Building Industry-University Research Partnerships: Corporate Perspective

John C. Tao

Air Products and Chemicals, Inc.

Air Products and Chemicals is a $5 billion company with almost 17,000 employees worldwide, operating in more than 30 countries. Our company supplies industrial gases and specialty chemicals directly to other industries, as opposed to consumers, and undertakes about $130 million of R&D, between 4 and 7 percent of which is done externally.

Why do external technology? Box 5.1 gives some of the broad reasons, focusing on the university partnership that Air Products has.

Certainly external technology development is something that wasn't done a whole lot 20 years ago. On the demand side, those of us in industry have to face faster development cycles. We're always driving to get the cost down—cheaper, faster, better. Moreover, everything is becoming more complex. Businesses and relationships are becoming more complex, and the problems are becoming more complex. We can hardly have all that expertise internally. At the same time, industry has reengineered, downsized, or right-sized, which makes it that much more difficult to do more with less.

BOX 5.1 Changing Environment for Research and Technology Development—Why Do External Technology?



• Faster development cycle time

• Global availability

• Cost and margin pressures

• Faster and higher growth of research knowledge

• Increasing complexity

• Small start-ups and spin-offs from universities

• Flat and shrinking technical expertise from reengineering

• Increasing talent pool


• Defense funding shifting to industrial uses

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 50
--> 5— Building Industry-University Research Partnerships: Corporate Perspective John C. Tao Air Products and Chemicals, Inc. Air Products and Chemicals is a $5 billion company with almost 17,000 employees worldwide, operating in more than 30 countries. Our company supplies industrial gases and specialty chemicals directly to other industries, as opposed to consumers, and undertakes about $130 million of R&D, between 4 and 7 percent of which is done externally. Why do external technology? Box 5.1 gives some of the broad reasons, focusing on the university partnership that Air Products has. Certainly external technology development is something that wasn't done a whole lot 20 years ago. On the demand side, those of us in industry have to face faster development cycles. We're always driving to get the cost down—cheaper, faster, better. Moreover, everything is becoming more complex. Businesses and relationships are becoming more complex, and the problems are becoming more complex. We can hardly have all that expertise internally. At the same time, industry has reengineered, downsized, or right-sized, which makes it that much more difficult to do more with less. BOX 5.1 Changing Environment for Research and Technology Development—Why Do External Technology? Demand Supply • Faster development cycle time • Global availability • Cost and margin pressures • Faster and higher growth of research knowledge • Increasing complexity • Small start-ups and spin-offs from universities • Flat and shrinking technical expertise from reengineering • Increasing talent pool   • Defense funding shifting to industrial uses

OCR for page 50
--> Figure 5.1 Technology transfer. In looking at technology external to the company that you are interested in, there are a number of things you can do (see Figure 5.1). You can start with doing nothing. You can just read about it. You can search around. You can attract some students to have them work or actually have them do the research at the university under contract. You can even acquire the technology or the company that has the technology. Thomas Manuel commented earlier on the global supply of technology. Universities in the United States are just one place outside of corporations where good science and engineering are done. There is always this increase and faster growth of research and knowledge. There are more and more spin-off companies coming out of the universities, and this talent pool is growing. Furthermore, there has been a shift in the federal government's spending from defense to more industrial-related R&D. At Air Products we do work at national laboratories. We also work with other companies, both large companies, in the case of Chemical Industry Environmental Technology Projects—a limited liability company that we formed with DuPont, Akzo Nobel—and with a lot of Small Business Innovation Research-type companies that bring a lot to the table from a feasibility standpoint in terms of new ideas. But in the remainder of this paper, I focus on the industry-university relationships and partnerships. Industry-University Relationships As shown in Box 5.2, the size and purpose of the research enterprise have grown substantially over the past several decades. Research during the pre-World War II era was mostly a search for new knowledge, with a small amount of government funding and very few research partnerships among the different sectors. During the post-Cold War era, however, federal agencies funded most of the research focused on defense and health. The tot al amount of research conducted in the United States increased substantially, and significant discoveries were made in electronics and biomedical research. At the same time, industrial research grew substantially as well. A few of the "high-tech" industries were born. In addition, the interactions between academia and industry grew, especially at the basic research level. Figure 5.2 shows the sources of funding for academic basic research from 1993 to 1997. It makes

OCR for page 50
--> BOX 5.2 The Changing Research Enterprise Research in the pre-World War II Era • Search for new knowledge • No substantial government funding • Few industrial science research labs • Little or no interaction between academic research and industry • No partnerships or alliances Research in the post-Cold War Era • Search for new knowledge and defense, economic, and health needs • Government is primary funder • Scientific research is large enterprise • Explosion of progress (especially electronics, medicine) • Growth in industrial research • Growth of technology-based industries • Increasing interactions between academia and industry and true partnerships Figure 5.2 Academic basic research funding sources, 1993 to 1997. the point that the federal government is still the major source of the university research money, and industry provides, at best, 10 percent. Figure 5.3 gives the breakdown of industry funding for academic research ranging from basic to applied to development. It shows that industry does not go to universities to conduct work downstream of research (i.e., development work is not the university's strength). When we have a pilot scale or when we do further applied work, we do that primarily in-house.

OCR for page 50
--> Figure 5.3 Industrial funding of academic research, 1993 to 1997. Survey from the Council of Chemical Research Let me shift gears and talk about the survey from the Council for Chemical Research (CCR) done in July 1997. This was a survey sent to 450 members of the CCR, mostly chairs of chemistry and chemical engineering in the United States, but we also included people overseas in Europe, Japan, and Taiwan. We received almost 200 responses (shown in Figure 5.4): about half of the responses were from the United States and the other half from overseas. The open squares in Figure 5.4 represent the current situation, and the open circles show what the respondents said there should be in the future. The higher the number, the higher the emphasis, as can be seen when collaboration moves from a scale of about 4.5 all the way to almost 9. The respondents would like to see participation from industry—which is very low right now, close to 2.5—go above 8. Figure 5.4 CCR survey, industrial participation. Respondents judged the current state of industrial participation in education programs to have the lowest emphasis and the greatest desire for improvement.

OCR for page 50
--> Figure 5.5 Graduate education. Substantial changes are desired in graduate education, with most countries' respondents believing that theory and fundamentals are overemphasized. The survey also looked at graduate education and how it prepared students for the work place (Figure 5.5). It looked at training in theory versus more practical and applied, and the arrow goes from right to left only for theory and fundamentals. Everything else shows significant moves to a high emphasis on teamwork, a little less than 5 all the way up to 9; industrial awareness is a little over 3 to close to 9. What Each Partner Brings to the Table As shown in Box 5.3, the universities obviously provide a lot of research and good work to complement industrial research, and the costs are lower. Industry seeks access to professors as consultants. We look at certain technology rights. We look for new ideas, for opportunities. Of course, we recruit the graduates. And we can look at the leveraging of our funding with the funding from the federal sector. BOX 5.3 What Universities Offer to Industry • Fundamental research and enhancement of own research • Additional high-quality scientific research • Costs—lower overheads • Access to consultants • Patent rights • New ideas and business opportunities • Recruitment of graduates • Leveraging of federal funding

OCR for page 50
--> BOX 5.4 What Industry Offers Universities • Access to world-class and important problems • Access to other activities of the innovation cycle (development, engineering, manufacturing, etc.) • Access to a different culture of thinking • Market awareness • Enriching teaching programs • Job opportunities • Research funding at the margin Industry, on the other hand, brings relevance (see Box 5.4). It brings some real live problems. University researchers can gain some access to the activities beyond the research they do. They can see how a product finally gets out the door downstream of the innovation cycle. They get access to an entirely different type of culture and thinking. They learn about the marketplace. All of these perspectives should help the universities in their teaching program, and, of course, the students gain industrial job opportunities. Also, the funding is approximately 10 percent when compared with federal funding. Technology Transfer There are a number of ways to transfer the technology from a university to industry (see Box 5.5). Much of the research is published. You can hire someone to tell you exactly what they know. You can hire students. You can give them contract research work. You can build an alliance, license the work, buy the spin-off company, or do some work with the spin-off company. There are different degrees of usefulness and different degrees of timing and value in terms of how much you gain. BOX 5.5 Technology Transfer from Universities • Publications   • Consulting   • Hiring graduates   • Contract research   • Research alliances   • Research centers and consortia   • Licensing   • Spin-off companies  

OCR for page 50
--> BOX 5.6 Industry-University Interaction   IP Value Timeliness Gifts Low Low Consultancy High Medium-High Consortia Low-Medium Low Contracts High High Hiring graduates Medium-High Medium As can be seen in Box 5.6 the intellectual property (IP) from a gift to a university is fairly low, and there is no control on the speed of results, whereas with a research contract or work in a consortium, you achieve different levels of timing and a different value. Partnership Models I now turn to some examples of partnerships, i.e., the different types of relationships possible in industry-university research (see Box 5.7). Professor Wakeham talked about the Air Products relationship with Imperial College. CCR has a partnership with Pennsylvania State University, and we are about to launch one with Georgia Institute of Technology. The Center of Interfacial Engineering at the University of Minnesota is an example of the multicompany-single university collaboration model. Perhaps the Massachusetts Institute of Technology (MIT) media BOX 5.7 Industry-University Research Partnership Models I. Single Company-Multiple Universities   — Hoeschst Celanese with Rutgers University, North Carolina State University, University of North Carolina   — Air Products and Chemicals, Inc. with Imperial College, Pennsylvania State University, Georgia Institute of Technology II. Multiple Company-Single University   — MIT Media Laboratory with Hewlett Packard, Phillips, Digital Equipment Corporation, and others   — Imperial with Air Products and Chemicals, Inc., British Petroleum, Rhone Polenc, Unilever, Fujitsu, Microelectronic and Computer Technology Corporation III. Multiple Companies and Multiple Universities   — Pennsylvania Infrastructure Technology Alliance with Lehigh University and Carnegie Mellon University

OCR for page 50
--> lab is a better known example. In the chemical industry, the Separation and Research Program at the University of Texas at Austin is well established. There are four multicompany-multiuniversity cases as well. The one that we are closest with is the Pennsylvania Infrastructure Technology Alliance. This involves Commonwealth of Pennsylvania money, and it goes to two former National Science Foundation (NSF) centers, one at Lehigh University and one at Carnegie Mellon University, and there are over two dozen industrial participants. Issues and Problems With any two partners, there are always problems and issues to deal with. There are differences in goals and values (see Box 5.8). Those of us in industry deal with the marketplace; we focus on return on investment. We are very cost conscious. We go out of business if we are not cost competitive. We are for-profit; we have to provide a return to our shareholders, and we always have to keep the results of innovation as a competitive advantage. That is how we make money. And, of course, timing is everything. On the university side, first on the list of goals is advancement of knowledge. Professors like to set their own objectives. They publish. They need to educate students. And the recognition of their work in the academic institution is very important. Their fame equates to power, at least at the researcher level. Timing is not that important. When it comes to innovation, again there are some differences (see Box 5.9). At the university, BOX 5.8 Differences in Goals and Values Industry Universities • Market driven • Advancement of knowledge • Emphasis on return on investment • Academic freedom • Cost conscious • Publication of results • Profit oriented • Education of students • Keeps results and competitive advantage • Recognition of research and institution • Timing is everything • Relaxed time frame and milestones BOX 5.9 Issues in Partnerships: Innovation Universities Industry • Driven by curiosity • Results driven • Conceptual breakthroughs • Maintenance of core technologies • Discontinuous technology • Continuous and incremental improvements • Displace technology • Capture markets • New markets • Market share of existing markets

OCR for page 50
--> innovation is driven by curiosity; researchers tend to make breakthroughs and their innovations often lead to discontinuous technologies or new technologies to replace existing ones. Innovation in industry is more results driven, with the focus on core technologies and incremental improvements. The improvements offer new markets and increase the market share of existing markets. Publication in an alliance or a partnership is usually not a problem (see Figure 5.6). In most cases we agree on a time period to review the manuscript. We want to have enough time to file a patent and obtain approval or at least some rights to delete any sensitive information. And in a lot of cases there is confidential information involved from the company, and there are typical clauses in any confidentiality statement. The Association of University Technology Managers (AUTM) does an annual survey of its members. Results of the latest survey as shown in Box 5.10 indicate that total university-sponsored research Figure 5.6 Common publications-related issues. BOX 5.10 AUTM Survey (1997) • Total university-sponsored research—$19.8 billion (+6%)   $2.2 billion (+15.8%) from industry • Patents   —Disclosure: 11,303 (+10%)   —Filed: 4,267 (+40%)   —Issued: 2,645 (+24%) • License income   —$483 million from U.S. universities   —79% from life sciences   —74% from top 15 of 132 universities

OCR for page 50
--> BOX 5.11 Common Intellectual Property Problems • Patent ownership and assignment • Royalty-free nonexclusive • Option period for royalty-bearing exclusive • First right of refusal • Up-front fees and annual minimums • Patenting costs and maintenance fees • Right to use all other know-how • Third-party funding and their rights (especially government) • Use of confidential and proprietary information is approximately $20 billion, up 6 percent. The percentage from industry is up even more (16 percent). Patents are increasing also, both in filing (+40 percent) and actual numbers issued (+24 percent). For licensing income, approximately 80 percent is from the life sciences, which is really biomedical and agrichemical. Close to three-quarters of the licensing profits are from the top 15 universities out of the 132 that responded. In fact, a new report done for NSF by Professor Yong Lee of Iowa State University, called ''University-Industry Collaboration on Technology Innovation," includes a survey of approximately 700 people, of whom over 400 are faculty members and about 150 are from industry.1 A question about what faculty members look for in a collaboration was, "What are the personal reasons for getting involved with industrial-sponsored projects?" The answer "getting funds for the graduate students and lab equipment" was close to 70 percent. "Gaining insights into their own research" also ranged close to 70 percent. "Testing the practical application of their own theory in research" scored 65 percent, and 91 percent ranked "to look for funds from industry" as the least important reason for collaboration. Intellectual Property Some common IP problems are listed in Box 5.11. First on the list of IP issues is "Who gets to own what?" Does the sponsoring party have to pay for it after the patents are issued? Is there an option period to get a license? Does the sponsor get the right of first refusal to an exclusive license? Do they have to pay any up-front fees? Who pays for the patenting costs and the maintenance fees? That may sound like a trivial matter, but it is not when you consider multiple applications and foreign patents. For some foreign patents, just the maintenance fee alone could be $15,000 for a few years. In regard to the rights to use the other know-how that came out of this work, what are the rights of the third party? And who gets to use what information afterwards? To answer some of these questions, we need to look at the possibilities of both sides. What are the possibilities for the universities if the company owns the IP? (see Box 5.12). There are three possibilities 1   Young S. Lee, "University-Industry Collaboration on Technological Innovation: Views from the Ivory Tower," Policy Studies Journal, 26(1): 69–84, 1998.

OCR for page 50
--> BOX 5.12 Company Owns Intellectual Property: Possibilities for University Rights Benefits • No rights without permission of company • No reward for commercial exploitation • Equal rights and free use for own purpose; negotiated rights for third parties • Subsequently negotiated reward • University can use for research; no rights for third parties • Preagreed upon reward on the "rights" and three possibilities on the "benefits." The university cannot get any rights without the permission of the company, can have equal rights, or can get some kind of limited rights to use it for research, but cannot do any licensing. And there are a range of benefits. There is no reward, some reward, or something that can be negotiated afterward or beforehand. When the university owns the IP, the company then has to pay for an exclusive license, pay for an nonexclusive license, or get a royalty-free, nonexclusive license (see Box 5.13). And again, you go through the same benefits on the sharing of royalty and licensing income from the third parties. Perhaps Figure 5.7 best summarizes how I look at the situation. It is really not black and white. I gave the two extreme cases above. One party owns the IP, and you have to cut a deal with the other party. The model to look at, from the type of partnership to the type of contractual relationship, asks the questions: How much is the funding, is it 100 percent, is it partial, how much of it is basic, how much of it is applied, and how much input and background knowledge does the company bring? Is there a model in which they just sign the check and do not get too involved until the results are out? In Europe, some of the universities will charge a fee. This gets to the issue of overhead. They actually make money. A decision in that case was, "Okay. We aren't going to argue with you on intellectual property. Who knows whether anything will come out of this research? Who knows whether there's going to be an invention? You pay us a fee, right up front, versus uncertainty later, and you get to own the IP. We know exactly what that' s worth." Very few U.S. universities ever think about that. BOX 5.13 University Owns Intellectual Property: Possibilities for Company Rights Benefits • Company pays for exclusive • Company receives preagreed upon reward from third-party licenses • Company pays for nonexclusive • Company receives subsequently negotiated reward • Company receives free nonexclusive • Company receives nothing from third parties

OCR for page 50
--> Figure 5.7 Spectrum of arrangements for university-industry partnerships. IPR, intellectual property rights.

OCR for page 50
--> BOX 5.14 What Makes a Good University-Industry Relationship? • Good historical relationship • Complementing and overlapping strengths in core technologies • Clear goals and roles • Good teamwork • Good communication • Good science • Mutually benefiting (share success and failure) • Intellectual property ownership agreed to BOX 5.15 The Bottom Line • Dollars in research contracts • Facilities, equipment, and materials used or provided • Background technologies used • Expertise provided by both • Time committed by team • Timely decisions • Recognition of people and organization • Meet each other's expectations The contract research with full overhead is another case. When you have cost sharing from a third party or, in most cases, with a federal agency, then Bayh-Dole kicks in. The university owns the IP rights, and you have to negotiate whether you get a nonexclusive, royalty-free license or if you have to pay for that license whether it is exclusive or not. Another trend that I believe will be increasing is to reverse the IP transfer. Increasingly companies will be giving away IP to the university for the university to exploit. And the company will get a tax benefit. We are going to see more and more of this. What Makes a Good Partnership? First, it is important to have a good historical relationship (see Box 5.14). Professor Wakeham talked about the Imperial College partnership with Air Products in which a former CCR vice president of engineering, an Imperial College alumnus, brought us together. CCR now has over 500 Pennsylvania State University graduates on campus. These alumni help. It also helps to have overlapping strength in core technologies, the kind of work industry does and the kind of research that professors are interested in. It is important to have very clear goals, good

OCR for page 50
--> teamwork, and good communication. You must always do good science. At the end of the day, that is the only result you get. You have to share the benefits, and you have to work out the IP issues ahead of time. What I learned is that the more you put in, the more you get out (Box 5.15). If all you do is write a check and wait for the results to come a year later, that is not going to work well. You can measure the amount of dollars you put in at an institute; you can look at the amount of facilities, materials, and the equipment you have built; and you can look at how much background work you put in, how much expertise the two parties put in, the commitment from the teams, and whether decisions are made quickly. Do you recognize the people when successes are realized, and, most importantly, do you meet the expectations of the parties going into the partnership? Discussion Henry Kohlbrand, Dow Chemical Company: Would you comment on the direction that your university relationships are going? Do you see the number of institutions that you are actively working with decreasing in favor of a few strategic relationships, or do you see a blend of strategic relationships and one-of-a-kind agreements? John Tao: It's the latter. We never ask our researchers to only work at Imperial College or Penn State or in the future at the Georgia Institute of Technology. Traditionally we have worked with about 40 different institutions, including those overseas. And we see those partnerships continue, but we built a few alliances that we are also committed to. And as Professor Wakeham said, these are long-term relationships, but they are not everything. David Sehetter, University of California, Irvine: You mentioned that 4 to 7 percent of your budget was for extramural activity. Do you see that going up in the future as was projected earlier? John Tao: Yes, for the reasons I stated: the drivers on the demand side as well as the supply side. We cannot possibly do everything in-house. Ashok Dhingra, Dupont: I found some things in your presentation that are common to what we are doing at Dupont. But in your case I can see that it was a prior relationship between your vice president of engineering who had some connections, which helped start the project. Do you give a certain amount of money to a university and then build projects based on that funding? In the case of Dupont and many other companies, the corporate funding is not there anymore as it was in the past. How do you get the businesses to participate in funding these programs? John Tao: A typical project or proposal coming in from a university has a staged approach. In the first stage, it may be very exploratory in which the feasibility has not even been established. Then it's fair to ask for corporate money to help get the project kick-started. Very quickly, once we move beyond feasibility, let's say, to the second year, we'll ask the business units to kick in at least half the money, and hopefully beyond that they take over. Ashok Dhingra: Do you see continued support in the areas of restructuring and reengineering and the changes in management? In other words, you start a program and then you want business to participate in it. Do you see this happening?

OCR for page 50
--> John Tao: Yes, and we have not had any problems. Fritz Kokesh, Massachusetts Institute of Technology: You mentioned that you have had some experience in which European universities will allow ownership if the price is greater than their cost. Could you expand on that? Is that across-the-board markup or is that on a per patent basis? How does that work? John Tao: In that one case at a university in Holland, it was 20 percent extra overhead, but it' s not per invention. Christopher Hill, George Mason University: Earlier I mentioned that the tax status of the university could be a problem for that scheme in the United States. You said that you had been able to do it here. How do you get the work done? Janet Osteryoung, National Science Foundation: What is this tax problem? Christopher Hill: Universities are nonprofit institutions. However, income from sports events, bookstores, and cafeterias may be taxed as unrelated business income. It has been said that if universities start to act as a contract research shop, particularly if they seek to earn a profit, it may put the tax-exempt status at some risk across the board. At least universities have been cautioned about this. David Schetter: There is the Tax Reform Act as well, which means that universities are actually legally precluded from prevaluing technology that has not yet been invented if the research was conducted in facilities built with tax-exempt bonds. It's the law. So you would love to do that sometimes, but it depends on whether the research is actually conducted, and it will challenge the tax-exempt status of a university. John Tao: I am not sure that I'm the best person to answer that question, but I know some people have done it with a separate subsidiary company. Christopher Hill: Many of us have created various subsidiaries to get around this problem. If you create a profit-making subsidiary, then you can do those things. John Tao: The American Chemical Society (ACS) has a very profitable subsidiary, Chem Abstracts, and they are a 501(c) 3 organization. Frank Feher, University of California, Irvine: When you do research at a university, the research is usually done by graduate students or postdoctorates who have a finite time at the universities. So you start to develop what could be a pipeline problem if you are going to be doing your research with essentially what are temporary workers. To maintain a flow of workers into the system, you must have prospects, good economic prospects, for workers at the other end. Would you care to comment on your model for research in which 80 percent of the research conducted at a university has some connection to industry? Are there going to be positions available for the people doing the work as temporary researchers, and can we keep the process going? John Tao: Well, the professor is still there.

OCR for page 50
--> Thomas Manuel, Council for Chemical Research, Inc.: I heard the question differently. The issue is if research is funded at universities at an increasing rate and if the workers are people who are getting their degrees so that they can come out of the university and do research somewhere else, then the research that industry was going to do in their own labs is now done in the university. If so, do you have a material balance? That's the old question of supply and demand of technical people. I have two observations: First, the postdoctorate pool is the surge tank on this in the short term. And second, if I recall, when Ned Heindel was ACS president, ACS did a big study of all this and found that everyone got a job somehow. What happens is that people who are in industry turn into salesmen. If they are not very successful, they become managers. And then finally they retire and work for nonprofits. So there are other avenues out of this great tank of people. In theory, it's a problem, and someday it may happen, but you have to ratchet up the ratios before this becomes very serious. Andrew Kaldor, Exxon: Do you have any examples of the type of reductions that you are able to achieve through university interactions? John Tao: I can't give you any specifics. Joseph Gordon, IBM: You implied that you had a sort of sliding scale for ownership of intellectual property and the amount that you are willing to pay for it. Do you actually have a range? And what fraction of your extramural research is in the form of a gift? What fraction is through contracts in which you own everything? John Tao: Let me put it this way. We have very few in gift form, and we have more on the contract side of it. Todd La Porte, University of California, Berkeley: You have gone through quite a long experience yourself on this. Could you talk about the surprises you've had in this experience? John Tao: Some of the problems are people issues. It has nothing to do with good science or good research. We had to deal with one case in which the communications broke down, and we finally realized what happened: There was going to be a publication, and it was a battle about whose name came first. But very rarely do we have difficult problems that cannot be solved once you get beyond the people problems. Todd La Porte: I didn't mean just negative surprises. John Tao: Yes, we have had good surprises. We had a project in which we were looking at specific interactions of gases in the electronic etching process. There was an unexpected result in that the addition of some CO2 gave a much better life to the tools. And there was an invention that no one expected. That is translating into dollars for us right now.