Ethyl diazoacetate synthesis in flow

Article abstract of DOI:10.3762/bjoc.9.211

Ethyl diazoacetate is a versatile compound in organic chemistry and frequently used on lab scale. Its highly explosive nature, however, severely limits its use in industrial processes. The in-line coupling of microreactor synthesis and separation technology enables the synthesis of this compound in an inherently safe manner, thereby making it available on demand in sufficient quantities. Ethyl diazoacetate was prepared in a biphasic mixture comprising an aqueous solution of glycine ethyl ester, sodium nitrite and dichloromethane. Optimization of the reaction was focused on decreasing the residence time with the smallest amount of sodium nitrite possible. With these boundary conditions, a production yield of 20 g EDA day^-1 was achieved using a microreactor with an internal volume of 100 μL. Straightforward scale-up or scale-out of microreactor technology renders this method viable for industrial application.

Full text of DOI:10.3762/bjoc.9.211

Ethyl diazoacetate synthesis in flow
Mariëlle M. E. Delville, Jan C. M. van Hest and Floris P. J. T. Rutjes*
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2013, 9, 1813–1818.
Radboud University Nijmegen, Institute for Molecules and Materials,
Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
doi:10.3762/bjoc.9.211
Received: 28 June 2013
Email:
Accepted: 15 August 2013
Published: 05 September 2013
Floris P. J. T. Rutjes* - f.rutjes@science.ru.nl
*
Corresponding author
This article is part of the Thematic Series "Chemistry in flow systems III".
Guest Editor: A. Kirschning
Keywords:
diazo compounds; diazotization; ethyl diazoacetate (EDA); flow
chemistry; microreactor technology
© 2013 Delville et al; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
Ethyl diazoacetate is a versatile compound in organic chemistry and frequently used on lab scale. Its highly explosive nature,
however, severely limits its use in industrial processes. The in-line coupling of microreactor synthesis and separation technology
enables the synthesis of this compound in an inherently safe manner, thereby making it available on demand in sufficient quantities.
Ethyl diazoacetate was prepared in a biphasic mixture comprising an aqueous solution of glycine ethyl ester, sodium nitrite and
dichloromethane. Optimization of the reaction was focused on decreasing the residence time with the smallest amount of sodium
nitrite possible. With these boundary conditions, a production yield of 20 g EDA day−1 was achieved using a microreactor with an
internal volume of 100 μL. Straightforward scale-up or scale-out of microreactor technology renders this method viable for indus-
trial application.
Introduction
Diazo compounds are frequently used versatile building blocks chemistry, including the inherently safe conducting of reactions
in organic chemistry [1,2]. From this class of compounds due to the small reactor dimensions, efficient heat transport and
diazomethane and ethyl diazoacetate (1, EDA) are arguably the excellent control over the reaction conditions [6-8]. While the
synthetically most useful ones. Due to the potentially explosive synthesis of diazomethane has been extensively explored in
nature of diazomethane and EDA [3-5], however, synthetic batch [9] and in continuous-flow reactors [10,11], EDA is
routes that involve large scale batchwise handling of such diazo synthesized via different routes in batch [12,13], but relatively
compounds is generally avoided in industrial processes. With little is known about continuous-flow approaches [14]. Consid-
the advent of continuous processing over the past decade, new ering the importance of EDA in a wide variety of reactions e.g.
approaches have appeared to conceptually change the way cyclopropanation, X–H insertion, cycloaddition and ylide for-
chemical synthesis is performed. In particular continuous-flow mation [13,15], and more recently, in the synthesis of valuable
microreactor technology offers multiple advantages over batch compound classes such as β-keto esters [16] and β-hydroxy-α-
1813

Beilstein J. Org. Chem. 2013, 9, 1813–1818.
diazocarbonyl compounds [17], we aimed to develop an inher- reactions. Straightforward scale-up or scale-out of microreactor
ently safe continuous-flow EDA process using microreactor and technology renders this method viable for industrial application.
separation technology.
Results and Discussion
Ethyl diazoacetate (1) can be synthesized in flow via different Flow synthesis
pathways. Bartrum et al. [18] published a flow synthesis of Ethyl diazoacetate (1) was synthesized from glycine ethyl ester
numerous diazo esters starting from the corresponding arylsul- (2) and sodium nitrite in a biphasic system of dichloromethane
fonylhydrazones, where the diazo moiety was installed through and an aqueous sodium acetate buffer. Dichloromethane was
elimination of the sulfone substituent. Additionally, Ley et al. chosen as the organic phase to dissolve the water insoluble
[
19] recently prepared a range of α-hydroxy acids in flow EDA, because of its low water uptake and low boiling point and
starting from the corresponding amino acids, involving diazoti- its compatibility with potential follow-up reactions. In principle,
zation of the amine to the diazonium salt in a biphasic system. however, any other organic solvent immiscible with water could
Inspired by Ley’s approach, which is significantly more atom be used. The pH of the buffer was set to 3.5 which had been
efficient than the sulfonylhydrazone pathway, we chose to identified by Clark et al. as the optimal pH for the reaction [12].
synthesize EDA (1) from glycine ethyl ester (2) using readily A schematic representation of the initial microreactor set-up is
available sodium nitrite [20] (Scheme 1). Although the diazoti- shown in Figure 1. The box with the dotted line indicates the
zation step itself resembles the first step of Ley’s hydroxy acid single-glass microreactor containing two mixing units M of the
synthesis, we specifically aimed to produce and isolate the folding flow type [21]. The reactor temperature was controlled
diazo product, which from there can be used for subsequent by a Peltier element and sensed by a Pt1000 temperature sensor.
reactions.
At the outlet of the microreactor, a back-pressure regulator
BPR, 40 psi) was attached to guarantee a liquid phase even
(
above boiling temperatures of the solvents. To ensure well-
defined reaction times during optimization experiments, neat
N,N-diisopropylethylamine (DIPEA) was added via syringe 4 to
efficiently quench the reaction. The collected product (60 μmol)
was analyzed by HPLC to establish the conversion of the reac-
tion.
Scheme 1: Synthesis of ethyl diazoacetate (1).
We intended to optimize the process focusing on decreasing the Univariate optimization
residence time in order to reduce solvent use and gain in Determination of the optimal conditions for the reaction started
throughput. Reaction temperature was considered less of an off with investigating the important reaction parameters via a
issue since in an industrial setting energy can generally be effi- univariate optimization. Based on knowledge obtained from
ciently regenerated. In-line phase separation was thought to EDA synthesis in batch [12] and other flow reactions [22,23],
greatly enhance the usefulness of the EDA flow synthesis. residence time, temperature and NaNO2 stoichiometry were
Therefore, the outlet of the microreactor was directly connected chosen as relevant parameters. Temperature was expected to
to membrane-based phase separator to obtain EDA in the have a large influence on the rate of the reaction. Shortening the
organic phase, which in principle can then be immediately used residence time to a minimum would minimize the risk of side
for either batch [13,15] or continuous-flow [16,17] follow-up reactions and reduce costs, and the reaction should be
Figure 1: Schematic representation of the microreactor setup.
1814

Beilstein J. Org. Chem. 2013, 9, 1813–1818.
Figure 2: Univariate optimization using 30 s, 15 °C and 1.5 equiv NaNO2 as standard.
performed with the smallest amount of NaNO2 possible. The These plots show a rather broad optimum for the conversion of
results of the univariate optimization are shown in Figure 2. glycine ethyl ester (2) into EDA (1). The decrease in the upper
EDA synthesis was shown to be fast, since within 200 seconds left corner of the second contour plot can be explained by the
complete conversion was obtained at 15 °C. Additionally, the high uncertainty of the model at the edge of the plots. As was
temperature shows a steep increase between 0–30 °C, indi- expected, temperature has a large influence on the reaction rate.
cating a large influence of both parameters on the reaction rate. The conversion into EDA shows a steep increase with
The amount of NaNO2 shows only a rather small influence. increasing temperature. High temperatures and increasing
Based on these univariate optimizations the experimental ranges amounts of NaNO2 decrease the residence time to a minimum
of the three parameters were determined to investigate the inter- of 20 seconds while still obtaining complete conversion. Not
relationships via a multivariate optimization.
surprisingly, the minimal amount of NaNO2 required is 1
equivalent. We aimed to reach complete conversion into EDA
(1) maintaining a short residence time with a minimum amount
Multivariate optimization
An experimental design based on a D-optimal algorithm was of sodium nitrite, possibly using higher temperatures. Based on
created from the aforementioned three parameters within their these boundary conditions, the optimal parameter settings were
respective ranges, namely 5–120 s, 0–60 °C and 0.7–1.5 equiv fixed at 20 seconds residence time, a temperature of 50 °C using
of NaNO2. Using MATLAB (MathWorks, R2007a), fifty data 1.5 equivalents of NaNO2. A triple-experiment was performed
points were selected of which the corresponding experiments to prove that this set of optimal parameters indeed provided
were performed in random order. The resulting HPLC yields complete conversion into EDA. The experiment was performed
were normalized and fitted to a third order polynomial model. in alternation with two other sets of parameters to rule out
In-house-developed FlowFit software [24] was used to calcu- potential memory effects. HPLC yields of 95, 96 and 95% for
late the best possible model fit. The results are visualized in the triple-experiment demonstrate the high reproducibility of
2
D-contour plots (Figure 3).
the system.
Figure 3: 2D-Contour plots of the multivariate optimization.
1815

Beilstein J. Org. Chem. 2013, 9, 1813–1818.
Figure 4: Phase separation using a Flow-Liquid–Liquid-Extraction module (FLLEX) directly coupled to the microreactor.
FLLEX module
equivalents of NaNO2. Repeating the EDA synthesis in flow
Having established a microreactor protocol for the continuous- employing the optimal reaction parameters showed complete
flow synthesis of EDA, the next issue was to separate the pro- conversion and high reproducibility of the results. Additionally,
duct from the biphasic system in which it was collected. In we successfully combined a plug-and-play microreactor setup
order to increase safety and decrease the hold-up of EDA, the with a commercially available membrane-based phase sep-
phase separation ideally had to be performed in flow as well. aration module to perform a direct in-line extraction of the pro-
Therefore, a Flow-Liquid–Liquid-Extraction module (FLLEX) duct. Even in our small set-up (internal volume 100 μL), we
[
25] was connected to the system [26,27]. The module utilizes a were able to generate approximately 20 g of pure EDA per day
hydrophobic Teflon membrane and two back-pressure regula- (11 wt % solution in CH2Cl2).
tors (BPRs) to create a pressure difference, which causes the
Experimental
organic layer, in this case dichloromethane, to pass through the
membrane resulting in phase separation. A schematic represen- Physical and spectroscopic measurements
tation of the whole setup is shown in Figure 4.
NMR spectra were acquired at ambient temperature with a
Bruker DMX 300 MHz spectrometer. 1H NMR spectra were
As the conversion into EDA was quantitative, quenching with referenced to TMS or to the residual solvent peak. HPLC
DIPEA was no longer required. Between the microreactor and analysis was performed using an Agilent 1120 Compact LC,
the FLLEX module some additional tubing was used to ensure C-18 column, 10% acetonitrile in MilliQ, 254 nm. Pyridine
complete partitioning of the compounds over the two phases. (internal standard) has a retention time of 1.75 min, EDA of
The back pressure of the FLLEX was set to 40 psi, similar to 9.67 min.
the BPR used previously, and a pressure difference of 0.14 bar.
Direct full separation of phases resulted in a clean organic phase Chip dimensions
containing 409 mg EDA (11 wt % solution in CH2Cl2, after 30 Three different microchips were used during the experiments.
min of collection) while all salts remain in the aqueous phase.
This corresponds roughly to an EDA production of 20 g day−1
and a space time yield of 100 kg day−1 dm−3 as compared to a
reported industrial scale batch process yielding EDA in 48 g
day−1 dm−3 [12].
1. Single borosilicate glass quench microreactor with an
internal volume of 92 μL, a channel width of 600 μm and
a channel depth of 500 μm.
2. Single borosilicate glass microreactor with an internal
volume of 100 μL, a channel width of 600 μm and a
channel depth of 500 μm.
Conclusion
EDA can be safely synthesized utilizing microreactor and sep-
aration technology starting from cheap and readily available
starting materials. Optimization of the reaction was aimed at
reaching complete conversion into EDA within a minimized
3. Single borosilicate glass quench microreactor with an
internal volume of 1 μL, a channel width of 120 μm and
a channel depth of 50 μm.
residence time using the smallest required amount of sodium Univariate optimization
nitrite, possibly applying higher temperatures. The optimal Solution A: Glycine ethyl ester hydrochloride (40 mmol, 5.6 g)
reaction conditions identified based on these criteria were a dissolved in 20 mL buffer 1. Solution B: CH2Cl2. Solution C:
residence time of 20 seconds, a temperature of 50 °C and 1.5 NaNO2 (60 mmol, 4.1 g) dissolved in 30 mL degassed MilliQ.
1816

Beilstein J. Org. Chem. 2013, 9, 1813–1818.
Table 1: Conditions of the univariate experiments using 30 s, 15 °C and 1.5 equiv NaNO2 as standard.
Time (s)
15
0
30
5
60
10
1.2
120
15
180
20
300
30
600
40
2
900
50
Temperature (°C)
Amount of NaNO2
60
90
0.7
1
1.4
1.5
1.8
Solution Q: Neat DIPEA. Buffer 1: Sodium acetate trihydrate Multivariate optimization
132 mmol, 18.0 g) and pyridine (7.5 mL, internal standard) Solution A: Glycine ethyl ester hydrochloride (40 mmol, 5.6 g)
dissolved in 70 mL MilliQ. Concentrated hydrochloric acid dissolved in 20 mL buffer 1. Solution B: CH2Cl2. Solution C:
37%, 12 M) was added until a pH of 3.5 was reached (17 mL), NaNO2 (60 mmol, 4.1 g) dissolved in 30 mL degassed MilliQ.
resulting in a buffer with a total volume of 105 mL. Solution Q: Neat DIPEA.
(
(
The flow rates and temperatures were set based on predeter- The flow rates and temperatures were set based on predeter-
mined conditions of residence times and temperatures (Table 1). mined conditions of residence times and temperatures (Table 2).
Experiments were performed in a glass microreactor with an Experiments with a residence time of 5 s were performed in a
internal volume of 92 μL. Solution Q was set at a flow rate 1/3 glass microreactor with an internal volume of 1 μL. For longer
of the flow rate of solution A. Each experiment had a collection residence times, a microreactor with an internal volume of
time equal to 30 μL of solution A. The product was collected in 92 μL was used. Solution Q was set at a flow rate 1/3 of the
1
mL of acetonitrile and analyzed by HPLC. Results are visual- flow rate of solution A. Each experiment had a collection time
ized in Figure 2. equal to 30 μL of solution A. The product was collected in
Table 2: Experiments for the multivariate optimization deduced from a D-optimal experimental design algorithm.
Exp#
Molar ratio Residence time (s) Temperature (°C)
Exp#
Molar ratio Residence time (s) Temperature (°C)
1
2
3
4
5
6
7
8
9
1.5
1.5
1.5
0.7
1.5
1.5
1.5
0.7
1.5
1.1
0.7
1.5
1.1
0.7
0.7
0.7
1.5
1.5
1.5
0.7
1.5
1.5
1.1
1.1
1.5
5
0
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
1.1
1.5
0.7
1.1
1.1
1.1
1.5
1.5
0.7
0.7
1.1
1.5
0.7
1.1
1.1
1.1
0.7
1.1
0.7
1.1
0.7
0.7
0.7
0.7
0.7
120
120
5
0
120
45
5
60
0
60
0
60
60
40
0
15
120
45
5
60
60
60
20
20
0
45
120
45
45
15
5
0
120
15
5
20
20
20
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
20
60
60
60
60
40
40
40
0
45
120
15
5
5
15
45
120
120
5
0
60
40
60
0
120
45
120
5
5
60
40
60
0
5
45
120
5
120
120
15
5
60
0
40
0
5
60
0
45
120
5
120
120
120
0
20
40
20
0
1817

Beilstein J. Org. Chem. 2013, 9, 1813–1818.
1
1.Struempel, M.; Ondruschka, B.; Duate, R.; Stark, A. Green Chem.
1
mL of acetonitrile and analyzed by HPLC. Results are visual-
2008, 10, 41–43. doi:10.1039/b710554a
ized in Figure 3 as 2D-contour plots.
1
2.Clark, J. D.; Heise, J. D.; Shah, A. S.; Peterson, J. C.; Chou, S. K.;
Levine, J.; Karakas, A. M.; Ma, Y.; Ng, K.-Y.; Patelis, L.; Springer, J. R.;
Stano, D. R.; Wettach, R. H.; Dutra, G. A. Org. Process Res. Dev.
2004, 8, 176–185. doi:10.1021/op034123q
FLLEX experiment
Solution A: Glycine ethyl ester hydrochloride (10 mmol, 1.4 g)
dissolved in 5 mL buffer 2. Solution B: CH2Cl2. Solution C: 13.Maas, G. Angew. Chem., Int. Ed. 2009, 48, 8186–8195.
doi:10.1002/anie.200902785
NaNO2 (15 mmol, 1.0 g) dissolved in 5 mL degassed MilliQ.
1
4.Thathagar, M.; Poechlauer, P.; Braune, S. Process for the production
of cyclopropane derivatives. WO2010055106, May 20, 2010.
5.Davies, H. M. L.; Beckwith, R. E. J. Chem. Rev. 2003, 103,
Buffer 2: Sodium acetate trihydrate (100 mmol 13.6 g)
dissolved in 80 mL MilliQ. Concentrated hydrochloric acid
1
(
37%, 12 M) was added until a pH of 3.5 was reached (7 mL).
Additional MilliQ was added to obtain a total volume of
00 mL of buffer.
2
861–2903. doi:10.1021/cr0200217
16.Bartum, H. E.; Blakemore, D. C.; Moody, C. J.; Hayes, C. J.
J. Org. Chem. 2010, 75, 8674–8676. doi:10.1021/jo101783m
1
1
7.Krishna, P. R.; Prapurna, Y. L.; Alivelu, M. Eur. J. Org. Chem. 2011,
089–5095. doi:10.1002/ejoc.201100496
5
Solution A (86.25 μL/min) was combined in a stainless steel
T-splitter with solution B (172.5 μL/min). The biphasic mixture
1
8.Bartrum, H. E.; Blakemore, D. C.; Moody, C. J.; Hayes, C. J.
Chem.–Eur. J. 2011, 17, 9586–9589. doi:10.1002/chem.201101590
immediately entered the glass microreactor (internal volume: 19.Hu, D. X.; O’Brien, M.; Ley, S. V. Org. Lett. 2012, 14, 4246–4249.
1
00 µL) where it was mixed with solution C (86.25 μL/min).
doi:10.1021/ol301930h
2
0.Monbaliu, J. C.; Jorda, J.; Chevalier, B.; Stevens, C. V.; Morvan, B.
Chim. Oggi 2011, 29, 50–52.
The reaction was performed at 50 °C. After the reaction, the
mixture was passed through 15 μL of FEP-tubing (ID =
See for the use of sodium nitrite in flow chemistry for example.
1.MacInnes, J. M.; Vikhansky, A.; Allen, R. K. W. Chem. Eng. Sci. 2007,
2
54 μm) before entering the FLLEX module where phases were
separated (40 psi, Δp = 0.14 bar). The set-up was stabilized for
min before collecting for 30 min. EDA was obtained as a
2
62, 2718–2727. doi:10.1016/j.ces.2007.02.014
2
22.Delville, M. M. E.; Nieuwland, P. J.; Janssen, P.; Koch, K.;
van Hest, J. C. M.; Rutjes, F. P. J. T. Chem. Eng. J. 2011, 167,
solution in CH2Cl2 (1.52 g). According to 1H NMR analysis,
clean EDA was obtained. Based on the residual solvent peak in
the 1H NMR spectrum it was calculated to be a 27 wt % solu-
tion of EDA in CH2Cl2 meaning 409 mg of pure EDA.
556–559. doi:10.1016/j.cej.2010.08.087
2
3.Delville, M. M. E.; van Gool, J. J. F.; van Wijk, I. M.; van Hest, J. C. M.;
Rutjes, F. P. J. T. J. Flow Chem. 2012, 4, 124–128.
doi:10.1556/JFC-D-12-00008
2
4.For more information about FlowFit software see:
http://www.futurechemistry.com
Acknowledgements
2
5.http://www.syrris.com.
We thank Dr. Pieter Nieuwland (FutureChemistry) for fruitful
discussions. The NWO-ACTS Process-on-a-Chip (PoaC)
program is kindly acknowledged for financial support.
2
6.Tricotet, T.; O’Shea, D. F. Chem.–Eur. J. 2010, 16, 6678–6686.
doi:10.1002/chem.200903284
27.Varas, A. C.; Noël, T.; Wang, Q.; Hessel, V. ChemSusChem 2012, 5,
1703–1707. doi:10.1002/cssc.201200323
References
1
.
Regitz, M. Synthesis 1972, 351–373. doi:10.1055/s-1972-21883
Zhang, Z.; Wang, J. Tetrahedron 2008, 64, 6577–6605.
doi:10.1016/j.tet.2008.04.074
2.
License and Terms
3
.
.
Clark, J. D.; Shah, A. S.; Peterson, J. C.; Patelis, L.; Kersten, R. J. A.;
Heemskerk, A. H.; Grogan, M.; Camden, S. Thermochim. Acta 2002,
This is an Open Access article under the terms of the
Creative Commons Attribution License
386, 65–72. doi:10.1016/S0040-6031(01)00760-2
4
Clark, J. D.; Shah, A. S.; Peterson, J. C.; Patelis, L.; Kersten, R. J. A.;
Heemskerk, A. H. Thermochim. Acta 2002, 386, 73–79.
doi:10.1016/S0040-6031(01)00761-4
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
5
6
7
8
9
1
.
.
.
.
.
Hosmane, R. S.; Liebman, J. F. Struct. Chem. 2002, 13, 501–503.
doi:10.1023/A:1020573723147
Wegner, J.; Ceylan, S.; Kirschning, A. Adv. Synth. Catal. 2012, 354,
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
17–57. doi:10.1002/adsc.201100584
Wiles, C.; Watts, P. Chem. Commun. 2011, 47, 6512–6535.
doi:10.1039/c1cc00089f
(http://www.beilstein-journals.org/bjoc)
Wirth, T., Ed. Microreactors in Organic Synthesis and Catalysis;
Wiley-VCH: Weinheim, Germany, 2008. doi:10.1002/9783527622856
Cohen, J. D. J. Chromatogr. 1984, 303, 193–196.
doi:10.1016/S0021-9673(01)96061-3
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.9.211
0.Ferstl, W. F.; Schwarzer, S.; Loebbecke, S. L. Chem. Ing. Tech. 2004,
6, 1326–1327. doi:10.1002/cite.200490233
7
1818