Part of the  

Solid State Technology

  and   

The Confab

  Network

About  |  Contact

Posts Tagged ‘oxide’

Picosun and Hitachi MECRALD Process

Friday, February 24th, 2017

thumbnail

By Ed Korczynski, Sr. Technical Editor

A new microwave electron cyclotron resonance (MECR) atomic layer deposition (ALD) process technology has been co-developed by Hitachi High-Technologies Corporation and Picosun Oy to provide commercial semiconductor IC fabs with the ability to form dielectric films at lower temperatures. Silicon oxide and silicon nitride, aluminum oxide and aluminum nitride films have been deposited in the temperature range of 150-200 degrees C in the new 300-mm single-wafer plasma-enhanced ALD (PEALD) processing chamber.

With the device features within both logic and memory chips having been scaled to atomic dimensions, ALD technology has been increasingly enabling cost-effective high volume manufacturing (HVM) of the most advanced ICs. While the deposition rate will always be an important process parameter for HVM, the quality of the material deposited is far more important in ALD. The MECR plasma source provides a means of tunable energy to alter the reactivity of ALD precursors, thereby allowing for new degrees of freedom in controlling final film properties.

The Figure shows the MECRALD chamber— Hitachi High-Tech’s ECR plasma generator is integrated with Picosun’s digitally controlled ALD system—from an online video (https://youtu.be/SBmZxph-EE0) describing the process sequence:

1.  first precursor gas/vapor flows from a circumferential ring near the wafer chuck,

2.  first vacuum purge,

3.  second precursor gas/vapor is ionized as it flows down through the ECR zone above the circumferential ring, and

4.  second vacuum purge to complete one ALD cycle (which may be repeated).

Cross-sectional schematic of a new Microwave Electron Cyclotron Resonance (MECR) plasma source from Hitachi High-Technologies connected to a single-wafer Atomic Layer Deposition (ALD) processing chamber from Picosun. (Source: Picosun)

The development team claims that MECRALD films are superior to other PEALD films in terms of higher density, lower contamination of carbon and oxygen (in non-oxides), and also show excellent step-coverage as would be expected from a surface-driven ALD process. The relatively density of these films has been confirmed by lower wet etch rates. The single-wafer process non-uniformity on 300mm wafers is claimed at ~1% (1 sigma). The team is now exploring processes and precursors to be able to deposit additional films such as titanium nitride (TiN), tantalum nitride (TaN), and hafnium oxide (HfO). In an interview with Solid State Technology, a spokesperson from Hitachi High-Technologies explained that, “We are now at the development stage, and the final specifications mainly depend on future achievements.”

The MECR source has been used in Hitachi High-Tech’s plasma chamber for IC conductor etch for many years, and is able to generate a stable high-density plasma at very low pressure (< 0.1 Pa). MECR plasmas provide wide process windows through accurate plasma parameter management, such as plasma distribution or plasma position control. The same plasma technology is also used to control ions and radicals in the company’s dry cleaning chambers.

“I’m really impressed by the continuous development of ALD technology, after more than 40 years since the invention,” commented Dr. Tuomo Suntola, and the famous inventor and patentor of the Atomic Layer Deposition method in Finland in 1974, and member of the Picosun board of directors. “Now combining Hitachi and Picosun technologies means (there is) again a major breakthrough in advanced semiconductor manufacturing.”

MECRALD chambers can be clustered on a Picosun platform that features a Brooks robot handler. This technology is still under development, so it’s too soon to discuss manufacturing parameters such as tool cost and wafer throughput.

—E.K.

Indium-free Perovskite TCOs Could Save Costs

Monday, January 4th, 2016

thumbnail

By Ed Korczynski, Sr. Technical Editor

Lei Zhang, et al. from Pennsylvania State University—with collaborators from Rutgers University and University of Toledo—have found two new families of transparent conductive oxides (TCO) based on “correlated” electrons in ternary oxides of vanadium. From reported first principles, the co-authors are confident they will find many other correlated materials that behave like strontium vanadate (SrVO3) and calcium vanadate (CaVO3), which could make flat panel displays (FPD) and photovoltaic (PV) modules more affordable.

The correlation relies on strong electron–electron interactions resulting in an enhancement in the carrier effective mass. Both SrVO3 and CaVO3 demonstrate high carrier concentration (>2.2 ×1022 cm−3), and have low screened plasma energies (<1.33 eV). The Figure shows that there is a transparency trade-off in using these new TCOs, since at nominal 10nm thickness they are more than twice as opaque as Indium tin oxide (ITO).

Optical transmission of free standing conductive oxide films at 550nm wavelength, accounting for reflection and interference, and averaged over the range of the visible spectra. (Source: Nature Materials)

ITO has been the dominant TCO used in FPD manufacturing, but the price of indium metal has varied over the range of $100-1000/kg in the last 15 years. Consequently, industry has long searched for a TCO made of less expensive and less variable direct materials. Currently vanadium sells for ~$25/Kg, while strontium is even cheaper. Lei Zhang, lead author of the Nature Materials article (http://dx.doi.org/10.1038/nmat4493) and a graduate student in assistant professor Roman Engel-Herbert’s group, was the first to recognize the application.

“I came from Silicon Valley where I worked for two years as an engineer before I joined the group,” says Zhang. “I was aware that there were many companies trying hard to optimize those ITO materials and looking for other possible replacements, but they had been studied for many decades and there just wasn’t much room for improvement.” Engel-Herbert and Zhang have applied for a patent on this technology.

The U.S. Office of Naval Research, the National Science Foundation, and the Department of Energy funded this R&D. “Now, the question is how to implement these new materials into a large-scale manufacturing process,” said Engel-Herbert. “From what we understand right now, there is no reason that strontium vanadate could not replace ITO in the same equipment currently used in industry.”

Electrons flow like a liquid

Correlated oxides are defined as metals in which the electrons flow like a liquid, unlike conventional metals such as copper and gold in  which electrons flow like a gas. “We are trying to make metals transparent by changing the effective mass of their electrons,” Engel-Herbert says. “We are doing this by choosing materials in which the electrostatic interaction between negatively charged electrons is very large compared to their kinetic energy. As a result of this strong electron correlation effect, electrons ‘feel’ each other and behave like a liquid rather than a gas of non-interacting particles. This electron liquid is still highly conductive, but when you shine light on it, it becomes less reflective, thus much more transparent.”

In the November 2007 issue of the prestigious Physical Review B (DOI:  10.1103/PhysRevB.76.205110), F. Rivadulla et al. reported on “VO: A strongly correlated metal close to a Mott-Hubbard transition.” Vanadium oxide (VO) has a rocksalt cubic crystal structure, and displays strongly correlated metallic properties with non-Fermi-liquid thermodynamics and an unusually strong spin-lattice coupling. The structural and electronic simplicity of 3D monoxides provides a basic understanding of highly correlated electron systems, while this new work with 2D ternary oxides is inherently more complex.

One positive aspect of the more complex perovskite structure of SrVO3 and CaVO3 is that it provides for intriguing device integration possibilities with other functional perovskite materials. PV devices based on thin-films of complex perovskites have demonstrated excellent photon-electron conversion efficiencies in labs, but commercial manufacturing has so far been limited by the lack of an inexpensive TCO that can be integrated into a moisture barrier. The templating effect of underlayers could allow for faster deposition of more ideal SrVO3.

—E.K.