Investigation on the formation and hydrolysis of N,N-dimethylcarbamoyl chloride (DMCC) in vilsmeier reactions using GC/MS as the analytical detection method

Article abstract of DOI:10.1021/op900018f

The formation of N,N-dimethylcarbamoyl chloride (DMCC) in Vilsmeier reactions (VRs) and the hydrolysis of DMCC during aqueous workup were investigated. The amount of DMCC formed was dependent on the chlorination reagent. The activity order was found to be

Full text of DOI:10.1021/op900018f

Organic Process Research & Development 2009, 13, 857–862
Investigation on the Formation and Hydrolysis of N,N-Dimethylcarbamoyl Chloride
(DMCC) in Vilsmeier Reactions Using GC/MS as the Analytical Detection Method
Martina Stare, Krzysztof Laniewski, Andreas Westermark, Magnus Sj o¨ gren, and Wei Tian*
Global Process R&D S o¨ dert a¨ lje, AstraZeneca, 151 85 S o¨ dert a¨ lje, Sweden
Abstract:
Scheme 1. Chlorine replacement of a hydroxyl group by
Vilsmeier reagent 1
The formation of N,N-dimethylcarbamoyl chloride (DMCC) in
Vilsmeier reactions (VRs) and the hydrolysis of DMCC during
aqueous workup were investigated. The amount of DMCC formed
was dependent on the chlorination reagent. The activity order was
found to be: thionyl chloride > oxalyl chloride > phosphorus
oxychloride. The concentrations of DMCC found in the tested
reactions were normally at a level of 0-20 ppm. In the presence
of a base the level was higher. DMCC hydrolyzed quickly under
aqueous workup conditions. No more than 3 ppm of DMCC could
be detected in the product after aqueous workup of the tested VR.
Our results clearly indicate that by following the described
procedure for VR it is perfectly safe from the health risk point of
view to operate this chemistry when preparing compounds for
pharmaceutical use. A very useful analytical procedure for
detection and quantification of trace amounts of DMCC formed
in the VR was developed. The method was based on derivatization
of DMCC with ethanol to form ethyl N,N-dimethylcarbamate and
analysis by GC coupled to a mass spectrometer in selected ion
monitoring mode. The analytical method was selective and showed
very good linearity. The limit of detection (LOD) and limit of
quantification (LOQ) of DMCC using standard addition analysis
with real Vilsmeier reaction matrices were determined to be LOD
6
by D. Levin in light of the fact that DMCC is a potential
carcinogen according to the International Agency for Research
on Cancer (IARC) evaluations. During the last five years an
increased interest regarding potential genotoxic impurities
7,8
(
PGIs) has emerged from health authorities and the phamaceu-
tical industry. PGIs have had such an impact that process
chemists must consider changing the reagent(s), choosing
another route or applying a control strategy to eliminate any
PGIs. At the moment there is inadequate evidence for carci-
nogenicity of DMCC in humans, although there is sufficient
evidence for carcinogenicity in animals. Thus, DMCC generated
in an organic process for the manufacturing of an API can be
defined as a PGI. The permitted level is dependent on the daily
9,10
dose, but DMCC is normally only allowed in ppm levels,
which puts challenges into the analytical methods.
The concerns around the formation of DMCC under the VR
conditions have resulted in restricted application of the VR in
the process development of pharmaceuticals even though little
is reported as to how much or about how much DMCC is really
generated during the reactions. In an ongoing project, we
decided to apply the VR in the penultimate step, the chlorination
of 2 for the synthesis of 5 (Scheme 2). In the process, DMF is
utilized both as a protecting agent for the formation of the
hydroxyl amidine 3 and as a catalyst for the Vilsmeier
chlorination of 3 to generate the chlorinated amidine 4 (Scheme
)
0.2 ppm and LOQ ) 0.7 ppm.
Introduction
The Vilsmeier reaction (VR), or Vilsmeier-Haack reaction,
1,2
has a large range of applications in organic synthesis. One of
the most important applications is the chlorination of an alcohol
as is shown in Scheme 1. The actual chlorination reactant is
the in situ generated Vilsmeier reagent 1, which is formed by
reaction of the chlorination agent and N,N-dimethylformamide
2
). In order to assess the health safety risk of the process before
scaling it up into the pilot scale we decided to determine the
amount of DMCC generated under the reaction conditions
employed.
A survey of the literature revealed that there was neither a
suitable analytical method nor data available for the detection
(
DMF), the latter normally present in catalytic amounts.3
N,N-dimethylcarbamoyl chloride, also called DMCC, is an
active alkylating agent and has been used as an intermediate in
the manufacture of a number of pharmaceuticals and pesti-
cides.The formation of DMCC under VR conditions was
4,5
reported as early as the 1960s. The health safety concerns
regarding the application of VR in organic synthesis was raised
(
(
(
6) Levin, D. Org. Process Res. DeV. 1997, 1, 182.
7) IARC Monograph 1999, Vol. 71, pp 531-543.
8) Lohman, P. H. M. Mutat. Res. 1999, 428, 237–254.
*
To whom correspondence should be addressed. Wei Tian, PR&D Astra-
(9) M u¨ ller, L.; Mauthe, R. J.; Riley, C. M.; Andino, M. M.; Antonis, D. D.;
Beels, C.; DeGeorge, J.; De Knaep, A. G.; Ellison, D.; Fagerland,
J. A.; Frank, R.; Fritschel, B.; Galloway, S.; Harpur, E.; Humfrey,
C. D.; Jacks, A. S.; Jagota, N.; Mackinnon, J.; Mohan, G.; Ness, D. K.;
O’Donovan, M. R.; Smith, M. D.; Vudathala, G.; Yotti, L. Reg.
Toxicol. Pharmacol. 2006, 44, 198–211.
Zeneca, R&D S o¨ dert a¨ lje SE-15185, Sweden. Telephone: +46 8 553 251 83.
Fax: +46 8 553 242 76. E-mail: wei.tian@astrazeneca.com.
(
(
(
1) Qian, D. Q.; Cao, R. Z.; Liu, L. Z. Youji Huaxue 2000, 20, 30–43.
2) Marson, C. M. Tetrahedron 1992, 48, 3659–3726.
3) Hepburn, D. R.; Hudson, H. R. J. Chem. Soc., Perkin Trans. I 1976,
7
54–757.
(10) Question & Aswers on the CHMP Guideline on the Limits on
Genotoxic Impurities. EMEA/CHMP/SWP/431994/2007, revision 1;
European Medicines Agency (EMEA): London, UK, 26 June 2008.
(
(
4) K u¨ hle, E. Angew. Chem., Int. Ed.Engl. 1962, 1 (12), 647–652.
5) Hasserodt, U. Chem. Ber. 1968, 101, 113–120.
1
0.1021/op900018f CCC: $40.75  2009 American Chemical Society
Vol. 13, No. 5, 2009 / Organic Process Research & Development
•
857
Published on Web 07/02/2009

Scheme 2. Chlorination of compound 2 using a Vilsmeier reaction, with in situ protection of the amino group by using DMF
Table 1. Detection of DMCC in the reaction of different chlorinating agents with DMF
molar ratio
(DMF: mhlorinating agent)
entry
chlorinating agent
solvent
temp. (°C)
time (min)
DMCC detected (ppm)
1
2
3
4
5
6
7
8
9
SOCl
SOCl
SOCl
POCl
POCl
POCl
SOCl
POCl
POCl
SOCl
POCl
POCl
2
2
2
3
3
3
2
3
3
2
3
3
neat
1:0.4
23
23
23
23
23
23
23
23
65
23
70
65
23
30
30
11
neat
1:1.1
6
neat
1:2.4
30
13
neat
1:0.4
30
1.1
0.8
1.0
2.4
1.9
neat
1:1.1
30
neat
1:2.4
30
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
1:2.4
30
b
1:2.7
30
1:2.7
180
30
∼7
a
10
11
12
13
:pyridine
:pyridine
:2
1:1:2.4
1:1:2.7
1:0.8:2.7
1:2.6
2400
b
30
180
120
17
104
(COCl)
2
Detected but not quantified
due to poor solubility
14
15
16
(COCl)
2
CH
CH
CH
2
Cl
2
Cl
2
Cl
2
2
2
1:2.6
1:2.6
1:2.6
23
23
23
30
30
30
2
POCl
SOCl
3
2
0.2
5.0
a
The amount of DMCC is estimated. It exceeded the linear range of this method. b Some emulsion was observed.
1
1
of DMCC in a VR matrix. Matienzo and co-workers reported
a direct analysis by GC for the determination of DMCC in air
for as low as ppb levels by concentrating the sample on an
3 2
(POCl ), and oxalyl chloride [(COCl) ], were investigated. The
results are summarized in Table 1. A neat mixture of DMF
and thionyl chloride (ratio 1:0.4; 1:1.1, and 1:2.4) at room
temperature after 30 min generated levels of DMCC at 11 ppm,
12
absorbing polymer before analysis. Rusch and co-workers
reported a quantitative analytical method for determination of
DMCC in air by derivatization of the sample with 4-(p-
nitrobenzyl)pyridine and spectrophotometric analysis of the
derivative. Alternative analytical techniques such as LC/MS and
GC/MS, without derivatization, had been investigated but
6
ppm, and 13 ppm (entries 1-3). Under similar reaction
conditions using similar equivalents of POCl the reactions
yielded 1.1 ppm, 0.8 ppm, and 1.0 ppm of DMCC, respectively
entries 4-6). The results from these experiments show that
3
(
increasing the equivalents of the chlorinating reagent does not
increase the amount of DMCC proportionally. Another impor-
13
suffered from corrosion problems.
3
tant observation is that POCl gave a considerable reduction of
the amount of DMCC obtained.
Results and Discussions
Quantitative Determination of DMCC under Various
Conditions. With the development of GC/MS in selected ion
monitoring (SIM) method we determined the amounts of
DMCC generated in the VR under different conditions. Reac-
tions of DMF with the three most commonly used chlorination
Most VRs will be run in a solvent, and the solvent of interest
to us was dioxane. When repeating the experiments above with
dioxane present, SOCl /DMF gave 2.4 ppm, and POCl /DMF
2 3
gave 1.9 ppm of DMCC. The analytical results showed that a
very low level of DMCC was found, and the difference between
2
agents, thionyl chloride (SOCl ), phosphorus oxychloride
2 3
SOCl and POCl is now very small (entries 7-8).
(
(
(
11) Matienzo, L. J.; Hensler, C. J. Am. Ind. Hyg. Assoc. J. 1982, 43 (11),
38–844.
12) Rusch, G. M.; LaMendola, S. L.; Kats, G. V.; Laskin, S. Anal. Chem.
An emulsion was observed in the mixture of POCl , DMF,
3
8
and dioxane at room temperature. To minimize the formation
of emulsion an experiment at increased temperature (65 °C
instead of 23 °C) was performed. The amount of DMCC was
then increased up to approx 7 ppm (entry 9).
1
976, 48 (14), 2259–2261.
13) When GC/MS was employed for the analysis of DMCC in the mixture
of thionyl chloride and DMF, the injector was severely corroded. When
LC/MS was used, the injector syringe was corroded.
8
58
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development

Scheme 3. Formation of DMCC in the reaction of DMF
with SCl and SO Cl
Table 2. Determination of the amount of DMCC formed in
the chlorination of compound 2
2
2
2
entry sampling stage temperature (°C) time (h) DMCC (ppm)
1
2
3
amidine 3, 4
5
5 (isolated)
60
70
23
3.0
0.5
na
87
0.9
2.8
Table 3. Experimental data from the hydrolysis of DMCC
in the presence of POCl
3
-3
t [s]
C
t
[mol/L] × 10
ln C
t
0
11.4100
10.1100
8.7700
7.4900
6.1990
5.3670
4.5160
3.7548
3.2324
2.5160
1.2733
0.8104
0.2569
0.0764
0.0316
-4.47
-4.59
-4.74
-4.89
-5.08
-5.23
-5.40
-5.59
-5.74
-5.98
-6.67
-7.12
-8.27
-9.45
-10.36
2
3
4
6
7
8
40
60
80
00
20
40
A base is often employed in VRs to neutralize the HCl
generated during the reaction. Pyridine is one of the most
commonly used bases in this regard. Our interest then turned
to investigating the effect that addition of a base, such as
pyridine, could have. The results show that it had a great impact
960
1
1
1
080
200
500
1800
2
3
3
400
000
600
2
on the formation of DMCC when SOCl and DMF were used
in dioxane. The amount of DMCC formed in this case was as
high as 2400 ppm (compare entries 7 and 10). An enhancement
Comparing the data in entries 14 to 16 we could find that the
reactivity of the three chlorination reagents for the formation
of the amount of DMCC could be detected also when POCl
was used, but not to the same extent (compare entries 9 and
1). When the pyridine was replaced by the substrate 2 that is
also a base, the amount of DMCC increased from 2.7 ppm to
04 ppm (compare entries 7 and 12), which indicates that the
3
of DMCC on reaction with DMF is: SOCl
2
(5 ppm) > (COCl)
2
1
(
2 ppm) > POCl
3
(0.5 ppm).
The formation of DMCC in the chlorination of compound
using POCl was investigated, and the results are found in
1
2
3
presence of a base increases the formation of DMCC. The
mechanism of the formation of DMCC under VR conditions
has not been thoroughly investigated. We believe that in the
case when thionyl chloride is used as the chlorinating reagent
two possible routes can lead to the formation of DMCC: one
involving sulfuryl chloride and one involving sulfur dichloride
as shown in Scheme 3. The reversible reduction-oxidation of
thionyl chloride can generate sulfuryl chloride and sulfur
Table 2. The reaction was performed in dioxane and was
complete within 3 h at 60 °C. The analysis of the reaction
mixture gave 87 ppm of DMCC (entry 1). On completion, the
reaction mixture was quenched with water and hydrolyzed for
30 min to remove the amino protection group. After the
hydrolysis the concentration of DMCC in the reaction mixture
was 0.9 ppm (entry 2), and the isolated product 5 (yield 90%)
contained only 2.8 ppm of DMCC (entry 3).
14
dichloride. The sulfuryl chloride reacts with DMF to form
complex 6 that on elimination of SO and HCl yields DMCC
as suggested by K u¨ hle. The sulfur dichloride forms complex
with DMF, which on elimination of S(s) and HCl yields
Hydrolysis of DMCC in Dioxane/Water under Acidic
Conditions. VRs are often followed by an aqueous workup
in an organic process. This offers an opportunity to reduce
the amount of DMCC once it is formed because of its
fast hydrolysis. Solvolysis of DMCC in pure water has
been carefully investigated over a wide ranges of tem-
2
4
7
5
DMCC. The formation of DMCC is an irreversible process
due to the formation of gaseous sulfur dioxide and solid sulfur.
This process is apparently base promoted by capture of the
proton generated during the reaction. This explains why there
was a significant amount of DMCC formed with sulfur
15
perature by Queen. The rate constant for the hydrolysis
in H O is >0.0025/s, and its half-life in moist air (50%
2
relative humidity) is approximately 3 h, both of which
precipitation in entry 10. This is in line with the observation
16
indicate a fairly rapid hydrolysis.
5
by Hasserodt where a reaction of SCl
2
and DMF gave as much
To get a better understanding of the hydrolysis of DMCC
under our process conditions, we performed a reaction at our
standard conditions without the substrate compound 2. Thus, a
as 25% of DMCC.
DMCC was detected in the reaction mixture of DMF and
oxalyl chloride in dioxane but not quantified due to the poor
solubility of the mixture in dioxane (entry 13). In order to
determine the amount of DMCC formed we replaced dioxane
mixture of DMCC, dioxane, and POCl was heated to 80 °C
3
and then quenched with water. Samples were taken from the
quenched reaction mixture and analyzed with the GC/MS
with CH
4). Since the reaction mixtures of DMF with both POCl
SOCl are soluble in dichloromethane we determined the
amount of DMCC in these systems as well (entries 15 and 16).
2 2
Cl
and found 2 ppm of DMCC in the solution (entry
method. The concentration of DMCC found at each time (C
and the ln C data are listed in the Table 3. The plot of ln C
against the reaction time is shown in Figure 1. The good linear
t
)
1
3
and
t
t
2
(
15) Queen, A. Can. J. Chem. 1967, 45, 1619–1629.
(
14) Evans, A. G.; Meadows, G. W. Trans. Faraday Soc. 1947, 43, 667–
74.
(16) Dangerous Properties of Industrial Materials Report; Van Nostrand
Reinhold: New York, 1987; Vol. 7 (1), pp 51-54.
6
Vol. 13, No. 5, 2009 / Organic Process Research & Development
•
859

GC/MS analyses in full scan mode of standard solutions of
EDMC and naphthalene were carried out to decide which ions
should be selected for SIM analysis. For each compound three
ions were selected. One ion was used for quantifications, and
the other two were chosen as qualifiers (see Experimental
Section for detailed information regarding the selected ions).
Figure 2 shows mass spectra of both EDMC and naphthalene.
Figure 3 shows the chromatograms from GC/MS-SIM analysis
of a sample obtained after one of the VRs.
The concentration of DMCC in the analyzed samples was
determined by standard addition analysis at four levels. This
analytical approach is used when sample matrix can cause
incorrect readings from a calibration curve based on pure
standards. For all four samples the area ratio of EDMC to IS
detected by GC/MS-SIM vs the amount of the sample were
plotted against the amount of added standard of DMCC. The
quantity of DMCC was calculated by extrapolating the calibra-
tion curve to a point at a zero signal, and the absolute value of
the x-intercept was the amount of the DMCC in the analysed
sample.
3 t
Figure 1. Hydrolysis of DMCC in the presence of POCl : ln C
plotted against time.
Table 4. Hydrolysis rate constant and half-life time of
DMCC under different conditions
temperature
rate constant half-life
-1
entry
(°C)
reaction conditions
k (s )
τ (min)
1
2
80
80
dioxane/ POCl
dioxane/ POCl
3
/ H
/ H
2
O
O
0.0017
0.0028
6.80
4.13
3
2
(
In the presence of 2)
3
50
dioxane/ POCl
3
/D
2
O
0.0010
11.55
Scheme 4. Derivatization of DMCC with ethanol for the
detection and determination of DMCC formed in the
Vilsmeier reaction
Qualification of the Analytical Method. SelectiVity. Since
the SIM mode with three characteristic ions recorded for the
analysed substance was used, the method was considered to be
selective, when the inspection of the intensity of the ions showed
consistency with the expected abundance ratio between them,
and the retention time was the same for the monitored ions.
Fulfilling these required conditions assured that the desired
compound was detected and quantified.
relationship between ln C
DMCC exhibits the expected first-order reaction: ln C
ln C . From this equation we could determine the hydrolysis
t
and t indicates that the hydrolysis of
t
) -kt
+
0
Linearity. The results from standard addition analysis showed
a very good linear correlation between the amount of DMCC
added and the area ratio of the detected EDMC and IS. The
-1
rate constant as k ) 0.0017 s . Similar experiments were
carried out in the presence of compound 2, and the hydrolysis
rate constant was found to be k ) 0.0028 s . The half-life (τ)
-1
2
correlation coefficient for all analysed samples was R > 0.9.
of DMCC under these conditions was 6.80 and 4.13 min,
Figure 4 presents an example of this linear response from the
GC/MS-SIM analysis of one sample from a VR studied during
the method development.
respectively, based on the equation τ ) ln 2/k. We have also
1
investigated the hydrolysis in the presence of POCl
3
by H NMR
and found the half-life of DMCC to be 11.55 min at 50 °C
entry 3, Table 4).
Limit of Detection and Limit of Quantification. Limit of
detection (LOD) and limit of quantification (LOQ) of DMCC
with the Vilsmeier reaction matrices were determined to be
LOD ) 0.2 ppm, and LOQ ) 0.7 ppm. These values of LOD
and LOQ were determined with the sample matrix at signal-
to-noise ratios (S/N) of S/N ) 3 for LOD and S/N ) 10 for
LOQ.
(
These results show that even though DMCC might be
formed during the Vilsmeier reaction, an aqueous workup can
eliminate it down to a very low level. It should also be possible
to capture and hydrolyse DMCC in a scrubber from the process
ventilation system to prevent air contamination.
Development of the Analytical Method. The analyses of
standard samples of the ethanol adduct of DMCC, i.e. ethyl
N,N-dimethyl carbamate (EDMC), using GC with flame ioni-
sation detection (FID) showed that this analytical technique
could be used for detection and quantification of DMCC.
However, when authentic samples from the Vilsmeier synthesis
were analysed, interferences from the sample matrix were
observed. Therefore, an analytical procedure involving deriva-
tization of DMCC with ethanol to form EDMC (Scheme 4)
and a subsequent analysis by GC/MS-SIM using an internal
standard (IS) was developed. A number of compounds were
tested as the IS with respect to peak shape and retention time.
Naphthalene was chosen as the most suitable compound. Using
GC/MS in SIM mode only few chosen ions were recorded, and
the interference from compounds eluting at the same time range
could be minimised or avoided. Moreover, the sensitivity of
the analysis was improved.
Conclusions
In the presented investigation a robust analytical procedure
for measuring DMCC, using derivatisation with ethanol and
GC/MS analysis in selected ion monitoring mode has been
developed. The performance of the analytical method shows
that it can accurately measure DMCC in Vilsmeier reaction
mixtures down to single digit ppm levels.
The analytical method enabled us to investigate the impact
of some common chlorinating agents on the formation of
DMCC in the Vilsmeier reaction. In all cases the amount of
DMCC was low (<15 ppm), and the established order was
thionyl chloride > oxalyl chloride > phosphorus oxychloride,
with the latter yielding the lowest amount of DMCC. It was
also found that, when an organic base was present in the reaction
mixture, it profoundly increased the amount of DMCC formed.
8
60
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development

Figure 2. Mass spectra of EDMC (the derivative of DMCC) and naphthalene (chosen as the internal standard), respectively, acquired
by GC/MS analysis in full scan mode of pure standard samples. Characteristic ions selected for EDMC to be analysed in SIM mode
were m/z 117, 89, and 72, and for naphthalene the following three ions were chosen: m/z 128, 102, and 64; m/z depicts mass-to-
charge ratio.
3
Figure 3. GC/MS analyses in SIM mode of the sample from a Vilsmeier reaction with POCl and DMF (1:1), at room temperature
for 4.5 h. Six chromatograms showing the response to the six analysed ions are presented. The concentration of DMCC in the
analysed sample was 4.4 ppm.
It has been shown that DMCC hydrolyses rapidly under
aqueous workup conditions. This provides an opportunity for
process chemists to design their processes in such a way that
the formed DMCC is decreased to acceptable levels during an
aqueous workup. Therefore, processes for manufacturing phar-
maceuticals based upon a Vilsmeier reaction can be developed
without risking the health of process operators or patients.
analytical purity. Compound 2 was synthesized according to
the published method.
17
Preparation of Ethyl N,N-Dimethylcarbamate (EDMC).
This compound was prepared in a similar fashion to the one
18
reported earlier by reacting DMCC (1 equiv.) with C H OH
2
5
(4 equiv) in the presence of pyridine (1 equiv) at 75 °C. On
completion, the excess of C H OH was removed by vacuum
2
5
distillation. The remaining product was partitioned between
diluted HCl and CH Cl . Pure EDMC was obtained by
Experimental Section
2
2
1
Materials. N,N-Dimethylcarbamoyl chloride (DMCC) and
naphthalene were purchased from Sigma-Aldrich with 99+%
purity. Other chemicals were obtained commercially with
distillation (139-142 °C/1 atm). Yield: 75%. Purity: 98%. H
NMR (400 MHz, CDCl ): δ 1.19 (t, J ) 7.3 Hz, 3H), 2.82 (s,
3
13
6H), 4.06 (q, J ) 7.3 Hz, 2H); C NMR (100 MHz, CDCl ):
3
δ 14.84, 35.95, 38.61, 60.84, 156.68. High resolution mass
(
(
17) Nordvall, G.; Ray, C.; Rein, T.; Sohn, D. PCT Int. Appl. WO
006107258 A1, 2006.
18) Valega, T. M. J. Org. Chem. 1966, 31 (4), 1150–1153.
+
5 2
spectrometry (HRMS): C H11NO , (M + H) calculated m/z
2
118.0854, found m/z 118.0846.
Vol. 13, No. 5, 2009 / Organic Process Research & Development
•
861

MS. Temperatures: MS quadrupole 150 °C, MS source 230
C.
Acquisition mode was set to Selected Ion Monitoring. Ions
°
monitored for EDMC from 6.0 to 10.5 min were m/z 72.0, 89.0,
and 117.0. Ion m/z 117.0 was used for quantifications, and m/z
72.0 and 89.0 were qualifiers. Ions monitored for naphthalene
(IS) from 10.5 to 20.2 min were m/z 64.0, 102.0, and 128.0.
For quantifications ion m/z 102.0 was chosen, and ions m/z 64.0
and 128.0 were selected as qualifiers.
Dwell time: 100 ms.
Vilsmeier Reactions. Neat Reactions. The reactions of
2 3
chlorinating agents SOCl , POCl and oxalyl chloride with DMF
in different molar ratios were carried out in approx 3 mL scale
in 7 mL glass test tubes.
Figure 4. Linearity of the analytical method for determination
of DMCC using standard addition analysis and GC/MS-SIM
technique. Data originate from a Vilsmeier reaction with POCl
3
Reactions in Dioxane. The chlorinating reagent (in excess)
was added to a solution of DMF in approx 5 mL dioxane. The
reactions were carried out in a 25 mL three-necked round-
bottom flask equipped with a thermometer and a magnetic stirrer
bar.
Reactions in the Presence of Base. The procedure was the
same as the one mentioned above but with one equivalent
pyridine added.
Reactions in Dichloromethane. The chlorinating reagent (in
excess), was added to a solution of DMF in dichloromethane
in three different 7 mL glass test tubes. The mixture was stirred
using a magnetic stirrer bar at the desired temperature.
and DMF (1:1), at room temperature for 4.5 h. The amount of
DMCC detected in this sample (0.564 g) was 2.48 µg, which
corresponded to a concentration of 4.4 ppm in the analysed
sample.
17
Conversion of Compound 2 to Compound 5. A mixture
of POCl (1.34 g, 8.74 mmol), DMF (0.28 g, 3.8 mmol), and
3
dioxane (8.26 g, 8.02 mL) was stirred for 30 min at room
temperature. Compound 2 (1.0 g, 3.3 mmol) was added to this
mixture. The whole mixture was then heated to 60 °C for 3 h
by which time the intermediate 3 was fully converted to the
intermediate 4 as confirmed by LC/MS. H O (2.0 mL) was then
2
added to the solution and heated at 70 °C for 30 min. After the
disappearance of the intermediate 4, the reaction mixture was
Reactions in the Presence of Compound 2. POCl
3
(1.34 g,
8.74 mmol) was added to a solution of DMF (0.28 g, 3.83
2
poured in cold H O (10 mL). Compound 5 was obtained by
mmol) in dioxane (8 mL) in a three-necked round-bottom flask.
After 30 min, compound 2 (1.0 g, 3.28 mmol) was added. The
whole mixture was then stirred with a magnetic stirrer bar at
filtration with a typical yield of 80% from compound 2.
1
Hydrolysis Studies by the H NMR Method. POCl
mL, 4.30 mmol) was added to D O (1.2 mL, 66.1 mmol), and
the mixture was stirred for 10 min. The hydrolyzed phosphorus
oxychloride in D O was added to a solution of N,N-dimethyl-
3
(0.4
2
65 °C for 3 h.
Sample Preparation for GC/MS-SIM Analysis. Four
2
samples (approximately 0.5 g) were taken simultaneously from
the reaction mixture and weighed. The samples were added to
four separate 5-mL test tubes with C H OH (1 mL) and then
2 5
heated to 70 °C. The test tubes were equipped with a magnetic
stirrer and septa penetrated with glass capillaries. The test tubes
were placed in a heating block. Acetonitrile (0.5 mL) was added
to one of the test tubes and standards of DMCC in acetonitrile
carbamoyl chloride (DMCC) (0.4 g, 3.72 mmol) in dioxane,
and the resulting mixture was stirred at 50 °C. Nine samples
were removed from the reaction mixture during 26 min and
1
analyzed quantitatively with H NMR for the determination of
the hydrolysis rate constant as well as the half-life time.
Hydrolysis Studies Using the GC/MS Method. A solution
3
of DMCC (31 mg, 0.3 mmol) and POCl (0.3 mL, 3.23 mmol)
in dioxane (25 mL) was heated to 80 °C. A sample (0.4 mL)
was removed from this hot solution and used as the zero
2
samples. H O (2 mL) was added to the hot solution, and the
whole mixture was stirred at this temperature for 60 min. The
samples were removed from the mixture for quantitative
determination of DMCC by the GC/MS method.
GC/MS-SIM Conditions. GC. The analyses were carried
out on an HP 6890 GC system combined with an HP 5973
mass selective detector and equipped with an Agilent DB-1701
column (14% cyanopropyl-phenyl-methylpolysiloxane), 30 m
(
0.5 mL) were added to the other three tubes. After one hour,
the mixture was cooled to room temperature. Saturated NaHCO
1-2 mL) and H O (1-2 mL) were added to each tube to reach
pH 8-9. Each sample was extracted with CH Cl (2 mL)
containing 50 ppm naphthalene (IS). The organic phase from
each of the four samples was dried over anhydrous Na SO
3
(
2
2
2
2
4
before being submitted to the GC/MS analysis.
Acknowledgment
We are grateful to Dr. Hefeng Pan in the Department of
analytic chemistry for her help with LC/MS analyses and the
HRMS measurements. We thank Dr. Hans-J u¨ rgen Federsel,
Prof. Tobias Rein, Malin Våger o¨ , and Dr. Ian Ashworth for
fruitful discussions and reviewing the manuscript.
×
0.25 mm, 0.25 µm film thickness.
Helium was used as carrier gas, at a constant flow rate of
1.0 mL/min.
Injection conditions: split injection; split ratio 50:1; injector
temperature 250 °C.
GC oven temperature program: 40 °C for 2 min, raised at
0 °C/min to 130 °C, then 40 °C/min to 300 °C, and kept at
00 °C for 5 min.
Received for review January 27, 2009.
OP900018F
1
3
8
62
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development