Developing efficient nucleophilic fluorination methods and application to substituted picolinate esters

Article abstract of DOI:10.1021/op5001258

This report describes nucleophilic fluorination of 3 and 5-substituted picolinate ester substrates using potassium fluoride in combination with additive promoters. Agents such as tributylmethylammonium or tetraphenylphosphonium chloride were among the bes

Full text of DOI:10.1021/op5001258

Article
pubs.acs.org/OPRD
Developing Efficient Nucleophilic Fluorination Methods and
Application to Substituted Picolinate Esters
Laura J. Allen,† Shin Hee Lee,† Yang Cheng,‡ Patrick S. Hanley,‡ Joseck M. Muhuhi,‡ Elisabeth Kane,‡
Stacey L. Powers,‡ John E. Anderson,‡ Bruce M. Bell,‡ Gary A. Roth,‡ Melanie S. Sanford,†
and Douglas C. Bland*,‡
†Department of Chemistry, University of Michigan, 930 N. University Ave, Ann Arbor, Michigan 48109, United States
‡Dow Chemical Company, 1710 Building, Midland, Michigan 48674, United States
S
* Supporting Information
ABSTRACT: This report describes nucleophilic fluorination of 3 and 5-substituted picolinate ester substrates using potassium
fluoride in combination with additive promoters. Agents such as tributylmethylammonium or tetraphenylphosphonium chloride
were among the best additives investigated giving improved fluorination yields. Additionally, the choice of additive promoters
could influence the potential formation of new impurities such as alkyl ester exchange. Other parameters explored in this study
include additive stoichiometry, temperature influence on additive degradation, solvent selection, product isolation by solvent
extraction, and demonstration of additive recycling.
adaptation was the ability to premix some of the heteroarene
substrates with tetrabutylammonium cyanide before adding
hexafluorobenzene, thus enabling in situ generation of
anhydrous TBAF. This strategy telescoped the fluorination
procedure into a single one pot process. Overall, the anhydrous
TBAF agent exhibited good solubility in a wide range of solvent
systems, thus delivering a potent fluoride source under much
milder room temperature conditions. Moving forward, we
wanted to apply the better solubility properties associated with
nucleophilic fluoride agents such as anhydrous TBAF toward
cheaper, less soluble metal fluorides such as KF. This paper will
discuss efforts to utilize additives that would enhance reactivity
of KF and apply this methodology toward less active
heteroarene substrates (e.g., specifically picolinate ester motifs).
INTRODUCTION
■
Aryl fluorides are extremely important structural motifs that are
featured prominently in pharmaceuticals, agrochemicals,
organic materials, and biological imaging agents.1 An increasing
number of fluorine-containing biologically active molecules
have been developed for pharmaceutical and agrochemical
applications (Figure 1). Despite rising pharmaceutical and
agrochemical prevalence, the selective fluorination of arenes
remains a difficult task. As a result, significant recent efforts
have focused on the development of new synthetic procedures
for the formation of CAryl−F bonds.2,3
Halex reactions are a class of nucleophilic substitutions where
a carbon−halogen bond (C−X1) is converted to new and
different carbon−halogen (C−X2). For this study, halex
reactions specifically refer to the conversion of aryl−Cl bonds
to aryl−F bonds using a nucleophilic fluoride source. Halex
reactions, typically run in polar aprotic solvents, require
extreme temperatures (greater than 150 °C) largely due to
the poor solubility of metal fluoride salts such as cesium
fluoride (CsF) and potassium fluoride (KF).4 KF is much more
economical than CsF but usually less reactive. Traditionally,
common halex substrates require an electron-withdrawing
group to activate the aryl ring towards nucleophilic attack.
These factors often present challenges for developing efficient
fluorination methods for aromatic and heteroaromatic motifs.
Recently, our efforts have focused on the development of
mild, selective methods to fluorinate aromatic systems. One
strategy developed in our laboratories utilizes the combination
of copper triflate Cu(OTf)2 and KF to convert a variety of aryl
potassium trifluoroborates to the corresponding fluorinated
products in good yield.5 This methodology has a wide range of
functional group compatibility (Figure 2). Additionally, we
developed methodology for high yielding fluorination of
heteroarenes utilizing adaptations to the anhydrous tetrabuty-
lammonium fluoride (TBAF) protocol.6 Unique to this
RESULTS AND DISCUSSION
■
To investigate reaction parameters (additives, solvent selection,
stoichiometry, temperature) for effective nucleophilic substitu-
tion using metal fluorides, monochloro picolinate esters 6 and 9
were chosen. Two complementary synthetic approaches for
these substrates are outlined (Scheme 1). Ethyl-3-chloro-6-
chloropicolinate (6) was prepared via the two-step procedure6
outlined in Scheme 1. Fisher esterification of 3,6-dichloropico-
linic acid followed by a Suzuki cross-coupling with phenyl
boronic acid afforded 6 in 38% yield (for two steps).
Preparation of 9 utilized 2,3-dichloro-6-(trichloromethyl)-
pyridine as the starting material. After purification of 7 by
recrystallization from methanol, treatment with concentrated
sulfuric acid followed by a solvent quench using 2-propanol
directly afforded 8 in good yield. A subsequent Suzuki cross-
coupling with phenyl boronic acid afforded 9 in 94% yield.
Special Issue: Fluorine Chemistry 14
Received: April 14, 2014
Published: July 4, 2014
© 2014 American Chemical Society
1045
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
Figure 1. Representative aryl fluoride containing actives.
Table 1. Fluorination of 6 using various additives at 130 °C
a
b
entry
additive
none
MX
% yield
1
2
3
4
5
6
7
CsF
KF
KF
KF
KF
KF
KF
34
0
none
18-crown-6
Bu4NCl
5
8
Bu4NBr
0
Me4NCl
2
Bu4N[(Ph)3SiF2]
9
a
Conditions: picolinate (1 equiv), PTC additive (2 equiv), MF (2
Figure 2. Mild and selective fluorination strategies for arenes and
heteroarenes.
b
equiv), DMSO, 130 °C. In-pot yield determination by 19F NMR or
GC.
Scheme 1. Synthetic approaches to picolinate model
substrates 6 and 9
a
transfer catalyst (PTC) agents used along with KF increased
yields of 10 above 10%, and mostly unreacted starting material
was observed. As yields of desired product 10 were consistently
low (presumably due to the C3 substitution pattern), our
attention next focused on using substrate 9, which presented a
better opportunity to fine-tune conditions towards enhancing
fluorination yields.
A much larger survey of PTC’s agents was conducted using 9
as a substrate (Table 2). The results of this screen showed
several trends. First, longer alkyl or aromatic chains on the PTC
additive (such as Bu vs Me) provided higher fluorination yields.
Second, PTC agents with a chloride anion generated relatively
higher yields than the same PTC additive paired with a different
anion. From the screening results, several promising leads,
tetrabutylammonium chloride (Bu4 NCl), bis-
(triphenylphosphoranylidene)ammonium chloride ([Ph3P
NPPh3]Cl), tetrabutylphosphonium chloride (Bu4PCl), and
tetraphenylphosphonium chloride (Ph4PCl), were identified as
additives that generated fluorination yields of >75% when
starting with 9.
Among the PTC candidates explored, Bu4NCl and Ph4PCl
are the most promising. Both of these reagents were
subsequently evaluated over a wider equivalent range with
the goal of using substoichiometric quantities of the PTC.
However, as shown in Figure 3, the yield of 12 diminished as
the equivalents of PTC was lowered. Fluorination yields greater
than 50% were still achieved when using 0.5 equiv of PTC
a
(a) EtOH, H2SO4, then Na2CO3, 73% yield; (b) PhB(OH)2,
Pd(Ph3P)4, KF, 52% yield; (c) H2SO4, 130 °C; then i-PrOH, 70 °C,
86% yield; (d) PhB(OH)2, Pd(Ph3P)2Cl2, KF, CH3CN, 94% yield.
Initial efforts surveyed promoters that could improve
nucleophilic substitution activity when used in combination
with metal fluorides. We aimed to increase the reactivity of KF
to be comparable to that of CsF through the use of additives.
To establish a control, a baseline fluorination yield was
measured for both CsF and KF without additives (Table 1
and 2; entries 1−2 respectively). CsF afforded higher observed
yields than KF for both substrates 6 and 9. To standardize
screening conditions and to investigate fluorination additives,
each reaction incorporated two equivalents of both potassium
fluoride and available phase transfer agents such 18-crown-6,7
quaternary ammonium salts,8 and phosphonium salts.9 In the
case of substrate 6 (Table 1), none of the surveyed phase
1046
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
utilizing a larger excess of PTC agents but at temperatures no
higher than 130 °C.
Table 2. Fluorination of 9 using various additives at 130 °C
With lead candidates chosen, our efforts next focused on
several additional parameters such as commercial availability of
PTC, solvent selection, PTC side chemistry, and product
isolation. These reaction parameters were explored to further
develop a scalable process that could efficiently fluorinate
challenging picolinate substrates such as 9.
a
b
entry
additive
MX
% yield
1
none
none
CsF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
KF
83
10
11
86
23
13
20
54
12
13
17
15
46
23
23
75
75
99
0
2
Bulk pricing10 was obtained for 50% Bu4NCl in water
(∼$25/kg). In contrast, bulk pricing for tributylmethylammo-
nium chloride (Bu3MeNCl), structurally similar to Bu4NCl, was
estimated to be ∼$8/kg as a 75 wt % solution in water, and
pricing for Ph4PCl was found to be ∼$40/kg. Thus, significant
savings in raw material costs would be possible if Bu3MeNCl
affords similar fluorination performance to Bu4NCl. Alter-
natively, cost savings would be achievable if a viable recycle
strategy for additives such as Ph4PCl could be developed. In
fact, treating 9 under similar conditions (two equivalents of
both KF and 98% Bu3MeNCl in DMSO at 130 °C) afforded a
75% in-pot yield of 12, which was comparable to previous
results using Bu4NCl as the additive. Several estimates were
modeled to gauge the relative cost ranges anticipated to prepare
12 in a single chemical step (Table 3). For this exercise, only
3
Bu4NF(X H2O)
Bu4NCl
4
5
Bu4NBr
6
Bu4NI
7
Me4NF
8
Me4NCl
9
Me4NBr
10
11
12
13
14
15
16
17
18
19
Me4NI
Bu4NCPF6
Bu4NOTf
Bu4N[(Ph)3SiF2]
Bu4NCN
Bu4NBF4
[Ph3PNPPh3]Cl
Bu4PCl
Ph4PCl
Table 3. Estimated relative cost scenarios for the
fluorination of picolinate 9
MePh3PBr
a
Conditions: picolinate (1 equiv), PTC additive (2 equiv), MF (2
a
b
entry
MF
PTC
none
PTC recycle
% yield
cost ($)
equiv), DMSO, 130 °C. In-pot yield determination by 19F NMR or
GC.
1
2
3
4
5
6
CsF
KF
KF
KF
KF
KF
none
none
0%
83
10
86
75
80
80
78
133
140
44
none
Bu4NCl
Bu3MeNCl
Ph4PCl
Ph4PCl
together with 2.0 equiv of KF in DMSO at 130 °C. When the
same range of equivalents was explored at a higher jacket
temperature set point of 150 °C, the yields increased at low
loadings of additive (∼30−40% at 0.1 equiv PTC). Diminished
yields were observed for both Bu4NCl and Ph4PCl at loadings
between 1.0 and 2.0 equiv when heated at 150 °C. These
results show that effective fluorination is achievable when
0%
0%
161
31
90%
a
Cost of reagents required for fluorination/kg of fluorinated product.
Figure 3. Temperature and stoichiometry influence on yields of 12 in DMSO.
1047
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
Figure 4. Observed fluorination yield of 12 for various common polar solvents.
Figure 5. Measured degradation of the Bu3MeNCl/KF mixture.
the cost of the raw materials required for fluorination was
included. It was assumed that the reaction solvent (DMSO)
would be recovered for recycling at an 80% rate. These cost
estimations (Table 3) suggested that Bu3MeNCl and Ph4PCl
(at a recycling rate of 90%) could be viable additives to
consider for fluorination studies.
optimal option due to lower toxicity. As will be discussed in a
later section of this report, sulfolane was also an effective
solvent for developing a product isolation strategy associated
with this fluorination approach.
As previously mentioned, good fluorination yields (>70%)
required adding an excess of PTC to promote KF efficiency.
Quaternary ammonium salts (such as Bu3MeNCl) are likely
degrading competitively at the elevated reaction temperatures
associated with this process and thus depend upon excess molar
equivalents to compensate. The stability of quaternary
ammonium salts has been previously studied, and typical
byproducts include the formation of trialkylamines.12 As part of
the current studies, we sought to better understand the degree
of Bu3MeNCl decomposition by measuring relative concen-
trations of active species as a function of time and temperature.
A thermal study evaluated a prepared mixture of Bu3MeNCl
and KF (45:55) in DMSO, and the solution was stirred for set
time intervals at ambient temperature, 90 and 120 °C,
respectively (Figure 5). ESI/LC/MS spectroscopy was
employed to approximate relative area ion % concentrations
Our initial additive screening efforts focused on reaction
parameters where DMSO was the solvent medium; however, it
became prudent to identify other process-friendly solvents that
could alleviate potential scale-up problems associated with
DMSO.11 Amide solvents (e.g., NMP) as well as less toxic
solvents (e.g., sulfolane) were evaluated under standard
screening reaction conditions, and the fluorination yields
were compared to DMSO results (Figure 4). For this effort,
98% Bu3MeNCl was chosen as the additive without further
purification or drying efforts. All of these solvents were
screened in a drybox (0.5 g of 9 at 130 °C for 24 h), and the
crude reaction mixtures were analyzed for in-pot yields by GC.
Overall, the observed fluorination yields exhibited a range
spanning 74−85%. While these solvents were all viable
candidates to replace DMSO, sulfolane was considered the
1048
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
Figure 6. Observed ester exchange products from the Bu3MeNCl additive.
for species such as quaternary ammonium cations as well as
tertiary amine byproducts.
observed ester exchange byproduct formation with respect to
larger scale fluorination batch runs using either Bu3MeNCl or
Ph4PCl promoters.
Two different quaternary ammonium cations (Bu3MeNX
and Bu2Me2NX) were detected in this mixture. After stirring at
ambient temperature for about 16 h, the relative concentrations
of Bu3MeNX and Bu2Me2X were 93.20% and 6.72%,
respectively. The relative concentration of two tertiary amines,
tributylamine (Bu3N) and dibutylmethylamine (Bu2MeN),
were both less than 0.1%. The mixture temperature was then
subsequently raised to 90 °C and then analyzed after stirring for
6.25 and 12 h, respectively. At 90 °C, very little change in
relative concentrations of either quaternary ammonium cation
or tertiary amine was observed. However, degradation was
detected as the mixture was subsequently heated to 120 °C for
an additional period of time. The relative Bu3MeNX
concentration dropped about 8% after stirring at 120 °C for
19.3 h. Correspondingly, the sum total concentration increase
of both Bu3N and Bu2MeN equaled 8% within the same time
span. Thus, this preliminary evidence suggested that degrada-
tion of active Bu3MeNX under typical halex reaction temper-
atures should occur slowly as a competing reaction pathway.
Besides thermal degradation under reaction conditions,
Bu3MeNCl can promote the formation of ester exchange
byproducts 13 and 14 (Figure 6). For example, partial
conversion to 13 was observed after extended heating of 9
with excess Bu3MeNCl in sulfolane at 130 °C. Water likely does
not play a direct role in ester exchange, as heating 9 in the
presence of “wet” Bu3MeCl gave less of 13 (Figure 6). A slow
increase in the concentration of 14 was also observed during
fluorination reactions, suggesting a similar ester exchange effect.
Fortunately, a strategy to circumvent ester exchange could
involve substituting another PTC that cannot deliver an alkyl
substituent in place of Bu3MeNCl. As proof of concept, heating
a mixture of 9 and Ph4PCl in sulfolane at 130 °C generated no
ester exchange byproducts; thus giving a viable means to shut
down this pathway. A later section of this report will discuss
Isopropyl-5-fluoro-6-phenylpicolinate (12) exists as a
viscous, colorless oil making product isolation from reaction
mixtures more challenging. Our efforts investigated an isolation
strategy that could alleviate the need for chromatography by
using a liquid extraction approach. A viable option would
ideally partition the reaction mixture between an organic
solvent and water. In principle, the organic solvent component
would prefer to extract only 12 and simultaneously reject more
polar reaction components such as PTC additive, reaction
solvent, and other polar species to the aqueous phase. To test
the feasibility of this extraction strategy, three organic solvents
were compared. Toluene, hexanes, and isopar C were selected
due to low miscibility with water as well as the ability to
solubilize 12. For this study, the reaction mixture (containing
∼4.5 wt % of 12) was partitioned between a 4:1 volume ratio of
the chosen organic solvent and water. The organic phase was
sequentially washed with first a dilute aqueous HCl solution (to
remove tertiary alkyl amine byproducts) followed by a final
water wash. Concentrating the organic extract afforded the
product oil, and then both % product recovery and purity were
estimated by GC analysis (Table 4).
Both toluene and hexanes demonstrated excellent (>95%)
recovery yield of 12 from the crude reaction mixture. Isopar C
Table 4. Solvent comparison of the product isolation
efficiency
extraction solvent
toluene isopar C hexane
a
% recovery of isolated 12 from rxn mix
95.4%
68.9%
87.8%
86.8%
97.2%
87.1%
a
calc. purity of 12 in isolated product
a
Purity and recovery yield estimated using GC analysis.
1049
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
Table 5. Results for reaction scale-up batches
PTC
(equiv)
KF
(equiv)
% conv.
of 9
% inpot yield
% Yield 12 in isolated
% yield 14 in isolated
% recov. 9 in isolated
a
ab
,
ab
,
ab
,
batch
PTC
12
prod.
prod.
prod.
c
1
Bu3MeNCl
Bu3MeNCl
Bu3MeNCl
Bu3MeNCl
Ph4PCl
2.0
1.0
1.9
2.0
2.0
2.0
3.0
2.0
2.0
2.0
95
91
87
95
99
94
81
75
72
76
69
68
72
59
90
3
3
4
5
0
0
5
8
d
2
d
3
12
5
e
f
4
NA
g
5
80
88
0
h
i
j
6
Ph4PCl
0.1
2.0
6
a
b
c
Conditions: Yields measured by GC assay. For the isolated product, the yield of component calculated based on 9 as the limiting reagent. Added
d
9, 98% Bu3MeNCl, KF, and dry sulfolane (<100 ppm water), 130 °C, 24 h. 75% Bu3MeNCl in water, sulfolane, 9, toluene; vacuum distillation; then
KF, 130 °C, 24 h. 98% Bu3MeNCl used as PTC. Reaction mixture taken directly through isolation. 98% Ph4PCl used as PTC. Aqueous Ph4PCl/
sulfolane/recovered 9 from batch 5 was reused. Added additional 10% mol of fresh Ph4PCl to existing Ph4PCl recovered from batch 5. Added fresh
e
f
g
h
i
j
KF (2 equiv) to existing KF recovered from batch 5.
exhibited slightly diminished capacity for extracting the desired
product (88% recovery); however, the purity of the isolated
product from isopar C was similar to purity obtained when
using hexanes to extract. Several observations were noted from
the isolated product mixtures. First, isolated product contained
very low concentrations of Bu3MeNCl regardless of the
extraction solvent utilized.13 Second, product isolated from
isopar C and hexanes appeared to not contain any residual
sulfolane. Strikingly, isolated product from the toluene
extraction run contained observable levels of sulfolane, which
also explained the much lower estimated purity (∼69%).
Overall, these preliminary results indicated aliphatic solvents
such as hexanes or isopar C offer good potential for efficient
product separation from the reaction mixture. With a better
understanding of side product chemistry and an isolation
strategy, we next proceeded to explore conditions to
demonstrate scale-up of this methodology.
This scale up demonstration specifically utilized additives
Bu3MeNCl and Ph4PCl and also leveraged aforementioned
process insights from the smaller scale studies. A total of six
batches at increased (18−26×) scale were performed using 9−
13 g of 9 (Table 5). Batch 1 involved directly scaling up
reaction conditions similar to drybox conditions. In this
example, all of the reaction components were carefully weighed
into a 200 mL Schlenk flask inside the drybox.14 The flask was
sealed using a rubber stopper to hold the inert atmosphere
from the drybox, and the vessel was subsequently transported
to a fume hood and heated at 130 °C for 24 h with magnetic
stirring. For the remaining scale-up examples (batches 2−6), a
more traditional reactor setup utilized a 500 mL three-neck
round-bottom flask fitted with mechanical stirring and vacuum
distillation capabilities. Batches 2 and 3 incorporated a 75 wt %
solution of Bu3MeNCl in water as the PTC additive. As
previously discussed, the moderate thermal stability of this PTC
agent required drying temperatures less than 120 °C to
minimize degradation. Azeotropically drying the reaction
mixture at reduced pressure using toluene as a cosolvent was
the approach demonstrated for batches 2 and 3. During the
azeotropic distillation, the vacuum was adjusted between 26
and 80 mmHg as the pot temperature was gradually raised from
ambient temperature to no higher than 120 °C. Under these
drying conditions, about 27 mL of toluene was used per 1 g of
starting material 9. Fortunately about 80% of the toluene could
be easily recovered after completing the drying step, and the
reaction mixture usually contained less than 300 ppm water.15
Dry KF16 was quickly added in one portion to the vessel under
a pad of nitrogen, and the mixture was subsequently heated to
130 °C for approximately 21 to 24 h. When using two
equivalents of 75 wt % Bu3MeNCl (e.g., batch 3), the
azeotropic drying required more toluene to obtain reaction
moisture levels below 300 ppm. Batch 4 incorporated process
conditions similar to batches 2 and 3 except solid 98%
Bu3MeNCl was substituted as the PTC additive. Batch 5 was
very similar to batch 4 except the PTC additive was substituted
with 98% Ph4PCl. Batch 6 investigated the feasibility of
recycling the aqueous Ph4PCl/sulfolane mixture recovered
from batch 5. In this run, the reactor was charged with the
aqueous PTC solution, and the water was removed by vacuum
distillation. LC analysis of the mixture indicated the presence of
leftover 12 from batch 5 (which explains the lower isolated
yield reported), and this material was carried into the
subsequent reaction steps associated with batch 6. For the
azeotropic drying step in batch 6, the reactor was loaded with a
fresh 10 mol % portion of Ph4PCl followed by 9 in toluene.
After completing the drying step, the reactor was charged with
fresh KF (2 equiv) and the reaction heated at 130 °C for 24 h.
Overall, batch results produced in-pot yields consistent with
conditions explored in the previous screening studies.
Fluorination performance also gave comparable yields regard-
less of whether solid PTC agent was utilized (e.g., batches 1, 4,
5, and 6) or an aqueous solution (batches 2 and 3). The
conversion of starting material 9 for all batches was good to
excellent, and most of unreacted 9 could be accounted for in
the isolated product mixture. As anticipated, ester exchange was
also observed for batches 1−4 employing Bu3MeNCl as the
additive. For these batches, the isolated product yield after
solvent extraction was predominantly a mixture of starting
material 9 and fluorinated esters 12 and 14. The observed yield
range for 12 and 14 in the isolated product was 68−76% and
3−5%, respectively. When the PTC additive was switched to
Ph4PCl (batch 5 and 6), only 12 was observed in comparable
yields and with no ester exchange byproduct formation. For
batch 5, the observed yield of 59% was slightly lower than for
the batches (1−4) following product isolation by solvent
extraction with hexanes. As previously mentioned, unrecovered
12 in the Ph4PCl recycle stream from batch 5 was subsequently
carried into batch 6 which accounted for the boost in observed
product yield. Toluene was utilized as the extraction solvent,
producing much better recovery of 12; however, sulfolane was
also retained in the product mixture (giving an estimated 45%
purity of 12 for batch 6). Further optimization of this extraction
procedure using toluene will be considered in future work.
1050
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
to 130 °C, and 2,3-dichloro-6-(trichloromethyl)pyridine (100.0
g, 0.4 mol) prewarmed to 60 °C was added in portions over 40
min. Degassing to the caustic trap was observed. The mixture
was stirred at 130 °C for 3 h and then cooled to 50 °C
overnight. The mixture was warmed to 70 °C, and isopropyl
alcohol (123.6 g, 2.1 mol) was added slowly via the addition
funnel over 55 min. Rapid degassing of HCl gas was observed.
The reaction mixture was stirred at 70 °C for an additional 1 h.
The reaction mixture was poured over crushed ice (400 g), and
the slurry was allowed to precipitate for 1 h in a refrigerator.
The solids were collected by vacuum filtration, and the cake
was washed with a 50:50 mixture of isopropyl alcohol/water
(100 g) and then by water (200 g). The white solid was allowed
to air-dry to constant weight. Isopropyl-5,6-dichloropicolinate
was obtained as a white solid (76.9 g, 0.3 mol, 86% yield, >98%
CONCLUSION
■
Efficient nucleophilic fluorination of 3 and 5-substituted
picolinate ester substrates was accomplished using potassium
fluoride in combination with additive promoters. Agents such
as tributylmethylammonium or tetraphenylphosphonium chlor-
ide were among the best additives screened improving
fluorination in-pot yields up to 81%. Additionally, the choice
of additive promoter influenced the formation of new
impurities such as alkyl ester exchange. Other parameters
explored in this study include additive stoichiometry, temper-
ature effects on additive degradation, solvent selection, product
isolation by solvent extraction, and demonstration of additive
recycling.
1
EXPERIMENTAL SECTION
purity). mp = 63−65 °C; H NMR (400 MHz, CDCl3-d3) δ
■
7.96 (d, J = 8.1 Hz, 1H); 7.88 (d, J = 8.1 Hz, 1H); 5.29 (hept, J
= 6.3 Hz, 1H); 1.40 (d, J = 6.3 Hz, 6H); 13C NMR (100.6
MHz, CDCl3) δ 162.9, 149.4, 146.7, 139.4, 134.5, 124.5, 70.4,
21.9; HRMS ESI+. (m/z): [M + H]+ calcd for C9H10Cl2NO2:
234.0083. Found: 234.0101
All stock solutions were made using volumetric glassware. All
liquid reagents were dispensed by difference from syringes. All
reagents were weighed out in a nitrogen-filled drybox with
exclusion of air and moisture, unless otherwise noted. All
reagents were purchased from common suppliers and dried
over P2O5 prior to use unless otherwise noted. NMR spectra
Preparation of Isopropyl 5-Chloro-6-phenylpicoli-
nate. In a 1 L 4-neck round-bottom flask equipped with a
mechanical overhead stirrer was charged KF (39.9 g, 688
mmol), water (125 mL), and acetonitrile (500 mL). The
mixture was stirred until all of the solids dissolved. To this
biphasic mixture was added phenylboronic acid (42.0 g, 344
mmol) and then isopropyl 5,6-dichloropicolinate (53.6 g, 229
mmol). The resultant suspension was sparged with N2 gas for
15 min, and then bis(triphenylphosphine)palladium chloride
(4.0 g, 5.7 mmol) was added. The bright yellow suspension was
sparged with N2 for further 15 min, and then the mixture was
heated to 55 °C. At the 28 h mark the reaction was deemed
complete by HPLC analysis. The heating mantle was removed
and the mixture cooled to ambient temperature. The mixture
was diluted with EtOAc (250 mL). The layers were separated,
and the organic layer was washed with water (2 × 100 mL).
The aqueous layer was extracted with ethyl acetate (2 × 100
mL) and the organic layers combined. The combined organic
layers were then washed with brine (1 × 100 mL) and SiO2
(161 g) added. The solvent was removed by a rotary
evaporator, and the crude product was purified by CombiFlash
chromatography (750 g column) to afford cream colored solid
1
were recorded on a 700 MHz (699.76 MHz for H; 175.95
MHz for 13C), 500 MHz (500.10 MHz for 1H; 125.75 MHz for
13C, 470.56 MHz for 19F), or 400 MHz (400.52 MHz for H;
1
100.71 for 13C; 376.87 MHz for 19F) NMR spectrometer with
the residual solvent peak (CDCl3: H: δ = 7.27 ppm, 13C: δ =
1
77.16 ppm) used as the internal reference unless otherwise
noted. Chemical shifts are reported in parts per million (ppm)
(δ) relative to tetramethylsilane. Multiplicities are reported as
follows: br (broad resonance), s (singlet), d (doublet), t
(triplet), q (quartet), hept (heptuplet), m (multiplet). Coupling
constants (J) are reported in Hz. NMR yields were determined
using trifluorotoluene (10 uL/rxn) as an internal standard. GC
inpot yields were determined using either 2-phenylpyridine or
biphenyl as an internal standard. Gas chromatography was
carried out on a Shimadzu 17A or Agilent 6850 Network GC.
HPLC was performed on a Agilent 1100 or PerkinElmer 200
series system series system. Purification by preparative
chromatography was performed using either CombiFlash Rf
200 or CombiFlash Rf Mm Hgent systems.
Purification of 2,3-Dichloro-6-(trichloromethyl)-
pyridine by Recrystallization.17 To a 1 L jacketed reactor
fitted with mechanical stirring, reflux condenser, and nitrogen
inertion was charged 2,3-dichloro-6-(trichloromethyl)pyridine
(420 g, 1.6 mol); methanol (218 g, 6.8 mol) was added, and the
mixture was stirred for 5−10 min with a jacket temperature at
40 °C. The solution was then bottom drained into a 1 L conical
flask and the flask was placed into a refrigerator. After allowing
the solution to stand overnight, the product precipitated out of
solution. The precipitate was suction filtered through a glass
fritted funnel, and the cake was allowed to fully deliquor. The
material was dried under vacuum to obtain 2,3-dichloro-6-
(trichloromethyl)pyridine as a white crystalline solid: (339 g,
purity was 95.7% by GC assay using biphenyl as an internal
1
weighing 59.7 g (94% yield). mp = 40−42 °C; H NMR (400
MHz, CDCl3) δ 7.98 (d, J = 8.2 Hz, 1 H), 7.91 (d, J = 8.2 Hz, 1
H), 7.81−7.76 (m, 2 H), 7.50−7.43 (m, 3 H), 5.32 (hept, J =
6.3 Hz, 1 H), 1.42 (d, J = 6.4 Hz, 6 H, CH3); 13C NMR (100.6
MHz, CDCl3) δ 164.1, 156.7, 146.8, 138.9, 137.6, 133.6, 129.8,
129.2, 128.1, 124.2, 69.8, 22.0; HRMS ESI+. (m/z): [M + H]+
calcd for C15H15ClNO2: 276.0786. Found: 276.0822.
General Procedure for Halex Additive Screening with
Ethyl-3-chloro-6-phenylpicolinate. A stock solution of
substrate (6) in DMSO was made in a volumetric flask to a
final concentration of 50 mg 6/0.5 mL DMSO. All solids
(metal fluorides/phase-transfer-catalyst/additives) were
weighed into a 4 mL vial equipped with a micro stir bar. The
stock solution of 6 in DMSO was then added (0.5 mL), and the
vial was sealed with a Teflon-lined screw cap. The reaction vial
was removed from the N2 drybox and placed on an IKA
heating/stirring plate with temperature probe, equipped with
an aluminum heating block. The reaction was heated/stirred at
the specified temperature (generally 130 or 150 °C) for the
given amount of time (generally 24 h). After the reaction was
1
standard, 77.3% recovery): mp = 37−40 °C; H NMR (400
MHz, CDCl3) δ 7.91 (s, 2H).
Preparation of Isopropyl-5,6-dichloropicolinate. To a
three-necked 500 mL round-bottom flask fitted with a
thermocouple, mechanical stirrer and a condenser vented to a
knockout pot, and then to an aqueous sodium hydroxide
scrubber was added concentrated H2SO4 (39.2 g, 0.4 mol) and
sulfolane (41.0 g, 0.3 mol). The reaction mixture was warmed
1051
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
complete, it was allowed to cool to room temperature, diluted
with methylene chloride, and internal standards (trifluoroto-
luene and 2-phenyylpyridine) were added. Yields were
determined by 19F NMR and GC.
ratio of starting material to isopropyl-5-fluoro-6-phenylpicoli-
nate to butyl-5-fluoro-6-phenylpicolinate. After cooling to <35
°C, the crude reaction mixture (191.2 g) was collected. GC
analysis (using biphenyl as an internal standard) indicated the
reaction mixture contained 3.7 wt % of isopropyl 5-fluoro-6-
phenylpicolinate which corresponded to a calculated 80.8%
inpot yield. The remaining 184.9 g of this crude reaction
mixture (which contained 3.7 wt % of isopropyl 5-fluoro-6-
phenylpicolinate) was partitioned between 203.9 g of water and
34.7 g of hexanes in a 500 mL separatory funnel. The mixture
was allowed to stand for 5 min, and then the top organic phase
was separated and saved. The bottom aqueous phase was
extracted with a second 36.0 g portion of hexanes. The
combined organic phases were sequentially washed with 50.6 g
of 1 M aqueous HCl and then 51.9 g of water. The top organic
phase was separated and concentrated on a rotovap at 40 °C
and <20 mmHg vacuum to remove light organics. This gave 7.6
g of light orange oil in 76.1% yield and 87.1% purity (as
determined by GC analysis using biphenyl as an internal
standard). The isolated product also contained 3.5 wt % butyl-
5-fluoro-6-phenylpicolinate (2.9% yield based on starting
material) and 6.1 wt % isopropyl-5-chloro-6-phenylpicolinate
(5.0% recovery of starting material).
Preparation of Isopropyl-5-fluoro-6-phenylpicolinate
Using Tributylmethylammonium Chloride via an Azeo-
tropic Drying Method. The reaction assembly consisted of a
three-neck 500 mL round-bottom flask fitted with mechanical
stirring, a heating mantle, and inertion was achieved by passing
ambient nitrogen through a packed bed of drierite. The
reaction vessel was fitted with a modified Dean−Stark
distillation head which contained a three-way stopcock on the
receiver arm, and this stopcock was set to divert collected
distillate to a subsequent reservoir 250 mL round-bottom flask.
To the reactor was loaded 29.4 g (93.4 mmol) of 75 wt %
tributylmethylammonium chloride in water followed by 175.2 g
(1.5 mol) of sulfolane that was premelted at 50−60 °C. To the
reaction mixture was added a solution of 13.0 g (47.2 mmol) of
isopropyl-5-chloro-6-phenylpicolinate in 132.8 g (1.4 mol) of
toluene. The reaction was evacuated (∼24 mmHg) and heated
from 22 to 92 °C, and toluene distillate (104.7 g) was collected
over a 42 min period. Karl Fisher analysis of the reaction
mixture indicated 2271 ppm of water. To the reaction vessel
was charged another 87.4 g (1.0 mol) of fresh toluene. The
reaction was evacuated (∼24 mmHg) and heated from 28 to 96
°C, and toluene distillate (76.7 g) was collected over an 18 min
period. Karl Fisher analysis of the reaction mixture indicated
413 ppm of water. To the reaction vessel was charged a final
86.80 g (0.94 mol) of fresh toluene. The reaction was evacuated
(∼24 mmHg) and heated from 27 to 83 °C to distill of the
toluene over a 20 min period. Karl Fisher analysis of the
reaction mixture indicated 133 ppm of water. After cooling the
vessel to below 40 °C under nitrogen inertion, 5.5 g (94.8
mmol) of anhydrous potassium fluoride was shot added
through a reaction port, and the mixture was heated to 130
°C with stirring. After 23 h, LC analysis indicated a 3:5:1 ratio
of starting material to isopropyl-5-fluoro-6-phenylpicolinate to
butyl-5-fluoro-6-phenylpicolinate. The reaction vessel contents
(209 g) were transferred to 1 L flask. GC analysis (using
biphenyl as an internal standard) indicated the reaction mixture
contained 4.2 wt % of isopropyl 5-fluoro-6-phenylpicolinate
which corresponded to a calculated 71.7% inpot yield. To the
reactor was loaded 202.6 g of water for rinsing. This water wash
was then transferred to the 1 L flask containing the reaction
General Procedure for Halex Additive Screening with
Isopropyl-5-chloro-6-phenylpicolinate. Isopropyl-5-
chloro-6-phenylpicolinate (50 mg), and all other solids (metal
fluorides/phase-transfer-catalyst/additives) were weighed into a
4 mL vial equipped with a micro stir bar. DMSO (0.5 mL) was
then added, and the vial was sealed with a Teflon-lined screw
cap. The reaction vial was removed from the N2 drybox and
placed on an IKA heating/stirring plate with temperature
probe, equipped with an aluminum heating block. The reaction
was heated/stirred at the specified temperature (generally 130
or 150 °C) for the given amount of time (generally 24 h). After
the reaction was complete, it was allowed to cool to room
temperature, diluted with DCM, and internal standards
(trifluorotoluene and 2-phenyylpyridine). Yields were deter-
mined by 19F NMR and GC.
Preparation of Isopropyl-5-fluoro-6-phenylpicolinate
Using Cesium Fluoride. The reaction was carried out in a
glovebox. To a glass jar equipped with a stir bar was added
isopropyl 5-chloro-6-phenylpicolinate (1.1 g, 3.9 mmol),
cesium fluoride (1.2 g, 8.0 mmol), and DMSO (anhydrous
grade, 8.5 g). The mixture was heated to 120 °C on a heating
block for 19 h. A sample was taken and analyzed by GC, which
indicated that this reaction was complete. The mixture was
cooled to room temperature. The salts were removed by
filtration and washed with 3 g of DMSO. The mixture was
poured into a separatory funnel with 12 g of water and
extracted with three 12 g portions of ethyl acetate. The organic
phase was then washed with three 12 g portions of water and
dried over anhydrous MgSO4. The solvent was removed with a
rotary evaporator. The concentrated crude product was purified
using column chromatography eluting with an ethyl acetate/
hexanes mixture (1:5) to give 0.9 g of isopropyl-5-fluoro-6-
phenylpicolinate as a pale yellow liquid (81% yield, ∼100%
purity by GC area, 98% purity by LC area). GC/MS results
were consistent with the chemical formula of isopropyl-5-
fluoro-6-phenylpicolinate (11): 70 eV EIMS (GC) m/z = 259
(8.3%), 174 (12.5%), 173 (100%), 172 (35.4%), 145 (9.4%);
1H NMR (400 MHz, CDCl3) δ 8.07 (m, 3H), 7.57 (dd, J =
10.4, 8.5 Hz, 1H), 7.53−7.44 (m, 3H), 5.33 (hept, J = 6.3 Hz,
1H), 1.44 (d, J = 6.3 Hz, 6H); 13C NMR (101 MHz, CDCl3) δ
163.9 (s), 159.2 (d, J = 267 Hz), 146.4 (d, J = 12 Hz), 144.6 (d,
J = 4.8 Hz), 134.7 (d, J = 5.6 Hz), 129.8, 129.2 (d, J = 5.9 Hz),
128.6, 125.5 (d, J = 5.5 Hz), 124.7 (d, J = 21 Hz), 69.7, 22.0;
19F NMR (376 MHz, CDCl3) δ −117.8.
Preparation of Isopropyl-5-fluoro-6-phenylpicolinate
Using Tributylmethylammonium Chloride in Schlenk
Equipment. A 200 mL Schlenk flask equipped with a magnetic
stir bar and thermowell fitted onto the 24/40 ground glass joint
was brought into a glovebox with nitrogen inertion. To this
vessel was sequentially charged 9.30 g (33.73 mmol) of 5-
chloro-6-phenylpicolinate followed by 16.00 g (67.84 mmol) of
98% tributylmethylammonium chloride followed by 4.0 g (68.9
mmol) of anhydrous potassium fluoride and then finally 169.6 g
(1.4 mol) of dry sulfolane (Karl Fisher <100 ppm). The
Schlenk flask was sealed with a rubber septum in the port
containing the stopcock, and the stopcock was left open (the
flask was essentially padded with nitrogen). The flask was then
removed from the glovebox and allowed to heat to 130 °C with
magnetic stirring. After 24 h, LC analysis indicated a 2:23:1
1052
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
mixture. This mixture was subsequently transferred to a 1 L
separatory funnel. The reaction vessel was rinsed with 35.5 g of
hexanes, and this rinse was transferred to 1 L separatory funnel.
After mixing, the phases were allowed to settle for 5 min, and
then the top organic phase was separated and saved. The
bottom aqueous phase was extracted with a second 35.0 g
portion of hexanes. The combined organic phases were
sequentially washed with 54.3 g of 1 M aqueous HCl and
then 51.4 g of water. The top organic phase was separated and
concentrated on a rotovap at 40 °C and <20 mmHg vacuum to
remove light organics. This gave 11.4 g of light orange oil in
68.4% yield and 73.5% purity (as determined by GC analysis
using biphenyl as an internal standard). The isolated product
also contained 4.9 wt % butyl-5-fluoro-6-phenylpicolinate
(4.2% yield based on starting material) and 14.0 wt %
isopropyl-5-chloro-6-phenylpicolinate (12.0% recovery of start-
ing material).
ASSOCIATED CONTENT
■
S
* Supporting Information
1H and 13C for new compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: dcbland@dow.com. Tel.: (989) 638-6946.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors would like to thank Greg Whiteker and Jim Renga
of Dow AgroSciences for helpful suggestions on Halex
methodology.
Preparation of Isopropyl-5-fluoro-6-phenylpicolinate
Using Tetraphenylphosphonium Chloride via an Azeo-
tropic Drying Method. The reaction assembly consisted of a
three-neck 500 mL round-bottom flask fitted with mechanical
stirring, a heating mantle, and inertion was achieved by passing
ambient nitrogen through a packed bed of drierite. The
reaction vessel was fitted with a modified Dean−Stark
distillation head which contained a three-way stopcock on the
receiver arm, and this stopcock was set to divert collected
distillate to a subsequent reservoir 250 mL round-bottom flask.
To the reactor was loaded a slurry of 13.0 g (47.2 mmol) of
isopropyl-5-chloro-6-phenylpicolinate and 35.7 g (95.1 mmol)
of 98% tetraphenylphosphonium chloride in 87.9 g (1.0 mol) of
toluene, and then an additional 41.7 g (0.5 mol) toluene flush
was added to the vessel. To the reaction vessel was charged
181.70 g (1.51 mol) of sulfolane that was premelted at 50−60
°C. The reaction was evacuated (16−21 mmHg) and heated
from 26 to 107 °C, and toluene distillate (105.2 g) was
collected over a 35 min period. Karl Fisher analysis of the
reaction mixture indicated 72 ppm of water. After cooling the
vessel to below 40 °C under nitrogen inertion, 5.7 g (98.5
mmol) of anhydrous potassium fluoride was shot added
through a reaction port, and the mixture was heated to 130
°C with stirring. After 23 h, LC analysis indicated a full
consumption of isopropyl-5-chloro-6-phenylpicolinate. The
reaction mixture was cooled to 80 °C (this avoided
tetraphenylphosphonium chloride from precipitating out of
solution) and the vessel contents (231 g) were transferred to 1
L flask. GC analysis (using biphenyl as an internal standard)
indicated the reaction mixture contained 4.2 wt % of isopropyl
5-fluoro-6-phenylpicolinate which corresponded to a calculated
80.2% inpot yield. The reactor was rinsed with 203.0 g of water
and this wash was then transferred to the 1 L flask containing
the reaction mixture. This mixture was subsequently transferred
to a 1 L separatory funnel. The reaction vessel was rinsed with
34.0 g of hexanes and this rinse was transferred to 1 L
separatory funnel. After mixing, the phases were allowed to
settle for 5 min, and then the top organic phase was separated
and saved. The bottom aqueous phase was extracted with a
second 37.0 g portion of hexanes. The combined organic
phases were washed with 51.7 g of water. The top organic
phase was separated and concentrated on a rotovap at 40 °C
and <20 mmHg vacuum to remove light organics. This gave 7.8
g of dark orange oil in 58.7% yield and 91.9% purity (as
determined by GC analysis using biphenyl as an internal
standard).
REFERENCES
■
(1) (a) Jeschke, P. Pest Manage. Sci. 2010, 66, 10. (b) Purser, S.;
Moore, P. R.; Swallow, S.; Gouverneur, V. Chem. Soc. Rev. 2008, 37,
320. (c) Hagmann, W. K. J. Med. Chem. 2008, 51, 4359. (d) Jeschke, P.
ChemBioChem 2004, 5, 570.
(2) (a) Barnette, W. E. J. Am. Chem. Soc. 1984, 106, 452.
(b) Differding, E.; Wehrli, M. Tetrahedron Lett. 1991, 32, 3819.
(c) Yamada, S.; Gavryushin, A.; Knochel, P. Angew. Chem., Int. Ed.
2010, 49, 2215. (d) Anbarasan, P.; Neumann, H.; Beller, M. Angew.
Chem., Int. Ed. 2010, 49, 2219.
(3) (a) Fier, P. S.; Hartwig, J. F. J. Am. Chem. Soc. 2012, 134, 10795.
(b) Fier, P. S.; Luo, J.; Hartwig, J. F. J. Am. Chem. Soc. 2013, 135, 2552.
(c) Mazzotti, A. R.; Campbell, M. G.; Tang, P.; Murphy, J. M.; Ritter,
T. J. Am. Chem. Soc. 2013, 135, 14012. (d) Hollingworth, C.;
Gouverneur, V. ChemComm. 2012, 48, 2929.
(4) (a) Wynn, D. A.; Roth, M. M.; Pollard, B. D. Talanta 1984, 31,
1036. (b) Gonzalez-Trueba, G.; Paradisi, C.; Scorrano, G. Gazzetta
Chim. Italiana 1996, 126, 457. (c) Vlasov, V. M. J. Fluorine Chem.
1993, 61, 193.
(5) (a) Ye, Y.; Schimler, S. D.; Hanley, P. S.; Sanford, M. S. J. Am.
Chem. Soc. 2013, 135, 16292. For other complementary copper
mediated fluorination strategy using an F+ oxidant, see: (b) Ye, Y.;
Sanford, M. S. J. Am. Chem. Soc. 2013, 135, 4648.
(6) (a) Allen, L. J.; Muhuhi, J. M.; Bland, D. C.; Merzel, R.; Sanford,
M. S. J. Org. Chem. 2014, 79, 5827. (b) Sun, H.; DiMagno, S. G. J. Am.
Chem. Soc. 2005, 127, 2050. (c) Sun, H.; DiMagno, S. G. Angew.
Chem., Int. Ed. 2006, 45, 2720.
(7) Liotta, C. L.; Harris, H. P. J. Am. Chem. Soc. 1974, 96, 2250.
(8) (a) Carpino, L. A.; Sau, A. C. J. Chem. Soc., Chem. Commun. 1979,
11, 514. (b) Sasson, Y.; Negussie, S.; Royz, M.; Mushkin, N. Chem.
Commun. 1996, 3, 297.
(9) Yoshida, Y.; Kimura, Y. Chem. Lett. 1988, 8, 1355.
(10) Source and pricing availability provided by John Kantzes of
SACHEM Inc.
(11) Wang, Z.; Richter, S. M.; Gates, B. D.; Grieme, T. A. Org. Process
Res. Dev. 2012, 16, 1994.
(12) Cox, D. P.; Terpinski, J.; Lawrynowicz, W. J. Org. Chem. 1984,
49, 3216.
(13) The Supporting Information contains NMR data which
illustrates the product purity obtained from using the three different
extraction solvents (toluene, isopar C, and hexanes).
(14) The hexanes extraction procedure was applied to this batch and
the results are summarized in Table 4.
(15) Halex reactions performed best with 9 when water levels were
below 300 ppm. Higher concentrations of water resulted in a lower
product yield and incomplete starting material conversion.
(16) For these experiments, KF (spray dried material purchased from
Aldrich) was further dried at 120 °C under a flow of nitrogen for 12 h.
The dried material was transferred to a nitrogen drybox and then
milled using a coffee blender.
1053
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054

Organic Process Research & Development
Article
(17) 2,3-Dichloro-6-(trichloromethyl)pyridine feed was initially
∼85% pure (contaminated with other chloropyridine byproducts).
Thus, recrystallization was required to get the purity greater than 95%
for the subsequent steps to prepare isopropyl-5,6-dichloropicolinate.
1054
dx.doi.org/10.1021/op5001258 | Org. Process Res. Dev. 2014, 18, 1045−1054