JP7275069B2 - Optical measurement system and method - Google Patents

Description

The present invention relates to a supercontinuum light source comprising an intermediate supercontinuum (SC) light source and a single-mode coupling unit, the supercontinuum light source being measured or otherwise analyzed by a measurement system, e.g. Suitable for use in systems in which the sample is illuminated by light originating from such a supercontinuum light source, the measurement system is configured to allow detection of light from the sample. The invention also relates to a system suitable for measuring at least one parameter related to an object, said system comprising a method for measuring at least one parameter related to the object of the measurement system, similar to the supercontinuum light source.

Optical measurement systems exist in many variations. Common to these systems is that a beam of light is directed to the sample and light from the sample is captured. The captured light may be light emitted from the sample in response to the incident beam, such as light reflected from the sample, light transmitted through the sample, and/or fluorescent light.

Octave-bandwidth supercontinuum (SC) has been successfully performed directly through nonlinear fibers such as microstructured, tapered standard, and tapered microstructured fibers by pumping the fiber with a pulsed laser (often in MOPA configuration) as input. generated. Such spectrally broad continuum light sources are used in many measurement systems, e.g., optical coherence tomography (OCT), optical frequency metrology, fluorescence microscopy, coherent anti-Stokes Raman scattering (CARS) potentially useful in anti-Stokes Raman scattering) microscopy, and two-photon fluorescence microscopy. Unfortunately, for those experiments, the large amplitude variations of conventional continuum light sources limit accuracy and/or sensitivity. Previous studies of SC generation have shown that the SC generation process is highly sensitive to quantum noise, technical noise, and certain parameters such as input wavelength, duration and chirp of the input laser pulse. Light sources derived from stable continuum would generally improve the utility of SC light sources.

Continuum generation in conventional holey photonic crystals or tapered single-mode long fibers is complex and leads to undesirable unevenly distributed noise and instabilities for different wavelength regions in the time and frequency domains. may contain distinct substructures in The amplitude of the continuum usually exhibits large fluctuations with a significant excess white noise background, which fluctuations can be revealed using fast detectors and RF spectrum analyzer (RFSA) measurements. can.

A common approach to wavelength conversion is to generate a supercontinuum, then spectrally slice a portion of the continuum and use this slice as the light source for the microscopy setup. However, the continuum chosen will likely contain large amplitude variations (noise), which may not be suitable for some applications.

In WO 2005/010000, the noise from an SC source is reduced by tapering a nonlinear fiber and using a femtosecond pulse source that induces so-called soliton splitting. The abstract of this patent states: "Cherenkov radiation (CR) introduced by a DMM and longitudinal variation of phase-matched states for four-wave mixing (FWM). enables the generation of low-noise supercontinuum." Tapering requires either post-processing techniques that can complicate the fabrication of SC light sources or changes in fiber diameter during fabrication, and the small cross section of the taper can limit the amount of light that can be safely transmitted. have a nature. Additionally, femtosecond pump sources are often relatively complex and expensive.

In (Patent Document 2), a structure having a basic structure capable of generating SC light and capable of shaping the spectral waveform of SC light, adjusting the power of SC light, or adjusting the repetition frequency of a pulse train including SC light is provided. A further comprising light source device is described. The light source device of US Pat. No. 6,200,000 includes an SC fiber pumped at a wavelength of about 1550 nm, and the repetition frequency of the SC optical pulse train from the light source is located between 1 MHz and 100 MHz. Throughout WO 2005/012000, noise is only discussed for single pulses, and it is explained that the noise characteristics of pulsed light P1 are unaffected. Regarding the noise characteristics of the SC light pulse train P2, it is noted that low noise detection is possible through synchronization with a photodetector configured external to the light source device. Noise spectra from SC light sources with different pump wavelengths are different, and therefore noise suppression may be different. WO 2005/010000 refers to a femtosecond pulse train P1. Such pump sources are often relatively complex and expensive.

U.S. Pat. No. 7,403,688 U.S. Patent Application Publication No. 2011/0116282

In view of the foregoing, it is an object of the present invention to provide a low-noise supercontinuum light source and, advantageously, a supercontinuum light source with reduced noise effects in the generated supercontinuum (SC). That is. Advantageously, the supercontinuum light source is suitable for use in optical measurement systems.

In embodiments, the present invention relates to a system suitable for measuring at least one parameter of an object, the system comprising a supercontinuum light source, and it is also an object to provide a method of measurement using the system. is.

These and other objects have been solved by the invention or its embodiments as defined in the claims and as described herein below.
It has been found that the invention and its embodiments have many additional advantages that will become apparent to those skilled in the art from the following description.

The supercontinuum light source of the present invention comprises a light source output, an intermediate supercontinuum light source, and a single-mode coupling unit, wherein the intermediate supercontinuum light source comprises:
a. a seed laser configured to provide a seed pulse with a pulse frequency F _seed ;
b. A pulse frequency multiplier (PFM) configured to double the seed pulse and convert F _seed into a pump pulse with a pulse frequency F _pump , wherein F _pump is higher than F _seed a large pulse frequency multiplier (PFM);
c. configured to receive the pump pulse and convert the pump pulse into supercontinuum light having a supercontinuum spectrum spanning about λ ₁ to about λ ₂ provided as an output of the nonlinear element; a nonlinear element, where λ ₂ −λ ₁ >about 500 nm.

The output from the nonlinear element is coupled to the single mode coupling unit to provide the output from the single mode coupling unit, and the light source output includes the output from the single mode coupling unit. A single mode coupling unit is configured to attenuate and shape the supercontinuum spectrum from the nonlinear element. Preferably, the F _pump is at least about 100 MHz, such as at least about 150 MHz, at least about 200 MHz, at least about 300 MHz, at least about 400 MHz, at least about 500 MHz, at least about 600 MHz, at least about 700 MHz, at least about 800 MHz, at least about 1 GHz. .

In a preferred embodiment of the frequency multiplier, said single mode coupling unit receives said supercontinuum light and divides it into a spectrum such that the output spectrum from said single mode coupling unit ranges from _λ3 to _λ4 . spectrally shaped output spectral output from the single mode coupling unit, wherein λ ₃ −λ ₄ >0, λ ₃ ≧λ ₁ , and λ ₄ ≦λ ₂ , and is different from the spectrum in the λ ₃ -λ ₄ wavelength region from an intermediate supercontinuum source.

The supercontinuum light source of the present invention has been found to have low noise resulting in a highly improved supercontinuum light source especially for applications where low noise is beneficial. The term "low noise" has otherwise been possible with prior art white light SC light sources operating at comparable power levels of output power in the spectral domain, for example when the light source is applied in a measurement system. significantly lower than would otherwise be possible with prior art supercontinuum light sources operating at comparable power levels of output power, and beyond the soliton splitting region. It is understood to mean average noise.

The seed laser of the intermediate supercontinuum source can for example be a modelocked fiber laser, preferably modelocked via SESAM, preferably the gain medium of said fiber laser is YtYb doped fiber, Er doped fiber and Er/Yb doped fiber.

In embodiments, the wavelength region "λ ₃ -λ ₄ " is greater than about 100 nm, such as greater than about 200 nm, greater than about 300 nm, or greater than about 500 nm. In embodiments, wavelength λ ₃ is less than about 1000 nm, such as less than about 900 nm, less than about 800 nm, less than about 700 nm, or less than about 600 nm. In embodiments, λ ₄ is greater than about 1070 nm, such as greater than about 1100 nm, greater than about 1200 nm, or greater than about 1300 nm.

In embodiments, the single-mode coupling unit includes one or more of the following: prisms, low-pass optical filters, high-pass optical filters, band-pass optical filters, and single-mode fibers. Advantageously, the single mode coupling unit is arranged to shape the spectrum from the intermediate supercontinuum light source into a Gaussian spectrum, a double peak spectrum or a flat top spectrum.

In an embodiment, the attenuation of the supercontinuum spectrum in said single-mode coupling unit is given by an optical power attenuation factor y, said optical power attenuation factor y being the ratio of optical power attenuation within the wavelength region of λ ₄ to λ ₃ . A measure, the optical power attenuation coefficient y is greater than about 2, such as greater than about 3, greater than about 4, greater than about 6, greater than about 8, greater than about 10.

In an embodiment, the single-mode coupling unit performs at least one of the following to perform said attenuation: i) a mismatch or mismatch in the output from the non-linear element to the single-mode coupling unit; iii) a broadband attenuation filter such as a neutral density filter or a broadband beam splitter;

In an embodiment, the single mode coupling unit has an input for coupling to the nonlinear element and a dichroic element at the input of the single mode coupling unit configured to transmit wavelengths below the threshold wavelength _λ5 . and _at least one of: a _prism , a low-pass optical filter, a high-pass optical filter or a band-pass optical filter; and a single-mode fiber, the output of which is a single a single mode fiber output from the one mode coupling unit. Advantageously, the dichroic element is a single mode fiber, said single mode fiber being a step index fiber or a microstructured fiber comprising microstructures in the form of air or low index glass material.

In embodiments, the total optical power at output from the single mode coupling unit is less than about 100 mW, such as less than about 50 mW, less than about 30 mW, less than about 20 mW.

In embodiments, the seed laser is configured to provide a seed pulse with a pulse width t _seed , said pulse width t _seed being greater than about 0.1 ps, such as greater than about 0.25 ps, and being about 0 >.5 ps, >0.75 ps, >1 ps, >2 ps, >3 ps, >5 ps, >10 ps, >20 ps, >50 ps, >100 ps Longer than about 200 ps, longer than about 300 ps, longer than about 400 ps, longer than about 500 ps, longer than about 1 ns.

In embodiments, the seed laser is configured to provide a seed pulse with a pulse width t _seed , said pulse width t _seed being less than about 1 μs, such as less than about 500 ns, less than about 200 ns, about less than 100 ns, less than about 50 ns, less than about 20 ns, less than about 10 ns, less than about 1 ns, less than about 500 ps, less than about 100 ps, less than about 50 ps, less than about 25 ps, less than about 20 ps, about Shorter than 15 ps, shorter than about 10 ps.

Advantageously, the nonlinear element is an optical fiber, such as a tapered and/or non-tapered microstructured fiber.
In embodiments, the intermediate supercontinuum light source includes a pulse compressor, such as a PBG fiber, said pulse compressor receiving pulses from said pulse frequency multiplier (PFM) and time-compressed pulses. configured to output to the non-linear element. Advantageously, the intermediate supercontinuum light source is a non-coherent light source.

The system is suitable for measuring at least one parameter relating to an object and comprises a supercontinuum light source of the invention and at least part of the output of said single mode coupling unit, e.g. The system further includes a detector for detecting light from the object.

Due to the supercontinuum light source of the present invention, including the low noise intermediate supercontinuum light source, a highly accurate optical measurement system is achieved.
In embodiments, the system includes a subject, where the subject is a part of a human or animal body, such as a mammalian eye or any part thereof. This allows in vivo and/or in vitro measurements of parts in the human or animal body.

Advantageously, the detector is greater than 50/F _pump , such as greater than 100/F _{pump, greater than 200/F pump ,} greater than 500/F _pump , greater than 1000/F _pump , greater than 5000/F _pump It has a long integration time.

In embodiments, the measurement system is a reflection mode measurement system configured to measure light reflected from said object, such as a system based on white light interferometry, optical coherence tomography (OCT). Advantageously, the system is based on time domain, frequency domain or swept source OCT.

In embodiments, the measurement system is used to diagnose age-related macular degeneration (AMD), diabetic retinopathy or glaucoma.

In embodiments, the measurement system is used for diagnostics associated with treatments that provide refractive correction of the eye, such as Laser Eye Surgery to Correct the Refractive Condition of the Eye (LASIK). In an embodiment, the measurement system is used to measure the boundaries of Bowman's layer inside the human eye.

A method of the invention for measuring at least one parameter related to a measurement object comprises providing a supercontinuum light source of the invention and a supercontinuum light source of the invention, such as all of the outputs of said single-mode coupling unit. illuminating an object to be measured with at least a portion of the output of said single-mode coupling unit of and detecting light from said object with a detector.

Advantageously, due to the high precision of the optical measurement system, the object is a human or animal body part, such as a mammalian eye or part thereof. This allows in vivo and/or in vitro measurements of parts in the human or animal body.

In the following, the invention will be described in relation to silica-based nonlinear fibers. However, as will be apparent to those skilled in the art, the present invention also applies to other materials such as fibers based on other materials (such as polymers, chalcogenides and fluoride glasses), nonlinear planar waveguides and gas-filled hollow core fibers. It includes SC light sources based on nonlinear elements of the type. For silica-based fiber parameters, material and/or waveguide-based parameters, such as dispersion and nonlinearity, must be adjusted accordingly.

Typically SCs are produced by applying a pulsed pump light source configured to pump a nonlinear fiber, such as the nonlinear fiber described above. A nonlinear process in the nonlinear element transforms the pump pulse into a supercontinuum that exits the fiber. Of particular interest is when substantial pump energy is delivered to wavelengths in nonlinear fibers that exhibit anomalous dispersion. because this greatly extends the achievable bandwidth. In particular, supercontinuum generation is efficient and described by Dudley et al., Rev. Mod. Phys., 78, 4, 2006. It is based on so-called modulation instability, in which the pump pulse is split into a series of short pulses (solitons) that allow the generation of a broad supercontinuum spectrum. In the normal dispersion regime, supercontinuum generation is primarily by self-phase modulation (SPM), which requires very high peak intensities to induce significant spectral broadening (e.g., >100 nm 10 dB bandwidth). caused.

Thus, in embodiments, the pump pulse and nonlinear fiber (i.e., nonlinear element) are configured such that the supercontinuum spectrum is generated primarily through modulation instability (MI)-induced splitting of the pump pulse. be done. That is, most of the input pulse power starts in the anomaly region or at wavelengths located sufficiently close together to allow initial spectral broadening via SPM to shift a substantial portion of the power into the anomaly region. be done. Preferably, more than 50%, such as more than 60%, more than 70%, more than 80%, more than 90%, more than 95%, 100% of the supercontinuum spectrum produced are MI and solitons produced by MI. generated through subsequent processes, including Any residual pump light exiting the nonlinear element is not considered part of the generated supercontinuum. In embodiments, these percentages are calculated as part of the total supercontinuum power. In embodiments, the percentage is calculated as a percentage of the bandwidth scanned by the supercontinuum.

The high nonlinearity of so-called Highly Nonlinear Fibers (HNLF) is generally the result of relatively small cross-sections that give rise to increased peak intensities, but more importantly, the dispersion of these fibers is , which is typically low and unusual for at least some of the wavelengths, the fiber guides, for example, at the pump wavelength. High nonlinearity guarantees a long effective nonlinear interaction length. Because peak power is maintained, peak power supports soliton formation and MI splitting. In embodiments, soliton formation and MI-induced splitting are the primary mechanisms in ultra-broadband light generation from nonlinear fibers. Other nonlinear processes such as self-phase modulation, cross-phase tuning, self-steepening, and Raman scattering do not require anomalous dispersion, but play a role as well.

The pump pulse and nonlinear element may be configured such that the center wavelength of the pump pulse is preferably in the anomalous dispersion region. Alternatively, the pump wavelength may be ZDW -150 nm or greater, ZDW -100 nm or greater, ZDW -50 nm or greater, ZDW or greater, ZDW +10 nm or greater, ZDW +20 nm or greater, ZDW +30 nm or greater, ZDW +50 or greater, ZDW +100 nm or greater, ZDW +150 nm or greater, etc., with moderate spectral broadening ( It is possible to be in the normal dispersion region (but close enough to the anomaly region) that can transfer a substantial portion of the pump energy to the anomaly region (eg, via SPM or Raman shift). In embodiments, the shape of the resulting supercontinuum spectrum can be largely controlled by varying the distance from the pump wavelength to the intersection between normal and anomalous dispersion. This is the so-called zero dispersion wavelength (ZDW).

The term "substantial pump energy shifted to the anomalous region" refers to greater than 30% of the pulse energy, such as greater than 50%, greater than 60%, greater than 70%, greater than 80%, greater than 90%, greater than 95%, 100% % is understood to mean that the pulse enters the abnormal region before being split.

Dudley et al., "Supercontinuum generation in photonic crystal fiber," Rev. Mod. Phys., Vol. 78, No. 4, ( 2006) p. 1159-1162, the supercontinuum is incoherent if modulation instability is the dominant process in splitting the pump pulse. Incoherent supercontinuum can be understood to arise from noise and thus the temporal and spectral stability of the generated light is compromised. According to the authors, pump pulses with soliton orders (N) in fibers with N<10 provide coherent supercontinuum, whereas pump pulses with N>30 provide incoherent supercontinuum. do. A value of 10≦N≦30 provides a transition between these two states, and supercontinuum spectra can be produced coherently or non-coherently, depending on the exact pump and fiber parameters. Here, the soliton order is defined as in (Equation 1).

ここで、ガンマは、ファイバ非線形性であり、Ｐ０は、パルスピーク電力であり、Ｔ０は、パルス幅であり、β_２は、ポンプ波長におけるファイバの群速度分散である。従って、この式は、短パルスが、より干渉的なスーパーコンティニウム及び従ってより低いノイズを規定するソリトン次数を低減することを確認する。

where gamma is the fiber nonlinearity, P0 is the pulse peak power, T0 is the pulse width, and _β2 is the group velocity dispersion of the fiber at the pump wavelength. This equation therefore confirms that short pulses reduce the soliton order, which defines a more coherent supercontinuum and thus lower noise.

Coherence can be dramatically reduced (and noise dramatically increased) for N>16. Increasing the value of N makes the modulation instability (which is quantum noise induced pulse splitting) faster than the deterministic soliton splitting process. Thus, the transition from soliton splitting to MI-induced splitting marks the separation between low-noise/high-interference and high-noise/low-interference. Fei Lu et al., "Generation of a broadband continuum with high spectral coherence in tapered single-mode optical fibers," Optics Express, Jan. 26, 2004, Vol. 2, No. 2, p. 347-353 (referenced in US Pat. No. 7,403,688 and having authors who agree with us), a short 50 fs pulse provides a relatively low N and the soliton order is tapered Further reduction by the ring provides high spectral coherence and low noise. "Super continuum generation for real time ultrahigh resolution optical coherence tomography", Proc. of SPIE, Volume 6102, 61020H, (200 6 years) in The supercontinuum was generated with a 95 fs pump pulse, and it is concluded that only spectra generated by pumping in the normal region have sufficiently low noise to be applicable. As mentioned above, such spectra are formed by SPM, a deterministic process, thus enabling the generation of low-noise, high-interference SCs.

In embodiments, the nonlinear fiber is not tapered. However, in embodiments, the present invention is combined with the noise reduction effect obtainable via tapering. A new type of tapered fiber suitable for SC generation is described in International Application PCT/Denmark Patent Application Publication No. 2011/050328.

However, if the soliton order of the pump pulse is substantially 16 or more, such as 18 or more, 20 or more, 22 or more, 24 or more, 26 or more, 28 or more, 30 or more, 40 or more, 50 or more, 75 or more, 100 or more , 200 or greater, 300 or greater, 400 or greater, 500 or greater, the present invention, in embodiments, provides incoherent or partially incoherent supercontinents. Enables the application of nium. Thereby, the supercontinuum generation process proceeds primarily through modulation instability.

In embodiments, a soliton order is defined when a pulse is split after, for example, shifting into an anomalous region and/or after traversing a tapered section of a fiber. In embodiments, the soliton orders are defined at the input of the pump pulse to the fiber.

In general, the spectral width of the generated SC depends on the peak power of the pump pulse, so for longer pulses the peak power cannot be reduced arbitrarily to reduce the soliton order. Longer pulses, such as pulses in the ps or ns regime, are often preferred as these pulses often allow simpler pump laser configurations compared to fs lasers. Accordingly, in embodiments, the present invention provides for pulse widths greater than about 0.1 ps, such as greater than about 0.25 ps, greater than about 0.5 ps, greater than about 0.75 ps, greater than about 1 ps, about longer than about 2 ps, longer than about 3 ps, longer than about 5 ps, longer than about 10 ps, longer than about 20 ps, longer than about 50 ps, longer than about 100 ps, longer than about 200 ps, longer than about 300 ps, longer than about 400 ps, about It enables longer pulse width applications, such as applications longer than 500 ps, longer than about 1 ns, longer than about 10 ns.

On the other hand, in SCs generated from very long pump pulses and CWs, pumping suffers from increased noise. While the present invention may reduce sensitivity to noise, it may also be preferable to reduce noise via pulse width reduction as well, so that in embodiments the seed laser may have a pulse width t _seed wherein the pulse width t _seed is less than about 1 μs, such as less than about 500 ns, less than about 200 ns, less than about 100 ns, less than about 50 ns, less than about 20 ns Short, less than about 10 ps, less than about 1 ns, less than about 500 ps, less than about 100 ps, less than about 50 ps, less than about 25 ps, less than about 20 ps, less than about 15 ps, less than about 10 ps.

The variable intervals referred to above may be combined to form closed intervals for pulse widths, such as pulse widths that are 0.1 ps to 1 μs, 0.25 ps to 100 ps, 1 ps to 50 ps.

As mentioned above, SCs are typically generated by applying a pulsed-pump light source. In the supercontinuum light source of the present invention, pump pulses are provided at a repetition rate F _pump that causes an amplitude modulation of the generated _{supercontinuum with the same frequency F pump .} On the other hand, the measurement system of the present invention typically has a measurement time longer than 1/F _pump , i.e., a measurement time over which the repetition rate is not determined and the measurement is integrated such that SC appears as CW radiation. apply to For that reason, pulsed lasers operating in the MHz range are often referred to as "quasi-CW". However, the pulsed nature of the supercontinuum reduces the effective measurement time in the presence of light. Thus, in embodiments, the SC light source is such that the F _pump is 100 MHz or higher, such as 150 MHz or higher, 200 MHz or higher, 300 MHz or higher, 400 MHz or higher, 500 MHz or higher, 600 MHz or higher, 700 MHz or higher, 800 MHz or higher, 1 GHz or higher. Apply a high repetition rate.

As explained further below, the pump laser system typically consists of a main laser oscillator, also called a seed laser, which has one or more options to boost the pulse power level from the seed laser. of optical amplifiers follow. That is, the pump laser may include a MOPA configuration. Depending on the type of seed laser, providing such high repetition rates may not be practical or possible. In an embodiment, the pump laser (also called pump laser system) includes a seed laser configured to provide a seed pulse with a pulse frequency F _seed lower than F _pump , and a seed laser configured to convert F _seed to F _pump . and one or more configured pulse frequency multipliers (PFM).

Preferably, the pulse frequency multiplier of the supercontinuum light source of the present invention includes a splitter for splitting at least one beam of seed pulses into a number of sub-beams, and a first beam configured to recombine at least some of the sub-beams. and preferably the pulse frequency multiplier further includes an adjustable attenuator configured to adjust at least one of the sub-beams.

Beam means here a pulse train.
The splitter can be any kind of splitter. Such splitters are well known in the art.

In embodiments, the pulse frequency multiplier includes an adjustable attenuator configured to receive at least one sub-beam. Preferably, the adjustable attenuator is configured to receive at least one sub-beam with power exceeding the average sub-beam power, and the pulse frequency multiplier optionally comprises a plurality of adjustable attenuators. and preferably each attenuator is configured to receive at least one sub-beam having pulses within a selected peak power range. Advantageously, in order to significantly reduce noise, the adjustable attenuator of at least one of the sub-beams is such that the pulses of the sub-beams combined in the first combiner have substantially identical peak power values. configured to receive the pulse and adjust it to a peak power value corresponding to the peak power value of the pulse in at least one other sub-beam;

In embodiments, the pulse frequency multiplier is configured to time delay at least one of the sub-beams. A time delay can be provided, for example, by first placing a path from the splitter to the combiner of one sub-beam that is shorter than the second path from the splitter to the combiner of the second sub-beam. Preferably, the pulse frequency multiplier is arranged to time delay the at least one sub-beam such that the pulses of the recombined sub-beams in the first combiner are preferably spaced at substantially uniform intervals. be done.

1 shows a schematic intermediate supercontinuum light source suitable for the present invention; An example of a supercontinuum spectrum (10) spanning from λ ₂ of about 460 nm to λ ₁ of about 2400 nm is shown. Fig. 3 shows an example of a pulse frequency modulator (PFM) of an intermediate supercontinuum light source according to the invention; Fig. 3 shows an example of a pulse frequency modulator (PFM) of an intermediate supercontinuum light source according to the invention; 2 shows a measurement setup suitable for measuring intensity noise in the spectrum of an SC light source, such as the SC light source of FIG. 1; An example supercontinuum spectral output from an intermediate supercontinuum light source 100 as well as an example spectral output from a single-mode coupling unit 300 are shown, respectively. 4 shows an exemplary spectral output from a single mode coupling unit; 4 shows an exemplary spectral output from a single mode coupling unit; 4 shows an exemplary spectral output from a single mode coupling unit; Figure 3 shows the average intensity noise of an intermediate supercontinuum source before and after compensation for spectrometer noise; Figure 3 shows the average intensity noise of an intermediate supercontinuum source before and after compensation for spectrometer noise; 1 shows an optical measurement system exemplified as an OCT system using an SC light source as the light source. Figure 2 shows an example of a single-mode coupling unit that includes a dichroic element that is a dichroic mirror, a dispersive element that is a prism, and a single-mode fiber configured to shape the spectrum. Figure 2 shows an example of a single mode coupling unit comprising a single mode fiber dichroic element, attenuation and/or shaping optics, and a second single mode fiber. Three examples of methods for attenuating optical power are shown. Three examples of methods for attenuating optical power are shown. Three examples of methods for attenuating optical power are shown.

FIG. 1a shows the configuration of a preferred intermediate supercontinuum light source 100 included in the supercontinuum light source according to the invention. A master oscillator (or seed laser) provides power along beam path 106 . The components are preferably fiber coupled, but may also be coupled via free space optics. Intermediate supercontinuum light source 100 includes two power amplifiers (PA1 and PA2) 102 and 104 . As noted above, these amplifiers are optional, but provide increased pulse energy and peak power compared to the output from seed laser 101 . The seed lasers 101, PA1 102 and PA4 104 are each pumped by diode lasers, however other pump sources such as electrical power sources could alternatively be used. Optional adjuster 105 is included to illustrate that the intermediate supercontinuum light source may include a feedback system. In this embodiment, a feedback loop is provided by a photodiode 109 that measures a portion of the output 108 and provides one or more beam-related parameters to a decision point 114 that adjusts the input to the nonlinear element 107. ,It is formed. Such an adjuster may, for example, be formed by an adjustable attenuator configured to adjust the optical power entering the nonlinear element 107 . Co-pending U.S. patent application Ser. No. 12/865,503 (which is hereby incorporated by reference) describes alternative configurations of the regulator 105 and photodiode 109, various embodiments of the regulator, focusing the beam on the photodiode, and the possibility of applying a feedback response to one or more of pump sources 110-112, various embodiments of feedback loops in SC light sources (see, eg, FIG. 1 and claims).

PFM 103 may be placed before the first amplifier, between amplifiers, and before the nonlinear fiber. In an embodiment, the pulse train saturates the amplifiers (PA1 and/or PA2) such that the peak power of the pulses from the amplifiers is constant regardless of their input power. In FIG. 1, the PFM is placed between two power amplifiers (PA1 and PA2 in this case). This can be preferred. This is because, in most cases, the PFA redistributes the optical power from the seed laser into a larger number of pulses, and when the seed laser's output pulse is relatively weak, the pulse train is efficiently amplified in subsequent amplifiers. This is because the PFA can have significant insertion loss so that it can generate a pulse train with an average power that is too low to be used. For this reason, it is preferred in embodiments to place the PFM after one or more amplifiers, such as between two amplifiers. On the other hand, placing a PFM after one or more amplifiers increases the nominal power loss due to such insertion loss. For this reason, it is preferred in embodiments to place the PFM in front of one or more amplifiers, such as between two amplifiers. This can also have the effect of reducing the peak power of pulses passing through one or more power amplifiers (or other components in the system), which in turn reduces nonlinearities in the pump laser system, etc. , may have one or more benefits. Such nonlinearities often have the effect of widening the pulse, which can result in a reduction in peak power levels to the nonlinear elements, which in turn can reduce the spectral width of the supercontinuum produced. In embodiments, multiple PFMs are applied, such as multiple PFMs separated by optical components such as optical amplifiers, attenuators, compressors or filters.

In embodiments, there is an upper limit to the allowable average optical power that illuminates the measurement object (also called sample). An example of such an application would be the retinal region of the mammalian eye, in particular the retina, where the subject is sensitive to light power (average and/or peak power) above a certain threshold (which would be the case for most biological samples). Including uses for the department. Examples of applications where the subject is an ophthalmic mammalian eye include OCT imaging of the retina or cornea and imaging using multiphoton fluorescence microscopy of the retina or cornea.

In embodiments, the output of the SC light source, or subsection thereof, should meet one or more of laser standards Classes 1, 1M, 2, 2M, 3R, 3B. In embodiments, the output power of the SC light source is such that the SC light source itself has an output AEL (acceptable emission level).

In embodiments, relatively low noise introduced by the pulse width is desirable, so that pulse widths in the range of 0.5 ps to 30 ps are preferred, such as pulse widths in the range of 1 ps to 20 ps, preferably 2 ps to 20 ps. In embodiments, the average optical power from the SC light source is less than 5 Watts output per ps pulse width, such as less than 3 Watts output per ps pulse width, less than 2 Watts per ps pulse width, less than 1 Watt per ps pulse width Therefore, increasing the average optical power for this system is undesirable. In one embodiment, the total average optical power in the visible region (400 nm-850 nm) is configured to be less than 100 mW, such as less than 50 mW, less than 30 mW, less than 20 mW. As noted elsewhere, reducing the average power after output from an SC light source is inappropriately complicated or impossible because the optical components needed to reduce the power modify the spectrum. often become.

As previously mentioned above, in embodiments the spectral width of the SC produced depends on the peak power of the pulse, at least up to some saturation level where further increases in peak power do not increase the spectral width. Also, the pump light to SC light conversion efficiency depends on the peak power, which means that for a fixed pulse width, the peak power (and corresponding average power) cannot be reduced at all. Below a certain value, the desired spectral width of the generated spectrum is compromised and eventually poor conversion efficiency results in too much unconverted pump light passing through the fiber, which can spoil the sample under observation. have a nature. Thus, in embodiments, resulting in minimal peak power, the insertion of PFM causes an increase in average optical output power relative to configurations in which PFM is omitted. This occurs because the peak power and pulse width remain constant while the pump pulse repetition rate is increased. In embodiments, the optical power is reduced by adjusting the pump energy supplied to the last power amplifier before the nonlinear element, but this compromises the resulting spectral width, as mentioned. there is a possibility.

In embodiments, reduction of the average optical power may be performed by introducing attenuation after a nonlinear element, such as attenuation of the beam away from the beam path or splitting a portion of the beam. Applications that require a tunable portion of the generated spectrum directed to the sample may apply the AOTF to perform such functions. In embodiments, the AOTF may be controlled to reduce the average amount of optical power directed to the sample. For applications requiring broadband illumination, such as, for example, in OCT imaging systems, it can be more difficult to apply optical components to the beam and/or damage said optical elements without disturbing the spectral shape. . In embodiments, the pump laser system includes a pulse compressor, such as a PBG fiber (hollow or solid core), configured to compress the pump pulse and thus increase peak power. This use of PBG fibers was discussed in PCT application WO2005041367. In embodiments, by increasing the peak power of individual pulses, the use of a pulse compressor allows the use of lower average optical power while preserving the spectral characteristics of the generated spectrum.

In principle, the PFM of the intermediate supercontinuum light source of the invention is any optical component suitable for receiving a pulse train at a certain repetition rate and transforming this input into a pulse train with a higher repetition rate. can be In embodiments, the input and output pulses have substantially the same pulse width and wavelength. In an embodiment, PFM works by splitting the pulse train at the input into multiple sub-pulse trains, each undergoing a different delay (optical path length) before being recombined. The relative delay causes the sub-pulse trains to shift in time when recombined, so that the combined pulse train contains more pulses than the input. For example, an input pulse train may be split into two sub-pulse trains (or sub-beams) where one pulse train is delayed with respect to the other. The repetition rate of the combined columns is then doubled. Preferably, the relative shift between beams corresponds to half the spacing between two pulses in the input pulse train. In embodiments, this principle is extended such that the input beam is first split into more than two sub-beams, such as two, three or four sub-beams, each delayed and recombined with respect to each other. . It is well known that optical splitters (or combiners) function in a symmetrical manner. Combining several light beams produces the same amount of output beam. In embodiments, only a single output is used/available, while optical power designated for other outputs is lost in the optical system. Therefore, in embodiments it is advantageous to cascade combiners/splitters, as described below in connection with FIG. 2b.

In embodiments, the invention provides a splitter for splitting a beam into sub-beams, an optional adjustable attenuator configured to receive the sub-beams, and a first combiner configured to combine the sub-beams. and PFM. Thus, the adjustable attenuator may be adjusted to compensate for manufacturing variations in the splitters and/or combiners as well as coupling variations, resulting in pulses with uniform peak power. A pulse train as can be generated. In embodiments, precise adjustment of peak amplitudes is not required and substantial differences between peak powers of recombined sub-beams are acceptable.

In embodiments, one or more of the splitters and combiners have non-uniform splitting ratios (

などであり、ここでｘはパーセンテージであり、例えば４５／５５、４０／６０、３５／６５又は３０／７０である）を有するように構成され、前記減衰器は、最も強力なサブビーム（又はビームの結合において、より大きく寄与するサブビーム）を受信するように構成され、これにより、サブビームが結合される場合に、より強力なサブビームが、他のサブビームと等しい電力レベルを供給するように減衰され得ることを保証し得る。それによって、ノイズは、サブビームが異なる電力レベルを有する状況と比較して、著しく低減される。

etc., where x is a percentage, e.g. sub-beam that contributes more heavily in the combination of the sub-beams, such that when the sub-beams are combined, the more powerful sub-beam can be attenuated to provide an equal power level to the other sub-beams We can guarantee that. The noise is thereby significantly reduced compared to situations where the sub-beams have different power levels.

In an embodiment, the PFM includes multiple attenuators each configured to receive a separate sub-beam. In embodiments, the splitter splits the beam into two sub-beams. In embodiments, the splitter divides the beam into more than 2, eg, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more sub-beams. In an embodiment, the first combiner further acts as a splitter to split the combined beam into a second sub-beam followed by a second combiner. In embodiments, the PFM includes an adjustable attenuator configured to receive one of said second sub-beams. This attenuator may be applied to adjust for coupling losses and other variations, as well as variations in the first combiner and the second combiner. In an embodiment, the second combiner is configured to have a non-uniform splitting ratio (and thus also a non-uniform combination of input beams), and the output from said adjustable attenuator has a larger portion configured to feed the output. Again, the second combiner can ensure that a pulse train with uniform power between pulses can be delivered by the PFM.

In embodiments, the PFM is formed by a free-space optical system such as a bulk beamsplitter. In embodiments, the PFM is formed by optical fiber splitters and/or combiners, which are often preferred for system cost and robustness.

Figure 1b shows an example of a supercontinuum spectrum (10) spanning from _λ2 , which is about 460 nm, to _λ1 , which is about 2400 nm. Spectra are obtained from NKT Photonics A/S product SuperK EXW-12.

Figures 2a and 2b show an example of a pulse frequency modulator (PFM) of an intermediate supercontinuum light source according to the invention. FIG. 2a shows an embodiment of PFM 200. FIG. An input beam (either in free space or through a fiber) enters the PFM at input 201 . Splitter 214 is illustrated as a 1×2 splitter, but may be any 1×N splitter, or even an M×N splitter. For an M×N splitter, multiple inputs may be combined, or alternatively only one input out of the M available inputs is used. A first splitter 214 splits the input beam into two sub-beams 202, 203 with a splitting ratio x ₁ /(1−x ₁ ). As discussed above, in embodiments the greater of x ₁ and (1−x ₁ ) is sent to adjustable attenuator 204 . In an embodiment, the attenuator is omitted, in which case _x1 is preferably about 0.5 (ie, 50%) so that variations in the peak power of the pulse train at output 207 can be minimized. Sub-beam 202 is fed into delay line 205 , which is preferably configured to delay sub-beam 202 by half the period between two pulses in input beam 201 . In embodiments, the delay line is adjustable to accommodate variations in the repetition rate of the input beam. In embodiments, small deviations from uniform spacing of pulses in the output beam (e.g., less than 75%, such as less than 50%, such as less than 25%, such as less than 15%, such as less than 10%, such as less than 5%, such as 1% less than ) can be tolerated because the delay line is fixed. Sub-beams 202 , 203 are combined at combiner 206 which provides output 207 . Combiner 206 has a division ratio of x ₂ /(1−x ₂ ). In an embodiment, either splitter 214 or combiner is configured to have a non-uniform splitting ratio, i.e., x ₁ or x ₂ deviates from 50%, and in this way attenuator 204 Beams 202 and 204 may be adjusted to contribute equally so that pulses at the input that are split into two pulses are recombined to have substantially the same peak power at the output, where "Substantially" means including within normal tolerances. The effect of PFM is a doubling of the pulse frequency of the input beam. Combiner 206 further has an output 208, which may or may not be a real, physically available output. However, if other optical losses (such as in couplings and attenuators) are neglected, the combiner should be less than the beam splitter/combiner so that the peak power is reduced to about 25% of the input peak power. Output 208 is included to illustrate the introduction of insertion loss due to inherent symmetry. In an embodiment, the output 208 beam is applied to monitor the beam and adjust the attenuator 204 .

FIG. 2b shows the PFM of FIG. 2a, but including a second combiner 213 so that the PFM provides a quadrupling of the pulse frequency. In principle, quadrupling can also be obtained by expanding splitter 214 to a 1×4 and combiner 206 to a 4×1 combiner. However, in this case the combiner would impose an insertion loss of approximately 75% due to the symmetry of the beam splitter, compared to the approximately 50% loss imposed by the second combiner 213 . The first delay line is preferably adjusted to half the input period at 201, the first delay line doubles the pulse repetition rate after combining at combiner 206, and the second delay line 212 is the It is preferably arranged to provide a delay of half or one quarter of the input period at 201 . The division ratio x ₂ /(1−x ₂ ) is set equal in an embodiment, and x ₁ and x ₃ are set so that attenuator 204 functions as described in connection with FIG. 2a. It is feasible and non-uniformly configured such that attenuator 211 may perform a similar function of compensating for variations in the division of 206 as in combining in combiner 213 . It is worth noting that further doublings can be obtained by adding couplers and further extending the PFM without increasing the insertion loss due to the symmetrical splitting.

Figure 3a shows a measurement setup in which the SC light source 1000 of the present invention is configured to illuminate the spectrometer rather than the measurement target. FIG. 3 a shows that the supercontinuum light source 1000 of the present invention includes an intermediate supercontinuum light source 100 and a single mode coupling unit 300 . The output from SC light source 1000 is the output from single mode coupling unit 300 . The output of intermediate SC light source 100 is the output from nonlinear element 107 (not shown in FIG. 3a). This output from intermediate SC light source 100 is coupled to the input to single mode coupling unit 300 . The output of the intermediate SC light source 100 is at least approximately the output from the nonlinear element (107 in FIG. 1a, not shown in FIG. 3a) of the intermediate supercontinuum light source (100 in FIG. 1a). Single mode coupling unit 300 includes adaptations in the form of spectral attenuation and/or shaping according to application requirements. In one embodiment, the SM coupling unit 300 comprises one of the embodiments in co-pending PCT Application PCT/Denmark Patent Application Publication No. 2011/050475, which is hereby incorporated by reference. See in particular any one of the items and/or claims, as well as the embodiments and variations thereof relating to Figures 5a, 6, 7, 8-10, 13-15 and 17-19.

FIG. 3b shows an example supercontinuum spectral output from the intermediate supercontinuum light source 100 (spectrum 10) as well as an example spectral output from the single-mode coupling unit 300 (spectrum 12), respectively. In this example, the spectrum after the single-mode coupling unit has a Gaussian distribution and extends from λ ₄ at about 650 nm to λ ₃ at about 950 nm. Therefore, Fig. 3b shows that the spectral shape after the single-mode coupling unit is different from the spectral shape in the same wavelength region from the intermediate supercontinuum source.

Figures 3c, 3d and 3e show example spectral outputs from the single-mode coupling unit 300, namely spectral shapes that are Gaussian (Figure 3c), flat-top (Figure 3d), and double-peaked distributions (Figure 3e), respectively. show. A double-peaked distribution can be advantageous when the output from the light source is sent through an optical element (such as an optical lens) with a Gaussian like transfer function before illuminating the target, and Illuminating the object with a flat-top distribution is advantageous.

In one embodiment, the spectral shape after the single-mode coupling unit is different than the spectrum in the same wavelength region from an intermediate supercontinuum source, such as a Gaussian, flat-top, or double-peaked distribution. Figures 4a and 4b show the measurement results from the set-up according to Figure 3a. An intermediate SC light source was constructed according to FIG.

Figure 4a shows the power of a 400-850 nm supercontinuum source as a function of power, using a Wasatch Cobra UD spectrometer (310) equipped with a Basler Sprint SPL4096-70km camera. Figure 1 shows the average intensity noise of the intermediate supercontinuum light source 100 (see Figure 1) measured from 790 to 870 nm. Figure 4a shows the average intensity noise after compensation for spectrometer noise, while Figure 4b shows the average intensity noise before compensation for spectrometer noise. FIG. 4a contains measurements for three different pump pulse frequencies (F _pump ), which are 80 MHz (curve 401), 160 MHz (curve 402) and 320 MHz (curve 403). It can be seen that the noise decreases as the pump pulse frequency increases. Intensity noise is compensated for noise added by the spectrometer.

FIG. 4b shows the intensity noise data from FIG. 4a before compensation for noise from the spectrometer. FIG. 4b contains measurements for three different pump pulse frequencies (F _pump ), which are 80 MHz (curve 411), 160 MHz (curve 412) and 320 MHz (curve 413). Again, it can be seen that the noise decreases as the pump pulse frequency increases.

MO101 is a mode-locked Yb fiber laser with an output having a center wavelength of about 1060 nm and a pulse width of about 6 ps. The laser is passively mode-locked via a SESAM to deliver pulses with a repetition rate of 80 MHz. This laser type is well suited for seeding. This is because the all-fiber configuration provides a laser that is robust and relatively simple to manufacture for a bulk optical setup. The maximum repetition rate is determined by how short the cavity can be made and the response characteristics of the SESAM. In practice, these limits often impose a practical upper limit on repetition rates of about 100 MHz. In embodiments, other gain media may be applied to provide other output wavelengths, and the pulse width and repetition rate may also be varied within limits discussed elsewhere.

In embodiments, the seed laser is a fiber laser, such as a mode-locked fiber laser, such as mode-locked via SESAM. The gain medium may be formed by any suitable laser gain, such as Yb-doped fiber, Er-doped fiber and Er/Yb-doped fiber. The seed laser may be, for example, a linear cavity laser or a ring laser.

The nonlinear medium 107 is a microstructure formed by a silica core surrounded by a hexagonal pattern of holes configured such that the core is formed by missing holes in the pattern. Structural PCF fiber. The fiber is constructed so that the ZDW of the fiber is relatively close to the pump wavelength because the substantial pump energy is delivered in the anomalous region of the fiber.

As in FIG. 1, optical fiber amplifier sets 102, 104 are arranged around the optional PFM. Without PFM, the pump system pumps the fiber at about 10 W, 8-10 ps at 80 MHz. By inserting a PFM according to Figure 2a, the repetition rate is increased to 160 MHz, and by inserting a PFM according to Figure 2b, the repetition rate is quadrupled to 320 MHz. Figures 4a and 4b were equipped with a Basler Sprint SPL4096-70km camera configured to measure the 790-870 nm spectral region with 4096 pixels, or approximately 0.02 nm/pixel. Experimental results obtained using a Wasatch Cobra UD spectrometer are shown. A measurement time of 12.9 μs was applied and the power variation measured at each pixel was recorded. Long and short measurement times, such as 1 μs to 1 ms or longer, are possible. Short measurement times are often desirable, such as for Fourier-domain OCT (see FIG. 4b) where real-time imaging is often required. In FIG. 4, the average relative standard deviation per pixel in the 790-870 nm spectral region is measured as a function of the visible part of the spectrum. The standard deviation and thus the intensity noise drops significantly as the pump pulse repetition rate is doubled and also when the pump pulse repetition rate is quadrupled for the equivalent of the average power in the visible region. It is observed that The amount of power in the visible range depends on how effectively the pump energy is converted to visible light which depends on the peak power of the pump pulse and the total amount of pump power (average power). Whereas in Figure 4a the estimated noise contribution from the spectrometer was subtracted, this is included in Figure 4b.

FIG. 5 shows an optical measurement system exemplified as an OCT system using an SC light source as the light source. The system shown in FIG. 5 is a Fourier-Domain OCT (FD-OCT) system according to the invention, in which the SC light source 1000 is therefore a suitable light source for the optical measurement system according to the invention. Applies. A 2×2 50/50 directional splitter/combiner (501) coupled to a light source and spectrometer (310) serving as detection on one side, and a lens (502), object (503) and reference on the other side. A reflector (504) constitutes the interferometer core of the OCT system. A line scan (depth profile of the sample) is performed by spectrometer measurement, where the measurement depth is determined by the spectral resolution and the spatial resolution in the sample is determined by the spectral width of the measurement. The beam is often scanned over the object to provide a 2D or 3D depth profile of reflectance in the sample. OCT is often a broad field with many variations in system configuration, all of which are expected to benefit from aspects of the present invention. The output spectrum is for example similar to FIG. 16 configured to provide a broad and tunable spectrum, FIG. 5a (single band Gaussian spectrum) and FIG. 6 (dual-band Gaussian spectrum), such that in one embodiment the SM combining unit is configured to shape the spectrum from the SC light source into a Gaussian spectrum. In one embodiment, the SM combining unit includes a filter configured to provide a Gaussian spectrum. A 50/50 coupler should be configured to handle a broad spectrum and is typically either a fused fiber coupler or a bulk optical coupler.

FIG. 6 shows an example of a single-mode coupling unit 300 that includes a dichroic element that is a dichroic mirror, a dispersive element that is a prism, and a single-mode fiber configured to shape the spectrum. Accordingly, FIG. 6 shows an example of how a single mode coupling unit 300 may be constructed. The output of intermediate supercontinuum light source 100 is directed to dichroic element 60 and dispersive element 61 . The mirrors and/or angle dispersive elements are connected to an electronic control 6 which allows rotation between these two elements. The system may also optionally include adjustable attenuation filter 62 and/or adjustable spatial filter 63 . The light is collimated by lens system 64 and focused by fiber 65, which in turn shapes the spectrum. The system may optionally include a broadband splitter 66 that sends a portion of the light to the output 67 and another portion of the light to the detection system 68 . Said detection system is connected to an electronic control system 6, which is again connected to the supercontinuum light source 100 and/or the dichroic element 500 in order to stabilize the output power. In one embodiment, the dispersive element is a prism. In one embodiment, fiber 65 is a single mode fiber such as a step-index fiber or a microstructured fiber. In one embodiment, collimating lens system 64 includes multiple lenses.

FIG. 7 shows an example of a single mode coupling unit 300 including a dichroic element which is a single mode fiber 60, attenuation and/or shaping optics 70, and a second single mode fiber 65. FIG.

In one embodiment, the first single mode fiber 60 has high loss above some threshold wavelength λ ₆ and thus acts as a spectral filter. In one embodiment, the attenuating and/or shaping optical element is selected from the list of prisms, optical low-pass, optical high-pass and optical band-pass filters, and neutral density filters.

Figures 8a-8c show three examples of methods for attenuating optical power in the supercontinuum light source of the present invention.
8a-8c, the supercontinuum light source is indicated by reference number 1000, while the intermediate supercontinuum light source is indicated by reference number 100 and the single mode coupling unit is indicated by reference number 300. ing.

In FIG. 8a, the single mode coupling unit 300 comprises an attenuation and shaping unit 81, the mode field diameter at the output of the attenuation and shaping unit 81 being equal to the mode field diameter of the second single mode fiber 82 Different from diameter. FIG. 8a thus shows the mode field diameter mismatch at the output of the damping and shaping unit 81 of the single-mode coupling unit 300 .

In FIG. 8b the single mode coupling unit 300 comprises a damping and shaping unit in the form of a shaping element 60 and a damping element 84. In FIG.
FIG. 8c shows an example where the attenuation in the single mode coupling unit 300 is obtained by having an optical splice with large loss 86 between the intermediate supercontinuum light source 100 and the input of the single mode coupling unit 300. show.

It should be emphasized that the term "including", as used herein, is to be interpreted as an open term. That is, it is to be understood to specify the presence of specifically stated features such as elements, units, integers, steps, components and combinations thereof, but one or more other stated features. Does not exclude the presence or addition of features.

Further, the term "substantially" is meant to include within normal tolerances.
All features of the invention, including regions and preferred regions, may be combined in various ways within the scope of the invention unless there are specific reasons not to combine such features.

Claims (12)

A supercontinuum light source,
a seed laser configured to provide a seed pulse having a pulse frequency F _seed ;
A pulse frequency multiplier (PFM) configured to convert a pulse with a pulse frequency F _seed into a pump pulse having a pulse frequency F _pump so as to double the seed pulse, wherein F _pump is higher than F _seed . a large, said pulse frequency multiplier (PFM);
a nonlinear element configured to receive the pump pulse and convert the pump pulse into a pulse of supercontinuum light, the nonlinear element comprising a microstructured optical fiber, the supercontinuum light comprising: a nonlinear element provided as an output of the nonlinear element and having a supercontinuum spectrum over a range greater than about 500 nm ;
The pulse frequency multiplier is
a splitter for splitting the seed pulse into first and second sub-beams having respective pulse frequencies F _seed , wherein the pulse frequency multiplier is configured such that one of the first and second sub-beams is one of the first and second sub-beams; the splitter configured to delay with respect to the other of the sub-beams of
A combiner/splitter cascaded after said splitter, said combiner/splitter combining said first and second sub-beams and forming a combined sub-beam greater than F _seed and greater than F _pump dividing into a plurality of further sub-beams each having a smaller pulse frequency, the pulse frequency multiplier being configured to delay one of the plurality of further sub-beams with respect to the other of the plurality of further sub-beams , the combiner/splitter;
a combiner for combining said plurality of further sub-beams into a beam having a pulse frequency F _pump . the seed laser configured to provide a seed pulse having a pulse width t _seed ;
The pulse width t _seed is longer than about 0.1 ps, such as longer than about 0.25 ps, longer than about 0.5 ps, longer than about 0.75 ps, longer than about 1 ps, longer than about 2 ps, longer than about 3 ps longer than about 5 ps, longer than about 10 ps, longer than about 20 ps, longer than about 50 ps, longer than about 100 ps, longer than about 200 ps, longer than about 300 ps, longer than about 400 ps, longer than about 500 ps, greater than about 1 ns The supercontinuum light source of claim 1, elongate. the seed laser configured to provide a seed pulse having a pulse width t _seed ;
wherein the pulse width t _seed is less than about 1 μs, such as less than about 500 ns, less than about 200 ns, less than about 100 ns, less than 50 ns, less than 20 ns, less than about 10 ns, less than about 1 ns, less than about 500 ps 3. The supercontinuum light source of claim 1 or 2, short, shorter than about 100 ps, shorter than about 50 ps, shorter than about 25 ps, shorter than about 20 ps, shorter than about 15 ps, shorter than about 10 ps. The supercontinuum light source of any one of claims 1-3, wherein the seed laser comprises a mode-locked Yb laser. 5. The supercontinuum source of claim 4, wherein the mode-locked Yb-laser comprises a fiber laser passively mode-locked via a semiconductor saturable absorber mirror (SESAM). A supercontinuum light source according to any preceding claim, wherein the pulse frequency multiplier includes a delay line for delaying one of the first and second sub-beams. A supercontinuum light source according to any preceding claim, wherein the pulse frequency multiplier comprises an attenuator for attenuating sub-beams. A supercontinuum light source according to any preceding claim, wherein one of said splitter and said combiner/splitter has a non-uniform splitting ratio. A supercontinuum light source according to any one of the preceding claims, further comprising first and second amplifiers, wherein said pulse frequency multiplier is arranged between said first and second amplifiers. . 4. The claim wherein the F _pump is at least about 100 MHz, such as at least about 150 MHz, at least about 200 MHz, at least about 300 MHz, at least about 400 MHz, at least about 500 MHz, at least about 600 MHz, at least about 700 MHz, at least about 800 MHz, at least about 1 GHz. 10. The supercontinuum light source according to any one of 1 to 9. 2. The supercontinuum light source is configured such that the pulses of supercontinuum light have a pulse width, and wherein the average optical power output from the supercontinuum light source is less than 5 watts output per ps pulse width. 11. The supercontinuum light source according to any one of 10. A supercontinuum light source according to any preceding claim, wherein the supercontinuum light source is configured such that the total average optical power in the range 400nm to 850nm is less than 100mW.

Images