characterization of zirconia– and niobia–silica mixture coatings produced by ion-beam sputtering

Characterization of zirconia– and niobia–silica mixturecoatings produced by ion-beam sputtering

Andrius Melninkaitis,1,* Tomas Tolenis,1,3 Lina Mažulė,1 Julius Mirauskas,1

Valdas Sirutkaitis,1 Benoit Mangote,2 Xinghai Fu,2 Myriam Zerrad,2

Laurent Gallais,2 Mireille Commandré,2 Simonas Kičas,3

and Ramutis Drazdys3

1Laser Research Center, Vilnius University, Saulėtekio al. 10, 10223 Vilnius, Lithuania2Institut Fresnel, CNRS, Aix-Marseille Université, Ecole Centrale Marseille,

Campus de St Jérôme, 13013 Marseille, France3State Research Institute for Physical Sciences and Technology, Savanoriu¸ pr. 231, 02300 Vilnius, Lithuania

*Corresponding author: [email protected]

Received 30 July 2010; accepted 13 September 2010;posted 29 October 2010 (Doc. ID 132641); published 10 December 2010

ZrO2–SiO2 andNb2O5–SiO2 mixture coatings aswell as those of pure zirconia (ZrO2), niobia (Nb2O5), andsilica (SiO2) deposited by ion-beam sputtering were investigated. Refractive-index dispersions, bandgaps,and volumetric fractions of materials in mixed coatings were analyzed from spectrophotometric data.Optical scattering, surface roughness, nanostructure, and optical resistance were also studied. Zirconia–silicamixtures experience the transition from crystalline to amorphous phase by increasing the content ofSiO2. This also results in reduced surface roughness. All niobia and silica coatings and theirmixtureswereamorphous. The obtained laser-induced damage thresholds in the subpicosecond range also correlateswith respect to the silica content in both zirconia– and niobia–silica mixtures. © 2010 Optical Societyof AmericaOCIS codes: 140.3330, 160.4670, 260.2065, 310.1620, 310.3840, 310.6860.

1. Introduction

Thin film coatings are widely used in many high-techapplications as a convenient way to improve optical,electrical, thermal, or mechanical properties of sur-faces. In laser optics applications, typically a stackdesign of periodically alternating low- and high-refractive-index dielectric layers of λ=4 thickness isused. However, rapid development of femtosecond la-ser technology requires much more complex coatingshaving more advanced design. Because of technologi-cal limitations, a very complex design can hardly beachievable by conventional low-energy evaporationtechniques, such as thermal or electron beam evap-oration. In addition, low optical scattering, controlla-

ble dispersion, narrow spectral width (in bandpassfilters), and high laser-induced damage threshold(LIDT) are also desirable features of the coatings forlaser technology applications. High-energy ion-beamsputtering (IBS) deposition is known to fulfill de-manding requirements and result in high opticalquality [1,2]. Moreover, recent progress in IBS co-sputtering process development also allows efficientdeposition of mixture coatings of different oxide ma-terials [3]. The idea of material mixing is attractivedue to the flexibility in obtaining materials with tai-lored optical constants. In practice, advantages ofmixed films also have been demonstrated: the in-crease in damage threshold [4,5], reduced opticallosses [6], and reduced internal stress [7]. Further-more, the usage of the IBS technique for materialmixing in combination with high-precision thick-ness and refractive-index control [8,9] opens up

0003-6935/11/09C188-09$15.00/0© 2011 Optical Society of America

C188 APPLIED OPTICS / Vol. 50, No. 9 / 20 March 2011

completely new capabilities. For example, rugate fil-ters [8,10,11] and broadband antireflection coatings[12] can be produced. Many alternative techniquesalso exist for the production of mixed material coat-ings, which are not the subject of the present study.For example, coevaporation from two sources [13,14],evaporation from solid solutions [15], chemical vapordeposition [16], and solgel mixing [17] can be used forproduction of mixed coatings. If the wavelength ofthe electromagnetic radiation is much larger thanthe particle size, the dielectric response of such com-pound materials can be characterized by applying so-called effective-medium approximation (EMA) [18].Some of the fundamental aspects of the EMA theoryof refractive index of mixed component materialshave been discussed by Aspnes [19]. EMA has alsobeen applied in our study for production and inter-pretation of metal oxide mixtures. Zirconia, niobia,and silica are the materials of major importancefor optical technology in the visible spectral region.For example, zirconia and niobia are widely usedfor the production of high refractive-index thin films,while SiO2 is frequently used as a low-refractive-index material. Some properties of zirconia–silicamixtures prepared with the IBS technique [7], e-beam deposition [20,21], and magnetron sputtering[22] were already investigated as well as those ofniobia–silica prepared by e-beam deposition [23,24].Nevertheless, the optical resistance of these mix-tures produced by the IBS technique are not studiedin detail at subpicosecond pulse durations. A deeperknowledge of the optical resistance properties inzirconia– and niobia–silica single layers could bevery helpful for further development of multilayer-mixture-based femtosecond optics. Therefore, theaim of this study was deposition and multiscale ana-lysis of IBS zirconia– and niobia–silica mixtures.

2. Experimental

A. Preparation of Thin Film Samples

All experimental coatings were deposited by using amodified IBS technique on 1mm thick fused-silica(UV grade KU1 glass) substrates from the same pol-ishing batch. An IBS coating plant from CuttingEdge Coatings GmbH was used for deposition. Theapparatus was equipped with two vacuum pumps:a combination of a cryopump with a mechanicalpump resulted in a base pressure of 3 × 10−5 Pa. Be-fore the deposition process, the vacuum chamber wasbaked at 50 °C for 1 h. To remove the impurity layer,the ion source was used for presputtering of the tar-get before deposition. During the process, oxygen gaswas supplied toward the substrate to ensure com-plete oxidation of the growing coating. The resultingworking pressure was 3 × 10−3 Pa. A radio-frequencygrid-system-based ion source was used to strike aplane metal zone target, consisting of two differentmaterials [high refractive index (Zr or Nb) and lowrefractive-index (Si), respectively] at an angle of in-cidence of 57 deg. Typical parameters of the main ion

source, using argon gas, were set to 1200V=130mA,resulting in deposition speeds of 1Å=s for low- and0:6Å=s for high-refractive-index materials, respec-tively. The target was mounted on a linear transla-tion stage to shift the ion–metal interaction zoneacross the two different coating materials. Mixingdifferent fraction materials was achieved by direct-ing the high-energy ion beam toward the intersectionbetween the zones (Fig. 1). The films were depositedon substrates held in a circular rack, which was ro-tated around its axis at a speed of 20 rpm. The thick-nesses of growing films were monitored by anintegrated broadband transmission optical monitor-ing system in the wavelength range of 400–1000nm[25]. Two sets of single-layer mixture samples wereprepared. All films were of the same optical thick-ness at the wavelength λ ¼ 1064nm and, namely,6λ=4nλ, where nλ is the refractive index. A total ofnine different compositions were investigated withthe silica fraction varying from 0% to 100% in desir-able increments of about 25%. The fraction of eachmaterial in the mixture was adjusted by covering ap-propriate parts of the target zones with the high-energy ion beam during the deposition process. Thecoating with approximately a 25% target value ofhigh-index material was named as the “high index—low/high—SiO2” sample. Target coatings of 50%/50% and 75%/25% are named as “half/half” and“high/low” samples, respectively. Later, the exactfractions were characterized by using effective-medium theories.

B. Characterization of the Thin Films

First, reflectance and transmittance (RT) measure-ments at close to a normal angle of incidence(AOI ¼ 8:5 deg) were performed in the spectralrange of 200–1200nm with a LAMBDA 950spectrophotometer (Perkin Elmer). Afterward, therefractive-index and physical-thickness analysis wascarried out in a low-absorptance spectral region with

Fig. 1. (Color online) Principle of the IBS process.

20 March 2011 / Vol. 50, No. 9 / APPLIED OPTICS C189

Optichar, the optical characterization software fromOptilayer [26]. Spectrophotometric analysis ofrefractive indices in combination with MaxwellGarnett [27], Bruggeman [28], and Lorentz–Lorenz[29,30] effective-medium theories are applied to esti-mate the volumetric material fraction in every inves-tigated coating. The bandgap (Eg) values known asclosely related to the LIDT in dielectrics were deter-mined from absorption spectra by using three differ-ent approaches: linear fitting of so called Tauc(EgT) and Cody (EgT) plots [31], as well as the inflec-tion point method (EgU) derived from the Urbachequation [32]. Furthermore, another important para-meter describing optical surface quality and, namely,root mean surface roughness has been investigated.Light scattering [33–36] as well as atomic forcemicroscopy (AFM) methods are frequently used forsurface roughness characterization. In this study,two independently developed systems for opticalscattering characterization were used at differentwavelengths. The first system having angle-resolvedscattering (ARS) measurement capability was devel-oped at the Institut Fresnel (France)[37]. This fullycomputer-controlled system is based on the collectionin the far field of scattered light by a detector placedon a rotating arm. The ARS measurements were per-formed using a monochromatic unpolarized light un-der normal incidence at the 633nm wavelength. Thebeam diameter (at 1=e2 intensity level) on the sampleplane was about 2mm. In the visible wavelengthrange, it allows one to reach the sensitivity of an an-gular scattering measurement within a seven-decaderange with respect to a Lambertian etalon sample.For direct measurements of total integrated scatter-ing (TIS), another measurement system has been de-veloped at the Laser Research Center of VilniusUniversity (Lithuania). This system is based on anintegrating Ulbricht sphere and was assembledaccording to the existing recommendations of inter-national ISO13696 standard [38]. The frequencydoubled pulses from a nanosecond Q-switchedNd:YVO laser were used at 532nm wavelengthand 1kHz repetition rate. The beam diameter (at1=e2 intensity level) on the sample plane was about0:47mm. The sample was mounted on the two-axisXY translation stage behind the Ulbricht sphere,thus allowing computer-controlled TIS mapping ofthe surface. The specularly reflected and transmittedbeams were directed out of the chamber and ab-sorbed by the beam dumps to ensure sufficient sen-sitivity of the measurement. As in previous ARSmeasurements, TIS values were measured with re-spect to a Lambertian etalon sample. To comparesurface roughnesses evaluated from TIS and ARSmeasurements, additional measurement series withan atomic force microscope (Quesant Q-scope 250)have been made. We used the contact mode imagingwith a scanned area of 10 μm× 10 μm. The root meansquare values of the roughness are given for all thesamples. X-ray diffraction (XRD) analysis was car-ried out with a Philips Xpert-MPD diffractometer

with an X’Celerator detector. The device works inparafocusing Bragg–Brentano geometry. The x-raysource was a ceramic x-ray tube with a Cu anode.The Kα spectral line was used for XRD analysis(Kβ filtered out by an Ni filter). Phase identificationwas done by using the Powder Diffraction File PDF-2database. Finally, laser-induced damage thresholdsof all the samples were tested in a one-on-one mode[39]. The pulse duration, wavelength, and beam dia-meter of the laser source were 530 fs, 1030nm, and74 μm, respectively, at (1=e2) level of maximal inten-sity. A differential interference contrast (Nomarski)microscope was employed to observe the laser-in-duced damage morphology after laser irradiation.Any laser-induced modifications of irradiated sitesthat can be resolved with a 100× magnification objec-tive were considered as “damaged.” More details onLIDT setup are given in [40].

3. Results and Discussion

A. Analysis of Spectrophotometric Data

The experimental RT curves are shown inFigs. 2 and 3. In all cases, measurements were per-formed under identical experimental conditions:samples were irradiated from the same beam sideat the same angle of incidence. The refractive indexand coating thickness analysis was carried out byusing classical Fresnel reflectance formulas and

Fig. 2. (Color online) Reflection and transmission of niobia–silicamixture coatings.

Fig. 3. (Color online) Reflection and transmission of zirconia–silica mixture coatings.


performing simulations of virtual RT spectra in asubstrate/monolayer optical system. For all the sam-ples, model parameters were virtually varied to fitthe experimental RT spectra. Sellmeier refractive-index dispersion and coating thickness correspond-ing to the best fit are considered as those of the realspecimen. It is worthy of note that refractive indicesand thicknesses extracted from transmission-onlydata slightly differed from those extracted from re-flection-only data. The obtained discrepancies possi-bly arise due to the systematic and calibration errorsof the spectrophotometer (wavelength accuracy of0:3nm and relative photometric error typically doesnot exceed 1%). The difference in refractive indicesand physical thicknesses extracted from reflection-and transmission-only data in both cases is less than1% over the whole wavelength range of analysis.Extinction k and absorption α coefficients were alsodeduced from an energy-conservation relation forlight by taking into account substrate losses, coatingthickness, and neglecting the scattering part:

Rþ Aþ Sþ T ¼ 1; ð1Þ

where R, A, S, and T are fractions of reflected, ab-sorbed, scattered, and transmitted light, respec-tively. The absorption bandgap values for the vastmajority of the samples were determined from calcu-

lated extinction k spectra, according to the Taucand Cody plot [31] and form absorption α spectra–inflection point [32] methods. For pure silicasamples, the absorption edge was out of spectropho-tometric measurement limits. However, the approx-imate values for similar materials can be found in theliterature [41,42]. The summarized mean values ofphysical thicknesses lph: and refractive indices n,as well as those of bandgaps Eg and high-indexmaterial fractions obtained by different methodsfrom spectrophotometric RT data are summarizedin Table 1.

As can be seen from Figs. 4 and 5, the refractive-index can be easily tuned between the values of purematerials. It can also be seen that the refractive in-dex of a pure silica coating was slightly higher thanthat of the substrate. The possible explanations forthis could be a very high density of coating, the pre-sence of some nonoxidized silicon or working gasatoms inside the coating, or even production of a verylow-concentration mixture with a small amount ofzirconia or niobia.

Next, the volumetric fraction f H ¼ X in eachðhigh refractive indexÞXðsilicaÞ1−X mixture systemis analyzed according to the so-called effective-medium theories. Maxwell Garnett (MG) [27][Eq. (2)], Bruggeman (BG) [28] [(Eq. (3)], andLorentz–Lorenz (LL) [29,30] [Eq. (4)] mixing formulasreproduced in [18] are used:

Table 1. Comparison of Sample Properties Obtained from Spectrophotometric Data

Sample lph: (nm) n1064 n1030 EgC (eV) EgT (eV) EgU (eV) f HðLLÞ f HðBGÞ f HðMGÞ

Pure Materials:Substrate 1mm 1.46 1.46 — — ≈9 [41] — — —

SiO2 1090 1.48 1.48 — — 8.3 [42] 0 0 0Nb2O5 717 2.23 2.23 3.43 3.46 3.92 1 1 1ZrO2 759 2.09 2.09 4.67 4.74 5.29 1 1 1

Mixed Nb2O5=SiO2:Low/high 939 1.71 1.72 3.87 3.94 4.80 0.386 0.336 0.316Half/half 838 1.92 1.92 3.70 3.74 4.37 0.664 0.604 0.592High/low 773 2.07 2.07 3.57 3.61 4.18 0.836 0.794 0.790

Mixed ZrO2=SiO2:Low/high 970 1.66 1.66 5.30 5.38 — 0.330 0.292 0.277Half/half 885 1.81 1.81 5.13 5.18 5.99 0.573 0.523 0.511High/low 811 1.96 1.96 4.88 4.94 5.61 0.792 0.752 0.748

Fig. 4. (Color online) Refractive indices of niobia–silicamixture coatings extracted by using the Sellmeier formula fitfrom RT data.

Fig. 5. (Color online) Refractive indices of zirconia–silica mixturecoatings extracted by using the Sellmeier formula fit fromRT data.


MG :εeff − εHεeff þ 2εH

¼ ð1 − f HÞεL − εHεL þ 2εH

; ð2Þ

BG : f HεH − εeffεH þ 2εeff

þ ð1 − f HÞεL − εeffεL þ 2εeff

¼ 0; ð3Þ

LL :εeff − 1εeff þ 2

¼ f HεH − 1εH þ 2

þ ð1 − f HÞεL − 1εL þ 2

; ð4Þ

where εeff , εH , and εL are the dielectric functions ofthe effective-medium (mixture) material, the high-refractive-index material, and the low-refractive-index material, respectively. f H is the volumetricmaterial fraction of the higher refractive-index mate-rial in amixed coating. The f H results calculated fromEMA theory are shown in Figs. 6 and 7. As canbe seen, the obtained fractions that we suppose tobe constantaredependenton thewavelengthatwhichthe analysis is made. In general, this fact evidencesthe limitations of the classical effective-medium the-ories and fails to describe IBS mixtures in the UVrange. In the visible range and the near-infraredrange, the dependence on wavelength is rather weak;however, allmodels show slightly differing results. Aswas shown by XRD measurements (Subsection 3.C),such discrepancies in zirconia–silica mixtures canpossibly be explained by changes in the crystalline

structure when the SiO2 content is varied. Similar re-sults were obtained for all the amorphous niobia–silicamixtures; therefore, it ismore reasonable to relyon the BG model that was less sensitive to the wave-length of analysis than those of the MG and LL mod-els. It is interesting to note that, in the case of similarniobia–silica mixture coatings deposited by reactiveelectron beam coevaporation, it was found by Janickiet al. [18] that the LL model gave the most reliableresults.

B. Evaluation of Surface Roughness by Optical Scatteringand AFM

Three different types of measurements were per-formed for characterization of surface roughness:angle-resolved scattering, total integrated scatter-ing, and scanning of surface by AFM. By using theARS setup, the so-called bidirectional reflectance dis-tribution functions (BRDFs) were recorded for everysample. The zone free from observable defects of ap-proximately 3mm2 size was illuminated by a colli-mated laser beam and the BRDF recorded on theangular range of incidence from 5 to 89 deg in2 deg angular steps. To compare obtained data withother measurements, the equivalent values of totalintegrated scattering coefficient S were derived fromthe BRDF. Later, root mean square roughnesses σRMSwere calculated from TISmeasurements according tothe following formula [35]:

Fig. 6. (Color online) Volumetric fraction of high-index materialcontent (X) in niobia–silica composite coating.

Fig. 7. (Color online) Volumetric fraction of high-index materialcontent (X) in zirconia–silica composite coating.

Fig. 8. AFM images of pure zirconia, niobia, silica, and mixture zirconia–, niobia–silica coatings.


σRMS ¼ffiffiffiffiSR

r·

λ4πn0 cosði0Þ

; ð5Þ

where λ is the wavelength, S is the TIS coefficient(part of scattered light with respect to the Lamber-tian etalon sample), R is the reflected part of light,n0 is the refractive index of the surrounding medium,and i0 is the angle of incidence. This equation wasderived by considering the inequality of σRMS ≪ λand also the fact that layers are fully correlated;thus, the surface of each layer perfectly replicates theroughness of the substrate. This condition is knownto be fulfilled for compact deposition technologies,such as IBS. Similarly, σRMS values were obtainedby direct measurement of S by using a second TISsetup at 532nm. In this case, statistics from 482 sur-face sites are taken for every sample. The illumi-nated area of each site was about 0:17mm2, andan 82mm2 area is interrogated in total. The medianvalue of S is taken from the collected statistics andhas been further used for comparisons. Finally, thesurface topography obtained by AFM is presentedin Fig. 8.

All the roughnesses obtained by different methodsare compared in Fig. 9. As can be seen, roughness cal-culated from AFMmeasurements correlates with theroughness calculated from total integrated scatteringmeasurements to within an order of magnitude andshow the same tendency for all coatings. However,the σRMS values of the above-mentioned measure-ments and, namely,AFMandTIS, have been obtainedby different spatial sampling frequency and cannot bedirectly compared [43,44]. The roughness of all IBSmixture coatings as well as those of pure niobiaand silica coatings are found to be similar to the sub-strate roughness, which is an expected result for IBStechnology. It is not the case for the zirconia IBSmonolayer, which exhibits an anomalous high rough-ness. The possible explanation for high zirconiasurface roughness ismicrocrystallization of themate-rial because grain structure can be also observed fromthe AFM image in Fig. 8.

C. X-Ray Diffraction Measurements

On the supposition that some of the coatings mightbe of different molecular/crystalline phase, x-ray dif-fraction measurements were performed. Figure 10presents the XRD results taken from the silica sub-strate and Nb2O5–SiO2 mixtures. No significant dif-ferences appear among these curves, showing that allthese samples are amorphous. In Fig. 11, results of asimilar test on zirconia–silica mixtures are shown. Inthis case, an XRD spectrum of pure zirconia showedsharp peaks, indicating a polycrystalline structure ofcoating. Some of the peaks can be attributed to amonoclinic and tetragonal crystalline structure ofZrO2, while the others correspond to Zr2O and ZrOstoichiometry. However, in general, there were alsopeaks that still could not be identified with anyknown crystal structure. Therefore, an ideal stoichio-metric ratio that matches all those peaks is notknown. The grain size of crystallites in a ZrO2 coat-ing has been roughly evaluated by using the Scherrerequation [45]:

d ¼ 0:89 · λδ · cosðθÞ ; ð6Þ

where λ is the wavelength of x rays (λcopperKα ¼0:154056nm), δ is the FWHM of the diffraction peak,and θ is the angle corresponding to the peak. The ty-pical grain size corresponding to different peaks of apure zirconia sample was estimated to be in therange from approximately 12 to 21nm. Of course,

Fig. 9. (Color online) Comparison of surface roughnessesestimated by different methods.

Fig. 10. X-ray diffraction patterns of substrate, pure niobia, puresilica, and niobia–silica mixture coatings.

Fig. 11. X-ray diffraction patterns of substrate, pure zirconia,pure silica, and zirconia–silica mixture coatings.


in general, the peak broadening could be attributednot only to the grain size but could also be affected byother factors, such as device parameters and internalstress. However, our aim was only rough estimationof the grain size, which is in agreement with AFMtopography data. By introducing a small amount ofSiO2 into zirconia, the crystalline structure of themixture experience a transition to amorphous phase.As can be seen, the high/low zirconia–silica sampleonly begins to crystallize showing a very broad peakbetween 28 and 36 deg. All the others samples ofhigher SiO2 concentration were also amorphous.

D. LIDT Measurements

Finally, the optical resistance of all samples was ex-amined in terms of a single shot LIDT. The measure-ments were performed by using laser pulses of 530 fsduration at 1030nm wavelength and a 74 μm beamdiameter (1=e2) in the target plane. For all samples,the obtained damage probability functions were verydeterministic, thus meaning “sharp” and reproduci-ble transition in damage probability between 0% and100% versus fluency. LIDT values clearly correlatewith the silica content in mixture materials withinthe limit values of pure materials. In both cases,either zirconia– or niobia–silica mixtures, the expo-nential decay dependence of LIDT versus f H is ob-served. The experimental LIDT data were fitted bya simple exponential relation:

LIDTðf HÞ ¼ A · exp�−f HC

�þ B: ð7Þ

Here, LIDTðf HÞ is the incident threshold fluency inJ=cm2, f H is the volumetric material fraction of high-er refractive-index material in %, and A, B, and C arefit parameters (here A and B are expressed in J=cm2

and their sum is equal to the threshold fluency of apure low-refractive-index coating material, C is ex-pressed in %). The results of LIDT data on mixturesand their exponential fits are summarized in Fig. 12.The obtained fitting parameters and their standarderrors are enclosed Table 2.

Another phenomenological relation has been pre-viously observed in titania–silica mixtures [5]. Theconclusion was made that the internal threshold

fluency of a standing wave follows the observedpower-law dependence on pulse duration and lineardependence on the absorption bandgap expressed ineV. This phenomenological formula was also pre-viously reported on pure metal-oxide coatings byMero et al. [42]:

Fðτ;EgÞ ¼ ðc1 þ c2 · EgÞ · τK : ð8ÞHere, Fth is the internal threshold fluency in J=cm2

determined from the standing-wave maximum of theelectromagnetic wave, Eg is the absorption bandgapof the material, and τ is the pulse duration (in fs)only, while the other parameters were fixed andset to c1 ¼ −0:16 J · cm−2 · fs−k, c2 ¼ 0:074 J · cm−2·fs−k eV−1, and K ¼ 0:3. To compare our data with theproposed relation, the ratio of the maximum internaland the incident intensity must be taken into ac-count. In our case, the maximum fluency in all thefilms is about 0.63 times the incident fluency inde-pendent of the material. For all samples, two of thestanding-wavemaxima are located at the air–coatingand the coating–substrate interface. The comparisonbetween the experimental results and the proposedmodel is made in Fig. 13. As can be seen, the firstambiguity issues are related to the definition ofEg. The inflection point method seems to be moreclose to the proposed model. Indeed, the LIDT versusthe EgU relation confirms that experimental valuesof pure materials are very close to those predicted bythe model. However, the results of mixtures clearlydeviate from the prediction by increasing silica con-tent in the mixture. This discrepancy could possiblybe explained by considering the wavelength differ-ence in LIDT testing. It was previously reported thatthe damage threshold could be wavelength depen-dent [46]. Another possible reason for discrepancycould be attributed to the recently reported transient

Fig. 12. (Color online) LIDT dependence on f H in mixed metaloxide coatings; here, the fluency of incident wave is used.

Table 2. Exponential Decay Fit Parameters of LIDT Data

Sample A (J=cm2) B (J=cm2) C (%)

Nb2O5=SiO2: 5:44� 0:12 0:66� 0:09 41:7� 2:5ZrO2=SiO2: 5:03� 0:06 1:11� 0:07 73:8� 2:4

Fig. 13. (Color online) Comparison of experimentally obtainedLIDT data and model prediction with respect to different bandgapdefinitions.


interference effects [47]. To confirm or negate theabove-mentioned assumptions, further research isnecessary to explain the differences arising betweenexperimental data on mixtures and the proposedmodel.

Damage morphologies of all samples were alsocompared in Fig. 14. A difference can be seen be-tween the pure and the mixture coatings. The transi-tion from “burned dots” type to “melting” type ofdamage is observed from pure silica to the mixturefor both zirconia– and niobia–silica mixtures. Allthe mixtures as well as pure niobia have a similar“melted” type of damage morphology. However, purezirconia results in a “delamination” type of damage,in which the coating is simply removed and the bot-tom of the crater seems to be undamaged. The differ-ences in damage morphology suggest that separate(material dependent) laser ablation mechanisms ex-ist in dielectric coatings at a subpicosecond timescale. On the other hand, the origins of the initial da-mage process are possibly very similar for all the in-vestigated coatings because there is no enormousdeviation in LIDT dependence on f H .

4. Conclusions

Zirconia–silica mixture coatings produced by the IBStechnique experience the transition from crystallineto amorphous phase by increasing the content ofglassy SiO2 in the mixture, thus resulting in en-hancement of LIDT. The relatively large roughnessarises only in the well-crystallized zirconia samplebecause it is difficult to form a flat surface from crys-tallite domains, while samples that remain amor-phous have smaller roughness. This relationshiphas been confirmed by XRD and AFM results andsupports the insights made in [7]. In the case ofniobia–silica mixtures, all samples were amorphous,and the addition of SiO2 resulted in enhancement ofLIDT as well. The obtained LIDT values correlatewith the silica content in the composite coating forboth zirconia– and niobia–silica mixtures. Experi-mental data of LIDT reveal exponential decay depen-dence on volumetric material content of higherrefractive-index material. Deviation from bandgapand pulse duration based on the model proposed in[42] is also observed. By analysis of the damage mor-phology, the assumption is made that more mech-anisms of coating ablation are involved in the sub-

picosecond laser-induced breakdown process. Theeffective-medium theories, namely, the MaxwellGarnett, Bruggeman, and Lorentz–Lorenz formulasused to describe volumetric fractions of materials inmixtures, gave wavelength-dependent results evenin the low-absorption region. This possibly couldbe related to different crystalline forms of pure andmixed coatings; therefore, application of the above-mentioned formulas might be inappropriate in someapplications where material concentration control iscritical.

The research leading to these results receivedfunding from the Seventh Framework Program ofthe European Commission (EC) (LASERLAB-EUROPE, grant agreement 228334). We also ac-knowledge financial support from the LithuanianFrench Program Gilibert (the Lithuanian ScienceCouncil and the French Foreign Ministry).

References1. P. J. Martin, “Ion-based methods for optical thin film deposi-

tion,” J. Mater. Sci. 21, 1–25 (1986).2. J. R. Sites, H. Demiryont, and D. B. Kerwin, “Summary

abstract: ion-beam sputter deposition of oxide films,” J. Vac.Sci. Technol. A 3, 656–656 (1985).

3. V. Janicki, D. Gäbler, S. Wilbrandt, R. Leitel, O. Stenzel,N. Kaiser, M. Lappschies, B. Görtz, D. Ristau, C. Rickers,and M. Vergöhl, “Deposition and spectral performance of aninhomogeneous broadband wide-angular antireflective coat-ing,” Appl. Opt. 45, 7851–7857 (2006).

4. M. Jupé, M. Lappschies, L. Jensen, K. Starke, and D. Ristau,“Applications of mixture oxide materials for fs optics,” inOptical Interference Coatings (Optical Society of America,2007), p. TuA6.

5. D. Nguyen, L. A. Emmert, I. V. Cravetchi, M. Mero,W. Rudolph, M. Jupe, M. Lappschies, K. Starke, and D. Ristau,“TixSi1−xO2 optical coatings with tunable index and their re-sponse to intense subpicosecond laser pulse irradiation,”Appl.Phys. Lett. 93, 261903 (2008).

6. S. Chao, W.-H. Wang, and C.-C. Lee, “Low-loss dielectricmirror with ion-beam-sputtered TiO2–SiO2 mixed films,”Appl.Opt. 40, 2177–2182 (2001).

7. B. J. Pond, J. I. DeBar, C. K. Carniglia, and T. Raj, “Stressreduction in ion beam sputtered mixed oxide films,” Appl.Opt. 28, 2800–2805 (1989).

8. M. Lappschies, B. Görtz, and D. Ristau, “Application of opticalbroadband monitoring to quasi-rugate filters by ion-beamsputtering,” Appl. Opt. 45, 1502–1506 (2006).

9. D. Ristau, H. Ehlers, S. Schlichting, andM. Lappschies, “Stateof the art in deterministic production of optical thin films,”Proc. SPIE 7101, 71010C (2008).

Fig. 14. (Color online) Damage morphology of investigated coatings.


10. C.-C. Lee, C.-J. Tang, and J.-Y. Wu, “Rugate filter made withcomposite thin films by ion-beam sputtering,” Appl. Opt. 45,1333–1337 (2006).

11. C.-J. Tang, C.-C. Jaing, K.-S. Lee, and C.-C. Lee, “Residualstress in Ta2O5–SiO2 composite thin-film rugate filters pre-pared by radio frequency ion-beam sputtering,” Appl. Opt.47, C167–C171 (2008).

12. H. Bartzsch, S. Lange, P. Frach, and K. Goedicke, “Graded re-fractive index layer systems for antireflective coatings and ru-gate filters deposited by reactive pulse magnetron sputtering,”Surf. Coat. Technol. 180–181, 616–620 (2004).

13. D. M. Sanders, E. N. Farabaugh, W. S. Hurst, andW. K. Haller,“Summary abstract: an advanced multi-chamber system forpreparation of amorphous thin films by coevaporation andtheir subsequent characterization by AES, ESCA, SIMSand ISS methods,” J. Vac. Sci. Technol. 18, 1308–1310 (1981).

14. H. Sankur, W. J. Gunning, and J. F. DeNatale, “Intrinsic stressand structural properties of mixed composition thin films,”Appl. Opt. 27, 1564–1567 (1988).

15. N. K. Sahoo and A. P. Shapiro, “Process-parameter-dependentoptical and structural properties of ZrO2MgO mixed-composite films evaporated from the solid solution,” Appl.Opt. 37, 698–718 (1998).

16. S. Larouche, H. Szymanowski, J. E. Klemberg-Sapieha,L. Martinu, and S. C. Gujrathi, “Microstructure of plasma-deposited SiO2=TiO2 optical films,” J. Vac. Sci. Technol. 22,1200–1207 (2004).

17. M. Yoshida and P. N. Prasad, “Sol-gel-processed SiO2=

TiO2/poly(vinylpyrrolidone) composite materials for opticalwaveguides,” Chem. Mater. 8, 235–241 (1996).

18. V. Janicki, J. Sancho-Parramon, andH. Zorc, “Refractive indexprofile modelling of dielectric inhomogeneous coatings usingeffective medium theories,” Thin Solid Films 516, 3368–3373(2008).

19. D. Aspnes, “Optical properties of thin films,” Thin Solid Films89, 249–262 (1982).

20. A. Feldman, E. N. Farabaugh,W. K. Haller, D. M. Sanders, andR. A. Stempniak, “Modifying structure and properties ofoptical films by coevaporation,” J. Vac. Sci. Technol. A 4,2969–2974 (1986).

21. N. K. Sahoo, S. Thakur, and R. B. Tokas, “Achieving superiorband gap, refractive index and morphology in composite oxidethin film systems violating the moss rule,” J. Phys. D 39,2571–2579 (2006).

22. D. H. Kuo and C. H. Chien, “Growth and properties of sput-tered zirconia and zirconia–silica thin films,” Thin Solid Films429, 40–45 (2003).

23. J. Sancho-Parramon and V. Janicki, “Effective medium the-ories for composite optical materials in spectral ranges ofweak absorption: the case of Nb2O5–SiO2 mixtures,” J. Phys.D 41, 215304 (2008).

24. J. Sancho-Parramon, V. Janicki, and H. Zorc, “Compositionaldependence of absorption coefficient and band-gap forNb2O5–SiO2 mixture thin films,” Thin Solid Films 516,5478–5482 (2008).

25. D. Ristau, H. Ehlers, T. Gross, and M. Lappschies, “Opticalbroadband monitoring of conventional and ion processes,”Appl. Opt. 45, 1495–1501 (2006).

26. URL: http://www.optilayer.com.27. J. C. Maxwell Garnett, “Colours in metal glasses and in me-

tallic films,” Philos. Trans. R. Soc. London Ser. A 203, 385–420 (1904).

28. D. A. G. Bruggeman, “Berechnung verschiedener physika-lischer konstanten von heterogenen substanzen. I. Dielektri-zitätskonstanten und leitfähigkeiten der mischkörper ausisotropen substanzen,” Ann. Phys. 416, 636–664 (1935).

29. A. H. A. Lorentz, “Ueber die beziehung zwischen der fortpflan-zungsgeschwindigkeit des lichtes und der koerperdichte,”Ann. Phys. 245, 641–665 (1880).

30. L. Lorenz, “Ueber die refractionsconstante,” Ann. Phys. 247,70–103 (1880).

31. C. Rotaru, S. Nastase, and N. Tomozeiu, “Amorphous phaseinfluence on the optical bandgap of polysilicon,” Phys. StatusSolidi A 171, 365–370 (1999).

32. F. A. Ponce, S. Srinivasan, A. Bell, L. Geng, R. Liu, M. Stevens,J. Cai, H. Omiya, H. Marui, and S. Tanaka, “Microstructureand electronic properties of InGaN alloys,” Phys. Status SolidiB 240, 273–284 (2003).

33. C. Amra, “Light scattering from multilayer optics. I. Tools ofinvestigation,” J. Opt. Soc. Am. A 11, 197–210 (1994).

34. C. Amra, “Light scattering from multilayer optics. II.Application to experiment,” J. Opt. Soc. Am. A 11, 211–226(1994).

35. J. M. Elson, J. P. Rahn, and J. M. Bennett, “Relationship of thetotal integrated scattering from multilayer-coated optics toangle of incidence, polarization, correlation length, and rough-ness cross-correlation properties,” Appl. Opt. 22, 3207–3219(1983).

36. A. Duparré and S. Kassam, “Relation between light scatteringand the microstructure of optical thin films,” Appl. Opt. 32,5475–5480 (1993).

37. C. Amra, D. Torricini, and P. Roche, “Multiwavelength(0:45–10:6 μm) angle-resolved scatterometer or how to extendthe optical window,” Appl. Opt. 32, 5462–5474 (1993).

38. International Organization for Standardization, “Optics andoptical instruments. Test methods for radiation scattered byoptical components,” ISO 13696 (International Organizationfor Standardization, 2002).

39. International Organization for Standardization, “Determina-tion of laser-induced damage threshold of optical surfaces—part 1: 1-on-1 test, ” ISO 11254-1 (International Organizationfor Standardization, 2000).

40. L. Gallais, B. Mangote, M. Zerrad, M. Commandré, A. Melnin-kaitis, J. Mirauskas, M. Jeskevic, and V. Sirutkaitis, “Laserinduced damage of hafnia coatings as a function of pulse dura-tion in the femtosecond to nanosecond range,” Appl. Opt. 50,C178–C187 (2011).

41. A. Zoubir, C. Rivero, R. Grodsky, K. Richardson,M. Richardson, T. Cardinal, and M. Couzi, “Laser-induced de-fects in fused silica by femtosecond IR irradiation,” Phys. Rev.B 73, 224117 (2006).

42. M. Mero, J. Liu, W. Rudolph, D. Ristau, and K. Starke, “Scal-ing laws of femtosecond laser pulse induced breakdown inoxide films,” Phys. Rev. B 71, 115109 (2005).

43. C. Deumié, R. Richier, P. Dumas, and C. Amra, “Multiscaleroughness in optical multilayers: atomic force microscopyand light scattering,” Appl. Opt. 35, 5583–5594 (1996).

44. M. Zerrad, C. Deumié, M. Lequime, and C. Amra, “An alter-native scattering method to characterize surface roughnessfrom transparent substrates,” Opt. Express 15, 9222–9231(2007).

45. A. L. Patterson, “The Scherrer formula for x-ray particle sizedetermination,” Phys. Rev. 56, 978–982 (1939).

46. M. Jupé, L. Jensen, A. Melninkaitis, V. Sirutkaitis, andD. Ristau, “Calculations and experimental demonstration ofmulti-photon absorption governing fs laser-induced damagein titania,” Opt. Express 17, 12269–12278 (2009).

47. L. Gallais, B. Mangote, M. Commandré, A. Melninkaitis,J. Mirauskas, M. Jeskevic, and V. Sirutkaitis, “Transientinterference implications on the subpicosecond laserdamage of multidielectrics,” Appl. Phys. Lett. 97, 051112(2010).


http://www.optilayer.com



characterization of zirconia– and niobia–silica mixture coatings produced by ion-beam sputtering

Documents