A Combined High-Throughput Screening and Reaction Profiling Approach toward Development of a Tandem Catalytic Hydrogenation for the Synthesis of Salbutamol

Article abstract of DOI:10.1021/acs.oprd.7b00261

A combined high-throughput screening and reaction profiling approach to the telescoping of two reductions in the synthesis of Salbutamol is described. Optimization studies revealed the beneficial effect of mildly acidic conditions, and the use of water as a cosolvent. Persistent formation of deoxygenated impurities using a Pd/C catalyst led to the initiation of reaction profiling studies, which revealed that the ketone intermediate formed after rapid debenzylation is the sole source of deoxygenated impurities, indicating that more rapid ketone hydrogenation should minimize this deoxygenation. A dual catalyst approach based on these insights has been developed, with both Pd/Pt and Ru/Pt catalyst systems as more selective than Pd-only systems. Based on reaction profiles that indicate the deoxygenation side reaction is first-order in the concentration of debenzylated ketone intermediate, Pt catalysts for rapid and selective ketone hydrogenation were paired with Pd and Ru catalysts known to perform selective debenzylation. Optimization of these dual catalyst processes led to conditions that were demonstrated on 20 g scale to prepare Salbutamol in 49% isolated yield after recrystallization.

Full text of DOI:10.1021/acs.oprd.7b00261

Subscriber access provided by University of Florida | Smathers Libraries
Full Paper
A Combined High-Throughput Screening and Reaction
Profiling Approach Toward Development of a Tandem
Catalytic Hydrogenation for the Synthesis of Salbutamol
David C. Leitch, Thomas F. Greene, Roisin O'Keeffe, Thomas
Lovelace, Jeremiah D Powers, and Andrew D. Searle
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.7b00261 • Publication Date (Web): 19 Oct 2017
Downloaded from http://pubs.acs.org on October 20, 2017
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 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
A Combined HighꢀThroughput Screening and
Reaction Profiling Approach Toward Development
of a Tandem Catalytic Hydrogenation for the
Synthesis of Salbutamol
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
David C. Leitch,*,†,‡ Thomas F. Greene,§ Roisin O’Keeffe,§ Thomas C. Lovelace,‡ Jeremiah D.
Powers,‡ and Andrew D. Searleǁ
†Chemical Catalysis Group, API Chemistry, GlaxoSmithKline, King of Prussia PA. ‡Catalysis
§
Center of Excellence, GlaxoSmithKline, Research Triangle Park, NC. Global
ǁ
Manufacturing and Supply, GlaxoSmithKline, Cork, Ireland. Global Manufacturing and
Supply, GlaxoSmithKline, Stevenage, UK.
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 30
1
2
3
4
Table of Contents artwork.
5
6
7
Legacy Two-Step Route
Salbutamol
OH OH
OH OH
OH OH
8
9
NaBH4
Pd/C, H2
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
HO
Ph
HO
Ph
N
H
N
HN
Cl
Pd/C, Pt/C, H2, mild acid
Direct, Telescoped Process
ACS Paragon Plus Environment
2

Page 3 of 30
Organic Process Research & Development
1
2
3
4
KEYWORDS. Salbutamol, hydrogenation, highꢀthroughput screening, tandem catalysis.
5
6
7
8
9
ABSTRACT. A combined highꢀthroughput screening and reaction profiling approach to the
telescoping of two reductions in the synthesis of Salbutamol is described. Optimization studies
revealed the beneficial effect of mildly acidic conditions, and the use of water as a coꢀsolvent.
Persistent formation of deoxygenated impurities using a Pd/C catalyst led to the initiation of
reaction profiling studies, which revealed that the ketone intermediate formed after rapid
debenzylation is the sole source of deoxygenated impurities, indicating that more rapid ketone
hydrogenation should minimize this deoxygenation. A dual catalyst approach based on these
insights has been developed, with both Pd/Pt and Ru/Pt catalyst systems as more selective than
Pdꢀonly systems. Based on reaction profiles that indicate the deoxygenation side reaction is
firstꢀorder in the concentration of debenzylated ketone intermediate, Pt catalysts for rapid and
selective ketone hydrogenation were paired with Pd and Ru catalysts known to perform selective
debenzylation. Optimization of these dual catalyst processes led to conditions that were
demonstrated on 20 g scale to prepare Salbutamol in 49% isolated yield after recrystallization.
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
3

Organic Process Research & Development
Page 4 of 30
1
2
3
4
Introduction.
5
6
7
8
Salbutamol (or albuterol, 3), a potent and short acting β2ꢀadrenergic agonist, is a vital
medicine for the treatment of bronchospasm resulting from a number of conditions, including
asthma and COPD.1 First described in the 1960s,2 Salbutamol is still widely used, and remains
among the most prescribed and highest volume medicines worldwide. Because of its continued
status as a vital medicine, improvements in synthetic efficiency are important to drive down
production costs and reduce waste generation, which will continue to ensure patient access into
the future. As a result, many research groups have reported racemic and enantioselective
syntheses of 3.3
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
Part of an existing synthetic route to Salbutamol is shown in Scheme 1, highlighting the final
steps.3a,b Transformation of the tertiary amino ketone 1 into 3 is achieved by a twoꢀstep
reduction sequence involving NaBH4 mediated ketone reduction and Pdꢀcatalyzed
hydrogenolysis of the benzyl protecting group; the overall yield for this twoꢀstep process is
~55%. While straightforward, both of these reductions could, in principle, be telescoped into a
single, oneꢀpot catalytic hydrogenation step. Replacement of the hydride reduction with a
catalytic step would be especially advantageous; in addition to elimination of stoichiometric
NaBH4, the current workꢀup and isolation protocol for the ketone reduction requires multiple
laborious liquidꢀliquid extractions to completely remove inorganic byproducts. Salmeterol, a
structurally similar β2ꢀadrenergic agonist can be made using such a tandem reduction strategy
with a dual Pd / Pt catalyst.4
ACS Paragon Plus Environment
4

Page 5 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Scheme 1. Synthesis of Salbutamol, with targetted tandem hydrogenation in blue.
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
Based on this established tandem reduction for synthesis of Salmeterol, we reasoned that a
similar strategy could improve the efficiency of Salbutamol synthesis. Herein we describe our
efforts to discover an optimal catalyst system, elucidate the various reaction pathways in this
tandem hydrogenation reaction, and to use these insights to develp a robust, scalable process that
delivers Salbutamol base 3 with equal or superior yield and purity to the current twoꢀstep
method. In order to achieve this, we applied a combined highꢀthroughput screening and reaction
profiling
approach to reaction space exploration. As a result of these smallꢀscale studies, this tandem
reduction sequence using a Pt/Pd dual catalyst system was demonstrated on 20 g scale with 49%
isolated yield after recrystallization.
Results and Discussion.
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 30
1
2
3
4
5
6
7
8
9
The effect of acidity on reaction outcome. In order to rapidly explore the major factors in
this tandem hydrogenation reaction, a statisticallyꢀdriven highꢀthroughput screening approach
was enacted. As part of the preliminary analytical work, we carried out synthesis of markers for
the various anticipated intermediates and byproducts. Since debenzylation of 1 is rapid relative
to ketone hydrogenation (vide infra), the preparation of 2° aminoketone 5 was attempted by
subjecting 1 to 1 atmosphere of H2 with a 5% Pd/C catalyst at room temperature in ethanol
solvent. After only one hour, 1 was completely consumed, giving a mixture of four products as
judged by HPLC analysis. LC/MS data enabled characterization of this mixture, which
contained compounds 3ꢀ6 (Scheme 2, top pathway); these included the desired Salbutamol base
(3), as well as the deoxygenated compound (4) and the respective ketone products (5 and 6).
Deoxygenation of benzylic alcohols is an unfortunately common reaction of heterogeneous
hydrogenation catalysts.5
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
This uncontrolled reactivity was attributed to the fact that the equivalent of acid in 1 was not
neutralized prior to hydrogenation. Subsequent work to prepare a pure sample of 3 followed the
two step borohydride reduction, hydrogenolysis process from Scheme 1. In this case,
debenzylation of 2 was extremely slow under analogous conditions (1 atmosphere H2, 5% Pd/C,
room temperature). Addition of one equivalent of acetic acid led to a dramatically accelerated
reaction, with no observed formation of the desoxy compound 4 (Scheme 2, bottom). From
these simple experiments, it is clear that the acidity of the reaction medium is a major factor in
all three types of hydrogenation: the desired debenzylation and ketone reduction, as well as the
undesired deoxygenation.
ACS Paragon Plus Environment
6

Page 7 of 30
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
Scheme 2. Acid strength effects on hydrogenation.
High-throughput screening. Based on the observation of a pronounced acid effect in our
preliminary experiments, several rounds of highꢀthroughput screening were designed to
modulate acid strength, through both base and solvent identity, as well as investigate the
importance of catalyst type. A library containing ~200 heterogeneous catalysts was employed
for this purpose. Because Pd/C alone is capable of effecting both debenzylation and ketone
hydrogenation, initial screening focused on developing a singleꢀcatalyst system for conversion of
1 to 3. H2 pressure was 2ꢀ4 barg, and temperature fixed at 30 °C.6
Several key insights were gleaned from these screens, which sought to simultaneously
maximize the amount of Salbutamol (3) formed, and minimize the degree of deoxygenation to
form 4 and 6. First, that a variety of Pdꢀbased catalysts are effective for dual reduction with
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 30
1
2
3
4
5
6
7
8
9
good selectivity, without the need for added Pt. In many cases, complete conversion can be
achieved in as little as 3 hours. Second, that the base used to partially neutralize the HCl salt 1 is
critical to achieving high reactivity and selectivity. Strong bases (KOH) induce slow or no
hydrogenation, while very weak bases (KOTFA, KH2PO4) promote significant deoxygenation, to
the point where in some cases 4 is the major product of the reaction. Organic nitrogen bases
(pyridine, N,Nꢀdimethylaniline) reduce reactivity, presumably by reversible inhibition through
coordination to Pd sites at the catalyst surface.7 Sodium and potassium acetate (acetic acid pKa =
4.76, aqueous) were identified as desirable bases in terms of both reactivity and selectivity.
Third, polar, protic solvents are preferred, with the general reactivity order MeOH > EtOH >
iPrOH. Finally, under optimal conditions, the amount of deoxygenation products tracks very
closely to the amount of 3 produced, consistently 4ꢀ7% when using NaOAc or KOAc.
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
With an effective solvent/base combination identified, we subjected our entire catalyst library
to the reaction conditions in an effort to find catalysts that would exhibit decreased
deoxygenation relative to standard 5ꢀ10% Pd/C systems. From these data, the ~200 catalysts
were categorized based on the reactivity observed.6 Unsurprisingly for a library this large and
diverse, nearly half of the catalysts achieved <50% conversion. About 25 systems achieved
selective debenzylation to 5, without appreciable ketone reduction or deoxygenation. This set
included rhodiumꢀ and rutheniumꢀbased catalysts in addition to several palladium systems; in
particular, 5% Ru/Al2O3 was a rather effective alternative to Pd/C for amine debenzylation;
while this Ru catalyst appears to require slightly elevated H2 pressures (no reaction observed at 1
atmosphere), the reduced cost of Ru relative to Pd could be advantageous. Several Ptꢀbased
catalysts were able to achieve modest ketone hydrogenation to 2, though the amount of 3
produced was low. Another ~25 catalysts, all based on Pd, achieved complete conversion, with
ACS Paragon Plus Environment
8

Page 9 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
>89% 3. In addition to several Pd/C systems, Pd catalysts with alternate supports such as
BaSO4, Al2O3, and CaCO3 performed particularly well. These nonꢀcarbon supports generally
have much lower surface area, leading to lower activity relative to Pd/C;8 however, a reduction in
activity could result in greater selectivity. From these ~200 catalysts, ten were chosen for further
study (Table 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
Table 1. Ten standout catalysts from library screen.
Mass
1
3
5
4 + 6
Entry Catalyst
Balance
(%)a (%)a (%)a (%)a
(%)a
1
2
5% Pd/C; E196 R/Wb
0.0
91.2
89.4
92.7
87.8
89.6
9.5
0.0
3.7
0.0
0.0
0.0
5.4
2.8
5.5
5.2
5.0
96.6
95.9
98.2
96.3
96.7
93.2
93.4
94.4
91.3
97.5
5% Pd/C; E213 R/Db
0.0
3
5% Pd/Al2O3; 5R325c
5% Pd/CaCO3; 5R405c,d
5% Pd/BaSO4; 5R29Ac,d
2% Pd/SiO2ꢀAl2O3; 2R31c
5% Ru/C; H105 RA/Wb
5% Ru/Al2O3; 5R698c
5% Ru, 0.25% Pd/C; 5R611c
5% Pd/C; E101 NE/Wb
0.0
4
3.3
5
2.1
6
0.0
83.7 0.0
43.3 0.0
70.1 0.0
59.4 0.0
7
49.3
23.4
31.3
3.0
0.8
8
0.9
9
0.6
10
88.0
0.6
5.9
a%Area from HPLC (210 nm), corrected for response factor vs. 3: 1: 1.6; 4: 1.0; 5: 1.2; 6: 1.2.
Remaining mass balance is comprised of multiple lowꢀlevel impurities that were not individually
b
quantified. Evonik. cJohnson Matthey. d20 wt% catalyst used.
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 30
1
2
3
4
5
6
7
8
The effect of water. As shown in Table 1, despite achieving excellent reactivity with a
number of systems, levels of the desoxy impurities 4 and 6 were still rather high in reactions that
proceeded to completion. As the telescoped hydrogenation would be the penultimate stage of the
API synthesis, further reducing the amount of these impurities was desired. As catalyst identity
had been fully explored, we then turned to base identity as a method to modulate reactivity and
selectivity. While there are countless conjugate bases of organic acids that would provide an
appropriate and tunable range of pKa values close to that of acetic acid, use of such bases was
deemed undesirable due to the additional costs and purification steps that could result; instead,
we investigated solvent composition. We had already established the reactivity order MeOH >
EtOH > iPrOH, and the importance of reaction medium acidity provides one possible rationale
for these observations: pKa values decrease in more polar media, meaning that the overall [H+] in
MeOH is higher than in iPrOH for the same acid.9 Thus, we investigated the effect of using
water as a coꢀsolvent with IMS (Industrial Methylated Spirits: 5% MeOH in EtOH) in varying
ratios for eight of the catalysts from Table 1.9b Two representative plots of product distribution
versus volume of water added are shown in Figure 1, for the Ru and Pd catalysts from Table 1,
entries 8 and 10 respectively.
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
10

Page 11 of 30
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
Figure 1. Plot of % area (HPLC, 210 nm, corrected for response factor vs. 3) versus water
content for the hydrogenation of 1 with 5% Pd/C (E101 NE/W, top) and 5% Ru/Al2O3 (5R698,
bottom).
Clearly, increasing water content has a dramatic effect on reactivity, leading to much higher
conversions to 3 in the case of 5% Pd/C (Figure 1, top), and 5 in the case of 5% Ru/Al2O3
(Figure 1, bottom). Each of the eight catalysts behaves differently in this experiment; however,
one striking commonality is that all of the systems operate better with some amount of added
water. On average, an IMS:water ratio of about 1:1 appears optimal; this is likely due to a
compromise between increasing acid strength and maintaining the solubilities of 1 and the
various intermediates. While water clearly has a positive effect on reactivity, its influence on
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 30
1
2
3
4
5
6
7
8
9
selectivity is not as pronounced. In several cases, as the water content is increased, the total area
of all impurities observed by HPLCꢀUV/Vis decreases; however, the total amount of desoxy
impurities (4 + 6) remains effectively constant.
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
Reaction Profiling of Pd-catalyzed hydrogenation. The consistent formation of small
amounts of deoxygenated compounds 4/6 under a wide range of conditions prompted a
mechanistic investigation of the hydrogenation process through reaction progress analysis.
Timeꢀcourse plots of the concentrations of 1ꢀ6 were generated by HPLC analysis of aliquots
withdrawn from the reaction mixture while maintaining stirring under an atmosphere of H2
(mesitylene was used as an internal standard). Initially, the hydrogenation of 1 (1.0 g) with a
lower loading of 5% Pd/C E101 NE/W (20 mg, 2 wt% versus 1) was studied in order to
determine if less catalyst could be used upon scaleꢀup. Following the progress of this reaction
reveals rapid debenzylation of 1 to give ketone 5, with a zeroꢀorder dependence on [1] (kobs
=
5.51 x 10ꢀ3 M minꢀ1; Figure 2). This is likely due to both very rapid amine debenzylation, and the
low H2 pressure used for these experiments (1 atm); stirꢀrate studies to assess massꢀtransport
effects were not explored. Under these conditions, further reduction of 5 to give 3 is not
observed, even over extended reaction times. While this result is disappointing, deoxygenation
to 4 or 6 is also not observed; thus, a lowꢀloading of Pd could be used as a selective
debenzylation component in a dualꢀcatalyst process if a suitable ketone hydrogenation system
could be identified (vide infra).
ACS Paragon Plus Environment
12

Page 13 of 30
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
Figure 2. Plot of normalized concentration versus time (using mesitylene as internal standard),
for the hydrogenation of 1 at 30 °C and 2 wt% charge of 5% Pd/C, clearly showing zeroꢀorder
dependence on [1] for the debenzylation reaction.
In order to collect reaction progress data on the entire hydrogenation process, a higher loading
of 5% Pd/C (10 wt% versus 1) was used under analogous reaction conditions; the timeꢀcourse
profiles and various expansions are shown in Figure 3. Several key insights have resulted from
this single experiment. First, debenzylation is again rapid and zeroꢀorder in [1] (Figure 3, top
right), with a rate of 7.66 x 10ꢀ5 M minꢀ1; the similar rates observed for debenzylation using 2
wt% and 10 wt% catalyst loadings are consistent with massꢀtransport limited kinetics for the
debenzylation. Second, ketone hydrogenation does not occur until debenzylation is complete.
This completely stepꢀwise reactivity can be explained by the much slower initial rate of ketone
hydrogenation (6.08 x 10ꢀ4 M minꢀ1), which is an order of magnitude less than debenzylation.
Alternatively, the starting material 1 may be more strongly bound to the catalyst surface than
intermediate 5, which would result in negligible rates of ketone hydrogenation until all of bound
1 is consumed. Third, the rate dependence on [3] for ketone hydrogenation is slightly less than
first order, with the reaction proceeding too quickly at high conversion for simple exponential
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 30
1
2
3
4
5
6
7
8
9
decay (Figure 3, bottom left). Fitting the first 1.5 halfꢀlives of [3] to a firstꢀorder mechanism
gives good correlation, while the overall reaction order appears to be complex, and empirically is
~0.8 by doubleꢀln plot analysis.6 This nonꢀinteger overall order could be due to changing acidity
over the course of the reaction (since the system is not strictly buffered, and the pKa values of 1ꢀ
H+, 3ꢀH+, and 5ꢀH+ are not exactly the same), and/or substrateꢀinhibition of the ketone
hydrogenation process. This latter possibility could result from catalyst surface saturation by
adsorbed 5, with dissociation of 5 required to reveal Pd sites for H2 activation.
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
Figure 3. Top left: full plot of normalized concentration versus time for the hydrogenation of 1
at 30 °C and 10 wt% charge of 5% Pd/C. Top right: expansion of debenzylation of 1 to give 5.
Bottom left: expansion of hydrogenation of 5 to give 3. Bottom right: expansion of
ACS Paragon Plus Environment
14

Page 15 of 30
Organic Process Research & Development
1
2
3
4
deoxygenation products, showing that all deoxygenation occurs from 5, with 3 stable under the
5
6
reaction conditions.
7
8
9
Finally, and perhaps most importantly, the formation of 4 and 6 only occurs from
deoxygenation of the ketone intermediate (Figure 3, bottom right). Compound 4 is formed by
hydrogenation of the ketone in 6, and not by direct deoxygenation of 3; in fact, compound 3 is
very stable under the reaction conditions, with no increase in deoxygenation observed over a
reaction time of more than 30 hours. We note that this result is not consistent with previous
observations that electronꢀdonating groups on the aryl ring accelerate deoxygenation,5a whereas
in our case, converting an electronꢀwithdrawing ketone into an electronꢀdonating benzyl alcohol
shuts down deoxygenation activity. This is a significant result, as the risk of overꢀreduction
under these conditions once the reaction is complete is negligible. In addition, the rate of the
deoxygenation process clearly depends on [5], so it should be possible to minimize
deoxygenation through rapid ketone hydrogenation. Scheme 3 summarizes the possible and
observed pathways deduced.
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
15

Organic Process Research & Development
Page 16 of 30
1
2
3
4
Scheme 3. Possible and observed reaction pathways in the Pdꢀcatalyzed hydrogenation of 1, with
5
6
nonꢀoperative pathways in grey.
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
Because of the possibility of substrateꢀinhibition in the conversion of 5 to 3, and the need to
more rapidly consume 5 to reduce deoxygenation, the hydrogenation of 1 was carried out at 55
°C and monitored over time. If desorption of 5 from the catalyst surface is necessary to effect
reduction, increasing reaction temperature should further increase the reaction rate by shifting
the adsorption/desorption equilibrium toward desorption, and thus decreasing inhibition.
Increasing temperature will also shift the acid/base equilibria to increase acidity, which should
also result in a faster reaction. Comparing the initial rates of ketone hydrogenation at 30 °C and
55 °C reveals a greater than 8ꢀfold increase, and a switch in reaction order from nearly firstꢀorder
in [5] to effectively zeroꢀorder in [5] (Figure 4). This means that the elementary steps of ketone
hydrogenation have increased in rate such that the reaction is now hydrogenꢀlimited, likely
ACS Paragon Plus Environment
16

Page 17 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
resulting from massꢀtransfer limitations. Rather than reaching completion in ~10 hours at 30 °C,
conversion to 3 is now complete in only one hour. Unfortunately, the degree of deoxygenation is
slightly worse under these conditions, with 4 at 9% by HPLC, even though the initial rate of
deoxygenation is only 1.4x faster than at 30 °C. This is because deoxygenation has also
switched to zeroꢀorder in [5], and therefore the ratio of 3 to 4+6 simply depends on the relative
initial rates of ketone reduction and deoxygenation, rather than the concentration dependences
observed at lower temperature. This is a key observation for further reaction development, in
that an increase in reaction temperature must be accompanied by an increase in H2 pressure to
avoid global massꢀtransfer limited kinetics. Finally, despite the more forcing conditions, 3 is still
stable toward deoxygenation, with all deoxygenation occuring from intermediate 5.
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
Figure 4. Left: full plot of normalized concentration versus time for the hydrogenation of 1 at 55
°C, showing zeroꢀorder dependence on [5] for ketone reduction. Right: expansion of
deoxygenation products, showing zeroꢀorder dependence on [5], and that all deoxygenation still
occurs from 5.
Development of a dual catalyst system. One method to further decrease the rate of
deoxygenation is to employ a dualꢀcatalyst process. This system would ideally exhibit
orthogonal reactivity: one catalyst should have low enough activity to only achieve
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 30
1
2
3
4
5
6
7
8
9
debenzylation, and the second catalyst should be able to reduce the ketone while minimizing
deoxygenation pathways. Based on the data collected in the comprehensive catalyst library
screen, several promising candidates were investigated. From the reaction profiling experiments,
it was apparent that reducing the loading of 5% Pd/C to 2 wt% would achieve selective
debenzylation; lowꢀloading 1% Pd/C catalysts also appeared promising, as did the 5% Ru/Al2O3
system.
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
In order to identify suitable ketone hydrogenation catalysts, a series of highꢀthroughput screens
were designed to evaluate 48 Ptꢀbased catalysts alongside several Pdꢀ and Ruꢀbased
debenzylation systems. Examining various reaction conditions, including solvent and base
identity, revealed that these dual catalyst systems operate best under the same general conditions
as the Pdꢀonly process, namely 1:1 IMS/water solvent, and KOAc as the base. This is a key
outcome, as development of many tandem catalyst systems are plagued by incompatibilities and
reaction rates that are not wellꢀmatched.10 Another key observation is that the Ptꢀcatalysts
investigated appear to perform ketone hydrogenation much more efficiently after N-
debenzylation. Control experiments with only Ptꢀcatalysts under analogous conditions gave
modest conversion of 1 to the alcohol 2 (~15ꢀ20%), with minimal formation of 3 by
debenzylation; in contrast, many dual catalyst systems achieve complete conversion to 3 in the
same length of time.6
Data from several highꢀthroughput screens were collated, and ten standꢀout Ptꢀbased catalysts
were chosen for further investigation. Table 2 details the results of pairing these ten Pt systems
with a 1% Pd/C catalyst under the previously optimized conditions. Clearly, these dual catalyst
systems exhibit superior selectivity to the Pdꢀonly reactions: typical desoxy content with the Pdꢀ
only system was 4ꢀ7%, whereas with a dual Pd/Pt system, as low as 1.2% deoxygenation can be
ACS Paragon Plus Environment
18

Page 19 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
achieved. In addition to dual Pd/Pt systems, we also investigated use of the aforementioned 5%
Ru/Al2O3 catalyst as the debenzylation component. To the best of our knowledge, Ruꢀbased
catalysts are not reported to be applicable to selective N–Bn hydrogenolysis; however, this
particular catalyst works well, even at modest H2 pressure. While control experiments revealed
that debenzylation with 5% Ru/Al2O3 does not occur at 1 atmosphere H2, increasing to only 1.5
barg enables catalysis. Table 3 outlines the results of tandem Ru/Pt catalysis for the
hydrogenation of 1. Once again, the levels of deoxygenation are lower than for Pdꢀonly
reactions, though not as low as for the Pd/Pt systems (Table 2); however, the ruthenium
screening results were obtained at lower H2 pressure, which may lead to lower selectivity.
Table 2. HTS results for dual Pd/Pt catalyst systems.
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
Mass
1
3
5
4 + 6
Entry Pt Catalyst
Balance
(%)a (%)a (%)a (%)a
(%)a
1
2
3
4
5
6
7
8
5% Pt/C; F105 N/Wb
0.0
91.2 1.5
89.0 2.7
94.1 1.7
90.3 2.1
91.2 1.2
91.1 1.6
92.9 0.9
93.2 1.0
2.8
2.4
2.0
2.6
1.2
2.5
1.7
1.9
95.5
94.1
97.8
95.0
93.8
95.5
95.5
96.1
5% Pt/C; F1015 RE/Wb 0.0
5% Pt/Al2O3; 5R94c
3% Pt/C; B103032ꢀ3d
1% Pt/C; 1.5R199c
5% Pt/C; 78ꢀ1611e
1% Pt/C; 38313f
0.0
0.0
0.2
0.3
0.0
0.0
3% Pt/C; 38317f
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 30
1
2
1% Pt/C; F1525 KT/Wb 0.0
5% Pt/C; F101KYA/Wb 0.0
91.0 2.3
93.8 0.9
1.3
2.1
94.6
96.8
3
4
5
6
9
10
7
8
9
a%Area from HPLC (210 nm), corrected for response factor vs. 3: 1: 1.6; 4: 1.0; 5: 1.2; 6: 1.2.
Remaining mass balance is comprised of multiple lowꢀlevel impurities that were not individually
b
quantified. Evonik. cJohnson Matthey UK. dJohnson Matthey US. eStrem. fAlfa Aesar.
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 3. HTS results for dual Ru/Pt catalyst systems.
Mass
1
3
5
4 + 6
Entry Pt Catalyst
Balance
(%)a (%)a (%)a (%)a
(%)a
1
2
3
4
5
6
7
8
9
10
5% Pt/C; F105 N/Wb
0.0
89.0 0.8
89.2 0.5
93.6 0.5
89.3 1.2
88.9 2.0
82.8 4.9
87.3 3.2
86.6 3.1
3.6
4.0
3.5
3.0
2.7
3.4
1.9
2.4
93.4
93.7
97.6
94.0
94.7
94.6
94.6
94.5
95.0
95.5
5% Pt/C; F1015 RE/Wb 0.0
5% Pt/Al2O3; 5R94c
3% Pt/C; B103032ꢀ3d
1% Pt/C; 1.5R199c
5% Pt/C; 78ꢀ1611e
1% Pt/C; 38313f
0.0
0.5
1.1
3.5
2.2
2.4
3% Pt/C; 38317f
1% Pt/C; F1525 KT/Wb 1.4
5% Pt/C; F101KYA/Wb 1.6
71.0 20.3 2.3
89.1 2.2 2.6
a%Area from HPLC (210 nm), corrected for response factor vs. 3: 1: 1.6; 4: 1.0; 5: 1.2; 6: 1.2.
Remaining mass balance is comprised of multiple lowꢀlevel impurities that were not individually
quantified. bEvonik. cJohnson Matthey UK. dJohnson Matthey US. eStrem. fAlfa Aesar.
ACS Paragon Plus Environment
20

Page 21 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Scale-up and isolation. Once this detailed picture of the hydrogenation pathways and catalyst
suitability was obtained through highꢀthroughput experimentation and smallꢀscale reaction
profiling experiments, our efforts shifted to scale this chemistry and develop a process to isolate
and purify 3. During the course of process optimization, it became apparent that removal of the
inorganic salt impurities would be a considerable challenge. KCl has nonꢀnegligible solubility in
ethanol/water mixtures;11 thus, use of even modest IMS/water ratios during the reaction severely
complicates isolation. Furthermore, upon neutralization to generate the free base of 3, an
equivalent of KOAc is generated, which has significant solubility even in waterꢀfree ethanol.12
Ordinarily, salts could be removed by water washes; however, 3 is very waterꢀsoluble, even in its
neutral form.
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
Thus, alternate reaction conditions are required for a feasible process. Using the insights
gained as a result of highꢀthroughput screening and mechanistic studies, we evaluated
substoichiometric inorganic carbonate bases to achieve a similar reaction acidity to that using
KOAc, and used IMS as the sole reaction solvent. Because the absence of water was anticipated
to dramatically reduce the reaction rate, the temperature was raised to 45 °C, and catalyst
loadings increased to compensate. Based on the reaction progress data from Figure 4, even a
small temperature increase should lead to a substantial rate increase. Table 4 summarizes the
key results from these experiments, with 0.25 equivalents of K2CO3 as the optimal base. While
the reactions do not reach complete conversion in 10 hours, the selectivity is increased, with
~1% deoxygenation observed. It should be noted that while these reactions do not have water as
a coꢀsolvent, there is undoubtedly some water present as a result of acid neutralization, and as
part of the catalysts (supplied as 50% w/w water).
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 30
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
Table 4. Substoichiometric K2CO3 as a base.
Mass
1
3
5
4 + 6
Entry Catalysts
1% Pd/C; 38293b
Balance
(%)a (%)a (%)a (%)a
(%)a
1
2
3
4
0.2
1.1
6.8
6.3
76.0 19.1 1.4
96.7
97.1
96.1
95.5
5% Pt/Al2O3; 5R94c
1% Pd/C; 38293b
90.1 4.9
1.0
3% Pt/C; 38317b
5% Ru/Al2O3; 5R698c
5% Pt/Al2O3; 5R94c
5% Ru/Al2O3; 5R698c
3% Pt/C; 38317b
57.9 30.8 0.6
84.8 3.9
0.5
a%Area from HPLC (210 nm), corrected for response factor vs. 3: 1: 1.6; 4: 1.0; 5: 1.2; 6: 1.2.
Remaining mass balance is comprised of multiple lowꢀlevel impurities that were not individually
quantified. bAlfa Aesar. cJohnson Matthey UK.
With confidence that a “waterꢀfree” reaction would still progress with a reasonable rate, we
scaledꢀup the conditions from Table 4, entry 2 for a 20 g demonstration run (eq. 1). The
hydrogenation was performed for 23 hours in order to ensure complete conversion of 1. Upon
ACS Paragon Plus Environment
22

Page 23 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reaction completion, the catalyst mixture and inorganic salts were removed by a simple filtration.
Recrystallization of the resulting crude product from an IMS/EtOAc system afforded 3 in 49%
isolated yield, with <0.1% of deoxygenated byproducts observed. While this isolated yield is
modest, it does compare well to the current 55% yield for the twoꢀstep sequence. Furthermore,
this new process is considerably more streamlined than the twoꢀstage reduction, and further
optimization of the isolation and recrystallization could certainly improve the isolated yield.
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
Conclusions.
Using a combined highꢀthroughput screening and mechanistic elucidation approach, we have
demonstrated that the tandem reduction of compound 1 to generate Salbutamol (3) can be
effected by a wide array of catalyst systems under mild conditions, thus shortening the synthetic
route by one step and greatly reducing the number of unit operations required. While a number
of Pd/C catalysts are able to perform both reductions in a tandem manner by using a weakly
acidic reaction medium, the consistent formation of deoxygenated impurities through benzyl
alcohol hydrogenolysis is problematic. Interestingly, these impurities only result from
deoxygenation of the ketone intermediate 5; no deoxygenation of Salbutamol (3) itself was
observed under the standard hydrogenation conditions investigated, even after prolonged
reaction times. As a result of these observations, we developed several dualꢀcatalyst systems that
combine highly selective debenzylation catalysts based on Pd or Ru with Pt catalysts that are
effective for ketone hydrogenation. In this manner, the amount of deoxygenated impurities can
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 30
1
2
3
4
5
6
7
8
9
be reduced to ~1% in the crude material. It should be noted that to achieve an effective isolation
of Salbutamol (3) from the tandem hydrogenation, reaction conditions needed to be changed to
eliminate water and switch from KOAc to K2CO3; while our initial highꢀthroughput screening
data helped to identify that these alternative conditions would be suitable, our more extensive
screening and reaction profiling experiments would have benefitted by incorporating this change
earlier in the project. A 20 g scale demonstration of this new tandem catalytic process was
carried out, giving 49% isolated yield of Salbutamol (3) after recrystallization. Further
optimization of the workꢀup, isolation, and purification of 3, particularly in the context of
residual inorganics, should enable a streamlined synthesis of this critical respiratory medicine.
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.
Materials. All common reagents and solvents were purchased from commercial sources and
used as received. Hydrogenation catalysts were obtained from the manufacturers or other
suppliers as noted in the main text tables. Compounds 1 and 5 (as an analytical marker) were
obtained from GlaxoSmithKline’s Global Manufacturing and Supply group; compounds 2ꢀ4
were prepared and characterized as described in the Supporting Information.
NMR Spectroscopy. All NMR spectra were acquired at ambient temperature on a Bruker 400
MHz or 500 MHz spectrometer. Solvents and frequencies for specific data acquisitions are
noted for each case in the Supporting Information.
HPLC. HPLC analysis was performed on Agilent 1100 or 1200 series instruments with diode
array detectors, though analysis was typically done with traces from a single wavelength. Two
HPLC methods were utilized during the course of this work.
ACS Paragon Plus Environment
24

Page 25 of 30
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Method A: Column: Zorbax SBꢀC18, 1.8 μm, 3 x 50 mm; column temp: 40 °C; flow rate: 1.0
mL/min; solvent gradient: ACN (0.05% TFA v/v) / H2O (0.05% TFA v/v), from 100/0 to 5/95
over 8 min; detection wavelength: 220 nm. Compounds 3 and 5 coꢀelute using this method, and
so a second HPLC method was developed (Method B).
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
Method B: Column: Zorbax SBꢀCN, 1.8 μm, 3 x 150mm; column temp: 25 °C; flow rate: 1.0
mL/min; solvent gradient: ACN (0.05% TFA v/v) / H2O (0.05% TFA v/v), from 100/0 to 5/95
over 15 min; detection wavelength: 210 nm.
SAFETY NOTE: Pressurized hydrogenation reactions should always be carried out in wellꢀ
maintained, leakꢀchecked pressure equipment located in a wellꢀventilated environment,
preferrably a dedicated pressure laboratory. Hydrogenation catalysts are pyrophoric materials,
particularly when dry, and especially after being exposed to hydrogen. Care should be taken in
handling the catalyst material after reaction completion. Never letting the catalyst cake dry
during filtration, and prompt quenching of the spent catalyst in a dilute aqueous oxidant such as
sodium bisulfite are critical to ensuring safe operation.
Procedure for high-throughput screening experiments. Smallꢀscale HTS reactions were
carried out with ~14 mg of 2 (~75% wt/wt, balance IMS, ~0.03 mmol), using ~1 mg of a
heterogeneous catalyst, and 0.15 mL solvent. Solid reagents and catalysts were automatically
dosed using a FlexiWeigh (Mettler Toledo) instrument into a 48ꢀwell array of 1.5 mL crimpꢀcap
HPLC vials containing micro stirbars. All liquid reagents and solvents were dispensed via
micropipette to the appropriate vials according to the experimental design. Once all materials
were charged, each vial was sealed with an aluminum crimpꢀcap containing a
PTFE/silicone/PTFE septum. The septa were pierced with a wideꢀbore needle to allow gas
ingress/egress. The vials (arrayed in a metal plate) were heated to the desired reaction
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 30
1
2
3
4
5
6
7
8
9
temperature with magnetic stirring under the desired pressure of hydrogen inside a Cat96
pressure reactor (HEL). Prior to commencing heating/stirring, the following purge cycles were
conducted: 3 x 5 barg nitrogen, 3 x 5 barg hydrogen. At the end of the desired reaction time,
stirring was stopped, the reactor was cooled to room temperature and the atmosphere replaced
with nitrogen following 3 x 5 barg purge cycles. Once the system returned to ambient pressure,
the plate was removed, and the caps removed. Each vial was diluted with 1 mL methanol. The
plate was placed on a stirplate to ensure complete dissolution of any organic products. The
resulting suspensions were centrifuged prior to analysis by HPLC. See supporting information
for full tables of reaction conditions and results for each highꢀthroughput screen.
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
Procedure for reaction profiling experiments. The following procedure corresponds to the
reaction progress plots from Figure 3, and is representative. A 100 mL, 2ꢀnecked RB flask
containing a stirbar was charged with 5% Pd/C (E101 NE/W, 50% w/w water, 100 mg) and
deionized water (2 mL). The flask was immersed in a silicone oil bath, and fitted with a gas
adapter connected to a dual manifold Schlenk line, and the remaining jointed sealed with a
septum. Stirring was commenced, and the headspace was briefly evacuated and refilled with
nitrogen three times. The headspace was then refilled with hydrogen three times. After the third
refill, the catalyst slurry was held under a slight flow of hydrogen (system vented through an oil
bubbler) with stirring for at least one hour. Then, the oil bath was heated to 30 °C, and a solution
of sodium acetate (0.185 g, 2.25 mmol) in deionized water (2 mL) was added to the reaction
mixture via syringe, followed by a solution of 1 (1.00 g, 74% w/w IMS, 0.74 g corrected, 2.04
mmol) and mesitylene (internal standard, 0.253 g, 2.10 mmol) in IMS (6 mL). Small aliquots
(~50 uL) were withdrawn via syringe and quenched by dilution into MeOH (~1 mL) as quickly
ACS Paragon Plus Environment
26

Page 27 of 30
Organic Process Research & Development
1
2
3
4
5
6
as possible. These samples were then filtered through a 0.45 um syringe filter before analysis by
HPLC.
7
8
9
After the 10 hour aliquot, the hydrogen flow was stopped, and the flask headspace filled with
nitrogen. The flask was removed from the oil bath and held at room temperature under nitrogen
without stirring overnight. The next morning, a reaction aliquot was taken to ascertain any
changes to composition overnight, which indicated no further reactions had taken place. The
headspace was swapped back to hydrogen, and the flask contents heated to 30 C to continue the
reaction profiling. At the end of the day, the septum was replaced with a ground glass joint, and
the reaction mixture stirred under static hydrogen overnight. The final 30 h timepoint was
obtained with an aliquot withdrawn the next morning.
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
20 g scale hydrogenation to prepare 3. This reaction was performed in a 500ml jacketed
glass hydrogenator with a Buchi H2 gas flow controller, and equipped with an agitator. Prior to
commencing, the reactor was pressure tested with 3 barg N2 to confirm a leakage rate of <0.02
barg over 1 hour. Then, compound 1 (25 g, 20% w/w IMS: 20 g corrected for residual solvent,
55.0 mmol) was charged to the hydrogenator. IMS (230 mL) was then added, and stirring
commenced in order to suspend 1. The jacket temperature was set so as to maintain an internal
temperature of 45 °C. Potassium carbonate (2.34 g, 0.25 eq) was then charged as a solid, and the
resulting slurry was stirred for 10 minutes. The catalysts were then charged as solids: 3% Pt/C
(4.0 g, 20 wt% vs. 1) and 1% Pd/C (4.0 g, 20 wt% vs. 1), using IMS (20mL) as a wash. The
hydrogenation reactor was then sealed, and inerted by purging with 3 cycles of 3 barg N2.
Stirring was stopped, and the vessel was then purged with 3 cycles of 3 barg H2. The reactor was
then charged with 1.5 barg H2, and stirring commenced at 1000 rpm. The hydrogenation was run
at 45 °C for 23 hours. Upon completion, the vessel was purged with N2, and the reaction mixture
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 30
1
2
3
4
5
6
7
8
9
was filtered through a GF/C filter to remove the catalyst. The reactor was charged with IMS
(200 mL), which was heated to 45 °C and applied to the catalyst cake as a cake wash. The
combined IMS filtrates were evaporated to dryness under reduced pressure in a 250mL jacketed
lab reactor. The resulting solid was reꢀdissolved in IMS (31 mL) at 30 °C. Ethyl acetate (142
mL) was added to the batch over approximately 1 hour at 30 °C. The mixture was left to age at
30 °C for 2 hours before cooling the resulting slurry to 0 °C. The slurry was stirred overnight at
0ꢀ5 °C before isolating the product 3 by filtration. The product cake was washed with cold ethyl
acetate (25mL) and dried under vacuum to yield 6.45 g (49%) of 3 with <0.1% of deoxygenated
impurities 4 or 6, and a solution assay by HPLC of 92% w/w (due to the presence of residual
inorganic salts).
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
ASSOCIATED CONTENT
Supporting Information. The following files are available free of charge.
Synthesis/characterization of intermediates 2ꢀ4 as analytical markers, charts of HTS data, and
supplemental concentration versus time plots (PDF).
AUTHOR INFORMATION
Corresponding Author
* david.c.leitch@gsk.com.
Present Addresses
Thomas C. Lovelace and Jeremiah D. Powers: Avista Pharma Solutions, 3501 Tricenter Blvd.,
Durham, NC 27713.
ACS Paragon Plus Environment
28

Page 29 of 30
Organic Process Research & Development
1
2
3
4
Author Contributions
5
6
7
8
9
The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
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
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
The authors thank all of the former members of the Catalysis Center of Excellence at Research
Triangle Park for insightful discussions and assistance throughout the course of this work.
REFERENCES
(1) PubChem Compound Summary for Salbutamol (National Center for Biotechnology
Information). https://pubchem.ncbi.nlm.nih.gov/compound/Salbutamol (accessed July 20,
2017).
(2) Hartley, D.; Jack, D.; Lunts, L. H. C.; Ritchie, A. C. Nature 1968, 219, 861ꢀ862.
(3) (a) Lunts, L. H. C.; Toon, P.; Collin, D. T. Phenylaminoethanol derivatives.
GB1200886A, September 23, 1966. (b) Collin, D. T.; Hartley, D.; Jack, D.; Lunts, L. H.
C.; Press, J. C.; Ritchie, A. C.; Toon, P. J. Med. Chem. 1970, 13, 674ꢀ680. (c) Babad, E.;
Carruthers, N. I.; Jaret, R. S.; Steinman, M. Synthesis 1988, 966ꢀ968. (d) Hong, Y.; Gao,
Y.; Nie, X.; Zepp, C. M. Tet. Lett. 1994, 35, 5551ꢀ5554. (e) Regla, I.; Reyes, A.; Korber,
C.; Demare, P.; Estrada, O.; Juaristi, E. Synth. Commun. 1997, 27, 817ꢀ823. (f)
Effenberger, F.; Jager, J. J. Org. Chem. 1997, 62, 3867ꢀ3873. (g) AyalaꢀMata, F.; Barreraꢀ
Mendoza, C.; JimenezꢀVazquez, H. A.; VargasꢀDiaz, E.; Zepeda, L. G. Molecules 2012,
ACS Paragon Plus Environment
29