矽晶圓離子佈植與微奈米結構之非破壞性光學檢測及其在光流體元件之應用

Nondestructive optical inspection for ion implantation processes and micro-/ nano-structures on Si wafers and its application on optofluidic device

蘇寶勻(Pao-Yun Su)

指導教授 ： 陳學禮

國立臺灣大學/工學院/材料科學與工程學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.10606

摘要

離子佈植在半導體製程中是最難監控的步驟之一，由於其造成的影響主要在材料或晶圓內部，一般只能使用破壞性的分析工具來觀察。因此，本論文提出一套有系統的非破壞性光學量測方法，以檢測離子佈植樣品。論文中分別使用了三種光學分析工具：紫外光-可見光光譜儀、橢圓偏光儀、拉曼光譜儀來進行佈植縱深及濃度分析。從光譜儀能夠簡易的辨別出離子佈植不同濃度樣品，而從橢圓偏光儀測得之橢圓參數能夠計算得到各個樣品的等效破壞層厚度與光學常數變化，拉曼光譜儀則擁有較多能夠調控的變因，包括激發雷射光波長、聚焦物鏡倍率、雷射光功率等。可根據紫外光-可見光光譜與橢圓參數的結果依照不同破壞程度的樣品選擇不同架構，利用衰減雷射光入射功率來改變其量測到的有效拉曼訊號深度，以此得到樣品內部的精確縱深分析。例如可以看到由表面至樣品深處分別為非晶矽、晶粒大小漸增的多晶矽至未遭受破壞的單晶矽。較低摻雜濃度(<1015 cm-2)的樣品可以利用穿透深度較淺的短波長及低功率的光來進行量測，而破壞較嚴重的高濃度(>1015 cm-2)高加速電壓樣品則可使用長波長的光作為激發雷射。本論文檢測佈植深度只有數奈米的10 keV、1 x 1013 cm-2佈植氟化硼樣品至深度數百奈米的200 keV、5 x 1015 cm-2離子佈植砷的各種樣品。實驗結果呈現從破壞程度極為嚴重的高加速電壓重摻雜到接面極淺的輕摻雜離子佈植的各種樣品皆能鑑別出來。預期可從量測已知參數的樣品建立出一資料庫，由此便能推測未知樣品的離子佈植參數，與目前現在所使用的破壞性檢測(如穿透式電子顯微鏡)相比，不但不需複雜的試片準備，且能大面積的快速非破壞性量測，為一非常方便且低成本的非破壞性檢測方法。
除此之外，本論文也利用此上述方法來檢測不同形貌的各種單晶矽微奈米結構，以鑑定溼式化學蝕刻對於單晶矽晶格的影響。再者，此單晶矽的微奈米複合結構能夠發展成一能即時偵測極低濃度的無管壁隱形微流道暨表面增強拉曼散射基板。由於此微奈米複合結構具有超疏水性質，其接觸角可以到達171.9度，可藉由表面親疏水性大幅差異來製作出完全無管壁微流道。藉由在超疏水表面選擇性鍍上親水性的金屬薄膜可完成無管壁微流道製程。可以簡單利用鍍膜遮罩或單次微影來控制流道的大小及形狀。在本論文中其微流道線寬能做到200微米，而同時流道處因金屬薄膜的電場增強現象使之兼具表面增強拉曼散射效果，此基板最低能量測到濃度為1 x 10-8 M的R6G溶液，且當待測液體在流動時能夠同時量測其拉曼訊號，即時偵測流動中液體是否含有待測分子，並能夠區分不同濃度的待測物。由於此微流道並無管壁，不需加裝幫浦來施加壓力推動液體，只需重力便能使液體流動，同時也無一般微流道所需的上蓋及管壁，在量測拉曼訊號時便不會受到上蓋及管壁結構的覆蓋及背景訊號影響，也較無景深對焦問題，也因流道各處皆具有表面增強拉曼散射效果，在任意位置皆能輕易進行拉曼光譜量測。

關鍵字

拉曼光譜縱深分析 ； 離子佈植 ； 超輕摻雜極淺接面檢測 ； 非破壞性光學檢測 ； 光流體元件 ； 微奈米複合結構 ； 親疏水性質 ； 表面增強拉曼散射

並列摘要

Ion implantation is one of the processes that are very difficult to be monitored in semiconductor manufacturing technology. Because it exclusively makes influence beneath the material surface, conventional monitoring generally relies on contact or destructive analysis tools. In this thesis, a nondestructive, optical method is systematically proposed for rapid inspection of implanted samples. Three analysis tools are used in this thesis—UV-visible spectrophotometer, ellipsometer and micro-Raman spectrometer. First, different implanted samples are simply discriminated by using a UV-visible spectrophotometer to record the reflectance and transmittance spectra. Second, spectroscopic ellipsometry gives information about the effective thickness and optical constants of the effective disordered layer in the implanted samples. Third, the setup of Raman spectroscopy is easily tunable, including the excitation wavelength, the magnification of object lens, and the incident power of laser. Depending on implantation parameters estimated by the spectral results from UV-visible spectrophotometer and ellipsometer, excitation wavelength and object lens are deliberately chosen. In addition, the probe depth of Raman spectroscopy will be close to the surface by attenuating the incident power of laser, which is helpful to perceive the depth profile of crystallinity. For example, the structures from surface to the interior can be observed in the order of amorphous silicon, increasing grain polysilicon, and the undamaged single crystal silicon. Short wavelength and low power laser are used to record the Raman spectra of a low-dose (<1015 cm-2) implanted sample, while the sample with both high-dose (>1015 cm-2) implantation and ion energy should apply longer excitation wavelength. In the thesis, the samples from 10 keV and 1 x 1013 cm-2 BF2+ implantation (only several nanometers implanted depth) to 200 keV and 5 x 1015 cm-2 As+ implantation (hundreds nanometers implanted depth) were inspected. The results represent that from the serious damage with high energy heavily doping to ultra-shallow junction with lightly doped implantation could be completely distinguished. A database can be established according to the spectral results recorded by the optical inspection tools. Comparing with the conventional destructive analysis tools, such as transmittance electron microscopy (TEM), complex preparation of specimen is unnecessary. And therefore these rapid, convenient, and nondestructive inspection methods can be used to find out the processing parameters of unknown implanted samples.
Besides, these methods are further applied on monitoring the micro- and nano-structures of Si formed by wet etching. And an invisible microchannel-based surface-enhanced Raman scattering (SERS) substrate is developed on the micro/nano hybrid structures, which provide an ultrasensitive, in situ detection of analytes. Due to the superhydrophobic property of the micro/nano hybrid structure, the contact angle can be 171.9 degrees. The invisible microchannel without solid pipe walls relies on the significant difference in the hydrophobicity of the hybrid structure. The hydrophilic channel on the superhydrophobic hybrid substrate is formed by depositing a metal thin film, which acts as the SERS hot spots simultaneously. In this thesis, the width of the microchannel could be achieved in 200 micrometer scale and the detection limit is R6G solution which in 1 x 10-8 M. The analyte solution can flow on the channel, and be taken Raman signals at the same time. Moreover, solutions in different concentration can also be discriminated. Because there are no pipe walls in this microchannel, the liquid can flow just depends on gravity so that exerting pressure to push forward by installing pump is not necessary. In addition, there is also no cover above the microchannel which have disadvantages of Raman background signals and difficulties in focusing. Moreover the SERS effect is provided everywhere in the microchannel, Raman spectra can be recorded easily at any positions.

並列關鍵字

Raman depth profile analysis ； ion implantation ； lightly doped ultra-shallow junction inspection ； nondestructive optical inspection ； optofluidic device ； micro/ nano hybrid structure ； hydrophobic property ； surface-enhanced Raman scattering (SERS)

參考文獻

1. H. Xiao, Introduction of Semiconductor Manufacturing Technology. Prentice Hall, 2000, p.282.

2. K. Suzuki, Analysis of ion implantation profiles for accurate process/device simulation: ion implantation profile database based on tail function. Sci. Tech. J, 2010, 46, 307-317.

3. V. Lavrentiev, J. Vacik, V. Vorlicek, V. Vosecek, Raman scattering in silicon disordered by gold ion implantation. Phys. Status Solodi B, 2010, 247, 2022-2026.

4. K. Fujiwara, M. Ohtani, K. Kanayama, H. Ogata, Quantitative Auger analysis of ion-implanted Boron and Arsenic in polycrystalline silicon. Surf. Sci. , 1976, 61, 436-442.

5. K. S. Jones, S. Prussin, E. R. Weber, A systematic analysis of defects in ion-implanted silicon. Appl. Phys. A, 1988, 45, 1-34.