JP6066876B2 - Method for manufacturing surface nanoscale, axial, and photonic devices - Google Patents

Description

This application claims the benefit of US Provisional Application No. 61 / 705,258, filed Sep. 25, 2012, and is hereby incorporated by reference.

  The present invention relates to a method of manufacturing a surface nanoscale axial photonic (SNAP) device, and more particularly, utilizes a “characterization and correction” process to determine the effective radius along the length of the SNAP device. It relates to locally modifying and creating the desired optical properties with sub-angstrom accuracy.

Miniature resonant photonic devices well known in the art are made from coupled high-Q cavities (eg, ring resonators, photonic crystal resonators, etc.). Resonance is the result of circulating a whispering gallery mode (WGM) created within a circular structure (such as around the outer periphery of an optical fiber), and the WGM moving around the outer periphery of the structure is repetitive at grazing incidence. Receives internal reflection. Light leakage can be very small within such a structure, resulting in a high intrinsic Q value. The Q factor is generally defined as a measure of energy loss relative to the energy stored in the resonator (or any type of vibrating device) and can be characterized by the center frequency of the resonator divided by its bandwidth (“ Typical values for “high Q” resonators are on the order of 10 ⁹ or more). Thus, the preferred “high Q” resonator is associated with a relatively narrow, sharp peak resonance feature.

  Conventional resonator structures are created by creating features whose size is greater than or equal to the wavelength of the propagating optical signal. For example, known rings or toroids or spheres are typically tens of microns in size. Such structures are typically created using lithographic techniques (eg, etching a silicon material to create a feature pattern) with the undesirable consequences of surface roughness. Lithography-related roughness results in scattering of the propagating optical signal, which reduces the Q value of the device. Furthermore, inaccuracies in conventional manufacturing processes limit the accuracy with which multiple devices can be coupled together to form more complex structures. While it would be useful to create a resonator with smaller dimensions (ie, sub-wavelengths) that provide certain advantages with respect to performance, such small dimensions impose additional manufacturing difficulties.

  Previously, we have used perturbations of the sub-wavelength size of the fiber radius to create a cavity resonator, and various complex coupled photonic micros inside and along the optical fiber. Developed a device. Multiple microstructures can be formed along a given length of optical fiber and combined together to create complex photonic microdevices. Details of this device structure can be found in US Publication No. 2012/0213474 dated 23 August 2012 and is incorporated herein by reference.

  However, when attempting to create a relatively long chain of such devices, its performance begins to be compromised by manufacturing errors, which grow with the chain length. One source of error may be nanometer-scale non-uniformity of the fiber radius, which then continues cumulatively and may affect all devices along the chain. Other sources of manufacturing error in creating a long chain of microresonators include, but are not limited to, fiber surface contamination, system alignment imperfections (i.e., to produce effective radius variation in the first example) System used), fluctuations in beam power used to produce effective radius variations, non-uniform doping profiles within the photosensitive fiber, and the like.

US Publication No. 2012/0213474

  The need remaining in the art is addressed by the present invention, which relates to a method of manufacturing a surface nanoscale axial photonic (SNAP) device, and more particularly, utilizing an in-line correction process, It relates to locally modifying the effective radius along the length of the SNAP device to create the desired optical properties with sub-angstrom accuracy.

  According to one embodiment, the present invention describes a method for characterizing and correcting the effective radius variation of a surface nanoscale axial photonic (SNAP) device comprising a plurality of separate optical microdevices. The method includes (1) characterizing the as-manufactured SNAP device, determining a local effective radius value for each optical microdevice, and calibrating the as-manufactured SNAP device to obtain a predetermined “process” (ie, Determining an appropriate correction factor, defined as the change in effective radius associated with the time-dependent annealing process or UV radiation exposure), and then (3) correcting individual microdevices by application of several processes And a step in which the number of processes applied to each microdevice is determined by a necessary correction amount and a correction coefficient determined in the calibration step. Several iterations of characterization and correction operations can be performed, and effective radius variations of less than angstrom are achieved. An apparatus for carrying out this method is also disclosed.

  In another embodiment, the present invention describes an apparatus that performs characterization and correction of resonance characteristics of surface nanoscale axial photonic (SNAP) devices. The apparatus applies a characterization stage that measures local resonant wavelength values for the individual optical microdevices forming the SNAP device, and a predetermined number of processes (annealing process or UV radiation exposure) on the individual optical microdevices. An exposure stage, wherein a predetermined number of processes is calculated based on known annealing energy and duration, or a known change in effective radius associated with UV radiation exposure.

  These and other embodiments of the present invention will become apparent during the following discussion by reference to the accompanying drawings.

  Reference is now made to the drawings, wherein like numerals represent like parts in the several views.

FIG. 3 illustrates an example configuration that supports whispering gallery mode (WGM) resonances within a tapered section of an optical fiber. FIG. 5 shows an alternative resonant structure created in a “bottle” tapered region of an optical fiber. FIG. 2 illustrates a surface nanoscale axial photonic (SNAP) device formed to include a plurality of microresonators that are distributed along the longitudinal extent of an optical fiber. FIG. 4 illustrates an exemplary annealing process that can be used to create the SNAP device of FIG. FIG. 4 illustrates an exemplary UV radiation process that can be used to create the SNAP device of FIG. 3 with a photosensitive optical fiber. FIG. 4 illustrates an exemplary calibration (characterization) and correction (exposure) configuration formed in accordance with the present invention to characterize and correct a resonant feature of a SNAP device, such as the device shown in FIG. FIG. 7 is an enlarged view of an exemplary portion of the calibration configuration of FIG. FIG. 8 (a) is a surface plot of a set of resonators after initial formation of a SNAP device, including a series of surface plots associated with a SNAP device including a set of 30 resonators formed in accordance with the present invention. FIG. 8 (b) is a surface plot of the same SNAP device after the calibration operation, and FIG. 8 (c) shows the SNAP device after performing individual correction operations for each microresonator according to the present invention. It is a surface plot. 8 includes a series of effective radius variation plots associated with the measured surface plot, plot 1 associated with FIG. 8 (a), plot 2 associated with FIG. 8 (b), and plot 3 associated with FIG. It is a figure relevant to c).

  FIG. 1 illustrates an exemplary configuration utilized to create a WGM in a tapered section of an optical fiber, as fully described by our co-pending application cited above. As shown in the figure, a section of the optical fiber 10 (hereinafter referred to as “device fiber 10”) is formed so as to include a tapered region 12, and the tapered region 12 is formed with a taper on the nanometer scale. That is, the radius of the device fiber 10 is formed to decrease on a nanometer scale depending on the length of the fiber. The exemplary length l of the tapered region 12 is generally on the order of the wavelength of the propagating optical signal (e.g., for a 1.3 or 1.5 μm input signal, a tapered region having a length l on the order of 1 micron is suitable. ) As described in more detail below, fiber radius modification (ie, “tapering”) is a physical change in the actual radius of the fiber, a local modification of the refractive index of the fiber, or a change in physical radius and refractive index. Both changes can be included, all referred to herein as changes in the “effective radius” of the optical fiber.

  Continuing with the description of FIG. 1, an optical microfiber 14 provides an input optical signal to the device fiber 10. In general, a “microfiber” is defined as an optical fiber having a diameter on the order of about 0.1 to 10 times the propagating wavelength, and for a 1.5 μm signal this is about 0.15 to 15 microns in diameter. Become. Any suitable type of optical waveguide device that produces evanescent coupling can be used to provide the input signal to the device fiber 10 and this discussion merely uses the term “microfiber” for convenience. Please understand that. Referring to FIG. 1, the optical microfiber 14 is sufficiently close to the device fiber 10 so that evanescent coupling occurs and at least a portion of the optical signal propagating along the microfiber 14 is transferred to the device fiber 10. Be placed.

  A light source 16 is shown as being used to introduce an optical signal 0 into the microfiber 14. As shown in FIG. 1, as the optical signal 0 propagates along the microfiber 14, a portion is evanescently coupled to the tapered region 12 of the device fiber 10, and the device fiber 10 in the device fiber 10 and the microfiber A WGM is created in the vicinity of the overlap with the fiber 14. The optical signal 0 continues to propagate along the microfiber 14 and is finally coupled to the detector 18, as discussed in detail below, the detector 18 measures the characteristics of the received signal and The resonant behavior in the fiber 10 is monitored.

The characteristic length of WGM along the device fiber 10 discussed below is on the order of Δz to 100 μm. As a result, nanometer-scale fiber radius variation along this length corresponds to a very large radius of curvature r on the order of 1 meter. In the configuration of FIG. 1, a resonance associated with the WGM is a turning point z _t, microfibers 14 are confined entirely between the point of attachment to the device fiber 10 (shown as in Fig. 1 point z _l) Can be shown.

FIG. 2 shows a configuration in which device fiber 10 includes a “bottle” region 20 instead of a tapered region 12 as shown in the device of FIG. Referring to FIG. 2, region 20 of device fiber 10 includes a first portion 22 having a monotonically increasing fiber radius followed by a second portion 24 having a monotonically decreasing fiber radius. Again, these monotonic increases and monotonic decreases are on the nanometer scale. The microfiber 14 is used to excite the WGM within the bottle region 20 as described above. In this configuration, the WGM is confined along the device fiber 10 between the turning points z _t1 and z _t2 .

In the present invention, it is assumed that device fiber 10 (specifically, tapered region 12, 22, or 24) has a small nanometer-scale radius variation, defined as follows.
r (z) −r ₀ = Δr (z)
Where r ₀ is the nominal fiber radius and z is defined as the fiber axis. The WGM excited by the microfiber 14 undergoes resonance in the vicinity of the wavelength defined by
λ = λ _q = 2πn _eff r ₀ / q
In the above equation, q is a large positive integer, and n _eff is the effective refractive index of WGM. In this analysis, the microfiber 14 has a wavelength λ _c near resonance, ie, | λ _c −λ _q | << λ _q / q, at the surface of the device fiber 10 at position z = z _1. It is considered to be a point source of emitted coherent light. For the configuration shown in FIG. 2, it can also be shown that the radius variation Δr (z) supports a fully localized state (also referred to as “bottle state”). With a conventional optical fiber having the same parameters as defined above, a Δλ _FSR value of about 2 pm is obtained with the configuration of FIG.

  The phenomenon described above has now expanded research into more complex devices based on the ability to generate WGM within a section of optical fiber with this type of effective radius variation. Specifically, surface nanoscale axial photonics (SNAPs) are created by smooth and dramatically small nanoscale variations in optical fiber radius and / or its refractive index (ie, “effective radius variation”). This is an emerging field of research on microscopic optical devices.

  FIG. 3 shows an exemplary SNAP device 30 of a plurality of N microscopic optical devices formed along a single optical fiber. In this example, the SNAP device 30 includes a plurality of N microresonators 32 disposed along the length of the optical fiber 34. As shown, resonators 32-1, 32-2,. . . , 32-N are formed along the fiber 34 and separated by an inter-resonator spacing of (for example) about 50 μm. In this device, the introduced WGM circulates around the outer periphery of the optical fiber 34 within the confinement of each separate resonator 32-i and propagates slowly along the longitudinal axis z of the optical fiber 34. The direction of light propagation of the SNAP device 30 is defined as that light traveling along the longitudinal axis z, and the fiber in each microresonator before the light begins to transfer its energy to the next microresonator in the sequence. As a result of the periodic rotation of light around the perimeter of the inevitably, it is “slow” (compared to the known speed of light).

Similar to the configuration described above, the optical signal is evanescently coupled via the microfiber 14 to a SNAP device 30 that is coupled between the light source 16 and the photodetector 18. In the particular embodiment shown in FIG. 3, the SNAP device 30 is formed from a “coreless” silica optical fiber 34 having a radius r ₀ on the order of 15-20 μm, for example. By introducing a local variation of the effective radius Δr (z) on the order of several nanometers (nm), a plurality of microresonators 32-1 to 32-N are formed.

  In most cases, SNAP devices are manufactured from drawn silica fiber, which results in low loss and high Q values similar to silica WGM microresonators (but that materials other than silica can be used) Is also contemplated). In addition, the characteristic axial wavelength of such SNAP devices is much longer than the wavelength of light, so that these SNAP devices can investigate basic characteristics of light (e.g., tunneling, point-source deactivation, dark state For example). However, other optical materials other than silica (eg, other oxide and non-oxide glasses, or crystalline materials such as silicon, germanium, indium phosphide) generally transmit appropriately the light emission wavelength of interest. It should be understood that a SNAP device can be formed from any material. Furthermore, such SNAP devices can take forms other than conventional cylindrical optical fibers (ie, spherical, annular, parabolic).

  The advantages of SNAP devices over substrate-based optical microdevices can be attributed to the ability to produce strong micron-scale light localization, along with nanometer-scale variations in optical fiber radius, as described above. Possible to achieve effective radius variation. This accuracy is significantly higher than that achieved with conventional lithographic manufacturing of microphotonic devices (ie, silicon substrate-based photonic devices). For comparison, the lithographic fabrication characteristic reproducibility is on the order of a few nanometers, whereas the SNAP device reproducibility of the present invention can be less than angstroms (at least an order of magnitude smaller).

To date, at least two different manufacturing methods have been utilized to introduce nanoscale effective radius variations for SNAP devices. As shown in FIG. 4, the first method is based on local annealing of an optical fiber using a CO ₂ laser beam. In this approach, a nanoscale modification of the fiber radius is created by the release of tension in the localized portion of the fiber that is exposed to the CO ₂ beam (ie, that portion of the fiber is annealed). This original tension in the fiber is a well-known phenomenon and is created during the process of pulling the fiber from the optical preform and “frozen” into place. Thus, application of the annealing beam relaxes the tension and modifies the effective radius at that point. An enlarged view of section A of optical fiber 40 is shown in FIG. 4 where the local variations in radius associated with performing annealing at different locations can be seen. By changing the focal length of the parabolic lens 42 and changing the beam output and exposure time of the laser source, the axial dimensions of the SNAP microresonator formed using this method can be modified. .

  As shown in FIG. 5, the second method is applicable to photosensitive optical fibers (eg, Ge-doped fibers) and uses UV radiation to modify the effective radius of the optical fiber. In the configuration shown in FIG. 5, the optical fiber 50 is formed to include a specific dopant that introduces photosensitivity to the fiber, and the dopant concentration is related to the amount of effective radius change that can be produced. In the configuration shown in FIG. 5, the UV beam passes through an appropriately configured amplitude mask 52 that causes interference fringes due to UV light in the photosensitive fiber 50. The presence of this UV radiation modifies the properties of the included dopant, and thus modifies the “effective” radius of the fiber 50 in the section where the radiation is applied. Amplitude mask details are determined to create the desired SNAP device structure. An enlarged view of section B of fiber 10 showing the change in effective radius introduced by UV is also shown in FIG.

The annealing method can be thought of as a “step and iteration” process, in which the CO ₂ beam performs annealing at a first position A, moves to a second position B and repeats the annealing process, Each position is associated with a desired microresonator position. Alternatively, multiple microresonator devices can be formed using a single UV exposure as long as the amplitude mask is appropriately formed to define the location of each separate device.

  Theoretically, any of these techniques can be used to form relatively long chains of microresonators along a single chain of optical support material (such as silica fiber) Errors limit the length of the exemplary chain to a relatively small number of microresonators. These errors include, for example, non-uniformity (in nm scale) of the original optical fiber radius, surface contamination at various locations along the fiber, annealing temperature, unintentional output fluctuations, UV exposure fluctuations, photosensitivity Including doping levels that produce While not a problem for larger devices, all of these and other factors contribute to introducing errors with nanoscale-level radius variations, thereby enabling the ability to create SNAP devices in the manner described above. Limited. This is because such SNAP devices rely on angstrom-ohm accuracy corrections to produce resonant behavior. Any small variation in one or more of these identified parameters limits the function of such devices.

  The present invention addresses this remaining problem by introducing a “characterization and correction” process in the manufacture of SNAP devices. As will be described in detail below, “processing” in the form of either (optionally) an annealing process or UV radiation exposure is a way to create the final SNAP device with sub-angstrom accuracy of its resonant features. The effective radius variation produced initially is used to locally “trim”. As will be described below, the details of the process used to apply the correction process are similar to those used to create the device in the first example.

In an exemplary embodiment related to utilizing an annealing process, the change in effective fiber radius is essentially linear between the time dependent exposure parameters of the annealing energy applied to the fiber section. It has been found that there is a relationship. For example, as discussed below, during the calibration portion of the characterization and correction process, processing consisting of 54 ms pulses from the CO ₂ beam produces a 0.054Å change in the effective radius of an optical fiber having a 19 μm radius. . Thus, by knowing the local errors present in each resonator along the chain, the correction part of the process of the present invention utilizes a series of annealing processes to individually adjust the radius of each resonator, Create the desired SNAP device.

  FIG. 6 illustrates an exemplary experimental setup 60 that can be used to calibrate and correct a SNAP device in accordance with the present invention. The illustrated setup 60 includes an exposure stage 62 and a characterization stage 64. This discussion assumes that the first example SNAP device was created using an appropriate method (as shown in FIGS. 4 and 5). Specifically, using a selected process, create a chain of 30 microresonators along the section of the (air core) optical fiber, and before creating the microresonator, the fiber has a nominal radius of 19 μm. Assume that you have. Each resonator was separated by a 50 μm spacing, and in this particular example it was desirable to create a set of resonators that all showed the same resonant wavelength λ of 1564.3 nm (with sub-angstrom accuracy).

  In order to determine the amount of correction that may be required to provide a set of 30 resonators with only a sub-angstrom variation of the resonant wavelength, the first step in the process of the present invention remains as fabricated. Characterizing a set of microresonators (ie, determining the resonant wavelength of each individual resonator). To that end, as shown in FIG. 6, a SNAP device is placed in the characterization stage 64 of the configuration 60, which measures the resonant wavelength of each individual microresonator that forms the SNAP device. Used to.

  As discussed in detail in our co-pending application, which is incorporated by reference, the characteristics of the resonant structure can be characterized by using a microfiber-based set of measurements of the created microresonator. it can. The characterization stage 64 performs this measurement by using a light source 66 to generate a light beam that is then coupled to an optical 68 fiber that tapers to the microfiber 70. As described above, the microfiber 70 contacts the surface S of the SNAP device 72 at the first location S1, so that a portion of the optical signal is evanescently coupled to the SNAP device 72 at location D1. In order for the signal to be coupled to microresonator locations along the SNAP device 72, the WGM introduces perturbations into the optical signal that circulates around the periphery of the SNAP device 72 and continues to propagate along the microfiber 70. The modified optical signal continues to pass through the microfiber 70 and is finally coupled to the detector 74 via the optical fiber 76. FIG. 7 is a close-up view of the microfiber 70 when it contacts the surface position S1 of the SNAP device 72. FIG.

  In order to fully characterize the set of microresonators formed along the SNAP device 72, the microfiber 70 translates relative to the longitudinal axis of the SNAP device 72 and is associated with each series of microresonators along the chain. A measurement is generated. For example, the microfiber 70 can be held by a securing assembly 90 that allows the microfiber 70 to move forward and backward (as shown). The SNAP device 72 itself can be held by a fixed assembly 92 attached to the translation stage 94. In performing the characterization, the microfiber 70 first touches the SNAP device 72 and an initial measurement is made. The microfiber 70 is then pulled back, translation of the microfiber 70 relative to the SNAP device 72 is performed, the microfiber 70 contacts the new surface location, and the microfiber is in several positions along the length of the SNAP device 72. The same applies when contacting.

  FIG. 8 (a) is a surface plot of a 30-chain SNAP device 72, which is formed by a series of measurements using the microfiber 70 as described above. These measurements were obtained with a wavelength resolution on the order of 1.3 pm, and the measurements are separated at 10 μm intervals along the length of the SNAP device 72. Recalling that this particular SNAP device 72 was intended to be formed as a series of 30 “uniform” microresonators, each representing the same resonant wavelength, the results of the initial manufacturing process were analyzed. can do.

  When examining the spectrum of FIG. 8 (a), it is shown that the spectrum shows a fundamental transmission (resonance) band near the desired wavelength λ = 1564.3 nm, followed by a band gap. The resonant (undesirable) wavelength variation Δλ is shown on the scale on the left vertical axis. In this particular example, the resonance wavelength variation is on the order of 6 mm. Curve 1 in FIG. 9 shows the undesired variation in resonance wavelength for the 30 microresonator chain described herein, where the horizontal axis represents the position of each microresonator along the distance along the SNAP device. Defined as a function of As discussed above, this undesirable variation is created by one or more errors in the manufacturing process, and in order to create a long chain of microresonators, these variations are significant, preferably sub-angstroms. It is desirable to reduce it to a level.

  In accordance with the present invention, the variation is reduced by first performing a calibration process to determine the effective radius change (ie, “correction factor”) associated with a given annealing pulse. The correction factor determined to modify the radius of the individual microresonator to improve the overall uniformity of the SNAP device. In fact, this “characterization and correction” process can be performed multiple times, with additional iterations used to fine tune the effective radius of each microresonator until the desired level of accuracy is achieved. . When creating the set of “uniform” resonators described, this “desired accuracy” is, for example, all 30 resonators vary by only ± 0.7 mm (standard deviation 0.12 mm). Creating a SNAP device that exhibits a certain resonant wavelength.

With reference to FIG. 6, an exemplary calibration process is performed by moving the SNAP device 72 to the exposure stage 62. Exposure stage 62 is shown as including a CO ₂ laser beam source 80, a CO ₂ laser beam source 80 sends a beam through the control beam shutter 82 and parabolic lens 84, the annealing beam Energy is focused to a desired location on the SNAP device 72. The exposure time and energy of each time-dependent annealing process is determined by the control shutter 82, and the beam spot size is controlled by the focal length of the parabolic lens 84. A series of controlled duration and output annealing processes are used to modify the effective radius of the set of microresonators along the chain. In some cases, an increased number of annealing processes are applied to each microresonator along the chain, a single process is applied to the first microresonator in the chain, and the two processes are chained. Applied in succession to the second microresonator in the series, a series of three processes applied to the third microresonator in the chain, and so on (i.e., m annealing processes are chained). Applied to the m th resonator along the line). In this particular example, each annealing process has an exposure length of 54 ms, and the relocation of the annealing process for each microresonator moves the translation stage 94 to move the SNAP device 72 incrementally. Controlled.

  At the completion of this calibration step, the SNAP device 72 is returned to the characterization stage 64, which uses the microfiber 70 to repeat the same resonant wavelength measurement process, with each microresonance along the chain. Find the corrected resonant wavelength of the instrument. FIG. 8 (b) shows the resonance then provided by such a “post-calibration” microresonator, and plot 2 of FIG. 9 shows the variation in effective radius along the chain for the measurement of FIG. 8 (b). .

Since the calibration process introduces known corrections for each individual resonator, the results of plots 1 and 2 can be evaluated to determine the effective radius variation associated with a single “process”. This is illustrated by plot 3 in FIG. 9, which is the difference between plots 1 and 2. Plot 3 clearly shows the linear relationship between the number of annealing treatments applied to the microresonator and the change in its effective radius (and hence the change in resonance wavelength). In this particular device, it has been found that a single 54 ms pulse (controlled by shutter 82) from the CO ₂ laser source introduces a change in effective radius +0.054 Å. Here, this value of + 0.054Å is defined as the “correction factor” for the remainder of the process of correcting the microresonator forming the SNAP device 72.

  In fact, after determining this correction factor, each separate resonator forming the SNAP device 72 uses the appropriate number of processes necessary to modify that particular resonant wavelength, thereby exposing the exposure stage. 62 is corrected. For example, assume that the resonant wavelength of resonator # 4 along the chain needs to be increased by 0.27 mm. In this case, based on the determined correction factor, a set of five treatments from the laser source appropriately adjusts its effective radius in a manner that produces the desired resonant wavelength. Resonator # 12 may only require two processes to correct its value, and so on. FIG. 8 (c) shows the result of performing this correction process and creating a set of 30 microresonators along the SNAP device 72 that exhibits a resonant wavelength variation of only about 0.56 mm.

  This example describes a correction process using an annealing procedure, but for photosensitive SNAP devices, a UV radiation process can be applied as well (ie, measurement, calibration, then correction) and a specific time-dependent UV radiation. It should be understood that a process is defined as a specific “treatment” associated with providing the desired amount of change in resonant wavelength.

  As described above, after performing the first set of corrections (using annealing or UV radiation), the SNAP device can be returned to the characterization stage 64 and its spectrum re-determined. Any further modifications related to “trimming” the effective radius of one or more resonator devices may be implemented. This process continues with additional iterations between the characterization stage and the exposure stage until the desired final characteristics of the SNAP device are achieved (in this case, a uniform resonance value along the entire chain) Can do. In fact, the experimental process has been found to produce a set of 30 microresonators with a uniformity on the order of ± 0.15 cm.

  In this particular experiment, it was desired to create a set of resonators that exhibited all of the same resonant wavelength. In general, it should be understood that a SNAP device can be formed to exhibit any desired configuration (perhaps each resonator has a different resonant wavelength, or tunable wavelength, or any combination of these possibilities).

  It should be understood that many variations of the calibration and correction process described above can be used. For example, the power and exposure time of the correction process can be modified, and a low power level produces a relatively small effective radius change, thus resulting in a “finer” resolution. In addition, the calibration process described above utilizes an incremental and increasing series of processes applied to each resonator along the chain, while another calibration process uses only a subset of the resonators. A correction coefficient can be obtained. In fact, referring to FIG. 9, it is clear that the linearity of plot 3 is such that an appropriate correction factor can be determined using a smaller number of resonators.

  Furthermore, the calibration and correction process of the present invention is believed useful with microresonators formed in any suitable optical waveguide material. Silica is just one such exemplary material. As long as the resonance can be measured and calibrated, the process of modifying the individual resonant wavelengths based on the determined correction factor makes it possible to achieve sub-angstrom accuracy.

Claims (21)

A method for characterizing and correcting effective radius variation of a surface nanoscale axial photonic (SNAP) device comprising a plurality of separate optical microdevices formed as a long chain of separate optical microdevices, comprising:
Characterizing the as-manufactured SNAP device and determining a local effective radius value of the plurality of separate optical microdevices along the long chain ;
Calibrating the as-manufactured SNAP device to determine a correction factor defined as a change in effective radius associated with a single predetermined time-dependent effective radius change process, and application of several effective radius change processes wherein the method comprising correcting each of the separate optical micro device along a separate long chain of light microdevice, processing number of the is applied to each of the separate optical microdevices, in the light microdevice by A method comprising a correction determined by a required effective radius correction amount and the correction factor determined in the calibrating step. The step of calibrating comprises:
Applying a predetermined number of processes to a number of said separate optical microdevices along a long chain of said separate optical microdevices, each optical microdevice undergoing a different number of processes about,
Measuring the local effective radius value of each of the processed optical microdevices, and based on the difference between the local effective radius value of the initially fabricated optical microdevice and the local effective radius value after processing The method of claim 1 including determining a correction factor.   3. The method of claim 2, wherein during the applying step, the number of processes applied to each optical microdevice increases incrementally.   The method according to claim 3, wherein the number of processes increases by one.   The method of claim 2, wherein each optical microdevice forming the SNAP device is processed during a calibration step.   The method of claim 1, wherein the effective radius change process is a time-dependent annealing process that introduces a physical change in local radius.   A laser source is used to perform the time-dependent annealing process, and the duration and output of the laser source output are controlled to define the amount of radius change produced by a local optical microdevice. 6. The method according to 6. The method of claim 7, CO ₂ laser source is used.   The method of claim 1, wherein the SNAP device is a photosensitive device and the effective radius changing process is a time-dependent UV radiation process that introduces a refractive index change and produces a change in effective radius.   The method of claim 9, wherein the duration and output of the UV source are controlled to define the amount of refractive index change produced by the local light microdevice.   The method of claim 1, wherein the SNAP device is formed along the length of an optical fiber.   The method of claim 11, wherein the SNAP device is formed along the length of an air-core optical fiber having a nominal radius on the order of tens of μm.   The method of claim 1, wherein the SNAP device is formed to include a plurality of separate optical microresonators.   14. The method of claim 13, wherein the characterizing step is used to determine a resonant wavelength of the optical microresonator as fabricated for each optical microresonator.   The method of claim 1, wherein the calibrating and correcting steps are repeated a predetermined number of times.   The method of claim 15, wherein the steps of calibrating and correcting are repeated until an effective radius variation error of less than 1 angstrom is achieved. An apparatus for characterizing and correcting resonance characteristics of a surface nanoscale axial photonic (SNAP) device comprising a plurality of separate optical microdevices formed as a long chain of separate optical microdevices, comprising:
A characterization stage for measuring a local resonant wavelength value for each of the separate optical microdevices formed along the long chain ;
An exposure stage that applies a predetermined number of effective radius changing processes to each of the separate optical microdevices formed along the long chain , wherein the predetermined number of processes comprises a known energy of each process and An exposure stage that is related to the duration and is calculated based on a known change in effective radius associated with a single effective radius change process . The characterization stage includes a light source and a detector, and an optical microfiber is disposed therebetween to support propagation of an optical signal from the light source to the detector, the optical microfiber being connected to the SNAP. The plurality of separate lights that are in contact with spaced locations along the length of the device, couple a portion of the propagating optical signal to the SNAP device and evanescent, and wherein the detector is formed along the long chain 18. The apparatus of claim 17, introducing a whispering gallery mode (WGM) that perturbs the propagating signal in a manner that measures a local effective radius value of a microdevice.   The exposure stage includes a laser source, a control beam shutter, and a parabolic lens, the laser source for providing a predetermined output annealing laser beam directed to the surface of the SNAP device; The control beam shutter is used to control the power level and exposure time of the laser beam on the surface of the SNAP device, and the parabolic lens controls the spot size of the laser beam on the surface of the SNAP device. The device according to claim 17, which is used for determining.   20. The apparatus of claim 19, wherein the control beam shutter produces a microsecond long annealing beam pulse.   The exposure stage includes a UV radiation source, a control beam shutter, and a parabolic lens, wherein the UV radiation source provides a beam of a predetermined power directed at the surface of the SNAP device; A control beam shutter is used to control the power level and exposure time of UV radiation on the SNAP device surface, and the parabolic lens is used to determine the spot size of radiation on the SNAP device surface. The apparatus according to claim 17.