by Chi-I Lang, VP of workflow products and applications, Intermolecular Inc.
June 30, 2011 - Today marked the conclusion of the American Vacuum Society’s 2011 Atomic Level Deposition Conference in Cambridge, MA. Unlike many conferences, hardly anyone left on "getaway day," and all the sessions saw solid attendance. Much of the subject matter today was very in-depth, as presenters dug into some of the nitty-gritty process details of various flavors of ALD and its applications, including solar cell and fuel cell production.
|Atomic Level Deposition Conference 2011|
|Day 1: Interface engineering, rabbit ears and Roy Gordon|
|Day 2: Manufacturability takes center stage|
|Day 3: Precursor needs, spatial ALD, and butterfly wings|
One cross-cutting trend, picking up a theme from Harvard Professor Roy Gordon’s talk on Sunday, was a broad desire for more access to the myriad precursors that can be used in ALD. Many presenters in the solar cell sessions discussed aluminum oxide deposition, but its popularity seemed to be largely due to the ready availability of precursors.
It’s important to recognize that from a chemical company’s perspective, making a new line of precursors is not a trivial task, especially in the face of uncertain markets. And choosing which ones to make is even more difficult, as there’s a definite chicken-and-egg situation. But the bottom line is that every time a speaker mentioned the need for more precursor options, lots of heads in the audience started nodding.
In addition, there is strong interest in precursors that are less expensive than the semiconductor-grade chemicals that make up the bulk of today’s offerings, which are not necessary for many applications. Moreover, today’s chemicals are optimized for use on silicon, and those exploring germanium and III-V materials would like equal attention to prevent surface-interface issues.
Another interesting dynamic is the jockeying for position between traditional time-dependent ALD, where the substrate remains stationary and throughput is dependent on pulse and purge efficiency, and spatial ALD, where the substrate is transported past stations that dispense precursors and oxidizers (typically, air curtains are used to prevent crosstalk between adjacent stations). Spatial processes seem to be emerging as the leading contender in cost-sensitive and high-throughput applications that do not require cleanroom conditions. Speaker C.S. Hwang of Seoul National University said he does not see a future for spatial ALD in semiconductor production, due to issues with particle contamination on moving objects.
Our last high-level observation is that ALD seems to be enabling a separation of chemical and mechanical engineering in device development. We saw a number of situations that leveraged ALD’s ability to produce films with very high mechanical strength; ALD deposition can be used to create an "exoskeleton" over less robust films, with excellent step coverage and uniformity.
- Harvard’s J. Heo, a student of Prof. Gordon, speaking on tin(II) sulfide deposition. The caliber of students in Gordon’s program is very high; Heo gave an extremely rigorous and in-depth review of film growth, including characterization and control methods.
- Several presentations on the use of aluminum oxide as a passivation layer on solar cells. W.M.M. Kessels of the University of Eindhoven outlined the interface defect density difference between using Al2O3 and SiO2 passivation. For Al2O3, he also pointed out a performance difference between PEALD and thermal ALD Al2O3 process, due to the plasma UV radiation in the PEALD process. Multiple papers on ALD’s use in dye-sensitized photovoltaics, including an interesting one by Pedro Cunha of Cambridge University exploring the use of titania to create gyroid-shaped structures for use in high surface-area cells. Down the road, these low-cost, flexible PV materials could be incorporated into the surface of electronic systems, allowing them to recharge themselves — sounds handy!
- An intriguing discussion of ALD’s use in replicating the surface structures of butterfly wings, by a multi-university team from Belgium. The work ties into the emerging field of bio-mimicry, which seeks to utilize naturally occurring structures for new applications. Butterfly wings have a complexly textured crystal structure on their surface, which scatters light; this structure may prove useful in solar cell development. ALD’s ability to create very fine conformal coatings allowed the researchers to make extremely precise casts of wing surfaces — a truly remarkable demonstration.
- Several presentations on fuel cell development, which leveraged ALD in a number of roles including cost-effective coatings for cathodes, catalyst development, and passivation.
- Additional comments by Prof. Hwang, on semiconductor applications of ALD. He was very bullish, saying that future DRAM generations will draw heavily on ALD, as no other technology can do what it does in terms of film quality and thickness and aspect ratio. He added fuel to the precursor discussion, pointing out that even similar compounds can produce very different results. He cited research comparing Ge(II) and Ge(IV), which have different valence charges; Ge(II) gave the right result, which is in line with our experience at Intermolecular. Hueng noted that if Ge(IV) is the only commercially available precursor, we are doomed — a strong message to chemical suppliers!
- A group of people who discussed the emerging field of molecular level deposition (MLD). This process seeks to improve on the slow deposition rates of ALD, and can achieve 3-8