**Spherical Activated Carbons with High Mechanical Strength Directly Prepared from Selected Spherical Seeds**

#### **Ana Amorós-Pérez, Laura Cano-Casanova, Mohammed Ouzzine, Mónica Rufete-Beneite, Aroldo José Romero-Anaya, María Ángeles Lillo-Ródenas \* and Ángel Linares-Solano**

MCMA Group, Department of Inorganic Chemistry and Materials Institute, University of Alicante, E-03080 Alicante, Spain; ana.amoros@ua.es (A.A.-P.); laura.cano@ua.es (L.C.-C.); ouzzine\_mohamed@yahoo.fr (M.O.); monica.rufete@ua.es (M.R.-B.); ajromero@ua.es (A.J.R.-A.); linares@ua.es (A.L.-S.)

**\*** Correspondence: mlillo@ua.es; Tel.: +34-965-90-35-45; Fax: +34-965-90-34-54

Received: 21 March 2018; Accepted: 4 May 2018; Published: 10 May 2018

**Abstract:** In the present manuscript, the preparation of spherical activated carbons (SACs) with suitable adsorption properties and high mechanical strength is reported, taking advantage of the retention of the spherical shape by the raw precursors. An easy procedure (carbonization followed by CO2 activation) has been applied over a selection of three natural seeds, with a well-defined spherical shape and thermal stability: *Rhamnus alaternus* (RA), *Osyris lanceolate* (OL), and *Canna indica* (CI). After the carbonization-activation procedures, RA and CI, maintained their original spherical shapes and integrity, although a reduction in diameter around 48% and 25%, respectively, was observed. The porosity of the resulting SACs could be tuned as function of the activation temperature and time, leading to a spherical activated carbon with surface area up to 1600 m2/g and mechanical strength similar to those of commercial activated carbons.

**Keywords:** spherical seeds; spherical activated carbons; activation; microporosity; mechanical properties

#### **1. Introduction**

Spherical activated carbons (SACs) are very interesting materials, which are attracting great attention because of their outstanding physical properties, such as wear resistance, mechanical strength, good adsorption performance, purity, low ash content, smooth surface, good fluidity, good packaging, low pressure drop, high bulk density, high micropore volume and tunable pore size distribution [1–5]. All these features make SACs suitable for various applications like blood purification, catalysts support, chemical protective clothing [2,6,7], in adsorption processes; both in gas phase (e.g., toluene, CO2, CH4 and H2) [5,8–10] and solution (e.g., phenol) [11], as supercapacitors [12,13], in medicine for poison adsorption in living organisms [14], as catalyst supports for hydrogenation reactions [15,16], etc.

SACs can be prepared using several methods: by polymerization reactions [17], by agglomeration from mixtures of resin and activated carbon [18] or by hydrothermal synthesis [19–23]. All these methods imply the use of expensive or synthetic precursors (such as aerogels [17], divinylbenzene-derived polymers [24] and urea/formaldehyde resin [25]). However, nowadays it is common to look for cheaper precursors, such as coals [4], lignocellulosic materials [19–21] and carbohydrates [22,26–29].

Herein we present the preparation of SACs with high mechanical strength and tunable porosity from spherical seeds using an easy, cheap and a well-known method. This simple route allows the valorization of inexpensive and available biomass precursors, such as not edible seeds, to convert them in potentially useful and valuable materials like SACs. In particular, we focused our interest

on the selected seeds that combine simultaneous spherical shape and thermal stability, and cover a wide range of diameters (from 1 to 7 mm). It should be highlighted that the size of the final materials would depend on the size of the used precursor. Among the tested spherical seeds, which accomplish these requirements (Table 1), the study was focused on three of them: *Rhamnus alaternus* (RA), *Osyris lanceolate* (OL), and *Canna indica* (CI) (Figure 1).


**Table 1.** Common and scientific names of the selected spherical seeds and their diameters.

**Figure 1.** Natural seeds used as precursors for SACs preparation, together with the carbonized and activated spherical materials prepared from them.

#### **2. Materials and Methods**

#### *2.1. Methodology*

#### 2.1.1. Carbonization Process

All the seeds were initially dried in an oven at 110 ◦C for 3 h. For the carbonization process, 2 g of such dried seeds were heated up to 850 ◦C, held for 2 h, in a horizontal furnace with N2 flow of 300 mL/min, and a heating rate of 5 ◦C/min. The corresponding carbonization yields are summarized in Table 2.

**Table 2.** Mechanical strength (expressed as the percentage of the remaining mass after sieving, SRM%, see Section 2.2.3.) of the natural seeds and carbonization yields, textural properties and mechanical strength (SRM%) of the materials after carbonization.


<sup>a</sup> SRM, remaining mass after sieving the natural spherical seeds, as percentage. <sup>b</sup> Yield, yield of carbonization process, as percentage. <sup>c</sup> VDR (N2), total micropore volume, obtained applying the Dubinin-Raduskevich method to data of N2 adsorption isotherm at −<sup>196</sup> ◦C. <sup>d</sup> VDR (CO2), narrow micropore volume, obtained applying the Dubinin-Raduskevich method to data of CO2 adsorption isotherm at 0 ◦C. <sup>e</sup> SRM, remaining mass after sieving the carbonized materials, as percentage.

#### 2.1.2. Activation Process

Carbonized seeds were activated using CO2 in order to develop their porosity, using a CO2 flow of 80 mL/min. To study the effect of temperature and time on the activation process, the samples were heated at 5 ◦C/min up to different temperatures: 800, 850, or 880 ◦C, and such temperatures were maintained for various fixed times, as described in Table 3.

**Table 3.** Activation conditions, activation percentages, SRM values and textural properties of some activated samples.


<sup>a</sup> SBET, BET surface area, obtained applying the BET method to data of N2 adsorption isotherm at −<sup>196</sup> ◦C. <sup>b</sup> VDR (N2), total micropore volume, obtained applying the Dubinin–Raduskevich method to data of N2 adsorption isotherm at −<sup>196</sup> ◦C. <sup>c</sup> VDR (CO2), narrow micropore volume, obtained applying the Dubinin–Raduskevich method to data of CO2 adsorption isotherm at 0 ◦C. <sup>d</sup> Vmeso, mesopore volume, obtained from N2 adsorbed as liquid at P/Po = 0.9 minus the adsorbed volume at P/P0 = 0.2 [30]. <sup>e</sup> SRM, remaining mass after sieving the activated materials, as percentage. <sup>f</sup> NM: not measured. <sup>g</sup> VN2–VCO2, difference between VDR (N2) and VDR (CO2).

#### *2.2. Characterization*

#### 2.2.1. Morphology

Morphology of the original, carbonized and activated samples was characterized by Scanning Electron Microscopy (SEM) in a JSM-840 microscope (JEOL, Tokyo, Japan) with a scintillator–photomultiplier type secondary electron detector.

#### 2.2.2. Surface Area and Pore Volumes

Textural characterization of precursors, carbonized and activated materials was performed using N2 adsorption at −196 ◦C [31] and CO2 at 0 ◦C [32] in a volumetric Autosorb-6B apparatus from Quantachrome. Before analysis, the samples were degassed at 250 ◦C for 4 h. The BET equation was applied to the nitrogen adsorption isotherm in the low-pressure region (relative pressure between 0.05–0.25) to get the apparent BET surface area, SBET [30]. The Dubinin–Radushkevich equation was applied to the nitrogen adsorption isotherm to determine the total micropore volume (VDR (N2) corresponding to micropores of size below 2 nm) and to the carbon dioxide adsorption isotherms to determine narrow micropore volume (VDR (CO2), corresponding to micropores of size below 0.7 nm) [33]. Mesopore volume, Vmeso, corresponding to pores between 2 and 20 nm, was estimated from N2 adsorbed as liquid at P/P0 = 0.9 minus the volume adsorbed at P/P0 = 0.2 [30]. The difference between VDR (N2) and VDR (CO2) was calculated as an estimation of the micropore size distribution [30,31].

#### 2.2.3. Mechanical Properties

The mechanical strength, defined as SRM%, was estimated by a method developed in our laboratory that consists of the evaluation of the sample mass remaining in a sieve after vigorous shaking (Figure 2). Thus, for each test, a known quantity of material was put in a cylindrical vial together with 15 stainless steel balls (Figure 2a). These vials were placed horizontally in a polymer mold used as immobilizing support (Figure 2b), which was placed in an electromagnetic sieve shaker CIPSA RP08 Ø200/203 for 20 min at power number 8 (equivalent to the shaking speed of 1.8 mm of vibration amplitude per second) (Figure 2c). Then, the samples were sieved using a sieve (300 μm) and the resulting residue (the sample not converted to dust in the sieving step) was weighed (Figure 2d). The mechanical strength was expressed as the percentage of the remaining mass after sieving (SRM%). The validation of this method was performed by analyzing the mechanical properties of several commercial activated carbons (Table 4), and such values were used as reference to confirm that the mechanical properties of our SACs are similar to those of their commercial counterparts. Note that the SRM values for the selected commercial ACs are in the range between 72% and 97%. The analysis of the mechanical properties was performed for the precursors, carbonized materials (Table 2) and for the activated ones (Table 3).

**Figure 2.** Depiction of materials and procedure for the mechanical strength evaluation of the samples: (**a**) a weighted sample was sieved using in 300 μm sieve and put in a vial; (**b**) 15 steel balls were also incorporated in the vial, which was then placed in a polymer mold; (**c**) the molds were placed in the sieve shaker during 20 min; (**d**) the sample was sieved again in a 300 μm sieve and the residue (the sample not converted to dust in the sieving step) was collected and weighed to calculate the sample remaining mass percentage (SRM%).


**Table 4.** Textural properties and SRM values of some commercial activated carbons.

<sup>a</sup> SBET, BET surface area, obtained applying the BET method to data of N2 adsorption isotherm at −<sup>196</sup> ◦C. <sup>b</sup> VDR (N2), total micropore volume, obtained applying the Dubinin-Raduskevich method to data of N2 adsorption isotherm at −<sup>196</sup> ◦C. <sup>c</sup> VDR (CO2), narrow micropore volume, obtained applying the Dubinin–Raduskevich method to data of CO2 adsorption isotherm at 0 ◦C. <sup>d</sup> Vmeso, mesopore volume, obtained from N2 adsorbed as liquid at P/P0 = 0.9 minus the adsorbed volume at P/Po = 0.2 [30].

#### **3. Results**

Figure 3 shows that the retention of the desired original spherical shape during the carbonization process was achieved for these three seeds, though this step generally led to a decrease in the diameter of the materials. Such variation depends on the type of seed: for RA the size was significantly reduced (around 3 mm, which represents 40% reduction), while CI and OL were only slightly shortened (in both cases about 1 mm, around 20%). This could be related with the intrinsic natural differences in the composition of the seeds.

**Figure 3.** SEM images of precursors, carbonized materials and final SACs. Experimental preparation conditions of the materials: (**a**–**c**) dried at 110 ◦C for 3 h; (**d**–**f**) carbonized at 850 ◦C for 2 h in 300 mL/min N2 flow; (**g**) activated at 800 ◦C for 30 h using 80 mL/min CO2 flow; (**h**) activated at 800 ◦C for 2 h in 80 mL/min CO2 flow; (**i**) activated at 800 ◦C for 5 h in 80 mL/min CO2 flow.

Table 2 reports carbonization yields and values of mechanical properties (SRM%) for CI, RA, and OL carbonized seeds, together with SRM values for the precursors, as reference. It shows that: (i) the carbonization yields ranged from 21 to 30%, which is in the range of typical values expected for lignocellulosic materials [34]; (ii) micropore volumes determined by CO2 adsorption were larger

than those measured by N2, indicating that the mean micropore sizes were below 0.7 nm [31], and (iii) SRM values were higher than 98.7%, indicating that the samples possess high mechanical resistance. Such porous texture, together with the mechanical properties, made these carbonized materials potentially useful as spherical carbon molecular sieves.

The conditions of the activation process were also optimized in order to obtain similar burn-off percentage (around 30%) and maintain the spherical morphology. Figure 3 shows that RA and CI seeds retained their original shape after activation, and their sizes were minimally affected (around 0.5 mm (8% reduction with respect to the size) before the activation, and 0.2 mm (5%), respectively). Only OL seeds were broken after activation, and this occurred for all the explored activation conditions. Hence, from OL only spherical carbonized materials could be prepared. It is important to mention that the carbonized and activated materials from both RA and CI remained physically intact (without cracks) and the same occurred for carbonized OL, whereas only the material obtained from OL, after the activation process showed cracks, that can be visually distinguished.

With respect to the activation yields (Table 3), CI was the more reactive candidate, since a shorter activation time was required to get 33% burn-off. For the RA precursor, as expected, the burn-off percentage at constant temperature increases proportionally with the reaction time. Interestingly, the desired activation percentage (33%) could also be achieved for RA using higher temperatures (850 ◦C instead of 800 ◦C) and shorter times (10 h instead of 40 h).

Regarding the textural properties, Table 3 shows that, for RA seeds, although there exists a linear relationship between the activation time and the burn-off percentage, no direct correlation was found when analyzing the effect of the activation time on the porosity development. For this precursor, the same burn-off percentage, 33%, and similar surface area values (about 880 m2/g) have been obtained using different combinations of activation temperature and time. Low activation times (10 and 30 h) led to materials with the mean micropore sizes below 0.7 nm, whereas the micropore size was around that value for larger activation time and temperature.

Similar experimental conditions screening for CI precursor highlighted that higher adsorption capacities can be developed from it, which reached 1616 m2/g when treating up to 880 ◦C for 3 h. For the CI activated material with surface area above 1600 m2/g, the fact that total micropore volume determined by N2 adsorption is much larger than that measured by CO2, is indicative of the average pore size above 0.7–1 nm [30].

By comparing SRM values for natural and carbonized materials in Table 2, it can be observed that natural precursors show slightly higher mechanical strength than the corresponding carbonized spheres. Table 3 contains the SRM values for SACs, indicating that their mechanical properties are only slightly reduced after the activation process. This is probably due to the high number of heteroatoms linked to the carbon material, and eliminated during the activation [35]. However, SRM values are in the range of 95% for samples with areas around 800 m2/g, and about 85% when the BET surface area surpasses 1600 m2/g, which indicated that the samples generally display significant mechanical properties that lie in the range of the selected common commercial references (CW, CK and ROX) (Table 4 and Figure 4).

**Figure 4.** SRM values for commercial activated carbons (black color) and for some activated seeds prepared in this work (orange color).

#### **4. Conclusions**

In this work spherical activated carbons with high mechanical strength and well-developed porosity were prepared while maintaining the spherical shape of the natural seeds, selected as carbon precursors. The three reported candidates: RA, OL and CI, could be successfully converted into spherical activated carbon materials using a well-known, simple and cheap method and, additionally, CI and RA maintained their original spherical shapes and integrity all along the activation process, avoiding breakage. Their diameter sizes were notably reduced after the carbonization step and only slightly affected by the activation process. The mechanical properties for all the activated materials were found to be similar to those of common commercial activated carbons with different morphologies (granular, spherical and pellets). Interestingly, depending on the precursor and/or on the activation conditions, significant differences in porosity development and micropore size distributions were obtained, reaching specific surface areas up to 1600 m2/g. The interesting properties of the prepared materials, together with their spherical morphology, make them interesting candidates for many applications.

**Author Contributions:** Á.L.-S. conceived and designed the experiments; A.A.-P., L.C.-C., M.O., M.R.-B. and A.J.R.-A. performed the experiments; A.A.-P., L.C.-C., M.O., M.A.L.-R. and Á.L.-S. analyzed the data; A.A.-P., M.O., A.J.R.-A., M.A.L.-R. and Á.L.-S. contributed reagents/materials/analysis tools; A.A.-P., A.J.R.-A., M.A.L.-R. and Á.L.-S. wrote the paper.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors thank the Spanish Ministry of Economy and Competitiveness (MINECO) and FEDER, project of reference CTQ2015-66080-R, GV/FEDER (PROMETEOII/2014/010) and University of Alicante (VIGROB-136) for financial support. Pilar Garcia Cardona, from Universidad Nacional de Colombia-Medellin is acknowledged for providing some of the spherical materials.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **One-Step Hydrothermal Synthesis of Zeolite X Powder from Natural Low-Grade Diatomite**

#### **Guangyuan Yao, Jingjing Lei, Xiaoyu Zhang, Zhiming Sun \* and Shuilin Zheng \***

School of Chemical and Environmental Engineering, China University of Mining and Technology (Beijing), Beijing 100083, China; slxw.yao1990@hotmail.com (G.Y.); 18811568019@163.com (J.L.); zhangxiaoyukdbj@163.com (X.Z.)

**\*** Correspondence: zhimingsun@cumtb.edu.cn (Z.S.); shuilinzheng8@gmail.com (S.Z.)

Received: 15 May 2018; Accepted: 25 May 2018; Published: 28 May 2018

**Abstract:** Zeolite X powder was synthesized using natural low-grade diatomite as the main source of Si but only as a partial source of Al via a simple and green hydrothermal method. The microstructure and surface properties of the obtained samples were characterized by powder X-ray diffraction (XRD), scanning electron microscopy (SEM), wavelength dispersive X-ray fluorescence (XRF), calcium ion exchange capacity (CEC), thermogravimetric-differential thermal (TG-DTA) analysis, and N2 adsorption-desorption technique. The influence of various synthesis factors, including aging time and temperature, crystallization time and temperature, Na2O/SiO2 and H2O/Na2O ratio on the CEC of zeolite, were systematically investigated. The as-synthesized zeolite X with binary meso-microporous structure possessed remarkable thermal stability, high calcium ion exchange capacity of 248 mg/g and large surface area of 453 m2/g. In addition, the calcium ion exchange capacity of zeolite X was found to be mainly determined by the crystallization degree. In conclusion, the synthesized zeolite X using diatomite as a cost-effective raw material in this study has great potential for industrial application such as catalyst support and adsorbent.

**Keywords:** diatomite; zeolite X; hydrothermal method; calcium ion exchange capacity

#### **1. Introduction**

Zeolites are crystalline aluminosilicates built from TO4 tetrahedra (T = Si and Al) with excellent properties of high surface area, uniform and precise microporosity, shape selectivity, high ion-exchange capacity, strong Brønsted acidity and high thermal and hydrothermal stability [1]. Therefore, zeolites have been widely used in many environmental and other industrial applications, such as ion exchange [2–5], catalysts [6–9], membrane separations [10–12] and adsorbents [13–17].

The principle raw materials used for the synthesis of the zeolites are different sources of silica and alumina, which are usually composed of sodium silicates, sodium aluminate, aluminum salts or colloidal silica. However, traditional methods for synthesizing zeolites typically involve chemical reagents as starting materials or crystallization from a gel or clear solution under hydrothermal conditions, which have the disadvantages of high cost, excessive waste, and unfriendly nature to the environment. Therefore, many attempts are underway for economical synthesis of zeolites. In general, natural aluminosilicate and silicate minerals and industrial solid wastes have been explored as silica and/or alumina source because they are cost-effective precursors and can lead to reduction of the synthesis costs. Until now, There have been many studies on synthesizing zeolites from natural minerals such as kaolinite [18–20], bentonite [19], feldspar [19,21] and other precursors [22–26].

Although zeolites have been synthesized from the solid wastes, such as fly ash [27–29], rice husk ash [30] and coal gangue [22], the uncertainty in their supplies and the impurity in their components may limit their practical application. Therefore, direct synthesis of zeolites from natural aluminosilicate and silicate minerals with abundant reserves in the earth has been pursued because of its great potential in reducing the generation of hazardous wastes, saving energy, and possibly altering the properties of the resulting zeolites [31]. However, most aluminosilicate minerals are inactive, which restricted their practical application in the zeolite synthesis. Additionally, even after the thermal activation, only part of the aluminum-oxygen bonds can be broken, which means that just part of the Al2O3 and a small amount of SiO2 only can participate in the zeolite synthesis.

Diatomite is an interesting silica material because of its relatively low cost, large reserves and its highly reactive amorphous state derived from silica skeletons of diatoms. Being silica rich, diatomite serves as the silica source in the synthesis of zeolites, which is cost-effective. In addition, diatomite is amorphous and highly reactive and therefore, unnecessary to transform it into a reactive state as done with many crystalline minerals [18–21]. Furthermore, parts of secondary building units of minerals could be preserved in the synthesis process. Therefore, it is necessary to explore detailed research on the preparation and formation mechanisms of zeolites from diatomite.

In the present work, we report on the synthesis of zeolite X from natural low-grade diatomite with high crystallization degree under hydrothermal conditions. Additionally, the effect of factors such as aging time and temperature, crystallization time and temperature, Na2O/SiO2 and H2O/Na2O ratio on the synthesis of zeolite in the SiO2-Al2O3-Na2O-H2O system were systematically investigated. The crystallization degree of the products was evaluated by XRD, SEM and CEC analysis. Meanwhile, the formation mechanism was also explored and discussed.

#### **2. Experimental**

#### *2.1. Materials*

Diatomite (Dt) is obtained from Linjiang City, Jilin Province, China. Its main chemical composition by wt % is: SiO2: 63.77%; Al2O3: 18.97%; Fe2O3: 1.48%; CaO: 0.48%; K2O: 0.16%; Na2O: 0.04%. It was grounded to a size smaller than 30 mesh and dried at 105 ◦C. Commercial zeolites X were purchased from Tianjin yuanli Reagent Co. (Tianjin yuanli, China). Sodium hydroxide, aluminum hydroxide (nordstrandite) and the other chemicals used in the experiments were purchased from Xilong Reagent Co. (Xilong, China). All chemicals were of analytical reagent grade and used without any further purification. Deionized water was used throughout this study.

#### *2.2. Preparation of Zeolite X*

The synthesis of zeolites from diatomite includes three processes as follows: gel formation, aging and crystallization. Initially, diatomite and Al(OH)3 were dispersed in NaOH solution under vigorous magnetic stirring to form a homogeneous dispersion. The amount of diatomite and Al(OH)3 were according to the 1.13 (molar ratio) of [Si/Al], and the amount of NaOH solution was according to the molar ratios of [Na2O/SiO2] and [H2O/Na2O]. Subsequently, the slurry was subjected to aging for 0–120 min at 30–60 ◦C. Then, the mixed solution was put into a Teflon-lined stainless steel autoclave. Finally, the container was closed and crystallized at 90–120 ◦C for 3–9 h. After that, the autoclave was cooled to room temperature naturally, and the samples were removed from the reactor, filtered, and washed with deionized water until the pH of the filtrate reached 6–7. Finally, the wet washed solids were dried overnight at 105 ◦C before further measurement and characterization.

#### *2.3. Characterization*

X-ray diffraction (XRD) analysis were performed on a D8 advance X-ray diffractometer (Bruker, Germany) equipped with Cu-Kα radiation (λ = 0.154056 nm) to identify the crystalline phase of the obtained X zeolite products. The samples were scanned in the 2θ range of 5◦ to 50◦ with a 0.02◦ step at a scanning speed of 4◦/min. The surface morphology of the samples was observed by S-4800 scanning electron microscope (Hitachi, Japan). Nitrogen adsorption-desorption isotherms were measured at 77 K using an ASAP 2020 instrument (Micromeritics, Norcross, GA, USA), after evacuation of the samples at 350 ◦C for

4 h. The specific surface area (SBET) and microporous volume (Vmicro) were calculated using the BET and t-plot methods, respectively. Pore size distribution curves were calculated by Barrett-Joyner-Halenda (BJH) method. The crystallization behavior of zeolites as well as the thermal properties of the composites was monitored and evaluated using a Mettler TGA/DSC 1 SF/1382 equipment (NETZSCH, Germany). The TGA/DSC measurements were carried out in air flow with a heating rate of 5 ◦C/min from 25–900 ◦C. Chemical composition of the sample was determined using wavelength dispersive X-ray fluorescence spectrometry (XRF, Shimadzu, Japan) on a Shimadzu XRF-1800 apparatus. Calcium ion exchange capacity (CEC) was determined as follows [22]: Typically, 0.5 g of zeolite sample was poured into 500 mL of 0.005 M CaCl2 solution and the mixture was shaken for 20 min at 35 ◦C. Then the filtrate was analyzed by the addition of calconcarboxylic acid and EDTA to determine the CEC of the samples.

#### **3. Results and Discussion**

#### *3.1. Starting Materials*

The XRD pattern (a) and SEM images (b,c) of raw diatomite are shown in Figure 1. According to the XRD pattern of diatomite, the broad reflection centered at 2θ = 15–30◦ was attributed to the amorphous silica, and the peaks at 2θ = 20.08◦ and 26.65◦ were ascribed to quartz. In addition, the peaks at 2θ = 11.88◦, 27.35◦ and 35◦–40◦ were characteristic to kaolinite-montmorillonite [32], which were the main Al source. As shown in Figure 1b,c, the diatomite exhibits highly porous cylinder-like or boat-like shape.

**Figure 1.** XRD pattern (**a**) and SEM images (**b**,**c**) of raw diatomite. Q = quartz; M = montmorillonite; K = kaolinite.

#### *3.2. Effect of Crystallization*

Crystallization conditions are important parameters that control the crystallization of zeolites. The crystallization conditions could also change the autogenous pressure in the autoclave and may alter the structure of the resulting zeolites. Then, a batch of experiments were carried out under different crystallization temperatures and times, while the other preparation conditions including aging temperature and time, Na2O/SiO2 and H2O/Na2O ratio were kept constant. Namely: 30 ◦C of aging temperature, 60 min of aging time, 40 of H2O/Na2O ratio and 3.0 of Na2O/SiO2 ratio. The XRD patterns of samples with different crystallization temperatures and times are shown in Figure 2a,b, respectively. It can be seen from Figure 2a that the sample at a crystallization temperature of 90 ◦C showed only one

strong diffraction peak at 2θ = 18.35◦, which can be attributed to aluminum hydroxide indicating that diatomite was dissolved in the alkaline solution, but the silica did not react with aluminum hydroxide. When the temperature was raised to 100 ◦C, typical diffraction peaks of X zeolite (JCPDS 38-0237) can be seen at 2θ = 6.10◦, 9.97◦, 11.76◦, 15.39◦, 18.46◦, 20.12◦, 23.24◦, 26.58◦ and 30.86◦ suggesting that zeolite X started to crystallize (Figure 2a). However, the diffraction peak of aluminum hydroxide can also be seen at 2θ = 18.35◦, which suggested an incomplete reaction of aluminum hydroxide and silica from diatomite. Highly crystalline single phase zeolite X was formed as the crystallization temperature was raised to 110 ◦C. When the reaction temperature was further raised to 120 ◦C, almost pure phase of zeolite X was observed along with some zeolite A (JCPDS 43-0142). It indicated that the higher crystallization temperature of 120 ◦C caused the transformation of some zeolite X into zeolite A. As shown in Figure 2b, no zeolite X was obtained at crystallization time of 3 h at 110 ◦C. With the increase of crystallization time, the crystallization degree of zeolite X was enhanced. However, longer reaction time caused the transformation of zeolite X into zeolite A at 110 ◦C. Therefore, suitable crystallization temperature and time are essential for the formation of high purity zeolite X.

**Figure 2.** XRD patterns showing the effect of (**a**) crystallization temperature at crystallization time of 5 h and (**b**) crystallization time at crystallization temperature of 110 ◦C on zeolite formation.

SEM images of samples obtained at different crystallization temperatures are shown in Figure 3. When the crystallization temperature was 90 ◦C, the reactants formed spherical aggregates of ill-defined particles. With the increase of crystallization temperature, the products gradually formed regular crystals with smooth faces. However, the crystallization temperature at 110 ◦C formed better octahedral crystal morphology along with uniform distribution of crystals (Figure 3e,f). In addition, amorphous material was hardly noticeable in the SEM images of as-synthesized zeolite X suggesting the formation of highly crystalline zeolite X. When the crystallization temperature was increased up to 120 ◦C, some cubic crystals appeared (Figure 3g,h), which could be attributed to the formation of zeolite A.

The morphologies of the samples prepared at different crystallization times at 110 ◦C are shown in Figure 4. When the crystallization time was 4 h, the synthesized products formed crystals with no distinct faces, and the particles were not uniform. When the crystallization time reached 5 h, the synthesized products formed regular crystals with smooth faces. When the crystallization time was up to 6 h, the synthesized products do not show smooth faces. The morphologies as determined by SEM with different crystallization temperatures and times are in accordance with the XRD results. When the crystallization temperature and time were 110 ◦C and 5 h, the prepared samples showed higher crystallization degree.

The influence of crystallization temperature and time on the CEC is presented in Figure 5a,b. Increasing temperature of crystallization up to 110 ◦C increased the CEC but by further increasing the temperature up to 130 ◦C, the CEC of prepared samples gradually decreased (Figure 5a). Furthermore, the CEC of samples firstly increased with the increase of crystallization time and then gradually decreased when the time was over 5 h (Figure 5b). Combined with the analysis of SEM and XRD, it is concluded that the CEC increased because of the enhancement of crystallization degree of zeolite X.

**Figure 3.** SEM images of zeolites obtained with crystallization temperature at (**a**,**b**) 90 ◦C, (**c**,**d**) 100 ◦C, (**e**,**f**) 110 ◦C and (**g**,**h**) 120 ◦C.

**Figure 4.** SEM images of zeolites obtained with crystallization time at (**a**,**b**) 4 h, (**c**,**d**) 5 h and (**e**,**f**) 6 h at 110 ◦C.

**Figure 5.** Effect of (**a**) crystallization temperature and (**b**) crystallization time on CEC.

#### *3.3. Effect of Aging*

Aging also played an important role in the nucleation of amorphous gel. During this stage, the structure and composition of the silica-alumina gel changed along with the aging conditions. Meanwhile, the aluminosilicate species included in the gel phase were also transformed [33]. To investigate the effect of aging conditions on the structure of the products, a batch of experiments under different aging temperatures and times were carried out by keeping the crystallization temperature and time constant at 110 ◦C and 5 h, respectively, and the alkalinity of the base solutions

at the initial values. The XRD patterns of samples prepared at different aging temperatures and times are shown in Figure 6a,b, respectively. The XRD peaks of the prepared products at different aging temperatures exhibited similar crystallinity of zeolite X (Figure 6a) indicating that aging temperature within this range played only a minor role, if any in the formation of zeolite X. However, the synthetic product without aging exhibited extremely weak diffraction peaks of zeolite X (Figure 6b). Also, there was a sharp diffraction peak at 2θ = 18.35◦, which could be attributed to the unreacted aluminum hydroxide. The diffraction peaks of the prepared products at different aging times exhibited the same typical features of zeolite X with high crystallization degree. However, when the aging time was up to 60 and 120 min, some obvious diffraction peaks of zeolite A appeared indicating that the aging time plays a significant role in the formation of zeolite prepared from diatomite.

**Figure 6.** XRD patterns showing the effect of (**a**) aging temperature at aging time of 60 min and (**b**) aging time at aging temperature of 30 ◦C on zeolite formation after treatment at 110 ◦C and 5 h.

The CEC values of samples with different aging temperatures and times are displayed in Figure 7a,b, respectively. As shown in Figure 7a, the CEC values of samples at different aging temperatures fluctuated little. The CEC values firstly increased with the increase of aging time and then gradually decreased when the time is over 30 min (Figure 7b). Hence, on the basis of XRD and CEC analysis, samples with high crystallization degree possessed high CEC value.

**Figure 7.** Effect of (**a**) aging temperature and (**b**) aging time on CEC of zeolites prepared upon hydrothermal treatment at 110 ◦C and 5 h.

#### *3.4. Effect of Alkalinity*

Another important parameter that controls the nucleation and crystal growth of zeolite is the alkalinity of the base solution, which includes H2O/Na2O and Na2O/SiO2 ratios [34]. Generally high alkaline concentration of the system could accelerate the dissolution of silicon and aluminum components in the precursor materials, which will shorten the induction period and nucleation time, and then speed up the crystallization rate. To investigate the effect of alkalinity on the structure of the products, a batch of experiments under different H2O/Na2O and Na2O/SiO2 ratios were carried out with the crystallization temperature and time of 110 ◦C and 5 h and the aging temperature and time of 30 ◦C and 30 min. The XRD patterns of samples with different H2O/Na2O and Na2O/SiO2 ratios are shown in Figure 8a,b, respectively. As the H2O/Na2O ratio increased, its crystallization degree was enhanced (Figure 8a). When the H2O/Na2O ratios were 40 and 45, the crystallization degree reached the maximum and there were also some zeolite A impurities. However, when the H2O/Na2O ratios were 45, a relatively sharp diffraction peak of aluminum hydroxide appeared (Figure 8a), which indicated that diatomite and aluminum hydroxide did not react completely. When the H2O/Na2O ratio was up to 50, there was only one sharp diffraction peak attributable to aluminum hydroxide. This reveals that higher H2O/Na2O ratios may inhibit the formation of zeolite X from diatomite, probably because low alkalinity reduces the dissolution of diatomite and results in a low conversion to zeolites. Figure 8b shows that the crystallization degree at different Na2O/SiO2 ratio was relatively high, which indicated that Na2O/SiO2 ratio within this range played minor effects in the formation of zeolite X.

**Figure 8.** XRD patterns showing the effect of (**a**) H2O/Na2O ratio and (**b**) Na2O/SiO2 ratio on zeolite formation upon hydrothermal treatment at 110 ◦C and 5 h.

SEM images of samples with different Na2O/SiO2 ratios are presented in Figure 9. Figure 9 shows that all samples with Na2O/SiO2 ratios of 1.3, 1.4 and 1.5 exhibited high crystallinity with smooth faces. On the other hand, the sample with Na2O/SiO2 ratio of 1.4 exhibited better octahedral structure and uniform distribution.

**Figure 9.** SEM images of zeolites obtained at Na2O/SiO2 ratio of (**a**,**b**) 1.3, (**c**,**d**) 1.4 and (**e**,**f**) 1.5 upon hydrothermal treatment at 110 ◦C and 5 h.

The influence of H2O/Na2O and Na2O/SiO2 ratios on the CEC is shown in Figure 10a,b. As shown in Figure 10a, the increase of H2O/Na2O ratio resulted in the increase of the CEC. However, when the H2O/Na2O ratio was higher than 40, its CEC gradually decreased. It can be attributed to the low alkalinity, which reduces the dissolution of diatomite and results in a low conversion to zeolites. As displayed in Figure 10b, the CEC at different Na2O/SiO2 ratios maintained a high value and fluctuate only a little. It is obvious that higher crystallization degree led to higher CEC value. Meanwhile, the calcium ion exchange capacity of the prepared zeolite X sample is 248 mg/g, which is higher than that of commercial zeolites X (232 mg/g).

**Figure 10.** Effect of (**a**) H2O/Na2O ratio and (**b**) Na2O/SiO2 ratio on CEC upon hydrothermal treatment at 110 ◦C and 5 h.

#### *3.5. TG-DTA and XRF Analysis*

The chemical composition of the synthesized X zeolite is presented in Table 1. The synthesized X zeolite is composed of Si, Al, and Na. Meanwhile, the Si/Al ratio is 1.21, which agrees with the theoretical value of 1.13 in the raw material. In addition, the contents of Si in the raw material and the synthesized X zeolite is 3.22 and 2.88 g, respectively, which is a highly efficient use of diatomite. Thermal behavior of the obtained zeolite X sample was investigated using simultaneous TG/DTA thermoanalytical techniques. Typical TG/DTA thermograms for the prepared zeolite X sample in the temperature range of 25–900 ◦C are shown in Figure 11a. TG results showed that the synthesized zeolite X products lost all its moisture (22 wt %) at temperatures lower than 350 ◦C. In general, this kind of weight loss was due to the removal of water adsorbed on the zeolite surface and that present in the zeolite channels. Furthermore, DTA curves showed that the endothermic peaks occurred at lower temperatures (150 ◦C) of the synthesized zeolite X sample, which could be assigned to the loss of adsorbed water. However, the exothermic peaks at the temperature of 820 ◦C could be attributed to the framework collapse and crystallization of NaAlSiO4 (JCPDS 52-1342), which can be attributed to the nepheline (Figure 11b). The TG-DTA analysis indicated that the synthesized X zeolite possessed excellent thermal stability.

**Figure 11.** TG-DTA curves (**a**) of the prepared zeolite X sample at crystallization time and temperature, aging time, and temperature, H2O/Na2O and Na2O/SiO2 ratio of 5 h and 110 ◦C, 30 min and 30 ◦C, 40 and 1.4 and XRD pattern (**b**) of the prepared zeolite X sample after calcinating at 820 ◦C.

**Table 1.** The composition of the prepared zeolite X sample at crystallization time and temperature, aging time, and temperature, H2O/Na2O and Na2O/SiO2 ratio of 5 h and 110 ◦C, 30 min and 30 ◦C, 40 and 1.4.


#### *3.6. N2 Adsorption Performance*

The N2 adsorption-desorption plot at 77 K for the prepared zeolite X is presented in Figure 12. Figure 12 shows a type I isotherm with the presence of steep nitrogen uptake at very low relative pressures (p/p0 = 0.03), which is attributed to the filling of micropores. Meanwhile, an obvious type H4 hysteresis loop (from 0.45 to 0.99) was observed, corresponding to the filling of uniform slit-shaped intercrystal mesopores, which was ascribed to the packing of zeolite crystals [35]. Thus, the zeolite X synthesized from diatomite possessed binary meso-microporous structure. The textural properties of the prepared zeolite X are summarized in Table 2. The specific surface area and total pore volume were up to 453 m2/g and 0.2838 cm3/g, respectively.

**Table 2.** Textural properties of the zeolite X.


**Figure 12.** Nitrogen adsorption/desorption isotherms at 77 K of as synthesized zeolite X.

A comparison of the X zeolites prepared from different Si and Al sources are summarized in Table 3. The surface area of zeolite X prepared here is relatively high. Another comparison of different methods in the preparation of zeolites with diatomite as Si source is presented in Table 4. It can be seen from Table 4 that most methods need acid activation and calcination, which are not environmentally friendly and of high-cost. In addition, these conventional methods also take a longer time to achieve the transformation from minerals to zeolites. Therefore, the one-step hydrothermal method proposed in this research is more environmentally friendly and cheaper.


**Table 3.** Comparison of different Si and Al sources on the surface areas of synthetic zeolite X.

**Table 4.** Comparison of different methods in the preparation of zeolites with diatomite as Si source.


#### **4. Conclusions**

In this work, the zeolite X was obtained from diatomite via a simple hydrothermal method. In addition, the optimum preparation conditions of zeolite X were 5 h of crystallization time, 110 ◦C of crystallization temperature, 30 ◦C of aging temperature, 30 min of aging time, H2O/Na2O ratio of 40 and Na2O/SiO2 ratio of 1.4. The prepared pure zeolite X with binary meso-microporous structure possessed remarkable thermal stability, high calcium ion exchange capacity of 248 mg/g and large surface area of 453 m2/g. Furthermore, it is shown here that the calcium ion exchange capacity of the samples is mainly determined by their crystallization degree and higher crystallization degree means higher calcium ion exchange capacity. Compared with the traditional synthesis techniques, the hydrothermal process developed here is simple, low-cost, and environmentally friendly. In addition, the high purity of zeolite X using natural low-grade diatomite as raw material may be useful for potential industrial application as catalyst support or adsorbent.

**Author Contributions:** Investigation, Visualization, Writing-original draft, G.Y.; Investigation, J.L.; Investigation, X.Z.; Supervision, Z.S.; Conceptualization, Supervision, S.Z.

**Funding:** This research is supported by the Young Elite Scientists Sponsorship Program by CAST (2017QNRC001), Yueqi Funding Scheme for Young Scholars (China university of Mining &Technology, Beijing) and the Fundamental Research Funds for the Central Universities (2015QH01 and 2010YH10).

**Acknowledgments:** The first author thanks the China Scholarship Council (CSC) for financial support.

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
