DSpace at VNU: Response Surface Optimization of Impregnation of Blended Amines into Mesoporous Silica for High-Performance CO2 Capture

DSpace at VNU: Response Surface Optimization of Impregnation of Blended Amines into Mesoporous Silica for High-Performan...

Energy & Fuels is published by the American Chemical Society 1155 Sixteenth Street N.W., Washington, DC 20036

Published by American Chemical Society Copyright © American Chemical Society However, no copyright claim is made to original U.S Government works, or works produced by employees of any Commonwealth realm Crown government in the course

Response Surface Optimization of Impregnation of Blended

Amines into Mesoporous Silica for High-Performance CO2

Capture

Duc Sy Dao, Hidetaka Yamada, and Katsunori Yogo

Energy Fuels, Just Accepted Manuscript • Publication Date (Web): 22 Jan 2015

Downloaded from http://pubs.acs.org on January 26, 2015

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication They are posted

online prior to technical editing, formatting for publication and author proofing The American Chemical

Society provides “Just Accepted” as a free service to the research community to expedite the

dissemination of scientific material as soon as possible after acceptance “Just Accepted” manuscripts

appear in full in PDF format accompanied by an HTML abstract “Just Accepted” manuscripts have been

fully peer reviewed, but should not be considered the official version of record They are accessible to all

readers and citable by the Digital Object Identifier (DOI®) “Just Accepted” is an optional service offered

to authors Therefore, the “Just Accepted” Web site may not include all articles that will be published

in the journal After a manuscript is technically edited and formatted, it will be removed from the “Just

Accepted” Web site and published as an ASAP article Note that technical editing may introduce minor

changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers

and ethical guidelines that apply to the journal pertain ACS cannot be held responsible for errors

or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

1

Response Surface Optimization of Impregnation of Blended Amines into Mesoporous Silica for High-

Duc Sy Dao,†,‡ Hidetaka Yamada,§ and Katsunori Yogo*,†,§

†Graduate School of Materials Science, Nara Institute of Science and Technology (NAIST), 8916-5 Takayama-cho, Ikoma, Nara 630-0192, Japan

‡Faculty of Chemistry, Hanoi University of Science, VNU, 19 Le Thanh Tong, Hoan Kiem, Hanoi, Vietnam

619-0292, Japan

*To whom correspondence should be addressed Tel: +81-774-75-2305; Fax: +81-774-75-2318

E-mail: yogo@rite.or.jp (K Yogo)

ABSTRACT: We used the response surface method (RSM) to optimize the conditions for the

impregnation of blended amines into mesostructured cellular silica foam (MSU-F) to prepare effective solid sorbents for CO2 capture The effects of the amounts of tetraethylenepentamine (TEPA),

were investigated The influences of these three independent variables and their interactions were determined using the RSM; the optimum amounts of TEPA, DEA, and MeOH were 1.33 g (7.03 mmol), 0.85 g (8.08 mmol), and 58.2 g, for the preparation of 3 g of sorbent Under the adsorption conditions 40

C and 100 kPa CO2, the optimum sorbent showed fast kinetics and an excellent CO2 adsorption of 5.64 mmol/g; the value predicted by the RSM model was 5.67 mmol/g Analysis of variance and the

coefficient of determination (R2 = 0.9525) showed the utility of the approach used in this study to optimize the conditions for preparing high-performance solid sorbents For the optimum sorbent, the

ACS Paragon Plus Environment

2

heat of adsorption was determined to be 78.9 kJ/mol CO2 at 40 C and 100 kPa, using a Calvet

calorimeter, indicating relatively strong chemisorption in the form of carbamate We also found that the

The highest CO2 adsorption, 6.86 mmol/g, was obtained at 50 °C and 100 kPa

1 INTRODUCTION

The atmospheric concentration of CO2, which is a primary greenhouse gas, is increasing, mainly

because of anthropogenic activities such as fossil fuel combustion To mitigate this problem, CO2

emissions need to be reduced significantly from their current level It is widely believed that the CO2

capture and storage (CCS) method is viable option in this context.1–6 However, one of the critical issues

in CCS deployment is reducing the energy costs, especially in the capture process High-performance

materials for CO2 capture have therefore been extensively explored and studied in recent years

One promising method for CO2 capture is the use of solid amine sorbents, which have the following advantages: (i) low energy requirements for sorbent regeneration because of the absence of

solvents, (ii) potentially high adsorption capacity because of the dense amine-containing structure, and

(iii) the low corrosion and toxicity caused by amines anchored to solid supports.7–12 Solid amine

sorbents can be prepared by the wet impregnation of liquid organic amines into porous supports or the

covalent grafting of amines onto porous surfaces using silane coupling agents.3,4,11,13 Wet impregnation

is a simple preparation method, by which a large amount of amine can be introduced into the pores of

the support, leading to a higher CO2 adsorption capacity compared with those of sorbents prepared using

grafting methods

Mesoporous silica materials, including hexagonal MCM-41,4,14–16 SBA-12,13 SBA-15,13,14 dimensional hexagonally ordered MSU-H,4 and mesocellular silica foam MSU-F4 have been used for

two-amine loading because of their high surface areas and thermal stabilities MSU-F, which has large pore

volumes, large pore sizes, and good pore interconnections, showed high CO2 adsorption performance

ACS Paragon Plus Environment

3

after wet impregnation of amines.4 The amines can be linear or branched polyamines such as tetraethylenepentamine (TEPA) or polyethyleneimine (PEI), which have been impregnated into mesoporous materials to attain high-contents of reactive sites, i.e., amino functional groups.3–5,8,10–12,17

Recently, we reported the development of solid sorbents using several mesoporous silica materials prepared by wet impregnation of amines and organic compounds with/without hydroxyl groups.4 The results showed that in addition to large mesopores, hydroxyl groups have positive effects

on the amounts of CO2 adsorbed Consequently, new sorbents with high performances in CO2

adsorption have been developed using large-pored MSU-F silica and polyamine/alkanolamine blends Among some amine blends, TEPA/diethanolamine (DEA) was the most effective The results also indicated that the molar ratio of amino to hydroxyl groups is a key determinant of the adsorption performance and, depending on this ratio, the amine blends showed synergistic effects with regard to the amount adsorbed

In this study, we determined the optimum conditions for the preparation of solid amine sorbents with high CO2 adsorption performances, and evaluated the effects of, and interactions among, the variables in sorbent preparation For the development of an acceptable process in the shortest time using the minimum number of experiments, time, and materials, we used the response surface method (RSM), which is a powerful mathematical and statistical technique for the investigation of various processes and can be widely used for designing experiments, building models, process modeling and optimization.18,19

A few applications of the RSM to the optimization of processes for CO2 capture by solid sorbents have been reported.19–21 Shafeeyan et al.19 developed models for calculating the optimum conditions for amination of activated carbon for the preparation of CO2 sorbents by investigating the effects of amination temperature and time on the CO2 adsorption/desorption performance Gil et al.20used the RSM as a tool for rapidly optimizing activation parameters such as activation temperature and burn-off degree to obtain the highest CO2 adsorption capacity of activated carbon Aziz et al.21 used a

ACS Paragon Plus Environment

4

fractional factorial design to functionalize mesoporous silica particles with aminopropyltriethoxysilane

and showed that the amount of water present during synthesis, the reaction time, pretreatment of the

silica with a mineral acid, and certain two-way interactions were important in the optimization of CO2

uptake

The main aim of this study was optimization of the conditions for the impregnation of MSU-F

the effects of the amounts of TEPA, DEA, and solvent (MeOH), and their interactions, on the amounts

evaluated the amount of CO2 adsorbed and its temperature dependence We also measured the heat of

adsorption using an isothermal calorimetric method

2 EXPERIMENTAL SECTION

2.1 Materials All chemicals were purchased and used without further purification TEPA [98%,

H(NHCH2CH2)4NH2] and MSU-F were purchased from Sigma-Aldrich (St Louis, MO, USA) DEA

[99%, (HOCH2CH2)2NH)] and MeOH (99.8%, CH3OH) were supplied by Wako (Osaka, Japan) He

2.2 Preparation of Solid Sorbents Amine-functionalized MSU-F sorbents were prepared

using the wet impregnation method A specific amount of MSU-F was added to MeOH and the mixture

was agitated ultrasonically for 3 min The required amounts of TEPA and DEA were then added and the

mixture was agitated for 3 min The solid sorbents were obtained by MeOH removal with a rotary

evaporator at 60 °C The prepared sorbents were denoted by TEPAx1-DEAx2/MSU-F/x3, where x1 and x2

in the preparation of 3 g of sorbent; for example, TEPA40-DEA30/MSU-F/75 represents the sorbent

prepared by loading a mixture of 1.2 g (40 wt%) of TEPA and 0.9 g (30 wt%) of DEA into the pores of

ACS Paragon Plus Environment

5

MSU-F (0.9 g, 30 wt%) with 75 g of MeOH

2.3 Material Characterization N2 adsorption–desorption isotherms were obtained using a surface area and porosimetry measurement system (ASAP 2420, Micromeritics, Norcross, GA, USA) Preadsorbed water and CO2 were removed by degasification at 40 C under a N2 flow for 6 h before the adsorption–desorption analysis To improve the data integrity, all the experiments, including measurements of CO2 adsorption isotherms, were set up with adequate equilibration intervals; filler rods were used to ensure sample accuracy by reducing the free-space volume Equilibration was reached when the pressure change per equilibration time interval was less than 0.01 of the average pressure during the interval The specific surface areas of the materials were calculated using the Brunauer–Emmett–Teller (BET) method The total pore volume was determined as the volume of liquid N2adsorbed at a relative pressure of 0.97 The pore size was determined by the Barrett–Joyner–Halenda method using the adsorption branch Thermogravimetric (TG) curves were obtained, using a Thermo

analysis, the samples (mass about 10 mg) were heated from approximately 30 to 1000 °C at a constant rate of 5 °C/min Elemental analysis was performed using an elemental analyzer (Perkin Elmer 2400II CHNS/O, Waltham, MA, USA) The heat of adsorption was measured using a Calvet C80 calorimeter (SETARAM Instrumentation, Caluire, France) The sorbent (about 200 g) was activated at 80 °C for 6 h under N2 at a flow rate of 50 cm3/min The sample was cooled to 40 °C, and then CO2 was passed through the sample at a rate of 45 cm3/min at atmospheric pressure The heat of adsorption at 40 °C and

100 kPa CO2 was determined based on the measured value of the total heat and the adsorption isotherm

2.4 Sorption Studies The amounts of CO2 adsorbed were measured at 40 C up to a pressure

100 kPa was used as the dependent variable in the optimization For investigating the effects of

ACS Paragon Plus Environment

6

measured using an ASAP 2020 instrument (Micromeritics) or a chemical adsorption analyzer

(ChemiSorb HTP, Micromeritics) in the adsorption temperature range from 20–80 °C Before the

adsorption–desorption measurements, degassing was performed at the adsorption temperature for 6 h

under vacuum to remove any preadsorbed moisture and gas

The pure CO2 adsorption kinetics was evaluated gravimetrically at 40 °C and atmospheric pressure using a Thermo Plus TG-DTA 8120 instrument (Rigaku) The sample (about 10 mg) was

placed in a Pt pan The sorbent was pretreated at 80 °C in N2 at a flow rate of 100 cm3/min for 4 h, and

then cooled to 40 °C The N2 flow was then switched to CO2 at a flow rate of 70 mL/min for 2 h

3 RESPONSE SURFACE OPTIMIZATION

The optimum conditions for the preparation of solid sorbents for high-performance CO2 adsorption were

determined by the RSM using Modde software (Umetrics, Umeå, Sweden) In this study, sorbent

preparation was performed based on factorially designed experiments The amounts of TEPA, DEA, and

were examined The experiments were conducted using the central composite design (CCD) method,

with three center points, and fitted to an empirical full second-order polynomial model Table 1 shows

the ranges and levels of independent variables used in this study; low, center, and high levels are

denoted by –1, 0, and +1, respectively This design also required experiments to enable prediction of the

responses outside the cubic domain, denoted by –1.682 and +1.682 The RSM model was constructed

by fitting the experimental data for the amounts of CO2 adsorbed with the following equation:

where Y represents the response, i.e., the amount of CO2 adsorbed at 40 C and 100 kPa, o is a

constant, i, ii, ij are the linear, quadratic, and interaction term coefficients, respectively, X is the i

ACS Paragon Plus Environment

(2)

where x actual , x high , and x low are the real values of the ith independent variable under the operating

conditions, and at the high level and low level, respectively

4 RESULTS AND DICUSSION

4.1 Material Characterization The textural parameters of MSU-F and all the sorbents were

determined using N2 adsorption–desorption isotherms Figure 1 shows the N2 adsorption–desorption isotherm and pore size distribution (inset) for MSU-F; values of 272 m2/g, 1.54 cm3/g, and 28 nm were obtained for the surface area, total pore volume, and pore diameter, respectively Table 2 shows the complete design matrix used for the preparation of all 17 sorbents with the corresponding results for the textural properties and amounts of CO2 adsorbed at 40 °C and 100 kPa After the amines were loaded into the support, the BET surface areas and total pore volumes of all the samples decreased from 272

m2/g and 1.54 cm3/g to less than 30 m2/g and 0.2 cm3/g depend on the amine loading, respectively, indicating that the support pores were filled with the amines

The thermal properties of the materials were determined using TG analysis The silica support showed good thermal stability, and only approximately 2 wt% was lost, because of moisture The TG data also confirmed that the amines used for sorbent preparation were completely loaded into the MSU-

their weight below 100 °C; this can be attributed to preadsorbed gases and solvent that could not be completely removed The weight of the remaining solids after thermal analysis confirmed that the

amines used for wet impregnation were completely loaded into the pores

ACS Paragon Plus Environment

8

4.2 RSM Model The effects of the operating parameters for the preparation of the solid sorbent

on the amount of CO2 adsorbed were investigated using the RSM CCD-designed experiments were

performed to visualize the effects of the independent factors on the response As shown in Table 2, the

amount of CO2 adsorbed varied from 3.85 to 5.63 mmol/g, depending mainly on the effects of the

concentration and composition of the amine blend;4 the amine efficiency, which was defined as the

molar ratio of CO2/N, was in the range 0.27–0.50

The effects of the three variables were evaluated from the experimental results in Table 2, and the following approximate response function was obtained:

Y = 5.62192 + 0.0830368X1 – 0.0345798X2 – 0.0208987X3 – 0.156328X12 – 0.619369X22

– 0.140422X32 – 0.295X1X2 – 0.215X1X3 – 0.1375X2X3 (3)

where Y is the amount of CO2 adsorbed at 40 °C and 100 kPa, and X1, X2, and X3 are the coded values of

the doses of TEPA, DEA, and MeOH, respectively

Figure 3 shows the relationship between the values predicted using eq 3 and the experimental values for the amount of CO2 adsorbed at 40 °C and 100 kPa The fit with the model represented by eq 3

was evaluated based on analysis of variance (ANOVA) and the coefficient of determination (R2) The

ANOVA results for the empirical second-order polynomial model are shown in Table 3 The F-value of

15.6045 indicates that the developed model is highly significant Furthermore, the p value of 0.001,

which is lower than 0.05, suggests that the model is statistically significant.21,22 The R2 value of 0.9525

means that 95.25% of the response variability is explained by the model and also confirms that the

model has good predictability, for which at least R2 = 0.80 is suggested.23 The reproducibility was

determined to be 0.998, using Modde software; this confirms the good predictability of the model.24

4.3 Effects of Independent Variables and Analysis of Response Surface The RSM model

developed in this study was used to study the effects of the TEPA, DEA, and MeOH doses on the

amount of CO2 adsorbed; the results are shown in Figure 4a, 4b, and 4c, respectively In the figure, each

ACS Paragon Plus Environment

9

factor is varied from a low to a high level, while the other two factors are kept at the center level Figure

32 to 40 wt% However, when the TEPA dose increased from 40 to 48 wt%, the response decreased to 5.33 mmol/g Similar results were obtained for the effect of the DEA dose, as shown in Figure 4b The effect of the MeOH dose was small compared with those of the TEPA and DEA doses (Figure 4c) This result indicates that amines are dispersed well into the pores of supports by MeOH and suggests the

robustness of the RSM model in this study The regression coefficients in eq 3 for X1, X2, X3, X12, X22,

and X32, confirm that the doses of TEPA (X1) and DEA (X2) had greater effects on the response than did

the MeOH dose (X3) The absolute values of the coefficients were largest for X1 and X22, for the linear and quadratic terms, respectively These results suggest that chemisorption by TEPA and DEA strongly contributes to the CO2 uptake; this can be represented by the following equations:

2(HOCH2CH2)2NH + CO2 (HOCH2CH2)2NH2+ + (HOCH2CH2)2NCOO (5)

of 0.23 and 0.51, respectively These values suggest that only one amino group in a TEPA molecule forms a carbamate as in eq 4., while the amine efficiency of DEA can be explained by the stoichiometry

amine efficiency is higher than TEPA This is because in DEA, instead of the amino group, the hydroxyl group can play a role in stabilizing the carbamate anion.4 In the present case, as stated above, the amine

efficiency for TEPAx1-DEAx2/MSU-F/x3 lies between these two limiting values (Table 2) The

coefficient in eq 3 for the X1X2 term is also significant, as it indicates interaction between TEPA and DEA; this is partly ascribed to the effect of hydroxyl groups.4

The contour and three-dimensional plots in Figure 5a and 5b show the effects of TEPA and DEA doses on the amount of CO2 adsorbed, with the MeOH dose kept at the center level As can be seen from these figures, the optimum region is for doses of TEPA and DEA of 4244 and 2830 wt%,

ACS Paragon Plus Environment

10

respectively, where the amounts of CO2 adsorbed reached values higher than 5.60 mmol/g These

responses can be explained as follows When the total amine doses are relatively low, the number of

amino group sites that can react with CO2 molecules increases with increasing amine dose, which results

in adsorption of high amounts of CO2 However, when the amine doses are too high, more amine sites

can be blocked; therefore, diffusion of CO2 gas in the sorbent is limited, resulting in adsorption of low

amounts of CO2 and low amine efficiencies.27,28 Furthermore, DEA gives a higher amine efficiency than

TEPA, but the density of amino groups is higher in TEPA than in DEA There is therefore an optimum

composition of TEPA and DEA

4.4 Optimization of Conditions Based on the developed model for simulating CO2 adsorption

at 40 °C and 100 kPa, the optimum conditions for obtaining maximum CO2 adsorption can be

determined using Modde software The model developed in this study predicted that conditions of 44.4

wt% TEPA (1.33 g), 28.4 wt% DEA (0.85 g), and 58.2 g of MeOH for preparing 3 g of solid sorbent

The sorbent TEPA44.4-DEA28.4/MSU-F/58.2 was prepared for verification of the optimum conditions The surface area and pore volume of the sorbent were 2.43 m2/g and 0.003 cm3/g,

respectively, suggesting that most of the pores in the material were filled by the amine blend The

amount of residual sorbent after thermal analysis was 27 wt%, which corresponded to the weight of the

support used for preparing the sorbent Details of the preparation conditions, textural properties of the

sorbent prepared under the optimum conditions, and the predicted and experimental values of the

amounts of CO2 adsorbed are listed in Table 4 The average amount of the separately prepared three

optimum sorbents was 5.64 mmol/g at 40 °C and 100 kPa CO2 This is higher than the previously

highest reported value (5.02 mmol/g) for CO2 adsorption under similar conditions by solid amine

sorbents prepared using a wet impregnation method.29 The predicted amount of CO2 adsorbed by the

optimum sorbent, calculated using eq 3, was 5.67 mmol/g The good agreement between the predicted

and experimental values confirms the validity of the developed model

ACS Paragon Plus Environment

11

4.5 Kinetics and Heat of Adsorption The pure CO2 adsorption kinetics for the optimum sorbent was investigated gravimetrically at 40 °C and atmospheric pressure; the results are shown in Figure 6a The figure shows that CO2 adsorption on the optimum sorbent can be divided into two parts

In the first stage, a sharp linear weight gain occurred, which was completed within the first minute,

was about 5 mmol/g The slower kinetics in the second stage can be attributed to diffusion resistance of

CO2 built up during the adsorption process.30 Our optimum sorbent showed similar kinetic behavior to those of solid amine sorbents prepared by impregnation of silica materials with PEI This shows that our sorbent has potential practical applications, with the benefit of shortened adsorption cycle times because

of the fast kinetics.30,31

a calorimeter The data are shown in Figure 6b A sharp exothermic heat flow peak was observed when the sorbent was exposed to CO2 because of the reaction between the amines and CO2 The average heat

of adsorption in terms of heat release per mole of CO2 was determined from three sample trials to be 78.9 ( 1.4) kJ/mol, based on the values of the heat released for CO2 adsorption and the amount of CO2

adsorbed at 40 °C and 100 kPa The results indicate relatively strong interactions between CO2 and amino groups in this system, resulting in the formation of carbamate species Although the value of the heat of adsorption obtained in this study was higher than those for physisorption on other adsorbents such as zeolites (about 50 kJ/mol or lower), it was in the same range as, or lower than, the reported values for amine-modified mesoporous silicas (up to about 95 kJ/mol).30–34 The moderate heat of adsorption of the optimum sorbent in this study may be the result of enriched levels of hydroxyl groups, because DEA addition caused an increase in the fraction of weakly adsorbed CO2 species.35

4.6 Temperature Effects The adsorption temperature is one of the most important factors

determining the amount of CO2 adsorbed on a solid sorbent.4 The RSM was used at a fixed temperature

ACS Paragon Plus Environment

12

of 40 °C, for the optimum sorbent, therefore the effect of temperature on the CO2 adsorption isotherm

was additionally investigated in the range 2080 °C As shown in Figure 7a, the adsorbed amount

increased with increasing temperature in the range 2050 °C, and reached a maximum value of 6.86

mmol/g at 50 oCunder pressure of 100 kPa The adsorbed amount of the optimum sorbent decreased

above 50 °C According to earlier reports,4,31 the CO2 sorption of amines depends strongly on the

balance between thermodynamic sorption and kinetic diffusion When the temperature increases in the

range 2050 °C, amines become more flexible and more accessible sorption sites for CO2 capture will

be exposed to CO2, resulting in higher CO2 adsorption The reaction between CO2 and amino groups is

exothermic, therefore the observed temperature dependence in the range 2050 °C suggests that CO2

adsorption is predominantly determined by diffusion kinetics rather than thermodynamic factors.4,7,8 It

should be noted that this kinetically limited adsorption amount is regarded as the practical capacity,

because this system reaches almost full capacity in the first stage, as shown in Figure 6 At higher

temperatures, the equilibrium is reversed and desorption is favored, leading to a decrease in the

adsorbed amount The amount adsorbed therefore decreases at adsorption temperatures of 60 °C or

higher

Besides the stability of sorbent in the presence of the humidity and flue gas impurities,36-38 the performance under low CO2 pressure is of great importance for industrial practical applications As

shown in Figure 7b, in the pressure range of 315 kPa, the adsorbed amount is still higher than 5

mmol/g at 50 oC, indicating that our optimum sorbent is promising for realistic post-combustion CO2

capture applications Effects of flue gas composition will be investigated in the future work

4 CONCLUSIONS

sorbents was achieved using the RSM The effects of the amounts of TEPA, DEA, and MeOH used in

the preparation of a MSU-F-based sorbent by the wet impregnation were investigated The doses of

ACS Paragon Plus Environment

13

TEPA and DEA and the ratio of these amines greatly affected the amount of CO2 adsorbed For the preparation of 3 g of solid sorbent, 1.33 g of TEPA, 0.85 g of DEA, and 58.2 g of MeOH were

100 kPa Under these conditions, the optimum sorbent showed fast kinetics and excellent CO2

adsorption, namely 5.64 mmol/g, with a heat of adsorption of 78.9 kJ/mol The adsorption temperature also largely governed the amount of CO2 adsorbed in this system The optimum sorbent gave a

ACKNOWLEDGMENT

The authors thank Mr Fumio Asanoma at the Nara Institute of Science and Technology (NAIST) for technical support with elemental analysis Financial support from the Ministry of Economy, Trade, and Industry (METI), Japan is gratefully acknowledged

REFERENCES

1 Youn, H K.; Kim, J.; Chandrasekar, G.; Jin, H.; Ahn, W S Mater Lett 2011, 65, 1772–1174

2 Peng, Y.; Zhao, B.; Li, L Energy Procedia 2012, 14, 1515–1522

3 Hiyoshi, N.; Yogo, K.; and Yashima, T Microporous Mesoporous Mater 2005, 84, 357–365

4 Dao, D S.; Yamada, H.; Yogo, K Ind Eng Chem Res 2013, 52, 13810–13817

5 Dao, D S.; Yamada, H.; Yogo, K Prepr Pap – Am Chem Soc., Div Fuel Chem 2014, 59 (1),

435–436

6 Song, C Catal Today 2006, 115, 2–32

7 Wang, J.; Chen, H.; Zhou, H.; Liu, X.; Qiao, W.; Long, D.; Ling, L J Environ Sci 2013, 25,