Batch-cleaning tools, both immersion and spray type, are among the few remaining batch systems in leading-edge wafer fabs. Batch tools still command the lion's share of the wet clean and surface preparation equipment markets. Despite claims by some industry experts in the early 1990s that wet tools would be eliminated from fabs by the end of the last century, batch systems continue to be mainstays in today's fabs. This is because of batch tools' continuing ability to meet demanding process specifications and operational requirements. The selection of these tools is driven by the same factors as other types of fab equipment: process performance, high productivity, high throughput, and low cost of ownership (COO).

Batch wet tools account for around 80% of the total wet clean market, with single-wafer spin/spray tools and scrubbers making up the remaining portion. Unfortunately for equipment makers, batch immersion tools tend to be extremely price sensitive, with low margins and generally high field service and support costs. It has also been difficult to establish and maintain any significant technical differentiation in the automated wet-station market. Consequently, running a profitable auto-wet tool business has been quite difficult.

Although there have been numerous advances in batch wet-cleaning tool performance in the past decade, the financial performance of the companies that manufacture those tools has seen little improvement. As a result, there has been significant consolidation in the market, with a few companies exiting the space. Unless one of the surviving suppliers develops a significant form of sustainable competitive differentiation based on patent-protected intellectual property, the difficult economic conditions will likely persist in the batch-cleaning space.

On the process technology front, wet-cleaning solutions for batch tools have evolved from concentrated chemistries into more-dilute mixtures, resulting in much lower chemical usage. Significant efforts have also been made to reduce DI-water consumption. Furthermore, better process control has resulted in improved process results. The push for higher productivity in auto-wet immersion tools has led to a shift from huge systems with many tanks to tools with much smaller footprints and fewer tanks.

Reduced cleans with dilute chemistries in new batch single-tank systems offer higher productivity and throughputs, shorter cycle times, lower chemical consumption, and an even smaller footprint than previous versions, increasing their appeal to 300-mm fabs. All major suppliers offer reduced-footprint/ single-tank tools, and more and more fabs have adopted them for critical cleaning processes.

The adoption of single-wafer cleaning technology has been under way since the early 1990s. Single-wafer scrubbers are widely used for particle removal in back-end-of-line (BEOL) processes. Stand-alone single-wafer spin/spray processors have been primarily employed for enabling side-selective cleaning applications, such as backside particle removal (prelitho clean), backside film strips (mainly in front-end-of-line [FEOL] steps), backside and bevel-edge cleans (BEOL copper contamination removal), and wafer thinning.

Generally, these side-selective single-wafer applications have not taken any significant market share away from batch tools because they have created new markets for new and unique cleaning applications. However, more-recent single-wafer tools, designed for some conventional BEOL cleaning applications, are in direct competition with batch tools. In particular, there is a growing interest in using single-wafer wet cleans following BEOL dry-resist-strip steps. This application promises to be the largest near-term growth area for single-wafer spin/spray processors and certainly will enhance their overall market position at the expense of batch tools.

Cryogenic aerosol particle removal has found some niche BEOL applications but has not been widely adopted as a viable mainstream cleaning technique. Other new and exotic single-wafer cleaning technologies under development, including laser-assisted cleaning, have yet to be proven in the field.

Historically, the comparatively low throughput and productivity, and the high cost of single-wafer cleaning systems have been barriers to the technology's broad adoption for critical applications. However, certain system makers have tried to address these issues by offering multichamber, single-wafer wet-processing systems. The expected increases in throughput and productivity of these multichamber single-wafer systems will help narrow the operational gap with batch tools.

Single-wafer tools also promise to reduce chemical and DI-water consumption, while at the same time offering the potential to improve process control and uniformity. The importance of the aforementioned factors has driven an increased level of interest in single-wafer wet technology as chipmakers map their strategies for converting to 300-mm wafers. Single-wafer cleaning technology is of particular interest to fabs with high product mixes and smaller average lot sizes, although fabs with high-volume runners may have less-compelling operational reasons to switch to single-wafer systems.

In certain applications, momentum is building for the integration of wet cleans with dry-process tools. The integrated-cleaning strategy potentially provides better process control, shorter overall cycle time, improved overall equipment efficiency, and better COO than the existing stand-alone approach. However, the successful introduction of a high-productivity tool is contingent on the availability of high-reliability cleaning modules.

This trend, if successful, would pose a serious threat to stand-alone single-wafer cleaning tools, as was demonstrated by the rapid integration of the post-CMP clean in the late 1990s and the demise of the stand-alone single-wafer post-CMP scrubber market. Integrated post-CMP cleaning was an enabling technology driven by process requirements and the need for dry-in/dry-out wafers. More recently, backside and bevel cleans have been integrated with copper plating tools. Dielectric etch and dry-strip tools have emerged as prime candidates for the next wave of wet-cleaning integration, particularly for BEOL applications. Integrating BEOL applications and cleaning is much sought after and has several (primarily operational) advantages. However, successful integration of dry and wet cleans in a cluster tool requires significant tool and process development effort. Issues that must be addressed include interfacing (dry) vacuum and (wet) atmospheric chambers, dry/wet cluster throughput matching, and process module integration.

The integration of wet precleans with thermal processes would provide better control over the quality of the gate stack. But the stringent requirements for an atomically clean surface/interface, the success of existing batch tools in this area, and the lack of truly capable and practical FEOL single-wafer clean modules have made this difficult to achieve. For instance, the move to single- chemistry solutions for FEOL applications was not successfully adopted by the industry, despite the tremendous performance of systems using these new processes.

One important operational reason for the industry's reticence to embrace drastic chemistry changes in the wet space is that many larger fabs often run multiple process technology nodes and multiple products concurrently on common wet-cleaning tools. In general, any change to the process of record that could potentially alter process, device performance, and reliability may require extensive and expensive process and product requalification, which would be time-consuming and quite cumbersome. Given the financial performance and complexity of the industry as a whole, this is just the kind of extra expense and high-risk action that senior industry executives and fab managers avoid.

In next month's Reality Check, I will continue my examination of the wafer-cleaning space. I will focus on the emergence of dry techbnologies such as supercritical carbon dioxide, and then take a high-altitude view of the whole wafer-cleaning landscape today.

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit

Questions/comments about MICRO Magazine? E-mail us at cheynman@gmail.com.

© 2007 Tom Cheyney
All rights reserved.