The solvent-free and catalyst-free conversion of an aziridine to an oxazolidinone using only carbon dioxide

Article abstract of DOI:10.1039/c1gc15850c

It has been found for the first time at room temperature that the reaction of an unactivated 2-alkyl or 2-aryl aziridine with carbon dioxide to generate the corresponding oxazolidinone does not need any form of catalysis or solvent to proceed in high yiel

Full text of DOI:10.1039/c1gc15850c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 3224
www.rsc.org/greenchem
PAPER
The solvent-free and catalyst-free conversion of an aziridine to an
oxazolidinone using only carbon dioxide†
Chau Phung, Rani M. Ulrich, Mostafa Ibrahim, Nathaniel T. G. Tighe, Deborah L. Lieberman and
Allan R. Pinhas*
Received 14th July 2011, Accepted 9th August 2011
DOI: 10.1039/c1gc15850c
It has been found for the first time at room temperature that the reaction of an unactivated 2-alkyl
or 2-aryl aziridine with carbon dioxide to generate the corresponding oxazolidinone does not need
any form of catalysis or solvent to proceed in high yield, especially when using high speed ball
milling.
According to recent reviews, in comparison to activated
aziridines, few papers have been published on the chemistry of
readily available unactivated N-alkyl aziridines. One reaction of
an N-alkyl aziridine is the insertion of carbon dioxide into a C–N
bond to generate an oxazolidinone, which is an important class
of compounds used as chiral auxiliaries, as metal ligands, and as
pharmaceuticals (specifically as antibiotics).1–3 Although carbon
dioxide is abundant, renewable, nonflammable, and inexpensive,
due to its stability,4,5 a difficult to synthesize catalyst, high
pressure (typically over 100 atm), and/or high temperature
(typically over 100 ◦C) generally are required for these insertion
reactions to proceed.6–9 In addition, most of these reactions only
work when the 2-position is substituted with a phenyl group;
they do not work when the 2-position is substituted with an
alkyl group. Put another way, in most reactions, compound 1
Scheme 1
will react to give 2, but compound 3 is stable.
While working on this project, it was reported that N-alkyl-2-
aryl aziridines (1) will generate the corresponding oxazolidinone
(2) with no catalyst.13 However, those reactions must be done at
120 ◦C and at 90 atm of CO2 pressure.
In this manuscript, it is shown that the reaction of an aziridine
with carbon dioxide to generate an oxazolidinone may be done at
low pressure and at room temperature in the absence of catalyst
and solvent.
For the past several years, we have been investigating the
reactions shown in Scheme 1 in which both types of unactivated
aziridines (1 and 3, R = alkyl) are converted to the corresponding
oxazolidinones at low pressure and temperature using a salt as
a catalyst.10–12 In a control experiment allowed to go for an
extended period of time (12 h vs. 4 h or less when a salt is used)
at room temperature, it was found that the reaction of compound
3 (Scheme 1, R = PhCH2 and alkyl = CH3) with CO2 (3 atm) in
THF using no catalyst gives a 7% yield of oxazolidinones 4 + 5.
This result is encouraging and led to an effort to find a method
to convert an aziridine plus CO2 into an oxazolidinone using no
catalyst.
Conversion performed in glassware with no solvent
As mentioned above, the reaction of compound 3 (R = PhCH2
and alkyl = CH3) with CO2 in THF using no catalyst gives a 7%
yield of oxazolidinones 4 + 5. Extending the reaction time or
increasing the pressure did not greatly improve the yield of this
reaction.
A solvent is generally used to dissolve the reactants, which
in turn facilitates the collisions that must occur in order to
transform the reactants into products. It was thought that a
solvent may not be necessary for this reaction because it involves
Department of Chemistry, University of Cincinnati, P. O. Box 210172,
Cincinnati, OH, 45221-0172, USA. E-mail: allan.pinhas@uc.edu;
Fax: 513-556-9239; Tel: 513-556-9255
† Electronic supplementary information (ESI) available: 1H and 13C
NMR spectral data for oxazolidinones 2, 4, 9, and 11. See DOI:
10.1039/c1gc15850c
3224 | Green Chem., 2011, 13, 3224–3229
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
a gas and a liquid. Thus, the aziridine plus CO2 reaction was
attempted in the absence of solvent.
Unlike the salt catalyzed reactions, the nitrogen of the
aziridine does not have to be substituted by an alkyl group.
Aziridine 6 (entry 3) gave the corresponding oxazolidinone 7 in
quantitative yield.
When aziridine 3 was stirred, with no catalyst or solvent, under
a CO2 pressure of 3 atm for 12 h, the yield of oxazolidinones 4
+ 5 increased from the 7% observed in THF to 37%. At 4 atm,
the yield increased to about 40%. In each case, compound 4 was
the major isomer and constituted 93% of the product mixture.
Due to the low yield, the reaction was allowed to go for a
longer period of time. At 4 atm, the yield increased from 40%
after 12 h, to 57% after 24 h, to 80% after 48 h, and to 89% after
8 days, with no effect on the regioselectivity.
Not surprisingly, given how much more quickly 2-phenyl
substituted aziridines react in comparison to 2-alkyl substituted
aziridines, when compound 1 (R = PhCH2 and aryl = Ph) was
subjected to CO2 gas at 4 atm and at room temperature, in the
absence of any solvent or catalyst, after 24 h, oxazolidinone 2
was generated in nearly quantitative yield.
These reactions are very sensitive to moisture. When a couple
drops of water were added to the reaction, the yield was about
half that observed under anhydrous conditions. The reason the
yield decreases with water may be due to the reaction of water
with carbon dioxide to generate carbonic acid, which in turn,
protonates the aziridine. This acid–base reaction would thus
greatly decrease the amount of nucleophilic aziridine present in
the reaction mixture.
To learn about the relative stereochemistry of the products,
two different 2,3-disubstituted aziridines were used. The reac-
tions of cis-2,3-dimethylaziridine 8 gave a nearly quantitative
yield of cis-stereoisomer 9 in the HSBM apparatus (entry 4).
Similarly, N-benzyl-7-azabicyclo[4.1.0]heptane 10 gave a nearly
quantitative yield of the corresponding cis oxazolidinone 11
(entry 5). Thus, these reactions proceed with net retention
of stereochemistry as was observed previously for the LiI
catalyzed reactions, in which a double inversion is known to take
place.10
Given the success of the 2,3-dialkyl aziridine reactions and
the ease of the 2-phenyl aziridine reaction, it is surprising that
2,3-diphenyl aziridine 12 gives back only starting material even
after shaking for an extended period of time (entry 6). This result
is the same as is observed when a salt is used as the catalyst.
One explanation is that the phenyl groups have enough electron
withdrawing ability that the nitrogen does not have the required
electron density to attack the CO2.
Consistent with this idea, a mild electron withdrawing sub-
stituent at the 2-position, such as a meta-chloro phenyl group,
generates only a trace amount of a compound with the correct
molecular weight for the product, even when shaken for an
extended period of time (entry 7).
When the nitrogen is substituted by a phenyl, regardless
if the 2-position is substituted by a phenyl (entry 8) or by
an alkyl group (entry 9), the starting aziridine is recovered
unchanged after an extended period of shaking. Even when
the nitrogen is substituted with a para-methoxy phenyl (entry
10), none of the corresponding oxazolidinone is obtained.
These results are similar to the results observed for the
high temperature-no catalyst reactions13 and the salt catalyzed
reactions.10
It is believed that these observations are due to the greatly de-
creased basicity and nucleophilicity of N-phenyl in comparison
to N-alkyl aziridine. Specifically, the pKa of an alkyl azirdine
is 8.3 and the pKa of para-methoxy aniline is 5.2.16,17 Thus, the
N-phenyl aziridine is much less basic, and thus less nucleophilic,
than an N-alkyl aziridine.
As was observed previously for the salt catalyzed reactions,
N-acyl aziridine 17 simply isomerizes to oxazoline 18 (entry 11),
regardless of the amount of carbon dioxide present.
Due to the nature of the shaking apparatus, 17 h is a
convenient period of time for each reaction. However, we were
interested in determining if a shorter reaction time was possible.
In other words, we wanted to determine the shortest time that
would still give a 95+% yield of the oxazolidinone. For 2-phenyl
aziridine 1, this time was found to be 6 h, and for 2-alkyl aziridine
3, this time was found to be 12 h.
High speed ball milling
In order to greatly decrease the time of the reactions, high speed
ball milling (HSBM) was investigated. Mechanical energy from
HSBM has been shown to give high yields and increased rates of
chemical reactions, especially those reactions done in the absence
of any solvent. The HSBM apparatus uses a small steel ball in
a steel vial with high speed shaking to achieve small particle
sizes through milling. Subsequent collisions of the milling ball
with the sides of the reaction vessel provide mechanical energy
to overcome the activation barrier of the reaction.14,15
These reactions are accomplished by using a 3.5 mL steel vial
with a 3.2 mm diameter milling ball. To the vial was added
approximately 150 mg of the aziridine and 3 g of dry ice as the
CO2 source. The vial was sealed and shaken in a paint shaker for
the appropriate length of time. Table 1 summarizes our HSBM
reactions.
Given the ease of the reaction in the glass flask, the first
reaction attempted in the HSBM was 2-phenyl substituted
compound 1 with dry ice for 17 h (entry 1). This conversion
generated compound 2 in nearly quantitative yield with only
one regioisomer.
Therefore, the reaction of 2-alkyl aziridine 3 (R = PhCH2 and
alkyl = CH3) with dry ice was subjected to the same HSBM
reaction conditions (entry 2). This conversion was completed
in 17 h in quantitative yield with over 93% of the product as
isomer 4. As a set of control experiments, when the HSBM
reaction conditions were used for the reaction of compound 3,
but with no shaking or no milling ball, the yields obtained were
similar to those reported above in the glass apparatus. In other
words, the shaking/mechanical energy decreases the time of the
reaction from over a week to over night.
Mechanism
Based on the stereochemical results, the nucleophilicity of
the nitrogen, and the mechanism of the salt catalyzed re-
action, there are two possible mechanisms for these trans-
formations. The initial step of each is the nucleophilic ni-
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 3224–3229 | 3225
©

View Article Online
Table 1 High Speed Ball Milling Reactions
Major product
oxazolidinone
Reaction
time (h)
Regioselectivity %
major product
Total % yield of
all products
Entry
1
Starting aziridine
17
100
quantitative
2
17
93
quantitative
3
17
100
quantitative
4
17
100
quantitative
5
17
100
quantitative
6
43
0
7
43
trace
3226 | Green Chem., 2011, 13, 3224–3229
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 1 (Contd.)
Major product
oxazolidinone
Reaction
time (h)
Regioselectivity %
major product
Total % yield of
all products
Entry
8
Starting aziridine
43
0
9
43
0
10
43
0
11
17
100
quantitative
trogen of the aziridine attacking CO2 to give compound
19. The mechanisms then differ because a different nucle-
ophile is used for subsequent steps. One mechanism is a
variation of the mechanism proposed for the high temperature-
no catalyst reaction,13 in which the nucleophile is a carboxylate,
and the other is a variation of the mechanism proposed in a
recent study using an amino acid as the catalyst for the aziridine
to oxazolidinone conversion,9 in which the nucleophile is an
aziridine. These proposed mechanisms are shown in Scheme 2
for prototypical compound 3. (The carboxylate in compound
19 cannot undergo a unimolecular reaction and attack the 3-
position because (1) this would be a 5-endo-tet reaction, which
is disfavored by Baldwin’s Rules, and (2) this mechanism would
give net inversion, which is inconsistent with entries 4 and 5 in
Table 1.)
Conclusion
It has been shown that the reaction of an unactivated aziridine
with CO2 does not need a catalyst, a solvent, high temperature,
or high pressure if mechanical energy is applied.
General experimental procedure
Into a 48 mL thick walled glass flask, with a stir bar, was added
aziridine 3 (0.18 g, 1.2 mmol) and 2 mL of THF, or just the
aziridine. The reaction mixture then was subjected to carbon
dioxide gas at room temperature for the appropriate length of
time. The pressure was monitored by a gauge and adjusted to 3
or 4 atm by bleeding off excess pressure at the beginning of the
reaction.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 3224–3229 | 3227
©

View Article Online
Scheme 2
After the reaction was completed, the THF was evaporated.
For those reactions in which no solvent was used, at the end
of the reaction, ether was added to the reaction mixture so it
could be transferred to a round bottom flask. The ether was
then evaporated.
As done previously, GC-MS and NMR spectroscopy were
used to analyze the reaction mixture. Spectroscopic data may
be found in references 10 and 18 (See ESI†). The product was
weighed to determine the yield. Each reaction was run at least
three times and the reported yields are the averages of these three
runs.
steel cap containing a Teflon gasket. For the non-shaking control
experiments, the vial remained stationary at room temperature
for the appropriate time period. For the shaking experiments,
the vial was shaken at 18 Hz, using a Spex 8000 M mixer, for
999 min (approx. 17 h). The reaction mixture was transferred
and analyzed spectroscopically as above. Spectroscopic data for
starting materials and for products may be found in references
10, 18, and in references cited therein (See ESI†). The product
was weighed to determine the yield. Each reaction was run at
least three times and the reported yields are the averages of these
three runs.
HSBM experimental procedure
Acknowledgements
Between 0.10 and 0.20 g of the aziridine were placed into a
stainless steel reaction vessel (total volume 3.5 mL) with a 3.2
mm diameter milling ball and an excess of dry ice (3.0 g, 68 mmol
CO2). (Because there is such a large excess of carbon dioxide,
we have found that it does not matter exactly how much of
the aziridine is used.) The vial was tightly sealed with a stainless
The authors thank Will Shearouse and Dan Waddell for their
help with the HSBM experiments. ARP thanks the faculty and
staff of the Department of Pharmacology and Toxicology and
the Department of Chemistry of the University of Louisville
for their hospitality during a sabbatical leave when much of this
manuscript was written.
3228 | Green Chem., 2011, 13, 3224–3229
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
8 Y. Du, Y. Wu, A.-H. Liu and L.-N. He, J. Org. Chem., 2008, 73,
4709.
9 H.-F. Jiang, J.-W. Ye, C.-R. Qi and L.-B. Huang, Tetrahedron Lett.,
2010, 51, 928.
References
1 G. Zappia, E. Gacs-Baitz, G. D. Monache, D. Misiti, L. Nevola and
B. Botta, Curr. Org. Synth., 2007, 81.
2 I. D. G. Watson, L. Yu and A. K. Yudin, Acc. Chem. Res., 2006, 39,
194.
10 M. T. Hancock and A. R. Pinhas, Tetrahedron Lett., 2003, 44, 5457.
11 C. Phung and A. R. Pinhas, Tetrahedron Lett., 2010, 51, 4552.
12 After reference 10 was accepted for publication but before it was
printed, a very similar manuscript was accepted: A. Sudo, Y.
Morioka, E. Koizumi, F. Sanda and T. Endo, Tetrahedron Lett.,
2003, 44, 7889.
3 X. E. Hu, Tetrahedron, 2004, 60, 2701.
4 T. Sakakura, J.-C. Choi and H. Yasuda, Chem. Rev., 2007, 107, 2365.
5 H. Arakawa, M. Aresta, J. N. Armor, M. A. Barteau, E. J. Beckman,
A. T. Bell, J. E. Bercaw, C. Creutz, E. Dinjus, D. A. Dixon, K. Domen,
D. L. DuBois, J. Eckert, E. Fujita, D. H. Gibson, W. A. Goddard, D.
W. Goodman, J. Keller, G. J. Kubas, H. H. Kung, J. E. Lyons, L. E.
Manzer, T. J. Marks, K. Morokuma, K. M. Nicholas, R. Periana, L.
Que, J. Rostrup-Nielson, W. M. H. Sachtler, L. D. Schmidt, A. Sen,
G. A. Somorjai, P. C. Stair, B. R. Stults and W. Tumas, Chem. Rev.,
2001, 101, 953.
13 X.-Y. Dou, L.-N. He, Z.-Z. Yang and J.-L. Wang, Synlett, 2010,
2159.
14 W. C. Shearouse, D. C. Waddell and J. Mack, Current Opinion in
Drug Discovery and Development, 2009, 12, 772.
15 M. A. P. Martins, C. P. Frizzo, D. N. Moreira, L. Buriol and P.
Machado, Chem. Rev., 2009, 109, 4140.
16 C. E. O’Rourke, L. B. Clapp and J. O. Edwards, J. Am. Chem. Soc.,
1956, 78, 2159.
6 Y. Wu, L.-N. He, Y. Du, J.-Q. Wang, C.-X. Miao and W. Li,
Tetrahedron, 2009, 65, 6204.
7 L. He, Y. Du, C. Miao, J. Wang, X. Dou; and Y. Wu, Front. Chem.
Eng. China, 2009, 3, 224.
17 See for example, http://yfaat.ch.huji.ac.il/jaguar-help/mank.html.
18 M. T. Hancock and A. R. Pinhas, Synthesis, 2004, 2347.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 3224–3229 | 3229
©