Angewandte
Communications
Chemie
a study of the stereochemical integrity of chiral 1,2-amino
alcohols in hydrogen-borrowing alkylations.[10] Any retention
of absolute stereochemistry from the starting amino alcohol
would provide a new route to enantioenriched products. Our
initial concern lay in the vulnerability of reactive intermedi-
ates i and ii (Scheme 1B), formed during alkylation, to the
basic reaction conditions: this was especially pertinent given
our previous work which had clearly showed racemization
adjacent to the alcohol with all carbon substituents.[4c]
We began with the hydrogen-borrowing reaction between
Ph* methyl ketone 2 and commercially available (S)-N-
benzyl-L-prolinol 1a, using conditions developed within our
group for reactions of Ph* ketones: [(cod)IrCl]2 A (2 mol%),
dppBz (4 mol%) and KOH (2.0 equiv.) in PhMe (1.0 M) at
858C for 16 hours (Table 1, entry 1).[4] Whilst we were
delighted to see the formation of the desired product 3a,
both the yield (20%) and e.r. (77:23) were unsatisfactory.
Inspired by recent developments in hydrogen borrowing
utilizing earth-abundant catalysts,[11] we screened a range of
transition metal complexes. Whilst Ru-MACHO catalyst B
led to an improvement in both yield (41%) and e.r. (81:19), its
manganese analogue C resulted in only decomposition of the
starting alcohol (Table 1, entries 2 and 3). The use of a mixture
of iron complex D and Me3NO (8 mol%) resulted in no
reaction, with prolinol 1a and Ph* methyl ketone both
appearing unchanged in the crude 1H NMR spectrum
(Table 1, entry 4). Upon switching to [Cp*IrCl2]2 E we
observed an increase in both yield (51%) and e.r. (82:18),
which was then used as a basis for further optimization
(Table 1, entry 5). Next, the reaction was conducted in the
absence of toluene solvent; this resulted in a further increase
in yield to 63%, with no change in e.r. (Table 1, entry 6).
Upon increasing the reaction temperature to 1108C, we
observed product formation in a similar 62% yield accom-
panied by a large drop in e.r. to 65:35 (Table 1, entry 7). As
a corollary, lowering the temperature to 658C did result in an
increase in e.r. to 88:12, but at the expense of conversion with
the yield dropping to 49% (Table 1, entry 8). In an attempt to
prevent deprotonation of the acidic a-protons, and preserve
enantioenrichment, we screened the more hindered alkoxide
bases KOtBu and NaOtBu. Although this modification
resulted in an increase in product e.r. to 90:10 and 96:4,
respectively, it was accompanied by marked decreases in yield
to 18% and 10% (Table 1, entries 9 and 10). Despite the low
yield, the high e.r. encouraged us to persist with NaOtBu and
we discovered that decreasing the equivalents of base first to
equimolar, and then to sub-stoichiometric amounts resulted
in a dramatic increase in yield to 63%—with only a slight
reduction in e.r. to 92:8 (Table 1, entries 11 and 12). At this
point it was observed that, in the absence of a solvent, the
change in base from KOH to NaOtBu was preventing full
mixing of reactants. A range of concentrations of tBuOH as
solvent were therefore screened, with a 2.5 M solution giving
the final optimized yield and e.r. of 72% and 94:6, respec-
tively (Table 1, entries 13, 14 and 15).
Table 1: Optimization of a 1,2-amino alcohol hydrogen-borrowing reac-
tion using (S)-N-benzyl-L-prolinol 1a as a model substrate.
With optimized conditions in hand, we set out to evaluate
the reactivity of other amino acid-derived alcohols. We were
therefore surprised to find that although (S)-N,N-dibenzyl-L-
alaninol 4b gave the corresponding N,N-dibenzyl ketone 6b
in 70% yield, the e.r. of the product was only 73:27. To probe
the effect of sterics on the reaction, we examined smaller and
larger nitrogen protecting groups for both alaninols 4a–c and
d2-glycinols 5a–c (Table 2).
In both cases, we saw a clear trend between increasing
steric bulk of the nitrogen protecting group and a decreasing
degree of intermediate deprotonation—indicated either by
racemization in products 6a–c, or by loss of deuterium
incorporation in d2-glycine derivatives 7a–c. The use of
a triphenylmethane (trityl, Tr) protecting group yielded the
best results—with alanine and d2-glycine-derived products 6c
and 7c being formed in 79% yield, 96:4 e.r. and 71% yield,
> 95% D, respectively. There are several advantages that
accompany the use of a Tr protecting group: i) prevention of
epimerization; ii) low cost of the TrCl precursor; iii) ease of
removal with either mild acid or hydrogenolysis.[12] With this
modification of the protecting group, we then embarked on an
evaluation of other amino acid-derived alcohols under our
optimized reaction conditions (Scheme 2).
Entry [M] Base
(equiv)
T
Solvent
Yield[b] e.r. 3a[c]
[8C] (M)
1
A[d] KOH (2.0)
85
85
85
85
85
85
PhMe (1.0 M)
20
41
77:23
81:19
2
3
4
B
C
KOH (2.0)
KOH (2.0)
PhMe (1.0 M)
PhMe (1.0 M)
PhMe (1.0 M)
PhMe (1.0 M)
no solvent
[e]
[e]
–
–
D[f] KOH (2.0)
–
–
[g]
[g]
5
6
7
8
E
E
E
E
E
E
E
E
E
E
E
KOH (2.0)
KOH (2.0)
KOH (2.0)
KOH (2.0)
KOtBu (2.0)
NaOtBu (2.0) 85
NaOtBu (1.0) 85
NaOtBu (0.5) 85
NaOtBu (0.5) 85
NaOtBu (0.5) 85
NaOtBu (0.5) 85
51
63
62
49
18
10
42
63
82:18
82:18
65:35
88:12
90:10
96:4
110 no solvent
65
85
no solvent
no solvent
no solvent
no solvent
no solvent
tBuOH (0.5 M) 50
tBuOH (1.0 M) 46
tBuOH (2.5M) 72
9
10
11
12
13
14
15
95:5
92:8
92:8
95:5
94:6
The alcohols derived from the hydrophobic amino acids
glycine, alanine, leucine, and phenylalanine were well toler-
ated—with good to excellent yields and high product e.r. (3b,
3c, 3d and 3e). The reaction proved consistent upon scale-up
and pleasingly, the formation of 3c, 3d and 3k could be
carried out on a gram scale. In contrast, valine-derived
[a] All reactions were performed on a 0.4 mmol scale. [b] Isolated yield.
[c] Determined by normal phase HPLC analysis using a chiral stationary
phase. [d] With 4 mol% dppBz. [e] Complex mixture formed. [f] With
8 mol% Me3NO. [g] No reaction. cod=1,5-cyclooctadiene; dppBz=1,2-
bis(diphenylphosphino)benzene; TMS=trimethylsilane; Cp*=penta-
methylcyclopentadienyl.
6982 www.angewandte.org ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 6981 –6985