Chromogenic Mechanism and Formation of Zonal Genesis of Raspberry-Red Grossular from the Sierra de Cruces Range, Mexico

Abstract

The raspberry-red grossular, discovered in the Sierra de Cruces in Coahuila, Mexico, is characterized by its zoned coloration, featuring a pink rim and a black mantle with a sharp color transition at the border. However, there is a notable lack of definitive and systematic identification characteristics pertaining to its special zones. The mineral chemical composition and chromogenic mechanism remain unsupported by empirical validation derived from specific experimental data. In this study, the gemological properties, chemical composition, and spectral characteristics are systematically analyzed to explore the chromogenic mechanism and formation of zonal genesis. The results of the X-ray diffraction pattern, Raman spectrum, and major elements’ composition show that the raspberry-red grossular samples are grossular with high purity. Mn ions are a direct coloring factor of the pink rim of the grossular samples, while Fe ions are chromogenic elements of the black mantle. The MnO content of the pink rim ranges from 0.15 wt% to 1.72 wt%. The FeO content of the black mantle ranges from 3.11 wt% to 5.09 wt%, which is generally higher than that of other parts. The trace element compositions reveal that the rim and core of samples were formed in an oxidative environment (δEu = 0.43–2.41), which could be derived from the hydrothermal metasomatic skarn (δ¹⁸O = 11.03–12.14); the mantles of samples were formed in a reducing environment (δEu = 0.42–0.85), which is consistent with the magmatic skarn (δ¹⁸O = 11.40–11.66). They also show that the surrounding rocks provide part of the compositional sources for the raspberry-red grossular and interact with the black mantle, which affects the formation of the pink rim. This study offers geological and mineral compositional insights, addressing a significant void in the study of raspberry-red grossular, and lays the foundation for follow-up investigations.

1. Introduction

Gem-quality garnet has garnered widespread favor among consumers worldwide due to its abundant colors, vivid hues, intense fire, and exceptional cost-effectiveness. The general chemical formula for garnet is A₃B₂[SiO₄]₃, wherein A represents divalent cations such as Ca²⁺, Mg²⁺, Fe²⁺, Mn²⁺ while B represents trivalent cations such as Al³⁺, Fe³⁺, Cr³⁺ [1]. Garnets are generally classified into two series: the aluminum series, which is primarily composed of small-radius divalent cations and Al³⁺ that form isomorphous series; the calcium series, which is characterized by large-radius divalent cation Ca²⁺-dominated isomorphous series [2]. Current research on grossular predominantly centers on hessonite, which exhibits an orange hue due to the presence of manganese [3,4], and tsavorite, which displays a vivid green coloration attributed to vanadium and chromium [5,6,7,8,9]. The research of grossular encompasses various aspects, including geological contexts, chromogenic mechanisms, and origin identification. In orange grossular garnet, a minor proportion of Al³⁺ at the B site of the trivalent cations can be substituted by Fe³⁺, resulting in a subtle reddish tint to the otherwise shallow orange color of the garnet. Notably, among the various types of grossular, there exists a unique variant known as raspberry-red grossular.

Raspberry-red grossular is a rare gem-quality grossular exclusively found in the Sierra de Cruces, Mexico [10]. It is characterized by its zoned coloration, featuring a pink rim and a black mantle, characterized by a pronounced color transition at the boundary. Previous research indicates that the pink rim of raspberry-red grossular is primarily composed of grossularite, while the black mantle exhibits high titanium content [11]. The raspberry-red grossular is approximately 0.11 wt% Mn³⁺ and the pink color of grossular may result from various chemical substitutions [12]. Notably, there have been no documented instances of raspberry-red grossular displaying a white core in conjunction with a pink rim and black mantle. Additionally, the mineral composition and chromogenic mechanisms have not been substantiated by empirical evidence derived from specific experimental data, resulting in a lack of conclusive findings.

This study aims to conduct a thorough gemological and spectral analysis, alongside an examination of the chemical compositions, to investigate the gemological properties and mineralogical characteristics of raspberry-red grossular from Coahuila, Mexico. We will explore the chromogenic mechanisms and the formation of zonal genesis in these raspberry-red grossular samples, which may provide significant insights into the geological formation conditions and address existing research gaps.

2. Geological Setting

The Sierra de Cruces Range (18°59′–19°43′ N, 99°00′–99°40′ W), located in the north of the Sierra Mojada Mountains in Coahuila, is the western border of the Mexican Basin (Figure 1). It is rich in strontianite, silver, copper, cobalt minerals, and a recently mined iridescent type of iridescent sphalerite.

The Sierra de Cruces Range is predominantly composed of Cretaceous limestone, alkaline diorite, and minor occurrences of dolomite. The formation of skarn in this area is attributed to contact metamorphism occurring at the interface between the plutonic body and the limestone [13]. The majority of skarn and its associated ore deposits are found in proximity to the contact zone between carbonate rocks and intermediate-acid intrusive bodies, with carbonate rocks serving as the primary host. Notably, calcareous skarn that yields raspberry-red grossularite is located within the carbonate-rich host rock, and its mineral assemblages typically include grossularite, calcite, quartz, and idocrase, along with minor quantities of wollastonite, scapolite, and epidote [13].

3. Materials and Methods

3.1. Sample Description

Two representative samples, designated Gro-1 and Gro-2, were selected based on their favorable crystallization quality and distinct morphological features. Both samples exhibit a pink coloration with zoned characteristics. The Gro-1 sample is characterized by a pink rim, a black mantle, and a white core, with diameters ranging from 9 mm to 1.2 mm (Figure 2A,B). In contrast, the Gro-2 sample features a pink rim surrounding a black mantle, with diameters approximately between 1.1 mm and 1.6 mm (Figure 2C,D). The rims of both samples display a zonal structure that varies from subtranslucent to opaque. The crystals exhibit a high degree of self-formation, presenting a typical rhombic dodecahedral shape, and possess relatively well-developed internal fissures. Due to the small size of the samples, both were sectioned to facilitate observation and experimentation.

3.2. Analytical Methods

The gemological evaluations were conducted at the Gemological Experimental Teaching Center of the School of Gemology, affiliated with the China University of Geosciences (Beijing, China), encompassing gemological microscopic examinations, laser Raman spectroscopic analysis, UV-Vis spectroscopy, infrared spectroscopic studies, alongside a series of routine diagnostic tests. These auxiliary diagnostic procedures encompassed the measurement of refractive index (RI) and birefringence by using a gemological refractometer after polishing the surface of the samples, the determination of specific gravity (SG) through the application of the hydrostatic weighing technique, and the observation of pleochroism utilizing a dichroic scope.

The UV-Vis fiber spectrometer, model UV-5000, manufactured by Nanjing Baoguang Detection Technology Co., LTD (Nanjing, Jiangsu, China), utilizes a comprehensive signal-balanced light source. It covers a spectral range from 300 to 800 nm, boasts a resolution of 1 nanometer, maintains a signal-to-noise ratio exceeding 1000:1, and the text time is 50 ms.

Raman spectroscopic analysis of the sample matrix was conducted using an HR-Evolution microlaser confocal Raman spectrometer (CRS), manufactured by Horiba, Kyoto, Japan. Prior to analysis, the instrument was calibrated using a single-crystal silicon standard. The Raman spectrometer was configured with a laser wavelength of 532 nm, an output power set to 100 mW, a spot size ranging from 1 to 5 μm, and a spectral resolution of 1 cm⁻¹. Each spectral scan was acquired with an exposure time of 20 s, incorporating three integrations, and covered a wavenumber range of 100 to 2000 cm⁻¹. The collected Raman data were then compared against reference spectra from the RRUFF database.

X-ray diffraction (XRD) analysis was carried out utilizing a D8 Advance diffractometer equipped with CuKα1 radiation, operating at a voltage of 40 kV and a current of 100 mA. A graphite monochromator was employed, and the instrument was configured for continuous scanning across an angular range from 3° to 90°, with slit settings of DS = SS = 1° and RS = 0.3 mm. The experiment was executed at a controlled temperature of 21 °C and a relative humidity of 10%, employing the rotating anode polycrystalline X-ray diffraction technique in accordance with the JY/T 009-1996 standards.

The electron paramagnetic resonance (EPR) findings were acquired utilizing a paramagnetic resonance spectrometer, the ESR/EPR model A300 manufactured by Bruker (Karlsruhe, Germany), at the Technical Institute of Physics and Chemistry of the Chinese Academy of Sciences (CAS, Beijing, China). These measurements were conducted under controlled temperature conditions of 298 K, employing a microwave power of 20.03 mW and a modulation frequency set at 100 kHz.

Backscattered electron imaging was performed at the National Mineral Rock and Fossil Specimens Resource Center (Beijing, China), utilizing a high-resolution ZEISS SUPRTM 55 field emission scanning electron microscope, capable of achieving a resolution of 0.8 nanometers. The acceleration voltage was adjustable within a range of 0.1 to 30 kilovolts, and the operational working distance was maintained between 13 and 18 mm.

For microbeam X-ray fluorescence (micro-XRF) analysis, a Bruker M4 TORNADO X-ray fluorescence imaging spectrometer (Bruker, Karlsruhe, Germany) was employed, configured with a voltage of 50 kilovolts and a current of 600 milliamperes. The analysis was conducted under vacuum, utilizing a beam spot size of 20 μm, with a dwell time of 20 milliseconds per point and a step size of 10 μm.

Major and trace element analysis was conducted at the Key Laboratory of Gold Mineralization Processes and Resource Utilization, Shandong Provincial Key Laboratory of Metallogenic Geological Process and Resource Utilization, Shandong Institute of Geological Sciences (Jinan, Shandong, China). The major element analysis was carried out using a Jeol JXA-8100 electron probe microanalyzer (EPMA) from JEOL Ltd., Tokyo, Japan. The experimental conditions included an accelerating potential of 15 kV, beam current of 20 nA, accounting time of 20 s, and a spot size of 10 µm for mineral phase analysis. Data correction was performed using a modified ZAF procedure, with standards including chromium oxide (Cr), diopside (Ca), jadeite (Na), nickel (Ni), olivine (Si, Mg), hematite (Fe), pyrope garnet (Al), rhodonite (Mn), rutile (Ti), and sanidine (K). Trace element analysis was conducted on an Agilent 7900 quadrupole inductively coupled plasma mass spectrometer (ICP-MS), combined with a GeoLas HD 193 nm ArF excimer laser for laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). LA-ICP-MS was manufactured by Coherent in Santa Clara, CA, USA. Laser ablation was performed at a repetition frequency of 5 Hz in a helium atmosphere, with concentration errors of most elements being less than or equal to 10%. The analysis was conducted on polished sample surfaces [14].

The O isotopic composition was examined at the Institute of Mineral Resources, which is part of the Chinese Academy of Geological Sciences located in Beijing, China. The analysis employed the 253 plus instrument manufactured by Thermoelectric Corporation, based in West Chester, PA, USA. A dual-channel injection technique was utilized during the analysis. The sample underwent a chemical process involving bromine pentafluoride to liberate oxygen, which was trapped in a vacuum-sealed tube containing a 5Å molecular sieve. The isotopic ratio of δ¹⁸O in the oxygen was then quantified using a 253 plus gas isotope mass spectrometer. The precision of the analysis for standard samples was found to exceed ±0.2‰ when referenced against the V-SMOW standard, demonstrating high external accuracy. For individual sample tests, the internal precision achieved was 0.05‰. The laboratory employed GBW04421 as the primary standard, which is recognized as a top-tier national reference material for silicon isotopes, featuring a δ³⁰Si_NBS-28 value of −0.02. Following calibration with the international benchmark NBS-28, the δ¹⁸O_V-SMOW value for GBW04421 was established at +10.92‰. For a comprehensive overview of the instrumental setup and conditions used, please refer to the source in [15].

4. Results

4.1. Gemological Properties

The fundamental gemological characteristics of the analyzed samples are presented in

Table 1. Gemological properties of garnet samples.

Sample Number	Gro-1	Gro-2
Property	Sierra de Cruces, Coahuila, Mexico
Color	Pink–black–white–black–pink	Pink–black–pink
Weight	3.32 g	3.18 g
Diaphaneity	Pink: subtranslucent; black and white: opaque
Luster	Glassy luster
RI	1.744	1.751
SG	3.6734	3.5825
UV	LWSW inertness
Internal features	Black solid inclusions; partially healed fractures

. Two distinct raspberry-red grossular samples were identified, both displaying a characteristic zonal structure. The Gro-1 sample features a pink rim, a black mantle, and a white core structure. The Gro-2 sample also possesses a pink rim, but it solely surrounds a black mantle. The internal composition of the Gro-2 sample is marked by zonal patterns and uneven distribution (Figure 3E,F), indicating a significant degree of idiomorphism. The transparency of these samples ranges from subtranslucent to opaque, characterized by the presence of developed internal fissures and a limited quantity of black solid inclusions (Figure 3A–D). Overall, the samples exhibited typical gemological properties associated with garnets.

4.2. Spectra Analysis

4.2.1. X-Ray Diffraction Analysis

X-ray diffraction testing was conducted on the rim, mantle, and core of Gro-1, with the results shown in Figure 4. The diffraction peak positions for the raspberry-red grossular samples from the core, mantle, and rim exhibit a high degree of consistency, with distinct splitting of the diffraction peaks. The intensity of the peaks is greatest at the rim, followed by the mantle, while the core displays the lowest peak intensity. The data obtained align closely with the standard reference card for grossularite, indicating a high degree of correlation. It suggests that the primary phase components of the raspberry-red grossular samples are grossular of high purity, with no detectable impurities present.

4.2.2. UV-Vis-NIR Spectra

The UV-Vis-NIR absorption spectra of different parts of raspberry-red grossular samples display similar patterns, as shown in Figure 5. Grossular does not possess a distinctive absorption spectrum, as the predominant divalent cation (Ca²⁺) does not directly influence the formation of the UV-Vis-NIR absorption spectrum. In this study, there are three distinct absorption bands at 326 nm, 516 nm, and 685 nm. The absorption band at 326 nm is presumed to be caused by the transition from a six-fold excited state to a four-fold excited state of Mn²⁺ [16], the broad absorption band between 480 nm and 600 nm is presumed to be the spin-allowed electronic transition from ⁵E_g to ⁵T_2g of Mn³⁺ [17]. The absorption band at 685 nm is attributed to the charge transfer of Fe²⁺→Fe³⁺ in the octahedral position [18].

4.2.3. Raman Spectra

The absorption peaks in the fingerprint region of Gro-1 exhibit a consistent pattern, as illustrated in Figure 6. The symmetric C-O stretching mode associated with CaCO₃ is observed near 1086 cm⁻¹, which is exclusively present in the core [19]. The Raman spectra reveal shifts in the (Si-O)_Str peaks at approximately 1006 cm⁻¹, 879 cm⁻¹, and 822 cm⁻¹, with the peak at 879 cm⁻¹ displaying the highest intensity. Additionally, the (Si-O)_Bend peaks are identified at around 628 cm⁻¹, 548 cm^−1, and 415 cm⁻¹. The R(SiO₄)⁴⁻ peak is noted near 375 cm⁻¹ and 328 cm⁻¹, while the T(X²⁺) peak appears at approximately 277 cm⁻¹ and 247 cm⁻¹, which closely aligns with the Raman characteristic peaks of grossular [20,21]. The characteristic peak positions across different regions are similar, which are in proximity to the end members of grossular. These features corroborate the findings from X-ray diffraction analysis. Furthermore, the Raman peaks corresponding to the mantle and core regions exhibit a shift towards lower wavenumbers, and the intensity of the bending vibrations within the range of 400 cm⁻¹ to 650 cm⁻¹ is relatively diminished. This phenomenon can be attributed to the larger ionic radius of Fe compared to Al, which facilitates molecular vibrations and results in a gradual shift of the characteristic peaks towards lower frequencies. There is a significant disparity in the bond angles between Si-O-Si and Y-O-Y. Specifically, the bond angles of grossular between O₍₁₎-Si-O₍₂₎ and O₍₁₎-Y-O₍₅₎ are smaller than those observed in andradite, whereas the bond angles between O₍₁₎-Si-O₍₃₎ and O₍₁₎-Y-O₍₄₎ are larger than those in andradite [22,23,24]. The variations in these bond angles, as reflected in the Raman spectrum, indicate a gradual transition in the bending vibration peaks of the crystal from grossular to andradite with increasing Fe content within the structure [25]. The remaining peaks are unrelated to the composition of grossular, but the mixed peaks are caused by impurities and background effects.

4.2.4. EPR Spectra

The EPR spectra results reveal a strong paramagnetic signal at g = 2.003, as depicted in Figure 7B, which is attributed to the capture of unpaired electrons at surface oxygen vacancies and other defect sites [26]. Additionally, the spectrum displays a g-factor of 4.33, indicative of the Fe³⁺ signal (Figure 7A) [27].

4.3. Chemical Compositions

4.3.1. Micro-X-Ray Fluorescence

Two samples of Gro-1 and Gro-2 were scanned by micro-XRF. The chemical compositions of raspberry-red grossular samples were analyzed, as shown in Figure 8. The contents of Fe, Zr, and Ti in the black mantle are obviously higher than those in the red rim, while the contents of Al and Mn in the black mantle are obviously lower than those in the red rim. There is some subtle zonation in the pink region that shows that, the higher the Mn content, the darker the pink color is (Figure 8). The increased Fe content in these bands is associated with a corresponding decrease in Al content, suggesting that Fe is substituted for Al, thereby contributing to the formation of the black region.

4.3.2. Mineral Major Element Analysis

Table 2. Major element composition (wt%) of raspberry-red grossular samples by EPMA.

Samples	Gro-1-R-1	Gro-1-R-2	Gro-1-R-3	Gro-1-R-4	Gro-1-M-1	Gro-1-M-2	Gro-1-M-3	Gro-1-M-4	Gro-1-M-5	Gro-1-C-1	Gro-1-C-2	Gro-1-C-3	Gro-1-C-4
SiO2	39.80	40.05	39.11	39.46	37.69	38.18	38.44	39.03	39.15	39.48	39.92	39.34	39.54
TiO2	0.09		0.77	0.62	0.98	0.84	0.78	0.37	0.88	0.04	0.08	0.09	0.08
Al2O3	19.43	21.84	17.34	17.05	13.69	14.21	14.45	16.12	17.06	20.23	19.15	19.75	18.84
Cr2O3	0.07	0.02	0.02		0.05	0.03				0.04	0.03	0.04
FeO	3.28	0.31	4.46	4.89	9.74	9.16	8.92	7.56	4.20	3.11	3.50	3.24	4.34
MnO	0.35	1.04	0.15	0.16	0.17	0.20	0.18	0.19	0.14	0.25	0.30	0.21	0.29
MgO	0.60	0.82	0.98	0.83	0.90	0.91	0.80	0.60	1.10	0.26	0.50	0.32	0.30
CaO	36.32	35.88	37.15	36.93	36.70	36.44	36.39	36.10	37.38	36.47	36.45	36.83	36.54
Total	99.93	99.95	99.97	99.94	99.92	99.97	99.96	99.96	99.92	99.89	99.92	99.84	99.93
Gro	86.23	93.22	79.19	79.04	61.84	64.31	65.96	72.62	78.94	88.48	86.15	87.19	84.19
And	11.15	2.70	21.05	20.07	40.33	36.23	33.93	25.77	21.68	10.30	11.46	12.82	14.70
Uvr	0.21	0.06	0.06		0.17	0.08				0.13	0.09	0.12
Samples	Gro-2-R-1		Gro-2-R-2	Gro-2-R-3		Gro-2-R-4	Gro-2-R-5	Gro-2-M-1	Gro-2-M-2	Gro-2-M-3		Gro-2-M-4	Gro-2-M-5
SiO2	40.31		40.19	39.95		40.10	40.11	39.00	39.03	39.18		39.16	38.93
TiO2				0.02		0.02	0.02	0.98	0.95	0.99		1.00	0.84
Al2O3	21.71		21.78	21.04		21.51	21.81	16.50	16.68	16.62		16.75	16.62
Cr2O3			0.01	0.07			0.04		0.05
FeO	0.34		0.16	1.25		0.39	0.20	4.79	4.45	4.51		4.28	5.09
MnO	1.15		1.31	1.65		1.72	1.27	0.11	0.16	0.17		0.14	0.13
MgO	0.79		0.79	0.55		0.71	0.77	1.30	1.27	1.21		1.34	1.14
CaO	35.62		35.73	35.37		35.48	35.74	37.28	37.38	37.31		37.30	37.23
Total	99.93		99.97	99.89		99.93	99.97	99.96	99.96	99.98		99.98	99.98
Gro	93.62		93.39	90.28		92.08	93.30	76.10	76.94	77.34		77.54	76.02
And	1.07		1.82	4.35		2.68	2.01	24.40	23.69	22.78		22.61	24.80
Uvr			0.03	0.19			0.11		0.14				0.01

Alm: Almandine, Sps: Spessartite, Pyr: Pyrope, Gro: Grossularite, And: Andradite, Uvr: Uvarovite, R: Rim, M: Mantle, C: Core

lists the major element compositions of raspberry-red grossular samples measured by EPMA. The contents of Fe²⁺ and Fe³⁺ were calculated by the cation method and electrovalence difference method, and then the coefficient of each cation in the crystal formula was calculated by the oxygen atom method [28,29,30]. The CaO content of Gro-1 was 35.88~37.38 wt% and the content of Al₂O₃ was within the range of 13.69 wt% to 21.84 wt%. The Al₂O₃ content of the core was significantly higher than that in the rim and mantle, and the MnO content of the core ranged from 0.14~1.04 wt%. Compared with the rim and core, the FeO content of the mantle is higher, ranging from 4.21 wt% to 9.74 wt%. The CaO content in Gro-2 ranges from 35.37 wt% to 37.38 wt%, and the Al₂O₃ content in the rim was 21.04~21.82 wt%, which is significantly higher than that in the mantle (16.50~16.75 wt%). The MnO content in the rim was 1.27~1.72 wt%, the MnO content in the mantle was within the range of 0.11 wt% to 0.17 wt%, and the FeO content in the mantle was 4.28~5.09 wt%, which is much higher than the FeO content of the rim (0.16~1.25 wt%).

4.3.3. Mineral Trace Element Analysis

Table 3. Trace element composition (ppm) of raspberry-red grossular samples by LA-ICP-MS.

Samples	Gro-1-R-1	Gro-1-R-2	Gro-1-R-3	Gro-1-R-4	Gro-1-R-5	Gro-1-M-1	Gro-1-M-2	Gro-1-M-3	Gro-1-M-4	Gro-1-M-5	Gro-1-C-1	Gro-1-C-2	Gro-1-C-3	Gro-1-C-4	Gro-1-C-5
Sc	9.07	0.19	10.7	0.75	0.11	22.8	29.4	0.36	5.60	6.16	0.98	0.51	0.93	0.96	8.13
V	81.2	16.6	66.6	6.15	1.32	69.9	62.5	59.7	58.2	64.7	9.63	9.52	13.3	9.03	21.3
Cr	2.05	0.006	2.64		0.46	10.3	16.0	0.51	1.34	1.30	1.01	0.32	0.47	0.86	0.64
Co	0.17	0.14	0.14	0.14	0.26	0.11	0.17	0.15	0.25	0.23	0.051	0.018	0.018	0.041	0.024
Ni	0.63	0.22	0.71		0.039	0.60	0.88	0.16	0.92	0.56
Zn	8.54	4.41	7.25	3.94	6.98	4.75	7.34	2.36	5.43	4.93	7.88	2.31	3.77	4.84	2.13
Sr	0.67	0.20	0.37	0.17	0.12	0.44	0.86	0.47	0.55	0.53	9.07	10.2	6.69	7.69	15.7
Y	45.8	6.32	63.0	4.20	1.54	107	80.5	15.6	32.4	70.9	15.0	37.4	20.5	8.70	46.1
Nb	40.9	11.1	13.3	0.89	0.24	2.36	9.63	15.1	10.7	17.0	1.41	0.79	2.00	0.82	1.95
La	0.87	3.87	0.63	3.81	3.85	0.53	1.25	1.94	1.67	2.07	2.04	2.32	1.09	1.43	4.80
Ce	10.6	25.6	8.16	24.8	15.3	6.88	15.0	21.9	18.6	22.2	6.87	7.65	6.35	6.24	11.3
Pr	3.19	5.04	2.61	4.27	1.49	2.19	4.26	5.86	5.34	5.94	1.02	1.05	1.22	0.94	1.50
Nd	25.6	22.9	20.9	15.7	3.55	19.0	31.6	39.5	37.7	39.9	4.91	5.16	6.52	4.40	7.30
Sm	10.6	3.26	10.1	1.28	0.44	10.2	12.2	10.5	12.1	13.9	1.55	2.86	1.71	1.30	3.27
Eu	1.66	0.85	1.52	0.59	0.09	1.63	1.96	2.24	2.10	2.16	1.09	1.23	1.23	1.04	1.32
Gd	10.7	1.81	11.7	0.61	0.37	13.9	13.4	6.22	8.61	13.1	2.14	5.66	2.39	1.34	4.49
Tb	1.59	0.21	1.97	0.081	0.051	2.69	2.15	0.69	1.13	2.08	0.38	1.05	0.47	0.25	0.99
Dy	9.56	1.19	12.2	0.58	0.28	19.5	15.2	3.65	6.84	13.4	2.59	6.63	3.36	1.49	7.28
Ho	1.75	0.22	2.49	0.12	0.049	3.98	3.13	0.59	1.24	2.75	0.52	1.18	0.69	0.27	1.58
Er	4.94	0.57	6.89	0.45	0.13	11.9	9.37	1.38	3.44	7.56	1.41	2.76	1.94	0.85	4.77
Tm	0.67	0.08	1.05	0.08	0.02	1.74	1.41	0.17	0.52	1.09	0.18	0.31	0.27	0.11	0.72
Yb	4.36	0.59	6.46	0.52	0.13	11.8	9.95	1.12	3.46	7.15	1.21	1.80	1.60	0.74	4.92
Lu	0.59	0.08	0.90	0.11	0.06	1.68	1.43	0.12	0.47	0.98	0.13	0.18	0.18	0.12	0.67
Hf		0.08		0.04	0.002			0.31	1.40	0.58	0.23	0.14	0.07	0.22	0.50
Ta	11.4	0.055	4.39	0.009	0.003	1.12	2.31	0.78	1.59	2.15	0.11	0.062	0.100	0.066	0.15
Pb	0.23	0.008	0.088	0.005		0.014	0.043	0.018	0.034	0.038	3.12	1.09	0.78	1.82	3.47
Th	73.9	0.99	26.0	0.25	0.069	2.55	9.67	4.53	6.97	12.8	0.27	0.23	0.25	0.15	0.40
U	1.43	2.86	0.74	2.89	1.57	0.63	1.17	1.02	1.29	1.78	0.32	0.20	0.30	0.19	0.96
Samples	Gro-2-R-1		Gro-2-R-2	Gro-2-R-3	Gro-2-R-4		Gro-2-R-5	Gro-2-M-1		Gro-2-M-2	Gro-2-M-3		Gro-2-M-4		Gro-2-M-5
Sc	0.28		0.22	0.057	0.028		0.28	2.55		3.64	3.44		3.35		1.86
V	3.04		2.85	1.91	1.73		2.98	73.1		75.2	77.3		76.0		76.8
Cr	0.45		1.28	0.41	0.15		0.74	2.77		2.83	1.50		2.07		0.32
Co	0.27		0.33	0.27	0.22		0.22	0.091		0.16	0.15		0.14		0.074
Ni			0.019		0.21		0.12	0.55		0.24	0.81		0.0024		0.46
Zn	8.70		8.99	7.26	6.92		7.08	4.93		7.63	5.28		5.39		3.25
Sr	0.16		0.21	0.20	0.19		0.18	0.35		0.36	0.37		0.39		0.32
Y	1.30		1.32	1.32	1.36		1.88	52.2		62.3	56.8		53.6		54.7
Nb	8.90		6.00	1.91	2.17		4.47	38.7		42.2	34.6		40.8		27.0
La	0.68		0.42	0.14	0.21		0.24	0.79		0.63	0.76		0.80		0.59
Ce	2.89		2.54	0.47	0.65		1.21	8.60		7.11	8.90		9.42		6.90
Pr	0.52		0.48	0.080	0.080		0.17	2.61		2.25	2.69		2.83		2.19
Nd	2.13		1.94	0.39	0.46		0.75	20.1		18.1	21.2		21.8		17.8
Sm	0.24		0.24	0.13	0.15		0.15	7.95		8.34	9.17		8.98		7.73
Eu	0.093		0.072	0.018	0.035		0.058	1.24		1.33	1.41		1.33		1.33
Gd	0.12		0.13	0.17	0.15		0.23	9.75		10.5	9.87		10.3		10.1
Tb	0.027		0.034	0.023	0.021		0.029	1.68		1.81	1.76		1.67		1.74
Dy	0.20		0.18	0.22	0.19		0.30	11.3		12.5	11.7		11.3		11.1
Ho	0.052		0.041	0.031	0.033		0.048	1.97		2.50	2.22		2.17		2.02
Er	0.083		0.13	0.14	0.11		0.21	5.28		6.53	6.22		5.43		5.61
Tm	0.019		0.019	0.017	0.018		0.034	0.65		0.91	0.80		0.75		0.74
Yb	0.17		0.11	0.14	0.15		0.27	4.10		5.31	4.60		4.26		4.47
Lu	0.027		0.032	0.032	0.016		0.040	0.52		0.64	0.60		0.52		0.50
Hf	2.89		2.03	0.16	0.16		2.01	5.96		7.57	6.89		8.15		1.78
Ta	0.96		0.64	0.076	0.11		0.50	7.77		9.94	8.00		8.08		5.51
Pb			0.019	0.011	0.0024		0.0081	0.096		0.079	0.094		0.11		0.059
Th	0.66		0.65	0.54	0.48		0.53	26.5		26.6	27.7		34.1		19.0
U	0.24		0.22	0.091	0.099		0.12	0.99		1.42	1.06		1.31		0.85

R: Rim, M: Mantle, C: Core.

lists the trace element composition of raspberry-red grossular samples measured by LA-ICP-MS.

The chondrite-normalized rare earth element (REE) patterns indicate that the black mantle exhibits a higher total REE (∑REE) concentration compared to both the rim and core, with the core displaying the lowest ∑REE content. The rims of the two samples demonstrate an enrichment in light REEs (LREE), characterized by (La/Yb)_N values ranging from 0.14 to 5.27, while the rims of the two samples exhibit a relatively flat pattern of heavy REEs (HREE), with (Gd/Yb)_N values between 0.756 and 2.59. The white core of Gro-2 shares similar characteristics with the rim, showing LREE enrichment with (La/Yb)_N values from 0.65 to 2.80. The white core of Gro-2 displays a depletion in HREE, as indicated by (Gd/Yb)_N values ranging from 0.58 to 0.99. In contrast, the black mantle displays a distinct pattern with HREE concentrations surpassing those of LREE. The mantles of both samples exhibit a flat trend in LREE, with (La/Yb)_N values between 0.03 and 1.25, while showing enrichment in HREE, as evidenced by (Gd/Yb)_N values ranging from 0.98 to 4.60.

4.3.4. Oxygen Isotope

The δ¹⁸O value at the rim of Gro-1 is 11.75 ‰, the δ¹⁸O value at the mantle of Gro-1 is 11.40 ‰, and the δ¹⁸O value at the core of Gro-1 is 12.14 ‰. The δ¹⁸O value at the rim of Gro-2 is 11.03 ‰ and the δ¹⁸O value at the mantle of Gro-2 is 11.66 ‰. There is not much difference between these five values, all of the values are close to the δ¹⁸O value of skarn and distinct from other rocks (Figure 9).

5. Discussion

5.1. Chemical Composition of Raspberry-red Grossular

End members refer to extreme points in a material composition. A mineral sequence may have two or more end member components, and different end members contribute varying proportions of components to the mineral, ultimately forming a specific mineral. For example, in the Ca series of garnet, the three end members are grossular–andradite–uvarovite. The different percentage compositions of these three end members lead to the formation of garnets with different properties. It is evident that the primary end member component of the two samples is grossular (Figure 10). The pink rim chemical component of the samples is grossular, with high purity. The mole fraction is basically above 85%, and the content of other end member components is low. All the detection points are very close to the end members of the grossular. In the black mantle, the mole fractions are mostly more than 70%. The second is the andradite and the mole fraction is more than 20%.

In the scanning electron microscope images of the two samples, a distinct distribution of bands was observed, due to the difference in composition. According to the results of XRD analysis, it is found that the samples in this study are both grossular. It can be excluded that the black mantle is caused by the inclusions. The 23-degree diffraction peak of the core is speculated to be related to Al₂O₃, while the 29.5-degree diffraction peak is speculated to be related to calcium carbonate from the surrounding rock (Figure 4). There are common quartz and calcite inclusions in the samples (Figure 11A,B), which may be from the surrounding rocks—skarn [37]. In addition, there is also some distinct zoning in the pink rim of the samples (Figure 11C,D), which reflects variations in the composition of the fluids.

5.2. Chromogenic Mechanism of Raspberry-Red Grossular

There are three distinct absorption bands at 326 nm, 516 nm, and 685 nm of UV-vis-NIR spectra (Figure 5). Grossular does not have a characteristic absorption spectrum because its main divalent cation Ca²⁺ does not contribute directly to the generation of the UV-Vis-NIR absorption spectrum. The absorption band at 326 nm is presumed to be caused by the transition from a six-fold excited state to a four-fold excited state of Mn²⁺ [18]. The broad absorption band between 480 nm and 600 nm in the green and yellow light region is presumed to be the spin-allowed electronic transition from ⁵E_g to ⁵T_2g, belonging to the d-d electron forbidden shift of Mn³⁺ [38]. The absorption band at 685 nm is attributed to the charge transfer of Fe²⁺→Fe³⁺ in the octahedral position [39]. The MnO content in the pink rim is significantly higher than that in the black mantle (Table 2), which further proves that the raspberry-red color of grossular is related to the presence of manganese. Specifically, higher concentrations of Mn are associated with a darker pink hue. It is further corroborated by the surface scanning images obtained from X-ray fluorescence spectroscopy (Figure 8).

The EPR results reveal a strong paramagnetic signal at g = 2.003, as depicted in Figure 8B. It is attributed to the capture of unpaired electrons at surface oxygen vacancies and other defect sites [26]. Additionally, the spectrum displays a g-factor of 4.33, indicative of the Fe³⁺ signal (Figure 7A) [27]. UV-Vis-NIR spectra also show that the absorption band at 685 nm is attributed to the charge transfer of Fe²⁺→Fe³⁺ in the octahedral position [18]. For Mn²⁺ ions located at octahedral positions, the g-factor of 2.0 suggests that the linear center of resonance absorption should be around the magnetic field intensity H = 3400 × 10⁴ T. It is accompanied by a six-fold hyperfine structure characteristic signal resulting from hyperfine interactions [27]. However, there is no six-fold hyperfine structure characteristic signal detected in the sample, indicating that Mn²⁺ is not present in the sample. Mn³⁺ and Mn⁴⁺, lacking unpaired electrons, cannot be detected as signals in EPR. So, the form of Mn in the sample may be Mn³⁺ or Mn⁴⁺. While the UV-Vis-NIR absorption band between 480 nm and 600 nm belongs to the d-d electron forbidden shift of Mn³⁺, the absence of a coinciding g-factor at g = 2.0 indicates that the manganese in the samples is likely to be in the form of Mn³⁺.

The contents of FeO and Al₂O₃ show the opposite trend (Figure 12). In general, the Al₂O₃ content is significantly higher than that of FeO. But in the mantles of the two samples, FeO content is higher than those of the rim and core, ranging from 4.205 wt% to 9.741 wt%. In the rim and core of the two samples, the Al₂O₃ content ranges from 13.20 wt% to 21.91 wt% and the FeO content ranges from 0.16 wt% to 9.25 wt%.

It is speculated that in the initial Fe-rich oxidizing environment, Fe³⁺ is isomorphous in the [AlO₆] octahedral lattice to replace a part of Al³⁺ to form an andradite component. Therefore, the Fe content in the black mantle is obviously higher than in the rim and core, which results in the formation of a black mantle in the center of the grossular. It proves that Fe is a direct factor for the coloring of the black mantle of raspberry-red grossular. With the stabilization of the environment, Fe content decreases, and Mn ions accumulate up to a certain amount in the crystal lattice, causing the rim of the grossular to appear raspberry red (Figure 8).

5.3. Formation of Zonal Genesis

The ∑REE content in the black mantle is higher than that in the rim and core, with the core having the lowest ∑REE content. The normalized HREE content in the black mantle of the two samples is higher than that of LREE. The rim and core are completely different, with a higher LREE content than HREE. HREE is enriched in the black mantle, and the HREE content gradually decreases from the mantle to the rim (Figure 13).

The primitive mantle-normalized spider diagram (Figure 14) shows that U, Ta, and Ce are enriched in Gro-1 samples, while La and Sr are depleted. Conversely, the Gro-2 sample is characterized by enrichment in Ta, Ce, and Hf, alongside a depletion in La and Sr. The overall trend of REE variation in both samples is largely consistent. In the Gro-1 sample, the trace element concentrations in the core are generally lower than those in the rim and mantle, while in the Gro-2 sample, the mantle displays a higher concentration of trace elements than in the rim. The trace element distribution patterns of the pink rims in both samples closely resemble those of the surrounding rocks. This observation supports the hypothesis that the pink rims were formed through interactions between the black mantle of the grossular and the adjacent surrounding rocks, leading to a reduction in HREEs within the pink rims.

Garnets with similar REE distributions have different Eu anomalies (δEu). The δEu value (δEu = Eu_N/(Sm_N×Gd_N)^1/2 [41]) serves as an indicator of the oxygen fugacity [42]. In garnet structures, REEs predominantly substitute for divalent cations such as Ca²⁺, Fe²⁺, Mg²⁺, and Mn²⁺ through isomorphism with trivalent REE³⁺ ions. In hydrothermal fluids, REEs typically exist as positively charged trivalent ions. However, europium is characterized by its variable valence. Under oxidizing conditions, europium predominantly exists as Eu³⁺, and its associated elements do not display significant anomalies. Conversely, under reducing conditions, europium primarily exists as Eu²⁺, leading to observable anomalies in its associated elements. The extent of the depletion is influenced by the redox state of the system [43]. Therefore, the reduction in oxygen fugacity leads to a decrease in the concentration of Eu3+ in the hydrothermal fluids, which in turn leads to a relative decrease in europium in garnet and a decrease in the δEu value.

The δEu value of Gro-1 exhibits a decreasing trend from the core to the mantle, followed by an increase from the mantle to the rim. This pattern suggests a corresponding decrease in the oxygen fugacity of the metamorphic hydrothermal fluids initially, which is then followed by an increase. In contrast, the δEu value of Gro-2 shows an increasing trend from the mantle to the rim, indicating an increase in the oxygen fugacity of the metamorphic hydrothermal fluids (Figure 15). Notably, both samples display the highest δEu values at the rims, signifying that these regions were formed during a period characterized by the highest oxygen fugacity of the metamorphic hydrothermal fluids [44,45].

The negative Eu anomaly in garnet may be controlled by the fractional crystallization of plagioclase. As the fractional crystallization process intensifies, the negative Eu anomaly becomes significant. The distribution pattern of REEs in garnet of skarn and the formation environment should be related as follows: Eu positive anomaly, LREE enrichment, and HREE depletion (mostly La_N/Yb_N > l, Eu/Eu* > 1) indicate the oxidative hydrothermal metasomatic skarn environment; Eu negative anomaly, LREE depletion, and HREE enrichment (mostly La_N/Yb_N < 1, Eu/Eu* <1) represent the reducing magmatic skarn environment [46,47,48,49]. It can be concluded that the pink rim and white core of the samples were formed in hydrothermal metasomatic skarn in an oxidative environment, while the black mantle was formed in magmatic skarn. In the pink rim of samples, there are also some bands which have changes in Fe, Al, and Mn contents (Figure 9). It is speculated that the bands are caused by hydrothermal metasomatism.

6. Conclusions

This study systematically investigated two types of raspberry-red garnets from the Sierra de Cruces Range in Mexico with pink rims and black mantles, differing only in the presence or absence of a white core. The samples are high purity grossular with no impurities. The possibility of the black core being colored by inclusions has been excluded. All samples have distinct regional distributions due to differences in composition. Based on the UV-Vis absorption spectrum and the X-ray fluorescence, it is speculated that Fe³⁺ is isomorphous in the [AlO₆] octahedral lattice to replace part of Al³⁺ to form an andradite component, which results in the black color in the mantle of the grossular. Mn ions are a direct factor for coloring of the rim of the grossular, which leads to the raspberry-red color. The trace element analysis results reveal that the rim and core of samples were formed in an oxidative environment, while the mantles of samples were formed in a reducing environment. The pink rims are formed by the interaction between the black mantle of the grossular and the surrounding rocks.