MEASURING APPARATUS FOR WEAK ELECTROMAGNETIC SIGNALS FROM A SAMPLE AT LOW FREQUENCIES, IN ADDITION TO A METHOD

Abstract

The invention relates to a measuring apparatus for detecting weak electromagnetic signals from a sample at low frequencies, specifically in the frequency range of 1 kHz-10 MHz, in particular, and to a measuring method. The problem addressed by the invention is that of providing an apparatus which can be used to detect weak electromagnetic signals from a sample, in particular in the frequency range of 1 kHz-40 MHz, with a good signal-to-noise ratio. For the solution, the measuring apparatus comprises an electromagnetic resonant circuit having a pick-up coil of low quality, a preferably tunable capacitance and a filter coil; the filter coil and the capacitance have a high quality of at least 100, advantageously at least 200, particularly preferably at least 500. Alternatively or additionally, the quality of the resonant circuit is at least 100, advantageously at least 200, particularly preferably at least 500. The quality of the filter coil and the quality of the capacitance exceed the quality of the pick-up coil, specifically at least by twice the amount. The measurement signal is then available at the two ends of the filter coil with a good signal-to-noise ratio.

Claims

1. A measuring device for detecting electromagnetic or magnetic alternating fields with an electromagnetic oscillating circuit, the device comprising a pickup coil (2), a tunable capacitance (4) and a filter coil (6), wherein at least one of the tunable capacitance (4), and filter coil (6), and the oscillating circuit have quality that is greater than 100 and wherein the quality of the filter coil (6) and the quality of the tunable capacitance (4) are at least two times greater than the quality of the pickup coil (2).

2. The device of claim 1, wherein the filter coil (6) of the oscillating circuit comprises a grounded center tapping (8).

3. The device of claim 1, wherein the filter coil (6) comprises two electric conductors that are twisted together, which are wound into the coil, wherein one respective end of each conductor is grounded.

4. The device of claim 1, wherein the filter coil (6) comprises a magnetic core (5), which in particular consists of ferrite.

5. The device of claim 1, wherein the filter coil (6) is shielded magnetically and electrically by a shielding (7).

6. The device of claim 5, wherein the shielding (7) is double-walled.

7. The device of claim 5 or 6, wherein the distance between the shielding (7) and the pickup coil (2) amounts to at least 1 cm.

8. The device of claim 1, further comprising a transfer line (3) between pickup coil (2) and capacitor as well as filter coil (6), which consists of two electric conductors that are wound around each other.

9. The device of claim 1, further comprising an excitation coil (12) for conducting EPR or NMR spectroscopy, which is separated from the pickup coil (2).

10. The device of claim 1, wherein the capacitance (4) is formed of a plurality of discrete capacitors that are connected in parallel.

11. The device of claim 1, wherein a differential amplifier (9), in particular an instrument amplifier, allows amplifying the voltage present at the filter coil (6).

12. The device of claim 11, wherein the differential amplifier is high resistance, specifically greater than 10.sup.6 ohm.

13. The device of claim 1, further comprising an outer shielding (10), in which the electric oscillating circuit comprising the pickup coil (2), the tunable capacitance (4) and the filter coil (6) are located and in particular moreover also an excitation coil (12), a transfer line (3) and/or an amplifier (9).

14. A method for detecting weak electromagnetic signals from a sample (1), the method comprising providing a measuring device for detecting electromagnetic or magnetic alternating fields with an electromagnetic oscillating circuit, the measuring device including a pickup coil (2), a tunable capacitance (4), and a filter coil (6), wherein at least one of the tunable capacitance (4) and filter coil (6) and the oscillating circuit have high quality that is greater than 100 and wherein the quality of the filter coil (6) and the quality of the tunable capacitance (4) are at least two times greater than the quality of the pickup coil (2), arranging the sample (1) within or on the surface of the pickup coil (2), locating the sample (1) in a static magnetic field (B.sub.0), and exciting the sample (1) via a coil (12), which is separated from the pickup coil (2).

15. The method of claim 14, further comprising measuring a signal with a frequency of 1 kHz to 40 MHz.

Description

[0036] It shows:

[0037] FIG. 1: Setup of an embodiment

[0038] FIG. 2: Electric conductors that are twisted together for a filter coil

[0039] FIG. 3: Signal-to-noise ratio of 0.5 cm.sup.3 benzene as a function of the number of windings nl of the pickup coil respectively input coil

[0040] FIG. 4: Signal-to-noise ratio of benzene as a function of the number of windings nl of the pickup coil under the use of filter coils with ferrite cores

[0041] FIG. 5: Signal-to-noise ratio of .sup.1H-NMR signals as a function of the volume Vs of a benzene sample

[0042] In the following, a preferred embodiment is shown in FIG. 1.

[0043] A pickup coil 2 of any structural form that is wound around a sample 1 is connected to an external high-quality filter coil 6 through a low-loss transfer line 3 and an tunable capacitance (capacitor) 4 of high quality (Q.sub.c˜10000). The transfer line comprises two electric conductors, which each consist either of a multi-wired copper line with large cross section (>1 mm.sup.2) and minimal capacity dielectrical loss due to a Teflon insulation or of thick low-resistance stranded wire (1000×0.05 mm) with negligible skin and proximity effect at frequencies below 1 MHz. The both conductors are wound around one another as shown, thus twirled or twisted together, respectively. In case of a Teflon insulation, every conductor is encased by Teflon. Additionally, the both conductors of the transfer line 3 that are twisted together are located in a Teflon casing.

[0044] The oscillating NMR signal that is recorded by the pickup coil (so called free induction decay, FID) is conducted via the transfer line 3 with low loss to the filter coil 6 with or without core 5. Due to the high quality Q.sub.E of the filter coil, the NMR signal is compared to the noise by the reducing factor Q.sub.red=Q.sub.E/(1+R.sub.I/R.sub.E)>>100 inflated with R.sub.I=alternating current resistance of the pickup coil and R.sub.E=alternating current resistance of the filter coil. The alternating voltage at the filter coil is conducted through two electric conducting connections into the inverting and non-inverting entrance of a differential amplifier 9 and further amplified there. In order to suppress the oscillating tendency of the very sensitive resonator in conjunction with the differential amplifier 9, a grounded center tapping 8 at the filter coil 6 is provided. Furthermore, the center tapping 8 provides for the draining of the technically unavoidable bias currents of the differential amplifier 9.

[0045] The filter coil 6 is separately magnetically shielded by means of a shielding 7.

[0046] Possible embodiments of filter coils with a quality Q.sub.E>>100 are for example specially wound low-loss ferrite cores, toroidally or cylindrically shaped coils with winding of stranded wire.

[0047] Especially during the use of ferrites, the shielding 7 consists of a simple or several fold Mu-metal shielding, which shields any DC and/or AC magnetic fields. On the other hand, the for her part heavy magnetic Mu-metal shielding should not affect the homogeneity of the B.sub.0 field. This is achieved by a sufficiently large distance between sample 1 and the Mu-metal shielding 7. The transfer line 3 is therefore at least ca. 10-50 cm long.

[0048] For toroidally or cylindrically shaped filter coils 6 without magnetic core 5, an electric shielding 7 is used for example made of copper sheet in order to shield electromagnetic disturbances in the frequency range of interest. The shielding 7 consists in this case in one advantageous embodiment of among each other electric insulated segments of copper and is advantageously composed such that eddy currents are avoided, which could reduce the Q.sub.E. Eddy currents are avoided in particular by appropriate arranged slits.

[0049] At a cylindrically shaped filter coil 6 without magnetic core 6, the critical eddy current would for example be a circular current about the cylindrical axis of the filter coil 6, which considerably reduces the inner magnetic field of the filer coil 6. These circular currents cannot flow, if the cylindrically shaped Cu shielding, which includes the filter coil 6, consists of two halves, which do not electrically touch each other along the shell of the cylinder. Further slits of the shielding 7 in order to suppress further circular currents on the surface of the shielding cylinder are advantageously provided in the sense of an improved quality Q.

[0050] For the purpose of optimizing the signal-to-noise ratio, the differential amplifier 9 is advantageously configured such that in the considered frequency range, the current and voltage noise (i.sub.n, e.sub.n) is lower than the Johnson noise (background) of the input circuit. The amplified NMR signal can be either directly digitally sampled via a fast A/D converter or analogously further processed via a lock-in amplifier 11 and analyzed via a computer. The analysis system can be equipped with a bandpass filter for the purpose of improvement of the signal-to-noise ratio.

[0051] Depending on the application and the environmental conditions, the pickup coil 2 can be a cylindrical inductor (coil), a saddle coil or a surface coil with dimensions of 0.01-10 cm. The quality and shape of the pickup coil 2 does thereby play only a minor role for the signal-to-noise ratio. This is one of the big advantages of the present invention, which allows to choosing the shape and design of the pickup coil 2 entirely freely. This is important for the feasibility of NMR examination under difficult conditions. The sample 1 with the nuclear spins is located inside or on the surface of the respective pickup coil 2, depending on whether it is a cylindrical inductor (coil), a saddle coil or a surface coil.

[0052] An excitation of the nuclear spin occurs by means of a larger saddle coil 12 in conjunction with a pulsed high frequency generator 13 and a high frequency amplifier 14. On the one hand, the saddle coil 12 is far enough (˜1 cm) away from the pickup coil 2 for the purpose of minimal coupling, and on the other hand, the main axis of the saddle coils (=direction rf field) is arranged orthogonally to the sensitive axis of the pickup coil 2.

[0053] The excitation of the sample 1 with a magnetic alternating field from the saddle coil 12 is practically entirely decoupled from the entire reception circuit. Therefore, the reception circuit does not require any more to be tuned to a specific impedance and can be optimized to a best possible signal-to-noise ratio independent from the transmission coil. The sample 1 is located in the homogenous magnetic field area B.sub.0 of a not shown electromagnet or Hallbach magnet, which can generate a typical field strength in the range of B.sub.0=10.sup.−4 T−1 T. This field range corresponds for protons to a range of the Larmor frequency of f=4 kHz-42 MHz. The components 1 to 9 are electrically two-fold shielded by means of two copper and/or aluminum shields 10 that are nested into each other.

[0054] In the embodiment shown in FIG. 1, the transmission circuit is decoupled from the reception circuit. The reception circuit is physically separated (spaced apart) in an area of the signal pickup (pickup coil 2 with sample 1) and an external resonator of high quality. The decoupling of the transmission from the reception circuit allows to optimize the reception circuit separately without consideration of an impedance adaptation between transmission and reception circuit. In particular, the reception circuit can advantageously have for example several MOhm impedance, which is in particular the case at parallel resonance circuits with very high quality. Due to the second condition, the pickup coil can be freely designed independent from all other elements and does not need to have high quality. The external resonator can be optimized for highest quality and therefore for maximal amplification of the NMR signal.

[0055] In the FIG. 2, two electric conductors 15 and 16 that are twisted together are shown, as they can be used for the transfer line 3 or for the filter coil 6. In case of filter coil 6, the wires 15 and 16 that are twisted together are wound into the coil. One end of the one electric conductor 15, such as for example the in the FIG. 2 left shown end, and the opposite end of the other conductor 16, thus then the right shown end, are electrically connected together and thereby form the center tapping 8, which is grounded. The remaining ends of the electric conductors 15 and 16 are then connected on the one hand with the preferably tunable capacitor 4 and on the other hand with an electric conductor of the transfer line 3.

[0056] A filter coil 6 with ferrite core 5 has the big advantage that it is relatively small and low-priced and qualities up to Q.sub.E˜500 are realizable. Ferrite cores are commercially available in different sizes and structural shapes, for example as toroid or in closed shapes (Pot circle; coaxial cavity resonator) in order to keep the magnetic stray losses as low as possible.

[0057] Two interwoven copper wires of equal length are in one embodiment wound with few windings around the ferrite core, wherein the number of windings n.sub.E depends on the detection frequency, typically of n.sub.E=8 at 8 to 500 kHz up to n.sub.E=104 at 20 kHz. The intertwined winding pair of multi-wired copper or braid wire minimizes skin and proximity effect and the therewith connected AC resistance considerably. A center tapping 8 is led out exactly in the middle between both interwoven winding halves. This avoids a possible oscillating tendency of the high-quality overall circuit during grounding. All filter coils with ferrite cores react extremely sensitive to external magnetic and electromagnetic fields and are therefore preferably shielded with a magnetic two-fold shielding made of Mu-metal (shielding factor˜1000). Because the both Mu-metal shields are also electric conductive, they shield besides from magnetic fields also from electromagnetic alternating fields. Different filter coils with ferrite cores were manufactured at five different frequencies with the following qualities: Q.sub.E=218 at 500 kHz, Q.sub.E=380 at 166 kHz, Q.sub.E=369 at 83 kHz, Q.sub.E=280 at 41 kHz and Q.sub.E=255 at 20 kHz. The AC-resistance, the inductance, capacitance respectively the quality of all filter coils as well as all relevant components such as transfer line and tuning capacitors were measured with an impedance spectrometer as a function of the frequency and optimized. Protons (.sup.1H) experiments with all five ferrites were conducted with a pre-polarized benzene (benzol) sample (V.sub.S˜0.3 cm.sup.3, N.sub.S=2×10.sup.22 protons, P.sub.n˜1.5×10.sup.˜6). The pickup coil 2 was characterized by the parameter D.sub.i=H.sub.c=1 cm, W.sub.I=63, d.sub.Cu=0.12 mm, and by their overall winding number n.sub.I. The measured signal-to-noise ratio (SNR) as a function of n.sub.I for the filter coil with ferrite at 500 kHz is shown in FIG. 3. The measurement points (black circles) are in good agreement with the theoretic expectation (continuous line). The parameters for the ferrite core at 500 kHz are: L.sub.E=116 μH, R.sub.I=1.77 Ohm and Q.sub.E=218. FIG. 3 clearly shows the initially linear increase of the signal-to-noise ratio with n.sub.I, then a maximum at n.sub.I.sup.max˜90-100 with signal-to-noise ratio=170, and finally the hyperbolic decrease, which is in agreement with a dominating skin effect and very weak proximity effect (˜3% of skin effect).

[0058] FIG. 4 shows the signal-to-noise ratio of benzene as a function of the number of windings n.sub.I of the pickup coil under use of filter coils with ferrite cores and after pre-polarization to P.sub.n=1.5×10.sup.−6. Circles represent measured values and continuous lines correspond to the expected course. The frequencies f (and quality Q.sub.E) amount to 500 kHz (Q.sub.E=218, e), 166 kHz (Q.sub.E=280, a), 83 kHz (Q.sub.E=369, b), 41 kHz (Q.sub.E=380, c) and 20 kHz (Q.sub.E=250, d). The parameters of the filter coil are equal as in FIG. 3. FIG. 4 confirms based on experimental results, which were measured at the five different frequencies 500, 166, 83, 41 and 20 kHz, that at constant quality of the filter coil, the signal-to-noise ratio only depends very weakly from the frequency. When decreasing the frequency f, the measured maximum of the signal-to-noise ratio shifts towards higher values of n.sub.I. From the maxima of the five curves the winding number n.sub.I.sup.max can be read: n.sub.I.sup.max=90, 140, 350, 750 for f=500, 166, 83, and 41 kHz. The most important result of FIG. 4 is that the maximal signal-to-noise ratio decreases very weakly with the frequency. This means that measurements can be conducted with ferrite cores in the entire frequency range of some kHz (NMR in the earth's magnetic field) up to several MHz with high signal-to-noise ratio.

[0059] FIG. 5 shows the signal-to-noise ration of .sup.1H-NMR signals as a function of the volume V.sub.s of a benzene sample. It was measured after pre-polarization to P.sub.n=1.5×10.sup.−6 at 500 kHz with a filter coil with ferrite core with Q.sub.E=218. The crosses correspond to measurements, which were all conducted with the same input coil (D.sub.I=H.sub.C=1 cm) but with different sample volumes (0.7-500 μL). The rhombi correspond to measurements, at which input coil and sample are shrunk together by the same factor (1 mm<D.sub.i=H.sub.C<1 cm). The straight lines correspond to the expected run.

[0060] FIG. 5 illustrates based on two .sup.1H-NMR experiments of benzene, how the signal-to-noise ratio is scaled with the volume of the sample. At the first experiment, the signal-to-noise ratio was measured at 500 kHz with a filter coil with ferrite core (Q.sub.E=218) and with a fixed predefined pickup coil (parameter: D.sub.i=H.sub.C=1 cm, n.sub.I=160, d.sub.CU=0.12 cm), but with different sample volumes V.sub.S (of 0.7 to 500 μL). At the second experiment, corresponding with the shrunk volume of the sample, also the volume respectively inner diameter D.sub.i and the height H.sub.C of the pickup coil were reduced, from D.sub.i=H.sub.C=1 cm to D.sub.i=H.sub.C=0.1 cm. The winding number n.sub.I=80 was here kept constant. It can be seen in FIG. 5 that the measured signal-to-noise ratio scales in the first case with signal-to-noise ratio ˜N.sub.S˜V.sub.S and in the second case with signal-to-noise ratio ˜N.sub.S.sup.1/2˜V.sub.S.sup.1/2 (N.sub.S=number of the spins). This weak volume dependency in the second case is reducible to the increased sensitivity ˜1/D.sub.i of the ever smaller becoming pickup coil. This result shows that with this method micro coil detected NMR spectroscopy at low frequencies and with μL size samples is possible. This is an important result in view of a miniaturization of mobile NMR spectroscopes. At frequencies below 10 MHz and for very small samples, the magnet, which generates the B-field, may be very small, portable, energy and cost efficiently manufactured. Energy efficiency means for example that a lithium-ion battery suffice to supply a small electro magnet with current. Cost efficiency manifests among others also during miniaturized permanent magnets, which need for their construction a comparatively far lower amount of magnetic material (rare earth).

[0061] At this point, it is also mentioned that the measured scaling signal-to-noise ratio ˜V.sub.S.sup.1/2 only applies for a winding number n.sub.I that is kept constant.