Enhanced activity of CuO/ZnO catalyst on the decomposition of dimethylhexane-1,6-dicarbamate into dimethylhexane-1,6-diisocyanate

Article abstract of DOI:10.1007/s11164-018-3382-4

To decompose dimethylhexane-1,6-dicarbamate (HDC) to hexamethylene diisocyanate (HDI), mixed CuO and ZnO catalysts with Cu/(Cu + Zn) ratio of 4 and 8% were prepared by coprecipitation (CP), sequential precipitation (SP) and incipient wetness impregnation (IW). The SP-derived CuO/ZnO catalysts showed higher HDC yields than those derived by CP and IW. The IW method produced CuO/ZnO catalysts consisting of larger CuO and ZnO particles compared to the two precipitation methods. The CP method led to substitution of Zn²⁺ by Cu²⁺ in the hydrozincite precursor phase, resulting in higher BET and Cu surface areas of CuO–ZnO catalysts due to intimate intergrowth of nano-sized particles. However, the inherent character of ZnO in the CP-derived catalysts was modified by interfacial contact between CuO and ZnO identified by UV–visible and Raman spectra. In contrast, the properties of CuO and ZnO, as well as the relatively large surface areas, were kept in the SP-derived catalysts owing to deposition of Cu precipitates to fully aged Zn precipitates. This is believed to be a benefit of the SP method for the reaction. Therefore, our preparation approach has great potential to be extended to various mixed oxide catalysts.

Full text of DOI:10.1007/s11164-018-3382-4

Res Chem Intermed
https://doi.org/10.1007/s11164-018-3382-4
Enhanced activity of CuO/ZnO catalyst
on the decomposition of dimethylhexane-1,6-
dicarbamate into dimethylhexane-1,6-diisocyanate
Jinsung Kim1 Cheonwoo Jeong1 Jae-Hong Lee1
Young-Woong Suh1,2
•
•
•
Received: 10 September 2016 / Accepted: 18 January 2017
Ó Springer Science+Business Media B.V., part of Springer Nature 2018
Abstract To decompose dimethylhexane-1,6-dicarbamate (HDC) to hexam-
ethylene diisocyanate (HDI), mixed CuO and ZnO catalysts with Cu/(Cu ? Zn)
ratio of 4 and 8% were prepared by coprecipitation (CP), sequential precipitation
(SP) and incipient wetness impregnation (IW). The SP-derived CuO/ZnO catalysts
showed higher HDC yields than those derived by CP and IW. The IW method
produced CuO/ZnO catalysts consisting of larger CuO and ZnO particles compared
to the two precipitation methods. The CP method led to substitution of Zn2? by
Cu2? in the hydrozincite precursor phase, resulting in higher BET and Cu surface
areas of CuO–ZnO catalysts due to intimate intergrowth of nano-sized particles.
However, the inherent character of ZnO in the CP-derived catalysts was modified by
interfacial contact between CuO and ZnO identified by UV–visible and Raman
spectra. In contrast, the properties of CuO and ZnO, as well as the relatively large
surface areas, were kept in the SP-derived catalysts owing to deposition of Cu
precipitates to fully aged Zn precipitates. This is believed to be a benefit of the SP
method for the reaction. Therefore, our preparation approach has great potential to
be extended to various mixed oxide catalysts.
Keywords Carbamate decomposition Á CuO/ZnO Á Sequential precipitation Á
Coprecipitation
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11164-
018-3382-4) contains supplementary material, which is available to authorized users.
Jinsung Kim and Cheonwoo Jeong have contributed equally to this work.
&
Young-Woong Suh
ywsuh@hanyang.ac.kr
1
2
Department of Chemical Engineering, Hanyang University, Seoul 04763, Republic of Korea
Research Institute of Industrial Science, Hanyang University, Seoul 04763, Republic of Korea
123

J. Kim et al.
Introduction
Since isocyanates are utilized as a key material for polyurethane, their manufacture
is important in chemical industry. Isocyanate compounds are usually synthesized by
a commercial phosgene (COCl2) process based on the reaction of amine with
phosgene releasing HCl. Although this process is very popular, it has several
problems such as the use of toxic gas and separation of corrosive HCl. Among non-
phosgene processes studied in recent years [1], the two-step route using carbamate
as a reactant is very attractive [1, 2]. However, a high selectivity to isocyanate is
very difficult because it can react with alcohol or itself yielding byproducts such as
dimer or polymer due to high reactivity. Therefore, efficient catalysts should be
developed in order to minimize byproducts and thereby increase isocyanate yield.
In our previous work coprecipitated CuO–ZnO catalyst was applied for thermal
decomposition of dimethylhexane-1,6-dicarbamate (HDC) to hexamethylene diiso-
cyanate (HDI) to investigate the effect of precursor catalyst characteristics on the
activity and properties of CuO–ZnO [3]. This catalyst is usually synthesized by the
coprecipitation method consisting of a series of steps including precipitation,
ageing, drying and calcination [4]. At the first step, primary particles are
precipitated by mixing an aqueous solution of Cu2? and Zn2? solution and a basic
solution of precipitation agent. In the ageing step, the resulting amorphous
precipitate is transformed into meso-structured crystalline particles [5]. After
drying, the sample called a precursor catalyst has several crystal structures
depending on the Cu/Zn ratio. Further calcination and reduction transform the dried
sample into nano-structured CuO–ZnO and highly dispersed Cu/ZnO catalyst,
respectively [6].
A number of variables in these steps are known to influence the catalytic
performance of reduced catalyst, which is called ‘‘chemical memory’’ [7]. Many
studies have been conducted to understand the relationship between the catalytic
activity and the properties of precursor catalyst [8]. However, the correlation
between the property and activity of oxide catalyst is less focused due to the well-
known crystal structure of CuO and ZnO. Similar to the Cu/ZnO system, CuO–ZnO
catalyst has a synergistic effect between Cu2? and Zn2?, which was investigated in
CO oxidation [9]. We also examined the promoting (Cu rich) and inhibiting (Zn
rich) effects of CuO–ZnO catalysts depending on Cu/Zn ratios in HDC decompo-
sition [3]. However, CuO–ZnO catalysts with no less than 10% Cu showed lower
activities than pure ZnO because of interfacial contact between CuO and ZnO
particles. Our further effort has been made on the enhanced activity of CuO–ZnO
catalysts containing less than 10% Cu. In this report mixed CuO and ZnO catalysts
with 4 and 8% Cu were prepared using two different precipitation methods, as well
as impregnation method for comparison. These catalysts were tested in HDC
decomposition (Scheme 1) and characterized by PXRD, SEM–EDS, UV–Vis and
Raman spectroscopy, BET and N2O–RFC. Consequently, discussion based on all
results was made to relate the interaction between CuO and ZnO with the catalytic
activity.
123

Enhanced activity of CuO/ZnO catalyst on the…
Scheme 1 Reaction scheme on decomposition of dimethylhexane-1,6-dicarbamate (HDC) to
hexamethylene monoisocyanate (HMI) and hexamethylene diisocyanate (HDI)
Experimental
Sample preparation
Mixed CuO and ZnO catalysts, denoted as X(y), were prepared via coprecipitation
(CP), sequential precipitation (SP) and incipient wetness impregnation (IW), where X
denotes the catalyst preparation method and the y value stands for the Cu atomic
percentage (100 9 [Cu]/{[Cu] ? [Zn]}). For the synthesis of CP(8) precursor, an
aqueous solution (totally 1.2 M and 175 ml) containing Cu(NO3)2Á3H2O (4.10 g) and
Zn(NO3)2Á6H2O (58.65 g) was added dropwise (14 ml min-1) into NaHCO3 solution
(0.1 M, 4200 ml) at 70 °C for 17 min under vigorous stirring. After complete
addition, the resulting suspension was aged at 70 °C for 90 min. The CP(4) and
CP(12) precursors, as well as ZnO precursor (hydrozincite, HZ) were prepared in the
same manner except the weights of Cu(NO3)2Á3H2O and Zn(NO3)2Á6H2O added. To
prepare SP(8) precursor, aqueous solutions of Cu(NO3)2Á3H2O and Zn(NO3)2Á6H2O
(each 1.2 M) were prepared separately, where the volume of water was 14 and
161 mL, respectively. Zn2? solution was first injected into NaHCO3 solution (0.1 M,
4200 mL) at 70 °C and aged at the same temperature for 60 min, followed by
injecting Cu2? solution and ageing for 30 min. The same procedure was used for
preparation of SP(4) except the water volumes of 7 and 168 ml for Cu2? and Zn2?
solutions, respectively. Then, the CP- and SP-derived precipitates were filtered and
washed repeatedly with deionized water. The obtained cake was dried at 105 °C
overnight, followed by crushing and sieving to\ 200 lm. Finally, the calcination
was carried out at 400 °C (ramping rate 5 °C min-1) for 3 h.
For IW method, the prepared ZnO was used as a support material. IW(4) and
IW(8) catalysts were prepared by addition of Cu2? solution (ca. 2 mL) containing
Cu(NO3)2Á3H2O (0.32 and 0.67 g, respectively) into ZnO (2 g) until ZnO surface
123

J. Kim et al.
was pretty wet. The impregnated sample was dried at 105 °C overnight and calcined
at 350 °C (ramping rate 5 °C min-1) for 3 h.
Characterization
Powder X-ray diffraction (PXRD) analysis was performed with a D8 Discover with
GADDS diffractometer (Bruker) using Cu Ka as a radiation source (40 kV and
40 mA). Field emission scanning electron microscopy (FE-SEM) images were
taken in a FEI Nova NanoSEM 450 microscope with an electron dispersive X-ray
spectrometer (EDS) operated in the range 10–15 kV after the samples were coated
by Pt. Ultraviolet–visible (UV–Vis.) spectra were obtained by a Perkin Elmer
Lambda 650S. Raman scattering measurement was performed with a LabRam
Aramis Micro Raman Spectrometer (Horriba Jobin–Yvon) using an Argon ion laser
(532 nm) with 5 mW excitation. The Brauner–Emmett–Teller (BET) surface area
(SBET) of the sample (ca. 0.1 g) was measured in a Micromeritics 3Flex after
pretreatment at 105 °C for 1 h under vacuum. For measuring Cu surface area (SCu),
N2O reactive frontal chromatography (N2O–RFC) experiments were conducted in a
BELCAT-B (BEL JAPAN, Inc). After reducing the sample (0.1 g) at 250 °C with
10% H2 in Ar (30 mL min-1) and cooling to 40 °C under He, 1% N2O in He
(5 mL min-1) was introduced and the released nitrogen (m/z = 28) was measured
by a MS detector (BEL-Mass). The calculation was based on the reaction
stoichiometry between copper and oxygen (Cu/O = 2) and the copper surface
density of 1.46 9 1019 Cu atom m-2
.
Activity test
The HDC decomposition experiment was conducted in a single-neck round-bottom
flask into which the solvent (MarlothermÒ SH, Sasol Germany Gmbh) 60 g, HDC
20 g and catalyst 0.025 g were charged. The reactor was connected to a distillation
column and further to a sample tube linked with a liquid nitrogen trap and a rotary
pump for the system under vacuum. When the column temperature was close to
110 °C, the reaction mixture was heated to 230 °C for ca. 30 min and then
maintained for 2 h. After cooling to room temperature, the sample was taken out,
mixed with N-methyl-2-pyrrolidone for GC calibration, and finally analyzed with a
Younglin YL6500 GC equipped with a FID detector and an HP-5 column. The HDC
conversion, and HDI selectivity and yield were calculated from the quantities of
HDC, HDI and HMI (mono-isocyanate product).
Results and discussion
Figure 1 shows the HDI yields obtained using the CP-, SP- and IW-derived catalysts
(HDC conversion and HDI selectivity in Table S1 of Supplementary Material). The
HDI yields of CP-derived oxides (68.9, 68.8, 64.8, and 27.0% for ZnO, CP(12),
CP(30) and CuO, respectively) were similar to our previous results [3], although the
reaction system was a little different. More importantly, the activities of CP(4) and
123

Enhanced activity of CuO/ZnO catalyst on the…
Fig. 1 HDI yields of the CP-, SP- and IW-derived catalysts, as well as CuO and ZnO
CP(8) were higher (HDI yield of ca. 75%) than that of ZnO. This indicates that the
small amount of Cu2? present in ZnO promotes the HDC decomposition. In our
previous work, the coprecipitated CuO–ZnO catalysts (Cu/Zn ratio = 10/90 and
30/70) showed the lower HDC conversion compared to physically mixed CuO and
ZnO catalysts, due to interfacial contact between individual CuO and ZnO particles
caused by high Cu2? incorporation into ZnO [3]. Therefore, the enhanced activities
of CP(4) and CP(8) catalysts can be explained by the presumption that when the
CuO content is small, the interaction between CuO and ZnO benefits the catalytic
activity even if their interfacial contact exists. This is supported by the activities of
SP- and IW-derived catalysts. For both types of CuO/ZnO catalysts, the activity
enhancement was confirmed: IW(4) and SP(4) exhibited the HDI yields of 74.0 and
75.3%, respectively, and the SP(8) catalyst showed the highest HDI yield of 83.8%.
This finding implies that the interaction between CuO and ZnO affecting the
catalytic activity can be controlled by adjusting the method to add Cu2? into ZnO
matrix.
To investigate the properties of the prepared catalysts and thereby understand the
main reason for enhanced activity of CuO-promoted ZnO, several characterization
works were conducted. Firstly, uncalcined CP- and SP-derived precursors were
characterized by PXRD analysis. In all diffraction patterns, the reflections of
hydrozincite (Zn5(OH)6(CO3)2 (JCPDS PDF#19–1458) at 13.07°, 32.90° and 36.19°
were mainly observed. The reflections of malachite Cu2(OH)2CO3 (JCPDS
PDF#41–1390) at 14.83° and 17.58° were developed by increasing the Cu content
only in the SP-derived precursors (Fig. 2a). It was reported that malachite is not
found for coprecipitated Cu, Zn precursors with small Cu contents because Zn2? is
substituted by Cu2? in the hydrozincite up to 30% Cu [7, 10], like PXRD patterns of
our CP-derived precursors. Zn2? substitution by Cu2? was examined at the shift of
˚
the (002)-reflection of hydrozincite: the shift from 33.32° (2.689 A) to 33.38°
(2.684 A) occurred for the CP-derived precursors whereas there was very little shift
˚
for the SP-derived precursors (Fig. 2b, c). Therefore, we believe that the
substitution of Zn2? by Cu2? is not possible by the SP method. Very differently,
123

J. Kim et al.
Fig. 2 PXRD patterns of the CP- and SP-derived Cu, Zn precursors in the 2h range of a 5° to 50° and b,
c 33.0° to 33.6°, where the dashed line indicates the (002)-reflection of hydrozincite (HZ)
the typical reflections of ZnO (JCPDS PDF#36–1451) at 31.72°, 34.38° and 36.21°
and those of Cu(NO3)3(OH) (JCPDS PDF#45–0594) at 12.76° and 25.61° were seen
in PXRD patterns of the uncalcined IW-derived precursors (Fig. S1).
Also, PXRD patterns of the prepared CuO–ZnO catalysts after calcination at
400 °C are shown in Fig. 3. The major reflections in all patterns were for ZnO
structure due to its large amounts, similar to PXRD results of uncalcined precursors.
When calculated by the Scherrer equation using the most intense (101)-reflection of
ZnO at 36.21°, ZnO crystallite sizes were larger for IW-derived catalysts
(29.1–30.9 nm) than for pure ZnO (19.8 nm), whereas those derived by SP
(16.6–17.4 nm) and CP (14.3–15.7 nm) were smaller. Similarly, the intensity of
CuO reflections at 35.49° and 38.82° decreased in the following order: IW [ SP [
CP. Since it was too hard to calculate the crystallite size of CuO due to very low
peak intensities, the Cu surface area was measured by N2O–RFC method. In case of
the CP-derived catalysts, the SCu value increased from 3.4 to 5.5 m2 g-1 as the Cu
content increased from 4 to 12%. In contrast, the SCu values of SP-derived catalysts
were close to 3.0 0.2 m2 g-1 and those of IW-derived catalysts were negligible,
irrespective of the Cu content. This indicates that the CP method produces
individual CuO and ZnO particles in smaller sizes compared to the SP and IW
methods due to intimate intergrowth of these particles explained in our previous
report [3].
123

Enhanced activity of CuO/ZnO catalyst on the…
Fig. 3 XRD patterns of the CP-, SP- and IW-derived catalysts, where the reflection near 2h of 38.80°
corresponding to CuO(111) in each pattern was enlarged
Fig. 4 a SEM images of the CP-, SP- and IW-derived catalysts, and b EDS mapping images of CP(8),
SP(8) and IW(8) represented by yellow and red dots for Cu and Zn atoms, respectively
123

J. Kim et al.
The PXRD results were confirmed by SEM–EDS analysis. Like the images of
Zn-rich oxides in binary Cu/Zn system [3], the FE-SEM images of CP-derived
catalysts show the particles of aggregated platelet-like morphology, where Cu and
Zn are homogeneously distributed (top row in Fig. 4). However, the two different
morphology types were found for the SP-derived catalysts; the aggregated platelet-
like morphology corresponds to ZnO and elongated cuboid-like to CuO, revealed by
EDS mapping (middle row in Fig. 4). Very different from these catalysts, the IW-
derived catalysts were composed of many aggregates at the particle surface in which
Cu species existed locally (bottom row in Fig. 4). This results possibly from the fact
that the original structure of ZnO is collapsed by addition (i.e., impregnation) of an
acidic Cu(NO3)2 solution into ZnO and further calcination at 350 °C to remove
nitrate ions. Because of this effect, the BET surface areas of IW(4) and IW(8)
catalysts were 7 and 3 m2 g-1, respectively. Note that the SBET value of pure ZnO
was 23 m2 g-1 and the SBET values of CP-derived catalysts (28–29 m2 g-1) were a
little higher than those of SP-derived ones (25–28 m2 g-1). Therefore, the CP
method would be considered to be better than the SP method in yielding more nano-
sized CuO and ZnO particles.
These results cannot explain the higher activity of catalysts derived by SP
compared to that of CP. Thus, we conducted UV–Vis and Raman spectroscopy
measurements. Figure 5 shows UV–Vis spectra of the prepared catalysts in the
range 200–800 nm. The absorption spectrum of CuO, showing a plateau in
250–720 nm [11], was very different from that of ZnO whose unique feature was a
sharp decline of the absorbance around 400 nm [11–13]. When the spectra of the
catalysts with the same Cu content were compared in each diagram of Fig. 5, the
absorbance of CP- and SP-derived catalysts were more intense than that of the
catalyst derived by IW in 250–720 nm. The more interesting finding was that as the
Cu content increased for CP-derived catalysts the spectrum approached that of CuO
even though small amounts of CuO were present. This was not found for SP- and
IW-derived catalysts still showing a step-down shaped curve with a little high
Fig. 5 UV–Vis spectra of the CP-, SP- and IW-derived catalysts, as well as CuO and ZnO
123

Enhanced activity of CuO/ZnO catalyst on the…
Fig. 6 Raman spectra of the CP-, SP- and IW-derived catalysts, as well as CuO and ZnO
baseline. Therefore, it is believed that the character of pure ZnO is somewhat
modified in the CP-derived catalysts.
Raman spectra of the prepared catalysts are presented in Fig. 6. The character-
istic Raman bands of ZnO were observed at 99.4 and 437.6 cm-1, and that of CuO
at 270.1 cm-1 [14]. The intensities of ZnO-related bands increased with the Cu
content for the IW-derived catalysts. In contrast, these intensities decreased for both
CP- and SP-derived catalysts, where more decrease was observed for the former.
This is due to size effect [15], hence explaining that the size of ZnO particles
increased in the following order of the preparation method: CP \ SP \ IW. This is
in good agreement with their PXRD results. An additional finding is a shift of
Raman band for E2Low mode [14] from 99.4 to 96.1 cm-1 for the CP-derived
catalysts with higher Cu contents, which was little observed for the SP- and IW-
derived catalysts (98.6–99.4 cm-1). It was reported that the lower energy side shift
of E2Low mode indicates the effect of doping on ZnO [14]. Such effect is caused by
interfacial contact of CuO and ZnO nanoparticles in the CP-derived CuO–ZnO
catalysts suggested in our previous report [3]. However, the SP method led to the
production of mixed CuO/ZnO catalysts retaining the property of pure ZnO due to
little intimate growth of CuO, where individual CuO and ZnO particles are
relatively small. Consequently, the activity difference between CP- and SP-derived
catalysts is explained by different ZnO nanostructure that is induced by Zn2?
substitution by Cu2? in ZnO matrix.
123

J. Kim et al.
Conclusion
We have prepared mixed CuO and ZnO catalysts with Cu contents of 4 and 8%
using three different methods and obtained their activities in the conversion of
dialkylcarbamate to the corresponding diisocyanate. The impregnated (IW) CuO/
ZnO catalysts showed the smallest BET surface areas due to larger CuO and ZnO
particles. In contrast, the coprecipitated (CP) CuO–ZnO catalysts contained the
largest BET and Cu surface areas due to intimate intergrowth of CuO and ZnO
nanoparticles. However, a change of unique ZnO character was observed in the CP-
derived catalysts, explaining interfacial contact between individual CuO and ZnO
particles, and thereby an activity decrease of CP-derived catalysts with high Cu
contents. The sequential precipitation (SP) method is a combination of IW and CP
methods in an elegant manner, enabling deposition of Cu precipitates to fully aged
Zn precipitates and thus keeping the properties of each CuO and ZnO in mixed
CuO/ZnO catalysts. This is responsible for the enhanced activities of the SP-derived
catalysts.
Acknowledgements This research was financially supported by the Fusion Research Program for Green
Technologies (NRF-2012M3C1A1054501), C1 Gas Refinery Program (NRF-2018M3D3A1A01018010)
and Basic Science Research Program (NRF-2016R1A6A1A03013422) through the National Research
Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning and by the
Ministry of Education.
References
1. O. Kreye, H. Mutlu, M.A.R. Meier, Green Chem. 15, 1431 (2013)
2. H. van den Berg, L. van der Ham, H. Gutierrez, S. Odu, T. Roelofs, J. de Weerdt, Chem. Eng. J.
207–208, 254 (2012)
3. C. Jeong, M.J. Hyun, Y.-W. Suh, Catal. Commun. 70, 34 (2015)
4. S. Schimpf, M. Muhler, in Synthesis of Solid Catalysts, ed. by K.P. de Jong (Wiley-VCH, Weinheim,
2009), p. 329
5. S. Zander, B. Seidlhofer, M. Behrens, Dalton Trams. 41, 13413 (2012)
6. M. Behrens, J. Catal. 267, 24 (2009)
7. B. Bems, M. Schur, A. Dassenoy, H. Junkes, D. Herein, R. Schlogl, Chem. Eur. J. 9, 2039 (2003)
8. M. Behrens, S. Zander, P. Kurr, N. Jacobsen, J. Senker, G. Koch, T. Ressler, R.W. Fischer, R.
¨
Schlogl, J. Am. Chem. Soc. 135, 6061 (2013)
9. D.M. Whittle, A.A. Mirzaei, J.S.J. Hargreaves, R.W. Joyner, C.J. Kiely, S.H. Taylor, G.J. Hutchings,
Phys. Chem. Chem. Phys. 4, 5915 (2002)
10. G.J. Millar, I.H. Holm, P.J.R. Uwins, J. Drennan, Faraday Trans. 94, 593 (1998)
11. J. Yu, S. Zhuang, X. Xu, W. Zhu, B. Feng, J. Hu, J. Mater. Chem. A 3, 1199 (2015)
12. T.R. Chetia, D. Barpuzary, M. Qureshi, Phys. Chem. Chem. Phys. 16, 9625 (2014)
13. C. Wang, D. Wu, P. Wang, Y. Ao, J. Hou, J. Qian, Appl. Surf. Sci. 325, 112 (2015)
14. R. Uadayabhaskar, B. Karthikeyan, J. Appl. Phys. 115, 154303 (2014)
15. Z. Sun, B. Zhao, J.R. Lombardi, Appl. Phys. Lett. 91, 221106 (2007)
123