Education On Microfabrication In Latin America And The ... - IEEE Xplore

IEEE TRANSACTIONS ON EDUCATION, VOL. 41, NO. 4, NOVEMBER 1998

253

Education on Microfabrication in Latin America and The Microelectronics Workshop at UNICAMP Jacobus W. Swart, Senior Member, IEEE

Abstract—This paper discusses the importance and the need to improve education on microfabrication technologies in the region of Latin America. A short historical overview of microfabrication and its current activities in the region are presented. Since the beginning of the 1960’s, industrial and academic activities in microelectronics have taken place in the region. However, the technological gap to the developed countries has been growing continuously. To change this scenario, a strong effort in education is needed. The education in microfabrication in the regular engineering courses is very limited and needs to be changed. A specific workshop at the State University of Campinas—UNICAMP, called “Oficina de Microeletrônica,” was set up in order to reduce the gap in education on microfabrication technologies. This workshop is discussed in detail. Index Terms— Integrated circuits, microelectronics, microfabrication, workshop.

I. INTRODUCTION

M

ICROFABRICATION technologies have seen an extraordinary evolution during the last 40 years, with its main application on microelectronic devices and circuits. Other applications, such as optoelectronics, photonics, sensors, actuators, and micromechanical devices and circuits, have also benefited greatly from this evolution. As a consequence, society has gone through a strong economical and social revolution. A democratic and idealistic view is to see the knowledge related to this technology and its benefits to be shared by the whole human society around the world. With this in mind, it is most important that this knowledge be spread through efficient educational activities, specially in regions that are technologically less developed. The objective of this paper is to contribute to a discussion on this goal, analyzing the situation and activities in Latin America, with a deeper focus on Brazil. First, a historical overview of microfabrication activities and a description about the industrial and educational environment in the region are presented in Section II. In Section III, a discussion about the applications and benefits of education in microfabrication techniques is presented. Following, in Sections IV and V, a description and discussion of education on microfabrication at regular undergraduate and graduate engineering courses and at a specific workshop at the State Manuscript received December 9, 1996; revised July 27, 1998. This work was supported by the Microelectronics Workshop at UNICAMP, the Brazilian Agencies FINEP, CNPq, CAPES and fthe Ibero-American programs Cyted and Ibership. The author is with DSIF/FEEC and CCS UNICAMP, Campinas, SP 13.083970, Brazil. Publisher Item Identifier S 0018-9359(98)08408-8.

Fig. 1. Photograph of a 2-kbit ROM memory developed at the University of São Paulo, Brazil, in the mid 1970’s.

University of Campinas, UNICAMP, are presented. Final conclusions are drawn in Section VI. II. HISTORICAL BACKGROUND AND REGIONAL ENVIRONMENT A. Brazil Since the beginning of the 1960’s, academic and industrial activities on semiconductor devices have taken place in Brazil. In 1962, a transistor plant was started in So Paulo by the company Philco. This was not only a device packaging plant but a full wafer processing facility. Later in the sixties, a Microelectronics Laboratory was set up at the University of São Paulo. This laboratory developed a MOS technology, among other activities, leading to a successful fabrication of a 2-kbit ROM in 1976 [1]. This development, although not the state of the art, was not far behind it at its time. Also laser research and fabrication, at UNICAMP and later at Telebrás, Brazil was in a good position at that time [2]. Figs. 1 and 2 illustrate some of the circuits developed at that time. Fig. 1 shows a picture of the 2-kbit ROM mentioned above and Fig. 2 shows a picture of a 8-bit three-phase BCCD, buried-channel charge couple device, with input and output stages, developed at the same laboratory. A home-made ion-implanter system was developed (by de Souza) and used for the channel doping of the BCCD (by the author).

0018–9359/98$10.00  1998 IEEE

254


international multi-project chip programs like CMP from France, Europractice from the European Community, IBERCHIP from Ibero-American countries or MOSIS from USA. Actually, it is observed that more institutions move from microfabrication activities to circuit design or software work. These activities are more “friendly” tasks, considering they are less dependent on a laborious and costly infrastructure. However, it is important to keep in mind that real life will always depend on hardware and that it is very helpful for an IC designer to have a good knowledge and view about the microstructure and the fabrication process he or she is using. B. Other Latin America Countries

Fig. 2. Photograph of an 8-bit three-phase BCCD developed at the University of São Paulo, Brazil, during the end of the 1970’s.

Since that time, microelectronics activities have been continued, but just at a stand by level, both at academic level as well as at industrial level. Different plants were set up by local as well by international companies. However, gradually and especially, after the beginning of the nineties, these companies have closed their manufacturing facilities. Just recently (mid 1996), the only IC wafer plant (5- m linear bipolar technology) closed its wafer processing line, keeping only its device mounting activity. One successful exception in this scenario is a small company named AEGIS, focused on the fabrication of power diodes and thyristors. This company was started by two former faculty members, one of the University of São Paulo and another from UNICAMP, and it is an indication that even in the competitive global economy, there is a place for small local companies in niche applications. This reduced activity in microelectronics has increased the technological gap to the first world countries. Despite this, an increase in human resources can be observed during this period. The number of experts in the area has increased, many of them with Ph.D. or postdoctoral work done abroad. There are about 100 universities and colleges with electrical engineering courses in Brazil. Among these, there are two universities that are reasonably set up for teaching and performing research in microfabrication activities. These are UNICAMP (State University of Campinas) in Campinas and USP (University of São Paulo) in São Paulo City, both in the State of São Paulo. Other universities, with some limited facilities are: UFRGS (Federal University of Rio Grande do Sul) in Porto Alegre, UFPe (Federal University of Pernambuco) in Recife. This number of academic laboratories is very modest when compared to the 30–100 estimated in the USA [3]. Besides this small number, it is observed that in general, like most of the laboratories in the USA [3], they are relatively underutilized. The activity of custom integrated circuit design on the other hand, is more diffused across the country (also over Latin America). Two institutions, FAPESP, a research funding agency of the State of São Paulo [4] and CTI, a federal research institute [5], sponsor the fabrication of prototype IC’s through

Mexico is the country with most industrial activities in Microfabrication, possibly due to its proximity to the USA. In the other countries of this group, there are no semiconductor plants. At the academic level, there are some examples: 1) the University of Santiago, Chile, had a microelectronics laboratory for BJT and MOS fabrication. This laboratory was closed and remains so since 1980. 2) in Caracas, Venezuela, the Instituto Venezolano de Investigaciones Cient´ıficas, another microelectronics laboratory was set up during the 1970’s and remained active up to 1982, when it was also closed. 3) Havana, Cuba, is another place that houses a facility for fabrication of semiconductor devices and small IC’s. 4) University of Mar del Plata in Argentina has a very simple laboratory that can process some MOS structures. 5) Mexico is without doubt, the country with the most academic microelectronic laboratories among these countries. We can mention facilities at three universities: Instituto Politécnico Nacional, in México City, Instituto ´ Nacional de Astrof´ısica, Optica y Eletrónica, in Tonantzintla and Universidad Autónoma de San Luis Potos´ı, in San Luis Potos´ı. We are not aware of the existence of other microelectronics laboratories in other Latin America countries. This information shows that the academic and industrial activities in microfabrication are very limited in Latin America. This fact has a severe effect on the general thinking of the decision-making community in these countries, including government education and research agencies, industrial policy promoters, and even academic faculty members. This thinking includes some of the following general thoughts. • Microelectronics is exclusively a complex and very expensive activity. • Microelectronics is essentially an activity for first-world countries. • Microelectronics means mainly production of microprocessor, memory and other digital circuits. Unfortunately, these ideas have been guiding investment and hiring policies by institutions and government agencies. These ideas have to be changed in order to develop in the region, high-tech fields involving microfabrication. These changes can only be achieved by appropriate educational programs. Political and economic issues, which are beyond the scope

SWART: EDUCATION ON MICROFABRICATION IN LATIN AMERICA

of this paper, are other aspects to be considered for these technological developments. However, a change in thinking will be a good start to encourage the economical and political forces to persuade the necessary technological development. III. APPLICATIONS AND BENEFITS OF EDUCATION IN MICROFABRICATION TECHNIQUES Although most of the development and production of microelectronics have been performed in first world and/or far-east countries, actions have to be taken to spread these technologies and its amazing benefits to other parts of the world, like Latin America, among others. As an important motivation for that, we can mention the following. • Electronics is becoming the largest economical market. Each country must try to share and benefit from this huge market. • Electronics is penetrating most human activities. • Microfabrication techniques can be applied to other devices besides electron devices. Microfabrication technology has evolved very rapidly during the last 40 years, leading to the present phase of ULSI integrated circuits, approaching the scale of a billion transistors per chip [6]. Besides this fantastic evolution, microfabrication technologies are and can be used for an increasing number of other devices or applications. These include optoelectronic devices, photonic devices, magnetic devices, sensors, and actuators of all kinds and applications, micromechanical devices [7] and also devices or micro-structures for biotechnology applications and/or experiments [8]. This broad and increasing spectrum of applications of microfabrication technologies shows the importance of offering education programs in this field. A larger population with knowledge in this field will result in the following benefits: 1) a change in thinking, changing the misleading views mentioned before, which means convincing people that: a.

b. c. d. e.

microfabrication is not magic but in fact, is feasible even in countries like those from Latin America; microfabrication is not needed only for the large markets of microprocessors and memories; a number of niche markets exist; new types of microstructure devices can be made leading to new markets; microfabrication is a very expensive activity when dealing with ULSI level IC’s but is affordable for a lot of other components and/or levels of integration.

2) an increase in the probability that new industries in the area will appear; 3) an increase in the interest of large corporations to establish microfabrication plants in the region, motivated by available capable human resources; 4) possibility to provide engineering services to places all over the world, based on an increasing global economy and availability of efficient new communications networks.

255

The above considerations justify the need to improve and spread out education in microfabrication technologies in countries like those from Latin America. We would like to stress that even circuit or microstructure designers would greatly benefit from a microfabrication course with hands-on experience, as is the case in developed countries. IV. EDUCATION AT UNDERGRADUATE GRADUATE ENGINEERING COURSES

AND

It can be observed that, in general, microfabrication technologies are not taught through experimental courses at the universities. This topic is only given through lecture classes in courses like electronic materials and/or electronic devices. Exceptions are some optional experimental courses at some of the few universities mentioned above that have microfabrication facilities. This blackboard approach to microelectronics teaching is even worse if we consider that most of the faculty members teaching these subjects have no hands-on experience in microfabrication. This shows that education on microfabrication is far from what it should be. To improve this picture, the following actions should be considered by the educational system. • Offer additional training for the faculty members. This could be solved through special workshops designed for faculty members, as for instance the one offered by San Jose State University [3] and others. • Install simple experimental set-ups at engineering schools, to demonstrate and perform basic microfabrication experiments. We consider that concepts such as the “paradigm experiment” and the “opportunistic laboratory management” [3] are specially suitable for this region. However, relying on donations from companies is far more difficult in Latin America as compared to the United States. • Organize experimental courses at the few universities with microfabrication facilities, for students from other universities all over the region. V. THE MICROELECTRONICS WORKSHOP AT UNICAMP, SP-BRAZIL In order to compliment microfabrication technology education and to encourage research and development in this area, UNICAMP has organized a series of workshops, called “Oficina de Microeletrônica.” The first four were organized by Mammana during 1977 to 1982, bringing speakers mainly from the United States and Europe. Recently these series of workshops have been resumed, by initiative of the author, in collaboration with an organizing committee (C. A. Reis, F. Damiani, M. Pimentel, P. J. Tatsch, and R. G. Pereira) and a large number of speakers (21) and laboratory instructors (22) [9]. The fifth “Oficina de Microeletrônica” took place at UNICAMP from July 15–26, 1996. These new workshops have been adapted to the present time, with the following characteristics. • The target audience is last-year undergraduate, graduate students, as well as professionals from industries and faculties.

256

• Activities are in the form of 50% lectures and 50% laboratory hands-on experiments. • The lectures are all given by faculty members of different local universities, each in its own area of expertise. As mentioned before, presently there are a reasonable number of local experts available, in contrast to the time of the first four workshops. • Manuscripts of the lectures and laboratory activities are given to the participants [9]. In the case of the fifth workshop, the subject was mainly microelectronics, including the following subjects. A. Lectures Electronic devices (BJT, MOSFET, MESFET, and power devices); fabrication processing steps (diffusion, ion implantation, epitaxy, thin films, lithography, and plasma etching); process integration (CMOS, MESFE,T and multilevel metallization); reliability and failure analysis; digital IC design; analog IC design; IC design syntheses; design for testability; and MMIC design. B. Experimental Activities All processing steps to fabricate a Schottky diode on GaAs were performed; electrical characterization of Schottky diodes were done and also electrical characterization of a CMOS test chip (processed in Barcelona, Spain) were performed; demonstration of IC design tools and design exercises on a Gate Array were done by the participants. The fifth workshop was a two-week period, with the first week focused on devices and fabrication and the second week was dedicated to circuit design. Due to the laboratory activities, the group of participants was limited to 30 persons. The participants came from five countries of Latin America, with most being from Brazil, from four different states. With the experience acquired in the previous workshops, we are able to and plan to expand the next workshops. These will be composed of two partially parallel groups. Group A will follow the same schedule as the previous fifth workshop, while group B will have in common 50% of the program of group A, namely the topics of electronic devices and microfabrication. The IC design module will be replaced for group B by a module on optoelectronic, photonic, sensor, actuator, and micromechanical devices, fabrication and characterization. In this manner, group A will be more specific for electronic engineering, while group B for material engineering and for applied physics. This expansion allows us to accept a total of 60 participants. As mentioned, these workshops aim to compliment the regular engineering and applied physics courses, especially where no microfabrication facilities are available. It offers a wide view of the area and even more important, an opportunity for students to have a hands-on experience. This can demystify some ideas that form barriers for new initiatives in this important technological area. It also allows current and future faculty members to have a better view about the subjects they do or will teach. Besides organizing the workshops, UNICAMP also offers regular under-graduate as well as


graduate courses on microfabrication techniques, with the same goals. VI. CONCLUSIONS Arguments are presented showing the importance of education on microfabrication also in countries like those from Latin America. This would result in many benefits and opportunities to increase participation in the huge and increasing microfabricated devices market. It is shown that the present education programs are very limited at most universities in the region. A workshop at the State University of Campinas, called “Oficina de Microeletrônica,” has been resumed recently and aims to reduce the gap in education on microfabrication technologies. ACKNOWLEDGMENT The authors are thankful for information about other Latin America countries received from Dr. N. Beltran from University of Santiago, Chile, and from Dr. J. L. H. Pérez from University of Puebla, México. Also collaboration received from CTI is acknowledged. REFERENCES [1] J. P. de Sousa, E. C. Rodriguez, and J. W. Swart, “A very simple Algate technology for high performance LSI Circuits,” Int. J. Electron., vol. 54, no. 1, pp. 155–159, 1983. [2] J. E. Ripper and R. C. C. Leite, “Physics in a developing country,” in Proc. Int. Conf. Phys. Ind., Dublin, Ireland, 1976, pp. 221–223. [3] P. S. Gwozdz, “NSF microfabrication workshops,” IEEE Trans. Educ., vol. 39, pp. 211–216, May 1996. [4] J. W. Swart and W. A. M. Van Noije, “The multi project chip program of FAPESP,” J. Solid-State Devices Circuits, vol. 4, no. 2, pp. 30–30, July 1996. [5] S. Finco, P. Serazzi, and C. I. Z. Mammana, “The Brazilian multiproject wafer,” J. Solid-State Devices Circuits, vol. 5, no.02, pp. 27–29, July 1997. [6] P. Singer, “Looking down the road to quarter-micron production,” Semiconductor Int., vol. 18, no. 1, p. 46, 1995. [7] A. B. Frazier, R. O. Warrington, and C. Friedrich, “The miniaturization technologies: Past, present, and future,” IEEE Trans. Ind. Electron., vol. 42, pp. 423–430, 1995. [8] R. Singhvi et al., “Engineering cell shape and function,” Science, vol. 264, pp. 696–698, 1994. [9] Lecture Notes of “V Oficina de Microeletrônica,” C. A. Reis, F. Damiani, J. W. Swart, M. Pimentel, P. J. Tatsch, and R. G. Pereira, Eds. Campinas, SP, Brazil: UNICAMP, July 15–26, 1996. Jacobus W. Swart (S’76–M’77–SM’90) was born in 1950 in The Netherlands. He received the B.Eng. and Dr.Eng. degrees in 1975 and 1981, respectively, from the Polytechnic School of the University of São Paulo, Brazil. After his Dr.Eng. degree, he worked at the following institutions: K. U. Leuven, Belgium, from 1982 to 1983, as a Postdoctoral Fellow; CTI, Campinas, 1984, as head of Process Engineering; LSIUniversity of São Paulo, 1985 to 1988, as Assistant Professor; SID Microeletrônica, 1986, as part-time Researcher; RTI, USA, 1991, as a Visiting Scientist and State University of Campinas, since 1988, presently as Full Professor and Head of the Center of Semiconductor Components. He worked on CCD, nMOS, and CMOS process integration, device characterization and modeling, gettering processes, rapid thermal processes, and silicide formation and characterization. His current research activities and interests include MESFET and HBT devices, process integration and circuit design on III-V materials, refractory gate metallization, CVD, and plasma and RTP processes. He has presented more than 100 papers in journals and conferences, more than 40 in international media. He has advised 21 Dr. and MSc. degree students. Dr. Swart is a member of ECS, SBMicro, and SBMO, and was President of SBMicro from 1988 to 1990 and reelected for the period of 1998 to 2000.