中文摘要：

提出了一种基于谱域相位分辨光学相干层析的纳米级表面形貌成像方法,由干涉光谱计算样品相邻两点的相位差,得到样品表面相位差分图,经过积分,重建样品表面形貌的定量分布.当相邻两点相位差的绝对值小于π,不产生相位包裹,避免了目前的干涉法相位解包裹存在的问题,将干涉法相邻两点相位差绝对值的限制条件由目前的π扩大到2π,提高了干涉法表面形貌成像的适用范围.参考面和样品置于同一平台之上,消除环境干扰及系统振动的影响,噪声幅度小于0.3 nm.通过对光学分辨率片及表面粗糙度标准样板的表面形貌成像,对本方法进行了验证,系统的轴向分辨率优于1 nm.

英文摘要：

Microscopic surface topography plays an important role in studying the functions and properties of materials. Microscopic surface topography measurement has been widely used in many areas, such as machine manufacturing, electronic industry and biotechnology. Optical interferometry is a popular technique for surface topography measurement with an axial resolution up to nanoscale. However, the application of this technique is hampered by phase wrapping, which results in a limited measurement range for this technique. Various digital algorithms for phase unwrapping have been proposed based on the phase continuity between two adjacent points. However, several significant challenges still exist in recovering correct phase with this technique. Optical coherence tomography （OCT） is a non-contact three-dimensional imaging modality with high spatial resolution, and it has been widely used for imaging the biological tissues. In this paper, we demonstrate a method for nanoscale imaging of surface topography by using common-path phase-resolved ~pectral domain OCT to reduce the influence of phase wrapping. The system includes a superluminescent diode with a central wavelength of 1310 nm and a spectral bandwidth of 62 nm, an optical fiber circulator, a home-made spectrometer, and a reference arm and a sample arm in common-path arrangement. The reference mirror and the sample under investigation are positioned on a same stage in order to further reduce the influence of ambient vibration. The phase difference between two adjacent points is calculated by performing Fourier transform on the measured interferometric spectrum. The phase difference distribution of the surface is obtained first. And then, the surface topography of the sample is constructed by integrating the phase difference distribution. In the traditional methods, phase wrapping occurs if the absolute value of the measured phase is greater than n. However, in the present method, phase wrapping occurs if the absolute value of the phase difference between two adja