Internal photoemission spectroscopy determination of barrier heights between Ta-based amorphous metals and atomic layer deposited insulators

Abstract

The energy barrier heights between two recently reported Ta-based amorphous metals (TaWSi and TaNiSi), TaN, and atomic layer deposited Al2O3 and HfO2 insulators are measured in metal/insulator/metal (MIM) structures with Au top electrodes using internal photoemission (IPE) spectroscopy. For Al2O3, the Ta-based metal barrier heights, φBn, increase with increasing metal work function, ΦM, for TaN, TaWSi, and TaNiSi, respectively. For HfO2, however, the barrier heights are relatively constant for all three metals φBn,TaNiSi ≈ φBn.TaWSi ≈ φBn.TaN. The difference between HfO2 and Al2O3 is attributed to enhanced Fermi-level pinning due to a larger dielectric constant. The slope parameter, S, was found to be roughly 0.89 and 0.44–0.69 for Al2O3 and HfO2, respectively. For devices with a TaWSi bottom electrode, a comparison was also made between Al and Au top electrodes. Significantly, smaller barrier heights were obtained with an Au top electrode than with an Al top electrode, 0.6 eV and 0.8 eV lower for HfO2 and Al2O3, respectively. IPE energy barriers are found consistent with current-voltage asymmetry of MIM diodes, whereas Schottky model predictions of barrier heights were inconsistent.

Determination of Hafnium Zirconium Oxide Interfacial Band Alignments Using Internal Photoemission Spectroscopy and X-ray Photoelectron Spectroscopy

“…φ Bn values are summarized in Table and range from 2 eV for the Ti/Pt top electrode to 3.4 eV for Pt and Au. For Au, Al, and TaN electrodes with HZO, we find electron energy barriers of 3.3, 2.6, and 2.7 eV, respectively, each approximately 0.2 eV lower than for the respective HfO 2 barriers measured by Jenkins et al, , pointing toward a slightly lower band gap for HZO as compared to HfO 2 . Considering the TaN electrodes for the nominally symmetric device, IPE measurements yield φ Bn = 2.7 and 2.5 eV for the bottom and top electrodes, respectively.…”

“…Finally, to correct for any image force or field-induced barrier lowering that may be present, zero-field barrier heights, φ Bn , were extrapolated from Schottky plots of the φ thresh values vs E 1/2 . IPE φ Bn values have an estimated accuracy of ±0.1 eV. ,, …”

“…The current was normalized by subtracting the dark current for each applied bias and then converted to quantum yield, Y . Using an algorithm, the spectral threshold, φ thresh (the photon energy at the onset of IPE), at each insulator electric field, E , was extrapolated from the most linear region of Y 1/2 vs h υ curves to the baseline x -intercept . The square root of Y was used to account for the energy distribution of carriers in the metal electrode.…”

“…Using an algorithm, the spectral threshold, φ thresh (the photon energy at the onset of IPE), at each insulator electric field, E, was extrapolated from the most linear region of Y 1/2 vs hυ curves to the baseline x-intercept. 25 The square root of Y was used to account for the energy distribution of carriers in the metal electrode. E was corrected for the built-in field, E bi , that arises due to the effective work function difference of the metal electrodes and was taken as the field at which electron emission switches from the top to the bottom electrode.…”

A reliable and accurate model of photoelectron yield spectrum and its applications

“…由于结是通过 金属、半导体、绝缘体中任意两个组成的界面系统, 对此界面而言光电产额谱起到重要的作用. 用光电 产额谱能够对结的电结构进行分析 [9] , 用光电产额 谱可以测量结中的狄拉克缺陷及费米能级 [10] , 应 用光电产额谱能研究结中的能带偏移 [11] , 从光电 产额谱能够获取结中的缺陷密度在能级的分布 [12] , 光电产额谱还能用来探索价带剖面 [13] . 不同物质 形成的结有界面, 由于界面的存在就有了势垒, 因 为光电产额在势垒高度处是零, 故确定这个势垒高 度的常用方法是光电产额谱 [14,15] .…”