Retardation behavior of hydration of calcium sulfate hemihydrate (bassanite) induced by sodium trimetaphosphate (STMP)

Wei Chen a, Yi-nan Wu a, Bingru Zhang a, Ying Wang a, Fengting Li *a and Zeming Qi b
aState Key Lab of Pollution Control and Resource Reuse Study, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, P.R. China. E-mail:
bNational Synchrotron Radiation Laboratory and College of Nuclear Science and Technology, University of Science and Technology of China, Hefei, Anhui, China

Received 12th December 2017 , Accepted 2nd February 2018

First published on 2nd February 2018

The sub-micro crystals of a bassanite intermediate were isolated through the introduction of sodium trimetaphosphate (STMP) during the hydration process of bulk bassanite. The particle size and morphology of precipitates were first checked by scanning electron microscopy. The particle size was found to be significantly reduced with increments of STMP, and the majority of the precipitates at a >80 mM STMP addition level had sub-micro size. The length, height, and width of the crystals were determined to be ∼1 μm, ∼150 nm, and ∼300 nm, respectively, by atomic force microscopy. Surface structural information at the molecular level was then obtained by diffuse reflectance infrared Fourier transform spectroscopy. Afterwards, the crystal structure evolution with different amounts of STMP was determined by wide angle X-ray diffraction. Both analytical results show that the final precipitated sub-micro rods (100 mM STMP) are almost all bassanite rather than gypsum. Moreover, the multipeak deconvolution of the WAXD patterns presents a quantitative analysis of the weight percent of bassanite in the final precipitates. The final concentration of bassanite is close to 100% (100 mM STMP). Based on these results, the hydration process of bulk bassanite can be described as the dissolution of original bulk bassanite, the formation of bassanite sub-micro rods, and finally the self-assembly of the sub-micro rods along the c-axis to form gypsum micro-crystals without the additive.

1. Introduction

Gypsum (CaSO4·2H2O) has been an indispensable building material even before recorded human history.1 It was used in building the Egyptian pyramids ∼5000 years ago, and even today, it is still widely used in various areas, such as construction and interior decoration.2 Its light weight and high strength ensure its great potential in construction areas. After calcination within 100–150 °C under atmospheric pressure,3,4 some crystal water molecules would be removed, and calcium sulfate hemihydrate (CaSO4·0.5H2O, bassanite, stucco or plaster of Paris) is obtained. As the most widely produced inorganic material,5 bassanite is the key raw material for preparing gypsum-related products through hydration with water, thereby attracting considerable attention for understanding fundamental phase transition.6–8 Previously, the formation of gypsum from bassanite could be described as a classical dissolution–reprecipitation process: the dissolution of original bassanite, nucleation of gypsum, and growth of gypsum nuclei.8,9

In addition to bassanite, numerous additives, such as carboxylic acids,8 phosphoric acid derivatives, and polymers,10 are widely used to control certain properties of gypsum products, i.e., accelerating or retarding the hydration rate and increasing the anti-sag performance.11,12 Therefore, understanding the influence of additives on the dissolution and precipitation mechanism of gypsum can help us optimize the macroscopic performance of calcium sulfate-related materials. Early studies showed that additives, i.e., sodium poly(acrylate), do not influence the dissolution of the hemihydrate but block the nucleation of gypsum through adsorption on bassanite surfaces.10 Techniques, such as Raman spectroscopy, have been used to monitor the kinetics of plaster hydration and the effect of a retarder.13 Moreover, the autocatalytic growth of gypsum needles during the first stage of the reaction is confirmed.

In a laboratory, however, free ions Ca2+/SO42− in supersaturated aqueous solution are preferred to investigate gypsum crystallization rather than bulk bassanite.14–20 Therefore, the majority of literature reports focus on the precipitation of gypsum, which is actually the second stage of the hydration process of bulk bassanite. Different from classical nucleation and growth theory in gypsum crystallization, recent work shows that the precipitation of Ca2+/SO42− in aqueous solution does not directly form gypsum but a metastable bassanite nanorod intermediate.21 These bassanite nanorods require further self-assembly along the c-axis and finally transform into gypsum crystals. This process is also described as crystallization by particle attachment (CPA), where the particle is not a monomeric species as described by classical nucleation and growth theory but a higher-order species with size ranging from multiple ions to nano-crystals.22–25 Meanwhile, Wang and Meldrum re-investigated the influence of additives on the precipitation of Ca2+/SO42− in aqueous solution.9 The additives were found to stabilize the bassanite nanorod intermediate. Despite numerous studies on the precipitation mechanism of Ca2+/SO42− in aqueous solution, most investigations on the hydration process of bassanite follow the classical dissolution–reprecipitation mechanism as mentioned above. Based on the non-classical gypsum crystallization model,21,26 after the dissolution of original bulk bassanite into free ions (Ca2+/SO42−), metastable bassanite would first form and then transform into stable gypsum crystals. The difficulty in confirming this hypothesis is the isolation of the bassanite nano-rod intermediate. The difference from previous systems was that the initial species were free ions (Ca2+/SO42−) in the supersaturated state, and no bassanite was observed in the original solution. For bulk bassanite as the original reaction material, isolating such a bassanite intermediate after the dissolution of original bulk bassanite is almost impossible because the whole system is actually a mixture of bulk bassanite and newly formed ones. Techniques, such as TR-cryo-TEM, have been used to observe the hydration process of bassanite.27 However, the in situ observation conducted with TR-cryo-TEM measurement requires the filtration of the undissolved bassanite, wherein only the supersaturated solution remains. Also, this is somewhat far away from the hydration of bulk bassanite.

The main objective of this work is to check the existence of the hypothesized metastable hemihydrate as discussed above. Inspired by the stabilization performance of additives in the precipitation of free ions,9 we found one additive, sodium trimetaphosphate (STMP, the chemical structure is shown in Fig. 1), that can significantly stabilize the bassanite intermediate for up to months. This is, to the best of our knowledge, the first report to validate the existence of an intermediate phase during the bassanite hydration process through the introduction of an appropriate additive. Meanwhile, the dissolution of the hemihydrate is indirectly confirmed due to the destruction of the original bulk bassanite and the formation of final sub-micro bassanite rods. Admittedly, such a sub-micro bassanite intermediate is obtained in the presence of STMP, but the direct confirmation of the existence of this phase in pure water still requires further investigation. The detailed crystallization mechanism of bassanite hydration is summarized in the Results & discussion section.

image file: c7ce02144e-f1.tif
Fig. 1 Concentration of P (blue line, right) and Ca2+ (black line, left) in the supernatant as characterized by ICP. The calculated P concentrations c(P) are shown as red triangles.

2. Experimental apparatus

2.1 Chemical reagents

Sodium trimetaphosphate (STMP) and β-bassanite were obtained from Sigma-Aldrich. All chemical reagents were used as received without further purification. Ultra-pure water (18.2 mΩ cm) was used throughout the experiments.

2.2 Hydration of bassanite

The hydration of bassanite was conducted at 20 °C with a bulk bassanite concentration of 50 g L−1. STMP at six different addition levels, ranging from 0 mM to 100 mM with an interval of 20 mM (relative to water) was used. STMP was first mixed with water until complete dissolution before the mixture was further mixed with bulk bassanite. The mixture was kept at ambient temperature for three days to ensure complete hydration of bulk bassanite. Afterwards, the precipitate was filtered and washed with ultra-pure water thrice to remove free-state STMP. The samples were then dried at 40 °C for 24 h before any characterization.

2.3 Characterization techniques

2.3.1 Inductively coupled plasma (ICP) measurements. The concentrations of Ca2+ and P (originating from STMP) were characterized using an inductively coupled plasma (ICP) emission spectrometer (720ES, Agilent, USA). Limited by the sensitivity of the ICP (cmax < 100 ppm), the initial supernatant was diluted 100 times before ICP measurements.
2.3.2 SEM characterization. SEM was conducted using a Philips XL-30 scanning electron microscope at an accelerating voltage of 20 kV. High-resolution SEM photos were captured with an Ultra 55 (Zeiss, Germany) field-emission scanning electron microscope. All samples were coated with gold before SEM characterization.
2.3.3 AFM measurements. All AFM experiments were carried out in air using a Veeco Dimension D3100 in tapping mode (TM) under ambient conditions. The precipitate was placed on mica directly.
2.3.4 Diffuse reflectance infrared Fourier transform (DRIFT). Diffuse reflectance infrared Fourier transform (DRIFT) experiments were performed using a Bruker IFS 66v FTIR spectrometer on an infrared beamline (BL01B) at the National Synchrotron Radiation Laboratory, China. A diffuse reflection accessory (Praying Mantis, Harrick) was used. Each scan was collected from 4000 cm−1 to 400 cm−1 with 128 scans at a resolution of 2 cm−1. OPUS IR analysis software was used to analyze the final spectra. Pure KBr powder was first used to obtain the background IR signal. Then all samples were added directly without dilution by KBr.
2.3.5 Wide angle X-ray diffraction (WAXD). Wide angle X-ray diffraction (WAXD) measurements were performed on a Bruker D8 Advance X-ray powder diffractometer (40 kV, 40 mA, CuKα1 radiation of λ = 1.54 Å) with 2θ ranging from 5–80° with a scanning rate of 0.02° s−1 at ambient temperature.

3. Results & discussion

3.1 Concentration of Ca2+ and P in the supernatant

STMP can serve as an anti-scale reagent to inhibit the precipitation of gypsum, which is unwanted in areas, such as oil/gas wells28 and desalination plants. In order to check the influence of STMP on the hydration process of bulk bassanite, the concentration of Ca2+ (c(Ca2+)) in the supernatant after filtration was characterized. Fig. 1 presents c(Ca2+) as characterized by ICP together with that of phosphate P (c(P)). By increasing the STMP addition level from 0 mM up to 100 mM, c(Ca2+) increases from 0.15 wt% to 0.31 wt%. c(Ca2+) is almost doubled with the introduction of STMP. This observation suggests that STMP can significantly increase the solubility (or concentration) of Ca2+ in the presence of SO42−, and on the other hand confirms the anti-scale ability of STMP in a gypsum scale system. When the STMP addition level increases up to 60 mM, c(Ca2+) almost reaches a plateau value (30 ppm), and further increments of STMP do not further increase c(Ca2+). Meanwhile, the phosphate P concentration c(P) increases almost linearly with the increment of STMP. The calculated c(P) of pure STMP in aqueous solution is also shown in Fig. 1 marked by red triangles. A minor difference exists between the c(P) of the supernatant and the calculated one of pure STMP. This finding indicates that the majority of STMP remains in aqueous solution rather than co-precipitating with gypsum crystals during the bassanite hydration process.

3.2 Crystal morphology of precipitates with different amounts of STMP

In addition to the investigation of the interaction between STMP and Ca2+ in the supernatant, the influence of STMP on the precipitated crystals was investigated by SEM. Fig. 2 presents the SEM photos of the initial bulk bassanite and the precipitates with different amounts of STMP. The original precipitate without STMP mainly displays a needle-like structure (with some plates and prisms), which is similar to that of gypsum crystals prepared through the Ca2+/SO42− supersaturated solution.3,15,29,30 With the increment of STMP, the original needle-like structure becomes irregular, and the coexistence of giant crystals and small particles is observed in the presence of 40 mM STMP (Fig. 2(d)). When the addition level of STMP reaches 80 mM, small crystals dominate the precipitation, and the particle size is significantly reduced as compared with the original needle-like crystals (Fig. 2(e–g)). The particle size of these small precipitates is further characterized by high-resolution field emission SEM as shown in Fig. 3.
image file: c7ce02144e-f2.tif
Fig. 2 SEM images of (a) original bassanite and gypsum (or bassanite) crystals with (b) 0 mM, (c) 20 mM, (d) 40 mM, (e) 60 mM, (f) 80 mM and (g) 100 mM STMP. The scale bar represents 20 μm.

image file: c7ce02144e-f3.tif
Fig. 3 (a and b) Field emission SEM photos of precipitates obtained with 100 mM STMP. (c) AFM height image of precipitates (100 mM STMP) together with the (d) cross-section as marked by the white dashed line. (e and f) TEM images of the precipitates and (g) the selected area electron diffraction (SAED) pattern.

Fig. 3(a and b) show the high-resolution SEM photos of the precipitated small particles (100 mM STMP). Small needle-like structures are observed with a length of ∼1 μm. The height and width of these particles are determined to be ∼150 and ∼300 nm, respectively, as characterized by AFM (Fig. 3(d)). The size of these sub-micron rods is considerably smaller than those of either the initial bassanite (Fig. 2(a)) or needle-like gypsum crystals (Fig. 2(b)). Such morphology is quite different from previously reported plate-like crystals after the hydration of bassanite with maleic acid and citric acid as additives.8 The crystal size is close to that of the aggregated bassanite nano-rod intermediate,21 and the needle-like structure represents crystal self-assembly and growth along the c-axis. The orientation of the crystals was further confirmed by TEM and SAED as shown in Fig. 3(e–g). The phase structure must be checked to rationalize the hydration process of bulk bassanite in the presence of STMP.

3.3 Characterization of functional groups of the precipitates through DRIFT

DRIFT was first applied to determine the functional groups. DRIFT, compared with conventional bench-top FTIR, mainly reflects surface IR signals.31–33 During DRIFT measurements, all powder precipitates were not ground and not mixed with KBr to obtain the surface IR spectra of the original crystal surface. As shown in Fig. 4, two dominant IR absorption regions, 3700–3000 and 1700–1500 cm−1, are obtained, which are assigned to the stretching mode of H2O (ν(H2O)) and the bending mode of H2O (δ(H2O)), respectively.7,34 For gypsum, δ(H2O) displays doublet absorption lines, and ν(H2O) exhibits a broad one. The doublet lines of gypsum are assigned to the Au (1686 cm−1) and Bu (1623 cm−1) modes of the v2 water bending mode, while only one v2 bending mode exists in bassanite. With the increment of STMP in the whole system, for δ(H2O), the left-side peak almost disappears with 100 mM STMP, which is similar to that for pure bassanite. For ν(H2O), the broad peak starts to become narrow when the addition level of STMP reaches 60 mM. Further increasing STMP leads to a narrow ν(H2O) absorption line and finally becomes similar to that for pure bassanite. The DRIFT spectra qualitatively show that the functional groups of the final precipitates with STMP are similar to those of pure bassanite. The direct and quantitative analysis of crystal structure evolution would be shown later through WAXD.
image file: c7ce02144e-f4.tif
Fig. 4 DRIFT spectra of gypsum precipitates with different amounts of STMP.

3.4 Phase determination through WAXD

The detailed phase structure was determined by WAXD as shown in Fig. 5(a). For pure gypsum (0 mM STMP), four dominant WAXD peaks are obtained and assigned to (020)g (11.97°), (021)g (21.06°), (040)g (23.75°), and (041)g (29.46°), matching well with JCPDS data (#33-0311).29,35,36 For pure bassanite, four distinct peaks are assigned to (200)b (14.64°), (020)b (25.33°), (400)b (29.68°), and (204)b (31.75°) based on JCPDS data (#41-0224).37,38 Diffraction peak (200)b starts to appear at 20 mM STMP, whereas the intensity of (020)g decreases as compared with pure gypsum. Increasing the addition level of STMP changes the crystal structure of the final precipitates from pure gypsum to a dominant bassanite structure. Based on the DRIFT and WAXD results, bassanite is predominantly formed in aqueous solution with 100 mM STMP. In order to obtain a quantitative result of the bassanite portion in the final precipitates, a multipeak deconvolution method was used as shown in Fig. 5(b). The fitting region is fixed within 2θ = 10–35° with eight WAXD diffraction peaks, and each of them is assigned to gypsum and bassanite as discussed above. After fitting, the red peaks represent gypsum, and the blue ones belong to bassanite. The concentration of bassanite in the final precipitates as a function of the STMP addition level is summarized in Fig. 5(c). The introduction of STMP increases the weight percent of bassanite in the final crystal, and under 100 mM STMP conditions, the precipitation is dominated by bassanite. In addition to DRIFT results, this result suggests that the bassanite intermediate is stabilized by STMP in aqueous solution.
image file: c7ce02144e-f5.tif
Fig. 5 (a) WAXD pattern of the precipitates with different concentrations of STMP; (b) multipeak deconvolution of the WAXD pattern (60 mM STMP) within 2θ = 10–35 °; (c) the bassanite content in the final precipitates as a function of the STMP addition level.

3.5 Stability of sub-micro bassanite induced by STMP

A chemical agent, such as polyacrylic acid, is found to stabilize bassanite in aqueous solution, which is attributed to the retardation of gypsum crystallization.9 However, the majority of the metastable bassanite nanorods would finally transform into gypsum within days. In order to confirm the influence of STMP on the stabilization of metastable bassanite, crystallization experiments with different crystallization times ranging from 1 d up to 36 d were conducted. As shown in Fig. 6, even up to 36 d, the bassanite crystal structure is predominantly formed in the final precipitates. This finding suggests that STMP can stabilize the sub-micro bassanite intermediate for up to months based on the current experimental results, which is considerably longer than previously reported results in the literature.
image file: c7ce02144e-f6.tif
Fig. 6 WAXD patterns of the precipitates obtained with 100 mM STMP after different times ranging from 1 to 36 days.

Meanwhile, comparison of the WAXD data of the sub-micro bassanite obtained in the current experiment with bulk bassanite shows that the relative intensity of (020)b is considerably higher than that of (200)b in sub-micro bassanite, while the intensities of these two peaks are comparable in bulk bassanite. Such a difference may originate from the varying growth rates of different crystal planes.39 (020)b stands out as the primary diffraction peak, suggesting that the growth rate of the (020)b plane is the fastest as compared with other planes. The origin of such a phenomenon may be attributed to the weak binding of STMP to the (020)b plane. Moreover, the original bulk bassanite has granular micro-crystals (Fig. 2(a)), and the final sub-micro bassanite shows a needle-like crystal morphology as discussed above. The relative intensity ratio of different WAXD diffraction peaks can reflect detailed crystal morphology information, such as average longitude and aspect ratio. Further investigation to correlate WAXD diffraction peaks and crystal morphology is required.

3.6 Initial hydration process of bassanite in the presence of STMP

Based on the crystal morphology change characterized by SEM and AFM and the interior crystal structure evolution as determined by DRIFT and WAXD, the sub-micro bassanite rod intermediate was clearly successfully prepared through the introduction of STMP. The initial bulk bassanite, before mixing with water, displays giant micro-particles with lengths of up to 50 μm, whereas the final bassanite crystal has a length of ∼1 μm as shown in Fig. 7. Given that the dissolution of the original bulk bassanite is not largely influenced by additives,10 the introduction of STMP has a major contribution to the precipitation process. Such a result, on the other hand, confirms the hypothesis as mentioned in the Introduction section: during hydration, bulk bassanite in aqueous solution undergoes dissolution to free ions (Ca2+/SO42−), then precipitates to the metastable bassanite rod intermediate, and finally forms stable gypsum (Fig. 6) if no additive is added. This process is also in agreement with the recently proposed CPA.22 The sub-micro rod intermediate found here should be the intermediate precursor before the formation of final gypsum. Admittedly, the current work only detects or isolates the newly-found sub-micro crystal intermediate rather than further hydration processes; thus, more direct evidence confirming this intermediate phase and the later hydration process is still required.
image file: c7ce02144e-f7.tif
Fig. 7 Schematic of the hydration process of bassanite in aqueous solution in the presence of STMP.

The precipitated bassanite rod can be stabilized in water for up to 36 days in the current experiment. The particle size and polymorph are found to significantly influence the stability of materials, such as alumina40 and titania,41 whose stability is dominated by enthalpy.21,42 Although the solubility of bulk bassanite is higher than that of gypsum, which is below 100 °C, the bassanite nanorod or nanoparticle intermediate is stable in supersaturated solution, suggesting higher stability as compared with bulk bassanite and gypsum. However, without STMP, pure gypsum can be obtained. This finding indicates that the ultra-high stability of bassanite sub-micro rods is not completely dominated by the particle size. STMP should also contribute to the stability. As shown by the ICP (Fig. 1) results, the introduction of STMP can significantly increase the solubility of bulk bassanite in aqueous solution, indicating a strong interaction between STMP and Ca2+. Therefore, STMP would strongly interact with Ca2+ on the surface of bassanite as shown in Fig. 7. The binding of STMP to the bassanite sub-micro rod surface further reduces the surface energy, thereby increasing the stability of the precipitated crystals. Meanwhile, since STMP is a salt, once (PO3)33− binds to the surface of a bassanite rod, the charged surface would further inhibit the self-assembly of these rods along the c-axis to form a gypsum crystal due to repulsion forces.21 Overall, both the constrained sub-micro particle size and charged surface induced by the (PO3)33− group of STMP increase the stability of bassanite sub-micro rods in aqueous solution.


Overall, we have successfully isolated a metastable bassanite micro-rod intermediate through the addition of STMP during the hydration process of bulk bassanite. The particle size of the precipitate was found to be significantly reduced with an increment of STMP, and the majority of the precipitates exhibit sub-micro size when the STMP addition level reaches 80 mM. The sub-micro crystal length, height, and width were determined to be ∼1 μm, ∼150 nm, and ∼300 nm, respectively, as characterized by AFM. The precipitates were further investigated by DRIFT and WAXD to obtain phase structure information. Both results show that the final precipitated sub-micro rods are actually metastable bassanite rather than stable gypsum. Quantitative analysis based on WAXD results shows that the final concentration of bassanite in the precipitated crystals reaches close to 100%. Based on the state-of-the-art gypsum crystallization mechanism,21–23 where bassanite nano-rods are determined to be the precursors before gypsum formation, the hydration process of bulk bassanite can be described as the dissolution of original bulk bassanite, formation of bassanite sub-micro rods, and finally self-assembly along the c-axis to form gypsum micro-crystals if no additive is added. As we mentioned in the Introduction section, the current results rely on the hydration of bassanite in the presence of STMP, and more direct evidence to confirm the above hypothesis is still required. Since bulk bassanite is widely used on an industrial scale (∼1011 kg y−1 of bassanite are produced for construction purposes),7,21 our findings clarify the detailed hydration process of bulk bassanite in aqueous solution and especially the intermediate phase transition process.

Conflicts of interest

There are no conflicts to declare.


This work was supported by the National Natural Science Foundation of China (21577100 & 21777119), the Science and Technology Commission of Shanghai Municipality (17230711600), the China Postdoctoral Science Foundation (2017M621531) and the Foundation of State Key Laboratory of Pollution Control and Resource Reuse (Tongji University) (PCRRY15007).

Notes and references

  1. W. A. Felsing and A. D. Potter, J. Chem. Educ., 1930, 7, 2788–2807 CrossRef CAS .
  2. A. E. Charola, J. Pühringer and M. Steiger, Environ. Geol., 2007, 52, 207–220 CrossRef .
  3. D. Freyer and W. Voigt, Monatsh. Chem., 2003, 134, 693–719 CrossRef CAS .
  4. V. Tydlitát, I. Medveď and R. Černý, J. Therm. Anal. Calorim., 2012, 109, 57–62 CrossRef .
  5. H. Weiss and M. F. Braeu, Angew. Chem., Int. Ed., 2009, 48, 3520–3524 CrossRef CAS PubMed .
  6. N. B. Singh, C. Vellmer and B. Middendorf, Indian J. Eng. Mater. Sci., 2005, 12, 337–344 CAS .
  7. N. B. Singh and B. Middendorf, Prog. Cryst. Growth Charact. Mater., 2007, 53, 57–77 CrossRef CAS .
  8. E. Badens, S. Veesler and R. Boistelle, J. Cryst. Growth, 1999, 198–199, 704–709 CrossRef .
  9. Y.-W. Wang and F. C. Meldrum, J. Mater. Chem., 2012, 22, 22055–22062 RSC .
  10. J. P. Boisvert, M. Domenech, A. Foissy, J. Persello and J. C. Mutin, J. Cryst. Growth, 2000, 220, 579–591 CrossRef CAS .
  11. Q. Yu, S. W. Sucech, B. E. Groza, R. J. Mlinac, F. T. Jones and P. J. Henkels, U.S. Pat., 6342284B1, 2002, vol. 190, pp. 181–190 Search PubMed .
  12. R. Arese, D. Martin and M. Rigaudon, U.S. Pat., US10537397, 2006, p. 1 Search PubMed .
  13. S. Joiret, F. Pillier and A. Lemarchand, J. Phys. Chem. C, 2014, 118, 28730–28738 CAS .
  14. Y. J. Wu, T. W. T. Tsai and J. C. C. Chan, Cryst. Growth Des., 2012, 12, 547–549 CAS .
  15. H. Fu, B. Guan, G. Jiang, M. Z. Yates and Z. Wu, Cryst. Growth Des., 2012, 12, 1388–1394 CAS .
  16. E. Ruiz-agudo and M. Ortega-huertas, Cryst. Growth Des., 2016, 16, 4950–4959 CAS .
  17. S. Karthika, T. K. Radhakrishnan and P. Kalaichelvi, Cryst. Growth Des., 2016, 16, 6663–6681 CAS .
  18. F. K. Crundwell, Cryst. Growth Des., 2016, 16, 5877–5886 CAS .
  19. K. Gupta, S. Singh and M. S. R. Rao, Cryst. Growth Des., 2016, 16, 3256–3261 CAS .
  20. U. Tritschler, M. Kellermeier, C. Debus, A. Kempter and H. Cölfen, CrystEngComm, 2015, 17, 3772–3776 RSC .
  21. A. E. S. Van Driessche, L. G. Benning, J. D. Rodriguez-Blanco, M. Ossorio, P. Bots and J. M. Garcia-Ruiz, Science, 2012, 336, 69–72 CrossRef CAS PubMed .
  22. J. J. De Yoreo, P. U. P. A. Gilbert, N. A. J. M. Sommerdijk, R. L. Penn, S. Whitelam, D. Joester, H. Zhang, J. D. Rimer, A. Navrotsky, J. F. Banfield, A. F. Wallace, F. M. Michel, F. C. Meldrum, H. Cölfen and P. M. Dove, Science, 2015, 349, aaa6760 CrossRef PubMed .
  23. J. J. De Yoreo and P. G. Vekilov, Rev. Mineral. Geochem., 2003, 54, 57–93 CrossRef CAS .
  24. L. Wang, X. Guan, R. Tang, J. R. Hoyer, A. Wierzbicki, J. J. De Yoreo and G. H. Nancollas, J. Phys. Chem. B, 2008, 112, 9151–9157 CrossRef CAS PubMed .
  25. R. Stach, P. Krebs, F. Jones and B. Mizaikoff, CrystEngComm, 2017, 19, 14–17 RSC .
  26. Y.-W. Wang, Y.-Y. Kim, H. K. Christenson and F. C. Meldrum, Chem. Commun., 2012, 48, 504–506 RSC .
  27. A. Saha, J. Lee, S. M. Pancera, M. F. Braeu, A. Kempter, A. Tripathi and A. Bose, Langmuir, 2012, 28, 11182–11187 CrossRef CAS PubMed .
  28. G. Marvin, U.S. Pat., 2777818, 1957 Search PubMed.
  29. W. M. M. Heijnen and P. Hartman, J. Cryst. Growth, 1991, 108, 290–300 CrossRef CAS .
  30. D. Bosbach and M. F. H. Jr, Chem. Geol., 1996, 132, 227–236 CrossRef CAS .
  31. A. Z. Graciano-Verdugo, E. Peralta, H. González-Ríos and H. Soto-Valdez, Anal. Chim. Acta, 2006, 557, 367–372 CrossRef CAS .
  32. R. Gnanasambandam and A. Proctor, Food Chem., 2000, 68, 327–332 CrossRef CAS .
  33. G. C. Shearer, S. Forselv, S. Chavan, S. Bordiga, K. Mathisen, M. Bjorgen, S. Svelle and K. P. Lillerud, Top. Catal., 2013, 56, 770–782 CrossRef CAS .
  34. P. S. R. Prasad, V. K. Chaitanya, K. S. Prasad and D. N. Rao, Am. Mineral., 2005, 90, 672–678 CrossRef CAS .
  35. W. F. Cole and C. J. Lancucki, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1974, 30, 921–929 CrossRef .
  36. M. Atoji and R. E. Rundle, J. Chem. Phys., 1958, 29, 1306–1311 CrossRef CAS .
  37. H. F. Follner, S. A. Wolter, A. Preusser, S. Indris and C. Silber, Cryst. Res. Technol., 2002, 37, 1075–1087 CrossRef .
  38. P. Ballirano, A. Maras, S. Meloni and R. Caminiti, Eur. J. Mineral., 2001, 13, 985–993 CrossRef CAS .
  39. Á. O. Caso, J. Cryst. Growth, 2013, 380, 169–175 CrossRef .
  40. J. M. McHale, A. Auroux, A. J. Perrotta, A. Navrotsky, S. S. Flaschen, R. C. Garvie, G. P. Johnston, R. Muenchausen, D. M. Smith, W. Fahrenholtz, S. J. Foltyn, S. Blonski, S. H. Garofalini, W. C. Mackrodt, R. J. Davey, S. N. Black, I. Manassidis, M. J. Gillan, S. Schneider, S. Ansell, G. F. Cerofolini, R.-S. Zhou, R. L. Snyder, D. J. Coster, J. J. Fripiat, M. Muscas, A. Auroux, A. Gervasini, A. Auroux, A. Bailey, B. D. Begg, E. R. Vance, J. Nowotny, H. F. Holmes, E. L. Fuller and R. B. Gammage, Science, 1997, 277, 788–791 CrossRef CAS .
  41. M. R. Ranade, A. Navrotsky, H. Z. Zhang, J. F. Banfield, S. H. Elder, A. Zaban, P. H. Borse, S. K. Kulkarni, G. S. Doran and H. J. Whitfield, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 6476–6481 CrossRef CAS PubMed .
  42. A. Navrotsky, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 12096–12101 CrossRef CAS PubMed .

This journal is © The Royal Society of Chemistry 2018