Open Access
Add to BibSonomyAbstract
The effect of iodine (I2) on the photoluminescence properties of Tm3+ ions in the As2S3 matrix was investigated. Results showed three strong emission bands at 1.22 µm (3H5→3H6), 1.46 µm (3H4→3F4) and 1.82 µm (3F4→3H6) under the excitation wavelength of 800 nm, indicating that I2 enables the solubility of Tm3+ ions in As2S3 glasses. The concentration ratio of I2 and Tm3+ were optimized and it revealed that the increasing of the concentration of I2 of 4 times in the glass increases the solubility of Tm ions three times. The effects of the I2 on the fundamental glass properties, including optical, thermal and structural characteristics, were explored as well.
© 2015 Optical Society of America
1. Introduction
One of the most interesting promises of chalcogenide glasses (ChG) is related to their capacity to be doped by rare-earth (RE) ions, with the aim to create new infrared (IR) radiation sources, lasers and amplifiers. The main advantage of ChG as a host matrix for RE ions is the low phonon energy (360–400 cm−1) comparing with fluoride (580 cm−1) or oxide (1100 cm−1) glasses [1]. This enables many RE transitions in the near and middle IR regions, which are normally quenched in silica or fluoride glasses.
The primary criteria in the search of an appropriate ChG host for the above-mentioned applications are the transparency in the visible and near IR regions, small optical losses at the pump and emission wavelengths, good thermal stability allowing fiber drawing and sufficient solubility for the RE ions, which is an important factor for ensuring the efficiency of an amplifier or laser. The difficulties to combine all of the required properties in a single ChG restrain their use in practical devices. A variety of chalcogenide systems, based on arsenic [2,3], germanium [4–6], gallium [7–10] and their combinations [11–20] were considered as host glasses for RE ions. Some of them have poor thermal stability (Ge–Ga–S [3,14,20] and Ge–La–S [21]), others exhibit insufficient solubility for RE ions (As–S [2] and Ge-S [4]).
Among all ChG systems, As2S3 glasses were most extensively studied because of their high thermal stability, broad transparency window (up to 7 µm) and ease of fiber fabrication [22]. Several demonstrations of RE doped As2S3 glasses have been made [2,3], however, it was established that only very small amount of REs can be incorporated in this glass matrix before the microscopic clustering and appearance of ion-ion interaction. RE ions require a large coordination number and the insufficient quantity of non-bridging sulfur, needed to coordinate the isolated REs in the network of As2S3, causes the clustering of ions.
According to previous reports, there are some glass modifiers, namely some metals (Ga, In) [23–27], which enhance the solubility of RE ions in ChG. The increased solubility of RE is related to the fact that the incorporation of Ga (In) provides compensation for the negative charge of free S2- ions by forming chemical bonds with RE ions. The effect of Ga in the solubility of Tm3+ ions in the glass matrix of As2S3 is reported in ref [26]. As it was shown three strong emission bands at wavelengths of 1.22 µm (3H5→3H6), 1.46 µm (3H4→3F4) and 1.82 µm (3F4→3H6) were obtained in the infrared emission spectra of Tm with the addition of Ga into the host. However, the glass formation region is very small in Ga-As-S system and only up to 3.5at% of Ga can be incorporated into the matrix [28] without crystallisation, which on its turn limits the solubility of higher concentrations of RE ions. Besides, optical losses in the metal (Ga/In) based chalcogenides are enhanced due to the multiphonon absorption of metal-chalcogen bonds.
Some reports indicated that halides (I2) [6,29–31] also can prevent the clustering of RE ions in ChG. Halide and chalcogenide glasses differ significantly in most aspects, including their methods of fabrication, basic glass and spectral properties. So, the incorporation of halogen elements into ChG (so-called chalcohalides) can reduce undesirable losses in the glass by mean of decreasing the number of metal-chalcogen bonds. This should lead to compromise between the improvement of optical properties and thermochemical instability incurred by the formation of weak metal-halogen bonds. Glass formation in I-As-S system was first reported by Flaschen et al. in 1960 [32]. The glass formation region is large, and glasses containing up to 33at% of I2 were successfully prepared.
The goal of this work is the study of the role of I2 in the solubility of RE ions in the As2S3 glass matrix and to obtain purified glass preform doped with RE ions. Different compositions of I-As-S glasses were used as host matrices for Tm3+ ions in order to find the optimal concentration ration of I2 and Tm for obtaining the highest possible luminescence efficiency. The effects of I2 and Tm on the optical, thermal, structural and photoluminescence properties of the glass are also investigated. In order to improve the optical quality of the glass matrix, a distillation of the glass matrix, with the optimal concentration of constituents, was performed under high vacuum and a Tm3+ doped high purity glass was produced.
2. Experiment
Glass synthesis: Tm3+ doped I-As-S bulk glasses were fabricated using the pre-purified As2S3 glass, solid iodine and metallic Tm3+. Traditional melt quenching method was applied. Glasses were synthesized in quartz ampoules sealed under high vacuum (10−5 Torr). The synthesis was carried out at the temperature of 750°C for 12 hours, and the glasses were cooled down with air. Samples were annealed at 130-160°C (depending on glass transition temperature) for 6 hours to remove the thermal stress. The glass rods were then cut into disks and polished with an abrasive silicon carbide disc to obtain flat and transparent bulk glasses.
Purification of glass matrix: ChG purification is performed in special furnace in a series of dynamic distillations under high vacuum. Our purification system consisted of three interconnected ampoules: reactor, receiver and trap (Fig. 1).
First, the glass matrix (without Tm) with desired chemical composition was prepared separately (at 600°C), loaded into the reactor and placed in the furnace. The Tm3+ was loaded in the receiver ampoule and maintained at room temperature. The system was attached to the high vacuum pump. The purification procedure included melting the I-As-S matrix at 250°C, followed by vacuum distillation of the melt. Extracted impurities condensed in the trap. Afterwards the receiver ampoule was sealed from both sides. Then the purified glass melt with Tm was homogenized in a rocking furnace at 750°C.
Characterisation: The emission spectra of Tm3+ doped I-As-S glasses were collected with a Horiba Jobin Yvon NanoLog 3-22-TRIAX spectrofluorometer. The samples were excited at 800nm wavelength with a 450 W continuous xenon lamp connected to a computer controlled monochromator. The luminescence spectra were measured with a Symphony II CCD detector operating in the infrared spectral region ranging from 1000nm to 2200 nm. Fluorescence lifetime measurements were also performed with a NIR PMT single channel detector with 1000 - 1700 nm wavelength range. The fluorescence decay curves were fitted with a first-order exponential decay to extract the emission lifetimes.
The IR transmission spectra from 1 to 20 µm were measured with a Perkin Elmer Frontier FT-IR/FIR Spectrometer using the factory-supplied MIR source and MIR TGS detector combination.
The refractive index was measured with a Metricon 2010/M prism coupler for different wavelengths (633nm, 972nm, 1308 nm and 1538 nm). The prisms 200-P-5 and 200-P-6 were used for these measurements. The measurement error was about ± 0.003.
The glass transition temperature (Tg) was measured with a Netzsch 404 F3 Pegasus differential scanning calorimeter (DSC). A 20 mg glass sample was placed in the aluminum crucible and heated up to 300 °C (10°C/min ramp) and the heat flow was compared with that of an empty crucible. The measurement error was ± 2°C.
The coefficient of linear thermal expansion of the samples was measured with Netzsch 402 F1 Hyperion thermomechanical analyzer. The samples (with the diameter of 10mm and height of 6mm) were brought into contact with push rod of fused silica, inserted in the furnace and heated up to 200°C with controlled temperature program. The length change in the sample was measured as a function of temperature by a highly sensitive inductive displacement transducer via a push rod. The measurement error was ± 0.05·10−6 1/°C.
The viscosity of glasses in the softening range was measured with Bansbach Easylift Theta US parallel plate high temperature viscometer. The disk of glass (with about 10 mm diameter and 6 mm high) is sandwiched between two parallel silica plates inside a well-insulated furnace. The samples were heated up to 300°C at a heating rate of 5°C/min, under the compressive load of 300mg. By means of recording the rate of the thickness change of the sample as a function of time (using a linearly variable differential transformer (LVDT)), the logarithm of viscosity is calculated by DilaSoft I program. The estimated error was about ± 3°C on the temperature values and ~10 ± 0.2 Poise on the viscosity values.
The X-ray diffraction (XRD) spectra of the glass samples were obtained with a Siemens D5000 X-ray diffractometer. The X-ray source was made of a Cu anode and emitted at a wavelength of 1.54 Å. XRD data were collected in reflection mode with Theta-Theta configuration (2Theta step = 0.02°, time step = 1.2 s) using a NaI scintillation counter detector. It was processed with the JADE 2.1 software based on the JCPDS database of International Centre for Diffraction Data (ICDD).
The Raman spectra were collected with a Renishaw inVia spectrometer coupled a Leica DM2700 microscope. A back-scattering geometry was used in the frequency range of 100cm−1-600cm−1. The excitation light source was vertically polarized He-Ne laser with 633nm wavelength and 17mW power. The laser beam was focused with a 50X long working distance objective, generating a sub-micron spot size containing a total power at the sample of approximately 5-10mW. The frequency uncertainty was estimated to be ± 2 cm−1. The deconvolution of Raman spectra were performed using curve fit function of Wire 4.1 software.
3. Results
3.1. Photoluminescence of Tm3+ in I-As-S matrix
As it is known from our previous work [26], no emission of Tm3+ ions was observed in the As2S3 glass matrix. The luminescence spectra of Tm3+ doped I-As-S glasses for excitation wavelength of 800 nm are presented in Fig. 2. As it can be seen, the I2 enables the solubility of Tm3+ ions in the As2S3 matrix. Three emission bands centered at 1.22 µm, 1.46 µm and 1.82 µm were observed in the spectra, corresponding to the optical transitions 3H5→3H6, 3H4→3F4 and 3F4→3H6, respectively (Fig. 2(b)) [33,34].
Fig. 2 Emission spectra of Tm3+ doped I-As-S glasses a.) 1%Tm doped I-As-S glasses; b.)Tm3+ energy levels diagram c.) Tm doped I2.5As39S58.5 glasses; d.) Tm doped I10As36S54 glasses (excitation at 800nm).
In order to find the maximal content of I2 that can be incorporated in the glass matrix of As2S3 to improve the Tm3+ emission properties, the concentration of I2 was gradually increased up to 15 at%, maintaining the S/As ratio equal to 1.5 (as in As2S3). The content of Tm3+ ions was fixed to 1 at% (presented atomic percentages are calculated). The emission properties of I-As-S glasses depending on the I2 concentration are presented in Fig. 2(a). With the increasing of I2 concentration from 2.5 to 10%, the intensity of luminescence gradually increased for all three emission bands. At 15% the intensity decreased. It was thus concluded that ratio I2/Tm3+ equal to 10:1 corresponds to the maximum ratio to improve the solubility of Tm ions in the glass.
To find the limit of solubility of Tm ions in the I-As-S system (depending on the content of I2 in the glass), two glass compositions with low and high content of I2, namely I2.5As39S58.5 and I10As36S54, were chosen as host matrices for Tm ions. The luminescence spectra of Tm3+ doped I2.5As39S58.5 and I10As36S54 glasses depending on Tm concentration are presented in the Figs. 2(c) and 2(d), respectively. As we can see in I2.5As39S58.5 glass matrix the maximum intensity is observed for 2at%Tm, so we can suppose that the solubility of Tm is limited to 2at% since above the 4%Tm the intensity of luminescence decreased. Contrariwise, in the I10As36S54 matrix the highest photoluminescence efficiency reached at 6 at% of Tm.
With increasing the amount of Tm up to 8% we can note that the intensity of emission decreased, indicating that the limit of solubility of Tm ions in this matrix is about 6%. From these results, we can conclude that the increasing of the concentration of I2 by a factor of 4 in the As2S3 glass matrix increases the solubility of Tm ions three times.
The fluorescence lifetimes of Tm3+ doped I-As-S glasses for excitation wavelength of 800 nm were measured, and no substantial changes have been obtained depending on the concentration of I2 and Tm. The lifetime of emission band at 1.22µm band is about 1,8·10−5 and at 1.46μm it is about 7,9·10−5. The measurement error was about ± 0.1·10−5.
As the glass matrix I10As36S54 has enhanced RE ions solubility, preserving good optical properties, further characterisations (optical, thermal and structural) are presented hereafter for this glass matrix as a function of Tm concentration.
3.2. Optical properties
The infrared transmission spectra of Tm3+ doped I-As-S glasses as a function of the concentration of I2 and Tm are shown in Fig. 3(a) and 3(b), respectively. The spectrum of un-doped and purified As2S3 glass is presented for comparison. The maximum transmission of I-As-S glasses is lower than that of As2S3, but the increase of I2 concentration in I-As-S leads to the increase of maximum transmission of glass and broadens the transmission range up to 9μm (Fig. 3(a)).
Fig. 3 IR transmission spectra of Tm doped I-As-S glasses a.) 1%Tm doped I-As-S glasses; b.) Tm doped I10As36S54 glasses (thickness is 2mm).
As it can be seen from Fig. 3(b) the increase of Tm concentration in the I10As36S54 glass leads to the decrease of maximum transmission of the glass. The absorption bands observed in the spectra of the doped glasses are attributed to impurities coming from traces of oxygen and hydrogen and they are assigned to S-H (3.96 um), O-H (3.06um), H2O (6.62um) and As-O (10.1um) [35]. This is due to the fact that the doped glasses were synthesized using commercial chemical products and have not been additionally purified during the preparation.
Dependence of the refractive indices of Tm doped I-As-S glass on the concentration of I2 and Tm are presented in Fig. 4(a) and 4(b), respectively. The incorporation of I2 in the As2S3 glass matrix and the increase of its content up to 15% gradually decreased the refractive index (Fig. 4(a)). Contrariwise, the increase of the concentration of Tm increased the refractive index (Fig. 4(b)). For 8%Tm the refractive index increased by ~0.1 compared with the host matrix I10As36S54.
Fig. 4 Refractive index of Tm doped I-As-S glasses a.) 1%Tm doped I-As-S glasses; b.) Tm doped I10As36S54 glasses (error ± 0.003).
3.3. Thermal properties
Figures 5(a) and 5(b) show the glass transition temperature (Tg), and the coefficient of thermal expansion (α) of Tm3+ doped I-As-S glasses versus the content of I2 and Tm, respectively. The addition of I2 in the As2S3 matrix and the increasing of its concentration gradually decreased the Tg of glass from 200°C to 114°C. Simultaneously, the coefficient of thermal expansion increased from 21.4·10−6 to 39.5·10−6. On the other hand with the increase of Tm content up to 8% in the matrix I10As36S54 the Tg gradually increases (Fig. 5(b)) from 133°C to 150°C. Meanwhile, the coefficient of thermal expansion decreased from 33.9·10−6 to 28.8·10−6.
Fig. 5 Glass transition temperature and coefficient of thermal expansion of Tm doped I-As-S glasses a.) 1%Tm doped I-As-S glasses; b.) Tm doped I10As36S54 glasses (Tg error is ± 2°C, α error is ± 0.005·10−6 1/°C).
Figure 6 presents the viscosity-temperature curves of Tm3+ doped I-As-S glasses in the glass softening range depending on the concentration of I2 and Tm. It may be seen that the extrusion viscosity of As2S3 is about 1010.3 at 234°C. With the increase of the content of I2 in the glass the viscosity-temperature curves have tended to shift gradually to a lower temperature and at 15%I2 the temperature of extrusion viscosity decreases up to 139°C (Fig. 6(a)). From these data it follows that at equal temperatures the viscosity of I2 based glasses is considerably less than that of As2S3. In a similar way, the addition of Tm began to increase the temperature of extrusion viscosity and for 8%Tm it reached to 190°C, comparing with I10As36S54 matrix (152°C) (Fig. 6(b)).
Fig. 6 Viscosity vs. inverse of absolute temperature for Tm doped I-As-S glasses in the softening range a.) 1%Tm doped I-As-S glasses; b.) Tm doped I10As36S54 glasses
Generally it is established that the suitable viscosity at which the fiber can be drawn is in the range of 106.5-105.5 Poises, so to find the temperature of fiber drawing for this system, the viscosity data were extrapolated into the range up to 105 Poise, as it is shown in Fig. 6. It revealed that the addition of up to 15%I2 in the As2S3 glass decreases the fiber drawing temperature by nearly 90°C. In contrast, the incorporation of up to 8%Tm increased the fiber drawing temperature of I10As36S54 glass matrix by 35°C.
Considering the linear relation between the log(η) and temperature (1/T), the activation energy was calculated by the following equation [36]:
where R is the gas constant. The results are presented in the Table 1. The obtained activation energy of As2S3 glass is equal to 32.72kcal/mole and with the addition of 2.5%I2 and 1%Tm in As2S3 it increases up to 37.59kcal/mole. However, the further increase of the content of I2 decreases the activation energy up to 21.6kcal/mole (at 15%I). Similarly, the activation energy of I10As36S54 matrix (M) is equal to 23.53kcal/mole and upon the addition of Tm (up to 8%) it increased gradually up to 29.04kcal/mole. This may indicate that at 2.5%I2 the contribution of 1%Tm on the properties of As2S3 is more significant compared to I2, which lead to the increase of activation energy.
Table 1. Activation energy of Tm doped I-As-S glasses vs content of I2 and Tm3+
3.4. Structural properties
To understand which kind of structural changes occur in the As2S3 glass matrix by the incorporation of I2 and which mechanisms are responsible for the luminescence of Tm3+ ions, structural studies were done on I-As-S system using X-ray diffraction (XRD) and Raman spectroscopy.
The XRD patterns recorded for Tm3+ doped I-As-S glasses are presented in Fig. 7. The patterns associated to 1%Tm doped I-As-S samples reveal broad diffraction lines indicating an amorphous character (Fig. 7(a)). Figure 7(b) shows the XRD spectra of Tm3+ doped I10As36S54 glasses depending on Tm concentration. As it can be seen the samples having ≤6at%Tm are completely amorphous (no apparent crystallization). However, for the concentration of Tm = 8at% small diffraction peaks, assigned to Tm2S3 crystalline phase (JCPDS card No. 44-1157), arise from the broad vitreous profile, which provide evidence in the favour of partial crystallisation of the glass at higher Tm concentrations.
Fig. 7 X-ray diffraction spectra of Tm doped I-As-S glasses a.)1%Tm doped I-As-S glasses; b.) Tm doped I10As36S54 glasses
Raman spectra (normalized at 342cm−1) of 1%Tm3+ doped I-As-S glasses depending on I2 concentration are presented in the Fig. 8. From the Raman spectra of As2S3 glass (Fig. 8(a)) we can observe 2 main bands centered at 342 cm−1 and 495 cm−1. The most intense broad band, centered at 342cm−1, is assigned to the symmetric stretching vibrational modes of [AsS3/2] regular pyramids [37–40]. The band at 495 cm−1 is associated to vibrational mode of S-S bridges in S2As-S-S-AsS2 fragment [37,39,41].
Fig. 8 a.) Raman spectra of 1%Tm3+ doped I-As-S glasses, b.) Raman bands intensities vs concentration of I2.
The addition of I2 to As2S3 gives rise to two bands centered at 188 cm−1, 211 cm−1, (Fig. 8(a)). Besides, with the incorporation of I2 the band at 495cm−1 increases. The band centered at 188cm−1 is attributed to the Tm-S bond (inset in Fig. 8(a)). According to Khiminets et al. the band at 210 cm−1 is assigned to stretching vibration of IAsS2/2 groups [42]. They investigated the Raman spectra of quasibinary join AsI3-As2S3 of the ternary As-S-I system and suggested that the incorporation of AsI3 in the matrix of As2S3 leads to the formation of IAsS2/2-S2/2AsI type of molecules, so called twisted chain structure. Whereas, Koudelka et al. propose that the band at 208cm−1 in the Raman spectra of As40-xS60Ix glasses is assigned to symmetrical stretching vibration of discrete AsI3 pyramidal molecules [43], claiming that the formation of discrete molecules is more favorable due to the low total energy of the system with symmetric groups (AsI3, As2S3 and S8 rings). They assumed that the structure is composed of AsS3/2 pyramidal units mixed with dissolved units of AsI3 and S8. In these Raman data no S8 ring (band at 475cm−1) was observed. With the addition of I2 we observed only an increase of 495cm−1 band (S-S bridges) accompanied by a decrease of the band at 342cm−1 (showing a dissociation of AsS3/2 pyramidal units). So we associate the band at 211cm−1 to IAsS2/2 units. The same twisted chain structure has been proposed by Hopkins et al. [44] when studying the structure of the I-As-S system by x-ray diffraction, showing that the arsenic is coordinated to two sulphur atoms and one iodine atom on average.
In order to understand the role of the addition of I2 on the solubility of RE into the matrix of arsenic sulfide, the Raman spectra of studied glasses were deconvoluted with Gaussian peaks using least squares approach. The intensity of each band was calculated by integrating the surface of the corresponding peak for all Raman modes present in the spectra. The resulting structural trends for each assigned structural units depending on I2 contents are presented in the Fig. 8(b). It can be seen that the band corresponding to the symmetric stretching of [AsS3/2] pyramids (342cm−1) decreases with the addition of I2 in the matrix, whereas the band assigned to AsIS2/2 structural units (211cm−1) significantly increases in amplitude. Simultaneously, the bands at 188cm−1 (Tm-S) and at 495cm−1 (S-S) increase with the increase of the I2 concentration.
Figure 9(a) and 9(b) show Raman spectra (normalized at 342cm−1) of Tm doped I10As36S54 glasses and intensities of observed bands depending on Tm concentration (1-8 at%). As we can see AsS3 pyramids stay almost intact upon the increase of Tm concentration up to 4%, whereas the band 188cm−1 (Tm-S) continues to increase slightly, simultaneously leading to a decrease of 495cm−1 (S-S). At higher concentration of Tm (>4 at%) it seems that the intensity of Tm-S band reaches saturation. However the band IAsS2/2 begins to decrease, and meanwhile an increase of AsS3 bands is observed.
Fig. 9 a.) Raman spectra of Tm3+ doped I10As36S54 glasses (dash lines represent deconvoluted curves) b.) Raman bands intensities vs concentration of Tm.
We expected that the dissociation of IAsS2/2 units and the reformation of AsS3 pyramids should lead to the formation of Tm-I linkages. The formation energy of TmI3 (0.477eV/atom (OQMD ID: 347708)) is much lower comparing to the formation energy of Tm2S3 (2.262eV/atom (OQMD ID: 23780)), so the formation of Tm-I bonds should be more advantageous. However no band associated to Tm-I (168cm−1, inset in Fig. 9(a)) was detected in the Raman spectra by deconvolution. Probably this band is small and just overlapped with the stronger bands at 188cm−1 and 211cm−1. This can be explained with the fact that the formation of Tm-S bonds is more favourable in these glasses because of the much smaller atomic masse of S comparing to I2. Additionally, we may note that the higher noises obtained in Raman spectra of TmI3 are related to the lower melting temperature of this compounds. The bands at 310 cm−1 and 380 cm−1 (Fig. 9(a)) included in the deconvolution of spectra correspond to interactions among the AsS3/2 pyramids [37].
3.5. Purification of Tm doped I-As-S glass
One of the key requirements of chalcogenide materials is purity, because it has decisive effect on their application. Hydrogen, oxygen, carbon and other impurities considerably reduce the transmission of chalcogenide glasses especially in middle infrared spectral region, which prevent their applications. Besides, the presence of OH- complex in the glass has an important implication in the luminescence efficiency of RE ions [45]. At high OH- concentrations, a direct energy transfer occurs from the excited ion to OH-, which has adverse effect on luminescence properties of the glass. So the purification of glass matrix is one of the most important steps in the procedure of fabrication of doped glasses.
Based on above shown results, the 2%Tm3+ doped I10As36S54 glass composition was chosen for performing an additional purification to improve the quality of the glass matrix. Figure 10 shows the infrared transmission spectra of above mentioned bulk glass synthesized with as-received chemicals and with purified matrix. As it can be seen with only one distillation cycle the impurity losses, attributed to the fundamental S-H vibrational (3.96um), S-H combinational (3.78um) [46] and H2O vibrational (6.62um) bands, are dramatically decreased in the glass. Additionally, the transparency of the purified glass is improved comparing to the non-purified.
Fig. 10 Infrared transmission spectra of 2%Tm3+doped I10As36S54 glass before and after purification.
It is known that the luminescence efficiency and the emission lifetime of RE ions in the glass strongly depend on the concentration of OH-groups in the matrix [45,47]. Since initially the concentration of OH-group was small in the matrix (Fig. 10), the additional purification did not change luminescence properties of this glass. The same intensity and the lifetime of emission of Tm3+ have been obtained for non-purified glass and the glass with purified matrix (not presented).
Previous study presented by our group [26] indicates that the Ga can dissolve Tm3+ ions in the matrix of As2S3 as well. However, the drawback of the Ga based chalcogenide glasses is the required high reaction temperature and long synthesis procedure, due to the low reactivity of Ga with sulfur. Also, Ga2S3 compound cannot be distillated during the glass purification, because of its high melting point (1090°C) and therefore it leads to the contamination of the sulfide with impurities originating from the walls of the ampoule. In contrast, the low melting temperature of I2 (113.7°) makes the purification of the raw materials (pre-processing) and the glass purification (post-processing) easier with less consuming time.
So we can conclude that the I2 can completely replace the Ga because of its sufficient solubility for the RE ions, larger transparency region in IR, enough thermal stability for fiber drawing and promising way to decrease optical loses in these glass matrices.
4. Discussion
First we will discuss the structure of Tm doped I-As-S glasses. Based on the Raman spectra of As2S3 glass the three-dimensional network of As2S3 glass can be interpreted using trigonal pyramidal units [AsS3/2], which are interconnected through As-S-S-As bridges. With the incorporation of I2 the AsS3/2 pyramidal units are disassociated by means of the formation of As-I bonds, which lead to the formation of AsIS2/2 structural units. So, we can suppose that our system mostly consists of the AsS3 pyramids mixed with AsIS2/2 twisted chains as it is schematically presented in the Fig. 11. This network configuration provides sufficient number of non-bridging sulfur for coordinating Tm ions in the rigid network of As2S3, which leads to the increase of the solubility of Tm ions in the glass. As we can see from Fig. 8b, the S-S bonds increase more slowly compared with AsIS2/2 and Tm-S bonds, showing that the large part of non-bridging sulfur was bonded with Tm.
When the content of I2 reaches 15% the intensity of AsIS2/2 band become higher than that of AsS3/2 (Fig. 8(b)), showing that the structure consists of more twisted chains than polymeric pyramids. This probably leads to the degradation of some fundamental properties of As2S3 glass, incurred by the formation weak metal-halogen bonds, which may cause the decrease of the luminescence of the Tm at 15at% of I2.
With the increase of Tm up to 6% the intensity of Tm-S band continues to increase (see Fig. 9(b)), showing that more Tm-S bonds are formed in the glass leading to an increase of luminescence. At higher concentrations of Tm (>6 at%) the formation of Tm-S bonds seems to reach saturation and luminescence intensity decreases (Fig. 1(d)). So we can assume that at 8at% of Tm3+ the ions begin to cluster because of the insufficient quantity of non-bridging sulfur. Therefore, the mean distance between Tm3+ ions in the lattice becomes small enough so the transfer of excitation energy to neighbouring ions becomes possible, which depopulates the excited states, leading to the decrease of the emission intensity.
According to Hook’s law, the frequency of bond vibration is inversely proportional to the atomic mass of elements in the glass. The incorporation of I2 and Tm (with much higher atomic masses comparing to As and S) results in a decrease in the frequency of bond vibration and therefore, the long-wavelength cut-off edge of transmission increases (Fig. 3(a)). Usually, the lower transmission is caused by the presence of inhomogeneities or the formation of structural defects (for example As-As or Tm-Tm homo-polar bonds) in the glass. We can suppose that the incorporation of 1%Tm in the I2.5As39S58.5 matrix has created some structural defects in the glass leading to the decrease of transmission (Fig. 3(a)). This could be explained by the fact that the contribution of Tm clusters in the properties of this composition is more significant comparing to I2, since the 2.5at%I is not enough to provide sufficient content of non-bridging sulfur for the dissolution of Tm ions. The higher activation energy of 1%Tm doped I2.5As39S58.5 glass is due to the same reason (Table 1). However, the increase of I2 content reduces the number of As-S bonds (responsible for higher absorption coefficient), by forming As-I bonds, and simultaneously supplies the sulphur necessary to form Tm-S hetero-bonds, leading to the formation of more homogeneous glass with less structural defects. This leads to the increase of the maximum transmission of the glass. The decrease of transmission of I10As36S54 with the further increasing of Tm concentration (Fig. 3(b)) in the glass is related to the formation of Tm2S3 crystals in the glass, as it is shown in the Fig. 7(b).
The introduction of halogen promotes the dissociation of the As-S structure favouring the destruction of the compact character of the glass network of As2S3, which may result in the decrease of its density [48], despite the higher atomic mass of I2 compared with sulfur and arsenic. This leads to the decrease of refractive index of the glass (as well as the viscosity) with the increase of I2 concentration (Fig. 4(a)). In agreement with the classical dielectric theory, the refractive index depends on the density and the polarizability of atoms in a given material. According to the Lorenz-Lorenz relation [49] larger is the atomic radius of the atom, larger will be its polarizability and consequently larger will be the refractive index (Fig. 4(b)). So in our case the increase of Tm concentration probably increases the polarizability [50] of the glass matrix, due to its much larger atomic radius leading to the increase of the refractive index.
It is known that in As2S3 glasses As-S bond strength is about 90.7 kcal/mole [51], and the incorporation of I2 in the glass eliminates the As from S-As-S bonds through the formation of As-I (70.9 kcal/mole) [51] and S-S (33 kcal/mole) lower strength bonds [22], which leads to the decrease of bonding strength of the glass and therefore of the Tg. The atomic distance As-S is about 2.28Å [52], whereas As-I distance is about 2.45Å [48] so the formation of As-I bonds in the glass network causes a deformation of the layer structure and an expansion of interlayer distance. Consequently, it becomes easy for the glass to expand thermally.
5. Summary
The incorporation of I2 in the network of As2S3 glasses increase the solubility of Tm3+ ions, through the dissociation of AsS3/2 pyramidal units and the formation of As-I bonds, which provides non-bridging sulfur for RE ions. Three strong emission bands of Tm3+ ions at wavelengths around 1.2 µm (3H5→3H6), 1.4 µm (3H4 →3F4) and 1.8 µm (3F4→3H6) were observed under the excitation by a lamp at 800 nm. The increase of the concentration of I2 by a factor of 4 in the glass increases the solubility of Tm ions three times. We have determined the best concentration ratio of elements in the glass to produce optimal physical, optical and thermal properties of the glass for achieving high stability, high luminescence efficiency and low losses.
We think that for the fabrication of efficient and compact sources of laser radiation operating in near infrared region, the RE doped I-As-S glass fiber can be considered as a serious contender.
Acknowledgments
The authors would like to thank the Canadian Excellence Research Chair program (CERC) on Enabling Photonic Innovations for Information and Communication for their financial support. The Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canada Foundation for Innovation (CFI) agencies are also acknowledged.
The work was performed according to collaboration in the frame of project of Russian Scientific Foundation RSF 15-12-20040.
References and links
1. A. K. Mairaj, Optical waveguides and lasers in improved gallium Lanthanum sulphide glass (Thesis, University of Southampton, June 2003) Chap. 1.
2. J. Heo, “Optical characteristics of rare-earth-doped sulphide glasses,” J. Mater. Sci. Lett. 14(14), 1014–1016 (1995). [CrossRef]
3. Y. Ohishi, A. Mori, T. Kanamori, K. Fujiura, and S. Sudo, “Fabrication of praseodymium-doped arsenic sulfide chalcogenide fiber for 1.3-um fiber amplifiers,” Appl. Phys. Lett. 65(1), 13–15 (1994). [CrossRef]
4. D. R. Simons, A. J. Faber, and H. de Waal, “GeSx glass for Pr3+-doped fiber amplifiers at 1.3 μm,” J. Non-Cryst. Solids 185(3), 283–288 (1995). [CrossRef]
5. B. Frumarová, P. Němec, M. Frumar, J. Oswald, and M. Vlček, “Synthesis and optical properties of the Ge–Sb–S:PrCl3 glass system,” J. Non-Cryst. Solids 256–257, 266–270 (1999). [CrossRef]
6. V. Krasteva, D. Machewirth, and G. Sigel Jr., “Pr3+-doped Ge–S–I glasses as candidate materials for 1.3μm optical fiber amplifiers,” J. Non-Cryst. Solids 213–214, 304–310 (1997). [CrossRef]
7. D. N. Payne, J. A. Medeiros Neto, D. W. Hewak, R. I. Laming, and B. N. Samson, “Emission at 1.3µm from dysprosium-doped Ga:La:S glass,” Electron. Lett. 30(12), 968–970 (1994). [CrossRef]
8. K. Kadono, H. Higuchi, M. Takahashi, Y. Kawamoto, and H. Tanaka, “Upconversion luminescence of Ga2S3-based sulfide glasses containing Er3+ ions,” J. Non-Cryst. Solids 184, 309–313 (1995). [CrossRef]
9. T. Schweizer, D. J. Brady, Y. D. West, T. Schweizer, and D. J. Bra, “Gallium Lanthanum Sulphide Fibers for Infrared Transmission,” Fiber Integr. Opt. 19(3), 229–250 (2000).
10. R. Reisfeld and A. Bornstein, “Absorption and emission spectra in chalcogenide glass of the composition 0.7Ga2S3·0.27La2S3·0.03Nd2S3,” Chem. Phys. Lett. 47(1), 194–196 (1977). [CrossRef]
11. J. Heo, W. Y. Cho, and W. J. Chung, “Sensitizing effect of Tm3+ on 2.9 μm emission from Dy3+-doped Ge25Ga5S70 glass,” J. Non-Cryst. Solids 212(2–3), 151–156 (1997). [CrossRef]
12. K. Abe, H. Takebe, and K. Morinaga, “Preparation and properties of Ge-Ga-S glasses for laser hosts,” J. Non-Cryst. Solids 212(2–3), 143–150 (1997). [CrossRef]
13. J. Heo and Y. B. Shin, “Absorption and mid-infrared emission spectroscopy of Dy3+ in Ge-As(or Ga)-S glasses,” J. Non-Cryst. Solids 196, 162–167 (1996). [CrossRef]
14. K. Wei, D. P. Machewirth, J. Wenzel, E. Snitzer, and G. H. Sigel Jr., “Spectroscopy of Dy3+ in Ge-Ga-S glass and its suitability for 1.3-microm fiber-optical amplifier applications,” Opt. Lett. 19(12), 904–906 (1994). [CrossRef] [PubMed]
15. Y. S. Han, J. Heo, and Y. B. Shin, “Cross relaxation mechanism among Tm3+ ions in Ge30Ga2As6S62 glass,” J. Non-Cryst. Solids 316(2–3), 302–308 (2003). [CrossRef]
16. V. G. Truong, A. M. Jurdyc, B. Jacquier, B. S. Ham, A. Q. Le Quang, J. Leperson, V. Nazabal, and J. L. Adam, “Optical properties of thulium-doped chalcogenide glasses and the uncertainty of the calculated radiative lifetimes using the Judd-Ofelt approach,” J. Opt. Soc. Am. B 23(12), 2588–2596 (2006). [CrossRef]
17. V. Moizan, V. Nazabal, J. Troles, P. Houizot, J.-L. Adam, J.-L. Doualan, R. Moncorgé, F. Smektala, G. Gadret, S. Pitois, and G. Canat, “Er3+-doped GeGaSbS glasses for mid-IR fibre laser application: Synthesis and rare earth spectroscopy,” Opt. Mater. (Amst) 31(1), 39–46 (2008). [CrossRef]
18. F. Prudenzano, L. Mescia, L. Allegretti, M. De Sario, F. Smektala, V. Moizan, V. Nazabal, J. Troles, J. L. Doualan, G. Canat, J. L. Adam, and B. Boulard, “Simulation of mid-IR amplification in Er3+-doped chalcogenide microstructured optical fiber,” Opt. Mater. (Amst) 31(9), 1292–1295 (2009). [CrossRef]
19. Z. Yang, L. Luo, and W. Chen, “The 1.23 and 1.47μm emissions from Tm3+ in chalcogenide glasses,” J. Appl. Phys. 99(7), 076107 (2006). [CrossRef]
20. K. Wei, D. P. Machewirth, J. Wenzel, E. Snitzer, and G. H. Sigel Jr., “Pr3+-doped Ge-Ga-S glasses for 1.3 μm optical fiber amplifiers,” J. Non-Cryst. Solids 182(3), 257–261 (1995). [CrossRef]
21. 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]
22. S. Tsuchihashi and Y. Kawamoto, “Properties and structure of glasses in the system As-S,” J. Non-Cryst. Solids 5(4), 286–305 (1971). [CrossRef]
23. A. B. Seddon, Z. Tang, D. Furniss, S. Sujecki, and T. M. Benson, “Progress in rare-earth-doped mid-infrared fiber lasers,” Opt. Express 18(25), 26704–26719 (2010). [CrossRef] [PubMed]
24. T. H. Lee, S. I. Simdyankin, J. Hegedus, J. Heo, and S. R. Elliott, “Spatial distribution of rare-earth ions and GaS4 tetrahedra in chalcogenide glasses studied via laser spectroscopy and ab initio molecular dynamics simulation,” Phys. Rev. B 81(10), 104204 (2010). [CrossRef]
25. C. Lin, S. Dai, C. Liu, B. Song, Y. Xu, F. Chen, and J. Heo, “Mechanism of the enhancement of mid-infrared emission from GeS2-Ga2S3 chalcogenide glass-ceramics doped with Tm3+,” Appl. Phys. Lett. 100(23), 231910 (2012). [CrossRef]
26. A. Galstyan, S. H. Messaddeq, V. Fortin, I. Skripachev, R. Vallée, T. Galstian, and Y. Messaddeq, “Tm3+ doped Ga–As–S chalcogenide glasses and fibers,” Opt. Mater. (Amst) 47, 518–523 (2015). [CrossRef]
27. H. Sakr, D. Furniss, Z. Tang, L. Sojka, N. A. Moneim, E. Barney, S. Sujecki, T. M. Benson, and A. B. Seddon, “Superior photoluminescence (PL) of Pr³⁺-In, compared to Pr³⁺-Ga, selenide-chalcogenide bulk glasses and PL of optically-clad fiber,” Opt. Express 22(18), 21236–21252 (2014). [CrossRef] [PubMed]
28. B. Kolimiets, N. Goryunova, and V. Shilo, Glassy State in Chalcogenides (Publishing House of the USSR Academy of Sciences, 1960), pp. 456–460.
29. D. Machewirth, K. Wei, V. Krasteva, R. Datta, E. Snitzer, and G. Sigel Jr., “Optical characterization of Pr3+ and Dy3+ doped chalcogenide glasses,” J. Non-Cryst. Solids 213–214, 295–303 (1997). [CrossRef]
30. M. F. Churbanov, I. V. Scripachev, V. S. Shiryaev, V. G. Plotnichenko, S. V. Smetanin, E. B. Kryukova, Y. N. Pyrkov, and B. I. Galagan, “Chalcogenide glasses doped with Tb, Dy and Pr ions,” J. Non-Cryst. Solids 326–327, 301–305 (2003). [CrossRef]
31. D. Shixun, P. Bo, Z. Pengjun, X. Tiefeng, X. Wang, N. Qiuhua, and Z. Xianghua, “The near- and mid-infrared emission properties of Tm3+-doped GeGaS–CsI chalcogenide glasses,” J. Non-Cryst. Solids 356(44–49), 2424–2428 (2010). [CrossRef]
32. S. S. Flaschen, A. D. Pearson, and W. R. Northover, “Low Melting Sulfide-Halogen Inorganic Glasses,” J. Appl. Phys. 31(1), 219–220 (1960). [CrossRef]
33. T. Schweizer, B. N. Samson, J. R. Hector, W. S. Brocklesby, D. W. Hewak, and D. N. Payne, “Infrared emission and ion–ion interactions in thulium- and terbium-doped gallium lanthanum sulfide glass,” J. Opt. Soc. Am. B 16(2), 308–316 (1999). [CrossRef]
34. B. J. Cole and M. L. Dennis, “Heavy metal modified silica glass fibers doped with thulium, holmium, and thulium-sensitized-holmium high quantum efficiencies.” United States Patent No: US 6,667, 257 (B2 23Dec-2003).
35. G. E. Snopatin, V. S. Shiryaev, V. G. Plotnichenko, E. M. Dianov, and M. F. Churbanov, “High-purity chalcogenide glasses for fiber optics,” Inorg. Mater. 45(13), 1439–1460 (2009). [CrossRef]
36. S. Suzuki, Y. Kamiya, Y. Suzuki, and T. Kobayashi, “Viscosity of Glass in the Transition Region in the Systems As-S, As-S-I and As-S-Tl,” J. Soc. Mater. Sci. Japan 21(221), 143–147 (1972). [CrossRef]
37. T. Wagner, S. O. Kasap, M. Vlcek, A. Sklenár, and A. Stronski, “Modulated-temperature differential scanning calorimetry and Raman spectroscopy studies of AsxS100−x glasses,” J. Mater. Sci. 33(23), 5581–5588 (1998). [CrossRef]
38. R. Golovchak, O. Shpotyuk, J. S. Mccloy, B. J. Riley, C. F. Windisch, S. K. Sundaram, A. Kovalskiy, and H. Jain, “Structural model of homogeneous As–S glasses derived from Raman spectroscopy and high-resolution XPS,” Philos. Mag. 90(34), 4489–4501 (2010). [CrossRef]
39. E. Bychkov, M. Miloshova, D. L. Price, C. J. Benmore, and A. Lorriaux, “Short, intermediate and mesoscopic range order in sulfur-rich binary glasses,” J. Non-Cryst. Solids 352(1), 63–70 (2006). [CrossRef]
40. M. S. Iovu, S. D. Shutov, A. M. Andriesh, E. I. Kamitsos, C. P. E. Varsamis, D. Furniss, A. B. Seddon, and M. Popescu, “Spectroscopic study of As2S3 glasses doped with Dy, Sm and Mn,” J. Non-Cryst. Solids 326–327, 306–310 (2003). [CrossRef]
41. F. Kyriazis, A. Chrissanthopoulos, V. Dracopoulos, M. Krbal, T. Wagner, M. Frumar, and S. N. Yannopoulos, “Effect of silver doping on the structure and phase separation of sulfur-rich As–S glasses: Raman and SEM studies,” J. Non-Cryst. Solids 355(37–42), 2010–2014 (2009). [CrossRef]
42. O. V. Khiminets, P. P. Puga, V. V. Khiminets, I. I. Rosola, and G. D. Puga, “Raman spectra of the As-S-I system of glasses,” J. Appl. Spectrosc. 28(4), 477–479 (1978). [CrossRef]
43. L. Koudelka and M. Pisárčik, “Raman spectra and structure of As-S-I system glasses,” J. Non-Cryst. Solids 64(1–2), 87–94 (1984). [CrossRef]
44. T. E. Hopkins, R. A. Pasternak, E. S. Gould, and J. R. Herndon, “X-ray diffraction study of arsenic trisulfide-iodine glasses,” J. Phys. Chem. 66(4), 733–736 (1962). [CrossRef]
45. V. P. Gapontsev, S. M. Matitsin, A. A. Isineev, and V. B. Kravchenko, “Erbium glass lasers and their applications,” Opt. Laser Technol. 14(4), 189–196 (1982). [CrossRef]
46. M. Zhu, X. Wang, C. Jiang, H. Xu, Q. Nie, P. Zhang, S. Dai, X. Shen, T. Xu, G. Tao, and X. Zhang, “Freely adjusted properties in Ge–S based chalcogenide glasses with iodine incorporation,” Infrared Phys. Technol. 69, 118–122 (2015). [CrossRef]
47. V. P. Gapontsev, Y. E. Sverchkov, A. K. Gromov, A. A. Izyneev, and V. B. Kravchenko, “Parameters and mechanism of elementary interaction of the electronic excitation of rare-earth ions with the local impurity oscillations,” 29(4), 210–213 (1979).
48. J. Heo, “Chalcohalide glasses for infrared fiber optics,” Opt. Eng. 30(4), 470–479 (1991). [CrossRef]
49. Y. Y. Huang, A. Sarkar, and P. C. Schultz, “Relationship between composition, density and refractive index for germania silica glasses,” J. Non-Cryst. Solids 27(1), 29–37 (1978). [CrossRef]
50. E. R. Shaaban, Y. A. M. Ismail, and H. S. Hassan, “Compositional dependence of the optical properties of amorphous Se80−xTe20Bix thin films using transmittance and reflectance measurements,” J. Non-Cryst. Solids 376, 61–67 (2013). [CrossRef]
51. Y.-R. Luo, Comprehensive Handbook of Chemical Bond Energies (CRC Press, 2007) p. 488.
52. D. Zhao, Atomistic Mechanisms and Kinetics of Photoinduced Changes in Chalcogenide Glasses (Thesis, Lehigh University 2011).
