Damage tests are carried out at 1064nm to measure the laser resistance of TiO2/Al2O3and HfO2/Al2O3 antireflection coatings grown by atomic layer deposition (ALD). The damage results are determined by S-on-1 and R-on-1 tests. Interestingly, the damage performance of ALD coatings is similar to those grown by conventional e-beam evaporation process. A decline law of damage resistance under multiple irradiations is revealed. The influence of growth temperature on damage performance has been investigated. Result shows that the crystallization of TiO2 layer at higher temperature could lead to numerous absorption defects that reduce the laser-induced damage threshold (LIDT). In addition, it has been found that using inorganic compound instead of organic compound as precursors for ALD process maybe effectively prevent carbon impurities in films and will increase the LIDT obviously.
© 2012 OSA
In high-power laser systems there are thousands of large-aperture (>400mm) coated optics, such as mirrors and polarizers . To meet the requirement, optical coatings must be fabricated with excellent uniformity, high accuracy and high laser damage resistance. Usually, dielectric thin films are mainly deposited by conventional electrical beam evaporation process. Although the optimization of e-beam process has been developed for decades, it is still very difficult to achieve super-high uniformity and accuracy over large-scale optics. Moreover, in traditional coating process, the complicated operation and unfavorable environmental control may cause some structural or stoichiometric defects in film and probably induce damage precursors, such as nodular and nano-scale size absorbers, which are the most important factors decreasing coatings damage resistance. Therefore, it is very important to develop an improved coating method to settle those problems. Atomic layer deposition (ALD) technique is a chemical gas phase coating deposition process which has been developed in nanotechnology [2,3] and semiconductor industry for nearly two decades . The unique saturated surface reaction during deposition process gives ALD several attractive features, such as accurate thickness control, high uniformity, perfect reproducibility, etc. Such excellent interesting advantages make ALD a rather promising technique in large-aperture optical coatings manufacture for high-power laser system .
In previous works, the optical properties of ALD optical coatings have already been studied extensively [6–10]. However, as another important aspect, the laser damage resistance hasn’t been thoroughly studied yet [11,12]. In this work, TiO2/Al2O3 and HfO2/Al2O3 antireflection coatings are prepared by ALD process and their damage performance is investigated via S-on-1 and R-on-1 damage tests. X-ray diffraction (XRD), atomic force microscope (AFM) and scanning electron microscope (SEM) are employed to analyze damage morphology and the damage mechanism of ALD coatings is discussed. The impact of growth temperature and carbon impurities on damage resistance of ALD coatings is also discussed in this work.
2. Experiment procedure
The deposition parameters of ALD coatings are listed in Table 1 . Taken Al2O3/ TiO2 antireflection (AR) coatings as an example, the designed thickness is 218.59 nm for Al2O3 layer and 39.06 nm for TiO2. The pulsing durations are 0.1s for Al (CH3)3(TMA) and 4s for TiCl4 in one cycle. The growth rate is 0.1nm/cycle for Al2O3 and 0.4nm/cycle for TiO2. Ar2 and N2 are chosen as purge gas for Al2O3 and TiO2, respectively. The purging time after each pulse is set to 1s with a flow rate of 200 sccm. In this work, Different deposition temperatures and precursors, organic or inorganic compounds, are chosen to investigate the influence of growth temperature and carbon impurities on damage performance of ALD coatings. For comparison, a couple of samples are prepared by e-beam evaporation process with the growth temperature of 220°C.
Experimental investigations are carried out on small optics damage test facility. The laser characteristics have been presented in detail elsewhere . In brief, the beam (5.5ns FWHM) generated from an Nd: YAG laser is lens-focused to form a spot with 0.67mm 1/e diameter on the sample plane. In our experiment, the incidence angle at sample surface is slightly off (3~5 degrees) normal so as to avoid additional affection from the exit surface reflection.
For S-on-1 tests, over 150 sites are chosen with an interval of 2mm on each sample. Each site will be irradiated with 1000 pluses at certain energy and its surface status will be recorded by online detection unit. For R-on-1 test, a series of laser pulses whose energy varies from low to high energy will irradiate on 50 sites of each sample until damage occurs. The initial fluence is set at 5% of single-pulse LIDT and the increment is about 0.5%. Every site will be examined by Nomarski microscope after irradiation to check if damage happens or not. SEM and AFM are employed to get more details about damage morphology and depth distribution, for a further understanding of damage mechanism.
3. Results and discussion
3.1 Damage performance of TiO2/Al2O3 coatings
Al (CH3) 3(TMA) and TiCl4 are chosen as precursors in ALD process for Al2O3 and TiO2. Before damage tests, the transmission and refractive index of all samples are measured and the result is show in Table 2 . The transmission of TiO2/Al2O3 AR coatings is over 99.8% at 1064nm. The refractive index of TiO2 and Al2O3 will increase along with growth temperature. Other optical properties have been presented in detail in ref . The results of single-pulse damage test are shown in Fig. 1 . It is interesting to find that ALD sample grown at 110°C has the highest LIDT, about 6.1J/cm2. Moreover, the results also indicate that the growth temperature could remarkably impact laser damage resistance. Actually samples grown at 110°C have a LIDT of 6.1J/cm2, and that is obviously higher than those grown at 280°C, about 3.2J/cm2. For a further understanding of this phenomenon, the 1-on-1 damage test has also been taken on single layer of Al2O3 and TiO2 under different growth temperature (Fig. 2 ). It could be immediately found that the LIDT of Al2O3 single layer is on the whole relatively higher than TiO2. It may be relative to their intrinsic properties. For example, the narrower band gap  makes TiO2 easier to absorb laser photons and more vulnerable than Al2O3 under higher intensity laser intensity . However, it is expectable to find that both Al2O3 and TiO2 single layer show the same trend: higher temperature leads to lower LIDT, with different decline amplitude.
One probable explanation for this phenomenon is the crystallization in film. The evidence of crystallization can be easily found by AFM image of TiO2 single layer surface shown in Fig. 3 . It can be obviously found that for the samples grown at 280°C, there are lots of grains with the scale of 100~150nm on surface and the roughness is about 1.851nm. However, for the samples grown at 110°C, no grains can be observed and the surface becomes smoother with a roughness of 0.415nm, rather close to the roughness of substrate. It is considered that the crystallization of TiO2 films at higher temperature will reduce film homogeneity. And the inhomogeneity particles or clusters formed by partial crystallization could play a role of potential damage sources, such as nano-scale absorbers [15,16], which may increase laser energy absorption and finally lead to the LIDT reduction. In addition, the results of XRD analysis in our previous study  further confirm the existence of crystallization at higher temperature. It has been found that the TiO2 film grown at 280°C shows a tetragonal crystal phase whereas at 110°C it’s still amorphous. However, for Al2O3 film, there is no evidence of crystallization and the reason of LIDT decrease with temperature increase is still unclear.
The TiO2/Al2O3 AR coating damage morphology has been investigated via Nomarski microscope, SEM and AFM (Fig. 4 ) to understand the damage mechanism of ALD coatings. The result indicates that the dominated damage morphology is craters with the diameter of 4~16um. Further AFM analysis shows that bottom of crater is relatively smooth and no visible damage sources could be found. It can be probably inferred from that the damage is relative to the temporal expansion which is triggered by some invisible absorbers, such as stoichiometric mismatch, crystalline clusters, high density electrical defects, etc. Crater depth distribution (Fig. 5 ) obtained by AFM clearly shows that most of craters are 240~270nm in depth, just the same to the film thickness. It probably means that the damage originates in TiO2 layer and then expands to Al2O3 layer. Therefore, it is considered that the laser damage performance of TiO2 layer plays a crucial role, and reducing growth temperature will effectively increase the LIDT of TiO2/Al2O3 coatings.
Laser damage resistance of TiO2/Al2O3 coatings under multi-pulse irradiation has been tested by S-on-1 and the result indicates that the LIDT of all samples will decrease with pulse number following an exponential decay rule, shown in Fig. 6 . This phenomenon has also been observed in various optical materials in previous studies [17–20] and there is an empirical formula below used to fit the experiment data
F (N) is the LIDT while N is the pulse number and Φ∞, A and B are the fitting parameters. In some case this formula can be used for roughly estimating the lifetime of thin films under multi-pulse irradiation [21,22] and the fitting result is also shown in Fig. 6. After a huge number of pulse irradiation, the LIDTs of TiO2/Al2O3 coatings will steady at 3.61J/cm2 and 1.14J/cm2 for growth temperature 110°C and 280°C, respectively. It is found that the LIDT of ALD coatings grown at 110°C is rather closed to e-beam coatings at beginning. However, with the growth of pulse number, the LIDTs of e-beam coatings decrease more rapidly than ALD ones. The experiment result shows that the LIDTs of ALD and e-beam coatings will reach a steady value of 3.6±0.32J/cm2 and 4.0±0.36J/cm2 after 1000-pulse irradiation, respectively.
In order to test the laser conditioning effect on ALD antireflection coatings, over 100 sites on each sample are chosen for R-on-1 test. The initial pulse fluence is set at 5% of single-pulse LIDT and the increment is about 0.1J/cm2 per pulse. The results are shown in Fig. 7 . It could be immediately found that the coatings grown by ALD have the same or even better laser conditioning effect than those grown by e-beam. Interestingly, several sites on ALD sample have rather high laser damage resistance, for instance, up to over 30J/cm2. This is probably attributed to the excellent performance of ALD on film defects mitigation. More work is needed in the future for quantitative characterization of defect distribution of ALD film, and further obtaining the relationship between laser conditioning effect and their defects.
3.2 Damage performance of HfO2/Al2O3 coatings
Generally speaking in ALD process, there could be several kind of precursor material for a certain film. For example, HfCl4 can be used instead of Hf (N (CH3) (C2H5)) 4 (TEMAH) to grow HfO2 film. Different precursor combination may result in different damage performance of thin films. Therefore in this section, organic and inorganic precursors listed in Table 1 have been used to grow HfO2/Al2O3 coatings and their damage behaviors have been discussed. For comparison, a couple of samples were prepared by e-beam evaporation process with the growth temperature of 220°C. The grown rate is 0.45nm/sec for HfO2 and 0.50nm/sec for Al2O3 film. The transmission and refractive index of all samples are listed in Table 3 .
The single-pulse damage tests are first performed and the results are shown in Fig. 8 . Here the samples grown with TEMAH are noted as “OP” short for organic precursor while “IP” represents those grown with inorganic precursor, HfCl4. The result shows that the 1-on-1 LIDT of OP sample is only 4.2J/cm2, much lower than IP sample with 7.4J/cm2 and e-beam sample with 8.8J/cm2. The key point of such difference in LIDT is possibly the carbon impurities in film. In ALD process, TMA or TEMAH reacts with steam to produce alumina, hafnia and methane. Methane should be drawn away by purge gas immediately. But unfortunately this process cannot always take away all methane. In that case there could remain a few carbon compounds remain in film. The intense absorption at 1064nm by carbon impurities will make the film more susceptible to damage. Therefore, e-beam samples have higher LIDT than those grown by ALD using organic precursors. In order to separate the influence of carbon impurities on different material, the HfO2 and Al2O3 single layer grown with organic and inorganic precursor have also been tested. HfCl4 and TEMAH are chosen as precursors for HfO2 film, while AlCl3 and TMA for Al2O3 film. The results (Fig. 9 ) show that the influence of carbon impurities on LIDT for Al2O3 and HfO2 film is rather different. It can be noted that there is a slight drop of LIDT from 13.4J/cm2 to 12J/cm2 for Al2O3 single layer while using organic precursor. And for HfO2 film, the LIDT decreases clearly from 45J/cm2 to 14.4J/cm2. The different decrease amplitude of LIDT for Al2O3 and HfO2 may be relevant to their intrinsic material properties. However, it is difficult to determine those differences since some important parameters are still unknown, such as carbon content in film.
Crater is the main morphology of damage sites on Al2O3/HfO2 antireflection coatings, shown in Fig. 10 . It also can be called as pinpoints for its high density. Some spumous structure around the crater indicates that there could be a thermal explosion process during the laser irradiation. Liking TiO2/Al2O3 AR coatings, the bottom of crater is so smooth that no visible damage precursors such as nodule ejection can be observed. With the crater-depth distribution achieved by AFM shown in Fig. 11 , it can be noticed that most of craters have a depth between 180nm and 200nm, just smaller than the thickness of Al2O3 layer. Therefore, it indicates that the damage of Al2O3/ HfO2 coatings must originate in Al2O3 layer, and improving the damage resistance of Al2O3 will directly increase the LIDT of Al2O3/ HfO2 coatings.
In order to test the laser conditioning effect, R-on-1 damage tests have been performed on HfO2/Al2O3 AR coatings grown by ALD and e-beam. From result (Fig. 12 ), it can be immediately found that the R-on-1 average LIDT of e-beam sample is about 38.7J/cm2, which is over twice as high as ALD samples with only 14~16J/cm2. Although HfCl4 has been used instead of TEMAH to reduce the content of carbon impurities, there may still some carbon left in film since the precursor for Al2O3 is always TMA in HfO2/Al2O3 AR coatings in our experiment. The carbon impurities could be the key factor causing laser conditioning effect decrease. Therefore, it is believed that the LIDT and laser conditioning effect will improve by using fully inorganic precursors in future.
In summary, laser damage resistance of TiO2/Al2O3 and HfO2/Al2O3 antireflection coatings grown by ALD has been thorough tested in this work. Results show that the damage performance of TiO2/Al2O3 coatings grown by ALD is rather close to those grown by e-Beam evaporation. However for HfO2/Al2O3 coatings, e-beam samples seem have higher damage resistance than ALD ones. An exponential decline law of LIDT has been observed under multi-pulse irradiation, and an empirical formula is used to estimate the lifetime of thin films. Further, the influence of growth temperature on damage performance has been investigated. The results indicate that the LIDT of TiO2/Al2O3 coatings will decrease along with the temperature. This Phenomenon could be related with the partial crystallization of TiO2 film over 250°C. The influence of carbon impurities on damage performance was also studied. Results indicated that choosing HfCl4 instead of TEMAH as precursor will increase the LIDT of HfO2/Al2O3 coatings. The possible reason is that the carbon impurities in HfO2 layer will increase the absorption at 1064nm and make film more susceptible to damage. Using inorganic precursors could make ALD coatings better damage performance.
This work was performed under the auspices of China Academy of Engineering Physics. And we acknowledge Picosun corp., Beneq corp. and Nanjing University for their help in sample preparation. The authors also want to thank Dr. S. Motokoshi for constructive discussion.
References and links
1. E. I. Moses, J. H. Campbell, C. J. Stolz, and C. R. Wuest, “The national ignition facility: the world’s largest optics and laser system,” Proc. SPIE 5001, 1–15 (2003). [CrossRef]
2. M. Ritala and M. Leskelä, “Atomic layer epitaxy—a valuable tool for nanotechnology?” Nanotechnology 10(1), 19–24 (1999). [CrossRef]
3. J. Maula, “Atomic layer deposition (ALD) for optical nanofabrication,” Proc. SPIE 7591, 75910S, 75910S-15 (2010). [CrossRef]
4. O. Sneh, R. B. Clark-Phelps, A. R. Londergan, J. Winkler, and T. E. Seidel, “Thin film atomic layer deposition equipment for semiconductor processing,” Thin Solid Films 402(1-2), 248–261 (2002). [CrossRef]
5. S. Zaitsu, S. Motokoshi, T. Jitsuno, M. Nakatsuka, and T. Yamanaka, “Large-area optical coatings with uniform thickness grown by surface chemical reactions for high-power laser applications,” Jpn. J. Appl. Phys. 41(Part 1, No. 1), 160–165 (2002). [CrossRef]
7. A. Szeghalmi, M. Helgert, R. Brunner, F. Heyroth, U. Gösele, and M. Knez, “Atomic layer deposition of Al2O3 and TiO2 multilayers for applications as bandpass filters and antireflection coatings,” Appl. Opt. 48(9), 1727–1732 (2009). [CrossRef]
8. S. S. Kim, N. T. Gabriel, W. B. Song, and J. J. Talghader, “Encapsulation of low-reflective-index SiO2 nanorods by Al2O3 with atomic layer deposition,” Opt. Express 15(24), 16285–16291 (2007). [CrossRef]
9. T. Pilvi, M. Ritala, M. Leskelä, M. Bischoff, U. Kaiser, and N. Kaiser, “Atomic layer deposition process with TiF4 as a precursor for depositing metal fluoride thin films,” Appl. Opt. 47(13), C271–C274 (2008). [CrossRef]
10. N. B. Abaffy, P. Evans, G. Triani, and D. McCulloch, “Multilayer alumina and titania optical coatings prepared by atomic layer deposition,” Proc. SPIE 7041, 70419 (2008).
11. Y. W. Wei, H. Liu, O. Y. Sheng, Z. C. Liu, S. L. Chen, and L. L. Yang, “Laser damage properties of TiO2/Al2O3 thin films grown by atomic layer deposition,” Appl. Opt. 50(24), 4720–4727 (2011). [CrossRef]
12. S. Zaitsu, S. Motokoshi, T. Jitsuno, M. Nakatsuka, and T. Yamanaka, “Laser damage properties of optical coatings with nanoscale layers grown by atomic layer deposition,” Jpn. J. Appl. Phys. 43(3), 1034–1035 (2004). [CrossRef]
13. 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]
15. 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]
16. S. Papernov and A. W. Schmid, “Laser-induced surface damage of optical materials: Absorption sources, initiation, growth, and mitigation,” Proc. SPIE7132, 71321J (2008).
17. A. Ciapponi, P. Allenspacher, W. Riede, J. Herringer, and J. Arenberg, “S on 1 testing of AR and HR designs at 1064nm,” Proc. SPIE 7842, 78420J, 78420J-6 (2010). [CrossRef]
18. L. Gallais, J. Y. Natoli, and C. Amra, “Statistical study of single and multiple pulse laser-induced damage in glasses,” Opt. Express 10(25), 1465–1474 (2002). [PubMed]
19. A. Melninkaitis, D. Miksys, R. Grigonis, V. Sirutkaitis, D. Tumosa, G. Skokov, and D. Kuzma, “Multiple pulse laser-induced damage of antireflection coated lithium triborate,” Proc. SPIE 5963, 59631I, 59631I-8 (2005). [CrossRef]
20. M. Mero, L. A. Emmert, and W. Rudolph, “The role of native and photoinduced defects in the multi-pulse subpicosecond damage behavior of oxide films,” Proc. SPIE 7132, 713209, 713209-10 (2008). [CrossRef]
21. J. W. Arenberg, “Life testing for laser optics: a first look,” Proc. SPIE 7504, 7504I (2009).
22. J. W. Arenberg, W. Riede, A. Ciapponi, P. Allenspacher, and J. Herringer, “An empirical investigation of the laser survivability curve,” Proc. SPIE 7842, 78421B, 78421B-8 (2010). [CrossRef]