Esmectita + Zeolitas ± Clorita ± Calcita: Prof (254.5 – 369.5) m

2 .1 .1 H istorical review

Laser action at 2.1/xm from the level o f holmium was first reported by Johnson et al in 1962^°. T h e narrow ab sorp tion bands o f the h o lm iu m io n , th o u g h , w ere unsuitable for efficien t flashlamp pumping. In 1966, Johnson et al reported^^ on im proved laser perform ance from the holm ium ion in w hich flashlam p ligh t was absorbed and transferred to the holmium ion via erbium, ytterbium and thulium ions. The main reason for the improvement was the broad absorption in the visible region o f the op tical spectrum o f the erbium ion and a cross-relaxation p rocess w hich transferred excitation to thulium and then to the upper laser state o f holm ium . H ow ever, e ffic ie n t laser operation was lim ited to c ry o g en ic tem peratures due to thermal population o f the terminal laser lev el, the highest sublevel o f which lies at only ~530cm'^ above ground.

In the 19 years following Johnson's early observations, there were no further reports o f a multiply doped crystal in which erbium, ytterbium and thulium are used to transfer flashlamp energy to holmium, only those in which erbium and thulium are used. There is no docum ented ev id e n ce for the ex clu sio n o f the ytterbium ion although a likely reason would be the absorption bands o f the ytterbium ion w hich, being centred around 0.9/xm , may have been considered to be beyond the output spectrum o f the pump sources therefore making its presence unnecessary.

The sensitivity o f the Er:Tm:Ho:YAG crystal, also known as alpha-beta (cxR) and alphabet h o lm iu m :YAG, to tem perature was dem onstrated by R em ski and Smith^^ by measuring the changes in the laser threshold energy between 90K and

300K . A slow rate o f change below 200K produced thresholds for a 50mm x 3mm diameter al5Ho: YAG crystal, pumped with a xenon flashlamp, o f about 8J. However, above 250K the rate o f change increased rapidly resulting in thresholds in the region o f 70J at 300K indicating the very strong dependence on temperature. Additionally, Johnson et al identified^' that, as a result o f the changes in thermal population o f the low er laser le v el, different transitions w ere observed depending on the operating temperature.

D espite the laser thresholds which increase with temperature, C hicklis et al reported respectable room temperature operation o f aCHoiYLF in both Fixed-Q and Q-Switched modes^*. Using a rotating mirror to modulate the resonator losses, Q-SwitCh energies in excess o f 500mJ were obtained for flashlamp discharge energies o f on ly lOOJ. T he authors attribute this perform ance to the in sen sitiv ity o f the fluorescent lifetim e o f the upper laser level to temperature in YLF allowing a greater amount o f energy to be stored. When C hicklis et al compared the Fixed-Q m ode performance o f the alphabet mixture in YLF and YAG hosts^^, they observed slope efficien cies in YLF over tw ice those observed in YAG. A dditionally the YAG was found to suffer from internal damage far sooner than the YLF crystal prompting them to encourage further work on the «15Ho:YLF based lasers.

Continuing work at cryogenic temperatures continued to show the promise o f the sensitiser scheme. Moving from Fixed-Q operation o f HoF^^'' to continuous-wave operation o f aBHo: YAG at 77K , D evor et af^ achieved 2 0 W average output pow er at an overall efficien cy o f 4%. U sing a rotating mirror Q -Sw itch, over 70% o f the CW p o w e r w as o b ta in ed in a se r ie s o f sp ik es 1 0 0 -3 0 0 n s in d u ration , further emphasising the excellent energy storage capabilities o f thé system. Further increases in the CW operating efficiency were subsequently reported by Beck and Giirs^^ who recorded an output o f 50W and a 6.5% slope efficiency from an aBHoiYAG crystal, cooled to 77K with liquid nitrogen. To date this remains one o f the most efficien t, lamp-pumped, CW laser ever reported.

Despite the recommendations o f Chicklis et al, investigative work continued on the sensitised Ho: YAG crystal, alongside that with the YLF crystal, due to the fa v o u r a b le r e p o r ts b y B e ck and G iirs and th e b e tte r h ig h p o w e r h a n d lin g characteristics compared to YLF, although cryogenic cooling was still required^'

The most significant breakthrough in the development o f a room temperature system operating around 2fxm occurred in 1985 when Antipenko""' published the first results o f room temperature operation o f a YAG crystal doped with chromium in

place o f erbium. For the first tim e, room temperature operation comparable to that obtained p rev io u sly on ly for the YLF crystal w as com bined with the better th e r m o -m e c h a n ic a l c h a r a c te r is tic s o f th e YAG c r y s t a l. S in c e 1 9 8 5 fu rth er improvements in the laser efficiency have continued to be reported, most, though not all, relying on the chromium:thulium:holmium:YAG (CTH: YAG) crystal.

Follow in g the early reports by Russian researchers^"*'^^ the first reports by western groups using flashlamp pumped CTH: YAG occurred in 1987 when Storm et aP^ reported a 0.75% slope efficien cy and a 45J threshold energy for a 3"x4mm <j)

crystal. To m axim ise the flashlamp light absorption and the energy transfer, but to m inim ise the reabsorption by the holmium ion, the concentrations o f the dopants were selected to be 2.5% chromium, 5.6% thulium and 0.36% holmium.

Rosenbaum et al^"* follow ed up the work o f Storm et al by investigating the role o f chrom ium in the sen sitisation p rocess and the e ffe c t o f the host m aterial. Compared to YSAG and YSGG host materials, the YAG host was found to provide superior cross transfer o f the excitation energy to the thulium levels thereby leading to improved laser performance. By reducing the chromium concentration to only 0.8% , slope efficiencies up to 3.8% were observed for free-running mode operation with a sim ilar threshold to that reported previously by Storm et al. The authors cite the reduction in thermal effects with a reduction in the chrom ium concentration. From the considerations o f Duczynski et al^*, though, the effects were probably also due to better cross-relaxation betw een the chrom ium and thulium ion s due to a reduction in the lattice strain which is caused by the substitution o f chromium ions into the aluminium sites, the former having a significantly larger ionic radius than the alu m in iu m w h ich it r ep la ce s. T h e p u rsu it o f h o sts in w h ich la rg e am ou n ts o f chromium can be incorporated without suffering lattice distortion effects has been reported^* a lth ou gh YAG rem ains the m ost co m m o n h o st due to its b etter thermo-mechanical properties.

Further im provem ents in the laser perform ance w ere reported soon after the work o f Rosenbaum et al by the same team o f researchers (and the sam e labs' in which Storm's work was carried out). Slope efficiencies o f 5.1% and laser threshold energies as low as 28J were reported by Quarles et The dopant concentrations in this system were identical to those used by Rosenbaum et al with the improvements arising from the ch oice o f a focusing geom etry pumping chamber in contrast to the diffuse nature o f the pumping used by Rosenbaum et al.

circumstances, approached operation at repetition rates above IHz was limited. M oulton et al com m ented on the reason for this in their posted paper at CLEO in 1989, measuring induced thermal lensing significantly stronger than that observed in the Nd:YAG crystal. However the dependence o f the population o f the terminal laser lev el has led som e researchers^^*^^ to suggest that heat accumulation is principally resp o n sib le for reductions in output energy observed at higher pow er lo a d in gs. R ecen tly , the le n sin g p aram eter has been m easured fo r a CTH: YAG rod^^ at betw een 4 Dioptres kW * and 7 .2 Dioptres kW* which supports the observations made by Moulton et al o f thermally induced lensing much greater than that reported for neodymium doped YAG. In the first, and long overdue, measurements o f the th erm o-m echanical properties o f the CTH: YAG crystal, Marion^^ m easured the thermal conductivity o f the CTH: YAG crystal at 0.0664 W cm'* K'*, approximately half o f the thermal conductivity o f Nd:YAG. This explains, in part at least, why a higher degree o f thermal lensing is observed for the CTH: YAG crystal compared to the Nd:YAG crystal.

The restrictions which strong thermal lensing place on laser resonator design can be ov erco m e to a llo w high repetition rate operation o f F ixed -Q CTH: YAG lasers. B ecker et al^^ have dem onstrated operation at repetition rates up to 2 IH z w ith o u t s ig n ific a n t r ed u c tio n s in eith er the la ser th resh old en erg y or the slo p e efficiency using small diameter rods. This is in apparent contradiction to the theory o f thermal lensing which predicts shorter induced focal lengths in smaller diameter rods. Thus, resonators containing smaller diameter rods would be expected to be more sensitive to thermal lensing than resonators containing larger diameter rods. However, Becker's results show that high repetition rate operation in small diameter rods can be achieved, due probably to a combination o f improved cooling conditions around the smaller rods and the lower fraction o f total flashlamp light being absorbed by the laser crystal. T he significant results o f Becker et al are only marred by their suggestion that it is necessary to obtain a critical, elevated temperature within the laser rod before efficient operation can be achieved. The effect on which this claim is b a se d , n a m ely the m axim um output en ergy b ein g ob ta in ed at 9 H z rather than repetition rates at either lower or higher rates, is almost certainly due to optimisation o f the resonator p erform ance as thermal lensing a llo w s an increasing number o f transverse modes to oscillate in the resonator, thereby extracting the energy stored in the rod more efficiently. Beyond the 9H z operating lev el, the resonator is sim ply beginning to exhibit 'rollover', a familiar phenomenon in continuous wave Nd:YAG laser designs^^ where strong lensing takes the resonator out o f a stable condition and the number o f transverse modes drops again, thereby reducing the efficien cy o f the energy extraction.

Recently, Hamlin et al have presented results^^ which im prove on both the efficien cy and repetition rate performance previously reported by Quarles et al and Becker et al respectively. Returning to a diffuse pumping chamber, but o f a different d esign to that used by R osenbaum , slo p e e ffic ie n c ie s o f 5.5% are reported for a 6 " x 5 m m 0 rod d o p ed to th e n ow stand ard d o p a n t c o n c e n tr a tio n s d e s c r ib e d e a r l i e r ^ U s i n g filter glasses in the pumping chamber to filter flashlam p light that would otherwise only contribute to heating the laser crystal, repetition rates up to 25H z are reported, again with a small diameter rod, with only a 12% reduction in the output energy compared to the IHz laser performance.

Despite reports o f laser action at 2.1/zin from the holmium ion as long ago as 1962, it is only sin ce 1985 that progress has been m ade towards e ffic ie n t room temperature operation. Several research groups have reported on the performance o f laser crystals based around the holmium ion in YAG, sensitised with chromium and thulium. However, the data gained from these systems is, in some cases, incomplete and in others needs corroboration.

2 .1 .2 Spectroscopy and Pum p Schem e

The addition o f chromium and thulium to crystals containing holmium enhances the absorption bands in the visible region o f the optical spectrum. Figure 2.1 shows the key le v els in the chrom ium , thulium and holm ium w hich play a p a rt in the laser process. o LU CL LASER 0 -L Tnr Tm' Tm

Most o f the flashlamp light is absorbed by the broad and transitions in chromium which overlap well with the discharge spectrum obtained from pulsed xenon flashlamps, Figure 2 .2 . Non-radiative relaxation from these levels^^ to the chrom ium le v el is fo llo w ed by a m ixture o f radiative and non-radiative energy transfer^^ to thulium, via the level, to excite the ^F^ level. Although a mixture o f transfer processes, Armagan et al have shown that the radiative parts o f the processes account for on ly 3% o f the total transfer and con seq u en tly the sequence is often described as being non-radiative^'^. A ccurate data on the transition tim e betw een donor and acceptor levels in Cr and Tm is not available for a YAG host. However, A lp a t'e v et al^^ have reported that, for a GSAG h o st, 90% o f the ch rom iu m excitation is transferred in approximately 10/xs. It might be assumed that the transfer time in a YAG crystal would be either similar or less due to its stronger crystal field. H ow ever, Bowm an et aP* are on ly ab le, from observation o f the gain recovery following Q-Switching, to predict a transfer time less than lOOfis.

3 . 0 2 . 5 2 . 0 Ho 1 . 5 1 . 0 Ho 0 . 5 0 . 0 8 0 0 1000 1200 MOO 1500 1800 2 0 0 0 400 6 0 0 W a v e l e n g t h ( n m )

F igu re 2 .2 Cr:Tm:Ho:YAG absorption spectrum and em ission envelope o f xenon flashlamps (after Teichmann et aP")

The ^F^ level then cross-relaxes with the ^H^ level to produce two excited ^H^ states. Again there is no data on the characteristic time for this splitting to take place in YAG and it is n ecessary to assum e a sim ilar value to the 20fxs reported for a YSAG host by Alpat'ev^®. A final cross-relaxation process then takes place between the Tm ^H^ level and the level in holmium which is the upper laser level for the 2 .1 /.tm la ser tra n sitio n . U sin g an A le x a n d rite laser, o p era tin g at 0 .7 8 5 /x m , to selectively excite the ^F^ thulium level, Kintz et al have shown~^ that the quantum yield in the upper laser level o f holm ium , as a result o f the tw o-for-one process in thulium, is as high as 1.76. Thus, thulium plays a vital role, not only in transferring

the energy absorbed by the chromium to the holmium ions, but also in enhancing the efficiency. In so doing, this process also reduces the amount o f heat generation in the crystal.

There has been much discussion on the degree o f coupling between the Tm and the H o ^ L levels^^'^^. Fan et af'^ have shown that the Tm and Ho ^L

4 7 4 7

states are therm ally cou p led , i.e . a Boltzm ann statistical an alysis is adequate to d escrib e the relative p opulations. Fan et al^"^ and B ow m an et al^' are in general agreem ent that the therm alisation tim e is approxim ately 20/iS although N o g in o v c la im s c h a r a c te r is tic tim e s o f 2 0 0 ^ s and 667/xs fo r Tm(^H^)->Ho(^I^) and Ho(^I.^)-^Tm(^H^) respectively^^. A practical result o f this is that the recovery time o f the le v e l w ill lim it the number o f Q -Sw itch p u lses w hich can be extracted, at reasonable power lev els, from a 300/^s long Fixed-Q pulse, to approximately three (assuming lOOfis are required to fully recover the gain from reference^*).

T he is split into 15 Stark com ponents although only a few o f the low er sub-levels contribute to the laser process^ \ Bowman et al have observed laser action at nine distinct w avelengths between 2.080/>tm and 2 . 128jnm and have additionally identified the energy o f the sublevels o f the manifolds between which the laser action occurs. Their data is reproduced in Table 2 .1.

Wavelength (fjLm) Upper State fcm'h Terminal State ('em'*! 2.1275 5221 522 2.1207 5221 503 2 .1107 5242 503