- Email: [email protected]
Adhesive As-S-Se-I immersion lenses for enhancing radiation characteristics of mid-IR LEDs operating in wide temperature range
Infrared Physics & Technology 78 (2016) 167–172
Contents lists available at ScienceDirect
Infrared Physics & Technology journal homepage: www.elsevier.com/locate/infrared
Regular article
Adhesive As-S-Se-I immersion lenses for enhancing radiation characteristics of mid-IR LEDs operating in wide temperature range Viktor A. Markov a,⇑, Alexandr V. Semencha a, Mikhail V. Kurushkin a, Dmitry V. Kurushkin a, Viktor A. Klinkov a, Andrey A. Petukhov b a b
Peter the Great St. Petersburg Polytechnic University, Polytechnicheskaya, 29, 195251 St. Petersburg, Russian Federation Microsensor Technology LLC, Postbox 100, 194223 St. Petersburg, Russian Federation
h i g h l i g h t s LED integral power as a function of refractive index for two lens shapes has been simulated. Chalcogenide melt adhesion force towards electronic engineering materials has been determined. LED stability against cracking in wide temperature range has been calculated and experimentally justified.
a r t i c l e
i n f o
Article history: Received 16 April 2016 Revised 27 July 2016 Accepted 28 July 2016 Available online 29 July 2016 Keywords: IR immersion lens Chalcogenide glass Mid-IR LED Adhesion force Stability against cracking
a b s t r a c t The influence of As-S-Se-I chalcogenide glass lenses on the integral and spectral power and pattern of LED radiation has been shown. Simulation of the influence of the refractive index on the integral power for two lens shapes has been performed. The wettability and adhesion force of As-S-Se-I melt has been determined for several electronic engineering materials. Mechanical stresses between chalcogenide glass and adjacent diode body materials have been calculated for 100 to 53 °C temperature range. Stability of the immersion lenses against cracking has been experimentally investigated for 150 to 53 °C temperature range. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Near and middle infrared devices are widely used for various applications in the 1–5 lm working range, such as chemical gas sensors, biophotonics and biomedicine [1]. Latest research in the field is dedicated to the development [2,3] and application [4,5] of IR semiconducting sources and detectors operating in wide temperature range [6,7]. However, their wide application is hindered by the low radiation output from the crystal due to the total internal reflection at the interface between the highrefractive crystal and air, significant radiation scattering and the expensiveness of suitable optical materials. Shaping the radiant crystal into a lens shape is an unreasonably expensive solution, therefore, there is a need in a material that
Abbreviations: CG, chalcogenide glass; CM, chalcogenide melt.
⇑ Corresponding author.
E-mail address: [email protected] (V.A. Markov). http://dx.doi.org/10.1016/j.infrared.2016.07.020 1350-4495/Ó 2016 Elsevier B.V. All rights reserved.
would transmit in the working range, possess a high index of refraction (in order to decrease the critical angle) and provide a full optical contact with the LED’s crystal. Polymeric materials are unsuitable due to the low index of refraction (usually less than 1.7). The production of lenses from monocrystalline materials is expensive, furthermore, their application demands the use of optical adhesives. The most efficient solution is to cover the LED with an immersion lens made of glass which is transmitting in the working range and possesses both low glass softening point (Ts < 100 °C) and a high refraction index (n > 2). We have chosen the As-S-Se-I chalcogenide glasses as satisfying the abovementioned requirements. Chalcogenide glass immersion lenses multiply the radiation output coefficient several times (depending on their index of refraction and shape) and alter the radiation pattern by giving the lens a different shape. Low Ts is determined by the application temperature of the lens not exceeding 200 °C to avoid the p-n transition degradation.
168
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
2. Materials and methods 2.1. Chalcogenide glass characteristics
has been determined in order to calibrate the method. The deviation did not exceed 7% [11]. Liquid/solid work of adhesion has been derived from Young– Dupre equation:
Chalcogenide glass As12.8S24.2Se53I10, has been used for immersion. The glass characteristics are the following: Ts = 53 °C, for refractive index dispersion see Fig. 1. Glass softening point was measured with the use of BÄHR DIL802 difference-dilatometer. Refractive index has been measured by interference fringes method [8] using an FSM-1201 FTIR spectrometer.
W LS ¼ rglass ð1 þ cos hÞ;
2.2. Lens application
CG thermal expansion coefficient has been measured with the use of BÄHR DIL802 difference-dilatometer (temperature range 100 to 53 °C, fused silica inert sample, liquid nitrogen cooling). Mechanical properties have been determined through measurements of the longitudinal ultrasonic wave velocities (VL) and transverse ultrasonic wave velocities (VT) analogous to those described in [13].
The lenses were applied on LMS34 LEDs (with kovar bodies) by method of dispensing chalcogenide melt (CM), T = 150–180 °C, lens shape was adjusted by varying the mass of the drop. 2.3. Radiation patterns, radiation spectra and integral power Radiation patterns have been plotted based on the data obtained with the use of an Lms36PD-05 photodiode and a rotating table. Monochromator MDR-41 with a Judson J10D-M204-R01M-60 detector have been used for the recording of radiation spectra. THORLABS IS200 series integration sphere and a Judson J10DM204-R01M-60 detector have been used to measure integral power.
where rglass = 111 mJ/m2 – the surface tension of the used CM, cos h – the wetting angle cosine. The determined surface tension is in good agreement with that of an analogous CM given in [12]. 2.5. Thermal and mechanical properties
2.6. Temperature testing Experimental testing of LED stability against cracking has been performed with the use of BÄHR DIL802 difference-dilatometer (temperature range 150 to 53 °C, LED as the experimental sample, liquid nitrogen cooling). On course of the experiment, the LED was periodically switched on in order to examine its workability. After low-temperature testing visual inspection was performed with the use of a Leica EZ4 HD microscope.
2.4. Wetting angle and surface tension 2.7. Simulation Wetting angle has been determined by means of Drop Shape Analysis method [9]. The drop mass did not exceed 0.05 g. The height and radius of the drop have been measured with the use of Leica EZ4 HD microscope. CM surface tension has been determined by means of pendant drop tensiometry method [10]. Surface tension of sulfur at 150 °C
The increment of the integral power (DP) has been calculated with the use of the following formula:
DP ¼
PCG 100%; P0
where P0 is the integral power of the LED without a lens, PCG - the integral power of the LED with a lens. Type A lens possesses the minimal size sufficient to completely immerse the radiant crystal and LED’s upper electric contact (Fig. 4, left). Type B lens is the one that has shown maximum experimental increase of integral power (Fig. 5, left). LEDs with two lens shapes have been simulated in Zemax software. Numerical simulation of the radiation output from the crystal and radiation patterns has been performed in the ‘‘non-sequential mode”. A rectangular radiation source, simulating a radiative p-n junction, was put inside of the LED crystal (InAs) of the size 350 350 250 mm. The wavelength of the radiation source matched the wavelength for the LED’s radiation peak intensity (3.4 lm). The influence of the diode body material absorption has been omitted (‘‘mirror” type material option). 3. Results and discussion 3.1. LED radiation patterns, radiation spectra and current-voltage characteristic before and after immersion
Fig. 1. Refractive index dispersion.
LED images with and without a lens are given in Fig. 2. LED radiation patterns, radiation spectra and current-voltage characteristic before and after immersion are given in Fig. 3. LEDs with immersion lenses possess enhanced spectral power and narrowed radiation patterns, current-voltage characteristic do not change.
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
169
Fig. 2. LED without a lens (daylight, left); LED with a lens (810 ± 5 nm, right).
3.2. The influence of the refractive index on the increment of the integral power To perform simulation, initial experimental data (radiation patterns and integral power) for two types of lenses has been acquired. In order to justify the applicability of the utilized models, simulation results have been compared to the experimental radiation patterns and the increment of the integral power (Figs. 4 and 5). Simulation results are in good agreement with experimental ones, the difference being motivated by the approximation of lens shapes with simple geometric shapes. Real lens shapes possess more complicated geometry. In order to evaluate the influence of the refractive index on the increment of the integral power we have simulated LEDs with type A and B lenses having a set of different indices of refraction (see Fig. 6). The non-linear dependences are motivated by the opposite contribution of two factors: (1) The increase of the angle of total internal reflection on the crystal/CG. (2) The decrease of the angle of total internal reflection on the CG/air interface. The optimal refractive index for the immersion lenses is between 2.6 and 2.8. Contrary to monocrystalline lenses of fixed refractive index values, there is an opportunity to alter the CG refractive index. 3.3. Adhesion force of CG melt towards electronic engineering materials
Fig. 3. Before and after immersion: radiation patterns (top), radiation spectra (middle), current-voltage characteristic (bottom).
One of the crucial properties of the appropriate for the production of As-S-Se-I immersion lenses CM is its wettability towards diode body and the semiconductor crystal. In case of low wettability (h < 90 °C) the CM incompletely covers the upper diode platform and, hence, the lens becomes asymmetrical. Furthermore, for type B lenses, low wettability can cause the CM flow off the upper diode platform. In [14] it has been shown that liquid/solid work of adhesion (WLS) is proportional to the corresponding solid/solid adhesion strength (WSS). It is expected that at high values of work of adhesion of the CM towards the diode body material the junction between the immersion lens and the adjacent diode body material will be more secure. We have determined the work of adhesion of CM towards several electronic engineering materials in order to qualitatively evaluate the relative difference in the corresponding adhesion strength of CG towards said materials.
170
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
Fig. 4. Simulated type A lens shape (left); radiation patterns and the increment of the integral power: experiment and calculation (right).
Fig. 5. Simulated type B lens shape (left); radiation patterns and the increment of the integral power: experiment and calculation (right).
Wetting angle and WLS of electronic engineering materials are given in Table 1. The highest wettability and maximal work of adhesion is observed for CM/kovar and CM/InAs which are used for the production of the diode body and the radiant crystal, respectively. Based on [14], it can be proposed that among abovementioned materials the adhesion strength is maximal for CG/kovar and CG/InAs. As expected, the lowest wettability has been proved to be for CM/fluoroplastic, hence, the latter can be used as a mold for lens production of desired geometry. 3.4. Stability against cracking
Fig. 6. Simulated LED integral power as a function of refractive index for type A and B lens shapes.
Table 1 Wetting angle and work of adhesion. No
Material
cos h
h, degrees
WLS, mJ/m2 (180 °C)
1 2 3 4 5 6 7 8 9 10 11 12
Kovar InAs TiAlN FR4 textolite ZrN NbN Amorphous graphite CrMoN Fused silica TiN Stainless steel Fluoroplastic
0.37 0.31 0.20 0.20 0.16 0.15 0.15 0.14 0.13 0.11 0.07 0.20
68.3 ± 0.9 71.5 ± 1.1 78.2 ± 0.4 78.4 ± 0.5 80.8 ± 0.6 81.2 ± 0.6 81.3 ± 0.1 81.9 ± 0.4 82.6 ± 0.3 83.7 ± 1.1 85.9 ± 0.7 100.9 ± 0.6
152 145 133 133 129 128 128 127 125 123 119 89
After application, the chalcogenide lens becomes an integral part of the LED and is bound to the diode body via adhesion force. Due to the fact that the CG and the adjacent materials possess different thermal expansion coefficients, mechanical stresses emerge within the CG/diode junction. The numerical value of those stresses depend both on the mechanical and thermal properties of the materials and the temperature range of exploiting. Upon reaching critical stresses, there is a possibility of lens cracking or separation from the diode body. We have calculated mechanical stresses within the junction between CG and LED body materials based on their mechanical and thermal properties. Stability against cracking has been experimentally tested for the 150 to 53 °C temperature range. In order to evaluate mechanical stresses in the junction between CG and diode body material (Fig. 7), we have used a method devised by [15]. Mechanical stresses have been calculated in an approximation of instant hardening (absence of relaxation) with the use of the formula:
r¼
EG ðaCG aÞ DT; 1 lCG ECG dCG þ 1 Ed
where r - mechanical stress, ECG – Young’s modulus, lCG – Poisson ratio, aCG – thermal expansion coefficient, dCG – density;
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
171
Fig. 8. Stability against cracking experimental apparatus. 1 – Pushrods, 2 – inert sample (fused silica), 3 – sample holder back panel, 4 – LED, 5 – CG, 6 – LED driver.
Fig. 7. CG dilatometer curve (top) and mechanical stresses (bottom).
Table 2 Diode body materials mechanical properties.
a
Fig. 9. Experimental change in length of the LED.
Material
a 106, 1/R
E, GPa
d, g/cm3
Source
Kovar Stainless steela Aluminiuma Fused silicaa
5 16.6 23.8 0,6
196 212 62,1 75
8.35 7.90 2.70 2.20
[16] [15] [15] [15]
For comparison.
Table 3 CG mechanical properties. VL, m/c
VT, m/c
lCG
ECG, GPa
GCG, GPa
dCG, g/cm3
a 106, 1/R
1860 ± 10
940 ± 10
0.33
9.1 ± 0.2
3.5 ± 0.1
3.86
Fig. 7 top
E, l, a, d – the corresponding properties of the diode body material, DT = (Ts T), Ts = 53 °C (see Tables 2 and 3). According to [15,17,18], chalcogenide glasses can withstand mechanical stresses up to approximately 100 MPa. It can be seen in Fig. 7 that mechanical stresses in the junction between CG and any of the diode body materials do not reach critical value in the 100 °C to 53° temperature range. In order to confirm theoretical calculations, experimental testing of LEDs has been performed in the 150 to 53 °C temperature range. The apparatus is described in Fig. 8. Experimental change in length of the LED (DL) is given in Fig. 9. Jumps of DL have not been registered upon cooling or subsequent heating to ambient temperature (see Fig. 9), LED remained workable on course of the whole experiment. This indicates that the electrical contacts stayed undamaged. Visual inspection has not shown lens defects.
4. Conclusion The application of As-S-Se-I melt upon a LED results in the formation of an immersion CG lens (Ts = 53 °C) due to surface tension force. LED damage does not occur due to low (less than 200 °C) application temperature. Lens shape can be adjusted by varying the mass of the applied melt. LEDs with immersion lenses possess enhanced spectral power and narrowed radiation patterns. A maximum increase of integral power (about ten times) has been achieved for type B lenses. LEDs with two lens shapes have been simulated in Zemax software. The good agreement of simulation results with experimental ones justify the applicability of the utilized models. The influence of the refractive index on the integral power has been simulated. It has been determined that the optimal refractive index for the immersion lenses, analogous to those described in the present work, is between 2.6 and 2.8 (k = 3.4 lm). The wettability of As-S-Se-I melt has been determined for several diode body materials. The highest wettability among abovementioned materials has been proved to be for kovar, while the lowest – for fluoroplastic, hence, the latter can be used as a mold for lens production of desired geometry. Mechanical stresses between CG and adjacent diode body materials (e.g. quartz glass, kovar, stainless steel, aluminium) have been calculated. It has been calculated that for 100 to 53 °C temperature range critical stress (100 MPa) is not reached with any of the materials. The calculation has been justified by the
172
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
experimental data (the LED does not crack in 150 to 53 °C temperature range). Acknowledgement The research was partially supported by FASIE. We kindly acknowledge Alfred B. Sinani for the determination of glass mechanical properties by ultrasonic method. References [1] A. Krier (Ed.), Mid-Infrared Semiconductor Optoelectronics, Springer, London, 2006, http://dx.doi.org/10.1007/1-84628-209-8. [2] B.E. Zhurtanov, N.D. Il’inskaya, A.N. Imenkov, M.P. Mikhaĭlova, K.V. Kalinina, M. A. Sipovskaya, et al., Low-noise photodiodes based on GaSb/GaInAsSb/ AlGaAsSb double heterostructures for the 1–4.8 lm spectral range, Semiconductors 42 (2008) 458–462, http://dx.doi.org/10.1134/ S1063782608040155. [3] N.V. Zotova, N.D. Il’inskaya, S.A. Karandashev, B.A. Matveev, M.A. Remennyi, N. M. Stus’, Sources of spontaneous emission based on indium arsenide, Semiconductors 42 (2008) 625–641, http://dx.doi.org/10.1134/ S1063782608060018. [4] A.P. Astakhova, A.S. Golovin, N.D. Il’inskaya, K.V. Kalinina, S.S. Kizhayev, O.Y. Serebrennikova, et al., High-power InAs/InAsSbP heterostructure leds for methane spectroscopy (k 3.3 lm), Semiconductors 44 (2010) 263–268, http://dx.doi.org/10.1134/S1063782610020235. [5] N.D. Stoyanov, K.M. Salikhov, K.V. Kalinina, S.S. Kizhaev, A.V. Chernyaev, Super low power consumption middle infrared LED-PD optopairs for chemical sensing, in: M.J.F. Digonnet, S. Jiang (Eds.), Proc. SPIE – Int. Soc. Opt. Eng., SPIE, 2014, p. 89821A, http://dx.doi.org/10.1117/12.2036277. [6] P.N. Brunkov, N.D. Il’inskaya, S.A. Karandashev, N.M. Latnikova, A.A. Lavrov, B. A. Matveev, et al., P-InAsSbP/n 0-InAs/n +-InAs photodiodes for operation at moderate cooling (150–220 K), Semiconductors 48 (2014) 1359–1362, http:// dx.doi.org/10.1134/S1063782614100066.
[7] P.N. Brunkov, N.D. Il’inskaya, S.A. Karandashev, A.A. Lavrov, B.A. Matveev, M.A. Remennyi, et al., InAsSbP/InAs0.9Sb0.1/InAs DH photodiodes (k0.1 = 5.2 lm, 300 K) operating in the 77–353 R temperature range, Infrared Phys. Technol. 73 (2015) 232–237, http://dx.doi.org/10.1016/j.infrared.2015.09.017. [8] F. Carreño, J.C. Martínez-Antón, E. Bernabeu, Optical interference method to obtain thickness and refractive indices of a uniaxial medium, Rev. Sci. Instrum. 65 (1994) 2489, http://dx.doi.org/10.1063/1.1144707. [9] Y. Yuan, T.R. Lee, Contact angle and wetting properties, 2013, pp. 3–34. http:// dx.doi.org/10.1007/978-3-642-34243-1_1. [10] J.D. Berry, M.J. Neeson, R.R. Dagastine, D.Y.C. Chan, R.F. Tabor, Measurement of surface and interfacial tension using pendant drop tensiometry, J. Colloid Interf. Sci. 454 (2015) 226–237, http://dx.doi.org/10.1016/j.jcis.2015.05.012. [11] D.L. Timrot, S.A. Traktueva, B.A. Alekseev, Surface tension of liquid sulfur, High Temp. 21 (1983) 671–676.. [12] T.D. Mel’nichenko, V.I. Fedelesh, T.N. Mel’nichenko, D.S. Sanditov, S.S. Badmaev, D.G. Damdinov, On the approximate estimation of the surface tension of chalcogenide glass melts, Glas. Phys. Chem. 35 (2009) 32–42, http:// dx.doi.org/10.1134/S1087659609010052. [13] G. Yang, B. Bureau, T. Rouxel, Y. Gueguen, O. Gulbiten, C. Roiland, et al., Correlation between structure and physical properties of chalcogenide glasses in the AsxSe1x system, Phys. Rev. B 82 (2010) 195206, http://dx.doi.org/ 10.1103/PhysRevB.82.195206. [14] V.Y. Traskine, Z.N. Skvortsova, Estimation of adhesive tensile and wearing strengths from work of adhesion on the liquid – solid interface, Vestn. Mosk. Univ. Seriya 2 Khimiya. 45 (2004) 376–381. . [15] N.O. Gonchukova, V.A. Ananichev, S.L. Ratushnyak, Calculation of stresses in seals of chalcogenide glasses with other materials, Glas. Phys. Chem. 32 (2006) 206–213, http://dx.doi.org/10.1134/S1087659606020131. [16] D.D.L. Chung, Composite Materials, Springer, London, 2003, http://dx.doi.org/ 10.1007/978-1-4471-3732-0. [17] G.M. Camilo, Mechanical properties of chalcogenide glasses: a review, Proc. SPIE – Int. Soc. Opt. Eng. 4940 (2003) 222–229, http://dx.doi.org/10.1117/ 12.488150. [18] E. Hartouni, J.J. Mecholsky, Mechanical properties of chalcogenide glasses, in: R.W. Schwartz (Ed.), Proc. SPIE – Int. Soc. Opt. Eng., 1986, pp. 92–97, http://dx. doi.org/10.1117/12.936421.
Contents lists available at ScienceDirect
Infrared Physics & Technology journal homepage: www.elsevier.com/locate/infrared
Regular article
Adhesive As-S-Se-I immersion lenses for enhancing radiation characteristics of mid-IR LEDs operating in wide temperature range Viktor A. Markov a,⇑, Alexandr V. Semencha a, Mikhail V. Kurushkin a, Dmitry V. Kurushkin a, Viktor A. Klinkov a, Andrey A. Petukhov b a b
Peter the Great St. Petersburg Polytechnic University, Polytechnicheskaya, 29, 195251 St. Petersburg, Russian Federation Microsensor Technology LLC, Postbox 100, 194223 St. Petersburg, Russian Federation
h i g h l i g h t s LED integral power as a function of refractive index for two lens shapes has been simulated. Chalcogenide melt adhesion force towards electronic engineering materials has been determined. LED stability against cracking in wide temperature range has been calculated and experimentally justified.
a r t i c l e
i n f o
Article history: Received 16 April 2016 Revised 27 July 2016 Accepted 28 July 2016 Available online 29 July 2016 Keywords: IR immersion lens Chalcogenide glass Mid-IR LED Adhesion force Stability against cracking
a b s t r a c t The influence of As-S-Se-I chalcogenide glass lenses on the integral and spectral power and pattern of LED radiation has been shown. Simulation of the influence of the refractive index on the integral power for two lens shapes has been performed. The wettability and adhesion force of As-S-Se-I melt has been determined for several electronic engineering materials. Mechanical stresses between chalcogenide glass and adjacent diode body materials have been calculated for 100 to 53 °C temperature range. Stability of the immersion lenses against cracking has been experimentally investigated for 150 to 53 °C temperature range. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Near and middle infrared devices are widely used for various applications in the 1–5 lm working range, such as chemical gas sensors, biophotonics and biomedicine [1]. Latest research in the field is dedicated to the development [2,3] and application [4,5] of IR semiconducting sources and detectors operating in wide temperature range [6,7]. However, their wide application is hindered by the low radiation output from the crystal due to the total internal reflection at the interface between the highrefractive crystal and air, significant radiation scattering and the expensiveness of suitable optical materials. Shaping the radiant crystal into a lens shape is an unreasonably expensive solution, therefore, there is a need in a material that
Abbreviations: CG, chalcogenide glass; CM, chalcogenide melt.
⇑ Corresponding author.
E-mail address: [email protected] (V.A. Markov). http://dx.doi.org/10.1016/j.infrared.2016.07.020 1350-4495/Ó 2016 Elsevier B.V. All rights reserved.
would transmit in the working range, possess a high index of refraction (in order to decrease the critical angle) and provide a full optical contact with the LED’s crystal. Polymeric materials are unsuitable due to the low index of refraction (usually less than 1.7). The production of lenses from monocrystalline materials is expensive, furthermore, their application demands the use of optical adhesives. The most efficient solution is to cover the LED with an immersion lens made of glass which is transmitting in the working range and possesses both low glass softening point (Ts < 100 °C) and a high refraction index (n > 2). We have chosen the As-S-Se-I chalcogenide glasses as satisfying the abovementioned requirements. Chalcogenide glass immersion lenses multiply the radiation output coefficient several times (depending on their index of refraction and shape) and alter the radiation pattern by giving the lens a different shape. Low Ts is determined by the application temperature of the lens not exceeding 200 °C to avoid the p-n transition degradation.
168
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
2. Materials and methods 2.1. Chalcogenide glass characteristics
has been determined in order to calibrate the method. The deviation did not exceed 7% [11]. Liquid/solid work of adhesion has been derived from Young– Dupre equation:
Chalcogenide glass As12.8S24.2Se53I10, has been used for immersion. The glass characteristics are the following: Ts = 53 °C, for refractive index dispersion see Fig. 1. Glass softening point was measured with the use of BÄHR DIL802 difference-dilatometer. Refractive index has been measured by interference fringes method [8] using an FSM-1201 FTIR spectrometer.
W LS ¼ rglass ð1 þ cos hÞ;
2.2. Lens application
CG thermal expansion coefficient has been measured with the use of BÄHR DIL802 difference-dilatometer (temperature range 100 to 53 °C, fused silica inert sample, liquid nitrogen cooling). Mechanical properties have been determined through measurements of the longitudinal ultrasonic wave velocities (VL) and transverse ultrasonic wave velocities (VT) analogous to those described in [13].
The lenses were applied on LMS34 LEDs (with kovar bodies) by method of dispensing chalcogenide melt (CM), T = 150–180 °C, lens shape was adjusted by varying the mass of the drop. 2.3. Radiation patterns, radiation spectra and integral power Radiation patterns have been plotted based on the data obtained with the use of an Lms36PD-05 photodiode and a rotating table. Monochromator MDR-41 with a Judson J10D-M204-R01M-60 detector have been used for the recording of radiation spectra. THORLABS IS200 series integration sphere and a Judson J10DM204-R01M-60 detector have been used to measure integral power.
where rglass = 111 mJ/m2 – the surface tension of the used CM, cos h – the wetting angle cosine. The determined surface tension is in good agreement with that of an analogous CM given in [12]. 2.5. Thermal and mechanical properties
2.6. Temperature testing Experimental testing of LED stability against cracking has been performed with the use of BÄHR DIL802 difference-dilatometer (temperature range 150 to 53 °C, LED as the experimental sample, liquid nitrogen cooling). On course of the experiment, the LED was periodically switched on in order to examine its workability. After low-temperature testing visual inspection was performed with the use of a Leica EZ4 HD microscope.
2.4. Wetting angle and surface tension 2.7. Simulation Wetting angle has been determined by means of Drop Shape Analysis method [9]. The drop mass did not exceed 0.05 g. The height and radius of the drop have been measured with the use of Leica EZ4 HD microscope. CM surface tension has been determined by means of pendant drop tensiometry method [10]. Surface tension of sulfur at 150 °C
The increment of the integral power (DP) has been calculated with the use of the following formula:
DP ¼
PCG 100%; P0
where P0 is the integral power of the LED without a lens, PCG - the integral power of the LED with a lens. Type A lens possesses the minimal size sufficient to completely immerse the radiant crystal and LED’s upper electric contact (Fig. 4, left). Type B lens is the one that has shown maximum experimental increase of integral power (Fig. 5, left). LEDs with two lens shapes have been simulated in Zemax software. Numerical simulation of the radiation output from the crystal and radiation patterns has been performed in the ‘‘non-sequential mode”. A rectangular radiation source, simulating a radiative p-n junction, was put inside of the LED crystal (InAs) of the size 350 350 250 mm. The wavelength of the radiation source matched the wavelength for the LED’s radiation peak intensity (3.4 lm). The influence of the diode body material absorption has been omitted (‘‘mirror” type material option). 3. Results and discussion 3.1. LED radiation patterns, radiation spectra and current-voltage characteristic before and after immersion
Fig. 1. Refractive index dispersion.
LED images with and without a lens are given in Fig. 2. LED radiation patterns, radiation spectra and current-voltage characteristic before and after immersion are given in Fig. 3. LEDs with immersion lenses possess enhanced spectral power and narrowed radiation patterns, current-voltage characteristic do not change.
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
169
Fig. 2. LED without a lens (daylight, left); LED with a lens (810 ± 5 nm, right).
3.2. The influence of the refractive index on the increment of the integral power To perform simulation, initial experimental data (radiation patterns and integral power) for two types of lenses has been acquired. In order to justify the applicability of the utilized models, simulation results have been compared to the experimental radiation patterns and the increment of the integral power (Figs. 4 and 5). Simulation results are in good agreement with experimental ones, the difference being motivated by the approximation of lens shapes with simple geometric shapes. Real lens shapes possess more complicated geometry. In order to evaluate the influence of the refractive index on the increment of the integral power we have simulated LEDs with type A and B lenses having a set of different indices of refraction (see Fig. 6). The non-linear dependences are motivated by the opposite contribution of two factors: (1) The increase of the angle of total internal reflection on the crystal/CG. (2) The decrease of the angle of total internal reflection on the CG/air interface. The optimal refractive index for the immersion lenses is between 2.6 and 2.8. Contrary to monocrystalline lenses of fixed refractive index values, there is an opportunity to alter the CG refractive index. 3.3. Adhesion force of CG melt towards electronic engineering materials
Fig. 3. Before and after immersion: radiation patterns (top), radiation spectra (middle), current-voltage characteristic (bottom).
One of the crucial properties of the appropriate for the production of As-S-Se-I immersion lenses CM is its wettability towards diode body and the semiconductor crystal. In case of low wettability (h < 90 °C) the CM incompletely covers the upper diode platform and, hence, the lens becomes asymmetrical. Furthermore, for type B lenses, low wettability can cause the CM flow off the upper diode platform. In [14] it has been shown that liquid/solid work of adhesion (WLS) is proportional to the corresponding solid/solid adhesion strength (WSS). It is expected that at high values of work of adhesion of the CM towards the diode body material the junction between the immersion lens and the adjacent diode body material will be more secure. We have determined the work of adhesion of CM towards several electronic engineering materials in order to qualitatively evaluate the relative difference in the corresponding adhesion strength of CG towards said materials.
170
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
Fig. 4. Simulated type A lens shape (left); radiation patterns and the increment of the integral power: experiment and calculation (right).
Fig. 5. Simulated type B lens shape (left); radiation patterns and the increment of the integral power: experiment and calculation (right).
Wetting angle and WLS of electronic engineering materials are given in Table 1. The highest wettability and maximal work of adhesion is observed for CM/kovar and CM/InAs which are used for the production of the diode body and the radiant crystal, respectively. Based on [14], it can be proposed that among abovementioned materials the adhesion strength is maximal for CG/kovar and CG/InAs. As expected, the lowest wettability has been proved to be for CM/fluoroplastic, hence, the latter can be used as a mold for lens production of desired geometry. 3.4. Stability against cracking
Fig. 6. Simulated LED integral power as a function of refractive index for type A and B lens shapes.
Table 1 Wetting angle and work of adhesion. No
Material
cos h
h, degrees
WLS, mJ/m2 (180 °C)
1 2 3 4 5 6 7 8 9 10 11 12
Kovar InAs TiAlN FR4 textolite ZrN NbN Amorphous graphite CrMoN Fused silica TiN Stainless steel Fluoroplastic
0.37 0.31 0.20 0.20 0.16 0.15 0.15 0.14 0.13 0.11 0.07 0.20
68.3 ± 0.9 71.5 ± 1.1 78.2 ± 0.4 78.4 ± 0.5 80.8 ± 0.6 81.2 ± 0.6 81.3 ± 0.1 81.9 ± 0.4 82.6 ± 0.3 83.7 ± 1.1 85.9 ± 0.7 100.9 ± 0.6
152 145 133 133 129 128 128 127 125 123 119 89
After application, the chalcogenide lens becomes an integral part of the LED and is bound to the diode body via adhesion force. Due to the fact that the CG and the adjacent materials possess different thermal expansion coefficients, mechanical stresses emerge within the CG/diode junction. The numerical value of those stresses depend both on the mechanical and thermal properties of the materials and the temperature range of exploiting. Upon reaching critical stresses, there is a possibility of lens cracking or separation from the diode body. We have calculated mechanical stresses within the junction between CG and LED body materials based on their mechanical and thermal properties. Stability against cracking has been experimentally tested for the 150 to 53 °C temperature range. In order to evaluate mechanical stresses in the junction between CG and diode body material (Fig. 7), we have used a method devised by [15]. Mechanical stresses have been calculated in an approximation of instant hardening (absence of relaxation) with the use of the formula:
r¼
EG ðaCG aÞ DT; 1 lCG ECG dCG þ 1 Ed
where r - mechanical stress, ECG – Young’s modulus, lCG – Poisson ratio, aCG – thermal expansion coefficient, dCG – density;
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
171
Fig. 8. Stability against cracking experimental apparatus. 1 – Pushrods, 2 – inert sample (fused silica), 3 – sample holder back panel, 4 – LED, 5 – CG, 6 – LED driver.
Fig. 7. CG dilatometer curve (top) and mechanical stresses (bottom).
Table 2 Diode body materials mechanical properties.
a
Fig. 9. Experimental change in length of the LED.
Material
a 106, 1/R
E, GPa
d, g/cm3
Source
Kovar Stainless steela Aluminiuma Fused silicaa
5 16.6 23.8 0,6
196 212 62,1 75
8.35 7.90 2.70 2.20
[16] [15] [15] [15]
For comparison.
Table 3 CG mechanical properties. VL, m/c
VT, m/c
lCG
ECG, GPa
GCG, GPa
dCG, g/cm3
a 106, 1/R
1860 ± 10
940 ± 10
0.33
9.1 ± 0.2
3.5 ± 0.1
3.86
Fig. 7 top
E, l, a, d – the corresponding properties of the diode body material, DT = (Ts T), Ts = 53 °C (see Tables 2 and 3). According to [15,17,18], chalcogenide glasses can withstand mechanical stresses up to approximately 100 MPa. It can be seen in Fig. 7 that mechanical stresses in the junction between CG and any of the diode body materials do not reach critical value in the 100 °C to 53° temperature range. In order to confirm theoretical calculations, experimental testing of LEDs has been performed in the 150 to 53 °C temperature range. The apparatus is described in Fig. 8. Experimental change in length of the LED (DL) is given in Fig. 9. Jumps of DL have not been registered upon cooling or subsequent heating to ambient temperature (see Fig. 9), LED remained workable on course of the whole experiment. This indicates that the electrical contacts stayed undamaged. Visual inspection has not shown lens defects.
4. Conclusion The application of As-S-Se-I melt upon a LED results in the formation of an immersion CG lens (Ts = 53 °C) due to surface tension force. LED damage does not occur due to low (less than 200 °C) application temperature. Lens shape can be adjusted by varying the mass of the applied melt. LEDs with immersion lenses possess enhanced spectral power and narrowed radiation patterns. A maximum increase of integral power (about ten times) has been achieved for type B lenses. LEDs with two lens shapes have been simulated in Zemax software. The good agreement of simulation results with experimental ones justify the applicability of the utilized models. The influence of the refractive index on the integral power has been simulated. It has been determined that the optimal refractive index for the immersion lenses, analogous to those described in the present work, is between 2.6 and 2.8 (k = 3.4 lm). The wettability of As-S-Se-I melt has been determined for several diode body materials. The highest wettability among abovementioned materials has been proved to be for kovar, while the lowest – for fluoroplastic, hence, the latter can be used as a mold for lens production of desired geometry. Mechanical stresses between CG and adjacent diode body materials (e.g. quartz glass, kovar, stainless steel, aluminium) have been calculated. It has been calculated that for 100 to 53 °C temperature range critical stress (100 MPa) is not reached with any of the materials. The calculation has been justified by the
172
V.A. Markov et al. / Infrared Physics & Technology 78 (2016) 167–172
experimental data (the LED does not crack in 150 to 53 °C temperature range). Acknowledgement The research was partially supported by FASIE. We kindly acknowledge Alfred B. Sinani for the determination of glass mechanical properties by ultrasonic method. References [1] A. Krier (Ed.), Mid-Infrared Semiconductor Optoelectronics, Springer, London, 2006, http://dx.doi.org/10.1007/1-84628-209-8. [2] B.E. Zhurtanov, N.D. Il’inskaya, A.N. Imenkov, M.P. Mikhaĭlova, K.V. Kalinina, M. A. Sipovskaya, et al., Low-noise photodiodes based on GaSb/GaInAsSb/ AlGaAsSb double heterostructures for the 1–4.8 lm spectral range, Semiconductors 42 (2008) 458–462, http://dx.doi.org/10.1134/ S1063782608040155. [3] N.V. Zotova, N.D. Il’inskaya, S.A. Karandashev, B.A. Matveev, M.A. Remennyi, N. M. Stus’, Sources of spontaneous emission based on indium arsenide, Semiconductors 42 (2008) 625–641, http://dx.doi.org/10.1134/ S1063782608060018. [4] A.P. Astakhova, A.S. Golovin, N.D. Il’inskaya, K.V. Kalinina, S.S. Kizhayev, O.Y. Serebrennikova, et al., High-power InAs/InAsSbP heterostructure leds for methane spectroscopy (k 3.3 lm), Semiconductors 44 (2010) 263–268, http://dx.doi.org/10.1134/S1063782610020235. [5] N.D. Stoyanov, K.M. Salikhov, K.V. Kalinina, S.S. Kizhaev, A.V. Chernyaev, Super low power consumption middle infrared LED-PD optopairs for chemical sensing, in: M.J.F. Digonnet, S. Jiang (Eds.), Proc. SPIE – Int. Soc. Opt. Eng., SPIE, 2014, p. 89821A, http://dx.doi.org/10.1117/12.2036277. [6] P.N. Brunkov, N.D. Il’inskaya, S.A. Karandashev, N.M. Latnikova, A.A. Lavrov, B. A. Matveev, et al., P-InAsSbP/n 0-InAs/n +-InAs photodiodes for operation at moderate cooling (150–220 K), Semiconductors 48 (2014) 1359–1362, http:// dx.doi.org/10.1134/S1063782614100066.
[7] P.N. Brunkov, N.D. Il’inskaya, S.A. Karandashev, A.A. Lavrov, B.A. Matveev, M.A. Remennyi, et al., InAsSbP/InAs0.9Sb0.1/InAs DH photodiodes (k0.1 = 5.2 lm, 300 K) operating in the 77–353 R temperature range, Infrared Phys. Technol. 73 (2015) 232–237, http://dx.doi.org/10.1016/j.infrared.2015.09.017. [8] F. Carreño, J.C. Martínez-Antón, E. Bernabeu, Optical interference method to obtain thickness and refractive indices of a uniaxial medium, Rev. Sci. Instrum. 65 (1994) 2489, http://dx.doi.org/10.1063/1.1144707. [9] Y. Yuan, T.R. Lee, Contact angle and wetting properties, 2013, pp. 3–34. http:// dx.doi.org/10.1007/978-3-642-34243-1_1. [10] J.D. Berry, M.J. Neeson, R.R. Dagastine, D.Y.C. Chan, R.F. Tabor, Measurement of surface and interfacial tension using pendant drop tensiometry, J. Colloid Interf. Sci. 454 (2015) 226–237, http://dx.doi.org/10.1016/j.jcis.2015.05.012. [11] D.L. Timrot, S.A. Traktueva, B.A. Alekseev, Surface tension of liquid sulfur, High Temp. 21 (1983) 671–676.