SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (46)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.
Soil Science Society of America Journal 63:1463-1471 (1999)
© 1999 Soil Science Society of America

DIVISION S-9-SOIL MINERALOGY

Color Identification of Iron Oxides and Hydroxysulfates

Use and Limitations

A.C. Scheinost{dagger},a and U. Schwertmanna

a Lehrstuhl für Bodenkunde, TU München-Weihenstephan, 85350 Freising, Germany

scheinos{at}udel.edu


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 
Colors are widely used to describe soil profiles but seldom to identify minerals. This study was conducted to determine whether color measurements with a colorimeter could be employed to identify Fe oxides and hydroxysulfates, either as monomineralic samples or in soils. We measured the color of 277 Fe oxide and hydroxysulfate mineral samples, as well as 309 soils and soil fractions. Discriminant analysis was employed to separate the minerals in three different color systems: CIE-Yxy, CIE-L*a*b*, and Munsell. Using the color of monomineralic samples, 100% of maghemites and jarosites could be correctly classified, 95% of goethites, 90% of feroxyhites, 84% of lepidocrocites, 83% of hematites, 56% of ferrihydrites, 50% of akaganéites, and 44% of schwertmannites. In soil samples, goethite (90% correct classifications) could be reliably distinguished from hematite–goethite mixtures (82% correct classifications), and from lepidocrocite–goethite mixtures (89%). The identification of ferrihydrites, akaganéites, and schwertmannites largely failed because of similar average colors and because of high color variability. Because the color variability is a mineral-intrinsic property rather than produced by errors of measurement, color is in principle unsuited for identifying ferrihydrites, akaganéites, and schwertmannites. However, color may be used to identify goethite, hematite, lepidocrocite, jarosite, maghemite, and feroxyhite with a relatively high reliability. The choice of color system is not crucial for this type of analysis.

Abbreviations: CIE, Commission Internationale d'Eclairage • DRS, diffuse reflectance spectroscopy • Fed, dithionite-citrate-bicarbonate extractable Fe • Feo, acid oxalate extractable Fe • XRD, x-ray diffraction


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 
SOIL COLOR has long been known to reflect Fe oxide composition and content. An increase in chroma is generally interpreted as increasing Fe oxide content. Likewise, Schwertmann (1993) explained how yellowish, orange, and reddish hues of soils are caused by Fe-bearing minerals including Fe oxides, hydroxides, oxyhydroxides, and hydroxysulfates. Even when organic matter darkens a soil, the yellowish hues of goethite can easily be distinguished from the reddish hues of hematite. Orange hues are often caused by ferrihydrite or lepidocrocite, whose identification are crucial in many soils because they indicate active or recent redox processes. These obvious color differences within soils have resulted in color being a fundamental means to delineate, classify, and sample soil units. The identification and selection of individual soil units is made possible by the eye's ability to discriminate even between small changes in color (Wyszecki and Stiles, 1982).

Qualitative and quantitative analysis of soil Fe oxides is of great pedological interest because Fe oxide content and mineralogy reflect duration and intensity of pedogenesis. In addition, Fe oxides greatly affect the adsorption capacity of soils with respect to oxyanions. Therefore, the knowledge of their spatial variability is crucial, for example, for the risk assessment of selenate and arsenate toxicity (Zhang and Sparks, 1990; Bowell, 1994) or for the site-adapted application of phosphate fertilizers (Torrent, 1987; Van der Zee et al., 1988; Scheinost and Schwertmann, 1995). Standard methods for the mineral analysis, such as x-ray diffraction (XRD) and selective dissolution of Fe oxides, are time-consuming, not very sensitive, and not suited for field evaluations.

The difficulties of the standard methods relative to the ease of identifying soil color has resulted in many attempts to predict the type and concentration of Fe oxides simply from Munsell color (Schwertmann, 1993; Cornell and Schwertmann, 1996). While substantial progress has been made in the identification and quantification of goethite and hematite in soils (Torrent et al., 1983; Barrón and Torrent, 1986; Scheinost and Schwertmann, 1995), much less is known about the color variability of other Fe minerals, and whether they can be reliably distinguished by color. Schwertmann and Lentze (1966) determined frequency distributions of the Munsell coordinates of 90 synthetic Fe oxides and 174 natural Fe oxide samples. The goethites were clearly separated from the hematites by hue, while goethites, lepidocrocites, and ferrihydrites had similar (natural samples) or overlapping hues (synthetic samples). The greater similarity of the natural samples was explained by the influence of other coloring soil components such as organic matter and Mn oxides.

Most studies on soil color employed the Munsell color system because of its ease of use, ready availability, and historical importance to soil science (Simonson, 1993). However, newer systems, like those created by the Commission Internationale d'Eclairage (CIE), may be better suited to represent a uniform color space, that is, a color space where the Euclidean differences between colors are equivalent to the human perception of those differences (Melville and Atkinson, 1985). Another advantage of these new systems is that they overcome the limitations of the Munsell color system's discrete color chips. Those limitations include:

1. The resolution of color assignments is limited by the resolution of the chips. Interpolation between chips is possible in principle, but difficult in practice as it has to be done in a three-dimensional space.

2. Due to metamerism, the assignment of a color may depend on the light source that is often not standardized, and on the physiological properties of the observer's color receptors.

3. Differences in gloss and other surface properties between the chips and the soil may influence the assignment.

These three factors cause a substantial scattering of color coordinates when the colors are determined repeatedly by one observer, or by several observers (Post et al., 1993). The limitations can be eliminated by using a colorimeter, which is an instrument capable of providing both accurate and precise color data (Torrent and Barrón, 1993). Another alternative is to calculate the color from diffuse reflectance spectra (Fernandez and Schulze, 1987); however, compared with this method, measurements with a colorimeter are faster, cheaper, and can be carried out in the field.

The objective of this study was to expand the work of Schwertmann and Lentze (1966), by using colorimeter measurements and a larger sample collection than was previously available. In addition to pure, synthetic or natural, Fe oxide and hydroxysulfate minerals (277 samples), we examined the color of 309 soil samples containing goethite, hematite, or lepidocrocite. For the discrimination of the groups (minerals or soils) we employed stepwise discriminant analysis (Jennrich, 1977). This statistical method allows separation of groups in n-dimensional space (n = number of discriminating variables, here the three color coordinates). The effectiveness of discrimination is given by correct classifications. By exactly quantifying the effectiveness of discrimination, discriminant analysis also allowed evaluation of different color systems.


   Materials and methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results and discussion
 Conclusion
 REFERENCES
 
The 277 monomineralic samples included nine minerals: the oxides hematite and maghemite; the oxyhydroxides goethite, lepidocrocite, akaganéite, feroxyhite, and ferrihydrite; the oxyhydroxysulfate schwertmannite, and the hydroxysulfate jarosite (Table 1) . Common features of these minerals are their yellow to red color caused by electronic transitions between the 3d orbitals of Fe(III) (Sherman and Waite, 1985), and their occurrence in soils or surface waters (Cornell and Schwertmann, 1996). The samples were either produced in the laboratory (synthetic) or collected in the field (natural). The purity of the minerals was determined by step-counted XRD analysis. The mineral samples were also characterized by determining dithionite-citrate-bicarbonate extractable Fe (Fed) (Mehra and Jackson, 1960; Holmgren, 1967) and acid oxalate extractable Fe, (Feo) (Schwertmann, 1964). The larger mineral groups — goethite, hematite, lepidocrocite, and ferrihydrite — represented a wide variation in conditions of formation. A full range of Al for Fe substitution was included in the goethite, hematite, lepidocrocite, and maghemite samples (Table 1). The 309 soil samples were separated into three classes according to their Fe oxide mineral composition (Table 2) . The relative mineral contents were determined by XRD using the untreated or the NaOH-treated clay fraction (Kämpf and Schwertmann, 1982a; 1982b), or by Mossbauer spectroscopy (Schwertmann et al., 1982a; 1982b; Murad and Schwertmann, 1988). The absolute Fe oxide contents were estimated from Fed and Feo.


View this table:
[in this window]
[in a new window]
  		Table 1 Synthetic (syn) and natural (nat) Fe oxides and hydroxysulfates used in this study. xAl gives the mole fraction of Al for Fe substitution, Al/(Al+Fe)

 

View this table:
[in this window]
[in a new window]
  		Table 2 Soils and soil fractions used in this study

 
Color of soils and monomineralic samples was measured with a model CR-300 Chroma Meter (Minolta, Osaka, Japan). This colorimeter uses a pulsed Xe light for a stable and uniform illumination of the sample. Three photocells measure the photons diffusely reflected by the sample through filters matching the CIE standard observer spectral response. We calibrated the colorimeter with a standard white plate (Minolta), and used an illumination type C (Commission Internationale d'Eclairage, 1978). In preparation for the color measurement, unfractionated soils, fine soils (<2 mm), and concretions were ground in an agate mortar until the color visually remained constant ({approx}10 min; Torrent and Barrón, 1993). The monomineralic samples and the clay fractions of the soil samples, which had been freeze-dried from suspension, were used without further treatment. For the analysis, the head of the colorimeter (CR-A33a), which is protected by a glass window, was set onto the powdered samples. The pressure against the samples was standardized only by the weight of the head ({approx}500 Pa). Less than 300 mg of sample was sufficient to cover the tube orifice (5 mm in diameter) with a thickness of at least 1 mm of sample. This thickness proved to be infinite with respect to visible radiation. After using the equipment for measuring hundreds of soil samples (not shown in this study), we observed scratches on the glass window protecting the sensor. Color measured through this scratched window deviated by <15% from color measured through a new window. Care was taken, therefore, to monitor a possible change with time. One goethite and one hematite sample were measured daily during the measuring period of 8 wk (n = 37). The color readings of these two samples were highly reproducible with standard deviations of <=0.08 units of hue, <=0.08 units of value, and <=0.12 units of value. No systematic change of color with time was observed.

The digital output of the colorimeter, consisting of the mean tristimulus values X, Y, and Z of three replicated readings, was transmitted to a personal computer through the serial port. The tristimulus values were converted into the coordinates of three color spaces, CIE-Yxy, CIE-L*a*b*, and Munsell. For the conversion into the first two systems, equations were taken from Wyszecki and Stiles (1982). However, no closed equations exist to convert the tristimulus values into Munsell coordinates. The conversion was therefore accomplished by using a computer code (written by F. Billmeyer, Purdue University; see Fernandez and Schulze, 1987). This code utilizes the digitized tables of Wyszecki and Stiles (1982), and linearly interpolates between the discrete colors given there. The stepwise discriminant analysis was performed with the computer program package Statistica (StatSoft, 1996). The a priori classification probabilities of all groups (minerals or soils) were set to equal values to improve the discrimination of small groups (StatSoft, 1996).

Color Spaces
In the CIE-Yxy color space, all colors lie within a plane resembling a shoe sole (Fig. 1) . An arbitrary point B within that plane is defined by the Cartesian coordinates x and y, where the redness increases with x, and the greenness increases with y. The lightness, Y, is perpendicular to x and y out of the plane. An alternative way to define B is by drawing a line through B and the achromaticity point, A. The intersection of that centripetal line with the perimeter gives the dominant wavelength, {lambda}d, (580 nm for point B in Fig. 1). The relative distance between A and the intersection with the perimeter gives the excitation purity, Pe (0.6 for point B). The dominant wavelength is therefore similar to the hue, and Pe is similar to the chroma of the Munsell system.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1 Color designation in the CIE-Yxy system. Colors lie within the solid lines resembling a shoe sole. Colors within that plane may be given either by the Cartesian coordinates x and y, or by the dominant wavelength, {lambda}d (nm), and the excitation purity, Pe. The lightness, Y, is normal to the plane

 
In order to produce color systems with Euclidean distances similar to those perceived by the eye, a wide variety of so-called uniform color spaces have been developed (Hunter and Harold, 1987). We selected two of them, the Munsell system, because of its traditional usage in soil science (Simonson, 1993), and the CIE-L*a*b* system, because it is the most frequently used modern system (Commission Internationale d'Eclairage, 1978; Melville and Atkinson, 1985). A short description of the (among soil scientists) less well-known CIE-L*a*b* system is given in the following. Color planes are defined in this system by the Cartesian axes a* and b*, which coincide at the achromaticity point (Fig. 2A) . The a* axis extends to the complementary colors red (+a*) and green (-a*), and the b* axis extends to the complementary colors yellow (+b*) and blue (-b*). A third axis normal to a* and b* defines the lightness, L*. Alternatively, the Cartesian coordinates a* and b* may be replaced by the polar coordinates H° (hue) and C* (chroma), resulting in a cylindrical coordination system comparable to the standard notation of the Munsell color system (Fig. 2B). The Munsell system, in turn, may be converted into a Cartesian system using the rectangular axes xM and yM instead of hue and chroma (Melville and Atkinson, 1985).



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2 Color designation in the CIE-L*a*b* system. A plane may be defined either (A) by a* and b* in Cartesian notation or (B) by chroma (C*) and hue (H°) in polar coordinates. The lightness axis, L*, is normal to the plane

 
The expression of color in cylindrical coordinates is better suited for its visualization, and is therefore used in the following discussion whenever the color description is of importance. On the other hand, the polar coordinates are unsuited for multivariate statistical analyses, because hue and chroma do not represent Euclidean distances. Thus, only the Cartesian coordinates were used in statistical calculations.


   Results and discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
Conclusion
 REFERENCES
 
Color of Pure Iron Oxides and Hydroxysulfates
The hues of the samples studied extended from 3.5 R (hematite) to 3.6 Y (jarosite) (Fig. 3 , Table 3) . The Munsell value varied between 2.3 (ferrihydrite) and 8.0 (jarosite), the chroma between 1.5 (hematite) and 9.9 (lepidocrocite). Hematite was clearly distinguishable from goethite, because all hematite samples were redder than 4.1 YR, and all goethite samples were yellower than 7.3 YR (Table 3). The jarosites were yellower and higher in value than the goethites. The remaining Fe oxides filled the gap between the average hue of hematite and of goethite, and strongly overlapped with each other, making a distinction solely by hue impossible. The average hues of hematite, ferrihydrite and lepidocrocite were close to the ones given by Schwertmann and Lentze (1966), but the average hue of goethite was 2.5 units yellower. This difference may be explained by the more precise color determination or by the larger sample set of our study.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3 Munsell coordinates of Fe oxides and hydroxysulfates

 

View this table:
[in this window]
[in a new window]
  		Table 3 Munsell colors of the Fe oxides and hydroxysulfates (median and range)

 
Both the red hue and the (correlated) low value of hematite are due to a strong absorption band at 530 nm (Scheinost et al., 1998). This band has been assigned to an electron pair transition fostered by the interaction of neighboring Fe atoms in face-sharing octahedra (Sherman and Waite, 1985). Other Fe oxides have both edge- and corner-sharing octahedra and hence larger Fe to Fe distances (Manceau and Combes, 1988). Consequently, the electron pair transition of these minerals is weaker (as shown for goethite by Scheinost et al., 1999), and the hues are more yellow than that of hematite. Jarosite has only corner-sharing octahedra and the largest Fe to Fe distances (Gaines et al., 1997). Therefore, the reflectance spectra of jarosite lack the electron pair transition band and show only the weak, spin-forbidden, 3d absorption bands of Fe(III) (Buckingham and Sommer, 1983). The resulting weak absorption across the visible range is responsible for the grayish-yellow color of jarosite (Fig. 3). The strong correlation between value and yellowness (r = 0.95 without maghemite, Fig. 3) may be explained by both the stronger electron pair transition bands of hematite and feroxyhite and by the brighter appearance of yellow colors (Hunter and Harold, 1987). Maghemite has a much lower value than would be expected from its hue (and its structure). This low value is caused by a broad Fe2+ to Fe3+ charge transfer band that centers at 1500 nm and drastically reduces the reflectance in the visible range (Scheinost et al., 1998).

Applying the discriminant analysis, the best separation of the mineral samples was achieved in the CIE-Yxy space (79% of all samples correctly classified), followed by the CIE-L*a*b* space (77%), and the (Cartesian) Munsell system (75%) (Table 4) . To check if this ranking was significant, we repeated the calculation for randomly selected data subsets. The classification results using the CIE-Yxy and the CIE-L*a*b* systems were always better than those by the Munsell system; however, the difference between the systems never exceeded 5%, and the distribution of correct classifications among the classes was always similar. This suggests that the choice of the color system is not crucial for the discrimination of the minerals. In the following, results will be shown for the CIE-Yxy system only.


View this table:
[in this window]
[in a new window]
  		Table 4 Comparison of the classification result for monomineralic samples in three color spaces

 
The discrimination was based on three discriminant functions, or roots (Table 5) . The first root achieved 88% of the final discrimination, the second root added another 8%, and the third root the remaining 4% (see cumulative probabilities in Table 5). All three roots were statistically significant. The standardized canonical variables show the relative contributions of the discriminating variables (i.e., the color coordinates) to the roots. The first and most important root is dominated by x and y, indicating that the discrimination result depends largely on these two variables corresponding with hue and chroma in the cylindrical coordination system. The minor importance of Y (or value) is probably due to the high correlation between hue and value (see above).


View this table:
[in this window]
[in a new window]
  		Table 5 Standardized coefficients for the canonical variables of monomineralic samples

 
The dark yellowish brown maghemites and the grayish-yellow jarosites were perfectly classified (Table 6) . Goethite, hematite, lepidocrocite, and feroxyhite were also successfully (though not perfectly) classified. However, ferrihydrite, akaganéite, and schwertmannite samples were only partly separated from each other (Table 6). The dependable discrimination of maghemite and jarosite is in line with their isolated location in the color space (Fig. 3), which, in turn, is due to the unique causes of their color (see above). Hematite samples have the reddest average hue, but the wide range in hue causes a slight overlap with feroxyhite. Goethites, with the yellowest hues (besides jarosite), interfere with the neighboring schwertmannites due to a relatively high variability in hue. The good classification result for the intermediate colors of lepidocrocite and feroxyhite is because of their low variability in hue and value. In contrast, intermediate color together with high variability leads to the poor classification results for ferrihydrite, akaganéite, and schwertmannite. Thus, the discrimination of the minerals with a similar cause of color (weak electron pair transition by edge- and corner-sharing octahedra) depends strongly on their variability.


View this table:
[in this window]
[in a new window]
  		Table 6 Classification result for monomineralic samples based on Yxy color coordinates.{dagger}

 
Color variation within a given mineral form has been explained by several reasons. For example, the color of hematite changed from yellow-red to blue-red as the diameter of crystals increased from 0.1 to 1.0 µm (Schwertmann, 1993). A similar effect is caused by the aggregation of hematite crystals (Torrent and Schwertmann, 1987). Barrón and Torrent (1984) found a positive correlation between Munsell value and the Al substitution of hematite. Goethite and akaganéite crystals appear darker as their size decreases (Schwertmann, 1993). Acicular hematite crystals have a more yellow hue than hexagonal hematite crystals (Hund, 1981). For the goethite samples of our study, relations between morphology and color have been investigated and published (Scheinost et al., 1999). In line with former results, the goethite samples increased in darkness and redness as the needle length decreased from 0.8 to 0.05 µm. The variation in needle length, in turn, was caused by either Al substitution or temperature of crystallization (Scheinost et al., 1999). Parameters like crystal size and shape, particle aggregation, Al substitution, or other impurities may also be responsible for the observed color variability of the other minerals, but have not been investigated. Moreover, transition metal for Fe substitution also greatly affects color by giving rise to additional crystal-field bands (Schwertmann, 1993; Scheinost and Schwertmann, 1997). However, because of their relative rareness in nature, samples with transition metal for Fe substitution were not included in our study, and are not responsible for the variability of the data presented here.

Color of Soils and Soil Fractions
The hue of goethitic soils was close to the hue of synthetic goethites (compare Fig. 3 with Fig. 4 , and Table 3 with Table 7) . Soils containing both hematite and goethite had hues between those of the pure hematite and goethite samples, and soils containing both lepidocrocite and goethite had hues between those of the pure lepidocrocite and goethite minerals. Thus, the hue of soils is largely determined by these three Fe oxides. Value and chroma of soils were substantially lower than those of the pure minerals. This is due to the reduction of value by soil organic matter, and the reduction of chroma by the dilution of the Fe oxide pigments with the soil matrix, which usually consists of minerals of high reflectance (Fernández and Schulze, 1992; Scheinost and Schwertmann, 1995). Nevertheless, the red hematitic soils were darker than goethitic soils in accordance with results for the pure minerals, confirming that not only soil organic matter but also Fe oxide minerals strongly influence the value. Similarly, the high chroma of lepidocrocite-containing soils is in line with the high chroma of pure lepidocrocite. In addition, the significantly higher Fe oxide content of soils with lepidocrocite or hematite may contribute to their higher chroma (Table 2).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 4 Munsell coordinates of soils

 

View this table:
[in this window]
[in a new window]
  		Table 7 Munsell colors of soil samples (median and range)

 
The result of the discriminant analysis for the soil samples is shown in Table 8 . About 90% of the soil samples were correctly classified using the CIE-Yxy coordinates (results were similar for the other two color systems). This discrimination is surprisingly successful if one takes into account that (i) the color of soils is influenced by additional color-bearing and colorless components, and (ii) most of the samples classified as hematite or lepidocrocite soils contain a substantial amount of goethite. The positive result may be explained by the clear separation of goethite and lepidocrocite by chroma, and by the pronounced red shift of soils by even small relative amounts of hematite (Fig. 5) .


View this table:
[in this window]
[in a new window]
  		Table 8 Classification result for soil samples based on CIE-Yxy coordinates.{dagger}

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5 Hue of hematitic soils as a function of the relative hematite content

 
Soils dominated by Fe minerals other than goethite, hematite, and lepidocrocite were not available for this study. Nevertheless, conclusions on the identification of these other Fe oxides in soils may be drawn from the results of the pure minerals. The unique color of jarosite should always allow for a reliable determination of acid sulfate soils (Fanning et al., 1993). In contrast, the unique (among Fe oxides) color of maghemite is similar to that of Mn oxides and soil organic matter, which may prevent the identification of maghemite in soils. The combination of color and simple field tests (e.g., a magnet for detecting maghemite and magnetite, H2O2 for detecting Mn oxides) or the use of diffuse reflectance spectroscopy (DRS) may allow for an unequivocal identification of maghemite. However, there is little chance that the other minerals (schwertmannite, ferrihydrite, akaganéite and feroxyhite) may be detected in soils unless they would occur in substantial quantities and separated from other Fe oxides (e.g., in concretions).

Comparison with Diffuse Reflectance Spectroscopy
The color of minerals is directly related to their wavelength-dependent reflectance of visible light. Therefore, our results are directly comparable with those of a related paper where second-derivative DRS was employed to identify pure Fe oxides and Fe oxides in soils (Scheinost et al., 1998). The mineral samples were better discriminated by color coordinates than by absorption band positions (88 vs. 66% correct classifications). Furthermore, soils containing lepidocrocite and goethite could be reliably discriminated from goethitic soils by color, while the absorption band positions did not allow for discriminating them. That means that color coordinates are more sensitive to small spectral changes than band positions. Only hematite and maghemite have band positions that are clearly separated from those of the other Fe oxides. Therefore, second-derivative DRS is more sensitive in detecting these two Fe oxides than color. In conclusion, band positions and color together offer a more reliable identification of Fe oxides in soils than one of the parameters alone. Because both parameters can be derived from diffuse reflectance spectra, spectrometers offer improved discriminations compared with colorimeters, but are much more expensive.

Classification of Samples with Unknown Composition
Classification functions derived by discriminant analysis allow classification of unknown samples. For this, one has to multiply the color coordinates of each sample by the factors given in Table 9 (for pure Fe oxides or hydroxysulfates) or Table 10 (for soils). The sum of each row (including the constant) will give the scores for each class. The unknown sample is most probably the mineral or soil which yields the highest scores. However, one has to keep in mind the limitations of the method as revealed by the correct classifications (Table 6, Table 8).


View this table:
[in this window]
[in a new window]
  		Table 9 Coefficients of the classification functions for monomineralic samples, based on CIE-Yxy coordinates

 

View this table:
[in this window]
[in a new window]
  		Table 10 Coefficients of the classification functions for soil samples, based on CIE-Yxy coordinates

 

   Conclusion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
REFERENCES
 
Goethite, akaganéite, and hematite samples are clearly separated by hue. Maghemites and jarosites can be distinguished by their extreme values, and lepidocrocites mainly by their high chromas. The other minerals with intermediate hues, values, and chromas — namely feroxyhite, ferrihydrite and schwertmannite — cannot be reliably identified by their color. Their discrimination is restricted, on one hand, by the similarity in color, which can be explained by similar Fe to Fe distances in the crystal structure. On the other hand, these minerals have a large color variability due to variation in particle size and shape, particle aggregation, crystal defects, or impurities, which further inhibits their classification based on color. Soils containing hematite or lepidocrocite can be reliably discriminated from soils containing only goethite, even when their Fe oxide composition is dominated by goethite. In addition, jarosite, feroxyhite, and maghemite may be identified in soils. However, it is unlikely that color can be used to identify ferrihydrite, akaganéite, and schwertmannite in soils.


   ACKNOWLEDGMENTS
 
We are deeply indebted to J.M. Bigham (Ohio State Univ.), D. Laird (USDA-ARS), and D.G. Schulze (Purdue Univ.) for their help with improving a first version of the manuscript. D.G. Schulze provided the computer code to calculate Munsell colors. This study was funded by the Deutsche Forschungsgemeinschaft.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 
{dagger} A.C. Scheinost current address: Dep. of Plant and Soil Sciences, Univ. of Delaware, Newark, DE 19717. Back

Received for publication July 30, 1998.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
N. Richter, T. Jarmer, S. Chabrillat, C. Oyonarte, P. Hostert, and H. Kaufmann
Free Iron Oxide Determination in Mediterranean Soils using Diffuse Reflectance Spectroscopy
Soil Sci. Soc. Am. J., January 21, 2009; 73(1): 72 - 81.
[Abstract] [Full Text] [PDF]

Home page
 Clay MineralsHome page
H. STANJEK and C. MARCHEL
Linking the redox cycles of Fe oxides and Fe-rich clay minerals: an example from a palaeosol of the Upper Freshwater Molasse
Clay Minerals, March 1, 2008; 43(1): 69 - 82.
[Abstract] [Full Text] [PDF]

Home page
 Eur J MineralHome page
A. Manasse and M. Mellini
Iron (hydr)oxide nanocrystals in raw and burnt sienna pigments
European Journal of Mineralogy, December 1, 2006; 18(6): 845 - 853.
[Abstract] [Full Text] [PDF]

Home page
 Journal of Sedimentary ResearchHome page
M. J. Kraus and S. T. Hasiotis
Significance of Different Modes of Rhizolith Preservation to Interpreting Paleoenvironmental and Paleohydrologic Settings: Examples from Paleogene Paleosols, Bighorn Basin, Wyoming, U.S.A.
Journal of Sedimentary Research, April 1, 2006; 76(4): 633 - 646.
[Abstract] [Full Text] [PDF]

Home page
 Clay MineralsHome page
E. MURAD and P. ROJIK
Iron mineralogy of mine-drainage precipitates as environmental indicators: review of current concepts and a case study from the Sokolov Basin, Czech Republic
Clay Minerals, December 1, 2005; 40(4): 427 - 440.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. J. D'Amore, S. R. Al-Abed, K. G. Scheckel, and J. A. Ryan
Methods for Speciation of Metals in Soils: A Review
J. Environ. Qual., September 8, 2005; 34(5): 1707 - 1745.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
M. P. F. Fontes and I. A. Carvalho Jr.
Color Attributes and Mineralogical Characteristics, Evaluated by Radiometry, of Highly Weathered Tropical Soils
Soil Sci. Soc. Am. J., June 2, 2005; 69(4): 1162 - 1172.
[Abstract] [Full Text] [PDF]

Home page
 Clay MineralsHome page
P. Nornberg, P. NORNBERG, U. SCHWERTMANN, H. STANJEK, T. ANDERSEN, and H. P. GUNNLAUGSSON
Mineralogy of a burned soil compared with four anomalously red Quaternary deposits in Denmark
Clay Minerals, March 1, 2004; 39(1): 85 - 98.
[Abstract] [Full Text] [PDF]

Home page
 American MineralogistHome page
E. Murad, E. Murad, and P. Rojik
Iron-rich precipitates in a mine drainage environment: Influence of pH on mineralogy
American Mineralogist, November 1, 2003; 88(11-12): 1915 - 1918.
[Abstract] [Full Text] [PDF]

Home page
 Clays and Clay MineralsHome page
J. Torrent, J. Torrent, and V. Barron
THE VISIBLE DIFFUSE REFLECTANCE SPECTRUM IN RELATION TO THE COLOR AND CRYSTAL PROPERTIES OF HEMATITE
Clays and Clay Minerals, June 1, 2003; 51(3): 309 - 317.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
R. Uusitalo and E. Turtola
Determination of Redox-Sensitive Phosphorus in Field Runoff without Sediment Preconcentration
J. Environ. Qual., January 1, 2003; 32(1): 70 - 77.
[Abstract] [Full Text] [PDF]

Home page
 Clays and Clay MineralsHome page
null Ji Junfeng, J. Ji, W. Balsam, J. Chen, and L. Liu
RAPID AND QUANTITATIVE MEASUREMENT OF HEMATITE AND GOETHITE IN THE CHINESE LOESS-PALEOSOL SEQUENCE BY DIFFUSE REFLECTANCE SPECTROSCOPY
Clays and Clay Minerals, April 1, 2002; 50(2): 208 - 216.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (46)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Scheinost, A.C.
Right arrow Articles by Schwertmann, U.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome