Optical fiber with a shaped photosensitivity profile for producing structures with photoinduced modulation of refractive index, in particular Bragg gratings

Abstract

The present disclosure relates to an optical fiber with shaped photosensitivity profile, comprising a nanostructured core composed of at least two types of glass rods, wherein at least one type of glass rods is doped with germanium. The invention relates also to a method for preparing an optical fiber with a core allowing for obtaining photoinduced refractive index modulation. Depending on their specific type, such optical fibers are applicable i.a. in laser generation and in amplification techniques (active optical fibers) and/or in optical fiber sensors and telecommunications applications (passive optical fibers).

Claims

1. An optical fiber with a shaped photosensitivity profile, adapted to guide and generate radiation with a wavelength A and to induce refractive index modulation, said fiber being provided with a cladding and a nanostructured core composed of longitudinal glass elements oriented along the fiber and forming a compact bundle, wherein transverse dimensions of the longitudinal elements are smaller than the wavelength , and the core is composed of at least two types of longitudinal elements differing in refractive index, characterised in that the longitudinal elements of the first type are made of GeO.sub.2-doped silica glass, and the longitudinal elements of the second type are made of pure silica.

2. The optical fiber according to claim 1, characterised in that the longitudinal elements of the third type are made of glass containing at least one active dopant.

3. The optical fiber according to claim 2, characterised in that the active dopant is selected from erbium, praseodymium, ytterbium, neodymium, thulium, holmium and others.

4. The optical fiber according to claim 1, characterised in that the refractive index of the longitudinal elements of one type has a value lower than or equal to the lowest value of the refractive index characteristics in the core cross-section, and the refractive index of the longitudinal elements of another type has a value higher than or equal to the highest value of the refractive index characteristics in the core cross-section.

5. The optical fiber according to claim 2, characterised in that it has a photonic cladding.

6. The optical fiber according to claim 1, characterised in that it is a birefringent optical fiber.

7. The optical fiber according to claim 1, characterised in that the transverse dimensions of the longitudinal elements are smaller than of the wavelength .

8. The optical fiber according to claim 1, characterised in that a structure with photoinduced modulation of refractive index is applied onto the longitudinal elements of the first type.

9. The optical fiber according to claim 8, characterised in that the structure with photoinduced modulation of refractive index is selected from the group comprising a Bragg grating and a long-period grating.

10. The optical fiber according to claim 8, characterised in that it is a multi-mode fiber.

11. A method for preparing an optical fiber with a core adapted for forming a structure with photoinduced refractive index modulation, characterised in that an UV laser beam is used to irradiate an optical fiber adapted to guide and generate radiation with a wavelength , the optical fiber being provided with a cladding and a nanostructured core composed of longitudinal glass elements oriented along the optical fiber and forming a compact bundle, wherein transverse dimensions of the longitudinal elements are smaller than the wavelength , and the core is composed of at least two types of longitudinal elements differing in their refractive index, whereby the longitudinal elements of the first type are made of GeO.sub.2-doped glass.

12. The method according to claim 11, characterised in that irradiation of the optical fiber by an UV laser beam is carried out by an interferometric method, a phase mask method, a point-by-point writing method, or an amplitude mask method.

Description

[0037] The disclosure will be presented now in greater detail in preferred embodiments, with reference to the accompanying drawing, in which:

[0038] FIG. 1 shows a cross-sectional structure of the optical fiber core according to the invention, with a parabolic distribution of the refractive index in the core;

[0039] FIGS. 2 a-f show a cross-sectional structure of a preform and a subpreform of the optical fiber of FIG. 1;

[0040] FIGS. 3 a-b show an image from a scanning electron microscope (SEM) of the optical fiber cross-section of FIG. 1;

[0041] FIGS. 4 a-e show the results of measurements of the second mode cut-off wavelength and attenuation for the exemplary optical fibers according to the invention;

[0042] FIGS. 5 a-b show the measured and calculated characteristics of dispersion D for optical fibers of an exemplary optical fiber according to the invention and change of the dispersion difference as a function of experimentally determined ZDW;

[0043] FIG. 6 shows a diagram of a measurement system for characterising Bragg gratings formed in an exemplary optical fiber according to the invention;

[0044] FIG. 7 shows a transmission spectrum of a Bragg grating with .sub.B=1061.5 nm, formed in an exemplary optical fiber according to the invention;

[0045] FIG. 8 shows a transmission spectrum of Bragg gratings with .sub.B close to the third transmission window (1550 nm), formed in an exemplary optical fiber according to the invention;

[0046] FIGS. 9 a-b show a temperature-dependent shift of the Bragg wavelength for an exemplary optical fiber according to the invention;

[0047] FIGS. 10 a-b show a temperature-dependent shift of the Bragg wavelength for an exemplary optical fiber according to the invention;

[0048] FIG. 11 shows an energy density-dependent photoinduced refractive index modulation for an exemplary optical fiber with a step profile of the refractive index according to the invention;

[0049] FIG. 12 shows a photo-induced refractive index modulation dependent on the distribution of GeO.sub.2-doped rods for the same power density;

[0050] FIG. 13 shows the efficiency of the refractive index modulation dependent on the distribution of GeO.sub.2-doped rods as a function of power density.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

[0051] Exemplary nGRIN (nano-structured Graded-Index) optical fiber according to the disclosure, the cross-section of which is shown in FIG. 1, was obtained as follows:

Fiber Preparation

[0052] Pure silica glass (Ohara SK1310) was formed into glass rods with a diameter of 0.45 mm. A step-index preform doped with GeO.sub.2 (Optacore), with a concentration of 8.5 mol % was also drawn to obtain glass rods having the same diameter. Next, a structured preform was stacked from glass rods layer by layer according to the pattern shown in FIG. 1. A total of 2107 glass rods were used, with 53 rods on the diagonal. The preform was thermally integrated and drawn to form subpreforms with a diameter of 1 to 3 mm. The subpreform was then placed inside a glass tube (made of pure silica glass from Heraeus), and the free space was filled with glass rods of the same type in order to obtain the desired size of the cladding of the final fibers in the range from 100 to 136 m. All drawing processes were carried out using a tower for drawing optical fibers, typically used for the preparation of silica glass fibers. Fiber #2 was additionally protected with a polymer coating to improve mechanical properties, which allowed to measure the cut-off wavelength of the second mode by bending method.

[0053] Table 1 below shows properties of types of glass and glass preforms used to prepare optical fibers.

TABLE-US-00001 TABLE 1 Type of glass low high refractive index refractive (preform having a step profile) index core cladding Parameter SK1310 F300 doped with Ge F300 refractive index, ne 1.4609 1.47207 1.46007 initial diameter [mm] 10 18.4 24 GeO.sub.2 concentration 0 8.5 0 [mol %] final rod diameter [mm] 0.45 0.45 effective GeO.sub.2 0 4.9* concentration [% mol] effective refractive index ne 1.4609 1.4678 *Due to the diffusion effect and to averaging the germanium concentration in the whole rod from the preform having a step profile of refractive index, the effective GeO.sub.2 concentration in doped rods is 4.9 mol %.

[0054] In FIG. 2 a cross-section is shown of a complex structural preform and a subpreform of nGRIN element made of rods of pure silica glass and germanium-doped glass. The structural preform is shown in perspective view (FIG. 2a) and in front view front (FIG. 2d). In the preform view of FIG. 2b each glass rod with a diameter of 0.45 mm can be distinguished. FIG. 2c shows a cross-section of the drawn subpreform with a diameter of 1.5 mm. FIGS. 2c and 2f show a cross-section of the subpreform in an enlarged view. Dark areas correspond to the non-doped silica, and bright onesto the core rods with a high dopant content (8.5 mol % of GeO.sub.2), formed of base preform with a step profile of the refractive index. A mosaic pattern designed for the fiber, shown in FIG. 2e was very well preserved in the actual structure of the drawn subpreform shown in FIG. 2f. Diameter of individual rods in the subpreform with a diagonal of 1.5 mm is about 28 m.

[0055] Table 2 below shows geometrical parameters of the prepared fibers.

TABLE-US-00002 TABLE 2 Fiber m [m] M [m] [m] #1 6.51 7.82 1.20 100.2 #2 7.15 7.72 1.08 135.8 #3 7.44 7.88 1.06 103.7 #4 7.88 9.85 1.25 125.7 #5 8.41 9.72 1.16 124.4

[0056] Diameter of the cladding was marked as . Cores of the prepared fibers were dimensioned using the major axis (M) and the minor axis (m) of the ellipse and ellipticity () just like in case of elliptical core fibers. Fibers #2 and #3 have cores most similar in shape to a circle, resulting in the smallest ellipticity. Fibers #1, #4 and #5 became elliptical during drawing, which may result in residual birefringence of fibers. Ellipticity of the prepared fibers does not exceed 1.25, thus, the birefringence is very small and therefore negligible.

[0057] FIGS. 3 a and b show a SEM image of cross-section of fiber #4 with a core radius of approximately 4 m and a cladding diameter of 125.7 m. Diameters of individual rods in the core of this fiber are about 190 nm. Dark areas in FIG. 3b correspond to pure silica glass, and bright areas correspond to Ge-doped silica glass rods.

Characterisation of nGRIN Fibers

[0058] In the first step, a modal analysis was carried out by observing the modal field distribution. The fiber #2 is at least two-mode for a wavelength of 0.85 m, but it is single-mode from at least 1.064 m. The optical fiber #3 for up to 1.064 m is still two-mode, and single-mode from at least 1.310 m (FIGS. 4 (a-b)). This method is not critical for estimating the cut-off wavelength (.sub.cut-off). Consequently, the .sub.cut-off value is contained between these wavelengths. Therefore, an additional measurement was carried out for the fiber #2. Higher order modes (HOM) were filtered from the output spectrum by bending (the fiber was protected by a polymer coating). Losses generated in the bent fiber are indicated in FIG. 4c. Initially, the straight fiber (continuous series) was bent with a radius R.sub.12 cm (dashed series) and with a smaller radius R.sub.21 cm (dotted series). The long-wave side of the attenuation peak located near the smaller wavelengths corresponded to the cut-off wavelength for mode LP.sub.11 and was .sub.LP11=0.95 m. As in case of a silica fiber having a step change of refractive index, such as e.g. SMF-28, bent with such a small radius (1 cm), the bending negatively affects also the fundamental mode (FM), what can be observed for the long-wave edge of the examined spectral range, where losses increase very quickly. Transmission for straight and bent fibers was measured using a broadband (600-1700 nm) optical spectrum analyser. In the next step, numerical aperture NA was determined for a wavelength of 1.55 m by means of a standard method and for all fibers it is 0.110.01.

[0059] FIG. 4 illustrates measurements of cut-off wavelength and attenuation. The intensity distribution of modes for all fibers was recorded using two CCD cameras: with a silicon sensor and with an additional phosphorus layer for testing the near-infrared mode at 1.55 m. FIGS. 4a-b show modal field distribution recorded with a CCD camera at different wavelengths and for different coupling conditions, for the fibers #2 and #3, respectively. FIG. 4c shows characteristics of losses due to bending, showing cut-off wavelength for mode LP.sub.11 in the fiber #2. FIG. 4d shows optical power at the output of long (continuous series) and short (dashed series) sections of the fiber #3, and FIG. 4e shows attenuation in a wide spectral range, calculated basing on FIG. 4d.

[0060] Attenuation was measured using a typical cut-off method and, in the considered spectral range from 0.7 to 1.7 m, it was at the same level of 0.05 dB/m for all fibers. Attenuation is higher only at the characteristic peak of absorption on OH ions, located at a wavelength of approximately 1.4 m, where its value is 0.5 dB/m (see FIG. 4e). The resulting attenuation is low, taking into account the laboratory conditions for fiber drawing and the standard quality of the glass rods used to prepare the fibers. It also shows that the assembling procedure introduced no serious impurities into the fiber.

[0061] Dispersion D of the prepared fibers was also examined. Measurements were carried out using a Mach-Zehnder interferometer with full compensation of optical elements in both arms. Numerical analysis of the dispersion of the prepared fibers was also carried out. The structures were implemented into the numerical model with the assumption of subpreform pattern (FIG. 2b) and the core size of the final fibers (Table 2). Simulations were also carried out for exemplary fibers with a parabolic distribution of refractive index and with a hexagonal core shape having a core diagonal corresponding to the average core size of the actual fibers. Ellipticity of the cores was neglected. The measured and calculated dispersion characteristics are shown by way of example for one optical fiber #4 in FIG. 5a. From the experimental curves measured for most fibers, the wavelength, for which the dispersion has a value zero, was determined (ZDW, zero dispersion wavelength) and the values of numerically estimated dispersion functions were calculated at these wavelengths. The differences are shown in FIG. 5b for the optical fibers #1, #3, #4 and #5.

[0062] Parameters D calculated for the actual structures are indicated by the dotted series named nGRIN in FIG. 5a. The experimental results are marked as continuous series. Parabolic distributions (dashed series) were also calculated with the assumption of averaged size of cores for each fiber, respectively. Nanostructuring with an arbitrary, parabolic distribution of refractive index with an inclusion size of /3 leads to a refractive index distribution very similar to the parabolic shape, which was confirmed numerically for the examined fibers. Experimental results allow to verify whether the dispersion shape is consistent with the numerical simulations' results. Small discrepancies in the dispersion parameter D, not exceeding 3.5 psnm.sup.1 km.sup.1, can be attributed to measurement uncertainty (FIG. 5b). The values of D correspond to a mismatch between the measured and the calculated zero-dispersion wavelengths. If differences in dispersion between experimental and ideal (parabolic) distribution are taken into account, the value of D is lower than 1.5 psnm.sup.1 km.sup.1. To sum up, the dispersion characteristics obtained experimentally is consistent with the expected one. It confirms that the drawing process is well-controlled and the method allows for preparation of fibers having predictable parameters.

[0063] Table 3 below shows a comparison of parameters of the optical fiber according to the invention (with nano-structured core nGRIN) and of the typical telecommunications optical fiber SMF-28.

TABLE-US-00003 TABLE 3 Parameter nGRIN SMF-28 Refractive index profile Gradient: parabolic step Core diameter [m] 7.6 (averaged) 8.2 Cladding diameter [m] 125 125 Numerical aperture NA 0.11 0.14 Modal field diameter at 1550 nm 9.8 10.4 [m] Cut-off wavelength [nm] 1100 1260 ZDW [nm] 1325 1347 1304 1324 Attenuation at 1550 nm [dB/m] 0.05 0.00018

[0064] As it can be seen from the data shown in the table above, the parameters are mostly comparable. The nGRIN fiber shows the largest discrepancy (two orders worse) for attenuation, which should be attributed to laboratory conditions for preparation and to the lower purity of materials used, compared to the commercial fiber SMF-28. The examined nGRIN optical fiber has a slightly smaller core (but may have exactly the same as SMF), which proportionally results in a slightly longer wavelength at which zero dispersion (ZDW) occurs. The main difference between nGRIN fiber and SMF-28 fiber is the gradient-parabolic profile of refractive index and significant reduction of cut-off wavelength values, which are attributed to the possibility of profiling modal characteristics of the optical fiber by nanostructuring. Due to the lower cut-off wavelength value, the optical fiber is single-mode for a wider wavelength range.

Bragg Grating Characterisation

[0065] In a fiber with a nanostructured core (the fiber #4 having geometrical parameters provided in Table 2 above), a Bragg grating was fabricated using the phase mask method, and its properties were subsequently characterised.

(a) Measuring System for Characterising the Bragg Grating

[0066] The fiber with the inscribed Bragg grating (BG) was placed in the measuring system shown in FIG. 6. Light from a broadband supercontinuum source (SC) was introduced into the fiber through a 20 magnification microscope lens and let out through a 40 magnification microscope lens. A mirror (M) allowed to measure the modal field distribution on a CCD camera. After removing it, light intensity at the output of the optical fiber was measured. For this purpose, the light was focused on the input of a single-mode optical fiber cable connected to an optical spectrum analyser (OSA). In order to characterise Bragg gratings in terms of temperature sensitivity, the fiber with a centrally located BG grating was placed on the Peltier module in order to implement a controlled temperature change in the range of 13.5 C. to 100 C., and the Bragg peak shift in the transmission spectrum recorded on the optical spectrum analyser was monitored.

(b) Transmission Spectrum of the Bragg Grating

[0067] The spectral transmission characteristics were measured for all the fabricated Bragg gratings. As shown in FIGS. 7 and 8, each of the four fabricated Bragg gratings showed a central wavelength of the Bragg peak at .sub.B0=1061.5 nm, .sub.B1=1553.1 nm, .sub.B2=1552.6 nm and .sub.B3=1535.5 nm, respectively. Spectral width values .sub.FWHM varied in the range of 250 to 900 m, and the transmission minimum PB corresponding to the Bragg wavelength varied in the range of 14 to 34 dB.

(c) Temperature-Dependent Shift of the Bragg Wavelength for BG at .SUB.B.=1061.5 nm

[0068] FIG. 9 shows Bragg peaks recorded for a heated fiber, whereby FIG. 9a shows the peak shift in the transmission spectrum, and FIG. 9b shows central wavelength shift estimated for the heated and cooled fiber. Temperature sensitivity of the grating is d.sub.B0/dT=10.80.1 pm/K or after normalising to the wavelength 10.2 K.sup.1. As the temperature rose the Bragg peak shifted towards longer waves.

(d) Temperature-Dependent Shift of the Bragg Wavelength for BG at .SUB.B2.=1552.6 nm

[0069] FIG. 10 shows Bragg peaks recorded for the heated fiber, whereby FIG. 10a shows the peak shift in the transmission spectrum, and FIG. 10b shows central wavelength shift estimated for the heated and cooled fiber. Temperature sensitivity of the grating is d.sub.B0/dT=16.20.3 pm/K or after normalising to the wavelength 10.4 K.sup.1. As the temperature rose the Bragg peak shifted towards longer waves.

Efficiency of the Refractive Index Modulation Dependent on the Dopant Distribution in an Optical Fiber with an Effective Step Change of the Refractive Index

[0070] Three optical fibers were considered, i.e. three distributions of GeO.sub.2-doped rods and non-doped rods in the optical fiber core with an effective step profile of the refraction index. In W1 fiber, GeO.sub.2 dopant is evenly distributed across the entire core cross-section in the form of 100% packing of rods with a germanium concentration of 3 mol %. For W2 fiber, 38% of rods in the entire core cross-section were replaced with pure silica rods, and the remaining 62% of rods were replaced with rods with a higher dopant concentration of 4.9 mol %. In case of W3 fiber, 64% of the core cross-section is composed of pure silica rods, and 36% of rods having an increased germanium concentration of 8.5 mol %. All fibers have the same average germanium concentration in the core of 3 mol % and all have a step profile of refractive index for the transmitted light, with a maximum value corresponding to germanium-doped silica in an amount of 3 mol %. In publication of J. Albert, B. Malo, K. O. Hill, F. Bilodeau, D. C. Johnson, and S. Theriault, Comparison of one-photon and two-photon effects in the photosensitivity of germanium-doped silica optical fibers exposed to intense ArF excimer laser pulses, [Appl. Phys. Lett 67, 3529 (1995)], it was shown that photoinduced efficiency of refractive index modulation in a time unit is dependent on the germanium concentration in the optical fiber core and on the laser energy density. FIG. 11 shows calculated characteristics showing the efficiency of refractive index modulation as a function of laser beam energy density for step-index optical fibers with a dopant concentration ranging from 3 to 9 mol %, according to the above-mentioned article. It has been also shown that refractive index modulation along the fiber is obtained in the three types of optical fibers discussed above, FIG. 12 for the same power density. It can be seen that for the W1 fiber which is identical to a typical telecommunication fiber in terms of core doping level and refractive index profile, this modulation is the smallest, whereas for the passive photo-sensitive fibers according to the invention the modulation is higher and increases non-linearly with increasing dopant concentration. It has been also shown how in these three types of optical fibers the efficiency of refractive index modulation depends on the number of Bragg grating periods, FIG. 13. Thus, the core nanostructuring allows profiling of fiber photosensitivity while maintaining the same propagation properties.