A lthough there is now considerahle choice in selecting nonlinear m aterials a,s a result of recent developm ents in the growing and availability of new crystals, striking a balance between the physical lim itation of a crystal and the dem ands of a particular ajiplication is, however, still a very im portant task.
Basically a m aterial m ust satisfy four criteria, if it is to be useful for frequency conversion, and these are adequate nonlinearity, wide optical transparency, significant birefringence for phase-m atching, and a sufficiently high dama,ge threshold.
The nonlinear m aterials th a t can m eet the above requirem ents and hence are currently widely used for OPOs and SFG are the n io b a te s , such as LiNbOa, LÜO3, and KNbOa; b o ra te s , such as LiBaOs and BaBgO^ (BBO); and the K T P is o m o rp h s , such as KTiOPO^ (KTP), K Ti0A s04 (KTA), CsTiAsO^ (OTA), and RbTiOAsO.! (RTA).
K T P and LBO are the two crystals which have been used in this research. Before sum m arizing the properties of the two crystals, a brief review on other crystals (as m entioned above) ma,y be very useful for comparison.
3.1 A review on som e useful crystals
3.1 .1 L ith iu m n io b a te (L iN b O a ) a n d L ith iu m io d a t e (L Ü O3)
LiNbOa is a unia,xial ferroelectric crystal of point group 3m. The m ain advantages of this crystal are its large nonlinearity (5.44 pm /V ) and the capability of non-critical phase-
C H APTE R 3. NONLINEAR M ATERIAL G9
m atching. A later discovery was th a t the addition of MgO to LiNbO.j greatly reduces the photo-refractive induced dam age and leads to uniform high-cpiaJity large-sized single crystals being available. At th e standard doping level of 5% MgO, the threshold of dam age has been reported to increase by two orders of m agnitude [1]. D am age thresholds of 610 and 340 M W cm “ ^ were reported for 1.064 and 0.532 //m radiation respectively for doped LiNbOa[2, 3]. However, a dam age threshold of approxim ately 30 to 50 MWcm~^ was also reported due to a tw o-photon absorption process. The absorption in lithium niobate is low in the region of 1fim (0.001/cm ), while increasing significantly by the tim e the wavelength falls to 0.532 fim (0,02/cm ). Type I phase-m atching with the scheme of wg = -fwg is the only effective geom etry in LiN bO a. In this geom etry, the effective nonlinear coeflicient[4] is:
dpj j — d i5 sin Q — r/22 cos B sin 3(f) (3.1)
We m ust choose (f) and 0 correctly so th a t the <-/i5 and f/22 add and not subtract for 0 ^ 90^'. As shown in Fig. 3.1, the calculated m axim um y of LiNbOa from the above form ula and using the d a ta presented in [4], is 6.09 pm /V , corresponding to a,n angle (9 of 63", and obviously to m aximise y requires th a t <f) — 30".
> 6
I
%
% o Q 0= 30" // /
/
/ / = —5. p: ï i / V d.22—2.76 p m /V. 0 10 20 30 40 50 60 70 80 90 P h ase m a tc h in g a n g le 6 (degree)Figure 3.1: The effective nonlinear coefficient of LiN bO a.
An im portant application of LiNbOa is in tem ])erature-tuned type I non-critically phase- m atched OPOs with a puinj) wavelength around 0.5 //.m, or in reverse, the non-critically
CH APTER 3. NONLINEAR MATERIAL 70
phase-m at died frequency-doubling of i fim radiation. Fig. 3.2 shows the tuning curve versus operating tem perature for a range of pump wavelengths in the range 0.5-0.0 ///m. Using the d a ta given in Ref.[4], the calculated d ^ff of this OPO geom etry is 5.44 p m /V . The range of pum p wavelengths is lim ited by the m axim um safe operating tem perature, and the idler wave can be tuned to reach only about 2 //.m, which is much less th an the transparency range of the m aterial (5 //-m).
3.0-1 T ype I NCPM LiNbO, a. X,=500 n m . 2.5- b. X,=600 n m . 3.2.0- M 1.5- > 1.0- 0.5- 0.0 0 50 100 150 Phase m a tc h in g tem p e ra tu r e Tp„ (”C)200 250 300
Figure 3.2: Tuning curves of LiNbO.3 OPO in type I NCPM geom etry for pum p wave lengths range 0.5-0.6 ftm.
If radiation of wavelength 1 fim is used to pum p a LiNbOa O PO , as shown in Fig. 3.3 (I), the tuning can be extended to 5 //.m. In this geometry, the calculated r/^// is still around 5 p m /V and for a 10 mm long crystal, the acceptance angle is 1.4 m illi-radian, and the walk-off angle is 2^\ An interesting question is w hether the OPO perform ance can be im proved by tem perature tuning? The answer is yes. Fig. 3.3 (II) illustrates the possibility of tem perature tuning the type I critically phase-m atched OPO pum ped at 1 fw i. This geom etry, we believe, has the following advantages: (1) the ’’Fresnel” loss due to angular tuning is avoided; (2) the tuning can cover the whole range from 1.35-5 pm , and there is more flexibility in wavelength range selecting from the com bination of both angle and tem perature tuning; (3) the phase-m atchable angle can be increased by increasing the tem perature, and this will lead to slightly large d^jf (see Fig. 3.1) and acceptance angle, and smaller walk-off angle.
CH APTE R 3. NONLINEAR MATERIAL 71 (I) 5.0-1 4.5- 4.0- 3.5- « 2.5- ^ 2.0- 1.5- 1.0 45 4 6 47 48 4 9 50
Phase m a tc h in g angle 6 (degree)