Experimental Study on the Damage of Optical Materials by Out of Band Composite Laser

: For the paper, experimental studies were performed on the damage of the Ge- and Si-based flat window by lasers out-of-band. The experimental results showed that lasers out-of-band can cause film damage and substrate damage to Ge and Si windows. The high-energy laser damage window mechanism mainly manifested as thermal effects. The composite laser damage thresholds for the substrate were an Si window of 21.6 J/cm 2 and a Ge window of 3 J/cm 2 . Compared with continuous laser and long pulse laser experimental results, it was found that the use of long pulse-continuous composite constitution could effectively reduce the damage threshold. Compared to the long-pulse laser, the composite laser could achieve similar damage effects with a smaller energy density.


Introduction
Laser-induced damage to optical components is a key research issue in high-energy laser emission systems, and it is also one of the key technologies that need to be resolved for the development of high-power optoelectronic countermeasure systems. Starting from the basic principle of the interaction between lasers and matter, a laser can interact with optical systems and optical elements through the laser thermal effect and laser-electron interactions. This provides a theoretical basis for a single-band laser to achieve full-band photoelectric loading. Based on this principle, researchers have proposed the concepts of "in-band damage" and "out-band damage." "In-band damage" refers to the damage of an optoelectronic system by a laser in its operating band. Researchers generally believe that the damage mechanism of "in-band damage" comprises the semiconductor band structure theory, thermoelectric effect, etc., and "in-band damage" has been widely used in contemporary optoelectronic countermeasures. "Out-of-band damage" refers to the damage of photoelectric systems by lasers outside the operating band. Earlier studies have suggested that optoelectronic components do not respond or respond weakly to the lasers outside the operating band. However, with the advancement of laser technology, more and more high-power and high-energy lasers have begun to be applied to high-power laser emission systems. This has caused the risk of damage to optical systems by lasers outside the high-energy band. Therefore, it is necessary to carry out systematic experimental research on the interaction between high-energy lasers outside the band and optical elements.
At present, the research reports on "out-band damage" have mainly focused on photodetectors. The related literature has conducted experimental studies on interference and damage of HgCdTe, InSb, and Si-CCD (charge coupled device) detectors [1][2][3][4][5][6][7]. The mainstream view believes that the mechanism of "out-band damage" is mainly the semiconductor band structure theory (CCD) and thermoelectric effects (Mid-wavelength infrared and Long-wavelength infrared). Some researchers have found that out-of-band lasers can also damage window mirrors in experiments of laser interference effects on optical systems. Wang [8] used a Deuterium fluoride laser to perform cumulative damage experiments on a visible light plane array CCD and found that multiple irradiations of the laser at different positions on the CCD surface and multiple irradiations at the same position damaged the K9 (A glass model) optical window. Wang [9] experimented with a 3.8 μm continuous-wave laser to destroy the ternary Photoconductive type HgCdTe detector system and found that the film and substrate damage occurred in the Ge window at the laser irradiation spot; they also found the internal filter had a melting phenomenon. Existing research believes that the key to "out-band damage" lies in whether the laser source has a sufficient damage ability, and the multi-mode composite laser has this characteristic. A multi-mode composite laser consists of lasers with different wavelengths, different systems, and different frequency changes that act on the target at the same time or alternately to obtain a better damage effect than a single continuous-wave or Pulsed-laser. Related studies have been conducted on composite lasers: Cheng [10] carried out an experimental study on the combined damage of a 1030 nm continuous laser and a 1064 nm pulsed laser and found that the "non-linear avalanche ablation" effect occurred under the combined or alternating effects of two lasers, which made the combined laser have a stronger ablation effect than the pulse laser. The experimental results showed that the average single-pulse ablation amount of the composite laser was 13 times that of the pulse laser. Wang [11] found, in a study of pulse-pulse composite lasers, that the increase in target damage was a result of an increased power density. The damage effect of composite lasers is related to the overlap of pulse time domains, and the damaging effect of composite lasers is better than that of long-pulse lasers. Jiao [12], using 1053 nm pulsed and 1064 nm multiple compound lasers to study the irradiation effect of steel found that the surface reflection of steel decreased with the increase of pulsed laser frequency. The pre-irradiation of steel plates with a long pulse laser can increase its absorption rate for subsequent lasering. Xiao [13,14] simulated the thermodynamic characteristics of the continuous-pulse composite laser irradiation of aluminum alloys. Through simulation, it was found that the composite laser could significantly increase the size of the molten pool and increase the center temperature of the irradiation spot. The longer the "preheating" time, the shorter the yield time of the material, the larger the plastic deformation, and the larger the yield range.
According to the analysis of the above literature, it is known that the research on the laser irradiation effects of composite lasers is still in its infancy. Almost all reported studies have used low-energy in-band laser sources. The research targets have mostly been photoelectric sensors and metal structural parts. Composite laser damage studies on optical components have not been reported. For this paper, high-power laser damage experiments were performed on a common Ge-based and Si-based flat window to provide technical support for the design of high-power laser emission systems.

Absorption and Scattering of Optical Film
A flat window is composed of surface optical films and substrates. The following assumptions are used to solve the reflectivity of optical films: The film does not absorb incident light, its refractive index is uniform, and both interfaces are smooth. These assumptions accurately reflect the optical properties of general dielectric films. An optical film made of a single-layer film has a limited optical performance. To meet a variety of optical requirements, multi-layer films are needed. Suppose an optical film composed of an m-layer film-then, the j-layer has a refractive index of nj, and the feature matrix of this layer of film can be expressed as: where δj = (2r/λ)njdj is the phase thickness of the j-th layer film. The combined feature matrix of the m-layer film is: where n0, ns, and λ are the refractive index of the incident medium, the refractive index of the substrate, and the laser wavelength, respectively. For transparent dielectric films, m12 and m21 are pure imaginary numbers, and m1 and m22 are pure real numbers. The reflection coefficient r and reflectance R of the multilayer film are: The reflectivity of the optical film can be solved by the above two equations [15][16][17].

Temperature Field of Optical Film
The direct factor leading to the destruction of the optical film under laser irradiation is the temperature rise caused by the thermal effect. To study the laser damage of optical film, the time and space distribution of the temperature field of optical film should be considered, and these can be determined by solving the heat conduction equation with the laser source term. The heat conduction equation of a multilayer dielectric film and its initial and boundary conditions can be expressed as: Boundary conditions: r→∞ or z→∞:T→0 Initial conditions: Ti(r,z,0) = T0(r,z) Among them, ci and ki are the specific heat capacity and thermal conductivity of the i-th layer film, respectively; H is the surface heat exchange constant, which is related to the surface thermal radiation and thermal convection; and gi is the laser energy deposited in the i-th film: where Pi(r,t,z) is the Poynting vector of the i-th film. The energy transmission between the membrane layers meets the principles of temperature continuity and heat flow balance (i = 1, 2, …) leads to: where Ri is the thermal resistance at the interface between the i and i + 1 films, zi and zi+1 are the positions of the interface, and Ti and Ti+1 are the temperatures of the two films at the corresponding interfaces. Solving the equations provides the temperature distribution of the optical film [17][18][19][20].

Optical Element Substrate Thermal Response
The steady system assumes that the semiconductor material is crystal and its internal heat conduction may be anisotropic. The crystal structure of semiconductor materials such as silicon germanium is a cubic crystal system, and the thermal conductivity of this crystal structure is the spherical tensor. It can be considered that the specificity of heat conduction in the same characteristic material is the same [21,22].
The expression of the heat flux density vector can be obtained from Fourier's law: where Kij is the thermal conductivity coefficient component; xj (j = 1,2,3) is the coordinate in the xyz direction; and ei is the basis vector in the xyz direction. Bring Equation 1 into the heat flux density expression to obtain the heat conduction equation in the semiconductor material: where ρ is the density and cp is the specific heat capacity at constant pressure. The heat flux density of the laser incident surface is: The incident laser has a Gaussian distribution, as follows: where I0 is the incident power density, r is the distance from the reference point to the center of the spot, and w is the spot radius.
Convection heat transfer on the rear surface of the target is expressed as: According to Equations (8)- (12) can solve the temperature distribution T(x, t) of a flat window under the action of a laser [23][24][25]. Due to the non-linear changes in the actual physical parameters, the temperature field is also unstable. Therefore, the finite element calculation method can be used to calculate the temperature distribution of a flat window under the action of a laser over time.

Experimental Study of Composite Laser Damaged Flat Window
A composite laser damage effect experimental platform was set up, as shown in Figure 1. The continuous and pulsed laser triggering was controlled by a digital delay pulse generator, and the laser parameters were monitored online by a spectroscope. Before starting the experiment, we used the coaxial indicating light to fix the spot position on the flat window. When ready, we used the digital delay pulse generator to control laser delay and trigger. During the experiment, the laser status was monitored with a beam quality analyzer and a power meter. The laser irradiation time was 5 s, and the average value was recorded by repeating the experiment ten times for each power and energy value. The experimental observation system consisted of a visible high-speed camera, an infrared thermal imager, and a plasma spectrometer. We determined whether the window was damaged according to the temporal and spatial distribution of the spot collected by the plasma spectrometer. We used the photodetector to monitor the window mirror for body damage. The model of the used CW laser was of the fundamental mode with a Gaussian distribution. The model of the used pulsed laser was dominated by the fundamental mode that was mixed with some higher-order modes. The laser parameters are shown in Table 1. The targets used in the experiments were silicon-based and germanium-based flat window. The substrate material parameters are shown in Table 2.   First, two kinds of window were used to measure the absorptivity. We used a 1.06 um continuous laser to irradiate the flat window and measuring the reflected power with a power meter, since the two substrates were not transparent to the Near-Infrared band; as such, the absorptivity shown in Table 3 could be obtained.

Continuous Laser Damage Flat Window Experiment
The continuous laser was used to verify the damage effect of the out-of-band laser on the flat window. Due to thermal stress deformation caused by laser irradiation, the spot of collimated light was shifted. The thermal deformation state of the flat window under continuous laser loading was observed with a scattering detection method-that shown in Figure 2a is before loading, that shown in Figure 2b is the spot shift caused by thermal distortion during loading, and that shown in Figure  2c is the spot shift before the lens body exploded.

Figure 2. Changes in scattered light spots caused by thermal distortion. (a) Irradiation begins (b)
Start to deform (c) Body explode By comparing the experimental results, it could be seen that the film color of the germanium window changed from 600 to 660 °C, and ripples were generated (Figure 3). At 660 °C, the film began to melt, and optical coating damage occurred. At 700 °C, substrate damage began. When the laser power reached 1500 W, the window body exploded. The film damage of the silicon window began to occur at 500 °C (Figure 4), and the damage of the film was similar to that of the germanium window mirror. When the laser power reached 1440 W, the substrate of the silicon window exploded at 950 °C.  When the laser was irradiated to the flat window, due to the asymmetrical distribution of the high-power laser beam and non-uniform absorption, the temperature gradient and thermal stress in the material were caused. The larger temperature gradient and thermoelastic stress first appeared in the irradiated area, which made the window material generate thermal distortion. When the thermal stress gradually accumulated and exceeded the strength limit of the material, the irradiated area first burst, eventually causing macro damage to the window [26,27]. It can be seen from Table 4 that the damage thresholds of the NIR lasers out of band on the Ge and Si windows were close. The reason for the body damage was that the substrate absorption rate increased after the optical coating melted and the temperature rose became faster, which caused substrate bursting. The damage of the optical film system was mainly caused by melting. It could be seen that the film system disappeared and the substrate was severely ablated when observed in the central area of irradiation. This was because when continuous laser light irradiated the surface of the window mirror, the thermal stability of the substrate material was better than that of the film. The thermal stress when the substrate was heated to 550-600 °C was not enough to cause the body damage of the window, but the optical film had begun to melt and be destroyed.

Long-Pulse Laser Damage Experiment
A long-pulse laser damage flat window experiment was performed. The absorption rates of a 1.03 μm laser for flat windows were measured. The measurement results are shown in Table 5. The absorption rate of the Si window for lasers in the 1.03 μm band was higher than that of the 1.06 μm band, and the absorption rate of the Ge window for lasers in the 1.03 μm band was slightly lower than the 1.06 μm band. We performed a S-on-1 single-pulse experiment, loaded five pulses, and observed the experimental results, as shown in Table 6 below. Table 6. Experimental results of the long-pulse laser. According to Table 6, when the laser energy reached 29 J, the Si window had coating damage, and at the same time, the plasma flash produced by the breakdown of electrons appeared in the irradiation area [28]. With the increase of energy, the damage of the optic film became more serious. The damage phenomenon is shown in Figure 5 and Figure 6.When the energy reached 44 J, substrate damage occurred. When the energy reached 80 J, the first pulse caused substrate damage. When the laser energy reached 11 J, coating damage occurred in the Ge window. The coating damage became more serious at 17-23 J, and when the energy reached 29 J, the first pulse caused damage to the substrate. The long-pulse laser and continuous laser irradiation on windows were mainly caused by surface thermal fusion damage, and plasma absorption was the main cause of damage.

Short-Pulse Laser Damage Experiment
A short-pulse laser damage window experiment was performed using the 1-on-1 method. The experimental results are shown in Table 7. Observing the damage phenomenon, it was found that, unlike the long-pulse damage, the coating damage caused by the short-pulse laser was shallower than that of the long pulse. Under the experimental conditions of this paper, only the Ge window showed slight substrate damage at the laser energy of 160 mj, though it showed no body damage (Figure 7 and Figure 8). This was because the short-pulse laser damaged the material for a short time, mainly due to the effect of the laser electric field, so it mostly appeared as film damage. Non-linear effects (such as non-linear absorption, non-linear refractive index, and self-focusing effects) induced by high electric fields when the high-power lasers acted on the windows were the main causes of damage. Since the laser intensity in the short-pulse laser irradiation area reached the order of GW/cm 2 , the self-focusing effect appeared. By combining the above three experimental results, it could be seen that short-pulse lasers outside the band could only cause mild film damage, while long-pulse lasers could cause body damage.

Composite Laser Damage Experiment
According to the results of the pulse laser damage to the window, although the short-pulse laser had a higher peak power, it did less damage to the window and had difficulty damaging the substrate. Therefore, this paper used a combination of a long-pulse and a CW laser to perform the damage experiment. The trigger time of the long-pulsed laser was controlled by digital delay pulse generator. When the current temperature of the thermal imager reached a preset value, the CW laser was turned off and the pulsed laser was triggered at the same time. The results of the composite laser experiment are shown in table 8 and table 9 below. The damage phenomenon is shown in Figure 9 and Figure 10. It could be seen from the experimental results of the Ge window that the composite laser could cause slight damage to the film system at the laser energy of 2 J and 145 °C. The results of the Si window experiments showed that the composite laser started to damage the optic film system at 17 J. When the pulsed laser energy is less than 17j, the continuous laser pre-heating temperature needs to exceed 300 ℃ to cause mild film damage ,which was consistent with the long-pulse laser damage mechanism discussed above. Increasing the continuous laser preheating temperature while maintaining the pulse laser energy at 17 J led to more serious optic film damage , and the substrate damage was caused when the preheating temperature reached 260 °C.    Table 10 summarizes the damage threshold data for different lasers. The giant-pulse laser waveform used in this experiment was of the spike and square mixed waveform. The energy of the spike pulse as about three-to-four times that of the square pulse. After the flat window was pre-heated by continuous light, the spike pulse part first acted on the window body. Its high peak energy was absorbed by the optical film and substrate. The optical film was instantly melted under the double temperature rise effect, and the residual energy was directly absorbed by the substrate and eventually caused damage to the substrate.

Conclusions
(1). Out-of-band high-energy lasers can damage Ge and Si windows. NIR laser damage to Ge and Si windows with medium wave Anti-Reflection coatings occurred at 500-600 °C The damage thresholds of the NIR continuous laser and the pulsed laser to the Ge window were divided into 2 kW/cm 2 and 37 J/cm 2 , respectively, and the damage thresholds of the NIR continuous laser and pulsed laser to the Si window were divided into 1.8 kW/cm 2 and 65 J/cm 2 . The high-energy laser damage mechanism of long-pulse laser and CW laser is mainly manifested as thermal effect, and it is easier to damage the substrate. The damage mechanism of the short-pulse laser as a high-power laser was mainly manifested by the laser electric field effect, but it was difficult to damage the substrate due to the laser's low energy.
(2). The pulse-continuous composite laser could effectively improve the damaging effect, and the long-pulse laser energy had a significant effect on the damage. The thresholds of the damage to the substrate by the composite laser were 21.6 J/cm 2 for the silicon-based flat window and 3.8 J/cm 2 for the germanium-based flat window.