Synthesis of Template-Free ZSM-5 from Rice Husk Ash at Low Temperatures and Its CO2 Adsorption Performance.

In this paper, a green synthesis method for ZSM-5 zeolite is explored to reduce the synthesis cost, environmental hazard, and reaction temperatures. For the ZSM-5 samples prepared at low temperatures, the influence of factors such as the hydrothermal temperature, crystallization time, and the number of seeds is systematically investigated. The adsorption isotherm of CO2 is used for fitting analysis of adsorption models and determination of the adsorption selectivity. The results show that the best one among the three samples presents the highest CO2 adsorption capacity of 2.39 mmol/g at 273 K and 15 bar. It is prepared with a hydrothermal temperature of 393 K, crystallization time of 7 days, and a seed crystal of 1 wt %. The dual-site Langmuir model can well describe the experimental data, indicating that double adsorption sites rather than the simple single-layer adsorption are dominant in samples. As the pressure increases, the adsorption capacity calculated by the model is much lower than the actual value with a deviation index of 12.5%. At a pressure of 1 bar, the optimal selectivity is attained with sample L-20, viz., CO2/N2 of 34.3 and CO2/O2 of 70.2. The green synthesis method reported in this research can be used to successfully prepare ZSM-5 zeolite, and it shows excellent CO2 adsorption performance. In addition, the use of low-cost raw materials and template-free synthesis methods will facilitate the large-scale application of green synthesis processes in the future.

Introduction

Zeolite Socony Mobil 5 (ZSM-5) is a typical MFI-type zeolite with two kinds of open-framework structures: one consists of eight five-membered rings as the basic structural unit and the other is a three-dimensional channel system of ten-membered rings as the main channel window.[1,2] It has great thermal stability, intrinsic acidity, and coking resistance.[3] Its unique open-framework structure makes it suitable for both catalysis and CO2 adsorption. Newsome et al.[4] believed that the lower the Si/Al ratio of a zeolite, the more favorable is its CO2 adsorption capacity. Therefore, the preparation of ZSM-5 with a low Si/Al ratio will have more significance in the field of CO2 adsorption.

The synthesis of ZSM-5 zeolite has been explored for many years, especially in terms of the synthesis of silicon sources, templating agents, and synthesis temperature. In terms of silicon sources, ZSM-5 was mostly synthesized from chemical silicon sources, such as tetraethyl orthosilicate.[5] To reduce the synthesis cost and protect the environment, it was found that kaolin[6−9] and rice husk[10] in the waste are rich in silicon, and the calcined rice husk ash (RHA) is rich in SiO2. Due to the low packing density of RHA, disposal in landfills or open spaces may cause serious problems related to the environment and human health.[11] Therefore, using RHA as a silicon source not only reduces the synthesis cost but also eases the environmental pressure. Barbosa et al.[12] successfully synthesized ZSM-5 with RHA as a silicon source. In terms of templating agents, although ZSM-5 has been synthesized by RHA, the use of a template could not be avoided.[13−15] A template usually plays an important role of structure-directing in the synthesis process of ZSM-5. The typical templates are always tetrapropyl ammonium hydroxide (TPAOH)[13] and tetrapropyl ammonium bromide (TPABr).[16] Both of them are organics and harmful to the environment and very expensive. So, it is necessary to explore a template-free method to synthesize ZSM-5. In addition, the absence of a template means that the sample does not require calcination to remove the template, which saves cost in terms of energy consumption. In recent years, many researchers have synthesized template-free ZSM-5, but the hydrothermal reaction temperature was high, around 443 K.[17−22] Generally, high temperatures mean high energy consumption; therefore, in terms of synthesis temperature, much work has been done to reduce the hydrothermal reaction temperature of ZSM-5. Barakov et al.[23] and Kadja et al.[24] have both successfully synthesized ZSM-5 at 373 and 363 K, respectively. Table shows the synthetic conditions of ZSM-5 zeolite in previous studies. Although the synthesis of ZSM-5 has a history of several decades, there are still shortcomings in its current synthesis process, such as high synthesis cost, use of chemical raw materials as silicon or aluminum sources, addition of organic templates, and high synthesis temperatures.

Table 1

Synthetic Conditions of ZSM-5 Zeolite in Earlier Studies

sample	silicon source (RHA, kaolin, or chemical silicon source)	template	synthesis temperature (K)
ZSM-5[25]	chemical silicon source	TPAOH	443
ZSM-5[26]	chemical silicon source	TPAOH	443
ZSM-5[27]	RHA	TPAOH	438
ZSM-5[1]	chemical silicon source	TPAOH	453
ZSM-5[3]	chemical silicon source	TPABr	413
ZSM-5[5]	chemical silicon source	TPAOH	443
ZSM-5[6]	kaolin	free	463
ZSM-5[28]	chemical silicon source	free	463
ZSM-5[29]	RHA	TPABr	383
ZSM-5[30]	RHA	TPABr	423
ZSM-5[18]	chemical silicon source	free	438
ZSM-5[31]	kaolin	TPABr	453

In this work, RHA and pseudo-boehmite were used as silicon and aluminum sources to synthesize ZSM-5, aiming at saving the synthesis cost. At the same time, ZSM-5 was successfully synthesized by the template-free method at a low temperature. This method reduced both energy consumption and environmental pollution. Then, the CO2 adsorption performance of ZSM-5 synthesized at a low temperature was studied at 15 bar.

This work is new with respect to the synthesis method and adsorption performance analysis. In terms of the synthesis method, this work used RHA as the silicon source, successfully synthesized ZSM-5 zeolite under the conditions of no organic template and a low temperature, and explored a green synthesis route for ZSM-5 zeolite. In terms of performance testing, this work explored the CO2 adsorption performance of ZSM-5 zeolite under high-pressure conditions (0–15 bar), which will be of great significance for its use in the field of high-pressure adsorption of CO2.

Results and Discussion

Characterization Analysis of L-m Zeolites

Figure shows the X-ray diffraction (XRD) patterns of the three samples. As can be seen, these three samples have the same characteristic diffraction peaks at 2θ = 8.0, 8.9, 13.3, 14.1, 14.9, 16.0, 23.2, 24.0, 24.4, 25.9, 26.9, 29.4, and 30.2°, which are in agreement with the MFI-type framework structure (ICDD PDF No. 44-0003).[32] Crystallinity is calculated by dividing the diffraction peak intensity by the total intensity. The crystallinities of samples L-20, L-30, and L-40 are 84.1, 77.3, and 69.9%, respectively. The crystallinity decreases with the increase of the Si/Al molar ratio. Between 20 and 25°, samples L-30 and L-40 also have a clear amorphous hump of SiO2, which is attributed to the increase in the crystallization time of ZSM-5 zeolite with the increase in the ratio of silicon/aluminum and part of the SiO2 not been fully converted. The XRD pattern indicates that ZSM-5 zeolite with different Si/Al ratios can be successfully synthesized under the chosen conditions. However, the incomplete reaction of excessive SiO2 in the raw material will adversely affect the finished product of ZSM-5 zeolite.

Figure 1

XRD patterns of L-m (m = 20/30/40) zeolites.

Figure shows the Fourier transform infrared (FTIR) spectra of the three samples. All three samples were dried under vacuum at 100 °C for 2 h before the FTIR measurement to avoid signal interference caused by the adsorbed water. The transmission bands of the three samples synthesized at 393 K are almost at the same wavenumbers as that of the seed sample R-20, which indicates that the four samples contain similar functional groups, further verifying the successful synthesis of ZSM-5 zeolite at low-temperature conditions. The presence of the bands at 3447 and 1634 cm–1 are mainly attributed to the tensile vibration and bending vibration bands of the O–H bond in the water molecules absorbed in the samples. The presence of the bands at 1080 and 799 cm–1 is mainly attributed to the symmetric and asymmetric tensile vibration of Si–O–Si and Si–O–Al tetrahedral structures.[33] The presence of the band at 550 cm–1 is considered to represent the double-loop vibration of the five-membered ring.[34−38] The presence of the band at 1225 cm–1 corresponds to the asymmetric stretching vibration of the typical five-membered ring structure in ZSM-5 zeolite.[39] With the increase of the Si/Al ratio, the intensity of the transmission bands at 550 and 799 cm–1 decreases significantly, indicating that the amount of ZSM-5 zeolite in the sample is reduced, which is consistent with the XRD characterization results.

Figure 2

FTIR spectra of L-m (m = 20/30/40) and R-20 zeolites.

Figure shows the scanning electron microscopy (SEM) images of the three samples. In appearance, all samples showed the standard prismatic morphology of ZSM-5, with consistent grain size and uniform distribution, indicating the successful preparation of ZSM-5. The average diameters of ZSM-5 zeolite in samples L-20, L-30, and L-40 are 0.80, 0.83, and 0.71 μm, respectively. With the increase of the Si/Al molar ratio, the ZSM-5 with the prismatic morphology in the samples gradually decreases.

Figure 3

SEM images of (a, b) L-20, (c, d) L-30, (e, f) L-40, and (g) SiO2.

Figure shows the N2 adsorption–desorption isotherm at 77 K and the pore size distribution (calculated by the 2D-NLDFT model[40]) of L-m analyzed in Table . The N2 adsorption amount of all three samples at 77 K increases rapidly in the low-pressure zone (P/P0 < 0.1), which is mainly due to microporous adsorption.[35] The adsorption amount first increases steadily and then increases rapidly in the high-pressure zone (P/P0 > 0.1), and a loop representing the mesoporosity is exhibited in each adsorption–desorption diagram, which is mainly due to mesoporous adsorption, macroporous adsorption, and adsorption between microparticles. So, both microporous and mesoporous structures are present in all three samples, which can also be seen in the pore size distribution maps of the three samples.[41] All three samples contain micropores of about 0.58 nm and extensive mesoporous structures. With the increase of the Si/Al molar ratio, the specific surface area of the sample, calculated by the t-plot method, decreases. At the same time, the total pore volume, micropore volume, and the micropore volume ratio of the samples also decrease. This is because with the increase of the Si/Al molar ratio, the formation of microporous structures takes more time. Besides, the sample L-40 contains an obvious mesoporous structure, which is mainly due to the unreacted amorphous SiO2 in the sample. The results show that the zeolite obtained in this study has excellent pore size uniformity, which is very conducive to the adsorption of CO2 with a specific size of molecules.

Figure 4

N2 adsorption–desorption isotherm at 77 K and the pore size distribution of (a, b) L-20, (c, d) L-30, and (e, f) L-40.

Table 2

Textural Properties of L-m Zeolitesa,b,c

sample	SBET (m2/g)	Smicropore (m2/g)	Sexternal (m2/g)	Vtotal (cm3/g)	Vmicropore (cm3/g)	Rmicropore (%)
L-20	284.8	211.8	73.0	0.157	0.110	70.1
L-30	246.3	185.7	60.6	0.154	0.098	63.6
L-40	182.9	133.0	49.9	0.138	0.071	51.4

SBET: specific surface area calculated by the Brunauer–Emmett–Teller model.

Smicropore: micropore area calculated by the t-plot model.

Sexternal: external surface area calculated by the t-plot model.

Adsorption Performance of L-m Zeolites

According to the investigation results, ZSM-5 zeolite samples with Si/Al molar ratios of 20, 30, and 40 were synthesized with a synthesis temperature of 393 K, crystallization time of 7 days, and a seeding ratio of 1 wt %. They are named L-m (m = 20/30/40).

Adsorption Regeneration Property

The adsorption regeneration property of the three samples was studied in this work. Without removing or reactivating the samples, cyclic adsorption and desorption were directly performed to observe the change in the adsorption amount of the samples. The relationship between the adsorption index of each sample and the number of cycles is given in Figure a. According to the results, for the L-20 zeolite, the adsorption amount in the vicinity of 15 bar for the five cycles was 2.39, 2.23, 2.21, 2.21, and 2.03 mmol/g, respectively. From the second cycle of adsorption and desorption, the adsorption indices of the sample were 93.43, 92.59, 92.28, and 85.01%. For the L-30 zeolite, the adsorption amounts were 2.11, 1.92, 1.91, 1.73, and 1.74 mmol/g, respectively, and the adsorption indices of the sample were 91.15, 90.68, 82.25, and 82.69%. For the L-40 zeolite, the adsorption amounts were 1.92, 1.58, 1.54, 1.75, and 1.60 mmol/g, respectively, and the adsorption indices of the sample were 82.49, 80.11, 91.35, and 83.43%. The cyclic adsorption–desorption performance of L-40 exhibited first a decrease and then an increase owing to the presence of a large amount of amorphous silica in L-40, which has not been converted into ZSM-5 zeolite crystals. According to the pore size distribution test result of sample L-40, the sample contains many large pores, which makes the cyclic desorption performance of the sample not sufficiently stable under high-pressure conditions. Furthermore, we conduct an error analysis on the cyclic adsorption performance of the three samples in Figure b. The results show that the cyclic adsorption performance of the L-20 sample is significantly better than those of the other two samples, which is mainly due to the perfect crystallization of the L-20 zeolite.

Figure 5

Cyclic adsorption–desorption performance of ZSM-5-L-m: (a) adsorption index and (b) error analysis.

Measurement of CO2 Adsorption Capacity

The CO2 adsorption isotherm of samples L-20, L-30, and L-40 was measured at temperatures of 273, 303, and 333 K in the pressure range of 0–15 bar. It is necessary to predict the equilibrium relationship between the adsorbent and the adsorbate to explore its interaction and adsorption mechanism.[42] The Langmuir, Freundlich, and dual-site Langmuir models were used to carry out the fitting analysis with the measured data.

Figure shows the CO2 adsorption isotherm fitted by models of L-m zeolites at 273, 303, and 333 K. Table gives the CO2 adsorption isotherm parameters of the three models of L-m at 273, 303, and 333 K. The adsorption amounts of the three samples decrease with the increase of the adsorption temperature. The adsorption amounts at a pressure of 1 bar of sample L-20 at temperatures of 273, 303, and 333 K were 2.02, 1.76, and 1.50 mmol/g, respectively; moreover, when the pressure increased to 15 bar, they increased to 2.39, 2.26, and 2.16 mmol/g, respectively. For sample L-30, they were 1.61, 1.25, and 1.12 mmol/g at 1 bar and 2.10, 1.82, and 1.67 mmol/g at 15 bar. For sample L-40, they were 1.27, 1.07, and 0.79 mmol/g at 1 bar and 1.70, 1.65, and 1.47 mmol/g at 15 bar. With the increase of the adsorption temperature, the adsorption amount of the L-20 sample decreased almost equally. But for the L-30 sample, when the adsorption temperature increased from 273 to 303 K, the adsorption capacity of the sample decreased significantly, while for the L-40 sample, the adsorption capacity decreased significantly when the temperature increased from 303 to 333 K. At the same time, the adsorption amount of the sample decreased with the increase of the Si/Al molar ratio under the same pressure conditions. This is because the SiO2 impurity in the sample was in an amorphous state and it failed to form a rich pore structure and did not have a specific CO2 adsorption site. According to Table , compared with ZSM-5 zeolites reported in other references, the L-20 sample described herein has a higher CO2 adsorption capacity with a lower synthesis cost.

Figure 6

CO2 adsorption isotherm and fitted model of (a, b, c) L-20, (d, e, f) L-30, and (g, h, i) L-40 at 273, 303, and 333 K.

Table 3

CO2 Adsorption Isotherm Parameters of the Models of L-m at 273, 303, and 333 K

model	parameter	L-20	L-30	L-40
Langmuir	qm (mmol/g)	2.148	1.789	1.509
	b0 (bar–1)	0.001	0.001	2.996
	Q (kJ/mol)	22.878	22.439	28.763
	R2	0.924	0.919	0.926
Freundlich	k (mmol/(g·bar1/n))	1.622	1.217	0.981
	n1	13.087	16.464	14.126
	n2	–1855.3	–3221.6	–2632.3
	R2	0.855	0.843	0.821
dual-site Langmuir	qm,A (mmol/g)	1.244	1.075	0.968
	b0,A (bar–1)	4.595 × 10–5	2.519 × 10–5	4.068 × 10–5
	QA (kJ/mol)	25.090	24.439	22.968
	qm,B (mmol/g)	1.092	0.958	0.751
	b0,B (bar–1)	1.861 × 10–4	6.425 × 10–4	8.047 × 10–9
	QB (kJ/mol)	33.681	26.829	56.280
	R2	0.995	0.980	0.991

Table 4

Comparison of CO2 Adsorption Capacity of ZSM-5 Zeolite at 1 bar

sample	Si/Al	adsorption temperature (K)	CO2 adsorption capacity (mmol/g)
ZSM-5[15]	25	273	2.03
ZSM-5[8]	50	298	1.79
ZSM-5[43]	130	308	1.45
ZSM-5[44]	27	323	1.75
ZSM-5[45]	27	303	1.34
ZSM-5a	20	273	2.02
		303	1.76
		333	1.50

This work.

According to Table , the correlation coefficients (R2) of all three fitting models are in the range of 0.821–0.995. For the same sample, the dual-site Langmuir model always fits better than the other models. It can be speculated that the Langmuir model cannot fit the adsorption isotherms of samples L-20, L-30, and L-40, as the three samples do not undergo single-layer adsorption. This is mainly due to the presence of incompletely crystallized SiO2 in the sample. The Freundlich isotherm model does not fit the isotherms of the three samples, which is considered to be suitable for the adsorption of different layers but for heterogeneous systems.[42] The dual-site Langmuir model always fits better than the other models because of the presence of incompletely crystallized SiO2 in the sample. The two adsorption sites are ZSM-5 with micropores and SiO2 with mesopores.

The adsorption capacities of samples were measured at 1 bar at different temperatures, and their adsorption capacities were predicted at higher pressures by fitting with different models. The CO2 adsorption isotherm of sample L-20 was measured at a temperature of 273 K in the pressure range of 0–1 bar and fitted with the dual-site Langmuir model. The CO2 adsorption isotherm parameters of the samples are listed in Table , and the correlation coefficient (R2) is 0.998. The deviation indices of sample L-20, calculated by eq , are shown in Figure . At all three temperatures, the absolute value of the deviation index is very large because the adsorption amount at low pressures is quite small. When the pressure is over 1 bar, the adsorption amount obtained by model fitting starts to deviate from that measured actually. When the adsorption temperatures are 273, 303, and 333 K, the absolute values of the adsorption deviation index of the sample at about 15 bar reached 12.5, 8.7, and 6.6%, respectively. All deviation indices are negative, meaning that the adsorption amount of the sample calculated by the model is much lower than the actual value. It is necessary to measure the real adsorption capacity of samples at high pressures rather than calculate it.

Table 5

CO2 Adsorption Isotherm Parameters of the DSL Model of L-20 at 1 bar

parameter	qm,A (mmol/g)	b0,A (bar–1)	QA (kJ/mol)	qm,B (mmol/g)	b0,B (bar–1)	QB (kJ/mol)	R2
L-20	1.235	9.599 × 10–5	25.460	0.868	6.285 × 10–4	32.316	0.998

Figure 7

Deviation indices of sample L-20.

Selectivity Measurements

To know more about the adsorption performance of samples synthesized at low temperatures, the selectivity of CO2/N2 and CO2/O2 was predicted in this paper. Due to the differences in size, polarity, and quadrupole moment of the CO2 and N2 or O2 molecules, they were selectively adsorbed by adsorbents. To obtain the selectivity of CO2/N2 and CO2/O2, the adsorption capacities of CO2, N2, and O2 for L-m zeolites of three silicon/aluminum ratios were first measured at 273 K. The obtained adsorption isotherm was fitted, and the fitting formula was used to calculate the selectivity of CO2/N2 and CO2/O2. The dual-site Langmuir model with the highest fitting degree of CO2 adsorption isotherms was selected to fit the adsorption isotherms of N2 on the three samples. The CO2/N2 adsorption isotherm parameters of the three samples are shown in Table . It can be seen that all of the correlation coefficients (R2) are beyond 0.999, meaning that the dual-site Langmuir model’s excellent fit in the low-pressure region provides the necessary accuracy for predicting selectivity.

Table 6

CO2/N2 Adsorption Isotherm Parameters of the Models of L-m at 273 K

sample	gas	qm,A (mmol/g)	bA (bar–1)	qm,B (mmol/g)	bB (bar–1)	R2
L-20	CO2	1.244	2.887	1.092	513.883	0.9949
	N2	1.250	0.212	0.442	2.372	0.9998
L-30	CO2	1.075	1.188	0.958	86.814	0.9803
	N2	1.221	0.153	0.435	2.062	0.9996
L-40	CO2	0.968	1.004	0.751	466.037	0.9911
	N2	0.818	0.239	0.275	2.823	0.9998

The selectivity of CO2/N2, calculated by eq , of the three samples are shown in Figure . It can be seen that the selective adsorption capacity of the three samples decreases with the increase of pressure. When the pressure is 1 bar, the selective adsorption capacities of samples L-20, L-30, and L-40 are 34.3, 30.3, and 30.5, respectively. So, the selective adsorption capacity of sample L-20 is much better than those of the other two samples. As the pressure increases, the selective adsorption capacity of the sample tends to be stable. For sample L-20, the selective adsorption capacity of CO2/N2 is around 15, and those of samples L-30 and L-40 are around 14 and 16, respectively.

Figure 8

CO2/N2 adsorption isotherm and the adsorption quantity ratio of L-m zeolites at 273 K: (a) L-20, (b) L-30, and (c) L-40.

The dual-site Langmuir model was also selected to fit the adsorption isotherms of O2 on the three samples. The CO2/O2 adsorption isotherm parameters of the three samples are shown in Table . It can be seen that all of the correlation coefficients (R2) are beyond 0.999, meaning that the selectivity of CO2/O2 predicted by the dual-site Langmuir model was also reasonable.

Table 7

CO2/O2 Adsorption Isotherm Parameters of the Models of L-m at 273 K

sample	gas	qm,A (mmol/g)	bA (bar–1)	qm,B (mmol/g)	bB (bar–1)	R2
L-20	CO2	1.244	2.887	1.092	513.883	0.9949
	O2	1.557	0.199	1.706 × 10–4	2.325 × 107	0.9998
L-30	CO2	1.075	1.188	0.958	86.814	0.9803
	O2	1.727	0.171	0.041	1.604	0.9998
L-40	CO2	0.968	1.004	0.751	466.037	0.9911
	O2	1.373	0.142	0.072	1.245	0.9995

The selective adsorption capacities of CO2/O2 calculated by eq of the three samples are shown in Figure . It can be seen that the selective adsorption capacities of the three samples decrease with the increase of pressure. When the pressure is 1 bar, the selective adsorption capacity of samples ZSM-5-L-20, ZSM-5-L-30, and ZSM-5-L-40 are 70.2, 49.7, and 52.8, respectively. So, the selective adsorption capacity of sample ZSM-5-L-20 is far better than those of the other two samples. As the pressure increases, the selective adsorption capacity of the sample tends to be stable. For sample ZSM-5-L-20, the selective adsorption capacity of CO2/O2 is around 18, and those of samples ZSM-5-L-30 and ZSM-5-L-40 are around 14 and 15, respectively.

Figure 9

CO2/O2 adsorption isotherm and adsorption quantity ratio of L-m zeolites at 273 K: (a) L-20, (b) L-30, and (c) L-40.

Conclusions

This paper explored the synthesis of ZSM-5 zeolite under low-temperature and template-free conditions using an inexpensive silicon source (rice husk ash) and aluminum source (pseudo-boehmite) and verified the feasibility of the green synthesis strategy. Based on the findings, ZSM-5 zeolite samples with Si/Al molar ratios of 20, 30, and 40 were finally synthesized at 393 K, with a crystallization time of 7 days and a seeding ratio of 1 wt %. Based on the characterization and performance measurement results of the three samples, the L-20 zeolite has better adsorption and selectivity performance. From the measurement results, when the temperatures were 273, 303, and 333 K, the adsorption amounts of sample L-20 near 15 bar were 2.39, 2.26, and 2.16 mmol/g, respectively. Through the fitting of the three models (Langmuir, Freundlich, and dual-site Langmuir), it is found that the ZSM-5 zeolite with three silica/alumina ratios synthesized at a low temperature is more biased toward adsorption with double sites. As the pressure increases, the adsorption capacity predicted by the model is lower than the actual measured value. For sample L-20, the selective adsorption capacities of both CO2/N2 and CO2/O2 decrease with increasing pressure, and the selective adsorption capacity at 1 bar is 34.3 and 70.2, respectively. In addition, the excellent regeneration capacity of ZSM-5 obtained in this work will also become a significant advantage for large-scale applications in the future.

Materials and Methods

Materials

The silicon source was from rice husk bought from a local farm. RHA was obtained by burning the rice husk (973 K, 6 h), and its chemical composition analysis (X-ray fluorescence, XRF) is shown in Table . Using RHA as a silicon source not only reduces the synthesis cost but also eases the environmental pressure. The alkali-dissolving acid extraction method reported in a prior study[46] was used to obtain the silica solid powder from RHA with a SiO2 content of about 98.5 wt % (Table ). The silica solid powder was used directly as a silicon source for the synthesis of ZSM-5. Pseudo-boehmite purchased from Guizhou Morui New Material Technology (China) was also chosen as an aluminum source in this paper, and its chemical composition analysis (XRF) is shown in Table . Besides, tetrapropyl ammonium hydroxide (TPAOH, 25 wt % in H2O), sodium chloride, sodium hydroxide, and hydrochloric acid (36–38%) were purchased from Sinopharm.

Table 8

Chemical Composition Analysis of RHA and Pseudo-Boehmite (wt %)

samples	RHA	SiO2 from RHA	pseudo-boehmite
SiO2	91.6726	98.2956	0.3053
Al2O3	0.6033	0.2274	99.4150
K2O	3.9910	0.0704
CaO	1.1797	0.0261	0.2097
P2O5	0.9388
MgO	0.5845
Fe2O3	0.5809	0.0580	0.0175
MnO	0.2812
Cr2O3	0.0912
SO3	0.0342	0.0408	0.0093
Ga2O3			0.0220
ZnO	0.0137	0.0085	0.0204
Cl	0.0117	0.6401
Na2O	0.0081	0.6127	0.0026

Synthesis Methods

The hydrothermal method was used for the synthesis of the ZSM-5 zeolite seed crystals. The ZSM-5 zeolite seed was synthesized with the raw material in the molar ratio of 20SiO2:8TPAOH:xAl2O3:400H2O:2.6xNaCl:0.5xHCl (x = 0.5/0.375/0.25) and was named R-m (m = 20/30/40). Pseudo-boehmite was completely dissolved with stirring in deionized water, and HCl was dropped into the solution to obtain sol particles of pseudo-boehmite. Then, TPAOH, NaCl, and SiO2 were added into the solution sequentially. Four hours later, the solution was transferred into a Teflon-lined stainless steel reactor at 438 K for 48 h. Finally, it was calcined at 773 K for 6 h to remove the organic template.

Then, ZSM-5 zeolite samples were synthesized under low-temperature and organic template-free conditions and were named L-m (m = 20/30/40). Different from the synthetic process above, the silicon and aluminum source solutions were prepared separately. Then, the silicon source solution was added to the aluminum source solution. After that, the above synthesized seed crystals were added to the solution an hour later. Other steps are similar except for calcination. By the addition of 0.5 g of seed crystals, about 4 g of ZSM-5 zeolite is synthesized and so the cost of using an organic template can be reduced by 87.5%.

The three main factors affecting the synthesis of ZSM-5 are synthesis temperature, crystallization time, and the dosages of seed crystals. To determine the influence of the temperature on the synthesis of ZSM-5 zeolite, samples were synthesized with a Si/Al molar ratio of 20 at different temperatures of 363, 378, and 393 K and were named L-20-t (t = 363/378/393). To determine the influence of crystallization time, ZSM-5 samples were synthesized with a Si/Al molar ratio of 30 at different crystallization times of 3, 5, and 7 days and were named L-30-xD (x = 3/5/7). To determine the influence of the dosage of seed crystals, samples were synthesized with a Si/Al molar ratio of 20 at different seeding ratios of 0, 1, 2, and 5 wt % and were named L-20-n (n = 0/1/2/5). Then, ZSM-5 zeolite samples with Si/Al ratios of 20/30/40 were synthesized at a low temperature, named L-m (m = 20/30/40), and their adsorption performance was studied.

Characterization Methods

The chemical composition analysis of RHA silica was performed by X-ray fluorescence (XRF) with a ZSX Primus ii spectrometer (Rigaku, Japan). A powder X-ray diffractometer (XRD-7000, Shimadzu) equipped with a Cu anticathode was used to measure the crystal phases of the samples. The measuring range was from 5 to 40° with a step of 0.02°. A Fourier infrared spectrometer (Cary 660 FTIR, Agilent) was used to measure the functional groups of samples in the range of 500–4000 cm–1. A scanning electron microscope (Ultra Plus, ZEISS) was used to observe the microcrystal morphology of samples. The nitrogen adsorption/desorption isotherms of the samples were recorded using a Physical adsorption apparatus (ASAP 2460, Micromeritics) at 77 K. The Brunauer–Emmett–Teller (BET) and T-plot models were used to analyze the specific surface area, pore size distribution, and the micropore volume.

Gas Adsorption Measurements

The adsorptive properties of samples were measured using a high-pressure physical adsorption apparatus (3H-2000PH, Beishide). Two pressure sensors are placed in the high-pressure physical adsorption apparatus: one for the 200 bar model US381-2-200BA and the other for the 10 bar model US381-2-10BA. Before installing in the sample tube, all samples were dried in a vacuum drying oven at 423 K for about 6 h. Then, 1 g of the sample was placed in the sample tube. Before the measurements, the samples were heated to 573 K for 12 h to degas for accurate measurements.

CO2 Adsorption Isotherm

In this paper, CO2 adsorption isotherms of samples L-20, L-30, and L-40 were measured at temperatures of 273, 303, and 333 K in the pressure range of 0–15 bar. The Langmuir, Freundlich, and dual-site Langmuir models were used to carry out the fitting analysis with the measured data.

The Langmuir model was originally used to describe the gas–solid phase adsorption on activated carbon, which was traditionally used to quantify and compare the performance of different adsorbents.[47] The Langmuir isotherm refers to homogeneous adsorption, which means that each molecule has a constant enthalpy and adsorption activation energy and the adsorbate does not migrate onto the surface plane.[48] It is expressed by the following equationswhere q is the adsorption quantity (mmol/g) at the absolute pressure p (bar), qm is the complete monolayer adsorption constant (mmol/g), bL is the Langmuir model constant (bar–1), which is calculated by eq , Q is the same amount of adsorption heat that is required for physical adsorption (kJ/mol), T is the reaction temperature (K), and bL,0 is a constant (bar–1).

Compared to the Langmuir model, the Freundlich model is an empirical model applied to multilayer adsorption.[42] It is expressed by the following equations[49]where q is the adsorption quantity (mmol/g) at the absolute pressure p (bar), k is the Freundlich model constant (mmol/(g·bar1/)), n is the heterogeneous adsorption surface coefficient calculated by eq , T is the reaction temperature (K), and n1 and n2 are constants.

The dual-site Langmuir model is a type of four-parameter model developed from the two-parameter Langmuir model, which is used to fit the adsorption isotherm. In the building of this model, it is considered that the heterogeneous surface of the adsorbent is a collection of different energy locations,[50] and it is expressed by the following equationswhere q is the adsorption quantity (mmol/g) at the absolute pressure p (bar), qm,A and qm,B are the maximum saturated adsorption quantities of the two different adsorption sites considered (mmol/g), bA and bB are the Langmuir model constants of the two different adsorption sites (bar–1), which are calculated by eqs and 8, respectively, QA and QB are the same amount of adsorption heat that is required for physical adsorption (kJ/mol), T is the reaction temperature (K), and b0,A and b0,B are constants (bar–1).

In this paper, the deviation index (D) is defined to express the adsorption quantity calculated by the model (qc, mmol/g) and that measured practically (q, mmol/g) and is expressed as follows

Selectivity

To determine the selective adsorption capacity (S) of CO2/N2 and CO2/O2 of the three samples, we measured the adsorption isotherms of CO2, O2, and N2 at 273 K. The obtained adsorption isotherms are fitted, and the fitting formula is used to calculate the selective adsorption capacity of CO2/N2 and CO2/O2 of the corresponding sample combined with the actual working conditions to predict the effect of the sample used in industrial flue gas separation. The industrial flue gas usually consists of N2 and CO2 in a volume ratio of 9:1.[51] The selective adsorption capacity of CO2/N2 is calculated by the following equationwhere q is the adsorption quantity (mmol/g) at the partial pressure p (bar), i represents CO2, and j represents N2 or O2.

To measure the adsorption regeneration property, the CO2 cyclic adsorption performance was measured at 273 K. The percentage of the regenerated adsorption amount and the initial adsorption amount was defined as the adsorption index.