3.1 Basic glass
The XRD patterns of the three groups of basic glasses are shown in Fig. 2(a). It can be seen from the figure that the sintered basic glass powder is mainly amorphous. By comparison, it is found that the basic glass prepared by scheme 1 has a small amount of crystallization. The appearance of crystallization makes it easy for the basic glass to generate internal stress in the subsequent heat treatment process, which adversely affects the properties of foamed glass-ceramics. Therefore, considering the stability of the basic glass comprehensively, it is more appropriate to choose schemes 2 and 3. The DTA curve of the basic powder shown in Fig. 2(b) shows that the nucleation temperature of the basic glass powder is about 730 ℃, and the crystallization temperature is about 1000 ℃.
According to the trial sintering experiment and the differential thermal analysis, the following nine groups of orthogonal experiments (Table 3) are determined to study the factors such as the amount of foaming agent and the heat treatment system.
Table 3
Results of the orthogonal experiment
Factor Group | Scheme | CaCO3 content (wt.%) | Crystallization temperature (°C) | Holding time (h) | Bending strength (Mpa) |
1 | A | 10% | 1000 | 1 | 78.172 |
2 | A | 20% | 1050 | 2 | 35.555 |
3 | A | 30% | 1100 | 3 | 13.815 |
4 | B | 10% | 1050 | 3 | 81.460 |
5 | B | 20% | 1100 | 1 | 42.137 |
6 | B | 30% | 1000 | 2 | 15.233 |
7 | C | 10% | 1100 | 2 | 76.195 |
8 | C | 20% | 1000 | 3 | 43.630 |
9 | C | 30% | 1050 | 1 | 11.037 |
k1 | 42.514 | 78.609 | 45.675 | 43.782 | / |
k2 | 46.273 | 40.44 | 42.684 | 42.324 | / |
k3 | 43.62 | 13.358 | 44.049 | 46.3 | / |
R | 3.759 | 65.251 | 2.911 | 3.976 | / |
Factor priority | CaCO3 content > Holding time > Scheme > Crystallization temperature |
3.2 Foamed glass-ceramics
3.2.1 The content of foaming agent
The degree of dispersion of bubbles in foamed glass-ceramics depends to a certain extent on the uniformity of the dispersion of the foaming agent in the basic glass powder [20]. Furthermore, the amount of foaming agent incorporated will greatly affect the distribution of bubbles. When the amount of foaming agent is too high, the number of microbubbles increases, which leads to the convergence of a large number of rising bubbles, and makes the foaming distributed in layers. The crystallization temperature of the basic glass powder also restricts the type and amount of foaming agent. The transformation of glass powder from solid to molten is an endothermic process. The DTA curve shows that the nucleation temperature and crystallization temperature are about 730 ℃ and 1000 ℃. According to comprehensive analysis, CaCO3 is selected as the foaming agent. This is because the foaming temperature of CaCO3 is compatible with the nucleation temperature and crystallization temperature of the basic glass powder. According to the research on the trial sintering experiment, the content of CaCO3 is selected as 10%, 20%, and 30 wt.%, respectively.
3.2.2 Crystallization temperature
The proper foaming temperature can control the size of the bubbles and the uniformity of the cell distribution. The foaming temperature of foamed glass-ceramics should be equivalent to the nucleation temperature of the melt, so the foaming temperature should be 730 ℃. The growth of crystal grains is mainly completed in the crystallization stage, and the crystallization temperature is too high or too low, which is not conducive to the precipitation of crystals. The crystallization temperature is preferably 1000 ℃, 1050 ℃, and 1100 ℃, respectively.
3.2.3 Holding time
The holding time mainly affects the pore size and the dispersion degree of the bubbles. Too long holding time will cause small bubbles to rise. The foaming agent has just begun to decompose into small bubbles, and then it has entered the cooling stage, which makes the foaming process end prematurely, resulting in large bulk density and poor mechanical properties of products [21]. The crystallization time will directly affect the size of the crystal grains, thereby affecting the properties of the foamed glass-ceramics. The research shows [22] that in order to obtain enough grain length, the holding time should not be less than 60 min. In summary, in this study, the foaming holding time is 60 min, and the crystallization holding time is 60 min, 120 min, and 180 min, respectively.
3.3 Orthogonal experiment analysis
Nine groups of orthogonal foamed glass-ceramics are used to test their bending strength, fracture toughness, and elastic modulus, to explore the effect of foaming agent content and heat treatment system on the mechanical properties of foamed glass-ceramics. The trends of the three mechanical properties are shown in Fig. 3 (a-c). According to the results shown in the trend picture, it can be seen that the bending strength, fracture toughness, and elastic modulus of foamed glass-ceramics show a decreasing trend with the increase of calcium carbonate content. This is because with the increase of foaming agent content, the number of large pores and connected pores will increase, resulting in uneven foaming, and the mechanical properties of the samples will become worse. Moreover, the effects of crystallization temperature and holding time in the heat treatment system on bending strength and fracture toughness show a trend of decreasing at first and then increasing. In general, the bending strength of the glass-ceramic sample will increase with the increase of its crystallinity within a certain range. This is because according to the microcrack strength theory proposed by Griffith [23], the critical strength of crystals can be expressed as a large number of growth in the glass body, which leads to a large number of uneven surfaces on the fault section, resulting in higher fracture surface energy. It can be seen from the following formula that the increase of fracture surface energy will improve the strength of the material.
σ=(E·γ/πC)1/2
Where E is the elastic modulus (GPa), γ is the fracture surface energy, (J·m− 2); C is the critical length of microcracks (mm).
According to the mechanical properties and the range analysis of nine groups of the orthogonal experiments, comprehensive considerations, the second group of basic glass is selected as the optimal scheme, that is, the calcium carbonate content is 10 wt.%, and the crystallization temperature is 1000 ℃. The holding time is 3 h.
3.4 Phase analysis
Figure 3d is the XRD pattern of the orthogonal experiment group. The results shown in the pattern show that the crystal type of the nine groups of samples is a single diopside phase (CaMgSi2O6). With the increase of the crystallization temperature and holding time, the intensity of the diffraction peaks of the samples increased continuously (1–3 experimental groups). There are some abnormal peaks in group 4, which may be mixed with a small amount of impurities. With the increase of crystallization temperature, the diffraction peak intensity of samples increases at first and then decreases, which is because the driving force for crystallization provided by the external temperature field decreases due to the short crystallization time [24], resulting in insufficient conditions for the crystal nucleus to grow up, which leads to the low crystal content of the crystallized samples. With the extension of the crystallization time, the driving force of crystallization also increases, the crystal nucleus fully develops and grows, and the crystal content of the sample after crystallization is higher. Therefore, the diffraction peak intensity of the sample will increase with the increase of the crystallization temperature within a certain range, but an excessively high crystallization temperature will cause the sample to remelt and suck back during the crystallization process, which leads to the increase of glass phase content, the diffraction intensity of the crystallization peak decreases.
3.5 Microscopic morphology analysis
The experimental results shown in Fig. 4 show that sample No.1 has a strong crystallization ability, and the formed crystals are massive, dense, and uniform, attached to the glass substrate, and the pore distribution is also relatively uniform. The crystallization ability of No.2, 3, and 5 samples is insufficient, a few massive crystals are precipitated, and the number of nucleation is not much, and the pores are large. Sample No.4 precipitated some massive crystals, with dense pore distribution and honeycomb shape. Sample No.6 precipitates a lot of crystals, which are granular, massive, and a small number of flakes, and the pores are evenly distributed. Sample No.7 has large pores and sparse distribution, and the precipitated crystals are massive and evenly distributed. Sample No.8 is ideal, with many and dense precipitated crystals, most of which are massive, and the pore distribution is relatively dense. The morphology of the No.9 sample is dominated by massive crystals, some of which are inserted into the glass matrix in sheet form and wrapped by massive crystals, with large pores and uniform distribution. SEM analysis shows that with the increase of nucleation temperature and nucleation time, the nucleation process of the glass matrix is relatively sufficient, and the number of nuclei and the content of the crystal phase of the samples increase. With the extension of the crystallization time and the increase of the crystallization temperature, the content of the main crystal phase will also increase, and the crystal grains of the main crystal phase aluminum diopside are interlaced with each other, and the arrangement is more uniform and dense.
3.6 Optimization scheme
Combined with the mechanical properties obtained by orthogonal experiment, XRD pattern, and SEM morphology analysis results, it can be known that the optimal scheme for synthesizing foamed glass-ceramics is to select the second group of basic glass, with CaCO3 content of 10 wt.%, the crystallization temperature of 1000 ℃ and holding time of 3 h. The XRD pattern of this group of samples shows that the main crystal phase is diopside (CaMgSi2O6), and the diffraction peak intensity is high, as shown in Fig. 5c. The SEM image of the sample shows that the precipitated massive crystal grains are more and evenly distributed, and the pore distribution is relatively dense (Fig. 5d). The adsorption-desorption isotherm curve of the sample shows type III, the multi-point BET specific surface area is 3.77 m2·g− 1, the single-point total pore volume is 0.013491 cc/g, and the single-point average pore radius is 217.80 Å (Fig. 5a-b). The mechanical properties of the samples prepared by the optimal scheme show that the bending strength, fracture toughness, and elastic modulus are 95.73 Mpa, 53.09 MPa·m1/2, and 28023.55 Mpa, respectively, which are superior to other orthogonal experimental groups, and the experimental results are consistent with the theoretical analysis of XRD and SEM.