Open Access
Add to BibSonomyAbstract
We report, for the first time, Er-doped Ge-Ga-Se films and waveguides deposited using co-thermal evaporation and patterned with plasma etching. The emission properties of the bulk glasses were studied as a function of Erbium doping, showing for the first time that there is a clear concentration quenching effect in the Ge-Ga-Se glasses with a linear radiative lifetime degradation slope of −0.48 ms/mol% Er from a low concentration lifetime of 1.7 ms, even when sufficient Gallium is present to ensure homogeneous distribution of the Erbium. A region between approximately 0.5 and 0.75 mol% Erbium however is shown to provide sufficient doping, good photoluminescence and adequate lifetime to envisage practical planar waveguide amplifier devices. Film emission properties at 0.7 mol% doping were studied and compared with the bulk counterpart showing adequate lifetimes and photoluminescence. Erbium doped films with ~0.8 dB/cm propagation loss at 1550 nm limited by Mie scattering off small particles ejected from the evaporation crucible were fabricated. Planar hybrid Er-Ge-Ga-Se/As2S3 rib waveguides fabricated through photolithography and plasma etching demonstrated propagation losses of ~2 dB/cm at 1650 nm limited by particulate scattering.
© 2014 Optical Society of America
1. Introduction
Chalcogenide glass (ChG)-based nonlinear optical waveguides have proven to be very useful for high-speed all-optical signal processing of telecommunications signals [1–4]. To achieve the best performance, materials with larger optical nonlinearity and waveguides with smaller mode area are essential since the magnitude of the waveguide optical nonlinearity per meter per Watt of input power, γ, is defined as γ = 2πn2/λAeff, where n2 is the third order optical nonlinear coefficient of the material, λ the wavelength of operation and Aeff the mode area of the waveguide. Recently, γ of 130-150 W−1m−1 was achieved in Ge11.5As24Se64.5 nanowires with 500x630 nm cross section and mode effective area ~0.24 μm2 [5], which is the highest value reported so far for any glass waveguide [6]. However, nanowires experience significant propagation losses because of the strong coupling of the optical fields to the etched sidewalls even when they have been fabricated using optimized processes that produce the lowest roughness [5]. Typically the losses achieved for chalcogenide nanowires to date are more than 1.5 dB/cm, which is in sharp contrast with lower values of 0.2-0.3 dB/cm in 850 nm thick As2S3 rib waveguides with a larger mode area of 1 μm2 [7] and 0.05 dB/cm in 2.5 μm thick rib devices with Aeff~7 μm2 [2]. These losses strongly limit the available conversion efficiency, which decays approximately with the inverse square of the loss [8]. Thus a method of overcoming the waveguide losses is desirable to enhance the efficiency of all-optical processing and enable techniques currently inaccessible due to insufficient nonlinearity.
One approach to remedy this situation is to employ rare earth (RE) doped waveguides in which optical gain can compensate the propagation losses. Net waveguide gain has been demonstrated in Silica, Alumina, Tellurium Dioxide, Tantulum Pentoxide, etc based devices employing the 4I13/2→4I15/2 transition of the trivalent state of Erbium at 1.55 μm [9–14]. Compared with oxide glass hosts, the lower characteristic phonon energy in ChGs reduces the impact of multi-phonon processes such as multi-phonon relaxation (MPR) or phonon-assisted parasitic transitions and also a considerably larger emission cross section is available. The challenge is that RE ions tend to form clusters when doped into ChGs at the concentration levels need for planar amplifiers/loss compensation (1020-1021 ions/cm3) and therefore the emission efficiency is often significantly suppressed. Recently it has been reported that, the addition of Ga can improve the solubility of RE ions in chalcogenide glasses and the optimal ratio between Ga and RE was established to be ~10:1 [15, 16].
Erbium doped ChG bulk glasses have been investigated somewhat, but the properties of Erbium doped thin films and waveguides based on ChG are not at all well understood. For example, K. Koughia et al. [16] studied the photoluminescence, optical absorption and structural properties of Ge–Ga–Se glasses. The results suggested that a high ratio of Ga to Er ensured homogeneous distribution of Er, whilst a low ratio leads to the formation of clusters. In the intermediate region, energy was shown to be able to effectively migrate from one ion to another. Er2S3 concentrations from 1.8 to 2.4 mol% doped into GeGaS glasses were prepared by D.T. Tonchev, and the glasses exhibited a broad emission band at ~1540 nm with PL decay lifetimes in the range 1.13 to 1.55 ms when pumped at 818nm, decreasing rapidly around these concentrations with Er content at 0.7 ms/mol% [17]. S.O. Kasap et al. examined the optical and photoluminescence properties of Er3+-doped GeGaSe glasses and calculated a pure radiative lifetime around 2.6 ms for the 4I13/2 to 4I15/2 transition using the Judd–Ofelt theory [18]. Allen et al. [19] studied the photoluminescence characteristics of a series of Er-doped ChGs and found that, GaGeAsSe had the shortest lifetimes of 1–1.5 ms for 980 nm pumping, whilst GaGeSe and GaGeS samples had the highest values of 2–4 ms.
High quality ChG thin films doped with Erbium have also been reported for a very few chalcogenide hosts. For example, thin As2S3 and As-S-Se films were formed by thermal evaporation and Erbium doping was obtained by subsequent ion implantation in [20]. Emission cross sections up to 15x10−21/cm2 (2x the best Alumino-Phospho-Silicate glasses) were measured with peak Er3+ emission at 1.54 um, and the 4I13/2 metastable lifetime was determined to be 2.3 ms with excitation at 983 nm. Takahiko et al. reported the properties of Ga-Ge-Se films on fused silica substrates deposited by sputtering, and the lifetime was 1.8-2.6 ms for the 1550 nm band when excited by a 973 nm laser [21]. In V. Nazabal’s reports [22], Ga5Ge20Sb10S(Se)65 and Ga5Ge23Sb5S67 films were fabricated by pulsed laser deposition (PLD), both physical and optical properties of the films are investigated. The fluorescence decays of 4I13/2 level (excited at 980 nm) of 3000 and 14,000 ppm Er-doped Ga5Ge20Sb10S65 films were found to be:0.45 ± 0.03 and 0.30 ± 0.03 ms, respectively. 1-μm-thick of Er-doped As40Ge10Se25S25 film was deposited on silicon substrates by radio frequency sputtering in [23]. The PL and photoluminescence excitation (PLE) spectroscopy of 4I13/2-4I15/2 Er3+ emission were recorded, from the results, Er3+ emission is excited by the Urbach absorption edge of the host glass rather than direct absorption by the Er3+ intra-4f shell transitions, providing valuable design flexibility in the choices of pump wavelengths.
To date, however, there have been very few demonstrations of fiber or waveguide amplifiers based on RE doped ChGs in spite of the promising progress that has been made on RE doped bulk glasses. The first laser action in a rare-earth doped chalcogenide glass fibre was reported by T. Schweizer in 1997 [24], where laser action at 1080 nm was obtained by pumping a 22 mm long Neodymium doped Gallium Lanthanum Sulphide glass fibre with a Ti-sapphire laser at 815 nm. In 2000, optical amplification at 1.34 μm with a gain coefficient of 0.81 dB/mW was achieved in a single-mode Pr3+ doped Ga-Na-S (GNS) fiber [25]. However, in the following decades until now, no significant progress has been reported beyond these two demonstrations.
In the present work, six pieces of Ge25Ga10Se65 glass with different Er concentrations were prepared and their emission properties investigated in order to determine the effect of Er concentration and establish a baseline for the performance of Erbium in these hosts to compare with thin films. Optimized composition thin films were then deposited using co-thermal evaporation and the optical properties and PL, lifetime, and film propagation loss investigated. Finally a hybrid planar rib waveguide for single mode operation was designed and fabricated through photolithography and plasma etching techniques, as the initial steps towards realizing planar waveguide amplifiers and lasers.
2. Experiments
In order to understand the basic properties of Erbium in these hosts and to provide a benchmark for film based materials, Erbium doped Ge25Ga10Se65 ChGs with different Er concentrations were prepared from high purity Germanium, Gallium, Selenium and Erbium metals. This composition was chosen as it was expected to be a stoichiometry that should produce thin films with properties close to the bulk glass [26] and that had sufficient Gallium to enable ~1% Erbium to remain unclustered. The required amounts of these raw materials were weighed inside a dry nitrogen glove box and loaded into a pre-cleaned quartz ampoule. The ampoule was then sealed under vacuum (~10−6 torr) using an Oxygen-Hydrogen torch, and introduced into a rocking furnace for melting of the contents at 950 °C. The melt was homogenized for a period not less than 12 hours, then the ampoule was removed from the rocking furnace and water quenched. The resulting glass boule was subsequently annealed at a temperature 20 °C below its glass transition temperature Tg of ~395 °C, before being slowly cooled down to room temperature.
Various methods have been employed to deposit chalcogenide thin films including by sputtering [21], pulsed laser deposition [27, 28], chemical vapour deposition (CVD) [29] or thermal evaporation [26]. Each method has its own advantages and drawbacks. However for Gallium and rare earth containing materials there are additional issues, for example, in a typical thermal evaporation chamber, the respective evaporation rates of rare-earth atoms from the host glasses are so different that thermal evaporation leads to nearly un-doped films [30], and a similar issue also occurred in Gallium containing materials [29].
For this reason, co-thermal evaporation was employed, where each element has its own source and evaporation rate monitor, and thus the evaporation rate of each element can be controlled accurately and independently to select any desired film composition. High purity Ge, Ga, Se and Er elements were used as starting materials, and evaporation was performed at a vacuum level around ~1x10−7 Torr. The evaporation temperature of each source was slowly increased to achieve the desired evaporation rate, which in turn will determine the final composition of the film. After all the desired evaporation rates were achieved, the shutter covering the wafers was opened and the evaporation started. The whole process was monitored by Quartz microbalance thickness monitors, one for each elemental source and one for the wafers. The film thickness and linear refractive index (n) were measured by a dual angle spectroscopic reflectometer (SCI FilmTek 4000) using a Tauc-Lorentz model. The final film composition was determined by energy-dispersive X-ray spectroscopy (EDX).
The room temperature photoluminescence spectra (PL) of the bulk glass samples were measured using a Horiba Jobin Yvon 64000 spectrometer employing an 830 nm laser as the excitation source. An InGaAs detector installed on the spectrometer recorded PL spectra up to 1600 nm. The 4I13/2 excited state lifetime was measured by pumping the samples with modulated ~10 ms long rectangular pulses from a fiber pigtailed laser diode at a wavelength of 1490 nm using the all fiber confocal arrangement previously described in Ref [13]. With this wavelength, the effect of the possible transitions involved in higher excited states on the measured 4I13/2 level lifetime is reduced significantly. PL intensity data were also collected in this setup using 1490 and 980 nm pumps. Absorption spectra for the glasses were also recorded using a Cary 5000 UV-Vis-NIR spectrophotometer.
3. Results and discussion
3.1 PL and lifetime of bulk glasses
Figure 1(a) shows the PL spectra of the bulk glasses with different Er concentrations excited with a continuous wave 830 nm laser. All the spectra have similar line shape and peak positions. The emission peak lay at 1538 nm, which is slightly red-shifted from the 1532 nm usually observed in oxide glass hosts [31]. It is clear that a shoulder is growing with increase of Er concentration. Shoulder growth accompanied by increasing emission bandwidth has previously been observed in a number of studies, e.g [16], is believed in those cases to be associated with reabsorption of the emitted 1550 nm light in highly doped materials. Here, however, there is no sign of bandwidth increase, which suggests a different mechanism is in play, perhaps related to the higher order ~1550 nm transition between the 4H11/2 to 4I9/2 transition possible due to the 800 nm pump wavelength. Further study is needed to unambiguously determine the mechanism here.
Fig. 1 Fluorescence spectra of Ge-Ga-Se glasses with different Er concentrations excited at 830 nm (a); Emission intensity at 1538 nm with 830 nm excitation and absorption as a function of Er concentrations (b); Emission intensity at 1538 nm with different Er doped concentrations excited at 1490 nm (c).
Figure 1(b) shows the PL intensity at 1538 nm vs concentration for the 830 nm pump. Interestingly it shows an initial supra-linear increase in PL with concentration before becoming sub-linear at high concentrations as would be expected from Ion-Ion effects. Figure 1(c) shows the PL intensity with a 1490 nm pump source which displays the normally expected behavior of reducing PL efficiency with increasing concentration, as was also seen with a 980 nm pump. The second line in Fig. 1(b) plots the measured sample absorption in dB/cm at the 1550 nm absorption peak, which is linear and verifies that the supra-linear curvature in the 830 nm PL result is not due to errors in the Er concentration at low values, and also that the Erbium is not forming physical clusters at high concentration [15]. The origin of this effect remains uncertain and is the topic of ongoing investigation. However it is clear that the performance is beginning to decline significantly at concentrations beyond 1%, which for this glass composition could be considered optimum on the basis of the data presented. While considering the 10 mol% Ga concentration in all the glasses and the optimal ratio of Ga/RE at 10:1 noted in prior work [16], the limit of the homogeneous distribution of Er should be ~1 mol%. Therefore it was reasonably expected that Er-ions start to form clusters at an Er concentration more than 1 mol%, leading to the gradual quenching of PL emission at 1550 nm [32, 33] which fits with the measured data.
Based on the absorption spectrum of the 0.1 mol% sample, McCumber theory was employed to calculate the emission cross-section, and the results are shown in Fig. 2.The McCumber theory and measured emission cross-section agree very well except for the minor deviation in the calculated data in the 1520 nm tail region. Also on Fig. 2 a maximum pump efficiency curve is plotted which represents the maximum inversion possible versus pump wavelength taking into account the effect of pump initiated stimulated emission under the resonant pumping in this region. The curve was calculated according to the method in [34]. At 1490nm the efficiency is around 71%. At lower wavelength, the efficiency is higher but the absorption cross-section is significantly reduced leading to lower overall pump absorption. The results also indicate that for a highly doped short waveguide amplifier, the optimum pump wavelength lies in the 1500-1510 nm region where the absorption is approaching half its peak value whilst the inversion efficiency remains above 60%, as has been successfully employed in for example TeO2 based waveguide amplifiers [13]. This is a significantly longer pump wavelength than in silica or other oxide based hosts.
Fig. 2 Normalized absorption, emission spectra and simulated emission cross-section based on M-C theory, and max pump efficiency with pump at 1490 nm of the 0.1 mol% doped bulk glass
The dependence of PL intensity versus pump power has also been investigated for high and low doped samples, 0.1 mol% and 2 mol% Erbium concentrations respectively, with excitation at 1490nm, and the results are shown in Fig. 3.It is clear in both of the samples that PL intensity is increasing with the increase of pump intensity, but with obviously different trends. For the 0.1 mol% doped material, PL intensity increases linearly with pump intensity in the measured range which is low enough in intensity that there is still a significant fraction of Er in the ground state and no saturation of the emitted 1550nm radiation occurs. For the 2 mol% doped sample there is instead a quadratic relationship over the same pump power range. This quadratic trend in the highly doped glass results from radiative and non-radiative ion-ion energy exchange interactions which are not present in the lower doped samples due to the larger average distance between ions. To determine the exact processes participating requires a much more thorough examination which was beyond the scope of this initial study, but which will be reported at a future date.
PL lifetime is another important parameter to evaluate the performance of Er-doped ChG glasses. Because of the low phonon energy in ChG, the multiphonon relaxation rate between Er 4I11/2 and 4I13/2 levels is very low and the state has a long lifetime, comparable to that of the 4I13/2 level [35]. Therefore, the lifetime of the 4I13/2 to 4I15/2 transition when pumped by a 980 nm laser would be expected to be significantly longer than that under 1490 nm laser excitation due to energy storage at the 4I11/2 level. To avoid this issue, a 1490 nm laser was employed to probe the PL, using the confocal experimental configuration to avoid reabsorption and stimulated emission artifacts [13]. It is well known that as the Er3+ ion concentration increases that the average distance between neighboring ions diminishes, enabling a variety of energy transfer effects to become relevant. As many of these effects occur between two or more Er3+ ions in the 4I13/2 metastable level, these excited ions do not contribute to creating photons in the desired state and therefore reduce efficiency and also result in a decrease of the lifetime of the desired state with increasing ion concentration as the energy transfer is a very fast ion-ion interaction. Typically these effects have been found to follow an empirical formula first proposed in [36]:
where τobs is the observed fluorescence lifetime, τ0 is the ideal fluorescence lifetime with zero concentration, ρ is the Er3+ ion concentration, Q is the quenching concentration and p referred as phenomenological parameter characterizing the steepness of the corresponding quenching curve. As a several of the effects encapsulated in this relationship also depend (sometimes) nonlinearly on the number of excited ions, there is also a dependence of the parameters on pump power.The luminescence decay curves were measured for the glass samples over a range of different pump intensities (from ~10 to ~600 KW/cm2), the parameters recorded being the measured 1/e lifetime, and the “intrinsic” lifetime (ie the supposed radiative lifetime) in a region typically between the second and third decade down on the initial luminescence intensity where power dependent effects should have fallen to negligible levels. The lifetimes vs pump power for each glass were then fitted with suitable polynomials to extrapolate the data back to zero pump power to enable the effective “zero pump power lifetimes” to be extracted. Figure 4 presents the data for both the 1/e and intrinsic lifetimes at effective zero power and at ~600 kW/cm2 pump intensity.
Fig. 4 1/e and intrinsic lifetime of 4I13/2 metastable state of Er3+ with different doping concentrations (a) at extrapolated zero pump power (b) at high pump power (~600 KW/cm2 intensity).
There are a number of clear trends from the data. Looking first at the intrinsic lifetime, then it is independent of the pump power (as expected) and has a clear linear dependence on concentration with a slope of −0.48 ms/mol % Er. The slope of the line is also significantly lower than the −0.7 ms/mol% reported in [17]. However this improvement in performance still lags the better oxide glasses where essentially concentration independent radiative lifetimes have been observed up to about 2 mol% Er [37]. It also needs to be reinforced that sufficient Gallium is present to ensure Erbium clustering did not occur according to prior research [16], and the absorption data also verified that clustering was absent. The concentration quenching present is therefore more intrinsic in nature, occurring even at low Er doping and high Ga:Er ratios, and reasons for this relatively strong dependence contrary to those in some oxide glasses remain to be elucidated.
Both the ”zero” power and high power 1/e lifetimes show a more significant dependence on concentration and both can be fitted very well by the empirical formula mentioned above as shown in Fig. 4. The Q value referred as quenching concentration was 1.1 for the zero pump power case and 0.545 for the high pump power based on the fitting. This indicates that mechanisms that depend upon the inverted population (co-operative upconversion and/or excited state absorption process for example) are strongly active in these glasses. The relatively strong power dependence indicates that care is required in the design of amplifier devices using these materials where the pump intensity has to be considered relative to the Er concentration, and the likely tradeoffs this will impose. Of particular concern is the very considerable lifetime shortening at high power and high concentrations, the region of operation often used in planar waveguide amplifiers. Further study is needed to determine the dominant mechanism at play and potential remedies for this behavior. The high power lifetime shortening in particular suggests that concentrations below 1 mol% Er are strongly preferable for waveguide amplifier devices, though a trade off with the propagation loss then will determine the optimum concentration and design. Concentrations in the range 0.5-0.75mol% Er might be considered optimum as they would be expected to show up to about a factor of two reduction in the 1/e lifetime under high pumping compared to the radiative value (which is tolerable), and also lie in the high efficiency part of the PL intensity vs pump power density curve as shown above.
3.2 Film lifetime and loss measurement
A film with a thickness of 1062 ± 5 nm and refractive index of 2.433 ± 0.002 at 1550 nm was achieved by appropriately careful elemental co-evaporation, and its composition was measured to be 24.60% Ge, 10.94% Ga, 63.74% Se and 0.71% Er which was in the optimal region based on the PL and lifetime measurements of the bulk glasses as noted above. An issue encountered during film evaporations was “spitting” of Gallium particles out of the evaporation crucible. As will be discussed this lead to a moderate density of small particles in the films.
The lifetime of the 4I13/2 level of this erbium doped film was measured using the all fiber confocal set up with a 1490 nm pump delivered to the edge of the film via a lensed tip fiber with a 2.5 μm spot size. The intrinsic lifetime of this film was 0.87 ms, which is somewhat shorter than the ~1.35 ms observed in the corresponding bulk glass. Several factors could account for this. Firstly, the film was fabricated under non-equilibrium conditions in which a population of homopolar and ‘wrong’ bonds could be created, and these ‘wrong’ bonds may change the local environment for the Er ions thereby degrading the performance. Secondly, local erbium clusters may also be formed during the evaporation as pure Er metal was used as a source, leading to the quenching of the emission [38]. Thirdly, the nanoscale homogeneity of the films is uncertain compared to the bulk glasses. The bulk materials which were quenched from an equilibrium melt were expected to be located in a region of the phase diagram where the glass is essentially single phase, rather than nanoscale phase separated. Given the clear issues with the Gallium source and the non-equilibrium nature of the film growth the film homogeneity cannot be guaranteed, and the “granular” nature often observed in evaporated thin films may also be relevant here [39]. Further research is required to determine the precise cause and improve the Er lifetime in the films.
The dependence of PL intensity and 1/e lifetime on pump power was also investigated in the obtained films, which is shown in Fig. 5.It is clear the emission has quadratic pump power dependence indicating higher order processes being present, which is similar to the bulk glasses. The 1/e lifetime also decays with increase of pump intensity as seen in the bulk glasses. To compare the 1/e lifetime of film and bulk glass, the observed lifetime empirical formula mentioned above was employed again to estimate the 1/e lifetime of a bulk sample with 0.7 mol% Erbium at a pump intensity around 300Kw/cm2. This produced an expected 1/e lifetime in the 0.7% doped glass of 0.74ms compared to the 0.52ms measured in the film. The lifetime reduction in the film under pumping is smaller than that seen in the radiative lifetime (27% vs 35%), and the shape of the decay curve in Fig. 5(b) indicates that further reductions at the highest pump intensities expected in a typical waveguide device (up to ~1MW/cm2) will be modest leaving a workable lifetime around 0.5 ms.
Measurements of film loss and Er absorption were performed using a prism coupler. The scattering streaks from the TE fundamental slab guided mode was imaged and captured using a cooled high sensitivity InGaAs camera at wavelengths between 1450 and 1650 nm and then the propagation loss was calculated using custom image processing software by analyzing the decay of the normalized and background corrected scattering streak vs distance. The resulting optical loss of the film as well as a fitted absorption curve based on Erbium doped bulk glass is shown in Fig. 6.
Fig. 6 Absorption spectrum of Er doped films measured by prism coupler with a fitted curve based on erbium doped GeGaSe bulk glass.
A minimum value of 0.8 dB/cm at 1650 nm was found outside the Erbium absorption band. Given the observed particles in the as deposited film, it is likely that the loss is dominated by scattering off these particles. To investigate the film quality and particle induced scattering loss, the deposited film was inspected using dark field microscopy with x20 and x100 magnifications to estimate the particle density and size. Custom software was used to count the particles in numerous images to ascertain an average particle density and also to estimate the size distribution. From the results, the density of particles was ~0.01 particles/μm2, with a mean particle diameter from 1.0 to 1.1μm. The size distribution was always tightly bounded though with varying shape but the upper and lower bounds were 0.9 and 1.15 μm diameter based on multiple measurements of different parts of the wafer. Mie scattering would be expected to dominate in this size range (with probably relatively low wavelength dependence), and coupled with the density of particles (ie a 1 mm wide beam typical of the setup used would encounter 10,000 particles per mm of propagation length) and their likely metallic nature of the Gallium droplets would be expected to induce non negligible loss as was observed. It is expected that the particles can be eliminated with further refinement of the evaporation setup, most likely by using a baffled evaporation source that has no line of sight between the evaporant and the wafer.
3.3. Waveguide fabrication and test
To date most high quality waveguides in ChG have been fabricated by dry etching. However, Erbium doped glasses are difficult to etch using chemistries that are effective on the chalcogenide hosts since the resulting Erbium compounds are not volatile. The involatile Erbium compounds can then form micro-masks resulting in a very rough etched surfaces [13]. Fortuitously, the refractive index of the film is rather close to that of a fully annealed As2S3 film (n = 2.435 at 1550 nm) and so rib waveguides could be made by adding a top layer of As2S3 and etching this. Rib waveguides with total thickness of 1.35 μm, 0.35 μm rib depth and 2 μm rib width were therefore designed for single-mode operation, while the overlaps for both TE and TM fundamental modes with the Erbium doped area for this structure were around 89%. Bilayer films to match these thicknesses were then deposited on a thermally oxidized 100 mm diameter Silicon wafer (2.0 μm thick oxide). The layer stack consisted of Er doped GeGaSe with a thickness of 1 μm, followed by a 0.35 μm thick As2S3 layer. The ribs were structured using contact lithography and standard positive Photoresist followed by Reactive Ion Etching (RIE) with CHF3 gas for the layer of pure As2S3. Finally, the waveguides were top clad with a 10 μm thick film of UV cured polysiloxane polymer. An optical micrograph of the final device cross section and simulated TE fundamental mode are shown in Fig. 7.
Fig. 7 Structure of a 2 μm waveguide and simulated TE fundamental mode (The overlap for TE mode and Er doped area are around 89%).
Cut-back measurements were then performed at 1650 nm. The result, which is shown in Fig. 8, indicates the total coupling loss (mode mismatch plus reflection) of this waveguide is around 2.9 dB per facet, while propagation losses were 1.99 dB/cm for the 2 μm wide waveguide in the TE mode. As shown above a considerable amount of this loss results from scattering off the particles embedded in the film. Assuming that the waveguide film and the film previously tested were similar in terms of particle density, there is however additional loss present. This could only have come from waveguide sidewall roughness or increased particle density. Based on prior experience using exactly the same fabrication techniques in deeper etched thinner rib waveguides [7], then increased particle density is likely the cause. Nonetheless, the methodology for fabricating waveguides has been demonstrated and with elimination of the particulate problem, much lower losses would be anticipated.
4. Conclusions
Broad PL emission band centered at 1538 nm was observed from a series of GeGaSe bulk glasses with different erbium concentrations. The dependence of PL intensity and lifetime with increasing Erbium concentration was presented and discussed, clear concentration dependences being present leading to continuous monotonic decay of the lifetime with increasing Erbium concentration. These effects were magnified at high concentration and pumping power with ~0.2ms lifetimes being obtained for 2% doped samples under high pumping, suggesting that highly doped materials may not be suitable for very short waveguide based amplifiers/lasers. Further study is needed to determine the dominant concentration and power dependent processes and what can be done to ameliorate their effects. GeGaSe films doped with 0.7 mol% Erbium as an optimized value where the concentration effects are tolerable were then prepared by co-thermal evaporation and the films exhibited reasonable radiative lifetime of 0.87 ms and 1/e lifetime under high pumping of ~0.5 ms. An acceptable first demonstration propagation loss below 0.8 dB/cm, limited by particulates in the film from spitting off the Gallium source was also obtained in the film. Planar hybrid Er-Ge-Ga-Se/As2S3 rib single mode waveguides were fabricated with propagation losses of ~2 dB/cm. Further development of the co-evaporation scheme is required to eliminate the particulate issues which should lead to sufficiently low waveguide propagation losses to build working 1550nm amplifiers.
Acknowledgments
This research was supported by the Australian Research Council Centre of Excellence for Ultrahigh bandwidth Devices for Optical Systems (project number CE110001018). Dr Zhiyong Yang is supported by ARC DECRA project DE120101036.
References:
1. B. J. Eggleton, B. Luther-Davies, and K. Richardson, “Chalcogenide photonics,” Nat. Photonics 5, 141–148 (2011).
2. S. J. Madden, D. Y. Choi, D. A. Bulla, A. V. Rode, B. Luther-Davies, V. G. Ta’eed, M. D. Pelusi, and B. J. Eggleton, “Long, low loss etched As(2)S(3) chalcogenide waveguides for all-optical signal regeneration,” Opt. Express 15(22), 14414–14421 (2007). [CrossRef] [PubMed]
3. M. D. Pelusi, F. Luan, S. Madden, D. Y. Choi, D. A. Bulla, B. Luther-Davies, and B. J. Eggleton, “Wavelength Conversion of High-Speed Phase and Intensity Modulated Signals Using a Highly Nonlinear Chalcogenide Glass Chip,” IEEE Photonic Tech. L 22(1), 3–5 (2010). [CrossRef]
4. M. Galili, J. Xu, H. C. H. Mulvad, L. K. Oxenløwe, A. T. Clausen, P. Jeppesen, B. Luther-Davis, S. Madden, A. Rode, D. Y. Choi, M. Pelusi, F. Luan, and B. J. Eggleton, “Breakthrough switching speed with an all-optical chalcogenide glass chip: 640 Gbit/s demultiplexing,” Opt. Express 17(4), 2182–2187 (2009). [CrossRef] [PubMed]
5. X. Gai, S. Madden, D. Y. Choi, D. Bulla, and B. Luther-Davies, “Dispersion engineered Ge11.5As24Se64.5 nanowires with a nonlinear parameter of 136 W⁻¹m⁻¹ at 1550 nm,” Opt. Express 18(18), 18866–18874 (2010). [CrossRef] [PubMed]
6. B. J. Eggleton, B. Luther-Davies, and K. Richardson, “Chalcogenide photonics,” Nat. Photonics 5, 141–148 (2011).
7. D. Y. Choi, S. Madden, D. A. Bulla, R. Wang, A. Rode, and B. Luther-Davies, “Submicrometer-Thick Low-Loss As2S3 Planar Waveguides for Nonlinear Optical Devices,” IEEE Photonic Tech L 22(7), 495–497 (2010). [CrossRef]
8. S. Madden, Z. Jin, D. Choi, S. Debbarma, D. Bulla, and B. Luther-Davies, “Low loss coupling to sub-micron thick rib and nanowire waveguides by vertical tapering,” Opt. Express 21(3), 3582–3594 (2013). [CrossRef] [PubMed]
9. W. J. Miniscalco, “Erbium-Doped Glasses for Fiber Amplifiers at 1500 nm,” J. Lightw. Tech. 9(2), 234–250 (1991). [CrossRef]
10. G. C. Righini, M. Brenci, G. N. Conti, S. Pelli, M. Ferrari, M. Bettinelli, A. Speghini, and B. J. Chen, “Integrated optical amplifiers based on rare-earth doped (RED) oxide glasses,” International Symposium on Photonic Glass (Ispg 2002) 5061, 36–42 (2003).
11. A. Jha, S. X. Shen, L. H. Huang, B. Richards, and J. Lousteau, “Rare-earth doped glass waveguides for visible, near-IR and mid-IR lasers and amplifiers,” J. Mater. Sci. Mater. Electron. 18(S1), 315–320 (2007). [CrossRef]
12. E. M. Yeatman, M. M. Ahmad, O. McCarthy, A. Vannucci, P. Gastaldo, D. Barbier, D. Mongardien, and C. Moronvalle, “Optical gain in Er-doped SiO2-TiO2 waveguides fabricated by the sol-gel technique,” Opt. Commun. 164(1-3), 19–25 (1999). [CrossRef]
13. K. Vu and S. Madden, “Tellurium dioxide Erbium doped planar rib waveguide amplifiers with net gain and 2.8 dB/cm internal gain,” Opt. Express 18(18), 19192–19200 (2010). [CrossRef] [PubMed]
14. A. Z. Subramanian, G. S. Murugan, M. N. Zervas, and J. S. Wilkinson, “High index contrast Er:Ta2O5 waveguide amplifier on oxidised silicon,” Opt. Commun. 285(2), 124–127 (2012). [CrossRef]
15. M. Munzar, C. Koughia, D. Tonchev, K. Maeda, T. Ikari, C. Haugen, R. Decorby, J. N. McMullin, and S. O. Kasap, “Optical properties of Er-doped Ga-x(Ge0.3Se0.7)(100-x) glasses,” Opt. Mater. 28(3), 225–230 (2006). [CrossRef]
16. K. Koughia, M. Munzar, D. Tonchev, C. J. Haugen, R. G. Decorby, J. N. McMullin, and S. O. Kasap, “Photoluminescence in Er-doped Ge-Ga-Se glasses,” J. Lumin. 112(1-4), 92–96 (2005). [CrossRef]
17. D. T. Tonchev, K. V. Koughia, Z. G. Ivanova, and S. O. Kasap, “Thermal and optical properties of erbium doped (GeS2)(75)(Ga2S3)(25) glasses,” J. Optoelectron. Adv. Mater. 9, 337–340 (2007).
18. S. Kasap, K. Koughia, G. Soundararajan, and M. G. Brik, “Optical and Photoluminescence Properties of Erbium-Doped Chalcogenide Glasses (GeGaS:Er),” IEEE J Sel Top Quant 14(5), 1353–1360 (2008). [CrossRef]
19. T. W. Allen, M. M. Hawkeye, C. J. Haugen, R. G. DeCorby, J. N. McMullin, D. Tonchev, K. Koughia, and S. O. Kasap, “Photoluminescence measurements of Er-doped chalcogenide glasses,” J. Vac. Sci. Technol. A 22(3), 921–924 (2004). [CrossRef]
20. J. Fick, E. J. Knystautus, A. Villeneuve, F. Schiettekatte, S. Roorda, and K. A. Richardson, “High photoluminescence in Erbium-doped chalcogenide thin films,” J. Non-Cryst. Solids 272(2-3), 200–208 (2000). [CrossRef]
21. K. M. Takahiko Imai, M. Fujita, and N. Saito, “Photoluminescence properties of erbium-doped amorphous gallium-germanium-selenium films fabricated by RF sputtering,” Physica Status Solidi (C) 6, 106-109 (2009).
22. V. Nazabal, P. Nemec, A. M. Jurdyc, S. Zhang, F. Charpentier, H. Lhermite, J. Charrier, J. P. Guin, A. Moreac, M. Frumar, and J. L. Adam, “Optical waveguide based on amorphous Er3+-doped Ga-Ge-Sb-S(Se) pulsed laser deposited thin films,” Thin Solid Films 518(17), 4941–4947 (2010). [CrossRef]
23. S. Ramachandran and S. G. Bishop, “Excitation of Er3+ emission by host glass absorption in sputtered films of Er-doped Ge10As40Se25S25 glass,” Appl. Phys. Lett. 73(22), 3196–3198 (1998). [CrossRef]
24. T. Schweizer, B. N. Samson, R. C. Moore, D. W. Hewak, and D. N. Payne, “Rare-earth doped chalcogenide glass fibre laser,” Electron. Lett. 33(5), 414–416 (1997). [CrossRef]
25. H. Tawarayama, E. Ishikawa, K. Yamanaka, K. Itoh, K. Okada, H. Aoki, H. Yanagita, Y. Matsuoka, and H. Toratani, “Optical amplification at 1.3 mu m in a praseodymium-doped sulfide-glass fiber,” J. Am. Ceram. Soc. 83(4), 792–796 (2000). [CrossRef]
26. D. A. P. Bulla, R. P. Wang, A. Prasad, A. V. Rode, S. J. Madden, and B. Luther-Davies, “On the properties and stability of thermally evaporated Ge-As-Se thin films,” Appl. Phys Adv. Mater. 96, 615–625 (2009).
27. R. A. Jarvis, R. P. Wang, A. V. Rode, C. Zha, and B. Luther-Davies, “Thin film deposition of Ge33As12Se55 by pulsed laser deposition and thermal evaporation: Comparison of properties,” J. Non-Cryst. Solids 353(8-10), 947–949 (2007). [CrossRef]
28. M. Frumar, B. Frumarova, P. Nemec, T. Wagner, J. Jedelsky, and M. Hrdlicka, “Thin chalcogenide films prepared by pulsed laser deposition - new amorphous materials applicable in optoelectronics and chemical sensors,” J. Non-Cryst. Solids 352(6-7), 544–561 (2006). [CrossRef]
29. A. Abrutis, V. Plausinaitiene, M. Skapas, C. Wiemer, O. Salicio, M. Longo, A. Pirovano, J. Siegel, W. Gawelda, S. Rushworth, and C. Giesen, “Chemical vapor deposition of chalcogenide materials for phase-change memories,” Microelectron. Eng. 85(12), 2338–2341 (2008). [CrossRef]
30. D. J. Coleman, P. Golding, T. A. King, and S. D. Jackson, “Spectroscopic and energy-transfer parameters for Er3+-doped and Er3+, Pr3+-codoped GeGaS glasses,” J. Opt. Soc. Am. B 19(9), 1982–1989 (2002). [CrossRef]
31. S. L. Li, G. Fu, and Y. K. Ye, “1.54 um Photoluminescence emission from Er-implanted SiO2 crystal and SiO2 glass,” Nucl. Instrum. Meth. B 307, 434–437 (2013). [CrossRef]
32. F. Auzel and P. Goldner, “Towards rare-earth clustering control in doped glasses,” Opt. Mater. 16(1-2), 93–103 (2001). [CrossRef]
33. Y. Guimond, J. L. Adam, A. M. Jurdyc, H. L. Ma, J. Mugnier, and B. Jacquier, “Optical properties of antimony-stabilised sulphide glasses doped with Dy3+ and Er3+ ions,” J. Non-Cryst. Solids 256, 378–382 (1999). [CrossRef]
34. W. J. Miniscalco, “Erbium-Doped Glasses for Fiber Amplifiers at 1500-Nm,” J. Lightw. Tech. 9(2), 234–250 (1991). [CrossRef]
35. C. C. Ye, D. W. Hewak, M. Hempstead, B. N. Samson, and D. N. Payne, “Spectral properties of Er3+-doped gallium lanthanum sulphide glass,” J. Non-Cryst. Solids 208(1-2), 56–63 (1996). [CrossRef]
36. X. Orignac, D. Barbier, X. M. Du, R. M. Almeida, O. McCarthy, and E. Yeatman, “Sol-gel silica/titania-on-silicon Er/Yb-doped waveguides for optical amplification at 1.5 mu m,” Opt. Mater. 12(1), 1–18 (1999). [CrossRef]
37. S. X. Shen, A. Jha, X. B. Liu, M. Naftaly, K. Bindra, H. J. Bookey, and A. K. Kar, “Tellurite glasses for broadband amplifiers and integrated optics,” J. Am. Ceram. Soc. 85(6), 1391–1395 (2002). [CrossRef]
38. Z. G. Ivanova, K. Koughia, D. Tonchev, J. C. Pivin, and S. O. Kasap, “Photoluminescence in Er-implanted amorphous Ge-S-Ga thin films,” J. Optoelectron. Adv. Mater. 7, 1271–1276 (2005).
39. D. Y. Choi, S. Madden, A. Rode, R. P. Wang, A. Ankiewicz, and B. Luther-Davies, “Surface roughness in plasma-etched As2S3 films: Its origin and improvement,” IEEE T Nanotechnol 7(3), 285–290 (2008). [CrossRef]
