Development and Proof of Concept for a Large-Scale Photoredox Additive-Free Minisci Reaction

Additive-Free Minisci Reaction

Article

Development and Proof of Concept for a Large-Scale Photoredox

Mark A. Graham,* Gary Noonan,* Janette H. Cherryman, James J. Douglas, Miguel Gonzalez,

Lucinda V. Jackson, Kevin Leslie, Zhi-qing Liu, David McKinney, Rachel H. Munday, Chris D. Parsons,

David T. E. Whittaker, En-xuan Zhang, and Jun-wang Zhang

Cite This: https://dx.doi.org/10.1021/acs.oprd.0c00483

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: New route development activities toward ceralasertib (AZD6738) have resulted in the discovery of an eﬃcient, acid

additive-free, photoredox Minisci reaction. Mechanistic understanding resulting from LED-NMR reaction proﬁling, quantum yield

measurements, and Stern−Volmer quenching studies have enabled optimization of the catalyst system, resulting in a signiﬁcant

enhancement in the rate of reaction. A large-scale continuous photoﬂow process has been developed, providing encouraging proof-

of-concept data for the future application of this technology in the clinical manufacture of ceralasertib.

KEYWORDS: Minisci, photoredox, photoﬂow, ceralasertib, AZD6738, LED-NMR

INTRODUCTION

advances in the scale-up of photochemical processes using

continuous ﬂow chemistry were also encouraging when

considering light-driven manufacturing processes for a drug

■

−3

Ceralasertib (AZD6738, 1)

is currently being evaluated in

multiple phase I/II clinical trials for the treatment of cancer.

Its structure, comprising a pyrimidine core decorated with a

chiral morpholine, a cyclopropyl sulfoximine, and an azaindole,

makes it a challenging molecule to synthesize on a large scale.

The incorporation of the thiomethylcyclopropyl moiety on the

pyrimidine ring is especially challenging from a synthesis

perspective. A signiﬁcant improvement in the large-scale

synthesis of 1 came with the recently disclosed development

of a scalable approach to the thiomethylcyclopropyl acid 2,

16−20

substance.

the approach of Fu, Sherwood,

We were particularly interested in exploring

21 22,23 24

and Opatz who have

recently reported eﬀective Minisci reactions using redox active

esters under visible light photocatalysis conditions. Crucially,

these reactions are redox-neutral; thus, we were optimistic that

this chemistry would be compatible with our substrate. Herein,

we describe the development work carried out to realize this

transformation, as well as the mechanistic and practical insights

gleaned from the work. We then demonstrate the proof-of-

concept scale-up of this reaction using a photoﬂow reactor,

conﬁrming our process as a viable option for the future

manufacture of an active pharmaceutical ingredient (API).

shown in Scheme 1. However, in an attempt to further

improve the route, we considered the desirable “shortcut”

shown in Scheme 1, which would allow for more eﬃcient

access to the dichloropyrimidine intermediate 3.

The classical Minisci reaction involves the addition of a

nucleophilic alkyl radical to a heteroaromatic compound.

Typical reaction conditions employ an oxidant, most often a

persulfate salt, to induce decarboxylation to an alkyl radical, as

well as a Brønsted or Lewis acidic reagent, which is thought to

be important to increase the electrophilicity of the hetero-

RESULTS AND DISCUSSION

■

Development of Photoredox Minisci Conditions.

Initially, we investigated the visible-light photoredox reaction

under batch conditions, as reported by Sherwood. This

involved the in situ generation of the N-hydroxyphthalimide

redox active ester 6, followed by the addition of camphorsul-

fonic acid and catalytic 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-

dicyanobenzene (4CzIPN) in DMSO and irradiation in an

EvoluChem photochemical reactor. Pleasingly, the desired

product 3 was formed, albeit in a modest isolated yield of 25%,

arene. Mai and co-workers recently reported a silver-catalyzed

Minisci-type alkylation of 2-chloropyridine using potassium

persulfate, while Shore reported Minisci allylations using a

variety of electron-deﬁcient pyrimidines. We hoped that

similar conditions would be applicable to our system while

being acutely aware of the relatively facile oxidation of sulfur in

thioethers. Encouragingly, we were able to demonstrate the

conversion of acid 2 to pyrimidine 3 using classical Minisci

conditions (Scheme 2), albeit with the accompanying

formation of the sulfur oxidation products.

despite full consumption of active ester 6 (Scheme 3).

Received: October 27, 2020

Visible light photocatalytic reactions have exploded in

popularity over the past decade, and with this increased

interest, a large number of classical radical chemistries have

1−15

been realized under irradiative conditions.

Recent

XXXX American Chemical Society

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

Organic Process Research & Development

Article

Scheme 1. Current Manufacturing Route to Ceralasertib: Minisci “Shortcut” Highlighted

Scheme 2. Promising Result Employing Classical Minisci Conditions

Scheme 3. Initial Evaluation of Sherwood’s Conditions

Scheme 4. Preparation of N-Hydroxyphthalimide Redox Active Ester 6

With this encouraging result in hand, we carried out

screening of a variety of photocatalysts in a range of solvents

under batch conditions, using a 96-well plate photochemical

reactor. To enable this screening work, we synthesized and

isolated the N-hydroxyphthalimide redox active ester 6 via the

corresponding acid chloride (Scheme 4).

catalyst 4CzIPN was an attractive choice from a cost and

sustainability perspective, so we decided to focus our initial

reaction optimization activities on this catalyst.

The use of an acid additive in Minisci reactions is believed to

be important for reactivity, as outlined by the mechanism

proposed by Sherwood and co-workers. We therefore

proceeded to evaluate a range of Brønsted and Lewis acid

additives in a batch photoreactor. A variety of both Brønsted

and Lewis acid additives resulted in the formation of 3, but

surprisingly, we observed an inverse correlation between

From this catalyst-screening work, we established that the

catalysts 4CzIPN and [Ir(dtbbpy) (ppy) ]PF resulted in the

highest conversion to the desired product, and we observed

DMSO to be the preferred solvent. The organophotoredox

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

Organic Process Research & Development

Article

conversion to 3 (as measured by HPLC relative to an internal

standard) and increasing strength of the acid additive (see

Table 1).

the desired product in reasonable yield, that we turned our

focus to increased reaction productivity or throughput, central

to this would be a move to a ﬂow photochemistry platform.

Development of the Photoredox Minisci Reaction in

Continuous Flow. The initial results in a batch photo-

chemical reactor were encouraging, but we were acutely aware

that the scale-up of photocatalytic reactions presents unique

Table 1. Eﬀect of Acid pK on Conversion of 6 to 3 Using

CzIPN

18,31,32

challenges.

light penetration into the reaction solution is limited by the

necessarily) high molar extinction coeﬃcient of the photo-

catalyst. Therefore, penetration beyond the ﬁrst few milli-

Because of the Beer−Lambert−Bouguer law,

(

meters of the reaction mixture is often not possible. For this

reason, chemists have generally turned to ﬂow reactor design

to provide the required short pathlength for irradiation of the

bulk of the solution, while maintaining the ability to process

pK of acid (in

product/internal standard (HPLC

area counts)

8,29

17−19,32−41

acid additive

water)

large quantities of the material.

triﬂuoroacetic

0.23

1.35

To render the current photo-Minisci reaction viable for the

manufacture of ceralasertib on a required scale of greater than

acid

oxalic acid

pyruvic acid

malonic acid

1.23

2.39

2.83

1.35

1.61

1.93

00 kg, we estimated that a minimum productivity of 5 kg/day

of pyrimidine 3 would be required. Thus, although our initial

focus was on reaction yield, we now focused on increasing the

reaction rate to attain shorter residence times and thus higher

productivity for a continuous ﬂow process. We investigated our

highest yielding batch conditions using DIPEA as an additive

and 4CzIPN as a photocatalyst in DMSO using a Vapourtec

UV-150 photochemical reactor with blue LEDs (450 nm). We

were delighted to observe a 50% solution yield of product 3

with a residence time of 50 min which enabled complete

consumption of redox active ester 6. For comparison, the use

of malonic acid as an additive required a longer residence time

of 100 min to achieve complete consumption of 6 and a similar

solution yield. Unfortunately, we also observed very signiﬁcant

darkening of the reaction mixture. We believed that this

darkening would lead to diminishing returns for increases in

residence time and to test this hypothesis, and to begin to

improve our understanding of the reaction, we carried out

some LED-NMR experiments.

Next, we conducted an experiment with triethylamine

hydrochloride as an acidic additive, showing that the current

Minisci reaction could proceed with a very weak acid.

Remarkably, the use of the ammonium salt produced a higher

solution yield using the EvoluChem photochemical reactor

(Table 2). Indeed, we had doubted that the role of acid in our

Table 2. Investigation of Alternative Additives

LED-NMR Reaction Proﬁling. LED-NMR is a data-rich

technique for the analysis of photocatalytic reactions in situ.

entry

additive

NMR solution yield (%)

malonic acid (2.0 equiv)

triethylamine hydrochloride (2.0 equiv)

no additive

This technique enabled us to gain detailed information on the

reaction rate and proﬁle (Figure 1), as well as highlight the

formation of reaction byproducts over the course of the

reaction.

Our reaction proﬁling conﬁrmed the increased rate of the

reaction when DIPEA (0.2 equiv) is used as an additive

N,N-diisopropylethylamine (0.2 equiv)

Yields based on proton NMR of the crude reaction mixture,

quantiﬁed against a known amount of an external standard (TCNB).

(

Figure 1). We were able to improve yield using an excess (3

reaction was to protonate the heterocycle as proposed in the

general mechanism of Sherwood, as our substrate, 2,4-

dichloropyrimidine, is a very weak base indeed, with a

equiv) of the cheap and readily available reagent 2,4-

dichloropyrimidine. However, it also became apparent that

the product 3 is oxidized to sulfoxide 4 over extended time

periods. Notably, oxidation to sulfoxide 4 was only observed

when the redox active ester 6 was almost or fully consumed.

This would be consistent with an alternative photoredox

catalytic cycle becoming operative only when the redox active

ester was consumed. In theory, the formation of sulfoxide 4

could be controlled by identifying the optimal ﬂow rate, but we

were interested in further improving the rate of the reaction to

improve throughput.

22,30

calculated pK of −2.84.

We therefore performed an

experiment with no acid additive present and surprisingly

found that the desired product was formed in higher yield than

with any of the carboxylic acid additives present. For this

system, it appears that we have a combination of a particularly

nucleophilic radical and an electrophilic arene, allowing for

radical addition to occur in the absence of activation.

Realizing that an acid additive was not required, we decided

to investigate other additives, primarily DIPEA. Interestingly,

we observed an improved solution yield of 50% after 21 h

using catalytic DIPEA as an additive (Table 2). The

mechanistic eﬀects of DIPEA were unknown at this point

but we decided to proceed to scale-up. It was at this point in

development, with reaction conditions in hand that produced

We next carried out a series of “same-excess”-type

experiments in an attempt to increase our understanding of

the reaction under our best conditions using DIPEA (0.2

equiv) as an additive. The “same-excess” experiment

demonstrated that the rate of the reaction at 50% consumption

of 6 is not equivalent to the rate of the reaction when the

Organic Process Research & Development

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

Figure 1. LED-NMR reaction pro ﬁ le of the 4CzIPN-catalyzed Minisci reaction with and without a DIPEA additive.

Article

Figure 2. “Same excess” and potential product inhibition experiments.

reaction is initiated from this point, with fresh reagents at

concentrations equivalent to 50% consumption of 6 (Figure 2,

blue data points vs red data points). This would usually suggest

that a product from the reaction is inhibiting the catalyst or

that the catalyst is becoming “deactivated”. However, our in

situ NMR studies had shown that the 4CzIPN catalyst was

stable over the course of the reaction, so true catalyst

deactivation or decomposition was not responsible for the

drop in rate. Spiking experiments with the reaction product

plus phthalimide at concentrations matching those expected at

50% consumption of redox active ester 6 conﬁrmed that the

presence of product 3 or phthalimide did not inhibit the rate of

the reaction (Figure 2). The key observation that the reaction

mixture becomes progressively darker as the reaction

progresses led us to believe that this is the cause of the

reduction in the reaction rate over time. The solution

darkening has the eﬀect of reducing the number of photons

that can reach the photocatalyst. Therefore, the concentration

Organic Process Research & Development

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

Article

Figure 3. Structures of organophotoredox catalysts synthesized and studied.

Table 3. Comparison of Solution Yields and Redox Potentials with Diﬀerent Catalyst Systems

photocatalyst

CzIPN

CzIPN + DIPEA

CzClIPN

CzBN

DPA2FBN

residence time (min)

solution yield (%)

E₁_/₂reduction (V vs SCE)

E₁_/₂oxidation (V vs SCE)

•+

100

−1.18 (PC /PC*)

+1.49 (PC /PC)

•

−

+1.43 (PC*/PC^•⁻)

−1.24 (PC/PC

)

•

•+

−0.93 (PC /PC*)

−1.42 (PC /PC*)

−1.60 (PC /PC*)

+1.79 (PC /PC)

•+

+1.41 (PC /PC)

•+

+1.41 (PC /PC)

•+

+1.24 (PC /PC)

•+

−1.60 (PC /PC*)

+1.24 (PC /PC)

•−

+0.92 (PC*/PC^•⁻)

DPA2FBN + DIPEA

−1.92 (PC/PC )

Residence time is controlled by the ﬂow rate in a 10 mL volume ﬂow photoreactor.

of the excited photocatalyst decreases over the course of the

reaction, while the total concentration of the photocatalyst

class of organic photocatalysts could be eﬀectively tuned by

changing the substituents around the benzonitrile or

(

ground-state and excited-state photocatalyst) remains con-

isophthalonitrile core. Using either diphenylamine or

carbazole moieties as electron-donating substituents, they

showed that highly reducing and highly oxidizing catalysts

could be accessed. We synthesized a number of these catalysts

(Figure 3, synthesis described in the Supporting Information)

and carried out trial reactions with them on the photoﬂow

reactor (Table 3).

stant. This situation is equivalent to catalyst deactivation and

results in a signiﬁcant drop in the reaction rate relative to what

would be expected at that stage of the reaction. With the

reaction rate being our key optimization parameter, darkening

of the reaction mixture over time represented a signiﬁcant

barrier toward carrying out this reaction on large scale.

DIPEA has the ability to quench a photoexcited catalyst, and

although this appeared beneﬁcial from a rate perspective, the

When we compare solution yields at the shortest residence

time of 4 min, it is notable that 2,4,6-tri(9H-carbazol-9-yl)-5-

chloroisophthalonitrile (3CzClIPN), the catalyst with the

lowest reduction potential (−0.93 V vs SCE), gives a much

lower solution yield than all other catalysts. Pleasingly, two of

these catalysts showed exceptional reactivity in the current

reaction, namely, 2,4,6-tris(diphenylamino)-3,5-diﬂuorobenzo-

nitrile (3DPA2FBN) and penta-9H-carbazol-9-ylbenzonitrile

(5CzBN). 3DPA2FBN gave a solution yield of 70% with a 13

min residence time and 43% with a 4 min residence time. In

general, more reducing photoexcited catalysts seemed to be

favorable for our reaction system. However, when DIPEA was

employed as an additive with 3DPA2FBN, a lower solution

yield of 21% was observed, even though this system should

produce a highly reducing catalyst (−1.92 V vs SCE). Again,

signiﬁcantly, more darkening of the reaction mixture was

observed when DIPEA was employed as an additive, when

compared to the additive-free reactions. A comparison of the

4CzIPN-DIPEA system to 3DPA2FBN was also performed by

resulting byproducts are likely to cause solution darkening.

Although we had carried out the photo-Minisci reaction with

several alternative photocatalysts, a range of strongly reducing

photocatalysts were absent from our screening work. The

redox potential of the N-hydroxyphthalimide ester 6 was

measured as −1.22 V versus SCE by cyclic voltammetry.

When we compare the oxidation potential of photoexcited

•

CzIPN in the oxidative quenching cycle (E₁_/₂PC /PC* =

1.18 V vs SCE) to the oxidation potential of the

corresponding radical anion formed by reductive quenching

−

•

−

of 4CzIPN with DIPEA (E₁_/₂PC/PC = −1.24 V vs SCE),

•

−

it appears that the more strongly reducing PC from the

reductive quenching cycle could have a positive impact on the

rate of the reaction. We wondered if the reaction rate could be

further increased by employing more reducing catalysts, also

obviating the need for the addition of DIPEA. A recent study

by Zeitler and co-workers showed that the redox potentials of a

Organic Process Research & Development

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

Figure 4. Comparison of 4CzIPN-DIPEA to 3DPA2FBN by LED-NMR.

Article

Figure 5. Proposed oxidative quenching cycle.

LED-NMR (Figure 4). Not only did we observe a signiﬁcant

increase in the reaction rate, we also observed no further

oxidation of the desired product 3 to sulfoxide 4 when the

the 3DPA2FBN gave a quantum yield of 0.61. Importantly, the

increased rate observed at the beginning of the reaction is

maintained for most of the reaction proﬁle for 3DPA2FBN,

whereas the reaction rate tails oﬀ dramatically for the 4CzIPN-

DIPEA system.

DPA2FBN catalyst was used (Figure 5). The redox potential

of the sulﬁde in product 3 was measured as + 1.90 V versus

SCE by cyclic voltammetry. The radical cation of 3DPA2FBN

Finally, we carried out Stern−Volmer quenching studies to

•

(

E₁_/₂PC /PC = +1.24 V vs SCE) is signiﬁcantly less oxidizing

ﬁnd out which reaction components were eﬀective quenchers

•

48,49

when compared to the radical cation of 4CzIPN (E₁_/₂PC /

PC = +1.49 V vs SCE) and is therefore potentially unable to

oxidize sulﬁde 3, a plausible ﬁrst step toward the formation of

the sulfoxide byproduct. As stated above, we believe that some

of the decrease in the rate observed for the 4CzIPN system is

due to reaction darkening.

for our excited photocatalysts.

Unsurprisingly, DIPEA

quenched the 4CzIPN catalyst very eﬀectively with a K of 85

−

(Table 4), presumably forming the 4CzIPN radical anion;

this is consistent with our hypothesis that the rate enhance-

ment in moving from 4CzIPN to 4CzIPN-DIPEA is due to the

formation of the more reducing and negatively charged radical

anion of the photocatalyst. Also, the Stern−Volmer quenching

Quantum Yield Measurement and Stern−Volmer

Quenching Studies. We also carried out quantum yield

constant for 3DPA2FBN with redox active ester 6 (K = 9.52)

42,47

measurements using LED-NMR actinometry

for both the

is more than four times greater than that for 4CzIPN (K =

CzIPN-DIPEA catalyst system and for 3DPA2FBN. The

CzIPN-DIPEA system gave a quantum yield of 0.21, whereas

2.32), which is consistent with a more rapid electron transfer

from 3DPA2FBN. This allows an eﬃcient redox neutral

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

Organic Process Research & Development

which to develop a scalable work-up.

Article

Table 4. Stern−Volmer Quenching Studies

Results of Large-Scale Photoﬂow Proof of Concept.

Our optimized but not scaled up additive-free conditions using

3DPA2FBN as a catalyst worked very eﬀectively using the

Vapourtec UV-150 photochemical reactor, aﬀording 3 in 56%

isolated yield when using 3 g of redox active ester 6. However,

the throughput of material is limited by the size of the reactor.

With a total reactor volume of 10 mL, a continuous ﬂow

process using our current optimized conditions at a ﬂow rate of

0.75 mL/min would deliver approximately 32 g of product 3

over a 24 h period. We wanted to evaluate the eﬀectiveness of

our process using larger-scale continuous photoﬂow equip-

ment. We were aware of the limitations of light penetration

ﬂow reaction solution yield

photocatalyst quencher Stern−Volmer K_Q

(%) (residence time)

CzIPN

DPA2FBN

CzBN

DIPEA

85.21

2.32

9.52

3.12

12.64

55 (50 min)

50 (100 min)

70 (13 min)

21 (13 min)

59 (13 min)

Yields based on proton NMR of the crude reaction mixture,

quantiﬁed against a known amount of an external standard (TCNB).

31,32

because of the Beer−Lambert−Bouguer law,

but we were

reaction with a proposed oxidative quenching cycle (Figure 5).

The ﬁnal oxidation to product 3 can be achieved either via

electron transfer to the radical cation of the photocatalyst or

via electron transfer to another molecule of redox active ester

starting material 6, thus resulting in a radical chain process.

Development of a Suitable Work-Up. Our photoredox

Minisci reaction using catalytic 3DPA2FBN was consistently

giving a DMSO solution of product 3 in 65−70% yield. We

believed that the yield and reaction rate could deliver a process

with a suitably high throughput for large-scale manufacturing

via continuous photoﬂow. However, for this reaction to be a

viable manufacturing process, we needed to develop a straight-

forward and potentially scalable work-up. We decided to trial

an aqueous work-up with an extraction into heptane.

Gratifyingly, we discovered that upon addition of the reaction

mixture to water (12.5 relative volumes) and heptane (25

relative volumes), >98% of the product 3 was observed in the

organic layer and >99% of the phthalimide, the major reaction

byproduct, resided in the aqueous layer. Also, almost all the

DMSO was washed into the aqueous layer. This work-up

yielded a heptane solution containing mainly 3, some

unreacted 2,4-dichloropyrimidine, and a trace of the

encouraged by the relative lack of solution darkening under our

additive-free conditions, and we believed that we had a

reasonably photon-eﬃcient reaction, having a quantum yield of

0.61. We initially evaluated our process in a photoﬂow reactor

with 6 mm diameter ﬂuorinated ethylene propylene (FEP)

tubing (internal diameter 4 mm), wrapped around 450 nm

LEDs (100 W). The solution was preheated to 50 °C, and the

pressure was controlled with a back-pressure regulator to

prevent the formation of slugs of CO gas (Figure 6).

Our initial evaluation using 4 mm internal diameter FEP

tubing was performed using 5 equiv of 2,4-dichloropyrimidine

and 2 mol % of 3DPA2FBN. A 15 mL/min ﬂow rate, giving a

residence time of 13 min, resulted in a solution yield of 24%

(

Table 5, entry 1). However, when the process was performed

at 6 mL/min to enable 33 min residence time, we observed an

encouraging solution yield of 62% (entry 2).

Using the same reagent stoichiometry but switching to FEP

tubing with 8 mm internal diameter and adjusting the ﬂow rate

accordingly to give an identical residence time of 33 min

resulted in a very encouraging solution yield of 65% (entry 3).

A comparison of 4 and 2.5 equiv of 2,4-dichloropyrimidine

charge using the 4 mm internal diameter tubing gave a slightly

better yield for 4 equiv, but this beneﬁt could be oﬀset with a

more laborious work-up to remove the excess reagent (entries

DPA2FBN catalyst. Based on the hydrophilicity of 2,4-

dichloropyrimidine compared to 3, we were able to remove

most of the excess 2,4-dichloropyrimidine by washing the

heptane solution with a mixture of water (15 relative volumes)

and DMSO (22.5 relative volumes) three times. Each

sequential wash removed approximately 50% of the residual

and 5). Maintaining the reagent stoichiometry but increasing

the overall reaction concentration resulted in a lower yield

entries 6 and 7), so we decided to use 25 relative volumes of

(

DMSO for scale-up. Finally, in an investigation of catalyst

loading using the 8 mm diameter FEP tubing, a signiﬁcant

drop-oﬀ in yield to 38% was observed when 0.5 mol % of

3DPA2FBN was used (entry 10), but 1 mol % of catalyst was

preferable to 2 mol % (entries 8 and 9). We chose to take the

optimal conditions of 2,4-dichloropyrimidine (2.5 equiv),

3DPA2FBN (1 mol %), and DMSO (25 relative volumes) to

scale-up using 250 g of redox active ester 6 in the 8 mm

,4-dichloropyrimidine into the aqueous layer, with loss of only

% of product 3 per wash. The heptane solution was then

switched to a propan-2-ol solution by performing a distillative

solvent swap. Finally, the product 3 was crystallized from

propan-2-ol using water as an antisolvent. Although more work

is required to optimize this work-up procedure for manufactur-

ing scale, the current protocol represents a solid basis from

Figure 6. Schematic diagram of the large-scale photoﬂow reactor.

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

Organic Process Research & Development

Article

Table 5. Evaluation of the Photoredox Minisci Reaction Using Large-Scale Photoﬂow Equipment

catalyst

equiv of 2

entry ,4-dichloro-pyrimidine

DMSO volume

(relative volumes)

tubing internal

diameter (mm)

tubing irradiation

volume (mL)

ﬂow rate

residence

solution yield at

(mol %)

(mL/min) time (min)

steady state (%)

2.5

200

100

200

100

12.5

0.5

Yields based on proton NMR of the crude reaction mixture, quantiﬁed against a known amount of an external standard (TCNB).

internal diameter tubing (irradiated length of 2 m). The

solution was pumped through the photoﬂow reactor at a ﬂow

rate of 3 mL/min to enable a 33 min residence time. The crude

solution yield for the output solution was 52% (measured by

tion and mechanistic understanding, resulting from LED-NMR

reaction proﬁling and Stern−Volmer quenching studies, have

enabled us to enhance signiﬁcantly the rate of the reaction and

avoid the phenomenon of reaction darkening. A change in

catalyst also removed the unwanted over-reaction sulfoxide

product. This process has been successfully scaled-up to 8 mm

diameter continuous photoﬂow equipment, providing proof of

concept on a pilot scale for this process to be used for multi-

kilogram per day manufacture of the API. We believe that the

current work demonstrates that visible-light photoredox

catalysis has a bright future in the large-scale manufacture of

pharmaceutical intermediates and API.

H NMR against an external standard), and this solution

contained approximately 4% of the unreacted redox active

ester 6. The output solution was added to water and heptane

and then the resulting heptane solution was washed with

water/DMSO (2:3) to remove excess 2,4-dichloropyrimidine.

Concentration of the resulting heptane solution yielded 118 g

of product 3 in 44% yield (purity 79% by H NMR assay).

Further recrystallization of the product from propan-2-ol (236

mL) and water (24 mL) aﬀorded 3 in 31% yield, although this

crystallization process is not fully optimized.

EXPERIMENTAL SECTION

The large-scale pilot photoﬂow experiments were performed

using a ﬂuorinated ethylene propylene (FEP) tubing ﬂow cell

■

This process is not yet fully optimized, but we were very

encouraged by the yield, reaction rate, and productivity. If this

process using the 8 mm internal diameter tubing was repeated

on the currently available pilot-scale equipment and run

continuously for 24 h, we estimate that 65 g/day of product 3

could be manufactured. However, the 8 mm diameter pilot-

scale photoﬂow equipment has a tube length of only 2 m. We

have the option to use larger-scale photoﬂow equipment with

an internal diameter of 8 mm but with a total length of 204 m.

Related photoﬂow equipment using UV LEDs for 2 + 2

cycloaddition was reported by Beaver and co-workers enabling

(

external diameter 10 mm, internal diameter 8 mm, wall

thickness 1 mm, irradiation volume 200 mL), equipped with a

with a back-pressure regulator. Irradiation of the FEP tubing

was performed with blue LEDs (450 nm, 600 LEDs in the

array, electric power 100 W, light beam geometry 90°,

positioned 10 mm from FEP tubing). The light irradiation

intensity was 34 mW/cm on the surface of the reactor, and

2.5 mW/cm was passed through the empty FEP tube. H

NMR spectra were recorded on a Bruker DRX 500 (500

a throughput of >5 kg/day. If we maintain the same

residence time of 33 min and increase the ﬂow rate

accordingly, we can extrapolate that the productivity of the

continuous ﬂow process to generate 3 would increase to

approximately 6.6 kg/day using the current conditions. Any

increase in concentration or optimization of stoichiometry and

yield could further improve the productivity of this process.

We believe that this scalable photoredox reaction is a viable

process for the future manufacture of a drug substance, and

work is currently underway to further optimize this process.

MHz). The central peaks of dimethylsulfoxide-d (DMSO-d ;

δH 2.50 ppm) was used as a reference. Spectral data are

reported as a list of chemical shifts (δ, in ppm) with a

description of each signal, using standard abbreviations (s =

singlet, d = doublet, m = multiplet, t = triplet, q = quartet, br =

broad, etc.).

Large-Scale Pilot Synthesis of 2,4-Dichloro-6-[1-

(

methylsulfanyl)cyclopropyl]pyrimidine (3). (1,3-Dioxoi-

soindolin-2-yl)1-methylsulfanylcyclopropanecarboxylate (6,

50 g, 1.0 equiv), 2,4-dichloropyrimidine (336.12 g, 2.50

equiv), and 3DPA2FBN (5.77 g, 0.01 equiv) were dissolved in

DMSO (6.25 L). The solution was sparge-degassed with

nitrogen for 10 min. The resulting solution was preheated to

50 °C and pumped through a ﬂow cell (8 mm internal

CONCLUSIONS

We have carried out successful development work on an

additive-free photoredox Minisci reaction. Catalyst optimiza-

■

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

Organic Process Research & Development

Complete contact information is available at:

Article

diameter, 2 m length), which was irradiated with blue visible

light (450 nm) at a ﬂow rate to enable a residence time of 33

min at steady state (back pressure 0.6−0.8 MPa). The output

solution was added dropwise to a stirred mixture of water (3.12

L) and heptane (6.25 L). The layers were partitioned and then

the organic layer was washed three times with a mixture of

water (3.75 L) and DMSO (5.625 L). The organic layer was

concentrated to yield 2,4-dichloro-6-[1-(methylsulfanyl)-

cyclopropyl]pyrimidine (3, 118 g, 44% yield). Assay 79% w/

Chris D. Parsons − Early Chemical Development,

Pharmaceutical Sciences, R&D, AstraZeneca, Macclesﬁeld

SK10 2NA, U.K.

David T. E. Whittaker − Early Chemical Development,

Pharmaceutical Sciences, R&D, AstraZeneca, Macclesﬁeld

SK10 2NA, U.K.

En-xuan Zhang − Asymchem Laboratories (Tianjin) Co. Ltd.,

TEDA, Tianjin 300457, P. R. China

Jun-wang Zhang − Asymchem Laboratories (Tianjin) Co.

Ltd., TEDA, Tianjin 300457, P. R. China

w ( H NMR).

Recrystallization of the crude product in propan-2-ol (236

mL) and water (24 mL) aﬀorded 2,4-dichloro-6-[1-

https://pubs.acs.org/10.1021/acs.oprd.0c00483

(

methylsulfanyl)cyclopropyl]pyrimidine (3, 67 g, 31% yield).

Assay 87% w/w ( H NMR). H NMR (500 MHz, DMSO, 27

Author Contributions

C) 1.42−1.55 (2H, m), 1.61−1.77 (2H, m), 2.14 (3H, s),

The manuscript was written by M.A.G. and G.N. through

contributions of all authors. All authors have given approval to

the ﬁnal version of the manuscript.

.03 (1H, s); C NMR (126 MHz, DMSO, 27 °C) 14.94,

3.02, 29.84, 117.66, 159.16, 161.92, 176.54; HRMS (ESI):

calcd for C H N SCl [M + H] , 234.9858; found, 234.9863.

Notes

The authors declare no competing ﬁnancial interest.

ASSOCIATED CONTENT

sı Supporting Information

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.oprd.0c00483.

■

ACKNOWLEDGMENTS

■

The authors would like to thank Gair Ford, Peter Hamilton,

Martin Jones, Rita Galan, Adam Clarke, Carl Mallia, Niall

McCreanor, Anne O’Kearney McMullan, Simone Tomasi,

Robert Cox, and Ian Ashworth for helpful discussions. The

authors would like to thank Emma Randles and Andrew Ray

for analytical support.

Details of screening experiments, batch photochemistry,

small-scale photoﬂow, H NMR spectra, Stern−Volmer

quenching studies, and quantum yield measurement and

cyclic voltammetry (PDF )

ABBREVIATIONS

API, active pharmaceutical ingredient; CSA, camphorsulfonic

acid; 5CzBN, penta-9H-carbazol-9-ylbenzonitrile; 3CzClIPN,

,4,6-tri(9H-carbazol-9-yl)-5-chloroisophthalonitrile; 4CzIPN,

AUTHOR INFORMATION

Corresponding Authors

Mark A. Graham − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Maccles ﬁ eld SK10 2NA, U.K.; orcid.org/0000-0001-

■

,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene; DCM, di-

chloromethane; DIC, diisopropylcarbodiimide; DIPEA, N,N-

diisopropylethylamine; DMAP, N,N-dimethylaminopyridine;

DMF, N,N-dimethylformamide; 3DPA2FBN, 2,4,6-tris-

186-9107 ; Email: mark.a.graham@astrazeneca.com

Gary Noonan − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Maccles ﬁ eld SK10 2NA, U.K.; Email: Gary.Noonan1@

astrazeneca.com

(

diphenylamino)-3,5-diﬂuorobenzonitrile; LED, light-emitting

diode; MS, mass spectrum; NMR, nuclear magnetic resonance;

PC, photocatalyst; SCE, saturated calomel electrode; SET,

single electron transfer; V, volt

Authors

Janette H. Cherryman − Chemical Development,

Pharmaceutical Technology & Development, Operations,

AstraZeneca, Macclesﬁeld SK10 2NA, U.K.

James J. Douglas − Early Chemical Development,

Pharmaceutical Sciences, R&D, AstraZeneca, Maccles ﬁ eld

SK10 2NA, U.K.; orcid.org/0000-0002-9681-0459

Miguel Gonzalez − Asymchem Laboratories (Tianjin) Co.

Ltd., TEDA, Tianjin 300457, P. R. China

Lucinda V. Jackson − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Macclesﬁeld SK10 2NA, U.K.

Kevin Leslie − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Macclesﬁeld SK10 2NA, U.K.

REFERENCES

■

(1) Foote, K. M.; Nissink, J. W. M.; McGuire, T.; Turner, P.;

Guichard, S.; Yates, J. W. T.; Lau, A.; Blades, K.; Heathcote, D.;

Odedra, R.; Wilkinson, G.; Wilson, Z.; Wood, C. M.; Jewsbury, P. J.

Discovery and Characterization of AZD6738, a Potent Inhibitor of

Ataxia Telangiectasia Mutated and Rad3 Related (ATR) Kinase with

Application as an Anticancer Agent. J. Med. Chem. 2018, 61, 9889−

(

907.

2) Foote, K. M., Nassink, J. W. M.; Turner, P. Morpholino

pyrimidines and their use in therapy. WO 2011154737 A1, 2011.

(3) Goundry, W. R. F.; Dai, K.; Gonzalez, M.; Legg, D.; O’Kearney-

McMullan, A.; Morrison, J.; Stark, A.; Siedlecki, P.; Tomlin, P.; Yang,

J. Development and Scale-up of a Route to ATR Inhibitor AZD6738.

Org. Process Res. Dev. 2019, 23, 1333−1342.

(

4) Henssen, A. G.; Reed, C.; Jiang, E.; Garcia, H. D.; von Stebut, J.;

Zhi-qing Liu − Asymchem Laboratories (Tianjin) Co. Ltd.,

TEDA, Tianjin 300457, P. R. China

MacArthur, I. C.; Hundsdoerfer, P.; Kim, J. H.; de Stanchina, E.;

Kuwahara, Y.; Hosoi, H.; Ganem, N. J.; Dela Cruz, F.; Kung, A. L.;

Schulte, J. H.; Petrini, J. H.; Kentsis, A. Therapeutic targeting of

PGBD5-induced DNA repair dependency in pediatric solid tumors.

Sci. Transl. Med. 2017, 9, 9078.

5) Graham, M. A.; Askey, H.; Campbell, A. D.; Chan, L.; Cooper,

K. D.; Cui, Z.; Dalgleish, A.; Dave, D.; Ensor, G.; Galan, R.; Hamilton,

David McKinney − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Macclesﬁeld SK10 2NA, U.K.

Rachel H. Munday − Chemical Development, Pharmaceutical

Technology & Development, Operations, AstraZeneca,

Macclesﬁeld SK10 2NA, U.K.

(

P.; Heﬀernan, C.; Jackson, L. V.; Jones, M. F.; Mulholland, K. R.;

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

Organic Process Research & Development

Pervez, M.; Popadynec, M.; Randles, E.; Tomasi, S.; Wang, S.

Article

Development and Scale-up of an Improved Manufacturing Route to

the ATR Inhibitor Ceralasertib. Org. Process Res. Dev. 2020,

DOI: 10.1021/acs.oprd.0c00482 .

(25) Details of by-products in Supporting Information.

(26) See Supporting Information for further details.

(27) Shang, T.-Y.; Lu, L.-H.; Cao, Z.; Liu, Y.; He, W.-M.; Yu, B.

Recent advances of 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene

(4CzIPN) in photocatalytic transformations. Chem. Commun. 2019,

55, 5408−5419.

(28) Dippy, J. F. J.; Hughes, S. R. C.; Rozanski, A. 498. The

dissociation constants of some symmetrically disubstituted succinic

acids. J. Chem. Soc. 1959, 2492.

(29) Braude, E. A.; Nachod, F. C. Determination of Organic Structures

by Physical Methods; Academic Press: New York, 1955.

(30) Calculated using Advanced Chemistry Development (ACD/

Labs) Software V11.02 (1994−2020 ACD/Labs).

(31) Moschetta, E. G.; Richter, S. M.; Wittenberger, S. J. Heuristics,

Protocol, and Considerations for Flow Chemistry in Photoredox

Catalysis. ChemPhotoChem 2017, 1, 539−543.

(32) Bonfield, H. E.; Knauber, T.; Levesque, F.; Moschetta, E. G.;

Susanne, F.; Edwards, L. J. Photons as a 21st century reagent. Nat.

Commun. 2020, 11, 804.

(

6) Graham, M. A.; Jackson, L. V.; Noonan, G. M.; Inglesby, P. A.;

Dave, D. P.; Cooper, K. G. Pharmaceutical process and intermediates.

WO 2020127208 A2, 2020.

7) Minisci, F.; Bernardi, R.; Bertini, F.; Galli, R.; Perchinummo, M.

Nucleophilic character of alkyl radicals VI. Tetrahedron 1971, 27,

575−3579.

8) Proctor, R. S. J.; Phipps, R. J. Recent Advances in Minisci-Type

Reactions. Angew. Chem., Int. Ed. Engl. 2019, 58, 13666−13699.

9) Mai, W.-P.; Sun, B.; You, L.-Q.; Yang, L.-R.; Mao, P.; Yuan, J.-

W.; Xiao, Y.-M.; Qu, L.-B. Silver catalysed decarboxylative alkylation

and acylation of pyrimidines in aqueous media. Org. Biomol. Chem.

10) Shore, D. G. M.; Wasik, K. A.; Lyssikatos, J. P.; Estrada, A. A.

015, 13, 2750−2755.

Minisci alkylations of electron-deficient pyrimidines with alkyl

carboxylic acids. Tetrahedron Lett. 2015, 56, 4063−4066.

11) Marzo, L.; Pagire, S. K.; Reiser, O.; Ko

nig, B. Visible-Light

(33) Cambie,

Noel, T. Applications of Continuous-Flow Photochemistry in Organic

D.; Bottecchia, C.; Straathof, N. J. W.; Hessel, V.;

Photocatalysis: Does It Make a Difference in Organic Synthesis?

Angew. Chem., Int. Ed. Engl. 2018, 57, 10034−10072.

Synthesis, Material Science, and Water Treatment. Chem. Rev. 2016,

12) Douglas, J. J.; Sevrin, M. J.; Stephenson, C. R. J. Visible Light

116, 10276−10341.

Photocatalysis: Applications and New Disconnections in the Synthesis

(34) Politano, F.; Oksdath-Mansilla, G. Light on the Horizon:

Current Research and Future Perspectives in Flow Photochemistry.

Org. Process Res. Dev. 2018, 22, 1045−1062.

of Pharmaceutical Agents. Org. Process Res. Dev. 2016, 20, 1134−

13) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Visible Light

147.

(35) VanHeyst, M. D.; Qi, J.; Roecker, A. J.; Hughes, J. M. E.;

Cheng, L.; Zhao, Z.; Yin, J. Continuous Flow-Enabled Synthesis of

Bench-Stable Bicyclo[1.1.1]pentane Trifluoroborate Salts and Their

Utilization in Metallaphotoredox Cross-Couplings. Org. Lett. 2020,

22, 1648−1654.

Photoredox Catalysis with Transition Metal Complexes: Applications

in Organic Synthesis. Chem. Rev. 2013 , 113 , 5322 −5363.

(

14) Romero, N. A.; Nicewicz, D. A. Organic Photoredox Catalysis.

Chem. Rev. 2016, 116, 10075−10166.

15) McAtee, R. C.; McClain, E. J.; Stephenson, C. R. J. Illuminating

Photoredox Catalysis. Trends Chem. 2019, 1, 111−125.

16) Beaver, M. G.; Zhang, E.-x.; Liu, Z.-q.; Zheng, S.-y.; Wang, B.;

Campeau, L.-C.; DiRocco, D. A.; Britton, R. Development of a Direct

(36) Halperin, S. D.; Kwon, D.; Holmes, M.; Regalado, E. L.;

Photocatalytic C-H Fluorination for the Preparative Synthesis of

Odanacatib. Org. Lett. 2015 , 17 , 5200 −5203.

(37) Reiser, O. Blue Chemistry: Trifluoromethylations Going Large!

Chem 2016, 1, 344−345.

(38) Beatty, J. W.; Douglas, J. J.; Miller, R.; McAtee, R. C.; Cole, K.

P.; Stephenson, C. R. J. Photochemical Perfluoroalkylation with

Pyridine N-Oxides: Mechanistic Insights and Performance on a

Kilogram Scale. Chem 2016, 1, 456−472.

(39) Yayla, H. G.; Peng, F.; Mangion, I. K.; McLaughlin, M.;

Campeau, L.-C.; Davies, I. W.; DiRocco, D. A.; Knowles, R. R.

Discovery and mechanistic study of a photocatalytic indoline

dehydrogenation for the synthesis of elbasvir. Chem. Sci. 2016, 7,

2066−2073.

(40) Bonfield, H. E.; Mercer, K.; Diaz-Rodriguez, A.; Cook, G. C.;

McKay, B. S. J.; Slade, P.; Taylor, G. M.; Ooi, W. X.; Williams, J. D.;

Roberts, J. P. M.; Murphy, J. A.; Schmermund, L.; Kroutil, W.; Mielke,

T.; Cartwright, J.; Grogan, G.; Edwards, L. J. The Right Light: De

Novo Design of a Robust Modular Photochemical Reactor for

Optimum Batch and Flow Chemistry. ChemPhotoChem 2019, 4, 45−

51.

(41) Steiner, A.; Roth, P. M. C.; Strauss, F. J.; Gauron, G.; Tekautz,

G.; Winter, M.; Williams, J. D.; Kappe, C. O. Multikilogram per Hour

Continuous Photochemical Benzylic Brominations Applying a Smart

Dimensioning Scale-up Strategy. Org. Process Res. Dev. 2020, 24,

2208−2216.

(42) Ji, Y.; DiRocco, D. A.; Kind, J.; Thiele, C. M.; Gschwind, R. M.;

Reibarkh, M. LED-Illuminated NMR Spectroscopy: A Practical Tool

for Mechanistic Studies of Photochemical Reactions. ChemPhotoChem

2019, 3, 984−992.

(43) Blackmond, D. G. Kinetic Profiling of Catalytic Organic

Reactions as a Mechanistic Tool. J. Am. Chem. Soc. 2015, 137,

10852−10866.

(

Lu, J.-p.; Tao, J.; Gonzalez, M.; Jones, S.; Tedrow, J. S. Development

and Execution of a Production-Scale Continuous [2 + 2] Photo-

cycloaddition. Org. Process Res. Dev. 2020, 24, 2139−2146.

(

17) Harper, K. C.; Moschetta, E. G.; Bordawekar, S. V.;

Wittenberger, S. J. A Laser Driven Flow Chemistry Platform for

Scaling Photochemical Reactions with Visible Light. ACS Cent. Sci.

(

019, 5, 109−115.

18) Noel, T. Photochemical Processes in Continuous-Flow Reactors:

From Engineering Principles to Chemical Applications, 1st ed.; World

Scientiﬁc Publishing Europe Ltd, 2017; p 271.

(

19) Corcoran, E. B.; McMullen, J. P.; Levesque, F.; Wismer, M. K.;

Naber, J. R.; McMullen, J. P.; Levesque, F.; Wismer, M. K.; Naber, J.

R. Photon Equivalents as a Parameter for Scaling Photoredox

Reactions in Flow: Translation of Photocatalytic C-N Cross-Coupling

from Lab Scale to Multikilogram Scale. Angew. Chem., Int. Ed. Engl.

(

020, 59, 11964−11968.

20) Grimm, I.; Hauer, S. T.; Schulte, T.; Wycich, G.; Collins, K. D.;

Lovis, K.; Candish, L. Upscaling Photoredox Cross-Coupling

Reactions in Batch Using Immersion-Well Reactors. Org. Process

Res. Dev. 2020, 24, 1185−1193.

21) Cheng, W.-M.; Shang, R.; Fu, M.-C.; Fu, Y. Photoredox-

Catalysed Decarboxylative Alkylation of N-Heteroarenes with N-

Acyloxy)phthalimides. Chemistry 2017, 23, 2537−2541.

22) Sherwood, T. C.; Li, N.; Yazdani, A. N.; Dhar, T. G. M.

Organocatalyzed, Visible-Light Photoredox-Mediated, One-Pot Min-

isci Reaction Using Carboxylic Acids via N-(Acyloxy)phthalimides. J.

Org. Chem. 2018, 83, 3000−3012.

(

23) Sherwood, T. C.; Xiao, H.-Y.; Bhaskar, R. G.; Simmons, E. M.;

Zaretsky, S.; Rauch, M. P.; Knowles, R. R.; Dhar, T. G. M.

Decarboxylative Intramolecular Arene Alkylation Using N-(Acyloxy)-

phthalimides, an Organic Photocatalyst, and Visible Light. J. Org.

Chem. 2019, 84, 8360−8379.

(44) Beatty, J. W.; Stephenson, C. R. J. Amine Functionalization via

Oxidative Photoredox Catalysis: Methodology Development and

Complex Molecule Synthesis. Acc. Chem. Res. 2015, 48, 1474−1484.

(45) Details in Supporting Information .

24) Kammer, L. M.; Rahman, A.; Opatz, T. A Visible Light-Driven

Minisci-Type Reaction with N-Hydroxyphthalimide Esters. Molecules

018, 23, 764.

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

Organic Process Research & Development

Article

46) Speckmeier, E.; Fischer, T. G.; Zeitler, K. A Toolbox Approach

To Construct Broadly Applicable Metal-Free Catalysts for Photo-

redox Chemistry: Deliberate Tuning of Redox Potentials and

Importance of Halogens in Donor −Acceptor Cyanoarenes. J. Am.

Chem. Soc. 2018, 140, 15353−15365.

47) Ji, Y.; DiRocco, D. A.; Hong, C. M.; Wismer, M. K.; Reibarkh,

M. Facile Quantum Yield Determination via NMR Actinometry. Org.

Lett. 2018, 20, 2156−2159.

48) Pitre, S. P.; McTiernan, C. D.; Scaiano, J. C. Understanding the

Kinetics and Spectroscopy of Photoredox Catalysis and Transition-

Metal-Free Alternatives. Acc. Chem. Res. 2016, 49, 1320 −1330.

49) Buzzetti, L.; Crisenza, G. E. M.; Melchiorre, P. Mechanistic

Studies in Photocatalysis. Angew. Chem., Int. Ed. Engl. 2019, 58,

730−3747.