Preparative synthesis via continuous flow of 4,4,5,5-tetramethyl-2-(3- trimethylsilyl-2-propynyl)-1,3,2-dioxaborolane: A general propargylation reagent

Article abstract of DOI:10.1021/op200180t

A scalable process for the preparation of 4,4,5,5-tetramethyl-2-(3- trimethylsilyl-2-propynyl)-1,3,2-dioxaborolane from trimethylsilylpropyne, isopropyl pinacol borate, and n-butyllithium is described. Problems associated with implementing a typical aqueous workup and batch process into production due to borolane ate equilibration and protonolysis are presented. To address these issues, a continuous-flow and distillation process was developed which efficiently produced 297 kg of the key propargylation reagent.

Full text of DOI:10.1021/op200180t

COMMUNICATION
pubs.acs.org/OPRD
Preparative Synthesis via Continuous Flow of 4,4,5,5-Tetramethyl-
-(3-trimethylsilyl-2-propynyl)-1,3,2-dioxaborolane: A General
Propargylation Reagent
2
,
†
†
‡
‡
‡
Daniel R. Fandrick,* Frank Roschangar, Chunyoung Kim, Byoung J. Hahm, Myoung H. Cha,
‡
‡
‡
†
†
†
†
Hyoun Y. Kim, Gyesang Yoo, Taeyun Kim, Jonathan T. Reeves, Jinhua J. Song, Zhulin Tan, Bo Qu,
Nizar Haddad, Sherry Shen, Nelu Grinberg, Heewon Lee, Nathan Yee, and Chris H. Senanayake
†
†
†
†
†
†
†
Chemical Development, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road/PO Box 368, Ridgefield, Connecticut,
6877-0368, United States
0
‡
Department of Process R&D Lab, SK Holdings, 140-1 Wonchon-dong, Yuseong-gu, Daejeon 305-712, Korea
S Supporting Information
b
necessary propargyl bromide required either the TMS protec-
tion of a propargyl bromide or a sequential TMS protection
1
1
ABSTRACT: A scalable process for the preparation of
,4,5,5-tetramethyl-2-(3-trimethylsilyl-2-propynyl)-1,3,2-
4
and substitution of a propargyl alcohol to the bromide inter-
12
dioxaborolane from trimethylsilylpropyne, isopropyl pina-
col borate, and n-butyllithium is described. Problems asso-
ciated with implementing a typical aqueous workup and
batch process into production due to borolane “ate” equili-
bration and protonolysis are presented. To address these
issues, a continuous-flow and distillation process was devel-
oped which efficiently produced 297 kg of the key propar-
gylation reagent.
mediate. Alternatively, a Matteson homologation approach has
also been pursued for the preparation of propargyl boronates
13
through the iodomethyl pinacol borate intermediate. Direct utiliza-
tion of trimethylsilylpropyne for the substitution with a borate
circumvents the multistep synthesis and purification of a halide
intermediate, thereby affording a more cost-effective approach
towards the valuable reagent. Herein, we report the multikilogram
synthesis of the propargyl borolane 1 through the use of a
continuous-flow process from trimethylsilylpropyne and isopropyl
pinacol borate.
’
INTRODUCTION
Organoboronic acids and esters are valuable synthetic building
’ RESULTS AND DISCUSSION
1
blocks. Recently, propargyl borolane 1, 4,4,5,5-tetramethyl-
Initial experiments toward borolane 1 were based on the
site selective functionalization of trialkylsilylpropynes through
the lithiated species 14 by Corey and Kirst and related
addition of lithiated trialkylsilylpropynes to trialkylboranes.
2
2
-(3-trimethylsilyl-2-propynyl)-1,3,2-dioxaborolane, demon-
strated significant utility as a versatile reagent for the site-selective
propargylations of carbonyl and imine species (Scheme 1). For
example, the zinc-catalyzed propargylation of aldehydes and
ketones with borolane 1 provided an operationally simple
14
15
NMR and IR characterization of the lithiated propyne 14
by Reich et al. showed a rapid equilibration between the
allenyl and propargyl lithium species wherein the allenyl
3
process for the synthesis of homopropargylic alcohols. Addi-
tionally, chiral homopropargyl amines were readily accessed
through the highly diastereoselective zinc-catalyzed propargyla-
16
derivate 14b was favored. Due to the trimethylsilyl substi-
tuent, reactions with this organolithium mixture typically
strongly favor the acetylenic product.
4
tion of tert-butanesulfinimines. Through the proper choice of
14,15
Direct addition of
ligands, the copper-catalyzed asymmetric propargylation of
the borate 11 to an in situ generated lithiated propyne 14
afforded the borolane 1 in <30% yield. Reversing the addition
by charging a solution of lithiated propyne 14 to a solution
of borate 11 at À25 °C furnished a reasonable yield of borolane
5
6
aldehydes and ketones was also demonstrated with remarkable
functional group tolerance and scope. Alternatively, conversion
of the propargyl boronate to the diethanolamine derivative
7
provided facile access to allenyl halides. Synthesis of organobor-
1
after an aqueous workup (Figure 1). Employing an azeotropic
onate esters may involve the addition of a Grignard reagent to a
trialkoxyborate wherein formation of the “ate” intermediate
prevents polyalkylation, thereby affording reasonable to high
distillation to dry the wet organic layer, typical for production
processes, with heptane resulted in significant decomposi-
tion and a low yield. Borolane 1 showed reasonable thermal
8
yields for the respective organoborolane. Hoffmann and co-
17
stability but demonstrated susceptibility to protonolysis
workers employed such a strategy for the synthesis of propargyl
under control experiments with protic solvents such as isopro-
panol, that provides a rationalization for the limited stability
borolane 1 with an allenyl Grignard reagent generated from a
9
propargyl bromide (Scheme 2). Although the process provided
a gram-scale synthesis of the key borolane, the use of the shock-
Special Issue: Continuous Processing 2012
10
sensitive propargyl bromide and exothermic, dissolving-metal
reduction presents safety and engineering issues that limit the
rapid implementation into a pilot plant. Furthermore, the
Received: July 3, 2011
Published: December 16, 2011
r 2011 American Chemical Society
1131
dx.doi.org/10.1021/op200180t
|Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Scheme 1. Applications of propargyl borolane 1
Scheme 2. Synthetic approaches for the preparation of propargyl borolane 1
1
9
during azeotropic drying (eq 1).
nonaqueous workup was pursued. The process required the
formal removal of lithium isopropoxide from the borolane “ate”
intermediate 15 (Figure 2). To facilitate the workup, the
byproducts from the quench should be readily removed by
filtration or have a relatively low boiling point to allow removal
during the solvent distillation. Accordingly, quenching the reac-
20
tion with acetic acid, acetyl chloride, methyl iodide and
21
Although laboratory batches can be readily prepared through
workup with a solid desiccant, the use of this technique in
production dramatically increases operational costs due to the
trimethylsilyl chloride were examined (Table 1). Although an
acetic acid quench afforded a reasonable yield on a 10-g scale, the
yield was significantly affected by the batch size. Alternatively,
quenching the process with acetyl chloride provided a
18
time-consuming nature of the operation. Accordingly, a
1
132
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Figure 1. Initial Process for the Preparation of Borolane 1.
Figure 2. Design of a non-aqueous workup for borolane 1.
a
Table 1. Screen of quenching reagents for the preparation of borolane 1
b
c
d
entry
quench reagent
borate 11 (g)
product 1 (g)
quality (wt %)
yield (%)
1
2
3
4
5
6
7
AcOH
AcOH
AcCl
10
50
10
50
55
10
46
10
39
13
64
77
À
80
33
80
72
63
À
64
20
78
71
69
AcCl
AcCl
e
3
CH I
<20
e
TMSCl
À
À
<40
a
b
c
1
1
.05 equiv n-BuLi and 1.05 equiv 12 to borate 11. 1.05 equiv to borate 11. Weight percent determined by H NMR assay with dimethyl fumarate as
an internal standard in C D . Isolated yield. Assay yield before workup by H NMR with dimethyl fumarate as an internal standard in C D
6 6 6 6.
d
e
1
consistent high yield for the borolane 1. The workup using
acetyl chloride simply required a solvent switch to heptane,
filtration to remove lithium chloride, and concentration to remove
the excess solvent and the isopropyl acetate byproduct. Further-
more, the batch after the quench with acetyl chloride can be held at
ambient temperature for up to 2 days with no noticeable impact on
the yield.
Although a scalable workup for the preparation of borolane 1
was established, the yield was significantly impacted by the
batch size. Control experiments revealed a significant effect of
the addition time for the charge of the in situ generated
lithiated propyne 14 to the borate required to maintain the
batch temperature between À15 to À25 °C (Table 2). While a
rapid addition in 15 min provided a 78% yield, increasing the
addition to an hour dramatically lowered the yield to <50%.
Under identical conditions with 150 g batches wherein
the addition time was increased from 0.75 to 1.0 h a reduction
in the yield from 57 to 48% (entries 3 and 4) was demonstrated.
Cooling capacity of pilot-plant reactors vary considerably
and slightly exothermic additions in typical production facilities
22
1
133
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
a
Table 2. Effect of addition time on the preparation of borolane 1
a
b
entry
borate 11 (g)
addition time (h)
assay yield 1 (%)
1
2
3
4
5
10
50
0.25
0.42
0.75
1
78
71
57
48
47
150
150
182
1.25
a
b
Duration for the addition of the lithiated trimethylsilylpropyne 14 to the borate THF solution. Assay yield after acetyl chloride quench determined by
H NMR in C D with dimethyl fumarate as an internal standard.
6 6
1
at À15 to À25 °C can require several hours. Accordingly, only
low to moderate yields are expected for the current process in
production.
decomposition of the desired product 1 to further contribute
to the reduced yield with prolonged additions. The borolane
“ate” intermediates are typically stabilized with lower reaction
1
9,21,23
A reasonable rationalization for the effect of the addition time
on the yield for the charge of lithiated trimethylsilylpropyne to
isopropyl pinacol borate relates to the limited stabilities of the
borolane “ate” intermediates and associated equilibration. After
the organometallic addition to the borate, a minimal effect on the
yield was observed when the batch was held at À25 °C for up to
temperatures,
but only a marginal improvement in yield
to 65% from 57% was obtained by conducting the addition
at À60 °C with a 45-min addition (eq 2). Due to the variable
yields with different addition times and the decreased effi-
ciency to remove heat from a batch with increasing batch
1
8
size, the batch process for the preparation of propargyl
borolane 1 is unlikely to afford reproducible and reasonable
yields upon scale.
2
h. Therefore, the addition time effect must affect the quality of
the lithiated propyne or reaction mixture during partial conver-
sion due to increasing the time with which each is held at À25 °C
throughout the addition. Although the lithiated propyne has
limited stability at À5 °C, a nearly quantitative yield for the
addition to p-anisaldehyde was obtained after holding the
lithiated reagent for 2 h at the reaction temperature of À25 °C
(Figure 3). Accordingly, the reduced yield due to the increased
addition time is not due to decomposition of the lithiated
propyne. Alternatively, organoborone “ate” intermediates
derived from the addition with an organolithium reagent are
known to dissociate a lithium alkoxide ligand and can lead to
The formation of organoborolanes typically involves the
addition of an organolithium or Grignard reagent to a
trialkoxyborate. Lithium and magnesium cations are expected
1
9,23
8
polyalkylation.
Accordingly, the prolonged addition time
increases the time when the borolane “ate” intermediate 15 is
exposed to the starting borate which due to the dissociation
can reasonably enable a borolane “ate” equilibration and
generation of the bis-isopropoxyborolane “ate” byproduct 17
to excerpt different complexations and stabilities between the
organotrialkoxyborate and alkoxide intermediates encountered
19
in the batch process. Accordingly, the equilibrium between the
intermediate organotrialkoxyborate and organoborolane inter-
mediates encountered during a batch process should be depen-
dent on the countercations utilized in the process. In an effort to
shift the equilibrium towards the “ate” intermediate for a batch
(
Scheme 3). The dissociation and equilibration then estab-
lishes a competition for addition to the starting borate 11 and
product propargyl borolane 1 upon additional charge of the
lithiated propyne to further reduce the yield for the desired
product by formation of the byproduct 16. Complicating the
mechanistic analysis is the observation that the starting borate
19,23
process (Scheme 3),
the additive magnesium chloride was
included to the isopropyl pinacol borate solution prior to the
addition of the in situ generated lithiated propyne 14aÀb. The
additive magnesium chloride, albeit poorly soluble in the reaction
system, appeared to significantly reduce the yield dependence on
the addition time of the lithiate propyne to the borate (Table 3).
The yield minimally decreased from 78% to 72% upon increasing
1
1 as a solution in THF oligomerizes into a gel upon addition
2
4
of lithium isopropoxide at the reaction temperature. This
oligomerization prevents direct observation of the borolane
“
ate” intermediates and byproducts and may also promote the
1
134
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Figure 3. Temperature stability of lithium trimethylsilylpropyne 14a. HPLC (220 nm) assay yield of 4 and 4b with toluene as an internal standard (error
5%). 4À5:1 site selectivity between 4 and 4b.
(
Scheme 3. Proposed equilibration and byproduct formation
the addition time from 0.5 to 2 h. However, when the modified
batch process with magnesium chloride was demonstrated on
a kilogram scale, the yield decreased to 64% with a 30-min
addition in comparison to a 78% yield for a 30 min addition on
a 182-g scale (entries 1 and 2). In addition to this drop in yield upon
scale-up, the process also presented operational challenges for
production. The magnesium chloride additive increased the mass
balance for the process and was observed to significantly increase the
filtration time for removal of the salts during the workup. These
complications and reduced yield upon scale-up presented reasonable
concerns to implement the modified batch process with magnesium
chloride on a pilot-plant scale.
Continuous-flow processes have gained popularity in chemical
production due to the ability to mitigate safety hazards, to
1
135
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
a
Table 3. Effect of addition time and additive on the preparation of borolane 1
a
b
entry
borate 11 (g)
addition time (h)
assay yield 1 (%)
1
2
3
4
5
182
1500
68
0.5
0.5
1.0
1.5
78
64
73
74
72
68
68
2.0
a
b
Duration for the addition of the lithiated trimethylsilylpropyne 14 to the borate THF solution. Assay yield after acetyl chloride quench determined by
H NMR in C D with dimethyl fumarate as an internal standard.
6 6
1
Figure 4. Continuous-flow process for the preparation of borolane 1.
prepare unstable compounds or intermediates, localize physical
the reaction followed by the quench with acetyl chloride and
deposit into a collection vessel.
effects, and achieve maximum productivity by achieving optimal
18
reaction conditions. Preparation of the propargyl borolane 1
through a continuous-flow process would eliminate the addition
time effect observed in a batch process by employing the
simultaneous addition of the lithiated propyne 14 and starting
borate 11. This process was designed in three phases
(
1
Figure 4). The first phase generates the lithiated propyne
4 by a continuous stream of a THF solution of the starting
Optimization of the continuous-flow process focused on
reagent stoichiometry, delay loop retention times and respective
temperatures. The baseline process utilized a À5 °C bath with
1.13 equiv of n-BuLi, 1.07 equiv of starting propyne 12, 1.13
equiv of acetyl chloride and a retention time of 61 s for the
lithiation and 21 s for the coupling. In contrast to a batch process
wherein a slight excess of the lithiated propyne was utilized to
trimethylsilylpropyne 12 with n-butyllithium. A delay loop
was installed after the mixture to allow for adequate time for
complete lithiation and minimize the formation of the butyl
borolane byproduct 18 (eq 3). The second phase conducts the
coupling with another delay loop to allow sufficient time for
1
136
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Table 4. Stoichiometry optimization for the continuous-flow
process
compensate for the decomposition pathways, the optimal stoi-
chiometry for a continuous-flow process was the desired 1 equiv
of lithiated propyne and 1 equiv of acetyl chloride to starting
borate 11 (Table 4). The reaction temperatures and retention
times for the lithiation and coupling were subsequently opti-
mized (Table 5). Complete lithiation of trimethylsilylpropyne as
indicated by <0.5 A% butyl adduct formation was achieved with a
lithiation temperature of 5 °C and a retention time of 40 s. The
optimal conditions for the coupling utilized a 21 s delay loop at
À5 °C before the acetyl chloride quench at the same tempera-
ture. After the quench, the reaction mixture can be held in the
collection vessel for up to 24 h at À5 °C. The lab-scale
continuous-flow process on a 50À100-g scale afforded a con-
sistent 64À70% yield for the propargyl borolane 1 before
distillation and after workup to remove the salts and solvents.
After preparation of the borolane 1 through a continuous-
flow process, the crude material was purified by distillation. The
crude material from the continuous-flow reaction was mixed
with heptane to precipitate the salts (LiCl), filtered, and the
filtrate concentrated to remove the majority of the solvents. For
smaller batches (<10 kg), a batch distillation was employed.
The filtrate from the filtration was distilled through a short path
distillation apparatus to remove the heavy impurities and
distillation through a fractional distillation column provided
the borolane 1 in >95 A% (GC-FID) and 88 wt % with an
a
b
c
equiv
GC IPC (A%)
entry
n-BuLi
AcCl
11
18
1
1
2
3
4
5
6
7
1.13
1.1
1.13
1.1
5.9
5
6.3
4.3
4.7
1.7
2.1
1
73.4
74.6
74.8
80
1.1
1.03
1.1
4.2
5.4
6
1.03
1.03
1
1.03
1
80.2
82.1
81.9
6.1
8.6
0.93
0.93
0
a
Process conducted in Figure 4 with a À5 °C lithiation, coupling, and
quenching temperature and 1.07 equiv of propyne 12 to borate 11.
Delay loop 1 and 2 retention time of 61 and 21 s, respectively.
b
c
Equivalents to starting borate 11. GC relative area by FID.
Table 5. Temperature and retention time optimization for
a
the continuous-flow process
d
temperature (°C)
delay loop (s)
GC IPC (A%)
b
c
b
c
entry
1
2
1
2
11
18
1
1
2
3
4
5
6
7
8
9
0
0
45
45
45
45
45
62
53
62
40
40
21
21
21
21
21
0
9.3
8.4
1.7
3.1
1.8
1.4
0.5
1
80.6
79.8
78.8
77.2
84.7
77.4
77.5
82.1
84.7
83.8
8
2% distillation recovery (Figure 5). For production batches
À5
À2
0
À5
À5
À5
À5
À5
À5
À5
À5
À5
(
>10 kg), pot distillation was not practical and exposes the
10.1
12.5
6.5
product to prolonged heating due to the time required to distill
the large volumes. Accordingly, a continuous distillation pro-
cess was developed with a series of thin film distillation
apparatuses to provide the propargyl borolane with reproduci-
ble quality (Figure 6). Due to the large capacity of the
continuous distillation equipment, hold up loss was inevitable
with larger losses with smaller batches. Distillation of the largest
batch (∼200 kg) afforded a 70% recovery, and larger batches
are expected to yield better recoveries.
5
À5
À5
À5
5
12.8
9.7
5
1.1
1
21
21
31
6.1
c
6.5
7.1
0.5
1
10
5
a
Process conducted in Figure 4 with 1.07 equiv of propyne 12 and 1
equiv of n-BuLi and AcCl to starting borate 11. Temperature and
retention time for the lithiation. Temperature and retention time for
the coupling. GC relative area by FID.
b
The continuous-flow process for the preparation of borolane 1
reproduced well from lab scale to production to afford an average
yield of 64À75% after workup to remove the salts and most of
c
d
Figure 5. Lab-scale batch distillation of borolane 1.
1
137
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Figure 6. Continuous distillation of propargyl borolane 1 through a series of thin film distillation apparatuses.
Table 6. Continuous-flow batches for propargyl borolane 1
a
b
c
GC-IPC by FID (A%). Crude product after concentration, heptane chase, and filtration. Percent 11, 18 and 1 by GC-FID (A%). Weight percent
1
d
e
6 6
assay determined by H NMR in C D with dimethyl fumarate as an internal standard. Distillation recovery yield. Due to large capacity of distilla-
f
tion equipment, hold up loss was inevitable with larger losses with smaller batches. Overall yield for reaction and distillation. ND = not determined.
SPD = short path distillation.
1
138
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
Figure 7. Continuous-flow system for lab-scale batches.
the solvents (Table 6). Due to the hold-up loss in the continuous
distillation process and relatively small batches in relation to the
equipment design, a moderate distillation recovery was obtained
in the production batches that were ∼100 kg. Overall, the
continuous-flow process and distillation produced 318 kg of
the propargyl borolane 1 in >91 wt % purity (297 kg at 100 wt %).
in MTBE (72 mL) was charged to the reaction at a rate to
maintain Tint = À20 to À15 °C. The quenched reaction was aged
for 1 h at Tint = À20 to À15 °C, then warmed to ambient
temperature over 0.5 to 1 h. The batch was concentrated in vacuo
to remove 1.82 L of solvent. The resulting mixture was charged
then distilled with MTBE (1 Â 1.0 L, 1 Â 700 mL) followed by
heptane (1 Â 1.0 L, 1 Â 350 mL). The solids were removed by
filtration and rinsed with heptane (2 Â 100 mL). The filtrate was
concentrated in vacuo to provide the crude borolane 1 as a
’
CONCLUSIONS
In conclusion, an efficient process for the production of
homogeneous orange oil (225.1 g). 1H NMR in C D with
6 6
4
,4,5,5-tetramethyl-2-(3-trimethylsilyl-2-propynyl)-1,3,2-dioxa-
dimethyl fumarate as an internal standard showed 80.8 wt % for a
reaction yield of 78%.
borolane from trimethylsilylpropyne, n-butyllithium, and isopro-
pyl pinacol borate was developed. The development of a
nonaqueous continuous-flow process circumvented the issues
with borolane “ate” intermediate equilibration and protonolysis
associated with a batch process, and typical aqueous workup
afforded over 300 kg of the propargylation reagent. Due to the
synthetic value of organoboronic esters in process chemistry,
continuous-flow technologies that provide a robust process for
the production of these valuable yet often relatively unstable
reagents are of significant utility.
The crude borolane 1 (747.6 g, 53.8 wt %) was distilled
through a fractional distillation system with ∼3À5 theoretical
plates at ∼2 mbar. Collection of the fractions with Tvap
=
68À79 °C with Tbath = 108À135 °C provided borolane 1 as a
yellow oil (260 g) in 92.8 wt % with a 60% distillation recovery.
1
H NMR (400 MHz, C
D
) 1.92 (2H, bs), 0.971 (s, 12 H), 0.189
6
6
13
(s, 9 H). C NMR (100 MHz, C
24.70, 3.7 (b), 0.46. EI-HRMS calcd for [M] C12
D
6 6
) 104.16, 83.80, 83.17,
+
H
H
23BO
20BO
Si
Si
2
2
+
238.1560, found 238.1577; calcd for [M À CH
23.1326, found 223.1331.
Continuous-Flow Process. The four feed solutions were
] C11
3
2
’
EXPERIMENTAL SECTION
prepared in dried vessels under argon. In the first vessel, a solution
of 1-trimethylsilyl-1-propyne (56.5 g, 0.503 mol) was prepared in
THF (<500 ppm water, 313 g). The second vessel was charged
with n-BuLi (2.5 M in hexanes, 134.4 g, 194 mL, 0.485 mol). The
third vessel was charged with THF (<500 ppm water, 284.4 g)
followed by 2-isopropoxy-4,4,5,5 tetramethyl-1,3,2-dioxaborolane
(87.6 g, 0.471 mol). The fourth vessel was charged with heptane
(<100 ppm water, 24.6 g) and acetyl chloride (38.0 g, 0.484 mol).
A continuous-flow system was constructed as delineated in
Figure 7. The indicated mixers and delay loops were immersed
into two separate chillers adjusted to the respective temperatures.
The four feed streams were charged into the specified mixtures
with nitrogen pressure at the respective flow rates metered by
mass flow controllers. The final mixed streams were collected
into a vessel under nitrogen with agitation at Tint NHT À15 °C.
After completion of the continuous-flow process, the reaction
solution was agitated at Tint NHT À15 °C for 1 h. The solution
Batch Process with MgCl . Anhydrous THF (<500 ppm
water, 733 mL) followed by 1-trimethylsilyl-1-propyne
154 mL, 117 g, 1.05 mol) were charged to a dried reactor under
argon. The batch was cooled to Tint = À25 °C. n-BuLi (2.5 M in
hexanes, 404 mL, 1.01 mol) was charged to the trimethylsilyl-
propyne solution at a rate to maintain Tint = À20 to À25 °C. The
reaction was aged for 1 h at Tint = À25 °C to afford the lithiated
propyne solution as a homogeneous yellow to orange solution.
2
(
While maintaining the lithiated propyne solution at Tint
=
À25 °C, the lithiated propyne solution was charged to an
anhydrous mixture of 2-isopropoxy-4,4,5,5-tetramethyl-1,3,2-di-
oxaborolane (200 mL, 182 g, 0.980 mol) and anhydrous magne-
sium chloride (93.3 g, 0.980 mol) in THF (<500 ppm water,
1
.0 L) at a rate to maintain the borolane solution at Tint = À25
to À15 °C. The reaction was aged for 1 h at T = À25 °C. An
int
anhydrous solution of acetyl chloride (71.8 mL, 79.3 g, 1.01 mol)
1
139
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140

Organic Process Research & Development
COMMUNICATION
was warmed to ambient temperature and can be held at this point
for up to 18 h.
(11) (a) Orsini, F.; Pelizzoni, F. Synth. Commun. 1984, 14, 805–816.
(b) Orsini, F. Synthesis 1985, 500–503. (c) Verkruijsse, H. D.; Brandsma,
L. Synth. Commun. 1990, 20, 3375–3378. (d) Trost, B. M.; Huang, X.
Chem.—Asian J. 2006, 1, 469–478. (e)Chen, D.-D.; Hou, X.-L.; Dai, L.-X.
Tetrahedron Lett. 2009, 50, 6944–6946.
The reaction was concentrated to ∼450 mL by distilling
4
75 mL of solvent to afford an orange to red mixture. The
mixture was charged and then distilled with heptane (500 mL);
this chase was repeated until the relative area ratio of heptane to
THF by GC-FID of the reaction mixture was >4: 1. The resulting
mixture was cooled to ambient temperature. The solids were
removed by filtration and rinsed with heptane (275 mL). The
filtrate was concentrated to a cloudy red oil (117.3 g, 61.5 wt %).
(
12) (a) MacInnes, I.; Walton, J. C. J. Chem. Soc., Perkin Trans. 2
987, 1077–1082. (b) Yao, G.; Steliou, K. Org. Lett. 2002, 4, 485–488.
c) Banerjee, M.; Roy, S. Org. Lett. 2004, 6, 2137–2140. (d) Schelper,
1
(
M.; De Meijere, A. Eur. J. Org. Chem. 2005, 582–592. (e) Bernard, S.;
Defoy, D.; Dory, Y. L.; Klarskov, K. Bioorg. Med. Chem. Lett. 2009, 19,
6127–6130.
(13) (a) Zweifel, G.; Pearson, N. R. J. Org. Chem. 1981, 46, 829–830.
(b) Soundararajan, R.; Li, G.; Brown, H. C. Tetrahedron Lett. 1994,
1
H NMR in C D with dimethyl fumarate as an internal standard
6
6
showed 61.5 wt % for a reaction yield of 64.3% yield.
The crude borolane 1 (92.6 g, 61.5 wt %) was distilled under
the conditions and equipment setup as depicted in Figure 5 to
afford the borolane 1 as a yellow oil (52.8 g, 88.0 wt %, 81.6%
distillation recovery).
35, 8961–8964. (c) Brown, H. C.; Roy, C. D.; Soundararajan, R.
Tetrahedron Lett. 1997, 38, 765–768.
(
(
14) Corey, E. J.; Kirst, H. A. Tetrahedron Lett. 1968, 9, 5041–5043.
15) (a) Pearson, N. R.; Hahn, G.; Zweifel, G. J. Org. Chem. 1982,
47, 3364–3366. (b) Wang, K. K.; Nikam, S. S.; Ho, C. D. J. Org. Chem.
1
983, 48, 5376–5377. (c) Brown, H. C.; Khire, U. R.; Narla, G. J. Org.
Chem. 1995, 60, 8130–8131. (d) Hernandez, E.; Soderquist, J. A. Org.
Lett. 2005, 7, 5397–5400. (e) Hernandez, E.; Burgos, C. H.; Alicea, E.;
Soderquist, J. A. Org. Lett. 2006, 8, 4089–4091.
’
ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures, com-
b
1
(16) Reich, H. J.; Holladay, J. E.; Walker, T. G.; Thompson, J. L.
plete characterization data for propargyl borolane 1, and copies of
13
J. Am. Chem. Soc. 1999, 121, 9769–9780.
H and C NMR spectra. This material is free of charge via the
Internet at http://pubs.acs.org.
(
(
17) See Supporting Information for DSC profile of borolane 1.
18) Anderson, N. G. Practical Process Research & Development;
Academic Press: San Diego, 2000; Chapters 1 and 13.
(19) Brown, H. C.; Cole, T. E. Organometallics 1985, 4, 816–821.
’
AUTHOR INFORMATION
(
20) Brown, H. C.; Srebnik, M.; Cole, T. E. Organometallics 1986, 5,
300–2303.
21) Raushel, J.; Sandrock, D. L.; Josyula, K. V.; Pakyz, D.; Molander,
G. A. J. Org. Chem. 2011, 76, 2762–2769.
22) Formation of isopropyl acetate was confirmed by GC/MS
Corresponding Author
daniel.fandrick@boehringer-ingelheim.com
2
(
(
’
REFERENCES
analysis of the reaction mixture.
(
1) Hall, D. G. Boronic Acids; Wiley-VCH: Weinheim, Germany,
(23) Brown, H. C.; Cole, T. E. Organometallics 1983, 2, 1316–1319.
(24) The gel was only observed on the control experiment when
lithium isopropoxide was charged to a solution of the isopropyl pinacol
borate. This gel was not observed during the batch or continuous-flow
processes.
2
005.
(
2) Fandrick, D. R.; Reeves, J. T.; Song, J. J. Process for Stereo-
selective Synthesis of Alkynyl Aryl Trifluoromethyl Alcohols via Reac-
tion of Silylalkynylborolanes with Trifluoromethyl Ketones. PCT Int.
Appl. WO/2010/141328, 2010.
(
3) Fandrick, D. R.; Fandrick, K. R.; Reeves, J. T.; Tan, Z.; Johnson,
C. S.; Lee, H.; Song, J. J.; Yee, N. K.; Senanayake, C. H. Org. Lett. 2010,
2, 88–91.
4) Fandrick, D. R.; Johnson, C. S.; Fandrick, K. R.; Reeves, J. T.;
Tan, Z.; Lee, H.; Song, J. J.; Yee, N. K.; Senanayake, C. H. Org. Lett. 2010,
2, 748–751.
5) Fandrick, D. R.; Fandrick, K. R.; Reeves, J. T.; Tan, Z.; Tang, W.;
1
(
1
(
Capacci, A. G.; Rodriguez, S.; Song, J. J.; Lee, H.; Yee, N. K.; Senanayake,
C. H. J. Am. Chem. Soc. 2010, 132, 7600–7601.
(
6) Fandrick, K. F.; Fandrick, D. R.; Reeves, J. T.; Gao, J.; Ma, S.; Li,
W.; Lee, H.; Grinberg, N.; Lu, B.; Senanayake, C. H. J. Am. Chem. Soc.
011, 137, 10332–10335.
7) Fandrick, D. R.; Reeves, J. T.; Tan, Z.; Lee, H.; Song, J. J.; Yee,
N. K.; Senanayake, C. H. Org. Lett. 2009, 11, 5458–5461.
8) Knochel, P.; Ila, H.; Korn, T. J.; Baron, O. Functionalized
2
(
(
Organoborane Derivatives in Organic Synthesis. In Handbook of Func-
tionalized Organometallics; Knochel, P., Ed.; Wiley-VCH: Weinheim,
2
005; Vol 1, Chapter 3.
9) (a) Hoffmann, R. W.; Brinkmann, H.; Frenking, G. Chem. Ber.
990, 123, 2387–2394. (b) Blais, J.; L’Honore, A.; Soulie, J.; Cadiot, P.
(
1
J. Organomet. Chem. 1974, 78, 323–337.
10) (a) Coffee, R. D.; Wheeler, J. J. Loss Prev. 1967, 1, 6–9.
b) Magin, R. W.; Herndon, R. C.; Sabahi, M.; Allen, R. H.; Brantley,
(
(
N. H.; Elnagar, H. Y.; Lin, R. W. Stabilized Propargyl Bromide
Compositions as Soil Fumigants. U.S. Pat. Appl. Publ. U.S.
20030194514, 2003. (c) Sabahi, M.; Magin, R. W.; Herndon, R. C.;
Elnagar, H. Y.; Brantley, N. H.; Lin, R. W.; Allen, R. H. Stabilization and
Use of Propargyl Bromide, U.S. Pat. Appl. Publ. U.S. 2003087024, 2003.
1
140
dx.doi.org/10.1021/op200180t |Org. Process Res. Dev. 2012, 16, 1131–1140