Method development for the determination of 1,1-dimethylhydrazine by the high-performance liquid chromatography–mass spectrometry technique

Article abstract of DOI:10.1177/1469066718761437

Unsymmetrical dimethyl hydrazine is highly toxic, carcinogenic compound, widely used for organic synthesis and drug development. Therefore, due to its high reactivity, direct analysis is problematic. Current study proposes to use derivatization reaction t

Full text of DOI:10.1177/1469066718761437

Original research paper
European Journal of Mass Spectrometry
(00) 1–8
The Author(s) 2018
0
!
Method development for the determination of
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1469066718761437
journals.sagepub.com/home/ems
1
,1-dimethylhydrazine by the high-performance
liquid chromatography–mass spectrometry
technique
1
2
1
Igors Susinskis , Peteris Mekss and Juris Hmelnickis
Abstract
Unsymmetrical dimethyl hydrazine is highly toxic, carcinogenic compound, widely used for organic synthesis and drug
development. Therefore, due to its high reactivity, direct analysis is problematic. Current study proposes to use derivatization
reaction to increase selectivity and sensitivity of high-performance liquid chromatography–mass spectrometry method.
Different derivatization agents were tested and optimal reaction media was found. Derivatization was performed
by using small amounts of reagents to lower the cost of analysis. The full validation of the method was performed
and it can be used in a routine control in pharmaceutical analysis. Method sensitivity is 0.15 ppm, and linearity range is
0.15–2.70 ppm.
Keywords
Unsymmetrical dimethyl hydrazine, meldonium, ketone, benzoyl chloride, electrospray ionization, derivatization,
high-performance liquid chromatography–mass spectrometry, tandem mass spectrometry, multiple reaction monitoring
(MRM)
Received 13 December 2017; accepted 5 February 2018
The developed method allows to control the pres-
ence of UDMH in anti-ischemia drug meldonium.
The method includes the nonreversible derivatization
Introduction
Unsymmetrical dimethylhydrazine (UDMH) is
a
widely used compound, mostly known as a rocket- of UDMH via nucleophilic substitution in high
1
13
fuel component and a reagent in drug synthesis. Due yield. Four derivatization reagents were tested: pro-
to high toxicity and cancerogenic properties of UDMH, panone, cyclopentanone, benzaldehyde, and benzoyl
its content must to be strictly controlled in drug chloride. The reaction mixture was purified by centri-
2
,3
forms.
Direct analysis of UDMH is complicated fugation. The resulting reaction product was injected in
due its high reactivity. Also, catalytic oxidation and the HPLC/MS/MS system.
decomposition of UDMH are possible on metallic sur-
4
Chromatographic separation occurs on a reverse-
faces. Degradation of this compound can occur during phase chromatography column with ion-pair reagent.
5
storage or in the sample before the injection. Most Due to high polarity of the reaction products, separ-
common methods for determination of UDMH in a ation of similar type of compounds is possible in hydro-
sample are based on compound derivatization, follow- philic interaction liquid chromatography (HILIC)
ing concentration, purification of mixture, and analysis mode as well. From the first sight, HILIC seems to be
1
4
by gas chromatography–mass spectrometry (GC/MS) a preferable technique for separation due to the
6
,7
techniques.
Also, modern methods of analysis absence of signal suppression influence of the ion-pair
report the usage of the desorption electrospray ioniza- reagent. However, sample purification from the
tion as well as a direct analysis in real-time desorption reagents is more complex due to high solubility of
8
atmospheric pressure chemical ionization. Some meth-
ods offer to use ion-selective electrodes for a fast ana-
lysis of UDMH, but the main problem of this method is
a lack of electrode manufacturer and possible negative
9
influence of different ions. There are also studies where
1
JSC ‘‘Grindeks’ ’, Riga, Latvia
2
University of Latvia, R ¯ı ga, Latvia
Corresponding author:
UDMH can be detected by spectrophotometry and
1
chromatography with a UV detector.
Igors Susinskis, JSC ‘‘Grindeks’’, Krustpils 53, Riga, LV-1057, Latvia.
Email: igors.susinskis@gmail.com
0–12

2
European Journal of Mass Spectrometry 0(00)
reagents in acetonitrile (the solvent and main compo- performed (see Table 1). Mobile phase flow was set to
ꢂ1
nent for mobile phase for HILIC). To optimize method 0.25 mL min
.
throughput and to reduce the cost of analysis small Positive ionization mode was used, capillary voltage
amounts of reagents were used. was set to 3.0 kV, cone voltage was set to 27 V, tem-
Derivatization of UDMH was performed with all peratures of ion source and desolvation gas were set to
ꢁ
ꢁ
four reagents and optimization of the reaction media 110 C, and 220 C, respectively. Desolvation and cone
ꢂ1
ꢂ1
was performed. Full validation of the developed gas flow rates were set to 220 L h and 20 L h ,
method was performed. respectively. For a method development, a full scan
mode, in the range of m/z 70–400 was used. To
obtain better sensitivity and selectivity, the final
method contains only one MRM channel with mass
transition m/z 165.1 ! m/z 44.1.
Experimental
Materials
Methanol (HPLC grade), acetonitrile (HPLC grade),
isopropanol (HPLC grade), propanone (99.9%
Optimization of the derivatization agent
purity), cyclopentanone (99% purity), benzaldehyde To check the possibility of the derivatization of
99% purity), benzoyl chloride (99% purity), N,N-dii- UDMH, the following reaction conditions were used.
sopropylethylamine (DIPEA) (99% purity), triethyla- The mixture of 2 mL of solvent (methanol or propylene
(
1
5
mine (TEA) (99.7% purity), UDMH (99% purity), glycol), 10 mL of UDMH, 100 mL of derivatization
propylene glycol (99.5% purity) were purchased from agent were sonicated for 20 min, and then centrifugated
Sigma–Aldrich. Heptafluorobutyric acid (HFBA) (99% (see Scheme 1).
purity) was purchased from Across Organics.
According to the available information, a reaction
between aldehydes and UDMH is rapid without heat-
ing, base or acid catalysts. To simplify the method, no
influence of temperature was tested.
When benzoyl chloride was used as a derivatization
1
5
Instrumentation
All the experiments were carried out on the Waters
2695 HPLC system, coupled with Waters Micromass reagent, then 100 mL of TEA was added to bind hydro-
Quattro Micro API triple quadrupole mass spectrom- gen chloride (see Scheme 2).
eter equipped with electrospray ionization (ESI) source.
Each reaction product was analyzed by HPLC/MS/
Separation was achieved on Waters X-Terra MS-C18 MS using conditions described in a previous section.
analytical column (150 ꢀ 3.0 mm, particle size 3.5 mm),
Optimization of the reaction solvent
ꢁ
column temperature was maintained at 25 C.
Composition of mobile phase A—0.1% solution of
HFBA in water and mobile phase B—0.1% solution Benzoyl chloride was used as a derivatization agent
of HFBA in methanol. Gradient elution was when optimization of the solvent was carried out.
Three reaction media were selected—water, aceto-
nitrile, and isopropanol in different proportions (see
Table 2).
Table 1. Mobile phase gradient program.
Time (min)
Mobile phase A (%)
Mobile phase B (%) MS/MS parameter optimization
0
0
0
7
90
65
40
25
90
90
10
35
60
75
10
10
Optimal cone voltage of ion source for N ,N -dimethyl-
benzohydrazide was optimized by direct syringe pump
injection, at flow rate 20 mL/min; maximal signal inten-
sity was achieved without fragmentation. For the
MRM method development, argon-filled collision cell
parameters were optimized. The daughter ion with the
most intense signal was searched. Fragmentation of
meldonium was performed in similar conditions.
1
1
1
2
2
8
9
5
Sample volume was 20 mL.
Scheme 1. Reaction between carbonyl compound and UDMH.

Susinskis et al.
3
Scheme 2. Reaction between benzoyl chloride and UDMH.
Table 2. Reaction media volumes.
Linearity and range
For linearity tests, 10 solutions were used in a range of
.15–2.70 ppm for UDMH content in meldonium.
Experiment No.
1.
2.
3.
4.
5.
–
0
Acetonitrile
Isopropanol
Water
2.0 mL 1.0 mL 1.0 mL
–
–
–
–
2.0 mL 1.0 mL
Accuracy recovery
–
–
1.0 mL
75 mL
50 mL
–
1.0 mL
75 mL
50 mL
Accuracy and recovery were tested at 10% (limit of
quantification value), 80%, 100%, and 120% of max-
imum allowed content of UDMH, compare for impur-
ity free spiked samples with standard solution at 100%
DIPEA
75 mL
75 mL
50 mL
75 mL
50 mL
Benzoyl chloride 50 mL
N,N-diisopropylethylamine (DIPEA) was used as an organic base instead of
triethylamine (TEA). Possibly not only DIPEA but also meldonium impacts
on pH of reaction media. Due to this reason, meldonium was added to each
reaction solution.
(1.5 ppm) allowed content of UDMH.
Precision
Such parameters such as system precision, method pre-
cision, and intermediate precision were tested. System
precision was tested by six injections of meldonium
Possibility to shorten analysis time
To reduce time of analysis, experiments with isocratic solution, spiked by UDMH. Meldonium solution was
elution were carried out. Only one compound appears prepared as described above. For method precision, six
on MRM channel, but elution of meldonium and independently prepared solutions were analyzed. For
DIPEA also occurs.
intermediate tests, the experiments were performed by
another analyst on similar equipment, and the results
were compared.
Method validation
A full method validation was performed according to
International Council for Harmonisation of Technical
Requirements for Pharmaceuticals for Human Use Stability of the method was tested by changing amount
Robustness
(
The most important parameters are described below.
The sample was prepared by mixing 133 mg of mel- of the meldonium-spiked sample, after preparing by
donium with 1 mL of water and 1 mL of isopropanol. derivatization process, and following clean up proced-
Seventy-five microliters of DIPEA and 50 mL benzoyl ure, was checked after storage during 48 h. Spiked at
chloride were added. Reaction mixture was stirred for 100% sample was used.
ICH) and European Pharmacopoeia (EF) guidelines. of ion-pairing reagent in mobile phase (ꢃ10%), analyt-
ꢁ
ical column, and column temperature (ꢃ5 C). Stability
3
0 s, and stored for 1 h at room temperature. Then, the
mixture was diluted with the mobile phase A to 20 mL,
stored for 1 h at room temperature, and centrifuged for
Stress studies
6
min at 16,000 g. For all experiments, UDMH was Due to ICH requirements, forced degradation of meldo-
used as a free base. Due to manufacturing process, nium was performed. The main purpose of stress studies
the presence of a counter anion is not possible in mel- was to prove that UDMH could not be formed in mel-
donium, so the salt of UDMH cannot be formed. donium substance during storage. Photodegradation,
The pH value of meldonium water solutions is usually oxidative degradation by hydrogen peroxide, thermal
8
form with possible counter anion.
–8.5. Also DIPEA acts as organic base and allow to degradation, hydrolytic degradation under alkali and
acidic conditions were carried out. Five percent hydro-
gen peroxide solution, 1 M NaOH solution, and 1 M
HCl solution were used. Photodegradation was per-
formed according to ICH guidelines under xenon arc
Sensitivity of the method
As criteria for sensitivity, a signal-to-noise ratio was light source. All other of degradation experiments were
chosen for impurity free spiked sample. performed in hermetically sealed vial for 2 h.

4
European Journal of Mass Spectrometry 0(00)
Figure 1. Chromatogram of four reaction products.
Table 3. Reaction products comparison.
1,1-dimethyl-2 -
(propan-2-ylidene) 2-cyclopentylidene-1, (2 Z)-2-benzylidene-1,
hydrazine 1-dimethylhydrazine 1-dimethylhydrazine
0
0
Product name
N ,N -dimethylbenzohydrazide
þ
[
M þ H] , m/z of ion
Retention time, min
101.2
8.97
1.31
781
127.2
11.96
1.12
843
149.2
18.33
1.11
1600
6.8
165.1
15.20
1.10
As
S/N ratio
8857
100.0
Relative response, compare to
0
7.9
8.2
0
N ,N -dimethylbenzohydrazide, %
Results
observed
only
when
1,1-dimethyl-2 -(propan-
-ylidene)hydrazine was formed. Comparison of the
2
reaction products is shown in Table 3.
Choice of the derivatization agent
0 0
Formation of N ,N -dimethylbenzohydrazide for
The reaction between UDMH and each derivatization
reagent resulted in a desired product. Gradient elution, derivatization of UDMH is preferable because of the
described before, allows to resolute all the compounds. higher signal-to-noise ratio and higher response (see
Unacceptable shape of the peak (see Figure 1) was Table 3). Moreover, only this component is solid,

Susinskis et al.
5
while all the other products are liquids. This fact allows dimethylbenzohydrazide in this media. Moreover,
to obtain dry product by vacuum evaporation. such nucleophilic substitution reaction better occurs
1
6
Moreover, due to high density of benzoyl chloride, in polar solvents.
compared to water, the sample can be simply purified
by centrifugation. So, N ,N -dimethylbenzohydrazide
was chosen for the next experiments.
0
0
MS/MS parameters optimization
Optimal cone voltage of 27 V was found. The most
intensive daughter ion m/z is 44.1, but optimal collision
energy is 16 eV, and dwell time is 0.2 s (see Figure 2).
Formation of ion with m/z 59.1 was observed during
Choice of the reaction solvent
Different yield of the derivatization reaction was
0
0
observed when different reaction solvents or concen- fragmentation of both N ,N -dimethylbenzohydrazide
trations were used. Reaction yield was compared to and meldonium.
0
0
in-house standard of N ,N -dimethylbenzohydrazide, by
comparing areas of peaks on the corresponding MRM
Sensitivity of the method
0
0
channel of N ,N -dimethylbenzohydrazide. Results of
these experiments are presented in Table 4.
Minimal chosen criteria of S/N ratio for LOQ are 10
The highest yield (91.7%) of the product was and 3 for LOD. Obtained values are 0.09 ppm and
obtained when 50% isopropanol solution in water 0.03 ppm, respectively. Baseline shows no fluctuation
was used. This high yield can be explained by good or drift near the peak of the interest (see Figure 3).
0
0
solubility of UDMH, meldonium, and N ,N -
Linearity and range
Every concentration was introduced in a system in trip-
licate. The obtained regression model is described by
Table 4. Yields of reactions in different reaction media.
2
linear function with correlation coefficient R ¼ 0.9990
Reaction media type
Acetonitrile, 2.0 mL
Mean yield (%)
RSD (%)
(see Figure 4); value is acceptable for pharmaceutical
analysis.
1
7
71.1
62.6
6.4
2.8
Acetonitrile, 1.0 mL (higher
concentration of reagents)
Accuracy recovery
50% Acetonitrile in water
solution, 2.0 mL
6.1
21.5
Every concentration was introduced in a system in
triplicate. Obtained overall recovery value is 97.1%,
with RSD 7.6%, but recovery values and relative stan-
dard deviation for every concentration are showed in
Table 5.
Isopropanol, 2.0 mL
44.3
91.7
6.3
2.9
50% Isopropanol solution
in water, 2.0 mL
0
0
Figure 2. Fragmentation pattern of N ,N -dimethylbenzohydrazide at collision energy 16 eV.

6
European Journal of Mass Spectrometry 0(00)
0
0
Figure 3. Chromatogram of N ,N -dimethylbenzohydrazide at S/N ratio 10.
1
1
200
000
y = (363 ± 4) x + (10 ± 6)
R² = 0.9990
8
00
Area, counts 600
4
2
00
00
0
0
.00
0.50
1.00
1.50
2.00
2.50
3.00
UDMH concentration, ppm
Figure 4. Calibration curve for UDMH in concentration from 0.15 ppm to 2.70 ppm.
Table 5. Recovery rates for different unsymmetrical dimethylhydrazine (UDMH) concentrations.
UDMH concentration
in analyzed solutions
ng/mL
Concentration of
maximal allowed
content (%)
UDMH concentration
in active pharmaceutical
ingredient (ppm)
Recovery
(%)
RSD
(%)
1
8
.0
.0
10.0
80.0
0.15
1.20
1.50
1.80
101.4
98.2
99.6
89.0
11.6
3.4
2.8
3.1
1
1
0.0
2.0
100.0
120.0
Table 6. Robustness experiments.
Precision
Absolute
deviation
of area (%)
RSD between areas of six injections in the system pre-
cision experiment is 2.4%. RSD is 3.6% in the method
precision experiment, and for intermediate precision
Type of experiment
experiment, RSD between both analysts’ results is 0.9% of HFBA in mobile phase
.7%.
0.7
0.7
0.3
6.9
5.0
4.6
2
1.1% of HFBA in mobile phase
Similar analytical column, another batch
ꢁ
Robustness
Column temperature 20 C
ꢁ
Column temperature 30 C
0
0
No significant influence on area of N ,N -dimethylbenzo-
hydrazide peak was observed during the robustness
Sample after 48 h of storage
experiments. Stability of derivatized sample allows to HFBA: heptafluorobutyric acid.

Susinskis et al.
7
Figure 5. Chromatogram of analyzable mixture under isocratic elution.
increase the method throughput and resolve main prob-
lem of UDMH oxidation in water solutions (see Table 6).
Conclusion
The developed method allows to use high activity of
UDMH in an appropriate way for analysis by simple
derivatization reaction. Moreover, mass increase of
Stress studies
As it was predicted, no formation of UDMH was derivatization product allows to obtain higher selectiv-
1
9
detected during any type of degradation.
ity and sensitivity. Methodology does not use such
concentration techniques as evaporation and allows to
use centrifugation for the purification of the reaction
media. LOQ of the method is about 0.15 ppm and lin-
Possibility to shorten analysis time
To achieve shorter analysis time, optimal isocratic sep- earity range is 0.15–2.70 ppm. After preparation pro-
aration conditions were found. It is possible to separate cess, analyzable sample is stable at least 48 h. All the
0
0
meldonoium, DIPEA, and N ,N -dimethylbenzohydra- data obtained during the validation allow to use the
zide by use of the same mobile phases in 75:25 propor- method in routine pharmaceutical control.
tions (see Figure 5).
1
8
Because of possible impurity presence in UDMH,
authors do not recommend to use isocratic elution, if We thank Olesja Koleda for support and additional work on
presence of another hydrazines must be controlled. this study.
Acknowledgments

8
European Journal of Mass Spectrometry 0(00)
Declaration of conflicting interests
9. Sobhanmanesh B and Najafi M. Selective determination
of unsymmetrical dimethylhydrazine on a Prussian blue
modified carbon paste electrode. J Braz Chem Soc 2015;
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
2
6: 451–457.
1
0. Smolenkov AD, Rodin IA and Shpigun OA.
Spectrophotometric and fluorometric methods for the
determination of hydrazine and its methylated analogues.
Anal Chem 2012; 672: 98–113.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
1
1
1
1. Bicchi C, Cordero C, Rubiolo P, et al. Determination of
daminozide residues in apple pulp using HPLCꢂDAD-
UV. J Agric Food Chem 2001; 49: 3548–3552.
2. Elias G and Bauer WF. Hydrazine determination in
sludge samples by high-performance liquid chromatog-
raphy. J Sep Sci 2006; 29: 460–464.
References
1
. Voloshchuk G, Yagupolskyi YL, Tanchuk YV, et al. The
problem of utilization (detoxification) unsymmetrical
dimethylhydrazine. Challenge for synthesis of mildronate
and its structure analogues. EPMA J 2014; 5: A92.
3. Danikiewicz W, Bien
Aromatic nucleophilic substitution (SNAr) reactions of
,2- and 1,4-halonitrobenzenes and 1,4-dinitrobenzene
´
kowski T, Kozlowska D, et al.
2
. U.S. Departament of health, education, and welfare.
Public Health Service, Center for Disease Control,
National Institute for Occupational Safety and Health.
Criteria for a recommended standard. Occupational expo-
sure to hydrazines. The National Institute for Occupational
Safety and Health (NIOSH) 1978; 78–172; 1–269.
1
with carbanions in the gas phase. J Am Soc Mass
Spectrom 2007; 18: 1351–1363.
1
4. Hmelnickis J, Pugovics O, Kazoka H, et al. Application
of hydrophilic interaction chromatography for simultan-
eous separation of six impurities of mildronate substance.
J Pharm Biomed Anal 2008; 48: 649–656.
3
4
5
. Vernot EH, MacEwen JD, Bruner RH, et al. Long-term
inhalation toxicity of hydrazine. Fundam Appl Toxicol
1
1
5. Siggia S and Stahl CR. Determination of aldehydes using
unsymmetrical dimethylhydrazine. Anal Chem 1955; 27:
1985; 5: 1050–1064.
. Lunn G and Sansone EB. Oxidation of 1,1-dimethylhydra-
zine (UDMH) in aqueous solution with air and hydrogen
peroxide. Chemosphere 1994; 29: 1577–1590.
. Pestunova O, Elizarova G, Ismagilov Z, et al.
Detoxication of water containing 1,1-dimethylhydrazine
1
975–1977.
6. Uggerud E. Reactivity trends and stereospecificity in
nucleophilic substitution reactions. J Phys Org Chem
2006; 19: 461–466.
by catalytic oxidation with dioxygen and hydrogen perox- 17. Kazakevich YV and LoBrutto R. HPLC for pharmaceut-
ide over Cu- and Fe-containing catalysts. Catalysis Today
002; 75: 219–225.
. Rutschmann M and Buser H. Determination of damino-
ical scientists. New Jersey, Hoboken: John Wiley & Sons,
Inc., ISBN: 978-0-471-68162-5 January 2007. DOI:
10.1002/0470087951.
2
6
zide and dimethylhydrazine residues in Swiss apple juice 18. Buryak AK, Serdyuk TM and Ul’yanov AV.
concentrates using gas chromatography-mass spectrom-
etry. J Agric Food Chem 1991; 39: 176–181.
. Smirnov RS, Rodin IA, Smolenkov AD, et al.
Determination of the products of the transformation of
Investigation of the reaction products of unsymmetrical
dimethylhydrazine with potassium permanganate by gas
chromatography-mass spectrometry. AV Theor Found
Chem Eng 2011; 45: 550.
7
unsymmetrical dimethylhydrazine in soils using chromatog- 19. Gao S, Zhang Z-P and Karnes HT. Sensitivity enhance-
raphy/mass spectrometry. Moscow, Russia: Department
of Chemistry, Moscow State University, 2010.
. Griffin T and Berger C. Three methods of detection of
hydrazines. NASA Tech Briefs 2010; 05–10: 7–8.
ment in liquid chromatography/atmospheric pressure
ionization mass spectrometry using derivatization and
mobile phase additives. J Chromatogr B 2005; 825: 98.
8