Tetramethylammonium Fluoride Tetrahydrate for SNAr Fluorination of 4-Chlorothiazoles at a Production Scale

of 4 ‑Chlorothiazoles at a Production Scale

Article

Tetramethylammonium Fluoride Tetrahydrate for S Ar Fluorination

Mai Khanh Hawk,* Sarah J. Ryan,* Xin Zhang, Ping Huang, Jing Chen, Chuanren Liu, Jianping Chen,

Peter J. Lindsay-Scott, John Burnett, Craig White, Yu Lu, and John R. Rizzo

Cite This: https://doi.org/10.1021/acs.oprd.1c00042

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: This article describes the use of tetramethylammonium ﬂuoride tetrahydrate (TMAF·4H O) for the large-scale

preparation of a challenging 4-ﬂuorothiazole. Commercially available TMAF·4H O was procured on a large scale and rigorously

dried by distillation with isopropyl alcohol and then dimethylformamide at elevated temperature. This method of drying provided

anhydrous TMAF [TMAF (anh)] containing <0.2 wt % water and <60 ppm isopropanol. The use of TMAF (anh) was essential for

production of the 4-ﬂuorothiazole. When the chlorothiazole starting material was treated with other anhydrous ﬂuoride sources,

poor conversion of the starting material or potential safety issues were observed. S Ar ﬂuorination using dried TMAF·4H O was

carried out at a 45.1 kg scale at 95−100 °C to produce 36.8 kg of 4-ﬂuorothiazole 1b.

KEYWORDS: ﬂuorination, S Ar, tetramethylammonium ﬂuoride, ﬂuorothiazole, heteroarene, anhydrous ﬂuoride

INTRODUCTION

demonstrated that anhydrous tetrabutylammonium ﬂuoride

TBAF (anh)] could be prepared in situ from tetrabutylam-

monium cyanide (TBACN) and hexaﬂuorobenzene (C F )

■

[

The substitution of hydrogen for ﬂuorine can result in

improved bioavailability or metabolic stability of bioactive

(Scheme 2A). This reagent was used for the room-

temperature ﬂuorination of a variety of electron-deﬁcient aryl

chlorides and nitroarenes. A 4-ﬂuoroimidazole was even

reported in the reaction scope, suggesting that this method

molecules. For this reason, the number of ﬂuorinated

heteroaromatics found in active pharmaceutic ingredients

APIs) and agrochemicals is increasing. Despite the frequency

that ﬂuorine is incorporated into bioactive heteroarenes, large-

scale production of these molecules remains challenging,

particularly under mild, process-friendly conditions.

One of the most common methods for the industrial-scale

(

11c

could be applied to synthetically challenging 4-ﬂuoroazoles.

However, this method requires expensive (C

F ) and toxic

6 6

(

TBACN) as stoichiometric reagents, limiting its feasibility on

an industrial scale. More recently, Sanford reported the room-

temperature ﬂuorination of electron-deﬁcient aryl halides and

nitroarenes using anhydrous tetramethylammonium ﬂuoride

[TMAF (anh)] (Scheme 2B). This reagent is more stable

than other anhydrous tetraalkylammonium salts. Unlike

preparation of heteroaryl ﬂuorides is nucleophilic aromatic

3,4

substitution (S Ar). This reaction involves the substitution

of an electron-deﬁcient aryl halide with a nucleophilic ﬂuoride

source via the Meisenheimer intermediate (Scheme 1). The

most commonly used ﬂuoride source for S Ar ﬂuorination is

TBAF (anh), it cannot undergo problematic Hofmann

an anhydrous alkali metal ﬂuoride. However, the low solubility

of alkali ﬂuorides under anhydrous conditions often results in

the requirement for high temperature and long reaction time.

S_NAr ﬂuorination is particularly challenging when used to

prepare ﬁve-membered azoles with ﬂuorine at the 4 or 5

elimination as it lacks the β-protons required for E2

elimination. For this reason, TMAF (anh) is commercially

available and can be stored in its anhydrous form under

ambient conditions.

We realized that TMAF (anh) would be an attractive

reagent to use for the large-scale production of a 4-

ﬂuorothiazole. However, even though TMAF (anh) is readily

available in gram quantities, it does not have good bulk

commercial availability, limiting its feasibility for use at a large

position. The high π-electron density at these positions

further exacerbates the requirement for forcing reaction

conditions, which can limit functional group tolerance and

lead to the formation of undesired impurities. In fact, there

are relatively few procedures of any kind that have been

reported for the preparation of 4-ﬂuoroazoles. Moreover,

many of these conditions involve the use of expensive

electrophilic ﬂuorinating reagents. Thus, there is a need for

the development of scalable, process-friendly methods for the

production of ﬁve-membered ﬂuoroarenes.

scale. To address this challenge, we have pursued an

approach that involves drying TMAF tetrahydrate (TMAF·

Received: February 4, 2021

Many recent reports have demonstrated that electron-

deﬁcient aromatics can undergo S Ar ﬂuorination under

mild conditions if a soluble anhydrous ﬂuoride source is

0−12

employed.

For example, the pioneering work of DiMagno

XXXX American Chemical Society

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

Article

Scheme 1. S Ar Fluorination Mechanism

Scheme 2. Methods for S Ar Fluorination Using Soluble, Anhydrous Nucleophilic Fluorination

Table 1. S Ar Fluorination of Chlorothiazole 2a and 2b

entry

anhydrous ﬂuoride

CsF

TMAF (anh) (4 equiv)

solvent

temperature (scale)

2 area %

1 area % (yield)

DMSO

DMF

130−135 °C (38 g)

130−135 °C (300 g)

2a 36.8

2a 40.2

2a 100

1a 58.5 (36%)

1a 48.2 (23%)

1a ND

DMF

100 °C

1a 94.4

1b 80.5 (69%)

1a ND

b 12

TMAF (anh) (4 equiv)

DMSO

DMF

100 °C

2a 100

2b 100

2a 100

2a 14.0

2b 5.0

2b 79.8

TBAF (anh)

1a ND

00 °C

TMAOPh + BzF

TMACl + CsF

DMF

DMSO

100 °C

110 °C

1a 75.8

1b 61.1 (32%)

1b 20.1

30 °C

TMACl + CsF

DMF

130 °C

Unless otherwise stated, reactions were completed on 50 mg of chlorothiazole 2a or 2b. Yield adjusted based on the amount of ﬂuorothiazole 1a

in the crude mixture of 1a and 2a. Isolated yield after ﬂash column chromatography. TBAF (anh) was prepared in situ by the reaction of TBACN

and C F .

H O) for the production of a challenging ﬁve-membered

procedure developed by the discovery chemistry group

involved the reaction of the Boc-protected chlorothiazole 2a

with CsF in dimethyl sulfoxide (DMSO) at elevated

temperature (Table 1, entries 1 and 2). Unfortunately, these

reaction conditions were not amenable to large-scale

production. When the reaction was completed with 38 g of

chlorothiazole 2a, the reaction mixture needed to be heated to

heteroaryl ﬂuoride (Scheme 2C). The readily available TMAF·

H O is rigorously dried by distillation with isopropyl alcohol

(

IPA) and then dimethylformamide (DMF) at elevated

temperature to provide TMAF (anh). The S Ar ﬂuorination

using dried TMAF·4H O was carried out at a 45.1 kg scale at

5−100 °C to produce 36.8 kg of aryl ﬂuoride 1b.

30−135 °C for 48 h to achieve greater than 50% conversion

RESULTS AND DISCUSSION

■

(

(Table 1, entry 1). When the scale was increased to 300 g, the

reaction mixture required 60 h of heating to achieve a similar

level of conversion (Table 1, entry 2). The resulting mixture of

We required 35 kg of the ﬁve-membered ﬂuoroarene 1

Scheme 2) as an intermediate for an API. The initial

Organic Process Research & Development

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Article

Figure 1. Scanning reaction calorimetry of the reaction of chlorothiazole 1b with TMACl and CsF in DMSO. The safety data were collected on a

.0 g scale reaction using 4 equiv of TMACl and 10 equiv of CsF in 20 vol of DMSO.

chlorothiazole 2a and ﬂuorothiazole 1a could only be

separated by supercritical ﬂuid chromatography (SFC).

obtain large quantities of TMAF (anh), we began to investigate

methods to prepare it in situ. Unfortunately, when TMAF

(anh) was prepared in situ according to the procedure of

Sanford, using tetramethylammonium phenoxide (TMAOPh)

and benzoyl ﬂuoride (BzF), we did not see any conversion of

Moreover, a signiﬁcant amount of decomposition was observed

under the forcing reaction conditions, resulting in poor mass

balance and low yield (Table 1, entries 1 and 2). These

conditions provided access to hundreds of grams of the target

ﬂuorothiazole for preclinical development. However, the low

conversion of the starting material, poor mass balance, and the

need for SFC puriﬁcation prompted us to investigate alternate

S_NAr ﬂuorination conditions.

We hypothesized that a soluble, anhydrous ﬂuoride source

would facilitate higher conversion under milder reaction

conditions. To test this hypothesis, the chlorothiazoles 2a

and 2b were reacted in the presence of commercially available

TMAF (anh). The reaction did not proceed at room

temperature (Table 1, entry 3). However, when the reaction

mixture was heated to 100 °C, we were delighted to see good

conversion of the starting material and high product area % in

the high-performance liquid chromatography (HPLC) chro-

matogram (Table 1, entry 4). DMF was selected as a solvent as

DMSO did not provide good conversion of the starting

material (Table 1, entry 5). Interestingly, when TBAF (anh)

was used as the ﬂuoride source under the same reaction

conditions, we did not see any conversion of the starting

material (Table 1, entry 6). We propose that this is due to

instability of TBAF (anh) at the elevated temperature required

12b

the starting material (Table 1, entry 7). We were initially

excited to ﬁnd moderate to good conversion of the starting

material when tetramethylammonium chloride (TMACl) was

used as a superstoichiometric phase-transfer reagent in the

presence of CsF as the stoichiometric ﬂuoride source in

7b,14b

DMSO (Table 1, entry 8).

However, reaction calorimetry

of these conditions indicated that the safety proﬁle of this

reaction was not desirable for scaleup (Figure 1). The reaction

mixture was heated in a high-pressure Hastelloy cell from room

temperature to 250 °C with a temperature increase of 0.5 °C/

min. Two exothermic peaks were observed: the ﬁrst peak (with

an onset temperature of 100 °C) is likely to be the heat of the

reaction of the desired reaction, and the second peak (onset

temperature of 193 °C) indicates the thermal decomposition

of the reaction mixture. Decomposition of DMSO is likely to

be accelerated under these conditions due to the reaction with

anhydrous ﬂuoride with the solvent to produce the dimsyl

anion and hydrogen ﬂuoride (HF). Deprotonation of DMSO

has previously been demonstrated to occur at 90 °C in the

presence of TMAF (anh), leading to the formation of benzyne

intermediates. It is well known that DMSO decomposes

7c,d

for thiazole S Ar ﬂuorination. This reagent can undergo

violently, and its decomposition is autocatalytic.

The onset

competing Hofmann elimination with the very basic ﬂuoride

anion in its anhydrous form, leading to the formation of

temperature of autocatalytic degradation of DMSO is lowered

when in the presence of the dimsyl anion and in the presence

biﬂuoride (HF ), which is signiﬁcantly less nucleophilic than

of HF. The thermal stability of DMSO is also known to

ﬂuoride.

decrease over time in the presence of halides. Therefore, it

Having established that TMAF (anh) could lead to full

conversion of the starting material at the lab scale, we turned

our attention to large-scale production. It became quickly

was not recommended to hold this reaction mixture in DMSO

at high temperature at a large scale.

In an attempt to improve the safety proﬁle of the TMACl

and CsF conditions, DMF was trialed as a solvent.

evident that TMAF (anh) could not be procured in bulk. To

Organic Process Research & Development

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Article

Figure 2. Scanning reaction calorimetry of the reaction of chlorothiazole 1b with TMAF (anh) in DMF. The safety data were collected on a 1.3 g

scale reaction using 6 equiv of TMAF in 24 vol of DMF.

Table 2. Methods for Drying TMAF·4H O

entry

drying solvent

reaction solvent

water content wt % [molar equiv]

2b area %

1b area % (isolated yield)

3 area %

none (TMAF·H O)

DMF

4.9% [7.0]

0.50% [0.7]

0.33% [0.5] (0.25% IPA)

7.8% [11] (4.4% IPA)

0.16% [0.2] (<60 ppm IPA)

0.9

65%

79 (40%)

17%

toluene

IPA

DMF

All reactions were conducted on a 2 g scale using 6 equiv TMAF·4H O. Reaction concentration after removal of the drying solvent is 24 mL/g

DMF/TMAF. Water content (wt %) remaining in the reaction mixture after removal of the drying solvent by distillation. Molar equivalents

relative to anhydrous TMAF. TMAF was isolated by crystallization from THF after azeotropic drying with IPA. Reaction was not completed due

to the high water content and residual IPA. Residual IPA was removed by distillation with DMF after azeotropic drying with IPA.

Unfortunately, this resulted in low conversion of the starting

material (Table 1, entry 9), presumably due to the lower

solubility of the CsF in DMF compared to DMSO (dielectric

constant DMSO = 47.0, DMF = 38.3). Fortuitously, the safety

proﬁle of TMAF (anh) in DMF indicated that this reagent was

suitable for large-scale production (Figure 2). As such, we

focused our attention on drying commercially available TMAF·

As expected, when TMAF·4H O was used “as is” without

any drying procedure in place, there was no conversion of the

starting material (Table 2, entry 1). Our initial attempt to drive

oﬀ the water bound to the TMAF involved exploration of

azeotropic distillation, followed by solvent swap to DMF to

conduct the S Ar ﬂuorination. The use of toluene as an

azeotropic solvent resulted in only 9 area % of the desired

product 2a (Table 2, entry 2). We suspect that the low yields

were due to residual water that remained in the TMAF

reaction solution after these drying methods were employed.

H O to generate large amounts of TMAF (anh) for the

production-scale S Ar ﬂuorination.

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

Article

Karl Fischer (KF) analysis of the solution of TMAF in DMF

that was obtained after solvent swap indicated that 0.50 wt %

water was present in the DMF reaction solution when toluene

was used as the azeotropic distillation solvent. To improve the

yield of the product, a more eﬃcient drying method needed to

controlled to <60 ppm by distillation with DMF (Table 2,

entry 5). After distillation with DMF, the water content also

remained very low (0.16%). This drying method provided

99.1% conversion of the starting material and 79 area % of the

desired product when analyzed by HPLC. The ﬂuorothiazole

was isolated in 40.0% yield.

be implemented. A crystal structure of TMAF·4H O obtained

by McLean and Jeﬀrey indicates that the water molecules are

We were delighted that TMAF·4H O could be successfully

very tightly bound in the TMAF·4H O crystal lattice via H-

dried on a lab scale. We next turned our attention to large-scale

production. Chlorothiazole 2b was selected for scaleup due to

ease in the downstream chemistry. This substrate was prepared

by cyclization of thiourea with ethyl chloroacetate to provide

the amino thiazole 4 (Scheme 3). Vilsmeier haloformylation

bonding-type interactions with ﬂuoride (Figure 3 ). We

with POCl and DMF then aﬀorded chlorothiazole 5, which

was acylated with acetic anhydride to provide 139.2 kg of the

target chlorothiazole 2b in 97.3% purity and 98.1 wt % assay

potency.

With chlorothiazole 2b in hand, the S Ar ﬂuorination was

completed with 45.1 kg of chlorothiazole. As such, 285.1 kg of

TMAF·4H O was transferred to a 2000 L reactor and 1375 kg

of IPA was added (Figure 4). The mixture was concentrated 6

times with IPA added back until the water content was

measured to be <0.2 wt %. The water content was checked

after the fourth distillation to be 0.5%, and after the sixth

distillation, the water content was 0.12%. These results are in

agreement with lab-scale experiments, which indicated that at

least four IPA distillation/add-backs were required to achieve

the desired water wt % (Chart 1). DMF was then added, and

the mixture was concentrated until the IPA content was

measured to be <60 ppm. This required multiple days under

high vacuum (<−0.09 MPa). Chlorothiazole 2b was then

added to the TMAF (anh)/DMF solution, and the mixture

was heated to 95−105 °C for 2 h to provide 36.8 kg of

ﬂuorothiazole in 70.4% purity (50.1 wt % purity) and 44%

yield. Two major impurities were characterized as remaining

chlorothiazole 2b (3 area %) and dimethylamino thiazole 6 (3

area %). The formation of dimethylamine 6 was not observed

in the lab-scale experiments. This suggested that DMF was

decomposing to dimethylamine during the prolonged dis-

tillation time required at a larger scale. The purity of the

ﬂuorothiazole was successfully upgraded by conversion to the

bisulﬁte adduct. The low mass balance is attributed to the

formation of unknown decomposition products throughout the

reaction progress. We suspect that the starting chloro-thiazole

and the ﬂuoro-thiazole undergo decomposition during the

reaction.

Figure 3. TMAF·4H O complex. Reproduced from ref 17, with

permission from AIP publishing.

expect that these interactions would still be present in solution.

To drive oﬀ the hydrate water molecules, we propose that they

could be exchanged with an alcohol to produce an alcohol

3,18

solvate of TMAF.

In 1990, Christe demonstrated that

TMAF (anh) could be isolated as a crystalline solid from the

TMAF isopropanol solvate by removal of the solvated

alcoholate under dynamic vacuum at elevated temperature.

To avoid the need for isolation and drying, we propose that

solvent exchange of the alcohol solution containing the TMAF·

ROH with DMF would then allow removal of the alcohol to

produce TMAF (anh). To test this hypothesis, we turned to

drying the TMAF·4H O with isopropanol (Table 2, entry 3).

The resulting TMAF isopropanol complex was then solvent-

exchanged to DMF to provide TMAF (anh). To our delight,

this reagent allowed successful conversion of the chlorothiazole

to provide the desired product (Table 2, entry 3). However, a

major impurity had formed under these reaction conditions.

This was identiﬁed as the isopropyl ether 3. To control this

impurity, the amount of residual IPA also needed to be

controlled. Our initial attempt to control the IPA was by

CONCLUSIONS

isolation of the dried TMAF·H O via crystallization from

■

tetrahydrofuran (THF). Unfortunately, in our hands, the

isolation of TMAF (anh) resulted in an increase in the water

content in the DMF reaction solution to 7.8% (Table 2, entry

In conclusion, we have developed a method for drying readily

available TMAF·4H O for the large-scale ﬂuorination of a

challenging ﬁve-membered heteroaryl ﬂuoride. TMAF·4H O

). We suspect that this is due to the handling of the extremely

was procured in bulk from Changzhou Huadong Chemical

Research Institute. The reagent was rigorously dried by

hydroscopic solid. Gratifyingly, the IPA content could be

Scheme 3. Synthesis of Chlorothiazole 2b

Organic Process Research & Development

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Article

Figure 4. Reaction ﬂow scheme for the production of ﬂuorothiazole 1b. Note: Repeat 4−5 until the KF of the mixture is below 0.2% and repeat 6−

until the IPA content is below 60 ppm.

Chart 1. wt % of Water after Five IPA Distillations/Add-

Back on a Lab Scale

TMAF·4H O was carried out at a 45.1 kg scale at 95−100 °C

to produce 36.8 kg of 4-ﬂuorothiazole 1b.

EXPERIMENTAL SECTION

■

General Experimental Section. Purchased starting

materials and reagents were used without further puriﬁcation.

HPLC analysis (IPC and purity testing) were performed on a

Shimadzu LC-20A instrument using a Waters XBridge C18

(

75 mm × 4.6 mm, 3.5 μm) column. Liquid chromatography/

mass spectrometry (LC/MS) analysis was performed on an

Agilent HPLC 1200 with a 6120 Quadrupole LC/MS detector.

H NMR (400 MHz) and C (100 MHz) spectra were

recorded on a Bruker Ultrashield 400, Avance II 400. Reaction

calorimetry studies were performed using a Setaram C80

calorimeter with a Hastelloy high-pressure cell.

Step 1: Preparation of 2-Aminothiazol-4-ol Hydrochloride

. Thiourea (77.0 kg, 101 mol) and ethanol (250.4 kg, 4.12 L/

kg) were charged into a 2000 L reactor, and the temperature

was adjusted to 70−80 °C. Ethyl chloroacetate (127.5 kg, 104

mol, 1.03 equiv) was added dropwise below 80 °C, and the

reaction mixture was stirred at 75−80 °C for 2−3 h. The

reaction mixture was next cooled to 40−50 °C, and a sample

was pulled for analysis to assure the reaction was complete.

The temperature was adjusted to 5−10 °C, and the mixture

was stirred for 1−2 h. The precipitate was collected by

ﬁltration, and the ﬁlter cake was washed with ethanol and dried

under vacuum at 50−55 °C for 10−12 h to provide 137.5 kg of

thiazole 4 as the hydrochloride salt in 88.3% yield (99.7%

distillation with IPA and then DMF at elevated temperature to

provide TMAF (anh) containing <60 ppm water and <60 ppm

isopropanol. The use of this reagent addressed the low

conversion of the starting material and product decomposition

that occurred when CsF was employed. The reaction

conditions that included 6 equiv of dried TMAF·4H O in 24

vol of DMF were monitored by reaction calorimetry. These

conditions were determined to be safe with no exothermic

events or pressure buildup. The S Ar ﬂuorination using dried

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

Article

purity, 99.1 wt % potency). H NMR (400 MHz, DMSO-d ,

was added back into the reactor to provide a ﬁnal

ppm): δ 10.95 (s, 1H), 9.65 (s, 2H), 4.11 (s, 2H); C NMR

400 MHz, DMSO-d , ppm): δ 178.70, 177.60, 37.23. ESI-MS

m/z: C H N OS [M + H] , 117.0123; found, 117.0121.

Step 2: Preparation of 2-Amino-4-chlorothiazole-5-

carbaldehyde 5. DMF (478.8 kg, 509 L, 7.5 L/kg), with a

water content of 0.03 wt %, was charged into a 2000 L reactor.

The solution was cooled to −5 to 2 °C, and POCl (204.1 kg,

33 mol, 2.3 equiv) was added dropwise, maintaining the

temperature below 5 °C. The reaction mixture was stirred at

5 to 2 °C for 0.5 h. Thiazole 4 (67.5 kg, 58 mol) was next

concentration of 6.1 L/kg relative to TMAF·4H O. The

(

temperature was adjusted to 40 °C, and the mixture was stirred

for approximately 20 min until all of the solid had dissolved.

The mixture was concentrated to a ﬁnal concentration of 300−

360 L, maintaining the temperature below 70 °C. This

azeotropic distillation (IPA add back and concentration)

procedure was repeated several times until KF analysis

indicated that the water content was <0.20%. Once the

desired water content was achieved, DMF (822 kg, 900 L, 3.15

−

L/kg relative to TMAF·4H O) was charged into the reaction

added, maintaining the temperature below 20 °C. Once the

addition was complete, the reaction mixture was heated to 50−

mixture. The mixture was concentrated to a ﬁnal concentration

of 300−360 L, maintaining the temperature below 90 °C. This

distillation (DMF addition and concentration) was repeated

until the residual IPA was measured to be <60 ppm. Once the

desired IPA content was achieved, the reaction temperature

was adjusted to 25−30 °C. Chlorothiazole 2b (45.1 kg, 22

mol) was charged into the reactor under nitrogen protection,

followed by DMF (442.0 kg, 451 L, 10 L/kg). The reaction

mixture was stirred at 25−30 °C for 30 min, and then, the

temperature was increased to 95−105 °C. The reaction

mixture was stirred at this temperature for 2−4 h before

cooling to 45−50 °C. A sample was pulled for analysis to

assure complete conversion. Once complete conversion had

been achieved, the reaction mixture was adjusted to 20−30 °C.

2-MeTHF (738.0 kg, 868 L, 19 L/kg) was charged into the

5 °C and stirred at this temperature for 17−18 h. A sample

was pulled for analysis to assure that the reaction was

complete. Once the reaction was complete, the temperature

was reduced to 25−30 °C and dichloromethane (DCM)

(

223.4 kg, 168 L, 2.5 L/kg) was added, followed by the

addition of water (1822 kg, 27 L/kg). The temperature was

adjusted to 0−5 °C, and the mixture was transferred to a 3000

L reactor, maintaining the temperature below 15 °C. The

temperature was next adjusted to 30−35 °C, and the mixture

was stirred for 16−18 h before cooling to 15−20 °C and

centrifuging the mixture. The wet cake was dried under

vacuum at 55−60 °C for 18−19 h to provide 60.45 kg of

chlorothiazole 5 in 76.8% yield (93.0% purity, 92.6 wt %

potency). H NMR (400 MHz, DMSO-d , ppm): δ 9.64 (s,

H), 8.74 (s, 1H); C NMR (400 MHz, DMSO-d , ppm): δ

79.90, 173.01, 147.32, 117.82. ESI-MS m/z: C H ClN OS

reactor, followed by charging of aqueous NH Cl (1362.0 kg,

[

19 L/kg). The reaction temperature was adjusted to 25−30 °C

and stirred at this temperature for 30−40 min. The layers were

separated, and the aqueous phase was extracted with 2-

MeTHF (730 kg, 858 L, 19 L/kg). The organic phases were

M + H] , 162.9733; found, 162.9737.

Step 3: Preparation of the Compound N-(4-Chloro-5-

formylthiazol-2-yl)acetamide 2b. Chlorothiazole 5 (109.21

kg, 67 mol) and DMSO (483 kg, 440 L, 4 L/kg) were charged

into a 3000 L reactor at 12−15 °C. Tetraethylammonium

combined and washed twice with saturated NH Cl (914.0 kg ×

2). The combined organic phase was concentrated to a ﬁnal

concentration of 1−2 L/kg, maintaining the temperature

below 55 °C. DCM (248 kg) was added, maintaining the

temperature below 55 °C, and then, n-heptane (282 kg, 414 L,

9.2 L/kg) was added. The crystallization temperature was

adjusted to 20−25 °C, and the mixture was stirred at this

temperature for 1−2 h. The precipitate was collected by

ﬁltration, and the ﬁlter cake was rinsed with n-heptane (26.0

kg, 38 L, 0.8 L/kg). The wet cake was dried under vacuum at

55−65 °C for 10−24 h to provide 36.8 kg of ﬂuorothiazole 1b

(

TEA) (281.4 kg, 278 mol, 4.15 equiv) was next added into

the reactor, maintaining the temperature below 20 °C. Once

the addition was complete, the temperature was adjusted to 5−

0 °C. Acetic anhydride (209.6 kg, 205 mol, 3.06 equiv) was

then added dropwise, maintaining the temperature below 10

C, and the reaction mixture was stirred at 15−20 °C for 2−3

h. A sample was pulled for analysis to assure the reaction had

reached completion. Once the reaction was complete, the

reaction mixture was cooled to 0−5 °C. Water was added

dropwise, maintaining the temperature below 40 °C. Once the

addition was complete, the temperature was adjusted to 5−10

in 44% yield (70.4% purity, 50.1 wt % potency). H NMR

(400 MHz, DMSO-d , ppm): δ 13.00 (1H, COHs), 9.87 (1H,

s, CONH), 2.22 (3H, s, CH); C NMR (400 MHz, DMSO-

d₆, ppm): δ 180.80, 171.09, 164.18, 162.10, 106.47, 23.22; F

C, and the reaction mixture was stirred at this temperature for

−8 h. The precipitate was collected by ﬁltration and washed

NMR (400 MHz, DMSO-d , ppm): δ 98.79. ESI-MS m/z:

with H O (596 kg, 8.9 L/kg). The wet cake was dried under

C H FN O S [M + H] , 189.0140.

vacuum at 60−65 °C for 20−24 h to provide 139.2 kg of

Preparation of Compound N-(4-Chloro-5-formylthiazol-

chlorothiazole 2b in 99% yield (98.0% purity, 98.1 wt %

2-yl)pivalamide 2a. Chlorothiazole 5 (21.3 g, 123 mmol) and

DCM (200 mL, 9.4 vol) were charged into a three-neck round-

bottom ﬂask. TEA (19.7 mL, 144 mol, 1.2 equiv) and 4-

dimethylaminopyridine (0.60 g, 4.9 mmol, 0.04 equiv) were

added. Boc-anhydride (26.8 g, 122.8 mmol, 1.0 equiv) was

then added, and the reaction temperature was adjusted to 15

°C. The reaction mixture was stirred for 16 h. After this time, a

sample was pulled for analysis to assure that the reaction had

reached completion. Once complete conversion was achieved,

water was added. The layers were separated, and the organic

layer was ﬂushed through a silica gel plug. The organic layer

was then concentrated to dryness, and the resulting solid was

dried under vacuum at 40 °C for 42 h to provide ﬂuorothiazole

potency). H NMR (400 MHz, DMSO-d , ppm): δ 13.00 (1H,

COHs), 9.90 (1H, s, CONH), 2.22 (3H, s, CH); C NMR

400 MHz, DMSO-d , ppm): δ 182.81, 170.76, 163.40, 144.90,

23.03, 23.18. ESI-MS m/z: C H ClN O S [M + H] ,

04.9839; found, 204.9845.

Step 4: Preparation of N-(4-Fluoro-5-formylthiazol-2-

(

6 5 2 2

yl)pivalamide 1b. TMAF·4H O (285.1 kg, 172 mol, 7.8

equiv) and IPA (1375 kg, 1740 L, 6.1 L/kg relative to TMAF·

H O) were charged into a 3000 L reactor. The temperature

was adjusted to 40 °C, and the mixture was stirred for

approximately 20 min until all of the solid had dissolved. The

mixture was then concentrated to a ﬁnal concentration of

00−360 L, maintaining the temperature below 70 °C. IPA

2a (14.42 g) was in 99% yield (99.4% purity). H NMR (400

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

Complete contact information is available at:

Article

MHz, CDCl , ppm): δ 9.97 (1H, s, CHO), 1.567 (9H, s,

Peter J. Lindsay-Scott − Eli Lilly and Company Limited,

Surrey GU20 6PH, U.K.; orcid.org/0000-0002-4721-

3594

John Burnett − Eli Lilly and Company Limited, Surrey GU20

6PH, U.K.

CO CCH ); C NMR (400 MHz, CDCl , ppm): δ 181.8,

66.2, 151.8, 145.0, 123.8, 84.7, 28.7. ESI-MS m/z:

C H ClN O S [M + 1] , 263.02.

Preparation of N-(4-Fluoro-5-formylthiazol-2-yl)-

pivalamide 1a. TMAF·4H O (32.5 g, 200 mmol, 10.5

Craig White − Eli Lilly and Company Limited, Surrey GU20

6PH, U.K.

equiv) was charged into a three-neck round-bottom ﬂask

followed by IPA (200 mL, 6.1 mL/g relative to TMAF·4H O).

Yu Lu − Small Molecule Design and Development (SMDD),

Lilly Technology Center North, Eli Lilly and Company,

Indianapolis, Indiana 46221, United States; orcid.org/

0000-0002-6411-9954

John R. Rizzo − Small Molecule Design and Development

(SMDD), Lilly Technology Center North, Eli Lilly and

Company, Indianapolis, Indiana 46221, United States

The reaction mixture was concentrated and azeotropically

distilled with IPA 10 times until the water content was

measured to be <0.2% (by KF analysis). DMF (200 mL, 6.1

mL/g) was next added to the reaction mixture. The mixture

was concentrated with DMF added back 10 times until the IPA

concentration was measured to be <60 ppm. The TMAF/

DMF solution was adjusted to provide a ﬁnal concentration of

https://pubs.acs.org/10.1021/acs.oprd.1c00042

0 mL/g DMF relative to compound 2a (100 mL).

Compound 2a (5.0 g, 19.0 mmol) was charged into the

ﬂask. The reaction mixture was heated to 100 °C for 2 h and

then cooled to 25 °C. 2-MeTHF (100 mL, 20 mL/g) was

Author Contributions

The manuscript was written through contributions of all

authors. All authors have given approval to the ﬁnal version of

the manuscript.

added, followed by addition of saturated NH Cl (150 mL, 30

mL/g). The mixture was stirred for 17 h, and then, the layers

were separated. The organic layer was concentrated to dryness,

and the residue was puriﬁed via ﬂash column chromatography

Notes

The authors declare no competing ﬁnancial interest.

(

2:1 petroleum ether/EtOAc). Fractions containing the

desired product were combined and concentrated to provide

ﬂuorothiazole 1a (1.4 g) as a yellow solid in 30% isolated yield

ACKNOWLEDGMENTS

■

(

96.2% purity). H NMR (400 MHz, CDCl , ppm): δ 9.933

The authors would like to acknowledge Dr. Kevin Cole for his

help in reviewing this manuscript.

1H, s, CHO), 1.567 (9H, s, CCH ); C NMR (400 MHz,

CDCl , ppm): δ 179.2, 164.9, 164.7, 164.1, 161.4, 151.7, 107.2,

07.1, 84.7. ESI-MS m/z: C H FN O S [M + 1] , 247.05.

9 11 2 3

ABBREVIATIONS

■

API, active pharmaceutical ingredient; BzF, benzoyl ﬂuoride;

KF, Karl Fischer; MF, metal ﬂuoride; NMR, nuclear magnetic

resonance; SFC, supercritical ﬂuid chromatography; S AR,

nucleophilic aromatic substitution; TBACN, tetrabutylammo-

nium cyanide; TBAF, tetrabutylammonium ﬂuoride; TMACl,

tetramethylammonium chloride; TMAF (anh), anhydrous

ASSOCIATED CONTENT

sı Supporting Information

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.oprd.1c00042.

■

¹H, ¹³C, and ¹⁹F NMR spectra of compounds 4, 5, 2a,

2b, 1a, and 1b and the development of preparative SFC

tetramethylammonium ﬂuoride; TMAF·4H O, tetramethylam-

monium ﬂuoride tetrahydrate; TMAOPh, tetramethylammo-

conditions (PDF)

nium phenoxide

AUTHOR INFORMATION

Corresponding Authors

■

REFERENCES

■

ller, K.; Faeh, C.; Diederich, F. Fluorine in

1) (a) Mu

Mai Khanh Hawk − Small Molecule Design and Development

Pharmaceuticals: Looking Beyond Intuition. Science 2007, 317,

(

SMDD), Lilly Technology Center North, Eli Lilly and

881 −1886. (b) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur,

V. Fluorine in Medicinal Chemistry. Chem. Soc. Rev. 2008 , 37 , 320 −

30. (c) Jeschke, P. The Unique Role of Halogen Substituents in the

Design of Modern Agrochemicals. Pest Manage. Sci. 2010, 66, 10−27.

d) Gillis, E. P.; Eastman, K. J.; Hill, M. D.; Donnelly, D. J.; Meanwell,

Company, Indianapolis, Indiana 46221, United States;

Email: hawk_mai_khanh_n@lilly.com

Sarah J. Ryan − Small Molecule Design and Development

(

SMDD), Lilly Technology Center North, Eli Lilly and

(

Company, Indianapolis, Indiana 46221, United States;

orcid.org/0000-0002-9633-5299;

N. A. Applications of fluorine in medicinal chemistry. J. Med. Chem.

2015, 58, 8315−8359. (e) Wang, B.-C.; Wang, L.-J.; Jiang, B.; Wang,

S.-Y.; Wu, N.; Li, X.-Q.; Shi, D.-Y. Application of Fluorine in Drug

Design During 2010 −2015 Years: A Mini Review. Mini-Rev. Med.

Chem. 2017 , 17 , 683 −692. (f) Kirk, K. L. Fluorination in Medicinal

Chemistry: Methods, Strategies, and Recent Developments. Org.

Process Res. Dev. 2008, 12, 305−321.

Email: Ryan_sarah_j@lilly.com

Authors

Xin Zhang − STA Pharmaceutical (WuXi STA), WuXi

AppTec., Shanghai 200131, People’s Republic of China

Ping Huang − STA Pharmaceutical (WuXi STA), WuXi

AppTec., Shanghai 200131, People’s Republic of China

Jing Chen − STA Pharmaceutical (WuXi STA), WuXi

AppTec., Shanghai 200131, People’s Republic of China

Chuanren Liu − STA Pharmaceutical (WuXi STA), WuXi

AppTec., Shanghai 200131, People’s Republic of China

Jianping Chen − STA Pharmaceutical (WuXi STA), WuXi

AppTec., Shanghai 200131, People’s Republic of China

(

2) (a) Zhou, Y.; Wang, J.; Gu, Z.; Wang, S.; Zhu, W.; Acena, J. L.;

Soloshonok, V. A.; Izawa, K.; Liu, H. Next Generation of Fluorine-

Containing Pharmaceuticals, Compounds Currently in Phase II −III

Clinical Trials of Major Pharmaceutical Companies: New Structural

Trends and Therapeutic Areas. Chem. Rev. 2016 , 116 , 422 −518.

b) Ilardi, E. A.; Vitaku, E.; Njardarson, J. T. Data-Mining for Sulfur

and Fluorine: An Evaluation of Pharmaceuticals to Reveal

Opportunities for Drug Design and Discovery. J. Med. Chem. 2014 ,

57, 2832−2842. (c) O’Hagan, D. Fluorine in Health Care:

Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

H Amination of Imidazopyridines with N -Fluorobenzenesulfonimide.

Article

Organofluorine Containing Blockbuster Drugs. J. Fluorine Chem.

J. Org. Chem. 2018 , 83 , 13991 −14000. (c) Albertshofer, K.; Mani, N.

S. Regioselective Electrophilic Fluorination of Rationally Designed

Imidazole Dericatives. J. Org. Chem. 2016 , 81 , 1269 −1276. (d) Hat-

field, J. M.; Eidell, C. K.; Stephens, C. E. Mono-and Trifluorination of

the Thiazole Ring of 2,5-diarylthiazoles using N -Fluorobenzenesulfo-

nimide (NFSI). Tetrahedron Lett. 2013 , 54 , 1025 −1028. (e) Yamada,

S.; Gavryushin, A.; Knochel, P. Convenient Electrophilic Fluorination

of Functionalized Aryl and Heteroaryl Magnesium Reagents. Angew.

Chem., Int. Ed. 2010, 49, 2215−2218.

(

010, 131, 1071−1081.

3) (a) Langolis, B.; Gilbert, L.; Forat, G. Fluorination of Aromatic

Compounds by Halogen Exchange with Fluoride Anions (“Halex”

Reaction). In Industrial Chemistry Library; Jean-Roger, D., Serge, R.,

Eds.; Elsevier: New York, 1996; pp 244−292. (b) Clark, J. H.; Wails,

D.; Bastock, T. W. Aromatic Fluorination ; CRC Press: Boca Raton, FL,

996. (c) Adams, D. J.; Clark, J. H. Nucleophilic Routes to Selectively

Fluorinated Aromatics. Chem. Soc. Rev. 1999, 28, 225−231.

d) Siegemund, G.; Schwertfeger, W.; Feiring, A.; Smart, B.; Behr,

(

F.; Vogel, H.; McKusick, B.; Kirsch, P. Fluorine Compounds, Organic.

In Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH:

Weinheim, 2016. (e) Furuya, T.; Klein, J. E.; Ritter, T. C −F Bond

Formation for the Synthesis of Aryl Fluorides. Synthesis 2010, 1804−

(10) (a) Allen, L. J.; Muhuhi, J. M.; Bland, D. C.; Merzel, R.;

Sanford, M. S. Mild Fluorination of Chloropyridines with in Situ

Generated Anhydrous Tetrabutylammonium Fluoride. J. Org. Chem.

2014, 79, 5827−5833. (b) Ryan, S. J.; Schimler, S. D.; Bland, D. C.;

Sanford, M. S. Acyl Azolium Fluorides for Room Temperature

Nucleophilic Aromatic Fluorination of Chloro- and Nitroarenes. Org.

Lett. 2015, 17, 1866−1869.

(11) (a) Sun, H.; DiMagno, S. G. Anhydrous Tetrabutylammonium

Fluoride. J. Am. Chem. Soc. 2005 , 127 , 2050 −2051. (b) Sun, H.;

DiMagno, S. G. Room-Temperature Nucleophilic Aromatic Fluori-

nation: Experimental and Theoretical Studies. Angew. Chem., Int. Ed.

821.

4) (a) Finger, G. C.; Kruse, C. W. Aromatic Fluorine Compounds.

VII. Replacement of Aromatic −Cl and −NO2 Groups by −F1,2. J.

Am. Chem. Soc. 1956 , 78 , 6034 −6037. (b) Boechat, N.; Clark, J. H.

Fluorodenitrations using Tetramethylammonium Fluoride. J. Chem.

Soc., Chem. Commun. 1993 , 921 −922. (c) Gottlieb, H. B. The

Replacement of Chlorine by Fluorine in Organic Compounds. J. Am.

Chem. Soc. 1936 , 58 , 532 −533. (d) Kuduk, S. D.; DiPardo, R. M.;

Bock, M. G. Tetrabutylammonium Salt Induced Denitration of

Nitropyridines: Synthesis of Fluoro-, Hydroxy-, and Methoxypyr-

idines. Org. Lett. 2005, 7, 577−579.

006 , 45 , 2720 −2725. (c) Sun, H.; DiMagno, S. G. Fluoride Relay: A

New Concept for the Rapid Preparation of Anhydrous Nucleophilic

Fluoride Salts From KF. Chem. Commun. 2007, 528−529.

(12) (a) Schimler, S. D.; Ryan, S. J.; Bland, D. C.; Anderson, J. E.;

Sanford, M. S. Anhydrous Tetramethylammonium Fluoride for

Room-Temperature SNAr Fluorination. J. Org. Chem. 2015, 80,

(

5) (a) Arimoto, S.; Kubota, M.; Kurahashi, M.; Ohira, D. Amide

Compound and use Thereof. Japanese Patent JP 2011001353 A, Jan

, 2011. (b) Uchibori, Y.; Umeno, M.; Yoshioka, H. Nucleophilic

2137 −12145. (b) Cismesia, M. A.; Ryan, S. J.; Bland, D. C.; Sanford,

M. S. Multiple Approaches to the In Situ Generation of Anhydrous

Tetraalkylammonium Fluoride Salts for SNAr Fluorination Reactions.

J. Org. Chem. 2017, 82, 5020−5026.

fluorination of chlorinated N-Heterocycles with Tetrabutylphospho-

nium Hydrogendifluoride and dihydrogentrifluoride. Heterocycles

992, 34, 1507−1510. (c) Forster, H.; Beck, G.; Klauke, E.; Santel,

13) (a) Christe, K. O.; Wilson, W. W.; Wilson, R. D.; Bau, R.; Feng,

H.-J.; Schmidt, R. R. Novel 4,5-di-substituted 1,3-thiazol-2-yloxyace-

J. A. Synthesis, Preparation, and Structures and Anhydrous

Tetramethylammonium Fluorinde and its 1:1 Adduct with Trans-3-

amino-2-butenenitrile. J. Am. Chem. Soc. 1990, 112, 7619 −7625.

tamide herbicides. U.S. Patent 4,929,743 A, May 29, 1990.

(

6) Metzger, J. V. 4.19Thiazoles and their Benzo Derivatives. In

Comprehensive Heterocyclic Chemistry; Katritzky, A. R., Rees, C. W.,

Eds.; Pergamon, 1984; pp 235−331.

(14) (a) Cox, D. P.; Terpinski, J.; Lawrynowicz, W. “Anhydrous ”

Tetrabutylammonium Fluoride: A Mild but Highly Efficient Source of

7) (a) Grushin, V. V.; Marshall, W. J. Fluorination of Nonactivated

Nucleophilic Fluoride Ion. J. Org. Chem. 1984, 49, 3216−3219.

Haloarenes via Arynes under Mild Conditions, Resulting from Further

Studies toward Ar −F Reductive Elimination from Palladium(II).

Organometallics 2008, 27, 4825−4828. (b) Allen, L. J.; Lee, S. H.;

Cheng, Y.; Hanley, P. S.; Muhuhi, J. M.; Kane, E.; Powers, S. L.;

Anderson, J. E.; Bell, B. M.; Roth, G. A.; Sanford, M. S.; Bland, D. C.

Developing Efficient Nucleophilic Fluorination Methods and

Application to Substituted Picolinate Esters. Org. Process Res. Dev.

(b) Elias, S.; Karton-Lifshin, N.; Yehezkel, L.; Ashkenazi, N.;

Columbus, I.; Zafrani, Y. Synthesis, Characterization, and Reactivity

of Thermally Stable Anhydrous Quaternary Ammonium Fluorides.

Org. Lett. 2017, 19, 3039−3042. Hong, C. M.; Xu, Y.; Chung, J. Y. L.;

Schultz, D. M.; Weisel, M.; Varsolona, R. J.; Zhong, Y.-L.; Purohit, A.

K.; He, C. Q.; Gauthier, D. R.; Humphrey, G. R.; Maloney, K. M.;

Lévesque, F.; Wang, Z.; Whittaker, A. M.; Sirota, E.; McMullen, J. P.

Development of a Commercial Manufacturing Route to 2-

Fluoroadinine, The Key Unnatural Nucleobase of Islatravir. Org.

Process Res. Dev. 2021, 25, 395−404.

014, 18, 1045−1054. (c) Yang, Q.; Sheng, M.; Li, X.; Tucker, C.;

Vásquez Céspedes, S.; Webb, N. J.; Whiteker, G. T.; Yu, J. Potential

Explosion Hazards Associated with the Autocatalytic Thermal

Decomposition of Dimethyl Sulfoxide and its Mixtures. Org. Process

Res. Dev. 2020, 24, 916−939. (d) Deguchi, Y.; Kono, M.; Koizumi, Y.;

Izato, Y.-i.; Miyake, A. Study on Autocatalytic Decomposition of

Dimethyl Sulfoxide (DMSO). Org. Process Res. Dev. 2020, 24, 1614−

(15) The largest bulk source of TMAF (anh) was 15 kg with a cost

of $15,000/kg. In comparison, the bulk cost of TMAF·4H O is $70/

kg to provide a total amount of 800 kg of material.

(

16) See Supporting Information for SFC puriﬁcation development

and conditions.

17) McLean, W. J.; Jeffrey, G. A. Crystal Structure of

Tetramethylammonium Fluoride Tetrahydrate. J. Chem. Phys. 1967,

7, 414−417.

18) For TBAF alcohol complexes and their reactivity see; Engle, K.

8) (a) Milner, P. J.; Yang, Y.; Buchwald, S. L. In-Depth Assessment

620.

of the Palladium-Catalyzed Fluorination of Five-Membered Heteraryl

Bromides. Organometallics 2015, 34, 4775−4780. (b) Fluorine in

Heterocyclic Chemistry: 5-Membered Heterocycles and Macrocycles;

Nenajdenko, V., Ed.; Springer: New York, 2014; Vol. 1. (c) Kirk,

K. L. Fluorinated Five-Membered Nitrogen-Containing Heterocycles.

In Fluorinated Heterocyclic Compounds: Synthesis, Chemistry, and

Applications; Petrov, V. A., Ed.; Wiley: Hoboken, NJ, 2009; pp 91−

(

M.; Pfeifer, L.; Pidgeon, G. W.; Giuffredi, G. T.; Thompson, A. L.;

Paton, R. S.; Brown, J. M.; Gouverneur, V. Coordination diversity in

hydrogen-bonded homoleptic fluoride-alcohol complexes modulates

reactivity. Chem. Sci. 2015, 6, 5293−5302.

58. (d) Shermolovich, Y. Fluorinated Five-Membered Heterocycles

(19) For drying of TMAF alcohol complexes in alcohol solvent see;

Containing Oxygen, Sulfur, Selenium, and Phosphorus. In Fluorinated

Heterocyclic Compounds: Synthesis, Chemistry, and Applications; Petrov,

V. A., Ed.; Wiley: Hoboken, NJ, 2009; pp 159−225.

Bland, D. C.; Yang, C. Method for Preparing Tetramethylammonium

Fluoride. U.S. Patent 20,180,050,980 A1, Feb 22, 2018.

(

9) (a) Grineva, A. A.; Filippov, O. A.; Nefedov, S. E.; Lugan, N.;

César, V.; Valyaev, D. A. Direct Access to IMesF and IMesF2 by

Electrophilic Fluorination of Abnormal N -Heterocyclic Carbenes.

Organometallics 2019, 38, 2330−2337. (b) Lu, S.; Tian, L.-L.; Cui, T.-

W.; Zhu, Y.-S.; Zhu, X.; Hao, X.-Q.; Song, M.-P. Copper Mediated C-