Industry Background

What is ALD?

Invented in the early 1970's by Finnish scientist Tuomo Suntola, ALD is a process by which thin film deposition occurs in a self-limited chemical reaction that results in atomic layer resolved growth. These films are superior to traditional CVD and PVD in terms of uniformity, conformality, and their pin-hole free quality.

 

Challenges and Opportunities

 

At the ALD 2005 conference in San Jose, CA held from August 7-10, invited speaker Dr. Sang Bom Kang of Samsung offered the following summary of the status of atomic layer deposition,

 

“atomic layer deposition has established its place as an alternative deposition technique in semiconductor industry especially for applications that require precise thickness controllability, low deposition temperatures, and excellent conformality. In order to extend the applications of ALD, methods to obtain higher quality films and higher productivity must be further investigated.”

Higher Quality Films:  This industry challenge is trying to be met with improved chemistries and the use of plasma processing to improve film quality in terms of density, adhesion, and contaminant levels as well as to be able to process at lower temperatures.  Many chemical and semiconductor companies as well as academic laboratories are engaged in active research to develop more and better precursor chemistries to improve film properties and broaden applications.  Equipment companies along with their customers are working to develop improved processes.  Atomicity has tried to help meet this challenge by developing a tool with flexibility in terms of temperature and flow control at the same time maintaining a very clean system in terms of trace element contamination and particle contamination.  This system allows process developers a wide range of options and approaches to develop the advanced processes of the future.

Higher Productivity:  This challenge has been largely divided in terms of higher throughput batch reactors vs. lower throughput single wafer reactors, each having their own issues.  In the large batch reactors where 25 or more wafers are processed at a time there is a high risk for product loss in the event of a system malfunction.  Also, there may be some significant variation across the batch reactor in terms of film quality.  Finally, wafer backside deposition also tends to occur in these type of reactors which may be undesirable.  Single wafer tools are currently only able to process ~20 wafers per hour resulting in a high cost of ownership/operation.  Atomicity has radically improved on the cycle rate without compromising on the deposition rate per cycle, resulting in a tool platform that has unprecedented throughput that is directly competitive with traditional CVD and PVD processing.  Further, Atomcity’s design reduces downtime with it ultra clean approach that maintains a high degree of separation between the precursors except at the point of application, the substrate.  This way, vacuum systems are able to operate for long periods between maintenance cycles.

Cost of Ownership:  While the CoO challenge was not explicitly stated in Dr. Kang’s summary above, it is a commonly appreciated reality in the semiconductor manufacturing world.  Atomicity has attempted to address this concern by developing a tool that has low maintenance and high precursor utilization at the same time providing unprecedented throughput possibilities.

Current Applications or Areas of Focus

 Semiconductor manufacturing

• DRAM – trench capacitor dielectric
• Gate – high-k gate oxide layer
• Metallization - Barrier, adhesion, and seed layer 
• Others?

Magnetic Media Storage (hard drives) 

• Magnetic disk read/write heads

Research and development

• Nanotechnology, MEMS, fuel cells, solar cells, etc.

Possible Future Applications

Optics - Filters, multiplexers, nano-waveguides, solid state lasers, etc.

Total Market

The total semiconductor capital equipment market fluctuates roughly between $20B-$30B per year with about 12% of that total, ~$2B-$3B, being from chemical vapor deposition (CVD) with some additional percentage from physical vapor deposition (PVD). Currently, given the widely accepted throughput and cost limitations of existing ALD platforms, ALD is expected to displace PVD and CVD in only some of the process areas where extremely demanding film quality is required. The conventional view is that this should result in an ALD semiconductor market on the order $300M/yr. Other existing applications include thin film heads (est. ~$50M/yr or less), nanotechnology (unknown but small initially), and basic research tools.  If ALD throughput and quality can exceed the traditional CVD and PVD approaches, Atomicity expects a much larger portion of the $2B-$3B per year deposition market to become dominated by ALD than what conventional market analysis suggests.

Major Players

 ASM, Applied Materials, Aixtron/Genus, Aviza, Novellus Systems, Tegal, Veeco, Vesta