Ultrafast Laser Inscription of Buried Waveguides in W-TCP Bioactive Eutectic Glasses

Dr. Daniel Sola thanks the Ministry of Economy and Competitiveness of the State General Administration under the project MINECO FIS2016-76163-R and the PIT2 program of the University of Murcia’s own research plan for the financial support of his contract.


Introduction
Since its invention by T.H. Maiman in 1960, the laser interaction with matter has been successfully investigated to find and develop new applications in fundamental, applied, technological and industrial research [1]. Laser materials processing is a versatile tool for materials synthesis, modification and structuring involving in many cases multidisciplinary approaches. Laser technology has been studied in the search for novel applications in research fields such as optics, photonics, energy, microelectronics, aerospace and biomedicine [2][3][4][5][6][7][8][9][10][11][12].
The acronym "LASER" stands for "Light Amplification by Stimulated Emission of Radiation" and was first introduced in the late 1950s [13]. Essentially, a laser device is a resonator which emits coherent light originated in a stimulated emission process. A laser resonator or oscillator is usually made up of three main parts: a gain medium which amplifies light by stimulated emission, a pumping source to provide external energy to stimulate the atoms of the gain medium to its excited states, and an optical cavity which supplies feedback to the laser light. The resulting radiation, the laser light, has unique properties such as a high degree of temporal and spatial coherence, and low divergence which confers it of exceptional advantages compared to conventional light sources [14].
Along the last 50 years, laser technology has progressed based on the advances achieved in the development of new gain media, pumping sources and cavity design, leading to many different types of lasers. Attending to the gain medium, lasers can be classified as gas laser, chemical laser, excimer laser, solid-state laser, semiconductor laser, dye laser, free electron laser and metal-vapor laser. According to the laser operation wavelength, the classification can also be as ultraviolet (UV) laser, visible and infrared (IR) laser. In particular, for an operation wavelength between 700 nm and 2 μm, the laser is referred as near-infrared (NIR) laser. Also, the classification can be carried out attending to the operation mode; continuous wave (cw) and pulsed lasers [14]. The output power of a cw laser is constant, whereas for a pulsed laser, the power varies pulse-by-pulse. In this latter case, the pulse energy, E p , and the peak power, P p , can be calculated as: where P out is the output power, f is the repetition rate and Δt the pulse duration. Therefore, an appropriate combination of both repetition rate and pulse duration may allow pulsed lasers to achieve peak powers much higher than cw lasers. To date, the peak power that a pulsed laser can reach is in the range of a Petawatt with extremely high intensities, in the order of 10 19 Wcm −2 , much higher than the atomic unit of intensity (3.5 10 16 Wcm −2 ). Hence, ultrashort pulsed lasers have recently attracted scientist's attention aiming at developing novel applications in fields such as biochemistry, spectroscopy, medicine, photonics and telecommunications due to the unique properties of the interaction between ultrashort laser pulses and matter [2][3][4]14].
After the demonstration of laser action in the ruby laser by Maiman in 1960, for which pumped flash lamp produced a burst of spikes, each several nanoseconds long, that lasted several hundred microseconds, scientists focused their efforts in developing pulsed laser systems with shorter pulse duration. Q-switching was successfully applied in the early 1960s to produce single pulses 10 nanoseconds long [15]. Next, mode-locking technique was first introduced in 1964 for which pulse duration was reduced to 100 picoseconds [16], and by 1981, pulse duration had been reduced down to 100 femtoseconds using the colliding-pulse mode-locking technique [17]. Sub-6 femtosecond pulses were achieved improving passive-mode-locking and dispersion compensation [18][19][20].
Mode-locking is a technique in which a fixed or locked phase relationship between the longitudinal modes of the laser cavity is induced. The longitudinal modes are made to interfere with each other giving rise to the emission of a train of short pulses whose duration may range from few femtoseconds to several picoseconds [21]. The temporal separation between these pulses is the time that a pulse takes to complete one round trip in the resonator cavity. As the duration of a mode-locked laser pulse is very short, according to the time-bandwidth product, its spectrum consists of a broad wavelength range. Mode-locking technique can be active or passive. In active mode-locking, an external signal induces a modulation of the intra-cavity laser radiation by using acousto-optics modulators. Passive mode-locking is based on the introduction in the laser cavity of some elements such as saturable absorbers, dyes or solidstate media which induce self-modulation of the laser light. The most common mode-locked femtosecond lasers are the Ti:Sapphire femtosecond lasers based on the Kerr-Lens Modelocking (KLM) mechanism, in which the Ti:Sapphire gain medium crystal placed inside the cavity acts as Kerr-effect medium [14,21]. Nowadays, typical commercial femtosecond laser systems produce sub-100 femtoseconds.

Ultrafast laser inscription
The exceptional characteristics of femtosecond laser pulses have allowed the development of novel applications in the framework of femtosecond laser processing. In particular, the fabrication of two-and three-dimensional permanent structures inside transparent optical materials to be applied in the fields of optics, photonics and telecommunication as key elements such as waveguides, photonic crystals, diffraction gratings, beam splitters, and so on.
In the ultrafast laser inscription (ULI) technique, femtosecond laser pulses are tightly focused beneath the surface. During the absorption process of ultrashort laser pulses, a high local spatial and temporal electronic and vibrational excitation densities are produced which together with the laser pulse duration, short compared to the relevant relaxation processes give rise to laser-induced nonlinear processes in the focal volume, such as two-photon or multiphoton absorption, inducing avalanche ionization in a very short time, leading to localized micro or sub-micrometric lattice damage, modifying the local refractive index and forming the so called "track" or "filament" [3].
The permanent refractive index change (Δn) induced in the focal volume has been found to be either positive or negative, depending on the processing parameters and the characteristics of the material. In addition to refractive index modification, a suitable configuration of the femtosecond laser pulses can be used to micromachine small areas by laser ablation or even to create nanovoids inside dielectric materials. The tight focalization in the focal spot can create extreme pressure and temperature conditions, with intensities significantly above the optical breakdown threshold and energy densities of several MJcm −3 into a submicron volume, several times higher than the strength of any material. This causes the solid to superheat yielding to the formation of a confined plasma, which explodes and generates a powerful shock wave that expands out of the focal volume and compresses the surrounding material, with pressures over 10 megabar (1 TPa) [45][46][47]. Therefore, an appropriate combination of the laser processing conditions in terms of numerical aperture (NA), diameter at the focal plane (d), laser wavelength (λ), pulse energy (E p ), pulse-width (Δt), repetition rate (f), and scanning speed (V) can be used to create a wide variety of laser-induced structures for waveguiding. Specifically, the energy delivered by the laser over each focal spot d is given by: The repetition rate is a critical parameter since the accumulation of multiple laser pulses over the same point will result in a local temperature increase if the period between consecutive pulses (1/f) is shorter than the cooling time: where D is the thermal diffusion coefficient. Thus, there is a critical repetition rate which determines two processing regimes; non-thermal and thermal regime for repetition rate lower and higher than the critical repetition rate [3]. For dielectric materials, thermal diffusivity is in the range of 10 −3 cm 2 s −1 . For a typical laser diameter at the focal plane of around 1 μm, the critical frequency ranges 100 kHz.
It is generally accepted the classification of waveguides structures into four categories according to the structure generated with the laser-induced filament [34]. In Type I waveguides, the laser beam generates a positive refractive index change, Δn > 0, directly in the irradiated area and the hence, the dimensions of the waveguide can be controlled by the amount of energy delivered on the focal volume. This type of waveguide can be induced in glasses since amorphous structure facilitates a positive refractive index change [3,4,23,36,37]. On the contrary, the mechanisms involved to be produced in crystals are complicated and entangled and hence, the type of refractive index change and the guiding axis cannot be envisaged. In addition, the damage created by the laser beam in the irradiated area gives rise to a non-desirable strong structural change of the material crystalline network. Furthermore, the waveguiding region is not stable with temperature, and the guidance is only supported in one polarization direction. Therefore, this type of waveguide has been reported in few crystalline materials such as LiNbO 3 , ZnSe and YCa 4 O(BO 3 ) 3 [35,[48][49][50][51].
In Type II or double-line configuration waveguides, the damage that can be produced by the laser beam in the focal volume is used to create a stress-induced region and a positive refractive index change between two close parallel tracks, normally between 15 and 30 μm. This type of configuration overcomes the drawbacks of Type I waveguide in crystalline materials so that waveguide structure is stable with temperature, the guidance can be achieved for both laser polarizations and in addition, preserves the luminescence of the bulk material.
The damage and the negative refractive index change originated by the ultra-intense and ultrashort laser radiation can be employed to surround a volume region of the bulk material. This is the origin of Type III or cladding waveguides for which the guiding core is adjusted and tailored by an appropriate position of the low-index tracks, placed few micrometers to each other to confine the light inside. The core diameter normally ranges between 25 and 150 μm so that it is possible the guidance of both monomode and multimode laser radiation for both TM and TE polarizations, from the visible to the IR region. These characteristics allow a better coupling to the waveguides, low propagation losses and a more efficient laser action in materials doped with Rare Earths (RE) [40][41][42][43][44][59][60][61][62].
Finally, in Type IV, waveguide also referred as ridge waveguides, the ultra-high intensity achieved by the femtosecond laser pulses is used to ablate the surface to produce ridge waveguides on planar waveguide substrates obtained by other methods. Therefore, the guiding features strongly depend on the planar waveguide substrate [63][64][65][66][67].
In addition to the ULI technique, the fabrication of optical waveguides has also been reported by other processing routes. Among these techniques, it is worth noting ion exchange (IE), annealed proton exchange (APE), molecular beam epitaxy (MBE), liquid phase epitaxy (LPE), and sol-gel [68]. Table 1 shows some examples of optical waveguides manufactured by these methods.
Nevertheless, when compared to these fabrication techniques, ULI has demonstrated to be a very powerful and robust technique due to its versatility, capable of manufacturing waveguides in a great variety of materials, from glasses to crystals, and flexible, allowing to inscribe a broad range of 3D structures and a large variety of waveguide configuration. Furthermore, the fabrication of integrated photonic devices by ULI has shown to be cost-effective and efficient so that a suitable configuration of the processing parameters may lead to waveguide losses below 1 dB/cm, resulting in a low laser operation threshold.

Ultrashort laser interaction with dielectrics
The high power density, in the order of tens of TW/cm 2 , achieved delivering laser pulses of few microjoules in short period of time, of few femtoseconds, focused beneath the surface in micron-sized areas, results in nonlinear interaction processes. Such high intensities may be manipulated to be employed for modifying the dielectric focal volume, inducing from weak refractive index changes to ultrahigh pressures which lead to void generation. These permanent structural modifications depend not only on the laser peak power but also on the focusing conditions, scanning speed, polarization and repetition rate.
The ultrashort laser-matter interaction process can be divided into three stages: generation of free carriers inside the material by non-linear processes such as multiphoton, tunnel ionization, or avalanche ionization, followed by energy relaxation processes and a subsequent modification of the material [3,4]. The absorption process of ultrashort laser radiation for wide bandgap dielectric materials cannot be explained on the basis of linear absorption, since the photon energy of commonly used femtosecond laser pulses with wavelength between UV and NIR does not have sufficient energy to be linearly absorbed. On the contrary, multiphoton absorption can excite an electron from the valence to the conduction band as long as mħω > E g , where m is the smallest number of photons for which the overall energy surpasses the bandgap energy E g . In addition, tunneling photoionization can also take place under an extremely strong laser electromagnetic field. This mechanism in dielectrics permits electron from the valence band to tunnel to the conduction band in a period of time shorter than the laser pulse. However, in most dielectric multiphoton, ionization dominates the excitation processes [78,79]. On the other hand, laser photons can be sequentially absorbed by electrons excited in the conduction band by means of free carrier absorption. When the energy of an electron in the excited state exceeds the bandgap energy, the ionization of another electron from the valence band can take place, resulting in two excited electrons at the conduction band minimum. These electrons can be excited again by free carrier absorption processes, and more valence electrons can be produced by the same mechanism, leading to the electronic avalanche [3,4]. The requirement for avalanche ionization is the existence of seed electrons in the conduction band, which can be provided by multiphoton or tunneling ionization or by thermally excited impurity or defect states.
Once the nonlinear photoionization and avalanche ionization create a free electron plasma, they transfer their energy to the lattice inducing three types of structural changes: a smooth refractive index modification for low pulse energies such as 100 nJ and 100 fs at 800 nm for 0.6 NA, a birefringent refractive index modification for pulse energies ranging 150-500 nJ and 100 fs at 800 nm for 0.6 NA, and microexplosions which result in void formation for pulse energies higher than 500 nJ and intensities greater than 100 TW/cm 2 [4,78,80], Figure 1. The type of structural change depends on the laser features (irradiance, repetition rate, polarization), working parameters (scanning speed, numerical aperture) and material properties (thermal conductivity, bandgap energy).

Bioactive glasses
Formation of glasses with improved properties and manufacturability plays an important role in many technologies. Various noncrystalline materials such as silica glasses, amorphous halides, semiconductors and metals have been used in areas for many engineering applications including biology, communications, electronics, and so on. Glass formation consists basically in avoiding crystallization by cooling from the molten state. The ability to glass formation depends on the melt composition that defines some parameters as glass transition and liquidus temperatures, on the cooling rate which has to be higher enough to avoid the nucleation and growth of the crystalline phases and also on the sample size. As the solidification technique used limits at some extent the cooling rate, the first question in the development of glasses is to determine the composition ranges in which glasses can be obtained under certain cooling conditions (quenching, directional solidification, etc.).
It is known that for multicomponent systems, a strong tendency to glass formation exists near the eutectic points because they have the lower liquidus temperature. Glass formation and its relation to eutectic growth have been investigated for different metal and ceramic alloy systems in terms of the competition between the growth of crystalline phases and the formation of the amorphous phase. The maximum growth rate of a eutectic structure with two or more phases is lower than the maximum growth rate of a single crystalline phase. Therefore, the locations of glass formation in some oxide, halide or metal systems lie near the eutectic area compositions, even though the glass formation also depends on the cooling rate. High (Left) birefringence cartography in optical fiber Ge-doped preform core [80], and (right) nanovoids in sapphire [46]. liquidus slopes, low mutual solubility of the phases (high eutectic range C αβ ), small diffusion in the melt and large solid/liquid interfacial energies favor the glass formation in competition with the eutectic formed through cooperative growth of the crystalline phases.
Bioactive glasses capable of forming tight chemical bonds to living tissues contain silicates and some of them also phosphates, the first providing a scarcely soluble matrix that compensates for the excess solubility of the latter. The ability of bonding to bone tissue is a result of their chemical reactivity of the surface in a physiological media. The formation of SiO 2 rich layer and calcium phosphate film on the surface of a bioglass implanted in a body and bond with living bone was reported by Hench [81]. Kokubo proposed the use of artificial body fluid as a simulated body environment to estimate the bioactivity of the glasses and ceramics [82].
De Aza et al. showed that glasses and ceramics in the systems containing CaO▬P 2 O 5 ▬SiO 2 can be designed to optimize biological and mechanical response. They have studied different microstructures in the wollastonite (CaO.SiO 2 )-calcium phosphate TCP (3CaO.P 2 O 5 ) system reporting the presence in the phase diagram of an invariant point at 1402°C and a composition of 80% mol W and 20% mol TCP [83]. It is possible to estimate the eutectic compositions from the values of melting temperature and heat of fusion of the components through the expressions [84]: where X l i is the mole fraction of component i in the liquid and T m i and H m i are melting temperature and molar heat of fusion of component i in the liquid. The eutectic composition and the eutectic temperature can be calculated with the expressions indicated above if the melting temperature and heats of fusion of the components are known, as shown in Table 2.
On the other hand, in order to form an amorphous solid, the melt must be cooled sufficiently fast to prevent the precipitation of the competitive crystalline phases, defining a minimum cooling rate. Lu and Liu [85] used the dimensionless parameter γ = Tx/(Tg + Tl) as an indicator of the glass-forming ability of the system, with Tx, Tg and Tl the crystallization onset, glass transition and melting temperatures, respectively. The relationship between this parameter and the critical cooling rate (Rc) for glass formation is approximately given by the expression: where the constants R o and γ o are, for inorganic glasses, 8 × 10 27 K/s and 0.421, respectively. Magallanes-Perdomo et al. described the devitrification and crystallization process of W-TCP eutectic glass, evidencing that the devitrification begins at 870°C with the crystallization of a Ca-deficient apatite phase [86]. Being the glass transition temperature 790°C, the γ parameter is 0.414 and the calculated critical cooling rate R c is about 3 K/s. This rate corresponds to a solidification rate of about 100 mm/h in the Laser Floating Zone (LFZ) technique for an experimental axial thermal gradient of 10 5 K/m, in good accordance with the growth rate of 150 mm/h reported by Pardo et al. for the fabrication of a W-TCP eutectic glass rod of 3 mm in diameter [87]. A detailed description of this technique can be consulted elsewhere [88,89].
This eutectic glass, produced by the laser floating zone technique, is chemically stable and has a high optical quality being able to be used as a matrix for luminescence active ions. This composition corresponds to an "invert" glass where the modifier content (Ca) is larger than the former content (Si + P). This term was introduced by Trapp and Stevels [90] because the traditional network forming oxides SiO 2 and P 2 O 5 form continuous molecular/ionic networks in normal conditions; however, when the network modifying oxides are in majority on the molar basis, the glasses are structurally inverted compared to conventional glasses. This structural inversion is reflected in the properties of the glasses. In addition, the optical properties of this glass doped with Nd 3+ and Er 3+ ions have been assessed, resulting in emission crosssections and lifetimes similar to those reported for the best commercial alkaline-silicate glasses [87]. Furthermore, the incorporation of trivalent rare-earth (RE) ions can also be used as local ordering probe due to the close relation between their spectroscopic properties and the local structure and bonding at the ion site, since the spectroscopic properties of trivalent rare-earth ions depend on the chemical composition of the glass matrix, which determines the structure and nature of the bonds, and thus spectroscopic characterization allows studying the local structure surrounding the RE ion and the covalence of the RE-O bond [91,92]. In particular, site-selective excitation and emission of Nd-and Eu-doped W-TCP glass ceramics provided information about the differences on the spectral features of amorphous and crystalline environments in this matrix, allowing the identification by means of Raman and LIBS of crystalline phases corresponding to dicalcium silicate (Ca 2 SiO 4 ) and apatite-like structures, both with well-known bioactive and biocompatible properties [93][94][95]. These results showed the potential applications of these materials obtained from the LFZ technique as luminescence bioprobes for in vitro applications and promote extended studies to other rare-earth ions, which can be used in biomedical applications such as multicolor bioprobes and biosensors among others.

Fabrication of photonic devices in W-TCP bioactive glasses
Both stress-induced and depressed cladding waveguides have been studied in Nd-doped CaSiO 3 ▬Ca 3 (PO 4 ) 2 bioactive glasses. In both cases, the guiding structures were manufactured by using the ULI technique. For this purpose, a tunable Ti:Sapphire laser emitting at the central wavelength of 795 nm with 120 fs laser pulses and 1 kHz repetition rate was used.  the guiding characteristics were studied modifying the pulse energy between 0.42 and 1.68 μJ, the scanning speed at 25 and 50 μm/s and the distance between filaments at 15 and 20 μm. Also, in this case, the structures were inscribed at 150 and 250 μm beneath the surface of the sample. As shown in Figure 2(a), the filament dimensions strongly depend on the energy delivered (E d ) by the laser beam, between 8.4 and 67 μJ, achieving lengths from 16 up to 40 μm. For the case of the cladding structure, both the distance between filaments and the scanning speed were set constant at 3 μm and 500 μm/s respectively, studying the guiding properties with the diameter of the core, between 20 and 150 μm, and the depth from the surface, 300 and 600 μm. Worth mentioning is the fact that the fabrication of cladding structures required lower delivered energy than in the case of double-filament waveguides, since the filaments were very close to each other and an excessive amount of energy might create cracks on the irradiated area and hence to fracture the sample. Specifically, in this case, filaments of the cladding structure were produced delivering 1.44 μJ over each focal spot. In addition, the interaction of the ultrashort laser radiation with the glass sample gave rise to an affected zone in the surroundings of the tracks, caused by the generation of color centers. It is well known that high intensity ultrashort laser pulses may modify the properties of the material at atomic scale by the photophysical and photochemical interactions which take place after the laser absorption [3]. These color centers also may affect to both the guiding and luminescence properties.
The guiding properties were assessed by characterization of the near-field intensity distribution with a continuous wave He-Ne laser at 633 nm. Both types of waveguides presented TE and TM propagation. Furthermore, all double-filament waveguides inscribed were found to be monomode, whereas only 20 μm diameter depressed cladding waveguides presented this feature. In addition, for double-filament waveguides, the width found in TE and TM modes was shorter along the x-axis than in the orthogonal one, but the modal area decreased with the pulse energy thus increasing the light confinement. Moreover, the best confinement conditions were achieved for both the shortest track distance (15 μm) and depth (150 μm), and the highest scanning speed (50 μm/s). Concerning the cladding waveguides, there were no remarkable differences depending either on the guided polarization or processing parameters. As an example, Figure 3 shows the near-field distribution intensity of TE mode in a double-filament waveguide inscribed 150 μm beneath the surface with 15 μm between tracks (a), and cladding waveguides (b) of 20 μm (left) and 30 μm (right) diameters.
The absorption generated by color centers, appeared during the laser inscription process was reduced by means of isochronal heat treatments thus decreasing propagation losses [96][97][98]. This annealing process produced an improvement of up to a 29% in the transmitted power in the range of 300-350°C for the double-line waveguide, as shown in Figure 4(a), whereas higher annealing temperatures led to a diminution of the stress-induced between filaments and therefore to a decrease in the refractive index change and consequently in the light confinement. In depressed cladding waveguides, the thermal treatment induced a continuous improvement of the transmitted intensity up to 650°C, based on a gradual removal of color centers and a partial recovery of the refractive index change caused by the release of the stress-induced by the ultrashort laser pulses. Annealing temperatures higher than 650°C reduced the optical barrier and the tunneling losses became non-negligible [99]. The analysis of the evolution of the modal intensity as a function of the thermal treatment temperature confirmed this mechanism, as shown in Figure 4(b), revealing that the light confinement kept almost unaltered until 600°C, whereas higher temperatures enlarged the modes hindering the light transmission.
Luminescence characterization carried out in both types of waveguides revealed that spectroscopic properties were well preserved in the volume of the waveguides and that annealing treatment at high temperature recombined the damaged produced in the irradiated areas, resulting in a worsening the light confinement. Fluorescence lifetimes, around 240-250 μs, and both luminescence spectra and μL map of the 4 F 3/2 → 4 I 11/2 transition in the bulk and in the waveguide's volume pointed out no significant modifications, Figure 5.
The refractive index modification was estimated on the basis of the refractive index distribution generated by femtosecond laser pulses in double-filament waveguides [100,101], for which: where Δn max is the maximum reduction on the refractive index in the track, and σ x and σ y are the width of the damaged region along the horizontal and vertical directions respectively, resulting in a Δn~5 × 10 −3 , which was found to be in the same range of previously reported for other waveguides inscribed by femtosecond laser writing, such as Nd:YAG ceramics, Nd:YVO 4 and fused-silica.

Conclusions
Fabrication of integrated photonic devices has been a hot topic in research since the first report on femtosecond laser written waveguides in glass by using the Ultrafast Laser Inscription Technique (ULI), in 1996. In this chapter, we have reviewed the fundamentals of this laser  processing technique, describing the main configurations for waveguide fabrication achieved to date, and the materials for which they have been applied. Also, a comparison to other techniques commonly used for waveguide fabrication has been included, pointing out the principal advantages of ULI when compared to these methods. Furthermore, the basis of the interaction between ultrashort laser pulses and dielectric media which may give rise to refractive index modification and hence, to waveguide fabrication has been outlined. In addition, this chapter briefly reviews the development of bioactive glasses related to their capability of forming tight chemical bonds to living tissues. This section is particularized to the development of CaSiO 3 ▬Ca 3 (PO 4 ) 2 bioactive eutectic glass and glass ceramics, and to their potential applications as luminescence bioprobes in biomedical applications when doped with rareearth ions. Finally, this chapter reviews the fabrication of double-filament and depressed cladding waveguides in Nd-doped CaSiO 3 ▬Ca 3 (PO 4 ) 2 bioactive eutectic glasses. Doublefilament waveguides presented an increase of the refractive index in the region between damaged areas, the so-called "tracks," inscribed with the ultrashort laser pulses. In cladding waveguides, tracks were inscribed close to each other to surround a volume of bulk material. Both types of waveguides supported TE and TM polarization guiding under illumination of cw He-Ne laser radiation. In particular, only the cladding waveguide with the smallest core, 20 μm, presented monomode transmitted light. Color centers formation was observed in both types of waveguides as a consequence of the laser interaction with the glass. Isochronal thermal treatments allowed decreasing the absorption by color centers and for the highest annealing temperatures, a release of the stress-induced in the laser treated area. In addition, it was shown that spectroscopic properties were well preserved in the volume of the waveguides. Finally, the refractive index modification, Δn, was estimated in 5 × 10 −3 , in the same range of previously reported works by the ULI technique.
Future work is expected to be done in the fabrication of integrated photonic devices in glasses and glass ceramics of the CaO▬P 2 O 5 ▬SiO 2 ▬MgO system by the ULI technique. Glasses and glass ceramics of this system are also bioactive. Therefore, we will ascertain and assess the optimal conditions for the fabrication of optical waveguides and their possible utilization as laser sources for biomedical applications.