beam: the race to make the laser

24 Optics & Photonics News ■ July/August 2005

BeamThe Race to Make the Laser

Jeff HechtForty-five years ago this

year, physicist Theodore

Maiman and his

colleagues succeeded

in making the first laser

work at Hughes Research

Laboratories in Malibu,

Calif. Maiman performed

the experiment on

May 16, 1960, using an

elegant ruby rod placed

in a spring-shaded flash

lamp. This article is

adapted from Jeff Hecht’s

new book describing the

science, politics and

passion surrounding

the great laser race;

Beam: The Race to Make

the Laser was published

in March 2005 by Oxford

University Press.

July/August 2005 ■ Optics & Photonics News 25

T he new Hughes ResearchLaboratories building in Malibuseemed “the palace of science” to

physicist Bob Hellwarth when he movedin along with Theodore Maiman andother laser pioneers. The hillside Malibucomplex had a movie-star’s view of thePacific. The floors were wood parquet,with long, narrow offices lining the outer walls and windowless labs onthe other side.

In April, Maiman moved into an officefacing the ocean. His colleagues CharlesAsawa and Irnee D’Haenens shared thenext office. On the other side, HaroldLyons had a larger office that came withhis rank of department manager.

Moving his laboratory from CulverCity to Malibu stalled Maiman’s progresson laser development by at least threeweeks. But as soon as he was able, theintense and competitive Maiman focusedas tightly on his quest as a laser beam. Hedidn’t know that his competitors WilliamBennett and Ali Javan at Bell Laboratorieshad measured the first gain in a helium-neon mixture, but he had plenty of otherreasons to worry about Bell. Ten timeslarger than Hughes Research, and withten times the funds, Bell Labs was theodds-on favorite.

Maiman had made a compelling casefor using ruby as a laser material, but hewas surrounded by doubters. ArthurSchawlow, who had published the firstdetailed proposal for building a laser with Charles Townes, had dismissed ruby as an unsuitable material becausehe thought it could not generate lightefficiently or continuously. OutsideHughes, Schawlow’s conventional wis-dom that ruby wouldn’t work prevailed.Even within the institution, Maiman felt a lingering air of skepticism frommanagement.

But he forged ahead in the belief thatSchawlow had not based his opinion onadequate data, and that ruby could gener-ate pulses that were good enough todemonstrate laser action.

The task before Maiman differed fromthat which faced Bennett and Javan.Their excited mixture of helium andneon had produced only feeble stimu-lated emission at an infrared wavelengthinvisible to the human eye. They neededto tweak the gain upward in every way

they could, and to design optics thatcould capture every bit of that stimulatedemission to produce a detectable beam.

The choice of ruby gave Maimansome big advantages. He could buy thecoiled flashlamp needed to excite thechromium atoms from a standard cata-log. General Electric made three modelsthat he calculated could produce enoughenergy, and he bought a few of each to beon the safe side. His early experimentsalso indicated that ruby converted thatexcitation energy efficiently into red fluorescence. Thus, he wouldn’t need tostruggle for months to measure a smallamount of amplification, as Bennett andJavan had done. Nor would he require aresident optical wizard like Don Herriottat Bell to make the ideal set of optics,which, even when perfectly adjusted,would extract a weak and invisible beam.Maiman fully expected to see the redruby beam from his laser.

What Maiman needed, however, was a way to concentrate the lamp’s intenseflashes of white light onto the laser rod,and to make sure the red light from therod was stimulated emission rather thanordinary fluorescence.

When he first thought about excitingruby with a bright continuous lamp,Maiman had planned to put the lamp atone focus of an elliptical cylinder and a ruby rod at the other. That approachdidn’t look promising for the coiled flash-lamps, so he looked for another focusingarrangement. Inspiration came from a

salesman who said the biggest of thethree coiled lamps was so intense that itcould ignite a piece of steel wool placednext to it. Maiman realized he didn’t need special optics to focus the light.He could simply slip the ruby crystalinside the spring-shaped lamp, where the coils would surround the rod with abright surface.

After making that decision, Maimandesigned a simple experimental appara-tus. His calculations showed that thesmallest of the three lamps should sufficeto drive a ruby laser, so he started withthe FT-506, which could fit in the palm ofhis hand. He needed a small ruby rod tofit inside, so he picked one 3/8 in. indiameter and 3/4 in. long—a cylinderroughly the size of a fingertip.

He had the Hughes shop polish theends of the ruby rod flat, perpendicularto the length of the rod and parallel toeach other. Then he had both ends coatedwith silver, the most reflective metalavailable. He scraped the silver off thecenter of one end, leaving a transparentopening so the beam could escape fromthe reflective cavity.

Maiman had an aluminum cylindermachined and polished to slip around the spring-shaped flashlamp. The shinyinside of the cylinder reflected lightemerging from the outside of the lampcoils back toward the ruby rod. The lampcoil absorbed some of the light, but somepassed between the widely spaced coils,delivering more pump light to the rod.

Mount

Silver coatingon end of ruby

Silver coatingwith hole for beam

Ruby rod

Electrode

Beam emergingfrom ruby

FlashlampReflective metal cylinder(cut-away view)

Structure of Maiman’s first laser.

Plugs that slipped into the ends of thealuminum cylinder held the ends of thelamp in place, and served as mountingpoints for the ruby rod and the electrodethat triggered the lamp to fire. One plughad an opening for the beam to emerge.

The whole package was the size andshape of a small water glass. Some lightfrom the flashlamp leaked out throughthe beam opening, but the cylinder con-fined most of the lamp’s blindingly brightflash, so it wouldn’t completely dazzle theexperimenters’ eyes. A power supply firedelectrical pulses, which electrical cablesdelivered to the lamp.

The ruby laser design was remarkablysimple compared to the helium-neon gaslaser being developing at Bell. Maiman’scalculations indicated that he had a gen-erous margin for error. He didn’t needthe best mirrors; simply silvering the endsof the ruby cylinder would suffice for afirst trial. Nor did he need the best crys-tal—which was fortunate because theoptical quality of his ruby rod was notvery good.

The aluminum cylinder wouldn’treflect all the light back onto the rod,but he calculated that it should reflectenough. The Hughes machine shop

made the parts that he couldn’t buy.Starting with the smallest of the lamps,Maiman hoped to avoid the need forcryogenic cooling to remove excessenergy. If it didn’t work the first time,he could try better mirrors, bigger lampsor both.

In contrast, Bell Labs knew they wereworking at the margins of possibility.They needed the best mirrors to push thehelium-neon laser over the threshold ofoperation. They didn’t see any reason toseal their experimental laser tube in acase. They didn’t expect a blinding glarefrom the tube, and leaving it open to theair would help dissipate the heat.

A critical part of the experiment wascollecting conclusive results. Nobodyhad made a laser before, so Maimanneeded to devise his own set of tests.It wasn’t enough to look for light;ruby glowed red with fluorescence whenilluminated by violet, blue or greenwavelengths. He needed a way to distin-guish clearly between ruby’s ordinaryred fluorescence and the coherent beamgenerated by the amplification of stimu-lated emission in a resonant cavity.Fortunately, theory predicted someimportant differences.

The most obvious was that laser lightshould be concentrated in a narrowbeam—the idea that had excited GordonGould and others. Simple fluorescence is spontaneous emission, so it shouldemerge from a sample of ruby in alldirections. Stimulated emission thatamplified light bouncing back and forthbetween a pair of mirrors should bedirected along a line between the mirrors.

Yet it wasn’t that simple for a rubyrod. If the entire volume of the rodglowed uniformly and none of that fluo-rescence was absorbed inside the ruby,the light would look brightest when youlooked at the thickest parts of the ruby.Thus, more light would emerge from theends of the rod than from the sides.

Maiman’s rod was short and stubby,but the package he had designed let himobserve only the light coming from theend. He hoped to see the light narrowinto a beam, but he didn’t expect it to betightly focused. He also knew that merelyprojecting a red spot onto a wall wouldn’tconvince skeptics. He needed to makeadditional quantitative measurementsthat other physicists could replicate.Reproducibility would be the acid test.

A subtler way to distinguish laser lightfrom spontaneous emission was to mea-sure the speed at which the red rubyemission dropped after the pulse fromthe flashlamp ended.

Chromium atoms quickly releasedsome of the energy they absorbed fromthe flashlamp, dropping into a metastablestate. If left alone, they spontaneouslyemitted the rest of the energy as redfluorescence in about 3 ms.

However, if other photons with theright energy came along during themilliseconds that the chromium atomsretained that extra energy, they couldstimulate the atoms to emit their extraenergy before they dropped to the lowerlaser level on their own.

As long as the ruby produced onlyspontaneous emission—ordinary fluo-rescence—the intensity of the red lightshould drop by a factor of two every 3 ms. If laser action switched on, stimu-lated emission would drain the lightenergy faster, producing a shorter, moreintense spike of light than ordinary fluorescence. The human eye couldn’trespond fast enough to sense that


BEAM: THE RACE TO MAKE THE LASER

Theodore Maiman and Irnee D’Haenens display the first laser a quarter century after theymade it.

Hughes Research Laboratory, courtesy AIP Emilio Segrè Visual Archives

Sharp spike of laser pulse above threshold

A small spike of stimulated emission at threshold

Fluorescence only below threshold


difference, but sensitive light detectorsand electronic instruments could mea-sure it on an oscilloscope screen. Whenlaser action began and stimulated emis-sion overwhelmed spontaneous emission,Maiman expected the peak of the pulseon the oscilloscope screen to becomeshorter and sharper, followed by a rapid decline. That was one test for aworking laser.

Another test was even subtler.Maiman expected stimulated emissionfrom ruby to span a much narrowerrange of wavelengths than spontaneousemission. Ruby fluoresces across a rangeof wavelengths that all look red to theeye, but they are not the same. The fluo-rescence is most likely to be at the centerof the range, with the probability drop-ping off at the sides, so the light intensityvaries in the same way.

Stimulated emission spans the samerange of wavelengths as the fluorescence,but it amplifies the initial fluorescence.That amplification process concentratesstimulated emission in a much narrowerrange of wavelengths, as if the emissioncurve was multiplied by itself manytimes. The gain is highest in the middleof the curve, so that wavelength is ampli-fied more than wavelengths to the side.

If the fluorescence intensity varies by afactor of two, the difference will double ateach round of stimulated emission. Laseremission should span a much narrowerrange of wavelengths than fluorescence,with the exact variation depending on thenature of the laser material and the reso-nant cavity. Maiman expected a goodspectrometer to show this line-narrowingeffect, shrinking the band of red fluores-cence narrow to a laser line, as shown inthe figure at the top right of the page.

Both pulse-shortening and line-nar-rowing are closely related to the exis-tence of a distinct threshold for laseraction. When input power is low, noth-ing happens beyond the emission ofa little fluorescence, which increasesslightly as power rises. When the inputpower escalates above a threshold value,stimulated emission takes over and thelaser turns on, with power increasingmuch more rapidly. The effect is likepushing an object along an increasingdownhill slope until it finally startsmoving on its own.

Stimulated emission becomes signifi-cant once the pump energy produces apopulation inversion. However, such aninversion is not enough to reach laserthreshold. Only when the gain from stim-ulated emission equals the internal lossesdoes the laser reach threshold. For exam-ple, if one percent of the light is lost oneach round trip of the laser cavity, andanother percent emerges as the beam, thegain on a round trip of the laser cavitymust be two percent in order to reachlaser threshold.

Once you exceed the threshold, thelight power increases sharply. If the risein power— the gain—is only one percentper pass, the power increases 64.4 percentafter 50 round trips and cascades upwardfrom there. After 100 round trips, thepower is 170 percent (a factor of 2.7)higher; after 500 round trips, it increasesby a factor of 145. Since the round trips

are at the speed of light, a nearly instanta-neous jump in power can be observed asthe laser exceeds threshold and turns on. The power doesn’t grow infinitelybecause only a limited number of atomscan be excited at any instant, and theamplification process saturates.

Maiman’s plan was to slowly crank upthe power applied to the flashlamp andwatch what happened. He expected to seeordinary red fluorescence at lower power,increasing gradually until he reachedthreshold. Above threshold, the laserwould turn on, emitting a much brighterred beam. Crossing laser threshold wasimportant, but he knew that accomplish-ing that alone would not convince skepti-cal physicists. He wanted to record bothpulse shortening and line narrowing onhis instruments.

Maiman and D’Haenens conductedtheir first test of the ruby laser on the


Ruby laser pulses rise and fall much faster than fluorescence produced by lower-powerillumination. When Maiman started with modest flashlamp pulses, he saw only the riseand slow fluorescence (bottom). At laser threshold, a small spike of stimulated emission(center). When the flashlamp power exceeded the laser threshold, stimulated emissionextracted energy from the ruby in a brief burst much shorter and more intense than thefluorescence he saw at lower flashlamp powers (top).

Wavelength

Above laser threshold,R1 line is narrower and much stronger than R2

Below laser threshold, bothlines are wide and R1 is onlyabout twice as strong as R2

R2 R1

R2 R1Wavelength

Laser operation narrowed the range of wavelengths and shifted the balance between thetwo red lines of ruby. Spontaneous emission from ruby (bottom) was fairly evenly dividedbetween two lines called R1 and R2. Above laser threshold, laser oscillation narrowed theR1 line and greatly increased its power, leaving the R2 line weak (top). Maiman had pre-dicted this effect, and he considered observing it proof of laser operation.

afternoon of May 16, 1960. Electricalcables connected a power supply to thetwo ends of their flashlamp. Anothercable linked the trigger electrode to asource of high voltage pulses synchro-nized with an oscilloscope. The heart ofthe power supply was a big capacitor,which slowly accumulated a charge andheld it until the trigger voltage ionizedthe xenon in the lamp and allowed thecapacitor to discharge its energy throughthe lamp.

Linking the trigger voltage to theoscilloscope synchronized the firing ofthe lamp with the trace drawn acrossthe oscilloscope screen, so Maiman andD’Haenens could compare the sequenceof events for many different pulses asthey changed the voltage across the lamp.The trace measured the amount of redlight registered by their sensor. It sweptacross the screen at a constant rate, withhorizontal markings measuring the

passage of time, and vertical markingsgauging voltage from the detector. Theresearchers set the trigger voltage to fire at a certain point in the cycle. A memoryinside the oscilloscope recorded the tracesfrom successive pulses.

They aligned the laser package to focus the ruby output into theirmonochromator, which spread the lightinto a spectrum. Then they directed athin slice of that spectrum to a sensitivephotomultiplier tube, where the photonstriggered a cascade of electrons that drovethe oscilloscope. The oscilloscope tracerose and fell with the strength of the sig-nal, measuring how much light reachedthe tube.

Each sweep showed how emissionfrom the ruby varied as the pulse of lightfrom the flashlamp hit it. Maiman hadbigger flashlamps in reserve, just in casethe little one didn’t have enough power topush ruby above the laser threshold.

They started firing the flashlamp withpulses of 500 V, a modest level for theirlamp, and well below the threshold forlaser action. When they fired the lamp,it flashed, and the oscilloscope showed a trace of light rising in power, thendropping in about 3 ms—just what theyexpected from ruby fluorescence. Theyadjusted the knobs and started turningup the voltage step by step, expecting the power in the ruby pulse to increaseincrementally.

The experiment was simple andrepetitive: Turn up the voltage, fire thepulse, and look at the oscilloscope trace.Today it would be computer controlled,but in 1960 Maiman and D’Haenens didit by hand, firing a single pulse that trig-gered both the flashlamp and the oscil-loscope scan. Gradually the power levelincreased, but the more powerful pulsesdropped with the same 3 ms decay time.The aluminum cylinder contained mostof the light from the flashlamp, butbright flashes of ruby fluorescence scat-tered red light through the room.

Maiman and D’Haenens kept theireyes on their instruments, but the redflashes inevitably reached their eyes. Thebrightness of the ruby pulses increasedsteadily as they turned up the voltage.Their instruments showed that theincrease was proportional to the extraenergy delivered to the lamp—a signthat everything was working properly.They continued to increase the voltage,and watch their instruments each timethey fired the flashlamp. The oscillo-scope trace was a heartbeat on thescreen. The flash brightness increased in small steps until they turned thepower supply above 950 V. Maiman saw the oscilloscope trace surge.

“The output trace started to shoot upin peak intensity and the initial decaytime rapidly decreased,” Maiman wrote.“Voilà. This was it! The laser was born!”

As the laser surged to life, its brilliantred glow permeated the room. The twomen had aimed the cylinder at a piece of white poster board, and Maiman’sred-dazzled eyes were focused on hisinstruments. The beam formed a redhorseshoe-shaped spot a few degreeswide on the poster board. D’Haenensjumped with joy. Maiman felt numb,




relieved that his gamble had paid off.They basked in the red light of glory.

Others in the atomic physics depart-ment came to take a look. Bob Hellwarth,who had brainstormed with Maiman onhow to recognize laser action, congratu-lated him. Harold Lyons was delighted:His department had beaten Bell Labs to a prize that Bell had spent far more timeand money seeking.

The next morning, Lyons showed up bright and early at Maiman’s office,with plans to put out a press release. ButMaiman was not ready to go public. Hewas proud of his achievement, but he alsoinsisted on precision. His first experimenthad left some doubt in his mind. He hadnot seen as sharp a jump in laser poweras he had expected at laser threshold. Hewanted to track down the cause of theanomaly. He suspected the problemstemmed from imperfections in the ruby crystal.

Maiman immediately ordered threenew ruby crystals, and asked the supplierto cut them into rods and polish theirends, because he wasn’t sure the Hughesshop could handle the hard ruby crystalsproperly. Indeed, Asawa soon found thatthe first ruby rod had internal flaws thatcaused light scattering. In addition, theend faces had not been made preciselyparallel, so the light did not bounce back and forth properly between them.Unfortunately, only a single company, adivision of Union Carbide, could supplythe ruby crystals, and Maiman wouldhave to wait weeks for delivery. For themoment he had to do the best he couldwith the original.

In the first experiment, Maiman andD’Haenens had concentrated on the easi-est measurement—the change in howfast the red emission dropped. Thedecrease they saw was consistent withlaser action, but Maiman sought moreevidence. He wanted to show that therange of wavelengths in the ruby outputnarrowed when stimulated emissiondominated above laser threshold.

He also wanted to verify the presenceof a line narrowing specific to laser light.Pink ruby fluoresces at two closely spacedred lines, R1 and R2, at wavelengths of694.3 and 692.9 nm, respectively, at roomtemperature. This happens because themetastable upper laser level is actually a

pair of closely spaced energy states. Whenthe chromium atom is in the lower state,it emits on the R1 line; when it is in theupper one, it emits at the shorter-wave-length R2 line. When ruby fluoresces, itemits about twice as much light at thelonger R1 wavelength; Maiman expectedstimulated emission to increase the dom-inance of the R1 line.

Maiman’s modest monochromatorcouldn’t make the required measure-ments. He needed an expensive high-resolution spectrograph. Another Hughes physicist, Ken Wickersheim,had just received exactly the instrumentthat Maiman needed, but he had a back-log of experiments to perform because hehad been waiting for it for six months.At Maiman’s request, Lyons pulled rank and commandeered the new instru-ment so that Maiman and Asawa couldperform additional tests. The furiousWickersheim, an old classmate ofAsawa’s, took an extended camping vaca-tion in the Sierra Nevadas to cool off.

Maiman and Asawa got results quickly.When Asawa ran the ruby output throughthe new spectrograph, he recorded theline narrowing that Maiman had pre-dicted. The light was so bright that it

initially blackened the photographicplates that they had used to record thespectrum. Asawa had to insert high-den-sity filters or spread the light in order torecord the spectrum properly. The spec-trograph also clearly separated the twoclosely spaced red lines. Asawa found thatthe R1 line was at least 50 times brighterthan the R2 line, confirming another ofMaiman’s predictions. Maiman andLyons were delighted; they had resultsgood enough to publish, and a big suc-cess to report.

After this second experiment, whichfinally convinced Maiman that the laserwas working properly, word spreadaround the lab in a matter of hours. ButMaiman still had to persuade the outsideworld that his laser worked, and thatwould pose a new round of challenges . . .

Jeff Hecht ([email protected]) is a contributingeditor at Laser Focus World, a correspondent for New

Scientist and author of City of Light: TheStory of Fiber Optics, Understanding FiberOptics and Beam: The Race to Make the Laser.

Sources

Primary sources: Theodore Maiman, The Laser Odyssey(Laser Press, 2000); interviews with Charles Asawa andIrnee D’Haenens.

Secondary sources: Telephone interviews with GeorgeBirnbaum, Richard Hellwarth and Bela Lengyel; docu-ments collected by Laser History Project.


Member

“The output trace

started to shoot up

in peak intensity and

the initial decay time

rapidly decreased,”

Maiman wrote.

“Voilà. This was it!

The laser was born!”

When it came time to announce Maiman’s laser, Hughes’s public relations firm hired a topphotographer to take pictures. The photographer liked to get shots of people behind theirinventions, but Maiman’s first laser was too small, so he insisted on using a larger flashlampand ruby rod. This photo was widely published as the “first” laser, and when other researcherstried to duplicate Maiman’s work, they used this larger flashlamp to pump their ruby rods.Although they had the wrong lamp, their experiments worked within a few weeks.

beam: the race to make the laser

Documents