A comparative study of laser induced damage of HfO2/SiO2 and TiO2/SiO2 mirrors at 1064 nm has been carried out. One TiO2/SiO2 mirror with absorption of 300 ppm and two HfO2/SiO2 mirrors with absorption of 40 and 4.5 ppm were fabricated using electron beam evaporation method. For r-on-1 test, all HfO2/SiO2 mirrors with low average absorption are above 150 J/cm2 at 10ns. However, the TiO2/SiO2 mirrors with high average absorption are just 9.5 J/cm2, which are probably due to the rather high absorption and rather low band gap energy. Meanwhile, all the samples were irradiated from front and back side respectively using the raster scan test mode. In case of front side irradiation, it is found that: for TiO2/SiO2 high reflectors, the representative damage morphologies are shallow pits that were probably caused by absorbing centers. However, for HfO2/SiO2 high reflectors, the dominant damage morphologies are micrometer-sized nodules ejected pits and the delamination initiating from the pits. The absorption of HfO2/SiO2 coatings is low enough to have minor influence on the laser damage resistance. In case of backside irradiation, the morphology of TiO2/SiO2 mirrors is mainly center melted pits that are thermal melting induced damage. Meanwhile, HfO2/SiO2 mirrors with isometrical fracture rings damage morphology are thermal induced stress damage.
© 2011 OSA
Optical coatings play an indispensable role in the high power laser systems. Except for the spectral performance, another demanding requirement is the laser damage threshold which directly influences the overall performance of the whole system. Therefore, a lot of efforts had been taken to improve the damage resistance of laser coatings during the past years, especially for the mirrors working at first harmonic [1–6]. It has been well accepted that the factors mainly affecting laser induced damage threshold (LIDT) of these mirrors at 1064 or 1053 nm are isolated micro-sizes nodule defects [7-8]. However, the absorption, including average absorption and localized absorption centers, also plays an important role to induce the laser damage of the high reflectors working at first harmonic . The average absorption can be measured by surface thermal lens technique expediently , but the detail characterization and detection of localized absorption centers are still in progress for further understanding of this kind of damage behavior .
As we known, the laser induced damage (LID) mechanism can be inferred through detecting the damage morphologies, and this method can be extracted from the research of several working groups and authors [12-13]. In this present paper a comparative damage characteristic study will be carried out. A TiO2/SiO2 high reflector with absorption of 300 ppm and two HfO2/SiO2 high reflectors with absorption of 40 and 4.5 ppm were prepared by electron beam evaporation method. To investigate the influence of coating materials and whole absorption on intrinsic or quasi-intrinsic LIDT, the laser damage resistance of these coatings was comparatively studied using r-on-1 test mode, and to better understand the influence of the isolated nodule defects or nano-sized absorbers and average absorption on the damage characteristics, the raster scan test mode from front side and rear irradiation has been carried out respectively. Meanwhile, possible damage mechanisms have been proposed to interpret the observed damage morphologies.
2.1 Sample preparation
In this study, three high reflectors were investigated: one TiO2/SiO2 high reflector (Sample T1) and two HfO2/SiO2 high reflectors (Sample H1 and Sample H2). All of these high reflectors were designed for using at 1064 nm with the incidence angle of 45 degree for the s polarization. TiO2/SiO2 high reflector was prepared using the e-beam evaporation method by another vender. HfO2/SiO2 high reflectors were fabricated by reactive e-beam evaporation in our laboratory. We used a cryopumped Optorun OTFC-1300 coater that was equipped with two JEOL electron guns and an ion-source. The designs were quarter wave stack reflector including a silica halfwave overcoat, optical monitoring was used to control the film thickness. Hafnium was used as starting material for the high index material. O2 gases were filled in during evaporation. The flow rate of O2 gas for Sample H1 and Sample H2 during hafnium evaporation was 35 and 70 sccm respectively, and the flow rate of O2 gas was 20sccm for silica during this study. All the samples were deposited in the 160 centigrade circumstance and the deposition rates was 0.3 nm/s and 1nm/s for hafnium and silica respectively. The two kinds of high reflectors were deposited on conventionally polished and normally cleaned fused silica (diameter: 30 mm; thickness: 5 mm).
2.2 Laser damage test
A Nd:YAG laser operating at 1064 nm with TEM00 mode was used in our damage tests. The Q-switched pulse width is 10 ns and the maximum output energy is 2 J/pulse at 10 Hz. The modulation of output laser energy is obtained by rotating a half-wave plate that is in front of a polarizer. Samples were mounted in a computer controlled three dimensional motor-driven stage, and they were irradiated for selected pulse energies. The laser beam was focused with a 1500 mm focal lens and the beam diameter is 1mm in the 1/e2 of the laser energy. Damage was determined if any visible change is observed by an in situ 45× microscope and subsequently identified by an off-line Nomarski microscope with 200× magnification.
3. Results and discussion
Before laser damage test, the surface thermal lens (STL) technique was adopted to measure the absorptance of samples in our research firstly. We tried to find some correlation between the laser damage and optical character of these high reflectors, especially the relationship between the laser damage and absorption of the samples. Table 1 shows the absorptance of samples at 1064 nm wavelength measured with the STL technique. It is found that Sample T1 has the biggest absorption with the result of 300 ppm at 1064nm in the 3 mm × 3 mm region. While Sample H2 has the quite low absorption, all of which are below 5 ppm in the scan region. And the Sample H1 with the result of 40 ppm has the middle absorption level in this research.
Due to the rather bigger testing beam radius and the scanning region, the absorption we got from the STL technique here is the average absorption of the testing samples. And, the sample with high absorption will also have the big existing probability of much more nano-absorbers, such as sub-stoichiometric, metal clusters, or high-density electronic defect areas, and then the laser damage introduced by localized absorbers will be easily produced. Moreover, the samples with bigger absorption prone to absorb more heat during the laser irradiation, then the thermal melting induced damage and thermal mechanic damage can be triggered more easily. Therefore, it is natural to predict that the samples with big absorption will certainly have low laser damage threshold.
3.2 Laser induced damage resistance
The laser damage resistance of the above mirrors was evaluated using R-on-1 and raster scan test. R-on-1 test was taken to find the influence of coating materials and whole absorption on intrinsic or quasi-intrinsic LIDT of these high reflectors . Raster scan prototype  that was proposed by Lawrence Livermore National Laboratory is used to investigate what are the predominant precursors for triggering laser damage. Moreover, the initial damage morphology and sequential damage growth procedure during raster scan test was carefully studied to infer the possible damage mechanism.
3.2.1 R-ON-1 Laser damage threshold
As for the r-on-1 measurement, nearly fifteen sites are tested, the damage probability has been defined as the ratio of the happened catastrophic damage sits among all test sites under this laser fluence. And the energy density is adjusted by rotating the half-wave plate gradually. The increment is about 0.2J/cm2 per pulse throughout the tests until the catastrophic damages happen.
Form the test results shown in Fig. 1 , we can find that the average damage threshold of sample T1 is only 9.5 J/cm2, meanwhile, sample H1 and H2 are 155.8 J/cm2 and 160.3 J/cm2 respectively, all of them are rather high than TiO2/SiO2 high reflector. One apparent reason for this huge difference of LIDT is that the bigger difference between TiO2/SiO2 and HfO2/SiO2 high reflectors. It is well known that the main damage mechanism in laser coating working at 1 μm is the thermal-mechanical damage caused by absorption, and then the high absorption of TiO2/SiO2 high reflector directly leads an abrupt decrease in laser damage threshold. Moreover, the band gap energy of TiO2 is only 3.3ev, which is much lower than 5.1ev of HfO2 . The lower band gap energy of TiO2 is more vulnerable to high intensity photons, this also contribute partly to the very low LIDT of TiO2/SiO2 high reflectors. Future work will study the LIDT of low absorbing TiO2/SiO2 high reflectors to distinguish the influence from absorption and band gap energy.
Although sample H1, with an average absorption of 40 ppm, has a higher absorption than sample H2, but they almost have the same damage threshold in our test. We are inclined to think that the average absorption of 40 ppm in HfO2/SiO2 high reflectors working at 1064 nm is small enough not to be a limiting factor to decrease LIDT.
3.2.2 Laser damage characteristic under Raster Scan
The damage morphologies of HfO2/SiO2 mirrors irradiated with nanoseconds laser pulse are commonly nodule ejected pits, flat bottom pits and delaminates [17-18], which are corresponding to different damage initiations and damage mechanisms. After r-on-1 testing, we also investigate the damage characteristics of these samples using raster scan method. A random 10mm × 10mm area has been tested for these samples. And the main attention has been paid to investigate the damage morphology of these samples, through which we can analyse and compare the damage precursor and the damage mechanism of these reflectors with different absorptions.
184.108.40.206. Raster scan form front surface
For sample T1, the laser irradiation fluence started from 2 J/cm2, and increased with a 0.5 J/cm2 increment until the damage happened at 7.5 J/cm2; but for sample H1 and sample H2, the irradiation fluence begun at 20 J/cm2, and increased with a 5 J/cm2 increment. Finally, sample H1 resisted the laser irradiation up to 115 J/cm2, while sample H2 shew a catastrophic damage at 120 J/cm2. The detail result of the damage threshold tested with raster scan method of these samples has been listed in Table 1.
During test of sample T1 (TiO2/SiO2 mirrors), there was no visible change of the coating surface when examined using Nomarski microscope with 200× magnification before the laser fluence increasing to 6 J/cm2. But at the fulence of 7.5 J/cm2, the laser energy was sufficient enough to create damage pits. As shown in Fig. 2 , the main damage morphologies of Sample T1 are shallow pits with the diameter about 20 um (as shown in Fig. 2 (a)), and there are many tiny pits and ejecting materials around the big round shallow pit (as shown in Fig. 2 (b)), which indicates there are so many different sized initial damage precursors exited in the coatings. And then, from the SEM and FIB picture shown in Fig. 2 (c) and Fig. 2 (d) we can find that except some fragmentary material left in the middle of the shallow pits, almost the whole outmost layer had been removed.
There are no visible damage precursors that can be found in the bottom of pits, the damage of TiO2/SiO2 coatings was probably caused by isolated invisible absorbing centers liking sub-stoichiometric oxides or high density electrical defects. Since the electric field of HR in the outmost several layers is relatively high, the absorbing centers in these layers especially in interface are more probable to induce laser damage. The material surrounded the absorber can be melted or fractured off by the plasma triggered by the laser irradiation on these absorbing centers. Except the contribution of the random distributed localized invisible absorbing centers to the damage formation, the big average absorption also aggravates the melting of TiO2/SiO2 outmost material.
Unlike sample T1, the damage morphologies of sample H1 and sample H2 are almost same, both are mainly nodules ejected pits and delaminates, as shown in Fig. 3 . In sample H1 and H2, there are no shallow pits found during the inspection using Nomarski microscope. Usually, we first find micrometer-sized nodules ejected pits, liking Fig. 3 (a). And then plasma scalds near the nodules ejected pits can be found, which are towards the laser beam incidence direction. Finally, delaminates can be inspected before the following fluence increases to catastrophic damage threshold, and the delaminate zone lies behind the pits, as shown in Fig. 3 (b), which has the same direction as the scalds. The presence of surface ripple delaminates (Fig. 3 b) suggests shock waves or plasma interference melted into the surface during the damage event. It has been found that, for HfO2/SiO2 reflectors (sample H1 and H2) with low absorption, the damage precursors are mainly micro-size nodules, and the absorption is low enough to have minor influence on the laser induced damage in this study.
As discussed above, the raster scan from coating side is an effective method to investigate the initial damage precursor and to find out what is the limited factor to LIDT. For HfO2/SiO2 mirrors, the visible nodule defects are the main cause to trigger the laser damage rather than the absorption, the absorption of 40ppm and 4ppm have the equivalent attribution to the laser damage mechanism, which are all can be ignored in HfO2/SiO2 mirrors. But for TiO2/SiO2 high reflectors, the invisible absorbing centers are the dominate reason to induce laser damage inferring from the damage morphology. Due to the lower band gap energy and high average absorption, the high density of absorbing centers directly introduce thermal melting damage at low damage threshold, at which the nodule defects cannot be triggered yet. However, this front irradiation mode can only extrapolate the occurrence probability of absorbing centers in the coatings rather than the contribution of average absorption to the laser damage. To further illustrate the influence of average absorption on damage characteristic of these mirrors more, all the samples were also raster scan tested with laser irradiation from the rear side.
220.127.116.11. Raster scan form rear side
For sample T1, the laser irradiation fluence started from 1 J/cm2, and increased with a 0.5 J/cm2 increment until the damage happened at 4 J/cm2; but for sample H1 and sample H2, the irradiation fluence begun at 5 J/cm2, and increased with a 5 J/cm2 increment. Finally, both sample H1 and H2 resisted the laser irradiation up to 30 J/cm2. The detail result of the damage threshold tested with raster scan method of these samples has been listed in Table 1.
Figure 4 shows the typical damage morphology of theses TiO2/SiO2 mirrors under rear irradiation, there are no visible precursors that trigger damage in the pits penetrating to the bottom of the coating. This observation indicates that the laser damage is interfacial damage between the coating and substrate, and the damage is initiated from invisible nano-sized absorbers that may be imbedded in coating material or coating and substrate interface.
The model describing laser damage initiating from nano-sized absorbers has been proposed early [11, 19]. When an absorber is exposed to a laser beam of enough fluence, it will create a strong plasma ball. The intense plasma can melt the surrounding materials, induce highly tensile stress, and cause the coating layers to delaminate and fracture off from the substrate. Especially to the high absorbing TiO2/SiO2 coatings, the absorbed heat generating from plasma ball is big enough to melt the material, which can be proved by the cross-section microscope of sample T1 (Fig. 4 b): Firstly, the smooth sidewall melted zone along the few bottom layers can be clearly seen, then follows the bigger and bigger fracture ring along the upward direction. Secondly, there is an obvious melted zone in the substrate. At the same time, the rather low melting point of TiO2 and SiO2 also increase the possibility of material melting. Therefore, to these TiO2/SiO2 mirrors, all discussed above can prove that the main damage mechanism is thermal melting damage caused by the high average absorption.
Figure 5 describes the representative laser damage morphologies of HfO2/SiO2 mirrors under rear irradiation, the morphologies captured by Nomarski and SEM microscopy are similar to TiO2/SiO2 ones, both the delamination happened from the interface between substrate and coating. The circular geometry of the coating damage pits, which is absent with melted zone but exiting in TiO2/SiO2 mirrors, indicate that the damage are mainly induced by nano-absorbers embedded in substrate rather than in coating material. Additionally, combined with low average absorption and high melting point of coating material, the damage morphology with isometrical fracture rings proves that the absorbing heat from plasma is not so big as to melt the surrounding material. Hence, the main damage mechanism of HfO2/SiO2 mirrors is thermal induced stress damage.
In summary, it is an effective method to use the raster scan testing from rear irradiation to investigate the influence of average absorption on damage morphology and damage characteristics. And the damage mechanism of TiO2/SiO2 mirrors are thermal melting induced damage, otherwise, damage cause of HfO2/SiO2 high reflectors are thermal induced high tensile stress fracture.
The absorption of three different high reflectors working at 1064nm was investigated; meanwhile, the laser damage performance of these samples was also analyzed. Table 1 shows the measurement results.
The influence of coating materials and whole absorption on intrinsic or quasi-intrinsic LIDT of these high reflectors has been investigated by r-on-1 test. Then the raster scan test mode from front surface and rear side respectively has been carried out to investigate the influence of the isolated nodule defects or nano-sized absorbers and average absorption on the damage characteristic. For high absorbing TiO2/SiO2 high reflectors, damages are always found when the laser fluence is as low as 9.5 J/cm2 for the r-on-1 testing mode, it is the result of rather high absorption and low band gap energy. And the main reason of damage from front side of TiO2/SiO2 mirrors under raster scan is isolated absorbing centers, while the damage morphologies from rear irradiation are high average absorption induced center melted pits, which indicates that the damage mechanism of TiO2/SiO2 mirrors irradiated from backside are thermal melting induced damage. For HfO2/SiO2 high reflectors, the average absorption less than 40 ppm is low enough to get a high laser induced damage threshold, which is more than 150J/cm2 using r-on-1 test method. And the dominant damage morphologies under raster scan mode from front surface are micro-sized nodules ejected pits and delaminates, it can be inferred that the visible nodule defects are the main cause to trigger the laser damage rather than the absorption. Meanwhile, the damage morphology with isometrical fracture rings from rear irradiation proves that the main damage mechanism of HfO2/SiO2 mirrors is thermal induced stress damage.
Authors wish to thank Jiangtao Lu for technical assistance in STL measurements. And also acknowledge support from the Program 863 of China, National Natural Science Foundation of China (No. 61008030).
References and links
1. M. R. Borden, J. A. Folta, C. J. Stolz, J. R. Taylor, J. E. Wolfe, A. J Griffin, M. D. Thomas, “Improved method for laser damage testing coated optics”, Proc. SPIE 5991, 59912A1 (2005)
2. F. Y. Genin and C. J. Stolz, “Morphologies of laser-induced damage in hafnia-silica multilayer mirror and polarizer coatings,” Proc. SPIE 2870, 439–448 (1996). [CrossRef]
3. M. R. Kozlowski, “Ri. Tench, R. Chow, L. Sheehan, “Influence of defect shape on laser induced damage in multilayer coatings,” Proc. SPIE 2253, 743–750 (1994). [CrossRef]
4. J. C. Stolz, J. Adams, M. D. Shirk, M. A. Norton, and T. L. Weiland, “Engineering meter-scale laser resistant coatings for the near IR”, Proc. SPIE 5963, 59630Y1–9 (2005)
5. J. Christopher, Stolz, Michael D. Thomas and Andrew J. Griffin, “BDS thin film damage competition”, Proc. SPIE 7132, 71320C1 (2008)
6. J. H. Campbell, R. A. Hawley-Fedder, C. J. Stolz, J. A. Menapace, M. R. Borden, P. K. Whitman, J. Yu, M. Runkel, M. O. Riley, M. D. Feit, and R. P. Hackel, “NIF optical materials and fabrication technologies: An overview,” Proc. SPIE 5341, 84–101 (2004). [CrossRef]
7. M. R. Kozlowski and R. Chow, “The role of defects in laser damage of multilayer coatings,” Proc. SPIE 2114, 640–649 (1994). [CrossRef]
8. C. J. Stolz, R. J. Tench, M. R. Kozlowski, and A. Fornier, “A comparison of nodular defect seed geometries from different deposition techniques,” Proc. SPIE 2714, 374–382 (1995). [CrossRef]
9. S. Papernov, and A. W. Schmid, “Laser-induced surface damage of optical materials: Absorption sources, initiation, growth, and mitigation”, Proc. SPIE 7132, 71321J1 (2008)
11. J. Dijon, T. Poiroux, and C. Desrumaux, “Nano absorbing centers: a key point in laser damage thin films,” Proc. SPIE 2966, 315–325 (1997). [CrossRef]
12. L. Jensen, M. Jupé, and D. Ristau, “UV damage mechanisms in oxide high reflectors,” Proc. SPIE 7132, 71320G, 71320G-12 (2008). [CrossRef]
13. S. papernov, A. W. Schmid, J. B. Oliver, and A. L. Rigatti, “Damage threshold and morphology of the front- and back-irradiated SiO2 thin films containing gold nanoparticles as artificial absorbing defects,” Proc. SPIE 6720, 67200G (2007). [CrossRef]
14. J. Hue, P. Garrec, J. Dijon, and P. Lyan, “R-on-1 automatic mapping: a new tool for laser damage testing,” Proc. SPIE 2714, 90–101 (1996). [CrossRef]
15. M. R. Borden, J. A. Folta, C. J. Stolz, J. R. Taylor, J. E. Wolfe, A. J. Griffin, and M. D. Thomas, “Improved method for laser damage testing coated optics”, Proc. SPIE 5991, 59912A1 (2005)
16. M. Mero, J. Liu, W. Rudolph, D. Ristau, and K. Starke, “Scaling laws of femtosecond laser pulse induced breakdown in oxide films,” Phys. Rev. B 71(11), 115109 (2005). [CrossRef]
17. F. Y. Genin, C. J. Stolz, and M. R. Kozlowski, “Growth of laser-induced damage during repetitive illumination of HfO2-SiO2 multilayer mirror and polarizer coatings,” Proc. SPIE 2966, 273–282 (1997). [CrossRef]
18. J.-Y. Natoli, F. Wagner, A. Ciapponi, S. Palmier, L. Gallais, and M. Commandré, “Non destructive evaluation on optical components for high power density applications,” Proc. SPIE 7101, 710118 (2008). [CrossRef]
19. X. Cheng, Z. Shen, H. Jiao, J. Zhang, B. Ma, T. Ding, and Z. Wang, “Laser damage resistance of dichroic mirrors at 532 nm and 1064nm,” Proc. SPIE 7842, 78420C, 78420C-8 (2010). [CrossRef]