A highly efficient resolution protocol for 2′-halo-α-methylbenzylamines

Article abstract of DOI:10.1016/j.tetlet.2007.04.120

A highly efficient resolution protocol for 2′-halo-α-methylbenzylamines is reported. Commercially available and inexpensive mandelic acid can be used for the resolution of the Br, Cl, and F derivatives to >99% de in a single crystallization. In addition, the reduction of acetophenone oximes using borane-dimethylsulfide is presented as a method for the preparation of racemic amine precursors.

Full text of DOI:10.1016/j.tetlet.2007.04.120

Tetrahedron Letters 48 (2007) 4589–4593
A highly efficient resolution protocol
0
for 2 -halo-a-methylbenzylamines
Liane M. Klingensmith, Kelly A. Nadeau and George A. Moniz*
Chemical Process R&D, Amgen, Inc., One Kendall Square, Bldg. 1000, Cambridge, MA 02139, United States
Received 17 April 2007; accepted 24 April 2007
Available online 29 April 2007
0
Abstract—A highly efficient resolution protocol for 2 -halo-a-methylbenzylamines is reported. Commercially available and inexpen-
sive mandelic acid can be used for the resolution of the Br, Cl, and F derivatives to >99% de in a single crystallization. In addition,
the reduction of acetophenone oximes using borane-dimethylsulfide is presented as a method for the preparation of racemic amine
precursors.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
the ease and cost of resolving the racemic amine might
outweigh the benefits of an asymmetric synthesis.
1
b
The process of resolving enantiomers through diastereo-
meric salt formation is a widely used method for obtain-
ing optically enriched acids or bases. As part of an
The literature contains many reports of the resolution
of various a-methylbenzylamines, including the use of
chiral acids such as 6-(1,2:3,4-di-O-isopropylidene-a-D-
1
ongoing development program, we required large quanti-
0
3
ties of 2 -halo-substituted a-methylbenzylamines (Fig. 1,
galactopyranosyl)hydrogen phthalate, isopropylidene
4
1
a–c) in high optical purity.
glycerol hydrogen phthalate, 3,4-dihydro-2H-1-benzo-
5
6
pyran-2-carboxylic acid, mandelic acid, substituted
6
a,7
8,9
There are numerous asymmetric synthetic strategies that
mandelic acids,
tartaric acid, N-Ac-L-leucine deriv-
atives, and malic acid.9b Surprisingly, there are few
10
afford enantiomerically enriched a-methylbenzylam-
2
0
ines. However, from our standpoint, a resolution-based
examples of the resolution of 2 -halo-a-methylbenzyl-
3
–5,6e,10
process was more practical for two reasons: (1) the de-
gree of enantiomeric enrichment required for our pur-
poses (>99% ee) might exceed the capacity of available
asymmetric methods. In this case, a secondary salt-
based chiral upgrade would still be required, and (2) if
a relatively inexpensive resolving agent could be found,
amines,
and those examples produced only modest
enantioselectivity, or required multiple crystallizations to
achieve high levels of enantioenrichment. Our goal was
to develop a general and efficient resolution protocol,
employing a single resolving agent that could be
applied to the series depicted in Figure 1.
NH2
2. Results and discussion
At the outset, a general and mild approach was needed
0
to obtain large quantities of racemic 2 -halo-substituted
X
a-methylbenzylamines, as racemate was not commer-
cially available. There is a large body of literature
regarding the preparation of racemic versions of these
compounds. Starting from the ketone, reductive amina-
X = Br, Cl, F
1
a-c
0
Figure 1. 2 -Halo-a-methylbenzylamines (1a–c).
11
tion, including the Leuckart reaction, can be used.
Oximes and oxime ethers can be converted to the corre-
sponding amine utilizing both catalytic and non-cata-
Keywords: Reduction; Halogenated; Resolution; Enantioselective.
1
2
*
Corresponding author. E-mail: gmoniz@amgen.com
lytic transfer hydrogenation agents,
lithium
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.04.120

4590
L. M. Klingensmith et al. / Tetrahedron Letters 48 (2007) 4589–4593
aluminum hydride,13 and various methods which use
the racemic amine. It was found that 2 equiv of borane-
dimethylsulfide effected the reduction efficiently and pre-
served the aryl–halogen bond, producing the desired
amine in good yields (Table 2). An acid/base extractive
workup yielded the desired amine without need for fur-
ther purification. This reduction process was demon-
strated on scale with 287 g of 2-fluoroacetophenone
oxime (3c) (Table 2, entry 3), and found to produce
amine with equivalent yield and quality to smaller scale
experiments.
1
4
combinations of metals and hydride reducing agents.
Borane-based reduction methods have been reported
1
5
less frequently in the literature, despite the fact that
they are mild in comparison to other reduction proto-
cols. We chose this as the starting point for our studies
since we expected minimal reduction of the carbon–halo-
gen bond with these relatively mild reagents.
Table 1 shows the results of oxime formation using an
optimized procedure for the condensation of the corre-
sponding ketone (2a–c) with O-benzylhydroxylamine,
which was found to be the optimal oxime substrate for
reduction to the amine. Both cis and trans regioisomers
were obtained as products, and notably were used with-
out further purification following an aqueous workup.
With an efficient process for racemic amine in hand, a
preliminary screen of several commercially available chi-
0
ral resolving agents was performed with 2 -fluoro-a-
methylbenzylamine (Table 3). Ethyl acetate was chosen
as the initial screening solvent to evaluate any solubility
difference between the diastereomeric salts while maxi-
mizing crystallization. In some cases, methanol was em-
ployed as a co-solvent to facilitate dissolution of the acid
With the oximes in hand, investigations into the reduc-
tion focused on utilizing mild boron reagents to produce
Table 1. Synthesis of 2-halo-O-benzylacetophenone oximes 3a–c
Table 2. Synthesis of a-methylbenzylamines 1a–c
.
O NH HCl
OBn
OBn
2
O
X
N
X
N
X
NH2
X
(
1.1 equiv)
) BH3. DMS (2 equiv)
THF, 24 h
1
NaOAc (1.1 equiv)
EtOH:H O (2.5:1), 80 °C
2
2) 4 M HCl
2
3
3
1
a
a
Entry
Ketone
X
Oxime
Yield (%)
Entry Oxime
X
Temperature (°C) Amine Yield (%)
1
1
3
2a
2b
2c
Br
Cl
F
3a
3b
3c
96
96
96
1
2
3
3a
3b
3c
Br rt
Cl rt
1a
1b
1c
72
68
66
b
b
F
40
a
a
Isolated yields.
Isolated yields, average of two reactions.
287 g reaction.
b
b
170 g reaction.
Table 3. Results of initial resolving agent screen
O
. HO
NH
F
NH2
F
2
Acid (1 equiv)
Solvent (0.36 M)
R
70 °C-rt
4c
1
c
a
Entry
Acid
Solvent
Yield (%)
de (%)
1
2
3
4
5
6
7
8
9
0
1
2
3
(S)-(+)-Mandelic acid
L-Malic acid
(S)-(+)-2-Phenylglycine
(S)-(ꢀ)-2-Pyrrolidine-5-carboxylic acid
N-Acetyl-L-phenylalanine
N-Acetyl-L-leucine
(S)-(ꢀ)-2-Pyrrolidine-5-carboxylic acid
(S)-(ꢀ)-2-Pyrrolidine-5-carboxylic acid
(S)-(+)-Mandelic acid
(S)-(+)-Mandelic acid
(S)-(+)-Mandelic acid
(S)-(+)-Mandelic acid
(S)-(+)-Mandelic acid
EtOAc
EtOAc
NA
NA
NA
NA
NA
NA
NA
NA
NA
22
5.7
0.4
NA
6.7
0.7
0.7
b
EtOAc/MeOH (5:3), 0.22 M
EtOAc
EtOAc/MeOH (10:3), 0.27 M
EtOAc/MeOH (10:3), 0.27 M
EtOH
MeOH
MeOH
EtOH
IPA
b
NA
b
NA
b
NA
1
1
1
1
93
89
98
99
41
39
34
EtOH/IPA (3:2), 0.72 M
EtOH/IPA (1:1), 0.72 M
a
Determined by chiral GC.
No crystallization observed.
b

L. M. Klingensmith et al. / Tetrahedron Letters 48 (2007) 4589–4593
4591
prior to the addition of the amine. As seen in Table 3
entries 1–6), initial upgrades were poor. However, both
Table 4. Expanded screen of scope using mandelic acid
(
O
mandelic acid and 2-pyrrolidine-5-carboxylic acid affor-
ded modest levels of enrichment, revealing a difference
in solubility between the two diastereomeric salts gener-
. HO
NH2
NH
X
2
(
S)-(+)-Mandelic-
Acid (1 equiv)
HO
1
6
ated with these two resolving agents.
X
Solvent, 70 °C-rt
1
4
The initial hits using mandelic acid and 2-pyrrolidine-5-
carboxylic acid were subjected to a more extensive sol-
vent screen (Table 3, entries 7–10), which revealed man-
delic acid to be a very effective resolving agent in
ethanol. The conditions for resolution utilizing mandelic
acid were optimized to afford a diastereoenrichment of
a
Entry
X
Solvent
Yield (%) de (%)
1
2
3
4
5
6
7
8
9
Cl IPA/EtOH (1:1), 0.32 M
Cl IPA, 0.4 M
Cl IPA/EtOH (1:6:1), 0.4 M
Cl IPA/EtOH (1:6:1), 0.4 M
Cl IPA/EtOH (1:7:1), 0.32 M
Br IPA/EtOH (1:1), 0.25 M
Br IPA, 0.5 M
Br IPAC, 0.5 M
Br IPAC/EtOH (3:1), 0.5 M
Br IPAC/EtOH (3:1), 0.33 M
Br IPAC/EtOH (3:1), 0.25 M
25
37
34
29
87
45
77
98
99
NA
NA
9.6
92
94
99
b
9
9% in a single crystallization from ethanol and iso-
b
25
propyl alcohol (entry 13). The solubility and nucleation
curves were generated for this salt and revealed a high
nucleation temperature and narrow metastable zone
c
c
NA
NA
67
(
Fig. 2).
34
37
35
b
b
1
1
0
1
0
Application of these conditions to the other 2 -halo-a-
a
methylbenzylamines of interest revealed that the optimal
Determined by chiral GC.
0
b
conditions for 2 -fluoro-a-methylbenzylamine did not
The reaction was first cooled to 30 °C and slurried for 24 h before
cooling to 23 °C.
directly translate to the other amine substrates. How-
ever, upon further optimization for each derivative,
mandelic acid continued to be an outstanding resolving
agent (Table 4).
c
No crystallization observed.
0
Qualitatively, nucleation and crystallization for the 2 -
60
chloro derivative were observed to be slower than that
of the 2 -fluoro series, and to take place at a lower tem-
Solution
Point
Nucleation point
0
50
perature. This was quantified by the solubility and
nucleation curves for this salt (Fig. 3), which confirmed
that nucleation of the (R)-2 -chloro-a-methylbenzyl-
one
Hold
40
30
20
10
0
Z
at 30°C
0
table
tas
e
amineÆ(S)-mandelate salt only commences at 30 °C at
concentrations relevant to the resolution protocol. It
was expected that holding the mixture at 30 °C would
promote nucleation of the desired diastereomer and sub-
sequent crystallization within the metastable zone, while
still maintaining temperature above the solubility limit
of the undesired diastereomer. The impact of this hold
time is demonstrated by entries 3 and 4 in Table 4, which
show that holding the mixture at 30 °C for crystalliza-
tion results in a substantial increase in % de with mini-
mal impact in % yield.
M
0
10
20
30
40
50
60
70
Temperature (°C)
0
Figure 3. Solubility and nucleation curves for (R)-2 -chloro-a-methyl-
benzylamineÆ(S)-mandelate salt (4b) in 1.7:1 IPA/EtOH.
1
20
00
Solution
Point
0
For the 2 -bromo derivative, use of isopropyl acetate
Nucleation point
1
(IPAC) as an antisolvent in combination with ethanol
was required to achieve the optimal balance of enrich-
ment and recovery. The solubility and nucleation curves
for this salt are shown in Figure 4. Although the nucle-
ation temperature is relatively high for this salt, in prac-
tice, nucleation was slow. Again, holding the
crystallization at 30 °C allowed sufficient nucleation
time at a slightly elevated temperature, affording very
good % de’s in a single crystallization.
8
6
4
2
0
0
0
0
0
e
n
o
Z
le
tab
tas
e
M
Final optimized conditions as well as absolute configu-
rations for the entire amine series are shown in Table
0
10
20
30
40
50
60
70
80
Temperature (°C)
1
7
5
. Under the optimized conditions, these resolutions
0
Figure 2. Solubility and nucleation curves for (S)-2 -fluoro-a-methyl-
benzylamineÆ(S)-mandelate salt (4c) in 1:1 IPA/EtOH.
achieve high diastereoenrichments and good yields in a
single crystallization, employing a readily available

4592
L. M. Klingensmith et al. / Tetrahedron Letters 48 (2007) 4589–4593
6
0
0
0
0
0
0
0
Acknowledgments
5
4
3
2
1
The authors wish to acknowledge Dr. Jason Tedrow
0
for a supply of (R)-2 -fluoro-a-methylbenzylalcohol,
Solution
Point
Nucleation point
Dr. Eric Bercot for helpful discussions, and Dr. Randy
Jensen for NMR support.
Supplementary data
Experimental procedures and full spectroscopic data for
all new compounds are available as supplementary data.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2007.04.120.
0
10
20
30
40
50
60
70
80
Temperature (°C)
0
Figure 4. Solubility and nucleation curves for (R)-2 -bromo-a-methyl-
benzylamineÆ(S)-mandelate salt (4a) in 3:1 IPAC/EtOH.
References and notes
and inexpensive ($25/kg)18 resolving agent. Demonstra-
tion of this process on larger scale produced 76.4 g of 2 -
fluoro-a-methylbenzylamine-mandelic acid salt (4c),
0
1. (a) Sheldon Roger, A. Chirotechnology: Industrial Synthe-
sis of Optically Active Compounds; Marcel Dekker: New
York, 1993; (b) Collins, A. N.; Sheldrake, G. N.; Crosby,
J. Chirality in Industry II; John Wiley & Sons Ltd: West
Sussex, England, 1997; (c) Kozma, David CRC Handbook
of Optical Resolutions via Diastereomeric Salt Formation;
CRC Press LLC: Boca Raton, Florida, 2002.
. For reviews see: (a) Kobayashi, S.; Ishitani, H. Chem. Rev.
999, 99, 1069; (b) Enders, D.; Reinhold, U. Tetrahedron:
Asymmetry 1997, 8, 1895.
with the expected yield and diastereoselectivity (Table
1
9
5
, entry 3).
0
In conclusion, the borane reduction of 2 -haloacetophe-
none oximes has been demonstrated to be a viable route
for large scale generation of racemic a-methylbenzylam-
ines. These materials can be resolved to >99% de in a
single crystallization using commercially available and
inexpensive mandelic acid.
2
1
3. Mereyala, H. B.; Fatima, L.; Pola, P. Tetrahedron:
Asymmetry 2004, 15, 585.
0
Table 5. Optimized resolutions of 2 -halo-a-methylbenzylamines (la–c)
O
.
HO
NH2
NH
X
2
(
S)-(+)-Mandelic-
Acid (1 equiv)
a
HO
b
X
Solvent, 70 °C-rt
1
4
c
d
Entry
1
Amine
Solvent
Product
Yield (%)
de (%)
Amine configuration
NH2
IPAC/EtOH (3:1), 0.25 M
IPA/EtOH (1:7:1), 0.32 M
4a
35
99.3
99.6
R
1
a
Br
NH2
2
4b
4c
25
34
R
S
1b
Cl
NH2
e
3
IPA/EtOH (1:1), 0.72 M
ꢀ99.1
1c
F
a
b
c
See Ref. 17.
Entry 1: 60 °C–rt, held at 30 °C for 24 h before cooling to 23 °C.
Isolated yields, average of two reactions, with a maximum yield of 50%.
Determine by chiral GC, average of two resolutions. See Supplementary data for absolute configuration analysis.
Performed with 108 g of racemate.
d
e

L. M. Klingensmith et al. / Tetrahedron Letters 48 (2007) 4589–4593
4593
4
. (a) Pallavicini, M.; Valoti, E.; Villa, L.; Piccolo, O.
Tetrahedron: Asymmetry 2001, 12, 1071; (b) Pallavicini,
M.; Valoti, E.; Villa, L.; Piccolo, O. Tetrahedron: Asym-
metry 1996, 4, 1117.
14. For representative examples see: (a) Itsuno, S.; Sakurai,
Y.; Ito, K. Synthesis 1988, 995; (b) Harrison, J. R.;
Moody, C. J.; Pitts, M. R. Synlett 2000, 1601; (c) Leeds, J.
P.; Kirst, H. A. Synth. Commun. 1988, 18, 777; (d)
Sotomayor, N.; Dominguez, E.; Lete, E. J. Org. Chem.
1996, 61, 4062; (e) Baruah, B.; Dutta, M. P.; Boruah, A.;
Prajapati, D.; Sandhu, J. S. Synlett 1999, 409; (f) Jarvest,
R. L.; Armstrong, S. A.; Berge, J. M.; Brown, P.; Elder, J.
S.; Brown, M. J.; Copley, R. C. B.; Forrest, A. K.;
Hamprecht, D. W.; O’Hanlon, P. J.; Mitchell, D. J.;
Rittenhouse, S.; Witty, D. R. Biorg. Med. Chem. Lett.
2004, 14, 3937; (g) Sasse, A.; Ligneau, X.; Rouleau, A.;
Elz, S.; Ganellin, R.; Arrang, J.-M.; Schwartz, J.-C.;
Schunack, W.; Stark, H. J. Med. Chem. 2002, 45, 4000; (h)
Bloom, J. D.; Dushin, R. G.; Curran, K. J.; Donahue, F.;
Norton, E. B.; Terefenko, E.; Jones, T. R.; Ross, A. A.;
Feld, B.; Lang, S. A.; DiGrandi, M. J. Bioorg. Med. Chem.
Lett. 2004, 14, 3401; (i) Martins, J. E. D.; Mehlecke, C.
M.; Gamba, M.; Costa, V. E. U. Tetrahedron: Asymmetry
2006, 17, 1817; (j) Koufaki, M.; Detsi, A.; Theodorou, E.;
Kiziridi, C.; Calogeropoulou, P. N.; Vassilopoulos, A.;
Kourounakis, A. P.; Rekka, E.; Kourounakis, P. N.;
Gaitanaki, C.; Papazafiri, P. Bioorg. Med. Chem. Lett.
2004, 12, 4835.
5
. Janssen, C. G. M.; Thijssen, J. B. A.; Verluyten, W. L. M.
J. Labelled Compd. Radiopharm. 1987, 8, 909.
6
. (a) Kinbara, K.; Harada, Y.; Saigo, K. J. Chem. Soc.,
Perkin Trans. 2 2000, 1339; (b) Sakai, K.; Hashimoto, Y.;
Kinbara, K.; Saigo, K. Bull. Chem. Soc. Jpn. 1993, 66,
3
414; (c) Roje, M.; Hamersak, Z.; Sunjic, V. Syn.
Commun. 2002, 22, 3413; (d) Kohara, T.; Hashimoto,
Y.; Saigo, K. Tetrahedron 1999, 55, 6453; (e) Nieuwenhuij-
zen, J. W.; Grimbergen, R. F. P.; Koopman, C.; Kellogg,
R. M.; Vries, T. R.; Pouwer, K.; van Echten, E.; Kaptein,
B.; Hulshof, L. A.; Broxterman, Q. B. Angew. Chem., Int.
Ed. 2002, 22, 4281; (f) Hagiya, T. K.; Harada, E. I.; Goto,
H. O. US Pat. 5,739,401.
7
. Kinbara, K.; Sakai, K.; Hashimoto, Y.; Nohira, H.; Saigo,
K. Tetrahedron: Asymmetry 1996, 6, 1539.
8
. (a) Theilacker, W.; Winkler, H. G. Ber 1954, 87, 690; (b)
Ault, A. Org. Syn. Coll. Vol. 1969, 5, 932; (c) Andre, E.;
Vernier, C. Compt. Rend. 1931, 193, 1192.
9
. (a) Ingersoll, A. W. Org. Syn. Coll. Vol. 1943, 2, 506; (b)
Ingersoll, A. W. J. Am. Chem. Soc. 1925, 47, 1168.
1
0. (a) Pickard, S. T.; Smith, H. E. J. Am. Chem. Soc. 1990,
12, 5741; (b) Wang, X.; Erickson, S. D.; Iimori, T.; Still,
W. C. J. Am. Chem. Soc. 1992, 114, 4128.
1. For a review see: (a) Moore, M. L. Org. React. 1949, 301;
15. (a) Carroll, F. I.; Brieaddy, L. E.; Navarro, H. A.; Damaj,
M. I.; Martin, B. R. J. Med. Chem. 2005, 48, 7491; (b)
Feuer, H.; Braunstein, D. M. J. Org. Chem. 1969, 6, 1917;
(c) Hassner, A.; Catsoulacos, P. Chem. Comm. 1967, 590;
(d) Carlier, P. R.; Du, D.-M.; Han, Y.-F.; Liu, J.; Perola,
E.; Williams, I. D.; Pang, Y.-P. Angew. Chem., Int. Ed.
2000, 39, 1775; (e) Tchertchian, S.; Hartley, O.; Botti, P. J.
Org. Chem. 2004, 69, 9208; (f) Hanna, P. E.; Grund, V. R.;
Anders, M. W. J. Med. Chem. 1974, 9, 1020; (g)
Yamazaki, N.; Atobe, M.; Kibayashi, C. Tetrahedron
Lett. 2002, 43, 7979; (h) Geyer, H. M., III; Martin, L. L.;
Crichlow, C. A.; Dekow, F. W.; Ellis, D. B.; Kruse, H.;
Setescak, L. L.; Worm, M. J. Med. Chem. 1982, 25, 340.
16. Solubility differences for diastereomeric salts are generally
manifested in a range of organic solvents. See: Wilen, S.
H.; Collet, A.; Jacques, J. Tetrahedron 1977, 33, 2725.
17. Use of less than 1 equiv of mandelic acid results in reduced
1
1
(
b) Pickard, S. T.; Smith, H. E. J. Am. Chem. Soc. 1990,
12, 5741; (c) Ingersoll, A. W.; Brown, J. H.; Kim, C. K.;
Beauchamp, W. D.; Jennings, G. J. Am. Chem. Soc. 1936,
8, 1808.
2. (a) Abiraj, K.; Gowda, D. C. Synth. Commun. 2004, 34,
99; (b) Garofalo, A.; Campiani, G.; Ciani, S. Am.;
1
5
1
1
5
Fiorini, I.; Nacci, V. Tetrahedron 1996, 52, 7745; (c)
Abiraj, K.; Gowda, D. C. J. Chem. Res. (S) 2003, 332; (d)
Doreswamy, B. B. H.; Mahendra, M.; Mantelingu, K.;
Sridhar, M. A.; Shashidhara Prasad, J.; Rangappa, K. S.
Indian J. Chem., Sect. B 2005, 44, 148.
3. For representative examples see: (a) Rahman, M. A.;
Fraser-Reid, B. J. Am. Chem. Soc. 1985, 107, 5576; (b)
Miyata, O.; Koizumi, T.; Asai, H.; Iba, R.; Naito, T.
Tetrahedron 2004, 60, 3893; (c) Krzeminski, M. P.;
Wojtczak, A. Tetrahedron Lett. 2005, 46, 8299; (d) Gilbert,
C. L. K.; Lisek, C. R.; White, R. L.; Gumina, G.
Tetrahedron 2005, 61, 8339.
0
yield (24% for 2 -fluoro).
18. Price quotation at 100 kg scale. Cost at 25 kg scale is $30/
kg.
19. See Supplementary data for a representative salt break
procedure.