Endurance performance of transmissive liquid crystal phase and polarization controllers for kW-class high-power lasers

This study developed a Liquid Crystal (LC) device that incorporated lasers for machining with output power of the order of kW. The use of sapphire as a substrate along with an appropriate cooling system facilitated the LCs in enabling the control of the phase and polarization of high-power laser light, which is not possible in case of conventional LCs. The exposure of LCs to high-power laser light results in heat accumulation, which causes the destruction of the LC cells. As photochemical damage has a lower occurrence probability at longer wavelengths, the promotion of heat radiation is an important strategy for the application of LCs to near infrared laser beams. The phase control of laser beams up to 6 kW was achieved using the proposed LC substrate as a cooling plate and optimizing the cooling system. Moreover, we compensated for the variation in the properties of the LC with temperature via adjustments to the drive voltage of the LC.


INTRODUCTION
Structured beams, which provide intensity, phase, and polarization distribution in beam cross-sections, have many applications in optical instruments. Applications involving cylindrical vector beams with distributed polarization and phase, such as optical vortices, azimuthally polarized light, and radially polarized light, have been proposed for various uses, including microscopy [1][2][3], laser processing [4], optical communications [5], optical trapping [6], and particle acceleration [7]. The devices used to facilitate electrical control of polarization and phase include electro-optic modulators, acoustooptic modulators, photoelastic modulators, and liquid crystal (LC) devices. LCs exhibit promise for such applications owing to their simple structure, electrical controllability, scalability for large size, and spatial distribution.
However, the application of LCs to high-power lasers is challenging because of their characteristics at a high temperature. A LC cell exhibits a structure in which the LC is injected into a narrow gap between two glass substrates, thus resulting in the deposition of thin films such as transparent electrodes and alignment films at the interface between the LC and glass. These materials are transparent and designed to minimize absorption of light. However, when irradiated via a high-power laser, even small amounts of light absorption can result in destructive heat generation. A previous study examined the resistance of LC devices to nanosecond pulsed lasers with high pulse energies. The durability of alignment films [8] and LCs [9] has been improved using highly photo-durable materials, and LC cells capable of withstanding nanosecond pulsed laser output of several joules have been developed [10]. Furthermore, studies have examined the destruction thresholds and mechanisms of transparent electrodes, alignment film materials, and LC substrates under light irradiation [11][12][13].
Rather than focusing on peak intensity, this study examined the durability of high-power CW lasers. The trend in laser processing has shifted from micromachining to macromachining; consequently, the required breakdown threshold for LCs is a pulse energy of high power, i.e., the wattage that the LC can withstand. We fabricated transmissive LC devices with enhanced heat dissipation and confirmed their resistance to high-power near infrared (NIR) fiber-lasers up to 6 kW via experiments. Moreover, we have succeeded in compensating for thermal-induced variations in the operating characteristics of the LCs.

PRINCIPLE
This study developed a transmissive static LC cell. Static LC are inexpensive, compact, and can be installed in any optical equipment via the simple insertion of an LC into its optical path. As shown in Fig. 1, the basic structure was similar to that of ordinary static LCs. However, the LC cell comprised substrates with high thermal conductivity to support high-power lasers. This experiment employed sapphire substrates, which exhibits a thermal conductivity that is 56 times greater than that of glass. When incident light passes via a LC cell, coincident light absorption results in the generation of a heat source. The heat generated propagates by the cell via thermal conduction. Furthermore, it is radiated via thermal conduction to the housing incorporating the LC cell as well as through heat transfer from the surface of the LC cell to the air; however, the heat radiation is negligible. Moreover, the temperature distribution of a LC cell in steady state was determined at the instance the heat balance reached equilibrium. Heat conduction to the housing occurred in the in-plane direction of the LC cell, whereas heat transfer to the air and heat radiation were in the out-of-plane direction. Therefore, the specifications of cells required for efficient cooling differed in terms of enclosure and natural air cooling. In general, the thickness of the LC layer was several micrometers, alignment film was several hundred nm, ITO was approximately 10-100-nm-thick, and the substrate was approximately several hundred micrometers thick. Although thermal conductivity is important for both the cooling processes, the thinner the substrate, the better it is in case of natural air cooling. For chassis cooling, a thicker substrate exhibited improved cooling performance owing to the easier in-plane heat conduction toward the edges of the substrate. In practice, rendering a thin substrate is challenging, and natural cooling is insufficient for high-power lasers. Therefore, assuming enclosure cooling, the thickness of the LC cell substrate must be sufficient to not prevent mounting. However, the heat flux in the out-of-plane direction is limited. For a particular temperature gradient, the heat flow rate is proportional to the product of thermal conductivity and heat transfer cross section. Therefore, the substrate, which is the thickest component of the LC cell, is an important element of heat conduction. Because the dimensions of the LC cell must not be significantly changed to satisfy the required performance and cost, the thermal conductivity must be increased to decrease the temperature of the LC. This study employed c-plane sapphire as a LC substrate with a thermal conductivity of 41 W/(m･K) (56 times higher than that of glass). Compared with other high thermal conductivity transparent substrates, c-plane sapphire provides a good balance between cost and performance; moreover, it is commercially viable as a product.

MATERIALS AND METHODS
The structure of the LC cell was prepared, as shown in Fig. 1, and then its resistance and characteristics under highpower laser irradiation were evaluated. To perform evaluations of durability to lasers of 100 W or less, a LC cell with dimensions of 17.2 mm × 19.7 mm with a substrate thickness of 0.5 mm was prepared. Soda-lime glass, borosilicate glass, silica, and sapphire were used as substrates, and a 35 mm 2 , 0.5-mm-thick sapphire substrate was used for the kWorder laser. Furthermore, ITO and polyimide alignment films, measuring 15-150 and 200 nm in thickness, respectively, were deposited on the LC substrate. The alignment films were rubbed such that the rubbing directions of the upper and lower substrates were antiparallel. This resulted in an electrically controlled birefringence (ECB) LC cell in which the birefringence phase was controlled via the drive voltage. This study performed simulations and experiments to examine the endurance performance and temperature elevation under natural cooling for lasers with a power of <100 W. The breakdown thresholds of four different substrates were measured, and the thermal lensing effect was evaluated. For kW-class lasers, a water-cooling system was developed, experimental durability tests were conducted, and the effect of temperature rise on polarization of LCs was modified. Furthermore, fiber lasers (YLR-100-SM, iPG LASER) and (FF6000i-model A, FANUC) were used for powers of< 100 W and >1 kW, respectively.

Light absorption in LC cells
Static LCs have transparent electrodes, alignment films, and LCs; however, there is a small amount of light absorption. To realize a quantitative evaluation of the optical absorption, transmittance and reflectance were determined using a spectrometer having an integrating sphere; the results are shown in Fig. 2. The absorption spectra were measured at the end of each fabrication process of the LC cell, beginning with the sapphire substrate, followed by the deposition of ITO and alignment film, and finally culminating with the injection of LC into the LC cell. Here, absorption for two substrates was included in the result of a cell in which LC was injected. This is because a pair of substrates was required for a LC cell.
In this study, NIR fiber laser with target laser wavelength of 1070 nm was used, which caused minimal optical damage to LC materials and necessitated the consideration of only the thermal effects. Moreover, this is the most popular wavelength for laser processing. The results demonstrated that the contributions of the ITO electrode and the LC were 0.5 and 1% at 1070 nm, respectively. Moreover, the contribution of the alignment film was negligible.

High-power laser resistance of LC cells without water cooling
The optical absorption rate obtained from the measurements was used to perform finite element method simulations and experiments regarding the temperature change in the LC under laser irradiation. The device size was 17.2 mm2, with the thicknesses of the substrate, ITO, alignment film, and LC layer being 0.4 mm, 150 nm, 200 nm, and 15 μm, respectively. Assuming that the laser was 2 mm in diameter and 35 W CW, 0.14 and 1.31% were absorbed by the ITO and LC layer, respectively, thus resulting in a heat generation of 0.049 and 0.50 W, respectively. The cooling conditions were natural convection cooling in air (293.2 K). Furthermore, the NI point temperature of the LC was 387.15 K. Figure 3 shows a plot of the temporal variation of the maximum and minimum temperatures of the LC and the substrate. An obvious difference between the glass-and sapphire-substrate LCs was observed, which reached 389.0 and 366.6 K in 6 and 300 s, respectively. Thus, the temperature of the sapphire substrate LC cell was lower than the NI point and maintained its function as a LC. However, the temperature of the glass substrate LC cell exceeded the NI point after 6 sec, indicating that the glass substrate LC cell lost its function as a LC. Note that additional simulations demonstrated that the temperature of glass substrate LC increased to >550 K in 300 s, indicating that irreversible destruction of the LC device will occur. Furthermore, there was a difference in the temperature distribution in the plane of the LC cell. In the LC layer, the difference between the highest and lowest temperatures in the cell was 7.4 K (@300 s) for the sapphiresubstrate LC cell and 93.8 K (@6 s) for the glass-substrate LC cell. These results indicate that the temperature distribution within the LC cell plane was homogenized by the thermal conduction of the sapphire substrate. Consequently, the LC temperature was reduced and the LC characteristics were rendered uniform as well as having the temperature distribution. Moreover, the temperature difference between the substrate and the LC was smaller for the sapphire-substrate LC cell, with the temperature at the center of the LC cell at its maximum being 0.15 and 0.18 K for the sapphire-and glass-substrate LC cells, respectively. These results confirmed the loss of the function of the glasssubstrate LC cell after only 6 s of beam irradiation. This can be attributed to the heat accumulating in both the in-and out-of-plane directions owing to the lack of thermal conductivity. respectively. The temperature of glass substrate cell was greater than 120 °C, which is beyond the temperature measurement range of our experiment. Moreover, it exceeded the NI point of the LC, thus indicating the difficulty in using the LC cell. However, the temperature of the sapphire substrate cell was 65.7°C, and the temperature distribution was uniform across the entire substrate. This indicated that the LC was within the available temperature range and that the properties of the LC is homogeneous within the plane because of the uniform temperature distribution. In this experiment, the temperatures of all substrates were lower than in the simulation, and the glass-substrate LC cells were not destroyed as simulated. This can be attributed to that the requirement of a small amount of contact with the jig to hold the LC cell. Moreover, from this point, cooling by heat conduction via contact occurred.
To detect defects caused by local temperature elevation, the optical system that is shown in Fig. 4(g) was employed. Figures 4(b), (c), and (e), (f) show the results for the glass-and sapphire-substrate LC cells, respectively. In Figures 4(b) and (e), the polarizers were removed to facilitate a visual inspection using transmitted light. Consequently, the phase change from the LC to the isotropic phase owing to temperature rise was easily observed visually in the glass-substrate LC cell. However, no change in appearance was observed for the sapphire-substrate LC cells. In Figs. 4(c) and (f), the polarization distribution was observed through a cross-Nicol arrangement of polarizers. For the glass -substrate LC cell, the center of the laser-irradiated area was isotropic and appeared black owing to the crossed-Nicol arrangement. Moreover, the brightness around the center changed as well, which can be attributed to the effect of the temperature gradient in the plane of the LC cell on the polarization characteristics of the LC. However, no such non-uniform polarization distribution is observed in the sapphire-substrate LC cell.
Subsequently, the destruction threshold was measured based on the determination of the intensity of laser incidence that caused irreversible damage to the LC. Figure 5(a) shows the breakdown thresholds of the LC cells prepared using four substrates. The destruction threshold and mechanism are dependent on the substrate. For the sapphire-substrate, the LC cell resisted breakage even at the maximum laser power. However, for soda-lime and borosilicate glass, mechanical breakdown of the substrates occurred owing to thermal stress followed by the internal breakdown of the LC cell. However, for silica glass, which exhibits a thermal conductivity close to that of glasses, the coefficient of thermal expansion was negligibly small and thermal stress fracture was not observed; nevertheless, the device became red-hot at approximately 80 W. Furthermore, the organic material inside the cell was destroyed by carbonization, thus resulting in thermal destruction. Figure 5(b) shows the results of surface temperature measurements while increasing the incident laser power for the silica-and sapphire-substrate LC cells in which no substrate cracking occurred. The temperatures were measured via a thermal imaging camera. The results demonstrated that the silica-substrate LC cell broke down after exceeding the NI transition temperature. However, the sapphire-substrate LC cell maintained a sufficiently low temperature and is thus expected to withstand power outputs of 100 W or more. Moreover, similar to the results presented in Fig. 4, the temperature distribution in the substrate is more uniform in the sapphire-substrate LC cell.  We examined the magnitude of the thermal lensing effect. Thermal lensing is a phenomenon in which a beam is focused through the thermal modulation of the refractive index. For an inhomogeneous temperature distribution in the LC, a distribution of refractive indices was generated in the substrate and LC layer, which changed the size of the beam. The thermal lensing effect was measured employing an experimental setup shown in Fig. 6(a). The laser beam was projected on a screen at a distance greater than 7 m with its beam diameter measured using a camera. Figure 6(b) shows the results. The black plot indicates the original beam diameter. The red, and blue plots show the beam diameters when the silica-, and sapphire-substrate LC cells inserted in the optical path, respectively. The results confirmed the larger thermal lensing effect of the silica-substrate LC cell than for sapphire. Moreover, the increase in the thermal lensing effect was accelerated, particularly when the laser power exceeded 30 W. The effect on beam diameter was smaller for the sapphire-substrate LC cell, regardless of the laser power. Furthermore, the effect of laser power on the beam diameter of the sapphire-substrate LC cell was smaller than that of the silica-substrate LC cell. These results are related to the LC temperature measurement results shown in Fig. 5(b). Even though the value of the material parameter dn/dT [14,15], which indicates the magnitude of the thermal lensing effect, was larger for sapphire than for silica, the thermal lensing effect was lower for sapphire. Therefore, it was found that the uniformity of temperature distribution due to sapphire's high thermal conductivity contributed significantly to the reduction of the thermal lensing effect.

Resistance of LC cells to kW lasers using water cooling
Natural cooling is not sufficient for the installation of a sapphire-substrate LC cell in the optical path of a high-power laser of the order of kW. Therefore, this study proposed an appropriate combination with a water-cooling system to ensure durability at laser powers of up to 6 kW. Moreover, we compensated for the inhomogeneous polarization distribution owing to the temperature distribution. Figure 7 shows the results of the water-cooled cooling system for the LC cell and the corresponding endurance test. Based on the absorption rate of the LC cell, the heat generated by the LC cell reached 87 W for 6 kW of incident light. Therefore, a chiller with a capacity of ~300 W was used for cooling. The laser beam irradiating the LC cell was magnified, and the beam diameter at the position where the laser was irradiated to the LC cell was determined to be 9 mm. The temperature of the cooling water was set at 15°C. A movie illustrating a kW laser endurance test conducted using this configuration is shown in Fig. 7(c). The camera and thermal imaging camera performed simultaneous recording. In this movie, the power was gradually increased from an initial value of 600 W to a maximum of 6 kW. Moreover, irradiation was applied for ~30 min to confirm the non-observance of any damage. Figure  7(d) shows the dependence of the temperature of the LC cell measured via the thermal imaging camera on the incident laser power. Thus, this experiment indicated that the temperature of the LC cell did not reach the NI point even under the action of an incident laser power of ~6 kW.
However, under kW-class laser irradiation, the temperature distribution within the surface of the LC cell was inhomogeneous even when using a powerful water-cooling system. Figure 8 shows an example of a 16-segments LC cell with radial and concentric divided segments, wherein a laser irradiated at the center resulted in an increase in the temperature and changed the phase difference within the LC. A glass substrate LC was used to provide a large temperature gradient. Figure 8(f) shows the laser irradiation of the LC and the optical system for observation. Figures  8(a) and (b) show the temperature distribution before and after irradiation, respectively, and Figs. 8(c) and (d) show the polarization camera images, respectively. In the polarization camera images, the color implies direction, and the darkness indicated circularity of the polarization. With increase in the temperature owing to laser irradiation, the birefringence phase shift decreased and the linear polarization was converted to circular polarization. Consequently, the voltage at the laser irradiated area was increased by 0.5 V to correct the linear polarization, as shown in Fig. 8(e).

CONCLUSION
Using a sapphire substrate with high thermal conductivity, this study realized an LC polarizing element capable of withstanding high-power lasers. The sapphire substrate efficiently radiated the heat generated inside the LC cell and maintained a low and uniform temperature distribution on the LC cell surface, thereby achieving a uniform polarization state. Furthermore, the sapphire substrate LC cells addressed lasers of 100 W or greater with natural cooling and up to 6 kW with water cooling. The temperature inhomogeneity in the LC cell because of laser irradiation resulted in inhomogeneous birefringence, which was compensated for through adjustments to the voltage applied to the liquid crystal.