We described the principle and the fabrication of a Nd:YVO4/KTP microchip for the linearly-polarized green laser and verified its availability by manufacturing and characterizing the green laser using the microchip. Under the driving condition having the modulation frequency of 60 Hz and the duty ratio of 25%, the laser showed the stable linear polarization, the maximum average power of 37 mW, yielding the high electrical-to-optical efficiency of 10.9%.
©2009 Optical Society of America
The microchip laser was first demonstrated in 1989 . It has a small solid segment cavity, which results in the intrinsic small size and robustness to the external mechanical shock. Since diode lasers are highly efficient, they are used as pump sources.
A microchip cavity may be made of a single gain medium in combination with other mediums. For obtaining efficient frequency conversion, a nonlinear optical material should be added to the microchip cavity. CW green laser which is essential in the laser display can be developed in this way. In laser display, three red, green, and blue lasers are essentially needed. Red and blue lasers are being commercially produced in the form of laser diode, while green one is not. Nd:YVO4/KTP composite microchip laser is a promising alternative to the green laser diode [2–5]. The laser usually needs the temperature-controlling unit for the high-power and stable operation. To decrease the total volume of the laser and the temperature-controlling unit, we have developed the ones which have the built-in tiny thermoelectric cooler (TEC) [4,5]. The first kind of one with just a TEC  has the advantage of simplicity, while the second kind of one with two TECs  has the enhanced characteristics because the pump laser diode and microchip can be all optimized owing to two independent TECs. On the other hand, microchip green lasers using common optically-contacted Nd:YVO4/KTP microchip cavities, usually produce beams with poor polarization . Their polarization ratios can be very low and directions of the major polarization axes are unpredictable. However, well-polarized green lasers are essential in many applications including the laser display system which has the polarization-dependent components, such as polarization beam combiner or beam scanner .
In this paper, we describe the reason that the common optically-contacted Nd:YVO4/KTP composite microchip is difficult to generate the polarized green light in detail and propose a kind of Nd:YVO4/KTP microchip for the polarized green light. And then we present the fabrication of the new Nd:YVO4/KTP microchip and the development of the stably-polarized green laser manufactured using it.
2. Nd:YVO4/KTP composite microchip for the polarized green laser
The common microchip for the high-power intracavity frequency doubling is made generally by optically-contacting the gain medium and the frequency-doubling medium, with their polished and uncoated surfaces faced together, as shown in Fig. 1 . Silicon pads are used for firmly fixing them. In the case for getting the green beam, Nd:YVO4 and KTP crystals have been widely used as the gain and frequency-doubling medium, respectively. Since KTP has the interaction of type II, principal axes of two crystals, Nd:YVO4 and KTP, are 45°-tilted to each other for the efficient frequency doubling. The harmonic beam generated from the forward-propagating oscillating fundamental beam keeps the linearly-polarized state because it passes out the output coupler directly. But the one from the backward-propagating oscillating one gets elliptically polarized because it makes a round trip in Nd:YVO4 before going out the output coupler. Since the interface between two mediums has the very low reflectance to the harmonic beam, it makes a round trip in Nd:YVO4 and propagates KTP again before going out the output coupler, resulting in the elliptical polarization because two crystals have the principal axes 45°-tilted to each other. As a result, the whole harmonic beam has the polarization which is not linearly polarized. Since Nd:YVO4 has the high birefringence (ne-no) of 0.233 at 532 nm , its half-wave-plate length which is determined with an expression of the wavelength(λ) and birefringence, λ/2(ne-no), is no more than 1.14 μm. Therefore, it is very difficult to adjust Nd:YVO4 length to the exact one.
To obtain the whole harmonic beam being linear polarized, the backward-generated one should not make a round trip in Nd:YVO4. To realize this requisite, we devised the microchip having the structure shown in Fig. 2 (a) . If the inner surface of KTP crystal is HR-coated to the harmonic beam and AR-coated to the oscillating fundamental one and the inner surface of Nd:YVO4 is AR-coated to the oscillating one, the backward-generated harmonic beam can go out the output coupler without any trip in Nd:YVO4 medium. This beam is combined to the forward-generated one having the same polarization to form the whole beam being totally linearly-polarized. The power of the whole beam depends upon the degree of the interference between two coupled beams and can be optimized through the constructive interference, which can be achieved by adjusting the temperature of the microchip. Since the constructive interference gives rise to the higher power than the pure mixture without interference, this type of microchip in which two beams can be fully constructively-interfered owing to the same polarizations can produce the higher-power beam than the common optically-contacted type of one in which two beams cannot be fully constructively-interfered in general. This power enhancement is more promoted because the backward-generated beam doesn’t have the loss due to the absorptive Nd:YVO4. To ensure that the two surfaces were parallel, a spacer needed to be used. Keeping them separated has additional advantages. Since the possibility of defects on surfaces which can be formed during the contact can be inherently removed, the damage threshold of the microchip can be increased and the oscillation threshold of the laser using the microchip can be decreased. The maximum power of the laser can also be increased.
According to the above method, we fabricated the microchip whose image is shown in Fig. 2 (b). The used Nd:YVO4 crystal (CASIX) has the Nd-doping concentration of 3 at %, surface area of 3.0*3.0 mm2, thickness of 0.5 mm, and parallelism of less than 20 arc second. One surface was coated to have the low reflectance of 5% at 808 nm and the high reflectance of more than 99.8% at 1064 nm and the other surface, to have the low reflectance of less than 0.1% at 1064 nm. The KTP crystal (CASIX) has the crystal angle cut for the second harmonic generation of 1064 nm, surface area of 3.0*3.0 mm2, thickness of 2.0 mm, and parallelism of less than 20 arc second. One surface was coated to have the low reflectance of less than 1% at 1064 nm and the high reflectance of more than 90% at 532 nm and the other surface, to have the high reflectance of more than 99.8% at 1064 nm and the low reflectance of less than 5% at 532 nm. Since the forward-propagating fundamental beam and the backward-propagating one propagate in the exactly-opposite direction, they have the same magnitude of effective nonlinear coefficient and generate the harmonic green beams having the same magnitude of power. Subsequently the backward-generated beam is reflected with the high reflectance of more than 90% in the inner surface of KTP. Therefore, two green beams can be combined with almost the same magnitude of power, although generally with the different phase. The spacer was made out of the crystalline quartz and has the thickness of about 0.25 mm, outer area of 3.0*3.0 mm2, U-shaped clear aperture, and parallelism of less than 1 arc second. Its aperture was made with the dicing saw and has the width of 1.3 mm and the height of about 1.8 mm. Two silicon pads were used for fixing three ones firmly.
3. Linearly-polarized microchip green laser
A linearly-polarized microchip green laser using the fabricated microchip was developed. The inner structure of the manufactured laser is shown in Fig. 3 . A pump laser diode and a fabricated microchip were mounted on the platform, which had two TECs to temperature-control the two key optical elements independently, with the gap of 100 μm. The pump laser diode was the small QA-type 808-nm one (Axcel Photonics) which emits the optical power of 500 mW at the driving current of about 540 mA. This small laser system was covered with a small cap having a window blocking IR beam for the efficient and precise temperature control.
For characterization, the green laser was operated in the modulation-mode condition of the frequency of 60 Hz and duty ratio of 25% which is common in the laser display. Peak value of the driving current was 600 mA, which was the possible maximum value of the used pump laser diode. Since Nd:YVO4 has the maximum absorption at the wavelength of 809 nm , the pump laser diode needed to be adjusted to the temperature at which it has the wavelength of this value for the highly-efficient lasing. We observed the output power of the green beam varying the temperature of the pump laser diode from 25°C to 60°C while keeping the one of the microchip constant at 25°C. The power optimum happened at 50°C. At this temperature, the pump laser diode had the voltage of 2.26 V and thus the electrically-consumed average power of 340 mW. Keeping the temperature of the pump laser diode at its optimum temperature, 50°C, we measured the output power of the green beam varying the temperature of the microchip from 25°C to 60°C and obtained the result shown in Fig. 4 .
Variation of the output power with the temperature is thought to be due to the birefringent filter effect  in the KTP crystal, the interference between the forward-generated harmonic beam and the backward-generated one, and the phase mismatch of the frequency doubling. Since KTP crystal has the principal axes 45°-tilted to ones of Nd:YVO4, it acts as a phase retarder to the oscillating 1064-nm beam. According to Sasaki’s study performed with 4.3-mm-thick KTP , 2.0-mm-thick KTP can be predicted to have the temperature period of 30°C. The back-generated harmonic beam interferes with the forward-generated one. Since the phase difference between them is a function of the optical length of KTP crystal, the degree of the interference changes with temperature of the microchip. Considering the thermo-optic property and the thermal expansion coefficient of KTP, the interference period can be predicted to be about 8°C. The temperature bandwidth of the second harmonic generation is about 120°C. It can be seen these predictions roughly agree with the experimental result. Local minimum doesn’t decrease to zero, which is maybe due to the limited destructive interference occurring because the laser is not the single-longitudinal one . Maximum average power of 37 mW was obtained. We can see the electrical-to-optical efficiency is 10.9%, recalling that the average electrical power of 340 mW was consumed in the pump laser diode. This value is a high one.
We observed the polarization of the laser as well as the output power and got the result shown in Fig. 5 . It could be seen that the output light was linearly polarized with the high polarization ratio and the constant polarization angle. The fluctuation in the polarization ratio is maybe due to not the intrinsic property of the laser but the limited measurement range of the power meter used in the experiment, which can be verified by the fact that at 50°C showing the low output power the polarization ratio was also low. The polarization angle of the green light was 45°, which was because the principal axes of KTP crystal are 45°-tilted to ones of Nd:YVO4 crystal. According to our experiments using several common microchips, the green beams generated from the common ones showed the polarization states which could have the low polarization ratio of about 2:1 in the bad case and sensitively changed with the microchip temperature, as theoretically predicted. Temperature change of no more than 6°C might cause the polarization ratio to change from 14:1 to 3.3:1 and the direction of the polarization major axis to change by 26°.
To measure the output-power stability, we operated the microchip green laser at the microchip temperature of 53°C and monitored the output power for 3 hours. We could see the laser was stable with the fluctuation of 0.14%, as shown in Fig. 6 .
We developed a kind of Nd:YVO4/KTP microchip for the polarized microchip green laser. By manufacturing and characterizing the microchip green laser using the microchip, it could be seen that our new Nd:YVO4/KTP microchip can generate the linearly-polarized green light stably regardless of the microchip temperature, unlike the common optically-contacted microchip generating the unstable light whose the polarization ratio and direction sensitively change with the temperature. Under the driving condition having the modulation frequency of 60 Hz, duty ratio of 25%, and average electrical pump laser diode power of 340 mW, the output green beam had the maximum average power of 37 mW, yielding the high electrical-to-optical efficiency of 10.9%.
We expect that the polarized microchip green laser using our Nd:YVO4/KTP microchip can be widely used in many applications including the portable or hand-held laser display system.
We would like to acknowledge Dr. Jeong Soo Kim of Phovel for packaging the green laser. This work was supported by Institute for Information Technology Advancement (IITA) in Ministry of Knowledge Economy (MKE) through Leading Edge R&D Program and by APRI-Research Program through Asian Laser Center Program provided by Gwangju Institute of Science and Technology (GIST).
References and links
1. J. J. Zayhowski and A. Mooradian, “Single-frequency microchip Nd lasers,” Opt. Lett. 14(1), 24–26 (1989), http://www.opticsinfobase.org/ol/abstract.cfm?URI=ol-14-1-24. [CrossRef] [PubMed]
2. N. MacKinnon and B. D. Sinclair, “A laser diode array pumped, NdYVO4/KTP, composite material microchip laser,” Opt. Commun. 105(3-4), 183–187 (1994). [CrossRef]
3. J. Watanabe and T. Harimoto, “Oscillating longitudinal-mode control of a microchip green laser by injection current,” Opt. Express 15(3), 965–970 (2007), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-15-3-965. [CrossRef] [PubMed]
4. C. Jung, B.-A. Yu, K. Lee, Y. L. Lee, N. E. Yu, D.-K. Ko, and J. Lee, “A compact diode-pumped microchip green light source with a built-in thermoelectric element,” Appl. Phys. Express, 2008, 1, 062005 (2008), http://apex.ipap.jp/link?APEX/1/062005/.
5. B.-A. Yu, C. Jung, I.-S. Kim, Y. L. Lee, N. E. Yu, and D.-K. Ko, “Temperature-controlled efficient microchip green laser for mobile projection displays,” Electron. Lett. 45(18), 943–945 (2009). [CrossRef]
6. E. Tamaki, Y. Hashimoto, and O. Leung, “Computer-to-plate printing using the Grating Light Valve device,” Proc. SPIE 5348, 89–97 (2004). [CrossRef]
7. CASIX, Inc., http://www.casix.com/product/prod_cry_ndyvo4.html.
8. Y. Mao, M. Huang, and C. Wang, “Temperature effect on emission lines and fluorescence lifetime of the 4F3/2 state of Nd:YVO4,” Chin. Opt. Lett. 2, 102–105 (2004), http://www.opticsinfobase.org/col/abstract.cfm?URI=col-2-2-102.
9. T. Sasaki, T. Kojima, A. Yokotani, O. Oguri, and S. Nakai, “Single-longitudinal-mode operation and second-harmonic generation of Nd:YVO(4) microchip lasers,” Opt. Lett. 16(21), 1665–1668 (1991), http://www.opticsinfobase.org/ol/abstract.cfm?URI=ol-16-21-1665. [CrossRef] [PubMed]