Process Development of a Suzuki Reaction Used in the Manufacture of Lanabecestat

Article abstract of DOI:10.1021/acs.oprd.8b00312

We developed a scalable Suzuki process for the synthesis of lanabecestat (+)-camsylate, an active pharmaceutical ingredient that was recently investigated in a Phase III clinical program for the treatment of early Alzheimer's disease. The evolution of this process culminated with the use of a stable and crystalline diethanolamine boronic ester that rapidly hydrolyses under the reaction conditions. Herein, we report that the liberated diethanolamine plays an important role in the catalytic process, with supporting evidence for an equilibrium between an unbound and bound palladium complex. Additionally, the diethanolamine acts as an internal scavenger during the crystallization of lanabecestat by increasing the solubility of the palladium species, obviating the need for a discrete scavenging step.

Full text of DOI:10.1021/acs.oprd.8b00312

Subscriber access provided by University of Sunderland
Full Paper
Process Development of a Suzuki Reaction
used in the Manufacture of Lanabecestat
Ian W. Ashworth, Andrew D. Campbell, Janette H. Cherryman, Jemma Clark, Alex
Crampton, Edward Eden-Rump, Matthew Evans, Martin F. Jones, Sahar Abbas, Rebecca
E. Meadows, Kathryn Skilling, David Whittaker, Robert Woodward, and Phillip Inglesby
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00312 • Publication Date (Web): 07 Nov 2018
Downloaded from http://pubs.acs.org on November 7, 2018
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 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 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.
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 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Process Development of a Suzuki Reaction used in
the Manufacture of Lanabecestat
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
Ian W. Ashworth, Andrew D. Campbell, Janette H. Cherryman, Jemma Clark, Alex Crampton,
Edward G. B. Eden-Rump, Matthew Evans, Martin F. Jones, Sahar McKeever-Abbas, Rebecca E.
Meadows, Kathryn Skilling, David T. E. Whittaker†, Robert L. Woodward, Phillip A. Inglesby*
Pharmaceutical Development and Technology, AstraZeneca, Silk Road Business Park,
Macclesfield, United Kingdom.
†Pharmaceutical Sciences, AstraZeneca, Silk Road Business Park, Macclesfield, United Kingdom.
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 28
1
2
3
4
For Table of Contents Only
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 28
Organic Process Research & Development
1
2
3
4
Keywords: Process Development – Suzuki – Kinetics – Scavenger – Manufacture – Lanabecestat.
5
6
7
8
9
Abstract
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 have developed a scalable Suzuki process for the synthesis of lanabecestat (+)-camsylate, an
active pharmaceutical ingredient that was recently investigated in a Phase III clinical program for
the treatment of early Alzheimer’s Disease. The evolution of this process has culminated with the
use of a stable and crystalline diethanolamine boronic ester which rapidly hydrolyses under the
reaction conditions. Herein, we report that the liberated diethanolamine plays an important role in
the catalytic process, with supporting evidence for an equilibrium between an unbound and bound
palladium complex. Additionally, the diethanolamine acts as an internal scavenger during the
crystallization of lanabecestat by increasing the solubility of the palladium species, obviating the
need for a discrete scavenging step.
Introduction
Lanabecestat, co-developed by AstraZeneca (AZD3293) and Eli Lilly and Company
(LY3314814), is a -secretase inhibitor and was recently investigated in a Phase III clinical
program for the treatment of early Alzheimer’s disease (AD). Amyloid precursor proteins (APPs)
are found within neurons and cleavage of these large membrane proteins result in elevated amyloid
levels within the brain. Amyloid accumulation is thought to play a key role in the progression of
AD and can result from changes in production, processing, and/or clearance of brain amyloid-β
(Aβ) levels. β-Site amyloid precursor protein cleaving enzyme 1 (BACE-1) is the first step in the
processing of APP to Aβ peptides, and its inhibition is an attractive target to stop the production
of Aβ.1
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 28
1
2
3
4
5
6
7
8
9
We recently disclosed the Phase III route to lanabecestat, which involved a late-stage Suzuki
coupling to install the pyridine side-chain (Scheme 1).2 The Suzuki coupling is a common reaction
used in the manufacture of pharmaceuticals and is the most common biaryl bond forming
reaction.3-5 The final stage, to generate the active pharmaceutical ingredient (API) (lanabecestat
(+)-camsylate) is a salt formation with (+)-camphor sulfonic acid.
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
Several factors need to be understood when developing a chemical process with precious metals
such as palladium. For example, the cost of the catalyst and subsequent loading can have a large
impact on the overall cost-of-goods (CoGs) of the API and typically, tight controls of palladium
levels in the final drug substance are necessary to ensure patient safety. There are several ways to
remove such metals from a chemical process, for instance, via solid-supported scavengers,
however, these steps can be expensive. Whilst the introduction of a palladium-catalyzed reaction
can pose challenges to a control strategy, the late introduction of an expensive fragment, such as
pyridine 2 (Scheme 1), is advantageous to cost.
Scheme 1. Suzuki coupling for the final step to Lanabecestat and associated known organic
impurities.
ACS Paragon Plus Environment
4

Page 5 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The Suzuki process that was originally developed for kilo-scale manufacture in a large-scale
laboratory (Process A) was not robust enough to deliver routinely on quality or quantity at full
scale production. The reaction was performed in n-BuOH and the work-up involved numerous and
lengthy phase separations with water to remove the potassium camsylate by-product. Furthermore,
it involved a solid-supported scavenger step, a challenging solvent swap from n-BuOH to n-
BuOAc, and an uncontrolled crystallization (precipitation) of lanabecestat, all of which were
undesirable for scale-up.
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
In 2015, we developed a new Suzuki process (Process B) for Phase III clinical supply. High-
throughput experimentation (HTE) was used to rapidly optimize the aryl bromide (1) and boronic
acid (2a) charges,6 the catalytic system (metal and ligand), solvent and base. The optimized process
(Process B) used a 0.05 mol equivalent excess of boronic acid (relative to aryl bromide 1),
Na2PdCl4, 3-(di-tert-butylphosphonium)propane sulfonate (DTBPPS) and potassium phosphate,
in aqueous ethanol at 70 °C. After removal of palladium on a solid-supported scavenger
(PhosphonicS SPM32), lanabecestat was isolated via the addition of water as an antisolvent.
Between 2015 and 2017, this process was successfully employed at AstraZeneca’s Macclesfield
pilot plant and delivered over 500 kg of lanabecestat in 15 batches.
A review of the manufacture highlighted several challenges associated with the testing, handling
and charging of the boronic acid (2a). Development work had shown that undercharging the
boronic acid led to an incomplete reaction and an overcharge gave rise to an over-coupled impurity
(3);6 both instances afford lanabecestat that does not meet the required specification. Charging was
considered a potentially critical process parameter (CPP) since it had a narrow range of ±0.02 mol
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 28
1
2
3
4
5
6
7
8
9
equivalents (relative to 1), and operating outside of this range was detrimental to the quality of the
API. In this instance, we considered that the equipment had a control range of ±0.04 mol
equivalents and since the boronic acid charge range did not extend beyond the plant control range,
the process posed a significant quality risk to future manufactures. Additional learning from the
manufacture and subsequent development work highlighted that analysis of the boronic acid was
challenging due to difficulties in determining, unambiguously and with accuracy, the relative
proportions of acid vs. boroxine (2e) present.6 Furthermore, the preceding manufacturing process
could not be confirmed as delivering solely either the boronic acid or boroxine species on a
consistent basis. In addition to this, the material was a fine, amorphous and difficult-to-handle
powder.
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
To circumvent the analytical and manufacturing challenges with the boronic acid, we
investigated boronic esters in the Suzuki reaction, including diethanolamine (2b), neopentylglycol
(2c) and pinacol (2d) esters.6-8 The diethanolamine (DEA) boronic ester (2b) was selected as the
coupling partner since; 1) the ester hydrolysis was assumed to be rapid (vide infra),9 2) the boronic
ester was a monomeric and crystalline species that could be assayed, and charged, with greater
accuracy than the boronic acid, 3) the preceding borylation process could easily be amended to
manufacture the boronic ester,10-12 and 4) the Suzuki reaction quality profile was comparable to
that delivered by the boronic acid.
Diethanolamine boronic esters, often referred to as DABO boronates or dioxazaborocanes, are
a class of N-coordinated cyclic boronic esters. They display tetrahedral geometry about boron and
are advantageously air-stable, showing superior stability to the parent boronic acid, and can be
used directly in subsequent Suzuki reactions. They are rarely used in synthetic organic chemistry,
particularly in the context of process chemistry and manufacturing, and typically have significantly
ACS Paragon Plus Environment
6

Page 7 of 28
Organic Process Research & Development
1
2
3
4
5
6
different physical properties from the parent boronic acid, making them easy to prepare without
necessitating chromatography.10-12
7
8
9
With the newly selected diethanolamine boronic ester, a second round of high-throughput
experimentation was initiated, identifying Pd(AmPhos)2Cl2 and Pd(dtbpf)Cl2 as being superior to
the previous catalytic system since both catalysts allowed a lower loading to be used as well as
showing improved impurity profiles. Additionally, the pre-formed complexes were considered
advantageous compared to the previous in-situ formed complex since there was no risk of over- or
undercharging the ligand relative to the metal; this can be particularly challenging when charging
relatively small quantities at production scale. The Pd(AmPhos)2Cl2 catalyst was chosen for further
development based on its cost, reactivity, availability and safety. Following the high-throughput
experiments, development work quickly showed that the catalyst loading could be reduced from
0.4 mol% with Na2PdCl4 (Process B) to 0.15 mol% with Pd(AmPhos)2Cl2 (Process C) and the
charge range for the boronic ester could be widened to ±0.05 mol equivalents with no detrimental
impact to quality. The increase in charge range beyond the plant control range meant that the
boronic ester charge was no longer considered potentially critical and it reduced the risk of failing
specification in future manufactures with respect to residual aryl bromide (1) and the over-coupled
impurity (3).6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A drawback to the switch from the boronic acid (2a) to the boronic ester (2b) was that the rate
of reaction was reduced; this was observed with all catalysts that were tested. An important factor
in obtaining high yield with acceptable quality in this Suzuki coupling was achieving a fast reaction
since a competitive non-palladium-catalyzed hydrolysis of the dihydroimidazole ring leads to an
amide impurity (4).6 The amide is an impurity controlled by the specification of drug substance
(lanabecestat (+)-camsylate) and its formation in the Suzuki reaction was time and temperature
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 28
1
2
3
4
5
6
7
8
9
dependent. Ultimately however, the extent of hydrolysis was reduced to an acceptable level by
increasing the reaction temperature to 80 °C (vs. 70 °C) and reducing the reaction time.
Nevertheless, it was important for robustness reasons that we understood the cause for differences
in rates between the boronic acid and ester, and we proposed that this could be due to one or more
of the following factors; 1) the boronic ester may be the reactive species and inherently less
reactive, 2) the boronic acid may be the reactive species and the boronic ester hydrolysis is slow
(Scheme 2), and 3) the expelled diethanolamine may be compromising the catalyst. A series of
studies was conducted to explore these potential effects.
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. Hydrolysis of the diethanolamine boronic ester 2b in water.
Hydrolysis of the Boronic Ester (2b) in Aqueous Ethanol. The amount of residual boronic
ester (vs. boronic acid) is important for the lanabecestat (+)-camsylate control strategy. The
boronic acid is Ames positive and is an impurity controlled by the specification of drug substance
(lanabecestat (+)-camsylate). The acceptable level of the boronic acid is set at <128 ppm and this
upper limit is based primarily on its mutagenicity and threshold of toxicological concern (TTC). It
was therefore critical to understand how the switch from the boronic acid (2a) to the boronic ester
(2b) impacted the control strategy. Specifically, it was important to determine if the boronic ester
was retained or hydrolyzed to the boronic acid, and consequently whether additional controls and
tests were required in the API.
ACS Paragon Plus Environment
8

Page 9 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
300 s mixing
240 s mixing
180 s mixing
120 s mixing
90 s mixing
60 s mixing
45 s mixing
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
30 s mixing
15 s mixing
10 s mixing
5 s mixing
Resolved signal for boronic ester (2b) at ~7.78 ppm
Resolved signal for boronic acid (2a) at ~7.66 ppm
1
Figure 1: Stacked H-NMR spectra between 5-300 seconds reaction time with highlighted
regions of interest for 2a and 2b.13 Reactions were run at 27 °C in a tube observing the organic
fraction.
The diethanolamine boronic ester (2b) was dissolved in ethanol (0.56 M 2b) and mixed with
aqueous K3PO4 (1.2 M), which were the standard reaction conditions, and the concentration of
each species (acid vs. ester) was measured vs. time.14 It was possible to observe the hydrolysis to
the boronic acid (2a) via 1H-NMR analysis (Figure 1). The conversion times and final equilibrium
positions were measured in a variety of conditions designed to simulate relevant process
conditions.
The results provide evidence that the system tends towards equilibrium, favoring the boronic
acid (2a), and that equilibrium is achieved in <2 minutes at 27 °C (Figure 2: Red data points). At
the process temperature for the Suzuki reaction (80 °C), the rate of approach to equilibrium should
be much quicker, meaning that an equilibrium position is reached before the Suzuki reaction
starts.15 In a separate experiment, diethanolamine was charged to a freshly prepared solution of 2b
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 28
1
2
3
4
5
6
7
8
9
ethanol (0.56 M 2b) in aqueous K3PO4 (1.2 M) and it was shown that this leads to an equilibrium
that is shifted in favor of the boronic ester (2b) (Figure 2: Blue data points), consistent with our
expectation (Scheme 2).
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
100
80
2b Std Conditions
60
2a Std Conditions
2b Std Conditions + DEA
40
2a Std Conditions + DEA
20
0
0
50
100
150
200
Time (s)
Figure 2: Hydrolysis of the diethanolamine boronic ester vs. time in ethanol/aqueous 1.2 M
K3PO4 (3:2) at 27 °C demonstrating the effect of added DEA upon the final equilibrium position.
1
No evidence was found in the H-NMR data (Figure 1) for the formation of the ring-opened,
partially hydrolyzed intermediate. We therefore defined the equilibrium constant KEQ (eq 1) based
on a direct conversion of the boronic ester (2b) to acid (2a) (Scheme 2).
[
] [ ]
2a . DEA
퐾EQ= [
(1)
2
] [
]
2b . H O
2
However, the partially aqueous conditions used in the process mean that the concentration of
water can be considered a constant and an apparent equilibrium constant, KEQ’ (eq 2) was therefore
used.
[
] [ ]
2a . DEA
2
[
]
퐾EQ' = 퐾EQ H2O =
(2)
[
]
2b
The experimental time course data for the approach to equilibrium were fitted to give a pseudo
first order forward reaction rate constant (kf) and the equilibrium constant KEQ’ (eq 2).16 A good
ACS Paragon Plus Environment
10

Page 11 of 28
Organic Process Research & Development
1
2
3
4
5
6
agreement was generally obtained between the data and the model (Figure 3). Selected values of
the best-fit constants and final equilibrium compositions are summarized in Table 1.
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 3: Result of fitting rate constants and equilibrium constants in DynoChem®.16 Reaction
approach to equilibrium with kf = 0.0559 s-1 and KEQ’ = 2.03 M vs. experimental data.
Table 1: Summary of fitted constants and equilibrium compositions under the reaction and
isolation conditions at 27 °C.
Reaction
Isolation
Isolation
(Experimental) (Predicted) (Experimental)
[H2O] (M)
kf (s-1)
18.6
0.056 ± 0.002
2.03 ± 0.06
0.034
38
38
-
0.234
8.47
KEQ’ (M)
[2a] (M)
[2b] (M)
% 2b
6.37
0.0205
0.0016
6.8
0.0209
0.0012
5.5
0.016
32
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 28
1
2
3
4
5
6
7
8
9
By applying eq 2 it was possible to estimate KEQ under the process conditions and then use this
value to predict KEQ’ and hence the equilibrium position at different water levels.17 A comparison
of the predicted and observed equilibrium positions (KEQ’) under the isolation conditions (Table
1) shows tolerable agreement between the predicted and actual results despite the difference
between the predicted and observed equilibrium constant. This difference is likely to be due to the
changes in the water/ethanol ratio in the solvent leading to changes in the bulk properties of the
reaction solvent and thus changing the rate and equilibrium constants.
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
Assuming that the 0.05 mol equivalent overcharge of the boronic ester used in the reaction
remains unreacted, it is possible to calculate how it will be speciated at the end of the Suzuki
reaction (18.6 M water in ethanol) and during the isolation (38 M water in ethanol). The tabulated
values (Table 1) show that the increased levels of water present in the isolation favors 2a over 2b,
relative to the reaction conditions.18
With the knowledge that under the isolation conditions the boronic ester (2b) is predominantly
hydrolyzed to the boronic acid (2a) (93.2% boronic acid, 6.8% boronic ester), it was concluded
that it was not necessary to develop a separate analytical method capable of detecting ppm levels
of the boronic ester. Consequently, we did not perform additional testing for residual levels of the
boronic ester or amend the specification of lanabecestat (+)-camsylate.
Diethanolamine Effect on Catalytic Cycle. We were interested in investigating the fundamental
reasons for the slower Suzuki reaction when the diethanolamine boronic ester was used because
of the potential wider implications for analogous esters in other Suzuki reactions.
With the knowledge that the boronic ester hydrolyses in <2 minutes under the Suzuki
conditions, the key question that we wanted to address was why is the rate of the Suzuki reaction
ACS Paragon Plus Environment
12

Page 13 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
slower for the boronic ester than the boronic acid? A by-product from the boronic ester hydrolysis
is diethanolamine (Scheme 2) and we questioned whether this could affect the catalysis.
Palladium-diethanolamine complexes are known and diethanolamine-based solid-supported
scavengers have been developed for metal removal.19,20 It was therefore conceivable that
diethanolamine was inhibiting the catalysis through the potentially reversible formation of a non-
productive palladium-diethanolamine complex.
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
100
80
60
Std Conditions
40
20
0
Std Conditions + 1eq DEA
Std Conditions + 1eq DEA + 2[Pd]
0
60
120
180
240
Time (minutes)
Figure 4: Effect of diethanolamine content on the rate of the Suzuki reaction at 80 °C.
The boronic ester (2b) was reacted under Process C conditions (at 80 °C) and passed its residual
starting material in-process control (IPC) test after 90 minutes (Figure 4: Blue line). Conversely,
when diethanolamine (1 equiv.) is charged to the reaction it is slower and requires over 240
minutes to reach its pass criteria (Figure 4: Red line). This result is consistent with deactivation of
palladium by the diethanolamine. We speculate that there is an equilibrium between an unbound
and active, and a bound and inactive palladium-diethanolamine species which, in the presence of
additional diethanolamine, shifts to the bound and inactive form. It was postulated that if an
equilibrium existed, charging 1 equiv. of diethanolamine whilst doubling the palladium loading
should give a reaction profile similar to the standard conditions. This was indeed the case and the
experimental observation was consistent with a palladium-diethanolamine deactivation hypothesis
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 28
1
2
3
4
5
6
(Figure 4: Gold line vs. Red line). Whilst additional work is required to unambiguously determine
if there is an equilibrium, these results are consistent with this hypothesis.
7
8
9
It was important to understand how the control strategy would evolve to adapt to the changes in
catalyst and catalyst loading (between Process B and Process C) prior to implementing them at
production scale. For example, when the boronic acid (2a) was reacted with 0.4 mol% Na2PdCl4
under Process B conditions, the amount of palladium in the isolated solid was 204 ppm (Table 2,
entry 1). This was not in line with our control strategy (<200 ppm in the API, lanabecestat (+)-
camsylate) since we had already determined that using lanabecestat from Process B in the API
stage (salt formation with (+)-camphorsulfonic acid) led to no reduction in palladium level in
lanabecestat (+)-camsylate. Hence, in Process B we implemented a solid-supported scavenger step
during the work-up of the Suzuki stage which reduced the palladium level to <40 ppm (entry 2).
The development of the new catalytic system for Process C (Pd(AmPhos)2Cl2) enabled the loading
to be reduced to 0.15 mol%, which lowered the theoretical maximum palladium level in the
isolated solid from 936 to 351 ppm. When the new catalyst was subjected to the same conditions
as Na2PdCl4, albeit with a different catalyst loading and reaction dilution, the level of palladium
in the solid was unexpectedly similar (entry 1 vs. 4). Interestingly, when simply changing the boron
species from the boronic acid (2a) to the diethanolamine boronic ester (2b), there was a 3.5-fold
reduction in palladium (from 166 ppm to 48 ppm) in the isolated material (entry 4 vs. 7).
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 2 Diethanolamine effect on palladium removal
:
Entry
2
Pd Loading /
Cat.
Additive
Pd in lanabecestat
(ppm)
Theor.a
Actual
ACS Paragon Plus Environment
14

Page 15 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
10
0.40 mol%
Na2PdCl4
2a
“
None
936
204c
40c
b
Solid
“
936
936
351
351
936
351
351
1755
3150
scavengerd
DEA
(1 equiv.)
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
“
“
96c
0.15 mol%
Pd(AmPhos)2Cl2
“
None
166
143
143
48
DEA
(1 equiv.)
“
“
0.40 mol%
2b
“
None
None
b
Na2PdCl4
0.15 mol%
Pd(AmPhos)2Cl2
DEA
(1 equiv.)
“
“
36
0.75 mol%
Pd(AmPhos)2Cl2
“
None
None
405
1255
1.50 mol%
Pd(AmPhos)2Cl2
“
a Theoretical maximum palladium in the isolated lanabecestat assuming 100% yield and no purge
of palladium. b Catalyst was modified with 0.4 mol% DTBPPS. c Process B was used (14 volumes)
in entries 1 and 3. All other entries used Process C (9 volumes). d PhosphonicS SPM32 (20 wt%)
or QuadraSil MP eco (12 wt%) was charged to the batch. The slurry was agitated and filtered
before crystallizing lanabecestat.
The reduction in residual palladium level when we changed to the diethanolamine boronic ester
may be explained by the formation of a palladium-diethanolamine complex. The diethanolamine-
bound complex is likely to have superior solubility in the ethanol/water mixture (isolation
conditions) compared to an unbound species. This coordination effect also supports our hypothesis
on the rate differences between the boronic acid and ester (vide supra); additional control
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 28
1
2
3
4
5
6
7
8
9
experiments were aimed at addressing this. Entry 3 was a repeat of Entry 1, albeit diethanolamine
was charged after the reaction has reached its IPC criteria; the residual palladium was reduced 2-
fold. This trend was not apparent across all reactions tested, for example, there was little difference
between Entries 4 and 5 which mimicked Entries 1 and 3 albeit with a different catalyst.
Additionally, Entry 5 was expected to have a similar palladium level to Entry 7 since the overall
balanced equations for each reaction were equal, however, it was 3-fold higher. Nevertheless, the
reduction in palladium loading and, more critically, the switch from the boronic acid (2a) to the
diethanolamine boronic ester (2b) led to a reduction in palladium level in lanabecestat and
advantageously removed the requirement for a scavenging step in the process.
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
Schemes 3-5 describe the final optimized conditions and process flow diagrams with associated
timelines for the development history of the Suzuki reaction used in the manufacture of
lanabecestat.
Scheme 3: Optimized Process C conditions for the manufacture of lanabecestat.
ACS Paragon Plus Environment
16

Page 17 of 28
Organic Process Research & Development
1
2
3
4
Scheme 4. Process flow diagrams showing the development history.
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 5. Development timelines for Processes B and C and manufacturing dates for Processes
A-C.
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 28
1
2
3
4
Summary
5
6
7
8
Process development between 2014 and 2018 successfully introduced an easily preparable
diethanolamine boronic ester in the Suzuki process which improved accommodation for charging
accuracy. High-throughput experimentation enabled rapid data collection across a broad chemical
space and quickly identified a new catalyst for the process. In addition, the challenging solvent
swaps, phase separations, solid-supported scavenger step and uncontrolled crystallization were all
removed from the process. These improvements positively impacted the overall environmental and
sustainability factors of the product; process mass intensity was reduced from 85 kg waste/kg
lanabecestat in Process A to 35 kg waste/kg lanabecestat in Process C. Additionally, the changes
provided a reduction in cost and time, and an increase in throughput. In 2018, Process C (Scheme
4) successfully manufactured 250 kg of lanabecestat in 4 batches.21 The process was operated at
100 kg scale (based on input of 1) and the palladium levels in lanabecestat, now without a discrete
scavenging step, were an average of 37 ppm (cf. Table 2). Conversion of lanabecestat to the drug
substance (lanabecestat (+)-camsylate) led to a further reduction in palladium (the average
palladium level per batch was 19 ppm) to levels that were 10-fold below that set by the
specification of drug substance (<200 ppm).
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
Changing from the boronic acid to the diethanolamine boronic ester required additional
screening and mechanistic work. This process understanding work facilitated development of a
simple and robust process with associated enhancements in the control of organic impurity levels.
The liberated diethanolamine impacted both the catalysis and the purging of palladium, and as a
result, we were able to remove the expensive and high risk solid-supported scavenging step from
the process. The knowledge we obtained allowed for a process that was successfully scaled 200-
fold, from 500 g in a laboratory to 100 kg in production and the demonstration at this scale has
ACS Paragon Plus Environment
18

Page 19 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
highlighted the value of diethanolamine boronic esters. All these factors will likely stimulate
further academic and industrial research into diethanolamine boronic esters as viable coupling
partners, alternative to boronic acids, in Suzuki reactions.
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
A process description for Process C is provided. In 2017, the process was operated on a 500 g
scale in an R&D lab (5 L vessel) and in 2018 it was successfully scaled to 100 kg in a pilot plant.
Preparation of lanabecestat. To a 5 L jacketed vessel, equipped with an overhead agitator,
temperature probe and condenser, was charged 1 (580.0 g, 924.4 mmol), 2b (223.5 g, 970.6 mmol),
Pd(AmPhos)2Cl2 (1.00 g, 1.39 mmol) and ethanol (3 mL/g, 1740 mL). The slurry was agitated at
20 °C and the oxygen was purged with a flow of nitrogen. Potassium phosphate tribasic (490.6 g,
2311 mmol) was dissolved in water (2 mL/g, 1160 mL) and the solution was charged to the 5 L
vessel. The biphasic mixture was warmed to 80 °C and agitated for 90 minutes. The biphasic
solution was cooled to 55 °C and the lower phase (~1 mL/g) was removed. Water (1 mL/g, 580
mL) was charged to supersaturate the solution and lanabecestat seed (0.15 wt%) was added. The
thin slurry was held for 30 minutes and water (4 mL/g, 2320 mL) was added at a constant rate over
3 hours. The slurry was cooled to 20 °C over 3 hours, held for 1 hour, filtered and washed
(displacement wash) with a mixture of water (1.6 mL/g, 928 mL) and ethanol (0.8 mL/g, 464 mL).
The wet cake was dried under reduced pressure at 40 °C to afford lanabecestat (379.0 g) as a
monohydrate in 93% yield (assay adjusted), 98.9% HPLC purity (area%) and 94.9% 1H-QNMR
assay (99.1% assay on anhydrous basis) as an off-white solid. 1H NMR (400 MHz, d6-DMSO) δ
8.65 (d, J = 2.3 Hz, 1H), 8.50 (d, J = 1.9 Hz, 1H), 7.88 (dd, J = 2.3, 1.9 Hz, 1H), 7.52 (dd, J = 7.8,
1.7 Hz, 1H), 7.39 (d, J = 7.8 Hz, 1H), 6.82 (d, J = 1.7 Hz, 1H), 6.52 (s, 2H), 3.19 (s, 3H), 3.08 (d,
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 28
1
2
3
4
5
6
7
8
9
J = 15.5 Hz, 1H), 2.99 (d, J = 15.5 Hz, 1H), 2.99 – 2.90 (m, 1H), 2.17 (s, 3H), 2.08 (s, 3H), 1.87-
1.77 (m, 2H), 1.51-1.36 (m, 3H), 1.30-1.11 (m, 2H), 1.02-0.92 (m, 1H). 13C NMR (101 MHz, d6-
DMSO): δ 161.5, 160.5, 149.9, 146.1, 145.0, 143.1, 135.8, 135.5, 134.1, 126.4, 125.9, 120.5,
120.2, 109.0, 90.3, 78.7, 76.5, 54.8, 52.2, 39.7, 30.0, 29.0, 28.5, 28.3, 14.2, 4.0.
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
Preparation of 2-(5-(prop-1-ynyl)pyridin-3-yl)-1,3,6,2-dioxazaborocane (2b).
To a vessel containing a solution of 3-bromo-5-(prop-1-ynyl)pyridine (697 kg, 8.49 w/w% in
toluene, 301.9 mols) was charged tetrahydrofuran (2.5 kg/kg, 147 kg) and triisopropylborate (75.5
kg, 401.4 mols) at 25 °C. The solution was agitated, cooled to −70 °C and n-hexyllithium (108 kg,
33 w/w% in hexanes, 390.8 mols) was added over 3 hours. The reaction was quenched on a 15 °C
stirred mixture of isopropyl acetate (2.7 kg/kg, 162 kg) and water (10.6 kg/kg, 633 kg). The bulk
mixture was warmed to 25 °C, the layers were separated and the lower aqueous layer was washed
twice with isopropyl acetate (2.7 kg/kg, 162 kg). The aqueous phase was charged to a clean vessel
and 2-methyl tetrahydrofuran (8.6 kg/kg, 513 kg) was added followed by 13% sulfuric acid (2
kg/kg, 126 kg). The layers were separated and the organic layer was washed with water (4 kg/kg,
240 kg). The organic solution was distilled to 6.5 kg/kg (390 kg) at 100 mbar/40 °C, and
azeotropically dried via continuous distillation with 2-methyl tetrahydrofuran (15.4 kg/kg, 924
kg). A solution of diethanolamine (13.5 kg, 128.4 mols) in isopropanol (0.1 kg/kg, 6.4 kg) was
added to the 2-methyl tetrahydrofuran solution at 25 °C and after 15 minutes 2b seed (0.15 wt%)
was added. A solution of diethanolamine (13.5 kg, 128.4 mols) in isopropanol (0.1 kg/kg, 6.4 kg)
was then added over 30 minutes and the slurry was held for 30 minutes. The slurry was cooled to
5 °C over 30 minutes, held for 1 hour and filtered. The filter cake was washed with methyl tert-
butyl ether (1.5 kg/kg, 90 kg) and a mixture of methyl tert-butyl ether/acetonitrile (1:1 v/v, 1.5
ACS Paragon Plus Environment
20

Page 21 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
kg/kg, 90 kg). The wet cake was dried under reduced pressure at 50 °C to afford 2b (51.15 kg) in
71% yield (assay adjusted), 100% HPLC purity (area%) and 96.2% HPLC assay as an off-white
solid. 1H NMR (400 MHz, d6-DMSO) δ 8.46 (d, J = 1.6 Hz, 1H), 8.36 (d, J = 2.3 Hz, 1H), 7.70 -
7.68 (m, 1H), 7.08 - 7.00 (m, 1H), 3.91 - 3.83 (m, 2H), 3.82 - 3.75 (m, 2H), 3.11 (tdd, J = 7.0, 9.2,
11.7 Hz, 2H), 2.91 - 2.84 (m, 2H), 2.04 (s, 3H). 13C NMR (101 MHz, d6-DMSO) δ 152.28, 149.50,
142.51, 118.88, 88.40, 77.67, 63.11, 50.64, 3.87.
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 Supporting Information is available free of charge on the ACS
Publications website.
Process descriptions and kinetic data (PDF).
AUTHOR INFORMATION
Corresponding Author. *E-mail: Phillip.Inglesby@AstraZeneca.com, tel.: +44 (0)1625 51 57
56.
ORCID. Phillip A. Inglesby: 0000-0003-3370-4656; Ian W. Ashworth: 0000-0001-7210-3875.
Notes. The authors declare no competing financial interest.
Author Contributions
The manuscript was written by Phillip A. Inglesby. All authors have given approval to the final
version of the manuscript.
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 28
1
2
3
4
5
6
Funding Sources. None
Notes. None
7
8
9
Any additional relevant notes should be placed here.
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
ACKNOWLEDGMENT
None
ABBREVIATIONS
None
REFERENCES
1. (a) Menting, K. W.; Claassen, J. A. β-Secretase Inhibitor; A Promising Novel Therapeutic
Drug in Alzheimer's Disease. Front. Aging Neurosci. 2014, 6, 165. (b) Vassar, R.; Kuhn, P.
H.; Haass, C. et al. Function, Therapeutic Potential and Cell Biology of BACE Proteases:
Current Status and Future Prospects. J. Neurochem. 2014, 130, 4. (c) Yan, R.; Vassar, R.
Targeting the β-Secretase BACE1 for Alzheimer's Disease Therapy. Lancet Neurol. 2014, 13,
319.
2. (a) Taylor, B.; Fernandez Barrat, C.; Woodward, R. L.; Inglesby, P. A. Synthesis of 2-
Oxopropanethioamide for the Manufacture of Lanabecestat: A New Route for Control,
Robustness, and Operational Improvements. Org. Process Res. Dev. 2017, 21, 1404. (b)
Cantillo, D.; Inglesby, P. A.; Boyd, A.; Kappe, C. O. Hydrogen Sulfide Chemistry in
Continuous Flow: Efficient Synthesis of 2-Oxopropanethioamide. J. Flow. Chem. 2017, 7, 29.
(c) Znidar, D.; Hone, C. A.; Inglesby, P. A.; Boyd, A.; Kappe, C. O. Development of a
Continuous-Flow Sonogashira Cross-Coupling Protocol using Propyne Gas under Process
ACS Paragon Plus Environment
22

Page 23 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Intensified Conditions. Org. Process Res. Dev. 2017, 21, 878. (d) Znidar, D.; Cantillo, D.;
Inglesby, P. A.; Boyd, A.; Kappe, C. O. Process Intensification and Integration Studies for the
Generation of a Key Aminoimidazole Intermediate in the Synthesis of Lanabecestat. Org.
Process Res. Dev. 2018, 22, 633.
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
3. (a) For recent reviews on the Suzuki coupling, see: (a) Roy, D.; Uozumi, Y. Recent Advances
in Palladium‐Catalyzed Cross‐Coupling Reactions at ppm to ppb Molar Catalyst Loadings.
Adv. Synth. Catal. 2018, 360, 602. (b) Maluenda, I.; Navarro, O. Recent Developments in the
Suzuki-Miyaura Reaction: 2010–2014. Molecules 2015, 20, 7528. (c) Johansson Seechurn
C.C.C.; Kitching, M.O.; Colacot, T.J.; Snieckus, V. Palladium‐Catalyzed Cross‐Coupling: A
Historical Contextual Perspective to the 2010 Nobel Prize. Angew. Chem. Int. Ed. 2012, 51,
5062.
4. For recent examples of production-scale Suzuki processes, see: (a) Hansen, M. M.; Kallman,
N. J.; Koenig, T. M. et. al. Double Heck Route to a Dibenzoxepine and Convergent Suzuki
Cross-Coupling Strategy for the Synthesis of an MR Antagonist. Org. Process Res. Dev. 2017,
21, 208. (b) Smith, M. J.; Lawler, M. J.; Kopp, N. et. al. Development of a Concise
Multikilogram Synthesis of LPA-1 Antagonist BMS-986020 via a Tandem Borylation–
Suzuki Procedure. Org. Process Res. Dev. 2017, 21, 1859. (c) Maity, P.; Reddy, V. V.; Mohan,
J. et. al. Development of a Scalable Synthesis of BMS-978587 Featuring a Stereospecific
Suzuki Coupling of a Cyclopropane Carboxylic Acid. Org. Process Res. Dev. 2018, 22, 888.
5. (a) Carey, J. S.; Laffan, D.; Thomson, C.; Williams, M. T. Analysis of the Reactions used for
the Preparation of Drug Candidate Molecules. Org. Biomol. Chem. 2006, 4, 2337. (b) Magano,
J.; Dunetz, J. R. Large-Scale Applications of Transition Metal-Catalyzed Couplings for the
Synthesis of Pharmaceuticals. Chem. Rev. 2001, 111, 2177. (c) Roughley, S. D.; Jordan, A.
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 28
1
2
3
4
5
6
7
8
9
M. The Medicinal Chemist’s Toolbox: An Analysis of Reactions Used in the Pursuit of Drug
Candidates. J. Med. Chem. 2011, 54, 3451. (d) Brown, D. G.; Boström, J. Analysis of Past
and Present Synthetic Methodologies on Medicinal Chemistry: Where Have All the New
Reactions Gone? J. Med. Chem. 2016, 59, 4443.
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
6.
7. For a review on selecting boron reagents in Suzuki reactions, see: Lennox, A. J. J.; Lloyd-
Jones, G. Selection of Boron Reagents for Suzuki–Miyaura Coupling. Chem. Soc. Rev. 2014, 43,
412.
8. The boronic esters 2c and 2d were stable, monomeric intermediates that were isolated as
solids. The pinacol ester (2d) was very slow to react in the Suzuki reaction and competing
hydrolysis to form amide (4) was significant. The neopentylglycol ester (2c) showed similar
rate and impurity profile to the boronic acid (typical relative rates were: 2a ≈ 2c > 2b >> 2d).
9. Whilst the hydrolysis of diethanolamine-boronic esters to boronic acids under acidic
conditions has been reported, we are unaware of any reports on hydrolysis rates under typical
Suzuki conditions. For an example of hydrolysis under acidic conditions, see: (a) Caron, S.;
Hawkins, J. M. Directed Ortho Metalation of Neopentyl Benzoates with LDA:ꢀ Preparation of
Arylboronic Acids. J. Org. Chem. 1998, 63, 2054. (b) Sun, J.; Perfetti, M. T.; Santos, W. L.
ACS Paragon Plus Environment
24

Page 25 of 28
Organic Process Research & Development
1
2
3
4
5
6
A Method for the Deprotection of Alkylpinacolyl Boronate Esters. J. Org. Chem. 2011, 76,
3571.
7
8
9
10. The diethanolamine boronic ester (2b) crystallized upon charging a solution of diethanolamine
in iso-propanol to a solution of the boronic acid/boroxine (generated in-situ via a Li-Br
exchange borylation sequence) in Me-THF. The boronic ester (2b) was isolated as a
monomeric, crystalline and stable solid. In 2018, 200 kg of 2b was manufactured in 4 batches
using this process. See Experimental section for further details.
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
25

Organic Process Research & Development
Page 26 of 28
1
2
3
4
5
6
7
8
9
11. For a review on diethanolamine boronic esters, see: Bonin, H.; Delacroix, T.; Gras, E.
Dioxazaborocanes: Old Adducts, New Tricks. Org. Biomol. Chem. 2011, 9, 4714.
12. For the preparation of diethanolamine esters from boronic acids or pinacol esters, see: (a)
Sharp, M. J.; Cheng, W.; Snieckus, V. Synthetic Connections to the Aromatic Directed
Metalation Reaction. Functionalized Aryl Boronic Acids by Ipso Borodesilylation. General
Syntheses of Unsymmetrical Biphenyls and m-Terphenyls. Tetrahedron Lett. 1987, 28, 5093.
(b) Hodgson, P. B.; Salinge, F. H. The Preparation of a Stable 2-Pyridylboronate and its
Reactivity in the Suzuki–Miyaura Cross-Coupling Reaction. Tetrahedron Lett. 2004, 45, 685.
(c) Bonin, H.; Delbrayelle, D.; Domonchaux, P.; Gras, E. Base Free Aryl Coupling
of Diazonium Compounds and Boronic Esters: Self-Activation Allowing an Overall Highly
Practical Process. Chem. Commun. 2010, 46, 2677. (d) Reilly, M. K.; Rychnovsky, S. D.
DABO Boronates: Stable Heterocyclic Boronic Acid Complexes for Use in Suzuki-Miyaura
Cross-Coupling Reactions. Synlett 2011, 16, 2392. (e) Bonin, H.; Leuma-Yona, R.; Marchiori,
B.; Domonchaux, P.; Gras, E. Highly Practical Boronic Acid Surrogates for the Suzuki–
Miyaura Cross-Coupling. Tetrahedron Lett. 2011, 52, 1132.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13. Although we could not unambiguously identify whether the protons of interest were at the 2-
or 6-position of 2a and 2b, the chemical shifts suggest that we were observing the proton at
the 2-position.
14. The reaction mixture was prepared by dissolving 2b in ethanol (0.56 M) and mixing it with
an aqueous solution of K3PO4 (1.2 M). This results in a biphasic mixture with much of the
water from the K3PO4 solution absorbed into the ethanol fraction. It is assumed, based on the
high ionic strength of the aqueous fraction, that all organic species are retained in the organic
ACS Paragon Plus Environment
26

Page 27 of 28
Organic Process Research & Development
1
2
3
4
5
6
fraction and that the volume increase of the organic fraction is due entirely to the addition of
water. This volume increase is used to calculate the molarity of the water in ethanol.
7
8
9
15. The Suzuki reaction does not start until the batch temperature is >70 °C and typical heat-up
rates in production are 0.5 °C/minute. Therefore, the boronic ester will have achieved
equilibrium much before the Suzuki reaction starts.
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
16. Dynochem® 2011, V.4.0.0.0, www.scale-up.com. Details of the model used may be found in
the Supporting Information.
17. An algebraic solution to the equilibrium problem was used to calculate the composition from
the equilibrium constant and starting concentrations. This can be found in the Supporting
Information.
18. The proven acceptable range for the boronic ester charge of 1.05 mol equivalents +/−0.05 mol
equivalents means that an excess of 0.10 mol equivalents may be charged. In this worst-case
scenario, the residual boron species will exist as 7.0% boronic ester (2b) and 93.0% boronic
acid (2a).
19. For examples of palladium-diethanolamine complexes, see: (a) Petrović, Z. D.; Djuran, M. I.;
Heinemann, F. W.; Rajković, S.; Trifunović, S. Synthesis, Structure, and Hydrolytic Reaction
of trans-Dichlorobis(diethanolamine-N)palladium(II) with N-Acetylated L-Histidylglycine
Dipeptide. Bioorg. Chem. 2006, 34, 225. (b) Petrović, Z. D.; Hadjipavlou-Litina, D.; Pontiki,
E.; Simijonović, D.; Petrović. V. P. Diethanolamine Pd(II) complexes in Bioorganic Modeling
as Model Systems of Metallopeptidases and Soybean Lipoxygenase Inhibitors. Bioorg. Chem.
2009, 37, 162. (c) Petrović, Z. D.; Petrović, V. P.; Simijonović, D.; Marković, S. Mechanistic
Pathways for Oxidative Addition of Aryl Iodides to the Low-Ligated Diethanolamine
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 28
1
2
3
4
Palladium(0) Complex in Phosphine-Free Heck Reactions J. Organomet. Chem. 2009, 694,
5
6
3852.
7
8
9
20. For examples of solid-supported diethanolamine acting as a palladium scavenger, see: (a)
Tsukamoto, H.; Suzuki, T.; Sato, M.; Kondo, Y. Chromatography-Free Pd-Catalyzed
Deprotection of Allyl Ethers using PS-DEAM as a Scavenger of Boronic Acids and Pd
Catalyst. Tetrahedron Lett. 2007, 48, 8438. (b) Antonow, D.; Cooper, N.; Howard, P. W.;
Thurston, D. E. Parallel Synthesis of a Novel C2-Aryl Pyrrolo[2,1-c][1,4]benzodiazepine
(PBD) Library. J. Comb. Chem. 2007, 9, 437.
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
21. To the best of our knowledge, these are the first reported examples of a multi-kilo manufacture
of a diethanolamine boronic ester and subsequent Suzuki coupling.
ACS Paragon Plus Environment
28