留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation

Jian-hua SHI Bing-chen HAN

史健华, 韩丙辰. 光程补偿近红外光透射反射干涉重构微结构内部形貌[J]. 中国光学, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
引用本文: 史健华, 韩丙辰. 光程补偿近红外光透射反射干涉重构微结构内部形貌[J]. 中国光学, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
SHI Jian-hua, HAN Bing-chen. Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation[J]. Chinese Optics, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
Citation: SHI Jian-hua, HAN Bing-chen. Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation[J]. Chinese Optics, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395

光程补偿近红外光透射反射干涉重构微结构内部形貌

doi: 10.3788/CO.20191202.0395
基金项目: 

国家自然科学基金青年基金 11705107

山西省科技攻关项目 2015031002-1

山西大同大学博士学位研究基金 2014B15

详细信息
    作者简介:

    史建华(1978—),男,副教授,山西宁武人,硕士研究生学历,硕士学位,现任教学评估与督导中心副主任、物电学院教师,主要从事光学干涉及图像处理方面的研究。E-mail:

  • 中图分类号: O435.1

Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation

Funds: 

the National Natural Science Foundation Youth Foundation of China(NSFC) 11705107

the Science and Technology Infrastructure Program of the Ministry of Science and Technology of Shanxi Province, China 2015031002-1

the Shanxi Datong University Research Foundation for Ph. D. 2014B15

More Information
    Author Bio:

    SHI Jian-hua(1978—), male, Associate Professor, from Ningwu County, Shanxi Province, Master′s Research Degree. Present:Deputy Director of the Teaching Evaluation and Supervision Center, and teacher at the Institute of Physics and Energy, mainly engaged in the research of optical processing involving image processing. E-mail:shijianhua_dtdx@163.com

    Corresponding author: SHI Jian-hua, E-mail:shijianhua_dtdx@163.com
  • 摘要: 高深宽比微结构的底部及侧壁形貌重构是微机电系统领域亟待解决的一个问题。本文提出光程补偿近红外光透射反射干涉技术重构微结构内部形貌的方法,所采用的近红外光干涉技术将白光干涉系统中的光源扩展至近红外光源,将反射干涉技术扩展至透射反射干涉技术,近红外光干涉测量系统由近红外光光源、干涉显微镜、红外光CCD、高精度压电陶瓷和数据采集系统组成。设计了具有两个台阶的GaAs半导体微结构待测样品,采用近红外光垂直扫描干涉法并通过光程补偿,重构了微结构的内部三维形貌,并与扫描电镜结果进行对比。光程补偿近红外光透射反射干涉技术测量的台阶相对高度分别为2.132 μm和0.766 μm,与扫描电镜和近红外光反射干涉测量结果基本一致,分别对应2.16%和2.68%的相对误差。测量结果表明,该测量系统能够测量高深宽比微结构底部及侧壁形貌。
  • 图  1  近红外光透射反射干涉仪测量高深宽比微结构底部及侧壁形貌

    Figure  1.  Bottom and side wall profile reconstruction of the microstructure with high aspect ratio measured by the near-Infrared light transmission reflection interferometer

    图  2  近红外光干涉系统示意图

    Figure  2.  Schematic of near-infrared light interferometer

    图  3  测量样品结构图:俯视图和剖面图

    Figure  3.  Structure of the test object: the top view and the cross-section

    图  4  近红外光反射干涉测量结果

    Figure  4.  Measurement results with near-infrared light reflection interference

    图  5  光程补偿近红外透射反射干涉测量结果

    Figure  5.  Measurement results using near-infrared light transmission reflection interference with optical path compensation

    表  1  Comparison of relative step heights for different measurement methods

    Table  1.   Comparison of relative step heights for different measurement methods

    Measurement method Step A/μm Error/% Step B/μm Error/%
    SEM 2.087 0.746
    Near-infrared light reflection interference 2.107 0.96 0.759 1.74
    Near-infrared transmission interference 2.132 2.16 0.766 2.68
    下载: 导出CSV
  • [1] TOTSU K, FUJISHIRO K, TANAKA S, et al.. Fabrication of three-dimensional microstructure using maskless gray-scale lithography[J]. Sensors and Actuators A:Physical, 2006, 130-131:387-392. doi:  10.1016/j.sna.2005.12.008
    [2] ZHANG T, GAO F, JIANG X Q. Surface topography acquisition method for double-sided near-right-angle structured surfaces based on dual-probe wavelength scanning interferometry[J]. Optics Express, 2017, 25(20):24148-24156. doi:  10.1364/OE.25.024148
    [3] HOU P X, LIU C, CHENG H M. Field Emission from Carbon Nanotubes[M]. Nanomaterials Handbook. 2nd Ed. CRC Press, 2017: 255-272.
    [4] MORITA S, GIESSIBL F J, MEYER E, et al.. Noncontact Atomic Force Microscopy[M]. Berlin:Springer, 2015.
    [5] 董恺琛, 娄帅, 姚杰, 等.脉冲激光沉积薄膜的残余应力测量[J].光学 精密工程, 2018, 26(1):70-76. http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201801011.htm

    DONG K CH, LOU SH, YAO J, et al.. Measurement of residual stresses in pulsed laser deposited thin films[J]. Opt. Precision Eng., 2018, 26(1):70-76.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201801011.htm
    [6] TSUDA Y, KITASAKO Y, SADR A, et al.. Effects of brushing timing after erosive challenge on enamel loss in situ: White light interferometer and nanoindentation study[J]. Dental Materials Journal, 2016, 35(4) 613-620. doi:  10.4012/dmj.2015-405
    [7] DAO T, THOMAS T, MARX D, et al.. Evaluation of non-destructive etch depth measurement for through silicon vias[C]. 2012 IEEE International Conference on IC Design & Technology, IEEE, 2012: 1-4.
    [8] ZHOU Y F, CAI H ZH, ZHONG L Y, et al.. Eliminating the influence of source spectrum of white light scanning interferometry through time-delay estimation algorithm[J]. Optics Communications, 2017, 391:1-8. doi:  10.1016/j.optcom.2016.12.006
    [9] XUE CH Y, LIU J, CHOU X J, et al.. White-light transmission reflection interference technology application in three-dimensional reconstruction method validation for microstructures[C]. Proceedings of the 3rd International Congress on Image and Signal Processing, IEEE, 2010, 2: 867-870.
    [10] 刘燕德, 万常斓.芝麻油掺伪的近红外透射光谱检测技术[J].农业机械学报, 2012, 43(7):136-140. doi:  10.6041/j.issn.1000-1298.2012.07.025

    LIU Y D, WAN CH L. Analysis of sesame oil adulteration using near infrared transmission spectroscopy[J]. Transactions of the Chinese Society for Agricultural Machinery, 2012, 43(7):136-140.(in Chinese) doi:  10.6041/j.issn.1000-1298.2012.07.025
    [11] 秦五昌, 汤修映, 彭彦昆, 等.基于可见/近红外透射光谱的孵化早期受精鸡蛋的判别[J].光谱学与光谱分析, 2017, 37(1):200-204. http://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201701044.htm

    QIN W CH, TANG X Y, PENG Y K, et al.. Identification of fertilized chicken eggs based on visible/near-infrared spectrum during early stage of incubation[J]. Spectroscopy and Spectral Analysis, 2017, 37(1):200-204.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201701044.htm
    [12] THOMAS D J. 3D white light interferometry assessment of robotic laser scalpel assisted surgery to minimise scar tissue formation[J]. International Journal of Surgery, 2017, 38:117-118. doi:  10.1016/j.ijsu.2016.12.037
    [13] XIAO Y, QIU L R, ZHAO W Q. Laser confocal cylindrical radius measurement method and its system[J]. Applied Optics, 2017, 56(23):6596-6602. doi:  10.1364/AO.56.006596
    [14] TAPILOUW A M, CHANG Y W, YU L Y, et al.. Reduction of batwing effect in white light interferometry for measurement of patterned sapphire substrates(PSS) wafer[J]. Proceedings of SPIE, 2016, 9960:996006. doi:  10.1117/12.2236874
    [15] TERESCHENKO S, LEHMANN P, GOLLOR P, et al.. Vibration compensated high-resolution scanning white-light Linnik-interferometer[J]. Proceedings of SPIE, 2017, 10329:1032940. doi:  10.1117/12.2270226
  • [1] WANG Hao-bing, TAO Jin, LV Jin-guang, MENG De-jia, LI Yang, ZHAO Yong-zhou, WANG Jia-xian, ZHANG Jun, QIN Yu-xin, WANG Wei-biao, LIANG Jing-qiu.  Research on the Enhancement of Absorption Properties of Silicon via Localized Surface Plasmons Resonance in Blue Light . 中国光学, 2020, 13(): 1-23. doi: 10.37188/CO.2020-0056
    [2] HAN Jing, GAO Yang, JIAO Wei-yan, FAN Guang-hua, GAO Ya-chen.  Mid-infrared plasmon regulation based on graphene nanoribbons . 中国光学, 2020, 13(3): 627-636. doi: 10.3788/CO.2019-0185
    [3] FAN Li-na, MA Jun-shan.  Design of resonant waveguide grating filter with reflection and transmission modes . 中国光学, 2020, 13(5): 1-11. doi: 10.37188/CO.2020-0072
    [4] Ye YU, Wen-yan LIN, Xue-kang PENG, Yu JIN, Zhi-jun WU, Yan CHEN.  Weak microcavity effect in tandem white organic light-emitting diodes employing an effective charge-generation layer of Liq/Al/HAT-CN . 中国光学, 2019, 12(2): 382-394. doi: 10.3788/CO.20191202.0382
    [5] Hong-yan LIU, Qiang SUN.  Finite element analysis of infrared thermal imaging for four-layers structure of human thigh . 中国光学, 2018, 11(2): 237-247. doi: 10.3788/CO.20181102.0237
  • 加载中
图(5) / 表 (1)
计量
  • 文章访问数:  426
  • HTML全文浏览量:  93
  • PDF下载量:  151
  • 被引次数: 0
出版历程
  • 收稿日期:  2018-06-05
  • 修回日期:  2018-07-13
  • 刊出日期:  2019-04-01

Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation

doi: 10.3788/CO.20191202.0395
    基金项目:

    国家自然科学基金青年基金 11705107

    山西省科技攻关项目 2015031002-1

    山西大同大学博士学位研究基金 2014B15

    作者简介:

    史建华(1978—),男,副教授,山西宁武人,硕士研究生学历,硕士学位,现任教学评估与督导中心副主任、物电学院教师,主要从事光学干涉及图像处理方面的研究。E-mail:

    通讯作者: SHI Jian-hua, E-mail:shijianhua_dtdx@163.com
  • 中图分类号: O435.1

摘要: 高深宽比微结构的底部及侧壁形貌重构是微机电系统领域亟待解决的一个问题。本文提出光程补偿近红外光透射反射干涉技术重构微结构内部形貌的方法,所采用的近红外光干涉技术将白光干涉系统中的光源扩展至近红外光源,将反射干涉技术扩展至透射反射干涉技术,近红外光干涉测量系统由近红外光光源、干涉显微镜、红外光CCD、高精度压电陶瓷和数据采集系统组成。设计了具有两个台阶的GaAs半导体微结构待测样品,采用近红外光垂直扫描干涉法并通过光程补偿,重构了微结构的内部三维形貌,并与扫描电镜结果进行对比。光程补偿近红外光透射反射干涉技术测量的台阶相对高度分别为2.132 μm和0.766 μm,与扫描电镜和近红外光反射干涉测量结果基本一致,分别对应2.16%和2.68%的相对误差。测量结果表明,该测量系统能够测量高深宽比微结构底部及侧壁形貌。

English Abstract

史健华, 韩丙辰. 光程补偿近红外光透射反射干涉重构微结构内部形貌[J]. 中国光学, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
引用本文: 史健华, 韩丙辰. 光程补偿近红外光透射反射干涉重构微结构内部形貌[J]. 中国光学, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
SHI Jian-hua, HAN Bing-chen. Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation[J]. Chinese Optics, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
Citation: SHI Jian-hua, HAN Bing-chen. Internal profile reconstruction of microstructures based on near-infrared light transmission reflection interferometry with optical path compensation[J]. Chinese Optics, 2019, 12(2): 395-404. doi: 10.3788/CO.20191202.0395
    • In recent years, high aspect ratio microstructures have been widely used in the field of MEMS due to their high sensitivity and large displacement[1]. While due to their small widths(about 1-10 μm), high depth(about 10-500 μm) and high aspect ratio(10:1 to 100:1), real-time online reconstruction of the bottom and sidewall features has become an urgent problem that needs to be solved[2]. At present, there are three main instruments for measuring the 3D shape of a microstructure: the scanning electron microscope[3](SEM:Scanning Electron Microscope), the atomic force microscope[4-5](AFM:Atomic Force Microscope) and the white light interferometer[6]. The SEM is a device that scans the sample with a very narrow electron beam. When measuring high aspect ratio structures, it is necessary to cut the microstructure sample to scan the profile. It is a destructive method of measurement with a high time requirement and a high cost. AFMs directly calculate the height of the sample surface using the force of a microprobe. Due to the particularity of high aspect ratio structures, it is necessary to change the tip structure to measure the inside of a structure during measurement. This is very difficult. White light interferometers are more suitable for measuring the surface topography of microstructures. There are many commercial devices, such as Polytec, Talysurf, Veeco, Zygo, etc., but for high aspect ratio microstructures, white light diffraction is more severe. Furthermore, since the light hits the object to be tested at an angle during measurement, it cannot pass through a narrow slit completely. It is therefore difficult to measure a structure′s bottom and sidewall topography[7-8]. In summary, based on the white light interference principle, semiconductor material(gallium arsenide GaAs or silicon Si) can be transmitted in the infrared light band. If infrared light is used instead of the white light in white light interference, an infrared light interference system could be constructed and a certain optical path compensation could be implemented. Optical path compensation using infrared light transmission and reflection interference technology could achieve bottom and sidewall shape reconstruction for high aspect ratio microstructure.

      In this paper, the near-infrared optical band is used as an example and a near-infrared light transmission reflection interference technology is proposed. The corresponding interferometric system consists of a near-infrared light source, an interference microscope, an infrared CCD, high-precision piezoelectric ceramic and a data acquisition system. In the beginning, the near-infrared optical interference topography reconstruction technology based on vertical scanning interferometry and the basic principle of GaAs material transmission in the infrared light band are studied. Secondly, a typical double-step GaAs material microstructure sample is designed and its internal shape of the measurements are performed. Comparing with the results obtained from SEM and the near-infrared light reflection interferometer, it shows that the optical path compensation near-infrared light transmission reflection interference system is effective. It was also found that the measured morphology is consistent with that measured by SEM method and near-infrared light reflection interference method. The results indicate that the internal shape of a high aspect ratio microstructure can be measured by proposed method.

    • Fig. 1 shows the near-infrared light transmission reflection interference measurement of bottom and sidewall topography of microstructure with high aspect ratio. It can be seen that the bottom and sidewall topography can be reconstructed by an optical path compensation method.

      图  1  近红外光透射反射干涉仪测量高深宽比微结构底部及侧壁形貌

      Figure 1.  Bottom and side wall profile reconstruction of the microstructure with high aspect ratio measured by the near-Infrared light transmission reflection interferometer

    • We use near-infrared light interference microscopy to generate interference fringes(Fig. 2 shows the schematic of near-infrared light interferometer). This interference system is an improvement on the Linnik optical interferometer with 168× measurement area magnification. The near-infrared light emitted by the near-infrared light source is split into two beams using a beam splitter. One is reflected by the reference mirror and the other meets the object to be tested whereupon it is reflected or transmitted. The interference fringes generated by the two beams are obtained by infrared CCD(pixel 320×256)[9]. High-precision piezoceramics is allowed for vertical movement in a range of 400 μm. Data acquisition and topographic reconstruction are completed by self-developed dedicated software. To avoid error in results, measurements are taken in a quiet and dark environment.

      图  2  近红外光干涉系统示意图

      Figure 2.  Schematic of near-infrared light interferometer

    • Vertical scanning interferometry(VSI) is a typical algorithm for reconstructing the three-dimensional topography of a microstructure for its high precision and fast operation. It measures relative height by selecting fringes at zero phase[10-12]. The basic principle of near-infrared light interference is similar to that of white light interference. Therefore, a near-infrared light source with a center wavelength of 1 170 nm and a wavelength range from 1 000 nm to 1 400 nm is selected.

      According to the principle of white light interference, the intensity of the near-infrared interferogram can be expressed using a standard relationship:

      (1)
      (2)

      where I1 and I2 are the intensities of the two beams, δ is the phase difference between I1 and I2, λ is the wavelength of the near-infrared light, and Δ is the difference in the optical paths of I1 and I2. For the reflection interference Δ=L1-L2, and for the transmission reflection interference Δ=L1L-(L2+nh), L1 and L2 are the reference optical path and the optical path of the object being tested in the reflection interference, ΔL is the reference optical path compensation in the transmission reflection interference, n is the refractive index of the transmission material(GaAs material has a refractive index of approximately 3.45 in the near-infrared band), and h is the thickness of the transmissive material. For the near-infrared broad-spectrum light source, the interference pattern obtained by the infrared light CCD can be regarded as the result of the interference superposition of different wavelengths. Therefore, the intensity distribution can also be expressed as[13]:

      (3)

      where I0 is the intensity of the interferogram, λc is the center wavelength of the near-infrared light, 2λb is the bandwidth of the spectrum, ψ(λ) is the energy distribution of the interference pattern near the wavelength λ, z is the relative position of PZT and zp is the position corresponding to the zero optical path difference[10]. Since the selected near-infrared light source has a very short coherence length, when each pixel is at the zero position of the optical path difference (z=zp), the intensity of the interference is at its highest and the three-dimensional topography of the microstructure can be reconstructed by recording the zp[10] corresponding to each pixel.

      Due to intrinsic absorption, the absorption process of electrons from the valence band to the conduction band gets excited. The intrinsic absorption wavelength of a semiconductor single crystal material is expressed by λg. When light illuminates the material, light with a wavelength greater than λg will penetrate, and light with a wavelength less than λg will be absorbed. As intrinsic absorption, the relationship between electron energy and the band gap is described as follows:

      (4)

      where h is the Planck constant, c is the speed of light and Eg is the band gap width. For the gallium arsenide(GaAs) material used in this experiment, the Eg was 1.43 eV and the minimum transmission wavelength was 0.86 μm.

    • First, standard micro-structures were designed to verify the feasibility and accuracy of the system. The standard samples were made of GaAs semiconductor materials with two different step heights(as shown in Fig. 3). In order to enhance the reflection effect, a layer of gold was plated on the surface of the GaAs and then polished. In order to verify the feasibility of the measurement system, the two steps were measured by SEM(step A and step B), producing measurements of 2.087 μm and 0.746 μm, respectively. As shown in Tab. 1, the SEM measurement results were used as a standard reference to verify the performance of the designed system.

      图  3  测量样品结构图:俯视图和剖面图

      Figure 3.  Structure of the test object: the top view and the cross-section

      表 1  不同测量方法台阶相对高度对比

      Table 1.  Comparison of relative step heights for different measurement methods

      Measurement method Step A/μm Error/% Step B/μm Error/%
      SEM 2.087 0.746
      Near-infrared light reflection interference 2.107 0.96 0.759 1.74
      Near-infrared transmission interference 2.132 2.16 0.766 2.68

      Next, the microstructure morphology was reconstructed by near-infrared light reflection interference technique. At this time, the optical path difference between the two paths was Δ=L1-L. Five phase-shifting interferograms were recorded to have a scatter stripe pattern. The piezoelectric ceramic stepping was 0.01 μm, corresponding to a phase step of 0.061 rad. The interferogram was shown in Fig. 4(a) and the three-dimensional shape obtained after the solution was calculated. The results were very consistent with the actual topography shown in Fig. 3. The accuracy of the measurement results is proven by the diagonal appearance of extracted pixels from (0, 256) to (256, 0), as shown in Fig. 4(b). The measurement results show that the relative heights of the two steps are 2.107 μm and 0.759 μm, which is close to the measurement results from the SEM, with relative errors of 0.96% and 1.74%, respectively. Due to the influence of the diffraction effect, the measured shape of the edge of the step is distorted, which is a common problem in interferometers[14-15].

      图  4  近红外光反射干涉测量结果

      Figure 4.  Measurement results with near-infrared light reflection interference

      Under the same interference microscope, the near-infrared light transmission reflection interference technology replaces the near-infrared light reflection interference. At the same time, the optical path of the object to be tested increases by nh. Because of this, optical path compensation is performed on the reference path, and ΔL=nh. In order to verify whether the near-infrared light is transmitted through the GaAs material, the interferogram of the different layers is obtained by adjusting the PZT position. As seen in Fig. 5(a), the upper interference fringes have no change at the position of the step and the interference fringes in the lower layer have changed. This proves that the infrared light has passed through the GaAs material. By adjusting the position of the reference mirror and controlling the piezoelectric ceramics using a computer with a scanning interval of 0.01 μm the interferogram was obtained. At the same time, the results were corrected due to the uneven upper surface of the transmitted layer. As shown in Fig. 5(a), the calculated three-dimensional topography is largely identical to the topography shown in Fig. 4(a). Also, to describe the experimental results, the diagonal topography from pixels (0, 256) to (256, 0) were extracted, as shown in Fig. 5(b). As shown in Tab. 1, the relative heights of the two steps are 2.132 μm and 0.766 μm, respectively, which are fundamentally consistent with the results of the SEM and the near-infrared light reflection interference, with relative errors of 2.16% and 2.68%, respectively. Due to the diffraction effect, the topography on the edge of the step is similarly distorted.

      图  5  光程补偿近红外透射反射干涉测量结果

      Figure 5.  Measurement results using near-infrared light transmission reflection interference with optical path compensation

    • A near-infrared light interference measurement system was built and the relative heights of steps were obtained using the near-infrared light transmission reflection interference technology. The internal three-dimensional shape reconstruction of a microstructure was achieved. The morphological measurement results of the standard sample were consistent with the SEM measurement results. When the near-infrared light transmission interference technology was used instead of the near-infrared light reflection interference technology, the system could be measured normally. The optical path compensation near-infrared light transmission reflection interference measurement results were 2.132 μm and 0.766 μm, respectively, corresponding to a 2.16% and 2.68% relative error, which is good enough to measure the height and high-aspect ratio microstructure. At present, the measurement error is large and there are two main reasons for this. The first is that the big difference between the near-infrared light and the white light spectrum introduces chromatic aberration. The second is that the reconstruction algorithm requires further improvement.

      ——中文对照版——

    • 近年来, 高深宽比的微结构以其灵敏度高、位移大等优点在微机电系统领域得到了广泛的应用[1], 而其宽度小(约1~10 μm)深度大(约10~500 μm)高深宽比(10:1至100:1)的特点, 使如何实时在线重构高深宽比底部及侧壁三维形貌成为一个亟待解决的问题[2]。目前, 测量微结构三维形貌的仪器主要有3种:扫描电子显微镜[3](Scanning Electron Microscope, SEM)、原子力显微镜[4-5](Atomic Force Microscope, AFM)、白光干涉仪[6]。SEM是用极狭窄的电子束去扫描样品, 在测量高深宽比结构时, 需要切开微结构制成剖面样品, 本身就是一种破坏、损伤性的测量方法, 而且测量时间长成本高; AFM是通过显微探针受力的大小直接换算出样品表面的高度, 由于高深宽比结构的特殊性, 测量时, 需要改进针尖结构使其深入到结构内部进行测量, 但针尖的改进面临很大的困难; 相比于其它两种仪器, 白光干涉仪更适合用于测量微结构的表面形貌, 而且已经有很多商用设备, 如Polytec、Talysurf、Veeco、Zygo等, 但对于高深宽比微结构形貌测量, 白光干涉仪也有其局限性:一方面白光衍射比较严重, 另一方面由于测量时光线与待测物品有一定角度, 光线无法完全通过窄狭缝, 因此很难测量此结构的底部及侧壁形貌[7-8]。综上所述, 基于白光干涉原理, 利用半导体材料(砷化镓GaAs或硅Si)在红外光波段可以透射的特性, 如果将红外光代替白光干涉的白光光源, 构建红外光干涉系统并实施一定的光程补偿, 采用红外光透射反射干涉技术, 即可实现高深宽比微结构的底部及侧壁形貌重构。

      本文以近红外光波段为例, 提出光程补偿近红外光透射反射干涉技术, 对应的干涉测量系统由近红外光源、干涉显微镜、红外CCD、高精度压电陶瓷和数据采集系统组成。首先研究了基于垂直扫描干涉技术的近红外光干涉形貌重构技术, 以及GaAs材料在红外光波段透射的基本原理, 其次设计了具有两个台阶的典型GaAs材料微结构样品, 并对其内部形貌进行测量, 最后通过与SEM和近红外光反射干涉的测量结果比较, 证明光程补偿近红外光透射反射干涉系统能够测量高深宽比微结构内部形貌。

    • 图 1所示为近红外光透射反射干涉测量高深宽比微结构底部及侧壁形貌示意图, 可知通过一定的光程补偿方法, 即可实现底部及侧壁形貌的重构。

    • 利用近红外光干涉显微仪产生干涉条纹, 图 2为近红外光干涉系统装置图及原理示意图。此干涉系统是基于168倍测量区域放大率的Linnik光学干涉仪进行改进的。近红外光源发射的近红外光通过分束器分成两束, 一路经参考镜后反射, 另一路经待测物品后反射或透射反射, 两束光产生的干涉条纹由红外光CCD(像素320×256)获取[9], 高精度压电陶瓷允许在400 μm的垂直范围内移动, 数据采集和形貌重建由自行编制的专用软件完成。为了避免将误差导入到结果中, 测量是在安静和阴暗的环境中进行的。

    • 垂直扫描干涉法(VSI)由于精度高、操作速度快, 是一种典型的微结构三维形貌重建算法, 其通过选择在零相位处的条纹测量其相对高度[10-12]。近红外光干涉基本原理与白光干涉的原理相似, 因此光源选择了中心波长为1 170 nm, 波长范围从1 000 nm到1 400 nm的近红外光光源。

      根据白光干涉原理, 可以用标准关系表示近红外光干涉图的强度:

      (1)
      (2)

      其中, I1I2是两光束的光强, δI1I2的相位差, λ是近红外光波长, ΔI1I2的光程差, 对于反射干涉情况Δ=L1-L2, 对于透射反射干涉情况Δ=L1L-(L2+nh), 其中, L1L2是反射干涉情况下参考路光程和待测物光程, ΔL为透射反射干涉情况下参考路补偿的光程, n为透射材料的折射率(GaAs材料在近红外光波段折射率约为3.45), h为透射材料的厚度。对于近红外光宽谱光源, 由红外光CCD获取的干涉图样可以看做是不同波长干涉叠加的结果, 因此强度分布也可以表示为[13]:

      (3)

      其中, I0是干涉图的强度, λc是近红外光中心波长, 2λb是光谱的带宽, ψ(λ)为在波长λ附近干涉图样的能量分布, z为PZT的相对位置, zp为零光程差对应的位置[10]。由于所选择的近红外光源相干长度很短, 当每个像素点在零光程差位置(z=zp)时, 干涉强度最大, 通过记录每个像素点对应的zp[10], 即可重构微结构的三维形貌。

      由于本征吸收, 电子从价带到导带的吸收过程被激发。半导体单晶材料的本征吸收波长用λg表示。当光照射材料时, 波长大于λg的部分光会穿透, 波长小于λg的光将被吸收。作为本征吸收, 电子能和能带隙的关系描述如下:

      (4)

      其中, h是普朗克常数, c是光速, Eg是带隙宽度。对于本实验所用的砷化镓(GaAs)材料, Eg为1.43 eV, 最小的透射波长为0.86 μm。

    • 首先, 设计了一个标准待测样本(Micro-structures), 以验证系统的可行性和准确性, 标准样品由GaAs半导体材料制作, 其有两个不同高度的台阶(如图 3所示)。为了增强反射效果, 在GaAs表面镀了一层金并进行了抛光处理。为了验证该测量系统的可行性, 利用SEM对两个台阶进行了测量(台阶A和台阶B), 其结果分别为2.087 μm和0.746 μm, 如表 1所示。其中, SEM测量结果做为一个参考标准来验证所设计系统的性能。

      其次, 利用近红外光反射干涉技术重构了微结构形貌, 此时两路光的光程差为Δ=L1-L。用分散条纹图案记录了5个相移干涉图, 压电陶瓷步进为0.01 μm, 对应0.061 rad的相位步进, 干涉图如图 4(a)所示。解算后所得的三维形貌与图 3所示的实际形貌非常一致, 提取从像素(0, 256)到像素(256, 0)的对角线形貌以来证明测量结果的准确性, 如图 4(b)所示。测量结果显示, 这两个台阶的相对高度是2.107 μm和0.759 μm, 接近于SEM测量结果, 分别对应0.96%和1.74%的相对误差。由于衍射效应的影响, 台阶边缘所测量形貌发生了失真, 这个是干涉仪常见的问题[14-15]

      最后, 在同样的干涉显微镜下, 用近红外光透射反射干涉技术代替近红外光反射干涉技术, 同时由于待测物品光程增加了nh, 因此在参考路进行光程补偿ΔL, 且ΔL=nh。为了验证近红外光是否透过GaAs材料, 通过调整PZT位置获得了不同层的干涉图样, 如图 5(a)所示。由图 5可以看出, 上层干涉条纹在台阶边缘位置没有任何变化, 而下层的干涉条纹发生了变化, 证明红外光已经透过GaAs材料, 调整参考镜的位置并通过计算机控制压电陶瓷以0.01 μm的扫描间隔获取干涉图样, 同时由于透射的上层表面不均匀, 对结果进行了矫正, 如图 5(a)所示。所解算的三维形貌与图 4(a)中所示的形貌基本一致。为了清晰地描述实验结果, 提取从像素(0, 256)到像素(256, 0)的对角线形貌, 如图 5(b)所示。表 1表明这两个台阶的相对高度分别为2.132 μm和0.766 μm, 这与扫描电镜和近红外光反射干涉测量结果基本一致, 分别对应2.16%和2.68%的相对误差。同样地, 由于衍射效应, 在台阶边缘上的形貌有失真现象。

    • 本文搭建了近红外光干涉的测量系统, 利用光程补偿近红外光透射反射干涉技术获得了台阶的相对高度, 实现了微结构的内部三维形貌重构。标准样品的形貌测量结果与SEM测量结果一致, 当用光程补偿近红外光透射反射干涉技术代替近红外光反射干涉技术时, 该系统可以正常测量。光程补偿近红外光透射反射干涉技术测量结果分别为2.132 μm和0.766 μm, 分别对应2.16%和2.68%相对误差, 结果表明其能够测量高深宽比微结构底部及侧壁形貌。目前测量误差较大, 主要原因有两方面, 一方面近红外光和白光光谱相差大引入了色差, 另一方面是重构算法需要改进。

参考文献 (15)

目录

    /

    返回文章
    返回