PRODUCTION OF THE SHORT-LIVED ISOTOPES 12 N AND 12 B

1. PRODUCTION OF THE SHORT-LIVED ISOTOPES 12N AND 12B IN THE 14N(,2n), 14N(,2p), AND 13C(,p) REACTIONS L.Z.Dzhilavyan1, A.I.Karev2, V.D.Laptev1, V.G.Raevsky2 1 Institute for Nuclear Research, Russian Academy of Sciences, Moscow, Russia 2 P.N.Lebedev Physical Institute, Russian Academy of Sciences, Moscow, Russia

2. The theme of this report is connected with 12N-nuclei. • For the first time their production in the reaction 12C(p,n) was found by L.W.Alvarez at pulsed proton linac. • Soon W.K.H.Panovsky and D.Reagan also produced them at pulsed electron linac in Stanford, using the photonuclear reaction: • n (IA:14N-99.63%,15N-0.37%; E thresh30.6 MeV) (I)

3. 1. The production and decay features of 12N-nuclei • Decay of 12N and ways of their production have a remarkable set of features (especially, when 12N are produced at electron accelerators): • 12N-nuclei have a rare short lifetime T1/211.0 ms, what helps to separate 12N-activity from another ones and to do it fast. • 12N-nuclei decay emitting -particles. It is possible to use “coincidence registration” of pairs of outgoing annihilation '-quanta (with E'0.511 MeV each, flying out in opposite directions) and thus to suppress effectively big part of background. • 12N-nuclei decay, emitting -particles with the very high E max 17 MeV. • 12N are the neutron-deficient nuclei with A  2(Z1)(while, except for 1H and 3He, for all stable nuclei A  2Z). This restricts strongly choice of reasonable reactions for production of 12N. • It is possible to apply photoproduction of 12N for rather thick objects (having connection with concentrations of nitrogen). • If an incident beam of γ-quanta has small divergence and “spot” on irradiated objects, then we have a common way for localization of N-concentrations. • If for detection of 12N-decays registration of (γ'-γ')-coincidences of annihilation ‘-photons is used, we may use it also to find localization for N-concentrations (together with the common method with achieving better accuracy and suppressing of background).

4. 2. The first variant of connected with 12N detectors of hidden explosives Alvarez suggested the idea, which unfortunately was not yet experimentally confirmed, to use for a photonuclear detector of hidden explosives (DHE) “in coincidence–registration” of annihilation '-quanta from decays of 12N, produced on a pulsed electronaccelerator in the reaction (I) on 14N, contained in big concentrations in widely used now explosives. It is supposed that this registration takes place for ~100 ms or even less between accelerator’s pulses. We call such DHE as DHE-1. DHE-1 was propose for the external not-destroyingcontrol of air-passengers’ baggage. But there is also a big interest to DHE for humanitarian demining and for testing of cargo. Unfortunately, for the latter 2 cases DHE-1 does not suit. For landmine searching it is impossible to register both annihilation γ'-quanta. For cargo containers there is too big absorption of outgoing '-quanta.

5. 3. The second variant of connected with 12N detectors of hidden explosives • At the further development of a photonuclear DHE, there was suggested to register single '-quanta. In this case there were made the moderately successful, but encouraging experimental tests, including our own ones,. We call here such DHE as DHE-2. The purpose of DHE-2the universal DHE for all 3 pointed out applications. DHE-2 can work at E' up to several MeV with significant decreasing (without taking into account advantages from the coincidental technique) of the corresponding in energy gamma-background level, with essential reduction of absorption of required '-quanta in irradiated objects and with an opportunity to use for '-quanta registration Čerenkov detectors with "cutting off" the big part of gamma-background at small energies. In this case for providing with useful signals there is some addition from the reaction (II), which produces also on 14N radioactive nuclei 12B (-decay; E max13 MeV; T1/220.4 ms [1]), having T1/2 and values of decay E' for registration, close to those for 12N-nuclei: • Bp (E thresh25.1 MeV). (II)At electron accelerators with Ee55 MeV for outgoing electrons (maximum for the pulsed race track microtrone, which is now under construction in LPI, for these studies), there are few reactions, which may produce radioisotopes with 7 ms T1/2 30 ms, and with E', close to those for 12N and 12B. Except for the reactions (I,II) such nuclei may be produced in some background reactions under bombardment by primary bremsstrahlung photons or by secondary neutrons, produced before in photoneutron reactions.

6. Among nuclei-products of these background reactions there may be(taking into account isotopic abundances, values of reaction thresholds, and, to some extents, chemical abundances and branching ratios for these reactions): • as 12N and 12B, but produced on some other nuclei-targets (not on nitrogen), and among these types of reactions there may be: the photonuclear reactions: • 13CBp (IA: 12C-98.89%, 13C-1.11%; E thresh17.5 MeV). (III) • 6ON3Hn (IA:16O-99.76%; Ethresh45.1 MeV), (IV) • 9F2N3n (IA: 19F-100%; E thresh45.6 MeV); (V) the reaction under thermal neutrons: • n11BB (IA: 11B-80.2%; the released energy 3.4 MeV); (VI) the reaction under fast neutrons(En thresh–neutron kinetic energies En at threshold): • nCBp (En thresh12.6 MeV); (VII) as some other nuclei-products, and among these reactions there may be: • 6O3B3p (-decay; E thresh43.2 MeV; T1/217 ms), (VIII) • 6O3O3n (-decay; E thresh52.1 МэВ; T1/29 ms). (IX) • For DHE-2 there may be serious background in several ms after beam, from (n,')-reactions. On the other hand, there may be background from giving radioisotopes with 0.1 sT1/21 s reactions, in particular, with 3 nucleon escape, especially from: • 2C9Li3p (-decay; E thresh46.8 MeV; T1/2178 ms), (X) • 2C9C3n (-decay; E thresh53.1 MeV; T1/2126.5 ms). (XI)

7. 4. The cross sections of the used reactions and some background reactionsTo optimize DHE, one needs data on , of reactions (I)(XI). The measured  of the reactions (III) and (VII) are presented on Fig. 1 and 2. Data on  of the reaction (VI) under thermal neutrons also may be found. Unfortunately, we did not find original experimental data on  of the reactions (I), (II), (IV), (V), (VIII)(XI).Fig. 1. Fig. 2.

8. The compiled data on  of the reactions (I)(III), (VIII), and (IX) were reported by W.P.Trower in 3 works. Data from the last of them are more complete as well; therefore, they are presented on Fig. 3. However, these data give rise to much criticism: 1) their sources are not indicated; 2) the data from these studies are contradictory; 3)  of the reaction (III) from the last work near their maximum are several times smaller than the well-consistent corresponding  from Fig. 1; 4) the claims for the „dynamic range” and validity of behavior at relatively high Ewith respect to  of the reaction (III), made in this work, are, at least, surprising (compare if only with the behavior of  (III) from Fig. 1, measured in the range from maximum of  to E30 MeV). A possible reason for the latter fact is that not measured  were reported, but some predicted ones. Fig. 3. 

9. So, we need experimental data on  of both useful reactions (I) and (II) and the background reactions (IV), (V), (VIII)(XI). • *(For convenience on this and the next pages we repeat the list of pointed out before reactions) • In this technique, some background may be eliminated due to the thresholds and characteristic E-ranges in  of the reactions (I) and (II) are much lower than those for the background reaction (IV), (V), (VIII)(XI) and another reactions, connected with escape of 3 nucleons and giving radioisotopes with 0.1 s T1/21 s. On the other hand, some background may be significantly reduced, because of the E thresholds and characteristic ranges of  for the reactions (I) and (II) are much higher, than those for reactions from the giant resonance range, including the (,p)-reaction (III) and reactions, giving the most of photoneutrons. • n (IA:14N-99.63%,15N-0.37%; E thresh30.6 MeV) (I) • Bp (E thresh25.1 MeV). (II) • 13CBp (IA: 12C-98.89%, 13C-1.11%; E thresh17.5 MeV). (III) • 6ON3Hn (IA:16O-99.76%; Ethresh45.1 MeV), (IV) • 9F2N3n (IA: 19F-100%; E thresh45.6 MeV); (V) n11BB (IA: 11B-80.2%; the released energy 3.4 MeV); (VI) • nCBp (En thresh12.6 MeV); (VII) 6O3B3p (-decay; E thresh43.2 MeV; T1/217 ms), (VIII) • 6O3O3n (-decay; E thresh52.1 МэВ; T1/29 ms). (IX) • 2C9Li3p (-decay; E thresh46.8 MeV; T1/2178 ms), (X) • 2C9C3n (-decay; E thresh53.1 MeV; T1/2126.5 ms). (XI)

10. The small fraction of B in typical tested objects permits to believe the background from the reaction (VI) to be not very interfering for this technique. The background from the reaction (VII) should be small because (i) the fraction of primary photoneutrons with EnEn thresh for this reaction is small and (ii) such neutrons with overwhelming probability first undergo scattering; in few of which their energy En decreases to values below En thresh. The reaction (VIII) is not dangerous due to its small . For background from the reactions (IV,V) situation seems to be the same. In principle, the reaction (IX) may give essential contribution, especially at searching of land-mines, since soil may contain up to 50% of O, but at Ee55 MeV this contribution should be small. All background contributions from reactions, giving long-lived isotopes (including the reactions (X,XI)) should be small too, and their influence may be taken into account by usual procedures of background subtraction. • n (IA:14N-99.63%,15N-0.37%; E thresh30.6 MeV) (I) • Bp (E thresh25.1 MeV). (II) • 13CBp (IA: 12C-98.89%, 13C-1.11%; E thresh17.5 MeV). (III) • 6ON3Hn (IA:16O-99.76%; Ethresh45.1 MeV), (IV) • 9F2N3n (IA: 19F-100%; E thresh45.6 MeV); (V) n11BB (IA: 11B-80.2%; the released energy 3.4 MeV); (VI) • nCBp (En thresh12.6 MeV); (VII) 6O3B3p (-decay; E thresh43.2 MeV; T1/217 ms), (VIII) • 6O3O3n (-decay; E thresh52.1 МэВ; T1/29 ms). (IX) • 2C9Li3p (-decay; E thresh46.8 MeV; T1/2178 ms), (X) • 2C9C3n (-decay; E thresh53.1 MeV; T1/2126.5 ms). (XI) • So, it seems that for DHE-2 the most serious background may be from the reaction (III), but, of course, it is very desirable to check the situation experimentally.

11. 5. The third variant of connected with 12N detectors of hidden explosives • “False signals”–one of the most important problem for all methods of explosive detection, which determines method’s efficiency, productivity, and practical importance. For DHE-1 or DHE-2 it is reduced to rejection of signals from another substances, containing like explosives N and/or C. • Earlier for DHE-2 there were suggested 2 ways for subtraction of connected with the reaction (III) background, based on its dependence on Е and Е'. In our opinion these ways have grave shortcomings. We suggest our own way. • N and C – the basic elements of widely-used explosives, and it is desirable to consider C- signals not as background, but as helpful ones in detecting and identification of explosives. • We suggested DHE-3, based on registration and analysis of time-distributions of events, connected with the reactions (IIII). Relative concentrations of 12N- and 12B-isotopes are unequivocally connected with relative concentrations of N and C in irradiated substances and define unique “portraits” of tested substances. Form of - event time-dependence is determined by values of T1/2 and by initial ratio of 12N- and 12B- produced quantities.

12. In common case a total number N{t} of 12N- and 12B- nuclei, which leave un-decayed to a moment t after an end of irradiation at t0, is: • N{t}  N0 (N-12) exp((N-12) t)  N0(B-12) exp((B-12) t), (1) • where: N0(N-12) , (N-12)(ln 2) / (T1/2)(N-12) and N0(B-12) , (B-12)(ln 2) / (T1/2)(B-12) – numbers of produced to a moment t  0 nuclei, the decay constants, and the half-lives for 12N and 12B, respectively; N0 N{t  0}  N0 (N-12)N0 (B-12). • Results of analysis of '-radiation time-distribution may give initial relative concentration of 12N and 12B and identify tested substance by parameter k: • kk(N-12)1 k(B-12), (2) • where: k(N-12){N0 (N-12) / N0 }, k(B-12){N0 (B-12) / N0  } (k-substance’s “portrait”). A measured in channel number of '-events from decay of 12N and 12B is : • n{t} (N-12)N0 (N-12)exp((N-12)t)t (B-12)N0 (B-12)exp((B-12)t)t, (3) • where t – a channel width for a measured time-distribution. • A common trouble for DHE-2 and DHE-3: in several ms after irradiation there is high '- background from reactions (n,'), initiated by photoneutrons. This background may give rise to serious distortions of results. There is a sense to use not all measured data on distributions n{t}, but only those which are after delay of several ms, with refusing from direct measurements of N0  and from direct calculations of k. Instead of that, we suggest to determine values of k, proceeding, for example, from two measured values n(t), for moments of time ti and tj. In this case, according to (3), we get system of two equations, which are linear ones with respect to N0 (N­12) andN0 (B-12): • n{ti}(N-12)N0 (N-12)exp((N-12)ti)t  (B-12)N0 (B-12)exp((B-12)ti)t. n{tj}(N-12)N0 (N-12)exp((N-12)tj)t  (B-12)N0 (B-12)exp((B-12)tj)t, (4)

13. From (4), we get N0 (N-12), N0 (B-12), and k={N0 (N-12)/N0 }. Repeating this procedure for all values of i and j = (i+1), we get a sequence of values of ki, corresponding to every interval of time-distribution. • To decrease statistical uncertainties at determination of k for express-analysis it may be useful to turn from the differential form of describing of a decay process to the integral form. From (1) in this case a difference between quantities of un-decayed nuclei to moments of time t1 and t2 (t2>t1): • N{t1}N{t2}N0 (N-12)[exp((N-12)t1)exp((N-12)t2)]N0 (B-12)[exp((B-12)t1)exp((B-12)t2)]∑ni, (5) • where ∑ni – a sum of events, registered in channels, corresponding to an interval (t2-t1). If to take two time intervals, then from equations (5) we get a system, analogous to (4), from solving of which we get k. • The suggested method may be practically useful, if only k-“portraits” of explosives and “false- substances” differ essentially. To answer to this question we made computer simulation of 12N- and 12B- production in objects of interest under bombardment them by bremsstrahlung, generated by incident on radiator electrons with Ee55 MeV. In spite of all criticism with respect to data on  of the reactions (I) and (II), as the first step, we decided to use data from Fig. 3. At small statistical uncertainties, there were found values of N0 (N-12), N0 (B-12), and N0  and calculated values, called here “true” ones – ktrue.

14. Some of found in such a way values of ktrue are presented in Table 1. We may see, that values of ktrue for explosives differ essentially from those for chemical compounds, which may be in baggage of air-passengers. Values of ktrue for explosives differ also essentially from those for objects, which have vegetable origin, what is very important at operations of humanitarian demining. Presented in Table 1 values of ktrue were calculated in approximation with doubtful data on some  and without taking into account some process’s details (in particular, dependence of physical efficiency on E and influence of absorption of '-quanta in cover-substances on E-spectra). At practice there is possibility for considerable improving of the situation by means of experimental calibrations on a real installation, at which there may be obtained and written into a base of data values of k for substances of interest. Table 1. 

15. 6. Work simulation for the third variant of connected with 12N detectors of hidden explosives • Differences of values of ktrue do not mean that at work of real DHE-3 accumulated counts in channels will be sufficient for calculating k with accuracy, ensured reliable identification of substances. That is why there was carried out computer simulations of DHE-3 in conditions, imitating baggage inspection in airports at the following main parameters of the installation:

16. Thefollowingscenarioofworkwasused: • 1. The γ'-detector is switched on with some delay after a beam pulse end, and ‘- pulses are accumulated in the histogrammic memory of the time-digital converter. The obtained histogram is written in the computer. • 2. Obtained according to “1.” data are normalized and summed up across all channels with receiving a number N* . If N*is less than threshold N* thresh, it is supposed, that in an irradiated region there is no explosives. • 3. The removed by the scanning device beam from the next accelerator pulse irradiates the next region of the tested object and initiate new accumulation of data and calculation of the new number N*. If in this case N*N* thresh, then the described above procedure of data processing for time-distributions is switched on, and the previous irradiation signals are used as background ones. • The presented below results of work simulation for DHE-3 were obtained at the following conditions: • ■ The connected with produced photoneutrons background signals have time-distribution, described by sum of two exponents. For the first exponent (T1/2)11.5 ms and sum of obtained signals (N0)1700; for the second exponent (T1/2)2  5 ms and (N0)2 800. • ■The useful signals appear from 50 g of trotyl (TNT), covered with the water layer with the thickness – 10 cm. For this registered signals N0 2100. • ■ Accumulation of '-detector data beguines at 4 ms after a beam pulse and continues till 19 ms. The channel width for the time-digital converter – 1 ms.

17. Fig. 4. Obtained by simulation distributions n{t}: 1–background (without explosives); 2–background and effect (i.e. with an explosive); 3–only effect.

18. Fig. 5. Obtained from results of simulation ksim{t}: 1– for 50 g of TNT; 2– for 50 g of C; 3–for background. It is seen, that while for N- and C-containing substances groups of points may be in good approximation presented by straight lines, parallel to t-axis, for background tangents to the approximating curve for the corresponding group of points form big angles  with t-axis

19. Table 2. Average values of ksim in comparison with the obtained before values of ktrue demonstrate with convincing effectiveness of the suggested method of substance identification in the conditions, close to work at the restricted quantityof usefui signals and the relatively high levels of background.Big difference of average values of tg() for considered cases is seen too. Therefore, if even because of some reasons at an object test a sum of registered signals receives a sharp increase, but values of k are strongly dependent on t, then it means that in this part of an object 12N and 12B were not produced and it’s not explosive.

20. This method permits to construct new devices with parameters much better than those for existing devices and open perspectives to make detection of explosives totally automatic with separation them from false objects. • At the same conditions it was also shown by computer simulation, that for DHE-3, installed on the moving platform, it is possible to detect explosives with weight 40 g in ground at depth 20 cm and during 8 hours to test 0.01 km2 of territory, what is approximately in 500 times more than what the commonly used at humanitarian demining manual method gives. • Use of stationary installations of a type in the airports permits to get a high efficient and fast acting detector of hidden explosives. For example, the inspection of one baggage unit in airport takes less than 2 s and probability of explosive detecting in a case is much higher than that for the used nowadays methods (different intrascopes, gas analyzers, trained animals and so on).

21. Conclusions • The algorithm for substance identification, using measured time-distributions of 12N- and 12B- activities, produced in photonuclear reactions and connected with relative concentrations of nitrogen and carbon in irradiated substances, was considered. It was shown by computer simulation that installations for disclosure and identification of hidden explosives based on photonuclear method can surpass significantly used now for these purposes devices on combined criterion sensitivity-ability for fast acting-trustworthiness. The method is under successful development in LPI RAS.