1 Introduction

Sodium metasilicates (water glass) are transformed into soluble silicic acids in organic solvents by the addition of inorganic acids and salting out [1]. Among the organic solvents, tetrahydrofuran (THF) shows high silicic acid extraction ability and solubility [2]. Silicic acid is a potential starting material because it is highly reactive in forming alkoxysiloxanes [3, 4], acetoxysiloxanes [5, 6], polymetalloxanes [7,8,9], and cubic-type silicates [10]. Silicic acid has been reported to consist mainly of polysiloxane with a 6-membered ladder structure [1], and the hydrolysis–condensates of tetraethoxysilane (TEOS) have a similar structure [11]. Hence, silicic acid can be regarded as an intermediate of the hydrolyzed TEOS.

Mesoporous silica is prepared by a combination of sol–gel science and surfactant template science [12, 13], with pore diameters ranging from 2 to 50 nm [14]. We previously reported the facile preparation of bulk porous silica using TEOS, and bulk porous silica-supported zinc oxide exhibited desulfurization performance [15]. The desulfurization facility adopts an up-flow process [16], and the use of powdered supporting agents is unsuitable because powders are released as suspended particulates by up-flow gas; therefore, preparation of bulk body is in some demand. Because the bulk porous silica-supported ZnO after desulfurization is disposed of, it is desirable to use raw material that is inexpensive and has a low environmental impact. Focusing on the silica source, the environmental impact and cost can be reduced by using water glass (7500 yen/25 kg) instead of TEOS (3200 yen/0.5 L) because, in the chemical industry, TEOS is produced from SiO2 through a multiple-step procedure, including the reduction of the SiO2 as an energy-intensive process [17, 18]. Porous silica prepared from water glass has been reported [19]; however, this method has disadvantages such as contamination of sodium salt and non-reuse of Pluronic P123. Silicic acid is more suitable for the preparation of porous silica to prevent the contamination of sodium salt and to recover Pluronic P123, compared to water glass. It seems possible to prepare porous silica at a lower cost because the relatively expensive surfactants can be reused. In this study, we investigated the environmentally friendly and low-cost preparation of bulk porous silica using silicic acid derived from water glass, recycling of Pluronic P123, and desulfurization ability of bulk porous silica-supported ZnO.

2 Experimental section

2.1 Reagents

THF (>99.5%), and hydrochloric acid (12 and 6 mol/L HCl aq.) were purchased from Kanto Chemical Co., Inc. (Tokyo, Japan). Zinc chloride (ZnCl2, 98%), a zinc standard solution (1000 mg/L), and magnesium sulfate (MgSO4, 98%) were purchased from FUJIFILM Wako Pure Chemical Corp. (Tokyo, Japan). Pluronic P123 (average Mn ~ 5800) was purchased from Sigma-Aldrich Japan (Tokyo, Japan). Water glass No.3 (SiO2:28–30%; Na2O: 9–10%) was purchased from Toso Sangyo Co., Ltd. (Tokyo, Japan). Water was purified using a PURELAB flex 5 water purification system from ELGA LabWater (High Wycombe, UK), and the resistivity was ~15 MΩ cm.

2.2 Measurements

Thermogravimetric analysis (TGA) was performed using a 2000SE instrument (Netzsch Japan, Kanagawa, Japan). The sample was heated to 1000 °C at a rate of 10 °C/min under airflow, and α-Al2O3 was used as the reference. The morphologies and elemental compositions of the milled porous silica were recorded using scanning electron microscopy (SEM) using a JCM-6000 instrument (JEOL, Tokyo, Japan) with an accelerating voltage of 15 kV, and transmission electron microscopy (TEM) using a JEM-2100F (JEOL) with energy-dispersive X-ray spectroscopy (EDX) using a JED-2300T (JEOL). The porous silica was deposited on a copper grid from an ethanol suspension and then dried. Nuclear magnetic resonance (NMR) spectra were recorded using a JNM-ECZ 400 spectrometer (JEOL; for 1H, 400 MHz). The chemical shifts were reported in ppm relative to chloroform-d (CDCl3) (for 1H, 7.26 ppm in residual chloroform) as the internal standards. The content of zinc atoms in the porous silica-supported ZnO was analyzed by inductively coupled plasma atomic emission spectrometry (ICP-AES) using an ICPE-9000 (Shimadzu, Kyoto, Japan) and was calculated from the difference between the concentrations of ZnCl2 aqueous solution before and after impregnation. The pore size distribution and specific surface area (SBET) of the milled porous silica were measured at −196 °C using a BELSORP-max surface area analyzer (MicrotracBEL, Osaka, Japan). Before analysis, the samples were evacuated for 12 h at 150 °C under vacuum. The specific surface areas and microporous and mesoporous distributions were calculated using the Brunauer–Emmett–Teller (BET) [20], micropore (MP) [21], and Barrett–Joyner–Halenda [22] methods. Small-angle X-ray scattering (SAXS) measurements of the milled porous silica sandwiched in a copper cell were performed using a SAXSess camera (Anton Paar Japan, Tokyo, Japan) equipped with a PANalytical PW3830 laboratory X-ray generator with Cu-K α radiation (0.154 nm, 40 kV, 50 mA) as the X-ray source. Gas chromatography (GC) was performed on a GC-8A column (Shimadzu) packed with a Sunpak-S 80–100 E-9550 column (Shinwa Chemical Industries Ltd., Kyoto, Japan) and a flame photometric detector. Helium was used as the carrier gas. The column temperature was programmed as follows: column temperature = 80 °C; injection temperature = 150 °C; detector temperature = 150 °C. The gas pressures of the carrier gas, hydrogen gas, and air were 200, 50, and 50 kPa, respectively.

2.3 Preparation of silicic acid/THF via water glass

Silicic acid/THF was prepared according to a previous work [10]. Water glass No. 3 (10 g), 6 mol/L HCl aq. (10 mL), and water (25 mL) was stirred at 500 rpm for 15 min in an ice bath. THF (50 mL) and NaCl (15 g) were added to salt out, and the mixture was stirred at 900 rpm for 30 min. The organic layer was then collected and dried over MgSO4 for 20 min. After filtration, the silicic acid/THF was obtained as a pale-yellow solution.

The concentration was determined as follows: (i) 2 mL of silicic acid/THF was collected, and the solvent was evaporated at room temperature for 24 h followed by at 50 °C for 5 h, (ii) the ceramic yield of the residue was determined by TGA, and (iii) the concentration was calculated from the weight change before/after evaporation and ceramic yield. The concentration was 0.89 ± 0.04 mol/L.

2.4 Preparation of bulk porous silica via silicic acid

A mixture of Pluronic P123 (Si/Pluronic P123 = 40–160) and 12 mol/L HCl aq. (10 g), and water (H2O/Si = 80–121) was stirred at 40 °C to dissolve Pluronic P123. The silicic acid/THF (0.89 ± 0.04 mol/L, 30 mL) was added to the solution, and the mixture was stirred at 40 °C and 600 rpm for 20 h. After stopping the stirring, the mixture was heated at 80 °C for 24 h. The white gel was filtered and washed with water, and then the washed gel was filled into a plastic vial (15 mmφ). The filled plastic vial was heated at 50 °C for 24 h. The solidified gel was then removed from the plastic vial. To recover and reuse Pluronic P123, it was subjected to Soxhlet extraction in THF for 48 h. The washed gel was calcined at 500 °C for 6 h at 1 °C/min to provide bulk porous silica.

2.5 Preparation of bulk porous silica-supported ZnO

Bulk porous silica (2 g) was impregnated in a 0.7 mol/L ZnCl2 aq. (100 mL) at room temperature for 24 h. The porous silica was removed from ZnCl2 aq. and dried at 50 °C for 24 h. The dried porous silica was heated to 500 °C at the rate of 1 °C/min and then calcined for 6 h to provide bulk porous silica-supported ZnO.

2.6 Desulfurization ability test

Desulfurization of bulk porous silica-supported ZnO was performed using a specific apparatus reported in the literature [23], as shown in Fig. 1. Bulk porous silica-supported ZnO (~2.3 g) was placed in a stainless tube (inside diameter: 10 mmφ). The inlet gas was mixed with nitrogen, 200 ppm H2S, and methane as an internal standard. The inlet gas flow rate was set at 86 mL/min. The bulk porous silica-supported ZnO with a height of 40 mm and a radius of 5 mm was used. The gas hourly space velocity (GHSV) was calculated by an equation: GHSV = flow rate (8.6 × 10−5 m3/min) × 60 min/h ÷ sample volume (3.14 × 10−6 m3) = 1640 h−1. The gas was passed into the sample at 120 °C. The outlet gas was collected in a gas bag (aluminum bag, 1 L; GL Science), and the concentration of H2S in the bag was determined by GC.

Fig. 1
figure 1

Desulfurization test equipment

Full size image

3 Results and discussion

3.1 Characterization of bulk micro–mesoporous silica

Bulk porous silica was produced from silicic acid, which was prepared by reacting water glass with HCl aq. and extracting with THF. The structure and properties of porous silica differ depending on parameters such as kinds of solvents and silica sources and the ratio of Pluronic P123 to silica source and water [24,25,26]. There are many reports on the preparation of mesoporous silica prepared from TEOS and Pluronic P123 [15, 27,28,29], but there are no reports on the preparation of bulk porous silica via silicic acid in THF. We first investigated the difference in appearance and properties at different ratios of Si/Pluronic P123 and H2O/Si (mol/mol), and bulk porous silica was obtained by a simple process as shown in Fig. 2. It is important to fill the filtered fresh swollen gel into plastic vials.

Fig. 2
figure 2

Preparation of bulk porous silica

Full size image

The preparation ratios of the silica source, Pluronic P123, and water are presented in Table 1, and the SBET values related to the preparation ratios are shown in Fig. 3. The optimum preparation ratio for the TEOS/EtOH system in previous report [15] is given in “cf.”, with Run 1 closely matching these ratios. The ratio varied in Runs 2–8.

Table 1 Preparation ratio of silica source, Pluronic P123, and water, and results on N2 adsorption
Full size table
Fig. 3
figure 3

Phase diagram of difference in SBET between Si/Pluronic P123 and H2O/Si

Full size image

The nitrogen adsorption–desorption isotherms and microporous-mesoporous distributions are shown in Fig. 4 and Table 1. The type IV hysteresis [30] was observed at the isotherms in Runs 6–8; in contrast, the unclear hysteresis is observed at the isotherms in Runs 1–4. A broad distribution was observed for the microporous and mesoporous distributions in Runs 1–4, whereas a sharp distribution was observed in Runs 6–8. The clear formation of micro- and mesopores using silicic acid depended on a high H2O/Si ratio (higher than 115), which seems to be a suitable ratio for the self-organization of Pluronic P123 in silicic acid/THF. The microporous volume was stronger than mesoporous volume, compared with mesoporous silica powder [31]. This tendency is similar to micro–mesoporous silica [32]; therefore, the bulk porous silicas are categorized into micro–mesoporous silica. The SBET had a maximum value of Si/Pluronic P123 = 100 and H2O/Si = 121 (Run 6). Bulk bodies with chalk-like hardness without cracking were obtained in Run 6, as shown in Figure S1. Run 6 was the best preparation using silicic acid and was reproducible.

Fig. 4
figure 4

Nitrogen adsorption–desorption measurements of the milled bulk porous silica: a isotherms, b micropore-size distribution, and c mesopore-size distribution

Full size image

The SEM, TEM, and SAXS images of the bulk porous silica of Run 6 are shown in Fig. 5. The SEM image is observed as particulate morphology in overall. SAX pattern is confirmed that not changed before/ after calcination. The uniform and regular mesoporous characteristics of the porous silica were confirmed using TEM, and the SAXS peak of the (100) plane was observed at q = 0.6 nm−1. The interplanar spacing (d100) was determined to be approximately 10 nm from TEM measurements and calculated as 10.4 nm using the equation d = 2π/q from SAXS data.

Fig. 5
figure 5

a SEM, (b) TEM, and (c) SAXS pattern of the milled bulk porous silica of Run 6

Full size image

Moreover, we tried the recovery and recycling of Pluronic P123. Table 2 shows the recovery rate by extraction using THF and the specific surface area of the bulk porous silica using recycled Pluronic P123. This rate was higher than 85%, and the specific surface areas of the bulk porous silica were similar. Hence, we recovered and recycled Pluronic P123 to reduce the production cost of bulk porous silica. The 1H NMR spectra of pristine and recovery Pluronic P123 were the same (Fig. S2).

Table 2 Results on recycling of Pluronic P123
Full size table

3.2 Desulfurization ability of bulk micro–mesoporous silica-supported ZnO

Bulk porous silica-supported ZnO was prepared by the impregnation of bulk porous silica in aqueous ZnCl2 followed by calcination at 500 °C. The ZnO content in the bulk porous silica-supported ZnO determined by ICP-AES analysis was 15.2 wt%. The SBET value and pore volume of bulk porous silica-supported ZnO were smaller than those of pristine bulk porous silica (Table 3); therefore, ZnO was deposited on the pores. The desulfurization ability of the bulk porous silica-supported ZnO was evaluated using the following method: (i) bulk porous silica was placed in a stainless tube, (ii) a mixed gas (flow rate 86 mL/min; GHSV 1640 h−1) of nitrogen and H2S (H2S concentration of 200 ppm) was circulated in the sample tube at 120 °C, and (iii) the outlet gas was collected, and the concentration of H2S was determined by GC analysis. The desulfurization ability is shown in Fig. 6. Bulk porous silica without ZnO showed no desulfurization ability; that is, SiO2 did not induce the absorption of H2S. The presence of ZnO showed desulfurization ability. The desulfurization reaction over ZnO occurred as ZnO + H2S → ZnS + H2O [29]. The desulfurization ability per gram of bulk porous silica-supported ZnO (SC/sorbent) and the desulfurization ability per gram of supporting ZnO in bulk porous silica-supported ZnO (SC/ZnO) were calculated using Eqs. (1) and (2), respectively:

$${\rm{SC}}/{\rm{sorbent}}=(Q\times {M}_{{\rm{H}}2{\rm{S}}}/{W}_{{\rm{sor}}}\times {V}_{{\rm{M}}})\times ({C}_{0}\times T{\rm{s}}-{\int }_{\!{\rm{o}}}^{T{\rm{s}}}C\left(t\right){\rm{d}}t)$$
(1)
$${\rm{SC}}/{\rm{ZnO}}=(Q\times {M}_{{\rm{H}}2{\rm{S}}}/{W}_{{\rm{sor}}}\times {V}_{{\rm{M}}})\times ({{\rm{C}}}_{0}\times T{\rm{s}}-{\int }_{\!{\rm{o}}}^{T{\rm{s}}}C\left(t\right){\rm{d}}t)/{C}_{{\rm{Zn}}}$$
(2)

where Q is the inlet flow of mixed gas (nitrogen and H2S; 0.086 L/min), MH2S is the molecular weight of H2S (34.06 g/mol), Wsor is the weight of the sorbent, VM is the molar gas volume (22.4 L/mol), C0 is the inlet concentration of H2S (200 ppm), C(t) is the outlet concentration of H2S (ppm), Ts is the breakthrough time at 0.5 ppm of H2S (min), and CZn is the content of ZnO in the sorbent (g/g) [15]. The SC/sorbent and SC/ZnO values (8.7 and 58.1 mg/g) were higher than those reported in our previous study (6.0 and 39.3 mg/g) [15]; this ability is similar to that of mesoporous silica powder-supported ZnO [33]. The SC/sorbent of bulk porous silica-supported ZnO showed three times larger than that of pure ZnO [23]. Desulfurization abilities for supported ZnO sorbent [15, 23, 33,34,35,36] are summarized in Table 4, and bulk porous silica-supported ZnO in this study showed comparable desulfurization performance to others. It is possible that preparing bulk porous silica with a three-dimensional structure such as KIT-6 or loading Fe2O3 or CuO together may provide higher desulfurization ability [37, 38], and further improvements are expected. The nitrogen adsorption–desorption and TEM-EDX results for bulk porous silica-supported ZnO before and after the desulfurization test are shown in Table S1 and Fig. 7. After the desulfurization test, a decrease in the specific surface area, mesopore size, and adsorption of sulfur atoms was observed. No change in the micropore size was observed; therefore, the desulfurization was faster in the mesoporous area.

Table 3 Results on N2 adsorption of bulk porous-silica and its supported ZnO
Full size table
Fig. 6
figure 6

Desulfurization ability of bulk porous silica-supported ZnO (a) desulfurization test, (b) sulfurization degree (GHSV 1640 h−1)

Full size image
Table 4 Summary of desulfurization ability of supported ZnO sorbent
Full size table
Fig. 7
figure 7

Nitrogen adsorption–desorption measurements of bulk porous silica before/after desulfurization and TEM-EDX

Full size image

4 Conclusion

In this study, we prepared bulk porous silica (ca. 10 mmφ) using silicic acid provided by water glass, which has low environmental impact and low cost. When the molar ratio of H2O/Si was 121 and Si/Pluronic P123 was 100, bulk porous silica with chalk-like hardness without cracking and having good surface area and reproducibility was obtained. The specific surface area, micropore size, and mesopore size were calculated to be 818 m3/g, 1.6 nm, and 6.2 nm, respectively, by nitrogen adsorption–desorption. Bulk porous silica was categorized into micro–mesoporous silica. The uniform and regular mesopore characteristics of the porous silica were confirmed by TEM, and the SAXS peak of the (100) plane was observed at q = 0.6 nm−1. Other H2O/Si and Si/Puronic P123 ratios gave porous silica that (i) could not form a bulk, (ii) did not have uniform and regular pores, (iii) or was very brittle. Moreover, Pluronic P123 was recycled, and the recovery rate was higher than 85% by maintaining a specific surface area similar to that of original bulk porous silica.

ZnO was supported on bulk porous silica for H2S adsorption. Bulk porous silica-supported ZnO was prepared by impregnation with ZnCl2 aq. followed by calcination in air. The ZnO content in the bulk porous silica-supported ZnO was 15.2 wt%. The desulfurization ability per gram of bulk porous silica-supported ZnO and the desulfurization ability per gram of supporting ZnO in bulk porous silica-supported ZnO were 8.7 and 58.1 mg/g, respectively. The desulfurization ability was similar to that of mesoporous silica powder-supported ZnO prepared using TEOS. After the desulfurization test, the specific surface area and mesopore size decreased, and adsorption of sulfur atoms was observed. No change in the micropore size was observed; therefore, the desulfurization reaction was faster in the mesoporous area. Further functionalization of porous silica bulk bodies using silane coupling agents and their separation/ adsorption is now under investigation.