Open Access
Add to BibSonomy
Abstract
We report on the feasibility of short-wavelength transmission window modification in anti-resonant hollow core fibers using post-processing by hydrofluoric (HF) acid etching. Direct drawing of stacked anti-resonant hollow core fibers with sub-micron thin cladding capillary membranes is technologically challenging, but so far this has been the only proven method of assuring over an octave-spanning transmission windows across the visible and UV wavelengths. In this study we revealed that low HF concentration allows us to reduce the thickness of the cladding capillary membranes from the initial 760 nm down to 180 nm in a controlled process. The glass etching rates have been established for different HF concentrations within a range non-destructive to the anti-resonant cladding structure. Etching resulted in spectral blue-shifting and broadening of anti-resonant transmission windows in all tested fiber samples with lengths between 15 cm and 75 cm. Spectrally continuous transmission, extending from around 200 nm to 650 nm was recorded in 75 cm long fibers with cladding membranes etched down to thickness of 180 nm. The experiment allowed us to verify the applicability and feasibility of controlling a silica fiber post-processing technique, aimed at broadening of anti-resonant transmission windows in hollow core fibers. A practical application of the processed fiber samples is demonstrated with their simple butt-coupling to light-emitting diodes centered at various ultraviolet wavelengths between 265 nm and 365 nm.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Ultra-violet (UV) light sources are used in numerous applications, for example in time-resolved spectroscopy [1–3], semiconductor industry [4,5], photocatalysis [6], or trapped-ion manipulation [7]. Deep-UV light from gas-filled fibers has been used for ultrafast molecular spectroscopy [8]. Notwithstanding, light guidance in the UV is challenging because of high attenuation or fragility of compatible optical fibers. Fused silica remains the most versatile material for optical fibers due to its exceptional mechanical robustness and high laser damage threshold, ultra-low loss in the telecommunications range, as well as mature technology and low cost. However, silica has high attenuation in the UV and in the mid-infrared spectral range. The development of non-silica glasses (soft glasses) such as telluride [9] and chalcogenide glasses [10,11], addressed the challenges inherent to optical fibers for application in the mid-infrared part of spectrum. The soft glass-based fibers have been used for supercontinuum light source development in the mid-infrared with output powers 1-2 order of magnitude higher than typical synchrotron infrared beamlines [12]. Choice of glasses for the UV wavelengths is limited, but practical UV transmitting step-index fibers are available and include UV-grade silica fibers or fluoride glass fibers, among which the most well know is the ZBLAN glass (ZrF4-BaF2-LaF3-AlF3-NaF) [13–15]. Photodarkening phenomena and generation of broadband-absorbing color centers in silica, as well as in soft glass fibers becomes a practical limiting consideration for application of such fibers with intense UV light handling [13]. Fluorophosphate glass also enables low attenuation extending in the UV down to a wavelength of 180 nm but similarly to fluoride glasses, the fiber drawing technology is challenging [16].
As an alternative, use of hollow core fibers has been a promising approach for UV light delivery. In a hollow core fiber, the light is confined in the air core with negligible overlap with the glass. Additionally, dispersive and nonlinear deformations of light pulses can be much lower than in solid glass core fibers, which enables, among others, handling of laser pulses with large energy in the UV spectral range [17,18]. The first demonstrations of hollow-core fibers for UV wavelengths have been realized using the kagomé fibers [19,20]. In one such implementation, a kagomé fiber with a core diameter (Dcore) of 30 µm supported multimode guidance in UV range with the loss of 2 dB/m at λ = 355 nm [19]. When the core diameter is further decreased, higher order modes would experience increased attenuation, and hence a small-core kagomé fiber has been demonstrated guiding UV light in the fundamental spatial mode over a meter-scale distance [20]. A silica hollow core fiber with Dcore = 20 µm and the cladding membranes thickness of 220–260 nm, exhibited loss of 3 dB/m at a wavelength of λ = 280 nm over the long-term operation of 14 hours [20]. Simplifications of the cladding structure have led to the single capillary ring cladding hollow core fibers in which guiding can be described with the anti-resonant reflecting optical waveguide model [21]. These antiresonant fibers (ARFs) have also been reported in UV delivery applications [17,18,22–24], offering transmission at wavelengths as short as 270 nm with attenuation of around 3 dB/m [23]. Single mode operation at the UV wavelengths has also been reported in ARFs [23], but in non-touching capillary cladding ARFs it requires in general stringent control of capillary diameter d to core diameter Dcore ratio (d/Dcore) as demonstrated by authors in [24]. The common feature of different ARFs reported for the UV wavelength range is then that the core is generally smaller – in the 10-20 µm range - and the cladding capillary membranes are thinner – in the 100-500 nm range – than the fibers reported and successfully implemented in applications involving mid-infrared wavelengths [25–29]. The resonance wavelengths in an ARF are given by the equation [30,31]:
where λm is the resonance wavelength of the order m, t is the thickness of the capillary membranes, n and nc are refractive indices of fused silica glass and the core medium, respectively (nc ≈ 1 for an air-filled hollow-core fiber). Thus, the thinner the cladding capillary wall t, the shorter are the resonant wavelengths and the closer their spectral spacing is. Therefore, in the spectral range of interest, with smaller t we obtain broader transmission windows related to lower order resonance wavelengths, compared to a fiber with thicker membranes. It has been shown with experimental results, that the capillary membrane thickness below roughly 500 nm is one of the key factors for achieving octave-spanning anti-resonant transmission windows over visible and near-infrared wavelengths, as well, and that decreasing this thickness contributes to blue-shifting of the transmission windows [32]. Such ultra-thin cladding capillary hollow core fibers have subsequently found their way into extensive research on generation, temporal recompression and handling of intense UV laser pulses or high energy near-infrared pulses [33–36]. Applications in UV supercontinuum generation in gas-filled structures have also been reported [37–39]. Fabrication of these extremely thin cladding capillary membranes using the conventional stack-and-draw method is however technologically challenging. Post-processing of drawn fibers can be implemented to thin the cladding membranes using the tapering technique and membrane thinning down to even below t = 100 nm at the thinner taper end have been demonstrated with silica ARFs [40].An alternative means for blue-shifting of transmission bands in ARFs is infiltration of the fiber hollow interior with low refractive index liquids [41,42]. The decrease of refractive index contrast causes the shifting of transmission windows toward the shorter wavelengths and the reduction of transmission window widths [42]. In such fibers, the material loss of the selected liquids is the main contributor to attenuation. Recently, ARFs infiltrated with liquids have been used for optofluidic applications [43–45], however, due to high loss, this modification of the ARF optical properties cannot be considered as an alternative for opening of the UV transmission bands in these structures.
Here, we report on combining the feasibility of liquid infiltration of an ARF inner microstructure, with the functionality of its post-processing, aimed at broadening and blue-shifting the fibers’ short-wavelength transmission windows. The work addressed this directly by demonstrating with technological detail an alternative method of reaching extreme geometric parameters in a hollow core fiber, which have been shown critical to the enhancement of its UV transmission characteristics. We specifically employ wet etching to thin the capillary cladding membranes in non-touching capillary cladding silica ARFs. In the experiments, various concentrations of hydrofluoric (HF) acid are pumped through ARF sample lengths between 15 cm and 100 cm using a microfluidic pressure pump. Previously, similar technique was used with photonic bandgap (PBG) fibers with comparably much more complex structure, and narrow-band near-infrared transmission was investigated in short fiber samples only (20 cm) [46]. We focus on much simpler fiber structure, in which the influence of geometric modification of cladding (thinning of capillary membranes) is straightforward to isolate from other influences, such us reduction or modification of geometry of the mesh rings in the PBG lattice. We note that Köttig et al. reported on use of a HF-etched ARF in a gas cell for UV light generation, however no details of the etching dynamics nor fiber parameters were provided [37].
This work was focused on the short-wavelength spectral range and broadening and blue-shifting of fiber transmission spectra is recorded over the UV, visible and near-infrared wavelengths up to 800 nm under different conditions involving HF concentration and exposure time. Scanning electron microscopy (SEM) images of etched fiber structures have been recorded for these varying sets of parameters. In combination, this allowed us to establish characteristic post-processing parameters like silica membrane etching rates (thickness reduction in time) and blue-shifting of transmission windows (spectral shift in time) as well as effective HF concentrations for non-destructive thinning of capillary membranes. Finally, we demonstrated a proof-of-concept application of the etched fibers as simple, butt-coupled fiberization of output of various UV light emitting diodes (LEDs).
2. Investigated fiber and experimental setup
2.1 Fabrication of the initial anti-resonant fibers
The initial ARF was developed in-house using the conventional stack-and-draw method. The fiber structure is shown in scanning electron microscopy (SEM) images in Fig. 1. It has a cladding formed by a single ring of 6 capillaries with outer diameter (d) of around 14.5 µm, and membrane thickness t = 760 nm. All the membrane thicknesses across the work are provided as averaged thicknesses over all six cladding capillaries. The range of geometric parameters of fibers, i.e. the diameters and thickness of the cladding capillaries in the pre-etched fibers was small due to conditions of the fiber drawing process. We assured the same drawing temperature and gas pressure fed into the preform during the entire drawing process. Therefore the recorded capillary membrane thickness variability was below 2% for any given capillary (i.e. along its circumference), and was within 3% between any two capillaries across the fiber’s section. The core diameter Dcore is larger than that of ARFs used for UV light delivery in earlier works [17,18,25], and consequently larger is the capillary-to-capillary separation. Admittedly, in a practical application such a large capillary separation is undesirable, because it contributes to increased attenuation. The fiber supports multimode light guidance in both the UV and visible parts of spectrum. These parameters are agreeably less favorable for shortwave and especially UV transmission applications, than the previous works. We note, that use of a large core fiber in this work was motivated, among others, by the feasibility to avoid very time-consuming, low-rate process of liquid infiltration of the fiber’s hollow core during etching. The advantage of this for our study is that we were able to investigate and demonstrate etching using different concentrations of HF in water. This in turn allowed us to draw conclusions on the role of this parameter in the shaping of such process characteristics like etching rates and processing times and etching uniformity along the fiber sample.
Fig. 1. SEM images of the initial hollow core fiber used in the etching experiments: (a) general view of the fiber microstructure, (b) close-up of one of the cladding capillaries.
2.2 Description of primary setups for wet-etching and transmission measurements
The system as shown in Fig. 2(a) is used to experimentally investigate the reduction of capillary thickness. A microfluidic pump with controlled pressurized gas is used to fill the liquid into the tested fiber. HF-water solutions with volume HF concentration of 0.2%, 0.4%, 2.5%, and 5.5% were used for the etching process. The procedure started with the etching solution fed from the reservoir into the hollow space of the fiber through a plastic tube labelled “A” in Fig. 2(a), until the fiber was completely infiltrated with the liquid. Next, after a specific etching time, tube A was replaced by tube B containing distilled water. In order to stop wet etching inside the fiber, distilled water was fed into the fiber through tube B for 30 minutes, to push the HF solution out of the ARF. Finally, pressurized air was blown into the fiber through tube C for 60 minutes to flush the residual water out. During the etching process, the reservoir containing the HF-water solution at the given concentration was constantly pumped using a microfluidic pressure controller (Fluigent, MFCS-EZ) with pressure setting of 80 kPa. Flow-rate of the liquid inside the fiber was measured using a flow-rate platform (Fluigent, Flowboard), and typical flow rates recorded over duration of an etching procedure are shown in Fig. 2(b). For example, when operating under pressure of 80 kPa, the flow rate was roughly 19 µl/min and 5 µl/min with a fiber sample length of 25 cm and 130 cm, respectively. The time of liquid infiltration into a 1-meter fiber sample was around 2 minutes. It has been shown, that the infiltration time in capillary fibers depends on the capillary diameter [47]. Based on this data, a 1 m long capillary fiber 14-15 µm in diameter (which is comparable to the cladding capillaries in our fiber) would be infiltrated in roughly 200 s, which is comparable to the infiltration time observed experimentally in our work and still significantly shorter than the 500-600 minutes long etching times with certain HF concentrations used in our procedures. Thus, the cladding capillaries in the fibers were not collapsed (i.e. were left open) during infiltration and etching.
Fig. 2. (a) Experimental setup for hollow core fiber etching; (b) Flow rate of liquid flow inside the fiber under applied pressure of 80 kPa and various lengths of the fiber sample.
The etching process was uniform at the fiber cross-section, i.e. all capillaries were etched similarly. We did not observe any gravitation-related etching nonuniformity of membrane thickness from capillary to capillary and variability of their thickness was comparable to the initial fiber at 3% (up to 2% for any single capillary along the circumference). It is to be noted, that relatively low etching concentrations were used in the experiments and the flow of the acid was fast through the sample. The lower end of the range of capillary membrane thicknesses obtained in the etching procedures described in the following sections was around t = 200 nm or t = 140 nm (depending on the given HF concentration and etching time), which corresponds to tolerances of t = 200 nm ±3 nm and t = 140 nm ±2 nm.
Fig. 3. Experimental setup used for measurement of transmission characteristics of ARFs in the visible and near-infrared wavelengths.
Optical characterization of all fiber samples in the first step involved transmission measurements and the experimental setup used for this is shown in Fig. 3. A supercontinuum light source was used as a reference signal, with spectral coverage in a wavelength range of 450 nm-2400 nm (Koheras SuperK Compact, Denmark). L1 is a microscope objective with a focal length of 11 mm and numerical aperture (NA) of 0.26. L2 is 10× microscope objective. The output beam from the fiber was coupled into a multimode fiber connected with spectrometers. Thorlabs CCS-200 (wavelength range of 200-900 nm) and an Optical Spectrum Analyzer (wavelength range of 600-1700 nm) were used to record the output spectra. Additionally, a compact spectrometer with 355 nm blaze wavelength (Ocean Insight HDX, USA) was used in measurements of UV transmission windows in the etched fibers.
3. Results and discussion
3.1 Wet etching of short hollow core fiber samples
Thickness of the cladding capillary membranes t, is gradually reduced as result of the etching. Figure 4 shows SEM images of etched ARF samples using 0.4% HF-water solution and etching time of 60 minutes, 300 minutes, 450 minutes, and 540 minutes. For the particular ARF used and the 0.4% HF solution concentration, we observed that 540 minutes is the maximum safe processing time, within which the fiber structure remains intact. Longer exposure times resulted in disintegration of one or more cladding capillaries, as shown in Fig. 5 for etching time of 630 minutes. The decrease of capillary thickness follows a roughly linear trend for all demonstrated concentrations of HF-water solution. The slope depends on the concentration of HF, as shown in Fig. 6. Use of low concentration HF-water solutions necessitates very long etching time. For example, in the case of HF of 0.2%, the ARF fiber was etched to cladding capillary thickness of t = 560 nm in 360 minutes. More concentrated HF solutions were used to investigate feasibility of decreasing the etching time. Infiltration of the ARF with HF-water solution of 0.4% HF resulted in cladding capillary thickness reduction from the initial t = 760 nm down to t = 440 nm in 300 minutes, and down to t = 220 nm in 540 minutes. Application of 2.5% HF solutions scales the capillary thickness down to t = 206 nm in 100 minutes. The most concentrated HF solution of 5.5% used in this work, etched the capillaries down to t = 300 nm in just 32 minutes. ARF samples of the initial fiber, each 15 cm long, were subjected to etching with the given HF-water solution concentrations for specific time durations, and after flushing out with distilled water and air, the etched samples were inspected for capillary thickness t corresponding to different etching times. The obtained data on capillary thickness change over etching time, shown in Fig. 6, follows roughly linear trend for every investigated concentration of HF-water solution. The calculated etching rates, expressed in units of nm of etched thickness over time (nm/min) are: 0.32 nm/min for the 0.2% solution, 0.95 nm/min for the 0.4% solution, 5.3 nm/min for the 2.5% solution, and 13 nm/min for the 5.5% solution. The thinnest capillaries obtained in these experiments were around 200 nm and the further decrease of capillary thickness was limited by nonuniformity of the cladding capillary diameters and thicknesses in the initial fiber. We also note, that for the two most concentrated HF solutions used in the work, 2.5% and 5.5%, the high etching rate made it difficult to stop the etching process efficiently by flushing the hollow microstructure with distilled water.
Fig. 4. SEM images of etched AR fibers and their cladding capillary with etching time of 60 mins (a, e), 300 mins (b, f), 450 mins (c, g), and 540 mins (d, h). The concentration of HF is 0.4%. Images were taken from the acid input end of fiber samples.
Fig. 5. SEM image of AR fiber etched by HF of 0.4% in 630 mins. Images were taken from the acid input end of fiber samples.
Fig. 6. The capillary thickness of etched AR fiber with the difference of concentration of HF and etching time.
In the next step, fiber samples of the initial (unetched) ARF and the samples etched using different HF solution concentrations at different processing times, were investigated for their transmission properties. The measurements were first conducted over 430-1700 nm wavelength range and the results are shown in Fig. 7. It is to be noted that the fibers supported multimode guidance. The complex structure of the plateaus of the individual anti-resonant transmission windows is assigned to the interference between the modes including the fundamental mode and the higher order modes. The transmission windows of the fibers are presented by their order and additionally indicated by the order of resonant wavelengths, such as the 2nd transmission band is located between resonant wavelengths given by m = 2 and m = 1 following Eq. (1). The initial fiber (capillary membrane thickness t = 760 nm) had three transmission bands in the investigated wavelength range labelled in Fig. 7 with orders 2nd, 3rd, and 4th. The first-order transmission window is beyond the investigated wavelength range, and it is not shown. Etching of the capillary thickness had a two-fold effect on the hollow core fiber transmission characteristics. Firstly, it resulted in blue-shifting of the resonant wavelengths and of the transmission windows. Measurements of transmission spectra of the fiber samples subjected to etching with different HF concentrations revealed, that this effect depended on both the HF concentration and etching time. In the case of the 2nd order transmission window, in the unetched fiber centered around a wavelength of 1250 nm, the rates of resonant wavelengths blue-shift established with the use of HF solutions with concentrations between 0.2% and 5.5% were between 0.54 nm/min up to as much as 19.5 nm/min, respectively. Rates established for the 3rd order transmission window, initially centered around a wavelength of 730 nm, were slightly lower. The etching rates obtained for each of the HF solution concentrations used in the etching procedures are gathered in Table 1. The second effect of etching on the transmission characteristics of the processed fibers is a practical consequence of blue-shifting of the resonant wavelengths. Thinning of the cladding capillary membranes in the cladding results in the blue-shifting of the resonant wavelengths, so the order m of the resonant wavelengths in a given wavelength range is decreased, and transmission windows in this wavelength range are broadened spectrally. The same reason is related to the fact, that the resonance wavelength blue-shift rate expressed in the unit of nm/min takes smaller values for the 3rd order transmission window, as compared to the 2nd order window for the same HF concentration and etching rate. The combined effects of ARF capillary etching on the transmission characteristics are shown in Fig. 7 for all investigated HF solution concentrations and etching times. Notably, for the 0.4% HF solution, the 3rd and 4th order transmission windows of the ARF shift towards the ultraviolet wavelengths from the visible, while the 1st order transmission window shifts from the near-infrared into the visible. Additionally, in the fiber etched using the 0.4% HF solution over 540 minutes, when the capillary membrane thickness is thinned down to t = 220 nm, the 1st order window covered the whole of the investigated wavelength range. We note, that in these measurements, the short-wavelength limit was imposed by the used reference light source – a high brightness supercontinuum source with shortwavelength spectral edge around 450 nm. In the following part of work, this source was replaced with either a broadband deuterium lamp or a set of LEDs operating at different UV wavelengths.
Fig. 7. Transmission windows of initial ARF and ARF samples etched using different HF concentrations and etching times: (a) 0.2% HF solution, (b) 0.4% HF solution, (c) 2.5% HF solution, (d) 5.5% HF solution. The fiber samples were around 15 cm long.
Table 1. The change of capillary thickness and blue-shifting transmission windows with different concentrations of HF.
3.2 Etching uniformity along the fiber length
In the previous section, we discussed how the etching effectiveness – understood as the feasibility to shift and broaden the ARF transmission windows – depended on the concentration of the HF solution and the etching time. From the process dynamics point of view, we are concerned with two effects. One, the etching first occurs at the acid input end of the fiber, and then the process gradually takes place along the fiber and away from the input end. Second, the concentration of the HF solution along the fiber changes due to the chemical reactions taking place. However, the acid solution was continuously pumped into the fiber during processing. The first effect results in that the etching time, i.e. the time during which the etching solution interacted with the glass capillary membranes at different sections of the fiber, was different – specifically it can be expected that the longest interaction, that is etching time, was at the acid input end of the fiber. However, taking into account that the time to fill the entire capillary was about 2 minutes and when etching with a 0.4% solution the etching ratio is about 0.95 nm/min, this effect can be neglected. The difference between the observed infiltration time of the fiber and the anticipated infiltration time of the cladding capillaries (120 s to 200 s respectively) is also an order to two orders of magnitude shorter than the etching times in the procedures investigated in this work. Thus the contribution of the non-collapsed cladding capillaries to etching non-uniformity can be assumed negligible for the longer etching procedures employing lower HF concentrations used in our experiments (although it can play a role for short etching times in general). The second and more important effect results in the fact that the fiber closer to the input end is always etched in a solution of higher concentration [48]. Consequently, the capillary at the input end can be expected thinner than that at the output end [49]. Indeed, in the case of high HF concertation, e.g. 2.5% and an etching time of 150 minutes, the capillary thickness at the input end of ARF was t = 180 nm, while in the same sample, at a length 40 cm from the acid input end, the capillary thickness was t = 345 nm – the respective SEM images are shown in Fig. 8. Thus the difference of capillary thickness between each end of the fiber is around 412 nm in case of a 1 m long sample. In other words, for the 1 m long fiber sample etched over 150 minutes, the capillary wall thickness at the acid input end and at the acid output end of fiber were t = 180 nm and t = 592 nm, respectively. For the etching process performed with a with lower HF concentration of 0.4% but over a longer time of 600 minutes, the capillary thickness at the acid input end was 140 nm, while at 40 cm and 100 cm ARF length from the input end, it was t = 250 nm, and 400 nm, respectively, as shown in Fig. 9. Therefore, for etching conditions with lower HF concentrations and longer processing time, the cladding capillary thickness difference between the acid solution input and output ends is almost a factor of 2 smaller, i.e. 260 nm, for a 1 m long ARF sample. These results allow to draw a conclusion, that while high HF concentration etching does indeed shorten the processing times, it also leaves large nonuniformity of geometric dimensions of the fiber microstructure along its length. On the contrary, diluting the HF solution necessitates longer etching time, but brings in the benefit of decreased nonuniformity of the cladding capillary thickness along the processed fiber. Thus in order to consider scaling up of the etched fiber lengths, dilution of the etching liquid and extension of the etching times should be considered to alleviate the limitation related to etching nonuniformity. This is even more pronounced when short sample lengths are subjected to etching. For example, in a 15 cm long sample, although etching nonuniformity can be already observed, it is limited to less than 3% of membrane thickness increase, from t = 345 nm to t = 355 nm in the case of etching with 0.4% HF solution over 450 minutes, as shown in SEM images in Fig. 10. On the contrary, when using the 2.5% HF solution in etching over only 90 minutes, the nonuniformity of membrane thickness in a 15 cm long sample exceeded 13%, with the thickness going up along the sample from t = 300 nm to t = 340 nm, as shown in Fig. 11 (from acid input toward acid output end of fiber). Therefore – for the procedure involving the more diluted etching solution and longer processing time – decreasing of the HF concentration and extending the etching time by roughly a factor of 4, decreased the membrane thickness nonuniformity by nearly an order of magnitude.
Fig. 8. Etched capillary thickness in a single ARF sample processed with 2.5% HF solution over 150 minutes as observed from the (a) acid input end and (b) at length of 40 cm from the acid input.
Fig. 9. Etched capillary thickness in a single ARF sample processed with 0.4% HF solution over 600 minutes as observed from the (a) acid input end, (b) at length of 40 cm from acid input (c) at length of 100 cm from acid input.
Fig. 10. Etching nonuniformity in 15 cm long fiber samples, 0.4% HF solution and 450 minutes etching time, (a),(c) acid input end of sample, (b),(d) acid output end of sample.
Fig. 11. Etching nonuniformity in 15 cm long fiber samples, 2.5% HF solution and 90 minutes etching time, (a),(c) acid input end of sample, (b),(d) acid output end of sample.
In the next step we investigated the influence of the longitudinal nonuniformity of capillary thickness on the transmission spectra of wet etched ARFs. Samples of an ARF of 75 cm length have been prepared by wet etching over 600 minutes using 0.4% HF-water solution. Then, transmission windows were measured using a broadband deuterium lamp (SLS204, Thorlabs, USA) as the reference light and transmission spectra were recorded using a short-wavelength compact spectrometer optimized for the UV-violet wavelengths (blaze wavelength 355 nm, Ocean Insight HDX, USA). The etched fiber was then cut into three 25 cm long sections and transmission spectra were recorded in each of the sections, as well. Firstly though, the transmission spectra were measured over two directions of propagation in the 75 cm long sample – coupling the light from the acid input and propagating towards the acid output, as well as when coupling from the acid output and propagating towards the input end. As shown in Fig. 12(a), the influence of nonuniformity of capillary thickness on the propagation direction is negligible, i.e. regardless of which fiber end is used for in-coupling of light, at the output transmission characteristic averaged over the loss mechanisms along the fiber is recorded. Impact of membrane thickness nonuniformity is revealed when transmission characteristics are recorded for each of the consecutive, 25 cm-long sections of the fiber: i.e. the first section after the acid input, the second (“middle”) section, and the final section towards the output of the original, 75 cm long sample. This is shown in Figs. 12(b)–12(d), from the first fiber sample (from the acid input) up until the final fiber sample (towards the acid output). The three transmission windows labelled 1-3 (circled labels in the plots) are visible in the 200-800 nm wavelength range in the first and the middle fiber section, while in the final fiber section, where the membrane thickness was etched the least, the 1st order window remains at wavelengths longer than 800 nm. The central wavelengths of the 2nd order transmission windows in the consecutive fiber sections are located at wavelengths of 435 nm, 480 nm, and 533 nm (from the first up until the last fiber section). The center wavelengths measured in the consecutive fiber sections for the 3rd transmission window are 289 m, 299 nm and 309 nm. Despite this evolution, the transmission in the etched fiber covers the short-wavelength, i.e. UV-visible spectral range without significant anti-resonances. The UV feature (intensity dip) at around 250 nm wavelength is not affected by thinning of the capillary membranes since it is not related to a resonance.
Fig. 12. Transmission windows of ARFs etched by HF of 0.4% over 600 minutes: (a) 75 cm long sample before cutting into three consecutive sections 25 cm long each, (b) the first section – the acid input section, (c) the second, “middle” section, (d) the final section towards acid output.
3.3 Butt coupling of wet-etched hollow core fibers with UV-emitting LEDs
The relative simplicity of the etching approach to UV-violet transmission improvement of ARFs motivated verification of the feasibility of their application in a comparably simple arrangement of fiber-pigtailed UV light sources. For this purpose, we tested transmission performance of the etched fiber samples with a series of UV-emitting LEDs M265D2, M300D3, and M365D2 (Thorlabs, USA), operating at central wavelengths of 265 nm, 300 nm, and 365 nm. The optical arrangement of this experiment was intentionally very simple, as shown in Fig. 13(a) to maximize reproducibility of measurement conditions upon replacement of the fiber samples. No bulk optical elements were used between the LEDs and the ARFs, i.e. the light was butt-coupled to the fibers directly from the LEDs. Output light was either imaged using a standard imaging camera or collected with a 600 µm core, UV-grade silica patchcord (Thorlabs M114l01-IC) and its spectrum recorded with a UV-optimized spectrometer (Ocean Insight HDX, USA). The spectrometer integration constant was maintained at the same level in all measurements and the spectrometer’s high UV sensitivity due to its 355 nm blaze wavelength and its broad dynamics range supported reproducibility of measurement conditions after replacement of the consecutive fiber samples. The ARF sample used here was etched over 480 minutes with the 0.4% HF solution and was 27 cm long. This study allowed to observe, that under very simple experimental conditions – also very simple to maintain from one fiber sample to the other i.e. from the unetched fiber to the etched fiber, we observed clear improvement of UV light intensity at the fiber output after wet etching, shown in the recorded spectra in Fig. 13(b). In each butt-coupling case, the guided mode was successfully excited as evidenced with mode field images. Admittedly, the share of cladding tube light was noticeable, as well, but this can be managed with fiber geometry of the initial fiber subjected to wet etching. It is to be noticed also that no coupling optics was used in an attempt to improve the coupling quality here. The experiment confirms suitability of the procedure to application in construction of very simple, pigtailed UV light sources.
Fig. 13. (a) Experimental setup used for UV LED butt-coupling tests of the etched ARFs; (b) Transmission spectra recorded over 27 cm long samples of the initial, unetched ARF and etched fiber, when the fibers were butt-coupled to UV-emitting LEDs M265D2, M300D3, and M365D2 operating at wavelengths around 265 nm, 300 nm and 365 nm. Mode field camera images were recorded for the etched fiber output.
3.4 Attenuation measurements in wet-etched fibers
In the final step, we measured attenuation of the original ARF prior to the etching procedures, and of the etched fibers, in the visible and the UV spectral range. The cut-back method was used over three consecutive cuts. The initial lengths of the fiber samples were 170 cm for the ARF before etching and 100 cm for the etched fiber. A tungsten-halogen lamp (SLS201L, Thorlabs, USA) with a wavelength range of 360-2600 nm and a deuterium lamp (SLS204, Thorlabs, USA) with a wavelength range of 200-700 nm were used as the reference light sources. A multimode fiber (Thorlabs-M14L01) with core diameter of 50 µm was used to guide reference light from each of the lamps to the tested fibers, as shown schematically in Fig. 14(a). The choice of fiber was motivated by roughly matching the mode field diameter between the multimode delivery fiber and the investigated ARFs. The multimode fiber patchcord to ARF coupling was arranged without any optics (simple butt-coupling). We note that the commercially available solarization-resistant multimode fibers for UV delivery usually have large core diameter (>100 µm), and their use in this experiment would result in large portion of reference light coupled into the cladding of the ARF, which would be highly undesirable. Use of a 50 µm core delivery fiber allowed efficient excitation of the guided modes of the ARF without noticeable excitation of the cladding modes, as shown in Fig. 14(b). Two compact spectrometers were used in the study, one optimized for the UV and short-wave visible (Ocean Insight HDX) and one for the long-wave visible and near-infrared (CCS-200, Thorlabs). The output beam from the tested fibers was directly coupled to the spectrometers as shown in the inset in Fig. 14(a).
Fig. 14. (a) Experimental setup for attenuation measurements; (b) mode profile image at a wavelength of 450 nm at the output of the etched fiber (0.4% HF, 480 minutes etching time).
Fig. 15. Comparison of anti-resonant transmission windows in the short-wave visible and UV wavelengths of fibers (a) prior to etching and (b) post-etching using 0.4% HF solution, the light colored concave trace without resonance bands is the halogen lamp output spectrum shown for reference; (c) comparison of measured attenuation spectra of the fibers prior to- and post-etching.
Fig. 16. Comparison of anti-resonant transmission windows in the short-wave visible and UV wavelengths of fibers (a) prior to etching and (b) post-etching using 0.4% HF solution, the light colored concave trace without resonance bands is the deuterium lamp output spectrum shown for reference; (c) comparison of measured attenuation spectra of the fibers prior to- and post-etching.
The main loss contributions in an ARF at the investigated wavelength range are confinement loss (CL), surface scattering loss (SSL) and micro-bend losses. Regular bending losses are neglected in this work because the investigated fibers were kept straight during measurements and the etching procedures. The CL loss depends on the core diameter and capillary to core diameter ratio d/Dcore. It has been shown, that for ARFs with 6 cladding capillaries, when the ratio d/Dcore decrease below 0.6 the CL increases [36]. In the case of the investigated fibers, d/Dcore is around 0.24, which explains generally higher attenuation than ARFs reported in previous works either experimentally [17,18] or with the use of numerical modelling [50–52]. The SSL loss depends on the roughness of inner glass surfaces of the fiber and varies inversely as λ3 [50], which makes this loss contribution important at visible and UV wavelengths. The micro-bend losses have been found to dominate over the SSL at short wavelengths and in fibers with increased capillary gap separation [53,54], thus their role in shaping the loss characteristics of our etched fibers cannot be ruled out. We note that the large core fiber used in this work was intentionally selected to carry out the etching experiments within a reasonable timeframe in order to reveal favorable combinations of etching times and HF-water solution concentrations for which the detrimental influence of the etching procedure would be minimal. Focused on these experiment-driven objectives, we have also omitted theoretical treatment of the etched fibers using numerical models.
The initial (unetched) ARF had around 5 dB/m attenuation in the near-infrared, which increased in the short-wavelength range. At the wavelength of 520 nm, its measured attenuation was 6.1 dB/m, as shown in Fig. 15(a). The etching process reduces the capillary thickness, which is beneficial to transmission characteristics, but nonuniformity of capillary thickness along the fiber is a source of loss. Transmission spectra obtained for the etched fiber show noticeable narrowing of anti-resonant windows towards their red-shifted spectral edges, which corresponds to our earlier discussion of transmission window narrowing due to longitudinal nonuniformity of capillary membrane thickness. Thus the increase of attenuation in the etched fibers to around 7-10 dB/m is assigned to this phenomenon, instead of surface scattering loss, although the latter is not entirely ruled out [55,56]. It is to be noted, that despite this loss figure, the transmission coverage in the etched fiber is improved over the fiber prior to etching with one single transmission band covering the entire visible wavelength range in the etched fiber.
At the short-wavelength visible and UV wavelengths (below 500 nm), the initial, unetched fiber has seven transmission bands in the investigated wavelength range of 200-500 nm with narrow spectral widths as shown in Fig. 16(a). Measured attenuation exceeds slightly 5 dB/m at 400 nm, but goes up to 8 dB/m at the wavelength of 225 nm. Measured transmission spectrum of the etched fiber shows only three resonance loss bands instead of seven as in the same ARF prior to etching, Fig. 16(b). The loss of this etched fiber is higher than that of the initial fiber due to, among others, the roughness of the silica surfaces increasing the scattering loss [50]. We note, that nonuniform etching and gradual “washing out” of anti-resonance bands also contributed to increased losses in the etched samples. The etched fiber had loss of 11dB/m and 14.5 dB/m at the wavelengths of 320 nm and 300 nm, respectively, which roughly doubles the loss of the unetched fiber in comparable wavelength range. We were not able to record broadband attenuation spectra in the etched fibers at wavelengths shorter than roughly 270 nm. The shortest wavelength for which a reproducible attenuation measurement was obtained in our experimental conditions was at 235 nm, which yielded 15 dB/m. Notably, this result was obtained at a wavelength roughly at a resonance in the un-etched, initial fiber, which confirms blue-shifting of anti-resonant transmission window, shown in Fig. 16(c) in reference to wavelength of 235 nm.
4. Conclusions
Reported work involved post-processing of anti-resonant hollow core fibers using HF solution wet etching. The procedure was aimed specifically at thinning of the fiber’s cladding capillary membranes from the initial 760 nm thickness, down to below 200 nm. Such membrane thickness is exceedingly difficult to obtain directly in a fiber drawing process. Optical characterization of the etched fibers was concentrated on the short-wavelength part of spectrum, because the initial, unetched fibers had capillary membrane thickness comparable to the visible light wavelength, making them moderately applicable for transmission in this wavelength range. It is to be highlighted that the purpose of the work was not the pursuing of a strict attenuation figure comparison to fibers with extreme geometric parameters – including the cladding membrane thickness below 200 nm – obtained directly in a stack-and-draw process, such as the fibers demonstrated e.g. by authors in [18]. Instead, we carried out a study involving postprocessing of a fiber with relatively thick capillary membranes using a wide range of process variables. Specifically, we changed the HF concentration in the etching solution from 0.2% to 5.5%. The etching results, including capillary thickness, attenuation and etching uniformity were recorded at different processing durations from 30 minutes to over 600 minutes of etching time. This provided a broad view of how a hollow core fiber’s geometry can be changed using very simple and easily reproducible processing, compared to the direct drawing of 100-200 nm thick membrane fibers directly. The etching procedures have shown, that transmission characteristics of such fibers in the visible and the UV can be improved due to blue-shifting and spectral expanding of the anti-resonant transmission windows, when the capillary membranes are thinned. For the most favorable combination of HF solution concentration of 0.4% and etching time of 540 minutes, we obtained a continuous transmission window covering the near-infrared (up to 1600 nm) down towards 450 nm without any resonant loss bands. In the short-wavelength part of the visible and UV (200 nm – 500 nm) etching resulted in decreasing the number of resonant bands from 7 to 3, i.e. 7 narrow transmission bands over 200-500 nm were replaced with three broad transmission windows. We investigated etching of fibers as long as 100 cm.
The influence of thickness nonuniformity of the capillaries was verified as well. While it is noticeable in the shape of the transmission spectra along the fiber, we noticed that it does not depend on the input direction, i.e. the transmission characteristic remains practically unchanged from one direction of propagation to the other along the fiber. Furthermore, the difference in membrane thickness after etching can be limited when processing with decreased HF solution concentrations and over extended periods of time, instead of using high HF concentrations and rapid etching. It is worth highlighting that this combination of processing parameters could be considered in etching of fibers with much smaller air cores, i.e. 15-17 µm in diameter, which has been found more suitable for UV transmission, than large air core fibers [17,18]. Successful processing of fibers at least 100 cm long also provided an opportunity to perform reasonably accurate attenuation measurements. While the length is too short for loss measurements according to commonly accepted step-index fiber standards, these measurements here were important, because the impact of etching on surface roughness in the processed fibers was unknown. The tests revealed, that etching increased fiber UV loss by up to 50%. This is related to decrease of surface quality of capillary membrane in anti-resonant hollow core fibers related to process of nonuniform glass surface etching with HF solutions [57]. Mitigation of loss can be obtain if HF solution will be supplemented by HNO3 or replaced by other solvents as KOH [58]. Very good surface quality can also be achieved with etching technology based on molten salts – this would involve use of alkali metal or alkaline earth metal cation salts consisting of aqueous solutions of sodium salts, lithium salts, calcium salts, magnesium salts or their mixtures [59]. In this case glass surface roughness is reduced down to below 0.5 [60].
Finally, we used the etched fibers in a simple feasibility test involving butt-coupling of UV emitting LEDs, in each case observing an improvement of excited mode intensity over the unetched fiber without use of any coupling optics. The proposed post-processing method is suitable for simple preparation of anti-resonant hollow core fibers for sensor and spectroscopy applications, in which confinement and spectrally broad transmission is required in the visible and the UV wavelengths.
Funding
H2020 Marie Skłodowska-Curie Actions (H2020-MSCA-ITN-2016 Grant No 722380, SUPUVIR); Fundacja na rzecz Nauki Polskiej (First TEAM, POIR.04.04.00-00-1D64/16); Narodowe Centrum Nauki (PRELUDIUM-18 UMO-2019/35/N/ST7/01768).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. A. Stolow, A. E. Bragg, and D. M. Neumark, “Femtosecond Time-Resolved Photoelectron Spectroscopy,” Chem. Rev. 104(4), 1719–1758 (2004). [CrossRef]
2. L. V. Besteiro, H. Zhang, J. Plain, G. Markovich, Z. Wang, and A. O. Govorov, “Aluminum Nanoparticles with Hot Spots for Plasmon-Induced Circular Dichroism of Chiral Molecules in the UV Spectral Interval,” Adv. Opt. Mater. 5(16), 1700069 (2017). [CrossRef]
3. M. Oppermann, B. Bauer, T. Rossi, F. Zinna, J. Helbing, J. Lacour, and M. Chergui, “Ultrafast broadband circular dichroism in the deep ultraviolet,” Optica 6(1), 56–60 (2019). [CrossRef]
4. B. Yaakobovitz, Y. Cohen, and Y. Tsur, “Line edge roughness detection using deep UV light scatterometry,” Microelectron. Eng. 84(4), 619–625 (2007). [CrossRef]
5. T. Manouras and P. Argitis, “High Sensitivity Resists for EUV Lithography: A Review of Material Design Strategies and Performance Results,” Nanomaterials 10(8), 1593 (2020). [CrossRef]
6. A. Jain, A. Kumar, H. Kaur, and V. Krishnan, “Strategic combination of ultra violet-visible-near infrared light active materials towards maximum utilization of full solar spectrum for photocatalytic chromium reduction,” Chemosphere 267, 128884 (2021). [CrossRef]
7. C. W. Chou, D. B. Hume, J. C. J. Koelemeij, D. J. Wineland, and T. Rosenband, “Frequency Comparison of Two High-Accuracy Al+ Optical Clocks,” Phys. Rev. Lett. 104(7), 070802 (2010). [CrossRef]
8. N. Kotsina, F. Belli, S. Gao, Y. Wang, P. Wang, J. C. Travers, and D. Townsend, “Ultrafast Molecular Spectroscopy Using a Hollow-Core Photonic Crystal Fiber Light Source,” J. Phys. Chem. Lett. 10(4), 715–720 (2019). [CrossRef]
9. H. Timmers, A. Kowligy, A. Lind, F. C. Cruz, N. Nader, M. Silfies, G. Ycas, T. K. Allison, P. G. Schunemann, S. B. Papp, and S. A. Diddams, “Molecular fingerprinting with bright, broadband infrared frequency combs,” Optica 5(6), 727–732 (2018). [CrossRef]
10. D. D. Hudson, S. Antipov, L. Li, I. Alamgir, T. Hu, M. E. Amraoui, Y. Messaddeq, M. Rochette, S. D. Jackson, and A. Fuerbach, “Toward all-fiber supercontinuum spanning the mid-infrared,” Optica 4(10), 1163–1166 (2017). [CrossRef]
11. C. R. Petersen, U. Møller, I. Kubat, B. Zhou, S. Dupont, J. Ramsay, T. Benson, S. Sujecki, N. Abdel-Moneim, Z. Tang, D. Furniss, A. Seddon, and O. Bang, “Mid-infrared supercontinuum covering the 1.4–13.3 µm molecular fingerprint region using ultra-high NA chalcogenide step-index fibre,” Nat. Photonics 8(11), 830–834 (2014). [CrossRef]
12. C. R. Petersen, P. M. Moselund, L. Huot, L. Hooper, and O. Bang, “Towards a table-top synchrotron based on supercontinuum generation,” Infrared Phys. Technol. 91, 182–186 (2018). [CrossRef]
13. T. Sylvestre, A. R. Ragueh, M. W. Lee, B. Stiller, G. Fanjoux, B. Barviau, A. Mussot, and A. Kudlinski, “Black-light continuum generation in a silica-core photonic crystal fiber,” Opt. Lett. 37(2), 130–132 (2012). [CrossRef]
14. M. Poulain, M. Poulain, and J. Lucas, “Verres fluores au tetrafluorure de zirconium proprietes optiques d'un verre dope au Nd3+,” Mater. Res. Bull. 10(4), 243–246 (1975). [CrossRef]
15. M. Poulain, M. Poulain, and J. Lucas, “Structure cristalline de SmZrF7. Relations structurales avec le type ReO3,” J. Solid State Chem. 8(2), 132–141 (1973). [CrossRef]
16. G. Galleania, Y. Ledemi, E. S. de Lima Filho, S. Morency, G. Delaizir, S. Chenu, J. R. Duclere, and Y. Messaddeq, “UV-transmitting step-index fluorophosphate glass fiber fabricated by the crucible technique,” Opt. Mater. 64, 524–532 (2017). [CrossRef]
17. S. F. Gao, Y. Y. Wang, W. Ding, and P. Wang, “Hollow-core negative-curvature fiber for UV guidance,” Opt. Lett. 43(6), 1347–1350 (2018). [CrossRef]
18. F. Yu, M. Cann, A. Brunton, W. Wadsworth, and J. Knight, “Single-mode solarization-free hollow-core fiber for ultraviolet pulse delivery,” Opt. Express 26(8), 10879–10887 (2018). [CrossRef]
19. S. Février, F. Gérôme, A. Labruyère, B. Beaudou, G. Humbert, and J.-L. Auguste, “Ultraviolet guiding hollow-core photonic crystal fiber,” Opt. Lett. 34(19), 2888–2890 (2009). [CrossRef]
20. F. Gebert, M. H. Frosz, T. Weiss, Y. Wan, A. Ermolov, N. Y. Joly, P. O. Schmidt, and P. St. J. Russell, “Damage-free single-mode transmission of deep-UV light in hollow-core PCF,” Opt. Express 22(13), 15388–15396 (2014). [CrossRef]
21. M. A. Duguay, Y. Kokubun, T. L. Koch, and L. Pfeiffer, “Antiresonant reflecting optical waveguides in SiO2-Si multilayer structures,” Appl. Phys. Lett. 49(1), 13–15 (1986). [CrossRef]
22. A. Hartung, J. Kobelke, A. Schwuchow, K. Wondraczek, J. Bierlich, J. Popp, T. Frosch, and M. A. Schmidt, “Double antiresonant hollow core fiber – guidance in the deep ultraviolet by modified tunneling leaky modes,” Opt. Express 22(16), 19131–19140 (2014). [CrossRef]
23. A. Hartung, J. Kobelke, A. Schwuchow, J. Bierlich, J. Popp, M. A. Schmidt, and T. Frosch, “Low-loss single-mode guidance in large-core antiresonant hollow-core fibers,” Opt. Lett. 40(14), 3432–3435 (2015). [CrossRef]
24. A. D. Pryamikov, A. F. Kosolapov, G. K. Alagashev, A. N. Kolyadin, V. V. Vel’miskin, A. S. Biriukov, and I. A. Bufetov, “Hollow-core microstructured ‘revolver’ fibre for the UV spectral range,” Quantum Electron. 46(12), 1129–1133 (2016). [CrossRef]
25. B. P. Uebel, M. C. Günendi, M. H. Frosz, G. Ahmed, N. N. Edavalath, J.-M. Ménard, and P. St.J. Russell, “Broadband robustly single-mode hollow-core PCF by resonant filtering of higher-order modes,” Opt. Lett. 41(9), 1961–1964 (2016). [CrossRef]
26. F. Yu, W. J. Wadsworth, and J. C. Knight, “Low loss silica hollow core fibers for 3–4 µm spectral region,” Opt. Express 20(10), 11153–11158 (2012). [CrossRef]
27. F. B. A. Aghbolagh, V. Nampoothiri, B. Debord, F. Gerome, L. Vincetti, F. Benabid, and W. Rudolph, “Mid IR hollow core fiber gas laser emitting at 4.6 µm,” Opt. Lett. 44(2), 383–386 (2019). [CrossRef]
28. M. Nikodem, G. Gomółka, M. Klimczak, D. Pysz, and R. Buczyński, “Demonstration of mid-infrared gas sensing using an anti-resonant hollow core fiber and a quantum cascade laser,” Opt. Express 27(25), 36350 (2019). [CrossRef]
29. P. Jaworski, K. Krzempek, G. Dudzik, P. Sazio, and W. Belardi, “Nitrous Oxide detection at 526 µm with a compound glass Antiresonant Hollow-Core Optical Fiber,” Opt. Lett. 45(6), 1326–1329 (2020). [CrossRef]
30. J.-L. Archambault, R. J. Black, S. Lacroix, and J. Bures, “Loss Calculations for Antiresonant Waveguides,” J. Lightwave Technol. 11(3), 416–423 (1993). [CrossRef]
31. C. F. Couny, F. Benabid, P. J. Roberts, P. S. Light, and M. G. Raymer, “Generation and photonic guidance of multi-octave optical-frequency combs,” Science 318(5853), 1118–1121 (2007). [CrossRef]
32. B. Debord, A. Amsanpally, M. Chafer, A. Baz, M. Maurel, J. M. Blondy, E. Hugonnot, F. Scol, L. Vincetti, F. Gérôme, and F. Benabid, “Ultralow transmission loss in inhibited-coupling guiding hollow fibers,” Optica 4(2), 209–217 (2017). [CrossRef]
33. P. St. J. Russell, P. Hölzer, W. Chang, A. Abdolvand, and J. C. Travers, “Hollow-core photonic crystal fibres for gas-based nonlinear optics,” Nat. Photonics 8(4), 278–286 (2014). [CrossRef]
34. F. Belli, A. Abdolvand, W. Chang, J. C. Travers, and P. St.J. Russell, “Vacuum-ultraviolet to infrared supercontinuum in hydrogen-filled photonic crystal fiber,” Optica 2(4), 292–300 (2015). [CrossRef]
35. J. C. Travers, T. F. Grigorova, C. Brahms, and F. Belli, “High-energy pulse self-compression and ultraviolet generation through soliton dynamics in hollow capillary fibres,” Nat. Photonics 13(8), 547–554 (2019). [CrossRef]
36. M. Michieletto, J. K. Lyngsø, C. Jakobsen, J. Lægsgaard, O. Bang, and T. T. Alkeskjold, “Hollow-core fibers for high power pulse delivery,” Opt. Express 24(7), 7103–7119 (2016). [CrossRef]
37. F. Köttig, F. Tani, C. M. Biersach, J. C. Travers, and P. St. J. Russell, “Generation of microjoule pulses in the deep ultraviolet at megahertz repetition rates,” Optica 4(10), 1272–1276 (2017). [CrossRef]
38. C. R. Smith, A. Moltke, A. I. Adamu, M. Michieletto, P. Bowen, P. M. Moselund, C. Markos, and O. Bang, “Low-noise tunable deep-ultraviolet supercontinuum laser,” Sci. Rep. 10(1), 18447 (2020). [CrossRef]
39. C. R. Smith, A. I. Adamu, M. Michieletto, and O. Bang, “Spectral broadening of ultraviolet dispersive waves in gas-filled hollow-core fiber using pump pulse modulation,” Opt. Lett. 45(24), 6744–6747 (2020). [CrossRef]
40. A. R. Pennetta, M. T. Enders, M. H. Frosz, F. Tani, and P. St. J. Russell, “Fabrication and non-destructive characterization of tapered single-ring hollow-core photonic crystal fiber,” APL Photonics 4(5), 056105 (2019). [CrossRef]
41. X. Liu, W. Ding, Y. Wang, S. Gao, L. Cao, X. Feng, and P. Wang, “Characterization of a liquid-filled nodeless anti-resonant fiber for biochemical sensing,” Opt. Lett. 42(4), 863–866 (2017). [CrossRef]
42. V. T. Hoang, D. Dobrakowski, G. Stępniewski, R. Kasztelanic, D. Pysz, K. X. Dinh, M. Klimczak, M. Śmietana, and R. Buczyński, “Antiresonant fibers with single- and double-ring capillaries for optofluidic applications,” Opt. Express 28(22), 32483–32498 (2020). [CrossRef]
43. M. Nissen, B. Doherty, J. Hamperl, J. Kobelke, K. Weber, T. Henkel, and M. A. Schmidt, “UV Absorption Spectroscopy in Water-Filled Antiresonant Hollow Core Fibers for Pharmaceutical Detection,” Sensors 18(2), 478 (2018). [CrossRef]
44. J. Yu, Y. Liu, Y. Wang, Z. Wang, X. Zhang, X. Liu, S. Gao, X. Wang, and P. Wang, “Optofluidic laser based on a hollow-core negative-curvature fiber,” Nanophotonics 7(7), 1307–1315 (2018). [CrossRef]
45. C. Wei, J. T. Young, C. R. Menyuk, and J. Hu, “Temperature sensor based on liquid-filled negative curvature optical fibers,” OSA Continuum 2(7), 2123–2130 (2019). [CrossRef]
46. Y. Y. Wang, P. S. Light, and F. Benabid, “Core-Surround Shaping of Hollow-Core Photonic Crystal Fiber Via HF Etching,” IEEE Photonics Technol. Lett. 20(12), 1018–1020 (2008). [CrossRef]
47. K. Nielsen, D. Noordegraaf, T. Sørensen, A. Bjarklev, and T. P. Hansen, “Selective filling of photonic crystal fibers,” J. Opt. A: Pure Appl. Opt. 7(8), L13–L20 (2005). [CrossRef]
48. J. Cheng, J. Wang, J. Hou, H. Wang, and L. Zhang, “Effect of polishing-induced subsurface impurity defects on laser damage resistance of fused silica optics and their removal with HF acid etching,” Appl. Sci. 7(8), 838 (2017). [CrossRef]
49. P. K. Wong, U. Ulmanella, and C. M. Ho, “Fabrication Process of Microsurgical Tools for Single-Cell Trapping and Intracytoplasmic Injection,” J. Microelectromech. Syst. 13(6), 940–946 (2004). [CrossRef]
50. F. Poletti, “Nested antiresonant nodeless hollow core fiber,” Opt. Express 22(20), 23807–23828 (2014). [CrossRef]
51. M. S. Habib, J. E. Antonio-Lopez, C. Markos, A. Schülzgen, and R. Amezcua-Correa, “Single-mode, low loss hollow-core anti-resonant fiber designs,” Opt. Express 27(4), 3824–3836 (2019). [CrossRef]
52. M. Bache, M. S. Habib, C. Markos, and J. Lægsgaard, “Poor-man’s model of hollow-core anti-resonant fibers,” J. Opt. Soc. Am. B 36(1), 69–80 (2019). [CrossRef]
53. H. Sakr, Y. Hong, T. D. Bradley, G. T. Jasion, J. R. Hayes, H. Kim, I. A. Davidson, E. N. Fokoua, Y. Chen, K. R. H. Bottrill, N. Taengnoi, N. V. Wheeler, P. Petropoulos, D. J. Richardson, and F. Poletti, “Interband Short Reach Data Transmission in Ultrawide Bandwidth Hollow Core Fiber,” J. Lightwave Technol. 38(1), 159–165 (2020). [CrossRef]
54. M. S. Habib, C. Markos, and R. Amezcua-Correa, “Impact of cladding elements on the loss performance of hollow-core anti-resonant fibers,” Opt. Express 29(3), 3359–3374 (2021). [CrossRef]
55. J. Neauport, C. Ambard, P. Cormont, N. Darbois, J. Destribats, C. Luitot, and O. Rondeau, “Subsurface damage measurement of ground fused silica parts by HF etching techniques,” Opt. Express 17(22), 20448–20456 (2009). [CrossRef]
56. H. Xiao, H. Wang, G. Fu, and Z. Chen, “Surface roughness and morphology evolution of optical glass with micro-cracks during chemical etching,” Appl. Opt. 56(3), 702–711 (2017). [CrossRef]
57. L. Wong, T. Suratwala, M. D. Feit, P. E. Miller, and R. Steele, “The effect of HF/NH4F etching on the morphology of surface fractures on fused silica,” J. Non-Cryst. Solids 355(13), 797–810 (2009). [CrossRef]
58. M. Pfiffer, P. Cormont, E. Fargin, B. Bousquet, M. Dussauze, S. Lambert, and J. Néauport, “Effects of deep wet etching in HF/HNO3 and KOH solutions on the laser damage resistance and surface quality of fused silica optics at 351 nm,” Opt. Express 25(5), 4607–4620 (2017). [CrossRef]
59. D.G. Lovering, Molten Salt Technology (Springer-Verlag, 1982).
60. H. Schröder, “Glass etching technology based on molten salts,” https://www.izm.fraunhofer.de/content/dam/izm/de/documents/Publikationen/Themenbroschueren/ge/pw_2019_en_RZ_web.pdf.
