Zn-Co bimetallic supported ZSM-5 catalyst for phosgene-free synthesis of hexamethylene–1,6–diisocyanate by thermal decomposition of hexamethylene–1,6–dicarbamate

Article abstract of DOI:10.1016/j.cclet.2017.03.015

A set of mono- and bimetallic (Zn-Co) supported ZSM-5 catalysts was first prepared by PEG-additive method. The physicochemical properties of the catalysts were investigated by FTIR, XPS, XRD, N₂ adsorption-desorption measurements, SEM, EDS and NH₃-TPD techniques. The physicochemical properties showed that the ZnCo₂O₄ spinel oxide was formed on the ZSM-5 support and provided effectual synergetic effect between Zn and Co species for the bimetallic catalyst. Furthermore, bimetallic supported ZSM-5 catalyst exhibited weak, moderate and strong acidic sites, while the monometallic supported ZSM-5 catalyst showed only weak and moderate or strong acidic sites. Their catalytic performances for thermal decomposition of hexamethylene–1,6–dicarbamate (HDC) to hexamethylene–1,6–diisocyanate (HDI) were then studied. It was found that the bimetallic supported ZSM-5 catalysts, especially Zn-2Co/ZSM-5 catalyst showed excellent catalytic performance due to the good synergetic effect between Co and Zn species, which provided a suitable contribution of acidic sites. HDC conversion of 100% with HDI selectivity of 91.2% and by-products selectivity of 1.3% could be achieved within short reaction time of 2.5?h over Zn-2Co/ZSM-5 catalyst.

Full text of DOI:10.1016/j.cclet.2017.03.015

Accepted Manuscript
Title: Zn-Co bimetallic supported ZSM-5 catalyst for
phosgene-free synthesis of hexamethylene–1,6–diisocyanate
by thermal decomposition of hexamethylene–1,6–dicarbamate
Authors: Muhammad Ammar, Yan Cao, Peng He, Li-Guo
Wang, Jia-Qiang Chen, Hui-Quan Li
PII:
S1001-8417(17)30093-1
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.cclet.2017.03.015
CCLET 4010
To appear in:
Chinese Chemical Letters
Received date:
Revised date:
Accepted date:
20-12-2016
21-1-2017
9-3-2017
Please cite this article as: Muhammad Ammar, Yan Cao, Peng He, Li-
Guo Wang, Jia-Qiang Chen, Hui-Quan Li, Zn-Co bimetallic supported ZSM-
5
catalyst for phosgene-free synthesis of hexamethylene–1,6–diisocyanate by
thermal decomposition of hexamethylene–1,6–dicarbamate, Chinese Chemical
Lettershttp://dx.doi.org/10.1016/j.cclet.2017.03.015
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Original article
Zn-Co bimetallic supported ZSM-5 catalyst for phosgene-free synthesis of
hexamethylene‒1,6‒diisocyanate by thermal decomposition of hexamethylene‒1,6‒
dicarbamate
Muhammad Ammar a, b, Yan Cao a, , Peng He , Li-Guo Wang , Jia-Qiang Chen , Hui-Quan Li
a
a
a
a,*
a
Key Laboratory of Green Process and Engineering, National Engineering Laboratory for Hydrometallurgical Cleaner Production Technology, Institute of
Process Engineering, CAS, Beijing 100190, China
University of Chinese Academy of Sciences, Beijing 100049, China
b
Graphical abstract
In this study, a phosgene-free route for the thermal decomposition of HDC to HDI was demonstrated over Zn-Co bimetallic
supported ZSM-5 catalyst. Bimetallic supported ZSM-5 catalyst revealed the deposition of ZnCo O on the ZSM-5 support.
2 4
Furthermore, bimetallic supported ZSM-5 catalyst exhibited greater performance for the thermal decomposition of HDC to HDI
than monometallic supported ZSM-5 catalyst.
ARTICLE INFO
ABSTRACT
Article history:
A set of mono- and bimetallic (Zn-Co) supported ZSM-5 catalysts was first prepared by
PEG-additive method. The physicochemical properties of the catalysts were investigated
Received 20 December 2016
Received in revised form 21 January 2017
Accepted 27 February 2017
Available online
by FTIR, XPS, XRD, N
TPD techniques. The physicochemical properties showed that the ZnCo
2
adsorption-desorption measurements, SEM, EDS and NH
3
-
2 4
O
spinel oxide
was formed on the ZSM-5 support and provided effectual synergetic effect between Zn
and Co species for the bimetallic catalyst. Furthermore, bimetallic supported ZSM-5
catalyst exhibited weak, moderate and strong acidic sites, while the monometallic
supported ZSM-5 catalyst showed only weak and moderate or strong acidic sites. Their
catalytic performances for thermal decomposition of hexamethylene‒1,6‒dicarbamate
Keywords: Hexamethylene‒1,6‒dicarbamate
(
HDC) to hexamethylene‒1,6‒diisocyanate (HDI) were then studied. It was found that
(HDC)
the bimetallic supported ZSM-5 catalysts, especially Zn-2Co/ZSM-5 catalyst showed
excellent catalytic performance due to the good synergetic effect between Co and Zn
species, which provided a suitable contribution of acidic sites. HDC conversion of 100%
with HDI selectivity of 91.2% and by-products selectivity of 1.3% could be achieved
within short reaction time of 2.5 h over Zn-2Co/ZSM-5 catalyst.
Hexamethylene‒1,6‒diisocyanate (HDI)
Thermal decomposition
Bimetallic supported ZSM-5 catalyst
Synergetic effect
1
. Introduction
HDI is one of the most extensively used aliphatic diisocyanate in the manufacture of polyurethane, paint, insulator, synthetic rubber,
adhesives [1, 2]. Traditionally, the synthesis route of HDI involves extremely toxic phosgene as a reagent at industrial scale [3]. This
route to produce HDI not only causes of environmental hazards due to toxicity of phosgene, but also produces seriously corrosive
byproduct HCl [2]. The drawbacks of this phosgene route divert attention to the development of phosgene-free route. Among other
phosgene-free routes for HDI synthesis, thermal decomposition of HDC to HDI is one of the most effective, environmentally benign and

Corresponding authors.
E-mail address: ycao@ipe.ac.cn (Y. Cao), hqli@ipe.ac.cn (H.-Q Li)

green route [4]. Thermal decomposition of HDC to HDI involves two steps, HDC decomposes to hexamethylene‒1‒carbamate‒6‒
isocyanate (HMI) as intermediate and then HMI decomposes to HDI as shown in Scheme 1.
Thermal decomposition of HDC to HDI is still in an immature state and so far it has faced the problems such as low HDC conversion,
low HDI selectivity and polymerization of highly active HDI. In order to improve this phosgene-free route and overcome its problems,
many researchers have focused on the use of catalysts, e.g. di-n-butyltin oxide [5], montromrillonite K-10 [6], ZnAlPO
4
[7], ZnO [8],
CuO/ZnO [9] and Co [10]. Concerning all the catalysts reported up to date, zinc and cobalt based catalysts have found relatively
2
O
3
active and selective to HDI. However, HDI yield is still not ideal (i.e. less than 90%) and the route suffers with long reaction time, which
results in high energy cost. Therefore, it is vital to modify the catalyst which can efficiently improve this route.
The catalytic properties of the catalyst can be improved by depositing active species on a support [11, 12]. The choice of the support
plays an important role, since support not only acts as a carrier of the active species, but also contributes to the performance of the
catalyst. It has been reported that the interactions between metal and support significantly affect the physicochemical properties of the
catalyst such as dispersion, size and metal crystallite distribution or acidity [13, 14]. ZSM-5 is a suitable choice to use as a catalyst
support due to its unique channel structure, large surface, thermal stability, shape-selectivity and acidity. The strength and distribution
of acidic sites in ZSM-5 have been shown to be altered upon loading metal species on the ZSM-5 [15]. Therefore, modified ZSM-5 as
catalyst is widely used in various moderate acid-catalyzed reactions [16]. Furthermore, it is reported that ZSM-5 modified by bimetallic
cations found much more active and selective catalyst as compared to monometallic modified ZSM-5 catalyst [17, 18]. It is clear from
previous work that both Zn and Co are active and selective to HDI. Therefore, it could be expected that Zn-Co bimetallic catalyst might
give better performance in the thermal decomposition of HDC. However, the catalytic performance of Zn-Co bimetallic supported ZSM-
5
catalyst for the thermal decomposition of HDC has not been reported and the influence of Zn and Co synergetic effect on the selectivity
of HDI has not been studied.
In this work, a phosgene-free route for the thermal decomposition of HDC to HDI was developed over Zn-Co bimetallic supported
ZSM-5 catalyst. For this, the monometallic (Co, Zn) and bimetallic (Zn-Co) supported ZSM-5 catalysts were prepared by PEG-additive
method and used in the thermal decomposition reaction using low boiling point solvent. The physicochemical properties of the catalysts
were thoroughly characterized by different techniques. The catalytic performances of the monometallic and bimetallic supported ZSM-
5
catalysts were studied. In addition, a correlation was explored between the contribution of acidic sites and catalytic performance of the
catalysts. The reusability of the catalyst was also evaluated.
2
. Results and discussion
2
.1. Catalyst characterization
Fig. 1 shows the FTIR spectra of the ZSM-5, monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts. The band at 3654,
-
1
3
442 and 3230 cm were ascribed to terminal hydroxyl groups Si-OH, hydroxyl groups on extra framework aluminum and strong
-
1
Bronsted acid sites Si-OH-Al groups on the zeolite’s surface [20]. The band at 1630 cm was assigned to water bending vibration and
observed in ZSM-5 and all monometallic and bimetallic supported ZSM-5 catalysts. The characteristic bands of MFI structure at 1220,
-
1
1
083, 795, 544 and 452 cm were observed in all catalysts, which ascribed to external asymmetric stretch, internal asymmetric stretch,
external symmetric stretch, double five ring, and T-O bending vibration of internal tetrahedral respectively [21, 22]. These results
confirmed that the deposition of mono and bi metal cations on the ZSM-5 had no influence on the framework of ZSM-5.
Compared with ZSM-5 spectrum (Fig. 1, curve a), two new bands at 579 and 670 cm-1 were observed in the 2Co/ZSM-5 spectrum
(
Fig. 1, curve b). These bands can be associated with the metal-oxygen bonding which resulted in after the metal loading on the ZSM-5.
-
1
3+
-1
2+
The band at 579 cm was ascribed to the Co octahedral coordination, while the band at 670 cm was ascribed to the Co tetrahedral
coordination in cubic spinel Co structure [23]. As can be seen, all the bimetallic supported ZSM-5 spectra (Fig. 1, curves c-e) also
showed the two new bands at 579 and 670 cm same as the 2Co/ZSM-5 spectrum. The FTIR results apparently suggested the deposition
of cubic spinel oxide of metals on bimetallic supported ZSM-5 catalysts. However, there was no new band appeared in the 2Zn/ZSM-5
spectrum (Fig. 1, curve f) as compared to ZSM-5 spectrum, since the band of ZnO at 450 cm was overlapped with the intensive
adsorption band of T-O bending vibration in MFI structure at 452 cm [24].
3
O
4
-
1
-
1
-
1
Fig. S1 in Supporting information shows the surface composition and oxidation state of monometallic and bimetallic (Zn-Co)
supported ZSM-5 catalysts analyzed by XPS. The XPS full-range survey spectra of monometallic and bimetallic supported ZSM-5
catalysts are shown in Fig. S1(A). The characteristic peaks of Zn and/or Co, O, Si and Al elements were detected in all the catalysts. Fig.
S1(B) shows the Zn2p spectra of monometallic (Zn) and bimetallic supported ZSM-5 catalysts. Two main peaks at 1022-1022.3 and
2
+
1
045.2-1045.4 eV were detected in Zn2p spectra, which were ascribed to Zn2p3/2 and Zn2p1/2, signifying Zn oxidation state. Fig. S1(C)
shows the Co2p spectra of monometallic (Co) and bimetallic supported ZSM-5 catalysts. The Co2p spectra displayed two peaks at 780.3-
2
+
3+
7
80.7 eV and 795.3-795.7, which were attributed to Co2p3/2 and Co2p1/2, indicating Co and Co oxidation state. Fig. S1(D) shows the
O1s spectra of monometallic and bimetallic supported ZSM-5 catalysts and could be deconvoluted to three components. The first
component at 529.9 eV corresponded to the surface lattice oxygen of the catalysts (O ), the second component at 531.1 eV corresponded
to the adsorbed oxygen of the catalysts (O ) and the last component at 532.7 eV corresponded to the lattice oxygen of the ZSM-5 support
) [25, 26].
α
β
(O
γ

Fig. 2 shows the XRD patterns of the ZSM-5, monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts. It can be seen that all
the catalysts showed peaks at range of 2θ = 7-9° and 2θ = 23-25°, which were corresponded to the typical peaks of ZSM-5 zeolite [27].
It implied that the crystalline structure of ZSM-5 remained intact after Zn and/or Co supported on ZSM-5. The XRD pattern of 2Zn/ZSM-
5
catalyst revealed the deposition of hexagonal phase of ZnO on the ZSM-5 support. The diffraction peaks of hexagonal phase of ZnO
found at 2θ = 37.7°, 34.7°, 36.2°, 47.5°, 56.5°, 62.7° and 67.9° correspond to the (100), (002), (101), (102), (110), (103) and (112) crystal
planes (JCPDS 89-1397) in the 2Zn/ZSM-5 catalyst respectively (pattern f). While, the diffraction peaks of the cubic spinel phase of
Co
crystal planes (JCPDS 42-1467) in the 2Co/ZSM-5 catalyst respectively (pattern b). It is noteworthy that the XRD pattern of Zn-
Co/ZSM-5 catalyst showed the position of diffraction peaks almost same as the position of diffraction peaks of cubic spinel Co in
XRD pattern of 2Co/ZSM-5 catalyst (pattern b and c). As shown earlier by XPS results that Zn was presented on the surface of Zn-
Co/ZSM-5 catalyst. However, there was no characteristic peak of ZnO phase detected in the XRD pattern of Zn-2Co/ZSM-5 catalyst.
Therefore, it clearly revealed the deposition of cubic spinel phase of ZnCo on the ZSM-5 support in Zn-2Co/ZSM-5 catalyst and was
in good agreement with JCPDS 23-1390. According to literature [28], cubic spinel ZnCo is isomorphic to the cubic spinel Co
3 4
O found at 2θ = 19.0°, 31.2°, 36.8°, 38.5°, 44.8°, 59.4° and 65.2° correspond to the (111), (220), (311), (222), (400), (511) and (440)
2
3 4
O
2
2
O
4
2
O
4
3 4
O
2
+
3+
crystal structure wherein Zn has tetrahedral coordination and Co has octahedral coordination in the lattice. Therefore, it can be
expected that interaction between divalent and trivalent metal cations and their respective position in the lattice of the cubic spinel
structure may be beneficial in improving the catalytic performance. In contrast, the diffraction peaks of the hexagonal phase of ZnO
appeared in the XRD patterns of Zn-Co/ZSM-5 and 2Zn-Co/ZSM-5 (pattern d and e) indicated the co-existence of cubic spinel phase of
ZnCo
2 4
O and hexagonal phase of ZnO.
Surface texture properties of the ZSM-5, monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts were measured by N
2
adsorption-desorption measurements and the data are listed in Table 1. It can be seen that the bare ZSM-5 revealed a high BET surface
area and total pore volume. With the addition of Zn and/or Co species on the ZSM-5 support, the resultant monometallic and bimetallic
(Zn-Co) supported catalysts showed a decrease in the BET surface area and total pore volume, which can be attributed to the blocking
of some pores and channels of ZSM-5 support by metal species [29]. In contrast, an increase in the average pore diameter was observed
for monometallic and bimetallic (Zn-Co) supported catalysts, indicating the location of some metal oxide species inside ZSM-5 pores
caused an enlargement [30]. It is noteworthy that the deposition of Co
average pore diameter compared with bare ZSM-5. This observation revealed that more Co
ZSM-5 pores as compared to ZnO. While, ZnO was mostly deposited on the exterior surface of the ZSM-5 support. As shown in Fig. S2
in Supporting information, N adsorption-desorption isotherms for ZSM-5, monometallic supported ZSM-5 and Zn-2Co/ZSM-5 catalysts
3
O
4
or ZnCo
2
O
4
on the ZSM-5 support led to a large increase in
3
O
4
or ZnCo species accommodated inside
2 4
O
2
exhibited type VI isotherms according to the IUPAC classification. The type of isotherms was not changed after the deposition of metal
species on the ZSM-5, which revealed the presence of still open mesopores on the outer surface. However, a decrease in the quantity of
adsorbed N was observed, indicating the pore blockage by metal species. Therefore, it could be deduced that Co O or ZnCo O with
2 3 4 2 4
cubic spinel structure blocked a large number of pores, resulting in the drastic decrease in BET surface.
Fig. 3. Surface SEM images of bare ZSM-5 (a), 2Co/ZSM-5 (b), Zn-2Co/ZSM-5 (c) and 2Zn/ZSM-5 (d) catalyst.
Fig. 4 shows the NH
quantitative results are listed in Table 1. According to the literature [31], the desorption peaks centered in the temperature range of 100-
50 °C, 225-350 °C and 400-500 °C could be attributed to weak, moderate and strong acidic sites respectively. NH -TPD profiles of the
3
-TPD profiles of the ZSM-5, monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts and the
1
3
ZSM-5 and monometallic supported ZSM-5 catalysts showed two major desorption peaks, whereas the bimetallic (Zn-Co) supported
ZSM-5 catalysts revealed three major desorption peaks. The ZSM-5 revealed two major desorption peaks at 112 and 425 °C, ascribing
to weak and strong acidic sites on the surface (profile a). As can be seen in Fig. 4, the TPD profiles were varied significantly with the
introduction of metal species on the ZSM-5. Interestingly, the introduction of Co species on the ZSM-5 demolished the weak acidic sites
and facilitated the formation of new strong acidic sites (profile b). No shift in the desorption peak of 2Co/ZSM-5 catalyst indicated that
the acidic strength was unchanged compared with the ZSM-5 and acidic sites distribution was only affected by Co loading. In a contrast,
the introduction of Zn species on the ZSM-5 poisoned the strong acidic sites of the ZSM-5 and facilitated the formation of new moderate
acidic sites in 2Zn/ZSM-5 catalyst (profile f).
3 3
Compared to the NH -TPD profile of the ZSM-5, the NH -TPD profiles of bimetallic (Zn-Co) supported ZSM-5 catalysts exhibited
an extra desorption peak at 270 °C, suggesting the presence of weak, moderate and strong acidic sites on the surface (profile c-e). This
reflects that the coexistence of Zn and Co species on the ZSM-5 effectively affected the acidic sites distribution of the catalyst through a
synergetic effect of both metal species. It is worth noting that a significant shift in desorption peak centered at 425-465 °C for the
bimetallic supported catalysts, suggesting that major change in the strong acidic sites strength occurred. The desorption peaks were found
to be shifted in high temperature with increasing Zn and decreasing Co proportion in bimetallic supported catalysts. Following the order
Zn-Co/ZSM-5 < 2Zn-Co/ZSM-5 < Zn-2Co/ZSM-5, the contribution of moderate and strong acidic sites increased and the contribution
of weak acidic sites decreased in the catalysts as shown in Table 1.
2
.2. Catalyst performance evaluation
The catalytic performances of the as-prepared catalysts for the thermal decomposition of HDC to HDI were evaluated and the results are
listed in Table. 1. A low HDC conversion of 89.2% with 49.0% HDI selectivity and 39.2% HMI selectivity was achieved in the presence of
ZSM-5. Interestingly, the introduction of cobalt and/or zinc species on the ZSM-5 gave dramatic increment of strongly active sites on the
catalysts, which resulted in an increase of not only HDC conversion but also HDI selectivity. Comparing both monometallic supported ZSM-5

catalysts, the catalytic performance of 2Co/ZSM-5 catalyst with 81.5% HDI selectivity was inferior to the 2Zn/ZSM-5 catalyst with 85.0% HDI
selectivity. These results apparently suggested that Zn species was comparatively more selective to HDI than Co species for the thermal
decomposition of HDC. Thus, it could be expected that the synergetic effect between Zn and Co species in bimetallic supported ZSM-5 catalyst
as confirmed by the characterization, might be beneficial to improve the catalytic performance. Therefore, different as-prepared bimetallic
supported ZSM-5 catalysts were investigated in the thermal decomposition of HDC to HDI under the same reaction conditions.
The bimetallic supported ZSM-5 catalyst showed obviously better selectivity to HDI as compared to monometallic supported ZSM-5
catalysts. The HDI selectivity was about 83.2% with 10.0% by-products selectivity over Zn-Co/ZSM-5 catalyst. Compared with Zn-Co/ZSM-5
catalyst, HDI selectivity increased and by-products selectivity further decreased after increasing Zn species and reached 87.1% and 7.4% over
2
Zn-Co/ZSM-5 catalyst respectively. On the other hand, when Co species were increased in the Zn-Co/ZSM-5 catalyst, the Zn-2Co/ZSM-5
catalyst revealed superior selectivity to HDI and inferior selectivity to by-products. The HDI selectivity reached 91.2% with only 1.3% by-
products selectivity after 2.5 h.
According to the literature [6], the modification of acidic properties of the catalyst for thermal decomposition of carbamate to
isocyanate could be beneficial for improving its selectivity to desired product. Therefore, the acidic properties of the catalysts were
correlated with HDI selectivity to gain further insight into the possible reason for performance. As can be seen in Table 1, the 2Zn/ZSM-
5
2
catalyst with the contribution of moderate acidic sites revealed higher HDI selectivity and slightly lower by-products selectivity than
Co/ZSM-5 catalyst with contribution of strong acidic sites. As discussed earlier, the synergetic effect between Zn and Co led the
simultaneous formation of weak, moderate and strong acidic sites in bimetallic supported ZSM-5 catalysts. It was observed that bimetallic
supported ZSM-5 catalysts were more selective to HDI than monometallic supported ZSM-5 catalysts due to the combined contribution
of weak, moderate and strong acidic sites. All the bimetallic supported catalysts tested in this study, the selectivity of HDI increased and
selectivity of by-products decreased as the contribution of moderate and strong acidic sites increased and contribution of weak acidic
sites decreased. This finding indicated that the co-contribution of moderate and strong acidic sites of the bimetallic supported ZSM-5
catalysts played a crucial role for the HDI selectivity. Thus, bimetallic supported ZSM-5 catalysts revealed better catalytic performance
than monometallic supported catalysts. Particularly, Zn-2Co/ZSM-5 catalyst showed superior catalytic performance in the thermal
decomposition of HDC to HDI, which might be due to the relatively average acidity and greater contribution of moderate and strong
acidic sites.
2
.3. Catalyst reusability
The recycling of the Zn-2Co/ZSM-5 catalyst was also investigated and results are shown in Fig. 5. The catalyst was collected after
o
the reaction by simple centrifugation, washed with chlorobenzene and dried at 120 C for 12 h then reused directly for the next run.
During the recycling experiments, it can be seen that the HDC conversion was unchanged. In contrast, a slight decrease in the HDI and
HMI selectivity and increase in by-products selectivity was observed in the second and third run. However, the catalyst revealed low
performance after being used in three consecutive runs. The HDI selectivity gradually decreased and HMI and by-products selectivity
increased in fourth and fifth run. The decrease in the performance of the Zn-2Co/ZSM-5 catalyst was possibly due to loss of active sites
on the catalyst surface during the thermal decomposition reaction. Therefore, XPS characterization of the Zn-2Co/ZSM-5 catalyst was
used to investigate the changes of the used catalyst. Fig. S4 in Supporting information shows the Zn2p and Co2p spectra of fresh and
used Zn-2Co/ZSM-5 catalyst. Compared with fresh Zn-2Co/ZSM-5 catalyst, the Zn2p and Co2p peaks of used Zn-2Co/ZSM-5 catalyst
shifted to higher binding energy, i.e. 1021.9-1045.1 eV to 1023.7-1046.9 eV and 780.7-795.7 to 782.3-797.3 eV, respectively. The higher
binding energies of both Zn and Co could be attributed to the weak interaction inside ZnCo
2 4
O phase due to electron-deficient state of
metal species in used Zn-2Co/ZSM-5 catalyst. It is reported that methoxy group (CH O-) can adsorbed on the Zn-site and atomic
3
hydrogen (H) can bonded to the O-site in the system containing ZnO species and transferred crystalline ZnO to amorphous form [8, 32].
Therefore, it could be deduced that ZnO formed intermediate species with methoxy group and atomic hydrogen during the thermal
decomposition reaction and leached out into the reaction mixture. The leaching of ZnO caused structural defects in the ZnCo
2 4
O cubic
spinel structure and resulted in loss of active sites on the surface of the catalyst and performance of the Zn-2Co/ZSM-5 catalyst.
3
. Conclusion
In summary, a phosgene-free route for the thermal decomposition of HDC to HDI was developed over Zn-Co bimetallic supported
ZSM-5 catalyst, using low boiling point solvent. The monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts with different Zn
and Co weight composition were prepared by the PEG-additive method. Their physicochemical properties were characterized by various
techniques and found that the introduction of Co and/or Zn on the ZSM-5 had no significant influence on the framework of ZSM-5. The
synergetic effect between Zn and Co over bimetallic supported ZSM-5 catalyst significantly affected the contribution of acidic sites. The
catalytic performances of different as-prepared monometallic and bimetallic supported ZSM-5 catalysts were investigated in the thermal
decomposition of HDC. The results showed that bimetallic supported ZSM-5 catalysts were more selectivity to HDI than monometallic
supported ZSM-5 catalysts. The catalytic performance of bimetallic supported ZSM-5 catalyst depended upon the contribution of weak,
moderate and strong acidic sites. Among all the catalysts, Zn-2Co/ZSM-5 catalyst revealed excellent performance with 91.2% HDI
selectivity and 1.3% by-products selectivity.
4. Experimental
4
.1. Materials

2 2 3
The ZSM-5 zeolite (SiO /Al O = 25) was purchased from the Catalyst Plant of Nankai University and used as support. Cobalt nitrate
hexahydrate (98.5%), Zinc nitrate hexahydrate (98.5%) Chlorobenzene (≥99.0%), Methanol (≥99.0%) and Biphenyl were purchased
from Sinopharm Chemical Reagent Co., Ltd., China. Polyethylene glycol (PEG) (MW = 400) was purchased from Aladdin Chemical
Co., Ltd., Shanghai, China. HDI (>98%) was purchased from Tokyo Chemical Industry Co., Ltd., Japan. All these reagents were used
as received without further purification. HDC was synthesized by the reaction of HDI with methanol (m(HDI):m(Methanol) = 1:5) at
3
5 °C for 12 h. HDC (>99%) was finally obtained by purification and confirmed by Mass Spectrometry and FTIR Spectroscopy. HMI
was obtained in the form of HDC, HMI and HDI mixture, by the reaction of HDI with methanol (m(HDI):m(Methanol) = 3:1) at 35 °C
for 1.5 h., since HMI was not commercially available.
4
.2. Preparation of catalysts
The monometallic and bimetallic (Zn-Co) supported ZSM-5 catalysts were prepared by PEG-additive method, according to literature
reported [19]. Prior to metals loading, ZSM-5 zeolite was calcined at 500 °C for 3 h under static air in a muffle furnace. The desired
amounts of Co(NO .6H O and/or Zn(NO .6H O and PEG (metal/PEG ratio 0.16) were added into a solvent containing methanol (30
3
)
2
2
3
)
2
2
mL) and deionized water (90 mL).The resultant solution was refluxed at 90 °C for 2 h. 3 g of ZSM-5 support was added into the solution
subsequently. The obtained suspension was heated at 100 °C with continuous stirring to evaporate the solvent. All the catalysts were
dried at 120 °C for 12 h and calcined at 600 °C for 6 h under static air in a muffle furnace. The monometallic supported ZSM-5 catalyst
was denoted as xZn/ZSM-5 and yCo/ZSM-5 and bimetallic supported ZSM-5 catalyst were denoted as xZn-yCo/ZSM-5, where x and y
denotes the weight composition of zinc and cobalt (‘1’ refers to 10wt% and ‘2’ refers to 20 wt% relative to the weight of the support).
For example, the catalyst with 10wt% Zn and 20wt% Co was named Zn-2Co/ZSM-5.
4
.3. Characterization of catalysts
Fourier-transform infrared spectra (FTIR) were recorded on a Bruker Tensor 27 spectrometer at room temperature. For the FTIR
-
1
characterizations, the catalysts were dispersed in KBr to make pellets. The spectra resolution was 4 cm . All the spectra were recorded
-
1
in the range of 4000-400 cm . Surface composition of the catalysts was examined by X-ray photoelectron spectroscopy (XPS) analysis.
XPS was performed under an ultrahigh vacuum using an ESCALAB 250Xi spectrometer with Al Kα radiation (1486.6 eV) and a
multichannel detector. The collected binding energies were calibrated using the C1s peak at 284.6 eV as the reference. The XRD
diffraction patterns of all the catalysts were obtained with a PANalytical Empyrean diffractometer using Cu Kα radiation and recorded
in the 2θ range of 5° to 90°. N
. The samples were outgassed at 523 K under vacuum for 3 h before recording their isotherms. The BET surface area of the samples
was calculated according to the Brunauer-Emmett-Teller theory (BET). Total pore volumes were evaluated at relative pressures (P/P
2
adsorption-desorption isotherms of the catalysts at 77 K were measured by using Quantachrome Autosorb-
1
o
)
closed to unity. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) were conducted using Hitachi
SU 8020 field-emission scanning electron microscope. The acidity of the catalysts was detected using the temperature-programmed
desorption of ammonia (NH
NH -TPD was carried out from 50 °C to 600 °C with a ramping rate of 10 °C/min, and the amount of desorbed NH
thermal conductivity detector (TCD).
3
-TPD) using the Micrometrics Autochem II 2920 unit equipped with the thermal conductivity detector.
3
3
was monitored by a
4
.4. Reaction procedure and product analysis
The thermal decomposition reaction was carried out in a 1 L stainless steel autoclave. In a typical reaction, HDC (19.6 g) and
chlorobenzene (560.0 g) were charged into the autoclave containing catalyst (0.25 g). Then, the autoclave was purged with N (99.9%)
at the flow rate of 800 mL/min for 15 min to ensure the complete removal of inner oxygen. The reaction mixture was then heated to
30 °C at the heating rate of 10 °C/min and held at the temperature with continuous mechanical stirring for 3 h. The reaction time started
once the autoclave reached the desired temperature. During the reaction, the methanol produced by the decomposition of HDC to HDI,
was continually removed from the reaction system using N flow at 800 ml/min. The pressure of the autoclave was 0.68 MPa. After the
2
2
2
reaction, the autoclave was cooled to room temperature and the product mixture sample was taken for analysis. The product mixture
collected was quantitatively analyzed by Shimadzu GC-2010 equipped with a Rtx-5 (30 m × 0.25 mm × 0.25 µm) capillary column and
flame ionization detector (FID). The injection and detector temperatures were 260 °C and 280 °C, respectively. The temperature program
for GC analysis was held at 150 °C for 6 min., from 150 °C to 230 °C at 10 °C/min and then held at 230 °C for 5 min. Nitrogen and
Biphenyl were used as a carrier gas and internal standard for quantitative analysis, respectively. HDC, HMI and HDI standard curves
were conducted under the same conditions. The HMI standard curve was accomplished by the mass balance technique. The relative
standard deviation (RSD) for the same sample through five sampling analyses by GC was less than 0.3%. HDC conversion (XHDC), HDI
yield (YHDI) and HMI Yield (YHMI) was calculated according to the Eqs. (1)-(3):
(1)
(2)

(3)
where CHDC is the concentration of HDC in the product sample, ppm; CHDI is the concentration of HDI in the product sample, ppm; CHMI
is the concentration of HMI in the product sample, ppm; mHDC is the mass of HDC fed to the autoclave, g; mps is the mass of product
sample in the solution for quantitative analysis, g; m
s
p
is the total mass of the solution for quantitative analysis, g; m is the total mass of
product, g; MHDC is the molecular weight of the HDC, g/mol; MHDI is the molecular weight of the HDI, g/mol; MHMI is the molecular
weight of the HMI, g/mol.
Acknowledgment
This work was supported by National Natural Science Foundation of China (Nos. 21476244 and 21406245) and Youth Innovation
Promotion Association CAS.
References
[
[
1] O. Kreye, H. Mutlu, M.A.R. Meier, Sustainable routes to polyurethane precursors, Green Chem. 15 (2013) 1431-1455.
2] E. Delebecq, J.P. Pascault, B. Boutevin, F. Ganachaud, On the versatility of urethane/urea bonds: reversibility, blocked isocyanate, and non-isocyanate
polyurethane, Chem. Rev. 113 (2013) 80-118.
[
[
3] J.S. Nowick, D.L. Holmes, G. Noronha, et al., Synthesis of peptide isocyanates and isothiocyanates, J. Org. Chem. 61 (1996) 3929-3934.
4] W.H. Lin, Y.S. Guo, S.A. Dai, An efficient one-pot synthesis of aliphatic diisocyanate from diamine and aiphenyl carbonate, J. Taiwan Inst. Chem. Eng. 50
(2015) 322-327.
[
[
[
5] T. Masuda, D. Saylik, L. Diebele, Production of aliphatic isocyanate, JP 6239826, 1994.
6] S.C. Miranda, C.C. Cabrero, E.F. Gutierrez, et al., Isocyanate production procedure, US 6639101, 2003.
7] D.L. Sun, J.Y. Luo, R.Y. Wen, J.R. Deng, Z.S. Chao, Phosgene-free synthesis of hexamethylene-1, 6-diisocyanate by the catalytic decomposition of
dimethylhexane-1, 6-dicarbamate over zinc-incorporated berlinite (ZnAlPO
8] M.J. Hyun, M. Shin, Y.J. Kim, Y.W. Suh, Phosgene-free decomposition of dimethylhexane-1, 6-dicarbamate over ZnO, Res. Chem. Intermed. 42 (2016) 57-
0.
9] C. Jeong, M.J. Hyun, Y.W. Suh, Activity of coprecipitated CuO/ZnO catalysts in the decomposition of dimethylhexane-1, 6-dicarbamate, Catal. Commun. 70
2015) 34-39.
10] Y. Cao, H.Q. Li, N.B. Qin, G.Y. Zhu, Kinetics of the decomposition of dimethylhexane-1, 6-dicarbamate to 1, 6-hexamethylene diisocyanate, Chin. J. Chem.
Eng. 23 (2015) 775-779.
11] Y. Yu, G.Q. Jin, Y.Y. Wang, X.Y. Guo, Synthesis of natural gas from CO methanation over SiC supported Ni-Co bimetallic catalysts, Catal. Commun. 31
4
), J. Hazard. Mater. 266 (2014) 167-173.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
7
(
(2013) 5-10.
12] L. Ren, T. Zhang, D.B. Liang, et al., Effect of addition of Zn on the catalytic activity of a Co/HZSM-5 catalyst for the SCR of NO
x 4
with CH , Appl. Catal. B:
Environ. 35 (2002) 317-321.
13] Y.N. Li, S.L. Liu, S.J. Xie, L.Y. Xu, Promoted metal utilization capacity of alkali-treated zeolite: preparation of Zn/ZSM-5 and its application in 1-hexene
aromatization, Appl. Catal. A: Gen. 360 (2009) 8-16.
2 3
14] A. Masalska, J.R. Grzechowiak, K. Jaroszewska, Effect of metal-support interactions in Ni/ZSM-5+Al O catalysts on the transformation of n-paraffins, Top.
Catal. 56 (2013) 981-994.
15] C.T. Hung, K.C. Lin, C.B. Wang, et al., Zeolite ZSM-5 supported bimetallic Fe-based catalysts for selective catalytic reduction of NO: effects of acidity and
metal loading, Adv. Porous Mater. 4 (2016) 189-199.
16] N. Rahimi, R. Karimzadeh, Catalytic cracking of hydrocarbons over modified ZSM-5 zeolites to produce light olefins: a review, Appl. Catal. A: Gen. 398
(2011) 1-17.
17] B. Rhimi, M. Mhamdi, V.N. Kalevaru, A. Martin, Synergy between vanadium and molybdenum in bimetallic ZSM-5 supported catalysts for ethylene
ammoxidation, RSC Adv. 6 (2016) 65866-65878.
18] X.J. Li, S.L. Liu, X.X. Zhu, et al., Effects of zinc and magnesium addition to ZSM-5 on the catalytic performances in 1-hexene aromatization reaction, Catal.
Lett. 141 (2011) 1498-1505.
19] M. Yao, N. Yao, Y. Shao, et al., New insight into the activity of ZSM-5 supported Co and CoRu bifunctional Fischer-Tropsch synthesis catalyst, Chem. Eng.
J. 239 (2014) 408-415.
20] L.F. Isernia, FTIR study of the relation, between extra-framework aluminum species and the adsorbed molecular water, and its effect on the acidity in ZSM-
5
steamed zeolite, Mater. Res. 16 (2013) 792-802.
21] X.H. Zhao, L. Wei, J. Julson, Z.R. Gu, Y.H. Cao, Catalytic cracking of inedible camelina oils to hydrocarbon fuels over bifunctional Zn/ZSM-5 catalysts,
Korean J. Chem. Eng. 32 (2015) 1528-1541.
22] S.H. Zhang, Z.X. Gao, S.J. Qing, S.Y. Liu, Y. Qiao, Effect of zinc introduction on catalytic performance of ZSM-5 in conversion of methanol to light olefins,
Chem. Pap. 68 (2014) 1187-1193.
[
[
[
23] C.W. Tang, C.B. Wang, S.H. Chien, Characterization of cobalt oxides studied by FT-IR, Raman, TPR and TG-MS, Thermochim. Acta 473 (2008) 68-73.
24] Z.J. Wang, H.M. Zhang, L.G. Zhang, et al., Low-temperature synthesis of ZnO nanoparticles by solid-state pyrolytic reaction, Nanotechnology 14 (2002) 11.
25] Z.Z. Zhu, G.Z. Lu, Z.G. Zhang, et al., Highly active and stable Co
2013) 1154-1164.
3 4
O /ZSM-5 catalyst for propane oxidation: effect of the preparation method, ACS Catal. 3
(
[
[
[
[
[
26] Q. Liu, L.C. Wang, M. Chen, et al., Dry citrate-precursor synthesized nanocrystalline cobalt oxide as highly active catalyst for total oxidation of propane, J.
Catal. 263 (2009) 104-113.
27] X. Huang, B. Hou, J.G. Wang, et al., CoZr/H-ZSM-5 hybrid catalysts for synthesis of gasoline-range isoparaffins from syngas, Appl. Catal. A: Gen. 408
(2011) 38-46.
28] A. Fisher, P. Goodall, M.W. Hinds, S.N. Nelms, D.M. Penny, Atomic spectrometry update. Industrial analysis: metals, chemicals and advanced materials, J.
Anal. Atom. Spectrom. 19 (2004) 1567-1595.
2+
29] B.M. Abu-Zied, Cu -acetate exchanged X zeolites: preparation, characterization and N
2011) 59-66.
2
O decomposition activity, Microporous Mesoporous Mater. 139
(
30] I. Othman, R. Mohamed, I.A. Ibrahim, M.M. Mohamed, Synthesis and modification of ZSM-5 with manganese and lanthanum and their effects on
decolorization of indigo carmine dye, Appl. Catal. A: Gen. 299 (2006) 95-102.
[
[
31] D. Li, P. Bui, H. Zhao, et al., Rake mechanism for the deoxygenation of ethanol over a supported Ni
32] P.C. Roy, W.H. Doh, S.K. Jo, C.M. Kim, Interaction of methanol and hydrogen on a ZnO (0001) single crystal surface, J. Phys. Chem. C 117 (2013) 15116-
5121.
2 2
P/SiO catalyst, J. Catal. 290 (2012) 1-12.
1

(f)
(e)
(d)
(c)
(b)
(a)
3
900 3600 3300
1600 1400 1200 11000 800 600 400
-
Wavenumber (cm )
Fig. 1. FTIR spectra of bare ZSM-5 (a), 2Co/ZSM-5 (b), Zn-2Co/ZSM-5 (c), Zn-Co/ZSM-5 (d), 2Zn-Co/ZSM-5 (e) and 2Zn/ZSM-5 (f) catalyst.

Co O
3 4

ZnO
.
ZnCo O

2
4









(
f)
.
.


. 
.
.
.
.

.
(e)
(d)
.
.




.
.

.
.

.


.
.
.
(c)
(b)



(a)
1
0
20
30
40
50
60
70
80
90
2
Theta (degree)
Fig. 2. XRD diffraction patterns of bare ZSM-5 (a), 2Co/ZSM-5 (b), Zn-2Co/ZSM-5 (c), Zn-Co/ZSM-5 (d), 2Zn-Co/ZSM-5 (e) and 2Zn/ZSM-5 (f) catalyst.
Fig. 3. Surface SEM images of bare ZSM-5 (a), 2Co/ZSM-5 (b), Zn-2Co/ZSM-5 (c) and 2Zn/ZSM-5 (d) catalyst.

112
4
65
270
4
50
4
25
(f)
(e)
(
d)
c)
(
(b)
(a)
0
100
200
300
400
500
600
Temperature (C)
3
Fig. 4. NH -TPD profiles of bare ZSM-5 (a), 2Co/ZSM-5 (b), Zn-2Co/ZSM-5 (c), Zn-Co/ZSM-5 (d), 2Zn-Co/ZSM-5 (e) and 2Zn/ZSM-5 (f) catalyst.
HDC Conversion
HMI Selectivity
HDI Selectivity
By-products Selectivity
1
00
8
6
4
2
0
0
0
0
0
Run 1
Run 2
Run 3
Run 4
Run 5
Fig. 5. Reusability of Zn-2Co/ZSM-5 catalyst in the thermal decomposition of HDC to HDI. Reaction condition: HDC concentration in chlorobenzene, 3.5%;
o
catalyst, 0.25 g; temperature 230 C; nitrogen flow rate, 800 mL/min; pressure, 0.68 MPa; time 2.5 h.

Scheme 1. Thermal decomposition of HDC to HDI.
Table 1
a
Physicochemical properties of catalysts and their catalytic performances for thermal decomposition of HDC .
BET
surface
Selectivity (%)
Total pore
Average
pore size
(nm)
HDC
Conversion
(%)
Total acidity
Weak
(%)
Moderate Strong
Catalyst
ZSM-5
volume
area
(mmol NH
3
/g-cat.)
(%)
(%)
3
(
cm /g)
HDI
HMI
By-products
2
(
m /g)
355.3
254.4
293.6
297.1
0.216
0.175
0.184
0.179
0.159
0.143
2.43
2.52
2.50
2.44
2.48
2.59
1.11
1.28
0.47
1.24
0.99
1.09
61.8
51.2
51.1
43.3
41.4
32.1
0.0
0.0
48.9
27.5
28.3
29.4
38.2
48.8
0.0
29.2
30.3
38.5
89.2
100
100
100
100
100
49.0
81.5
85.0
83.2
87.1
91.2
39.2
5.9
3.9
6.8
5.5
7.5
11.8
12.6
11.1
10.0
7.4
2
2
Co/ZSM-5
Zn/ZSM-5
Zn-Co/ZSM-5
2
Zn-Co/ZSM-5 255.6
Zn-2Co/ZSM-5 221.5
1.3
a
Reaction condition: HDC concentration in chlorobenzene, 3.5%; catalyst, 0.25 g; temperature 230 °C; nitrogen flow rate, 800 ml/min; pressure, 0.68 MPa; time
.5 h.
2
Fig. 3 shows the surface SEM images of ZSM-5, monometallic supported ZSM-5 and Zn-2Co/ZSM-5 catalysts. The bare ZSM-5 had
typical orthogonal crystal shape. While for monometallic and Zn-2Co/ZSM-5 catalyst, all the catalysts consisted of agglomerate particles.
Compared to 2Co/ZSM-5 and Zn-2Co/ZSM-5 catalyst, more loose metal oxide particles were observed on the surface of 2Zn/ZSM-5
catalyst which resulted in high surface area of 2Zn/ZSM-5 catalyst than other catalysts. The surface EDS images of monometallic
supported ZSM-5 and Zn-2Co/ZSM-5 catalysts showed the signal of Zn and/or Co, O, Si and Al, suggesting that metal species are also
present on the exterior surface of ZSM-5 support as shown in Fig. S3 in Supporting information. The weight percent of Zn and/or Co
obtained by EDS analysis was in accordance with the surface texture properties. As the weight percent of metal species increased on the
exterior surface of the ZSM-5, BET surface area and total pore volume also increased and followed the order of 2Zn/ZSM-5 > 2Co/ZSM-
5
> Zn-2Co/ZSM-5. This observation further confirmed that Co
3 4 2 4
O and ZnCo O species were mostly deposited inside ZSM-5 pores as
compared to ZnO and ZnO species existed on the exterior surface of ZSM-5 which caused less decrease in BET surface area.
<InlineImage4><InlineImage5>