Open Access
Add to BibSonomyAbstract
We report the use of laser ablation of metal targets onto a glass substrate as a way of producing waveguiding devices. In the geometry employed, the nanosecond pulses used for the ablation pass through the glass substrate, and are focused on the metal surface, which is located in close proximity with the substrate. We present measurements of the refractive index profile obtained with this technique, and present a discussion of the physical mechanisms that produce the profiles measured.
©2006 Optical Society of America
1. Introduction
Glass waveguides are suitable candidates for a class of passive integrated optical components, such as star and access couplers, power combiners, and wavefront sensors. They are also helpful in certain switching applications requiring low speeds. The use of glass as the building block for such planar waveguiding devices is attractive for reasons such as low loss, immunity to optical damage, compatibility with commercial fibre interfacing, as well as potential low production costs.
A wealth of methods has been proposed for fabricating integrated waveguiding devices, ion exchange [1], ion implantation [2, 3, 4] lithographic methods, and more recently optical damage by ultrafast laser pulses [5, 6]. Another method that has been recently proposed for making waveguiding structures is the laser backwriting technique [7]. In this technique, the output of a pulsed laser is focused onto the surface of a metal target which is placed in close proximity to the back surface of the glass substrate. The laser pulses, which actually travel through the substrate before reaching the surface of the metal, ablate the metal creating a plasma with the detached ions, which then travel back to the substrate face. The expected effect is the inclusion of the ablated metallic ions into the substrate, without significant damage to the surface of the substrate. The inclusion of the metal ions, and the associated induced stress in the glass matrix will generate the desired index change. The main advantage of this method is that it allows the direct writing of structures such as dividers, directional couplers, et cetera, without the need for complicated lithographic procedures. Ablation of metal targets in close proximity to a substrate has been used in a similar way in the laser induced forward transfer (LIFT) technique [8], but for other purposes.
On the other hand, the fabrication of optical waveguides and other integrated optics components requires a detailed knowledge of the material parameters involved. In particular, the precise determination of the refractive index profile is important for a characterization of waveguiding properties, for a proper control of the fabrication process, and as a basis for numerical calculations of the performance of such waveguiding devices.
Several methods have been proposed for determining the refractive index profile of gradient index elements, including bound near-field, axial and transverse interferometric, and the refracted near-field methods. Among all these, the refracted near-field method [9, 10] presents several advantages over other methods: high resolution and accuracy, minimal sample preparation needs, straightforward interpretation of results, and being able to measure bidimensional profiles of different structures.
In this article we present experimental results for the localmodification of the refractive index of glass substrates by a laser backwriting technique. We also present a characterization of the resulting refractive index profile, using the refracted near-field (RNF) technique. A discussion is also made of the possible mechanisms that generate the observed refractive index profile.
2. Experimental results
Waveguides are produced by focusing nanosecond laser pulses onto a metallic target which is in close contact with a glass substrate, as shown in Fig. 1. The substrate employed was a 2mm thick pyrex slide. Pyrex is a borosilicate glass, its composition being approximately 80% SiO2, and 13% B2O3, with a small percentage of other oxides. The target was a 1mm thick stainless steel foil placed in contact with the substrate. The light used to perform the ablation was provided by a diode-pumped Q-switched Nd:YAG laser (Rofin, E-line model) producing 12ns (FWHM) pulses at the fundamental wavelength (1064nm) with a repetition rate that can be adjusted in the 5–50KHz range. The laser is directed into the sample and tightly focused onto the metal-glass interface using a lens with a 60mm focal length. The position of the irradiation spot on the target is controlled by a galvanometer system, that scans the beam at a constant speed of 50mm/s. In this way, a straight line of ablated material was deposited onto the glass surface, and into the glass matrix. The translation speed and pulse energy were controlled as to have a good metal removal rate while minimizing damage to the glass surface. This was corroborated by measuring the surface profile of the substrate with a surface profilometer.
After laser exposure, the samples were treated to remove any metallic thin films that could have formed in the surface of the substrate. This was done by placing the samples in an aqua regia bath for 1 hour. Refractive index profiles were measured before and after the cleaning procedure, without finding any significant differences.
The resulting refractive index profile was measured using the RNF technique. The technique consists of end-coupling light to the waveguide with a microscope objective having a numerical aperture larger than that of the waveguide. The substrate containing the waveguide is supported on a glass block (see Fig. 2), with the waveguide facing the block, and an index matching fluid layer between the block and the substrate. In this way, part of the light will be guided, while part will escape the guide, and pass through the supporting block to form a hollow cone. The inner part of this hollow cone contains leaky modes whose contribution is difficult to assess, whereas the outer part contains purely refracted light. Because of this, a disk is included to block the leaky modes, and to allow only the refracted light to pass. A set of lenses is used to gather the refracted light cone, which is made to impinge onto a large area detector. Under these conditions, and assuming that the refractive index change δn(r), defined through n(r)=ns+δn(r) (where ns is the refractive index of the substrate), is small, it is possible to show [9] that the power that crosses the disk P(θ) is approximately proportional to the refractive index at the point of incidence, that is:
where a 1 and a 2 are constants that depend on the specific experimental conditions, the actual values of these constants having to be determined by a calibration procedure. The calibration is easily achieved by correlating the observed power levels corresponding to the known refractive indices of the immersion liquid, and either one of the glass block, or the guide substrate. The accuracy achieved depends on how accurately these reference values are known. In the present case we chose to use the refractive index of the glass block since it is made of pyrex glass as well (n=1.473 at λ=589.3nm), the immersion liquid we used has a reported nD=1.450±0.0002 for the sodium D line (λ=589.3nm) at 20°C. The refractive index value has to be corrected for the laser wavelength employed, λ=632.8nm in this case, and this is done by a linear interpolation with the nC, and nF values of the liquid.
Figure 3 shows a calibrated raster scan data set obtained for a waveguide produced by the method previously described. The fabrication parameters employed were: a laser scanning rate of 50mm/s, a laser repetition rate of 25KHz, an energy per pulse of 124µJ, and the metal target was a stainless steel foil. Figures 4 and 5 show linear profiles measured across the waveguide (in the z direction), and parallel to the substrate surface (x direction), respectively. The spa-tial resolution of the set-up can be deduced from the sharpness of the recorded profile at the waveguide-liquid interface.
Fig. 3. Measured bidimensional refractive index profile. The uniform region to the left corresponds to the glass block, while the vertical thin low n value strip corresponds to the immersion liquid.
Fig. 4. Measured refractive index profile crossing from the liquid, into the waveguide, and into the substrate (z direction). The dotted line shows the extrapolated refractive index values of the glass block, matching liquid and substrate.
Fig. 5. Measured refractive index profile parallel to the substrate surface, about 5mm into the waveguide (x direction).
3. Discussion
Figure 3 shows that a refractive index change was obtained in two well defined regions of the substrate, one region confined close to the substrate surface, and another less localized, with a longer penetration depth. This is easier to observe in Fig. 4, that shows a well defined peak with a ~10µm width, and another, smaller maximum further into the substrate, that tapers very slowly. There are some features localized around the x=120µm, z=140µm region, but these are very probably due to imperfections on the polished side face of the substrate. The maximum refractive index change observed is approximately δn≃0.02. A section of the index profile, taken parallel to the substrate surface, displayed in Fig. 5, shows an approximately gaussian index change with a 60µm width (HWHM). The gaussian cross section observed for the refractive index correlates well with the gaussian distribution of the ablating beam.
One possible effect of the fabrication process is the alteration of the surface by the deposition process. Surface profilometry scans were performed using a commercial surface profilometer (Dektak3 profilometer). These scans show that while there is some surface damage, this does not exceed 100nm in height.
The observed refractive index change is believed to arise from the inclusion of metallic ions from the plasma produced by the laser irradiation, into the glass matrix. Spatially resolved energy dispersive x-ray analysis(EDX), seems to be consistent with this. The EDX analysis was carried at a few selected points on the substrate end face, close to the edge, and around the area where laser irradiation was performed. The data in table 1 shows that significant concentrations ofCr, Fe,Cu and Zn ions, was only found from analysis at points 1,2,5 and 7, shown in figure 6. All these points lay within 5µm from the edge, which correlates very well with the observed refractive index change.
The more extended, weaker refractive index change observed can be due to a glass matrix modification produced by the free electrons in the laser produced plasma, which could penetrate deeper into the substrate. Direct optical damage produced by the incoming focused laser beam could also explain this feature, but more work should be done to clarify this point.
In order to see if the broad region of index change could be removed, thermal annealing of the waveguide was performed, in the form of repeated heating cycles with soaking at 450°C for 5 hours, but no significant changes were observed.
While the refractive index change achieved is large enough to ensure that waveguiding of light takes place, we have not been able to obtain efficient out-coupling of light from the sample. This is probably due to the relatively large area of the modified region (10×60µm 2), and to the presence of a second broader index change region that penetrates deep into the substrate. The lateral dimensions of the waveguide (in the x-direction) can be reduced by a more careful focusing of the pulses onto the target surface. Different combinations of pulse energy and scan rate can be used to control the actual refractive index change, the penetration depth, and the longitudinal uniformity of the index change. It also remains to be seen whether annealing can remove the more extended refractive index change area, which is probably the most important obstacle for light confinement in the z direction. Optimization of these parameters is underway and will probably allow the production of efficient waveguiding structures.
Table 1. Chemical composition (given as percentages) of the zones analyzed at different distances from the substrate edge.
4. Conclusions
In conclusion, we present a novel technique for the production of waveguiding structures in glass substrates, based on laser ablation of a metallic target, and we show the refractive index profile characterization of a waveguide produced by this method. We find a relatively large (δn~0.02) refractive index change in a 10×60µm 2 region, and a second weaker, more extended region. The refractive index change observed is associated with the inclusion of metallic ions from the target into the substrate glass matrix. This simple, dry method, can become an interesting alternative for producing different waveguiding structures.
Acknowledgments
This work has been supported by Ministerio de Educación y Ciencia, Spain, under contract FEDER/TIC2003-03041. Antonio Castelo has been supported by a Formación de Personal Investigador (FPI) grant. The GRIN Optics Group of the University of Santiago de Compostela, Spain, is a NEMO/EU (Network of Excellence on Micro-Optics) partner. Germán de la Fuente would like to acknowledge a fellowships received from Ordenación Universitaria. Raúl Rangel-Rojo wants to thank la Xunta de Galicia, Consellería de Educación for awarding him a visiting researcher grant, and CONACYT-México for partial funding through project no. 46492.
References and links
1. S.I. Najafi, “Introduction to Glass Integrated Optics,” Artech House Inc. Massachusetts, 1992.
2. E. Flores-Romero, G. V. Vázquez, H. Márquez, R. Rangel-Rojo, J. Rickards, and R. Trejo-Luna, “Planar waveguide lasers by proton implantation in Nd:YAG crystals,” Opt. Express 12, 2264–2269 (2004). [CrossRef] [PubMed]
3. K. Liu, E.Y.B. Pun, T.C. Sum, A.A. Bettiol, J.A. van Kan, and F. Watt, “Erbium-doped waveguide amplifiers fabricated using focused proton beam writing,” Appl. Phys. Lett. 84, 684–686 (2004). [CrossRef]
4. F. Chen, X.L. Wang, X.S. Li, L.L. Hu, Q.M. Lu, K.M. Wang, B.R. Shi, and D.Y. Shen, “Ion-implanted waveguides in Nd3+-doped silicate glass and Er3+/Yb3+ co-doped phosphate glass, Appl. Surf. Sci. 193, 92–101 (2002). [CrossRef]
5. K.M. Davis, K. Miura, N. Sugimoto, and K. Hirao, “Writing waveguides in glass with a femtosecond laser,” Opt. Lett. 21, 1729–1731 (1996). [CrossRef] [PubMed]
6. S. Nolte, M. Will, J. Burghoff, and A. Tuennermann, “Femtosecond waveguide writing: a new avenue to three-dimensional integrated optics,” Appl. Phys. A. 77, 108–11 (2003). [CrossRef]
7. A. Castelo, D. Nieto, M.T. Flores-Arias, C. Gómez-Reino, and X. de la Fuente, “Laser backwriting process on glass via ablation of metal targets,” in Lasers and Applications, Proc. SPIE 5958, K.M. Abramski, A. Lapucci, and E.F. Plinski, eds., 600–607, (2005).
8. J. Bohandy, B.F. Kim, and F.J. Adrian, “Metal deposition from a supported metal film using an excimer laser,” J. Appl. Phys.60, 1538–1539.
9. R. Göring and M. Rothhardt, “Application of the refracted near fied technique to multimode planar and channel waveguides in glass,” J. Opt. Commun. 7, 82–85 (1986). [CrossRef]
10. P. Oberson, B. Gisin, B. Huttner, and N. Gisin, “Refracted near-field measurements of refractive index and geometry of silica-on-silicon integrated optical waveguides,” Appl. Opt. 37, 7268–7272 (1998). [CrossRef]
