Safe Scale-Up of a Hydrazine Condensation by the Addition of a Base

Article abstract of DOI:10.1021/op4002577

Herein we describe an observation where an exothermic event encountered during the safety evaluation of the scale-up of the synthesis of 4-chloro-1H-indazol-3-amine was mitigated upon the addition of a base. The 100 C adiabatic temperature rise was attributed to the hydrazine condensation reaction, which could cause the batch to self-heat beyond the onset temperature of the exothermic decomposition of the reaction mass. Switching from 1-methyl-2-pyrrolidinone to a lower-boiling-point solvent was explored, but that alone did not guarantee the safe operation in the event cooling is lost. The reaction byproduct, HCl, was identified as a possible cause for the decreasing onset temperature of hydrazine monohydrate. The addition of a base to the reaction mixture increased the onset temperature and decreased the severity of the observed decomposition of the reaction mass. By the introduction of sodium acetate as a base in combination with a lower-boiling-point solvent, safe operating conditions for the process were identified. This base-stabilizing effect has been observed with other hydrazine condensation reactions in our laboratory.

Full text of DOI:10.1021/op4002577

Article
pubs.acs.org/OPRD
Safe Scale-Up of a Hydrazine Condensation by the Addition of a Base
Zhe Wang,* Steven M. Richter, Jorge Gandarilla, Albert W. Kruger, and Michael J. Rozema
Process Research and Development, AbbVie Inc., 1 North Waukegan Road, North Chicago, Illinois 60064, United States
S
* Supporting Information
ABSTRACT: Herein we describe an observation where an exothermic event encountered during the safety evaluation of the
scale-up of the synthesis of 4-chloro-1H-indazol-3-amine was mitigated upon the addition of a base. The 100 °C adiabatic
temperature rise was attributed to the hydrazine condensation reaction, which could cause the batch to self-heat beyond the
onset temperature of the exothermic decomposition of the reaction mass. Switching from 1-methyl-2-pyrrolidinone to a lower-
boiling-point solvent was explored, but that alone did not guarantee the safe operation in the event cooling is lost. The reaction
byproduct, HCl, was identified as a possible cause for the decreasing onset temperature of hydrazine monohydrate. The addition
of a base to the reaction mixture increased the onset temperature and decreased the severity of the observed decomposition of
the reaction mass. By the introduction of sodium acetate as a base in combination with a lower-boiling-point solvent, safe
operating conditions for the process were identified. This base-stabilizing effect has been observed with other hydrazine
condensation reactions in our laboratory.
solutions of hydrazine containing >40 mol % hydrazine are also
flammable.4
INTRODUCTION
■
The condensation reaction between 2,6-dichlorobenzonitrile
and hydrazine was initially optimized using 1-methyl-2-
pyrrolidinone (NMP) to generate our target intermediate 4-
chloro-1H-indazol-3-amine (1, chloroaminoindazole). As
shown in Scheme 1, two reaction pathways were postulated.
In order for a process to be scaled up in a pilot plant, a
process safety evaluation is typically performed. This usually
comprises measuring heats of reaction, determining the
propensity for gas generation, screening the thermal stabilities
of process intermediates and reaction mixtures, and identifying
potential runaway reactions.
Scheme 1. Chloroaminoindazole Formation
In the case of chloroaminoindazole, the thermal stabilities of
reactants, solvents, reaction mixtures, intermediates, and
products were screened by calorimetric methods to identify
the onset temperatures and the severity of exothermic events.
The heat of reaction, measured with heat flow calorimetry, was
used to determine the adiabatic temperature rise (ATR) of the
reaction. The ATR provides a measure of the thermal potential
of the desired reaction. If cooling is lost at the process
temperature (Tp), the temperature could increase by the ATR
to the maximum temperature of the synthesis reaction
(MTSR). Secondary reactions such as exothermic decom-
positions can occur if the MTSR is higher than their onset
temperature. Various instruments can be used to detect the
onset temperature (Tonset). The onset temperature of a reaction
is dependent upon the sensitivity of the instrument used, and a
reaction may proceed below the detected onset temperature,
but at a rate below the detection limit of the instrument.
Insufficient removal of the heat from an exothermic reaction
can cause the reaction mass to uncontrollably self-heat. As the
temperature increases, the rate of the reaction increases,
eventually reaching a maximum rate. The temperature at
which this time to reach the maximum rate is 24 h is defined as
TD24. The criticality index introduced by Stoessel5 allows the
classification of potential runaway reaction scenarios based on
Under the optimized reaction conditions, the formation of
intermediate A was favored, and its conversion to product was
fast.1 By comparison, path B required excess hydrazine, and
intermediate B was converted to the product only after
extended reaction times. The chloroaminoindazole product/
intermediate B ratio at the end of the reaction was
approximately 20:1. The safety evaluation findings were
summarized in the discussion of process development of
ABT-869 by Kruger et al.1 Details of the process safety
evaluation are discussed in this work.
Hydrazine is toxic, and the 8 h time-weighted exposure limit
is 0.1 ppm. Personal exposure to hydrazine through skin
contact, ingestion, and inhalation should be minimized.2
Hydrazine is flammable over a very wide range, from 4.7% to
100% in air. Once ignited, hydrazine can burn by exothermic
decomposition in the absence of air or another oxidant.3 Water
Special Issue: Safety of Chemical Processes 13
Received: September 19, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
the relative levels of Tp, MTSR, TD24, and a maximum technical
temperature, which is frequently the boiling point of solvent,
Tb. The criticality index categorizes the upset scenarios into five
classes ranging from the least critical (1) to the most critical
(5), as shown in Figure 1. The processes in classes 1 to 3 do
To attempt to identify the source of the exotherm, the
thermal stabilities of the starting material, reaction mixture, and
product were screened using a differential scanning calorimeter
(DSC). No exothermic event was detected in the isolated
starting material or product. As displayed in Figure 3, the DSC
analysis of the reaction mixture showed exothermic events from
87 to 202 °C and above 215 °C.
Figure 3. DSC results for the fresh NMP reaction mixture.
Figure 1. Criticality classification of chemical processes.
On the basis of the process temperature of 110 °C, the
exothermic event detected from 87 to 202 °C was most likely
due to the desired reaction. The RC1 and DSC results
indicated that the ATR of the reaction was sufficient to increase
the temperature of the reaction mass to near the onset
temperature of the second and more energetic exothermic
decomposition. These observations caused us to perform a
thorough investigation of the energetics of the system.
Additional thermal stability testing using the more sensitive
accelerating rate calorimeter (ARC) was conducted in order to
obtain a more accurate onset temperature and monitor pressure
effects. A fresh reaction mixture (with the charges shown in
Table 5 in the Experimental Section) was tested with the ARC,
and the results are summarized in Table 1 and shown in Figures
4, 5, and 6.
not have the thermal potential to reach the onset of the
secondary reaction through self-heating alone. The processes in
classes 4 and 5 have the thermal potential to self-heat to the
decomposition onset temperature.
RESULTS AND DISCUSSION
■
The heat of reaction was measured with a Mettler-Toledo
reaction calorimeter (RC1) to calculate the MTSR during a loss
of cooling event. In the test, after thermal equilibrium was
reached at the reaction temperature (110 °C), hydrazine
monohydrate was added over a period of 30 min into the
reactor containing the 2,6-dichlorobenzonitrile starting material
and NMP (with the charges shown in Table 9 in the
Experimental Section). The measured heat of reaction was
−237 kJ/mol of 2,6-dichlorobenzonitrile, and the correspond-
ing ATR was 100 °C. The MTSR, calculated as summation of
the process temperature and ATR, was 210 °C. The RC1
results are shown in Figure 2.
Table 1. ARC Test Results for the Fresh Reaction Mixture
a
range of exotherm (°C)
estimated ATR (°C)
45−120
130−210
>280
116
160
>30
a
Corrected for thermal inertia, ϕ = 1.78.
Exothermic events were detected at 45, 130, and 280 °C,
which were in line with the DSC results. Since the ARC is more
sensitive than the DSC in detecting exothermic events, the first
exothermic event was detected at 45−120 °C and is presumably
due to the desired reaction. Exothermic decomposition
occurred during the event at 130 °C, as indicated by the
noncondensable gas generation accompanying the second
exotherm. Because of the complexity of the decomposition,
as indicated by the multiple peaks in the self-heating rate, the
determination of TD24 was not attempted for this secondary
reaction, and Tonset was used for the safety analysis instead.
Since the MTSR is 210 °C, the heat generated by the desired
reaction can increase the temperature to the second exotherm if
excess heat is not removed. Using a high-boiling-point solvent
Figure 2. RC1 results for hydrazine condensation.
B
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 4. Reaction mixture (NMP): ARC test temperature and
pressure history.
Figure 7. Reaction mixture (NMP): VSP2 test temperature and
pressure history.
Figure 5. Reaction mixture (NMP): ARC test temperature and
pressure rise rates.
Figure 8. Reaction mixture (NMP): VSP2 test temperature and
pressure rise rates.
runaway would proceed at larger scales. The results are shown
in Figures 7 and 8.
An exothermic event, presumably due to the desired
hydrazine condensation reaction, was detected at 70 °C.
Extrapolation of the data to the ARC detection sensitivity of
0.02 °C/min gave a temperature of ∼40 °C, in agreement with
the onset temperature observed in ARC testing. The event led
to violent exothermic decomposition. The maximum temper-
ature and pressure for the half-full test cell were 316 °C and
>1250 psig, respectively. The recorded maximum temperature
and pressure rates exceeded 1150 °C/min and 6400 psi/min,
respectively.
At 202 °C, the boiling point of NMP, the reaction was in the
process of running away with the self-heat rate of 65 °C/min.
Relying on solvent evaporative cooling to remove the reaction
heat at such a high reaction rate was not considered practical.
The reaction did not proceed to practical levels at lower
temperatures, and thus, reducing reaction temperature was not
an option.
In an effort to take advantage of energy removal via
evaporative cooling, solvents with boiling points near the
reaction temperature were screened. The intent of this effort
was that in the event of loss of cooling or uncontrolled heating,
the energy generated by the reaction could be removed via
evaporation of the solvent while maintaining a lower reaction
rate. Adding this thermal barrier would reduce the criticality
class from 5 to 4. Pyridine, with a boiling point of 115 °C, was
identified as the most favorable alternative solvent to NMP in
Figure 6. Reaction mixture (NMP): ARC test temperature and
pressure (corrected for pad gas).
such as NMP places this reaction in criticality class 5, with
enough thermal potential to reach the exothermic decom-
position and no thermal barrier to overcome.5
The reaction was observed to become organic−aqueous
biphasic as the reaction proceeds. To examine agitated
conditions, a thermal stability test was performed using the
Vent Sizing Package 2 (VSP2) and a closed test cell (with the
charges shown in Table 6 in the Experimental Section). In
addition to the availability of agitation, this low-thermal-inertia
calorimeter provides a more realistic simulation of how the
C
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
terms of yield, selectivity, and reactivity to test the hypothesis.
The reaction using pyridine was examined using VSP2 (with
the charges shown in Table 7 in the Experimental Section). A
sufficient amount of solvent was charged to remove the
generated heat of the desired reaction by evaporative cooling.
The test cell was vented to atmospheric pressure to test
whether this barrier could prevent the reaction mass from
reaching the onset temperature of exothermic decomposition.
The vented vapor was condensed and collected on a balance.
The results are shown in Figures 9 and 10.
Figure 11. DSC results for the reaction components.
generated during the reaction, the effects of HCl on the thermal
stability of the product as well on hydrazine were examined.
The HCl salt of the product was stable up to 280 °C. The
hydrazine hydrochloride and hydrazine dihydrochloride salts
were both stable up to >250 °C (see Figure 12). When the
Figure 9. Reaction mixture (pyridine) vented to ambient: VSP2 test
temperature, pressure, and collected condensate history.
Figure 12. DSC results showing the effect of HCl and K2CO3 on the
thermal stability of hydrazine monohydrate.
Figure 10. Reaction mixture (pyridine) vented to ambient: VSP2 test
temperature and pressure rise rates.
product was exposed to hydrazine hydrochloride or hydrazine
hydrate, the thermal stability was reduced, but the onset
temperatures were still approximately 200 °C, which is
significantly higher than the onset observed in the reaction
mixture. Although the isolated crystalline intermediate B was
thermally stable up to 260 °C, the solution containing 50%
intermediate B in hydrazine, solvent, and water showed an
exothermic event at 70 °C.
DSC results for the reaction mixtures sampled at different
stages of the reaction are shown in Table 2. The reaction
mixture did not exhibit any significant exotherm after addition
of 27% of the hydrazine charge. Reaction mixture samples taken
at the completion of hydrazine monohydrate addition and 2, 3,
and 6 h after the addition exhibited exothermic decomposition
at temperatures as low as 120 °C.
After evaporation of most of the solvent, exothermic
decomposition still occurred and resulted in temperature and
pressure excursions with magnitudes of 100 °C and 30 psi,
respectively. The maximum observed temperature and pressure
rise rates were 3000 °C/min and 2500 psi/min, respectively.
The onset temperature of the exothermic decomposition was
83 °C, as determined by extrapolating the self-heat rate to 0.02
°C/min. This indicated that the vaporization of solvent alone
did not provide an adequate barrier to prevent reaching the
decomposition. However, because of the partial safety benefit it
offered, pyridine was still preferred over NMP and therefore
used in the further investigation.
To identify the source of the exothermic decomposition, a
series of thermal stability tests using the DSC were performed.
As shown in Figure 11, no exotherms were observed up to 300
°C in the starting material and product. Since HCl was
Since the reaction is biphasic, DSC testing was also
conducted for samples taken from the aqueous and organic
phases 3 and 6 h after hydrazine monohydrate addition. The
D
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
amount of HCl to hydrazine monohydrate, an energetic
exothermic event with onset temperature near 180 °C was
observed. Upon introduction of a base, K2CO3, this exothermic
event was eliminated. It seems that the presence of HCl
significantly reduces the thermal stability of hydrazine
monohydrate. However, the introduction of a base improved
(or restored) the thermal stability of the mixture after exposure
to HCl.
Interestingly, increasing the HCl concentration to 0.5, 1, or 2
equiv had a smaller effect on the thermal stability. In these
cases, the onset temperature of exothermic decomposition was
reduced only to ∼250 °C, similar to the onset temperature of
hydrazine monohydrochloride. DSC results for the effect of
HCl on hydrazine monohydrate are presented in Figure 12.
Since intermediate B was also identified in the aqueous
phase, its thermal stability and the influences of adding
hydrazine monohydrate and a base on the thermal stability
were also investigated by DSC testing. As shown in Figure 13,
Table 2. DSC Results for the Reaction Mixture (Pyridine as
Solvent, Addition of Hydrazine at 110 °C)
a
range of
exotherm (°C)
sample
ΔH (J/g)
reaction
mixture
after adding 27% of the
hydrazine
endothermic
105−150
after adding all of the
hydrazine
−1590
120−276
2 h after adding hydrazine −1243
3 h after adding hydrazine −1384
6 h after adding hydrazine −651
120−254
155−281
131−278
138−253
solid obtained by concentrating the
reaction mixture (6 h after adding
hydrazine)
−835
a
Plots showing the DSC results are provided in the Supporting
Information.
results are shown in Table 3. The organic portions of the
reaction mixture did not exhibit any exothermic events. The
Table 3. DSC Results for Aqueous and Organic Samples of
a
the Reaction Mixture (Pyridine as Solvent)
range of
exotherm
(°C)
sample
ΔH (J/g)
organic
layer
3 h after adding hydrazine
6 h after adding hydrazine
3 h after adding hydrazine
stable
stable
−2036
−1977
up to 255
up to 220
134−284
134−281
aqueous
layer
3 h after adding hydrazine,
retested 2 days later
6 h after adding hydrazine
−2081
−1879
119−270
140−275
6 h after adding hydrazine,
retested 2 days later
6 h after adding hydrazine,
retested 2 days later and filtered
with 2 micron filter paper
−2088
126−280
solid
concentrated from organic layer
(6 h after adding hydrazine)
−874
121−295
Figure 13. DSC results showing the effect of hydrazine monohydrate
and K2CO3 on the thermal stability of intermediate B.
concentrated from aqueous layer
(6 h after adding hydrazine)
exothermic
>90
a
Plots showing the DSC results are provided in the Supporting
Information.
no exothermic event was detected in the intermediate B solid
up to 260 °C. After addition of hydrazine monohydrate, an
exothermic event with an onset temperature near 150 °C was
detected. By the further introduction of a base, K2CO3, the
severity of the exothermic event was reduced to 5% of the case
without the base. It was suspected that with hydrazine
monohydrate, the dissolved intermediate B underwent the
cyclization reaction to form the product and to generate HCl.
This process would accelerate with increasing temperature
during the DSC testing. The generated HCl reduces the
thermal stability of hydrazine monohydrate.
The effect of the addition of base additives to the reaction
mixture on the onset temperature of exothermic decomposition
was also evaluated. On the basis of DSC testing, the aqueous
layers of reaction mixtures using KOAc, NaOAc, K2CO3, or
Na2CO3 as the base exhibited no significant exotherms up to
300 °C. The onset temperatures of the exotherms observed
with K2HPO4 and NaHCO3 were below 300 °C but were still
much higher than without base. Organic bases also had a
stabilizing effect but were less effective than the inorganic bases
screened. The results are summarized in Table 4. The DSC
results showing the effect of sodium acetate on the stability of
the reaction mixture aqueous layer are shown in Figure 14.
solid obtained by concentrating the organic phase taken at 6 h
exhibited a mild exotherm starting at ∼120 °C. All of the
aqueous portions of the reaction mixture, even samples retested
2 days later, exhibited strong exothermic decompositions at
temperatures as low as 119 °C. Through HPLC analysis and
mass balance, it was concluded that the aqueous layer consisted
of hydrazine, hydrazine hydrochloride, and the formed
intermediate B.
In a process safety review of a similar process using an aryl
fluoride, this thermal instability of the reaction mixture was not
observed. In the case of the aryl fluoride, a base was added to
prevent etching of glassware by the HF byproduct. The
possibility that the presence of this base may also have affected
the thermal stability of the reaction mixture led us to explore
the effect of adding a base to the aryl chloride reaction.
Since the only reaction component identified in the aqueous
phase in significant concentration was hydrazine, its thermal
stability in the presence of HCl and base was studied with DSC.
No significant exothermic event was observed for hydrazine
monohydrate up to 300 °C. During the hydrazine condensation
reaction, 0.1 molar equiv of HCl relative to hydrazine is
generated as a byproduct of the reaction. After addition of this
E
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
decomposition detected at 200 °C did not cause a pressure
excursion. The results are shown in Figures 15 and 16.
Table 4. DSC Results for Samples of the Aqueous Layer of
the Reaction Mixture in Screening of Bases (in Pyridine)
Tonset
(°C)
base used in the reaction
potassium carbonate
potassium acetate
sodium carbonate
sodium acetate
>300
>300
>300
>300
270
(solid concentrated from aq layer with potassium phosphate
tribasic)
potassium phosphate dibasic
potassium phosphate tribasic
sodium bicarbonate
255
250
210
210
205
205
200
120
sodium phosphate dibasic heptahydrate
triethylamine
potassium bycarbonate
N-methylmorpholine
no added base
Figure 15. Reaction mixture with NaOAc: VSP2 test temperature and
pressure history.
Results using other bases can be found in the Supporting
Information.
Figure 16. Reaction mixture with NaOAc: VSP2 test temperature and
pressure rise rates.
With the updated safety procedures in place, namely, adding
sodium acetate as a base, using pyridine as the solvent, and
including sufficient venting, the hydrazine condensation process
was safely and successfully scaled-up into multikilogram pilot
plant runs. Taking these measures increases the onset
temperature of the decomposition reaction, changing the
criticality classification from 5 to 3. The severity of the
decomposition reaction is also reduced.
Figure 14. DSC results showing the effect of NaOAc on the aqueous
sample of the reaction mixture.
Introducing a base such as KOAc, NaOAc, K2CO3, or
Na2CO3 into the reaction mixture effectively increased the
onset temperature of decomposition to above 300 °C. A weak
base was preferred in order to minimize the dimerization of
product and potential hydrolysis of the nitrile starting material
to the amide. NaOAc was considered to be favorable in the
scale-up because of its high solubility and effectiveness in
increasing the thermal stability of the reaction while delivering
product of good quality.
A closed-cell VSP2 test on the reaction using pyridine as the
solvent and NaOAc as the base (with the charges shown in
Table 8 in the Experimental Section) was performed to verify
whether the heat generation from the hydrazine condensation
reaction would lead to an exothermic decomposition. The
initial exotherm due to the hydrazine condensation reaction
brought the temperature to 140 °C under adiabatic conditions
with corresponding vapor pressure at about 50 psig. A second
exotherm was detected at 200 °C during the subsequent heat−
wait−search step. It is also noted that the exothermic
CONCLUSION
■
The synthesis of 4-chloro-1H-indazol-3-amine via the hydrazine
condensation reaction at 110 °C in NMP was determined to be
unsafe to scale up because, in the event of a loss of cooling, the
heat of reaction could heat the reaction above the onset
temperature of violent exothermic decomposition of the
reaction mass. Employing a low-boiling-point solvent alone
would not guarantee a safe operation. The aqueous phase of the
biphasic reaction mixture was identified as being the source of
the unstable components (potentially a combination of
hydrazine, hydrazine hydrochloride, and intermediate B.) The
thermal stability of hydrazine monohydrate and therefore of the
reaction mixture appears to be reduced by the presence of HCl.
In fact, 0.1 equiv of HCl in hydrazine monohydrate was enough
to lower the onset temperature of decomposition from >300 °C
to around 180 °C. Introduction of a base (sodium acetate) in
F
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
combination with a low-boiling-point solvent (pyridine)
reduced the criticality classification from class 5 (no inherent
safeguards to prevent decomposition) to class 3 (sufficient
thermal potential to reach the boiling point of the solvent but
not enough to reach the onset of the decomposition) by
enhancing the thermal stability of the reaction mass. After safe
operating conditions for the hydrazine condensation reaction
for a loss of cooling upset contingency were identified, the
process was successfully scaled up in pilot plant runs.
Although we cannot disclose structures, the base stabilization
effect has proven effective in other hydrazine condensation
processes examined in our laboratory. In all instances, indazole
entities were prepared with either hydrazine or methylhy-
drazine as the nucleophile. With respect to the SNAr reaction,
our observations include bromo, fluoro, and chloro displace-
ments followed by condensation on a ketone, aldehyde, or
nitrile. Invariably in every case, the exothermic degradation
stage was reached, and addition of base postponed the onset
temperature of the decomposition.
self-heating was detected, the calorimeter maintained adiabatic
conditions by offsetting heat losses through the guard heater
and tracked the exotherm to the temperature or pressure limits
defined for the test.
Test Charges for the Reaction Using NMP as the
Solvent in the Closed-Cell VSP2 Test. Charges for VSP2
test using NMP as the solvent are shown in Table 6.
Table 6. Test Charges for the Reaction Using NMP as the
a
Solvent for the Closed-Cell VSP2 Test
material
amount (g)
2,6-dichlorobenzonitrile
1-methyl-2-pyrrolidinone
hydrazine monohydrate
10.43
32.05
12.50
a
Thermal inertia ϕ = 1.25.
Test Charges for the Reaction Using Pyridine as the
Solvent in the Open-Cell VSP2 Test with Venting to
Ambient. Charges for the vented VSP2 test using pyridine as
the solvent are shown in Table 7. The test cell was equipped
EXPERIMENTAL SECTION
■
Equipment Used. Differential scanning calorimetry was
performed using a Mettler-Toledo DSC821e or DSC823e
calorimeter and sealed glass ampules. Accelerating rate
calorimetry was performed using a Columbia Scientific ARC
and either glass or titanium test cells. VSP2 testing was
performed using Fauske & Associates Vent Sizing Package 2.
Heat of reaction testing was performed using a Mettler-Toledo
reaction calorimeter (RC1) or an Omnical SuperCRC
isothermal calorimeter.
Table 7. VSP2 Test Charges Using Pyridine as the Solvent
with Venting to Ambient
material
amount (g)
2,6-dichlorobenzonitrile
pyridine
7.07
35.93
22.14
hydrazine monohydrate
with two thermocouples, one located ∼1/4″ from the bottom
and the other located near the bottom closer to the outer test
cell wall. The test cell was vented through a 1/16″ line. The
condensate was collected in a flask located on a balance. The
vent valve was opened, and the condensate mass was measured
during the test. The heat−wait−search steps were performed
from 60 °C. An exothermic event was detected at 80 °C.
Test Charges for the Reaction Using Pyridine as the
Solvent and NaOAc as the Base in the Closed-Cell VSP2
Test. Charges for the VSP2 test using pyridine as the solvent
and NaOAc as the base are shown in Table 8. The heat−wait−
search steps were performed from 70 °C. The exothermic event
detected at 80 °C heated the reaction to 140 °C. The heat−
wait−search was restarted with 20 °C steps.
General Procedure for DSC Testing. Samples were first
loaded into and sealed in glass ampules with a capacity of
approximately 50 μL and ramped in the DSC furnace from 0 to
300 °C at 10 °C/min.
General Procedure for ARC Testing. The sample was
loaded into a pressure bomb made of titanium or glass with a
capacity of 10 mL. The temperature and pressure of the test
were recorded. The bomb was inserted into the calorimeter and
heated in heat−wait−search steps of 10 °C starting at 30 °C.
When self-heating was detected, the calorimeter maintained the
airspace around the bomb at the adiabatic condition and
tracked the exotherm to 300 °C or until a pressure of 1200 psia
was reached.
ARC Test Charges. Charges for the reaction using NMP as
the solvent for the ARC test are shown in Table 5.
Table 8. VSP2 Test Charges Using Pyridine as the Solvent
and NaOAc as the Base
Table 5. Test Charges for the Reaction Using NMP as the
Solvent in the ARC Test
material
amount (g)
2,6-dichlorobenzonitrile
pyridine
6.04
28.34
3.35
material
amount (g)
2,6-dichlorobenzonitrile
1-methyl-2-pyrrolidinone
hydrazine monohydrate
0.8131
2.4706
0.9706
sodium acetate
hydrazine monohydrate
18.40
General Procedure for VSP2 Testing. A closed
cylindrical stainless steel test cell (∼120 mL) was placed in a
4 L containment vessel. The test cell was enclosed by a test
heater, a layer of insulation, and a guard heater, which in turn
was enclosed by thermal insulation. Mixing was provided by a
magnetic stir bar. The sample temperature and pressure as well
as the containment pressure were recorded. The VSP2 test cell
with the charges shown in the following tables was loaded into
the VSP2 vessel and heated in heat−wait−search steps. When
Procedure for Heat of Reaction Measurement Using
RC1 with NMP as the Solvent at 110 °C. A 1 L reactor was
used for the testing. Mixing was provided by a mechanical
stirrer. The test charges are shown in Table 9. The reactor with
2,6-dichlorobenzonitrile and NMP was heated to 110 °C. After
the system reached thermal equilibrium at 110 °C and a
calibration was performed, hydrazine monohydrate was added
in 30 min.
G
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 9. RC1 Reaction Test Charges
material
amount (g)
2,6-dichlorobenzonitrile
1-methyl-2-pyrrolidinone
hydrazine monohydrate
96.9
298.1
112.0
ASSOCIATED CONTENT
* Supporting Information
DSC results. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: Zhe.Wang@abbvie.com.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The work presented in the article was sponsored by AbbVie.
AbbVie contributed to the design, research, and interpretation
of data and the writing, reviewing, and approval of the
publication. Z.W., S.M.R., J.G., A.W.K., and M.J.R. are
employees of AbbVie.
REFERENCES
■
(1) Kruger, A. W.; Rozema, M. J.; Chu-Kung, A.; Gandarilla, J.;
Haight, A. R.; Kotecki, B. J.; Richter, S. M.; Schwartz, A. M.; Wang, Z.
Org. Process Res. Dev. 2009, 13, 1419−1425.
(2) Simpson, D. K. 22nd Department of Defense Explosives Safety
Seminar, 1985.
(3) Troyan, J. E. Ind. Eng. Chem. 1953, 45, 2608−2612.
(4) Kuethe, J. T.; Childers, K. G.; Peng, Z.; Journet, M.; Humphrey,
G. R.; Vickery, T.; Bachert, D.; Lam, T. T. Org. Process Res. Dev. 2009,
13, 576−580.
(5) Stoessel, F. Thermal Safety of Chemical Processes: Risk Assessment
and Process Design; Wiley-VCH: Weinheim, Germany, 2008.
H
dx.doi.org/10.1021/op4002577 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX