RICERCA ITALIANA     

Integrated amplifiers e laser sources within soliton waveguides in Er:LiNbO3

Università degli Studi di Roma "La Sapienza"

Abstract

The present project will realise laser sources at 1.55 um, integrated within single-mode soliton waveguide, buried inside single crystals on lithium niobate doped in volume by erbium. Both single and arrays of lasers will be realised.
Even if waveguide integrated laser have been already realised in the past (the first Er:LiNbO3 laser was built in 1992), their construction, because of the fabrication techniques, was always performed close to the surface of a bulk material, that for this reason got the name "substrate", i.e. lower layer. Indeed until now only superficial structures have been realised, because the whole integrated optics is indeed superficial: any test to get buried waveguides inside any substrate gave unsatisfactory results, being not able to overpass a depth for the external surface larger than 100-200 um. Only recently such a limit has been overcome by using new writing techniques coming from nonlinear optics. In fact, within nonlinear materials like photorefractive lithium niobate, spatial soliton formation, i.e. formation of self-confined undiffracting beams, was both theoretically and experimentally demonstrated. Such beams permanently modify the material refractive index writing, as a consequence, a perfect single-mode waveguide. These waveguides will be used to realise integrated lasers within erbium doped lithium niobate.
Soliton waveguides start a new technology for using the whole volume of a material, because they can be written everywhere inside a bulk, without restrictions. Thus, such a technology implies many advantages from the points of view of both material and possible geometries: a) regarding the material, rare-earth volume-doping instead of surface-doping is more homogeneous and efficient; b) regarding the waveguides, the refractive index profiles are self-written by light and consequently they are optimised for single-mode propagation without losses; c) regarding laser, either single cavities or array of independent lasers can be constructed on the same chip.
The Consortium will have the following tasks: to grow up the crystals, to characterise them, and to use them for solitons and lasers.
By using the Czochralski technique, particularly important for the quality of the obtainable products, high-purity and high-optical quality single-crystals of lithium niobate will be grown up, with volume doping of erbium. Particular attention will be given to the structural and optical characterisation of the obtained samples, either because the specific literature is not that deepen, either because the consortium can optimise the growing up procedures according to the properties found during the characterisation and use stages. Then soliton waveguides will be written and completely characterised measuring the index profile, losses and luminescence efficiency. Moreover single-crystal fibres of erbium doped lithium niobate will been grown up, trying also to write solitons inside them, that may act as the fibre core for the light.
As soon as the crystals, and eventually the fibres, are characterised, laser sources will be considered with suitable cavity geometries. Arrays of lasers within the same chip will be investigated too.
Such a project is strategically relevant for the applications of integrated lasers and especially of laser arrays in many different fields, ranging from telecommunication, to displays, biological and bioengineering applications as for example photometric microanalysis of biological samples as arrays. <<<

Principal Investigator

Eugenio FAZIO Università degli Studi di ROMA "La Sapienza"

Research Objectives

The proposed project would realise laser sources at 1.55 um, integrated in single-mode waveguides with soliton refractive index profile, buried inside single-crystals of lithium niobate doped with erbium in volume by means of the Czochralski technique. Both single sources and arrays of laser cavities will be realised in the same crystal.
The project would also write soliton waveguides in single-crystal fibres of lithium niobate doped with erbium grown by means of the MicroPullingDown technique. Such soliton waveguides would act as cores for the fibres, within which the light can be propagated and optically amplified.
The primary task of the entire project will be caught up by following intermediate advance steps of the job, that they can therefore be defined as:
1) growing up of single crystals of lithium niobate doped in volume with erbium by means of the Czochralski technique. During the initial step the growing of single-crystal samples of lithium niobate with concentrations of erbium, distributed in homogenously in the inner volume, ranging between 0 and 1mol%. Such samples must have optical and crystalline quality comparable with undoped crystals of the same material available in the market. Secondary task, in case of successful growing of single-crystals with the previous characteristics, is the determination of the growing protocol in order to obtain reproducible samples. In a second step of such a task, using the protocol set up before, crystals with larger erbium concentrations, ranging between 1 and 2mol%, will be tried to be grown.
2) single ferroelectric domain poling of the grown crystals will be tried together with the crystal cut according to the needed dimension.
3) Complete characterization of optical the structural properties of the material, with particular attention to the problem of the charge and energy transfers from erbium ions to the host lattice and vice versa will be analysed. From these studies one expects of being able to develop an analytical code for simulating of the interaction process radiation-material to be used for the numerical simulations of the soliton formation and of the laser action in the cavity to be realised.
4) Realization of spatial solitons in erbium doped lithium niobate and characterization of the soliton channels like single-mode waveguides. Great attention will be given to the characterization of the mode of propagation of light within the guides varying the wavelength. This measure will allow to a priori quantify the modal overlapping between the pump-beam mode and the emitted-beam mode from the source that will be experienced latter on during the laser realisation.

5) Growing of erbium doped lithium niobate fibres by means of the MicroPullingDown technique, with crystallographic direction and chemical composition controlled along the whole fibre length.
6) Poling to single ferroelectric domain of single-crystal fibre of lithium niobate.
7) Realization of spatial solitons in erbium doped lithium niobate fibres, and their use as fibre cores for the light propagation.
8) Growing and characterization of lithium niobate crystals with double doping, with erbium and with other elements (Yb, Ce, Ag, Fe.....) as sensitizers of the light emission process and as intensifiers of the photorefractive effect of the material. At the end of this step one or more species of co-dopants will be chosen, the ones showing the best performances and contemporarily the best doping concentration for the amplification and the laser effect.
9) Design and realization of the laser resonator with dimensioning of the optical components as well as of the eventual cooling system to use.
10) Spectroscopic and power characterizations of the amplified spontaneous emission (ASE) occurring inside the soliton waveguides up to the ignition of the realized laser cavity.
11) Characterization of the generated laser beam and determination of the temporal stability of the realized cavity. All the procedures able to optimise the emission efficiency, as well as the quality of the mode and the stability of the delivered intensity will be carried out. For such optimization, the temporal stability of the soliton waveguides will be analysed as well as further stabilization techniques (like for example the application of bias fields during the lasing action) or the employment of new materials with special doping (as an example using one larger percentage of iron). Various stechiometric compositions of erbium and of other sensitizers in the lithium niobate will be studied, in order to improve the characteristics of the lasing.
12) Realisation on the same chip of arrays of lasers contemporarily and independently working. <<<

Timescale

24 months

National and international background

The exploding demand for internet access, telecommunications and broadband services is pushing for lightwave transmission, that had intrinsically shown much larger capacity than electronic technologies, which shortly will not likely be able to comply with the increasing request of high-speed systems. In order to meet this demand, more and more sophisticated and compact optical components are required, such as bulk- or directly fibre-type devices. The key point of this technology is based on "optical waveguides", which mean zones of larger refractive index than the surrounding, in order to confine the light inside by means of total internal reflection phenomenon. The present technology has successfully developed both 1-dimensional and 2-dimensional structures, while the 3-dimensional ones are still unexplored, being any suitable ad innovative technology for volume-waveguide writing still unavailable. Different techniques are usually adopted for constructing 1-dimensional and 2-dimensional structures: for example optical fibres are drawn from a larger preform until the desired dimension is reached; optical waveguides instead can either be grown as thin films on substrate or be realised by chemical-physical treatment of the external surface of the material to locally modify the refractive index. Such a procedure can use suitable masks for getting planar or channel waveguides. However the possibility of increase the number of complex functions on the same chip will greatly advantage a complete and more efficient integration: it might allow in fact to reduce the fabrication costs, because of the simultaneous growing of the whole integrated structure, might reduce the number of manual interconnections increasing consequently the affordability, compactness and efficiency of the final chip. Thus, a growing pulse toward the developing of new materials for light amplification, guiding and switching. Along this line, the simultaneous combination of many properties, such as electro-optic, acousto-optic, piezoelectric, photo-elastic, photovoltaic, second order optical nonlinearity effects, make lithium niobate (LiNbO3) one of the most suitable materials for integrated optics [1].
Actually, thanks to these properties, lithium niobate crystals are already successfully used in integrated devices for electronic applications (integrated circuits for signal processing in videotape-recorders and TV's, acousto-otpic modulators, Pockels cells, Faraday rotators, optical dispersion compensators, optical wavefront converters, optical parametric amplifiers or optical frequency generators and so on…)
Recently a new technique has been demonstrated [2] for writing optical waveguides inside the material volume and not anymore close to the external surface. Such a technique take advantages of nonlinear optic phenomena, such as the generation of "spatial solitons". By using this new technology fully 3-dimensional optical circuitry within a material volume can be realised.
The expression "spatial soliton" means here a light beam that does not diffract during propagation. This phenomenon is possible because the light beam, thanks to the optical nonlinearity of the material in which it propagates, locally increases its refractive index. In such a way it writes a real waveguide inside which the light is self-confined, compensating the natural diffraction. The spatial solitons were theoretically demonstrated in1964 [3] and experimentally observed in 1972 [4]. Since then, a large characterization activity occurred, in order to study the soliton properties as function of the guest material. In the present project we shall investigate lithium niobate crystals, a photorefractive material. Photorefractive solitons were theorised in 1992 [5] and experimentally observed one year later in strontium barium niobate crystals doped with rhodium [6]. However only in 2004 spatial solitons in lithium niobate were observed [2]. This delay is mainly due to the competition, inside lithium niobate, of the photorefractive effect, that locally increases the refractive index allowing the soliton formation, and the photovoltaic effect, that locally decreases the refractive index avoiding the beam self-confines. Anyhow, adopting a new technique based on the application of intense static electric biases, bright spatial solitons were indeed observed even in this material. It was even demonstrated [7] that in ferroelectric materials like lithium niobate, the refractive index modification induced by the soliton beam can be permanently fixed behaving as a perfect single-mode waveguide. Solitons can indeed write waveguides inside the volume of the guest material. Thus the technique has many advantages with respect to the traditional techniques:
1. spatial solitons write perfect single-mode waveguides, either channel or planar ones;
2. the refractive index profile, generally squared hyperbolic secant in shape, is self-optimised by light for single-mode propagation and consequently has low dispersion and no-losses. In traditional waveguides the index profile can vary either by step or following the trend of the writing chemical process. In both cases it is not optimised for propagation, or better the light that must adapt itself for propagating inside the waveguide. This modal adaptation causes losses and dispersion, contrarily to soliton waveguides whose profile is chosen and written by the light itself.
3. Soliton technology is a low cost one: it does require large facilities like for the traditional writing procedures by only low power lasers (mW-W);
4. complex 3-dimensional integrated circuits can be realised by using such volume waveguides and especially by using the innumerable cross-interaction properties of solitons [8]. The first 3-dimensional simple realisations were obtained by writing in the same chip arrays of 9x9 [9] and 32x32 [10] arrays of soliton waveguides, over transversal dimensions not larger than 2.3x2.3 mm^2.
The present research project follows exactly this road, developing single and multiple (as arrays) laser sources, within lithium niobate crystals doped with erbium, by using soliton waveguide technology. Within the last decade, erbium ions have played a key role for optical technologies of the light amplification at 1.55 um, 3rd-telecom window. In fact, within interurban networks, regeneration of optical signals must be performed to compensate for glass fibre losses, that can be performed all-optically, instead of electro-optically, using Er3+ fibre and waveguide amplifiers. This was a revolutionary technology for high-speed optical networks, allowing high-capability optical interconnections on a world scale.
The band structure of Er3+ allows to pump the energy level 4I11/2 by a semiconductor laser at 980 nm, that consequently fills up the 3I13/2 state by carrier relaxation from which radiative emission of energy is possible towards the level 4I15/2 at 1.55 um light wavelength, 3rd window of the optical fibres. Thus, using the erbium emission optical integrated amplifiers have been designed and realised, as a derivation of fibre amplifiers. With respect to them, integrated amplifiers can include many optical elements in the same chip, form lasers to multiplexers and demultiplexers, modulators, interconnections and so on… A significant reduction of the device dimensions and fabrication costs is then derived.
By using erbium3+ properties, a full family of integrated lasers can be developed, with emission either at 1.55 um or at the other erbium lines, in the visible or IR spectral regions. Actually it is well known that erbium ions can emit at different wavelengths for laser action (being 550nm the most efficient), when used as doping for materials like YAF, LiYF4, glass fibres and so on… Such potential miniaturisation has inspired researches on new guest materials for Er3+, among which lithium niobate, for its nonlinear optical properties, is a preferable candidate to realise CW as well as pulsed lasers, DFB or acousto-optically tuneable. Typically lithium niobate integrated amplifiers and lasers need a double doping: with erbium, as active material, and with titanium, for writing the waveguide inside which laser emission and amplification occur. Such Ti-Er co-doping is produced by thermal in-diffusion from thin film [11-14]. However it was observed that Ti and Er diffusion processes follow different trends, i.e. generate different profiles inside lithium niobate, one for the index profile (given by titanium) and one for optical gain (given by erbium). Such difference gives many disadvantages: i) a small overlapping between the waveguide mode and the amplifying region occurs, and consequently the obtainable optical gain decreases while the noise increases; ii) only a surface doping instead of a volume doping is possible, and consequently only superficial circuitry can be realised; iii) light polarisation rotation occurs inside such waveguides because of the diffusion-induced birefringence; iv) the co-doping modifies the structural properties of the material, avoiding strong doping. In fact, because the lifetime of the Er3+ excited state is usually in the millisecond range (depending on concentration) the absorption and emission cross-sections are relatively small (typically 10^(-21)/10^(-20) cm^2). Subsequently reasonable values of optical gain (3 dB) can be reached with high erbium concentrations, at least as high as 10^20/10^21 atom/cm2, (densities of 0.1-1 at%) possible if the host material has a high erbium solubility. In this case Er3+-Er3+ distance can be too close that electric dipole-dipole interactions are excited, reducing the gain performance of the optical amplifier. This is a clear limit of erbium lasers and amplifiers realised until now. Other solutions were proposed to replace Ti in-diffusion, such as the ion implantation of light elements such as C, N and O but again only surface planar configuration were available even if the optical losses were strongly reduced [15-16].
The present project will introduce 6 novelties with respect to the previous realisations: 1) volume waveguides will be realised instead of surface ones; 2) thus, by using Czochralski (Cz) technique [17] single-crystals of lithium niobate will be grown up directly doped in volume by erbium, opposite of the previous techniques where the doping was realised by a post treatment; 3) the waveguides will be written by the laser pump beam, in order to get index profile and pumping by the same process; 4) small index variation, that will allow a larger dimension of the waveguide mode in order to cover a larger active volume; 5) both single cavities and arrays of independent lasers will be realised; 6) both bulk-shaped (by and fibre-shaped (by using the MicroPullingDown technique (MPD) [18]) single crystals will be analysed, where the soliton waveguide can represent the propagation core.
Both Cz and MPD techniques grow up high optical quality crystals, doped in volume, with chosen crystallographic orientation, that will be cut and polished of the required dimensions.
By the Cz method the crystals develop from liquid phase, through solidification at the solid/liquid interface [17]. Such Cz technique shows many advantages: 1) versatility and adaptability: the crystal orientation, composition and dimensions can be precisely selected a priori; 2) no constrains are used to grow the crystal boules, main requirement for obtaining low dislocation density or streak-free crystals. 3) in principle any element can be introduced in the crystals during the growth, and co-doping with multiple species can be easily achieved. Thus, the technique is so extremely precise that it grows the commercially available lithium niobate crystals.
The MPD technique is instead relatively new even if it has been adopted to grow optoelectronic materials, both semiconductors or insulators, for many applications like IR waveguides, scintillators, superconductors, frequency converters, microlaser, photorefractive crystals [19-20]. Such a technique uses a small crucible where the base material is fused and then pulled downwards by means of a seed and a rod. Consequently single-crystal and single-domain fibres are obtained, with low dislocation densities, diameters ranging from 10 um to few mm's; it is fast, relatively simple and cheap. In fact fibres are grown up from a small initial volume (tens of cm3) with high speed (0.5-20 mm/min) that allow a uniform radial distribution of components and high doping rates (larger than conventional methods).
By using the high crystalline and optical qualities of single-crystal volume-doped samples and by using the performances of soliton the waveguides, volume integrated amplifiers and lasers will be realised in erbium doped lithium niobate.
The present project demonstrate many innovations in the framework of the Italian research for the following tasks: a) growing of lithium niobate single-crystals doped in volume by erbium; b) growing of erbium doped single-crystal fibres; c) realisation of spatial solitons in amplifying photorefractive media; d) realisation of solitons in photorefractive fibres; e) realisation of amplifiers and lasers within single-mode soliton waveguides; f) realisation of volume integrated lasers within optically active lithium niobate and, last by not least, g) realisation of laser arrays on the same chip. A fundamental aspect of the project is the interaction of the research units, with high specialisation in their research fields and complementarily of knowledge's, that allow to optimise the materials according to their amplification and photorefractive uses. <<<