热激电流谱测试技术(TSC)是宽禁带半导体深能级缺陷非常有效的测试方法, 能够精准获得缺陷类型, 深度(Ｅａ, ｉ), 浓度(Ｎｉ)以及俘获截面(σｉ)等重要物理信息。 研究了基于变升温速率的Arrhenius公式作图法和同步多峰分析法(SIMPA)对热激电流谱数据处理的差异及影响规律。 结果表明, Arrhenius公式作图法在陷阱能级深度确认方面较为准确, 但需要通过多次变升温速率提高其准确性, 实验操作周期较长, 并且无法分解热激电流谱峰重叠的情况。 相比之下, 同步多峰分析法能够通过单次温度扫描的数据处理得到Ｅａ, ｉ, Ｎｉ及σｉ等陷阱参数, 所需实验周期较短。 但β, Ｅａ, ｉ, σｉ和载流子迁移率寿命积(μt)等参数的选择对谱峰的位置, 幅值及峰宽影响较大。 初值的设定对拟合结果和数据吻合程度影响显著。 此外, 红外透过成像结果表明, 头部样品Te夹杂相的浓度较低且呈现明显的带状分布, 而尾部样品夹杂相浓度呈均匀分布。 通过对比不同样品的热激电流谱测试结果发现, 尾部样品浅能级陷阱浓度远高于头部样品, 且其低温光电导弛豫过程呈现明显的曲线变化规律。 这一研究结果表明, Te夹杂相的分布及浓度可能会导致晶体内部浅能级缺陷的浓度变化, 并且浅能级缺陷对光激发载流子的俘获时间更长, 去俘获时间更短。

Thermally stimulated current (TSC) spectroscopy is a quite effective method for the defects studies in wide bandgap semiconductors, from which the physical information, i. e. defect types, activation energy (Ｅａ, ｉ), concentration (Ｎｉ) and capture cross-section (σｉ), can be given. The discrepancy and effects of heating-rate-dependent Arrhenius method and simultaneous multiple peak analysis (SIMPA) on the data processing results of TSC were studied in this work. The results indicated that Arrhenius method were more accurate in terms of the thermal activation energy of different traps. However, more heating rate, which meaned longer test cycles, were needed to maintain the accuracy. And also, this method could not deal with conditions of traps over-lap. In contrast, the SIMPA method could obtain the trap signatures (Ｅａ, ｉ, Ｎｉ, σｉ) with only one heating rate. However, parameters, i. e. β, Ｅａ, ｉ, σｉ and carrier mobility and lifetime product (μt), had significant effects on the peak position, height and width, which directly influenced the results of the fitting curve. Furthermore, the IRTM images of sample from the head of the ingots showed lower concentration of Te inclusions with belt-like distribution compared to that from the tail. Through the investigation of TSC spectroscopy, the concentration of shallow levels were much higher in the sample from the tail of the ingots than that from the head. The low temperature persistent photoconductivity (PPC) experiments showed a curvilineal variation for the tail sample. The results showed that the concentration and distribution of Te inclusions could probably result in the concentration variation of shallow trap centers, which present longer trapping time and shorter de-trapping time of optical excited carriers in crystal.