Iridescent Iron Oxides

Abstract

Iridescent iron oxides known as rainbow hematite and turgite are found in a variety of localities worldwide and display a variety of beautiful interference colors. Usually, there is a thin layer of nanocrystals containing aluminum and phosphorous coats, such as with hematite, although one example of aluminum with arsenic is presented. Infrared spectra of samples with thicker layers of these nanocrystals show absorption in the phosphate region. The thin films that range from tens to hundreds of nanometers thick are believed to cause the color. High-resolution secondary electron imaging shows that the thin film consists of nanocrystals arranged in three directions (120° apart). The rod-shaped crystals have a width from 5 to 35 nm and display the same morphology on all samples, irrespective of if they contain phosphorus or arsenic. The minute crystals have failed to produce either an X-ray powder diffraction pattern, an electron back-scatter diffraction pattern in SEM, or an electron diffraction ring pattern under TEM. Raman spectra are dominated by hematite features. Infrared ATR spectra of the bulk material show mostly hematite with occasional minor amounts of goethite and phosphate.

1. Introduction

Occasionally, hematite is brilliantly colored with what appears to be a thin film on its surface that produces colors that range from solid and uniform to thin and resembling the interference colors of an oil slick (Figure 1). Such specimens with thin, iridescent coatings are frequently referred to as iridescent hematite. Sometimes, iridescent specimens with thicker and brighter coloring have been referred to as ‘turgite’. Formerly, turgite was considered to be a mineral that was comprised of a hydrated hematite, ‘Fe₂O₃·½H₂O’. Posnjak and Merwin [1] argued that turgite was, instead, a mixture of hematite and goethite and Tochilin [2] characterized turgite as a mixture of goethite, ‘hydrogoethite’ and ‘hydrohematite’.

In some cases, the iridescence is associated with dense, crystalline hematite; in other cases, it is associated with the surface of hard, botryoidal hematite or goethite. In the case of a ‘rainbow hematite’ currently in commerce, it is associated with a friable polycrystalline hematite that consists of a mass of weakly cemented hematite crystals that range from 10’s to 100’s of µm in width and up to a few tens of micrometers in thickness.

Numerous localities where iridescent hematite occurs are documented, and specimens from many more localities have been offered for sale in commerce or illustrated on the internet. A sampling of significant reported localities includes Bisbee, Arizona; Durango, Mexico; Cumberland, England; Diamond Hill, Rhode Island; Elba, Italy; Goldfield, Nevada; and Minas Gerais, Brazil.

The color of some of these materials has been thought to be due to a thin coating, but the detailed examination of the coating was first presented examining the Brazilian rainbow hematite [3,4] with the conclusion that the coating could be an aluminum phosphate. The Brazilian material was more recently characterized further, leading to the conclusion that rather than a coating, the entire sample was made up of layers of hematite with incorporated aluminum and phosphorous [5]. It is the purpose of this work to further characterize this highly visible and, at times, beautiful phenomenon from a variety of localities.

2. Experimental Details

Nine samples investigated in this study (Table 1) were chosen for obvious iridescence on the surface of crystals or for what appear to be obvious, strongly colored surfaces of the sample. All of them came labeled as hematite except for CIT-12889, which was presented as hematite with turgite. The colors of the iridescent films are often reminiscent of the first- and second-order interference colors of oil slicks on water. Several of the samples are so brightly colored that, initially, they appeared to be coated with a uniformly colored phase over dimensions of several centimeters. However, for all these samples, regions could be found where the colors showed gradational change into different colors on the surface of the sample.

A ZEISS 1550VP field emission scanning electron microscope (FE-SEM) (ZEISS Group, Oberkochen, Germany) with an in-lens detector was used for high-resolution secondary electron (SE) imaging. Low voltage (<5 kV) or 10 kV accelerating voltage imaging was performed on uncoated samples and carbon-coated samples. Low-voltage SE imaging is important for studying the morphology of nanocrystals due to the smaller depth of beam penetration [4]. An Oxford INCA 300 X-ray energy dispersive spectrometer (EDS) (Oxford Instruments, High Wycombe, UK) on the FESEM was used for elemental analysis at 10 kV in spot mode. Quantitative analyses were made with the XPP matrix correction procedure [6]. Electron backscatter diffraction (EBSD) analysis of in-situ crystals at a nanometer scale was attempted using an HKL EBSD system on the FESEM, operated at 20 kV and 0.3 to 1 nA in a focused beam with a 70° tilted stage and at 5 to 10 mm working distance. Electron probe microanalyses were obtained with a JEOL 8200 instrument (JEOL Ltd., Tokyo, Japan) at 15 keV and 30 nA with a 20 μm beam defocus. Standards used were hematite (Fe Ka), anorthite (Al Ka) and apatite (P Ka). The CITZAF matrix correction [7] was employed for quantitative analyses.

X-ray fluorescence (XRF) analyses of the samples were carried out with an INAM Expert 3L XRF unit with a Ti target using 60 s scans and standardless quantitative analyses with the INAM software.

Raman spectra were obtained with a Renishaw M1000 system with 3 mW or less of 514.5 nm laser power at the sample through a 50× lens. Infrared spectra were conducted with both a Nicolet Magna 860 FTIR with a Continµum infrared microscope and with a Themo-Nicolet iS50 FTIR with a SensIR DuroScope attenuated total reflection (ATR) accessory. Spectra were collected with the micro-ATR attachment and by direct transmission through individual coated hematite crystals. To reduce interference fringes, spectra of some crystals were obtained by placing the crystal on a thin KBr plate and coating the crystal with Halocarbon Products Corp. fluorocarbon series 1.8 oil, which is transparent in the 4000–2000 cm⁻¹ region.

Other analytical methods that failed to produce satisfactory results were investigated. SIMS analyses were obtained on a Cameca 50L Nano SIMS. Standards were samples of synthetic berlinite (AlPO₄), augelite (Al₂PO₄(OH)₃) and trolleite (Al₄(PO₄)₃(OH)₃). TEM studies were conducted using a Philips EM430 STEM operated at 300 kV. Finely ground samples were dispersed in ethanol and loaded on a holey carbon TEM grid. Particular effort was devoted to obtaining ring patterns from the coating material but failed.

To test the possibility that the phosphate is present not as an aluminum phosphate phase, but rather as an absorbate on the surface of an aluminum compound, a series of exchange tests were conducted. Aliquots of a dark blue portion of sample 1462 were held at 35 °C for 92 h in solutions of 0.1 M K₂HPO₂, 0.1 M KH₂PO₂, 1.0 M K₂HPO₄, 1 M HNO₃ and pure water. An additional sample was held at 98 °C for 30 min in 1.5 M HNO₃ to test both the exchange and solubility of the nanocrystals.

3. Results

3.1. Visual Properties

Microscopic examination of sample 1462 shows occasional places where the irridescence is missing (Figure 2) and the specular reflectance of the underlying hematite shows through. This observation is consistent with the proposal that the color on sample 1462 comes from a coating and is not something that carries through the bulk of the sample.

3.2. Field Emission SEM-EDS Studies

Low-magnification images show that individual crystals of hematite sample 1462 are coated with a thin film of material (Figure 3). EDS analysis shows that the coating on the sample contains aluminum and phosphorous. Accurate elemental analyses by this method are difficult because the electron beam penetrates the thin film and interrogates a significant amount of the material underlying the thin film. High-resolution secondary electron imaging reveals that the iridescent portion of sample 1462 has a surface film that, based on their external morphology and uniform appearance over large areas, is comprised of rod-shaped nanocrystals arranged in three directions (120° apart), as shown in Figure 3c.

High-resolution imaging and EDS analyses reveal that most of the other iridescent specimens also have a surface film comprised of similar rod-shaped nanocrystals arranged in three directions (120° apart), such as sample CIT-11952, shown in Figure 4. Again, in this sample and most of the others, Al and P are found in the coating of phosphorous nanocrystals (possibly present as aluminum phosphate) and absent in the area between the nanocrystals. Among these samples, the ratio of Al to P ranges from 2.21 to 3.47. Higher Al to P ratios (6.9 to 7.1) were observed with the Pico Mine sample (2377), which also had higher Al and Si contents expressed in the XRF analyses. The primary exception was sample 2380 from Quartz Mountain, Oregon (Figure 5), which has nanocrystals containing Al and As coating the sample (possibly present as aluminum arsenate). The composition has Al and As in the atomic ratio of 1.98 to 3.03, with no P substitution for As.

The rod-shaped crystals typically have a diameter from 5 to 35 nm. The thickness of the surface coating typically ranges from 200 nm to 500 nm. They are particularly well developed on hematite samples comprised of aggregates of individual, flat crystals.

Samples with botryodial surfaces (e.g., samples 2378, 12889, 1960) also have the nanocrystals on their surface. In these cases, the crystals are not as well ordered as in the case of the flat crystals, and they may form thicker, six-rayed clusters (Figure 6).

The three comparison samples that had no evidence of iridescent colors showed no evidence for crystals or a foreign phase on the surface of the hematite. In particular, in contrast with samples that have iridescence, they did not display higher concentrations of either Al or P.

3.3. Electron Probe Studies

Electron probe microanalyses were obtained on eleven grains of 1462 consisting of 35 distinct analytical points. All analyses of the surface layer were dominated by iron because the volume of excitation greatly exceeds the thickness of the coating. Al and P were in the 1 to 3% range (as oxides). Other elements, such as Si, were absent. In a region of a hematite crystal with no visible coating, the Al₂O₃ values fell to 0.1% and the P₂O₅ values fell to 0.03% or less. The Al to P ratio is relatively constant for grains taken from a single hand specimen. For example, one sample has Al to P ratios that vary from 2.8 to 3.1. Individual colored zones from that sample vary slightly in the ratio: green (3.1), blue (3.1), deep blue (3.0), yellow-red (2.9). The Al to P ratios do vary from hand specimen to hand specimen. For example, grains from a different sample have Al to P ratios that range from 3.7 to 4.0. These ratios compare to 3.02, the grand average of all SEM-EDS analyses of 1462.

3.4. EBSD and TEM Studies

Several attempts to obtain EBSD analyses failed to capture a diffraction pattern, most likely due to the small size of the nanocrystals. We do not have evidence that the coating material is beam sensitive in the SEM. However, from the EBSD analyses, it is possible to determine that when the nanocrystals are arranged on the (0001) surface of hematite, they are arranged in three directions (Figure 4), as revealed by EBSD analysis on sample CIT-11952, where the nanorods (120° apart) are parallel to the [10$\overline{1}$0], [$\overline{1}$100] and [0$\overline{1}$10] directions (i.e., the directions on the (0001) plane normal to {10$\overline{1}$0} of hematite). Electron diffraction patterns from the coating material were not obtained under TEM. All attempts to obtain ring diffraction patterns failed. The samples were sensitive to rapid damage in the TEM’s electron beam.

3.5. X-Ray Fluorescence Analysis

Nine samples were examined with an INAM Expert 3L XRF unit. If the iridescent portion of the sample could not be placed directly over the detection unit, the iridescent material was either scraped off or a small piece about 1 mm thick of the iridescent portion of the sample was broken off and examined with the iridescent portion in direct contact with the detection unit. The area analyzed was about 2.5 mm in diameter but again, the X-ray beam interrogated an area more than a mm into the bulk sample. The results appear in Table 2.

3.6. Vibrational Spectroscopy

3.6.1. Raman Spectroscopy

A spectrum obtained on the bare hematite portion of sample 1462 gave an excellent hematite pattern. When the hematite pattern was subtracted from spectra of the coated portion, in addition to hematite features, this revealed only one clear new feature at 92 cm⁻¹ and a possible very weak band near 309 cm⁻¹ that overlapped with hematite features. The phases producing these features remain unidentified. No features from either carbonate, phosphate or hydroxide were identified in the Raman spectra. Spectra of iridescent portions of samples 1960, 2378, 2380 and 12889 produced hematite patterns. Better results were obtained from the infrared ATR spectra.

3.6.2. Infrared Spectra

All infrared transmission spectra of individual hematite crystals of sample 1462 are dominated by an interference pattern that arose from the high refractive index of the thin hematite crystals. Spectra run under oil reduced the interference pattern to about half of its original intensity but did not eliminate it. Many, but not all, of the spectra showed a broad band under the interference fringes that maximizes at about 3160 cm⁻¹ (Figure 7). When the average height of each interference band is plotted, the spectral pattern suggests that there are two overlapping bands, the stronger one at 3160 cm⁻¹ and the other at about 3495 cm⁻¹. These are in the region where OH and water typically absorb. This is also a region where some AlO(OH) polymorphs absorb and most resemble the OH spectrum of boehmite. The spacings of the interference fringes in Figure 7, if derived from hematite, correspond to a thickness of 22 μm, assuming n_λ of 2.2 in the infrared OH region [8], which is in the range of the measured thickness (7 to 25 µm) of typical crystals of sample 1462.

Unfortunately, due to the interference fringes, none of the spectra produced definitive results that would unambiguously demonstrate that OH was a part of the nanocrystals. The attempts to use micro-ATR methods on sample 1462 failed to produce a spectrum that differed from hematite.

ATR spectra were obtained on all the iridescent samples by either scraping or breaking off a thin portion of the samples with iridescence and grinding them to a powder for placement on the ATR apparatus. Inevitably, some of the underlying material was part of the sampling. All the spectra were dominated by hematite features, with the exception of 12889, which had a large goethite component. Samples 1960, 2378 and 2522 had minor amounts of goethite in the pattern and the sample had 2377 traces of goethite. Samples with thicker amounts of the fibrous material in the SEM tended to show significant absorption in the region where phosphate ions absorb. Figure 8 shows the absorption of two phosphate standards compared to an iridescent sample. Figure 9 shows the spectra of samples with obvious phosphate features in the 1000–1050 cm⁻¹ region. Samples that displayed only thin coatings of the fibrous Al-P material tended to provide little or no indication of phosphate in the infrared spectra. Sample 2380 with arsenate was devoid of absorption in the 1000–1050 cm⁻¹ phosphate region.

3.7. SIMS

The atomic ratio of Al to P was determined for the calibration standards and on three grains of sample 1462. The SIMS analysis confirmed that both Al and P were present in the coating, but gave both higher and variable Al to P ratios (8–14) than were obtained from either the SEM-EDS analyses of the SIMS sample or the electron probe and SEM-EDS analyses of other samples of 1462. This is most likely due to differences between the sample (low density of packing of the nanocrystals) and the standards (dense crystalline material) and how they formed a plasma and were ionized. An SEM-EDS Al/P value of 3.45 was obtained on the sample of 1462 used for the SIMS analysis.

3.8. Additional Description of Samples

Nanorods are not found on sample 2522 (Figure 10). This sample has brilliant colors on its surface that were particularly rich in the deep blue region. SEM and EDS analysis show that there are sub-micrometer thick layers of different composition on the very surface of the sample (Figure 11). In addition to the dominant Fe and O, EDS analysis of the surface layer reveals near-percent amounts of Al and Si, as well as minor amounts of Cu. EDS analysis of the underlying hematite gives only O and Fe. Goethite occurs in some of the valleys (Figure 12). A few micrometer-sized fluorapatite crystals were also found.

The crystals on the surface of the Pico Mine sample (2377) displayed oriented crystals approaching 400 nm in length, but also appeared to grade into stacks of plates that were about 15 nm thick and about 350 nm wide (Figure 13). This is the sample with the highest Al to P ratio measured in this study.

3.9. Anion Exchange Tests

The appearance of coatings does not change after immersion for 10 min in concentrated hydrochloric acid, or by 30 min in 98 °C, 1.5 M nitric acid. Likewise, samples recovered from the 92 h, 35 °C experiments with potassium phosphates and diluted acids showed no visually obvious change in color. The SEM images of these samples showed that the nanocrystals were still intact in all the samples.

The Al content of the sample heated in 1.5 M HNO₃ for 30 min was lower (0.66 wt% Al₂O₃ and 0.22 wt% P₂O₅) than the other samples from this locality that were analyzed in the electron probe (typically 2.5–3.0 and 1.1–1.5 wt% respectively), and the Al to P ratio was higher (4.2). The color was not visibly changed by the acid treatment.

4. Discussion

4.1. Origin of Color

Several observations indicate that on several of the samples, the color arises from interference effects. The color smoothly varies on the surface of some samples, matching the way the color of an oil film varies when on water on a pavement. The color on a particular spot of a sample will vary somewhat depending on the angle of reflection. The layers of nanocrystals form thin films at a sub-micrometer scale that are assumed to have an index of refraction that differs from hematite. In other mineral systems, crystals of sub-micrometer thickness, non-periodic layers of crystals and compositional differences at the sub-micrometer scale have been shown to cause interference colors [9,10]. Thus, we conclude that the films cause interference colors in the iridescent samples. We note that in a previous study, Lin et al. [5] concluded that the Brazilian rainbow hematite from the Andrade Mine did not come from a coating on hematite crystals, but rather was entirely composed of layers of nanorods of an Al,P-containing hematite. Likewise, Heany et al. [11] concluded that the origin of color in iridescent turgite was from quasi-periodic void layers in the botryoidal goethite. We were not able to confirm these conclusions in any of our samples.

4.2. Mineralogy

The Al to P ratio of about 3 to 1 observed for the thin film on iridescent hematite and the presence of phosphate absorption in the infrared raises the possibility that the nano crystals are a distinct aluminum phosphate crystalline phase. We note that all currently known aluminum phosphates with this Al to P ratio contain water in their formulas. Our spectroscopic studies did not provide definitive evidence for water of hydration and the stability of the phase in the SEM further makes it unlikely that it is hydrous. One phase that has an Al to P ratio of 3 to 1 is evansite, Al₃(PO₄)(OH)₆·6H₂O, but the infrared pattern of the iridescent samples in the phosphate region is not a good match to that of evansite (Figure 8). Thus, we are unable to associate this phase with a currently known phosphate mineral phase.

There is no obvious morphological difference between the crystals in the coatings containing aluminum and phosphate and the crystals containing aluminum and arsenate. In each case, the atomic ratios of Al/P and Al/As are generally of the same order. All our attempts to identify the nanocrystal phase have failed. The variable ratio of Al to P (or As) raises the possibility that the nanocrystals may not be a single aluminum phosphate (or arsenate) phase. Lin et al. [5] concluded that Al in the needles was incorporated into the crystal structure of hematite rather than in a distinct phosphate phase. They proposed that the nanocrystals are hematite substituted with Al and possible P with a stoichiometry of Fe_1.81Al_0.23P_0.03O₃. They recognized that the phosphorous could occur as a minor impurity either in the hematite structure or as atoms attached to the surfaces of the nanocrystals. A study of the Raman spectra of aluminum-substituted hematite [12] showed that there were small band shifts and weak new features that appeared in such material, but we were unable to conclusively find such features in the spectra of our samples.

Infrared spectra do not provide reproducible evidence of either OH or H₂O in the coating. Nevertheless, apparent bands in the OH region were seen often enough to provide an estimate of the upper limit of OH (or H₂O) concentration. From the infrared spectrum of Figure 7, an estimate of the water (or OH) content can be made. Assuming a density of 3.2 (intermediate between diaspore and boehmite) and a coating thickness of 300 nm on both sides of the hematite crystals, the absorbance in the OH region of about 0.042 gives about 2 wt% H₂O, assuming a molar absorption coefficient of 100 for the OH region of the nanocrystals. Although there are many uncertainties with such a calculation, the calculation does not suggest that a stoichiometric hydroxyl phase is in the coating (compared to diaspore, which has about 15 wt% H₂O or “evansite” with 42%).

4.3. Genesis

The development of the nanocrystals appears to be related to fluids that flow through the hematite. The color is developed on exposed surfaces, and on cracks or fractures in the hematite, or within highly porous samples such as 1462. The orientation of the nanocrystals suggest that the nucleation of the crystals is templated by the hexagonal hematite structure on (0001) along the [10$\overline{1}$0], [$\overline{1}$100] and [0$\overline{1}$10] directions.

Not all iridescent hematite samples show the nanocrystals that contain aluminum and phosphate (or arsenate) in SEM images. Three iridescent samples did not show these crystals in SEM images but, instead, were coated with other substances. In the EDS analysis, sample 12889 (Arizona) contained Na, Al, silicate and S in the approximate ratio of 0.2:0.3:1.0:0.1. Sample 14296 (Elba) contained Na, Al, silicate and S in the approximate ratio 0.4/0.9/1.0/0.2. Sample 2522 (Inyo) contained Cu, Al and silicate in highly variable ratios with minor amounts of P, S, As, Ca (e.g., Cu:Al:Si:P:S:Ca:As = 0.3:0.3:1:2.8:0.2:0.1:0.2).

In general, iridescent hematite has thin (200 to 500 nm) layers on the surfaces of the material. In our studies, coatings of nanocrystals of a phase containing aluminum and phosphate were most commonly observed. The ubiquity of these coatings suggests that the formation of these nano-phases is a thermodynamically favorable process. These coatings also represent a source of phosphorous in iron ores that could have deleterious effects on the final product and may need to be removed from the ore. Some samples contained goethite in the iridescent layer, consistent with the concept of iridescent ‘turgite’. Even those contain notable amounts of phosphate in the iridescent region and some displayed crystals with a similar morphology to the coatings on hematite crystals.

Author Contributions

G.R.R. conducted the spectroscopic and wet-chemistry experiments. C.M. conducted the electron-beam experiments. Both authors contributed figures and contributed to the drafting of the manuscript. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The authors confirm that the data supporting the findings of this study are available within the article. Additional spectra and electron microscope image are available from the authors.

Acknowledgments

Rock H. Currier (Jewel Tunnel Imports) drew the authors’ attention to this problem and provided numerous research samples. Robert Housley, Benny Fenn, PA Schroeder and Paul M Adams provided additional samples. Mineracoes Brasileiras Reunidas S.A. is thanked for permission to collect samples at their mines.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. An assortment of colorful iridescent samples: 1462: Mina do Andrade, Brazil; 1960: El Salvador Mine, Mexico; 2377: Pico Mine, Brazil; 2378: Graves Mountain, Georgia; 2380: Quartz Mountain, Oregon; 2252: Whittaker Mine, Inyo County; 11952: Alaska; 12889: Dequesne, Arizona; 14296: Elba Island, Italy. Illustrated regions range from 3 to 9 cm wide.

Figure 2. Surface of Brazilian rainbow hematite showing both the iridescent colors and small areas where the iridescent color is absent in areas where the coating appears to be missing.

Figure 3. (a) Low magnification SEM image of sample 1462 showing hematite crystals covered with a thin layer of the aluminum–phosphorous-containing phase. The crystal region in the center has an area on the top side where the coating is missing. (b) EDS X-ray analyses of the coated region showing Al and P and of the area missing the coating showing the absence of Al and P. (c) Higher magnification SEM images of the boundary (a) where the aluminum–phosphorous-containing coating is adjacent to an area where only hematite is seen. (d) A high-resolution image of the rod-like crystals oriented at 120 degrees.

Figure 4. High magnification SEM images of the partially developed layer of aluminum phosphate nanocrystals coating a single hematite crystal on a blue region of sample CIT-11952.

Figure 5. SEM image of nanocrystals containing both aluminum and arsenic on the sample from Quartz Mountain, Oregon (2380). (a) Lower magnification. (b) Green portion at high magnification.

Figure 5. SEM image of nanocrystals containing both aluminum and arsenic on the sample from Quartz Mountain, Oregon (2380). (a) Lower magnification. (b) Green portion at high magnification.

Figure 6. SE images of the yellow region on an iridescent sample from Graves Mountain, Georgia (2378). (a) Low magnification view of a fragment of the coating; (b) high magnification view of the nanocrystals.

Figure 6. SE images of the yellow region on an iridescent sample from Graves Mountain, Georgia (2378). (a) Low magnification view of a fragment of the coating; (b) high magnification view of the nanocrystals.

Figure 7. Transmission infrared pattern of a single flake of deep blue, iridescent hematite (1462) that shows apparent broadband in the 3600–2800 cm−1 region admixed with strong interference fringes.

Figure 8. ATR infrared spectra of two phosphate standards, evansite (Al3(PO4)(OH)6·6H2O) and wavellite (Al3(PO4)2(OH)3·5H2O), compared to the phosphate region in iridescent sample 1960.

Figure 9. Expanded views of the ATR spectra of samples showing absorption in the phosphate ion region (950–1100 cm−1) and goethite features.

Figure 10. Optical photograph of Inyo County sample 2522 showing a region with multiple colors.

Figure 11. SEM image of the surface material on Inyo sample 2522 in a blue region showing material that analyzed as dominantly Fe and O plus minor Al and Si.

Figure 12. SEM image of the fibrous goethite in the valleys in a blue-green region of the surface material on Inyo sample 2522.

Figure 13. Crystals and plates on the surface of a green-colored region of hematite from the Pico Mine (2377).

Table 1

Iridescent Samples Examined.

Sample	Locality	Description
GRR 1462	Mina do Andrade, João Monlevade, Minas Gerais, Brazil	brilliant, variably colored polycrystalline, friable, schist-like material
GRR 1960	El Salvador Mine, Sierra Mojada, Mexico	deep blue and green coating on a massive, fine-grained sample
GRR 2377	Pico Mine, Nova Lima, Minas Gerais, Brazil	thin coloration on a fracture surface of a hand-sized sample
GRR 2378	Graves Mountain, Lincoln County, GA, USA	intensely colored hand specimen with a brilliant play of colors
GRR 2380	Quartz Mountain, Douglas County, OR, USA	variably colored coating with strong greens
GRR 2522	Whittaker Iron Mine, Inyo County, CA, USA	strongly colored coating in a variety of colors
CIT-11952	Unspecified locality in Alaska, AK, USA	thinly coated sample
CIT-12889	Dequesne, Santa Cruz County, AZ, USA	moderately bright primarily green but with smaller regions of other colors
CIT-14296	Rio Marina, Elba Island, Livorno Province, Italy	thinly coated 8 to 12 mm crystals mostly lavender to blue color
Samples with no evidence of iridescent coatings
CIT-2051	Bouse, La Paz County, AZ, USA	5 mm crystals, specularite variety
CIT-8820	Broughton-in-Furness, Cumbria, UK	well-crystallized, 3–8 mm flakes on quartz
CIT-10446	Cherry Mine, Ishpeming, MI, USA	polycrystalline, approximately 1 mm flakes, specularite variety.

Table 2

X-Ray fluorescence analysis of the area near the iridescent portion of the sample.

	1462	1960	2377	2378	2380	2522	11952	12889	14296
	Brazil	Mexico	Pico	Georgia	Oregon	Inyo	Alaska	Arizona	Italy
O	30.6	31.4	30.7	34.4	31.0	31.0	30.9	30.9	30.6
Mg	0.73	0.8	-	-	0.7	1.72	-	-	-
Al	0.8	0.13	0.72	3.67	0.93	-	0.38	0.28	0.37
Si	0.11	0.46	0.63	5.65	1.25	-	1.18	1.41	0.69
P	0.17	0.034	0.053	0.37	0.021	0.021	0.039	0.048	-
S	0.01	0.13	0.057	0.3	1.25	0.032	0.18	0.096	0.11
Ti	-	-	0.073	-	-	0.019	-		-
V	-	-	-	-	-	-	-	0.039	-
Mn	0.23	4.77	0.34	0.71	0.68	0.74	0.25	0.097	0.54
Fe	67.32	62.23	67.39	54.09	63.15	66.30	66.93	64.93	67.51
Cu	-	-	-	0.008	0.004	0.069	0.031	0.13	-
Zn	-	0.19	0.004	-	0.008	0.041	0.014	0.26	0.008
Ga	-	0.003	-	0.004	0.003	-	-	-	-
Ge	-	0.01	-	-	-	-	-	-	-
As	-	0.2	0.001	0.005	1.93	0.022	0.007	0.33	0.004
Se	-	-	-	0.006	-	-	-	-	-
Sr	-	0.003	0.007	0.42	0.005	-	-	0.003	-
Y	-	-	0.006	0.006	0.004	-	-	-	-
Zr	-	-	0.004	0.002	0.112	-	-	-	-
Nb	-	-	-	-	0.008	-	-	-	-
Mo	-	0.006	-	-	-	-	-	0.21	-
Sn	-	-	-	0.006	0.021	-	-	-	-
Ba	-	-	-	0.3	0.08	-	-	-	-
W	-	-	-	-	-	-	-	-	0.12
Hg	-	-	-	-	-	0.012	-	-	-
Pb	-	-	-	-	-	-	-	0.44	-

References

1. Posnjak, E.; Merwin, H.E. The hydrated ferric oxides. Am. J. Sci.; 1919; 47, pp. 311-348. [DOI: https://dx.doi.org/10.2475/ajs.s4-47.281.311]

2. Tochilin, M.S. Existence of turgite. Dokl. Akad. Nauk. SSSR; 1951; 81, pp. 75-76.

3. Ma, C.; Rossman, G. Low voltage FESEM of geological materials. Microsc. Microanal. Microanal.; 2003; 9, pp. 990-991. [DOI: https://dx.doi.org/10.1017/S1431927603444954]

4. Ma, C.; Rossman, G.R. Low voltage FESEM of geological materials. Microsc. Today; 2006; 14, pp. 20-22. [DOI: https://dx.doi.org/10.1017/S1551929500055140]

5. Lin, X.; Heaney, P.J.; Post, J.E. Iridescence in metamorphic “rainbow” hematite. Gems Gemol.; 2018; 54, pp. 28-39.

6. Pouchou, J.L.; Pichoir, F. Quantitative analysis of homogeneous or stratified microvolumes applying the model “PAP”. Electron Probe Quantitation; Heinrich, K.F.J.; Newbury, D.E. Plenum Press: New York, NY, USA, 1991.

7. Armstrong, J.T. CITZAF: A package of correction programs for the quantitative electron microbeam X-ray analysis of thick polished materials, thin films, and particles. Microbeam Anal.; 1995; 4, pp. 177-200.

8. Glotch, T.D.; Christensen, P.R.; Sharp, T.G. Fresnel modeling of hematite crystal surfaces and application to Martian hematite spherules. Icarus; 2006; 181, pp. 408-418. [DOI: https://dx.doi.org/10.1016/j.icarus.2005.11.020]

9. Ma, C.; Gresh, J.; Rossman, G.R.; Ulmer, G.C.; Vicenzi, E.P. Micro-analytical study of the optical properties of Mexican rainbow and sheen obsidians. Can. Mineral.; 2001; 39, pp. 57-71. [DOI: https://dx.doi.org/10.2113/gscanmin.39.1.57]

10. Ma, C.; Rossman, G.R.; Miller, J.A. The origin of color in “fire” obsidian. Can. Mineral.; 2007; 45, pp. 551-557. [DOI: https://dx.doi.org/10.2113/gscanmin.45.3.551]

11. Heaney, P.J.; Post, J.E.; Chen, S.A.; Clark, T.; Wenzel, T.; Jacucci, G.; Vignolini, S. Painting a Rainbow with Ochre; Iridescence in Botryoidal Goethite. Proceedings of the GSA Annual Meeting; Indianapolis, IN, USA, 4–7 November 2018; Paper No. 282-12

12. Zoppi, A.; Lofrumento, C.; Castellucci, E.M.; Sciau, P. Al-for-Fe substitution in hematite: The effect of low Al concentrations in the Raman spectrum of Fe₂O₃. J. Raman Spectrosc.; 2008; 39, pp. 40-46. [DOI: https://dx.doi.org/10.1002/jrs.1811]