Handling Hydrogen Peroxide Oxidations on a Large Scale: Synthesis of 5-Bromo-2-nitropyridine

Article abstract of DOI:10.1021/acs.oprd.6b00433

5-Bromo-2-nitropyridine was prepared from the corresponding amine via hydrogen peroxide oxidation in large scale production. This transformation initially showed low conversion, high impurity content and lack of reproducibility in lab trials. Parallel to process development, safety studies were conducted to investigate the stability of oxidant mixture, its composition and the oxidation reaction itself by reaction and adiabatic calorimetry. The resulting robust reaction conditions and appropriate safety boundaries allowed to develop a reproducible, safe protocol for the implementation of this chemistry on large scale, obtaining consistent results throughout the campaign.

Full text of DOI:10.1021/acs.oprd.6b00433

Subscriber access provided by Fudan University
Communication
Handling hydrogen peroxide oxidations on large
scale: synthesis of 5-bromo-2-nitropyridine
Alessandro Agosti, Giorgio Bertolini, Giacomo Bruno,
Christian Lautz, Thomas Glarner, and Walter Deichtmann
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00433 • Publication Date (Web): 07 Feb 2017
Downloaded from http://pubs.acs.org on February 7, 2017
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 free 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 accessible to all
readers and 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.
Organic Process Research & Development 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 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Handling hydrogen peroxide oxidations on large
scale: synthesis of 5ꢀbromoꢀ2ꢀnitropyridine
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
60
Alessandro Agosti‡*, Giorgio Bertolini‡, Giacomo Bruno‡, Christian Lautz# ,Thomas Glarner#,
Walter Deichtmann#
‡OLON SpA, Segrate (MI), Italy
# F. HoffmannꢀLa Roche AG, CHꢀ4070, Basel, Switzerland
ABSTRACT
5ꢀBromoꢀ2ꢀnitropyridine was prepared from the corresponding amine via hydrogen peroxide
oxidation in large scale production. This transformation initially showed low conversion, high
impurity content and lack of reproducibility in lab trials. Parallel to process development, safety
studies were conducted to investigate the stability of oxidant mixture, its composition and the
oxidation reaction itself by reaction and adiabatic calorimetry. The resulting robust reaction
conditions and appropriate safety boundaries allowed to develop a reproducible, safe protocol for
the implementation of this chemistry on large scale, obtaining consistent results throughout the
campaign.
KEYWORDS
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 32
1
2
3
4
5
6
oxidation, hydrogen peroxide, nitroꢀaromatic, RC1 calorimetry, ARC, thermal stability,
process safety
7
8
9
INTRODUCTION
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
60
Oxidation reactions represent an important tool available for the synthetic organic chemist.
Albeit a large number of methods and reagents are present in the literature, their application on
large scale still represents a challenge1. The main reason might lie in the intrinsic danger
associated with a class of reactions in which oxidant, fuel and energy are present at the same
time. Among the various oxidant alternatives, the use of hydrogen peroxide2 can certainly be
considered advantageous for its low cost and environmentally benign byꢀproducts. Nevertheless,
hydrogen peroxide must be handled with care as its highly exothermic decomposition is
associated with large amounts of gas and steam generation with consequent overpressurization.
Moreover, peroxideꢀbased oxidants are known to interact with metal ion contaminants, which
catalyze their decomposition as witnessed by several accidents reported in the literature3. In case
of oxidation of (hetero)aromatic amine to the corresponding (hetero)aromatic nitro derivative, an
additional problem is often caused by the generation of unstable functional groups which might
be present in the reaction mixtures (e.g. partially oxidized Nꢀcontaining compounds such as Nꢀ
oxide, diazo and the like). For these reasons, a thorough safety investigation needs to be
conducted already at small preparative scale and certainly prior to scaleꢀup4.
In this paper, we describe the methodological work that led, within a small
timeframe, to the development and industrialization of the intrinsically dangerous conversion of
5ꢀbromoꢀ2ꢀaminopyridine into the corresponding 5ꢀbromoꢀ2ꢀnitropyridine by means of hydrogen
peroxide oxidation. To ensure a rigorous safety approach it was necessary to establish the
boundaries within which the process could be safely operated. The systematic approach
ACS Paragon Plus Environment
2

Page 3 of 32
Organic Process Research & Development
1
2
developed by Gygax5 with the definition of the “cooling failure scenario” and applied by
Stoessel6 in the Criticality Classes categorization was very helpful. The definition of the
following parameters helps in categorizing the process from Criticality Class 1 to Criticality
Class 5 (with 5 being the most dangerous. Figure 1):
3
4
5
6
7
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
60
1) Tp: Process Temperature (temperature at which the process is operated under desired
conditions).
2) ∆Tad. : Adiabatic Temperature Rise of the synthesis reaction (temperature increase
under adiabatic conditions in case of cooling failure).
3) MTSR: Maximum Temperature of the Synthesis Reaction (Maximum temperature
attainable under adiabatic conditions after cooling failure; TP + ∆Tad ).
4) MTT: Maximum Temperature for Technical Reasons (The maximum temperature that
the system can tolerate. In case of reactions run at atmospheric pressure the boiling
point of the solvent is used).
5) TMR: Time to Maximum Rate (time required to reach maximum decomposition rate
under adiabatic conditions).
6) TD,24: Temperature at which the Time to Maximum Rate equals to 24 hours.
This methodology allows, through a set of questions and experimental investigations, to
numerically quantify, for the decomposition reaction, the severity (defined as ∆Tad., dec.) and the
probability (defined as TMR). A scenario can therefore be constructed in which, in case of
cooling failure, the desired exothermic reaction will result in temperature rising from TP to the
Maximum Temperature of the Synthesis Reaction (MTSR= TP + ∆Tad). At MTSR a further
decomposition may occur adiabatically with a certain time span before completion (TMR). The
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 32
1
2
3
4
5
6
safety limit can arbitrarily be defined as the temperature at which the TMR is equal to 24h
(TD,24).
7
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
60
Figure 1: Criticality Class description. Reproduced with permission from Ref.6. Copyright
(2008, WILEYꢀVCH)
The development of the process conditions and the safety investigation needed to be conducted
in parallel in order to meet the following criteria:
• Decompositions, if any, should take place at temperatures high enough to allow a
sufficient time frame (Time to Maximum Rate) for intervention before complete
decomposition.
• Synthetic reactions will take place with conversion rates high enough to prevent
reagent accumulation and consequently heat accumulation, a threat in case of loss of
cooling.
RESULTS AND DISCUSSION
Synthesis
A two step procedure was utilized for the synthesis of the target compound 5ꢀbromoꢀ2ꢀ
nitropyridine (Scheme 1). The use of an unusual oxidation of (hetero)aromatic amine to the
correspondent (hetero)aromatic nitro derivative was dictated by the required substituents position
on the pyridine ring.
ACS Paragon Plus Environment
4

Page 5 of 32
Organic Process Research & Development
1
2
3
4
Scheme 1: Reaction sequence for the synthesis of 5ꢀbromoꢀ2ꢀnitropyridine (3)
5
6
7
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
60
Nꢀbromosuccinimide (NBS) bromination of 2ꢀaminopyridine 1 in acetonitrile at 0ꢀ5 °C delivered
5ꢀbromoꢀ2ꢀaminopyridine with good regioselectivity (> 20:1). The brominated aminoꢀpyridine 2
was then reacted with the oxidant mixture yielding the desired compound 3 after dilution with
water and isopropanol.
Bromination reaction
The first step required the bromination of 2ꢀaminopyridine 1. Literature already described this
8
9
transformation with several reagents including NBS7, Br2 and NaBr/NaBrO3 with H2SO4 in
various solvents including acetonitrile, dimethylformamide and acetone. To the best of our
knowledge, no large scale synthesis of this molecule has been reported so far. We chose to use
NBS as reagent combination with acetonitrile as solvent/reagent combination due to the inherent
stability of the acetonitrile/Nꢀbromosuccinimide mixture.10 Moreover, handling of solid NBS on
large scale was deemed to be favored over molecular bromine.
First attempts were conducted by addition of solid NBS (1.05 mol equiv. with respect to 2ꢀ
aminopyridine) to a cooled (ꢀ5 °C) solution of starting material in acetonitrile (10 v/w with
respect to 2ꢀaminopyridine). As quite foreseeable, polybromination was observed and LC/MS
analysis confirmed the presence of a dibromoꢀamine (up to 15% area HPLC) whose structure
was assigned to 3,5ꢀdibromoꢀ2ꢀaminopyridine 4. Reaction calorimetry (Figure 2) indeed showed
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 32
1
2
3
4
5
6
7
8
9
that addition of a single portion of NBS resulted in an instantaneous heat generation. The sharp
heat spike together with the potential presence of large concentration gradients are the likely
cause of high level of impurity. Moreover, after approximately 70% of the total amount of
brominating agent was added, a precipitate formed (identified as succinimide byꢀproduct)
resulting in a thick suspension and consequently difficult heat exchange. All these observations
made the application on large scale, where both stirring and heat dissipation are more difficult,
undesirable.
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
60
Figure 2. Reaction calorimetry NBS solid addition
In order to overcome the problem and supported by the already mentioned stability of NBS
solution in acetonitrile, we modified the protocol to accommodate the addition of the
brominating agent (1.05 mol equiv. with respect to 2ꢀaminopyridine) as acetonitrile solution (6
w/w with respect to NBS) to a solution of 2ꢀaminopyridine in the same solvent (1.6 w/w with
respect to 2ꢀaminopyridine) in approximately 1h (Figure 3).
ACS Paragon Plus Environment
6

Page 7 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
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
60
Note: end of addition corresponding to 1.05 equiv. of NBS
Figure 3. Reaction calorimetry NBS addition as acetonitrile solution
Lower amounts of the dibromoimpurity (4% area HPLC vs. 14% area HPLC when solid NBS
used) were obtained when this protocol was followed and a higher temperature could be used (2ꢀ
5 °C vs. previous ꢀ5 to ꢀ8 °C) (Table 1).
Table 1. Comparison of impurity profile for NBS addition as a solid and as solution in
acetonitrile
Trial
Reaction
temperature
Dibromo impurity
Reaction end
Dibromo
impurity
Isolated
yield
(°C)
Isolated solid
(% area HPLC)
0.87 %
(% mol)
(% area HPLC)
Solid NBS addition
ꢀ10 to ꢀ5
0 to +5
14 %
4 %
75.3 %
77.8 %
NBS acetonitrile
solution
0.05 %
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 32
1
2
3
4
5
6
7
8
9
The importance of efficient heat removal in the bromination reaction was supported by the fact
that the formation of dibromo impurity increased with larger batch sizes, a peculiar behavior
already evident on laboratory scale (0.3% HPLC area dibromoꢀimpurity on 15 g vs. 4.9% HPLC
area dibromoꢀimpurity on 140 g scale).
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
60
After the addition of small amount of water (2.5 w/w) to solubilize the succinimide byproducts
and charcoal treatment to remove discoloration, the product was isolated by concentration under
reduced pressure of the reaction mixture to low volume (2 vol. with respect to the starting 2ꢀ
aminopyridine). The isolation continued with addition of water and distillation to partially
remove the remaining acetonitrile, followed by addition of water to achieve a total of 5 v/w (with
respect to the starting 2ꢀaminopyridine) and a composition of approximately 4:1 w/w ratio of
water and acetonitrile. Water positively influenced the process in several ways as it allowed the
complete solubilization of the succinimide (byꢀproduct of NBS bromination) and decreased the
solubility of the desired compound (2ꢀaminoꢀ5ꢀbromopyridine 2 is 10 times less soluble in 4:1
water/acetonitrile compared to pure acetonitrile), also creating a favorable azeotropic mixture for
acetonitrile removal.11
From a process safety perspective, it was necessary to investigate the stability of the
concentrated reaction mixture by Accelerating Rate Calorimetry (ARC).12 The decomposition
onset was detected at 195 °C (∆Tad: +132 K, corrected for Φꢀfactor13: 2.37) and the calculated
TD,24 was 165ꢀ175 °C. Considering the process temperature of 2ꢀ5 °C and the adiabatic
temperature rise related to the batch process (undesired addition of all the reagent at once), the
transformation (synthesis and isolation) can be assigned to a Criticality Class 1 (Table 2).
Table 2: Process parameters for Criticality Class definition. Bromination reaction and isolation
Parameter
Value
Comments
ACS Paragon Plus Environment
8

Page 9 of 32
Organic Process Research & Development
1
2
3
4
5
6
TP
2ꢀ5 °C
36 °C
Process defined
From reaction calorimetry,
worst case batch addition
MTSR
7
8
9
MTT
82 °C
Acetonitrile boiling point
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
60
TD,24
165ꢀ175 °C
ARC on concentrated mixture
Criticality Class: 1
Introduction of oxidation reaction
In order for the reaction to take place, the amount of oxidant necessary to convert the amino
group into the corresponding nitro group equals to a minimum of 3:1 molar ratio of hydrogen
peroxide/amine. To the best of our knowledge, no large scale examples of this transformation
have been reported. Nevertheless, various literature examples exist involving the use of hydrogen
peroxide 30% and concentrated sulfuric acid (>95%) in various w/w ratios. The addition modes
reported are either by preꢀmixing the H2O2 with sulfuric acid and adding a solution of amine in
H2SO4 or adding the preꢀmixed oxidant on the amine solution.15 In general terms, it would be
14
advisable to add the oxidant (ideally only hydrogen peroxide) to the amine solution so that a
potential exothermic event could be controlled by the oxidant dosing. In order to understand the
energetic potential of the oxidation reaction, adiabatic calorimetry was conducted by first
simulating a reaction in which all the reagent is added in a single portion (Figure 4).
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 32
1
2
3
4
5
6
7
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
60
Figure 4. ARC experiment, “batchꢀtype” oxidation
An undesired “batchꢀtype” reaction will quite clearly result in large heat generation (∆Tad: +107
K, corrected for Φꢀfactor: 1.85) with large Self Heating Rate (17 °C/min, corrected for Φꢀfactor:
1.85) and pressure generation rate (2 bar/min). The rate of this reaction is such that the TMR at
30 °C is 35 min. This is indicative of the energy inherent to the reaction and the potential
consequences of an unoptimized oxidation process are clearly unacceptable.
An initial set of experiments in which hydrogen peroxide 30% was added to the amine was
designed. A solution of 5ꢀbromoꢀ2ꢀaminopyridine 2 in sulfuric acid 98% (4ꢀ6 v/w with respect to
starting amine) was prepared and the H2O2 30% solution (1.3 equiv. corresponding to 3.6 mol
ratio) added in 1h while maintaining the temperature below 20 °C. Much to our regret, no
conversion was observed at the end of the addition and full conversion necessitated 40 h at 20
°C. When the reaction temperature was increased to 30 °C, the reaction was completed after 8ꢀ10
h. Furthermore, the formation of red fumes was observed. These vapors were found to contain
ACS Paragon Plus Environment
10

Page 11 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
molecular bromine which was identified by reduction with thiosulfate solution and precipitation
of the corresponding AgBr through silver nitrate addition. We presumed that a competitive
decomposition pathway might lead to generation of Br2 through oxidative decomposition of the
pyridine ring into low molecular weight fragments, in turn being oxidized liberating molecular
bromine. This hypothesis was reinforced by the observation that large amounts of dibromoꢀ
pyridine are formed although 5ꢀbromoꢀ2ꢀaminopyridine 2 free from dibromo impurity was used
as starting material (Scheme 2). The reaction product proved stable under the oxidation
conditions probably due to the electronwithdrawing nature of the nitro group, therefore the
decomposition is most likely occurring on the starting material generating unidentified low
molecular weight fragments.
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
60
Scheme 2.Proposed oxidative decomposition pathway
These reaction conditions were considered unacceptable because the reaction proceeded with
large reagent accumulation (little to no conversion at the end of the addition) and large amounts
of the starting material was decomposed. Moreover, the mass balance of the reaction was only
50%.
As already mentioned, another possibility is to use a preꢀmixed solution of H2SO4 95ꢀ98% /
H2O2 30% in a w/w ratio between 1:1 to 6:1 (also known as Caro’s acid or Piranha solution)
whose active oxidizing species is considered to be peroxosulfuric acid (H2SO5). Indeed, a
preliminary experiment using a H2SO4 95ꢀ98% / H2O2 30% mixture (1.3 equiv. of oxidant) was
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 32
1
2
3
4
5
6
7
8
9
conducted. Addition of amine solution over the oxidant in 1h showed promising results with only
22% of starting material left at the end of the addition. Furthermore, the extent of bromine fumes
and consequently the amount of dibromo impurity 4 was greatly reduced (5 – 8% HPLC area at
the end of addition vs. 18% area HPLC under the previous conditions).
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
60
Role of sulfuric acid in the oxidation
Encouraged by these results, we modified the reaction conditions in order to increase the
amount of oxidant to 2.2 mol equiv. (always considering the preꢀmixed oxidant as reagent of
choice for the reasons described in the previous paragraph) and increasing the addition time to 6
h, resulting in complete conversion of the starting material. 5ꢀbromoꢀ2ꢀaminopyridine 2 was
dissolved in 7.5 w/w sulfuric acid. The optimal ratio of sulfuric acid 98% and H2O2 30% was
experimentally found to be ≥ 5.6:1 w/w. This corresponds to a final sulfuric acid concentration
of 82.6% w/w after mixing. A lower ratio, corresponding to more diluted sulfuric acid, invariably
resulted in slow reactions and higher sideꢀproduct formation. This might be explained with the
need of having sulfuric acid with a concentration high enough to form the active H2SO5 species
and act as a dehydrating agent as well. We performed calorimetric experiments in order to
measure the amount of heat released when mixing H2SO4 with either hydrogen peroxide 30%
solution or with the same amount of pure water. We then compared the results with calculated
values for the hydration of sulfuric acid at different concentrations.16 By considering the
following equation for the formation of Caro’s acid (H2SO5):
H2SO4 + H2O2 ꢀ H2SO5 + H2O
we can calculate the hydration enthalpy in both experiments and compare it with the
experimental data (Table 3).
Table 3: enthalpy of mixing of sulfuric acid with pure water and 30% H2O2 solution
ACS Paragon Plus Environment
12

Page 13 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
H2O addition
H2O2 addition
85.1 %
82.9 %
Final H2SO4 concentration
Calculated enthalpy (@25 °C)
Measured enthalpy (@10 °C)
25.53 kJ/mol H2SO4
25.99 kJ/mol H2SO4
22.61 kJ/mol H2SO4
22.53 kJ/mol H2SO4
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
60
By comparing the data, it appears evident that the H2SO5 hypothesis holds true and, therefore, we
may consider the peroxosulfuric acid to be the species present in solution. Nevertheless it is fair
to say that this only represents an indirect proof of the formation of said species.
The critical role played by the sulfuric acid was also supported by the following experimental
observations:
1) Direct addition of hydrogen peroxide 30% over a solution of 5ꢀbromoꢀ2ꢀaminopyridine 2
resulted in a slower reaction, as previously mentioned, and in high impurity levels.
2) When more diluted sulfuric acid (50ꢀ80%) was used, reactions invariably met with failure.
Finally, a reaction run with sulfuric acid/hydrogen peroxide ratio (3:1 vs. 5.6:1 w/w), a more
diluted oxidant solution, produced a large amount of an impurity which was isolated from the
mother liquor and identified as the “azoꢀoxy” derivative in Scheme 3. Furthermore, an original
sample of this impurity was synthetically prepared and characterized (see Supporting
Information).
Scheme 3. Structure of azoꢀoxy impurity
Taken all together these considerations are the foundation of our reaction optimization and are
closely linked with the safety studies.
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 32
1
2
3
4
Safety investigation of the oxidation reaction
5
6
7
8
It appears evident that the preparation and the use of the Piranha solution, even at laboratory
scale, requires great attention. A thorough investigation must include the thermal stability of
reagents, reaction intermediates and product together with reaction calorimetry to assess the
thermal behavior of the desired reaction and the potential presence of heat accumulation.
Furthermore, the stability of the oxidant mixture must be evaluated under different process
conditions in order to incorporate process deviations such as the presence of metals (rust and the
like) and organic residues. The latter are known to be extremely dangerous when hydrogen
peroxide oxidations are run on large scale (see Reference 2). We therefore embarked into
detailed investigations whose findings are presented here.
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
60
Stability of 5ꢀbromoꢀ2ꢀaminopyridine solution and its preparation
Stability of pure 5ꢀbromoꢀ2ꢀaminopyridine 2 solid was studied by DSC analysis showing an
onset of 237 °C with a decomposition energy of 353 J/g. The 5ꢀbromoꢀ2ꢀaminopyridine solution,
analyzed by DSC, showed good thermal stability with decomposition onset at 362 °C. Reaction
calorimetry, conducted on the preparation of amine solution in sulfuric acid, showed the typical
pattern of an acid/base reaction. The total heat of reaction was calculated to be 119 kJ/molpyridine 2
corresponding to an adiabatic temperature rise of +53 K. Clearly, the preparation of the solution
will be made by portionwise addition of the solid (no less than 10 aliquots) in order to maintain
the temperature within the range (20 ± 10 °C).
Preparation and stability of H2SO4 / H2O2 solution
Quite intuitively, the handling of the oxidizing reaction represents the most daunting challenge
of the project and the actual scaleꢀup depended on the possibility to establish reasonable margins
of safety for this step and for the subsequent oxidation reaction. After having determined the
ACS Paragon Plus Environment
14

Page 15 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
necessary amounts of acid and oxidant, we began to study the preparation of the oxidant solution
by reaction calorimetry. The scope of the experiment is twoꢀfold: firstly the knowledge of the
presence of potential heat accumulation was deemed mandatory; secondly knowing the actual
amount of heat released (in terms of total heat and in terms of the heat of reaction per unit
volume, respectively) was necessary in order to match them with the cooling capacity of the
industrial reactor.
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
60
Two reaction calorimetry experiments were conducted by modifying the addition rate (30 min
and 40 min) of hydrogen peroxide solution (Figure 5, only one addition time experiment
reported).
Figure 5: Reaction calorimetry of H2SO4/H2O2 reaction mixture preparation. 40 min addition
time
Note: end of addition to 5.6/1 w/w ratio of H2SO4/H2O2
Table 4: Relevant reaction calorimetry parameters for H2SO4 / H2O2 solution preparation
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 32
1
2
3
4
5
6
7
8
9
T
(°C)
10
Dosing
(min)
40
Max.T
(°C)
118
ꢀHrx.
(kJ/mol)
138
ꢀTad.
(°C)
108
ꢀTad.corr.
% acc.
(°C)
0
1
/
10
30
141
114
124
1
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
60
No heat of accumulation was detected and both experiments showed the typical heat flow
profile of an instantaneous reaction with heat flow decreasing due to dilution. The heat of
reactions was 138ꢀ141 kJ/molperoxide respectively. In case of an unwanted batch addition and loss
of cooling, the adiabatic temperature rise would be close to 120 °C, causing unacceptable
consequences. In order to meet the average cooling capacity of an industrial glass lined reactor
(approx. 15 W/L), the addition time in plant should be between 9ꢀ11 h with the possibility to
increase the addition rate towards the end of the addition due to dilution effects and
corresponding increase of the heat exchange area.
Thermal stability of the oxidant mixture was investigated in great details and information on the
material compatibility were gathered.
Table 5: Stability of oxidant solution under different experimental conditions
Entry Experiment
Onset
(°C)
Energy
(J/g)
Material
Comments
ꢀP
(bar)
gold
plated
st.steel
DSC
1
Potential coating
corrosion
46
30
160
95
n.a.
n.a.
(sample 1)
Sample 2
gold
platedst.
steel
DSC
2
analyzed after
Sample 1;more
corrosion present
(sample 2)
3
4
C8017
ARCb
68
45
135 a
n.a.
20
glass
Inert material
Hastelloy C22
not completely
inert towards
Hastelloy
C22
>500 c
ACS Paragon Plus Environment
16

Page 17 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
(TD,24: 18 °C)
93
oxidants
5
ARCb
>1000 c
15
Glass
Inert material
(TD,24: 58 °C)
9
a: lower decomposition energy might be due to the small amount of material used in a large
test cell
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
60
b: ARC = Accelerating Rate Calorimeter
c: decomposition energy is underestimated as the decomposition did not reach completion.
Instrumental safety limit was reached.
DSC analyses utilizing gold plated high pressure crucibles were performed in duplicate.
Interestingly, by comparing Entry 1 and 2, it is evident that the detected onset decreased by more
than 10 °C. The sample in Entry 2 was left in the crucible for approximately 1 h prior to analysis.
Thus we can infer that a partial corrosion of the gold coating by the oxidant solution would
expose it to the presence of the underlying stainless steel, thus causing the metal ions to catalyze
the decomposition and hence the lower onset. Analogously, an adiabatic calorimetry experiment
conducted with Hastelloy® C22 test cells reported a slightly higher onset temperature (compared
to DSC experiment, Entry 1), although Hastelloy C22 possesses limited chemical resistance
towards strongly oxidizing acidic mixtures. Additionally, ARC system allowed to verify the
large pressure generation associated with the oxidant decomposition. The experiment conducted
with the C80 thermal screening equipment with a glass test cell showed a decomposition onset of
68 °C. The observation that metal ions could contribute to lowering the decomposition onset was
eventually verified by performing the adiabatic experiments using glass test cells (Entry 5).
Onset increased to 93 °C and, accordingly, the TD,24 rose to 58 °C. Therefore it appears evident
that a reactor train free from any metal ion contamination is of utmost importance.
Table 6: Process parameters for Criticality Class definition. Preparation of the oxidant mixture
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 32
1
2
3
4
Parameter
Value
Comments
5
6
7
8
TP
10 °C
11 °C
Process defined
No reagent accumulation
observed
MTSR
MTT
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
60
200°C
58 °C
Sulfuric acid solution
Metalꢀfree impurity situation
(Metal contamination present)
TD,24
(18 °C)
Criticality Class: 2
Oxidation reaction calorimetry and stability of the oxidation reaction
After having determined the parameters influencing the oxidant mixture decomposition
and the range of temperature in which to operate, we set out to investigate the reaction outcome
by means of Reaction Calorimetry experiments. In the development section we drew the
attention towards the need to have a direct correlation between the addition of the 5ꢀbromoꢀ2ꢀ
aminopyridine 2 and the formation of the product. If this correlation is established, the net result
will be an intrinsically safer process as, in case of cooling failure, interrupting the dosing will
result in immediate decrease of reaction heat. Again, two experiments were conducted by
varying the addition rate of 5ꢀbromoꢀ2ꢀaminopyridine solution (3h and 6h, respectively) to a
solution of oxidant at 20 °C (Figure 6).
Figure 6: Reaction calorimetry of oxidation reaction (6h addition time showed)
ACS Paragon Plus Environment
18

Page 19 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
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
60
Table 7: Reaction calorimetry data (Oxidation reaction)
% acc.
T
(°C)
20
Dosing
Max.T
(°C)
114
ꢀHrx.
(kJ/mol)
897
ꢀTad.
(°C)
94
ꢀTad.corr.
(°C)
28
(h)
3
32 %
23 %
20
6
878
90
63
21
The extent of the accumulation decreased with the increasing addition time (32% and 23%
respectively). This would result in a maximum reaction temperature of 41°C in case of 6h
addition time. It is important to mention that the maximum of accumulation was found at the end
of the amine addition.
In order to have a broad knowledge of the process we also studied the stability of the
final reaction mixture prior to quench. Both DSC analysis and ARC showed very moderate
decomposition energies (25 ꢀ 40 J/g) at an onset around 46ꢀ50 °C. The Criticality Class is
illustrated in the following table (Table 8).
Table 8: Process parameters for Criticality Class definition. Oxidation reaction
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 32
1
2
3
4
Parameter
Value
Comments
5
6
7
8
TP
25 °C
46 °C
Process defined
No reagent accumulation
observed. (RC1 calorimetry,
see Supporting Information)
MTSR
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
60
Teb.
TD,24
200 °C
58 °C
Sulfuric acid solution
Metalꢀfree impurity
(18 °C)
(Metal contamination present)
Criticality Class: 2
It is important to mention that the Criticality Class definition holds true if and only if no metal
ion contaminants are present. In case of any contamination, the process becomes a Criticality
Class 5 even if the amount of reagent accumulation is relatively low (approx. 20%).
The calorimetry results, together with HPLC analyses, clearly showed that the desired
oxidation is actually completed 2h after the end of addition and that the accumulated heat from
this point on might be attributed to the decomposition of the excess of oxidant. Nevertheless, it
was decided to continue stirring the mixture for a total of 18 h prior to quench. The reason is
twofold: first, the impurity profile changed and the 3,5ꢀdibromoꢀ2ꢀaminopyridine was slowly
converted into the corresponding oxidized nitro derivative which purged much better during
isolation. Second, the quench could be performed on a solution containing a very small amount
of oxidant left due to its almost complete decomposition.
Product isolation
As already mentioned, after the oxidation reaction the product was isolated by directly
adding the reaction mixture into water (20 w/w with respect to 5ꢀbromoꢀ2ꢀaminopyridine),
followed by addition of isopropanol18 (8.5 v/w with respect to 5ꢀbromoꢀ2ꢀaminopyridine). The
solid obtained was isolated by filtration, washed with water until the rinsing was at neutral pH,
ACS Paragon Plus Environment
20

Page 21 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
followed by final isopropanol rinsing (0.63 v/w). To ensure complete removal of any peroxide
intermediate, sodium metabisulfite (1 w/w with respect to 5ꢀbromoꢀ2ꢀaminopyridine 2) was
added to the water phase prior to quench followed by semiꢀquantitative peroxide test (limit <10
ppm). Isopropanol was necessary to ensure good filterability and a better impurity purging. As
expected, the quench was highly exothermic and the gas generation (SO2) from metabisulfite
acidification must be controlled through the use of a scrubber. Nevertheless, reaction calorimetry
showed no degree of accumulation (see Supporting Information for RC1 calorimetry graph), so
the process could be considered perfectly dose controlled.
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
60
The quenched mixture showed no significant decomposition when analyzed with DSC and C80.
Quite expectedly, the presence of nitro group is associated with high decomposition energy of
the 5ꢀbromoꢀ2ꢀnitropyridine (average 1490 J/g, DSC analyses). The final product was obtained
after a slurry in hot isopropanol on the combined crude material from the small batches.
Industrial scaleꢀup results
After having established a robust protocol and understood the effect of the heat generation on
the amount of the dibrominated impurity, the reaction was successfully run on full scale with two
batches of 305 kg and 360 kg, respectively. Longer addition time (9h vs. 3h on laboratory scale)
were necessary in order to ensure the best temperature control. The batches were isolated in
yields and purity comparable if not better than those obtained on the largest 140 g laboratory
sample (Table 9).
Table 9. Bromination reaction. Comparison between plant and lab trial results
Experimental data
Isolated wet solid (kg)
HPLC purity (% area)
PlantꢀBatch #1
305
PlantꢀBatch #2
360
Laboratory scale
n.a.1
95.52 %
95.07 %
95.08 %
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 32
1
2
3
4
5
6
diꢀbromo aminopyridine (% area)
Isolated yield (% mol)
4.48 %
77.4%
4.93 %
81.5 %
4.91 %
75.4%
7
8
1: On laboratory scale the solid was dried and the yield calculated on the dry product
9
Particular attention was dedicated to the oxidation reaction. In order to ensure complete
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
60
cleanliness of the reactor train from metal impurities and residual organic solvents from cleaning
procedures, glassꢀlined reactors were utilized after extensive washing with water. Addition of
hydrogen peroxide 30% over concentrated H2SO4 was performed via Teflonꢀcoated metering
pump. Hydrogen peroxide solution was delivered in multiple small drums (25 kg each) thus
avoiding accidental overcharge. Analogously, the process was designed in such a way that only a
single aliquot of 5ꢀbromoꢀ2ꢀaminopyridine in sulfuric acid could be added. The solution of the
amine prepared in a reactor was discharged in aliquots (corresponding to 1/10 of the total
amount), in a dedicated plastic drum. Therefore, even in case of accidental uncontrolled addition
of the whole quantity contained in a single drum, this would correspond to a temperature
increase of only +9K. The oxidation reaction was conducted in several different batches in order
to allow sufficient void volume in the reactor containing the oxidant solution while maintaining a
reasonable productivity. ARC data allowed to determine that, in case of decomposition, the
maximum pressure generated inside the reactor will not exceed the maximum design pressure of
the reactor.
Therefore, a total of 8 batches of the oxidation step were run. The crude material was
combined and slurried in hot isopropanol, delivering an average isolated yield of 48.6% mol
compared with 52% on laboratory scale with the same HPLC purity (99.96% HPLC area) (Table
10).
Table 10. Oxidation reaction. Comparison of plant and lab trial results
ACS Paragon Plus Environment
22

Page 23 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Experimental data
Plantꢀ(final)
305
Laboratory scale
n.a.
Isolated wet solid (kg)
HPLC purity (% area)
99.92 %
0.08 %
99.96 %
diꢀbromo nitropyridine (% area)
Isolated yield (% mol)
0.04 %
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
60
48.6 %
52.9 %
(after isopropanol slurrying)
The slightly lower yield might be attributed to the longer addition time required for the
preparation of the oxidant mixture, which in turn resulted in greater decomposition. Furthermore,
an incomplete precipitation was observed during the preparation of the first batch resulting in
yield decrease. The precipitation was indeed improved in the subsequent 7 batches by increasing
the amount of water for the dilution/precipitation.
CONCLUSION
In summary, we have developed an intrinsically safer process of an otherwise dangerous
reaction by conducting a rigorous safety investigation combined with process development.
Optimization of the reaction conditions with particular emphasis on the role played by the
sulfuric acid allowed the definition of the operating range. The transformations could be scaledꢀ
up to full production scale and delivered, throughout the campaign, results comparable to those
obtained in the laboratory.
We do believe that performing early stage safety investigation together with more
“classical” process development will allow the quick identification of the critical process
parameters and bring obvious positive consequences in an everꢀchanging, fastꢀpaced industrial
environment such as the one of today’s pharmaceutical industry.
EXPERIMENTAL SECTION
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 32
1
2
3
4
5
6
7
8
9
All reactions were carried out under an atmosphere of dry nitrogen unless otherwise specified.
All reagents and solvents were used as received without further purification unless otherwise
specified. Reaction calorimetry was conducted either on Mettler Toledo RC1e RTCal reaction
calorimeter or Mettler Toledo Optimax equipped with HFCal reaction calorimeter module.
Magnetic resonances were recorded using a 300 MHz Varian Mercury spectrometer. Chemical
shift is reported in ppm against the solvent residual peaks. LC/MS analyses were recorded using
an Agilent Technologies 6500 series QꢀTOF (for Electro Spray Ionization) or AB Sciex (for
Atmospheric Pressure Chemical Ionizazion).
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
60
HPLC method
Instrument: Agilent 1260 equipped with quaternary pump or equivalent.
Column: Phenomenex Luna PhenylꢀHexyl; 4.6 x 150 mm, 5 ꢁm or equivalent.
Column temperature: 35°C
Samples tray: 10°C
Eluent A: buffer pH=7.0, 0.01M KH2PO4 and 0.01 M Na2HPO4
Eluent B: methanol
Flow rate: 1 mL/min
Wavelength: 210 nm
Injection volume: 5 ꢁL
Diluent: eluent A/eluent B 50:50
Gradient
Time
0
Phase A
70
Phase B
30
2
70
30
20
22
10
90
10
90
ACS Paragon Plus Environment
24

Page 25 of 32
Organic Process Research & Development
1
2
3
4
5
6
22.1
26
70
70
30
30
7
8
5ꢀbromoꢀ2ꢀaminopyridine (2)
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
60
In a 6300 L, glass lined reactor (R1), 2ꢀaminopyridine 1 (200 kg, 2.13 kmol) was added.
Acetonitrile (316 kg) was added and the mixture was stirred at room temperature until complete
dissolution. In a second, 6300 L, glass lined reactor (R2), NꢀBromosuccinimide (397 kg, 2.23
kmol, 1.05 equiv.) was added followed by acetonitrile (2372 kg) and the mixture stirred at 20 °C
until complete dissolution. Reactor R1 was cooled to ꢀ5 °C and the solution in R2 was added to
R1 in no less than 6h while maintaining the temperature below 5 °C. The reactor R2 and the
transfer line were rinsed with acetonitrile (200 kg). The mixture was stirred for 1h at 5 °C,
warmed to 20 °C and maintained for 1h at this temperature. The mixture was then sampled for In
Process Control (IPC). After IPC was met (2ꢀaminopyridine ≤ 1.5% HPLC area), water (500 kg)
was added to solubilize the precipitate, followed by activated charcoal (10 kg). The mixture was
stirred for 1h at 20ꢀ25 °C and subsequently filtered through a cellulose layer in reactor R2. The
reactor R1 and transfer line were rinsed with water (100 kg) and acetonitrile (79 kg). The
solution was concentrated under reduced pressure with a jacket temperature of 50ꢀ60 °C until
approximately 3400 L of solvent mixture were distilled. Water (2000 kg) was slowly added and
the mixture was cooled down to 0ꢀ5 °C in no less than 2h. The solid product was isolated via
centrifugation and the solid cake was washed with water (700 kg). The resulting solid (305 kg)
of wet 5ꢀbromoꢀ2ꢀaminopyridine was used without drying after determination of water content
(KF titration and Loss on Drying). HPLC purity: 95.5% (KF: 4.8%, LOD= 5.8%, for information
only)
1H NMR (DMSO):
δ
in ppm (integral, multiplicity, Hz)
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 32
1
2
3
4
7.94 (1H, d, 2.54), 7.50 (1H, dd, 8.82, 2.59), 6.43 (1H, dd, 8.83, 0.55) 6.13 (2H, bs)
5
6
7
13C NMR (DMSO):
δ in ppm (type of carbon)
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
60
158.65 (Cq), 147.86 (CꢀH), 139.19 (CꢀH), 110.03 (CꢀH), 105.12 (Cq)
LC/MS data (ESI)
172.9713 ([M+H]+), 174.9693 ([M+H]+), 558.8497 ([3M+K]+)
5ꢀbromoꢀ2ꢀnitropyridine (3)
In a 6300 lt glass lined reactor (R1), sulfuric acid (491 kg) was added and cooled below 10 °C.
5ꢀbromopyridineꢀ2ꢀamine 2 (65 kg dry material, 0.375 kmol, 69.5 kg wet) was added in small
portions in order to maintain the temperature below 15 °C. The preparation of the amine solution
must be performed prior to the oxidant preparation. In a 6300 L glass lined reactor (R2), sulfuric
acid 98% (1474 kg) was added and the reactor cooled to 0ꢀ5 °C. Hydrogen peroxide 30%
solution (263 kg, 2.32 kmol H2O2, 6.2 mol ratio, 2.05 equiv.) was dosed through a teflonꢀcoated
pump while maintaining the temperature below 10 °C. In case of temperature increase, the
dosing must be interrupted and restarted only when the temperature is below the setpoint.
Additional sulfuric acid (10 kg) was added to rinse the dosing line and the mixture was stirred
for 1h at 10ꢀ15 °C. The amine solution in R1 was discharged in aliquots of approximately 70 kg
in a dedicated plastic drum. The aliquot was added in R2 through the same Teflonꢀcoated dosing
pump, pervioulsy rinsed with concentrated sulfuric acid, while keeping the temperature below 25
°C and, in any case, the overall addition time must not be faster than 6h. In case of temperature
increase, the addition must be stopped and restarted only when the internal temperature is below
20 °C. Additional sulfuric acid 98% (20 kg) was added to rinse the addition line and pump. The
ACS Paragon Plus Environment
26

Page 27 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reaction was stirred at 20ꢀ25 °C for no less than 14h whereupon a sample for In Process Control
was taken. After IPC limit was met (2ꢀaminoꢀ5ꢀbrompyridine ≤ 1.0 %), the reaction mixture was
slowly transferred in a glass lined reactor (R3) in which a water (1460 kg) and metabisulfite (65
kg) solution was previously prepared. Upon addition a precipitate formed. The addition was
performed at a rate allowing the temperature in R3 to remain below 15 °C and, in any case, not
faster than 6h. R2 was rinsed with sulfuric acid (20 kg). After that isopropanol (515 kg) was
added and the mixture stirred at 20ꢀ25 °C for no less than 4h. The mixture was further cooled to
0ꢀ5 °C and stirred for 2h. The product was isolated via centrifugation and the wet product rinsed
with water until rinsing water resulted in a neutral pH. A final isopropanol rinsing (41 kg) was
applied and an average of 40 kg of wet crude 5ꢀbromoꢀ2ꢀnitro pyridine 3 was obtained (KF
titration and Loss on Drying were performed in order to estimate the amount of dry product).
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
60
The combined wet solids obtained in the 8 batches of crude product produced in the campaign
(305 kg estimated dry material, 330 kg wet material) were charged in a glass lined reactor.
Isopropanol (1164 kg) was added and the mixture stirred at refluxing temperature (78ꢀ82 °C) for
no less than 1h. The mixture was cooled to 0ꢀ5 °C and aged at that temperature for no less than
2h. The product was isolated via centrifugation and rinsed twice with isopropanol (111 kg). The
wet solid was dried in a static tray drier at maximum 40 °C under reduced pressure for no less
than 12h. The process delivered 301,7 kg of the title compound. HPLC purity: 99.92% (KF:
0.07%, limit ≤ 2.0%)
1H NMR (DMSO):
δ in ppm (integral, multiplicity, Hz)
8.83 (1H, dd, 2.36, 0.55), 8.51 (1H, dd, 8.63, 2.37), 8.26 (1H, 8.63, 0.58)
13C NMR (DMSO):
δ
in ppm (type of carbon)
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 32
1
2
3
4
5
6
155.61 (Cq), 150.12 (CꢀH), 143.59 (CꢀH), 127.04 (Cq), 120.45 (CꢀH)
LC/MS data (APCI):
7
8
9
202.9 [M+H]+ , 204.8 [M+H]+
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
60
SUPPORTING INFORMATION
Adiabatic calorimetry graphs, Self Heating Rate and calculation of TMR are reported in the
Supporting Information. They also include detailed description of the preparation of azoꢀoxy
impurity 5 and spectral characterization of the compounds.
ACKNOWLEDGMENTS
We wish to thank Dr. Dan Fishlock (Roche) for the preparation and characterization of the
“azoꢀoxy” impurity and the precious scientific discussions. We are grateful to Dr. Tania
Minaglia, Maria Francesca Pizzi and Elena Colombo (Olon) for method development and
analytical support. We wish to thank the Derivados Quimicos production team for conducting the
large scale synthesis. Finally, we are grateful to Mrs. Lucia Gigante and Mr. Christian Pasturenzi
of INNOVHUBꢀStazione Sperimentale per le Industrie, for the helpful discussions and for
conducting the C80 experiment.
AUTHOR INFORMATION
Corresponding author
Email: aagosti@olonspa.it
Note: All the authors declare no conflict of interest
REFERENCES
ACS Paragon Plus Environment
28

Page 29 of 32
Organic Process Research & Development
1
2
3
4
5
6
1
Caron, S.; Dugger, R. W.; Gut Ruggeri, S.; Ragan, J.A.; Brown Ripin, D.H. Chem. Rev.,
7
8
2006, 106, 2943
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
60
2 Jones, C. W., Application of hydrogen peroxide and derivatives, RSC, 1999
3 J. Loss Prev. in Proc. Ind., 2006, 19(4), 307ꢀ311; Lacoursiere, J.P. Process Saf. Progr., 2005,
24, 115
4 Urben, P.; Bretherick, L.; Bretherick’s Handbook of Reactive Chemical Hazards (6th Edition),
London: Elsevier Ltd., 2006; Crowl, D.A.; Louvar, J.F. Chemical Process Safety: Fundamentals
with Applications (3rd International Edition), 2011
5 Gygax, R. Chem. Eng. Progr., 1988, 43, 1759
6 Stoessel, F.; Thermal Safety of Chemical Processes, WileyꢀVCH, 2008
7
For examples of NBS bromination in different solvents: DMF: a) Silpa, L.; Niepceron, A.;
Laurent, F.; Brossier, F.; Penichon, M.; EnguehardꢀGueffier, C.; Abarbri, M.; Silvestre, A.;
Petrignet, J. Bioorg. Med. Chem. Lett. 2016, 26, 114 Acetonitrile b) Leboho, T. C.; Giri, S.;
Popova, I., Cock, I.; Michael, J.P.; de Koning, C.B. Bioorg. Med. Chem. 2015, 23, 4943 c)
Leboho, T. C; van Vuuren, S.F.; Michael, J.P.; de Koning, C.B. Org. Biomol. Chem. 2014, 12,
307 Acetone d) PinkertonZhang, Y.; Song, H.; Liu, T.; Du, M.; Xu, S.; Li, X.; Liang, X. CN
103755628
8
For examples of bromine reactions in different solvents: DMF a) Chen, X. CN 103664763
acetic acid b) Ismail, M.A.H.; Abou El Ella, D.A.; Abouzid, K.A.M.; Mahmoud, A.H Bioorg.
Med. Chem. Lett., 2012, 70, 2455
ACS Paragon Plus Environment
29