On the possibility of using nuclear magnetic resonance to detect explosives on the human body.

On the possibility of using NQR to detect explosives on the human body.

GRECHISHKIN Vadim Sergeevich, Doctor of Physical and Mathematical Sciences, Professor
SHPILEVOI Andrey Alekseevich, Candidate of Physical and Mathematical Sciences, Associate Professor
BURMISTROV Valery Ivanovich

ON THE POSSIBILITY OF USING NQR TO DETECT
EXPLOSIVES ON THE HUMAN BODY

In the last decade, due to the increased threat from terrorist organizations and the increase in the number of local military conflicts in developed countries, large funds have been allocated to the development of various methods for detecting and identifying explosives.

This work is mainly carried out in two directions: humanitarian demining and the development of baggage control systems.

However, recently the problem associated with the detection of explosives on the human body has become increasingly acute.

And here it turns out that the methods more or less successfully used for the purposes of baggage control and detection of hidden explosives in soil and buildings [1 – 4] are not very effective or are not applicable at all for use in this case.

It is clear that the methods of X-ray and neutron analysis are not applicable due to their danger to human health. The use of radio wave detectors and non-linear radars is also questionable.

In addition to their insufficient noise immunity, it is necessary to take into account the fact that microwave radiation of such intensity can both harm human health and affect the performance of surrounding equipment.

The use of gas analytical devices and chemical express tests requires the presence of a certain amount of explosives in the air or on the surface being examined, which greatly limits their use.

The effectiveness of using mine detection dogs depends on many difficult to control factors and does not allow achieving the desired result, especially when vacuum packing explosives in a plastic bag.

In this regard, devices created on the basis of the phenomenon of nuclear quadrupole resonance (NQR) [5] have a number of undeniable advantages [6].

They do not require direct contact with explosives, the intensity and frequency of electromagnetic radiation required to excite the sample under study are low and do not affect human health or the performance of electronic devices.

The NQR frequency is unique for each chemical compound and is virtually independent of the admixture of other chemicals, a slight change in temperature or humidity.

With proper design of equipment, the minimum mass of the substance to be detected can be only a few tens of grams with a detection reliability of 97% [7].

The state of affairs in this area of ​​research also does not correspond to modern requirements.

There are devices for checking baggage and mail: QR-160, QR-500, Qscan 2000, manufactured by Quantum Magnetics (USA), NQR mine detectors developed by Logis and the State Enterprise Research Institute of Instrument Engineering named after V.V. Tikhomirov. Quantum Magnetics tested a NQR mine detector in 1998-2000.

Thus, the research directions are the same – humanitarian demining and baggage control, although the NQR method could be one of the most effective in detecting explosives on humans.

The following methods can be used to create a system for detecting explosives on the human body:

Local NQR [8];
NQR using a volumetric coil [9, 10];
Remote dual NQR (RDNQR) method [11].

In remote observation of NQR, open oscillatory systems are used for irradiating the system under study and recording its response: a flat spiral coil and coils wound on RF ferrites [12].

Since the exciting coil is located at some distance from the sample under study, the amplitude of the exciting RF field at its location, especially under conditions of strong interference, can be comparable to the noise level.

In this case, extraneous electromagnetic fields with frequencies close to the NQR frequency can have a significant effect on NQR signals.

To clarify the nature and extent of these changes, let us consider a NQR system with spin I = 1 (nitrogen nuclei 14N), excited by an RF field and noise simultaneously.

Noise is taken into account both during the action of RF pulses and in the time intervals between them, when the system tends to an equilibrium state. White Gaussian noise was chosen as a noise model, since in this case it is possible to perform averaging [13] and solve the problem analytically for single-frequency excitation of the spin system under study.

To solve the Neumann equation at the moments of action of RF pulses, the canonical transformation method (transition to a “rotating coordinate system”) was used together with classical averaging. In the intervals between RF pulses, when the system relaxes against the background of noise, it is necessary to use the method of fictitious spin operators [14, 15].

As a result, expressions for the induction signals (SI) and echo signals (SE) were obtained. No averaging was performed over different directions in the powder.

In order for the induction and echo signal amplitudes to be maximal at a given RF field amplitude, it is necessary to use pulses of a certain (optimal) duration.

The time interval between pulses must be greater than T2, so that by the time the next pulse acts, the signals from the previous one are attenuated.

On the other hand, this time interval should be significantly less than the irreversible attenuation times T1, T2, otherwise echo signals will not be observed.

Thus, having selected the optimal values ​​of the durations of the excited pulses and the time intervals between them, it is possible to observe NQR signals. However, in the presence of noise, these simple patterns cease to be fulfilled.

Let's consider hexogen as an example.

At room temperature, the transition frequency n+ = 5192 kHz, relaxation times T1 = 10 ms, T2 = 8 ms, RF field strength B1 = 1.3 mT, and the duration of the interval between pulses t = 1 ms.

After the first RF pulse of duration t1 has ended, a decaying free induction signal is observed.

In the absence of noise, the optimal pulse duration is t1opt = 31 μs.

If this pulse excites the test substance in the presence of noise, the signal amplitude will depend on the relative magnitude of the noise spectral power (NSP) X = S/B1, where B1 is the amplitude of the magnetic field induction in the RF pulse, and S is the spectral density of the noise power. This dependence is shown in Fig. 1.

Fig. 1. Dependence of the SI amplitude on the noise level

It is evident that as the noise level X increases, the SI amplitude begins to decrease. However, this decrease becomes noticeable only when the noise level is comparable to the amplitude of the RF field in the pulse.

This occurs because the value of t1opt changes in the presence of noise (Fig. 2).

Fig. 2. Dependence of the optimal duration of the excitation pulse on the noise level

However, a noticeable change in the optimal pulse duration also occurs only at a sufficiently high noise level.

If, when changing the noise, the pulse duration is changed so that it is always optimal, then the amplitude of the induction signal remains unchanged (Fig. 3).

Fig. 3. Dependence of the SI amplitude on the noise level at optimal pulse durations

The optimal values ​​of the first and second pulse durations for the spin echo signal in the absence of noise are t1opt = 31 μs, t2opt = 62 μs.

The dependence of the signal amplitude on the noise for these values ​​of the excited pulse durations is shown in Fig. 4.

Fig. 4. Dependence of the SE amplitude on the noise level

It is evident that the amplitude of the SE can vary within very wide limits with an insignificant (on the order of several percent) change in the relative value of the noise dispersion. With a significant increase in the noise level, the signal value decreases significantly.

As for the induction signal, this is due to the dependence of the optimal durations of the first and second excited pulses on the noise (Fig. 5).

a)

б)
Fig. 5. Dependence of the optimal duration of the first (a) and second (b) pulses on noise

Based on this, we can conclude that such behavior of the SE is due to a more complex dependence of the optimal durations of excitation pulses on noise.

Unlike the SI, if the pulse durations are changed so that they are always optimal when the noise changes, the SE amplitude will still change very strongly with a small change in noise (Fig. 6).

Fig. 6. Dependence of the SE amplitude on the noise level at optimal pulse durations

The dependences of the SE amplitude on the time interval between pulses for different values ​​of X at t1opt = 31 μs, t2opt = 62 μs are shown in Fig. 7.

a)

б)

c)
Fig. 7. Dependence of the SE amplitude on t for different noise values: X = 0 (a); X = 0.025 (b); X = 0.1 (in)

In the presence of noise, the dependence of the SE amplitude on t exhibits oscillations, the frequency of which increases proportionally to the increase in the noise level. This leads to the fact that even with correctly selected pulse durations taking into account the noise and with an unsuccessfully chosen value of t, the signal will not be observed.

This is explained by the fact that in the presence of noise, the optimal durations of the excitation pulses t1 and t2 and the value of t between them are related by a complex dependence, the nature of which changes with a change in X (Fig. 8, 9).

a)

b)
Fig. 8. Dependence of t1opt on t for different X: X = 0.01 (a); X = 0.1 (b)

a)

b)
Fig. 9. Dependence of t2opt on t for different X: X = 0.01 (a); X = 0.1 (b)

Such different dependence of induction and echo signals on noise is explained as follows. When SI is formed, noise affects the irradiated sample for a fairly short period of time equal to the duration of the RF pulse, which has a value of the order of several tens of microseconds.

Therefore, the influence of noise becomes noticeable only at its large values.

The echo signal is excited by two short pulses separated by a time interval t of the order of units to tens of milliseconds.

The signal itself appears after a time t after the end of the second pulse.

The effect of noise during these two fairly long time intervals t determines the more complex and strong dependence of the echo signal on noise.

Thus, in order to minimize the effect of noise on NQR signals, it is necessary to use the shortest and most powerful RF pulses and shorter intervals between them.

The main disadvantage of NQR is the low signal intensity. To overcome this, signal accumulation is usually used.

To reduce the time spent on detecting NQR signals during accumulation, multi-pulse sequences are used [16].

From the above reasoning, it is clear that the influence of noise on the parameters of NQR signals obtained as a result of using multi-pulse sequences will be even more complex and more significant than on single echo signals.

Thus, it turns out that the amplitudes of local NQR signals depend on parameters whose values ​​cannot be controlled under conditions of significant interference. The slightest change in the external electromagnetic environment can lead to such a change in these parameters that the amplitude of the NQR signals will become close to zero.

In addition, isolating NQR signals from noise at a high noise level also presents a serious problem [17].

Based on the above, it can be argued that the local NQR method can hardly be used to create a device for detecting explosives on the human body, which should operate at airports, railway stations, etc., i.e. in places with an increased RF background.

The equipment for remote NQR was described in [11].

The quadrupole moments of light nuclei, such as nitrogen 14N, are significantly smaller than those of heavy nuclei, and their quadrupole transitions, accordingly, are located in a lower frequency region [18].

This circumstance often turns out to be significant, since the sensitivity of direct NQR methods decreases with frequency and already at n = 1 MHz the observation of absorption lines is associated with significant difficulties.

In addition, the composition of four main explosives: TNT, RDX, HMX and PETN includes NO2 groups. The frequency range of 14N NQR from NO2 groups in organic compounds falls into the range of medium-wave radio broadcasting stations, which leads to a decrease in the sensitivity of the direct detection method.

The situation is further complicated by the presence of various chemically nonequivalent nuclei in the sample and crystalline nonequivalence, which leads to line multiplicity. In addition, NO2 groups in compounds included in explosives, such as TNT, can rotate at room temperature, broadening the NQR line [19].

All this leads to the impossibility of observing NQR signals by the direct method even with accumulation and even in the absence of noise. In this situation, the remote DNQR method, which is characterized by high sensitivity, is very suitable.

To demonstrate this advantage, the spectra of trinitrotoluene and ammonium nitrate were obtained at room temperature after five accumulations (Figs. 10, 11).

Fig. 10. TNT spectrum at room temperature after five accumulations

Fig. 11. Spectrum of ammonium nitrate at room temperature after five accumulations

The quadrupole coupling constants eQqzz and the asymmetry parameter h for these compounds, calculated from the obtained spectra, are given in Table 1.

Table 1. NQR frequencies and quadrupole coupling parameters for TNT and ammonium nitrate at room temperature

In addition, since the NQR signal is registered indirectly by the change in the NMR signal from protons, it can be assumed that the effect of noise during excitation of the quadrupole system will be much less than in the case of the direct method, which is due to the high sensitivity of double resonance. This reduces the requirements for shielding the transceiver system from external electromagnetic influences compared to the direct method.

Thus, it can be argued that the broad capabilities of double nuclear quadrupole resonance methods in detecting narcotic and explosive substances clearly justify the costs and difficulties associated with the design of equipment of this kind.

The works [10, 20] describe NQR equipment for detecting explosives and narcotics in baggage.

It was shown that when using a direct pulse method with a 110-liter coil at a power in the RF pulse of P = 5 kW, 30 g of hexogen is detected in a few seconds. Based on a device designed for baggage control, it is possible to quickly and inexpensively make a device that allows for the effective detection of explosives on the human body.

The basis of such a device is a coil with a volume of 110 — 160 liters, the axis of which is perpendicular to the surface of the earth. This coil is the inner part of a cylindrical cabin where a person is placed.

The outer part of this cabin shields the coil from the effects of external RF fields.

Since the detection time of explosives in such a device does not exceed several tens of seconds, and the device itself has a fairly high sensitivity, this method will allow for a quick and mass check of passengers for explosives and drugs, which is necessary due to the current situation in the country.

Since there are already serially produced NQR devices for baggage control, the development of the electronic part of this device will not take much time.

Tests conducted with a 110-liter solenoid coil in the screen made it possible to detect 30 g of hexogen on the body of a person who entered the device in 10 seconds.

Thus, based on the above considerations, the following can be stated. For the creation of a device designed to detect explosives on the human body that give a sufficiently strong NQR signal (hexogen, octogen), the direct pulse NQR method using a well-shielded volume coil is most suitable.

The remote DNQR method can be used in cases where the sensitivity of the direct method is not sufficient to detect explosives, such as trinitrotoluene or drugs.

However, further research is required to create a device based on remote DNAR that would work in real conditions.

Literature.

1. Petrov S. I. On the assessment of the possibility of detecting explosives and devices containing them.//Special equipment, 2001, No. 4, pp. 16 – 20.
2. Petrenko E. S. Means of searching for explosive objects by indirect signs.//Special equipment, 2002, No. 2, pp. 28 – 34.
3. Petrenko E. S. Some Features of Searching for Explosives and Explosive Objects Using Dogs, Gas Analytical Devices, and Chemical Express Tests.//Special Equipment, 2002, No. 4, pp. 20–24.
4. Shelkov V. A. Contactless Method for Detecting Explosives and Narcotics.//Special Equipment, 2000, No. 6, pp. 48–49.
5. Grechishkin V. S. Nuclear Quadrupole Interactions in Solids. Moscow, Nauka, 1973.
6. Bely Yu.I. et al. Equipment for combating terrorism based on the effect of nuclear quadrupole resonance.//Special equipment, 2002, No. 2, pp. 35 — 39.
7. Grechishkin V.S. News of universities. Physics. 1992, No. 7, pp. 62 — 65.
8. Anferov V.P., Grechishkin V.S., Sinyavsky N.Ya. Nuclear spin resonance. New methods. L., Leningrad State University, 1990.
9. Grechishkin V.S., Grechishkina R.V., Emelianov O.S. News of universities. Physics, 1996, No. 10, p. 98 – 99.
10. Grechishkin V. S. J. Appl. Phys. 1994, v. A58, p. 63 – 65.
11. Grechishkin V. S., Grechishkina R. V., Shpilevoy A. A., Persichkin A. A., Heo Hong. Optics and Spectroscopy, 2003, Vol. 94, No. 3, pp. 392–393.
12. Grechishkin V. S., Sinyavskii N. Ya. Usp. Phys., 1993, Vol. 163, No. 10, pp. 95–119.
13. Akhmanov S. A. Introduction to Statistical Radiophysics and Optics, Moscow, 1981.
14. Wokaun A., Ernst R. R. Selective excitation and detection in multilevel spin systems: Aplication of single transition operators, J. Chem. Phys., 1977, 67, p. 1752 – 1758.
15. Vega S. Fictitious spin ? operator formalism for multiple quantum NMR, J. Chem. Phys., 1978, 68, p. 5518 – 5527.
16. Ermakov V. L., Osokin D. Ya. Radiospectroscopy, 1985, Issue 16, pp. 31 – 44., Perm, PSU.
17. Grechishkin V. S., Anferova L. V. Using the Bohr principle for NQR signals in mine clearance.//Special Equipment, 2004, No. 3, pp. 42 49.
18. Grechishkin V. S., Shpilevoy A. A. Uspekhi Fizicheskikh Nauk, 1996, No. 7, pp. 763–776.
19. Grechishkin V. S., Sinyavskii N. Ya., Mozzhukhin G. V. Izvestiya vuzov. Physics. 1992, No. 7, pp. 58–61.
20. Grechishkin V. S. Appl. Phys. 1992, v. A55, pp. 505–507.