J. Cent. South Univ. (2017) 24: 1627-1636

DOI: 10.1007/s11771-017-3568-8

Metallogenic mechanism of Pingguo bauxite deposit, western Guangxi, China: Constraints from REE geochemistry and multi-fractal characteristics of major elements in bauxite ore

CAO Jing-ya(曹荆亚)^{1, 2}, WU Qian-hong(吴堑虹)^{1, 2}, LI Huan(李欢)³,

OUYANG Cheng-xin(欧阳承新)^{1, 2, 4}, KONG Hua(孔华)^{1, 2}, XI Xiao-shuang(奚小双)^{1, 2}

1. Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring of Ministry of Education (Central South University), Changsha 410083, China;

2. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China;

4. Engineering Research Center of Hunan Earthquake Administration, Changsha 410001, China

Central South University Press and Springer-Verlag Berlin Heidelberg 2017

Abstract: Major and REE geochemistry and multi-fractal analysis of two types of bauxite (primary bauxite and accumulated bauxite) ores were studied in Pingguo bauxite orefield in western Guangxi, China. The results of geochemical data show that the accumulated bauxite has a feature of high Al₂O₃ whereas relative low Fe₂O₃ and SiO₂ contents compared to the primary bauxite. The similar chondrite-normalized rare earth element (REE) patterns illustrate that they have a cognate relationship. However, the negative Ce anomalies of primary bauxite and positive Ce anomalies of accumulated bauxite indicate that the ore-forming system changed from reducing environment to oxidation environment. The results of multi-fractal spectrum and parameters of Al₂O₃, Fe₂O₃ and SiO₂ between primary bauxite and accumulated bauxite show that the distinct multi-fractal spectrum parameters reflect the different grade distribution between accumulated and primary bauxite ores. Metallogenic process from primary bauxite to accumulated bauxite is accompanied by the loss of diffluent elements (e.g., Si and S) and enrichment of stable elements (e.g., Al and Fe) in the surface environment. Among the rest, the migration mechanism of iron during the evolutionary process from primary ore to accumulated ore can be described as combined leaching and chemical weathering action with participation of sulfur.

Key words: geochemistry; multi-fractality; boxing-counting method; metallogenic mechanism; Pingguo bauxite

1 Introduction

Although there are many standards of classification for bauxite deposits, classification based on different basements present in beneath bauxite deposits is most widely used. It is well accepted that bauxite deposits can be divided into two types: karstic and lateritic. Bauxite deposits overlying paleokarstic surfaces of carbonates are called karstic bauxites, and those lying on alumosilicate rocks are named lateritic bauxites [1].

Western Guangxi holds one of the important bauxite resources in China, accounting for 16% of China’s bauxite reserves. There are two different types of karstic bauxite deposits in western Guangxi: one is primary bauxite (PB) formed during the Permian by sedimentation, and the other is accumulated bauxite (AB, also called Salento-type bauxite) which is transformed from the primary bauxite after a series of processes such as disintegration, oxidation and re-sedimentation [2]. Geochemical and mineralogical characteristics, source of ore-forming materials and metallogenic environment of these bauxite deposits have been examined by a number of researchers in recent years, and some important achievements have been made [3, 4]. Studies indicate that the drain of silicon, magnesium, alkalis and other diffluent elements with the residual enrichment of aluminum, iron and other immobile elements is necessary prerequisite during the process of bauxitization [5-7]. It is widely accepted that weathering and leaching processes lead to the erosion of silicon and other elements from PB; consequently, aluminum gets enriched passively and then bauxite quality is improved [8, 9]. However, most previous studies mainly focused on the descriptive distribution statistics of metallogenic elements (e.g., Al₂O₃, Fe₂O₃ and SiO₂) using traditional methods [10], while few of them tried to characterize this process by using quantitative data analytics. Studies on the migration mechanisms of Al₂O₃, Fe₂O₃ and SiO₂ (especially Fe₂O₃) during ore evolution from primary bauxite to accumulated bauxite are still the weak link in our understanding. What is the spatial distribution pattern for Al₂O₃, Fe₂O₃ and SiO₂ in PB and AB? How should we describe the enrichment degree of Al₂O₃, Fe₂O₃ and SiO₂ quantitatively? What is the migration mechanism of Fe during the process from PB to AB? This study examines these questions by analyzing major and REE geochemistry contents and the multi-fractal characteristics of Al₂O₃, Fe₂O₃ and SiO₂ contents in Pingguo bauxite ore, hoping to offer some new insights into the metallogenic mechanism of Pingguo-type accumulated bauxite.

2 Geological background and deposit geology

The important bauxite metallogenic belt, located in Western Guangxi, southwest portion of China (Fig. 1), is composed by Pingguo, Jingxi and Debao ore fields. This fourth-largest bauxite metallogenic belt in China lies in the transition zone between Yangtze Block and Cathaysian Block. Long periods of tectonic inactivity from the Upper Devonian to the Middle Permian lead to the ultra-thick accumulation of shallow sea carbonate sediments in this area [3]. The subsequent Dongwu movement caused a large-scale regression event in the continental interior of South China beginning in the Middle Permian [11]. Consequently, the limestone of Middle Permian Maokou Formation was exposed at the surface and the original soil and regolith were converted from carbonates by chemical weathering, thereby providing abundant materials for bauxite formation [12]. Alternatively, it has been proposed that volcanic eruptions in the Emei area contributed some,or even the main part, of the ore-forming materials [13]. In the late Permian, abundant laterite (considered metallogenic materials) together with the large-scale transgression event and an appropriate paleoclimate contributed to the formation of primary bauxite deposits in western Guangxi. Cenozoic tectonic events lifted the deposited bauxite-bearing layers which were exposed and broken- up and accumulated in the karstic depressions and the primary bauxite deposits during the events and the primary bauxite deposits were transformed into accumulated bauxite ores [2].

Fig. 1 Simplified geological map of Pingguo bauxite orefield, Guangxi, China

The Pingguo bauxite orefield is located in the western Guangxi Province and consists of Nadou, Taiping, and Guohua deposits. The main exposure strata in this area are Devonian sandstone and carbonates, Carboniferous carbonates, Permian limestone, Triassic sandstone and carbonate, as well as Quaternary glutenite (Fig. 1). A series of NW trending anticlines controlled distribution pattern of orebodies. The primary bauxite orebody which lies between the Late Permian Heshan Formation and the Middle Permian Maokou Formation has a varied thickness (0.5-3m) due to the uneven carbonate basement (Fig. 2(a)). The accumulated bauxite ores with laterites, which have a complicated configuration, are strictly controlled by geomorphology and karstification. The accumulated bauxite ores occur in the karst depressions within the Quaternary laterite profiles; from bottom to top the ore-bearing layer is composed of a lateritic layer, ores mixed with laterites, and clay stratum ((Fig. 2(b)). Both primary and accumulated ores show a pisolitic,oolitic honeycomb or massive structure (Figs. 2(c) and (d)). XRD analysis indicates that diaspore is the main mineral for PB, followed by anatase and goethite, with kaolinite, hematite, muscovite and pyrite as rare minerals (Fig. 2(e)). In contrast, the major minerals of AB are diaspore, hematite, kaolinite, anatase, gibbsite, goethite and rutile (Fig. 2(f)).

3 Methodology

3.1 Major geochemistry of Al₂O₃, Fe₂O₃ and SiO₂ compositions

In this work, 99 samples of PB were collected from 17 drills, and 1527 samples of AB were collected from over a hundred shallow wells. The sampling area can be found in Fig. 1. According to the national standards of China, Al₂O₃, Fe₂O₃ and SiO₂ compositions of each sample were determined by EDTA titrimetric method, potassium dichromate titration method and gravimetric-molybdenum blue method. The analytical methods follow the procedures described by MENG et al [14], LIU and XU [15] and WANG [16], respectively.

Fig. 2 Field and micro-photographs characters of bauxite ore

3.2 REE geochemistry

Nine PB samples and six AB samples collected from Nadou mining area were chosen for rare earth elements (REE) analysis which was carried out in the Key Laboratory of Metallogenic Prediction of Nonferrous Metals, Central South University, Changsha. The samples were analyzed by using inductively coupled plasma mass spectrometry (ICP-MS) VG PQ-EXCELL with analytical uncertainties generally better than 5%.

3.3 Multi-fractal method description：boxing-counting method(BCM)

HALSEY et al [17] proposed the multi-fractal analysis of moments approach, and several methods for multi-fractal calculation have gained wide acceptance. Among them, the boxing-counting method (BCM) is widely used due to its computational simplicity and high reliability [18]. Based on generalized dimensions and singularity spectrum functions, the computation procedures of BCM can be summarized as follows.

Firstly, suppose that R is the length of the ore grade sequence, and R is divided into n nonoverlapping boxes. Measurement p_i in each box can be defined as

                                    (1)

where m_i is the sum of the ore grades in box i; m_τ is the sum of the total ore grades in the sequence; i=0, 1, 2, …, +∞; n(r)=R/r. Then the partition function x(q) is defined as

                                 (2)

where q is the factor of weight and is also an index variable.

Then, we define the mass exponent τ(q) which was first introduced by HENTSCHEL and PROCACCIA [19], and τ(q) can be calculated by the following formula:

                           (3)

Consequently, we can obtain the singularity exponent a(q) and function of multi-fractal spectrum f(a) by the Legendre transform and formulas which are expressed as follows:

                               (4)

                           (5)

where a is the index of singularity strength, which can be used to characterize the singularities of ore grade sequences. The values of a_min and f(a_min) can be calculated when the weight q is maximal; a_max and f(a_max) are derived when q is minimal. When q is equal to 0, the curve of the multi-fractal spectrum reaches the culmination. The shape and the singularity spectrum of f(a)-curve imply crucial information about the internal distribution features of the study object. Generally speaking, the shape of the multi-fractal spectrum curve is concave downward. The widths of multi-fractal spectrum curves Δa in the spectrum can be defined by the following equation:

Δa=a_max-a_min                                (6)

Meanwhile, the height difference Δf is the essential parameter in multi-fractal analysis. Here, we can obtain the values of Δf from formulas below:

Δf=f(a_max)-f(a_min)                             (7)

BCM is carried out over the massive Al₂O₃, Fe₂O₃ and SiO₂ grade data of samples from PB and AB. The function of multi-fractal spectrum is calculated by Eqs. (1)–(7) and value range of q is taken from –5 to 5 with increment at 0.5. The associated r values are 0.025, 0.05, 0.1, 0.2, 0.25 and 0.5, respectively. Accordingly, computational procedure is done by the VBA program for excel.

4 Results

4.1 Statistics of Al₂O₃, Fe₂O₃ and SiO₂ compositions

Table 1 illustrates the main descriptive statistics for Al₂O₃, Fe₂O₃ and SiO₂ contents and related parameters of all samples. The chemical compositions of samples from PB are dominated by Al₂O₃ 30.7%-73.21% (average 51.98%), Fe₂O₃ 3.04%-55.36% (average 11.50%), and SiO₂ 2.53%-55.36% (average 12.16%). Ratios of Al-to-Si and Al-to-Fe range from 1.45 to 42.56 and 0.92 to 10.91, respectively, with average values at 7.99 and 6.63. Samples from AB contrast significantly, with distinct chemical compositions: Al₂O₃ 35.58%- 71.6% (average 67.17%), Fe₂O₃ 1.04%-20.28% (average 3.42%), and SiO₂ 0.49%-13.01% (average 10.39%). Ratios of Al-to-Si and Al-to-Fe range from 1.26 to 65.69 and 1.35 to 36.39, respectively, with average values of 22.65 and 8.28. Based on the data in Table 1, we can conclude that the Al concentration increases and the ore quality improves from PB to AB. However, we do not have sufficient data to accurately map the distribution of chemical compositions in the sample area.

4.2 REE compositions

The results of the REE analyses are listed in Table 2.

Table 1 Contents of Al₂O₃, Fe₂0₃ and SiO₂ and related characteristic elemental ratios at Pingguo bauxite deposit

Table 2 Rare earth element contents in samples from Pingguo Bauxite, western Guangxi.

The samples of PB have variably high total REE contents (106×10^-6 to 692×10^-6), and the chondrite- normalized REE patterns display right-dipping distribution patterns (Fig. 3(a)). LREE/HREE ratios of PB range from 1.9 to 5.19, with a medium negative Eu anomalies (δEu=0.45 to 0.58) and significant negative or weak positive Ce anomalies (δCe=0.36 to 1.09). In general, the samples of AB have the similar REE distribution patterns (Fig. 3(b)) and the REE contents range from 106×10^-6 to 530×10^-6. LREE/HREE ratios of PB range from 0.97 to 3.93, with a medium negative Eu anomalies (δEu=0.51 to 0.76) and significant positive Ce anomalies (δCe=1.13 to 2.78). In summary, both the PB and AB are characterized by enrichment of LREE and relatively right-dipping distribution patterns, with a deep “V” shape.

4.3 Multi-fractal characteristics

The results of multi-fractal analysis using BCM method are listed in Table 3. Δa and Δf are important parameters in multi-fractal analysis. Δa can provide a measure of multi-fractal spectrum. The higher value of Δa represents a heterogeneous pattern of grades distribution, while, on the contrary, the lower value of Δa indicates that the pattern of grade distribution is homogeneous. In other words, higher Δa values imply more clustered distributions and higher degrees of singularity [21, 22]. A positive Δf value means that the multi-fractal spectrum is right-hooked, indicating that smaller values within the dataset hold a dominant position. In contrast, a negative Δf value represents a left-hooked multi-fractal spectrum, implying that the dataset is dominated by larger values [21, 22]. So, a higher Δa and lower (negative) Δf are the essential parameters for identifying a better mineralization area or a higher quality deposit [21].

Fig. 3 Chondrite-normalized REE patterns of PB (a) and chondrite-normalized REE patterns of AB (b) (Normalized values for chondrite are from Taylor and McLennan [20])

Table 3 Multi-fractal parameters of Al₂O₃, Fe₂O₃ and SiO₂ contents in Pingguo bauxite deposit

As shown in Figs. 4(a) and (b), the spectrum curves of τ(q) and f(a) are concave functions with q, which confirm the multi-fractal behavior of these sequences. The Δa values of Al₂O₃, Fe₂O₃ and SiO₂ in AB are higher than those in PB respectively, whereas the Δf values of Al₂O₃ and Fe₂O₃ in AB are negative and obviously lower than those in PB. However, the Δf values of SiO₂ in both PB and AB are positive and Δf values of PB are larger.

5 Discussion

5.1 Significance of multi-fractal analysis of PB and AB

As a significant branch of nonlinear mathematical analysis, fractal methods are used to describe the distribution of inhomogeneous or irregular objects, and have been demonstrated to be very useful to reveal the extent of inhomogeneity or irregularity [23-26]. Multi-fractal analysis is applied to describe the object requiring many fractal indices to characterize its scaling properties. In recent years, due to the development of computer science and powerful software (GIS in particular), multi-fractal analysis has been widely used in geo-science fields such as oil and gas indices determination [27], resource potentiality appraisement [28] and geological phenomena discrimination [29, 30]. With the application of multi-fractal theory, a new relationship between geoscience and computer science, with great promising prospects, has been established.

Multi-fractal analysis reveals that the singularity and heterogeneity of multi-fractal distributions of Al₂O₃, Fe₂O₃ and SiO₂ in AB are higher than those in PB, indicating that some specific grades of Al₂O₃, Fe₂O₃ and SiO₂ in AB are higher than those in PB. The smaller negative Δf values of Al₂O₃ and Fe₂O₃ and larger positive Δf values of SiO₂ in AB indicate that the high grades of Al₂O₃ and Fe₂O₃ have predominance in AB over those in PB while low grades of SiO₂ in both PB and AB dominate. Based on the analysis above, the conclusion can be drawn that the contents of Al₂O₃ and Fe₂O₃ get enriched and the content of SiO₂ gets depleted during the evolutionary process from PB to AB.

Fig. 4 Multi-fractal spectra of Al₂O₃, Fe₂O₃ and SiO₂ in Pingguo bauxite deposit

In order to better understand the PB to AB evolutionary process and the degree of clustering of Al₂O₃, Fe₂O₃ and SiO₂, we performed a contrastive analysis using a new parameter, Δw which is defined as

                            (8)

As summarized in Table 2, the Δw values of Al₂O₃, Fe₂O₃ and SiO₂ are -0.680, -0.291 and 0.056, respectively. This implies that although Al₂O₃ and Fe₂O₃ get enriched, the enrichment degree of Al₂O₃ is higher than that of Fe₂O₃. Consequently, we can speculate that the Fe₂O₃ content is likely diluted compared with Al₂O₃. Moreover, the multi-fractal analysis can give quantitative evidence regarding the variation of Al₂O₃, Fe₂O₃ and SiO₂ contents from PB to AB.

5.2 Metallogenic mechanism of bauxite from PB to AB

The distribution of aluminum, iron, silicon and other elements in PB and AB are a function of their respective geochemical characters, and the similar REE distribution patterns show the homologous relation between PB and AB. In general, aluminum, iron and silicon will migrate away during the process of weathering from PB to AB. However, vast losses of silicon and other elements lead to the enrichment of aluminum and iron passively. Previous studies have placed emphasis on the loss of silicon and the diffluent elements [31], while further investigation is still necessary to understand the active geochemical mechanisms of iron in the transformation from PB to AB. Although a general understanding of variation of Al₂O₃, Fe₂O₃ and SiO₂ during the ore-forming process was obtained by multi-fractal analysis, more evidence is still needed. It is clearly demonstrated that the Al₂O₃ contents display a more negative relationship with Fe₂O₃ and SiO₂ in AB than those in PB (Figs. 5(a)-(d)). Moreover, the multi-fractal analyses above indicate that a massive loss of iron occurs during the transition process. Losses of silicon and iron play significant roles in improving qualities of AB and while we know that SiO₂ can be leached out in the way of kaolinite, and the loss mechanism of kaolinite has been researched deeply [2]. The loss mechanism of iron is rather complicated and it needs more attention. Recent studies have shown that the metallogenic physical-chemical environment has a great impact on the evolution of bauxite minerals [1]. As a variable valence rare earth element, Ce is sensitive to changes in chemical environment [32, 33]. The occurrence of positive Ce anomalies in AB and negative Ce anomalies in AB are due to variation in the oxidation state of Ce (from Ce³⁺ to Ce⁴⁺). Obviously, the different Ce anomalies in PB and AB illustrate that the ore-forming process of PB was likely to be under deoxidized conditions and AB was formed in an oxidation environment. Moreover, pyrite is found in PB (Fig. 2(e)) and iron minerals mainly occur as ferric iron in AB (Fig. 2(f)). So, the conclusion can be drawn that the evolutionary process from PB to AB took place under the conditions of epigenetic oxidation. In addition, sulfur removal from PB could be a significant factor in this transition, and the probable reactions of this process can be expressed in the chemical equation bellow:

4FeS₂+15O₂+8H₂O=4FeO(OH)+8H₂SO₄          (9)

Sequentially, another chemical reaction may proceed:

2FeO(OH)+3H₂SO₄=Fe₂(SO₄)₃+4H₂O           (10)

Fig. 5 Plots of Al₂O₃ vs Fe₂O₃ of PB (a), Al₂O₃ vs SiO₂ of PB (b), Al₂O₃ vs Fe₂O₃ of AB (c) and Al₂O₃ vs SiO₂ of AB (d)

Since Fe₂(SO₄)₃ dissolves easily in water, it can be carried to the bottom of AB by leaching process. However, due to the alkaline barrier formed by the underlying carbonates, iron sulfates were precipitated and oxidized into hematite. Eventually, deposition and weathering resulted in a layer of red clay at the bottom of sequences. Based on the assumption of an improvement in the intensity of the leaching and improved drainage conditions, iron, silicon and other impurities were eroded away during the evolutionary process from PB to AB. Thus, the qualities of AB were greatly improved and this made AB a significant bauxite type in western Guangxi.

6 Conclusions

1) The chemical compositions of PB and AB have different features. Compared to PB, AB is characterized by high Al₂O₃ contents, high ratios of Al₂O₃-to-SiO₂ and Al₂O₃-to-Fe₂O₃ with low compositions of Fe₂O₃ and SiO₂.

2) Chondrite-normalized REE patterns of both PB and AB display right-dipping distribution characteristics with LREE abundances and negative Eu anomalies. However, the negative Ce anomalies of PB and positive Ce anomalies of AB indicate that the ore-forming environment was transformed from a reducing environment to an oxidizing environment.

3) The multi-fractal boxing-counting method was used for the analysis of Al₂O₃, Fe₂O₃ and SiO₂ contents of primary and accumulated bauxite ore in the Pingguo bauxite deposit, western Guangxi. Al₂O₃, Fe₂O₃ and SiO₂ contents of bauxite present a characteristic multi-fractal distribution. During the evolutionary process from a primary bauxite deposit to an accumulated bauxite ore, Al₂O₃ and Fe₂O₃ contents get enriched whereas SiO₂ content is depleted.

4) Plausible geochemical mechanisms for the evolution of AB from PB were discussed. Ore quality of accumulated bauxite is improved due to continuous draining of silicon and iron. However, the migration of iron is via the transformation from pyrite in primary bauxite to soluble iron sulfates in accumulated bauxite. This metallogenic process is of great significant and makes the accumulated bauxite into an important ore in western Guangxi, with high Al₂O₃, high Al-to-Si, low Fe₂O₃ and low SiO₂.

References

[1] WANG Qing-fei, DENG Jun, LIU Xue-fei, ZHANG Qi-zuan, LI Zhong-ming, KANG Wei, CAI Shu-hui, LI Ning. Review on research of bauxite geology and genesis in China [J]. Geology and Exploration, 2012, 48(3): 430-448. (in Chinese)

[2] YU Wen-chao, WANG Rui-hui, ZHANG Qi-lian, DU Yuan-sheng, CHEN Yue, LIANG Yu-ping. Mineralogical and geochemical evolution of the Fusui bauxite deposit in Guangxi, South China: From the original Permian orebody to a Quarternary Salento-type deposit [J]. Journal of Geochemical Exploration, 2014, 146: 75-88.

[3] LIU Xue-fei, WANG Qing-fei, DENG Jun, ZHANG Qi-zuan, SUN Si-lei, MENG Jian-yin. Mineralogical and geochemical investigations of the Dajia Salento-type bauxite deposits, western Guangxi, China [J]. Journal of Geochemical Exploration, 2010, 105(3): 137-152.

[4] WEI Xiao, JI Hong-bing, LI Dao-jing, ZHANG Feng-lei, WANG Shi-jie. Material source analysis and element geochemical research about two types of representative bauxite deposits and terra rossa in western Guangxi, southern China [J]. Journal of Geochemical Exploration, 2013, 133: 68-87.

[5] . Geological and geochemical evolution of the Bolkarda06i bauxite deposits, Karaman, Turkey: Transformation from shale to bauxite [J]. Journal of Geochemical Exploration, 2013, 133: 118-137.

[6] . Al (Fe, Ti, Si)-mobility and secondary mineralization implications: A case study of the karst unconformity diasporite-type bauxite horizons in Milas (), Turkey [J]. Journal of African Earth Sciences, 2011, 60(3): 175-195.

[7] MONGELLI G, BONI M, BUCCIONE R, SINISI R. Geochemistry of the Apulian karst bauxites (southern Italy): Chemical fractionation and parental affinities [J].Ore Geology Reviews, 2014, 63: 9-21.

[8] LIU Xue-fei, WANG Qing-fei, ZHANG Qi-zuan, FENG Yu-wen, CAI Shu-hui. Mineralogical characteristics of the superlarge Quaternary bauxite deposits in Jingxi and Debao counties, western Guangxi, China [J]. Journal of Asian Earth Sciences, 2012, 52: 53-62.

[9] ZHU Rui-qin, XI Xiao-shuang, WU Qian-hong, YANG Zhen. Karst landform evolution and mineralization of accumulative bauxite in Pingguo Bauxite Ore, Guangxi [J]. Journal of Central South University: Science and Technology, 2011, 42(3): 744-751. (in Chinese)

[10] ZHU Rui-qin, XI Xiao-shuang, WU Qian-hong, YANG Zhen. Enclosed environment for mineralization of accumulative bauxites in Pingguo, Guangxi [J]. Geotectonica et Metallogenia, 2011, 35(3): 386-393. (in Chinese)

[11] HE Bin, XU Yi-gang, CHUNG Sun-ling, XIAO Long, WANG Ya-mei. Sedimentary evidence for a rapid, kilometer-scale crustal doming prior to the eruption of the Emeishan flood basalts [J]. Earth and Planetary Science Letters, 2003, 213(3): 391-405.

[12] LIU Chen-sheng, JIN Zhong-guo, GUO Jian-hua. Facies analysis of sedimentary bauxite deposition in freshwater of Wuzhengdao in northern Guizhou [J]. Journal of Central South University: Science and Technology, 2015, 46(3): 962-969. (in Chinese)

[13] DENG Jun, WANG Qing-fei, YANG Shu-jun, LIU Xue-fei, ZHANG Qi-zuan, YANG Li-qiang, YANG, Yue-heng. Genetic relationship between the Emeishan plume and the bauxite deposits in western Guangxi, China: Constraints from U–Pb and Lu–Hf isotopes of the detrital zircons in bauxite ores [J]. Journal of Asian Earth Sciences, 2010, 37(5): 412-424.

[14] MENG Ya-dong, SUN Lu-xin, FU Xiao-qiang, YANG Gui-song, SHEN Kai-hong. Determination of aluminum in bauxites by fluorate replacement-EDTA titrimetry [J]. Rock and Mineral Analysis, 2008, 27(6): 475-476. (in Chinese)

[15] LIU Guan-long, XU Jun-hong. Rapid determination of titanium and iron in ilmenite by potassium dichromate titrimetry [J]. Metallurgical Analysis, 2012, 32(3): 74-76. (in Chinese)

[16] WANG Jing. Chemical analysis method for magnesia and magnesia-alumina refractories-gravimetric-molybdenum blue photometric method for determination of silicon dioxide content [J]. China's Refractories, 2007, 16(1): 39-41.

[17] HALSEY T, JENSEN M, KADANOFF L, PROCACCIA I, SHRAIMAN B. Fractal measures and their singularities: The characterization of strange sets [J]. Physical Review A, 1986, 33(2): 501-511.

[18] AGTERBERG F. Fractals and spatial statistics of point patterns [J]. Journal of Earth Science, 2013, 24(1): 1-11.

[19] HENTSCHEL H, PROCACCIA I. The infinite number of generalized dimensions of fractals and strange attractors [J]. Physica D: Nonlinear Phenomena, 1983, 8(3): 435-444.

[20] TAYLOR S, MCLENNAN S. The continental crust: its composition and evolution [M]. Oxford: Oxford Press, 1985.

[21] WANG Qing-fei, DENG Jun, LIU Huan, WANG Yan-ru, SUN Xiang, WAN Li. Fractal models for estimating local reserves with different mineralization qualities and spatial variations [J]. Journal of Geochemical Exploration, 2011, 108(3): 196-208.

[22] DENG Jun, WANG Qing-fei, WAN Li, LIU Huan, YANG Li-qiang, ZHANG Jing. A multifractal analysis of mineralization characteristics of the Dayingezhuang disseminated-veinlet gold deposit in the Jiaodong gold province of China [J]. Ore Geology Reviews, 2011, 40(1): 54-64.

[23] CHENG Qiu-ming. Multifractal modelling and spectrum analysis: Methods and applications to gamma ray spectrometer data from southwestern Nova Scotia, Canada [J]. Science in China, 2006, 49(3): 283-294.

[24] LIU Ning-qiang, YU Chong-wen. Analysis of onset and development of ore formation in Dajishan tungsten ore area, Jiangxi Province, China [J]. Journal of Earth Science, 2011, 22: 67-74.

[25] AGTERBERG F. Multifractals and geostatistics [J]. Journal of Geochemical Exploration, 2012, 122(11): 113-122.

[26] SIDDIQUI S, SOLDATI M, CASTALDINI D. Appraisal of active deformation from drainage network and faults: inferences from non-linear analysis [J]. Earth Science Informatics, 2015, 8(1): 233-246.

[27] XIE Shu-yun, BAO Zheng-yu, SU Jiang-yu, JIA Xian-qiao. Multifractal properties of oil and gas indexes in the Southern Aixieke-Sangtamu Region of Xinjiang, China [J]. Geological Review, 2005, 51(1): 107-112. (in Chinese)

[28] CHENG Q M, AGTERBERG F, BONHAM G. A spatial analysis method for geochemical anomaly separation [J]. Journal of Geochemical Exploration, 1996, 56(3): 183-195.

[29] ZHANG Yan, ZHOU Yong-zhang, WANG Lin-feng, WANG Zheng-hai, HE Jun-guo, AN Yan-fe. Mineralization-related geochemical anomalies derived from stream sediment geochemical data using multifractal analysis in Pangxidong area of Qinzhou-Hangzhou tectonic joint belt, Guangdong Province, China [J]. Journal of Central South University, 2013, 20(1): 184-192.

[30] MAMELI P, MONGELLI G, OGGIANO G, DINELLI E. Geological, geochemical and mineralogical features of some bauxite deposits from Nurra (Western Sardinia, Italy): Insights on conditions of formation and parental affinity [J]. International Journal of Earth Sciences, 2007, 96(5): 887-902.

[31] MONGELLI G. Ce-anomalies in the textural components of Upper Cretaceous karst bauxites from the Apulian carbonate platform (southern Italy) [J]. Chemical Geology, 1997, 140(1): 69-79.

[32] WANG Xiao-mei, JIAO Yang-quan, DU Yuan-sheng, LING Wen-li, WU Li-quan, CUI Tao, ZHOU Qi, JIN Zhong-guo, WENG Shen-fu. REE mobility and Ce anomaly in bauxite deposit of WZD area, Northern Guizhou, China [J]. Journal of Geochemical Exploration, 2013, 133: 103-117.

[33] ABEDINI A, CALAGARI A. REE geochemical characteristics of titanium-rich bauxites: The Permian Kanigorgeh horizon, NW Iran [J].Turkish Journal of Earth Sciences, 2014, 23(5): 513-532.

(Edited by YANG Hua)

Cite this article as: CAO Jing-ya, WU Qian-hong, LI Huan, OUYANG Cheng-xin, KONG Hua, XI Xiao-shuang. Metallogenic mechanism of Pingguo bauxite deposit, Western Guangxi, China: Constraints from REE geochemistry and multi-fractal characteristics of major elements in bauxite ore [J]. Journal of Central South University, 2017, 24(7): 1627-1636. DOI: 10.1007/s11771-017-3568-8.

Foundation item: Project(GX2007CAQB01) supported by the Key Research Project of Aluminum Corporation of China Limited; Project(41502067) supported by the National Natural Science Foundation of China

Received date: 2015-12-16; Accepted date: 2016-08-01

Corresponding author: WU Qian-hong, Professor; Tel: +86-731-88877077; E-mail: qhwu19@163.com