C-N and N-H Bond Metathesis Reactions Mediated by Carbon Dioxide

Article abstract of DOI:10.1002/cssc.201500318

Herein, we report CO₂-mediated metathesis reactions between amines and DMF to synthesize formamides. More than 20 amines, including primary, secondary, aromatic, and heterocyclic amines, diamines, and amino acids, are converted to the corresponding formamides with good-to-excellent conversions and selectivities under mild conditions. This strategy employs CO₂ as a mediator to activate the amine under metal-free conditions. The experimental data and in situ NMR and attenuated total reflectance IR spectroscopy measurements support the formation of the N-carbamic acid as an intermediate through the weak acid-base interaction between CO₂ and the amine. The metathesis reaction is driven by the formation of a stable carbamate, and a reaction mechanism is proposed.

Full text of DOI:10.1002/cssc.201500318

DOI: 10.1002/cssc.201500318
Full Papers
CÀN and NÀH Bond Metathesis Reactions Mediated by
Carbon Dioxide
Yehong Wang,[a] Jian Zhang,[a] Jing Liu,[a] Chaofeng Zhang,[a, b] Zhixin Zhang,[a] Jie Xu,*[a]
Shutao Xu,[a] Fangjun Wang,[a] and Feng Wang*[a]
Herein, we report CO2-mediated metathesis reactions between
amines and DMF to synthesize formamides. More than
20 amines, including primary, secondary, aromatic, and hetero-
cyclic amines, diamines, and amino acids, are converted to the
corresponding formamides with good-to-excellent conversions
and selectivities under mild conditions. This strategy employs
CO2 as a mediator to activate the amine under metal-free con-
ditions. The experimental data and in situ NMR and attenuated
total reflectance IR spectroscopy measurements support the
formation of the N-carbamic acid as an intermediate through
the weak acid–base interaction between CO2 and the amine.
The metathesis reaction is driven by the formation of a stable
carbamate, and a reaction mechanism is proposed.
Introduction
The selective N-formylation of amines to formamides is a signif-
icant reaction in organic, medicinal, and biological chemistry.[1]
Formic acid[2] or in situ-generated formic acid,[3] DMF,[4] alde-
hydes,[5] and their derivatives have been used as formylation
reagents. However, the majority of examples require high tem-
peratures, water-sensitive reagents, or complex workups and,
thus, are unsuitable for use with fine chemicals, pharmaceuti-
cals, and molecules bearing diverse functional groups, such as
proteins and thiophenes.
different to previous ones that required activated formylation
reagents, metal-containing catalysts, and strong additives. In
addition, the reaction can be performed in air (ꢀ400 ppm CO2
in the atmosphere) with comparable results. Moreover, it has
broad substrate compatibility.
Reactions induced by weak acid–base interactions play a key
role in living systems and in the chemical industry.[6] For exam-
ple, CO2 generated from cellular respiration is expired in part
through the reversible formation of a carbamate between CO2
and the amino groups of hemoglobin.[7] Furthermore, the reac-
tions of primary and secondary amines with CO2 are used to
capture CO2 through the formation of a zwitterion first and
then a carbamate.[8] Encouraged by these previous studies, we
proposed the possibility of utilizing the weak interactions be-
tween CO2 and amines to functionalize NÀH bonds. We report
here the results on the N-formylation of amines mediated by
CO2 through a metathesis reaction [Eq. (1)]. Usually, CÀN bond
activation requires metal catalysts.[9] This method is markedly
Results and Discussion
First, we optimized the reaction temperature and time. The
temperature dependence of the conversion of benzylamine in
DMF under 1 bar of CO2 gas is summarized in Figure S1. The
onset temperature for the reaction has been reported to be
more than 808C or even as high as 1508C.[10] In this study, the
onset temperature (ꢀ408C, defined as the 30% conversion
temperature) is clearly much lower. The conversion increased
with increasing reaction temperature and reached 92% after
4 h at 1008C. The selectivity for N-benzylformamide is also pre-
sented in Figure S1. An increased reaction temperature of
608C or higher offered almost pure N-benzylformamide. The
reaction did not proceed under Ar gas. The reaction in air of-
fered comparable results in 48 h. In 1973, Kraus reported that
the formylation of aliphatic amines occurred in pure DMF.[11]
Up to 100 h was required to complete the reaction. Our study
revealed that the reaction conducted by Kraus might be medi-
ated by the slow absorption of atmospheric CO2 into the reac-
tion mixture.
[a] Y. Wang, J. Zhang, J. Liu, C. Zhang, Z. Zhang, Prof. Dr. J. Xu, Dr. S. Xu,
Dr. F. Wang, Prof. Dr. F. Wang
State Key Laboratory of Catalysis (SKLC)
Dalian National Laboratory for Clean Energy (DNL)
Dalian Institute of Chemical Physics (DICP)
Chinese Academy of Sciences
457 Zhongshan Road, Dalian 116023 (PR China)
E-mail: xujie@dicp.ac.cn
wangfeng@dicp.ac.cn
The progress of the reaction as monitored by GC is dis-
played in Figure 1. The formylation of benzylamine was con-
ducted in DMF at 1008C under 1 bar of CO2. The conversion of
benzylamine increased significantly to 86% in 2 h and then
slowed down and reached 93% in 8 h and >99% in 24 h.
[b] C. Zhang
University of the Chinese Academy of Sciences
No.19A Yuquan Road, Beijing 100049 (PR China)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201500318.
ChemSusChem 2015, 8, 2066 – 2072
2066
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 1. Time-on-stream profile of the formylation of benzylamine to ben-
zylformamide. Reaction conditions: benzylamine (1.5 mmol), DMF (2 mL),
1 bar CO2, 1008C.
Figure 2. Relationship between amine conversion and pKa(NÀH) values. Reac-
tion conditions: amine (0.6 mmol), DMF (2 mL), 1 bar CO2, 1008C, 2 h.
Through the whole reaction, the selectivity for N-benzylform-
amide was >99%.
A reaction with isotopically labelled [D7]DMF was conducted
to clarify the reaction route (Figure S2). The molecular ion (M+)
peak of N-benzylformamide (normal m/z=135) increased to
m/z=136 (labeled), and the M+ peak of dimethylamine (m/z=
45) increased accordingly to m/z=51 (Figures S3 and S4). We
also used 13CO2 instead of CO2 in this reaction system. There
was no increase for the molecular ion peak of N-benzylforma-
mide and it remained at m/z=135. However, if the C atom
came from 13CO2, the M+ peak of N-benzylformamide should
have an m/z value of 136. The results confirm that (1) a meta-
thesis reaction occurs between the amine and DMF, (2) DMF is
the formylation reagent, and (3) CO2 is a mediator, not a reac-
tant. In some other studies, although CO2 was added and for-
mamides were products, the real reaction involved the hydro-
genation of CO2 to formic acid, which functioned as the formy-
lation reagent.[3h,i,5a,12] In contrast, the CO2-mediated method
requires no metal catalyst and no hydrogen. Moreover, it can
be used to prepare 13C-labeled formamides.
thumb, it helps to determine which kind of amine can be con-
verted by this method.
We then studied the substrate scope (Table 1). Sterically un-
hindered linear aliphatic amines with chains of up to 12
carbon atoms were converted to the corresponding forma-
mides with >99% selectivity and >99% conversion (1–5);
these results are better than those reported previously.[10,14] Ali-
cyclic cyclohexylamine and cyclopentylamine were converted
satisfactorily to the corresponding formamides (4 and 5). No
conversion was observed for dipropylamine and diethylamine
(6 and 7). However, a hydroxy-terminated amine (8) and piperi-
dine (9) achieved 23% and >99% conversion, respectively,
with >99% formamide selectivity. These amines were different
at the terminal group. Steric as well as electronic and solvation
effects might affect their activity. The two terminal hydroxy
groups might form hydrogen bonds and, thus, expose more
space for the NÀH bond activation. Piperidine, which could be
viewed as a molecule in which the terminal groups are
bonded, achieved better results. In contrast, a piperidine for-
mamide was obtained with only 36% yield in a B(OCH2CF3)3–
DMF system.[10] To show the practical application of the proce-
dure, a gram-scale synthesis of dodecylamine (2.78 g) was per-
formed. The reaction offered 97% isolated yield of dodecylfor-
mamide (3).
Heterocyclic amines such as furfurylamine and 2-thiophene-
methylamine were converted to formamides (10 and 11, re-
spectively) with selectivities of >99%. To the best of our
knowledge, this method represents a rare example of the for-
mylation of a thiophene amine because sulfur may poison
most catalytic metal centers. Several benzylamines were con-
verted with excellent formamide selectivities (12–18). A slight
substituent effect was evidenced by the lower conversion of
the CF3-substituted benzylamine than for the F-substituted
one. The less-nucleophilic aniline was inactive (19) because of
its weak basicity (Figure 2). The formylation of the sterically un-
hindered hexane-1,6-diamine afforded 99% of the diforma-
mide (20). A worse result than that for hexane-1,6-diamine was
We then investigated the dependence of the initial amine
conversion (2 h) on the pKa(NÀH) values. The larger the pKa(NÀH)
value, the stronger the amine basicity. A plot of pKa(NÀH) versus
conversion produces symmetric linear fits that intercept at the
maximum point of pKa(NÀH) =9.9 (Figure 2). The left end point
was determined by extending the left half line to y=0, which
intercepted the x axis at pKa(NÀH) =8.6. The right end point was
determined by dipropylamine [pKa(NÀH) =10.6]. The experimen-
tal data showed that the metathesis reaction relies greatly on
the amine pKa(NÀH), and only amines within the pKa(NÀH) range
8.6–10.6 were converted into formamides. This explains why
the reactions with aniline and some diamines were sluggish.[13]
The basicity of the amine affects the strength of the interaction
between the N atom of the amine and the C atom of CO2:
weak basicity [such as aniline, pKa(NÀH) <8.6] results in an insta-
ble amine–CO2 adduct; strong basicity [pKa(NÀH) >10.6] gener-
ates an interaction that is too stable, and the CO2 is bound
strongly. The amines in Figure 2 were selected after consider-
ation of both the pKa(NÀH) and steric hindrance; therefore, some
amines fall out of this range. However, as a simple rule of
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2067
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
methods remains a hot topic.[16] In this study, we tested the la-
beling of protein amine groups with DMF through metathesis
reactions. We added a little water to promote solubility but
sacrificed some activity. After the reaction, both unreacted ar-
ginine (m/z=175) and formylated arginine (m/z=203) were
detected by means of matrix-assisted laser desorption/ioniza-
tion–time of flight (MALDI-TOF)/TOF MS (Figure 3A). The reac-
tion was further quantified by LC–MS (Figure 3B) using an
identical amount of lysine as internal standard. After the reac-
tion, the relative intensity of arginine decreased greatly. The ar-
ginine conversion was 81%, as analyzed by LC–MS.
Table 1. Substrate scope of the reaction.[a]
Structure Compound
Conversion Selectivity
[%]
[%]
1
2
3
n=3
n=5
n=11
>99
>99
> 99
>99[b]
>99
>99
4
>99
>99[b]
>99
5
>99
6
7
n=1
n=2
trace
trace
–
–
We further applied the method to more-complex amines
from trypsin-digested bovine serum albumin (BSA). Trypsin is
a pancreatic serine endoprotease that hydrolyzes peptide
bonds specifically at the carboxyl side of arginine and lysine
residues and leads to two classes of peptides. One is the pep-
tide with the arginine residue at the carboxyl side, in which
only the amine group of the N-terminus and the guanidine
group of the arginine residue could be formylated, and the
other is the peptide with the lysine residue at the carboxyl
side, in which the amine group of the N-terminus and the
lysine residue could be formylated. For the former, 57% of the
amine groups of the N-terminus, 14% of the guanidine
groups, and 0% of both were formylated (Figure S5). For the
later, these data were 67%, 72%, and 46%, respectively. For
8
23
>99
9
>99
>99
>99
>99
10
11
>99
94
12 X=H
>99
95
92
93
96
>99
94
90
>99
68
13 X=p-Cl
14 X=o-Cl
15 X=p-F
16 X=p-CF3
17
18
91
98
>99[b]
98
19
20
trace
–
>99
>99
21[c] carried out in CO2
22[d] catalyzed by CeO2 instead
of CO2
>99
70
72
4
23[e]
>99
>99
68
24
>99
[a] Reaction conditions: amine (1.5 mmol), DMF (2 mL), 1008C, 24 h, 1 bar
CO2; results were analyzed by GC and are presented as conversion/selec-
tivity. [b] 48 h. [c] The remaining product was 1,4-diazepane-1-carbalde-
hyde (selectivity 28%). [d] Amine (1.5 mmol), DMF (2 mL), CeO2 (100 mg),
1508C, 24 h.[4k] 1,4-Diazepane-1-carbaldehyde selectivity 96%. [e] Pipera-
zine-1-carbaldehyde selectivity 32%.
obtained for an alicyclic diamine with >99% conversion and
72% selectivity for the diformamide and 28% for the monofor-
mamide (21). This might be due to the robust structure of the
alicyclic diamine. However, this method was still superior to
one with solid CeO2 catalyst, for which only 70% conversion
and 4% of the diformamide was obtained (22).[4k] The formyla-
tion of piperazine offered 68% diformamide and 32% mono-
formamide (23). For dimethyl-substituted piperazine, monofor-
mamide was obtained with >99% selectivity (24).
Carbamylation reactions play important roles in multiple bio-
logical processes, such as the transportation of amino acids
and protein synthesis in living systems.[15] To date, the protein
formylation methods employed use expensive labeling re-
agents and complex procedures, but the development of new
Figure 3. (A) MALDI-TOF results for arginine with a-cyano-4-hydroxycinnamic
acid (CHCA) as the matrix. The y axis for arginine was vertically adjusted.
(B) LC–MS results for the N-formylation of arginine.
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2068
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
histidine, 28% of the side imidazole groups were formylated,
which shows its inertness. The reaction was governed by the
pKa(NÀH) rule (Figure 2) and the effect of steric hindrance. Usual-
ly, acetylation labeling through chemical reactions can only be
performed for the primary amino group on the protein or pep-
tide N-terminus and the lysine side chain. In this study, the
method can be performed not only on the N-terminus and
lysine but also on the guanidine group of the side chain of ar-
ginine. Although the CO2-mediated metathesis reaction is less
efficient than the mature acetylation method, it provides a pos-
sible next-generation labeling strategy, especially for amines
with coexisting functional groups sensitive to metal ions and
strong acids or bases.
In previous studies, DMF and its derivatives were activated
by strong additives such as HClO4, iodine, NaOMe, and RSO2Cl.
In our study, CO2 is a weakly acidic molecule and cannot acti-
vate DMF. We believed CO2 would activate amines and con-
firmed this through IR and NMR spectroscopy characteriza-
tions.
Figure 5. Stacked in situ FTIR spectra for initial 30 min. Reaction conditions:
benzylamine (3 mmol), DMF (24 mL), 1 bar CO2, 1008C.
for dipropylamine and d=158 ppm for benzylamine were as-
signed to the carbonyl C atoms of the N-carbamic acids [Fig-
ure 6A and B].[17] The discernible low-field shifts of the C
atoms next to the N atoms from d=52–48 ppm for dipropyla-
mine and from d=46–44 ppm for benzylamine indicated the
gain of electron density after the formation of the N-carbamic
acids. The chemical shift at 39 ppm for both amines was due
to the trace residue of normal DMSO in the [D6]DMSO solvent.
The presence of the N-carbamic acid was again confirmed by
1H NMR spectroscopy in [D6]DMSO. After the CO2 bubbling, the
d(OÀH) of the N-carbamic acid appeared at 7.7 ppm, whereas
the d(CÀH) of dipropylamine at 2.4 ppm was upshifted to
3.0 ppm (Figure 6C), which was assigned to the d(CÀH) of the
N-carbamic acid.[17] After the bubbling of CO2 into benzyla-
mine, the d(CÀH) of the amine at 3.7 ppm shifted to 4.2 ppm,
which was indicative of d(CÀH) of the N-carbamic acid. The
newly formed d(OÀH) of the N-carbamic acid appeared at
8.7 ppm (Figure 6D). Signals at d=3.4 and 1.1 ppm were at-
tributed to ethanol contamination from the tube for CO2 bub-
bling. The signal at d=2.5 ppm originated from normal DMSO
in [D6]DMSO.
The attenuated total reflectance (ATR)-IR spectra of benzyla-
mine were measured before and after CO2 was bubbled
through it at room temperature (Figure 4A). The CO2 dissolved
in the benzylamine, as indicated by a peak at n˜ =2340 cmÀ1
for the asymmetric stretch of CO2. Two peaks at n˜ =1720 and
1648 cmÀ1 were assigned to the ÀCOOH vibration of the N-car-
bamic acid.[17] No carbamate was formed, as indicated by the
absence of an absorption at n˜ ꢀ1570 cmÀ1.[17] For DMF with
CO2 bubbling at room temperature (Figure 4B), the spectrum
resembled that of pure DMF, which indicates that no interac-
tion occurred between DMF and CO2. More detailed reaction
information for benzylamine, DMF, and CO2 was obtained by
monitoring the initial 30 min reaction through ATR-IR spectros-
copy (Figure 5). The bubbling of CO2 generated three peaks at
n˜ =1231, 1542, and 1720 cmÀ1, which indicated the formation
of the N-carbamic acid. The formation of N-benzylformamide
was evidenced by the peaks at n˜ =1673 and 1382 cmÀ1.
The formation of the N-carbamic acids of dipropylamine and
benzylamine after CO2 bubbling in [D6]DMSO was further char-
acterized by 13C NMR spectroscopy. The signals at d=157 ppm
We then studied the solvent effect by 13C NMR spectroscopy
(Figure 7). The addition of CO2 into a mixture of benzylamine
and dipropylamine in [D6]DMSO and [D7]DMF at room temper-
ature resulted in 100% conversion to the N-carbamic acids.
This was illustrated by the presence of two new C=O signals at
d<160 ppm (159 and 158 ppm in [D6]DMSO, 160 and
159 ppm in [D7]DMF, respectively), in good agreement with the
results in Figure 6. In contrast, only a single signal appeared at
d=163 ppm in CDCl3, but it was stronger than the signals re-
corded in [D6]DMSO and [D7]DMF. The value was larger than
the normal shift of the C atoms of carbamic acids (less than
160 ppm). This signal originated from ammonium carbamate
+
[PhCH2NHCOOÀ (CH3CH2CH2)2NH2 ], the adduct of the amine
and CO2 in CDCl3; thus, only one signal was discerned.[17]
The reaction of an amine and CO2 to form an N-carbamic
acid (CA) is rapid and has a reaction rate k1 of 104 mÀ1 sÀ1. The
reverse reaction is slow with k1’=101 mÀ1 sÀ1.[18] Previous stud-
ies mentioned that the CÀN bond might undergo an exchange
reaction of the carboxyl group with CO2.[19] The formation of
Figure 4. In situ ATR-IR spectra of (A) benzylamine and (B) DMF, flushed with
Ar gas and in the presence of CO2.
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2069
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 6. 13C and 1H NMR spectroscopy characterization of dipropylamine (A and C) and benzylamine (B and D) in Ar or CO2.
1
CA was confirmed by H NMR, 13C NMR, and ATR-IR spectrosco-
py. The key reaction is that the CA further reacts with DMF to
generate the formamide (FA) product and release CO2, which
was also confirmed by IR and NMR spectroscopy. The reaction
of DMF with CA may involve a CÀN bond-metathesis transient
state. We isolated pure ammonium carbamate zwitterion (AC)
and reacted it with DMF (Figure S6). As the reaction selectivity
was >99%, we used the FA production rate as the reaction
rate. The initial rate of 0LmolÀ1 sÀ1 might infer that AC was
converted to CA during this period. Further, the reaction rates
decreased with time with an average rate k2 of 10À6 LmolÀ1 sÀ1.
This rate was much smaller than k1; therefore step 2 is the
rate-determining step.
Therefore, the initial reaction quickly generated CA and then
AC, as an amine reservoir (Figure 8). The FA was generated
from CA and DMF. The addition of a small amount of water ir-
reversibly hydrolyzed CA to carbonate. A control reaction in
a 1:1 volumetric ratio mixture of DMF and water at 1008C for
Figure 7. 13C NMR spectra of a mixture of benzylamine and dipropylamine in
CDCl3, [D6]DMSO, or [D7]DMF after CO2 bubbling.
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2070
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
gradual disappearance of the gel
indicated the end of reaction,
which was also tracked by GC
(Agilent 7890A). The reaction solu-
tion was cooled to room tempera-
ture, and then a sample of the re-
action mixture (ꢀ0.2 mL) was ana-
lyzed by GC–MS (Agilent 7890A-
5975C). In most cases, the reac-
tion was very clean and, thus, the
conversion of amine and the se-
lectivity for formamide were ob-
tained from the normalized GC in-
tegration peak areas.
General characterizations and
analyses: All liquid NMR spectros-
copy experiments were performed
Figure 8. The possible reaction mechanism.
with a Bruker Avance III-400 spec-
trometer with [D6]DMSO, [D7]DMF,
or CDCl3 as the deuterated solvent for locking. 1H and 13C NMR
spectroscopy experiments with proton decoupling were conducted
at 100.62 MHz with 200 scans and a 2 s recycle delay. The chemical
shifts were referenced to tetramethylsilane (TMS). In situ ATR-IR
spectra were recorded with a Mettler Toledo React IR 4000 spec-
trometer with a DiCom probe. The IR spectra were recorded with
[benzylamine]=0.75m in DMF after CO2 degassing for at least
2 min under continuous stirring.
Bovine serum albumin (BSA) digestion: The BSA was denatured
with 8m urea/100 mm tetraethylammonium bromide (TEAB,
pH 8.0), reduced by 10 mm dithiothreitol at 608C for 1 h, and alky-
lated with 20 mm iodoacetamide in darkness at room temperature
for 30 min, followed by dilution with 100 mm TEAB buffer (pH 8.0)
and digestion with trypsin at 378C for 16 h. Finally, the BSA digests
were desalted with C18 solid-phase extraction (SPE) columns and
lyophilized for use.
4 h gave 24% conversion of benzylamine, in comparison with
92% conversion in pure DMF. Moreover a reaction of the sepa-
rated carbonate with DMF generated no product. These facts
indicate that carbonate is not the reaction intermediate and its
production is irreversible. The steric hindrance of the methyl
groups (R1, R2) connected to the N atom of the amine interrupt
the interaction of DMF with the amine and retard this reaction.
The metathesis product dimethylamine, as the leaving group
of DMF, has a pKa(NÀH) value of 10.6, which is large enough to
form a stable carbamic acid product with CO2 under the reac-
tion conditions and drive the reaction to the right. After the re-
action, an increase in temperature or flushing with N2 would
remove dimethylamine and CO2 gas; therefore, the reaction is
clean and convenient.
MALDI-TOF MS analysis: The MALDI-TOF MS experiments were
performed using an AB Sciex 5800 MALDI-TOF/TOF mass spectrom-
eter (AB Sciex) equipped with a pulsed Nd–YAG laser at 355 nm in
reflective positive-ion mode. The sample (ꢀ0.5 mL) and matrix
(0.5 mL; 25 mgmLÀ1 2,5-dihydroxybenzoic acid in 50% acetonitrile/
H2O) were spotted on the MALDI plate for MS analysis. The liquid
chromatography coupled with mass spectrometry/mass spectrom-
etry (LC–MS/MS) experiments were performed using a Thermo Q
Exactive mass spectrometer (Thermo) with a nanospray ion source
and a U3000 RSLCnano system (Thermo). After lyophilization, the
samples were redissolved with 0.1% formic acid solution and
loaded on a C18 capillary trap column (200 mm i.d., 4 cm) packed
with C18 AQ beads (5 mm, 120 Š, Daison) and separated by a capilla-
ry analysis column (75 mm i.d.) with C18 AQ beads (3 mm, 120 Š,
Daison). The buffers used for the online analysis were 0.1% (v/v)
formic acid in water and 0.1% (v/v) formic acid in acetonitrile, the
flow rate was 300 nLminÀ1 for nanoflow LC–MS/MS analysis. A gra-
dient from 5 to 35% (v/v) acetonitrile was achieved in 15 min for
arginine samples and 90 min for BSA samples. The MS and MS/MS
spectra were collected by higher-energy collision-induced dissocia-
tion (HCD) at 28% energy in a data-dependent mode with one MS
scan followed by 10 MS/MS scans. The RAW files collected by Xcali-
bur 2.1 were converted to MGF files by Proteome Discoverer
(v1.2.0.208, Thermo) and searched with Mascot (version 2.3.0,
Matrix Science). Cysteine carboxamidomethylation was set as
a static modification of 57.0215 Da, and formylation and methio-
nine oxidation were set as variable modifications of 27.9949 Da
and 15.9949 Da, respectively. The mass tolerances were 10 ppm
and 0.5 Da for the parent and fragment ions, respectively. A maxi-
Conclusions
We have reported the use of CO2 as a mediator in the transfor-
mation of amines to their corresponding formamides in DMF.
The metathesis reaction is conducted without the use of any
metals. This will provide a useful method for the formylation
and protection of amine groups in a simple, green, and eco-
nomical way.
Experimental Section
Methods: All chemicals were analytical grade and used as pur-
chased without further purification. Most of the chemicals were
purchased from Aladdin Chemicals, except DMF, aniline, and dieth-
ylamine, which were purchased from Tianjin Kemiou Chemical Re-
agent Co., Ltd., n-butylamine and cyclohexylamine were purchased
from Sinopharm Chemical Reagent Co., Ltd., and 2,6-dimethylpyr-
azine was purchased from J&K chemicals.
Catalytic reactions: The reactions were conducted in a 50 mL
glass batch reactor (heavy wall, maximum pressure: 0.6 MPa, Syn-
thware Glass). Typically, benzylamine (1.5 mmol), DMF (2 mL), and
a stirring bar were placed in the reactor, which was then charged
with CO2 gas (1 bar) from a gas cylinder. The reaction solution
changed gradually into a thick white gel over several minutes. The
reactor was sealed tightly with a Teflon stopper and then im-
mersed in an oil bath preheated at the desired temperature. The
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2071
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[5] a) C. L. Allen, B. N. Atkinson, J. M. J. Williams, Angew. Chem. Int. Ed. 2012,
51, 1383–1386; Angew. Chem. 2012, 124, 1412–1415; b) B. N. Atkinson,
A. R. Chhatwal, H. V. Lomax, J. W. Walton, J. M. J. Williams, Chem.
Commun. 2012, 48, 11626–11628; c) S. Joseph, P. Das, B. Srivastava, H.
Nizar, M. Prasad, Tetrahedron Lett. 2013, 54, 929–931; d) A. E. Pasqua, M.
Matheson, A. L. Sewell, R. Marquez, Org. Process Res. Dev. 2011, 15,
467–470; e) C. Zhang, Z. J. Xu, T. Shen, G. L. Wu, L. R. Zhang, N. Jiao,
Org. Lett. 2012, 14, 2362–2365.
mum of two missed cleavages was allowed, and the false discovery
rate for peptide identifications was controlled to <1%.
Acknowledgements
This work was supported by the National Natural Science Foun-
dation of China (21303189, 21273231, 21422308), and Hundred
Person Project of the Chinese Academy of Sciences.
[6] B. M. Brandsen, T. E. Velez, A. Sachdeva, N. A. Ibrahim, S. K. Silverman,
Angew. Chem. Int. Ed. 2014, 53, 9045–9050; Angew. Chem. 2014, 126,
9191–9196.
[7] W. H. Schaefer, Curr. Drug Metab. 2006, 7, 873–881.
[8] a) B. Dutcher, M. Fan, A. G. Russell, ACS Appl. Mater. Interfaces 2015, 7,
2137–2148; b) M. X. Tan, Y. Zhang, J. Y. Ying, ChemSusChem 2013, 6,
1186–1190; c) J. R. Switzer, A. L. Ethier, E. C. Hart, K. M. Flack, A. C.
Rumple, J. C. Donaldson, A. T. Bembry, O. M. Scott, E. J. Biddinger, M. Tal-
reja, M.-G. Song, P. Pollet, C. A. Eckert, C. L. Liotta, ChemSusChem 2014,
7, 299–307; d) C. S. Srikanth, S. S. C. Chuang, ChemSusChem 2012, 5,
1435–1442; e) S. Saravanamurugan, A. J. Kunov-Kruse, R. Fehrmann, A.
Riisager, ChemSusChem 2014, 7, 897–902.
[9] a) J. Hu, Y. Xie, H. Huang, Angew. Chem. Int. Ed. 2014, 53, 7272–7276;
Angew. Chem. 2014, 126, 7400–7404; b) S. Guo, B. Qian, Y. Xie, C. Xia, H.
Huang, Org. Lett. 2011, 13, 522–525; c) Y. Xie, B. Qian, P. Xie, H. Huang,
Adv. Synth. Catal. 2013, 355, 1315–1322.
Keywords: carbon dioxide · formamide · metathesis · reaction
mechanisms · synthetic methods
[1] a) G. A. Olah, L. Ohannesian, M. Arvanaghi, Chem. Rev. 1987, 87, 671–
686; b) C. L. Allen, J. M. J. Williams, Chem. Soc. Rev. 2011, 40, 3405–
3415.
[2] a) D. Habibi, M. Nasrollahzadeh, H. Sahebekhtiari, J. Mol. Catal. A 2013,
378, 148–155; b) A. Chandra Shekhar, A. R. Kurnar, G. Sathaiah, V. L.
Paul, M. Sridhar, P. S. Rao, Tetrahedron Lett. 2009, 50, 7099–7101; c) B. A.
Aleiwi, K. Mitachi, M. Kurosu, Tetrahedron Lett. 2013, 54, 2077–2081;
d) A. Hamasaki, Y. Yasutake, T. Norio, T. Ishida, T. Akita, H. Ohashi, T. Yo-
koyama, T. Honma, M. Tokunaga, Appl. Catal. A 2014, 469, 146–152;
e) V. K. Das, R. R. Devi, P. K. Raul, A. J. Thakur, Green Chem. 2012, 14,
847–854.
[10] R. M. Lanigan, P. Starkov, T. D. Sheppard, J. Org. Chem. 2013, 78, 4512–
4523.
[11] M. A. Kraus, Synthesis 1973, 361–362.
[3] a) C. J. Gerack, L. McElwee-White, Chem. Commun. 2012, 48, 11310–
11312; b) H. Tumma, N. Nagaraju, K. V. Reddy, J. Mol. Catal. A 2009, 310,
121–129; c) P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc.
1994, 116, 8851–8852; d) W. Y. Hao, G. D. Ding, M. Z. Cai, Catal.
Commun. 2014, 51, 53–57; e) X. J. Cui, Y. Zhang, Y. Q. Deng, F. Shi,
Chem. Commun. 2014, 50, 189–191; f) O. Jacquet, C. Das Neves Gomes,
M. Ephritikhine, T. Cantat, J. Am. Chem. Soc. 2012, 134, 2934–2937; g) Y.
Li, I. Sorribes, T. Yan, K. Junge, M. Beller, Angew. Chem. Int. Ed. 2013, 52,
12156–12160; Angew. Chem. 2013, 125, 12378–12382; h) K. Beydoun,
T. vom Stein, J. Klankermayer, W. Leitner, Angew. Chem. Int. Ed. 2013, 52,
9554–9557; Angew. Chem. 2013, 125, 9733–9736; i) C. Das Neves
Gomes, O. Jacquet, C. Villiers, P. ThuØry, M. Ephritikhine, T. Cantat,
Angew. Chem. Int. Ed. 2012, 51, 187–190; Angew. Chem. 2012, 124,
191–194.
[4] a) S. Ding, N. Jiao, Angew. Chem. Int. Ed. 2012, 51, 9226–9237; Angew.
Chem. 2012, 124, 9360–9371; b) J. Muzart, Tetrahedron 2009, 65, 8313–
8323; c) T. B. Nguyen, J. Sorres, M. Q. Tran, L. Ermolenko, A. Al-Mourabit,
Org. Lett. 2012, 14, 3202–3205; d) R. Vanjari, B. Kumar Allam, K. N.
Singh, RSC Adv. 2013, 3, 1691–1694; e) S. N. Rao, D. C. Mohan, S. Adi-
murthy, Org. Lett. 2013, 15, 1496–1499; f) P. R. Giles, C. M. Marson, Tetra-
hedron 1991, 47, 1303–1310; g) Y. J. Liu, X. L. Xu, Y. M. Zhang, Synlett
2004, 445–448; h) G. Pettit, E. Thomas, J. Org. Chem. 1959, 24, 895–
896; i) M. Suchy, A. A. H. Elmehriki, R. H. E. Hudson, Org. Lett. 2011, 13,
3952–3955; j) K. Takahashi, M. Shibagaki, H. Matsushita, Agric. Biol.
Chem. 1988, 52, 853–854; k) Y. Wang, F. Wang, C. Zhang, J. Zhang, M.
Li, J. Xu, Chem. Commun. 2014, 50, 2438–2441.
[12] a) E. Blondiaux, J. Pouessel, T. Cantat, Angew. Chem. Int. Ed. 2014, 53,
12186–12190; Angew. Chem. 2014, 126, 12382–12386; b) Q. Liu, L. Wu,
R. Jackstell, M. Beller, Nat. Commun. 2015, 6, 5933.
[13] In some cases, steric hindrance should be considered, for example, for
dipropylamine. Overall, the effect of pKa(NÀH) should play a key role here.
[14] N. Ortega, C. Richter, F. Glorius, Org. Lett. 2013, 15, 1776–1779.
[15] C. Francklyn, P. Schimmel, Nature 1989, 337, 478–481.
[16] a) T. Jiang, X. F. Zhou, K. Taghizadeh, M. Dong, P. C. Dedon, Proc. Natl.
Acad. Sci. USA 2007, 104, 60–65; b) D. Mader, M. Liebeke, V. Winstel, K.
Methling, M. Leibig, F. Gotz, M. Lalk, A. Peschel, BMC Microbiol. 2013,
13, 7.
[17] a) Z. J. Dijkstra, A. R. Doornbos, H. Weyten, J. M. Ernsting, C. J. Elsevier,
J. T. F. Keurentjes, J. Supercrit. Fluids 2007, 41, 109–114; b) K. Masuda, Y.
Ito, M. Horiguchi, H. Fujita, Tetrahedron 2005, 61, 213–229.
[18] W. Conway, X. Wang, D. Fernandes, R. Burns, G. Lawrance, G. Puxty, M.
Maeder, Environ. Sci. Technol. 2012, 46, 7422–7429.
[19] a) C. S. McCowan, M. T. Caudle, Dalton Trans. 2005, 238–246; b) Y.
Otsuji, N. Matsumura, E. Imoto, Bull. Chem. Soc. Jpn. 1968, 41, 1485–
1485.
Received: March 3, 2015
Revised: April 3, 2015
Published online on June 3, 2015
ChemSusChem 2015, 8, 2066 – 2072
www.chemsuschem.org
2072
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim