Heterogeneous Cobalt-Catalyzed Direct N-Formylation of Isoquinolines with CO2 and H2

Article abstract of DOI:10.1002/cctc.201601682

Isoquinolines (IQs) are an abundant feedstock, and N-formyl-1,2,3,4-tetrahydroisoquinolines (FTHIQs) are valuable fine chemicals and key intermediates. Herein, we report for the first time the Co⁰/ZnCl₂-catalyzed direct N-formylation of IQs by using CO₂ with H₂ to produce FTHIQs. It was discovered that the Co catalyst and ZnCl₂ worked synergistically in catalyzing the N-formylation reactions, and moderate to high yields of the desired products could be obtained, depending on the nature of the substrates. The Co⁰ catalyst could be reused at least five times without a notable decrease in activity. A possible reaction mechanism is proposed on the basis of control experiments.

Full text of DOI:10.1002/cctc.201601682

DOI: 10.1002/cctc.201601682
Full Papers
Heterogeneous Cobalt-Catalyzed Direct N-Formylation of
Isoquinolines with CO2 and H2
Zhenhong He,[a] Hangyu Liu,[a, b] Huizhen Liu,*[a, b] Qingli Qian,[a] Qinglei Meng,[a]
Qingqing Mei,[a, b] and Buxing Han*[a, b]
Isoquinolines (IQs) are an abundant feedstock, and N-formyl-
1,2,3,4-tetrahydroisoquinolines (FTHIQs) are valuable fine
chemicals and key intermediates. Herein, we report for the first
time the Co0/ZnCl2-catalyzed direct N-formylation of IQs by
using CO2 with H2 to produce FTHIQs. It was discovered that
the Co catalyst and ZnCl2 worked synergistically in catalyzing
the N-formylation reactions, and moderate to high yields of
the desired products could be obtained, depending on the
nature of the substrates. The Co0 catalyst could be reused at
least five times without a notable decrease in activity. A possi-
ble reaction mechanism is proposed on the basis of control
experiments.
Introduction
N-formyl-1,2,3,4-tetrahydroisoquinoline and its derivatives
(FTHIQs) are important fine chemicals and key intermediates
for the synthesis of medical intermediates and alkaloids.[1] Gen-
erally, classical approaches for the synthesis of FTHIQs involve
N-formylations of 1,2,3,4-tetrahydroisoquinolines (THIQs) with
C1 sources, such as CO, CO2, CH3OH, HCHO, or DMF.[2] Synthesis
of FTHIQs directly from isoquinolines (IQs) is attractive[3] be-
cause IQs are abundant, are present in coal and oil shale, and
are the byproducts of petroleum refining processes.[4] The
transformation of IQs into value-added products is an attrac-
tive route for efficient utilization of carbon resources.[5]
synthesis of FTHIQs that would involve selective hydrogenation
of IQs and N-formylation of THIQs; however, it has not been re-
ported. The selective hydrogenation of the heteroaromatic
compounds, isoquinolines, is difficult because of the high reso-
nance stability of these substrates and the potential poisoning
of the catalysts by either the substrates or the hydrogenated
products. To improve the reaction, various strategies have
been developed,[8] including the activation of N-heteroaromatic
compounds by formation of their salts over various homoge-
neous precious Ir metal complex catalysts.[9] In addition, several
heterogeneous catalysts have also been developed for the se-
lective hydrogenation of IQs.[10]
CO2 is an economical, abundant, and renewable C1 building
block for the production of value-added chemicals.[6] N-Formy-
lation of various amines has been studied widely with CO2 as
the C1 source and H2 or PhSiH3 as the reducing agent in recent
years.[7] Hydrogen is a commonly used reducing agent because
it is abundant, economical, and non-toxic and the only byprod-
uct is H2O. Ding and co-workers have reported the N-formyla-
tion of a series of saturated amines with H2 and CO2 catalyzed
by a ruthenium-based homogeneous catalyst.[7l] Shi and co-
workers have found that a heterogeneous palladium catalyst
was also active for this kind of reactions.[7j] The N-formylation
of IQs with CO2 and H2 would be an ideal route for the
Herein, we report the direct N-formylation of IQs to form
FTHIQs by using CO2 and H2 for the first time. It was found
that a Co0 catalyst, which is abundant and relatively low
cost,[11] could catalyze the reaction efficiently in the presence
of ZnCl2. A detailed study indicated that the Co0 catalyst and
ZnCl2 work synergistically in promoting the reaction by the co-
ordination of ZnCl2 and the nitrogen atom of the IQs.
Results and Discussion
We initially studied the N-formylation of isoquinoline (1a) with
CO2 and H2 to form N-formyl-1,2,3,4-tetrahydroisoquinoline
(3a) and screened the catalytic system. The Co0 catalyst was
prepared by reduction of Co3O4 with H2 at 4008C for 2 h (the
details are given in the Experimental Section). The reduction
temperature for catalyst preparation was determined by tem-
perature-programmed reduction (TPR) measurement (Fig-
ure 1a), in which the peaks at around 231 and 3578C could be
attributed to the reduction of Co3O4 to CoO and CoO to Co, re-
spectively.[12] The results indicated that, after reduction, the Co
species had been reduced to Co0, which was confirmed by
characterization with X-ray diffraction (XRD) and X-ray photo-
electron spectroscopy (XPS), as illustrated in Figure 1. In the
XRD patterns (Figure 1c) of the Co3O4 catalyst, the peaks at 19,
[a] Dr. Z. He, H. Liu, Prof. Dr. H. Liu, Dr. Q. Qian, Dr. Q. Meng, Q. Mei,
Prof. Dr. B. Han
Beijing National Laboratory for Molecular Sciences (BNLMS)
Institute of Chemistry
Chinese Academy of Sciences
100190 Beijing (China)
Fax: (+86)106-256-282-1
E-mail: liuhz@iccas.ac.cn
hanbx@iccas.ac.cn
[b] H. Liu, Prof. Dr. H. Liu, Q. Mei, Prof. Dr. B. Han
University of Chinese Academy of Sciences
Beijing (China)
Supporting information for this article can be found under:
https://doi.org/10.1002/cctc.201601682.
ChemCatChem 2017, 9, 1 – 7
1
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
that the particle size of the catalyst was in the range of 20–
100 nm (Figure 1e).
The conversion of 1a was very low and the main product
was 1,2,3,4-tetrahydroisoquinoline (2a) if only Co0 catalyst was
used (Table 1, entry 1). Desired product 3a was not detected if
pure ZnCl2 was used as the catalyst (Table 1, entry 2). With
a combination of Co0 catalyst and ZnCl2 (Co0/ZnCl2), the con-
version of 1a and the yield of 3a could reach 97 and 94%, re-
spectively (Table 1, entry 3). If MnCl2 was used, the conversion
of 1a was 85% with a 30% yield of 3a (Table 1, entry 4). The
conversions of 1a were 52 and 58% in the presence of KCl
and LiCl, respectively (Table 1, entries 5 and 6), which were
higher values than that observed with pure Co0 catalyst. How-
ever, in these cases, the yields of 3a were still low. Addition of
NiCl2, CuCl2, or AlCl3 resulted in lower conversions than that
with pure Co0 catalyst, which indicated that these salts inhibit
the reaction remarkably (Table 1, entries 7–9). The substrate
conversion could reach up to 96 and 95% in the presence of
Zn(acac)2 and Zn(NO3)2, respectively, and the yields of 3a were
69 and 51%, respectively (Table 1, entries 10 and 11). The re-
sults indicated that Zn(acac)2 and Zn(NO3)2 could promote the
selective hydrogenation of 1a effectively, whereas their promo-
tion effect for the N-formylation was weaker than that of ZnCl2
(Table 1, entries 3, 10, and 11). If ZnI2 was used, the conversion
of 1a and the yields of 2a and 3a were all low (Table 1,
entry 12), which indicated that the reaction was remarkably
suppressed by the iodine ion. The poor performance was pos-
sibly a result of strong coordination between IÀ ions and the
Figure 1. a) TPR test of Co3O4. c) XRD patterns of Co3O4, Co0 catalyst, and
Co0 catalyst after reuse five times. b) and d) XPS spectra of the Co0 catalyst
and Co3O4, respectively. e) TEM image of the Co0 catalyst. f) TEM image of
the Co0 catalyst after reuse five times.
Table 1. N-Formylation of isoquinoline with CO2 and H2 over various catalysts.[a]
31, 36.9, 38.5, 44.8, 55.6, 59.2, and 65.28 can be as-
signed to the (111), (220), (1111), (2212), (400),
(422), (511), and (440) indices of Co3O4 (powder dif-
fraction file no. 43-1003), respectively. The peaks at
41.6, 44.4, 47.4, and 75.98 for the Co0 catalyst can be
attributed to the (100), (002), (101), and (110) indi-
ces of Co0 (powder diffraction file no. 05-0727).[6e,13]
These results indicated that the Co species was re-
duced to Co0. The XRD spectrum of the Co0 catalyst
after it had been reused five times indicated that the
main crystalline phase did not change significantly.
Broad peaks appeared at around 238 because of the
amorphous structure. In the XPS spectrum of Co3O4,
the peak at 780 eV is attributed to the Co3+ 2p3/2
configuration and a small peak at around 781.5 eV is
ascribed to the Co2+ 2p3/2 configuration. The other
spin-orbit component, 2p1/2, appears at 794.9 and
796.8 eV for Co3+ and Co2+, respectively. The two
peaks at 786.5 and 803.3 eV are Co2+ shake-up satel-
lite peaks of Co3O4 (Figure 1d). In Figure 1b, the
peaks at 779.2 and 794.6 eV can be assigned to the
Co0 2p3/2 and Co0 2p1/2 configurations, respectively.[14]
The results indicated that, after reduction with hy-
drogen at 4008C, the Co0 species was formed. The
CoII/III species were a result of oxidation during the
catalyst preparation and characterization. The trans-
mission electron microscopy (TEM) image showed
Entry
Cat.
Additive
Conversion
[%]
Yield[b] [%]
2a
3a
4a
1
2
3
4
5
6
7
8
Co
–
Co
Co
Co
Co
Co
Co
Co
Co
Co
Co
–
39
0
34
0
2
54
44
49
6
5
0
94
30
8
9
12
6
0
ZnCl2
ZnCl2
MnCl2
KCl
0
<1
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
97
85
52
58
18
13
11
96
95
40
74
0
0
12
0
0
0
LiCl
NiCl2
CuCl2
AlCl3
Zn(acac)2
Zn(NO3)2
ZnI2
ZnCl2
ZnCl2
ZnCl2
ZnCl2
ZnCl2
ZnCl2
ZnCl2
–
7
7
9
4
[c]
10
11
12
13[d]
14
15
16
17
18
19
20
21
26
44
26
42
0
0
9
0
0
69
51
14
32
0
0
3
0
0
Co
CoO
Co3O4
Fe
Mn
Cu
In
0
31
40
0
53
48
0
0
0
Co/ZnO
Co
84
88
ZnCl2·nH2O
[a] Reaction conditions: substrate (0.5 mmol), catalyst (5 mg), additive (0.02 mmol),
CO2 (2 MPa), H2 (8 MPa), THF (2 mL), 1708C, 16 h. [b] Yield was determined by GC with
toluene as an internal standard. [c] acac: acetyl acetone. [d] ZnCl2 (0.01 mmol).
&
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
active Co sites and/or Zn2+ ions. These results demonstrated
that the counterions of zinc salts affected the catalytic per-
formances greatly; this is probably related to their different co-
None of the desired product 3a was detected in the absence
of CO2, which indicated that CO2 was the C1 source for the re-
action (Table 2, entry 1). If the H2 pressure was 8 MPa, the yield
of 3a increased remarkably from 43 to 71% with an increase
in CO2 pressure from 0.5 to 1 MPa, respectively (Table 2, en-
tries 2 and 3). In addition, the pressure of H2 had a great
impact on the conversion of 1a and the yield of 3a, both of
which increased with increasing H2 pressure (Table 2, entries 5,
6 and 4). The effect of reaction time on the catalytic per-
formance was studied as well. Only 12% yield of 3a was gen-
erated with a time of 4 h. This yield increased to 67 and 86%
after the time was extended to 8 and 12 h. At the same time,
the hydrogenated product 2a increased from 20 to 28% and
then reduced to 12% (Table 2, entries 8–10). The results re-
vealed that 2a could be transformed into 3a, and the 3a yield
was thus increased by prolonging the reaction time. Based on
the above results, we speculated that the N-formylation of iso-
quinoline with CO2 and H2 occurs in two steps, that is, the se-
lective hydrogenation of 1a to 2a and the N-formylation of 2a
with CO2 and H2 to 3a. Fortunately, no N-formyl-decahydroiso-
quinoline or decahydroisoquinoline were detected under the
reaction conditions we studied. The results indicated that Co0/
ZnCl2 exhibited high selectivity for the N-formylation of
isoquinoline.
To test the versatility of the Co0/ZnCl2 catalytic system, some
typical substituted IQs with electron-donating and electron-
withdrawing groups were studied and the results are given in
Table 3. The IQs with electron-donating groups, namely 5-me-
thoxy- and 5- and 6-methylisoquinolines, were effectively con-
verted into the desired products, with yields of 98, 88, and
92%, respectively (Table 3, entries 1–3). 3-Methyl-isoquinoline
(1e) was converted into 3e with 79% yield, whereas 1-methyl
isoquinoline (1 f) afforded 3 f with a relatively lower yield even
after 48 h (Table 3, entry 5), which can probably be attributed
to steric hindrance from the methyl group at the 1 position.
The chlorine-substituted IQs 1g and 1h could be formylated
to form the corresponding products after 36 h with yields of
83 and 63%, respectively (Table 3, entries 6 and 7). The results
indicated that the Co0/ZnCl2 catalytic system was compatible
for substituted IQs with electron-donating and electron-with-
drawing groups. The isoquinoline-N-oxide (1i) was converted
into 3a with a high yield of 99% (Table 3, entry 8), which indi-
cated that the oxidized substrate did not affect the catalytic
performance and this approach is not sensitive to oxygen.
However, the activity of Co0/ZnCl2 for the direct N-formylation
of quinoline was very low, and only 7.8% yield of N-formyl-
1,2,3,4-tetrahydroquinoline (3j) was obtained, which is lower
than that of isoquinoline (Table 3, entry 9).
1
ordination abilities and was proved by the H NMR spectra in
Figures S1–S3 in the Supporting Information. The above results
showed that Co0 and ZnCl2, and/or isoquinolines, had an excel-
lent synergistic effect for catalyzing the direct N-formylation of
isoquinolines by using CO2 with H2.
The ZnCl2 loading affected the reaction dramatically. If
2 mol% of ZnCl2 was added, 32% yield of 3a was obtained,
which was increased to 94% by increasing the loading to
4 mol% (Table 1, entries 13 versus 3). No product was detected
over CoO or Co3O4 (Table 1, entries 14 and 15), which indicated
that the Co0 species was the active component for the reac-
tion. Combinations of ZnCl2 with Fe0, Cu0, Mn0, and In0 species
were checked as well. However, they all exhibited poor catalyt-
ic performances in the reaction (Table 1, entries 16–19). The ac-
tivity of Co/ZnO was higher than that of the Co catalyst alone:
the conversion of 1a was 84% and the yield of 3a was 53%
(Table 1, entries 20 versus 1). The results indicated that Zn was
important to the reaction. However, the interaction of ZnO and
1a was weaker than that of ZnCl2 and 1a, which made the ac-
tivity of Co/ZnO lower than that of Co/ZnCl2 (Table 1, entries 20
versus 3). If hydrated zinc chloride was used, 88% conversion
was exhibited and the yield of 3a was 48% (Table 1, entry 21).
The results indicated that the hydrogenation and N-formyla-
tion over the Co0/ZnCl2 catalytic system proceeded independ-
ently. The hydrated water affected the hydrogenation of iso-
quinolines slightly but mainly suppressed the N-formylation
step, possibly as a result of the effect of water on the coordi-
nation of zinc chloride and isoquinoline (Table 1, entry 21).
The results above indicated that Co0/ZnCl2 was the best can-
didate for the N-formylation of isoquinoline with CO2 and H2.
The effect of reaction conditions on the reaction was studied,
and the results are given in Table S1 in the Supporting Infor-
mation and Table 2. The optimized temperature was 1708C
and the best solvent among those checked was THF (for de-
tails, see Table S1 the Supporting Information). The pressures
of CO2 and H2 were also important parameters for the reaction.
Table 2. Effect of reaction conditions on Co0/ZnCl2-catalyzed direct N-for-
mylation of isoquinoline with CO2 and H2.[a]
Entry
PCO /PH
[MPa]
t
[h]
Conversion
[%]
Yield[b] [%]
2
2
2a
3a
4a
1
2
3
4[c]
5
6
7
8
9
0/8
0.5/8
1/8
2/8
2/4
2/6
2/10
2/8
2/8
2/8
16
16
16
16
16
16
16
4
98
99
97
97
53
71
98
32
95
98
98
56
26
2
6
12
2
20
28
12
0
43
71
94
47
59
95
12
67
86
0
0
0
<1
0
0
1
0
0
0
To find out the intermediate of CO2 conversion in the N-for-
mylation reaction, other C1 sources, namely CO, HCOOH, and
HCOONa, were checked with 1a as the starting material
(Table 4). If CO was employed, 1a gave a rather low yield of
3a (1%; Table 4, entry 1). Notably, the hydrogenation activity
of 1a to form 2a was poor in this case, which was probably
a result of catalyst poisoning by CO. With consideration of the
fact that HCOOH can dissolve the Co0 catalyst, HCOONa was
also surveyed. However, only a small amount of 3a was
8
12
10
[a] Reaction conditions: substrate (0.5 mmol), catalyst (5 mg), ZnCl2
(0.02 mmol), THF (2 mL), 1708C. [b] Yield was determined by GC with tol-
uene as an internal standard. [c] From Table 1, entry 3 for comparison.
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
3
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Briefly, 1a was transformed into 2a by selective hy-
drogenation via the I–III species, and 2a then react-
ed with CO2 and H2 via IV to form 3a. The Co0 cata-
lyst and ZnCl2 work cooperatively, which results in
high activity in the reaction. This synergy not only
enhances the selectivity for the hydrogenation of 1a
to 2a but also the N-formylation of 2a to 3a. The
high selectivity of hydrogenation of 1a to 2a is
probably caused by coordination between ZnCl2 and
1a,[15] which has been preliminarily proved by the
1H NMR spectra in Figures S5–S7 in the Supporting
Information. In the 1H NMR spectrum of 1a, the
chemical shift of the H atom in the 1 position on 1a
can be observed at d=9.26 ppm, which is shifted to
d=9.52 ppm after the introduction of ZnCl2. Impor-
tantly, the anion acting as a ligand to the Zn2+ ion
modulates the catalytic performance and affects the
1H NMR spectrum of the H atom in the 1 position of
isoquinoline remarkably, and the ClÀ ion was the
best one for the reaction. It is believed that the
easily formed carbamic acid or carbamate[16] formed
from 2a with CO2 may be an intermediate in the N-
formylation and is then hydrogenated to the formyl
amine over the Co0/ZnCl2 catalyst (Scheme 1). The
high activity is related to the appropriate coordina-
tion abilities of ZnCl2 and the Co0 catalyst. The coor-
dination of ZnCl2 and the nitrogen atom may not
only be able to activate the N-heteroaromatic ring
but also to eliminate the coordination of 1a and/or
2a with the active catalyst sites, and it therefore
promotes both the hydrogenation of 1a and the N-
formylation of 2a.
Table 3. Scope of substrates for direct N-formylation catalyzed by the Co0/ZnCl2 cata-
lyst.[a]
Entry
1
Substrate
Product
Yield[b] [%]
98 (91)
1b
1c
3b
3c
2
88
3
4
1d
1e
3d
3e
92 (88)
79
5[c]
1 f
3 f
51
6[d]
1g
3g
83 (79)
7[d]
1h
1i
3h
3a
63
8
99 (95)
9
1j
3j
8
[a] Reaction conditions were the same as those for entry 3 in Table 1, unless men-
tioned otherwise. [b] Determined by GC–FID with toluene as an internal standard.
Yields of isolated products are given in brackets. [c] t=48 h. [d] t=36 h.
To determine whether Co leached into the solvent
in the reaction, a hot filtration test was performed,
and the result is shown in Figure 2a. The yield of 3a
was not changed after the Co0 catalyst was
Table 4. Control experiments for N-formylation of 1a and 2a with other
C1 sources.[a]
Entry
C1 source
Conversion
[%]
Yield[b] [%]
2a
3a
4a
1
2
3
CO
HCOOH
HCOONa
12
15
8
11
5
3
1
10
5
0
0
0
[a] The reaction conditions were same as those for entry 3 of Table 1,
except that CO2 was replaced by another C1 source (3 equiv., relative to
the substrate). [b] CO pressure was 2 MPa.
produced by using HCOONa or HCOOH as the C1 source
(Table 4, entries 2 and 3). These results demonstrated that the
above C1 sources were probably not the intermediates for the
N-formylation of 1a with CO2 and H2. Additionally, our experi-
ments showed that CO and HCOOH were not detected in the
reaction of CO2 and H2 over the Co0/ZnCl2 catalytic system.
Based on the above experimental study, a possible reaction
mechanism has been proposed for the N-formylation of IQs
with CO2 and H2 over the Co0/ZnCl2 catalytic system, and the
catalytic cycle, with 1a as an example, is shown in Scheme 1.
Scheme 1. Proposed reaction mechanism for N-formylation of isoquinoline
with CO2 and H2 catalyzed by the Co0/ZnCl2 catalyst system.
&
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
4
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Catalyst preparation and characterization
The Co0 catalyst was prepared by reducing of Co3O4 with H2 at
a flow rate of 30 mLminÀ1 at 4008C for 2 h. After the reduction,
the catalyst was cooled to room temperature. The catalyst was
then passivated in a flow of N2 containing 1% O2 (30 mLminÀ1).
The Co/ZnO catalyst was prepared by impregnation–reduction of
Co(NO3)2 on a ZnO support with NaBH4 as the reducing agent. The
catalyst was characterized by TPR, XRD, XPS, and TEM. The TPR test
was performed in a conventional atmospheric quartz flow reactor
(3 mm in diameter; AutoChem II 2950 HP Chemisorption Analyzer,
MICROMERITICS) to detect the effluent gas with a thermal conduc-
tivity detector (TCD). Before the test, the catalyst was heated at
3008C for 30 min in Ar at a flow rate of 50 mLminÀ1. The tempera-
ture was then reduced to 508C and 10% H2/Ar mixed gas was
passed through the catalyst at a flow rate of 30 mLminÀ1. The tem-
perature was linearly raised from 50 to 8008C at a heating rate of
108CminÀ1. The XRD experiment was performed on a Rigaku D/
max 2500 instrument with nickel-filtered CuKa radiation (l=
0.154 nm) operated at 40 kV and 20 mA. The XPS spectra were
measured by using an ESCALab 220I-XL electron spectrometer
from VG Scientific with 300 W AlKa radiation and a hemispherical
energy analyzer. The binding energies were calibrated with the C1s
level of adventitious carbon at 284.8 eV as the internal standard
reference. The TEM images of the catalysts were measured by
using a JEOL-1011 electron microscope operated at 120 kV.
Figure 2. a) Hot filtration test; &: typical reaction, *: the Co0 catalyst was
filtered off. Reaction conditions were similar to those given in Table 1,
entry 3. b) Reusability of the Co0 catalyst.
removed. This indicated that the Co0 catalyst was a highly effi-
cient heterogeneous catalyst for N-formylation of IQs. Notably,
the Co0 catalyst exhibited good stability in the reaction and
could be reused at least five times without apparent loss of ac-
tivity (Figure 2b), and these results further confirmed that the
catalyst worked in a heterogeneous way. The XRD and TEM
characterizations of the fresh and the five-times reused Co0
catalyst indicated that the main crystalline phase and morphol-
ogy did not change significantly (Figure 1c and f). Combined
with the above results, it was determined that the Co0 species
was the active site, and most of Co0 species was still in the Co0
state after the reaction.
N-Formylation of isoquinoline with CO2 and H2
The reaction was conducted in a Teflon-lined stainless-steel reactor
of 16 mL volume with a magnetic stirrer. The pressure was deter-
mined by a pressure transducer (FOXBORO/ICT, Model 93), which is
accurate to Æ0.025 MPa. In a typical experiment, suitable amounts
of Co0, ZnCl2, substrate, and solvent were loaded into the reactor.
The reactor was sealed and purged with CO2 to remove the air at
ambient temperature and was then placed in an air bath at the de-
sired temperature. CO2 was charged to the desired pressure, H2
was added to the final pressure, and then the stirrer was started at
800 rpm. Upon reaction completion, the reactor was placed in ice
water and the gas was released slowly. The reaction mixture was
analyzed by GC–MS and GC with toluene as an internal standard.
GC analysis was performed on an Agilent Technologies 7890B
system equipped with an HP-5 column. GC–MS analysis was con-
ducted on an Agilent 7892B/MSD 5975C system equipped with
a HP-5MS column. NMR spectra were obtained with a Varian
INOVA 400 MHz spectrometer. Chemical shifts are reported in ppm
relative to the deuterated solvent.
Conclusions
In conclusion, we have demonstrated the first approach for
direct N-formylation of IQs with CO2 and H2 to produce FTHIQs
and the Co0/ZnCl2 catalytic system has been found to be excel-
lent for this kind of reaction. Moderate to very high yields of
the desired products can be obtained, depending on the struc-
tures of the substrates. The Co0 catalyst and ZnCl2 work coop-
eratively to catalyze the N-formylation of THIQs. The Co0 cata-
lyst works in a heterogeneous manner and can be reused at
least five times without a considerable decrease in activity. This
work opens the way for the synthesis of formamides from
feedstocks with pyridine structures.
Experimental Section
Chemicals
The reusability of the Co catalyst was tested. After the reaction,
the reaction mixture was centrifuged and the solid catalyst was re-
covered and washed with THF (3ꢁ5 mL). The catalyst was reused
directly for the next run after it had been dried at 608C for 5 h in
a vacuum oven.
Chemicals were sourced as follows: Co3O4 (99.5%, J&K Chemicals),
CoO (99.0%, Sinopharm Chemical Reagent Co., Ltd), CuCl2 (98.5%,
J&K Chemicals, AlCl3 (99%, J&K Chemicals), NiCl2 (98%, J&K Chemi-
cals), LiCl (99%, J&K Chemicals), Fe (98.0%, Sinopharm Chemical
Reagent Co., Ltd), Cu (99.99%, Sinopharm Chemical Reagent Co.,
Ltd), In (99.99%, Alfa Aesar), Mn (99.6%, Alfa Aesar), HCOOH
(98.0%, Sinopharm Chemical Reagent Co., Ltd), and HCOONa
(97.0%, Sinopharm Chemical Reagent Co., Ltd). CO2 (99.99%) and
H2 (99.99%) were provided by Beijing Analytical Instrument Com-
pany. All other chemicals were purchased from commercial sources
and used without further purification.
Under the optimized reaction conditions, a hot filtration experi-
ment was carried out to study whether the catalyst worked in
a
heterogeneous way. In the experiment, the reaction was
quenched after 4 h by placing the reactor in ice water, and the
solid catalyst was separated by filtration. The filtrate was analyzed
by GC and then treated under similar conditions for 12 h, and then
the reaction mixture was analyzed again by GC.
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Dielmann, J. Am. Chem. Soc. 2016, 138, 1840–1843; m) J. Chen, L. Fali-
vene, L. Caporaso, L. Cavallo, E. Y.-X. Chen, J. Am. Chem. Soc. 2016, 138,
5321–5333.
Acknowledgements
This work was supported by the National Natural Science Foun-
dation of China (21603235, 21373234, and 21533011), the Chi-
nese Academy of Sciences (QYZDY-SSW-SLH013), and the Recruit-
ment Program of Global Youth Experts of China.
[7] a) T. V. Q. Nguyen, W.-J. Yoo, S. Kobayashi, Angew. Chem. Int. Ed. 2015,
54, 9209–9212; Angew. Chem. 2015, 127, 9341–9344; b) K. Motokura,
N. Takahashi, A. Miyaji, Y. Sakamoto, S. Yamaguchi, T. Baba, Tetrahedron
2014, 70, 6951–6956; c) S. Zhang, Q. Mei, H. Liu, H. Liu, Z. Zhang, B.
Han, RSC Adv. 2016, 6, 32370–32373; d) S. Kumar, S. L. Jain, RSC Adv.
2014, 4, 64277–64279; e) X. Frogneux, O. Jacquet, T. Cantat, Catal. Sci.
Technol. 2014, 4, 1529–1533; f) O. Jacquet, X. Frogneux, C. D. N. Gomes,
T. Cantat, Chem. Sci. 2013, 4, 2127–2131; g) L. Hao, Y. Zhao, B. Yu, Z.
Yang, H. Zhang, B. Han, X. Gao, Z. Liu, ACS Catal. 2015, 5, 4989–4993;
h) C. Das Neves Gomes, O. Jacquet, C. Villiers, P. Thuꢃry, M. Ephritikhine,
T. Cantat, Angew. Chem. Int. Ed. 2012, 51, 187–190; Angew. Chem. 2012,
124, 191–194; i) O. Jacquet, C. D. N. Gomes, M. Ephritikhine, T. Cantat, J.
Am. Chem. Soc. 2012, 134, 2934–2937; j) X. Cui, Y. Zhang, Y. Deng, F.
Shi, Chem. Commun. 2014, 50, 189–191; k) Q. Bi, J. Lin, Y. Liu, S. Xie, H.
He, Y. Cao, Chem. Commun. 2014, 50, 9138–9140; l) L. Zhang, Z. Han, X.
Zhao, Z. Wang, K. Ding, Angew. Chem. Int. Ed. 2015, 54, 6186–6189;
Angew. Chem. 2015, 127, 6284–6287; m) A. Tlili, E. Blondiaux, X. Frog-
neux, T. Cantat, Green Chem. 2015, 17, 157–168; n) P. G. Jessop, Y.
Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1994, 116, 8851–8852.
[8] a) B. Balakrishna, J. L. NfflÇez-Rico, A. Vidal-Ferran, Eur. J. Org. Chem.
2015, 5293–5303; b) A. Alvanipour, L. D. Kispert, J. Mol. Catal. 1988, 48,
277–283; c) F. Chen, A.-E. Surkus, L. He, M.-M. Pohl, J. Radnik, C. Topf, K.
Junge, M. Beller, J. Am. Chem. Soc. 2015, 137, 11718–11724.
[9] a) Z.-S. Ye, R.-N. Guo, X.-F. Cai, M.-W. Chen, L. Shi, Y.-G. Zhou, Angew.
Chem. Int. Ed. 2013, 52, 3685–3689; Angew. Chem. 2013, 125, 3773–
3777; b) A. Iimuro, K. Yamaji, S. Kandula, T. Nagano, Y. Kita, K. Mashima,
Angew. Chem. Int. Ed. 2013, 52, 2046–2050; Angew. Chem. 2013, 125,
2100–2104.
[10] a) Z. Wei, Y. Chen, J. Wang, D. Su, M. Tang, S. Mao, Y. Wang, ACS Catal.
2016, 6, 5816–5822; b) H. Okazaki, K. Onishi, M. Soeda, Y. Ikefuji, R.
Tamura, I. Mochida, Bull. Chem. Soc. Jpn. 1990, 63, 3167–3174.
[11] B. Su, Z.-C. Cao, Z.-J. Shi, Acc. Chem. Res. 2015, 48, 886–896.
[12] a) W. Song, A. S. Poyraz, Y. Meng, Z. Ren, S.-Y. Chen, S. L. Suib, Chem.
Mater. 2014, 26, 4629–4639; b) H. Karaca, J. Hong, P. Fongarland, P.
Roussel, A. Griboval-Constant, M. Lacroix, K. Hortmann, O. V. Safonova,
A. Y. Khodakov, Chem. Commun. 2010, 46, 788–790; c) Z. Han, H.
Zhang, B. Dong, Y. Ni, A. Kong, Y. Shan, ChemistrySelect 2016, 1, 979–
983.
[13] H. Zhou, J. Song, H. Fan, B. Zhang, Y. Yang, J. Hu, Q. Zhu, B. Han, Green
Chem. 2014, 16, 3870–3875l.
[14] W. Guo, Y. E. L. Gao, L. Fan, S. Yang, Chem. Commun. 2010, 46, 1290–
1292.
[15] M. Li, J. Nie, D. Xu, Acta Crystallogr. Sect. E 2010, 66, m876.
[16] Q. Zhou, Y. Li, J. Am. Chem. Soc. 2015, 137, 10182–10189.
Keywords: carbon dioxide
catalysis · hydrogenation · isoquinolines
· formylation · heterogeneous
[1] a) C. J. Suckling, J. A. Murphy, A. I. Khalaf, S. Zhou, D. E. Lizos, A. N.
van Nhien, H. Yasumatsu, A. McVie, L. C. Young, C. McCraw, P. G. Water-
man, B. J. Morris, J. A. Pratt, A. L. Harvey, Bioorg. Med. Chem. Lett. 2007,
17, 2649–2655; b) D. E. Lizos, C. Mckerchar, J. Murphy, Y. Shiigi, C. Suck-
ling, H. Yasumatsu, S.-Z. Zhou, J. Pratt, B. Morris (Mitsubishi Pharma Cor-
poration), WO 2004/087124 A1, 2004; c) A. Romanens, G. Belanger, Org.
Lett. 2015, 17, 322–325; d) A. I. Meyers, L. M. Fuentes, J. Am. Chem. Soc.
1983, 105, 117–118.
[2] a) P. Preedasuriyachai, H. Kitahara, W. Chavasiri, H. Sakurai, Chem. Lett.
2010, 39, 1174–1176; b) H. Lv, Q. Xing, C. Yue, Z. Lei, F. Li, Chem.
Commun. 2016, 52, 6545–6548; c) N. Ortega, C. Richter, F. Glorius, Org.
Lett. 2013, 15, 1776–1779.
[3] a) S. I. Murahashi, Y. Imada, Y. Hirai, Tetrahedron Lett. 1987, 28, 77–80;
b) L. Tao, Q. Zhang, S. Li, X. Liu, Y. Liu, Y. Cao, Adv. Synth. Catal. 2015,
357, 753–760.
[4] S.-I. Murahashi, Y. Imada, Y. Hirai, Bull. Chem. Soc. Jpn. 1989, 62, 2968–
2976.
[5] a) J. Jin, D. W. C. MacMillan, Angew. Chem. Int. Ed. 2015, 54, 1565–1569;
Angew. Chem. 2015, 127, 1585–1589; b) M. Wan, H. Lou, L. Liu, Chem.
Commun. 2015, 51, 13953–13956; c) Y. Siddaraju, M. Lamani, K. R.
Prabhu, J. Org. Chem. 2014, 79, 3856–3865; d) N. Okugawa, K. Moriya-
ma, H. Togo, Eur. J. Org. Chem. 2015, 4973–4981; e) L. Shi, Z. Ye, L. Cao,
R. Guo, Y. Hu, Y. Zhou, Angew. Chem. Int. Ed. 2012, 51, 8286–8289;
Angew. Chem. 2012, 124, 8411–8414.
[6] a) M. Hulla, F. D. Bobbink, S. Das, P. J. Dyson, ChemCatChem 2016, 8,
3338–3342; b) A. V. Bavykina, E. Rozhko, M. G. Goesten, T. Wezendonk,
B. Seoane, F. Kapteijn, M. Makkee, J. Gascon, ChemCatChem 2016, 8,
2217–2221; c) P.-Z. Li, X.-J. Wang, J. Liu, J. S. Lim, R. Zou, Y. Zhao, J. Am.
Chem. Soc. 2016, 138, 2142–2145; d) M. He, Y. Sun, B. Han, Angew.
Chem. Int. Ed. 2013, 52, 9620–9633; Angew. Chem. 2013, 125, 9798–
9812; e) Z. He, Q. Qian, J. Ma, Q. Meng, H. Zhou, J. Song, Z. Liu, B. Han,
Angew. Chem. Int. Ed. 2016, 55, 737–741; Angew. Chem. 2016, 128,
747–751; f) Z. He, Q. Qian, Z. Zhang, Q. Meng, H. Zhou, Z. Jiang, B. Han,
Philos. Trans. R. Soc. A 2015, 373, 20150006; g) Q. Qian, J. Zhang, M. Cui,
B. Han, Nat. Commun. 2016, 7, 11481; h) X. Wang, M. Nakajima, R.
Martin, J. Am. Chem. Soc. 2015, 137, 8924–8927; i) L. Wu, Q. Liu, I.
Fleischer, R. Jackstell, M. Beller, Nat. Commun. 2014, 5, 3091; j) S. Moret,
P. J. Dyson, G. Laurenczy, Nat. Commun. 2014, 5, 4017; k) M. Cui, Q.
Qian, Z. He, Z. Zhang, J. Ma, T. Wu, G. Yang, B. Han, Chem. Sci. 2016, 7,
5200–5205; l) F. Buß, P. Mehlmann, C. Mꢂck-Lichtenfeld, K. Bergander, F.
Manuscript received: December 22, 2016
Revised: February 16, 2017
Accepted Article published: March 1, 2017
Final Article published: && &&, 0000
&
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Z. He, H. Liu, H. Liu,* Q. Qian, Q. Meng,
Q. Mei, B. Han*
&& – &&
Heterogeneous Cobalt-Catalyzed
Direct N-Formylation of Isoquinolines
with CO2 and H2
Taking a direct route: N-Formylation of
isoquinolines with CO2 and H2 repre-
sents a step-economical route for the
synthesis of N-formyl-tetrahydroisoqui-
nolines. A heterogeneous cobalt catalyst
(Co0) and ZnCl2, which together have an
excellent cooperative effect in the cata-
lytic process, can achieve the reaction
with moderate to high yields.
ChemCatChem 2017, 9, 1 – 7
www.chemcatchem.org
7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ