Pedogenic Iron Oxides in Iron-Rich Oxisols Developed from Mafic Rocks

ABSTRACT

Despite the considerable amount of information on the mineralogical characteristics of pedogenic Fe oxides in Brazilian soils, there are few studies on Fe-rich soils developed from mafic rocks with taxonomic identities at lower categorical levels. This study evaluated the mineralogical characteristics of pedogenic Fe oxides in B horizons (Bw) of Fe-rich Oxisols developed from several mafic rocks in the state of Minas Gerais, Brazil. The Bw horizons were sampled at a 0.8-1.0 m depth in 13 Ferric and Perferric Rhodic Oxisols along with a Mesoferric Typic Oxisol originating from basalt, gabbro, tuffite, amphibolite and itabirite in Minas Gerais. The selected soils have taxonomic identities up to the fourth categorical level of the Brazilian System of Soil Classification. In the laboratory, the following analyses were made: a) powder X ray diffraction (XRD) of the clay fraction before and after selective concentration of Fe oxides by silicate alkaline dissolution (5 mol L^-1 NaOH); b) selective chemical dissolution of the clay fraction by citrate-bicarbonate-dithionite (CBD), acid ammonium oxalate (AAO), and sulfuric acid (H₂SO₄ 1.8 mol L^-1); c) quantitative estimation of minerals in the clay fraction through allocation of phases from the XRD patterns, magnetic susceptibility of the clay fraction, and quantification of elements after sulfuric acid digestion (H₂SO₄ 1:1) of the air-dried fine earth and treatment of the clay fraction with CBD; and d) estimation of the mean crystal size (MCS), specific surface area (SSA), and isomorphic Al-substitution (IS) of hematite, goethite, and maghemite from the XRD patterns obtained from concentrates of Fe oxides. The results showed that estimation of Fe content of maghemite by selective dissolution with 1.8 mol L^-1 H₂SO₄ may not be accurate enough to realistically reflect the maghemite contents in the soil sample. The Al content extracted may also be influenced by other minerals that are sources of this element. Hematite crystals were predominantly placoid in shape in all Rhodic Oxisols and had smaller SSA compared to goethite, which showed both isodimensional and asymmetric habit. Higher crystallinity of maghemite and the IS values generally lower than those of hematite and goethite suggest that in well-drained soils derived from mafic rocks, the IS phenomenon in maghemites seems to result from pedogenetic advancement after its formation from magnetite oxidation.

hematite; maghemite; crystallographic characteristics; Al-substitution; selective dissolution

INTRODUCTION

Knowledge of the conditions under which Fe oxides (oxides, hydroxides, and oxyhydroxides) are formed in pedogenic and geochemical environments is critical for interpretation of current and past processes in soils (Kämpf and Curi, 2000). These minerals are involved in various soil phenomena, such as in adsorption of anions, cations, and organic compounds, as well as in fixing P and heavy metals. In addition, they are linked to some morphological properties, such as color, structure, and magnetism, inducing the formation of aggregates and cementing of other important soil components, namely nodules, concretions, plinthites, petroplinthites, ortsteins, etc. (Schwertmann and Taylor, 1989).

Goethite (Gt), responsible for yellow color, is the most common antiferromagnetic Fe oxide, followed by hematite (Hm), which causes red coloring in soils (Schwertmann and Taylor, 1989; Kämpf and Curi, 2000; Cornell and Schwertmann, 2003; Fontes and Carvalho Jr., 2005). Maghemite (Mh), ferromagnetic Fe oxide, is common in several soil classes in tropical and subtropical environments (Schwertmann and Taylor, 1989). Although it generally occurs in small amounts in some soil classes, Mh provides high magnetization to soils originating from basic igneous rocks (Fabris et al., 1997a; Costa et al., 1999) and other Fe-rich materials (Costa et al., 2014; Carvalho Filho et al., 2015). In Rhodic Oxisols developed from basic rocks, the Mh content increases considerably and may account for up to 40 % of the mass of the Fe oxide content in the clay fraction (Costa et al., 1999).

Methods of identification, quantification, and characterization of various mineral phases of Fe oxides are particularly important. Some of the main methods used include X ray diffraction (XRD – powder method) and selective chemical dissolution with citrate-bicarbonate-dithionite of sodium (CBD) (Mehra and Jackson, 1960), acid ammonium oxalate (AAO) (McKeague and Day, 1966), and sulfuric acid (Schwertmann and Fechter, 1984).

Despite the high geochemical affinity of Fe with various metals, Al³⁺ is the main isomorphic metal cation substituent in the structure of Fe oxides (IS). The amounts 0.33, 0.15, and 0.33 mol of IS mol^-1 are established limits for Gt, Hm, and Mh, respectively (Schwertmann and Kämpf, 1985; Kämpf and Curi, 2000; Cornell and Schwertmann, 2003). Isomorphic substitution is a natural phenomenon and is more a rule than an exception in pedogenic Fe oxides of tropical climate. Generally, higher degrees of isomorphic substitution are associated with weathered soils and with greater Al activity, resulting in considerable changes in the size of the unit cell and crystallinity of the mineral particle (Fontes and Weed, 1991).

Regarding size, Gt usually features an isodimensional crystallization habit and its spherical habit is confirmed by electron microscopy techniques in tropical soils (Schulze and Schwertmann, 1984; Schwertmann and Kämpf, 1985; Melo et al., 2001; Correa et al., 2008) and by similar values for mean crystal size (MCS) in the d₁₁₀ and d₁₁₁ directions obtained by the Scherrer equation (Klug and Alexander, 1954; Fontes and Weed, 1991). However, Melo et al. (2001) also found different MCS values for some samples, and growth was greater in the d₁₁₀ direction. Hematite commonly has the form of circular plates (Schwertmann and Kämpf, 1985; Fontes and Weed, 1991; Melo et al., 2001). Maghemite is usually formed from the transformation of the solid state of other Fe oxides and, for the most part, it adopts the growth habit of its precursor (Cornell and Schwertmann, 2003).

Although several studies have been conducted on Fe oxides in Brazil, few were able to gather a significant volume of soils originating from different parent materials with taxonomic similarities to the Brazilian (SiBCS - Santos et al., 2013) and North American (Soil Survey Staff, 2014) Classification Systems. The Fe-rich Rhodic Oxisols (Soil Survey Staff, 2014) occupy significant areas in the regions of the Central Plateau and the Ferriferous Quadrilateral in Brazil, especially those developed from basalt, tuffite, and itabirite, where, in addition to dystrophic and acric properties, typically feature high saturation magnetization, although with differences between soils developed from the same parent material (Carmo et al., 1984; Ferreira et al., 1994; Fabris et al., 1997a; Costa et al., 1999; Silva et al., 2005; Silva et al., 2010; Cervi et al., 2014; Costa et al., 2014; Carvalho Filho et al., 2015).

Once the mineralogical variations in pedogenic Fe oxides are determined by the parent material and pedoenvironmental conditions, it is hypothetically assumed that will be possible to observe quantitative and crystallographic differences in these minerals even between groups of taxonomically similar soils. Thus, this study evaluated the mineralogical characteristics of pedogenic Fe oxides in B horizons (Bw) of Fe-rich Oxisols developed from several mafic rocks in the state of Minas Gerais, Brazil.

MATERIALS AND METHODS

Samples were collected from the 0.8-1.0 m depth in different regions of the state of Minas Gerais in the Bw horizons of 13 Rhodic Oxisols originating from various parent materials: ferric (total contents of Fe₂O₃ between 180 and 360 g kg^-1 - LV₁, LV₂, LV₃, LV₅, LV₆, LV₁₀, LV₁₂, and LV₁₃) and perferric (total contents of Fe₂O₃ higher than 360 g kg^-1 - LV₄, LV₇, LV₈, LV₉, and LV₁₁). A mesoferric Typic Oxisol was also collected (total contents of Fe₂O₃ between 80 g kg^-1 and 180 g kg^-1 - LVA₁) (Figure 1, Table 1). According to the Brazilian System of Soil Classification (SiBCS) (Santos et al., 2013), the taxonomic names ferric, perferric, and mesoferric are based on the Fe content extracted after sulfuric digestion of air-dried fine earth (ADFE), which is obtained from air drying and breaking up soil samples, and then sieving them through a 2-mm mesh (Donagema et al., 2011).

Figure 1
Localization of the soils with their respective parent materials.

After separation of particle size fractions (Jackson, 1979), minerals of the clay fraction were estimated by the allocation method proposed by Resende et al. (1987) with ALOCA software (Moura Filho et al., 1995). The chemical elements of minerals in the clay fraction [extracted by sulfuric digestion with ADFE (1:1 ratio with H₂SO₄) (Donagema et al., 2011) and by five successive extractions with dithionite-citrate-bicarbonate of sodium (CBD) at 80 °C (Mehra and Jackson, 1960)] were allocated in the minerals identified by X ray diffraction (XRD) in the same clay fraction without treatment (natural clay). Diffractograms of X rays were obtained in the X’Pert Pro Panalytical device with CoKα radiation, operating at 40 kV and 30 mA. The scan range was 4 to 50 °2θ, with an interval of 0.01 °2θ s^-1.

Magnetic susceptibility per mass unit was quantified by measuring at low frequency (0.47 kHz) through a Bartington MS2 system (Bartingon Instruments LTD, Oxford, England), coupled to an MS2B sensor (Dearing, 1999). Determining magnetic susceptibility in the clay fraction allowed the Mh rate in soil to be estimated (Resende et al., 1987).

In addition to the sequential extractions with CBD for complete removal of pedogenic Fe oxides, including the crystalline (basically Gt, Hm, and Mh) and low crystalline (ferrihydrite - Fr) (Fe_d) forms (Mehra and Jackson, 1960), natural clay samples were also subjected to a single extraction with 0.2 mol L^-1 acid ammonium oxalate (AAO) at pH 3 for removal of only Fe from the low crystallinity phases (Fe_o) (McKeague and Day, 1966). For selective Fe dissolution of Mh, the clay fraction was subjected to only one extraction with 1.8 mol L^-1 H₂SO₄ at 80 °C for 2 h (Fe_H2SO4) (Schwertmann and Fechter, 1984, modified by Costa et al., 1999). The contents of Fe and Al solubilized by the CBD, AAO, and H₂SO₄ extractants were determined by atomic absorption spectroscopy.

Concentrates of Fe oxides were obtained by selective dissolution of silicate minerals in the clay fraction, using 5-mol L^-1 NaOH solution under heating (Norrish and Taylor, 1961; Kämpf and Schwertmann, 1982). The mineral components of concentrated samples were identified by XRD with scanning from 4 to 80 °2θ. The samples were ground in an agate mortar along with 10 % silicon, used as an internal standard for correction of distortions (position and width of half-height of peaks), and mounted on slides excavated without guidance (powder).

The Hm/(Hm+Gt) ratio was estimated based on reflex areas of d₀₁₂ of Hm and d₁₁₀ of Gt in Fe oxide concentrates. The formula used for calculation was suggested by Resende et al. (1987): [Hm/(Hm + Gt)] = 4 × A_Hm012/(4 × A_Hm012 + A_Gt110), in which A is an area of the respective reflexes.

The mean crystal size (MCS) was calculated from the width of the half-height of reflexes d₁₁₀ and d₁₁₁ of Gt, d₁₀₄, d₁₁₀, and d₀₁₂ of Hm, and d₂₂₀ and d₄₀₀ of Mh, using the Scherrer equation (Klug and Alexander, 1954). The instrumental error correction was made using the adjustment equation established by Melo et al. (2001) from the data presented by Klug and Alexander (1954) for low and moderate Bragg angles (°2θ). Next, values of the specific surface area (SSA) were determined for Gt and Hm. For Gt, the isodimensional shape (sphere) was adopted (Schulze and Schwertmann, 1984; Schwertmann and Kämpf, 1985; Fontes and Weed, 1991; Melo et al., 2001): SSA_Gt = (1049/MCS₁₀₀) - 5 (m² g^-1), in which MCS₁₀₀ = d₁₁₀ × 0.42 nm (Schulze and Schwertmann, 1984; Schwertmann and Kämpf, 1985). For Hm, the shape of round plates was adopted (Schwertmann and Kämpf, 1985; Fontes and Weed, 1991; Melo et al., 2001), and the geometric shape of a cylinder was used: SSA_Hm = 2 × (r + h) × 10³/r × h × d (m² g^-1), in which r = MCS₁₁₀ × 0.71/2, h = MCS₀₁₂ × 0.59, and d = 5.26 g cm^-3 (Schwertmann and Kämpf, 1985).

According to the Vegard Law, the occurrence and magnitude of isomorphic substitution of Fe by Al (IS) may be quantified from the regression between the volume of the mineral unit cell and the degree of IS [Al/(Fe+Al), mol mol^-1]. IS in the structure of Gt, Hm, and Mh was estimated by the position of the reflexes of these minerals obtained by XRD in the concentrates of Fe oxides. The position of the silicon peaks in each sample was used for instrumental corrections. In Gt, IS was calculated using the equation Al (mol mol⁻¹) = 17.30 - 5.72 × c₀, in which the c₀ size of the unit cell in Gt is obtained from d₁₁₀ and d₁₁₁ reflexes by c₀ = (1/d²₁₁₁ - 1/d²₁₁₀)^0.5 (Schulze, 1984). For Hm, the equation Al (mol mol^-1) = 31.09 - 6.17 × a₀ was used, in which the a₀ size of the unit cell in Hm is obtained by d₁₁₀ reflexes for a₀ = 2 × d₁₁₀ (Schwertmann et al., 1979). In Mh, IS was estimated by the equation Al (mol mol⁻¹) = [0.8343 - a₀] / 2.22 × 10^-4, in which a₀ corresponds to the size of the unit cell in Mh perpendicular to plane d₂₂₀ (Schwertmann and Fechter, 1984).

The Pearson correlations between the variables analyzed were determined using the statistical software Statistic v.9 (Carvalho Filho et al., 2015).

RESULTS AND DISCUSSION

Mineralogical composition of the clay fraction

The mineralogical constituents identified in the clay fraction were kaolinite, goethite, hematite, maghemite, gibbsite, anatase, and rutile at different rates, even in taxonomically related soils originating from the same parent material (Table 1). In quantitative terms, these results disagree with the general idea of mineralogical homogeneity for Oxisols (Resende et al., 1987).

The oxide dominance in the clay fraction, essentially gibbsite, hematite, maghemite, and anatase, is a consequence of the significant loss of exchangeable cations and silica (ferralitization), promoting accumulation of residual minerals in soils, as well as the desilicated nature of the parent materials (Carmo et al., 1984; Melo et al., 2001; Costa et al., 2014; Carvalho Filho et al., 2015). Only LV₁ showed a significant amount of kaolinite, possibly due to the influence of detrital sediments on the soil surface in South America (King, 1956).

Hematite (Hm) is the main antiferromagnetic Fe oxide in soils and its maximum content was observed in LV₁₁ (perferric) (226.2 g kg^-1), formed from itabirite (Table 1). However, a considerable amount of goethite (Gt) was also observed in LV₁₁, which was not expressed in color because of the high pigmenting power of Hm (Resende, 1976; Fontes and Carvalho Jr., 2005). The same behavior was observed in LV₁₂. Curi and Franzmeier (1987) reported the presence of significant quantities of Gt, higher than Hm, in a Rhodic Oxisol formed from itabirite. The authors highlight that the lithogenetic heritage of Gt, concentrated in its pigmenting effect on the coarse fraction (sand and silt), explains the redder colors in soils formed from basalt, with Hm contents greater than Gt contents.

Table 1
Soil classification according to the Brazilian System of Soil Classification – SiBCS and U.S. Soil Taxonomy (SSS) and quantitative estimate of the mineralogical assemblage of the clay fraction

SiBCS(1)	SSS(2)	K	Gb	Gt	Hm	Mh	An/Rt(3)
		----------------------------------------------g kg-1----------------------------------------------
Latossolo Vermelho Distroférrico típico - LV1	Rhodic Hapludox	433.1	67.2	23.4	134.4	41.5	22.4
Latossolo Vermelho Acriférrico típico - LV2	Rhodic Acrudox	175.9	338.5	16.5	93.3	91.0	27.0
Latossolo Vermelho Acriférrico típico - LV3	Rhodic Acrudox	209.2	273.5	12.2	98.3	91.3	16.8
Latossolo Vermelho Perférrico típico - LV4	Rhodic Acrudox	139.4	361.3	23.5	117.2	80.3	20.5
Latossolo Vermelho Acriférrico típico - LV5	Rhodic Acrudox	105.2	437.7	24.5	121.8	57.9	37.3
Latossolo Vermelho Distroférrico típico - LV6	Rhodic Hapludox	277.2	276.8	52.5	88.3	11.3	54.1
Latossolo Vermelho Perférrico típico - LV7	Rhodic Hapludox	54.4	228.2	12.0	64.1	161.8	49.7
Latossolo Vermelho Perférrico típico - LV8	Rhodic Acrudox	127.1	291.9	3.4	110.6	123.6	52.8
Latossolo Vermelho Perférrico típico - LV9	Rhodic Hapludox	71.4	189.4	n.d.	62.3	208.6	31.3
Latossolo Vermelho Distroférrico típico - LV10	Rhodic Hapludox	154.6	173.0	8.6	130.1	121.8	50.9
Latossolo Vermelho Perférrico típico - LV11	Rhodic Acrudox	21.9	138.2	85.9	226.2	217.1	8.3
Latossolo Vermelho Distroférrico típico - LV12	Rhodic Hapludox	221.1	183.3	60.5	173.7	20.8	17.7
Latossolo Vermelho Acriférrico típico - LV13	Rhodic Acrudox	256.1	259.8	39.3	112.6	41.4	5.6
Latossolo Vermelho-Amarelo Ácrico típico - LVA1	Typic Acrudox	196.8	435.3	52.9	64.6	1.3	48.3

• ⁽¹⁾ Santos et al. (2013); ⁽²⁾ Soil Survey Staff (2014); ⁽³⁾ Mineralogical composition estimated by allocation from X ray patterns, sulfuric digestion, CBD, and magnetic susceptibility (Resende et al., 1987). K: kaolinite, Gb: gibbsite, Gt: goethite, Mh: maghemite, Hm: hematite, and An/Rt: anatase and/or rutile; n.d.: not detected.

The occurrence of Hm and Gt oxides in soils is generally assigned to neoformation from the weathering of primary minerals containing Fe in their structure, or inherited directly from the parent material, as reported for Hm (Barbosa and Lagoeiro, 2010; Costa et al., 2014; Ávila et al., 2015) and Gt (Curi and Franzmeier, 1987) of ferriferous formations in the Ferriferous Quadrilateral in Minas Gerais. However, the predominance of Hm or Gt is controlled by the environmental conditions of soil formation (Curi and Franzmeier, 1984; Schwertmann and Taylor, 1989; Kämpf and Curi, 2000). The higher Fe content, common in soils from mafic rocks, associated with a lower degree of moisture, favored Hm formation and the predominance of this mineral over Gt in all soils studied. In contrast, LV₆ (3.5 YR hues) and LVA₁ (5 YR hue) had nearly equal Hm and Gt contents, indicating wetter pedoenvironmental conditions.

Maghemite (ferromagnetic Fe oxide - Mh) occurred in all soils, even if virtually inexpressive in LVA₁ and LV₆ (Table 1). In Brazil, for soils developed from mafic rocks, Mh formation is more associated with magnetite oxidation inherited from the parent material (Curi and Franzmeier, 1987; Fontes and Weed, 1991) as well as transformation from other magnetic phases, such as magnesiumferrite commonly found in tuffite and in the sand fraction of soils originating from it (Fabris et al., 1997b; Silva et al., 2005). However, under conditions of redox oscillations that occur during the wetting and drying cycles, fine magnetite grains may also be formed and subsequently oxidized, giving rise to Mh (Maxbauer et al., 2016). Although less common, another formation of Mh under field conditions is from transformation of other pedogenic Fe oxides (Hm and Gt) when under high temperatures in the presence of organic C (Curi and Franzmeier, 1987; Fontes and Weed, 1991; Costa et al., 1999; Fontes et al., 2000). The low expression of Mh in LV₆ and LVA₁ might be attributable to lithological variations and/or preferential dissolution of Mh rather than Hm in wetter pedoenvironments (Curi and Franzmeier, 1984).

The presence of Mh, in some cases at higher rates than Hm or Gt (LV₇, LV₈, and LV₉), highlights the need to quantitatively use the Mh contents or the magnetic properties of soils as criteria for taxonomic separation, especially for the Fe-rich Rhodic Oxisols (Carvalho Filho et al., 2015). In these cases, there is inconsistency in the exclusive use of color, dictated by the ratio Hm/(Hm + Gt) as criterion of the second-level category (suborder) in SiBCS (Santos et al., 2013). This consideration also applies to international classification systems, such as Soil Taxonomy (Soil Survey Staff, 2014) and the World Reference Base (IUSS, 2015), which do not have criteria for separating soil according to magnetic properties.

Anatase and/or rutile, Ti oxides in soils developed from common igneous and metamorphic rocks (Taylor et al., 1983), also occurred in the clay fraction of all the soils studied (Table 1). The genesis of these minerals may be related to weathering of minerals, such as ilmenite, or linked to the release of Ti during the weathering of silicates, such as pyroxene, amphibole, and biotite (Taylor et al., 1983). Anatase may also originate from Ti-magnetite (Taylor et al., 1983), fairly common in tuffite soils (Fabris et al., 1994). Soils formed from tuffite (LV₅, LV₆, LV₇, LV₈, LV₉, LV₁₀, and LVA₁) had higher contents of anatase/rutile, exceeding the maximum limit of 30 g kg^-1 generally found in the clay fraction of soils (Jackson, 1964). Rutile from XRD of Fe oxide concentrates also occurred in these soils (Figure 2). This shows the high influence of the parent material on Ti oxide content in the soil clay fraction.

Figure 2
X ray patterns from the clay fraction after treatment for Fe oxide concentration. (1) Internal standard used.

Selective dissolution of Fe oxides

In general, all soils exhibited high contents of Fe_d, reaching 480.5 g kg^-1 of clay in soil LV₁₁ (perferric). High Fe_d contents were also found by other authors for soils formed from itabirite (Table 2) (Curi and Franzmeier, 1987; Costa et al., 2014; Carvalho Filho et al., 2015). The lowest contents of Fe_d and Fe_H2SO4 were observed in LV₆ and LVA₁, suggesting both lower Fe content in the parent material, due to tuffite heterogeneity (Guimarães, 1955), and deferrification processes occurring in wetter pedoenvironments (Curi and Franzmeier, 1984; Schwertmann and Taylor, 1989; Kämpf and Curi, 2000).

Table 2
Silicon, Al, Fe, and Ti contents determined by sulfuric digestion of ADFE (Sis, Als, Fes , and Tis) and Fe and Al obtained in the clay fraction by sequential extractions with citrate-bicarbonate-dithionite (CBD) (Fed and Ald), only one extraction with acid oxalate ammonium (AAO) (Feo and Alo) and 1.8 mol L-1 H2SO4 (FeH2SO4 and AlH2SO4) all presented as oxides

Soil	Sulfuric digestion				CBD		AAO		1.8 mol L-1 H2SO4		Feo	Fed	FeH2SO4	FeH2SO4	ISd(1)	ISH2SO4(2)
	SiO2	Al2O3	Fe2O3	TiO2	Fe2O3	Al2O3	Fe2O3	Al2O3	Fe2O3	Al2O3	Fed	Fes	Fed	Fes
	-------------------------------------------g kg-1-----------------------------------------														---mol mol-1---
LV1	201.4	222.3	310.7	69.7	189.5	25.5	10.2	8.5	85.5	119.1	0.05	0.61	0.45	0.28	0.174	0.686
LV2	81.8	303.5	345.1	71.8	186.2	20.8	12.0	10.4	77.3	173.0	0.06	0.54	0.42	0.22	0.149	0.778
LV3	97.3	272.0	330.0	71.7	189.9	19.7	10.2	10.0	80.0	199.9	0.05	0.58	0.42	0.24	0.139	0.797
LV4	64.8	303.1	370.8	71.5	206.6	23.0	11.2	10.2	75.1	176.0	0.05	0.56	0.36	0.20	0.148	0.786
LV5	48.9	340.3	293.0	60.0	189.1	23.8	8.2	10.2	73.7	196.7	0.04	0.65	0.39	0.25	0.164	0.807
LV6	128.9	307.3	235.4	69.6	129.7	39.4	10.1	13.6	67.0	146.3	0.08	0.55	0.52	0.28	0.322	0.774
LV7	25.3	178.5	459.2	107.3	228.8	26.0	19.4	10.2	84.6	142.4	0.08	0.50	0.37	0.18	0.151	0.725
LV8	59.1	250.6	372.0	84.3	227.8	23.9	17.4	9.6	78.3	129.7	0.08	0.61	0.34	0.21	0.141	0.722
LV9	33.2	158.2	438.4	107.6	264.8	40.4	35.4	15.3	93.7	142.1	0.13	0.60	0.35	0.21	0.193	0.704
LV10	71.9	187.7	315.6	79.3	246.1	36.3	22.4	12.2	78.8	117.7	0.09	0.78	0.32	0.25	0.187	0.700
LV11	10.2	138.3	579.4	21.1	480.5	126.2	25.6	11.1	94.9	91.5	0.05	0.83	0.20	0.16	0.291	0.602
LV12	102.8	235.5	286.2	28.3	220.1	38.9	8.7	12.9	71.1	106.7	0.04	0.77	0.32	0.25	0.216	0.702
LV13	119.1	287.5	257.4	16.3	172.7	29.4	5.3	8.0	73.4	161.4	0.03	0.67	0.42	0.29	0.210	0.775
LVA1	91.5	373.1	161.4	48.3	102.4	30.9	3.1	9.8	26.4	166.4	0.03	0.63	0.26	0.16	0.320	0.908

• ⁽¹⁾ Isomorphic substitution of Fe by Al calculated from the molar ratio of the elements extracted by CBD, in which IS_d = [Al_d/(Al_d + 0.483Fe_d)] (Cornell and Schwertmann, 2003); ⁽²⁾ Isomorphic substitution of Fe by Al calculated from the molar ratio of the elements extracted by 1.8 mol L^-1 H₂SO₄ in which IS_H2SO4 = [Al_H2SO4/(Al_H2SO4 + 0.483Fe_H2SO4)] (Cornell and Schwertmann, 2003).

At least five successive extractions with CBD were required for total removal of the Fe contained in pedogenic Fe oxides (Figure 3). The slope of curves of cumulative Fe_d contents were lower in soils with greater participation of Gt (LVA₁ and LV₆). Conversely, Inda Junior and Kämpf (2003) highlighted that two successive extractions with CBD, heated at 80 °C, are enough to extract almost all (>98 %) pedogenic Fe oxides, regardless of the oxide content in the samples. This was not confirmed by Gualberto et al. (1987) and Corrêa et al. (2015), who reported the need for more than four extractions, due to the significant influence of crystallinity and isomorphic substitution of Fe by Al on the stability of Fe oxides in soils (Inda Junior and Kämpf, 2005; Correa et al., 2008; Corrêa et al., 2015; Carvalho Filho et al., 2015). Fontes (1988) observed that the slopes of the curves of cumulative Fe were higher as Hm and Gt showed lower isomorphic substitution of Fe by Al. The author concluded that only two extractions would not be enough to extract most aluminous Gt.

Figure 3
Cumulative iron content throughout the five sequential extractions with citrate-bicarbonate-dithionite in the clay fraction.

The Fe contents obtained from sulfuric digestion (Fe_s) (r = 0.92, p<0.01, n = 14), Fe_d (r = 0.80, p<0.01, n = 14), and Fe_H2SO4 (r = 0.72, p<0.01, n = 14) correlated with the Mh contents, which does not occur when these materials correlate with the ratios Fe_H2SO4/Fe_s and Fe_H2SO4/Fe_d. These observations confirm the association of Mh with the Fe content in the parent material (Costa et al., 1999; Silva et al., 2010; Carvalho Filho et al., 2015) and show the low efficiency of the method of selective dissolution with 1.8 mol L^-1 H₂SO₄ to quantify Fe from Mh. Costa et al. (1999) made changes to the original method of Schwertmann and Fechter (1984) by suggesting 2 h of extraction in an acidic solution rather than 7.5 h for different soil classes and parent materials. Inda et al. (2013) used 45 min for Argissols. This shows the high empiricism of the method and emphasizes the need to conduct further studies to establish appropriate protocols for Fe-rich Oxisols.

The Fe contents extracted by ammonium oxalate (Fe_o), usually lower than 20 g kg^-1 (Table 2), resulted in a low Fe_o/Fe_d ratio, indicating the dominance of crystalline forms of Fe oxides, characteristic of more developed Oxisols. With the exception of tuffite-originated soils (LV₆, LV₇, LV₈, LV₉, and LV₁₀), values <0.06 for the Fe_o/Fe_d ratio were observed. Kämpf et al. (1988) established the lower limit of 0.03 for the order of Oxisols in different regions of Brazil. The highest ratio values were established for most tuffite soils, especially in LV₉ (0.13), which reflects the lower degree of evolution of these soils compared to the others with shorter oxide crystallization time. The main implications of low crystallinity minerals in soils are related to relevant participation in their chemical properties, especially with regard to sorption reactions (Fontes and Weed, 1996).

The Fe_d/Fe_s ratio, which indicates the participation of Fe forms linked to pedogenic Fe oxides in relation to total Fe sulfuric digestion (Fe_s) (Rolim Neto et al., 2009), was low, except for LV₁₁. In Oxisols, the ratio should be between 0.8 and 1.0 (Kämpf et al., 1988). Resende et al. (2011) attribute this result to the high Fe_s values arising from the breakdown of other forms of more crystalline Fe in the coarse fractions (sand and silt from ADFE). However, in Oxisols, the lowest Fe_d/Fe_s ratios were always associated with yellowish soils, due to preferential dissolution of Mh and Hm and greater resistance of Gt, which was largely replaced by Al for reduction by CBD (Torrent et al., 1987; Fontes, 1988).

The values for isomorphic substitution of Fe by Al (IS_d), calculated by the ratio of molar concentration Al_d/(Al_d+Fe_d), ranged from 0.139 to 0.322 mol mol^-1 (Table 2), consistent with results commonly found by other authors in studies on highly weathered soils from tropical and subtropical regions (Curi and Franzmeier, 1987; Fontes and Weed, 1991; Kämpf and Curi, 2000; Inda Junior and Kämpf, 2005; Correa et al., 2008; Corrêa et al., 2015) and also within the maximum limits established for the structures of Gt and Mh (0.33 mol mol^-1) and Hm (0.15 mol mol^-1) (Cornell and Schwertmann, 2003). The indirect relationship between slope of the curves of the cumulative Fe_d contents (Figure 3) and the IS_d values confirm the stabilizing effect of Al on Gt structure (Torrent et al., 1987; Fontes, 1988; Inda Junior and Kämpf, 2005; Carvalho Filho et al., 2015) in LV₆ and LVA₁.

The positive correlation between contents of Al_d and Fe_d (0.85, p<0.01, n = 14), as well as the negative correlation between contents of Al_d and Al obtained by sulfuric digestion (Al_s) (r = -0.53, p<0.05, n = 14), indicate that the Al contents obtained by successive extractions with CBD are predominantly from the structure of Fe oxides, with negligible performance of other possible sources of this element, such as gibbsite and/or kaolinite (Inda Junior and Kämpf, 2003). This hypothesis is consistent with the absence of correlation between IS_d and Al_s.

Conversely, higher contents of Al_H2SO4, not correlated with those of Fe_H2SO4, suggest the contribution of Al sources, besides Mh, extracted after selective dissolution with 1.8 mol L^-1 H₂SO₄. In addition, the values for isomorphic substitution of Fe by Al (IS_H2SO4) calculated by the ratio of molar concentration Al_H2SO4/(Al_H2SO4+Fe_H2SO4) were higher than those for the Mh structure (Cornell and Schwertmann, 2003) and correlated with the contents of Al_s (r = 0.90, p<0.01, n = 14), reinforcing the hypothesis of contribution of other Al sources (Costa et al., 1999).

Crystallographic characteristics of Fe oxides

The values of the Hm/(Hm+Gt) ratio between 0.55 and 1.00 indicate the higher rate of Hm in relation to Gt in all soils, explaining hues redder than 3.5 YR, with the exception of LVA₁ (5 YR hue) (Table 3). The main factors that favor Hm formation in the soil, and therefore compete to increase this ratio, are greater Fe contents in the parent material, high temperatures, lower humidity, lower organic matter content, and higher pH values (Schwertmann and Taylor, 1989). In this study, excluding the LV₁₁ sample, the value of the linear correlation coefficient between the Hm/(Hm+Gt) ratio and the Fe_d and Fe_s contents increased from 0.24 (ns, n = 14) to 0.88 (p<0.01, n = 13) and from 0.44 (ns, n = 14) to 0.80 (p<0.01, n = 13), respectively. This fact, in addition to indicating that the Fe content in the parent material was essential in the formation of Fe oxides (Melo et al., 2001; Carvalho Filho et al., 2015), suggests that LV₁₁ (Fe_s = 570.4 g kg^-1), originating from itabirite, shows significant amounts of Gt obtained from the parent material (Curi and Franzmeier, 1987).

Table 3
d-spacing (d) corrected, mean crystal size (MCS), specific surface area (SSA), isomorphic substitution of Fe by Al (IS) in goethite (Gt), hematite (Hm), and maghemite (Mh), and Hm/(Hm + Gt) ratio for the clay fraction

Soil	d-spacing							MCS							MCS ratios(2)			Hm/(Hm+Gt)	SSA		IS
	Gt110	Gt111	Hm012	Hm104	Hm110	Mh220	Mh400	Gt110	Gt111	Hm012	Hm104	Hm110	Mh220	Mh400	A	B	C		Gt	Hm	Gt	Hm	Mh
	------------------------------------------------------------nm------------------------------------------------------------																		---m2 g-1---		----mol mol-1----
LV1	0.416	0.243	0.369	0.271	0.251	0.295	0.208	26.57	14.86	24.33	13.93	30.83	38.91	47.93	1.79	1.55	0.81	0.85	89.0	43.9	0.175	0.057	0.033
LV2	0.416	0.242	0.372	0.269	0.251	0.294	0.208	17.03	13.82	33.54	20.04	39.74	45.41	36.04	1.23	1.45	1.26	0.85	141.7	36.4	0.296	0.154	0.125
LV3	0.414	0.242	0.366	0.269	0.251	0.294	0.208	15.22	16.00	33.55	20.04	39.73	27.46	24.29	0.95	1.45	1.13	0.89	159.1	36.5	0.270	0.130	0.194
LV4	0.413	0.242	0.367	0.270	0.251	0.294	0.208	15.23	13.23	29.89	20.03	34.82	54.49	47.94	1.15	1.42	1.14	0.83	159.0	42.3	0.253	0.106	0.163
LV5	0.413	0.242	0.367	0.269	0.251	0.294	0.207	19.36	11.29	24.58	20.04	34.82	30.44	32.14	1.71	1.73	0.95	0.83	124.0	34.8	0.171	0.125	0.175
LV6	0.414	0.243	0.367	0.269	0.250	n.d.	n.d.	38.02	14.47	22.60	11.55	23.47	n.d.	n.d.	2.63	1.27	n.d.	0.63	60.7	70.4	0.176	0.188	n.d.
LV7	0.415	0.242	0.367	0.270	0.251	0.294	0.208	20.78	17.89	38.27	20.04	34.83	23.01	36.05	1.16	1.11	0.64	0.84	115.2	54.1	0.207	0.147	0.129
LV8	0.416	0.243	0.367	0.269	0.251	0.294	0.208	66.47	46.55	22.60	18.75	46.30	30.44	24.29	1.43	2.50	1.25	0.97	32.6	18.1	0.150	0.126	0.174
LV9	n.d.	n.d.	0.368	0.269	0.251	0.294	0.208	n.d.	n.d.	17.15	11.55	18.94	21.31	18.43	n.d.	1.35	1.16	1.00	n.d.	82.0	n.d.	0.148	0.151
LV10	0.416	0.243	0.367	0.270	0.251	0.294	0.208	26.78	23.58	19.49	15.76	31.04	27.46	26.42	1.14	1.94	1.04	0.94	88.3	34.7	0.191	0.145	0.108
LV11	0.417	0.242	0.367	0.270	0.251	0.294	0.207	17.03	12.68	20.92	17.63	18.94	21.31	19.61	1.34	1.10	1.09	0.72	141.7	100.1	0.260	0.167	0.192
LV12	0.414	0.242	0.366	0.269	0.251	0.294	0.207	20.78	13.22	19.49	13.03	23.47	34.17	59.26	1.57	1.47	0.58	0.74	115.2	60.7	0.224	0.171	0.127
LV13	0.416	0.242	0.367	0.269	0.251	0.295	0.208	20.78	19.03	26.97	18.76	31.04	18.57	19.60	1.09	1.40	0.95	0.74	115.2	48.0	0.227	0.146	0.085
LVA1	0.414	0.242	0.368	0.269	0.251	n.d.	n.d.	33.32	19.04	29.89	12.00	21.72	n.d.	n.d.	1.75	0.89	n.d.	0.55	70.0	108.7	0.207	0.095	n.d.

• ⁽¹⁾ d-spacing corrected using silicon as an internal standard. Hm/(Hm + Gt), MCS, SSA and IS ratios were calculated from the X ray patterns of the concentrated Fe oxides; n.d. not detected; ⁽²⁾ Ratios calculated from the MCS values: A = MCS_Gt110/MCS_Gt111, B = MCSa/MCSc = (MCS_Hm110 × 0.72)/(MCS_Hm012 × 0.59), C = MCS_Mh220/MCS_Mh400.

The d-spacing (d) for Fe oxides (Hm, Gt, and Mh) in various directions (hkl) was corrected using Si as an internal standard (Table 3). The values were homogeneous overall; however, formed on the same parent material. The d₁₁₀ and d₁₁₁ reflexes of Gt were measured in all soils except for LV₉ (Figure 2).

The values of MCS_Gt, perpendicular to planes d₁₁₀ and d₁₁₁, showed relative agreement with the values observed by other authors for Brazilian soils (Schwertmann and Kämpf, 1985; Curi and Franzmeier, 1987; Fontes and Weed, 1991; Melo et al., 2001; Carvalho Filho et al., 2015). The results for the MCS_Gt110/MCS_Gt111 ratio ranged from 0.95 to 2.63 (Table 3) and showed differences among the crystal morphologies in this mineral. Although some samples showed values of this ratio close to one, in others, the values are far higher (over 1.4), contrary to the common trend of equidimensionality of Gt along the three crystallographic axes (xyz) (Schwertmann, 1988). According to Schwertmann (1988), acicular crystals, in which MCS_Gt110 is greater than MCS_Gt111, are observed in Gt, displaying preferred growth along the z direction. However, the occurrence of both isodimensional and acicular Gt is common in Brazilian soils (Schwertmann and Kämpf, 1985; Fontes and Weed, 1991; Melo et al., 2001; Cornell and Schwertmann, 2003).

There is a great range of estimated values for SSA_Gt (32.2 to 157.2 m² g^-1), as well as for IS_Gt (0.150 to 0.296 mol mol^-1) (Table 3), both within the range of common values in pedogenetic environments (Carvalho Filho et al., 2015). The influence of IS on crystal size reduction of Gt (Schulze and Schwertmann, 1984; Schwertmann and Kämpf, 1985) was confirmed by the significant correlation between the values of IS_Gt and SSA_Gt (r = 0.81, p<0.01, n = 13) (Figure 4a). The same result was observed correlating IS_Gt with MCS_Gt110 (r = -0.69, p<0.01, n =13) (Schwertmann and Kämpf, 1985; Fontes and Weed, 1991; Melo et al., 2001).

Figure 4
Relationship between specific surface area (SSA) and isomorphic substitution of Fe by Al (IS) values in Gt (a); and relationship between IS values in Hm and Gt (b). **: significant at 1 %.

The great variation in SSA_Gt values in tuffite soils highlights the occurrence of Gt with different degrees of crystallinity between soils originating from the same parent material. In this case, besides the possibility of variation in pedoenvironmental conditions favorable to the occurrence of IS, tuffites are basic unhardened heterogeneous rocks, with various proportions of pyroclastic debris and, consequently, different elementary levels (Guimarães, 1955), which may also cause variations in mineral crystallinity. Another aspect is the clear difference of the soil under the influence of detrital materials in the South American plain (LV₁), where SSA_Gt and IS_Gt values were well below those observed in the other basalt soils.

There was no correlation between IS_Gt and the contents of Al_s, disagreeing with the general trend observed in Brazilian soils (Costa et al., 2014). The direct relationship between IS_Gt and Al activity in the soil indicates a pedogenetic environment with strong desilication, where Al would be more susceptible to be incorporated into the Gt structure (Schwertmann and Kämpf, 1985). However, generalizations should be avoided because the intensity of this phenomenon cannot be based solely on Al availability in the system. Gibbsite formation is competitive to IS in Gt, that is, environments with high Al contents may prioritize gibbsite formation and reduce IS in Fe oxides.

In general, higher values of MCS_Hm were observed compared to those obtained for Gt (Table 3), mainly in the d₁₁₀ direction. This was also verified by other authors (Schwertmann and Kämpf 1985; Kämpf et al., 1988; Fontes and Weed, 1991; Melo et al., 2001, Carvalho Filho et al., 2015), indicating greater development of crystals in the x-y direction compared to the z direction. According to Fontes and Weed (1991) and Melo et al. (2001), this pattern of growth also suggests greater crystallinity of Hm compared to Gt.

Values of the MCSa/MCSc ratio (MCSa = MCS₁₁₀ × 0.72; MCSc = MCS₀₁₂ × 0.59) higher than the unit in all soils, except for LVA₁, indicate predominance of placoid morphology for Hm.

IS_Hm values ranged from 0.057 to 0.188 mol mol^-1 (Table 3). These results are lower than those observed for Gt, except in LVA₁. This reinforces the hypothesis that when these two minerals occur simultaneously in soils, Hm tends to be smaller than IS (Schwertmann et al., 1979; Schwertmann and Kämpf, 1985; Singh and Gilkes, 1992). Singh and Gilkes (1992), based on the linear regression coefficient adjusted between IS_Gt and IS_Hm (line slope = 0.5), reported that Hm may accommodate approximately half of Al present in Gt in soils when these minerals coexist, which was also observed in this study (Figure 4b). However, the occurrence of IS_Hm values higher than half the values observed for Gt indicates that in soils with significant Hm contents, under favorable conditions, the occurrence of this phenomenon in Hm may be more effective.

In addition to the higher values of MCS_Hm, SSA_Hm values lower than those for Gt were observed, with few exceptions (LV₆ and LVA₁), in agreement with the results observed by Melo et al. (2001) and Carvalho Filho et al. (2015). The lack of correlation between IS_Hm and SSA_Hm, as well as between IS_Hm and MCS_Hm012, and MCS_Hm104 and MCS_Hm110, show high variability of the factors involved in Hm crystallinity in the Oxisols studied.

In general, the MCS_Mh values were higher than those found for Hm and Gt, which reflects better crystallinity of Mh in soils derived from mafic rocks where their genesis is more associated with oxidation of magnetite inherited from the parent material (Curi and Franzmeier, 1987; Fontes and Weed, 1991; Costa et al., 1999). With values close to one, the MCS_Mh220/MCS_Mh400 ratio indicates equal dimensions in the d₂₂₀ and d₄₀₀ directions of Mh.

IS_Mh values ranged from 0.033 to 0.194 mol mol^-1, similar to the values found by other authors in tropical and subtropical soils (Schwertmann and Fechter, 1984; Fontes and Weed, 1991; Costa et al., 1999; Carvalho Filho et al., 2015) and generally lower than those estimated for Hm and Gt. Thus, it can be suggested that in well-drained soils from mafic rocks, IS in Mh would occur throughout pedogenesis from its direct transformation after oxidation of magnetite inherited from the parent material.

The IS values estimated by mathematical equations of Schulze (1984), Schwertmann et al. (1979), and Schwertmann and Fechter (1984) for Gt (IS_Gt), Hm (IS_Hm), and Mh (IS_Mh), respectively, were not consistent with those obtained through chemical dissolution by CBD (IS_d). Correa et al. (2008) also did not observe correlation between the IS_Gt determined by chemical methods and that estimated by mathematical equations in soils of the Barreiras Group and, therefore, highlighted the need for studies to demonstrate the relationships between the crystallographic parameters of Gt and their chemical properties.

CONCLUSIONS

The occurrence of Mh contents higher than those of Hm and Gt in some Rhodic Oxisols underscores the importance of this mineral for soils originating from mafic rocks and suggests the need to reflect on the use of magnetic properties as criteria for taxonomic classification.

The estimate of Fe from Mh obtained for selective dissolution with 1.8 mol H₂SO₄ may not correspond to actual existence in the soil. In addition, the contents of Al extracted may be heavily influenced by other minerals that are sources of Al.

IS_Gt was higher than IS_Hm in all Rhodic Oxisols. This phenomenon has a strong influence on SSA_Gt; however, for Hm, other factors seem to exert a greater effect on crystallinity.

Hm crystals predominantly displayed a placoid shape in all Rhodic Oxisols and lower SSA_Hm in comparison to Gt crystals, which occurred both in the isodimensional and acicular forms.

The higher Mh crystallinity and IS_Mh values generally lower than those of Hm and Gt suggest that in well-drained soils originating from mafic rocks, the IS phenomenon in Mh seems to be a result of pedogenetic advances after its formation from magnetite oxidation.

REFERENCES

• Ávila CF, Lagoeiro L, Barbosa PF, Graça L. EBSD analysis of rhombohedral twinning in hematite crystals of naturally deformed iron formations. J Appl Cryst. 2015;48:212-19. https://doi.org/10.1107/S1600576714025928
  » https://doi.org/10.1107/S1600576714025928
• Barbosa PF, Lagoeiro L. Crystallographic texture of the magnetite-hematite transformation: Evidence for topotactic relationships in natural samples from Quadrilátero Ferrífero, Brazil. Am Miner. 2010;95:118-25. https://doi.org/10.2138/am.2010.3201
  » https://doi.org/10.2138/am.2010.3201
• Carmo DN, Curi N, Resende M. Caracterização e gênese de Latossolos da Região do Alto Paranaíba (MG). Rev Bras Cienc Solo. 1984;8:235-40.
• Carvalho Filho A, Inda AV, Fink JR, Curi N. Iron oxides in soils of different lithological origins in Ferriferous Quadrilateral (Minas Gerais, Brazil). Appl Clay Sci. 2015;118:1-7. https://doi.org/10.1016/j.clay.2015.08.037
  » https://doi.org/10.1016/j.clay.2015.08.037
• Cervi EC, Costa ACS, Souza Junior IG. Magnetic susceptibility and the spatial variability of heavy metals in soils developed on basalt. J Appl Geophys. 2014;111:377-83. https://doi.org/10.1016/j.jappgeo.2014.10.024
  » https://doi.org/10.1016/j.jappgeo.2014.10.024
• Cornell RM, Schwertmann U. The Iron Oxides: structure, properties, reactions, occurrences and uses. 2nd ed. Weinheim, Germany: Wiley-VHC Verlag GmbH and Co. KGaA; 2003.
• Correa MM, Ker JC, Barrón V, Fontes MPF, Torrent J, Curi N. Caracterização de óxidos de ferro de solos do ambiente tabuleiros costeiros. Rev Bras Cienc Solo. 2008;32:1017-31. https://doi.org/10.1590/S0100-06832008000300011
  » https://doi.org/10.1590/S0100-06832008000300011
• Corrêa MM, Ker JC, Araújo Filho JC, Camêlo DL. Formas de ferro, silício e, ou, alumínio na gênese de fragipãs e horizontes coesos dos tabuleiros costeiros. Rev Bras Cienc Solo. 2015;39:940-9. https://doi.org/10.1590/01000683rbcs20140679
  » https://doi.org/10.1590/01000683rbcs20140679
• Costa ACS, Bigham JM, Rhoton FE, Traina SJ. Quantification and characterization of maghemite in soils derived from volcanic rocks in southern Brazil. Clays Clay Miner. 1999;47:466-73.
• Costa SAD, Ker JC, Simões DFF, Fontes MPF, Fabris JD, Andrade FV. Pedogênese e classificação de Latossolos desenvolvidos de itabiritos no Quadrilátero Ferrífero, MG. Rev Bras Cienc Solo. 2014;38:359-71. https://doi.org/10.1590/S0100-06832014000200001
  » https://doi.org/10.1590/S0100-06832014000200001
• Curi N, Franzmeier DP. Effect of parent rocks on chemical and mineralogical properties of some Oxisols in Brazil. Soil Sci Soc Am J. 1987;51:153-8. https://doi.org/10.2136/sssaj1987.03615995005100010033x
  » https://doi.org/10.2136/sssaj1987.03615995005100010033x
• Curi N, Franzmeier DP. Toposequence of Oxisols from the Central Plateau of Brazil. Soil Sci Soc Am J. 1984;48:341-6. https://doi.org/10.2136/sssaj1984.03615995004800020024x
  » https://doi.org/10.2136/sssaj1984.03615995004800020024x
• Dearing J. Environmental magnetic susceptibility: Using the Bartington MS2 System. 2nd ed. Keniloworth: Chi Publishing; 1999.
• Donagema GK, Campos DVB, Calderano SB, Teixeira WG, Viana JHM, organizadores. Manual de métodos de análise do solo. 2a ed. rev. Rio de Janeiro: Embrapa Solos; 2011.
• Fabris JD, Coey JMD, Jesus Filho MF, Santana DP, Goulart AT, Fontes MF, Curi N. Mineralogical analysis of a weathering mantle derived from tuffite. Hyperfine Interact. 1994;91:751-7. https://doi.org/10.1007/BF02064602
  » https://doi.org/10.1007/BF02064602
• Fabris JD, Jesus Filho MF, Coey JMD, Mussel WN, Goulart AT. Iron-rich spinels from Brazilian soils. Hyperfine Interact. 1997a;110:23-32. https://doi.org/10.1023/A:1012619331408
  » https://doi.org/10.1023/A:1012619331408
• Fabris JD, Mussel WN, Coey JMD, Jesus Filho MF, Goulart AT. Mg-rich iron oxide spinels from tuffite. Hyperfine Interact. 1997b;10:33-40. https://doi.org/10.1023/A:1012611029591
  » https://doi.org/10.1023/A:1012611029591
• Ferreira SAD, Santana DP, Fabris JD, Curi N, Nunes Filho E, Coey JMD. Relações entre magnetização, elementos traços e litologia de duas sequências de solos do Estado de Minas Gerais. Rev Bras Cienc Solo. 1994;18:167-174.
• Fontes MPF. Iron oxide mineralogy in some Brazilian Oxisols [Thesis]. Raleigh: North Carolina State University; 1988.
• Fontes MPF, Carvalho Jr. IA. Color attributes and mineralogical characteristics, evaluated by radiometry, of highly weathered tropical soils. Soil Sci Soc Am J. 2005;69:1162-72. https://doi.org/10.2136/sssaj2003.0312
  » https://doi.org/10.2136/sssaj2003.0312
• Fontes MPF, Oliveira TS, Costa LM, Campos AAG. Magnetic separation and evaluation of magnetization of Brazilian soils from different parent materials. Geoderma. 2000;96:81-99. https://doi.org/10.1016/S0016-7061(00)00005-7
  » https://doi.org/10.1016/S0016-7061(00)00005-7
• Fontes MPF, Weed SB. Iron oxides in selected Brazilian Oxisols: I. Mineralogy. Soil Sci Soc Am J. 1991;55:1143-9. https://doi.org/10.2136/sssaj1991.03615995005500040040x
  » https://doi.org/10.2136/sssaj1991.03615995005500040040x
• Fontes MPF, Weed SB. Phosphate adsorption by clays from Brazilian Oxisols: relationships with specific surface area and mineralogy. Geoderma. 1996;72:37-51. https://doi.org/10.1016/0016-7061(96)00010-9
  » https://doi.org/10.1016/0016-7061(96)00010-9
• Guimarães D. Contribuição ao estudo dos tufos vulcânicos da Mata da Corda. Belo Horizonte: Instituto de Tecnologia Industrial; 1955. (Boletim, 18).
• Gualberto V, Resende M, Curi N. Química e mineralogia de Latossolos, com altos teores de ferro, da Amazônia e do Planalto Central. Rev Bras Cienc Solo. 1987;11:245-52.
• Inda AV, Torrent J, Barrón V, Bayer C, Fink JR. Iron oxides dynamics in a subtropical Brazilian Paleudult under long-term no-tillage management. Sci Agric. 2013;70:48-54. https://doi.org/10.1590/S0103-90162013000100008
  » https://doi.org/10.1590/S0103-90162013000100008
• Inda Junior AV, Kämpf N. Avaliação de procedimentos de extração dos óxidos de ferro pedogênicos com ditionito-citrato-bicarbonato de sódio. Rev Bras Cienc Solo. 2003;27:1139-47. https://doi.org/10.1590/S0100-06832003000600018
  » https://doi.org/10.1590/S0100-06832003000600018
• Inda Junior AV, Kämpf N. Variabilidade de goethita e hematita via dissolução redutiva em solos de região tropical e subtropical. Rev Bras Cienc Solo. 2005;29:851-66. https://doi.org/10.1590/S0100-06832005000600003
  » https://doi.org/10.1590/S0100-06832005000600003
• International Union of Soil Sciences/Working Group Word Reference Base – IUSS/WRB. World Reference Base for Soil Resources 2014, update 2015 International soil classification system for naming soils and creating legends for soil maps. Rome: FAO; 2015. (World Soil Resources Reports, 106).
• Jackson ML. Soil Clay mineralogical analysis. In: Rich CI, Kunze GW, editors. Soil clay mineralogy. Chapel Hill: Univ. North Carolina Press; 1964.
• Jackson ML. Soil chemical analysis - advanced course: A manual of methods useful for instruction and research in soil chemistry, physical chemistry of soils, soil fertility, and soil genesis. 2nd ed. Madison, WI: USA Dept. Soil Sci. Univ. of Wisconsin; 1979.
• Kämpf N, Curi N. Óxidos de ferro: Indicadores de ambientes pedogênicos e geoquímicos. Tópicos Cienc Solo. 2000;1:107-38.
• Kämpf N, Resende M, Curi N. Iron oxides in Brazilian Oxisols. In: 8th International Soil Classification Workshop: Classification, Characterization and Utilization of Oxisols; 1988; Rio de Janeiro. Rio de Janeiro: Embrapa-SNLCS/SMSS/USDA-SCS/UPR; 1988. p.71-7.
• Kämpf N, Schwertmann U. The 5-M-NaOH concentration treatment for iron oxides in soils. Clays Clay Miner. 1982;30:401-8.
• King LC. A Geomorfologia do Brasil Oriental. Rev Bras. Geogr. 1956;18:147-263.
• Klug HP, Alexander LE. X-ray diffraction procedures for polycrystalline and amorphous materials. New York: John Wiley and Sons; 1954.
• Maxbauer DP, Feinberg JM, Fox DL. Magnetic mineral assemblages in soils and paleosols as the basis for paleoprecipitation proxies: A review of magnetic methods and challenges. Earth Sci Rev. 2016;155:28-48. https://doi.org/10.1016/j.earscirev.2016.01.014
  » https://doi.org/10.1016/j.earscirev.2016.01.014
• Mckeague JA, Day JH. Dithionite and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. Can J Soil Sci. 1966;46:13-22.
• Mehra JP, Jackson ML. Iron oxides removal from soils and clays by a dithionite-citrate-bicarbonate system buffered with bicarbonate sodium. Clays Clay Miner. 1960;7:317-27.
• Melo VF, Fontes MPF, Novais RF, Singh B, Schaefer CEGR. Características dos óxidos de ferro e de alumínio de diferentes classes de solos. Rev Bras Cienc Solo. 2001;25:19-32. https://doi.org/10.1590/S0100-06832001000100003
  » https://doi.org/10.1590/S0100-06832001000100003
• Moura Filho G, Resende M, Cruz CD. ALOCA – Programa de estimativa de minerais do solo. Versão 1.0. In: 25° Congresso Brasileiro de Ciência do Solo; 1995; Viçosa. Viçosa: SBCS; 1995. (Software)
• Norrish K, Taylor RM. The isomorphous replacement of iron by aluminium in soil goethites. J. Soil Sci. 1961;12:294-306. https://doi.org/10.1111/j.1365-2389.1961.tb00919.x
  » https://doi.org/10.1111/j.1365-2389.1961.tb00919.x
• Resende M. Mineralogy, chemistry, morphology and geomorphology of some soils of the Central Plateau of Brazil [Thesis]. West Lafayette: Purdue University; 1976.
• Resende M, Bahia Filho AFC, Braga JM. Mineralogia da argila de Latossolos estimada por alocação a partir do teor total de óxidos do ataque sulfúrico. Rev Bras Cienc Solo. 1987;11:17-23.
• Resende M, Curi N, Ker JC, Rezende SB. Mineralogia de solos brasileiros: interpretação e aplicações. 2a ed. Lavras: Universidade Federal de Lavras; 2011.
• Rolim Neto FC, Schaefer CEGR, Fernandes Filho EI, Corrêa MM, Costa LM, Parahyba RBV, Guerra SMS, Heck R. Topolitossequências de solos de Alto Paranaíba: atributos físicos, químicos e mineralógicos. Rev Bras Cienc Solo. 2009;33:1795-809. https://doi.org/10.1590/S0100-06832009000600028
  » https://doi.org/10.1590/S0100-06832009000600028
• Santos HG, Jacomine PKT, Anjos LHC, Oliveira VA, Oliveira JB, Coelho MR, Lumbreras JF, Cunha TJF. Sistema brasileiro de classificação de solos. 3a ed. Rio de Janeiro: Embrapa Solos; 2013.
• Schulze DG. The influence of aluminium on iron oxides. VIII. Unit-cell dimensions of Al-substituted goethites and estimation of Al from them. Clays Clay Miner. 1984;32:36-44.
• Schulze DG, Schwertmann U. The influence of aluminium on iron oxides. X. Properties of Al-substituted goethites. Clay Miner. 1984;19:521-39.
• Schwertmann U. Some properties of soil and synthetic iron oxides. In: Stucki JW, Goodman BA, Schwertmann U, editors. Iron in soils and clay minerals. Dordrecht: D. Reidel Publishing Company; 1988. p.203-50.
• Schwertmann U, Fechter H. The influence of aluminium on iron oxides: XI. Aluminium substituted maghemite in soils and its formation. Soil Sci Soc Am J. 1984;48:1462-3. https://doi.org/10.2136/sssaj1984.03615995004800060054x
  » https://doi.org/10.2136/sssaj1984.03615995004800060054x
• Schwertmann U, Fitzpatrick RW, Taylor RM, Lewis DG. The influence of aluminium on iron oxides. Part II. Preparation and properties of Al-substituted hematites. Clays Clay Miner. 1979;27:105-12.
• Schwertmann U, Kämpf N. Properties of goethite and hematite in kaolinitic soils of Southern and Central Brazil. Soil Sci. 1985;139:344-50.
• Schwertmann U, Taylor RM. Iron oxides. In: Dixon JB, Weed SB, editors. Minerals in soil environments. Madison: Soil Science Society of America; 1989. p.379- 438.
• Silva AR, Souza Junior IG, Costa ACS. Suscetibilidade magnética do horizonte B de solos do Estado do Paraná. Rev Bras Cienc Solo. 2010;34:329-37. https://doi.org/10.1590/S0100-06832010000200006
  » https://doi.org/10.1590/S0100-06832010000200006
• Silva FD, Couceiro PRC, Fabris JD, Goulart AT, Ker JC. Magnesioferrita e caminho pedogenético de transformação de óxidos de ferro magnéticos em dois perfis de solo derivados de tufito da região do Alto Paranaíba (MG). Rev Bras Cienc Solo. 2005;29:763-75. https://doi.org/10.1590/S0100-06832005000500012
  » https://doi.org/10.1590/S0100-06832005000500012
• Singh B, Gilkes RJ. Properties and distribution of iron oxides and their association with minor elements in the soils of south-western Australia. J Soil Sci. 1992;43:77-98. https://doi.org/10.1111/j.1365-2389.1992.tb00121.x
  » https://doi.org/10.1111/j.1365-2389.1992.tb00121.x
• Soil Survey Staff. Keys to soil taxonomy. 12th ed. Washington, DC: United States Department of Agriculture, Natural Resources Conservation Service; 2014.
• Taylor RM, Mckenzie RM, Fordham AW, Gillman GP. Oxide minerals. In: Soils: an Australian viewpoint. London: Division of Soils, CSIRO, Melbourne Academic Press; 1983. p. 309-34.
• Torrent J, Schwertmann U, Barrón V. The reductive dissolution of synthetic goethite and hematite in dithionite. Clay Miner. 1987;22:329-37.

Publication Dates

• Publication in this collection
  2017

History

• Received
  17 Aug 2016
• Accepted
  1 Nov 2016