Page 28 - ChipScale_Jan-Feb_2020-digital
P. 28
pixel size (24µm) (Figure 8c). The data the demands of full-production wafer 5. J. M. Trujillo-Sevilla, O. Casanova-
was collected in a single snapshot with fabs. The current WFPI system is made Gonzaleza, S. Bonaque-González, J.
an exposure time of 0.1s. for samples with a maximum diameter of Gaudestad, J. M. Rodríguez-Ramos,
The data collected using the above 50mm. With different optics, however, “High-resolution wavefront phase
procedu re was analyzed th rough one can easily make a system that works sensor for silicon wafer metrology,”
the WFPI algorithm to calculate the on a full 300mm diameter sample, which SPIE Photonics West, San Francisco
wavefront phase and the associated is currently being built, that will collect CA, USA (Feb. 2019).
amplitude (Z height). The filtered more than 10 million data points on a 6. J. W. Goodman, Introduction
topography data reveals the higher spatial 300mm wafer. to Fourier Optics, Roberts and
frequencies, and clearly showing the low- As WFPI has shown to be a very Company Publishers (2005).
amplitude depth map associated with NT good candidate for FEOL and FBEOL 7. K. Harding, ed., Handbook of
and roughness caused by wafer polishing. applications, using the correct spatial Optical Dimensional Metrology,
filtering technique, it can also perform CRC Press, 2013.
Summary global wafer geometry measurements 8. C. Beitia, M. Abdel, M. Cordeau,
The comparison between WFPI on patterned wafers during the BEOL S. Godny, S. Petitgrand, D. Alliata,
and chromatic confocal microscopy process to help process-i nduced “Optical profilometry and AFM
was done using standard mirrors with overlay errors during lithography steps measurements comparison on low
known bow. The results showed that [1]. This work is currently in progress amplitude deterministic surfaces,”
WFPI has higher speed and lower noise with very good initial results; the first Adv. Semi. Manufacturing Conf.
than the chromatic confocal microscopy official results are expected in the first (ASMC), 2019.
methodology. The current demo system half of 2020. 9. B. S. Chun, K. K. Kim, D. Gweon,
is made for samples with a maximum “Three-dimensional surface profile
diameter of 50mm. However, with Acknowledgment measurement using a beam scanning
different optics one can easily make This article is based on a paper chromatic confocal microscope,”
a system that works on a full 300mm presented at the International Wafer Level Review of Scientific Instruments,
diameter sample. Also, to improve the Packaging Conference (IWLPC) 2019. 80(7), 073706 (2009).
number of data points, one can use 10. S. Cha, P. C. Lin, L. Zhu, P. C. Sun,
a higher pixel image sensor without References Y. Fainman, “Nontranslational
increasing the data acquisition time 1. https://www.jedec.org/standards- three-dimensional profilometry by
very much—this is something that will documents chromatic confocal microscopy
help when collecting data for a complex 2. P. Vukkadala, K. T. Turner, J. K. with dynamically configurable
warpage situation. Sinhaa, “Impact of wafer geometry micromirror scanning,” Applied
WFPI was proven to reveal roughness on CMP for advanced nodes,” Jour. Optics, 39(16), 2605-2613 (2000).
by applying a high-pass filter on the of The Electrochemical Soc. (ECS), 11. https://www.cybertechnologies.
global topography data to analyze the 2011. com/en/products/ct-300/
higher spatial frequencies. WFPI has 3. SEMI M43-1109, “Guide for 12. https://nanovea.com/profilometers/
a lateral resolution of 24µm and an reporting wafer nanotopography,” 13. SEMI M1-0918 “Specification
amplitude sensitivity at 0.3nm. This www.semi.org (Oct. 2009). for polished single-crystal silicon
means that, while collecting data on the 4. J. Trujillo, J. M. Ramos-Rodriguez, wafers,” www.semi.org (Sept. 2018).
entire wafer in a single image snapshot, J. Gaudestad, “Wavefront phase
WFPI has the potential to become the imaging of wafer warpage,” Inter.
only wafer geometry technique capable of Wafer Level Packaging Conf.
revealing NT and roughness of the entire (IWLPC) San Jose, CA, USA (Oct.
wafer at a speed fast enough to satisfy 2018).
Biographies
Juan M. Trujillo-Sevilla is Chief Optical Engineer at Wooptix S. L., La Laguna, Tenerife, Spain. He leads the
optical research team focusing on light field and wavefront phase imaging, bringing these technologies to the
semiconductor metrology fab market. He was also a researcher on plenoptic imaging at U. de La Laguna where
he obtained his PhD as well as a Master’s in Bioengineering and a Bachelor’s in Electronics Engineering.
Email trujillo@wooptix.com
Jose M. Ramos-Rodriguez is the co-founder, CEO, and CTO of Wooptix S. L., La Laguna, Tenerife,
Spain—for which he raised Series A funding from Intel Capital. He received his PhD from IAC (Instituto de
Astrofísica de Canarias, Tenerife, Spain) and has been an Associate Professor at the U. of La Laguna since 2001, from which
he is currently on leave to grow Wooptix and commercialize his technology developments.
Jan Gaudestad is VP of Business Development at Wooptix, based in San Francisco, CA USA. His work focuses on
developing a new wafer geometry technique for in-line semiconductor manufacturing using wavefront phase imaging.
Previously, he was Director of BD at uSens, a VR startup. He received his MBA from Santa Clara U., a Master’s in Physics
from U. of Maryland, College Park, and his Master’s in Physics from the Norwegian U. of Science and Technology.
26 Chip Scale Review January • February • 2020 [ChipScaleReview.com]
26