Evaluation and Development of Practical Routes to an Enantiomerically Pure C 2-Symmetric Diamine Building Block

Article abstract of DOI:10.1021/op400292m

Several routes to an enantiomerically pure C₂-symmetric diamine were evaluated and modified to scalable methods. A Zn/Me₃SiCl- mediated reductive coupling of an imine was found to be superior to the other methods investigated, allowing us to safely prepare the enantiomerically pure diamine also on a large scale. One key step in this method was a highly efficient resolution of a stereoisomeric mixture of the diamine through salt formation with (-)-dibenzoyl-l-tartaric acid. The enantiomerically pure C ₂-symmetric diamine obtained was further used as a key building block for the synthesis of potent Kv1.5 channel blockers.

Full text of DOI:10.1021/op400292m

Subscriber access provided by DUESSELDORF LIBRARIES
Full Paper
Evaluation and Development of Practical Routes to an
Enantiomerically Pure C2 Symmetrical Diamine Building Block
Staffan Per Olof Karlsson, Jan Lindberg, and Henrik Sörensen
Org. Process Res. Dev., Just Accepted Manuscript • Publication Date (Web): 21 Nov 2013
Downloaded from http://pubs.acs.org on November 24, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Organic Process Research & Development is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Evaluation and Development of Practical Routes to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
an Enantiomerically Pure C Symmetrical Diamine
2
Building Block
Staffan Karlsson,* Jan Lindberg and Henrik Sörensen
Medicinal Chemistry, Cardiovascular & Metabolic Diseases iMed, AstraZeneca R & D Mölndal,
S-431 83 Mölndal, Sweden
SYNOPSIS
ABSTRACT: Several routes to an enantiomerically pure C symmetrical diamine were evaluated
2
and modified to scalable methods. A Zn/Me SiCl mediated reductive coupling of an imine was
3
found to be superior to the other methods investigated allowing us to safely prepare the
enantiomerically pure diamine also on a large scale. One key step in this method was a highly
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
efficient resolution of a stereoisomeric mixture of the diamine through salt formation with (ꢀ)ꢀ
dibenzoylꢀLꢀtartaric acid. The enantiomerically pure C symmetrical diamine obtained was
2
further used as a key building block for the synthesis of potent Kv1.5 channel blockers.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION: Enantiomerically pure C symmetrical diamines are attractive targets for
2
1
use
as
chiral
ligands
in
asymmetric
synthesis.
In addition to the importance of these in this field, we also identified such types of symmetrical
diamines as building blocks for the synthesis of potent Kv1.5 ion channel blockers. To be able to
run a full toxicology program we needed several hundreds of grams of the symmetrical diamine
(S,S)ꢀ1 which further could be substituted on the secondary amines to give the desired API AZDꢀ
2502 (Figure 1). We considered to utilize the commercially available diamine (S,S)ꢀ2 as starting
material. However, since this diamine was very expensive and only a few suppliers were
available we sought a scalable route to obtain the diamine (S,S)ꢀ1 or (S,S)ꢀ2. This would secure
material needed to select a candidate for treatment of arrhythmia.
HN
NH
H N
2
N
N
NH2
N N
N
N N
N
(S,S)-1
(S,S)-2
AZD-2502
Figure 1. Building blocks for the synthesis of Kv1.5 ion channel blockers such as AZDꢀ2502.
RESULTS AND DISCUSSION
A literature survey of methods to synthesize diamines of type 1 and 2 revealed that several
2
methods were available (Figure 2). To prepare screening compounds for the lead optimization
ACS Paragon Plus Environment
2

Page 3 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
3
program, an approach via an epoxide was used (approach A, Figure 2). Although this approach
was suitable for the preparation of a range of diverse screening compounds within a short period
of time, it suffered from low selectivity in the coupling step to the epoxide and the necessity to
separate enantiomers through chromatography at a later stage. Also, we were concerned with the
use of hexaethylphosphorous triamide in the coupling step since this reagent is transformed into
hexaethylphosphoric triamide (HEPA) which is an analogue to the well known carcinogenic
HMPA. Furthermore, the reagent was expensive and not easily accessible. In the evaluation of
routes to the diamine (S,S)ꢀ1 we initially chose five approaches (approach BꢀF) depicted in
Figure 2 which we believed would have the greatest chance to be scaleꢀup friendly.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
Ar
O
O
NH2
Ar
Ar
H
Ar
NH OAc, AcOH
4
N
N
N
H
Ar
Ar
1
2
) LDA
Approach C
Approach D
) Oxidation
1) Reduction
2) Hydrolysis
O
O
Ar
S N
N S
ArLi
R HN
1
NHR1
Ar
1
) Reduction
HN
Ar
N
NH3
ArCHO
Ar
2
) Hydrolysis
Ar
Approach B
Approach E
Zn, TMSCl
O
Approach A
R1
Ar
N
Ar
Ar
O
R NH
1
2
2
P(Et N)3
O
Approach F
Ar
Ar
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 2. Various approaches available to obtain the C symmetrical diamines 1 or 2.
2
APPROACH BꢀC.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Among the six methods depicted in Figure 2, we first believed that the method described by
2
e
Deng et al. (Approach B) would be suitable for our purpose because it would furnish the
enantiomerically pure diamine (S,S)ꢀ1 without having to resolve enantiomeric mixtures.
However, despite several attempts using various conditions with Grignard and lithium
carbanions, we were not able to obtain a satisfactory yield and diastereoselectivity for this
transformation. We decided to abandon this approach and instead turn to another asymmetric
b
method described by Yaozhong et al. (Approach C).2 In this approach a chiral auxiliary (i.e
camphor) is condensed with 2ꢀaminomethylpyridine. The resulting imine undergoes an oxidative
coupling mediated by various oxidants to give diastereomerically enriched diimines in good
selectivity, which after hydrolysis furnish the diamine (S,S)ꢀ2. However, very soon we realized
that the method did not give the result that we expected. The reaction was performed several
times using various oxidants such as Fe(III), 1,2ꢀdibromoethane and Br but we were not able to
2
reproduce the result obtained by Yaozhong et al. Incomplete conversion of starting material and
a very low diastereoselectivity were obtained and it was clear that this method was not suitable
for further scaleꢀup work. We realized that a nonꢀasymmetric method followed by resolution of
enantiomers had to be considered to obtain the desired diamine (S,S)ꢀ1 or (S,S)ꢀ2.
APPROACH D.
f
In a paper by Corey et al.,2 a nonꢀasymmetric method is described where an aromatic
diketone is condensed with cyclohexanone and NH OAc to give the corresponding diimine
4
(Approach D). The authors were then able to reduce the diimine to the corresponding transꢀ
ACS Paragon Plus Environment
4

Page 5 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
diamine. After hydrolysis, the corresponding racemic primary diamines could be liberated and
then resolved into pure enantiomers through salt formation with tartaric acid. We decided to
evaluate this route further but starting from 2ꢀpyridinecarboxaldehyde. Using this approach we
were able to secure 1.2 kg of the dihydrochloride salt of the racemic diamine (±)ꢀ2 x 2HCl after
some development work according to Scheme 1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 1. Approach D .
O
OH
O
i
ii
iii
N
N
iv
N
N
N
H
N
N
90%
74%
60%
N
N
OH
O
3
4
5
x 2 HCl
H N
NH2
N
2
v
vi
HN
NH
N
N
N
vii
HN
NH
N
6
8%
steps
81%
2 steps
N
N
N
N
N
2
6
(±)-2 x 2HCl
(±)-1
7
a
Reaction conditions: (i) KCN, H O/pyridine, reflux; (ii) I , NaOAc, EtOH; (iii)
2
2
o
o
cyclohexanone, NH OAc, HOAc, 40 C; (iv) Mg, MeOH, 40 C; (v) HCl/2ꢀpropanol; (vi)
4
propionaldehyde, NaBH(OAc) ; (vii) 3M HCl, 1ꢀpropanethiol.
3
The diketone 4, although commercially available, seemed to be quite labile and we were not
able to purchase pure material in sufficient amount from any supplier. Therefore we prepared
4
this freshly via the enediol 3 obtained from the commercially available 2ꢀ
pyridinecarboxaldehyde. The oxidation of 3 to 4 was initially carried out according to a
5
procedure involving NaBrO generated from Br . However, we found that the procedure could
2
be simplified by carrying out the oxidation with I in equally high yield, and thus avoiding the
2
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
handling of large amounts of bromine. The diketone 4 was then condensed with cyclohexanone
and ammonia to give the diimine 5 that was ready for a stereoselective reduction. We found that
magnesium (0) as reducing agent in MeOH gave the most consistent results and we could
abandon other inconvenient reducing methods such as Li(0) and Na(0) reductions in liquid NH3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
which had to be carried out at very low temperatures (ꢀ78 C) to avoid reduction of the pyridine
rings. The obtained trans diastereomer 6 was then treated with HCl which hydrolyzed the N,Nꢀ
ketal leaving the dihydrochloride salt of the pure diamine (±)ꢀ2 x 2HCl as an insoluble
crystalline solid in 70% yield. At this stage we wanted to resolve the enantiomers prior to further
manipulations. We wanted to avoid the use of chromatography and therefore we screened a
selection of chiral acids and we were finally able to resolve the racemate through salt formation
with (R)ꢀmandelic acid providing the dimandelate salt of the diamine (S,S)ꢀ2 in 96% ee (Scheme
2). However, the resolution was challenging and control over the stoichiometry of acid/amine
was essential to obtain high enantiomeric purity of the product. This was due to a switch in
relative solubilities of the diastereomeric salts when having a 1:1 salt (n = 1) compared with a
1:2 salt (n = 2, Scheme 2). Moreover, it was found that a too high loading of the mandelic acid
1
resulted in an almost racemic product. When analyzing this salt by H NMR it was found that the
ratio of acid/amine exceeded 2 (n > 2, Scheme 2) meaning that probably a third molecule of
mandelic acid was part of the salt through hydrogen bonding making the crystallization even
more troublesome.
Scheme 2. Resolution of (±)ꢀ2 through salt formation with (R)ꢀmandelic acid, n = 2.
ACS Paragon Plus Environment
6

Page 7 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
On a small scale (20 g), the 1:2 mandelate salt of (S,S)ꢀ2 isolated after one crystallization from
EtOAc/CH Cl was reductively alkylated followed by a hydrolysis to give (S,S)ꢀ1 following the
2
2
procedure described for the racemate in Scheme 1. We were able to determine the enantiomeric
purity of (S,S)ꢀ1 by chiral HPLC (97:3 er). The scale up (800 g) of the resolution of (±)ꢀ2
through salt formation with (R)ꢀmandelic acid failed and a very low enantiomeric purity was
obtained of the diamine (S,S)ꢀ2 for reasons as described above. We were at that time under
pressure to deliver material for the toxicology studies and to save time we decided to move on
6
with the racemic material and separate the enantiomers at a later stage through chromatography.
Using the approach as described in Scheme 1, we were able to prepare 394 g of the diamine (±)ꢀ
1
. After separation of the enantiomers via chiral chromatography followed by the final synthetic
step we were able to successfully deliver the desired API in time for the initial toxicology
studies. However, we were a bit concerned of many of the steps for further scaleꢀup. In addition
to the long linear sequence to (±)ꢀ1 (total overall yield = 22%), the reduction of 5 to 6 with Mg
was highly exothermic and timeꢀconsuming on a larger scale. Moreover, the resolution of the
racemate (±)ꢀ2 was not as straightforward as we first expected and a lot of efforts would
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
probably be needed to develop a working process. Therefore, as a backꢀup we searched for a
shorter route where such issues could be avoided.
APPROACH E.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
It is known that 2ꢀpyridinecarboxaldehyde in the presence of an aqueous solution of ammonia
7
undergoes condensation furnishing the cisꢀdihydroimidazole 8 (Scheme 3). By treatment with a
7
strong base, 8 epimerizes to give the transꢀdihydroimidazole 9. In our hands, this procedure was
easily reproduced and gave 9 in 87% yield. By some modifications of the original procedure we
were able to develop a scaleꢀup friendly method that gave consistent results. A potential route to
the enantiomerically pure diamine (S,S)ꢀ2 from 9 could then be via separation of the
8
enantiomeric mixture of 9 mediated by acetylmandelic acid as has been shown for an analogue.
However, it would require harsh conditions to hydrolyze the mandelic amide (HBr in AcOH) and
this operation was foreseen to be impractical on a large scale. Instead our plan was to reduce 9 to
give the corresponding imidazolidine 10 followed by resolution of the enantiomers at a later
9
stage. The reduction devised by Corey and Kühnle using aluminum amalgam in wet THF was
rejected for environmental reasons. We reasoned that acylated 9 may be easier to reduce but after
1
0
having tried several conditions on Bocꢀprotected 9 we turned back on finding alternatives for
reducing 9. An initial futile attempt using Na/Naphthalene was followed by a much more
successful combination of zinc and acetic acid that after some optimization cleanly gave the
desired aminal 10. We found that the imidazolidine 10 was stable under alkaline conditions but
rapidly hydrolyzed under acidic conditions (pH 1ꢀ2). Our plan was to utilize the 2ꢀ
pyridinecarboxaldehyde unit as a protective aminal group to avoid excessive alkylation in the
following reductive alkylation step.
ACS Paragon Plus Environment
8

Page 9 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
a
Scheme 3. Approach E .
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reaction conditions: (i) 26% NH (aq.), THF; (ii) Kꢀamylate (0.02 equiv.), THF; (iii) Zn,
3
HOAc, 2ꢀmethylꢀTHF; (iv) propionaldehyde , DIPEA, sodium triacetoxyborohydride, THF; (v)
HCl, 1ꢀpropanethiol (6 equiv.); (vi) (ꢀ)ꢀdibenzoylꢀLꢀtartaric acid (50 mol%), EtOAc/heptane.
Treatment of 10 with propionaldehyde in the presence of sodium triacetoxyborohydride gave
the alkylated product 11 in 69% yield. In an attempt to optimize the route for further scaleꢀup we
investigated if the reduction of 9 and alkylation of 10 could be telescoped. Thus we wanted to
use the MeꢀTHF solution of 10 after the extractive workꢀup directly in the next step. However,
we found that the reductive alkylation was best performed using EtOAc as solvent and it was
also an advantage to use reasonably dry solvents. When run with the crude wet MeꢀTHF
solution of 10, incomplete conversions were obtained and the reaction mixtures in some cases
turned into a unfavorable twoꢀphase mixture. Thus we decided to concentrate the MeꢀTHF
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
solution of 10 prior to the reductive alkylation step and use EtOAc as solvent instead. What
remained to be done to reach the diamine (±)ꢀ1 was the hydrolytic removal of the 2ꢀ
pyridinecarboxaldehyde moiety. This was easily accomplished by direct treatment of the acidic
aqueous layer, after the workꢀup of 11, with propanethiol to trap the aldehyde formed after
hydrolysis. In order to resolve the enantiomers we screened for chiral acids that would furnish a
salt with suitable properties (Table 1).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Table 1. Salt screen performed in order to find a suitable acid for the resolution of (±)ꢀ1.
b
c
Entry
1
Resolving acid
Type of salt
1:1
Yield
49%
Er 1
97:3
d
2
ꢀꢀ
ꢀꢀ
ꢀꢀ
d
3
4
5
ꢀꢀ
ꢀꢀ
20%
ꢀꢀ
ꢀꢀ
0%
ꢀꢀ
1:1
d
ꢀꢀ
6
1:1
~ 10%
0%
ACS Paragon Plus Environment
10

Page 11 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
d
7
8
ꢀꢀ
ꢀꢀ
ꢀꢀ
1:2
25%
92:8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
b
Salt formation in EtOAc or EtOAc/heptane mixtures on a small scale (0.3 mmol). 50ꢀ70
c
d
mol% of the resolving acid was used. Determined using chiral HPLC. No solid material
obtained.
Fortunately we found that both dibenzoyltartrate (1:1 salt, entry 1) and Nꢀacetyl leucine (1:2
salt, entry 8) formed a salt with desired properties, the former salt yielding a much better result.
On a 5 g scale, we obtained the dibenzoyltartrate salt of (S,S)ꢀ1 in 49% yield with a slightly
lower enantiomeric ratio (88:12 er) of the free amine as compared with the small scale
experiment (97:3 er, entry 1, Table 1). One more recrystallization from EtOAc gave the
1
1
stereoisomerically pure (S,S)ꢀ1 Lꢀdibenzoyltartrate salt.
The approach depicted in Scheme 3 was applied starting from 0.5 kg of the 2ꢀ
pyridinecarboxaldehyde. We were able to obtain 144 g of the diamine (±)ꢀ1 corresponding to an
overall yield of 25% from 2ꢀpyridinecarboxaldehyde. No chromatography was necessary in the
sequence as the intermediates 8 and 9 both had favorable properties allowing for an efficient
crystallization to take place. Furthermore, careful adjustment of pH was used in the extractive
workꢀup of (±)ꢀ1 which raised the chemical purity to an acceptable level.
Despite the successful results obtained above, the approach E still suffers from being a
relatively long linear sequence with a low overall yield. Thus, we were eager to see if we could
find a route that was shorter allowing for a more efficient synthesis of the desired diamine (S,S)ꢀ
1.
APPROACH F.
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In a paper by Alexakis et al., they describe the synthesis of various C symmetrical diamines of
2
a
type 1 mediated by Zn/Me SiCl (approach F).2 Although the diastereoselectivity in these
3
reactions in most cases were low to moderate resulting in a low yield of the desired isomer after
purification, the desired product was obtained in one single step from easily prepared imines. If
this method could be successfully applied on imine 12, only a resolution of the diasteromeric
mixture and the enantiomers would remain to reach diamines of type 1. Therefore, we put some
efforts to investigate this approach. The starting imine 12 was easily prepared according to
Scheme 4 in almost quantitative yield.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Scheme 4. Approach F .
O
O
O
O
O
O
O-
O-
+
NH2+
O
NH HN
i
N
ii, iii
H N
2
v
N
N
N
N
N
N
98%
10%
2
steps
(
S,S)-1
12
(S,S)-1
L-dibenzoyltartrate salt
v,iii
SO42-
H2+
N
H2+
N
ii, iv
3
7%
2 steps
9:1 dr
N
N
9
(
±)-1
sulfate salt
a
Reaction conditions: (i) 1ꢀpropylamine (1.05 eq), EtOH; (ii) Zn powder (0.5 equiv.), TMSCl
(1.5 equiv.), CH CN, extractive workꢀup; (iii) (ꢀ)ꢀdibenzoylꢀLꢀtartaric acid, CH CN, > 99% ee,
3
3
9
8:2 dr; (iv) H SO , CH CN; (v) Extraction with MTBE/NH (aq.) solution.
2
4
3
3
ACS Paragon Plus Environment
12

Page 13 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
In a first attempt to reductively couple imine 12, we found that the crude mixture when analyzed
1
by H NMR consisted of a 70:30 mixture of diastereomers which was a slightly better result than
a
1
that reported for an analogue.2 A 41% assay w/w by H NMR of the desired isomer was
recorded which corresponded to an effective yield of 40%. Although a method exists for the
isomerisation of the mesoꢀ to the (±)ꢀisomer we believed that this method should not be practical
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
on a large scale. Since we wanted to avoid purification of the crude mixture through
chromatography, we searched for a suitable acid that would separate the diasteromers and
remove other impurities upon salt formation. A salt screen was set up and we found that
sulphuric acid formed the corresponding diastereomerically pure bidentate sulfate salt of (±)ꢀ1
in 37% overall yield. The experiment was repeated on a 1.3 mol scale with similar result. Pure
enantiomer (S,S)ꢀ1 could be obtained after resolution of the free based racemic diamine (±)ꢀ1
with (ꢀ)ꢀdibenzoylꢀLꢀtartaric acid as described above. Encouraged by these results we wanted to
investigate if the process could be further improved to allow separation of diastereomers and
enantiomeric salt resolution in only one crystallization step. On a small scale, we were delighted
to see that after treatment of the crude mixture from the reductive coupling of the imine 12 with
(ꢀ)ꢀdibenzoylꢀLꢀtartaric acid acid in CH CN, both a diastereomeric separation and enantiomeric
3
resolution were observed and the almost stereoisomerically (ee > 99%, dr 98:2) pure (S,S)ꢀ1 Lꢀ
1
3
dibenzoyltartrate salt was isolated in a 15% overall yield (i.e. from 2ꢀpyridinecarboxaldehyde).
Through comparison of retention times on chiral HPLC with authentic material prepared from
commercially available diamine (S,S)ꢀ2 (transformation (S,S)ꢀ2 to (S,S)ꢀ1, Scheme 1), we were
able to assign the absolute stereochemistry of the diamine 1 as S,S.
Before we were ready for further scaleꢀup using this approach we wanted to solve a few issues
in the reductive coupling step: First of all, in the original procedure, an excess of Zn (Zn:imine,
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
:1) is used which of course led to a large amount of zinc (0) waste after the reaction is complete.
This excess of zinc resulted in a more troublesome workꢀup and the quenching of the excess
could also be a potentially hazardous operation. In an attempt to reduce the amount of zinc used
we found that a stoichiometry of 1/0.5 (imine/Zn) was enough to reach almost full conversion in
the reductive coupling resulting in a simple workꢀup with no or little remaining unreacted zinc in
the reactor. Another issue that had to be addressed before further scaleꢀup work was to remove
all zinc salts liberated prior to salt formation because the product diamine (±)ꢀ1 was found to
bind zinc very tightly which then interfered in the crystallization step. This was solved through
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
washing of the crude mixture with an EDTA/25% NH solution which efficiently removed all
3
zinc salts and also facilitated the workꢀup. Another thing that we wanted to avoid prior to further
scaleꢀup was the use of the toxic 1,2ꢀdibromoethane used for the activation of the zinc.
Fortunately we found that this activation procedure was not necessary as we got the same result
using the zinc as such from a fresh bottle. Moreover, the exclusion of 1,2ꢀdibromoethane as
activation reagent gave rise to a more straightforward workꢀup in which the formation of black
unwanted precipitates were avoided. However, we noticed that these improvements were at the
expense of the diastereoselectivity in the coupling step since now an approximately 60:40 ratio
of the two diastereomers were obtained. Using the optimized conditions as described above we
were able to produce 77 g of the almost stereoisomerically pure (S,S)ꢀ1 Lꢀdibenzoyltartrate salt
(ee > 99%, dr 98:2) in 10% overall yield from the imine 12. This was our preferred method
among the three main routes (approach DꢀF) evaluated on a large scale (Table 2).
Table 2. Comparison of the three approaches (DꢀF) to obtain diamine (±)ꢀ1.
Approach D Approach E Approach F
Cost and accessibility of raw materials
+
+
+
ACS Paragon Plus Environment
14

Page 15 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Number of steps
Overall yield
7
22%
ꢀ
5
25%
+
3
a
b
c
d
34% (10%)
Ease of operation
+
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Highly exothermic steps
Yes
No
No
a
b
Overall yield to crude diamine (±)ꢀ1 with assay 74% w/w. Overall yield to crude diamine
c
d
(±)ꢀ1 with assay 82% w/w. Overall yield to (±)ꢀ1 sulfate salt. Overall yield to (S,S)ꢀ1 Lꢀ
dibenzoyltartrate salt.
CONCLUSION
Several routes towards the C symmetrical diamine (S,S)ꢀ1 have been evaluated, both
2
asymmetric as well as nonꢀasymmetric ones. It turned out that a Zn/TMSCl mediated reductive
coupling approach of imine 12 was the preferred method. Despite being nonꢀasymmetric and
also low yielding this route was chosen for further scaleꢀup. This is for reasons as follows: It is a
short and simple sequence to execute on a large scale and the pure desired enantiomer can easily
be isolated after a single salt crystallization. Moreover, the methods are chromatography free
processes and the cost of raw materials is very low. The obtained C symmetrical diamine (S,S)ꢀ1
2
can be used as a building block to synthesize new potent ion channel blockers. The diamine
might also be useful as a chiral ligand in asymmetric syntheses.
EXPERIMENTAL SECTION
General. All materials were purchased from commercial suppliers and used as such without
further purification. Zinc was supplied from SigmaꢀAldrich (powder, < 10 ꢁm, cat. no. 20998ꢀ8).
The reactions were performed under an atmosphere of nitrogen in reactors equipped with
overhead stirrer and connected to an external heating Huber aggregate. In process controls (IPCs)
1
were recorded using either HPLC or H NMR analyses on the crude mixtures. Diastereomeric
1
ratios were determined using H NMR integration. Assays of final products and crude mixtures
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
were determined by
H
NMR integration using either benzylbenzoate or
octamethylcyclotetrasiloxane as internal standards. Enantiomeric purity of compound (S,S)ꢀ1
was determined by chiral HPLC analysis (column Chiralpak IA, 250 x 4.6 mm, eluent heptane/2ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
propanol/Et N 70/30/0.1). High resolution mass spectrometry was performed on a Waters XEVO
3
qTOF instrument, mass precision ± 5 ppm. NMR measurement was performed on a Bruker
Avance III spectrometer. LC/MS analyses were recorded on a Waters ZMD instrument, LC
column Terra MS C (Waters); detection with a HP 1100 MSꢀdetector diode array.
8
1,2-Di(pyridin-2-yl)ethane-1,2-dione (4). A reactor was charged with 1,2ꢀdi(pyridinꢀ2ꢀ
yl)etheneꢀ1,2ꢀdiol 3 (2.68 Kg, 12.5 mol), sodium acetate (2.26 Kg, 27.5 mol) and EtOH (6.0 L).
To the resulting suspension was added over 2.5 h a solution of I (3.17 Kg, 12.5 mol) in EtOH
2
(20 L) at 20 °C. The stirring was continued at 20 °C for 20 min. Then, 5% aqueous Na S O (800
2
2
3
mL) was added. 20 L of the solvent was distilled off at 30 ꢀ 40 mbar with the mantle temperature
set at 40 °C. The suspension was further stirred at 20 °C for an additional 1 hour. The product
was filtered off and washed with cold EtOH (2 x 5 L). The solid product was dried under
1
reduced pressure to yield 4 (1.96 Kg, 100% w/w, 9.24 mol) as a white solid. H NMR was in
1
4
accordance with that previously published.
2
,3-Di(pyridin-2-yl)-1,4-diazaspiro[4.5]deca-1,3-diene (5). A reactor was charged with 4 (1.96
Kg, 9.24 mol), ammonium acetate (5.70 Kg, 73.9 mol), cyclohexanone (1.24 L, 12.0 mol) and
AcOH (9.3 L). The resulting mixture was stirred at 40 °C for 17 hours. Water (18.5 L) and
CH Cl (7 L) were added. The phases were separated and the aqueous layer extracted with
2
2
CH Cl (8 L). The combined organic layers were washed with 3.8M NaOH (18.5 L) and water (4
2
2
L). The organic layer was concentrated to yield crude 5 (2.70 kg, assay 60% w/w, 5.57 mol) as a
1
brown syrup. H NMR (400 MHz, CDCl ): δ 1.67–2.01 (m, 10H), 7.27 (ddd, J = 7.5, 4.8, 1.3
3
ACS Paragon Plus Environment
16

Page 17 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Hz, 2H), 7.74 (dt, J = 7.5, 1.7 Hz, 2H), 7.80 (dt, J = 7.8, 1.2 Hz, 2H), 8.46 (ddd, J = 4.8, 1.7, 1.0
1
3
Hz, 2H), C NMR (100 MHz, CDCl ) δ 24.0, 25.5, 34.2, 105.2, 123.8, 124.0, 136.3, 148.7,
3
+
1
52.3, 163.6. HRMS (ESI+) m/e 291.1626 [(M + H) , calcd for C H N 291.1610].
1
8
19
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
±)-1,2-Di(pyridin-2-yl)ethane-1,2-diamine dihydrochloride [(±)-2 x 2HCl]. A mixture of 5
(2.69 kg, 60% w/w, 5.56 mol) in MeOH (10 L) was stirred in a 25 L reactor. Magnesium
turnings (40 g, 1.65 mol), and I (2.7 g, 10 mmol) were added which resulted in gas evolution
2
and an exothermic reaction. More magnesium turnings were added in 75 – 100 g portions over 7
hours (500 g, 20.6 mol). The temperature of the reaction mixture was maintained in the interval
2
7 – 40 °C controlled by the rate of addition of the magnesium and adjustment of the mantle
temperature. After complete addition of the magnesium, the mantle temperature was set at 25 °C
and the stirring was continued for an additional 14 hours. The resulting mixture of crude 6 was
cooled to 15 °C and 5.6M HCl in 2ꢀpropanol (10.9 L, 61 mol) was added over 2 hours. The
resulting suspension was stirred at 20 °C for 21 hours. A second portion of MeOH (5 L) was
added and stirring was continued at 20 °C for 4 days. The solid product was filtered off, washed
with MeOH (3 x 2.5 L) and dried under reduced pressure to yield (±)ꢀ2 x 2HCl (1.24 kg, 87%
1
w/w, 3.76 mol) as an offꢀwhite solid. H NMR (400 MHz, D O): δ 4.89 (s, 2H), 7.16 (dt, J = 7.9,
2
1
.0 Hz, 2H), 7.30 (ddd, J = 7.7, 4.9, 1.1 Hz, 2H), 7.66 (dt, J = 7.8, 1.8 Hz, 2H), 8.45 (ddd, J =
4
.9, 1.7, 0.9 Hz, 2H).
(±)-N1,N2-Dipropyl-1,2-di(pyridin-2-yl)ethane-1,2-diamine [(±)-1]. A reactor was charged
with NaBH(OAc) (763 g, 3.60 mol) and CH Cl (4 L). The mantle temperature was set at 20 °C
3
2
2
and DIPEA (836 mL, 4.80 mol) was added followed by the addition of propionaldehyde (871
mL, 12.0 mol). A solution of (±)ꢀ2 x 2HCl (396 g, 87% w/w, 1.20 mol) and DIPEA (836 mL,
4
.80 mol) in CH Cl (10 L) was added over 30 minutes and the stirring continued for 1 hour.
2 2
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Water (2.0 L) and 3.8M HCl (6 L) were added to the resulting crude mixture of compound 7 over
3
5 minutes and the stirring continued for 30 minutes. To the aqueous layer was added 1ꢀ
propanethiol (650 mL, 7.18 mol). The resulting mixture was stirred at 60 °C for 4 hours and at
0 °C for an additional 5 hours. The mixture was washed with CH Cl (2 x 5 L). To the aqueous
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
2
layer was added over 40 minutes a solution of 3.8M NaOH (10 L). The resulting emulsion
pH11) was heated to 30 °C and extracted with CH Cl (3 x 2.5 L). The combined organic layers
(
2
2
were concentrated to yield the diamine (±)ꢀ1 (394 g, 74% w/w, 0.977 mol) as a brown oil. The
NMR and MS data were in accordance with those given for (S,S)ꢀ1 below.
o
N-(Pyridin-2-ylmethylene)propan-1-amine (12). With a temperature set at 0 C, the reactor
was charged with propanꢀ1ꢀamine (318 g, 5.37 mol) and EtOH (99.5%, 2 L). 2ꢀ
pyridinecarboxaldehyde (548 g, 5.12 mol) was added over a 15 minute period keeping the
o
temperature between + 8 to + 23 C (exothermic). The brown homogenous solution was stirred
1
5
o
for 14 hours and then concentrated at 40 C. Traces of water were azeotropically removed with
1
toluene (3 X 200 ml) to give 12 as a dark brown liquid (744 g, 96% w/w, 4.82 mol). H NMR
(
400 MHz, CDCl ) δ 0.98 (t, J = 7.4 Hz, 3H), 1.72ꢀ1.82 (m, 2H), 3.65 (dt, J = 6.9, 1.2 Hz, 2H),
3
7.29ꢀ7.33 (m, 1H), 7.71ꢀ7.77 (m, 1H), 7.98ꢀ8.02 (m, 1H), 8.37ꢀ8.39 (m, 1H), 8.64ꢀ8.67 (m, 1H).
13
C NMR (101 MHz, CDCl ) δ 11.8, 23.8, 63.3, 121.1, 124.5, 136.4, 149.4, 154.7, 161.7.
3
(±)-N,N’-Dipropyl-1,2-di(pyridin-2-yl)ethane-1,2-diammonium sulphate [(±)-1 sulfate
salt]. A mixture of zinc (< 10 ꢀM , 91g, 1.39 mol) and dibromoethane (12.4 g, 66 mmol) in
o
acetonitrile (2 L) was heated at reflux for 5 minutes. At 28 C, chlorotrimethylsilane (7.2 g,
o
0
.066 mol) was added within 1 minute and stirred for 30 minutes at 28 C. The mantle
o
temperature was set at 20 C and Nꢀ(pyridinꢀ2ꢀylmethylene)propanꢀ1ꢀamine 12 (209 g, 96%
w/w, 1.35 mol) was added within 10 minutes. The reaction temperature was maintained in the
ACS Paragon Plus Environment
18

Page 19 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
o
interval + 27 to + 30 C during this addition. After 5 minutes of stirring, chlorotrimethylsilane
(
216 g, 1.99 mol) was added within 20 minutes keeping the temperature between +27 and +31
°
C (exothermic!). The greenꢀbrown suspension was then stirred at ambient temperature for 35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
o
minutes. The mixture was cooled to + 1 C and the reaction is then carefully quenched
(exothermic!) through the addition of a solution of 25% NH solution (aq., 2 L) keeping
3
o
temperature below 20 C. MTBE (2.5 L) was added and the biphasic mixture was stirred for 15
minutes followed by separation of the layers. The dark brown organic layer was concentrated to
a viscous oil. To this was added MTBE (1 L) and 0.27M EDTA solution (disodium salt, 1 L)
followed by the addition of 25% NH solution (0.5 L). The biphasic mixture was stirred for 18
3
hrs followed by separation of the layers. The dark brown organic layer was washed with a
mixture of 0.27M EDTA solution (disodium salt, 0.5 L) + 25% NH solution (aq, 250 ml). The
3
pooled aqueous layer was extracted with MTBE (0.5 L). The pooled organic layer was
concentrated to give the crude diamine as a viscous oil (181 g, 41% w/w of the desired isomer by
1
H NMR, diastereomeric ratio 70/30, 0.249 mol). The crude oil was dissolved in acetonitrile (900
mL) and sulfuric acid (45.5 g, 0.46 mol) was added. The green suspension obtained was heated
o
to reflux and the suspension was then allowed to reach 20 C during 1 hour and then stirred at
that temperature for an additional 1 hour. The thick green suspension was filtered and washed
with CH CN (5 X 100 mL). After drying under reduced pressure the desired (±)ꢀ1 sulfate salt
3
was obtained as a pale green solid (107 g, 90% w/w, 0.243 mol, dr ~ 99:1) with CH CN as the
3
1
main impurity (7% w/w). H NMR (400 MHz, DMSO) δ 0.81 (t, J = 7.4 Hz, 6H), 1.41ꢀ1.61 (m,
4H), 2.42ꢀ2.64 (m, 4H), 4.34 (s, 2H), 6.93ꢀ6.96 (m, 2H), 7.25ꢀ7.29 (m, 2H), 7.56ꢀ7.61 (m, 2H),
13
8
.56ꢀ8.59 (m, 2H). C NMR (100 MHz, CDCl ) δ 11.3, 20.8, 48.0, 65.0, 123.5, 124.3, 136.8,
3
149.5, 155.2.
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
1S,2S)-N1,N2-Dipropyl-1,2-di(pyridin-2-yl)ethane-1,2-diamine (-)-dibenzoyl-L-tartrate
salt [(S,S)-1 L-dibenzoyltartrate salt]. The reactor was charged with zinc (< 10 ꢀm, 71.4 g,
o
1
.09 mol) followed by the addition of acetonitrile (2.5 L). Mantle temperature was set at 20 C
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and (E)ꢀNꢀ(pyridinꢀ2ꢀylmethylene)propanꢀ1ꢀamine 12 (337 g, 96% w/w, 2.18 mol) was added
within 4 minutes. No exotherm was observed. Chlorotrimethylsilane (356 g, 3.27 mol) was
o
added during 40 minutes while maintaining the reaction temperature between + 23 to 30 C. A
green suspension was obtained which was further stirred at ambient temperature for 30 minutes.
1
H NMR analysis of the crude mixture indicated full conversion of the imine. The mantle
o
temperature was set at 0 C and an aqueous solution of 25% NH solution (1 L) was added during
3
10 minutes. CAUTION: The reaction is initially exothermic! The reaction temperature was kept
o
below + 39 C during this addition. MTBE (2.5 L) was added and the biphasic mixture was
stirred for 1 hour. The pale yellow aqueous layer was discarded and the dark brown organic layer
was concentrated to almost dryness. The residue was dissolved in MTBE (1 L). An aqueous
solutions of EDTA (disodium salt, 0.27M, 1.5 L) + 25% NH solution (aq., 250 mL) were added
3
and the biphasic mixture was stirred over night. The aqueous dark green layer was discarded.
The organic layer was washed with water (500 mL) followed by concentration to a dark brown
oil (326 g, 35% w/w of the desired isomer, 0.382 mol). Water was azeotropically removed with
MTBE (3 X 200 mL). The crude oil was dissolved in acetonitrile (2 L). The mixture was heated
o
to 55 C and (ꢀ)ꢀdibenzoylꢀLꢀtartaric acid (196 g, 0.55 mol) was added in small portions. The
o
dark brown homogenous solution was then allowed to attain 20 C during 1 hour. The thin
suspension obtained was further stirred at room temperature for 4 hours. The mixture was then
filtered and the salt collected was washed with CH CN (5 x 150 mL). After drying under reduced
3
pressure, the titled compound was obtained as a pale green solid (77 g, 97% w/w, 114 mmol), dr
ACS Paragon Plus Environment
20

Page 21 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
9
8:2 by H NMR, ee > 99% by chiral HPLC. H NMR (400 MHz, DMSO) δ 0.76 (t, J = 7.4 Hz,
6H), 1.35ꢀ1.55 (m, 4H), 2.36ꢀ2.44 (m, 2H), 2.48ꢀ2.56 (m, 2H), 4.28 (s, 2H), 5.67 (s, 2H), 6.96ꢀ
1
3
6
.99 (m, 2H), 7.20ꢀ7.24 (m, 2H), 7.46ꢀ7.65 (m, 8H), 7.91ꢀ7.95 (m, 4H), 8.50ꢀ8.53 (m, 2H). C
NMR (100 MHz, DMSO) δ 11.2, 20.8, 47.8, 65.2, 72.3, 123.2, 124.1, 128.7, 129.3, 129.6, 133.4,
36.5, 149.3, 155.7, 164.9, 168.1.
1S,2S)-N1,N2-Dipropyl-1,2-di(pyridin-2-yl)ethane-1,2-diamine (S,S)-1. An analytical
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(
sample was prepared from the above (S,S)ꢀ1 Lꢀdibenzoyltartrate salt through extraction with
MTBE/NH (aq) solution followed by concentration of organic layer to a pale yellow viscous oil.
3
1
H NMR (400 MHz, CDCl ) δ 0.84 (t, J = 7.4 Hz, 6H), 1.38ꢀ1.51 (m, 4H), 2.31ꢀ2.41 (m, 4H),
3
1
3
3
.86 (s, 2H), 6.89ꢀ6.92 (m, 2H), 6.97ꢀ7.01 (m, 2H), 7.34ꢀ7.39 (m, 2H), 8.45ꢀ8.47 (m, 2H). C
NMR (101 MHz, CDCl ) δ 11.8, 23.4, 50.0. 69.3, 121.7, 123.4, 135.5, 149.2, 161.6. HRMS
3
+
20
D
(
ESI+) m/e 299.2246 [(M + H) , calcd for C H N 299.2236]. [α]
– 1 (c = 1, CH CN), dr
3
18
27
4
9
8:2, ee > 99%.
Cis-(±)-2,2',2''-(4,5-dihydro-1H-Imidazole-2,4,5-triyl)tripyridine (8). A solution of 2ꢀ
o
pyridinecarboxaldehyde (500 g, 4.67 mol) in THF (3 L) was heated to 50 C and a solution of
ammonia (aq., 26% w/w, 0.92 L, 12.6 mol) was added within 5 minutes. The homogenous brown
o
o
solution was then stirred at 50 C for 3 days. The mixture was cooled to 20 C, 30% brine (1.4 L)
was added and the phases were separated. The organic phase was concentrated to a dark brown
solid. Water was azeotropically removed with toluene (3 x 300 mL). To the crude solid was
added CH Cl (400 mL) and isopropyl acetate (1000 mL). The brown suspension was then
2
2
o
allowed to stir over night followed by cooling at 0 C for 30 minutes. The mixture was filtered
and the solid was washed with isopropyl acetate (500 mL). After drying under reduced pressure,
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the titled compound was obtained as a beige solid (305 g, 98% w/w, 0.99 mol). Analytical data
7
was in accordance with those given in the literature.
Trans-(±)-2,2',2''-(4,5-dihydro-1H-Imidazole-2,4,5-triyl)tripyridine (9). Compound 8 (305
g, 98% w/w, 0.99 mol) was suspended in dry THF (1 L) and a solution of potassium 2ꢀ
methylbutanꢀ2ꢀolate (1.7M in THF, 12 mL, 20 mmol) was added within 1 minute. The mixture
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
was stirred for 1.5 hour after which H NMR analysis of the crude mixture indicated 96%
conversion to the desired transꢀisomer. Heptane (100 mL) and seeding crystals were sequentially
added. An immediate and quite rapid crystallization initiated. The thin suspension was stirred at
o
2
0 C for 30 minutes. More heptane (1 L) was added and the beige suspension was stirred over
night. The mixture was filtered and the solid collected was washed with MTBE (3 x 200 mL).
After drying under reduced pressure the title compound was obtained as a beige solid (268 g,
7
98% w/w, 0.87 mol). Analytical data were in accordance with those given in the literature.
Trans-(±)-2,2',2''-(imidazolidine-2,4,5-triyl)tripyridine (10). Compound 9 (268 g, 98%
w/w, 0.87 mol) and acetic acid (255 ml, 4.45 mol) were mixed in 2ꢀmethylꢀTHF (4.0 L). The
o
suspension was heated to 40 C and zinc (< 10 ꢀm, 70 g, 1.07 mol) was added over a 10 minute
period. The mixture was stirred for 2 days after which full conversion of the starting material had
1
been obtained ( H NMR). The mixture was cooled to 10 °C and a solution of sodium hydroxide
(3.75 M, 2.0 L, 7.5 mol) was added. The twoꢀphase mixture was stirred for 15 minutes followed
by removal of the inhomogenous aqueous layer. To the organic layer was added a solution of
EDTA (aq., 0.26 M, 1.2 L). The pH was adjusted to 8.9 using sodium hydroxide solution (aq.,
3.75 M, 85 mL). The aqueous phase was discarded. The organic layer was washed with brine
(500 mL) followed by concentration of the clear organic layer. The residue was concentrated
with 2 x 200 ml 2ꢀmethylꢀTHF to give the product as a red oil (255 g, 80% w/w, 0.67 mol).
ACS Paragon Plus Environment
22

Page 23 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
+
1
HRMS (ESI+) m/e 304.1570 [(M + H) , calcd for C H N 304.1557]. H NMR (600 MHz,
1
8
17
5
CDCl ) δ 4.52 (d, J = 6.7 Hz, 1H), 4.55 (d, J = 6.7 Hz, 1H), 5.41 (s, 1H), 7.15 – 7.19 (m, 2H),
3
7
.21 (d, J = 7.8 Hz, 1H) 7.24 – 7.28 (m, 2H), 7.54 – 7.63 (m, 3H), 7.72 (dt, J = 7.5, 1.2 Hz, 1H),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
8
.62 (m, 2H), 8.67 (d, J = 4.8 Hz, 1H). C NMR (151 MHz, CDCl ) δ 161.22, 159.35, 159.19,
3
149.47, 149.45, 149.45, 149.37, 136.63, 136.30, 136.27, 129.01, 128.20, 125.28, 123.28, 123.13,
1
22.71, 122.40, 122.11, 122.07, 78.69, 72.00, 70.02.
Trans-(±)-2,2',2''-(1,3-dipropylimidazolidine-2,4,5-triyl)tripyridine (11). Propionaldehyde
(195 g, 3.36 mol) and DIPEA (348 g, 2.70 mol) were added to a solution of 10 (255 g, 80% w/w,
72 mmol) in ethyl acetate (3.0 L). At 18 °C sodium triacetoxyborohydride (364 g, 1.72 mol)
6
was added portion wise over a period of 40 minutes. The reaction is slightly exothermic and the
temperature rose to 26 °C during the addition. The mixture was stirred for 16 hrs at 20 °C. The
reaction was quenched by the addition of water (1 L) after which a rise in temperature from 20
to 29 °C was noticed. After stirring for 4 minutes the phases were separated and the organic
phase was washed with another 600 ml of water. The resulting solution was used as such in the
16
next step. An assay showed 69% effective yield . A small aliquot was concentrated and used for
1
recording of the NMR spectra. H NMR (400 MHz, CDCl ) δ 0.56 (dt, J = 7.4, 3.3 Hz, 6H), 1.05
3
–
5
1.26 (m, 4H), 1.99 – 2.12 (m, 2H), 2.72 – 2.86 (m, 2H), 4.33 (d, J = 5.4 Hz, 1H), 4.54 (d, J =
.4 Hz, 1H), 5.19 (s, 1H), 7.11 – 7.17 (m, 2H), 7.19 – 7.26 (m, 1H), 7.39 – 7.46 (m, 1H), 7.61 –
1
3
7.79 (m, 3H), 7.9 – 7.98 (m, 2H), 8.4 – 8.45 (m, 1H), 8.5 – 8.58 (m, 2H). C NMR (101 MHz,
CDCl ) δ 163.80, 163.38, 160.85, 149.26, 148.67, 148.16, 136.59, 136.34, 136.24, 123.80,
3
1
23.18, 122.81, 122.24, 122.07, 121.80, 87.73, 76.24, 73.89, 55.76, 48.09, 22.27, 20.95, 11.72,
1.68.
1
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
±)-N1,N2-Dipropyl-1,2-di(pyridin-2-yl)ethane-1,2-diamine [(±)-1]. To 3.7 L of the
extraction mixture containing 11 from the previous step (465 mmol) was added 390 ml 12 M
hydrochloric acid. The mixture was stirred for 5 minutes followed by separation of the layers. To
the aqueous layer was added 1ꢀpropanethiol (253 ml, 6 eq) and the mixture was heated at 50 °C
for 18 hrs. MTBE (900 ml) was added and the pH was adjusted to 5.95 using 3.8 M sodium
hydroxide. The aqueous layer was washed with 900 ml of MTBE. The pH of the aqueous layer
was adjusted to 12.3 by addition of 200 mL of 3.8 M sodium hydroxide solution and then
extracted with 900 ml MTBE. The organic layer was washed with 100 ml of water and then
filtered through celite (Celpure C300). After concentration, the desired product (±)ꢀ1 was
obtained as an oil (143.8 g, 82% w/w, 0.395 mol). Spectroscopic data were in accordance with
those given for (S,S)ꢀ1 above.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ASSOCIATED CONTENT
1
13
Supporting Information. H NMR, C NMR, HRMS, optical rotation, and HPLC analyses.
This material is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*
Eꢀmail: staffan.po.karlsson@astrazeneca.com
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
ACS Paragon Plus Environment
24

Page 25 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
We thank the lead optimization group for valuable discussions, our separation science laboratory
group for determination of enantiomeric excesses, our NMR specialists for assistance with
structure elucidation, and our analytical specialists for recording analytical data such as HRMS
and HPLC purity analyses.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
1
See for example: (a) Ohkuma, T.; Doucet, H.; Pham, T.; Mikami, K.; Korenaga, T.; Terada,
M.; Noyori, R. J. Am. Chem. Soc. 1998, 120, 1086ꢀ1087. (b) Trost, B. M.; Bunt, R. C. J. Am.
Chem. Soc. 1994, 116, 4089ꢀ4090. (c) Pospisil, P. J.; Carsten, D. H.; Jacobsen, E. N. Chem. Eur.
J. 1996, 2, 974ꢀ980. (d) Alakonda, L.; Periasamy, M. J. Organometallic Chem. 2009, 694, 3859ꢀ
3863. (e) Cortez, N. A.; Aguirre, G.; ParraꢀHake, M.; Somanathan, R. Tetrahedron Lett. 2009,
50, 2228ꢀ2231. (f) Maillet, C.; Praveen, T.; Janvier, P.; Minguet, S.; Evain, M.; Saluzzo, C.;
Tommasino, M. L.; Bujoli, B. J. Org. Chem. 2002, 67, 8191ꢀ8196. (g) Berkessel, A.; Schröder,
M.; Sklorz, C. A.; Tabanella, S.; Vogl, N.; Lex, J.; Neudörfl, J. M. J. Org. Chem. 2004, 69,
3050ꢀ3056. (h) Annunziata, R.; Benaglia M.; Caporale, M.; Raimondi, L. Tetrahedron:
Asymmetry 2002, 2727ꢀ2734. (i) Enders, D.; Meiers, M. Synthesis 2002, 2542ꢀ2560. Bennani, Y.
L.; Hanessian, S. Chem. Rev. 1997, 97, 3161ꢀ3195 (j) Hilgraf, R.; Pfaltz, A. Adv. Synth. Catal.
2005, 347, 61ꢀ77.
2
See for example: (a) Alexakis, A.; Aujard, I.; Mangeney, P. Synlett 1998, 873ꢀ874. (b)
Guilan, L.; Wenhao, H.; Yuzi, M.; Yaozhong, J.; Yang, T.ꢀK. Synth. Commun. 1994, 24, 3115ꢀ
122. (c) Lin, G.ꢀQ.; Xu, M.ꢀH.; Zhong, Y.ꢀW.; Sun, X.ꢀW. Acc. Chem. Res. 2008, 41, 831ꢀ840.
d) Zhong, Y.ꢀW.; Izumi, K.; Xu, M.ꢀH.; Lin, G.ꢀQ. Org Lett. 2004, 6, 4747ꢀ4750. (e) Sun, X.;
3
(
Wang, S.; Sun, S.; Zhu, J.; Deng, J. Synlett 2005, 2776ꢀ2780. (f) Corey, E. J.; Lee, D.ꢀH.;
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Sarshar, S. Tetrahedron Asymm. 1995, 6, 3ꢀ6. (g) Bette, V.; Mortreux, A.; Ferioli, F.; Martelli,
G.; Savoia, D.; Carpentier, J.ꢀF. Eur. J. Org. Chem. 2004, 3040ꢀ3045. (h) Westerhausen, M.;
Bollwein, T.; Karaghiosoff, K.; Schneiderbauer, S.; Vogt, M.; Nöth, H. Organometallics 2002,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1, 906ꢀ911. (i) Westerhausen, M.; Bollwein, T.; Makropoulos, N.; Schneiderbauer, S.; Suter,
M.; Nöth, H.; Mayer, P.; Piotrowski, H.; Polborn, K.; Pfitzner, A. Eur. J. Inorg. Chem. 2002,
89ꢀ404. (j) Westerhausen, M.; Bollwein, T.; Makropoulos, N.; Rotter, T. M.; Habereder, T.;
3
Suter, M.; Nöth, H. Eur. J. Inorg. Chem. 2001, 851ꢀ857. (k) Betschart, C.; Schmidt, B.; Seebach,
D. Helv. Chim. Acta. 1988, 71, 1999ꢀ2002. (l) Roskamp, E. J.; Pedersen, S. F. J. Am. Chem. Soc.
1987, 109, 3152ꢀ3154. (m) Kim, H.; Nguyen, Y.; Yen, C. P.ꢀH.; Chagal, L.; Lough, A. J.; Kim,
B. M.; Chin, J. J. Am. Chem. Soc. 2008, 130, 12184ꢀ12191. (n) Vögtle, F.; Goldschmitt, E.
Chem. Ber. 1976, 109, 1ꢀ40. (o) Mangeney, P.; Tejero, T.; Alexakis, A.; Grosjean, F.; Normant,
J. Synthesis 1988, 3, 255ꢀ257. (p) Campos, P. J; Arranz, J.; Rodríguez, M. A. Tetrahedron 2000,
56, 7285ꢀ7289. (q) Dutta, M. P.; Baruah, B.; Boruah, A.; Prajapati, D.; Sandhu, J. S. Synlett
1998, 857ꢀ858.
3
Mark, V. J. Am. Chem. Soc. 1963, 85, 1884ꢀ1885.
4
Buehler, C. A.; Addleburg, J. W.; Glenn, D.M. J. Org. Chem. 1955, 20, 1350ꢀ1355.
5
Chang, H. S.; Woo , J. C.; Lee, K. M.; Ko, Y. K.; Moon, S.ꢀS.; Kim, D.ꢀW. Synth. Comm.
2002, 32, 31ꢀ35.
6
We later found that the enantiomers efficiently could be resolved at a later stage [i.e (±)ꢀ1]
through salt formation with (-)ꢀdibenzoyl tartrate described below in approach EꢀF.
ACS Paragon Plus Environment
26

Page 27 of 27
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
7
Larter, M. L.; Phillips, M.; Ortega, F.; Aguirre, G.; Somanathan, R.; Walsh, P. J. Tetrahedron
Lett. 1998, 39, 4785ꢀ4788.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
Braddock, D. C.; Redmond, J. M.; Hermitage, S. A.; White, A. J. P. Adv. Synth. Catal. 2006,
3
48, 911ꢀ916.
9
Corey, E. J.; Kühnle, F. N. M. Tetrahedron Lett. 1997, 38, 8631ꢀ8634.
1
1
0
1
H /PdꢀC in HOAc, LiBH , NaBH , Mg/MeOH, and BH ꢀTHF.
2
4
4
3
In order to avoid the risk of acylation of the amines by the EtOAc we later on changed
solvent to CH CN (see approach F).
3
1
1
2
3
Alexakis, A.; Aujard, I.; Mangeney, P. Synlett 1998, 875ꢀ876.
It was found that heating the salt in CH CN at reflux for several hours, resulted in
3
deterioration.
14
Rakotomalala, M; Katz, M; Voisin, E; Pace, T. C. S; Bohne, C.; Williams, V. E. Can. J.
Chem. 2011, 89, 297ꢀ302.
1
1
5
6
1
By H NMR full conversion of the aldehyde was indicated after 1 hr.
By HPLCꢀMS the material contains 8% of the pyridine/propyl transaminalized product, 7,
(
m+1 = 339). The assay given corresponds to the sum of 7 and 11.
ACS Paragon Plus Environment
27