Data processing method, data processing apparatus, and data processing program

Abstract

To quickly know the success or failure of measurement in quantum sensing using an NV center.SOLUTION: A data processing method (4) according to the present invention developed for solving the previously described problem includes measurement data preparation processing (42) for preparing measurement data of fluorescence emitted by an NV center, the measurement data being acquired by irradiating a diamond having the NV center with an electromagnetic wave based on a predetermined pulse sequence, likelihood function preparation processing (51) for preparing a likelihood function of a probability distribution having, as a parameter, a fluorometry model for the NV center, the fluorometry model being defined based on a measurement target model including a measurement target physical quantity as a parameter and the predetermined pulse sequence, and likelihood calculation processing (52) for calculating a likelihood of the parameter based on the measurement data and the likelihood function.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to quantum sensing, and more particularly to a method, an apparatus, and a program for processing data obtained by quantum sensing. [Background technology]

[0002] One of the lattice defects in the crystal structure of diamond is the NV center, a complex defect in which a nitrogen atom and an atomic vacancy are adjacent to each other. The NV center can capture electrons. When an NV center captures one electron, it has three electrons provided by the three carbon atoms adjacent to the atomic vacancy, two electrons provided by the nitrogen atom, and the captured electron. The electron's spin (spin quantum number S = 1, spin magnetic quantum number M = 0, ±1) is stable. In quantum sensing, two quantum states of the spin magnetic quantum numbers are generally used. For example, when considering a two-level system with M = 0 and M = -1, the quantum state |ψ〉 of the electron spin of the NV center can be expressed as follows, where |0〉 is the quantum state with M = 0 and |1〉 is the quantum state with M = -1, and θ and φ are real numbers (the same applies to a two-level system with M = 0 and M = 1):

number

[0003] When the electron spin of an NV center is irradiated with microwaves (referred to herein as "resonant microwaves") having a frequency close to the frequency (resonant frequency) corresponding to the energy difference between the quantum states of Ms=0 and Ms=-1, the quantum state of the electron spin changes depending on the time for which the resonant microwaves are irradiated. In other words, by irradiating the electron spin of the NV center with resonant microwaves for a predetermined time, the quantum state of the electron spin of the NV center can be changed from Ms=0 to Ms=-1, or can be changed to a superposition state of the quantum state of Ms=0 and the quantum state of Ms=-1 (intuitively, for example, if Ms=0 is an upward-pointing electron spin and Ms=-1 is a downward-pointing electron spin, the electron spin is in a state where it is lying on its side). In the case of the electron spin of an NV center, the coherence time (the time during which the superposition state is maintained) is relatively long, and the superposition state is maintained for approximately several milliseconds at room temperature.

[0004] Furthermore, when the electron spin of an NV center is excited with a laser beam of a certain wavelength (e.g., 532 nm) (hereinafter referred to as the "initialization / readout laser beam" for the purpose described later), the electron spin of the NV center transitions to the ground state (direct transition) while emitting fluorescence (at a wavelength of approximately 637 nm). While the quantum state of the electron spin remains unchanged in a direct transition, an electron spin with Ms = -1 stochastically transitions to the ground state with Ms = 0 through a separate process (nonradiative transition) without emitting light. Therefore, the intensity of the fluorescence emitted when the electron spin of an NV center is excited with the initialization / readout laser beam varies depending on the quantum state of the electron spin. (For example, when the electron spin of an NV center is irradiated with resonant microwaves to bring it to the quantum state of Ms = -1 and then repeatedly irradiated with the initialization / readout laser beam, the total fluorescence intensity is approximately 30% smaller than the total fluorescence intensity obtained by repeatedly irradiating the initialization / readout laser beam in the quantum state of Ms = 0.) Furthermore, as mentioned above, when an electron spin of Ms=-1 is excited by the initialization / readout laser beam, a non-radiative transition occurs with a high probability. Therefore, the quantum state of the electron spin of the NV center after excitation by the initialization / readout laser beam becomes Ms=0, regardless of the quantum state before excitation by the initialization / readout laser beam.

[0005] These properties can be utilized to utilize NV centers as sensors for quantum sensing. Quantum sensing is a sensing technology that utilizes quantum mechanical phenomena to measure local values ​​of physical quantities or obtain information about trace amounts of a target object. For example, quantum sensing using NV centers to measure AC magnetic fields involves irradiating the NV center with an initialization / readout laser beam to initialize the quantum state of the electron spin to Ms = 0. Then, irradiating the NV center with resonant microwaves for a predetermined time manipulates the quantum state of the electron spin, placing it in a superposition state of Ms = 0 and Ms = -1. In this state, the magnetic field around the NV center is obtained by measuring the phase φ. The NV center is then irradiated with resonant microwaves again to manipulate the quantum state, and irradiated with an initialization / readout laser beam to read out the phase φ as a difference in fluorescence intensity. This allows for the measurement of local values ​​of the magnetic field or the acquisition of information about trace amounts of a target object (e.g., atomic type, distance, spin, etc.). As mentioned above, the coherence time of the NV center is relatively long, on the order of several milliseconds at room temperature, so that phase changes can be measured with high sensitivity even in a weak magnetic field.

[0006] Quantum sensing (e.g., the measurement of magnetic fields through phase as mentioned above) is performed based on a predetermined pulse sequence (a time series of the number of times, intervals, duration, intensity, and phase of the initialization / readout laser light or resonant microwave irradiations). Creating an appropriate pulse sequence can improve the sensitivity of the sensor and the frequency selectivity for the AC component of the measurement target. For example, a pulse sequence that irradiates a π / 2 pulse of resonant microwaves followed by N π pulses (where N is a natural number) at a fixed time interval τ functions as a narrow bandpass filter with a center frequency of 1 / (2τ) and a bandwidth of 1 / (Nτ), enabling the acquisition of spectral data with a linewidth of several hundred kHz.

[0007] In quantum sensing, which uses NV centers to measure AC magnetic fields, one technique with superior frequency resolution is the quantum heterodyne method (Qdyne method) (e.g., Non-Patent Document 1). In the Qdyne method, a predetermined pulse sequence is performed, and the measured fluorescence intensity is associated with a clock signal to create time-series data. The time-series data is then converted into spectral data, which allows the frequency of the AC magnetic field to be calculated with high precision. To obtain the spectral data, for example, a fast Fourier transform (FFT) is used. The frequency resolution of the Qdyne method depends only on the stability of the clock signal, and spectral data with a linewidth of approximately several hundred μHz can be obtained. [Prior art documents] [Non-patent literature]

[0008] [Non-Patent Document 1] Simon Schmitt et al., “Submillihertz magnetic spectroscopy performed with a nanoscale quantum sensor”. Science vol.356, pp832-837. May 26, 2017 Summary of the Invention [Problem to be solved by the invention]

[0009] Quantum states are inherently observed with statistical fluctuations (probability). Therefore, quantum sensing involves repeating the same measurement many times (e.g., 10 5 The measurements must be taken over an extended period of time (typically from a few hours to a few months).

[0010] In such long-term measurements, even if measurements are started under appropriate measurement conditions (measurement environment such as temperature and humidity, and the state of the measurement device), the measurement conditions may change during the measurement. However, in quantum sensing, the same measurement is repeated many times to measure the quantum state, so measurements may continue under inappropriate measurement conditions.

[0011] In quantum sensing for acquiring spectral data as described above, in addition to the reasons mentioned above, in order to more reliably identify the frequency components of the measurement target, a pulse sequence is set to average out noise (e.g., shot noise of the measuring instrument) contained in the time-series data to increase the S / N ratio, and measurements are carried out over a long period of time. Analysis using the spectral data is carried out after such long-term measurements have been completed.

[0012] As a result, in quantum sensing that acquires spectral data, it is sometimes impossible to notice measurement failures (for example, measurement conditions that are outside the allowable range, failure to measure signals of the desired frequency, failure to obtain information about the object being measured, etc.) during the measurement, and measurements that actually failed can continue for long periods of time. In addition, because the process of converting time-series data into spectral data using FFT takes a considerable amount of time, it is not possible to quickly confirm the success or failure of the measurement. These issues significantly reduce the efficiency of quantum sensing research and development using the NV Center.

[0013] The problem to be solved by the present invention is to enable quick detection of the success or failure of a measurement in quantum sensing using an NV center. [Means for solving the problem]

[0014] In order to solve the above problems, the data processing method according to the present invention comprises: a measurement data preparation process for preparing measurement data of fluorescence emitted from a diamond having an NV center, the measurement data being obtained by irradiating the diamond with electromagnetic waves based on a predetermined pulse sequence; a likelihood function preparation process for preparing a likelihood function of a probability distribution having as parameters a measurement object model including the physical quantity of the measurement object as a parameter and a fluorescence measurement model of the NV center defined based on the predetermined pulse sequence; a likelihood calculation process for calculating the likelihood of the parameter based on the measurement data and the likelihood function; Includes.

[0015] In addition, a data processing device according to the present invention, which is made to solve the above problems, comprises: A memory unit; an input receiving unit that receives inputs of measurement data of fluorescence emitted from the NV center, the measurement data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a measurement object model that includes the measurement object physical quantity as a parameter and a fluorescence measurement model of the NV center that is defined based on the predetermined pulse sequence; a likelihood calculation unit that calculates the likelihood of the parameter based on the measurement data and the likelihood function; Equipped with.

[0016] In order to solve the above problems, a data processing program according to the present invention is provided that executes the following steps: A memory unit; an input receiving unit that receives inputs of measurement data of fluorescence emitted from the NV center, the measurement data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a measurement object model that includes the measurement object physical quantity as a parameter and a fluorescence measurement model of the NV center that is defined based on the predetermined pulse sequence; The likelihood calculation unit operates to calculate the likelihood of the parameter based on the measurement data and the likelihood function. [Effects of the Invention]

[0017] In the present invention, the likelihood of a parameter is calculated based on a likelihood function of a probability distribution whose parameters are the measurement data of fluorescence emitted by the NV center, obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, and a fluorescence measurement model of the NV center, which is defined based on a measurement object model that includes the physical quantity to be measured as a parameter and the predetermined pulse sequence. A person implementing the present invention can refer to the calculated likelihood and determine that the measurement is successful if a likelihood peak is present (e.g., in the measurement of an AC magnetic field, a magnetic field with a frequency around the peak is measured). Conversely, if an anomaly is found in the likelihood (e.g., in the measurement of an AC magnetic field, a downward convex peak is obtained or no peak is obtained), the measurement is determined to be unsuccessful. According to the present invention, the presence or absence of a likelihood peak or anomaly can be confirmed even when the number of measurement data is small during measurement, and the processing time is short. Therefore, a person implementing the present invention can know the success or failure of the measurement even during the measurement. Furthermore, a person implementing the present invention can know the success or failure of the measurement in a shorter time.

[0018] Therefore, according to the present invention, in quantum sensing using NV centers, it is possible to quickly know whether a measurement has been successful or not. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a diagram showing an outline of a configuration example of a quantum sensing system according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram showing an example of an input screen (tab) displayed to allow a user to input a physical quantity to be measured and a measurement method. [Figure 3] FIG. 1 shows an example of displaying data obtained by CW ODMR. [Figure 4] FIG. 10 is a diagram showing an example of displaying data obtained by Rabi vibration measurement. [Figure 5]This is a diagram showing the pulse sequence used to measure the AC magnetic field generated by nuclear spins around the electron spin of an NV center using Hahn echo measurements. [Figure 6] FIG. 1 is a diagram showing a schematic diagram of how the quantum state of the electron spin of an NV center is manipulated by a pulse sequence related to Hahn echo measurement. [Figure 7] FIG. 1 is a diagram showing the correspondence between pulse sequences and clocks in the Qdyne method. [Figure 8] FIG. 10 is a diagram showing an example of a screen (tab) that displays the stochastic process and likelihood function used in the likelihood function setting unit, while allowing the user to change the probability distribution and likelihood function as needed. [Figure 9] FIG. 10 is a diagram showing an example of an input screen (tab) displayed for allowing a user to make settings for calculating likelihoods and to give instructions for executing likelihood calculation processing. [Figure 10] This figure shows likelihood curves and FFT spectra obtained by processing measurement data up to that point, for each elapsed time since the start of measurement. [Figure 11] FIG. 10 is a diagram showing the results of calculating likelihood for noise data. [Figure 12] FIG. 10 is a diagram illustrating an outline of the configuration of a data processing system according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, a quantum sensing system including an embodiment of a data processing method, a data processing device, and a data processing program according to the present invention will be described with reference to the drawings. Note that this embodiment describes an exemplary embodiment of the present invention, and does not represent the only embodiment in which the present invention can be practiced.

[0021] <Example of quantum sensing system configuration> An example of the configuration of a quantum sensing system 1 according to this embodiment will be described. Fig. 1 is a diagram showing an outline of an example of the configuration of a quantum sensing system 1 according to this embodiment. As shown in Fig. 1, the quantum sensing system 1 includes a measurement unit 2, a control and processing unit 4, and a user interface 6.

[0022] The measurement unit 2 comprises a diamond element 20 having an NV center, a static magnetic field source 21, a pulse generator 22, a microwave (MW) source 23, a microwave waveguide 24, a microwave switch 25, a laser source 31, an acousto-optic modulator (AOM) 32, a dichroic mirror 33, an objective lens 34, a fluorescence filter 35, an imaging lens 36, a pinhole 37, and a single-photon detector 38.

[0023] The diamond element 20 has an NV center inside the crystal. As mentioned above, an NV center is a complex defect in the diamond crystal structure in which a nitrogen atom and an atomic vacancy are adjacent to each other, and the electron spin exists stably when one electron is captured. When the NV center in this state is irradiated with, for example, a 532 nm wavelength laser light (initialization / readout laser light), it emits fluorescence (wavelength around 637 nm) with an intensity corresponding to the quantum state of the electron spin. The quantum state of the NV center's electron spin is affected by magnetic fields generated by the surrounding environment (e.g., minute leakage currents in electronic circuits) and by magnetic fields generated by trace amounts of the object to be measured (e.g., nuclear spin or electron spin of biomolecules, etc.). Therefore, by measuring the quantum state of the NV center's electron spin as a difference in fluorescence intensity, it is possible to measure the local magnetic field around the NV center and obtain information about the surrounding environment of the NV center and trace amounts of the object to be measured (e.g., the presence or absence of defects in the electronic circuit, the type and distance of atoms, spin, etc.).

[0024] The shape of the diamond element 20 is not particularly limited, but it can be, for example, sheet-like or plate-like. If the diamond element 20 is placed on a movable stage (not shown), the position can be easily adjusted during measurement. Furthermore, only one NV center may be formed in the diamond element 20, or two or more NV centers may be formed.

[0025] The static magnetic field source 21 applies a static magnetic field to the NV center included in the diamond element 20. When a static magnetic field is applied to the NV center, the degeneracy between the quantum state of Ms=-1 and the quantum state of Ms=1 of the electron spin of the NV center is resolved (Zeeman splitting), so that the transition between Ms=0 and Ms=-1 (or Ms=1) can be easily selected, and the quantum state of the electron spin of the NV center can be treated as a two-level system. In this specification, for convenience of explanation, a two-level system of Ms=0 and Ms=-1 is considered as the quantum state of the electron spin of the NV center. As the static magnetic field source 21, for example, a known magnet such as a neodymium magnet can be used. The static magnetic field source 21 preferably incorporates a mechanism that can adjust the direction and strength of the magnetic field so that a static magnetic field of appropriate strength can be applied parallel to the quantization axis of the NV center.

[0026] The pulse generator 22 generates a pulse signal based on parameters such as a preset pulse generation interval time, a delay time from a pulse start signal, and a pulse duration. A known pulse generator can be used as the pulse generator 22. In this embodiment, as shown in FIG. 1 , the pulse generator 22 sends an on / off electrical signal to the microwave switch 25 and the AOM 32 based on the parameters, thereby realizing a pulse sequence consisting of laser light and microwaves.

[0027] The microwave source 23 generates microwaves of a preset frequency. A known microwave generator can be used as the microwave source 23. The generated microwaves are irradiated to the diamond element 20 by a known method. For example, in this embodiment, the microwaves generated by the microwave source 23 are transmitted through the microwave waveguide 24, pulsed by the microwave switch 25, and then passed through a copper wire (not shown) installed on the diamond element 20, thereby irradiating the diamond element 20 with pulsed microwaves.

[0028] The laser source 31 generates laser light of a preset wavelength. The generated laser light is pulsed by the AOM 32 and irradiated onto the diamond element 20. In this embodiment, the pulsed laser light is reflected by the dichroic mirror 33 towards the objective lens 34, and then enters the objective lens 34 and is focused, and is irradiated onto a focal point on the diamond element 20. The laser source 31 generates and irradiates, for example, initialization / readout laser light.

[0029] The fluorescence emitted from the NV center when irradiated with the initialization / readout laser light passes through the objective lens 34, dichroic mirror 33, fluorescence filter 35, imaging lens 36, and pinhole 37 in this order, and is detected by the single-photon detector 38.

[0030] The laser source 31, AOM 32, dichroic mirror 33, objective lens 34, fluorescence filter 35, imaging lens 36, pinhole 37, and single-photon detector 38 can be realized, for example, by using a known confocal microscope.

[0031] The control and processing unit 4 includes, as functional blocks, a control unit 41, an input receiving unit 42, a memory unit 43, a model setting unit 44, a resonance frequency identifying unit 45, a pulse width identifying unit 46, a pulse counting unit 47, a display processing unit 48, a spectrum data creating unit 49, a likelihood function setting unit 51, a likelihood calculation unit 52, and a measurement success / failure determining unit 53. The control and processing unit 4 is realized, for example, as software executed on a personal computer (PC) and an electronic circuit (for example, an FPGA) that operates in cooperation with the PC. The function of each unit will be described together with the operation of the quantum sensing system 1.

[0032] The user interface 6 includes an input unit 61 and a display unit 62. The user interface 6 can be, for example, a keyboard, a mouse, a screen, etc. of a PC on which software that functions as the control and processing unit 4 is installed.

[0033] <Operation of quantum sensing system> The operation of the quantum sensing system 1 according to this embodiment will now be described. When the quantum sensing system 1 is started, the control unit 41 first prompts the user to input the physical quantity to be measured (measurement target physical quantity) and the measurement method for measuring the measurement target physical quantity through the input unit 61, for example, by displaying a predetermined input screen on the display unit 62. FIG. 2 is a diagram showing an example of an input screen (tab) displayed on the display unit 62 to prompt the user to input the measurement target physical quantity and the measurement method. The control unit 41 specifies the measurement target physical quantity and the measurement method by prompting the user to input the measurement target physical quantity and the measurement method from pull-down menus on the input screen, for example, as shown in FIG. 2. Options displayed in the pull-down menu can be stored in advance in the storage unit 43, and can be added or deleted as appropriate. When the user inputs the measurement target physical quantity and the measurement method, the input accepting unit 42 accepts the input and outputs it to the model setting unit 44.

[0034] In this embodiment, for example, the frequency of an AC magnetic field and the Qdyne method can be input as a combination of a physical quantity to be measured and a measurement method. Note that the combinations of physical quantities to be measured and measurement methods that are the subject of the data processing method according to the present invention are not limited to this, and the combinations shown in Table 1 can also be used, for example. The measurement methods will be described later. [Table 1]

[0035] The control unit 41 allows the user to input a mathematical expression or numerical model (measurement object model) that includes the input measurement object physical quantity as a variable (parameter), and a pulse sequence (a time series representation of the number of times, intervals, irradiation time, intensity, and phase of irradiation of a predetermined electromagnetic wave, such as an initialization / readout laser beam or a resonant microwave, etc.) to be executed by the input measurement method. When the user inputs a measurement object model or pulse sequence, the input accepting unit 42 accepts these inputs and outputs them to the model setting unit 44. When the user inputs an instruction not to input a measurement object model or pulse sequence (for example, by checking the box for "Set measurement object model as default" as shown in FIG. 2), the input accepting unit 42 accepts the input and outputs it to the model setting unit 44. In response to the input, the model setting unit 44 reads out a measurement object model that corresponds to the input measurement object physical quantity from among the measurement object models pre-stored in the storage unit 43 as a default measurement object model. The same applies to pulse sequences.

[0036] In this embodiment, the measurement object model that includes the frequency ν of the AC magnetic field as a parameter can be, for example, an AC magnetic field model S expressed by the following equation: Here, A is the amplitude of the AC magnetic field model, and φ0 is the initial phase of the AC magnetic field model S, which are input as predetermined values ​​via the input unit 61 or read out from the storage unit 43. Note that the measurement object model is not limited to this, and a linear combination of an orthogonal function system such as a trigonometric function (Fourier series) can also be used.

number

[0037] In this embodiment, the pulse sequence executed by the Qdyne method can be, for example, a pulse sequence for XY8-k measurement, and a filter function F corresponding to the pulse sequence, which is expressed by the following equation, can be used: xy8k ↑ (in the main text of this specification, a vector is expressed in this way) can be input through the input unit 61. Here, F xy8k ↑ represents a square wave, and dt xy8kis the time step of the pulse sequence for xy8-k measurement, τ represents the period of the Rabi oscillation of the NV center, and k represents the number of repetitions of the XY8 sequence (a pulse sequence in which π pulses are irradiated eight times to rotate the NV center by π [rad] around the x-axis or y-axis). Note that there are no limitations on the pulse sequence or filter function, and various ones can be used depending on the measurement method to be implemented.

number

[0038] The model setting unit 44 sets a model (fluorescence measurement model) that represents the measured intensity of fluorescence emitted by the NV center, based on the measurement object model and pulse sequence input via the input unit 61 or read as default from the storage unit 43. In this embodiment, the fluorescence measurement model can be set, for example, by the following procedure, based on the filter function corresponding to the AC magnetic field model and pulse sequence described above.

[0039] First, the model setting unit 44 can set a magnetic resonance signal model B expressed by the following equation, for example, based on the AC magnetic field model and the filter function. Note that integration can be performed numerically.

number

[0040] Subsequently, the model setting unit 44 can set a quantum state model Q of the electron spin of the NV center based on the magnetic resonance signal model B. The quantum state model Q of the electron spin of the NV center can be expressed by, for example, the following equation. Here, γ NV is the gyromagnetic ratio of the NV center, and can be input by the user via the input unit 61 or can be a value stored in advance in the storage unit 43.

number

[0041] Thereafter, the model setting unit 44 can set, for example, the rate (arrival rate) λ at which photons emitted from the NV center are detected as the fluorescence measurement model. The arrival rate λ can be expressed, for example, by the following equation. Here, λ BG is the background photon arrival rate, and P max and P min are the average numbers of photons detected when the quantum state of the electron spin of the NV center is Ms = 0 and Ms = -1, respectively. For these, too, the user can input values ​​through the input unit 61, or values ​​stored in advance in the storage unit 43 can be used.

number

[0042] The model setting unit 44 can also display a predetermined input screen (not shown) on the display unit 62 to allow the user to input the magnetic resonance signal model B, the quantum state model Q of the electron spin of the NV center, and the arrival rate λ through the input unit 61. Furthermore, the fluorescence measurement model is not limited to the one described above, and can be set appropriately based on the physical quantity to be measured and the pulse sequence.

[0043] The model setting unit 44 outputs the created fluorescence measurement model to the likelihood function setting unit 51. The operation of the likelihood function setting unit 51 will be described later.

[0044] After the physical quantity to be measured and the measurement method have been identified, in parallel with or before or after the operation of the model setting unit 44 described above, the control unit 41 causes the resonant frequency specifying unit 45 to specify the resonant frequency of the electron spin of the NV center of the diamond element 20, and causes the pulse width specifying unit 46 to specify the microwave irradiation time required to set the quantum state of the electron spin of the NV center to a specific state. The resonant frequency specifying unit 45 specifies the resonant frequency of the electron spin of the NV center of the diamond element 20 by reading out a value stored in advance in the memory unit 43, having the user input it through the input unit 61, or having the measurement unit 2 perform, for example, continuous wave optically detected magnetic resonance (CW ODMR) or pulsed optically detected magnetic resonance (pulsed ODMR). The pulse width specifying unit 46 specifies the microwave irradiation time required to set the quantum state of the electron spin of the NV center of the diamond element 20 to a specific state by reading out a value previously stored in the memory unit 43, having the user input it through the input unit 61, or having the measurement unit 2 perform Rabi oscillation measurement based on the resonance frequency specified by the resonance frequency specifying unit 45. The specified resonance frequency and irradiation time are stored in the memory unit 43. Here, as an example, an overview of specifying the resonance frequency based on CW ODMR and specifying the irradiation time based on Rabi oscillation measurement will be described.

[0045] [CW ODMR] CW ODMR is a technique for measuring the intensity of fluorescence emitted by an NV center by placing a diamond element 20 so that the target NV center is included in the observation field of an objective lens 34, irradiating the diamond element 20 with microwaves while sweeping the frequency from a microwave source 23, and irradiating it with initialization / readout laser light from a laser source 31. The fluorescence emitted from the NV center passes through the objective lens 34, dichroic mirror 33, fluorescence filter 35, imaging lens 36, and pinhole 37 in this order, and is detected by a single-photon detector 38. The detection result of the single-photon detector 38 is counted by a pulse counter 47 in the control / processing unit 4, and then stored in a memory unit 43 in a predetermined format and output to an input receiving unit 42. The input receiving unit 42 receives input of intensity measurement data of the fluorescence emitted from the NV center (data obtained by counting the number of detected photons) and outputs it to a resonant frequency identifying unit 45 and a display processing unit 48. FIG. 3 shows an example of data obtained by CW ODMR plotted by the display processor 48 and displayed on the display unit 62. The display processor 48 can plot measurement data and also plot approximate curves (dashed lines) in the figure. As described above, when a quantum state of Ms = −1 is achieved by irradiating a resonant microwave and then irradiating an initialization / readout laser beam repeatedly, the total fluorescence intensity is approximately 30% smaller than the total fluorescence intensity obtained by repeatedly irradiating an initialization / readout laser beam in a quantum state of Ms = 0. The resonant frequency determination unit 45 utilizes this fact to determine the resonant frequency by detecting dips in the measurement data of the fluorescence intensity obtained by CW ODMR.

[0046] [Rabi vibration measurement] Rabi oscillation measurement is a technique in which the laser source 31 irradiates the NV center with initialization / readout laser light to initialize the quantum state of the electron spin of the NV center to Ms = 0, the microwave source 23 irradiates with resonant microwaves for a predetermined time, and then the laser source 31 again irradiates with initialization / readout laser light to measure the intensity of fluorescence emitted from the NV center. This procedure is repeated while changing the irradiation time of the resonant microwave using a microwave switch. Figure 4 shows an example of data obtained by Rabi oscillation measurement being plotted by the display processing unit 48 and displayed on the display unit 62. As shown in Figure 4, the fluorescence intensity of the NV center, i.e., the quantum state of the NV center's electron spin (in other words, the electron energy state), oscillates between two levels, Ms = 0 and Ms = -1, with respect to the irradiation time of the resonant microwave. 4, near the maximum point of the periodic fluctuation of the fluorescence intensity, the electron spin of the NV center is in a quantum state of Ms = 0 (upward, i.e., corresponding to θ = 0), and near the minimum point, the electron spin of the NV center is in a quantum state of Ms = -1 (downward, i.e., θ = π). Utilizing this, the pulse width specifying unit 46 specifies a time equivalent to half the period of the Rabi oscillation as the irradiation time of the resonant microwave required to change the quantum state of the NV center's electron spin from Ms = 0 to Ms = -1 (from θ = 0 to θ = π) (the pulse corresponding to this irradiation time is referred to as a π pulse). Alternatively, the pulse width specifying unit 46 specifies a time equivalent to one-quarter of the period of the periodic fluctuation as the irradiation time of the resonant microwave required to change the quantum state of the NV center's electron spin from Ms = 0 to a superposition state of Ms = 0 and Ms = -1 (from θ = 0 to θ = π / 2) (the pulse corresponding to this irradiation time is referred to as a π / 2 pulse).

[0047] After the operations of the resonant frequency specifying unit 45 and the pulse width specifying unit 46 described above, the control unit 41 instructs each unit to sequentially measure the intensity of the fluorescence from the NV center based on the pulse sequence output to the model setting unit 44.

[0048] As shown in Table 1, the data processing method according to the present invention can process measurement data acquired by various measurement techniques. CW ODMR and Rabi oscillation measurement have already been explained, but here we will explain the Qdyne method, which is the measurement technique according to this embodiment, and Hahn echo measurement, which is a more basic measurement technique.

[0049] [Hahn echo measurement] Figure 5 shows the pulse sequence used to measure the AC magnetic field generated by the nuclear spins surrounding the electron spin of the NV center using Hahn echo measurements. Figure 6 shows a schematic diagram of how the quantum state of the electron spin of the NV center is manipulated by the pulse sequence shown in Figure 5.

[0050] When the control unit 41 instructs execution of a pulse sequence, the laser source 31 irradiates the diamond element 20 with initialization / readout laser light, and initializes the quantum state of the electron spin of the NV center to Ms=0 (FIG. 6(a)).

[0051] Next, the microwave source 23 irradiates the diamond element 20 with a π / 2 pulse of resonant microwaves, causing the quantum state of the electron spin of the NV center to be in a superposition state of Ms=0 and Ms=-1 (Fig. 6(b)). In Fig. 6, the π / 2 pulse corresponds to an operation of rotating the electron spin by π / 2 [rad] around the y axis. In this state, the quantum state of the electron spin of the NV center accumulates an AC magnetic field with a phase φ1 during the time τ until the microwave source 23 irradiates the π pulse of resonant microwaves (Fig. 6(c)). Note that in the superposition state, for example, an AC magnetic field B generated by nuclear spins in the vicinity of the NV center during the time τ nuc The phase φ accumulated by (t) changes the gyromagnetic ratio of the electron spin of the NV center to γ e is expressed by the following formula:

number

[0052] After a time τ has elapsed since the microwave source 23 irradiated the diamond element 20 with a π / 2 pulse, a subsequent π pulse is irradiated. In Figure 6, this π pulse corresponds to an operation of rotating the electron spin by π [rad] around the y axis. By irradiating the π pulse, the electron spin of the NV center becomes symmetrical with respect to the y axis (the phase based on the positive direction of the x axis is π-φ1) (Figure 6(d)). In this state, the quantum state of the electron spin of the NV center accumulates the AC magnetic field as a phase φ2 during the time τ until the microwave source 23 again irradiates a π / 2 pulse of the resonant microwave (Figure 6(e)). At this time, the phase of the electron spin is π-(φ1+φ2) based on the positive direction of the x axis.

[0053] After a time τ has elapsed since the microwave source 23 irradiated the diamond element 20 with a π pulse, it again irradiates it with a π / 2 pulse. By irradiating the π / 2 pulse, the quantum state of the electron spin of the NV center forms an angle θ = φ = φ1 + φ2 with respect to the quantization axis (z axis) (Figure 6(f)). In this state, the laser source 31 irradiates the diamond element 20 with initialization / readout laser light again, and the emitted fluorescence is detected by the single-photon detector 38, so that the quantum state of the electron spin of the NV center, more specifically the angle θ, can be read out as a difference in fluorescence intensity.

[0054] As shown in Figure 5, the AC magnetic field B nucWhen the phase of (t) is reversed, φ2 accumulates in the opposite direction to φ1, resulting in a larger φ=φ1+φ2. Conversely, when the phase of the AC magnetic field does not reverse before and after a π pulse, as in the case of a DC magnetic field, φ2 accumulates in a direction that cancels out φ1 (the same direction as the accumulation of φ1), resulting in a smaller φ=φ1+φ2. In other words, in Hahn echo measurement, if the interval between π / 2 pulses and π pulses is τ, the AC component of the AC magnetic field with a frequency of 1 / 2τ can be selectively detected. Therefore, when performing Hahn echo measurement, spectrum data of the AC magnetic field can be obtained by executing a pulse sequence that scans the value of τ. In this embodiment, the spectrum data creation unit 49 can read measurement data from the storage unit 43 and create spectrum data.

[0055] [Qdyne method] The Qdyne method offers superior frequency resolution compared to the previously described measurement methods. Figure 7 shows the relationship between the pulse sequence of the Qdyne method and the clock signal in the control unit 41. When implementing the Qdyne method in the quantum sensing system 1 according to this embodiment, for example, an XY8-k measurement pulse sequence is performed, and the measured fluorescence intensity is associated with the clock signal in the control unit 41 to generate time-series data, which is then stored in the storage unit 43. In this embodiment, the spectral data generator 49 converts the time-series data into spectral data, for example, by fast Fourier transform (FFT), thereby enabling highly accurate calculation of the frequency of the AC magnetic field. The frequency resolution of the Qdyne method depends only on the stability of the clock signal, and spectral data with a linewidth of approximately several hundred μHz can be obtained.

[0056] As already explained, conventional quantum sensing involves repeating the same measurement multiple times to measure the quantum state, which can lead to measurements being continued under inappropriate measurement conditions. Furthermore, in quantum sensing that acquires spectral data, as mentioned above, analysis using the spectral data is performed only after the long-term measurement is completed, which can result in measurements continuing for a long period of time when they actually failed. In particular, with the Qdyne method, the process of converting time-series data into spectral data using FFT takes a considerable amount of time, making it impossible to quickly confirm the success or failure of the measurement. These issues significantly reduce the efficiency of quantum sensing research and development using NV centers.

[0057] In contrast, in the quantum sensing system 1 according to this embodiment, the likelihood calculation unit 52 calculates the likelihood for each value of the physical quantity to be measured based on measurement data acquired, for example, by the Qdyne method and the likelihood function set by the likelihood function setting unit 51. The user can refer to the calculated likelihood and determine that the measurement is successful if a likelihood peak is detected (e.g., in the measurement of an AC magnetic field, a magnetic field having a frequency around the peak is measured). Conversely, if an anomaly is detected in the likelihood (e.g., in the measurement of an AC magnetic field, a downward convex peak is obtained or no peak is obtained), the user can determine that the measurement has failed. As will be described later, the presence or absence of a likelihood peak or anomaly can be confirmed even during measurement when the amount of measurement data is small, and the processing time is short. Therefore, the user can know the success or failure of the measurement even during the measurement, and can know the success or failure of the measurement in a shorter time. The operations of the likelihood function setting unit 51 and the likelihood calculation unit 52 are described below.

[0058] The likelihood function setting unit 51 receives input of the fluorescence measurement model (in this embodiment, the arrival rate λ) from the model setting unit 44, and sets a likelihood function of a probability distribution (stochastic process) that uses the fluorescence measurement model as a parameter. The stochastic process can be, for example, a non-uniform Poisson process. Because the likelihood decreases exponentially with the number of measurement data and can result in underflow, and because calculating the likelihood by addition rather than multiplication requires less computational effort, it is preferable to use a logarithmic likelihood function as shown in the following formula as the likelihood function. Here, Y↑ is a vector whose elements are the time-series data of the number of photons actually detected, and θ↑ is a family of parameters including the physical quantity to be measured. Y k is the number of photons detected at the measurement time k, and λ k is the arrival rate at k. Since the same stochastic processes and likelihood functions (logarithmic likelihood functions) may be used for different physical quantities to be measured and different measurement methods, the likelihood function setting processing unit 51 can read out those stored in advance in the storage unit 43 and set them as the likelihood functions in order to reduce the burden on the user.

number

[0059] Note that the likelihood function setting unit 51 can display the stochastic process and likelihood function used in the likelihood function setting unit 51 by, for example, displaying a predetermined display / input screen on the display unit 62 as shown in Fig. 8, while allowing the user to change the probability distribution (stochastic process) with the fluorescence measurement model as a parameter and the likelihood function of the stochastic process as needed via the input unit 61. For example, in Fig. 8, when changing the stochastic process and likelihood function used in the likelihood function setting unit 51, the user can mark an X in the "Cancel default" box and change the stochastic process and likelihood function using a pull-down menu or by manual input.

[0060] After the likelihood function setting unit 51 sets the likelihood function, the control unit 41 can display a predetermined input screen on the display unit 62, and allow the user to make settings for calculating the likelihood through the input unit 62 or to give instructions to execute the likelihood calculation process. Fig. 9 is a diagram showing an example of an input screen (tab) displayed on the display unit 62 to allow the user to make settings for calculating the likelihood or to give instructions to execute the likelihood calculation process. As an example, the input screen displays three methods for causing the likelihood calculation unit 52 to calculate the likelihood.

[0061] In the first method, the user can cause the likelihood calculation unit 52 to calculate the likelihood using the measurement data up to the present time. When the user places an X in the "Calculate likelihood using measurement data up to the present time" box displayed on the input screen, enters a range of values ​​for the parameter (physical quantity to be measured) for calculating the likelihood, and issues a command to execute the likelihood calculation process, the control unit 41 accepts the command and causes the likelihood calculation unit 52 to calculate the likelihood using the measurement data up to the present time. Specifically, the likelihood calculation unit 52 reads all measurement data from the start of measurement to the present time from the memory unit 43 and also reads the likelihood function from the likelihood function setting unit 51. Assuming that the fluorescence measurement data P of the NV center conforms to the fluorescence measurement model P(λ), the likelihood calculation unit 52 calculates the likelihood that the fluorescence measurement data of the NV center will be obtained from the fluorescence measurement model for each value of the parameter within the input range, as shown in the following equation.

number

[0062] According to the first method, the user can calculate the likelihood for each parameter value at a desired timing, such as while the measurement unit 2 is sequentially measuring the fluorescence intensity of the NV center based on the pulse sequence output to the model setting unit 44, or after the measurement is completed. Therefore, the user can check the success or failure of the measurement at a desired timing.

[0063] In the second method, the user can specify in advance the timing to start likelihood calculation, the number of times to repeat likelihood calculation, the time interval between each iteration, etc. Note that the "number of times to repeat likelihood calculation" here refers to the number of times to repeat the operation, with one calculation for each parameter value in the input range counted as one iteration. When the user places an "X" in the "Reserve likelihood calculation timing, etc." box displayed on the input screen, inputs the time to start likelihood calculation (the elapsed time from the start of measurement), the number of times to calculate likelihood, the time interval between likelihood calculations, and the range of parameter values ​​for likelihood calculation, and issues an instruction to register the input, the control unit 41 accepts the instruction and causes the likelihood calculation unit 52 to start likelihood calculation at the input time and calculate the likelihood the input number of times and at the input time interval.

[0064] According to the second method, the user can confirm the success or failure of the measurement at a more accurate timing by setting the timing for calculating the likelihood in advance depending on the physical quantity to be measured and the measurement method.

[0065] In the second method, it is preferable to repeat the calculation of the likelihood multiple times at predetermined time intervals. While the measurement is being carried out, the likelihood is calculated multiple times to check the success or failure of the measurement and the frequency components being measured, thereby enabling quick detection of changes in the measurement environment, etc. In particular, since the NV center is extremely small as an object to be irradiated with laser light or microwaves, even a slight deviation in the positional relationship between the diamond element 20 and the wire or objective lens 34 can cause the measurement to fail. In such a situation, if the success or failure of the measurement can be confirmed by periodically referring to the latest likelihood calculation results, the efficiency of research and development will be greatly improved.

[0066] When likelihood calculation is repeated at predetermined time intervals, the likelihood can be updated successively by applying the latest measurement data to the previous likelihood calculation result. Therefore, even if measurements are performed over a long period of time and the amount of measurement data increases, the amount of calculation per calculation does not change. This is in contrast to FFT, in which the processing time increases as the amount of time-series data increases. Therefore, when using the second method, it is preferable to start likelihood calculation simultaneously with the start of measurement and perform it at regular time intervals (for example, every second).

[0067] In the third method, the user can cause the likelihood calculation unit 52 to calculate the likelihood using measurement data acquired during a predetermined period. When the user places an "X" mark in the "Select the range of data to be used and calculate the likelihood" box displayed on the input screen, inputs the measurement data acquisition period and the range of values ​​of the parameter (physical quantity to be measured) for calculating the likelihood, and issues a command to execute the likelihood calculation process, the control unit 41 accepts the command and causes the likelihood calculation unit 52 to calculate the likelihood using the measurement data acquired during the input period. Note that in FIG. 9, the "Select the range of data to be used and calculate the likelihood" box has not been marked with an "X," so it is not possible to input a value.

[0068] According to the third method, after the measurement is completed, the user can divide the entire measurement period into several periods and calculate the likelihood using the measurement data obtained in each period, allowing the user to easily check whether there are any periods in which the measurement failed, and if so, which periods.

[0069] The calculation of likelihood is suitable for parallel calculation, and the calculation time can be reduced by using a computer equipped with parallelized electronic circuits.

[0070] The likelihood calculation unit 52 can output the calculated likelihood to the measurement success / failure determination unit 53. The measurement success / failure determination unit 53 receives the likelihood input from the likelihood calculation unit 52 and detects whether or not a likelihood peak (maximum value of likelihood) exists. If a likelihood peak exists, the measurement success / failure determination unit 53 determines that a magnetic field of a frequency near the peak has been measured, and can display on the display unit that the measurement has been successful, or can notify by sound via a speaker (not shown) or the like. If no likelihood peak exists, or if a dip exists, or if the likelihood oscillates, the measurement success / failure determination unit 53 determines that some abnormality has been detected in the measurement, and can display on the display unit that the measurement has failed, or can notify by sound via a speaker (not shown) or the like. The user can easily know whether the measurement was successful by referring to the display or notification.

[0071] Furthermore, the likelihood calculation unit 52 can output the calculated likelihood to the display processing unit 48. The display processing unit 48 can receive an input of the likelihood from the likelihood calculation unit 52, and can plot and display on the display unit, for example, a graph with frequency on the horizontal axis and likelihood on the vertical axis. By referring to this display, the user can more easily know whether the measurement was successful or not. [Example]

[0072] Hereinafter, a first example of a quantum sensing system including an embodiment of a data processing method, a data processing device, and a data processing program according to the present invention will be described. In this first example, a homoepitaxial diamond film having a thickness of 20 μm was grown on the upper surface of a 5 mm × 5 mm × 1 mm Ib type high-pressure, high-temperature single crystal diamond substrate by plasma-assisted chemical vapor deposition. The growth conditions were a gas pressure of 140 Torr, a microwave power of 1.2 kW, and a microwave power density of 120 to 180 W / cm. 3 The conditions were: methane concentration 1%, oxygen concentration 0%, total flow rate 200 SCCM, and substrate temperature 800±10°C. The diamond film was then irradiated with an accelerating energy of 2.5 keV and a fluence of 1.5×1011 cm -2After implanting a single ion of 15N at 1000°C, the material was heat-treated in 10% hydrogen gas at 1000°C to form NV centers. Individual NV centers were located at depths between 2nm and 11nm, with approximately 50% located within 4nm from the surface.

[0073] In the first example, a homemade scanning confocal fluorescence microscope (CFM) was used. The initialization and readout laser light was generated by a green 532 nm laser light source (Changchun New Industries optoelectronics Technology, MGL-III-532 nm 300 mW-1%) and pulsed with an AOM (Gooch & Housego, AOMO 3350-120). The objective lens (Olympus, MPLAPON 50×) was mounted on a piezo stage (PI, NanoCube P-611.3S) to scan the diamond element, which was attached to the sample holder. Microwave pulses from an analog radio frequency (RF) signal generator (Keysight, E4428C) were pulsed by an RF circuit consisting of a phase shifter, two switches, and a combiner (Mini-Circuits, ZX10Q-2-25-S+, ZASWA-2-50DR+, ZX10-2-442-S+) and amplified by a power amplifier (Mini-Circuits, ZHL-16W-43-S+). The two switches were controlled by a data timing generator (Textronics, DTG5274). The microwave pulses were irradiated through a copper nanowire placed on the surface of the diamond element. A static magnetic field strength of 30 mT was applied to the NV center by a 450 mT neodymium magnet placed behind the sample holder. Fluorescence emitted from the NV center was detected by a single-photon detector (Laser Components, COUNT-100C) after passing through a long-pass filter (≥650 nm). The photon counting protocol was implemented on an FPGA board (Digilent, Cora Z7-10) and transmitted as binary data to a desktop computer. The various components of the CFM were managed using the Qudi software package.

[0074] In the first example, a 2 MHz AC magnetic field was measured using a coil wound around a diamond element. The coil had a loop diameter of 38 mm, a wire diameter of 0.40 mm, and 50 turns. In the first example, the Qdyne method was used as the measurement method. The FFT was performed in Python using the Scipy library. The maximum likelihood estimation (MLE) was implemented in Python using the JAX library according to [Equation 3] to [Equation 9], and the calculation was horizontally split between two GPUs (Nvidia RTX 3060) running at 64-bit precision. The likelihood curve was created by calculating the log likelihood for all frequencies from 1.5 MHz to 2.5 MHz in 1 kHz steps.

[0075] Figure 10 shows the likelihood curves and FFT spectra obtained by processing the measurement data up to that point, at various elapsed times since the start of measurement. As shown in Figure 10, the likelihood curves are more stable than the FFT spectra across all elapsed times, and their peaks are clearer. The likelihood peak is correctly located around 2.0 MHz when the elapsed time is 0.3 seconds and the number of measurement data is 12,758, and this is also the case for subsequent elapsed times. Conversely, the FFT spectrum does not show a clear peak, even when the elapsed time is 150 seconds and the number of measurement data is approximately 6 million. The reason why the response of the FFT is slower than that of the MLE is that the FFT purely mathematically converts all measurement data from the time domain to the frequency domain, while the MLE calculates the likelihood when the measurement data originates from a predefined model.

[0076] In the stochastic process of reading out the quantum state of the electron spin of an NV center using the Qdyne method, MLE uses a model-based approach to account for various sources of randomness, such as quantization error and shot noise due to APD overlap, leading to a stable likelihood curve over time. Because MLE can account for the inherent randomness in the measurement, the number of data points required to detect the correct frequency peak is several orders of magnitude fewer than FFT. Therefore, the success or failure of the measurement can be quickly determined, even during the measurement.

[0077] In the first example, only one frequency is estimated using a model of an AC magnetic field consisting of a single frequency, but in principle, it is possible to estimate the presence or absence of multiple frequencies by creating a more detailed model. Also, the frequency resolution of the MLE can be improved by improving the model.

[0078] Also, the frequency of the FFT spectrum shown in the figure is demodulated, while the frequency axis of the MLE likelihood curve is not. Because the FFT spectrum obtained using the Qdyne method is in the kHz range due to the nature of the Qdyne method, an offset must be added to the FFT frequency to shift it to the MHz range. Demodulation requires offset estimation, which is often done heuristically and may result in errors. In contrast, MLE uses a measurement model, so the calculated frequency can be treated as the frequency of the magnetic field being measured. Because MLE directly inputs frequency into the likelihood function or log-likelihood function, MLE can cover a wider frequency range than FFT. [Example]

[0079] A second example of a quantum sensing system including an embodiment of a data processing method, a data processing device, and a data processing program according to the present invention will be described below. In the first example, a 2 MHz AC magnetic field was prepared as the magnetic field to be measured, whereas in the second example, noise data was prepared instead of the AC magnetic field. The second example was carried out to verify whether noise would be erroneously detected as an AC magnetic field by MLE.

[0080] Figure 11 shows the results of calculating the likelihood of the noise data based on the model prepared in the first example. In contrast to the first example, a dip is observed around 2 MHz. This dip is thought to indicate a lack of similarity between the data and the model. The presence of a dip in the range from approximately 1.8 MHz to 2.2 Hz and the flatness of the remaining areas is thought to be due to the frequency filtering characteristics of the XY8-k measurement pulse sequence. In other words, this is thought to be because the pulse sequence of the XY8-k measurement performed using the Qdyne method can selectively detect this frequency range. Therefore, in this frequency range, a peak appears when a certain frequency closely matches the observed value, and a dip appears when it does not. According to our research, the measurement model used in MLE becomes less likely to falsely detect noise (becomes robust) when the number of measurement data exceeds 200,000 (typically a measurement time of approximately 5 seconds).

[0081] [Variations] The quantum sensing system 1 according to this embodiment can be modified in various ways. For example, the measurement unit 2 can include a scanning mirror for scanning the diamond element with laser light. Furthermore, the measurement unit 2 can use an arbitrary waveform generator (AWG) as a microwave source instead of a microwave switch to generate pulsed microwaves. Furthermore, the microwave irradiation device can be a microwave antenna or an on-chip coplanar waveguide instead of the wire described above.

[0082] Furthermore, the embodiments of the data processing method, data processing device, and data processing program according to the present invention are not limited to the examples described above. For example, in the above-described embodiments, the data processing method, data processing device, and data processing program according to the present invention are included in the quantum sensing system 1 and include the measurement unit 2. However, the measurement unit 2 is not essential for the data processing method, data processing device, and data processing program according to the present invention. FIG. 12 is a diagram showing an outline of the configuration of a data processing system according to a modified example. The data processing system 10 includes a data processing unit 40, a user interface 6, and a data transmission unit 7. The data processing unit 40 includes, as functional blocks, an input receiving unit 42, a memory unit 43, a model setting unit 44, a display processing unit 48, a spectral data creation unit 49, a likelihood function setting unit 51, a likelihood calculation unit 52, and a measurement success / failure determination unit 53. These functions are similar to those of the functional blocks of the control / processing unit 4 described above, and therefore will not be described here. The user interface 6 includes an input unit 61 and a display unit 62. The data transmission unit 7 may be a measurement device or a data management computer (database).

[0083] In the data processing system 100 according to the modified example, when a user issues an instruction to process predetermined measurement data via the input unit 61, the input accepting unit 42 accepts input of the measurement data from the data transmitting unit 7 and performs the likelihood calculation process described above. Details of this process are the same as those performed by the control and processing unit 4 described above, and therefore will not be described here.

[0084] [Aspect] It will be apparent to those skilled in the art that the above-described exemplary embodiments are examples of the following aspects.

[0085] (Item 1) A data processing method according to one aspect of the present invention comprises: a measurement data preparation process for preparing measurement data of fluorescence emitted from a diamond having an NV center, the measurement data being obtained by irradiating the diamond with electromagnetic waves based on a predetermined pulse sequence; a likelihood function preparation process for preparing a likelihood function of a probability distribution having as parameters a measurement object model including the physical quantity of the measurement object as a parameter and a fluorescence measurement model of the NV center defined based on the predetermined pulse sequence; a likelihood calculation process for calculating the likelihood of the parameter based on the measurement data and the likelihood function; Includes.

[0086] (Item 2) A data processing device according to one aspect of the present invention comprises: A memory unit; an input receiving unit that receives inputs of measurement data of fluorescence emitted from the NV center, the measurement data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a measurement object model that includes the measurement object physical quantity as a parameter and a fluorescence measurement model of the NV center that is defined based on the predetermined pulse sequence; a likelihood calculation unit that calculates the likelihood of the parameter based on the measurement data and the likelihood function; Equipped with.

[0087] (Item 9) A data processing program according to one aspect of the present invention includes a program for causing an electronic computer to: A memory unit; an input receiving unit that receives inputs of measurement data of fluorescence emitted from the NV center, the measurement data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a measurement object model that includes the measurement object physical quantity as a parameter and a fluorescence measurement model of the NV center that is defined based on the predetermined pulse sequence; The likelihood calculation unit operates to calculate the likelihood of the parameter based on the measurement data and the likelihood function.

[0088] The data processing method according to paragraph 1, the data processing device according to paragraph 2, and the data processing program according to paragraph 9 calculate the likelihood of the parameter based on a likelihood function of a probability distribution whose parameters are a fluorescence measurement model of the NV center defined based on a measurement object model that includes the physical quantity to be measured as a parameter and the predetermined pulse sequence, and the measurement data of the fluorescence emitted by the NV center is obtained by irradiating a diamond having an NV center with electromagnetic waves based on the predetermined pulse sequence. A person implementing the data processing method according to paragraph 1, the data processing device according to paragraph 2, and the data processing program according to paragraph 9 can refer to the calculated likelihood and determine that the measurement is successful if a peak in the likelihood exists (e.g., in the measurement of an AC magnetic field, a magnetic field having a frequency around the peak is measured). Conversely, if an abnormality is found in the likelihood (e.g., in the measurement of an AC magnetic field, a downward convex peak is obtained or no peak is obtained), the person implementing the data processing method according to paragraph 1, the data processing device according to paragraph 2, and the data processing program according to paragraph 9 can determine that the measurement is unsuccessful. According to the data processing method of paragraph 1, the data processing device of paragraph 2, and the data processing program of paragraph 9, the presence or absence of likelihood peaks and likelihood anomalies can be confirmed even when the amount of measurement data is small during measurement, and the processing time required is short. Therefore, those who implement the data processing method of paragraph 1, the data processing device of paragraph 2, and the data processing program of paragraph 9 can know the success or failure of the measurement even during the measurement. Furthermore, those who implement the data processing method of paragraph 1, the data processing device of paragraph 2, and the data processing program of paragraph 9 can know the success or failure of the measurement in a shorter time.

[0089] (Item 3) The data processing device according to item 3 is the data processing device according to item 2, further comprising a control unit that operates the likelihood calculation unit at a predetermined timing during the period in which the predetermined pulse sequence is being executed.

[0090] According to the data processing device of paragraph 3, by processing the measurement data at a predetermined timing during the period in which the pulse sequence is being executed, the user can know the success or failure of the measurement at a predetermined timing during the measurement.

[0091] (4) The data processing device according to paragraph 4 is the data processing device according to paragraph 2, further comprising a control unit that operates the likelihood calculation unit a predetermined number of times at predetermined time intervals during the period in which the predetermined pulse sequence is being executed.

[0092] According to the data processing device of paragraph 4, while the measurement is being carried out, the likelihood is calculated a predetermined number of times at predetermined time intervals to check the success or failure of the measurement and the frequency components being measured, thereby enabling quick detection of changes in the measurement environment, etc.

[0093] (Item 5) The data processing device according to item 5 is the data processing device according to item 2, wherein the likelihood calculation unit calculates the likelihood based on measurement data acquired during a predetermined period of time.

[0094] According to the data processing device of paragraph 5, for example, after the measurement is completed, the user can divide the entire measurement period into several periods and calculate the likelihood using the measurement data obtained in each period, and can easily check whether there are any periods in which the measurement failed, and if so, which periods.

[0095] (Clause 6) The data processing device according to clause 6 is the data processing device according to clause 2, further comprising a likelihood output unit that outputs the likelihood calculated by the likelihood calculation unit in a predetermined format.

[0096] According to the data processing device of paragraph 6, the user can more easily know whether the measurement was successful or not by referring to the likelihood output in a predetermined format.

[0097] (Clause 7) The data processing device according to clause 7 is the data processing device according to clause 2, further comprising a determination unit that determines whether or not a peak exists in the likelihood calculated by the likelihood calculation unit.

[0098] According to the data processing device of paragraph 7, the user can more easily know whether the measurement was successful or not by referring to the determination result by the determining unit.

[0099] (Item 8) The data processing device according to item 8 is the data processing device according to item 2, further comprising a spectrum data creating unit that creates spectrum data of the magnetic field to be measured based on the measurement data.

[0100] According to the data processing device of paragraph 8, the user can more accurately know the frequency components of the magnetic field to be measured by referring to the spectrum data created based on the measurement data in addition to the likelihood estimated by the likelihood calculation unit. [Explanation of symbols]

[0101] 1...Quantum sensing system 10...Data processing system 2...Measuring part 20...Diamond element 21...Static magnetic field source 22...Pulse generator 23...Microwave source 24...Microwave waveguide 25...Microwave switch 31...Laser source 32...Acousto-optic modulator 33...Dichroic mirror 34...Objective lens 35...Fluorescence filter 36...Imaging lens 37...Pinhole 38...Single photon detector 4...Control and processing section 40...Data processing unit 41...Control unit 42...Input reception section 43...Storage section 44...Model setting section 45...Resonant frequency identification section 46...Pulse width specification section 47...Pulse counter 48...Display processing unit 49...Spectral data creation section 51...Likelihood function setting unit 52...Likelihood calculation unit 53...Measurement success / failure determination section 6. User Interface 61...Input section 62...Display section 7...Data transmission unit

Claims

1. a measurement data preparation process for preparing measurement data of fluorescence emitted from an NV center obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence; a likelihood function preparation process for preparing a likelihood function of a probability distribution having as a parameter a fluorescence measurement model of the NV center defined based on a measurement object model including the physical quantity of the measurement object as a parameter and the predetermined pulse sequence; a likelihood calculation process for calculating the likelihood of the parameter based on the measurement data and the likelihood function; Data processing methods, including:

2. A memory unit; an input receiving unit for receiving inputs of measurement data of fluorescence emitted from an NV center, the fluorescence data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a fluorescence measurement model of the NV center defined based on a measurement object model that includes the physical quantity of the measurement object as a parameter and the predetermined pulse sequence; a likelihood calculation unit that calculates the likelihood of the parameter based on the measurement data and the likelihood function; A data processing device comprising:

3. moreover, The data processing device according to claim 2 , further comprising a control unit that operates the likelihood calculation unit at a predetermined timing during a period in which the predetermined pulse sequence is being executed.

4. moreover, 3. The data processing device according to claim 2, further comprising a control unit that operates the likelihood calculation unit a predetermined number of times at predetermined time intervals during a period in which the predetermined pulse sequence is being executed.

5. The data processing device according to claim 2 , wherein the likelihood calculation section calculates the likelihood based on measurement data acquired during a predetermined period.

6. moreover, The data processing device according to claim 2 , further comprising a likelihood output unit that outputs the likelihood calculated by said likelihood calculation unit in a predetermined format.

7. moreover, The data processing device according to claim 2 , further comprising a determination unit that determines whether or not a peak exists in the likelihood calculated by said likelihood calculation unit.

8. moreover, 3. The data processing device according to claim 2, further comprising a spectrum data creating unit that creates spectrum data of the magnetic field to be measured based on the measurement data.

9. Electronic computers, A memory unit; an input receiving unit for receiving inputs of measurement data of fluorescence emitted from an NV center, the fluorescence data being obtained by irradiating a diamond having an NV center with electromagnetic waves based on a predetermined pulse sequence, the type of physical quantity to be measured, and the predetermined pulse sequence; a likelihood function setting unit that sets a likelihood function of a probability distribution having as a parameter a fluorescence measurement model of the NV center defined based on a measurement object model that includes the physical quantity of the measurement object as a parameter and the predetermined pulse sequence; a data processing program that operates as a likelihood calculation unit that calculates the likelihood of the parameter based on the measurement data and the likelihood function;

Images