A Survey of the Borrowing Hydrogen Approach to the Synthesis of some Pharmaceutically Relevant Intermediates

Article abstract of DOI:10.1021/acs.oprd.5b00199

The use of the "borrowing hydrogen strategy" in the synthesis of a number of typical pharmaceutical intermediates has been investigated. The main aim of this work was to investigate the scope and limitations of current methodology using standard laboratory techniques in an industrial context. Some interesting and significant results were achieved across a diverse set of complex substrates; however several drawbacks with this approach were identified, such as the high loading, poor turnover, and susceptibility to substrate inactivation of the catalysts. These are areas which are highlighted for future investigation and improvements.

Full text of DOI:10.1021/acs.oprd.5b00199

Page 1 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
A Survey of the Borrowing Hydrogen Approach to the
Synthesis of some Pharmaceutically Relevant
Intermediates
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
John Leonard,* A. John Blacker,† Stephen P. Marsden,† Martin F. Jones, Keith R. Mulholland and
Rebecca Newton,#
Pharmaceutical Development, AstraZeneca, Charter Way, Macclesfield, SK10 2NA, U.K.
*Author for correspondence, Telephone: +44 (0)7710201419. Eꢀmail: john.leonard1@astrazeneca.com.
#Current address; Paterson Institute for Cancer Research, The University of Manchester, Wilmslow
Road, Manchester, M20 4BX.
†Institute of Process Research and Development (iPRD) and School of Chemistry, University of Leeds,
LS2 9JT, U.K.
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 33
Table of Contents Graphic
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
ACS Paragon Plus Environment
2

Page 3 of 33
Organic Process Research & Development
Abstract
1
2
3
4
5
6
7
8
9
The use of the “borrowing hydrogen strategy” in the synthesis of a number of typical pharmaceutical
intermediates has been investigated. The main aim of this work was to investigate the scope and
limitations of current methodology using standard laboratory techniques in an industrial context. Some
interesting and significant results were achieved across a diverse set of complex substrates; however
several drawbacks with this approach were identified, such as the high loading, poor turnover and
susceptibility to substrate inactivation of the catalysts. These are areas which are highlighted for future
investigation and improvements.
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
Keywords: Commercial reagents and literature conditions used to evaluate “Borrowing hydrogen”
methodology for pharmaceutical intermediates.
Yields are high for simple R/R’, but challenges identified for typical pharma substrates.
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 33
Introduction
1
2
3
4
5
6
7
8
9
A crossꢀindustry survey published in 2006 revealed that ‘heteroatom alkylation and arylation’ is the
most common category of reaction used by process R&D functions in the preparation of drug candidate
molecules.1 Analysis showed that Nꢀsubstitution accounts for 57% of such transformations, and of these
36% are achieved by the reaction of an amine with an alkylating agent and 20% by reductive amination.
The ACS Green Chemistry Institute Pharmaceutical Roundtable has also published a perspective on key
research areas.2 Activation of alcohols for nucleophilic substitution was one of the most important areas
in the category of ‘currently used reactions that require improvement’.2 These two papers highlight the
fact that Nꢀalkylation is one of the most commonly employed transformations in the pharmaceutical
industry, even though these reactions are inherently inefficient. Whereas direct coupling of an amine
with an alcohol would lead to water as the only byꢀproduct, conversion of an alcohol to an alkylhalide,
or other activated form, adds at least one additional chemical step and leads to undesirable waste
products. Reductive amination is a common alternative to alkylation but this may employ reagents that
have poor atom economy (e.g. stoichiometric borohydride salts) and additional chemical steps are also
required for alcohol oxidation and then reduction of an imine intermediate. Hydrogenation can
sometimes be an atom efficient alternative for the reduction step, but the additional alcohol oxidation is
still required and substrate scope is limited. Poor chemoselectivity is a general problem for reductive
aminations and the risk of byꢀproduct formation through overꢀalkylation is also a limitation of both
alkylation and reductive amination methodology, which can compromise yield and may lead to
additional purification procedures. Alkylating reagents may also have genotoxic properties (PGIs), so
in the preparation of pharmaceutical intermediates their use is stringently controlled to ensure that they
are completely purged during the processes involved in active pharmaceutical ingredients (API)
manufacture.3 Any methodology that offers the complete avoidance of alkylating agents in
pharmaceutical manufacture would therefore be beneficial, as the need for dedicated analytical and
purging procedures would be circumvented.
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
ACS Paragon Plus Environment
4

Page 5 of 33
Organic Process Research & Development
Direct nucleophilic substitution of an alcohol by amine is an attractive alternative to alkylation, as water
would be the only byꢀproduct and alcohols themselves are usually nonꢀtoxic, readily available and
inexpensive. First reported in 1981,4 in recent years there has been a resurgence of interest in using
homogeneous catalysis to achieve direct Nꢀalkylation of amines using alcohols as the formal
electrophilic partner (Scheme 1).5 This redoxꢀneutral approach to alkylation belongs to the class of
reactions termed ”borrowing hydrogen” and is potentially of significant interest to the pharmaceutical
industry in that alkylating agents or the stoichiometric use of unstable aldehydes (in reductive
aminations) might be avoidable.
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
Scheme 1. “Borrowing Hydrogen” Approach to Alkylation of Amines with Alcohols
While the “borrowing hydrogen” approach has been widely studied from an academic perspective, there
are a number of challenges that need to be addressed before this reaches the status of a general
technique which could be readily adopted by the pharmaceutical industry. The majority of “borrowing
hydrogen” reactions are conducted at high temperatures, often with high loadings of precious metal
catalysts, and high concentrations of reactants in relatively nonꢀpolar aromatic solvents such as toluene
and xylene. Whilst this is satisfactory for smallꢀscale reactions using simple, low molecular weight
(often liquid) reagents, the technique may not necessarily transfer to largeꢀscale reactions employing
substrates of the type most commonly encountered in pharmaceutical development programme: high
molecular weight, densely functionalised and often solid, with limited solubility. Some progress has
been made in addressing the shortcomings, for example the use of water as a solvent has been successful
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 33
in some reactions,6 and more active catalysts enabling lower temperature reactions (down to roomꢀ
temperature under solventꢀfree conditions) have been described.7 Furthermore, a Pfizer research team
have recently described the optimisation and scaleꢀup of an iridiumꢀcatalysed alkylation of a benzylic
amine with a primary alcohol.8 This work identified an advantage in carrying out the reaction in
pressure equipment to allow for heating above the azeotrope temperature of toluene/water, providing a
higher and more consistent yield, and allowing catalyst loadings to be reduced to <0.05 mol% iridium.
This groundbreaking application of the technology notwithstanding, the substrates for this specific
reaction are not necessarily representative of the breadth of substrates likely to be encountered across
the industry. The substrates used in this example are soluble to high concentrations in a nonꢀpolar
solvent (toluene) and, unlike many pharmaceutical and agrochemical intermediates, they contain
relatively little by way of functionality that might compete as metal binding sites. The current paper
describes a programme of work aimed at evaluating the scope and performance of “borrowing
hydrogen” chemistries with complex substrates from across a range of AstraZeneca development
projects. We believe that the results of these broad scoping studies are complementary to those of the
Pfizer group, and will both provide further support for the general applicability of a hydrogen borrowing
approach to the synthesis of APIs, and also point out where limitations could benefit from more
fundamental research.
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
Results and Discussion
A survey of the AstraZeneca portfolio of development compounds revealed a number of examples
containing amines as structural subꢀunits which were constructed either by reaction of an amine with an
alkylating agent or by reductive amination. Several of these examples appeared amenable to a
borrowing hydrogen sequence as an alternative. These were considered for inclusion in the study on the
basis of availability of appropriate substrates and an assessment of the apparent viability of the resulting
alternative route to the API.
ACS Paragon Plus Environment
6

Page 7 of 33
Organic Process Research & Development
From the outset, in each case, we wished to assess quickly whether a ‘borrowing hydrogen’ approach
would be viable using simple standard laboratory techniques, avoiding any requirement for sealed tube
or pressure equipment where possible. If preliminary reactions proved successful, optimisation was
considered on the basis of project priority; where unsuccessful, some investigation into possible causes
was undertaken. The primary emphasis was to screen suitable examples against available catalyst
systems rather than to develop bespoke conditions for each. In this way we hoped to gain further insight
into the scope and limitations of this redox neutral methodology for amine synthesis in the context of
typical API preparations.
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
Our strategy was to focus mainly on the catalyst systems developed by Williams (2.5 mol% [Ru(pꢀ
cymene)Cl2]2/5 mol% DPEphos/1M in toluene – “Catalyst A”)9 and by Fujita and Yamaguchi (2.5
mol% [IrCp*Cl2]2/5 mol% NaHCO3/10M in toluene – “Catalyst B”)10. The rationale behind this was
that a broad range of substrates have been shown to be compatible with both of these systems.
Moreover, they are both appealing from a process development perspective: the catalyst/ligand
combinations are available ‘offꢀtheꢀshelf’, airꢀstable, and both protocols usually employ a 1:1 ratio of
amine : alcohol. Exceptions to this were made where alternate catalyst systems had been reported to be
applicable to closelyꢀrelated substrates, in which case these were prioritised for investigation. In this
work, both the amines and alcohols studied were often novel and therefore little was known about the
likely compatibility of the substrates under the reaction conditions. Model reactions were often
therefore employed in order to ascertain whether the coupling partner in question would undergo
reaction with an amine or alcohol known to participate in ‘borrowing hydrogen’ type alkylations e.g.
piperidine or benzyl alcohol, respectively.
For clarity of presentation, the various case studies are collated according to the type of alcohol being
used as the alkylating agent.
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 33
Case Study 1 – Methanol
1
2
3
4
5
6
7
8
9
Amino alcohol 1 is a final intermediate in the synthesis of potent SRC kinase inhibitor AZꢀI 2 and is
introduced via baseꢀmediated displacement of the aryl fluoride 3 prior to formation of the difumarate
salt of the API (Scheme 2).11
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
Scheme 2. Role of Piperazinyl Alcohol 1 in the Synthesis of AZ-I (2)
Amine 1 had previously been prepared by one of two conventional approaches: alkylation of Nꢀ
methylpiperazinewith 2ꢀchloroethanol or reductive methylation of piperazine 4 with formaldehyde
(Scheme 3). The reductive amination approach was eventually optimised to provide the product in high
yield (93%), under carefully controlled conditions, involving removal of excess formaldehyde by
conversion to volatile diethylmethylamine, by a sequential reductive amination with diethylamine. An
alternative approach to the latter transformation based upon borrowing hydrogen reactions of methanol
was therefore investigated.
There are relatively few examples of Nꢀmethylations using methanol under borrowing hydrogen
conditions using homogeneous metal catalysts, and relatively forcing conditions are usually required,
presumably reflecting the less favourable oxidation of methanol relative to higher alcohols.4a,12,13
However, Del Zotto has reported that ruthenium(II) cyclopentadienyl complexes promote Nꢀmethylation
o
of amines in methanol solvent in a sealed vessel at 100 C.14 Applying modified conditions (reaction
carried out in an open vessel at reflux) to piperazine 4 resulted in smooth conversion of starting
material to product, allowing straightforward isolation by distillation in excellent yield (89% on 60
mmol scale). Reducing the catalyst charge from 1 mol% led to unacceptably long reaction times
ACS Paragon Plus Environment
8

Page 9 of 33
Organic Process Research & Development
(incomplete after 24 hours at reflux). Interestingly, no alkylation was observed if alcohols other than
methanol were used (e.g. ethanol, isopropanol, benzyl alcohol) which may explain why no selfꢀ
condensation of the hydroxylꢀbearing piperazine 4 was observed. Use of this catalyst may be generally
applicable to methylation of secondary amines and provide a useful alternative to methods using
potentially carcinogenic formaldehyde. It is noteworthy that Del Zotto also reports that primary amines
bisꢀmethylate but that anilines are unreactive.14
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
Scheme 3. N-Methylation Routes to Amino Alcohol 1
Case Study 2 – Primary Alcohols
Aniline 5 is a significant subꢀunit of antiꢀhepatitis candidate AZꢀII 6, a potent inhibitor of hepatitis C
virus replication. Compound 5 is currently prepared by alkylation of sulfonamide 7 (Scheme 4)15 with pꢀ
nitrobenzyl bromide 8. Reduction of the nitro group in the product 9 then furnishes the corresponding
aniline 5, ready for acylation to assemble the drug substance.
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 33
Scheme 4. Role of Aniline 5 in the Synthesis of AZ-II (6)
1
2
3
4
Br
5
6
7
8
HN
N
O2N
8
NSO2Pr
NSO2Pr
O2N
9
7
.TFA
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
H2, Pt/C
O
N
CF3
NSO2Pr
N
O
H
N
NSO2Pr
H2N
5
HN
6 (AZ-II)
O
The amination of pꢀnitrobenzyl alcohol 10 with aniline under borrowing hydrogen conditions has been
reported using iridium catalysts,6d,10,16 but initial investigations aimed at a potential borrowing hydrogen
approach towards aniline 5, using 10 with piperidine as a model amine, were not encouraging (Scheme
5a). Very little reaction was observed with Catalyst A and although some reaction occurred with
Catalyst B there was also evidence for reduction of the nitro group: the use of ruthenium17 and iridiumꢀ
based18 catalysts to effect tandem nitroarene reduction/alkylation using alcohols as both the terminal
reductant and alkylating agent is documented. However, we were unable to achieve clean reduction of
the nitro group in this reaction. An alternative strategy based upon alkylation of a preꢀformed aniline
was therefore sought to circumvent the incompatibility of the reactive nitro function with the catalyst
system. Direct alkylation of pꢀaminobenzyl alcohol was discounted on the basis that competing
alkylation of the unprotected amine may occur but pleasingly, the Bocꢀprotected analogue 11 performed
well in model studies in the alkylation of piperidine, using either Catalyst A or B. Each catalyst (2.5
mol%) was trialled in toluene at 125 °C (sealed tube) for 17 h, after which time no starting alcohol was
apparent, and both catalysts gave satisfactory levels of conversion (79% and 71% respectively) as
ACS Paragon Plus Environment
10

Page 11 of 33
Organic Process Research & Development
assayed by NMR analysis using TCNB (1,2,4,5ꢀtetrachloroꢀ3ꢀnitrobenzene) as an internal standard
(Scheme 5b). We were therefore encouraged to examine the behaviour of the monosulfonyl piperazine
7 found in the real system.
1
2
3
4
5
6
7
8
9
Scheme 5. Model Alkylations of p-Nitrogen-substituted Benzylic Alcohols with Piperidine
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
OH
10
O2N
N
NH
NH
(a)
(b)
O2N
2.5 mol% Cat A or Cat B,
Toluene, 125 oC, 17 h
Not observed
N
OH
11
BocHN
BocHN
2.5 mol% Cat A or Cat B,
Toluene, 125 oC, 17 h
Cat A 79% (NMR)
Cat B 71% (NMR)
Initial experiments assessing the alkylation involved separate preparation of the free base of 7 prior to
use in model alkylation reactions with benzyl alcohol. However, results with either Catalyst A or B
were capricious. This led to speculation that residual TFA (as a result of the freeꢀbasing procedure) or
another trace impurity in sulfonamide 7 may be influencing the outcome of the reaction. Williams has
reported the direct use of dimethylammonium acetate as a substrate in borrowing hydrogen alkylations9
so attempts were made to use the TFA salt 7 directly. Pleasingly, it was shown that 7 could be alkylated
efficiently with benzyl alcohol in the presence of 0.75 eq DIPEA when using Catalyst A. The same
conditions were then applied to effect alkylation of TFA salt 7 with alcohol 11 in excellent yield (92%
on a 10 mmol scale); deprotection of intermediate 12 using TFA in dichloromethane then returned the
required aniline 5 (89%, Scheme 6). This represents an important proofꢀofꢀprinciple for the application
of borrowing hydrogen methodology in the synthesis of polyfunctional drug intermediates.
Additionally, the ability to use amine salts directly in a borrowing hydrogen approach is very attractive
given that amine intermediates are frequently isolated as salts. A further important observation from a
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 33
process development perspective is that significant coloration is observed during the course of these
reactions. In this example, dark brown coloration remained in the product even after silica gel
chromatography. This may have implications with regard to both metal contamination and carryꢀ
through of colour to API.
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
Scheme 6. Borrowing Hydrogen Approach to Intermediate 5.
We next examined the alkylation of a more highly functionalized amine with a simple aliphatic primary
alcohol. Anaesthetic Ropivacaine (13) may be prepared by alkylation of piperidine 14 with
bromopropane (Scheme 7).19 Replacement of the alkylating agent at this late stage of the synthesis
would be attractive and we therefore investigated the use of a borrowing hydrogen reaction with
propanol.
Scheme 7. N-Alkylation Route to Ropivacaine (13)
However, attempts to effect this transformation were unsuccessful using either Catalyst A or B. In all
cases, analysis by GC/GCMS revealed essentially unreacted starting materials. However, trace levels of
two components were also observed, each with MH+ 273 i.e. 2 less than that required for piperidine 13
ACS Paragon Plus Environment
12

Page 13 of 33
Organic Process Research & Development
(MW = 274). We therefore suspected that the initial alcohol oxidation step had occurred but that an
alternative pathway was preventing return of the hydrogen back to the desired iminium intermediate,
thus stalling the reaction. We therefore investigated the reaction of 14 with propanal, and observed
clean conversion to the two products with MH+ = 273 that had been observed as the minor components
above. Subsequent treatment with sodium triacetoxyborohydride failed to effect conversion to the
desired product, supporting the notion that the resistance of this intermediate towards reduction was
responsible for the borrowing hydrogen reaction stalling. The structure of the major component was
shown to be aminal 15, as confirmed by single crystal xꢀray diffraction (Figure 1 – enantiomer shown).
The minor component was not isolated but was presumed to be the diastereomeric aminal 16, given the
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
1
similarity between its H NMR spectrum and that of aminal 15 (Scheme 8). It therefore seems that in
the catalyzed process, nucleophilic capture of the iminium ion derived from reaction of 14 with propanal
by the pendant amide to form the mixture of aminals was faster than intermolecular reduction by the
relatively low concentration of metal hydride species in solution, and that once formed the aminals are
resistant to ringꢀopening. To probe this further, piperidine 14 was treated with propanal in a ‘direct’
reductive amination using sodium triacetoxyborohydride,20 i.e. with all components mixed without prior
formation of the iminium ion. In this case, Ropivacaine 13 was returned quantitatively: evidently, the
relatively high concentration of the active hydride reducing agent in this case can compete effectively
with capture by the amide group. The results obtained suggest that borrowing hydrogen approaches
may not be applicable to certain substrates bearing nucleophilic functionality close to the intermediate
iminium ion, although the possibility remains that catalyst tuning to find a species more active in the
reduction of iminium ions than systems A or B might lead to success.
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 33
Scheme 8. Iminium Generation and Competing Internal Nucleophilic Capture vs Reduction
1
2
3
4
5
6
7
8
9
i) nPrOH, 2.5 mol% [Ru(p-cymene)Cl2]2,
5 mol% DPEphos, Toluene, heat
H
N
H
N
H
O
O
H
N
N
N
+
N
H
or
O
ii) EtCHO/ DCM/ RT
(no in situ reducing agent)
14
15
16
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
NaBH(OAc)3 (1.4 eq),
DCM
EtCHO, DCM,
in situ NaBH(OAc)3
H
O
H
N
N
13
Figure 1. X-Ray Crystal Structure of Aminal 15
Case Study 3 – Secondary Alcohols
We next examined the applicability of borrowing hydrogen reactions of secondary alcohols, proceeding
via the generally less electrophilic ketones. Piperazine dihydrochloride salt 17 is a key intermediate in
the preparation of the histamine H3 receptor antagonist AZꢀIII 18 (Scheme 9). It is generated by
reductive amination of NꢀBoc piperazine 19 with cyclobutanone, followed by deprotection; acylation
with acid 20 using EDCI gives 18.21
ACS Paragon Plus Environment
14

Page 15 of 33
Organic Process Research & Development
Scheme 9: Synthesis of AZ-III (18)
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
We therefore elected to examine the behaviour of 19 in a borrowing hydrogen alkylation with
cyclobutanol. Pleasingly, good conversion was observed by GC analysis, with catalyst A giving
superior performance over catalyst B (Table 1). These initial screening experiments were run at a
loading of 2.5 mol%: attempts to reduce the catalyst loading gave unreliable results. In line with the
current route, the intermediate Bocꢀprotected amine 21a was not isolated; addition of HCl/IPA to the
reaction mixture effected deprotection and precipitation of the dihydrochloride salt 17 in near
quantitative yield (Scheme 10). Coloration of the product, presumably due to residual metal, was
observed to be a problem in this unoptimised procedure, but it was considered that this reaction would
have been a viable alternative for scaleꢀup and manufacture of intermediate 17.
Scheme 10. Borrowing hydrogen route to piperazine 17
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 33
1
2
3
4
5
6
7
8
9
This facile combination of a secondary amine and secondary alcohol was very pleasing but unexpected,
so we decided to explore the breadth of the reaction by studying the reaction of Bocꢀpiperazine 19 with
other cyclic secondary alcohols The use of secondary alcohols with Catalyst A is reported to be more
challenging than when using primary alcohols and requires more forcing conditions (refluxing xylene).9
We therefore investigated the reaction of 19 with a range of cyclic alcohols, analysing for conversion by
GC (Table 1). For catalyst A, cyclobutanol was found to be a markedly better substrate than higher
homologues, while for catalyst B cyclopentanol performed best. It can be anticipated that the oxidation
of cyclobutanol is more challenging than for higher cyclic alcohols owing to the ring strain associated
with introducing an sp2 centre into a small ring. The resulting ketone will be notably more electrophilic
than less strained congeners, and thus hemiaminal formation would be favoured, but elimination to form
the iminium ion (frequently the rateꢀlimiting step in imine formation22) would again be more difficult
than for higher alcohols. Finally, reduction of the cyclobutanoneꢀderived iminium would be expected to
be more facile owing to relief of ring strain. These results suggest that for Catalyst A at least the latter
step is significant in the overall rate of the reaction, and cyclobutanol can perhaps be regarded as an
‘outlier’ in terms of secondary alcohol performance. Indeed, the very clean reaction observed here
between a secondary amine and cyclobutanol may provide a useful approach for preparation of
cyclobutylamines.
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
Table 1. Alkylation of Piperazine 19 with Cyclic Alcohols
ACS Paragon Plus Environment
16

Page 17 of 33
Organic Process Research & Development
Alcohol
n
Product
Cat A (1M)
Cat B (10M)
1
2
3
4
5
6
7
8
9
10
% Conversiona
% Conversiona
Cycobutanol
Cyclopentanol
Cyclohexanol
Cycloheptanol
1
2
3
4
100
54
8
44
84
57
70
21a
21b
21c
21d
35
a Determined by GC Area
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
Case Study 4 – 1,n-Diols
As detailed above, amino alcohol 1 is an intermediate in the synthesis of AZꢀI (2).11 It has previously
been prepared by alkylation of Nꢀmethylpiperazine with 2ꢀchloroethanol. Given the prevalence of betaꢀ
amino alcohols as targets or intermediates in pharmaceutical synthesis, we were interested to investigate
whether Nꢀalkylation of amines with ethylene glycol under borrowing hydrogen conditions9,12c,f,23 was a
viable alternative to the use of 2ꢀhaloethanols. We therefore attempted to aminate ethylene glycol with
Nꢀmethylpiperazine 22 using a modified literature procedure.23a Piperazine 22 was rapidly consumed to
return an 88:12 ratio (as determined by GC) of the desired amino alcohol 1 and the unwanted tetramine
23 arising from amination of both alcohols of the ethylene glycol (Scheme 11). It is possible that the
ratio could have been improved with further optimisation but this approach was not pursued with this
particular substrate as the alternative methylation approach described in Scheme 3 appeared more
promising in terms of ease of purification.
Scheme 11. Amination of Piperazine 22 with Ethylene Glycol
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 33
Given this encouraging result, the same approach was investigated towards the synthesis of
investigational compound AZꢀIV (24) by amination of ethylene glycol with the advanced piperazine
intermediate 25, but in this instance we were disappointed to find that no trace of product was observed
(Scheme 12). It was postulated that the prevalence of potential bidentate metal binding sites in
piperazine 25 might deactivate the catalyst. This was supported by a spiking experiment where it was
shown that the reaction of piperazine 22 and ethylene glycol no longer proceeded in the presence of
piperazine 25. This highlights a general problem area when considering a borrowing hydrogen strategy:
catalyst poisoning is more likely when using densely functionalised intermediates.
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
Scheme 12. Failed Borrowing Hydrogen Approach to AZ-IV (24)
We next investigated the use of higher diols in amine alkylation chemistry. The use of 1,nꢀdiols to
effect cyclative dialkylation of primary amines has been well documented, 6,7b,9,24 and we were keen to
investigate whether this would be efficient in the context of an advanced pharmaceutical intermediate.
(S)ꢀAminoester AZꢀV (26) is an investigational muscarinic antagonist which is prepared from
enantiomerically pure piperidine 27, itself prepared as a racemate by substitution of methyl 2ꢀbromoꢀ2ꢀ
phenylpropionate 28 with piperidine, followed by chromatographic resolution of the enantiomers
(Scheme 13).25
ACS Paragon Plus Environment
18

Page 19 of 33
Organic Process Research & Development
Scheme 13. Current Route to AZ-V (26)
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
Given the availability of enantiomerically pure (S)ꢀamino ester 29, we were intrigued to see whether
cyclative double alkylation could be effected under borrowing hydrogen conditions to give 27. The
formation of benzyl piperidine by reaction of benzylamine with 1,5ꢀpentanediol is known using catalyst
B,6 and we verified that this reaction was successful in our hands (89% benzyl piperidine by GC area,
unoptimised) before examining the hindered quaternary amino ester 29. As expected, the cyclisation is
adversely affected by increased steric bulk around the amine, returning only traces of the desired (S)ꢀ
piperidine 27 (5% by GC area) (Scheme 14).
These reactions were all performed in toluene at 1M concentration whereas many reactions using
Catalyst B are run under much more concentrated conditions e.g. 10M. Therefore, the reaction was
repeated at 10M concentration and also in the absence of solvent which greatly improved the conversion
to (S)ꢀpiperidine 27 (87% by GC area for neat reaction; incomplete at 10M). However, the isolated
yield from the neat reaction was disappointing (56%) based on the apparent conversion. Analysis of the
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 33
1
crude product by H nmr was not informative due to a number of overꢀlapping signals. The reaction
was repeated with an internal standard present and the GC yield was determined to be 68%, whereas the
profile seemed much better (92% by GC area). This led to speculation that involatile byꢀproducts, not
detectable by GC, were being formed. This is perhaps not surprising since the reaction is performed
neat: at higher concentrations undesired intermolecular alkylations are more likely to compete with the
desired intramolecular cyclisation step. The reaction was scaled to 20 mmol and subjected to an
extractive workꢀup to return (S)ꢀpiperidine 27 (3.67 g, 87% GC area; theoretical yield 4.94 g). Clearly,
additional optimisation is required but this is a very encouraging result given the steric demands of the
substrate.
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
Scheme 14. Borrowing Hydrogen Approach to (S)-Piperidine 27
Case Study 5 – Glycolic Acid Derivatives
Access to 2ꢀsubstituted acetamide 30 was required as part of an investigation into alternative synthetic
routes to the orally active antiꢀproliferative agent AZꢀVI difumarate (31). Compound 30 is currently
obtained by alkylation of 4ꢀaryloxypiperidine 32 with chloroacetamide 33 (Scheme 15).26
ACS Paragon Plus Environment
20

Page 21 of 33
Organic Process Research & Development
Scheme 15. Current Route to AZ-VI (31)
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
The replacement of the chloroacetamide 33 with the glycolic amide 34 was therefore examined. This is
a challenging substrate for borrowing hydrogen alkylation, given the likely difficulties in the oxidation
of glycolic derivatives to the highly reactive glyoxylic derivatives, and indeed there is only one report of
a method for such transformations in the literature.27 In fact, no reaction at all was observed when using
Catalyst A with alcohol 34. Some reaction was observed using Catalyst B, however, and the best
conditions (Scheme 16) returned the desired acetamide 30 in 45% yield, after purification by
chromatography (on 2 mmol scale). While clearly still requiring some optimization, this is a highly
challenging transformation and illustrates further the potential for replacement of highly reactive
halides.
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 33
Scheme 16. Amination of Glycolic Amide 34
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
Case Study 6 – Amino Alcohols
Triamine 35 is an intermediate in the synthesis of novel antiarrhythmic agent AZꢀVII (36), and is
currently prepared by alkylation of monoprotected diamine 37 with mesitylsulfonate ester 38 (Scheme
17).28 Replacing the mesitylsulfonate with a direct alkylation using NꢀBoc ethanolamine 39 would
therefore save a synthetic operation (sulfonate ester formation) and we therefore investigated the
feasibility of the transformation. This example is also a key exercise to challenge the scope of the
hydrogen borrowing catalyst systems, which appear to be compromised by substrates containing
multiple heteroatoms (nitrogen and/or oxygen), particularly when they are separated by 2 carbon atoms.
Such substrates can potentially chelate the metal centre of the catalyst, leading to catalyst ‘poisoning’,
as well as suffering from potential sideꢀreactions associated with tautomerisation of αꢀheterosubstituted
aldehydes and imines. 9
ACS Paragon Plus Environment
22

Page 23 of 33
Organic Process Research & Development
Scheme 17. Current Route to AZ-VII (36)
1
2
3
4
5
O
O
S
NHBoc
O
6
7
8
9
38
BnN
O NH
37
BnN
O N
NHBoc
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
35
F
O
N
O N
NHBoc
36
NC
For the preparation of our desired intermediate 35 we required to react together substrates 37 and 39;
however, since each of these compounds, as well as the product 35, have the potential to compromise
the catalysts, we decided to carry out model studies to test the viability of each substrate separately in
hydrogen borrowing reactions (Scheme 18). In the first instance we examined the reaction of diamine
37 with benzyl alcohol, which is known to be a good substrate in redoxꢀneutral alkylations. Both the
ruthenium system (Catalyst A) and the iridium system (Catalyst B) performed well, providing almost
complete conversion to dibenzylated product 40 after 19 h at reflux in toluene. As noted for previous
reactions, a higher concentration of the reaction with the iridium catalyst (10M) was required to give a
similar rate as to the ruthenium catalyst at 1M concentration. The conversions for this reaction were
estimated to be between 80ꢀ90% based on GC/GCMS and on NMR analysis of the crude reaction
products and we judged that amine 37 is a suitable substrate for hydrogen borrowing alkylation with
simple alcohols, even though it contains 3 heteroatoms conformationally disposed towards bidentate
coordination modes.
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 33
In a second model reaction we tested the viability of NꢀBocꢀethanolamine 39 as an alcoholic substrate
for hydrogen borrowing alkylation. Using piperidine as a simple model amine substrate, no significant
reaction was observed using iridiumꢀbased Catalyst B, but some conversion to the desired 41 was
observed with rutheniumꢀbased Catalyst A, estimated to be about 65% conversion by GC. This suggests
that protected ethanolamine 39 is a more problematic substrate than standard aliphatic alcohols.
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
Scheme 18. Model Alkylation Reactions with Amine 37 and Alcohol 39
Following this modest success with model studies, the best conditions identified were applied to the
desired reaction of 37 with 39. The reaction mixture was complex and contained only trace quantities of
a compound identified as the desired protected triamine 35 by GC/MS comparison with an authentic
sample (Scheme 19). The main new product in the reaction mixture was identified as dibenzylamine 40
by comparison of GCMS data and NMR signals in the crude reaction mixture to data from a sample
produced from the previous model reaction between diamine 37 and benzyl alcohol. The formation of
diamine 40 is informative. From the previous model study, using piperidine, it is clear that N-Boc ꢀ
ethanolamine 39 is not a particularly good borrowing hydrogen substrate and it seems that the
ACS Paragon Plus Environment
24

Page 25 of 33
Organic Process Research & Development
dehydrogenation of benzylamine 37 occurs in preference to (or competition with) oxidation of the
alcohol. Iminium ion 42 might then act as an alkyating agent for 37, to provide dibenzylamine 40
1
2
3
4
5
6
following hydride transfer.
Dehydrogenation and subsequent dimerisation has previously been
demonstrated for primary and secondary amines.16 The potential for amine dehydrogenation to compete
(particularly with activated benzylic amines) when alcohol dehydrogenation is problematic and may be
a general issue that needs to be considered in planning borrowing hydrogen alkylations of
polyfunctional substrates.
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
Scheme 19. Attempted Synthesis of Amine 35
Conclusions
In this broad scoping study, we hope to have demonstrated some of the opportunities and current
limitations of borrowing hydrogen chemistry for the alkylation of amines avoiding potentially genotoxic
alkyl halide or sulfonate intermediates. The impressive application of this technology by Pfizer in the
synthesis of PFꢀ03463275 notwithstanding,8 it appears that the demands of complex, polar
polyfunctional intermediates common to pharmaceutical targets do still impose some limitations on the
broad applicability of the chemistry. Nevertheless, we have found significant encouragement from
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 33
some key examples that this technology has the potential to deliver improved processes for
pharmaceutical manufacture. The method still offers good opportunities for avoiding alkylating agents
as intermediates and works relatively reliably for simple aliphatic alcohols (MeOH, 1ry and 2ry).
Structural incompatibilities with the chemistry are more obviously visible, however, in complex
polyfunctional molecules typified by pharmaceutical intermediates rather than the relatively simple
lipophilic substrates typically examined in methodological development studies. The catalyst
incompatibilities with heteroatom rich heterocyclic structures were demonstrated eg by spiking
experiments that show catalyst poisoning by 25 in the otherwise successful alkylation of piperazine 22
by ethylene glycol. Additionally, neighbouring group participation in polyfunctional molecules can lead
to interception of reactive intermediates, as in the case of formation of the stable imidazolidinone 15.
Finally, the reaction conditions employed in this scoping study highlight the need for more active and
tolerant catalysts, capable of operating at low loadings (from the standpoints of both economics and
metal residue issues –as highlighted here by colouration issues), and across a broad range of solvents
(such as those often required for relatively insoluble advanced intermediates) and processꢀrelevant
concentrations. We hope that this study will serve to encourage others in the process chemistry
community to investigate the adoption of this technology, and to challenge those engaged in catalyst
development to investigate the next generation of hopefully more active and robust catalysts that can
tolerate the demands of the molecules outlined herein.
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
Experimental Section
General. Starting materials, reagents and solvents were obtained from commercial suppliers and used
without further purification. NMR spectra were obtained using Bruker DPX 400, DRX 500 and Avance
600 instruments; 1H spectra were measured with reference to an internal standard of tetramethylsilane at
13
0 ppm and C spectra were measured with reference to the CDCl3 signal at 77.0 ppm or the DMSO
signal at 39.5 ppm. GC analyses were performed with an Agilent 6890N instrument and HPLC analyses
ACS Paragon Plus Environment
26

Page 27 of 33
Organic Process Research & Development
with an Agilent 1100 instrument. High resolution mass spectra were run on a Bruker MicrOTOFQ
LCMS instrument and the major ion was analysed using positive ion electrospray ionisation.
1
2
3
4
5
2-(4-Methylpiperazin-1-yl)ethanol (1) [RuCl(Cp)(PPh3)2] (436 mg, 0.6 mmol, 1 mol%) was added to
a solution of piperazine 4 (7.81 g, 60.0 mmol) in methanol (240 ml) under N2. The reaction was stirred
at RT for 10 minutes, then heated to reflux (bath at 85 °C) for 24 hours. Methanol was removed on the
rotary evaporator and the residue was then purified by bulbꢀtoꢀbulb distillation under reduced pressure
(oven at 110 °C; ca 2 mbar) to return amino alcohol 1 (7.68 g, 89% th) as a pale yellow oil.
Spectroscopic analysis was in agreement with the reported data.11
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
4-((4-Propylsulfonyl)piperazin-1-yl)methyl)aniline (5) [Ru(pꢀcymene)Cl2]2 (156 mg, 0.25 mmol, 2.5
mol%), DPEphos (275 mg, 0.50 mmol, 5.0 mol%), sulphonamide trifluoroacetate salt 7 (3.06 g, 10.0
mmol), alcohol 11 (2.23 g, 10.0 mmol), diisopropylyethylamine (0.70 g, 7.50 mmol) and PhMe (10 ml)
were stirred under N2 at RT for 10 minutes. The reaction was heated at reflux for 21 hours, then cooled
to RT and concentrated under reduced pressure to return the crude product which was purified by flash
column chromatography, eluting with 2% MeOH in DCM, to return intermediate 12 (3.67 g, 92% th) as
a brown oil which slowly solidified. 1H NMR (400 MHz, 300 K, CDCl3) δ 1.05 (3H, t, J7.40), 1.52
(9H, s), 1.81ꢀ1.88 (2H, m), 2.48ꢀ2.51 (2H, m), 2.85ꢀ2.89 (2H, m), 3.26ꢀ3.29 (4H, m), 3.47 (2H, s), 6.47
13
(1H, Br s), 7.18ꢀ7.23 (2H, m), 7.30.7.36 (2H, m); C (100 MHz, 300 K, CDCl3), δ13.18, 16.80, 28.33,
45.75, 50.73, 52.54, 62.14, 80.57, 118.50, 129.75, 132.14, 137.54, 152.76; HRMS (electrospray +ve
ion) calcd for [M+H]+ C19H32N3O4S 398.2108, found 398.2099.
TFA (7.65 mL, 100.0 mmol) was added to a solution of intermediate 12 (3.67 g, 9.23 mmol) in DCM
(50 mL) at RT. The reaction was stirred overnight then concentrated under reduced pressure. The
residue was redissolved in PhMe (50 mL), then extracted with 2M HCl (3 x 50 mL). The combined
aqueous phases were washed with DCM (50 mL), then basified with 2M NaOH and extracted with
DCM (3 x 50 mL). The combined organic phases were dried over MgSO4, filtered and concentrated
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 33
1
under reduced pressure to return aniline 5 (2.45 g, 89% th) as a brown oil which slowly solidified. H
NMR (400 MHz, 300 K, d6ꢀDMSO ) δ 1.04 (3H, t, J7.60), 1.74 2H, sex, J7.60), 2.39ꢀ2.48 (4H, m), 3.05
(2H, t J7.60), 3.14ꢀ3.26 (4H, m), 3.36 (2H, s), 5.04 (2H, s), 6.56 (2H, ~d, J8.20), 6.98 (2H, ~d, J8.20);
13C (100 MHz, 300 K, CDCl3) δ12.80, 16.30, 45.20, 48.95, 51.85, 61.45, 113.75, 124.20, 129.70,
147.65. Other data was also in agreement with the reported data.15
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
1-Cyclobutylpiperazine dihydrochloride (17) [Ru(pꢀcymene)Cl2]2 (383 mg, 0.63 mmol, 2.5 mol%),
DPEphos (673 mg, 1.25 mmol, 5.0 mol%), NꢀBoc piperazine 12 (4.66 g, 25.0 mmol), cyclobutanol
(1.98 g, 25.0 mmol) and PhMe (25 ml) were stirred under N2 at RT for 10 minutes. The reaction was
heated at reflux for 16 hours, then cooled to RT. IPA (42 ml) and 5ꢀ6N HCl in IPA (25 ml, 125ꢀ150
mmol) were added to the brown suspension which was then heated to 50 °C for 2.5 hours. The
suspension was cooled to RT and stirred overnight (for convenience). The solid was isolated by
filtration, displacement washed with IPA (50 ml), then dried under reduced pressure at 50 °C to return
piperazine 14 (5.17 g, 97% th) as a light brown solid. Spectroscopic analysis was in agreement with the
reported data.20
(3S,8aS)-2-(2,6-Dimethylphenyl)-3-ethylhexahydroimidazo[1,5-a]pyridine-1(5H)-one
(15)
Propanal (616 µL, 8.6 mmol) was added to a solution of piperidine 14 (2.00 g, 8.6 mmol) in DCM (30
mL) at RT and stirred overnight. The reaction mixture was concentrated under reduced pressure, then
purified by flash column chromatography, eluting with 1:1 EtOAc:ihexane, to provide an analytical
sample. The cleanest fraction only was concentrated under reduced pressure to return aminal 15 (370
1
mg, 16% th) as a pale yellow solid which was recrystallised from EtOH. H NMR (400 MHz, 300 K,
CDCl3) δ 0.66 (3H, t, J7.5), 1.30ꢀ1.76 (6H, m), 1.92 (1H, m), 2.12 (1H, m), 2.20 (3H, s), 2.25 (3H, s,),
2.26 (1H, m), 2.79 (1H,dt, J10.5, 2.5), 3.13 (1H,dt, J10.5, 3.5), 4.21 (1H, m), 7.04 (1H, m), 7.08ꢀ7.17
13
(2H, m); C (100 MHz, 300 K, CDCl3) δ 9.40, 18.80, 19.20, 23.60, 24.00, 25.05, 25.70, 48.70, 64.85,
80.40, 128.05, 128.45, 128.75, 133.80, 135.80, 138.15, 171.80; HRMS (electrospray +ve ion) calcd for
C17H25N2O, 273.1961, found 273.1965.
ACS Paragon Plus Environment
28

Page 29 of 33
Organic Process Research & Development
(S)-Methyl 2-phenyl-2-(piperdin-1-yl)propanoate (27) (S)ꢀMethyl 2ꢀaminoꢀ2ꢀphenylpropanoate 29
(358 mg, 2 mmol), 1,5ꢀpentanediol (210 µL, 2 mmol), [IrCp*Cl2]2 (40 mg, 0.05 mmol, 2.5 mol%) and
NaHCO3 (9 mg, 0.1 mmol, 5.0 mol%) were stirred under N2 at RT for 10 minutes in a carousel tube.
The tube was then sealed and heated to 125 °C for 18 hours. The reaction was cooled to RT and the
crude product was purified by flash column chromatography, eluting with 9:1 isohexane:EtOAc, to
return piperidine 27 (276 mg, 56% th) as a colourless oil. Spectroscopic analysis was in agreement with
the reported data.25
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
2-(4-(5-Cyano-2-methoxyphenoxy)piperidin-1-yl)-N-methylacetamide (30) [IrCp*Cl2]2 (40 mg,
0.05 mmol, 2.5 mol%), NaHCO3 (9 mg, 0.1 mmol, 5.0 mol%), piperidine 32 (465 mg, 2 mmol), alcohol
34 (178 mg, 2 mmol) and PhMe (2 ml) were stirred under N2 at RT for 10 minutes in a carousel tube.
The tube was then sealed and heated to 125 °C for 21 hours. The reaction was cooled to RT, then
concentrated under reduced pressure to return the crude product which was purified by flash column
chromatography, eluting with 2% MeOH in DCM, to return acetamide 30 (270 mg, 45% th) as a
colourless solid. Spectroscopic analysis was in agreement with the reported data.26
Acknowledgements
We would like to thank AstraZeneca Chemical Development scientists for suggesting the drug project
examples that were investigated during this work and for useful discussions throughout the project.
Supporting Information Available
XꢀRay Crystal Structure data for Aminal 15 is available. This information is available free of charge via
the internet at http://pubs.acs.org/.
ACS Paragon Plus Environment
29

Organic Process Research & Development
Page 30 of 33
References
1
2
3
4
1. Carey, J. S.; Laffan, D.; Thomson, C.; Williams, M. T. Org. Biomol.Chem. 2006, 4, 2337ꢀ2347.
5
6
7
2. Constable, D. J. C.; Dunn, P. J.; Hayler, J. D.; Humphrey, G. R.; Leazer, Jr., J. L.; Linderman, R. J.;
8
9
Lorenz, K.; Maley, J.; Pearlman, B. A.; Wells, A.; Zaks, A.; Zhang, T. Y. Green.Chem. 2007, 9, 411ꢀ
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
420.
3. Robinson, D. I. Org. Proc. Res. Dev. 2010, 14, 946ꢀ959.
4. (a) Grigg, R.; Mitchell, T. R. B.; Sutthivaiyakit, S.; Tongpenyai, N. J. Chem. Soc., Chem. Commun.
1981, 12, 611–612 (b) B.ꢀT. Khai, C. Concilio, G. Porzi, J. Org. Chem. 1981, 46, 1759–1760. (c) Y.
Watanabe, Y. Tsuji, Y. Ohsugi, Tetrahedron Lett. 1981, 22, 2667 –26705.
5. For reviews, see: (a) Hamid, M. H. S. A.; Slatford, P. A.; Williams, J. M. J.; Adv. Synth. Catal. 2007,
349, 1555ꢀ1575 (b) Nixon, T. D.; Whittlesey, M. K.; Williams, J. M. J. Dalton Trans. 2009, 753-762
Yamaguchi, R; Jujita, K.; Zhu, M. Heterocycles 2010, 81, 1093ꢀ1140 (d) Dobereiner, G. E.; Crabtree,
R. H. Chem. Rev. 2010, 110, 681ꢀ703. (e) Guillena, G.; Ramon, D. J.; Yus, M. Chem. Rev. 2010, 110,
1611–1641. (f) Gunanathan, C.; Milstein, D. Science 2013, 341, 249ꢀ257
6. (a) Saidi, O.; Blacker, A. J.; Lamb, G. W.; Marsden, S. P.; Taylor, J. E.; Williams, J. M. J.; Org.
Process Res. Dev. 2010, 14, 1046ꢀ1049; (b) Saidi, O.; Blacker, A. J.; Farah, M. M.; Marsden, S. P.;
Williams, J. M. J.; Chem. Commun. 2010, 46, 1541ꢀ1543; (c) Kawahara, R.; Fujita, K.ꢀi.; Yamaguchi,
R.; J. Am. Chem. Soc. 2010, 132, 15108ꢀ15111; (d) Kawahara, R.; Fujita, K.ꢀi.; Yamaguchi, R.; Adv.
Synth. Cat. 2011 353, 1161ꢀ1168; (e) Qu, P. P.; Sun,. C. L.; Ma, J.; Li, F.; Adv. Synth. Cat. 2014 356,
447ꢀ459.
7. (a) Li, J.ꢀQ.; Andersson, P. G.; Chem. Commun. 2013, 49, 6131ꢀ6133; (b) Enyong, A. B.; Moasser,
B.; J. Org. Chem. 2014¸79, 7553ꢀ7563.
ACS Paragon Plus Environment
30