Study on the Correlation between Fe2+ and Peridot's Yellow Green Color and Quality Evaluation of Color Based on CIE1976 L*a*b* Uniform Color Space

Jun Tang, Ying Guo, Xiangxiang Lai, Yimin Liu, Qi Wu2016 · DOI

Abstract

Based on the CIE1976 $L^*a^*b^*$ uniform color space, the colors of 54 peridot samples were measured under a D65 light source. Chemical analysis identified all samples as forsterite and chrysolite. Results indicate that increasing $Fe^{2+}$ concentration shifts peridot's color toward yellow, though $Fe^{2+}$ shows no significant correlation with chroma. Using K-means clustering, the yellow-green colors were categorized into eleven groups, with ANOVA confirming significant differences in $L^*$, $a^*$, and $b^*$ values ($p < 0.001$). A linear discriminant function model developed via Fisher discriminant analysis achieved a 98.1% accuracy rate. Finally, a seven-level color grading scale was established based on lightness and chroma: Fancy Vivid (FV), Fancy Intense (FI), Fancy Deep (FDe), Fancy Light (FL), Fancy Dark (FDa), Light (L), and Very Light (VL).

Introduction

In the gemstone industry, color is the primary determinant of value. For peridot, a vibrant yellow-green gemstone, evaluation has historically relied on the subjective expertise of gemologists. However, human perception is inconsistent, and descriptive terms like "yellow-green" lack the precision needed for standardized trade. While ferrous iron ($Fe^{2+}$) is known to be the primary chromophore in peridot, the precise quantitative relationship between iron concentration and specific color attributes—such as lightness, hue, and saturation—has remained poorly defined. This study aims to replace subjective assessments with a scientific, quantitative framework for peridot color characterization and quality grading.

Methods and Materials

The study analyzed 54 high-purity peridot samples, ranging from 4mm to 8mm in size. The research methodology integrated chemical analysis with advanced colorimetry:

  • Chemical Composition: An Electro Probe Microanalyzer (EPMA) was used to determine the mass fractions of $FeO$ and $MgO$. The samples were identified as forsterite and chrysolite, with a calculated crystal chemical formula of $(Fe_{0.16-0.22}Mg_{1.74-1.85}Ni_{0.00-0.01})_{1.94-2.02}Si_{0.99-1.03}O_4$.
  • Color Measurement: Using an X-Rite SP62 spectrophotometer under a D65 standard light source, each sample's color was mapped into the CIE1976 $L^*a^*b^*$ uniform color space. This system quantifies Lightness ($L^*$), green-redness ($a^*$), and blue-yellowness ($b^*$).
Distribution of peridot samples within the CIE1976 L*a*b* color space.
  • Statistical Modeling: K-means clustering grouped the samples into 11 distinct mathematical color categories. Fisher discriminant analysis was then used to develop a linear model for automatic classification. Analysis of Variance (ANOVA) confirmed that the differences in color coordinates between these categories were statistically significant ($p < 0.001$).

Results and Discussion

The Impact of Iron on Color

The analysis revealed a clear correlation between chemical composition and perceived color. As the concentration of $Fe^{2+}$ increases, the hue angle ($h^\circ$) decreases, indicating a shift from green toward yellow. Interestingly, $Fe^{2+}$ concentration showed no significant correlation with chroma (color saturation), suggesting that while iron determines the hue, other factors likely influence the vividness of the color.

Correlation between Fe2+ concentration and hue angle (h°), showing the shift toward yellow as iron content increases.

Furthermore, iron has an indirect effect on lightness ($L^*$). Because yellow tones naturally possess higher lightness than green tones in this color space, the iron-induced shift toward yellow results in a corresponding increase in the gemstone's lightness.

Quantitative Grading System

The Fisher discriminant model achieved a 98.1% accuracy rate in correctly classifying peridot samples. To make these findings practical for the gem trade, the researchers condensed the 11 mathematical clusters into a 7-level grading system based on lightness and chroma, mirroring the "Fancy" grading system used for colored diamonds:

  1. Fancy Vivid (FV): High chroma and medium lightness; the "gold standard."
  2. Fancy Intense (FI)
  3. Fancy Deep (FDe)
  4. Fancy Light (FL)
  5. Fancy Dark (FDa)
  6. Light (L)
  7. Very Light (VL): High lightness but very low chroma, appearing grayish.

Conclusion

This research successfully transitions peridot color evaluation from subjective observation to a quantitative, reproducible science. By linking $Fe^{2+}$ content to specific coordinates in the CIE1976 $L^*a^*b^*$ space and applying machine learning for classification, the study provides a robust framework for gemstone grading. Future work should expand the dataset to include diverse global sources and evaluate color stability under varying light temperatures to further refine this standardized system.

Summary updated

Abstract

Based on CIE1976 L * a * b * uniform color space, 54 peridot samples' color were measured under D65 light source. The component analysis results showed that all samples are forsterite and chrysolite. It showed that with the increase of Fe 2+ , peridot's color turns to yellow. But Fe 2+ has no significantly correlation with chroma. By applying clustering analysis and discriminant analysis, the results showed that peridot's yellow green color were divided well into eleven categories by K-means clustering analysis, and ANOVA showed L * , a * and b * values among various categories were significantly different (sig0.001). A linear discriminant function model of the eleven categories of peridot's yellow green color was established by Fisher discriminant analysis which had excellent performance of discriminant with the model checking accuracy of 98.1%. Based on lightness and chroma, peridot's yellow green color was graded into seven levels, consisting of Fancy Vivid(FV), Fancy Intense(FI), Fancy Deep(FDe), Fancy Light(FL), Fancy Dark(FDa), Light(L), Very Light(VL).
BibTeX
@misc{tang2016study,
  author = {Jun Tang and Ying Guo and Xiangxiang Lai and Yimin Liu and Qi Wu},
  title = {{Study on the Correlation between Fe2+ and Peridot's Yellow Green Color and Quality Evaluation of Color Based on CIE1976 L*a*b* Uniform Color Space}},
  year = {2016},
  doi = {10.2991/emcpe-16.2016.123},
}
Figure