10 1021@acs Oprd 0c00172 PDF

Subscriber access provided by Uppsala universitetsbibliotek
Full Paper
Development of a Large-Scale Cyanation Process Using Continuous
Flow Chemistry en Route to the Synthesis of Remdesivir
Tiago Vieira, Andrew Stevens, Andrei Chtchemelinine, Detian Gao, Pavel Badalov, and Lars Heumann
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.0c00172 • Publication Date (Web): 21 May 2020
Downloaded from pubs.acs.org on May 21, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,

Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 30 Organic Process Research & Development
1
2
3
4
5
6
7
8
Development of a Large-Scale Cyanation Process
9
10
11
12
Using Continuous Flow Chemistry en Route to the
13
14
15
16
Synthesis of Remdesivir
17
18
19
20
21 Tiago Vieira, †* Andrew C. Stevens,†* Andrei Chtchemelinine,‡ Detian Gao,† Pavel Badalov,†
22
23 and Lars Heumann ‡
24
25
26
†Gilead Alberta ULC, 1021 Hayter Road, Edmonton, Alberta T6S 1A1, Canada
27
28
29 ‡Gilead Sciences, Inc. 333 Lakeside Drive, Foster City, California 94404, United States
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 ACS Paragon Plus Environment
1
Organic Process Research & Development Page 2 of 30
1
2
3
Table of Contents Graphic
4
5
NH2 NH2 NH2
6
N N N
7 O
N
N
TFA, TMSOTf,
TMSCN O
N
N
O
O O
N
BnO BnO N
8 OH
DCM, -30 °C N
O HN P O
O N
9 BnO OBn Flow Chemistry BnO OBn HO OH
10 1 2 remdesivir
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
2
1
2
3
ABSTRACT: The implementation of cyanation chemistry at manufacturing scales using batch
4
5
6 equipment can be challenging due to the hazardous nature of the reagents employed, and the
7
8
9 tight control of reaction parameters, including cryogenic temperatures, that help to afford
10
11 acceptable selectivity and conversion for the desired reaction. Application of continuous flow
12
13
14 chemistry offers a means to mitigate the risk associated with handling large amounts of
15
16
17 hazardous reagents and to better control the reaction parameters. A case study describing the
18
19 cyanation of a glycoside using continuous flow chemistry towards the synthesis of the drug
20
21
22 candidate remdesivir is presented.
23
24
25
26 KEYWORDS: Continuous flow chemistry, cyanation, remdesivir, COVID-19
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
3
1
2
3
Introduction
4
5
In response to the 2014 outbreak of Ebola virus disease in Libera, Guinea and Sierra Leone,
6
7
8 Gilead Sciences Inc., in collaboration with the Centers for Disease Control and Prevention
9
10
11 (CDC) and the U.S. Army Medical Research Institute for Infectious Diseases (USAMRIID),
12
13 identified an adenosine nucleotide analogue that demonstrated strong antiviral activity against
14
15
16 multiple strains of the Ebola virus in vitro and in animal models (remdesivir, Scheme 1).1
17
18
19 Remdesivir has also demonstrated broad-spectrum antiviral activity against coronaviruses,
20
21 including SARS-CoV and MERS-CoV in primary human cell culture models, and in vivo
22
23
24 models of viral pathogenesis.2 More recently, remdesivir has been found to possess in vitro
25
26
inhibition of the novel 2019 coronavirus (SARS-CoV-2) responsible for the outbreak of
27
28
29 COVID-19.3 Based on this observed activity towards coronaviruses, remdesivir has been
30
31
32 applied, on compassionate use grounds, in response to the outbreak of COVID-19 infection4,5
33
34 and clinical trials have been initiated.6,7
35
36
37 The synthesis of remdesivir has been previously reported8 and proceeds through a 6-step
38
39
40 sequence (Scheme 1) which includes the installation of a cyano group through the
41
42 stereoselective addition of cyanide to the 1’-position of riboside 1. In the initial process, this
43
44
45 transformation, required cryogenic conditions (-78 °C) and had the potential to liberate
46
47
48
significant quantities of hydrogen cyanide. Pre-clinical and clinical evaluation of remdesivir
49
50 would necessitate production of multi-kilogram quantities of compound 2, requiring the
51
52
53 establishment of a safe, robust process for its synthesis. Herein we report the development of
54
55
continuous flow conditions for the preparation of 2.
56
57
58
59
4
1
2
3 Scheme 1. Synthetic Route to Remdesivir
4
5
6 NH2 NH2 NH2
7 N
TfOH, TMSOTf,
N
O
N
N N N
8 BnO
O N TMSCN
BnO
O N O O N
OH O HN P O
9 DCM, -78 °C N
O N
BnO OBn BnO OBn
10 HO OH
11 remdesivir
1 2
12
13
14
Results and Discussion
15
16 Batch Mode Preparation of Compound 2
17
18
19 Two manufacturing processes for the preparation of compound 2 were developed: i) a batch
20
21
cyanation process for pre-clinical and early-stage clinical remdesivir demands; ii) a cyanation
22
23
24 process using continuous flow for advanced phase clinical trials and commercial
25
26
27 manufacturing.
28
29 As previously reported,8 the cyanation reaction is performed through combination of
30
31
32 compound 1 (Scheme 1) in dichloromethane (DCM) with trifluoromethanesulfonic acid
33
34
35 (TfOH), trimethylsilyl trifluormethanesulfonate (TMSOTf), and trimethylsilyl cyanide
36
37 (TMSCN) while maintaining the reaction temperature at -78 °C. These reaction conditions
38
39
40 afforded a clean reaction profile, with good diastereoselectivity (Table 1, entry 1). The first
41
42
43
challenge to be addressed was the requirement of cryogenic reaction temperatures. When the
44
45 reaction was conducted at -15 °C, a rapid and significant decomposition of the starting material
46
47
48 was observed along with reduced diastereoselectivity for the transformation (Table 1, entry 2).
49
50 Alternative Brønsted acids were evaluated and trifluoroacetic acid (TFA) was found to be a
51
52
53 promising lead (Table 1, entries 3–7), providing good solution purity of compound 2 and
54
55
56 acceptable diastereoselectivity. Decreasing the reaction temperature to -30 °C was found to
57
58
59
5
1
2
3
increase the solution purity as a result of decreased degradation of reactive intermediates,
4
5
6 while providing a modest improvement to the diastereoselectivity (Table 1, entry 5 vs entry
7
8
9 8). The order of addition was also found to impact the diastereoselectivity. For example, adding
10
11 the TMSCN prior to the addition of TFA and TMSOTf significantly eroded the
12
13
14 diastereoselectivity of the reaction (Table 1, entry 5 vs entry 9). Reagent equivalents were
15
16
17 investigated and observed to play a role for reaction performance, as increasing the quantities
18
19 of both TMSOTf and TMSCN to 6 equivalents increased the selectivity of the reaction,
20
21
22 affording a high diastereoselectivity of 93:7. (Table 1, entries 10–12).
23
24
25
With the optimized conditions in hand, a larger scale experiment was performed (Table 1,
26
27 entry 13). During the scale-up experiment, TMSOTf and TMSCN were each added over
28
29
30 approximately 30 minutes to maintain the reaction at the target temperature of -30 °C. The
31
32
30 min addition time resulted in significant decomposition of the reaction mixture and a lower
33
34
35 diastereoselectivity (Table 1, entry 13). This finding was addressed by first combining the
36
37
38 starting material 1 in DCM with TFA at -30 °C, followed by adding a mixture of TMSOTf and
39
40 TMSCN in DCM that had been pre-cooled to -30 °C. This new order of addition prevented the
41
42
43 starting material from decomposing while maintaining acceptable diastereoselectivity (Table
44
45
46 1, entry 14). The stoichiometry of TMSOTf was revisited, and it was found that a decrease
47
48 would result in compromised diastereoselectivity (Table 1, entry 15).
49
50
51
52
53
54
55
56
57
58
59
6
1
2
3
Scheme 2. Preparation of Compound 2
4
5
6 NH2 NH2 NH2
7 N N N
8 O
N
N
Brønsted Acid
TMSOTf, TMSCN O
N
N O
N
N
BnO BnO BnO
9 OH
N
+ N
DCM, temperature
10 BnO OBn BnO OBn BnO OBn
11 1 2 3
12 Undesired Diastereomer
13
14
15 Table 1. Optimization of Cyanation Batch Mode Reaction Conditions
16
17
18 Brønsted Acid TMSOTf TMSCN Reaction Solution Purity In-Process d.r.
19 Entry Input of 1
Acid (equiv) (equiv) (equiv) Temperature (°C) of 2 (%)a Content of 3 (%)a (2:3)
20
21
22
1 100 mg TfOH 2 2 3 -78 90.4 3.6 96: 4
23
24 2 100 mg TfOH 2 2 3 -15 32.3 13.3 71:29
25
26 3 100 mg TsOH 2 4 4 -15 41.0 5.8 88:14
27
28 4 100 mg MsOH 2 4 4 -15 16.4 7.8 68:32
29
30 5 100 mg TFA 2 4 4 -15 54.1 9.5 85:15
31
32 6 100 mg HCO2H 2 3 3 -15 11.1 5.7 66:34
33
34
7 100 mg AcOH 2 3 3 -15 2.4 1.4 64:36
35
36
37
8 200 mg TFA 3 3 3 -30 75.7 8.7 90:10
38
39 9b 200 mg TFA 2 3 3 -15 34.8 23.0 60:40
40
41
42
43
44
46 7
47
1
2
3
10 200 mg TFA 1 1 3 -30 48.6 17.4 74:26
4
5
6 11 200 mg TFA 1 3 3 -30 84.0 8.4 91: 9
7
8 12 200 mg TFA 3 6 6 -30 88.6 6.8 93: 7
9
10 13c 66 g TFA 3 6 6 -30 69.2 8.0 90:10
11
12 14d 200 mg TFA 3 6 6 -30 83.2 14.7 85:15
13
14 15d 200 mg TFA 2 4 6 -30 78.3 20.4 79:21
15
16
Reaction Conditions: Compound 1 in DCM was cooled to the indicated temperature and the indicated Brønsted acid, TMSOTf,
17
18 and TMSCN were added sequentially. a Purity was determined by LC analysis. b Reaction Conditions: Compound 1 in DCM was
19 cooled to the indicated temperature and TMSCN was added, followed by TFA and TMSOTf. c Addition of TMSOTf and TMSCN
20
were performed over about 30 min each. d Reaction Conditions: TFA was added to a solution of compound 1 in DCM at -30 °C
21
22 and a solution of TMSCN and TMSOTf in DCM pre-cooled to -30 °C was subsequently added.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
46 8
47
1
2
3
Having identified viable conditions for the formation of compound 2, attention was directed
4
5
6 towards isolation of the product. On the basis of a number of laboratory trials in solvent
7
8
9 screening (over 40 single solvent or solvent combinations were investigated), it was
10
11 determined that compound 2 exhibited favorable solid-state properties and could be selectively
12
13
14 crystallized from either a mixed solvent system (such as ethyl acetate and heptane) or a single
15
16
17 solvent (such as 2-propanol or toluene). Following additional development studies, including
18
19 hazards assessments and safety studies, the batch cyanation process was successfully executed
20
21
22 for the early deliveries of remdesivir (Table 2). This enabled acceleration of the program at a
23
24
25
time when there was an urgent need of remdesivir for clinical evaluation due to the escalating
26
27 Ebola virus epidemic. For the planned large scale batch size, two key challenges needed to be
28
29
30 addressed: improvement of the safety measures due to the increased operational hazards
31
32
associated with the larger quantities of acidic reaction mixtures containing cyanide; and the
33
34
35 increased operational time required to combine large quantities of compound 1/TFA with
36
37
38 TMSOTf/TMSCN which could negatively impact diastereoselectivity and yield.
39
40 Table 2. Batch Mode Preparation of 2
41
42
43 Reaction d.r.
44 Input Crystallization Product Product d.r.
Batch Yield (%)
45 (kg) (2:3) Conditions Purity (%)a (2:3)
46
47
48 1 EtOAc /
9.6 84:16 70 100.0 99.9:0.1
49 heptane
50
51 2 22.0 86:14 IPA 57 99.0 99.7:0.3
52
53 3 23.1 N/Ab Toluene 71 99.2 99.1:0.9
54
55
56
57
58
59
9
1
2
3
Reaction Conditions: TFA (3 equiv) was added to a solution of compound 1 in DCM at -30 °C
4
5 and a solution of TMSCN (6 equiv) and TMSOTf (6 equiv) in DCM pre-cooled to -30 °C was
6 subsequently added. a Purity was determined by LC analysis. b Reaction samples not analyzed
7 to minimize exposure to hazardous reaction mixtures.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
10
1
2
3
Feasibility Studies for the Preparation of Compound 2 via Continuous Flow Chemistry
4
5
Based on the above challenges associated with further increasing batch size, alternative
6
7
8 processing modes were evaluated. Isolated literature reports discuss the use of continuous flow
9
10
11 chemistry to perform C-glycosylations,9 O-glycosylations,10 and the use of cyanide in flow
12
13 chemistry directly11 or by using a cyanide precursor.12 This mode of processing offers several
14
15
16 unique advantages to traditional batch processing, including enabling large scale chemistry to
17
18
19 be performed with reduced reaction volumes, and improved temperature control and mixing
20
21 efficiency.13-17 Application of these advantages of continuous flow chemistry could, in
22
23
24 principle, address the challenges associated with the batch cyanation process to prepare
25
26
compound 2. For instance, smaller quantities of hazardous reaction mixture would be present
27
28
29 during the reaction and could be rapidly quenched once the reaction stream exited the flow
30
31
32 reactor, thereby mitigating a portion of the safety risks associated with the process. The
33
34 continuous flow chemistry manufacturing mode would also enable processing facilities to
35
36
37 perform the process at about -30 °C with relatively inexpensive cooling units. An appropriately
38
39
40 designed flow reactor would improve mixing and cooling compared to batch mode chemistry,
41
42 both of which could impact the process (Table 1, entries 13 and 14). Lastly, the process was
43
44
45 believed to be viable for continuous flow chemistry as the reaction components were all clear
46
47
48
solutions (once dissolved in DCM) and the reaction mixture was not observed to become
49
50 heterogeneous throughout the course of the reaction.
51
52
53 The development of the continuous flow chemistry process was first investigated on a
54
55
VapourtecTM R-series system consisting of four HPLC pumps and 1/16” PFA tubing. The initial
56
57
58
59
11
1
2
3
configuration can be found in Figure 1, where compound 1 is sequentially combined with TFA,
4
5
6 followed by TMSOTf, and then TMSCN. Based on the fast reaction times observed during
7
8
9 development of the batch mode process, the residence times were initially set at 0.5 min, 0.4
10
11 min, and 1.8 min for the reaction of compound 1 with TFA, TMSOTf, and TMSCN,
12
13
14 respectively. These initial conditions resulted in high conversion and good diastereoselectivity
15
16
17 (Table 3, entry 1). It appeared that the Brønsted acid (TFA) was not necessary at small scale
18
19 for the continuous flow chemistry process (Table 3, entry 2). The results at larger scale
20
21
22 (discussed later) suggested that the presence of TFA was important for process robustness.
23
24
25
Evaluation of the residence times indicated that 0.5 min was optimal for the reaction with
26
27 TMSOTf, and that a 2 min residence time was sufficient for the reaction with TMSCN (Table
28
29
30 3, entry 2 vs entry 3–4). Decreasing the equivalents of reagents was found to negatively impact
31
32
reaction performance, with a decrease in both conversion and diastereoselectivity under the
33
34
35 flow chemistry conditions, regardless of whether TFA was present or absent (Table 3, entries
36
37
38 9, 10, and 11).
39
40 With acceptable continuous flow conditions established, an in-house model system was
41
42
43 assembled, consisting of three diaphragm metering pumps and 3/16” PFA tubing. The more
44
45
46 powerful diaphragm pumps (compared to the VapourtecTM HPLC pump system) allowed a
47
48 significant increase in throughput (ca. 317 g/h consumption of compound 1). Using this system,
49
50
51 the process was tested at 100 g input of compound 1, affording high diastereoselectivity (94:6)
52
53
and a solution purity of compound 2 of 93.5% (Table 3, entry 12). Subsequent isolation of the
54
55
56 quenched reaction stream afforded compound 2 in 84% yield and 99.8% purity. This larger
57
58
59
12
1
2
3
scale test demonstrated the viability of the continuous flow process and enabled the shift from
4
5
6 batch mode to continuous flow for future preparation of compound 2.
7
8
9 Figure 1. A Schematic Drawing of Continuous Flow Chemistry Setup
10
11
12 Compound 1 TMSOTf TMSCN
in DCM in DCM in DCM
13
14
15
Pre-cooling Pre-cooling Pre-cooling
16 Loop Lloop Loop
17
18
19 Pre-cooling
20 Loop
Reaction Reaction Reaction
21 Loop 1 Loop 2 Loop 3
22
23 TFA
Temperature controlled to -40 °C
24 in DCM KOH(aq) quench
25
26 Table 3. Development of Continuous Flow Chemistry Conditions
27
28
29 TFA / Residence Time (min)
30 In-Process Solution In-Process
TMSOTf / d.r.
31 Entry Content of 1 Purity of 2 Content of 3
32 TMSCN Reaction Reaction Reaction (2:3)a
(%)a (%)a (%)a
33 (equiv) Loop 1 Loop 2 Loop 3
34
35 1 3/6/6 0.5 0.4 1.8 0.5 89.3 6.3 93:7
36
37 2 0/6/6 — 0.5 2.0 0.6 93.4 5.6 94:6
38
39 3 0/6/6 — 0.5 3.9 1.3 92.1 6.0 94:6
40
41 4 0/6/6 — 2.5 2.0 2.1 91.6 5.8 94:6
42
43 5 0/6/6 — 2.4 3.7 1.7 90.5 7.1 93:7
44
45 6 0/6/6 — 0.5 2.0 4.0 81.5 9.0 90:10
46
47 7 0/6/6 — 0.4 1.8 3.8 87.4 7.8 92:8
48
49 8 0/6/6 — 2.1 1.8 0.3 77.8 6.4 92:8
50
51 9 1/3/3 0.8 0.7 2.9 3.7 81.9 13.1 86:14
52
53 10 0/3/3 — 0.7 2.9 4.7 69.1 15.3 82:18
54
55
56
57
58
59
13
1
2
3 11 2/2/2 0.5 0.4 1.9 5.7 74.4 10.8 87:13
4
5 12b 0/6/6 — 0.5 2.0 0.7 93.5 5.6 94:6
6
7
8 Reactions were performed at 1 g input of compound 1 on a VapourtecTM R-Series flow
9 chemistry system using 1/16” (OD) PFA tubing. Reaction components were combined
10 according to Figure 1. Residence times were achieved through increasing the volume of the
11
12 appropriate reaction loop while maintaining the flow rates constant. Reaction temperatures
13 were maintained at about -40 °C through immersion of the flow reaction loops in an IPA/dry
14 ice bath. Only steady state reaction streams were used to analyze in-process reaction profiles.
15 a Purity was determined by LC analysis of an aliquot of the quenched reaction mixture.
16
b Experiment was performed at 100 g input of compound 1 on an in-house build system
17
18 consisting of 3 diaphragm pumps and 1/4” (OD) PFA tubing. The quenched reaction stream
19
20
was isolated via crystallization to afford compound 2 in 84% yield and 99.8% purity by LC.
21
22 Kilo-Lab Scale Preparation of Compound 2 via Continuous Flow Chemistry
23
24
Following the successful 100 g demonstration with the in-house model continuous flow
25
26
27 chemistry system (Table 3, entry 12), an improved system was designed for the preparation of
28
29
30 20 kg of compound 2. The following criteria were considered necessary during the design of
31
32 this continuous flow reactor: improved safety for larger-scale manufacturing, compatibility of
33
34
35 the reagents and product with the materials of construction of the reactor system,
36
37
38 heating/cooling requirements, mixing requirements, and increased throughput relative to the
39
40 lab-scale system. A stainless-steel reactor was designed and fabricated (Figure 2a), which
41
42
43 improved the durability of the reactor and helped the prevention of potential loss of
44
45
46
containment. Stainless-steel (SS) was selected as the material of construction, as it was found
47
48 to be compatible with the individual reaction components or mixtures thereof (elemental
49
50
51 impurities due to leaching were not detected in the reaction streams by ICP-MS). With regards
52
53
to the cooling requirements for the continuous flow chemistry system, the heat of reaction
54
55
56 was found to be -133 kJ/mol for the reaction of compound 1 with TMSOTf, and -68 kJ/mol for
57
58
59
14
1
2
3
the reaction of this stream with TMSCN. These exotherms would give adiabatic temperature
4
5
6 rises of 9 K and 4 K, respectively, which could be easily controlled by bath cooling when using
7
8
9 the SS reactor. Quenching of the reaction mixture was accomplished in a semi-batch mode by
10
11 constant delivery of the reaction mixture into a caustic aqueous mixture. A complete system
12
13
14 that was assembled based on the established conditions is shown in Figure 2b.
15
16
17
Figure 2. Kilo Lab Flow Chemistry Equipment a) Flow Reactor. b) Kilo Lab Flow Reactor
18
19 Assembly
20
21
22 a) b
23 )
24
25
26
27
28
29
30
31
32
33
34
35 With the SS continuous flow system, an experimental run was performed with 100 g input
36
37
38 of compound 1 (without the use of TFA) to confirm the system performance. The results
39
40
41 obtained were within expectations, although a slight increase in the residual amount of starting
42
43 material (2.5%), along with a decrease in diastereoselectivity (91:9) (Table 4, entry 1). At the
44
45
46 time, these lower than expected results were attributed to diffusion at the beginning and end
47
48
49
of the experiment, as a significant amount of non-steady state reaction mixture was collected
50
51 for this smaller trial run using the SS reactor system. With the operation of the SS system
52
53
54 confirmed, the process was piloted at 2.75 kg scale (compound 1 input) using the preferred
55
56
57
58
59
15
1
2
3
conditions identified in Table 3, entry 12 with 0 equiv TFA, however, more pronounced
4
5
6 challenges were encountered. Incomplete conversion of the starting material and a significant
7
8
9 reduction in diastereoselectivity were observed (Table 4, entry 2). Investigation into
10
11 equipment setup and execution with the previously used systems (small-scale experiments
12
13
14 with the VapourtecTM system and 100 g experiments with either the in-house 3/16” PFA
15
16
17 reactor system or the SS reactor system) and the 2.75 kg pilot run with the SS reactor system,
18
19 when TFA was not added to the reaction mixture (Table 3 entries 11–12 and Table 4 entry 1
20
21
22 vs Table 4 entry 2), suggested that the reactor configuration and materials of construction were
23
24
25
not the cause of the change in reaction performance.
26
27 An alternative explanation for the discrepancy between the small-scale experiments and
28
29
30 2.75 kg pilot run was sought. It was hypothesized that residual water could have generated
31
32
trace amounts of TfOH through hydrolysis of TMSOTf. The generated TfOH could then act as
33
34
35 a catalyst, driving both the selectivity and the conversion of the transformation. This effect
36
37
38 would likely have a larger impact on smaller scale reactions (i.e., 1 g or even 100 g) with
39
40 decreasing impact as the scale increased (i.e., 2.75 kg). To test this hypothesis, TfOH was
41
42
43 introduced into the TMSOTf/DCM stock solution of a 150 g reaction using the SS reactor
44
45
46 system, and it was found that the conversion and diastereoselectivity were fully restored (Table
47
48 4, entry 3). The use of TFA in the reaction mixture was assessed further, and experiments at
49
50
51 150 g scale were conducted with the SS reactor system (Table 4, entries 4–6). It was found that
52
53
0.5 mol equiv TFA afforded acceptable conversion and selectivity during the process (Table 4,
54
55
56
57
58
59
16
1
2
3
entry 5). To accommodate potential fluctuations in the flow rates of the three-pump system
4
5
6 and increase the robustness of the process, 1.0 equiv TFA was selected for subsequent runs.18
7
8
9 During implementation of this change, the TFA/TMSOTf feed line experienced increased
10
11 back-pressure and ultimately plugged. After additional experimentation to evaluate the source
12
13
14 of the plugging, it was discovered that TFA (melting point of -15.4 °C) had limited solubility
15
16
17 in the TMSOTf/DCM mixture at -40 °C and was likely crystallizing during pre-cooling of the
18
19 feed line. To circumvent this issue, the TFA/TMSOTf/DCM feed pre-cooling loop was
20
21
22 removed. Additionally, the reaction temperature was adjusted from -40 °C to -30 °C to
23
24
25
decrease the risk of TFA crystallizing from the TFA/TMSOTf/DCM feed solution. These
26
27 changes did not have substantial impact on the performance of the reaction, as the solution
28
29
30 purity of compound 2 (and diastereoselectivity of the reaction) was similar despite the increase
31
32
in reaction temperature and removal of the TFA/TMSOTf/DCM pre-cooling loop (Table 4,
33
34
35 entries 5 and 6).
36
37
38 Table 4. Performance of the Stainless Steel Continuous Flow Reactor System
39
40
In-Process Solution In-Process
41 Reaction TFA / TMSOTf / d.r.
42 Entry Input of 1 Content of 1 Purity of 2 Content of 3
Temp (°C) TMSCN (equiv) (2:3)
43 (%) (%)a (%)a
44
45 1 100 g -40 0/6/6 2.4 87.7 8.8 91:9
46
47 2 2.75 kg -40 0/6/6 14.1 63.6 18.4 78:22
48
49
3b 150 g -40 0.5 / 3 / 3 1.4 86.8 7.7 92:8
50
51
52 4 150 g -40 0.1 / 6 / 6 3.9 77.3 13.3 85:15
53
54 5 150 g -30 0.5 / 6 / 6 1.2 86.4 7.1 92:8
55
56
57
58
59
17
1
2
3
6 150 g -30 1/6/6 0.9 88.7 7.9 92:8
4
5
6 Reactions were performed with the SS continuous flow reactor system displayed in Figure 2.
7 Reactions were performed with residence times of 0.5 min for the reaction of compound 1
8 with TFA/TMSOTf and 2.0 min for the reaction with TMSCN. Reaction temperatures were
9
10 maintained at about -40 °C through immersion of the flow reaction loops in an IPA/dry ice
11 bath. a Purity was determined by LC analysis of an aliquot of the quenched reaction mixture.
12 b TfOH was used in place of TFA.
13
14
Under these newly established conditions, the subsequent 5 runs with 2.75 kg input of
15
16
17 compound 1 using the SS reactor system provided high solution purity and diastereoselectivity
18
19
20 (Table 5, entries 2 to 6). The quenched effluents from the multiple runs were combined and
21
22 13.0 kg of compound 2 was isolated in 66% yield with 99.7% purity.
23
24
25 Table 5. In-Process Results for the Stainless Steel Continuous Flow Reactor Runs
26
27
28 In-Process Solution In-Process
TFA / TMSOTf / d.r.
29 Entry Scale (kg) Content of 1 Purity of 2 Content of 3
30 TMSCN (equiv) (2:3)
(%)a (%)a (%)a
31
32
1 0/6/6 2.75 14.1 63.6 18.4 78:22
33
34
35 2 1/6/6 2.76 3.3 88.3 5.0 95:5
36
37 3 1/6/6 2.75 2.4 91.3 4.7 95:5
38
39 4 1/6/6 2.75 0.4 93.0 5.6 94:6
40
41 5 1/6/6 2.75 2.1 90.8 5.7 94:6
42
43 6 1/6/6 2.76 0.7 92.6 5.6 94:6
44
45
46
Reactions were performed with the SS continuous flow reactor system displayed in Figure 2.
47 Reactions were performed with residence times of 0.5 min for the reaction of compound 1
48 with TFA/TMSOTf and 2.0 min for the reaction with TMSCN. Reaction temperatures were
49
maintained at about -30 °C through immersion of the flow reaction loops in an IPA/dry ice
50
51 bath. a Purity was determined by LC analysis of an aliquot of the quenched reaction mixture.
52
53 Plant Manufacturing Scale Preparation of Compound 2 via Continuous Flow Chemistry
54
55
56
57
58
59
18
1
2
3
The results from the gram scale to kilogram scale runs using the various continuous flow
4
5
6 reactor systems, along with the process parameters applied to each system served as the basis
7
8
9 for design of the plant continuous flow reactor system. The following major factors were used
10
11 to define the plant system.
12
13
14 First, the plant system needed to be well contained to ensure safe handling of the cyanide-
15
16
17 containing reaction mixture. To address this concern, the flow reactor and ancillary equipment
18
19 were constructed from stainless-steel and underwent rigorous pressure testing prior to use.
20
21
22 Second, the plant system needs to have the capacity to produce hundreds of kilograms of
23
24
25
compound 2 in reasonable time frame. Table 6 presents a comparison of process parameters for
26
27 the different systems. The interaction between residence time, reactor volume and throughput
28
29
30 (consumption of compound 1) can be clearly observed. From this table, it suggests that a
31
32
stainless-steel tube reactor with a volume of 2.5 L would achieve a throughput of
33
34
35 approximately 2 kg/h, enabling processing of 200 kg of compound 1 in about 4 days. This
36
37
38 duration was deemed acceptable for the manufacturing requirements. In order to consistently
39
40 achieve the desired flow rates of each of the three feeds, three diaphragm pumps constructed
41
42
43 of stainless steel and PTFE, equipped with back pressure regulators, were selected as the feed-
44
45
46 delivery devices and were calibrated prior to use.
47
48 Lastly, the plant system needs to have sufficient cooling capacity and efficiency. The cooling
49
50
51 capacity would be required both to adjust the reagent solutions to about -30 °C prior to
52
53
reaction, as well as rapidly scavenge heat generated during the reaction to ensure high-quality
54
55
56 product streams are generated. These factors were mitigated through use of the SS construction
57
58
59
19
1
2
3
of a tube-in-shell reactor system, enabling rapid heat transfer, along with use of a high capacity
4
5
6 heating/cooling unit.
7
8
9 Figure 3. A Schematic Drawing of the Plant Manufacturing Scale Continuous Flow Reactor
10
11
12 TFA / TMSOTf TMSCN
13 in DCM in DCM
14
15
16 Pre-cooling
Loop
17
18 Compound 1
in DCM
19
20
21 Pre-cooling Reaction Reaction
Loop Loop 1 Loop 2
22
23 Temperature controlled to -30 °C
24
KOH(aq) quench
25
26
27
Table 6. Key Parameters for Continuous Flow Reactor Systems
28
29
30 Residence Time
31 Reactor Volume
(min)
32 System Material of Throughput of
Entry
33 Description Construction Compound 1
Reaction Reaction Reaction Reaction
34
Loop 1 Loop 2 Loop 1 Loop 2
35
36
37 Vapourtec
1 PFA 0.5 2.0 2 mL 10 mL 9 g/h
38 R-Series
39
40 In-House PFA
2 PFA 0.5 2.0 59 mL 300 mL 317 g/h
41 Reactor System
42
43 Kilo-lab SS Stainless
3 0.5 2.0 60 mL 300 mL 317 g/h
44 Reactor System Steel
45
46 Plant Scale Stainless
4 0.5 2.0 0.5 L 2.0 L 1.98 kg/h
47 Reactor System Steel
48
49
50
51 Following the fabrication and installation of the plant scale continuous flow reactor system,
52
53
54 a pilot run was performed at 86 kg input of compound 1 to confirm operability of the plant
55
56
57
58
59
20
1
2
3
system for manufacturing of compound 2. As observable in Figure 4, after a brief induction
4
5
6 period whereby steady-state operation of the system was obtained, a consistent reaction profile
7
8
9 was observed throughout the duration of the flow chemistry run. Numerical results can be
10
11 found in Table 7 (entry 1), demonstrating the plant scale system afforded similar reaction
12
13
14 diastereoselectivity and solution purity of compound 2 when compared to the SS kilo lab
15
16
17 system (Table 7, entry 1 vs Table 5, entries 2–6). The quenched reaction mixture was carried
18
19 through an aqueous workup in the same fashion as the kilo lab batch (see experimental details)
20
21
22 to afford a stock solution of compound 2 in DCM. Following solvent exchange into toluene,
23
24
25
the product was crystallized and filtered to obtain 68 kg of compound 2 (Table 7, entry 1). This
26
27 delivery enabled preparation of remdesivir to support clinical evaluation of the drug candidate.
28
29
30 Two additional deliveries were subsequently performed using the same plant scale continuous
31
32
flow reactor system, at 280 kg and 250 kg input of 1, to provide 204 kg and 198 kg of compound
33
34
35 2, respectively (Table 7, entries 2 and 3).
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
21
1
2
3 Figure 4. 86 kg Flow Chemistry Reaction Profile
4
5
6
7 Flow Chemistry Reaction Profile
8 100
9
10 90
LC Purity of Major Reaction
11 80
Components (% area)
12 70
13 60
14
50
15 Compound 1
16 40
Compound 2
17 30
18 Compound 3
20
19
10
20
21 0
22 0 5 10 15 20 25 30 35
23 Flow Chemistry Run TIme
24
25 Reaction profile obtained through sampling an aliquot of the quenched reaction mixture and
26 LC analysis.
27
28 Table 7. In-Process and Final Results for Plant Scale Batches to Generate Compound 2.
29
30
31 Input In-Process Solution In-Process Output
32 d.r. Yield Purity
Entry Weight Content of Purity of Content of Weight of
33 (2:3) (%) (%)a
of 1 (kg) 1 (%)a 2 (%)a 3 (%)a 2 (kg)
34
35
36 1 86 0.9 88.6 6.1 94:6 68 78 99.6
37
38 2 280 0.7 89.6 5.9 94:6 204 72 99.7
39
40 3 250 0.5 89.8 5.9 94:6 198 78 99.9
41
42 Reaction Conditions: See experimental details for reaction and isolation conditions. a Purity
43 was determined by LC analysis.
44
45
Conclusion
46
47
48 The development of batch cyanation process to prepare compound 2 and enable
49
50 development of the clinical candidate remdesivir was described. Translation of the batch-mode
51
52
53 chemistry into a continuous flow process was accomplished, providing improved control over
54
55
the reaction conditions and increased diastereoselectivity. Four unique continuous flow
56
57
58
59
22
1
2
3
chemistry reaction systems were evaluated, including two systems that were utilized for large-
4
5
6 scale manufacture of compound 2. The preparation of about 500 kg of compound 2 through
7
8
9 continuous flow chemistry was reported, enabling preparation and clinical evaluation of
10
11 remdesivir for the treatment of Ebola virus and coronavirus diseases.
12
13
14
15 Experimental Details
16
17
18 Commercially available solvents and reagents were used as received without further
19
20
purification. LC data were collected on Waters Acuity or Waters Acuity H-Class UPLC
21
22
23 instruments with detection by UV. UPLC conditions were as follows: Acuity UPLC BEH C18,
24
25
26 130 Å, 1.7 μm, 2.1 x 100 mm, 75–5% gradient of 20 mM ammonium acetate at pH 9.5 in water
27
28 and 100% acetonitrile; flow rate 0.4 mL/min; acquisition time, 12 min; UV at 245 nm. Purity
29
30
31 was calculated through % area normalization.
32
33
34 Preparation of Compound 1
35 NH2
36
37 N
38 NH2 i) TMSCl, PhMgCl, N O
O BnO O
BnO N
39 i-PrMgCl, THF, -20 °C
OH
N
40 N
ii) NdCl3, n-Bu4NCl, 5, THF, -20 °C BnO OBn
41 I N BnO OBn
42 4 1 5
43
44
Anhydrous neodymium (III) chloride (1.0 equiv), n-tetrabutylammonium chloride (1.0 equiv),
45
46 and tetrahydrofuran (THF, 9 volumes) were combined and the resulting mixture was
47
48
49 concentrated to about 4.5 volumes under ambient pressure at about 66 °C. THF (4.5 volumes)
50
51 was charged and the concentration was repeated. The mixture was cooled to about 22 °C and
52
53
54 2,3,5-tri-O-benzyl-D-ribono-1,4-lactone (compound 5, 282 kg, 1.0 equiv) was charged. After
55
56
57
58
59
23
1
2
3
about 30 min the mixture was cooled to about -20 °C and held. In a separate reaction vessel, 7-
4
5
6 iodopyrrolo[2,1-f][1,2,4]triazin-4-ylamine19 (compound 4, 1.10 equiv) and THF (5 volumes)
7
8
9 were combined and cooled to about 0 °C. Chlorotrimethylsilane (1.10 equiv) was added slowly
10
11 and, after about 30 min the mixture was cooled to about -10 °C. Phenylmagnesium chloride (2
12
13
14 M in THF, 2.17 equiv) was added slowly and, after about 30 min, the reaction mixture was
15
16
17 cooled to about -20 °C. iso-Propylmagnesium chloride (2 M in THF, 1.13 equiv) was added
18
19 slowly and, after about 2 h, the Grignard reaction mixture was transferred into the
20
21
22 lactone/NdCl3/n-Bu4NCl/THF mixture. The resulting mixture was agitated at about -20 °C and,
23
24
25
after about 8 h, a solution of acetic acid (1 volume) and water (4 volumes) was added. The
26
27 resulting mixture was warmed to about 22 °C and iso-propylacetate (i-PrOAc) was added. The
28
29
30 layers were separated and the organic layer was washed sequentially with aqueous potassium
31
32
bicarbonate and aqueous sodium chloride. The organic layer was concentrated under vacuum
33
34
35 to about 4.5 volumes and i-PrOAc was charged. The organic layer was washed with water
36
37
38 twice and concentrated to about 4.5 volumes under vacuum. i-PrOAc was charged and the
39
40 concentration was repeated. The mixture was filtered and the filtrate was concentrated under
41
42
43 vacuum to about 3 volumes. Methyl tert-butyl ether (MTBE) was charged and the mixture was
44
45
46 adjusted to about 22 °C. Seed crystals of compound 1 were charged, followed by n-heptane,
47
48 and the mixture was cooled to about 0 °C. The solids were isolated by filtration and rinsed
49
50
51 forward with an MTBE/n-heptane mixture. The resulting solids were dried under vacuum to
52
53
afford compound 1 in 69% yield and 100% purity. Analytical data was consistent with data
54
55
56 previously reported in the literature.8
57
58
59
24
1
2
3
Preparation of Compound 2
4
5
Safety Notice: The below chemistry was conducted using TMSCN which may generate HCN
6
7
8 upon exposure to water, protic solvents, or in the presence of acids. HCN is a highly toxic,
9
10
11 colorless, flammable liquid that boils at 25.6 °C. The SDS should be consulted and, in
12
13 consultation with appropriate environmental, health and safety personnel, adequate
14
15
16 precautions should be taken. Additional precautions may include items such as the following:
17
18
19  written documentation outlining handling procedures for reagents, stock solutions,
20
21 process streams, waste, and loss of containment or exposure
22
23
24  higher level of personal protective equipment use
25
26
 cyanide detectors
27
28
29  trained response personnel with access to cyanide antidote
30
31
32 Batch process to prepare compound 2: To a solution of compound 1 (23 kg, 1.0 equiv) in
33
34 DCM (10 volumes) pre-cooled to about -40 °C was charged trifluoroacetic acid (3.0 equiv),
35
36
37 followed by a solution of TMSOTf (6.0 equiv) and TMSCN (6.0 equiv) in DCM (5 volumes),
38
39
40 pre-cooled to about -30 °C, while maintaining the internal temperature below about -25 °C.
41
42 The reaction mixture was agitated at below about -30 °C for no less than 10 minutes and
43
44
45 quenched into a pre-cooled (about -10 °C) solution of KOH (35.8 equiv) in water (14 volumes).
46
47
48
The bi-phasic mixture was warmed to ambient temperature. The organic layer was separated
49
50 and washed three times with aqueous sodium chloride. The organic phase concentrated under
51
52
53 vacuum to about 4 volumes, diluted with DCM (8 volumes) and concentrated under vacuum
54
55
to about 6 volumes. The mixture was filtered and the filtrate was concentrated to about 4
56
57
58
59
25
1
2
3
volumes. The concentrate was diluted with toluene (19 volumes) and concentrated under
4
5
6 vacuum to about 14 volumes at about 50 °C. The mixture was heated to about 90 °C, cooled to
7
8
9 about 55 °C, and seeded with compound 2. The mixture was agitated at about 55 °C for about
10
11 one hour and cooled to about 0 °C over about 6 hours. The solids were isolated by filtration
12
13
14 and the filter cake was washed with toluene (3 volumes). The solids were dried under vacuum
15
16
17 at about 50 °C to afford compound 2 in 71% yield and 99.2% purity.
18
19 Continuous flow chemistry process to generate compound 2: Stock solutions of compound 1
20
21
22 (250 kg, 1.0 equiv) in DCM (15.0 volumes) (Feed 1), TMSOTf (6.0 equiv) and TFA (1.0 equiv)
23
24
25
in DCM (4.4 volumes) (Feed 2), and TMSCN (6.0 equiv) in DCM (4.5 volumes) (Feed 3), were
26
27 prepared in separate reactors or feed vessels. Feed 1 was pumped at a flow rate of
28
29
30 approximately 504 mL/min through a pre-cooling loop at about -30 °C, and Feed 2 was pumped
31
32
at a flow rate of approximately 207 mL/min. Feeds 1 and 2 were combined in Reaction Loop
33
34
35 #1 at about -30 °C for about 30 seconds. The effluent was then combined with Feed 3 (pumping
36
37
38 at approximately 189 mL/min through a pre-cooling loop at about -30 °C) in Reaction Loop #2
39
40 at about -30 °C for about 2 minutes. The effluent of the combined feeds was collected directly
41
42
43 into a vessel containing a solution of KOH (19.7 equiv) in water (8 volumes) at about -10 °C.
44
45
46 The mixture was adjusted to about 22 °C, then 2-propanol was added and the layers were
47
48 separated. The organic layer was washed with aqueous sodium chloride twice and
49
50
51 concentrated. The resulting solution was filtered. Toluene was charged to the filtrate and the
52
53
mixture was concentrated. The mixture was heated to about 55 °C, then cooled to about 0 °C.
54
55
56 The resulting slurry was filtered, rinsed with toluene and dried at about 60 °C to afford
57
58
59
26
1
2
3
compound 2 in 78% yield with 99.9% purity. Analytical data were consistent with data
4
5
6 previously reported in the literature.8 Waste streams from the cyanation process (both
7
8
9 laboratory and plant scale) were monitored for cyanide content and were treated with bleach
10
11 until cyanide levels were below the limit of detection by cyanide test strips.
12
13
14 AUTHOR INFORMATION
15
16
17 Corresponding Authors
18
19 *E-mail: tiago.vieira@gilead.com, andrew.stevens@gilead.com
20
21
22
23 ACKNOWLEDGMENT
24
25
26 The authors would like to thank Anna Chiu, Dragos Vizitiu, Jeff Ng, Kevin Allan, Mark
27
28 Scott, Richard Yu, and Zhongxin Zhou for thoughtful discussions and feedback during
29
30
31 manuscript preparation.
32
33
34
35
36
37
References
38
39
40 ( ) Warren, T. K.; Jordan, R.; Lo, M. K.; Ray, A. S.; Mackman, R. L.; Soloveva, V.; Siegel, D.;
41
42
43 Perron, M.; Bannister, R.; Hui, H. C.; et al. Therapeutic efficacy of the small molecule GS-5734
44
45 against Ebola virus in rhesus monkeys. Nature, 2016, 531, 381–385.
46
47
48
49 (2) Sheahan, T. P.; Sims, A. C.; Graham, R. L.; Menachery, V. D.; Gralinski, L. E.; Case, J. B.;
50
51 Leist, S. R.; Pyrc, K.; Feng, J. Y.; Trantcheva, I.; et al. Broad-spectrum antiviral GS-5734 inhibits
52
53
54
55
56
57
58
59
27
1
2
3
both epidemic and zoonotic coronaviruses. Sci. Transl. Med., 2017, 396, doi:
4
5
6 10.1126/scitranslmed.aal3653.
7
8
9
10
(3) Wang, M.; Cao, R.; Zhang, L.; Yang, X.; Liu, J.; Xu, M.; Shi, Z.; Hus, Z.; Zhong, W.; Xiao,
11
12 G. Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus
13
14
15 (2019-nCoV) in vitro. Cell Research, 2020, DOI: 10.1038/s41422-020-0282-0.
16
17
18 (4) Holshue, M. L.; DeBold, C.; Lindquist, S.; Lofy, K. H.; Wiesman, J.; Bruce, H.; Spitters, C.;
19
20
21 Ericson, K.; Wilkerson, S.; Tural, A.; et al. First Case of 2019 Novel Coronavirus in the United
22
23
24
States. N. Engl. J. Med., 2020, DOI: 10.1056/NEJMoa2001191
25
26
27 (5) Grein, J.; Ohmagari, N.; Shin, D.; Diaz, G.; Asperges, E.; Castagna, A.; Feldt, T.; Green, G.;
28
29
30 Green, M. L.; Lescure, F.-X.; et al. Compassionate use of remdesivir for patients with severe
31
32 COVID-19. N. Eng. J. Med., 2020, DIO: 10.1056/NEHMoa2007016.
33
34
35
36 (6) Gilead Sciences Statement on the Company’s Ongoing Response to the 2019 Novel
37
38 Coronavirus. Gilead Sciences Inc. Press Release, January 31, 2020.
39
40
41 https://www.gilead.com/news-and-press/company-statements/gilead-sciences-statement-on-
42
43
44 the-company-ongoing-response-to-the-2019-new-coronavirus (accessed 2020-02-05).
45
46
47 (7) Gilead Sciences Initiates Two Phase 3 Studies of Investigational Antiviral Remdesivir for
48
49
50 the Treatment of COVID-19. Gilead Sciences Inc. Press Release, February 26, 2020.
51
52 https://www.gilead.com/news-and-press/press-room/press-releases/2020/2/gilead-sciences-
53
54
55
56
57
58
59
28
1
2
3
initiates-two-phase-3-studies-of-investigational-antiviral-remdesivir-for-the-treatment-of-
4
5
6 covid-19 (accessed 2020-03-05).
7
8
9
10
(8) Siegel, D., Hui; H. C., Doerffler, E.; Clarke, M. O.; Chun, K.; Zhang, L.; Neville, S.; Carra,
11
12 E.; Lew, W.; Ross, B.; et al. Discovery and Synthesis of a Phosporamidate Prodrug of a
13
14
15 Pyrrolo[2,1-f][triazin-4-amino] Adenine C-Nucleoside (GS-5734) for the Treatment off Ebola
16
17
and Emerging Viruses. J. Med. Chem. 2017, 6, 1648–1661.
18
19
20
21 (9) Mukaiyama, T.; Kobayashi, S. A convenient synthesis of C-α-D-ribofuranosyl compounds
22
23
24
from 1-O-acetyl-2,3,5-tri-O-benzyl-β-D-ribose by the promotion of triphenylmethyl
25
26 perchloriate. Carbohydr. Res. 1987, 171, 81–87.
27
28
29
30 (10) Xolin, A.; Stévenin, A.; Pucheault, M.; Norsikian, S.; Boyer, F.-D.; Beau, J.-M.
31
32 Glycosylation with N-acetyl glycosamine donors using catalytic iron(III) triflate: from
33
34
35 microwave batch chemistry to a scalable continuous-flow process. Org. Chem. Front. 2014, 1,
36
37
38
992–1000.
39
40
41 (11) Monteiro, J. L.; Pieber, B.; Corrêa, A. G.; Kappe, C. O. Continuous synthesis of hydantoins:
42
43
44 Intensifying the Bucherer-Bergs reaction. Synlett 2016, 27, 83–87.
45
46
47 (12) Brahma, A.; Musio, B.; Ismayilova, U.; Nikbin, N.; Kamptmann, S. B.; Siegert, P.; Jeromin,
48
49
50 G. E.; Ley, S. V.; Pohl, M. An orthogonal biocatalytic approach for the safe generation and use
51
52 of HCN in a multistep continuous preparation of chiral O-acetylcyanohydrins. Synlett, 2016,
53
54
55 27, 262–266.
56
57
58
59
29
1
2
3
(13) Hessel, V. Novel Process Windows – Gate to Maximizing Process Intensification via Flow
4
5
6 Chemistry. Chem. Eng. Technol. 2009, 32, 1655–1681.
7
8
9
10
(14) Gutmann, B.; Cantillo, D.; Kappe, C. O. Continuous-Flow Technology – a Tool for the
11
12 Safe Manufacturing of Active Pharmaceutical Ingredients. Angew. Chem., Int. Ed. 2015, 54
13
14
15 (23), 6688–6728.
16
17
18 (15) Hartman, T. L.; McMullen, J. P.; Jensen, K. F. Deciding Whether to go with the Flow:
19
20
21 Evaluating the Merits of Flow Reactors for Synthesis. Angew. Chem., Int. Ed. 2011, 50 (33),
22
23
24
7502–7519.
25
26
27 (16) Plutschack, M. B.; Pieber, B.; Gilmore, K.; Seeberger, P. H. The Hitchhiker’s Guide to
28
29
30 Flow Chemistry. Chem. Rev. 2017, 117 (18), 11796–11893.
31
32
33 (17) Baumann, M.; Moody, T. S.; Smyth, M.; Wharry, S. A Perspective on Continuous Flow
34
35
36 Chemistry in the Pharmaceutical Industry. Org. Process Res. Dev. 2020, DOI:
37
38 10.1021/acs.oprd.9b00524
39
40
41
42 (18) TFA was selected over TfOH due to the less corrosive nature and lower cost.
43
44
45 (19) Alternatively, 7-bromopyrrolo[2,1-f][1,2,4]triazin-4-ylamine can be used in place of
46
47
48 7-iodopyrrolo[2,1-f][1,2,4]triazin-4-amine under the same reaction conditions.
49
50
51
52
53
54
55
56
57
58
59
30

10 1021@acs Oprd 0c00172 PDF

Uploaded by

Copyright:

Available Formats

You might also like

10 1021@acs Oprd 0c00172 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 1021@acs Oprd 0c00172 PDF

Uploaded by

Copyright:

Available Formats

Subscriber access provided by Uppsala universitetsbibliotek

is published by the American Chemical Society. 1155 Sixteenth Street N.W.,

You might also like