6. Ruby laser built from scratch.
This laser was for a future experiment to explore the possibility of sum frequency mixing Ruby 694.3nm with Nd:YAG 1064nm 4th harmonic 266nm to produce 193n, notwithstanding the non-availability of SFM NLO.
Ruby lasers rarely come up on eBay and when they do they are usually expensive, probably because nowadays they are of historical interest compared to the more mundane Nd:YAG.
A particular feature of ruby is it needs all light that can be thrown upon it and even minor shadows can cause lasing to cease. The helix flash lamp idolised after Theodore Maiman invented the first ruby laser using one in 1960, is a poor choice for this reason (more below).
In 2014 I bought one of the then few affordable ($100) used ruby rods from eBay Israel, advertised as 8mm dia but turning out to be 6mm, for which I got a 25% refund. This rod was pink along its entire length, complicating any mechanism holding it in place that would inevitably restrict light to the rod.
Around 2015, used ex-USSR ruby rods began to appear on eBay with bubble defects that would have caused them to be swapped out for new replacements, together with poorly stored new rods with chips rendering them equally useless. Not expecting new rods to appear, I bought one with minimal defects. Unlike the Israel rod, this had a ruby red centre section and transparent ends, satisfying ruby's lust for light: the transparent section allows opaque clamps, O-rings, etc. to be fitted without obstructing flash lamp light from reaching the chromium doped lasing section.
Then out of the blue in 2020 cheap new USSR ruby rods arrived accompanied by compatible USSR flash lamps See bottom of [Flash Lamps]. Inevitably, over time prices began to rise but ever more exotic rods appeared, perhaps culminating in 2021 when I saw a new 640mm x 45mm Nd:YAG rod for a mere $300 that sat there for several weeks. Temptingly beautiful, but who could use such a behemoth?
The sudden influx of affordable new USSR ruby rods meant I could replace my less than ideal earlier purchases, which is why I ended up with a third rod.
However a downside of USSR rods is a low Chromium doping level of around 0.02% to 0.03% vs the 'normal' 0.05%. This affects the lasing threshold, meaning the threshold is reached sooner and more light input produces less laser power output. Even so, I still hope to produce 1J of 694.3nm from rod #3.
DRIVE ELECTRONICS
Amongst my ruby rod driver research I came across Ben Krasnov's home built ruby laser blog and asked him how he chose his rod: http://benkrasnow.blogspot.co.uk/2014/01/ruby-laser-design-process.html
Ben said he'd used an Excel spreadsheet created by Doug Little and sent me a copy, also available at the Sam's excellent laser website: http://www.repairfaq.org/sam/laser/pfncalc1.xls
[O28], Mass Spectrometry Basics, 'Chromium [Cr] ions produce the red colour of ruby because they absorb blue/green light from white light, leaving unabsorbed red light to be transmitted.'
RUBY ROD STOCK
Unlike #2 and #3, the Israel rod is red throughout. Sold with the description 102mm x 8mm but on arrival found to be only 6mm diameter.
Rod #1 Cr area = 102mm x 6mm dia, πxDxL = 1923mm² [Israel]
#2: USSR used rod, total length 120mm, with a 75mm red section in the middle of two transparent end sections that lack Chromium, and which are for mounting the rod:
Rod #2 Cr area = 75mm x 8mm dia, πxDxL = 1885mm² [ex-USSR] ( 2% less area than rod #1)
#3: USSR new rod
Perfectly matches EG&G flash lamp arc length. The total length is 180mm and as with #2, there are two transparent end sections:
Rod #3 Cr area = 120mm x 8mm dia, πxDxL = 3016mm² [ex-USSR] (36% more area than rod #1)
The rods below are totally unsuitable due to their various defects, however the round ones are a similar construction to rods #2 and #3 above.
All of these poorly stored ex-USSR rods are 10mm dia x 180mm long with a 120mm long Chromium section:


Several of the USSR ruby rods have unconventional shapes although all have clear ends; some have a rectangular section and others have rings of bevelled edges; some a combination of both. I have seen it mentioned these were designed to work with helical flash lamps but I am not convinced due to the likelihood of shadows between the bevels. I suspect instead the bevels are intended to project more light into the rod from a conventional linear flash lamp.
These rods are unsuitable for lasers due to damage from poor storage but they were relatively inexpensive and I bought a few to investigate deep UV absorption [See FUTURE EXPERIMENT *TBD*]. They came from eBay seller rusgems based in Thailand, both on eBay and their website rusgems.com, and I highly recommend them. They also supplied these rod photos.






Helical (spiral) flash lamps are NOT a good idea
The first ruby laser was created by Theodore Maiman at Hughes Research laboratories on 16 May 1960,
using a 10mm dia x 15mm long ruby rod in a helical flashlamp: https://en.wikipedia.org/wiki/Ruby_laser
http://www.repairfaq.org/sam/lasercps.htm
'Note that while a helical flash lamp may look "cool", it's generally easier to deal with linear flash lamps for pumping laser rods - too much light energy is lost to reflections from the coils and out the ends of the cavity even if a perfect reflector surrounds them. Additionally, the long discharge path means helical flash lamps will tend to have larger Ko values and thus require higher voltage to achieve the same pulse width. If the helix can't get close to the Ruby rod (touching), it might require close to 1000 joules for the Ruby to lase compared to few hundred joules for a linear flash lamp.'
Applied Photonics, C.Yeh:
'The repulsive force between adjacent coil turns of a helical flash lamp may mechanically flex the glass tube and crack it, limiting both pulse width and input energy.'
EXCITING THE ROD
A useful reference for ruby laser design is [O17]: Design and Optimisation of Ruby Lasers in the USSR.
http://en.wikipedia.org/wiki/Laser_pumping
'The rod and the lamp are relatively long to minimize the effect of losses at the end faces and to provide a sufficient length of gain medium. Longer flash lamps are also more efficient at transferring electrical energy into light, due to higher impedance.[2] However, if the rod is too long in relation to its diameter a condition called "pre-lasing" can occur, depleting the rod's energy before it can properly build up.[3] Rod ends are often antireflection coated or cut at Brewster's angle to minimize this effect.[4] Flat mirrors are also often used at the ends of the pump cavity to reduce loss
Another configuration uses a rod and a flash lamp in a cavity made of a diffuse reflecting material, such as Spectralon or powdered barium sulfate. These cavities are often circular or oblong, as focusing the light is not a primary objective. This doesn't couple the light as well into the lasing medium, since the light makes many reflections before reaching the rod, but often requires less maintenance than metalized reflectors.[6] The increased number of reflections is compensated for by the diffuse medium's higher reflectivity: 99% compared to 97% for a gold mirror.[7] This approach is more compatible with unpolished rods or multiple lamps.'
RUBY ABSORPTION PEAKS AT 400nm & 560nm
Examining the graph below, where:
|| = incident light parallel to c-axis
T = incident light perpendicular to c-axis
AU = arbitrary unit
<=200nm intensity AU >> 3.6 regardless of T or ||
~260nm intensity AU = 0.85 T, 1.0 ||
404nm intensity AU = 3.2 T, 2.8 ||
554nm intensity AU = 2.8 T, 1.4 ||
From this graph appears ruby rod is best irradiated into the rod end T at 404nm and along the rod || at 554nm, but 532nm intensity is only 1.0 T so there is no real advantage to end pumping by 2HG 532nm Nd:YAG.
Below left, Northrup-Grumman's ruby absorption plot Below right, [O48] Page 6, Fig.2


UV ABSORPTION
Interestingly, absorption below 200nm goes off the scale. At first I wondered if it might have been a 'run-in' where the spectrometer began to work, rather than real data. Eventually after much fruitless searching, I emailed Northrop-Grumman not expecting to receive a reply, but instead one came almost immediately:
'The spectrum we have online is typical of Ruby. It shows a significant rise in the 200nm region, which is normal. The band-edge for sapphire is ~190nm or so and by doping sapphire with chromium (thus making Ruby) there will be an increase in the near band-edge absorption.'
I later found the following paper [O8] citing specific ruby absorption DUV wavelengths in terms of eV:
https://aip.scitation.org/doi/10.1063/1.1726965
'The absorption spectrum of thin samples of ruby with 0.5 to 0.005 weight % of Cr2O3 has been measured up to 8.9eV at room temperature with linearly polarized light. The spectrum shows an absorption peak at 7eV [177nm] followed by humps around 8eV [155nm] with absorption cross section and oscillator strength at the 7eV [177nm] peak ~7×10-17 cm² and ~0.3, respectively. The absorption with E [perpendicular to] c is generally larger than E||c in the vacuum-uv region. The absorption between 6.8eV [182nm] and 8.6eV [144nm] is interpreted as a charge-transfer transition around the chromium ion because of its large oscillator strength and some similarity to the ultraviolet reflectance spectrum of Al2O3 crystal.
The excitation spectrum of ruby cubes with several chromium concentrations has been measured to 11eV at room temperature with linearly polarized light. Structure and polarization effects corresponding to those of the absorption spectrum are observed. The quantum efficiency in the region 7eV [177nm] to 9eV [138nm] is comparable to that in the blue absorption band of ruby and is essentially concentration independent. Above 9eV the relative quantum efficiency drops about two orders of magnitude and is concentration dependent, with low quantum efficiency for samples of low chromium concentration.'
Furthermore, the XLS calculated %PI (Population Inversion) threshold equation is greatly affected by wavelength. I'm guessing the value of 475nm for ruby absorption may be derived from the area of the curve for ruby absorption, so takes into account ALL of the above peaks. The flash lamp optical characteristics are also included, see in the embedded comments: top RHS red triangles. The graph implies there is great absorption ~200nm and low level 200nm is generated by the flash lamp.
Summary:
a) The quantum efficiency (the efficiency of converting absorbed photons into emitted photons) in the 7eV to 9eV (177nm to 138nm) region is comparable to that of the blue absorption band of ruby (404nm peak) and is independent of chromium concentration.
b) Above 9eV (<138nm), the quantum efficiency drops sharply (by about two orders of magnitude) and becomes dependent on chromium concentration, with lower efficiency for samples with lower Cr³? content.
(which of course my USSR rods are).
DeepSeek (Feb 2025) added: 'In the deep UV region,
'1. High Absorption Efficiency: In this region, the absorption efficiency is typically very high due to the strong interaction of UV photons with the lattice. The exact absorption spectrum may include sharp peaks or edges corresponding to specific electronic transitions.'
2. Cooling the sample reduces thermal broadening of absorption peaks, allowing for more precise identification of electronic transitions; it’s common for such studies to be conducted at low temperatures e.g. Nitrogen-cooled 77K (-195°C) to reduce thermal broadening and reveal finer details in the absorption spectrum.
3. Host Lattice absorption: Below 200nm, the absorption of ruby (Al₂O₃:Cr³⁺) and sapphire (Al₂O₃) is dominated by electronic transitions in the aluminum oxide lattice itself, rather than the Cr³⁺ impurities. These transitions involve the valence band (oxygen 2p orbitals) and the conduction band (aluminum 3s and 3p orbitals).'
I appended the nm equivalent of the eV above using the following app that converts eV to nm:
http://halas.rice.edu/conversions
9.0eV = 138nm, 8.9eV = 139nm, 8.6eV = 144nm, 8.0eV = 155nm, 7.0eV = 177nm, 6.8eV = 182nm
The deep UV absorption band is of interest because it is far more sensitive than any of the other wavelengths, and my EG&G Xenon flash lamps have a quartz wall which passes UV.
DOES UV RADIATION DAMAGE RUBY?
I needed to determine how destructive UV radiation is to ruby. If the answer is it can survive a thousand shots, that is more than enough to complete my experiments.
US National Bureau of Standards, Electron Paramagnetic Resonance Intensity Standard SRM-2601, [O27]
http://archive.org/details/standardreferenc2605chan, Page 45:
'5.5 Precautions Concerning Visible, UV, X-Ray and Nuclear Radiations
It is well known that ruby suffers damage from intense radiation [8], Usually nuclear radiations are not encountered in EPR [Electron Paramagnetic Resonance] experiments. However, it is possible that x-ray, UV, or strong visible light may be used in certain applications. This may cause damage or may induce undesirable fluorescence.
Damage to ruby can be detected by a change of color, from pink to orange. This can be best observed against a dark background, preferably by comparison with an undamaged ruby SRM [Standard Reference Material]. Exposure to room light at room temperature for several days will usually bleach the discoloration. The effects of low levels of irradiation on the EPR signal is small and reversible [8].
In optical EPR experiments using UV irradiation, the ruby will usually fluoresce, giving off the prominent R-lines (at 694nm). Some of the common varnishes or household cements used to attach the SRM to the cavity or holder, also fluoresce when excited by UV radiation.'
Page 47: '8. R.Wenzel, J.Halpin, F.Campbell, Report NRL-7063 (AD-704069), Naval Res. Lab., 20-Feb-70'
Unfortunately I can find no online copy of NRL-7063.
Laser Induced Damage in Optical Materials, 1978:
http://books.google.co.uk/books?id=tQXxIOc1QEkC&pg=PR9&lpg=PR9&dq=is+ruby+damaged+by+uv?
Laser Induced Damage in Optical Materials: Proceedings of a Symposium Sponsored by the American Society for Testing and Materials and by the National Bureau of Standards: 1979
http://books.google.co.uk/books?id=9OqjlgG2oE8C&pg=PA167&dq=is+ruby+damaged+by+uv?
The overall impression is that yes, ruby is damaged by UV radiation, but I have not been able to quantify how much. It seems I shall have to experiment myself, which is why I bought those old speckled ex-USSR ruby rods. In addition, I'll use them to determine the relative fluorescence intensity of the rod over its absorption bands.
However http://en.wikipedia.org/wiki/Laser_pumping: 'laser host materials are chosen to have a low absorption; only the dopant absorbs. Therefore any light at frequencies not absorbed by the doping will go back into the lamp and reheat the plasma, shortening lamp life.'
So there is a compromise between UV radiation damaging the rod as well as being re-admitted back into the lamp and damaging that too.
DUAL RUBY + Nd:YAG FROM SAME FLASHLAMP
Could I increase ruby absorption by adding a 532nm NLO in the optical path from the Nd:YAG to the ruby?
Paper [N32] compares intracavity vs extracavity NLO generated 532nm from a Q-switched Nd:YVO4 laser with the aim of producing 266nm and concludes extracavity produces more power vs intracavity due to diminished optical timing of the fromer. Although not tried with ruby, I expect similar results.
FLOW TUBES
In order to protect both the rod and flash lamp from breakage, they are usually mounted inside separate glass flow tubes that are sufficiently large to allow coolant to flow through them and over the rod and flash lamp.
A flow tube is often made of glass or quartz doped with a material that absorbs and blocks the transmission of UV light which is generally damaging to lasing media and as we discovered above, this includes ruby. The dopant converts the UV by fluorescence, to wavelengths best suited to the absorption bands of the lasing material.
Cerium is the doping agent best suited for the ruby rod:
Spectroscopic Properties of Cerium-Doped (Ce3+) Aluminosilicate Glasses':
http://www.osapublishing.org/ome/fulltext.cfm?uri=ome-5-4-720&id=312757
From this paper it is evident cerium blocks below 350nm with absorption ~350nm to 400nm and fluoresces ~440nm to 540nm depending on concentration.
Below left - quartz doped with Samarium and Cerium modifies the wavelengths of transmitted light:
a) Cerium is best suited to ruby,
b) Samarium is best suited to Nd:YAG.

Frosted flow tubes increase gain. http://en.wikipedia.org/wiki/Laser_pumping: 'Parasitic modes occur when reflections are generated in directions other than along the length of the rod, which can use up energy that would otherwise be available to the beam. This can be a particular problem if the barrel of the rod is polished. Cylindrical laser rods support whispering gallery modes due to total internal reflection between the rod and the cooling water, which reflect continuously around the circumference of the rod. Light pipe modes can reflect down the length of the rod in a zig-zag path.
If the rod has an anti-reflection coating or is immersed in a fluid that matches its refractive index, it can dramatically reduce these parasitic reflections. Likewise, if the barrel of the rod is rough ground (frosted) or grooved, internal reflections can be dispersed.[8]'
'Pumping with a single lamp tends to focus most of the energy on one side, worsening the beam profile. It is common for rods to have a frosted barrel to diffuse the light, providing a more even distribution of light throughout the rod. This allows more energy absorption throughout the gain medium for a better transverse mode. A frosted flow tube or diffuse reflector, while leading to lowered transfer efficiency, helps increase this effect, improving the gain.[9]'
How close is water to the refractive index of ruby?
DeepSeek:
'Water has a refractive index of 1.33, while ruby's refractive index is 1.76-1.77.
Ruby's refractive index is about 32%-33% higher than that of water
Ruby has a much higher refractive index, meaning it bends light more strongly than water.'
Otoh FC-77 is even worse:
'FC-77 has a refractive index of 1.29
Ruby's refractive index is about 36-37% higher than that of FC-77.'
Whilst neither is ideal, FC has been used successfully and water is slightly better than FC.
THE COLDER THE BETTER
Ruby benefits from a low operating temperature: 'Internal thermal limits of ruby laser performance' [O26] Page 4: 'The lower the operating temperature of the ruby the higher the efficiency of the device.'
Cooling a Ruby laser below room temperature can increase output energy; cooling to 0°C vs 20°C can increase it by perhaps 50%: http://en.wikipedia.org/wiki/Ruby_laser
See also [O26], Internal Thermal Limits of Ruby Laser Performance, Page 4:
'The lower the operating temperature of the ruby the higher the efficiency of the device.'
The flow tube would therefore be best placed around the rod and the cooling liquid directed through it first before cooling the flash lamp which will naturally raise the coolant temperature.
Reflector
http://en.wikipedia.org/wiki/Ruby_laser
'The rod and the lamp are relatively long to minimize the effect of losses at the end faces and to provide a sufficient length of gain medium. Longer flash lamps are also more efficient at transferring electrical energy into light, due to higher impedance.[2] However, if the rod is too long in relation to its diameter a condition called "pre-lasing" can occur, depleting the rod's energy before it can properly build up.[3] Rod ends are often antireflection coated or cut at Brewster's angle to minimize this effect.[4] Flat mirrors are also often used at the ends of the pump cavity to reduce loss
Another configuration uses a rod and a flash lamp in a cavity made of a diffuse reflecting material, such as Spectralon or powdered barium sulfate. These cavities are often circular or oblong, as focusing the light is not a primary objective. This doesn't couple the light as well into the lasing medium, since the light makes many reflections before reaching the rod, but often requires less maintenance than metalized reflectors.[6] The increased number of reflections is compensated for by the diffuse medium's higher reflectivity: 99% compared to 97% for a gold mirror.[7] This approach is more compatible with unpolished rods or multiple lamps.
Aluminium is a far better reflector of UV than gold and will reflect deep UV to the cerium doped quartz flow tube around the rod; silver does not extend down to UV but is otherwise similar:


The curve above left shows the the reflection efficiency of several materials over optical bandwidth; the curve above right shows the effective bandwidth of aluminium vs silver reflector for a Xe flashlamp.
EXCITING THE RUBY LASER WITH A DIODE LASER
From the ruby curve, of some interest is the option to pump it from laser diode 808nm second harmonic 404nm to match its peak absorption around 400nm. However the graph from the Giant Pulse thesis (above) indicates 400nm pumping is more efficient at the rod end and 560nm is better pumped into the rod along its length.
The 13 May 2009 edition of Laser Focus World revealed the Klastech GmbH LD pumped Crescendo ruby laser, although since being acquired by Power Technology Inc in 2013, I've found no further mention of it:
The LHC project uses flashlamps primarily due to their lower cost.
EXCITING THE RUBY LASER WITH A FLASHLSMP
Of interest below are my hand drawn curves derived from several papers, comparing the emission bands of krypton and xenon flashlamps with the absorption bands for ruby, Alexandrite, Ti:Sapphire and Nd:YAg lasers; all have different requirements.
However the following patent reveals a Xenon Krypton flash lamp is best suited to ruby:
https://patents.google.com/patent/US3440559
'...a xenon filled gas tube is not as effectively efficient a light stimulator for a ruby laser as is desirable. This is because the emission spectrum of xenon does not match the absorption spectrum of ruby, which has two peak absorption regions in the visible spectrum at approximately 570nm and 420nm whereas xenon has emission peaks at approximately 467.1nm, 462.4nm and 450nm....a gas filling of xenon and krypton [is better]. A tube with a filling of 50% krypton and 50% xenon has been found to be a substantial improvement over a known xenon filled tube and experiment indicates that still better results follow reduction of the krypton content substantially below 50%. Krypton has emission peaks at approximately 587nm and 557nm, the 587nm line peak coinciding with the second absorption peak of chromium doped ruby...a reduction in the neighbourhood of 25% of the energy required for ruby excitation is obtainable with XeKr vs a purely Xe filled tube.'[text modified for readability]
Something I shall clearly have to do, is determine if my EG&G flash lamps are Xenon (most likely), krypton, or Xenon Krypton.
Xenon and Krypton spectral curves are readily available but I have not yet found a XeKr curve. If I can't find a XeKr flash lamp, could I run both a Xe and a Kr lamp in series for the same rod?
From [O8], 177nm is ruby's UV absorption peak.
Deepseek: 'A Xenon Flash Lamp can emit at 7 eV (177 nm) but is unlikely to emit strongly at 8 eV (155 nm) due to window material and gas absorption limitations.
A Xenon-Krypton Flash Lamp ishas minimal emission at 8 eV (155 nm).
If you need strong emission at 7 eV (177 nm) and 8 eV (155 nm), you would typically use a specialized VUV light source, such as a deuterium lamp:
Light Source Nominal Intensity at 7 eV (177nm) [photons/(s·nm·sr)]
Deuterium Lamp 10^14-10^15
Xenon Flash Lamp 10^12-10^13
Xenon-Krypton FL 10^12-10^13'



Above left, Flash lamp output at visible wavelengths best suited to ruby significantly increases when operated at high currents (graphs from Perkin and Elmer flash lamp catalogue [O11]), but this in turn limits their life; UV absorption is high for ruby and quartz admits UV, but UV eventually damages rods.
Below, Perkin Elmer flash lamp Xe and Kr flash lamp spectral emission graphs:

The reason why I eventually settled for a rod with a 120mm long ruby section is because towards the start of the project I bought two unknown EG&G flashlamps that look very similar to the Perkin Elmer QXF series: 'liquid cooled xenon filled high average power medium peak power':
6mm bore, Ko 26.7, Ke 184000W, 16kV trigger, Iavg1k1A, Pavg 4713W, 1ms, 800V-2800V.
However as noted above, Xenon krypton flash lamps may be optimum for ruby.
FLASH LAMP DRIVER CALCULATIONS
Below left: from CORD Module 3-2 Pulsed Flash Lamps and Power Supplies: Flash lamp explosion energy as a function of pulse duration, expressed as the duration at one-third peak amplitude of the electrical input to the lamp, with the inside bore diameter as a parameter. (G. A. Hardway, MicroWaves 5, 46, 1966)
https://pe2bz.philpem.me.uk/Lights/-%20Laser/Info-902-LaserCourse/c04-04/mod04_04.htm
Right: CORD Fig 31 Flash lamp life as a function of the ratio of the input energy vs explosion energy.


I think this is Ben Krasnow. At 08:19 he references the explosion energy graph above left for his DIY ruby laser. https://www.youtube.com/watch?v=VSdUQ4oNKiI
RUBY LASER CONSTRUCTION
Having acquired rod #3 to match the EG&G flash lamps, I sketched the assembly (below left), deciding to make the end blocks out of Delrin, a hard plastic often used in similar applications. The aluminium reflector would be held within the blocks by four steel threaded rods at each corner of the Delrin block, slack taken up by a sheet of neoprene rubber at each end. The assembly was intended to be cooled by my stock of C-74 perfluorocarbon (see Cooling Loop section) which would run through the reflector with no need for flow tubes.




Above: Clockwise,the machining drawings for the two Delrin ends to the ruby cavity, the overall dimensions of the completed cavity, and the completed assembly view (from my logbook).
Below, the finished blocks:



RUBY LASER REFLECTOR
Whilst the Delrin blocks were being machined, I began to make the reflector.
Laser flash lamp reflectors can take several forms, largely depending upon the number of flash lamps but an elliptical tube is more efficient than a round tube because it focuses the light onto the rod:
https://en.wikipedia.org/wiki/Laser_pumping
'Smaller ellipses create fewer reflections, (a condition called "close-coupling"), giving higher intensity in the centre of the rod.[1] For a single flash lamp, if the lamp and rod are equal diameter,
an ellipse that is twice as wide as the height is the usually the most efficient at imaging the light into the rod.'
My experimental ruby laser cavity tube is an aluminium pipe intended for an automotive exhaust fitting. Understandably its inside doesn't have the nice chrome finish that was on the outside but this isn't an issue: the reflectivity curves below reveal aluminium is a far better reflector of UV than chrome and indeed gold, normally used for Nd:YAG cavities.
I cut the aluminium tube to size with a simple pipe cutter then I made an electro-polisher using 80% phosphoric acid in a pint beer glass and various pieces of polypropylene block (inert in the acid) as a scaffold to hold the tube in place. With only crude wood cutting tools at my disposal, the finish is poor but usable. I put a smaller aluminium rod in the centre extending the full length of the tube to be polished, and used it for the positive terminal. The ruby tube is the negative terminal and to prevent the outside being etched as well, I covered that in adhesive backed paper masking tape.




The small tube also doubled up as an aerator using a small pump feeding into the top and an airstone at the base, secured by crimping the end of its tube. To get the starting temperature up to 74°C, I placed the acid filled glass on top of a heating block, and initially used an electrophoresis PSU but although it had some effect, much more current was required.


I hooked up my industrial automotive battery charger from the garage and set to 12V, the solution took 22A from the supply, melting the insulation on the feeble croc clips and burning the masking tape:



After rinsing and drying, I finished off the inside of the tube to a mirror like surface by rubbing it down with Autosol aluminium polish and a soft cloth.
Below left: after electro-polishing Below right: after hand polishing


CHANGE OF PLAN
The assembly was intended to be cooled by my stock of C-74 perfluorocarbon but my research into cooling fluids took place after this decision, see [Cooling Loop], as I had previously earmarked the FC-74 for a PC overclocking project when I believed it be be inert. My new research proved otherwise, and discovering ruby benefits from cold also came later. As they say, the benefit of hindsight is wonderful. These new discoveries meat I will now be cooling the laser with water, and I will require flow tubes to protect the flash lamps and rod (although I'll still run the UV absorption experiment). The twin flowtube design allows freshly cooled water to first flow across the ruby rod before passing through the flash lamp flowtube. This in turn means I will have to re-machine the Delrin blocks to accommodate the flowtubes, or perhaps abandon them and start again. However finding affordable flow tubes for the rod and flash lamps is difficult, and this is now compounded by my desire to research the benefits of DUV.
Furthermore initially it was my intention to use the two EG&G flash lamps to build the Nd:YAG before the MK580KK appeared. When I came to consider building the ruby, they seemed the ideal choice and I dimensioned the blocks to accommodate them. Now, much later, I think they are too powerful and operating flash lamps below their rated power is not good for them either, see [O11]: PE flash lamps catalogue, plus I only have two of them.
However like the ruby rods, in 2021 there was now a glut of very cheap ex-USSR laser flash lamps on eBay, and I bought a dozen to experiment with. They seem very crude in comparison to the EG&G lamps, but they are so cheap they are expendable whereas the EG&G lamps are not. The USSR flash lamps also come with specifications, lacking from the EG&Gs whose parameters I have only guessed.
To compound the issues, I tried to compress the reflector towards an ellipical profile using a large vice but it warped, ruining it. Later I discovered a simple approach to determine the optimum ellipse using a larger reflector, but this was reliant on being machined from a solid block of aluminium. For now all I can produce is a polished round pipe.
It also seemed more practical to redesign the reflector for the USSR flash lamps instead but after Putin invaded Ukraine all Russian sources on eBay were banned and the design is now in a state of flux. This is why still in 2025 I haven't got much further with construction.
MIRRORS
From: http://www.repairfaq.org/sam/lasercps.htm
'The HR can be either a totally reflecting **HE** dielectric mirror or a roof prism or corner reflector, AR coated for the laser wavelength if possible. A high quality first surface aluminized mirror might survive for a few shots or more but with the high intra-cavity power of a PSS laser, possibly not for long. And, probably not at all if Q-switched! The OC mirror really needs to be one designed for PSS laser.Without a Q-switch, figure something like 70 percent reflectivity at the lasing wavelength.With a Q-switch, it can be as low as 30 percent
I'd suggest avoiding "resonant" type mirrors in favor of high energy dielectric types. While resonant optics have no coatings to be damaged, their internal optical element spacing and alignment are absolutely critical (if disassembled for any reason, will likely be ruined). And, their condition or even effective reflectivity can't easily be determined. The reflectivity is also likely to be too low for useful operation without a Q-switch. Although the original ruby laser, circa 1960, used an aluminum output mirror with a hole in the middle, this is probably not a great idea (but it wouldn't hurt to try!).'
Below left: Typical ruby OC reflectivity vs output power; 70% is peak, but 50% is more stable
from [O19], CORD Module 3-5, Pulsed Solid-State Laser Systems:

Chris Hardy (Kigre) gave the following advice which should help me drive my unknown flash lamps: 'Using a good hall-effect current probe (Such as Pearsons), a high voltage probe (Tektronix), and oscilloscope (Tek, etc) you can determine the lamp's impedance at any time in the pump pulse. The most convenient way is to connect the probes to the scope and the scope to a computer to calculate the impedance. You can also measure the input energy as the integration of lamp voltage and current over time (the pump pulse) using a scope and computer. Some fancy scopes can do this without a computer if they have advanced math capability. Published formulas for Ko (lamp impedance) are pretty accurate.'
FURTHER UV ABSORPTION RESEARCH NOTES
2025 Deepseek (October 2023 dataset):
Q Can a ruby rod be made to lase when irradiated by 200nm?
No, a ruby rod cannot be made to lase when irradiated by 200 nm light. Here's why:
-
Ruby Laser Basics: A ruby laser is a type of solid-state laser that uses a synthetic ruby crystal (aluminum oxide doped with chromium ions, Cr³⁺) as the gain medium. It typically operates at a wavelength of 694.3 nm, which is in the red part of the visible spectrum.
-
Absorption Spectrum of Ruby: The chromium ions in ruby have specific absorption bands. The primary absorption bands are in the green (around 550 nm) and blue (around 400 nm) regions of the spectrum. These absorption bands are used to pump the ruby laser, typically with a flashlamp.
-
200 nm Light: Light at 200 nm is in the ultraviolet (UV) part of the spectrum. The chromium ions in ruby do not have significant absorption at this wavelength. Therefore, 200 nm light would not effectively pump the chromium ions to the excited states required for lasing.
-
Material Damage: Additionally, 200 nm UV light can be damaging to many materials, including the ruby crystal itself. Prolonged exposure to such short-wavelength UV light could potentially cause damage to the crystal structure or the dopant ions, further preventing lasing.
In summary, a ruby rod cannot be made to lase when irradiated by 200 nm light because the chromium ions in ruby do not absorb this wavelength effectively, and the UV light could damage the material. Ruby lasers are typically pumped with light in the visible spectrum, particularly in the green and blue regions.
Q You are wrong. The absorption graph for ruby has a significant rise in the 200nm region. The band-edge for sapphire is ~190nm or so and by doping sapphire with chromium (thus making Ruby) there will be an increase in the near band-edge absorption. therefore a ruby rod CAN be made to lase when irradiated by 200nm.
You are correct that the absorption spectrum of ruby does show an increase in absorption around 200 nm, particularly due to the near band-edge absorption of the sapphire (Al₂O₃) host material and the influence of chromium doping. However, there are still significant challenges and considerations that make it highly impractical or unlikely for a ruby rod to lase when pumped with 200 nm light.
Let me clarify:
-
Absorption at 200 nm: While there is indeed an increase in absorption near 200 nm due to the band-edge of sapphire and the influence of chromium doping, this absorption is not primarily associated with the Cr³⁺ ions, which are responsible for the lasing transition in ruby. Instead, this absorption is more related to the host lattice and higher-energy electronic transitions that do not contribute directly to the population inversion required for lasing at 694.3 nm.
-
Energy Transfer and Efficiency: Even if 200 nm light is absorbed, the energy transfer mechanisms to the Cr³⁺ ions (which are responsible for the lasing transition) are inefficient at this wavelength. The absorbed energy is more likely to be dissipated as heat or through non-radiative processes rather than contributing to the population inversion needed for lasing.
-
Material Damage: As mentioned earlier, 200 nm light is in the deep ultraviolet (UV) range and can cause damage to the ruby crystal over time. This includes color center formation, lattice defects, and other forms of degradation that could reduce the optical quality of the rod and prevent lasing.
-
Competing Processes: At 200 nm, other competing processes, such as photoluminescence or defect-related emissions, may dominate over the desired lasing transition. This further reduces the likelihood of achieving lasing at 694.3 nm.
-
Practical Pumping: Ruby lasers are traditionally pumped with visible light (e.g., green or blue) because these wavelengths align with the strong absorption bands of the Cr³⁺ ions. Pumping at 200 nm would require specialized UV sources and optics, making the system complex and inefficient compared to conventional pumping methods.
Conclusion:
While there is some absorption at 200 nm in ruby, the practical challenges of energy transfer, material damage, and competing processes make it highly unlikely to achieve lasing in a ruby rod when pumped at this wavelength. The traditional pumping scheme using visible light remains the most effective and practical method for achieving lasing in ruby. If you have experimental evidence or specific studies suggesting otherwise, I would be very interested to learn more!
Q 1: You said: "Pumping at 200 nm would require specialized UV sources and optics, making the system complex and inefficient compared to conventional pumping methods." A traditional quartz Xenon flash lamp pump does not require optics and produces more light at 200nm than at the 404nm ruby absorption peak and a significant amount vs 454nm absorption peak.
2: You said: "there is some absorption at 200 nm" Your word 'some' is misleading because the absorption at 200nm exceeds that of any other wavelength peak for ruby.
3: My application is for single shot pulses several seconds apart and will not exceed 100 pulses per day. Whilst it is true UV damages ruby (and any flash lamp), the low pulse rate ensures acceptable rod and flash lamp life. Normally ruby flash lamps have a cerium coating to block UV below 250nm but I believe with uncoated quartz I can significantly increase ruby emission by combining visible absorption peaks with the full band of UV that is absorbed by ruby.
4: Can you find any information that computes the relationship between UV intensity and ruby rod damage?
5a: My absorption graph (UBC_1969_A6_7, page 6 Fig 2) stops at 200nm but at this point absorption is off the scale so it clearly absorbs well below 200nm.
5b: Can you point me to any research or graph that shows UV absorption intensity of ruby below 200nm? 6: Below which UV wavelength does absorption of ruby change from Cr³⁺ ions, to mechanisms that do not contribute directly to the population inversion?
Thank you for the detailed clarification and context. Let me address your points systematically and provide the best information I can based on your requirements and questions:
1. Pumping at 200 nm with Xenon Flash Lamps
You are correct that a traditional quartz Xenon flash lamp emits significant UV light, including at 200 nm, and does not require additional optics for pumping. The emission spectrum of a Xenon flash lamp indeed extends into the deep UV, and the intensity at 200 nm can be substantial. If the quartz envelope is uncoated (i.e., not blocked by cerium or other UV-absorbing coatings), the lamp will transmit this UV light efficiently. This could theoretically allow for additional pumping of the ruby rod via UV absorption, complementing the visible absorption peaks (404 nm and 554 nm).
2. Absorption at 200 nm
I agree that my use of the word "some" was misleading. The absorption of ruby at 200 nm is indeed very strong, often exceeding the absorption at the visible peaks (404 nm and 554 nm). This is due to the near band-edge absorption of the sapphire host material and the influence of chromium doping. Your observation that the absorption at 200 nm is "off the scale" on your graph (UBC_1969_A6_7, page 6, Fig 2) aligns with the known behavior of ruby in the UV region.
3. UV Damage and Low Pulse Rate
Your application for single-shot pulses with long intervals (several seconds apart) and a low daily pulse count (=100 pulses/day) significantly reduces the risk of UV-induced damage to the ruby rod and flash lamp. While UV can cause color center formation and lattice defects in ruby, the low duty cycle ensures that the cumulative damage remains within acceptable limits for your use case. Uncoated quartz flash lamps will indeed allow more UV to reach the ruby rod, potentially increasing the overall pumping efficiency by leveraging both UV and visible absorption.
4. Relationship Between UV Intensity and Ruby Rod Damage
Unfortunately, I could not find specific quantitative studies that compute the exact relationship between UV intensity and ruby rod damage. However, the following general principles apply:
UV Damage Mechanisms: UV light can create color centers (e.g., F-centers) and lattice defects in ruby, primarily due to the high energy of UV photons. These defects can degrade the optical quality of the rod over time.
Damage Threshold: The damage threshold depends on the UV wavelength, pulse energy, repetition rate, and cooling of the rod. For your low-duty-cycle application, the damage is likely minimal.
Empirical Data: Most ruby laser systems use cerium-coated flash lamps to block UV below 250 nm, as this is known to extend the lifetime of the rod. However, in your case, the trade-off between increased pumping efficiency and reduced rod lifetime may be acceptable.
If you are designing a system, I recommend conducting empirical tests to quantify the damage over time under your specific operating conditions.
5. UV Absorption Below 200 nm
5a. Absorption Below 200 nm
Your observation that the absorption at 200 nm is "off the scale" suggests that ruby continues to absorb strongly below 200 nm. This is consistent with the near band-edge absorption of sapphire (Al2O3), which has a bandgap of ~9.9 eV (~125 nm). Doping with chromium (Cr³?) introduces additional absorption features, but the host material's band-edge absorption dominates at these shorter wavelengths.
5b. Research or Graphs on UV Absorption Below 200 nm
I could not locate specific graphs or studies showing ruby's absorption below 200 nm in the literature I have access to. However, the following points may help:
The absorption below 200 nm is likely dominated by the sapphire host material's band-edge transition, which becomes increasingly strong as the photon energy approaches the bandgap.
Chromium doping may introduce additional absorption features, but these are typically weaker compared to the host material's intrinsic absorption in this region.
You may need to consult specialized spectroscopy studies or databases (e.g., NIST, or journals like Journal of Applied Physics) for detailed data on ruby's absorption in the far-UV region.
6. Transition from Cr³? Absorption to Non-Lasing Mechanisms
The transition from Cr³?-related absorption to non-lasing mechanisms occurs when the photon energy exceeds the energy levels associated with the Cr³? ions. Specifically:
Cr³? Absorption: The Cr³? ions in ruby have strong absorption bands in the visible region (404 nm and 554 nm) and weaker absorption in the UV region (e.g., 200 nm). These transitions contribute to the population inversion required for lasing at 694.3 nm.
Host Material Absorption: Below ~250 nm, the absorption is increasingly dominated by the sapphire host material's band-edge and defect-related transitions. These mechanisms do not contribute directly to the population inversion for lasing.
Threshold Wavelength: The exact wavelength at which Cr³? absorption becomes negligible depends on the crystal's electronic structure, but it is generally accepted that absorption below ~250 nm is dominated by the host material.
Summary
A ruby rod can indeed be pumped at 200 nm, and the strong absorption at this wavelength may enhance the overall pumping efficiency when combined with visible absorption peaks.
Your low-duty-cycle application minimizes the risk of UV-induced damage, making uncoated quartz flash lamps a viable option.
While I could not find specific data on ruby's absorption below 200 nm or quantitative damage thresholds, the general behavior of ruby in the UV region supports your approach.
The transition from Cr³?-related absorption to non-lasing mechanisms occurs below ~250 nm, where the host material's absorption dominates.
If you have access to specialized spectroscopy databases or journals, I recommend searching for far-UV absorption data on ruby or sapphire to refine your design further. Let me know if you'd like help with specific search strategies!
=================================
Merlin (2021 database):
while ruby does absorb light below 200nm, the absorption-to-emission efficiency (aka quantum efficiency, QE) drops below 50% around the 300nm to 350nm range
Non-Radiative Losses:
If the absorbed photons have energies that do not correspond efficiently to the necessary transitions (e.g., if they are too high or too low), the likelihood of non-radiative relaxation paths increases. High-energy photons (like those at 200 nm) may not promote effective transitions and can lead to heating or damage within the crystal structure rather than productive laser emission.
The bandgap of Sapphire: Approximately 8.8 eV
Corresponding Wavelength: Approximately 141 nm (rounded to two decimal places)
Q+++DeepSeek: Merlin AI states sapphire has a bandgap of ~8.8 eV. Where did you (DeepSeek) get your answer of 9.9eV from?
The discrepancy in the reported bandgap values for sapphire can originate from different sources and interpretations of the material's properties. Sapphire (Al2O3) is a wide bandgap semiconductor with commonly reported bandgap values in the range of 7.8 eV to 9.9 eV, depending on the specific crystalline structure and conditions under which the measurements were made.
8.8 eV: This value is often cited in older literature or certain references, possibly reflecting specific conditions or types of sapphire.
9.9 eV: The value of approximately 9.9 eV, corresponding to ~125 nm, is widely accepted in more recent research and provides a commonly referenced bandgap for sapphire derived from various experimental methods including optical absorption measurements.
Q+++DeepSeek stated its main database is from 2023.
Merlin: Which year is your general database based on?
My knowledge is based on information available up until September 2021. This includes scientific literature, textbooks, and other sources available before that cutoff date.
text
µ Ω ± ° ⌠ ⌡ ∫ │ ─ √ φ θ Θ ∂ δ ζ ξ ς λ ψ ω τ µ Ω ∆ Δ ∑ ∏ π Ξ ○ ≠ ³ ² ±