Growth of High Quality Al-Doped CsLiB6O10 Crystals Using Cs2O-Li2O-MoO3 Fluxes

High quality and large size Al-doped CsLiB6O10 (CLBO) single crystals have been successfully grown by top-seeded solution growth (TSSG) technique using Cs2O–Li2O–MoO3 fluxes. The advantages of this newly developed flux system were investigated by viscosity measurements and growth experiments. Al-doped CLBO presents a very high transmittance in the visible region and the weak absorption values at 1064 nm along a and c axes are only 140 and 50 ppm/cm, respectively. The measured LIDT of Al-doped CLBO at λ = 1064 nm and τ = 5.0 ns is 5.10 GW/cm2. Moreover, Al-doped CLBO exhibits an apparent enhancement of the hygroscopic nature in contrast with the undoped crystal as determined by the humidity experiments. Finally, a high fourth harmonic generation (FHG) conversion efficiency of 63% utilizing Al-doped CLBO has been achieved by a picosecond mode-locked Nd:YAG laser, the results also reveal that Al doping has no obvious impact on the FHG conversion efficiency.


Introduction
Nonlinear optical (NLO) crystal is one type of functional crystal that has been widely used in optical communication systems, industrial lasers, and microelectronic devices, owing to its various properties including electro-optical, nonlinear optical, photorefractive effects, etc. [1][2][3]. Recently, considerable research has been conducted on borate series nonlinear optical crystals for harmonic generation in high-power solid-state ultraviolet (UV) lasers due to their excellent properties such as relatively high tolerance to laser-induced damage, large nonlinear optical coefficients, moderate birefringence, and wide transparency range in the UV region. For instance, several borate crystals, such as β-BaB 2 O 4 (BBO) [4], LiB 3 O 5 (LBO) [5], CsB 3 O 5 (CBO) [6], CsLiB 6 O 10 (CLBO) [7,8], KBe 2 BO 3 F 2 (KBBF) [9], YCa 4 O(BO 3 ) 3 (YCOB) [10], and YA1 3 (BO) 4 (YAB) [11] have attracted a great deal of attention. Among these crystals, CLBO is a newly developed borate crystal and has been successfully applied for the fourth and fifth-harmonic generation of a Nd:YAG laser [12,13]. After firstly reported by researchers in 1995 [7,8,[14][15][16], CLBO is considered as a highly promising NLO crystal due to its small walk-off angle and relatively large phase-matching angular, spectral, and temperature acceptance bandwidths compared with other borate NLO crystals [12,17]. Additionally, in comparison with other commercially available NLO crystals such as BBO and LBO, CLBO possesses incomparable advantages, since it can be easily grown from congruent or near-congruent melts at extremely fast growth rate by top-seeded solution growth (TSSG) method [16,18].
Since CLBO melts congruently at 848 • C, many growth techniques could be utilized to grow CLBO single crystals. The TSSG method has been commonly used to grow CLBO crystals. Currently,

Crystal Growth
The starting material was synthesized from high purity reagents Cs 2 CO 3 , Li 2 CO 3 , H 3 BO 3 , and MoO 3 at an appropriate ratio with the addition of a certain amount of Al 2 O 3 . After being accurately weighed and mixed homogeneously, it was then melted in a Ø 90 × 80 mm 3 platinum crucible at 900 • C in several batches. The crucible was placed in a three-zone resistance furnace controlled by a programmable Eurotherm (model 818) temperature controller. Then, the mixture was heated to approximately 50 • C above the melting point, followed by stirring with a platinum stirrer for 48 h to enable the solution to melt completely and mix uniformly. The saturation temperature was precisely determined by seeding measurements method, and the suitable vertical temperature gradient was about 1 • C/6 cm. A high quality [001] oriented seed crystal was slowly introduced into the solution surface at a temperature of 1 • C higher than the saturation temperature and maintained at this temperature for half an hour. The growth rate was carried out at a very slow cooling rate in the range of 0.05-0.1 • C/day in the initial stage. The growing crystal was rotated at a rate of 20-50 rpm with the rotation direction inverted every 2 min. After the growth finished, the crystal was carefully drawn out of the solution surface and cooled to room temperature at a rate of 5-10 • C/h.

Characterization
The viscosity of growth solution was measured by a DVII + Pro viscometer (Brookfield Co., Middleboro, MA, USA). Powder X-ray diffraction (XRD) measurement was verified by a computer-automated Bruker D8 Focus diffractometer equipped with Cu Kα radiation (λ = 1.54056 Å). The as-grown Al-doped crystals were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) for the determination of Al 3+ concentrations. The transmission spectrum was recorded using a Perkin-Elmer Lambda 900 UV-Vis-near-infrared (NIR) spectrophotometer in the wavelength range of 185-1800 nm at room temperature. Optical weak absorption testing at 1064 nm was performed by the photo-thermal common path interferometer (PCI) method. The laser-induced damage threshold of the crystal was determined by a Continuum Surelite II pulsed Nd:YAG laser system. The surface etching morphology of CLBO crystal induced by humidity was observed on the (001) planes using an optical microscope (Nikon SMZ 745T, Nikon Co., Kanagawa, Japan). In the laser output experiment, a 532 nm Nd:YAG laser was utilized as the fundamental source, and then doubled to 266 nm for the measurement of conversion efficiency.

Crystal Growth and Viscosity Measurement
As previously reported [38], the molar ratio of Cs 2 O:Li 2 O:B 2 O 3 :MoO 3 for CLBO crystal growth was in the range of 1:(1.0-1.5):(3.0-6.0):(0.5-1.0). According to our growth results of various solution compositions, we determined that the appropriate molar ratio was 1:1:3:1. After the appropriate molar ratio of this flux system was identified, undoped and 2.5%-10% Al-doped CLBO were successfully grown with growth duration of 30-40 days. The as-grown undoped and Al-doped crystals were free from visible cracks and inclusions, as shown in Figure 1. The detailed growth conditions and results are summarized in Table 1. It is clear that the cooling range varies from 3.2 to 4.1 • C in this flux system, apparently wider than that of self-flux system (1-1.5 • C) [14,19], indicating that this flux system is conducive to the growth of large size crystals.
The volatility and the viscosity-two crucial parameters during crystal growth-both depend on solution temperature and composition. Low volatility and viscosity are beneficial to mass and heat transfer as well as to maintain the stability of solution components. Hence, a flux system with low volatility and viscosity is vitally important to grow high quality crystal.   Table 1 shows that the measured saturation temperature was in the range of 800-811 °C, which was remarkably lower than that of self-flux system (837-844 °C) [39]. Consequently, this flux system can reduce volatility so as to avoid component deviation caused by the relatively high evaporation of Cs. In the case of Al-doped CLBO grown from self-flux system, it is difficult to accurately determine saturation temperature due to the relatively great fluctuation of saturation temperature induced by volatiles. Additionally, the seed crystal can be easily etched by the volatiles from the solution, resulting in the growing crystal falling. However, in our growth experiments utilizing MoO3 flux, these disadvantages could be effectively eliminated. The viscosity of a Cs2O-Li2O-B2O3-MoO3 system in the temperature range of 780-880 °C has been determined by the present authors for the first time, and the results are illustrated in Figure 2. As can be seen from the figure, the viscosity increases gradually with the increase of Al doping concentration at the same temperature. Or rather, the viscosity increases rapidly from 198.4 to 239.5 cP with increasing Al doping concentration only from 2.5% to 5% at 800 °C, but it changes slightly with Al doping concentration in range of 0%-2.5% and 5%-10%. The viscosities of undoped, 2.5%, 5%, and 10% Al-doped CLBO growth solutions at their corresponding saturation temperature are 182. 5, 198.4, 205.3, and 209.8 cP, respectively. Obviously, there is a tremendous decrease in viscosity compared with self-flux system (the viscosity beyond 1000 cP) [40], which often inevitably leads to low growth repeatability and the formation of inclusions in crystal. The above results confirm that the introduction of MoO3 can enormously reduce the viscosity of the growth solution. Thus, the Cs2O-Li2O-B2O3-MoO3 system has a great advantage over the Cs2O-Li2O-B2O3 system in that it possesses a clearly lower volatility and a significantly lower viscosity.   Table 1 shows that the measured saturation temperature was in the range of 800-811 • C, which was remarkably lower than that of self-flux system (837-844 • C) [39]. Consequently, this flux system can reduce volatility so as to avoid component deviation caused by the relatively high evaporation of Cs. In the case of Al-doped CLBO grown from self-flux system, it is difficult to accurately determine saturation temperature due to the relatively great fluctuation of saturation temperature induced by volatiles. Additionally, the seed crystal can be easily etched by the volatiles from the solution, resulting in the growing crystal falling. However, in our growth experiments utilizing MoO 3 flux, these disadvantages could be effectively eliminated. The viscosity of a Cs 2 O-Li 2 O-B 2 O 3 -MoO 3 system in the temperature range of 780-880 • C has been determined by the present authors for the first time, and the results are illustrated in Figure 2. As can be seen from the figure, the viscosity increases gradually with the increase of Al doping concentration at the same temperature. Or rather, the viscosity increases rapidly from 198.4 to 239.5 cP with increasing Al doping concentration only from 2.5% to 5% at 800 • C, but it changes slightly with Al doping concentration in range of 0%-2.5% and 5%-10%. The viscosities of undoped, 2.5%, 5%, and 10% Al-doped CLBO growth solutions at their corresponding saturation temperature are 182. 5, 198.4, 205.3, and 209.8 cP, respectively. Obviously, there is a tremendous decrease in viscosity compared with self-flux system (the viscosity beyond 1000 cP) [40], which often inevitably leads to low growth repeatability and the formation of inclusions in crystal. The above results confirm that the introduction of MoO 3 can enormously reduce the viscosity of the growth solution. Thus, the Cs 2 O-Li 2 O-B 2 O 3 -MoO 3 system has a great advantage over the Cs 2 O-Li 2 O-B 2 O 3 system in that it possesses a clearly lower volatility and a significantly lower viscosity.

XRD and ICP Analysis
The powder X-ray diffraction (XRD) measurement was performed to verify the structure of as-grown Al-doped crystals. As displayed in Figure 3, we found that the positions and intensities of main diffraction peaks are well in accordance with these of undoped CLBO crystal. The results confirm that Al doping do not change the essential structure of CLBO.
In order to determine the effective Al 3+ segregation coefficient, ICP analysis was carried out to measure the Al 3+ concentration in the crystal. The effective segregation coefficient K can be calculated by the following equation: k = Cs/C0, where Cs and C0 correspond to Al 3+ concentration in the crystal and Al 3+ concentration in the growth solution, respectively. Based on the measured Al 3+ concentration in the as-grown crystal, the calculated effective Al 3+ segregation coefficients of 2.5%, 5%, and 10% Al-doped CLBO crystals are 0.028, 0.034, and 0.031, respectively, which are slightly higher than those of previous study [39].

Transmission Spectrum
For investigation of Al doping on the transmittance, the as-grown crystals were cut along (001) wafer and optically polished to about 1 mm thickness for spectrum measurement. The transmission spectrum of undoped and Al-doped CLBO crystals ranging from 185 to 1800 nm are presented in Figure 4. As is evident from Figure 4, both undoped and Al-doped crystals exhibit extremely high transmittance and exceed 92%. Moreover, there was no absorption band in the range of 200-380 nm, demonstrating that Mo ion did not crystallize into the crystal, and thus the ultraviolet cutoff wavelength could not be shifted [35]. Consequently, it can be concluded that as-grown crystals present excellent optical quality and Al doping has no adverse effect on the transmittance.

XRD and ICP Analysis
The powder X-ray diffraction (XRD) measurement was performed to verify the structure of as-grown Al-doped crystals. As displayed in Figure 3, we found that the positions and intensities of main diffraction peaks are well in accordance with these of undoped CLBO crystal. The results confirm that Al doping do not change the essential structure of CLBO.

XRD and ICP Analysis
The powder X-ray diffraction (XRD) measurement was performed to verify the structure of as-grown Al-doped crystals. As displayed in Figure 3, we found that the positions and intensities of main diffraction peaks are well in accordance with these of undoped CLBO crystal. The results confirm that Al doping do not change the essential structure of CLBO.
In order to determine the effective Al 3+ segregation coefficient, ICP analysis was carried out to measure the Al 3+ concentration in the crystal. The effective segregation coefficient K can be calculated by the following equation: k = Cs/C0, where Cs and C0 correspond to Al 3+ concentration in the crystal and Al 3+ concentration in the growth solution, respectively. Based on the measured Al 3+ concentration in the as-grown crystal, the calculated effective Al 3+ segregation coefficients of 2.5%, 5%, and 10% Al-doped CLBO crystals are 0.028, 0.034, and 0.031, respectively, which are slightly higher than those of previous study [39].

Transmission Spectrum
For investigation of Al doping on the transmittance, the as-grown crystals were cut along (001) wafer and optically polished to about 1 mm thickness for spectrum measurement. The transmission spectrum of undoped and Al-doped CLBO crystals ranging from 185 to 1800 nm are presented in Figure 4. As is evident from Figure 4, both undoped and Al-doped crystals exhibit extremely high transmittance and exceed 92%. Moreover, there was no absorption band in the range of 200-380 nm, demonstrating that Mo ion did not crystallize into the crystal, and thus the ultraviolet cutoff wavelength could not be shifted [35]. Consequently, it can be concluded that as-grown crystals present excellent optical quality and Al doping has no adverse effect on the transmittance. In order to determine the effective Al 3+ segregation coefficient, ICP analysis was carried out to measure the Al 3+ concentration in the crystal. The effective segregation coefficient K can be calculated by the following equation: k = C s /C 0 , where C s and C 0 correspond to Al 3+ concentration in the crystal and Al 3+ concentration in the growth solution, respectively. Based on the measured Al 3+ concentration in the as-grown crystal, the calculated effective Al 3+ segregation coefficients of 2.5%, 5%, and 10% Al-doped CLBO crystals are 0.028, 0.034, and 0.031, respectively, which are slightly higher than those of previous study [39].

Transmission Spectrum
For investigation of Al doping on the transmittance, the as-grown crystals were cut along (001) wafer and optically polished to about 1 mm thickness for spectrum measurement. The transmission spectrum of undoped and Al-doped CLBO crystals ranging from 185 to 1800 nm are presented in Figure 4. As is evident from Figure 4, both undoped and Al-doped crystals exhibit extremely high transmittance and exceed 92%. Moreover, there was no absorption band in the range of 200-380 nm, demonstrating that Mo ion did not crystallize into the crystal, and thus the ultraviolet cutoff wavelength could not be shifted [35]. Consequently, it can be concluded that as-grown crystals present excellent optical quality and Al doping has no adverse effect on the transmittance.

Laser-Induced Damage Threshold
Laser-induced damage is one of the major limitations associated with the development of high-power solid-state lasers, because it can seriously restrict the available input power. Bulk laser-induced damage is mainly related to crystal quality, so we measured the bulk LIDT of as-grown crystal for the purpose of evaluating its prospect in practical applications. The CLBO crystal sample with dimensions of 5 × 5 × 5 mm 3 was employed and two (001) faces were optically polished. The bulk LIDT was tested by means of the 1-on-1 test procedure using a Q-switched Nd:YAG laser (1064 nm) operating in a longitudinal single mode. The pulse repetition rate and pulse duration were 1 kHz and 5 ns, respectively. During the measurement, ten sites were chosen as test points at each pulse fluence, and laser-induced damage was confirmed to have occurred by visual observation of a laser spark. The pulse fluence which causes the corresponding damage probability of 0% is defined as the laser-induced damage threshold of the sample. Figure 5 shows the measured damage probability of undoped and Al-doped CLBO crystals. The results show that the LIDT of undoped and Al-doped CLBO crystals along the c axis are 33.98 J/cm 2 (6.80 GW/cm 2 ) and 25.48 J/cm 2 (5.10 GW/cm 2 ), respectively. The LIDT of newly-grown undoped CLBO significantly exceed that of conventional undoped CLBO (5.72 GW/cm 2 at 1064 nm along the c axis) [15]; the enhancement of LIDT may be attributed to high crystallinity and low dislocation density of crystal [19,20,41]. Hence, the results further prove the high quality of as-grown crystals and their promising applications in high-power solid-state lasers.

Optical Weak Absorption
The PCI system is a precise weak absorption measurement apparatus based on surface thermal lensing technique. The probe beam crosses the pump beam at a certain angle. As the crystal sample

Laser-Induced Damage Threshold
Laser-induced damage is one of the major limitations associated with the development of high-power solid-state lasers, because it can seriously restrict the available input power. Bulk laser-induced damage is mainly related to crystal quality, so we measured the bulk LIDT of as-grown crystal for the purpose of evaluating its prospect in practical applications. The CLBO crystal sample with dimensions of 5 × 5 × 5 mm 3 was employed and two (001) faces were optically polished. The bulk LIDT was tested by means of the 1-on-1 test procedure using a Q-switched Nd:YAG laser (1064 nm) operating in a longitudinal single mode. The pulse repetition rate and pulse duration were 1 kHz and 5 ns, respectively. During the measurement, ten sites were chosen as test points at each pulse fluence, and laser-induced damage was confirmed to have occurred by visual observation of a laser spark. The pulse fluence which causes the corresponding damage probability of 0% is defined as the laser-induced damage threshold of the sample. Figure 5 shows the measured damage probability of undoped and Al-doped CLBO crystals. The results show that the LIDT of undoped and Al-doped CLBO crystals along the c axis are 33.98 J/cm 2 (6.80 GW/cm 2 ) and 25.48 J/cm 2 (5.10 GW/cm 2 ), respectively. The LIDT of newly-grown undoped CLBO significantly exceed that of conventional undoped CLBO (5.72 GW/cm 2 at 1064 nm along the c axis) [15]; the enhancement of LIDT may be attributed to high crystallinity and low dislocation density of crystal [19,20,41]. Hence, the results further prove the high quality of as-grown crystals and their promising applications in high-power solid-state lasers.

Laser-Induced Damage Threshold
Laser-induced damage is one of the major limitations associated with the development of high-power solid-state lasers, because it can seriously restrict the available input power. Bulk laser-induced damage is mainly related to crystal quality, so we measured the bulk LIDT of as-grown crystal for the purpose of evaluating its prospect in practical applications. The CLBO crystal sample with dimensions of 5 × 5 × 5 mm 3 was employed and two (001) faces were optically polished. The bulk LIDT was tested by means of the 1-on-1 test procedure using a Q-switched Nd:YAG laser (1064 nm) operating in a longitudinal single mode. The pulse repetition rate and pulse duration were 1 kHz and 5 ns, respectively. During the measurement, ten sites were chosen as test points at each pulse fluence, and laser-induced damage was confirmed to have occurred by visual observation of a laser spark. The pulse fluence which causes the corresponding damage probability of 0% is defined as the laser-induced damage threshold of the sample. Figure 5 shows the measured damage probability of undoped and Al-doped CLBO crystals. The results show that the LIDT of undoped and Al-doped CLBO crystals along the c axis are 33.98 J/cm 2 (6.80 GW/cm 2 ) and 25.48 J/cm 2 (5.10 GW/cm 2 ), respectively. The LIDT of newly-grown undoped CLBO significantly exceed that of conventional undoped CLBO (5.72 GW/cm 2 at 1064 nm along the c axis) [15]; the enhancement of LIDT may be attributed to high crystallinity and low dislocation density of crystal [19,20,41]. Hence, the results further prove the high quality of as-grown crystals and their promising applications in high-power solid-state lasers.

Optical Weak Absorption
The PCI system is a precise weak absorption measurement apparatus based on surface thermal lensing technique. The probe beam crosses the pump beam at a certain angle. As the crystal sample

Optical Weak Absorption
The PCI system is a precise weak absorption measurement apparatus based on surface thermal lensing technique. The probe beam crosses the pump beam at a certain angle. As the crystal sample is irradiated by a pump beam, the refractive index change can occur in the heated area. Accordingly, the probe beam suffered phase distortion in this area induced by common-path point diffraction interference. The perceived phase distortion signals of the probe beam are collected by a photodetector behind an aperture, and then are accurately processed by the lock-in amplifier and transmitted to the computer. Samples with size of 5 × 5 × 5 mm 3 were optically polished for this measurement.
The optical weak absorption curves of undoped and Al-doped CLBO crystals at the critical wavelength of 1064 nm were obtained, as presented in Figures 6 and 7, respectively. The two highest peaks are regarded as the weak absorption of two end surfaces of the crystal, while the stable area between the two peaks represents the measured bulk weak absorption of the crystal. As seen from the two figures, the weak absorption values of undoped CLBO crystal along a and c directions were only 80 and 50 ppm/cm, respectively, while the weak absorption values of Al-doped CLBO crystal along a and c directions were approximately 140 and 50 ppm/cm, respectively. The measured weak absorption values are dramatically lower compared to previous investigation (600 ppm/cm for a direction and 150 ppm/cm for c direction) [42], as listed in Table 2. To our knowledge, the measured weak absorption values in the present work are by far the minimum levels reported in articles. Since the weak absorption values are closely correlated with the optical quality of the crystal, the newly grown crystals possess outstanding optical quality. is irradiated by a pump beam, the refractive index change can occur in the heated area. Accordingly, the probe beam suffered phase distortion in this area induced by common-path point diffraction interference. The perceived phase distortion signals of the probe beam are collected by a photodetector behind an aperture, and then are accurately processed by the lock-in amplifier and transmitted to the computer. Samples with size of 5 × 5 × 5 mm 3 were optically polished for this measurement.
The optical weak absorption curves of undoped and Al-doped CLBO crystals at the critical wavelength of 1064 nm were obtained, as presented in Figures 6 and 7, respectively. The two highest peaks are regarded as the weak absorption of two end surfaces of the crystal, while the stable area between the two peaks represents the measured bulk weak absorption of the crystal. As seen from the two figures, the weak absorption values of undoped CLBO crystal along a and c directions were only 80 and 50 ppm/cm, respectively, while the weak absorption values of Al-doped CLBO crystal along a and c directions were approximately 140 and 50 ppm/cm, respectively. The measured weak absorption values are dramatically lower compared to previous investigation (600 ppm/cm for a direction and 150 ppm/cm for c direction) [42], as listed in Table 2. To our knowledge, the measured weak absorption values in the present work are by far the minimum levels reported in articles. Since the weak absorption values are closely correlated with the optical quality of the crystal, the newly grown crystals possess outstanding optical quality.    is irradiated by a pump beam, the refractive index change can occur in the heated area. Accordingly, the probe beam suffered phase distortion in this area induced by common-path point diffraction interference. The perceived phase distortion signals of the probe beam are collected by a photodetector behind an aperture, and then are accurately processed by the lock-in amplifier and transmitted to the computer. Samples with size of 5 × 5 × 5 mm 3 were optically polished for this measurement.
The optical weak absorption curves of undoped and Al-doped CLBO crystals at the critical wavelength of 1064 nm were obtained, as presented in Figures 6 and 7, respectively. The two highest peaks are regarded as the weak absorption of two end surfaces of the crystal, while the stable area between the two peaks represents the measured bulk weak absorption of the crystal. As seen from the two figures, the weak absorption values of undoped CLBO crystal along a and c directions were only 80 and 50 ppm/cm, respectively, while the weak absorption values of Al-doped CLBO crystal along a and c directions were approximately 140 and 50 ppm/cm, respectively. The measured weak absorption values are dramatically lower compared to previous investigation (600 ppm/cm for a direction and 150 ppm/cm for c direction) [42], as listed in Table 2. To our knowledge, the measured weak absorption values in the present work are by far the minimum levels reported in articles. Since the weak absorption values are closely correlated with the optical quality of the crystal, the newly grown crystals possess outstanding optical quality.

Humidity Experiment
In order to investigate the influence of Al doping on the hygroscopic behavior of CLBO crystal, a humidity experiment was conducted at a constant temperature in a humidity chamber. Two optically polished samples with dimensions of 4 × 4 × 3 mm 3 -namely, A (undoped CLBO crystal) and B (5% Al-doped CLBO crystal)-were kept in the program-controlled chamber at relative humidity of 60% and temperature of 25 • C for different durations. The surface micrographs of the (001) plane are given in Figure 8 (under an optical microscope with 100 magnification). From Figure 8, some regular etch pits can be clearly observed on the edges of (001) plane in sample A with an increase in duration time from 0 to 1200 h. In contrast, there were no observable etch pits in sample B, even over a period of 50 days. The results demonstrate that Al doping could substantially improve the hygroscopic nature of CLBO.

Humidity Experiment
In order to investigate the influence of Al doping on the hygroscopic behavior of CLBO crystal, a humidity experiment was conducted at a constant temperature in a humidity chamber. Two optically polished samples with dimensions of 4 × 4 × 3 mm 3 -namely, A (undoped CLBO crystal) and B (5% Al-doped CLBO crystal)-were kept in the program-controlled chamber at relative humidity of 60% and temperature of 25 °C for different durations. The surface micrographs of the (001) plane are given in Figure 8 (under an optical microscope with 100 magnification). From Figure  8, some regular etch pits can be clearly observed on the edges of (001) plane in sample A with an increase in duration time from 0 to 1200 h. In contrast, there were no observable etch pits in sample B, even over a period of 50 days. The results demonstrate that Al doping could substantially improve the hygroscopic nature of CLBO.

266 nm UV Laser Conversion Efficiency
A picosecond mode-locked Nd:YAG laser (pulse width: 25 ps, pulse repetition frequency: 10 Hz) was employed as the fundamental light source for the laser output and fourth harmonic generation (FHG) conversion efficiency measurements. Samples of uncoated and optically polished undoped and Al-doped CLBO crystals with dimension of 3 × 3 × 10 mm 3 were cut along the angles of (θ, φ) = (61.6°, 45°) and θBrewster = 56.3° according to type-I phase matching direction. Figure 9 displays the conversion efficiency of the 266 nm radiation as a function of 532 nm input peak power density. Remarkably, the FHG conversion efficiency from 532 nm to 266 nm of CLBO increases gradually with increasing the 532 nm input peak power density. The maximum conversion efficiencies of undoped and Al-doped CLBO crystals can reach 65% and 63% at their corresponding peak power density of 325 MW/cm 2 , respectively. Despite the FHG conversion efficiency of Al-doped CLBO crystal is slightly lower than that of undoped CLBO crystal, the difference between them being very small. The decrease of conversion efficiency may be connected with the reduction of the crystalline perfection due to Al doping [43]. The results illustrate that Al doping has no obvious influence on the FHG conversion efficiency of CLBO.

266 nm UV Laser Conversion Efficiency
A picosecond mode-locked Nd:YAG laser (pulse width: 25 ps, pulse repetition frequency: 10 Hz) was employed as the fundamental light source for the laser output and fourth harmonic generation (FHG) conversion efficiency measurements. Samples of uncoated and optically polished undoped and Al-doped CLBO crystals with dimension of 3 × 3 × 10 mm 3 were cut along the angles of (θ, φ) = (61.6 • , 45 • ) and θ Brewster = 56.3 • according to type-I phase matching direction. Figure 9 displays the conversion efficiency of the 266 nm radiation as a function of 532 nm input peak power density. Remarkably, the FHG conversion efficiency from 532 nm to 266 nm of CLBO increases gradually with increasing the 532 nm input peak power density. The maximum conversion efficiencies of undoped and Al-doped CLBO crystals can reach 65% and 63% at their corresponding peak power density of 325 MW/cm 2 , respectively. Despite the FHG conversion efficiency of Al-doped CLBO crystal is slightly lower than that of undoped CLBO crystal, the difference between them being very small. The decrease of conversion efficiency may be connected with the reduction of the crystalline perfection due to Al doping [43]. The results illustrate that Al doping has no obvious influence on the FHG conversion efficiency of CLBO.

Conclusions
In summary, undoped and Al-doped CLBO crystals free from inclusions were successfully grown using the newly developed Cs2O-Li2O-MoO3 fluxes. The Cs2O-Li2O-B2O3-MoO3 system could drastically lower the viscosity of the growth solution compared to that of extensively used self-flux system. The as-grown crystals exhibited excellent quality. The measured weak absorption values of undoped and Al-doped CLBO were remarkably smaller than previous studies. The LIDT of undoped CLBO came up to 6.80 GW/cm 2 -19% higher than that of conventional undoped CLBO. Our results further illustrate that Al doping can significantly enhance the resistance of CLBO to water. The undoped and Al-doped CLBO with FHG conversion efficiencies greater than 63% have been realized, and Al doping has a weak influence on FHG conversion efficiency. Therefore, the newly grown Al-doped CLBO crystal has been demonstrated to be a highly promising NLO crystal for UV light generation.

Conclusions
In summary, undoped and Al-doped CLBO crystals free from inclusions were successfully grown using the newly developed Cs 2 O-Li 2 O-MoO 3 fluxes. The Cs 2 O-Li 2 O-B 2 O 3 -MoO 3 system could drastically lower the viscosity of the growth solution compared to that of extensively used self-flux system. The as-grown crystals exhibited excellent quality. The measured weak absorption values of undoped and Al-doped CLBO were remarkably smaller than previous studies. The LIDT of undoped CLBO came up to 6.80 GW/cm 2 -19% higher than that of conventional undoped CLBO. Our results further illustrate that Al doping can significantly enhance the resistance of CLBO to water. The undoped and Al-doped CLBO with FHG conversion efficiencies greater than 63% have been realized, and Al doping has a weak influence on FHG conversion efficiency. Therefore, the newly grown Al-doped CLBO crystal has been demonstrated to be a highly promising NLO crystal for UV light generation.