石墨是理想的无机非金属材料, 具有高化学稳定性、 良好的导电性、 较好的耐磨性等优点, 被广泛应用于现代化学工业及其他诸多领域。 由于石墨是难熔物, 其中的微量杂质元素使用普通化学法或常规仪器分析法均难以准确检出。 常用的火法-电感耦合等离子质谱(ICP)方法检测石墨存在的问题是: （1）在高温灼烧期间个别元素容易损失; （2）在加酸溶解灰化组分过程中部分杂质氧化物仍无法溶解完全。 因此, 很多学者开始研究利用固体进样法来测定石墨中的杂质元素含量。 辉光放电质谱法（GDMS）是将辉光放电源（GD）与质谱分析方法（MS）联用的一种技术, 采用固体进样方式, 具有样品前处理简单、 基体效应小、 检出限低、 灵敏度高等优点, 在国内外已成为部分高纯金属和半导体材料分析领域的标准方法。 灵敏度因子值(RSF)是一个用于校正GDMS分析结果的系数, 对于GDMS分析而言, 大部分元素在不同基体中仍然存在较明显的基体效应。 要将GDMS分析作为一种定量分析方法, 需要采用与基体匹配的标准物质来校正RSF, 但目前大多数GDMS分析均采用仪器厂家提供的标准相对灵敏度因子（RSFStd）进行测定, 只能获得半定量分析结果。 研究了采用GDMS直接测定石墨样品中9种杂质元素含量的方法, 通过对放电条件等参数的优化选择, 确定了石墨分析的最佳放电条件（电流强度为55 mA, 放电气体流量为450 mL·min-1）。 在此条件下采用半定量法（RSFStd）测定了石墨参考样品中Mg, Cr, Ni, Ti, Fe, Cu, Al, Si, Ca共九种杂质元素含量, t检验结果表明, 多数元素测量结果与参考值存在显著差异。 要获得更为准确的结果, 需要获得各元素相应的RSFx以建立定量分析方法。 通过实验, 考察了不同的电流强度和放电气体流量对九种元素RSF值的影响, 讨论了影响因素产生的原因。 实验结果表明, 电流强度和放电气体流量对大部分元素的RSF值都有较大的影响, 其中放电气体流量对RSF值影响最大, 各元素的RSF值变化幅度在15%和405%之间。 在选定条件下采用RSFx值定量分析了石墨材料中Mg, Cr和Ni等九种杂质元素含量, 检测结果的t检验sig值均大于0.05, 表明测定结果与参考值无显著性差异, 方法的准确度有了显著提高; 测定结果的精密度（RSD）介于3.2%～9.9%之间, 方法可满足4N纯度以上高纯石墨材料的分析。

Graphite material is an ideal inorganic non-metallic material with high chemical stability, good conductivity, and good wear resistance. As graphite is a refractory substance, it is difficult to test the trace element contents by using common chemical methods or conventional instrumental analyses. Common problems in the graphite analysis by the Fire-ICP method are as follows: (1) the individual elements are easily lost during the high temperature burning pretreatment process, and (2) the graphite cannot be dissolved completely during the process of adding acid dissolution. Therefore, many scholars began using solid analytical technology to determine the impurity contents in graphite. Glow discharge mass spectrometry (GDMS) is a technology combining glow discharge power supply (GD) with mass spectrometry (MS). It has advantages of simple pretreatment, weak matrix effect, low detection limit and high sensitivity. It has become a division of high pure metal and semiconductor materials at home and abroad. Relative sensitivity factor (RSF) is a coefficient used to correct GDMS analysis results. For GDMS analysis, most elements still have obvious matrix effect in different matrixes. In order to make GDMS analysis as a quantitative analysis method, it is necessary to use the standard material matching the matrix to correct the RSF. However, most of the GDMS analysis is based on the standard relative sensitivity factor (RSFStd) provided by instrument manufacturers and only semi quantitative analysis results can be obtained. This paper describes an analytical method to determine the content of 9 elements in graphite materials using GDMS. Through the optimization of the discharge conditions, the suitable discharge conditions of graphite were determined (current intensity as 55 mA, discharge gas flow rate as 450 mL·min-1). Under optimized analytical conditions, 9 impurities (Mg, Cr, Ni, Ti, Fe, Cu, Al, Si and Ca) were determined. The result of t-test showed that there was significant difference between the results of most elements and the reference value. In order to obtain more accurate results, the corresponding RSFx of each element was required to establish quantitative analysis methods. Through experiments, the effects of different current intensity and discharge gas flow on the RSF value of 9 elements were investigated, and the causes of the influencing factors were discussed. The experimental results showed that the current intensity and discharge gas flow have a great influence on the RSF value of most elements. The discharge gas flow has the greatest influence on the RSF value, and the RSF value of each element varies between 15% and 405%. Under selected conditions, the content of 9 impurity elements, such as Mg, Cr and Ni in graphite materials, was quantitatively analyzed by RSFx value. The t-test sig value of the test results was more than 0.05, indicating that there was no significant difference between the measured results and the reference value, and the accuracy of the method was significantly improved. The relative standard deviations (RSD) were between 3.2% and 9.9%. The method can meet the need for the analysis of high purity graphite materials above 4N purity.