A new method to determine composition of sphalerite without secondary pollution based on CIELAB color space

Currently, most of the methods for mineral materials analysis generate secondary pollution, which is detrimental to human health. For instance, traditional methods for sphalerite analysis in the zinc (Zn) smelting industry including chemical titration, atomic absorption spectrometry, and inductively coupled atomic emission spectroscopy. Colored indicators and toxic heavy metals are used in the analytical processes, causing severe pollution. For some methods, liquid is transformed into gaseous plasma, which is more dangerous to human health. Due to large quantities of sphalerite being used, secondary pollution cannot be ignored. This study proposes a green analysis method for the detection of sphalerite based on colorimetry, which does not generate secondary pollution. The results show that the strong substitution ability of iron (Fe) for Zn contributes to their inverse correlation in contents. The lattice parameters decrease with the increasing Fe content, resulting in a darker coloration. Here, key colorimetry parameters of L*, a*, and b* show clear linear correlations with the Zn and Fe contents. Compared with traditional approaches, this new method is environmental friendly with high sensitivity and accuracy. The relative error and relative standard deviation were less than 10% and 5%, respectively. This study provides a significant reference for nonpollution determination of other mineral materials.


INTRODUCTION
The analysis of mineral materials is very important for the development of the green smelting process. 1 There has been an increasing amount of research for methods of mineral materials analysis in the last a few years. 2,3owever, most of them are accompanied by secondary pollution, which is detrimental to human health, and require high levels of expertise and technical skills from the testing personnel. 4Due to the global Zn ore stone production approaching 600 million tons per year, secondary pollution and damage to human health resulted from sphalerite analysis cannot be ignored. 5resently, the composition of sphalerite can be determined using two series of methods.The first series of the method involves the dissolution of the mineral under high temperature or high-pressure conditions, which is followed by the quantification of the target element's content based on its concentration in the resulting liquid phase.Prominent techniques of this category include chemical titration, 6 atomic absorption spectrometry (AAS), 7 inductively coupled plasma mass spectrometry (ICP-MS), 8 and inductively coupled plasma-optical emission spectrometer (ICP-OES). 9The second series of methods relies on discerning the characteristic spectral lines generated by internal electronic transitions induced by absorbed radiation energy, which facilitates direct and swift identification of the target element content in the mineral.Prominent techniques of this category include Raman spectroscopy, 10 scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS), 11 X-ray fluorescence spectroscopy (XRF), 12 and laser-induced breakdown spectroscopy (LIBS). 13ost previous studies have only used two or three methods to determine Fe or Zn contents in natural sphalerite, while rarely considering secondary pollution or damage to human health caused by different analysis methods. 14or example, the chemical titration, AAS, ICP-MS, and ICP-OES methods entail the dissolution of the mineral in strongly acidic solutions, such as HNO 3 or HCl. 8,9Notably, colored indicators, such as xylenol orange and toxic heavy metals of copper sulfate, are added to the sample during chemical titration.This analytical process is widely used by companies as a national standard method, 15 leading to severe organic and heavy metal pollution.In AAS, ICP-MS, ICP-OES, and LIBS methods, samples in their liquid or solid state are transformed to gaseous plasma under high temperatures.This change in state is even more dangerous to human health, since toxic heavy metals such as Zn, Pb, and As, originally in the aqueous state or solid state, are transformed into a gaseous state that is more easily taken in by human. 16In contrast, XRF is accompanied by stronger radiation compared to the former approach.
This work proposes a clean and facile method for analyzing composition of sphalerite based on colorimetry, which generates no secondary pollution and damage to human health.CIELAB color space based detection technology has been widely used in food, industrial production, and other fields.This technology is not only environmental friendly, but also possesses high sensitivity and accuracy.Driven by the advent of advanced optoelectronic sensing, electronic advancements, and computer innovations, the realm of color theory and technology have experienced profound development, with colorimetric analysis based on rigorous color measurement taking the forefront.
In 1971, the International Commission on Illumination released an upgraded version of Hunter Lab, CIELAB, a nonlinear transformation of XYZ.The Euclidean distance between two colors is equivalent to their perceived distance (for distances less than 10 units).CIELAB can convincingly demonstrate the relationship between color and the subject under test.In contrast to other color spaces, CIELAB is designed to simulate human vision and is very effective in detecting subtle color changes.Gardis et al. implemented a partial least squares regression model based on the variation in CIELAB L* (luminosity), a* (ranging from red to green), and b* (ranging from yellow to blue) values during the beef jerky drying process to predict the moisture content and has great potential for real-time monitoring. 17Márcia et al. analyzed cassava starch using CIELAB parameters to construct a classification tree for identifying and evaluating potential adulteration of cassava starch with turmeric or metanil yellow. 18Our study proposed a relationship between the concentration (C) of the target substance in the liquid and color space coordinates (L*, a*, and b*) and the maximum absorption wavelength (λmax) of the target substance at different concentrations.Compared with offline laboratory monitoring, the relative error of this monitoring method was less than 2%. 19David introduced a novel technique by employing image processing to rapidly capture ore colors, thus facilitating the classification of gold and silver grades in the context of silver mining. 20In spite of this, due to the differences and complexity in the crystallization degree, chemical composition, and occurrence state of minerals, no studies have been conducted on sphalerite.Exploring the colorimetry-based detection method for the directly determining of sphalerite still faces significant challenges.
The method proposed by this study reveals the main coloring mechanisms of sphalerite based on the characterizations of phase composition, particle morphology, and spectroscopic characteristics of sphalerite.The contribution and correlation of Zn and Fe to color variations were investigated.The effects of three main colorimetry parameters, that is, L*, a*, and b*, on the quantitation of Zn and Fe contents in sphalerite were compared.As Fe atoms replace Zn atoms in ZnS, sphalerite exhibits distinct color variations, and key colorimetry parameters of L*, a*, and b* show clear linear correlations with the Zn and Fe contents in sphalerite, L* achieves the best quantitative performance.Compared to the conventional methods, this method is rapid, facile, nondestructive, low cost, environmental friendly, and easily operatable and does not produce secondary pollution.Leveraging the CIELAB color space as an alternative tool for evaluating the Zn and Fe contents in sphalerite can effectively reduce the amount of pollutants, decrease the difficulty and pressure of final treatment, and improve energy utilization efficiency.Promoting the effective utilization and green sustainable development of sphalerite resources is of great significance.

Fe, Zn, and S contents by chemical titration
Chemical titration is widely used by companies as a standard method in China.To determine the Fe, Zn, and S contents, we used chemical titration to analyze the contents of Zn and Fe in the sphalerite.The Zn, Fe, and S contents of ZFS-1 to ZFS-6 are shown in Figure 1.As can be seen, Zn content is noticed to increase from 40.47% to 51.67%, and Fe content decreases from 14.76% to 1.56%.S content fluctuates from 33.53% to 36.11%, with no incremental or decremental trend.Meanwhile, Zn content and Fe content exhibit a negative linear correlation, and the coefficient of determination is high reaching 0.97, suggesting the increasingly less substation of Zn by Fe from ZFS-1 to ZFS-2.Notably, the Zn content exhibited a pronounced negative linear correlation with Fe content, as evidenced by y Fe = − 1.12x Zn + 59.31, which had a high coefficient of determination (R 2 ) of 0.97.The ratio of the relative atomic masses between Zn and Fe was 1.16, and the absolute slope, |k|, was remarkably similar at 1.12.This proximity suggests the likelihood that Fe has substituted Zn.
Although the instrument used in this method is simple, the application range is wide and the result accuracy is high, the operation is complicated, time-consuming and requires high levels of expertise and technical skills from the testing personnel.It is not suitable for production control analysis.This method dissolves the solid in an acidic liquid under high temperature and adds large amounts of colored indicators, such as dimethylphenol orange and toxic heavy metal copper sulfate, and then indirectly obtains the content of the solid according to the content of the target element in the liquid, causing organic severe and heavy metal pollution.

2.2
Composition information based on solid-phase characterization

Information from XRD and Raman spectroscopy
Based on the XRD patterns in Figures 2A, it could be reached that (i) all six ZFSs exhibit a cubic crystal structure, (ii) all six ZFSs are highly crystallized, (iii) the predominant mineral phases are ZnS and Zn 3 Fe 2 S 5 , with a small amount of SiO 2 .It is worth noting that the Zn 3 Fe 2 S 5 signal is getting weaker from ZFS-1 to ZFS-6, indicating the Fe content decreases from ZFS-1 to ZFS-6.This result is consistent with the above mentioned conclusion that Zn substation for Fe tends to weaken from ZFS-1 to ZFS-6.According to Figure 2B, the cell parameter (a 0 ) increases from 5.3988 Å (ZFS-1) to 5.4786 Å (ZFS-6), which represents a 1.46% reduction.Generally, a 0 of the lightcolored sample deviates slightly from theoretical values and is slightly smaller, whereas a 0 of the dark-colored sample surpasses the theoretical value.Particularly, with an increment in the Fe content, a 0 tends to decrease for six ZFSs, indicating a slight contraction of the crystal lattice structure.This is possibly owing to the smaller radius of Fe atom compared to that of Zn atom, resulting in enhanced interatomic bonding.Meanwhile, Fe-S bond is shorter than Zn-S bond, further strengthening the interaction between Fe and S atoms and causing S atoms bonded to Fe atoms to shift toward Fe atoms (as shown in Figure 2C). 10Therefore, in terms of geometric structure, the crystal structure becomes more compact with increasing Fe content.Notably as the number of Fe atoms replacing Zn atom increases and the Fe content increases, the color of sphalerite samples darkens.It is deduced that such a compact arrangement can affect the intrinsic density and then affect the flotation energy consumption of sphalerite.Importantly, these findings were consistent with the results obtained using the titration method.
Raman spectra of the six ZFSs are shown in Figure 2D, which are similar to those reported in the literature. 7The Raman spectrum is split into two fundamental modes: the longitudinal optical (LO) mode at ∼349 cm −1 (the vibrations are parallel to the exciting light) and the transverse optical (TO) mode at ∼273 cm −1 (perpendicular vibrations). 7,10,11In terms of atom bonding, the LO mode (349 cm −1 ) was assigned to the Zn-S vibration, [7][8][9][10] while the Raman lines at 298 cm −1 were assigned to the Fe─S vibration mode, both cations being in tetrahedral coordination. 21If the LO mode is the fundamental vibration of M─S bonds, one notices that the former shifts to wavenumbers that are lower when M═Fe and higher when M═Zn.The explanation lies in the size of the atoms: The contents of Zn (A), Fe (B), and S (C) obtained by chemical titration, the relationship between Zn content and Fe content (D) from ZFS-1 to ZFS-6.Each sample was tested three times to take an average value.
Fe 2+ has a smaller ionic radius than Zn 2+ , therefore Fe forms shorter and, consequently, stronger bonds with sulfur.These bonds require higher energies to vibrate; thus, the peaks appear at lower frequencies. 17he variation in the color intensity among the different series of sphalerite samples did not significantly affect the peak positions.The scattering peaks located at 273 cm −1 and 349 cm −1 were attributed to the Zn-S vibrations, which corresponded to LO mode of the Zn-S bonds.A new scattering peak appears at 298 cm −1 , which was associated with the Fe-S vibration originating from the substitution of Zn atom by Fe atom, and their intensities vary about the Fe content, a correlation should be made between the intensities of these three peaks.Despite having a lower abundance, Fe-S signal is strong owing to the large scattering cross-section.As the color of sphalerite deepened, which indicated an increase in Fe content within the sphalerite, the substitution of Zn with Fe in the lattice structure became more significant.This led to a decrease in the atomic mass and consequently resulted in a blue shift in the wavenumber of the spectral bands.The relative intensities of these three peaks were closely related to Fe concentration in the samples.The substitution of Fe altered the crystal cell parameters, thereby influencing the Raman spectra.These findings are consistent with the results obtained using the XRD and chemical titration methods.

2.2.2
Information from SEM and EDS analyses SEM images and EDS analytical results of the six ZFSs are shown in Figure 3.For all six ZFSs, the particle sizes are heterogeneously distributed, and the particles exhibit diverse morphologies and distinct boundaries, which are directly correlated with the crushing and milling processes.Two points were selected in different areas of each sphalerite for EDS analysis.All of the six samples have strong Zn, Fe, and S peaks.A quantitative analysis of the chemical elements (points 1-12) was conducted for the different mineral particles within the sphalerite.The results indicated a significantly nonuniform distribution of Zn, Fe, and S on the surfaces of six sphalerite samples, with considerable variations in the elemental content at different points.The Zn content ranged from 39.8% to 57.0%, Fe content ranged from 3.6% to 16.4%, and S content ranged from 31.2% to 36.0%(Figure 3A-F).
Generally, the content of the target element in the analyzed area can be perceived up to 10 -14 .The measurement diameter is usually minimized to 1 μm and maximized to 500 μm.We note that the analytical value of one point of the SEM-EDS can only represent the composition of the sphalerite, not the composition of the entire mineral particle; it cannot be used to describe the overall composition of sphalerite.As the distribution of elements in sphalerite is not uniform, the results from "point for surface" vary.Therefore, there are considerable variations in the elemental content at different points.
For all six ZFSs, the EDS mapping results show that Fe and Zn distributions are heterogeneous and highly variable among different particles.Fe enters into sphalerite lattice via primarily two distinctive forms: the quasi-isomorphic substitution for Zn atom, giving birth to expansive solid solution denoted as (Fe x Zn 1-x )S; the concomitance of FeS within sphalerite structure. 22,23Particularly, for Zn, it is mostly enriched in ZFS-1 and least in ZFS-6.On the contrary, Fe is mainly enriched in ZFS-6 and least in ZFS-1.For S, its abundance is negligible and varied among the six ZFSs.Moreover, regions with high Zn content corresponded to regions with low Zn content.The measured results of Zn, Fe, and S contents do not agree with the results of chemical titration.This explains the significant differences in Fe and Zn measurements across different surface areas of the samples.A comparison of the particles within the sphalerite showed that characteristic diffraction peaks of Zn, Fe, and sulfur elements were observed in each sphalerite sample by EDS, which was consistent with the XRD analysis (Figure 3).This indicates the predominant presence of sphalerite and Fe-bearing sphalerite forms, which is consistent with the XRD results.

2.3
Fe and Zn contents by CIELAB color space colorimetry

Establishment of chromaticity model
The sphalerite samples used in this study are from Hunan Province, as shown in Figure 4.They are roughly the same size and have different colors.According to Figure 5A, the reflectance of different-grade sphalerite samples increased with the wavelength from 360 to 740 nm.The reflectance increased from a minimum of 18.8% to 38.7%, increasing by 105.85%.This indicates that the samples exhibited strong absorption in the shorter wavelength range (less than 400 nm) and relatively weaker absorption in the more extended wavelength range (greater than 400 nm).At the same wavelength, sphalerite exhibited a significant enhancement in light absorption and lower reflectance in  the visible light range with increasing Fe content.Owing to the proximity in radii between Fe and Zn ions within the crystalline lattice of sphalerite, Fe often engages in isomorphic substitution, permeating the lattice structure and exerting direct influences on the crystallographic arrangement.As the Fe content increases, discernible alterations manifest within the lattice parameters. 24Concurrently, the electronic transitions of Fe's outermost electrons on the sphalerite lattice induce profound chromatic manifestations.This is attributed to the interaction between Fe and light, which caused electrons in the lower-energy 3d orbital shell to transition to higher 3d orbital levels.The electron transfer of a crystalline field-type orbit featuring a nonbonded orbit 2e to Fe 2+ generated absorption at 500 nm.As a result, strong absorption bands and coloration occurred in the visible light region, which aligned with the characteristics observed in the Raman spectra, as for the three stimulus values of XYZ, representing the color produced by the interaction of visible light with sphalerite.As shown in Figure 5B, all samples exhibited varying degrees of reduction in the XYZ values.This reduction is consistent with the gradual decrease in Fe content, which contributed to the change of color characteristics, as previously discussed.
Based on Figure 5C Sensitivity represents the degree of change (ΔL * , Δa * , Δb * ) due to a change in the substance to be measured in a unit concentration of samples (Δ).As shown in Figure 5A, under different wavelength conditions, the reflectance () between the samples showed different trends.The dispersions of (), the corresponding reflectance values at different wavelengths, are calculated separately.The results show that the largest dispersion of for different samples is achieved at around 540 nm.It can be seen that the standard color matching function ȳ() has Gaussian-like form, whose maximum is also achieved at around 540 nm, indicating that ȳ() is more sensitive for those neighborhood wavelengths.From Equations ( 2) and ( 4), it can be seen that L* is positively correlated to the product of () and ȳ().The high sensitivity of () as discussed above will be further enhanced for L* by the multiplier of ȳ(), which is also sensitive at around 540 nm.Based on Figure 5C, D, F, G, I, and J, Figure 6

Advantages over other methods
Figure 7 presents a comprehensive comparison of the CIELAB color space and the chemical titration, ICP-OES, and XRF methods of evaluating sphalerite properties.Significant advantages were observed in various aspects based on the proposed nondestructive direct measurement method using the CIELAB color space model.The accuracy and precision were both less than 10%, which meet general requirements.The measurement range was wide, and the measurement time was fast, that is, as low as 5 s, resulting in low costs.Although the chemical titration method demonstrates high accuracy and precision in determining element contents, its complex preprocessing steps, such as dissolution/filtration and dilution, are cumbersome, time-consuming, and labor-intensive.Moreover, it requires suitable indicators to determine the endpoint, which inevitably reduces the operability of the testing process.ICP-OES is a relatively expensive precision instrument that offers high accuracy and precision for determining elemental contents, but has a narrow linear range.Moreover, although it exhibits high accuracy at low concentrations, the accuracy decreases as the concentration increases and even exceeds 10%.Although handheld XRF enables the direct measurement of element types in compounds with a short measurement time of up to 30 s, its accuracy and precision are relatively poor and may exceed 10%.This limitation prevents this method from satisfying the requirements of high-precision quantitative analysis in practical scenarios.In summary, the new approach based on a nondestructive direct measurement model using the CIELAB color space exhibited excellent performance for the six indicators.It can be used for the rapid, nondestructive, and accurate determination of the contents of the main and minor components of sphalerite.Furthermore, owing to its environmental friendly characteristics, it has great potential for further development and application.

CONCLUSIONS
1.This study proposes a green analysis method for direct measurement of Zn and Fe contents in natu-ral sphalerite that is environmental friendly, highly accurate, and has no secondary pollution was established based on the CIELAB color space comparative method.2. As Fe atoms replace Zn atoms in ZnS, the sphalerite exhibits distinct color variations.Colorimetry parameters L*, a*, and b* showed significant linear correlations with the Zn and Fe contents in sphalerite.The quantitative method and model based on color parameter L* demonstrated suitable sensitivity, as well as high accuracy and precision.3.This new method has the potential for future expansion to the detection of other types of mineral components, indicating significant prospects for its the development and application.

EXPERIMENTS
Detailed experimental materials and methods can be found in the Supporting Information section.

F
I G U R E 2 (A) XRD patterns of sphalerites with different Zn and Fe contents; (B) Zn, Fe, and S contents as a function of cell parameter; (C) schematic diagram of the substitution of Zn with Fe; and (D) Raman spectra of sphalerites with different Zn and Fe grades.

F I G U R E 4
Physical photos of sphalerite samples with different grades.
, F, and I, the colorimetry parameters of L*, a*, and b* exhibited a significant linear correlation with the Zn content.The goodness of fit, R 2 L* , reached a maximum of 0.97 and R 2 a* reached a minimum of 0.90, where R 2 L* > R 2 b* > R 2 a* ≥ 0.90.Similarly, as shown in Figure 5D, G, and J, the colorimetry parameters of L*, a*, and b* also demonstrated a notable linear relationship with the Fe content.The goodness of fit R 2 L* reached a maximum of 0.98 and R 2 a* reached a minimum of 0.93, where R 2 L* > R 2 b* > R 2 a* ≥ 0.93.However, as shown in Figure 5E, H, and K, there was no discernible correlation between the colorimetry parameters of L*, a*, and b* and the S content.
reveals the following results: K L* = 0.25, K a* = 0.002, K b* = 0.01, K L* >> K b* > K a* for Zn and K L* = 0.21, K a* = 0.001, K b* = 0.009, K L* >> K b* > K a* for Fe.Similarly, colorimetry parameter a* and b* has lower sensitivity compared to L*.Table 1 lists the linear fitting parameters for the test data in Figure 5C, D, F, G, I, and J.The contents of Fe and Zn in sphalerite determined by CIELAB color space colorimetry were in good agreement with the results based on chemical titration.F I G U R E 5 Reflectance spectra (A) and three stimulation values XYZ (B) of sphalerite samples with different grades.Correlation between the Zn, Fe, and S contents and color parameters L*, a*, and b*, where (C), (F), and (I) show the correlations of Zn content with L*, a*, and b*; (D), (G), and (J) show the correlations between Fe content and L*, a*, and b*; and (E), (H), and (K) show correlations of S content with L*, a*, and b*.

F I G U R E 6
Comparison of the indicators, including k (A), MRE (B), RSD (C), and RMSE (D) obtained at L*, a*, and b* in different mass concentration ranges.TA B L E 1 Linear fitting parameters (y = ax + b).

F I G U R E 7
Comprehensive comparison of the performance of the direct determination of Zn concentrate by CIELAB color space, ICP-OES, chemical titration, and XRF.
Zn and Fe contents compared to a* and b*.By combining the information from Figure6B-D, the values of MRE Zn , MRE Fe , RMSE Zn , RMSE Fe , RSD Zn , and RSD Fe were obtained: MRE Zn > MRE Fe ≥ 3.05%, RMSE Zn > RMSE Fe ≥ 0.58% (wt.%), and RSD Zn > RSD Fe ≥ 4.35%.Colorimetry parameter L* can be used to quantify the Zn and Fe contents in sphalerite.The quantification model performed better for Fe than Zn, and the results were similar to the chemical titration method for determin- ing the Zn and Fe contents.Therefore, the CIELAB color space provides a suitable approach with adequate sensitivity, accuracy, and precision for quantifying the Zn and Fe contents in sphalerite.This method has significant research implications for achieving the efficient utilization of sphalerite resources and promoting its sustainable development in resource extraction.