DOI: 10.1007/s003400200814 Appl. Phys. B 74, 415–418 (2002)
Lasers and Optics
Applied Physics B
y.-f. chen1,✉
y.p. lan2
Comparison between c-cut and a-cut Nd:YVO
4
lasers passively Q-switched with a Cr
4
+
:YAG
saturable absorber
1Department of Electrophysics, National Chiao Tung University, 1001 TA Hsueh Road, Hsinchu,
Taiwan, 30050
2Institute of Electro-Optical Engineering, National Chiao Tung University, 1001 TA Hsueh Road, Hsinchu,
Taiwan, 30050
Received: 10 December 2001/Revisedversion: 22 January 2002 Published online: 14 March 2002 • © Springer-Verlag 2002
ABSTRACTComparison between c-cut and a-cut Nd:YVO4 mi-crochip lasers passively Q-switched with a Cr4+:YAG saturable absorber is experimentally made. The lower emission cross sec-tion of the c-cut Nd:YVO4 crystal can enhance the passive Q-switching effect to produce a peak power 10 times higher than that obtained with the a-cut crystal. The experimental result fur-ther reveals that a c-cut Nd:YVO4crystal is a very convenient material for short-pulse (sub-nanosecond) and high-peak-power (> 10 kW) lasers.
PACS42.60.B; 42.65; 42.70
1 Introduction
Compact, all-solid-state pulse lasers are of poten-tial interest for numerous applications such as ranging, remote sensing, and microsurgery [1]. Diode-pumped Q-switched microchip lasers are compact efficient solid-state lasers with a diffraction-limited output beam. Passive techniques that use saturable absorbers have the advantages of potentially lower cost and simplicity in fabrication and operation since they require no high-voltage or RF drivers. In recent years, Cr4+:YAG crystals have been successfully used as passive Q-switches for a variety of gain media such as Nd:YAG [2], Nd:YVO4[3], and Nd:GdVO4crystals [4]. So far, only
pas-sively Q-switched Nd:YAG/Cr4+:YAG lasers can generate extremely short (< 1 ns) high-peak-power (> 10 kW) pulses. The relatively narrow absorption band of a Nd:YAG crys-tal, however, sets stringent requirements on the spectrum of the pump diodes. Although microchip lasers passively Q-switched with a semiconductor saturable absorber mirror (SESAM) typically generate much shorter pulses compared to those using bulk-crystal absorbers, the pulse energy of Nd-doped microchip lasers was typically around 100 nJ [5].
YVO4and GdVO4 crystals belong to the group of oxide
compounds crystallizing in a zircon structure with a tetrag-onal space group. The four-fold-symmetry axis is the crys-tallographic c axis. Perpendicular to this axis are the two indistinguishable a and b axes. The uniaxial Nd:YVO4 and
✉ Fax: +1-886/35-729134, E-mail: [email protected]
GdVO4 crystals show strong polarization-dependent
fluo-rescence emission due to the anisotropic crystal field. In a Nd:YVO4 crystal, for example, the stimulated emission
cross section parallel to the c axis,σ= 25 × 10−19cm2, is
four times higher than that orthogonal to the c axis,σ⊥= 6.5 × 10−19cm2, for the emission wavelength at 1064 nm [6].
A larger stimulated emission cross section normally results in a lower pumping threshold for cw or actively Q-switched laser operation. Therefore, the conventional Nd:YVO4or GdVO4
crystals are cut along the a axis, i.e. so-called a-cut, to use the stimulated emission cross section ofσto dominate the laser oscillation.
On the contrary, the large stimulated emission cross sec-tion may be disadvantageous for a passively Q-switched laser. For good passive Q-switching, the saturation in the absorber must occur before the gain saturation in the laser crystal (the second threshold condition) [7]. From the analysis of the coupled rate equations, the criterion for good passive Q-switching is given by [8]: ln 1 T02 ln 1 T02 + ln1 R + L σgs σ A AS γ 1− β (1)
where T0is the initial transmission of the saturable absorber,
A/Asis the ratio of the effective area in the gain medium to
that in the saturable absorber, R is the reflectivity of the output mirror, L is the nonsaturable intracavity round-trip dissipative optical loss,σgsis the ground-state absorption cross section of
the saturable absorber,σ is the stimulated emission cross sec-tion of the gain medium,γ is the inversion reduction factor with a value between 0 and 2 as discussed in [9], andβ is the ratio of the excited-state absorption cross section to that of the ground-state absorption in the saturable absorber. Since theσ
value of the Nd:YVO4crystal is comparable to theσgsvalue
of the Cr4+:YAG crystal (∼ (20 ± 5) × 10−18cm2[7]), using a Cr4+:YAG crystal as a saturable absorber in a-cut Nd:YVO4
lasers generally produces a longer pulse width and a lower peak power, usually less than 1 kW [3, 4].
When a Nd:YVO4 crystal is cut along the c axis, i.e.
c-cut, the effective stimulated emission cross section is dom-inated byσ⊥ instead of σ. Sinceσ⊥ is four times smaller thanσ, the c-cut Nd:YVO4 crystal may be more
appropri-416 Applied Physics B – Lasers and Optics
ate than the a-cut Nd:YVO4crystal with a Cr4+:YAG crystal
as a saturable absorber, as indicated in the criterion of (1). In this work, we make an experimental comparison between
c-cut and a-cut Nd:YVO4 lasers passively Q-switched with
a Cr4+:YAG saturable absorber. The results reveal that a c-cut Nd:YVO4 crystal is a very convenient material for
short-pulse (< 2.5 ns) and high-peak-power (> 5 kW) lasers. Using a Cr4+:YAG crystal with an initial transmission of 60%, more than 21-kW peak output power was obtained at the incident pump power of 2.4 W. The corresponding pulse width and the repetition rate were 0.85 ns and 13.5 kHz. Although the ab-sorption bandwidth of the c-cut Nd:YVO4crystal is slightly
narrower than that of the a-cut one with the same doping con-centration, it is still three times wider than that of the standard Nd:YAG crystal and could be efficiently pumped by a laser diode without thermal regulation.
Figure 1a and b are schematics of the present resonators for cw and passively Q-switched lasers, respectively. The pump source is a 2.5-W fiber-coupled diode-laser array (Co-herent) with a core size of 200µm and a numerical aperture of 0.18. The output wavelength of the diode laser ranges from 807 to 810 nm at 25◦C. A focusing lens with 20-mm focal length is used to re-image the pump beam onto the laser crys-tal. The waist diameter of the pump beam was about 100µm. The laser crystal is 5-mm long and doped with 1.0% Nd3+ concentration. One side of the Nd:YVO4crystal was coated so
as to be nominally highly reflecting at 1064 nm (R> 99.8%) and anti-reflection-coated at 808 nm (T> 90%). The other side was anti-reflection-coated at 1064 nm (R< 0.2%). The output coupler mirror used in the cw laser was a 1-m radius-of-curvature concave mirror with 85% reflectance at 1064 nm. In the passively Q-switched laser experiment, one side of the Cr4+:YAG crystal was partial-reflection-coated at 1064 nm (R= 75%) so as to be an output coupler. The other side of the Cr4+:YAG crystal was anti-reflection-coated at 1064 nm (R< 0.2%). Several Cr4+:YAG crystals with different initial trans-missions (T0= 0.80, T0= 0.70, and T0= 0.60) were used in
the experiment. The final transmissions for Cr4+:YAG crys-tals can be approximated as(T0)β, where the parameterβ was
measured to be around 0.23. The nonsaturable losses can be
laser diode
laser diode a
b
FIGURE 1 a Schematics for cw Nd:YVO4laser. b Schematics for pas-sively Q-switched Nd:YVO4/Cr4+:YAG lasers
estimated to be 1− (T0)β. The total cavity length was
approxi-mately 6–7 mm. The pulse temporal behavior was recorded by a LeCroy 9362 oscilloscope (500-MHz bandwidth) and a fast Si PIN photodiode with a rise time of∼ 0.35 ns. For all results shown the output was operating in a TEM00mode.
Figure 2 shows the cw output power of a-cut and c-cut Nd:YVO4 lasers as a function of incident pump power. As
is seen, thresholds of 0.28 W and 0.95 W and slope effi-ciencies of 63% and 73% were obtained for the a-cut and
c-cut Nd:YVO4crystals, respectively. A higher threshold for
the c-cut Nd:YVO4crystal was expected because of a lower
stimulated emission cross section, compared with the a-cut crystal. Nevertheless, a higher slope efficiency for the c-cut Nd:YVO4crystal indicates that its intrinsic loss is distinctly
smaller than that of the a-cut crystal. For the measurements at different output couplings the intrinsic losses were estimated to be 2.8% and 0.5% for the a-cut and c-cut Nd:YVO4
crys-tals, respectively.
Figure 3 shows the average output power of a-cut and c-cut Nd:YVO4passively Q-switched lasers as a function of
inci-dent pump power with a saturable absorber of T0= 0.80. It
can be seen that the threshold pump powers for passive Q-switched behavior are 1.0 and 1.6 W for the a-cut and c-cut Nd:YVO4 crystals, respectively. Although the threshold of the c-cut crystal is higher, the corresponding slope efficiency of 53.3% is significantly higher than 23.3% obtained in the
a-cut Nd:YVO4 passively Q-switched laser. Figure 4 shows
the pulse-repetition rate versus the incident pump power for both lasers. It is seen that increasing the pump power increases the pulse-repetition rate in the c-cut Nd:YVO4/Cr4+:YAG
passively Q-switched laser, whereas the repetition rate in the a-cut case is nearly independent of the pump power. Figures 5 and 6 illustrate the pulse energy and the pulse width, respectively, versus the incident pump power for both
,QFLGHQW SXPS SRZHU : 0.0 0.5 1.0 1.5 2.0 2.5 3.0 &: RX WSX WSRZH U : 0.0 0.3 0.6 0.9 1.2 1.5 DFXW FFXW
FIGURE 2 The cw output power of a-cut and c-cut Nd:YVO4lasers as a function of incident pump power
CHENet al. Passively Q-switched laser 417 ,QFLGHQW SXPS SRZHU : 0.0 0.5 1.0 1.5 2.0 2.5 3.0 $ YH UDJH RX WS XW SRZ HU : 0.0 0.1 0.2 0.3 0.4 0.5 0.6 DFXW FFXW
FIGURE 3 The average output power of a-cut and c-cut Nd:YVO4 pas-sively Q-switched lasers as a function of incident pump power with a sat-urable absorber of T0= 0.80 ,QFLGHQW SXPS SRZHU : 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3X OV H UHSHWL WLRQ N+] 0 20 40 60 80 100 120 DFXW FFXW
FIGURE 4 The pulse-repetition rate versus the incident pump power for a-cut and c-a-cut Nd:YVO4passively Q-switched lasers
lasers. Like typical systems, such as a Nd:YAG/Cr4+:YAG passively Q-switched laser, both the pulse energy and the pulse width are insensitive to the pump power in the c-cut Nd:YVO4/Cr4+:YAG passively Q-switched laser. In contrast
to this insensitivity, increasing the pump power substantially increases the pulse energy and decreases the pulse width in the
a-cut Nd:YVO4/Cr4+:YAG passively Q-switched laser. This
atypical behavior has been discussed in [10] and is attributed to the effect of the pumping rate. As estimated from Figs. 5 and 6, at a pump power of 2.4 W the peak powers are 0.57 and 5.8 kW for the a-cut and c-cut Nd:YVO4cases, respectively.
This result confirms that the c-cut Nd:YVO4 crystal is more
,QFLGHQW SXPS SRZHU : 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3X OV H HQHUJ \ µ - 0 3 6 9 12 15 18 DFXW FFXW
FIGURE 5 The pulse energy versus the incident pump power for a-cut and c-cut Nd:YVO4passively Q-switched lasers
appropriate than the a-cut crystal for the passively Q-switched microchip laser with a Cr4+:YAG saturable absorber.
In the c-cut Nd:YVO4/Cr4+:YAG passively Q-switched
laser, a decrease in the initial transmission of the absorber re-sulted in an increase of the pulse energy and a decrease of the pulse width, thus leading to a higher peak power. When the Cr4+:YAG crystal with an initial transmission of 60% was used, the pulse energies reach a level of 18µJ with a pulse width of 0.85 ns at a pump power of 2.4 W. The
correspond-,QFLGHQW SXPS SRZHU : 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3X OV H Z LGWK Q V 0 5 10 15 20 25 30 35 DFXW FFXW
FIGURE 6 The pulse width versus the incident pump power for a-cut and c-cut Nd:YVO4passively Q-switched lasers
418 Applied Physics B – Lasers and Optics
QVGLY µVGLY
D
E
FIGURE 7 Oscilloscope traces of a train of output pulses; the lower trace is an expanded shape of a single pulse, showing a 0.85-ns width (FWHM)
ing peak power exceeds 21 kW at a pulse-repetition rate of 13.5 kHz and the conversion efficiency of the pump power into the average output power was about 10.1%. The inter-pulse time jittering is generally less than ±4%. A typical oscilloscope trace is presented in Fig. 7. The pulse-to-pulse amplitude fluctuation of the Q-switched pulse train was found to be less than ±5%. The output-beam quality was nearly the same for both laser cavities and the M2 parameter has
been estimated to be< 1.2 over the complete output power range with the algorithms of the knife-edge technique. Pre-viously, Zayhowski and Dill [2] demonstrated a passively Q-switched Nd:YAG/Cr4+:YAG microchip laser in which 11-µJ pulses of 0.337-ns duration at a pulse-repetition rate of 6 kHz were produced at a pump power of 1.2 W and the corresponding conversion efficiency was 5.5%. Compared
with the Nd:YAG/Cr4+:YAG microchip laser, the present results have a comparable peak output power and a higher optical conversion efficiency. This is mainly due to the fact that the c-cut 1% doped Nd:YVO4crystal has lower intrinsic
losses [6] and its effective stimulated emission cross section is close to that of the Nd:YAG crystal in magnitude. Referring to the pulse duration, a two-fold pulse shortening should be feas-ible in our set-up using a cavity with a shorter c-cut 2% doped Nd:YVO4crystal.
In summary, the studies of the present performance indi-cate that a c-cut Nd:YVO4 crystal is a very convenient
ma-terial for short-pulse (sub-nanosecond) and high-peak-power (> 10 kW) lasers. It is demonstrated that 18-µJ pulses of 0.85-ns duration at a pulse-repetition rate of 13.5 kHz can be generated at a pump power of 2.4 W. The main advantages of the c-cut Nd:YVO4crystal are the combination of its high
ab-sorption cross section, wide abab-sorption bandwidth, and low intrinsic losses. In addition to the c-cut Nd:YVO4crystal, we
believe that the results presented here can be usefully applied to the c-cut Nd:GdVO4crystal for the design of reliable and
simple sub-nanosecond (and even< 500 ps) lasers for various applications.
ACKNOWLEDGEMENTS The authors thank the National Sci-ence Council of the Republic of China for financially supporting this research under Contract No. NSC-90-2112-M-009-034.
REFERENCES
1 J.A. Morris, C.R. Pollock: Opt. Lett. 15, 440 (1990) 2 J.J. Zayhowski, C. Dill, III: Opt. Lett. 19, 1427 (1994)
3 Y. Bai, N. Wu, J. Zhang, J. Li, S. Li, J. Xu, P. Deng: Appl. Opt. 36, 2468 (1997)
4 C. Li, J. Song, D. Shen, N.S. Kim, J. Lu, K. Ueda: Appl. Phys. B 70, 471 (2000)
5 B. Braun, F.X. Kärtner, U. Keller, J.P. Meyn, G. Huber: Opt. Lett. 21, 405 (1996)
6 H. Nagamoto, M. Nakatsuka, K. Naito, M. Yamanaka, K. Yoshida, T. Sasaki, T. Kanabe, A. Nakai, S. Saito, Y. Kuwano: Laser Res. 18, 87 (1990) [in Japanese]
7 A.E. Siegman: Lasers (University Science, Mill Valley, CA 1986) p. 1024
8 G. Xiao, M. Bass: IEEE J. Quantum Electron. QE-33, 41 (1997) 9 J.J. Degnan, D.B. Coyle, R.B. Kay: IEEE J. Quantum Electron. QE-34,
887 (1998)
10 M.A. Jaspan, D. Welford, G. Xiao, M. Bass: Conf. Dig. CLEO-2000, San Francisco, 7–12 May 2000, p. 454
