Development of High-Resolution Nuclear Emulsion Plates for Synchrotron X-Ray Topography Observation of Large-Size Semiconductor Wafers

Characterization of defects in semiconductor wafers is essential for the development and improvement of semiconductor devices, especially power devices. X-ray topography (XRT) using synchrotron radiation is a powerful methods used for defect characterization. To achieve detailed characterization of large-size semiconductor wafers by synchrotron XRT, we have developed nuclear emulsion plates reaching a high-resolution and wide dynamic range. We have shown that higher-resolution XRT images could be obtained using emulsions with smaller iodobromide crystals, and demonstrated clear observation of threading edge dislocations in a SiC epitaxial layer having small contrast. Furthermore, we demonstrated XRT image acquisition for almost all of a 150-mm SiC wafer with one plate. Our development will contribute to advances in electronic materials, especially in the field of power electronics, in which defect characterization is important for improving the performance and yield of devices.


Introduction
Defects in semiconductor devices, especially power devices, adversely affect their performance and reliability. [1][2][3][4][5][6][7][8][9][10] Thus, great efforts have been made in crystal growth processes to reduce the defects in semiconductor wafers. [11][12][13][14][15][16] However, certain amounts of dislocations are included in compound semiconductor wafers, such as silicon carbide (SiC) and gallium nitride (GaN). Therefore, it is important to characterize the dislocations in semiconductor wafers. Compound semiconductor wafers have been characterized by various methods. [17][18][19][20][21][22][23][24][25][26] Among the various characterization methods, x-ray topography (XRT) is a powerful method of defect imaging for semiconductor wafers. [27][28][29][30][31] Thanks to their high photon flux and collimation and large beam size, synchrotron x-ray beams have been used for detailed analysis of the defects in wafers with high resolution. [32][33][34] Since there is no large-size digital x-ray detector with high resolution, researchers often use x-ray films and nuclear emulsion plates as detectors, and digitally record the image using an optical microscope with transmitted light. [35][36][37] Although large-size [more than 6 in (c.15 cm) square] x-ray films are commercially available, it is difficult to characterize the detailed features of defects, 1 3 such as threading edge dislocations (TEDs) in SiC having small contrast, in XRT images due to the lack of resolution. Although commercial nuclear emulsion plates, for which the area is limited to a 3-in (c.7.6 cm) square, can manage to observe TEDs, emulsion plates with larger area and higher resolution are necessary for detailed analysis of large-size SiC wafers. [38][39][40] Therefore, the use of holography films and the development of high-resolution nuclear emulsion plates have been pursued for detailed observations. 38,41 In the field of particle physics, nuclear emulsion plates have long been used and developed for the detection of cosmic rays. Muography (cosmic-ray muon radiography) is used to visualize the interior density distribution of largescale structures. [42][43][44][45][46][47][48][49] Recently, Morishima et al. discovered a large void in Khufu's Pyramid by observation of cosmic-ray muons using their developed nuclear emulsion plates. 49,50 In the present study, we developed a high-resolution nuclear emulsion plate for synchrotron XRT and demonstrated the observation of 150-mm SiC wafers by applying the technology used in the fabrication of nuclear emulsion plates for muon radiography.

Experimental
Nuclear emulsions comprise silver iodobromide crystals with different target diameters (R target ) embedded in gelatin. Details of the fabrication method for the emulsion plates are shown in Ref. 50 We prepared emulsion plates with an area of 1 inch × 3 inch (c. 2.5 cm x 7.6 cm) for the observation of small specimens and with an area of 150 × 150 mm 2 for 150-mm SiC wafers. The thickness of the emulsion layer is 20 μm. XRT observations with glazing incident geometry using a monochromatic x-ray beam with energy of 8.27 keV were conducted in the BL8S2 in the Aichi Synchrotron Radiation Center and the BL09 in the SAGA Light Source with an x-ray beam of approximately 150 mm in width. 34 The applied g vector was − 1−128. Specimens for the observation were 4H-SiC epitaxial wafers diced to 10 × 10 mm 2 and 150-mm bare wafers. Both wafers were off-cut at an angle 4° from the (0001) basal plane. The epitaxial layers on the wafers were of N type with a nitrogen concentration of 1.0 × 10 16 cm −3 and thickness of 10 μm. The particle size of iodobromide crystals in the emulsion was evaluated by transmission electron microscopy (TEM) using a JEM-2010F microscope (JEOL) operated at 200 kV. X-ray topography images were digitally recorded by an optical microscope with transmitted light with a pixel resolution of approximately 0.7 μm/pixel. Figure 1 shows the TEM images of iodobromide crystals with different target diameters in the prepared nuclear emulsion. Although the size of the iodobromide crystals with R target = 100 nm was approximately 70 nm, which is a bit smaller than the target diameter, the actual diameter of the crystals was almost the same as the target diameter. Under electron beam illumination, the silver iodobromide reacted with electrons by a reaction similar to image development: Figure 2 shows the evolution of iodobromide crystals with R target = 300 nm under electron beam illumination. Horn-shaped silver precipitates were formed and the successive formation of small silver precipitates resulted in collapse of the crystal shape after a few minutes. Thus, although we quickly took an image after illumination by the electron beam, the shapes and sizes of the iodobromide crystals in the TEM images were somewhat influenced by the reaction shown in Eq. 1. We are currently trying to establish a precise method for the evaluation of iodobromide crystal diameters without reaction with electrons. Figure 3 shows the XRT images of the SiC epitaxial wafers recorded on emulsion plates with crystals of different target diameters and a commercially available plate (Ilford). The XRT image recorded on the emulsion plate with the target iodobromide crystal diameter of 100 nm is much clearer than that recorded on the other emulsion plates. In these images, linear contrasts of basal plane dislocations (BPDs) on (0001), large white circular contrasts of threading screw dislocations (TSDs) having a c-component of the Burgers vector, and small contrasts of TEDs having the Burgers vector of 1/3 < 11-20 > were observed. The contrasts of the TEDs are different depending on the direction of the Burgers vector. For example, some TEDs, such as TED-I in Fig. 3a, were imaged as bright spots, while other TEDs, such as TED-II in Fig. 3a, were imaged as dark spots. The contrasts of dislocations were more clearly observed by our 100-nm emulsion plates than by the other plates; in particular, dark-contrast TEDs, which are often difficult to observe with commercially available emulsion plates, were definitely observed. Figure 4 shows XRT images of TEDs with the six different Burgers vector directions recorded on our 100-nm plates and on commercially available nuclear emulsion plates. All the types of TEDs were clearly observed by our 100-nm plates. In particular, the TEDs exhibiting dark contrasts were much more clearly recorded on our 100-nm plates than on commercially available nuclear emulsion plates. Furthermore, the dynamic range of the image seems to be wide, and the BPD-II in Fig. 3a, which are in the substrate beneath the epitaxial layer  and are difficult to image on commercially available emulsion plates, were observed by our 100-nm emulsion plates. Figure 5 shows the XRT images of a 150-mm SiC wafer recorded on our 150 × 150 mm 2 emulsion plate with R target = 100 nm and the appearance of the recorded emulsion plate. Note that 7,739 microscope images were acquired and stitched together as one image. We successfully observed almost a whole 150-mm wafer in one plate with high resolution and wide dynamic range, as shown in Fig. 5b. The contrasts of TEDs were clearly observed, and it is possible to identify the direction of the Burgers vector from the images. The exposure time for acquiring XRT by our 100-nm emulsion plates is almost the same as that for the commercially available x-ray films. With almost the same exposure time, the XRT images were much more clearly observed for our high-resolution nuclear emulsion plates than for commercially available x-ray films.

Results and Discussion
The results of the present study indicate that small size iodobromide crystals can obtain XRT images with high resolution as well as a wide dynamic range without reduction in the throughput of synchrotron experiments. Furthermore, our technology can be applied in the XRT observation of larger size wafers if a wide synchrotron x-ray beam is available. Our development will contribute to advances in electronic materials, especially in the field of power electronics, in which defect characterization is important for improving the performance and yield of devices.

Conclusions
A nuclear emulsion plate for XRT observation of largesize semiconductor wafers with high resolution and wide dynamic range was developed based on the technology established for muon radiography. We have shown that smaller size iodobromide crystals lead to a higher-resolution XRT image, and demonstrated clear observation of TEDs in a SiC epitaxial layer having small contrasts in an XRT  image, enabling us to confidently identify the direction of the Burgers vector. Furthermore, we have prepared nuclear emulsion plates with a size of 150 × 150 mm 2 and demonstrated XRT image acquisition for almost all of a 150mm SiC wafer with one plate. Our development of nuclear emulsion plates for XRT observation will contribute to the fields of electronics and crystal growth both in science and industry.

Note
Recently, a similar commentary with the same title without figures and references was published in Journal of Imaging in Interventional Radiology (Prime Scholars). 51 Although the contents themselves are very similar, our current paper shows the figures as evidence of our development and logically explains our development in detail with sufficient references.
Author Contributions SH and KM conceived and designed the present study with AT. NK prepared the nuclear emulsion plates with KM. TN prepared the specimens for TEM observation under the guidance of KM. TEM observation was conducted by SH with TN. XRT observation was conducted by SH, AT, KH. and KI. The manuscript was written by SH with feedback from all the authors.
Funding This paper was supported in part by a project, JPNP20004, commissioned by the New Energy and Industrial Technology Development Organization (NEDO).

Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.