Development and scale-up of an efficient miyaura borylation process using tetrahydroxydiboron

Article abstract of DOI:10.1021/acs.oprd.6b00345

A revised Miyaura borylation process has been developed using tetrahydroxydiboron that avoids the use of bis(pinacolato) diboron and hence the need to hydrolyze the resulting boronic ester to its corresponding acid. The process was greatly simplified and successfully scaled-up in the pilot plant on a 65 kg scale, reducing plant cycle time and resulting in a 47% overall cost reduction. Furthermore, methodology for the study of the oxygen sensitivity of the process is reported that allowed for optimization of the amount of tetrahydroxydiboron and catalyst used. These studies also demonstrated an oxygen-induced decomposition of tetrahydroxydiboron.

Full text of DOI:10.1021/acs.oprd.6b00345

Subscriber access provided by University of Otago Library
Full Paper
Development and scale up of an efficient Miyaura
borylation process using tetrahydroxydiboron
Sing R. Gurung, Christopher Mitchell, Jie Huang, Marco Jonas, Josiah D Strawser, Elena
Daia, Andrew Hardy, Erin O'Brien, Frederick Hicks, and Charles Dimitrios Papageorgiou
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00345 • Publication Date (Web): 20 Nov 2016
Downloaded from http://pubs.acs.org on November 23, 2016
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 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Development and scale up of an efficient Miyaura borylation process using
tetrahydroxydiboron
1
2
1
1
2
Sing R. Gurung , Christopher Mitchell , Jie Huang , Marco Jonas , Josiah D. Strawser , Elena
3
2
2
2
2
Daia , Andrew Hardy , Erin O’Brien , Frederick Hicks , Charles D. Papageorgiou *
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
Albany Molecular Research Inc, 21 Corporate Circle, Albany, NY, 12203
Takeda Pharmaceuticals International Co., Process Chemistry, 40 Landsdowne St., Cambridge, MA, 02139
Carbogen AMCIS Ltd, 303 Clayton Lane, Manchester, M11 4SX
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table of Contents Graphic
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HO
OH
OH
B B
Br
OH
B
N
O
HO
NH2
NH2
N
O
Pd
P
HO
Cl
NH2
Pdꢀ168, KOAc
MeOH.
tꢀBu
tꢀBu
tꢀBu
1
2
Pd-168
2
ACS Paragon Plus Environment

Page 3 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Keywords: tetrahydroxydiboron, Miyaura borylation, oxygen sensitivity, ethylene glycol, cost
analysis
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
A revised Miyaura borylation process has been developed using tetrahydroxydiboron that avoids
the use of bis(pinacolato) diboron and hence the need to hydrolyze the resulting boronic ester to
its corresponding acid. The process was greatly simplified and successfully scaledꢀup in the pilot
plant on a 65 kg scale, reducing plant cycle time and resulting in a 47 % overall cost reduction.
Furthermore, methodology for the study of the oxygen sensitivity of the process is reported that
allowed for optimization of the amount of tetrahydroxydiboron and catalyst used. These studies
also demonstrated an oxygenꢀinduced decomposition of tetrahydroxydiboron.
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Boronic acids are not only extremely useful reagents in a number of crossꢀcoupling reactions, but
they are also of great interest in medicinal chemistry. Their synthesis typically relies on one of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1, 2
two main methods.
The most atom efficient and cost effective approach involves a metalꢀ
halogen exchange of the corresponding halide using an organolithium or organomagnesium
3
reagent followed by quench with a trialkyl borate. However, this method presents limitations
with respect to functional group tolerance, as many pharmaceutically relevant substrates contain
base sensitive moieties within their structures. An alternative, more versatile strategy involves a
transition metalꢀcatalyzed borylation that provides access to a wide range of aryl and
heteroarylboron derivatives with a high degree of complexity. These methods are less cost
effective and atom efficient typically utilizing palladium (Pd) catalysts and bis(pinacolato)
4
ꢀ7
diboron (B Pin ) as the boron source.
The resulting boronic esters are then typically
2
2
hydrolyzed, in an additional step to their corresponding boronic acids, a process that can be
8
challenging given its reversible nature, and that affords large amounts of pinacol waste.
Removal of this pinacol waste from the reaction mixture can also be troublesome, requiring
laborious extraction and distillation techniques as the development of suitable crystallization
processes is often plagued by oiling issues. Finally, reaction conditions need to be identified that
minimize the competing deꢀhalogenation and homoꢀcoupling of the aryl halide starting material,
that can significantly impact the overall process yield and isolated product purity.
Figure 1. Synthesis of arylboronic acid 2 and related impurities
4
ACS Paragon Plus Environment

Page 5 of 33
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
Arylboronic acid 2 is a key starting material in the manufacturing process for TAKꢀ117, a
selective PI3Kα inhibitor currently in Phase 1b clinical trials (Figures 1 & 2). Treatment of the
bocꢀprotected aryl bromide 1 with stoichiometric quantities of secꢀBuLi under a wide range of
conditions followed by a quench with B(OMe) resulted predominantly in deꢀbromination. This
3
indicates that the metalꢀhalogen exchange may be faster than the deprotonation of the free
hydrogen on the nitrogen, this internal proton source effectively quenching the desired lithiated
species. A rapid reverse addition of the arylꢀbromide to an excess of secꢀBuLi was shown to
enable both complete lithiumꢀhalogen exchange and nitrogen deprotonation. However the rapid
8
heat evolution associated with this approach makes it unsuitable for batch processing.
Therefore a Pd mediated process was developed, outlined in Figure 2 that employed B Pin as
2
2
the borylating agent and utilized a relatively high loading of bis(triphenylphosphine)palladium
II) chloride catalyst (0.01 eq.). Aryl bromide 1, B Pin and KOAc were charged to a clean and
(
2
2
dry reactor, flushed with nitrogen followed by 15 volumes of 2ꢀBuOH. Oxygen was removed
from the suspension by conducting 3 vacuum/ N purge cycles before adding the catalyst. The
2
reaction mixture was then heated to 100 °C for 1ꢀ2 hours after which time an inꢀprocess control
(
IPC) sample was removed to confirm that the overall conversion was within the set specification
of noꢀlessꢀthan (NLT) 99.0%. The reaction mixture was cooled to 20 °C and a polish filtration
was conducted to remove the generated KBr salts. The solids were washed with 2ꢀBuOH and the
filtrate concentrated to 6 volumes. Concentrated HCl was finally added to mediate the hydrolysis
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of the ester that took place within 4ꢀ6 hours at 85 °C. Once the hydrolysis was complete
(achieved a NLT 98.0% conversion), the suspension was cooled to 20 °C, filtered and the wetꢀ
cake washed with 2ꢀBuOH and water. The wetꢀcake was reꢀslurried in water to remove the
residual inorganics (mainly boric acid) and then filtered and washed with water and 2ꢀBuOH
before being dried under vacuum at 45 °C.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Original B Pin ꢀmediated Miyaura borylation of aryl bromide 1
2
2
Clean, dry, nitrogen flushed
reactor
Cool to 20 °C ± 5 °C
Compound 1 (1eq.)
Bisꢀ(pinacolato)diboron (1.1 eq.)
2ꢀButanol (15 L/kg)
Isolate solids via filtration
Pd(PPh ) Cl (0.01 eq.)
3
2
2
KOAc (3 eq.)
Filtrate
Organic wash
2ꢀButanol (1 L/kg)
Water (5 L/kg)
Heat to 100 °C ± 5 °C
for 1.5ꢀ2 hrs
Reslurry solids for 1 hr
Cool to 20 °C ± 5 °C
Isolate solids via filtration
Filtrate
Polish Filtration
Water (1 L/kg)
2ꢀButanol (1 L/kg)
Inorganic residues
Aqueous wash
Organic wash
2ꢀButanol (2 L/kg)
Distil to 6 volumes
2
ꢀButanol
Dry under vacuum at 45 °C
1
2N HCl (3.5 eq.)
Heat to 85 °C ± 5 °C
for 4 ꢀ6 hrs
Arylboronic acid 2
A total of 5 batches were manufactured according to this process on a 20ꢀ50 kg scale that
afforded product with good overall purity (95.8ꢀ97.5%), with the main impurities being the deꢀ
6
ACS Paragon Plus Environment

Page 7 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
brominated (3) and homoꢀcoupled (4) side products that were each observed, during IPC
analysis, in 7ꢀ13% (Figure 1). The process yield was also low and variable, between 60 and 75%,
which was of great importance as the boronic acid is a significant cost contributor, accounting
for approximately 19 % of the overall active pharmaceutical ingredient (API) cost. As
highlighted in Figure 2, the process was extremely laborious with a large number of unit
operations that resulted in a long average batch cycle time of approximately 9.25 days. There
were also significant challenges with determining the end point of the hydrolysis step as
appropriate conditions to quench the reaction during sample preparation for IPC analysis could
not be identified, due to the reversible nature of the boronate hydrolysis. This resulted in a
number of batches being stopped before the desired endꢀpoint had been reached. Finally,
conversion was highly variable on scale, taking 2 to greater than 12 hrs, and in some cases it
stalled and needed additional B Pin and catalyst to reach full conversion.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
2
This paper describes the subsequent studies conducted toward developing and scaling up into the
pilot plant an alternative more atom economical, higher yielding and robust process utilizing
tetrahydroxydiboron (BBA). BBA has only recently been reported by Molander et al. in the
direct metalꢀcatalyzed synthesis of arylboronic acids and there is only one report of its use in the
9
ꢀ14
large scale manufacture of boronic acids.
This method utilized Buchwald’s second generation
XPhos PdꢀG2 catalyst that lead to a more rapid in situ formation of the requisite Pd(0) species.
EXPERIMENTAL SECTION
General high throughput screening experimental procedure. All reactions were performed
using 0.75 mmol (160 mg) of aryl bromide 1. All solid reagents were placed in 24 x 150 mm
Radley glass reaction tubes. The tubes were placed in the Radley Carousel Reaction Station
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
equipped with a cooling reservoir and lid, and were evacuated under vacuum and purged with
nitrogen for a total of 5 cycles. Solvent (25 vol.) and liquid reagents were then added using
syringes, and the resulting mixtures were heated to 90 °C for a total of 16 hours. For all
reactions, 3 eq. of base and corresponding borylating reagent were used. The reaction mixtures
were then analyzed by HPLC.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General experimental procedure. All reactions were performed in a 3ꢀneck 100 mL cylindrical
jacketed reactor equipped with an overhead stirrer, temperature probe and condenser. Aryl
bromide 1 (5.0 g, 23.4 mmol, 1 eq.), BBA (1.0ꢀ6.0 eq.), KOAc (2.5ꢀ3.0 eq.) and Pdꢀ168 (0.01ꢀ
0
.25 mol %) were charged to the reactor that was then subjected to three vacuum/ N purge
2
cycles. MeOH (50.0 mL, 10 vol.), followed by ethylene glycol (3.0 vol.) if applicable, was
charged and agitation was initiated at 200 rpm. Three additional vacuum/ N purge cycles were
2
conducted and the reaction mixture was heated to the desired temperature over 0.5ꢀ3 hrs. When
the reaction was complete, as determined by HPLC analysis, water (50 mL, 10 vol.) was charged
over 1 hr and the resulting slurry aged for 2ꢀ16 hrs. The slurry was filtered under vacuum and the
wetꢀcake washed with MeOH:Water/ 1:1 (25 mL, 5 vol.) and water (25 mL, 5 vol.) before being
dried under vacuum at 50 °C to afford boronic acid 2 as an offꢀwhite solid.
Oxygen sensitivity studies. All reactions were performed in a 0.3 L vacuum jacketed dish
bottom reactor equipped with an overhead stirrer, temperature probe, MettlerꢀToledo Ingold
Inpro 6850iG gas oxygen sensor (part number: 52 206 435), MettlerꢀToledo Ingold Inpro 6850i
dissolved oxygen sensor (DO, part number: 52 206 121) and Mettler Toledo Easysampler 1210
(part number: 30083901). The oxygen composition in the reactor headspace was controlled with
a Brooks 0254 controller (part number: 0254AB1B15A) that controlled two Brooks SLA5850
8
ACS Paragon Plus Environment

Page 9 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
thermal mass flow controllers calibrated for each N2 and O2 (part number:
SLA5850S1BAB1C2A1, Figure 3).
The O headspace composition of the reactor was set to the desired level and a slight positive
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
pressure was maintained in the reactor (by setting the total gas flow rate to 100 mL/min), which
was typically higher when the reagents were charged than during the course of the reaction (total
gas flow rate was typically 30 mL/min during the course of the reaction) to minimize the
introduction of air into the reactor headspace. Aryl bromide 1 (20.0 g, 93.9 mmol, 1 eq.) and
KOAc (23.0 g, 234.8 mmol, 2.5 eq.) were charged to the reactor followed by MeOH (150 mL,
7.5 vol.) and ethylene glycol as applicable (60 mL, 3.0 vol.), and the reaction mixture agitated at
the desired agitation speed for 30 min to allow for the liquid and headspace O concentrations to
2
equilibrate. A mixture of BBA (10.5 g, 117.3 mmol, 1.25 eq.) and Pdꢀ168 (24 mg, 0.0470 mmol,
0.05 mol%) was charged, and the reaction temperature was quickly adjusted to 35 °C taking
advantage of the reaction exotherm to reach the setꢀpoint. Sampling was initiated using the
Easysampler (a total of 24 samples were taken) and the reaction mixture aged for a total of 24
hrs, after which time it was discarded.
Figure 3. Experimental setꢀup for conducting the oxygen sensitivity studies
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Mettler Toledo
EasySampler
Probe
Brooks Mass Flow
Controllers:
N2
Gas O Sensor
2
O2
Brooks 0254
Controller
Purge Gas Inlet
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
.3L Vacuum
Jacketed Glass
Reactor
Gas Mixing Zone
Pilot plant experimental procedure. Scaleꢀup was performed in a 2800 L GLMS Pfaudler
jacketed reactor equipped with temperature and pressure probes, a MettlerꢀToledo Ingold Inpro
6
850iG gas oxygen sensor and a MettlerꢀToledo Ingold Inpro 6850i dissolved oxygen sensor.
Each oxygen sensor was connected to an M400 Type 2 singleꢀchannel multiparameter
transmitter and the data was recorded using a Yokogawa datalogger. The oxygen composition in
the reactor headspace was modified and controlled using two 40ꢀ400 mL/min flow meters
manufactured by PLATON (part number: PGU #24SS/Vit) one for each N and compressed air.
2
The reactor was charged with aryl bromide 1 (65 kg, 305 mol, 1 eq.), KOAc (74.6 kg, 760 mol,
2.5 eq.) and MeOH (475 L, 7.5 vol.) followed by ethylene glycol (51.3 L, 3 eq). During the
charge of these materials, a positive N pressure was maintained by flowing N at 10L/min to
2
2
minimize the introduction of air into the reactor headspace. At the end of the charge, stirring was
initiated at 54 rpm and the mixture was cooled to 5 °C. A cycle of five vacuum/ N purge cycles
2
was then performed and the oxygen content was confirmed to be 0.18 ppm in solution and 0.4 %
in the headspace.
1
0
ACS Paragon Plus Environment

Page 11 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The agitator was stopped and a mixture of BBA (34.1 kg, 380.4 mol, 1.25 eq.) and Pdꢀ168 (78 g,
1
52 mmol, 0.05 mol %) was charged. Five additional vacuum/ N purge cycles were conducted,
2
the stirring was reꢀinitiated at 54 rpm, and the reaction temperature was adjusted to 35 °C.
The mixture was aged at this temperature for 12 hrs after which time IPC analysis indicated that
it had reached full conversion (100.0 %). Water (650 kg, 10 vols) was charged over 1 hr and the
mixture was cooled to 20±5 °C over 1 hr, stirred at this temperature for 1 hr and sampled for the
determination of the crystallization endꢀpoint. In process analysis showed that the concentration
of the product in the mother liquors was below the specification of NMT 3.0 mg/mL (2.5
mg/mL), and therefore the product was filtered. The wetꢀcake was washed with a 1:1 mixture of
MeOH:water (97.5 kg, 3.0 vol.) and then three times with water (97.5 kg, 3 x 3.0 vol.) and
finally dried under vacuum at 45 °C to afford 50.7 kg of boronic acid 2 (91.1%) as an off white
solid with an HPLC purity of 98.8 % containing 1.4 % of compound 3 and 0.4 % of compound 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
B NMR analysis. 200 mg of BBA was placed in a 25 mL round bottom flask and the
headspace was displaced with nitrogen (if applicable) via a sequence of vacuum/ N pressure
2
cycles. In a separate 25 mL round bottom flask, KOAc, Pdꢀ168 and/ or ethylene glycol were
added, as applicable followed by 9 mL of MeOD. The mixture was then purged with nitrogen or
air for 1 hour before being transferred to the flask containing the BBA. The resulting mixture
was agitated at ambient temperature until fully dissolved and a sample transferred to an
11
appropriate quartz NMR tube for B NMR analysis using a Bruker 400 MHz instrument and a
relaxation time of 1.0 sec. The extent of degradation was calculated according to equation 1.








1

% degradation = 
×2× B(OH) ×100
(1)
∫
3



BBA× + B OH 
2
(
)


∫
∫
3






1
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HPLC Method. The HPLC method used for IPC analysis as well as for analyzing the purity of
boronic acid 2 employed a Symmetry Shield RP1 Column, 100 Å, 3.3 µm, 4.6 x 150 mm
maintained at 25 °C. A 20 mM NH OAc Buffer/ CH CN, 90/10 (v/v) solution was used as
4
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mobile Phase A and a 20 mM NH OAc Buffer/ CH CN, 10/90 (v/v) solution as Mobile Phase B.
4
3
The total flow rate was set to 1.0 mL/min, the injection volume was 5.0 ꢁL and the detection was
carried out at 285 nm. The total method analysis time was 20 min. A gradient was used starting
at 100 % of Mobile Phase A, moving to 69 % after 5 min, 50 % after 7 min and ending at 38 %
at 10 min. The latter composition was maintained for 2 min after which time the eluent was
switched to 100 % Mobile Phase B for 4 min. Finally, the column was reꢀequilibrated at 100 %
Mobile Phase A for a total of 4 min.
The sample preparation procedure was as follows: 1) for IPC analysis, 25 ꢁL of the reaction
mixture was diluted with MeOH:H O/ 70:30 in a 2 mL volumetric flask; 2) for purity analysis,
2
2
5 mg of boronic acid 2 was dissolved in a 50 mL volumetric flask using MeOH:H O/ 70:30.
2
Compounds 2, 3 and 1 eluted at a relative retention time (RRT) of 1.40, 1.86 and 1.98 min
respectively.
RESULTS AND DISCUSSION
High throughput screen. Pinacolborane (HBPin), neopentylglycolborane (HBNpt), bisboronic
acid (BBA) and tetrakis(dimethylamino)diboron (B (NMe ) ) were screened using Pdꢀ100
2
2 4
9
, 15ꢀ
(Pd(PPh ) Cl ), Pdꢀ106 (PdCl (dppf)—CH Cl ) and the second generation SPhos palladacycle.
3
2
2
2
2
2
19
Three solvents were screened at 25 vol. (2ꢀBuOH, toluene and 1,4ꢀdioxane) using 2 mol % of
the catalyst, 3 eq. of KOAc and 3 eq. of the borylating reagent at 90 °C. HBPin and HBNpt did
not lead to any significant conversions and mostly deꢀbromination was observed with HBPin
1
2
ACS Paragon Plus Environment

Page 13 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
while the aryl bromide 1 was recovered with HBNpt. B (NMe ) resulted in poor conversions
2
2 4
(<40 %) and high levels of 3. All reactions using BBA in 2ꢀBuOH went to completion (>99 %
conversion), however with relatively high levels (5ꢀ15 %) of both deꢀbromination and homoꢀ
coupling.
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
BBA was selected and a high throughput catalyst screen was conducted using a 2 mol % loading
investigating both preformed and in situ generated catalysts (using a Pd(OAc) /L ratio of 1:2).
2
The catalyst loading for the hits out of this screen was then progressively lowered to 0.5 mol %.
20
Pdꢀ127 (PdCl (dcypf), Pdꢀ116 (Pd[P(tBu) ] ) and Pdꢀ168 emerged as the front runners from
2
3 2
this screen affording >99 % conversion with ~5 % of 3. Pdꢀ127 afforded significantly higher
levels of 4 than either Pdꢀ116 or Pdꢀ168 (7 % vs. 3 %).
The catalyst loading was further reduced to 0.25 mol %, and the 3 hits were subjected to a
solvent screen investigating 1ꢀBuOH, EtOH and MeOH. The levels of both compounds 3 and 4
were reduced to 1ꢀ3 % for Pdꢀ118 and Pdꢀ168 in MeOH, while high levels (9.1 %) of compound
3
were observed for Pdꢀ127.
Further reduction of the catalyst loading to 0.1 mol % and 0.05 mol % demonstrated that Pdꢀ168
was the more active catalyst and was subsequently selected for further process optimization. At
0.05 mol % a 96 % conversion was observed for Pdꢀ168, with 4.1 % of 3 and 2.4 % of 4,
compared to only 45 % conversion for Pdꢀ118.
Evaluation of the screening hit. The conditions identified from the screen were successfully
scaled up to 5 g (Table 1, Entry 8) using 0.25 mol % of the Pdꢀ168 catalyst. However, a large
excess of BBA was used (2.0 eq.) which is a major cost contributor ($750/ kg) for the boronic
acid (Figure 11). Reduction of the amount of this reagent resulted, as expected, to larger levels of
1
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
homoꢀcoupling (4; Table 1, Entries 7ꢀ9 & 4), which was found to be an impurity that was
difficult to purge during the isolation of the product and in downstream processing, and therefore
needed to be controlled to noꢀmoreꢀthan (NMT) 3.0 %. Furthermore, the amount of compound 3
formed as a result of deꢀbromination increased from 4.3 % to 8.4 %, as the BBA stoichiometry
increased from 2 to 6 eq. Using only 1.0 eq., resulted in a 50.7 % conversion and therefore, 2.0
eq. offered the optimal compromise between cost, conversion and reaction impurity profile.
Under these conditions the reaction tolerated up to 2.5 vol % water, reaching complete
conversion within 4 hrs affording 4.7 % and 3.5 % of compounds 3 and 4 respectively.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Reaction mixture at the end of the reaction: A) at 65 °C, B) at 35 °C, 1.0 eq. 1, 3.0 eq.
KOAc, 0.25 mol % Pdꢀ168, 2.0 eq. BBA, 10 vol. MeOH, 4 hrs
A)
B)
1
4
ACS Paragon Plus Environment

Page 15 of 33
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
Significant base and Pd catalyzed protoꢀdeboronation of 2 was observed at 65 °C, which could
account for the varying levels of 3 observed. However, under the reaction conditions at 25 °C,
boronic acid 2 was found to be stable for >22 hrs, suggesting that elevated temperatures should
be avoided for this reaction. It was also observed that at 65 °C a large amount of black
particulates were formed that were very difficult to clean from the reactor walls, requiring
extensive mechanical scrubbing (Figure 4a). This material was thought to be a result of the
polymerization of BBA and not the precipitation of a Pd species, as it was not soluble in nitric
2
1
acid and its quantity appeared to be related to the amount of BBA used. It was then postulated
that lower temperatures may reduce the amount of the polymerization and hence black
particulates. Gratifyingly, complete conversions were achieved within 16 hrs at temperatures as
low as 35 °C (Table 1, Entries 1ꢀ3 & 5ꢀ6), and at these lower temperatures the levels of
impurities 3 and 4 were further reduced. Indeed the postulated polymerization was significantly
reduced, but not completely suppressed, as demonstrated by the presence of fewer black
particulates (Figure 4b). At 25 °C, the reaction rate was greatly reduced and 94% conversion was
obtained after 42 hrs, so that 35 °C appeared to be the lowest practical reaction temperature
(Table 1, entry 5).
1
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Effect of temperature and amount of BBA on the conversion and reaction impurity
profile ꢀ1.0 eq. 1, 3.0 eq. KOAc, 0.25 mol % Pdꢀ168, BBA, 10 vol. MeOH
1
LCAP
Entry
BBA Time/ T/ °C
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
eq.
1.0
1.5
1.5
1.5
2.0
2.0
2.0
3.0
6.0
hrs
2
3
4
1
1
2
3
4
5
6
7
8
9
4
35
35
45
65
25
35
65
65
65
49.1
94.0
91.9
85.3
90.1
94.0
92.8
93.3
90.5
1.3
2.4
3.6
6.1
2.5
3.6
4.3
5.8
8.4
1.6
2.5
3.3
6.9
0.8
1.3
2.5
0.7
0.6
47.7
0.3
ND
1.4
5.3
0.3
ND
ND
ND
16
16
16
42
4
4
4
4
1
.
LCAP = Liquid Chromatography Area Percent
In the absence of a suitable inꢀprocess analytical method for monitoring BBA during the reaction,
a series of experiments were conducted using a stoichiometric amount of the reagent. This
allowed for better discrimination of the key process parameters as well as understanding of its
11
stability. BBA has been shown by Molander et al. using B NMR to decompose almost
9
instantaneously in the presence of various Pd(II) species to boric acid. The kinetics of this
degradation were greatly reduced by preꢀforming the Pd(0) catalyst and using an excess amount
of ligand to ensure its efficient formation. Therefore it was postulated that the degradation was a
result of the Pd catalyst and that a semi batch process could improve conversions at lower BBA
stoichiometries. Accordingly 1.0 eq. of BBA was dissolved in 4 vol. of MeOH that was then
added to a solution of 1, KOAc and catalyst in 6 vol. of MeOH. This reaction only went to
1
6
ACS Paragon Plus Environment

Page 17 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
3
5.7 % conversion vs. 50.7 % in the case where MeOH was added to the solid reagents.
Moreover, when the BBA solution in MeOH was aged for 16 hrs prior to being added to the
remainder of the reagents, the conversion dropped to 27.4 %, highlighting its instability in the
reaction solvent. Alternative nonꢀprotic solvents were also considered and after an extensive
solubility screen, BBA was found to be only appreciably soluble in DMAc. The borylation was
unsuccessful in pure DMAc but found to proceed to 53.6 % conversion when a solution in 4 vol.
of DMAc was added to 10 vol. of MeOH. However, when the DMAc solution was added over 6
hrs, only a 33.2 % conversion was observed offering no advantage over using a pure MeOH
solvent 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
Figure 5. Tentative BBA decomposition pathway
Stability of BBA. Accelerating Rate Calorimetry (ARC) was performed on BBA in the solid
state in order to determine its thermal stability. The results indicated a catastrophic
21
decomposition reaction at 85 °C possibly as a result of rapid polymerization. The time to
maximum rate under adiabatic conditions (TMR ), which is a measure of the probability of
ad
occurrence of a decomposition reaction, was calculated to be 165 hr at 35 °C and only 5 hr at
6
5 °C further supporting the need for a lower reaction temperature. It was noted that upon
addition of BBA to MeOH effervescence was observed, presumably as a result of the evolution
of hydrogen gas which raised into question its stability in the solvent used (Figure 5). Boronic
acids are well known to be unstable even at room temperature, to transesterify with alcohols and
1
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
readily form dimeric and cyclic trimeric anhydrides, making their in process monitoring
2
2
particularly challenging. Extensive NMR studies were performed in an attempt to better
understand its stability (Figure 6 and Table 2). It was found that BBA significantly degraded to
boric acid after prolonged aging at room temperature under all conditions explored. In general
oxygen, water and Pdꢀ168 significantly accelerated its hydrolysis, while KOAc appeared to have
no impact. In the presence of Pdꢀ168, 63 % of BBA had degraded by the time the first NMR
sample was taken and it was fully hydrolyzed within just 4 hours (Entry 5, Table 2). The addition
of ethylene glycol has been shown to result in the in situ formation of a much more stable
boronate ester that can significantly accelerate reaction rates and allow for the amount of BBA
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
and catalyst used to be reduced. Indeed this greatly increased its stability so that in the presence
of water after 20 hours 13.8 % of BBA was hydrolyzed compared to 47.4 % in the absence of
ethylene glycol (Entries 4 & 3, Table 2). Similar trends were observed for air and Pdꢀ168.
Therefore it was suspected that BBA was hydroscopic and that it was the residual water that was
responsible for its hydrolysis when dissolved in MeOH (e.g. Table 2, entry 1). Karl Fischer
analysis resulted in 7.1 % water which corresponded to 3.0 % surface water once the water
generated from the transesterification occurring in MeOH was accounted for.
1
1
Figure 6. B NMR of BBA in MeOH under nitrogen after 21 hrs of aging
1
8
ACS Paragon Plus Environment

Page 19 of 33
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
BBA
B(OH)3
1
1
Table 2. Stability studies of BBA using B NMR.
Entry
Conditions Ethylene
Glycol
H O
KOAc Pdꢀ168 Time/
hr
Degradation
/ %
2
1
2
3
4
5
6
7
Inert
Inert
Inert
Inert
Inert
Inert
Air
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0
21
0
2.0
19.4
3.8
20
19.4
7.4
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0
20
47.4
3.8
0
20
13.8
63.0
99.3
21.9
88.0
54.1
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0
4
0
4
0
ꢀ
1
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
3
55.4
16.7
33.3
66.2
68.4
33.3
43.2
8
9
Air
Air
Air
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0
20
ꢀ
0
20
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0
20
Effect of ethylene glycol. As expected, ethylene glycol was found to have a profound effect on
both the reaction rate and amount of BBA and Pdꢀ168 required to achieve complete conversions.
Under the reaction conditions, BBA presumably exists as a mixture of species. Once ethylene
glycol is added the equilibrium will shift in favor of the more stable bidentate ethylene glycol
ester which is anticipated to be less Lewis acidic than the methyl ester or the boronic acid as a
result of the favorable overlap of the BꢀO σ bond with the vacant B p orbital. Indeed, adding just
1
eq. of ethylene glycol with respect to the BBA used, increased the overall conversion from
48.4 % to 85.6 %, while 5 eq. resulted in >99 % conversion after just 6 hrs of reaction.
Use of ethylene glycol therefore allowed for the amount of BBA required to achieve the target
conversion specification of NLT 99.0 % to be reduced from 2.0 eq. to 1.15 eq., greatly
improving the efficiency of the process (Figures 7A & 11). In its absence conversion dropped
dramatically when less than 1.75 eq. of BBA were used to afford <95.9 % conversion. If stalled,
due to for example poor BBA quality, the reaction was found to be readily recovered by simply
adding more BBA. Therefore, 1.25 eq. of BBA was selected to afford a reasonable safety margin
from the determined edge of failure. The catalyst loading could also be reduced 5ꢀfold (from
0
.25 mol % to 0.05 mol %) with no impact on the observed impurity profile, reaction rate and
overall conversion (Figure 7B; 85.7 % vs. 85.0 % respectively), which was otherwise impossible
to achieve.
2
0
ACS Paragon Plus Environment

Page 21 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Figure 7. Effect of ethylene glycol on: A) the BBA stoichoimetry ꢀ1.0 eq. 1, 3.0 eq. KOAc, BBA,
15.2 eq. ethylene glycol (if applicable), 0.05 & 0.25 mol % Pdꢀ168, 10 vol. MeOH, 35 °C, 4 hrs,
200 RPM; B) the catalyst loading, ꢀ1.0 eq. 1, 3.0 eq. KOAc, 1.0 equiv. BBA, 15.2 eq. ethylene
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
glycol (if applicable), Pdꢀ168, 10 vol. MeOH, 35 °C, 4 hrs, 200 RPM
A)
B)
Effect of oxygen. The process in the absence of ethylene glycol, using 2.0 eq. of BBA and 0.25
mol % Pdꢀ168 catalyst was not found to be particularly sensitive to oxygen, tolerating up to 5 %
without an appreciable impact on the overall conversion or reaction impurity profile (Figure 8A).
2
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Above 5 %, the conversion dropped significantly and when run under air (20.9 % O ) the
2
reaction stalled at 63.0 % conversion. On the other hand, when ethylene glycol was used, that
allowed for the amount of BBA and catalyst used to be reduced significantly, the process was
found to be greatly dependent on the reactor headspace oxygen levels and a significant drop in
the overall conversion was observed at levels above 0.5 % (Figure 8B). For example, under an
atmosphere of 2.0 % O , the conversion dropped to 91.4 %, while under 5.0 % O the reaction
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
stalled at just 74.6 % conversion. Therefore the amount of O in the headspace was determined to
2
be a key process parameter, and a specification of NMT 0.5 % O was set. Interestingly, oxygen
2
did not appear to impact the levels of deꢀbromination and homoꢀcoupling, which were 1.2ꢀ3.0 %
and <0.5 % respectively for all conditions investigated.
Figure 8. Effect of oxygen on the arylboronic acid 2conversion: A) under the optimal process
conditions without using ethylene glycol ꢀ1.0 eq. 1, 3.0 eq. KOAc, 2.0 eq. BBA, 0.25 mol % Pdꢀ
168, 10 vol. MeOH, 35 °C, 4 hrs, 200 RPM; B) under the optimal process conditions using
ethylene glycol ꢀ1.0 eq. 1, 2.5 eq. KOAc, 1.25 eq. BBA, 15.2 eq. ethylene glycol, 0.05 mol %
Pdꢀ168, 10 vol. MeOH, 35 °C, 4 hrs, 200 RPM
A)
B)
2
2
ACS Paragon Plus Environment

Page 23 of 33
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
Effect of agitation speed. Surprisingly the reaction agitation speed only had an impact on
conversion when the process was run under a headspace containing some amount of O . For
2
example when run under an atmosphere of 0.5 % O , the overall conversion dropped from
2
98.0 % to 91.8 % as the agitation speed increased from 250 to 600 RPM (Figure 9). Addition of
supplemental BBA alone was sufficient to push the reaction to completion, suggesting that the
catalyst was active and that the reaction performance was compromised as a result of the
degradation of the BBA. This mixing dependency suggested that the observed degradation was
O ꢀmediated, as at higher agitation speeds the mass transfer of the O gas to the liquid is higher
2
2
leading to faster dissolution of the gas into the liquid layer and therefore faster degradation.
Figure 9. Effect of agitation speed on the arylboronic acid 2conversion at different O headspace
2
levels ꢀ1.0 eq. 1, 2.5 eq. KOAc, 1.25 eq. BBA, 15.2 eq. ethylene glycol, 0.05 mol % Pdꢀ168, 10
vol. MeOH, 35 °C, 4 hrs
2
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scale-up. The previously described labꢀscale experimentation demonstrated how the dissolved
oxygen (DO) level is expected to be a key process parameter upon scaleꢀup. The speed and
efficiency with which a target DO level can be obtained is a factor of both the headspace oxygen
content, which provides the driving force for gas exchange, and the gasꢀliquid massꢀtransfer
coefficient (k a) which determines the rate at which gas is able to move across the gasꢀliquid
L
phase boundary. The choice of the inertion technique and operating parameters are crucial in
2
3
obtaining an acceptable DO level. The two most common approaches for the inertion of a
reactor at any scale are either a flowꢀthrough of inert gas to dilute the headspace or the use of
variableꢀpressure cycles to exchange the headspace. Both are typically performed while the
liquid mass is being agitated to maximize the gasꢀliquid massꢀtransfer rate. The first, headspace
2
4
ACS Paragon Plus Environment

Page 25 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
flowꢀthrough method, can take many hours on scale as the inert gas flow rates are typically low
compared to the large headspace volumes. Variableꢀpressure cycles are typically quicker and use
less nitrogen to achieve a target headspace oxygen concentration, however when vacuum is used
there is the risk of losing solvent via evaporation as well as pulling in air through the inevitable
leaks that are present in a typical plant reactor. Pressurization with inert gas followed by release
to atmosphere avoids pulling air into the reactor but is typically not possible when lowꢀpressure
and/or glass components are used as part of the reactor. On the basis of this, agitated variableꢀ
pressure vacuum and release cycles were selected as the inertion technique. The vacuum setꢀ
point was 200 mbar, to minimize solvent loss (the vapor pressure of methanol at 20 °C is 130
mbar) and the nitrogen gas flow rate was set to 400 L/min in order to vigorously mix the
headspace and ensure homogeneity and efficient dilution. To ensure the effectiveness of this
proposed inertion sequence an atꢀscale solvent pilot run was performed. For this the plant reactor
was flushed with nitrogen before 1,800 L of methanol was charged. After commencing agitation
at the intended setꢀpoint of 54 rpm, vacuum was applied and the headspace pressure reduced to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
200 mbar i.e. above the boiling point. After a 5 minute hold the reactor was backfilled with
nitrogen at a flowrate of 400 L/min and held once more for 5 minutes. This cycle was repeated a
further four times and ultimately yielded a dissolved oxygen content of 0.85 ppm and a
headspace oxygen level of 0.2 %. This was deemed to be sufficiently below the 0.5 %
requirement where at a labꢀscale the reaction had been demonstrated to proceed to full
conversion without stalling. Further to this, additional inertion runs were performed which
allowed the reactor k a coefficient to be calculated.
L
Figure 10 shows a typical data set of dissolved oxygen collected across a total of five vacuum
and nitrogen release cycles. The behavior of the probe as well as its response time when
2
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
experiencing variable pressure is not well characterized and must be considered when analyzing
the data in order to extract, for example, the k a. However it can clearly be seen that each
L
successive vacuumꢀrelease cycle reduces the dissolved oxygen level, and the extent with which it
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
drops diminishes with each successive cycle as the headspace O level decreases and the driving
2
force is reduced.
It is also worth noting that once BBA is charged to the reaction the dissolved oxygen sensor fails
to return a value. This is suggesting that the presence of the boron species either depolarizes the
sensor or otherwise interferes with the electrochemical reaction resulting in a failure to produce a
reading. Due to this effect the amphoteric probe cannot be used to monitor dissolved oxygen
levels over the course of the reaction and is solely used to confirm the inertion of the reaction
solvent prior to the reagent addition. The gaseous O sensor installed in the headspace continues
2
to read throughout the reaction and its value is used to provide assurance that the blanket of inert
gas remains within the desired specifications throughout the reaction.
Figure 10. Typical dissolved oxygen data collected across a total of five vacuum and nitrogen
release cycles in a 2,800 L reactor charged with 1,800 L of MeOH and agitated at 54 rpm
2
6
ACS Paragon Plus Environment

Page 27 of 33
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
The validated inertion protocol was successfully applied on pilot scale at 65 kg and achieved,
after five cycles a headspace oxygen level of 0.2 % and a dissolved oxygen level of 0.18 ppm.
The borylation reaction met all specifications with a >99.9 % conversion to boronic acid 2 and a
yield of 91.4 % with an HPLC purity of 98.8 %. The levels of deꢀbrominated and homoꢀcoupled
impurities were within the typical ranges at 1.4 % and 0.4 %, respectively.
rd
Cost analysis. The 3 generation ethylene glycolꢀBBA process was shown to afford significant
cost savings over the previous synthetic routes (Figure 11). The data that showed this reduction
was obtained directly from the contract manufacturer and included separate line items for
individual raw material costs and total processing costs. While the exact factors and each one’s
significance in reducing the processing cost are unknown (being proprietary CMO costing
2
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
information), a number of improvements made to the process are likely large contributors.
Beyond the obvious reduction in BBA stoichiometry and Pd loading, the ethylene glycol process
was significantly streamlined and reduced the number of unit operations compared to the
previous iterations (Figures 2 & 12). The 2ꢀbutanol distillation was removed, as was the polish
filtration previously needed to remove inorganic residues. The synthesis directly afforded the
boronic acid and bypassed the need to hydrolyze it. In making these changes the number of
vessels required dropped from 2 to 1 and the overall processing time from 9.25 days to 2.5. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
st
nd
yield also improved from 60 % in the 1 generation B Pin process to 80 % in the 2 generation
2
2
rd
BBA process and finally to 90 % with the 3 generation ethylene glycolꢀBBA process – this
achieves an overall increase in efficiency and throughput. When considering solely the
nd
processing costs per mole the 2 generation BBA process achieved a 59 % reduction compared
st
rd
to the 1 generation B Pin process while the 3 generation ethylene glycolꢀBBA process
2
2
nd
achieved a 63 % reduction. When the savings in raw material costs are also included, the 2
generation BBA process was found to offer a total 35 % reduction in cost per mole as compared
st
rd
to the 1 generation B Pin process while the 3 generation process offered a total 47 % cost
2
2
st
reduction compared to the 1 generation B Pin process.
2
2
st
Figure 11. Costs analysis of the three borylation processes: 1 generation, being the original
nd
rd
B Pin ; 2 generation, being the original BBA mediated process; 3 generation, being the
2
2
modified with ethylene glycol BBA process
2
8
ACS Paragon Plus Environment

Page 29 of 33
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
CONCLUSION
A Miyaura borylation process utilizing BBA was developed to afford a key fragment of TAKꢀ
17, an API currently being developed by Takeda Pharmaceuticals as an oncolytic agent. The
1
11
stability of BBA was evaluated under the reaction conditions using a combination of B NMR,
calorimetry and targeted experiments using a stoichiometric amount of BBA, and found to be
greatly improved by the addition of ethylene glycol. Data was presented to suggest an O2ꢀ
mediated degradation of BBA that was determined to be the root cause for the observed
inconsistent reaction conversions. Methodology was developed for both the study and scaleꢀup
of O sensitive processes that enabled the tight control of both the dissolved and headspace O2
2
concentrations, allowing for the identification of suitable O specifications. Finally, the revised
2
process was demonstrated to be significantly more efficient than the original process that utilized
B Pin , offering a cost saving of 47 %.
2
2
Figure 12. Revised BBAꢀmediated Miyaura borylation of aryl bromide 1
2
9
ACS Paragon Plus Environment