Open Access
Add to BibSonomy
Abstract
Semiconductors are promising candidates as fiber materials for transmission in the mid-infrared (IR) region. This paper reports the lowest measured losses for a Ge-core, borosilicate cladded optical fiber produced by the ‘rod-in-tube’ method that is scalable for manufacturability. The fibers were drawn in a mini draw tower at 1000°C. The relatively low drawing temperature and the optimal location of the Ge core in the preform minimize the time the molten Ge interacted with the borosilicate cladding limited diffusion from the cladding to the core. The coefficient of thermal expansion (CTE) match between the core and cladding led to fibers that were crack- and pore- free. Transmission electron microscopy studies showed the Ge core to have good crystalline quality with low impurity content. The optical properties of the fibers were characterized in the mid-IR region with a quantum cascade laser and transmission losses were found to be in the 3.1-9.1 dB/cm range in the spectral range of 5.82-6.28 μm, with an average loss value of 5.1 dB/cm.
© 2017 Optical Society of America
1. Introduction
In the last decades, significant progress has been made in lasers [1], detectors [2] and fibers [3,4] operating at mid-IR wavelengths. Fibers that are capable of low loss transmission in the mid-IR spectral region play an important role in light delivery for a variety of applications including biomedical surgery [5], chemical sensing [6] and defense systems [7], as well as for more advanced functionalities enabled by nonlinear optics in this wavelength range [8].
Silica, the ubiquitous material used for transmission of near-IR wavelengths, has enormous losses in the mid-IR region [9]. This has led to investigations of suitable materials for mid-IR transmission, and several material systems have been studied, including halide crystals [10], fluoride glasses [11] and chalcogenide glasses [4]. Although most of these materials have low transmission losses at mid-IR wavelengths, they have not become the preferred solution due to issues such as property degradation over time [7], even at room temperature.
Semiconductors, such as silicon, germanium, zinc selenide, and indium-antimonide, have low transmission losses at mid-IR wavelengths. They also have crystalline structure that, in theory, can lead to lower scattering losses and have stable mechanical and optical properties at room temperature [12,13]. This has led to several studies on semiconductor-core fibers [14–17]. Of particular interest are Ge-core fibers, since Ge is transparent all the up to 14 μm [18], thus spanning both the Mid Wave Infrared (MWIR) and Long Wave Infrared (LWIR) regions of the electromagnetic spectrum [19,20].
Two techniques have been used to fabricate Ge-core, glass-clad fibers. The ‘rod in tube’ method involves conventional drawing of a preform consisting of a Ge rod placed in a silica tube, and is amenable to producing long fiber lengths [21]. The high-pressure chemical vapor deposition (HP-CVD) method involves CVD deposition in pre-drawn glass tubes [22]. Although this technique has successfully produced amorphous and crystalline Ge-core fibers that transmit light in the mid-IR spectrum [12,22,23], it is not scalable to a manufacturing setting due to the limitation of the length of fiber that can be fabricated.
Two choices of the glass used as cladding for Ge-core fibers fabricated by the rod-in-tube technique have been explored; borosilicate and silica. Ballato and associates have fabricated silica-cladded Ge-core canes (300 µm core diameter) using the rod-in-tube method. They reported mid-IR transmission in these canes due to low oxygen impurity content in the Ge-core, which was attributed to volatilization of GeO2 at the high drawing temperature of 1925°C [21]. However, scaling these dimensions down to those of typically flexible fibers that transmitted mid-IR light was not achieved. Moreover, these canes exhibited significant cracking due to CTE mismatches between the core and the cladding [21]. Gibson and associates have addressed the cracking issue in silicon-core, silica cladded fibers by introducing an intermediate layer of CaO that accommodated the thermal strain by forming a fine-structured eutectic at the Si-core/silica-cladding interface, while also acting as an oxygen getter [24]. Whether similar techniques may be used with Ge-core fibers is not yet known. Moreover, it is unclear how the addition of this extra processing step impacts the manufacturability of the fabrication process. Ballato and associates have also fabricated Ge-core fibers with borosilicate glass cladding by the rod-in-tube method [25]. The Raman scattering response from these fibers by Wang et al. indicated that the core is crystalline Ge [26]. However, no transmission loss measurements were reported through the 4-mm samples that were characterized. The fibers were not successful in transmitting light in the mid-IR spectrum due to a reported 4 wt% oxygen impurity content in the core [25].
Thus, to the best of our knowledge, there has been no reported mid-IR transmission with measurement of transmission losses in crack-free Ge-core, borosilicate glass cladded fibers fabricated by the rod-in-tube method which is scalable to commercial fiber manufacturing without the added intermediate layer needed for silica glass cladding. In this paper, we report on the fabrication of borosilicate glass-clad, crystalline Ge core fibers that guide mid-infrared light. This was achieved by simulating and experimentally verifying the location of the neck formation in the preform and placing the Ge rod slightly above this location to minimize the interaction time between the molten Ge core and the borosilicate cladding before fiber drawing occurred. Loss measurements were conducted with a Quantum Cascade Laser (QCL) operating in the 6-μm wavelength range, of fundamental importance to the studies of proteins, nucleic acids, and phospholipids, as well as to the study polymer composites [27–29]. The record low loss (~5 dB/cm) we report validates our fabrication technique as a promising manufacturable pathway for the realization of Ge-core optical fibers.
2. Experimental
Fabrication by the ‘rod in tube’ method [30] involves core drilling a 10 mm long and 3 mm outer diameter (OD) Ge rod from a 10 × 10 × 50 mm3 single crystal Ge sample of 99.999% purity (Lattice Materials LLC, Bozeman, MT). The Ge rod is placed inside a borosilicate glass tube with 3 mm inner diameter (ID) and 9 mm OD, and sealed on both sides with 3 mm OD borosilicate glass rods of the same composition to prevent oxidation. Additional borosilicate glass tubes with increasing diameters were added concentrically as needed to adjust the area ratio of the core and cladding. This area ratio is a parameter that is preserved during the fiber drawing process. Duran (Schott Glass, Germany) borosilicate glass (13 wt% B2O3, 4 wt% Na2O + K2O, 2 wt% Al2O3, balance – SiO2) was chosen for its CTE match with Ge (CTE of borosilicate glass is 3.3 × 10−6/K and CTE of Ge is 5.65 × 10−6/K) [31,32]. Also, borosilicate glass has a low softening point (826°C) that allows low temperature drawing, and potentially reduces impurity diffusion from the glass cladding into the core.
Fibers were drawn in a mini draw tower designed and assembled in-house at Boston University. The drawing temperature was chosen to be 1000°C, which is slightly above the softening point of the borosilicate glass and the melting temperature of the Ge core (938°C). In order to minimize the time the molten core interacts with the borosilicate glass, a cold preform was inserted in a furnace that was already at temperature, since the time for the preform to heat up to temperature was significantly smaller than the time required to ramp a cold furnace to the drawing temperature. In order to locate the position where necking starts, simulations of the preform deformation was carried out. The temperature dependent viscosity of the borosilicate glass [33] was used in the simulation, and a Gaussian temperature profile was assumed for the furnace, based on the measured temperature profile. Intuitively, the location of the hottest temperature in the furnace corresponds to the softest glass. However, the stresses at the top of the preform is the highest, since it has the most weight of the preform below it pulling down due to gravitational forces. Thus, the likely location of necking initiation is expected to be somewhere in between the location of maximum temperature and maximum stress, i.e., at a point above the location of maximum temperature in the furnace. The simulations were carried out by coupling a heat transfer model using COMSOL Multiphysics® to a mechanical deformation model using MATLAB®, and good agreement was found between simulated and experimental values of both the location and time of neck formation. These simulations allowed the Ge-core to be placed slightly above the region of neck formation to minimize the interaction of the molten core with the borosilicate cladding at temperature. Typically, tens of meters of fibers were drawn from each preform.
After drawing, the fibers were placed in epoxy, and the ends were polished to observe the fibers in cross section in an optical microscope (Nikon Eclipse LV 150, Japan). In order to study the diffusion profile at a high spatial resolution, electron transparent transmission electron microscopy (TEM) samples containing the fiber core/cladding interface were prepared using a focused ion beam (FIB; FEI, Oregon) based ‘lift-out’ technique. Elemental dot maps and composition profiles across the core-cladding interface were obtained by scanning/transmission electron microscopy (S/TEM; FEI, Oregon) and energy dispersive x-ray (EDX) spectroscopy at 200 kV, using a 1 nm diameter electron beam. Electron transparent samples were also made by standard polishing and ion milling techniques for high-resolution TEM (HRTEM) studies of the interface. The crystallinity of the samples was also investigated by x-ray diffraction (XRD; Bruker, Wisconsin), by reciprocal space mapping (RSM) with a 1 mm diameter x-ray beam, using a 2-D detector (Vantec 500, Bruker, Massachusetts). Optical transmission measurements were carried out in the mid-IR spectrum by Fourier transform infrared spectroscopy (FTIR; Bruker, Massachusetts) and by using a QCL (Daylight Solutions, California).
3. Results and discussion
The microstructure (crystallinity and diffusion profile between core and cladding) and the optical (mid-IR transmission) properties of the fibers have been characterized, as described below.
3.1 Microstructural characterization
A finely polished facet of drawn fiber with a 40µm Ge core diameter is shown in Fig. 1. The fiber is embedded into hardened epoxy in order to polish the facets at both ends. Figure 2(a) shows a TEM bright-field micrograph of the Ge-core/borosilicate glass-clad interface of fiber, along with EDX elemental dot maps of Ge, O and Si (Fig. 2(b)-2(d)), indicating that the glass/cladding interface is quite sharp. Figure 2(e) shows an EDX line-scan across a core/cladding interface region. Figure 2(e) shows that the diffusion of the cladding components (Si, O, B, K, Na and Al) diffusion from cladding to core is minimal with the diffusion distances in the submicron regime. The low oxygen content of the core is likely due to the low drawing temperature and is encouraging for mid-IR transmission since oxygen is known to increase mid-IR transmission losses in fibers.
Fig. 1 Optical micrograph of a polished fiber cross-section. The bright center is the Ge core, surrounded by the borosilicate glass cladding. The most outer part is hardened epoxy, into which the fiber was embedded to improve facet quality during polishing.
Fig. 2 (a) TEM bright-field micrograph of core/cladding interface of fiber. EDX dot maps of (b) Ge (c) O and (d) Si. (e) EDX line scan showing the composition profile of Si, Ge, O, B, K, Na, and Al across the core/cladding interface. The diffusion of oxygen from the cladding to the core is minimal.
Figure 3(a) shows an HRTEM micrograph of core-cladding interface in the fiber. The crystalline quality of the Ge core is evident in the micrograph. The crystallinity of the Ge core was further investigated by RSM with a 1 mm x-ray beam, using a 2D detector. The scan, shown in Fig. 3(b) shows a single (111) peak orientation, indicating that the region interrogated by the 1 mm beam is a single crystal. The 2θ value of the peak location (27.22) indicated that the Ge core is in slight tension (0.31%) compared to unstrained Ge. This is consistent with a recent report of slight red shifting of the T2g Raman mode frequency from a Ge fiber core deposited by high-pressure CVD compared to an unstrained Ge reference [23].
Fig. 3 (a) High resolution TEM micrograph of the core and cladding interface, showing the crystalline nature of the Ge core. (b) Reciprocal space mapping of the Ge core with 1 mm X-ray beam and a 2D detector, showing a single (111) peak.
3.2 Optical characterization
Cleaving of the Ge-core optical fibers did not lead to good quality facets due to cracking of the core during the process. Thus, it was necessary to mount the fibers in epoxy, and polish both ends to get good quality facets for optical measurements. This limited the size of the fibers lengths for optical measurements to ~10 mm. Transmittance through Ge-core fibers were measured using a QCL, as described in Section 3.2.2.
3.2.1 Transmittance measurements by FTIR
In order to characterize the mid-IR transmittance through a Ge core/borosilicate cladded fiber imbedded in hardened epoxy, it is important to characterize the transmittance through all the materials in the sample, before transmission can be exclusively attributed to the Ge-core. To this end, a borosilicate glass rod was embedded in hardened epoxy (~10 mm thick), and both ends were polished to a fine finish, exposing the cross-section of the borosilicate glass rod at both ends. The sample was then characterized by FTIR in the range of 1.7-10 μm wavelengths by focusing the beam on the glass rod as well as on the hardened epoxy. Figure 4 shows the transmission results of borosilicate glass and hardened epoxy. The borosilicate glass stops transmitting at wavelength above 3.3 μm whereas the epoxy stops doing so at a wavelength of 2.1 μm. This dictates that any transmission through a Ge-core fiber at the spectral range of QCL (5.82-6.28 μm) has to be through the Ge-core.
Fig. 4 Wavelength dependent FTIR spectroscopy results showing lack of mid-IR transmission through borosilicate glass and hardened epoxy.
3.2.2 Transmittance measurements in fiber by QCL
The QCL optical characterization setup (Fig. 5) consists of a 6 mm focal length ZnSe objective for the input coupling, two 25.4 mm focal length ZnSe convex lenses for collecting the fiber output, and gold-coated mirrors, configured to obtain a flat-phase front Gaussian beam of 13-µm diameter at the input fiber facet, ensuring that the entire beam is coupled into the fiber under test. Transmission characteristics of the fibers were measured with the QCL in the range of 5.82-6.28 μm (1592-1718 cm−1). The input power was set at ~3.9 mW at the peak emission wavelength, which in conjunction with the glass absorption filter ensured that the output power from the fiber did not saturate the mercury cadmium telluride (MCT) detector. The transmitted signals were read by a lock-in amplifier phase-locked to the 100 kHz repetition rate of the QCL.
Figure 6(a) shows the horizontal and vertical scans across the fiber core. For transmission loss measurements, the input beam was positioned at the maximum intensity location in Fig. 6(a). An 8.63 mm long fiber sample was used for the initial measurement, after which the sample was polished down to 8.25 mm and 7.91 mm for additional measurements. Figure 6(b) shows the un-normalized output signals from the three lengths of the sample as well as the un-normalized QCL output signal measured without any sample in the light path. In all these measurements, the same filter was used to ensure that the detector did not saturate. The sharp dips in the output signals were correlated primarily to known rotationally narrowed absorption peaks in water vapor (as listed in the HITRAN Database [34]) in the free space portions of the beam path. For each un-normalized QLC output signal (Fig. 6(b)), a reduced data set was created by only choosing points away from the absorption peaks. These reduced data sets were fit to a 4th order polynomial. Figure 6(c) shows the fit for the 8.63 mm fiber data.
Fig. 6 (a) Horizontal and vertical scans of fiber core with QCL. (b) Transmission characteristics in the QCL range of Ge fibers of three different lengths of 8.63 mm, 8.25 mm, and 7.91 mm, as well as the laser output signal with no sample in the path of the QCL. The sharp dips are due to atmospheric absorption peaks. (c) Best fit of a 4th order polynomial to the transmission spectrum of the 8.63 mm fiber, using a reduced data set of points away from the atmospheric absorption peaks.
The differences between the fiber lengths being small, the transmission loss was calculated by comparing the un-normalized output signal of each individual length to the detector signal without the sample in the path of the laser. The loss can be expressed by
where L is the length of the fiber, Pfiber and Pbackground are fiber output power and background power respectively. The three measurements led to three sets of wavelength-dependent loss data, which was found to range from 3.1 to 9.1 dB/cm. Since the wavelength range of the measurement is much too narrow to deduce any meaningful trends with respect to wavelength dependence, an average of all the calculated loss values (over a wavelength range of 5.82-6.28 μm for fiber lengths of 8.63 mm, 8.25 mm and 7.91 mm) was computed to be 5.1 dB/cm. The aforementioned measurements accounted for Fresnel reflections at the input and output of the fiber (calculated to be 34% from the refractive index discontinuity between Ge and air), and all the remaining losses were attributed to the fiber. This implies that the calculated loss values are an upper bound of the actual losses in the Ge-fiber core; hence, the actual losses of our fibers may be even lower than the record value we report here.Measurement of the transmission loss in a 99.999% pure, 3 mm diameter single crystal Ge rod using the same technique revealed a loss of 0.4 dB/cm at a wavelength of 6.1 µm. Obviously, the higher loss in the Ge-core of the fiber is the result of the fiber drawing process, and can be attributed to one or more of the following effects; i) diffusion of impurities from cladding to the core, ii) evanescent light fields that penetrate the highly absorbing borosilicate cladding layer, and iii) formation of grain boundaries in the core.
4. Conclusions
Ge-core borosilicate glass-clad semiconductor fibers were fabricated for mid-IR transmission. The fibers were drawn by the ‘rod in tube’ method at 1000°C, which is much lower than the drawing temperature of silica fibers. S/TEM and EDX studies showed that minimal impurity content, likely a result of the low drawing temperature. Transmission measurements were carried out using an FTIR over a 1.7-10 µm range, and a QCL over a 5.82-6.28 μm range. Fiber losses were calculated to be in the 3.1-9.1 dB/cm range in the range of the QCL wavelengths, with an average loss value of 5.1 dB/cm. These are the lowest reported loss values for a Ge-core, borosilicate cladded optical fiber produced by the ‘rod-in-tube’ method that is scalable to manufacturing settings.
Funding
National Science Foundation (NSF) (Grant Number CMMI-1301108, 2013); AFOSR (Grant Number FA9550-14-1-0165).
Acknowledgements
The authors would like to thank Dr. H. Durmaz for discussion on FTIR, Prof. M. Sander and A. Totachawattana for advice on QCL measurements, and Dr. A. Nikiforov and Dr. Y. Yu for assistance with the TEM studies.
References and links
1. J. Faist, F. Capasso, D. L. Sivco, C. Sirtori, A. L. Hutchinson, and A. Y. Cho, “Quantum cascade laser,” Science 264(5158), 553–556 (1994). [CrossRef] [PubMed]
2. B. F. Levine, K. K. Choi, C. G. Bethea, J. Walker, and R. J. Malik, “New 10 μm infrared detector using intersubband absorption in resonant tunneling GaAlAs superlattices,” Appl. Phys. Lett. 50(16), 1092–1094 (1987). [CrossRef]
3. T. Miya, Y. Terunuma, T. Hosaka, and T. Miyashita, “Ultimate low-loss single-mode fibre at 1.55μm,” Electron. Lett. 15(4), 106–108 (1979). [CrossRef]
4. T. Kanamori, Y. Terunuma, S. Takahashi, and T. Miyashita, “Chalcogenide glass fibers for mid-infrared transmission,” J. Lightwave Technol. 2(5), 607–613 (1984). [CrossRef]
5. J. S. Sanghera, L. B. Shaw, L. E. Busse, V. Q. Nguyen, P. C. Pureza, B. C. Cole, B. B. Harrison, I. D. Aggarwal, R. Mossadegh, F. Kung, and D. Talley, “Development and infrared applications of chalcogenide glass optical fibers,” Fiber Integr. Opt. 19(3), 251–274 (2000).
6. J. Heo, M. Rodrigues, S. J. Saggese, and G. H. Sigel Jr., “Remote fiber-optic chemical sensing using evanescent-wave interactions in chalcogenide glass fibers,” Appl. Opt. 30(27), 3944–3951 (1991). [CrossRef] [PubMed]
7. J. A. Harrington, Infrared fibers and their applications (SPIE, 2003).
8. D. D. Hudson, S. A. Dekker, E. C. Mägi, A. C. Judge, S. D. Jackson, E. Li, J. S. Sanghera, L. B. Shaw, I. D. Aggarwal, and B. J. Eggleton, “Octave spanning supercontinuum in an As2S3 taper using ultralow pump pulse energy,” Opt. Lett. 36(7), 1122–1124 (2011). [CrossRef] [PubMed]
9. J. M. Senior and M. Y. Jamro, Optical Fiber Communications: Principles and Practice (Pearson Education, 2009).
10. A. Sa’ar, F. Moser, S. Akselrod, and A. Katzir, “Infrared optical properties of polycrystalline silver halide fibers,” Appl. Phys. Lett. 49(6), 305–307 (1986). [CrossRef]
11. D. Szebesta, S. T. Davey, J. R. Williams, and M. W. Moore, “OH absorption in the low loss window of ZBLAN (P) glass fibre,” J. of non-Crys. Solids 161, 18–22 (1993).
12. A. C. Peacock, J. R. Sparks, and N. Healy, “Semiconductor optical fibers: progress and opportunities,” Laser Photonics Rev. 8(1), 53–72 (2014). [CrossRef]
13. F. H. Suhailin, L. Shen, N. Healy, L. Xiao, M. Jones, T. Hawkins, J. Ballato, U. J. Gibson, and A. C. Peacock, “Tapered polysilicon core fibers for nonlinear photonics,” Opt. Lett. 41(7), 1360–1363 (2016). [CrossRef] [PubMed]
14. R. He, P. J. A. Sazio, A. C. Peacock, N. Healy, J. R. Sparks, M. Krishnamurthi, V. Gopalan, and J. V. Badding, “Integration of gigahertz-bandwidth semiconductor devices inside microstructured optical fibres,” Nat. Photonics 6(3), 174–179 (2012). [CrossRef]
15. J. Ballato, T. Hawkins, P. Foy, C. McMillen, L. Burka, J. Reppert, R. Podila, A. M. Rao, and R. R. Rice, “Binary III-V semiconductor core optical fiber,” Opt. Express 18(5), 4972–4979 (2010). [CrossRef] [PubMed]
16. M. Ordu, J. Guo, S. Erramilli, S. Ramachandran, and S. Basu, “Germanium-core borosilicate glass-cladded semiconductor fibers for mid-IR applications,” in CLEO: Science and Innovation, OSA (2016), paper STu4R–1.
17. M. Ordu, J. Guo, B. Tai, J. Bird, S. Ramachandran, and S. Basu, “Processing and optical properties of Ge-core fibers,” in Ceramic Materials for Energy Applications VI: Ceramic Engineering and Science Proceedings Vol. 37(6), (2017), 85.
18. C. Claeys and E. Simoen, Germanium-Based Technologies: From Materials to Devices (Elsevier, 2007).
19. D. Harris, “Durable 3-5 μm transmitting infrared window materials,” Infrared Phys. Technol. 39(4), 185–201 (1998). [CrossRef]
20. R. Soref, “Mid-infrared photonics in silicon and germanium,” Nat. Photonics 4(8), 495–497 (2010). [CrossRef]
21. J. Ballato, T. Hawkins, P. Foy, S. Morris, N. K. Hon, B. Jalali, and R. Rice, “Silica-clad crystalline germanium core optical fibers,” Opt. Lett. 36(5), 687–688 (2011). [CrossRef] [PubMed]
22. P. Mehta, M. Krishnamurthi, N. Healy, N. F. Baril, J. R. Sparks, P. J. A. Sazio, V. Gopalan, J. V. Badding, and A. C. Peacock, “Mid-infrared transmission properties of amorphous germanium optical fibers,” Appl. Phys. Lett. 97(7), 071117 (2010). [CrossRef]
23. X. Ji, R. L. Page, S. Chaudhuri, W. Liu, S. Yu, S. E. Mohney, J. V. Badding, and V. Gopalan, “Single-Crystal Germanium Core Optoelectronic Fibers,” Adv. Opt. Mater. 5(1), 1600592 (2017). [CrossRef]
24. E. F. Nordstrand, A. N. Dibbs, A. J. Erakar, and U. J. Gibson, “Alkaline Oxide Interface Modifiers for Silicon Fiber Production,” Opt. Mater. Express 3(5), 651–657 (2013). [CrossRef]
25. J. Ballato, T. Hawkins, P. Foy, B. Yazgan-Kokuoz, R. Stolen, C. McMillen, N. K. Hon, B. Jalali, and R. Rice, “Glass-clad single-crystal germanium optical fiber,” Opt. Express 17(10), 8029–8035 (2009). [CrossRef] [PubMed]
26. P. Wang, C. Charlton O’Mahony, T. Lee, R. Ismaeel, T. Hawkins, Y. Semenova, L. Bo, Q. Wu, C. McDonagh, G. Farrell, J. Ballato, and G. Brambilla, “Mid-infrared Raman sources using spontaneous Raman scattering in germanium core optical fibers,” Appl. Phys. Lett. 102(1), 011111 (2013). [CrossRef]
27. P. R. Griffiths and J. A. de Haseth, Fourier Transform Infrared Spectrometry, 2nd Ed. (Wiley, 2007).
28. R. Bhargava, “Infrared Spectroscopic Imaging: The Next Generation,” Appl. Spectrosc. 66(10), 1091–1120 (2012). [CrossRef] [PubMed]
29. R. Adato, A. A. Yanik, J. J. Amsden, D. L. Kaplan, F. G. Omenetto, M. K. Hong, S. Erramilli, and H. Altug, “Ultra-Sensitive Vibrational Spectroscopy of Protein Monolayers with Plasmonic Nanoantenna Arrays,” Proc. Natl. Acad. Sci. U.S.A. 106(46), 19227–19232 (2009). [CrossRef] [PubMed]
30. J. M. Ballato and R. R. Rice, US Patent 9,069,117 (30 June 2015).
31. Schott, Duran Borosilicate Glass Catalog (2014).
32. M. E. Straumanis and E. Z. Aka, “Lattice parameters, coefficients of thermal expansion, and atomic weights of purest silicon and germanium,” J. Appl. Phys. 23(3), 330–334 (1952). [CrossRef]
33. A. Fluegel, “Glass viscosity calculation based on a global statistical modeling approach,” Glass Technology-European J. of Glass Science and Technol. Part A 48(1), 13–30 (2007).
34. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. Chris Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, E. A. Cohen, L. H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A. Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, and G. Wagner, “The HITRAN2012 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transf. 130, 4–50 (2013). [CrossRef]
