Selective laser melting (SLM) is an additive manufacturing (AM) technology that uses a high-power laser beam to melt metal powder in chamber of inert gas. The process starts by slicing the 3D CAD data as a digital information source into layers to create a 2D image of each layer. Melting pool was formed by using laser irradiation on metal powders which then solidified to consolidated structure. In a selective laser melting process, the variation of melt pool affects the yield of a printed three-dimensional product. For three dimensional parts, the border conditions of the conductive heat transport have a very large influence on the melt pool dimensions. Therefore, melting pool is an important behavior that affects the final quality of the 3D object. To meet the temperature and geometry of the melting pool for monitoring in additive manufacturing technology. In this paper, we proposed the temperature sensing system which is composed of infrared photodiode, high speed camera, band-pass filter, dichroic beam splitter and focus lens. Since the infrared photodiode and high speed camera look at the process through the 2D galvanometer scanner and f-theta lens, the temperature sensing system can be used to observe the melting pool at any time, regardless of the movement of the laser spot. In order to obtain a wide temperature detecting range, 500 °C to 2500 °C, the radiation from the melting pool to be measured is filtered into a plurality of radiation portions, and since the intensity ratio distribution of the radiation portions is calculated by using black-body radiation. The experimental result shows that the system is suitable for melting pool to measure temperature.
We have developed a EUV scatterometer using a focused high-order harmonic generation (HHG) source for nano-scale
grating measurement. The coherent light source with multiple discrete wavelengths of 25-35 nm was pumped by a
tabletop Ti:sapphire laser system. A charge-couple-device (CCD) camera directly records the diffraction image of the
zero and the first order diffraction information from the grating samples. The grating structure can be reconstructed base
on the calculations from the location and the intensity distribution of diffraction pattern.
Overlay metrology for stacked layers will be playing a key role in bringing 3D IC devices into manufacturing. However, such bonded wafer pairs present a metrology challenge for optical microscopy tools by the opaque nature of silicon. Using infrared microscopy, silicon wafers become transparent to the near-infrared (NIR) wavelengths of the electromagnetic spectrum, enabling metrology at the interface of bonded wafer pairs.
Wafers can be bonded face to face (F2F) or face to back (F2B) which the stacking direction is dictated by how the stacks are carried in the process and functionality required. For example, Memory stacks tend to use F2B stacking enables a better managed design. Current commercial tools use single image technique for F2F bonding overlay measurement because depth of focus is sufficient to include both surfaces; and use multiple image techniques for F2B overlay measurement application for the depth of focus is no longer sufficient to include both stacked wafer surfaces. There is a need to specify the Z coordinate or stacking wafer number through the silicon when visiting measurement wafer sites. Two shown images are of the same (X, Y) but separate Z location acquired at focus position of each wafer surface containing overlay marks. Usually the top surface image is bright and clear; however, the bottom surface image is somewhat darker and noisier as an adhesive layer is used in between to bond the silicon wafers. Thus the top and bottom surface images are further processed to achieve similar brightness and noise level before merged for overlay measurement.
This paper presents a special overlay measurement technique, using the infrared differential interference contrast (DIC) microscopy technique to measure the F2B wafer bonding overlay by a single shot image. A pair of thinned wafers at 50 and 150 μm thickness is bonded on top of a carrier wafer to evaluate the bonding overlay. It works on the principle of interferometry to gain information about the optical path length of the stacked wafers, to enhance the image contrast of overlay marks features even though they are locating in different Z plane. A two dimensional mirror-symmetric overlay marks for both top and bottom processing wafers is designed and printed in each die in order to know and realize the best achievable wafer to wafer bonding processing. A self-developed analysis algorithms is used to identify the overlay error between the stacking wafers and the interconnect structures. The experimental overlay results after wafer bonding including inter-die and intra-die analysis results will be report in the full paper. Correlation of overlay alignment offset data to electrical yield, provides an early indication of bonded wafer yield.
We have demonstrated a full-field IR wavelength scanning interferometry system for the adhesive thickness
measurement which in between the temporary bonded wafer and a carrier wafer. The illumination wavelength can be
varied and selected by tilting the angle of interference filter along the main optical axis. The varying wavelength was
calibrated by a commercial spectrometer. By combining the phase-shifting technique and the spectrum curve fitting
method, the total thickness variation (TTV) of the adhesive layer and the adhesive thickness distribution map can be
obtained. The experimental results showed that the TTV of the adhesive is 3.76 μm within the area of 110 mm diameter. The thickness variation is in a range from 16.47 μm to 20.23μm.
As a type of optical measuring apparatus, the charge-coupled diode (CCD) camera provides the capability of increasing the speed of measurement by inspecting an area with only one shot. However, the CCD camera's high-variation range of reflectivity presents an exceptional challenge for the optical measurement established on the surface. We present a method that could enable one to acquire an image with a high-dynamic range in one shot without any reduction in spatial resolution. Because of the sufficient signal-to-noise ratio, the method presented could perform the robustness of the phase-retrieving algorithm, and the surface topography could be measured more accurately.