Open Access
Add to BibSonomy
Abstract
We present the inscription of narrow-linewidth fiber Bragg gratings (FBGs) into different types of multicore fibers (MCFs) using ultrashort laser pulses and the phase mask technique, which can act as notch filters. Such filters are required, e.g., to suppress light emitted by hydroxyl in the Earth’s upper atmosphere, which disturbs ground-based observation of extraterrestrial objects in the near infrared. However, the inscription into a commercially available seven-core fiber showed a quite large core-to-core deviation of the resonance wavelength of up to 0.45 nm. Two options are presented to overcome this: first, we present the photo-treatment of the FBGs to tune the resonance wavelength, which allows for sufficient resonance shifts. Second, adapted MCFs containing 12 cores, arranged on a circle, are fabricated. For this, two different fabrication procedures were investigated, namely, the mechanical drilling of the preform for a rod-in-tube version as well as a stack-and-draw approach. Both adapted MCFs yielded significant improvements with core-to-core wavelength variations of the FBGs of only about 0.18 nm and 0.11 nm, respectively, sufficient to fulfill the requirements for astronomical filter applications as discussed above.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. INTRODUCTION
The inscription of fiber Bragg gratings (FBGs) into multicore fibers (MCFs) is a hot topic in current research and development. Driven by specific applications, the requirements are very different. For sensing applications, e.g., for bending sensors [1,2], the resonance wavelengths of the different cores do not have to overlap or are even intended to not overlap with each other. Only the shift in the resonance wavelengths during bending or temperature changes is recorded and translated into the corresponding properties. As a filter element, for instance, in Raman spectroscopy, the accuracy of the wavelength matching is not that crucial as long as it provides the required attenuation and covers the pump laser wavelength [3,4]. However, in the case of filtering the hydroxyl emission lines (OH lines) from the Earth’s upper atmosphere, wavelength matching with a small bandwidth is very important. Herein, a large number of narrow OH lines spread across the whole near infrared must be filtered prior to entering the spectrograph [5]. Until now, only Lindley et al. presented a suitable single filter element in the outer six cores of a seven-core fiber [6]. Scaling the number of cores seemed to be very challenging and until now was not successful in terms of optical quality [7].
But why even use MCFs if it seems to be challenging inscribing FBGs into their cores? In applications where light gathering is important as in the aforementioned Raman spectroscopy and OH line suppression, efficient coupling into single mode (SM) fibers is not feasible due to the small core sizes (few micrometers). Therefore, multimode fibers with core diameters of tens of micrometers or even 100 micrometers are used. However, for spectral analysis, SM fibers allow to increase the achievable resolution, as modal interference is ruled out [8,9]. Furthermore, the inscription of FBGs into multimode fibers cannot provide the required single notch, as each guided mode has its different resonance wavelength [10,11]. Therefore, the collected light in the multimode fiber must be split up into SM fibers using a photonic lantern [12]. If the number of cores is larger than or at least equals the number of guided modes in the multimode fiber, this distribution can be performed at low losses. Usually, this results in 10 or more SM fibers. To shrink the device size while keeping the performance parameters, an MCF with embedded SM cores can be used. Additionally, because FBGs are sensitive to strain and temperature, the structured fibers must be stabilized to avoid a drift in the notch filters. This task is dramatically simplified and even more efficient if only a single MCF must be stabilized in contrast to a large number of individual SM fibers.
Several different techniques have been used for inscribing FBGs into MCFs. Using UV lasers requires a photosensitive fiber core. Additionally, in the case of the typical hexagonal core arrangement, the holographic UV inscription technique suffers from shadowing effects of the upper cores with respect to the beam traveling direction and focusing properties of the cylindrical fiber surface [4,6,7]. This focusing of the cylindrical fiber surface can be reduced by inserting the whole fiber in a D-shaped capillary, as demonstrated by Lindley et al. [6]. Still, the intensity reduction with increasing penetration depth due to linear absorption of UV radiation makes it challenging to scale the number of cores with this technique. Another approach focuses the UV laser light into specific fiber cores to avoid this limitation [13]. So far, this localized structuring has been applied only for sensing and not for writing narrowband notch filters with high attenuation.
Another technique for FBG inscription is based on the use of ultrashort laser pulses and its nonlinear absorption inside the focal volume [11,14]. This allows for a very precise structuring of a variety of transparent materials, in particular, optical fibers. Herein, no photosensitivity is required [15]. The point-by-point technique uses a microscope objective to focus the inscribing laser beam into the fiber core. This allows for a tailored FBG inscription, as the location and the period can be freely selected as desired [16]. In this case, the period is determined by the repetition rate of the laser beam in conjunction with the translation speed of the fiber underneath the pulsed laser beam. This can be used to address the individual cores of an MCF with individual FBGs [17]. However, the inscription is based on the generation of small voids inside the fiber core, which causes scattering losses and coupling into cladding and radiation modes [18,19]. This can be avoided using either plane-by-plane inscription, where the focal region gets extended to modify a larger cross section [20], or by the line-by-line technique, where a line is written across the fiber core [21,22]. Both methods use a homogeneous refractive index structuring to reduce scattering losses and due to the larger core coverage, avoiding coupling into radiation and cladding modes. However, the alignment is very critical, and in the case of the line-by-line technique, the inscription duration is very long.
The method of applying a phase mask to produce the periodic refractive index pattern enables a simplified material structuring of the fiber core, as the interference pattern is fixed by the phase mask [23,24]. However, the stability and reproducibility come at the expense of a fixed period. Different techniques exist to overcome this restriction; here, we mention this only for completeness, as they also come with some disadvantages [24–27]. Inscribing FBGs into an MCF using ultrashort laser pulses has been presented before, but the results were not satisfactory [7,28]. The reason is slight manufacturing deviations of the different cores of the MCF resulting in inequality of the Bragg resonance, as the effective refractive index differs from core to core. The origin of the resonance wavelength variation is not yet fully understood and has probably many origins [29]. However, one major influence is the preform used, as any inhomogeneity translates by the scale down ratio into the fiber. Therefore, the individual core materials that form the fiber cores later must be chosen carefully. Another important influence is the arrangement of cores within the fiber. Lindley et al. showed that if they consider only the outer six cores of their seven-core fiber, the resonances overlap to a certain value [6]. Most likely, the inner core sees a different stress field due to the fabrication process, and/or the surrounding cores cause a different environment in comparison to the outer cores. However, no further progress has been published since 2014 for either scaling towards more cores or more notch filters per core.
For application in astrophotonic filtering, the targeted OH lines have a bandwidth around 0.005 nm with a broad background ranging up to 0.1 nm, due to their Lorentzian shape [30]. Therefore, the requirements for a notch filter are a bandwidth around 0.2 nm to 0.3 nm and a suppression of 20 dB to 30 dB [30]. The wavelength matching must assure full coverage of the OH line. Thus, having a bit broader resonance above 0.2 nm relaxes the target wavelength accuracy to about 0.1 nm.
In this work, we present the inscription of FBGs into MCFs and discuss the different aspects of the spectral properties realized. We focus our investigations on the resonance wavelength variation between the different cores for single-resonance FBGs, as this restricts the application as a notch filter in general. In the upcoming section, we describe the inscription of FBGs into a seven-core fiber by applying ultrashort laser pulses and the phase mask technique. During our experiments, we faced the issue of strong core-to-core variations of the resonance wavelength. To overcome this, we demonstrate two possibilities: first is the post-processing of the written FBGs in an individual single core of the seven-core MCF, tuning the resonance towards longer wavelengths. Second, we present the fabrication of an innovative multicore design that provides almost identical fiber core properties by arranging the cores on a circle centered inside the fiber. This should therefore ensure identical surroundings for all cores as well as negligible cross talk between the cores. The number of cores in these fibers was almost doubled, as we increased it from seven to 12. For demonstration, two 12-core fibers were realized using two different fabrication technologies. One fiber was fabricated by the traditional stacking technology, referred to as “stacked MCF” in the following. For the second fiber, 12 holes were drilled into a glass rod and subsequently filled with prepared core material. This fiber will be referred to as “drilled MCF” in the following. These new fibers are then used to inscribe homogeneous FBGs into all cores.
2. INSCRIPTION OF FBG INTO AN MCF
The inscription of the FBGs is performed using a titanium–sapphire laser (Spectra Physics, Spitfire ACE) delivering 100 fs pulses at a repetition rate of 1 kHz, centered at a wavelength of 800 nm. The laser beam has a Gaussian shape with a diameter of 10 mm (determined at $1/{e^2}$). To achieve a localized refractive index modification inside the desired fiber core, we use an aspherical cylindrical lens (asphericon GmbH) with a focal length of 12 mm (which correlates to a numerical aperture of about 0.38). To avoid issues with the cylindrical fiber surface, we rotate the fiber in such a way that the targeted core is always at the top. For this purpose, we use high-precision air-bearing rotation stages to minimize the remaining fiber movement. These rotation axes are mounted onto a 2D translation stage (Aerotech GmbH, PlanarDL200, ANT130LZ, ABRS150MP). We move this stage including fiber and phase mask under the laser beam to scan perpendicular to the fiber axis across the fiber core covering the whole core. Next, the fiber is rotated to address the following core accordingly. The rotation of the fiber can be aligned with an accuracy of ${\pm}{1^ \circ}$, which enables precise structuring of the individual cores. Be aware that this accuracy comes from the alignment procedure; the precision of the rotation axes is well below (${\pm}3 \;{\rm arcsec}$). The inscription scheme is depicted in Fig. 1. We choose to process the upper core because once the rotation goes beyond 180°, we would disturb our inscription with the already modified cores if we inscribe into the lower core. For all inscriptions, we removed the coating prior to structuring. Note that this is not required [31,32], but we wanted to avoid any potential influence arising from inhomogeneous coating sections. The phase masks used have a thickness of 1 mm and are all in-house fabricated by electron beam lithography. These masks provide periods of around 1070 nm resulting in first-order FBGs in the range of 1500 nm to 1600 nm and are optimized to diffract light into the ${\pm}{1}$st order.
Fig. 1. Inscription strategy for the MCF. The laser focus scans across the upper core (1 and 2). Afterwards, the fiber is rotated (3) and the inscription is repeated (4) until all cores are structured (5).
The transmission spectra are obtained by probing the FBG with a super-continuum white light source (NKT Photonics, SuperK) and detecting the transmission spectrum by an optical spectrum analyzer (OSA) (Yokogawa AQ6375). The OSA has a resolution of 20 pm and a sampling of 4 pm. The light is coupled into a single core of the MCF using butt-coupling of a single core fiber with the MCF. We detect the transmission signal of an individual fiber core using free-space out-coupling and coupling into a single core and SM fiber for transporting the light to the spectrometer. The free-space coupling is required to select the light from a single core for analysis. The reflection spectra are measured by an interrogator (MicronOptics sm125), which contains a sweeping laser source and a detector providing a resolution below 5 pm. The reflection was measured only for the seven-core fiber, as here a fan-out was available, allowing for a simple detection scheme.
3. RESULTS FOR THE SEVEN-CORE FIBER
For the first inscription tests, we chose a commercially available MCF (fibercore, SM-7C1500) containing six cores with a core-to-core spacing of 35 µm arranged in a hexagon and an additional central core. The cores have a diameter of 6.1 µm and are designed to guide the light in an SM from 1520 to 1650 nm (${\rm NA} = 0.21$). The cladding has a diameter of 125 µm.
A. Inscribing the FBGs
For the inscription, we applied the technique described above, scanning across the fiber core with 0.5 mm/min over $20 \;{\unicode{x00B5}{\rm m}}$ and rotating the fiber by 60° to address the following core. The phase mask in this case had a period of 1050 nm. For this fiber, a fan-out was available, enabling us to measure the reflection spectrum of the individual cores directly. The transmission signal was obtained using the free-space coupling scheme described in Section 2.
Figure 2 shows a transmission measurement of the FBGs in the outer six cores. The modulation in the transmission spectra originates from the white light source, which varies in its intensity over time. The resonances are nevertheless clearly visible. In these first experiments, only the outer six cores were addressed. The wavelength variation across the different cores amounts to about 0.45 nm, which correlates to an effective refractive index difference of the guided modes of about $4.3 \cdot {10^{- 4}}$. This variation is due to inhomogeneity of the fiber and not due to the FBG inscription, since it is reproducible no matter which core we start the inscription with.
It is interesting to note that when we inscribe an FBG also in the central core, its resonance lies spectrally between the resonances of the outer cores and is not separated as observed by Lindley et al. [6]. However, one has to be aware that in our case, the spread of the resonance wavelengths is much larger than those of Lindley et al.
To summarize, the wavelength variation of this seven-core fiber is too large for direct application in OH filtering. Having in mind that we would like to fabricate narrow-bandwidth filters at the same wavelength in all cores of an MCF, we have to overcome this spread of the individual cores either by applying a technique to correct for the wavelength variation or by using a more homogeneous fiber. The first solution option is addressed in the next section.
B. Tuning the Resonance of an Individual Core
Tuning of the resonance wavelength after inscription of the FBG can be done by post-processing. The variation in the effective refractive index of $4.3 \cdot {10^{- 4}}$ is lower than what has been demonstrated by different groups using photo-treatment of the inscribed FBG. The photo-treatment of FBGs has been demonstrated with UV and femtosecond pulsed radiation [33,34]. The post-processing with UV radiation was used for various applications such as creating a phase-shift inside the FBG or for optimizing the dispersion of chirped FBGs [35]. Recently, ultrashort pulsed radiation has been applied to tune the resonance wavelength of FBGs [36].
The post-processing is similar to writing the FBG but without the phase mask, to produce a constant refractive index change. Figure 3 sketches the evolution of the refractive index change and the saturation effect occurring. The reason for the saturation is the limited refractive index change that can be realized in fused silica. The saturation will lead to an increase in the transmission at the resonance, which weakens the filter application. However, providing a pre-compensation, for instance, by inscribing longer gratings, the final attenuation at the resonance ends up in an acceptable range. For a strong FBG with a transmission of around ${-}20\; {\rm dB}$, we measured a transmission increase of about 4 dB for a wavelength shift of 0.255 nm [28]. In this range, the transmission increase scales roughly linearly with the tuning of the resonance with a slope of 15.9 dB/nm. However, these results were obtained in a single core fiber.
Fig. 3. Sketch of the photo-treatment of an FBG and the occurring saturation effect, which lowers the reflectivity. (a) Refractive index profile of the periodic FBG structure; (b), (c) increase in the effective refractive index in the FBG region and the increasing saturation effect during the photo-treatment.
Here, we apply a similar strategy of post-processing to our seven-core fiber to increase the resonance wavelength. Because we aim for tuning individual cores of the seven-core fiber, we use the same high numerical aperture lens with 12 mm focal length as before for the FBG inscription. Initially, we inscribe FBGs into all cores of the MCF (parameters: phase mask period of 1057 nm, pulse energy of 200 µJ, repetition rate of 100 Hz, scanning speed of 0.1 mm/min; several scans were performed for equal grating strength). After measuring the reflection of the gratings, we addressed core #1 for post-processing. Note that this choice is arbitrary since all outer cores are equal for post-processing purposes. The post-processing was performed with the same repetition rate at a pulse energy of 230 µJ. The higher pulse energy for post-processing is required because the interference pattern of the phase mask increases the local intensity inside the fringes of constructive interference, which is no longer the case when the phase mask is removed.
A comparison of the inscribed grating before (blue) and after post-processing (red) is shown in Fig. 4. For tuning the center wavelength, we achieved a shift of 0.44 nm (correlates to an index change of $4.2 \cdot {10^{- 4}}$). Additionally, the other cores stayed unaffected. The reflection spectrum of the modified core shows the appearance of slight modulations after the post-processing. The origin of the modulation is the Gaussian beam profile used for post-processing, resulting in a Gaussian shaped constant refractive index change on top of the periodic refractive index modulation. However, the influence on the filter performance is negligible as it is more than 10 dB weaker than the resonance.
The large variation in resonance wavelengths already present for the outer cores requires a significant resonance shift to achieve a matching notch filter. This becomes impractical when we want to inscribe multiple FBGs into the fiber. In addition, it is important to note that the central core cannot easily be post-processed, as the inscribed refractive index modulations in the outer cores disturb the laser beam. Thus, we decided to progress with an MCF with an adapted geometry to ease the FBG inscription process. Additionally, this new fiber design should provide more equal effective refractive indices across the different cores for better wavelength matching of the FBGs to avoid the need for post-processing. The design and realization of this new MCF is described in the next section.
4. FABRICATION OF THE MULTICORE FIBER
The first step towards an MCF with equal core properties is the arrangement of the cores within the fiber. We decided to distribute the cores on a circle, because this enables an equal surrounding for all cores during all processes. For example, the potential stress fields arising from the fiber drawing process should be circularly symmetric and thus equal for all cores. The design thus consists of 12 cores equally distributed on a circle. This enables relatively large core-to-core distances of ${\gt}30 \;{\unicode{x00B5}{\rm m}}$ to prevent cross talk between neighboring cores as well as an outer barrier of ${\gt}30 \;{\unicode{x00B5}{\rm m}}$ for a total outer diameter of ${\lt}250 \;{\unicode{x00B5}{\rm m}}$. The resulting fiber design is shown in Fig. 5, and the corresponding dimensions are listed in Table 1.
Fig. 5. Sketch of the fiber design. The green, gray, and blue circles represent the cores, cladding, and coating, respectively. The fiber dimensions are listed in Table 1.
Table 1. Fiber Dimensions for Drilled and Stacked MCFs
For fiber fabrication, we followed two different approaches, based on drilling and on “stack and draw.” The drilling method is a subtractive procedure where holes are drilled into a fused silica rod and afterwards filled with preforms forming the final cores. The stack and draw method is an additive manufacturing procedure, where the preform is built from rods and tubes. As core material commercially available Ge-doped fused silica with an NA of 0.15 was chosen for both fibers. This material is typically used as preforms for fibers in telecom applications. The core material was pre-characterized in terms of refractive index and diameter. Sections with reduced variations were selected to increase the homogeneity of the individual cores in the final fibers. Next, both procedures for preform fabrication and fiber drawing are discussed.
A. Subtractive Preform Fabrication—Drilling
First, a fiber according to the design in Fig. 5 was realized with the drilling method as described in the following. A macroscopic preform with linearly up-scaled fiber dimensions was manufactured by drilling 12 holes, each having a diameter of 3 mm, into a silica rod (diameter of 25 mm) using a deep hole drilling machine, in a ring geometry. A photograph of the drilled preform is shown in Fig. 6. After drilling, mechanical and chemical residuals were removed in several cleaning steps, including purging with hydrogen fluoride, to ensure contamination free surfaces, before elongating the structure at the drawing tower, resulting in an outer diameter of 9.5 mm [Fig. 6(b)]. The holes, now 1 mm in diameter, were filled with resized rods of the preselected telecom grade Ge-doped preform. From this assembly, the final fiber was drawn. The final fiber geometry is listed in Table 1. In the following, we refer to this fiber as drilled MCF.
Fig. 6. (a) Photograph of the drilled structure composed of 12 holes (Ø 3 mm) located on a circle (Ø 18.54 mm) within the fused silica rod. (b) Microscope image after stretching to preform dimensions.
B. Stack and Draw
As an alternative to the drilled fiber, we also fabricated a second fiber by stack and draw. Again, a macroscopic preform was manufactured. This time it was realized by stacking together differently sized rods according to the sketch in Fig. 7. The orange rods in Fig. 7 illustrate the 12 resized telecom grade preforms including the Ge-doped cores. After cleaning all parts, the stack was built inside a jacket tube. Then the stack was collapsed to a solid preform by heat treatment and finally drawn to fiber. Table 1 gives an overview of the geometric dimensions obtained. This fiber has a cladding diameter of $220\; {\unicode{x00B5}{\rm m}}$, which is $20 \;{\unicode{x00B5}{\rm m}}$ smaller than the drilled fiber. However, this has negligible influence on the final FBG inscription and operation. In the following, we refer to this fiber as stacked MCF.
Fig. 7. Layout for the stacked MCF. All filled rods are pure fused silica, except for the orange ones where the Ge-doped core material is included.
5. RESULTS FOR THE 12-CORE FIBERS
Next, we present and compare the properties of the FBGs inscribed into the two 12-core MCFs. Here, the same technique is applied as presented in Section 2 and shown in Fig. 1, where the focal spot is scanned across the individual fiber cores with rotations of just 30° to address all 12 cores instead of only the outer six cores as before. As presented in the previous section, the cores are arranged on a circle, which ensures that the core is located always at the same position when the fiber is rotated. For evaluation and for comparison, we inscribe similar gratings into all cores and scan only perpendicular to the fiber axis across the fiber cores to ensure a homogeneous inscription. We evaluate the resonance wavelengths and the correlated effective refractive indices of the fiber cores to analyze the achieved homogeneity of the 12-core MCF. Therefore, we monitor the transmission using butt-coupling for the input of the light into the fiber and a free-space coupling for output and light transport to the spectrometer. Again, the super-continuum white light source was used for characterization.
A. FBG Inscription into the Drilled MCF
The inscription into the drilled MCF was performed at a repetition rate of 1 kHz with a pulse energy of 250 µJ. The phase mask had a period of 1074 nm. The cores were scanned multiple times with 1 mm/min over a length of $30 \; {\unicode{x00B5}{\rm m}}$ across the fiber core to achieve equal transmission values for the resonances. The transmission spectra of all cores are depicted in Fig. 8. The inscribed FBGs show a maximum wavelength variation of 0.18 nm between the lowest and highest resonance wavelength, which correlates to a maximum effective refractive index variation of $1.7 \cdot {10^{- 4}}$ between all cores. Because the transmission spectra have a slight modulation when they filter more than 10 dB, we considered the notch center for evaluating the resonance wavelength and not the minimum transmission. Table 2 lists the FBG properties (center wavelengths, bandwidths, and transmission values). The gratings have a transmission loss at the resonance of about ${-}20\; {\rm dB}$ and are centered around 1556 nm. The bandwidths are rather large with about 1.0 nm due to the short grating length inside the fiber core. The grating length and index change are evaluated based on equations considering the measured bandwidth and the transmission value [37]. The FBGs have an approximate length of 2.3 mm, and the modulated refractive index change is about $6.5 \cdot {10^{- 4}}$. The broadband losses are difficult to estimate due to slow intensity fluctuations of the white light source. However, a rough estimation assumes that the losses must be below 0.2 dB/cm. Loss measurements of FBGs inscribed into a standard fiber yielded values of around 0.08 dB/cm for similar refractive index changes [38].
Table 2. Properties of FBGs Inscribed into the Cores of the Drilled MCFa
B. FBG Inscription into the Stacked MCF
Next, we inscribed similar FBGs into the 12-core stacked MCF. We use again a repetition rate of 1 kHz but a reduced pulse energy of $220\; {\unicode{x00B5}{\rm J}}$. Due to the smaller fiber diameter, the focal spot gets smaller, and thus less pulse energy is required to achieve similar structuring. The phase mask had a period of 1081 nm. The cores are modified by scanning multiple times over a range of $30 \;{\unicode{x00B5}{\rm m}}$ with a velocity of 1 mm/min to realize equal transmission values.
Again, we evaluated the resonance wavelengths at the notch center. The measured transmission spectra are shown in Fig. 9, and the results are listed in Table 3. The resonances are centered around 1565.8 nm and have a bandwidth in the range of 1.1 nm. The resonance wavelength variation is even lower in comparison to the drilled MCF and measures 0.11 nm, which correlates to a maximum effective refractive index variation across all cores of about $1.0 \cdot {10^{- 4}}$. As before, the resonances have a transmission value of around ${-}20\; {\rm dB}$, which results from an induced refractive index change of around $7.2 \cdot {10^{- 4}}$ and a structure length of around 2.1 mm.
Fig. 9. Transmission spectra of the 12-core stacked MCF. The slight modulation around the resonances is due to intensity variations of the white light source.
Table 3. Values of FBGs Inscribed into the Cores of the Stacked MCFa
C. Comparison of the Two 12-Core MCFs
The comparison of the spectral properties of both 12-core fibers shows no significant difference. Both fibers show minimal cladding mode excitation by the FBGs, and the resonance itself has a clear strong attenuation based on its Gaussian modification profile. However, if we compare the core-to-core resonance differences, the stacked fiber has lower variations in the resonance wavelengths and thus is better suited for filter application.
For the inscription of the FBG, we first had to find the right orientation for addressing the cores. When observing the fiber during rotation, the drilled MCF was much easier to align in comparison to the stacked MCF. Our guess is that because the cladding of the drilled MCF is one full block of glass, it refracts the light only when passing through the cores. For the stack and draw technique, several rods are used that—even if not intentionally doped—refract the light partially and thus cause a similar pattern to the rods with Ge doping. Here we conclude that the drilled MCF is easier to align.
D. Extended FBG Lengths for Lower Bandwidths
Until now, we have investigated the performance of the different MCFs. Because we always performed only scans across the fiber cores but did not extend the grating lengths, the bandwidths were always relatively large for the purpose of our application. As we presented in the Introduction, filtering of OH lines requires a bandwidth in the range of 0.2 nm [30].
As the stacked MCF provided the lowest resonance wavelength variation, we continued with this fiber to inscribe FBGs with reduced bandwidths. Reduction of the bandwidth while maintaining high suppression can be achieved only by extending the length of the gratings [37]. Therefore, we reduced the pulse energy to $170\; {\unicode{x00B5}{\rm J}}$ and performed single scans across the fiber core at different positions along the fiber axis. Each scan with a velocity of 1 mm/min across the fiber core was $30 \; {\unicode{x00B5}{\rm m}}$ long. Each step along the fiber axis was spaced by $140 \; {\unicode{x00B5}{\rm m}}$, and a total number of 101 steps were performed, resulting in a total length of 14 mm. Including the approximately 2 mm modification length induced by each scan across the fiber, we ended up with a total grating length of 16 mm. The induced refractive index change ranged from $1.3 \cdot {10^{- 4}}$ for ${-}30\; {\rm dB}$ transmission and increased to roughly $2.2 \cdot {10^{- 4}}$ for the strongest resonance #11. In this case, a phase mask with a period of 1088 nm was used.
The measured transmission spectra are shown in Fig. 10, and Table 4 lists the parameters. The resonances vary 0.12 nm in their resonance wavelengths. The slightly larger variation in comparison to previous results might be caused due to a stronger variation of the grating strengths. The wavelength variation translates into an effective refractive index difference of $1.1 \cdot {10^{- 4}}$. The bandwidths are lower than 0.38 nm, and the transmission exceeds ${-}30\; {\rm dB}$. The parameters vary slightly because the fiber cores are not always structured identically. Due to the short focal length, the Rayleigh length ($2{z_0} \approx 3.3 \; {\unicode{x00B5}{\rm m}}$), and thus the modification extension in the beam propagation direction, is smaller than the core size. Therefore, a minimal variation in the focal position causes a reduction in the overlap of the modification with the fiber core, resulting in a reduced coupling strength. Additionally, the slight offset leads to weak cladding mode coupling, which occurs around 1573 nm. For the strongest resonances (in particular #6 and #11), we could not measure the exact transmission value due to noise limitations. Therefore, these gratings appear as though having a flat resonance, which is most likely not the case.
Fig. 10. Depicted are the transmission spectra of the FBGs in all 12 cores of the stacked MCF. The upper graph shows the wide range of transmission, while the lower one is a close-up demonstrating the low bandwidths and strong attenuation.
6. CONCLUSION
We demonstrated the inscription of FBGs for filtering applications into different MCFs. Initially, we used a commercially available seven-core fiber that showed quite large variations in the effective refractive index across the different cores ($\Delta n = 4.3 \cdot {10^{- 4}}$). To reduce this variation, we demonstrated a post-processing technique by photo-treatment of individual cores of the MCF. We tuned the resonance wavelength up to 0.44 nm towards longer wavelengths by inducing an additional constant refractive index change to only a single core while keeping the other cores unaffected.
To reduce the refractive index variation across the cores and improve the inscription process, we designed and manufactured optimized MCFs. These MCFs contained 12 cores equally distributed on a circle. We demonstrated two different techniques of producing those MCFs to ensure a successful fabrication with low index variation. The first contains a rod-like cladding precursor, and the core holes are drilled into that bulk material. The core material is inserted into the holes and finally the fiber is drawn. The second method uses the commonly known stack and draw technique. For both fibers, commercially available Ge-doped silica was used as the core material, which was preselected in terms of minimized refractive index and diameter variations. This preselection procedure in combination with the distribution of the cores on a circle ensures a similar surrounding and enables reproducible FBG inscription. Here, we demonstrated that the lowest variation in the effective refractive index is only about $1.0 \cdot {10^{- 4}}$. This allowed us to inscribe narrow FBGs into that fiber providing a suitable filter element. The results demonstrate that both fiber fabrication techniques, namely, drilling and stacking, deliver comparable results. To the best of our knowledge, this is the first direct comparison of the drilling technique and the common benchmark stacking technology.
The next step will consist of the inscription of multiple FBGs into the fiber to scale the number of filtered resonances to match the targeted application as OH notch filters for astronomy.
Funding
Bundesministerium für Bildung und Forschung (03Z1H534); Deutsche Forschungsgemeinschaft (GRK 2101, NO 462/18-1); European Regional Development Fund (2017 FGI 0014, EffiMOF); European Social Fund (2018 FGR 0099, RATI); Freistaat Thüringen (2018 FGR 0099, RATI).
Acknowledgment
The authors thank Uwe D. Zeitner, Thomas Siefke, Markus Walther, and Martin Heusinger for stimulating discussions on phase mask designs and manufacturing of excellent phase masks. T. A. Goebel acknowledges support from the IMPRS-PL within the international Ph.D. program.
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. M. Gander, W. MacPherson, R. McBride, J. Jones, L. Zhang, I. Bennion, P. Blanchard, J. Burnett, and A. Greenaway, “Bend measurement using Bragg gratings in multicore fibre,” Electron. Lett. 36, 120 (2000). [CrossRef]
2. H. Zhang, Z. Wu, P. P. Shum, R. Wang, X. Q. Dinh, S. Fu, W. Tong, and M. Tang, “Fiber Bragg gratings in heterogeneous multicore fiber for directional bending sensing,” J. Opt. 18, 085705 (2016). [CrossRef]
3. S. Dochow, I. Latka, M. Becker, R. Spittel, J. Kobelke, K. Schuster, A. Graf, S. Brückner, S. Unger, M. Rothhardt, B. Dietzek, C. Krafft, and J. Popp, “Multicore fiber with integrated fiber Bragg gratings for background-free Raman sensing,” Opt. Express 20, 20156–20169 (2012). [CrossRef]
4. M. Becker, A. Lorenz, T. Elsmann, I. Latka, A. Schwuchow, S. Dochow, R. Spittel, J. Kobelke, J. Bierlich, K. Schuster, M. Rothhardt, and H. Bartelt, “Single-mode multicore fibers with integrated Bragg filters,” J. Lightwave Technol. 34, 4572–4578 (2016). [CrossRef]
5. J. Bland-Hawthorn, S. C. Ellis, S. G. Leon-Saval, R. Haynes, M. M. Roth, H.-G. Löhmannsröben, A. J. Horton, J.-G. Cuby, T. A. Birks, J. S. Lawrence, P. Gillingham, S. D. Ryder, and C. Q. Trinh, “A complex multi-notch astronomical filter to suppress the bright infrared sky,” Nat. Commun. 2, 581 (2011). [CrossRef]
6. E. Lindley, S.-S. Min, S. G. Leon-Saval, N. Cvetojevic, J. S. Lawrence, S. C. Ellis, and J. Bland-Hawthorn, “Demonstration of uniform multicore fiber Bragg gratings,” Opt. Express 22, 31575–31581 (2014). [CrossRef]
7. R. Haynes, T. A. Birks, J. Bland-Hawthorn, J. L. Cruz, A. Diez, S. C. Ellis, D. Haynes, R. G. Krämer, B. J. Mangan, S.-S. Min, D. F. Murphy, S. Nolte, J. C. Olaya, J. U. Thomas, C. Q. Trinh, A. Tünnermann, and C. Voigtländer, “Second generation OH suppression filters using multicore fibers,” Proc. SPIE 8450, 845011 (2012). [CrossRef]
8. J. Bland-Hawthorn and P. Kern, “Molding the flow of light: photonics in astronomy,” Phys. Today 65(5), 31–37 (2012). [CrossRef]
9. N. Cvetojevic, N. Jovanovic, S. Gross, B. Norris, I. Spaleniak, C. Schwab, M. J. Withford, M. Ireland, P. Tuthill, O. Guyon, F. Martinache, and J. S. Lawrence, “Modal noise in an integrated photonic lantern fed diffraction-limited spectrograph,” Opt. Express 25, 25546–25565 (2017). [CrossRef]
10. C. Lu and Y. Cui, “Fiber Bragg grating spectra in multimode optical fibers,” J. Lightwave Technol. 24, 598–604 (2006). [CrossRef]
11. J. U. Thomas, C. Voigtländer, R. G. Becker, D. Richter, A. Tünnermann, and S. Nolte, “Femtosecond pulse written fiber gratings: a new avenue to integrated fiber technology,” Laser Photon. Rev. 6, 709–723 (2012). [CrossRef]
12. S. G. Leon-Saval, T. A. Birks, J. Bland-Hawthorn, and M. Englund, “Multimode fiber devices with single-mode performance,” Opt. Lett. 30, 2545–2547 (2005). [CrossRef]
13. S. L. Jantzen, R. H. S. Bannerman, A. Jantzen, P. L. Mennea, D. H. Smith, J. C. Gates, L. J. Boyd, P. G. R. Smith, and C. Holmes, “Individual inscription of spectrally multiplexed Bragg gratings in optical multicore fibers using small spot direct UV writing,” Opt. Express 28, 21300–21309 (2020). [CrossRef]
14. K. Itoh, W. Watanabe, S. Nolte, and C. B. Schaffer, “Ultrafast processes for bulk modification of transparent materials,” MRS Bull. 31(8), 620–625 (2006). [CrossRef]
15. S. J. Mihailov, C. W. Smelser, D. Grobnic, R. B. Walker, P. Lu, H. Ding, and J. Unruh, “Bragg gratings written in all-SiO2 and Ge-doped core fibers with 800-nm femtosecond radiation and a phase mask,” J. Lightwave Technol. 22, 94–100 (2004). [CrossRef]
16. A. Martinez, M. Dubov, I. Y. Khrushchev, and I. Bennion, “Direct writing of fibre Bragg gratings by femtosecond laser,” Electron. Lett. 40, 1170–1172 (2004). [CrossRef]
17. A. Wolf, A. Dostovalov, K. Bronnikov, and S. A. Babin, “Arrays of fiber Bragg gratings selectively inscribed in different cores of 7-core spun optical fiber by IR femtosecond laser pulses,” Opt. Express 27, 13978–13990 (2019). [CrossRef]
18. M. L. Åslund, N. Jovanovic, N. Groothoff, J. Canning, G. D. Marshall, S. D. Jackson, A. Fuerbach, and M. J. Withford, “Optical loss mechanisms in femtosecond laser-written point-by-point fibre Bragg gratings,” Opt. Express 16, 14248–14254 (2008). [CrossRef]
19. J. U. Thomas, N. Jovanovic, R. G. Becker, G. D. Marshall, M. J. Withford, A. Tünnermann, S. Nolte, and M. J. Steel, “Cladding mode coupling in highly localized fiber Bragg gratings: modal properties and transmission spectra,” Opt. Express 19, 325–341 (2011). [CrossRef]
20. A. Theodosiou, A. Lacraz, M. Polis, K. Kalli, M. Tsangari, A. Stassis, and M. Komodromos, “Modified fs-laser inscribed FBG array for rapid mode shape capture of free-free vibrating beams,” IEEE Photon. Technol. Lett. 28, 1509–1512 (2016). [CrossRef]
21. K. Zhou, M. Dubov, C. Mou, L. Zhang, V. K. Mezentsev, and I. Bennion, “Line-by-line fiber Bragg grating made by femtosecond laser,” IEEE Photon. Technol. Lett. 22, 1190–1192 (2010). [CrossRef]
22. R. J. Williams, R. G. Krämer, S. Nolte, M. J. Withford, and M. J. Steel, “Detuning in apodized point-by-point fiber Bragg gratings: insights into the grating morphology,” Opt. Express 21, 26854–26867 (2013). [CrossRef]
23. D. Grobnic, C. W. Smelser, S. J. Mihailov, R. B. Walker, and P. Lu, “Fiber Bragg gratings with suppressed cladding modes made in SMF-28 with a femtosecond IR laser and a phase mask,” IEEE Photon. Technol. Lett. 16, 1864–1866 (2004). [CrossRef]
24. C. Voigtländer, R. G. Becker, J. U. Thomas, D. Richter, A. Singh, A. Tünnermann, and S. Nolte, “Ultrashort pulse inscription of tailored fiber Bragg gratings with a phase mask and a deformed wavefront [invited],” Opt. Mater. Express 1, 633–642 (2011). [CrossRef]
25. J. D. Prohaska, E. Snitzer, S. Rishton, and V. Boegli, “Magnification of mask fabricated fibre Bragg gratings,” Electron. Lett. 29, 1614–1615 (1993). [CrossRef]
26. T. A. Goebel, C. Voigtländer, R. G. Krämer, C. Matzdorf, M. Heck, D. Richter, A. Tünnermann, and S. Nolte, “Wavelength tuning of through-coating-written fiber Bragg gratings,” in Conference on Lasers and Electro-Optics (2017), paper CM-4-5.
27. J. Habel, T. Boilard, J.-S. Frenière, F. Trépanier, and M. Bernier, “Femtosecond FBG written through the coating for sensing applications,” Sensors 17, 2519 (2017). [CrossRef]
28. T. A. Goebel, M. Weissflog, R. G. Krämer, M. Heck, D. Richter, and S. Nolte, “Tuning multichannel filters based on FBG in multicore fibers,” Proc. SPIE 10908, 109080O (2019). [CrossRef]
29. S. C. Ellis, S.-S. Min, S. G. Leon-Saval, and J. Bland-Hawthorn, “On the origin of core-to-core variations in multi-core fibre Bragg gratings,” Proc. SPIE 10706, 174 (2018). [CrossRef]
30. S. C. Ellis and J. Bland-Hawthorn, “The case for OH suppression at near-infrared wavelengths,” Mon. Not. R. Astron. Soc. 386, 47–64 (2008). [CrossRef]
31. D. Grobnic, S. J. Mihailov, C. W. Smelser, and R. T. Ramos, “Ultrafast IR laser writing of strong Bragg gratings through the coating of high Ge-doped optical fibers,” IEEE Photon. Technol. Lett. 20, 973–975 (2008). [CrossRef]
32. M. Bernier, F. Trépanier, J. Carrier, and R. Vallée, “High mechanical strength fiber Bragg gratings made with infrared femtosecond pulses and a phase mask,” Opt. Lett. 39, 3646–3649 (2014). [CrossRef]
33. J. Canning and M. Sceats, “π-phase-shifted periodic distributed structures in optical fibres by UV post-processing,” Electron. Lett. 30, 1344–1345 (1994). [CrossRef]
34. A. Shamir and A. A. Ishaaya, “Effect of femtosecond photo-treatment on inscription of fiber Bragg gratings,” Opt. Lett. 41, 765–768 (2016). [CrossRef]
35. M. Sumetsky, P. I. Reyes, P. S. Westbrook, N. M. Litchinitser, B. J. Eggleton, Y. Li, R. Deshmukh, and C. Soccolich, “Group-delay ripple correction in chirped fiber Bragg gratings,” Opt. Lett. 28, 777–779 (2003). [CrossRef]
36. A. Halstuch and A. A. Ishaaya, “Characterizing the effect of femtosecond photo-treatment on the center wavelength of fiber Bragg gratings,” Opt. Express 26, 18990–18997 (2018). [CrossRef]
37. T. Erdogan, “Fiber grating spectra,” J. Lightwave Technol. 15, 1277–1294 (1997). [CrossRef]
38. T. A. Goebel, M. Heusinger, R. G. Krämer, C. Matzdorf, T. O. Imogore, D. Richter, U. D. Zeitner, and S. Nolte, “Femtosecond inscription of semi-aperiodic multi-notch fiber Bragg gratings using a phase mask,” Opt. Express 28, 35682–35694 (2020). [CrossRef]
