J. Cent. South Univ. (2012) 19: 1741-1752

DOI: 10.1007/s11771-012-1201-4

Sources and control factors of rare earth elements in Late Permian mudstones, Southwest China

XIAO Ming-guo(肖明国)^{1, 2}, Zhuang Xin-guo(庄新国)², Yi Wei(易炜)², Mao Wan-hui(毛婉慧)²

1. Key Laboratory of Tectonics and Petroleum Resources (China University of Geosciences),

Ministry of Education, Wuhan 430074, China;

2. Institute of Sedimentary Basin and Mineral, Faculty of Earth Resources, China University of Geosciences,Wuhan 430074, China

? Central South University Press and Springer-Verlag Berlin Heidelberg 2012

Abstract: The material sources and control factors of rare earth elements (REEs) for 25 borehole bulk samples from the Late Permian Longtan Formation in Mount Huaying (borehole number: ZK10-6), Sichuan Province, South China, were investigated. All samples were determined by inductively coupled plasma mass spectrometry (ICP-MS). The chondrite-normalized distribution patterns of mudstone samples are uniform. All samples belong to the light rare earth element (LREE)-rich type and are enriched in LREEs relative to heavy rare earth elements (HREEs). The distribution curves of REEs in mudstone are highly similar to Mount Emei basalt and the three periods of REEs enrichment correspond to three Mount Emei basalt eruption cycles in Longtan period. The results indicate that REE patterns are not controlled by materials from the seawater or land plants. The predominant sources of REEs are from terrigenous material as indicated by negative Eu anomaly. So, the sources of REEs are controlled by terrigenous material, and the Mount Emei basalt is the predominant source of terrigenous material. Thus, transgression-regression is another control factor of REEs enrichment.

Key words: rare earth elements (REEs); control factors; material sources; transgression-regression; Mount Emei basalt; Late Permian

1 Introduction

Rare earth elements (REEs) have been used as tracers to identify sources and epigenetic modi?cation of mineral matter. In addition, the potential economic value of REEs in coal measure stratum has attracted much attention. Also, mudstone records of rare earth elements have been used widely as an oil and gas geochemical tracer to investigate the petroleum and natural gas exploration. The sources of REEs in mudstone are diverse, with one or several dominant sources in a specific coal-forming and petroliferous basin. According to the research methods and previous studies [1-4], a total of twenty-five samples of Mount Huaying were studied to shed a light on the material source of REEs in Mount Huaying mudstone.

Meanwhile, Sichuan Basin is a favorable enrichment area for hydrocarbons in China, and Longtan Formation is an important hydrocarbon source rock bed and coal bed. The relationship between Mount Emei basalt and mineralization attracts great concern by more and more researchers, but little attention had been paid to the relationship between Mount Emei basalt and REEs of mudstone. In addition, a great deal of geochemical studies have been done on the Late Permian in the past, but the research object was coal but not mudstone. However, CHEN et al [5-8] showed that multi-tuffs interbedded with coal measure strata and interbedded kaolinite-mudstones appeared in many mudstone members, which were transformed and altered by acidic or alkaline volcanic ashes. To understand the relationship between REEs enrichment in mudstone and Mount Emei basalt, the authors conducted systematic REEs geochemical studies for the Late Permian mudstone from Mount Huaying in eastern Sichuan Province, China.

2 Geological setting

The Mount Huaying in the eastern part of Sichuan Province (see Fig. 1), Southwestern China, is an important energy base in this province. From the paleogeographical point of view, the Mount Huaying district is located in the north west part of the Yangtzi platform. The Late Permian Mount Huaying area is formed as part of the South China interior epicontinental sea basin in the Late Permian and is located in the northwest part of this epicontinental sea basin (see    Fig. 2). The sedimentary facies in this area mainly consist of carbonate platform connecting to tidal-flat sediments. All the mudstones of Longtan Formation show evidence of marine influence in the depositional environment [4]. From top to bottom in the borehole ZK10-6, the number of limestone beds in the Late Permian reduces considerably, and the number of mudstones and the thickness of mudstones beds increase (see Fig. 3).

Fig. 1 Location map of study area, borehole sampling point and simplified geological map showing study area, Mount Huaying, South China

Fig. 2 Paleogeographic map of Late Permian in South China (modified from Ref. [4]): 1-Old land; 2-Subaqutic uplift; 3-Carbonate platform; 4-Littoral sediments with dominant carbonate; 5-Coarse clastic sediments in the fault trough; 6-Bay sediments; 7-Clastic seashore sediments; 8-Fluvial plain and lake sediments; 9-Delta sediments; 10-Alluvial fan sediments; 11-Direction of source clastic; 12-Direction of transgression

The Longtan Formation was divided into five members in the Mount Huaying from bottom to top (Long 1-Long 5, as shown in Fig. 3). The Longtan Formation is made up of limestone with frequent flint nodules, siltstone, fine sandstone, mudstone, thin argillaceous limestone, and coal seams with common pyrite and siderite nodules. Mudstone is mainly present in Long 1, Long 3 and Long 5 member (containing twelve-layer mudstone beds with accumulative thickness of 24.2 m) (see Fig. 3).

3 Samples and analytical procedures

All samples were sampled vertically in a same borehole ZK10-6 (see Table 1 and Fig. 3). A total of 25 representative mudstone bulk samples from Mount Huaying were taken from fresh faces. The samples were collected and stored in plastic bags to prevent contamination and weathering. The types of samples of stratigraphic position are shown in Fig. 3. Except for some special purpose studies, fresh rock without alteration was collected for samples. Acquisition position should try to avoid all kinds of contact zones, alteration zones and fractured zone. Samples collection should be perpendicular to its direction. If the study within the same layer of rock composition varies along the  direction, we can go a certain distance along its system of collecting samples. The specific method of samples collection followed the Chinese National Standard for Collecting Borehole Samples GB 482—1995 [9]. The bulk samples were air-dried, milled and split until a representative split of 0.5 kg was obtained. The samples for chemical analyses were pulverized to less than 75 μm and dried for 12 h in a desiccator. All samples were acid digested following a two-step digestion method devised to retain volatile elements in solution [10]. The resulting solutions were analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) for REEs in the Institute of Earth Sciences, Barcelona, Spain. The analysis accuracy (RSD) was estimated to be less than 5%. Chondritic values by HERRMANN [11] were used for normalization with the modification of REEs. The concentration of REEs in the mudstone samples from Mount Huaying, as well as in basalt, seawater, coral and chondrite are listed in Table 1.

Fig. 3 Upper Permian Longtan Formation composite geological columnar section of borehole ZK10-6 in Mount Huaying (Geologic age from 2009 geologic time scale, reported by International Commission on Stratigraphy)

Table 1 REE contents (in μg/g) in 25 mudstone samples from Mount Huaying and associated geochemical parameters

4 Results and discussion

4.1 Sources of REEs

4.1.1 Marine

The chondrite-normalized REE pattern of modern seawater is characterized by two main features: negative Ce anomalies and relative enrichment of the heavy REE. Positive La anomalies and high Y/Ho ratios could also be added to the list. These characteristic features, which should be ubiquitous in past oceans, too, were seawater chemistry similar to today. All samples of study area do not have this feature. At the same time, seawater is enriched with both LREE and HREE (see Table 1). However, the REEs concentration is very low. Cerium is negative anomaly in seawater, because Ce³⁺ is oxidized into Ce⁴⁺, and is preserved in solution in the form of CeO₂, but other REEs retain +3 in the ocean conditions. In all the 25 samples, δ(Ce) values are in the range of 0.94-1.14, and their arithmetic mean is 1.05. Ce shows very low anomaly (see Table 1), indicating that REEs in mudstone from Mount Huaying are not derived from marine material.

The concentration of REEs in seawater and coral in modern ocean and in mudstone from Mount Huaying are listed in Table 1. Figure 4 is drawn from reference values of Table 1 (seawater and coral). From Fig. 4, it can be seen that the concentrations of REEs in mudstone are significantly higher than those in seawater and corals, including ∑REE, HREE and LREE. So, REEs in mudstone from Mount Huaying is unlikely to come from ocean where the concentration of REEs is too low.

A study of REEs in representative continental shelf seabed sediment in Chinese Bohai Sea, Yellow Sea, East Sea and South Sea by ZHAO et al [13] shows that REEs in modem continental shelf seabed sediments are mainly from mainland weathered rock, instead of seawater or biological debris, because the concentrations of REEs in the latter is too low. Also, Refs. [7, 14] show that the concentrations and distribution patterns of REEs in coal bearing formation are less affected by seawater. In fact, input of terrigenous material in?uenced the concentrations and distribution patterns of seawater and coral. To conclude, REEs in mudstone of Mount Huaying are not mainly derived from sea material.

4.1.2 Terrestrial plant

The concentration of REEs in terrestrial plants is very low. The concentration of La or Ce is no more than 1 μg/g in dry plants. The average concentration of La is 130 μg/g, and that of Yb is 13 μg/g in ash plant. An approximate calculation by ESKENAZY [15] shows that only 2%-3% of the total La concentration (an average of 4.8 μg/g) in coal or mudstone is possibly derived from plants in Pirin Deposit in Bulgaria. The abundance of La in this study area is 23.72-71.18, arithmetic mean is   39.27 μg/g (see Table 1), and only a small quantity of REEs may be derived from coal-forming plants according to the calculation model in Ref. [15].

Besides, ZHANG et al [16] concluded that there is no fractionation of REEs during uptake by roots from a soil solution, but they also reported a fractionation of REEs during transfer from roots to leaves. WANG et al [17] observed REE fractionations in plants studied with a depletion of heavy REEs (HREEs) in some and a HREE enrichment in others. A fractionation of REEs with an enrichment in middle REEs in wheat organs was also reported by DING et al [18]. The interpretations of many of these studies were based on investigations of plant–soil systems under open-?eld conditions. These complex characteristic features which should be ubiquitous in past plants, too, were plants chemistry similar to today. So, the concentration and distribution patterns of REEs in terrestrial plants are variational, complex and different, and vary with plants growth, growth pattern, different parts of plants, varied clay substrates and so on. So, the comparison of the REE patterns from plants and mudstone in study area shows less significance.

Fig. 4 Chondrite-normalized REE patterns of mudstone samples (mean value of 25 samples), seawater and coral

4.1.3 Terrigenous material

The major inorganic minerals in mudstone are clay mineral, quartz and pyrite, but the predominant one is clay mineral. In the clay mineral, kaolinite is the most important one [19]. REEs are bound predominantly to the lithophile of the mineral matter in the mudstones. Correlation analysis and cluster analysis show that REEs of study area are positively correlated with Al (see    Fig. 5). Although the number of samples is limited, the main purpose is to provide a preliminary analysis, and the correlation coefficient is 0.58, characterized by positive correlation. Clay mineral in mudstone is mechanically transferred from land and shows its terrigenous characteristic, and its content roughly reflects the availability of terrigenous material. It is generally believed that clay minerals play a great role in the source of REEs if the content of Al is high. The REEs content is high in the clay mineral. The chondrite-normalized distribution patterns are uniform. They are characterized by a negative Eu anomaly whose mean value is 0.61 (varying in the individual samples from 0.44 to 0.68). It is strongly supposed that the Eu anomaly is inherited from terrigenous materials.

Fig. 5 Correlation between ∑REE and Al content from studied samples

Therefore, it is suggested that the source of REEs in mudstone is mainly influenced and controlled by terrigenous material sources in Late Permian from Mount Huaying.

4.2 Relationship between REEs in mudstone and Mount Emei basalt

As previously stated, REEs source is terrigenous material. Then, what kind of terrigenous material is it? Study on this topic is not only for further trace study of REEs, but also for the verification of the above-mentioned conclusions on material sources.

4.2.1 LREE, HREE and ∑REE

The comparison of LREE, HREE and ∑REE in mudstone and Mount Emei basalt are shown in Fig. 6. It can be seen that the distribution curves are well similar. They are all enriched in light rare earth elements relative to heavy rare earth elements, and HREE depletion is the clearest in mudstone. However, such similarity just provides possible relationship between them. To demonstrate clear relationship between them, further studies are necessary.

4.2.2 REEs distribution patterns

Figure 7 shows the chondrite-normalized REE patterns for mudstone compared to Mount Emei basalt according to Table 1. It can be seen that REEs distribution curves are similar not only among mudstone (it indicates that basin in Huaying City is similar in deposition process and evolution), but also between mudstone and Mount Emei Basalt, in which REEs normalized value decreases gradually with elements from La to Lu. As a whole, the chondrite- normalized distribution patterns of mudstone samples are very similar, with slightly steep LREE and relatively flat HREE patterns. The LREE/HREE ratio is 5.99-11.81, and the mean value is 9.79, which means relatively enriched in LREEs. They are characterized by a negative Eu anomaly whose mean value is 0.61. The results on negative δ(Eu) anomaly suggest that REEs in the Mount Huaying mudstone are controlled by terrigenous materials.

The Mount Emei Continental flood basalts (ECFB), a type of large igneous provinces (LIPs), erupt in Late Permian in Mount Emei-Panzhihua-Kunming area (southwest of study area), west Sichuan (see Fig. 8). Further, Mount Emei basalt eruption is the most important thermal event in the Late Paleozoic western margin of the Yangtze plateform. The Mount Emei flood basalt is a multiple layer of igneous rock from large mantle plume volcanic eruption. These eruptions are so important, which are associated with the so-called end-Guadalupian (Permian) extinction. The duration of the Mount Emei LIP corresponds with the period from Maokou age to Longtan age in Mount Huaying (253.8- 260.4 Ma, as shown in Fig. 3), because the termination age of Mount Emei volcanism is about (257±4) Ma and intermittence and eruption cycle. So, voluminous flood volcanism is potentially responsible for global climate change and mass extinction during geologic time of Longtan period and REEs enrichment of study area [12, 20-21].

Furthermore, it is worth mentioning that Mount Huaying is located at about 300 km northeast from the Mount Emei LIPs (see Fig. 8). References [17, 19, 22-23] show that the major source in Late Permian coal bearing formation in Southwestern China is basalt, and REEs content is mainly related to the fine heavy minerals which are derived from weathered basalts, such as apatite mineral, tourmaline, barite, rutile, zircon, or celestite- yttrium phosphate rock-monazite portfolio. Moreover, according to Ref. [3], the temporal link and geochemical analyses suggest that the clay mineral of Longtan formation mudstone at the Early-Middle–Late Longtan period boundary is genetically related to the Mount Emei silicic volcanism.

Fig. 6 LREE, HREE and ∑REE distribution curves of mudstone in Mount Huaying and Mount Emei basalt

Therefore, the authors suggest that Mount Emei basalt not only controls REEs distribution pattern, but also the ultimate source of REEs for the mudstone in Late Permian from Mount Huaying.

4.2.3 Relationship between REEs enrichment and Mount Emei basalt

Comparison of LREE, HREE and ∑REE amongst the 25 mudstone samples from Table 1 shows that REEs are relatively enriched in Long 1 member (S18, S19, and S21-25 samples), Long 3 member (S8 sample) and Long 5 member (S2, S6 samples), a total of 10 samples. The arithmetic mean of LREE in 25 samples is 200.46×10^-6, but that in Long 1 member is 234.74×10^-6, in Long 3 member is 222.49×10^-6 and in Long 5 member is 208.79×10^-6. The arithmetic mean of HREE in 25 samples is 20.99×10^-6, but that in Long 1 member is 29.69×10^-6, in Long 3 member is 23.71×10^-6 and in Long 5 member is 23.13×10^-6. The arithmetic mean of ∑REE in 25 samples is 221.45×10^-6, but that in Long 1 member is 311.77× 10^-6, in Long 3 member is 246.20×10^-6 and in Long 5 member is 231.92×10^-6. As already observed, the REEs enrichment is characterized by a general upward decreasing in the LREE, HREE and ∑REE contents. The general upward decreasing trend of REEs enrichment coincides with an upward weakened Mount Emei volcanism. In other words, the period of three REEs enrichment corresponds to three eruption cycles during Longtan period. The depletion of REEs from bottom to top (Long 1-Long 5) indicates that the ability of Mount Emei volcanic eruption is weaker.

Similar to studies on neighbour region reported in Refs. [24-25], basalt eruption in Longtan Formation is classified into 2-3 cycles too. Compared with the former cycle, the eruption of the latter is weakened. In the geological period of Longtan, lithofacies include a complete system of land system, hybrid system, carbonate platform system, and slope basin system from west to east in Sichuan Basin. According to the distribution of rough gravel Mount Emei basalt, it can be inferred that there are marly volcanic eruption mouths  in Emei County, Chongqing City and Guiyang City (see Fig. 8); local southeast of basalt distribution edges into the sea. Mount Huaying and other places have started to take the shape of a barrier, which controls the formed coal, oil and gas basin.

Fig. 7 Chondrite-normalized REE abundances of 25 mudstone samples and basalt

Fig. 8 Distribution map of Mount Emei LIPs [20-21]

So, the relative enrichment of REEs in ten samples are closely related to the eruption of Mount Emei basalt, and are controlled by the Mount Emei basalt. This vertical change is related to three cycles of volcanic eruption and effusion ability in Longtan age (see Fig. 3).

4.3 Transgression-regression influences

Mount Huaying is located in the sea-land transition in Late Permian and transgression-regression cycle is essential to form coal basin. The frequent transgression- regression makes REEs enrichment change. In the transgression, the role of clay minerals (weathered crust of Mount Emei basalt) in REEs enrichment is counteracted and reduced by seawater invasion; as a result, REEs contents are low. However, during the regression, the role of clay minerals in REEs enrichment is few or free counteracted, so REEs contents are high. However, REEs normalized value pattern in mudstone samples are similar in Mount Emei basalt (see Fig. 7). This shows that the REEs of samples is controlled by weathered basaltic crust, so they have similar REEs patterns; nevertheless, REEs contents are influenced by the transportation modes and the transportation scale of clay minerals, which are controlled by the transgression- regression cycle. Therefore, the transgression-regression cycle in Late Permian coal measured from Mount Huaying can be inferred by analyzing the content variation of REEs. Accordingly, it can be divided into two major transgression and regression cycles (see   Fig. 9).

From bottom to top, the first is the low sea level, and the coal-forming environment is the regression. The role of marine decreases and the contribution of weathered basaltic crust to REE enrichment is paramount. As a result, REEs contents in S21, S22, S23, S24, and S25 samples are high in this period (Long1 member, early Longtan period). Clay mineral is transported into basin by the mechanical suspension, wherein most of REEs are preserved in the process of the coal-formation, so REEs content increases. As the transgression occurs and sea level begins to increase after S21 sample, the role of weathered crust decreases in the study area, namely, the role of clay minerals in REEs enrichment decreases. From Fig. 3, it can be seen that REEs contents in S20-S9 samples that form in the transgression environment begin to decrease. Then, S8 is formed when regression occurs. So, REEs contents in S8-S6 increase. Regression environment makes basalt full weathering and mature clay minerals are developed. The clay materials are transported to the coal-forming basin and peat, into the coal; as a result, REEs contents are enriched in this period. Then, a transgression happens, and the scale of regression is strong. The interbedded calcareous mudstone and interbedded limestone are the direct evidence of the transgression, which are found in this layer. After that, seawater gradually subsides in the study area, the role of weathered basaltic crust in REEs enrichment increases, and accordingly, REEs contents increase. REEs content in S2 sample increases to a high level. The above analysis from the transgression- regression cycle is basically consistent with previous studies. This shows that REEs content of coal measure is in a good relationship with transgression-regression, i.e., REEs content is low when transgression occurs, and REEs content is high when regression occurs (see Fig. 9). Also, the content changing trend of La, Ce, Nd elements is the same as the sea-level change.

Fig. 9 Sketch map showing correlation between REEs contents variation and transgression-regression cycles

It deserves noting that the characteristic of REEs contents variation in the coal-bearing strata according to transgression-regression has a special background, i.e., its source is Mount Emei basalt of high contents of REEs. There will not be very good correlation between REEs content variation and the transgression-regression if the sources are not basaltic weathered crust of high contents of REEs. In addition, if the REEs contents of weathered crust are low, i.e., when they are not obviously different from REEs contents of marine sediment, it is difficult to find a good correlation between REEs content variation and transgression-regression cycle. Therefore, the method is effective only as a source area that has a special geochemical characteristic.

5 Conclusions

1) The chondrite-normalized pattern of REE is characterized by slightly steep LREE, relatively flat HREE patterns and a negative Eu anomaly. Fractionation of REE in the sedimentary environment is displayed by a vertical decrease in the LREE/HREE ratio and by a very low Ce anomaly. The Eu-depletion in the mudstone is attributed to the influence of terrigenous materials. The very low Ce anomaly is attributed to the influence of continental materials and shallowness of the depositional basin.

2) Mount Emei basaltic eruption provides the major material source for REEs enrichment, and intermittence and cycle of basaltic eruption provide opportunities for REEs evolution in mudstone. Besides, the general upward weakening of the Mount Emei volcanic force corresponds to upward decrease in LREE/HREE ratio and content of REEs. Moreover, the sedimentary rhythm in Longtan Formation is synchronous with intermittence and cycle of basaltic eruption.

3) REEs normalized values in mudstone and in Mount Emei basalt have similar distribution curves and the negative Eu anomaly, which shows Mount Emei basalt controls REEs patterns and enrichment. The studies of REEs enrichment are consistent with previous sedimentologic and stratigraphic studies in reflecting the source materials, paleoenvironment and so on.

4) The REEs enrichment in the mudstone of Mount Huaying is related to marine regression. Because regression environment makes basalt full weathering, mature clay minerals are well developed. The clay materials are transported to the coal-forming basin, into peat and mudstone; as a result, REEs contents are enriched in Long 1 member, Long 3 member and Long 5 member.

Acknowledgments

We are grateful to thank anonymous reviewers for their constructive comments which greatly improved the manuscript. We also thank our Spanish co-workers (Institute of Earth Sciences, Barcelona, Spain) for the ICP-MS measurements. Field assistance by XING Feng-cun, ZHANG Feng, FU Xiao-dong, ZHANG Wen-tao are acknowledged.

References

[1] Schatzel S J, STEWART B W. Rare earth element sources and modi?cation in the Lower Kittanning coal bed, Pennsylvania: Implications for the origin of coal mineral matter and rare earth element exposure in underground mines [J]. International Journal of Coal Geology, 2003, 54(3/4): 223-251.

[2] Seredin V V. Rare earth element-bearing coals form the Russian Far East deposits [J]. International Journal of Coal Geology, 1996, 30(1/2): 101-129.

[3] HE Bin, XU Yi-gang, Zhong Yu-ting, Guan Jun-peng. The Guadalupian-Lopingian boundary mudstones at Chaotian (SW China) are clastic rocks rather than acidic tuffs: Implication for a temporal coincidence between the end-Guadalupian mass extinction and the Emeishan volcanism [J]. Lithos, 2010, 119(1/2): 10-19.

[4] ZHUANG Xin-guo, QUEROL X, ALASTUEY A, PLANA F, MORENO N, ANDR?S J M, WANG Jie-lin. Mineralogy and geochemistry of the coals from the Chongqing and Southeast Hubei coal mining districts, South China [J]. International Journal of Coal Geology, 2007, 71(2/3): 263-275.

[5] Chen Zhong-shu, WU Yu-dong. Late Permian Emeishan basalt and coal-bearing Formation in Southern Sichuan area [J]. Coal Geology of China, 2010, 22(2): 14-18. (in Chinese)

[6] Feng Ren-wei, Wang Xin-zhi, Zhang Fan, Yang Yue-ming, LI Yue-gang, QI Yan-jun, Zhang Ruo-xiang. The study on reservoir property and characteristics of the Emeishan basalts of Zhougongshan and its neighbour area in the Southwest Sichuan [J]. Acta Sedimentologica Sinica, 2008, 26(6): 913-924. (in Chinese)

[7] Dai Shi-feng, Ren De-yi, LI Shan. Geochemical and mineralogical anomalies of the Late Permian coal in the Zhijin Coalfield of Southwest China and their volcanic [J]. International Journal of Coal Geology, 2003, 55(2/3/4), 117-138.

[8] Wang Li-ting, LUO Jin-hui, Wang Chang-wei, Wang Min. Geology and coal-accumulating regularity of the paralic coal fields of Late Permian in western Guizhou [J]. Guizhou Geology, 1993, 10(4): 291-300. (in Chinese)

[9] LI Zhen-xiang, YUAN Jun, LI Xiao-bo. Compilation of coal commonly used standard [M]. Beijing: Coal Industry Press, 2000: 192-200.

[10] Querol X, Whateley M K G, Fernandez-Turiel J L, Tuncali E. Geological controls on the mineralogy and geochemistry of the Beypazari lignite, Central Anatolia, Turkey [J]. International Journal of Coal Geology, 1997, 33(3): 255-271.

[11] Herrmann A G, WEDEPOHL K H. Yttrium and lanthanides, handbook of geochemistry [M]. Berlin: Springer-Verlag, 1970.

[12] HE Bin, XU Yi-yang, Huang Xiao-long, LUO Zhen-yu, SHI Yu-he, Yang Qi-Jun, YU Song-yue. Age and duration of the Emeishan flood volcanism, SW China: Geochemistry and SHRIMP zircon U–Pb dating of silicic ignimbrites, post-volcanic Xuanwei Formation and clay tuff at the Chaotian section [J]. Earth and Planetary Science Letters, 2007, 255(3/4): 306-323.

[13] Zhao Yi-yang, Wang Jin-tu, Qin Chao-yang, CHEN Yu-wei, WANG Xian-dang, WU Ming-qing. Rare earth elements in continental shelf sediments of the China seas [J]. Acta Sedimentologica Sinica, 1990, 8(1): 37-43. (in Chinese)

[14] Ure A M, Bacon J R. Scandium, yttrium and the rare earth contents of water lily (Nuphar lutea) [J]. Geochim Cosmochim Acta, 1978, 42(6): 651-652.

[15] Eskenazy G M. Rare earth elements in a sampled coal from the Pirin deposit, Bulgaraia [J]. International Journal of Coal Geology, 1987, 7(3): 301-314.

[16] Zhang Z Y, Wang Y Q, Li F I, Xiao H Q, Chai Z F. Distribution characteristics of rare earth elements in plants from a rare earth ore area [J]. Journal of Radioanalytical and Nuclear Chemistry, 2002, 252(3): 461-465.

[17] Wang Z, LU P, Liu D. Initial assessment of the effects of rare earth elements on soil integrity [C]// GUICHERIT R, ZHU W, Ed. Proc 2nd Sino-Dutch Workshop, Environmental Behavior and Ecotoxicology of Rare Earth Elements and Heavy Metals. La Delft: TNO Publ, 1997: 1-14.

[18] DING Shi-bing, LIANG Tao, Zhang Cao-sheng, YANG Jun-cai, Zhang Zi-li. Accumulation and fractionation of rare earth elements (REESs) in wheat: Controlled by phosphate precipitation, cell wall absorption and solution complexation [J]. Journal of Experimental Botany, 2005, 56(420): 2765-2775.

[19] DU Mei-xia, Zhuang Xin-guo. A characteristic study of rare earth elements in southern china Late Permian and Late Triassic coals [J. Coal Geology of China, 2006, 18(2): 22-26.

[20] XU Yi-yang, Chung Sun-lin, Jahn B M, WU Gen-yao. Petrologic and geochemical constraints on the petrogenesis of Permian-Triassic Emeish an flood basalts in southwestern China[J]. Lithos, 2001, 58(3/4): 145-168.

[21] JASON R A, Thompson G M, Zhou M F, Song X Y. Emeishan large igneous province, SW China [J]. Lithos, 2005, 79(3/4): 475-489.

[22] Zheng Liu-gen, Liu Gui-jian, Zhang Hao-yuan, Gao Lian-fen, Xue Jian. Study on geochemistry of rare earth elements (REEs) in Permian coal from Huaibei coalfield [J]. Geological Journal of China Universities, 2006, 12(1): 41-52. (in Chinese)

[23] CHI Ru-an, LI Long-feng, WANG Dian-zuo. Studies of ion exchange equilibrium in clay minerals of adsorbed rare earth [J]. Journal of Central South University: Science and Technology, 1991, 22(2): 142-148. (in Chinese)

[24] Chen Wen-yi, Liu Jia-ren, Wang Zhong-gang, Zheng Qi-ling. Study on lithofacies palaeogeography during the Permian Erneishan basalt explosion in Guizhou Province [J]. Journal of Palaeogeography, 2003, 5(1): 17-28. (in Chinese)

[25] HE Bin, XU Yi-yang, Wang Ya-mei, Luo Zhen-yu. Sedimentation and lithofacies paleogeography in southwestern China before and after the Emeishan ?ood volcanism: New insights into surface response to mantle plume activity [J]. Journal of Geology, 2006, 114(1): 117-132.

(Edited by DENG Lü-xiang)

Foundation item: Project(40839910) supported by the National Natural Science Foundation of China

Received date: 2011-03-29; Accepted date: 2011-05-27

Corresponding author: XIAO Ming-guo, PhD Candidate; Tel: +86-18627075352; E-mail: xiaomingguo@163.com