利用磁場式感應耦合電漿質譜儀精確測量碳酸鹽類之鎂鈣及鍶鈣比值

Abstract

我們已成功建立一套準確度達千分比等級之碳酸鹽類Mg/Ca及Sr/Ca的磁場式感應耦合電漿質譜儀分析技術。利用800瓦無線電波功率的冷電漿技術減低頻譜干擾，並搭配高敏感度和高穩定度的溼式標本導入系統。鎂鈣和鍶鈣比值（24Mg/43Ca和86Sr/43Ca）是藉由脈衝式計數器量測之24Mg, 43Ca 及86Sr離子訊號強度比值計算而得，並使用標準樣品穿插校正法校正儀器的質量分化及低頻飄移的現象。而除鎂鈣及鍶鈣比值之外，標本中的鋁鈣（Al/Ca）、錳鈣（Mn/Ca）、鐵鈣（Fe/Ca）、鎘鈣（Cd/Ca）和鋇鈣（Ba/Ca）比值也可以同時測量。在本方法中發現到有三種基質效應，包含（1）硝酸濃度影響各種元素的訊號強度及比值與測量，當硝酸由1-6%增加時，元素比值變化趨緩，最後選擇使用5%之硝酸做為實驗時固定濃度;（2）鎂鈣比值測定受到鈣濃度影響，當鈣濃度介於0.5-2.5 ppm時，鎂鈣比值會造成2.5%的變化，當鈣濃度高於3 ppm時，此基質效應僅有輕微的影響；（3）不同的鍶鈣比值亦會對鎂鈣比值造成0.18% mmol/mol的偏差，主要原因是受到Sr++ 二價離子的形成干擾到43Ca+ 離子強度進而影響到鎂鈣比值，所幸，此影響只會對有孔蟲和珊瑚的鎂鈣比值造成約小於0.1%的偏差。此方法的測量準確度可達0.1-0.2%，標準溶液和標本的重覆測試結果顯示鎂鈣比值的分析再現性為0.45%，鍶鈣比值則為0.35%（2RSD)；此結果相對於有孔蟲鎂溫度計的誤差為0.05 ℃，珊瑚鍶溫度計的誤差為0.5 ℃。本方法的標本需求量僅為2 µg，且分析產能為每小時10個標本，加上簡便的事前準備工作，本技術優於其他元素比值分析分法。

A permil-level analytical technique for determination of carbonate Mg/Ca and Sr/Ca ratios using inductively coupled plasma sector-field mass spectrometry (ICP-SF-MS, Thermo Electron ELEMENT2) has been established. Cold plasma technique with a RF power of 800-W were employed to diminish spectral interference. A wet introduction system with Scott-type double-pass spray chamber and an ESI micro-nebulizer with a sample uptake rate of 50 µl/min were used, which can offer low blank and high stability. Intensities of ion beams, 24Mg, 43Ca and 86Sr, were detected in pulse-counting mode. Mg/Ca and Sr/Ca ratios were determined directly from intensity ratios using external, matrix-matched standard to correct for mass discrimination and low-frequency ratio drift. Al/Ca, Mn/Ca, Fe/Ca, Cd/Ca, Ba/Ca and Th/Ca were also determined. Three different matrix effects are observed. (1) Acid effect on the elemental intensities and ratios decreases with increasing HNO3 from 1-6%. 5% HNO3 was suggested be used for routine measurement. (2) For the effect of Ca level, Mg/Ca ratio varies by 2.5% with [Ca] from 0.5-2.5 ppm and changes insignificantly for [Ca] higher than 3 ppm. (3) Different Sr/Ca slightly affects Mg/Ca determination with a slope of 0.18% per mmol/mol. It is caused by the formation of 86Sr double charged ions, which biases the measurement of 43Ca+ and affects the accuracy of Mg/Ca analysis. This effect causes only 0.1% or less for Mg/Ca ratio in both foraminiferal and coral samples. Replicate measurements made on standards and samples show that our method can give an internal precision of 0.1-0.2% (2RSD). Reproducibility of 0.45% for Mg/Ca and 0.35% for Sr/Ca can be attained. The corresponding errors are 0.05 ℃ for foraminiferal Mg thermometer and 0.5 ℃ for coral Sr thermometer. Advantages of this new technique include a small sample size of only 2 µg and high throughput of 10 samples per hour.