Gradiometric volumetric flux concentrator for ultrasensitive magnetic detection and squid-based magnetic detection system using this flux concentrator

Abstract

The invention relates to a flux concentrating device (5) arranged upstream of an ultra-sensitive magnetometer (3), which is intended to receive an external magnetic flux and transmit a concentrated flux at the inlet of the magnetometer (3), and is characterized by containing a primary gradiometric volumetric magnetic sensing antenna (50) cooled to a cryogenic temperature and a flux transformer (55) arranged between the primary sensing antenna (50) and the magnetometer (3), which has a primary winding (6) connected to the primary sensing antenna (50).

Description

FIELD OF INVENTION

[0001] The present invention relates to a primary antenna intended for use as a flux concentrator for an ultrasensitive magnetometer. This antenna is a volumetric gradiometer made of copper. The invention also relates to equipment incorporating this antenna: SQUID magnetic detection devices, or any other type of ultrasensitive magnetometer (NV centers, giant magnetoresistance sensors, atomic magnetometers), MRI based on SQUID detection, biomagnetism, or more generally, ultrasensitive near-field magnetometry. STATE OF THE ART

[0002] The SQUID (Superconducting Quantum Interference Device) is a micrometer-sized magnetometer consisting of a loop of superconducting material intercepted by two Josephson junctions. The technologies and materials used vary considerably, but two main families are distinguished: low-temperature (low Tc) SQUIDs, operating below 10 K, offering superior performance but requiring more intensive cryogenics, and high-temperature SQUIDs, with lower performance and more complex fabrication, but capable of operating at the temperature of liquid nitrogen (77 K). The SQUID behaves like a fluxmeter, meaning it produces a sinusoidal voltage V(Φ) that depends on the magnetic flux Φ passing through it.These are magnetometers with very high performance: among their many advantages, the first are their extreme sensitivity (detection threshold approaching 1 ft.Hz -1 / 2) [1], their wide operating bandwidth, up to more than 100 MHz [2], and the preservation of their sensitivity down to very low frequencies, including in direct current (DC) fields. A detailed description of the operation of SQUIDs is given in John Clarke's book [3].

[0003] Due to their small size, SQUIDs are widely used in the near field, that is, to measure local magnetism at small scales, for example, to map the magnetic domains of certain materials. Using SQUIDs in the far field, particularly for MRI experiments or biomagnetism measurements, generally requires a flux concentrator. This is a conventional magnetic antenna in series with an input coil placed above the SQUID. The magnetic flux is captured by the primary antenna and then transmitted to the SQUID via the input coil. In short, the flux concentrator can be understood as virtually increasing the surface area of ​​the SQUID, allowing it to capture a large magnetic flux despite its small size.These flux concentrators can have various geometries, made from metals such as copper or superconducting materials such as Niobium-Titanium wires. An overview of the various implementations of flux concentrators is given in Fagaly's review [4] or in more detail in the chapter of Weinstock's book [5].

[0004] Until now, the common feature of flux concentrator implementations has been the use of primary surface antennas, most often gradiometric. Let's assume that the magnetic signal to be measured is emitted isotropically into space by a source S. A surface antenna only intercepts a fraction of the signal emitted by the source, while a volume antenna encompasses the source S to collect the maximum signal. Until now, in the context of magnetic detection using SQUIDs at low critical temperatures, surface antennas have always been preferred. This is because the goal is to have an antenna with the lowest possible electrical resistance to minimize Johnson-Nyquist noise. Flux concentrators are therefore most often made of superconducting material, for example, niobium-titanium wire, and must therefore be cooled to 4 K.For this reason, the primary antenna must be confined to the bottom of a cryostat, which necessitates the use of a surface antenna.

[0005] In the case of high-temperature SQUIDs, it is not possible to use superconducting wire antennas as with low-temperature SQUIDs because the materials used are ceramics, which are poorly suited to wire fabrication. The most widespread technologies are superconducting planar antennas, although some experiments have used copper volume antennas coupled to the SQUID.

[0006] Since the SQUID coupled with a primary antenna is a very sensitive device, it is also very vulnerable to ambient electromagnetic noise. To mitigate this, the primary antenna can be configured as a gradiometer. Let's consider the simple case of a first-order axial gradiometer. In this case, the antenna consists of two loops with the same axis rotating in opposite directions, separated by a distance b. This device is less sensitive to the signal of interest than a simple loop, because the second loop rotating in the opposite direction will tend to compensate for the signal received by the first loop. However, this system will tend to reject noise from sources located at a distance R >> b.

[0007] Indeed, from the perspective of the noise source, the two loops of the gradiometer are equidistant, and therefore the signal received by the gradiometer from the noise source cancels out between the two loops. In fact, the system thus described is called a gradiometer because it is sensitive to the gradient of the magnetic field, rather than to the magnetic field itself. Many other gravimetric systems with more complex geometries exist, sensitive to higher-order derivatives of the magnetic field.

[0008] The higher the order of the gradiometer, the more effective the noise rejection property, but the less sensitive the system will be to the signal of interest: a compromise must therefore be found. Generally speaking, it should be remembered that if the field emitted by a source decreases as R -3< with the distance R At the source, the gradient decreases as R -4< , and the derivative of order k the field decreases as R- 3< -k< . A detailed description of the different implementations of surface gradiometers can be found in the Fagaly review cited above.

[0009] There are various implementations of SQUID systems with flow concentrators. Below are listed some prior art examples classified by technology type.

[0010] For low Tc gradiometers, reference should be made to the documents cited [6] to

[11] .

[0011] For high Tc gradiometers, reference should be made to the documents cited

[12] to

[18] .

[0012] However, these flux concentrators do not solve the problem of the lack of sensitivity in current SQUID detection systems. The aim of the invention is to propose a new flux concentrator which, combined with a SQUID detector, makes it possible to create a magnetometer with significantly higher sensitivity than current magnetometers, while also being robust against ambient electromagnetic noise. REFERENCES

[0013] 1. Stolz et al., Supercond. Sci. Technol. 12, 806-808 (1999). 2. D. Drung et al., dc SQUID readout electronics with up to 100 MHz clo8ed-loop bandwidth, in IEEE Tran8action8 on Applied Superconductivity, vol. 15, no. 2, pp. 777-780, June 2005 3. J. Clarke and AI Bragin8ki, The SQUID handbook vol. 1, Wiley-VCH. 4. Fagaly et al., Review of Scientific Instruments 77, 101101 (2006). 5. SQUID sensors: Fundamentals, Fabrication and Applications, NATO ASI Series, Springer Science, 1995, pages 117 - 178. 6. Seton 1999. Seton et al. Magnetic resonance materials in physics, biology and medicine 8 116-120 (1999). Planar or axial low Tc gradiometer for low-field MRI. 7. Stolz 1999. Stolz et al., Supercond. Sci. Technol. 12, 806-808 (1999). Low Tc planar geometry gradiometers. 8. Stolz 2001. IEEE Transactions on applied superconductivity 11, vol. 1 (2001). Planar low Tc second-order gradiometer. 9. JP2010148578A. Axial gradiometers for magnetoencephalography. 10.US2013271142A1. SQUID MRI system comprising multiple low-Tc second-order gradiometers. 11. WO2006052236A1. Low-Tc SQUID MRI system comprising one untuned second-order gradiometer. 12. Tavrin 1994. Tavrin et al., Supercond. Sci. Technol. 7 265-268 (1994). Axial second-order gradiometer capable of operating without shielding for magnetocardiography. 13. Oisjoen 2008. Oisjoen et al. Supercond. Sci. Technol. 21, 034004 (2008). High-Tc gradiometer for magnetic immunoassay experiments. 14. Schmidt 2004. Schmidt et al. Exploration Geophysics 35 297-305 (2004). Triaxial high-Tc SQUID gradiometer allowing measurement of the three components of the magnetic field. 15. Schultze 2001. Schultze et al., Supercond. Sci. Technol 15 120-125 (2002). Planar gradiometer for use without shielding. 16. DE19509230A1. High-Tc gradiometer for the detection of biomagnetic fields. 17. CN106814338.Noise reduction system based on a first-order gradiometer for nuclear magnetic resonance applications, detection performed with a second-order gradiometer. 18. CN107430174A. System ensuring the thermal stability of an axial gradiometer with SQUID. 19. Nardelli 2020. Nardelli et al. EPJ Quantum Technology 7:11 (2020). Implementation of an axial gradiometer for magnetoencephalography. Magnetic detection is performed in this case with atomic magnetometers. In this specific case (no SQUID), the gradiometer consists of two atomic magnetometers, without a flux concentrator. 20. Chen 2011. Chen et al. Journal of Applied Physics 110 093903 (2011). Low-field MRI detected by a high-Tc SQUID system including a non-gradiometric volumetric primary coil (solenoid). 21. Hout et al. J. of Mag. Res. 24 vol. 1, 71-85 (1976). 22. AIP Advances 8, 075126 (2018). 23. WO 2009 / 023303 A2 (Penanen et al.). Description of the invention

[0014] The invention proposes to address the problem of lack of sensitivity of current SQUID detection systems by means of a copper volume gradiometric antenna, enabling the collection of more signal while maintaining robustness against noise, as defined by claim 1.

[0015] This objective is achieved with a flux concentrator device placed upstream of a magnetometer ultrasensitive, intended to receive an external magnetic flux as input and deliver a concentrated flux at the input of said magnetometer, characterized in that it comprises a primary gradiometric volume magnetic detection antenna cooled to a cryogenic temperature and a flux transformer disposed between the primary detection antenna and the magnetometer, said flux transformer having a primary winding connected to said primary detection antenna.

[0016] In a particular configuration of the invention, the primary antenna comprises two antennas in series respectively internal and external, having the same inductance and arranged so that the current flowing through the external antenna is in the opposite direction to that flowing through the internal antenna.

[0017] In a particular configuration of the invention, the primary antenna comprises two antennas in series respectively internal and external, having the same total oriented surface and arranged so that the current flowing through the external antenna is in the opposite direction to that flowing through the internal antenna.

[0018] The internal antenna can be arranged inside the external antenna and has a length and diameter respectively smaller than the respective length and diameter of the external antenna.

[0019] The internal and external antennas can respectively be made from internal and external conducting wires, the internal conducting wire having a smaller cross-section than the external conducting wire.

[0020] In other configurations of the invention, the primary detection antenna further comprises at least one intermediate antenna arranged in series between the external antenna and the internal antenna.

[0021] The flux concentrator device may further include means for electromagnetically shielding the primary detection antenna.

[0022] These electromagnetic shielding methods can be passive or active, in which case they employ at least one external noise compensation coil.

[0023] The primary detection antenna can be cooled by means for cooling the primary detection antenna by circulating a cryogenic liquid such as liquid nitrogen, or the primary detection antenna can be disposed in a cryostat comprising a pulsed tube and a gaseous helium circuit.

[0024] According to other aspects of the invention, a nuclear magnetic resonance device, a magnetic resonance imaging device, a magnetic detection device implemented in biomagnetism applications is proposed, the detection system of which integrates a flux concentrator device according to the invention.

[0025] One can also provide equipment for measuring very small variations in static magnetic field or a broadband magnetic detection equipment, or a magnetic detection or nuclear magnetic resonance equipment used for quality control in the food industry, the detection system of which incorporates a flux concentrator device according to the invention.

[0026] According to yet another aspect of the invention, a radio frequency (RF) detection and acquisition system based on SQUID is proposed, intended in particular for integration into a nuclear magnetic resonance (MRI) device, comprising: a flux concentrator device according to the invention, a SQUID device, arranged to be subjected to the flux from said flux concentrator device, and to deliver a secondary detection signal, a cryogenic device provided for cooling the SQUID device, a secondary detection signal processing stage emitted by the SQUID device, to deliver an analog acquisition signal, comprising a flux-locked loop (FLL) provided for linearizing the response of the SQUID device.

[0027] The flux concentrator device implemented in this system may advantageously include an internal area provided for the primary detection antenna to receive a part of the body of a human or animal subject. DESCRIPTION OF THE FIGURES

[0028] [ Fig.1 ] There [ Fig.1 ] presents different geometries of volumetric antennas considered within the framework of the present invention; [ Fig. 2 ] There [ Fig. 2] presents, using the saddle geometry as an example, the different orders of gradiometer that can be implemented; [ Fig.3 ] There [ Fig.3 ] presents a possible implementation of the antenna in the context of a SQUID magnetic detection experiment for MRI. DETAILED DESCRIPTION

[0029] All the geometries represented in [ Fig.1 ] have in common that they are gradiometric and volumetric. One can consider a saddle-type geometry, Helmholtz, solenoid, or any other type of volumetric geometry. It is possible to consider higher-order gradiometers for these geometries, as described in [ Fig. 2 In solenoid geometry such as Helmholtz, the current moves in opposite directions between the inner and outer parts of the antenna.

[0030] The exact dimensions of the geometries represented in [ Fig. 2The diagrams are not followed to make them easier to read. The antenna wiring is designed to alternate the direction of the current from one stage to the next.

[0031] The primary detection antenna implemented in a flux concentrator according to the invention has a gradiometric structure that can be of different orders, as illustrated by the three geometries of orders 0, 1 and 2 respectively shown in [ Fig. 2 ].

[0032] The first geometry 50 is of the saddle type, well known to those skilled in the art for its performance, particularly in terms of spatial homogeneity. The diameter of the saddle antenna, for example, is equal to 1.5 times its length. The zero-order gradiometric volume antenna 50 comprises a coil 53, 54 with two connection terminals 51, 52.

[0033] In the second geometry 60, the first-order primary detection antenna includes an external saddle-shaped antenna 65,66 in series with an internal saddle-shaped antenna 63,64 included in the external antenna, and two connection terminals 61,62.

[0034] The second-order gradiometric volume primary detection antenna 70 comprises an external saddle-shaped antenna 77,78 containing an intermediate antenna 75,76 into which an internal antenna 73,74 is inserted. The external, intermediate and internal antennas respectively are connected in series so that the primary detection antenna 70 has two connection terminals 71,72.

[0035] The external, internal or intermediate antennas in these different configurations, for example, have only one turn of wire.

[0036] The other geometry (b) is the first-order gradiometric version of the saddle antenna. This volumetric antenna 5' consists of two sub-antennas 51, 52 connected in series. The first, internal antenna 51 has a saddle geometry and, in this example, features two turns of wire.

[0037] The dimensions of the system and the orientation of the antenna wires are chosen so that the external and internal parts of the antenna have the same total oriented surface area, and the current in the internal part flows in the opposite direction to the current in the external part.

[0038] With this configuration, a saddle-shaped gradiometric antenna can reject noise from sources located at a distance much greater than the antenna's dimensions, while benefiting from the homogeneity properties of the saddle geometry. A detailed description of the principle of gradiometric antennas can be found in the article by RL Fagaly [4].

[0039] The antenna shown in the example of the [ Fig.3] illustrating a SQUID-based MRI detection system 1 is a 0th-order gradiometer (or magnetometer) comprising a primary detection antenna 50 having a saddle geometry and integrated into a cryostat 56, but any other volume antenna can be used. The sample (a) 2 from which an MRI image is to be made is placed at the center of the primary antenna (b) 50, here with a saddle geometry. The sample can, for example, be a human knee that is inserted inside the primary antenna 50. The dashed part (c) 11 represents a set (not shown) of polarization coils and gradients necessary to perform the MRI experiment. These coils are controlled by a console (d) 12. The signal from the primary antenna 50 is transmitted to the SQUID 3 via a flux transformer ( L 1, L 2) 6.7; 55 used to perform impedance matching between the primary antenna 50 and the input coil 8 of the SQUID Li.

[0040] Since this is a resonant antenna, in the case where the antenna is resonant, a coupling capacitor 9 of capacitance C a is connected in parallel with the primary winding 6 of the flux transformer 55.

[0041] The signal at the output of the SQUID 3 on the output coil L feed 10 is then amplified via a preamplifier (LNA) 40, and the stability of the operating point is ensured by a feedback loop (FLL) 41. The amplified signal 42 then passes through an analog-to-digital conversion stage (not shown). Example of antenna geometry calculation

[0042] Let us first consider, as an example, the case of a saddle-shaped antenna. This antenna has a cylindrical geometry, and its length a is related to the diameter D by a = 3 < DIn this example, a minimum diameter space of 18 cm must be maintained at the center of the antenna to accommodate a knee. Including a margin of 6 cm To account for the thickness of the antenna and the cryostat enclosure that cools it, a diameter will be used. D = 24 cm and therefore a length a = 36 cm.

[0043] We now consider the cross-section of the antenna conductor. Copper is used here as the conductor, which allows for simpler cryogenics than for superconducting antennas. The goal is to have the lowest possible resistance antenna to minimize Johnson-Nyquist noise, therefore requiring the largest possible conductor cross-section.

[0044] However, due to the thickness of the skin (noted δ ), it is not relevant to consider conductor diameters greater than 2 δAt the operating frequencies of interest to us, between 10 kHz and 100 kHz, the skin thickness of copper varies between 600 µ m and 200 µ m. Taking the value as a reference δ = 500 µ m, therefore we have a maximum diameter for our antenna d = 1 mm.

[0045] To build a volume gradiometer, the idea is to put in series two antennas of total surface areas oriented S1 and S2 opposite for a quasi-uniform magnetic field. S → j = ∬ Antenne j dS →

[0046] The current flowing through the external antenna is in the opposite direction to that flowing through the internal antenna. Thus this system, represented in the [ Fig.1 ], functions as a first-order gradiometer in that it rejects noise from distant sources by compensating between the internal and external antennas.

[0047] To construct a second-order volume gradiometer, the same principle is followed as for the surface gradiometers described in Weinstock's book [5]. Three antennas are connected in series; the inner and outer antennas have the same total oriented surface area, while the intermediate antenna has twice the total oriented surface area and in the opposite direction. This latter constraint is achieved by circulating the currents in opposite directions from one stage to the next. S → j = ∬ Antenna j dS → , S → 2 = − 2 S → 1 = − 2 S → 3

[0048] For higher gradient orders, the same construction principle is followed: the ratios between the total surface areas of each stage and the direction of current flow are given by the coefficients and the sign of the discrete expansion of the nth derivative. For example, for the second order, we have ∂ 2 B x ∂ x 2 ≃ 1 ϵ 2 B x + ϵ − 2 B x + B x − ϵ for a distance 1 small in front of the x position.

[0049] Consider the internal antenna. We seek to calculate the effective oriented surface area of ​​a saddle-shaped antenna of length a int = 36.0 cm and diameter D int = 24.0 cm.

[0050] We can refer to the [ Fig.1 For clarity, such an antenna has a total oriented surface area Sint = 0.150 m², oriented along the axis passing through the centers of the two antenna panels. Now, considering an external antenna with the same angle and a diameter Dext = 30.0 cm, we must have, maintaining the equality of the total surface areas, aext = 28.8 cm. An alternative is to double the internal antenna by winding the wire a second time. In this case, the effective surface area of ​​the internal antenna is Sint = 0.299 m², and it is possible to meet the constraint aext = 3 / 2 Dext by taking Dext = 34 cm and aext = 51 cm.

[0051] The total inductance of the antenna in the first case is estimated atL = 4.7 µ H. We know that to have the maximum sensitivity of the SQUID + antenna device, the inductance of the primary antenna must be equal to the inductance of the input coil of the SQUID.

[0052] This fact is described in publication FR2012642. This condition can be met to within a factor of 2, with very little impact on sensitivity, using the commercial SQ2600 device from STAR Cryoelectronics, which has an input inductance L i = 2580 nH. Johnson-Nyquist noise in the primary antenna

[0053] Using a detector as sensitive as the SQUID is not without its drawbacks: its extreme sensitivity makes it very vulnerable to external interference. The noise picked up by the primary antenna + SQUID system essentially has two origins: Noise originating from sources external to the device. The majority of the noise spectral density is located in the band ranging from DC to a few kHz. The contribution to noise from microphonics generated by the mechanical vibrations of the pulsed tube used for cryogenics is particularly noteworthy. Internal noise within the device. This is primarily Johnson-Nyquist noise, typically of much lower amplitude than external noise. However, to maximize the performance of the SQUID, the aim is to minimize the flux noise caused by the input coil to keep it below the intrinsic flux noise of the SQUID. Noise in 1 / f is negligible at the frequencies at which we work (between 10 and 100 kHz).

[0054] As mentioned previously, in most existing implementations, the primary antenna is superconducting to limit Johnson noise. However, using a superconducting antenna is not without its drawbacks. The application according to the invention requires that the antenna be placed in a lightweight cryostat that encloses the sample. This cryostat cannot be cooled to the temperatures required for using niobium antennas that need to be thermalized at 4K. One solution would be to use high-Tc materials for the antenna, such as YBCO. Such materials are ceramics, and it is not possible to simply fabricate a wire antenna with them. Furthermore, the layered crystalline structure makes the fabrication of non-planar devices very difficult, if not impossible.

[0055] For these reasons, a copper antenna configuration is preferred, ideally made of OFHC copper to further reduce resistivity. It is assumed that this antenna has been cooled to a temperature T = 40 K. At this temperature, a medium-quality copper has a resistivity of approximately ≈ 6.10⁻¹⁰ Ω.m. Considering the volume gradiometric saddle antenna described in the previous section, the internal wire length is 245 cm, with a wire diameter of 0.7 mm, and the external wire length is 259 cm, with a wire diameter of 1.0 mm. This antenna has a total resistance of approximately R ≃ 1 , 5 mΩ

[0056] The RMS value of the Johnson-Nyquist noise associated with this antenna is given by V n = 4 Rk B T 1 / 2 where T is the temperature and k B the Boltzmann constant. For T = 40 K, we obtain V n ≃ 1,8 . 10 − 12 V / Hz

[0057] We want to measure the impact of this voltage noise on the signal captured by the SQUID. To do this, let's start by noting that the total impedance Z of the antenna at the frequency of interest ω is simply Z = Lω since at the frequency of interest, f 0 ≈ 42 kH'., we have R << Lω.

[0058] The gradiometric antenna has an inductance L ≈ 5 5 µ H. The noise running through the antenna is then simply I n = V n 2 πLf 0 ≃ 1 , 25 pA / Hz

[0059] This current noise, present in the SQUID's input coil, produces flux noise via the coupling between the SQUID and its input coil. The STAR Cryoelectronics SQ2600 model, which exhibits coupling between the SQUID and the input coil, is used as an example. Mi = 10 Φ 0 / µ A. The noise in flux caused at the SQUID by the primary antenna is therefore Φ n sq = M i I n = 12 , 5 μ Φ 0 / Hz which stto be compared to the intrinsic noise of the SQUID, therefore for this model less than 5 µ Φ 0 / Hz.

[0060] The calculated noise is therefore of the same order of magnitude, which leaves a possibility of optimizing the system by going below the noise threshold of the SQUID, in particular by choosing a better quality copper. Temperature

[0061] Referring to the figure by S. Calatroni, arXiv 2006.02842vl, depending on the purity, the resistivity of Copper at 40 K varies from 2.10 -10< to 1.5.10 -9< Ω.m. This temperature can be reached using a cryogenic system employing a pulsed tube coupled with a circulation of gaseous Helium, for example the RP-082B2S model from Sumitomo.

[0062] However, it is much simpler to use liquid nitrogen cooling initially. At 77 K, the resistivity of copper varies from 1 x 10⁻⁹ to 2.5 x 10⁻⁹ Ω·m. A factor of 2 to 5 could be gained on the resistivity, which translates to a factor of 1.5 to 3 on the flux noise at the SQUID, by using good quality copper at 40 K, compared to nitrogen cooling. Current in the primary antenna

[0063] As calculated previously, the noise in the primary antenna is on the order of I n = ≃ 1 , 25 pA / Hz

[0064] It remains to be seen whether this noise is sufficiently low to detect the magnetic NMR signal without noise averaging. To calculate the current in the antenna, the calculation of Hoult et al., 1976

[21] is implemented. By the principle of reciprocity, the electromotive force E induced in an antenna of arbitrary shape C by a magnetization sample M0 in rotation in the xy plane is given by E = ∫ échantillon ∂ ∂ t B → 1 . M → 0 dV Or B 1 is the field induced by antenna C when a unit current flows through the antenna. If B Since 1 is homogeneous over the volume of the sample, the integral is calculated simply: E = Kω 0 B 1 xy M 0 V s cos ω 0 t

[0065] K is a constant that takes into account field inhomogeneities, B xy< is the component of B 1 orthogonal to the polarization field B 0. For an MRI experiment of the knee, a sphere of radius 4 cm is used as the volume of interest, i.e., Vs = 2.68 x 10-4 m3. Magnetization M 0, within the framework of the biological NMR experiment, is given by M 0 = nγ 2 ℏ 2 I I + 1 B 0 3 k B T Or n is the number of spins per unit volume, I is the quantum number associated with the eigenvalue of the quantum operator J 2. From equation 9, we can deduce the intensity of the current flowing through the antenna, given by i =E / Z, Or Z ( ω ) is the antenna impedance, which is simply Z ω = R 2 + L 2 ω 2 ≃ Lω

[0066] L being the total inductance of the system (input coil - primary antenna).

[0067] Then, following the framework of Wenfeng Wu et al.

[22] , we determine the field B 1 by a saddle-type coil. In this case, the parameters to consider are Φ = 2 π / 3 for opening the coil, R = 12.0 (resp. 12.7) cm, h = 36 (resp. 38, 1) cm for the internal (resp. external) antenna.

[0068] The field produced by this coil is, by construction, perpendicular to B0. Therefore, for a unit current B 1 xy = 2 μ 0 N πR s − 1 s − 1 / 2 + s 3 / 2 sin Φ / 2

[0069] We first consider simply the case of the internal coil, with N = 1 turn. s = 1 + ( h / D ) 2< is a function of the antenna's structure constant. In this case, s ≈ 3.25. Therefore, with the aforementioned parameters, we obtain: B 1 xy , int = 55 , 5 μT for the internal antenna. In the case of the external antenna, we obtain B 1 xy , ext = 52 , 5 μT

[0070] Consider an equivalent field as seen by the gradiometer B 1xy = 3, 0 µ T. The magnetization of the sample, on the other hand, is given by M 0 = 3,17 . 10 − 6 A . m 2 with the value of n for water, n = 6.6 x 10²⁸ m⁻³. The value of the electromotive force is then E = 6,8 . 10 − 10 V

[0071] Taking into account a resolution in At a frequency of 1 Hz, this result can be compared to typical noise. V n = 1.8 x 10⁻¹² < V / HzThe signal-to-noise ratio is 380, which is more than sufficient for clinical MRI scans under 10 minutes. However, other noise sources must be taken into account, particularly those from the readout chain (amplifiers, analog-to-digital converters) as well as external noise sources, which will inevitably degrade the signal-to-noise ratio.

[0072] This detection system according to the invention finds its use in several applications: Magnetic resonance imaging (MRI)

[0073] Indeed, very low-field MRI experiments (below 10 mT) require the use of highly sensitive detectors to compensate for the lack of signal. Most existing SQUID MRI experiments use surface-type primary antennas, most often second-order axial gradiometric antennas, as disclosed by

[23] . The use of a volumetric antenna allows for the capture of more signal and thus increases the signal-to-noise ratio of the device, and ultimately the image quality. The gradiometric nature of the antenna helps to mitigate the effects of surrounding electromagnetic noise, also increasing the signal-to-noise ratio. Biomagnetism.

[0074] The primary antenna according to the invention can be used to detect signals from NMR, magnetic relaxometry, superparamagnetism, or can be useful in magnetic immunoassays where the sensitivity of the detection device is a crucial factor. The antenna according to the invention can also be used for magnetoencephalography and magnetocardiography experiments.

[0075] Other applications of the present invention may aim to measure small variations in magnetic fields, particularly for submarine detection, or other defense applications, especially in electronic warfare. Use in nuclear magnetic resonance (NMR) in the food industry, particularly for quality control, is also foreseeable.

[0076] Of course, the present invention is not limited to the embodiments just described and many other modes of implementation can be envisaged without departing from the scope of the invention defined by the attached claims.

Claims

1. Flux concentrator device (5) arranged upstream of an ultrasensitive magnetometer (3), which device is provided to receive an external magnetic flux as input and to deliver a concentrated flux as input to said magnetometer (3), characterized in that the device comprises a primary gradiometric volumetric magnetic detection antenna (50,60,70), cooled to a cryogenic temperature, and a flux transformer (55) arranged between the primary detection antenna (50,60,70) and the magnetometer (3), said flux transformer (55) having a primary winding (6) connected to said primary detection antenna (50,60,70).

2. Concentrator device according to the preceding claim, characterized in that the primary antenna (60) comprises two antennas (63,65), namely an internal and external antenna, in series, having the same inductance and arranged so that the current flowing through the external antenna (65) is in the opposite direction to that flowing through the internal antenna (63).

3. Concentrator device according to claim 1, characterized in that the primary antenna (60) comprises two antennas (63,65), namely an internal and external antenna, in series, having the same total oriented surface and arranged so that the current flowing through the external antenna (65) is in the opposite direction to that flowing through the internal antenna (63).

4. Concentrator device according to the preceding claim, characterized in that the internal antenna (63) is arranged inside the external antenna (65) and has a length and diameter less than the length and diameter, respectively, of the external antenna.

5. Concentrator device according to the preceding claim, characterized in that the internal and external antennas (63,65) are made from conducting wires, namely internal and external conducting wires, the internal conducting wire having a cross-section smaller than the cross-section of the external conducting wire.

6. Concentrator device according to any of claims 2 to 4, characterized in that the primary detection antenna (70) further comprises at least one intermediate antenna (75) arranged in series between the external antenna (77) and the internal antenna (73).

7. Concentrator device according to any of the preceding claims, characterized in that it further comprises means for electromagnetically shielding the primary detection antenna.

8. Concentrator device according to the preceding claim, characterized in that the electromagnetic shielding means are passive.

9. Concentrator device according to claim 7, characterized in that the electromagnetic shielding means are active and implement at least one external noise compensation coil.

10. Concentrator device according to any of the preceding claims, characterized in that it further comprises means for cooling the primary detection antenna by circulating a cryogenic liquid such as liquid nitrogen.

11. Concentrator device according to any of claims 1 to 8, characterized in that the primary detection antenna is arranged in a cryostat comprising a pulse tube and a gaseous helium circuit.

12. Nuclear magnetic resonance equipment, the detection system of which incorporates a flux concentrator device according to any of the preceding claims.

13. Magnetic resonance imaging equipment, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

14. Magnetic detection equipment implemented in biomagnetism applications, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

15. Equipment for measuring very small variations in a static magnetic field, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

16. Broadband magnetic detection equipment, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

17. Magnetic detection equipment used for quality control in the agri-food industry, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

18. Nuclear magnetic resonance equipment used for quality control in the agri-food industry, the detection system of which incorporates a flux concentrator device according to any of claims 1 to 11.

19. System (1) for radio frequency (RF) detection and acquisition based on SQUID, provided in particular to be integrated into a nuclear magnetic resonance (MRI or NMR) apparatus, comprising: • a flux concentrator device (5) according to any of claims 1 to 11, • a SQUID device (3), arranged to be subjected to the flux from said flux concentrator device (5), and to deliver a secondary detection signal, • a cryogenic device (56) provided to cool the SQUID device (3), • a stage (4) for processing the secondary detection signal emitted by the SQUID device (3) to deliver an analog acquisition signal, comprising a flux-locked loop (FLL) provided to linearize the response of the SQUID device (3).

20. System (1) according to the preceding claim, characterized in that the flux concentrator device (5) comprises a zone provided within the primary detection antenna (50) to receive a part (2) of the body of a human or animal subject.