Dehydrogenative Synthesis of Carbamates from Formamides and Alcohols Using a Pincer-Supported Iron Catalyst

Article abstract of DOI:10.1021/acscatal.1c02718

We report that the pincer-ligated iron complex (iPrPNP)Fe(H)(CO) [1, iPrPNP- = N(CH2CH2PiPr2)2-] is an active catalyst for the dehydrogenative synthesis of N-alkyl- and N-aryl-substituted carbamates from formamides and alcohols. The reaction is compatible with industrially relevant N-alkyl formamides, as well as N-aryl formamides, and 1°, 2°, and benzylic alcohols. Mechanistic studies indicate that the first step in the reaction is the dehydrogenation of the formamide to a transient isocyanate by 1. The isocyanate then reacts with the alcohol to generate the carbamate. However, in a competing reaction, the isocyanate undergoes a reversible cycloaddition with 1 to generate an off-cycle species, which is the resting state in catalysis. Stoichiometric experiments indicate that high temperatures are required in catalysis to facilitate the release of the isocyanate from the cycloaddition product. We also identified several other off-cycle processes that occur in catalysis, such as the 1,2-addition of the formamide or alcohol substrate across the Fe-N bond of 1. It has already been demonstrated that the transient isocyanate generated from dehydrogenation of the formamide can be trapped with amines to form ureas and, in principle, the isocyanate could also be trapped with thiols to form thiocarbamates. Competition experiments indicate that trapping of the transient isocyanate with amines to produce ureas is faster than trapping with an alcohol to produce carbamates and thus ureas can be formed selectively in the presence of alcohols. In contrast, thiols bind irreversibly to the iron catalyst through 1,2 addition across the Fe-N bond of 1, and it is not possible to produce thiocarbamates. Overall, our mechanistic studies provide general guidelines for facilitating dehydrogenative coupling reactions using 1 and related catalysts.

Full text of DOI:10.1021/acscatal.1c02718

pubs.acs.org/acscatalysis
Research Article
Dehydrogenative Synthesis of Carbamates from Formamides and
Alcohols Using a Pincer-Supported Iron Catalyst
Tanya M. Townsend, Wesley H. Bernskoetter,* Nilay Hazari,* and Brandon Q. Mercado
Cite This: ACS Catal. 2021, 11, 10614−10624
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: We report that the pincer-ligated iron complex
−
(iPrPNP)Fe(H)(CO) [1, iPrPNP− = N(CH2CH2PiPr2)2 ] is an
active catalyst for the dehydrogenative synthesis of N-alkyl- and N-
aryl-substituted carbamates from formamides and alcohols. The
reaction is compatible with industrially relevant N-alkyl forma-
mides, as well as N-aryl formamides, and 1°, 2°, and benzylic
alcohols. Mechanistic studies indicate that the first step in the
reaction is the dehydrogenation of the formamide to a transient
isocyanate by 1. The isocyanate then reacts with the alcohol to
generate the carbamate. However, in a competing reaction, the isocyanate undergoes a reversible cycloaddition with 1 to generate an
off-cycle species, which is the resting state in catalysis. Stoichiometric experiments indicate that high temperatures are required in
catalysis to facilitate the release of the isocyanate from the cycloaddition product. We also identified several other off-cycle processes
that occur in catalysis, such as the 1,2-addition of the formamide or alcohol substrate across the Fe−N bond of 1. It has already been
demonstrated that the transient isocyanate generated from dehydrogenation of the formamide can be trapped with amines to form
ureas and, in principle, the isocyanate could also be trapped with thiols to form thiocarbamates. Competition experiments indicate
that trapping of the transient isocyanate with amines to produce ureas is faster than trapping with an alcohol to produce carbamates
and thus ureas can be formed selectively in the presence of alcohols. In contrast, thiols bind irreversibly to the iron catalyst through
1,2 addition across the Fe−N bond of 1, and it is not possible to produce thiocarbamates. Overall, our mechanistic studies provide
general guidelines for facilitating dehydrogenative coupling reactions using 1 and related catalysts.
KEYWORDS: iron, transition-metal catalysis, dehydrogenative coupling, carbamate formation, pincer ligands, reaction mechanism
The majority of methods to synthesize carbamates
industrially involves toxic reagents, such as isocyanates,6
carbon monoxide,7 and phosgene or its derivatives.8
Isocyanates are the most commonly utilized starting material
and can be readily coupled with alcohols to produce the
desired carbamate (Figure 2a). Depending on the substituents
INTRODUCTION
■
Carbamates are key monomers in polyurethane polymers,
which are frequently used as surface coatings, foams, and
textiles, and are also widely utilized as pharmaceuticals and
agrochemicals.1,2 For example, carbendazim is a commercially
available broad-spectrum fungicide that is used in agriculture
and forestry (Figure 1),3 while insecticides featuring
carbamates, such as carbaryl, are utilized as alternatives to
organophosphate insecticides.4 Additionally, carbamates are an
important class of protecting groups for amines in synthetic
chemistry, including the Boc (t-butyloxycarbonyl), CBz
(carbobenzyloxy), and Fmoc (9-fluorenylmethyloxycarbonyl)
groups.5
̈
on the starting materials, a catalyst, such as a Bronsted or Lewis
acid, or a bulky amine base, may be required (Figure 2a).9
Unfortunately, the direct use of isocyanates is not optimal
because they are toxic by inhalation, ingestion, and contact. In
particular, methyl isocyanate, which is used in the industrial
synthesis of insecticides and polyurethanes, is extremely toxic,
flammable, and boils at 39 °C.10 Methyl isocyanate also
requires careful storage as it reacts violently with water and
readily trimerizes in the presence of a catalytic amount of a
Received: June 16, 2021
Revised: July 27, 2021
Published: August 11, 2021
Figure 1. Examples of commercially available carbamates.
https://doi.org/10.1021/acscatal.1c02718
© 2021 American Chemical Society
ACS Catal. 2021, 11, 10614−10624
10614

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
base or metal salt (Figure 2d), both of which are highly
exothermic processes that can lead to runaway reactions.
then reacts with the alcohol to form a carbamate. The Milstein
system represents an important proof-of-principle but can only
catalyze the reactions with N-aryl substituted formamides. The
extension of the method to formamides with N-alkyl
substituents, such as N-methylformamide (NMF), which can
act as a source of methyl isocyanate, would increase its
potential utility. Further, there are likely benefits associated
with sustainability and toxicity, if the precious metal ruthenium
catalyst is replaced with a system based on a more sustainable
first-row transition metal.17
We and others have reported that the pincer-supported iron
(pre)catalyst, (iPrPNP)FeH(CO) [1, iPrPNP = N-
(CH2CH2PiPr2)2−], along with the related dihydride
(iPrPNHP)Fe(H)2(CO) [2, iPrPNHP = HN(CH2CH2PiPr2)2],
is highly active for the dehydrogenation of a range of substrates
including formic acid,18 primary and secondary alcohols,19
glycerol,20 N-heterocycles,21 and ammonia-borane,22 as well as
the dehydrogenative synthesis of lactones,23 lactams,23
amides,24 and ureas.25 In particular, we hypothesized that the
dehydrogenative synthesis of ureas from methanol and excess
amine is relevant to the development of a first-row system for
the dehydrogenative preparation of carbamates. In urea
synthesis, 1 is proposed to couple an alcohol and an amine
to generate a formamide, which undergoes dehydrogenation to
an isocyanate intermediate which is then trapped with an
amine to generate urea. We hypothesized that we could use 1
to generate isocyanates in situ from formamides and then trap
the isocyanates with an alternative nucleophile, such as an
alcohol or thiol, to form carbamates or thiocarbamates,
respectively (Figure 3). Herein, we report the first example
of dehydrogenative carbamate synthesis with a first-row
transition-metal catalyst. The system is active for a variety of
N-aryl- and N-alkyl-substituted formamides, including NMF,
and 1°, 2°, and benzylic alcohols. We also describe mechanistic
studies, which enable us to understand the relative ability of 1
to dehydrogenate amines, alcohols, and formamides and
mediate the trapping of isocyanates with amines and alcohols.
This fundamental work is expected to guide the design of
future dehydrogenative reactions using 1 and related catalysts.
Figure 2. (a−c) Representative existing methods for carbamate
synthesis. (d) Exothermic trimerization of isocyanates. (e) Compar-
ison of dehydrogenative synthesis of carbamates from formamides and
alcohols with catalyst A (previous work15) or 1 (this work).
Alternative methods to synthesize carbamates that do not
utilize isocyanates have been developed but are currently too
expensive to be utilized on an industrial scale. For example, the
treatment of di-substituted carbonates with amines in the
presence of NaH generates carbamates (Figure 2b).11 These
reactions can give high yields of carbamates with both aryl and
alkyl substituents, but the use of NaH limits the functional
group tolerance, and the carbonate starting materials can be
expensive to prepare. In a related reaction, Beller and co-
workers treated ureas with alcohols to generate carbamates
using iron salts as catalysts, but this reaction is limited to
unsubstituted ureas, which allow the reaction to be driven by
releasing NH3.12 A newly developed route to synthesize alkyl
carbamates involves treating CO2 with an amine and alcohol in
the presence of a base (Figure 2c).13 However, this method
requires high pressures of CO2 or supercritical conditions14
and the carbamate is often formed as a part of a mixture of urea
and carbonate products, which must be separated.
Figure 3. Dehydrogenative synthesis of ureas, carbamates, or
thiocarbamates from formamides.
In a conceptually different approach, the Milstein group
recently reported the first examples of the dehydrogenative
synthesis of carbamates from formamides and alcohols using
ruthenium catalyst A (Figure 2e).15 This method is atom
efficient, as H2 is the only byproduct, and is attractive because
formamides are considerably less toxic, volatile, and flammable
than isocyanates and are easier to synthesize.16 The proposed
mechanism involves the initial dehydrogenation of the
formamide in situ to generate a transient isocyanate, which
RESULTS AND DISCUSSION
■
Catalysis. Initial attempts to evaluate the ability of 1 to act
as a catalyst for the dehydrogenative formation of carbamates
were performed using the substrates N-cyclohexylformamide
and cyclohexanol, both of which are commercially available.
Reactions performed under the optimized conditions for urea
synthesis catalyzed by 1 (0.5 mol % 1, THF, 120 °C)25 did not
lead to carbamate formation (see Table S1 in Supporting
10615
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Information). Instead, higher boiling solvents, such as xylenes
or toluene, which enable the reaction to be performed at
higher temperatures, were required for the production of
carbamate 3a (Table 1). For example, a reaction between N-
cyclohexylformamide and 5 equiv of cyclohexanol in xylenes
generates 3a in 64% yield, while the same reaction performed
in toluene produces 3a in 76% yield (entries 1 and 2).
Reactions performed below the boiling point of the solvent
produced trace amounts of 3a (see Table S5 in Supporting
Information), presumably because at reflux H2 that is
generated as a reaction byproduct has limited solubility and
is rapidly released into the headspace of the flask.26 Decreasing
the catalyst loading from 5 mol % 1 to 1 mol % caused a sharp
drop in yield to 32% (entry 3), while decreasing the
equivalents of cyclohexanol resulted in a slight increase in
the yield (entries 4 and 5). This increase in yield is likely not
related to the suppression of competitive alcohol dehydrogen-
ation19b because in all the reactions only very small quantities
(<5%) of cyclohexanone were observed. This indicates that
alcohol dehydrogenation is not a major side reaction and
instead we propose that the alcohol can promote the
decomposition of the catalyst, as 1 is known to be unstable
in the presence of acidic species.18,19b,c,27 Increasing the
temperature from 150 to 170 °C did not improve the yield
(entry 6) and the optimal concentration of formamide was 0.5
M (entries 5,7, and 8). Several additives, including Brønsted
bases and Lewis acids, which could in principle promote either
the dehydrogenation of the formamide or the nucleophilic
attack of the alcohol on the isocyanate, were evaluated in the
catalytic reaction and in nearly all the cases a significant
decrease in the yield of 3a was observed (see Table S3 in
Supporting Information). Overall, the optimal conditions for
the dehydrogenative synthesis of 3a can be found in entry 5 of
Table 1 and result in an 86% yield. This is the first example of
the dehydrogenative synthesis of carbamates catalyzed by a
first-row transition metal and the first example on any metal
which involves a formamide with an N-alkyl substituent.15
Given our preliminary success in reactions involving N-
cyclohexylformamide and cyclohexanol, we explored the
substrate scope for the dehydrogenative coupling using a
range of formamides and alcohols under our optimized
conditions (Figure 4). Initially, the formamide substrate was
varied using cyclohexanol as the alcohol substrate. The
reaction is compatible with 1° (3b, 3c), 2° (3a, 3d), and 3°
(3e) N-alkyl-substituted formamides. Notably, the NMF-
derived carbamate 3b, which presumably involves the
generation of methyl isocyanate as an intermediate, is formed
in 40% yield (by NMR spectroscopy) under our optimized
conditions. However, the yield can be increased to 58%
(isolated) by adding an additional 5 mol % of 1 after 24 h and
leaving the reaction for another 24 h. The yield for this two-
step process is significantly higher than that obtained using a
single addition of 10 mol % of 1, which gives a yield of only
47% (by NMR spectroscopy, see Table S7 in Supporting
Information). A comparison of the formation of the related
products 3c, 3d, and 3e suggests that steric factors are
important and the 1° alkyl formamide (3c) is coupled more
efficiently than the 2° alkyl formamide (3d), which is coupled
more efficiently than the 3° alkyl formamide (3e). We were
also able to utilize an N-allylic formamide to form carbamate 3f
with a terminal alkene in 82% yield. The presence of a terminal
alkene is valuable for facilitating further synthetic trans-
formations. Our catalyst is also compatible with an N-benzylic
formamide, and carbamate 3g is generated in 83%. N-aryl
formamides, also known as formanilides, were also evaluated as
substrates, although only moderate yields are observed under
the standard catalytic conditions (see Table S8 in Supporting
Information). The yields of 3h−3k can be increased by adding
an additional 5 mol % of 1 after 24 h and leaving the reaction
for another 24 h. Although no clear electronic trend is
observed with either an electron-donating group (3j) or an
electron-withdrawing group (3k) on the aryl ring, the
respective carbamates are generated in 55 and 46% yields.
Finally, we are able to generate product 3i, which contains an
aryl chloride, in 56% yield. This is significant because it can
enable further functionalization of the carbamate through
palladium-catalyzed cross-coupling.
We evaluated the scope of the alcohol substrate using N-
cyclohexylformamide as the formamide. Initially, when
comparing the yields of products with 1° alcohols (4a−4d),
we observed a correlation between the boiling point of the
alcohol and the yield of the reaction. Under our optimized
conditions with 1 equiv of alcohol, methanol gave no yield of
4a, while 1-octanol gave a 47% yield (4d). The yield with
methanol could be increased by using 20 equiv of methanol,
which resulted in a 37% yield of 4a. Similarly, for both 1-
propanol (4b) and 1-butanol (4c), more than 1 equiv of the
alcohol was required to generate a significant amount of the
product. We propose that more equivalents of lower boiling
alcohols are required to ensure there is a sufficient
concentration of alcohol in the solution at 150 °C. In all the
reactions with 1° alcohols, ester formation, which results from
the dehydrogenation of the alcohol,19b is observed in between
5 and 20% yields (see Table S10 in Supporting Information).
This indicates that it is easier to dehydrogenate 1° alcohols
than the 2° alcohol cyclohexanol. Similarly, no significant
dehydrogenation is observed when 2-butanol is used as a
substrate, although due to the low boiling point of 2-butanol
20 equiv are required to generate 66% yield of 4e. Benzyl
alcohol is also compatible with the reaction and a 35% yield of
carbamate 4f is generated. However, selectivity was a problem
in this reaction and a 30% yield of benzyl benzoate was
observed, indicating that alcohol dehydrogenation is a major
Table 1. Catalytic Optimization of Dehydrogenative
Carbamate Synthesis with 1
formamide
conc. (M)
equiv.
CyOH
yield carbamate
formamide
recovered (%)
a
entry
(%)
b
1
0.5
0.5
0.5
0.5
0.5
0.5
1.0
0.25
5
5
5
2.5
1
1
1
1
64
76
32
82
86
81
59
68
22
18
48
11
11
17
25
24
2
3
c
4
5
6
d
7
8
a
NMR yields are the average of two trials and were determined using
a naphthalene standard. Residual alcohol was not quantified due to its
volatility, which led to partial loss during workup. Cyclohexanone was
b
c
observed in <5% yield in all the entries. Solvent = xylenes. 1 mol %
of 1 used. Temperature = 170 °C.
d
10616
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 4. Isolated and NMR yields for dehydrogenative carbamate synthesis catalyzed by 1. NMR yields are given in parenthesis after the isolated
yields and are the average of two trials. They were determined using a naphthalene or hexamethylbenzene standard. Residual alcohol or NMF was
not quantified due to their volatility, which led to partial loss during workup. Optimized conditions: (a) 48 h of total time, additional 5 mol % 1
added after 24 h, (b) 5 equiv of alcohol used, and (c) 20 equiv of alcohol used.
Figure 5. (a) 1,2-Addition of formamides to 1. (b) Solid-state structure of 5Me with thermal ellipsoids at 30% probability. Most hydrogen atoms
have been omitted for clarity. The iron hydride was located in the difference map and freely refined. Selected bond lengths (Å) and angles (°):
Fe(1)−P(1) 2.2198(9), Fe(1)−P(2) 2.2215(9), Fe(1)−C(2) 1.721(3), Fe(1)−N(1) 2.098(3), Fe(1)−N(2) 2.077(2), C(2)−Fe(1)−N(2)
172.96(13), P(1)−Fe(1)−P(2) 167.57(4), and O(1)−C(1)−N(1) 126.6(3).
side reaction. An alcohol with an oxygen containing hetero-
cycle is tolerated, and a yield of 87% of 4g was obtained.
Sterically bulky 3° alcohols, more acidic alcohols such as
phenols, and N-containing heterocycles do not generate any
carbamate products (see page S35 in Supporting Information).
Overall, our system is tolerant of a variety of N-alkyl and N-
aryl formamides, as well as 1°, 2°, and benzylic alcohols. It is
the first example of a system that couples methanol and benzyl
alcohol and is unique in its ability to dehydrogenate N-alkyl
and N-benzylic formamides, which are not compatible with the
previously described ruthenium system.15 Although 1 can
couple N-aryl formamides, it is not as active as the ruthenium
system, which is not unusual for first-row transition-metal-
based systems. Nevertheless, its ability to couple previously
unreactive substrates that are industrially relevant, such as
NMF, is likely to assist in the development of practical systems
for the dehydrogenative synthesis of carbamates and is
complimentary to the ruthenium system.
Mechanistic Studies. In order to understand the ability of
1 to catalyze the dehydrogenative synthesis of carbamates, we
performed a mechanistic study. Initially, we explored the
stoichiometric reactivity of 1 with different N-alkyl forma-
10617
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 6. (a) Reaction of 5Bn at 100 °C. (b) Synthesis of 6Bn and 6Cy from the reactions of 1 with benzyl and cyclohexyl isocyanate, respectively.
Solid-state structures of (c) 6Bn and (d) 6Cy with thermal ellipsoids at 30% probability. All the hydrogen atoms not bound to iron and the second
identical molecule in the asymmetric unit cell of 6Cy are omitted for clarity. The iron hydrides were located in the difference maps and freely
refined. Selected bond lengths (Å) and angles (°) for 6Bn: Fe(1)−P(1) 2.2128(12), Fe(1)−P(2) 2.2053(12), Fe(1)−N(1) 2.057(3), Fe(1)−C(2)
1.729(4), Fe(1)−N(2) 2.056(3), N(1)−C(1) 1.311(5), N(2)−C(1) 1.510(5), P(2)−Fe(1)−P(1) 166.20(5), C(2)−Fe(1)−N(2) 176.81(17),
N(1)−Fe(1)−N(2) 65.48(13), and O(1)−C(1)−N(1) 135.2(4). Selected bond lengths (Å) and angles (°) for one of the molecules of 6Cy (see
the Supporting Information for other molecules): Fe(1)−P(1) 2.2041(12), Fe(1)−P(2) 2.1974(12), Fe(1)−N(1) 2.064(3), Fe(1)−C(2)
1.727(4), Fe(1)−N(2) 2.057(4), N(1)−C(1) 1.304(6), N(2)−C(1) 1.504(5), P(2)−Fe(1)−P(1) 164.22(5), C(2)−Fe(1)−N(2) 175.99(18),
N(1)−Fe(1)−N(2) 65.68(14), and O(1)−C(1)−N(1) 134.1(4).
mides. It has previously been reported that acidic substrates
such as alcohols24 and formanilide28 undergo 1,2-addition
reactions across the Fe−N bond of 1. N-alkyl formamides,
which are predicted to be slightly less acidic than formanilide
or alcohols, such as methanol,29 also add across the Fe−N
bond of 1. For example, the reaction of 1 with 1 equiv of NMF
or N-benzylformamide at room temperature in C6D6 generates
5Me or 5Bn, respectively, in essentially quantitative yield (Figure
5a). At room temperature in C6D6, there is no evidence that
the formamide is dehydrogenated to form an isocyanate and
both 5Me and 5Bn are indefinitely stable. In fact, 5Me was
isolated in 92% yield and characterized by multinuclear NMR
and IR spectroscopy, as well as single-crystal X-ray diffraction
(Figure 5b). The 1H NMR spectra of 5Me and 5Bn both contain
downfield resonances at 11.43 and 11.03 ppm, respectively,
which are consistent with the newly formed N−H bond on the
iPrPNHP ligand engaging in hydrogen bonding with the
carbonyl oxygen of the amido ligand.18 A hydrogen bonding
interaction is also present in the solid-state structure of 5Me as
there is a short contact between the hydrogen atom of the N−
H on the iPrPNHP ligand and the carbonyl oxygen from NMF.
The geometrical parameters associated with the solid-state
structure of 5Me are similar to those in related six-coordinate
iron complexes supported by iPrPNHP ligands,18,30 although it
is notable that the Fe−N1 bond distance in 5Me is shorter than
in the related complex formed by the 1,2-addition of
formanilide across the Fe−N bond of 1 [2.098(3) vs
2.1287(9) Å].28 This is likely caused by steric factors, as
NMF is less bulky than formanilide. The 1,2-addition of NMF
is reversible because when a d8-toluene solution of 5Me is
heated in an NMR spectrometer at 100 °C;31 peaks consistent
with 1 (12% of the iron) and NMF are present along with
those of 5Me (88% of the iron). Further, consistent with an
equilibrium process, when the sample is cooled back to room
temperature only 5Me is observed.
In contrast to the full conversion observed in the reaction
between 1 and 1 equiv of NMF or N-benzylformamide at room
temperature, a reaction between 1 and 1 equiv of N-
cyclohexylformamide gives an 85:15 mixture of the starting
materials and 5Cy, the 1,2-addition product of N-cyclo-
1
hexylformamide to 1 (Figure 5a). The H NMR spectrum of
the reaction mixture displays a triplet at −22.0 ppm, which we
assign to the iron hydride of 5Cy, based on the similar chemical
shifts associated with the iron hydrides of 5Me and 5Bn (−21.3
10618
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 7. Conversion of 6Cy to 6Bn at 120 °C.
and −22.1 ppm, respectively). The 31P NMR spectrum
displays a signal at 89.3 ppm, which is similar to the shifts
assigned to the iPrPNHP ligands for 5Me and 5Bn (93.3 and 90.5
ppm, respectively). In agreement with the 1,2-addition of N-
cyclohexylformamide being an equilibrium process at room
temperature, the addition of further equivalents of N-
cyclohexylformamide increases the relative amount of 5Cy
(see Figure S55 in Supporting Information). Given the
similarly predicted acidity of N-cyclohexylformamide and
NMF,29b,c the increased favorability of the 1,2-addition of
NMF is likely related to steric factors as the bulky cyclohexyl
group is more likely to clash with the isopropyl substituents on
the iPrPNHP ligand. Again, at room temperature, there is no
evidence for the dehydrogenation of N-cyclohexylformamide
to form cyclohexyl isocyanate.
is in the form of 6Cy, the product of the cycloaddition of
cyclohexyl isocyanate with 1 (see Figures S56 and S57 in
Supporting Information). At 100 °C no 5Cy is observed,
consistent with the unfavorability of the 1,2-addition of N-
cyclohexylformamide to 1 (vide supra). When this sample was
cooled to room temperature, approximately 5% of the iron is in
the form of 2, consistent with the 1,2-addition of H2
(generated from the dehydrogenation of N-cyclohexylforma-
mide) to 1. Presumably, in this case, 2 is observed because
there is more conversion to isocyanate (compared to 5Bn) and
therefore more H2 generated, so not all of 2 reverts back to 1
and H2.18,32 We also independently demonstrated that 1 reacts
with cyclohexyl isocyanate to form 6Cy at room temperature,
which we were able to isolate and characterize using NMR
spectroscopy and single-crystal X-ray diffraction (Figure 6b,d).
The solid-state structure of 6Cy is analogous to 6Bn, as would be
expected because the only change is the alkyl substituent of the
isocyanate, which is not close to the iron center.
Our stoichiometric studies demonstrate that 1 is able to
dehydrogenate formamides to isocyanates but that this
reaction requires temperatures of approximately 100 °C.
Further, the product is not a free isocyanate, as 1 traps the
isocyanate via cycloaddition. Interestingly, at 100 °C, solutions
of the cycloaddition products 6Bn or 6Cy are almost completely
unreactive in the presence of H2, alcohols, or amines over
several days (see page S64 in the Supporting Information).
This suggests that 6Bn or 6Cy do not release free isocyanate into
the solution at these temperatures, as amines and alcohols
would be expected to rapidly react with any free isocyanate to
generate ureas and carbamates, respectively (vide infra).
Consistent with this observation, DFT calculations indicate
that the cycloadditions of benzyl or cyclohexyl isocyanate to 1
are highly thermodynamically favorable. The formation of 6Bn
from 1 and benzyl isocyanate is favored by 11.1 kcal/mol,
while the formation of 6Cy from 1 and cyclohexyl isocyanate is
favored by 8.2 kcal/mol. We propose that it is more
thermodynamically favorable to form 6Bn because it is less
sterically crowded than 6Cy. At 120 °C, we see evidence that
6Cy can generate free cyclohexyl isocyanate. Specifically, the
treatment of 6Cy with 5 equiv of benzyl isocyanate results in
the complete conversion to 6Bn and cyclohexyl isocyanate
(Figure 7),34 as predicted from DFT calculations. However,
this reaction is slow and requires 16 h, suggesting that
isocyanate release from 6Cy is not rapid even at this
temperature.
Heating a solution of 5Bn in d8-toluene at 100 °C in an NMR
spectrometer for 5 min results in the formation of a 1:1
equilibrium mixture of 5Bn and 1 and N-benzylformamide.
However, over an hour, the growth of a new iron hydride (6Bn,
16% of the iron) is observed as evidenced by the appearance of
1
a new triplet at −20.9 ppm in the H NMR spectrum (Figure
6a). We propose that at 100 °C, 1 dehydrogenates some of the
N-benzylformamide to generate benzyl isocyanate and iron
dihydride 2. In the absence of an overpressure of H2, 2 is
unstable and loses H2 by 1,2-elimination across the Fe−N
bond to form 1.18,32 Consistent with this proposal, H2 is
observed in the 1H NMR spectrum when the solution is cooled
to room temperature.33 The free benzyl isocyanate undergoes
cycloaddition across the Fe−N bond of 1 to form 6Bn, which is
a similar reaction to the previously observed cycloaddition
between 1 and CO2.27 To rigorously confirm that 6Bn is
formed, we independently synthesized 6Bn through the
reaction of 1 and an excess of benzyl isocyanate at room
temperature (Figure 6b). The reaction also produces an
isocyanate trimer and 2 equiv of benzyl isocyanate are required
to obtain the full conversion of 1 to 6Bn. Separation of 6Bn from
the trimer is challenging and sequential crystallizations were
required to generate only small quantities of the pure material.
6Bn was characterized by single-crystal X-ray diffraction, and
the solid-state structure indicates that the geometry around the
iron center is a distorted octahedral (Figure 6c). The 4-
membered ring containing the isocyanate is a kite-like shape.
Both Fe−N bonds are approximately the same length, Fe(1)−
N(1) is 2.057(3) Å and Fe(1)−N(2) is 2.056(3) Å, which is
surprising because the N-donor associated with the isocyanate,
N(1), is formally an X-type donor, while the N-donor
associated with the iPrPNHP ligand, N(2), is formally an L-
type donor. Presumably, in both cases, there is some distortion
from the expected Fe−N bond lengths because the nitrogen
atoms are constrained in a metallacyclic ring.
In catalysis, we propose that carbamates are generated via
the nucleophilic attack of alcohols on free isocyanates. To
probe this process, we performed a series of experiments
between cyclohexyl isocyanate and 1°, 2°, 3°, benzylic, and
phenolic alcohols, both in the presence and absence of 1
(Table 2). Only a small number of alcohols form carbamates in
the absence of 1 and in these cases, low yields are obtained. In
contrast, 1°, 2°, benzylic, and phenolic alcohols all form
carbamates in good yields in the presence of 1, but no
Heating an equilibrium mixture containing 5Cy results in
similar reactivity to that observed with 5Bn. Specifically, after an
equilibrium mixture containing 5Cy was heated for 1.5 h at 100
°C in d8-toluene, 67% of the iron is in the form of 1, while 33%
10619
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
carbamate product is observed with tBuOH, presumably due to
steric factors. This is consistent with our catalytic results. In
reactions between isocyanates and alcohols catalyzed by 1,
cyclohexyl isocyanate likely undergoes a cycloaddition reaction
with 1 to form 6Cy, as this process is facile at room temperature
(vide supra). In an independent reaction, we showed that 6Cy
is a catalyst for carbamate formation from benzyl isocyanate
and cyclohexanol (Figure 8). Importantly, only product 3g,
which arises from a direct reaction between benzyl isocyanate
and cyclohexanol is observed, and there is no evidence for the
formation of the N-cyclohexyl carbamate (3a), suggesting that
6Cy does not lose cyclohexyl isocyanate under these conditions.
The catalytic performance of 6Cy is directly comparable to 1,
although the mechanism through which 6Cy catalyzes the
coupling of the isocyanate and alcohol is unclear. Most likely,
despite being coordinatively saturated, 6Cy is able to facilitate
the reaction by acting as a Lewis acid, as it is known that Lewis
acids can catalyze the formation of carbamates from alcohols
and isocyanates.9 Interestingly, our results on the reaction of
isocyanates and alcohols indicate that carbamate generation
occurs rapidly at room temperature rather than at the elevated
temperatures (150 °C) required for catalysis. Based on the fact
that formamide dehydrogenation to isocyanate also occurs at
lower temperatures than the catalytic reaction (Figure 6), this
suggests that the high temperatures required in catalysis are to
liberate the isocyanate that is trapped in the form of the
cycloaddition product, 6, to allow 1 to catalyze further
formamide dehydrogenation.35
carbamates from formamides and alcohols shown in Figure 9.
The productive pathway for carbamate synthesis catalyzed by 1
involves steps A (formamide dehydrogenation) and B
(isocyanate and alcohol coupling), although for some
substrates the catalyst is likely not required for step B.
Numbered steps i−iv are equilibria or side reactions that we
have identified. Initially, 1 dehydrogenates the formamide to
form an isocyanate and 2 (step A). As evidenced from our
stoichiometric studies (vide supra), this step is slower than
step B (isocyanate and alcohol coupling), although neither are
the turnover-limiting step as the reaction is inhibited by a slow
off-cycle process (vide infra). Prior to dehydrogenation,
formamide can undergo reversible 1,2-addition across the
Fe−N bond of 1 to generate 5 (step i). Although it is possible
that 5 is an intermediate in formamide dehydrogenation, based
on computational precedent from alcohol dehydrogenation
using 1,19b,c it is likely an off-cycle species. In fact, given the
similar polarity of alcohols and formamides, we propose that
the mechanism of the dehydrogenation of formamides is likely
analogous to that previously described for alcohols.19b,c
Specifically, the formamide approaches the iron catalyst so
that the protic N−H of the formamide can interact with the
basic nitrogen of the iPrPNP ligand and the hydridic C−H of
formamide can interact with the iron. In principle, this could
occur in either a stepwise or concerted fashion involving
metal−ligand cooperativity and would result in the eventual
release of the isocyanate and the formation of new Fe−H and
N−H bonds to give 2. We independently demonstrated that
5Me is a kinetically competent precatalyst (see Table S11 in
Supporting Information) and it is noteworthy that at higher
temperatures the equilibrium between 1, NMF, and 5Me shifts
toward 1 and NMF (vide supra), likely making this process less
relevant under the high-temperature conditions used in
catalysis.
Table 2. Carbamate Formation from Isocyanate and Alcohol
with and without 1
Once the isocyanate is formed it can react with 1, which has
been reformed via the 1,2-elimination of H2 from 2, to form a
cycloaddition product, 6 (step ii). This is an off-cycle process
and we independently confirmed that 6Cy is a kinetically
competent precatalyst at 150 °C (see Table S11 in Supporting
Information). Additionally, on the basis of the two NMR
experiments under modified reaction conditions, we propose
that 6 is the catalyst resting state. Specifically, a d8-toluene
solution containing N-cyclohexylformamide, cyclohexanol, and
5 mol % 1 was heated to 150 °C for 1 h. An NMR spectrum
recorded after the sample had returned to room temperature
indicated that 89% of the iron is in the form of 6Cy and
approximately 11% has decomposed to (iPrPNHP)Fe(CO)2
and free iPrPNHP, as typically observed in dehydrogenation
reactions using 1.18,19 After 24 h, the ratio of 6Cy to free
iPrPNHP is approximately 1.3:1 and no (iPrPNHP)Fe(CO)2 is
observed. However, a solid, which is likely particulate iron(0),
is present in the bottom of the reaction flask, which is
indicative of catalyst decomposition. In a separate experiment,
no catalyst yield carbamate 5 mol % 1 yield carbamate
a
a
entry
R−OH
(%)
(%)
1
2
3
4
5
6
7
CyOH
BnOH
0 (10)
21 (22)
0
74 (79)
59 (82)
81
MeOH
1-BuOH
2-BuOH
tBuOH
PhOH
0 (3)
35
0
55 (82)
85
0
9
75
a
NMR yields, which are the average of two trials, were determined
using a naphthalene or hexamethylbenzene standard. Yields in
parentheses were obtained after 1 h. Residual alcohol or isocyanate
were not quantified due to their volatility, which led to partial loss
during workup.
On the basis of our stoichiometric and catalytic studies, we
propose the mechanism for the dehydrogenative synthesis of
Figure 8. Carbamate formation from benzyl isocyanate and cyclohexanol without a catalyst, or with 5 mol % 1 or 6Cy.
10620
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 9. Proposed mechanism for the conversion of formamides and alcohols to carbamates catalyzed by 1.
a
Table 3. Competitive Dehydrogenative Synthesis of Urea, Carbamate, or Thiocarbamate with 1
entry
Cy-YH
temp. (°C)
yield urea (%)
yield carbamate (3a) (%)
yield thiocarbamate (%)
formamide recovered (%)
1
2
3
4
none
120
120
120
150
99
94
0
0
0
90
0
Cy-OH
Cy-SH
Cy-OH
0
0
72
26
a
All yields are the average of two trials. The yield of urea is after isolation, whereas the yields of carbamate, thiocarbamate, and formamide
recovered were determined using NMR spectroscopy in the presence of a naphthalene standard.
Figure 10. Exchange of 1,3-dicyclohexylurea and cyclohexanol with 3a and cyclohexylamine to generate a thermodynamic mixture. The ratio of
urea/carbamate was determined using NMR spectroscopy in the presence of a naphthalene standard.
an identical sample was heated inside an NMR spectrometer to
100 °C.31 After 10 min, the formation of 6Cy was observed and
the corresponding hydride signal continued to grow in
intensity over the course of the reaction. These studies
strongly suggest that 6 is the catalytic resting state.
Information). It is likely that the use of a sealed vessel helps
inhibit this side reaction, as alcohol dehydrogenation catalyzed
by 1 was only reported to occur efficiently in an open system
where nitrogen was constantly being bubbled through the
solution.19b
At the elevated temperatures under which catalysis is
performed, 6 can liberate free isocyanate, which reacts with
the alcohol to form the carbamate (step B). However, in
principle the alcohol can undergo two other processes. In an
analogous fashion to the formamide substrate, the alcohol can
undergo a 1,2-addition reaction across the Fe−N bond of 1 to
form 7 (step iii). At catalytically relevant temperatures, this
process is unlikely to be significant because even at room
temperature the equilibrium for the 1,2-addition of alcohols to
1 often favors 1 and the free alcohol.24 We propose that the
reason the 1,2-addition of formamides is more favorable than
the 1,2-addition of alcohols, even though alcohols are typically
more acidic, is because formamide can form a hydrogen bond
with the ligand, which stabilizes the product. Additionally, the
alcohol can be dehydrogenated to form a ketone or aldehyde
(step iv). In reactions with 2° alcohols, this does not appear to
be a major competing reaction, as in all cases we observe <5%
of the dehydrogenated alcohol. In contrast, in reactions with 1°
alcohols and benzyl alcohol, a larger amount of alcohol
dehydrogenation is observed (see Table S10 in Supporting
Comparative Reactivity of Amines, Alcohols, and
Thiols. In a previous study, we proposed that when
formamides are dehydrogenated by 1 to form isocyanates,
they are trapped by amines to generate ureas.25 We were
interested in understanding the relative reactivity of amines,
alcohols, and also thiols under our catalytic conditions as this
could provide guidance about whether urea, carbamate, or
even thiocarbamate products could be selectively formed.
Initially, a catalytic reaction was performed between N-
cyclohexylformamide and 1 equiv of cyclohexylamine at 120
°C using 5 mol % 1 (Table 3, entry 1). This generates a 99%
yield of 1,3-dicyclohexylurea. The same reaction was then
repeated in the presence of 1 equiv of cyclohexanol (entry 2).
In this case, a 94% yield of 1,3-dicyclohexylurea was observed,
with no evidence for the formation of any carbamate product.
This suggests that the amine more efficiently traps the
isocyanate intermediate than the alcohol. The same competi-
tion experiment was then repeated at 150 °C (entry 4) because
this is the temperature at which we observe a substantial
carbamate formation from formamides and alcohols (Table 1).
10621
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Figure 11. (a) Synthesis of 8 via 1,2-addition of thiols to 1. (b) Solid-state structure of 8Bn with thermal ellipsoids at 30% probability. The
crystallographic model resides on a mirror plane. Hydrogen atoms not bound to iron or nitrogen are omitted for clarity. The iron hydride was
located in the difference map and freely refined. Selected bond lengths (Å) and angles (°): Fe(1)−P(1) 2.2073(3), Fe(1)−C(1) 1.7227(16),
Fe(1)−N(2) 2.0862(14), Fe(1)−S(1) 2.3643(4), C(1)−Fe(1)−N(2) 173.90(7), and P(1)−Fe(1)−P(1)’ 168.777(19).
This reaction produced a 72% yield of 1,3-dicyclohexylurea
and also a 26% yield of 3a. Independent control reactions
indicate that at 150 °C, 1,3-dicyclohexylurea and cyclohexanol
will exchange to form 3a and cyclohexylamine and similarly, 3a
and cyclohexylamine exchange to form 1,3-dicyclohexylurea
and cyclohexanol (Figure 10). In both cases, approximately a
3:1 mixture of 1,3-dicyclohexylurea/3a was observed, indicat-
ing that this is a thermodynamic distribution of products. This
is in good agreement with DFT calculations, which predict that
1,3-dicyclohexylurea and cyclohexanol are 1.10 kcal/mol lower
in energy than 3a and cyclohexylamine. These results indicate
that although selective urea synthesis in the presence of
alcohols is possible, it will be difficult to achieve selective
carbamate synthesis in the presence of amines. Further, they
suggest it will be challenging to develop a one-pot reaction for
the dehydrogenative synthesis of carbamates starting from
amines and alcohols. This is because once the alcohol and
amine have been coupled to form a formamide and then the
formamide dehydrogenated to form an isocyanate, the
isocyanate will selectively react with the amine to generate
urea rather than with the alcohol to generate carbamate. Thus,
our strategy of using isolated formamides is likely the best
route for dehydrogenative carbamate production. Finally, a
mechanistic consequence of the more efficient trapping of
isocyanates with amines compared to alcohols is that the
isocyanate cycloaddition product is not the catalyst resting
state in urea formation because the isocyanate reacts with the
amine before the iron catalyst.25 As a result, catalytic urea
formation occurs at a lower temperature than carbamate
formation.
consistent with hydrides at −17.9 and −18.1 ppm, respectively,
and only one resonance was observed in the 31P NMR spectra
at 94.2 and 93.6 ppm, respectively. 8Bn was characterized by
single-crystal X-ray diffraction (Figure 11b), which indicates a
distorted octahedral geometry around the iron center. A Fe−S
bond distance of 2.3643(4) Å is consistent with the bond
length observed in other phosphine-bound iron sulfur
compounds.41 In contrast to the 1,2-addition of alcohols and
formamides, samples of 8 are stable under vacuum, and the
reaction appears to be irreversible. This is in agreement with
the increased acidity of thiols and indicates that they are not
viable substrates in dehydrogenative catalysis using 1.
CONCLUSIONS
■
We have demonstrated that 1 is an active catalyst for the
dehydrogenative synthesis of carbamates from a variety of N-
alkyl- and N-aryl-substituted formamides, including NMF. The
system is compatible with 1°, 2°, and benzylic alcohols. Our
system is the only first-row transition-metal catalyst for the
dehydrogenative synthesis of carbamates and the only system
compatible with N-alkyl formamides. Mechanistic studies
indicate that the reaction proceeds via a two-step pathway
involving the dehydrogenation of the formamide to an
isocyanate and H2, followed by the reaction of the isocyanate
with an alcohol to generate the carbamate. However, catalytic
performance is hindered by an off-cycle process in which the
active catalyst is sequestered in an inactive form as a result of a
reversible cycloaddition reaction between 1 and the transient
isocyanate. High temperatures are required in catalysis to
release the isocyanate from iron. Competition experiments
indicate that the transient isocyanate that is generated is
preferentially captured by amines to produce ureas rather than
by alcohols to produce carbamates. Thus, the selective
formation of carbamate from formamide and alcohols using
1 is not possible when amines are present. Additionally, it is
not possible to extend our strategy to thiocarbamates as thiols
irreversibly react with 1. On the basis of our results, we
propose that a key factor in developing systems that are
compatible with acidic nucleophiles, such as phenols or thiols,
and are able to operate at lower temperatures, is to design
catalysts that are less basic than 1. This is because the basicity
of the N atom of the iPrPNP ligand in 1 means that it is highly
susceptible to protonation with acidic substrates and off-cycle
cycloaddition reactions with isocyanates. Future work in our
group will investigate modulating the ligand to promote a
wider range of dehydrogenative reactions.
Thiocarbamates are synthetically valuable as they can be
used as herbicides,36 fungicides,37 and pesticides,38 as well as
pharmaceuticals including anesthetics39 and antiviral agents.40
However, an attempt to synthesize a thiocarbamate via a
reaction between N-cyclohexylformamide and cyclohexane-
thiol catalyzed by 1 under the optimized conditions for
dehydrogenative carbamate production did not yield any
thiocarbamate. Further, a competition experiment between N-
cyclohexylformamide and 1 equiv of cyclohexylamine and
cyclohexanethiol at 120 °C using 5 mol % 1 did not generate
either the urea or the thiocarbamate product (Table 3, entry
3). This suggests that the thiol may be causing catalyst
deactivation. Consistent with this observation, a stoichiometric
reaction between 1 and benzylthiol or cyclohexanethiol rapidly
produced the isolable complexes 8Bn or 8Cy, respectively, from
the 1,2-addition of thiol across the Fe−N bond (Figure 11a).
1
The H NMR spectra of 8Bn and 8Cy contained resonances
10622
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
Synthesis, 5th ed.; John Wiley & Sons, Inc.: New Jersey, 2014; pp
895−1193.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acscatal.1c02718.
■
sı
(6) (a) Loev, B.; Kormendy, M. F. An Improved Synthesis of
Carbamates. J. Org. Chem. 1963, 28, 3421−3426. (b) Rand, L.; Thir,
B.; Reegen, S. L.; Frisch, K. C. Kinetics of alcohol-isocyanate reactions
with metal catalysts. J. Appl. Polym. Sci. 1965, 9, 1787−1795.
(c) Frisch, K. C.; Rumao, L. P. Catalysis in Isocyanate Reactions. J.
Macromol. Sci., Part C 1970, 5, 103−149. (d) Raspoet, G.; Nguyen,
M. T.; McGarraghy, M.; Hegarty, A. F. The Alcoholysis Reaction of
Isocyanates Giving Urethanes: Evidence for a Multimolecular
Mechanism. J. Org. Chem. 1998, 63, 6878−6885.
General methods and instrumentation methods; repre-
sentative procedure for catalysis; catalytic optimization;
isolation and characterization of carbamates and iron
complexes; substrates incompatible in catalysis; exper-
imental details for mechanistic studies; X-ray crystallo-
graphic details; and computational details and coor-
dinates (PDF)
(7) Mizuno, T.; Takahashi, J.; Ogawa, A. Synthesis of 2-
Oxazolidones by Sulfur-Assisted Thiocarboxylation with Carbon
Monoxide and Oxidative Cyclization with Molecular Oxygen under
Mild Conditions. Tetrahedron 2002, 58, 7805−7808.
(8) Raucher, S.; Jones, D. S. A Convenient Method for the
Conversion of Amines to Carbamates. Synth. Commun. 1985, 15,
1025−1031.
(9) Ibuka, T.; Chu, G.; Aoyagi, T.; Kitada, K.; Tsukida, T.; Yoneda,
F. A Convenient Lewis Acid Catalyzed Preparation of Carbamates
from Secondary Alcohols and Isocyanates. Chem. Pharm. Bull. 1985,
33, 451−453.
X-ray crystallographic information (CIF)
AUTHOR INFORMATION
Corresponding Authors
Wesley H. Bernskoetter − The Department of Chemistry, The
University of Missouri, Columbia, Missouri 65211, United
States; orcid.org/0000-0003-0738-5946;
■
Email: bernskoetterwh@missouri.edu
Nilay Hazari − The Department of Chemistry, Yale University,
New Haven, Connecticut 06520, United States; orcid.org/
0000-0001-8337-198X; Email: nilay.hazari@yale.edu
(10) Bucher, J. Methyl Isocyanate: A Review of Health Effects
Research Since Bhopal. Fundam. Appl. Toxicol. 1987, 9, 367−379.
(11) Angeles, E.; Santillán, A.; Martínez, I.; Ramírez, A.; Moreno, E.;
Salmón, M.; Martínez, R. A Simple Method for the Synthesis of
Carbamates. Synth. Commun. 1994, 24, 2441−2447.
Authors
Tanya M. Townsend − The Department of Chemistry, Yale
University, New Haven, Connecticut 06520, United States
Brandon Q. Mercado − The Department of Chemistry, Yale
University, New Haven, Connecticut 06520, United States
(12) Pena-López, M.; Neumann, H.; Beller, M. Iron-Catalyzed
̃
Reaction of Urea with Alcohols and Amines: A Safe Alternative for
the Synthesis of Primary Carbamates. ChemSusChem 2016, 9, 2233−
2238.
(13) Chaturvedi, D.; Ray, S. Versatile Use of Carbon Dioxide in the
Synthesis of Carbamates. Curr. Org. Chem. 2006, 137, 127−145.
(14) Ion, A.; Doorslaer, C. V.; Parvulescu, V.; Jacobs, P.; Vos, D. D.
Green Synthesis of Carbamates from CO2, Amines and Alcohols.
Green Chem. 2008, 10, 111−116.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acscatal.1c02718
Notes
The authors declare no competing financial interest.
(15) Bruffaerts, J.; von Wolff, N.; Diskin-Posner, Y.; Ben-David, Y.;
Milstein, D. Formamides as Isocyanate Surrogates: A Mechanistically
Driven Approach to the Development of Atom-Efficient, Selective
Catalytic Syntheses of Ureas, Carbamates, and Heterocycles. J. Am.
Chem. Soc. 2019, 141, 16486−16493.
ACKNOWLEDGMENTS
■
N.H. and W.H.B. acknowledge support from the U.S.
Department of Energy, Office of Science, Basic Energy
Sciences, Catalysis Science Program, under Award DE-
SC0018222. T.M.T. thanks the NSF for support as an NSF
Graduate Research Fellow. All the computational work was
supported by the facilities and staff of the Yale University
Faculty of Arts and Sciences High Performance Computing
Center. We thank Jeremy Weber for assistance in IR
spectroscopy, Dr. Fabian Menges for assistance with mass
spectrometry, and Emily Barth for valuable advice on isolating
carbamates.
(16) Gerack, C.; McElwee-White, L. Formylation of Amines.
Molecules 2014, 19, 7689−7713.
(17) (a) Chirik, P. J.; Gunnoe, T. B. A Meeting of Metals-A Joint
Virtual Issue between Organometallics and ACS Catalysis on First-
Row Transition Metal Complexes. ACS Catal. 2015, 5, 5584−5585.
(b) Furstner, A. Iron Catalysis in Organic Synthesis: A Critical
̈
Assessment of What It Takes To Make This Base Metal a
Multitasking Champion. ACS Cent. Sci. 2016, 2, 778−789.
(c) Egorova, K. S.; Ananikov, V. P. Toxicity of Metal Compounds:
Knowledge and Myths. Organometallics 2017, 36, 4071−4090.
(18) Bielinski, E. A.; Lagaditis, P. O.; Zhang, Y.; Mercado, B. Q.;
REFERENCES
■
Wurtele, C.; Bernskoetter, W. H.; Hazari, N.; Schneider, S. Lewis
̈
(1) Ulrich, H. Urethane Polymers. Kirk-Othmer Encyclopedia of
Chemical Technology; John Wiley & Sons, Inc.: New York, 2006.
(2) (a) Ray, S.; Chaturvedi, D. Application of organic carbamates in
drug design. Part 1:anticancer agents - recent reports. Drugs Future
2004, 29, 343−357. (b) Ray, S.; Pathak, S. R.; Chaturvedi, D. Organic
carbamates in drug development. Part II: antimicrobial agents -
Recent reports. Drugs Future 2005, 30, 161−180.
Acid-Assisted Formic Acid Dehydrogenation Using a Pincer-
Supported Iron Catalyst. J. Am. Chem. Soc. 2014, 136, 10234−10237.
(19) (a) Alberico, E.; Sponholz, P.; Cordes, C.; Nielsen, M.; Drexler,
H.-J.; Baumann, W.; Junge, H.; Beller, M. Selective Hydrogen
Production from Methanol with a Defined Iron Pincer Catalyst under
Mild Conditions. Angew. Chem., Int. Ed. 2013, 52, 14162−14166.
(b) Chakraborty, S.; Lagaditis, P. O.; Förster, M.; Bielinski, E. A.;
Hazari, N.; Holthausen, M. C.; Jones, W. D.; Schneider, S. Well-
Defined Iron Catalysts for the Acceptorless Reversible Dehydrogen-
ation-Hydrogenation of Alcohols and Ketones. ACS Catal. 2014, 4,
3994−4003. (c) Bielinski, E. A.; Förster, M.; Zhang, Y.; Bernskoetter,
W. H.; Hazari, N.; Holthausen, M. C. Base-Free Methanol
Dehydrogenation Using a Pincer-Supported Iron Compound and
Lewis Acid Co-catalyst. ACS Catal. 2015, 5, 2404−2415.
(3) Singh, S.; Singh, N.; Kumar, V.; Datta, S.; Wani, A. B.; Singh, D.;
Singh, K.; Singh, J. Toxicity, Monitoring and Biodegradation of the
Fungicide Carbendazim. Environ. Chem. Lett. 2016, 14, 317.
(4) Lewis, K. A.; Tzilivakis, J.; Warner, D. J.; Green, A. An
International Database for Pesticide Risk Assessments and Manage-
ment. Hum. Ecol. Risk Assess. 2016, 22, 1050−1064.
(5) Wuts, P. G. M. Protection for the Amino Group. Chapter 7:
Protection for the Amino Group in Greene’s Protective Groups in Organic
10623
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624

ACS Catalysis
pubs.acs.org/acscatalysis
Research Article
(20) Sharninghausen, L. S.; Mercado, B. Q.; Crabtree, R. H.; Hazari,
N. Selective Conversion of Glycerol to Lactic Acid with Iron Pincer
Precatalysts. Chem. Commun. 2015, 51, 16201−16204.
Certain N-Carboxymethyl Dithiocarbamic Acid Derivatives. Ann.
Appl. Biol. 1966, 57, 33−51. (b) Erian, A. W.; Sherif, S. M. The
Chemistry of Thiocyanic Esters. Tetrahedron 1999, 55, 7957−8024.
(38) (a) Chin-Hsien, W. Phase-Transfer-Catalysed Preparation ofS-
Alkyl Thiocarbamates. Synthesis 1981, 622−623. (b) Kochansky, J.;
Cohen, C. F. Insecticidal Activity of Dialkyl Carbamates and
Thiocarbamates. J. Agric. Entomol. 1990, 7, 293−304.
(21) Chakraborty, S.; Brennessel, W. W.; Jones, W. D. A Molecular
Iron Catalyst for the Acceptorless Dehydrogenation and Hydro-
genation of N-Heterocycles. J. Am. Chem. Soc. 2014, 136, 8564−8567.
(22) Gluer, A.; Förster, M.; Celinski, V. R.; Schmedt auf der, J.;
̈
(39) Wood, T. F.; Gardner, J. H. The Synthesis of Some
Dialkylaminoalkyl Arylthiourethans and Thioureas1. J. Am. Chem.
Soc. 1941, 63, 2741−2742.
Holthausen, M. C.; Schneider, S. Highly Active Iron Catalyst for
Ammonia Borane Dehydrocoupling at Room Temperature. ACS
Catal. 2015, 5, 7214−7217.
(40) Goel, A.; Mazur, S. J.; Fattah, R. J.; Hartman, T. L.; Turpin, J.
A.; Huang, M.; Rice, W. G.; Appella, E.; Inman, J. K. Benzamide-
Based Thiolcarbamates: A New Class of HIV-1 NCp7 Inhibitors.
Bioorg. Med. Chem. Lett. 2002, 12, 767−770.
(23) Pena-López, M.; Neumann, H.; Beller, M. Iron(II) Pincer-
̃
Catalyzed Synthesis of Lactones and Lactams through a Versatile
Dehydrogenative Domino Sequence. ChemCatChem 2015, 7, 865−
871.
(41) (a) Carroll, M. E.; Chen, J.; Gray, D. E.; Lansing, J. C.;
Rauchfuss, T. B.; Schilter, D.; Volkers, P. I.; Wilson, S. R. Ferrous
Carbonyl Dithiolates as Precursors to FeFe, FeCo, and FeMn
Carbonyl Dithiolates. Organometallics 2014, 33, 858−867. (b) Takács,
J.; Markó, L.; Párkányi, L. Synthesis and Some Reactions of
Fe(CO)2(dppe)(SR)2 and Fe3(CO)6(SAr)6 Complexes. The Crystal
Structure of cis,cis,cis-Fe(CO)2(dppe)(SPh)2. J. Organomet. Chem.
1989, 361, 109−116. (c) Wander, S. A.; Reibenspies, J. H.; Kim, J. S.;
Darensbourg, M. Y. Synthesis and Characterization of a Stable
Iron(II) Hydride-Thiolate Complex: (PhS)Fe(H)(CO)2(P(OPh)3)2.
Inorg. Chem. 1994, 33, 1421−1426.
(24) Lane, E. M.; Uttley, K. B.; Hazari, N.; Bernskoetter, W. Iron-
Catalyzed Amide Formation from the Dehydrogenative Coupling of
Alcohols and Secondary Amines. Organometallics 2017, 36, 2020−
2025.
(25) Lane, E. M.; Hazari, N.; Bernskoetter, W. H. Iron-Catalyzed
Urea Synthesis: Dehydrogenative Coupling of Methanol and Amines.
Chem. Sci. 2018, 9, 4003−4008.
(26) In agreement with the concentration of H2 in solution affecting
the rate of catalysis, a decrease in carbamate yield is observed as the
size of the reaction vessel decreases (see Table S6 in Supporting
Information).
(27) Zhang, Y.; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.;
Williard, P. G.; Mercado, B. Q.; Hazari, N.; Bernskoetter, W. H. Iron
Catalyzed CO2 Hydrogenation to Formate Enhanced by Lewis Acid
Co-Catalysts. Chem. Sci. 2015, 6, 4291−4299.
(28) Jayarathne, U.; Zhang, Y.; Hazari, N.; Bernskoetter, W. H.
Selective Iron-Catalyzed Deaminative Hydrogenation of Amides.
Organometallics 2017, 36, 409−416.
(29) (a) Ballinger, P.; Long, F. A. Acid Ionization Constants of
Alcohols. II. Acidities of Some Substituted Methanols and Related
Compounds1,2. J. Am. Chem. Soc. 1960, 82, 795−798. (b) Riddick, J.
A.; Bunger, W. B.; Sakano, T. K. Organic Solvents: Physical Properties
and Methods of Purification, 4th ed.; John Wiley & Sons, Inc.: New
York, 1986; p 1343. (c) Desai, S. D.; Kirsch, L. E. The Ortho Effect
on the Acidic and Alkaline Hydrolysis of Substituted Formanilides.
Int. J. Chem. Kinet. 2015, 47, 471−488.
(30) Koehne, I.; Schmeier, T. J.; Bielinski, E. A.; Pan, C. J.; Lagaditis,
̈
P. O.; Bernskoetter, W. H.; Takase, M. K.; Wurtele, C.; Hazari, N.;
Schneider, S. Synthesis and Structure of Six-Coordinate Iron
Borohydride Complexes Supported by PNP Ligands. Inorg. Chem.
2014, 53, 2133−2143.
(31) Due to limitations on the temperature that can be reached
inside the NMR probe, this experiment was performed at 100 °C,
even though catalysis is performed at 150 °C.
(32) Chakraborty, S.; Dai, H.; Bhattacharya, P.; Fairweather, N. T.;
Gibson, M. S.; Krause, J. A.; Guan, H. Iron-Based Catalysts for the
Hydrogenation of Esters to Alcohols. J. Am. Chem. Soc. 2014, 136,
7869−7872.
(33) Simnick, J. J.; Sebastian, H. M.; Lin, H.-M.; Chao, K.-C.
Solubility of Hydrogen in Toluene at Elevated Temperatures and
Pressures. J. Chem. Eng. Data 1978, 23, 339−340. At elevated
temperature we assume that H2 is not observed in the 1H NMR
spectrum because of the reduced solubility of H2 as temperature
increases. See for example:
(34) The benzyl isocyanate trimer is also formed in this reaction.
(35) Based on our observation that 6 is an active catalyst for the
formation of carbamates from alcohols and isocyanates, we propose
that the generation of 6 only hinders the 1st step in catalysis, the
dehydrogenation of formamide to isocyanate.
(36) Chen, Y. S.; Schuphan, I.; Casida, J. E. S-Chloroallyl
thiocarbamate herbicides: mouse hepatic microsomal oxygenase and
rat metabolism of cis- and trans-[14C:O]diallate. J. Agric. Food Chem.
1979, 27, 709−712.
(37) (a) Heyns, A. J.; Carter, G. A.; Rothwell, K.; Wain, R. L.
Investigations on Fungicides: The Systemic Fungicidal Activity of
10624
https://doi.org/10.1021/acscatal.1c02718
ACS Catal. 2021, 11, 10614−10624