A facile method of transforming FGD gypsum to α-CaSO4·0.5H2O whiskers with cetyltrimethylammonium bromide (CTAB) and KCl in glycerol-water solution

A facile method to transform flue gas desulfurization gypsum (FGD gypsum) to α- calcium sulfate hemihydrate (α-HH) whiskers with high aspect ratios mediated by cetyltrimethylammonium bromide (CTAB) and KCl in glycerol-water solutions was studied. Addition of KCl facilitated the dissolution of calcium sulfate dihydrate (DH) and created a much higher supersaturation, which could come into being a larger driving force for the phase transformation from DH to α-HH. CTAB as the crystal modifier can significantly promoted 1-D growth of α-HH whiskers along the c axis and the presence of 0.25% CTAB (by weight of FGD gypsum) resulted in the increase of the average aspect ratio of α-HH whiskers from 28.9 to 188.4, which might be attributed to the preferential adsorption of C16H33(CH3)3N+ on the negative side facets of α-HH crystal.

Recent scientific researches find that because the alcohol aqueous solution has a low water activity, the transformation from DH to α-HH can be realized in alcohol water solution under mild conditions 30,31 . Compared with the autoclave method, this process with the advantages of mild reaction conditions and no corrosion to equipment is more favorable to the continuous industrial production. Although the transition from DH to α-HH is kinetically unfavorable in alcohol water solution, addition of small amount of non-lattice cations (such as K + , Na + , Mg 2+ and Zn 2+ ) can significantly shorten the transition time and improve the transition efficiency 32 .
In this paper, we introduced a facile method to prepare α-CaSO 4 ·0.5H 2 O (α-HH) whiskers with small diameters and high aspect ratios by use of FGD gypsum in glycerol-water solutions. In the preparation process, CTAB and KCl were used as the crystal modifier and phase transformation accelerator, respectively, and the corresponding mechanism was studied.

Experimental
Materials. The FGD gypsum, of which the chemical compositions were given in Table 1, was received from Panzhihua Iron & Steel Co., Ltd., Sichuan Province, China. Glycerol, KCl and Cetyltrimetyl Ammonium Bromide (CTAB) were purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China.
Experimental procedure. Pretreatment of FGD gypsum. The gypsum was calcinated at 100 °C for 5 h in a drying oven, then mixed with enough deionized water and stirred (100 rpm) at room temperature for half hours. After complete hydration, the gypsum was filtrated and dried at 60 °C.
Preparation of α-HH whiskers. Glycerol (65 wt%)-water solutions with a certain amount of KCl and CTAB were firstly added into a three-necked flask equipped with a reflux condenser on top of it, and the solution was stirred with a magnetic stirrer at a constant rate of 150 rpm and preheated to 90 °C in an oil bath. Then 50 g FGD gypsum (10 wt % solid content) after pretreated was added into the reactor. During the reaction, 20 mL hot slurry was withdrawn at certain time intervals. Half of the sample was filtrated immediately, washed three times with boiling water and then rinsed once with ethanol before it was dried at 60 °C for 2 hours in an oven. The other 10 mL slurry was filtered by a syringe filter with a 0.2 μm cellulose membrane for Ca 2+ concentration determination.
Characterization. The chemical compositions of FGD gypsum were investigated by using X-ray fluorescence spectroscopy (XRF, Axios mAX, PANalytical B.V., Netherlands). The structures of the samples were determined by X-ray diffraction (XRD D8 Advanced, Bruker, Germany) using Cu Kα radiation (λ = 1.54178 Å), with a scanning rate of 5° min −1 and a scanning 2θ range of 5° to 70°. The morphology of the samples were characterized with the field-emission scanning electron microscopy (SEM, JSM-6490LV, JEOL, Japan) and the high resolution transmission electron microscopy (HRTEM, JEM-2100F, JEOL, Japan) equipped with the selected area electron diffraction (SAED). The surfaces of the whiskers were analyzed by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Fisher, USA) with a Al Kα photon energy of 1486.6 eV. C 1 s peak at 284.8 eV, which is related to the carbon adsorbed on the surface during the exposure of the samples to the ambient atmosphere, was used as reference for all spectra. The interaction between CTAB and the crystal surfaces was analyzed by Flourier Translation Infrared Spectroscopy (FT-IR, IRAffinity-1, Shimadzu, Japan) with a resolution of 4 cm −1 over the frequency range of 400-4000 cm −1 . For thermogravimetry and differential scanning calorimetry (TG-DSC, STA 8000, PerkinElmer, USA), 10 mg sample after dried was put into an Al 2 O 3 crucible with a lid under temperature programming from 60 °C to 400 °C at a heating rate of 10 °C/min under nitrogen gas atmosphere. The concentration of Ca 2+ in the filtrate was determined by the inductively coupled plasma-atomic emission spectrometry (ICP-AES, PS-6, Baird, USA). The average lengths and aspect ratios of the whiskers for each sample were counted up about 150 whiskers by Image-Pro plus 6.0 from SEM images with the magnifications of 500-4000.

Results and Discussion
Influence of KCl on the rate of phase transformation from FGD gypsum to α-HH. In the process of preparation of α-HH whiskers by using FGD gypsum under facile conditions, K + was used to accelerate the phase transformation rate 32 . As shown in Fig. 1, XRD patterns of the crystal products withdrawn at interval time were used to track the phase transformation process. The diffraction peaks of α-HH are at 2θ = 14.72°, 25.64°, 29.70°, 31.86° and FGD gypsum (DH) at 2θ = 11.61°, 20.69°, 23.35° and 29.08°.
Without KCl (Fig. 1a), the transformation from FGD gypsum to α-HH began at 360 min and completed within 420 min. When 1% KCl (by weight of FGD gypsum) (Fig. 1b) was added into the mixed solutions, the transformation started at 45 min and completed at 90 min. When KCl concentration was increased to 3% (Fig. 1c), the gypsum began to dehydrate at 15 min and the phase transformation completed within 1 h. Further increasing KCl concentration to 5% (Fig. 1d) and 7% (Fig. 1e), the time of complete transformation from FGD gypsum to α-HH decreased to 45 min. And as KCl concentration increased, α-HH generated by FGD gypsum would continue to dehydrate and be more easily transformed into anhydrous calcium sulfate (AH).
In order to further illustrate the difference of phase transformation rate with different KCl concentration (1%, 3% and 5% KCl), the crystal water content of samples withdrawn at certain time intervals was measured by TG-DSC technology, and the results are as follows. The water content of samples withdrawn at certain time intervals was cited to indicate whether FGD gypsum (20.93 wt%) has been completely transformed into α-HH (6.21 wt%). Figure 2 shows the water content change of samples prepared in glycerol (65%)-water mixtures with 1%, 3% and 5% KCl, respectively. As depicted in Fig. 2, in the presence of 1% KCl, the transformation started between 30 min and 45 min, and the crystal water content generally decreased over time and reached 6.21 wt% at 90 min, suggesting a complete transformation. When KCl concentration was increased to 3%, the FGD gypsum began to dehydrate between 0 and 15 min and the complete transformation needed 60 min. Further increasing KCl concentration to 5%, the water content of samples withdrawn at the same time intervals was lower than that of 3%, implying that the phase transformation rate with 5% KCl was faster. And the complete transformation only needed 45 min.
The promoting effect of KCl on the transformation rate should be attributed to the supersaturation (S HH ) of α-HH in the solution. A higher supersaturaion provided a larger driving force for α-HH nucleus formation, resulting in a shorter time of the phase transformation from FGD gypsum to α-HH. S HH is defined as:  where a, c, γ and K sp, HH are the activity, ion concentration, activity coefficient and thermodynamic equilibrium constant, respectively. In the research, the ratio of the supersaturations with K + ions to that without K + ions was used to illustrate the influence of K + ions on the supersaturations of α-HH:  where 0 represents the experimental data without K + ions. The concentration of Ca 2+ and SO 4 2− deriving from the dissolution of FGD gypsum are equal. Because the temperature was fixed at 90 °C and the concentration of KCl was much lower than that of glycerol, water activity and K sp, HH could be considered as constant.
The ionic activity coefficient was calculated by Debye-Hückel law 33 : where I, z i and a i are the ionic strength, ionic charge and ion size parameter, respectively. A, B and C are the Debye-Hückel constants. A and B are given by 34 : where ρ, ε r and T stand for the density, relative permittivity of the solution, and the thermodynamic temperature, respectively. Since ρ = 1.176 g/cm 3 and ε r = 58.283 in glycerol (65 wt%)-water solution at 90 °C, A = 0.6427 kg 0.5 ·mol −0.5 and B = 3.749E + 07 kg 0.5 ·mol −0.5 ·cm −1 32 . And the results are shown in Fig. 3.
As depicted in Fig. 3, as KCl concentration increased, Ca 2+ concentration and S HH /S 0 HH experienced an upward trend. And when KCl concentration increased from 0 to 7%, Ca 2+ concentration grew from 4.02 × 10 −3 to 1.69 × 10 −2 mol·kg −1 and S HH /S 0 HH from 1.00 to 17.67. It's clear that addition of KCl facilitated the dissolution of FGD gypsum and provided more available lattice ions, which created a much higher supersaturation and thus came into being a larger driving force for the phase transformation.  ( Fig. 4a), the crystal products with a diameter of 0.4-1.6 μm, an average length of 23.1 μm and aspect ratio of 28.9 were formed. When CTAB concentration was lower than 0.25%, the average lengths of the crystals increased and the average diameters decreased gradually with CTAB concentration increasing. Hence the aspect ratios  Continuing increasing the concentration of CTAB, the length of the crystal products rose unobviously and the diameter underwent a slight increase, which led to a slight decrease in the aspect ratio. The cause of this phenomenon might probably be the influence of high CTAB concentration on the solute diffusion. According to Song's research 25 , the diffusion coefficient of the solute decreases with the increase of CTAB concentration, indicating that the solute diffusion is prohibited by CTAB. Thus the mass of solute delivered from the solution to crystal surface decreases simultaneously, inhibiting the growth rate of calcium sulfate whisker and resulting in the decrease in the aspect ratio.

Influence of CTAB on the morphology of α-HH
To confirm the interactions between CTAB and the crystal surfaces of the α-HH crystals, the FT-IR spectra were explored as shown in Fig. 6. The peaks at 3612, 3558, and 1620 cm   XPS was used to characterize the surface adsorption of CTAB on the crystal products. Figure 7 shows the typical XPS wide-scan spectra of α-HH whiskers formed with different concentration of CTAB. The existence of N 2p spectra at 400 eV in line b, c and d indicated the presence of CTAB in the surface of the hydrothermal products. The XPS spectra of C1s core level from all four α-HH whisker samples are shown in Fig. 8. The C 1 s spectra of α-HH crystals prepared without CTAB as shown in Fig. 8a could be fitted with three components: The major component at 284.8 eV should be attributed to the majority of carbon contamination on the crystal surface 35 . The component located at 286.5 eV should be associated with C-O in the glycerol [35][36][37][38] , implying the adsorption of trace amount of glycerol on the crystal surfaces 18,39,40 . And the high binding energy component of C1s at 288.9 should be derived from the carbonate (CO 3 2− ) of the limestone existed in FGD gypsum 41,42 . The C 1 s spectra of α-HH whiskers formed with different CTAB concentration (Fig. 8b-d) could be fitted with four components. The C1s peaks at 284.8 eV should be attributed to the carbon atoms in the hydrocarbon chain of CTAB 36,43 , as well as the majority of carbon contamination on the crystal surface. And the peaks at 286.0 eV should have arisen from the carbon atoms bonded to nitrogen in CTAB [44][45][46] , which demonstrated the adsorption of the modifier on the crystal surface. Compared to the C-O peak of C1s XPS spectra from α-HH crystals prepared without CTAB, the C-O peaks of C1s XPS spectra from α-HH whiskers formed in the presence of CTAB shifted to the higher   binding energy, up to 286.8 eV, which might result from the interaction between glycerol and CTAB adsorbed on the crystal surfaces. As shown in Fig. 8, the intensity of the major component (284.8 eV) of C1s XPS spectra enhanced gradually with the increase of CTAB concentration, which also proved the adsorption of CTAB on the whisker surface. Table 2 shows the surface atomic distribution of carbon and nitrogen of α-HH whiskers formed in the presence of different CTAB concentration. From Table 2, we can see that as the increase of CTAB concentration, the surface atomic concentration of N and C generally enhanced, demonstrating adsorption of CTAB on the crystal surfaces.
TEM, high-resolution TEM (HRTEM) and SAED were performed to explore the morphology and structure of the crystal products further. As shown in Fig. 9, the interplanar spacing of the lattice fringes along the whiskers' growth direction was 0.595 nm, corresponding to the (002) plane (d(002) = 0.599 nm) of α-HH, which indicated that the crystals grow along the (001) direction. The electronic diffraction spots in the SAED pattern could be indexed to the [110] zone axis of α-HH, further confirming the preferential growth of the crystals along the (001) direction.
As shown in Fig. 10, α-HH lattices are composed of repeating, ionically bonded Ca and SO 4 atoms in chains of -Ca -SO 4 -Ca -SO 4 -. And SO 4 is a tetrahedral structure in which each S atom is covalently bonded to four O atoms 47,48 . The chains' structure may account for the fact that α-HH normally grows in the 1D shape. These chains are hexagonally and symmetrically arranged and form a framework parallel to the c axis. And along the c axis there exist continuous channels with a diameter of about 4.5 Å, where one water molecule is attached to every two calcium sulfate molecules 49 . The crystal structure makes the distribution of Ca 2+ denser on the top facets of {111} and the distribution of SO 4 2− denser on the side facets of {110} and {100}, which makes {111} facets positively charged and {110} and {100} facets negatively charged 47,49 . Therefore, the positive C 16 H 33 (CH 3 ) 3 N + is adsorbed more easily on the side facets of α-HH crystal, which promotes the 1-D growth of α-HH whiskers along the c axis, as shown in Fig. 11.

Conclusion
α-HH whiskers with high aspect ratios could be prepared by using FGD gypsum in glycerol-water solutions with CTAB and KCl under facile conditions. KCl played a key role in speeding up the transformation rate while CTAB acted as a crystal modifier to promote 1-D growth of α-HH along the c axis. The presence of 5% KCl could result in a significant increase in the supersaturation of α-HH, which dramatically reduced the time of complete phase transformation from 7 h to 45 min. And as KCl concentration increased, α-HH generated by FGD gypsum would continue to dehydrate and be more easily transformed into anhydrous calcium sulfate (AH). The existence of 0.25% CTAB in the crystallization reaction system could lead to a significant increase of the average aspect ratio from 28.9 to 188.4, which might probably be attributed to preferential adsorptioin of positive C 16 H 33 (CH 3 ) 3 N + on the side facets of α-HH crystal and promote the 1-D growth of α-HH whiskers along the c axis. And continuing increasing CTAB concentration, the aspect ratios suffered a slight decrease.