Catalytic Conversion of CO₂/H₂ Into Alcohol using Modified Lampung – Zeolite as Catalyst

Introduction

In respond to fossil fuels depletion and concern on environmental impacts, primarily due to CO₂ produced from their utilization, conversion of CO₂ into various chemicals useful as alternative and renewable fuel has been intensively investigated in the last few decades.  For this purposes, different catalysts has been tested and in general, the results obtained indicate that catalysts play very important roles in governing the yield  and type of product resulted. As an example, Cu₂₀Mo₂C/M41 catalyst abled to produce 5.5% yield of methanol at a CO₂ conversion of 8.8%¹. Other groups of research using different kinds of catalysts such as Fe-Zn-Zr/HY composite catalyst abled to obtain isoalkanes at 340^oC as much as 20% CO₂ conversion, 45% selectivity to i-C₄, and 9% yield²; CuO-ZnO-TiO₂ catalyst  operated at 493 K gave 35% conversion of CO₂ with the selectivity of 52.3% to ethanol³; K/Cu-Zn-Fe catalyst abled to produced ethanol at 300^oC as much as 90% CO₂ conversion and 85% selectivity to ethanol⁴; Co-Fe bimetallic catalyst operated at 200 – 270^oC  obtained 27% of CO₂ conversion and 71% of CH₄ selectivity ⁵.

Another type of materials that exhibit catalytic activity is natural zeolite, that can be found in many different places around the world. In Indonesia, one of the regions that has natural zeolite deposit is Lampung Province. Due to their availability and chemical as well as physical characteristics, natural zeolites have been used as catalyst for different reactions, for examples, pure phase β-zeolite operated at 120^oC and atmospheric pressure abled to obtain 90% yielf of ethyl acetate from 80% acetic acid conversion⁶; Pacitan-Natural zeolite as catalyst support for transesterification of palm oil abled to produce 96% yield of both methyl palmitate and stearate at 60^oC⁷; and  H-Mordenite type of zeolite operated at 423K abled to produce 12% glucose from 0.05g Cellulose in 5 mL water ⁸.

To improve their catalytic performance, natural zeolites are commonly activated or modified by addition of other active sites.  Several studies revealed that addition of active sites significantly increased catalytic activity of natural zeolites. For examples,  clinoptilolite type of natural zeolite impregnated by KOH and operated at 350 – 440^oC and abled to gain 90% yield of liquid fuel from natural triacylglicerol⁹; modified natural zeolite from Kidul mountain of java island abled to convert ethanol as much as 15% and produced 3% yield of diethyl ether at 140^oC¹⁰; and  modified natural clinoptilolite  from Bayah – Indonesia operated at 60 – 80^oC and abled to obtain almost 100% glycerol carbonate  from 25% conversion of glycerol¹¹. In hydrogenation of CO₂ reaction, modified natural zeolite with 5% Nickel abled to produce methane with the 96% selectivity and 10% conversion of CO₂ at 300^oC reaction temperature ¹².

In recognizing the important role of additional active sites on performance of natural zeolites, in this study, Lampung zeolite was modified by addition of different amounts of iron oxides as active sites.  The modified zeolites were subsequently calcined at 600^oC and then tested as catalysts for conversion of  CO₂ into alcohol at a temperature range of 200 – 400^oC.

Material and Methods

The chemical used  in this study, natural zeolite, nitric acid,  Fe-nitrates, and NH₃ are reagents grade  obtained from Merck. The main equipments used in this study are Fourier Transform Infrared (FTIR) spectrophotometer (Shimadzu Prestige-21), Scanning Electron Microscope (SEM, Philips-XL), Philips X-ray diffractometer (XRD) model PW 1710 with Cu-Kα radiation, and Gas Chromatography (Agilent  7890B) with Flame Ionization detector.

Modification of Natural Zeolite

Modification of natural zeolite was carried out in two steps. The first treatment was carried out by soaking the sample in 0.1N nitric acid solution for 2 hours with continuous shaking. The mixture was filtered and the solid was repeatedly rinsed with distilled water  to remove the excess of acid.  At the second step, the solid was calcined   at 600^oC  for  2 hours  (increased by 2^oC min^-1), and then impregnated with 1 – 7% Fe₂O₃ by soaking zeolite in Fe-nitrates solution and heated at 80^oC until water solvent evaporated. Finally, the samples were recalcined at 600^oC for 2 hours.

Characterizations

XRD Analysis

The XRD powder patterns were obtained in a Philips  diffractometer Model PW 1710, using Cu K_a radiation and operating at 40 kV and 40 mA. The scanning range  of 2q angle was set from 5 to 70^o, with a step size of 0.02^o/ s. The crystalline phases identification of the samples was analysed and  calculated by using Search and Match  Program (Version 2.1) and Rietveld Method, respectively ¹³.

FTIR Analysis

Sample was treated in vaccumized dessicator after heating at 120^oC  and then, the vapour of liquid pyridine was allowed to contact the sample located inside of vaccumized dessicator for 24 hours. After taking out from dessicator and allowing on air for 2 hours, a small amount of sample was finally mixed to KBr powder (sample/KBr ratio less than 1%) to form a pellet at ambient temperature and was ready to be analysed by putting it on quartz cell of  FTIR (Shimadzu) in the wavenumber range of 400 – 4000 cm⁻¹. This procedure had already been used in the previous work ^{14, 15}.

SEM Analysis

A small quantity of the sample was mounted on a stub with an adhesive coated by 40 – 60 nm of  gold or other conducting materials. Then, the crystal morphology and size of the samples were observed ^{16, 17}.

Results and Discussion

Structural Analysis

Materials that have been prepared were characterized using some instruments such as  X – ray diffraction, (FTIR) spectroscopy, and  TEM as explained as follow:

X-Ray Diffraction Analysis

Qualitative XRD analysis was conducted by comparing the diffraction lines of  zeolite sample with those in standard from the  ICCD  files and showed that the major phase obtained for the sample is clinoptilolite phase. The minor phases was also obtained such as Quartz and cristobalite as shown in Figure 1 below. This washed  zeolite, however, cliptonilolite is accordance with JCPD File  # 39-1383¹⁸, which has a monoclinic structure.

Furthermore,  to  obtain quantitative information  from the zeolite, Rietveld analysis is used to obtain unit cell parameters and distinguish the phase composition clearly.  As shown in Figure 2, one of  the XRD Rietveld analysis has been plotted for the zeolite.  This result was satisfied by examining the goodness of fitting value (χ² ≤ 4).

Figure 1: Diffractogram  of natural zeolite after washing with 0.1N nitric acid in 2 hours and then calcining at  600oC measurements. Click here to View figure

 

Figure 2: XRD Rietveld Plot for zeolite sample after washing and calcining at 600oC. Click here to View figure

 

The Observed data  are shown by the (+) red sign, and the calculated data by a solid line (Green). The vertical lines (Yellow, Green and purple, indicates hkl indices of three phases), The purple line below the  vertical lines is the difference profile.

In addition, this fitting plot indicates the accomplishment of the Rietveld refinement, in which the different plot between calculated and observed patterns show relatively minor fluctuation. This result also proved that the output of the refinement can be used to provide further information such as phase composition, unit cell parameters, and atomic position in space.

Table 1: Rietveld Refinement results of  Lampung  Zeolite (R_wp = 21.51; R_p = 16.97 ; χ² = 1.317)

No.	Phase  Identification	Unit Cell Parameters							Crystal System	Weight Percentage (% wt)
		a ( Å )	b ( Å )	c ( Å )	V ( Å3 )	α	β	γ
1.	Clinoptilolite	19.586	19.177	7.888	2661.6	90o	116.7o	90o	Monoclinic	65.9
2.	Quartz	4.985	4.985	5.287	113.7	90o	90o	120o	Hexagonal	31.6
3.	Cristobalite	17.632	17.941	7.400	2097.0	90o	90o	90o	Tetragonal	2.5

 

Table 1 shows phase composition, unit cell parameters, crystal system, and weight percentage from Rietveld refinement with  XRD data for zeolite sample washed and calcined at 600^oC for 2 hours. Evidently, the weight percentage for clinoptilolite, quartz, and crystoballite phases is 65.9, 31.6, 2.5, respectively. The relatively high quantity of quartz is probably affected by both acid washing and calcining at 600^oC  since natural zeolite dehydrated  when it had been heated above 500^oC^13,19. Then, structrural phases, the samples, with different iron oxides content were characterized and the X-ray diffraction patterns of the samples were compiled in Fig. 3. As can be shown in Fig. 3, the diffractogram are generally similar, with only some differences in the intensity of certain position of the representative peaks of zeolite (clinoptilolite form). For instanse, at 2q of 32.93^o as a major peak of Fe₂O₃ (PDF-330664) is more pronounced as quantity of Fe₂O₃ – dopant  increased, as particular for 7% Fe₂O₃ dopant. In a ragion of 20^o through 33^o, as shown in Fig.3, represent clinoptilolite type and then the representative peaks (i.e intensity and 2q) changed slightly as dopant content augmounted.  The effect of dopant addition into zeolite parameter cells is investigated by Rietveld calculation.

Figure 3: Diffractograms of Fe2O3 doped zeolite with variation of dopant: (a). 1% Fe2O3, (b). 3% Fe2O3, (c). 5% Fe2O3, and (d). 7% Fe2O3 and a reference of Fe2O3 phase (black – bar ;  PDF330664). Click here to View figure

 

To evaluate the effect of the iron oxides quantities on the stability of Lampung – zeolite.

Quantitative analysis of modified zeolite using Rietveld method is presented in Table 2 below, and analysis of Rietveld calculation has shown a very good result since its Goodness of Fitting value (c² or GOF) had been obtained satisfying the required values of £ 4. The effect of Fe₂O₃ addition into zeolite, proved that structure of zeolite is not sufficiently influenced by the formation of hematite – Fe₂O₃ crystalline phase. In the other word, the formation of hematite-Fe₂O₃ is not happened exactly inside the channels of zeolite since the crystalline structure of a washed natural zeolite  has still existed. This evident is also found by Cheng et al., while loading of Fe and Ni into HZSM – 5 catalyst ²⁰.

Table 2: Rietveld refinement results of  Zeolite doped

No.	Zeolite added with	Unit cell Parameters				Crystalline Phase(s) dominant
		a(Å)	b(Å)	c(Å)	c2 (GOF)*
1	0% Fe2O3	19.586	19.197	7.888	1.317	clinoptilolite
2	1% Fe2O3	19.588	19.196	7.898	1.429	clinoptilolite
3	3% Fe2O3	19.587	19.198	7.888	1.547	clinoptilolite
4	5% Fe2O3	19.589	19.197	7.889	1.671	clinoptilolite
5	7% Fe2O3	19.587	19.199	7.886	1.537	Clinoptilolite & Hematite

 

Microstructure Analysis

The surface characteristic is very important to the catalytic application involving the surface interaction (adorption – desorption) process. For this reason, the samples investigated in this study were characterized using SEM technique¹⁷. The SEM micrographs of some samples are shown in Fig. 4. As can be seen in Fig. 4, The SEM micrographs of washed-zeolite and some Fe₂O₃ doped zeolite samples display clearly a porous  structure.  The topography of these samples told that its texture is relatively homogeneous and amorphous. Also, its morphology suggesting there is the effect of Fe₂O₃ dopant on its shape and channel.  However, by careful inspection of certain region, the varied diameter size of channel and some ductiles spots are observed. Furthermore, some ductiles has a diameter of 0.1 – 3 mm (red-arrow) and other have diameter even higher than 3 mm. Also there are some brittle channel parts (green-arrow) . This brittle channel part is happened  probably because of  washing, dopant-adding and calcining process.  In general the Fe₂O₃ addition into zeolite did not damage the channel structure of zeolite as can be seen.

Figure 4: Micrographs of washed zeolite (a) 3% Fe2O3 doped zeolite (b) and 5% Fe2O3 doped Zeolite (c) after calcining at 600oC for 2 hours (Mag. = 2.00k x), from left to right. Click here to View figure

 

Acidity Analysis

To identify the functional group present in the sample which is primarily to know the existence of Lewis and Bronsted-Lowry acid sites, the Fourier transform infra-red spectroscopy is applied in this study. These characteristics are very important to determine the performance of catalyst. The FTIR spectrum of washed – zeolite as an example is shown in Fig. 5.

Figure 5: Infra red spectra of washed zeolite after treating with Ammonia. Click here to View figure

 

As can be seen in Fig. 5 above, the absorption bands representing both  acid sites were proved by the existing of Zeolite – NH₃ (as Lewis acid site) and/or  zeolite – NH₄⁺(as BrØnsted – Lowry acid site) vibrations. In general, wavenumber region of 3100 – 3700 cm^-1 refers to stretching vibrations and wavenumber below  2500 cm^-1 relates to bending and rocking vibrations. In Fig. 5, Lewis acid sites appeared to wavenumber of 1636.95 cm^-1 and little peaks in region of 1450 – 1550 cm^-1. However, BrØnsted – Lowry acid sites  appeared at wavenumber of 1400 – 1440 cm^-1. Furthermore region below 1200 cm^-1 belongs to fingerprint pattern of zeolite. Strong stretching vibrations of the bond Si – O – Si, Al – O – Si,  and Al – O – Si appeared at the wavenumber range of 1000 – 1136 cm^-1 and 789 cm^-1. These peaks indicate that tetrahedral linkage arrangement in tetrahedron (Al,Si)O₄ still remain after washing and calcining at 600^oC. The peaks that were appeared at 476 and 608 cm^-1 show the bending vibrations of tetrahedral linkage.

Figure 6: FTIR spectra of some zeolites doped with: (a) 1% Fe2O3, (b) 3% Fe2O3, and (c) 5% Fe2O3 after treating with ammonia. Click here to View figure

 

In the Fe₂O₃ doped zeolites samples investigated, as can be seen in Fig. 6, the spectra are pratically similar in terms of the absorption bands for acid sites characteristics, with only some minor differences in intensities and wavenumbers, depending on the quantity of Fe₂O₃ doped into zeolite. The existence of Lewis acid sites is indicated by the vibration bands resulted from stretching vibration, located at 1628 cm^-1 for 1% Fe₂O₃ doped zeolite; 1650 and 1632 cm^-1 for 3% Fe₂O₃ doped zeolite, and then 1648 and 1628 cm^-1 for 5% Fe₂O₃ doped zeolite. While the existence of BrØnsted-Lowry acid sites is shown by the vibration bands resulted from stretching vibration,  located at 1466 and 1400  cm^-1 for 1% Fe₂O₃ doped zeolite; 1443 and 1404 cm^-1 for 3% Fe₂O₃ doped zeolite, and then 1466 and 1408 cm^-1 for 5% Fe₂O₃ doped zeolite. By comparing the intensities of the absorption bands corresponding to Lewis and BrØnsted-Lowry acid sites, it can be concluded that acid property of the Fe₂O₃ doped zeolite samples is dominated by the Lewis acid sites, and there is a tendency that the Lewis acid strength relatively increases with increased Fe content in the samples. The increase of acid properties  is also found while  Fe and Co was loaded in SiO₂²¹. In finger print region of spectra, the absorption band representing stretching and bending vibrations of Fe-O were detected at 613, 677, and 663 cm^-1 for 1%, 3%, and 5%  Fe₂O₃ doped, respectively^{22, 23}.

Figure 7: Alcohol results from CO2/H2 reaction using zeolite washed as catalyst. Index 1, 2, 3, and 4 refers to methanol, ethanol, propanol and butanol, respectively and standard deviation 2.4%. Click here to View figure

 

Catalytic Activity

Hydrogenation of carbon dioxide was run in the reactor (U-type) with the variation of temperature, 200-400^oC and the gas flow of 3 L/hour  consisting CO₂/H₂ ratio of 0.25. The results are displayed in Fig. 7 and Fig. 8 below.

Figure 8: Alcohol results from CO2/H2 reaction using varied Fe2O3 doped zeolite as catalyst. Index 1, 2, 3, and 4 refers to methanol, ethanol, propanol and butanol, respectively and standard deviation 2.4%. Click here to View figure

 

As can be seen in Fig. 7, zeolite washed is active to convert CO₂ into alcohol eventhough its activity was low, 1000 ppm  of methanol at 200^oC was obtained. It can be explained that CO₂ and H₂ molecules was attached strong enough to form C₁ ionic molecule on the surfaces of catalyst before forming CH₃OH molecules. In higher temperature, 300 and 400^oC, its activity became very low. It can be implied that as temperature increase the contact time between CO₂ and H₂ molecules into zeolite surfaces shorter so that there are no sufficient interaction to produce more methanol.  Other alcohols such as ethanol, propanol, and butanol were also formed but its amount was too small compared to methanol formation at 200^oC. Besides insufficient interaction, it can also be explained that  the most  methanol formed was oxidized in advance as temperature increased.

Scheme 1: Illustration of alcohols formation in CO2/H2 reaction on Lampung Zeolite Catalyst Click here to View figure

 

The effect of Fe₂O₃ dopant in zeolite activity to convert CO₂ into alcohol is shown in Fig. 8. In general, the activity of zeolite is abruptly increased by adding Fe₂O₃. Eventhough, Fe₂O₃ itself that has been prepared with the same method did not show a good activity.  In addition, by looking at certain temperature, for instance 200^oC, the alcohol result is quite well distributed using 5% Fe₂O₃ doped Zeolite. The well distribution of alcohol results (i.e. CH₃OH, C₂H₅OH, C₃H₇OH, and C₄H₉OH) using 5% Fe₂O₃ doped zeolite can be explained that both CO₂ and H₂ molecules adsorbed on the surface of this catalyst have enough time to form  C⁺, C₂⁺ , C₃⁺ and C₄⁺ species, so higher alcohol can be obtained as illustrated on Scheme 1 below ^{24, 25}.  Another reason why a longer chain of alcohol was not formed greater than that of methanol is the Lewis acid sites of zeolite located at relatively high wavenumber, about 1640 cm^-1, plays an important role in binding CO₂ and H₂ molecules on the surface. Therefore, further oxidation is happened.

Conclusion

It can be concluded that modified zeolite of Lampung is active as a catalyst for converting CO₂ and H₂ into alcohol in the temperature range of 200 – 400^oC. By adding and varying Fe₂O₃ content into zeolite, its activity is ameliorated and the higher alcohol result is formed. Regarding on its acid behaviour,  it can be implied that the activity of modified Lampung-zeolite becomes better as acid characteristic more pronounced.

Acknowledgement

The author wish to thank and appreciate the Directorate General Higher Education Republic of  Indonesia for research funding provided through Competitive Research Grant program with a contract no. 267/H26/PL/2016.

References

1. Liu, X.; Song, Y.; Geng, W.; Li, H.; Xiao, L.; Wu, W. Catal. 2016, 6, 75 – 87.
   CrossRef
2. Ni, X.; Tan, Y.; Han, Y.; Tsubaki, N. Catal. Commun. 2007, 8, 1711 – 1714.
   CrossRef
3. Xiao, J.; Mao, D.; Guo, X.; Yu, J. Energy Technol. 2015, 3, 32 – 39.
   CrossRef
4. Li, S.; Guo, H.; Luo, C.; Zhang, H.; Xiong, L.; Chen, X.; Ma, L. Catal. Lett. 2013, 143, 345 – 355.
   CrossRef
5. Muthu, K.; Gnanamani, G.J.; Hamdeh ,H.H.; Shafer, W.D.; Liu, F.; Hopps, S.D.; Thomas, G.A.;  Davis, B.H.  ACS Catal. 2016, 6, 913 – 927.
   CrossRef
6. Yue, Y.; Liu, H.; Zhou, Y.; Bai, Z.; Bao, X. Appl. Clay Scie. 2016, 126, 1 – 6.
   CrossRef
7. Kusuma, R.I.; Hadinoto, J.P.; Ayucitra, A.; Soetaredjo, F.E.; Ismadji, S.  Appl. Clay Scie.  2013,74, 121 – 126.
   CrossRef
8. de Vyvwe, S.V.; Peng, L.; Geboers, J.; Schepers, H.; de Clippel, F.; Gommes, C.J.; Goderis, B.; Jacobs, P.A.; Sels B.F. Green Chem. 2010,12, 1560.
   CrossRef
9. Cvengroš , J.; Buzetzki, E.; Švaňová, K. 44^th Intern. Petrol. Conf. 2012, Bratislava, Slovak Republic, September 21 – 22^nd, 1 – 8.
10. Widayat, W.; Roesyadi, A.; Rachimoellah,  M. Internat. J. Sci. Eng. 2013, 4, 6 – 10.
11. Mahdi, H.I.; Irawan, E.; Nuryoto, N.; Jayanudin, J.; Sulistyo, H.; Sediawan, W.B.; Muraza, O. Waste and Biomass Valor. 2016, 7, 1349 – 1356.
    CrossRef
12. Bacariza, M.C.; Garcia, I.; Westermann, A.; Ribeiro, M.F.; Lopes, J.M.; Henriques, C. Topics in Catal. 2016, 59, 314 – 325.
    CrossRef
13. Hutching, G.J.; Comninos, H.; Copperthwaite, R.G.; van Rensburg, R.; Hugo, M.  J. Appl.Crystallogr. 1969, 2, 65 – 71.
    CrossRef
14. Parry E.P., J. Catal. 1963, 2, 371 – 379.
    CrossRef
15. Situmeang,  R.; Manurung, P.; Sulistiyo, S.T.; Hadi, S.; Simanjuntak, W.; Sembiring, S.  Asian J. Chem. 2015, 27, 1138 – 1142.
    CrossRef
16. Drbohlavova,  J.; Adam, V.; Kizek, R.; Hubalek, J.;  Int. J.  Mol. Sci. 2009, 10, 656 – 673.
    CrossRef
17. Hanke L.D., Handbook of Analytical Methods for Materials. Materials Evaluation and Engineering Inc. Plymouth, USA  .2001,35 – 38.
18. Gottardi G., Galli E., Natural Zeolite: Pottasium Sodium Calcium Alumminum Silicate hydrate (Clinoptilolite), Springer – Velag, Tokyo. 1985, 341.
19. Breger, I.A.; Chandler, J.C.; Zubovic, P.  The Am. Miner. 1970, 55, 826 – 840.
20. Cheng, S.; Wei, L.; Julson, J.; Muthukumarappan, K.; Kharel, P.R. Fuel Proc. Tehcn. 2017, 167, 117 – 126
    CrossRef
21. Cheng, S.; Wei, L.; Julson, J.; Rabnawaz, M. Energ. Conv. Manag. 2017, 150, 331 – 342.
    CrossRef
22. Zhiqiang, W.; Hongxia, Q.; Hua, Y.; Cairong, Z.; Xiaoyan, Y. J. Alloys. Compd. 2009, 479, 855 – 858.
    CrossRef
23. Kim, J.S.; Ahn, J.R.; Lee, C.W.; Murakami, Y.; Shindo, D. J. Mater. Chem. 2001, 11, 3373 – 3376.
    CrossRef
24. Kienneman, A.; Boujana, S.V.S.; Diagne, C.; Chaumette, P. Stud. Surf. Sci. Catal. 1993, 75B, 1479 .
    CrossRef
25. Jung, K.T.; Bell, A.T. J. Catal. 2001, 204, 339 – 347.
    CrossRef

---

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

---

This work is licensed under a Creative Commons Attribution 4.0 International License.