Miniaturized pulse electron paramagnetic resonance spectrometer

Abstract

A miniaturized spectroscopy system includes: a spectrometer integrated circuit including: one or more on-chip sensors configured to: generate pulses of an oscillating magnetic field, and detect pulse electron paramagnetic resonance (EPR) signals in a first mode of operation, and one or more on-chip timing circuits to control the generation of the pulses.

Description

MINIATURIZED PULSE ELECTRON PARAMAGNETIC RESONANCE SPECTROMETERSTATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0001] This invention was made with government support under Grant No(s). N00014-21-1- 4005, awarded by the Office of Naval Research (ONR). The government has certain rights in the invention.BACKGROUND

[0002] The present invention relates generally to the field of electron paramagnetic resonance spectroscopy.SUMMARY

[0003] One aspect relates to a miniaturized spectroscopy system. The miniaturized spectroscopy system includes: a spectrometer integrated circuit including one or more on- chip sensors configured to: generate one or more pulses of an oscillating magnetic field, and detect one or more pulse electron paramagnetic resonance (EPR) signals in a first mode of operation.

[0004] Another aspect relates to a device. The device includes: an on-chip EPR spectrometer including: one or more sensing cells, each sensing cell comprising a sensor oscillator, one or more amplifiers, and one or more mixers, and a controller including a memory storing instructions thereon that, when executed by one or more processors, cause the one or more processors to: generate, using a sensor oscillator of one of the one or more sensing cells, a plurality of pulses of an oscillating magnetic field at a first frequency to excite one or more electrons, turn off the sensor oscillator to cease generation of pulses, sense, using the one or more sensing cells, an alternating current (AC) signal produced by the one or more electrons indicative of the excitation of the one or more electrons, amplify, using the one or more amplifiers, the AC signal, convert, using the one or more mixers, the AC signal into a baseband signal, and detect the baseband signal.

[0005] Still another aspect relates to a method for detecting electron paramagnetic resonance (EPR) signals. The method includes generating, using one or more sensor oscillators, a plurality of pulses of an oscillating magnetic field at a first frequency to excite one or more electrons, turning off the one or more sensor oscillators to cease generation of pulses, sensing, using one or more sensing cells, an alternating current (AC) signal produced by the one or more electrons indicative of the excitation of the one or more electrons, amplifying, using one or more amplifiers, the AC signal, converting, using one or more mixers, the AC signal into a baseband signal, and detecting the baseband signal..BRIEF DESCRIPTION OF THE DRAWINGS

[0006] These and other aspects and features of the present implementations are depicted by way of example in the figures discussed herein. Present implementations can be directed to, but are not limited to, examples depicted in the figures discussed herein. Thus, this disclosure is not limited to any figure or portion thereof depicted or referenced herein, or any aspect described herein with respect to any figures depicted or referenced herein.

[0007] FIG. 1 is a diagram illustrating the Zeeman effect, according to an exemplary embodiment.

[0008] FIG. 2 is a diagram of a continuous wave (CW) electron paramagnetic resonance (EPR) detection concept, according to an exemplary embodiment.

[0009] FIG. 3 is a diagram of a principle of pulse EPR, according to this disclosure.

[0010] FIG. 4A is a diagram of a CW EPR detection concept that may be used by an integrated circuit, according to an exemplary embodiment.

[0011] FIG. 4B is a diagram of a pulse EPR detection concept that may be used by an integrated circuit, according to an exemplary embodiment.

[0012] FIG. 5A is schematic of a ringdown period in a cross-coupled inductor-capacitor (LC) oscillator, according to an exemplary embodiment.

[0013] FIG. 5B is schematic of a modification of the schematic of FIG. 5A using injection locking and a switch, according to an exemplary embodiment.

[0014] FIG. 6A is a schematic showing inductive and substrate coupling from a reference oscillator into a sensor oscillator, according to an exemplary embodiment.

[0015] FIG. 6B is a schematic showing the reference oscillator of FIG. 6A operating at a reduced frequency to reduce the inductive and substrate coupling, according to an exemplary embodiment.

[0016] FIG. 6C is a schematic showing a subharmonic direct-conversion architecture, according to an exemplary embodiment.

[0017] FIG. 7 is a schematic diagram of a dual-mode EPR integrated circuit, according to an exemplary embodiment.

[0018] FIG. 8 is a schematic diagram of a portable spectroscopy system, according to an exemplary embodiment.

[0019] FIG. 9A is a circuit diagram of a sensing oscillator cell of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0020] FIG. 9B is chart showing the cell of FIG. 9A generating a pulse with and without injection-locking and ringdown control, according to an exemplary embodiment.

[0021] FIGS. 10A-10B are circuit diagrams of a reference quadrature voltage controlled oscillator (QVCO) of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0022] FIG. 11 is a circuit diagram of a phase interpolation stage of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0023] FIG. 12A is a circuit diagram of a radiofrequency operational transconductance amplifier (RFOTA) stage of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0024] FIG. 12B is a circuit diagram of three of the RFOTA stages of FIG. 12A, according to an exemplary embodiment.

[0025] FIG. 13 is a circuit diagram of a subharmonic mixer of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0026] FIG. 14 is a circuit diagram of a variable gain amplifier (VGA) stage of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0027] FIG. 15 is a circuit diagram of a quadrature frequency-modulation (FM) demodulator used for CW EPR detection in the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0028] FIG. 16A is a diagram of an inductor having a standard spiral configuration showing the magnetic coupling from it to sensor inductors, according to an exemplary embodiment.

[0029] FIG. 16B is a schematic of a figure-eight inductor with reduced magnetic coupling compared to the inductor of FIG. 16A, according to an exemplary embodiment

[0030] FIG. 17 is a schematic of a pulse sequencing digital logic, according to an exemplary embodiment.

[0031] FIG. 18 is a timing diagram of enable signals for the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0032] FIG. 19 is an illustration of the integrated circuit of FIG. 7, according to an exemplary embodiment.

[0033] FIG. 20 is a chart showing the power consumption of the integrated circuit of FIG. 19, according to an exemplary embodiment.

[0034] FIG. 21 is a diagram of a portable spectroscopy system, according to an exemplary embodiment.

[0035] FIG. 22A is graph showing oscillator phase noise for a reference QVCO, according to an exemplary embodiment.

[0036] FIG. 22B is a plot of the reduction of phase noise in a sensor VCO enabled by injection locking, according to an exemplary embodiment.

[0037] FIG. 23 is schematic showing free induction decay, according to an exemplary embodiment.

[0038] FIG. 24A is a plot of a free induction decay signal, according to an exemplary embodiment.

[0039] FIG. 24B is a Fourier transform of the plot of a FIG. 24 A, according to an exemplary embodiment.

[0040] FIG. 25 is a plot of a continuous wave EPR spectrum, according to an exemplary embodiment.

[0041] FIG. 26A is a plot of free induction decay signals, according to an exemplary embodiment.

[0042] FIG. 26B is a Fourier transform plot of the free induction decay signals of FIG. 25A, according to an exemplary embodiment.

[0043] FIG. 27A is diagram showing a two-pulse inversion recovery, according to an exemplary embodiment.

[0044] FIG. 27B is a plot of two-pulse inversion recovery, according to an exemplary embodiment.

[0045] FIG. 28A is a diagram of the Hahn echo concept, according to an exemplary embodiment.

[0046] FIG. 28B is a diagram of echoes generated with anthracite and coffee and a plot with an echo amplitude and the delay between pulses, according to an exemplary embodiment.

[0047] FIG. 29A is diagram of a three-pulse stimulated echo concept, according to an exemplary embodiment.

[0048] FIG. 29B is a diagram of pulse sequence to produce a stimulated echo, according to an exemplary embodiment.

[0049] FIG. 30 is a plot of a continuous wave spectrum of DPPH and anthracite, according to an exemplary embodiment.

[0050] FIG. 31 is a plot of a continuous wave spectrum of TEMPO in glycerol, according to an exemplary embodiment.

[0051] FIG. 32 is a flow diagram illustrating a method of EPR spectrometry, according to an exemplary embodiment.DETAILED DESCRIPTION

[0052] Aspects of this technical solution are described herein with reference to the figures, which are illustrative examples of this technical solution. The figures and examples below are not meant to limit the scope of this technical solution to the present implementations or to a single implementation, and other implementations in accordance with present implementations are possible, for example, by way of interchange of some or all of the described or illustrated elements. Where certain elements of the present implementations can be partially or fully implemented using known components, only those portions of such known components that are necessary for an understanding of the present implementations are described, and detailed descriptions of other portions of such known components are omitted to not obscure the present implementations. Terms in the specification and claims are to be ascribed no uncommon or special meaning unless explicitly set forth herein. Further, this technical solution and the present implementations encompass present and future known equivalents to the known components referred to herein by way of description, illustration, or example

[0053] Electron paramagnetic resonance (EPR) is a spectroscopic technique for direct detection and characterization of free radicals (e.g., species containing unpaired electrons). EPR may also be referred to as electron spin resonance (ESR). EPR can detect and characterize paramagnetic species (e.g., free radicals) with high sensitivity and specificity. EPR may be suited for a wide range of biomedical applications. Applications include, for example, chemical characterization, site-directed spin labeling (e.g., to investigate protein structure) and detection of reactive oxygen species (e.g., in-vivo detection of harmful reactive oxygen species). Two major experimental methods of EPR spectroscopy are continuous wave (CW) EPR and pulse EPR.

[0054] Additional applications include, for example, studies of metal-containing enzyme complexes, in-vivo / in-vitro detection and imaging of radical species, and investigation of biomolecular structure and dynamics in conjunction with spin-labeling. EPR may also be utilized in the development of materials, for example, in defect characterization in semiconductors and organic materials. EPR may also be applied in quantum information science, particularly in manipulating and controlling spin qubits such as nitrogen-vacancy (NV) color centers in diamond. A limitation of current EPR technology is the bulky spectrometer, which limits portability and allows only serial measurements. As such,miniaturized EPR spectrometers may enable additional uses for EPR, such as leveraging the unique capability of EPR for radical detection and characterization as a robust sensing modality for point-of-care diagnostics, which can bring portable, inexpensive, and highly sensitive analysis equipment directly into a physician’s office or the patient’s home. Various embodiments of EPR systems are integrated on-chip, with sensing components integrated into a microchip, also known as an integrated circuit (IC). Current on-chip EPR systems include capabilities to perform CW EPR, but not pulse EPR. Further, current on-chip CW EPR systems do not include an FM demodulator on-chip, but rather implement the demodulator separately (e.g., on a circuit board).

[0055] CW spectrometer integrated circuits (ICs), many of which are compatible with miniaturized permanent magnets for portable use, may be used in on-off modulation or chopping for enhanced signal-to-noise ratio (SNR), high-field EPR for increased sensitivity and spectral resolution, and large sensitive volumes using multi-element sensing surfaces. However, pulse EPR spectrometer ICs have yet to see similar levels of development or on- chip integration. Integrated inductor-capacitor (LC) oscillators are a current method of excitation and sensing for on-chip CW EPR, owing to their simplicity, ease of design, and compatibility with standard integrated circuit fabrication processes. In various embodiments, an inductor of an LC oscillator may be a coil. In various embodiments, the coil comprises a part of a resonator used for EPR excitation and detection. Using an LC oscillator may also be an attractive approach to perform pulse EPR, since it can be turned on and off to generate pulses, and the resulting magnetization signal from a sample will induce a voltage on the LC tank. However, the slow startup and ringdown of oscillations in the LC tank have precluded the use of LC oscillators as pulse sensors. To avoid this, keeping the oscillator on and frequency-shifting onto and off of resonance to perform pulse-like measurements may be performed. The magnetization signal, however, is encoded in nonlinear amplitude and frequency modulation of the oscillation at a high baseband frequency, which makes reconstructing the EPR signal potentially difficult.

[0056] Accordingly, the systems and methods described herein provide a system for performing pulse EPR on a chip (e g., on a microchip scale). In some embodiments, the system may be configured to enable both CW EPR and pulse EPR on the same chip. CW can be achieved with simpler spectrometer architectures and generally offers higher sensitivities than pulse-mode measurements, while pulse EPR can provide more detailed information on spin dynamics and interactions that are unobtainable via CW. Technically and beneficially,this reduces the size of the spectrometry system, allowing EPR to be performed in smaller environments (e.g., point-of-care diagnostics). Additionally, the system includes an on-chip demodulator for CW EPR, thus further reducing a scale of the system.

[0057] An oscillator integrated into an integrated circuit (IC) chip may be utilized as an EPR sensing cell to perform CW EPR and / or pulse EPR. As will be described herein, a spectrometer IC utilizes pulse EPR spectroscopy on a highly integrated and low-power platform with on-chip multi-pulse sequencing. In various embodiments, the spectrometer IC may generate an EPR frequency of, for example, 14GHz with on-chip sensing oscillators, and these oscillators may also be used to perform CW EPR spectroscopy. A subharmonic architecture is described herein, which may enable the use of an oscillator as a flexible on- chip sensing cell and provide direct detection of transverse magnetization after pulses. Specifically, the subharmonic architecture may avoid a large voltage signal associated with a generated pulse being picked up and detected when a smaller free induction decay (FID) signal (e.g., the desired signal to be detected) is being generated. In various embodiments, pulse and CW EPR may be integrated on-chip (e.g., on a microchip), either alone or together on a single chip. EPR may utilize a plurality of EPR sensors to perform the methods described herein. In various embodiments, EPR experiments may be conducted on a larger scale relative to on-chip embodiments.

[0058] Referring to FIGS. 1-4 together, a concept of dual-mode EPR is illustrated, according to an example embodiment. The dual-mode EPR may be utilized by an integrated circuit (e.g., CW and pulse EPR methodologies combined on a single chip).

[0059] Referring now to FIG. 1, a diagram 100 illustrating the general principle of EPR is shown, according to an example embodiment. Particularly, the Zeeman effect on electrons is shown, which guides the principle and applications of EPR. When unpaired electrons 102 are placed within a magnetic field 104 (e.g., Bo), the electrons undergo splitting in energy levels. That is, the EPR signal originates from the interaction of a magnetic moment of unpaired electrons with a static magnetic field. As shown in FIG. 1., the magnetic field may be equal to 0 (e.g., no magnetic field is applied or the electrons are not placed in the magnetic field) prior to splitting. In one embodiment, where an unpaired electron with a spin quantum number S = 1 / 2 as depicted in FIG. 1, the spin energy states are degenerated when the external magnetic field Bo= 0. Upon placement in the magnetic field 104, the energy of the electrons may be split. That is, when Bo> 0, a splitting into two discrete energy levels occurs (e.g.,the Zeeman effect). The magnetic field may be a static magnetic field. In some embodiments, the magnetic field 104 may be generated using a permanent magnet. A permanent magnet may be used for generating a static or uniform magnetic field over a small region or area (e g., below a threshold size). In other examples, the magnetic field may be generated using electromagnets. An electromagnet may be used for generating a static or uniform magnetic field over a large region or area (e g., above a threshold size).

[0060] Further, the energy split may occur between a lower energy state and a higher energy state. Transitions between energy levels can be induced by electromagnetic radiation with a magnetic componentoscillating at a frequency co0that satisfies the resonance condition. That is, a photon energy of the radiation matches the gap between the two spin states:"o = ~YB0(1) where <u0is known as the Larmor frequency and y is the gyromagnetic ratio, y is in general given by g{3 / h, where {3 is the Bohr magneton, g is a proportionality constant known as the ^-factor, and h is the reduced Planck constant, y is approximately 2n X 28 GHz / T for a free electron with g — 2.0. Interactions of the unpaired electron with its environment modulate the spin energy levels, altering the EPR spectrum. For example, the hyperfine interaction between an electron spin and a nearby nuclear spin leads to splitting of the electron spin energy levels, yielding multiple spectral lines.

[0061] FIG. 1 shows electrons having a higher energy level 106 with magnetic moment(s) aligned nearer a north pole 112 of the magnetic field 104 and electrons having a lower energy level 108 with magnetic moment(s) aligned nearer a south pole 114 of the magnetic field 104.. As shown in FIG. 1, when the electrons split energy levels, the change in energy 110 is equal to Planck’ s constant, h, times a fundamental frequency (fo), which is proportional to the strength or value of the magnetic field Bo. That is, as the strength of the magnetic field 104 increases, the change in energy level also increases. EPR may utilize this Zeeman effect, thereby indicating that EPR may be performed at different magnetic field strengths and / or based on the energy difference 110. For example, in some embodiments, the strength of the magnetic field 104 may be 0.2T, 0.5T, LOT, 1.25T. 1.4T, 2.0T, etc. Further, the strength of the magnetic field 104 may determine a frequency as which a spectrometer system operates (e.g., the frequency changes based on magnetic field strength). For example, energy transitions may be excited by applying an oscillating field Bi with a frequency coo determined by Bo. Additionally, y may be calculated according to the equation 116, by dividing thefrequency (Do by the magnetic field Bo. In some embodiments, y may be equal to a value of approximately 28GHz / T.

[0062] Referring now to FIG. 2, a schematic 200 illustrating continuous wave (CW) EPR is shown, according to an example embodiment. The CW EPR detection concept of FIG. 2 may be used by an integrated circuit (IC). In CW EPR, a constant, low level (e.g., below a threshold) power excitation may be constantly output. That is, the sample is irradiated continuously with electromagnetic radiation. In CW EPR, the static magnetic field magnitude, Bo, or the frequency fo of a second oscillating field Bi, (e.g., which may be equal to coo / 27r), is swept while the other is held constant to generate the EPR spectrum. Bo may be swept, for example, when a large spectrometer that utilizes electromagnets is performing EPR. In chips that perform CW EPR, the frequency fo or coo may be swept. The magnetic field and / or frequency may be swept, for example, with frequency or field modulation to enhance sensitivity through phase-sensitive detection. Because the frequency depends upon the magnetic field strength, when the frequency and magnetic field strength values are aligned or substantially aligned (e.g., equal or substantially equal), a graph 202 having a characteristic shape (e.g., that of graph 202) is generated. The graph 202 is the CW EPR signal where the frequency is swept with a fixed field, as in some embodiments.

[0063] Additionally, the schematic 202 includes sinusoids 204 and 206. In some embodiments, as described above, in order to measure the spectrum 202 (e.g., frequency response) the external magnetic field may be fixed and frequency may be swept. In other embodiments, the frequency may be fixed and the magnetic field is swept. Sweeping the magnetic field may shift a resonant frequency (e.g., a Larmor frequency) at which the electrons exhibit energy splitting. To achieve a high sensitivity, a sinusoid modulation is applied to the magnetic field or the frequency. Thus, sinusoid 204 illustrates a sinusoidal modulation of a frequency and sinusoid 206 illustrates a demodulated CW EPR signal, from which the spectrum 202 is obtained. That is, the sinusoid 206 is the output of the frequency demodulators of the spectrometer in response to the modulation input 204. In frequency shiftbased detection, the sinusoid modulation may result in a frequency shift being detected and modulated. This occurrence may permit phase-sensitive detection. Phase-sensitive detection is a sensitive or low-noise detection mechanism that permits extraction of the amplitude and phase of the single (e g., fixed) frequency of the demodulated signal. As such, modulation ofthe fixed variable may permit phase-sensitive detection. In various embodiments, the phasesensitive detection may be accomplished with lock-in amplification.

[0064] In various embodiments, Bo may be swept when utilizing commercial instruments. In various embodiments, coo may be swept when utilizing a CW EPR chip. A spectrum may be recorded when either one of Bo or coo is swept. FIG. 2 illustrates a recorded spectrum 202 when frequency is swept. The spectrum 202 is a frequency vs. frequency shift plot. Additionally, as stated above, CW EPR may utilize low-power constant excitation of electrons. Absorption of the sample of electrons may be measured. In various embodiments, CW EPR may have a higher sensitivity relative to pulse EPR. In various embodiments utilizing chips, an oscillator may be used for CW EPR by keeping the oscillator on via a lower current bias and sensing a frequency shift in the resonant tank of the oscillator.

[0065] Referring now to FIG. 3, a pulse EPR detection concept is shown, according to an example embodiment. FIG. 3 includes a schematic 300 illustrating an operating principle of pulse EPR according to an example embodiment. In pulse EPR,pulses are used to manipulate the electron spins, and the evolution of the magnetization after one or more pulses is recorded and analyzed to yield the desired information. Due to the magnetic field, unpaired electrons develop a net magnetization (shown at equilibrium 302) which allows the electrons to behave as a single magnetic dipole. For example, when in the applied magnetic field, all electrons may align and produce a net magnetization. When pulses are applied to the electrons, as shown at excitation 304, the magnetization of each electron may tilt away from a vertical axis 308 (e g., an axis pointing north to south), shown in FIG. 3 as tilt 306. Further, sinusoid 310 indicates a pulse, shown as magnetic field (e.g., pulses) Bi, that is generated via excitation from a spectrometer. The amount by which the magnetization is tipped away from the axis 308 may depend upon the length of the generated pulse. For example, a longer duration of the pulse may correspond to a greater deviation (e.g., tilt) from a vertical plane and into a horizontal transverse plane. Additionally, a rate of the tilt may be expressed as cm. Additionally, cm may govern nutation 312 (e.g., an amount or rate of tilt). As shown, cm is dependent upon a strength of the magnetic field Bi.

[0066] Due to the energy gap 110 and Larmor frequency described above with respect to FIG. 1, the magnetization of the electrons may rotate after the electrons have been tiled away from the vertical axis 308. Such rotation is shown as precession 314. After the Bl pulses have been turned off or otherwise ceased, the magnetization may decay, and the electron may againpoint vertically (e.g., north to south), as shown at relaxation 316. For example, a tilt 318 decays so the electron is vertically aligned. The decay of the tilt may be sensed. For example, the element of the electron’s tilt that enters a transverse plane (shown at 320), may be sensed by a coil 322 of a spectrometer. In some embodiments, the coil 322 may be any type of resonator. For example, in one embodiment, the coil 322 may be a coil on a chip (e.g., a microchip). It should be understood that the resonator used to sense the decay in the transverse plane may be any type of resonator, such as a planar resonator. For example, in some embodiments, such as in spectrometers that utilize electromagnets, excitation and sensing may be performed by a type of cavity (e.g., a microwave resonator). In embodiments where a permanent magnet is utilized, excitation and sensing may be performed by a cavity resonator, planar resonator, loop-gap resonator.

[0067] Referring now to FIGS. 4A and 4B, dual-mode EPR on a chip is shown, according to an example embodiment. Specifically, both pulse and CW EPR may be performed on-chip using LC cross-coupled voltage-controlled oscillator (VCO) sensing cells, as illustrated in FIGS. 4A and 4B. Modifications to a cross-coupled oscillator may lead to fast oscillator startup and ringdown, thereby enabling generation of short pulses and detection of free induction decay (FID) signals after the pulses with minimal dead-time. In an example embodiment, and as will be described in greater detail with respect to at least FIGS. 5 and 7, a subharmonic architecture is utilized for pulse EPR. Specifically, an on-chip reference oscillator is used to injection-lock the sensor for fast startup and reduced phase noise of pulses. The reference is also used as the local oscillator (LO) for 2x-subharmonic downconversion of a received magnetization signal, which is detected directly as an induced voltage in the sensor tank. CW detection is performed via an on-chip frequency-modulation (FM) demodulator. An example spectrometer IC (e.g., having 14GHz frequency and Bo— 0.5T) may include two sensing cells, as is described below with respect to FIG. 19. The spectrometer may be used for pulse EPR with a plurality of pulses.

[0068] FIG. 4A shows a schematic 400 illustrating pulse EPR functionality implemented on a chip, according to an example embodiment. A circuit 402 may include a plurality of transistors configured as an electronic oscillator. In various embodiments, an oscillator may be, as shown in FIG. 4A, a cross-coupled inductor-capacitor (LC) voltage-controlled oscillator (VCO) including an LC tank resonator. A VCO is described in greater detail herein. The VCO 402 may allow a chip (e.g., a microchip on which EPR is performed) to performCW EPR. The VCO 402 may include a plurality of variable capacitors or varactors to vary the frequency of oscillation. For example, a first varactor may be used to cause a frequency sweep to the frequency of the spectrum 200, as shown by line 406. Furthermore, a second varactor may be used to cause a frequency modulation to the frequency of the spectrum 200, as shown by line 404.

[0069] Referring now to FIG. 4B, a schematic 450 illustrating pulse EPR functionality implemented on a chip is shown, according to an example embodiment. The schematic 450 includes a circuit 452. The circuit 452 includes an inductor (e.g., coil) 454 and transistors 456 and 458. The circuit 452 is a voltage controlled oscillator (VCO). In various embodiments, the VCO 452 may be the same circuit as VCO 402 as shown in FIG 4A. The VCO 452 may allow a chip (e.g., a microchip on which EPR is performed) to output or generate excitation pulses and sense EPR. The VCO 452 is turned on by enabling a current source 460 that permits current flow through the circuit 452 and allows oscillations to build up within the VCO 452. By pulsing the current source 460 (e.g., by turning the current source on and off), the pulses shown in the graph 462 can be generated. After forming such pulses, the VCO 452 is turned off. The magnetization described in FIG. 3 induces a voltage onto the coil 454, thereby allowing the VCO 452 to sense the magnetization signal from the EPR that is imparted on the coil 454. The magnetization signal may be transmitted to or sensed by the coil 454 as a voltage. The chip may therefore measure the EPR signal.

[0070] In some embodiments, the excitation may be delivered as a single pulse (e.g., as in a free induction decay). In other embodiments, the excitation may be delivered as multiple pulses (e.g., as in an inversion recovery experiment). The graph 464 may show voltage over time (e g., time domain data) data indicative of the magnetization signal. The time domain graph 464 may be converted (e.g., via a Fourier transform) into a frequency domain spectrum 466. In the frequency domain, information displayed may be analogous or similar to data provided by a CW EPR signal. In the frequency domain, information may include a number of peaks of the spectrum 466. The peaks in frequency may describe or indicate characteristics of the sample.

[0071] The graphs 468 illustrate aspects of a Fourier transform. Using pulse EPR data, a graph 470 may be generated that is or is indicative of an absorption curve. When the graph 470 is differentiated, the dispersion curve 472 may be generated. The graph 470 may be similar in shape to the graph 202.

[0072] In pulse experiments, Bi pulses may be applied to manipulate electron spins. FIG. 4B shows Bi pulses or excitations 462. The pulses are sinusoidal and may be short (e g., on a nanosecond timescale). Responsive to each of the Bi pulses, a transient evolution of the magnetization may be generated. The responses may indicate free induction decay (FID) and / or an echo, shown in graph 464. Magnetic relaxation of a sample of manipulated electrons may also be recorded. As shown in FIG. 4B, a Fourier transform may be taken of the response in the time domain to obtain the spectrum 466 of the spin relaxation. The Fourier transform may correspond to an absorption spectrum 470. From the absorption spectrum 470, a dispersion spectrum 472 may be generated. Pulse EPR may reveal dynamics of weakly coupled nuclei and relaxation that may not be visible with CW EPR. Additionally, pulse EPR may utilize a shorter acquisition time than that of CW EPR. In various embodiments, pulse EPR may utilize cryogenic temperatures for samples with short relaxation times. As described herein, the VCO 452 may be turned on and off to generate pulses, and may also be used as a sensor to sense the voltage corresponding to the magnetization decay. Further, when operated in a CW mode, the VCO 452 may be used to analyze a frequency shift of the voltage corresponding to the CW EPR signal.

[0073] In various embodiments, CW and pulse EPR experimental methods may be complementary. For example, CW EPR may offer greater sensitivity while pulse EPR may enable more detailed characterization of coherence and relaxation of the spin system. “EPR on a chip” systems may be used as portable and / or low cost biosensors. A chip may be, for example, a microchip or microfluidic chip capable of performing processes done in a laboratory on a smaller scale. For example, an EPR on a chip may be a biosensor used for a point-of-care healthcare application. Various advancements have been made in CW EPR spectrometer ICs. CW EPR spectrometer ICs may be compatible with small permanent magnets, therefore enabling portable use as, for example, point-of-care biosensors. Improvements include, for example, high-volume sensing arrays, high-field EPR, and chopping for enhanced sensitivity.

[0074] Referring now to FIG. 5A, use of a conventional cross-coupled LC oscillator 500 is shown, according to an example embodiment. The cross-coupled LC oscillator 500 may be used to generate B1pulses via oscillating AC current in the inductor (L). To perform pulse EPR, a sensor may generate short pulses (e.g., on the order of 10ns in length) at o>0to excite the sample. The sensor oscillator 500 may subsequently be turned off by disabling atail current source. The oscillator 500, when off, may detect the weak magnetization signal, which may be on the order of lOOnV to lOOuV. The magnetization is measured from the voltage induced across the LC tank by the magnetization. In various embodiments, a conventional LC oscillator may be associated with slow startup and ringdown of oscillations, which may limit achievable pulse lengths and may prevent detection of the free induction decay (FID) signal immediately after the pulse. Detecting the FID signal immediately after the pulse may be particularly important in an embodiment where samples have relaxation times on the order of nanoseconds or lower, and / or in embodiments in which it is desirable to reduce an EPR frequency to operate at lower Bowith small permanent magnets (e g , in which case ringdown takes longer for the same tank quality factor Q).

[0075] In some embodiments, to decrease startup and ringdown times, the Q of the resonator (e.g., the LC tank) may be decreased. However, a lower Q value may decrease an AC current, and therefore a corresponding B±magnitude, achievable with a given oscillator bias current. Furthermore, increased inductor parasitic resistance decreases the signal-to-noise ratio (SNR) of detection.

[0076] Referring now to FIG. 5B, a modified LC oscillator 550 is shown, according to an example embodiment. The modified LC oscillator 550 may include injection-locking 552 and a ringdown switch 554. The modified LC oscillator 550 may address limitations described above with respect to FIG. 5A. For example, to speed up the ringdown after the pulse once the oscillator has been turned off, the switch 554 may be included across the LC tank and may be closed after the pulse to briefly short-out the oscillator nodes and dissipate energy. The size of the switch device 554, and thus its on-resistance, may be optimized for the fastest decay in oscillations after the oscillator is turned off.

[0077] To boost startup and ensure a stable center frequency, the sensor oscillator may be injection-locked with a frequency reference. The frequency reference may have a low phase noise. Thus, injection-locking may also reduce the phase noise of pulses generated by the sensor oscillator, as the phase noise of an injection-locked oscillator may depend upon the noise of the injecting source. The reference may also be used as the local oscillator (LO) for downconverting the received magnetization signal to baseband (e.g., as will be described herein).

[0078] Referring now to FIG. 6A, a schematic 600 showing inductive and substrate coupling is shown, according to an example embodiment. Specifically, FIG. 6A shows inductive andsubstrate coupling from a reference oscillator 602 running at a frequency fo. The frequency fo is the same frequency as that used for EPR by a sensor oscillator 604. The coupling from a reference oscillator 602 may corrupt a pulse EPR signal when a sensor oscillator 604 is off, as will be described herein.

[0079] For a higher level of integration, a frequency reference may be realized as an on-chip oscillator (e.g., oscillator 602). The frequency may be set by enclosing the oscillator 602 in a phase-locked loop (PLL) with an external (e.g., off-chip) reference clock. However, when the oscillator 602 operates at the same frequency fo as the sensor 604 (e.g., the EPR frequency), substrate coupling from the reference 602 into the sensor devices and tank may obscure the weak magnetization signal while the sensor oscillator 604 is off. Furthermore, while an LC oscillator may be desirable for the reference due to a reduced phase noise, utilizing an LC oscillator for the reference may introduce inductive coupling into the sensor tank. As shown in FIG. 6A, inductive coupling may cause the EPR signal from the sample to be indistinguishable from the voltage on the sensor oscillator 604 resulting from coupling. For example, when the reference oscillator runs at the same frequency fo as the sensor 604, inductive and substrate coupling may cause an oscillation at fo to appear on the voltage of the sensor oscillator 604. The amplitude of such an oscillation may be larger than the actual pulse EPR magnetization signal, causing the actual signal to be undetectable. . Furthermore, a direct-conversion architecture using an LO frequency equal to the EPR frequency may experience self-mixing at a mixer. The self-mixing is caused by a frequency content at fo at a RF port of the mixer, which arises due to the LO frequency being equal to fo. The magnitude of the voltage induced in the sensor tank by any of the coupling mechanisms described herein may exceed the signal level of the EPR magnetization.

[0080] Referring now to FIG. 6B, a schematic 600 showing the reference oscillator 602 operating at a reduced frequency is shown, according to an example embodiment. To mitigate the challenges described with respect to FIG. 6A, the reference oscillator 602 may instead be operated at / REF= f0 / 2 (e.g., half the EPR frequency) as shown in FIG. 6B. Frequency doublers 606 may be utilized to inject into the sensor, and the reference may drive a 2x- subharmonic mixer (as will be described herein) to downconvert the magnetization signal to a baseband frequency, for example, zero, also known as direct current (DC). The operation of a pulse EPR spectrometer utilizing a subharmonic direct-conversion architecture 680 is shown in FIG. 6C. With the reference oscillator 602 operating at the 2x-subharmonic frequency, coupling of f0into the sensor arises from second-harmonic content in theoscillation of the reference oscillator. The amplitude of such content may be significantly smaller than that at the fundamental frequency. Similarly, f0content at an RF port of a mixer caused by the LO arises from asymmetries in the mixer, which can be minimized with techniques such as, for example, common-centroid layout.

[0081] Additionally, with double-balanced mixers, four LO phases (0°, 90°, 180°, 270°) may be used to downconvert a magnetization signal to baseband. In an example embodiment, the downconversion may be to DC. Quadrature detection may be utilized in this embodiment. As such, the mixer may utilize 8 LO phases to produce quadrature baseband outputs. This may be achieved by using a quadrature VCO (QVCO) for the reference oscillator and utilizing phase interpolators to generate the 8 LO phases from the 4 reference phases, as will be described herein with respect to FIGS. 7-15.

[0082] Referring now to FIG. 7, a block diagram 700 of a dual-mode EPR IC is shown, according to an example embodiment. FIG. 7 illustrates an implementation of circuit blocks for realizing the subharmonic direct-conversion architecture for pulse EPR described above with respect to FIG. 6B, in addition to incorporating compatibility with CW-mode operation. The dual mode EPR IC includes a reference quadrature VCO (QVCO) 702, a subharmonic injection buffer 704, one or more sensor cells 706, each including a sensor VCO 707, phase interpolators 708, an RFOTA 710, a 4-stage VGA 712, one or more subharmonic mixers 714, and a quadrature FM demodulator 716. The components of FIG. 7 are described in greater detail with respect to FIGS. 8-15. Further, the sensor cells 706 may be similar to the oscillator 452 described with respect to FIG. 4B.

[0083] Generally, the components 702-716 may generate pulses used in pulse EPR. The pulses may be used to excite a sample (e.g., a sample containing the electrons that are excited). The pulses may subsequently be turned off to remove the pulse energy from the resonator such that the VCO 707 may sense the generated pulses. Particularly, the QVCO 702 may produce a reference signal that injection locks a sensor cell 706 using the subharmonic injection buffer circuit 704 to generate the pulses appropriately. The digital state machine 718 may regulate and / or control timing of the generated pulses. For example, pulse widths, pulse durations, spacing between pulses, etc. may be accurately controlled such that desired pulse sequences are generated and an accurate pulse EPR signal is elicited from the sample.

[0084] Within the sensor cell 706, the VCO 707 may sense pulse EPR signals produced by samples positioned above the inductor of the VCO. These signals include, for example, freeinduction decays (FIDs) and echo signals. The generated pulses energize a sample by and tilt the magnetization of unpaired electrons with respect to FIG. 3. The precession induces a voltage in the VCO 707 at the same frequency that was used to energize the sample. For example, when the magnetic field has a value of 0.5T, the frequency may be 14GHz. The frequency value may depend on the value of the magnetic field. The strength of the magnetization of the sample may be smaller than the magnetic field strength Bi of the excitation pulse (e.g., orders of magnitude smaller). As such, a low noise receiver (e.g., the RFOTA 710) may be utilized to amplify the magnetization signal produced by the sample to enable detection of the signal. The RFOTA 710 may be a low noise amplifier or amplifier chain that amplifies a same frequency as the frequency emitted by the sample. For example, the RFOTA 710 may be three RFOTAs that are linked or chained to amplify the amplitude of a voltage of the signal coming from the sensor cell 706

[0085] Additionally, the frequency of the RFOTA 710 may be a radio-frequency value (e.g., 14GHz). The frequency may be downconverted to a baseband or low frequency to permit additional signal amplification and subsequent analog to digital conversion. Specifically, the subharmonic passive mixers 714 may multiply the signal emitted from the sample (after amplification with the RFOTA 710) with a local oscillator signal (e.g., generated by the reference QVCO 702), and the resulting multiplied signal may correspond to the magnetization signal shifted down in frequency. The signal may be downshifted to a DC or near DC value to permit detection of the amplified and downconverted signal. The VGA 712 may further amplify the signal once it has been downconverted by the mixers 714. The quadrature FM demodulator 716 may be used to enable CW EPR detection.

[0086] Further, the reference QVCO 702 may operate at a subharmonic frequency with respect to the EPR frequency. That is, the sensing oscillator (e.g., the VCO 707) may be running at a first frequency (e.g., 14GHz) and is turned on and off to produce pulses having a center frequency equal to the first frequency (e.g., 14GHz). The reference QVCO 702 may run at a second frequency that is subharmonic relative to the first frequency (of the VCO 707). The subharmonic mixer (e.g., the QVCO 702) may mix the signal a certain number of times until the frequency is equal to the desired baseband frequency (e.g., zero, also referred to as DC). For example, when the subharmonic mixer is operating at 7GHz and the VCO 707 is operating at 14GHz, the subharmonic mixer may mix the signal twice to downconvert it to DC. This process may occur because of the difference in orders of magnitude between thefree induction decay signal (e g., on a nanovolt to microvolt scale) and the excitation pulse signal (e.g., on a volt scale). If the reference QVCO 702 and the sensor VCO 707 operated at the same frequency, on-chip coupling (e g., through a silicon substrate, magnetic coupling resulting from coils on the chip acting as resonators for the oscillators, etc.) the sensor cell 706 may parasitically pick up a larger signal from the QVCO 702 that saturates and drowns out the sample signal that is to be detected. As such, the lower frequency of the QVCO 702 may be doubled and used to inject the VCO 707 with the subharmonic injection buffer circuit 704, and mixed down twice to downconvert the signal to a DC signal. The subharmonic buffer 704 is turned off when not in use to avoid the 14 GHz signal remaining on the chip.

[0087] FIG. 8 illustrates a portable spectrometer 800, according to an example embodiment. The portable spectrometer 800 may include one or more printed circuit boards (PCBs). The portable spectrometer 800 may include an EPR IC 802, a first phase locked loop (PLL) 804 having a phase-frequency detector and charge pump (PFD+CP) used for pulse EPR, a second phase locked loop 806 having a PFD+CP used for CW EPR, first and second analog-to-digital converters (ADC) 810 and 812, a field programmable gate array (FPGA) 814, a direct-digital synthesis (DDS) block 808 to generate frequency modulation, and a lock-in amplifier 820. In various embodiments, the dual-mode EPR IC and the portable spectrometer may be part of the same system. The dual mode EPR IC may utilize a subharmonic direct-conversion architecture. The IC may further utilize on-chip digital pulse sequencing. The QVCO 702 in the IC may include a varactor which is controlled by the first PLL 804, and the sensor VCO 706 may include a varactor which is controlled by the second PLL 806.

[0088] The first phase locked loop (PLL) 804 and the second phase locked loop 806 may keep different oscillators and the chip 802 running or operating at a known frequency. When the chip 802 is operating in a pulse mode, the first PLL 804 may be turned on and the second PLL 806 may be turned off. Conversely, when the chip 802 is operating in a CW mode, the second phase locked loop 806 and the circuit 808 may be turned on. When the first PLL 804 is turned on, an external programmable oscillator 818 may be set such that the first PLL 804 causes the QVCO 702 to operate at a frequency half of that of the VCO 707for performing pulse EPR. The oscillator may also be tunable such that the frequency can be swept over a certain range and EPR can be performed at different frequencies. The QVCO 702 may injection lock a sensor oscillator 706 during the generation of pulses. A sensor oscillator 706 may turn off and sense the magnetization signal produced in response to the generated pulses.In some embodiments, the spectrometer 800 may include a plurality of sensor cells 706 (e.g., two sensor cells). One sensor cell 706 may be used at a time, such that at any given time, one sensor cell 706 is injection-locked by the QVCO 702 to produce pulses and / or turned off to sense magnetization. The output of the pulse detection chain (e.g., the output from the RFOTA amplifiers 710, the subharmonic mixer 714, and the VGA 712) may then be received by the ADC 810 on the PCB The first ADC 810 may be configured specifically to convert signals from pulse EPR. The second ADC 812 may be configured to specifically convert signals from CW EPR. The ADC 810 may receive the signals and record the voltage signals over time. The ADC 810 may store and digitize the signals and transmit them for processing. The ADC 810 may receive four pulse outputs. Specifically, the chip 802 may output two signals, each having a differential (e.g., positive and negative component), causing the ADC 810 to receive four separate signals. The two differential signals may be an in-phase (I) signal and a quadrature (Q) signal. Both the I and Q signals may be needed to record the magnetization signal from pulse EPR. In various embodiments, the chip 802 may output any number of signals that are received by the ADC 810.

[0089] When the chip 812 is operated in a CW mode, the first PLL 804 and the QVCO 702 are turned off, leaving the sensor 706 turned on. The second PLL 806 is used to sweep the frequency (e.g., as described in FIG. 2). During a CW mode operation, the direct digital synthesis (DDS) block 808 includes a DDS 816. The DDS 816 generates a sine wave of a certain frequency. The frequency may be, for example, 100kHz (e.g., the modulation frequency of FIG. 2). The sine wave may be an input frequency modulation sine wave used to perform phase detection. The sine wave generated by the DDS 816 may be input into a modulation varactor in the sensor 706. On output, there is a lock-in amplifier 820 on the PCB into the second ADC 812. The sine wave generated by the DDS 816 may be used as the reference signal for the lock-in amplifier 820. The FPGA 814 is a programmable logic board described and shown in greater detail with respect to FIG. 21.

[0090] Referring now to FIG. 9A, a schematic 900 of a sensor VCO 902 (e.g., the sensor cell 706 and the VCO 707) and a subharmonic injection buffer circuit 904 (e g., the subharmonic injection buffer 704) of the dual -mode EPR IC of FIG. 7 is shown, according to an example embodiment. As described above, the QVCO 702 produces a subharmonic frequency (e.g., half of the EPR frequency). The subharmonic injection buffer 904 illustrates a process for doubling the subharmonic frequency of the QVCO 702 to achieve the EPR frequency f0(e.g.,how a 7GHz frequency is doubled to achieve the 14GHz EPR frequency), thereby allowing the subharmonic frequency to be injected into the oscillator 707. Frequency doublers utilize the 4 reference oscillator phases (e.g., I+, I-, Q+, Q-) to generate a differential signal at fQto be injected into the sensor. The injection circuitry 904 is disabled once the pulse ends to ensure that no large signals at f0are present on the IC during detection. High-voltage tolerant devices are used in the sensor to raise the differential swing on the tank nodes to, for example, 5Vpp, thereby increasing an achievable B1field strength. For CW operation, two sets of varactors are included in the sensor LC tank, one for sweeping the EPR frequency and the other for applying frequency modulation.

[0091] The sensor cell 902 includes an injection input block 906. The injection input block 906 feeds into a cross coupled oscillator 908, causing the injection of the frequency f0to occur. Additionally, as shown, the injection input block 906 may include one or more transistors. For example, as shown in FIG. 9A, the injection input block 906 may have three sets of transistor pairs to realize a controllable injection strength with three bits of resolution. Each of the transistor pairs may be turned on and off to vary a strength of the injection. When all pairs are turned off, injection into the VCO 707 is turned off, particularly when the VCO 707 is turned off to detect the generated pulses.

[0092] As shown, the VCO 707 may include an inductor 910, a modulation varactor 912, and a sweep varactor 914. The modulation varactor 912 may be implemented as one or more varactors (e.g., metal-oxide-semiconductor varactors). The sweep varactor 914 may be implemented as one or more varactors. The inductor 910 may be the coil described above with respect to FIGS. 4A and 4B that senses the magnetization of the electrons in a sample. The modulation varactor 912 and the sweep varactor 914 are inputs to which the CW EPR components described with respect to FIG. 8 (e.g., the second PLL and the DDS block) are connected. This allows the frequency to be swept and modulation to be applied. The VCO 707 may also include an enable switch 916. The enable switch 916 may be switched on briefly to dissipate energy from the VCO 707 after a pulse is generated. This may allow a pulse to have a quick fall time, and no energy is left in the resonator due to the generated pulse. The signal coming from the sample may then be able to be measured with a short (e.g., less than a threshold value) amount of dead time. Dead time may be the time between when the pulse is sent and when the free induction decay (FID) can be detected. If energy from the pulse cannot be quickly pulled from the resonator after the pulse, because the pulse is orders ofmagnitude larger in amplitude than the expected response magnetization signal, the receive may be saturated and the FID may be unable to be detected until the energy from the pulse is out of the resonator. Thus, the switch 916 allows the energy to be quickly removed from the resonator once the pulse has been sent, so that energy from the sample can be detected.

[0093] For pulse experiments (e.g., pulse EPR), the IC may generate short Bi pulses at an initial frequency (coo) and subsequently detect a weak magnetization signal. In various embodiments, an oscillator (e.g., the VCO 707) may be used for the dual-mode EPR IC to form pulses for the pulse mode EPR. The oscillator 707 may switch on and off to form the pulses. The oscillator 707 may also perform a frequency sweep via tuning inputs to enable the CW EPR mode of the dual-mode chip. In various embodiments, the pulses may be on an order of, for example, 10 nanoseconds for a wide pulse excitation bandwidth. An on-chip inductor-capacitor (LC) resonator, including the inductor 910 and the capacitors in the oscillator, may be used as a sensing cell to form the pulses and / or sense relaxation (e.g., detect pulse signals) through an induced voltage. In some embodiments, an LC VCO may be used as the sensing cell and may be switched on and off to generate pulses of an oscillating magnetic field. Timing circuitry may be included in the chip to control a duration of the pulses and lengths of time between multiple pulses. LC VCOs and / or other oscillators may have long (e.g., slow) startup and turn-off or ringdown times that can exceed the short pulse lengths (e.g., on the order of 10ns) that may be used for high excitation bandwidth, particularly for a high-Q tank, thus generating pulses that may not be conducive to pulse EPR on a chip.

[0094] Therefore, circuit techniques may be introduced to a sensing oscillator cell to enable both pulse and CW EPR operation in one sensing cell. For example, a switch 916 may be introduced in the sensor VCO tank to mitigate ringdown after a pulse is ceased. In various embodiments, the switch 916 may briefly short out the tank to dissipate energy after a pulse. The switch 916 may briefly close to rapidly dissipate stored energy, thus enabling detection of free induction decay (FID) signals without dead-time, as discussed herein, to decrease the startup time of the oscillator / VCO, the VCO may be injection-locked with an on-chip reference oscillator. For example, injection-locking may boost or increase a startup speed of the oscillator. Decreasing starting and stopping (e.g., startup and ringdown, respectively) times may allow accurate pulses of an oscillating magnetic field to be generated. The reference oscillator may remain on even when the sensor is off. Therefore, as described herein, the reference oscillator (e.g., QVCO 702) may be operated at fd2 (e.g., 7GHz when fois 14GHz) to avoid parasitic substrate coupling. The parasitic substrate coupling may obfuscate a weak EPR signal. In various embodiments, the reference oscillator may be implemented as a quadrature LC VCO. In various embodiments, frequency doublers may be activated during the pulse to inject into the sensor VCO. The injection circuitry of FIG. 9A may be disabled once the pulse from the sensor VCO ends. This may ensure that no large signals at coo are present in the IC during pulse detection

[0095] Referring now to FIG. 9B, a graph 950 is shown that illustrates a pulse generation with and without injection locking, according to an example embodiment. For example, graph 950 shows an AC current in the sensor inductor for a pulse (e.g., 10ns) with the VCO free- running (shown as waveform 952) versus utilizing injection-locking and the ringdown switch (shown as waveform 954). As shown in FIG. 9B, waveform 954 demonstrates improved startup and ringdown relative to the waveform 952.

[0096] Referring now to FIG. 10, a reference oscillator 1000 is shown, according to an example embodiment. The reference oscillator 1000 may be the same as or similar to the reference quadrature VCO (QVCO) 702. In various embodiments, the reference QVCO 1000 may be a parallel-coupled LC QVCO. The reference QVCO 1000 may operate or run at a frequency half of the EPR frequency. In some embodiments, the QVCO 1000 may operate at any other frequency value that is less than the value of the EPR frequency. The circuits 1002 and 1004 are connected to each other such that the circuits operate in time quadrature. The 1+ and I- voltages shown in each circuit 1002 and 1004 are 90 degrees off in phase relative to the Q+ and Q- voltages. Two circuits may be utilized to generate two signals, each at half of the EPR frequency. The quadrature signals may enable frequency doubling and subharmonic downconversion using the subharmonic mixer.

[0097] The reference QVCO 1000 may utilize figure-8 inductors to help suppress inductive coupling of common-mode second-harmonic content into the sensor cell(s) or sensor tanks. In pulse mode, the QVCO 1000 may be incorporated into a phase locked loop (PLL) The reference QVCO 1000 may be incorporated via, for example, an external phase-frequency detector and charge pump (PFD / CP), and on-chip frequency dividers. The on-chip frequency dividers may clock on-chip digital pulse-forming logic.

[0098] Referring still to FIG. 10, to help suppress the coupling of second-harmonic content to the sensor arising from the flow of common-mode current through the center-tappedinductors, figure-8 inductors are used (e.g., as shown in FIG. 16A and 16B). The figure-8 inductors may occupy a smaller area compared to a conventional symmetric inductor with the same inductance.

[0099] FIG. 11 shows a circuit diagram of active tunable phase interpolators 1100. The phase interpolators 1100 may be the same as or similar to the phase interpolators 708 of FIG. 7. The phase interpolators 1100 may generate local oscillator signals needed for the subharmonic mixer 714. FIG. 11 includes a diagram 1102 illustrating the concept behind operation of the phase interpolators 1100. For example, as shown, a phase interpolator 1103 has a first associated sine wave 1104 having a phase of 0 degrees and a second associated sine wave 1106 having a phase of 90 degrees. By combining the sine waves 1104 and 1106, a resulting output sine wave 1108 may have a phase of 45 degrees. As shown in circuit 1110, eight phase interpolator circuits may be included (e.g., to make up the phase interpolator 708). The eight phase interpolators may take four phases (e.g., 0 degrees, 90 degrees, 180 degrees, and 270 degrees) from the QVCO (e.g., the I+, I, Q+, and Q- signals shown at block 1112. From those four phases, the phase interpolators may generate eight phases, each spaced 45 degrees apart (e.g., 0 degrees, 45 degrees, 90 degrees, 135 degrees, 180 degrees, 225 degrees, 270 degrees, and 315 degrees). That is, 45°-offset phases may be generated by adding 0° and 90°-phase currents with active circuits, thereby avoiding area costs and potentially higher mismatch of, for example, passive polyphase filters.

[0100] Digital tuning of each interpolator is provided to account for process variations. The outputs of each interpolator may also be buffered with digital inverters to drive the LO ports of the subharmonic mixer. The eight signals may be used for subsequent conversion by the subharmonic mixer. A plurality of local oscillator (LO) phases at roo / 2 may be generated from the QVCO 702 (or 1000) with the active tunable phase interpolators 1100. The 8x LO phases may be utilized at the subharmonic mixer stage. The subharmonic mixer stage will be described in greater detail with respect to FIG. 13.

[0101] FIG. 12A shows a tuned amplifier stage 1200, according to an example embodiment. The stage 1200 may be part of a radiofrequency operational transconductance amplifier (RFOTA). That is, the RFOTA may include a plurality of amplifier stages 1200. For example, as shown in FIG. 7, the RFOTA 710 includes three stages. The stage 1200 may be a low noise tuned amplifier (e.g., an RF amplifier). The RFOTA tanks may be shorted while the pulse is active to prevent sensor coupling from imposing dead-time. As such, the tuned amplifier stage1220 may include a disable switch 1200. In standard RFOTAs, the amplifier may have a resonator 1204 that includes an inductor and a capacitor that is able to increase amplification at a frequency range of interest. However, the resonator 1204 may also store energy. Thus, while the generated pulse is on (e.g., a sensor cell is outputting a pulse), the resonator 1204 is saturated (e.g., energy is built up from the pulse). When the energy is not emptied from the resonator 1204, the time it takes for the energy to decay on its own (e g., without intervention) may be too long, causing the RFOTA to be unable to properly amplify a low amplitude signal that is being received by the RFOTA. Thus, the resonator 1204 of each stage 1200 of the RFOTA may be shorted using the switch 1202 to “reset” the stage and prepare the stage to properly amplify the signal. The switch 1202 may short the resonator 1204 of each stage 1200 during the pulse. The switch 1202 may also short the resonator 1204 after the pulse is sent. As such, the RFOTA may amplify the EPR pulse signal generated by the sensor VCO. The RFOTA may be connected to the sensor tank (e.g., as shown in FIG. 7).

[0102] FIG. 12B illustrates an RFOTA 1250, according to an example embodiment. The RFOTA 1250 amplifies the pulse EPR signal before the mixer, providing a voltage gain of, for example, 22.1dB minimum with a 3dB bandwidth of 860MHz over a tuning range of 1.9GHz. A first stage 1252 of the RFOTA 1250 may include two active cores (e.g., one for each sensor), to select between the sensors. To prevent coupling of f0content into the RFOTA tanks while the sensors are generating pulses, bias currents of each stage may be disabled and switches in the first two stages may short out the tanks (e.g., as discussed above with respect to FIG. 12A).

[0103] FIG. 13 illustrates a circuit diagram 1300 of the subharmonic mixer, according to an example embodiment. Though reference is made throughout to a subharmonic mixer, it should be understood that any type of mixer (e.g., a non-subharmonic mixer) may be utilized in the spectrometer. As shown in FIG. 13, a subharmonic 1300 mixer may comprise two mixer circuits (e.g., two chains of mixers), 1301a and 1301b. Each mixer circuit 1301a and 1301b may correspond to each quadrature output I and Q. That is, the mixer circuit 1301 may correspond to the I quadrature output and the mixer circuit 1301b may correspond to the Q quadrature output. Each mixer circuit 1301a and 1301b is comprised of two double-balanced passive mixer stages cascaded or chained such that the circuits connect. For example, in mixer 1301a of the circuit 1300, a mixer stage comprising transistor pairs 1306a and 1306c in section 1306 are connected to a mixer stage comprising transistor pairs 1310a and 1310c in section 1310. Likewise, in mixer 1301b, a mixer stage comprising transistor pairs 1306b and1306d in section 1306 are connected to a mixer stage comprising transistor pairs 1310b and 13 lOd in section 1310. The EPR frequency f0may be first downconverted to f0 / 2 and odd harmonics, and then to DC. For example, as shown in plot 1304, prior to downconversion, one frequency fQexists. As shown in plot 1306, the frequency is downconverted to0 / 2, 30 / 2, 50 / 2, etc. As shown in plot 1312, the signal is downconverted to a DC value (e.g., 0V) and even harmonics of / 0 / 2 (e g., f0, 2f0, etc.). The mixers 1301a and 1301b may be highly linear, low-noise, and cause a voltage conversion loss of, for example, 4.1 dB.

[0104] By cascading the two circuits, the RF signal (e.g., the amplified pulse signal) can be downconverted to a baseband (e.g., DC) value. As shown, the circuit 1300 includes I and Q outputs used to perform quadrature detection. The circuit 1300 may include n separate pairs of outputs (e.g., two pairs of outputs) mixed by signals 90 degrees apart. The baseband frequency may be 0Hz DC. As such, both I and Q may be needed to completely read out or detect the signal.

[0105] After the pulse signal passes through the RFOTA of FIG. 12A, one or more subharmonic mixers 1300 (e.g., passive double-balanced 2x-subharmonic mixers) may downconvert the signal to DC with a quadrature output. For example, the mixers may downconvert a 14GHz pulse signal to a DC signal. The subharmonic mixers may utilize the plurality of LO phases generated by the phase interpolators of FIG. 11. The use of a subharmonic mixer, with a reference oscillator or QVCO running at one-half the EPR frequency providing the local oscillator signals to the mixer, may avoid parasitic coupling of the EPR frequency coo into the sensors while sensing the pulse signal. Parasitic coupling may decrease performance (e.g., worsen the minimum detectable pulse EPR signal level) of the EPR system. The subharmonic mixer may downconvert the EPR signal, which is oscillating at an EPR frequency, to a baseband frequency. The baseband frequency may be, for example, 0Hz. In various embodiments, the reference oscillator used to inj ection-lock the sensor during pulses may be used to generate local oscillator signals for the subharmonic mixer. Generation of the local oscillator signals may be done through the phase interpolators described with respect to FIG. 11.

[0106] FIG. 14 illustrates a variable gain amplifier (VGA) stage 1400, according to an example embodiment. A chain of VGA stages may follow each mixer channel of FIG. 13. Further, output buffers (shown in FIG. 8 as buffers 718) may follow the chain of VGA stages. The gain (e.g., multiplication factor) by which the voltage is amplified may be tunable (e.g.,using resistors having varying resistance values). The VGA 1400 may include one or more (e.g., four) resistive-feedback stages. A single stage is shown in FIG. 14. The gain of each stage is adjustable via switched resistors in the feedback path. To compensate for DC offsets, a bank of current sources may be provided to selectively pull current out of the output node of each VGA 1400 stage in the chain. Each VGA 1400 may be followed by a pseudodifferential output buffer. Including the buffers, the gain of each chain may be adjustable (e.g., from 21dB to 55dB with a 67MHz 3dB bandwidth at a maximum setting).

[0107] A quadrature demodulator 1500 is shown in FIG. 15, according to an example embodiment. The quadrature demodulator 1500 may be implemented on the IC and used for the CW EPR mode of the dual-mode IC. The quadrature demodulator 1500 may be the same as or similar to the quadrature demodulator 716 of FIG. 7. For example, the quadrature demodulator 1500 may be fully integrated within (e.g., embedded on) a microchip on which pulse and CW EPR are performed. When the IC is operating in the CW mode, the PLL 806 may be on and may be closed using varactors in the sensor cell(s) (e.g., by sweeping the varactor 914) with respect to FIG. 8). Closing the PLL via varactors in the sensor cell(s) may allow an EPR frequency fo to be swept via the external frequency reference, and the QVCO 702 may be disabled. Frequency modulation (FM) may be applied to a second set of varactors (e.g., the modulation varactor 912) with a narrower tuning range. Demodulation may be performed, for example, by the on-chip quadrature demodulator, allowing a wide frequency sweep range. The demodulator may include an LC phase discriminator 1502, a mixer 1504, a lowpass filter 1506, one or more operational transconductance amplifier (OTA) stages 1508 (e.g., three cascaded OTA stages), and a buffer 1510, as shown in FIG. 15. The demodulator may further include one or more OTA cores 1512. For example, the quadrature demodulator 716 may include three cascaded OTA stages 1508, and each OTA stage (and the output buffer) may have an associated OTA core 1512.

[0108] In CW mode, the PLL 806 is closed around the sensor, using a set of varactors (e.g., the sweep varactor 914) in the sensor tank. This allows the EPR frequency to be swept by varying the external frequency reference. The sensor VCO 707 is operated at a reduced current bias for lowerto avoid saturating the sample, and FM is applied via the smaller set of varactors, which have a narrower tuning range. To prevent the PLL 806 from opposing and canceling out the FM input, the PLL 806 bandwidth may be kept below the modulation frequency. Demodulation is performed by an on-chip quadrature demodulator (e g., the quadrature demodulator 716), which allows a wide frequency sweep range. The demodulatorinput is taken from the RFOTA output in order to buffer and limit the sensor output to fullswing. The demodulated signal after the filter is a sinusoid whose amplitude is proportional to the frequency deviation in the sensor. It is amplified by a chain of 3* capacitive-feedback OTA stages and an output buffer.

[0109] Referring now to FIGS. 16A and 16B, an inductor 1600 and an inductor 1650 are shown, according to an example embodiment. The inductor 1600, shown in FIG. 16A, may occupy a larger area relative to the inductor 1650, shown in FIG. 16B. The inductor 1650 may be a figure-8 inductor as described above with respect to FIG. 10. The figure-8 inductor may be utilized in on-chip EPR where a smaller surface area may be desired. Common-mode current flowing in the inductor 1600, as shown in FIG. 16A, may have stronger magnetic coupling to sensor inductors compared to the same current flowing in the inductor 1650. The common-mode current may have a frequency equal to an EPR frequency (e.g., 14GHz). The figure-8 inductor may be utilized in on-chip EPR where a smaller magnetic coupling may be desired.

[0110] Referring now to FIG. 17, a pulse-sequencing state machine 1700 is shown, according to an example embodiment. On-chip digital pulse sequencing logic controls the sensor and RFOTA for pulse operation. As shown in FIG. 17, four counters 1702-1708 are clocked by clocks divided down from the reference oscillator (e g., the QVCO 702). The counters 1702- 1708 may generate a sensor bias and injection enable signals (e g., using counter 1704 with 10 bits and a frequency of REF / 4) and the ringdown switch control signal (e.g., using the counter 1702 with 6 bits and a frequency of / REF / 4). The counters 1702-1708 may generate delays between pulses in a sequence (e.g., using the counter 1706 with 1 Gbits and a frequency °f REF / 16) The counters 1702-1708 may also set a shot repetition time (SRT) (e.g., using the counter 1708 with 20 bits and a frequency of / REF / 64). The SRT may be a rate at which sequences of one or more pulses are repeated for signal-averaging.

[0111] The RFOTA is enabled after each pulse, with programmable delays for the enabling of each stage and releasing of the tank shorting switches after the end of the pulse. With the reference at, for example, 7GHz, the pulse sequencing achieves resolutions of, for example, 600ps in pulse lengths and 2.3ns in delays between pulses within a sequence, and pulse lengths up to, for example, 585ns and delays up to 150ps are achievable. The minimum SRT is not limited by hardware. As such, SRT ~ lp.s may be used for room-temperature experiments with relaxation times ~ 100ns. In pulse mode, the reference oscillator isincorporated into a PLL via an external phase-frequency detector and charge pump (PFD / CP) and on-chip frequency divider outputs with respect to FIG. 8. A trigger input to the IC is asserted to begin generating pulse sequences. To interface to external data acquisition circuitry, the IC provides a REF / 16 clock and a synchronization signal that is synchronized with the pulse output being active. FIG. 18 shows a timing diagram 1800 illustrating the timing of enable signals for the sensor and the RFOTA, according to an example embodiment.

[0112] In various embodiments, a prototype spectrometer IC may be fabricated. The following paragraphs describes an exemplary prototype spectrometer. FIG. 19 shows an exemplary chip 1900. For example, the chip 1900 may illustrate an example implementation of the above-de scribe EPR components on a microchip to perform on-chip dual EPR. For example, the spectrometer 1900 may be fabricated in a bulk 65nm CMOS process. The spectrometer may occupy a given area of the chip. For example, the spectrometer may occupy an area of 2x1.5mm2. In various embodiments, a plurality of sensing cells (e.g., sensing VCOs), shown as sensors 706a and 706b, may be on the chip 1900. For example, the sensor 706 (and VCO 707) may be implemented as two sensors 706a and 706b on the chip 1900 (e.g., EPR sensor A and EPR sensor B in FIG. 19). Each of the two sensors may be different in size. For example, a first EPR sensor (e.g., EPR sensor A, sensor 706a) may include an on- chip inductor having a diameter of 200pm and a second EPR sensor (e.g., EPR sensor B, sensor 706b) may include an on-chip inductor having a diameter of 140pm. The reference QVCO may be tunable from, for example, 6.2 to 7.6GHz, the sensor VCOs may be tunable from, for example, 12.9 to 15.1 GHz (Sensor A) and 12.7 to 15.1 GHz (Sensor B), and the CW demodulator may be tunable from, for example, 12.8 to 14.9GHz. The range of achievable frequencies may be, for example, 12.8 to 14.9GHz for CW mode and 12.8 to 15GHz for pulse mode. The minimum power consumption of one sensor VCO for CW operation may be 2.1mW (Sensor A).

[0113] The inductors may be responsible for the sensing, pulse generation, and / or CW generation in each sensor VCO. An implemented fo may correspond to a given Bo. For example, the sensing frequency fo may be equal to 14GHz, corresponding to a Bo equal to approximately 0.5T. In various embodiments, Bo may be a 0.5T biasing field with a portable magnet. An array of two independent sensor cells may be implemented with different inductor diameters. For example, a first inductor diameter may be 140pm and a second inductor diameter may be 200pm. A variety of inductor diameters may enable multiplexedexperiments and / or a choice of larger sensing area versus higher Bi magnitude. On-chip pulse control may achieve one or more resolutions. For example, resolutions may of 600ps in pulse length and 2.3ns in delays between pulses in a sequence may be achieved. In various embodiments, the pulse length or pulse timing resolution may be comparable to other pulse systems. A repetition rate for averaging may be limited by relaxation properties of the sample. This may thereby allow shot repetition times (SRT) below a given threshold (e.g., 1 ps for room -temperature experiments) with spin-lattice relaxation times (Ti) on the order of, for example, 100ns. Oscillator performance may be measured using divided-down digital outputs generated from the reference QVCO and sensor cells. In various embodiments, divisors of 16 and 256 may be used for the reference QVCO and sensors, respectively. In various embodiments, the QVCO may achieve a given phase noise when locked by the PFD / CP. For example, phase noise may be 99.7dBc / Hz at a 100kHz offset when locked by the PFD / CP at 7GHz. The EPR sensor B may achieve 85.5dBc / Hz at alOOkHz offset from 13.93GHz (e.g., while free-running) and 92.8dBc / Hz at a 100kHz offset from 14GHz (e.g., while injection- locked during the generation of a pulse). The values may not include potential degradation introduced by the dividers and buffers. FIG. 20 shows charts 2002 and 2004, each illustrating a power consumption breakdown. For example, chart 2002 illustrates a power consumption breakdown for the chip 1900 operating in the pulse EPR mode, and chart 2004 illustrates a power consumption breakdown for the CW EPR mode.

[0114] Referring now to FIG. 21, an example implementation of the EPR system is shown, according to an example embodiment. As shown in FIG. 21, the system includes a permanent magnet 2102, an FPGA 2104, and a data acquisition PCB 2106 including an EPR IC 2108. In various embodiments, the system may further include a computer. The IC 2108 is encapsulated and mounted onto the PCB 2106. The PCB may interface with a portion of the permanent magnet 2102. Additionally, the FPGA 2104 may control different aspects of the EPR system For example, the FPGA 2104 may set up the IC 2108, configure digital logic in the IC 2108 for, for example, pulse timings, cause the IC 2108 to generate pulses at specific times, etc. The IC 2108 may configure or determining pulse timings, and the FPGA 2104 may indicate a delay time between pulses, a duration of each pulse, etc. The FPGA 2104 may also program different components of the EPR system, such as phase locked loops, programmable oscillators, ADCs, etc.EXAMPLE EXPERIMENTAL RESULTS

[0115] Referring now to FIGS. 22A-32, example experimental results are shown and described. It should be understood that the data values described herein are exemplary only and are not meant to be limiting.

[0116] Referring now to FIG. 22A, a graph 2200 showing an oscillator phase noise is shown, according to an example embodiment. Graph 2200 shows an oscillator phase noise measured through a 16x divider output of the reference QVCO, locked by a PLL (e.g., the PLL 804) at 7GHz. Digital divided-down frequency outputs with divisors of 16 and 256 for the reference QVCO and sensors, respectively, are used to measure the oscillator performance. The reference achieves a phase noise of -99.7dBc / Hz at 100kHz offset when locked by the PFD / CP at 7GHz. Sensor A achieves -82.8dBc / Hz at 100kHz offset from 13.932GHz (free- running) and -92.8dBc / Hz at 100kHz offset from 14GHz (injection-locked during pulse). Sensor B achieves -85.5dBc / Hz at 100kHz offset from 13.929GHz (free-running) and - 92.8dBc / Hz at 100kHz offset from 14GHz (injection-locked during pulse).

[0117] Referring now to FIG. 22B, a graph 2250 showing an oscillator phase noise is shown, according to an example embodiment. Graph 2250 shows an oscillator phase noise measured through a 256x divider output of a sensor VCO. Specifically, the reduction of sensor phase noise during excitation (generation of the pulse) by injection locking with the reference oscillator (e.g., the QVCO) is shown in FIG. 22B. For example, the phase noise with injection locking, shown by line 2252, is generally less than the phase noise without injection locking, shown by line 2254. Thus, the reference oscillator is shown to be a stable reference and injection locking with the QVCO can reduce phase noise power of a sensor VCO. The IC may consume a total of amount of power when sensing in pulse mode and / or in CW mode. For example, the IC may consume 98mW when sensing in pulse mode and 39mW in CW mode (e.g., as shown in FIG. 20). In various embodiments, in CW mode, power consumption may be less than that of CW-only or single-mode EPR ICs.

[0118] In various embodiments, the spectrometer may be incorporated into a portable sensing platform with the PFD / CPs, CW FM source, ADCs, and an FPGA (e.g., the spectrometer of FIG. 8). In various embodiments, characterization of the spectrometer may be performed with various substance. For example, characterization of the spectrometer may be performed with solid 2,2-diphenyl-l-picrylhydrazyl (DPPH) powder. An electromagnet of a spectrometer (e.g., a Bruker EMX) may be used to provide Bo. An FID of a sample may be obtained. The sample may have a given volume of, for example, ~600 pm3. The sample may be obtainedusing a sensor of the IC with a given pulse and length. For example, the sample may be obtained using a sensor with a 140pm inductor (e.g., sensor 706b shown in FIGS. 7 and 19) with a 7t / 2 pulse of 23ns length. In some embodiments, Bo may be equal to 4992.5G, the SRT may be 1ms, and 1,000 averages may be used

[0119] Referring now to FIG. 23, a diagram 2300 illustrating the concept of FID is shown, according to an example embodiment. In FID, a single pulse is applied to tip the net magnetization away from equilibrium into the transverse plane. The magnetization precesses about the z-axis at the Larmor frequency coo (or / o). Viewed from a reference frame rotating at coo, the pulse rotates the magnetization away from the z-axis. The combination of Larmor precession with the rotation results in nutation when viewed in a stationary frame. A pulse of ideal length (a nil pulse) tips the magnetization by 90 degrees, so it lies entirely in the transverse plane. Afterwards, due to variations in the environment of the unpaired electrons, some spins will precess faster and some will precess slower than coo. This dephasing causes the magnitude of the net magnetization in the transverse plane to decay exponentially, producing the FID signal.

[0120] FIG. 24A shows a graph 2400 indicating a measurement of the FID of the sample, according to an example embodiment. The graph 2400 includes time-domain data measured with an ADC (e.g., the ADC 810) from baseband outputs of the IC after the signal travels through a pulse chain (e.g., amplification, mixing, etc.). The region 2402 indicates a region during which the pulse is sent (e.g., the pulse is generated or transmitted within this time). As shown, the pulse generates a voltage signal that exponentially decays. The graph 2400 shows decay of both the in-phase channel 2404 and the quadrature channel 2406 of the outputs.

[0121] FIG. 24B shows a Fourier transform 2450 of the FID of FIG. 24A, according to an example embodiment. The Fourier transform 2450 includes a real portion 2452 and an imaginary portion 2454. In various embodiments, the pulse length may be optimized for signal amplitude and an excitation bandwidth. For example, the pulse length may be optimized for a signal amplitude corresponding to Bi = 3.9G and an excitation bandwidth of 44MHz. A g-factor may be measured for the FID as a result. The measured g-factor from the FID of FIGS. 16A and 16B may be, for example, 2.00355. The measured g-factor may agree with a room temperature value for DPPH. For example, the measured g-factor of 2.00355 may correspond to a DPPH value of 2.0036. A shot repetition time may be lowered (e.g., to ~500ns) without observable reduction in the FID amplitude. Using a measured signal-to-noiseratio and a DPPH spin concentration, a measured limit of detection can be determined (e.g., 4.6* 109spins).

[0122] The SNR may be determined by dividing the initial amplitude of the FID by the RMS noise over the detection duration. The dead-time, determined as the duration between the transient at the end of the pulse to the start of the signal appearing as a smooth exponential decay, may be approximately 12ns (e g., inclusive of the limited bandwidth of the VGAs and output buffers).

[0123] FIG. 25 illustrates a graph 2500 showing a continuous wave signal obtained from the quadrature demodulator, according to an example embodiment. Specifically, graph 2500 illustrates a CW EPR frequency vs. magnitude plot for the sensing cells 706a and 706b. The graph 2400 includes data measured with an ADC (e.g., the ADC 812) from a lock-in amplifier (e.g., the lock-in amplifier 820) following quadrature demodulator outputs of the IC. In various embodiments, the difference in peak location between lines 2502 and 2504 may be due to uncertainty in positioning the system in the magnet. In various embodiments, the measured CW spin sensitivity of the sensor 706b with the same -600 m3sample as used for FID measurement in FIGS. 24A and 24Bis 2.9* 109spins per Hz. As shown by graph 2500 and graph 2400, for the same sensing cell (e.g., cell 706a or cell 706b), both CW and pulse EPR can be enabled. In various embodiments, a CW spectra with a modulation frequency of 100kHz and modulation depths of 1.8MHzpp (e g., using a 200 pm sensor with an estimated Bi of around 0.9G) and 1.2MHzpp (140pm, -0.8G) may be obtained.

[0124] In various embodiments, a setup may be used with an NMR reference permanent magnet (e.g., a Metrolab PM1055-050N) for portable operation, as shown in FIG. 21. The permanent magnet may be miniaturized. The permanent magnet may have a given nominal field, for example around 510mT. In various embodiments, the EPR IC may be mounted on one or more printed circuit boards (PCBs). The PCBs may be compatible with the permanent magnet that generates the static magnetic field. The PCBs may allow the chip to be located close to or inside a magnet, such that the magnetic field is at an appropriate field strength and is uniform. The size of the PCB may be minimized and constrained by the circuits, external to the spectrometer IC, that may be needed to support the IC, such as voltage regulators and the PFD / CPs.

[0125] Referring now to FIGS. 26A and 26B, FIDs with DPPH and varying fo are shown, according to an example embodiment (e.g., using the 140pm sensor, 23ns pulse, 1000averages). Specifically, FIG. 26A shows a time-domain FID 2600. Pulses may be generated where the center frequency is on resonance or off resonance (e.g., by less than a threshold amount) from the Larmor frequency (e.g., the resonant frequency determined by a magnetic biasing field). When the center frequency is on resonance, an exponential FID (e.g., as shown by line 2602) is produced. When the center frequency is off resonance, the signal may respond at the resonant frequency. Thus, when the signal is mixed down, using a local oscillator off the resonant frequency, a low frequency oscillation that modulates the exponential decay may be generated (e.g., as shown by lines 2604a-2610a). FIG. 26B shows a Fourier transform 2650 of the FID plot 2600 of FIG. 26 A. As shown by plot 2650, each spectrum 2602b-2610b has a different associated resonant frequency (e.g., the spectrum 2602b corresponds to the line 2602a in FIG. 26A). The amplitude is maximum with the FID that is most on-resonance (zero baseband frequency) and decreases as the frequency is moved off-resonance. In various embodiments, the EPR frequency may be first adjusted to be on-resonance with the sample using FIDs before the measurements of interest are performed.

[0126] The voltage recorded in the FID may be the voltage of the pulse outputs of the chip. There may be two sets of pulse outputs, in-phase (I) and quadrature (Q), as described herein. FIG. 26B shows a frequency vs. magnitude spectra according to this disclosure. The amplitudes shown in FIG. 26A may decrease as fo is moved off-resonance.

[0127] Referring now to FIG. 27A, a diagram 2700 illustrating a two-pulse inversion recovery concept is shown, according to an example embodiment. As the transverse component of the magnetization decays after a pulse, the magnetization eventually recovers to pointing along the +z direction by a longitudinal relaxation. The time constant governing this longitudinal relaxation is the spin-lattice relaxation time T . To measure T15a sequence of 2 pulses can be used in an inversion recovery experiment. First, a double-length u pulse is used to invert the magnetization along the z axis. Applying a n / 2 pulse after a delay T from the first pulse will produce an FID. When plotted versus T, the FID amplitude follows an exponential curve with a time constant of T1.

[0128] Referring now to FIG. 27B, data 2750 of an inversion recovery is shown, according to an example embodiment. In inversion recovery, for the second pulse, an ideal tip angle of magnetization of electrons may be 90 degrees. A pulse length of % / 2, shown in FIG. 27B as pulse 2752, may achieve such a tip angle. As such, the length of the pulse may be tuned to achieve maximum FID signal that corresponds to a 90 degree tip angle or rotation. Prior togeneration of the TT / 2 pulse, a n pulse 2754 may be generated. The it pulse may be double in length relative to the length of the 7t / 2 pulse. Additionally, by varying the length of time between two pulses (e g., delay), shown in FIG. 27B as r, the amplitude of the FID signal after a last pulse may increase and exponentially settle over time (e.g., as shown in plots 2756 and 2758). A rate of exponential increase or rise may be associated with a property of the sample, referred to as Ti. Ti may be referred to as a spin-lattice relaxation time or a longitudinal relaxation time. For example, inversion recovery may be performed using a 7t - it / 2 sequence and 1000 averages with DPPH (200 pm sensor, 82 — 41ns) and anthracite (140 pm, 46 — 23ns) as shown in FIG. 27B. The FID amplitude versus the delay between pulses may follow an exponential with a time constant of Ti. The fits may yield experimental Ti values of 81ns (DPPH) and 170ns (anthracite).

[0129] Referring now to FIG. 28A, a diagram 2800 of the Hahn echo concept is shown, according to an example embodiment. Sequences of multiple pulses may be constructed to produce spin echo signals, providing an alternative to detecting FIDs and enabling much more detailed investigation and manipulation of the unpaired electron(s) and the interacting environment. In a 2-pulse Hahn echo experiment, a TT / 2 — n sequence is used. As illustrated in Fig. 15(a), after the initial pulse, the FID decays as dephasing occurs. After a delay T, the n pulse flips the spins by 180°. Thereafter, as dephasing continues, the spins converge rather than spreading apart, refocusing to produce an echo after another delay of T.

[0130] FIG. 28B includes data 2850 illustrating the Hahn echo concept described in FIG. 28A. For example, Hahn echo with anthracite using the 140pm 706b, f0= 14.3GHz, 24.6 and 49.2ns pulse lengths, T = 198ns, and 100,000 averages is shown in chart 2852. The magnitude * l2+ Q2over time for nl?. — T sequences of samples of three different commercially available roasted ground coffee is shown in chart 2854. The same parameters (Sensor B, / 0= 14.32GHz, 24.6-49.2ns pulses, T = 117ns, and 1,000,000 averages) were used for all three samples. The echo intensity increases with the degree of roasting, in agreement with works that have found correlation between EPR signal intensity and roasting temperature. These echo experiments were performed with SRT = 10 ps, and the inset in chart 2854 and the magnitudes of chart 2856 are lowpass-filtered (20 MHz) in software to remove coupling on the printed circuit board from the clock divider outputs. The pulse lengths were optimized for maximum echo amplitude with fixed T. For these experiments, the echo peaks are located at T after the midpoint of the n pulse, as expected.

[0131] As with the FID in inversion recovery, the amplitude of the echo decays exponentially as T is increased. The time constant is the phase memory time TM, which depends on the spinspin relaxation as well as spectral diffusion. TM may be a property of or associated with a sample. Chart 2856 plots the echo amplitude versus r for the anthracite sample using the same pulses and number of averages as chart 2853. From the fit, an experimental TM value of 55ns is obtained.

[0132] Referring now to FIGS. 29A and 29B, a stimulated echo concept is shown, according to an example embodiment. A sequence of three n / 2 pulses produces a stimulated echo, which requires the presence of all three pulses, in addition to Hahn echoes after each pair of pulses. As illustrated in diagram 2900 of FIG. 29A, the second pulse rotates the disk of dephased spins into the yz plane. As dephasing continues, only the longitudinal components of the magnetizations will remain. The third pulse flips these components back into the transverse plane, which refocus to produce the stimulated echo. FIG. 29B shows a chart 2950 showing a sequence of three 20ns pulses with an anthracite sample, using Sensor B, SRT = lOps, and 1,000,000 averages. The stimulated echo occurs at a time of TXafter the last pulse, where TXis the delay between the first two pulses. In addition to the stimulated echo, FIDs after each pulse, as well as Hahn echoes corresponding to two pairings of pulses, are visible.

[0133] Referring now to FIGS. 30-31, CW experiments are shown as described.

[0134] FIG. 30 shows the CW spectra 3000 (obtained using the 200pm sensor 706a and the same settings as described above of DPPH and anthracite (10 averages)). CW spectra are obtained with the portable magnet using Sensor A with=~ 0.9G and 1.8MHzpp modulation at 100kHz. As seen in FIG. 30, the spectra of DPPH and anthracite (solid, 10 averages each) have a single spectral line. Owing to the mechanism of integrated CW EPR detection, wherein FM is applied in detecting a relative shift in the resonant frequency of an LC tank, the spectrum resembles the derivative of a dispersion-like curve. FIG. 31 shows a graph 3100 showing a CW spectrum for the spin probe TEMPO (a ImM sample in 90% glycerol with 100 averages). As shown in FIG. 31, the spectrum of TEMPO exhibits three peaks, owing to hyperfine coupling of the unpaired electron with the nearby14N nuclear spin. The peak whose frequency lies below resonance (equivalent to field above resonance in a field-swept spectrum) is lower in amplitude versus the other two peaks, which is consistent with spectra of TEMPO in viscous fluids. The ^-factor is approximately 2.01, and the peak spacing is 48MHz (equivalent to 17 Gauss), which is in reasonable agreement with reported values of the hyperfine coupling constant A.

[0135] Referring now to figure 32, a method 3200 is shown, according to an example embodiment. The method 3200 may be for detecting EPR signals. The method 3200 may be performed on a chip (e.g., a microchip). The chip may have associated digital logic (e.g., as described in FIG. 17) such that the chip, through this digital logic, performs the processes described herein. For example, when operating the chip, commands may be sent from a computer to the FPGA (or another controller) on the system (e g , chip), which then loads a configuration into a memory on the chip. The chip may execute the processes described herein according to the configuration.

[0136] At process 3202, a plurality of pulses of an oscillating magnetic field are generated. The pulses are generated using one or more sensor oscillators (e.g., the sensor oscillator 707) by successively turning a current source of the sensor oscillator on and off. The pulses may be generated at or have a first frequency (e.g., fo). The pulses may excite one or more electrons. In some embodiments, the sensor oscillators may be implemented as on-chip inductors. In some embodiments, the oscillator may be an inductor capacitor (e.g., LC) cell. In some embodiments, the chip may include one or more timing circuits to control the generation of the pulses.

[0137] The method 3200 may include generating, using a reference oscillator (e.g., the QVCO 702), a signal having a second frequency. The second frequency may be less than the first frequency (e.g., half of the first frequency). For example, the first frequency of the sensor oscillator may be 14GHz, and the second frequency of the QVCO may be half of the frequency of the sensor oscillator (e.g., 7GHz). The method 3200 may include increasing, using a frequency doubler (e.g., frequency doubler 606), the second frequency of the signal to match the first frequency. The method 3200 may further include injection locking the first frequency of the one or more sensor oscillators with the second frequency of the reference oscillator. Injection locking may be performed using the injection buffer 704. Injection locking may decrease a startup time of the one or more sensor oscillators and / or reduce a phase noise of the generated one or more pulses. The generation of the reference signal may be performed concurrently with the generation of the pulses at process 3202.

[0138] At process 3204, the one or more sensor oscillators are turned off. Turning off the sensor oscillators may cease generation of pulses.

[0139] At process 3206, an alternating current (AC) signal is sensed using one or more sensing cells (e.g., the sensing cell 706). The sensing cells may include the sensoroscillator(s). The AC signal may be produced by the one or more electrons upon excitation of the electrons from the pulses. The AC signal may be indicative of the excitation of the one or more electrons. In various embodiments, the AC signal is at least one of a: free induction decay (FID) signal or echo signal.

[0140] Process 3206 may further include removing (e.g., dissipating), using one or more switches (e.g., switch 916) in the one or more sensor oscillators 707, energy from the one or more sensor oscillators. Removing the energy may allow the FID signal to be sensed (e.g., without interference from a stronger signal generated by a reference oscillator). Each switch of the one or more switches may be coupled to a sensor oscillator of the one or more sensor oscillators. In some embodiments, the switches may reduce a time between generation of a pulse and detection of a pulse EPR signal.

[0141] At process 3208, one or more amplifiers may amplify the AC signal. The amplifiers may be one or more RFOTA stages 710. For example, the RFOTA may be a three stage amplifier, such that three amplifiers amplify the AC signal.

[0142] At process 3210, one or more mixers convert the AC signal into a baseband signal. The one or more mixers may be the subharmonic mixers 714. Converting the AC signal may include downconverting the signal from a first frequency (e.g., an EPR frequency) to the baseband frequency. The baseband signal may be, in some embodiments, a DC signal. In some embodiments, one or more phase interpolators (e g., phase interpolators 708) may generate local oscillator signals for the one or more mixers. Further, in some embodiments, one or more amplifiers (e.g., different than the RFOTA amplifiers that amplify the signal at process 3208) may amplify the baseband frequency.

[0143] At process 3212, the baseband signal is detected. The sensing cells may detect the baseband signal.

[0144] In some embodiments, the method further includes detecting CW EPR signals. The chip may detect CW EPR signals in a second mode of operation, where a first mode of operation detects pulse EPR signals, as described in processes 3202-3212. As such, the method may include generating, using the one or more sensor oscillators, a steady-state signal of the oscillating magnetic field. The method 3200 may further include sweeping the steadystate signal across a plurality of frequencies, and identifying a frequency shift indicative of excitation of the one or more electrons. To enable CW EPR detection, the chip may includea frequency-modulation demodulator (e.g., the quadrature FM demodulator 716) to demodulate frequency-modulated CW signals.

[0145] The disclosure is not limited by the embodiments shown in the drawings and described in the specification, which are given by way of example and not of limitation, but only in accordance with the scope of the appended claims.

[0146] Upon reading the foregoing description in light of the accompanying drawings, those skilled in the art will understand the concepts of the disclosure and will recognize applications of these concepts not particularly addressed herein. Those skilled in the art will recognize improvements and modifications to the preferred embodiments of the present disclosure. All such improvements and modifications are considered within the scope of the concepts disclosed herein and the claims that follow.

[0147] As utilized herein, the terms “approximately,” “about,” “substantially,” and similar terms are intended to have a broad meaning in harmony with the common and accepted usage by those of ordinary skill in the art to which the subject matter of this disclosure pertains. It should be understood by those of skill in the art who review this disclosure that these terms are intended to allow a description of certain features described and claimed without restricting the scope of these features to the precise numerical ranges provided. Accordingly, these terms should be interpreted as indicating that insubstantial or inconsequential modifications or alterations of the subject matter described and claimed are considered to be within the scope of the disclosure as recited in the appended claims.

[0148] It should be noted that the term “exemplary” and variations thereof, as used herein to describe various embodiments, are intended to indicate that such embodiments are possible examples, representations, or illustrations of possible embodiments (and such terms are not intended to connote that such embodiments are necessarily extraordinary or superlative examples).

[0149] The term “coupled” and variations thereof, as used herein, means the joining of two members directly or indirectly to one another. Such joining may be stationary (e.g., permanent or fixed) or moveable (e g., removable or releasable). Such joining may be achieved with the two members coupled directly to each other, with the two members coupled to each other using one or more separate intervening members, or with the two members coupled to each other using an intervening member that is integrally formed as a single unitary body with one of the two members. If “coupled” or variations thereof are modified by an additional term(e.g., directly coupled), the generic definition of “coupled” provided above is modified by the plain language meaning of the additional term (e.g., “directly coupled” means the joining of two members without any separate intervening member), resulting in a narrower definition than the generic definition of “coupled” provided above. Such coupling may be mechanical, electrical, or fluidic. For example, circuit A communicably “coupled” to circuit B may signify that the circuit A communicates directly with circuit B (i.e., no intermediary) or communicates indirectly with circuit B (e.g., through one or more intermediaries).

[0150] References herein to the positions of elements (e.g., “top,” “bottom,” “above,” “below”) are merely used to describe the orientation of various elements in the FIGURES. It should be noted that the orientation of various elements may differ according to other exemplary embodiments, and that such variations are intended to be encompassed by the present disclosure.

[0151] While various circuits with particular functionality are shown and described, it should be understood that any number of circuits for completing the functions described herein. For example, the activities and functionalities described herein may be combined in multiple circuits or as a single circuit. Additional circuits with additional functionality may also be included. Further, the circuits may further control other activity beyond the scope of the present disclosure.

[0152] As mentioned above and in one configuration, the “circuits” may be implemented in machine-readable medium for execution by one or more of various types of processors. Executable code may, for instance, comprise one or more physical or logical blocks of computer instructions, which may, for instance, be organized as an object, procedure, or function. Nevertheless, the executables need not be physically located together, but may comprise disparate instructions stored in different locations which, when joined logically together, comprise the circuit and achieve the stated purpose for the circuit. Indeed, a circuit of computer readable program code may be a single instruction, or many instructions, and may even be distributed over several different code segments, among different programs, and across several memory devices. Similarly, operational data may be identified and illustrated herein within circuits, and may be embodied in any suitable form and organized within any suitable type of data structure. The operational data may be collected as a single data set or may be distributed over different locations including over different storage devices, and may exist, at least partially, merely as electronic signals on a system or network.

[0153] While the term “processor” is briefly defined above, the term “processor” and “processing circuit” are meant to be broadly interpreted. In this regard and as mentioned above, the “processor” may be implemented as one or more processors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), digital signal processors (DSPs), or other suitable electronic data processing components structured to execute instructions provided by memory. The one or more processors may take the form of a single core processor, multi-core processor (e.g., a dual core processor, triple core processor, quad core processor, etc.), microprocessor, etc. In some embodiments, the one or more processors may be external to the apparatus, for example the one or more processors may be a remote processor (e.g., a cloud based processor). Alternatively or additionally, the one or more processors may be internal and / or local to the apparatus. In this regard, a given circuit or components thereof may be disposed locally (e.g., as part of a local server, a local computing system, etc.) or remotely (e.g., as part of a remote server such as a cloud based server). To that end, a “circuit” as described herein may include components that are distributed across one or more locations.

[0154] Embodiments within the scope of the present disclosure include program products comprising computer or machine-readable media for carrying or having computer or machine-executable instructions or data structures stored thereon. Such machine-readable media can be any available media that can be accessed by a computer. The computer readable medium may be a tangible computer readable storage medium storing the computer readable program code. The computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, holographic, micromechanical, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer readable medium may include but are not limited to a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), an optical storage device, a magnetic storage device, a holographic storage medium, a micromechanical storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, and / or store computer readable program code for use by and / or in connection with an instruction execution system, apparatus, or device. Machine-executable instructionsinclude, for example, instructions and data which cause a computer or processing machine to perform a certain function or group of functions.

[0155] The computer readable medium may also be a computer readable signal medium. A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including, but not limited to, electrical, electro-magnetic, magnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport computer readable program code for use by or in connection with an instruction execution system, apparatus, or device. Computer readable program code embodied on a computer readable signal medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, Radio Frequency (RF), or the like, or any suitable combination of the foregoing.

[0156] In one embodiment, the computer readable medium may comprise a combination of one or more computer readable storage mediums and one or more computer readable signal mediums. For example, computer readable program code may be both propagated as an electro-magnetic signal through a fiber optic cable for execution by a processor and stored on RAM storage device for execution by the processor.

[0157] Computer readable program code for carrying out operations for aspects of the present disclosure may be written in any combination of one or more other programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer readable program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone computer- readable package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0158] The program code may also be stored in a computer readable medium that can direct a computer, other programmable data processing apparatus, or other devices to function in aparticular manner, such that the instructions stored in the computer readable medium produce an article of manufacture including instructions which implement the function / act specified in the schematic flowchart diagrams and / or schematic block diagrams block or blocks.

[0159] Although the figures and description may illustrate a specific order of method steps, the order of such steps may differ from what is depicted and described, unless specified differently above. Also, two or more steps may be performed concurrently or with partial concurrence, unless specified differently above. Such variation may depend, for example, on the software and hardware systems chosen and on designer choice. All such variations are within the scope of the disclosure. Likewise, software implementations of the described methods could be accomplished with standard programming techniques with rule-based logic and other logic to accomplish the various connection steps, processing steps, comparison steps, and decision steps.

[0160] It is important to note that the construction and arrangement of the apparatus and system as shown in the various exemplary embodiments is illustrative only. Additionally, any element disclosed in one embodiment may be incorporated or utilized with any other embodiment disclosed herein.

Claims

CLAIMS1. A miniaturized spectroscopy system comprising: a spectrometer integrated circuit comprising: one or more on-chip sensors configured to: generate one or more pulses of an oscillating magnetic field; and detect one or more pulse electron paramagnetic resonance (EPR) signals in a first mode of operation.

2. The miniaturized spectroscopy system of claim 1, wherein the one or more on-chip sensors are configured to at least one of: generate an oscillating magnetic field, sweep a frequency of the magnetic field, or detect continuous-wave EPR signals in a second mode of operation.

3. The miniaturized spectroscopy system of claim 1, wherein the one or more on-chip sensors are implemented as oscillators comprising: one or more on-chip inductors configured to: generate the one or more pulses of the oscillating magnetic field; and detect the one or more pulse electron paramagnetic resonance (EPR) signals.

4. The miniaturized spectroscopy system of claim 3, wherein the spectrometer integrated circuit further comprises: one or more on-chip timing circuits to control the generation of the pulses; and a plurality of switches; and wherein each switch of the plurality of switches is coupled to one of the one or more sensor oscillators and is configured to at least one of: dissipate stored energy during the pulse generation or reduce a time between one of the one or more pulses and the detection of the one or more pulse EPR signals.

5. The miniaturized spectroscopy system of claim 4, wherein the one or more on-chip sensor oscillators are injection-locked when generating the one or more pulses, and wherein injection-locking at least one of: decreases a startup time of the one or more sensor oscillators or reduces a phase noise of the generated one or more pulses.

6. The miniaturized spectroscopy system of claim 3, further comprising: a frequency -modulation demodulator configured to demodulate frequency- modulated continuous wave signals.

7. The miniaturized spectroscopy system of claim 1, further comprising: one or more mixers configured to downconvert the pulse EPR signal from an EPR frequency to a baseband frequency; a plurality of phase interpolators to generate local oscillator signals for the one or more mixers; and a reference oscillator configured to injection-lock the one or more on-chip sensors.

8. The miniaturized spectroscopy system of claim 7, further comprising: one or more amplifier stages configured to amplify, at the EPR frequency, the pulse EPR signal prior to downconverting, by the one or more mixers, the pulse EPR signal; and one or more amplifiers configured to amplify, at the baseband frequency, after downconverting the pulse EPR signal by the one or more mixers, the downconverted pulse EPR signal.

9. A device comprising: a chip comprising: an on-chip EPR spectrometer comprising: one or more sensing cells, each sensing cell comprising a sensor oscillator); one or more amplifiers; and one or more mixers; and a controller comprising a memory and digital logic that causes the chip to: generate, using a sensor oscillator of one of the one or more sensing cells, a plurality of pulses of an oscillating magnetic field at a first frequency to excite one or more electrons; turn off the sensor oscillator to cease generation of pulses; sense, using the one or more sensing cells, an alternating current(AC) signal produced by the one or more electrons indicative of the excitation of the one or more electrons; amplify, using the one or more amplifiers, the AC signal;convert, using the one or more mixers, the AC signal into a baseband signal; and detect the baseband signal.

10. The device of claim 9, wherein the plurality of pulses are generated by: successively turning on and off a current source of the sensor oscillator.

11. The device of claim 9, wherein the on-chip EPR spectrometer further comprises a frequency -modulation demodulator configured to enable continuous-wave EPR; and wherein the digital logic causes the chip to: generate, using the sensor oscillator, a steady-state signal of the oscillating magnetic field; sweep the steady-state signal across a plurality of frequencies; and identify a frequency shift indicative of excitation of the one or more electrons.

12. The device of claim 11, wherein the device is configured to operate in a dual pulse EPR and continuous-wave EPR mode.

13. The device of claim 9, wherein: the AC signal is a pulse EPR signal that is at least one of: a free induction decay (FID) signal or an echo signal; the on-chip EPR spectrometer further comprises: one or more switches; and each switch of the one or more switches is coupled to the sensor oscillator of the one or more sensing cells, wherein the digital logic causes the chip to: remove, using one or more switches in the one or more sensor oscillators, energy from the one or more sensor oscillators to sense the FID signal.

14. The device of claim 9, wherein the on-chip EPR spectrometer further comprises: a reference oscillator; and a frequency doubler coupled to the sensor oscillator and the reference oscillator; and wherein the digital logic causes the chip to:generate, using the reference oscillator, a signal having a second frequency, the second frequency less than the first frequency; increasing, using the frequency doubler, the second frequency of the signal to match the first frequency; and injection lock the first frequency of the sensor oscillator with the second frequency of the reference oscillator.

15. The device of claim 14, wherein the sensor oscillator is injection-locked when generating the plurality of pulses, and wherein injection-locking is configured to at least one of: decreases a startup time of the VCO or reduces a phase noise of the generated plurality of pulses.

16. A method for detecting electron paramagnetic resonance (EPR) signals comprising: generating, using one or more sensor oscillators, a plurality of pulses of an oscillating magnetic field at a first frequency to excite one or more electrons; turning off the one or more sensor oscillators to cease generation of pulses; sensing, using one or more sensing cells, an alternating current (AC) signal produced by the one or more electrons indicative of the excitation of the one or more electrons; amplifying, using one or more amplifiers, the AC signal; converting, using one or more mixers, the AC signal into a baseband signal; and detecting the baseband signal.

17. The method of claim 16, wherein each of the one or more sensing cells comprises one of the one or more sensor oscillators.

18. The method of claim 16, further comprising: generating, using the one or more sensor oscillators, a steady-state signal of the oscillating magnetic field; sweeping the steady-state signal across a plurality of frequencies; and identifying a frequency shift indicative of excitation of the one or more electrons.

19. The method of claim 16, wherein the AC signal is at least one of a: free induction decay (FID) signal or echo signal; andwherein the method further comprises: removing, using one or more switches in the one or more sensor oscillators, energy from the one or more sensor oscillators to sense the FID signal.

20. The method of claim 16, further comprising: generating, using a reference oscillator, a signal having a second frequency, the second frequency less than the first frequency; increasing, using a frequency doubler, the second frequency of the signal to match the first frequency; and injection locking the first frequency of the one or more sensor oscillators with the second frequency of the reference oscillator.

Images