Glass large-core optical fibers

Various embodiments include devices and devices that use optical fibers such as optical fibers with large core dimensions and optical fibers that support single-mode propagation, as well as fibers such as lasers, amplifiers and laser-based material handling systems. Regarding the system.

Single-mode optical fibers provide a flexible transfer medium for high quality light beams. Traditional single-mode fibers generally have a core diameter of less than 9 μm. However, the small core diameter of conventional single-mode fibers is not well suited for the transmission of high power light beams.

High-intensity beams propagating through these fibers can cause strong non-linear effects such as self-phase modulation, Raman scattering, and Brillouin scattering. Self-phase modulation can lead to pulse distortion. Raman and Brillouin scattering can lead to large power losses during transmission.

Recently, fiber lasers based on the rare earth ion and Raman effects have advantages in many applications over their solid equivalents, and their power levels continue to improve. One limitation for further scaling of output power from a single-mode fiber laser or amplifier is non-linearity. However, the upper limit can be extended by designs that use much larger core diameters.

Single mode fiber optics are needed. It may also be desirable for those fibers to operate over a very large wavelength range. Ease of use and manufacture is also a consideration that can lead to increased use.

<p><patcit num="1"><text>US Provisional Patent Application No. 60 / 975,478</text></patcit><patcit num="2"><text>US Provisional Patent Application No. 61 / 086,433</text></patcit><patcit num="3"><text>U.S. Patent Application No. 10 / 844,943</text></patcit><patcit num="4"><text>U.S. Patent Application No. 11 / 134,856</text></patcit><patcit num="5"><text>U.S. Patent Application No. 11 / 323,177</text></patcit><patcit num="6"><text>U.S. Pat. No. 5,818,630</text></patcit></p>

<p><nplcit num="1"><text>Birks et al., Optics Letters, Vol. 22, No. 3, July 1, 1997, pp. 961-963</text></nplcit><nplcit num="2"><text>Martinez, F., Husey, CD, "(E) ESI determination from mode-field diameter and refractive index profile measurements on single-mode fiber", IEEE Proceedings V 135, pp. 202-210 (1988)</text></nplcit><nplcit num="3"><text>Knight et al., "Properties of photonic crystal fiber and the effective index mode", J.Opt.Soc.AM.A, Vol. 15, pp. 748-752 (1998)</text></nplcit><nplcit num="4"><text>Mortensen et al., "Modal cutoff and the V parameter in photonic crystal fibers", Opt. Lett. 28, 1879-1881 (2003)</text></nplcit><nplcit num="5"><text>Mortensen Optics Express, Volume 10, pp. 341-348, 2002</text></nplcit><nplcit num="6"><text>Kuhlmey et al., Optics Express, Volume 10, pp. 1285-1290, 2002</text></nplcit><nplcit num="7"><text>Limpert et al., Optics Express, Vol. 14, pp. 2715-2720, 2006</text></nplcit><nplcit num="8"><text>Schreiber et al., Optics Express, Vol. 13, pp. 7622-7630, 2005</text></nplcit><nplcit num="9"><text>Introduction to Fiber Optics, Cambridge University Press, 1998, page 153, Ghatak, "Butt coupling between single mode fibers"</text></nplcit></p>

Various embodiments include single mode fibers and large core fibers. In some embodiments, the fiber can operate over a large wavelength range.

Various embodiments also include fibers containing clad features, including materials that can produce very low relative index differences with respect to the clad material on which the clad features are located (eg, fluoridated quartz glass). Explained. This relative index of refraction difference is (n₁-n₂) / N₁May be characterized by, where n₁Is the index of refraction of the clad material that contains the clad features, n₂Is the refractive index of the clad feature.

In various embodiments, to achieve mode filtering and single-mode propagation, the configuration of clad features, such as the size and spacing of clad features, allows confinement of the basic mode and nevertheless a second mode. And may be chosen to allow higher-order mode leaks.

It is generally known that bending loss performance deteriorates with a low relative index of refraction difference. This tendency was confirmed by experiments and simulations. Surprisingly, however, it was also found that a significant improvement in mode filtering could be achieved with a low relative index difference, while bending loss performance was still reasonable. In particular, a small relative index difference can allow for better mode filtering and ease of use and manufacture, while allowing a relatively large but reasonable bending loss.

Thus, in various embodiments, the fiber is formed with the core at least partially enclosed by clad features. Clad has a refractive index n₁The first clad material with and the second index of refraction n₂Includes clad features formed from a second clad material with n₂&lt;n₁And the relative index of refraction difference (n)₁-n₂) / N₁Is small.

In some embodiments, the relative index of refraction difference (n) is very small.₁-n₂) / N₁However, it allows for improved mode filtering and sufficiently low mode loss.

In one embodiment, the relative index of refraction difference is about 4.5 × 10.^-3May be less than. In one embodiment, the relative index of refraction difference is about 1 × 10.^-3May be less than. In one embodiment, the relative index of refraction difference is about 8 × 10.^-4May be less than.

In some embodiments, the configuration may achieve higher-order mode losses of at least about 1 dB / m. In some embodiments, the configuration may achieve higher-order mode losses of at least about 5 dB / m. In some embodiments, the configuration may achieve higher-order mode losses of at least about 10 dB / m.

In some embodiments, the fiber diameter may be configured to achieve low loss in basic mode while increasing loss in higher order modes. For example, in various embodiments, the loss in basic mode may be less than or equal to about 0.1 dB / m, and in some embodiments the loss in higher order modes may be greater than or equal to about 1 dB / m. In other embodiments, the loss in basic mode may be less than or equal to about 1 dB / m, and in some embodiments the loss in higher order modes may be greater than or equal to about 10 dB / m.

In some embodiments, the fiber has a bend radius of 10 cm or more.

In some embodiments, the fiber may be "whole glass". The core, the clad material on which the clad features are located, and the clad features may include glass over at least a portion of the length of the fiber.

In various embodiments, the fiber design is a leak channel fiber (LCF), a photonic crystal fiber, an endless single-mode photonic crystal fiber (&gt; 500 nm or 1000 nm) that facilitates higher-order mode leakage. It has a wavelength range that exceeds, eg, has a wavelength range of at least 458 nm to 1550 nm) and / or may include large core designs.

In at least one embodiment, the optical fiber comprises a core surrounded by several glass features, the position, shape and dimensions of the glass features being at least 1 dB / m for single mode operation and the second mode of the fiber. Realize the loss.

In at least one embodiment, the fiber comprises a core surrounded by many second glass features arranged in a regular or irregular array in the first clad material, the position, shape and dimensions of the features. Enables single-mode operation, and the relative index of refraction difference is about 4.5 x 10^-3Is less than. In some embodiments, the relative index of refraction difference is about 1.0 × 10.^-3Is less than.

In at least one embodiment, the fiber comprises a core surrounded by a plurality of second glass features arranged in a regular or irregular array in the first clad material, the position, shape and number of features. Allows single-mode operation over a wide wavelength range (eg at least about 500 nm) with a relative index difference of about 4.5 x 10^-3Is less than. In some embodiments, the wavelength range is 500 to 1000 nm, for example 800 nm. In some embodiments, the relative index of refraction difference is about 1.0 × 10.^-3Is less than.

In at least one embodiment, the fiber is configured to produce a large discriminatory mode loss between higher order modes and basic modes. For example, the loss in primary mode may be about 0.1 dB / m or less, and in some embodiments the loss in higher mode may be about 1 dB / m or more.

At least one embodiment propagates at least one lower order mode having one wavelength (eg, a basic mode), while at least one said at said wavelength to a higher order mode having said wavelength. Includes an optical fiber that limits the propagation of the higher order modes by producing higher losses than the lower order modes. Optical fiber has a refractive index n₁The first clad region containing the material with, and the index of refraction n₂It may include a second clad region containing one or more clad features comprising a material with, and a core region at least partially enclosed by the second clad region. The clad feature is configured to substantially confine the propagation of the lower order modes to the core region. Clad feature is low relative index of refraction difference Δ_c= (n₁-n₂) / N₁The loss of at least one higher order mode is at least about 1 dB / m, including the material that produces. In some embodiments, the loss of at least one higher order mode is at least about 5 dB / m.

In various embodiments, the optical fiber has a hexagonal cross section that is orthogonal to the length of the fiber. The hexagonal cross section has rounded corners. In some embodiments, the fiber may be a double clad fiber, where the hexagonal perimeter of the fiber forms a reflective boundary in the region that induces excitation. Such a double clad fiber with a hexagonal cross section with rounded corners may be beneficial for excitation mode mixing. In the case of a perfect circular boundary, there is a skew ray that does not pass through the central impurity-added core. Certain embodiments with hexagonal boundaries either prevent such modes from being present or reduce the presence of such modes so that at least the amount of light passing through the center of the core can be increased. it can.

In various embodiments, the clad feature has a lower index of refraction than the background clad material on which the clad feature is placed, with impurities added.

At least one embodiment has a first clad material and a first index of refraction n.₁Includes fibers with. The clad features are located in the first clad material. The clad feature of the second clad is the second refractive index n₂have. At least one clad feature is non-rotational symmetric. Preselected outer diameters and relative spacing between features can result in higher-order mode losses that are at least about 5 dB greater than the fundamental mode loss. A mode profile that can be measured at fiber output is at least a substantial part of the center of the mode profile, for example between half-power points or 1 / e.²Between the intensity points, it is almost limited by diffraction.

Rare earths may be added to the core to form an optical gain medium. Various embodiments of the present invention may be utilized in laser amplifiers, lasers, short or ultrashort pulse generators, Q-switched lasers and other systems.

In various embodiments, the fiber comprises a first clad region containing a first clad material having a first index of refraction n1. The fiber also comprises a second clad region containing one or more clad features that include a second clad material having a refractive index of n2. The fiber also includes a core region that is at least partially enclosed by a second clad region. The first and second clad regions may be configured such that the core region propagates at least one lower order mode. In at least one embodiment, the dimensions of the first clad region and one or combination of the dimensions, arrangements and numbers of clad features result in higher-order mode losses that are greater than the basic mode losses.

Various embodiments may include a fiber amplifier system or fiber laser that produces pulses limited by near diffraction of high peak power. As an example, the output pulse energy may be in the range of about 1 μJ to 10 mJ, or about 100 μJ to 1 mJ. The output pulse width is less than 1 ns and may occur at repetition rates above 1 kHz. The output pulse may not have the distortion associated with the non-linear effect. For example, in some embodiments, the fiber optic amplifier system comprises an optical pulse source, an optical excitation source, and an optical fiber disclosed herein that supplies an optical pulse with an input space profile that is largely diffracted. Including any of. In some such embodiments, at least a portion of the core region of the fiber is added with rare earths. The fiber may further include an excitation clad configured to receive excitation energy from the photoexcitation source and an excitation guide to transfer a portion of the received excitation energy to the core region. The fiber is to receive the pulse and further amplify the pulse in order to generate an output pulse with an output space profile that is largely diffracted and at least one output pulse with energies in the range of about 10 μJ to about 10 mJ. May be configured.

Some embodiments of the fibers disclosed herein may include large core all-glass photonic crystal fibers (PCFs) with two or more layers of clad features.

In some embodiments, the LCF or PCF may be configured such that the basic mode is induced within a portion of the core that has a non-uniform index of refraction, such as a parabolic index change. The diameter of the mode can be a portion of the core diameter of the LCF, eg about 50%.

In various embodiments, the all-glass LCF may be configured such that the basic mode is induced within a portion of the core having a substantially uniform index of refraction.

In at least one embodiment, the LCF or PCF may be made by a method comprising adjusting the index of refraction profile of the LCF or PCF so that the basic mode is induced in at least a part of the core. Stress optical effects may be used to adjust the index of refraction profile. In some methods of making an optical fiber, the method determines at least one change in mode confinement loss and mode index of refraction in response to changes in the dimensions of the first clad region of the fiber, and Includes selecting the dimensions of the first clad region to produce a ratio of at least one higher mode loss to at least one lower mode loss greater than about 10.

In at least one embodiment, the all-glass fiber may include a first clad material with a first coefficient of thermal expansion. The clad feature is formed from a second clad material with a second coefficient of thermal expansion. The clad features may be placed within the first clad material. There may be a local increase in index of refraction in the region adjacent to the clad feature. The core region, which is bounded by the first inner layer of the clad feature, has a relative index difference of about 1.0 × 10 within the core.^-3It may have a non-uniform index of refraction profile such as less than. This index of refraction difference may be sufficient to induce the basic mode in at least a portion of the core, for example by a refractive index waveguide mechanism.

Various embodiments may include a large core all glass leak channel fiber (LCF). In at least one embodiment, the whole glass LCF may include a core with a diameter ranging from about 30 μm to about 200 μm. In at least one embodiment, the all-glass LCF may produce higher-order mode losses that are approximately 1 dB / m greater than the basic mode losses. In other embodiments, the all-glass LCF may produce higher-order mode losses that are approximately 5 dB / m greater than the basic mode losses.

Various embodiments may include a large core all-glass polarization preserving LCF or PCF.

An embodiment of the polarizing holding fiber comprises a first clad region containing a first clad material having a first refractive index n1 and a second clad material having a second refractive index n2 smaller than n1 1. Approximately 4.5 × 10 characterized by a second clad region containing one or more clad features and (n1-n2) / n1^-3It has a smaller relative index of refraction difference. The second clad region further contains a stress element having a refractive index different from that of the second refractive index. The fiber also includes a core region that is at least partially enclosed by a second clad region. The first and second clad regions, where the dimensions of the first clad region and one or combination of the dimensions, arrangement and number of clad features and stress elements, are more than the loss of basic mode while preserving beam polarization. The core region is configured to propagate the basic mode and retain beam polarization in such a way as to result in greater higher order mode loss.

In various embodiments, the ratio of higher mode loss to basic mode loss may be substantially maximized by preselecting the dimensions of the first clad region. In at least one embodiment, the change in mode confinement loss vs. first clad may increase or substantially maximize the ratio. In at least one embodiment, the dimension may be the diameter of the first clad region.

Various embodiments may include laser-based material processing systems that utilize fiber lasers, fiber amplifiers, and / or transmission fibers. The laser, amplifier, or transmission fiber may include an all-glass large core LCF or PCF. Material processing may be performed with spot sizes ranging from about 1 μm to about 250 μm.

Some embodiments may utilize a "whole glass" design, or at least a configuration that reduces the number of bulk optics. Integrated configurations may be used in fiber lasers, fiber amplifiers, or fiber-based transmission systems. For example, bulk optics may not be needed to couple energy into a large mode amplifier. Single-mode or multi-mode input fibers may be coupled to large-mode fibers to deliver pulses from the light source. The large mode fiber and the input fiber may be joined in such a way that the power coupled from the core of the input fiber to the large mode fiber is substantially greater than that obtained by butt coupling. For example, during bonding, the core of the input fiber may be expanded, for example by applying one or more electric arcs to heat the ends of the fiber. In addition, the output mode of the input fiber may be substantially matched to the mode of the large mode fiber.

In some embodiments, the power coupling may be high enough to limit the coupling loss to less than about 3 dB. In some embodiments, the power coupling may be high enough to limit the coupling loss to less than about 1.5 dB.

Embodiments of the present invention may be used in systems that utilize infrared, visible, and / or ultraviolet wavelengths to modify metal, semiconductor, and dielectric materials based on lasers. The fiber, laser, and / or amplifier embodiments described herein may be used in laser radar systems, material handling systems, telecommunications systems, and many other systems.

<figref num="1-a">Leakage channel optical fiber and parameters ie pitch , hole diameter d, core radius ρ and fiber diameter 2ρ₀It is a side view and a cross-sectional view schematically showing. The example fiber also includes a coating and a clad region beyond the hole. Leakage channel fibers may include large air holes to guide the optical mode. In at least one embodiment of the invention, the leak channel fiber comprises a fully "all-glass" design, including commercially available fluoridated silica as a second clad material (other than air holes), the first. A small relative index of refraction Δc is generated between the clad material of 2 and the other clad material. (Using glass clad features can result in larger clad features compared to cases where the core diameter is the same but the clad features include air holes.)</figref><figref num="1-b">Leakage channel optical fiber and parameters ie pitch , hole diameter d, core radius ρ and fiber diameter 2ρ₀It is a side view and a cross-sectional view schematically showing. The example fiber also includes a coating and a clad region beyond the hole. Leakage channel fibers may include large air holes to guide the optical mode. In at least one embodiment of the invention, the leak channel fiber comprises a fully "all-glass" design, including commercially available fluoridated silica as a second clad material (other than air holes), the first. A small relative index of refraction Δc is generated between the clad material of 2 and the other clad material. (Using glass clad features can result in larger clad features compared to cases where the core diameter is the same but the clad features include air holes.)</figref><figref num="1-c">FIG. 5 is a cross-sectional view schematically showing a double clad polarized light holding (PM) fiber including an itellibium-added large core and a multi-material clad according to an embodiment of the present invention.</figref><figref num="1-d">Shows the shape of other clad features that may be used. Cross sections of various different clad features are shown, but clad feature designs are not limited to these.</figref><figref num="2">A graph showing the estimated confinement loss and mode index vs. d / of a leak channel optical fiber in an example similar to Figure 1-b, where the clad material in which the clad features are located is infinite in diameter and covered. It was assumed that it was not. Relative index difference Δ_c(For example, a slight change in the index of refraction) is Δ_c= (n₁-n₂) / N₁Given in n₁Is the index of refraction of the background clad material (silica), n₂Is the index of refraction of the material in the hole. In this example, the refractive index of the quartz glass is assumed to be 1.444 at the design wavelength of the example of 1.05 μm, Δ for fused silica._c=8.3×10^-4Is. This graph shows the confinement loss and mode index of refraction of a leak channel fiber with a core diameter of 50 μm in both basic and second modes at a wavelength of 1.05 μm.</figref><figref num="3">If there is fluoridated silica in the hole and the core diameter is 50 μm, but constant d / = 0.675, then n₁Estimated confinement loss and mode index vs. relative index difference Δ at a wavelength of 1.05 μm corresponding to FIG. 2 for fused quartz of = 1.444._cIt is a graph which shows.</figref><figref num="4">For the leak channel fiber of Figure 1 with a core diameter of 50 μm, n₁Estimated design upper and lower limits of d / for the relative index difference between the fused silica of 1.444 and the fluoridated quartz glass (the second clad material that replaces the hole) (desirable for primary and higher order modes) It is a graph which shows the loss level (based on 0.1 dB / m and 1 dB / m, respectively) at a wavelength of 1.05 μm. This graph also shows the change in the refractive index difference Δn between the mode refractive indexes of the basic mode and the second mode as a function of Δc.</figref><figref num="5">It is a graph which shows the estimated bending loss and the mode refractive index of a basic mode and a 2nd mode with respect to a bending radius. This graph shows a core diameter of 50 μm, d / = 0.673, a wavelength of 1.05 μm, and the relative index difference Δ between two glasses._c=8.3×10^-4The simulation results of bending in two directions are shown for a leak channel fiber that has.</figref><figref num="6">Relative index difference between two glasses at a wavelength of 1.05 μm Δ_c=8.3×10^-4It is a graph which shows the upper limit and the lower limit of the design of d / and the estimation dependence of Δn with respect to the leakage channel fiber with.</figref><figref num="7">Core diameter of 48.6 μm, d / = 0.8, and relative index difference Δ between two glasses_c=8.3×10^-4It is a photograph (microscopic image) showing a cross section of a leak channel fiber manufactured with.</figref><figref num="8">A measurement mode profile obtained along two orthogonal directions using the manufactured leak channel fiber (core diameter of 48.6 μm) in Figure 7 (eg, a near-field distribution of modes) is shown.</figref><figref num="9">It is a graph which shows the measured bending loss vs. bending radius obtained by some bending radii of the manufactured leak fiber of FIG. 7 and the curve fit based on those measured values.</figref><figref num="10">Using a manufactured leak channel fiber with a core diameter of 110 μm, the measurement mode profile obtained along two orthogonal directions (eg, mode near field distribution) is shown.</figref><figref num="11">FIG. 5 is a graph showing measured bend loss vs. bend radii obtained at several bend radii of the manufactured leak channel fiber of FIG. 10 with a core diameter of 110 μm and curve fit based on those measurements.</figref><figref num="12">It is a photograph (microscopic image) of the produced fiber having a non-circular shape, and the glass feature of the second clad is almost rotationally symmetric.</figref><figref num="13-a">Of other manufactured fibers that have a non-circular shape and a 152 μm core, and also contain six highly non-rotationally symmetric (eg, non-circular) low index clad features and a clad region beyond these features. It is a photograph (microscopic image). However, these features are located on a rotationally symmetric path around the core, eg, along a ring.</figref><figref num="13-b">The measurement mode profile (mode "near field" distribution) obtained along two orthogonal directions using the manufactured fiber of Figure 13-a is shown.</figref><figref num="13-c">It is a graph which shows the measured bending loss vs. bending radius obtained by some bending radii of the manufactured fiber of FIG. 13-a and the curve fit based on those measured values.</figref><figref num="14-a">Photographs of six low index clad features with a non-circular shape, 168 μm core, non-circular / highly non-rotationally symmetric and other manufactured fibers containing a clad region beyond these features. (However, these features are arranged in a rotationally symmetric path around the core, for example in a ring.)</figref><figref num="14-b">The measurement mode profile (mode "near field" distribution) obtained along two orthogonal directions using the manufactured fiber of Figure 14-a is shown.</figref><figref num="15">It is a graph which shows the estimated confinement loss and the mode refractive index vs. fiber diameter. The profiles of the basic mode and the secondary mode corresponding to the two orthogonal directions are superposed to show the corresponding changes.</figref><figref num="16">It is a graph which shows the estimated confinement loss and the mode refractive index vs. the fiber diameter, and shows the loss which changes as the fiber diameter changes, and the coating refractive index different from the coating of FIG.</figref><figref num="17">It is a graph which shows the estimated confinement loss and the mode refractive index vs. the coating refractive index.</figref><figref num="18">It is a graph which shows the estimated confinement loss and the mode refractive index vs. wavelength.</figref><figref num="19">It is a cross-sectional view schematically showing a fiber having holes arranged in a hexagonal shape of three layers (N = 3), and definitions of pitch , hole diameter d, and core radius ρ.</figref><figref num="20">It is a graph which shows the simulated confinement loss vs. d / of the basic mode and the second mode. Embodiments of the examples of the present invention may include two or three layers (N = 2, N = 3) of fibers. These graphs show the corresponding second derivative of the loss in the second mode. In this example, Δn = 1.2 × 10^-3And λ / normalized wavelength = 10^-2Is.</figref><figref num="21">It is a graph showing the normalized wavelength λ / vs. d / , and shows the region of single-mode operation of air holes and all-glass fibers.</figref><figref num="22">A graph showing the effective mode index of refraction vs. fiber diameter, showing basic mode, secondary mode, some higher order modes, and discontinuities referred to herein as "anti-intersections". Mode profiles for some calculated examples are also shown. 22A-22F are larger diagrams showing the mode profile of the calculated example of FIG.</figref><figref num="23">It is a graph showing the confinement loss and the mode refractive index vs. fiber diameter, and further shows the relative confinement loss of the basic mode and the secondary mode at the anti-intersection. Mode profiles for some calculated examples are also shown. 23A-23D are larger views showing the mode profile of the example of FIG.</figref><figref num="24">It is a graph which shows the confinement loss and the loss ratio | wavelength of a secondary mode and a basic mode. Mode profiles for some calculated examples are also shown.</figref><figref num="24A">It is a larger view showing the mode profile of the example of FIG.</figref><figref num="24B">It is a larger view showing the mode profile of the example of FIG.</figref><figref num="25">It is a graph which shows the performance of the manufactured leakage mode fiber. Mode profiles for some examples are also shown. The measured critical bend radii at 2 dB / m are plotted in the inset. 25A and 25B are larger views showing the mode profile of the example of FIG. 25. FIG. 25C is a larger view of the inset of FIG. 25 showing the measured critical bend radii at 2 dB / m.</figref><figref num="26">It is a graph which shows the performance of the manufactured optical fiber in a fiber amplifier system. FIG. 26 also shows the cross section of the fiber and the output mode.</figref><figref num="26A">It is a larger view of the cross section of the fiber shown in FIG.</figref><figref num="26B">Output mode shown in FIG. 26 is a larger view of the de.</figref><figref num="26C">An embodiment of a large mode area fiber containing a core added with rare earth ions, which may be used in a fiber amplifier or laser excited by a multimode excitation source, is schematically shown.</figref><figref num="26D">An embodiment of a laser-based material processing system utilizing a large-mode area fiber amplifier is schematically shown.</figref><figref num="27">FIG. 26 is a three-dimensional diagram showing the measured two-dimensional index profile of the fiber of FIG. 26, showing an increase in index of refraction near the low index feature.</figref><figref num="28A">It shows some properties of a three-layer manufactured all-glass LCF. FIG. 28A shows a cross section of the fiber.</figref><figref num="28B">It shows some properties of a three-layer manufactured all-glass LCF. FIG. 28B shows the measured two-dimensional index profile.</figref><figref num="28C">It shows some properties of a three-layer manufactured all-glass LCF. Figure 28C shows some mode profiles measured at several different wavelengths in the range 780 nm to 1100 nm.</figref><figref num="29">A mode field measurement obtained at several wavelengths using a fiber drawn from the same preform as the LCF in Figure 28A is shown.</figref><figref num="30">It is a graph which shows a part of the 2D refractive index profile of FIG. 28B.</figref><figref num="31">A cross section of a manufactured LCF with a two-layer clad feature is shown. Here, d / 0.6.</figref><figref num="32">It is a graph which shows the confinement loss of the basic mode and the secondary mode of LCF which has a two-layer clad characteristic.</figref><figref num="33A">It is a cross section before joining which schematically shows the embodiment of a large core LCF and a single mode fiber having an impurity-added core.</figref><figref num="33B">FIG. 33A is a cross section schematically showing the LCF and the single mode fiber of FIG. 33A after joining, and schematically shows the widened mode field of the single mode fiber.</figref><figref num="33C">The mode profile of the LCF output for the cladding features is schematically shown, showing the propagation of the basic core mode and the LCF waveguide.</figref><figref num="33D">It is a photograph showing the output of the manufactured LCF, and the LCF and the single mode fiber were joined as shown in FIG. 33B. This photo shows the basic core mode for the clad features and demonstrates LCF waveguide.</figref>

As mentioned above, various embodiments of the invention include single mode fibers and large core fibers.

In general, conventional single-mode fibers generally have a core diameter of less than 9 μm. Traditional fibers support more modes when the core diameter is larger.

However, photonic crystal fibers can allow single-mode operation with slightly larger core diameters by effectively reducing the equivalent numerical aperture of the fiber at the expense of its ability to bend. The photonic crystal fiber contains a large array of air holes in a generally hexagonal array in the cladding. If the hole size is small enough, the photonic crystal fiber can also operate as a single-mode optical fiber over the entire wavelength spectrum (eg, from 458 nm to 1550 nm). Such fibers are sometimes referred to as endless single-mode fibers. See, for example, Birks et al., Optics Letters, Vol. 22, No. 3, July 1, 1997, pp. 961-963.

In leak channel fiber (LCF), the optical mode is guided by a few large clad features such as air holes. Air holes may have intervals that allow light energy to leak from higher-order modes. Basic modes may be supported by leaking higher-order modes.

Aspects and designs of leak mode fibers, "holly fibers", photonic crystal fibers, photonic bandgap fibers and the like are disclosed in the following disclosures assigned to the assignee of the invention. That is, (i) US Patent Application No. 10 / 844,943 filed on May 13, 2004 under the name "Large Core Holey Fibers" published as US Patent Publication No. 2005/0157998 on July 21, 2005. No., (ii) Filed on May 30, 2005, entitled "Single Mode Propagation in Fibers and Rods with Large Leakage Channels," published as US Patent Publication No. 2006/0263024 on November 23, 2006. US Patent Application No. 11 / 134,856, (iii) Published as US Patent Publication No. 2006/0193583 on August 31, 2006, and is currently US Pat. No. 7,209,619 "Photonic Bandgap" U.S. Patent Application No. 11 / 323,177 filed on December 30, 2005, entitled "Fibers." The disclosures of U.S. Patent Applications Nos. 10 / 844,943, 11 / 134,856, and 11 / 323,177 (and their corresponding publications) are incorporated herein by reference in their entirety.

Both photonic crystal fiber and leak channel fiber designs use air holes in the glass matrix and can benefit from large refractive index differences between air and glass. This refractive index difference is the relative refractive index difference Δ_c= (n₁-n₂) / N₁Characterized as. Where n₁Is the index of refraction of the background clad material (first clad material), n₂Is the refractive index of the material in the hole (second clad material). The large index of refraction difference between the holes in the fiber and the rest (other clad material) allows for high mode confinement and excellent bending loss performance (low bending loss) for single mode operation. Relatively large relative index of refraction difference Δ_cAlso allows for adequate mode filtering.

However, in order to seal the air holes from long-term environmental exposure, a small portion of the two ends of the fiber is heated to crush the air holes. These treated end faces often distort the output mode pattern due to non-uniform crushing along the fiber. Also, the size of the air hole is controlled during fiber drawing by a combination of hole pressurization, drawing temperature, preform supply rate and fiber drawing rate. This process can limit reproducibility, especially for large core photonic crystal fibers where the dimensions of small air holes with high surface tension are very sensitive to drawing conditions.

Embodiments of the present invention generally allow improvements in leak channel fibers, photonic crystal fibers, endless single-mode photonic crystal fibers, and other designs. Such improvements may include manufacturing advantages.

In various embodiments described herein, the clad comprises clad features that include at least a second material placed in the first clad material. (In some embodiments, two or more materials are used for the clad features. For example, in some embodiments, four fluoridated silica features and two boron-added silica stress rods are used.) First. The material has a refractive index n₁The second material has a refractive index n₂have. Certain properties of this fiber are known to be affected by the relative index of refraction of the first and second materials.

For example, it is generally known that bending loss performance deteriorates in the case of a low relative index of refraction difference. This tendency was confirmed by experiments and simulations. Surprisingly, however, it was also found that while significant improvements in modal filtering can be achieved in the case of low relative index differences, bending loss performance remains reasonable.

Such low index differential fibers may include glass clad features disposed in the glass material. The glass clad feature advantageously eliminates, or at least reduces, the difficulties associated with producing air holes in the clad. Moreover, a hexagonal cross section with rounded corners is easily formed on the fiber containing the glass clad features disposed in the glass clad material. Such fibers with a hexagonal cross section are used when mode-mixing the excitation light propagating through a double clad fiber configured to guide the excitation light using at least a partial hexagonal side wall. Show advantages.

Simulation results and measurements obtained from manufactured fibers with reduced index differences show that single-mode operation is achievable with sufficient mode filtering and sufficient bending loss performance for many applications. Various embodiments of the fiber are generally configured to result in higher mode losses such that the energy radiated from the output of the fiber corresponds to the lower mode. For example, the fiber may produce an output with a mode profile that is close enough to the basic mode.

As used herein, single-mode and multi-mode fibers are consistently defined in the definitions used for traditional non-holly fibers. In traditional fibers, single-mode and multi-mode fibers are generally defined in terms of numerical aperture, which is π (numerical aperture) (core diameter) / for stepped index fibers. Equal to wavelength. For non-step index fiber, the numerical aperture and core diameter can be calculated using the equivalent values for step index fiber (eg Martinez, F., Husey, CD, "(E) ESI determination from mode". -field diameter and refractive index profile measurements on single-mode fibers , IEEE Proceedings V See 135, pp. 202-210 (1988)). For fibers that satisfy the V &lt;2.4 relationship, the power in the basic mode is significantly greater than the optical power in the next higher mode. Instead, on fibers with V&gt; 2.4, at least the next mode above the basic mode may have significantly more power than the basic mode. Therefore, single-mode and multi-mode traditional fibers are clearly defined by the relationship of V &lt;2.4 and V&gt; 2.4, respectively. V = 2.4 is a cutoff for propagation in any mode except the lowest order mode.

In the holy fiber, the numerical aperture can be known by the difference in the refractive index between the core and the clad. However, the core diameter, which is an equivalent value for a step refractive index fiber, is difficult to calculate. Various references (eg, (1) Knight et al., "Properties of photonic crystal fiber and the effective index mode", J.Opt.Soc.AM.A Vol. 15, pp. 748-752 (1998) and (2) Mortensen et al. , "Modal cutoff and the V parameter in photonic crystal "fibers", Opt. Lett. 28, pp. 1879-1881 (2003)), if the core diameter is equalized to the pitch or the distance between holes , then any other than single mode. The Vs that are also cutoffs for mode propagation are 2.5 (see, eg, Knight et al.) And π (see, eg, Mortensen et al.). For the various embodiments described herein, it is not critical whether the V cutoff is 2.405, 2.5 or π. Various embodiments of holi fibers described herein have a much larger core radius than is possible with conventional optical fibers that support propagation in a single optical mode. Therefore, we take advantage of recent work in this art that multimode fibers are defined as the case of V&gt; π, and the core diameter is equal to the pitch or average pitch of the fibers. Conversely, a single-mode fiber is defined herein as a fiber with V &lt;π.

Holy fibers may be designed to cause losses for certain modes. The hole size, bridge, and number of holes may be chosen to cause loss for higher mode propagation, for example, in multimode fibers where V&gt; π. Due to the reduced number of holes, light in higher modes may not be trapped in the core and may leak out of the fiber. Such losses incurred in multimode fibers with V&gt; π include V over π, including mode filtering achieved, for example by bending the fiber to cause losses for higher-order mode propagation. Similar to traditional non-holly multimode fiber with numbers. (For example, the mode filter is "Single-mode Amplifier and Compressors Based on Multi-mode." It is described in US Pat. No. 5,818,630 issued to Fermann et al. On October 6, 1998, entitled "Fibers"). Sufficient bending may be applied to cause loss for each of the higher modes than the basic mode, as the basic mode is the only mode propagating a bent multimode fiber. Similarly, multimode holi fibers with Vs greater than about π are designed to cause losses for these higher modes so that propagation in the higher modes is attenuated. Sometimes. The fiber does not need to be bent to achieve mode filtering.

Figures 1-a and 1-b show leak channel fibers. An example fiber optic 100 includes a core 101 surrounded by clad features 102, such as 6 holes. These features have a diameter of d and a center-to-center spacing , also known as pitch. The core has a diameter defined as the spacing 2ρ between the closest holes. Fiber diameter is 2ρ₀Is. In this example, the first clad region 103 is formed beyond the low index clad feature 102 and has a coating 104 added. The standardized hole diameter d / is chosen so that the leakage loss in the second mode is significantly higher than that in the basic mode. This allows effective single-mode operation with a much larger core diameter than is possible with conventional fiber optics by using this built-in mode filtering.

Figures 1b and 1c schematically show circular holes as clad features. Similarly, fiber diameter 2ρ₀Is shown with respect to the exemplary circular fiber. Clad features may be non-circular and the shape of the fiber may include at least non-circular portions, as will be apparent from the examples of manufactured fibers below. For example, some features or clad shapes may be close to hexagons, octagons, or rotationally asymmetric, including straight and / or curved parts. In some embodiments, the shape of the clad may be irregular and not a well-defined standard form.

One possible definition of the diameter of a polygon is the maximum distance between the vertices of any pair, which corresponds to the longest diagonal of the polygon. Furthermore, with respect to feature dimension d, d / 2 is generally considered to be the distance from the center of the feature to the edge along the line connecting the centers of the two closest neighbors, unless otherwise specified. The center of the feature may be calculated as the "center of mass" or "center of gravity" for non-circular or asymmetric features.

As an example, fiber diameter numbers are generally associated with the outer edge of the first clad 103 as shown in Figure 1-b, which clad 103 is also used as an excitation guide in some embodiments. There are times. In the circular fiber example, the diameter is 2ρ₀Is. This diameter is the maximum dimension to the outer edge of the first clad 103 along a line passing through the center of the core 101.

In at least one embodiment, at least one of the clad features 102 is a second having a refractive index lower than that of either the core 101 or the other clad material (eg, the first clad material) 103. Made from material glass. Some embodiments comprise a "whole glass" design in which both the clad feature 102 and the material on which the clad feature is located include glass.

For example, very much between fluoridated silica (clad feature 102 uses fluoridated silica or other suitable material instead of all air holes) and other clad material where the clad feature is located. Small relative index difference, for example approximately 8.3 × 10^-4Δ_cOccurs. The clad material 102 (second clad) generally has a refractive index slightly lower than that of the clad material 103 (first clad) of the optical fiber 101. Surprisingly, the low-loss single-mode operation of both the photonic crystal fiber and the leak channel fiber has a relative index difference of 7 × 10 between the two clad materials.^-3It can occur in a state that is substantially smaller than. In some embodiments, the relative index of refraction difference Δ_cIs only 2x10 for large core fibers^-4Is. The relative index of refraction may be smaller as well. In some embodiments, Δc is about 1 × 10^-3Less than, or about 4.5 x 10^-3May be less than. In the fiber embodiments described herein, the relative index of refraction is determined at a nominal wavelength of 1.05 μm. Numerical simulations were generally performed based on a design wavelength of 1.05 μm, but the wavelength dependence was quantified and found to be weak.

In addition, all-glass photonic crystal fibers, endless single-mode optical fibers and leak-channel optical fibers may allow ease of use and ease of manufacture comparable to conventional optical fibers. As explained above, the advantages of such "all-glass" fibers include sufficiently low bending loss, improved reproducibility of manufactured products and performance due to the absence of air holes, and For example, there are fiberglass cross-sectional shapes that easily take the shape of a hexagon made with rounded corners.

In general, a very small relative index of refraction allows sufficient mode filtering for single-mode operation, while also achieving reasonable bending loss performance. Larger Δ_cMeans excellent mode filtering and bending performance, but a sufficiently small relative index difference allows control of single mode and bending loss. Also, as explained above, the reduced or minimal useful index of refraction difference between the index of refraction of the first background clad material and the index of refraction of the second clad material in the holes is the manufacturability and Significantly improves performance. Relative index difference Δ_c= Approximately 8.3 × 10^-4The results obtained by making several leak channel fibers with fluoridated silica with fluoridation replacing the air holes used in the previous design are described below. The embodiment with the manufactured leak channel fiber has been determined to produce sufficient bending loss performance for many applications.

As explained above, the bending loss performance is known to be worse in the case of a low relative index difference, which has been confirmed by the applicant's experiments and simulations. However, the applicant found a significant improvement in mode filtering using the low relative index difference and found that the bending loss performance was reasonable. This design, such as the "all-glass" design, allows for improved mode filtering with clad features compared to comparable leak channel fibers with air holes.

In some embodiments with a much smaller index of refraction difference between the two clad materials, widely available high-purity fused silica glass can be used as the first clad glass 103, while fluorine or / And other high-purity quartz glass with added boron can be used as the second clad glass 102. At low levels of fluorine and boron addition, the two clad materials have excellent mechanical, chemical, physical and thermal coexistence. Commercially available fluoridated silica may be used as the material for holes 102, and high-purity quartz glass may be used as the material for the first clad 103. Other materials and designs are also possible.

Figure 1c is another example of a whole glass design, a double clad polarized ytterbium-added large core fiber. (As mentioned above, excitation mixing may be enhanced when such a double clad array is contained in a fiber with a non-circular eg hexagonal cross section.) Ytterbium addition region inside core 101. 105 is 2d₀It has a diameter of 1 and a refractive index that is closely matched to the surrounding glass. The two stress elements 106 have a coefficient of thermal expansion substantially different from that of the surrounding glass and a refractive index different from that of the other low index features 102. The stress element 106 may also have different dimensions and sizes than the other low index features 102. The stress element 106 can be made from boron-added quartz glass. In this example, the glass 107 is a low index glass that provides an excitation clad and can be made from fluorine and / or boron-added quartz glass.

Assembling the fibers of FIG. 1 may generally be done using standard methods of assembling leak fiber designs. For example, a fluoridated silica rod is first inserted into a silica tube and stretched into a stick with the desired diameter and desired ratio of fluoridated silica and quartz glass. The stick-like material is then stacked with the stick-like silica in a hexagonal stack in the desired configuration. The stack is then inserted into a silica tube and drawn to the fiber. Vacuum in a silica tube is sometimes used in combination with low drawing temperatures to produce non-circular (eg, hexagonal fiber) fibers. Ytterbium-added silica rods and / and stress rods are sometimes used in a stacked state to make ytterbium-added fibers and / and polarization-retaining fibers. In general, a low relative index "all-glass" assembly may simplify the overall manufacture of leaky fibers, for example, as described above.

Figure 1-d shows the shape of other clad features that may be used. As shown, the cross section of the clad feature is not limited to a circular shape. Figure 1-d shows some other possible forms 120-127, and these forms can also be used. Therefore, the composition of the clad feature, for example, the shape, dimensions, material, refractive index, etc. of the clad feature may change. The variation is not limited to the perimeter of the clad features, but may also include internal features and designs. For example, cross sections 128, 130 and 132 of FIG. 1-d show one or more such inclusions 129, 131 and 133 having a different refractive index than the material in which the inclusions are embedded. Moreover, these internal features may vary in shape, size, arrangement, material, refractive index, etc., for example. Yet other designs are possible.

Figures 2-18 show various simulation results and characterize some manufactured fibers based on measurements and other information. For example, fibers with different geometries, core sizes, diameters, and coatings are mentioned. The calculation results corresponding to the graphs of FIGS. 2-6 were obtained on the assumption that the clad region 103 is infinite (eg, infinite diameter) and there is no coating 104. Some calculation results demonstrate the effect of finite diameter and coating index changes. In addition, the simulation was generally performed based on the design wavelength of 1.05 μm, but the wavelength dependence was quantified and found to be weak. Fused quartz glass is assumed to have a general value index of refraction n = 1.444 close to the design wavelength, which may actually vary somewhat due to normal manufacturing tolerances. An example showing the manufactured fiber and the measured mode profile is also included.

Figure 2 shows the confinement loss and mode index of refraction in basic and secondary modes on a leak channel fiber with a core diameter of 50 μm at a wavelength of 1.05 μm. The refractive index of quartz glass is assumed to be 1.444. Relative index difference Δ lower than silica_c=8.3×10^-4Commercially available fluoridated silica with is used for six clad features. Curves 201, 202, 203 and 204 are the confinement loss of the basic mode, the confinement loss of the second mode, the mode refractive index of the basic mode, and the mode refractive index of the second mode, respectively. Confinement losses 201 and 202 are plotted for various normalized hole diameters d / . The confinement loss increases toward the small clad feature diameter, and the confinement loss in the second mode, curve 202, is approximately 27 times the confinement loss in basic mode, curve 201, when expressed in dB / m. To my surprise, Δ_c=8.3×10^-4It is shown that the small relative index of refraction difference does not impair the level of built-in mode filtering. A maximum basic mode loss of 0.1 dB / m is used to find the lower limit 205 of the normalized hole diameter d / , and a minimum second mode loss of 1 dB / m to find the upper limit 206 of the normalized hole diameter d / . Was used.

Figure 3 shows the confinement of n = 1.444 fused quartz and the glass (clad feature) in the hole to the index difference for a leak channel fiber with a core diameter of 50 μm and d / = 0.675 at a wavelength of 1.05 μm. It shows the dependence of loss and mode index of refraction. Curves 301, 302, 303 and 304 are the confinement loss of the basic mode, the confinement loss of the second mode, the mode refractive index of the basic mode, and the mode refractive index of the second mode, respectively. The confinement loss increases as the index of refraction difference decreases.

Figure 4 shows the d / design for a leak channel fiber with a core diameter of 50 μm at a wavelength of 1.05 μm to the difference in index of refraction between the fused quartz n = 1.444 and the glass (clad feature) in the hole. Shows the dependence of the upper and lower limits. A maximum basic mode loss of 0.1 dB / m is used to find the lower limit of the normalized hole diameter d / , and a minimum second mode loss of 1 dB / m is used to find the upper limit of the normalized hole diameter d / . Was done. Curves 401, 402 and 403 are the lower limit of the normalized hole diameter d / , the upper limit of the normalized hole diameter d / , and the refractive index difference Δn between the mode refractive indexes of the basic mode and the second mode, respectively. In addition to the upper and lower hole diameters of the design increasing towards smaller relative index differences, the design space also becomes smaller and smaller towards smaller relative index differences, but the design space is , In this example, only Δ_c=3.5×10^-4It is still available with a relative index difference of about.

Figure 5 shows a core diameter of 50 μm, d / = 0.673, a wavelength of 1.05 μm, and a relative index difference Δ._c=8.3×10^-4The bending losses of the basic mode and the second mode corresponding to the two directions of bending are shown for the leak channel fiber having. AA bending direction 501 is when the bending surface crosses the center of the two clad features. The BB bending direction 502 is when the bending surface does not cross the clad feature. Curves 503, 504, 505, 506, 507 and 508 are the loss of the basic mode corresponding to the AA bending direction, the loss of the basic mode corresponding to the BB bending direction, and the loss of the second mode corresponding to the AA bending direction, respectively. The loss of the second mode corresponding to the BB bending direction, the mode refractive index of the basic mode, and the mode refractive index of the second mode. It can be seen from the loss curves 503, 504, 505 and 506 of the basic mode and the second mode that the loss changes slowly above the critical bend radius and increases very rapidly below the critical bend radius. The second mode has a larger critical bending radius than the basic mode. When operating with coil radii below the critical bending radius of the second mode and above the critical bending radius of the basic mode, extremely high second mode losses and at the same time low basic mode losses may be obtained.

Figure 6 shows the relative index of refraction difference Δ at a wavelength of 1.05 μm._c=8.3×10^-4For leak channel fibers with, the dependence of d / on the upper and lower core diameters of the design is shown. Curves 601, 602 and 603 are the lower limit of the normalized clad feature diameter d / , the upper limit of the normalized clad feature diameter d / , and the index of refraction difference Δn between the mode refractive index of the basic mode and the second mode. As in the example in Figure 4, a maximum basic mode loss of 0.1 dB / m is used to find the lower bound of the normalized hole diameter d / , and 1 dB / m to find the upper bound of the normalized hole diameter d / . The minimum second mode loss was used. It can be seen that the lower and upper limits of d / in the design space tend towards smaller clad feature diameters for larger core diameters. Large clad features are used for small core diameters. In this configuration, the relative clad feature diameter can generally only approach 1, so design space is absent for core diameters less than approximately 24 μm.

The design limits of Figure 6 relate to straight fibers. Since wrapping is often used in practice, the design space in Figure 6 is often moved towards a slightly larger d / so that the fiber exhibits better bending performance. Also, for example, "limits" for predetermined performance are described herein, but certain embodiments exceed these limits, for example, when different performance is required, or for other reasons. It may be in a state of being sick.

FIG. 7 is a photograph (microscopic image) showing a cross section of a manufactured leak channel fiber with a core diameter of 48.6 μm and d / = 0.8. The effective area of the fiber is about 1340 μm²Is. Clad features include fluoridated silica, and the material surrounding the clad features includes fluorine-free quartz glass. This combination of glasses has a relative index difference Δ_c=8.3×10^-4Bring. The preform is stacked with silica rods to give a relative clad characteristic diameter d / = 0.8 (six rods made by crushing a silica tube onto a fluorine-added silica rod), then stretched. Made by making it smaller and sticky. Two fibers are drawn with diameters of 244 μm and 552 μm, each about 1340 μm.²Core diameter of 48.6 μm and about 5940 μm with effective area of²Produced a core diameter of 110 μm with an effective area of. A cross section of a manufactured leak channel fiber 700 with a core diameter of 48.6 μm is shown in FIG. 7, where core 701 is substantially surrounded by six low index fluoridated silica regions 702. The clad region beyond the low index clad feature.

FIG. 8 shows the measured short-field mode pattern 800 at the output of the fiber 700. Mode profiles along the horizontal and vertical directions are also shown in 801 and 802, respectively.

FIG. 9 shows the bending loss 901 obtained by measuring the output of the fiber 700 while winding the fiber 700 around coils of various diameters. Curve 900 is the fit to the measurement data 901. The fiber 700 can be wound around a coil with a diameter of 30 cm, for example, without causing significant loss.

FIG. 10 shows a short-field mode pattern measured from the output of a fiber with a diameter of 110 μm. Mode profiles along the vertical and horizontal axes are shown at 1001 and 1002, respectively. In this case, nearly 7 meters of fiber was wound around a single coil diameter of nearly 2 meters for measurement.

FIG. 11 shows the measured bending loss 1100 of the fiber of FIG. 10 along with the fit 1101 to the measured data 1100. The minimum bend radius in this example is about 60 cm.

Further simulation and measurement results were obtained for fibers with non-circular shapes, different diameters, coatings, and other variations.

FIG. 12 shows another manufactured fiber 1200 with a non-circular, nearly hexagonal shape. The cross section of the fiber 1200 shown in FIG. 12 has six sides with rounded corners in between. This hexagonal geometry facilitates excitation mode mixing in double clad fibers. The core 1201 and the low index feature 1202 are also non-circular and nearly hexagonal. The clad region 1203 extends beyond the low index clad feature 1202.

The fiber 1200 of FIG. 12 with a hexagonal outline but with rounded corners is contrasted with the circular geometry of the fiber 700 of FIG. As explained above, this hexagonal shape with rounded corners of the fiber 1200 is useful for excitation mode mixing in a double clad configuration where the outer shape of the fiber forms an excitation guide. In the example preform fabrication process, a hexagonal stack is formed with the fluoridated rod and the silica rod in the desired configuration. This stack is slightly melted along its length in a hot furnace before being inserted into a larger silica tube. A slight vacuum is used inside the outer tube during fiber drawing. This vacuum is sufficient to help form the hexagonal shape of the fiber and features. In some embodiments, a vacuum of -1.27 cmHg (-0.5 inHg) to -7.62 cmHg (-3 inHg) may be used at a drawing temperature of 2000 degrees Celsius. With proper adjustment of the drawing temperature, lower and higher vacuums will likely produce similar results.

FIG. 13a shows another fiber 1300, which corresponds to the end view and has a non-circular shape with fairly large boundary deformation and a non-rotationally symmetric (eg, non-circular) low index of refraction feature. However, these features are located on a rotationally symmetric path around the core, eg, along a ring. Fiber 1300 has six sides with small indentations and rounded corners between them. The manufactured fiber 1300 includes a core 1301 with a diameter of 152 μm, which is defined as the distance between sides. The fiber also has six non-circular low index features 1302 and a clad region 1303 beyond these low index features. This form is useful for enhancing excitation mixing. As explained above, in a fiber with a circular cross section, skew rays avoiding the impurified center of the fiber may propagate through the fiber. The deviation from the circular cross section can increase the number of rays passing through the center of the fiber to which the gain medium has been added, thereby increasing the efficiency of excitation.

FIG. 13b shows the measured short-field mode profile 1310 of fiber 1300, along with the vertical and horizontal mode profiles 1311 and 1312, respectively. This mode is 1.4 measured M²have. Inadequate mode filtering may be the cause of the reduced beam quality. Almost 7 meters of fiber was wound around for measurement to a single coil diameter of approximately 2 meters, as shown in Figure 10.

Figure 13-c shows the bending loss. Curve 1320 is the bending loss measured for fiber 1300. The fiber can be bent to a bending radius of 0.5 m without sacrificing significant bending loss.

FIG. 14a corresponds to the end view and shows other manufactured fibers. The manufactured fiber 1400 has a core 1401 with a diameter of 168 μm. There are six low index features 1402 and a region 1403 beyond these low index features. The fiber 1400 also has a non-circular shape with fairly large boundary deformations and a non-rotationally symmetric (eg, non-circular) low index of refraction feature. However, these features are located on a rotationally symmetric path around the core, eg, along a ring. Fiber 1400 has six sides with small indentations and rounded corners between them. This form is also useful for enhancing excitation mixing.

Figure 14b shows the short-field mode pattern 1410 measured from fiber 1400. 1411 and 1412 are mode profiles along the vertical and horizontal directions, respectively. Effective mode area is approximately 13,900 μm²Is. This mode measures 1.12 near the theoretical limit M²Have a value. Straight fibers with a length of approximately 0.9 meters were used.

FIG. 15 is a graph showing estimated containment loss and mode index vs. fiber diameter. The profiles of the basic mode and the secondary mode corresponding to the two orthogonal directions are superposed to show the corresponding changes. These graphs show the calculated basic loss 1500, basic mode index of refraction 1501, secondary mode loss 1502 and secondary mode index of refraction 1503. Curves 1504, 1505, 1506 and 1507 are basic mode profiles along the horizontal and vertical directions of solid and dotted lines for fiber diameters of 179 μm, 207.5 μm, 251.5 μm, and 259 μm, respectively. Curves 1508, 1509 and 1510 are secondary mode profiles along the horizontal and vertical directions of solid and dotted lines for fiber diameters of 196.5 μm, 210 μm, and 247.5 μm, respectively. A coated index of refraction of 1.54 is assumed. On non-circular fibers, the 1500 and 1502 peaks are likely to widen.

This calculation demonstrates that the losses in both basic and secondary modes vary with fiber diameter. At a certain diameter, the mode is more confined to the center of the fiber and exhibits low loss (indicated by 1505 and 1509). On the other hand, at some other diameters, the mode is less confined to the center (eg 1504, 1506, 1507 in basic mode) and shows higher losses.

At a diameter that causes higher loss for the secondary mode and lower for the basic mode, or at a diameter chosen to occur, as the choice of diameter can affect the loss. Or it may be beneficial to operate with a diameter optimized (or almost so) to occur.

FIG. 16 is a graph further showing the estimated confinement loss and the mode index of refraction vs. fiber diameter, showing the loss that changes as the fiber diameter changes. FIG. 16 shows the calculated basic loss 1600, basic mode refractive index 1601, secondary mode loss 1602 and second mode refractive index 1603. When the fiber diameter changes, the loss changes for both the basic mode and the second mode. For analysis, n₀= 1.37 + j^*10^-8The complex-covered refractive index was assumed.

FIG. 17 is a graph showing the estimated confinement loss and the modal index of refraction versus the coated index of refraction. Figure 17 shows the calculated basic mode loss 1700, basic mode index 1701, secondary mode loss 1702 and secondary mode index 1703 vs. coated index n.₀Is shown. Once again, the imaginary part 10^-8A lossy coating with was used. A fiber diameter of 247 μm is assumed. Line 1704 corresponds to the coating being matched to a second clad material that surrounds the clad features and contains glass with a refractive index of 1.444. Lines 1705 and 1706 are the levels of basic and secondary mode losses when the coating index is matched to a second clad material that surrounds the clad features and contains glass, respectively. It can be seen that the difference loss between the secondary mode and the basic mode is much larger during operation when the coating index is different from the index of the glass surrounding the clad features.

FIG. 18 is a graph showing the estimated confinement loss and the mode index of refraction vs. wavelength. The basic mode loss 1800, the basic mode index of refraction 1801, the secondary mode loss 1802 and the secondary mode index of refraction 1803 calculated for different wavelengths are weak, which would make it possible to realize a practical fiber. It shows wavelength dependence. Where n₀= 1.54 and the fiber diameter is 247 μm.

Some embodiments of the present invention may include photonic crystal fibers in addition to leak channel fibers. Photonic crystal fibers are described in Birks et al., Optics Letters, Vol. 22, No. 3, July 1, 1997, pp. 961-963.

One application of photonic crystal fibers is to design endless single-mode fibers, where single-mode operation is achieved over a wide range of wavelengths (eg, 458 nm to 1550 nm). Wavelength range above 500 nm or above 1000 nm, such as wavelength range up to). This is achieved by limiting the relative hole diameter to less than 0.406. Experiments have demonstrated in this region that there is more light in the glass at shorter wavelengths (relative to holes) and therefore a higher effective clad index of refraction. Therefore, as the wavelength becomes shorter, the effective NA of the fiber decreases. This dispersion of such photonic crystal fibers can advantageously be used to extend the single-mode operation of the fiber to shorter wavelengths.

In particular, this dispersion is contrary to the tendency of classic single-mode fibers to become multi-mode at shorter wavelengths. For example, the V number is inversely proportional to the wavelength. Therefore, as the wavelength decreases, the V number increases and exceeds 2.405, but at 2.405, conventional fibers are multimode at these smaller wavelengths.

In photonic crystal fibers with small holes, the effective index of refraction increases at smaller wavelengths, allowing more light to be collected into the background glass (for clad features). This dispersion compensates for the inverse dependence on wavelengths of the V number, which would otherwise limit the range of single-mode operation to higher wavelengths. Therefore, this dispersibility of the clad index of refraction in these photonic crystal fibers extends the fiber's single-mode operation to shorter wavelengths, allowing endless single-mode operation, which is a conventional single-mode operation. Not possible with fiber.

In addition to allowing single-mode operation, photonic crystal fibers can also form large-core single-mode fibers. This is due to the ability of photonic crystal fibers to achieve a very small numerical aperture (NA) by controlling the air hole diameter in the cladding. Clad is a composite whose effective index of refraction is determined by air filling factor and wavelength. Therefore, the design of the air holes, such as size and density, controls the effective index of refraction and numerical aperture.

Photonic crystal fibers (PCFs) and leak channel fibers (LCFs) have a number of holes as a differentiating factor. The PCF generally has a large number of holes and the number of layers (eg, rings) is generally N 3. The PCF may also have smaller holes than the LCF.

An example of a N = 3 fiber 1900 is shown in Figure 19, which has a core 1901 surrounded by an array of air holes 1902 arranged in a hexagonal pattern containing three rings or layers. There is. In addition, the hole diameter d, the center-to-center distance also called pitch , and the core diameter 2ρ are shown.

Some embodiments of the invention are placed in "whole glass" (eg, in a glass matrix material) in either an endless single-mode fiber configuration, a large core fiber configuration, or both. Glass clad features) may be substantially included, and glass features 1902 (not air holes) are compared to the relative index of refraction of the material in which the glass clad features are located, or the rest of the fiber. Has a very low relative index of refraction difference compared to. Only 8.3x10 as described below^-4However, perhaps even lower relative index differences may be used to achieve these fibers. Design guidelines for various embodiments, including all-glass endless single-mode fibers and large-core photonic crystal fibers, are provided herein.

The cutoff of the secondary mode can be examined to analyze the single mode operating region. In conventional fibers, the cutoff of the second mode can be easily obtained by a well-defined number of V values, for example V = 2.405 for step index fibers. Physically, this is a transition point where all internal reflection conditions in the second mode are just not met. For photonic crystal fibers with composite cladding, this point may not be just as easily defined.

Previous studies have analyzed the modal cutoff of air-hole photonic crystal fibers. The initial analysis was performed by examining the transition of the effective mode area of the fiber to find the transition point where the second mode is no longer substantially constrained in the core (Mortensen Optics Express, Vol. 10, 341-348). Page, 2002). Kuhlmey et al. (Kuhlmey et al., Optics Express, Vol. 10, pp. 1285-1290, 2002) found that a clearer transition point was in the confinement loss of the second mode. This transition point can be pinpointed by looking for the peak of the second derivative of the second mode confinement loss.

Figure 20 shows λ / = 10^-2Relative index of refraction difference Δc = 8.3 × 10^-4Demonstrate this process of locating the second mode cutoff of the fiber. Curves 2001, 2002, 2003, 2004, 2005, and 2006 show the second mode confinement loss in the dual layer fiber, the basic mode confinement loss in the dual layer fiber, and the second mode in the dual layer fiber, respectively. In the second derivative of the loss, the second mode confinement loss in the three-layer fiber, the basic mode confinement loss in the three-layer fiber, and the second derivative of the second mode loss in the three-layer fiber. is there. The strong dependence on d / , that is, the transition point of the second mode loss from the locally constrained region to the loose dependence, that is, the delocalized mode, is shown by curves 2001 and 2004. This point is pinpointed by the peaks of curve 2003 for N = 2 and curve 2006 for N = 3. The number of layers used does not affect the position of the transition point, but narrows the peak of the second derivative of the second mode loss. Absolute loss is even less with fibers with more layers. In contrast, as the mode becomes less localized, the mode becomes more sensitive to perturbations such as microbends and large bends, which rapidly attenuates the mode in the actual fiber.

Kuhlmey's results (Kuhlmey et al., Optics Express, Vol. 10, pp. 1285-1290, 2002) are shown in Figure 21 with curve 2101. The second mode would be effectively a cutoff to the left of curve 2101. The endless single-mode operation of the air-hole photonic crystal fiber is applicable when d / &lt;0.406. Curve 2100 is Δ_c=8.3×10^-4This is the second mode cutoff in the case of the relative index difference of. The upper part of FIG. 21 relates to a small core diameter with a large λ / , while the lower part of FIG. 21 relates to a large core with a small λ / . As can be seen from Figure 21, the endless single-mode operation is Δ at d / &lt;0.425._c=8.3×10^-4Can be achieved in the case of. In the large core region, at the bottom of FIG. 21, curves 2100 and 2101 are in close proximity, but in the small core region, at the top of FIG. 21, the curves 2100 are moving more towards the larger d / . A small core diameter, ie λ / &gt; 9.17 × 10^-2Can achieve single mode operation for all d / .

The calculated second mode cutoff is the relative index difference Δ_c=8.3×10^-4Corresponds to. Δ_c=8.3×10^-4The second mode cutoff with a greater relative index difference falls between the curves 2100 and 2101, Δ_c=8.3×10^-4The second mode cutoff with a smaller relative index difference will be below curve 2100.

Single-mode operation can be achieved with photonic crystal fibers, which have large clads with many periodic clad features, but with worse bending performance due to smaller d / . I'm making a sacrifice. On the other hand, leak channel fibers that require only a few clad features, such as one to three layers of holes, can achieve better bending performance with a sufficient level of mode filtering.

In addition, the core can be formed as shown in FIG. 19 by omitting more holes in the hexagonal array, such as seven holes, rather than just one hole. (See, for example, Limpert et al., Optics Express, Vol. 14, pp. 2715-2720, 2006 and Schreiber et al., Optics Express, Vol. 13, pp. 7622-7630, 2005). A small hole size of approximately d / &lt;0.25 is used in the design in some embodiments. Similar all-glass designs with low index differences are also possible, but may possibly include holes slightly larger than those used in the case of air holes.

The bending ability of the fiber becomes worse as the core diameter increases. In one embodiment, at 2ρ&gt; 100 μm, a large bending diameter of&gt; 1 m, for example, would be used. Amplifiers incorporating these fibers are primarily designed for use with straight fibers with a short length of less than 1 meter with a highly concentrating impurity-added core. This allows straight or near straight configurations to make this dimension practical. In some cases, the fiber may take the form of short rods with a diameter of 0.3 mm to a few mm. Leakage channel fibers may be designed to operate in a straight line with sufficient mode filtering. This may be done by reducing the diameter of the clad features, resulting in a smaller value ratio d / .

In the above example, silica and fluoridated glass were used as the first clad material and the second clad material, respectively, but by adding a dopant or combination of dopants such as fluorine and boron to the quartz glass. 5 × 10^-3The relative refractive index difference up to can be obtained.

In some embodiments, silica can be used for the first and second materials, but only one of these materials may be impurity-added, including, for example, impurity-added silica. .. In various embodiments, for example, the clad features are impurinated and thus have a lower index of refraction than the background clad material in which the clad features are located.

In other embodiments, a combination of, for example, germanium, phosphorus, tin, titanium, etc. is added to the background material, while the feature is, for example, silica to which no impurities are added. Impurities can also be added to both the background material and the clad features to create a refractive index difference. In other embodiments, impurity-added silica is used for the first and second materials, but may be used at different dopants or dopant levels for the first and second materials.

For example, germanium and / and phosphorus-added quartz glass may be used as the first clad material to achieve a relative index difference of as much as 3%, or fluorine and / and boron-added quartz glass may be the second. May be used for clad materials. These glasses can be made to be of high purity using vapor deposition techniques and can coexist physically, mechanically, chemically and thermally.

Other quartz glasses may also be used. Moreover, other non-quartic glasses such as phosphate glass, terluate glass, chalcogenide glass, bismus glass, fluoride glass and others have also been used to achieve low relative index difference designs. There are times. In some examples, a slightly modified version of the same glass with a slightly lower index of refraction may be used as the second clad material to achieve good coexistence. Other materials may also be used.

Preferred embodiments presented herein for achieving large core fibers include clad features with a low index of refraction. It is worth noting that features with a much larger index of refraction function as if the features were air holes.

Additional LCF characteristics, simulations, and embodiments Figures 15-18 illustrate examples showing confinement loss vs. fiber diameter, coated index of refraction, and wavelength. The diameter is optimized so that the choice of diameter can affect the loss and that it causes both a higher loss for the secondary mode and a lower loss for the basic mode (or almost so). It was disclosed that the operation in the state where it is done may be beneficial. Finite fiber diameters, relative variations in fiber diameter, fiber optic coverage, and wavelength can also affect the mode profile in a variety of ways.

A variety of suitable alternatives are available for coating the fiber, but commonly used coatings are either high index polymer coatings or low index coatings. High index polymer coatings are commonly used for standard optical fibers, while low index coatings allow total internal reflection at the interface between the low index coating and the clad region 103 to allow excitation-guided operation. In the state, it is used to form a double clad fiber that supports multimode excitation.

Back to the LCF in Figure 1-b, between the core formed by the six low index features 102 and the first clad region 103 beyond these six low index features. There may be strong optical coupling.

Further analysis of fiber performance with the coating 104 included is of great benefit, without giving any particular theory. The choice of high index or low index coating affects the propagation characteristics of the mode. Various embodiments may utilize the LCF as the gain fiber to produce high peak or average power. High power multimode excitation is commonly used in combination with double clad fibers.

Seeing Figure 1-b again, an example of an LCF gain fiber is LCF100, where the coating 104 is a low index polymer coating, and multimode excitation is at the interface between the clad 103 and the low index coating 104. Induced by the total internal reflection of. A single spatial mode laser beam is guided within the core 101, at least a portion of which core is supplemented with active ions.

When considering coated fibers, many modes are induced with such double clad LCFs. Figure 22 shows 2ρ = 50μm, d / = 0.7, Δn = 1.2 × 10^-3, And n_CoverFor LCF = 1.37, some of the calculated lower order modes are shown, and these results were obtained by varying the fiber outer diameter. The dispersed refractive index of the background quartz glass was λ = 1.05 μm, which was simulated using a standard empirical formula.

As pointed out above and shown in Figures 15-18, diameter selection can also have a significant impact on confinement loss in basic and higher modes. The basic core mode is represented by wire 2201 (which is substantially horizontal in Figure 22), the second core mode is represented by wire 2202, and the wire 2202 corresponding to the second core mode is prominent at various fiber diameters. Shows a discontinuity. Similar discontinuities occur along that part associated with line 2201, but cannot be clearly identified. All other modes are represented by lines 2203, which appear as a group of curves that are virtually strongly dependent on the fiber diameter, and the mode index of refraction increases overall with increasing fiber diameter. Simulations were performed for a fiber diameter of 125 μm, shown as 2204 in cross section, to a fiber diameter of 300 μm, shown as 2205 in cross section. The basic mode profile 2206 and the second mode profile 2207 with a fiber diameter of 125 μm are shown as an insert view in FIG. 22 and as an enlarged view in FIGS. 22A and 22F, respectively.

The basic core mode of traditional fibers is generally LP₀₁Or TEM₀₁Corresponds to a mode similar to. For example, mode 2206 at 125 μm is a TEM.₀₀Similar to (see Figure 22A). The basic mode is generally the mode with the highest effective index of refraction.

FIG. 22 shows that as the fiber diameter increases, the basic core mode of the LCF is no longer the basic mode. For example, above about 165 μm, the shape of the basic mode changes as illustrated in mode 2208 (see Figure 22B). In addition, the basic mode of the LCF at approximately 235 μm in diameter has a highly non-Gaussian profile, with most of its power in the second clad region 103 beyond the six low index features, mode 2211. (See Figure 22D). In other words, with a sufficiently large fiber diameter, the basic core mode is the higher order mode of the LCF fiber. In any case, unless otherwise specified, a basic mode as specified herein corresponds to a basic core mode.

However, exciting and propagating the basic core mode generally does not require a detailed analysis of each LCF, or anything beyond conventional optical systems. For example, TEM₀₀Excitation in a mode similar to this only requires focusing an input beam with a pattern that is properly correlated with this mode on the core. Closer-to-extreme matching is generally not required, as unwanted higher-order modes are suppressed while propagating the LCF.

The location of FIG. 22 with obvious discontinuities is called "anti-intersection" (eg, modes intersect each other) and is meaningful for relative confinement loss. Strong and weak anti-intersections are shown in FIG. 22, corresponding to relatively large and small magnitude discontinuities of the mode index of refraction, respectively. No two separate modes can have the same mode index of refraction, which results in anti-intersections. Anti-intersections correlate with relative confinement losses in various modes, as shown in FIG. 23, which are further described in the following paragraphs. Strong anti-intersections correspond to strongly interacting modes, for example 2208 and 2209 with similar spatial features. Such modes have a mode index of refraction with close values. One example is the basic core mode 2208 at about 170 μm or the second core mode 2209 at about 135 μm. At weak anti-intersections, for example, at intersection mode 2210 at about 195 μm, spatial changes in mode are more pronounced. Therefore, anti-intersections may be used to enhance mode discrimination and are important in some fiber embodiments.

Many stronger anti-intersections are present in the second core mode. Thus, in embodiments utilized in fiber amplifiers or lasers, the secondary mode may have a greater portion of its power in the gainless cladding. More importantly, in some cases, more reliance on the glass coating boundary for second mode induction is not only perturbation-induced coupling at the glass coating interface, but also coatings by large and microbending and Increase power leakage to other modes.

The above observation further supports that it may be beneficial to choose a fiber diameter that produces both higher losses for the secondary mode and lower losses for the basic mode. In various embodiments, the fiber diameter is chosen such that the second core mode is at a strong anti-intersection. As a result, higher-order mode suppression and significantly improved single-mode operation occur in the core.

While acknowledging that low index coatings are common in double clad configurations, it is worthwhile to further analyze how the choice of low or high index coatings affects propagation. In a double clad LCF with a lossless low index coating, all waveguide modes are theoretically lossless. At the boundary between the glass 103 and the coating 104, total internal reflection occurs for many of the waveguide modes. All modes are induced, but less from the internal clad feature 102 and the mode induced by the interface between the glass 103 and the coating 104 is much more susceptible to large and small bends. The susceptibility can be caused by much larger spatial presence and interfacial defects between the glass and the coating. Power is coupled between these modes and reduces the effective propagation distance of these modes in the core. Under these conditions, strong single-mode propagation in the core occurs only in basic core mode. When the core is supplemented with active ions, only the basic core mode is strongly amplified. However, this is not a decisive factor in the suppression of higher-order modes.

Unlike LCFs with a low index of refraction coating, LCFs with a high index of refraction coating are prone to leakage in all modes. Modes that rely heavily on low index coatings for induction and less on internal cladding features leak out suddenly. In this case, the containment loss gives a good indication of how the mode is induced by the internal clad features and thus provides a good indication of mode discrimination in propagation. When the higher index of refraction coating is replaced with the lower one despite some loss, the respective mode profiles remain intact. Therefore, the analysis in the case of a high index of refraction coating provides a good measure of the mode strength in the case of a low index of refraction coating.

Except for some transmission loss, the peak position of the relative confinement of the LCF in the example shown in FIG. 22 is not significantly affected by the coating choice. n_CoverThe confinement loss of the same LCF in FIG. 22 with a high index coating of = 1.54 and λ = 1.05 μm is shown in FIG. Lines 2301 and 2303 are the effective refractive indexes of the basic mode and the second mode, respectively. The dashed line 2302 and the solid line 2304 are the confinement losses in the basic mode and the second mode, respectively. The peak of confinement loss is from a strong anti-intersection. It is interesting to note that the peak remains in the same position as shown in FIG. 22 with a low index of refraction coating. The basic mode profiles of the examples at and away from the anti-intersection are shown in FIGS. 23 as inserts 2306 and 2305, respectively, and in FIGS. 23B and 23D, respectively, as enlarged views. Secondary mode profiles of examples at and away from the anti-intersection are shown as plug-in FIGS. 2308 and 2307 in FIG. 23 and enlarged views in FIGS. 23A and 23C, respectively.

With reference to FIG. 23 again, a high ratio of confinement loss between the secondary mode and the basic mode can be achieved with some fiber diameters, which in this example is maximized at about 215 μm in diameter. If confinement loss is the only factor to consider, in one embodiment a clear choice for maximizing loss in higher order modes is that this ratio is for small changes in diameter. The fiber diameter is about 215 μm, knowing that it can change rapidly. A significant increase in mode discrimination can be achieved with either a high index coating or a low index coating by designing the fiber with a fiber diameter to operate at a strong second core mode anti-intersection. ..

However, certain design considerations may result in choosing a fiber diameter that produces non-maximum losses in higher order modes in order to achieve other benefits. In various embodiments, the first clad size of the fiber, eg, the circular first clad diameter 2ρ.₀The choice may be based on cost and various design goals. For example, in an amplifier or laser, a design consideration is to excite the gain medium. Due to the relatively high cost of high-brightness excitation diodes, larger fiber diameters corresponding to mode 2308 may be preferred in some embodiments. FIG. 23 shows that the ratio of confinement loss between the secondary mode and the basic mode does not maximize at larger diameters and further exceeds 100: 1. The larger fiber diameter facilitates excitation photocoupling to the excitation guide for relatively low brightness excitation diodes.

Conversely, if the design goal is to increase or maximize peak pulse or CW power, high-brightness diodes and state-of-the-art transfer optics can be used to energy smaller diameter fibers. May be used to combine energy to an approximate diameter that increases or maximizes secondary mode loss and decreases or minimizes basic mode loss. Transfer optics may include small optics for delivering energy from a single bright diode or array or multiple diodes. If the fiber diameter is very large, for example above 300 μm in some embodiments, there are other considerations that affect beam propagation. Thus, some LCF embodiments may utilize inexpensive diodes and large excitation clad dimensions, while others may utilize smaller diameter excitation guides and brighter diodes. is there. In some embodiments, smaller diameter excitation guides and brighter diodes produce increased excitation power over the reduced length of the fiber, resulting in high peak power without significant non-linear effects.

Therefore, in various embodiments, fiber diameters in the range of about 125 μm to 300 μm may be utilized, and the ratio of basic mode confinement loss to secondary mode confinement loss may generally exceed 10. Or it may be in the range of about 10 to 100. In some embodiments, the confinement loss ratio can exceed 100. For example, the confinement loss ratio can be up to about 500, or for fibers several meters in length (eg, 1m, 2m, 5m), it can possibly approach 1000. Higher-order mode losses can exceed 10-20 dB over a length of 5 m and can be in the range of about 1-4 dB / m, 3-6 dB / m or other ranges.

The excitation energy may be induced by total internal reflection at the interface between the clad region 103 and the coating 104, as shown in FIG. 1b. In various embodiments, the low index glass 107 may form an excitation clad, as shown in FIG. 1c. In some embodiments, the excitation cladding may be formed as an array of air holes.

In some embodiments, the clad region 103 may be utilized as an excitation guide. Region 103 may have a diameter in the range of about 150-400 μm and may exceed about 500 μm. Core dimensions can range from about 50 to 150 μm. In some embodiments, the core dimension may be about 20-40% of the excitation guide dimension and also to maximize the loss of higher order modes and / or minimize the bending loss. It may be small enough.

Examples and properties of LCFs with a low index of refraction coating have been used to illustrate the properties of double clad fibers. Other suitable alternatives may be used. For example, in at least one embodiment, low index glass may be used as an alternative to the low index coating. This glass may be high concentration fluoridated silica. Other alternatives may be one or more layers of air holes that replace the low index coating. With reference to FIG. 1B again, the layer of air holes may be formed to replace the coating 104. In addition, additional silica (not shown) may be added beyond the highly fluoridated silica and air holes. Many variants are possible.

Also, the wavelength sensitive characteristics of LCF are interesting. FIG. 24 shows an example of an LCF operating at a strong second core mode anti-intersection with a fiber diameter of 268 μm. Confinement losses 2401 and 2402 in basic mode and second mode were simulated for a range of wavelengths, respectively. The basic mode profile 2405 and the second mode profile 2406 are shown in the inset in FIG. 24 and in the enlarged view in FIG. 24A. Figure 24 shows 337 loss ratios of 2403α._2nd/ α_FMIs achieved at λ = 1.05 μm, and α_2nd/ α_FMDemonstrates well over 100 over a range of 200 nanometers and improves by at least an order of magnitude compared to a standard LCF loss ratio of 2404, without taking into account the region beyond the low index feature. Represents. Resonance enhancement as used herein means enhanced secondary mode suppression, as shown, for example, in FIGS. 23 and 24. However, in some fiber embodiments, it is the pre-selected diameter, not the wavelength, that provides this enhancement. Enhanced higher-order suppression is obtained in some embodiments by choosing at least the appropriate diameter and / or generally operating the LCF at a secondary mode anti-intersection. For example, operating this 268 μm diameter fiber embodiment at a wavelength of about 1.05 μm, as shown in FIG. 24, provides significant suppression of higher order modes (eg, approximately 337 α)._2nd/ α_FM)。

Examples of passive and active LCFs produced and performance data The following paragraphs provide examples of the performance of both passive PCF and LCF power amplifiers. Some interesting parameters include changes in beam quality and loss due to bending, wavelength dependence of propagation, and changes in index of refraction due to stress optics.

Using the well-known stack-and-draw technique, where rods of the same diameter are stacked in a hexagonal shape to form a bundle, they have a variety of core diameters from 30 to 180 μm and d / from 0.7. Several LCFs of 0.9 were produced. The bundle is then inserted into the tube and drawn to the fiber with proper atmosphere and vacuum control inside the tube. Ytterbium-added fibers have also been produced. Non-circular cross section 2407 is shown in Figures 24 and 24B. LCF is 2ρ = 52.7μm, d / = 0.8, and n_Cover= Has a coating of 1.37. The glass outer diameter is 254.2 μm (eg, between planes measured from a substantially flat surface to a substantially flat surface on the opposite side), and the effective mode area is approximately 11 dB at 976 nm. 1548 μm with measurement excitation absorption of / m²Was calculated. Ytterbium ions are added to the central part of the core. Ytterbium-added glass was made using a process specially developed to produce glass with excellent index uniformity and accurate index matching with the rest of the fiber. These steps maintain the desired mode pattern as well as strong single-mode operation in this example.

In the first test of the strength of single-mode propagation of this LCF, the output of a loosely wound 6 m fiber was paralleled to a coil about 50 cm in diameter and imaged by a CCD camera while adjusting the input conditions. It was. Note that the LCF in this example is a double clad ytterbium-added fiber, but this fiber was used in an unexcited configuration in this experiment. Nevertheless, if the higher order modes had much larger propagation loss even in the unexcited double clad amplifier fiber, the output mode pattern would be even when the injection conditions were adjusted to never be optimal. It was stable and no higher-order modes were observed. During this first test, the output mode pattern remained stable and remained in basic mode throughout the manual handling of the fiber by applying pressure between the two fingers and further bending.

In order to characterize the wavelength dependence of propagation, the output mode in the wavelength range of 480 to 1100 nm after propagation of a super-continuous light source passing through the length of 20 cm of this LCF was captured by a CCD, and the upper inset and the figure of FIG. It is shown as images 2408 to 2414 on 24B. The responsiveness of the silicon CCD was so low above 1100 nm that no data were available. Clear, wide strip-like single-mode operation above 800 nm is evident, while below 800 nm there is evidence of an increase in the content of higher-order modes towards shorter wavelengths. Figure 24 shows that the simulated secondary mode is gradually transmitted towards shorter wavelengths, reaching a loss of about 1 dB / m at 800 nm with this LCF with a 50 μm core. Shown. This LCF has a low index of refraction coating, while the simulation shows good agreement with the experiment, even though the simulated fiber has a high index of refraction coating.

Figure 25 shows 2ρ = 101 μm, d / = 0.8 and n_CoverNumerical characteristics and performance data for a passive LCF with a = 1.54 coating are shown in Insert Figure 2501. The effective mode area of this LCF is 5117 μm²Was calculated. A cross section 2502 of this fiber is shown in plug-in FIGS. 2501 and 25A. The length of this LCF, which is about 6m long, was loosely wound around a 1m coil. M²Is commonly used for beam quality measurements. 1 M²A value corresponds to a "perfect" Gaussian beam profile, and all practical beams have an M of&gt; 1.²Have a value. ASE source and Spiricon M²M measured using -200²Is M²_x= 1.26 and M²_y= 1.29. Figure 25 also shows 2ρ = 183.3 μm, d / = 0.8 and n._CoverThe characteristics and performance of other LCFs with a = 1.54 coating are shown in Insert Figure 2504. A similar conventional single-mode fiber optic (SMF) cross section 2507 is also shown for comparison with cross section 2505 shown in insets 2504 and 25B. The effective mode area of this LCF is 15861 μm²It is calculated that it is a record effective mode area, which is considered to be an improvement of more than two orders of magnitude compared to the conventional single mode optical fiber. M measured using a 1 m straight fiber ASE source²Is M²_x= 1.22 and M²_y= 1.23. The measurement mode patterns 2503 and 2506 for the output of these fibers are also shown in the inserts 2501 and 2504 of FIG. 25, respectively, and further shown in FIGS. 25A and 25B, respectively.

FIG. 25 shows bending loss measurements 2508, 2509, 2510 and 2511 for four LCFs with core diameters of 35 μm, 40 μm, 50 μm, and 101 μm, respectively. The fibers were set in prefabricated annular grooves of various diameters. The transmission at each coil diameter was then measured after the output mode pattern was confirmed. Absolute transmission of fiber was measured by a separate cutback measurement. Absolute measurements were then used to recalculate the relative bending loss measurements. The bending ability of the LCF in some embodiments diminishes very rapidly as the core diameter increases. This effect is the ability of the waveguide mode to go through the bends, how quickly the spatial pattern can be changed without breaking up as the mode propagates, for example, the adiabatic transition can be maintained. Is basically related to what is involved. As the mode grows, this ability to change diminishes very quickly in some cases. The measured critical bend radius of 2 dB / m is plotted as a white circle in the inset of Figure 25 and Figure 25C, ρ.^2.5It fits perfectly with the dependency.

The examples shown in FIGS. 24-25 corresponded to passive fibers, for example, no excitation beam was injected during the test.

It is quite interesting to use a large core LCF for high peak pulse or CW power applications. As an example, the United States published as Figures 10-13 and 16-21 and U.S. Patent Publication No. 2006/0263024 entitled "Single Mode Propagation in Fibers and Rods with Large Leakage Channels" and filed May 30, 2005. The relevant text of Patent Application No. 11 / 134,856 describes various devices and systems / configurations. For example, FIG. 11 of U.S. Patent Application No. 11 / 134,856 illustrates an embodiment of a large mode area fiber containing a rare earth ion-added core used in a fiber amplifier or laser excited by a multimode excitation source. It is shown schematically.

Referring to FIG. 26C of the present application, FIG. 26C illustrates an embodiment of a large mode area fiber containing a rare earth ion added core, which may be used in a fiber amplifier or laser excited by a multimode excitation source. Shown schematically. In this embodiment, the fiber has straight input and output ends 2651, 2652 and a rounded portion in between. Multimode excitation 2655 is used to excite an amplifier or laser using the coupling lens 2654. The input beam 2656 is emitted into the fiber 2650 through the lens 2653. Output 2657 is separated by a dichroic mirror 2658.

First, to test fiber performance in the active region, a single-stage fiber amplifier was assembled based on the 3 m ytterbium-added LCF shown and described in connection with FIGS. 24 and 24A, 24B. FIG. 26 shows the characteristics and performance of this fiber with cross section 2602 (also shown in FIG. 26A) in plug-in Figure 2601. The end of the amplifier fiber was cleaved at about 8 degrees. The amplifier was seeded at a repetition rate of 25 kHz (pulse energy of 5 μJ) and excited in the opposite direction by a microchip laser operating at 1062 nm with a 125 mW output. The measured pulse duration of the microchip laser was 600 ps. Amplifier performance 2603 is summarized in Figure 26. The measured amplification slope efficiency is about 57%. A maximum power of 7 W (320 μJ) was obtained before the start of lasing at an amplifier gain peak of approximately 1026 nm (eg, the available 1062 nm input deviated from the gain peak). M measured at 7W output power²(Shown in plug-in Figure 2605) is M in two orthogonal orientations, respectively.²_x= 1.17 and M²_yIt was = 1.18. Output mode 2606 is also shown in the inset of FIG. 26 and in FIG. 26B.

The potential for handling high peak power with this LCF was tested using a two-stage LCF fiber amplifier configuration. Each stage of the LCF amplifier contains straight input and output ends, with a rounded portion with a coil diameter of approximately 0.5 m between them. The amplifier output was limited by the available excitation power. No obvious traces of self-phase modulation and Raman scattering were observed, so behavior below the non-linear threshold of the fiber was apparent. A spectrum 2605 measured at 400 μJ pulse energy is shown in Figure 26. The amplified natural emission level was more than 40 dB lower for the output spectrum 2605 than the amplified signal level.

In various embodiments, the LCF amplifier and / or laser may allow operation with pulse energies in the range of 10 μJ to 10 mJ, with some embodiments having a preferred operating range of from about 100 μJ to about 1 mJ. May be possible. CW or pulse operation may be used to generate pulses, for example at a repetition rate of KHz or MHz. In at least one embodiment, the output beam may generally have a low numerical aperture and is particularly well suited for fiber transfer of the output beam for applications involving, for example, welding, cutting, and marking. There may be. In various embodiments, the LCF fiber may be used to generate ultrashort pulses, for example in a chirped pulse amplification (CPA) system. Micromachining applications use full-fiber systems, along with suitable beam shaping and magnifying optics, for high beam quality, eg, M in the range of about 1.1 to 1.5.²It may be done using a high peak power pulse of value. In some embodiments, frequency converters may be utilized to generate visible or near-ultraviolet pulses of high peak power. In some embodiments, nearly diffraction-limited spot sizes in the range of approximately 1 μm to 50 μm may be produced.

Various laser or amplifier embodiments may be utilized in all-fiber designs to generate high peak power pulses in the femtosecond, picosecond, and nanosecond regions where non-linear effects are negligible. .. In one example embodiment, a core diameter of about 70 μm to 100 μm or greater may be used to generate nanosecond pulses with energies from a few mJ to about 10 mJ. Ultrashort pulses with an exemplary pulse width of about 1-10 ps may be generated with output energies in the range of about 10 μJ to several hundred μJ.

FIG. 26D schematically illustrates an exemplary laser-based material processing system that may be configured with the large mode area amplifier system 2670 of FIG. 26C. System 2670 may also be configured as a multi-stage amplifier (not shown). Optical system 2672 delivers a nearly diffractively restricted beam to target 2675. Spot size depends on the application, but can generally range from about 1 μm to 250 μm. Optical system 2672 may also include a scanning mirror or other suitable beam positioning device. Goal 2675 may also be mounted on a translation and / or rotary positioning table. Controller 2677 coordinates the operation of the laser, optical system, and any positioning mechanism.

Further characterization of this embodiment of the LCF fiber is influenced by various properties of the LCF fiber, including, for example, the size and / or spacing of the clad features, the coefficient of thermal expansion of the clad and / or the clad features, and so on. Included measurements of changes in the LCF index of refraction.

FIG. 27 shows the measured two-dimensional index of refraction 2700 of this LCF, showing the core 2702 with an impurity-added central portion 2703 and the low index of refraction feature 2701. The portion 2704 around the low index of refraction feature 2701 has an increased index of refraction. This increase in index of refraction can occur due to the different material properties of silica and fluoridated silica, especially due to the different coefficients of thermal expansion δT. Since the fibers are drawn at high temperatures, fluoridated silica with a higher coefficient of thermal expansion tends to shrink more than the surrounding quartz glass. However, this shrinkage is limited by the surrounding silica. At room temperature, the fluoridated silica is under tensile force in the fiber and the surrounding silica is under compression. This stress causes a change in the stress refractive index due to the stress optical effect.

In some fiber embodiments, local changes in the index of refraction may occur as a result of the characteristics and the different thermal properties of the first clad material. However, in some example experiments with LCF fibers, the LCF waveguide mechanism was observed. In some cases, the size, arrangement, and number of features of relatively large features provide LCF waveguide as the dominant mechanism.

In some embodiments, the size and arrangement of features can affect the index of refraction profile of the core region (or other fiber region), eg, increase the relative variation in index of refraction. Increasing the size and spacing of features (eg, expanding the overall dimensions at a constant rate) generally results in larger relative index changes (eg, larger maximum index modulation). The net index of refraction change caused by the stress optics effect may include offsetting the contribution of nearby features, and in some cases the net result depends on the feature spacing. For example, smaller features placed closer to each other generally result in reduced index modulation. In various embodiments, the size and / or spacing of the features may be preselected to adjust the index of refraction profile of the clad and / or core regions. In some embodiments, the material containing the clad and clad features may be preselected to adjust the index of refraction profile of the clad and / or core regions. For example, in some embodiments, these materials are selected at least in part based on the value of the coefficient of thermal expansion of the material. In some embodiments, the clad feature may include fluoridated silica and the clad may include silica. In other embodiments, other materials, such as impurity-added glass and / or non-impurity glass, may be used.

Example of manufactured LCF: N = 2--mode loss and bending loss In various embodiments where the clad features are arranged in a single layer (eg, a ring), as in Figures 1b and 1c, typical values for d / are about 0.65 to 0.9, 0.7 to 0.9, or It may be in the range of 0.75 to 0.85. In some embodiments, at least a second layer of features (N 2) may be located beyond the clad feature 102, and in some cases the general value of d / is about. It may be in the range of 0.3-0.9, 0.4-0.8, 0.5-0.7, or 0.5-0.8. Other ranges of d / may be used for any of the layers of clad features. When two or more layers of clad features are used, the ratio d / may (but need not be) different for each layer of clad features.

FIG. 31 shows a cross section of a manufactured LCF with a two-layer clad feature 3102 with d / = 0.6 that substantially surrounds the core 3101 and an outer clad layer 3103. This LCF has a core diameter of 45 μm and a fiber diameter of about 400 μm. Such a design realizes the LFC of this example with a much improved higher order loss but slightly higher bending loss. The calculated containment loss of the basic and secondary modes of the example LCF with the two-layer clad feature is shown in Figure 32. The basic mode confinement loss 3201 and the secondary mode confinement loss 3202 are plotted for various d / . The outer clad is not included in the calculations in this example. The clad features are assumed to be located in an infinite clad. It can be seen from Figure 32 that hundreds of confinement loss ratios can be easily achieved over a fairly large range of d / . In some embodiments, this ratio can be further improved by more carefully choosing the outer clad dimensions. Some LCF embodiments (N&gt; 2) that include additional layers can further improve the loss ratio, but in some cases at the expense of worse bending loss performance.

Manufactured PCF and example test results Returning to FIG. 19, this figure shows an embodiment of an N = 3 LCF that is distinguished from conventional photonic crystal fibers by several layers of holes.

Endless single-mode PCFs with air holes in the cladding were first reported by Birks et al. On the unique property of significantly expanding the single-mode (SM) region of conventional optical fibers. The direct result of this report is a significant increase in the core diameter of SM fibers. This is beneficial for increasing the peak power of commercial fiber lasers, which is limited by the non-linear effect of tight core confinement.

However, some embodiments of fibers with air holes have some defects. The size of the small air holes in the endless SM fiber is critically dependent on the balance of strong surface tension and pressure during the fiber drawing process and is therefore difficult to reproduce. Air holes must also be sealed in the device to prevent environmental pollution, which can often have a negative impact on mode quality, especially in large core fibers. All-glass PCFs are certainly attractive, both in terms of ease of manufacture and the possibility of adopting similar processing techniques well developed for traditional fibers.

As pointed out above, the size and arrangement of features, the thermal properties of the material, and one or more of the other elements are in the index of refraction profile of the core region (or other fiber region) (alone). , Or in various combinations), for example, to increase relative variation. Local changes can cause refractive index (non-PCF) waveguides. In some embodiments, if this amazing waveguide mechanism is not properly taken into account, the resulting output mode will deviate from the desired or expected form. When properly taken into account, refractive index waveguides have the potential to provide new and interesting waveguides for use with respect to PCFs. The following examples show the effect of index modulation on the waveguide and mode profile in the PCF fiber embodiment.

All-glass PCFs were made with core diameters of d / = 0.35 and 47 μm. A cross section of the fiber 2800 is shown in FIG. 28A and its index of refraction profile 2801 is shown in FIG. 28B. The low index feature 2802 is shown along the index ring 2804, which is elevated around each low index feature 2802 due to the mismatch of thermal properties. The elevated index ring 2804 also creates a high index section in the core 2803. The length of this fiber was kept straight throughout the output mode being measured at various wavelengths. Modes 2810, 2811, 2812, 2813, 2814 and 2815 at wavelengths 780 nm, 800 nm, 910 nm, 980 nm, 1000 nm and 1100 nm are shown in Figure 28C, respectively. Due to the presence of shorter wavelength cutoffs, PCF waveguides become weaker towards shorter wavelengths. The example fiber embodiment is less satisfactorily derived below 780 nm and gives a maximum normalized diameter of 2ρ / λ60.

A portion of the preform for the fiber shown in Figure 28A was drawn into a fiber with a diameter of about 700 μm and a core diameter of about 130 μm. The elevated index of refraction at the center of the fiber, about 80 μm in diameter, begins to induce the single mode shown in FIG. Modes 2901, 2902, 2903, 2904, 2905 and 2906 were measured at wavelengths of 780 nm, 800 nm, 850 nm, 910 nm, 1000 nm and 1050 nm, respectively. Basic mode operation is very strong at 1 μm, with higher mode content below 850 nm. Modes 2911, 2912, 2913, 2914, 2915 and 2916 are modes captured while adjusting the injection conditions at a wavelength of 1 μm. In this example, no other mode can be induced in this adjustment range. In any case, it is very clear from image 2920 that the mode is not guided by normal PCF waveguide. Image 2920 was taken with a cross section of the fiber exposed to light. In FIG. 29, it can be clearly seen that mode 2921 does not extend to the low index feature 2922. Mode 2921 is also substantially centered within the core region of the fiber, the boundaries of which region is defined by the inner layer of the low index feature 2922. The mode shape does not represent the characteristics of the clad feature, for example as illustrated in mode profile 2814 of FIG. 28C, and the mode profile shape of FIG. 28C shows the mode induced by the clad feature. ..

The change in index of refraction within the cross section of the fiber along the line across the center of the fiber with some fluorine-added rods has been measured and is shown in FIG. The elevated index core 3001 has a diameter of 2ρ. Also, the index of refraction recess 3002 due to the fluoridated glass is shown along with the index ring 3003, which is increased by the stress effect around the fluoridated glass. In some embodiments, the index of refraction change in a portion of the core may be approximately parabolic. This change in index of refraction may lead to the induction of the basic mode in a part of the core having a non-uniform index of refraction. For example, the mode diameter can be a portion of the core diameter, for example about 50%.

Non-PCF waveguides were a surprising result of this experiment. This is whether the conventional PCF waveguide on a large core fiber is a fiber embodiment in which only small air holes are formed in the first clad region, or a configuration in which a material other than glass is used. It implies that the holes may be limited to gas-filled configurations. This result also implies that some possible PCF designs are less favorable for all-glass large core fibers. In some large core embodiments, glass with a well-matched coefficient of thermal expansion may be utilized. Reducing the core size to, for example, 50 μm, and in some cases, can generally improve performance.

Unlike various LCF embodiments, the sequences and relatively small feature sizes chosen for this PCF example increased the local index of refraction change. This local change may be used for non-PCF waveguides.

In at least one embodiment, the all-glass fiber may include a first clad material with a first coefficient of thermal expansion. An additional layer N 2 of clad features may be placed in the first clad material, and these features may be smaller in size than the typical LCF clad feature size. Clad features may include a second clad material with a second coefficient of thermal expansion. There may be a local increase in index of refraction adjacent to the clad features. In addition, the boundaries of the core region may be defined by the first inner layer of clad features. A portion of the core region may exhibit a non-uniform index profile and form a refractive index gradient, as shown in FIG. With reference to FIG. 30, the measured exemplary relative index difference from the peak of the core region 3001 with the increased index to the minimum 3005 is about 5 × 10.^-4Less than, generally about 1x10^-3May be less than. The increased local index of refraction above each micro 3005 corresponds to the transition to the low index clad feature. The local gradient from peak to minimum is large enough to cause basic mode index waveguides in at least part of the core region. The relative index difference may be caused by the stress optical effect.

In various embodiments, the diameter of the large core fiber can range from about 30 μm to 200 μm. Applications of such fibers may be found to broaden the spectrum of input pulses, for example in high power chirped pulse amplification systems, nonlinear amplifiers, and continuous light generators. Such high peak power pulses are strong enough to exceed the non-linear threshold of the fiber medium. In some embodiments, the preamplifier or power amplifier may be formed by adding impurities to the core.

Various embodiments may utilize a "whole fiber" design, or at least a configuration that reduces the number of bulk optics. Such embodiments can include any of the exemplary "whole glass" designs described above. Integrated configurations may be used with N = 1, 2, 3 or more layers, polarized or unpolarized holding fibers, various excitation guide sequences, coatings, and various cladding configurations. Integrated designs may be used for lasers and amplifiers, including multi-stage designs. Various integrated designs can be particularly beneficial for fiber arrays, such as high power coherent arrays. The device may be utilized in exemplary applications including material handling, laser radar, and telecommunications.

In some embodiments, various combinations of amplifiers and transmission fibers may be joined and nevertheless efficiently combine energy while maintaining beam quality. For example, the lens 2653 in Figure 26C or other bulk optics may not be needed. Junctions may be used to efficiently couple the input beam to the large core fiber 2650 while preserving beam quality that is largely limited by diffraction.

FIG. 33A (not magnified at a constant rate) schematically shows an LCF3320 with a large core 3315 with rare earths added and a clad feature 102. A single-mode (or several-mode) fiber 3300 with a core 3305 will be joined to the fiber 3320. In this example, the outer diameters of the fibers 3300 and 3320 are approximately equal, but the fibers may be chosen with unequal outer diameters. The fibers will be joined near plane 3310.

As an example, a commercially available CORNING ® H11060 single-mode specialty fiber may be used as the single-mode fiber 3300. Fiber parameters include a mode field diameter of about 6.2 μm, a length of about 1.2 m, a numerical aperture of about 0.14, and a cutoff wavelength of 920 nm ± 50 nm. The LCF3320 may be a manufactured all-glass fiber with a length of about 80 cm, a core of 53 μm, a corresponding mode field diameter of about 43 μm, and a numerical aperture much smaller than the fiber 3300. In this example, the LCF3320 will be aligned and joined to the single mode fiber 3300.

The maximum power coupling expected for a direct butt coupling between two single-mode fibers may be calculated as follows (eg, Introduction to Fiber Optics, Cambridge University Press, 1998, p. 153, Ghatak. , See "Butt coupling between single mode fibers"). T₀= (2 * ω₁* ω₂)²/ (ω₁²+ ω₂²)², DB_loss= -10log₁₀(T₀) Where ω₁And ω₂Is the diameter of each mode. If the two fibers 3300 and 3320 mentioned above are directly coupled (butt-coupled), a loss of approximately 11 dB would be expected. Therefore, some mechanism is needed to widen the mode field or otherwise align the mode field to prevent such loss.

In one example, the junction was made using a light source to allow measurement and alignment of the cores of the two fibers. As a first step, the basic mode was excited and confirmed by forming an image on the CCD camera. The camera was then replaced with an optical power meter and the loss was measured. The joining process was carried out by heating the area near plane 3310 while establishing the alignment of the two fibers.

A surprisingly efficient binding sequence has been demonstrated. Without giving any particular theory, the results showed the dopant of the core material 3305a diffused during the junction, increasing the core size of the fiber 3300. Multiple electric arcs (not just a single arc) can be used to heat the ends of the fibers and join the fibers together (eg, melt bonding in this example), the mode field of the single-mode fiber 3300. Was made even larger. A small loss of 2.8 dB at a wavelength of approximately 1060 nm and a reduced loss of approximately 1.5 dB at a wavelength of 1.3 μm were observed. The observed loss is much smaller than what might be obtained with a direct coupling (eg, 2.8 dB vs. 11 dB in this example). A measurable change in efficiency may be found on a fiber-by-fiber basis, but the demonstrated reduction in loss is significant.

FIG. 33B schematically shows the LCF3320 and single-mode fiber of FIG. 33A after bonding. The dimensions of core 3305a are shown to be wider than core 3305. As a result, the input beam can be transmitted from the fiber 3300 to the LCF3320 with low loss and without significant sacrifice in mode quality.

FIG. 33C schematically shows an example of a mode profile for a clad feature, exemplifying basic core mode propagation and LCF waveguide. Almost 1 / e²The beam diameter of 3330 is shown, which corresponds to almost 80% of the core dimensions.

FIG. 33D is a photograph showing the output of the manufactured LCF, in which case the LCF and the single mode fiber were joined as shown in FIG. 33B. The picture shows the basic core mode for the clad features and demonstrates LCF waveguide. The mode diameter for the clad feature corresponds well to that schematically shown in Figure 33C.

Figures 33A-33C schematically show one example where a single-mode fiber and an LCF have the same outer diameter. LCF junctions are less restrictive and the diameter of a single mode fiber can differ from that of an LCF.

In some embodiments, intermediate fibers are used and the junction may be made at each end. Such sequences can allow efficient binding to LCFs with larger cores than the 53 μm cores in the above example.

Large core fibers can be used in a variety of devices, systems and applications. For example, large core fibers can be used as passive transmission fibers for high peak power pulsed lasers as well as CW lasers. Rare earths may be added to the core to operate with a fiber laser or laser amplifier. High peak power amplification is often used for ultrashort pulses with picosecond or femtosecond pulse widths, for example in the configuration of direct amplification systems and chirped pulse amplification systems.

As an example, related to Figures 10-13 and 16-21, and US Patent Application No. 11 / 134,856 filed May 30, 2005 entitled "Single Mode Propagation in Fibers and Rods with Large Leakage Channels". The text is specifically referenced and these are various devices and systems that can be modified to take advantage of the photonic crystal fiber, endless single-mode photonic crystal fiber and leak channel fiber configurations disclosed therein. The configuration of is shown. For example, embodiments of the present invention may be utilized in many optical systems including amplifiers, lasers, short or ultrashort pulse generators, Q-switched lasers and other systems. Embodiments of the present invention may be used in systems that utilize infrared, visible, and / or ultraviolet wavelengths to modify metal, semiconductor, and dielectric materials based on lasers. Other uses are also possible.

Various alternative configurations are also possible. For example, components (eg, layers) may be added, removed, or rearranged. Similarly, processing and method steps may be added, removed, and rearranged.

Accordingly, the invention described herein has been disclosed in the context of certain preferred embodiments and examples, but the invention goes beyond specifically disclosed embodiments to other alternative embodiments of the invention. It will be appreciated by those skilled in the art that it extends to forms and / or uses and obvious modifications and equivalents of the present invention. Moreover, although some modifications of the present invention have been shown and described in detail, other modifications within the scope of the present invention will be readily apparent to those skilled in the art based on this disclosure. .. It is also believed that various combinations or partial combinations of specific features and embodiments of embodiments have been made and may still be included within the scope of the invention. It should be understood that the various features and aspects of the disclosed embodiments can be combined or replaced with each other in order to form the various modes of the disclosed invention. Therefore, the scope of the invention disclosed herein should not be limited by the particular disclosed embodiments described above.