitrile
effect of substitution on the aryl rings showed that ortho-,
meta-, and para-substituents (Me, Br, OMe, Cl, and COOMe),
being either electron-donating or withdrawing in nature, were
all well tolerated. Catalytic hydration of naphthalene/
benzylfuran/indole-substituted cyanohydrins were also inves-
tigated, and good yields and TONs were obtained in these
cases (21−23). To our knowledge, except lactonitrile
(2)16−19,22,23 and mandelonitrile (3),22,23 hydration of the
remaining aldehyde-derived cyanohydrins depicted in Table
1A has not been reported. In addition, cyanohydrins 2, 9, 12−
14, 16, 20, and 23 were hydrated in gram scales.
hydration, which is similar to that proposed by Cadierno and
28
́
Lopez for [RuCl2(η6-C6H6)(PMe2OH)]. Guided by the
proposed mechanism, a highly efficient catalyst that could
solve the challenging direct hydration of cyanohydrins should
meet the following requirement: electron-rich bisphophine
ligands with smaller bite angles25,26 than 1,1′-ferrocenendiyl-
bis(diphenylphosphine) (DPPF) could enhance the nucleo-
philicity of the hydroxyl group in the trans PR2OH ligand that
is more sterically accessible for the five-membered metallacycle
and subsequent hydrolysis. Therefore, as shown in Figure 7A,
we envisioned that combining PMe2OH and electron-rich
bidentate ligands with smaller bite angles, such as 1,2-
b i s ( d i a r y l p h o s p h i n o ) b e n z e n e , 3 3 − 3 5 1 , 2 - b i s -
(diphenylphosphino)ethane,35−38 or 1,8-bis(diarylphosphino)-
naphthalene,39 would make potentially active catalysts for
hydration of cyanohydrins. According to the hypothesis,
platinum catalysts A−G were synthesized, and the X-ray
structures of complexes A, D, and F showed that their P−Pt−P
bite angles are 91.0, 87.0, and 90.5°, respectively (Figure 7B),
all of which are smaller than that of previously reported
catalysts H and I (P−Pt−P bite angles for H: 98.5°; for I:
100.0°).22 For comparison, the efficiencies of catalysts A−H (2
mol %) were assessed in the hydration of mandelonitrile (3)
(Figure 7B). To our delight, all the newly synthesized Pt
catalysts A−G show much higher activities than previously
reported optimal catalyst H.22 Particularly, catalyst B that
harbors electron-donating 1,8-bis[bis(5-methyl-2-furanyl)-
phosphine]naphthalene is able to completely hydrate man-
delonitrile (3) to mandelamide (3a) within 10 min. Pleasingly,
increasing the electron-donating ability of the bidentate ligand
results in even higher activity of catalyst A and further
accelerates the hydration reaction.
Ketone-derived cyanohydrins are far more challenging to be
hydrated than aldehyde-derived cyanohydrins. The enhanced
activity of catalyst A permits the hydration of a wide range of
ketone-derived cyanohydrins, including α,α-dialkyl, α,α-
alkylaryl, and α,α-diaryl-substituted cyanohydrin, to proceed
at room temperature with good TONs (Table 1B). The TONs
for catalytic hydration of acetone cyanohydrin 5 and
cyclohexanone cyanohydrin 6 by catalyst A were largely
improved compared to those obtained under the catalysis of
catalysts H and I.22 Catalytic hydration of other α,α-dialkyl-
substituted cyanohydrins including 24, 25, and 26 that had yet
to report to direct hydration was achieved by 0.5 mol % of
catalyst A with good yields and TONs. Furthermore,
mesterolone-derived cyanohydrin 27 was investigated, and its
direct hydration proceeds smoothly with 1 mol % of catalyst A,
providing the corresponding α-hydroxyamide in 77% yield as a
single diastereomer. Alkyl aryl disubstituted cyanohydrins 28
and 29 were then explored, and their hydration reactions also
occur at ambient temperature with high efficiency. α,α-Diaryl-
substituted cyanohydrins were the least stable and often
rapidly decomposed into the corresponding benzophenone
and HCN at room temperature. The previously reported Pd-
catalyzed transfer hydration was the only method that can
successfully convert these kinds of challenging cyanohydrins
into their α-hydroxyamides.23 We were pleased to find that
catalyst A was effective in hydrating a wide range of bulky
diaryl-substituted cyanohydrins (30−38) at room temperature,
including heteroaryl aryl disubstituted cyanohydrins (37 and
38) that have no literature precedency.
Substrate Scope. Following the successful establishment
of the catalytic system for hydration of cyanohydrins, the scope
of this reaction protocol was studied. As compiled in Table 1,
catalyst A displays noticeably enhanced catalytic activity
toward various cyanohydrins and sterically encumbered nitriles
with good to excellent TONs under mild conditions (Table
1A, aldehyde-derived cyanohydrins: 210-2600 TON; Table 1B,
ketone-derived cyanohydrins: 33-383 TON; Table 1C, steri-
cally hindered nitriles: 371-4000 TON). Used as low as 0.04
mol %, catalyst A allows hydration of lactonitrile (2) to
complete in 2 h with significantly higher TONs (2450 TON)
than previously observed for catalyst I (463 TON),22 Ru
complexes (20 TON),16−18 Parkins catalysts (110 TON),15
and a Pd-catalyzed transfer hydration (48 TON),23 and the
calculated turnover frequency of 1225 h−1 is much higher than
those reported.15−18,22,23 Then, cyanohydrins bearing a variety
of alkyl chain substitutions such as butyl (9), iso-propyl (10),
cyclo-propyl (11), cyclo-pentyl (12), cyclo-hexyl (13), and
phenylethyl (14) were hydrated, and their corresponding α-
hydroxylamides were obtained with excellent TONs. Aryl-
substituted cyanohydrins (3, 15−23) were less reactive and
thermally less stable than their aliphatic analogues; however,
they were all successfully hydrated by 0.2 mol % of Pt catalyst
A with good yields and TONs. Particularly, enantiomerically
enriched mandelonitrile (R)-(3) does not lose the stereo-
chemical fidelity under the hydration conditions. Exploring the
To demonstrate synthetic generality of this catalytic
protocol, hydration of sterically hindered nitriles (39−46,
Table 1C) was conducted with 0.1 mol % of catalyst A at
ambient temperature, and the corresponding amides were
isolated in excellent yields and TONs. Notably, catalyst A
allows hydration reaction of trimethylacetonitrile 39 to
proceed with significantly higher TONs (4000 TON) than
previously observed for the Parkins catalyst (160 TON).6f
CONCLUSIONS
■
In summary, we have developed a new cationic Pt catalyst for
general and practical hydrolysis of cyanohydrins to afford α-
hydroxyamides. X-ray crystallography and 18O-labeling experi-
ments proved the catalytic cycle that involves a five-membered
metallacyclic intermediate and subsequent hydrolysis via
attacking of the phosphorus of the PMe2OH ligand by H2O.
Rational design of the ligand structure led to the discovery of
Pt catalyst A, bearing the electron-rich, suitable-bite-angle
bisphosphine ligand, which demonstrates high activity toward
hydration of cyanohydrins. The catalytic protocol features mild
reaction conditions, low catalyst loadings, and broad substrate
scopes (both aldehyde- and ketone-derived cyanohydrins),
affording the corresponding α-hydroxyamides with TONs that
8723
ACS Catal. 2021, 11, 8716−8726