Physics of Atomic Nuclei, Vol. 66, No. 7, 2003, pp. 1211–1218. Translatedfrom Yadernaya Fizika, Vol. 66, No. 7, 2003, pp. 1251–1258. Original Russian Text Copyright c 2003 by Gangrsky, Zhemenik, Maslova, Mishinsky, Penionzhkevich, Sz¨oll¨os. NUCLEI Experiment
Independent Yields of Kr and Xe Isotopes in the Photofission of Heavy Nuclei
Yu.P.Gangrsky*, V.I.Zhemenik,N.Yu.Maslova, G. V. Mishinsky, Yu. E. Penionzhkevich, and O. Szoll¨ os¨ Joint Institute for Nuclear Research, Dubna, Moscow oblast, 141980 Russia Received July 15, 2002
Abstract—The yields of Kr (A =87–93)andXe(A = 138–143) primary fission fragments produced in 232Th, 238U,and244Pu photofission upon the scission of a target nucleus and neutron emission were measured in an experiment with bremsstrahlung from electrons accelerated to 25 MeV by a microtron, and the results of these measurements are presented. The experimental procedure used involved the transportation of fragments that escaped from the target by a gas flow through a capillary and the condensation of Kr and Xe inert gases in a cryostat at liquid-nitrogen temperature. The fragments of all other elements were retained with a filter at the capillary inlet. The isotopes of Kr and Xe were identified by the γ spectra of their daughter products. The mass-number distributions of the independent yields of Kr and Xe isotopes are obtained and compared with similar data on fission induced by thermal and fast neutrons; the shifts of the fragment charges with respect to the undistorted charge distribution are determined. Prospects for using photofission fragments in studying the structure of highly neutron-rich nuclei are discussed. c 2003 MAIK “Nauka/Interperiodica”.
INTRODUCTION At the same time, a wider range of experimental data and new theoretical approaches to describing The distributions of product fragments with re- them are required for obtaining deeper insight into spect to their mass numbers A and charge numbers Z the dynamics of the fission process. Measurement of are among the main properties of the nuclear-fission the isotopic and isobaric distributions of fragments process. The shape of these distributions is controlled originating from reactions induced by γ rays is one by the dynamics of the fission process from the saddle of the promising lines of investigation in this region to the scission point. This intricate process depends of nuclear fission since such reactions possess some on a number of factors, including the relief of the special features: energy surface, the configurations of nuclear shapes γ at the instant of scission, and the nuclear viscosity (1) The interaction of radiation with nuclei is of collective motion. Therefore, measurements of the purely electromagnetic, and its properties are well isotopic and isobaric distributions of fragments (that known; therefore, an adequate calculation of this in- is, those in A at given Z and those in Z at given A, teraction can be performed without resort to addi- respectively) provide important data on the dynamics tional model concepts. of the fission process. The distributions of fragments (2) In the energy range 10–16 MeV, the interac- produced after fissile-nucleus scission and the emis- tion of photons with nuclei is of resonance origin (this sion of neutrons (primary fragments) are the most is the region of a giant dipole resonance); the energy of informative of them. this resonance corresponds to the frequency of proton oscillations with respect to neutrons in a nucleus. However, measurement of such distributions in- volves some difficulties. Each of the techniques used (3) The absence of the binding energy and of the has its limitations and can be applied efficiently only Coulomb barrier allows one to obtain fissile nuclei of to a certain range of fission fragments. Therefore, iso- any (even extremely low) excitation energy immedi- topic and isobaric distributions of primary fragments ately after photon absorption. have received adequate study only for the fission of the (4) Over a wide range of γ-ray energies, the Th, U, and Pu isotopes that is induced by low-energy angular-momentum transfer to the irradiated nucleus neutrons and for the spontaneous fission of 252Cf. undergoes virtually no change—it is as low as 1 in Data on these distributions are compiled in [1, 2]. the dipole absorption of photons. These features peculiar to reactions induced by γ *e-mail: [email protected] rays allow one to obtain new information about the
1063-7788/03/6607-1211$24.00 c 2003 MAIK “Nauka/Interperiodica” 1212 GANGRSKY et al.
Ar 238U,and244Pu fission induced by γ rays of energy corresponding to a giant dipole resonance and to compare these yields with similar data obtained in neutron-induced fission. The measurements of the Chamber 238 – cumulative yields of these isotopes in Uphoto fis- e γ sion are reported in [9]; preliminary results on the independent yields of Xe fragments in 232Th and 238U fission are given in [10]. Capillary Pump EXPERIMENTAL PROCEDURE The extraction of Kr and Xe isotopes from the bulk of photofission fragments is based on the use of their special chemical properties, which differ significantly Cryostat from the properties of other elements and which re- Fig. 1. Layout of the experimental setup. veal themselves in a number of phenomena, includ- ing the adsorption of their atoms on a filter and the walls of the capillary through which the fragments are nuclear-fission process. For example, experiments transported from the irradiated target to radioactive- that studied the photofission of nuclei furnished a radiation detectors. The point is that Kr and Xe inert gases are efficiently adsorbed only at liquid-nitrogen wide set of data on the structure of the fission barrier ◦ and the shape of the potential surface [3]. The above temperature (−195.8 C), while all other fragments features of γ-ray interaction with nuclei are expected are adsorbed even at room temperature. It is precisely to manifest themselves in the process of fission- this property that is employed in our experimental fragment formation as well. This can be illustrated setup, which comprises a reaction chamber, a cryo- by the results presented in [4], according to which the stat, and Teflon capillaries for a gas flow. The layout γ-ray-energy dependence of the yield of fragments of the setup is displayed in Fig. 1, and a more detailed (115Cd, 117Cd) from symmetric fission has a rather description of it is given elsewhere [11]. unusual character, with a maximum at 6 MeV and a The irradiated targets were placed in the reaction kink at 5.3 MeV. chamber, which had the shape of a cylinder 30 mm in However, data on the isotopic and especially height, its inner diameter being 40 mm. The cylinder isobaric distributions of photofission fragments (first walls had a thickness of 1.5 mm. The inlet and out- of all, on their independent yields) are considerably let holes for a buffer gas were located symmetrically scarcer than data on reactions induced by neutrons opposite each other along a diameter of the reaction and charged particles. We can mention only the chamber. One or two targets placed at the end faces of results of the experiments performed in Ghent (Bel- the cylinder were used in the experiments. The prop- gium) for a limited set of nuclei [5–8]. erties of the targets are presented in Table 1 (these are The objective of this study is to measure the in- the chemical composition, admixtures, dimensions, dependent yields of fragments (these are the isotopes layer thickness, and the substrates). of Kr and Xe inert gases) produced directly in 232Th, An inert gas was supplied from a gas-container to the reaction chamber through a polyethylene capillary 4 mm in diameter. The gas pressure in the chamber Table 1. Properties of the targets used was adjusted by a container valve and measured with a manometer at the chamber inlet. Chemically pure Target 232Th 238U 244Pu He, Ar, and N2 at a pressure of 1 to 2 atm were used as buffer gases. The choice of inert gas and pressure Material Metal U3O8 PuO2 in the chamber was determined by the conditions of Content 100 99.85 96 the experiment (the efficiency of fragment collection, (%) the rate of their transportation, the levels of the γ- radiation background caused by the activation of the Admixtures (%) – 0.15 (235U) 4(240Pu + 242Pu) inert gas owing to microtron bremsstrahlung). Thickness 25 3 0.3 2 The fission fragments were transported, together (mg/cm ) with the buffer gas, from the reaction chamber to the Dimensions (mm) ∅15 ∅20 and 30 10 × 10 cryostat through a Teflon capillary of length 10 m and inner diameter 2 mm by means of evacuation Substrate – Al Ti with a pump. A fibrous filter placed at the capillary
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 INDEPENDENT YIELDS OF Kr AND Xe 1213
A = 91 93 140 142 Se 0.37 s ↓ Br 0.64 s Br 0.1 s I 0.86 s I 0.2 s ↓ ↓ ↓ ↓ Kr 8.6 s Kr 1.3 s Xe 14 s Xe 1.2 s ↓ ↓ ↓ ↓ Rb 58 s Rb 5.8 s Cs 64 s Cs 1.8 s ↓ ↓ ↓ ↓ Sr 9.5 h Sr 7.4 min Ba 13 d Ba 11 min ↓ ↓ ↓ ↓ Y 5.8 d Y 10.5 h La 40 h La 92 min ↓ ↓ ↓ ↓ Zr stable Zr 1.5 × 106 y Ce stable Ce stable
Fig. 2. Examples of the chains of β decays of the fission fragments that include Kr and Xe isotopes. The isotopes for which we measured γ spectraaregiveninboldfacetype. inlet efficiently adsorbed all fission fragments, with the cryostat. Therefore, only Kr and Xe fragments that exception of Kr and Xe. were produced directly at the instant of the scission A spirally twisted copper tube placed in a Dewar of a fissile nucleus reached the cryostat, and their vessel containing liquid nitrogen was used as a cryo- yields can be considered to be independent. Frag- stat. The total length of the tube was 3 m, its inner ments of other elements could appear in the cryostat diameter was 2 mm, and its walls had a thickness only after the β decay of Kr and Xe isotopes, and of 0.5 mm. This design of the cryostat enabled us to their yields must also correspond to the independent maintain the temperature at a level close to −200◦C yields of the inert gases. It was the yields of precisely and, hence, to ensure the condensation of Kr and Xe these nuclides that were measured in the experiment, (their temperatures of condensation are −157◦Cand because they had longer half-lives and more easily −112◦C, respectively). In this way, it was possible to observed γ lines. Short-lived fragments (forerunners separate Kr and Xe isotopes efficiently and quickly of Kr and Xe) that decayed in the reaction chamber from the bulk of fission fragments, which remained on had low independent yields and did not distort the the filter and the capillary walls. A comparison of the Kr and Xe yields. Figure 2 shows examples of β- spectra of γ radiation from fragments in the target, at decay chains involving the Kr and Xe isotopes; the the filter, and in the cryostat revealed that more than nuclides that were used to determine the yields in 90% of fragments not associated with the β decay question are given in boldface type. The properties chains involving Kr and Xe were retained by the filter of the radioactive decay of these nuclides (half-lives and that more than 50% of Kr or Xe fragments were T1/2,energiesEγ , intensities Iγ of measured γ lines) stopped in the cryostat. are listed in Table 2 [12]. The present approach to measuring independent yields enabled us to improve The time it takes to transport Kr and Xe fragments the accuracy of measurements and to observe frag- from the reaction chamber to the cryostat was de- −3 termined in dedicated tests that involved measuring ments whose yields are as low as 10 of the number the pressure drop in the chamber for the case where of fission events. the valve was closed while the pump was operating. This time, which comprised the time of fragment dif- EXPERIMENTAL RESULTS fusion to the outlet hole and the time of transportation through the capillary, ranged between 0.5 and 1.0 s, The experiments measuring the independent its specific value being dependent on the type of gas yields of Kr and Xe fragments were performed in and its pressure [11]. It was the time period that a bremsstrahlung-photon beam from the MT-25 determined the lower limit on the half-lives of the microtron of the Flerov Laboratory for Nuclear Re- fragments under study (about 0.3 s). If the half-lives actions at the Joint Institute for Nuclear Research of the fragments that were the forerunners of Kr and (JINR, Dubna). The description of this microtron and Xe were greater than the time of their diffusion to its basic properties were presented in [13]. The energy the outlet chamber hole (0.1–0.2 s), they were also of accelerated electrons was 25 MeV (which deter- adsorbed on the filter, making no contribution to the mined the endpoint energy of the bremsstrahlung- yield of the Kr and Xe fragments transported to the photon spectrum); the beam current was 15 µA. A
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1214 GANGRSKY et al.
Table 2. Properties of the radioactive decay of the nuclides explored in our experiments
Primary Daughter T T E ,keV I ,% fragment 1/2 isotope 1/2 γ γ 89Kr 3.2 min 89Rb 15 min 1032 58 1248 43 91Kr 9s 91Sr 9.5 h 750 23.6 1024 33.4 92Kr 2s 92Sr 2.7 h 1384 90.1 92Y 934 13.9 93Kr 1s 93Y 10.5 h 267 7.32 138Xe 14 min 138Cs 32 min 1436 76.3 2218 15.2 139Xe 40 s 139Ba 83 min 166 23.6 140Xe 14 s 140Ba 12.8 d 537 24.4 140La 40 h 1596 95.4 141Xe 1.8 s 141Ba 18 min 190 46 141La 4h 1355 1.64 142Xe 1.2 s 142La 93 min 641 47.4 2398 13.3 143Xe 0.8 s 143Ce 33 h 293 42.8
water-cooled tungsten disk 4 mm thick was used as radiation; Iγ is the relative intensity of a γ line in the a converter. Behind the converter, there was a 20- decay of the measured nuclide; α is the coefficient of mm-thick aluminum shield preventing electrons from the internal conversion of γ radiation; f(t) is the time entering the reaction chamber. The electron beam had factor that takes into account the decay of Xe and Kr the shape of an ellipse whose horizontal and vertical nuclei in the course of their transportation to the cryo- axes were 7 and 6 mm, respectively. stat and the accumulation and decay of the measured Typical periods of target irradiation were 30 min. nuclides; N is the number of decaying nuclei in the After a time interval of 5 min, the copper spiral was layer from which fission fragments escape; and η is removed from the Dewar vessel and brought to the the intensity of γ rays in the bremsstrahlung spec- γ -radiation spectrometer located in a room that was trum that induce the nuclear-fission process being protected from microtron radiation; the β decay of Kr considered. This method was used to measure, for the and Xe isotopes occurred within this period. first time, the independent yields of four Kr isotopes The areas of the γ lines associated with the nu- (A =89–93) and six Xe isotopes (A = 138–143)in clides of the daughter products of the β decay of the Kr 232 238 244 and Xe isotopes were obtained from the γ-radiation Th, U,and Pu photofission. spectra measured with this detector. These γ-line Since the Kr and Xe isotopes being studied were areas and the yields of the corresponding isotopes are obtained within a single exposure and since the pro- related by the equation cedure of subsequent measurements was identical for S (1 + α) Nf (t) Y (A)= , (1) all of them, the relative yields of these isotopes can tIγε1ε2ε3η be obtained from the measured γ-line areas by using where S is the γ-line area after background subtrac- the aforementioned parameters Iγ , α, ε3,andf(t) without recourse to the parameters ε and ε , which tion; t is the time of measurements; ε1, ε2,andε3 1 2 are the efficiencies of, respectively, the transportation are not known to a sufficient precision. The yields of Kr and Xe fragments along the capillary, their ad- obtained by this method and normalized with respect sorption in the cryostat, and the detection of their γ to the 91Kr and 139Xe isotopes are given in Table 3.
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 INDEPENDENT YIELDS OF Kr AND Xe 1215
Table 3. Independent yields of Kr and Xe fragments
232Th(γ,f) 238U(γ,f) 244Pu(γ,f) Fragment Yrel Yfr Yrel Yfr Yrel Yfr 89Kr 0.15(2) 0.10(2) 0.29(1) 0.18(2) 91Kr 1.00 0.62(5) 1.00 0.60(5) 92Kr 0.95(1) 0.59(5) 0.80(2) 0.46(4) 93Kr 0.25(1) 0.15(2) 0.25(2) 0.15(2)
137Xe 0.27(3) 0.16(2)∗ 138Xe 0.85(1) 0.53(4) 0.65(3) 0.38(4) 0.84(4) 0.47(4) 139Xe 1.00 0.63(5) 1.00 0.59(5) 1.00 0.56(5) 140Xe 0.85(7) 0.53(5) 0.94(7) 0.56(5) 1.03(8) 0.57(5) 141Xe 0.19(1) 0.12(1) 0.53(3) 0.31(3) 0.73(4) 0.41(4) 142Xe 0.09(1) 0.06(1) 0.26(2) 0.16(2) 0.46(4) 0.26(3) 143Xe 0.08(2) 0.05(1) 0.24(4) 0.14(2) Note: The value labeled with an asterisk was borrowed from [6].
The absolute values of the independent yields of DISCUSSION OF THE RESULTS 91Kr and 139Xe isotopes can be deduced by employ- ing the known data on their yields in the reactions The curves in Fig. 3 have a typical shape charac- induced by γ rays [6–8, 14] and by fission-spectrum terizedbyamaximumataspecificvalueofA and a and 14.7-MeV neutrons [2] (the isotopic distributions smooth decrease to the right and to the left of it. Such are similar for these reactions, and the yields are close inmagnitude[15]).Onthebasisofananalysisof these data, the independent 91Kr and 139Xe yields Y, (fission events)–1 divided by the sums of the yields of all A =91and 139 –1 10 Kr Xe isobars (fractional yields), respectively, were assigned values that are given in Table 3, and these values were used to determine the fractional independent yields for the rest of the measured Kr and Xe isotopes with 232Th 232Th allowance for the variation in the total yields of frag- 10–2 ments with a specific mass number and the emission of delayed neutrons from the fragments (these results are also presented in Table 3). The above yields di- vided by the total number of fission events are shown in Fig. 3 versus the mass number. 10–3 238U 238U ×10 × Owing to the fact that the bremsstrahlung spec- 10 trum is continuous, these yields are associated with a particular range of excitation energies. This range can 244Pu × be obtained from the experimentally measured exci- 10–4 100 tation function for the reaction 238U(γ,f) [16] (the 90 92 94 138 140 142 144 excitation functions for the remaining fissile nuclei A were assumed to be similar to that) and from the bremsstrahlung spectrum calculated for the specific Fig. 3. Mass-number distributions of the independent yields of Kr and Xe isotopes in the photofission of 232Th, conditions of the present experiment [17]. The mean 238U,and244Pu nuclei. The circles represent experi- excitation energy appears to be 12.5 MeV, and the mental data, while the dashed curves correspond to their half-width of the distribution is about 5 MeV. approximation by Gaussian curves.
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1216 GANGRSKY et al.
Table 4. Parameters of the isotopic distribution of Kr and Xe fragments
N − Z Kr Xe References Reaction N A¯ σ A¯ σ 232Th(γ,f) 0.366 91.3(2) 1.1(1) 138.9(2) 1.2(1) This study 238U(γ,f) 0.370 91.1(2) 1.3(1) 139.4(2) 1.5(1) This study 138.9(3) [5] 244Pu(γ,f) 0.373 139.7(2) 1.8(2) This study 235U(γ,f) 0.357 89.4(3) 1.3(1) 137.4(4) 1.4(1) [5] 233U(n, f) 0.352 89.3(1) 1.5(1) 137.8(1) 1.5(1) [1] 235U(n, f) 0.364 90.1(1) 1.5(1) 138.4(1) 1.6(1) [1] 238U(n, f) 0.379 91.5(1) 1.6(1) 139.5(1) 1.8(2) [1]
Table 5. Shifts of charges of Kr and Xe fragments with respect to the undistorted charge distribution
Z A¯ ν ∆Z Reaction Z0/A0 Kr Xe Kr Xe Kr Xe Kr Xe 232Th(γ,f) 0.388 36 54 91.3(2) 138.9(2) 1.0 0.8 −0.19(8) +0.20(8) 238U(γ,f) 0.387 36 54 91.1(2) 139.4(2) 1.0 0.8 −0.36(9) +0.25(8) 244Pu(γ,f) 0.385 54 139.7(2) 0.8 +0.13(6) curves can be described by a Gaussian distribution, (ii) The standard deviation of the distribution shows a sizable growth with increasing charge num- A − A¯ 2 Y (A)=K exp − , (2) ber of the fissile nucleus (by way of example, we 2σ2 indicate that, in going over from 232Th to 244Pu, it in- creases by a factor of 1.5). This implies a considerable where K is the normalization factor, A¯ is the mean distinction between the yields of the most neutron- mass number, and σ is the standard deviation. rich fragments. For example, the yield of 143Xe in The experimental values of the yields were ap- 244Pu photofission is an order of magnitude higher proximated by such distributions, which are shown in than that for 232Th, and this difference grows fast with Fig. 3, the fitted values of their parameters being given increasing number of neutrons in the fragment. ¯ 238 in Table 4 (the A value for Uis in agreement with (iii) Deviation of the experimental values of the that obtained in [7]). The corresponding parameters yields from the Gaussian distribution, as well as their 235 233 for Uand U fission induced by thermal neutrons odd–even differences, is small and does not go beyond and for 238U fission induced by 14.7-MeV neutrons the uncertainties of the measurement. [2] are also presented for the sake of comparison. A In 232Th photofission, the Kr and Xe fragments comparison of these parameters leads to the following are conjugate—that is, they are produced in a single conclusions: fission event. In comparing the sum of their mean (i) The mean mass number for Kr and Xe frag- mass numbers [A¯(Xe) = 138.9, A¯(Kr)=91.3]with ments exhibits a slow growth with increasing Z and the mass number of the fissile nucleus (A¯ = 232), A of the fissile nucleus and is close to the A¯ value one can assess the number of neutrons emitted from in 238U fission induced by 14.7-MeV neutrons, but these fragments. It appears to be ν =1.8(2). If it is it is noticeably greater than that in 235Uand 233U considered that the fractional yields of 91Kr and 139Xe fission induced by thermal neutrons. This suggests are, respectively, 0.62(5) and 0.63(5) (see Table 3), the growth of A¯ with increasing neutron excess in the this value is in satisfactory agreement with the known fissile nucleus [this excess can be characterized by the values of the number of fission neutrons from the frag- ratio (N − Z)/N , whose values are quoted in Table 4]. ments with a specific mass number and their depen-
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 INDEPENDENT YIELDS OF Kr AND Xe 1217 dence on the excitation energy [15, 18]. It follows from fission. It is obvious that photon-absorption-induced these data that the ratio of the numbers of neutrons dipole vibrations of the electric charge of nuclei are from a light and the complementary heavy fragment is damped completely by the instant of fragment for- 1.3 in the fragment-mass-number range under study mation. Probably, some other of the aforementioned and that it varies only slightly in the excitation-energy features of γ-radiation interaction with nuclei can range being considered. This corresponds to the value affect fragment formation at lower excitation energies of ν =1.0 for Kr isotopes and to the value of ν =0.8 (in the vicinity of the fission barrier or below it). for Xe isotopes. The results of this study give sufficient grounds By using the above values of ν, we can assess to conclude that photofission of nuclei is an effi- the shift of the fragment charge with respect to the cient means for obtaining the most neutron-rich nu- undistorted charge distribution, which corresponds to clides. A high yield of such nuclides is caused by the ratio Z0/A0 of the charge number of the fissile a low excitation energy of nascent fission fragments nucleus to its mass number. The expression for the (and, consequently, by a small number of neutrons charge shift in the fragment of charge number Z has emitted from them), by a high penetration capacity the form of bremsstrahlung radiation (this makes it possible Z to employ large amounts of fissile substance), and ∆Z = 0 A¯ + ν − Z. (3) A by a rather large standard deviation of the isotopic 0 distribution of fission fragments. Therefore, photofis- The ∆Z values obtained by this method for the mean sion provides a promising method both for explor- mass numbers of Kr and Xe fragments in 232Th, 238U, ing the nuclear structure of fission fragments and and 244Pu photofission are presented in Table 5. They for obtaining intense beams of accelerated neutron- appear to be close to the corresponding ∆Z values in rich radioactive nuclei by using these fragments. In- the neutron-induced fission of heavy nuclei [1, 19]. vestigation of reactions induced by such nuclei is of A wide variety of models have been used to de- great interest since they permit obtaining radically scribe the isotopic and isobaric distributions of fis- new information about the properties of nuclei, which sion fragments. In recent years, the diffusion model is inaccessible with other methods. A large num- [20, 21] has become widely accepted. Within this ber of projects devoted to the acceleration of fission model, the breakup of a fissile nucleus is a quasiequi- fragments have been developed in recent years or librium process with respect to the charge mode, so are being executed at present. These include DRIBs that the statistical approach can be used to describe project, which is being carried out at the Laboratory the charge distribution. In this case, the collective for Nuclear Reactions at JINR. In DRIBs, fission motion of nucleons, which manifests itself as dipole fragments are obtained by irradiating a thick uranium isovector vibrations, will be the main mechanism of target with bremsstrahlung from the microtron and formation of the isobaric charge distribution of frag- are accelerated at the 4-m isochronous cyclotron [23]. ments. However, the single-particle motion of nu- cleons can also make a nonvanishing contribution. ACKNOWLEDGMENTS The characteristic time of both kinds of motion (about 10−22 s) is much shorter than the time it takes for We are grateful to Yu.Ts. Oganessian, M.G. Itkis, the nucleus to descend from the saddle to the scission and S.N. Dmitriev for support of our studies; to point (about 3 × 10−20 s). The criterion of potential- Ya. Kliman and V.V. Pashkevich for stimulating energy minimum is used to determine the fragment discussions; and to A.G. Belov and G.V. Buklanov charge. The deformation dependence of the potential for assistance in performing the experiments. energy was calculated on the basis of the liquid-drop This work was supported by the Russian Founda- model with allowance for shell corrections. For the tion for Basic Research (project nos. 01-02-97038, fission of nuclei in the Th–Pu region, the calculations 01-02-16455) and INTAS (grant no. 2000-00463). of the isobaric charge distributions within this model showed that the shift of the charge with respect to the undistorted charge distribution is 0.2–0.4 [22]. These REFERENCES values of ∆Z are consistent with the experimental 1. A. C. Wahl, At. Data Nucl. Data Tables 39, 1 (1988). data obtained in our study of photofission, as well as 2. T. R. England and B. E. Rides, Preprint La-Ur-94- with those from other studies of the fission process 3106, LANL (Los Alamos, 1994). induced by various particles. 3. Yu. M. Tsipenyuk, Yu. B. Ostapenko, G.N.Smirenkin,andA.S.Soldatov,Usp.Fiz. The results of our experiments reveal that the iso- Nauk 144, 3 (1984) [Sov. Phys. Usp. 27, 649 (1984)]. topic distributions of fragments originating from the 4. B. V. Kurchatov, V. I. Novgorodotseva, V. A. Pchelin, photofission of heavy nuclei are similar to the corre- et al.,Yad.Fiz.7, 521 (1968) [Sov. J. Nucl. Phys. 7, sponding distributions obtained in neutron-induced 326 (1968)].
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1218 GANGRSKY et al.
5. E. Jacobs, H. Thierens, D. De Frenne, et al.,Phys. 16. J. T. Cadwell, E. J. Dowly, B. L. Berman, et al.,Phys. Rev. C 19, 422 (1979). Rev. C 21, 1215 (1980). 6. D. De Frenne, H. Thierens, B. Proot, et al.,Phys. 17. Ph.Kondev,A.Tonchev,Kh.Khristov,andV.E.Zhu- Rev. C 26, 1356 (1982). chko, Nucl. Instrum. Methods Phys. Res. B 71, 126 7. D. De Frenne, H. Thierens, B. Proot, et al.,Phys. (1992). Rev. C 29, 1908 (1984). 8. H. Thierens, A. De Clercq, E. Jacobs, et al.,Phys. 18. J. Terrell, Phys. Rev. 127, 880 (1962). Rev. C 29, 498 (1984). 19. A. C. Wahl, R. L. Fergusson, D. R. Nethaway, et al., 9. F. Ibrahim, J. Obert, O. Bajeat, et al.,Eur.Phys.J.A Phys. Rev. 126, 1112 (1962). 15, 357 (2002). 10.Yu.P.Gangrskiı˘ , S. N. Dmitriev, V. I. Zhemenik, 20.G.D.Adeev,I.I.Gonchar,V.V.Pashkevich,et al., ´ et al.,Fiz.Elem.´ Chastits At. Yadra 6, 5 (2000). Fiz. Elem. Chastits At. Yadra 19, 1229 (1988) [Sov. J. 11.Yu.P.Gangrskiı˘ ,V.P.Domanov,V.I.Zhemenik, Part. Nucl. 19, 529 (1988)]. et al.,Prib.Tekh.Eksp.´ 3, 67 (2002). 21. G. D. Adeev, Fiz. Elem.´ Chastits At. Yadra 23, 1572 12. E.BrouneandR.B.Firestone,Table of Radioactive (1992) [Sov. J. Part. Nucl. 23, 684 (1992)]. Isotopes (Wiley, New York, 1986). 22. G. D. Adeev and T. Dossing, Phys. Lett. B 66B,11 13. A. G. Belov, Preprint No. D-15-93-80 (Joint Inst. for Nucl. Res., Dubna, 1993). (1977). 14. J. C. Hogan and A. E. Richardson, Phys. Rev. C 16, 23. Yu. Oganessian, Dubna Radioactive Ion Beams, 2296 (1977). Project, Int. Report FLNR, Joint Inst. for Nucl. Res. 15.Yu.P.Gangrskiı˘ , B. Dalkhsuren, and B. N. Markov, (Dubna, 1998). Nuclear-Fission Fragments (Energoatomizdat,´ Moscow, 1986). Translated by E. Kozlovskii
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 Physics of Atomic Nuclei, Vol. 66, No. 7, 2003, pp. 1219–1228. Translated from Yadernaya Fizika, Vol. 66, No. 7, 2003, pp. 1259–1268. Original Russian Text Copyright c 2003 by Kadmensky, Rodionova. NUCLEI Theory
Angular Distributions, Relative Orbital Angular Momenta, and Spins of Fragments Originating from the Binary Fission of Polarized Nuclei S.G.KadmenskyandL.V.Rodionova Voronezh State University, Universitetskaya pl. 1, Voronezh, 394693 Russia Received June 17, 2002; in final form, November 5, 2002
Abstract—The angular distributions of fragments originating from the binary fission of odd and odd–odd nuclei capable of undergoing spontaneous fission that are polarized by a strong magnetic field at ultralow temperatures and from the low-energy photofission of even–even nuclei that is induced by dipole and quadrupole photons are investigated. It is shown that the deviations of these angular distributions from those that are obtained on the basis of the A. Bohr formula make it possible to estimate the maximum relative orbital angular momentum of fission fragments, this estimate providing important information about the relative orientation of the fragment spins. The angular distributions of fragments originating from subthreshold fission are analyzed for the case of the 238U nucleus. A comparison of the resulting angular distributions with their experimental counterparts leads to the conclusion that the maximum relative orbital angular momentum of binary-fission fragments exceeds 20, the fragment spins having predominantly a parallel orientation. The possibility is considered for performing an experiment aimed at measuring the angular distributions of fragments of the spontaneous fission of polarized nuclei in order to determine both the spins of such nuclei and the maximum values of the relative orbital angular momenta of fission fragments. c 2003 MAIK “Nauka/Interperiodica”.
1. INTRODUCTION the fragments in question emit prompt photons. Upon photon emission, the fragment spins assume values The problem of determining the relative orbital of J that correspond to the states of final fragments. angular momenta l and spins J (i =1,2)offission if i We denote by J and J the total spins of, respec- fragments is one of the key problems in the physics in iγ tively, prompt neutrons and prompt photons emitted of binary nuclear fission. The conservation of the total from the ith fragment. From the spin-conservation angular momentum J of a fissile nucleus in the fission law, it then follows that J = J + J + J .Experi- process leads to the vector relation i in iγ if ments devoted to measuring the multipole orders and J = l + F; F = J1 + J2. (1) multiplicities of prompt photons emitted by fission fragments revealed [2] that the mean values J¯iγ of It was shown in [1] that, because of the validity of the total spins J carried away by photons can be the adiabatic approximation in the asymptotic region iγ J¯ ≈ 6 8 of a fissile system, where primary fission fragments estimated (in units) at iγ – . are already formed, their deformation parameters hav- In the case of predominantly antiparallel orien- ing nonequilibrium values in this region, rotational tation of fragment spins, which follows from the bands characterized by rather wide sets of values of mechanism proposed by Mikhailov and Quentin [3], the fragment spins J1 and J2 are excited in these who considered the orientation pumping of fragment fragments. In the course of a further evolution, pri- spins, the mean value of the total spin of primary mary fission fragments go over to excited states that fragments, F¯, is close to zero. Therefore, the mean have equilibrium values of the deformation parame- value of the relative orbital angular momentum of ters, but which retain the initial values of the spins fragments is ¯l ≈ J, so that the values of ¯l are small for J1 and J2. At the next stage, these states of the typical values of the spin J of fissile nuclei undergoing fragments undergo transitions, via the emission of, spontaneous or low-energy induced fission (these first, prompt neutrons and, then, prompt photons, to values do not exceed a few units, being much less predominantly the ground states of final fragments than the mean spins J¯1 and J¯2). that are recorded in experiments. Upon the emission This conclusion is at odds with another impor- of prompt neutrons, the spins of fission fragments tant result that was obtained, within the quantum- 0 0 change values, from Ji to Ji , Ji being the initial val- mechanical theory of binary nuclear fission [1], from ues of the fragment spins for the stage within which an analysis of the angular distributions of fragments
1063-7788/03/6607-1219$24.00 c 2003 MAIK “Nauka/Interperiodica” 1220 KADMENSKY, RODIONOVA originating from the fission of polarized nuclei. These theory. Concurrently, the concept of angular distribu- distributions depend on the solid angle Ω=(θ,ϕ) tions of fission fragments in the internal coordinate determining, in the laboratory frame, the direction of frame of a fissile nucleus was also introduced in [1]. theradiusvectorR of the relative motion of fission The objective of the present study is to analyze, − fragments, R = R1 R2,whereR1 and R2 are the within the approaches proposed in [1], the angu- coordinates of the centers of mass of, respectively, the lar distributions of fragments originating from the first and the second fragment. The A. Bohr formula low-energy photofission of even–even nuclei into two (see [4]) was successfully used in describing these fragments that is induced by dipole and quadrupole angular distributions for the binary fission of polarized photons (the angular distributions for this case have nuclei [4] and in describing the coefficient of P -even already been investigated experimentally) and the an- and P -odd asymmetry in the angular distributions of gular distributions for the as-yet-unexplored case of fission fragments [5–8]. It was shown in [1] that, for the binary fission of J>1/2 odd and odd–odd nuclei this formula to be approximately valid, it is neces- capable of undergoing spontaneous fission that are sary that fission fragments have coherently coupled polarized by a strong magnetic field at ultralow tem- relative orbital angular momenta l lying in a rather peratures, these analyses being aimed at deducing wide range 0 ≤ l ≤ lm; that is, the maximum possible information about the values of the relative orbital value lm of the relative orbital angular momentum angular momenta and spins of fission fragments. must be quite large. Therefore, the mean value ¯l is also large; if ¯l is much greater than the spin J of a fissile nucleus, formula (1) then leads to the relation 2. ANGULAR DISTRIBUTIONS ¯≈ ¯ | | ¯ ¯ OF FRAGMENTS ORIGINATING l F = J1 + J2 > J1, J2, whence it follows that the FROM THE BINARY FISSION ¯ ¯ spins J1 and J2 are predominantly parallel to each OF POLARIZED NUCLEI other, but this contradicts the results presented in [3]. Since the nuclear-fission process induced by spe- 0 The upper limit lm on lm can be deduced from cific particle species (n, α, γ, ...) proceeds through the the law of conservation of the fissile-nucleus spin formation of a compound nucleus [4] having a spin 0 [see Eq. (1)]: lm = J1m + J2m + J. In order to assess J and a spin projection M onto the z axis in the the maximum values Jim of the fragment spins, we laboratory frame (this axis is chosen to be aligned will make make use of the relation Jim =(Jin)m + with the incident-beam axis), the differential cross section for the emission of fragments from the binary (Jiγ )m +(Jif )m, which follows from the law of con- servation of the fragment spin. If we assume that the fission of nuclei can be represented in the form values of the maximum possible spin carried away by dσf (Ω) σf (θ) ≡ (2) photons [(Jiγ)m]differ by a few units from the mean dΩ value J¯ obtained in [2] for this spin, consider that J J iγ Γ (JK) the maximum value of the total spin carried away = σ (JM) f T J (θ), Γ(J) MK by neutrons [(Jin)m] is about a few units, and use J M=−J K=0 values of about a few units for the fissile-nucleus spin σ (JM) J and for the maximum spins (J ) of final fission where is the cross section for the formation if m of a compound nucleus, Γ (JK) is the partial-wave fragments, we can derive an estimate of the upper f 0 0 ≈ decay width of this nucleus with respect to the chan- limit lm on lm: lm 25. It follows that the possible nel characterized by a specific value of the projection values of the relative orbital angular momenta l of K of the spin J of the axisymmetric compound nu- fission fragments are constrained by the condition ≤ cleus onto the z axis of the internal coordinate frame, l lm < 25. and Γ(J) is the total decay width of the compound With the aid of the methods proposed in [9, 10], nucleus. a quantum-mechanical theory of the nuclear-fission In the K channel, the angular distribution process was developed in [1] by introducing the con- J TMK (θ) of fission fragments that is normalized to cepts of fission phases and the amplitudes of partial- unity is usually calculated on the basis of the A. Bohr wave fission widths and by relying on the adiabatic formula [4]; that is, approximation in describing the asymptotic region of 2J +1 a fissile system, the idea of A. Bohr [4] that transition T J (θ)= (3) MK 8π fission states formed at the saddle point of the defor- mation potential of a fissile nucleus play a crucial role × J 2 J 2 DMK (ω) + DM,−K (ω) , in the fission process being taken into account in that β=θ
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 ANGULAR DISTRIBUTIONS, RELATIVE ORBITAL ANGULAR MOMENTA 1221
J where DMK (ω) is a generalized spherical harmonic Such a situation can be simulated mathematically [1] that depends on the Euler angles α, β, γ ≡ ω char- by describing the distribution of θ by a formula of the acterizing the orientation of the axes of the internal type in (4) under the condition that the quantity lm in coordinate frame with respect to the axes of the labo- it takes finite, albeit rather large, values. In that case, ratory frame. In applying this formula, it is considered nonzero values of the angle θ lie within cones whose 2 that DJ (ω) is independent of the angles α and axes correspond to the values of θ =0and θ = π and MK ≈ γ. Formula (3) is based on the qualitative physical as- whose opening angle at a cone apex is ∆θ 1/lm. sumption [4] according to which fission fragments are The investigation performed in [1] by using the emitted only along or against the direction of the sym- adiabatic approximation in the asymptotic region of metry axis (z axis) of the fissile nucleus being con- a fissile system showed that the generalization of the sidered. This means that, in the internal coordinate A. Bohr formula (3) to this case can be represented as frame, the angular distribution of fission fragments ¯J 2J +1 ∓ TMK (θ)= (5) has a delta-function-like character of the form δ(ξ 16π2 1),whereξ =cosθ , θ being the angle between the 2 2 × dω DJ (ω) + DJ (ω) F 2 (θ ), vector R and the symmetry axis (z axis) of the fissile MK M,−K lm nucleus. Taking into account the azimuthal symmetry where the function F 2 (θ) is a smeared delta function of the angular distribution of fission fragments, going lm over from the argument θ to the argument ξ in the of the form in (4) with an amplitude; that is, spherical harmonic Yl0(θ ), redenoting the resulting Flm (θ ) (6) Y (ξ) function by the symbol l0 as before,√ and using l m the completeness of the set of functions 2πYl0(ξ ) l = b(lm) Yl0(ξ )Yl0(1) 1+ππ1π2(−1) . ξ −1 ≤ ξ ≤ 1 in the space of ( ), we can recast the l=0 aforementioned delta functions into the form In expression (6), the factor [1 + ππ π (−1)l],where lm 1 2 δ(ξ ∓ 1) = 2πY (ξ )Y (±1) (4) π, π1,andπ2 are the parities of, respectively, the l0 l0 parent nucleus, the first fission fragment, and the sec- l=0 ond fission fragment, takes into account the parity- lm conservation law, while the quantity b(l ) is deter- ± m = (2l +1)πYl0(ξ )Pl( 1), mined from the condition requiring that the angular l=0 distribution in (5) be normalized to unity. As a result, where P (ξ) is a Legendre polynomial [P (+1) = 1, we obtain l l − l l 1/2 Pl(−1) = (−1) ] and the quantity lm is considered in m 2 (2l +1) l the limit l →∞. Formula (4) reflects the quantum- b(lm)= 1+ππ1π2(−1) . m 4π mechanical uncertainty relation between the operator l=0 of the square of the orbital angular momentum of a (7) particle, l2,andthesquareoftheangleθ, which char- Formula (5) reduces to formula (3) for lm →∞. acterizes the direction of the radius vector R of this But at finite and rather small values of lm,thean- particle. From this relation, it follows that the angle gular distribution of fission fragments in (5) differs θ can be specified exactly only if the orbital angular significantly from the angular distribution of fission momentum l oftheparticleisabsolutelyuncertain fragments in (3). ≤ ≤∞ (0 l ). With an eye to applying the above formulas to Since only finite values of the spins J1 and J2 and a more general case, we will analyze the angular of the relative orbital angular momentum l of fission distributions of fission fragments in (3) and (5) at fragments may appear in the nuclear-fission process, arbitrary values of J, M,andK.Byemployingthe formula (3) is only approximate. At the same time, definition of a complex-conjugate generalized spheri- an analysis of the angular distributions of fragments cal harmonic [4], ∗ − originating from the fission of polarized nuclei [4] and DJ (ω)=(−1)M K DJ (ω) , (8) an analysis of the coefficients of P -even and P -odd MK −M,−K asymmetries in the angular distributions of fission and the multiplication theorem for D functions, fragments [5–8] on the basis of expression (3) re- DJ1 (ω) DJ2 (ω) (9) veal that this expression reflects actual properties of M1K1 M2K2 the nuclear-fission process. This means that angular J= J1+J2 JM1+M2 JK1+K2 J distributions of fission fragments do not vanish only = C C DM +M ,K +K (ω), J1J2M1M2 J1J2K1K2 1 2 1 2 at values of the angle θ that are close to 0 and π. J=|J1−J2|
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1222 KADMENSKY, RODIONOVA we can recast the A. Bohr formula (3) for the angular allowance for the orthogonality property of D func- distribution of fission fragments into the form tions [4], we obtain T J (θ)= B0 P (ξ), (10) | ˜J |2 | |2 MK JMKL L dω AM,±K (ω,Ω) = b(lm) (17) L lm lm where × jM−m j±K jM−m CJlM−mCJl±K0CJlM−m 0 L B = 1+(−1) l=0 l=0 jm JMKL (11) 2 2J +1 L0 L0 M+K j±K 8π 2l +1 2l +1 ∗ × C − C − (−1) × C ± Y (Ω) 8π JJM M JJK K Jl K0 2j +1 4π 4π l m l l and L takesonlyevenvaluesintherangefrom0 × Ylm(Ω)[1 + ππ1π2(−1) ][1 + ππ1π2(−1) ]. to 2J. We note that, by virtue of the properties of Clebsch–Gordan coefficients, the quantity B0 is By using the multiplication theorem for spherical JMKL harmonics [4], equal to 1/4π at L =0. m ∗ (−1) Let us now reduce the normalized (to unity) an- Ylm (Ω) Ylm (Ω) = gular distribution of fission fragments in (5) to the 4π × L0 L0 form of an expansion in Legendre polynomials that is (2l +1)(2l +1)Cll00Cllm−mPL (ξ), analogous to the expansion in (10). We introduce the L notation we can now recast expression (17) into the form J ≡ J AM,±K ω,Ω DM,±K (ω) Flm (θ ) (12) dω|A˜J (ω,Ω) |2 = |b(l )|2 (18) and represent the angular distribution (5) as M,±K m 2J +1 lm lm ¯J jM−m j±K jM−m j±K TMK (Ω) = (13) × C C C C 16π2 Jl M−m Jl ±K0 JlM−m Jl±K0 l=0 l=0 jmL × J 2 J 2 dω AM,K ω,Ω + AM,−K ω,Ω . L0 L0 (2l +1)(2l +1) × C C − P (ξ) ll 00 ll m m L 2(2j +1) Making use of the Wigner transformation [4] for go- l l ing over from the internal coordinate frame to the × [1 + ππ1π2(−1) ][1 + ππ1π2(−1) ]. laboratory frame, This formula can be significantly simplified by per- l∗ Ylk(Ω )= Dmk(ω)Ylm(Ω), (14) forming summation over the projection m of the rel- m ative orbital angular momentum l with the aid of the formulathatmakesitpossibletoevaluatesumsofthe we can recast the function F (θ) (6) into the form lm products of Clebsch–Gordan coefficients [11], Flm (θ )=b(lm) (15) cγ eε dδ Cabαβ Cdbδβ Cafαϕ l m αβδ × l∗ − l Dm0(ω)Ylm(Ω)Yl0(1)[1 + ππ1π2( 1) ]. =(−1)b+c+d+f (2c +1)(2d +1) l=0 m Taking into account relation (15) and the represen- abc × eε tation of a complex-conjugate D function in the Ccfγϕ , form (8), we can rewrite expression (12) as efd J ˜J AM,±K ω,Ω = AM,±K (ω,Ω) (16) abc where isaWigner6j coefficient, and the lm j= J+l jM−m j±K efd = b(lm) CJlM−mCJl±K0 l=0 m j=|J−l| property of symmetry of Clebsch–Gordan coefficients under the permutation of angular momenta, × j − m DM−m,±K(ω)( 1) Ylm(Ω)Yl0(1) JM J1−M1 2J +1 J2M2 l C =(−1) C − × [1 + ππ π (−1) ]. J1J2M1M2 J1JM1 M 1 2 2J2 +1 2J +1 Substituting expression (16) into (13) and perform- − J2+M2 J1M1 =( 1) CJ J−M M . ing integration with respect to the Euler angles with 2J1 +1 2 2
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 ANGULAR DISTRIBUTIONS, RELATIVE ORBITAL ANGULAR MOMENTA 1223
Table 1. Values of the coefficients αJK, βJK,andγJK for the A. Bohr angular distribution (10) and for the angular distribution (20) in the problem of the low-energy photofission of nuclei
lm α10 α11 α20 α21 α22 β10 β11 β20 β21 β22 γ20 γ21 γ22
∞ 0 0.75 0 1.25 0 0.75 −0.375 0 −0.625 0.625 0.9375 −0.625 0.15625
30 0.024 0.74 0.060 1.200 0.030 0.714 −0.357 −0.022 −0.580 0.592 0.854 −0.569 0.142
20 0.04 0.73 0.087 1.163 0.044 0.697 −0.349 −0.033 −0.560 0.576 0.815 −0.543 0.136
19 0.04 0.73 0.091 1.159 0.046 0.695 −0.348 −0.034 −0.557 0.574 0.809 −0.539 0.135
10 0.07 0.72 0.162 1.088 0.082 0.652 −0.326 −0.061 −0.503 0.533 0.709 −0.473 0.118
9 0.07 0.72 0.177 1.072 0.089 0.643 −0.322 −0.066 −0.492 0.525 0.688 −0.459 0.115
5 0.12 0.69 0.283 0.965 0.139 0.577 −0.289 −0.105 −0.411 0.463 0.538 −0.358 0.090
4 0.14 0.68 0.332 0.915 0.169 0.545 −0.272 −0.122 −0.372 0.434 0.468 −0.312 0.078 Expression (18) then assumes the form 2L +1 j+J × (2l +1) (−1) 2J +1 | ˜J |2 | |2 dω AM,±K (ω,Ω) = b(lm) (19) l l × [1 + ππ1π2(−1) ][1 + ππ1π2(−1) ]. lm lm At L =0, the quantity B is equal to 1/4π. × j±K j±K L0 JM JMKL CJl±K0CJl±K0Cll00CJLM0 l=0 l =0 jL 3. ANGULAR DISTRIBUTIONS ljJ × (2l +1)(2l +1) OF FRAGMENTS ORIGINATING FROM THE LOW-ENERGY PHOTOFISSION JLl 2 OF NUCLEI 2L +1 j+J+L Let us consider the case of the low-energy photo- × (−1) PL(ξ) 2J +1 fission of an even–even target nucleus in the ground l l state, its total angular momentum being I =0.Inor- × [1 + ππ1π2(−1) ][1 + ππ1π2(−1) ]. der to describe the differential photofission cross sec- Since l and l have the same parity, it follows from tion as a function of the angle θ between the directions the properties of the Clebsch–Gordan coefficient of fission-fragment emission, which are defined with L0 respect to the axis of the incident-photon beam, one C that L can assume only the even values of L = ll 00 can employ formula (2) at M = ±1, since the photon 0, 2,...,2J. Upon the substitution of (19) into (13), ±1 ¯J helicity is , and set the value of the compound- the angular distribution TMK (Ω) of fission frag- nucleus spin to J =1 in the case of dipole pho- ments, which is normalized to unity, can be reduced toabsorption and to J =2in the case of quadrupole to a form that is analogous to (10). Specifically, we photoabsorption. have Since the angular distribution of fission fragments ¯J TMK(θ)= BJMKLPL(ξ), (20) that is specified by Eqs. (3) and (10) and the angular L distribution that is specified by Eqs. (5) and (20) are invariant under the substitution of −M for M, where formula (2) can be represented in the form lm lm 2 2 2J +1 2 jK σf (θ)=a0 + b0 sin θ + c0 sin (2θ) , (22) B = |b(l )| C (21) JMKL 16π2 m Jl K0 l=0 l=0 jL where the coefficients a , b ,andc are defined as 0 0 0 a = α P (JK); b = β P (JK); jK L0 JM ljJ 0 JK 0 JK × C C C (2l +1) JlK0 ll 00 JLM0 JK JK JLl (23)
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1224 KADMENSKY, RODIONOVA
Table 2. Values of the coefficients G20, a,andb for the A. Bohr angular distribution (10) and the angular distribution (20) in the problem of the low-energy fission of nuclei at G22 =0and G22 =1
G22 =0 G22 =1 lm
a G20 a G20
∞ 0 0.73 ± 0.04 0 1.17 ± 0.07
30 0.096 ± 0.002 0.82 ± 0.04 0.098 ± 0.003 1.35 ± 0.08
20 0.15 ± 0.006 0.87 ± 0.05 0.15 ± 0.005 1.43 ± 0.08
10 0.29 ± 0.013 1.06 ± 0.06 0.29 ± 0.014 1.79 ± 0.11
Experiment [14] 0.09 ± 0.024 0.09 ± 0.024 c0 = γJKP (JK), rather small values of lm,thecoefficients αJK, βJK, JK and γJK change significantly upon going over from the A. Bohr formula (10) to formula (20). with +1 Our further analysis of low-energy photofission Γ (JK) 238 P (JK)= f σ (JM), (24) will be performed for the fission of U nuclei that is Γ(J) M=−1 induced by bremsstrahlung from 5.2-MeV electrons immediately below the fission threshold. We will in- while the coefficients αJK, βJK,andγJK are super- vestigate the quantity positions of the coefficients BJ1KL for L =0, 1, 2. σf (θ) It can be seen from Table 1 that, in the case of W (θ)= ◦ , (25) σf (90 )
W(θ) which characterizes the asymmetry in the angular distribution of fission fragments with respect to the 1.6 photon-beam axis. From Eqs. (25) and (22), it follows that the asymmetry W (θ) is given by expression (22), where the coefficients a0, b0,andc0 must be replaced a b 1.2 by the coefficients a = 0 , b = 0 ,andc = a0 + b0 a0 + b0 c 0 , which satisfy the relation a + b =1. 0.8 a0 + b0 The experimental values of the coefficients a and c are aexpt =0.04 ± 0.04 and cexpt =1.02 ± 0.07 [12], 0.4 aexpt =0.10 ± 0.035 and cexpt =0.91 ± 0.08 [13], aexpt =0.09 ± 0.024 and cexpt =0.907 ± 0.045 [14], expt +0.06 expt ± and a =0.08−0.03 and c =0.97 0.10 [15], 0306090 whence we can see that the values cexpt change θ , deg only slightly from measurements of one experimental group to measurements of another group, but that Fig. 1. Asymmetry W (θ) in the angular distribution expt 238 a of fragments originating from Uphotofission induced the values change more sharply, remaining, expt expt by bremsstrahlung photons whose endpoint energy is however, much smaller than b and c .Ascan 5.2 MeV,this asymmetry being defined with respect to the be seen from Table 1, the coefficients αJK are positive photon-beam axis. The curves depict theoretical results for all positive values of lm; therefore, the smallness of for various values of the maximum relative orbital angular the coefficient a mayonlybeduetothesmallnessof momentum of fission fragments: (solid curve) l = ∞ m P (11) P (21) (A. Bohr formula), (dashed curve) lm =20, and (solid the quantities and appearing as factors curve labeled with a diamond) lm =10.Opencircles in front of the coefficients α11 and α21, which are large represent experimental data reported in [14]. in magnitude. Retaining, in (23), only the K =0and
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 ANGULAR DISTRIBUTIONS, RELATIVE ORBITAL ANGULAR MOMENTA 1225 K =2channels, we can then find for the coefficients displayed along with the analogous asymmetry ob- a, b,andc that tained experimentally in [14], the A. Bohr formula (10) ∞ α + α G + α G corresponding to lm = is unable to reproduce a = 10 20 20 22 22 ; (26) the experimental asymmetry in the vicinity of the d angle θ =0. β10 + β20G20 + β22G22 b = ; If we compare the interval 20
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1226 KADMENSKY, RODIONOVA
– J θ TJJ( ) 0.16 J = 3/2
J = 1
0.08
0 60 120 180 θ, deg
¯J Fig. 2. Angular distributions TJJ (θ) (J = 3/2, 1) of fragments originating from the fission of polarized nuclei for various values of the maximum relative orbital angular momentum of fission fragments: (solid curve) lm = ∞ (A. Bohr formula), (dashed curve) lm =20, (solid curve labeled with a diamond) lm =10, and (solid curve labeled with a circle) lm =5.
– J θ TJJ( ) 0.3
J = 3
0.2 J = 2
0.1
0 60 120 180 θ, deg
¯J Fig. 3. Angular distributions TJJ (θ) of fragments originating from the fission of polarized nuclei for various values of the maximum relative orbital angular momentum of fission fragments: (solid curve) lm = ∞ (A. Bohr formula), (dashed curve) lm =20, (solid curve labeled with a diamond) lm =10, and (solid curve labeled with a circle) lm =5.
T J (θ) (10) are both symmetric with respect to an of undergoing protonic decay [9], the latter having JJ ◦ angle of θ =90◦ and take maximum values at θ =0◦ maxima at an angle of θ =90 . The deviations of ◦ (180 ). With increasing spin J of the parent nucleus, the distributions in (20) from the A. Bohr distribu- T J ◦ T J ◦ T¯J ◦ T¯J ◦ tions in (10) are the most pronounced at angles of the ratios JJ (0 ) JJ (90 ) and JJ (0 ) JJ (90 ) ◦ ◦ ◦ grow rather fast, which makes it possible to use the θ =0 (180 ) and 90 , growing in magnitude with angular distributions in (10) and (20) to determine increasing spin J of the parent nucleus and with the value of the spin J. We note that the angular decreasing maximum value lm of the relative orbital distributions in (20) and (10) for polarized nuclei are angular momentum of fission fragments. At rather ¯J qualitatively different from the angular distributions of high values of lm, the angular distributions TJJ (θ) protons emitted by polarized odd–odd nuclei capable of fission fragments in (20) change only slightly upon
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 ANGULAR DISTRIBUTIONS, RELATIVE ORBITAL ANGULAR MOMENTA 1227
Table 3. Deviation from the A. Bohr angular distribution in the problem of the spontaneous fission of polarized nuclei
T¯(0◦) − T (0◦) T¯(90◦) − T (90◦) T (90◦) ◦ , % ◦ , % ◦ , % lm T (0 ) T (90 ) T (0 )
J =1 J =3/2 J =2 J =3 J =1 J =3/2 J =2 J =3 J =1 J =3/2 J =2 J =3
20 −2.33 −3.49 −4.36 −5.72 2.34 6.96 14.83 48.62 0.5 0.25 0.125 0.03
19 −2.45 −3.66 −4.57 −6.00 2.45 7.31 15.56 51.00 0.5 0.25 0.125 0.03
10 −4.35 −6.52 −8.15 −10.69 4.32 13.04 27.70 90.92 0.5 0.25 0.125 0.03
9 −4.77 −7.14 −8.90 −11.70 4.78 14.27 30.36 99.70 0.5 0.25 0.125 0.03
5 −7.70 −11.54 −14.41 −18.85 7.69 23.07 49.10 161 0.5 0.25 0.125 0.03
4 −9.10 −13.63 −17.02 −22.24 9.10 27.27 58.02 191 0.5 0.25 0.125 0.03 going over from even values of l to their odd neighbors interaction via strongly nonspherical Coulomb and (that is, those that differ from their even counterparts nuclear potentials, ensures the emergence of such ◦ by unity), as is illustrated in Table 3 at angles of 0 high values of lm and, hence, of the total fission- and 90◦. Therefore, the angular distributions in (20), fragment spin F . as well as the A. Bohr angular distributions in (10), It would be of great interest to perform experi- are virtually independent of the parities of the parent ments devoted to studying angular distributions of nucleus and fission fragments. It can be seen from fission fragments originating from the fission of odd Table 3 that the relative deviations of the angular dis- nuclei capable of undergoing spontaneous fission that ◦ ◦ tributions in (20) from those in (10) at θ =0 (180 ) are polarized in a strong magnetic field at ultralow ◦ and 90 grow with decreasing lm, reaching 20% for temperatures. We note that similar experiments that ◦ ◦ ◦ θ =0 (180 ) and 191% for θ =90 at lm =4 and studied the angular distributions of alpha particles J =3. The emergence of such large deviations of emitted by nuclei undergoing alpha decay that were the angular distributions in (20) from the angular polarized in strong magnetic fields at ultralow tem- distributions in (10) at low values of lm indicates that peratures were successfully performed in [16, 17]. experiments aimed at studying the angular distribu- Because of small branching fractions of the fission tions of fragments originating from the spontaneous process with respect to alpha decay or ε capture for fission of polarized nuclei would make it possible to known odd and odd–odd spontaneously fissile nuclei, determine the lower limit on lm. it is necessary to overcome serious difficulties in order to record angular distributions for such nuclei with a high statistical accuracy. Nonetheless, such experi- 5. CONCLUSION ments are quite promising from the point of view of From the analysis of angular distributions of fis- their theoretical treatment, since, in contrast to what occurs in the case of induced fission, one transition sion fragments that was performed here for the case fission state characterized by only one value of K where parent nuclei are polarized, it was deduced that manifests itself in such experiments. the maximum relative orbital angular momentum of fission fragments produced under such conditions is It also seems important to continue investigating the angular distributions of photofission fragments rather high, lm > 20. It follows that the deviations of the angular distributions of fission fragments from for various nuclei and various endpoint energies of incident photons. that predicted by the A. Bohr formula (3) must be quite modest, so that this formula appears to be ap- In addition, it would be of interest to obtain an plicable in many cases. experimental corroboration of the conclusion drawn here that the spins of fragments originating from the At the same time, the fact that values obtained for binary fission of nuclei are predominantly parallel. the maximum relative orbital angular momentum lm of fission fragments exceed 20 indicates that it is nec- essary to consider a specific physical mechanism that, ACKNOWLEDGMENTS at the stage of the scission of a fissile nucleus into We are grateful to V.E. Bunakov, W.I. Furman, and fission fragments and the stage of their subsequent A.A. Barabanov for stimulating discussions.
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1228 KADMENSKY, RODIONOVA This work was supported by INTAS (grant no. 99- 9. S. G. Kadmensky and A. A. Sanzogni, Phys. Rev. C 00229) and by the foundation Universities of Russia 62, 054601 (2000). (project UR-01.01.011). 10. S. G. Kadmensky, in Proceedings of the IX Inter- national Seminar on Interaction of Neutrons with Nuclei, Dubna, 2001, p. 128. REFERENCES 11.D.A.Varshalovich,A.N.Moskalev,andV.K.Kher- 1. S. G. Kadmensky, Yad. Fiz. 65, 1424 (2002) [Phys. sonskiı˘ , Quantum Theory of Angular Momentum At. Nucl. 65, 1390 (2002)]. (Nauka, Leningrad, 1975). 2. Yu. N. Kopach et al., Phys. Rev. Lett. 82, 303 (1999). 12. A. S. Soldatov, G. N. Smirenkin, et al., Phys. Lett. 3. I. N. Mikhailov and P. Quentin, Phys. Lett. B 462,7 14, 217 (1965). (1999). 13. N.S.Rabotnov,G.N.Smirenkin,et al.,Yad.Fiz.11, 4. A. Bohr and B. Mottelson, Nuclear Structure (Ben- 508 (1970) [Sov. J. Nucl. Phys. 11, 285 (1970)]. jamin, New York; Amsterdam, 1969, 1975; Mir, 14. A. V. Ignatyuk, N. S. Rabotnov, et al.,Zh.Eksp.´ Moscow, 1971, 1977), Vols. 1, 2. Teor. Fiz. 61, 1284 (1971) [Sov. Phys. JETP 61, 684 ´ 5. A.K.Petukhov,G.V.Petrov,et al.,Pis’maZh.Eksp. (1971)]. 30 30 Teor. Fiz. , 324 (1979) [JETP Lett. , 439 (1979)]. 15. L. J. Lindgren, A. Alm, and A. Sandell, Nucl. Phys. A 6. O. P. Sushkov and V. V. Flambaum, Usp. Fiz. Nauk 298, 43 (1978). 136, 3 (1982) [Sov. Phys. Usp. 25, 1 (1982)]. 16. J. Wouters et al., Phys. Rev. Lett. 56, 1901 (1986). 7. V.E. Bunakov and L. B. Pikel’ner, Fiz. Elem.´ Chastits At. Yadra 29, 337 (1997). 17. P. Schuurmans et al., Phys. Rev. Lett. 77, 4420 8. A. Barabanov and W. Furman, Z. Phys. A 357, 441 (1995). (1997). Translated by A. Isaakyan
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 Physics of Atomic Nuclei, Vol. 66, No. 7, 2003, pp. 1229–1238. Translated from Yadernaya Fizika, Vol. 66, No. 7, 2003, pp. 1269–1278. Original Russian Text Copyright c 2003 by Ishkhanov, Orlin. NUCLEI Theory
Semimicroscopic Description of the Gross Structure of a Giant Dipole Resonance in Light Nonmagic Nuclei B. S. Ishkhanovand V. N. Orlin Institute of Nuclear Physics, Moscow State University, Vorob’evy gory, Moscow, 119899 Russia Received April 15, 2002; in final form, August 12, 2002
Abstract—A simple model is formulated that makes it possible to describe the configuration and defor- mation splittings of a giant dipole resonance in light nonmagic nuclei. The gross structure of the cross sections for photoabsorption on 12C, 24Mg, and 28Si nuclei is described on the basis of this model. c 2003 MAIK “Nauka/Interperiodica”.
1. INTRODUCTION was first considered in 1964 by Neudatchin and A giant dipole resonance in light nonmagic nuclei Shevchenko [8] and was then investigated in detail has a number of special features: (i) a very large by the authors of [9–11]. Nonetheless, the problem of theoretically describing the gross structure of a width (up to 15 MeV in 23Na [1]), (ii) a manifest giant resonance in light nonmagic nuclei has yet to intermediate structure that varies from one nucleus be solved conclusively. to another strongly, and (iii) a rather complicated gross structure. In order to explain these features, it In the present study, we make an attempt at de- is sufficient to recall that, in contrast to what we have scribing the gross structure of a giant dipole reso- in heavy nuclei, shell effects in light nuclei are much nance within the semimicroscopic model of vibrations stronger than collective effects; as a result, the os- that was proposed in [12]. For the sake of simplicity, cillator strengths of single-particle dipole transitions we take into account only the deformation and the are not concentrated predominantly in one or two (if configuration splitting of a giant dipole resonance, the nucleus being considered is deformed) collective which are closely interrelated in light nonmagic nuclei states. (see, for example, [13]). For N = Z nuclei, the isospin In the present study, we consider a number of splitting of a giant dipole resonance can addition- problems connected with describing the gross struc- ally be taken into account by using recipes proposed ture of a giant dipole resonance in light nonmagic in [2–5]. nuclei. This structure is formed, first, by the defor- mation splitting of a giant dipole resonance; second, 2. CONFIGURATION SPLITTING OF by its configuration splitting, which is due to the fact SINGLE-PARTICLE E1 TRANSITIONS IN that, in light nuclei, the energies of single-particle E1 LIGHT NUCLEI transitions between principal (oscillator) shells (from a filled inner shell to the valence shell and from the As was indicated above, principal (oscillator) valence shell to an unfilled outer shell) strongly differ shells in light nuclei are separated by unequal energy from one another; and, finally, by the isospin splitting gaps. This is illustrated in Fig. 1, which displays of a giant dipole resonance. data on the energies of the knockout of s-, p-and Each of the above types of giant-resonance split- d-wave protons from the first three oscillator shells ting has received quite an adequate study in the corresponding to the excitation of N =0,1,and2 literature. For example, a theoretical analysis of the oscillator quanta [9]. spectrum of various isospin E1 modes was performed There is a wide scatter of the experimental val- as far back as 1965–1968 in [2–5], where simple ues of the energies EN l in the figure. This is due theoretical estimates were obtained for the isospin not only to errors in the relevant measurements but splitting of a giant dipole resonance. Ten years earlier, also to the effect of spin–orbit splitting of single- Danos [6] and Okamoto [7] successfully explained particle levels (for l =0 ) and (in unfilled shells) the the deformation splitting of a giant dipole resonance effect of nucleon–nucleon correlations on these lev- within a hydrodynamic model. The configuration els. It can be seen from the figure that, for nuclei splitting of a giant resonance in light nonmagic nuclei heavier than oxygen, the spacing between the “0s”
1063-7788/03/6607-1229$24.00 c 2003 MAIK “Nauka/Interperiodica” 1230 ISHKHANOV, ORLIN
Energy, MeV
s
s
s 60 s s s s 0s s s
p s p p s p 40 p s p 11 s p p s s s p 1p p p 10 ? s p 7 p d 7 s 8 p p 2d 11 s p p d 20 s 6 s ? ? d p d p d d ? 10 p d 11 p p ? d s s ? d s s d s s p p s s s s s s s s 2s 7 p d s s s 7 11 p s f s d s f p s d f 10 p s 6 p 0 He Li Be B C N O F NaMg Al Si P S Cl Ar K Ca Sc Ti V Cr Mn Fe CoNi Zn Atomic number
Fig. 1. Energies of the knockout of s-, p-, and d-wave protons from the first three oscillator shells [9]. and “1p” curves remains virtually unchanged as the independent of the number of nucleons occurring in charge number Z is increased. For nuclei lighter than this shell and in higher lying shells. oxygen—that is, in the region where the 1p shell is not Analyzing experimental data on the energies E1p, closed—these curves somewhat approach each other, E2d,andE2s and considering that the 2s-wave pro- which may be attributed to the effect of residual forces tons shift the centroid of the 2d2s shell only slightly, (for example, pairing), which shift the energy E1p we can arrive at the conclusion, in a similar way, that upward. We are interested here in the mean spacing the mean energy of single-particle transitions from between the 0s and 1p single-particle oscillator shells. the filled 1p shell to the 2d2s shell is also independent This quantity is not expected to depend on proton of the number of protons occurring in states that lie higher than the 1p shell. pairing. In estimating the difference E0s − E1p over 4 → 16 In all probability, this property is common to all the He O segment, where the filling of the closed shells. It implies a significant spatial separa- 1p shell occurs, it is therefore necessary to take into tion of nucleons occurring in different principal shells account only data on odd-Z nuclei (see Fig. 1). After of a nucleus. Indeed, it is clear that, if nucleons be- − this correction, the quantity E0s E1p appears to be longing to outer shells are situated by and large far- approximately constant for any values of Z greater ther from the center of the nucleus being consid- than that for helium. This means that, for a first ered than nucleons belonging to inner shells, then approximation, the mean energy of single-particle the energy of a nucleon transition from a filled inner transitions from the filled 0s shell to the 1p shell is shell to the next higher lying shell will be determined
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 SEMIMICROSCOPIC DESCRIPTION OF THE GROSS STRUCTURE 1231 predominantly by the mean field generated by “inter- of the nucleus, the greater the configuration splitting. nal” (core) nucleons. A spatial separation of shells In light nonmagic nuclei, there can occur a situation is directly corroborated by fluctuations of the radial where a major part of nucleons reside in the valence dependence of the nuclear density ρ(r) (see, for ex- shell, while, in heavy nuclei, the greater part of the ample, [14]). nuclear mass is always concentrated in filled inner Let us now estimate the mean spacings between shells. It therefore comes as no surprise that the con- the 0s, 1p,and2d2s shells. This can be done by figuration splitting of a giant dipole resonance man- two methods: either with the aid of experimental data ifests itself primarily in light nonmagic nuclei. (This presented in Fig. 1 or with the aid of the formula [15] statement can be illustrated by considering the exam- ple of the 12Cand 159Tb nuclei, for which the above (0) ≈ −1/3 ωcore 41Acore MeV, (1) (0) (0) ratio of the mean transition energies, ω2 / ω1 , (0) takes values of about 1.44 and 1.06, respectively.) where ωcore is the frequency parameter of the os- cillator potential simulating the mean field that is generated by Acore core nucleons (Acore =4for the 3. DESCRIPTION OF THE MODEL USED → 0s 1p single-particle transition, and Acore =16for In this section, we will first present fundamentals → the 1p 2d2s transition). of the semimicroscopic model of dipole vibrations that Using the approximating curves in Fig. 1, we find, was proposed in [12] and then show how it can be in the first case, that the 1p shell is separated by generalized to take into account the configuration and approximately 22.5 MeV from the 0s shell and by the deformation splitting of a giant dipole resonance. approximately 18.5 MeV from the 2d2s shell. In the second case, we instead obtain 25.8 and 16.3 MeV, respectively. It can easily be seen that the discrep- 3.1. Semimicroscopic Model of Dipole Vibrations ancy between the two types of estimates does not Basic properties of a nuclear dipole resonance can exceed the uncertainty in the approximating curves be explained in terms of the interplay of single-particle “0s” and “1p.” Indeed, perfect agreement between the nucleon excitations and the isovector dipole field that estimates can be attained if we draw the first curve is generated by these excitations and which, for vibra- somewhat higher and the second curve somewhat tions along the x axis, is given by lower (there are sufficient grounds for doing this). A | | + We now have all the required data at our dis- F = (2tzx)i = α 2tzx β aα aβ + h.c., (3) posal for estimating the configuration splitting of i=1 α>β single-particle dipole transitions. In a nonmagic + nucleus, two types of single-particle E1 transitions where aλ and aλ are the operators of, respectively, are possible: type-1 transitions proceed from the creation and absorption of a nucleon in a single- valence shell of the nucleus to an empty outer particle state |λ and tz = ±1/2 is the isospin variable shell, while type-2 transitions proceed from the last for a nucleon. filled inner shell to the valence shell. The nuclear The application of the dipole operator F to the shell mean field in which valence nucleons move can be ground state of a nucleus (physical vacuum |0 )gen- approximately described in terms of the oscillator erates a superposition of particle–hole (1p1h) states; potential corresponding to the oscillator-quantum that is, (0) ≈ −1/3 energy of ω 41A MeV, where A is the | | | + | mass number of the nucleus. Therefore, the mean F 0 = α 2tzx β aα aβ 0 . (4) (0) ≈ (0) α>β energy of type-1 transitions is ω1 ω .On the other hand, the mean energy of type-2 transi- In the case of heavy and medium-mass spheri- tions can be calculated by formula (1), which yields cal nuclei, the state F |0 is approximately an eigen- (0) (0) −1/3 state of the single-particle Hamiltonian H ,thecor- ω ≈ ωcore ≈ 41Acore MeV. 0 2 responding excitation energy being Thus, the ratio of the mean energies for two types − of single-particle E1 transitions in nonmagic nuclei ω(0) ≈ 41A 1/3 MeV, (5) can be approximated by the expression since the energy scatter of single-particle E1tran- ω(0) A 1/3 sitions is much less than ω(0). The excitation F |0 2 ≈ . (2) (0) is formed by a great number of single-particle exci- ω Acore 1 tations. In view of this, the operator F can be repre- From this formula, it can be seen that the greater sented in the form ∗ the contribution of valence nucleons to the total mass F = f (0)c(0)+ + f (0) c(0), (6)
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1232 ISHKHANOV, ORLIN where c(0)+ and c(0) are the operators of creation This estimate of the energy of dipole vibrations is in and annihilation of a vibrational quantum of energy good agreement with the experimental values of the ω(0), these operators approximately satisfying com- giant-dipole-resonance energy for heavy nuclei. mutation rules for bosons, and f (0) is the probability In the representation of normal vibrations (eigen- amplitude for the production of such a quantum by the states), the dipole-moment operator F has the form ∗ dipole field F . F = fc+ + f c, (15) The amplitude f (0) can be estimated with the aid where f is the probability amplitude for the excitation of the classical sum rule of normal vibrations. This amplitude is related to the 2 (0) ω(0)|f (0)|2 = A, (7) amplitude f by the equation 2M ω|f|2 = ω(0)|f (0)|2, (16) or it can be calculated directly by using the relation | (0)|2 | + | | | | |2 from which it follows that the canonical transforma- f = 0 F F 0 = α 2tz x β , (8) tion (11) has no effect on the oscillator strength of αβ dipole transitions. where M is the nucleon mass and where summation is performed over filled (β) and free (α) single-particle 3.2. Inclusion of the Configuration Splitting levels. of a Giant Dipole Resonance Single-particle dipole vibrations cannot be con- sidered as normal vibrations (eigenstates) of a nu- Ifthereisaconfiguration splitting of E1transi- cleon system about the equilibrium position, since tions, the dipole operator F must be broken down into they effectively interact with one another via the exci- two terms, tation of the isovector single-particle potential. This F = F1 + F2, (17) circumstance can be taken into account by supple- where the first term menting the single-particle Hamiltonian H0 with that κ 2 (1) | | + for dipole–dipole forces ,F /2, whereupon the vi- F1 = α 2tzx β aα aβ + h.c. (18) brational Hamiltonian assumes the form α<β 1 H = ω(0)c(0)+c(0) + κF 2. (9) exhausts single-particle transitions between the va- 2 lence and outer shells, while the second term By using the semiempirical Weizsacker¨ mass for- (2) | | + mula, the coupling constant for dipole–dipole inter- F2 = α 2tzx β aα aβ + h.c. (19) action can be expressed [12] in terms of the symmetry α<β V potential as takes into account transitions from a filled inner shell V to the valence shell and the transitions inverse to κ = , (10) 4Ax2 these. Following the same line of reasoning as in Sub- where x2 ≈R2/5 ≈ (1.2A1/3)2/5 fm2 is the mean- section 3.1, we introduce the operators of creation x (0)+ (0)+ (0) (0) square value of the coordinate of intranuclear nu- (c , c ) and annihilation (c , c )ofsingle- cleons. 1 2 1 2 particle quanta whose energies are (in MeV) With the aid of the linear canonical transformation (0) ≈ −1/3 (0) ≈ −1/3 c+ = Xc(0)+ − Yc(0) (X2 − Y 2 =1), (11) ω1 41A and ω2 41Acore (20) the Hamiltonian in (9) can be reduced to the diagonal (see Section 2); further, we employ the approximation form (0) (0)+ (0)∗ (0) Fi = fi ci + fi ci (i =1, 2), (21) H = ωc+c + const, (12) where the probability amplitudes for the excitation of where single-particle vibrations are given by ω = ( ω(0))2 +2κ ω(0)|f (0)|2 (13) | (0)|2 (i) | | | |2 fi = α 2tz x β (i =1, 2). (22) is the energy of a normal mode of dipole vibrations. α<β From (5), (7), (10), and (13), we find at V ≈ The next step consists in representing the vibra- 130 MeV that tional nuclear Hamiltonian in the form −1/3 ω ≈ 80A MeV. (14) H = H1 + H2 + H12, (23)
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 SEMIMICROSCOPIC DESCRIPTION OF THE GROSS STRUCTURE 1233 where of the calculations, which reproduces the calcula- (0) (0)+ (0) 1 tions described in Subsection 3.1, the Hamiltonian in H = ω c c + κ F 2 (24) Eq. (23) reduces to the form 1 1 1 1 2 1 1 2 is the Hamiltonian that describes normal vibrations H = ω c+c + κ F F + const, (29) for type-1 dipole transitions, i i i 12 1 2 i=1 (0) (0)+ (0) 1κ 2 + H2 = ω2 c2 c2 + 2F2 (25) where ci and ci are the operators of, respectively, cre- 2 ation and annihilation of a quantum of eigenvibrations generated by the Hamiltonian H ; is the analogous Hamiltonian for type-2 dipole tran- i sitions, and (0) 2 κ (0)| (0)|2 ωi = ( ωi ) +2 i ωi fi (30) H12 = κ12F1F2 (26) is the energy of such a quantum; is the operator that takes into account the interaction + ∗ of dipole transitions of the two types. Fi = fici + fi ci (31) In order to take into account the possible distinc- is operator Fi in the representation of the eigenstates tions for three types of 1p1h 1p1h interactions— of the Hamiltonian Hi;andfi is the probability ampli- +| 1 1, 2 2,and1 2—we have introduced three tude for the excitation of vibrations of the ci 0 type, different coupling constants for dipole–dipole forces this amplitude being related to the amplitude f (0) by κ κ κ i ( 1, 2,and 12). This can be justified by the follow- the equation ing considerations. According to the estimate in (10), | |2 (0)| (0)|2 the coupling constant for dipole–dipole forces is pro- ωi fi = ωi fi . (32) −5/3 portional to A . This mass dependence is obvi- Let us estimate the ratio of the energies ω and κ 2 ously appropriate for the constant 1 and, to some ω . We assume that the fraction q of the valence κ κ 1 extent, for the constant 12, but the constant 2 is shell is not occupied and that its fraction (1 − q) is −5/3 most likely to be proportional to Acore ,sinceouter accordingly filled. Using the dipole sum rule, we then (valence) nucleons have but a slight effect on the obtain oscillator excitations of nucleons belonging to a filled 2 ω(0)|f (0)|2 ≈ qA , (33) inner shell (see the analysis of data on quasielastic 2 2 2M core nucleon knockout in Section 2). Moreover, the in- 2 (0)| (0)|2 ≈ − teraction within the group of type-1 excitations is ω1 f1 (A qAcore). expected to be additionally suppressed in relation to 2M the interaction within the group of type-2 excita- By using these estimates and relations (20), (27), tions, since the former occurs at the nuclear periphery, and (30), we obtain where the mean density of nuclear matter is lower. 1/3 K Further, it can also be assumed that the intergroup ω2 ≈ A 1+0.0247q 2 − K . interaction is weaker than the intragroup interaction ω1 Acore 1+0.0247(1 qAcore/A) 1 because the overlap integral is small for type-1 and (34) type-2 particle–hole states. The configuration splitting of a giant dipole res- Taking the aforesaid into account, we represent onance does not play a significant role either in the the coupling constants for dipole–dipole interaction case where the valence shell is nearly empty (q → 1) in the form or in the case where it is nearly filled (q → 0). But in κ K −5/3 κ K −5/3 other cases, it can be seen from relations (28) that the 1 = 1A , 2 = 2Acore , (27) second factor on the right-hand side of (34) is close to −5/3 κ12 = K12A , unity. Therefore, we assume that (0) K K K ω ω A 1/3 where the factors 1, 2,and 12 satisfy the relations 2 ≈ 2 ≈ . (35) (0) ω1 ω Acore K12 K1 K2 ≈ 100 MeV. (28) 1 In the representation specified by Eqs. (23)–(26), We diagonalize the Hamiltonian in (23) in two the coupling constants κ1, κ2,andκ12 play the role steps: first, we find the eigenstates of the Hamilto- of free parameters of the model. Upon recasting the nians H1 and H2, whereupon we take into account Hamiltonian H into the form (29), it becomes possi- the interaction of these states. After the first step ble to choose parameters in an alternative way: ω1,
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1234 ISHKHANOV, ORLIN
5/3 ω2,andK12 = κ12A . This parametrization has a 3.3. Generalization to the Case of Deformed Nuclei number of advantages: on one hand, the number of In a spherical nucleus, normal dipole vibrations in independent variable parameters decreases by virtue three mutually orthogonal directions are degenerate of the condition in (35); on the other hand, the region in energy. In view of this, we have only considered where phenomenological dipole–dipole forces are ap- above vibrations along the x axis. In a deformed plied, which are only used to describe a relatively weak spheroidal nucleus, this degeneracy is partly removed. interaction of different configuration modes, shrinks. In that case, it is necessary to distinguish between The Hamiltonian in (29) can be diagonalized by vibrations along the symmetry axis of the nucleus means of the linear canonical transformation (z axis) and vibrations along an orthogonal direction 2 (say, along the x or the y axis). Each of these types + + − cˆi = (Xijcj Yijcj), (36) of vibrations is described in precisely the same way j=1 as vibrations along the x axis of a spherical nucleus (see Subsection 3.2). It is only necessary to employ, where the expansion coefficients X and Y satisfy ij ij in Eqs. (18), (19), and (22), the single-particle states the orthogonality conditions |α and |β of the deformed single-particle potential 2 and to take into account the effect of deformation ∗ − ∗ (XijXkj YijYkj)=δik, (37) on the oscillator frequencies (20) of single-particle j=1 (0) excitations—that is, to multiply the quantities ω1 2 (0) and ω by the factor 1 − 4δ/3 (by δ , we mean (X Y − Y X )=0. 2 ij kj ij kj here the deformation parameter that was introduced j=1 in [15]) for longitudinal vibrations (vibrations along The eigenenergies of the Hamiltonian H are given the z axis) or by the factor 1+2δ/3 for transverse by vibrations (vibrations along the x and y axes). 2 2 2 2 2 2 − 2 2 2 Relation (35) can now be recast into the form ω1 + ω2 ∓ ( ω1 ω2) ωˆi = (38) || ⊥ 1/3 2 4 ω2 ≈ ω2 ≈ A || ⊥ , (42) ω Acore 1/2 1/2 ω1 1 κ2 | |2 | |2 +4 12 ω1 f1 ω2 f2 , || || where ω1 and ω2 are the energies of, respectively, type-1 and type-2 intermediate normal vibrations for ≤ where i =1, 2 and ωˆ1 ωˆ2. the longitudinal mode of a giant dipole resonance, The dipole operator F reduces to the form ⊥ ⊥ while ω1 and ω2 are the analogous quantities for 2 its transverse mode [compare with the energies ω1 ˆ + ˆ∗ F = (ficˆi + fi cˆi), (39) and ω2 of the Hamiltonian in (29)]. i=1 The equalities in (42) make it possible to reduce ˆ the number of parameters varied in describing the where fi are the probability amplitudes for the excita- structure of a giant resonance in deformed nuclei to +| tion of the normal vibrations cˆi 0 . These amplitudes three; for these, one can use, for example, the energies can be found from the equation || ⊥ K ω1 and ω1 and the constant 12. | ˆ|2 | |2 2 2 − 2 2 ωˆi fi =[ ω1 f1 ( ωˆi ω2) (40) | |2 2 2 − 2 2 + ω2 f2 ( ωˆi ω1) 4. APPLICATION TO LIGHT NONMAGIC κ | |2 | |2 2 2 − 2 2 − 2 2 NUCLEI +4 ω1 f1 ω2 f2 ]/(2 ωˆi ω1 ω2) (i =1, 2). The model considered above was used to describe the gross structure of a giant dipole resonance in Finally, the relations the 12C, 24Mg, and 28Si nuclei. These are three self- 2ωˆ2 − 2ω2 conjugate nonmagic nuclei, for which the effects of |X |2 −|Y |2 = i 2 , (41) both the configuration and the deformation splitting i1 i1 2 2 − 2 2 − 2 2 2 ωˆi ω1 ω2 of a giant dipole resonance are of importance. 2ωˆ2 − 2ω2 We employed the following computational scheme. |X |2 −|Y |2 = i 1 i2 i2 2 2ωˆ2 − 2ω2 − 2ω2 First, the energies and oscillator strengths of normal i 1 2 excitations for longitudinal and transverse dipole vi- specify the contribution of type-1 and type-2 config- brations were calculated on the basis of the formalism +| urations to the dipole states cˆi 0 (i =1, 2). developed in Section 3. After that, the resulting dipole
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 SEMIMICROSCOPIC DESCRIPTION OF THE GROSS STRUCTURE 1235 resonances, whose number is equal to four with to reproduce, in the calculations, the positions of the || allowance for the degeneracy of vibrations along the x centroids E and E⊥ of dipole states for the longitu- and y axes, were approximated by Lorentzian curves. dip dip dinal and transverse modes of vibrations. After that, a Below, we expound on the details of our computa- different value was chosen for the parameter K12,and tional procedure. the procedure was repeated. || ⊥ The energies Edip and Edip were calculated with 4.1. Choice of Single-Particle Potential the aid of the relations In order to calculate the single-particle states |α , ⊥ || 2Edip + Edip we employed the Nilsson spheroidal potential [16], Edip = , (46) setting its parameters to 3 ⊥ E 2 − 4 dip z 1+2δ /3 ω(0) =41A 1/3 1 − δ, (43) = = . z 3 || x2 1 − 4δ/3 Edip − 2 ω(0) = ω(0) =41A 1/3 1+ δ, The giant-resonance energy E ,whichappears x y 3 dip in these relations, was estimated by the formula [18] 2 δ = δ/(1 + δ), 2 2 3 − 1+π η E ≈ 86A 1/3 [MeV], where δ is the parameter of the quadrupole deforma- dip 1+10π2η2/3+7π4η4/3 tion of a nucleus, (47) 3 Q δ = 0 . (44) where the quantity η was defined in (45). 2 4 Z r The best agreement with experimental data for 2 K ≈ Here, Q0 is the internal quadrupole moment and r the nuclei considered here was obtained at 12 is the mean-square radius of the nuclear-charge dis- 25 MeV. As might have been expected, this value, tribution. which characterizes the strength of interaction of dif- ferent dipole configurations, is rather small in relation For the nuclei considered here, the parameter δ to a value of about 100 MeV, which follows from the was calculated theoretically by means of the proce- estimate in (10). In order to assess the sensitivity of dure described in [17]. This procedure leads to over- calculations to the choice of value for the constant estimated values of the parameter δ for light nuclei, 28 K12, we have performed calculations for the Si nu- since it disregards the effect of nuclear-surface dif- cleus, employing various values of this constant. The 2 2 2 fuseness on x , y ,and z . The value corrected results are given in Fig. 2. with allowance for this effect can be obtained by the || ⊥ formula The intermediate energies ω1 and ω1 do not have an independent physical meaning. Nonetheless, 1+2π2η2 + 2 π4η4 15 it is interesting to note that, in varying the relevant δcorr ≈ δ , (45) 1+ 10 π2η2 + 7 π4η4 model parameters, they are grouped around values 3 3 that follow from formulas of the type in (30) if, in where η ≡ a/R0,witha ≈ 0.55 fm and R0 ≈ estimating the coupling constant for dipole–dipole 1.07A1/3 fm being, respectively, the diffuseness pa- forces [see Eq. (10)], use is made of a symmetry rameter of the nuclear surface and the distance potential V that faithfully reproduces the position between the center of the nucleus and the point at of the giant dipole resonance [compare the results which its density decreases by a factor of 2. in (14) and (47)]. The root-mean-square distinctions For the 12C, 24Mg, and 28Si nuclei, the eventual between the two types of estimates of the energies || ⊥ values of the parameter δ proved to be 0.13, 0.28, and ω1 and ω1 are less than 1 MeV. This confirms that 0.15, respectively. the approximations made in the model are physically justified. 4.2. Varying Model Parameters 4.3. Approximation of Resonance Widths || We varied the following three parameters: ω1 , ⊥ K Our calculations revealed (see Figs. 3–5) that ω1 ,and 12 (see Subsection 3.3). First, the value K type-1 and type-2 dipole configurations are weakly of the parameter 12 was fixed, whereupon the en- mixed in light nonmagic nuclei. On the other hand, || ⊥ ergies ω1 and ω1 were varied in such a way as the damping of vibrations in light nuclei is governed
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1236 ISHKHANOV, ORLIN
σ, mb σ, mb
20 (a) 50
10 25
0 0 10 15 20 25 30 E, MeV 20 (b)
Fig. 2. Results of our calculations for the 28Si nucleus at various values of the parameter K12. Points correspond to the experimental values of the photoabsorption cross 10 section from [19], while the dashed, the solid, and the dotted curve represent the results of the calculations with K12 =40, 25, and 10 MeV,respectively.
0 by different mechanisms for vibrations of different types: these are predominantly the emission of an 10 20 30 40 excited particle from a nucleus for type-1 vibrations E, åeV and the spreading of a collective mode over a large number of noncollective nuclear states interacting Fig. 3. Structure of the giant dipole resonance in the 12C with this mode for type-2 vibrations. It follows that, nucleus according to the calculations (a)withand(b) depending on the type of dominant configurations, the without allowance for the configuration splitting of the +| giant dipole resonance. Points represent the experimental normal excitation cˆi 0 (i =1–4) of a light nucleus values of the photoabsorption cross section from [20]. The ↑ may be assigned either the emission width Γ ( ωˆi) or description of the curves and histograms presented in this ↓ figure and in Figs. 4 and 5 below is given in the main body the spreading width Γ ( ωˆi). of the text. For 16 A 240 nuclei, the total width of a giant dipole resonance can be approximated by the for- mula [18] Each of these figures consists of two panels (a, b). In the panels carrying the label a, the experimental 0.0293 1+(π2/3)η2 Γ(E) ≈ 1 − 3η E2, (48) values of the photoabsorption cross section (points) 1+π2η2 1+π2η2 are contrasted against the results of the calcula- tion based on the model described above (curves and where, by η,weimplythesamequantityasin histograms). The panels labeled with b display the Eqs. (45) and (47). results of the calculation that was performed under As is well known, the approximate relation Γ ≈ the assumption that a giant dipole resonance is split Γ↑ +Γ↓ holds. We introduce the notation ν ≡ Γ↑/Γ. only into deformation modes whose energies are de- We then find that Γ↑(E) ≈ νΓ(E) and Γ↓(E) ≈ (1 − termined by relations (46) and (47). The thick solid ν)Γ(E),whereΓ(E) is given by expression (48). curves represent the theoretical values of the pho- In the present study, the dipole-resonance widths toabsorption cross section. In each case, two thin were estimated by using the values of ν =0.6,0.4, solid curves show the contributions to this cross sec- tion that come from, respectively, longitudinal and and 0.3 for 12C, 24Mg, and 28Si, respectively. In se- transverse dipole vibrations. Two dashed curves rep- lecting these values, we have taken into account the resent the configuration splitting of the giant dipole experimental trend toward a decrease in the contribu- resonance. The histograms depict the distributions of tion to Γ(E) from the emission width with increasing the oscillator strengths of dipole states (in arbitrary mass number A. units), the narrow and broad rectangles correspond- ing to, respectively, longitudinal and transverse dipole 5. DISCUSSION OF THE RESULTS modes. The open (closed) part of a rectangle repre- The basic results of our calculations are given in sents the contribution to a given state from type-1 Figs. 3–5. (type-2) configurations.
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 SEMIMICROSCOPIC DESCRIPTION OF THE GROSS STRUCTURE 1237
σ, mb σ, mb (a) 40 (a) 50
20 25
0 0 40 (b) ( ) 50 b
20 25
0 0 15 20 25 30 10 15 20 25 30 E, åeV E, åeV
Fig. 4. Structure of the giant dipole resonance in the Fig. 5. Structure of the giant dipole resonance in the 28Si 24Mg nucleus according to the calculations (a)withand nucleus according to the calculations (a)withand(b) (b) without allowance for the configuration splitting of the without allowance for the configuration splitting of the giant dipole resonance. Points represent the experimental giant dipole resonance. Points represent the experimental values of the photoabsorption cross section from [21]. values of the photoabsorption cross section from [19].
Figures 3–5 clearly demonstrate that the configu- from an inner filled shell [10]. The present calculation ration splitting of a giant dipole resonance plays a very basically confirms this point of view, but with one important qualification: the main contribution to the important role in the formation of its gross structure high-energy region of a giant dipole resonance is in light nonmagic nuclei. Indeed, experimental data formed by a collective mode that emerges from single- foranyofthenucleiconsideredherecannotbeade- particle excitations under the effect of residual forces quately reproduced without taking this phenomenon rather than by individual type-2 1p1h configurations. into account (compare panels a and b). The effect of Indeed, the residual interaction shifts this state (see the configuration splitting of a giant dipole resonance extreme right rectangles in Figs. 3–5) by aproxi- is especially pronounced in nuclei characterized by a 12 28 mately 8 MeV upward with respect to the energy small deformation ( C, Si). However, it can easily position of type-2 single-particle transitions. be singled out in the structure of the cross section for From the histograms presented in Figs. 3–5, it photoabsorption on the strongly deformed nucleus of 24 can be seen that type-1 and type-2 dipole configura- Mg as well (see Fig. 4). tions interact with each other rather weakly. This im- At the same time, the formation of the gross struc- plies that normal vibrational modes are formed owing ture of a giant dipole resonance is also greatly af- primarily to interactions of configurations belonging fected by the deformation splitting of a giant dipole to the same type. resonance (even in weakly deformed nuclei). This be- Finally, it should be noted that the main strength comes obvious in comparing the rectangles of the his- of dipole transitions from an inner shell to the valence tograms with the special features of the experimental shell (that is, of type-2 transitions) is associated with photoabsorption cross sections. the transverse mode of dipole vibrations. In the literature, the high-energy tails of the cross sections for photoabsorption on nonmagic 1p-and 6. CONCLUSIONS 2d2s-shell nuclei (see the notation in Section 2) are The model developed in the present study has been usually associated with single-particle excitations used to clarify the global features of the photoabsorp-
PHYSICS OF ATOMIC NUCLEI Vol. 66 No. 7 2003 1238 ISHKHANOV, ORLIN tion cross section in the mass range 10