Manganese dioxide [MnO₂, analytical reagent (AR)] and hydrofluoric acid [HF, guaranteed reagent (GR)] were purchased from Aladdin (Shanghai, China). Nitric acid (HNO₃, GR) and hydrochloric acid (HCl, GR) were purchased from Beijing chemical works (Beijing, China).

The cementitious materials in the test consisted of fluorite, sandstone, gypsum, bauxite, and limestone, which were provided by BBMG Liushui Environmental Protection Technology Co., Ltd. cement plant (Beijing, China).

The chemical characterization was determined by XRF (Axios Max, PANalytical, Holland). The mineralogical phases were identified by XRD (Ultima IV, Rigaku, Japan) using 2θ angle range of 5.0°–80° with a scan rate of 0.2 s per step and a step size of 0.02. The CuKα radiation (=1.5418 Å) was generated at 40 kV and 40 mA. The heavy metals were measured by Inductively Coupled Plasma-Mass Spectrometry (ICP-MS, NexION 300X, Perkin Elmer, United States).

Raman spectra were recorded with a JY HR800 Raman spectrometer (HORIBA, Japan), equipped with a ×50 objective lens (numerical aperture (NA) = 0.5) and a He−Ne laser with 514 and 633 nm wavelength. The laser power values measured in the experiments were obtained from a power meter (0.75 mW at the samples with a spot area of approximately 1.4 μm²). The Raman band of a silicon wafer at 520.8 cm⁻¹ was used to calibrate the spectrometer. The data integration time was 5 s for two accumulations.

The XPS analysis was carried out using an AXIS Supra X-ray photoelectron spectrometer with a monochromatic Al Kα (hν = 1486.7 eV) X-ray source operating at 225 W. The XPS spectra were measured under a 2 × 10⁻⁸ Pa vacuum, and the high resolution spectra were measured with a step size of 0.1 eV and pass energy of 20 eV. A standard Shirley background was used for all analyses. The binding energies were calibrated using the adventitious carbon peak at 284.8 eV.

Manganese ions were introduced into the raw materials (RMs) in the form of MnO₂ at doses of 0.0, 0.5, 1.0, 2.0, 3.0, 4.0, and 5.0 wt%, which were referred to as CM-Blank, CM-0.5%, CM-1.0%, CM-2.0%, CM-3.0%, CM-4.0%, and CM-5.0%, respectively. The composition of the cement designed for the study is presented in Table 1. All materials were ground in a vibrating ball mill for 10 min and then the powder was pressed into cylindrical blocks (diameter = 7 mm, height = 5 mm) at 100 kN to ensure a more uniform clinkering process. The cylindrical blocks were placed in a corundum crucible and heated up to 1450°C at a rate of 15°C/min, sintering for 30 min. The resulting clinkers were cooled rapidly outside the furnace and ground as homogenized clinker material (CM) powder for further analysis.

The chemical composition of the CM is shown in Table 2. The major constituents were oxides of calcium (Ca), aluminum (Al), and silicon (Si), The total wt% of CaO + Al₂O₃ + SiO₂ were 84.50%, 82.87%, 81.70%, 80.39%, 78.42%, 78.01%, and 75.88% in the blank CM sample and 0.5–5% Mn doped clinker samples, respectively.

The XRD patterns of Mn doped clinker samples are shown in Figure 1. The major component was Ca₄Al₆O₁₂SO₄. The characteristic peaks of Ca₂SiO₄ and Ca₃SiO₅ indicated the formation of a clinker mineral phase after the high temperature clinker calcination process. The XRD results were consistent with the data in Table 2, i.e., the dominant elements were Ca, Al, and Si. When the XRD results of the blank clinker sample were compared with those for Mn doped clinker samples, no new phases or obvious changes of crystal structure were apparent. Moreover, the referenced diffraction pattern of MnO₂ also did not exist in the clinker, even when the MnO₂ dose was 5.0%. It was therefore speculated that the Mn ion might be incorporated into the crystal lattice, and was unable to change the clinker crystal structure.

The leaching content of Mn in CM was determined according to Chinese Standard GB/T 30810-2014 (Test methods for leachable ions of HMs in cement mortar) and the results are shown in Table 3 (General Administration of Quality Supervision I a Q, P.R. China, 2014). The solidification ratio was calculacated to evaluate the retention of heavy metals in the clinker samples after cement clinker calcination process. High solidification ratio indicates high content of heavy metals solidified in clinker and has low risk to environment pollution (Gao et al., 2017). The solidification ratios of 0.5%–5% Mn doped clinker samples was calculated from the ICP-MS results and were 83.84%, 87.43%, 84.29%, 82.04%, 83.97%, and 82.62%, respectively (Miao, 2018). The solidification ratios calculated according to the XRF results were 86.89%, 85.19%, 77.17%, 77.52%, 77.69%, and 78.21%, respectively. The solidification ratio of Mn in cement was high when the Mn content was 0.5% and 1%, which indicated the effective fixation in cement clinker. The solidification ratio was lower when the Mn dose exceeded 1%. In general, the satisfactory solidification ratio provided evidence of the immobilization of Mn in the formation of cementitious materials.

Figure 2A shows the Raman spectra of Mn doped clinker samples recorded with 514 nm laser excitation. The signals at 443 and 612 cm⁻¹ were due to the υ2 and υ4 bending from [SO₄]²⁻ groups, the peak at 521 cm⁻¹ was due to the υ1 [AlO₄]⁵⁻ tetrahedra in belite, and the 834 cm⁻¹ band was assigned to the υ1 [SiO₄]⁴⁻ symmetric stretching in alite (Kloprogge et al., 2002; Potgieter-Vermaak et al., 2006; Black and Brooker, 2007; Black, 2009). The signals below 300 cm⁻¹ were assigned to Ca-O bonds (Torrens-Martin et al., 2013). A sharp band at 990 cm⁻¹ was assigned to the υ1 [SO₄] ²⁻ vibrations in the blank CM sample. However, the band weakened and disappeared when the Mn dose was 2%, and there was a band formed at 960 cm⁻¹ as the Mn content increased. This phenomenon may be due to the mineral phase transformation from yeelimite to belite, which was caused by the addition of Mn. Moreover, the peak at 628 cm⁻¹ that was assigned to MnO₂ was not found in clinker samples, proving the solidification effect of Mn after the high temperature calcination in the mineral formation process.

For 633 nm laser excitation (Figure 2B), the band at 991 cm^–1, which corresponded to yeelimite, disappeared in the 2% Mn doped clinker sample. Two major characteristic peaks could be observed at 1375 cm^–1 and 1406 cm^–1 when the dosage of Mn was below 3%, which could be assigned to vibrations of silicate materials (Hasanzadeh et al., 2016). Those two peaks became broader and red-shifted as the Mn content increased, which indicated a poor crystalline phase of the CM. The impurities in clinker mineral lattice and disordered silicates, aluminosilicates minerals may result in fluorescence emissions and background which became a big challenge to obtaining true Raman bands (Dyer et al., 1993; Bonen et al., 1994; Newman et al., 2005). The foreign Mn ions present in the clinker samples at a content exceeding 3% may broaden the spectra and cause a mineral phase disorder of the cementitious materials. Although the Mn content exceeded 3%, the characteristic MnO₂ peak was still not found in CM samples.

Overall, the increased addition of Mn influenced the crystallinity of the clinker; however, the Mn ion could be effectively immobilized and fixed in cementitious materials according to the leaching solidification ratio, XRD, and Raman spectroscopy results.

The chemical state and constituent composition of clinker samples were evaluated by XPS. The XPS narrow spectra of C 1s, Mn 2p, O 1s, and Si 2p for blank and 1% Mn doped CM samples are shown in Figure 3, and the spectra of the other Mn doped CM samples are provided in Supplementary Figure S1. The C 1s peaks identified at 284.8, 285.7, and 289.6 eV were the binding energies of contamination adventitious carbon, which accounted for the C-C, C-O-C, and O-C=O components, respectively (Kalina et al., 2014; Singh et al., 2020). The bonding energy situated at 293.2 eV accounted for C-F, which may form due to the presence of fluorite in the RMs. There were two peaks identified at 642.6 and 653.4 eV in Mn 2p, which corresponded to Mn-O in MnO₂. In the O 1s spectrum, the 529.8 eV peak energy was attributed to Ca-O and Mn-O, and the 532.0 eV peak energy was attributed to Si-O (Wang et al., 2016). The typical peak at 102.7–102.5 eV was assigned to SiO₂, and the peak at 101.9–101.7 eV was assigned to Si in belite (Estokova et al., 2018; Singh et al., 2020). The other O 1s spectrum for Mn doped CM samples is provided in Supplementary Figure S2.

Figure 4 shows the stacked Mn 2p spectra of MnO₂ and Mn doped clinker samples. For the Mn doped CM samples, it was found that the Mn 2p peak shifted to a higher bonding energy as the Mn dose increased. The changes of the chemical bonding environments can result in shifts in photoelectron energy, thus the drift of the Mn 2p peak was accordingly speculated as the changes in bonding structure (Black et al., 2003; Estokova et al., 2018). The Mn 2p peak shifted to a higher bonding energy indicating the progressive disordering of the silicate structure which may cause by the incorporation effect of silicate and/or aluminosilicate towards Mn (Rheinheimer and Casanova, 2014).

Furthermore, Supplementary Figure S3 shows the transformation of the Si 2p spectra. Compared with a blank sample, the peak proportion of 102.7 eV increased but was lower than that of 101.8 eV in the 0.5% Mn doped CM sample. As the Mn content increased, the peak proportions of the two peaks were the same as in the 1% Mn doped CM sample, but the peak proportion at 102.7 eV was higher than at 101.8 eV when the Mn content was 2%. This trend stabilized in the 3%–5% Mn doped CM samples. This phenomenon of the two peaks reflected the transformation of the SiO₂ and belite phases, which could explain the optimal solidification ratio of the 1% Mn doped CM sample shown in Table 3.

This change in the Si 2p spectra was hypothesized to be caused by the differences in the chemical environment of Mn doped CM samples, which was caused by the solidification of Mn in clinker (Nath et al., 2022). The addition of Mn influenced the formation of the silicate phase and the increasing Mn content generated an increase in the SiO₂ content and reduction in the belite phase, which may cause a disordered crystal structure. However, the XRD pattern and the solidification ratio were still satisfactory when the Mn content was 5%.