Patent Application
20190056266
Broadband light sources, super-continuum sources, and Mid-Infrared Fiber Light (MIRFIL) sources are described that generate wavelength in the mid-infrared (mid-IR being wavelengths substantially between 2 to 5 microns) based on nonlinear processes in optical fibers. Examples of nonlinear processes in optical fibers include super-continuum (SC) generation, modulational instability (MI), cascaded Raman wavelength shifting (CRWS), and four-wave mixing (4WM). Examples of optical fibers include fused silica fibers, fluoride fibers, chalcogenide fibers, and tellurite fibers.
Current techniques of generating mid-IR light include the use of optical parametric oscillators (OPOs) or optical parametric amplifiers (OPAs). However, OPOs and OPAs are generally expensive, complicated, and involve moving parts that are prone to misalignment. Alternative techniques for generating mid-IR light involve the use of quantum cascade lasers (QCL). However, QCL's are generally difficult to operate at wavelengths shorter than about 4.4 microns, they put out low output powers, they have relatively low efficiency, and they often required pulsed operation or cryogenic cooling.
A simpler technique for generating mid-IR light is to use laser diodes to pump optical fibers. The MIRFIL can exemplary involve the generation of mid-IR light in optical fibers by pumping with a variety of lasers including laser diodes, solid state lasers, or cladding-pumped fiber lasers. In one embodiment, SC generation is achieved to simultaneously generate a wide band of wavelengths, which can advantageously be used to mimic the black body radiation of hot metal objects or to perform spectral fingerprinting to identify one or more chemical species. The fiber based MIRFIL can be lighter, more robust, more compact, simpler and less costly than the OPA or OPO alternatives. Moreover, the MIRFIL can produce a single spatial mode with minimal requirements for optical alignments. In a preferred embodiment, nanosecond pulses are used to generate mid-IR light. In addition, the MIRFIL approach leverages the enormous investment in telecommunications technologies and the mature fiber platform.
In one embodiment, a spectroscopy system includes a light source comprising an input light source, including one or more semiconductor diodes, configured to generate an input beam that comprises a wavelength shorter than 2.5 microns, one or more optical amplifiers configured to receive at least a portion of the input beam and to form an amplified optical beam having a spectral width, wherein at least a portion of the one or more optical amplifiers comprises a cladding-pumped fiber amplifier, and a nonlinear element configured to receive at least a portion of the amplified optical beam and to broaden the spectral width of the received amplified optical beam to 100 nm or more through a nonlinear effect forming an output beam, wherein the output beam is pulsed. A filter is coupled to at least one of a lens and a mirror configured to receive at least a portion of the output beam, and to deliver at least a portion of the received output beam to a sample. A detection system comprises one or more detectors configured to receive at least a part of the output beam reflected or transmitted from the sample, wherein the detection system is configured to use a lock-in technique with the pulsed output beam, wherein the spectroscopy system is adapted to perform non-contact detection of chemical species within the sample.
Embodiments may include a spectroscopy system including a light source having an input light source, including one or more semiconductor diodes, configured to generate an input beam that comprises a wavelength shorter than 2.5 microns, one or more optical amplifiers configured to receive at least a portion of the input beam and form an amplified optical beam having a spectral width, and a nonlinear element configured to receive at least a portion of the amplified optical beam and to broaden the spectral width of the received amplified optical beam to 100 nm or more through a nonlinear effect forming an output beam, wherein the output beam is pulsed. At least one of a lens and a mirror is configured to receive at least a portion of the output beam, and to deliver at least a portion of the received output beam to a sample. A detection system comprises one or more detectors configured to receive at least a part of the output beam reflected or transmitted from the sample, wherein the detection system is configured to use a lock-in technique with the pulsed output beam, wherein the spectroscopy system is adapted to perform spectral fingerprinting to measure a chemical composition of the sample.
In one embodiment, a spectroscopy system comprises a light source comprising an input light source, including one or more semiconductor diodes, configured to generate an input beam that comprises a wavelength shorter than 2.5 microns, one or more optical amplifiers configured to receive at least a portion of the input beam and form an amplified optical beam having a spectral width, wherein at least a portion of the one or more optical amplifiers comprises a cladding-pumped fiber amplifier, and a nonlinear element configured to receive at least a portion of the amplified optical beam and to broaden the spectral width of the received amplified optical beam to 100 nm or more through a nonlinear effect forming an output beam, wherein the output beam is pulsed. the system includes at least one of a lens and a mirror configured to receive at least a portion of the output beam, and to deliver at least a portion of the received output beam to a sample, and a detection system comprising one or more detectors configured to receive at least a part of the output beam reflected or transmitted from the sample, wherein the detection system is configured to use a box-car averager with the pulsed output beam, and wherein the spectroscopy system is adapted to measure a chemical composition of the sample.
In one embodiment, an optical system for use in an imaging procedure includes one or more semiconductor diodes configured to generate an input beam, wherein at least a portion of the input beam comprises a wavelength shorter than about 2.5 microns, one or more optical amplifiers configured to receive at least the portion of the input beam and to communicate an intermediate beam to an output end of the one or more optical amplifiers, and one or more optical fibers configured to receive at least a portion of the intermediate beam and to communicate at least the portion of the intermediate beam to a distal end of the one or more optical fibers to form a first optical beam. A nonlinear element is configured to receive at least a portion of the first optical beam and to broaden a spectrum associated with the at least a portion of the first optical beam to at least about 50 nm through a nonlinear effect in the nonlinear element to form an output beam with an output beam broadened spectrum. A subsystem includes one or more lenses or mirrors configured to receive a received portion of the output beam and to deliver a delivered portion of the output beam to a sample to perform imaging for characterizing the sample. The subsystem includes an Optical Coherence Tomography (OCT) apparatus comprising a sample arm and a reference arm, wherein the delivered portion of the output beam has a temporal duration greater than approximately 30 picoseconds, a repetition rate between continuous wave and Megahertz or higher, and a time averaged intensity of less than approximately 50 MW/cm2. The output beam has a time averaged output power of about 20 mW or more.
In yet another embodiment, an optical system for use in an imaging procedure having a plurality of semiconductor diodes, each of the diodes configured to generate an optical beam, a beam combiner configured to receive at least a portion of the optical beams from the plurality of semiconductor diodes and to generate a multiplexed optical beam, an optical fiber configured to receive at least a portion of the multiplexed optical beam and to communicate the at least a portion of the multiplexed optical beam to form an intermediate beam having at least one wavelength, and a light guide configured to receive at least a portion of the intermediate beam and to propagate the at least a portion of the intermediate beam to form an output beam. A subsystem includes one or more lenses or mirrors configured to receive a received portion of the output beam and to deliver a delivered portion of the output beam to a sample to perform imaging for characterizing the sample, wherein the subsystem comprises an Optical Coherence Tomography (OCT) apparatus comprising a sample arm and a reference arm. The delivered portion of the output beam has a temporal duration greater than approximately 30 picoseconds, a repetition rate from continuous wave to Megahertz or higher, and a time averaged intensity of less than approximately 50 MW/cm2. The output beam has a time averaged output power of about 20 mW or more.
Another embodiment includes an optical system for use in a spectroscopy procedure having one or more semiconductor diodes configured to generate an input beam, wherein at least a portion of the input beam comprises a wavelength shorter than about 2.5 microns, one or more optical amplifiers configured to receive at least the portion of the input beam and to communicate an intermediate beam to an output end of the one or more optical amplifiers, and one or more optical fibers configured to receive at least a portion of the intermediate beam and to communicate at least the portion of the intermediate beam to a distal end of the one or more optical fibers to form a first optical beam. A nonlinear element is configured to receive at least a portion of the first optical beam and to broaden a spectrum associated with the at least a portion of the first optical beam to at least about 50 nm through a nonlinear effect in the nonlinear element to form an output beam with an output beam broadened spectrum. A subsystem includes one or more lenses or mirrors configured to receive a received portion of the output beam and to deliver a delivered portion of the output beam to a sample to perform spectroscopy for characterizing the sample based on chemical composition, wherein at least a part of the sample comprises skin or tissue. The delivered portion of the output beam has a temporal duration greater than approximately 30 picoseconds, a repetition rate between continuous wave and Megahertz or higher, and a time averaged intensity of less than approximately 50 MW/cm2. The output beam has a time averaged output power of about 20 mW or more.
One embodiment includes an optical system for use in a spectroscopy procedure having a plurality of semiconductor diodes, each of the diodes configured to generate an optical beam, a beam combiner configured to receive at least a portion of the optical beams from the plurality of semiconductor diodes and to generate a multiplexed optical beam, an optical fiber configured to receive at least a portion of the multiplexed optical beam and to communicate the at least a portion of the multiplexed optical beam to form an intermediate beam having at least one wavelength, and a light guide configured to receive at least a portion of the intermediate beam and to propagate the at least a portion of the intermediate beam to form an output beam. A subsystem includes one or more lenses or mirrors configured to receive a received portion of the output beam and to deliver a delivered portion of the output beam to a sample to perform spectroscopy for characterizing the sample based on chemical composition, wherein at least a part of the sample comprises skin or tissue. The delivered portion of the output beam has a temporal duration greater than approximately 30 picoseconds, a repetition rate from continuous wave to Megahertz or higher, and a time averaged intensity of less than approximately 50 MW/cm2. The output beam has a time averaged output power of about 20 mW or more.
One embodiment of a broadband light source comprises one or more laser diodes capable of generating a pump signal with a wavelength shorter than 2.5 microns and a pulse width of at least 100 picoseconds. The one or more laser diodes are coupled to one or more optical amplifiers, which are capable of amplifying the pump signal to a peak power of at least 500 W. A first fiber is further coupled to the one or more optical amplifiers, wherein the pump signal wavelength falls in an anomalous group-velocity dispersion regime of the first fiber, wherein the pump signal is modulated using a modulational instability mechanism in the first fiber, and wherein different intensities of the pump signal can cause relative motion between different parts of the modulated pump signal produced through modulational instability in the first fiber. A nonlinear element is coupled to the first fiber, and the nonlinear element is capable of broadening the pump optical spectral width to at least 100 nm through a nonlinear effect in the element.
In another embodiment, a mid-infrared light source comprises one or more laser diodes comprising a wavelength and a pulse width of at least 100 picoseconds. One or more optical amplifiers are coupled to the pump signal and are capable of amplifying the pump signal. Further, one or more fibers are coupled to the optical amplifiers. In the fibers, the pump signal wavelength falls in the anomalous group-velocity dispersion regime for at least a fraction of the one or more fibers, and the pump signal is modulated using a modulational instability mechanism. A nonlinear element is coupled to the one or more fibers and is capable of generating a super-continuum with a substantially continuous spectrum from at least the pump signal wavelength out to 2.6 microns or longer and wherein the nonlinear element introduces less than 10 decibels of power loss at 2.6 microns.
A further embodiment involves a method of generating broadband light by generating a pump signal, wherein the pump signal comprises a wavelength shorter than 2.5 microns and a pulse width of at least 100 picoseconds. The method further comprises the step of amplifying the pump signal to a peak power of at least 500 W, modulating at least a fraction of the pump signal using a modulational instability mechanism, and broadening the pump optical spectral width to at least 100 nm using a nonlinear effect.
In yet another embodiment, a MIRFIL can use technologies that have been developed for telecommunications. For example, the pump laser can be a laser diode followed by multiple stages of optical amplifiers. The pump can use continuous wave (CW) or quasi-CW light, which may comprise pulses broader than approximately 100 picoseconds. In a preferred embodiment, the mid-IR light generation may occur in an open loop of fiber, preferably a fiber that transmits light into the mid-IR. Advantageously, only a short length of fiber can be used, such as less than about 100 meters, preferably less than about 20 m, and even more preferably less than about 10 m. With this configuration, wavelengths can be generated in the fiber beyond approximately 1.8 microns, preferably beyond approximately 2.2 microns, and even more preferably beyond 2.5 microns.
In a particular embodiment, a MIRFIL can use a laser diode driven pump laser that outputs CW or quasi-CW pulses (greater than approximately 100 picoseconds) followed by a series of fibers, wherein the first length of fiber can be made from fused silica and can be used to break the CW or quasi-CW light into pulses based on the modulational instability (MI) or parametric amplification effect, and then another length of mid-IR fiber, such as ZBLAN, fluoride, tellurite, or a semiconductor waveguide can be used to broaden the spectrum, through the nonlinearity in the medium and a mechanism such as self-phase modulation. In a preferred embodiment, some curvature in the temporal domain can help to generate the super-continuum by causing relative motion between the MI generated pulses. Also, there can advantageously be exchange of energy between MI generated pulses through the Raman effect in the medium. The design of such a MIRFIL can be that the MI-induced pulse break-up may occur primarily in the first section, and the nonlinear spectrum generation may occur primarily in the second section. In a preferred embodiment, the length of the fused silica fiber can be under 10 meters, and the length of the mid-IR fiber can be less than 20 meters.
In another embodiment, super-continuum (SC) generation from the visible or near-IR wavelength range can be accomplished using nanosecond pulse pumping. The SC generation can exemplary be initiated using modulational instability (MI). In a preferred embodiment, the seed for MI may arise from the amplified spontaneous emission from the optical amplifiers or from a near-IR light source, such as a laser diode. In a particular embodiment using fused silica fiber, the SC can cover the wavelength range substantially between approximately 0.8 microns to approximately 2.8 microns. In another particular embodiment using ZBLAN fluoride fiber, the SC can cover the wavelength range substantially between approximately 0.8 microns to approximately 4.5 microns. With control of the fiber loss from the material or from bend induced loss, as well with tailoring the composition of the fluoride fiber, the long wavelength edge of the SC may be pushed out to 5.3 microns or longer. In a preferred embodiment, it may be valuable to add a wavelength conversion stage. In addition, it may be advantageous to have a pulse compression stage following the MI pulse break-up.
In yet another embodiment, wavelength conversion into the mid-IR wavelength range can be achieved based on four-wave mixing (4WM) in fibers. 4WM usually requires phase matching, and a new window for phase matching permits phase matching into the mid-IR. In a preferred embodiment, the phase matching wavelengths can be tuned by adjusting the fiber dispersion profile and tuning the seed wavelength in the near-IR. In a particular embodiment, a solid core or photonic crystal fiber can be used with a tailored dispersion profile, a seed wavelength from a laser diode or a tunable laser in the near-IR can be used to convert light from a near-IR pump to the mid-IR wavelength range.
In another embodiment, the power for the MIRFIL can be scaled up by using a higher power pump laser, such as a cladding pumped fiber amplifier, a cladding pumped fiber laser or a solid state diode-pumped light source. Based on the damage threshold of the particular fiber employed, the core size of the fiber can also be increased to increase the power throughput and output power.
The fiber based mid-IR light source may be an enabling technology for a number of applications. For example, the broadband mid-IR light source may be useful for infrared counter-measures for aircraft protection. Also, the SC light source could be used in chemical sensing, for non-contact or remote sensing of firearms, weapons, drugs. The SC source could also be used for industrial chemical sensing, such as in advanced semiconductor process control, combustion monitoring, or chemical plant process control. Other potential applications include bio-medical imaging and ablation. Moreover, the broadband SC light source could advantageously be used in an optical coherence tomography configuration for semiconductor wafer imaging or defect location. In addition, the broadband light source could be instrumental for applications in the last mile solution, such as fiber to the home, node, neighborhood, curb, premise, etc. More specifically, the broadband light source could enable wavelength division multiplexed or lambda passive optical networks.
Depending on the specific features implemented, particular embodiments of the present invention may exhibit some, none, or all of the following technical advantages. Various embodiments may be capable of covering other wavelength ranges or multiple wavelength ranges. For example, SC generation can cover the visible wavelength range from approximately 0.4 microns to 0.6 microns by using a dual pumping scheme. Some embodiments may be capable of generating bands of wavelengths rather a continuous range of wavelengths, and the bands of wavelengths may also be tunable or adjustable.
Other technical advantages will be readily apparent to one skilled in the art from the following figures, description and claims. Moreover, while specific advantages have been enumerated, various embodiments may include all, some or none of the enumerated advantages.
To provide a more complete understanding of the present invention and certain features and advantages, thereof, reference is made to the following description taken in conjunction with the accompanying drawings in which:
As required, detailed embodiments of the present invention are disclosed herein; however, it is to be understood that the disclosed embodiments are merely exemplary of the invention that may be embodied in various and alternative forms. The figures are not necessarily to scale; some features may be exaggerated or minimized to show details of particular components. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limiting, but merely as a representative basis for teaching one skilled in the art to variously employ the present invention.
Mid-IR light can be generated based on super-continuum in fused silica fibers and mid-IR fibers. Nonlinear waveguides other than fibers can also be used to generate the super-continuum. In one embodiment, SC has been demonstrated experimentally from ˜0.8 to ˜4.5 microns in ZBLAN fluoride fibers and from ˜0.9 to ˜2.8 microns in high-nonlinearity (HiNL) fused silica fiber. The SC originates for laser diode pumping, and modulational instability (MI) initiated SC generation leads to a significant simplification by reducing or eliminating the need for expensive, complicated, mode-locked lasers. In another embodiment, three orders of cascaded Raman wavelength shifting (CRWS) can be observed in sulfide-based chalcogenide fibers below the damage threshold. In one particular embodiment, the pump source comprises a laser diode followed by several stages of erbium-doped fiber amplifiers, in some cases also including a mid-stage modulator. Since in this embodiment the SC or CRWS occurs in meters to 10's of meters of fiber, the entire mid-IR light source can be compact, lightweight, inexpensive and rugged. Although particular experimental conditions are described in the following, other configurations, materials and fiber types can be used within the scope of the invention.
Fiber Dispersion can Determine SC Vs. Cascaded Raman Shifting
To organize and explain the experimental results in various types of fibers tested, a theoretical framework is first established. Various nonlinear processes are observed in fibers, included CRWS and MI. In turn, MI can give rise to the generation of broadband SC. Whether CRWS or SC occurs first in a fiber depends on the wavelength of the pump or the shifted pump with respect to the zero dispersion wavelength Λ0. When the pump is at a wavelength shorter than the zero dispersion wavelength (so-called normal dispersion regime), then CRWS can be first observed. When the pump is at a wavelength longer than the zero dispersion wavelength (so-called anomalous dispersion regime), then MI and SC can be first observed. When the pump lies in the normal dispersion regime, it can experience CRWS, but when the cascaded Raman order shifts into the anomalous dispersion regime, then MI and SC can occur. Thus, to understand the nonlinear spectrum generated in fibers, the position of the zero dispersion wavelength can indicate the expected behavior.
In the Raman effect, a strong pump beam coupled into the fiber can shake the glass matrix, which emits vibrational mode (so-called optical phonons), and then can provide gain to longer wavelengths. The Raman effect can be self-phase matched, and hence the process does not generally require tuning and it can be more-or-less independent of wavelength (the gain coefficient does scale inversely with wavelength, however). One attribute of the Raman effect is that a number of optical phonons can be emitted, or the wavelength can be shifted down through a cascaded Raman process sequentially to longer and longer wavelengths. This emission of a plurality of phonons to shift more than one Raman order is the CRWS phenomena that can be observed in the normal dispersion regime.
Either pure continuous wave (CW) light or quasi-CW light, such as nanosecond or longer pulses, are generally unstable when launched in the anomalous dispersion regime. In particular, the interaction between the nonlinearity and anomalous dispersion can break the quasi-CW inputs into a train of solitons in a process called modulational instability. MI can be considered as a parametric four-wave-mixing process in which the non-linearity explicitly enters the phase matching condition. Note that MI for a single pump wavelength generally phase matches in the anomalous dispersion regime. When MI occurs the peak powers reached in the fiber can be much higher than the powers launched in the quasi-CW light, since the quasi-CW background is usually compressed into short pulses. Further, a curvature in the pulse in time or a range of intensities can lead to collision and energy exchange between the MI-generated pulses, which can be advantageous for SC generation.
For pumping in the anomalous dispersion regime, the combined effects of MI and stimulated Raman scattering can lead to SC generation. MI can cause the break-up of the CW light into short temporal pulses such that those nonlinear phenomena that normally occur for pulsed pumping conditions can also contribute to the SC generation. In contrast, for normal dispersion pumping, CRWS generally occurs first in the fiber, since MI does not generally phase match for a single pump wavelength. As the higher Stokes orders fall into the anomalous-dispersion regime, MI can occur and lead to SC generation.
Because of the relevance of the group velocity dispersion (GVD) for determining the nonlinear behavior observed, a brief review is provided of fiber dispersion. GVD arises because different frequencies of light travel at different speeds in an optical fiber. The total GVD in the fiber is generally the sum of the material dispersion and the waveguide dispersion Dtotal≈Dm+Dw. The zero dispersion wavelength of a fiber Λ0 corresponds to the wavelength where the total dispersion crosses through zero.
The waveguide dispersion arises because the mode distribution between the fiber's core and cladding changes with wavelength. In a solid core fiber, the waveguide dispersion is usually negative, and, therefore, can generally shift the zero dispersion wavelength to longer wavelengths. It should be noted that in microstructure fibers, the zero dispersion wavelength can be shifted to any desired wavelength. Therefore, microstructure fibers can be useful for matching the zero dispersion wavelength to the laser wavelength when the laser wavelength falls outside of the usual telecommunications window between ˜1.3 and ˜1.6 microns.
The material dispersion curves for different fibers tested in the exemplary experiments are illustrated in
Based on the theoretical discussion above and the exemplary use of a pump wavelength near ˜1553 nm, the behavior for different fiber types illustrated in
Because of the relevance of modulational instability (MI), a brief review is also provided next. MI can be a parametric four-wave-mixing process in which the anomalous group velocity dispersion and nonlinearity generally work together to turn CW or quasi-CW light into short pulses. MI can lead to a significant system simplification in SC generation because the CW or quasi-CW light can evolve into narrow and high peak power soliton pulses, thus reducing or limiting the need for mode locked lasers for SC generation. As a simple illustration,
The dynamics of the collision of soliton pulses can be quite complicated, and it may be easier to illustrate through computer simulations. As an example,
The simulations of
Whether CRSW or MI is first seen in the fiber generally depends on the threshold for the different nonlinearities. For instance, in fused silica fiber the MI threshold can be ˜5× less than the Raman threshold. There are generally two components for the Kerr nonlinearity n2: ˜⅘ of n2 is electronic in nature (instantaneous and arising from the UV resonances), while ˜⅕ of n2 is from Raman-active vibrations (imaginary part of this is the Raman gain coefficient). Whereas MI usually benefits from the full n2, CRWS usually only benefits from the Raman-active vibrations. However, the Raman effect is generally self-phase matched, while MI usually requires phase matching. Moreover, MI phase matches for a single pump wavelength typically in the anomalous group velocity dispersion regime. Therefore, in the normal dispersion regime CRWS usually has a lower threshold and MI does not usually phase match for a single pump wavelength. On the other hand, in the anomalous dispersion regime MI can have a lower threshold, and it can initiate the pulse break-up that gives rise to SC generation through a combination of SRS, self-phase modulation and other nonlinearities.
Although particular examples of physical phenomena have been described for MI or CRWS and subsequent SC generation, other techniques are also possible for MI, CRWS and SC. As one particular example, the use of multiple pump wavelengths can lead to MI and SC even in the normal dispersion regime. As another example, the pulse break-up through MI may not need to be as thorough as shown in the simulations to lead to SC generation. Thus, many other combinations of physical phenomena can lead to MI, SC and CRWS.
As one particular embodiment, experiments have been conducted on nonlinear wavelength shifting in different fiber types, including fused silica (high-nonlinearity fiber—HiNL—as well as dispersion compensating fiber), chalcogenide fibers (arsenic-tri-sulfide), and fluoride (heavy metal fluoride ZBLAN—ZrF4—BaF2—LaF3—AlF3—NaF). For the experimental embodiment, the pump set-up uses a pulsed laser diode (˜1.8 nsec pulses near 1553 nm) followed by several stages of erbium-doped fiber amplifiers. Peak powers up to ˜3 kW can be generated by altering the duty cycle of the laser diode from a few hundred kHz down to 5 kHz.
The experimental configuration for the pump laser used to test the different fibers is shown in
The set-up 700 emulates a Q-switched laser system. Although the laser diode cannot easily be Q-switched, the EDFAs have a long upper state lifetime and store up energy between pulses. Thus, a low duty cycle during which the laser diode is turned on can lead to a larger energy per pulse. However, an ASE problem arises because when the laser diode is off, the optical amplifier continues to be pumped and ASE is emitted by the EDFAs. This ASE leads to inefficiency because it can deplete some of the energy from the power amplifier, which ideally would store up more energy before the next laser diode pulse passes.
To solve this problem, one solution that can be used is to block the ASE during the times that the laser diode is off. As a starting point, the experimental configuration of
A second problem that may arise in the last stage power amplifier is nonlinear fiber effects, which then can limit the useable power from the pump system. The last amplifier stage comprises as an example two WDM couplers (for coupling in and removing any residual 980 nm pump) surrounding a highly-doped, large core size, single spatial mode EDFA gain fiber. The gain fiber is selected in this instance to minimize the nonlinear limitations in the final amplifier stage. For example, the high doping level means that a short fiber length can be used, and the large core size means that the intensity is kept as low as consistent with a single spatial mode. In this particular embodiment, the power amplifier uses a ˜1.2 m length of gain fiber, and a forward pump is used in addition to a backward pump. With this set-up, the peak power for 5 kHz repetition rate approaches ˜2.5 kW without any significant nonlinear effect in the amplifier.
One aspect of using nanosecond pulses with peak powers up to ˜2-3 kW is that the average power can be scaled up by increasing the repetition rate and using larger lasers, such as cladding pumped fiber amplifiers or lasers. For example, average powers in the range of 1 kW to 15 kW are available from commercially available cladding-pumped fiber lasers. Moreover, fiber lasers can be modulated or Q-switched or a modulator can be placed after the fiber laser to generate nanosecond pulses relatively easily. Cladding pumped fiber lasers can operate at a number of wavelengths. For instance, ytterbium-doped cladding pumped fiber lasers operate near 1 micron, erbium-doped cladding pumped fiber lasers operate near 1.55 microns, and thulium-doped cladding pumped fiber lasers operate near 2 microns. Alternatively, the laser could a solid state laser or a diode-pumped laser. Although a few examples of high power lasers are mentioned above, many other lasers can be used consistent with the scope of this disclosure.
In one particular embodiment, a high power pump can comprise a seed laser diode that may be modulated followed by several stages of amplifiers that are single-mode fibers or cladding pumped fiber amplifiers. For example, the first stage pre-amplifier can be a single-mode fiber, such as an erbium-doped fiber amplifier. Then, the power amplifier can comprise one or more stages of cladding-pumped fiber amplifiers. In a cladding pumped fiber amplifier, the pump propagates through a fiber cross-sectional area that is typically larger than a signal cross-sectional area. A cladding-pumped fiber amplifier may comprise a fiber with a core for a signal and the pump that zig-zags through the signal and provides gain. As one particular example, the cladding-pumped fiber amplifier may be a double clad fiber, with the signal propagating through the core and the pump propagating through the inner cladding. For high gain systems, the cladding-mode fiber amplifier may also be a large mode area fiber, which generally means that the core is large enough to support several modes. The cladding-pumped fiber amplifier can be doped with erbium or a combination of erbium and ytterbium. Spectral and/or temporal filters may be advantageously used between different amplifier stages to control the level of amplifier spontaneous emission. Also, particularly in the last few stages of amplification, it may be advantageous to counter-propagate the pump from the signal, thereby reducing the nonlinear effects in the amplifier.
In one preferred embodiment, the pump laser 800 can be a modulated laser diode 820 followed by a parametric amplifier 860. For example,
The signal seed originates in the lower arm 880 from a 1550 nm laser diode 820, such as a DBR, DFB or Fabry-Perot laser diode. The light from the seed laser diode 820 is pre-amplified in a single-mode, EDFA 840 in this embodiment. Then, the about 1550 nm seed is boosted in a power amplifier, which in this preferred embodiment is an optical parametric amplifier 860 (OPA). The OPA 860 comprises a periodically-poled lithium niobate crystal in a preferred embodiment. The OPA crystal can be in length between several millimeters to several centimeters. The pump at 1064 nm 850 and the seed at 1550 nm 880 are made collinear through the OPA crystal 860, and the 1550 nm light is power amplified through the OPA process. Although one example of the OPA is discussed here, many other power boosting methods can be used within the scope of this disclosure.
Another aspect of the laser is that it is relatively simple to modulate the mid-IR light or the SC light. Rather than implementing a mid-IR modulator or a very broadband modulator, the modulation can be done on the pump laser. For example, the pump laser can be modulated directly (i.e., modulating the pump laser diodes or the power supply) or externally modulated (i.e., place a modulator after the pump laser). Then, in the SC generation process, the modulation is transferred to the entire broadband spectrum. In other words, all of the optical processing can be performed at the pump wavelength, and then the light can be shifted to other wavelengths in the last step. This approach is particularly attractive when the pump laser is at a telecommunications wavelength, since many modulator technologies are available for telecom wavelengths.
The pump laser described above is just one embodiment of the pump laser, but many other pump lasers can be used. For example, the pump laser can be a cladding-pumped fiber amplifier or laser, a diode-pumped solid state laser, or either of the lasers followed by a cascaded Raman wavelength shifter. The cascaded Raman wavelength shifter may be an open loop piece of fiber, or cascaded resonators formed by placing gratings on one or more ends of the fiber. Thus, many different configurations for the pump laser can be used consistent with the SC or wavelength conversion process.
To generate light in the mid-IR, one exemplary strategy is to test a number of different kinds of fibers. The starting point may be to use fused silica (SiO2) fiber, since it is of the highest quality and because it is the best characterized fiber. For example, fused silica fiber is the basis of most fiber optics communications. In addition, fused silica fiber has among the highest damage threshold (˜50 MW/cm2), which means that it can extend to the highest output powers. Moreover, the physics of fused silica is well understood and the parameter values can be measured carefully, permitting detailed understanding of the mechanisms behind SC generation and wavelength conversion. Furthermore, the dispersion can be tailored in fused silica fiber, different types of fibers can be spliced together to create a particular dispersion profile, and more exotic fiber geometries, such as photonic crystal fibers or microstructure fibers, can be implemented in fused silica.
Although fused silica is the starting point, the transmission of fused silica is limited in the mid-IR. Therefore, with the understanding gained from fused silica, the next step in the strategy can be to use fibers that transmit in the mid-IR. One attractive candidate for mid-IR transmission is ZBLAN fluoride fibers. These fibers have been made single and multi-mode for over 25 years. They have been used extensively in telecommunications, for example as praseodymium-doped fiber amplifiers and erbium-doped fluoride fiber amplifiers. The fluoride fibers also have relatively low loss and relatively high damage threshold (depending on the impurity concentration, typically between ˜10 to ˜20 MW/cm2). Furthermore, by adjusting the composition of the fluoride fibers, the low-loss transmission band can be extended out to between ˜4.5 microns and ˜5.5 microns.
Beyond ZBLAN fluoride fibers, other fibers or waveguides could also be candidates for mid-IR light generation. As one example, tellurite fibers (TeO2) can be used as mid-IR fibers. Tellurite glass compositions show enhanced Raman scattering behavior. By optimizing these oxide glass compositions with heavy-metal-oxides, fiber can be made that have high nonlinearity with transparency in the mid-IR wavelength range. Moreover, the tellurite fibers have been measured to have a damage threshold of ˜13GW/cm2. Other examples of mid-IR fibers include chalcogenide fibers (telluride, sulfide, selenide, as particular examples), sapphire fibers, AgBrCl fibers, etc.
As another example, silicon or other semiconductor waveguides could be used to generate mid-IR light. Silicon waveguides are expected to transmit light over the entire mid-IR wavelength band. Also, by making curves or S-type (i.e., waveguide going back and forth three times on a chip), relatively long lengths (i.e., several or tens of centimeters) of waveguide can be used. The use of silicon or other semiconductor waveguides is particularly effective if a pre-stage of fused silica fiber is first used to initiate the MI (discussed further below). In this case, the semiconductor waveguide serves primarily as the transparent, nonlinear element to lead to spectral broadening.
Although particular fiber types or waveguide structures have been described for advantageously generating super-continuum, other materials, compositions and guided wave structures can be used consistent with the disclosure.
The exemplary experiments use different lengths of fused silica fiber, which are a series of high-nonlinearity (HiNL) fibers made by Corning. The fibers have a zero dispersion wavelength between ˜1500 nm and ˜1950 nm. Some of the fibers had extra drying steps to remove to the extent possible the OH content, using steps that are commercially done for SMF-e Corning fiber. Lengths ranging from 1 to 13 m provide the broadest width of super-continuum in these particular experiments. Although these particular lengths were used in the experiments, other lengths can be used within the scope of the disclosure.
As an example of the SC spectrum from fused silica fiber,
To reduce the effects of bend-induced loss, the fiber can be laid out loosely. To reduce the effects of water absorption, the fibers can be dried using techniques used in commercial fibers, such as Corning's SMF-28e fiber or Lucent's (now OFS Fitel's) All-Wave fiber. To test this hypothesis, a new batch of HiNL fibers were made that were treated using the extra drying steps. As an example,
The edge of the SC spectrum could potentially be due to the vibrational absorption in the fused silica glass. If the edge of the spectrum is limited by the fiber loss, then it would be consistent that the spectrum might extend to longer wavelengths if the fiber length were to be reduced. However, sufficient fiber length is required to generate the full spectrum as well. In other words, there is a minimum length required to generate the spectrum, but then further propagation in the lossy fiber only reduces the long wavelength edge of the spectrum.
To understand this fiber loss limited spectrum further, a series of experiments were conducted at a pump peak power of ˜3 kW. First, an ˜2 m of standard single-mode fiber (SMF-28) is used to cause the nanosecond pulses to break up through MI. In fact, at these power levels SC can be already generated in the SMF fiber, with a reach out to ˜2500 nm. Then, the output of the SMF fiber is coupled to different lengths of HiNL fiber with a zero dispersion wavelength near 1544 nm. The data 1000 is illustrated in
The spectra of
The data of
The second step of the strategy or recipe can be to use a nonlinear element with at least partial transparency over the wavelength range of interest to broaden out the spectrum and to smoothen the spectrum into a SC. As one example, the dominant mechanism in the second stage can be self-phase modulation, where the nonlinearity for the high peak power pulses leads to spectral broadening. In addition, the Raman effect can also be effective in transferring energy from the short wavelength side to the long wavelength side, or more generally from shorter wavelengths to longer wavelengths. In the case of fused silica fiber, the second stage can be a relatively short length of HiNL fiber, as shown in
Although a two stage strategy or recipe is given as an example, more steps can be used to optimize the SC generation. For example, the first stage can be a set of fibers spliced or coupled together to achieve a particular dispersion profile. In one preferred embodiment, the fibers can be coupled to achieve a dispersion decreasing or dispersion increasing profile. Moreover, a number of stages of the nonlinear element can be used. In one preferred embodiment, the transparency region can be expanded in subsequent stages. In another embodiment, single mode as well as multimode fibers can be used in combination to obtain a high output power from the SC generation.
Although these experiments suggest that fused silica can generate SC out to ˜2.8 microns, the composition of fused silica can be altered to potentially achieve a wider wavelength range. As one particular example, fibers could be made from synthetic fused silica. For synthetic fused silica, there is a drop in transmission between ˜2.6 to ˜2.8 microns, which is probably due to the water absorption (OH absorption). The transmission through this wavelength range could be increased by using extra drying steps to minimize the OH content. Note, however, that there can be a transmission window between approximately 3 and 4 microns. Thus, with an appropriate fused silica composition, it may be possible to generate SC out to ˜3.6 to ˜4 microns. This is just one example of varying the composition, but other compositions of fused silica could also be advantageous for SC generation into the mid-IR.
For mid-IR generation, fibers that have lower loss than fused silica include chalcogenide fibers, tellurite fibers, and fluoride fibers. One of the more mature of the fluoride fibers is the heavy metal fluoride ZBLAN (ZrF4—BaF2—LaF3—AlF3—NaF). One advantage of the fluoride fiber is that the loss coefficient can be more than two orders-of-magnitude lower than chalcogenide fibers over the wavelength range between ˜2-5 μm. The Raman gain coefficient can be about ˜2-3× larger than in fused silica fiber. Moreover, the peak of the Raman gain falls at ˜600 cm−1, and fluoride fibers tend to be more mature technology with higher laser damage thresholds and no evidence of photo-darkening. For example, Alcatel and others made erbium-doped amplifiers and praesodynium doped amplifiers based on ZBLAN fiber in the 1980's and 1990's.
Three lengths of ZBLAN fluoride fiber were obtained for exemplary experiments of SC generation. The first fiber is 45 m long with a core diameter of ˜5.7 microns and a cladding diameter of 125 microns and a cut-off wavelength of ˜1.25 microns. The second fiber is a 85 m length of fiber with a core diameter of ˜6.5 microns and a cut-off wavelength at ˜1.75 microns. A third fiber is ˜20 m long with a core diameter of ˜7 microns, a cladding diameter of 125 microns, and a cut-off wavelength of ˜2.75 microns (the longer cut-off is achieved by using a higher numerical aperture of ˜0.3). For all these fibers the loss between 1.25 and 2.7 microns can be less than 10 dB/km (0.01 dB/m). There is a loss peak around 1 micron and another loss peak centered around 2.9 microns, and at these peaks the loss is between 30-50 dB/km. For the third fiber, the attenuation out to 4 microns is measured to be under 1 dB/m, and the attenuation beyond 4 microns is 1 dB/m at 4.3 microns, 2.25 dB/m at 4.5 microns, and 8 dB/m at 4.8 microns. Based on the experience from fused silica fiber, SC generation should advantageously have a long wavelength edge out to where the fiber has a loss of ˜1 dB/m to ˜2 dB/m. Therefore, the SC generation may be able to reach out to ˜4.5 to ˜4.6 microns in the ZBLAN fibers.
The loss spectra measured over a limited wavelength range is shown in
As the pump power is increased in the fluoride fiber, Raman wavelength shifting is first observed experimentally. Then, SC generation occurs after the Raman order crosses to the long wavelength side of the zero dispersion wavelength. For the 85 m length of fluoride fiber of
One main difference between the fused silica SC and the fluoride SC is the wavelength range expected. Whereas the glass transmission in fused silica would appear to limit the SC range to below 3 microns, because of the low loss in the fluoride fibers out to approximately 5 microns, the SC can continue to wavelengths longer than 3 microns. Moreover, since the dispersion slope is less in the fluoride fibers compared to fused silica (
The hypothesis is that the sharp wavelength edge observed in the exemplary ZBLAN fluoride fiber SC experiments arise from fiber bend induced losses. i.e., As the wavelength increases, the mode diameter increases and more of the mode penetrates into the cladding and is weakly guided. Several data points support the hypothesis of the spectral edge arising from the bend induced loss in the fiber. First, when the fluoride fiber was wound on a ˜8 inch spool, the edge of the spectrum reached to ˜3400 nm. When the same fiber was loosely laid in a drum, the spectral edge shifted toward longer wavelengths out to ˜3600 nm. Second, the bend induced loss is measured at 3.3 microns. For a bend diameter of 50, 100 and 200 mm, the percent loss at 3.3 microns is 85%, 3% and 1%, respectively. Therefore, the SC in the ZBLAN fluoride fiber should cover a wider range of the mid-IR when the bend induced is better controlled.
By using an appropriate length of fused silica pre-stage before the fluoride fiber, the length of fluoride fiber can be reduced and the spectral extent can be optimized. As an example, consider the second fluoride fiber (specifications in
The spectrum 1340 corresponding to using 5 m of the fluoride fiber from
The third fluoride fiber (characteristics in
The complete calibrated spectrum 1490 from ˜7 m of the third fluoride fiber following ˜1 m of SMF pre-stage is shown in
One significant feature of the SC can be a high spectral density over a wide wavelength range. For example, the spectrum 1490 in
Another feature of the SC 1490 is the remarkably smooth spectrum over a wide spectral range. Because of the relatively stable pump laser input to the SC fiber, it is believed that shot-to-shot the spectrum is the same. In fact, this is a valuable attribute for spectroscopy. However, during the pump pulse, the less than 2 nsec pulse probably has a range of intensities. The different values of the intensity may in turn be responsible for different parts of the spectrum. As a consequence of averaging over all the values of the intensities, the resulting spectrum may be quite smooth. This hypothesis also suggests a method of tailoring or adjusting the spectral shape of the SC. One way would be to use wavelength filters, such as gain equalizers or dynamic gain equalizers. However, another technique could be to modulate the time domain of the pump pulse, and then this temporal modulation could translate on to the spectrum as different parts of the pulse contribute to different parts of the spectrum.
There can be a number of techniques used to expand the long wavelength edge of the SC generation in optical fibers. In one embodiment, the composition of the fluoride glass can be changed to permit transmission out to longer wavelengths. The fibers described thus far are zirconium fluoride glass, with a exemplary composition for the ZBLAN of (mole %): ZrF4 (57), BaF2 (34), LaF3 (5), AlF3 (4). For the ZBLAN or more generally the zirconium fluoride fibers, the transmission edge of 1 dB/m at 4.3 microns is fairly common, and it the IR edge does not shift very easily. On the other hand, fluoride glass fiber that does not contain zirconium fluoride fiber or other short-IR-edged compounds may enable transmission to longer wavelengths. By changing the composition, the long wavelength edge can be found to extend beyond ˜5.4 microns. Therefore, if the SC generation were implemented in such a fiber, the edge of the SC might be expected to reach beyond 5 microns. For long wavelength performance, the cut-off wavelength for the fiber should probably be beyond 2 microns, preferably beyond 2.5 or 3 microns, to control the bend induced loss at the longer wavelengths. The core size can also be advantageously relatively small (e.g., less than a core diameter of 12 microns, more preferably less than 10 microns) to reduce the power requirements for the SC generation. However, larger core sizes may also be used to increase the overall output power from the SC spectrum.
Other embodiments of fluoride fiber can also be used to extend the long wavelength edge or to optimize the shape of the SC spectrum. In one embodiment, the pump wavelength could be made closer to the zero dispersion wavelength of the fiber, or a cascaded Raman shifted order of the pump could fall closer to the zero dispersion wavelength of the fiber. In a preferred embodiment, the pump or the shifted pump wavelength would fall slightly to the long wavelength side of the zero dispersion wavelength. This would lead to MI with a broad gain spectral width. In another embodiment, a hybrid configuration of different fluoride fibers could be used to effectively taper the core size of the chain, either downward or upward. In yet another embodiment the wavelength dependence of the core and cladding material can be selected so that the numerical aperture (NA) increases with increasing wavelength. For a step-index fiber, the NA=sqrt (n12−n22), where n1 is the index of the core and n2 is the index of the cladding. Therefore, if the difference between the two indices of refraction increases with increasing wavelength, then the NA will increase. As the NA is increased, the waveguide will be better guiding and the effect of bend induced loss will be lowered.
As an alternative, fibers made from different materials can also be used to increase the wavelength extent of the SC. Another option for mid-IR fibers are tellurite (TeO2) glass fibers. Recently, there has been growing interest in the TeO2-based glasses because of their strong nonlinear properties and capacity for doping with high concentrations of rare-earth elements. Hence, these glasses can be appropriate for a wide range of devices including lasers, amplifiers, and mid-IR wavelength converters. Several preliminary studies have been reported in the literature regarding the glass properties. For example, depending on the doping details, the Raman gain coefficient can range from 30 times larger than fused silica to 45 to 95 times larger than fused silica. In addition, the Raman gain band in the TeO2 glasses can be up to a factor of two wider in bandwidth than fused silica. Moreover, the damage threshold for the TeO2 glasses is measured to be approximately 15-20 GW/cm2, which is about a factor of two or three smaller than fused silica at 50 GW/cm2. For the tellurite fibers the nonlinearity can be strongly dependent on the material composition, and the zero dispersion wavelength can also vary with material composition. In addition, the tellurite fibers may transmit light at least out to 4 microns, and even out to 5 microns in bulk glass. According to some reports, at 4 microns the theoretical background loss (i.e., material loss) can be somewhere above 10 dB/m. The minimum loss in tellurite fiber would be around 3 microns, and the value of the loss should be between 5-10 dB/m in the fiber at 3 microns. In yet other embodiments, materials made in waveguides may be advantageous for mid-IR light generation. For example, if the pulse break-up first occurs in fused silica fiber, then the nonlinear spectral broadening for SC generation can occur in silicon or other semiconductor waveguides.
Given that only certain range of fiber parameters are available in the fluoride fibers and that step-index fiber can only provide limited control over the dispersion profile, an additional degree of freedom for the mid-IR fibers may be helpful. The use of two pump wavelengths may provide this optimization option. With the two pump case, MI can occur with either pump in the anomalous or normal dispersion regime. Thus, whereas for the single pump case MI phase matches when the pump is in the anomalous dispersion regime, the addition of a second pump relaxes this constraint. As an example,
Another aspect of the MIRFIL is that the average power can be increased to >500 mW from the current ˜20 mW average power. For the higher powers, one change could be to use a higher power pump laser. Examples of higher power pump lasers include solid-state lasers, diode-pumped laser systems including solid state lasers, cladding pumped fiber amplifiers and lasers, and optical parametric oscillators or amplifiers. To improve the efficiency and power, longer wavelength (˜2 microns) and higher power solid state lasers or cladding pumped fiber amplifiers or lasers can also be used. For instance, holmium or thulium lasers provide light near 2 microns in wavelength. As the powers are increased, another change can be to use larger core size fibers, so that the intensities can remain below the damage threshold while the overall output power can be increased. For example, different core sizes of fluoride fibers are already commercially available. In addition, the HiNL fused silica fibers could possibly be pulled to larger sizes, although care will be needed not to change the zero dispersion wavelength in these fibers.
Although a number of embodiments of using fluoride fibers to generate SC into the mid-IR are described, other configurations and fiber types can also be used to alter the shape of the SC spectrum or to extend the wavelength range of the SC generation.
The results in the fused silica and fluoride fibers suggest a more general model of optimizing SC generation or wavelength conversion (further discussed in a few sections below). One example of the generalized model is illustrated in
The first stage may be used to cause break-up 1620 of the CW or quasi-CW light into pulses or solitons through the MI effect. The first stage 1620 can advantageously be implemented in optical fibers, and for a single pump wavelength the MI phase matches in the anomalous group velocity dispersion regime. If the pulses are nanosecond (i.e., longer than approximately 100 psec, or even longer than about 30 psec) or quasi-CW light, there may be enough intensity modulation to cause collisions between different soliton pulses. Otherwise, in a preferred embodiment an intensity modulator can be used to create a distribution of intensities, which in turn can lead to a collision between soliton pulses. The intensity modulation may also help to create a smooth spectrum, due to the distribution of pump intensities.
In some cases, it may be further advantageous to have a mid-stage 1660 after the MI-initiated pulse break-up stage. This mid-stage, for example, can have a pulse sharpener 1660, which helps to compress the soliton pulses and/or create more modulation sidebands in the frequency domain. Examples of the mid-stage include optical fibers, dispersion decreasing fibers, tapered fibers, grating compressors, or other examples of pulse compressors, whether they are implemented in optical fibers or bulk optics. This mid-stage can additionally help by increasing the peak intensity of the pulses. As such, the mid-stage can also include optical amplifiers.
The second stage can then include a nonlinear element for SC generation 1630 or wavelength conversion 1670. The non-linear element can help to generate SC or new wavelengths based on four-wave mixing processes. For SC generation, the nonlinearity in the second stage can give rise to spectral broadening through self-phase or cross-phase modulation. Although the nonlinear properties of this second stage is one of the important parameters, it may also be desirable to have some dispersion to cause pulse walk-off or pulse motion. Such pulse motion may help to smoothen the spectrum or create even higher peak intensities. It may also be advantageous for the second stage to be at least partially transparent over the wavelength of interest. For example, for mid-IR conversion, it may be advantageous for the second stage to be transparent over much of the mid-IR wavelength range. Examples of the second stage include different optical fibers, including HiNL, ZBLAN, fluoride, tellurite, chalcogenide, or even semiconductor doped glasses or waveguides.
Although most of the experiments presented have used mid-IR fibers or fused silica fibers, in the more generalized model other elements such as semiconductor waveguides or nonlinear crystal material could be used in the second stage. As one particular example, a silicon waveguide could be used as the second stage. The nonlinearity in silicon is about four-orders-of-magnitude higher than in silica fiber. The band gap in silicon is around 1.1 microns, so silicon is transparent (at least in a linear sense) for wavelengths longer than 1.1 microns and throughout the mid-IR wavelength range. Therefore, it is advantageous to have a pump wavelength below the band gap of silicon. However, for a pump wavelength between approximately 1.1 and 2.2 microns, the pump will experience two-photon absorption (TPA). In turn, the carriers generated through TPA can induce free-carrier absorption.
One method to overcome the TPA-induced free carrier absorption is to embed the silicon waveguide in a P-I-N diode configuration, particularly with the PIN diode reverse biased. As an illustration, the waveguide may fall in the I (intrinsic) region, and the electric field from the reverse biased diode can help to quickly sweep out the electrons and holes created by the TPA effect. Although this technique reduces the free carrier absorption, it does not prevent the TPA effect. Furthermore, the silicon waveguide in a PIN diode can be enhanced in a number of ways For instance, the length of the waveguide can be extended by using multiple zigzags, such as in a S-configuration. Moreover, the pump light can be multiple passed by placing coatings on the semiconductor wafer or mirrors around the wafer. In a preferred embodiment, one side of the chip may be coated for high reflectivity, while the other side can be anti-reflection coated or dichroic coated. Another advantageous configuration can modulate the applied voltage to the PIN diode to control the loss in the waveguide. As an example, this modulation could control the long wavelength edge of the SC spectrum or could be used to put codes onto the SC spectrum.
The silicon PIN waveguide is just one example of the nonlinear element 1630 or 1670 that could be used for SC generation. There are many other semiconductor or other materials that could alternatively be used. For instance, a waveguide can be made in a wide-gap semiconductor, where the band gap is at shorter wavelength than the TPA edge. This would avoid the TPA problem, thereby removing the necessity of using a PIN for carrier sweep-out. Alternately, a more atomic-like material can be used, such as quantum dots, so the material does not have a conduction band and the associated TPA problems. Moreover, other nonlinear crystals could be used, such as lithium niobate or periodically-poled lithium niobate. Furthermore, different fiber configurations could be used. For example, a hollow core fiber or capillary could be used that is filled with a nonlinear liquid, such as C52. Other fiber types could also be used, such as tellurites, chalcogenides, or photonic crystal fibers.
Chalcogenide fibers represent another alternative of fiber types for mid-IR light generation. Examples of chalcogenide fibers include sulfide (typically transmitting out to approximately 6 microns), selenide (typically transmitting out to approximately 9 microns) and telluride (typically transmitting out to 11 microns). Technical feasibility has been demonstrated for cascaded Raman wavelength shifting in chalcogenide fibers. In a particular embodiment, samples of arsenic-tri-sulfide fibers were obtained. The testing started with a 20 m length of fiber number WS-884, which has a slight selenide doping, a core size of approximately 6.5 microns, and a numerical aperture of ˜0.22. For example,
To generate and measure the spectrum beyond the second Raman cascade order 1770, the light from the mid-IR fiber can preferably be sent to an optical spectrometer that is optimized for the near to mid-IR. In particular, a 0.3 m spectrometer is used that has a grating with 300 grooves/mm. The numerical aperture for the fiber output is optimized to couple into the spectrometer using lenses that are transmitting in the mid-IR, such as calcium fluoride lenses. The detector used is a modified InGaAs detector, which has high sensitivity out to 2.6 microns. To minimize the effect of the water absorption line around 1.9 microns, a dry nitrogen as is used to purge the interior of the spectrometer.
The results from the chalcogenide fibers could be improved using a number of techniques. Different fiber sizes will be tested to see if the fiber core is more uniform or continuous in the larger core size fibers. Gallium on the two ends of fiber can be used to test for the guiding properties of the lowest order mode in different fiber lengths. The ends of the sulfide fiber may also be encapsulated to remove heat and, thereby, to increase the damage threshold.
As another alternative, selenide fibers could be used, which are interesting because they should have an order of magnitude larger Raman gain coefficient compared to the fibers tested. One question is the value of the damage threshold power for the selenide fibers. If the damage threshold is the same in the selenide fibers as the sulfide fibers that have been tested, then a significant improvement in CRWS might be expected. However, the index-of-refraction variation with temperature ∂n/∂T can be positive in the chalcogenide fibers, and the value can be an order of magnitude larger in the selenide fibers compared to sulfide fibers. Therefore, one concern may be that catastrophic self-focusing might occur in the selenide fibers due to thermal effects from light absorption. In addition, the selenide fibers have a band gap of ˜750 nm, which is closer to the pump wavelength than the sulfide fibers (band gap around ˜520 nm). Thus, a second concern arises from photo-darkening effects arising from two-photon absorption. To overcome photo-darkening concerns, it might be worth trying a hybrid approach, where light is first shifted in fused silica out to ˜2-2.8 microns, and then the light is coupled into the chalcogenide fibers for further shifting. An alternative approach will be to pump the chalcogenide fibers with thulium lasers (either fiber based or solid state lasers), so the shifting starts from around 2 microns. Although particular schemes are described for CRWS in chalcogenide fibers, a myriad of other techniques and materials can be used for generating light using CRWS into the mid-infrared.
There are applications, such as spectral fingerprinting, where SC generation can be very valuable. Also, SC could benefit infrared counter-measures (IRCM), because it becomes virtually undefeatable because the broad spectrum mimics the black body radiation from hot metal objects. However, there are many cases where only a narrow band of frequencies in the mid-IR may be desired. For example, laser ablation typically only uses a band of frequencies, and IRCM traditionally uses three frequency windows in the mid-IR. For these cases where only a few mid-IR wavelengths are required, SC can be inefficient, since the energy may be spread over a wide spectral range. Wavelength conversion of the pump wavelength to a set of frequencies in the mid-IR would be significantly more efficient.
Because of the similarity of the experimental set-up and the same underlying physics at work, one question is when does SC generation occur and when does wavelength conversion occur. The MI process can be used to convert the CW or quasi-CW (e.g., nanosecond pulses) to short pulses required for many of the nonlinear phenomena, thereby reducing or eliminating the need for mode locked lasers. Also, for the single pump wavelength case MI phase matches in the anomalous group-velocity dispersion regime. Therefore, the first step for either SC or wavelength conversion can be to propagate the light in a length of anomalous dispersion fiber (i.e., soliton regime of the fiber). The main difference in outcome may depend on how long the pulses are permitted to propagate in the soliton regime of the fiber.
To distinguish SC generation from the wavelength conversion processes, it is worth first examining the onset of the SC generation process. As an example,
In order to observe wavelength conversion through four-wave-mixing (4WM), the MI break-up of the pulses as seen in
Depending on the wavelength conversion mechanism, different strategies can be used to avoid the collision and super-pulse creation of
Four-wave mixing is a four-photon process where two pump photons combine to produce a Stokes wavelength (longer wavelength) and an anti-Stokes wavelength. One aspect of 4WM is that phase matching is required between the four waves. For instance, the wave vector mismatch is given by
and the conversion efficiency is given by
Normally, high efficiency for the 4WM process can be obtained near the zero dispersion wavelength. However, a new regime of phase matching can be advantageously used that enables mid-IR light generation, since this new regime is distant from the zero dispersion wavelength. As one example, the 4WM wave vector mismatch 1800 is calculated and plotted in
As the zero dispersion wavelength and the dispersion profile of the fiber is changed, the position for this mid-IR light wavelength conversion can change. For instance, the following table shows different examples of the calculated and measured 4WM peak for different fibers measured.
Experimental confirmation can also be seen of this new regime of phase matching for 4WM in different fused silica and fluoride fibers. As a particular example, the experimental set-up for testing 4WM wavelength conversion is illustrated in
The data above suggests a procedure for wavelength conversion of light into the mid-IR wavelength range, particularly when there is a target wavelength desired for a particular application. First, the dispersion of the fiber can be tailed to phase match at a target wavelength. The fiber dispersion can be tailored by changing the zero dispersion wavelength, adjusting the dispersion slope, or perhaps by using more exotic fibers such as micro-structure fibers that can have more than one zero dispersion wavelength. Then, if the pump is for example within the telecommunications band, then tune the wavelength of the pump laser to obtain the correct target wavelength. The pump laser could be a tunable laser, or the pump laser could be laser diodes of different wavelengths, for example laser diodes that are on the ITU wavelength grid. With the appropriate adjustment of the phase matching condition, then introduce a seed laser at the anti-Stokes wavelength. Since the anti-Stokes wavelength falls in the near-IR wavelength range, one example of a seed laser would be laser diodes. With the introduction of the anti-Stokes wavelength, mid-IR light on the Stokes side should result, so long as the fiber can transmit light at the particular mid-IR wavelength. Thus, as an example light out to ˜2.7 microns might be generated in fused silica fiber, light out to ˜4.4 micron might be generated in ZBLAN fluoride fiber, and light out to ˜5.5 microns might be generated in the extended band fluoride fiber. In a preferred embodiment, a fused silica fiber pre-stage can be used to generate pulses through MI, and then the wavelength conversion would be in fiber where the pump wavelength falls in the normal dispersion regime. Although one particular method of wavelength converting light into the mid-infrared regime is suggested, numerous other techniques can be used within the scope of the present disclosure.
Several differentiators for the MIRFIL fiber-based sources include: Maturity of underlying technology; for SC, emulate black body radiation or attractive source for spectral fingerprinting or last mile solutions in telecommunications; for wavelength conversion, simple tuning over wide wavelength range Excellent beam quality (M2<1.4, as an example); advantages of fibers, such as compact, robust, lightweight, and no moving parts; potential room temperature operation with flexible repetition rate from CW to MHz or higher; and power scalable to ˜10 W or more by using larger core size fibers and higher pump powers.
On this last point, the scalability of the power by pumping with a cladding pumped fiber laser can be quite attractive. As an example, in the last several years the CW power from cladding pumped fiber lasers has increased from 10's of Watts to a time-average power of 15 kW in 2005. Moreover, pumping with a cladding pumped fiber laser could enable an all-fiber integrated MIRFIL. The SC generation or wavelength conversion fibers (whether one, two or more stages) could be coupled to this pump unit using either fusion splicing, mechanical splicing, or free space or bulk optical coupling. Then, the resulting unit could be an all-fiber, high power MIRFIL. As mentioned before, cladding pumped fiber lasers can operate at exemplary wavelengths near 1 microns, 1.55 microns or 2 microns, depending on the dopants in the fiber. Although a particular monolithically integrated MIRFIL is illustrated, many other configurations and pumping techniques can be used within the scope of the present disclosure.
The MIRFIL may be used for applications where light in the mid-IR wavelength range (exemplary 2 to 5 microns) is advantageous. For example, the mid-IR is known as the spectral fingerprint region, because many chemicals have their rotational and vibrational resonances at least in part in the mid-IR wavelength range. Also, the mid-IR can be important for heat sensing, since black body radiation from “hot objects,” such as plumes or hot metal, falls at least in part in the mid-IR. Moreover, for applications in the life sciences, laser ablation near 3.6 or 6.45 microns could be advantageous, since the protein and amide group absorption can exceed the water absorption. Also, mid-IR light near the peak of the water absorption could lead to high-resolution photo-acoustic imaging, which can be important for applications such as laser keratectomy. These are exemplary applications of mid-infrared light sources, but many other applications fall within the scope of the present disclosure.
The early adaptors of the MIRFIL laser technology may be in military related markets for chemical sensing and infrared counter-measures (IRCM). However, there are also commercial markets for the same kind of MIRFIL laser units. For example, a similar laser that is used for chemical sensing can be used in the commercial sector for industrial chemical plant control, advanced semiconductor processing, combustion monitoring and bio-medical diagnostics. Similarly, a similar laser that is used for IRCM can be used in the commercial sector for bio-medical laser ablation.
The first application to use the MIRFIL may be chemical sensing systems products. In particular, the wavelengths of IR absorption bands are characteristic of specific types of chemical bonds and every molecule has a unique IR spectrum (fingerprint). IR spectroscopy finds its greatest utility for identification of organic and organo-metallic molecules. There are three IR spectroscopy technologies employed in point detectors: Fourier transform IR (FTIR) spectroscopy, photo-acoustic infrared spectroscopy, and filter based IR spectroscopy.
The SC broadband source could be particularly useful for spectral fingerprinting. In several chemical sensing detection systems, a narrow line width, tunable laser may be used to perform spectral fingerprinting. Instead of this approach, the SC based spectral fingerprinting can be much more like white light spectroscopy. In other words, the SC may permit simultaneous monitoring over a wide spectral range. In one embodiment, the spectra at several wavelengths can be used to advantageously identify a chemical species. In another embodiment for absorption or reflection spectroscopy, several wavelengths of the absorption or reflection can be measured either simultaneously or in some time sequential fashion. Then, the relative magnitudes at different wavelengths or a particular spectral pattern of absorption or reflection can be pattern matched to identify the chemical of interest. Such a technique has the potential of having high selectivity, since the monitoring can be accomplished over a wider spectral range and since the spectral pattern matching can compare a number of features.
An exemplary system 2000 for performing spectral fingerprinting or using the light source is illustrated in
Systems such as 2000 can be used for chemical sensing for military applications as well as industrial plant monitoring systems. For example, chemical sensing can be used to detect chemical warfare agents, which are chemical substances that are intended for use in warfare or terrorist activities to harm people through their physiological effects. The most common chemical agents include nerve agents, blister agents and arsenical vesicants. Moreover, chemical sensing can be used for weapons detection, since residue from gun power can be sensed using remote or non-contact optical spectroscopy techniques. In addition, toxic industrial materials are chemicals other than chemical warfare agents that have harmful effects on humans. These are used in a variety of settings such as manufacturing facilities, maintenance areas, and general storage areas.
In one embodiment, the spectral fingerprinting system can be used for firearms detection. For example, firearms detection can be implemented by searching for the composition of gun powder. One chief ingredient in smokeless gun powder is nitro-cellulose, which has clear spectral features centered around 2.86 microns and 3.45 microns. Although there are also lines at 6 microns on beyond, many chemicals have a lot of lines in that wavelength range, so it may be difficult to separate one chemical from another. Beyond nitro-cellulose, there are also a number of additives in smokeless gun powder, an example of which is provided in the table below.
Diphenyl amine, which is used extensively as a stabilizer, shows clear spectral signatures centered around 2.94, 3.33 and 3.85 microns. Dibutyl phthalate, which is used as a plasticizer, shows an absorption peak around 3.4 to 3.55 microns. Lead azide, which is used as a primer, has a peak absorption around 4.8 microns. Other examples of primers include tetracene (broad absorption between approximately 2.8 and 4 microns), barium nitrate (absorption peaks near 2.94 and 4.2 microns), and strontium nitrate (absorption peaks around 2.94 and 4.15 microns). Thus, many of the components of smokeless gun powder have signatures in the mid-IR between 2 to 5 microns.
In another embodiment, the spectral fingerprinting system can be used for IED (improvised explosives detection) or weapons detection. Many of the explosives have modified benzene rings, and the benzene rings have a resonance around 3.2 microns. Although there are a lot of absorption lines from 6-10 microns and in the terahertz region, it may be difficult to sort out one chemical from another (i.e., there may be too much interference, leading to poor selectivity). Cleaner, more discrete, signatures are seen in the mid-IR, so although the level of absorption may not be as great, the selectivity may be better. Examples of explosives and their approximate mid-IR lines include the following:
There are other applications in chemical sensing for the spectral fingerprinting system as well. For example, the system can be used for drug detection or chemical weapons agent detection. As an illustration, drugs such as cocaine, methamphetamine, MDMA (ecstasy) and heroin have distinct optical spectral signatures in the wavelength range from 2-5 microns. In one embodiment, the use of a broadband source covering a large fraction of the mid-IR between 2 to 5 microns can be used to advantageously detect various drugs. Moreover, many of the chemical weapons agents, such as sarin, cyclosarin, soman, tabun, sulfur mustard, nitrogen mustard, VX and lewisite, have absorption features in the 3 to 4 micron window, particularly centered around 3.3 microns. Thus, non-contact, remote detection of drugs, weapons, firearms, and chemical agents could advantageously be implemented with a spectral fingerprinting system utilizing the SC source.
Beyond chemical sensing, another application for the high power version of the MIRFIL in military and homeland security might be in IRCM, particularly for the commercial air fleet. For instance, much of the black body radiation falls in the wavelength range covered by the SC sources described between ˜1 microns and ˜4.5 microns. In one embodiment, the SC spectrum could be carved or shaped using spectral filters to resemble the spectrum for hot metal or plume.
Other chemical sensing applications for the SC source or wavelength conversion source include semiconductor process control, combustion monitoring and industrial chemical. For example, the chemicals in the semiconductor growth chamber can be monitored to provide a real time feedback signal to an advanced process control engine. By using the SC, a number of chemical species can even be monitored simultaneously. Examples of chemicals that are relevant for semiconductor processing include monitoring HCl and HBr for plasma etching or monitoring CxFy for gate etching. Alternatively, the chemicals in combustion chamber can be monitored using spectral fingerprinting. Most applications relevant to gas dynamic and combustion flows are based on absorption by low-molecular weight molecules with well resolved transitions—such as O2, H2O, CO, CO2, NO, NO2, OH, NH3, HF, H2S, CH4, as particular examples. Because of current limitations arising from a lack of convenient mid-infrared sources, the absorption measurements today for chemical sensing may be performed on overtone and combinational vibrational absorption bands, which typically fall in the near IR where laser diodes are available. However, typical line strengths of these transitions are two or three orders-of-magnitude below the fundamental vibrational transitions in the mid-IR. Therefore, by using SC in the mid-IR wavelength range, a much stronger signal can potentially be obtained by operating at the fundamental wavelength of the transitions.
Another application for the MIRFIL based on SC generation or wavelength conversion is in bio-medical ablation or imaging. As an example, the protein absorption dominates over water absorption between ˜3.6 microns and ˜4 microns and again near 6.1 microns and 6.45 microns. By using laser ablation in one of these windows, the protein can be denatured (for example, by relying on the amino acid absorption) before boiling the water, thereby resulting in less collateral damage. One example of the value of avoiding the collateral damage could be in cosmetic surgery. For instance, cosmetic surgery is often used to remove wrinkles or unwanted skin or tissue, but discoloration or scars from heating might be undesirable. By denaturing the protein with minimum collateral damage, the unwanted skin or tissue or wrinkles could be removed without scaring or skin discoloration. To achieve the wavelength range of interest, SC generation or wavelength conversion could be used based on 4WM. This is one example of a mid-infrared light source for biomedical applications, but many other configurations can be used within the scope of the present disclosure.
The above example uses mid-IR light in applications at wavelengths where the protein absorption exceeds the water absorption. However, there are several instances where the optimal use of the mid-IR light can be at wavelengths where the water absorption dominates. In a particular embodiment, the mid-IR light can be used at a wavelength of strong water absorption, such as close to 2.9-3.1 microns, so that a short ultrasonic or acoustic wave can be launched for high-resolution ultrasound imaging. The wavelength of strong water can be selected to minimize the absorption length of the mid-IR light in the water. In a preferred embodiment, the pulse width of the mid-IR light is under 100 nsec, under 10 nsec or under 2 nsec. These wavelengths and pulse widths are exemplary, but many other ranges of values can be used.
For the short pulses and absorption lengths, the resulting wave then acts as an acoustic impulse. As an example, one particular embodiment where the acoustic impulses can be beneficial is in precise cornea thickness measurements (pachymetry) made during planning for laser keratectomy. Precise thickness measurements can be obtained with high-frequency ultrasound. The use of optical pulses at wavelengths of high water absorption to create the acoustic pulses lends itself to a non-contact procedure for ultrasonic measurements. On benefit of using an all-optical method to generate the acoustic wave can be that it enables simple integration with laser ablation systems. Thus, measurements and laser ablation can be done in one procedure sequentially without need for moving instruments or patients. More generally, laser-induced ultrasonics operating near the water absorption lines can be used to map out many different materials and biological systems. For these types of application, it could be more advantageous to wavelength conversion based on 4WM, so only a narrow band of wavelengths near the water absorption are generated, rather than the entire SC spectrum.
Another potential application for the SC generation can be to use the MIRFIL in Optical Coherence Tomography (OCT) systems used in bio-medical imaging and diagnostics. Over the past two decades, OCT has been established as a diagnostic technique for minimally invasive, high-resolution, cross-sectional imaging in a variety of medical fields. The OCT system comprises a broadband, low-coherence light source, a fiber-based Michelson interferometer, a sample scanning and positioning stage, and a detector followed by electronics. OCT is analogous to conventional ultrasonic pulse-echo imaging, except that it does not require direct contact with the tissue that is being investigated and it measures echo delay and the intensity of the back-reflected infrared light rather than acoustic waves from internal tissue structures. The light source used in OCT helps to determine the instrument properties in terms of the spectral bandwidth (axial resolution), center wavelength (penetration depth), power density (data acquisition time), cost and size. The SC light source that provides broad bandwidth without using a mode locked laser could lead to micron level resolution for OCT systems without using an expensive light source.
OCT is usually used for biological systems. However, the SC light source could also be advantageously be used with OCT for sub-surface defect detection in semiconductors, ceramics, or other solid state materials. As an illustration, OCT could be used to inspect silicon wafers before they are processed. This could permit sorting of the wafers (i.e., charge a premium for better wafers) and avoiding the cost of processing poor quality wafers. Alternately, OCT could be used to inspect multi-layered structures. By using SC light beyond 1.1 microns, which falls below the bandgap of silicon, the light can penetrate into the chip or wafer. Also, by using longer wavelengths, the scattering loss is reduced. Furthermore, because of the broadband spectrum, the depth resolution of OCT can be at the sub-micron level. Thus, sub-micron to several micron sized defects could be inspected using the SC-light source based OCT.
Typical OCT systems operate point-by-point, which may be too slow for some of the wafer or chip inspection applications. The speed is limited both because imaging is done point-by-point, as well as because one arm of the OCT is moved to achieve the depth resolution. As an alternative, methods used in spectral domain OCT can be used to avoid moving one arm of the interferometer, and by using a line scan the point-by-point scanning could be avoided. As an example, the light from the output of the SC source could be stretched onto a line using a cylindrical lens or an appropriate optical lensing system. The resulting line of light could be split using a beam splitter to a reference arm with a reference mirror or sample and the sample arm. The sample can be located in the sample arm, and the sample can be moved below the light to scan line by line. The return beams from the reference arm and sample arm can be recombined at the beam splitter, and an imaging lens can then be used to image it into a spectrometer. In one particular embodiment, the spectrometer could be dispersive optics, such as a grating or a lens, which could take each point of light and spread it into a spectrum to be detected by a detector array. By processing the multi-spectral data from each spatial point, the location of the reflection from the sample can be detected. Thus, instead of using a movement of the reference arm, the Fourier transform of the interference data may be processed to obtain the height of the reflection.
Yet another application of the SC source is in the so-called “last mile solution” in telecommunications. The last mile solution includes the technologies related to fiber-to-the-X (FTTx), where X can exemplary be home, node, neighborhood, curb, or premise. As one example, the SC source can be an enabling technology for wavelength division multiplexed passive optical networks (lambda-PONS). In a lambda-PONS based FTTx system, each location can receive one or more wavelengths. A challenge for lambda-PONS is the multi-wavelength light source, which may reside at the central office or other telecommunications location. The SC source can advantageously provide a potentially low-cost solution for the multi-wavelength light source.
In one particular embodiment, the SC source can be coupled to one or more modulators and a wavelength division multiplexer to implement the FTTx multi-wavelength light source. As an example, the SC source could advantageously emit wavelengths covering the low loss window in optical fibers, advantageously between 1250 nm and 1750 nm. For this example, the entire SC source could be implemented in fused silica fiber. Then, the output from the SC source can be separated into multiple wavelength channels, using, for example, a wavelength division multiplexer. Each wavelength can then be modulated using a modulator. The modulated wavelength signals can then be combined and coupled to the output fiber for propagation over the FTTx system. In addition, in the FTTx system the power splitters may be replaced with wavelength division multiplexers. Examples of wavelength division multiplexers include arrayed waveguide gratings, waveguide grating routers, dielectric coated beam splitters, and bulk optical gratings.
In another embodiment, the FTTx multi-wavelength light source could comprise a SC source coupled to a dispersive pulse stretcher, one or more high-speed modulators, and a wavelength division multiplexer to separate the wavelengths. Advantageously, the SC source could advantageously emit wavelengths covering the low loss window in optical fibers, particularly between 1250 nm and 1750 nm. The dispersive pulse stretcher can then broaden the SC pulse, spreading the wavelengths so the channels occupy different time slots. The one or more high speed modulators can be used to time sequentially encode the different channels, and then the wavelength division multiplexer is used to separate the wavelength channels in the FTTx system.
Although the present invention has been described in several embodiments, a myriad of changes, variations, alterations, transformations, and modifications may be suggested to one skilled in the art, and it is intended that the present invention encompass such changes, variations, alterations, transformations, and modifications as falling within the spirit and scope of the appended claims.
This application is a continuation of U.S. application Ser. No. 15/662,527 filed Jul. 28, 2017, which is a continuation of U.S. application Ser. No. 15/248,230 filed Aug. 26, 2016, now U.S. Pat. No. 9,726,539 issued Aug. 8, 2017, which is a continuation of U.S. application Ser. No. 14/861,755 filed Sep. 22, 2015, now U.S. Pat. No. 9,476,769 issued Oct. 25, 2016, which is a continuation of U.S. application Ser. No. 14/715,960 filed May 19, 2015, now U.S. Pat. No. 9,400,215 issued Jul. 26, 2016, which is a continuation of U.S. Ser. No. 14/186,171 filed Feb. 21, 2014, now U.S. Pat. No. 9,077,146 issued Jul. 7, 2015, which is a continuation of U.S. application Ser. No. 14/071,983 filed Nov. 5, 2013, now U.S. Pat. No. 8,971,681 issued Mar. 3, 2015, which is a continuation of U.S. application Ser. No. 13/750,556 filed Jan. 25, 2013, now U.S. Pat. No. 8,670,642 issued Mar. 11, 2014, which is a continuation of U.S. application Ser. No. 13/241,900 filed Sep. 23, 2011, now U.S. Pat. No. 8,391,660 issued Mar. 5, 2013, which is a Continuation of U.S. application Ser. No. 12/366,323 filed Feb. 5, 2009, now U.S. Pat. No. 8,055,108 issued Nov. 8, 2011, which is a Continuation of U.S. application Ser. No. 11/599,950 filed Nov. 15, 2006, now U.S. Pat. No. 7,519,253, issued Apr. 14, 2009, which claims priority to U.S. Provisional Patent Application Ser. No. 60/738,389, filed Nov. 18, 2005, the disclosures of which are hereby incorporated by reference in their entirety.
This invention was made with Government support under Contract No. W31P4Q-05-C-0159 awarded by the U.S. Army Aviation and Missile Command and Contract No. W911NF-04-C-0078 awarded by the U.S. Army Robert Morris ACQ CTR. The Government has certain rights in the invention.
| Number | Date | Country | |
|---|---|---|---|
| 60738389 | Nov 2005 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15662527 | Jul 2017 | US |
| Child | 16028473 | US | |
| Parent | 15248230 | Aug 2016 | US |
| Child | 15662527 | US | |
| Parent | 14861755 | Sep 2015 | US |
| Child | 15248230 | US | |
| Parent | 14715960 | May 2015 | US |
| Child | 14861755 | US | |
| Parent | 14186171 | Feb 2014 | US |
| Child | 14715960 | US | |
| Parent | 14071983 | Nov 2013 | US |
| Child | 14186171 | US | |
| Parent | 13750556 | Jan 2013 | US |
| Child | 14071983 | US | |
| Parent | 13241900 | Sep 2011 | US |
| Child | 13750556 | US | |
| Parent | 12366323 | Feb 2009 | US |
| Child | 13241900 | US | |
| Parent | 11599950 | Nov 2006 | US |
| Child | 12366323 | US |
