JLT-20479-2017

Abstract— A few-moded silica-based optical fiber fabricated from core materials that possess intrinsically low optical nonlinearities is reported. Specifically, the 8 µm core, 125 µm cladding diameter silicate fiber was composed of a strontium aluminosilicate oxyfluoride core with a fused silica cladding and was fabricated using the molten core method. Relative to conventional optical fibers, reductions of ∼6.3 dB in Brillouin gain coefficient (gB), ∼0.9 dB in Raman gain coefficient (gR), and ∼2.2 dB in thermo-optic coefficient (TOC) were realized as was a “silica-like” nonlinear refractive index (n2) with a value of ∼3x10-20 m2/W. The role of each core material constituent on parameters that drive optical nonlinearities is discussed to provide a materials solution route for low nonlinearity fiber systems. Materially addressing optical nonlinearities represents a simpler and more effective approach to mitigating power-scaling limits in high energy fiber laser systems compared to the geometric approaches employed using microstructured fibers.

Index Terms—High energy lasers, optical fiber, stimulated Brillouin scattering, stimulated Raman scattering, thermo-optic coefficient, nonlinear refractive index

I. INTRODUCTION

PTICAL nonlinearities serve as limitations to continued optical power scaling in high energy fiber-based laser

systems. Paramount amongst these parasitic phenomena are stimulated Brillouin scattering (SBS), stimulated Raman scattering (SRS), and nonlinear refractive index, n2, related wave-mixing phenomena (e.g., four-wave mixing, FWM, and self-phase modulation, SPM) [1]. In order to combat these performance limitations, fiber designers have developed ever-more complex structures that aim to spread the propagating mode out over a larger cross-sectional area so as to reduce the power density in the fiber below the threshold power for each parasitic effect. Such “large mode area” (LMA) fibers are often designed to behave as “effectively

O

Manuscript received March 23, 2017. This work was supported in part by the US Department of Defense Joint Technology Office through contracts W911NF-05-1-0517, FA9550-07-1-0566, W911NF-12-1-0602, FA9451-15-D-0009/0001 and FA9451-15-D-0009/0002 (JB and PD) as well as by the Sirrine Foundation (MC, CJK, TWH, and JB).

M. Cavillon, C. J. Kucera, T. W. Hawkins, and J. Ballato are with the Center for Optical Materials Science and Engineering Technologies, Clemson University, Clemson, SC 29634, USA, and also with the Department of Materials Science and Engineering, Clemson University, Clemson, SC 29634 USA (e-mail: mcavill@clemson.edu; courtnk@clemson.edu; hawkin2@clemson.edu; jballat@clemson.edu).

P. Dragic is with the Department of Electrical and Computer Engineering,University of Illinois at Urbana-Champaign, Urbana, IL 61801 USA (e-mail: p-dragic@illinois.edu).

A. F. Runge, and A C. Peacock are with the Optoelectronics Research Centre, University of Southampton, Highfield, Southampton, Hampshire SO17 1BJ, UK (e-mail: a.f.runge@soton.ac.uk; acp@orc.soton.ac.uk).

single mode” waveguides by controlling the losses associated with the higher order modes. Based on their inherently and inextricably multimode nature, such designs introduce additional parasitic phenomena, such as the thermally mediated Transverse Mode Instability (TMI) where, at some threshold power, the beam modal distribution randomizes and becomes dynamic. TMI diminishes beam quality and is presently the main cause of power scaling restrictions in high energy laser applications [2].

This paper presents results based on a second approach to mitigating nonlinear limitations in optical fiber lasers; namely attacking such nonlinearities at their fundamental origin – the material through which the light propagates. Though less well studied within the fiber laser community, such a materials approach offers a more powerful yet simpler and more scalable way to address nonlinearities in comparison to the greater complexity and cost of sophisticated fiber designs [3], [4]. As an example, it is known that the magnitude of stimulated Brillouin scattering is proportional to the Brillouin gain coefficient, gB, which is related to the material density (ρ), refractive index (n), acoustic velocity (Va), Brillouin spectral linewidth (Δν), and the Pockels photoelastic coefficient (p12). The latter parameter, p12, is of particular interest since it is the only factor in determining gB that can take a null value thus permitting, in a suitable material, the complete eradication of SBS. Using this approach, nearly 20 dB suppression in SBS has been achieved using simple core/cladding optical fibers possessing intrinsically low photoelasticity [5]. Raman scattering is an intrinsic material property that can also be reduced using intrinsically low Raman gain materials, coupled with a fabrication process with rapid quench rates that promote a more disordered core glass. This approach has led to a reduction of ∼3 dB in Raman gain coefficient (gR) relative to silica [6]. Finally, thermal effects such as TMI or thermal lensing (TL) are directly proportional to the thermo-optic coefficient (TOC) [2], [7], [8]. Therefore, a reduction, and at best suppression, of TL is expected with low-TOC materials. To date, however, to the best of the Authors knowledge, each of these reductions has been achieved individually in a fiber; i.e., there is no report of simultaneous reductions to SBS, SRS, and TOC in a single fiber from a common core material. This is the purpose of this work.

In order to materially attack each of these nonlinearities collectively, a silica cladding and silicate core glass optical fiber in the ternary strontium fluoride (SrF2) – alumina (Al2O3) – silica (SiO2) precursor system was fabricated. Alumina and alkaline earth oxides in silicates are known to exhibit intrinsically low gB values [5], [9], [10] and the multicomponent and rapidly quenched nature of molten-core-derived fibers yields a reduction in gR. Further, the

Oxyfluoride core silica-based optical fiber with intrinsically low nonlinearities for high energy

laser applications

M. Cavillon, C. J. Kucera, T. W. Hawkins, A. F. J. Runge, A. C. Peacock, Senior Member, IEEE, P. D. Dragic, Member, IEEE, J. Ballato, Fellow IEEE

1

JLT-20479-2017

introduction of fluorine into silicate glasses (in this case through the SrF2) lessens the linear and nonlinear refractive indices [11]–[13], making the fibers fewer-moded and possibly reducing wave-mixing phenomena through of the reduction in n2. Lowering the TOC can be achieved through the introduction of materials with reduced or negative TOC values, such as the alkaline earth fluoride family [14] and, specifically, SrF2 as employed here in a silica matrix. Further, alumina plays a crucial role in glass formation since it prevents the well-known phase separation exhibited in binary alkaline earth silicates [15], [16]. The use of a silicate core and a silica cladding has the added advantage of providing high strength and enhanced thermomechanical stability as is critical to the fieldability of such a fiber.

II.EXPERIMENTAL PROCEDURE

A. Fiber fabricationThe fiber was fabricated using the molten core method

[17]. Briefly, a mixture of high purity commercially available SrF2 and Al2O3 powders (Alfa Aesar; Al2O3, 𝛼-phase: 99.997 % purity, SrF2: 99.99 % purity) was mixed and inserted into a pure silica capillary tube of 3 mm inner and 30mm outer diameter. This preform was drawn at a temperature of about 2000 °C such that the precursor core materials melt, as is characteristic of the molten core method. The preform then was drawn directly into a fiber and the molten core is quenched as the fiber cools and forms a glassy state facilitated by the high cooling rates (>2000 °C/s). Additional details on the fabrication method can be found in [3], [18]. The fiber diameter was 125 µm and 800 m of fiber was collected. The fiber was coated with a conventional UV-curable acrylate coating (Desolite 3471-3-14, DSM Desotech). As the fiber is drawn, silica from the cladding dissolves into the molten core due to thermally activated mass transport, yielding a graded-index core whose composition contains the aforementioned SiO2 along with the original precursor SrF2-Al2O3 mixture. Hence, this fiber is said to be a SrF2-derived aluminosilicate oxyfluoride fiber; designated “SrAlSi-F” hereafter. For completeness and for the benefit of the reader, data from a SrO-derived aluminosilicate (i.e., SrAlSi) analog, previously fabricated and reported in [10], is used here for comparison and discussion as to the role of the fluorine.

B. Fiber characterizationScanning electron microscopy (SEM) analysis was

performed on a representative cross section of the fiber using a Hitachi-6600 SEM, at 25 kV, in order to ensure the absence of phase separation in the core and to determine core and cladding diameters. The cross-sectional composition of the fiber was determined by an energy dispersive x-ray (EDX) probe coupled to the SEM.

The refractive index profile (RIP) was measured transversely at a wavelength of 950 nm through the side of the fiber using a Fourier Transform interferometer [19]. Previous investigations have suggested that the wavelength dependence of the core refractive index difference (Δn) is

particularly small between 950 nm and 1550 nm [20], therefore introducing insignificant error to the present analysis. Attenuation of the fiber was determined using the cut-back method at a wavelength of 1534 nm.

Raman spectra of the fiber core materials (this reported fiber as well as an analogous fiber from [10] that did not possess fluorine) were obtained using a commercial Raman microscope (alpha300, WItech) in a backscattering geometry, utilizing a 532 nm pump source with a focused beam diameter of about 1µm and a photon collection time of 120 s per datum. Since gR is determined relative to fused silica (more details provided in Section D), a data point was taken in the cladding of SrAlSi-F fiber, which served as a reference value for pure fused silica.

Methodologies used to characterize the Brillouin (elastic) properties of the fiber are provided in greater detail elsewhere [20], [21]. Briefly, a heterodyne approach is applied to measure the Brillouin Gain Spectrum (BGS) of the fiber. A pre-amplified narrow linewidth signal (∼100 kHz linewidth at 1534 nm) served as the pump source, while the back-scattered signal is collected through a circulator at the input of the fiber under test (FUT). The Rayleigh back-scattered pump and Brillouin signals are mixed on a fast detector and the resultant electrical signal produced is examined utilizing an electrical spectrum analyzer (ESA). The same apparatus was used to measure the temperature and strain dependencies of the Brillouin frequency shift (BFS), by applying either heat (via a thermal bath) or strain (stretching a mechanically well-secured fiber) to the FUT. For both temperature and strain, the dependence of the BFS was found to be linear [22], [23].

In order to estimate gB, the relative Brillouin scattering strength (amplitude of the Brillouin back-scattered signal) from a known length of control fiber (whose gB is known) is compared with that from the FUT, also of known length [10]. A P2O5-doped silica control fiber was utilized for this measurement since it is very well mode-matched to the fundamental mode of the FUT and its BGS does not overlap with that of the FUT. The mode field diameters (MFDs, calculated from the RIPs) and all losses associated with the measurements were carefully determined as both are necessary for an accurate estimation of gB.

As in [10], a ring laser apparatus was used to characterize the refractive index dependence on temperature and longitudinal strain (thermo-optic coefficient or TOC and strain-optic coefficient or SOC, respectively). More specifically, the ring laser was constructed with a segment of FUT forming part of the cavity. Measurements of the change in laser free spectral range (FSR) when the FUT is subjected to either heat or strain allows the determination of the dependence of the fiber (modal) refractive index on these values. For disambiguation, the TOC = dn/dT and the SOC = p12−ν(p11+p12) where pij are the Pockels photoelastic coefficients and ν is Poisson’s ratio. Multiplying the SOC by -n3/2 yields dn/dε.

Finally, the nonlinear refractive index, n2, was estimated via a simple SPM measurement. In these experiments, high peak power pulses were propagated through the fiber to experience spectral broadening. The source used for this

2

JLT-20479-2017

measurement was a passively mode-locked fiber laser emitting linearly polarized, 750 fs soliton pulses, centered at 1540 nm. The optical pulses were coupled into 4.8 m of SrAlSi-F fiber and the output spectra for different input powers were recorded using an optical spectrum analyzer (OSA). Experimentally, the power coupling efficiency was optimized by checking the output power spectrum and adjusting the input polarization, in the goal to propagate only a single mode through the fiber; the amount of power depleted on the other polarization is considered neglectable. To obtain a value for n2, the measured spectra were compared with numerical modelling using the standard nonlinear Schrödinger equation [24], assuming single mode operation.

III. EXPERIMENTAL RESULTS

A. Fiber propertiesSEM imaging of the fiber cross section was performed and

a representative micrograph is provided in Fig. 1. The core and cladding diameters were determined to be 8 µm (using RIP, Fig. 3) and 125 µm (using SEM imaging, Fig. 1), respectively, and no evidence of phase separation was observed in the core.

Figure 1. Scanning electron microscope (SEM) image of the silica-cladding, strontium aluminosilicate oxyfluoride (SrAlSi-F) core fiber. Note

the simple core/cladding geometry in comparison to conventional large mode area (LMA) fibers studied for high energy laser (HEL) applications.

EDX compositional analysis on the same cross-section of fiber showed a graded profile composed of silicon, strontium, aluminum, fluorine and oxygen (Fig. 2). Again, the silicon and some of the oxygen are incorporated into the molten core during the draw due to dissolution/diffusion of silica from the cladding. In addition, the results are presented in atom percent (At. %) since the exact bonding nature of the fluorine is presently unknown (i.e., it has yet to be determined whether the fluorine remains bound to the strontium or preferentially reacts with the aluminum or silicon). In alkaline earth aluminosilicate glasses, studies have shown that fluorine can react and bond to other species in the glass

(e.g., Sr or/and Si or/and Al) depending on the glass composition and temperature [25]–[30]. In the present fiber, the measured fluorine content was ∼2.72 At.% (2.36 Wt.%) at the core center, which is higher than that typically found in CVD-derived silica fibers. From the EDX data, the atomic ratio of [F]/[Sr] was determined to be about ∼0.91 at the center of the core, a reduction by about half from the initial stoichiometric value of 2 in the SrF2 precursor. This reduction is believed to be due mainly to the formation of volatile SiF4 when SrF2 is in contact with SiO2 in the core melt, similar to CaF2-containing silicate slags [31]–[33].

Figure 2. Compositional profile across the SrAlSi-F fiber core. Not shown for reasons of clarity is the oxygen concentration, %O. However, %O (At.%) =

100 – [%F + %Sr + %Al + %Si].

The measured refractive index profile (RIP) across the fiber core is shown in Fig. 3, relative to the pure silica cladding. Also plotted for comparison is the RIP from the non-fluorine containing SrAlSi fiber of Ref. [10]. The SrAlSi-F fiber exhibits a Δn that was about 2 times lower than the SrAlSi fiber at the core center due to a lower doping concentration (i.e., higher silica content) and, additionally, to the presence of the fluorine. Using the well-known expression for the dimensionless V number and assuming that the index is uniform across the 8 µm core (as a limiting approximation), the SrAlSi-F fiber exhibits a V number value of ∼5.29, whereas this value is ∼10.12 for the SrAlSi fiber of [10]. The origin of this lower V number for SrAlSi-F fiber is principally arises from its reduced Δn, combined with its small core size.

3

JLT-20479-2017

Figure 3. Measured refractive index profile (RIP), relative to fused silica (cladding) for the SrAlSi-F fiber (this work) and a SrAlSi fiber ([10]).

The beneficial effect of fluorine in reducing n in these oxyfluoride glass systems is discussed in section C, and can be employed as an efficient route to fabricate few-mode (such as are typically found in high power fiber laser applications), and ultimately, single mode fibers. As is known, the introduction of Al and Sr oxides to the silica core matrix increases n, and thus the waveguide numerical aperture (NA), compromising single mode operation. There are a number of approaches possible to reduce the number of modes in the fiber. For example, a pedestal design can be implemented [34], with the addition of an index-raising inner cladding layer glass to reduce the NA of a central core. If the use of silica cladding offer great advantages, other cladding glass systems can be use in order to reduce fiber NA [35]. While simpler core-clad step index fiber designs are preferred from a fabrication standpoint, the proposed materials are still amenable to more complicated LMA fibers. In other words, these intrinsically low nonlinearity materials are also compatible with microstructured fiber designs and so a tandem approach of materials, coupled with design could significantly increase optical nonlinearity thresholds.

TABLE IFIBER PROPERTIES OF SrAlSi-F AND SrAlSi FIBERS

Value SrAlSi-F SrAlSi****Core diameter (µm) 8 10.5Cladding diameter (µm) 125 127Δn (10-3)* 35.64 ± 0.005 74.8V number** 5.29 10.12Mode Index, nm (1534 nm; room temp. and zero strain)*** 1.4673 ± 0.007 1.5109

Mode Effective Area (10-12 m2)*** 18.62 ± 0.04 6.10

Attenuation Coefficient (dB/m) at 1534 nm 0.65 ± 0.02 2.68

*Value at the core center.** Calculated using 𝑉 = (2𝜋a/𝜆) ∗ (𝑛𝑐𝑜𝑟𝑒 2 − 𝑛𝑐𝑙𝑎𝑑2)1/2, where a is the core radius. This assumption holds for a step index fiber, and these numbers are only used as indicators. When V<2.405, the fiber is single moded.***For fundamental mode, and calculated from the RIP**** Reproduced from [10].

Attenuation losses and other fiber properties are reported in Table I including measurement errors where appropriate. It should be noted that the high attenuation loss (0.65 dB/m)

relative to conventional LMA fibers is a limiting factor in the practicability of such fibers. However, previous fibers fabricated via the molten core method have shown reduced attenuation losses to the order of ∼0.1 dB/m [36], as the purity of the initial precursors used during the process is expected to grandly contribute to the waveguide attenuation losses. Further efforts to develop higher purity precursor materials are necessary to drive down attenuation losses.

B. Brillouin and thermal/acoustic propertiesThe Brillouin-related properties of the SrAlSi-F fiber, along

with other physical properties, are summarized in Table II; again with a comparison to the non-fluorine-containing SrO-derived aluminosilicate fiber (SrAlSi) analog of [10]. Fig. 4 shows the Brillouin Gain Spectrum (BGS) for the present fiber at two different temperatures (22 °C and 91 °C). These spectra are extremely well-represented by single-Lorentzian fitting functions.

Figure 4. Measured and normalized Brillouin Gain Spectrum (BGS, data

point) for two different temperatures (22 °C, 91 °C), at zero-strain for SrAlSi-F fiber. The data are fitted (solid lines) with Lorentzian curves. The tail

observed at ∼11 GHz is the signature of the SMF used in the measurement apparatus.

The measured gB for the SrAlSi-F fiber was 0.56×10-11

m/W, which is ∼6.3 dB lower than conventional fibers (∼2.4×10-11 m/W [37]). However, its value is larger than that for the SrAlSi fiber for the following reasons. First, the SrAlSi-F fiber contains more silica and, therefore, fewer intrinsically Brillouin reducing compounds. Secondly, this fiber is acoustically guiding, as opposed to the anti-guiding SrAlSi fiber, due to the former exhibiting a lower longitudinal acoustic velocity in the core than in the pure silica cladding. Accordingly, there is no contribution from the waveguide design to gB reduction (waveguide-based acoustic attenuation), as was the case for SrAlSi fiber [10].

The thermo-optic coefficient, TOC, was found to be 0.63x 10-5 K-1, which is ∼ 2.2 dB lower than fused SiO2 (∼1.04×10-

5 K-1) and ∼1.25 dB lower than the SrAlSi fiber (∼0.84×10-5

K-1 [10]). The low intrinsic value of the TOC can be attributed to the presence of fluorine in the fiber core. Indeed, the TOC of Al2O3 and SiO2 glassy constituents are both found to be positive, and only that for SrO is negative [10]. This suggests that a lower Sr content, and consequently a higher Si

4

JLT-20479-2017

content (this is the case for SrAlSi-F compare to SrAlSi), should give a higher nominal value of TOC.

TABLE IISUMMARY OF BRILLOUIN AND THERMAL/ACOUSTIC PROPERTIES

OF SrAlSi-F AND SrAlSi FIBERSValue SrAlSi-F SrAlSi**

BRILLOUIN PROPERTIESν (GHz) 10.754 ± 0.0005 11.975Va (m/s) (acoustic mode value) 5621 ± 28 6079Thermal Coefficient (MHz/K) +0.391 ± 0.02 -0.064Strain Coefficient (GHz/ε) 26.9 ± 0.7 24.4Δν (MHz) 52.0 ± 0.3 123Δνintrinsic (MHz)* 52.0 ± 0.3 87.2gB (10-11 m/W) 0.56 ± 0.4 0.11gB (dB relative to SMF-28) -6.32 -13.4

THERMAL/ACOUSTIC PROPERTIESTOC (10-5 K-1) 0.63 ± 0.03 0.84TOC (dB relative to silica) -2.18 -0.93SOC (dimensionless) 0.151 ± 0.005 0.102

* Insignificant acoustic waveguide contribution to the spectral width.**Reproduced from [10].

Since the opposite trend is observed, we attribute fluorine to be partially responsible for the reduction of the TOC. There is no evidence that F remains attached to Sr post fiber fabrication, as noted earlier, and additional work needs to be carried out to better understand how and in what proportion F participates to an intrinsically low TOC in these oxyfluoride glasses. For completeness, a change in TOC may also be induced by the specific core/cladding fiber geometry, where the core is thermo-mechanically constrained inside the SiO2

in a similar way as in [38]. A forthcoming paper will be dedicated to providing an insight into the structure of the glassy core as well as how it relates to changes in optical and material properties presently determined.

C. Effect of fluorine on glass properties

TABLE IIIMATERIAL PROPERTIES OF SrAlSi-F AT CORE

CENTER*Value SrAlSi-F SrAlSi

No F**Effect of

F***[Al2O3] (mole %) 3.02 3.11 -[SrO] (mole %) 8.47 8.71 -[F] (mole %) 2.71 0 -Index Difference (10-3) 34.81 37.48 -0.89Acoustic Velocity (m/s) 5592 5773 -66.8Δνintrinsic (MHz) 53.0 44.8 +2.91TOC (10-5 K-1) 0.56 0.85 -0.11

*Core center is defined to be the central region of the 4-layer approximation.**Assumes F dilutes the aggregate glass. The molar fraction of the remaining constituents increases with the removal of F.***In units of per mole % of F.

In an effort to gain insight into the role of F in the system, this section presents some mathematical analysis of the results presented in Tables I and II. Previous modeling efforts have focused on the effect of constituents on several material properties of multicomponent glasses [9], [10]. However, since the coordination of F is not currently known, a few key physical properties of the glass are calculated assuming the compositions of Fig. 2 at core center, but assuming no F in the system. Then, these calculations are compared with

measured data to deduce how the presence of F modifies the system.

For brevity, specific details of this ‘additivity’ model, which includes the effect of mode overlap with the material distribution, will not be provided here, but instead can be found in great detail in [3], [5], [20], [21]. These compositions and properties are summarized in Table III. Note that since F dilutes the system, the relative proportion of SrO and Al2O3 drops with the addition of F.

The dependence of refractive index difference on [F] (fluorine concentration in mole %) is somewhat less than what is found in [39], [40] (~ -1.5×10-3 per mole% F when F is added to silica). While this value is somewhat a function of fiber fictive temperature [39], it could be a first hint that F is not necessarily coordinated to Si within this oxyfluoride glass. The dependence of the acoustic velocity on [F] also differs from the data presented in [40] (~ -47.8 m/s/[F]) in that it is roughly 40% stronger (more negative). This could be evidence that F forms at least some AlF3 compounds since its acoustic velocity (extrapolated from [41] for AlF3 to be ~ ½ that of Al2O3) would have the largest reduction relative to its oxide form. Low-acoustic-velocity AlF3 replacing high-velocity Al2O3 in the glass system could therefore account for the larger reduction in overall acoustic velocity relative to F-doped silica. Interestingly, pure, bulk SrF2 has an acoustic velocity (~5400 m/s) [42] comparable to silica and the presence of SrF2 would be expected to raise the acoustic velocity relative to the presence of the same quantity of SrO [10]. That said, the observed data may also be an embodiment of an admixture of fluoride environments (AlF3

+ SrF2 + oxides).Continuing scrutiny of the data, the Brillouin spectral

width is found to increase with increasing [F], however to the best of our ability, literature data could not be found with which a comparison could be made. The addition of F to SiO2

decreases TOC by roughly -0.09×10-6 K-1[F]-1 [43], representing an order-of-magnitude difference from the observation provided in Table III. While Group II fluorides do have negative thermo-optic coefficients, they are roughly equal but opposite in sign to that of silica. Literature values include -13×10-6 K-1 [44] and -15.9×10-6 K-1 [45] for SrF2. This is insufficient to explain the strong effect of F on TOC and therefore it is conceded that this is currently not yet well understood. However, in summary, the advantages in the addition of F to the system clearly include a reduction in the refractive index difference and TOC, the latter possibly offering reduced susceptibility to TMI [2], [4].

D. Nonlinear refractive indexThe output spectra resulting from high power pulse

propagation within the SrAlSi-F fiber are shown in Fig. 5, obtained for four different input powers (orange). These results clearly display the effects of nonlinear spectral broadening due to SPM. Comparing these with the results of the numerical model (black), the nonlinear refractive index is estimated to be around 3×10-20 m2/W. Errors are estimated to be ±10% for this measurement. Although this value is similar to what is typically reported for fused silica [46], it is worth

5

JLT-20479-2017

recalling that the linear refractive index of the SrAlSi-F fiber core is comparatively higher.

Figure 5. Experimental (orange dots) and simulated (black line) self-phase modulation (SPM) spectra for four different input pulse peak powers; 4.8m of

SrAlSi-F fiber were used.

E. Raman properties

Figure 6. Compilation of the corrected and normalized Raman spectra of SrAlSi-F and SrAlSi (from [10]) fibers relative to the pure silica cladding (taken from SrAlSi-F fiber). Spontaneous Raman spectra of SrAlSi-F and SrAlSi fibers are corrected similarly to that in Ref. [47] and normalized

relative to the highest peak of fused silica (∼440 cm-1).

The normalized spontaneous Raman spectra for the SrAlSi-F and SrAlSi fibers are shown in Fig. 6. Typical features have been identified for fused silica (data point taken in cladding) [48], [49]. The main peak at ∼440 cm-1 is attributed to Si-O-Si stretching modes. Peaks at ∼490 cm-1

and ∼600 cm-1 correspond to the so-called defect lines, and other features, at ∼800 cm-1 (bending modes of Si-O bonds), 1060 cm-1 and 1200 cm-1 (associated with some asymmetric and symmetric stretching vibrations) are also present. The presence of Sr as a glass modifier correlates with the reduction in the relative strength of these well-known features due to the subsequent depolymerization of the SiO2

network (and decrease in SiO2 content). This structural modification also promotes the appearance of several additional peaks, specifically one around 1050 cm-1, which is

characteristic of the formation of non-bridging oxygen (NBO) ions [48]. The introduction of alumina results in a broadening of the Raman peaks, as observed in [6]. This feature is more pronounced for SrAlSi than for SrAlSi-F, due to the higher aluminum content in the former fiber. The broadened linewidths of the Raman peaks, therefore, further supports the existence of a more highly disordered glass structure as well as the SEM observation that the rapid cooling rate mitigated phase separation in these otherwise unstable glass compositions.

If fluorine in the SrAlSi-F fiber were principally bound to Si (formation of Si-F bonds), the Raman spectra would exhibit peaks around 950 and 480 cm-1, as reported in [50]–[52]. However, spectral overlap with the Si-O and NBO peaks at similar frequencies, coupled with the relatively low fluorine content relative to the other glass constituents, renders assignment of the fluorine peak contribution more challenging. Further, as noted above, the fluorine may not be exclusively bound to the silicon. Therefore, should the F be distributed among the Si, Al, and Sr, then the relative Raman contributions from F would be further reduced. The exact nature of the fluorine bonding is a topic of on-going study.

Raman gain coefficients, gR, are measured relative to the one for SiO2 (taken to be equal to 1) and correspond to the maximum intensity of the highest peak of the scattering bandwidth (situated around 440 cm-1). Thus, we obtain gR

values for SrAlSi-F and SrAlSi fibers that are lower than silica by ∼0.91 dB and ∼2.15 dB, respectively.The SrAlSi fiber exhibits a lower Raman gain coefficient than the SrAlSi-F fiber, likely due to its lower silica content (and thus a lower Si-O-Si peak intensity) even though the presence of fluorine in the later fiber should contribute to an intrinsically lower gR, all other things being equal [53]. It should be noted that the low Raman response in these aluminosilicate glasses contrasts with a typical increase of Raman cross section intensity when considering other common oxide additives to SiO2, such as GeO2, P2O5 or B2O3 [54]. Therefore, since the introduction of alkaline earth and alumina compounds into silicate glasses reduces gR relative to fused silica, using these materials can be considered as an effective and straight-forward approach to low SRS fibers. In the limit of very large incorporation of dopants into the silica matrix (and therefore, lower silica content), one may expect that peaks other than the one at ∼440 cm-1 will begin to dominate the Raman scattering cross section intensity, likely in the high energy region characteristic of NBOs formation. This may then ultimately be a limiting factor to further reduction of gR. Moreover, glass formation and fabrication of systems with low silica content can become more challenging.

IV. CONCLUSION

Reported here are the properties of a simple silica cladding, oxyfluoride core optical fiber that exhibits intrinsically low Brillouin, Raman, and thermo-optic properties based entirely on the materials from which it is made. Specifically, measured reductions of about 6 dB, 1 dB, and 2 dB in Brillouin gain coefficient (gB), Raman gain coefficient (gR), and thermo-optic coefficient, respectively, were realized as was a “silica-like” nonlinear refractive

6

JLT-20479-2017

index. Whether the fluorine atoms remain bound to the strontium or react with the alumina or silica during the molten core fabrication is presently unknown. Without question, though, fluorine influences each of the nonlinearities that are parasitic to fiber laser performance at high power while also reducing the effective refractive index of the resultant core glass. Accordingly, the chemical disposition of the fluorine during fiber draw, the resultant glass structure, and methods to increase the fluorine concentration in the final fiber are all topics worthy of future study.

ACKNOWLEDGMENT

The authors wish to thank Dr. A. Yablon (Interfiber Analysis, LLC, Sharon, MA, USA) for refractive index profile measurements and J. Tang (Clemson University, Clemson, SC, USA) for help with the Raman measurements. The authors also thank Prof. R. Stolen (Clemson University) for thoughtful discussions on nonlinear interactions in optical fibers.

REFERENCES

[1] J. W. Dawson, M. J. Messerly, R. J. Beach, M. Y. Shverdin, E. A. Stappaerts, A. K. Sridharan, P. H. Pax, J. E. Heebner, C. W. Siders, and C. P. J. Barty, “Analysis of the scalability of diffraction-limited fiber lasers and amplifiers to high average power.,” Opt. Express, vol. 16, no. 17, pp. 13240–13266, 2008.

[2] A. V Smith and J. J. Smith, “Mode instability in high power fiber amplifiers.,” Opt. Express, vol. 19, no. 11, pp. 10180–10192, 2011.

[3] J. B. P. Dragic, “Rethinking Optical Fiber: New Demands, Old Glasses,” J. Am. Ceram. Soc., vol. 96, no. 9, pp. 2675–2692, 2013.

[4] J. Ballato and P. Dragic, “Materials development for next generation optical fiber,” Materials (Basel)., vol. 7, no. 6, pp. 4411–4430, 2014.

[5] P. Dragic, T. Hawkins, P. Foy, S. Morris, and J. Ballato, “Sapphire-derived all-glass optical fibres,” Nat. Photonics, vol. 6, no. 9, pp. 629–635, 2012.

[6] P. D. Dragic and J. Ballato, “Characterisation of Raman gain spectra in Yb:YAG-derived optical fibres,” Electron. Lett., vol. 49, no. 14, pp. 895–897, 2013.

[7] L. Dong, “Stimulated thermal Rayleigh scattering in optical fibers,” Opt. Express, vol. 21, no. 3, pp. 915–918, 2013.

[8] L. Dong, “Thermal lensing in optical fibers,” Opt. Express, vol. 24, no. 17, pp. 19841–19852, 2016.

[9] P. Dragic, C. Kucera, J. Furtick, J. Guerrier, T. Hawkins, and J. Ballato, “Brillouin spectroscopy of a novel baria-doped silica glass optical fiber,” Opt. Express, vol. 21, no. 9, pp. 10924–10941, 2013.

[10] M. Cavillon, J. Furtick, C. J. Kucera, C. Ryan, M. Tuggle, M. Jones, T. W. Hawkins, P. Dragic, and J. Ballato, “Brillouin Properties of a Novel Strontium Aluminosilicate Glass Optical Fiber,” vol. 34, no. 6, pp. 1435–1441, 2016.

[11] J. W. Fleming and D. L. Wood, “Refractive index dispersion and related properties in fluorine doped silica.,” Appl. Opt., vol. 22, p. 3102, 1983.

[12] H. Wehr and D. Wiechert, “Refractive index and density of fluorine doped silica prepared by the PCVD process,” Mater. Res. Bull., vol. 21, no. 5, pp. 559–566, 1986.

[13] K. Nakajima and M. Ohashi, “Dopant dependence of effective nonlinear refractive index in GeO2- and F-doped core single-mode fibers,” IEEE Photonics Technol. Lett., vol. 14, no. 4, pp. 492–494, 2002.

[14] M. J. Weber, Handbook of optical materials. CRC Press, 2003.[15] V. McGahay and M. Tomozawa, “The origin of phase separation in

silicate melts and glasses,” J. Non. Cryst. Solids, vol. 109, no. 1, pp. 27–34, 1989.

[16] B. Hatta and M. Tomozawa, “Effect of Al2O3 on phase separation of SiO2-Nd2O3 glasses,” J. Non. Cryst. Solids, vol. 354, no. 27, pp. 3184–3193, 2008.

[17] J. Ballato and E. Snitzer, “Fabrication of fibers with high rare-earth concentrations for Faraday isolator applications.,” Appl. Opt., vol. 34, no. 30, pp. 6848–54, 1995.

[18] B. S. Morris and J. Ballato, “Molten-core fabrication of novel optical fibers,” Am. Ceram. Soc. Bull., vol. 92, no. 4, pp. 24–29, 2013.

[19] A. D. Yablon, “Multi-wavelength optical fiber refractive index profiling by spatially resolved fourier transform spectroscopy,” J. Light. Technol., vol. 28, no. 4, pp. 360–364, 2010.

[20] P. Law, A. Croteau, and P. D. Dragic, “Acoustic coefficients of P2O5-doped silica fiber : the strain-optic and strain-acoustic coefficients,” Opt. Mater. Express, vol. 1, no. 4, pp. 686–699, 2011.

[21] P. Law, Y. Liu, A. Croteau, and P. D. Dragic, “Acoustic coefficients of P2O5-doped silica fiber: acoustic velocity, acoustic attenuation, and thermo-acoustic coefficient,” Opt. Mater. Express, vol. 1, no. 4, p. 686, 2011.

[22] T. Horiguchi, T. Kurashima, and M. Tateda, “Tensile Strain Dependence of Brillouin Frequency Shift in Silica Optical Fibers,” IEEE Photonics Technol. Lett., vol. 1, no. 5, pp. 107–108, 1989.

[23] D. Culverhouse, F. Farahi, C. N. Pannell, and D. a. Jackson, “Potential of stimulated Brillouin scattering as sensing mechanism for distributed temperature sensors,” Electron. Lett., vol. 25, no. 14, p. 913, 1989.

[24] G. P. Agrawal, Nonlinear Fiber Optics, 2nd ed. Academic Press, 1995.

[25] A. K. Swarnakar, A. Stamboulis, D. Holland, and O. Van Der Biest, “Improved prediction of young’s modulus of fluorine-containing glasses using MAS-NMR structural data,” J. Am. Ceram. Soc., vol. 96, no. 4, pp. 1271–1277, 2013.

[26] A. Stamboulis, R. G. Hill, and R. V. Law, “Characterization of the structure of calcium alumino-silicate and calcium fluoro-alumino-silicate glasses by magic angle spinning nuclear magnetic resonance (MAS-NMR),” J. Non. Cryst. Solids, vol. 333, no. 1, pp. 101–107, 2004.

[27] R. G. Hill, A. Stamboulis, and R. V. Law, “Characterisation of fluorine containing glasses by 19F, 27Al, 29Si and 31P MAS-NMR spectroscopy,” J. Dent., vol. 34, no. 8, pp. 525–532, 2006.

[28] T. J. Kiczenski and J. F. Stebbins, “Fluorine sites in calcium and barium oxyfluorides: F-19 NMR on crystalline model compounds and glasses,” J. Non. Cryst. Solids, vol. 306, no. 2, pp. 160–168, 2002.

[29] C. Dalou, C. Le Losq, B. O. Mysen, and G. D. Cody, “Solubility and solution mechanisms of chlorine and fluorine in aluminosilicate melts at high pressure and high temperature,” Am. Mineral., vol. 100, no. 10, pp. 2272–2283, 2015.

[30] T. J. Kiczenski and J. F. Stebbins, “The effect of fictive temperature on the structural environment of fluorine in silicate and aluminosilicate glasses,” J. Am. Ceram. Soc., vol. 89, no. 1, pp. 57–64, 2006.

[31] M. Persson, S. Seetharaman, and S. Seetharaman, “Kinetic Studies of Fluoride Evaporation from Slags,” ISIJ Int., vol. 47, no. 12, pp. 1711–1717, 2007.

[32] F. Shahbazian, D. Sichen, K. C. Mills, and S. Seetharaman, “Experimental studies of viscosities of some CaO–CaF2–SiO2 slags,” Ironmak. Steelmak., vol. 26, no. 3, pp. 193–199, 1999.

[33] Z. Zhang, G. Wen, P. Tang, and S. Sridhar, “The Influence of Al2O3/SiO2 Ratio on the Viscosity of Mold Fluxes,” ISIJ Int., vol. 48, no. 6, pp. 739–746, 2008.

[34] P. Laperle, C. Paré, H. Zheng, and A. Croteau, “Yb-Doped LMA Triple-Clad Fiber for Power Amplifiers,” Proc. SPIE, vol. 6453, pp. 1–11, 2007.

[35] D. Litzkendorf, S. Grimm, K. Schuster, J. Kobelke, A. Schwuchow, A. Ludwig, J. Kirchhof, M. Leich, S. Jetschke, J. Dellith, J. Auguste, and G. Humbert, “Study of Lanthanum Aluminum Silicate Glasses for Passive and Active Optical Fibers,” Int. J. Appl. Glas. Sci., vol. 331, pp. 321–331, 2012.

[36] J. Ballato, T. M. Tritt, and M. J. Matthewson, “On the Fabrication of All-Glass Optical Fibers From Crystals,” J. Appl. Phys., vol. 105, no. 5, 2009.

[37] P. D. Dragic, “Estimating the effect of Ge doping on the acoustic damping coefficient via a highly Ge-doped MCVD silica fiber,” J. Opt. Soc. Am. B, vol. 26, no. 8, p. 1614, 2009.

[38] P. D. Dragic, C. Ryan, C. J. Kucera, M. Cavillon, M. Tuggle, M. Jones, T. W. Hawkins, A. D. Yablon, R. Stolen, and J. Ballato, “Single- and few-moded lithium aluminosilicate optical fiber for athermal Brillouin strain sensing.,” Opt. Lett., vol. 40, no. 21, pp. 5030–3, 2015.

7

JLT-20479-2017

[39] H. Kakiuchida, N. Shimodaira, E. H. Sekiya, K. Saito, and A. J. Ikushima, “Refractive index and density in F- and Cl-doped silica glasses,” Appl. Phys. Lett., vol. 86, no. 16, pp. 1–3, 2005.

[40] Y. Koyamada, S. Sato, S. Nakamura, H. Sotobayashi, and W. Chujo, “Simulating and designing Brillouin gain spectrum in single-mode fibers,” J. Light. Technol., vol. 22, no. 2, pp. 631–639, 2004.

[41] H. A. A. Sidek, S. Rosmawati, M. K. Halimah, K. A. Matori, and Z. A. Talib, “Effect of AlF3 on the density and elastic properties of zinc tellurite glass systems,” Materials (Basel)., vol. 5, no. 8, pp. 1361–1372, 2012.

[42] S. Alterovitz and D. Gerlich, “Third-Order Elastic Moduli of Strontium Fluoride,” Phys. Rev. B, vol. 1, no. 6, pp. 2718–2723, 1970.

[43] A. Koike and N. Sugimoto, “Temperature Dependences of Optical Path Length in Inorganic Glasses,” Reports Res. Lab. Asahi Glas. Co., Ltd, vol. 56, pp. 1–6, 2006.

[44] Tropf, “Temperature-dependent refractive index models for BaF2, CaF2, MgF2, SrF2, LiF, NaF, KCl, ZnS, and ZnSe,” Engineering, no. 5, pp. 1369–1373, 1995.

[45] F. Druon, S. Ricaud, D. N. Papadopoulos, A. Pellegrina, P. Camy, J. L. Doualan, R. Moncorgé, A. Courjaud, E. Mottay, and P. Georges, “On Yb:CaF2 and Yb:SrF2: review of spectroscopic and thermal properties and their impact on femtosecond and high power laser performance [Invited],” Opt. Mater. Express, vol. 1, no. 3, p. 489, 2011.

[46] M. Weber, D. Milam, and W. Smith, “Nonlinear refractive index of glasses and crystals,” Opt. Eng., vol. 17, no. 5, pp. 463–469, 1978.

[47] M. D. O’Donnell, K. Richardson, R. Stolen, C. Rivero, T. Cardinal, M. Couzi, D. Furniss, and A. B. Seddon, “Raman gain of selected tellurite glasses for IR fibre lasers calculated from spontaneous scattering spectra,” Opt. Mater. (Amst)., vol. 30, no. 6, pp. 946–951, 2008.

[48] P. F. McMillan, “Structural Studies of Silicate Glasses and Melts-Applications and Limitations of Raman Spectroscopy,” Am. Mineral., vol. 69, pp. 622–644, 1984.

[49] F. L. Galeener, “Raman and ESR studies of the thermal history of amorphous SiO2,” J. Non. Cryst. Solids, vol. 71, no. 1–3, pp. 373–386, 1985.

[50] M. Yoshikawa, K. Iwagami, N. Morita, T. Matsunobe, and H. Ishida, “Characterization of fluorine-doped silicon dioxide film by Raman spectroscopy,” Thin Solid Films, vol. 310, no. 1–2, pp. 167–170, 1997.

[51] A. Alessi, S. Girard, M. Cannas, S. Agnello, A. Boukenter, and Y. Ouerdane, “Influence of drawing conditions on the properties and radiation sensitivities of pure-silica-core optical fibers,” J. Light. Technol., vol. 30, no. 11, pp. 1726–1732, 2012.

[52] Y. N. P. Dumas, J. Corset, W. Carvalho, Y. Levy, “Fluorine doped vitreous silica analysis of fiber optic preforms by vibrational spectroscopy,” J. Non. Cryst. Solids, vol. 47, no. 2, pp. 239–242, 1982.

[53] C. Fukai, K. Nakajima, J. Zhou, K. Tajima, K. Kurokawa, and I. Sankawa, “Effective Raman gain characteristics in germanium- and fluorine-doped optical fibers.,” Opt. Lett., vol. 29, no. 6, pp. 545–547, 2004.

[54] F. L. Galeener, J. C. Mikkelsen, R. H. Geils, and W. J. Mosby, “The relative Raman cross sections of vitreous SiO2, GeO2, B2O3, and P2O5,” Appl. Phys. Lett., vol. 32, no. 1, pp. 34–36, 1978.

M. Cavillon received the Engineering degree in materials science from the Department of Materials Science and Sustainable Development at Ecole Supérieure d’Ingénieurs de Recherche en Matériaux, Dijon, France, in 2014. He is currently working toward the Ph.D degree in materials science and engineering at Center for Optical Materials Science and Engineering Technologies and the Department of Materials Science and Engineering, Clemson University, Clemson, SC, USA. He is currently a Research Assistant for Prof. Ballato and is investigating properties of various materials in order to reduce nonlinear effects in optical fiber systems.

C. J. Kucera received her BS in Ceramic and Materials Engineering from Clemson University in 2009. In 2007, she joined the Ballato group working on a variety of light emissive nanoparticles and optical composites. After graduation, she became a Research Associate at the Center for Optical Materials Science and Engineering Technologies (COMSET) where she continues to focus on a variety of light emitting materials. She currently has over 40 publications.

T. W. Hawkins earned a BS in Ceramic Engineering in 1999 from Clemson University. He completed his M. S. in Materials Science from Clemson University in 2005. Currently he is working towards his PhD in Materials Science. In 2012, he took over the role as Optical Fiber Laboratories Director at COMSET. Wade is currently an author on over 60 published scientific papers.

A. Runge received his BSc and MSc in Physics from the Université de Franche-Comté, Besançon, France in, in 2009 and 2011, respectively, and his obtained his PhD from the University of Auckland, New Zealand in 2015. He is currently a Research Fellow at the Optoelectronics Research Centre (ORC), working on nonlinear silicon photonics.

A. C. Peacock (M’14-SM’15) is a Professor of Photonics within the Optoelectronics Research Centre (ORC) at the University of Southampton. She obtained her BSc and MSc in Physics from The University of Auckland (New Zealand) in 1999 and 2001, respectively, and her PhD from the ORC in 2004. In 2007 she was awarded a five-year Royal Academy of Engineering Research Fellowship, following which she established the Nonlinear Semiconductor Photonics group, where the focus of the research is on the design and development of novel semiconductor waveguides. She is a fellow of The Optical Society (FOSA) and the Institute of Physics (FInstP) and currently holds an EPSRC research fellowship.

P. Dragic (M’05) received the PhD degree in 1999 from the University of Illinois at Urbana-Champaign. His thesis work focused on the development of high power fiber lasers for LIDAR instrumentation, and he has continued to work on next-generation laser sources for optical remote sensing. Part of this emphasis is on high-performance fiber lasers and the mitigation of non-linear optical fiber phenomena that limit scalability in these systems. Dr. Dragic has over 100 archival journal and conference papers, is a holder of 15 US patents, and has been involved with the launch of several start-up companies that focused on various optical technologies. He is currently an Assistant Professor in the Department of Electrical and Computer Engineering at the University of Illinois at Urbana-Champaign. He is also Chief Research Officer at Neolight Technologies, LLC of Chicago, Illinois.

J. Ballato (M’97-SM’12-F’17) is a professor of materials science and engineering at Clemson University where he is the inaugural holder of the Sirrine Endowed Chair in Optical Fiber. He earned a B.S. in Ceramic Science and Engineering (1993) and the Ph.D. in Ceramic and Materials Engineering (1997) from Rutgers, The State University of New Jersey. He has published more than 350 technical papers and holds 34 U.S. and foreign patents. Among numerous other honors, his collaborative work on Anderson-localizing optical fiber was chosen as one of Physics World’s Top Ten Breakthroughs for 2014. He is a Fellow of the Institute of Electrical and Electronics Engineers (IEEE), the Optical Society of America (OSA), the International Society of Optical Engineering (SPIE), and the American Ceramic Society (ACerS) as well as being an elected member of the World Academy of Ceramics and the US National Academy of Inventors.

8