There is nothing like a little hands-on experience to help someone understand a particular technology better and more fully. Last week I spent two days in Indianapolis learning about how cleaning formulations are made, at the invitation of Steve Dwyer, Business Director Electronic Materials Dynaloy LLC, A subsidiary of Eastman Chemical Company. The company is in the business of developing solvent formulations for the removal of photoresists and polymer residues left behind during such processes as copper pillar bumping, TSV DRIE etch and temporary bond/debond for 2.5D and 3D IC processes.
Steve and the rest of his team, including Kim Pollard, Technology Manager, Diane Scheele, Market Development Manager, and Maryellen Mantyla, Marketing Communications Manager, Additives and Functional Products for Eastman, brought me up to speed on Dynaloy’s approach to clean, and the company’s recent activities in the 3D IC space. While an in-depth feature story is in the works and is forthcoming, I wanted to share some of my initial take-aways and lessons learned from this visit.
1. It’s not about clean; it’s about surface preparation. In semiconductor manufacturing, the clean step in any process flow is often discussed as an afterthought, rather than flagging it as a step all its own. It requires time and the right chemistry to not only remove the residue in question (polymer from TSV etch, or photoresist from copper pillar bumps), but to do it without compromising the underlying metallization or polyimide layers that need to remain in tact, all the while keeping the next process step clearly in mind.
2. It’s more complex than you think. Some cleaning processes are simple, requiring only the photoresist strip solvent provided by the photoresist manufacturer. That works in removing the original formulation of the photoresist. But what happens when the properties of the original formulation change during processing? And what about the cleaning processes themselves? Is the first wafer that goes into an immersion bath experiencing the same solvent chemistries as the last? That’s where the term “bath life” comes in. This refers to how many wafers come out clean from the same solvent bath; indicating when it needs to be changed. I like to compare it to washing dishes in a sink of soapy water. Considering what could redeposit on the dishes, I’m careful to wash glassware first, then silver, plates and pots and pans last. It may be necessary to empty the sink and refill it to get those stubborn dishes really clean.
3. All cleaning formulations are not created equal. When considering the cost of cleaning, it’s important to understand the difference between cost per clean rather than cost per gallon of a particular formulation. I compare it to how many loads of wash can you get from one detergent compared with another? Why do some formulations cost more than others for what appears to be the same quantity of product? Likely because more consideration went into the development process. Attention must be paid to such details as what is the problem? What are you removing from it? What are the limitations? What is the next step in the process? Add to that – where is it going? In a worldwide industry, travel-ability of a formulation is crucial. Can it stand up to elevated or freezing temperatures? The responses to these questions vary not only from customer to customer, but also situation to situation. Ultimately, the answers require customized solutions.
4. Safety first! As awareness grows throughout the world about the hazardous properties of some chemistries despite their efficacy, the semiconductor industry is taking measures to eliminate or limit their use. Chemical companies are charged with finding alternatives to these products. Dynaloy has been following a robust roadmap for eliminating the worst offenders (e-series glycol ethers, TMAH, and DMSO) from their products by 2014.
These points provide the foundation of Dynaloy’s approach to clean. Stay tuned to a more in-depth look at how Dynaloy achieves this. ~ F.v.T.