Step-efficient access to chiral primary amines

Article abstract of DOI:10.1055/s-0032-1317589

Routes to enantioenriched amines are outlined that employ reductive amination and carbanion addition methods. The strategies require either one or two reaction steps from prochiral carbonyl compounds for the synthesis of the corresponding chiral primary amines. Georg Thieme Verlag Stuttgart New York.

Full text of DOI:10.1055/s-0032-1317589

FEATURE ARTICLE
▌153
feature article
Step-Efficient Access to Chiral Primary Amines
Chiral Primary Amines
Thomas C. Nugent,* Sofiya M. Marinova
Department of Chemistry, School of Engineering and Science, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany
Fax +49(421)2003229; E-mail: t.nugent@jacobs-university.de
Received: 02.10.2012; Accepted after revision: 23.10.2012
methods are step-economic: two reaction steps from a ke-
tone or aldehyde to a chiral primary amine. It is of further
note that we have recently developed a one-pot synthesis
of chiral primary amines from ketones.
Abstract: Routes to enantioenriched amines are outlined that em-
ploy reductive amination and carbanion addition methods. The
strategies require either one or two reaction steps from prochiral
carbonyl compounds for the synthesis of the corresponding chiral
primary amines.
Ti(Oi-Pr)4-based reductive amination
Key words: chiral amine, enantiopure amines, reductive amina-
tion, cuprate addition, imine reduction
NH2
NH2
RS
NH2
R
RL
RL= i-Pr; RS = n-Pr
RL= Ph; RS = n-Pr
n
R = t-Bu, c-Hex,
i-Pr, Ar, i-Bu
Introduction
n = 1, 2
Chiral primary amines are flexible building blocks that are
often relied upon for the synthetic design of pharmaceuti-
cal drugs and alkaloid natural products by academic
groups and process research chemists, while medicinal
chemists need them for biologically relevant diversifica-
tion of drug templates. Driven by these demands, the syn-
thesis of chiral amines (non-amino acid) has undergone a
broad transformation over the last twenty years.1 A litera-
ture search of methods for the synthesis of chiral amines
will identify many potentially suitable and diverse meth-
ods, but challenges remain, e.g. reaction step efficiency,
functional group compatibility, and a lack of access to
some chiral amine structural classes. From a larger per-
spective, a one-step method that will enable the stereogen-
ic intermolecular insertion of nitrogen into advanced
intermediates is desirable but not yet feasible.
Yb(OAc)3-based reductive amination
NH2
R
Ph
R =
n
2
n is an integer > 0
carbanion method
NH2
m
n
n ≠ m and are integers > 0
(all substituents are n-alkyl)
Figure 1 Accessible generic amine structural classes and some spe-
cific examples
The methods outlined here (Scheme 1) have a practical
advantage: their technical simplicity, but both require a
hydrogenolysis deprotection step. If low Pd/C loadings
(0.5 mol%) are desired for the hydrogenolysis step, a pres-
sure vessel capable of 15–20 bar (218–290 psi) H2 and
heating at 50–60 °C is required. If such an apparatus is not
available, then inordinately high Pd/C loadings (5–10
mol%) with 4 bar (60 psi) H2 at room temperature will
often suffice.2,3 A glovebox or specialized catalysts or
syntheses thereof are not required, instead only commod-
ity catalysts and reagents that are routinely used on an in-
dustrial scale are necessary.
Scope and Limitations
Figure 1 provides a summary of the amine structural
classes that can be accessed using the methods summa-
rized in this feature article. By ‘access’ we mean the chiral
primary amine can be arrived at in very good overall
yield, typically 60–80%, with good-to-high enantiomeric
excess (80–99%) from a prochiral ketone or aldehyde.
The synthetic strategies and most often used reaction con-
ditions for arriving at chiral primary amine products, such
as those shown in Figure 1, are given in Scheme 1. Most
of our research has focused on the reductive amination
strategy, but when this strategy has failed to provide
>80% ee, the carbanion strategy generally provided good-
to-excellent yields and enantiomeric excesses; in short the
methods complement one another. Importantly, both
The generic structures of Figure 1 do not address functional
group compatibility, but the following co-existing moi-
eties are possible: carbamates, urethanes, amides, aceton-
ides, ethers, silyl ethers, and sterically hindered esters.4
Further functional group compatibility needs to be ex-
plored, and limitations will certainly be encountered. For
example, titanium(IV) isopropoxide based reductive
amination methods will transesterify unhindered ester
moieties, e.g. methyl or ethyl esters, into isopropyl es-
SYNTHESIS 2013, 45, 0153–0166
Advanced online publication: 02.01.2013
DOI: 10.1055/s-0032-1317589; Art ID: SS-2012-Z0768-FA
© Georg Thieme Verlag Stuttgart · New York
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X

154
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
R1
Pd/C (0.5 mol%), MeOH
2 (15–20 bar), 55 °C
H2N
(R)-4
R2
H
aldehyde
starting
R1
Ph
NH2
H2N
Ph
materials
R2
Ph
Ph
N
H
O
O
Ti(Oi-Pr)4 (5 mol%), then
Ti(Oi-Pr)4 or Yb(OAc)3 (1.1 equiv)
(R,R)-3
R1
H
R2
carbanion strategy
reductive amination strategy
R1
R2
2
–78 °C, R2CuMgBr⋅MgBrI
Ni, Pd, or Pt, H2 (10 bar), 22 °C
1
N
H
R2
ketone
starting
materials
Ph
NH2
H2N
Ph
(S,S)-3
R1
Pd/C (0.5 mol%), MeOH
H2 (15–20 bar), 55 °C
H2N
(S)-4
R2
Scheme 1 Synthetic strategies for accessing chiral amines as illustrated in Figure 1
ters.4d,5 Regarding non-compatible functionality, haloaro- hold potential for incorporation into our reductive amina-
matics would be hydrogenolyzed under the reductive tion methodology.10
amination conditions,6 and we were surprised to find that
It is important to note that the reductive amination meth-
ods discussed here can be extended to the reaction of α-
chiral ketones with achiral amines. This is noteworthy be-
an aromatic fluoride underwent partial defluorination (6–
10%).7,8 Although not examined, the hydro-dehalogena-
tion problem would potentially extend to alkyl halides, al-
cause it is the first demonstration, to the best of our knowl-
though alkyl fluorides, in our limited experience, are
edge, of reductive amination allowing the stereochemical
integrity of an α-labile chiral ketone to be preserved. For
example, ketone (S)-5 (96% ee), an α-stereolabile chiral
tolerated.9 Finally, reaction conditions have been reported
that ameliorate the hydro-dehalogenation problem and
Biographical Sketches█
Thomas C. Nugent was product synthesis under the then entered the pharmaceu-
born in 1967 in London direction
of
Tomáš tical industry as a process
(UK) but raised in Califor- Hudlický in 1995. From research chemist, working
nia and Vermont (USA). He 1996–1997 he held a post- for DSM and Pfizer (San
attended Virginia Polytech- doctoral position at the Uni- Francisco, CA, USA). In
nic Institute and State Uni- versity of Liverpool (UK) 2004 he became an assistant
versity (Blacksburg, VA, investigating peptide-based professor of chemistry at
USA) and completed his epoxidation methodology Jacobs University (Bremen,
Ph.D. studies in natural with Stanley M. Roberts. He Germany).
Sofiya M. Marinova was School in Gabrovo, Bulgar- area of asymmetric synthe-
born in 1992 in Veliko Tar- ia. Currently, she is a second sis with a recent interest in
novo, Bulgaria. In the sum- year chemistry B.Sc. stu- organocatalysis, which she
mer of 2011, she received dent at Jacobs University is pursuing within the labo-
her high school diploma Bremen, Germany. Her re- ratory of Professor Nugent.
from Aprilov National High search interests are in the
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
155
ketone, was successfully reductively aminated to give di- isopropoxide based reductive amination to give bis[(R)-1-
amine 6 without loss of the initial stereointegrity. Thus, phenylethyl]amine (3a) in very good yield and high dia-
we were able to demonstrate the most efficient synthesis stereomeric excess (Figure 2), and it is predicted that sim-
of (2S,3S)-7 to date, a pivotal substance P receptor antag- ilar product profiles would be noted when examining
onist building block (Scheme 2).4a,11 Reductive amination para- or meta-substituted acetophenones. The reductive
of ketone (S)-5 using the known Brønsted acid methods amination of ortho-substituted acetophenones with di-
resulted in gross racemization.4a This general tactic, reac- verse electronic character (F, CF3, Me, OMe) was pur-
tion of 5 to give 6 (Scheme 2), should be amenable to the sued, and successfully led to the formation of N-(1-
reductive amination of other enantioenriched α-chiral ke- phenylethyl)-1-(o-substituted phenyl)ethanamines 3b–e
tones, e.g. hydroxy ketones, alkoxy ketones, etc., for more (Figure 2) in ≥95% diastereomeric excess with good yield
step-efficient chiral drug or natural product synthesis.
(67–97%). Importantly, products 3b–e (see general proce-
dure A and procedures 3b–e, experimental section) could
be smoothly converted into the corresponding primary
amines in ≥95% ee with good overall yield from the cor-
responding ketones (58–61%, ortho-OMe was 92%).7
This work represents one of less than a handful of reports
on the subject of how to access these hindered (ortho-sub-
stituted phenyl)ethanamines 4b–e (not shown).
O
O
L-tartaric acid
Ph
Ph
Ph
dynamic kinetic resolution
90% yield, 96% ee
N
N
Ph
Ph
(S)-5
rac-5
H
N
1. Ti(Oi-Pr)4, BnNH2, THF
Pt/C, H2 (4.1 bar), 12 h
Ph
2. crystallization: n-hexane/toluene
61% yield (from ketone)
99% ee, 99% de
N
N
Ph
Ph
6
H
3a
NH2
Ph
85% yield, 95% de
F
Pd/C, H2 (4.1 bar), 50 °C
OMe
MeOH, HCl or MeSO3H
95% yield, 99% ee
N
N
Ph
N
Ph
Ph
(2S,3S)-7
H
H
3b
3c
79% yield, 98% de
Scheme 2 Stereochemical preservation of an α-labile chiral ketone,
(S)-5, during reductive amination: synthesis of a key quinuclidine in-
termediate 7
97% yield, 98% de
CF3
N
Ph
N
Ph
The forward-going discussion examines the use of achiral
carbonyl compounds for chiral primary amine synthesis
(Scheme 1). Both strategies (reductive amination and
carbanion addition) proceed through secondary amines 3.
When mediocre diastereoselectivity is noted, salt forma-
tion, e.g. with phthalic acid or hydrogen chloride, fol-
lowed by crystallization allows high diastereomeric
enrichment of amines 3 with good-to-excellent yields.12 It
is important to note that the diastereomeric excess of the
secondary amines 3 and the corresponding enantiomeric
excess of the primary amines 4 are consistent for all meth-
ods examined.13,14
H
H
3e
3d
crude, 74% de
(67% yield, 95% de)a
crude, 88% de
(73% yield, >99% de)a
a After one crystallization, yield from starting ketone.
Figure 2 Titanium(IV) isopropoxide based reductive amination of
aryl methyl ketones
Concerning methyl alkyl ketone substrates, as the steric
bulk of the alkyl substituent is gradually reduced, e.g.
from cyclohexyl (3f: 98% de) to isobutyl (3h: 93% de),
high diastereomeric excess is maintained (Figure 3). Fur-
ther reduction in the steric bulk, e.g., to the 2-phenylethyl
moiety (product 3i, Figure 3) resulted in good diastereo-
meric excess (80% de), and yet further reduction, e.g. to
an n-hexyl substituent, led to mediocre diastereoselectivi-
ty (3j: 72% de). All of these products 3f–j were synthe-
sized using general procedure B (experimental section).
Within this substrate category, alkyl methyl ketones, only
pinacolone (tert-butyl methyl ketone) failed to provide the
desired product under the standard reaction conditions
(general procedure B). This limitation was overcome by
replacing the Raney nickel hydrogenation catalyst with
Pt/C, allowing formation of amine 3k with 87% de
(Figure 3).
Stoichiometric Titanium-Based Reductive Amina-
tion
The yields and diastereoselectivities for chiral amine
structures 3a–e arrived at using the titanium(IV) isopro-
poxide based reductive amination methods with prochiral
methyl aryl ketones are shown in Figure 2. Acetophenone
(1a) is not a practical substrate to study if the goal is the
corresponding primary amine, nonetheless it offers direct
insight into the stereoelectronic effects of an aromatic
substituent on the carbonyl carbon, and the starting point
for understanding aryl methyl ketone substrate diastereo-
selectivity. Acetophenone (1a) underwent titanium(IV)
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

156
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
HN
Ph
HN
Ph
N
Ph
N
Ph
N
Ph
H
H
H
3f
3g
3h
79% yield, 93% de
90% yield, 98% de
78% yield, 98% de
3m
crude, 89% de16
64% yield, 97% dea
Raney Ni
substrates
3l
76% yield, 87% de
Pt/C
N
Ph
N
Ph
substrate
H
H
3i
3j
90% yield, 72% de
HN
Ph
89% yield, 80% de
HN
Ph
76% yield, 87% de
3o
92% yield
89% de13
Pt/C used
N
H
Ph
3n
(Raney Ni failed)
82% yield, 92% de
3k
Pd/C
substrates
NH2
NH2
Figure 3 Titanium-based reductive amination of alkyl methyl ke-
tones
The following conclusions can be drawn and used to pre-
dict the diastereoselectivity of yet unexamined alkyl
methyl ketone substrates. When the alkyl group contains
α- or β-branching high product diastereomeric excess
(90–98%) will occur, γ-branching can be expected to pro-
vide good diastereomeric excess (~80%), while δ-branch-
ing or n-alkyl substituents will lead to mediocre
diastereomeric excess (~70%). Pinacolone (tert-butyl
methyl ketone) and similar quaternary carbon bearing
methyl ketones, e.g. 1-adamantyl, can be expected to give
very good diastereomeric excess (85–90%) when using
platinum on carbon as the heterogeneous hydrogenation
catalyst.
4p
4q
64% yield,b 76% ee
76% yield,b 92% ee
a After one crystallization, yield from starting ketone.
b Overall yield from starting ketone.
Figure 4 Less frequently examined chiral amines
Note that indan-1-one (five-membered-ring analogue of
α-tetralone) was not examined. It is subjectively predicted
that the 2-oxo variants of these five-, six-, and seven-
membered cyclic ketones, e.g. 2-tetralone, would result in
low diastereomeric excesses. It is important to note that no
Brønsted acid based reductive amination methods are ca-
pable of producing the products given in Figure 4, this
will be elaborated shortly.
The prior chiral amine examples (Figures 2 and 3) em-
ployed methyl ketone starting materials, but the method
can be extended to acyclic or cyclic ketones containing
one -branched and one unbranched carbonyl substituent,
and still maintain high diastereomeric excess (85–95%,
Figure 4).7,14,15 For these sterically hindered ketones, ele-
vated reaction temperatures and/or hydrogenation pres-
sures, and/or prestirring of the reaction components
before hydrogen are required, see 3l–o and 4p,q (experi-
mental section).16 Within this category, the cyclic alkyl
aryl ketone subcategory, -tetralone (1p) and benzosuber-
one (1q), was unique in that yields of the corresponding
secondary amines (3p and 3q, not shown) could not be
properly quantified because partial hydrogenolysis, for-
mation of 4p and 4q, was always noted during their reduc-
tive amination. To ameliorate this problem, the crude
reductive amination products of α-tetralone (1p) and ben-
zosuberone (1q) were hydrogenolyzed to convert their 3/4
product mixtures into one primary amine product 4 and
then the yields and enantiomeric excesses were deter-
mined.
From the combined body of examples delineated above
(Figures 2–4), it is, perhaps, self-evident that increases in
the difference in the steric bulk between the two ketone
substituents lead to increases in the diastereomeric excess,
while small differences in steric bulk between the two ke-
tone substituents lead to low diastereoselectivity. For ex-
ample, octan-3-one (not shown) provided the secondary
amine product with ~15% de, this is not surprising since
an ethyl group is being differentiated from an n-pentyl
group. [Note: the carbanion addition method, discussed
shortly, provides a very similar product, 3t, with 81% de
(Figure 7).] Using this combined information, qualitative
predictions can be made regarding the titanium(IV) isop-
ropoxide reductive amination method. For example, a ke-
tone bearing one β-branched substituent and one n-alkyl
substituent, e.g. 2-methylheptan-4-one, can be predicted
to give the secondary amine product with 75–80% de. Fi-
nally, the titanium-based reductive amination method has
been extended to the reductive amination of keto esters
(not shown).5
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
157
The workup for the titanium(IV) isopropoxide reductive
amination methods is not trivial and is now discussed.
Quenching these reactions with water or aqueous acid
(0.5–1.0 M HCl) leads to lower isolated yields, but treat-
ment with aqueous 1.0 M sodium hydroxide overcomes
this problem. Thus rigorous stirring under these basic
workup conditions are recommended for up to one hour to
allow the titanium alkoxide intermediates to collapse to ti-
tanium dioxide and in doing so liberate any associated
product. For base-sensitive substrates aqueous sodium
carbonate can be used,5 but the stirring period may need to
be increased. The produced titanium dioxide (fine partic-
ulate) represents a filtration challenge that slows or even
stops celite filtrations. For preparations of greater than 10
grams this presents a technical obstacle that would likely
prevent large-scale (industrial) applications; a creative so-
lution was recently introduced by the Department of
Chemical Development at Boehringer Ingelheim Pharma-
ceuticals.17 When working up reactions that employ stoi-
chiometric quantities of titanium(IV) isopropoxide they
add inexpensive N,N,N′,N′-tetrakis(2-hydroxyethyl)ethyl-
enediamine (EDTE) to convert all of the titanium species
into an organic soluble EDTE–Ti(IV) complex. This al-
lows high-yield recovery and removes the filtration prob-
lem. Use of this modified workup is highly recommended
when scaling up (>10 g) the reductive amination reactions
described here.
HN
Ph
HN
Ph
3j
3r
Yb: 86% yield, 87% de
Ti: 90% yield, 72% de
Yb: 81% yield, 85% de
Ti: 82% yield, 71% de
HN
Ph
HN
Ph
3i
3s
Yb: 87% yield, 89% de
Ti: 89% yield, 80% de
Yb: 80% yield, 80% de
Ti: 94% yield, 74% de
Figure 5 Stoichiometric ytterbium(III) acetate versus titanium(IV)
isopropoxide based reductive amination
Brønsted Acid Based Reductive Aminations
Brønsted acid mediated reductive amination with alkan-2-
ones occurs under mild conditions: ketone (1.0 equiv),
(R)- or (S)-phenylethylamine (1.1 equiv), acetic acid (20
mol%), methanol, Raney nickel, hydrogen (8 bar), 23
°C.19 For these unhindered substrates (n-alkyl methyl ke-
tones) similar yields and the same diastereomeric excess-
es are observed when using the titanium(IV) isopropoxide
method: ketone (1.0 equiv), (R)- or (S)-phenylethylamine
(1.1 equiv), titanium(IV) isopropoxide (1.25 equiv), vari-
ous solvents (e.g., hexane, CH2Cl2, THF, etc.), Raney
nickel, hydrogen (8 bar), 23 °C. The conclusion is that for
n-alkyl methyl ketones, the most cost-effective and simple
method is the Brønsted acid (acetic acid) based method, if
products are required with high diastereomeric excess the
ytterbium(III) acetate based method should be used.18
Increased Diastereoselectivity: Ytterbium(III) Ace-
tate Based Reductive Amination
Regarding alkyl methyl ketones, we noted earlier that al-
kyl carbonyl substituents containing no branching or γ- or
δ-branching, provided mediocre diastereoselectivity (71–
80%). For this niche subset of unhindered alkyl methyl
ketones, it was found that replacing titanium(IV) isopro-
poxide with ytterbium(III) acetate allowed fast reductive
aminations with elevated diastereoselectivity (80–89%)
(Figure 5 and the experimental section, general procedure
C).18 The most dramatic changes in diastereoselectivity
were observed for n-alkyl methyl ketones. For example,
reductive amination of octan-2-one with ytterbium(III)
acetate provided secondary amine 3j (Figure 5) with 87%
de, a 15% unit change in diastereomeric excess versus the
titanium(IV) isopropoxide method. It should be noted that
the same reductive amination albeit with a Brønsted acid,
results in mediocre diastereomeric excess (72%).18 We
speculate that the ytterbium(III) acetate facilitates in situ
cis-to-trans imine isomerization before the reduction oc-
curs. A likely route is via an in situ formed ytterbium-
based hemiaminal intermediate that could then undergo:
(1) pyramidal inversion at nitrogen, or (2) C–N σ bond ro-
tation (180°), before collapsing to the more stable trans
imine.18
When the alkyl methyl ketone substrate contains an α- or
β-branched alkyl group, the decision becomes more nu-
anced. For example, using the acetic acid method, isobu-
tyl methyl ketone (4-methylpentan-2-one) requires
elevated temperatures (50 °C), and cyclohexyl methyl ke-
tone requires elevated temperatures (50 °C) and increased
hydrogen pressure (20 bar) as does acetophenone [H2 (30
bar), 50 °C] to maintain reasonable yields and reaction
times under Brønsted acid catalysis.19 All of these sub-
strates are reductively aminated with titanium(IV) isopro-
poxide (1.25 equiv) at 23 °C and using a hydrogen
pressure of 8 bar (general procedure A). If the optimized
Brønsted acid methods are applied to the starting ketones
for the amine products shown in Figure 4, the yields range
between 0–25%, regardless of the combination of hydro-
gen pressure, temperature, and reaction time employed.19
In conclusion, the use of titanium(IV) isopropoxide has
greatly expanded the possible ketone substrates for reduc-
tive amination, while the Brønsted acid methods are re-
stricted to unhindered ketone substrates and always give
the same diastereoselectivity as the titanium(IV) isopro-
poxide method. It is interesting to note that using 10 mol%
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

158
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
ytterbium(III) acetate gives almost identical reaction dialkylcuprate complex, boron trifluoride etherate, and
product profiles as the acetic acid method.18
low reaction temperatures (–78 °C); but the approach does
conveniently provide amine structures 3 from prochiral
aldehydes 2 in one-pot (Scheme 1 and Figure 7).21
Sequential Reductive Amination–Hydrogenolysis:
One-Pot Synthesis of Primary Chiral Amines
Ph
HN
H
Using hydrogen to reduce the in situ formed imine during
reductive amination results in a simple question: Can the
hydrogenolysis catalyst be added after the reductive ami-
nation is complete and permit a one-pot prochiral ketone
to chiral primary amine synthesis? To our surprise, prior
literature examples of one-pot sequential reductive ami-
nation–hydrogenolysis were not found.20
3t
81% yield, 81% de
Ph
HN
H
Hydrogenolyses are overwhelmingly performed with pal-
ladium on carbon in an alcohol solvent, specifically meth-
anol, while our reductive aminations are performed in
non-alcoholic solvents [Ti(Oi-Pr)4 method] or in mixtures
thereof [Yb(OAc)3 method]. Regardless, in 2005 we add-
ed palladium on carbon at the end of a reductive amina-
tion, and under various conditions of catalyst loading,
temperature, and hydrogen pressure, we only observed
mixtures of the secondary and primary amine products; in
short, we failed. In 2010 we noted that hydrogenolysis re-
actions of secondary amines 3 readily occurred in isopro-
pyl acetate. We then examined isopropyl acetate for its
suitability as a reductive amination solvent (reaction rate
and diastereoselectivity). Once this was positively con-
firmed, the one-pot reaction became reasonably trivial to
execute, Figure 6 provides the demonstrated examples
(general procedure D).7
3u
85% yield, 86% de
Ph
HN
H
3v
85% yield, 90% de
Figure 7 Carbanion-based amine synthesis
All of the acyclic reductive amination products noted in
Figures 2–6 can be synthesized using the carbanion meth-
od, but it would only make sense to do so when high en-
antiomeric excess is provided and the reductive amination
method simultaneously provides, or can be predicted to
provide, low enantiomeric excess (0–70%). We have con-
sequently limited our carbanion discussion to amine struc-
tures that cannot be achieved in useful enantiomeric
excess via reductive amination, e.g., nonan-3-amine and
undecan-5-amine. The carbanion method provides the
precursor secondary amines 3t and 3u for these primary
amines with good diastereomeric excess (81% and 86%,
respectively, Figure 7). The diastereoselectivity noted for
structure 3v (90% de, Figure 7) would be predicted to ex-
ceed that via reductive amination.22 General procedure E
describes the synthesis of 3t, 3u, and 3v.21
OMe NH2
CF3 NH2
NH2
4e
4b
4d
72% yield, 96% ee
60% yield, 90% ee
70% yield, 71% ee
NH2
NH2
NH2
Ph
5
4m
4i
4j
76% yield, 89% ee
70% yield, 85% ee
88% yield, 82% ee
The conclusion here is that good-to-high diastereomeric
excess can be achieved when only small differences in the
steric bulk exist between the two α-substituents, e.g. in 3u
(Figure 7) between an n-hexyl and an n-butyl moiety. This
contrasts dramatically with the reductive amination meth-
od which would provide essentially no diastereoselectivi-
ty. The outlined carbanion addition method should also be
amenable to the use of arylcuprates. If correct, this would
be important because high diastereomeric excess could be
obtained for a difficult to access category of chiral amines,
i.e., those with two large α-substituents. For example, two
different aromatic moieties (one from the cuprate and one
from an aromatic aldehyde) could represent RL and RS in
Scheme 3.
Figure 6 Examples of one-pot chiral amine synthesis
The Carbanion Method: A Complementary
Approach to Amine Structures 3
A strong undercurrent in chemistry is the development of
more environmentally friendly ‘green’ methods. In this
context, the reductive amination methods outlined above
go in the correct direction, e.g., by using only a slight ex-
cess (1.1–1.25 equiv) of the nitrogen source, (R)- or (S)-
phenylethylamine, and molecular hydrogen. The carban-
ion method (Scheme 1) presented here would not be con-
sidered green, it requires three equivalents of a
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
159
Scheme 3 Stereocontrol: reductive amination and carbanion addition methods
nically simple methods are produced, the chiral amine
methods offered here will be valuable because of their re-
action step-efficiency, simplicity, exclusive use of com-
Stereocontrol (Reductive Amination and Carbanion
Addition)
For both the reductive amination and carbanion addition mercially available catalysts and reagents, and robustness.
methods, the stereocontrol is dictated by the geometry of
This summary was written to provide an overall perspec-
the in situ formed chiral ketimine (reductive amination) or
tive of the strengths and weaknesses of the presented
aldimine (carbanion addition) intermediates (Scheme 3).
methods, and Figure 1 and Scheme 1 provide the overall
strategy and accessible amine categories. This should en-
able researchers with short timelines (number of reaction
The ketimines will exist predominately in the trans form
and hydrogen will add from the face containing the meth-
yl group of the auxiliary. As the steric bulk between the
steps and overall yield of lower importance: medicinal
two carbonyl substituents (starting reductive amination
chemists) or those requiring step-efficiency (number of
ketone) becomes less pronounced, more of the cis isomer
reaction steps and overall yield of high importance: pro-
cess research chemists) to reach quick project decisions
regarding the applicability, or lack thereof, for the pre-
sented methods.
(not shown) will be present, and lower diastereomeric ex-
cess will result for the secondary amines 3. For the aldi-
mines, the trans form is expected to dominate regardless
of the alkyl or aryl moiety residing on the starting alde-
The cumulative examples shown here (Figure 2–6) dem-
hyde, and the carbanion will attack from the less hindered
onstrate that greatly increased substrate breath (and at
‘methyl’ face. A second, but subordinate, factor influenc-
times stereoselectivity) can now be achieved, and this rep-
ing the stereoselectivity of carbanion addition is the steric
resents a significant advance for the general field of re-
bulk of the carbanion. Increasing the size of the carbanion
ductive amination. With this perspective, reductive
results in increased diastereoselectivity. It is readily ap-
amination should find a greater role in the design of natu-
ral products and pharmaceutical drugs, and we have al-
preciated that using the carbanion method, either diaste-
reomer of the secondary amine 3 can be selectively
formed as the major product using only one configuration
ready implemented these methods to access new
organocatalyst templates.23 Finally, where reductive ami-
of the auxiliary. This is accomplished by switching
nation substrate limitations exist, the carbanion approach
whether the carbanion or the carbonyl substituent on the
may be a valuable alternative.
starting aldehyde is RL or RS as noted in Scheme 3.
General procedures A–E and procedures for 3a–e,i,k–o,r,s–v and
4b,d,e,i,j,m–p are detailed below. Please note the chiral auxiliary
used here is invariably referred to as (R)- or (S)-phenylethylamine
Conclusion
or (R)- or (S)-α-methylbenzylamine (α-MBA). Compounds that
were excluded are synthesized using the general methods as noted
in the discussion section of this manuscript. Please note that all
Forward-looking methods for chiral amine synthesis are
currently being developed,1 but until optimized and tech-
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

160
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
compounds discussed here have been previously described and
characterized by either our group or other research groups. We char-
acterized the following for the first time: 3k, 3l, 3n, 3o, 3t, 3u, 3v,
4p (trifluoroacetamide), 4q (trifluoroacetamide). When we were not
the first group to characterize a compound, we nonetheless have
bed of celite, and the celite subsequently washed with MeOH
(3 × 15 mL). The combined filtrates were then evaporated to dry-
ness (rotary evaporate at ≤25 °C for short periods due to the semi-
volatile nature of the secondary amine products 3), and CHCl3 (20
mL) and aq 1.0 M NaOH (15 mL) were added and this mixture
stirred for 90 min. After transferring to a separatory funnel, the
CHCl3 layer was removed, and the aqueous layer further extracted
with CHCl3 (3 × 15 mL). The combined CHCl3 extracts were fil-
tered through a small bed of celite (removes turbidity) and the celite
subsequently washed with CHCl3 (2 × 15 mL). The filtrate was then
washed with sat. NH4Cl (2 × 20 mL) [removes residual small quan-
tities of α-MBA], then with brine (1 × 20 mL), dried (MgSO4), fil-
tered, and evaporated to dryness (rotary evaporate at ≤25 °C for
short periods) to obtain the crude product (this material is used to
determine the de). Purification by chromatography (silica gel) was
performed as noted in general procedure A.
provided the 1H NMR text data (and often, but not always, the 13
C
NMR text data) and have always included a copy of the chromato-
graphic (HPLC or GC) data and the NMR spectra in the supporting
information of the original manuscripts.
Titanium(IV) Isopropoxide Based Reductive Amination; Gen-
eral Procedure A
In an anhydrous solvent (0.50–0.83 M), a prochiral ketone 1 (2.50–
5.00 mmol), Ti(Oi-Pr)4 (1.25 equiv), and (R)- or (S)-α-MBA (1.10
equiv) were combined and stirred at r.t. for 30 min at r.t. A hetero-
geneous catalyst, Raney Ni [100 wt%, pre-triturated with EtOH (2
×) and then the anhydrous reaction solvent (2 ×)], Pt/C (0.30 mol%),
or Pd/C (0.23 mol%) was then added and the vessel pressurized with
H2 [120 psi (8.3 bar)] and stirred at r.t. At complete conversion
(<4.0 area% of ketone and imine by GC, see specific products for
exact reaction times), the mixture was worked up by transferring it
to an Erlenmeyer flask with EtOAc (~50 mL) or CH2Cl2 (~50 mL)
and rigorously stirring it with aq 1.0 M NaOH (20 mL) for 1 h. [For
an alternative workup allowing scale-up beyond 15 g, see ref.17].
The heterogeneous mixture was then filtered through a bed of celite
and the celite subsequently washed with EtOAc or CH2Cl2. The fil-
trate was concentrated (rotary evaporator bath temperature ≤ 25 °C)
to remove the low boiling organics and the remaining aqueous soln
was then extracted with CH2Cl2 (3 × 15 mL). Note: the temperature
of the rotary evaporator bath was maintained at or below 25 °C
when working with the free amine products 3 to suppress the loss
(yield reduction) of these semi-volatile compounds. The combined
organic extracts were dried (Na2SO4), filtered, and concentrated (ro-
tary evaporator bath, T ≤ 25 °C). Diastereomeric excesses were then
determined by analysis of the crude product. The isolated yield of
the combined diastereomers was determined after flash chromatog-
raphy using the following protocol. Due to the high volatility of
some of the amine products 3, the desirable column fractions were
combined in a round-bottom flask and aq 4.0 M HCl (~3.0 equiv)
was added. After brief but vigorous stirring (<10 min), the soln of
amine diastereomeric hydrochloride salts was rotary evaporated
(rotary evaporator bath can now be as high as 50 °C because the
amine salts are not volatile). The resulting solid or viscous liquid
was high vacuum dried until a constant weight was achieved. The
major diastereomer, (R,R) or (S,S), was isolated in pure form only
after further careful flash chromatography of the free amine diaste-
reomeric mixture.
Pretreatment of Yb(OAc)3
After using several bottles of Yb(OAc)3, which is sold and de-
scribed as a semihydrate (Sigma-Aldrich catalog no. 544973), we
noted that it was sometimes free-flowing while other bottles from
the same production lot were not. As a consequence, before using
Yb(OAc)3 for any of our research, 10-gram quantities were first
high vacuum dried at 80 °C to constant weight (~12 h). Yb(OAc)3
treated in this way was used for the optimal results shown in general
procedures C and D. The dried Yb(OAc)3 could be stored in a dry
screw cap glass bottle at r.t., and this container could be repeatedly
opened to the atmosphere for short periods of time (at least 6 times
without detrimental effect) and the desired quantity of Yb(OAc)3
weighed out without the need for a glovebox. In this way, constant
and repeatable results were always observed.
One-Pot Synthesis of Enantioenriched Primary Amines; Gener-
al Procedure D
In a reaction vessel under N2, Yb(OAc)3 (2.2 mmol, 1.1 equiv) [see
note about Yb(OAc)3 as found in general procedure C] or Ti(Oi-Pr)4
(2.4 mmol, 1.2 equiv) was added, followed by the solvent (2.0 mL,
1.0 M), ketone (2.0 mmol, 1.0 equiv), and (S)-phenylethylamine
(2.4 mmol, 1.2 equiv). This was stirred for 30 min, at r.t., and then
a reaction solvent slurry (2.0 mL) of Raney Ni [100 wt% based on
the ketone, triturated first with EtOH (3 × 2 mL) and then with the
reaction solvent (3 × 2 mL)] was added. Alternatively, when
Pd/Al2O3 (1.5 mol%) was the heterogeneous hydrogenation cata-
lyst, additional reaction solvent (2.0 mL) was added because a slur-
ry was not used to add the Pd/Al2O3. [For clarity: the final molarity
for all reactions was 0.5 M. Note: all hydrogenation catalyst wt% or
mol% numbers are based on the starting ketone.] The reaction ves-
sel was then pressurized with H2 (10–20 bar) at 24 °C, 35 °C, or 50
°C, depending on the specific substrate, with stirring. After stirring
for 12–48 h, <4 area% of the ketone substrate and imine intermedi-
ate (combined) remained by GC analysis, and Pd/C (1.0 mol%) or
Pd(OH)2 (1.0 mol%) was added. The reaction was then heated at 50
or 55 °C under H2 (20 bar). Importantly, the reaction was worked up
when 5–6 area% of the secondary amine starting material still re-
mained, which was most often at 48 h. Workup: dilute with EtOAc
(40 mL) and add aq 1.0 M NaOH (30 mL). After vigorous stirring
for 4 h in a septa-capped Erlenmeyer flask, the heterogeneous two-
phase soln was filtered through a bed of celite, and the celite subse-
quently washed with EtOAc (3 × 10 mL). [Note: the primary
amines are semi-volatile to volatile, do not aspirate the filtration
step more than required.] The aqueous layer was further extracted
with EtOAc (3 × 20 mL), and the combined organic layers were
then treated with aq 4.0 M HCl (~4 equiv). This was concentrated
until a volume of ~0.5 to 1.0 mL and generally had a yellow color.
If the chemical purity of the product was not >95 area% (GC), then
extractive purification with CHCl3 and CH2Cl2 was helpful. Dis-
tilled H2O (50 mL) and CHCl3 (50 mL) were added and the CHCl3
layer was further extracted with distilled H2O (2 × 10 mL). To the
combined aqueous layers, CH2Cl2 (50 mL) was added, and then the
aqueous phase was made strongly basic with NaOH (pH >11) with
Titanium(IV) Isopropoxide Based Reductive Amination; Gen-
eral Procedure B
General procedure A was used, with the following refined points:
Heterogeneous hydrogenation catalyst: Raney Ni (70–100 wt%);
anhydrous solvent: hexane, CH2Cl2, THF, or t-BuOMe; reaction
times: 6–11 h.
Reductive Amination with Stoichiometric Ytterbium(III) Ace-
tate; General Procedure C
In a dry reaction vessel, Yb(OAc)3 (0.96 g, 2.75 mmol, 1.1 equiv)
was added and subsequently evacuated under high vacuum for 5
min before flooding with N2, anhyd MeOH (2.5 mL, 1.0 M) was
then added. To this soln a prochiral ketone 1 (2.5 mmol, 1.0 equiv)
and (S)-α-MBA (0.35 mL, 2.75 mmol, 1.1 equiv) were added and
the mixture was subsequently stirred at r.t. for 20–30 min. A THF
slurry of Raney Ni [100 wt% based on the ketone, pre-triturated
with EtOH (3 ×) and then with anhyd THF (3 ×) before addition]
was transferred to the mixture using anhyd THF (2.5 mL, final mo-
larity of reaction soln was 0.5 M) and the reaction vessel was pres-
surized with H2 [120 psi (8.3 bar)]. After 12 h at 22 °C, <3 area% of
the ketone and imine intermediate remained by GC analysis, and the
mixture was then diluted with MeOH (15 mL), filtered through a
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
161
4.0 M NaOH. After removal of the CH2Cl2 layer, the basic aqueous
layer was further extracted with CH2Cl2 (2 × 30 mL). To the com-
bined CH2Cl2 layers was added aq 4.0 M HCl (~3 equiv) and this
was concentrated (rotary evaporator, bath temperature initial: 30
°C; final: ~70 °C) providing an off-white solid that was dried under
high vacuum until a constant weight was achieved (usually over-
night). To assess the enantiomeric excess of the product, a sample
of the primary amine HCl salt (1 mmol) was neutralized (aq NaOH–
EtOAc) and then converted into the corresponding trifluoroacet-
amide, analysis using GC allowed the enantiomeric excess to be de-
termined.
ea% (GC). Purification by flash chromatography (silica gel,
hexanes–EtOAc–NH4OH, 83:15:2); yield: 79%. Attempts to crys-
tallize this crude product failed because the defluorinated byproduct
could not be fully removed. GC methods and product retention
times are available.7
(S,S)-3c
Previously characterized; free base.
1H NMR (400 MHz, CDCl3): δ = 7.3–7.0 (m, 9 H), 3.76 (q, J = 6.7
Hz, 1 H), 3.51 (q, J = 6.6 Hz, 1 H), 1.62 (br s, NH, 1 H), 1.31 (d, J =
6.7 Hz, 3 H), 1.28 (d, J = 6.7 Hz, 3 H).
Bis[(R)-1-phenylethyl]amine (3a)
(S,S)-3c·Phthalic Acid Salt
1H NMR (400 MHz, CDCl3): δ = 8.5–6.9 (m, 13 H), 4.45 (q, J = 6.7
Hz, 1 H), 4.12 (q, J = 6.6 Hz, 1 H), 1.81 (d, J = 6.8 Hz, 3 H), 1.75
(d, J = 6.8 Hz, 3 H).
Following general procedure A using EtOAc (5.0 mL, 0.5 M),
Ti(Oi-Pr)4 (3.13 mmol, 1.25 equiv), acetophenone (1a, 2.50 mmol,
1.0 equiv), (S)-α-MBA (2.75 mmol, 1.1 equiv), Raney Ni (300 mg,
100 wt%); H2 (8.3 bar), 24 °C, 8 h. Flash chromatography (EtOAc–
heptane–NH4OH, 25:74:1) gave the mixture of diastereomers (480
mg, 85%, 95% de) as a colorless viscous oil. GC methods and prod-
uct retention times are available.14
13C NMR (100 MHz, CDCl3): δ = = 171.8, 161.4, 158.9, 136.4,
134.2, 133.5, 131.4, 130.7, 130.6, 129.3, 129.1, 128.1, 125.4, 124.5,
124.3, 121.5, 116.0, 115.8, 57.5, 49.5, 49.4, 21.2, 20.8.
(S)-1-Phenyl-N-{(S)-1-[2-(trifluoromethyl)phenyl]ethyl}ethan-
amine (3d)
(R,R)-3a
Previously characterized; free base.
Following general procedure A using i-PrOAc, Ti(Oi-Pr)4 (3.2
mmol, 1.6 equiv), 1-[2-(trifluoromethyl)phenyl]ethanone (1d, 2.0
mmol, 1.0 equiv), (S)-α-MBA (2.4 mmol, 1.2 equiv), Pd/Al2O3 (1.5
mol%); H2 (15 bar), 35 °C, 24 h. Crude product: 88% de (1H NMR).
Crude product was purified via crystallization with phthalic acid
(1.0 equiv) in i-PrOH (4 volumes) and heptane (4 volumes); yield:
73% (from 1d); >99% de (1H NMR). GC analysis did not allow sep-
aration of these diastereomers. Further product information is avail-
able.7
1H NMR (400 MHz, CDCl3): δ = 7.21–7.35 (m, 10 H), 3.49 (q, J =
6.6 Hz, 1 H), 1.55 (br s, 1 H), 1.27 (d, J = 6.6 Hz, 6 H).
(S)-1-(2-Methoxyphenyl)-N-[(S)-1-phenylethyl]ethanamine
(3b)
Following general procedure A using i-PrOAc, Ti(Oi-Pr)4 (4.75
mmol, 1.25 equiv), 1-(2-methoxyphenyl)ethanone (1b, 3.0 mmol,
1.0 equiv), (S)-α-MBA (3.3 mmol, 1.1 equiv), Raney Ni (100 wt%);
H2 (10 bar), 24 °C, 16 h. After workup and high vacuum drying, the
product was determined to be of high chemical purity (GC and 1H
NMR, ≥95%), no column chromatography or crystallization was re-
quired; yield: 97%; 98% de. GC methods and product retention
times are available.7
(S,S)-3d
Previously characterized; free base.
1H NMR (400 MHz, CDCl3): δ = 7.9–7.2 (m, 9 H), 4.07 (q, J = 6.4
Hz, 1 H), 3.49 (q, J = 6.7 Hz, 1 H), 1.68 (br, 1 H), 1.24 (d, J = 6.7
Hz, 6 H).
(S,S)-3b
Previously characterized; free base.
(S,S)-3d·Phthalic Acid Salt
1H NMR (400 MHz, CDCl3): δ = 7.4–6.8 (m, 9 H), 3.77 (m, 4 H),
3.54 (q, J = 6.6 Hz, 1 H), 2.18 (br, 1 H), 1.32 (d, J = 6.7 Hz, 3 H),
1.29 (d, J = 6.7 Hz, 3 H).
1H NMR (400 MHz, CDCl3): δ = 8.5–7.3 (m, 13 H), 4.55 (q, J = 6.5
Hz, 1 H), 4.23 (q, J = 6.7 Hz, 1 H), 1.79 (d, J = 6.8 Hz, 3 H), 1.73
(d, J = 6.6 Hz, 3 H).
13C NMR (100 MHz, CDCl3): δ = 157.6, 146.1, 133.5, 128.3, 128.0,
127.6, 126.9, 126.7, 120.7, 110.8, 55.4, 55.2, 51.6, 25.3, 22.9.
13C NMR (100 MHz, CDCl3): δ = 171.8, 136.7, 134.1, 133.5, 131.4,
129.3, 129.2, 128.6, 128.5, 127.8, 57.8, 53.0, 22.6, 21.7.
(R,S)-3b
(S)-N-[(S)-1-(2-Methylphenyl)ethyl]-1-phenylethanamine (3e)
Following general procedure A using i-PrOAc, Ti(Oi-Pr)4 (3.2
mmol, 1.6 equiv), 1-(2-methylphenyl)ethanone (1e, 2.0 mmol, 1.0
equiv), (S)-α-MBA (2.4 mmol, 1.2 equiv), Pd/Al2O3 (1.5 mol%); H2
(15 bar), 35 °C, 24 h. Crude product: 74% de. The crude product
was purified via crystallization with phthalic acid (1.0 equiv) in
i-PrOH (3 volumes) and heptane (4 volumes); yield: 67% (from 1e);
95% de (GC). A second crystallization provided; yield: 48% (from
1e); 99% de. GC methods and product retention times are avail-
able.7
Minor diastereomer; free base.
1H NMR (400 MHz, CDCl3): δ = 7.3–6.8 (m, 9 H), 4.1 (q, J = 6.8
Hz, 1 H), 3.77 (s, 3 H), 3.69 (q, J = 6.7 Hz, 1 H), 1.65 (br, 1 H), 1.36
(d, J = 6.7 Hz, 3 H), 1.33 (d, J = 6.7 Hz, 3 H).
(S)-N-[(S)-1-(2-Fluorophenyl)ethyl]-1-phenylethanamine (3c)
Pt/C Results
Following general procedure A using i-PrOAc, Ti(Oi-Pr)4 (2.5
mmol, 1.25 equiv), o-fluoroacetophenone (1c, 2.0 mmol, 1.0
equiv), (S)-α-MBA (2.2 mmol, 1.1 equiv), Pt/C (0.5 mol%); H2 (20
bar), 35 °C, 24 h. Crude product (40% de) and the defluorinated
compound, <0.80 area% (GC), was noted. Crude product 3c was
purified via crystallization with phthalic acid (1.0 equiv) in i-PrOH
(6 volumes), providing 3c phthalic acid salt; yield: 56% (from 1c);
99% de (GC). More details regarding the crystallization and the GC
methods and product retention times are available.7
(S,S)-3e
Previously synthesized; free base.
1H NMR (400 MHz, CDCl3): δ = 7.5–7.0 (m, 9 H), 3.78 (q, J = 6.7
Hz, 1 H), 3.50 (q, J = 6.8 Hz, 1 H), 1.98 (s, 3 H), 1.65 (br, 1 H), 1.32
(d, J = 6.4 Hz, 3 H), 1.22 (d, J = 6.5 Hz, 3 H).
(S,S)-3e·Phthalic Acid Salt
1H NMR (400 MHz, CDCl3): δ = 10.37 (br, 2 H), 8.5–7.0 (m, 13 H),
4.30 (q, J = 6.3 Hz, 1 H), 4.14 (q, J = 6.4 Hz, 1 H), 1.81 (d, J = 6.8
Hz, 3 H), 1.70 (d, J = 6.8 Hz, 3 H), 1.53 (s, 3 H).
Raney Ni Results
Following general procedure A using i-PrOAc, Ti(Oi-Pr)4 (2.5
mmol, 1.25 equiv), o-fluoroacetophenone (1c, 2.0 mmol, 1.0
equiv), (S)-α-MBA (2.2 mmol, 1.1 equiv), Raney Ni (100 wt%); H2
(20 bar), 35 °C, 24 h. Crude product (97% de); from multiple reac-
tions we noted the defluorinated byproduct was present in 6–10 ar-
13C NMR (100 MHz, CDCl3): δ = 171.9, 136.7, 135.9, 135.5, 134.3,
133.5, 131.4, 130.9, 129.3, 129.1, 128.4, 128.2, 127.5, 126.3, 57.0,
52.4, 21.3, 21.2, 18.5.
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

162
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
(2R)-3,3-Dimethyl-N-[(R)-1-phenylethyl]butan-2-amine (3k)
Following general procedure B using pinacolone (1k, 5.00 mmol),
(R)-α-MBA, (1.10 equiv), and Ti(Oi-Pr)4 (1.25 equiv) was pre-
stirred for 2 h (neat), then hexane (6.0 mL, 0.83 M) and Pt/C (60 mg,
0.3 mol%) [added in 3 equal portion, thus Pt/C (20 mg) at t = 0 h, 3
h, and 6 h] (20 mg) were added and pressurized with H2 [120 psi
(8.3 bar)]; 50 °C. Reaction time: 20 h. Crude product: 87% de. Flash
chromatography (cyclohexane–EtOAc–NH4OH, 95:1:4) gave a
mixture of diastereomers as a colorless viscous oil; yield: 0.78 g
(76%). The major diastereomer was separately obtained by further
flash chromatography using the same eluent system. GC methods
and product retention times are available.14
vated temperature was inadvertently dropped from the original ex-
perimental description7, but it is absolutely required and is
consistent with our earlier description15 for a similar substrate
(3l).] DCE (2 mL) was then added followed by a toluene (2 mL)
slurry of Raney Ni (100 wt%). [Note: the DCE allows high de to be
achieved, but this solvent did not allow a free-flowing slurry of
Raney Ni. After our earlier literature description for 3l, we found
that a toluene slurry of Raney Ni was excellent for transferring the
Raney Ni (large-gauge needle/syringe) and importantly did not low-
er the de. Use of toluene as the sole reaction solvent resulted in re-
duced de.] Pressurize with H2 (20 bar), 50 °C, 48 h. Crude product:
89% de (GC). The crude product was converted into 3m·HCl and
crystallized (cyclohexane, 16 volumes); yield: 64% (from 1m);
97% de (GC). GC methods and product retention times are avail-
able.7
(R,R)-3k
Free base; Rf = 0.47 (cyclohexane–EtOAc–NH4OH, 91:6:3).
IR (KBr): 3427 cm–1.
(S,S)-3m
1H NMR (400 MHz, CDCl3): δ = 7.19–7.36 (m, 5 H), 3.77 (q, J =
6.5 Hz, 1 H), 2.29 (q, J = 6.5 Hz, 1 H), 1.27 (d, J = 6.5 Hz, 3 H),
0.84–0.89 (m 12 H).
13C NMR (100 MHz, CDCl3): δ = 147.6, 128.2, 126.7, 126.6, 59.5,
57.0, 34.7, 26.5, 23.6, 15.9.
MS (EI): m/z (%) = 205.1 (M+, 16), 190.1 (24), 148.1 (76), 105.1
(100), 86.1 (12), 77, (8), 44 (52).
Free base; Rf = 0.50 (hexanes–EtOAc–NH4OH, 90:8:2).
1H NMR (400 MHz, CDCl3): δ = 7.31–7.19 (m, 5 H), 3.84 (q, J =
6.6 Hz, 1 H), 2.15 (m, 1 H), 1.87 (m, 1 H), 1.36–1.05 (m, 10 H),
0.87–0.79 (m, 9 H).
13C NMR (100 MHz, CDCl3): δ = 146.6, 128.2, 126.7, 126.6, 59.0,
55.2, 29.9, 29.0, 28.7, 24.8, 22.8, 18.9, 16.9, 14.0.
(1R)-2,2-Dimethyl-N-[(R)-1-phenylethyl]cyclopentanamine
(3n)
Anal. Calcd for C14H23N: C, 81.89; H, 11.29; N, 6.82. Found: C,
81.65; H, 11.13; N, 6.87.
The reaction was prepared according to general method A with the
following noted refinements: ketone 1n (5.00 mmol) prestirred for
3 h (neat) with (R)-α-MBA and Ti(Oi-Pr)4; then EtOH (6.0 mL, 0.83
M) was added; H2 [120 psi (8.3 bar)]; Pt/C (60.0 mg, 0.3 mol%)
[added in 4 equal portion, thus Pt/C (15 mg) at t = 0 h, 2 h, 4 h, and
8 h]; 50 °C. Reaction time: 20 h. Crude product: 92% de. Purifica-
tion by column chromatography (cyclohexane–EtOAc–NH4OH,
88:10:02) provided the mixture of diastereomers as a colorless vis-
cous liquid, which was then converted into the HCl salt; yield: 1.04
g (82%). The analytically pure major (R,R)-3n diastereomer was
obtained by further column chromatography (cyclohexane–EtOAc–
NH4OH, 90:8:2) of the diastereomeric free amine mixture. GC
methods and product retention times are available.14
(3R)-2-Methyl-N-[(R)-1-phenylethyl]hexan-3-amine (3l)
A mixture of 2-methylhexan-3-one (1l, 4.15 mL, 30.00 mmol, 1.00
equiv), (R)-α-MBA (4.07 mL, 31.5 mmol, 1.05 equiv), and Ti(Oi-
Pr)4 (97% grade, 26.6 mL, 90.0 mmol, 3.00 equiv) in anhyd DCE
(20 mL) was stirred at 900 rpm under argon at 40 °C for 2 h. Raney
Ni (2.8 g) suspended in DCE (10 mL) was added and this was hy-
drogenated at 120 psi (8.3 bar) with vigorous stirring (1200 rpm) at
60 °C. After 24 h, 13 area% of the starting material remained, and
after 35 h (3 area% of the starting material remaining) the mixture
was quenched with aq 2.5 M NaOH (30 mL). After stirring for 2 h,
the soln was diluted with CH2Cl2 (50 mL), filtered through celite,
and the celite was washed with CH2Cl2 (3 × 70 mL). The organic
layer was removed from the filtrate and the aqueous layer further
extracted with CH2Cl2 (2 × 20 mL). The combined organic extracts
were dried (Na2SO4), filtered, and evaporated to dryness and kept
under high vacuum to remove excess (R)-α-MBA. The resulting oil,
secondary amine 3l, was then treated with 4.0 M ethereal HCl (15
mL), filtered, and dried to afford 3l·HCl; yield: 5.82 g (76%); 87%
de).
(R,R)-3n
Free base; Rf = 0.64 (cyclohexane–EtOAc–NH4OH, 75:20:5) [mi-
nor diastereomer (S,R)-3n: Rf = 0.50].
IR (KBr): 3441, 2955, 2864, 1465, 1134, 700 cm–1.
1H NMR (400 MHz, CDCl3): δ = 7.34–7.20 (m, 5 H), 3.86 (q, J =
6.4 Hz, 1 H), 2.53 (m, 1 H), 1.78–1.33 (m, 6 H), 1.30 (d, J = 6.4 Hz,
3 H), 1.02 (s, 3 H), 0.86 (s, 3 H).
For a small-scale example, details of chromatography (silica gel),
GC methods, and product retention times are available.15
13C NMR (100 MHz, CDCl3): δ = 147.2, 128.2, 126.7, 126.6, 65.6,
57.2, 40.9, 39.9, 32.0, 28.1, 24.6, 21.1, 19.7.
MS (EI): m/z (%) = 217 (22) [M+], 202 (54), 160 (58), 146 (10), 105
(100), 56 (26).
(R,R)-3l
Free base; Rf = 0.51 (cyclohexane–EtOAc–NH4OH, 90:6:4) [minor
diastereomer (S,R)-3l, Rf = 0.40].
IR (KBr): 3432, 2957, 1462, 1360 cm–1.
Anal. Calcd for C15H23N: C, 82.89; H, 10.67; N, 6.44. Found: C,
83.04; H, 10.57; N, 6.50.
1H NMR (400 MHz, CDCl3): δ = 7.18–7.38 (m, 5 H), 3.84 (q, J =
6.5 Hz, 1 H), 2.16 (m, 1 H), 1.86 (m, 1 H), 1.26 (d, J = 6.5Hz, 3 H),
1.13 (m, 4 H), 0.85 (m, 3 H), 0.78 (m, 6 H).
13C NMR (100 MHz, CDCl3): δ = 146.5, 128.1, 126.7, 126.6, 58.7,
55.0, 32.4, 28.8, 24.8, 19.6, 18.9, 16.7, 14.2.
MS (EI): m/z (%) = 218 (6) [M+], 205 (18), 204 (62), 105 (100), 72
(79).
(1R)-1-Phenyl-N-[(R)-1-phenylethyl]butan-1-amine (3o)
Reaction details: The reaction was prepared according to general
method A with the following noted refinements: ketone 1o (5.00
mmol), (R)-α-MBA, and Ti(Oi-Pr)4 prestirred for 2 h in EtOAc
(10.0 mL, 0.5 M). Raney Ni (500 mg, 100 wt%) was added, pres-
surized with H2 (8.3 bar), and heated at 35 °C, reaction time: 15 h.
We originally reported a 94% de for this product14 but all subse-
quent research on this product (by our group) has shown this is not
the case and we have observed a range of de (84–94%), see ref. 13
for more details. The mixture of diastereomers was isolated as a col-
orless viscous liquid; yield: 1.17 g (92%). An analytically pure sam-
ple of the (R,R)-3o diastereomer was obtained by repeated
recrystallization of the HCl salt of the diastereomeric mixture
Anal. Calcd for C15H25N: C, 82.13; H, 11.49; N, 6.39. Found: C,
82.01; H, 11.40; N, 6.51.
(S)-2-Methyl-N-[(S)-1-phenylethyl]heptan-3-amine (3m)
Reaction details: Ti(Oi-Pr)4 (6.0 mmol, 3.0 equiv), 2-methylheptan-
3-one (1m, 2.0 mmol, 1.0 equiv), (S)-α-MBA (2.4 mmol, 1.2 equiv)
were prestirred neat at 50 °C for 4 h. [Note: this prestirring at ele-
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
163
(hexane–EtOH). GC methods and product retention times are avail-
able.14
cedures) and conversion into the hydrochloride salt gave 4e·HCl;
yield: 70%. The high chemical purity was confirmed by GC analy-
sis of the free amine, >96% purity. GC methods, product retention
times, and ee determination (trifluoroacetamide GC analysis) are
available.7
(R,R)-3o
Free base; Rf = 0.60 (cyclohexane–EtOAc–NH4OH, 15:83:02).
IR (KBr): 3445, 2958, 1602, 1492, 1453, 1368, 1125, 761, 700
cm–1.
1H NMR (400 MHz, CDCl3): δ = 7.34–7.15 (m, 10 H), 3.47 (q, J =
6.4 Hz, 1 H), 3.28 (t, J = 6.8 Hz, 1 H), 1.60 (m, 2 H), 1.51 (m, 1 H),
1.24 (d, J = 6.8 Hz, 3 H), 1.08 (m, 1 H), 0.77 (t, J = 7.2 Hz, 3 H).
(S)-4e·HCl
Previously characterized.
1H NMR (400 MHz, CDCl3): δ = 8.71 (br, 3 H), 7.6–7.1 (m, 4 H),
4.63 (br, 1 H), 2.35 (s, 3 H), 1.62 (br d, 3 H).
13C NMR (100MHz, CDCl3): δ = 145.8, 144.9, 128.3, 128.25,
127.2, 126.7, 59.8, 54.8, 41.0, 25.1, 19.6, 14.0.
MS (EI): m/z (%) = 253 (4) [M+], 238 (25), 210 (100), 105 (86), 91
(33), 77 (14).
(S)-4-Phenylbutan-2-amine (4i)
Following general procedure D. Reductive amination stage:
Yb(OAc)3 (2.2 mmol, 1.1 equiv) [see pretreatment procedure at the
end of general method C], anhyd MeOH (2.0 mL, 1.0 M), anhyd
THF [2.0 mL, 1.0 M (used for Raney Ni transfer)], 4-phenylbutan-
2-one (1i, 2.0 mmol, 1.0 equiv), (S)-α-MBA (2.2 mmol, 1.1 equiv),
stirred 20 min, then Raney Ni (100 wt%) was added, pressurized
with H2 (10 bar), 24 °C, 12 h. Hydrogenolysis stage: Open reactor,
Pd(OH)2 (1.0 mol%) was added, pressurized with H2 (20 bar), 50
°C, 48 h. Acid-base workup (see general procedures) and conver-
sion into the hydrochloride salt gave 4i·HCl; yield: 70%. The high
chemical purity was confirmed by GC analysis of the free amine,
>96% purity. GC methods, product retention times, and ee determi-
nation (trifluoroacetamide GC analysis) are available.7
Anal. Calcd for C18H23N: C, 85.32; H, 9.15; N, 5.53. Found: C,
84.75; H, 9.24; N, 5.48.
(S)-1-(2-Methoxyphenyl)ethanamine (4b)
Following general procedure D. Reductive amination stage: i-PrOAc
(4.0 mL, 0.5 M), 1-(2-methoxyphenyl)ethanone (1b, 2.0 mmol, 1.0
equiv), (S)-α-MBA (2.4 mmol, 1.2 equiv), Ti(Oi-Pr)4 (2.4 mmol,
1.2 equiv), stirred for 20 min, then Raney Ni (100 wt%) was added,
pressurized with H2 (10 bar), 24 °C, 24 h. Hydrogenolysis stage:
Open reactor, Pd/C (1.0 mol%) was added, then pressurized with H2
(20 bar), 48 h, 55 °C. Acid-base workup (see general procedures)
and conversion into the hydrochloride salt gave 4b·HCl; yield: 72%.
The high chemical purity was confirmed by GC (a small portion
was converted into the free amine and shown to be >95% pure (GC).
GC methods, product retention times, and ee determination (tri-
fluoroacetamide GC analysis) are available.7
(S)-4i·HCl
Previously characterized.
1H NMR (300 MHz, CDCl3): δ = 8.47 (br, 3 H), 7.26–7.16 (m, 5 H),
3.31 (br, 1 H), 2.76 (t, J = 7.8 Hz, 2 H), 2.13 (m, 1 H), 1.93 (m, 1
H), 1.43 (d, J = 6.4 Hz, 3 H).
(S)-Octan-2-amine (4j)
(S)-4b·HCl
Previously characterized.
Following general procedure D. Reductive amination stage:
Yb(OAc)3 (2.2 mmol, 1.1 equiv) [see pretreatment procedure at the
end of general method C], anhyd MeOH (2.0 mL, 1.0 M), anhyd
THF [2.0 mL, 1.0 M (used for Raney Ni transfer)], octan-2-one (1j,
2.0 mmol, 1.0 equiv), (S)-α-MBA (2.0 mmol, 1.0 equiv), stirred for
20 min, then Raney Ni (100 wt%) was added, pressurized with H2
(10 bar), 24 °C, 16 h. Hydrogenolysis stage: Open reactor, Pd/C (1.0
mol%) was added, pressurized with H2 (20 bar), 50 °C, 45 h. [Note:
The same result could also be obtained with Pd(OH)2 (1.0 mol%),
H2 (20 bar), 50 °C, 45 h]. Acid-base workup (see general proce-
dures), and conversion into the hydrochloride salt gave 4j·HCl;
yield: 88%. The high chemical purity was confirmed by GC analy-
sis of the free amine, >99% purity. GC methods, product retention
times, and ee determination (trifluoroacetamide GC analysis) are
available.7
1H NMR (400 MHz, CDCl3): δ = 8.61 (br, 3 H), 7.4–6.8 (m, 4 H),
4.68 (q, J = 6.8 Hz, 1 H), 3.81 (s, 3 H), 1.68 (d, J = 6.8 Hz, 3 H).
13C NMR (100 MHz, CDCl3): δ = 156.8, 130.1, 127.7, 125.7, 121.0,
110.7, 55.6, 47.3, 18.6.
(S)-1-[2-(Trifluoromethyl)phenyl]ethanamine (4d)
Following general procedure D. Reductive amination stage: i-PrOAc
(6.0 mL, 0.5 M), 2′-(trifluoromethyl)acetophenone (1d, 3.0 mmol,
1.0 equiv), (S)-α-MBA (3.6 mmol, 1.2 equiv), Ti(Oi-Pr)4 (4.8
mmol, 1.6 equiv), stirred for 20 min, then Pd/Al2O3 (1.5 mol%) was
added, pressurized with H2 (20 bar), 35 °C, 24 h. Hydrogenolysis
stage: Open reactor, Pd(OH)2 (1.0 mol%) was added, then pressur-
ized with H2 (20 bar), 50 °C, 48 h. Acid-base workup (see general
procedures) and conversion into the hydrochloride salt gave
4d·HCl; yield: 60%. The high chemical purity was confirmed by
GC analysis of the free amine, >98% purity. GC methods, product
retention times, and ee determination (trifluoroacetamide GC anal-
ysis) are available.7
(S)-4j·HCl
Previously characterized.
1H NMR (400 MHz, CDCl3): δ = 8.32 (br s, 3 H), 3.29 (m, 1 H), 1.78
(m, 1 H), 1.70 (m, 1 H), 1.41–1.28 (m, 11 H), 0.86 (t, J = 6.4 Hz, 3
H).
(S)-4d·HCl
13C NMR (100 MHz, CDCl3): δ = 48.6, 35.0, 31.6, 28.9, 25.6, 22.6,
18.8, 14.1.
Previously characterized.
1H NMR (400 MHz, CDCl3): δ = 8.93 (br, 3 H), 8.0–7.4 (m, 4 H),
4.83 (br, 1 H), 1.71 (br, 3 H).
13C NMR (100 MHz, CDCl3): δ = 136.8, 133.1, 129.0, 127.9, 127.8
(q), 126.3, 124.0 (q), 42.5, 21.7.
(S)-2-Methylheptan-3-amine (4m)
Following general procedure D. Reductive amination stage: Ti(Oi-Pr)4
(6.0 mmol, 3.0 equiv), 2-methylheptan-3-one (1m, 2.0 mmol, 1.0
equiv), (S)-α-MBA (2.4 mmol, 1.2 equiv) were prestirred neat at 50
°C for 4 h. [Note: this prestirring at elevated temperature was inad-
vertently dropped from the original experimental description,7 but
it is absolutely required and is consistent with our earlier
description15 for a similar substrate (3l).] DCE (2 mL) was then
added followed by a toluene (2 mL) slurry of Raney Ni (100 wt%).
[Note: the DCE allows high de to be achieved, but this solvent did
not allow a free-flowing slurry of Raney Ni. After our earlier liter-
ature description for 3l, we found that a toluene slurry of Raney Ni
(S)-1-(2-Methylphenyl)ethanamine (4e)
Following general procedure D. Reductive amination stage: i-PrOAc
(6.0 mL, 0.5 M), 1-(2-methylphenyl)ethanone (1e, 3.0 mmol, 1.0
equiv), (S)-α-MBA (3.6 mmol, 1.2 equiv), Ti(Oi-Pr)4 (4.8 mmol,
1.6 equiv), stirred 20 min, then Pd/Al2O3 (1.5 mol%) was added,
pressurized with H2 (20 bar), 35 °C, 36 h. Hydrogenolysis stage:
Open reactor, Pd(OH)2 (1.0 mol%) was added, then pressurized
with H2 (20 bar), 55 °C, 48 h. Acid-base workup (see general pro-
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

164
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
was excellent for transferring the Raney Ni (large-gauge needle/
syringe) and importantly did not lower the de. Use of toluene as the
sole reaction solvent resulted in reduced de.] Pressurized with H2
(20 bar), 50 °C, 48 h. Hydrogenolysis stage: Open reactor, Pd/C (1.0
mol%) was added, pressurized with H2 (20 bar), 50 °C, 45 h. Acid-
base workup (see general one-pot procedure), and conversion into
the hydrochloride salt gave 4m·HCl; 88 area% chemically pure. The
corresponding free amine is semi-volatile. To overcome this, the
salt was added to a small separatory funnel containing aq 1 M
NaOH (100 mL) and CHCl3 (10 mL). The CHCl3 layer was glass pi-
pette removed and added to a column of silica gel pre-flushed with
petroleum ether. The basic organic phase was once more extracted
with CHCl3 (5 mL) and this was also added to the silica gel column.
With the free amine now on the column, petroleum ether was added
and the solvent pushed down to the level of the silica gel. The po-
larity of the eluent was slowly raised to a final mixture of EtOAc–
petroleum ether (2:3). The fractions containing the product (GC
verified) were combined and aq 4.0 M HCl (~3.0 equiv) was added.
This was rotary evaporated and high vacuum dried until a constant
weight was achieved; yield: 76%. The high chemical purity was
confirmed by GC analysis of a small portion of the free amine,
>98% purity. GC methods, product retention times, and ee determi-
nation (trifluoroacetamide GC analysis) are available.7
1H NMR (400 MHz, CDCl3): δ = 7.13–7.24 (m, 4 H), 6.56 (br s, 1
H), 5.16–5.20 (m, 1 H), 2.76–2.89 (m, 2 H), 2.07–2.13 (m, 1 H),
1.87–1.91 (m, 3 H).
13C NMR (100 MHz, CDCl3): δ = 157.1, 156.7, 156.4, 156.0, 138.0,
134.4, 129.6, 128.6, 128.1, 126.7, 120.3, 117.4, 114.5, 111.7, 48.5,
29.5, 29.0, 19.7.
MS (EI): m/z (%) = 243 (2), 146 (3), 130 (100), 129 (16), 115 (6),
91 (4).
Anal. Calcd for C12H12F3NO: C, 59.26; H, 4.97; N, 5.76; Found: C,
59.29; H, 4.95; N, 5.64.
(S)-6,7,8,9-Tetrahydro-5H-benzo[7]annulen-5-amine (4q)
To benzosuberone (1.00 equiv, 2.50 mmol 0.37 mL) was added
Ti(Oi-Pr)4 (0.92 mL, 3.13 mmol, 1.25 equiv) and (S)-α-MBA
(0.37 mL, 2.88 mmol, 1.15 equiv). This mixture was prestirred (5 h)
neat in the reaction vessel at r.t. EtOAc (5.0 mL) was added, then
Pd/C (24.0 mg, 0.23 mol%; pre-activated at 60 °C/0.3 mbar, 1 h)
and the reaction was pressurized with H2 (8.3 bar) at r.t. for 36 h.
Workup and hydrogenolysis stage were as for 4p to give, after col-
umn chromatography, 4q·trifluoroacetamide; yield: 64% (from
benzosuberone); 76% ee. GC methods, product retention times, and
ee determination (trifluoroacetamide GC analysis) are available.14
(S)-4q·Trifluoracetamide
Rf = 0.39 (heptane–EtOAc, 84:16).
(S)-4m·HCl
Previously characterized.
IR (KBr): 3293, 2929, 1694, 1556, 1183 cm–1.
1H NMR (400 MHz, CDCl3): δ = 8.32 (br s, 3 H), 3.29 (m, 1 H), 1.78
1H NMR (400 MHz, CDCl3): δ = 7.14–7.21 (m, 4 H), 6.62 (br s, 1
H), 5.18–5.22 (t, J = 8 Hz, 1 H), 2.78–2.93 (m, 2 H), 1.56–2.03 (m,
6 H).
(m, 1 H), 1.70 (m, 1 H), 1.41–1.28 (m, 11 H), 0.86 (t, J = 6.4 Hz, 3
H).
13C NMR (100 MHz, CDCl3): δ = 48.6, 35.0, 31.6, 28.9, 25.6, 22.6,
13C NMR (100 MHz, CDCl3): δ = 156.6, 156.3, 155.9, 155.5, 141.2,
139.4, 130.5, 127.9, 126.6, 125.9, 120.2, 117.4, 114.5, 111.6, 54.6,
36.0, 33.2, 27.7, 27.3.
18.8, 14.1.
(S)-1,2,3,4-Tetrahydronaphthalen-1-amine (4p)
A mixture of 1-tetralone (1p, 0.33 mL, 2.50 mmol, 1.00 equiv),
Ti(Oi-Pr)4 (0.92 mL, 3.13 mmol, 1.25 equiv), and (S)-α-MBA (0.35
mL, 2.75 mmol, 1.10 equiv) was prestirred (1 h) neat in the reaction
vessel at 23 °C. t-BuOMe (5.0 mL) and Pd/C (24.0 mg, 0.23 mol%)
were added, and the reaction pressurized H2 (8.3 bar), 40 °C, 27 h.
Acid-base workup (see general procedures) with rotary evaporator
concentration at r.t. gave the crude product (954.0 mg, a pale yellow
liquid of predominately 3p with some 4p) to which was added
MeOH (7.0 mL), AcOH (0.90 mL, 15.20 mmol, 4.00 equiv), and
Pd/C (970.0 mg, 6.0 mol%), followed by pressurization with H2 (4.1
bar), 23 °C, 14 h. The product was filtered through celite, the celite
washed with MeOH (2 × 20 mL), and aq 1.0 M HCl (5.0 mL) was
added. The soln was concentrated until the nonvolatile water re-
mained, and this was washed with Et2O (2 × 20 mL) to remove neu-
tral impurities. The remaining aqueous acid layer was basified with
aq 2.0 M NaOH (10 mL) to pH 10–12, and extracted with CH2Cl2
(3 × 30 mL). The combined CH2Cl2 extracts were then washed with
sat. NaCl (2 × 10 mL) and dried (Na2SO4). After filtration, 3.0 M
HCl in Et2O (2.00 equiv) was added and the mixture was then evap-
orated to dryness to obtain the crude hydrochloride salt of the pri-
mary amine 4p (747.6 mg). CH2Cl2 was added (0.50 M) to the salt,
followed by Tf2O (2.3 mL, 16.3 mmol, 4.00 equiv) and Et3N (2.8
mL, 20.35 mmol, 5.00 equiv), then gently refluxed. After 40 min the
mixture was quenched with sat. NaHCO3 (2 × 10 mL), the CH2Cl2
layer was dried (Na2SO4), filtered, and evaporated to dryness to ob-
tain the crude amide which was then purified by flash chromatogra-
phy (heptane–EtOAc, 19:1) to give 4p as white solid; yield: 462 mg
(76% overall from 1p); 92% ee. GC methods, product retention
times, and ee determination (trifluoroacetamide GC analysis) are
available.14
MS (EI): m/z (%) = 257 (6), 256 (10), 228 (4), 144 (100), 143 (7),
129 (34), 116 (12), 91 (7).
Anal. Calcd for C13H14F3NO: C, 60.7; H, 5.49; N, 5.44. Found: C,
60.47; H, 5.38; N, 5.56.
Cuprate Addition for the Synthesis of Amines 3; General Pro-
cedure E
A round-bottom flask containing anhyd CuBr (3.0 equiv) was gen-
tly heated (<80 °C) by heat gun under high vacuum for 5 min,
flushed with N2, cooled to r.t., and then THF or Et2O (30.0–35.0
mL, in total 39.0–44.0 mL, 0.10–0.08 M) was added. This soln was
cooled to –45 °C and the alkylmagnesium chloride in THF (6.0
equiv) was added over 5 min. This soln was then further stirred at
–45 °C for 15 min, and then cooled to –78 °C. BF3·OEt2 (for 3t: 3.0
equiv; for 3u and 3v: 2.0 equiv) was then added over 2 min, and the
soln was stirred for 5 min before the addition of a prestirred (r.t., 0.5
h) soln of the aldehyde 2 (5.0 mmol, 1.0 equiv), THF or Et2O (7.0
mL), Ti(Oi-Pr)4 (73 μL, 0.25 mmol, 0.05 equiv), and (S)-α-MBA
(0.68 mL, 5.25 mmol, 1.05 equiv) via cannula over 20–25 min. Ad-
ditional THF or Et2O (2.0 mL) was added to the residual imine and
the resulting soln was added via cannula to the reaction flask. The
reaction soln was then stirred at –78 °C for 2 h, and then at –45 °C
for 1 h. The reaction was then quenched by the addition of sat.
NH4Cl–25% NH4OH (7:3 ratio, 30 mL in total) at –45 °C for 5 min.
The cooling bath was removed and stirring continued for another 90
min at r.t. Et2O (20 mL) was added, and the biphasic soln was stirred
for 5 min. The mixture was then filtered through a bed of celite and
the celite subsequently washed with Et2O (3 × 25 mL). The aqueous
phase was extracted with Et2O (3 × 15 mL). The combined organic
extracts were washed with sat. NH4Cl (2 × 25 mL), then with brine
(2 × 20 mL), dried (Na2SO4), filtered, and then evaporated to dry-
ness to obtain the crude product (the de was measured at this point).
Details of the synthesis of 3t–v regarding GC methods, product re-
tention times, ee determination (trifluoroacetamide GC analysis),
(S)-4p·Trifluoroacetamide
Rf = 0.41 (heptane–EtOAc, 84:16).
IR (KBr): 3295, 2935, 1697, 1550, 1181 cm–1.
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Chiral Primary Amines
165
and the use of THF as a replacement solvent for Et2O are avail-
able.21
13C NMR (100 MHz, CDCl3): δ = 146.6, 128.2, 126.7, 126.6, 55.1,
52.4, 44.4, 34.8, 27.7, 24.9, 24.6, 23.5, 22.8, 22.6, 14.0.
MS (EI): m/z (%) = 247 (2) [M+], 232 (4), 190 (100), 106 (10), 105
(82), 86 (48).
HRMS (70 eV): m/z [M+] calcd for C17H29N: 247.2300; found:
247.2295.
(3S)-N-[(S)-1-Phenylethyl]nonan-3-amine (3t)
Following general procedure E using heptanal (0.67 mL, 5.0 mmol,
1.0 equiv), CuBr (2.15 g, 15.0 mmol, 3.0 equiv), 2.0 M EtMgCl in
THF (15.0 mL, 30.0 mmol, 6.0 equiv), BF3·OEt2 (1.91 mL, 15.0
mmol, 3.0 equiv), Et2O (44.0 mL, 0.08 M). Crude product: 81% de.
Purification by flash chromatography (silica gel, hexanes–EtOAc–
NH4OH, 92:6:2) gave the mixture of diastereomers as a colorless
viscous liquid, which was then treated with ethereal HCl to obtain
the hydrochloride salt; yield: 1.14 g (81%).
Acknowledgment
We are grateful for financial support from Jacobs University
Bremen. This work has been performed within the graduate pro-
gram of Nanomolecular Science.
(S,S)-3t
Free base; Rf = 0.48 (hexanes–EtOAc–NH4OH, 84:14:2).
IR (KBr): 3425, 3025, 2959, 2856, 1453, 1368, 1210, 1119, 1028,
700, 553 cm–1.
References
1H NMR (400 MHz, CDCl3): δ = 7.31–7.19 (m, 5 H), 3.86 (q, J =
6.8 Hz, 1 H), 2.29–2.23 (m, 1 H), 1.43–1.17 (m, 16 H), 0.88 (t, J =
6.8 Hz, 3 H), 0.80 (t, J = 7.2 Hz, 3 H).
13C NMR (100 MHz, CDCl3): δ = 146.5, 128.2, 126.6, 55.5, 55.0,
33.1, 31.8, 29.6, 27.1, 25.1, 24.8, 22.6, 14.1, 10.1.
MS (EI): m/z (%) = 247 (2) [M+], 232 (7), 218 (45), 162 (58), 114
(30), 105 (100), 58 (28).
HRMS (70 eV): m/z [M]+ calcd for C17H29N; 247.2300; found:
247.2293.
(1) For very recent reviews, see: (a) Xie, J.-H.; Zhu, S.-F.; Zhou,
Q.-L. Chem. Soc. Rev. 2012, 41, 4126. (b) Jones, S.; Warner,
C. J. A. Org. Biomol. Chem. 2012, 10, 2189. (c) Mathew, S.;
Yun, H. ACS Catalysis 2012, 2, 993. (d) Xie, J.-H.; Zhu, S.-
F.; Zhou, Q.-L. Chem. Rev. 2011, 111, 1713. (e) Tufvesson,
P.; Lima-Ramos, J.; Jensen, J. S.; Al-Haque, N.; Neto, W.;
Woodley, J. M. Biotechnol. Bioeng. 2011, 108, 1479.
(f) Robak, M. T.; Herbage, M. A.; Ellman, J. A. Chem. Rev.
2010, 110, 3600. (g) Fleury-Brégeot, N.; de La Fuente, V.;
Castillón, S.; Claver, C. ChemCatChem 2010, 2, 1346.
(h) Hesp, K. D.; Stradiotto, M. ChemCatChem 2010, 2,
1192. (i) Nugent, T. C.; El-Shazly, M. Adv. Synth. Catal.
2010, 352, 753. (j) Koszelewski, D.; Tauber, K.; Faber, K.;
Kroutil, W. Trends Biotechnol. 2010, 28, 324.
(2) We have demonstrated this for primary amines 4p and 4q,
i.e., Pd/C (6.0 mol%), AcOH, low hydrogen pressures (4
bar), r.t., allowed hydrogenolysis; see the procedure for 4p
(experimental section) for an example. Note that the
racemization of chiral amines is possible in the presence of
Pd, see ref. 3.
(5S)-N-[(S)-1-Phenylethyl]undecan-5-amine (3u)
Following general procedure E using heptanal (0.67 mL, 5.0 mmol,
1.00 equiv), CuBr (2.15 g, 15.0 mmol, 3.0 equiv), 2.0 M BuMgCl
in THF (15.0 mL, 30.0 mmol, 6.0 equiv), BF3·OEt2 (1.27 mL, 10.0
mmol, 2.0 equiv), Et2O (44.0 mL, 0.08 M). Crude product: 86% de.
Purification by flash chromatography (silica gel, hexanes–EtOAc–
NH4OH, 92:6:2) gave the mixture of diastereomers as a colorless
viscous liquid, which was then treated with ethereal HCl to obtain
the hydrochloride salt; yield: 1.32 g (85%).
(3) Parvulescu, A. N.; Jacobs, P. A.; De Vos, D. E. Chem.–Eur.
J. 2007, 13, 2034.
(S,S)-3u
Major product, the diastereomers were not separable via chroma-
tography (silica gel); the mixture was characterized as the free base;
Rf = 0.55 (hexanes–EtOAc–NH4OH, 84:14:2).
(4) For supportive examples, see: (a) Nugent, T. C.; Seemayer,
R. Org. Process Res. Dev. 2006, 10, 142. (b) Neidigh, K. A.;
Avery, M. A.; Williamson, J. S.; Bhattacharyya, S. J. Chem.
Soc., Perkin Trans. 1 1998, 2527. (c) Bhattacharyya, S.;
Chatterjee, A.; Williamson, J. S. Synlett 1995, 1079.
(d) Seebach, D.; Hungerbuehler, E.; Naef, R.;
IR (KBr): 3444, 3025, 2957, 2856, 1603, 1466, 1368, 1118, 909,
700, 556 cm–1.
1H NMR (400 MHz, CDCl3): δ = 7.33–7.20 (m, 5 H), 3.86 (q, J =
6.4 Hz, 1 H), 2.32–2.27 (m, 1 H), 1.34–1.12 (m, 20 H), 0.90–0.82
(m, 6 H).
Schnurrenberger, P.; Weidmann, B.; Zueger, M. Synthesis
1982, 138.
(5) For the reductive amination of keto esters, see: Nugent, T.
C.; Ghosh, A. Eur. J. Org. Chem. 2007, 3863.
13C NMR (100 MHz, CDCl3): δ = 146.4, 128.2, 126.6, 54.9, 53.9,
34.4, 33.6, 31.8, 29.6, 27.9, 25.1, 24.8, 22.8, 22.6, 14.1, 14.0.
(6) See the following reference which strongly implies that
chloro-, bromo-, and presumably iodoaromatic moieties
would be hydrogenolyzed when using our methods:
Bringmann, G.; Geisler, J.-P.; Geuder, T.; Künkel, G.;
Kinzinger, L. Liebigs Ann. Chem. 1990, 795.
(7) Nugent, T. C.; Negru, D. E.; El-Shazly, M.; Hu, D.; Sadiq,
A.; Bibi, A.; Umar, M. N. Adv. Synth. Catal. 2011, 353,
2085; and supporting information therein.
(8) See ref. 7, p 2087 for the text discussion of compound 2a.
(9) See ref. 7, compounds 2d and 3d (Tables 1 and 2).
(10) See ref. 6, p 797, in reference to aromatic halide removal:
‘This well-known hydro-dehalogenation can be prevented
only by application of a special catalyst mixture Rh2O3/PtO2
(3:2), which thus allows the preparation of the corresponding
secondary amine 9m with a still significant stereoselectivity
(see Table 1).’
(S)-2-Methyl-N-[(S)-1-phenylethyl]octan-4-amine (3v)
Following general procedure E using 3-methylbutanal (0.54 mL,
5.0 mmol, 1.0 equiv), CuBr (2.15 g, 15.0 mmol, 3.0 equiv), 2.0 M
BuMgCl in THF (15.0 mL, 30.0 mmol, 6.0 equiv), BF3·OEt2 (1.27
mL, 10.0 mmol, 2.0 equiv), Et2O (44.0 mL, 0.08 M). Crude product:
90% de. Purification by flash chromatography (silica gel, hexanes–
EtOAc–NH4OH, 93.5:3.5:3) gave the mixture of diastereomers as a
colorless viscous liquid, which was then treated with ethereal HCl
to obtain the hydrochloride salt; yield: 1.21 g (85%).
(S,S)-3v
Free base; Rf = 0.48 (hexanes–EtOAc–NH4OH, 83:15:2).
IR (KBr): 3441, 3026, 2956, 2930, 2628, 1467, 1367, 1154, 1118,
700, 558 cm–1.
1H NMR (400 MHz, CDCl3): δ = 7.33–7.19 (m, 5 H), 3.86 (q, J =
6.4 Hz, 1 H), 2.38–2.32 (m, 1 H), 1.63–1.55 (m, 1 H), 1.33–1.12 (m,
12 H), 0.89 (d, J = 6.8 Hz, 3 H), 0.84 (t, J = 6.8 Hz, 3 H), 0.77 (d,
J = 6.8 Hz, 3 H).
(11) Process for the preparation of (S,S)-cis-2-benzhydryl-3-
benzylaminoquinuclidine: Nugent, T. C.; Seemayer, R. WO
2004035575, 2004.
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 153–166

166
T. C. Nugent, S. M. Marinova
FEATURE ARTICLE
(12) See refs. 7 and 4a for examples of success using
crystallization to impart high de, in particular in the
supporting information of those references.
h) before pressurization with H2. See the experimental
section provided here for the procedures for 3m and 4m for
further details.
(13) From greater than 20 examples, one substrate remains an
oddity regarding the inconsistency of its stereoselectivity, in
particular 3o (see Figure 4). We have previously reported its
de as 94%, see ref. 14, but more recent unpublished research
results of ours have shown that the de, and consequently the
ee for this substrate, actually ranges between 84–94% de
(average = 89% de). It is difficult to speculate on the source
of the problem because other alkyl aryl ketones, e.g. o-
substituted acetophenones, required the same reaction
conditions or slightly harsher (albeit using a different
solvent), and show no discrepancies in stereoselectivity.
Racemization events can occur in the presence of Pd/C at
high temperatures, e.g., see ref. 3.
(14) Nugent, T. C.; Ghosh, A. K.; Wakchaure, V. N.; Mohanty,
R. R. Adv. Synth. Catal. 2006, 348, 1289; and supporting
information therein.
(15) Nugent, T. C.; Wakchaure, V. N.; Ghosh, A. K.; Mohanty,
R. R. Org. Lett. 2005, 7, 4967; and supporting information
therein.
(17) Reeves, J. T.; Tan, Z.; Han, Z. S.; Li, G.; Zhang, Y.; Xu, Y.;
Reeves, D. C.; Gonnella, N. C.; Ma, S.; Lee, H.; Lu, B. Z.;
Senanayake, C. H. Angew. Chem. Int. Ed. 2012, 51, 1400.
(18) Nugent, T. C.; El-Shazly, M.; Wakchaure, V. N. J. Org.
Chem. 2008, 73, 1297.
(19) See ref. 18, p 1302, the text and footnote 34, for specific
examples.
(20) We cannot rule out the possibility that a one-pot reductive
amination–hydrogenolysis has been previously achieved,
but our attempts to use key word and phrase searches failed
to identify anything regarding this topic in the literature. To
see our research in this area, see ref. 7.
(21) Wakchaure, V. N.; Mohanty, R. R.; Shaikh, A. J.; Nugent, T.
C. Eur. J. Org. Chem. 2007, 959.
(22) Here our educated guess is that 3v would be produced in
~85% de using reductive amination with the appropriate
ketone precursor.
(23) (a) Nugent, T. C.; Umar, M. N.; Bibi, A. Org. Biomol. Chem.
2010, 8, 4085. (b) Nugent, T. C.; Bibi, A.; Sadiq, A.; Shoaib,
M.; Umar, M. N.; Tehrani, F. N. Org. Biomol. Chem. 2012,
10, 9287.
(16) The experimental description, ref. 7, for compounds 3m and
4m lacked an important detail, i.e., prestirring of the reagents
and starting materials neat at elevated temperature (50 °C, 4
Synthesis 2013, 45, 153–166
© Georg Thieme Verlag Stuttgart · New York