History of Gas Lasers Part 1—Continuous Wave Gas Lasers

Jeff Hecht

In this first of a two-part series, Jeff Hecht relives the excitement that accompanied the development of the first gas lasers to generate continuous-wave beams.

eveloping the concept of “light amplification by the stimulated emission of radiation” in a resonant cavity was a crucial step D on the road to the laser. But making a working laser required finding a suitable laser medium. Theodore Maiman’s ruby laser proved that optically pumped solids were viable laser materials, but gases were also attractive candidates because their properties were well-understood. Seven months after Maiman’s success with ruby, the helium-neon laser became the first type to emit a continuous beam rather than pulses.

The quest for continuous wave lasers Most early laser developers sought four-level laser materials that could sustain a steady population inversion so that they could generate a continu- ous-wave (CW) beam. Years of gas spectroscopy experiments had generat- ed extensive tables of spectral lines, which could be mined for promising transitions. Developers studied two approaches to producing population inversions in gases—optical pumping and discharge excitation of the gas.

16 | OPN Optics & Photonics News www.osa-opn.org 1047-6938/10/01/0016/8-$15.00 ©OSA Courtesy of William Bennett

Don Herriott, Ali Javan and William Bennett (left to right) with the first helium-neon laser at Bell Labs. (Although Bell Labs officially banned alcohol, the beaker in Herriott’s hand holds a celebratory liquid supplied by their technician, Ed Ballik.)

January 2010 | 17 Developed in the early 1950s by As Herriott fiddled flat mirrors precisely parallel to each Alfred Kastler at the École Normale other at opposite ends of an 80-cm tube. Supérieure in Paris, optical pumping restlessly with a They weathered mishaps, including can directly excite specific transitions. mirror adjustment, melting a laser tube and destroying Optical pumping of alkali metal vapors mirror coatings. After hours of seem- was suggested for laser excitation both in Javan glanced at an ingly fruitless tests on a new tube, the Gordon Gould’s patent applications and oscilloscope screen three talked as heavy snow fell outside in the pioneering 1958 paper on infrared late in the afternoon of December 13, and optical masers by Charles Townes and saw the type of 1960. As Herriott fiddled restlessly with and Arthur Schawlow. The concept of a mirror adjustment, Javan glanced optical pumping was attractively simple, signal they had sought at an oscilloscope screen and saw the but its efficiency was low in low-density all day. Herriott had type of signal they had sought all day. gases or metal vapors, and the reactiv- Herriott had hit the sweet spot of cavity ity of alkali metal vapors created major hit the sweet spot of mirror alignment. experimental complications. cavity mirror alignment. Sometimes such momentary successes Discharge pumping was well estab- vanish as mysteriously as they appear, lished for generating red light in neon but this one was stable. Their mono- tubes and exciting mercury atoms to help with the experiments, he persuaded chromator showed that the laser was emit ultraviolet light in fluorescent Bell to hire William Bennett, who had oscillating on a predicted line at 1,153 lamps. Gould suggested discharge recently finished his dissertation on col- nm. After they made a few more adjust- excitation in his patent applications, lisions of the second kind at Columbia. ments, word spread through the lab and and he eventually received a patent on However, Bennett couldn’t start until a stream of visitors came to see the first collisional excitation. Yet discharge the summer of 1959. continuous-wave gas laser. After a few pumping posed challenges, including In the meantime, Oxford Univer- days of experiments, including sending selective excitation of the desired states sity physicist John Sanders came to their voices across the room by speaking and assuring gas purity. Bell for an eight-month sabbatical and close to a cavity mirror, the three sub- The first researcher to examine dis- decided to test another idea that Javan mitted a paper to Physical Review Letters. charge pumping for lasers in depth was had suggested—discharge excitation of Ali Javan, who earned his Ph.D. under pure helium. Lacking time for a detailed The red helium-neon laser Townes at Columbia for research on analysis, Sanders zapped the gas, but he The New York Times relegated the first microwave spectroscopy. In mid-1958, failed to see the cascade of stimulated gas laser to p. 39, next to a story about he interviewed for a job at Bell Labs, emission he had hoped for on a 668-nm X-raying a tusk of a baby walrus named where Schawlow told him about the laser helium line. Despite the slow progress, Ookie. But the feat generated widespread concept. Intrigued, Javan rushed back Bell encouraged both Sanders and Javan interest among scientists, and the Army to Columbia that afternoon and began to publish their preliminary results in Signal Corps at Fort Monmouth, N.J., investigating laser concepts. By the time the summer of 1959. The company sug- contracted Bell to build another helium- he started work at Bell in August, Javan gested this because it wanted to discour- neon laser. had convinced himself that discharge age the Pentagon from classifying all Bell’s basic research department was pumping was the best route to practical laser research after having awarded TRG busy trying to develop more new lasers, gas lasers. Inc. a $1-million contract to build a laser so managers routed the contract to Alan Javan proposed a two-step excitation based on Gould’s proposals. White and Dane Rigden in the explor- process. First, electrons would collide Javan and Bennett put in long hours atory development department. The two with helium atoms, exciting them to studying the helium-neon system. They had been working on gaseous electronic a higher energy level. Then stimulated enlisted the help of Bell optics special- devices. Thanks to the earlier experiment helium atoms would transfer their extra ist Donald Herriott to build a high- and the availability of Brewster windows energy to the less-abundant neon atoms, reflection optical cavity to push their and concave mirrors, they finished the exciting them to metastable states with low-gain laser above threshold. Progress new He-Ne laser much faster and were energies close to those of the excited was slow because they were exploring able to add improvements such as fine- helium—a process called collisions of completely new territory. They had to tuning the discharge. the second kind. Javan expected this to develop ways to measure energy transfer, Curious about what more they could produce a population inversion on an energy-state lifetimes and laser gain as do, they began further experiments. “It infrared neon transition. He proposed well as to make high-reflectivity mirrors was all done evenings and weekends, a step-by-step plan to demonstrate and that could survive within a discharge because our regular work had to go on verify gain, then try to build a laser. To cavity. Then they had to align a pair of as usual,” White recalls. But they didn’t

18 | OPN Optics & Photonics News www.osa-opn.org Alan White at Bell Labs, working on a red helium-neon laser in a very cluttered laboratory, which captures the place well.

Courtesy of Alan White mind because they were excited to be multimode beam bright enough to see for an MIT professorship in mid-1961. working on lasers and coherent light. on the laboratory wall. Unsure how Then a young Bell physicist, C. Kumar They improved stability and reduced long the tube would last, they called in N. Patel reduced helium pressure in a noise, substituting a hot-filament direct- more witnesses. helium-neon laser and observed lasing current discharge for the radio-frequency The visible red beam excited every- in pure neon. discharge that Javan and Bennett had one, including management. “Almost Bennett, Walter Faust and Ross used. Output power increased, reveal- immediately large amounts of money McFarlane then joined with Patel to ing previously unseen details in the laser came to us, and there was no need to search for more new laser gases and spectrum—including a new metastable work nights or weekends,” White recalls. lines. Concave mirrors and the avail- helium state. Thinking that the new Reported in 1962, the red helium-neon ability of red helium-neons for align- state might excite a laser emitting on the laser became the most familiar gas laser, ment made experiments easy. They 632.8-nm neon line, they ordered mir- widely used in classroom demonstra- observed laser action on the atomic lines rors with peak reflectivity in the red. The tions, laboratory experiments, hologra- of neon-oxygen, argon-oxygen, and evening after the mirrors arrived, White phy and construction alignment. pure argon, krypton and xenon. They recalls, “We put the first gas in the tube, measured emission on dozens of lines lined up the concave mirrors, and bingo, Other early gas lasers in the visible and infrared, including it went. We were three excited people,” some beyond two micrometers. Bennett including a witness they had invited in The basic research group’s helium-neon returned to Yale in the fall of 1962, but case their idea worked. experiments showed that excited neon by early 1963 Bell had counted more They first saw only a little sparkle atoms could transfer energy to oxygen than 150 laser lines and showed that gas when they looked down the laser tube molecules, and so they wondered if a discharges could readily produce popula- (in a practice that would horrify any mixture of neon and oxygen could lase. tion inversions. Yet only a tiny fraction modern safety officer). Adjusting the Bennett saw a population inversion in of the input energy emerged as light, mirrors and the discharge made the a pure neon discharge after Javan left and output power was limited. It took

January 2010 | 19 enough power for anything you wanted to do in the laboratory,” he recalled. Reaching that power level and 20 percent efficiency was enough to interest military laser-weapon developers because it promised much better efficiency, power and heat dissipation than solid- state lasers could deliver. The Pentagon began classifying high-power CO2 research, so Patel, a non-citizen lacking a clearance, decided to stay with spectro- scopic research.

Laser companies and ion lasers

Stephen Jacobs Gas laser development quickly spread with an early cesium beyond Bell Labs, which, under terms laser at TRG. of a 1950s consent decree, had to license

Courtesy of Stephen Jacobs its inventions to other companies. Soon after being founded in September 1961, a 15-m tube to generate 150 mW from a 1985 interview, recalling surprise that Spectra-Physics teamed with the well- helium-neon. his calculations were so close to the mea- established Perkin-Elmer to manufacture Optical pumping of alkali-metal sured results. “It worked marvelously helium-neon lasers. They exhibited a vapors proved a disappointment. well,” he said. “We got tens of milliwatts 1.15-µm version selling for about $8,000 Townes’s students at Columbia quickly on the first shot.” He then realized that in March 1962, and sales jumped after abandoned their cesium laser project diatomic molecules should also work, they introduced a red version six months th after Bell demonstrated the helium-neon and he tried carbon monoxide, which later. In June 1963, they sold their 75 laser. The Pentagon didn’t think TRG’s also lased. laser, and the two companies went their metal-vapor laser research was worth Molecular nitrogen soaks up dis- separate ways. classifying, so it was the only project charge energy efficiently, so Patel added Industry was also quick to recognize the potential of 10.6-µm CO lasers for that Gould could work on after having it to CO2, hoping for energy transfer 2 been denied a security clearance. After from the long-lived first excited state noncontact cutting and drilling of non- metals. Spectra-Physics founder Eugene painstakingly measuring population of N2 to an upper level of CO2. Power Watson saw the possibilities the first inversions and optical amplification in jumped from 10 mW with pure CO2 to cesium, Steve Jacobs and Paul Rabinow- 10 W from the gas mixture, the highest time he saw a CO2 laser at a 1965 meet- itz detected oscillation on a 7.18-µm CW power that had then been seen from ing, and when the Spectra-Physics board cesium line in early 1962. Gould read a laser. Adding helium also helped. “By refused to approve his plans to develop success on their faces when he walked mid-1965, I had a 200-watt continuous- CO2, Watson quit to establish Coherent Radiation Laboratories (now Coherent into his office one Monday morning, wave CO2 laser, which was more than saying, “Well, I’ll be damned. You Inc.). The new company landed a con- made it work!” tract to build a 100-W laser and set up After painstakingly shop in Watson’s home. Within months, they had the laser up and running, and Molecular gas lasers measuring population they demonstrated it by cooking paint In 1963, Patel realized that molecular inversions and optical on the garage door of an obnoxious lasers might convert more input energy neighbor across the street. into light than atomic lasers because amplification in ces- Laser companies contributed their molecular transitions were much closer own innovations. Early helium-neon to the ground state. He started study- ium, Steve Jacobs lasers could be short-lived, so Spectra- ing carbon dioxide because he thought and Paul Rabinowitz Physics co-founder Earl Bell tried add- the multiple series of vibrational states ing mercury vapor to the gas mixture in three-atom molecules should allow detected oscillation on to extend the laser’s lifetime. He saw metastable states. He calculated that a 7.18-µm cesium line a green glow near the cathode, which CO2 should lase near 10 µm. “It did the hadn’t appeared in ordinary He-Ne first time we tried,” he remembered in in early 1962. lasers, and he thought that might lead

20 | OPN Optics & Photonics News www.osa-opn.org to a mercury laser. Applying a standard continuous helium-neon discharge The first continu- ous wave argon-ion source across a long tube laced with lasers made at Bell mercury revealed nothing new. He Labs. The complex recalled later, “I then decided to dis- tube construction charge a high-voltage capacitor charged was necessary in by a neon sign transformer through the order to keep the electric discharge tube at 120 Hz and—Wow!” First he from going the saw laser emission on a new red-orange wrong way. line; later in the day, he saw a green laser line. The laser was pulsed, but the power levels were encouragingly high, and, with visible lines few and far between, Bell and his Spectra colleague Arnold Bloom were excited. Initially they thought the emis- sion came from neutral mercury, but comparing measurements with wave- length tables revealed the lines were from mercury ions. That was a surprise; ions hadn’t been considered suitable for lasers because they were high above the neutral ground state. But it was encour- aging because ions tend to have higher Courtesy of Colin Web transition energies than neutral atoms, offering the potential for shorter-wave- what was ostensibly a helium-mercury enough to identify all 10 lines. Bennett length lasers. laser. We now had a blue line at 4,880 and Convert discovered the argon lines The mercury laser never found a Ångstroms in addition to the red and independently, but Bridges published commercial niche because it didn’t green lines from mercury.” first. “Lines just tumbled out all over the operate continuously in a conventional Anxiety mixed with his excitement. place” in tests of krypton, xenon and discharge. Its lasting impact was what “You see something unexpected and rare-gas mixtures, he says, but he lacked inspired research into other ion lasers by furthermore you don’t quite know cavity optics to produce the ultraviolet many people, including William Bridges how you produced it. Maybe it will go lines of neon ions. at Hughes Research Laboratories, Ben- away, and you won’t get it back again.” Even before his paper appeared in nett at Yale, Dane Rigden at Perkin- Worried that the new line might be print, Bridges told Gordon at Bell Labs Elmer, Gene Gordon of Bell Labs, Steve from an unknown contaminant in his about the pulsed argon-ion laser. A Jarrett at TRG, Guy Convert at CSF in welding-grade argon, he left technician few weeks later, Gordon stunned him France, and Grant Fowles and William Bob Hodges to watch the laser run as by calling to announce that “we’ve got Silfvast at the University of Utah. he searched Hughes’s library, where he ours going continuous wave.” Bell had found the line probably came from used its own high-performance mir- ionized argon. rors and a capillary discharge only a Argon-ion lasers Unable to remove all the mercury millimeter in diameter, yielding current After Bridges got his own pulsed from the tube, Bridges rush-ordered densities 25 times higher than the helium-mercury laser running, he began a new tube and tested it with pure 5-mm Hughes tubes. investigating energy transfer. First he argon. Identifying the 10 argon lines he However, Bell’s success came only replaced helium with neon and demon- observed required mounting a series of after the intense heat from the dis- strated a neon-mercury laser. Then he relay mirrors in the halls to route the charge melted uncooled glass and quartz added argon as a buffer gas, but he put beam through a few hundred feet of tubes. Water cooling quartz solved that in too much and couldn’t get the mer- halls separating the immobile laser from problem, but the discharge then pumped cury laser lines. On February 14, 1964, the equally immobile high-resolution the gas all to one end of the tube. When he pumped out the tube, flushed it and spectrometer. Working at night when they added a return loop, the discharge put helium back into the tube to check the lab and the halls were empty, he went the wrong way, so a Bell glassmaker mirror alignment. He recalled, “To our and Hodges measured wavelengths devised an elaborately curved return to surprise, we had a new line going in to a few hundredths of an angstrom, block the discharge, allowing CW laser

January 2010 | 21 A giant 1.5-kW carbon-dioxide laser built by Hughes Research Laboratories, sitting on top of big sheets of plywood.

Courtesy of Bill Bridges action, recalls Colin Webb, who worked powder everywhere: The disks looked replacement, L’Esperance paid $25 to with Gordon. like nothing so much as burned barbe- a nearby crane operator, who slipped Bennett made a long-pulse argon cue briquettes.” Indeed, the graphite had it flawlessly inside, where the surgeon laser, and when he described it at a New burned, because it hadn’t been degassed used it to develop a laser treatment that York conference, he said CW operation properly, but that was solved by induc- has preserved the vision of millions of would require impossible amounts of tion heating in a high vacuum. people with diabetic retinopathy. power. Gordon, who was in the audience, The brightness of ion lasers at visible stood up to announce that Bell had made wavelengths earned ion lasers some Metal vapor lasers its own argon-ion laser pulsed with a important applications, but their tough The helium-mercury laser also inspired one-in-three duty cycle. “We switch it design requirements caused problems. At the discovery of other metal-vapor on in the morning and switch it off at Hughes, Bridges developed argon lasers lasers by Fowles and Silfvast at Utah. night,” he said. Bridges then used the for an Air Force night reconnaissance They first tried to make a bismuth laser Bell design with a larger power supply to system; the results were good, but it to study hyperfine structure, which is generate 80-mW CW from krypton and never went into production because the particularly strong in the heavy metal. xenon at Hughes. cooling system didn’t meet requirements Silfvast developed a simple quartz tube The argon laser was off and running. for installation in military planes. apparatus to vaporize bismuth and test However, commercial development was Eye surgeon Francis L’Esperance the vapor for pulsed laser action, but he a challenge for an ion laser with transi- ran into a different problem when he grew discouraged after a few months of tions so far above the ground state that ordered one of the first commercial tests found no laser lines. In early 1965, only about 0.05 percent of the input argon lasers from Raytheon. At more he decided to try zinc and cadmium, energy emerges in the beam. Jarrett, than two meters in length, it was too big which looked like good laser prospects who developed a segmented graphite to fit into the elevators at the Columbia- because of their electron configurations. bore for a white-light krypton-ion laser Presbyterian Medical Center in New He tried zinc first. “The very first as a co-founder of Coherent, recalled York. They hired a rigger to hoist the time I turned it on I got this turquoise, that initial tests were discouraging: massive laser through a window, but he blue-green transition at 492.4 nm to “There were signs of erosion and graphite dropped it. When Raytheon shipped a lase,” he recalled. Overjoyed, he hunted

22 | OPN Optics & Photonics News www.osa-opn.org down Fowles at a faculty meeting, and In the mid-1960s, the fluoride emitting at 2.6 to 3.0 µm, and the professor came running. After a deuterium fluoride emitting at 3.6 to few days of studying zinc, they tried Advanced Research 4.0 µm; the latter wavelengths were bet- cadmium, which also lased, although Projects Agency ter transmitted by the atmosphere. Both not on the now-familiar 441.6-nm blue have reached megawatt-class powers, cadmium-ion line. Other metals they had contractors but the latest megawatt-class laser test- could vaporize followed, including lead, build gigantic pulsed bed, the Airborne Laser, has shifted to which emitted a 723-nm line so strong the chemical oxygen-iodine laser, which that it lased even with blue-reflecting discharge-driven emits at 1.3 µm, allowing smaller optics mirrors on the laser tube. It was the first and better beam transmission. in a family of high-gain, self-terminating CO2 lasers with pulsed neutral atom lasers that includes average power well Looking back copper and manganese. (For more on pulsed gas lasers, make sure to read the above a kilowatt. As the first continuous-wave lasers, gas second part of this history in the Febru- lasers laid the foundation for today’s ary OPN.) laser industry. The venerable red helium- The blue He-Cd line came later, neon laser was the first to be widely used when Silfvast tried adding helium and laser weapon development, and it so in industry, and it was the standard a weaker electric discharge to make a impressed military brass that they kept demonstration laser for decades. Ion pulsed laser. A few months after mov- the results classified until 1970. lasers pioneered important applications ing to Bell Labs in August 1967, he Gas dynamic lasers eventually topped in ophthalmology, biomedical instru- made the first continuous-wave He-Cd out at a few hundred kilowatts in the ments and printing. CW gas lasers are laser by running a steady low-current Air Force Airborne Laser Laboratory giving way to diode and solid-state discharge at the proper vapor pressure, built in the 1970s. By then, high-energy lasers for most visible and near-infrared military laser development had shifted using equipment that wasn’t available applications, but the CO2 laser remains at Utah. In 1972, Silfvast made the to chemical lasers that produced other dominant for industrial applications at helium-selenium laser, which emitted vibrationally excited molecules. The longer infrared wavelengths. It’s been a first was hydrogen chloride, emitting at simultaneously on up to 46 lines, but long and remarkable run. t never proved practical. 3.7 µm, demonstrated by J.V.V. Kasper and George C. Pimentel in 1965 at the Jeff Hecht ([email protected]) is a University of California at Berkeley. Member High-power gas lasers science and technology writer based in Military developers preferred hydrogen Auburndale, Mass., U.S.A. In the mid-1960s, the Advanced Research Projects Agency had contrac- [ References and Resources ] tors build gigantic pulsed discharge- >> A.L. Schawlow and C.H. Townes. “Infrared and Optical Masers,” Phys. Rev. 112, 1940 (1958). driven CO2 lasers with average power >> A. Javan. “Possibility of producing of negative temperature in gas discharge,” Phys. Rev. Lett. well above a kilowatt. The big break- 3, 87-9 (1959). through in CW gas laser power—to >> J.H. Sanders. “Optical maser design,” Phys. Rev. Lett. 3, 86-7 (1959). tens of kilowatts—came in 1966 when >> A Javan et al. “Population inversion and continuous optical maser oscillation in a gas dis- Edward Gerry and Arthur Kantrowitz charge containing a He-Ne mixture,” Phys. Rev. Lett. 6, 106-10 (1961). of the Avco Everett Research Laboratory >> W. Sullivan. “Bell shows beam of ‘talking’ light,” New York Times, Feb 1, 1961, p. 39. >> A.D. White and J.D. Rigden. “Continuous gas maser operation in the visible,” Proceedings in the Boston suburbs demonstrated a IRE 50, 1697 (1962). radically new design, the gas dynamic >> W.B. Bridges. “Laser oscillation in singly ionized argon in the visible spectrum,” Appl. Phys. CO2 laser. Lett. 4, 128-130 (1964); erratum Appl. Phys. Lett 5, 39 (1964). Their inspiration was realizing >> J.V.V. Kasper et al. “HCl chemical laser,” Phys. Rev. Lett 14, 352 (1965). that extracting only 0.1 percent of the >> E.T. Gerry. “Gasdynamic lasers,” IEEE Spectrum 7(11), 51 (1970). gigawatt-class power from a rocket >> G. Gould. U.S. Patent 4,161,436, “Method of energizing a material,” issued Jul 17, 1979. >> G Gould. U.S. Patent 4,704,583, “Light amplifiers employing collisions to produce a popula- engine would produce a megawatt-class tion inversion,” issued November 3, 1987. laser. Their plan was to burn a carbon- >> J.L. Bromberg. The Laser in America 1950-1970, MIT Press, Cambridge, 1991. containing fuel and expand the hot gas >> J. Hecht. Laser Pioneers, Academic Press, 1991. at high velocity through nozzles into >> N. Taylor. Laser: The Inventor, the Nobel Laureate, and the 30-Year Patent War, Simon & a low-pressure laser cavity, producing Schuster, N.Y., 2000. >> S.M. Jarrett. “Early Ion Laser Development,” Opt. Photon. News 15(10), 24 (2004). a population inversion in CO2 mol- ecules. Their success at generating 50 >> J. Hecht. Beam: The Race to Make the Laser, Oxford, New York, 2005. kW in 1966 inspired a new round of >> J. Hecht. “Half a Century of Laser Weapons,” Opt. Photon. News 20(2), 14-21 (2009).

January 2010 | 23