Open Access
Add to BibSonomyAbstract
Material ablation by excimer laser micromachining is a promising approach for structuring sol-gel materials as we demonstrate in the present study. Using the well-known direct etching technique, the behaviour of different hybrid organic/inorganic self-made sol-gel materials is examined with a KrF* laser. Ablated depths ranging from 0.1 to 1.5 µm are obtained with a few laser pulses at low fluence (<1 J/cm2). The aim is to rapidly transfer surface relief multi-level diffractive patterns in such a substrate, without intermediate steps. The combination with the 3D profilometry technique of coherence probe microscopy permits to analyse the etching process with the aim of producing multi-level Diffractive Optical Elements (DOE). Examples of four-level DOEs with 10 µm square elementary cells are presented, as well as their laser reconstructions in the infrared.
©2008 Optical Society of America
1. Introduction
Diffractive optical elements (DOEs) are a complex transmittance employed to modify different characteristics of an incident light beam. There are several classes of elements in this field; the ones of best interest to us here are the computer-generated Fourier kinoforms that present a relief modulation which change the incident wave phase [1]. DOEs are atypical from ordinary optical components in that they can be extended to many new functions such as direct imaging [2,3], optical interconnects [4], optical read heads, beam shaping [5–7] and beam separation with laser sources [8]. The present work is mainly concerned with Fourier multi-level DOEs operating in the transmissive mode with infrared laser sources, particularly in 1–2 µm band, which offers many advantages for laser machining in the infrared spectrum such as multipoint drilling [9] and brazing [10], pattern marking and certain thermal treatments [5]. The Fourier element focuses its image at infinity and requires an intermediate lens in order to observe the diffraction reconstruction. During the 1990s, very large scale integration (VLSI) technologies developed in microelectronics were successfully extended to the fabrication of micro-optical components. Today, projection photolithography followed by dry etching is the standard answer for making diffractive structures in the visible spectrum using dielectric materials. Diamond turning can only be provided for components with axial or circular symmetry. Other techniques, such as laser-beam writing (LBW) or electron-beam writing (EBW), achieve multi-level or continuous surface relief with high resolution [11,12]. These solutions are complex and costly, though, and do not satisfy the requirements for rapid prototyping, which is of growing interest to manufacturers today. In fact, standard photolithography or beam writing techniques in which many parameters must be controlled (such as the temperature, the humidity, the time of baking, the insolation and development time, etc…) make it very difficult to be employed for rapid prototyping and small production where simplicity, rapidity and efficiently are required.
Laser ablation is today an efficient micromachining technique. Different authors have still proposed using direct ablation to create DOEs in polymers or quartz, but only for operating in visible light and low power activities [13–17]. The use of an organic polymer as the substrate for KrF* laser ablation is more common than quartz or glass, because the laser fluence requirement is lower and the resulting roughness is currently better. However, in the case of excimer lasers, different answers have been proposed for laser ablation with ns-, ps- or fs-pulses of dielectric materials such as additional absorbing materials deposited on the sample, liquid baths, temporary plasmas and adsorbed layers [16,17]. The fabrication of multi-level surface relief structures with a micron resolution inside a thin transparent layer remains a challenge for certain applications. KrF* laser ablation at 248 nm with low roughness on fused silica and quartz has recently been demonstrated in the literature with a particular setup [18,19]. Although the roughness is comparable to that of a polished surface, the required fluence is higher than that of an organic polymer.
Hybrid organic/inorganic sol-gel materials are well-adapted candidates for optical applications, due to their properties and their patternability. Several fabrication processes for these devices have been proposed today. The photolithography process is among the most known. The hybrid sol-gel material, exposed to ultraviolet light through masks, undergoes photopolymerization. The exposed zones become thus resistant to organic solvents, which makes it possible to remove the unexposed zones by dissolution. The realization of waveguides and gratings by these means has been demonstrated [20–22]. Other local densification techniques, such as laser beam scanning [23] and electron-beam lithography [24], are also employed with direct writing into the material. For all of these methods, the creation of chemical bonds in the exposed pattern induces a condensation of the material, which permits to carry out a surface relief. Organic/inorganic sol-gel materials have here been investigated with the ablation process and multi-level structuring keeping in mind fast prototyping, small production, flexibility, low cost and simplicity. This alternative method for the patterning of hybrid sol-gel films by means of UV-laser micromachining has been conceived with low laser fluence and low cost thin layer deposition on dielectric substrate. The aim was also producing in a fast handling way multi-level micro-optics and three-dimensional profiles applied in the infrared range of the spectrum. This original method avoids the expensive fabrication of chromium masks and limits the number of steps of the process, which does not require any development or cleaning after operation. Besides, this approach of great interest shows us that laser ablation is well adapted to precisely control the etching depth in sol-gel material by the laser fluence and/or the number of laser pulses. Furthermore, the sol-gel process allows easily to carry out particular characteristics or to functionalize the material by simple addition of organic species such as dyes, photo cross-linkable functions or molecules exhibiting non linear optical properties. In this article, hybrid organic/inorganic sol-gel films were processed by KrF* laser ablation with ns-laser pulses at 248 nm. The association of various precursors (Silicon and Titanium) enables the tuning of the optical and mechanical properties of our material. The organic network, initiated by a functionalized precursor, provides elasticity and more resistance to cracks in the final material, even when prepared in high thicknesses (>10 µm). The organic part also induces optical absorption in the UV range which, associated with the low heat capacity, allows the ablation of the sol-gel material at a very low laser fluence. In particular the impact of the titanium ratio and of the annealing temperature of the sol-gel process on the laser ablation rate have been investigated. The present paper also describes the synthesis of the sol-gel materials, the characteristics and ablation properties of our materials as function of different parameters, and brings the experimental evidence that laser micromachining of hybrid sol-gel materials permits the fabrication of effective optical devices.
2. Preparation and characterization of sol-gel material
Two liquid silica precursors were chosen, [3-(methacryloxy)propyl]trimethoxysilane (MAPTMS) and tetraethyl orthosilicate (TEOS), and mixed with a titanium oxide precursor (titanium isopropoxide). Two types of concentrations are presented in the present paper with different quantities of titanium: the first one, referred hereafter as solution A, with a molar ratio of 7 MAPTMS/1 TEOS/2 Ti and the second solution, referred as solution B, with twice as much titanium, i.e. a molar ratio of 7 MAPTMS/1 TEOS/4 Ti. The organic network in the final material is organized by the methacrylate functionalized precursor. It has been demonstrated that the introduction of titanium enhances the refractive index of the films [25]. The different precursors are first partially hydrolyzed. After heating at 40°C for 24 hours to obtain maturation and evaporation of volatile solvents, the solution is dried during 2 hours at 50°C in a vacuum oven (10 mbar). The addition of a controlled amount of solvent (butanol-1) is employed to adjust the viscosity for the spin coating of the solution onto clean borosilicate glass or fused silica substrates. This operation permits also to avoid the impermeable crust effect due to rapid elimination of volatile solvents at the surface. The films are annealed at 100°C for 24 hours to allow polymerization, additional condensation and densification of the matrix. Several annealing temperatures have been investigated with solution B: 100°C (sample B1), 130°C (sample B2) and 160°C (sample B3).
The transmission spectra are measured with a Perkin Elmer Lambda 900 UV/VIS/NIR spectrometer. The effect of the titanium amount and of the annealing temperature on the transmission spectrum of the sol-gel on a glass substrate is examined on Fig. 1. The borosilicate glass substrate thickness is 1.0 mm. Its spectrum is taken as a reference for comparing the other measured spectra. The film thickness is about 3 µm depending on the sample. The different samples present the same type of curve: the transmission ratio is near zero in the UV range whereas it rises to around 90% in the visible and near infrared range. The absolute measurement error with the spectrometer is estimated at 2%. In the case of sample A, the maximum transmission is attained starting at 350 nm, which means that the absorption in the UV range is slightly broader than for the reference spectrum. The absorption in the UV range is attributed to the methacrylate functionalized precursors of the sol-gel film. The laser ablation efficiency is consequently increased in the UV range (KrF* excimer laser emitting at 248 nm). For sample B1, this maximum transmission is attained at 400 nm. The quantity of titanium can then narrow the material bandwidth in the UV range. Insert in Fig. 1, the measured absorption for sample B1 and B3 deposited on a fused silica substrate, as well as the absorption of the fused silica substrate alone. Absorption in the UV can clearly be attributed to the sol-gel layer. Below 300 nm, the absorption signal of sample B1 and B3 saturates. One can furthermore notice an increased absorption in case of sample B3 compared to sample B1, indicating the impact of the annealing temperature. The 90% transmission in the visible and near infrared range offers interesting and evident characteristics for optical devices designed in these wavelength ranges, because the transmission is mainly limited by the Fresnel losses. The impact of the annealing temperature on the transmission spectrum is clearly visible in this experiment. With sample B2 and B3, an increase in the annealing temperature further narrows the bandwidth in the visible range, because the maximum transmission ratio starts at 560 nm and 600 nm respectively.
Fig. 1. Transmission spectra of the glass substrate, sol-gel film with a small titanium doping (A), a sol-gel film with a high titanium doping and annealed at 100°C (B1), 130°C (B2) and 160°C (B3). Insert: absorption of fused silica substrate, sol-gel film B1 and sol-gel film B3 deposited on fused silica substrates.
The refractive index of sol-gel film has been measured with a spectroscopic ellipsometer Jobin-Yvon Model UVISEL. For sample B1, the measurement gives refractive indices of 1.56 and 1.54 at 633 nm and 1064 nm respectively, with a measurement error of ±0.01. The index step with borosilicate glass at the wavelength of 633 nm is thus about 4×10-2. The refractive index of sample A with low titanium ratio is 1.54 and 1.52 at 633 nm and 1064 nm respectively. The increase in the refractive index between sample A and B1 (respectively with low and high amounts of titanium) is attributed to the doubling of the titanium molar ratio during the film preparation. High inhomogeneities for refractive indices were measured in samples B2 and B3, due to anisotropy in the films. This anisotropy can be explained by a consequence of the high annealing temperature, which may achieve stresses in the deposited film during hardening. M-lines measurements confirmed the presence of anisotropy in samples B2 and B3. In the case of a 160°C annealing temperature, a mean ordinary refractive index of 1.572 and a mean extraordinary refractive index of 1.569 were evaluated: the influence of the annealing temperature on the increase of the refractive index seems to be not negligible.
The heat capacity of the silica/titanium sol-gel films has been measured by differential scanning calorimetry (DSC), TA Instruments model TAG 1000. The measured heat capacity for sample A is 1.55±0.60 Jg-1K-1 at 25°C. For comparison, this value is similar to the 1.46 Jg-1K-1 of pure polymethylmetacrylate (PMMA) [12]. Low heat capacity of sol-gel material has many interests for laser micromachining since working with low laser fluence and high speed is then accessible. The thermal stability is high for a hybrid sol-gel material: damage has been observed from 270°C [26]. It is also of interest to estimate the heat affected zone (HAZ) of the sol-gel material in case of laser ablation. The HAZ is given by expression (1) where D is the thermal diffusivity and τ the pulse duration (6 ns with our KrF* excimer laser). The measured heat capacity of the sol-gel film is close to the value of PMMA. Our sol-gel material is however an organic/inorganic matrix, close to silica. To estimate its thermal conductivity, we can consider that the value is comprised between that of pure silica (1.4 W.m-1.K-1) and that of PMMA (0.2 W.m-1.K-1). Also for the sol-gel density, we can consider an intermediate value between pure silica (2.2 g.cm-3) and PMMA (1.2 g.cm-3). This leads to an estimation of the thermal diffusivity of about 10-7 m2/s. With pulse duration of 6 ns, the estimated HAZ is then about some tens of nanometers.
In order to use the sol-gel material with high power lasers in the infrared spectrum, sample resistance to laser fluence has been tested at 1064 nm. The source was a Nd:YVO4 laser with a 15 ns pulse duration at a frequency of 50 kHz and a 600 µm diameter beam (HIPPO model Spectra-Physics). No deteriorations were observed on a 3 µm thick sol-gel film of type B1 after 5 minutes exposure to a mean irradiance of 6500 W/cm2, corresponding to a peak power of 24500 W.
3. Micro-structuring of hybrid sol-gel substrate using laser ablation
Our experimental setup employs a ProMaster laser micromachining station from Optec implemented with a 248 nm pulsed ATL KrF* excimer laser. An optical projection system, containing an Optec three lens imaging system optimized for a wavelength of 248 nm, is exploited with a reduction factor of ten. The optical numerical aperture is 0.105 and the practical resolution is near to 1.5 µm. The maximum emitted fluence is 20 mJ, which can be controlled by a calibrated Optec attenuator (model AT 4020). The optical setup is presented on Fig. 2.
The sample to be machined is fixed via a vacuum chuck on XY moving stages. The working resolution of the moving stages is about 1 µm and the repeatability is typically between 2 and 5 µm. The pulse duration is 6 ns, the repetition rate ranges from 1 to 300 Hz and the output energy density at the workplace ranges from 0.09±0.04 J/cm2 to 5.10±0.30 J/cm2. Gas flow (helium or argon) assistance may be usable with an adapted injector nozzle. This is employed in order to reduce debris re-deposition on the processed area. The pattern ablated on the sample reproduces the shape of the mask aperture. Several mask patterns, with characteristic dimensions that range between 2500 µm and 100 µm, are available and fabricated with a photolithographic process and multilayer deposition such as an Aluminium-Chromium-Fused Silica mask. Since micro-optics commonly consist of a pattern composed of small elementary square cells, our masks contain only a single square having a side dimension corresponding to the elementary cell size taking into account the magnification of the projection setup. The whole pattern of a kinoform structure for example is made by scanning the substrate using the micropositionning stages and ablating each elementary cell at the required positions. This solution is well adequate to rapid prototyping and small productions. No homogenization system of the excimer laser beam has been employed in our system.
The process to demonstrate the multi-level structuring of a substrate is simple: the laser fluence is defined and changed with the calibrated attenuator. Several impacts are then realized by varying only the number of laser shots to show the ability of multi-level structuring. This consists also in the calibration of the substrate. The ablation depends on the attenuation and each laser shots plays then the role of an etching tool. The possibility of electronic control of the Optec attenuator is the main difference with other ablation techniques. By this way, different depths can be ablated inside of the sol-gel substrate using calibration curve, and the whole structure can be patterned in either a single step (by modifying the fluence in real time) or by multi-shots (by etching the different steps with several shots). The profile carried out in the sol-gel layer is measured by coherence probe microscopy (CPM) [27] after the fabrication process. This non destructive profilometry technique is capable of giving precise results, rapidly and easily. It requires no sample preparation and proposes precise measurements on many different types of material, including transparent layers greater than 0.5 µm in thickness. Figure 3(a) gives an example of a measured profile after laser ablation in a sol-gel layer deposited on a glass substrate. This test demonstrates that the wall angle is very sharp and the roughness at the surface of the sol-gel film is low. The measured R t roughness (peak-valley) is 10.6 nm for the film surface. Very small defects and debris are observed. The value of the wall angle (86°) is higher than the optimal angle (75°) calculated by rigorous electromagnetic tools like Fourier modal method for the standard aim of diffractive optical elements (maximal diffraction efficiency in the first order in the case of kinoforms) [1]. Seven shots are required with a fluence of 0.9 J/cm2 to go through the 3.04 µm thick sol-gel film.
Fig. 3. (a) Profile with a very sharp edge after laser ablation in the sol-gel substrate. The estimation of the wall angle is 86°. The thickness of the film is 3.04 µm. The laser fluence is 0.9 J/cm2. (b) Profile of the sol-gel film etched with one to five laser shots at 0.9 J/cm2. Insert: measured ablated depth as a linear function of the number of laser shots.
Figure 3(b) presents different profiles realized with multi-shots in a sol-gel layer. A first observation can be made on the irregular zones in the bottom of the ablated zones. These irregularities are comparable in the different shots visible in the Fig. 3(b): we explain the progression of this error by the low beam quality of the excimer laser. Different tests confirm this approach, and the sol-gel material and its preparation have no incidence of theses irregularities. The roughness achieved by the excimer laser ablation process is increased with the root square of the number of shots as demonstrated in [28], and in our case, the defaults depend not on the type of sol-gel. For small thickness, this roughness has no dramatic consequences in the reconstruction, however noise is observed as we will see in the next paragraphs. A larger depth is current on the circumference of the ablated pattern with our micromachining station. The geometrical characteristics depend on the fluence and on the laser number of pulses which arrive on the film. The bottom total roughness after a laser shot of 0.9 J/cm2 is approximately 70 nm. These defects are probably coming from problems of homogeneity but also on hot spots in the laser beam. In order to obtain the calibration curves and the ablation rate as a function of the laser fluence, series of several impacts were carried out between 0.3±0.1 J/cm2 and 1.6±0.1 J/cm2. The ablated depth taken into account is the mean value of three shots carried out with equivalent experimental conditions. Generally, series of one to five shots with same fluence are used in our process as shown in Fig. 3(b). Inserted in Fig. 3(b), the measured ablated depths present a linear behaviour with the number of laser shots. This has been observed for the whole investigated range of fluences. The best resolutions are achieved with the lowest fluence (0.3 J/cm2), where the ablated depth varies from 0.13±0.03 µm to 0.40±0.08 µm for one to five laser shots. This profile demonstrates that the control of the ablated depth is achieved with a good precision and can provide multi-level structures in a sol-gel material.
4. Influence of process parameters
The ablation rate as a function of the laser fluence is given in Fig. 4 in the case of two types of titanium ratio. If the fluence is lower than 0.3 J/cm2, the sol-gel film is not etched by ablation with the KrF* excimer laser. Between the two sol-gel compositions, the one with the highest titanium ratio (sample B1) has a higher density. Different behaviours of the ablation rate in the UV spectrum are described in the literature, and especially for organic polymers [28–30]. The logarithmic regime for low fluence close to the ablation threshold is commonly accepted with the expression (2) where d is the ablation rate (µm/shot), F is the fluence, F T is the ablation threshold and αeff is effective absorption coefficient.
For sample B1, the ablation threshold F T after a fitting operation of the experimental data is 0.12±0.01 J/cm2 and the effective absorption coefficient α eff is equal to 6.7±0.4 µm-1. In the case of sample A (low titanium ratio), the ablation threshold F T is 0.13±0.01 J/cm2, and the effective absorption coefficient α eff equal to 2.9±0.2 µm-1. For polymers etched by excimer laser, the ablation threshold is small, generally between 20 to 200 mJ/cm2, and the effective absorption coefficient is found between 1 and 10 µm-1 [28]. As the two hybrid sol-gel materials synthesized in this study present coefficients in this range of values, we can assume that their ablation behaviour at low fluence is close to that of polymers. The effective absorption coefficient used in the expression (1) is generally different from the absorption coefficient of the material (calculated with the extinction coefficient): this coefficient has apparently no notable influence on the ablation characteristics [29,30].
Fig. 4. Ablation rate as a function of the laser fluence for two sol-gel film compositions: sample B1 with a high titanium doping and sample A small titanium doping.
Figure 5 presents the etched depths in sol-gel films annealed at three different temperatures: 100°C, 130°C or 160°C. The depth decreases if the annealing temperature increases. For a laser fluence equal to 0.9 J/cm2, the ablated depth with five laser shots is 1.25±0.11 µm in the film annealed at 100°C, 1.11±0.10 µm in the film annealed at 130°C and 1.00±0.06 µm in the film annealed at 160°C. The difference is mainly significant when the number of laser pulse is increased. One can notice that an increased annealing temperature smoothens up the sol-gel surface, reducing surface defects and increases volume homogeneity. The annealing temperature is a process parameter which is easy to modify in order to regulate the ablation rate with a better precision.
Fig. 5. Depth etched as a function of the laser shots with a fluence of 0.9 J/cm2. Different cases of annealing temperature are compared in this figure.
In order to characterize the influence of the assistance gas, three series of tests were carried out for one to five laser shots at 0.9 J/cm2 with respectively argon under 1 bar pressure, compressed air under 1 bar pressure and without assistance gas. The gas is brought in the machining zone with a coaxial tube. A pressure of one bar is common in laser micromachining processes. Table 1 gives the ablated depth as a function of the number of laser shots. All parameters being identical, the minimum etched depth is obtained with the argon gas whereas the maximal value is achieved without assistance gas. With five laser shots, the ablated depth is 1.21±0.07 µm with argon gas, 1.28±0.11 µm with compressed air, 1.36±0.12 µm without assistance gas. The standard deviations of the depth measurements are the smallest when the argon is used as assistance gas. The consequences are a better precision and a good repeatability in the laser micromachining process of sol-gel materials. In fact, the assistance gas does not have a dramatic influence on the final result: this is the reason why we did not use the gas to assist the micromachining process of the realized DOEs.
Table 1. Depths etched in a sol-gel layer as a function of the number of laser shots with a fluence of 0.9 J/cm2. Comparison between different assistance gases used during the ablation process.
5. Surface relief etching and diffractive optical elements production
Figure 6 shows common squares etched in a sol-gel layer (optical microscope images). Each one will constitute the elementary pattern (or pixel) for the diffractive optical element. The lateral size is 50 µm, 25 µm and 10 µm respectively from left to right. In the fabrication process of diffractive optics to be used at 1 µm, the patterns having 10 µm and 25 µm lateral sizes are mainly employed. The surface of the etched sol-gel material is not strongly modified by the waste, as visible in Fig. 3(a): the material vaporisation is then very fast and has minimal repercussions. The form of the smallest pixel appeared extremely sensitive to a variation of the sample positioning out of the focal plan.
Fig. 6. Optical microscope images of square patterns etched in a sol-gel film. Dimensions of the square from left to right are 50 µm, 25 µm and 10 µm.
A movie of the laser micromachining process of a 4-level surface relief diffractive optical element is available. Figure 7 is a representative frame from this movie. The elementary pattern is 50 µm square.
Fig. 7. Frame of the movie showing the laser micromachining process of a 4-level surface relief diffractive optical element. File size: 4 Mb. Media 1.
Typical examples of multi-level kinoforms carried out in a sol-gel film are shown on Fig. 8. A first 32×32 pixel element with 4 levels is visible on Fig. 8(a) (optical microscope image) and a 3D profile measured with CPM is presented on Fig. 9 where the three machined depths are quite distinct. One can note in Fig. 9 the small spacing between two consecutive pixels, due to an ablated elementary pattern smaller than the specifications. The spacing is here 2 µm for a pixel dimension of 25 µm. This spacing can be easily eliminated by taking it into account in the manufacturing process, by modifying the coordinates of the pixels to be ablated. An 8-level structure machined with a 10 µm pixel in a sol-gel layer is visible on Fig. 8(b). The coloring of the sol-gel film observed in the etched zones is achieved by the oxidation of the titanium present in the material. The use of titanium oxide allows us to decrease the ablation rate. If necessary, it would be possible to remove the titanium from the sol-gel formulation, while working on a suitable annealing (temperature and duration) of the material to obtain the desired condensation.
Fig. 8. (a) Optical microscope image of a 4-level 32×32 pixel DOE machined in a sol-gel layer. The size of the pixel is 25 µm. (b) Detail of an 8-level structure realized in a sol-gel layer. The size of the pixel is 10 microns.
The surface relief of the diffractive optics used in transmission is related to the phase profile calculated to obtain the required optical function. For a surface relief discretized in N levels, the depth h corresponding to one step is given by expression (3), for a refractive index n at the wavelength λ:
In case of a 4-level DOE machined in a sol-gel layer, the measured depths for each step are shown in table 2. The mean depth measured for the second and third levels are in the tolerance of ±25 nm. In the case of the first level, the mean depth is slightly out this interval of acceptable tolerance. As still observed on the calibration curve of sol-gel materials, the roughness increases linearly with the ablated depth.
Table 2. Comparison between measurements and theoretical depths in the case of 4-level DOE machined in a sol-gel layer.
Fig. 9. 3D profile measured by Coherence Probe Microscopy of a 4-level 32×32 pixel phase DOE machined in a sol-gel layer. The size of the pixel is 25 microns.
6. Reconstruction with diffractive optical elements etched in sol-gel substrate
The different realized sol-gel DOEs were tested in the infrared spectrum, with a Nd:YVO4 solid-stated laser emitting at 1.064 µm (HIPPO model Spectra-Physics). Figure 10 presents different cases of reconstructions for DOEs machined with a 10 µm pixel. A 100 mm focal length lens is placed directly after the diffractive component. We can then naturally imagine machining DOEs on a sol-gel layer directly deposited on a plano-convex lens. The reconstruction is visualized by means of a plane screen placed on the optical axis and recorded by a simple CCD camera (Basler model 104k). The pulsed laser emission (frequency 30 kHz, pulse duration 15 ns) has been chosen in a low mean power situation (1.50 W). The laser beam is lineary polarized and its diameter is 600 µm. The incident beam power density is 530 W/cm2. The M2 beam quality factor is specified smaller than 1.2 by Spectra-Physics. The divergence is lower than 3.0 mrad. The beam has been employed directly at the output of the cavity, without spatial filtering.
Two reconstructions of a pattern containing eight dots on an off-axis circle are realized with this laser. The reconstructions of the 128×128 binary and 4-level machined DOEs are visible respectively in Fig. 10 (a) and (b), where the encoded forms are clearly visible. The DOEs have been machined with a 10 µm pixel. On Fig. 10(a), the pattern is visible on both sides of the zero order with quite the same intensity, which is normal for a binary DOE. The zero order is very important and the diffraction efficiency is here approximated to some percents with a large noise. Theoretical diffraction efficiency estimated from the scalar diffraction theory is 40.5 % in the first diffracted order (order of interest) for a 2-level surface relief; it is 81 % in the case of a 4-level surface relief. One can notice that the incident beam diameter does not optimally match the dimension of the diffractive components. However, the reconstruction of DOE machined by laser ablation in sol-gel substrate is demonstrated. The blur in the reconstruction can be explained with the roughness and the micromachining errors in the realized structures.
Fig. 10. (a) Reconstruction with a Nd:YVO4 laser of a binary 128×128 pixels DOE containing 8 dots on an off-axis circle. (b) Reconstruction with a Nd:YVO4 laser of a 4-level 128×128 pixels DOE containing 8 dots on an off-axis circle. The pixel size is 10 µm in both cases.
7. Conclusion
We bring the evidence that diffractive optical components can be fabricated in a hybrid sol-gel material by ablation process, evidence that is not found at present in the literature. The characterization and the preparation of the sol-gel material have been presented in detail to specify the advantages of this material. The micro-structuring with excimer laser and the ablation process have also been analyzed with the ablation rate as a function of the laser fluence, to match a sufficient control in the etching depth. The influences of the titanium ratio and different conditions of use, in particularly the annealing process, have also been shown. Examples of diffractive optical elements with surface relief modulation have been carried out and analysed by coherence probe microscopy. Finally the reconstructions demonstrate also that our solution is low cost and flexible. One can rapidly envisage prototypes for different wavelengths. These results need now to be confirmed with infrared laser beams operating at industrial applications.
Acknowledgements
The authors are grateful to P. Meyrueis and E. Fogarassy for providing insights on this subject and fruitful discussions. The authors would like to thank InESS Laboratory for financial support of this work and D. Montaner for helpful discussions. We also acknowledge the use to different high power lasers IREPA LASER research centre, Pôle API, 67400 Illkirch-Graffenstaden, France.
References and links
1. P. Meyrueis and B. Kress, Digital Diffractive Optics (Chichester Wiley, 2000).
2. T. Levola, “Diffractive optics for virtual reality displays,” J. Soc. Inf. Disp. 14, 467–475 (2006). [CrossRef]
3. O. Cakmakci and K. Rolland, “Comparative analysis of doublets versus single-layer diffractive optical elements in eyepiece or magnifier design,” Appl. Opt. 46, 8140–8148 (2007). [CrossRef] [PubMed]
4. J. Yang, X. Su, and P. Xu, “Perfect shuffle transform based on a microblazed grating array,” Appl. Opt. 46, 210–215 (2007). [CrossRef] [PubMed]
5. C. Haupt, M. Palhke, R. Krupka, and H. J. Tiziani, “Computer generated microcooled reflection holograms in silicon for material processing with CO2 laser,” Appl. Opt. 36, 4411–4418 (1997). [CrossRef] [PubMed]
6. T. G. Jabbour and S. M. Kuebler, “Axial field shaping under high-numerical-aperture focusing,” Opt. Lett. 32, 527–529 (2007). [CrossRef] [PubMed]
7. J. S. Liu, A. J. Caley, and M. R. Taghizadeh, “Diffractive optical elements for beam shaping of monochromatic spatially incoherent light,” Appl. Opt. 45, 8440–8447 (2006). [CrossRef] [PubMed]
8. A. Hermerschmidt, S. Krüger, and G. Wernicke, “Binary diffractive beam splitters with arbitrary diffraction angles,” Opt. Lett. 32, 448–450 (2007). [CrossRef] [PubMed]
9. M. Ekberg, M. Larsson, A. Bolle, and S. Hard, “Nd:YAG laser machining with multilevel resist kinoforms,” Appl. Opt. 30, 3604–3606 (1991). [CrossRef] [PubMed]
10. K. Fuse, K. Ebata, T. Okada, K. Kurisu, and M. Shiozaki, “Optical properties of ZnSe diffractive optical elements for spot array generation”, in Proceedings of International Congress on Applications of Lasers and Electro-Optics (2000), pp 179–186.
11. M. T. Gale, M. Rossi, and J. Pedersen, “Fabrication of continuous relief micro-optical elements by direct laser writing in photoresists,” Opt. Eng. 33, 3556–3566 (1994). [CrossRef]
12. E. B. Kley, “Continuous writing profile by electron and optical lithography,” Microelectron. Eng. 34, 261–298 (1997). [CrossRef]
13. N. A. Vainos, S. Mailis, S. Pissadakis, L. Boutsikaris, P. J. M. Parmiter, P. Dainty, and T. J. Hall, “Excimer laser use for microetching computer-generated holographic structures,” Appl. Opt. 35, 6304–6319 (1996). [CrossRef] [PubMed]
14. G. P. Behrmann and M. T. Duignan, “Excimer laser micromachining for rapid fabrication of diffractive optical elements,” Appl. Opt. 36, 4666–4674 (1997). [CrossRef] [PubMed]
15. K. Callewaert, Y. Martelé, L. Breban, K. Naessens, P. Vandaele, R. Baets, G. Geuskens, and E. Schacht, “Excimer laser induced patterning of polymeric surfaces,” Appl. Surf. Sci. 208–209, 218–225 (2003). [CrossRef]
16. Y-T. Chen, K. Naessens, R. Baets, Y.-S. Liao, and A. A. Tseng, “Ablation of Transparent Materials Using Excimer Lasers for Photonic Applications,” Opt. Rev. 12, 427–441 (2005). [CrossRef]
17. A. A. Tseng, “Recent developments in micromachining of fused silica and quartz using excimer lasers,” physica status solidi (a) 204, 709–729, (2007). [CrossRef]
18. R. Böhme, D. Hirsch, and K. Zimmer, “Laser etching of transparent materials at a backside surface adsorbed layer,” Appl. Surf. Sci. 252, 4763–4767 (2006). [CrossRef]
19. J. Ihlemann, “Ultraviolet laser ablation patterning of oxide films for optical applications,” Opt. Eng. 44, 0511081–0511085 (2005). [CrossRef]
20. P. J. Moreira, P. V. S. Marques, and A. P. Leite, “Hybrid Sol-Gel Channel Waveguide Patterning Using Photoinitiator-Free Materials,” IEEE Photon. Technol. Lett. 17, 399–401 (2005). [CrossRef]
21. S. I. Najafi, T. Touam, R. Sara, M. P. Andrews, and M. A. Fardad, “Sol-Gel Waveguide and Grating on Silicon,” J. Lightwave Technol. 16, 1640–1646 (1998). [CrossRef]
22. D. Blanc and S. Pelissier, “Fabrication of sub-micron period diffraction gratings in self-processing sol-gel glasses,” Thin Solid Films 384, 251–253 (2001). [CrossRef]
23. W. X. Yu, X.-C. Yuan, N.Q. Ngo, W.X. Que, W.C. Cheong, and V. Koudriachov, “Single-step fabrication of continuous surface relief micro-optical elements in hybrid sol-gel glass by laser direct writing,” Optics Express 10, 443–448 (2002). [PubMed]
24. W. C. Cheong, X.-C. Yuan, V. Koudriachov, and W. X. Yu, “High sensitive SiO2/TiO2 hybrid sol-gel material for fabrication of 3 dimensional continuous surface relief diffractive optical elements by electron-beam lithography,” Optics Express 10, 586–590 (2002). [PubMed]
25. L. Li, L. Qinghua, Y. Jie, Z. Zikang, P. Daocheng, and W. Zongguang, “Preparation and properties of photosensitive polyimide/titania-silica hybrid materials,” Mater. Sci. Eng. C 22, 61–65 (2002). [CrossRef]
26. R. Buestrich, F. Kahlenberg, M. Popall, P. Dannberg, R. Müller-Fiedler, and O. Rösch, “ORMOCERs for Optical Interconnection Technology,” Journal of Sol-Gel Science Technology 20, 181–186 (2001). [CrossRef]
27. P.C. Montgomery, A. Benatmane, E. Fogarassy, and J.P. Ponpon, “Large area, high resolution analysis of surface roughness of semiconductors using interference microscopy,” Mater. Sci. Eng. B 91–92, 79–82 (2002). [CrossRef]
28. P. E. Dyer, “Excimer laser polymer ablation: twenty years on,” Appl. Phys. A 77, 167–173 (2003).
29. T. Lippert, M. Hauer, C.R. Phipps, and A. Wokaun, “Fundamentals and applications of polymers designed for laser ablation,” Appl. Phys. A 77, 259–264 (2003).
30. T. Lippert, Polymers and Light, Advances in Polymer Science 168 (Springer, 2004).
