Chapter 2From Germanium to SiliconA History of Changein the Technologyof the Semiconductorsby Philip SeidenbergSilicon has been the dominant semiconductor materialsince the middle 1960s. Today, probably 95% of allsemiconductors are fabricated in silicon, yet the firsttransistor was a germanium device. Until 1960 most designengineers preferred germanium to silicon for computer logiccircuits, when, suddenly, germanium was out, and siliconwas in. What caused this abrupt shift to silicon? An answerto this question requires some understanding of how andwhy solid-state scientists went about their research.This chapter explores the technical choicesconcerning the use of germanium and silicon assemiconductor materials made by scientists and engineersduring the period from 1947, the year the transistor wasinvented, until 1960 when the design shift from germaniumto silicon occurred. During this period, the scientists andengineers at Bell Telephone Laboratories (BTL) were theworld's leading investigators of the properties ofsemiconductors. With few exceptions, scientific research on
Seidenbergthe transistor was dominated by BTL; from 1947 through1959, twenty-four of the forty-four papers cited aspioneering papers in semiconductor device technology werewritten by BTL researchers.! The remaining twentyoriginated from fourteen entities, split evenly between thecorporate world and the academic/government complex. Infact, it was not until the end of the 1950s that other industriallaboratories begin to challenge the dominance of BTL insemiconductor research and development. This is in starkcontrast with the relatively minor role that BTL and itsmanufacturing arm, Western Electric, played insemiconductor manufacturing and sales. Through 1958Western Electric produced 1.315 million transistors,2representing less than 2% of the total amount of transistorsproduced by the semiconductor industry during this period. 3In spite of its secondary role as a semiconductorsupplier, the course of semiconductor research was shapedby the scientific and engineering achievements at BTL duringthe 1950s. While most of the semiconductor industryconcentrated on manufacturing germanium transistors anddiodes during this decade, BTL spent most of its researchdollars on silicon devices. This research, coupled with theresponse by the rest of the industry, led to the developmentof the surface-stabilized silicon transistor and diode, whichin turn signaled the replacement of the commerciallydominant germanium by silicon as the leading semiconductormaterial. In this chapter, I will argue that the precipitatingcause for silicon replacing germanium, particularly incomputer logic, was the use of a silicon dioxide film tostabilize the surface of silicon semiconductors. This was themost important factor in the development of the planarprocess, wh,ich led first to the dominance of the silicontransistor and then to the integrated circuit.36
From Germanium to SiliconThere is very little historical research on theimportant change from germanium to silicon. Mosthistorians exploring the semiconductor field recognize thesignificance of the planar process, but treat lightly theprogress leading to this innovation. The seminal work onsemiconductor electronics refers very briefly to thestabilization of the silicon surface4 and another establishedhistorical work devotes slightly over half a page to the"surface stabilization of silicon by the oxide maskingprocess."5 An important book on technology transfer insemiconductors chronicles the changeover from germaniumto silicon but offers no insight on causes. 6 None of thesetexts indicates that the replacement of germanium by siliconwas an important event in the history of technology, despitethe fact that many scholars consider the transistor as the mostpervasive invention of the last half of the twentieth century.Furthermore, the works that examine progress made in thephysics of semiconductor surfaces do not expound on itsinfluence on technology or industry. The volume onelectronics technology and the physical science in BTL's AHistory of Engineering and Science in the Bell Systemoffer a non-technical overview of what happened in surfacephysics and processing during the 1950s, yet BTL'shistorians present no explanation or interpretation of whatthe replacement of germanium by silicon meant tosemiconductor manufacturers and their computercustomers. 7When they do contemplate the demise of thegermanium semiconductor, historians often ascribe it to theinfluence of the Defense Department. It has been argued thataggregate funding figures for transistor development can beused as a gauge for military influence on semiconductor37
Seidenbergtechnology, and that military semiconductor technologyspilled over into the commercial economy. 8 Although themilitary did fund silicon transistor development during the1950s, both military and commercial computermanufacturers almost exclusively used germanium diodelogic during this period because germanium diodes could bemass produced more reliably and less expensively than couldany other semiconductor. For example, the MinutemanMissile Program, which was probably the most importantmilitary program for semiconductor manufacturers in the late1950s, used germanium diode logic in its guidance system.Consider; also, that digital logic, the preferred choice amongcircuit designers for commercial computers, required largenumbers of inexpensive and reliable switching elements.The germanium gold-bonded diode fit their requirements.The tube and electromechanical relay were large powerhungry devices. The transistor, a three terminal device,suffered from low yields and unreliable performance whencompared to the two-terminal diode. Germanium diodeswere high-yield economical electronic switchesmanufactured in volume by reputable and reliable vendors.Clearly, the question of which material, germanium orsilicon, would dominate semiconductors was not driven bymilitary considerations.Much of the literature about the military's earlyinfluence on the semiconductor field was written during the1980s, during the Reagan years of military spending. AsAlex Roland has observed, "A kind of presentism movesand infects much of this (1980s) literature." Aggregateresearch funding by the Defense Department favored silicondevices, but it was germanium that the military purchased forhigh-volume logic requirements. Even' at BTL, a virtual38
From Germanium to Siliconhotbed of silicon research, practically all semiconductorsmanufactured through 1958 used germanium as the basic orstarting material.Semiconductor development has been drivenprimarily by the study of the physics and chemistry of thebody and surfaces of its starting material. The body-theregion of the material away from the surface-is ahomogenous periodic structure. Because it poses lesserobstacles to theoretical comprehension than the surface, ithas been easier to analyze and measure. For a long timesolid-state theorists implicitly assumed that the properties ofsemiconductors were determined strictly by what happenedin the body of the material. The surface-construed here asthe boundary area between the body of the semiconductorand any other medium-does not have the symmetry andperiodicity of the body. Only a few atomic layers deep, thesurface influences profoundly the semiconductor action inthe body. These atomic layers of material foreign to thebody often are difficult to identify and neutralize.Early efforts to understand semiconductor actiondepended largely on empirical inquiry. Theoretical conceptswere not formulated until the introduction of quantummechanics into semiconductors and the explanation for thebehavior of electrons in metals and semiconductors.AT&T' s corporate strategy for basic research at BTL gelledin the period 1925 to 1935 after the introduction of the newquantum physics. The idea of applying basic science totechnology became embedded in BTL strategy. As historianLillian Hoddeson has noted, "by the mid-1930's theproblems, approaches and atmospheres of fundamentalresearch at Bell Labs were remarkably similar to those inuniversity laboratories."9 Mervin Kelly set the research and39
Seidenbergdevelopment (R&D) parameters for BTL after World War II.Kelly, the vice-president of BTL, decided to focus R&Dactivities on semiconductors in order to develop a solid-stateamplifier for telephone applications. Before the war, Kellyhad been interested in substituting electronic relays formechanical relays in telephone exchanges."lO The SecondWorld War changed BTL's focus from peacetime research towartime engineering.The war accelerated semiconductor development asscientists were transformed into engineers to work ongermanium and silicon crystal rectifiers for radar and radioequipment. Scientists and engineers selected germanium andsilicon for military research because they were elementalsemiconductors with an orderly atomic structure.Germanium and silicon crystallize in a diamond lattice,consisting entirely of one type of atom. They possess a lesscomplex structure than selenium or certain intermetalliccompounds such as copper oxide, which were usedextensively in rectifiers from the 1920s. Scientists at BTLanticipated easier chemical processing and fewer structuraldefects in materials that showed the least crystal complexity.After World War II, BTL continued its pre-warresearch strategy. In the post-war social context, technologymeant progress. Radar had saved Britain from the Luftwaffein 1940. The atomic bomb had crushed Japan in 1945.Technology could win the peace as it had won the war, withthe United States lighting the way for the world. FrederickE. Terman, the Stanford University professor who headedtop-secret radar research at Harvard University during thewar, summed up the feeling of the scientific communitywhen he said World War II showed "science and technologyare more important to national defense than masses of men"/40
From Germanium to Siliconand "how essential the electron was to our type ofcivilization."11 It is in this environment that BTL embarkedupon a sustained research-and-development effort whichdrove progress in the semiconductor industry during the1950s.Solid-state physicists interested in doing research inbasic science in the post-war period were faced with onlyone real choice if they preferred to work in industry ratherthan academia; BTL was the premier industrial laboratory forsolid-state research. The management at BTL was intent onattracting talented researchers and creating an atmosphere offellowship12 and the result was a research facility whichcould be compared favorably with those of the leadinguniversities in the United States. 13 George F. Dacey, whoworked for William Shockley, recalls that Kelly wasinsistent on recruiting only the best people that could befound. Shockley, who had joined BTL in 1936, already hadachieved a reputation in the field and served as a beacon toattract additional bright solid-state scientists. 14 MorganSparks, who took over Shockley's group in the mid-1950s,was impressed with Shockley's great insight as a theoreticalphysicist. Shockley was not too concerned with themathematics of a problem as long as he understood most ofthe figures. Shockley also evinced a great interest in deviceapplications 15 and appeared to approach a problem with aright mix of the theoretical with the practical. Dacey andSparks recall the frequent meetings, both formal andinformal, held at BTL to discuss the almost daily appearanceof new semiconductor phenomena ensuing from solid-stateresearch. Progress was so rapid in semiconductors that itdominated the meeting schedules. By 1950, Shockley hadabout twenty people in his solid-state physics group, which41
Seidenbergultimately grew to nearly one hundred by 1960 under thedirectorship of Dacey, one of Shockley's successors.Typical of the people drawn to BTL after 1950 wasC.G.B. Garrett. Garrett, who was a graduate of CambridgeUniversity, and an instructor in physics at HarvardUniversity, joined BTL in 1952 to do research onsemiconductor surfaces because he felt that it could result inprofessional recognition. Moreover, BTL could afford topay well and the laboratory atmosphere was collegial.Technology diffused easily and rapidly among all researchand development groups. Regularly scheduled, topical andad hoc meetings characterized the diffusion of semiconductorknowledge within BTL. It is in this ambiance thatsemiconductor research flourished. 16A tabular view of the major milestones in transistorresearch and development between 1947-1959 reveals theimpact of BTL's research in this area (see Appendix 1). Ofthe ten major events instrumental in the demise ofgermanium as the preeminent starting material, sevenoriginated at BTL while the other three were derived fromBTL research activities. Of course, there was considerablymore happening in semiconductor research and developmentat BTL during this time besides these important events. Forexample, BTL researchers developed the phototransistor(J.N. Shive, 1950), the junction field-effect transistor (G.C.Dacey and I.M. Ross, 1952), the thermocompression wirebonder (O.L. Anderson, H. Christensen and P. Andreatch,1956), and other significant process and product innovationsnot directly connected to the history of the replacement ofgermanium by silicon.One should not conclude these milestones intechnological innovation were part of a continuous stream of42
From Germanium to SiliconBTL's successes, unaccompanied by individual failures andfrustrations. However, the rate of technological progress atBTL during this period did allow the semiconductor industryto grow to about one half billion dollars by 1960. 17Compare this to the inability of solid-state scientists andengineers to agree on common concepts for a basis ofresearch in the fifteen years from 1931, when the theory ofthe electronic semiconductor based upon quantum mechanicswas advanced. In a review of semiconductor research in a1955 issue of the Proceedings of the Institute of RadioEngineers, G.L. Pearson and Walter Brattain noted that "ittook about fifteen years for the full light to dawn" and thatthere were "many blind spots in the working concepts aboutsemiconductors in the nineteen thirties."18The quality and quantity of the researchers at BTL,operating in an atmosphere of easy communication andunfettered disclosure, contributed considerably to thesemiconductor industry's high rate of growth during the1950s. The strategy of the management to encourage theresearchers to be independent promoted basic solid-stateresearch. Brattain remembers that after many years ofresearch trying "to understand what was really going on inthe simplest of semiconductors, silicon and germanium," hebegan "to lose faith." However, he felt "no pressure frommanagement to continue or to change fields.".What had influenced Shockley and his twocolleagues, Brattain and Bardeen, to join BTL andconcentrate on the study of semiconductor surfaces? In1972, in an issue of the Bell Laboratories Recordcelebrating the 25th anniversary of the transistor, the threescientists recalled the personal interests that led to thediscovery of the transistor. The leader of the group, William43
SeidenbergShockley, who obtained his doctorate by studying thebehavior of electrons in crystals, was attracted to BTL in1936 by the opportunity to work with C.J. Davisson, aneminent physicist in the field of electron behavior. 19 WalterBrattain, the oldest of the three, evinced an early interest inquantum mechanics and entered the field of surface physicsat BTL in 1929 after receiving his doctorate. In 1937Brattain met Davisson when the latter used Brattain'slaboratory to demonstrate his work on electron behavior.John Bardeen, who joined BTL in 1945, had no experiencewith semiconductors, but prior to the war he had beeninterested in the theory of metals. He was anxious to "returnto solid-state physics after five years at the Naval OrdnanceLaboratory in Washington, D.C."20 Trained originally as anelectrical engineer, he had returned to school at Princetonafter several years in industry to obtain his doctorate inmathematical physics, drawn there by the opportunity tostudy quantum theory under Einstein.All three scientists attributed the success of theirresearch to the supportive atmosphere created by the BTL inwhich management allowed those in research to spend manyyears, much money, and the talents of many people toexplore solid-state physics. From their recollections, itappears that all three were motivated by an interest in surfacephysics and electron behavior, and an opportunity to studyand work with eminent scientists in their chosen fields. Thephysical reality of the microscopic world of atoms openednew vistas for students of physics, particularly thoseexploring electron behavior in solids. On the 25thanniversary of the transistor, John Bardeen recalled that heand Walter Brattain discovered the transistor as a result offollow-up experiments on some consequences of Bardeen's44
From Germanium to Siliconpublished investigations of semiconductor surfaceproperties. 21 Shockley, who with Bardeen and Brattainreceived the Nobel prize in 1956 for the invention of thetransistor, had theorized in 1939 that localized states canexist at the surface of a semiconductor. 22 However, solidstate physicists did not realize the importance of thesesurface states at the time. The accepted theories ofsemiconductor action formulated prior to World War IIassumed that rectification was closely related to the polarityand magnitude of the contact potential at the interfacebetween a conductor (metal) and a semiconductor, orbetween two different s,emiconductors. The polarity andmagnitude of the contact potential depended upon thedifference in work functions of the metal and thesemiconductor. However, experimental measurements didnot confirm the predictions of the theory. There was littlecorrelation between the work function of the metal and thecontact potential.This perplexing phenomenon was made whollyintelligible by Bardeen's hypothesis in 1947. Surfaceproperties rather than the contact potential at the interface ofthe metal and the semiconductor were the governing factorsfor device rectification. Bardeen theorized that electronsfrom the body of the semiconductor material become trappedand immobilized at the surface, repelling other electrons inthe conduction band. This produced a layer of depletedconductivity below the surface. 23 He further theorized thatsurface states occur when the body of the semiconductormaterial is abruptly terminated, allowing electrons and holes(the absence of electrons) in the forbidden energy gapbetween the valence band and the higher-energy conductionband. Normally, in the body of the semiconductor material,45
Seidenbergelectrons exist in the conduction band and holes in thevalence band. Between these two bands is the forbidden gapwhere neither electrons nor holes reside unless trapped byimpurities (surface states tend to act like impurities inpermitting electrons and holes to have energies in theforbidden gap). Trillions of surface states exist on a squarecentimeter of semiconductor material, dominating theelectrical properties of the structure. These surface stateswork against any electric fields or currents applied inaccordance with circuit design and also work to neutralizeany desired imbalance in electrons and holes by tending toreturn the semiconductor to its condition before theapplication of the electric field or currents. Realizing thatthese surface effects can dominate the body properties of thesemiconductor, Bardeen and Brattain conducted experimentsto change the surface potential of germanium. It was duringthis research that they invented the transistor24 (Appendix2a). Thus, the mutual interest of these two physicists inelectron behavior in solid states led them to this importantdiscovery.The inven
Silicon has been the dominant semiconductor material since the middle 1960s. Today, probably 95% ofall semiconductors are fabricated in silicon, yet the first transistor was a germanium device. Until 1960 most design engineers preferred germanium to silicon for computer logic circuits, when, suddenly, germanium was out, and silicon was in. What caused this abrupt shift to silicon? An answer to .
Part One: Heir of Ash Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Chapter 12 Chapter 13 Chapter 14 Chapter 15 Chapter 16 Chapter 17 Chapter 18 Chapter 19 Chapter 20 Chapter 21 Chapter 22 Chapter 23 Chapter 24 Chapter 25 Chapter 26 Chapter 27 Chapter 28 Chapter 29 Chapter 30 .
TO KILL A MOCKINGBIRD. Contents Dedication Epigraph Part One Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Part Two Chapter 12 Chapter 13 Chapter 14 Chapter 15 Chapter 16 Chapter 17 Chapter 18. Chapter 19 Chapter 20 Chapter 21 Chapter 22 Chapter 23 Chapter 24 Chapter 25 Chapter 26
The CMOS Process - photolithography (1) Silicon Wafer Silicon Wafer SiO 2 1μm Silicon Wafer photoresist (a) Bare silicon wafer (b) Grow Oxide layer (c) Spin on photoresist Lecture 3 - 4 The CMOS Process - photolithography (2) Silicon Wafer (d) Expose resist to UV light through a MASK Silicon Wafer (e) Remove unexposed resist Silicon Wafer
DEDICATION PART ONE Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 PART TWO Chapter 12 Chapter 13 Chapter 14 Chapter 15 Chapter 16 Chapter 17 Chapter 18 Chapter 19 Chapter 20 Chapter 21 Chapter 22 Chapter 23 .
A remind of the basic properties of silicon will be of great importance to understand well the silicon nanocrystals properties. 2 PHYSICAL PROPERTIES OF SILICON Silicon is, on Earth, the most abundant element after oxygen; weestimates that the Earth's surface is composed about 26% of silicon.  This
Intel's Silicon Photonics Research Innovating with lowInnovating with low-cost silicon to create new optical devicescost silicon to create new optical devices 1st Continuous Wave Silicon Raman Laser (Feb '05) Hybrid Silicon Laser (Sept. '06) Silicon Modulators 1GHz ( Feb '04) 10 Gb/s (Apr '05) 40 Gb/s (July '07) 8-channel integrated
Strained germanium thin film membrane on silicon substrate for optoelectronics . Donguk Nam, 1,* Devanand Sukhdeo, 1. Arunanshu Roy, 1. Krishna Balram, 1
Introduction to AutoCAD Academic Resource Center . What is CAD? Computer Aided Drafting Autodesk is the most popular drawing program Many student versions available for free online at students.autodesk.com o AutoCAD o Architecture o Mechanical o Revit o Inventor o Civil o MEP o etc. Capabilities: o 2D line drawings o 3D constructions o Rendering o Part Assemblies . Workshop Goals .