Ion lasers-the early years - Selected Topics in Quantum Electronics, IEEE Journal on

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

885

Ion Lasers—The Early Years

William B. Bridges

, Life Fellow, IEEE

Invited Paper

Abstract—

This paper is a personal, anecdotal history of the dis-

covery and early development of ion lasers, particularly the argon

ion laser. A brief discussion of the mechanisms that make this laser

work, and the engineering challenges and developments that make

it practical are included. Some early applications in night recon-

naissance and imaging are included.

I. W

HAT

HIS

APER IS

BOUT

HEN I was invited to write a paper on the early days

of ion lasers for this special issue, I hesitated. So much

was done by so many people so long ago, that it seemed impos-

sible to put it all into a single paper with any kind of balance.

In the end, I decided to write about the early days from a per-

sonal perspective, rather than try to include all developments in

all the organizations. If I can convey the rampant excitement of

those early years that surrounded the ion laser, both its devel-

opment as a device and its prospects for application, then I will

have succeeded, even if the perspective is that of a researcher

in a large aerospace company, driven by military applications.

Someone else will have to write the story of the rise and fall of

the commercial laser manufacturers and applications that even-

tually came to dominate the marketplace (and all the interesting

lawsuits over intellectual property that ensued).

Perhaps this perspective is not too bad for the very early days

of ion lasers, since most of the players were researchers in large

systems organizations: Hughes, Bell Telephone Laboratories,

Raytheon, RCA. Only Spectra Physics, then a relatively small

company, newly formed by people from Varian Associates but

growing rapidly as a successful manufacturer of helium–neon

lasers, was “commercial.” Bankrolling the research and devel-

opment was the Department of Defense. (Even Spectra Physics

had contracts from the Air Force). This was psychologically a

different era. It never crossed my mind to leave Hughes and seek

venture capital to start up a small ion laser manufacturing com-

pany, something that would happen in an instant if it were dis-

covered today! So perhaps this paper can serve as a window on

those earlier times, as well as a partial record of the technical

history.

I have included the mistakes and side alleys, as well as suc-

cesses and the mainstream. I believe it is important for engineers

to know that progress is not made in a simple straight line of

successes (although that is always what we write in proposals

Manuscript received September 26, 2000.

The author is with the California Institute of Technology, Pasadena, CA

91125 USA.

Publisher Item Identifier S 1077-260X(00)11498-4.

for new work!) I have chosen to include people and their first-

hand interactions in this account. History is made by people, not

organizations or projects or programs. The risk of including in-

dividuals in any narrative is that you will leave someone out. I

admit up front that I have “left out” literally hundreds of people,

even from these earliest days (check the bibliographies of any

of the summary articles listed in the References). To these, I can

only say “I’m sorry...But those were really great times, weren’t

they?”

II. I

N THE

EGINNING WAS

ERCURY

Strictly speaking, the first ion laser was Maiman’s ruby laser

in 1960, since the energy levels involved are those of triply ion-

ized chromium, Cr

, albeit in an aluminum-oxide crystal host.

But this paper is about ions that lase in a gaseous or vapor form,

and the first of these was singly ionized mercury, Hg

, demon-

strated in late 1963 by W. Earl Bell of Spectra Physics [1]. He

obtained oscillation at 615 nm in the red–orange, and 568 nm

in the green, as well as two infrared lines at 735 and 1058 nm,

in a pulsed helium–mercury discharge. These four wavelengths

were known to originate from energy levels in singly ionized

mercury, making them members of the “second spectrum” of

mercury, Hg II. The reported output power was of the order of

tens of Watts for microseconds, and the repetition rate was high

enough to make the laser output quite visible by scattering from

dust in the air. Fig. 1 appeared (in color) in a Spectra Physics ad-

vertisement in 1964, and shows Bell posed next to a laboratory

mercury ion laser with a striking green laser beam exiting. (This

figure is one of many in my “A Boy and His Laser” photograph

collection.)

Since this was the first green laser, it created a bit of

excitement, at least in the organization I then worked for, the

Hughes Research Laboratories (HRL) of the Hughes Aircraft

Company, and I decided to make one in my laboratory. My

scientific “excuse” was that Bell had not speculated in his

paper on the mechanisms involved that lead to the oscillation. I

had been studying mechanisms, particularly those in mixtures

of gases, so it was a natural extension to look at this new laser.

Recall that the well-known red helium–neon laser uses resonant

collisional exchange between the helium singlet metastable

levels and neon atoms to preferentially excite the neon upper

level. However, many other neutral noble gas transitions were

known to oscillate in mixtures with helium without any particular

resonant process involved. In 1963, I was looking at mechanisms

in the strong 3.5-

m neutral line in xenon, and concluded

that helium was helpful but not necessary for oscillation or

886

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

Fig. 1.

W. Earl Bell and his green mercury ion laser (from a color brochure

from Spectra Physics, circa 1965).

high optical gain [2]. Thus it was natural to try to figure out

what role, if any, the helium played in the helium–mercury

ion laser.

I had the HRL glass shop build a discharge tube similar to

Bell’s, and on February 7, 1964, I duplicated Bell’s pulsed os-

cillation on 615 and 568 nm with a small amount of mercury

in a helium discharge. The mercury pressure was controlled by

immersing a glass side-arm in a dewar of liquid nitrogen. I used

a high-voltage high-current pulse generator, left over from ear-

lier microwave vacuum tube research at Hughes. This pulser

could generate square voltage pulses up to eight microseconds in

length, without the decaying current “tails” typical of a simple

capacitor discharge. The first thing I noted was that the green

and red lines began to oscillate several microseconds after the

discharge current ended, that is, in the discharge “afterglow,”

and the green and red lines did not oscillate simultaneously in

time. This suggested that some kind of transfer of excitation

from atom to atom was involved in the excitation process, rather

than electron collision. The “afterglow” nature of the output

suggested that high-power CW operation was not likely, since

the absence of electrons seemed to be required. Actually, my

thinking along these lines was by analogy with the helium–neon

system, in which continuous low-power operation can be ob-

tained, but much higher powers can be obtained in the afterglow

of pulsed discharges (see, e.g., [3, Sec. 2.3.1.4]).

Early the next week, I swung by Bell Telephone Laborato-

ries (BTL) on the way to business visits at NASA Goddard and

Wright–Patterson AFB. I visited Eugene I. Gordon at Bell, and

discussed my Hg laser results with him, noting that the afterglow

output suggested that this would not make a good candidate for

a CW laser. He agreed. I told Gene that I was planning on trying

neon instead of helium, to see if there was a resonance in the

transfer of excitation. He told me that J. Dane Rigden, a former

BTL colleague now at Perkin–Elmer Corporation had also built

a pulsed He–Hg

laser and had substituted neon for the helium,

with the result that the mercury ion lines continued to oscillate.

It seemed clear that there was nothing “resonant” about helium

collisions. He suggested that I try argon as well as neon as a fur-

ther test of this “nonresonance.”

When I got back to HRL at the end of that week, I tried

Rigden’s experiment myself. The resulting neon–mercury dis-

charge also produced the 568-nm line (but not the 615-nm line),

thus demonstrating that resonant collisions with helium were

not involved, at least for the green line. The appearance of this

laser was quite striking, since the green laser output now exited a

bright pink discharge. As the “clincher” experiment mentioned

above, I tried using argon instead of helium as the buffer gas.

While I had bottles of spectroscopic-grade helium, neon, and

xenon already on my laser process station from previous exper-

iments, I did not have argon. It would have taken a week or so

to have the glass blower attach a flask of spectroscopic-grade

argon, so I asked my technical assistant, Robert B. Hodge, to

see if he could scare up a cylinder of welding-grade argon. This

we attached to the process station via a rubber hose. For reasons

I do not understand, this experiment to make an argon–mercury

ion laser failed. Others succeeded [4], [5]. But thinking that I

had already proven the point with neon, I simply pumped out

the discharge tube and refilled it with helium. It was Valentine’s

Day, February 14, 1964.

III. A

HEN

HERE

RGON

It took a couple of flushes with helium to turn the discharge

from the blue of ionized argon to the violet–white of helium.

The red and green outputs of the laser were again displayed on

the laboratory wall through a transmission grating at the end of

the optical bench. But now there was another line on the wall

that appeared “turquoise” and with a brightness equal to that

of the red and green mercury lines! A quick measurement with

the 0.5-meter scanning spectrometer in the laboratory gave the

wavelength as 488 nm (with about 0.2-nm accuracy). The time

behavior was strikingly different. This blue line oscillated ex-

actly with the current pulse, while the red and green mercury

lines remained in the afterglow. Given the dubious purity of the

welding tank and rubber hose filling system, it was not immedi-

ately certain that the line originated from argon. After the min-

utes it took to do the above measurements, I left Hodge in the

laboratory to look after the laser (we were afraid to turn it off),

and headed for the library downstairs at HRL. The MIT Wave-

length Tables [6] listed a line at this wavelength, but without

notation as to the source: Ar I (neutral) or Ar II (singly ion-

ized). I could not lay my hands quickly on a decent reference

for the second spectrum of argon. (I did not discover the ex-

cellent monograph by L. Minnhagen [7] in the Swedish journal

Arkiv Physik until some months later.)

In about half an hour, I returned to my laboratory, where the

laser was still running. At this point, we decided to flush the

discharge tube a few more times with helium, and indeed, the

blue 488-nm line gradually went away, leaving only the red and

green lines of mercury. Then we reproduced it by going through

the same process as before, with argon from the welding tank.

We also noted that we could make the Hg II lines disappear by

freezing out the mercury in a side-arm of the discharge tube

BRIDGES: ION LASERS—THE EARLY YEARS

887

Fig. 2. The author and argon ion laser number two, February, 1964.

with liquid nitrogen. At this point, I went running through the

hallways, dragging my supervisor, Donald C. Forster, and others

down to the lab to see “the blue laser.” By the end of the day,

we had an order placed with the HRL glass shop for an identical

tube, but with no mercury!

The new tube was ready in a couple of days, and we filled it

with helium (mostly) and argon (just a little bit). To our surprise,

we now had

five

wavelengths oscillating, from green (515 nm)

to blue (476 nm), including the previously observed 488-nm

“turquoise” line. Evidently, the “contamination” from mercury

had degraded the discharge conditions for the other four lines in

the previous tube. All of these lines oscillated coincident with

the discharge current pulse. The total peak power output was

estimated on the order of 1 W, with the discharge running 30 A

current in 5-

s pulses at 70 pps. Fig. 2 is another “A Boy and His

Laser” photo, of this tube about a week after it was turned on for

the first time. (Incidentally, the Smithsonian Institution called

years later to see if “the first argon ion laser” was available for

donation. We told them it was contaminated with mercury and

we had thrown it away. They were uninterested in “tube number

two.”)

It was clear that we needed more accurate wavelength mea-

surements than we could obtain with the 0.5-m scanning in-

strument in our laboratory. A 2-m photographic Bausch and

Lomb (B&L) spectrometer was available at the other end of the

building. It was also clear that it would be difficult to move ei-

ther the laser (attached to its process station) or the spectrom-

eter. But this was a laser! We directed the beam out the lab door,

down one hall, around a corner, down another hall, through an-

other lab to a second hall, and finally into the room housing the

B&L instrument. The path used seven mirrors in all, attached to

the metal walls with magnets or C-clamps, over 300 feet of total

path length. Of course, we were trying to keep all this “quiet”

until we could publish our results. We made the measurements

after hours, with telephone contact from our laser lab to James

K. Neeland running the B&L spectrometer.

But of course the HRL management knew of our discovery.

One of the HRL Department managers, Ronald Knechtli,

had left for a technical meeting in Europe, and was visiting

Thompson CSF Laboratories in France. We received a telegram

from him stating that he’d seen a laser with red, green, and blue

lines there in the laboratory of Guy Convert, and that we had

better publish quickly! And this is indeed what we did.

It is now clear that ionized argon was “an idea whose

time had come.” I am aware of three independent discoveries

of oscillation in ionized argon besides that at HRL. Convert

and his coworkers at CSF were doing the same thing I was

doing, adding argon to a mercury ion laser discharge. His

initial publication in Compt. Rendus [5] appeared a week

before my letter in

Applied Physics Letters

[8], [65], but he

reported only the existence of a blue line, without a correct

identification. He correctly identified the 488-nm line in a later

publication [9]. The second independent discovery was made

by Prof. William Bennett and his group at Yale University

[10], who were looking at argon as an ion laser possibility on

theoretical grounds, not as an additive to a mercury discharge.

Their “sudden perturbation” excitation process is described in

[10]. Bennett had submitted an earlier version of this paper to

Applied Physics Letters

, and it had the misfortune to “collide”

with mine in the Editor’s office. The editor rejected the Yale

paper, citing mine as priority. Reference [10] is actually a

second, more extensive, description of their work. The third

independent discovery was made, appropriately enough, by

W. Earl Bell, who was doing the same things that Convert and

I were doing, attempting to find out how the helium–mercury

laser really worked. Stanford Professor Robert L. Byer, then

a new graduate student at Stanford, and a new employee at

Spectra Physics, was given the task of identifying the blue line

that Earl had found. Byer identified the lines, but my paper

had just appeared, so Bell and Byer did not submit a formal

publication [4]. As far as I know, I was the only discoverer

to submit a patent disclosure on noble-gas ion lasers [11].

IV. A

TILL

ORE

OLORS

During the course of our “long-distance spectroscopy,” we

were able to coax ten total visible wavelengths out of singly

ionized argon [8], and also 15 lines in singly ionized krypton

and six lines in singly ionized xenon [12]. In addition, we had

also recorded on the spectroscopic plates several lines that did

not seem to belong to argon, krypton, or xenon, and this was

a puzzle. In particular, there were four lines in the blue–violet

region that were strong, and occurred on plates for more than

one noble gas. I recall sitting in my living room at home and

searching through the MIT wavelength tables, when it struck me

that these four lines were from singly ionized oxygen! We had

888

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

used oxide-coated thermionic cathodes in our discharge tube.

The nickel cathode and filament assembly had been scavenged

from worn-out pulse tubes, Eimac 4PR60’s, the same vacuum

tubes used in our pulser. To rebuild these cathodes, they were

cleaned and then sprayed with a standard “oxide cathode” mix-

ture of powdered barium, strontium, and calcium carbonates in a

nitrocellulose binder. Under vacuum, the cathodes were heated

to drive off the binder, then decompose the carbonates to ox-

ides, and finally, liberate a monolayer of barium. This is a stan-

dard vacuum tube cathode, and HRL was well furnished with

this technology. In operation, these cathodes give off a little

oxygen as the barium evaporates. In a normal vacuum tube, this

is collected by a chemical “getter” built into the tube. Here, the

oxygen collected in the discharge and lased! And there were also

nitrogen and carbon atoms, left over from the nitrocellulose and

the carbonates, and these also lased as ions! We listed all these

lines in a more extensive publication on 118 different laser lines

we observed [13].

The laser wavelengths recorded on the photographic plates

were measured under a microscope for an estimated accuracy of

a few times 0.001 nm. This actually proved critical in identifying

one of the original ten argon laser lines (502 nm), which had a

“logical but incorrect” assignment only 0.05 nm away from an

“illogical but correct” assignment. These measurements were

made by Arthur N. Chester and myself, usually after normal

working hours. Chester was then a Howard Hughes Doctoral

Fellow, assigned part-time to our group at HRL, with classes at

Caltech during the day. He would check out the spectroscopic

references at the Caltech library during the day, make copies,

and then drive to HRL in Malibu for the evening. His major was

theoretical high-energy physics, and I was an electrical engi-

neer. Neither of us had any experience in practical spectroscopy,

so we “learned on the job.” I am certain that if a real spectro-

scopist could have looked over our shoulders, we would have

provided no end of amusement. We had recorded the iron spec-

trum on each plate (from a discharge between two nails near

the slit of the B&L spectrometer), along with the laser lines.

We worked our way across the iron spectrum, identifying each

line from the tables in my high school copy of the Handbook

of Chemistry and Physics [14]. Then we would interpolate the

laser wavelength between the nearest iron lines. Since iron is

such a “rich” spectrum, we could make this measurement very

accurately. But also because it is such a “rich” spectrum, we had

to identify scores of iron lines between each laser line. This took

some time. A year after we did this, I discovered, tucked away

on a shelf in the HRL library, the Vatican photographic atlas

of the iron spectrum, all conveniently broken into segments in-

tended to match your spectrographic plate so you do not have to

identify any iron lines, just match the pictures!

One group of lines remains a puzzle, even to this day. When

we went to higher and higher pulsed currents in xenon, we noted

some strong lines in the blue and green [13]. These were not to

be found in the spectroscopic literature as far as we could deter-

mine. I was visiting Washington, DC, in July 1964 and stopped

by the National Bureau of Standards (NBS) to visit Charlotte

Moore Sitterly, who was widely known for her publications of

atomic and ionic energy levels. [15]. I had used the references

from the 1930s by Curtis J. Humphreys at NBS for the identifi-

Fig. 3. Periodic table of the elements, with those that work as ion lasers shown

shaded (as of 1982).

cation of lines in Xe II and Xe III [16]–[18], and I thought that

he might have left some unpublished lists with Dr. Sitterly. As it

was, she did not have any listings for the unknown xenon lines

we had observed, but said, “Oh, Curtis took all his notes with

him when he left for the Naval Ordnance Laboratory facility in

Corona, CA.” When I returned to California, I telephoned him,

and he confirmed he had observed all the xenon wavelengths

we had seen lase, but that he had not published them, since he

could not identify them. Many other groups have reproduced

these laser lines and puzzled over them (see [3, Section 2.2.3]

for references to work through 1979). But no one has identi-

fied them beyond the initial guess that they are from “Xe IV”

the spectrum of triply ionized xenon. Even a laser product pro-

ducing these lines at 2 W average power was marketed [19]!

Later, more “Xe IV” and possibly “Xe V” and “Kr IV” lines in

the vacuum ultraviolet were added to this list of puzzles by Mar-

ling at the Lawrence Livermore National Laboratories [20].

By mid 1964, there were many researchers working on ion

lasers with all sorts of ions. Since then, almost everything that

is a gas at room temperature or can be vaporized easily, or can

be produced by dissociating a gas molecule, or can be sputtered

into a discharge, has been made to lase as an ion. These elements

tend to cluster on the right-hand side of the Periodic Chart, as

seen in Fig. 3. Art Chester and I collected what had been re-

ported by ourselves and others by mid 1964 [21], again in 1971

[22], and again in 1982 [23]. The excitation mechanisms for var-

ious ion lasers, as they were understood in 1979, were described

in [3, Sections 2.2 and 2.4], which also has about 350 literature

citations for ion lasers through 1979. The most recent listing of

ion laser wavelengths is [24]. Table I gives a time line for the

discovery of the various gaseous ion lasers indicated in Fig. 3.

Given the hundreds of wavelengths originating from scores of

ions, it may seem surprising that only the noble gases, particu-

larly argon and krypton, saw extensive commercial application.

Neon and xenon among the noble gases were worth a try in the

early days, but both discharges proved to be tough to tame, and

tended to tear up the walls of the discharge tube structure. And

they offered little advantage over argon and krypton in power

output or wavelength selection. The only other ion laser that re-

BRIDGES: ION LASERS—THE EARLY YEARS

889

TABLE I

IME

INE FOR

ISCOVERY OF

ASEOUS

ASERS

1963–1980

ceived serious commercial development was the cadmium ion

laser, originally discovered by William T. Silfvast [25] during

his doctoral research at the University of Utah, and developed

later when he joined Bell Laboratories [26]. This is largely due

to the more benign nature of the discharge conditions that op-

timize the laser’s performance. The helium–cadmium ion laser

operates more in the regime of the helium–neon laser discharge

than the argon ion laser discharge. See [3, Sec. 2.4] for a discus-

sion of mechanisms and a guide to the literature.

The chemical reactivity of many materials precludes their use

in a practical commercial product. This is especially true of the

Group VIIA halogens, which have many lasing wavelengths in

the UV and visible, as well as the Group IIIA through VIA el-

ements. Some of these are solid at room temperature, but can

be obtained from gaseous molecules, for example, S from SF

[27] or P from PF

[28]. Other solid materials can be sput-

tered in hollow-cathode discharges, such as Ni, Cu, Ag, and Au

[29]–[32].

There are fewer ion lasers on the left-hand side of the periodic

chart. Only the Group IIA elements lase as ions. The Group IA

elements are conspicuous by their absence. Given the ease with

which sodium can be used in a discharge, it is puzzling why no

ion laser lines have been seen from the various ions of sodium.

The spectrum of singly ionized sodium, Na II, is similar to the

spectrum of neutral neon, Ne I, and that of doubly ionized potas-

sium, K III, should be similar to that of singly ionized argon.

Perhaps it is the lower ionization potential of the neutral Na or

K that keeps the electron temperature in the discharge lower than

in the corresponding noble-gas discharges, and thus hinders the

electron-collision excitation of the ionic levels. In the 1960s,

HRL was engaged in the development of electric space propul-

sion using cesium ions. We had a lot of cesium metal readily

available, so it was a natural experiment to put some in a dis-

charge tube and excite it under the same conditions as a simple

pulsed argon ion laser. We used an existing glass discharge tube.

Unfortunately, the cesium tended to disappear quickly into the

glass walls of the discharge tube! We decided such a laser would

not be practical, even if we were successful, so we abandoned

the effort.

I have always considered the discovery of still more ion lasers

a bit of unfinished business. In 1965, Art Chester and I consid-

ered how easy it would be to make ionized radon lase. The only

thing that stood in the way was the HRL management, who re-

fused to consider the required “hot lab.” (Radon has a half-life

of four days, so you have to work quickly!). The demonstra-

tion of laser oscillation on the higher ionization states of the

right-hand side of the Periodic Chart should only require higher

discharge currents. Likewise, I think lasing on the ions of Group

IA elements is only a matter of technical persistence. Even the

metals of Group IIIB through Group VIIIB, and the refractory

metals of group VIII should produce some laser action, provided

you can sputter the atoms into the discharge. But the biggest

obstacle to discovering new ion laser lines is likely disinterest!

Anything new will just be a laboratory curiosity. The field has

moved on, with argon and krypton serving in the trenches for

35 years, waiting for a more efficient diode-pumped solid-state

laser to displace these monsters from the technical workplace.

I believe there is room for an army of graduate students to dis-

cover new ion lasers, but I also believe they would have trouble

finding employment after graduation!

V. C

ONTINUOUS

PERATION

Ion lasers would have remained a laboratory curiosity if they

could be operated only as pulsed discharges. However, con-

tinuous operation is possible and practical, although that fact

was not evident at first. My early pulsed tubes used very low

pressures of argon, typically 50 mtorr in a 3-mm-diameter dis-

charge tube. Discharge currents were tens of amperes at first,

then 100–200 A or more. The voltage drop along a 1-m tube

was of the order of 1000 V. So initially, we were dealing with

10-kW pulses at the minimum. This posed much different tech-

nical problems than a typical helium–neon discharge.

The first continuous operation was obtained by Eugene I.

Gordon and Edward F. Labuda at Bell Telephone Laboratories,

and published by Gordon, Labuda, and Bridges [33]. The se-

quence of events leading up to the publication are a bit unusual,

and are as follows: I had discussed my Hg II laser results with

Gordon in early February on a visit to BTL, as mentioned above,

and I met him again at an IEEE Committee meeting Easter week

of March 1964. At that meeting I gave him a preprint of both my

argon paper [8] and my krypton/xenon paper [12], and we dis-

cussed my results. I noted that argon lased with the current pulse

whether in pure argon, or argon with neon or helium as a buffer

gas, suggesting direct electron excitation was a mechanism. I

also opined that argon might be a candidate for continuous op-

eration, but that the power required would be much too high

(exhibiting my “helium–neon” state of mind).

890

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

Gordon had more faith than I had! A week or so after the

meeting, I received a telephone call from Gordon, saying that

they had argon oscillating continuously, with about 1 A in a

small-bore discharge tube. The output was less than a milliwatt

but was increasing rapidly with current up to the maximum

the tube could handle. Gordon generously volunteered to add

my name to the paper he and Labuda were writing, but I told

him I would like to contribute more to the work, and that

we had a much bigger power supply at our disposal. Gordon

and Labuda started writing, and Hodge and I started making a

suitable tube at HRL. At first, we tried an air-cooled fused-silica

discharge tube, but the discharge “got away from us,” with

a spectacular display of a traveling fireball that went up and

down the length of the tube, destroying it utterly in seconds.

The next attempt took a few days to build, and resulted in

“tube C” in the publication. This tube was a water-cooled

fused-silica tube, 2.5 mm in inside diameter and about 20 cm

long, and it would survive up to 15 A or so of discharge

current. We obtained about 50-mW output total on the argon

blue–green ion lines. But we could get 80 mW out each end

on the blue–green lines of krypton, and similar outputs on the

lines of xenon. The draft of the paper went back and forth by

mail over a few weeks, to collect the measurements on “Tube

A” and “Tube B” at BTL, and “Tube C” at HRL, and then off to

Applied Physics Letters

. By another unfortunate coincidence,

the paper by Bennett and his group on “Quasi-CW” operation

of argon [10], the extension of their previously rejected paper

that had collided with my original argon paper, now collided

with our CW paper in the APL Editor’s office. This time, the

Editor (wisely, in my opinion) decided not to reject the Bennett

group’s paper again, and the two were published back-to-back

in the May issue of APL. From time to time, people have

asked me how it came to be that groups at Bell Labs and

Hughes “cooperated” in the demonstration of the first CW

argon ion laser. The above is the story.

VI. H

ASERS

ORK

It is impossible to give all the details in one paper. Instead, I

will attempt to summarize briefly what we know about argon ion

laser mechanisms, most of which we learned in the last half of

the 1960s. For a guide to the literature about other ion lasers, see

[3, Sections 2.2 and 2.4], for example, or the extensive reviews

by Dunnand Ross [34] and Davis and King [35].

Research in the early years was directed at determining the

pathways by which the upper laser levels were excited and the

lower laser levels destroyed. Closely coupled to this was research

to determine the gas discharge parameters, such as electron

temperature, ion temperature, electron density, and the atomic

parameters of collision cross section, transition probabilities,

etc. The goal was, of course, to make a mathematical model

that would predict the observed behavior with diameter, length,

current, pressure, etc. It is fair to say we did not completely

succeed, although we know enough to make successful ion

lasers, which is the engineering goal in any case.

The first pathway proposal was published by Bennett

et al.

[10], and is illustrated in Fig. 4(b). They proposed that the

upper laser level was excited from the argon ground state (37

(a)

(b)

(c)

(d)

Fig. 4.

Four different proposed excitation pathways for argon ion laser upper

levels.

eV) by collision with single high-energy electron and that the

preferred upper level would be the 4p

level, based on

a “sudden perturbation” calculation. This would predict that

the 476-nm line of Ar II would oscillate. In fact, this is the

strongest line when a pulsed discharge is operated at very

low gas pressure, the condition which produces a high value

of E/p (electric field to gas pressure ratio) that yields high

electron energies in a discharge. However, it does not explain

why the 488 and 515-nm lines oscillate, much less why they

are the strongest lines in the usual operation of the CW argon

ion laser, at much smaller E/p.

The “two-step” scheme illustrated in Fig. 4(a) was first pro-

posed by Labuda, Gordon and Miller [36], based on the obser-

vation that well above threshold the laser output power varies

as the square of the discharge current, suggesting that two suc-

cessive electron collisions were involved. The electron energies

implied by Fig. 4(a) are still pretty high compared to the elec-

tron temperatures later measured for typical argon ion laser dis-

charges, however (a few electronvolts). One can argue that step

1 may, in fact, be a very complicated series of substeps involving

much lower electron energies, ultimately resulting in having the

ion ground state population proportional to the discharge current

(which is really required by the overall condition of charge neu-

trality for the discharge in any case). But step 2 would require

electron energies greater than 20 eV. In fact, typical collision

cross sections for electric-dipole-allowed transitions may peak

as much as 100 eV above the threshold energy. Even worse, the

ion ground state is a 3p

state and has the same parity as the

4p upper laser levels, thus making step 2 a “forbidden” process.

Of course, no process is really forbidden, but the cross section

BRIDGES: ION LASERS—THE EARLY YEARS

891

should be much lower. However, for forbidden transitions, the

cross section typically peaks just above threshold, not 100 eV

above threshold, thus favoring 20 eV electrons, not 120 eV elec-

trons (see [37] for actual measurements of these cross sections).

The scheme illustrated in Fig. 4(c) is a refinement by Labuda

et al.

[38]. The “first electron” is again the sum of the processes

that create an ion ground state proportional to the discharge cur-

rent. Labuda

et al.

measured large ion metastable level popu-

lations that are closely coupled to the ion ground state due to

the rapid electron collisional creation and destruction processes.

Thus these metastable level populations also are proportional to

the discharge current. Some of these metastable levels are con-

nected to the 4p upper laser levels by allowed transitions re-

quiring only 3-eV electrons for excitation. This is the “second

electron” that results in the square law variation with discharge

current.

The above model seemed to explain the observations, until

Robert I. Rudko, a graduate student of Prof. C. L. Tang at Cor-

nell, measured the spontaneous radiation intensities of all the

transitions cascading into the 4p upper laser levels and all those

coming out of the 4p levels, in a nonlasing discharge. Rudko

and Tang concluded that about 50% of the population of the 4p

levels resulted from this cascade from higher lying states, as in-

dicated in Fig. 4(d) [39]. A. Stevens Halsted and I repeated this

measurement for one of our laboratory tubes, and concluded the

percentage was about 30% [40], not as great but still a very sig-

nificant fraction. The conclusion was that all the processes indi-

cated in Fig. 4 contribute, with relative contributions changing

according to the particular discharge geometry and conditions.

Many measurements and calculations of the plasma condi-

tions in argon ion laser discharges were made early on, espe-

cially by Gordon’s group at Bell Labs, and Sobolev’s group at

the Lebedev Institute [41], as well as the group at HRL. Ref-

erence [3] gives a guide to the extensive literature from these

groups and others. A short summery of plasma characteristics

is given in [42], by Chester, Halsted, Parker, and myself. The

actual numbers vary with discharge geometry, operating pres-

sure and current, but ballpark figures for a smaller laser are:

Electron temperature 2–3 eV (23 000–35 000K);

Ion temperature 0.2 eV (2300K);

Gas temperature 0.15 eV (1700K);

Electron density: a few

(plasma frequency of

50 GHz);

Neutral density: a few

Those familiar with discharge physics will recognize that ex-

periments on plasmas with these parameters are definitely not

easy. There was clearly a relationship between the ability to do

such experiments and the engineering of practical ion laser dis-

charge techniques and materials. In this case, scientific research

and engineering development went hand in hand.

To make matters worse, such discharge parameters result in

strong “pumping” of the gas in the discharge by ion drift (to-

ward the cathode) and electron collision momentum transfer

(toward the anode). Gas pressure gradients can develop in ei-

ther direction, depending on the discharge current, and these

can also be unstable. Gordon and Labuda were the first to rec-

ognize this problem, and also provided the answer in a “gas

return path” from one end of the tube to another, outside the

main gas discharge column to equalize the gas pressure [43]. At

HRL, Halsted made extensive measurements of gas return path

parameters [40], [42] and Chester, then in Gordon’s group at

BTL did likewise [45]. However, the problem is not as simple

as equalizing the pressure. Experiments made during the course

of developing a high-power UV ion laser [44] indicate that fully

equalizing the pressure between the active region and the elec-

trode regions is not a good thing; instead, the gas flow geometry

must equalize the pressure in the two electrode regions, but pro-

mote a differential between the active (small diameter) bore re-

gion, and the electrode regions. Modern commercial disk bore

ion lasers have carefully controlled gas flow geometries, and

these are usually proprietary with each manufacturer.

VII. A

RGON

ASER

NGINEERING

A. The Pulsed Era

Hughes was anxious to get some publicity for the discovery of

the argon ion laser, and the first opportunity we had to exhibit

it was at the Hughes Corporate booth at the July 1964 Amer-

ican Institute of Aeronautics and Astronautics (AIAA) show in

Washington, D.C. This was hardly the ideal place to exhibit a

new laser, but it was the first up. Robert Hodge and I built a small

pulsed ion laser for the show. The laser output was about 1 W

peak, and about 1 mW average over the blue–green argon lines.

Few visitors to the booth had a clue what they were seeing. The

typical question was “Where’s the ruby?” The next opportunity

we had for an exhibit was the IEEE WESCON show in Los An-

geles in August 1964. We set up the same laser for a much more

technically appreciative audience, and got a lot of attention. Un-

fortunately, we also received some unwanted attention from the

National Radio Company booth next to the Hughes booth. Na-

tional was exhibiting their latest communications receiver, and

our unshielded pulser wiring, which had radiated unnoticed at

the AIAA show (largely devoid of electronics exhibits) was now

blasting the receiver next door! We were off the air for a day,

shielding everything, but returned the next day with the RFI

gone.

This small pulsed argon ion laser was the basis for a “product”

from HRL. We had so many requests for a blue laser from inside

HRL and elsewhere at Hughes, that Hodge and I decided to en-

gineer a version suitable for manufacture. We handbuilt several

of these units in 1964 as the HRL “Model 40” before the Hughes

Electron Dynamics Division (EDD) took over the effort as the

Hughes “Model 3040H” in a much classier-looking package.

These small pulsed tubes were originally very short lived, since

the discharge sputtered the fused-silica walls rather badly, espe-

cially in the “throat” region, and the sputtered material “buried”

the argon gas. Thus the 50 mtorr of argon gas fill would be gone

in 100 hours or so of operation. With Hughes EDD, we devel-

oped a cheap gas refill system, consisting of measured “doses”

of argon sealed in glass ampoules, contained inside a soft copper

tubular appendage. When the laser “went dry,” you could break

one of these ampoules by squeezing the copper tube with a pair

of pliers. This may seem pretty crude, but it is no different in

principle from the electronically controlled valve “burp” system

still used in commercial argon ion lasers!

892

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

Given the short life of our tubes, you can imagine our

enthusiasm for a customer that approached us in 1966 with a

requirement for an argon ion laser that “only needed to operate

for a single pulse.” The folks at Los Alamos wanted us to build

a pulsed laser for instrumentation in an underground nuclear

test. They needed 10 W or more output for a single 50

s pulse.

Hodge and I thought this would be a piece of cake, so we took

the job and built a longer, larger diameter pulsed tube. Los

Alamos ordered four tubes, so they could do some life testing

in the lab and have spares. They developed their own “high

reliability” pulsed power supplies. Some months later, Hughes

EDD decided to “productize” this more powerful laser as the

model 3041H. Still later, we received a panic phone call from

Los Alamos. They had “life tested” all four lasers, and now

they were all used up, gas ampoules and all. We referred them

to A. Gene Peifer at EDD to work out a “loan” of the 3041H

prototype. After all, what could go wrong? The laser was to be

located safely in the surface instrument shack. Unfortunately,

this was one of those tests that did not quite go as planned, and

the “test” swallowed the 3041H prototype. Los Alamos had to

buy that fifth laser after all.

B. CW Argon Ion Laser Engineering and Applications

Of course, the pulsed ion lasers were only stopgap devices,

to serve niche applications and to act as standins for the real

challenge: a practical CW ion laser. In that, we were truly in at

the beginning. Even before our CW paper [33] appeared, our

colleagues at the Hughes Aerospace Group in Culver City, CA,

knew about our results and came to Malibu for a visit. They had

an application in mind that required “high” CW power in the vis-

ible. A year before, Perkin–Elmer Corporation had developed

a line-scanning night reconnaissance system using a “heroic”

100-mW helium–neon laser, two 2-m discharge tubes in a folded

path. In such a system, the laser beam is directed downward,

and a high-speed rotating mirror scans the beam from side to

side across the flight path. The aircraft motion supplies the scan

along the aircraft track. The received reflected light intensity

is recorded on film as “video” with a similar raster scan, and

the result is a long strip map of the ground overflown by the

aircraft, just like a daytime strip camera photo. A history of

the Perkin–Elmer line scanning systems is given in [46]. The

Hughes group was very interested, but concluded that 100 mW

of red light was not really enough and that there was no further

development possible for the helium–neon laser. Thus, when we

had demonstrated more than 100 mW of blue–green output, they

decided they should now develop a similar system.

Our first joint argon laser objective for the fall of 1964 was

200-mW output in a package rugged enough to fly. By July,

we had over 200 mW out of a simple water-cooled fused silica.

Of course, now that the systems group was serious, they were

building additional margin into their design, and asked for 400

mW. Since the output power seemed to be increasing as the

square of the discharge current, it seemed reasonable to us that

we could meet this objective, too, and we did. But when they

upped the ante to two watts, we decided it was time to have a talk

with their optical designer! It turned out that he was not familiar

with Gaussian beams, so he had chosen to expand the beam and

chop off the tails, so he could work with a more familiar “top

hat” transverse distribution! I recall we had to “educate” him

about the inherent merits of Gaussian beams. But we also gave

the systems people their two watts.

It became clear that fused-silica walls in the discharge tube

were not going to allow an indefinite increase in the discharge

current, and the sputtering by ions accelerated across the dis-

charge wall sheath (10 to 20 eV) resulted in severe erosion and

gas cleanup, even at the CW operating pressure of about 0.5

torr. Since our department at HRL, headed by Donald C. Forster,

also did research and development in millimeter-wave vacuum

tubes [47], it was only natural to feel that a proper ion laser

should be an all-metal–ceramic device. We undertook the de-

sign of such a device in the summer of 1964, in parallel with the

push toward higher powers using fused-silica tubes made in the

HRL glass shop. Most metal–ceramic vacuum tubes use poly-

crystalline aluminum oxide (alumina) as the ceramic, and this is

where we began. It was not long before we were breaking alu-

mina discharge tube bores right and left. Unfortunately, when

the discharge strikes, the bore has to conduct 200–400 W of heat

per centimeter of length, radially outward toward the coolant

jacket. We originally thought thermal shock was the culprit, but

later we concluded that it was stress caused by the thermal gra-

dient in the poorly conducting alumina.

We spent months trying to improve the design, and learn

how to “turn the discharge on slowly.” Still, the end result was

invariably broken tubes. The date for the first flight test of a

prototype Hughes system was fast approaching, so we stuffed

a laboratory-type fused-silica laser (complete with an argon

bottle and a refilling valve) into the box the system’s people had

built for us, and hoped for the best. The system was flown in

1965 in the bomb bay of a B-47 from Wright–Patterson AFB,

and performed beautifully. I no longer have the imagery from

those tests, but I do recall that we caught the image of a military

service mechanic asleep spread-eagle on the tarmac beside his

plane one night overflying the flight line. To make

our

nights

even less restful, in the middle of this test program, the Air

Force brass decided that all the WPAFB test pilots needed

landing practice, and the B-47 with our system aboard was

chosen as the training aircraft. Our “fragile” fused-silica laser

did touch-and-go’s at WPAFB for two weeks...and survived!

While we were engaged in developing a “flight-qualified”

CW ion laser, others were doing the same for commercial sales.

As far as I can tell, Raytheon was the first to advertise a CW

argon ion laser for sale in late 1964. I have a brochure and letter

in my file, dated February 1965, offering the Raytheon LG-12

for $27 500. Their design used a glass-blown fused-silica dis-

charge tube and produced 1 W. The operating life was guaran-

teed for 100 hours, the laser head weighed 80 lbs, and the power

supply 600 lbs!

During 1964 and 1965, I had several telephone conversations

with Earl Bell at Spectra Physics, who was engaged in their CW

argon ion laser development project. Mostly, we commiserated

on approaches that did not work. I told him of the thermal shock

problems with standard alumina, and he told me of his problems

getting a straight bore of beryllium oxide (beryllia). This neg-

ative view of beryllia was unfortunate for both of us, since we

BRIDGES: ION LASERS—THE EARLY YEARS

893

Fig. 5. Schematic view of a segmented metal-bore ion laser discharge tube developed at Bell Telephone Laboratories in 1965 (from [36]).

both later made successful beryllia-bore discharge tubes, but for

1964–1965, we avoided the material.

A major breakthrough in CW discharge engineering was

made in 1965 by Labuda, Gordon, and Miller at Bell Labs

[36]. While it seems counterintuitive, you can make the walls

of the discharge tube out of metal, provided you break the

length of the discharge tube into short electrically floating

segments. Labuda made a design with short molybdenum

tubes, supported in the center of disk-like cooling fins, as

shown in Fig. 5. The assembly of disks was then fitted into

a large-diameter fused-silica vacuum envelope, with the fins

radiating heat through the envelope.

Fresh from his Ph.D. at Stanford, A. Stevens Halstead

joined HRL and the ion laser research and development

activity. Shortly thereafter, we adapted Labuda’s design to

use molybdenum “slugs” contained in a one-inch i.d. alumina

vacuum envelope. Thus we would have our all-metal–ceramic

tube, without exposing the alumina to the thermal shock from

striking the discharge. But we were not home free, since

deposits of sputtered molybdenum “shorted out” the slugs at

the alumina wall, thus limiting operating life. The search for

a lower sputtering rate material lead naturally to tungsten, but

fabrication of machined shapes in tungsten is not easy. HRL

had experience in machining precise tungsten shapes from

sintered porous tungsten, with the pores filled with copper as

a machining lubricant. The copper is boiled out in a vacuum

furnace after machining. We tried “slugs” made of this material,

which proved to be low sputtering indeed. Unfortunately, some

copper always remained in the pores, and gradually evaporated

out and deposited on the envelope walls, again shorting out the

discharge.

About this time, James R. Fendley, Jr., and Karl G. Hern-

qvist at RCA Laboratories reported an argon ion laser made with

graphite “slugs” that had even lower sputtering yield than tung-

sten [48]. This we also tried, and found that the graphite was

a low sputtering material as advertised, but it had a tendency

to flake off powder with time. While this might not have been

a problem in a commercial laboratory tube sitting on a work-

bench (RCA and later Coherent sold quite a few), it was not

going to be suitable for an airborne-qualified laser! We tried a

number of variations including pyrolytic graphite and pyrolytic

graphite overcoated with CVD tungsten. None of these bore ma-

terials and configurations worked very well. All of the above en-

gineering developments are documented in a series of Air Force

reports, and summarized in two final reports [40], [49].

The next breakthrough in ion laser engineering was the

discovery by Huchital and Rigden that the discharge walls,

whether continuous insulating material or metal segments,

could be replaced with a series of thin metal disks with holes

[50]. Of course, this only worked if a strong magnetic field

was used to confine the discharge, but such fields were already

used to improve the power output. With this geometry, low

894

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

Fig. 6.

Photograph of a metal/ceramic argon ion laser discharge tube using radiation cooled tungsten disks, prior to assembly. This tube was develope

d at the

Hughes Research Laboratories for the AN/AVD-3 night reconnaissance system, 1967–1970.

sputtering tungsten could now be used, since very little ma-

chining is required to make a disk with a hole in it. This was

the structure we adopted for the airborne-qualified laser for the

follow-on Hughes development night reconnaissance system,

the AN/AVD-3. A photo of the discharge tube parts prior to

assembly is shown in Fig. 6, with the series of thin tungsten

disks (about one inch in diameter) stacked for insertion into the

alumina vacuum envelope. The metal can at the left contained

the thermionic cathode. This system was successfully flown in

1968–1970 with excellent results, but it was never funded for

production. (The Air Force did not like the idea of modifying

RF-4’s to route several gallons per minute of laser coolant

through bulkheads from the camera compartment in the nose to

the heat exchangers in the wing fuel tanks.) With the demise of

this application as a driver for development, we ended the ion

laser work at HRL. Halsted left HRL to join the Hughes Elec-

tron Dynamics Division and continued with the development of

a third-generation airborne argon ion laser using a beryllia-bore

discharge tube for the Air Force Materials Laboratory (AFML)

at WPAFB. But since there was no longer a system application,

that effort ended with the delivery of 13 lasers to AFML [51].

That design eventually became the first Hughes commercial

CW argon ion laser product, the Hughes EDD 4066H, in 1975.

However, EDD withdrew from the argon market completely

about 1976.

VIII. H

EROIC

UMBERS

No report on the early days would be complete without men-

tioning the winners in the power race. Many groups, including

HRL, made it to the 10–20-W level by 1966, but the overall

U.S.A. winner was Jim Fendley and J. J. O’Grady at RCA, who

delivered a 100–W CW output laser to the U. S. Army [52] in

1970. Their tube design used a 2.5-m stack of radiation-cooled

graphite disks, with 6.4-mm diameter holes and a discharge cur-

rent of 90 A. The overall international winner is V. I. Donin and

his group in the Soviet Union, who reported a 500-W CW output

argon ion laser in 1973 [53]. This tube was 16 mm in diameter

and 1.5 m long, and ran at 390-A discharge current with an effi-

ciency of 0.2%. The walls used alumina coatings on aluminum

disks that were stacked and directly water cooled. As far as I

know, this is still the record for a gaseous ion laser.

IX. S

OME

DEAS

HAT

ORK

In 1965, Earl Bell reported a novel excitation method for

pulsed ion lasers. He used the discharge bore and a gas return

path as a rectangular-shaped one-turn secondary winding of a

ferrite core transformer [54]. This had the advantage of being

an “electrodeless” discharge, eliminating any problems with the

anode and cathode of a conventional dc discharge (which also

meant that chemically active elements like the halogens could be

used in the discharge). When he ran into problems with beryllia

in his CW laser development, he extended this “one-turn” sec-

ondary to CW lasers, using high-power r–f at 41 MHz, air-core

coupled with a few turns of heavy copper tubing. This was the

basis of Spectra Physics’ first commercial argon ion laser [55].

However, this product only lasted a year or so due to insuffi-

cient operating life. While the Brewster’s angle windows were

located in side arms away from the main “turn” of the discharge,

there was sufficient r–f floating around to produce mild dis-

charges in these side arms. Initially this was deemed to be a good

BRIDGES: ION LASERS—THE EARLY YEARS

895

thing, to have the discharge “wash” the windows of any con-

tamination. Unfortunately, it also ion-etched the windows and

turned them into negative lenses in a couple of hundred hours,

so that the optical cavity would become unstable [56]!

Argon ion lasers have been excited by several different

means, ranging from microwave-driven discharges [57], mi-

crowave cyclotron resonance [58], and

-pinch discharges

[59]–[61], among others. Even a theta-pinch fusion machine

was used to make an argon ion laser. Unfortunately, the power

into the pinch had to be throttled to ridiculously low powers

to make the conventional Ar II and Ar III lines lase. At its

“normal” fusion-research level, the emission (nonlasing) was

all in the short wavelength ultraviolet on unidentified lines from

very high ionization states of argon [62]. For the last 35 years,

the simple dc discharge has been the workhorse of commercial

argon ion lasers.

The engineering development of dc discharges for ion lasers

is fraught with ideas (amusing in retrospect) that did not work

out, and we recount a few here.

In our frustration with thermal shock in alumina as a dis-

charge bore material in 1964, we made a laser discharge tube

with a boron nitride (BN) bore, press fit into a water-cooled

copper vacuum jacket. Solid boron nitride, made from

hot-pressed BN powder, seemed an ideal material, since it

was very resistant to thermal shock and easily machinable. It

lived up to our expectations in these regards. Unfortunately the

discharge would erode the powdered material in a few tens of

hours, and the powder would be electrostatically suspended in

a small “cloud” on the optical axis, thus shutting off the optical

path through the tube. We only built one.

During early environmental testing of the laser for the

AN/AVD-3, we discovered that the laser output would disap-

pear at high altitudes in the test chamber. It could be regained

by realigning the mirrors. The problem was that we had

oriented the Brewster’s angle windows on the two ends of the

tube to “face up,” since this made them easy to clean in the

laboratory. This turns the tube into a weak negative prism, with

the environmental air as the “denser” material. Removing the

air simply misaligned the cavity. The solution was to make

one window “face up” and the other “face down” so they

were exactly parallel. We note that all commercial ion lasers

have both windows “face up” for easy cleaning, requiring

realignment with changing altitude.

In the early all-fused-silica discharge tubes, the coolant in

the water jacket is exposed to intense ultraviolet from the dis-

charge. This was brought home to us painfully, when the plant

engineering staff at HRL shut us down. It seems that this UV

light polymerized the anti-algae agent they added to the cooling

water to keep the cooling tower clean, and we had been creating

“sludge” in their system for months. Needless to say, they were

very unhappy with us when they finally discovered what was

happening.

X. O

THER

ARLY

ASER

PPLICATIONS

In the period 1964 to 1970, many applications using gaseous

ion lasers were pursued by research and development groups

worldwide. These included projection color TV, high-speed data

recording on film, eye surgery (see the excellent review by E.

I. Gordon elsewhere in this special issue), dye laser pumping,

even disco light shows! Hughes being an aerospace firm, our

customers were usually government agencies.

In these early years, HRL built several “one-off” CW argon

ion lasers for various applications. We built a 1-W laser in 1965

for a prototype laser space communications system developed

by the Hughes Aerospace Group for NASA Houston. This

system used digital polarization modulation at 30 Mb/s to

transmit one video channel. When demonstrated over the

“four-mile range” at the Hughes Culver City facility, we had

to add 30-dB optical attenuation to keep the detectors from

saturating. We estimated this system would have provided

a 30-Mb/s channel to the moon. We built a CW argon ion

laser system for NASA Huntsville in 1966 for space illumi-

nation experiments from a small telescope dome there. The

biggest problem with this particular system was building a

coolant-to-air heat exchanger that would work 100 feet away

from the laser and the dome, and would not freeze in the

Huntsville winters.

About 1968, the Hughes systems people embarked on a dif-

ferent airborne application, a forward-looking imaging system

that used a 2-W argon ion laser scanned in a TV-like raster,

with a synchronously scanned receiving telescope. HRL was

again enlisted to provide a qualified laser for this system. Our

laser design was essentially the same as that for the AN/AVD-3

line scanner, but the package had to be different because the

laser was now to be mounted on the outer skin of the aircraft! It

seems that there was not enough room in the nose for the scan-

ning optics and the laser too, so we were relegated to the slip-

stream. Fig. 7 is a picture of a Douglas A-26 of the “Hughes Air

Force,” with the system installed, and our laser mounted on the

outside (arrow). The beak-like assembly is an optical window

that allowed the system to be pointed at angles between straight

ahead and straight down. One night this system was flown from

the Hughes airstrip in Culver City, CA, to San Diego, to ren-

dezvous with and image a submarine surfaced off shore. This

was done, and then the pilot decided to fly back to Culver City

with the system pointed straight down, using the TV display of

the highway to chart his way back. A later version of this system

was to add range-finding capability on any pixel of the raster by

pulsing the laser to ultraviolet output for that pixel [63]. This

latter system was not built, although the concept was demon-

strated at HRL.

From time to time in 1965–1970, we would get requests

for proposals for underwater applications. The argon ion

blue–green lines are in the optical “window” for water centered

around 470–500 nm (depending on where the water is located

geographically). We would respond with designs ranging from

simple modifications of our airborne systems to proposed de-

velopment of 100-W and 1000-W output ion lasers. (Operation

from an underwater platform seemed attractive since infinite

cooling capacity was readily available, or so we imagined.)

None of these ideas got beyond the proposal stage. Eventually

cooler heads would prevail and realize that light attenuates very

rapidly when it decays exponentially with distance, with

lengths of 5 to 20 meters, rather than as an inverse square law.

896

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 6, NO. 6, NOVEMBER/DECEMBER 2000

Fig. 7.

Forward-looking laser imaging system which used a TV-like raster

scan. The argon ion laser is mounted on the outside of the aircraft skin (arrow).

This system was built by the Hughes Aircraft Company, Aerospace Division,

CA, 1968.

One last laser “application” is worth special mention. In 1967,

Prof. C. O. Alley of the University of Maryland considered

directing a ruby laser at the Hughes/JPL/NASA Surveyor VII

spacecraft, then sitting on the surface of the moon. That craft

had successfully completed all its mission requirements and had

(unexpectedly) survived the lunar night. It had a video camera

aboard to view “digging in the lunar soil,” but the camera could

also be oriented to view the earth. And it was just sitting there,

with nothing to do. Michael Shumate of Caltech’s JPL sug-

gested to Alley that he needed a CW ion laser rather than a

pulsed ruby laser, since the chances of the Surveyor camera’s

video scan coinciding with the arrival of a pulse were next to

zero. Alley called all around the country looking for a laser

to borrow, but most groups (including us) had only laboratory

lash-ups at that early date. He eventually found one to borrow at

Spectra Physics, one of their r–f excited units. However, all his

telephone calls stimulated an interest in participating, even if it

meant dealing with our laboratory “kluges.” With Shumate as

our team leader, we moved our laboratory 2-W tube up to Table

Mountain near Wrightwood, CA, where JPL had a small obser-

vatory with a 24-inch telescope. This was all done “after hours”

Fig. 8. Video frame of the earth, taken from the Surveyor VII spacecraft on

the moon, January 20, 1968. Several lasers are pointed at the camera from the

earth, but only the ones from Table Mountain (left arrow) and Kitt Peak (right

arrow) can be detected.

at HRL, and at no cost, since this was not an official program of

any sort for us.

We were not the only ones so moved to participate on their

own! On the night of the test in January 1968, several groups

were pointing telescopes and argon ion lasers at the moon (and,

hopefully, at the location of the Surveyor spacecraft): Alley was

at Kitt Peak National Observatory, using the McMath 60-inch

Solar telescope and a borrowed 2-W Spectra Physics laser; a

group at Raytheon Research Laboratories at Waltham, MA, had

a laboratory laser with 60-W output, with the beam directed

out a window to a 4-inch portable telescope in the parking lot

(with ten undergraduate students from Wesleyan University to

help); a group at NASA Goddard, Greenbelt, MD, using a 10-W

RCA laser and an existing 5.5-inch satellite tracking telescope; a

group at Perkin-Elmer, Norwalk, CT, using a 2-W laser built by

Perkin–Elmer and Mr. R. Perkin’s personal 24-inch telescope; a

group at MIT Lincoln Laboratories, using a loaned 3.5-W Spac-

erays laser, and an 3-inch telescope with az–el gimbels; and our

gang at Table Mountain, CA, with a 2-W HRL-built laser and

the 24-inch telescope. Fig. 8 shows the result. The two white

spots indicated by the arrows are Table Mountain on the left

and Kitt Peak on the right. The groups on the east coast of the

U. S. are not seen (and further image processing by JPL failed to

bring them out). By the time the experiment was executed, the

pre-dawn sunlight provided an unfortunate background for the

East Coast groups. Another contributing factor was the skill re-

quired in pointing the telescope to the proper spot on the moon.

BRIDGES: ION LASERS—THE EARLY YEARS

897

Both our group and Alley’s had skilled astronomers directing

the telescope through a dichroic mirror at lunar features known

to be at the Surveyor location. After the experiment, the as-

tronomer directing our telescope let me look at these “features,”

which appeared to me as only a vague blur. I do not know how

he saw “craters.” A brief summary of this experiment is given

in [64].

XI. P

HILOSOPHICAL

BSERVATION

My intent in writing this paper was not to give a polished

summary of the theory and practice of ion lasers, but rather to

provide a historical account of “how it happened,” at least from

one person’s viewpoint. Some record should exist of the fits and

starts and dead-end ideas as well as the successes that lead to

the present state of an art. It should give some comfort to the

newcomer who may be intimidated by the very pat-sounding

development of science and technology typically given in the

classroom. Good people make mistakes, and learn from them.

And there is a tremendous thrill when everything finally works!

CKNOWLEDGMENT

The author would like to thank the people at HRL:

A. N. Chester, P. O. Clark, V. Evtuhov, H. R. Friedrich,

A. S. Halsted, R. B. Hodge, G. N. Mercer, J. K. Neeland,

J. V. Parker, M. R. Smith, R. Smith, D. Sobieralski, the De-

partment Manager D. C. Forster, and three different Laboratory

Directors during 1964–1972: M. R. Currie, L. M. Field,

and G. F. Smith. At Hughes Electron Dynamics Division:

D. E. Crane, D. D. Hallock, W. P. Kolb, A. G. Peifer. At the

Hughes Aerospace Group: J. C. Hill. At Bell Laboratories,

E. I. Gordon, E. F. Labuda, R. C. Miller, W. T. Silfvast,

C. E. Webb. At Spectra Physics: W. E. Bell, A. L. Bloom,

J. P. Goldsborough, R. C. Rempel. At RCA: J. R. Fendley, Jr.,

K. G. Hernqvist. At the Lebedev Institute: N. N. Sobolev; At

WPAFB: R. Firsdon, W. C. Schoonover. At JPL: M. S. Shu-

mate.

EFERENCES

[1] W. E. Bell, “Visible laser transitions in Hg

,”

Appl. Phys. Lett.

, vol. 4,

pp. 34–35, Jan. 15, 1964.

[2] W. B. Bridges, “High optical gain at 3.5-



m in pure xenon,”

Appl. Phys.

Lett.

, vol. 3, pp. 45–47, Aug. 1, 1963.

[3]

, “Atomic and ionic gas lasers,” in

Methods of Experimental

Physics, Volume 15: Quantum Electronics, Part A

. ser. L. Marton, C. L.

Tang, Ed. New York: Academic, 1979, pp. 31–166.

[4] W. E. Bell. Communicated to me in July 2000 by R. L. Byer, Stanford

University, who as a student had worked with Bell at Spectra Physics, ,

1964.

[5] G. Convert, M. Armand, and P. Martinot-Lagarde, “Effet laser dans des

mélanges mercure-gaz rares,”

Compt. Rend. Acad. Sci. Paris

, vol. 258,

pp. 3259–3260, Mar. 23, 1964.

[6] G. R. Harrison,

M. I. T. Wavelength Tables

. New York: Wiley, 1952.

[7] L. Minnhagen, “The spectrum of singly ionized argon,”

Ark. Fys.

, vol.

25, pp. 203–284, 1963.

[8] W. B. Bridges, “Laser oscillation in singly ionized argon in the visible

spectrum,”

Appl. Phys. Lett.

, pp. 128–130, Apr. 1, 1964.

[9] G. Convert, M. Armand, and P. Martinot-Lagarde, “Transitions laser vis-

ibles dans l’Argon ionisé,”

Compt. Rend. Acad. Sci. Paris

, vol. 258, pp.

4467–4469, May 4, 1964.

[10] W. R. Bennett Jr., J. W. Knutson Jr., G. N. Mercer, and J. L. Detch,

“Super-radiance, excitation mechanisms and quasi-CW oscillation in the

visible Ar

laser,”

Appl. Phys. Lett.

, vol. 4, pp. 180–182, May 15, 1964.

[11] W. B. Bridges, “Ionized noble gas laser,” U. S. Patent 3 395 364. As-

signed to Hughes Aircraft Company, July 30, 1968.

[12]

, “Laser action in singly ionized krypton and xenon,”

Proc. IEEE

vol. 52, pp. 843–844, July 1964.

[13] W. B. Bridges and A. N. Chester, “Visible and UV laser oscillation at

118 wavelengths in ionized neon, argon, krypton, xenon, oxygen, and

other gases,”

Appl. Opt.

, vol. 4, pp. 573–580, May 1965.

[14]

Handbook of Chemistry and Physics

, 32nd ed., C. D. Hodgman, Ed.,

Chemical Rubber Co., Cleveland, OH, 1950, pp. 2304–2319.

[15] C. E. Moore,

Atomic Energy Levels

, ser. National Bureau of Standards

Circular 467: U. S. Government Printing Office, 1949, 1952, 1958, vol.

I–III.

[16] C. J. Humphreys, “Second spectrum of xenon,”

J. Res. Natl. Bur. Stand.

(U.S.)

, vol. 22, pp. 19–53, Jan. 1939.

[17]

, “Third spectrum of xenon,”

J. Res. Natl. Bur. Stand. (U.S.)

, vol.

16, pp. 639–648, June 1936.

[18] C. J. Humphreys, W. F. Meggers, and T. L. deBruin, “Zeeman effect in

the second and third spectra of xenon,”

J. Res. Natl. Bur. Stand. (U.S.)

vol. 23, pp. 683–699, Dec. 1939.

[19] Anonymous, “LGI.37 high power pulsed optical quantum generator,”

The 1970 World’s Fair

, Osaka, Japan, specification sheet handed out

at the Soviet Union laser exhibit.

[20] J. B. Marling, “Ultraviolet ion laser performance and spectroscopy—I.

New strong noble-gas transitions below 2500 A,”

IEEE J. Quantum

Electron.

, vol. QE-11, pp. 822–834, Oct. 1975.

[21] W. B. Bridges and A. N. Chester, “Spectroscopy of ion lasers,”

IEEE J.

Quantum Electron.

, vol. QE-1, pp. 66–84, May 1965.

[22]

, “Ion lasers,” in

Handbook of Lasers With Selected Data on Optical

Technology

, R. J. Pressley, Ed.

Cleveland, OH: CRC Press, 1971, pp.

242–297.

[23] W. B. Bridges, “Ionized gas lasers,” in

Handbook of Laser Science and

Technology, Vol. II—Gas Lasers

, M. J. Weber, Ed.

Boca Raton, FL:

CRC Press, 1982, pp. 171–269.

[24] M. J. Weber,

Handbook of Laser Wavelengths

. Boca Raton, FL: CRC

Press, 2000.

[25] G. R. Fowles and W. T. Silfvast, “Laser action in the ionic spectra of

Zinc and Cadmium,”

IEEE J. Quantum Electron.

, vol. QE-1, p. 131,

June 1965.

[26] W. T. Silfvast, “Efficient CW laser oscillation at 4416 A in Cd(II),”

Appl.

Phys. Lett.

, vol. 13, pp. 169–171, Sept. 1, 1968.

[27] H. G. Cooper and P. K. Cheo, “Ion laser oscillations in sulfer,” in

Proc.

Physics of Quantum Electronics Conf.

San Juan, Puerto Rico, 1965,

pp. 690–697.

[28] P. K. Cheo and H. G. Cooper, “UV and visible laser oscillations in fluo-

rine, phosphorus, and chlorine,”

Appl. Phys. Lett.

, vol. 7, pp. 202–204,

Oct. 1, 1965.

[29] K. Jain, “A nickel ion laser,”

Appl. Phys. Lett.

, vol. 34, pp. 845–846,

June 15, 1979.

[30] J. R. McNeil, G. J. Collins, K. B. Persson, and D. L. Franzen, “CW laser

oscillation in Cu II,”

Appl. Phys. Lett.

, vol. 27, pp. 595–598, Dec. 1,

1975.

[31] J. R. McNeil, W. L. Johnson, G. J. Collins, and K. B. Persson, “Ultravi-

olet laser action in He-Ag and Ne-Ag mixtures,”

Appl. Phys. Lett.

, vol.

29, pp. 172–174, Aug. 1, 1976.

[32] R. D. Reid, J. R. McNeil, and G. J. Collins, “New ion laser transitions

in He-Au mixtures,”

Appl. Phys. Lett.

, vol. 29, pp. 666–668, Nov. 15,

1976.

[33] E. I. Gordon, E. F. Labuda, and W. B. Bridges, “Continuous visible laser

action in singly ionized argon, krypton, and xenon,”

Appl. Phys. Lett.

vol. 4, pp. 178–180, May 15, 1964.

[34] M. H. Dunn and J. N. Ross, “The argon ion laser,” in

Progress in

Quantum Electronics

, J. H. Sanders and S. Stenholm, Eds. New York:

Pergamon, 1976, vol. 4, pp. 233–269.

[35] C. C. Davis and T. A. King, “Gaseous ion lasers,” in

Advances in

Quantum Electronics

, D. W. Goodwin, Ed. New York: Academic,

1975, vol. 3, pp. 169–469.

[36] E. F. Labuda, E. I. Gordon, and R. C. Miller, “Continuous-duty argon

ion lasers,”

IEEE J. Quantum Electron.

, vol. QE-1, pp. 273–279, Sept.

1965.

[37] I. P. Zapesochnyi, A. I. Imre, A. I. Daschenko, V. S. Vukstich, F. F.

Danch, and V. A. Kel’man, “Experimental investigation of the exci-

tation of Ar II and Kr II in electron-ion collisions” (in English),

Sov.

Phys.—JETP

, vol. 36, pp. 1056–1060, June 1973.

[38] E. F. Labuda, C. E. Webb, R. C. Miller, and E. I. Gordon, “A study of

capillary discharges in noble gases at high current densities,” in 18th

Gaseous Electronics Conference, Minneapolis, MN, 1965.