InsightFab
Knowledge Base/Process Window: DOE Applications in Semiconductors
Semiconductor Process6 min read

Process Window: DOE Applications in Semiconductors

This practical article recounts a painful experience where a CPK report of only 1.08 caused immediate silence. It clearly explains what a "Process Window" is, its critical link to process quality and yield, and uses a dart-throwing analogy to demystify process robustness and how CPK assesses process stability. Essential reading for anyone interested in process optimization or yield improvement.

That day, when the CPK report came out, the whole room fell silent for three seconds

I still remember years ago, when a new process was run on our equipment, everyone was full of confidence. The next day, as soon as the CPK report came out, the whole room instantly fell silent; you could only hear the director take a deep breath. The data showed the CPK was only 1.08. Yes, you heard that right, 1.08. This number basically told us: although this batch wasn't on the verge of scrap, its quality was already walking a tightrope. The director's face turned ashen and he immediately said, "This Process Window definitely has problems, go back and do DOE!" At that moment, I knew that in the days ahead, I would have to wrestle with this "Process Window" thing.

Where did the problem lie?

To put it plainly, the Process Window is simply how much "tolerance" your process has. Imagine you're throwing darts; the bullseye is your target value, and the Process Window is the area where you can score. The more stable your process, the larger this scoring area will be, so even if your hand shakes occasionally, you can still land within the safe range. A CPK of only 1.08 means our process is like aiming at a very small target, where a slight deviation can easily lead to a miss.

So the key is, we need to find which parameters have the greatest impact, and then find an "optimal operating point" so that your process, under various variables, can stably remain within specifications. This isn't just about adjusting parameters to the target value; more importantly, it's about being "noise-resistant."

How is it actually done?

We typically use DOE (Design of Experiments) to find this Process Window. Frankly speaking, DOE is a systematic way to "mess around" with your process parameters and then observe the results.

For example, if we suspect that etch depth is affected by "RF Power" and "Gas Flow," we would do this:

  1. Select factors and levels: RF Power at three levels: 100W, 120W, 140W; Gas Flow at three levels: 50sccm, 60sccm, 70sccm.
  2. Design the experiment: We won't run through all combinations (that's too exhausting); we usually use statistical software (like JMP or Minitab) to design the experiment. For example, set up a 3x3 full factorial experiment, totaling 9 runs.
  3. Run experiments, collect data: Run samples for each set of parameters, then measure the etch depth.
  4. Analyze results: Use software to analyze which parameters have the greatest impact on etch depth, whether there are interaction effects, and then generate a "response surface plot." This plot is like a mountain; the peak is your optimal operating point, and the slope is your Process Window.

With this data, you'll know that at RF Power 125W and Gas Flow 65sccm, the etch depth is most stable, and even if RF Power fluctuates by ±5W, the etch depth will still be within our required range of 2000Å ± 100Å. At this point, your CPK value has the chance to increase from 1.08 to 1.33 or higher. Just imagine, reducing DPMO from 6210 directly to 2700—how much difference that makes in yield!

The Most Common Pitfalls

What are the most common pitfalls? It's setting "too few parameters" or "too narrow levels." Some newcomers might think, since my target is 120W, I'll only test 119W, 120W, 121W. What's the result? The resulting data shows no trend at all because you haven't spread out the variables, so you can't see its true range of influence. To put it bluntly, it's "not messing around enough." You need to be a bit braver, widen the range of parameters a bit, to test where its true limits lie. Of course, this doesn't mean adjusting randomly to the point of scrap, but at least make sure the results are distinctive.

There's another pitfall, which is only looking at single factors and ignoring "interaction effects." Sometimes adjusting RF Power alone doesn't do much, and adjusting Gas Flow alone is also fine, but when both vary simultaneously, the results can be vastly different. In DOE, this will indicate significant interaction effects. At this point, you should look for the "saddle point" or "valley" on the response surface plot, rather than just the peak.

One Thing You Can Do Today

Open your process parameter sheet and circle the two factors you believe have the most influence.

Want to try it yourself?

Every tool mentioned in this article is available on InsightFab — just upload a CSV to analyze.

Go to Tools

More in this category

Yield Engineering: From Die Yield to Wafer Yield

Ever wondered about wafer fab yield numbers but felt lost? This practical article introduces you to the world of wafer yield, using a vivid "pineapple field" analogy to clarify die yield, wafer yield, and what happens to defective chips. You'll learn how to pinpoint root causes from data, determining whether issues stem from equipment or processes, making crucial semiconductor knowledge accessible.

Yield Loss Classification: Random Defects vs. Systematic Defects

This article addresses the common factory challenge of unstable yield by differentiating between 'random defects' and 'systematic defects.' It provides a clear guide on how to identify the root cause of yield fluctuations, enabling more targeted problem-solving and significantly reducing time and cost spent on investigations.

Process Capability Assessment for Thin Film Deposition (PVD/CVD)

This article recounts a challenging experience where a newly promoted senior engineer faced intense pressure from management over a thin film deposition report showing a CPK of 1.08. It then provides an accessible introduction to thin film deposition, explaining common techniques like PVD and CVD. Readers will learn the critical importance of evaluating thin film process capability and the numerous intricacies involved in manufacturing these fundamental layers in chip production.