Highly Efficient and Selective N-Formylation of Amines with CO2 and H2 Catalyzed by Porous Organometallic Polymers

Article abstract of DOI:10.1002/anie.202011260

The valorization of carbon dioxide (CO₂) to fine chemicals is one of the most promising approaches for CO₂ capture and utilization. Herein we demonstrated a series of porous organometallic polymers could be employed as highly efficient and recyclable catalysts for this purpose. Synergetic effects of specific surface area, iridium content, and CO₂ adsorption capability are crucial to achieve excellent selectivity and yields towards N-formylation of diverse amines with CO₂ and H₂ under mild reaction conditions even at 20 ppm catalyst loading. Density functional theory calculations revealed not only a redox-neutral catalytic pathway but also a new plausible mechanism with the incorporation of the key intermediate formic acid via a proton-relay process. Remarkably, a record turnover number (TON=1.58×10⁶) was achieved in the synthesis of N,N-dimethylformamide (DMF), and the solid catalysts can be reused up to 12 runs, highlighting their practical potential in industry.

Full text of DOI:10.1002/anie.202011260

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: Highly Efficient and Selective N-Formylation of Amines with CO2
and H2 Catalyzed by Porous Organometallic Polymers
Authors: Yajing Shen, Qingshu Zheng, Zhe-Ning Chen, Daheng Wen,
James H. Clark, Xin Xu, and Tao Tu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202011260
Link to VoR: https://doi.org/10.1002/anie.202011260

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
Highly Efficient and Selective N-Formylation of Amines with CO
and H Catalyzed by Porous Organometallic Polymers
2
2
[
a]
[a]
[b]
[a]
[d]
[a,b]
Yajing Shen, Qingshu Zheng, Zhe-Ning Chen, Daheng Wen, James H. Clark, Xin Xu* and
Tao Tu*[a,c]
[
a]
Dr. Y. Shen, Dr. Q. Zheng, D. Wen, Prof. Dr. X. Xu and Prof. Dr. T. Tu
Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials
Department of Chemistry, Fudan University
2005 Songhu Road, Shanghai 200438, China
E-mail: taotu@fudan.edu.cn; xxchem@fudan.edu.cn
Z.-N. Chen, Prof. Dr. X. Xu
[
b]
Collaborative Innovation Center of Chemistry for Energy Materials
MOE Laboratory for Computational Physical Science, Fudan University
2005 Songhu Road, Shanghai 200438, China
E-mail: xxchem@fudan.edu.cn
Prof. Dr. T. Tu
College of Chemistry and Molecular Engineering
Zhengzhou University
[
[
c]
d]
100 Kexue Avenue, Zhengzhou 450001, China
Prof. Dr. J. H. Clark
Green Chemistry Centre of Excellence
University of York
York YO105DD, UK
Supporting information for this article is given via a link at the end of the document.
Abstract: The valorization of carbon dioxide (CO
2
) to fine chemicals
capture and
produce formamides is of great importance in pharmaceuticals,
pesticides, herbicides, and functional materials.[4] N,N-
dimethylformamide (DMF) is the best known formamide and is
widely used as a powerful polar aprotic solvent with rapidly
increasing global annual market demands. In industry, DMF is
mainly produced by a strong-base-catalyzed N-formylation of CO
gas with dimethyl amine.[5] How to replace CO with more
is one of the most promising approaches for CO
2
utilization. Herein we demonstrated a series of porous organometallic
polymers could be employed as highly efficient and recyclable
catalysts for this purpose. Synergetic effects of specific surface area,
iridium content, and CO
excellent selectivity and yields towards N-formylation of diverse
amines with CO and H under mild reaction conditions even at 20
2
adsorption capability are crucial to achieve
2
2
2
sustainable but inert CO to access DMF and other formamides
ppm catalyst loading. Density functional theory calculations revealed
not only a redox-neutral catalytic pathway but also a new plausible
mechanism with the incorporation of the key intermediate formic acid
has become a field of great interest.
In comparison with the hydroamination of CO with toxic and
2
expensive silane and borane as reductive reagents, transition-
metal-catalyzed N-formylation of various amines by CO and
molecular dihydrogen (H ) is much more appealing (Figure 1). As
early as 1970, Haynes established the first catalytic system for
DMF synthesis from CO and H with a turnover number (TON) of
, 200.[6] A breakthrough on this field was realized by Noyori using
RuCl [P(CH ] as the catalyst, up to 370, 000 TON was
obtained in supercritical CO under 80 atm H
.[7] Subsequently, by
using Ru-complex containing bi-dental phosphine 1,2-
via a proton-relay process. Remarkably, a record turnover number
2
6
(
TON = 1.58 × 10 ) was achieved in the synthesis of N,N-
2
dimethylformamide (DMF), and the solid catalysts can be reused up
to 12 runs, highlighting their practical potential in industry.
2
2
1
[
2
3 3 4
) ]
2
2
Introduction
a
bis(diphenylphosphino)-ethane (dppe), Baiker further increased
TON to 740, 000, however, a supercritical medium at high
pressure (130 atm) was still essential for the transformation. The
most successful and practical homogenous catalytic case until
now was reported by Ding and co-workers.[9] In the presence of a
Global energy consumption of fossil fuel leads to not only
[8]
consistent carbon dioxide (CO
consequences on climate change.[1] To alleviate this global issue,
using CO as a cheap, readily available, abundant and renewable
source and realizing its valorization to produce fine and bulky
2
) emission but also detrimental
2
C
1
Ru-pincer-type complex under a relative low CO
2 2
and H pressure
chemicals are highly desirable, since this provide a sustainable
and economical approach for carbon capture and utilization.[2]
However, due to its inherent inertness, industrial processes based
(
35 atm for each) in tetrahydrofuran (THF), up to 71% yield and
up to 599, 000 TON could be achieved (Figure 1a). Although, a
heterogeneous catalytic system benefits from the easy product
isolation and catalyst reusability. The efficiency, recoverability
and stability of the known heterogeneous catalysts[10] are far
below the requirements for commercial utilization even at high
catalyst loadings (Figure 1b).
on the utilization of CO
2
as a synthon is still a challenging task.[2b,
In contrast, hazardous and flammable carbon monoxide (CO) is
3
]
broadly utilized in diverse carbonylation reactions due to its high
reactivity. For instance, hydrogenation of CO with amines to
1
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
2 2
Figure 1. Catalytic N-Formylation of dimethylamine with H and CO with diverse catalysts.
To solve the above problems, by combining the advantages of
homogenous and heterogenous catalytic systems, we have
previously reported a series of linear N-heterocyclic carbene
iridium (NHC–Ir) coordination polymers[11] via the “self-supportingˮ
strategy.[12] Due to the "isolated effect", the possible clusterization
of NHC–Ir molecules are prohibited,[13] and the resulting linear
self-supported catalysts exhibited excellent catalytic activity (up to
(along with formic acid as byproduct), and the catalyst could only
be reused for 5 runs even at high catalyst loadings.
Very recently, we demonstrated that solid porous
organometallic polymers (POMPs) formed by direct knitting of the
robust bis-NHC-Ir complexes exhibited extremely high activity,
selectivity and reusability towards hydrogenation of levulinic acid
to γ–valerolactone, and up to 1,010,000 TONs could be
achieved.[ In light of their tunable hierarchical porous structures,
large specific surface areas, single-site catalytic center within a
robust polymeric matrix, we conceived that POMPs NHC-Ir may
15]
9
(
9% yield, Figure 1b)), wide substrate scope and good reusability
up to 10 times) in the N-formylation reaction.[11a] Unfortunately,
only moderate TONs (730) were achieved, probably due to the
restriction of the extremely low specific surface area and the
nonporous feature within the linear catalyst. Therefore, the design
and fabrication of robust porous catalysts based on viable
homogenous catalysts with a large specific surface areas may
2
be beneficial to the adsorption of CO and the direct N-formylation
of dimethyl amine to DMF. Here in this work, we show that, by
tuning the ratio between the selected bis-NHC-Ir complex and the
crosslinking reagents, the specific surface areas and the porosity
of POMPs-NHC-Ir complexes, as well as their adsorption capacity
2 2
accelerate the adsorption of CO /H and facilitate the mass and
heat transfer, thus leading to a significant enhancement of TONs,
especially for the desired DMF synthesis. During our manuscript
2
for CO ,can be strongly enhanced, leading to excellent selectivity
and activity towards N-formylation of diverse amines with
excellent functional group tolerance under mild reaction
2
preparation, Yoon realized the hydrogenation of CO to DMF
under 80 atm pressure at 140 oC with TON up to 160,000, by
conditions. Remarkably, a record TON (1.58 × 10 ) was achieved
6
14]
using
a
ruthenium-grafted biphosphine porous polymer.[
for DMF synthesis, and the solid POMP-NHC-Ir catalysts can be
reused for up to 12 runs without obvious losses of their catalytic
activity or selectivity even at 20 ppm catalyst loading, highlighting
their practical potential in industry.
However, due to the air and moisture-sensitive nature of the Ru-
biphosphine motif and the randomly grafted Ru on the polymer
matrix, the yield (80%) and selectivity were not very satisfactory
2
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
Table 1. Fabrication of POMP-NHC-Ir 2a-h by direct knitting with different ratios between complex 1 and benzene.
Entry[a]
POMP-NHC-Ir
Benzene [equiv.]
Ir content [wt%][b]
16.12
S
BET [m /g]
2
[c]
CO
23
21
21
21
29
31
32
24
3
Yield [%]
64
2
uptake [cm /g]
1
2
3
4
5
6
7
8
2a
2b
2c
2d
2e
2f
0
19
1/9
1/6
1/3
3
17.35
10
63
15.88
13
75
13.86
20
78
7.67
450
656
660
704
90
6
4.51
89
2g
2h
9
3.16
90
15
2.89
85
[
a] Reaction conditions: a mixture containing 0.1 mmol of bis-NHC-Ir complex 1, benzene, 20 equiv. FeCl
3
, and 20 equiv. FDA in 1 mL of dry dichloroethane were
heated in a sealed Schlenk tube at 80 °C for 24 h under N atmosphere. [b] All the Ir contents of POMPs were calculated based on the ICP-AES analysis and given
2
2
in wt%. [c] BET specific surface areas were calculated based on N adsorption isotherms at 77 K.
microporous within the matrix of POMPs (Figure 2d, Figure
S26).[18] These outcomes demonstrated that the porosity and
Results and Discussion
adsorption capacity of CO could be easily regulated by adjusting
2
the ratio between bis-NHC-Ir complex 1 and benzene during the
knitting process.
In light of its high catalytic activity and stability, bis-NHC-Ir
complex 1 was selected as a precursor to fabricate a series of
POMPs with benzene and formaldehyde dimethyl acetal (FDA) in
the presence of a catalytic amount of the Lewis acid FeCl
3
.[15-16]
According to our previous optimization of this mild one-step
Friedel-Crafts alkylation reaction, the initial amounts of the
crosslinking reagents and Lewis acid catalyst, FDA and FeCl ,
3
were both set as 20 equiv. of bis-NHC-Ir complex 1.[15] Concerning
the possible impact of the ratio between the crosslinking reagent
benzene and complex 1 on their morphology, porous structure,
[
15-17]
their CO
2
adsorption ability of the resulting POMPs,
the
amounts of benzene during the direct knitting were systematically
regulated (0-15 equiv., Table 1). Pleasingly, a series of POMP-
NHC-Ir 2a-h with different porosity and Ir contents were
successfully obtained in good to excellent yields (64%-90%,
Figure 1c and Table 1).
With these solid POMPs in hand, detailed characterizations of
them were then carried out. Transmission electron microscopy
(
TEM) and scanning electron microscope (SEM) revealed that
similar lamellar morphologies with obvious macropores existed in
POMPs-NHC-Ir 2e-g (Figure 2a-b and Figures S1-10). The
porosities of these solids were investigated by nitrogen adsorption
measurements (Table 1, Figure 2c, Figures S12-19). As expected,
the BET specific surface areas of POMP-NHC-Ir 2a-h were
generally increased along with the amounts of benzene added
2
during the knitting process. Up to 704 m /g BET specific surface
areas could be achieved for solid 2h with 15 equiv. benzene
addition (Table 1, entry 8). The pore size distribution calculated
by the quenched solid density functional theory (QSDFT)
indicated the coexistence of both micropores and mesopores
within the matrix of POMPs. And the percentage of micropores
increased with increasing benzene addition (Figure 2d). In light of
Figure 2. (a, b) Transmission electron microscopy (TEM) and high-angle
annular dark-field (HAADF) images of POMP-NHC-Ir 2f with the corresponding
element distribution maps as insets. (c) Nitrogen adsorption and desorption
isotherms of POMPs 2e-2g. (d) The corresponding pore width distribution of
POMPs 2e-2g determined by quenched solid density functional theory (QSDFT).
2
(e) CO adsorption and desorption isotherms of POMPs 2e-2g at 298 K. (f) X-
ray photoelectron spectroscopy (XPS) spectra of Ir(I) in POMPs 2e-2g.
the possible impact of CO
N-formayltion process, the CO
also performed (Figure 2e). Similarly, the CO
2
adsorption capacity of POMPs on the
adsorption measurements were
adsorption capacity
2
The iridium contents of these POMPs 2a-h were further
analyzed by inductively coupled plasma atomic emission
spectroscopy (ICP-AES). As shown in Table 1, more benzene
leads to a lower iridium content within the matrix of POMPs. High-
2
of POMPs increased along with the increase of the benzene
3
amount (Figures S20-27). Up to 32 m /g CO
2
was adsorbed by
solid 2g, which might be attributed to the increased percentage of
3
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
angle annular dark-field scanning transmission electron
microscopy (HAADF-STEM) with energy dispersive X-ray (EDX)
spectroscopy revealed that single-site iridium atoms rather than
metal nanoparticles were observed in the matrix of all fabricated
POMPs (Figures S5, S7, and S10). X-ray photoelectron
spectroscopy (XPS) showed almost identical Ir(I) binding
energies between POMP-NHC-Ir 2e-g and bis-NHC-Ir 1,
indicating that the structure of the homogeneous precursor bis-
NHC-Ir 1 is well preserved after knitting. This conclusion is further
supported by the solid state 13C NMR study of bis-NHC-Ir complex
yields were observed for solids 2c and 2d (78% and 67%,
respectively) with similar specific surface area (13-20 m /g) but
low Ir content (15.88% and 13.86%, respectively). Further
reducing the Ir content by increasing the amount of benzene
2
gradually enhanced the BET specific surface area and CO
2
2
adsorption capacity. As a result, solid 2f (with 656 m /g and 4.51%
of Ir content) exhibited the highest yield (>99% with >99%
selectivity). Inferior catalytic performances were observed for
solids 2g (86%) and 2h (78%) by further reducing the Ir content
and increasing the BET specific surface area and the CO
2
1
’ and POMP-NHC-Ir 2f’, in which the carbene carbon atoms are
adsorption ability. These outcomes clearly indicate that the
synergetic effect of the BET specific surface area, Ir content, and
1
3
labelled ( CNHC, see the Supporting Information for details). In
their 13C NMR spectra in both cases, there is a dramatic
enhancement at 189 ppm as compared to that in the unlabelled
CO
2
adsorption capacity are of crucial importance towards the N-
and CO
formylation of dimethylamine with H
2
2
.
samples, which corresponds to the 13
CNHC signals (Figure S31).
Although further reducing the total pressure (H
and CO
, 40-
2
2
o
Besides the common peaks at 150 ppm, 100 ppm and 36 ppm for
the aromatic rings, COD and methyl groups, respectively, an extra
peak at 48 ppm was found for solid 2f’. This was attributed to the
bridging methylene formed by FDA after knitting, which was
consistent with our previous study.[15] Therefore, increasing the
ratio of benzene during the knitting process had almost no impact
on the chemical environment around the iridium centre of the
NHC-Ir complex. All these observations clearly indicated the
POMP-NHC-Ir 2a-h with tunable porosity, well-defined
coordination structure were readily achieved via direct knitting of
molecular NHC-Ir complex 1. In addition, according to ICP-AES,
XPS (Figure S28) and EDS mapping (Figure S8), the obtained
POMPs after Soxhlet extraction were basically free from Fe.
With structure-defined POMP-NHC-Ir 2a-h possessing
regulated porosity in hand, their catalytic activities towards the
60 atm) or the reaction temperature (100 C) and time (12 hours)
led to significant decreases in the yields of DMF (49-90%, Table
S1, entries 2-5), a 85% yield of DMF was still achieved under the
standard reaction conditions (Table S1, entry 6), when the
catalyst loading of solid 2f was lowered to 0.001 mol% (S/C =
85000). To our delight, further decreasing the catalyst loading
could improve TON values (Table S1, entries 7-9), and a record
TON 1,581,588 was achieved for this challenging transformation,
o
albeit with 80 atm H
9).
2
/CO
2
for 96 hours at 140 C (Table S1, entry
In light of the insolubility of solid POMP-NHC-Ir in DMF, the
reusability of the solid 2f was then tested. The catalyst 2f could
be easily recovered by centrifugation and decantation, then the
recovered solid could be directly used in the next run without any
additional activation steps. Pleasingly, the solid catalysts could be
reused for 12 times with quantitative yield, indicating the
robustness of solid 2f (Figure 3c). The robustness and reusability
of the solid 2f were further supported by the analysis of filtrate and
recovered catalyst via inductively coupled plasma atomic
emission spectroscopy (ICP-AES), TEM, XPS and Fourier
transform infrared spectra (FT-IR) after each run. ICP-AES study
showed a slight iridium loss at the first run (< 0.7% of the original
iridium amount in solid 2f) and a negligible leaching was found in
the subsequent runs (Table S2). The trace amount of metal
leaching during the recycling tests might be attributed to both the
good solubility of DMF for the thimbleful low-molecular-weight
oligomer firmly entrapped in the polymer matrix and insufficient
centrifugation of the trace amount of tiny nano-sized solid
particles of 2f. TEM images showed similar morphologies
between the fresh prepared and recovered solid 2f (Figure S11),
further HAADF-STEM with EDS mapping images revealed no
iridium nanoparticles were formed with the recovered solid 2f after
the 12th runs (Figure S11). FT-IR analysis revealed that the
auxiliary ligand CO could be partially dissociated from the iridium
center after several recycle runs (Figure S30). Similar to the
molecular bis-NHC-Ir complexes, the detachment of CO ligands
did not affect the catalytic efficiency of solid 2f, for which excellnet
yield and selectivity could still be achieved at the 12th run. XPS
revealed that the Ir species in the recovered 2f were still in +1
state.[ The slight negative shift (0.6 eV) of the Ir(I) binding
energy (Figure S29) could be attributed to the ligand
dissociation/exchange of CO.[15]
hydrogenation of CO
2
with dimethylammonium dimethyl-
carbamate to DMF were then explored. As shown in Figure 3a, in
the presence of 0.002 mol% catalyst loading under a CO
partial pressure of 40/40 atm at 120 C for 24 hours, POMP-NHC-
2
/H
2
o
Ir 2a-h exhibited much higher catalytic activity than their
1
homogeneous precursor bis-NHC-Ir complex 1. H NMR and GC
analyses of the reaction mixtures revealed that water was the only
byproduct, in consistency with the previous reports.[9,11a] For
comparison, other known viable homogeneous catalysts 4-6 for
the N-formylation were also tested under the optimized reaction
conditions, and much lower yields were attained (Figure 3a). The
pincer complex PNP-Ru 4[9] only gave out 27% yield of DMF
under the otherwise identical reaction conditions. Its analogue 5
exhibited even lower activity (10% yield). Other commercially
available catalysts, like Vaska’s catalyst 6, resulted in trace
amount of DMF. In the case of our previously reported linear
nonporous self-supported NHC-Ir coordination polymer 3a with a
BET surface area of 8 m /g[11a], only 36% yield was achieved at
such low catalyst loading. All these outcomes indicated that the
preserved chemical environment around the iridium centre, large
specific surface area, hierarchical porous structure of POMP-
NHC-Ir were all necessary to achieve satisfactory outcomes in the
2
2 2
challenging catalytic N-formylation reaction with CO and H .
To gain further insight of the structure-function relationship of
the solid POMP-NHC-Ir-catalyzed N-formylation, the catalytic
performances of solids 2a-h were carefully investigated,
correlating with the iridium content, BET specific surface area,
19]
and CO
2
adsorption capability (Table 1, Figure 3b and 3d). The
The insolubility of the solid POMP-NHC-Ir means that the
catalytic progress of the transformation could be directly
monitored by H NMR spectra without any interference caused by
solids 2a and 2b with high iridium content (16.12% and 17.35%)
but small BET specific surface area (10-19 m /g) produced DMF
2
1
in yields of 87% and 91%, respectively. Significant decreases in
4
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
2 2
Figure 3. (a, b) Catalyst optimization in the N-formylation of dimethylamine to DMF with 20 ppm catalyst loading under 40 bar H and CO
at 120 oC for 24 hours.
(c) Reusability of solid POMP-NHC-Ir 2f (carried out on a 40 mmol scale with 20 ppm catalyst under the optimized reaction conditions). (d) Correlation of the BET
specific surface area, the CO
2
adsorption capacity, and the Ir content within the POMP-NHC-Ir matrix 2a-h.
1
the solid catalyst. According to H NMR analysis (Figure S32), the
yield of DMF after 3 hours was 58% and the possible reaction
intermediate formic acid (HCOOH) could be easily observed. By
extending reaction time, the yield of DMF was gradually increased
along with decreasing HCOOH. After 24 hours, HCOOH was
completely consumed and converted into DMF quantitatively. We
believe there are possibly two steps occurred during the N-
conditions. After addition of the dimethylamine salt, the formed
acid was readily converted to DMF at 120 C in 85% yield even
o
without catalyst addition (Figure 4a). It has to be pointed out,
-1
besides achieving rates as high as 7,524 h , delightedly, up to
227,711 TON could be achieved in the hydrogenation step to
formic acid with DBU as a base. This TON value is much higher
than the previous reports achieved by the viable homogenous and
heterogeneous catalytic systems (TONs: 1,604~24,300, details
see Table S3), further highlighting the catalytic efficiency of solid
2f.
formylation of dimethylamine with H
be the hydrogenation of CO leading to HCOOH formation. And
the second step may be the generated acid directly reacting with
dimethylamine under heating conditions to produce the desired
DMF.
In order to verify our hypothesis, several control experiments
were performed. Initially, the model reaction without amine
addition was carried out under the optimized reaction conditions
and no reaction happened (Figure 4a). Therefore, we believed
that amine might act as not only a reaction partner but also a base
during the transformation. In light of the low nucleophilicity but
strong basicity of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), a
control experiment with DBU instead of dimethyl amine was
carried out under the otherwise identical reaction conditions.
Formic acid HCOOH was successfully obtained, suggesting that
2 2
and CO . The first step may
2
To further understand the activation and transformation of CO
2
in the presence of solid catalyst 2f at the molecular level, we
performed density functional theory (DFT)[20] calculations on the
plausible pathway of N-formylation of dimethylamine by CO
2
and
H
2
to DMF based on
a
catalytic model of active motif
+
[Ir(NHC)
2
(CO)
2
] . According to the control experiments, this N-
formylation is composed of two parts, which might be calculated
separately: i) generation of formic acid from CO and H (Figure
2
2
4b), and ii) production of DMF by reaction of dimethylamine with
the as-generated HCOOH (Figure 4c).[21] Intriguingly, our DFT
calculations disclose that dimethylamine is crucial in the
generation of formic acid. As shown in Figure 4b, the activation of
2
the solid catalyst 2f could realize the activation of CO under basic
5
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
Figure 4. (a) Control experiments i: the reaction were carried out with 20 ppm solid 2f in 2 mL MeOH under 40/40 atm H
2
/CO
2
at 120 oC for 24 h; ii: The reaction
2
was carried out with 20 mmol dimethylamine salt and 154 mmol HCOOH (an equal amount with 40 atm CO in the 125 mL autoclave). (b-c) Computational
investigation on the reaction mechanism. Free energy profiles for the generation of formic acid at first, and DMF subsequently. The computational details are
provided in the Supporting Information. (d) Proposed reaction mechanism.
H
2
is the initial step, where the Ir center and N atom of
with one formic acid molecule is kinetically unfeasible with a free
energy barrier as high as 46.8 kcal/mol (Figure S33b). A proton-
relay pathway that incorporates the second formic acid molecule
can dramatically reduce the intrinsic reorganization energy of the
transition state, which facilitates the reaction kinetics. As formic
acid can be generated from the adduct DMFA after decomposition
by heating, the existence of some extra formic acid molecules to
assist the proton-relay is conceivable. As shown in Figure 4c,
generation of DMF from the N-formylation of formic acid and
dimethylamine is kinetically feasible with the incorporation of the
second formic acid (TS4), where the apparent activation free
energy is reduced to 25.1 kcal/mol.
dimethylamine both act as the hydride acceptors in a synergistic
way. This elementary step is kinetically feasible with a free energy
barrier of 16.8 kcal/mol (TS1). It is worth noting that H activation
2
typically undergoes an oxidative addition pathway, which
generates a NHC-Ir dihydride intermediate, and is unfavorable
with a higher free energy barrier of 23.4 kcal/mol (see TS1’ in
Figure S33a). The generated quaternary ammonium species in
LM1 can further act as a hydride donor in the subsequent
carbonylation with CO
free energy barrier of 19.8 kcal/mol (TS2). Along with the
regeneration of the catalytic species [Ir(NHC) (CO)
]+, the adduct
2
, leading to the HCOOH formation with a
2
2
salt dimethylammonium formate (DMFA) is formed. These
outcomes clearly demonstrate the role of dimethylamine in the
With all the information in hand, a proposed mechanism is
given in Figure 4d. H is first activated by the catalytic center of
2
activation of H
2
and CO
2
, further supporting the observation in the
solid 2f with a synergistic assistance of dimethylamine, while the
control experiments that no formic acid species is detected in the
absence of base. Moreover, it is intriguing to see that, with the
participation of dimethylamine, such a catalytic process is a
redox-neutral reaction, where the Ir(I) center retains its apparent
valence during the whole catalytic cycle. This redox-neutral
feature lends a support to the fact that the POMP-NHC-Ir solids
did not collapse during the N-formylation, highlighting the
structural robustness of these catalysts.
resultant quaternary ammonium species acts as a hydride donor
2
in the subsequent carbonylation with CO , leading to the
[
22]
formation of the key intermediate HCOOH in the adduct form of
DMFA,[
21b]
this pathway is also in accordance with the previous
to formic acid. DMFA further
reports on the hydrogenation of CO
2
reacts with one more HCOOH molecule via proton-relay,
eventually leading to the formation of DMF. This mechanism is
1
supported by the time-dependent H NMR spectra of the reaction
Hypothetically, after the adduct DMFA formation, the target
molecule DMF would be directly produced from the N-formylation
of dimethylamine and formic acid. However, our DFT calculations
suggest that the N-formylation of one dimethylamine molecule
mixture (Figure S32). The DMFA signal firstly appeared at 8.5
ppm (formyl CH), then the peaks at 8.0 ppm corresponding to
HCOOH gradually disappeared along with the increase of DMF
signals (7.9 ppm, formyl CH).
6
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
With the excellent outcomes achieved in the N-formylation of
dimethyl amine salt for DMF production, the compatibility of our
newly developed catalytic system with other amines was then
explored. Pleasingly, both primary and secondary amines reacted
well and the corresponding formamides were achieved in good to
excellent yields. As shown in Table 2, diverse cyclic amines
including morpholine, 1-phenylpiperazine, piperidine and 1,2,3,4-
tetrahydroisoquinoline were all well converted and in quantitative
yields to the corresponding formamides (92-99%, 10-13). In
contrast, a five member ring was slightly less active, so that when
pyrrolidine was employed, only a moderate yield was obtained
carbonyl, amide, alkenyl, as well as ether, proved to be suitable
substrates for the N-formylation with CO and H under the
2 2
standard reaction conditions. Amlodipine, which is used to treat
coronary artery disease and hypertension could be successfully
transformed into the corresponding formamide 24 in 81% yield
with hydrogen sensitive ester, alkenyl and halogen group
completely retained. Another important drug molecular
desloratadine, an antihistamine that has been used to treat the
symptoms of allergies, was also well tolerated in this catalytic
system and a moderate yield was obtained (60%, 25). Other
molecules containing piperazine fragment all delivered the
desired products in excellent yields (85-99%, 26-30), further
demonstrating the high activity, selectivity and broad substrate
scope of our solid molecular catalytic system.
(
68%, 14). High yields were observed in the cases of acyclic
symmetric secondary amines. To our delight, sterically
encumbered amines like diisopropylamine also reacted well to
give excellent yields (e.g. 94%, 15), whereas, only 43% yield was
achieved when N,N-diethyl amine was used (16), probably due to
its low boiling point. An aminoalcohol containing unprotected
hydroxyl group, diethanolamine, was also a suitable substrate
and the desired product was afforded in excellent yield leaving the
hydroxyl groups intact (91%, 17). The substrate scope was readily
extended to primary amines and good to excellent yields could
also be obtained in all the tested cases. Both linear and cyclic
aliphatic primary amines were suitable and readily converted into
the corresponding formamides with satisfactory to excellent yields
Conclusion
In conclusion, a series of porous organometallic polymers
(
POMP-NHC-Ir) with regulated porous structure and enhanced
adsorption capability of CO have been readily fabricated via
2
direct knitting of molecular bis-NHC-Ir complexes by tuning the
ratio of the molecular precursor and the crosslinking reagent
benzene. Synergetic effects of single-site iridium center, specific
(
77-99%, 18-22). Remarkably, the primary amine bearing a
surfance area and adsorption capability of CO
importance in achieving excellent selectivity and yields in the N-
formylation of diverse amines with CO and H under base- and
2
are of critical
heterocyclic substitute was also smoothly transformed into the
formamide 23 in 77% yield. In view of the importance of
aminoalcohols in various fields, several aminoalcohols were also
studied in this transformation also leading to excellent yields (19,
2
2
solvent-free conditions. Primary, secondary amines as well as
some functional molecules containing sensitive groups such as
halogen, ester, carbonyl, amide, alkenyl were are all tolerated.
20), highlighting the high efficiency and good functional group
tolerance of our new protocol.
6
Remarkably, a record TON (1.58× 10 ) was achieved in DMF
Table 2. Reaction scope with primary and secondary amines catalyzed by solid
synthesis by this protocol, and the solid POMP-NHC-Ir catalysts
could be readily recovered without any additional treatments and
reused for up to 12 runs without obvious loss of catalytic activity
or selectivity. Control experiments and DFT calculations not only
reveled a redox-neutral catalytic pathway, confirming the high
activity and stability of the solid POMP-NHC-Ir catalysts, but also
provided a new plausible N-formylation reaction mechanism
where dimethylamine helps to form formic acid, by-passing the
energetically unfavourable oxidative addition process; while the
incorporation of the second formic acid facilitates the final DMF
production via a proton-relay process. All these outcomes and the
excellent performance of the solid molecular catalysts highlighted
their potential for industrial applications.
[
a]
2f.
Acknowledgements
Financial supports from the National Key R&D Program of China
(
Nos. 2016YFA0202902 and 2018YFA0208600), the National
Natural Science Foundation of China (Nos. 21861132002,
1871059, 21572036 and 21688102), and Department of
2
Chemistry, Fudan University are gratefully acknowledged.
Keywords: Carbon dioxide • N-formylation • DMF • Porous
organometallic polymers • N-heterocyclic carbene iridium
complex
[
a] Reactions were carried out with 10 mmol scale: amines were dissolved in 2
o
mL MeOH and stirred at 120 C for 24 h. [b] Isolate yield. [c] Reaction was
carried out with 1 mmol scale: amines were dissolved in 2 mL MeOH and stirred
at 120 C for 24 h. [d] 2 mL THF was used as solvent.
o
[
[
1]
2]
a) D. Shindell, C. J. Smith, Nature 2019, 573, 408-411; b) S. J. Davis, K.
Caldeira, H. D. Matthews, Science 2010, 329, 1330-1333.
Encouraged by these excellent results, we turned our attention
to some functional molecules. To our delight, all the reagents
bearing different functional group such as halogen, ester,
a) J. Wu, T. Sharifi, Y. Gao, T. Zhang, P. M. Ajayan, Adv. Mater. 2019,
31, 1804257; b) J. Artz, T. E. Müller, K. Thenert, J. Kleinekorte, R. Meys,
7
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
A. Sternberg, A. Bardow, W. Leitner, Chem. Rev. 2018, 118, 434-504; c)
M. He, Y. Sun, B. Han, Angew. Chem. Int. Ed. 2013, 52, 9620-9633;
Angew. Chem. 2013, 125, 9798-9812.
[
[
3]
4]
a) J. Klankermayer, S. Wesselbaum, K. Beydoun, W. Leitner, Angew.
Chem. Int. Ed. 2016, 55, 7296-7343; Angew. Chem. 2016, 128, 7416-
7467; b) Q. Liu, L. Wu, R. Jackstell, M. Beller, Nat. Commun. 2015, 6,
5933.
a) P. Ryabchuk, K. Stier, K. Junge, M. P. Checinski, M. Beller, J. Am.
Chem. Soc. 2019, 141, 16923-16929; b) P. G. M. Wuts in Greene's
Protective Groups in Organic Synthesis, 5 ed. (Ed.: P. G. M. Wuts), John
Wiley & Sons, Inc., 2014, p. 991; c) E. Balaraman, C. Gunanathan, J.
Zhang, L. J. W. Shimon, D. Milstein, Nat. Chem. 2011, 3, 609-614.
a) S. Ding, N. Jiao, Angew. Chem. Int. Ed. 2012, 51, 9226-9237; Angew.
Chem. 2012, 124, 9360-9371; b) H. B. H. Kieczka in Ullmann's
Encyclopedia of Industrial Chemistry (Ed.: H. B. H. Kieczka), 2011.
P. Haynes, L. H. Slaugh, J. F. Kohnle, Tetrahedron Lett. 1970, 11, 365-
[
5]
[
[
6]
7]
3
68.
P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1994,
16, 8851-8852.
1
[
[
8]
9]
O. Kröcher, R. A. Köppel, A. Baiker, Chem. Commun. 1997, 453-454.
L. Zhang, Z. Han, X. Zhao, Z. Wang, K. Ding, Angew. Chem. Int. Ed.
2015, 54, 6186-6189; Angew. Chem. 2015, 127, 6284-6287.
[
[
10] a) Q.-Y. Bi, J.-D. Lin, Y.-M. Liu, S.-H. Xie, H.-Y. He, Y. Cao, Chem.
Commun. 2014, 50, 9138-9140; b) X. Cui, Y. Zhang, Y. Deng, F. Shi,
Chem. Commun. 2014, 50, 189-191.
11] a) Y. Zhang, J. Wang, H. Zhu, T. Tu, Chem. Asian J. 2018, 13, 3018-
3021; b) Z. Sun, J. Chen, T. Tu, Green Chem. 2017, 19, 789-794; c) Y.
Liu, Z. Sun, C. Huang, T. Tu, Chem. Asian J. 2017, 12, 355-360; d) Z.
Sun, Y. Liu, J. Chen, C. Huang, T. Tu, ACS Catal. 2015, 5, 6573-6578;
e) S. M. Sarkar, Y. Uozumi, Y. M. A. Yamada, Angew. Chem. Int. Ed.
2011, 50, 9437-9441; Angew. Chem. 2011, 123, 9609-9613.
[
[
12] a) Z. Wang, G. Chen, K. Ding, Chem. Rev. 2009, 109, 322-359; b) K.
Ding, Z. Wang, X. Wang, Y. Liang, X. Wang, Chem. Eur. J. 2006, 12,
5188-5197; c) L.-X. Dai, Angew. Chem. Int. Ed. 2004, 43, 5726-5729;
Angew. Chem. 2004, 116, 5846-5850.
13] a) J. Wu, L. Shen, S. Duan, Z.-N. Chen, Q. Zheng, Y. Liu, Z. Sun, J. H.
Clark, X. Xu, T. Tu, Angew. Chem. Int. Ed. 2020, 59, 13871-13878;
Angew. Chem. 2020, 132, 13975-13982; b) J. Campos, L. S.
Sharninghausen, R. H. Crabtree, D. Balcells, Angew. Chem. Int. Ed.
2014, 53, 12808-12811; Angew. Chem. 2014, 126, 13022-13025; c) L.
S. Sharninghausen, J. Campos, M. G. Manas, R. H. Crabtree, Nat.
Commun. 2014, 5, 5084.
[
14] G. H. Gunasekar, S. Padmanaban, K. Park, K.-D. Jung, S. Yoon,
ChemSusChem 2020, 13, 1735-1739.
[
[
[
15] Y. Shen, Q. Zheng, H. Zhu, T. Tu, Adv. Mater. 2020, 32, 1905950.
16] L. Tan, B. Tan, Chem. Soc. Rev. 2017, 46, 3322-3356.
17] a) J. Li, D. Jia, Z. Guo, Y. Liu, Y. Lyu, Y. Zhou, J. Wang, Green Chem.
2017, 19, 2675-2686; b) W. He, F. Zhang, H. Li, Chem. Sci. 2011, 2, 961-
966.
[
[
18] a) B. Li, Z. Guan, W. Wang, X. Yang, J. Hu, B. Tan, T. Li, Adv. Mater.
012, 24, 3390-3395; b) B. Li, R. Gong, W. Wang, X. Huang, W. Zhang,
2
H. Li, C. Hu, B. Tan, Macromolecules 2011, 44, 2410-2414.
19] a) C. Crotti, E. Farnetti, S. Filipuzzi, M. Stener, E. Zangrando, P. Moras,
Dalton Trans. 2007, 133-142; b) W. J. Louw, D. J. A. De Waal, T. I. A.
Gerber, C. M. Demanet, R. G. Copperthwaite, Inorg. Chem. 1982, 21,
1667-1668.
[
[
20] I. Y. Zhang, X. Xu, Y. Jung, W. A. Goddard, Proc. Natl. Acad. Sci. U. S.
A. 2011, 108, 19896-19900.
21] a) Y. Lu, Z. Liu, J. Guo, S. Qu, R. Zhao, Z.-X. Wang, Chem. Commun.
2017, 53, 12148-12151; b) N. M. Rezayee, C. A. Huff, M. S. Sanford, J.
Am. Chem. Soc. 2015, 137, 1028-1031.
[
22] a) K. Nakamae, T. Nakajima, Y. Ura, Y. Kitagawa, T. Tanase, Angew.
Chem. Int. Ed. 2020, 59, 2262-2267; Angew. Chem. 2020, 132, 2282-
2287; b) M. V. Vollmer, J. Ye, J. C. Linehan, B. J. Graziano, A. Preston,
E. S. Wiedner, C. C. Lu, ACS Catal. 2020, 10, 2459-2470.
…
8
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202011260
RESEARCH ARTICLE
Entry for the Table of Contents
Porous organometallic polymers (POMPs) with controllable porosity and enhanced CO
direct knitting strategy. Excellent selectivity and activity with a record turnover number (TON = 1.58 × 10 ) in the N-formylation
2
adsorption capability were fabricated via the
6
reaction with CO
2
and H
2
to N,N-dimethylformamide (DMF) were achieved under mild reaction conditions, and the POMPs could be
reused up to 12 runs.
9
This article is protected by copyright. All rights reserved.