CALIFORNIA

INSTITUTE Of

TECHNOLOGY

E N GI N E E

Volume LX I,

N II1llber 3,

I 9 9 8

IN THIS ISSUE

• Prosecuting

Science

Simulating

Molecules

Observing

Anniversaries London Bridge? The Manhattan skyline? No, it's an oil refinery. figuring out how to squeeze more gasoline rrom a barrel or crude is just one project a group or theoretical chemists at Caltech have taken on in a series or academic­ industrial collaborations. for more on this partner­ sh ip, see the story on page 20. Photo courtesy or Chevron Corporation.

\ California Inst i tute

of Technology

2 Random Wa l k

10 Scie ntific Fraud and Misconduct in American Po litica l Cu lture : Reflections on the Baltimore Case - by Dan ie l J. Kevles Excessive zeal on the part of legislators and the media hyped public suspicion of scientists' behavior.

20 " H ave Method , Wi ll T rave l" - by Douglas L. Smi t h Computational chemistry ventures into the Real World as Cal tech theorists tackle industrial problems.

30 Tw o A s tronomical Anniv ers aries : Palomar at 50 Still going strong , the hale and hearty Hale Telescope hits the half century mark.

36 ... and OVRO at 40 Radio astronomers parry, too. On t he Cover: The six IO.4·meter antennas (one 42 Boo ks - The Earth in Tf{rmoil by Kerr y Si eh a n d S i mon LeVay is hidden behind the others here) of the Owe ns 43 Faculty Fil e Valley Radio Observat ory millimeter· wave array can receive signals from up t o Engineering & Science (ISSN 0013-7812) is published Warren G. Goda billi on light years away. 10 quarterly at the Califo rnia Institute ofTcchnology, 1200 President of the Alumni AHociatioll Designed by the late 80b East California Boulevard, Pasadena, CA 9 1125. Annual J. Ernest N unnally subscription $10.00 domestic, $20.00 foreig n air mai l; Leighton (8S '41 , MS '44, Vice Presidelll for /mlilute Relations single copies $3.00. Send subscriptions to Cal tech 1-7 1, Robert 1. O 'Rourke PhD '47), t he millimeter· Pasadena, CA 91125. T hird class postage paid at Pasade­ A ssociate Vice P' -esideI1 J!o rlmtilllte Relatiom wave array· is t he latest na, CA. All rights reserved. Reproduction of m aterial co n~ tained herei n forbidden wi thout authorizat ion. © 1998, STAFF: Editor - J ane D ietrich addition to OVRO, which California Instit ure of Technology. Published by Ca ltcch M allagillg Editor - Douglas Smith celebrated its 40th and the Alumni Association. Telep hone: 626· 395-3630. Copy Editors - Sabra Akmal, Michael Farquhar, anniversary in October. PlcruRE CREDITS; Cover, 36--10 - Jane Dieu ich; 2 - Al J ulie H akewill Forster; 2- 3,6--7 - JPL; 4,7, 9, 31 - Doug Smith; 5 - Frank COl1tribJltillg \f1ritm - Ryan Poquette, Robert Tindol Palomar Observatory also Monzon; Erika Oller; 9.30,43,44 - Bob Paz; 12 - Bettyann BJlsillesJ M anager - Debbie Bradbury celebrated its 50th; t he Kevles; 14 - Ken Heinen; AP/Widc World Phoros; 17 - Seth Circulation iHtlllager - Susan Lee Resnick; 22 - Chevro n Corp.; 22-25 - Mario Blanco; 26-29- Photographer - Robert Paz two stories begin on Jason Perry; 3 1 - Moore Montgomery; 32 - J im McClanahan; page 30. 36-38 - J ohn Dohan; 39 - Curtis Phillips Visit Caltech on the Web at http://www.caltcch.edu Random Walk

S URV E YIN G M ARS

Ri ght, top: Mars's largest volcano, BROAD BUILDING ANNOUNCED Olympus Mons, is as big as Arizona and taller than three Everest s. A "shield" volcano, form ed by the gentle oozing of very fluid lava, Olympus Mons is very nearly as flat as a pancake-once one surmounts the cliffs ringing t he volcano, typical slopes are 2- S degrees. Mars Global Surveyor (MGS) took this enhanced-color image on a cool, crisp, Martian- winter morning. Note the clouds lapping against the mountain's fl a nks on the east (top) side. Bottom: Martian volcanoes come in all sizes. This is the first-ever close-up view of one of the small Eli Broad, one of Southern California's most prominent shield volcanoes that dot the civic and business leaders, has donated $18 million to create rhe Broad Center for the Biological Sciences, which will provide planet. This one has a two­ 100,000 square feet of space fo r 10 research groups char will kilometer-long, ISO-met er.deep, work at rhe cutting edge of the biological sciences. The new paramecium-shaped caldera at its building, shown above in a preLiminary re ndering. will be located in the northwest quadrant of campus near the Beckman summit, and t he whole volcano Institute. As rhe corn erstone capital p roject of the Biological would fit unnoticed in Olympus Sciences Initiative, the build ing will provide crucial infra­ Mons's 7S-kilometer-diameter structure underpinning [he Ins titute's new capabilities fo r caldera. magnetic imaging, s[(uctllcal chemistry, and genetics. Broad's gift is [he largest donation so fa r in Cal tech's new Biological Sciences Iniciative, which aims to raise $100 million for new filCultyand resources. A cotal of $56 million has been raised since the initiative was announced this past May.

2 ENGINE£KING & S(IENCE No. 3 Left: Mars's north polar cap as seen by MGS on September 12, early in the Martian spring. The permanent cap, believed to be water ice mixed with dust, has a scalloped, layered look due to channels eroded on its surface. (It actually has layers, but they're far too fine to be seen here.) The ground throughout this image is covered with carbon dioxide frost,

Mars may not have life, but it does which appears pink rather than

have fossils of a kind-the six~ white-probably for the same kilometer crater in the MGS image reason that springtime snow on at right used to be buried under Earth is often a dirty brown. about three kilometers of Martian

bedrock, part of which is still visi ~ ble in the upper left. The crater reemerged when catastrophic floods carved the gargantuan channel system called Kasei Vallis more than a billion years ago. (A similar flood carved the Ares Valles, where Pathfinder landed.) The crater survived because its rim stuck up above the flood, which didn't last long enough to erode a breach. The other two images, Above: This MGS close~up of • which show ever larger regions of Stickney, Phobos's largest crater­ Kasei Vallis, are from Viking I. at 10 kilometers, nearly half the diameter of Phobos itself-shows features as small as four meters (the length of a VW Beetle), and is one of the highest-resolution views ever of Mars's moons. At the same time, MGS was mapping Phobos's surface temperature, as shown. The fact that the night side gets so cold so quickly (Phobos rotates in seven hours) hints that the surface is covered with a meter-thick layer of very fine particles, which would lose heat rapidly. In other words, Phobos is hip-deep in powder from millions of years of meteor impacts, and getting around on it is going to be a royal pain, even though the gravitational field is only 1/1000 that of Earth's.

ENGINEE RI N G & SCI ENCE NO. ] Left: Make way! Make way! Make way for the Sojourner rover! This diorama showcases a system for clearing intersections in advance of ambulances, fire trucks, and other emergency vehicles (note the police car behind the rover). The vehicle carries a transponder that commandeers the traffic signals and causes a warning sign D OUBL E M ANHATTAN (arrowed) to flash at cross traffic.

The stuff on the chips in your com pu ter is essentially WHAT NEXT-TECHNO-TROUSERS fROM NASA? laid out in two dimensions, like a city. Oh, sure, the buildings stick up, and there are subways and pipes and Gening technology that kids. And an infrared camera cables underground (and the was developed at taxpayer is being used co detect cherry devices on a chip stick up, expense into the commerc.ial pits (o r parts thereof) in pie while some conneccors are arena is all the rage these filling. buried beneath the surface), days, but the Technology Speaking of cherries, a JPL­ but you can't stack the Mfiliates Program at JPL has designed imaging system Empire State Building atop been doing it since before it periodically scans the the Chrysler Building. Now became fashionable. This Constitution, the Bill of Cal tech researchers have decade-old program has Rights, and the Declaration found a way to layer multiple agreements with some 120 of Independence fot signs of copies of the entire borough companies, and about 200 deterioration. Another of Manhattan, as it were, in technology-transfer projects system was used co examine silicon. Thomas McGill (MS under way or completed. rhe Dead Sea Scrolls in rhe far '65, PhD '69), the Fletcher A lot of these projects are infrared, revealing lettering J ones Professor of Applied what you might expect: that had faded co invisi bility Physics, and colleagues reporr [obmic surgical assistants chat at sharrer wavelengths. the work in the October issue have a steadier hand cban any And 1PL's experience in of theJOItrnal o/Vacuum Science human; remote sensing of dealing with rhe icy cold of and Techllology B. hazardous waste sites, oil space has led to more efficient The method employs a spills, and land-use patterns; refrigerated display cases for widely used process called purring on CD-ROM the supermarkets, and a superior molecular-beam epitaxy to positions of the sun, moon, inslJaco r for the mail-order grow carefully controlled Above: Jo Pitesky of JPL's Commer­ and p lanets (what aStrono­ meat business. layers of various materials on mers call an ephemeris) from Then there's the fun stuff. cialization Program Office holds a chip. The method begins 3000 BC ro AD 3000; and There are the H ot Wheels, more spinoffs-the Mars Path­ with a silicon wafer, onto which an insulating layer of integrating Global Position­ of course. ]PL radar data has finde r and John Glenn Hot Wheels ing System data inca the next been used to create a colleCt­ cerium dioxide just a few Action Packs; a rearview mirror generation of the nation's air­ ible three-dimensional model acorns thick is grown. Then a traffic control system. But of the asteroid Toutatis, with that automatically dims in single crystal of silicon is some 1PL spinoffs have taken one of Castalia on the way. proportion to the amount of grown back onto the cerium dioxide, resulting in a three­ odder twists. 1PL scientists and engineers headlight glare; and a world map, For example. an arnal are providing technical infor­ dimensional device contain­ thermometer based on 1PL mation for Crusade, a new TV published by the National ing a cerium dioxide insulator infrared-sensor technology is series from the producers of Geographic Society, that has a and having a fresh layer of crystalline si licon on cop. now in widespread use. The Babyloll 5. And data from political map on one side and a thermometer, which gives a Pathfinder's weathet station The wafer is then ready co high-resolution photomosaic (each reading in seconds and is held have been converted into begin the process again. to the patient'S ear instead of sounds that, woven into the pixel is one square kilometer) on Layer upon layer of devices being inserted somewhere, is strains of1. S. Bach, are out the other. may be grown, one after especially useful on fidgety on CD as Winds of Mars. [J another, on the same chip.

4 ENGINEERING & SCIENCE NO. "The implications are enabled by rhis new tech­ major contributor in 10 to very significant," says McGill. nique, including resonant 15 years. " "For years there have been tunneling diodes. These In addition to McGill predictions that progress devices can be exploired for and J ones, rhe authors of the will eventually stop in sili­ novel memory storage, as well paper are Visiting Associate con electronics because rhe as used to enhance the in Applied Physics Edward devices will have been shrunk performance of numerous Croke III (PhD '91), Member as much as th ey can. Bur this other microelectronic c.ircuits. of the Professional Staff Carol new technology could allow "The silicon industry is a Garland, and Visiting you to get the functionality $100 billion industry," says Associate in Applied Physics increase by stacking instead McGill, "This could be a Ogden Marsh, n -RT of shrinking." McGill says the group has stacked only a single extra layer of silicon so far. How­ ever, the key is the demon­ saarion rhat the cerium oxide is indeed acting as an insula­ tor, and thar the silicon on top is one single crystal, suitable for further growth. "In principle, YOll can stack forever," he says. There's still time to get a set of According to grad student punk rats, suitable for framing, of Joel J ones, anorher member your very own. Artist Erika Oller of the ream, this technique is especially interesting because has graciously given us permission ir also allows the fabrication to sell prints at cost- approxi­ of a new group of novel mately $25 each. Call us at silicon devices. "We've already fabricated a primitive (626) 395-6730 for details. runnel-switched diode from the mulrilayered chips," Jones explains. "This is a single device that exhibits If you hit enough balls out of the parl<, one of them is bound to bounce off memory. At a given voltage, the ivory tower. Franklin Monzon, a grad student in Professor of Physics you can have twO different Michael Roukes's lab, used a scanning electron microscope to create what stable currents depending he calls "the world's smallest tribute to the world's biggest slugger." (In on how you've swirched rhe device." This phenomenon case your window in the tower doesn't face the ballpark, this would refer is called negative differential to Mark McGwire's hitting a record-breaking 70 home runs last summer.) resistance, allowing two Monzon drew t hese size_Xl.smali jerseys on a gallium arsenide wafer using current srates of different electron beam lithography, a standard chipmaking technique. The scale bar amperage to exisr at the same voltage. Similar effects can across the bottom of the photo is 10 millionths of a meter long. It would be found in other devices take more than a billion of these nanojerseys to cover McGwire's real jersey.

ENGINEERING & SCIENCE NO. • ------

Right: Voyager images of Gany­ mede showed two basic types of GAll LEO GALLERY terrain: dark, heavily cratered regions; and bright regions of parallel grooves that are presum- ably much younger. This view of the boundary between a bright, grooved region called Philae Sulcus and the dark area called Galileo GOOD-BYE, CRUEL WORLD Regio superimposes a Galileo image, with a resolution of about 92 meters per pixel, on the best Biologists at MIT and proteins arc normally guiet, Voyager shot of the region, with a Cal tech have uncovered the but can be readily triggered chemical details of a mecha­ by death signals in such a way resolution of about 1.4 kilometers nism char cells use to commit that the cell digests itself. per pixe l. Surprisingly, Galileo sees su icide. Caicech President The three proteins are known no overall brightness difference D avid Baltimore's lab at That as CED-3, CED-4, and CED- Ocher Institute of Technology 9. None of these protei ns between the two landforms­ has succeeded in describing alone will kill ce ll s, the bright and dark patches occur in how roundworms known as research shows, but the three both. The bright patches tend to nematodes kill off unwanted interact in such a way that occur on north- and east-facing cells. The work is especially CED-4 can signal CED-3 to interesting, Baltimore says, begin the des truction process, slopes, which, in t his north polar because human beings have while CED-9 acts as an inhib­ region, get the least sun and hence very similar pCOCei IL'i CO chose itor to CED-4. The general the most frost. The grooves in the causing cell suicide in nema­ outline of this particular "bright" terrain, however, are very todes and, in fact, hi s lab pathway of apoprosis, as cell can oft en substitute human sui cide is technica lly ca ll ed, visible; the "dark" terrain proved proteins with the same was discovered by MIT to be gently rolling hills. results. "All cells contain professor Roben H orvi tz rhe machinery to commit some years ago, but the suicide," he says. "You ca n details have never been see chis in a w ide variety of understood until now, says events, such as a tadpole's Baltimore. The human resorption of its tail, local equivalent of these nematode ischemia in a stroke victim's protei ns are Apaf- i , which is brain, and tiss ue des truction very sim ilar to CED-4; BcI-2, after a hean attack. Cel l which is a homolog of CED- suicide is also one of the great 9; and mammalian cysteine procections against cancer." protease zymogens that are The hallmark of cancer is analogous to CED-3. uncontrollable cell prolifera­ The work, which was t ion, so precancerous cells, supported by the National whose regularory machinery Institutes of Health, appeared has been damaged in such a in the August 28 issue of the way that the cells might soon journal Science. In addition to begin to run· amok, freq uent­ Balt imore, the authors are ly Opt to immolate them­ Xiaolu Yang and Howard Y. selves for the good of the Chang. Yang is currently at organism. the University of Pennsy lva­ The mechanism involves nia; Chang, a PhD, is now three stable proteins fou nd finis hing his MD at H arvard. in nematode cells. These O-RT

6 [IIGIHfEllNG & HIEHC[ No.3 Above: In this banner hanging from the front porch of JPL's Space Flight Operations Facility, the Galileo team wished bon voyage to the Lab's newest spacecraft, Deep Space One, which was launched on October 24. Deep BEAM ME Up, Above: This "catcher's mitt" on Space One's mission is to test a dozen innovative technologies, including an SCOTTY Jupiter's icy moon Europa is about ion-propulsion system, for possible use on future spaceflights. 100 kilometers wide, as measured across the palm to the tip of the An incernational collabora­ thumb. The mitt appears to be tion of physicists has suc­ ceeded in the first true tele­ made of frozen slush and seems to portation of a quantum state. bulge up from the adjacent In the O ctober 23 iss ue of surface, which is bent downward Sci",,,. H. Jeff Kimble. and cracked, most visibly along the Va lentine Professor and pro­ fessor of physics, and col­ lower left (southwest) side. This leagues reported transporting feature may be the result of an a quantum state of l.igh[ from upwelling of viscous, icy "lava," or one side of an optical bench perhaps even liquid water, melting to the other without travers­ Above: The Great Red Spot is Jupiter's best-known storm, but several ing any physical medium in its way to the surface from the between. Instead of actually smaller ones are noteworthy, too. For the last half century, scientists have ocean that is hypothesized to exist propagating the light beam been watching three white ovals, each about two-thirds Earth's diameter, below the surface. across the bench, [he physi­ that lie to the south of the Red Spot. In early October, two of the ovals cists exploited a phenomenon merged to form this new Earth-sized storm, second in power in the solar known as "quannun en­ tang lement," the quintessen­ system only to the Red Spot itself. Unfortunately, the collision itself tial ingredient in the emerg­ happened on jupiter's night side, out of view. ing fi eld of quantum informa­ tion science. Teleporration of Left: Wait a minute-this isn't this kind was first proposed Europa! As a matter of fact, it's theoretically by IBM scientist Charles H. Bennett and the Monterey 8ay Aquarium, where colleagues in 1993. a probe that may one day look for The Cal tech experiment is life on Europa took its maiden the first quantum teleporta­ voyage in August. Then, in Octo­ tion to be performed with a high degree of fidelity, which ber, JPL scientists lowered the describes how well a receiver, probe into Lo'ihi, an undersea "Bob," can reproduce quan­ volcano orr the Big Island of tum states from a sender, "Alice." (Quantum teleporta­ Hawaii. Next comes Antarctica's Above: This composite view of Jupiter's ring system, taken when Galileo was tion was recently announced Lake Vostok, .wh ich is (overed with in the planet's shadow and in the plane of the rings, proved that the very by tWO independent labs in ice four kilometers thick. The fine dust that forms the rings comes from comets and other things hitting Europe, bur neither experi­ probe currently carries a camera four of Jupiter's moons-the moons' orbital distances are shown in yellow. ment achieved a fidelity sufficient to unambiguously and a thermometer, with chemical Rj stands for Jupiter's radius (71,398 ki lometers). And the "gossamer" demonstrate the use of and spectroscopic instruments rings' thickness, marked by the red and green lines, coincides with the quantum entanglement under development for next year. deviation of Amalthea's and Thebe's orbits from the rings' plane. between Alice and Bob.)

ENGINEERING & SCIENCE NO. 7 The work uses an exotic form his half of the beam Bob's service to teleport conceived the scheme. "Our form of light known as a into one that mimics Alice's quantum States without work is an important step "squeezed vacuum," whic h is second beam, resurrecting at revealing these states to toward the realization of split in such a way thar Alice a distance the origi nal , the company. And Victor networks fo r distribut ing and Bob each receive a beam unknown quantum state. can independently assess the quantum information-a that i:; the quantum mechani­ A unique feature of the qual i ty of the service Alice kind of 'quantum Internet.'" cal "twin" of the other. [See experiment is a thi rd parry, and Bob provide." In recent years, Kimble's E&S, Summer 1993.) The "Vi ctor," who verifies various The experiment shows group has worked on several photons in these beams share aspects of the protocol per­ that the strange connections aspects of quantum comput­ information that has no inde­ formed by Alice and Bob. that exist between entities ing, including showing that pendent existence in either It is Victor who generates in the quantum realm can be individual photons ca n beam alone. Because of the and sends an input to Alice employed for tasks that have interact CO form a quantum H eisenberg Uncertai nty for teleporration, and who no counterpart in the "classi­ logic gate-the basic element Principle, this information­ afterward inspects the output cal," macroscopic world of a microprocessor chip. the numeric value of a quan­ from Bob to judge its fidelity known to our senses. "The Kimble's work suggests rhat tum variable such as the with the original input. "It's distance was only a meter, the quantum nature of light amplitude of the beam, for sort of like having a quantum but the scheme would work may someday be exploi ted for example-is indeterminate telephone company managed JUSt as well over much larger the quantum processi ng and until it is meas rned. The act by Alice and Bob,'· says d istances," says coauthor distribution of information of measurement, however, Kimble. "Having opened an Samuel Braunstein (MS '85 , to accomplish tasks otherwise "freezes" the variable simulta­ account with an agteed upon PhD ·88), of [he University impossible for classical neously in both beams. Thus, protocol, a customer (in this of Wales in Bangor, United syscems. D -RT through quantum entangle­ case, Victor) uses Alice and Kingdom, who, with Kimble, ment, the act of measurement in one place can influence the quantum state of light in another. Such beams-cal1ed EPR beams for the Einstei n­ Podolsky-Rosen paradox of 1935, which conjectured their existence-are among the strangest predictions of METHUSELAH REVISITED quaonun mechanics. In prin­ ciple, it would be possible for a measurement on an EPR beam at Cal tech to alter the Cal tech biologists have dis­ tali y, the fruit flies have an mated over many generations quantum state of the beam 's covered a gene that increases increase in life span," says to produce a strain of flies twin at MIT-a spooky the life span of fruit flies by Benzer, who has won the that lived longer. In selective "action at a distance" that one-third when mutated. Crafoord Prize of [he Royal breeding, eve n though the led Einstein to reject the Dubbed "[he Me[huselah Swedish Academy of Sc iences effeCts are obviously genetic idea that quantum mechanics gene," after the Biblical for hi s work on the relation of in nature, it is difficult to might be a fundamental figure who lived for 969 genes co behavior in the fru it pinpoint which genes are physical law. years, the gene's discovery fly Drosophila 1Ilelanogaster. responsible. In this case, once Alice was announced in the "If we take the mutation out In contrast, once an indi­ and Bob have received their October 30 issue of the again, the life span goes back vidual gene is identified as spatiall y separate, bur quan­ journal Science. According to to normal." important to aging, the gene tum-entangled halves of the the authors, postdocs Yi -J yun This is not the first time can be cloned by molecular EPR beam, Alice performs Lin and Laurent Seroude, and a gene has been found that methods and its specific ce rtain measurements on a Seymour Benzer, the Boswell affects an organism's life span, function studied, implying second light beam she wishes Ptofessor of Neuroscience, Benzer says, explaining that that it eventuall y might be CO [eleporr and on her half of Emeritus, the research the roundworm Caenorhabditis used to manipulate rhe aging [he EPR beam. This demoys strengthens the view that elegans can also be given a process. "Very often indeed, the second beam, but she such a gene or genes might longer life in [he lab through fruit fl y genes have human sends both sets of measure­ also be found in humans. genetic manipulation. Nor homologues," Benzer says. ments to Bob via a non­ The work also lends addi­ is this the first time that "The basic idea of our quantum means (over an tional credence to the view scientists have demonstrated research is to use the fr uit fly ordinary wi re, for example). tbat the wear and tear of that the life spans of fruit fli es as a model system and look Si nce Bob's half of [he EPR aging can be exacerbated by are genetically determined. for human equivalents." beam is now "frozen" in the molecular stresses, such as However, the earlier work "Now it's inescapable that state in which Alice measured tissue-damaging free radicals. on Drosophila focused on aging is regulated deliber­ her half, he can use the co tre­ "If we mutate the gene, selective breeding- indi­ ately by genes," said UC San lations she sent him to rrans- which we can do experimen- vidual long-lived fli es were Francisco molecul ar geneticist

8 ENGINEERING & SCI ENCE NO. Above: Who says only undergrads can pull a good prank? The spiffy teakwood furniture outside the Red Door inspired some grad Above: The Winnett Student Center got a major face-lift over the summer. students, who, with help from The new exterior harmonizes with the Andalusian look of other buildings Blacker, Ricketts, and Fleming along the Olive Walk. The remodeled interior features a bookstore twice Houses, one night replaced the the size of the old one, a computer store called Caltech Wired, and a chairs with 56 toilets from campus much-expanded Red Door Cafe. storage-the Housing Office having just finished rehabbing a bunch of bathrooms. Top, right: (from left) Custodians Bill Schouten and Efrain Hernandez and elevator mechanic Moty Zahavi take the porcelain seats in stride, as it were. Bottom, right: Kate Finigan Cynrhia Kenyo n in a Science speculate that it may code manages the Red Door. news brief describing the for a protei n that is pare Benze t group's results. "Since of a signaling pathway that it happens in boch worms and controls how well cells deal fruit fli es , you have to be with stress . This would crazy noc to think it won't explain the fact that the happen in vercebrates. " flies with a Methuselah gene The Cal[ech ream si ngled mutation can better with­ WE'RE NUMBER ONE! AND NUMBER Two, AND our rh e Methuselah gene by stand such external s([esses NUMBER THREE • •• manipulating a small, as food deprivation, excessive cransposable piece of DNA heat, and exposure to oxida­ that can cause mutations at tive damage. the gene in which it lands. Kenyon, who identified The magazine Science \'(/atch peri od ically ranks universities by The researchers then tracked the daf-2 gene that incteases their scientific "impact," as meas ured by how often their papers the mutated fli es to their longevity in roundworms, are ci ted in other people's papers, which shows what people in natural deaths. DrosophiLa said in the news review that the field think is important work. In the most recenr rankings, normally live about 60 to rhe Benzer lab's res ults give Caitech neuroscience was # 1 in its fi eld in the nation, chemistry 80 days, bur [he fli es wi[h a experts in rhe field another and materials science were #2; economics #3; computer science, mutation at the Methuselah gene co work with. "Now," geosciences, and physics #4; astrophysics #5; biology/biocJ:1emis­ gene lived more than 100 she says, "we have another try and mathemarics #6; and engi neering #7. days. These flies were also experimental sys rem co. Jr 's not surprising that the physical sciences placed well, but better able to resist various invesrigate" for understand­ for the economics prog ram-which followed the University of types of stress that can cause ing how a gene or several Chicago and That Other Institute of Technology, borh of which aging in fli es or kill them genes can affect an organism's have substantial economics departments- to score so highly is olltright. From the identified life span. CI- RT quite a tribute. Especially consid ering chat the prog ram is less sequence of the Methuselah than 25 years old, and even raday has onl y a dozen faculty gene DNA, [he biologises members and a handful of grad students and postdocs. 0

ENGINEE RIN G & SCI EN CE NO. • Nobelist Entangled in Fraud Case President of Rockefeller U. Retracts Scientific Paper Resigns as Head ofRockefeller U. That NIH Office Says Contains Elhricatcd Data r

cry ~ criuus 'he scroll)-

Nobel Scientist Apologizes to Whistle-Blower Co-Author of Flawed Study Concedes His Oversight Was /nadequat,

As.soc:Ut!d Press nology, where O'Toole was her re­ Office of Scientific Integrity cc David Baltimore, a Nobel laure. search associate. . eluded that Imanishi-Kari "repe" ate who was co-author of a study When O'Toole cha llenged the edJy presented false and misleadi: that prompted a prolonged inves­ accuracy of data reported by he. information" and made statemen tigation of possible scientirtc mis­ boss in the article in Cell, there she knew to be fa lse. conduct. h~s apologized to the . were investigations first at MIT Baltimore responded then that l ' young scientist who prompted the and later at Tufts University. findings raised "very serious que investigation and admitted that he where Imanishi-Kari now works. tions," He withdrew the Cell paper. made mistakes in defending the The university investigations un­ flawed work of a co-author. In his 14-page statement to N: . covered only minor errors. O'Toole investigators. Baltimore acknol.­ Baltimore. now president of was fired. edged that "Cor too long- he had d Rockefe, "-.ll er University in New York, When NIH and the House Energy - -...... - , ...... - .... '" ten~~ ..t .h~ . now ~~s~.ed~~.~~. e.~p~.r_._

10 ENGINEERING & I(IEN(E NO. Scientific Fraud and Misconduct In• American Political Culture: Reflections on the Baltimore Case

by D an i el J. Kevles The so-called Baltimore case fi rst surfaced pub­ having defended Imanishi-Kari . As the result of licly in 1988, but it had been simmering in part of an appeal, however, inJune 1996 Imanishi-Kari the biomedical research community for almost cwo was exonerated on all charges of fraud that had years. The case originated in May 1986 with a been leveled against her, and Baltimore was widely young scientist named Margot O'Toole. She was recogni zed for the courage he had shown-and for a postdoctoral fellow in the laboratory of Thereza the costs he had borne-in resolutely defending Imanishi-Kari, a ce llular immunologist then on her for a decade. the faculty at MIT. O'Toole found seriolls fault During the same period, several other cases of with a paper chac Imanishi-Kari and five co­ scientific fraud and misconduct achieved compa­ authors, including David B a l [imoL"t~, had pub­ rable salience, if not so prolonged a life . And in lished in the journal CelL in April of that year. almost everyone of them, the defendants were O'Toole protested chac the paper's cencral claim ultimately found not g uilty. Eac h had been was nO[ supported by rhe raw data. Investigations victimized by procedural flaws that denied them of her contentions at Tufts Medical School and at elemental rights of due process. Taken together, the scientific fraud cases of the late 1980s and early 19905 are puzzling, fot twO reasons. First, W hy d id scj~ntjfic fraud become such a sal ient issue in American political many observers say, and I think they are right, that scientific fraud is rare. Second, at the time culture during recent years ? scientific fraud as such compelled little if any attention in other scienrifically vital nations. Why did scientific fraud become such a salient MIT concluded that the paper suffered only from issue in American political culture during recent minor, inconsequential errors and that O'Toole's years? quarrels with it amounted to a sciencific dispute One answer is that scientific fraud and miscon­ that could only be resolved by further research. duct were made and sustained as a public issue What kept the case simmering was the zealous primarily by Congress man John Dingell. DingeU pursuit of it by twO scientists at the National did not stop with the 1988 hearing. He held Institutes of H ealth, Walter Stewart and Ned additional hearings on scientifi c fraud, focusing Feder, and they reported it to the staff of Con­ on the case oflmanishi-Kari, in 1989, 1990, and gressman John DingelJ, a Michigan Democrat. 1991. These hearings occurred with sllch regular­ Dingell in turn brought it to the public's atten­ ity in sllccessive springs that Imanishi-Kari's tion in April 1988, when it was featuted in hear­ lawyer got mixed feelings about the blooming of ings that he called "Fraud in N IH Grant Pro­ the cherry blossoms in Washington; their opening grams." In January 1989, a panel appointed by meant it was time to testify, yet again. N IH to investigate the matter cleared Iman ishi­ Another answer is the media. The Baltimore Kari of fraud, but later that year the new Office of Case, which both advanced attention to the issue Scientific Integrity reopened the investigation. In of misconduct and was taken to be exemplary of 1991 the OSI found Imanishi-Kari tentatively it, provided a delicious target for the press. The g uilty of fraud, a conclusion that was reiterated whistle-blower, Margot O'Toole, was a compel­ by its renamed successor, the Office of Research ling, highly atticulate figure, a woman who, as Integrity. In 1991, Baltimore was forced to resign one of Ding ell's staffers put it, just "reeked with the presidency of Rockefeller University for integrity" O'Toole's dogged insistence that she

ENGINEERING & i(I ENCE NO. " Nobelist Entamr1ed in Fraud Case Resigns as Head ofRockefeller U.

1 r I' IP J I ~

d Il IIIH'II It! v II d'l It Onl( I, Ii ;,'1 XII 111, ,I r j 1I 1 t 11 I I" I\'IHI 1\\'ll!V J IV lind, I

I" It I I! '\ I 1,,_,1 I, \)11';11(' I \tI Iii II I " I r I ,tll! I lid I' 1101111 III IIf Ill, IIIIl"lllIl";\1 lit II \1(' 1111111\ I \ ,,," l'llt t 11IIt t ,Ill I ~ It h IIIIV('I" .tIlt' I J It III! 1 11111.l1f' ul J'"1111111 t\' hv tl \('! '1"n!1/1! ..ta ... ;;;;i ~ , In\! , IHU II II "'\' II' 'lAo III It

n I fo.., II \ I t, r I ( I

June 23, 1996: Thereza Imanishi-Kari and David Baltimore celebrate her was right and Imanishi-Kari was wrong lay at the and loan scandals, evidence of waste and profiteer­ exoneration on all charges heart of the affair. In the media, she became a ing in the huge defense buildup, the sale of politi­ two days earlier, A year symbol of the heroic young scientist who takes a cal favors in episodes such as Absca m, the indul­ stand against the system and prevails over power­ gence in influence peddling, and the Iran/Contra later Imanishi-Kari was ful figures like David Baltimore. In a column controversy. Corruption in high places has been a associate professor with in Time magazine, the commentator Barbara constant motif of Ameri ca n life, but some eras tenure at Tufts University, Ehrenreich captured the overall take of most of the have been worse than others. The '80s will likely and Baltimore had been media. "Baltimore pooh-poohed O'Toole's be recorded in the history books as among the evidence and stood by while she lost her job. worst because, as in the era of Ulysses S. Grant, named president of Then, as the feds closed in, he launched a bold, the corruptions of public life seemed to echo Caltech. misguided defense of the sanctity of science." She loudly the Bonfire-ofthe-Val1ities values- greed and added, "What he lost sight of, in the smugness of careerism often coupled to lying and deception­ success, is that truth is no respecter of hierarchy or that were so manifest in the private sphere, in­ fame, It can come oue of the mouths of mere cluding the regions that touch on public interests. underlings, like the valiant O'Toole." This is nor to indict the '80s but to place rhe Yet neither John Dingell nor the media would emergence of the fraud and misconduct issue in have gorten anywhere with the fraud issue if it its larger historical frame. That issue was, in fact, hadn't resonated with broader concerns in Ameri ­ first broached publicly in the late 1970s in several ca n life. The fact of the matter is that in the post-Watergate-flavored articles in Science maga­ 1980s it struck a chord in recent developments in zine by Nicholas Wade and William Broad, who American political culture and the relationship of collected their case studies into a book that was science-in particular, biomedical science-to published in 1982 under the title Betrayers of the those developments. TrlJth. And the issue was not introduced into Congress by John Dingell. It was first explored * * * on Capirol Hill in April 1981, in hea rings titled "Fraud in Biomedica l Research," by then-Repre­ Remember the '80s? At the opening of the sentative Al Gore, the Tennessee Democrat. decade, some of the most powerful forces operat­ Most of the Gore hearings and most of Broad ing in American political cu lture were energized and Wade's book were devoted to tales of indi­ in the legacy of Watergate and the war in Viet­ vidual cases of fraud and misconduct. Certain nam. Both had generated a deep distrust of public common themes ran through rhem: whistle­ and governmental authority and institutions. blowers were not listened to and some suffered That distrust by no means diminished during the retaliation; academic institurions tended to be 1980s. Irs continuation was fueled by the savings desultory in their response to whistle-blowing

11 ENGINEEKIMG & SCIENCE NO. ment intrusion into the practice of research. Congressman Gore, a sturdy moral liberal, had gotten involved with policy making for recombi­ nant DNA when he entered the House, in 1977. He held that the corruption of fraud raised serious questions about the ability of scientists to deal reliably with the "ethical judgments" that now As an ally of business and a ward of government, science, too, was vulnerable confronted biological science "in great magnitude" to suspicion, and the biomedical sciences were especially so. During the era of -by which he meant genetic engineering, in­ cluding the "spectre of cloning." At the 1981 Vietnam and Watergate, they had prospered steadily, more than did the hearings, Congressman Gore emphasized the im­ portance of maintaining the American people's physical scjences, obtaining enotmous support from public and private sources. trust in science, especially given the billions of dollars they were providing ie. Gore said he sus­ pected that one reason for rhe persistence of scien­ tific misconduct was "the apparent reluctance of people high in the science field to take these matters very seriously." The principal sc ientific witnesses were not reassuring on the point. They were Philip Handler, the head of the National challenges and to be reluctant to jeopardize grants Academy of Sciences, and Donald Fredrickson, by acting on those challenges and, when they did the head of NIH. Handler acknowledged that the act on them, tended to deliver minimal punish­ scientific community, both at large and in its local ments, if any at all. Taken together, such out­ institutions, had "never adopted standardized pro­ comes appear to have prompted Gore, as well cedures of any kind to deal with these isolated as Broad and Wade, to pose questions about the events {of misconduct). We have no courts, no scientific enterprise in the U.S., questions that sets of courts, no understandings among ourselves resonated with the distrust of authority and dis­ as to how any ooe such incident shall be treated." satisfaction with the emerging culture of the '80s. Both H andler and Fredrickson implied that such As an ally of business and a ward of government, courts or understandings were unnecessary and science, too, was vulnerable to suspicion, and the might even be dangerous. They insisted that biomedical sciences were especially so. During the fraud in science was rare and that false claims were era of Vietnam and Watergate, they had prospered exposed by the scrutiny that scientists gave each steadily, more than did the physical sciences, ob­ other's work. Fredrickson contended that the raining enormous support from public and private perpetration of fraud shook scientists "to our very sources. But that period was also rhe time of the core": when fraud did occur, the penalties were wars over recombinant DNA. Recombinant DNA harsh, the equivalent of "excommunication" for opened the door to the transformation of organ­ the sinner. isms-plants, animals, and possibly even human The biomedical community's post-1970s super­ beings-at the core of their hereditary essence. sensitivity to government controls was evident in Many people, including many scientists, found the testimony. Fredrickson called it "frightening" these prospects unsettling. A number of biolo­ that the larger, lay culture might intrude "roughly" gists worried that recombinant micro-organisms upon scientific affairs while "failing to understand might threaten life or health or whole ecosystems. the requirements of the scientific method or the Some questioned the reconfiguring of life itself as fact that its own correctives are in place. " Al­ an act of hubris that would lead [Q unpredictable though Fredrickson acknowledged that NIH had and dangerous consequences. to be sensitive and responsive to changing public In the later 1970s, local and state governments perceptions of science, he held that the agency and the United States Congress geared up to "certainly cannot guarantee the behavior of scien­ legislate tough, mandatory restrictions on such tists, or certify the quality of their work through research. Norton Zinder, a biomedical scientist a whole system of independent analyses or fraud at Rockefeller University, denounced some of the squads, or even special statutes." recombinant regulatory bills in Congress for The Handler-Fredrickson defense prompted setting up "vast bureaucracies, cumbersome Congressman Robert Walker, a Republican mem­ licensing, harsh penalties and tedious reporting ber of the Gore subcommittee from Pennsylvania, procedures." By 1980, biomedical scientists had to find "a certain amount of arrogance" in "a lot of beaten back the threat of intrusive bureaucratic the testimony." He doubted that a policy of "self­ controls and obtained even a major relaxation of policing" was adequate for oversight of the rapidly the strict guidelines that NIH itself had initially burgeoning biomedical community. Alluding to imposed on recombinant research. Nevertheless, Abscam, he warned: "We in politics would like to the recombinant DNA wars left many of them think that the people who stuck $50,000 into skittish about anything that smacked of govern- their pockets at some townhouse here in Washing-

ENGINEERING & ICIENCE NO. " Rep. John Dingell of Detroit, whose House committee claimed jurisdiction over NIH, used his Subcommittee on Oversight and Investigation to pursue the Imanishi­ Kari case.

ton are an aberration in our profession, coo. It doesn't mean ... that we should not be conscious of the need co clean up problems of that kind. Walker predicted that if the press co ntinued to report these aberrations and nothi ng was done, the "credibility [of science) will decline pretty quickly and the public will think of everybody in science as they think of everybody in politics-that is, as somehow a little crooked." In the co urse of revising the health services act in 1985, Congress required institutions expecting to receive grants on behalf of their faculty to establish administrative processes for dealing with reports of fraud in biomedical research sponsored by the Public Health Service, the parent agency of NIH. The PHS promulgated the necessary specific regulations in June 1986, bur in the meantime, American universities had been dilatory in establishing reliable local proce­ dures. By the time John Dingell began to take an interest in the matter, in 1988, seven years after the Gore hearings, NIH itself had gotten around to assigni ng about 1.5 staff people to oversee the matter in all its grantee instirutions. By the late 1980sJohn Dingell was a hero to Margot O'Toole and Walter Stewart (middle) and Ned many Americans, especially those who opposed the decade's careering and corruption, To be sure, Feder, the two NIH staff scientists who brought O'Toole's Dingell, whose district is in the area of greater complaint to Dingell's attention, testify at the Dingell Detroit, fought hard co protect the auco in

14 ENGINEERING & SCIENCE NO. thing tbat moves, burns, or is sold," including Congressman Dennis Eckart, of Ohio, another securities markets, energy, railroads, tele­ member of Dingell's subcommittee, expUcitly communications, defense contractors, consumer identified the seeming trends in science with the protection-and the NIH. '80s culture of corruption, noting that defense Dingell, who is the son of a New Deal Demo­ contractors tell us they have systems to carch cratic congtessman, is shrewd and knowledgeable fraud, but the rest of us know those systems don't in the ways of Capitol Hill, and during the 1980s work. "What is at issue here, much in the same he used his committee chairmanship to counter way [as] within the defense industry, over at some of the chief injuries that careerism and greed NASA, over at the accounting profession, over at in and out of government were doing to the publ.ic savings and loans, are the adequacy of safeguards interest. He aggressively exploited the chairman­ which will give you and me and the public con­ ship of his committee's Subcommittee on Over­ fidence that waste, fraud, abuse, or misconduct do sight and Investigation to raise questions about not occur with taxpayers' dollars ." Dingell and his former Reagan aide Michael Deaver's influence subcommittee colleagues said they initiated the peddling; to skewer EPA administrator Anne hearings on fraud in N IH grant programs so as to Burford for selling environmental policy to the expose the ongoing flaws in the system and get the highest bidder; to expose the Pentagon for appar­ academic community to heal itself. ent extravagances Ii ke the $640 toilet seat; and to Dingell had a special way of doing such public reveal that General Dynamics deemed a dog's business, however. He was not known as a legis­ kennel fees a defense contracting expense worthy lative iniriaror, and he did not hold hearings of federal reimbursement. Civil libertarians, mainly to gather information for the purpose of especially those of a conservative bem, deplored drawing up bills. It is a longstanding tradition in Dingell's methods, including trial by press leak, Congress that committees use the instrument of intimidation of witnesses, and the pit-bull tactics hearings w get the attention of the executive of some of his huge staff. But many people tended branch and to p ressure it to change policy without to overlook the prosecutorial and persecutorial legislation. DingeU was a man of that tradition. means he used, partly because he was so powerful He used the hearing room to spotlight an issue by but also, 1 think, because in the circles of those out probing into someone, or some case, or some prac­ of sorts with the political culture of the 1980s, he tice exemplifying the matter. When he and his pursued the right enemies. staff began gearing up to take on the issue of But then the Dingell subcommittee took on scientific misconduct, they followed their normal scientific fraud. Whatever N IH and its grantee operating procedure, which was to seek out an universities might have done since the Gore ongoing high-profile case that would show the hearings to establish reasonable procedures for system still in need of repair. They found it in dealing with fraud and misconduct, Dingell and the challenge that Margot OToole had raised his staff believed, not without reason, that it was against MIT, Tufts, Thereza Imanishi-Kari, and flabby and inadequate. More cases of fraud had David Baltimore. Dingell's subcommittee pursued the matter relentlessly-by subpoena, by leak, and by bully­ ing. "We do not wear lace on our drawers as we conduct our investigations," Dingell told a re­ "We do not wear lace on our drawers as we conduct our investigations ," porter. ''I'm not paid ro be a nice guy. I'm paid to look after rhe public interest. Our purpose is Dingell told a reporter. ''I'm not paid to be a nice guy. I ' m paid to look simply to compel universities and scientisrs ro after the public interest." clean up their act and ro see to it that public money is properly spent." Ostensibly to that end, the D ingell subcommittee pursued the Imanishi­ Kari case implacably. surfaced in the press, along with accompanying The congressional criticism that began in April evidence of dilatory institutional responses. Two 1988 had, in the meantime, prompted N IH ro of Dingell's key staff members-Peter Stockton initiate reforms of the way it handled allegations and Bruce Chafin- tell of people coming to them of scienrific misconduct. In the fall of 1988, the with stories about scientific corruption going agency announced that it was contemplating unpunished and research institutions covering the creation of an office of scientific integrity and up to protect themselves. Chafin holds that invited comment on the scope and definition of the system fai led to deter wrongdoers or to take scientific misbehavior. Contemplation turned into whistle-blowers seriously. In 1988, Dingell action in the face of moves on Capitol H ill that declared that he was "shocked ro find our that the winter, including a draft bill from John Dingell's National Institutes of Health relies completely subcommittee, to legislate the creation of such an upon the universities to investigare themselves. office wirh powers that would include random Apparently here we have the possibility of the fox audits of the notebooks of N IH grantees. James actively investigating the chicken coop." Wyngaarden, then head of NIH, worried that

ENGINEERING & SCIENCE NO. IS O'Toole said several times that she deserved no special credit for her coura-

geous whistle-blowing, that she did only what an honest sc ie ntist should do,

But at other times she engaged in a kind of self-fashioning that resonated with

the SOrt of heroes that many people hungered for in the 1980s.

congressional aCtion would give control of mis­ NIH "to deal with cases of questioned science." conduct cases to an inspector general, a lawyer NIH officials were concerned that the agency not who knew nothing about science. "We needed appear to be remiss. "We were taking a lot of hits a preemptive strike," he said later. On March 8, from Dingell on this with the public," Kimes 1989, NIH created an Office ofScienrific Integ­ notes. "Wyngaarden was extremely worried about rity on its own. OSI would monitor inquities at the politics of it because those are things that its grantee instirutions, conduct its own investiga­ could impact all of NIH. It was a very uncomfort­ tions when necessary, and be run by scientists. able situation." Wyngaarden picked Brian Kimes to head the OSI had nor issued a set of rules and procedures. office temporarily, until a regular director could Kimes says that the staff was so busy that they had be found. Kimes is a biochemist by training who to make them up as they went along. Kimes loves science and the NIH, where he had been a himself wanted OSI investigations to be open, research administrator in the National Cancer In­ with everyone having access to all the data and stiwre for almost 15 years. He took the job of testimony. In August, he assured Baltimore and setting up OS1 relucmndy, telling Wyngaarden Imanishi-Kari that once the allegations were com­ that he would stay only until November 1, and pletely formulated, they would be made available that "it was not my ambition to send scientists up to "all subjects of the investigation," together with the river." What could send scientists up the river all related evidence. He told Imanishi Kari's was spelled out in the definition of scientific mis­ lawyer, Bruce Singal, that Imanishi-Kari would conduct that the Public Health Service issued in have "every opportunity to see and comment" on midsummer. It included "fabrication, falsifica­ all the relevant information and that she could be tion, plagiarism." It also included "other practices confident that OSI's inquiry would be "conducted that seriously deviate from those that ate com­ independently" of the subcommittee's. Dingell, monlyaccepted within the scientific commu­ however, was reportedly very upset by the sce­ nity"-a feature that a number of scientists had nario. To Kimes, it was evident that Congressman objected to, no doubt fearing that it raised "serious Dingell "did not want the process to be open; he risks of undue pressure for scientific conformity," wanted us to investigate." as an official in the Office of Management and Toward the end of October, OSI revised its plans Budget had put it when the proposition first came for handling the allegations against Imanishi-Kari. up. Many scientists were likely relieved, however, They would be presented to OSI and its scientific by the Public Health Service's stipulation that panel, not to the coauthors. Both would evaluate misconduct "does not include honest error or the evidence she provided and then conduct an honest differences in interpretations or judgments investigation, talking with the other principals of data." in the dispute. The new dispensation meant that The small OSI staff was burdened with a Imanishi-Kari would be given only indirect access backlog of 80 to 100 cases. "We were over our to the allegations and no access at all to the ma­ heads ... ," Kimes says. "We had to figure our terial evidence or testimony against her. where the political issues were and where the Imanishi-Kari was, to all intents and purposes, scientific issues were. " But they knew that one prevented from mounting a genuine defense. The case dominated all the rest. Congressman Dingell OSI combined the duties of investigator, prosecu­ had told Wyngaarden that he regarded the matter tor, judge, and jury and pursued them in the man­ of Imanishi-Kari as a "crucial test" of the ability of ner of the Star Chamber. During the investiga-

I. ENGINEERING & HIEN{E NO. tion, it did not provide the accused a list of speci­ fic charges. The charges "are revealed to you as time goes on," a lawyer familiar with the agency's ways remarked. "It is as if there is an indictment you can't see." Both during and after the investi­ gation, lmanishi-Kari was denied the right to see the evidence, the right to learn what witnesses said, and the right to cross-examine them. The burden of proof was on the accused rather than on the accuser. The N IH's Office of Scientific Integrity was as Orwellian in procedure as it was III name. In early 1991, OSl put its tentative conclusion that Imanishi-Kari was guilty of fra ud into a draft repoft. The report, though strictly confidential, was leaked to many newspapers, provoking a flood of media response. Analysis of it, along with the pre-leak representation and commentary on the case, reveals just how integral belief in scientific fraud and misconduct had become to the post­ Watergate political culture of the 1980s. The Margot O'Toole trope of Watergate itself ran through what was said by influential representatives of the national press, such as The New York Times, which savaged Baltimore for "stonewalling" and, likening the long affair to a huge cover-up, declared that "the Baltimore case started with apparent fraud by a sing le scientist and soon led to a widespread denial of wrongdoing by almost everyone in a position to right the wrong." The report also happened to be leaked at the same time that the university overhead scandals were breaking, particularly the alleged ripping off of the government by Stanford University. A number of observers melded the Baltimore case tOgether with misuse of public funds. Like a reporter for Chemical and Engineering News, they held that "a major problem is that universities today are not only knowledge centers but also financial conglomerates," with all the " incentives to corner-cutting behavior that mark such enterprises. But no single person did more to tie the case to the features of post-Watergate political culture than Margot O'Toole-not by design bur simply by refracting it through the language and suspi­ O'Toole said several times that she deserved no cions of the culture of Watergate, which came special credit fot her coutageous whistle-blowing, naturally to het, and by having been porttayed as a that she did only what an honest scientist should victim of powetful scientific and institutional do. But at other times she engaged in a kind interests by Congressman Dingell, his staff, and of self-fashioning that resonated with the sort the press. "Here was someone chewed up by the of heroes that many people hungered for in the system," Bruce Chafin says . "Here was a perfect 1980s. In an interview with a sympathetic re­ object lesson. " O'Toole herself indicted Tufts, porter for Mirabella, a prominent women's maga­ MIT, and the Cell paper's coauthors for responding zine of fashion and chic, she contrasted herself to to her complaints with "the knee-jerk reaction of David Weaver, who was another of the coauthors cover-up" and with displaying "contempt for the on the Cell paper when he was a postdoc with labor of people trying to tepeat the work" of the Baltimore: "We both could have made the choice experiment. In an article in the Chronicle 0/ Higher to keep quiet for the sake of our own careers. He Ed,tCation, she claimed that the Tufts and MIT did and I didn't .... The science was more impor­ investigators "told me the corrections I proposed tant to me than the career and ... the career was [to the paper] were 'out of the question' because mote important to him than the science." they could adversely affect the authors' careers and Parts of the media, including some with enor­ financial support." mous influence in scientific affairs, were willing

ENGINtERING & HIEN(E NO. 17 and enthusiastic collaborators in the representation points of his piece, which was that the investiga­ of the case as exemplary of a scientific Watergate. tion was desig ned to save the reputation of David Some could almost be termed collaborators with Baltimore at the expense of Margot O'Toole. Congressman Dingell. DingeU's subcommittee Baltimore could take care of himself. Imanish.i­ shrewdly leaked documents, many of them con­ Kari was far less able to defend herself against a fidential, (0 favored reporters, who published the fun damental point that Hilts reported about the committee's version of events, and published them Cell paper .in a profile of her in the Times on June uncritically. Donald Kennedy, the former presi­ 4, 1991. To quote from his article: "The central dent of Stanford who was uncritically bnualized claim of the paper depends on how many mice by the press during the overhead controversies at showed the unexpeCted antibody properties. But his university, recently observed, "Press accounts the statement in the paper that said this work was frequently refer to unnamed subcommittee mem­ done, she has admitted, was false. 'We did not do bers and quote fro m documents obtained before it,' she said." Hilts's story continued: "But in they were made public. W hen such material explanation she said a similar characterization was comes regularly from one side of a controversy, it done on orher mouse samples." amounrs to news management- and no respon­ The fact of the maner is that an enormous sible reporter should be captured in that way." amount of work was done by Imanishi-Kari that The capture is partly explained by the over- demonstrated that the mice had abnormal anti­ body properties. What was not done was an additional test-the isotyping of the antibodies "When such matetial comes regularly from one side of a controversy, generated by one set of mouse cells. That data would have added weight to the paper's central it amounts to news management-and no responsible reporter should be claim, but it was not decisive for it. a crucial point captured in that way." of the science that Hilts missed entirely. He also did not report that the statement about the iso­ typing had gorten into the paper inadvertently­ as the resu! t of a miscommunication between whelming enthusiasm of many journalists since Baltimore and Imanishi-Kari in the drafting of it Watergate for investigative reporting. Since the to the effect that the test had been done on that set 1960s, that trend has come to mark some science of cells when, in fact, it had been done, but on reporting, and it has created a genre of science another set. The bottom line is that, by not pay­ reporting that displays the strengths and weak­ ing attention to the science, Hilts's story misrepre­ nesses, including the often unfairly injurious sented Imanishi-Kari as having declared that the weaknesses, that are characteristic of investigative test was crucial to the experiment's central claim, reporting in general. I don't mean ro disparage which she had not and which it was nor. His story investigative reporting. The public interest has also gave the impression that in the published been advanced by it. Bur it would seem that paper she had asserted knowingly and deliberately investigative reporting in science, like investiga­ that she had performed the test. lmanishi-Kari tive reporting in general, demands getting the protested Hilts's misrepresentation in a letter to facts right. Equally important in this special the science editor of the Times, but she got neither genre, it demands getting the science right, roo. an apology nor a correction. Even though the leaked draft report was just that-a draft- Baltimore and Imanishi-Kari were * * * tried and conviCted in the press. In this affair- as likely in other high-profile cases of scientific mis­ Recall that it was the legacy of the war over conduct-influential reporters often got both the recombinant DNA that helped prompt the emer­ nonscientific and the scientific facts wrong. I will gence of fraud as an issue in American political cite just one example, Philip Hilts of The New culture. Before that, people had tended to think Yr)rk Times, but it is a salient one, because the Times that science was perhaps the least corruptible in­ is arguably the most influential paper in the coun­ stitution in American life, certainly less corrupt­ try, and its science reporting is excellent on the ible than certain branches of the evangelical whole. Donald Kennedy has indicted Hilts for churches. One of the things that prompted Al bias in the Baltimore case, pointing to an attack Gore to call his hearings in 1981 was the evidence, anicle that Hilts published in The New Republic in mounting as he saw it, that science, in fact, might 1992 and alluding to his important coverage of be corruptible, and that the potential for fraud dle case in the newspaper. Hilts has responded raised serious questions about the ability of scien­ that no one has pointed to any specific bias in his tists to deal reliably with the fraught ethical ques­ coverage of the case. I don't know anything about tions that the advance of molecular biology was Philip Hilts's morives, but I do know that his raising. Philip Handler claimed that there was article in The New Republic was riddled with errors "absolutely no relationship" between fraud, even as about how the case was investigated at MIT, all of a minor problem, and what Gore called "the other them misrakes that reinforced one of the main ques tions involving ethical judgments which now

18 ENGINEERING & SCIE N CE NO. science in the American political culture of the late 20th century. As such, it is something that the scientific community cannot afford to ignore, no matter how rarely fraud may occur. U

Dan Kevles first wrote about this case for The New Yorker ill May 1996, arguing that Thereza lmanishi­ Kari was not guilty and that David Baltimore had been right to defend her. The following month she was eX01lerated on afl counts. He then expanded that article into a book, The Baltimore Case: A Trial of Politics, Science, and Character, which was published by \V, Scientific fraud is a kind of late-20th-cclltury morals charge, in a branch or \v, Norton early this fall. The book made the bestseller morality that large parts of the public perceive as central to human health and list of the Los Angeles Times, which, in a ,'eview, praised the book as a "brilliant, unsparing and human fate . metiCllloll.fly 1'e.rearched account of a controversy that helped reshape how scientific misconduct is handled in the United States." The New York Times cafled it "a splendid .ftudy of a major contemporary scientific

scandal. 1) confront science in great magnitude." But Gore Kevles, the J. O. andJllliette Koepfli Professor of the asked: "If the scientists who have been found in Httlllanities, i.r head of Caltech's program il1 Science, this recently reported outbreak of seemingly Ethics, and Public Policy. He has been a member of the unconnected events to be making these kinds of Caltech famlty since 1964, after receiving his AB and judgments with respect to arguably small matters, PhD degrees from Princeton, and is the altthor of several shouldn't we be concerned about the decisions books and ntllllerOtt.f articles and reviewJ. they are making elsewhere?" Handler said, no we should not, whereupon Gore retorted, "Well, I disagree strongly. " My guess is that, at least tacitly, many people continue to disagree strongly. Scientists have long resembled a secular priesthood in American culture...----.originally as mediators between man and God's universe, but now as explorers, interpreters, and com manders of nature's powers. It should nor be surprising that many people display high trust and confidence in science, which they do. But it should not be surprising either that many develop comparably high suspicion of it if evidence--or a claim of evidence-emerges that its celebrated concern for truth is violated, particularly among biomedical scientists. To be sure, much of the concern with scientific fraud stems from the fact that biomedical science is lavishly supported with public money. Never­ theless, the exposure of fraud in the biomedical sciences is tantamount to revelations of pederasty among priests perhaps more than of embezzling among bankers. Scientific fraud is a kind of late- 20th-century morals charge, in a branch of mo­ rality that large parts of the public perceive as central to human health and human fate. The fact that it's freighted with all that such charges connote-violation of a moral code and exploita­ tion of innocents-helps explain why scientific fraud and misconduCt has become a powerful issue in the United States, a country that more than any The book can be ordered from the Caltech Bookstore for other expects its public shamans to behave con­ $29.95 pillS $5.00 for "hipping and handling. Send sistently with their dedared moral commitments. check or credit card number to Caltech Bookstore, Mail The public perception of both the behavior and Code 1-51, Pasadena, CA 91125. Ororderthebook powers of biomedical scientists may be exagger­ by phone at 1-800-514-2665, or by '-lIIaii at ated, but it is a fact of life for the fortunes of [email protected].

ENGINHRING & HIENCE NO. 19 10 ENGINEERING & HIEHE NO. "Have Method, Will Travel" by Douglas L. Smith

If it weren't fo r rhe oil in your engine, it would putational chemistry these days. (In fact, this seize up in just a few miles of freeway d ri ving. year's Nobel Prize in chemistry went to Walter Trust me-I know chis for a face, Bm if it weren't Kohn and John Pople fot developi ng the math­ for a dash of zinc dithiophosphate in that oil , the ematical and programming techniques that make engine would grind its camshaft and valve train computational chemistry possible.) W hat sets the down to nubbins within a few thousand miles. Goddard group apart is that it applies chemical This additive has been standard equipment in all theory to real , industrial problems-real enough high-quali ty motor oi ls si nce the late '405, says that industry underwrires half of Goddard's William A. Goddard III (PhD '64), the Ferkel research. Ptofessor of Chemistry and Applied Phys ics, but Most universi ty chemists prefer well-defined Chevron's refinery in nobody knew why it prevented wear. "People problems with sparse data (it's eas ier to get your Pascagoula. Mississippi. weren't that interested in why, because it always own data published that way), whereas industrial The United States worked. So why waste your money fig uring our chemists are often handed ill-defined problems how? Problem solved." The problem is rapidly wi th lots of data- sometimes several decades' consumed some 123 billion coming unsolved, however, because federal law worth. In the case of the chewed-up camshafts, gallons of gasoline in 1997, requires that by the year 2000, catalytic converters Chevron's fi les bulged with some 50 years of according to the American will have to last 200,000 miles. They don't las t engi ne-wear experiments. People would take an Petroleum Institute-at 20 that long now, because the phosphorus in the engine apart, measure all the clearances with a additive slowly poisons [he catalyst, Furthermore, micrometer, put everything back together again, miles per gallon, enough the additive slowly gees burned along with the oil, top off the oil , and run it on a tes t stand fo r 60 to drive the family minivan and the oxidized zinc becomes parr of the parcicu­ days. Then they'd tear the engine apart agai n, to Neptune and back 439 late as h that goes out of the tailpipe and into the compare the clearances to the previous measure­ smog. So the oil induscry is sc rambling to invent ments, change the oil , and do it all over again. times. Photo courtesy of a replacement-a phosphorus- and zi nc-free ash­ This isn't a very fast method, and jt's not cheap, Chevron Corporation. less inhibitor. Now, replacing something when either-a complete set of tests on one inhibitor you don't know how it works is a pretty good costs $150,000. The relative wear-resistance of trick, so in 1995, the fo lks at Chevron Research various additives had been ranked this way, and Technology Company in Richmond, Califor­ providing performance trends among similar nia called on Goddard, the director of the Materi­ molecules. als and Process Sim ul ation Center (MSC) at But now Chevron needed to know if some Caltech's Beckman Institute, to help them out. unknown, radically different molecule might be As the name implies, the center does simula­ better. In the search for a fast, cheap predictive tions of mare rials, chemistry, and biology, making method, Goddard recalls, "They tried pouting predictions before the experiment is actually done. oil over an iron ball and rubbing it. T hey tried It's the only chemistry lab at Cal tech where the spectroscopy; they tried lots of different things, video tubes outnumber the test tubes. The center but nothing correlated with the engine experi­ runs reactions in supercomputers, while collabo­ ments. So they turned to us, and for less per year rating experimentalists and engineers then per­ than the cOSt of one engi ne experiment, we came form the physical experimencs and check the up with a model they could use." results. Now, you can't throw a brick in a chemis­ The first thing the Goddard group did was step try department without hitting a theorist, and back and rake a fres h look at all the data, and try there are a goodish number of people doing COffi- to work out what might be going on "from the

ENGINEERING & SCIE NCE NO. 21 Above: The crystal structure of hematite predicted by the MSC's model. Iron atoms are rust-colored; oxygen atoms are red. Crystallographic data had shown that each iron atom is bound to six adjacent oxygen atoms, and that the bonds came in two lengths. It had been assumed that the bonds were all of roughly equal strength, but the model showed Above: A top view of an that the longer bonds (dashed lines) were fundamentally different in character, and engine on a test stand. significantly weaker, than the shorter bonds. This is why hematite flakes-the weak bonds The arrows point to the between the layers break more easily than the strong bonds within a layer. camshaft. Below: A single cam, close viewpoint of what's plausible to us as chemists." So postdoc Sunder Ramachandran, Baa-Liang up. The parallel scratches This step back before plunging in is vital, because Tsai (PhD '97), and Blanco developed a quantum­ on what should be a nice, what the induscrial collaborator thinks is the mechanical model of hematite from first prin­ shiny, smooth surface are problem may nor reaLly be the problem. The ciples, which, as faithful readers of E&S will know, theorists already had an inkling of the real prob­ means the dreaded Schrodinger equarion, which the wear marks. lem on this one, having just finished designing provides a complete mathematical description of some improved corrosion inhibitors for Chevron's every particle in any atom. Undergraduates have Petroleum Technology Center at La Habra, just 30 been solving the Schrodinger equation for the one­ miles from Calrech. Crude oil contains stuff that proton, one-electron hydrogen atom since the tends to adhere to and react with the inside surface 1920's, but as you add more particles to the of a pipe, rusting it away until it eventually begins system the number of terms in the calculation to leak, causing costly oil-well shutdowns. And it quickly spirals out of control. Recalls Goddard, turns out that keeping scuff from sticki ng to a "When I started in this business some 30 years surface has a lot more in common with keeping a ago, three atoms was difficult. We've made a lot surface from wearing away than you might think. of progress since then. Now it's hard to do more In the corrosion case, the Chevron chemists had than about 50 atoms or so." Fortunately, hematite assumed that the dissolved ions in the oil were the has a simple, repeating crystalline structure, so the instigators, and tried to trap or somehow disable investigators were able to model a unit cell, as the them. Nobody paid much attention to the prop­ basic rcpeating structure is called, with a f~w erties of the pipe itself, and when they did look dozen iron and oxygen atoms. The model hema­ at it, says MSC project manager Mario Blanco, tite formed two-dimensional layers, explaining the "People concentrated on modeling pure steel flaky nature of the real crystal, and the crystallo­ surfaces. We conjectured, and later found to be graphic coordinates of the atoms in the model correct, that hematite was the mechanistically matched those derived from the real thing by televant surface." Hematite is one of iron oxide's X-ray diffraction studies-two reassuring reality several guises, and a thin skin of it forms naturally checks. Once quantum mechanics had nailed on the pipe's interior-and on the interior surfaces down all the atoms' exact positions, the trio of an internal combustjon engine-as soon as of scientists, aided by Member of the Beckman there's a trace of moisture. In other words, unless Institute Siddarth Dasgupta, derived a mathemati­ you're working in the vacuum of outer space, the cal representation of a force field that numerically hematite layer will form even as the pipeline described every dimple and dome in the iron-oxide (or engine) is bei ng put together. surface, and predicted the force with which that

11 ENG I NEERING & HIENCE NO. The best inhibitOrs snapped info place like Lego b loc ks, paving the surface onc

molecule deep-forming a self-assembled monolayer, in the parlance of the

trade.

surface wou ld attran or repel any orher atOm anywhere in space. Such a force-field model, which is much sim pler compucationally than the original quantum-mechanical model, can be scaled up to perhaps a million aroms more than enough for this project. This million-atom level-called the nanoscale, because a cluster of a million atoms is about 50 nanometers in size-is where all the action is tOday, according to Goddard. "[ncegrated-circuit components ate starring co get down to the nano­ meter scale, so rhe applied physicists and electrical engineers are approaching it from above. And chemists are starring to manipulate structures Above: The best corrosion-inhibiting molecules look like tadpoles wearing party hats, with that assemble themselves into objects in that round heads and long, floppy tails. Carbon atoms are shown in gray, hydrogen atoms in size range, so chern ists and materials scientists white, and nitrogen atoms in blue. The nitrogen atom (arrowed) in the tadpole's chin is are approaching it from below. And molecular the part that actually sticks to the hematite surface. biologists-a rhinovirus, the common cold virus, has half a million atoms in its outer coat. We've Below: Atoms don't really have solid surfaces, but if you pretend that they do, it's easier to done simulations of thar. So it's really the frontier, see how they pack. As the tadpoles line up with the "grain" of the surface, their tails and the computational techniques and modeling overlap like shingles. The dashes are the nitrogen-iron bonds. methods that apply are nor very we ll understood. Below us is quantum mechanics, and above us are finite-element analysis and other standard engi­ neering methods. So it's open territory for theorists." With the force-field model of the hematite surface up and running, the next step was to pour water molecules and a candidate corrosion inhibi­ tor on it and see what wou ld happen. The best inhibitors snapped into place like Lego blocks, paving the surface one molecule deep-forming a self-assembled monolayer, in the parlance of the trade. The molecules resembled tadpoles, having round heads and floppy, oily tails 12 to 20 carbon atoms long. Each tadpole would plant itself face­ first onto an iron atom, shoving as ide the water molecules already nuzzling there. And as rank upon rank of radpoles lined up with the "grain" of the surface, their tai ls overlapped like shingles and created a waterproof roof that kept the corrosive ions at bay. "The self-assembled mono-

ENGINEE RING & I{IENCE NO . 13 Zinc dithiophosphate is more avian chan amphibian- the dithiophosphate

portion looks like an eagle in flight, with the zinc atom clutched between its

talons.

layers form an exquisite, geometrically regular pattern," says Blanco. "Molecules with the wrong shape don't fit the underlying structure and are not inhibi tOrs ac all." Cou ld the same mechanism be involved in wear inhibitors? The two types of molecules are vastly differenr. Z inc dithiophosphate is more avian than amphibian-the dithiophosphate portion looks like an eagle in flighc, with the zi nc atOm clutched between its talons. The wi ngs are stub­ by, branched three-carbon hydrocarbon units called isopropyls. So postdoc Shaoyi Jiang released a flock of dithiophosphate eagles into the hematite model, and confi rmed whac had only been suspect­ ed- that once the swooping eagle struck, it let go of the zinc acorn and sank its talons intO an Zinc dithiophosphate looks exposed iron atom instead. Says Goddard, "We like an eagle in night, with calculated the monolayer's cohesive energy, and the zinc atom (blue) found that it correlated inversely wich the cam clutched between its wear." In other words, the eagle that gripped the tightest protected the best. But Jiang discovered talons. Atoms of sulfur are that unlike the tadpoles' overlapping tails, the shown in yellow; phospho8 eagles' wings lay flat across the surface, wingtip rus is purple. Actually, touching wingtip. So he tried different wings on the same body, and behold! the isopropyl wings fit because a zinc atom has bes t. Smaller ones didn't cover the surface com­ four binding sites, there pletely, exposi ng some iro n atoms to wear. And are always two eagles on larger hydrocarbons got in each other's way, li ke These graceful trefoils, like a window in a Gothic cathedral, opposite sides of the zinc, fledglings gecring toO big for the nest. And JUSt appear when the hematite model is seen from a.bove. In as overgrown fl edglings tend to fall out, so did sort of like a face card in a the left view, only the very top layer-the surface atoms to chese dithiophosphates have a lower cohesive ener­ poker deck. If you tore gy. "So we did that the first yeat," Goddard says. which corrosive molecules (or corrosion inhibitors) can the card almost in half and "And Chevron was very pleased, because it was the bind-is shown. The dashed lines mark the "unit cells," or twisted it 90 degrees, that first time that they had anything that actuall y repeating units, of the crystal structure. Each successive correlated with the engine-wear experimenrs." is-the eagles are flying in In the contract's second year, Chevron suggested layer is offset by half a unit cell, so when a second layer is perpendicular directions. 13 new families of potential wear inhibi tors, and added (right), the trefoils turn to arabesques and the asked the Goddard group to predict how these cathedral becomes a mosque. A third layer (not shown) untried additives would work. (The MSC works converts the arabesques to hexagons. in three-year contracts, cancelable at the cnd of eac h year if the client doesn'r feel sufficienr progress has been made--to date, not one of

24 EN GINEEKING & SCIEN(E No.3 Sarah Sanders (BS '98), worki ng as a Summer Undergraduate Researc h Fellow (SURF), obliged, making the model much easier to use in the process. Unfortunately, many molec ules that work well in the model may be too expensive to be practical, so now, in the projeCt's final year, the theorists are collaborating with sy nthetic chemists at Chevron to try to pick compromise molecules. "We'll show them a molecule that we think will have really good properties," says Blanco. "And they'll say, 'No, we can't make that. But we can make this, which is kind of close--do YOLI think it will work?'" In rbe meantime, postdoc Yanhua Zhou and Tahir (agin, the MSC's director of simulation technology, are adding more derail to the model. Says Goddard, "We're applying pres­ sure and looking at the dynamics of shear as one hematite surface moves past another, just as in a rea l camshaft. We are learning how shear resis­ tance works. This is the first time that such dynamical studies have ever been done, thanks to some co mputational breakthroughs that Tahir has made." twO dozen or so contracts has ever been ca nceled .) The key to the group's creativity, says Blanco, Jiang ran all 13 f."milies through the model, "is the presence of other researchers doing unre­ How the isopropyl eagles and one family emerged as the leading co mender. lated work. Self-assembled monolayers-thio­ Instead of zinc and phosphorus, these birds were alkanes on gold-were a curiosity back when we stick to the hematite. In built around an environmentally friendly carbon­ started looking at the co rrosion problem, but Jim the figure at left, the nitrogen body. However, a little more resting Gerdy (PhD '96) was doing some work on them. bonds between the sulfur revealed that the isopropyl wings weren't goi ng A hallway meeting combined my knowledge of to fly any longer. Because rhe new body contained how surfactanrs work and Jim's ideas on monolay­ talons and the iron atoms more atoms, it sat up higher on the hematite, so ers in a way that appli ed to these projects. To are shown as dashed lines. longer, floppier chains of six or seven ca rbon atoms someone on the outside, things would appear a Note how the top two were needed to drape down and completely cover lot messier (and to some extent they still are), but layers of the hematite the surface. we could foclls on the essential features of these Jiang rated the 13 body types in his second end­ compounds and learn how to improve on them. surface are actually of-the- year presentation to the Chevron folks , and We call this process 'finessing the problem.'" distorted by the presence explained how the wings altered the performance. To that end, the group is home to about 20 grad of the inhibitor molecule. After the talk, he and Goddard lea rned that students and as many postdocs-roughly equal The figure at right shows Chevron had gone ahead and done some experi­ numbers of physicists, chemists, and materials ments already, based on preliminary results Jiang scientists, with a sprinkling of applied physicists, how the isopropyl wings had given rhem earl ier. The best-performing body biologists, and biochem ists thrown in for spice. cover the surface and in these experiments proved to be the one that the And if ever a project called for finesse, it was shield the iron atoms from model had predicted, and the wings behaved in another assignment from Chevron that the group real life the same way that they had in the model. tOok on at about the same t i me: helping to wring harm. "The boss of the whole operation was a couple of the most gasoline from a barrel of crude oil. feet off the ground," Goddard reca lls. "He rook Whether you're filling up your Ferrari or topping a risk to fund us , and we came back with a model off your Toyota, you expect what's going into your that ac tually made useful predictions. The tank to be pretty consistent. So does your· engine, Chevron group included several very good organic which is designed to burn gasoline with a specific chemists, but they hadn't been thinking about octane rating. But no two samples of crude oil are wear inhibitors as chemists because they had no completely ali ke. Even crude from a single well model fo r what propercies of the molecule were changes as the well gets pumped down, says Anil imporcant. So they'd try things similar to what Patel, lead planning engineer at Chevron Research they had already studied, instead of reasoning, and Tec hnology Company-the dregs are always 'Well, on this surface, this molecule is likely to thicker, as anyone who has ever drunk homemade do this, and it's likely to fit in this way.' So they chocolate milk through a straw knows. To try co asked us to put together a graphic interface that limit the variabiLity, most refineries were origi­ would let them look at our results on their own nally designed specificall y for the crude from a computers, see how the molecul es fit, and start given region, or even a particular oil field. But thinking about them as chemists." Blanco and in coday's globa l economy, the crude can come

ENGINEERING & SCIENCE trio. gasoline." And the higher rhe naphtha's oc tane The naphtha fraction of rating, the more of the low-octane stuff you can gasoline is a wild blend with it, and the more gasoline you get. The naphtha goes to the reformer, which is not a assortment of hydrocar­ school for wayward youths but a reactor that heats bons containing between the naphtha till it vaporizes. T he reformer then five and eight carbon blows the vapor through a catalyst that teartanges low-octane molecules into high-octane ones. atoms per molecule. some While in the caralyst, those 265 species partici­ of which are shown at left. pate in half a dozen diffe rent families of compet­ In these diagrams, each ing reactions. NO[ al l of the ourcomes are good­ line segment represents a a reaction that improves one molecule's oc tane rating may lessen another's . carbon-carbon bond, with Each spec ies undergoes eac h teaction at a a carbon atom and a full different rate, so the operator adjusts the reformer's complement of hydrogen settings to maximize the good reactions and mini­ atoms sitting at each mize the bad ones, based on a quick-and-dirty analysis of that batch of naphrha. Until recently, vertex and endpoint. the lab gave the operator four numbers: the per­ centage of paraffin s (linear and branched molecules from anywhere. So refineries are always resecci ng with no double bonds), olefins (the same mol­ their process conuols to maximize the amount of ecules, but with double bonds), naphthenes consistent product that can be made from this (circular molecules built around a five- or six­ inconstant raw material. Chevron wanted a model membered ring of carbon atoms), and aromatics; that would use an analysis of the feedstock to p lus a boiling-point curve and a spec ific-gravity predict how the controls should be set, which measurement that together provided a rough idea is not as easy as it sounds. To see why. we need of the size distri bution of the molecules. T he to take a quick look at how gasoline is made. operator then set rh e reactor's temperature, hydro­ Crude oil is a mishmash of molecules containing gen pressure, and feed rate based on gut instinct up to 60-something carbons, so the refinery dis­ and the accumulated wisdom of the plane, often tills the crude into fractions containing succes­ enshrined in hand-drawn graphs taped up in the sively larger molecules with successively higher control-room window. T hese performance curves boiling points. The so-called "naphtha" fraction, were worked out through trial and error, says with five to eight carbons per molec ule, becomes Patel. "It was strictly an empirical model. with no a major component of gasoline. Actually, gasol ine chemistry behind it. If the new fe ed was radically includes hydrocarbons containing up to 12 carbon different. they'd do a pilar-plant run on it first. atoms, because the distillat ion process is not 100 Bur if it was just a litcIe different, say maybe two percent efficient and a few heavier molecules sneak percent more paraffi ns than the las t batch. they'd in. In fact, because of the voracious demand. a lot say, 'Hey, I can handle this,' and JUSt go ahead and of the even heav ier stuff eventually becomes gasoline, toO, through a process called "cracking," which breaks big molecules down into smaller ones. More on that later. The naphtha fractio n includes 265 distinct species of hydrocarbons. There are linear mol­ p;n/l'"~, Dd>ydro

16 EH GlllEElIHG & H I EHCE HO. try it." But raday's inscrumental methods can give \ CH ==CH/ H H you tons of data instancly-for example, that this H H : : I I morning's naphtha has 6.3 percent unbranched ~ ~ paraffins with five carbon atoms, 4.7 percent Melal Meln l Mcud single-branched olefins with 12 carbons, 3.2 percent benzene, and so on. (The breakdown isn't This sample paraffin so detailed above 12 carbons, but that's another isomerization reaction story.) T here aren't enough windows in the shows how the reforming control room to POSt 265 graphs, and even if you had an infinite sheet of glass, there's precious little process works. The cata­ dara at that level of detail CO plot. Pilot-plant lyst actually uses two studies are expensive and time-consuming, so different components (the gathering process-control data hasn't kept pace with analytical advancements. Chevron estimated zigzag surface). mixed that it would cost several million dollars to do all together in a heteroge­ the experiments needed to find just 20 different neous mass. In the first rate constants over the three reactor parameters, step (above), a rheniuml says Goddard, "so they decided, for a small fraction of that cost, to see if we could predict platinum component strips these things from theory. And I told them there's a pair of hydrogen atoms about a 50-50 chance that we could do it." off of a paraffin, convert­ T he 265 species mentioned above don't begin co reflect the full complex ity of rhe problem. The ing it into an olefin, which catalyst rearranges a molec ule by converting it to then takes wing again. an ion, which you can think of as having a hole where something ought to be. Other parts of the molecule slide inco the hole to try to fill it, but this just moves the hole around. Eventually, the ion spits out a hydrogen ion, becoming a stable H. H. I I molecule with a double bond and detaching itself ~ ~ Acid Acid Acid Acid from the catalyst. The original ionized site can be anywhere along the ca rbon backbone, so the num­ Next, the olefin slams into the hydrogen-ion-studded ber of different ions increases with the length of the backbone. H ow the molecule morphs depends surface of a second, "acid" component. The olefin's double on where the hole is, adding another layer of bond picks up a proton, or hydrogen ion, converting the calculations. If you have up to eight carbon atoms hydrogen into a neutral atom and leaving the now-ionized in your feedstock, you have 540 different hydro­ olefin stuck to the catalyst by what used to be the other carbon species-including the intermediate ions and the byproducrs with fewer than five carbon end of the double bond. The ion flops around like a break atoms-that you have [ 0 keep track of in your dancer as other parts of the molecule try to fill the ionic calculations. (Even though the naphtha fraction hole. Eventually, a new double bond establishes itself and StartS at five catbons, you can't neglect the small er the rearranged olefin departs. molecules-some of the reaction paths cause the parent molecule to break off a ca rbon atom . If an eight-carbon molecule becomes a seven-carbon molecule, that's fine, but if a five-carbon molecule loses a carbon, it becomes butane-a gas, not gasoline.) These 540 species can undergo 883 H H possible reac tions. And the first set of products I I ~ ~ can hit the catalyst again downstream to undergo Mt;:lal MClal a second se t of reactions, and so on. So the first simplification was to limit the problem to molecules with six or fewer carbon And finally, the newly acorns. This brought the number of species 'down branched olefin can have to 8S, and the number of possible reactions down another run-in with the to lOS- much more manageable numbers. metallic component, Second, "We had to assume that the catalysts were perfeer," says Goddard. "That they didn't emerging as a paraffin get plugged up, and thar they could hydrogenate once again but with a and dehydrogenate, and proton ate and deproton­ higher octane rating. ate when they wanted to. Catalysis is our busi-

ENGINEERING & SCI ENCE NO. 27 0.7 r-- 0.6 !------V < 0.5 ~ OA / •8 < 0.' 8 I(!' 0.2 "- ....., 0.1 ~ ..... i--- 0 "- o 10 20 30 40 50 60 TIme (sec)

ness, and normally we would have looked in detail we'd need three years, and it took two." "Tt would at what goes on on rhe catalytic surface, but thar be cool to push the model fuerher," says Perry. would not have been practical here. We could not "But we've now moved on to modeling the crack­ Above: The pared-down have dealt with this unbelievably complicated set ing process, which is where the real action is, in reaction network (or six or of reactions if we'd had to worry about the surface collaboracion wich Michael Shippey [PhD '78J, coo." In the end, the catalyst was represented by a senior research chemist at Chevron. lnstead less carbon atoms. seven parameters: one for each of five types of ion of doing a few hundred species going up to eight Again, reactions with rearrangement- the "reaction families" mentioned catbon atoms, we're modeLing the reactions of double arrows can proceed earlier-that measured how much of a kick it took millions of species goi ng all the way up to 60 in either direction. The to get that rearrangement statted, and two that ca rbon atoms. The accollocing gets a little reflected the catalyst's abiliry to ionize rhe tricky, but the principles are the same." color coding matches the hydrocarbon at two different temperatures. These projects are typical of how the group graph above right, which Jason Perry (PhD '94) mapped che pared-down works. When an industrial partner signs on, shows how the model network of possible reactions and used quanrum Goddard hires a postdoc to work on the project mechanics and the ideali zed catalyst to calculate full-time, and assi gns a senior staff scientist at the evolves over time at a the rate constant for each link in the network, MSC to direct rhe effort. The Cal tech people and 0 temperature of 950 F and crearing a model that would predict rhe product the industry people meet monthly to update each a hydrogen pressure of 350 d istribmion for any six-carbon feedstock at any other on what's going on. Information flows in psig. or pounds per square co mbination of reactor settings. (Perry, a freelance both directions, and a cas ual comment by one quantum chemist-yes, there are such things!­ person can spark a revelation elsewhere (or dredge inch {gauge)-in other and president of First Principles Research, was up a forgotten tidbit of information). "But you've words, pounds per square a grad student of Goddard's.) The researchers gar to be eyeball [0 eyeball," says Goddacd. "If inch above normal checked these predictions against real data taken you don't have this SOrt of direct , continuous atmospheric pressure. at two temperarures, twO pressures, and three interaction, you don't get the rig ht information." elapsed times (the equivalent of flow rate-higher And sometimes the onl y way to get that informa­ flows mean the feedstock is in contact with the tion is to talk around the problem in a way that e­ catalyst for less time). Once the catalyst param­ mails and phone calls just don't do. Furthermore, eters had been adjusted to correctly predict the it's always the same industry people at each meet­ six-carbon products, rhe researchers successfully ing-an important point fo r project continuity expanded the model to predict product distribu­ and effi cient transfer of knowledge; a collaboration tions for up to eight carbon atoms. is only as good as the institutional memory of the As with the wear-inhibitOr project, the research­ collaborators. ers developed a graphic interface before giving the Some face-to-face interactions go much farther model to the engineers. "It's the simplest model than monthly meetings. The MSC iss ues a to use they've ever had," says Perry. "Yet it gives standing invitation to its industrial partners to the mosr detailed results. And Chevron uses lots come to Cai tech and work in its labs fo r a wh.ile. of different catalysts, bur this model could apply Yongchun Tang, a sen ior research chemist at to all of them with the appropriate tweaking-the Chevron 's La Habra fac il ity and leader of their seven parameters all have physical meaning, so it's molec ular modeling group, spent several months easy to adjust them." Patel is now us ing the mod­ at the MSC back around rhe turn of the decade, el for process design on the pilot-plant level. when the collaboration was fi rst getting off the "This was a real success," says Goddard. "We said ground. "We've been so successful in our model-

18 ENGINEER IN G & SCIENCE NO. will travel,'" Goddard resumes. '''Have method, will simulate.' We looked fo r whatever problems our method would be useful for. These days, ir's C8 Feed T=950F P",350psig CB Feed Te950F P.. 350psig Theory Experiment 'what's the problem? We'll solve it.' 0.7 0.7 "I could spend hours JUSt talking about applica­ 0.6 06 tions, even though that's only about 20 to 30 CI-C5 CI -CS 0.5 0.5 percent of what we do, because that's the parr that io- F t'. "-... most non-theorists are going to see as interesting 0.4 ~ 0.4 e, 03 K 0.3 i'x:: --- and important. And perhaps 25 to 30 percent of j 7 C8A u~ our theory work I wouldn't even tell to my closest 0.2 / 0.2 / C8A C8N5 C8nP C81P [chemist] fr iends, because they wouldn't under­ 0.1 k£..- 0.1 6 e'NS- 60P 61P stand why we're doing it. They'd say, 'It's naive o ~ o 20 30 40 50 60 o 10 20 30 40 50 60 to think YOLI can do that problem,' or, 'Ie 's already nme (sec) Time {sec} been done, so why would you want to do it again?' or, 'You've made the wrong assumptions,' or something. Like any kind of research enterprise, The model's predictions for ing efforts because our scientists have worked at some parts of it underwri te the rest of it at any an eight-carbon-atom the MSC," Tang says. "It's a critical link. Being paine in time. You JUSt have to keep working on things that wi II generate the new knowledge feedstock matched experi­ in the MSC all owed me to see how its methods could be applied to other problems we are work­ that wi ll be useful years from now. mental reality pretty well, ing on, and then to bring the relevant people in. "Ie 's actually evolved into a really good training tracking the rise and fall in We are way ahead of the other oil companies as a ground for graduate students and postdocs, concentration of the result." (The projects mentioned in this article are because today's world is very different from what just a sampling of the ones Chevron and the MSC it was 20 years ago. Typically then, a person got various hydrocarbon have collaborated on over the yeats.) a PhD in basic science, and joined a company like fam ilies_ The abbreviations Goddard's original notion was to have hi s Exxon or Chevron, and gradually learned the com­ are as follows: eJl' is theorists talking to the theorists in industry. pany's technology, gradually learned what was interesting to the company, and eventually got everything with x carbon But there aren't that many Out there any more­ changes in management strategies have caused into a position to co ntribute. These days, if yo u're atoms, A is aromatics, nP most of the companies that maintained showcase not useful on a projec t from Day One, they're nor is unbranched paraffins, iP research facilities as corporate crown jewels back in going to hire you. So our grad students and post­ is one-branched paraffins, the '60s and '70s to abandon or severely curtail docs have to understand what the technological issues are in order to know how to use their and NS is five-carbon them. Like the Cold War-driven space race, they appear to have been a prestige item whose time knowledge of basic science. They have to know rings. has passed. It's become much bener for the bor­ how to ask the right questions. There's no pre­ rom li ne to farm your research out. "So whar we scription; yo u have to learn how to think about do now is interact wi th engineers, who are problems. And you only learn to think about interested in solurions." them by being exposed to them. It's good for Goddard is a theori st, but he's also a pragmatist. them to see how we struggle with defining the He uses these corporate contracts- which now problem. Mostly, in science, the important account for about 50 percent of hi s budget- as a advances are in figuring out what the questions vehicle to advance his theorerical work on compu­ are. And the graduate students get the pleasure of tational methods and verify the resu lts. Says seeing how interes ted somebody other than their Goddard, "We continue to push the frontiers of advisor is in their research. Because the advisor is the methods, but by and large the methods them­ always interested. But to have other people who se lves aren't being fund ed. T hat's just part of our really think it's neat stuff, and to have students see overhead. If you say in a grant proposal that your how it's directly related to some important tech­ aim is to develop a new method, it's very easy for nology, that's good for them to see. So we think the reviewers to poke holes in you. Someone will it's developed into a good model fo r education in say, 'Oh, the ass umptions are bad,' or, 'It's not addition to one fo r doing research ." going to be practical,' or, 'You won't get the Bes id es the lab's corporate cliem s, says Siddarth answers well enough ... ", Adds Blanco, "The Dasgupta, "We're reall y a resource for the whole trouble is, when you propose method development Caltech community. We do a lor of custom soft­ alone, there are so many paths you can take. But ware development, and people in a lot of different our work is all based on using that method to fields are finding us useful. For example, if some solve a real-life problem. Then the cri tical issues geophysicists want to find out what happens to come our naturally as you work on the problem, some kind of rock under tremendous press ure, but and we focus on solving those issues." And if they don't have the software to si mulate it, they those issues call for a new method, they assign can corne here and work with us. They just have a postdoc to develop ie. to give us a half-page summary of what they want "I think it's fair to say that 20 years ago, you to do, and we give them the resources and training co uld have characterized what we did as 'have gun, to do it." n

ENG I NEE RI N G & SCIE N CE NO. 2. tronomic

Left: Progress continues on the giant dome and its shutter in December 1937. Above: The Hale Telescope dome today.

]0 ENGINEEKIH G & S(IENCE NO. Palomar at 50

On June 3, 1948, abollt a thousand invited guests (and reporters) witnessed the dedication of the world's largest telescope-the 20D-inch, newly Right: Among the christened Hale Telescope on Palomar Mountain. dignitaries who spoke at On June 3, 1998, a smaller group of about 200 gathered for a dinner beneath the telescope's giant the dedication of the Hale horseshoe mount to celebrate the birrhday of the Telescope were President magnificent instrument, still healthy and working Lee DuBridge (foreground, hard ar 50. If its lifetime were to be measured fro m concep­ with Mrs. Hale). and tion, however, it would be 70. It was in June representatives of the 1928 that George Ellery Hale received assurance Rockefeller Foundation, the of funding from the Rockefellet Foundation's Carnegie Institution of International Education Board to allow Caltech to start building the telescope. No one then dreamed Washington, (altech's that it would take 20 yea rs (with a war interven­ Board of Trustees, and the ing) to solve the eng ineering problems of pouring, Observatory Council. (fansporting. polishing, and supporting so large a piece of glass. The mirror's progress and the telescope's construction captured the imagination of the nation and the world. Even its housing, in a dome that rivaled Rome's Pantheon in size, inspired awe. The New York Times, in commemo­ Left: Palomar's current rating the 50th anniversary, called it a "cathedral of science," a phrase quoted by Caltech President director, Wallace Sargent, David Baltimore in his opening remarks at the the Ira Bowen Professor of recent dinner. Astronomy (far left), talks Probably the only person who attended both celebrations was Jesse Greenstein, now the to journalists touring the DuB ridge Professor of Astrophysics. Eme~itus. In 200-inch telescope in his [alk to the 1998 audience, he noted that he conjunction with the 50th had arrived at Cal tech JUSt a few weeks before the anniversary celebration. dedicarion of the 200-inch to take the position of "organizer" of astronomy. His only professional colleague was Fritz Zwicky. "a man of great originality," and the as (fonomy office in Robinson was filled with World War II debris. Within a few years of his arrival, however, Greenstein would be able to add a professorship per year for a dozen years, and Caltech's stature in astronomy grew to fit its telescope. Gteenstein began observing with the 200-inch

ENGINEERING & SCIENCE NO. 31 Jesse Greenstein at the prime focus of the Hale Telescope, his home for 30 nights a year over 40 years.

in 1952 and continued co use it fo r about 30 nights a year fo r the nex t 40 years. He also "fell in love with this 500-con monster of steel and glass, still one of the great instruments of the world." "There are hundreds of astronomers from all over the world who share my personal affection for that gray giant," Gree nstein said. "J[ has performed miracles, observed the faintest objects possible, and seen co the edge of the expanding universe- an edge that remains ever more unreachable as larger instruments are built." Augustus Oemler, director of the Observatories of the Carnegie Institution of Was hington, outlined "50 Years of Resea rch at Palomar" for the anniversary guests. Carnegie had been Caltech's Technology marches on. The top picture dates from 1978, and is the first ever taken of the partner in the Palomar O bservacory until 1979, when Carnegie turned its attentio n to building rings around Uranus. (The rings had been discovered in March 1917, when Uranus occu lted newer telescopes. Duri ng those 50 years the Hale a star named SAO 158681.) The planet was shot at two infrared wavelengths-one at which Telescope dominated as tronomy. Already in 1952, the planet was relatively bright, and one at which it was darker. Subtracting the two left the 200-inch had doubled the size of the visible universe, and Allan Sand age, PhD '53, of just the rings. But the process wasn't nearly that straightforward. There were no two­ Carnegie co ntinued to refin e the Hubble constant dimensional CCD arrays in those days, so the Hale Telescope was repeatedly scanned across and define the age and size of the universe. the planet and the light directed onto a pair of single-element detectors, generating two Sandage also observed on the telescope a new class channels of analog intensity data that were recorded on magnetic tape. The tape was of objects, dubbed quasi-stellar radio sources, or quasars, whicb Maanen Schmidt determined had played through a machine that subtracted one channel from the other and output the huge red shifts and were the most distant and difference as a strip chart. The strip chart was converted to digital format with a ruler and luminoLis objects in the universe. The 200.-inch a pendl by Keith Matthews, member of the professional staff, and the digital data punched also was used to explain the evolution of stars, to on cards that were fed into a computer. The computer interpolated the data, filling in the discover large-scale structures in the universe, and to identify tad io sources with optical counterparts. gaps between the points, and output the (digital) results onto another magnetic tape, Technology has enabled [he Hale Telescope [0 which was fed into a Geology Department computer that had a video display-a rarity back keep up wi th the times. In the '70s, charge­ thenj most computers talked to their operators through teletype machines. A Polaroid coupled devi ces (CCDs), which greaely improved sensitivity to light, revolutioni zed telescopes. The photo was taken of the video screen, and that's what you see here. The entire process took "4-shooter," a CCD ca mera built for the 200-inch months. By contrast, the recent infrared image of Jupiter's rings above took about 20 telescope, was developed at Caitech and was one of seconds to produce. the first adaptations of the technology co ground­ based astronomy. And the upgrading continues. "The observatory is nO[ a museum- it is a vital facility which is

11 EII G INE H IN G & H lfNCE N O . • .' " • .'~ " ' 0 I ; '0 , • r _' j __ 0 0 -.•. . () '. , ·.. .,.:t . ~ci~, . .. , :0 ••.. a _ (£) ,., . , , 0 , . .' • • • It ·· '.or; ... .. '. '0 , " Ia . "- .. ' 0.." • •.., 4(), ,, ~ . 0 , • Professor of Astronomy ,- I .• ' 0 · .... • '. Charles Steidel (PhD '90) is , , ' '0 , . , ~~ . , • ,~ , ,'. ," "", using the Hale Telescope to ~ ,:' • • v ;: . • , • map the universe as it was .. '00 at about one-tenth its 0 , ,,-9: :.. present age . The star field , . '~9J~gs , ' . , · at left contains some b', , ~ , . - -,. 2,000 galaxies, including " . .. . Q' : ", . en larged and the track f " ' ..-. .' the sky through three lowered four feet to , C/' O . 0 different filters provides a get it out. tf) 0 O ' b quick and easy way of separating galaxies that go way back-in both senses!-from more recent ones.

Right: After it was shipped across country from Corning. New York, as crowds lined the tracks to watch, the giant mirror spent nine years (and occupied 180,000 man hours) being ground and polished in Caltech's optical shop (later reincarnated as the synchrotron building). About 30 tons of abrasives were used to grind it down from 20 tons of Pyrex to less than 15 tons.

EH GI HEE RI NG & H I E N (( NO . H8 Hy

Left: A spectrum of 3(273, the Rosetta Stone of quasars (top), and a reference spectrum (bottom). Note the relative positions of the lines labeled Ho. Hy. and Hp. The key to matching objects seen through different telescopes is to have accurate location data. Early I'adio telescopes could only plot a source to within several arc minutes, but in 1962, the moon occulted 3(273 several times, giving its position to within an arc second. Astronomers got a Comparison J 1 '--'~ ___ J 1 Spectrum H8 Hy Hil . second lucky break when the Hale Telescope was trained on 1 the refined coordinates, in that the spectrum of the object 3889 A 5016 A 6030 A they saw there was redshifted by 16 percent-unusually far for something that looked like a star (it isn't), but not so far that Maarten Schmidt (below, left) couldn't eventually identify the spectral lines.

.. ) ,. ( _0_0 __ .> , ...... :,'11 .~ u r .140 0161. 0.1 0 J . ~~, .. t!... . I '"> • ~~:'F.... · i. . ~ • • _ ,------I

''1 If)

<. • .; ,.~ '. ..'> ._ ...... • -0'" . "~ ""' ,--.. ' '_. ... -­.. I...... : ~t"'.... _ •• ~ • • ...... :~,.. : .. 'I' ~" .. . ~~:r . '";\ : \ • oIL" ,...... ~.

-, .... ., .,., • . ...., I, . ~I # ' _

left: Research Fellow Don Hamilton and the Norris Spectrograph, which was built for the Hale Telescope in 1990. A robotic arm (at Hamilton's right elbow) deploys optical fibers anywhere in the instrument's field of view to carry light to a CCD detector; up to 176 spectra can be collected simultaneously, each by its own fiber. Above: Postdoc Todd Small and Professor Wallace Sargent are using the instrument to construct detailed maps of the large-scale distribution of galaxies out to redshifts of 0.5, or about 4 billion light years. This data slice from the autumn sky starts in the upper left­ hand corner, at redshift 0.0; each succeeding row picks up where the one above it left off. The colored dots are groups and clusters of galaxies, packed at least five times more dense­ ly than the field as a whole. Black dots are galaxies without nearby neighbors. The slice is

about as thick as it is wide (r), 50 in three dimensions it's a very tall, thin pyramid.

• 4 ENG I NEERIN{j & HIENCE NO • continually being upgraded and improved," said Wallace Sargenr in his talk on "The Next 50 Years. " Sargent is the Bowen Professor of Astron­ omy and current director of the Palomar Observa­ tory. He noted the role of computers and de­ scribed the adaptive optics system developed by new Palomar partners )PL and Cornell, which should provide "a dramatic improvement in image quality over small areas in the sky." Because of these efforts, "there is absolutely no doubt that the 200-inch today is a better telescope than it was 50 years ago. And how many 50-year-olds can say that?" The Hale Telescope is no longer the biggest in the world; the two 10-meter W. M. Keck Tele­ scopes (another Caltech undertaking) in Hawaii are. In describing some recent astronomical discoveries, Sargent emphasized the collaboration between Palomar and Keck. In the discovery of the first brown dwarf, of optical counterparts to gamma-ray bursts, and of newly forming galaxies in the very early days of the universe, Palomar first Below: In the media identified the objects as interesting, whereupon they were followed up and confirmed spectroscopi­ spotlights, the mirror is cally at the Keck Observatory. rolled out and prepared Sargent warned of the danger that the new ror its final journey to the generation of great telescopes might freeze grad top or Palomar Mountain. students and postdocs out of hands-on experience with the instruments themselves, whereas Palomar That journey (bottom), will be able to keep alive this important astro­ Above: Gamma-ray bursts suffered rrom the same identity which took place on nomical tradition. He closed with the reaffirma­ crisis that quasars did, until BeppoSAX, an Italian/Dutch November 19,1947, tion of a statement then-President Lee DuB ridge made at the 1948 dedication: satellite, began providing accurate position inrormation. required two pushers and "The word 'dedicate' in the English language On May 8,1997, BeppoSax saw a gamma-ray burst ror a puller to get up the means to set apart by a promise. It is essentially which the Hale telescope was able to find an optical sleet-slick road. synonymous with 'consecrate,' which means to counterpart (marked with an arrow, and still increasing in make holy by a special act. The word has more rhan a formal or material significance. It also brightness two days arter the burst). A spectrum taken at carries spiritual implications. It is in this sense the Keck Observatory on May I I, by a team led by actually that we do today set aside this temple of Professor or Astronomy and Planetary Science Shri learning, and promise that it shall be devoted henceforth to deepening man's intellectual and Kulkarni, proved that the source is more than halrway spiritual understanding." across the observable universe rrom us, meaning that the burst was among the most energetic events in the universe. There's still debate as to whether all gamma-ray bursts are as distant, so it's too early to tell ir this is the Rosetta Stone or gamma-ray bursts.

ENGINHRIIIG & ICIEIICE NO . lS ------

Left: One of the 90-foot dishes (foreground) is lined with renecting steel mesh berore being hoisted onto its pedestal in 1958. Above: In 1998 the same antenna has been joined by the 130·foot (right) and the millimeter-wave array (left).

•• EHG IHEERltlG & ICIEHCE HO • • • • and OYRO at 40

Farther down the spectrum, and a few hundred miles narch of Palomar, [he Owens Valley Radio Observarory marked its 40th anniversary with a Right (top): Surveying the day-long celebration October 3. There are empty Owens Valley site; actually three generations of telescopes at OVRO. The two 90-fooc (27 -meter) antennas were the cattle were gone but dedicated in 1958; the 130-foot (40-metet) dish the telescopes had not yet was finished in 1968; and the array of six mjUime­ appeared. ter-wave antennas, each 34 feet (lOA meters) in Center: Both 90-foot diameter, on a T-shaped track, was begun in 1977 and completed in 1995. dishes and the living A couple of hundred guests gathered in a wind­ quarters and shop (with whipped tent (no capacious dome here) to hear the White Mountains in several sessions of talks on OVRO's history and the background) can be current work. After greetings from President David Baltimore, relayed by Provost Steve seen here in 1959-after Koonin, Gotdon Stanley, who was the obser­ the dedication but before vatory's director from 1965 to 1975, spoke about the north arm of the L­ the past. He described driving up past the (electronically noi sy) Mojave desert, through Red shaped track was begun. Rock Canyon and into the Owens Valley in an Bottom: OYRO's first almost unbearable combination of heat and dust, director, Gordon Stanley, discovering the remote, and relatively noiseless, and current director, site in 1955: "By the time we arrived at Lone Pine, I was becoming excited. Here rhe magnifi­ Professor of Astronomy cence of the Sierra Nevada reveals itself for the Anneila Sargent, enjoy the first time; the outside temperature was now a desert air during a coffee seemingly tolerable 115 degtees, and the lack of break at the 40th birthday stations was noticeable on the car radio." Driv­ ing down a dirt road running east out of Big Pine, celebration. nestled up against the White Mountains, the Sierra Nevada ri sing up on the other side of the valley, he ended up in a remote piece of desert where the 130-foot antenna now stands. Provi­ dentially, the land belonged to the Los Angeles Department of Water and Power (but that's another story), and Stanley recalled how he stopped at their Independence office on his way back to Pasadena to start negotiating for a square mile of land. Stanley spoke about the early days at OVRO and about John Bolton, its founding director. He credited Bolton's vision for the

ENGINEEIING & SCI ENCE NO. .7 The 90-foot antennas were originally constructed to study the structure of radio galaxies in the distant reaches of the universe. Now they are being used to trace the sun's magnetic fields.

observatOry's later accomplishments, even though Bolton had returned to his native Australia in 1960, where he died in 1993. (His widow, Letty, traveled back for the anniversary celebration.) In conclusion, Stanley, who also calls Australia his home, quoted an Australian "bush" poet, A. B. Patterson: u "And the bush had friends to meet him and Solar nares, which erupt ~ their kindly voices greet him out of tangled magnetic iii In the mutmur of the breezes and the river v;> fields, emit x-rays and z on its bars, radio waves from a layer of And he saw rhe vision splendid of the sunlig ht plains extended the sun's atmosphere And at night the wondrous glory of the called the corona. The top everlasting stars." image, from the Yohl{oh Anneila Sargent, OVRO's current directOr, thanked him and said she had been waiting for spacecraft, shows the 30 years to hear that line quoted at a conference. nare's x-ray emissions, Most of the observatory's other previous directors while the bottom panel were also in attendance, including Robbie Vogt plots OVRO observations at (acring direcco r 1980- 81), Tony Readhead (1981- 1986), and Nick Scoville (1986-96). Many of the various radio frequencies, / original graduate srudents who worked at OVRO overlaid on magnetic field (as well as later ones) also recurned fo r the occa­ data for the sun's sion, and several gave talks on the early work on the twO 90-foot dishes, which, linked together as photosphere, or visible an interferometer, formed for a time rhe largest surface, a few thousand and most sensirive instrument of its kind in the kilometers below. world (E&S, Spring 1994). Its sensitivity enabled the extremely accurate measurement of the posi­ tion of radio sources, making possible their identi­ E-W [arc seel fication with optical counterparts. Among the

38 ENGINEERING & \( lEN ( E NO . J 998 TIME (Minutes After 1600 UT)

Dal e Gary, visiting asso­ ciat e in astrophysics, has been using the 90-root a ntennas to study former grad students who spoke was Bob Wi lson, sunspots, which are also PhD '62, known in his OVRO days as the Galaxy Man. Later, while at Bell Labs, Wilson co­ associated with strong discovered the cosmic background radiarion, electromagnetic activity. fat which he shared the Nobel Prize in 1978. Thi s is a cross-sectional In later years, the two oldest antennas might view or a magnetic loop have gone begging for something to do, according to Sargent, but with new receivers and signal derived rrom OVRO solar processing equipment, Professor of Astrophysics, data, showing how its Emeritus, H al Zirin cleverly adapted them to solar radio emissions stretch as tronomy. In their current incarnation as pan of a five-element array, they use the microwave spec­ rrom one sunspot to the [[urn co study the structure and activity of rhe other. Although the solar atmosphere, in particular the corona's mag­ ve rtical dimension here is netic fields. rrequency, it can be Besides the anniversary of the two original dishes, 1998 was also the 30th birthday of the thought or as height, 130-ft. antenna, an occasion that was poimed lower rrequencies out by Marshall Cohen, professor of astronomy, correlating with higher emeritus, who had come up with the idea of the heights. anniversary celebration. This huge antenna was built as a prototype for what was to be an array of eight large dishes, which never came to pass, Cal tech President l ee having been beaten out by the Very Large Array in DuBridge dedicates the New Mexico. (The VLA starred with Jodie Foster 130-foot antenna in in rhe 1997 film COlltact; OVRO's 130-fr. dish had its ow n 15 minutes of fame in The Arrival in October 1968 (above), and 1996.) The 130-fr. has served as one srarion in visitors get a ra re Very Long Baseline Interferometry (VLBT), an opportunity to. clamber array of antennas worldwide. It was always one of around it in 1998 (left). the most important, sajd Cohen~ since it was large and on the West Coast, it was used in the longest baselines, which have the weakest signals. As­ tronomers use ViBI to study the structure of distant quasars and radio galaxies. Starting in 1978, the 130-ft. antenna also carried out observations of the microwave back­ ground radiation, which has led to what Professor of Astronomy Tony Readhead described as a child of OVRO. Thar's rhe Cosmic Background Imaget (see E&S, No. 4, 1996). On a mountaintop in Chile, far from its parent, the e Bl will look for

ENGINEERING & SCIEN(E NO. •• .,

s ., - '""' '0 ·0·

1i"'~ ·oeW O'".J!"'1f1"6 18~ '8',( 1l1-" 1/~·1 18'1 O ~'J e""'e '6 le~ f) \II~ ~ 18',) 16' 2 H!~l

o (1950)

Professor of Astronomy Anneila Sargent and her group have been employing the millimeter array to search for other solar systems-observing stars that resemble the young sun (T Tauri stars) with surrounding disks of dust and gas out of which planetesimals, and eventually planets, might form. The image above is of a star named LkCa IS, about SOO light years away, which has a very promising protoplanetary disk. The image on the left shows contours of intensity, and the center picture represents motion­ blue gas is approaching us and red gas is moving away. Comparison with computer simulations shows that LkCa 15 is surrounded by a disk about seven times the size of our solar system that contains enough material to make all the planets, and is rotating like the planets rotate around the sun.

Above: Bob Leighton works on the first dish of the millimeter array in 1974 at Caltech. His ingenious "Tinkertoy" design made it possible to take the telescope apart and reassemble it at Owens Valley. Left: In 1998 guests at the anniversary celebration tour one of Leighton's 1O-meter dishes, which now number six (far left). The sixth dish, built with a gift from the Kenneth T. and Eileen l. Norris Foundation, was added in 1995.

40 ENGI N EERI NG & S{ I EN C E NO. 50

•o " ~ 0

- 50 The Whirlpool Galaxy, MS I, a large nearby sp iral galaxy, is shown at right as 60 o -50 seen by the millimeter- Arc Sec wave interferometer's observations of the carbon fluctuations in the very early universe that might monoxide (CO) molecule, explain the lumpy structure that we see today, which is used to trace the with clumps of galaxies spread unevenly about the cosmos. Carbon monoxide imaging of the extremely luminous galaxy location of the dense to The newcomers Owens Valley are Bob Arp 220 was carried out in the new ultra-high resolution molecular gas douds that Leighron's IO.4-meter millimeter-wave antennas, a configuration of the OVRO millimeter array. These new are the birthplaces of stars marvel of creative engineering. Leighron, the Valentine Professor of Physics, emeritus, who died observations uncovered the presence of two counter­ in galaxies and to study in 1997 (his widow, Matge, attended the anniver­ rotating disks of gas, believed to be the remnants of two the association of these sary parry), built the dishes out of panels machined merging galaxies in Arp 220's nucleus. The upper plots locations with the spiral to a particular figure and held up by a sys tem of show a color-coded intensity map and the gas motions. structure. This image 1,400 Struts tapped together with dowel pins­ "the world's best Tinkertoy set," according to The bottom diagram illustrates a theoretical model based shows the gas's velocity David Woody, OVRO's assistant direcror fo r site on the observations. (red for receding and blue development. First constructed at Calrech (they for approaching), as were designed for potential assembly by astronauts derived from the doppler in space), they were then reassembled at OVRO. The six dishes may be moved again, this time to sh ift of the CO emission Harkless Flats-at 9,000 feet in the Inyo Moun­ line. The overall rotation tains, 5,000 feet above Owens Valley-fat better of MS I 's galactic disk can imaging of the microwave radiation. The millimeter-wave artay is so sensitive that it be seen, as well as the can pick up signals 10 billion light years away. streaming motions Among the closer subjects it is observing are the associated with the spiral spiral structure of nearby galaxies, galactic nuclei, Arp 220 System arms. The images on this colliding galaxies, the interstellar clouds that supply the fue l for star formation, and the circum­ page are the work of a stellar disks around nearby stars that may be team led by Nick Scoville, incubating planetary systems like our own. nuclear disks the Francis Moseley In closing the talk sessions, Sargent announced that two graduate fellowships in radio astronomy, /\ Professor of Astronomy. named for John Bolcon and Bob Le ighcon, had been es tablished, thanks to a gift from Gotdon ~~ Arp 220 W Moore, chair of Cal tech's Board of Trustees. Arp 220 e The telescopes were open for tours, or JUSt scrambling around on your own, fo llowing the talks, and the anniversary celebration concluded outer disk with ~YRO's annual party, featuring live music and dancing and an almost-full moon. 0

ENGINEERING & I{JEHCE NO. J 41 Books

Our planer seems co be on a caused by the collapse of a geologic activity around contInual crash course, as the volcano summit in the couese the world. Of course, this is jigsaw puzzle- like places char of a major eruption). For entirely necessary in a world compose the earth's crust (00 - more than 60 years, the floor where continental collision srandy rearrange themselves, of the caldera rose gradual! y, has produced such odd creating and destroying ocean indicating the ptesence of occurrences as "bits of Africa floors and mountain ranges, magma pushing lip from that somehow ended lip in and triggering earthquakes underneath. Bue then, with­ Massachuse tts. " and volcanoes. Out an eruption, the floor As may be expected, the But geologists' present began subsiding again. The mass ive forces that create knowledge of cec tonic motion same situation is going on earthquakes and volcanoes does not answer all the ques­ at present in the Long Valley so met i mes produce some cions about earthquakes and caldera in Mammoth Lakes, interesting side effects, such volcanoes, as Kerry Sieh and California, where "the rescl ess as Old Fairhful, a hot spor Simon LeVay make abun­ magma seems to be consider­ whose steamy volcanic dantly clear in theif book, The Ing Its opoons. processes are a delight rather w. H. Freeman a nd Earth in Turmoil. Sieh, profes­ The book stares on the west than a fright to those who Company, 1998, sor of geology at Calcech, and coast of the United States, an experi ence them. And dead 324 pages LeVay, a well-known neuro­ area known for its seism ic forests-suffocated by the sc.iemis[ and science writer, activity, especially on the San excessive amount of carbon offer a history of the ea rch 's Andreas faule, "a seismologi­ dioxide that seeps to the destructive and conscc llcrive cal celebrity." Sieh has surface-reveal the presence processes for both the lay studied the San Andreas for of underlying magma. reader and rhe geology buff. more than 20 years, and his The writing itself is as The book reads like a seri es research, along with the cohesive as the processes it of mysteries, setting the scene findi ngs of his peers, com­ describes. Told in a visual of a geologic occurrence­ poses three chapters of the style, the narrative also in­ such as the eruption of Mount book, offering a glimpse of cludes plenty of drawings, St. Helens in 1980-and the past and present activity maps, and photographs. Sieh tben following the tra il of of the fault, as well as dis­ and LeVay augment their tale dues back to the cause. In cussing future possi bilities. wirh re-creations of ancient some cases, the mystery is From the west, the narrative legends and chilling first­ never solved, for nature is a travels far and fast, spreading person historical accounts. tticky wtiter who introduces q uickly to the opposite coast Take the case of one Missis­ unforeseen plot twists into like the lava that flows in the sippi Valley resident in 181 2, tbe geologic narrative on a H awaiian Islands-a turbu­ who described his experiences regular basis. lent region that Sieh and du ring a particularly severe In other cases, the m ys tery LeVay save for the cataclysmic aftershock on rhe New lies in the fact that something end. Although the book Madrid faulr. should have occurred but focuses mainly on the conti­ "Carbonized wood .. . didn't, as in the case of the nental U.S. and H awaii, the which was ejected to the Yellowstone caldera (a crater authors also give snapshots of hei ght offrom 10 to 15 feet ,

42 [NGINUR I N G & SCltNCE N O . Faculty File

GREENSTEIN LECTURESHIP BEGINS HONORS AND AWARDS

and fell in a black shower, Giuseppe Attardi, the Thomas McGill (MS '65, mixed with the sand which Steele Professor of Molecular PhD '69), Jones Professor of its rapid morion had forced Biology, has teceived a Applied Physics, has received along; at the same time, the Gairdner Foundation Interna­ the Defense Advanced roaring and whistling pro­ tional Award for Achieve­ Research Projects Agency's duced by the impetuosi ty of ment in Medical Science for Best Technical Development the air escaping from its con­ his "pioneering contribution of the Year Award for the finement, seemed to increase to our understanding of the work described in "Double the horrible disorder of the structure of the human Manhattan" in this issue's trees which everywhere en­ mitochondrial genome and its "Random Walk" section. countered each other, being role in human disease." Associate Professor of blown up, cracking and lance Davis, the Harkness Computer Science Peter splitting, and falling by Professor of Social Science, Schroder is one of 24 "prom­ thousands at a time. In the has been awarded the Alice ising young sc.ientific meantime, the surface was Hanson J ones Prize in researchers" to be awarded a sinking, and a black liquid Economic History. He and five-year fellowship from the was rising up to the belly of coauthors Robert Gallman David and Lucile Packard my horse, who stood morion­ and Karin Gleiter were Foundation. Schroder's Jesse Greenstein less, struck with terror." honored for their book, In research involves "modeling, Even though the book Search of the Leviathan: simulation, and visualization offers up more frightening Technology) Labor, Productivity of large problem sizes on tales of human! ty's relative A lectureship has been and PfOjits in American workstation-class computers." powerJessness against the established in honor of J esse Whaling, 1816-1906. John Seinfeld, Nohl forces of nature, in the end its Greenstein, DuBridge Professor of Aeronautics Professor and professor of purpose is not to terrify, but Professor of Astrophysics, Morteza Gharib (PhD '83) chemical engineering, and to educate. For besides Emeritus. The first annual has been elected a fellow of chair of the Division of bringing tl1e reader up to Greenstein Lecture was given the American Physical Engineering and Applied speed on what geologists on October 28, nine days Society. Science, received the Fuchs know about the earth's after Greenstein's 89th William Johnson, the Award at the International evolutionary processes, the birthday. The speaker was Mettler Professor of Engineer­ Aerosol Conference in· authors also discuss some Princeton University's James ing and Applied Science, will Edinburgh, Scotland, in steps humans ca n take to Gunn (PhD '66), who built receive the MRS Medal at the September. survive in such a hostile the "4-shooter" CCD camera December meeting of the Paul Wennberg, associate environment. for the Hale Telescope and Materials Research Society. professor of atmospheric After all, when you Uve on revolutionized ground-based He is credited with "the chemistry and environmental such a volatile, shifting optical astronomy. In his development and fundamen­ engineering science, has been surface as we do, putting the lecture, "The Sloan Digital tal understanding of bulk selected by the National pieces of the geologic puzzle Sky Survey: First Light and metallic glass-forming Science and Technology together is the first step Commissioning," he affec­ alloys," leading to a new class Council to receive the toward a safer future. 0 tionately referred to Green­ of structural materials for Presidential Early Career stein as his astronomical advanced engineering Award for Scientists and Ryan Poqtlette "father. " 0 applications. Engineers. A new award, it

ENGINEE R ING & SCIENCE NO. 4. HONORS AND AWARDS (CONTINUED) GOODSTEIN GETS OERSTED MEDAL

recognizes up-and-coming David Goodstein, professor of pbysics and applied physics, scientists and engineers "who G illoon Distinguished Teaching and Service Professor, and vice show exceptional potential for provost, has been selected the 1998 Oersted Medalist by the leadership at the frontiers of American Association of Physics Teachers. The Oersted Medal, scientific knowledge." established in 1936, recognizes a teacher for notable contribu­ The 1997-98 teaching tions to the teaching of physics. This puts Goodstein in awards of the Associated excellent company-past medalists include Carl Sagan, Explora­ Students of Caltech (ASClT) torium founder Frank Oppenheimer, and Nobel Laureates H ans have gone to Professor of Bethe and r. 1. Rabi. Only twO other Caltech faculty have ever Biology Marianne Branne-e­ been so honored: Roben Millikan (i n 1940) and Richard

Fraser, Professor of Applied Feynman (in 1972). D David Goodstein Mathematics Donald Cohen, Professor of Physics Andrew Lange, Professor of Biology Ellen Rothenberg, and Mellon Visiting P rofessor of Humanities John Miles. Recipients of honorable mentions are Professor of Biology J ames Bower, C ALTECH TAKES Two OF S IX LAWRENCE AWARDS Associate Professor of Physics Emlyn Hughes, Professor of Planerary Science and Geology Bruce Murray, Lecturer in Computer Science The U.S. D epartment of university has produced two and Electrical Engineering Energy has named Steven WInners. " Glen George, and Visiting Koonin (BS '72), vice Koonin developed a tech­ Professor of History Timothy ptesident and provost and nique for modeling atomic Breen. Named for teaching­ professor of theoretical nuclei on parallel computers assistant awards are Travis physics, and Ahmed Zewail, that is widely used in nuclear Williams, a senior in chemis­ the Pauling Professor of physics and astrophysics. try this year, and Alexa Chemical Physics and And his theory of quantum Harter, a graduate student in professor of physics, the interference effects between physics, and, for honorable recipients of this year's E. O. emitted pions is one of the Lawrence Award, given for most important techniques mentions, Michael Siu, a Steven Koonin graduate student in chemis­ excep tional contributions to for studyi ng hot nuclear try, and Ben Miller, who the development, use, or matter, including simulations received his bachelor's degree control of nuclear energy. of densities and temperatures in mathematics this past The award has been given similar to those soon after the J une. since 1959 in memory of Big Bang. Recipients of the 1997-98 Ernest Orlando Lawrence, the Zewail pioneered the bur­ Graduate Student Council Nobel Prize-winning nuclear geoning field of femrochem­ Teaching.and Mentoring physicist who invented the istry, in which molecular Awards are, for classroom cyclotron. Koonin and reactions are viewed in teaching, Professor of Zewail will join four other extremely short pulses of Aeronautics and Applied Americans at the J anuary 15, laser light. This creates stop­ Mechanics Ares Rosakis and, 1999, awards ceremony in motion photographs, as .it for mentoring, Scott Fraser, Washington, D.C. were, of chemical bonds in Rosen Professor of Biology. Secretary of Energy Bill the act of breaki ng and Teaching-assistant awards Richardson commended forming- the so-called have gone to Zvonimir Cal tech for Koonin's and "transition state," which Bandic, a graduate student in Zewai I 's advancements in had long been hypothesized applied physics, and Ayhan nuclear science, saying, "This to exist but had never before Irfanoglu, a graduate student is only the second time in the been seen directly. rJ in civil engineering. 0 Ahmed Zewail award '5 38 years that one

44 ENGINEER I NG & SC I ENCE NO. Offlce of Gift and Estate Planning

greater gift and estate tax world cannot be maintained savings," he says . Unlike a without leadership in tech­ charitable remainder trust, nical fields . If I'm not will­ where the donor receives ing to support Caltech with annual distributions for life my background, who will?" and the remainder goes to the The unrestricted nature of the Institute, a lead trUSt works income from this trust is par­ the opposite way. The Insti­ ticularly welcome because it tute receives an annual dis­ allows the Institute to direct tribution from the trust over support to the area of greatest a set period of time. When need. the trust ends, the principal It is clear that Goss and his goes to noncharirable bene­ wife Mildred, members of ficiaries, usually the donor's The Associates since 1986, heirs. T he tax advantage for are committed to excellence the donor is that, for gift and in education. They have put estate tax purposes, the IRS their three children through calculates the gift to the private universities and paid donor's heirs at its discounted the tuition for their six present value when the lead grandchildren to attend pres­ trust is established. Years tigious private institutions. Dr. W ilbur Goss, formerly "We have a family tradition LE ADING WITH later when the trust ends, assistant director of the Ap­ the assets distributed to the of trying to support private LEAD TRUSTS plied Physics Laboratory at designated heirs are likely to universities because we feel Johns Hopkins University, be worth much more, bur no they have a strong role to has spent much of his retire­ additional gift or estate tax is play," he says. ment studying, investing, assessed. "Many people my age, if and putting his knowledge Goss figures that the lead they have been prudent with into practice. H e says he's trust is the most tax-efficient their finances, will have cash made more money in retire­ way to transfer assets to the flow in excess of their needs," ment than in his professional Institute and to heirs. "J Goss says. "This is a window life. regard Cal tech as the premier of opportunity to make a sub­ Goss first learned about technical institute in the stantial contribution to Cal­ charitable lead trusts while country," says Goss, who tech without jeopardizing doing research at a law has a PhD in physics. "The estate settlement plans. I library. "The more I learned, fu ture of the country is asso­ have everything to gain and the better it looked," he says. ciated with what Caltech is nothing to lose." "The concept of a lead trust doing. Our leadership in the intrigued me because it seemed to meet my needs and desires." As a result, Goss created the first lead trust at Chris Yates, JD Caltech in 1992, and he is Susan A. Walker, CFP currently considering the Office or Girt and Estate Planning establishment of another. "This is an especially good Calirornia Institute or Technology time to think about lead Mail Code 105-40 trusts because the IRS dis­ Pasadena, Calirornia 91125 count rate (used to calculate (626) 395-2927 the tax bend! ts of a lead trust) is at an all-time low, [email protected] and that results in even

I'

!