JPWO2015030202A1 - Optical measuring apparatus, optical measuring method, and microscopic imaging system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a minute signal with high sensitivity by effectively removing noise in optical measurement. An optical measurement apparatus includes a first light source that emits pump light that excites a sample, a second light source that emits probe light that measures changes in the excited sample, and probe light that detects the probe light. And a beam splitter that divides the reference probe light, a first input port that receives the detection probe light that has passed through the excited sample, and a second input port that receives the reference probe light that does not pass through the excited sample. And a balance detector having an output terminal for outputting an electric signal representing a difference between the detection probe light and the reference probe light, and a lock-in for detecting a desired signal at a predetermined lock-in frequency from the output of the balance detector. And an amplifier.

Description

  The present invention relates to an optical measurement technique for performing optical measurement with high sensitivity and high resolution using pump light and probe light, and a microscopic imaging system using the optical measurement technique.

  Unlike an atomic force microscope, an optical microscope can observe an object in a non-contact manner. In addition, unlike an electron microscope, living body activity can be observed alive.

  One optical measurement is a pump-probe measurement method using pump light and probe light. Pump light is incident on the sample to excite molecules in the sample, and another laser beam (probe light) is incident on the sample to detect a change in transmittance (or reflectance). The pump-probe measurement method can measure the dynamics of the photoexcited state of molecules with high time resolution, and is widely applied to the measurement of optical materials and biological samples.

  The measurement that performs the pump-probe measurement with the nanoscale spatial resolution under the microscope is the pump-probe microscopic measurement. In this method, a microscopic image of the pump-probe signal can be obtained by scanning the beam position on the sample and acquiring data at each point.

For example, when a molecule to be measured absorbs light, stimulated emission of the molecule, excited state absorption, bleaching due to the decrease of the molecule in the ground state due to excitation, changes in the refractive index near the molecule due to temperature rise, etc. Contributes to transmittance change. The amount of change in the probe light intensity is 10 ⁻⁶ to 10 ⁻⁷ per molecule. The volume of the focal plane of the laser beam that passes through the objective lens of the optical microscope is 10 ⁻¹³ cm ³ , and the number of molecules present therein is limited. Therefore, the signal obtained by the pump-probe microscopic measurement is very high. small.

  There is a lock-in detection method as a method for measuring a weak signal. This method takes a product of a trigonometric function of a set frequency (lock-in frequency) and a signal, converts the signal of that frequency to a low frequency, and then extracts a signal component by a low-pass filter. In this respect, lock-in detection effectively acts as a narrowband bandpass filter. That is, noise observed at a frequency different from the lock-in frequency is removed, and a signal near the lock-in frequency is detected, thereby improving the signal-to-noise ratio. In addition, by taking a long time constant of the low-pass filter, the bandwidth can be narrowed and noise can be removed more effectively. In the pump-probe measurement method, the pump light intensity is mainly modulated by the lock-in frequency, and the modulation of the probe light by the pump light irradiation is detected using a lock-in amplifier.

  However, noise observed near the lock-in frequency cannot be removed in principle. In addition, if the time constant of the low-pass filter is increased, the response to changes in the signal is delayed, and a long time measurement is required to obtain an image by measuring at each point while scanning the sample. In addition to the lock-in detection method, a more fundamental countermeasure is required to detect the weak signal of the pump-probe microscopic measurement buried in the noise.

  In the pump-probe method, in order to detect a slight change in the probe light after passing through the sample, it is effective to improve the signal-to-noise ratio to remove the influence of the fluctuation of the probe light intensity. In general, the fluctuation of the laser light intensity is dominant in the low frequency region, and the influence is less at higher frequencies. For this reason, it has been proposed to reduce the influence of noise by performing pump-probe microscopic measurement at a high lock-in frequency of several megahertz (MHz) to several tens of megahertz (MHz) (for example, Non-Patent Document 1 and 2).

The method for improving the signal-to-noise ratio by setting the lock-in frequency high has the following problems or limitations.
(1) The detection signal cannot be measured unless the response of the molecule of the sample follows the lock-in frequency (the modulation frequency of the pump light). For example, when measuring the stimulated emission signal of a fluorescent molecule, the upper limit of the lock-in frequency is determined by the value of the fluorescence lifetime. Further, when considering the triplet state of the fluorescent molecule, the lifetime is microseconds. Therefore, modulation of light intensity at several MHz or more causes a decrease in signal intensity. Therefore, microseconds caused by reaction intermediates, triplet excited states, thermal diffusion processes of nanostructures, structural changes of polymers and proteins, etc., which are required in the fields of biological function analysis, medicine and materials science, etc. The relatively slow time-scale photoexcitation dynamics of 1 to millisecond cannot be measured.
(2) When pulse light is used as the light source, the pulse repetition frequency must be larger than the lock-in frequency. Moreover, since it is limited to a light source with high intensity stability between pulses, the types of light sources that can actually be used are limited. At present, the light source used in the pump-probe microscope is mainly limited to a titanium sapphire oscillator having a repetition frequency of several tens of MHz in the near-infrared wavelength, and an optical parametric amplifier or the like is used to obtain light of a specific wavelength. A large-scale device for wavelength conversion is required outside. In particular, supercontinuum light, which is a white light source in the entire visible light region, has a large intensity fluctuation between pulses, and thus noise cannot be removed even by a high-frequency lock-in detection method.
(3) Semiconductor lasers are small and highly efficient, and devices having various oscillation wavelengths have been developed in recent years, and are optimal as continuous wave light sources for laser microscopes. However, it is necessary to suppress the influence of the return light of the optical system in order to reduce noise. In addition, it is necessary to stabilize operating conditions such as operating current and temperature to suppress mode hops in which the longitudinal mode of laser oscillation fluctuates in an unstable manner. Providing such a stabilization mechanism is costly. In addition, the quantum noise intrinsic to the laser action has a resonant peak in the gigahertz (GHz) band, and a special mechanism such as optical feedback is required to suppress the quantum noise with a semiconductor laser.

  In addition, the structure which uses a balance detector for the laser measurement of a gas sample instead of a pump-probe measurement is known (for example, refer patent document 1).

  On the other hand, the size of the structure that can be observed with a normal optical microscope is determined by the diffraction limit of light and is about half of the wavelength. Recently, various super-resolution optical microscopes that exceed the diffraction limit of light have been proposed. . In a STED (stimulated emission depletion) microscope proposed in 1994, a single laser beam having an annular intensity distribution at the focal plane is used to suppress emission of molecules at the focal edge of the excited fluorescent molecules. (Stimulated emission light) is incident to achieve super-resolution. With a STED microscope, it is possible to observe a biological sample with a resolution about 4 to 5 times the diffraction limit (50 nm or less), but a high-intensity laser of several hundred mW or more is necessary to efficiently induce stimulated emission. It is.

  In addition, a method has been reported that uses an ultrashort pulse laser as a light source to combine a ring beam and two-photon excitation (fluorescence microscopy, current measurement microscopy) to suppress side lobes and improve resolution (for example, Non-Patent Documents 3 and 4). However, two-photon excitation requires a near-infrared light source having a wavelength twice that of one-photon excitation, making it difficult to improve resolution compared to a visible light source, and is difficult to put into practical use at high cost.

US Pat. No. 5,134,276

W. Min, S. Lu, S. Chong, R. Roy, G.R.Holtom, and X.S.Xie, “Imaging chromophores with undetectable fluorescence by stimulated emission microscopy”, Nature, 461, 1105-1109 (2009). L. Tong, Y. Lie, B.D. Dolash, Y. Jung, M.N.Slipchenko, D.E.Bergstrom and J.X.Cheng, Nature nanotechnology 7 (1), 56-61 (2012) S. W. Hell, et al. “Annular aperture two-photon excitation microscopy”, Optics Communications 117 (1995), 20-24 K. A. Serrels, et al. “70 nm resolution in subsurface optical imaging of silicon integrated-circuit using pupil-funcion engineering“, Applied Phyusics Letter 94, 073113 (2009)

  Therefore, it is an object to provide a high-sensitivity and high-resolution pump-probe measurement technique that reduces noise with a simple structure regardless of the type of light source.

  In order to solve the above problems, in optical measurement using pump light and probe light from an arbitrary light source, it is highly sensitive to weak signals caused by light excitation by performing lock-in detection after performing balanced homodyne detection. Detect with.

  In addition, high resolution measurement exceeding the diffraction limit is realized by making the pump light and the probe light into an annular beam.

Specifically, the optical measurement device according to one aspect of the present invention is:
A first light source that emits pump light for exciting the sample;
A second light source for irradiating probe light for measuring a change in the excited sample;
A beam splitter that divides the probe light into detection probe light and reference probe light;
A first input port that receives the detection probe light that has passed through the excited sample, a second input port that receives the reference probe light that does not pass through the sample, and a difference between the detection probe light and the reference probe light A balance detector having an output for outputting an electrical signal representing
A lock-in amplifier for detecting a desired signal at a predetermined lock-in frequency from the output of the balance detector;
Have

  As a preferred configuration example, the probe light is modulated at the first modulation frequency, the pump light is modulated at the second modulation frequency, and the lock-in amplifier is balanced with a reference signal having a difference frequency between the first modulation frequency and the second modulation frequency. The signal is locked in at the difference frequency with the output of the detector as an input.

  As another preferred configuration example, pump light having a first wavelength and probe light having a second wavelength different from the first wavelength are formed into an annular beam, and the pump light and the probe light formed into the annular beam are sampled. And detecting a balance between the probe light transmitted through the sample and the reference probe light not transmitted through the sample, and then lock-in detection of a desired signal at the predetermined lock-in frequency.

  Laser noise can be efficiently removed regardless of the type of light source, and a minute photoexcitation signal can be detected with high sensitivity. Further, by using an annular beam for the pump light and the probe light, it is possible to realize high resolution measurement exceeding the diffraction limit.

It is a schematic block diagram of the optical measuring device of 1st Embodiment. It is a figure which shows the structural example of the light source applied to the apparatus of FIG. It is a figure which shows another structural example of the light source applied to the apparatus of FIG. It is a figure which shows the noise cancellation effect by a balance detector. It is a figure which shows the noise cancellation effect by a balance detector. It is a figure which shows the pump-probe microscopic image by the optical measuring device of 1st Embodiment. It is a schematic block diagram of the optical measuring device of 2nd Embodiment. It is a figure explaining the intensity distribution in the focal plane of an annular beam, and the point spread function of a pump-probe signal. It is a figure which shows the side lobe suppression effect of 2nd Embodiment compared with the case of the two-photon excitation by pump light. It is a pump-probe microscopic image of gold nanoparticles. It is a figure which shows the cross-sectional intensity | strength of the horizontal direction (x direction) and vertical direction (y direction) of the image of FIG. It is a figure which compares the pump-probe microscopic image of the fluorescent bead when using an annular beam with the pump-probe microscopic image using a circular beam. It is a figure which shows the cross-sectional intensity | strength of the excitation spectrum and fluorescence spectrum of Nile red which label the fluorescent bead of FIG. 12, and the I-I 'line of FIG.

With reference to the drawings, an optical measurement technique and configuration according to the embodiment will be described.
<First Embodiment>
FIG. 1 is a schematic configuration diagram of an optical measurement apparatus 1 according to the first embodiment. The optical measuring device 1 includes a pump light source 11, a probe light source 12, an autobalance detector 20, and a lock-in amplifier 38. The pump light source 11 emits pump light (L ₁ ) for photoexciting the sample 10. The probe light source 12 emits probe light (L ₂ ) for detecting a change in transmittance of the sample 10 due to light excitation. The pump light and the probe light are respectively guided to spatial filters 13 and 14 including an objective lens and a pin fall in order to adjust the spatial mode and expand the beam diameter. The probe light L ₂ is split by the beam splitter 15.

One of the divided probe lights enters the reference light detection port 21 of the _autobalance light detector 20 through the neutral density filter 17 as reference light (L _2ref ). The other probe light is spatially superimposed on the pump light as detection light (L _2det ) via the dichroic mirror 18 and enters the sample 10 via the objective lens 31. When the target molecule absorbs light, stimulated emission of the molecule, absorption of the excited state, bleaching due to a decrease in the ground state molecule accompanying excitation, and the like contribute to a change in transmittance of the detection probe light. The transmitted light from the sample 10 is collimated by the condenser lens 32.

The pump light after passing through the sample 10 and the fluorescence generated from the sample 10 are blocked by the color glass filter or the narrow band interference filter 33, and only the probe light (detection signal light L _2det ) _carrying the detection information is _autobalance detected. It enters the signal light detection port 22 of the device 20. At the reference light detection port 21, the reference light _L2ref enters a photodiode (not shown) and is converted into a current. Similarly, at the signal light detection port 22, the detection signal light _L2det is incident on another photodiode (not shown) and converted into a current.

  Inside the autobalance detector 20, the reference photocurrent is divided into two, and subtraction is performed between one reference photocurrent and the detection signal photocurrent. The division ratio of the reference photocurrent is feedback controlled so that the DC component of the current after subtraction becomes zero. By subtracting the reference photocurrent and the detection signal photocurrent, common mode noise (laser noise) between the two channels is canceled. Further, by controlling the division ratio of the reference photocurrent so that the DC component after subtraction becomes zero, the imbalance between the intensity ratios of the detection signal light and the reference light can be canceled. That is, the transmitted light during beam scanning fluctuates due to the refractive index of the sample 10 and the nonuniformity of stationary absorption, etc., causing an imbalance in the intensity ratio between the detection signal light and the reference light. It can be canceled by ratio feedback control.

  The output from the autobalance detector 20 is connected to a lock-in amplifier 38, and a signal carrying information resulting from photoexcitation of the sample 10 is detected. An image of the sample 10 can be constructed by attaching the sample 10 to the movable stage 35, scanning the sample position, and detecting data at each point. The optical measurement apparatus 1 includes a computer 36 and a stage controller 34 in order to control scanning. The computer 36 is connected to a stage controller 34 via a general-purpose output port of a lock-in amplifier 38 or a DA converter 37, and controls an actuator (not shown) of the movable stage 35. The computer 36 also controls the lock-in amplifier 38 to collect measurement data.

A lock-in reference signal is input to the lock-in amplifier 38. As will be described later, the lock-in reference signal of the embodiment is generated by the difference frequency component (| ω ₁ −ω ₂ |) of the modulation signal of the probe light and the modulation signal of the pump light. In the autobalance detector 20, the upper limit frequency of the feedback control circuit for the division ratio of the reference photocurrent is larger than the sample scanning speed (measurement time per pixel), and the lock-in reference signal input to the lock-in amplifier 38. It is set smaller than the frequency. As a result, the auto-balancing detector 20 simultaneously cancels the laser noise and the intensity ratio imbalance between the detection signal light and the reference light during the sample scan, and only the pump-probe signal caused by the optical excitation by the lock-in amplifier 38. Can be detected.

  FIG. 2 shows a light source system 2 used in the optical measuring device 1 of FIG. In the example of FIG. 2, a pump-probe microscope system is realized using a continuous wave laser.

  Two continuous wave semiconductor lasers 74 and 64 having different wavelengths are used as light sources for pump light and probe light. In order to perform lock-in detection, a signal generator 41 that modulates the light intensity of the semiconductor laser 64 and a signal generator 42 that modulates the light intensity of the semiconductor laser 74 are arranged.

  The modulation signal from the signal generator 41 is input to the semiconductor laser 64 through the distributor 43, the amplifier 62, and the bias tee 63. The amplifier 62, the bias tee 63 and the semiconductor laser 64 are accommodated in the shield case 61, and a laser power source 65 is connected to the bias tee 63. Similarly, the modulation signal from the signal generator 42 is input to the semiconductor laser 74 through the distributor 44, the amplifier 72, and the bias tee 73. The amplifier 72, the bias tee 73 and the semiconductor laser 74 are accommodated in the shield case 71, and a laser power source 75 is connected to the bias tee 73. By using the shield cases 61 and 71, electrical interference between the semiconductor laser 64 and the semiconductor laser 74 due to unnecessary radiation is suppressed.

  In synchronization with the modulation signal generated by the signal generator 41, the modulation signal is applied from the outside to the bias current that drives the semiconductor laser 64 for probe light. The bias current is applied from the laser power source 65 to the semiconductor laser 64 through the bias tee 63. In synchronization with the modulation signal generated by the signal generator 42, the modulation signal is applied from the outside to a bias current that drives the semiconductor laser 74 for pump light. The bias current is supplied from the laser power source 75 via the bias tee 73. With this configuration, modulation can be performed over a wide band from kilohertz (kHz) to gigahertz (GHz).

The output branched from the distributor 43 and the output branched from the distributor 44 are used for generating a lock-in reference signal of the lock-in amplifier 38. The distributor 43 connects the signal from the signal generator 41 to the frequency mixer 49 through the 30 dB attenuator 45 and the amplifier 47. The distributor 44 connects the signal from the signal generator 42 to the frequency mixer 49 through the 30 dB attenuator 46 and the amplifier 48. Attenuator 46 eliminates the effect on attenuates omega ₂ components by signal generator 42 modulated signal of the signal generator 41 (ω _1). The attenuator 45 attenuates the ω ₁ component by the signal generator 41 to remove the influence of the signal generator 42 on the modulation signal (ω ₂ ). The low frequency filter 51 extracts the difference frequency component (| ω ₁ −ω ₂ |) from the output of the frequency mixer 49 and connects it to the reference signal input port 39 of the lock-in amplifier 38.

In this way, the probe light and the pump light are modulated with different frequencies (ω ₁ , ω ₂ ), and the difference frequency component (| ω ₁ −ω ₂ |) of the change in the probe light intensity after passing through the sample 10 is locked in. To detect. By modulating the probe light together with the pump light, each modulation frequency can be changed in a state where the difference frequency is fixed, and the response of the sample molecule can be widened without using a high-speed detector or high-speed lock-in amplifier. Can be measured over time. Further, by modulating the probe light, the oscillation longitudinal mode can be changed to a multimode and noise can be reduced.

  FIG. 3 is a schematic view of another light source system 3 used in the optical measurement apparatus 1 of FIG. In the example of FIG. 3, a supercontinuum light source 81 is used. The supercontinuum light source 81 is a white pulse light source 81 that can obtain a broad wavelength of about 450 nm to 2 μm. Infrared light contained in the white pulse light source 81 is removed by the infrared cut filter 82. Further, since the polarization of the white pulse light source 81 is random, the polarization prism 83 is used to extract only light of a specific linearly polarized light. Further, using the dichroic mirror 84, light having a wavelength of 505 nm or more is transmitted and used as probe light, and light having a wavelength shorter than that is reflected and used as pump light.

  As the transmitted light of the dichroic mirror 84, only light in a specific wavelength band is selected by the interference filter 85 and emitted as probe light. The reflected light of the dichroic mirror 84 is guided to the light intensity modulator 87 by the mirror 86, and the light intensity of the pump light is modulated. The light intensity modulator 87 is controlled by a signal generator 89 via a light intensity modulation controller 88. A signal distributed from the signal generator 89 is used as a lock-in reference signal of the lock-in amplifier 38.

  In the example of FIG. 3, the white light source 81, the dichroic mirror 84, and the interference filter 85 constitute a probe light source. The white light source 81, the dichroic mirror 84, and the light intensity modulator 87 constitute a pump light source. The lock-in reference signal has a modulation frequency of pump light.

  FIG. 4 shows the laser noise canceling effect of the autobalance detector 20 incorporated in the optical measurement apparatus 1 of FIG. 4A shows the noise spectrum of the probe light when the light source system 2 including the continuous wave semiconductor lasers 64 and 74 shown in FIG. 2 is used. FIG. 4B shows the supercontinuum (white pulse shown in FIG. 3). ) A noise spectrum of probe light when the light source system 3 including the light source 81 is used.

  In FIG. 4A, the upper spectrum is the spectrum before balance, and the lower spectrum is the spectrum after balance. A semiconductor laser having an oscillation wavelength of 655 nm is used as the continuous wave semiconductor laser 64, and the intensity of the probe light incident on the signal light detection port 22 of the autobalance detector 20 is 640 μW. Laser noise is reduced by about 20 dB in the 0 to 200 kHz region.

In FIG. 4B, the upper spectrum is the spectrum before balance, and the lower spectrum is the spectrum after balance. Only the probe light having a wavelength of 640 ± 5 nm is detected using the interference filter 85 of FIG. The pulse repetition frequency of the supercontinuum light source 81 is 40 MHz. The intensity of the probe light incident on the signal light detection port 22 of the autobalance detector 20 is 640 μW. The supercontinuum light source 81 is an unstable light source with extremely large fluctuations between pulses (maximum 20% (rms)). When the autobalance detector 20 is not operated, the laser noise is about −50 dBm. On the other hand, when the autobalance detector 20 is operated, the laser noise is reduced to -85 dBm. At this time, the light intensity-to-noise ratio is 10 ⁻⁶ at the maximum. The intensity change of the probe signal by optical excitation is 10 ⁻⁷ to 10 ⁻⁶ per molecule, and the signal from a small number of molecules in a minute volume in the beam focus can be measured. The influence of transmittance fluctuation during beam scanning is also canceled by the automatic adjustment mechanism (feedback control of the division ratio of the reference photocurrent) in the autobalance detector 20, and it can be seen that only the signal resulting from photoexcitation can be detected with high sensitivity. .

  FIG. 5 is a measurement result of the sample 10 using the optical measurement apparatus 1 of the embodiment. 5A is a microscopic image when the autobalance detector 20 is operated, and FIG. 5B is a microscopic image when the autobalance detector 20 is not operated. FIG. 5C is a diagram illustrating the signal intensity along the line X-X ′ in FIGS. 5A and 5B.

  Sample 10 used was a mouse T cell labeled with a CdSe / ZnS quantum dot having a peak fluorescence wavelength at 650 nm regardless of the presence or absence of autobalance treatment. As the light source, the pump light source 11 and the probe light source 12 having the continuous wave semiconductor laser shown in FIG. 2 were used. The pump light and the probe light have oscillation wavelengths of 488 nm and 660 nm, respectively. Quantum dots are photoexcited with pump light. The wavelength of the probe light overlaps with the fluorescence spectrum of the quantum dot, and the detected pump-probe signal is mainly due to stimulated emission. The pump light and the probe light are modulated at 3 MHz and 3.12 MHz, respectively, and the frequency of the lock-in reference signal is set to two difference frequencies (120 kHz).

  The output of pump light and probe light incident on the sample 10 is 1 mW and 210 μW, respectively. The lock-in amplifier 38 has a time constant of 5 ms, a pixel size of 0.4 μm × 0.4 μm, and an image size of 20 μm × 20 μm. When the autobalance detector 20 is not operated in FIG. 5B, the output of the signal light port 22 is directly connected to the lock-in amplifier 38.

When the autobalance detector 20 is operated in FIG. 5 (A), a clear image of quantum dots labeled on the surface of the cell is obtained, but in FIG. 5 (B), the cell is visually recognized. I can't. Comparing the signal intensity of the XX ′ section in FIG. 5C, the background noise of the image is 1.5 × 10 ⁻⁶ (rms) when the autobalance function is used, whereas the balance function When not used, it is 9.8 × 10 ⁻⁶ (rms). From this, it can be seen that even when a continuous wave semiconductor laser not equipped with a special stabilization mechanism is used as a light source, a minute signal that has been buried in background noise and cannot be detected can be detected with high sensitivity.

  FIG. 6 is a pump-probe image for evaluating the spatial resolution of the microscopic image obtained by the optical measurement apparatus 1. In the pump-probe optical measurement apparatus 1 of the embodiment, the signal intensity is determined by the product of the pump light and the probe light. Due to this non-linear characteristic, when two laser beams overlap each other, a signal from a molecule at a focal center portion where the light intensity is strong is selectively detected. That is, the detection area from the sample 10 can be limited to a narrow range, and super-resolution exceeding the diffraction limit of light can be realized.

  In order to evaluate the spatial resolution of the optical measurement apparatus 1 of the embodiment, a pump-probe microscopic image of non-luminous gold nanoparticles was measured. When a laser is incident on gold nanoparticles, light is absorbed by plasmon resonance, and the decrease in the distribution of the ground state at that time changes the transmittance of the probe light. A pump-probe microscopic image was obtained using a sample in which a gold nanoparticle aqueous solution having a size of 20 nm was dropped on a glass substrate and dried as the sample 10.

  6A shows an image when the sample 10 is scanned on a plane (xy plane) perpendicular to the optical axis, and FIG. 6B shows a plane parallel to the optical axis (xz plane). It is an image (in the depth direction) when scanned. Bright points indicated by arrows in FIG. 6A are individual gold nanoparticles, and the size is 284 nm ± 32 nm in full width at half maximum. Since the beam diameter sizes calculated by the diffraction theory are 313 nm and 341 nm for the pump light and the probe light, respectively, it is confirmed that the super-resolution is realized by the pump-probe measurement of the embodiment.

  On the other hand, from FIG. 6B, the resolution in the optical axis direction (depth direction) is 1.1 ± 0.1 μm. The spot sizes in the optical axis direction calculated by the diffraction theory are 1.08 μm and 1.18 μm for the pump light and the probe light, respectively, which are almost equal to the measured values. Improved beam collimation method, especially improved uniformity of beam intensity at the back focal plane of the objective lens, further improving in-plane resolution and super-resolution in the optical axis direction (depth direction) Can be expected.

  As described above, by using an autobalance detector, even when a general-purpose semiconductor laser or a white light source is used, the laser noise and the background noise of the light source are offset, and the sample and the scanning mechanism are caused to be uneven. The imbalance between the reference light and the detection signal light is reduced. With this configuration, a minute information signal can be extracted with high sensitivity in pump-probe measurement without using a special light source or light source stabilization mechanism.

Furthermore, the response of the sample can be measured over a wide frequency range by changing the modulation frequency while keeping the difference frequency between the pump light and the probe light constant. Further, by using the difference frequency component between the pump light and the probe light as the lock-in reference signal, a low-noise signal can be detected without depending on the high-frequency lock-in detection method.
Second Embodiment
FIG. 7 is a schematic configuration diagram of the optical measurement apparatus 101 according to the second embodiment. In the second embodiment, high resolution measurement exceeding the diffraction limit is realized by using an annular beam for the pump light and the probe light. As shown in FIG. 7 (A), the optical measuring apparatus 101, the irradiation with the first light source 111 for irradiating pumping light _{L 1} to photoexcitation of the sample 130, the probe light _{L 2} which detect the change of the sample 130 by photoexcitation A second light source 112, a balance detector 123, and a lock-in amplifier 124.

In the second embodiment, a semiconductor laser with a wavelength of 488nm as a light source 111 of the pump light _{L 1,} a semiconductor laser with a wavelength of 640nm as a light source 112 of the probe light _{L 2.} The light sources 111 and 112 are both inexpensive continuous wave semiconductor lasers that output visible light.

The pump light L ₁ and the probe light L ₂ pass through the spatial filters 113 and 114, respectively, and their spatial modes are adjusted, and are superposed coaxially by the dichroic mirror 117. The dichroic mirror 117 totally reflects the pump light L ₁ and transmits the probe light L ₂ . A part of the probe light L ₂ guided to the dichroic mirror 117 is split by the beam splitter 115 and input as reference light to the reference light port 121 _b of the balance detector 123 via the multimode fiber (MMF) 122.

The pump light L ₁ and the probe L ₂ superimposed by the dichroic mirror 117 pass through the filter 120 having an annular aperture. The filter 120 functions as a beam shaper that generates an annular beam. In the second embodiment, as shown in FIG. 7B, a circular aluminum seal 202 having a diameter φ of 6 mm is attached to a circular glass plate 201 with an antireflection coating, thereby producing a filter 120 having an annular portion. The thickness of the annular portion through which pump light _{L 1} and the probe light _{L 2} is a 1.2mm as one example. Although it is desirable to insert the filter 120 for generating an annular beam at an arbitrary position on the optical path between the dichroic mirror 117 and an objective lens 128 described later, the pump light L ₁ and the probe light L ₂ are respectively connected to the circular ring. After forming into a beam, it may be superimposed on the same axis.

The annular beam generated by the filter 120 enters the objective lens 128 via the polarization beam splitter 125 and is guided to the sample 130. The pump-probe light transmitted through the sample 130 is collected by the condenser lens 129, and only the probe light L ₂ is incident on the other port 121 a of the balance detector 123 via the band pass filter 132 and the lens 133.

The balance detector 123 is preferably an autobalance detector similar to that used in the first embodiment in order to cancel laser noise and measure with high sensitivity at the shot noise limit. In the balance detector 123, the reference light (part of the probe light) input to the port 121b and the detection light after passing through the sample 130 input to the port 121a are converted into currents, respectively. The reference photocurrent is divided into two, and one of the reference photocurrent and the detected photocurrent is subtracted. The division ratio of the reference photocurrent is feedback controlled so that the DC component of the current after subtraction becomes zero. The subtraction of the reference beam current and the detection signal photocurrent, common mode noise (laser noise) between the two channels is canceled, the component of the change that occurred in the sample 130 is output by the excitation of the pump light L _1. By controlling the division ratio of the reference photocurrent so that the DC component after subtraction becomes zero, it is possible to cancel the imbalance between the intensity ratios of the detection signal light and the reference light. The transmitted light during beam scanning fluctuates due to the refractive index of the sample 130, the nonuniformity of stationary absorption, etc., and causes an unbalance of the intensity ratio of the detection light and the reference light. It can be canceled by control.

The output of the balance detector 123 is connected to the input of the lock-in amplifier 124. The intensities of the pump light L ₁ and the probe light L ₂ are modulated by ω ₁ and ω ₂ , respectively, and the difference frequency component | ω ₁ −ω ₂ | of the transmitted light that is the pump-probe signal is the lock-in amplifier 124. Is detected in a heterodyne manner. In this embodiment, ω ₁ is 1 MHz and ω ₂ is 1.015 MHz. By modulating the injection current to the laser diode used for the light sources 111 and 112, the light intensity of the light source 111 and the light intensity of the light source 112 are modulated.

The lock-in amplifier 124 detects a beat signal having a frequency | ω ₁ −ω ₂ |. This beat signal, i.e., the pump-probe signal, is generated by the interaction between the pump field and the probe field in the sample 130, and carries information resulting from photoexcitation in the sample 130. The reference signal used for lock-in detection is a signal generated from the difference frequency between the modulation frequency ω ₁ used for intensity modulation of the light source 111 and the modulation frequency ω ₂ used for intensity modulation of the light source 112.

The sample 130 is fixed to the piezo stage 131, and the pump-probe signal at each point is detected and collected while scanning the sample 130, and processing such as mapping is performed by an arithmetic device (not shown) (see the control computer 37 in FIG. 1). Thus, a microscopic image can be obtained. A CMOS camera 126 is used to confirm the sample position, and only the probe light L ₂ is incident on the CMOS camera 126 by the band-pass filter 127.

  The operation principle will be described with reference to FIGS. FIG. 8A is a schematic diagram of an annular beam, and FIG. 8B is a diagram showing an intensity distribution at a focal plane and a point spreading function (PSF) of a pump-probe signal.

  The spatial resolution of the optical microscope is determined by the spot size (diffraction limit value) at the focal plane when the light beam is incident on the objective lens, and its intensity distribution depends on the shape of the incident beam. When the annular beam is incident on the objective lens 128, the main peak of the light intensity distribution at the focal plane becomes sharper than in the case of a normal circular beam, but the side lobes appearing around the main peak also become strong. This phenomenon is as shown by the thick solid line (ring pump light) and the broken line (ring probe light) in FIG. The inventors have thought that the spatial resolution can be improved by suppressing the contribution from the side lobes.

In FIG. 8 (A), D the outer diameter of the ring in a pupil plane of the pump light L ₁ and the probe light L _2, .epsilon.d an inner diameter, the focal length of the lens and f. At this time, the intensity distribution in the radial direction r at the focal plane is given by equations (1) and (2) from the scalar diffraction theory.

Here, I _c (r) is an intensity distribution at the focal point when a circular beam (ε = 0) is incident, and is expressed by Expression (3) using the Bessel function J ₁ (x).

NA in the formula (2) is the numerical aperture of the lens, and is given by NA = nD / f where n is the refractive index in the sample 30. λ is the wavelength of incident light.

On the other hand, the PSF of the pump-probe signal is the product of the intensity distribution of the pump light and the probe light. In FIG. 8B, the wavelengths of the pump light L ₁ and the probe light L ₂ are set to 488 nm and 660 nm and NA = 0.95, respectively, for the circular beam (ε = 0) and the circular beam (ε = 0.8), the intensity distribution in the focal plane and the PSF are simulated.

  When attention is paid to individual circular beams, the main peak at the center is sharper than in the case of a circular beam, but the side lobes are also increased at the same time. The intensity ratio SLR (side lobe ratio) is about 15%, and this side lobe adversely affects image resolution in a normal fluorescent microscope.

  On the other hand, since the PSF of the pump-probe signal (thick one-dot chain line in FIG. 8B) is the product of the intensity of the two beams, the strong part is dominant and the SLR is reduced to 1%. And the main peak becomes sharper. The full width at half maximum is 164 nm, which is 1 / 2.1 times that of the diffraction limit value 0.51λ / NA = 344 nm of the circular probe beam. In addition, in the case of a circular beam, the full width at half maximum of the pump-probe PSF (thin dotted line in FIG. 8B) is 212 nm, which is 1 / 1.3 times, and the resolution is improved.

The resolution when the wavelengths of the annular pump light L ₁ and the annular probe light L ₂ are 500 nm and 640 nm, respectively, is improved to 110 nm, and when the wavelengths are 400 nm and 520 nm, the resolution is improved to 90 nm. The optical measurement method of the embodiment can select an optimum wavelength according to an observation target, and has a high degree of freedom in measurement.

FIG. 9 is a schematic diagram for explaining the effects when two-color pump light and probe light are used. As shown in FIG. 9A, in the pump-probe method, the observed signal is proportional to the product of the intensities of the pump light L ₁ and the probe light L _2, and exists in the central portion of the main peak having a high intensity. A signal is detected preferentially from the molecule. Since the position of the side lobe peak depends on the wavelength, side lobes appear at different positions by using two colors having different wavelengths for the pump light L ₁ and the probe light L ₂ . When the wavelengths of the pump light L ₁ and the probe light L ₂ are different (for example, the wavelength of the pump light L ₁ is 488 nm and the wavelength of the probe light L ₂ is 660 nm), the side lobe when the two annular beams are overlapped Can be reduced to about half compared to the case of a single wavelength.

  On the other hand, in the conventional two-photon excitation measurement method of FIG. 9B, the pump light that excites two-photons is irradiated and multiplied by the sidelobe peak that appears at the same position, so the SLR is 2.3%. Become. In the first place, the two-photon excitation method is fluorescence detection by irradiation with an ultrashort pulse (femtosecond to picosecond) with a near infrared wavelength, and therefore the configuration and measurement principle are different from those of the present invention. The two-photon excitation method requires a light source requiring maintenance by an expensive and highly technical engineer such as a titanium sapphire laser, and the resolution is lower than that of a visible light source, making it difficult to put into practical use and commercialization.

  In the second embodiment, the effect of the side lobe on the focal plane is reduced to 1/10 or less with a low-cost and simple configuration by observing the pump-probe signal using an annular beam in a visible region having different wavelengths. Suppress and improve resolution about 2 times.

  10 and 11 show a measurement example 1 of the second embodiment. In Measurement Example 1, gold nanoparticles in the sample are observed with the optical measurement device 101 in FIG. FIG. 10A is a pump-probe microscopic image of gold nanoparticles dispersed on a glass substrate, and FIG. 10B is a microscopic image of a single gold nanoparticle when an annular pump-probe beam is incident. FIG. 10C is a microscopic image of a single gold nanoparticle when a circular pump-probe beam is incident.

  As a sample, a material in which gold nanoparticles are three-dimensionally dispersed in 20 μm thick polyvinyl alcohol (PVA) is prepared. The gold nanoparticle located near the plane scanned with the annular pump-probe light has a stronger pump-probe signal intensity.

Gold nanoparticles have a size of 20 nm and absorb light by plasmon resonance at the wavelength of the pump light L ₁ (488 nm). At this time, very small changes in refractive index due to temperature rise of the gold nanoparticles vicinity, the probe light L ₂ by refraction scattering causes a change in light transmission, the pump - which contributes to the probe signal.

  The intensity of the pump light incident on the sample is 0.7 mW, and the intensity of the probe light is 0.3 mW. The lock-in amplifier has a time constant of 0.5 ms, a dwell time per pixel of 1 ms, and an image resolution of 300 × 300 pixels.

  When the pump-probe microscopic image of one gold nanoparticle with an annular beam in FIG. 10B is compared with the pump-probe microscopic image of one gold nanoparticle with a circular beam in FIG. By detecting lock-in after detection, the gold nanoparticles can be clearly observed in FIG. 10C, but by using an annular beam in FIG. 10B, the outline of the gold nanoparticles becomes clearer. It has become. This is because the main peak becomes sharper due to the superposition of the annular beams having different wavelengths, while the side lobes are canceled out, and both the resolution and the contrast are improved.

  11A shows the cross-sectional strength in the horizontal direction (x direction) of the microscopic images of FIGS. 10B and 10C, and FIG. 11B shows the cross-sectional strength in FIGS. 10B and 10C. ) Shows the cross-sectional intensity in the vertical direction (y direction or polarization direction of incident light) of the microscopic image. A black circle is a cross-sectional intensity when an annular beam is used (FIG. 10B), and a triangular mark is a cross-sectional intensity when a circular beam is used (FIG. 10C). A solid line connecting the data points indicates a Gaussian fit.

  Since the convergent beam has a numerical aperture close to 1, the resolution in the x direction is about 5% higher than that in the y direction (the polarization direction of incident light). From the fitting using the Gaussian function, in FIG. 11A, the full width at half maximum when the annular pump-probe beam is used is 162 ± 4 nm, and the full width at half maximum when the circular pump-probe beam is used is 232 ±. 3 nm. In FIG. 11B, the full width at half maximum when an annular pump-probe beam is used is 196 ± 6 nm, and the full width at half maximum when a circular pump-probe beam is used is 245 ± 7 nm. According to this measurement result, a sufficiently high resolution can be obtained even with a circular pump-probe beam by lock-in detection after balanced homodyne detection, but the resolution is further 20-30% higher than that of a circular beam by using an annular beam. improves. This value almost coincides with the above theoretical calculation.

  Further, since the diffraction limit value of the probe light (full width at half maximum of the spot size at the focal plane) is 344 nm, an annular pump-probe beam is used as in the second embodiment, so that an ordinary scanning optical microscope can be used. The resolution is improved about twice.

  12 and 13 show a measurement example 2 of the second embodiment. In Measurement Example 2, fluorescent beads in the sample are observed with the optical measurement apparatus 101 in FIG. FIG. 12A shows a pump-probe microscopic image of fluorescent beads dispersed on a glass substrate when an annular beam is incident, and FIG. 12B shows a circular shape in the same region on the glass substrate as a comparative example. It is a pump-probe microscopic image of a fluorescent bead when a beam is incident.

  The size of the fluorescent beads is 1 μm, and it is labeled with Nile red having a fluorescent spectrum peak at 600 nm. The excitation spectrum of Nile Red is the left curve in FIG. 13A, and the fluorescence spectrum is the right curve in FIG. As shown in FIG. 13A, the wavelength of the probe light overlaps the long wavelength side of the fluorescence spectrum. In this case, the change in the intensity of the probe light due to the stimulated emission gain mainly contributes to the pump-probe signal.

  FIG. 13B is a diagram showing the cross-sectional strength along the line I-I ′ in FIGS. 12A and 12B. Black circles indicate the pump-probe signal intensity when the circular beam of FIG. 12A is used, and triangular marks indicate the pump-probe signal intensity when the circular beam of FIG. 12B is used. By detecting lock-in after balanced homodyne detection, a sufficiently sharp spectrum can be obtained even with a circular beam, but by using an annular beam, the rise is further sharpened. It can be seen that when the annular beam is used, the resolution is further improved as compared with the case where the circular beam is used.

  From the results of Measurement Examples 1 and 2, it can be seen that the optical measurement method of the second embodiment can be applied to non-fluorescent materials such as gold nanoparticles and fluorescent materials. Since a minute state change can be measured in a non-contact manner using laser light in the visible region, it can also be applied to in vivo observation and measurement.

  In the configuration and method of the second embodiment, two colors of laser light are incident on the sample, and the pump-probe signal from the sample is measured to suppress the side lobe near the focal point of the convergent ring beam, thereby reducing the resolution. Can be improved. By performing imaging measurement while scanning the beam position, the spatial resolution can be improved by a factor of two compared to a normal optical microscope.

Similar to the first embodiment, low-noise and high-sensitivity measurement can be performed by lock-in detection after balanced homodyne detection. In addition, since light of two colors having a shorter wavelength than the near-infrared region is used through the first and second embodiments, a continuous-wave semiconductor laser that is low-cost, space-saving and easy to maintain can be used as a light source. The second embodiment further has the following advantages.
(1) Artifacts can be reduced because complex image calculation processing for obtaining a super-super resolution image is not required.
(2) A special fluorescent dye is not required, and even weak fluorescent and non-fluorescent substances such as metal nanoparticles that are difficult to measure by a known super-resolution method can be measured.
(3) Super-resolution measurement of dynamics of photoexcited molecules (electron energy relaxation, recombination of photoexcited carriers, thermal diffusion, etc.) is possible.

  This application has priority based on Japanese Patent Application No. 2013-17895 filed on August 30, 2013 and Japanese Patent Application No. 2014-128920 filed on June 24, 2014. It is claimed and incorporated herein by reference in its entirety.

DESCRIPTION OF SYMBOLS 1,101 Optical measuring apparatus 2, 3 Light source system 10, 130 Sample 11, 111 Pump light source (1st light source)
12, 112 Probe light source (second light source)
20, 123 Balance detector 21, 121b Reference light detection port 22, 121a Signal light detection port 38, 124 Lock-in amplifier 120 Filter (ring shaper) having an annular aperture

Claims (10)

A first light source that emits pump light for exciting the sample;
A second light source for irradiating probe light for measuring a change in the excited sample;
A beam splitter that divides the probe light into detection probe light and reference probe light;
A first input port for receiving the detection probe light transmitted through the excited sample; a second input port for receiving the reference probe light not transmitted through the excited sample; the detection probe light and the reference probe; A balance detector having an output terminal for outputting an electric signal representing a difference in light;
A lock-in amplifier for detecting a desired signal at a predetermined lock-in frequency from the output of the balance detector;
An optical measuring device. The probe light is modulated at a first modulation frequency;
The pump light is modulated at a second modulation frequency;
The lock-in amplifier receives the reference signal of the difference frequency between the first modulation frequency and the second modulation frequency and the output of the balance detector as inputs, and locks in the signal at the difference frequency. The optical measuring device according to claim 1. The second light source selects light of a predetermined wavelength from a white light source and irradiates it as the probe light,
The first light source selects light other than the predetermined wavelength from the white light source and irradiates it as the pump light,
The lock-in amplifier receives a reference signal of a modulation frequency of a light intensity modulation signal that modulates the intensity of the pump light and an output of the balance detector, and detects the lock-in at the modulation frequency. The optical measuring device according to claim 1, wherein The balance detector is obtained by dividing a reference photocurrent obtained by photoelectrically converting the reference probe light by a predetermined division ratio, and photoelectrically converting the divided one current component and the detection probe light. Convert the difference from the detected photocurrent into a voltage signal and output it,
The optical measurement apparatus according to claim 1, wherein the predetermined division ratio is feedback controlled so that a direct current component included in the voltage signal is minimized.   The upper limit frequency of the feedback control is larger than a speed at which the sample is scanned with the pump light and the detection probe light, and smaller than the lock-in frequency of the lock-in amplifier. Optical measuring device. A beam shaper for shaping the pump light and the probe light into an annular beam;
A first optical system for causing the pump light and the probe light formed into the annular beam to enter the sample;
The optical measuring device according to claim 1, further comprising: A second optical system for coaxially superimposing the pump light and the probe light;
Further comprising
The optical measurement apparatus according to claim 6, wherein the beam shaper is inserted between the first optical system and the second optical system.   The optical measurement apparatus according to claim 1, wherein the first light source and the second light source are semiconductor lasers that emit laser light in a visible region having a shorter wavelength than the near infrared region. The optical measuring device according to any one of claims 1 to 8,
A scanning mechanism for scanning the sample relative to the pump light and the probe light;
An arithmetic processing unit that collects and processes the signals detected at each scanning point on the sample by the lock-in amplifier;
A microscopic imaging system characterized by comprising: Irradiate the sample with pump light to excite the sample,
Dividing probe light into detection probe light and reference probe light, irradiating the sample simultaneously with the irradiation of the pump light with the detection probe light,
The detection probe light transmitted through the sample is received by the first input port of the balance detector,
Receiving the reference probe light that does not pass through the sample at a second input port of the balance detector;
From the balance detector, an electrical signal representing a difference between the detection probe light and the reference probe light is output,
Detecting a desired signal at a predetermined lock-in frequency from the output of the balance detector;
An optical measurement method comprising steps.