Open Access
Add to BibSonomy
Abstract
A novel random laser, integrating a passive optical fiber with a phase separated aluminosilicate core–silica cladding as the feedback medium, is proposed and presented. The core exhibits greatly enhanced Rayleigh scattering, therefore requiring a significantly reduced length of scattering fiber (4 m) for lasing. With a Yb-doped fiber as the gain medium, the fiber laser operates at 1050 nm with low threshold power and possesses an output that can be amplified through conventional means. Furthermore, the laser was found to have a high degree of spatial coherence, spectral broadening with increasing input power, and temporal spectral variation. The facile setup and results herein pave the way for further study and applications based on low threshold random fiber lasers.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Random lasers have drawn considerable attention in recent years. This new breed of light source employs optical feedback from random or disordered scatterers as opposed to the conventional approach which makes use of external ‘mirrors’ [1], consequently forming a well-defined laser oscillator cavity. Random lasers have a host of applications ranging from speckle-free imaging [2,3], to long distance point sensing [4,5], among others [6–10]. Random fiber lasers (RFLs) are one such form of lasers which use fibers and fiber-based components to achieve random lasing. Some RFLs use extrinsic scattering mechanisms for feedback, such as randomized Bragg gratings [11,12], random scattering or reflecting centers [13], or nanocomposite polymer fibers [14]. Other RFLs take advantage of intrinsic scattering mechanisms, including distributed Rayleigh scattering, to provide feedback [15–18]. The RFLs that use Rayleigh scattering as their feedback mechanism typically require long fiber lengths (∼ on the km scale) [16–18] to provide sufficient distributed feedback due to the inherently small degree of Rayleigh scattering in conventional fibers. Consequently, they also typically make use of distributed nonlinear gain mechanisms, such as Raman amplification, which result in high threshold powers (∼W) [15,16,18], atypical output wavelengths [15,18], and may require the use of pumps with unconventional, or less-common wavelengths [16]. Randomized FBG based RFLs can be short in length and exhibit mode selectivity [11,12], but require specialized post-processing of the fiber in the fabrication of the gratings. Rayleigh-based systems, on the other hand, do not show mode selectivity and are attractive since they require no post processing, but the abovementioned typically long fiber lengths preclude compactness. While there are subtle differences between the two laser types, the focus of this work is on the latter.
By way of example, integrating a single mode, 25 km Rayleigh/Raman-based random laser system with a multimode fiber at the output, speckle-free imaging recently was shown [3]. Considering straight-forward Rayleigh-based demonstrations such as these, there is a need for a more practical and compact solution that is easy to implement, operates at a low threshold power, and that offers versatility in terms of amplification of its output. Thus, there is merit in the development of optical fibers with large-scale scattering to significantly reduce the requisite length of the scattering fiber, while also contending with the other aforementioned requirements. Here, scattering enhancements have been accomplished through the fabrication of fibers with heterogeneous core compositions resulting from spinodal phase separation of the glass [19]. Coupled with the use of rare earth doped fibers as the gain media, these fibers enable the construction of a new kind of compact RFL. Further, since the phase separation is intrinsic to the core composition and naturally occurs during the molten core fiber fabrication [20], the process affords scalability and practicality. Utilizing this fiber, a simple, compact RFL is constructed and explored, and its characteristics including power, spectrum, coherence, and temporal statistics are quantified. The laser utilizes Yb-doped fiber as the gain medium and an FBG reflector, which provides flexibility in the laser emission spectrum.
2. Phase separated optical fibers (P-SOF)
2.1 Fiber fabrication
In the proposed RFL, a passive, phase separated aluminosilicate optical fiber (P-SOF) provides random distributed feedback. It was fabricated using the molten core method (MCM) [20]. In short, pure alumina (Al2O3) powder was inserted into a telecommunications-grade silica capillary preform tube (3 mm inner / 30 mm outer diameters) that serves as the fiber cladding after drawing. This powder-in-tube preform then was drawn at a temperature of about 2100 °C. At these temperatures, the core melts, and the silica cladding softens, enabling a direct transition to fiber. As SiO2 from the cladding dissolves into the molten core during the draw, natural immiscibilities in the SiO2 – Al2O3 system lead to heterogeneities in the resultant core as the fiber is drawn and cools. The fiber was drawn with a targeted cladding diameter of 125 µm, and coated with a UV-curable conventional acrylate, yielding a total fiber diameter (including the buffer) of approximately 250 µm. It is noted that the conventional geometry of this fiber design provides the additional advantage of easier splicing with commercially available fibers and fiber-based components. By taking advantage of the draw conditions during fiber fabrication (essentially draw temperature in this case), nano-scale phase separation could be promoted in these fibers possessing cores with lower silica content, by conventional fiber compositional measures [19]. More details on the phase separation process can be found in Ref. [19], where this was investigated for the aluminosilicate system.
2.2 Fiber characterization
Phase separation was observed via electron microscopy. In short, glass core and cladding slices were milled using a focused ion beam coupled with an electron beam microscope (HITACHI NB5000) and thinned down to a 100 nm thickness. Subsequently, the same HITACHI NB5000 was set in a scanning transmission electron microscope configuration (STEM, Bright Field mode, 30 kV) and the nanostructures arising from phase separation were observed. In this case, it was found that spinodal phase separation (shown on the left side in Fig. 1) dominates in the fabricated aluminosilicate core optical fiber. The resulting attenuation of the P-SOF was determined to be ∼18 dB/m at 1050 nm (lasing wavelength) which can principally be attributed to scattering, although some absorptive impurity loss likely also is present [21].
Fig. 1. Scanning transmission electron microscope (STEM) image of the aluminosilicate fiber core (left) and pure silica cladding (right). Spinodal phase separation is apparent in the aluminosilicate fiber, with darker granular regions indicating areas of greater SiO2 concentration relative to Al2O3 [19]. No such nano-structuring is apparent for the pure silica cladding region.
The refractive index profile (RIP) was measured transversely through the side of the fiber at a wavelength of 980 nm using a spatially resolved Fourier transform interferometer [22] and is shown in Fig. 2. The core composition was determined using energy-dispersive X-ray analysis (EDX) and is also provided in Fig. 2. The core was found to contain only alumina and silica, as expected. The maximum refractive index difference, taken between the fiber core center and its silica cladding, was measured to be 0.0587, and the maximum alumina molar concentration was found to be ∼ 19.7%. The diameter of the core was estimated to be about 16.2 µm from the RIP line-scans.
Fig. 2. 1-D linescan extracted from 2-D RIP data and the molar composition in the fiber core from the EDX measurement as a function of the radial position.
Using the RIP data, modal analysis of the P-SOF was performed using an in-house solver to further understand the fiber’s mode distribution. Of particular importance is the implication of mating this fiber with a conventional single mode fiber in the laser system, specifically with 1060-XP (Coherent-Nufern, USA), which is hereon referred to as the ‘connecting fiber.’ At a wavelength of 1050 nm, the P-SOF was confirmed to be multimoded although the overlap between the fundamental (LP01) modes in the 1060-XP fiber and in the P-SOF was found to be very high (∼ 99.1%), as can be discerned from Fig. 3. While the fundamental mode could efficiently be excited upon launching into the P-SOF, strong Rayleigh scattering is expected to couple the LP01 mode quickly into higher-order modes (HOMs), both in the backward and forward directions. This indicates that the overall loss is dominated by the feedback path from the P-SOF, principally due to mode mismatch losses post-scattering (Rayleigh dominated as shown below). In other words, light will scatter into a range of guided modes within the P-SOF, but only the fundamental mode can couple back into the single mode fiber. As will be shown, this loss mechanism, in addition to the Rayleigh scattering loss itself, leads to a significant reduction in the laser slope efficiency. A P-SOF, therefore, designed to have fewer modes, or, ideally, to be single moded at 1050 nm, could significantly improve the performance of any integrated lasing system based on feedback from such phase separated glass cores. However, this can be challenging, as it is the large concentration of alumina, an index-raising dopant, that facilitates phase separation.
Fig. 3. Modal overlap between the fundamental modes (LP01) of the P-SOF and 1060-XP fiber [Mode profiles simulated using RIP data, integrated power normalized].
The nano-scale phase separation observed in the core from the STEM images indicates that scattering should be dominantly Rayleigh, as opposed to Mie, near the lasing wavelength (∼ 1 µm). In order to confirm this conjecture, a scattering profile was measured. More specifically, a segment of fiber was passed axially perpendicular across an aperture (with 1 mm diameter) and green light from a 532 nm doubled Nd:YAG laser was launched into one end. The spatial intensity pattern emerging from the aperture then was characterized by using a Si detector that was attached to a rotation stage (with angular positions, θ, of 0° and 180° representing scattering in the forward and backward direction, respectively). This secured equidistant rotation of the sensor about the aperture. The result of this measurement (normalized) is shown in Fig. 4. The angular dependence of the theoretical normalized Rayleigh scattering intensity(I) is given below [23].
Equation (1) is shown plotted in Fig. 4, visually demonstrating good agreement with measured data. To further quantify this, the equation ($f(\theta ) = \mathrm{\alpha}(1 + {\cos ^2}\theta )$) was fitted to measured data (θ from 40°-140°) via the Nonlinear Least Squares method. The value of the proportionality constant, α, was found to be 0.491 with a corresponding adjusted R-square value of 0.915, indicating a good fit, therefore pointing to the dominance of Rayleigh scattering in this fiber.Fig. 4. Normalized scattering pattern measured at 532 nm for the P-SOF (blue) and the normalized Rayleigh pattern from theory (green). The reduction in scattering approaching 0° and 180° is due to recapture of the scattered light by the fiber.
3. Random laser setup
As will be shown, the enhanced Rayleigh scattering in the P-SOF allows the use of a relatively short length of the P-SOF for distributed random feedback and enables the separation of the gain medium (commercial Yb-doped fiber, YDF) from the feedback medium. Integrating a YDF (or a rare earth doped fiber, generally) for gain also offers the possibility of amplifying the output using successive YDF-based amplifying stages. Turning now to the laser configuration, Fig. 5 shows a block diagram of the setup. A 75 cm length of commercial single mode YDF (Liekki, Yb 1200-4/125 fiber, Thorlabs USA) is pumped through a wavelength division multiplexer (WDM) by a single mode fiber-coupled laser centered at 976 nm (Lumentum, USA). This ultimately constrained the maximum power that can be launched into the system to less than 1W. Feedback at the output end is provided by a fiber Bragg grating (FBG, 50.68% reflectivity, 1049.78 nm center wavelength, and 1.199 nm spectral width, O/E Land, Canada). The FBG is used to select the output wavelength as well as to control the maximum linewidth achievable by the random laser. The feedback from the other end comes from either (A) flat-cleaving the WDM fiber, which essentially acts as a 3.4% Fresnel reflector, or (B) the spliced P-SOF. The former (Case A) was used to establish a benchmark non-random laser for comparison purposes. In the case of the latter (Case B), the fiber length is simply chosen to be long enough (4 m) such that no light emerges from the back-fiber facet. Splicing is performed with a standard telecom splicer for all fibers comprising the system, including that between the WDM and scattering fiber.
Fig. 5. Random laser configuration. The commercial Yb-doped fiber (YDF) is end-pumped by a laser diode operating at 976 nm and the output is taken at the FBG end. An isolator was used to prevent parasitic system lasing from the pump laser facet.
4. Experimental results and discussion
4.1 General lasing characteristics
The output power versus input power curves for the two cases (A and B as defined earlier) are shown in Fig. 6. The input power is defined here as the pump power entering the Yb doped fiber after taking the mode mismatch loss (coupling efficiency theoretically estimated to be 87.4% in going from the WDM to YDF) into account. The random laser was found to have significant amplified spontaneous emission (ASE) and this power was integrated (ASE power fractions ranging from 0.31-0.36 at input powers well above Pth [Pin > 4Pth]) and subtracted from the measurement to provide strictly the lasing output power as shown in Fig. 6. In the case of the random laser (with P-SOF, Case B), observed was a threshold power (Pth) of 62.1 mW and a maximum output power of ∼ 3.3 mW. The low slope efficiency (1.31%) is primarily due to the intrinsic loss associated with the P-SOF and mode mismatch loss between the P-SOF and connecting fiber (1060-XP), as described in Section 2. It is also noted that for both cases, most of the optical power exits not through the output (FBG) end, but rather through the back end of the system. For Case A, this is a flat-cleave with approximately 96.6% transmission, so the power exiting the back end of the laser is roughly equal to the intracavity power. In the case of the P-SOF, this light is scattered from the fiber in a pattern governed by Eq. (1).
Fig. 6. Output power vs input power plot for both cases a) without P-SOF but instead with a flat-cleaved WDM (cavity laser) [Case A], b) with P-SOF spliced to the WDM (random laser) [Case B].
To further understand the lasing characteristics, a model based on Ref. [24] was developed. This application generates slope efficiency ($\eta $) and Pth using the YDF characteristics, YDF length and feedback reflectivity values as inputs. This model was first validated by comparing the experimental and simulated results corresponding to Case A and excellent agreement was achieved. Next, given the slope efficiency and Pth obtained from the random laser power curve, an effective power reflectivity (Reff) can be assigned to the scattering fiber. Reff is defined to be the equivalent reflectivity of a conventional reflector that results in the same lasing characteristics as those (Pth and $\eta $) of the random laser. Its value was adjusted until the model matched experimental results $\delta ({{P_{th}}} )< 6\; mW,\delta (\eta )< 0.25\%$, giving rise to a value of 0.056%. Extrapolating from back scattering measurements found in [19] at 1550 nm, the estimated back reflectance at 1050 nm is in reasonable agreement to that obtained from the laser simulation. This is greater than the Fresnel reflectivity of the interface between the connecting fiber and the P-SOF (estimated to be ∼0.03%), indicating that distributed reflective feedback from the scattering fiber has a lower lasing threshold than the Fresnel reflection. This computation also provides a metric to model and design a more efficient lasing system.
Any random spatial sequence of distributed reflectors, in this case in the P-SOF, in terms of the Fourier representation can be thought of as a sum of many very weak regular gratings (randomly distributed) with fixed periods [16]. Extending this analogy, the overall laser multifrequency output then can be thought of as an aggregation of many monochromatic lasers with arbitrary phase and amplitude. The weak nature of the random gratings indicates that the mode structure in the output spectrum would be dominated by the threshold in each of these individual lasers, rather than the gratings themselves. Figure 7 shows the characteristic temporally dynamic activity in the output spectrum for the random laser, which was observed across the entire range of input powers available. This temporal instability is due to the competition of the lasing modes/frequency components that are formed between the FBG and the randomly distributed reflectors. Although presently constrained in terms of pump power (maximum input power ∼ 308 mW), literature suggests the possibility of temporally stable operation at very high input pump powers [16,25,26]. Figure 7 also provides a visual understanding of the ASE contribution (inset figure gives the full broadband spectrum of the laser).
Fig. 7. Laser spectra (single acquisition, non-averaged) at two different instances in time, t1 and t2, both spectra taken with Pin = 129 mW. (Inset Figure: Broadband spectrum of Random Laser for same Pin.)
Next, the variation of the random laser linewidth with input power is explored. As is clearly apparent, there is spectral broadening with increasing input power (shown in Fig. 8). Since there also is temporal variation (see Fig. 7), these spectra were averaged 25 times. The observed spectral broadening can be attributed to the greater number of wavelength components rising above threshold. We use the term wavelength components since the conventional concept of cavity modes do not apply in this case.
Fig. 8. Averaged output spectra (25 acquisitions) at different input powers, P1=111 mW, P2 = 212 mW, and P3 = 307 mW.
4.2 RF beat spectra
In order to further verify that the configuration was truly operating in a random state, the RF beat spectra at the output of the laser for both the case with P-SOF spliced onto the WDM (Case B) and flat-cleaved WDM case (Case A) were analyzed. This was done by connecting the laser output to a Si avalanche photodetector and viewing the resulting spectra on an electrical spectrum analyzer to observe any beating between cavity modes, should they exist. Random lasing is associated with a distinct lack of well-defined cavity modes whereas a conventional laser can have many close, equally spaced cavity modes [26]. Figure 9 shows the comparison between both cases and it can clearly be seen that the beating between cavity modes is suppressed in the random laser (Case B) whereas in the cavity-moded case (Case A), a number of cavity modes exist with a spacing very close to the estimated free spectral range (Fabry-Pérot cavity) of about 16 MHz. Essentially, the random laser (Case B) has a “modeless” spectrum consisting of random frequency components. It should be noted that ASE was not removed from the signal in acquiring the data shown in Fig. 9. ASE is a broad continuum and any associated beat noise will have a small power spectral density, and therefore negligible impact on the measurements over a 1 GHz range.
Fig. 9. RF beat spectra taken at Pout=1.25 mW for the laser without the P-SOF (WDM flat-cleaved, blue) and with the spliced P-SOF (green) [100 acquisitions averaged].
4.3 Statistics of spectral intensity fluctuations
Whilst the RF beat spectra provide evidence in support of random lasing, more recently the statistical behavior of intensity fluctuations, i.e the deviation from pre-lasing Gaussian to Lévy-like statistics around the onset of random lasing [27–30], has been established as characteristic behavior expected from a random laser. This phenomenon has been observed in random fiber lasers based on randomized Bragg gratings [28], more recently in a random laser based on Rayleigh feedback with the Raman effect as the gain mechanism [29], and even in plasmonic random lasers [30]. In Section 4.1, temporal spectral variation was discussed, however, here the spectral intensity/power fluctuations are quantified and studied using time domain measurements to garner further evidence in support of the notion that this setup is randomly lasing. This is done by collecting spectral intensity values at different instants of time at a particular wavelength from the Yokogawa AQ6370D Optical Spectrum analyzer. The wavelength chosen in this case 1049.6 nm, which is well within the lasing spectrum at all input powers, and a spectral resolution of 0.1 nm was used. The histograms of the spectral intensity values generated by the spectrum analyzer (about 6000 points each) for chosen input powers were then fitted to Lévy α-stable distributions (4 parameters [27]) to ascertain the value of the Lévy exponent α. An α value of 2 corresponds to a Gaussian distribution whereas an α value less than 2 corresponds to a Lévy distribution. The variation of α with normalized input power (as a fraction of the threshold power) is as shown below (Fig. 10). It is noted that since the pump is temperature stabilized, the pump power variations are negligible and the spectral intensity fluctuations stem mainly from laser behavior. Three statistical regimes of spectral intensity/power fluctuations are observed for the RFL system, these are pre-lasing Gaussian (α = 2), Lévy statistics (0 <α <2) around the threshold, and Gaussian statistics (α=2) well above the threshold (Pin/Pth >2.7). The observation of 3 distinct statistical regimes and a sharp drop off in α at the threshold power is consistent with the trends observed in previous works employing fiber based random lasers [28,29], indicating that the fiber lasing setup is randomly lasing.
Fig. 10. Variation of the Lévy exponent α with the normalized input power (normalized to the threshold power).
4.4 Coherence measurements
The coherence of the random laser was investigated using Young’s double slit experiment [31]. The commercially available, vertical double slit (3B Scientific Corporation, USA) had a slit width of 0.15 mm and a slit spacing of 1 mm. The laser output was collimated to a diameter that illuminated both slits symmetrically. The resulting fringes were recorded at a distance of approximately 37 cm from the slits with a CMOS machine vision camera. A neutral-density filter was used to prevent saturation when required. First, a central image was obtained for both cases, as shown in Fig. 11. Then, the camera was translated in the transverse direction (parallel to the plane of the slits) to record the fringe visibility as a function of distance from the central position. The resulting extended fringe pattern shown in Fig. 12 was generated by stitching together those images at different positions, thereby inducing a few discontinuities. Note also that the central fringe in Fig. 12 has an almost constant intensity value due to saturation of the camera. This is on account of the higher power required to observe weaker higher fringe orders, accomplished by decreasing the optical attenuation. Finally, to quantitatively and qualitatively evaluate the degree of coherence, the mutual coherence function ($\gamma $) is obtained from the fringe data. In this measurement, the intensity on both slits is assumed to be equal and $\gamma $ reduces to the Fringe Visibility (V) given by [32]:
where Imax and Imin are the maximum and minimum intensities of the interference fringes, respectively.Fig. 11. Interference fringes observed from the outputs of the [left] cavity-moded laser and the [right] random laser (Top: fringe pattern camera image; Bottom: spatially averaged fringe data) [Pout ∼ 1.6 mW].
Fig. 12. Spatially averaged fringe pattern intensity (0-255) as a function of horizontal position (x) for (a) cavity-moded laser (without P-SOF) and (b) random laser (with P-SOF); Corresponding spectra for (c) cavity-moded laser and (d) random laser (Pout ∼ 3.3 mW).
The results shown in Fig. 11 confirm that the random laser is spatially coherent, just as is observed with the cavity-moded laser (Case A), since clearly defined fringes are exhibited. This can be explained by the fact that the transverse mode is constrained to the fundamental since the YDF, WDM fiber, and the connecting fiber are all single moded at 1050 nm. Though many spatial modes may be excited by scattering in the P-SOF, only the random frequency components corresponding to the fundamental spatial mode can couple back efficiently into the system. These components experience gain and concurrently can lase. Essentially, this configuration does not allow for random behavior to manifest in the transverse spatial dimension. However, a legitimate comparison of the degree of spatial coherence can be made by considering the visibility of only the central fringe. This was found to be 0.68 and 0.74 for the random and cavity-moded lasers, respectively, at equivalent output powers of 1.6 mW. The central fringe visibility was found to be slightly less in the case of the random laser possibly due to the greater ASE observed. However, it is noted that the central fringe visibility is relatively high in both cases, which is characteristic of a spatially coherent laser source [31].
In Fig. 12, only select fringe orders, or rather fringe groupings, are visible. The sinc2 spatial envelope function is a consequence of the finite slit width used, as opposed to the ideal scenario (infinitesimal slit width) where every order would be visible. Decay of the fringe visibility with the fringe order is observed, with the random laser exhibiting fewer distinguishable fringe groups and a faster decay of fringe visibility (depth of modulation of the fringe patterns). This observation is fully in line with its broader linewidth, which necessarily implies a lower coherence length (temporal coherence). The output power was maintained the same for both cases (∼ 3.3 mW) in order to standardize the comparison.
4.5 Cut-back experiments
In order to study the impact of P-SOF length on lasing characteristics, cut-back experiments were performed. In particular, the RF beat spectra were recorded as a function of P-SOF length. In the first study, a flat-cleave at the end of the P-SOF provides additional reflective feedback (Fresnel reflection) from the fiber-to-air interface. From Fig. 13, as the length of the P-SOF decreases, more peaks with larger amplitudes begin to appear in the RF beat spectra, indicating the presence of a greater number of stronger cavity modes. This signifies that the relative strength of the reflective (Fresnel) feedback increases with decreasing P-SOF length, consequently increasing the number of cavity modes. As a side note, it also was observed (but not shown here) that the number of peaks increases with increasing input pump power. This is to be expected, as more cavity modes surpass the threshold. Observed temporal instability in the amplitude and position of these peaks in the spectra indicate competition between different cavity modes for dominance. At very short P-SOF lengths, the RF beat spectra begin resembling a purely cavity-moded laser, also as expected, since now the Fresnel reflection completely dominates over the distributed Rayleigh feedback. The threshold powers however did not change significantly across the experiment.
Fig. 13. RF beat spectra (10MHz- 1 GHz) at Pin = 307 mW for different cutback lengths (with Fresnel reflection concatenated, legend indicates length of P-SOF spliced) [100 acquisitions averaged].
A second cut-back experiment was performed, this time with an angled cleave (20°) at the P-SOF fiber end in order to remove the effects of feedback due to Fresnel reflection. The laser behavior is random across the entire range P-SOF fiber lengths and input pump powers over threshold, as is evidenced by the fact that the RF beat spectra possess no peaks corresponding to cavity mode beating. By way of example, Fig. 14 shows the RF beat spectra for a P-SOF length of 1.5 m at two different input powers (and cleaved at an angle). The other observation here (Fig. 14, inset figure) is that the threshold power increases as the length decreases (though the Pth becomes asymptotic at lower lengths). This is further evidence supporting random lasing from the phase separated fiber, since as Reff decreases the threshold increases.
Fig. 14. RF beat spectra (10 MHz – 1 GHz) at different input powers for spliced P-SOF length of 1.5 m (angled cleave at fiber end) [100 acquisitions averaged]. (Inset Figure: Pth vs length of angle cleaved P-SOF spliced onto the system).
To conclude, it is also noted that a length of 2.45 m is the minimum length of the fabricated P-SOF required for efficient random lasing. This length was identified empirically to be the point where laser threshold and concomitant increase in the light emerging from the angled-cleaved facet (scattering fiber side), while at the same time the P-SOF length, are all minimized. In other words, longer P-SOF fiber lengths did not lead to any further reduction in Pth.
5. Conclusion
In conclusion, a novel random laser at 1050 nm with a maximum output power of 3.3 mW using a phase separated aluminosilicate core – silica cladding fiber (P-SOF) as the reflecting medium has been demonstrated. The P-SOF was fabricated with the molten core method and was found to exhibit nano-scale phase separation and, therefrom, enhanced Rayleigh scattering. The salient features of this random laser include significantly reduced required P-SOF length rendering an ultra-short cavity setup (∼ few meters, in contrast to the ∼100s of meters to km required when using conventional fiber setups) [15–18,26], laser linewidth and lasing wavelength control through the use of an FBG, output characteristics allowing for conventional YDF-based amplification, and low threshold power. Measurements were made confirming the random lasing action through RF beat spectrum measurements. The random lasing action was also confirmed from the trends in the Lévy exponent α obtained from spectral intensity variation measurements. This random laser was found to have a high degree of spatial coherence and the temporal coherence was found to be consistent with the linewidth. Cutback experiments were also carried out, shedding light on the evolution of lasing behavior with P-SOF length and the impact of feedback on the lasing characteristics. The minimum length of P-SOF required for maximum Rayleigh-distributed feedback was also determined to be ∼ 2.5 m. The present work establishes this compact random laser configuration. Whilst random fiber lasers based on Rayleigh scattering is an up-and-coming field with research still ongoing to find novel applications, this setup potentially lends itself well for use in speckle-free imaging [3] and internally modulated lasers with reduced distortion [33]. Power scaling efforts and investigations into other glass families are currently on-going.
Funding
Joint Directed Energy Transition Office (N00014-17-1-2546); U.S. Department of Defense.
Disclosures
The authors declare no conflicts of interest.
References
1. R. V. Ambartsumyan, N. G. Basov, P. G. Kryukov, and V. S. Letokhov, “Non-resonant feedback in lasers,” Prog. Quantum Electron. 1, 107–185 (1970). [CrossRef]
2. B. Redding, M. Choma, and H. Cao, “Speckle-free laser imaging using random laser illumination,” Nat. Photonics 6(6), 355–359 (2012). [CrossRef]
3. R. Ma, Y. J. Rao, W. L. Zhang, and B. Hu, “Multimode Random Fiber Laser for Speckle-Free Imaging,” IEEE J. Sel. Top. Quantum Electron. 25(1), 1–6 (2019). [CrossRef]
4. Z. N. Wang, Y. J. Rao, H. Wu, P. Y. Li, Y. Jiang, X. H. Jia, and W. L. Zhang, “Long-distance fiber-optic point-sensing systems based on random fiber lasers,” Opt. Express 20(16), 17695–17700 (2012). [CrossRef]
5. M. Fernandez-Vallejo, M. Bravo, and M. López-Amo, “Ultra-long laser systems for remote fiber Bragg gratings arrays interrogation,” IEEE Photonics Technol. Lett. 25(14), 1362–1364 (2013). [CrossRef]
6. D. S. Wiersma, “The physics and applications of random lasers,” Nat. Phys. 4(5), 359–367 (2008). [CrossRef]
7. X. H. Jia, Y. J. Rao, Z. N. Wang, W. L. Zhang, Y. Jiang, J. M. Zhu, and Z. X. Yang, “Towards fully distributed amplification and high-performance long-range distributed sensing based on random fiber laser,” Proc. SPIE 8421, 842127 (2012). [CrossRef]
8. D. V. Churkin, S. Sugavanam, I. D. Vatnik, Z. Wang, E. V. Podivilov, S. A. Babin, Y. Rao, and S. K. Turitsyn, “Recent advances in fundamentals and applications of random fiber lasers,” Adv. Opt. Photonics 7(3), 516–569 (2015). [CrossRef]
9. Q. Song, S. Xiao, Z. Xu, V. M. Shalaev, and Y. L. Kim, “Random laser spectroscopy for nanoscale perturbation sensing,” Opt. Lett. 35(15), 2624–2626 (2010). [CrossRef]
10. Y. Li, P. Lu, X. Bao, and Z. Ou, “Random spaced index modulation for a narrow linewidth tunable fiber laser with low intensity noise,” Opt. Lett. 39(8), 2294–2297 (2014). [CrossRef]
11. M. Gagné and R. Kashyap, “Demonstration of a 3 mW threshold Er-doped random fiber laser based on a unique fiber Bragg grating,” Opt. Express 17(21), 19067–19074 (2009). [CrossRef]
12. W. L. Zhang, R. Ma, C. H. Tang, Y. J. Rao, X. P. Zeng, Z. J. Yang, Z. N. Wang, Y. Gong, and Y. S. Wang, “All optical mode controllable Er-doped random fiber laser with distributed Bragg gratings,” Opt. Lett. 40(13), 3181–3184 (2015). [CrossRef]
13. X. Wang, D. Chen, H. Li, L. She, and Q. Wu, “Random fiber laser based on artificially controlled backscattering fibers,” Appl. Opt. 57(2), 258–262 (2018). [CrossRef]
14. S. Sebastian, C. L. Linslal, C. P. G. Vallabhan, V. P. N. Nampoori, P. Radhakrishnan, and M. Kailasnath, “Random lasing with enhanced photostability of silver nanoparticle doped polymer optical fiber laser,” Laser Phys. Lett. 11(5), 055108 (2014). [CrossRef]
15. H. Zhang, P. Zhou, H. Xiao, and X. Xu, “Efficient Raman fiber laser based on random Rayleigh distributed feedback with record high power,” Laser Phys. Lett. 11(7), 075104 (2014). [CrossRef]
16. S. K. Turitsyn, S. A. Babin, A. E. El-Taher, P. Harper, D. V. Churkin, S. I. Kablukov, J. D. Ania-Castañón, V. Karalekas, and E. V. Podivilov, “Random distributed feedback fibre laser,” Nat. Photonics 4(4), 231–235 (2010). [CrossRef]
17. Q. Meng, H. Wu, B. Han, J. Li, and Z. Wang, “LD-Pumped Random Fiber Laser Based on Erbium-Ytterbium Co-Doped Fiber,” Photonic Sens. 10(2), 181–185 (2020). [CrossRef]
18. H. Zhang, H. Xiao, P. Zhou, X. Wang, and X. Xu, “Random distributed feedback Raman fiber laser with short cavity and its temporal properties,” IEEE Photonics Technol. Lett. 26(16), 1605–1608 (2014). [CrossRef]
19. M. Cavillon, P. Dragic, B. Greenberg, S. H. Garofalini, and J. Ballato, “Observation and practical implications of nano-scale phase separation in aluminosilicate glass optical fibers,” J. Am. Ceram. Soc. 102(3), jace.16099 (2018). [CrossRef]
20. J. Ballato and A. C. Peacock, “Perspective: Molten core optical fiber fabrication- A route to new materials and applications,” APL Photonics 3(12), 120903 (2018). [CrossRef]
21. K. C. Kao and G. A. Hockham, “Dielectric-fibre surface waveguides for optical frequencies,” Proc. Inst. Electr. Eng. 113(7), 1151–1158 (1966). [CrossRef]
22. A. D. Yablon, “Multi-wavelength optical fiber refractive index profiling by spatially resolved Fourier transform spectroscopy,” J. Lightwave Technol. 28(4), 360–364 (2010). [CrossRef]
23. A. Bucholtz, “Rayleigh-scattering calculations for the terrestrial atmosphere,” Appl. Opt. 34(15), 2765–2773 (1995). [CrossRef]
24. C. Barnard, P. Myslinski, J. Chrostowski, and M. Kavehrad, “Analytical model for rare-earth-doped fiber amplifiers and lasers,” IEEE J. Quantum Electron. 30(8), 1817–1830 (1994). [CrossRef]
25. A. A. Fotiadi, “Random lasers: an incoherent fibre laser,” Nat. Photonics 4(4), 204–205 (2010). [CrossRef]
26. S. A. Babin, A. E. El-Taher, P. Harper, E. V. Podivilov, and S. K. Turitsyn, “Tunable random fiber laser,” Phys. Rev. A 84(2), 021805 (2011). [CrossRef]
27. R. Uppu and S. Mujumdar, “Lévy exponents as universal identifiers of threshold and criticality in random lasers,” Phys. Rev. A 90(2), 025801 (2014). [CrossRef]
28. B. C. Lima, A. S. Gomes, P. I. Pincheira, A. L. Moura, M. Gagné, E. P. Raposo, C. B. de Araújo, and R. Kashyap, “Observation of Lévy statistics in one-dimensional erbium-based random fiber laser,” J. Opt. Soc. Am. B 34(2), 293–299 (2017). [CrossRef]
29. J. Li, H. Wu, Z. Wang, S. Lin, C. Lu, E. P. Raposo, A. S. Gomes, and Y. Rao, “Lévy spectral intensity statistics in a Raman random fiber laser,” Opt. Lett. 44(11), 2799–2802 (2019). [CrossRef]
30. J. H. Hsiao, S. W. Chen, B. Y. Hung, K. Uma, W. C. Chen, C. C. Kuo, and J. H. Lin, “Resonant energy transfer and light scattering enhancement of plasmonic random lasers embedded with silver nanoplates,” RSC Adv. 10(13), 7551–7558 (2020). [CrossRef]
31. Y. Deng and D. Chu, “Coherence properties of different light sources and their effect on the image sharpness and speckle of holographic displays,” Sci. Rep. 7(1), 5893 (2017). [CrossRef]
32. B. Redding, M. A. Choma, and H. Cao, “Spatial coherence of random laser emission,” Opt. Lett. 36(17), 3404–3406 (2011). [CrossRef]
33. M. Bravo, M. Fernandez-Vallejo, and M. Lopez-Amo, “Internal modulation of a random fiber laser,” Opt. Lett. 38(9), 1542–1544 (2013). [CrossRef]
