以攜帶式X射線螢光光譜儀測定蛇紋岩土壤特徵元素之可行性

Abstract

蛇紋岩土壤具有高濃度的鎂、鉻與鎳濃度，且鈣鎂比值低，因此鈣、鎂、鉻與鎳可視為是蛇紋岩土壤的特徵元素。高濃度的鉻、鎳與特殊的鈣、鎂比例可能對農業生產與環境品質造成潛在風險，若可於現場立即得知是否為蛇紋岩土壤即可提高土壤管理與環境品質決策的效率。傳統實驗室分析以氫氟酸消化法消解樣品測定樣品元素濃度，實驗過程危險且繁瑣耗時，分析費用昂貴。攜帶式X射線螢光光譜儀 (Portable X-ray fluorescence spectrometry, pXRF) 為非破壞式分析，可快速並同時測定多種元素。藉由測定鈣、鎂、鉻與鎳含量，pXRF可作為現場快速判別蛇紋岩土壤的潛在工具。本研究以pXRF、氫氟酸消化法與王水消化法 (鉻與鎳) 測定蛇紋岩土壤化育層之特徵元素濃度以探討方法間差異，另以pXRF測定剖面中深度為每10公分一段之土壤，探討不同前處理對pXRF測定之影響，並了解不同採樣方式對於剖面元素分布之差異。三個蛇紋岩土壤剖面採自台東地區 (n = 21)，每個剖面另採集深度為每10公分一段之土壤 (n = 38)，分為風乾前、風乾後過10 mesh篩網與風乾後過100 mesh篩網等三種前處理方式，並以pXRF (Olympus Delta Premium 2000，以標準品NIST 2709、2710與2711於Geochem mode建立檢量線) 測定元素含量。結果顯示pXRF測定供試土壤中的特徵元素濃度和氫氟酸消化法之結果達顯著相關 (p < 0.01)，和王水消化法測得的鉻與鎳濃度亦達極顯著相關 (p < 0.001)。輕元素 (鈣與矽) 的直線相關性相較於其它元素較不理想，可能受Auger effect與標準品土壤之母質影響；鎂與鉻則可能因基質干擾與光譜干擾導致直線迴歸方程式偏離y = x，其中鉻也可能受礦物結構與其它元素不同而影響測值，此外以pXRF測定時亦受樣品水分含量與均質程度影響。若以化育層土壤描繪剖面的元素分布，可利用pXRF測定化育層間元素濃度劇烈變化之處，協助描繪更接近實際剖面的元素分布。本研究證實pXRF對於蛇紋岩土壤的分析能力佳，可作為現場測定蛇紋岩土壤元素濃度的輔助參考工具。相較於其它母質土壤，臺灣的蛇紋岩土壤之鈣鎂比值相對低 (小於1.0)，鉻、鎳含量極高且兩元素的相關性佳，於現場使用可快速鑑別為蛇紋岩土壤而非人為污染土壤，節省大量時間並符合經濟效益。

Serpentine soils are characterized by high Mg, Cr and Ni content and low Ca/Mg values. Hence, Ca, Mg, Cr and Ni are considered as the characteristic elements of serpentine soils. Excessive Cr and Ni and low Ca/Mg in serpentine soils may contribute to the potential risk of agricultural production and environmental quality. Identification of serpentine soils in the field rapidly is important for the efficiency of soil management and the environmental quality decision. Conventional measurement of these elements is determined in the laboratory after hydrogen fluoride (HF) digestion or aqua regia extraction. However, this technique is time-consuming, dangerous, and expensive. Portable X-ray fluorescence spectrometry (pXRF) can provide rapid screening and simultaneous multi-element determination for non-destructive analysis. Therefore, pXRF has the potential to rapid screening of serpentine soils by determining Cr, Ni, Ca and Mg in the field. The objectives of this study were to evaluate the analytical ability of pXRF by comparing HF digestion and aqua regia digestion (Cr and Ni) by determining horizon samples of serpentine soils, to evaluate the analytical abilities of pXRF for the screening of different pre-treatments of 10-cm interval samples from soil profiles (n = 38), and to understand the vertical distribution of Ca, Mg, Cr and Ni in the soil profiles by different sampling approaches. Three serpentine soil profiles were collected in Taitung (n = 21). In addition, 10-cm interval samples (n = 38) were divided into three parts of subsamples. They were fresh, sieved less than 10 mesh and sieved less than 100 mesh ones for quantifying by pXRF (An Olympus Delta Premium 2000 XRF analyzer was calibrated by NIST 2709,2710 and 2711 in Geochem mode). The experimental results indicated that the significant correlation (p < 0.01) between pXRF and HF digestion by determining characteristic elements. An excellent performance of Cr and Ni quantification between pXRF and aqua regia digestion with significant correlation (p < 0.001). However, light elements (Ca and Si) are not as relative expected linear relationships due to Auger effect and the parent material of standard reference materials. The linear regression of Mg and Cr deviated from y = x because of spectral interference and matrix interference. Besides, Cr may also be affected by the difference of mineral structure compare to other elements. Furthermore, water content and homogeneity of samples have an effect on quantifying elements by pXRF as well. pXRF can determine the elemental concentrations with dramatic variation between horizon samples, which can support to achieve the true distribution of elements in the soil profile by horizon samples. This study verifies pXRF has excellent analytical ability to serpentine soils, which can be auxiliary equipment for quantifying elements in serpentine soils in the field. Comparing to other parent material soils, serpentine soils in Taiwan with low Ca/Mg (< 1.0), high Cr and Ni, and great correlation between Cr and Ni, which can rapidly identify serpentine soils rather than anthropogenic polluted soils by pXRF with saving considerable time and affording cost.