Acquisition of partial grain orientation information using optical microscopy

Abstract

Optical microscopy (OM) is often the first technique used to characterize surfaces of materials because of its simplicity and low cost. However, the information obtained by OM is usually limited to the geometry and spatial distribution of microstructural features. We present an advanced OM technique to perform quantitative microstructure analysis of polycrystalline metals and to assess partial grain orientation information from measurements of reflectance over a range of incident light directions. Our technique relies on the collection of a series of optical micrographs under controlled illumination conditions. We find a marked direction-dependence of reflectance and ascribe it to sub-micron facets on the surfaces of individual grains. These surface facets are, in turn, correlated to grain orientation. We automate our technique via digital processing and numerical analysis of micrographs, enabling it to generate a wealth of microstructure data while remaining simple to use and inexpensive to implement.

Introduction

Optical microscopes are among the most ubiquitous characterization tools in research laboratories. They are often the simplest and most immediate choice for imaging the surface of many different materials. In polycrystals, optical microscopy (OM) may be used to characterize the geometry and distribution of microstructural features with dimensions of ∼1 μm and above [1], such as crystal grains [2], precipitates [3], or cracks [4]. However, many elements of polycrystalline microstructures—such as the crystallographic orientations of grains or features with dimensions well below ∼1 μm—cannot be directly assessed through OM. To characterize such quantities—which are pivotal to interpreting and predicting material properties—researchers usually rely on electron microscopy (EM), even though it is much more resource-intensive than OM [5], [6].

We show that computationally-driven, quantitative analysis of surface reflectance enables quantitative assessment of polycrystalline microstructures and markedly extends the range of microstructure information that may be obtained via OM. In particular, we demonstrate the acquisition of partial grain orientation information through the characterization of sub-micron surface roughness using OM. While OM cannot replace EM for fundamental microstructure studies, our techniques present a clear advantage over EM for rapid, inexpensive analyses of large surface areas without the need for high vacuum.

Electron backscatter diffraction (EBSD) is an EM-based technique which is routinely employed to perform quantitative metallography measurements [5] and to map grain crystallography [7]. Being able to perform EBSD-like measurements using OM would offer many advantages in the study of materials structure. In fact, capturing an optical micrograph is orders of magnitude faster than acquiring an EBSD scan and requires simpler, less expensive equipment. OM generally has a larger field of view than EM and thus may be used to characterize surfaces of much larger area than EBSD. Moreover, OM measurements may be carried out in water or air and do not require vacuum. Thus, characterization of polycrystalline microstructures by OM has the potential to enable routine, facile, and high-volume microstructure data acquisition [8]. It may be employed as a non-destructive technique to improve material reliability, for instance by assessing microstructure variability of as-processed components, or by monitoring the temporal evolution and integrity of ones already in service.

The hallmark of our method is measurement of the intensity of the reflected light as a function of both sample microstructure and incident light direction. For this reason, we name our method directional reflectance microscopy (DRM). We use a custom-made apparatus that allows controlling the light source orientation with respect to the imaged sample and we analyze the collected optical micrographs using MATLAB. We compare our DRM measurements with EBSD analyses using polycrystalline pure nickel (Ni) as a case-study material. We demonstrate that DRM may be used to perform quantitative metallography measurements—such as measurement of grain statistics—as well as to assess microstructural characteristics that are usually only resolvable by electron or X-ray techniques—such as sub-micron differences in surface roughness and partial crystallographic grain orientation information.

Section 2 describes sample preparation and the development of the apparatus used in our work. The remainder of the paper is divided into two main parts. In the first (section 3), we describe how to construct a DRM dataset and how to use it for quantitative microstructure analysis. We show that grain statistics obtained by DRM are more accurate than ones found by conventional EBSD analysis, in some cases. In the second part (section 4) we demonstrate the capability of DRM to characterize crystallographic orientation of grains by quantifying grain reflectance variations as a function of the incoming light direction. We conclude with a discussion of our findings in section 5.