Electric and magnetic dipole strength in 66 Zn

The dipole strength of the nuclide 66 Zn was studied in photon-scattering experiments using bremsstrahlung produced with electron beams of energies of 7.5 and 13.4 MeV at the γ ELBE facility as well as using quasimonoenergetic and linearly polarized photon beams of 30 energies within the range of 4.3 to 9.9 MeV at the HI γ S facility. A total of 128 J = 1 states were identiﬁed, among them 9 with 1 + and 86 with 1 − assignments. The quasicontinuum of unresolved transitions was included in the analysis of the spectra and the intensities of branching transitions were estimated on the basis of simulations of statistical γ -ray cascades. As a result, the photoabsorption cross section up to the neutron-separation energy was determined and compared with predictions of the statistical reaction model. The experimental M 1 strengths from resolved 1 + states are compared with results of large-scale shell-model calculations.


I. INTRODUCTION
Experimental and theoretical studies of photon strength functions have attracted growing interest in the past years [1,2].These functions describe average transition strengths in the energy region of the quasicontinuum of nuclear levels and are important ingredients for calculations of reaction cross sections within the statistical reaction model [3].Such calculations are used, for example, to obtain information about radiative neutron-capture cross sections of unstable nuclides for modeling the astrophysical s process [4,5].Experimental photon strength functions are determined by using various nuclear reactions, such as light-ion induced reactions, neutron capture (n, γ ), proton scattering (p, p ), and photon scattering (γ , γ ).In the latter, also called nuclear resonance fluorescence (NRF), angular momentum L = 1 is preferentially transferred by the incident photons and, for example in even-even nuclei, mainly states with spin and parity quantum numbers J π = 1 + and 1 − are excited [6].Energy-integrated scattering cross sections of these states and the photoabsorption cross section are determined from the intensities of the respective γ transitions back to the ground state.In this case, the transition energy in the strength function f (E γ ) is identical to the excitation energy in the photoabsorption cross section σ γ and the two quantities are connected via the relation σ γ = g(π hc) 2 E γ f (E γ ) with g = (2J x + 1)/(2J 0 + 1), where J x and J 0 are the spins of the excited and ground states, respectively.
The standard electric dipole (E 1) strength functions used in statistical-reaction-model calculations are Lorentz functions adjusted to (γ , n) reaction data that represent the isovector giant dipole resonance (GDR).Below the neutron-separation energy S n , several additional structures have been found on top of the low-energy tail of the GDR.Just below S n , a pygmy dipole resonance (PDR) appears [7][8][9], which is characterized by intense E 1 transitions and develops with increasing neutron excess [10].The magnetic dipole (M1) spin-flip resonance evolves at around 8 MeV and the M1 scissors mode evolves around 3 MeV in deformed nuclei [11].These structures manifest their presence, for example, in J π = 1 + states excited from the 0 + ground state.However, according to the Brink-Axel hypothesis [12,13], they can also be built on excited states.Indeed, resonance-like structures at energies of the scissors mode have been observed in strength functions obtained from light-ion induced reactions that include many transitions between excited states of various spins [14][15][16].A surprising feature in the shape of the dipole strength function is an upbend toward low energies (E γ 2 MeV), which has been observed in several nuclides using light-ion induced reactions, for example in 56,57 Fe [17,18], in Mo isotopes [19], in 105,106 Cd [20], and in Sm isotopes [15,16].This low-energy enhancement of dipole strength has been described in shellmodel calculations as being caused by large B(M1) transition strengths of transitions linking close-lying states of various spins at high excitation energy [21][22][23][24].For the J = 1 states populated in (γ , γ ) experiments, this effect is also expected, but difficult to observe because the low-energy γ rays have small intensities compared to those of the ground-state transitions and are usually hidden in the huge background of photons scattered by atomic interactions with the target material [25].However, the usual goal of (γ , γ ) experiments is the determination of scattering and absorption cross sections that are derived from the intensities of ground-state transitions.
In recent photon-scattering studies using broad-band bremsstrahlung at the γ ELBE facility [26] of the Helmholtz-Zentrum Dresden-Rossendorf (HZDR), Germany, as well as using quasimonoenergetic, polarized γ rays produced at the high-intensity γ -ray source (HIγ S) [27], operated by the Triangle Universities Nuclear Laboratory (TUNL) in Durham, North Carolina, USA, we observed enhanced E 1 strength in the PDR region from about 6 to 10 MeV, for example in nuclides around N = 50 [28][29][30], in nuclides below N = 82 [10,31], in 181 Ta [32], and in Pb isotopes [33,34].To investigate the possible influence of the PDR on cross sections and reaction rates, the experimental strength distributions were used as an input for the statistical reaction model [35,36].The experiments at HIγ S using polarized photon beams allow a distinction between the E 1 and M1 contributions to the photoabsorption cross section.An investigation of 128 Xe and 134 Xe including strength in the quasicontinuum of unresolved states proved that the main part (≈90%) of the photoabsorption cross section in the PDR region is of E 1 character whereas the M1 cross section gives only a relatively small contribution [37].This relation seems to change when going to lighter nuclides, such as the ones in the iron-nickel region.In 50 Cr [38], 52 Cr [39][40][41], 56 Fe [42,43], and 58,60 Ni [42,44,45], several strong isolated M1 excitations have been observed.To study the relation between E 1 and M1 strengths in this mass region for varying properties such as nuclear deformation, we studied a closed-shell nuclide, 54 Fe, and an open-shell nuclide, 66 Zn.While the results for 54 Fe were published elsewhere [46], we report the results for 66 Zn in this work.
In an earlier NRF experiment on 66 Zn, seven J = 1 states were observed at excitation energies between 3.3 and 4.8 MeV [47].In another experiment, the photoexcitation of the 7368 keV level by capture γ rays from lead was studied [51].
We performed experiments at γ ELBE using electron energies of 7.5 and 13.4 MeV to determine integrated scattering cross sections and spins of excited states and the photoabsorption cross section including contributions from the quasicontinuum of states up to S n = 11.1 MeV, and experiments at HIγ S using photon beams of energies from 4.3 to 9.9 MeV to determine the parities of the excited states.

A. The photon-scattering method
In photon-scattering experiments, the energy-and solidangle-integrated scattering cross section I s of an excited state at the energy E x is deduced from the measured intensity of the respective transition to the ground state.It can be determined relative to known integrated scattering cross sections.In the present experiments, we used the integrated scattering cross sections I s (E B x ) of states in 11 B [52] and their angular correlations including mixing ratios [53] as a reference: Here, I γ (E γ , θ ) and I γ (E B γ , θ ) denote the efficiency-corrected measured intensities of a considered ground-state transition at energy E γ and of a ground-state transition in 11 B at E B γ , respectively, observed at an angle θ to the beam.W (E γ , θ ) and W (E B γ , θ ) describe the angular correlations of these transitions.The quantities N N and N B N are the areal densities of nuclei in the 66 Zn and 11 B targets, respectively.The quantities γ (E x ) and γ (E B x ) stand for the photon fluxes at the energy of the considered level and at the energy of a level in 11 B, respectively.
The integrated scattering cross section is related to the partial width of the ground-state transition 0 according to where σ γ γ is the elastic scattering cross section, E x , J x , and denote energy, spin and total width of the excited level, respectively, and J 0 is the spin of the ground state.If a given level deexcites to low-lying excited states (inelastic scattering) in addition to the deexcitation to the ground state (elastic scattering), then the branching ratio b 0 = 0 / of the ground-state transition has to be known to deduce 0 .The γ -ray intensities and, hence, the deduced quantities I s and 0 are also distorted if a level is populated from higher-lying levels.This feeding can be reduced by choosing beam energies not far above the considered levels.In experiments with quasimonoenergetic photons, a few levels are excited in a narrow energy window and feeding of these levels is excluded.Spins of excited states can be deduced by comparing experimental ratios of γ -ray intensities, measured at two angles, with theoretical predictions.The optimum combination includes angles of 90 • and 127 • to the beam direction, because the respective ratios for the spin sequences 0-1-0 and 0-2-0 differ most at these angles.The expected values are W (90 • )/W (127 • ) 0-1-0 = 0.74 and W (90 • )/W (127 • ) 0-2-0 = 2.15 taking into account opening angles of 16 • and 14 • of the detectors placed at 90 • and 127 • , respectively, in the setup at γ ELBE [26].
Parity quantum numbers of excited states can be determined by measuring the linear polarizations of the groundstate transitions.These are deduced from experimental asymmetries of intensities measured parallel or perpendicular to the polarization plane of the incident photon beam in the experiments at HIγ S.

B. The target
The target consisted of 1.4993 g of zinc, enriched to 98.00% in 66 Zn, and shaped to a disk of 20 mm in diameter.
In the experiments at γ ELBE, the target was combined with 0.300 g of boron, enriched to 99.5% in 11 B, and also formed to a disk of 20 mm in diameter.The known integrated scattering cross sections of levels in 11 B were used to determine the photon flux.

C. Detector response
For the determination of the integrated scattering cross sections according to Eq. (1) the efficiencies of the detectors and the photon flux are needed.The detector response was simulated using the program package GEANT4 [54][55][56].The reliability of the simulations was tested by comparing simulated spectra with measured ones as described, for example, in Refs.[57][58][59][60][61].The determination of the absorption cross section requires in addition a correction of the experimental spectra for photons scattered by atomic processes induced by the impinging photons in the target material, and for ambient background radiation, which is described in Sec.III.
The absolute efficiencies of the high-purity germanium (HPGe) detectors in the setup at γ ELBE were determined experimentally up to 2.4 MeV from measurements with a 226 Ra calibration source.For interpolation, an efficiency curve calculated with GEANT4 and scaled to the absolute experimental values was used.A check of the simulated efficiency curve up to about 9 MeV was performed via various (p, γ ) reactions at the HZDR Tandetron accelerator.The efficiency values deduced from these measurements agree with the simulated values within their uncertainties [62].Similar results were obtained for the resonances at 4.44 and 11.66 MeV in 12 C populated in the 11 B(p, γ ) reaction at the TUNL van de Graaff accelerator [63].

D. Experiments with bremsstrahlung at γELBE
The nuclide 66 Zn was studied in two experiments at γ ELBE [26].Bremsstrahlung was produced using electron beams of 7.5 and 13.4 MeV kinetic energy.In the measurement at 7.5 MeV, the electron beam hit a niobium foil of 5 μm thickness acting as a radiator at an average current of about 670 μA.In the measurement at 13.4 MeV, the niobium foil had a thickness of 12.5 μm and the average current was about 490 μA.A 10 cm thick aluminum absorber (beam hardener) was placed behind the radiator to reduce the low-energy intensity of the bremsstrahlung spectrum in the measurement at 13.4 MeV.The photon beam was collimated by a 260-cm-long pure-aluminum collimator with a conical borehole of 8 mm in diameter at the entrance, 90 cm behind the radiator, and 24 mm in diameter at the exit.The target, placed 200 cm behind the collimator exit, was irradiated with a typical flux of about 10 9 s −1 in a spot of 38 mm in diameter.Scattered photons were measured with four high-purity germanium (HPGe) detectors with a full-energy efficiency of 100% relative to a NaI detector of 7.6 cm in diameter and 7.6 cm in length.All HPGe detectors were surrounded by escape-suppression shields made of bismuth germanate (BGO) scintillation detectors of 3 cm in thickness.Two HPGe detectors were placed vertically at 127 • relative to the photon-beam direction and a distance of 32 cm from the target.The other two HPGe detectors were positioned in a horizontal plane at 90 • to the beam and a distance of 28 cm from the target.Absorbers of 8 mm Pb plus 3 mm Cu were placed in front of the detectors at 90 • and of 3 mm Pb plus 3 mm Cu in front of the detectors at 127 • .Spectra of scattered photons were measured for 74 and 120 h in the experiments at 7.5 and 13.4 MeV electron energy, respectively.Part of a spectrum including events measured with the two detectors placed at 127 • relative to the beam at an electron energy of 13.4 MeV is shown in Fig. 1.
The absolute photon fluxes in the two measurements at γ ELBE were determined from intensities and known integrated scattering cross sections of transitions in 11 B. For interpolation, the photon flux was calculated using a bremsstrahlung computer code [64] based on the Born approximation with Coulomb correction [65] and including an atomic screening correction [66].In addition, the flux was corrected for the attenuation by the beam hardener.The calculated flux curves were scaled to the experimental values obtained at the energies of levels in 11 B. The experimental flux values and the calculated curves are presented in Fig. 2.
The measurements at two electron energies allowed us to identify inelastic transitions feeding low-lying from highlying levels.Transitions found in the measurement at E kin e = 7.5 MeV are assumed to be ground-state transitions.Transitions additionally observed up to 7.5 MeV in the measurement at E kin e = 13.4MeV are considered to be inelastic transitions from high-lying to low-lying excited states.By comparing the respective spectra, these inelastic transitions were sorted out.Besides, there is a number of transitions with energies that fit the difference between the energy of a higher-lying level and the first excited 2 + state.These transitions are also assumed to be inelastic transitions, if their intensity is smaller than that of the ground-state transition from the considered higher-lying level.The remaining ground-state transitions were used to derive the corresponding level energies, the integrated scattering cross sections of the states, and spin assignments deduced from angular distributions of the ground-state transitions, that are compiled in Table I.The integrated scattering cross sections of levels up to E x = 7.0 MeV were taken from the measurement at 7.5 MeV electron energy, because they are affected by feeding intensities in the 13.4 MeV measurement.To illustrate this influence, values of the two measurements are compared with each other and with the results of previous   work [47] in Table II.One sees that the feeding intensities decrease toward high energy and become negligible above about 5 MeV, which corresponds to the findings of previous studies [68,69].The transitions considered as inelastic transitions are listed in Table III together with the corresponding ground-state transitions and the ratios of their intensities.

E. Experiments with monoenergetic and linearly polarized photon beams at HIγS
Monoenergetic photon beams are produced at HIγ S by Compton back-scattering of a high-intensity free-electron laser (FEL) beam from an electron beam in the TUNL storage ring.Presently, the energy of the backward scattered photons can be tuned over a wide energy range, from about 1 to 100 MeV, by changing the energy of the electron beam and the FEL wavelength [27].The polarization of the FEL photons, defined by the magnetic field of the undulators, is mostly b Quantity deduced from the integrated scattering cross section determined from the measurement at 13.4 MeV electron energy according to Eq. ( 2).c Quantity deduced from the integrated scattering cross section determined from the measurement at 7.5 MeV electron energy (see Table I) according to Eq. ( 2).d Widths of the ground-state transitions 0 taken from Ref. [47].
preserved during the Compton backscattering process due to a negligible recoil effect, leading to the production of intense photon beams with nearly 100% polarization.The measurements at HIγ S were carried out at photonbeam energies of 4.3, 4.4, 4.55, 4.7, 4.85, 5.0, 5.15, 5.3, 5.45, 5.6, 5.75, 5.9, 6.1, 6.3, 6.5, 6.7, 6.9, 7.1, 7.3, 7.5, 7.7, 7.9, 8.15, 8.4, 8.65, 8.9, 9.15, 9.4, 9.65, and 9.9 MeV.The energy spread (FWHM) of the beam was about 3% of the energy using a 30.5-cm-longlead collimator with a cylindrical hole of 1.9 cm diameter positioned 60 m downstream from the collision point of the electrons with the FEL photons.The measuring time was on average about 3 h for each selected energy.The photon beam impinged onto the target with a flux of about 5 × 10 6 s −1 for the lowest energies and up to 3 × 10 7 s −1 for the highest energies.Scattered photons were measured with two HPGe detectors placed at polar angles of 90 • to the beam.One Counts per keV 66 Zn(γ,γ') HIγS E γ = 7.9 MeV

E1 M1
FIG. 3. Parts of spectra of photons scattered from 66 Zn, measured during the irradiation with a quasimonoenergetic polarized photon beam of 7.9 MeV at HIγ S. The spectrum plotted in red was measured with the vertical detector and contains E 1 radiation, whereas the spectrum plotted in blue was measured with the horizontal detector and contains M1 radiation.
detector of 80% relative efficiency was positioned vertical and one of 100% relative efficiency horizontal to the beam tube.The type of radiation (E 1 or M1) was deduced from a comparison of the intensities of the transitions measured at the different azimuthal angles relative to the polarization plane [53,70].In the present setup, E 1 radiation is detected preferentially in the vertical detector and M1 radiation in the horizontal one.As an example, spectra measured at a beam energy of 7.9 MeV are given in Fig. 3.The spectra allow a clear distinction between E 1 and M1 radiation.Azimuthal asymmetries A = (I γ H − I γ V )/(I γ H + I γ V ) deduced from the intensities I γ H measured in the horizontal detector and I γ V measured in the vertical one are given in Table I together with the resulting parities assigned to the emitting states.

III. DETERMINATION OF THE DIPOLE-STRENGTH DISTRIBUTION
The determination of the dipole-strength distribution and the related photoabsorption cross section requires the knowledge of the intensity distribution of the ground-state transitions and their branching ratios.As these cannot be derived directly from the measured spectra, we applied statistical methods to discriminate between γ rays from nuclear excitations and photons scattered by atomic processes, and to disentangle the intensity distributions of elastic and inelastic transitions in the quasicontinuum of nuclear levels.
First, a spectrum of the ambient background adjusted to the intensities of the transitions from 40 K and 208 Tl decay in the in-beam spectrum was subtracted from the measured spectrum.To correct the measured spectrum for the detector response, spectra of monoenergetic γ rays were calculated in steps of 10 keV by using the simulation code GEANT4.Starting from the high-energy end of the experimental spectrum, the simulated spectra were subtracted sequentially (spectrumstripping method) [71].
The background radiation produced by atomic processes in the 66 Zn target was obtained from a GEANT4 simulation.The simulated atomic background is compared with the responsecorrected spectrum in Fig. 4. The atomic background amounts in average to only a few percent of the intensity in the spectrum, but coincides with that above the neutron threshold, which proves the right magnitude.This behavior is similar to that found in previous studies [28][29][30]32,33,37,59,72] and shows that the experimental spectrum contains a considerable amount of nuclear strength in a quasicontinuum.This is formed by a large number of unresolved transitions with small intensities that are the result of the increasing nuclear level density at high energy in combination with the finite detector resolution.Because of the different orders of magnitude, the nuclear intensity distribution resulting from the subtraction of the simulated atomic background is not very sensitive to uncertainties of the latter, for which we assume 5%.The nuclear intensity distribution contains ground-state (elastic) transitions and, in addition, branching (inelastic) transitions to lower-lying excited states as well as transitions from those states to the ground state (cascade transitions).The different types of transitions cannot be clearly distinguished.However, for the determination of the photoabsorption cross section and the partial widths 0 , the intensities of the ground-state transitions are needed.Therefore, contributions of inelastic and cascade transitions have to be subtracted from the spectra.We corrected the intensity distributions by simulating γ -ray cascades from the levels in the entire energy region using the code γ DEX [60,73].This code works analogously to the strategy of the code DICEBOX [74] developed for (n, γ ) reactions, but in addition it includes also the excitation from the ground state.In the present simulations, level schemes (nuclear realiza-tions) including states with J = 0, . . ., 5 were created.Known low-lying levels were taken into account up to about 3.5 MeV.Partial widths were varied in the individual nuclear realizations applying the Porter-Thomas distribution [75].Level densities were calculated by using the constant-temperature model [76] with the parameters T = 1.42 (3) MeV and E 0 = 0.04 (11) MeV adjusted to experimental level densities [77].In the individual nuclear realizations, the values of T and E 0 were varied randomly within a Gaussian distribution with a σ corresponding to the uncertainties given in Ref. [77].The parity distribution of the level densities was modeled according to the information given in Ref. [78].
The first input for the photon strength function simulations were assumed to be Lorentz-shaped.For the E 1 strength, a sum of three Lorentz functions (TLO) that account for a triaxial deformation of the nucleus was used with parameters described in Refs.[79,80].In the present case, deformation parameters of β 2 = 0.22 [81] and γ = 29 • [82] were applied.The parameters for the M1 and E 2 strengths were taken from global parametrizations of M1 spin-flip resonances and E 2 isoscalar resonances, respectively [83].Spectra of γ -ray cascades were generated for groups of levels in energy bins of E = 100 keV.Starting from the high-energy end of the intensity distribution, that contains ground-state transitions only, the simulated intensities of the ground-state transitions were normalized to the experimental ones in the considered bin.The intensity distribution of the branching transitions was subtracted from the total intensity distribution.Applying this procedure step by step for each energy bin moving toward the low-energy end of the spectrum, one obtains the intensity distribution of the ground-state transitions.Simultaneously, the branching ratios b 0 (E ) of the ground-state transitions are deduced for each energy bin.In an individual nuclear realization, the branching ratio b 0 (E ) is calculated as the ratio of the sum of the intensities of the ground-state transitions from all levels in E to the total intensity of all transitions depopulating those levels to either any low-lying levels and the ground state [10,29,30,32,60,73].Branching ratios b 0 (E ) , averaged over the many nuclear realizations from the present cascade simulations, are illustrated in Fig. 5.
The uncertainty of the number of counts N (E ) in an energy bin of the experimental intensity distribution was deduced as where b(E → E ) is the branching intensity from bin E to bin E .We transform N (E ) to the scattering cross section according to with the quantities defined in Eq. ( 1), the absolute detector efficiency (E ), the measuring time t, and the bin width E .The absorption cross section in each bin is obtained as σ γ (E ) = σ γ γ (E )/b 0 (E ) for each nuclear realization.Finally, the absorption cross sections of each bin were obtained by averaging over the values of the individual nuclear realizations.
The simulations were performed iteratively, where the strength function obtained from an iteration step was used as the input for the next step.We note that the simulations are FIG. 5. Average branching ratios (solid curve) with their uncertainty ranges (dashed curves) resulting from the simulations of statistical γ -ray cascades up to S n as described in the text.little sensitive to the shape of the first input strength function, which was tested, for example, in Refs.[68,73].The iteration was stopped when the input strength function and the output strength function were in agreement within their respective uncertainties, which was achieved in the fourth iteration step.The cross section obtained in the last iteration step is adopted as the final absorption cross section.The uncertainties of the cross-section values include statistical uncertainties of the spectrum, the given uncertainties of the efficiency and the subtracted simulated background spectra, uncertainties of the flux resulting from the integrated cross sections of the 11 B levels, and the given uncertainties of the level-density parameters.Toward low energy, the uncertainties increase due to the use of the spectrum-stripping method and the cross sections do not converge.Therefore, cross sections cannot be determined below an excitation energy of 5.4 MeV in the high-energy measurement.The absorption cross section obtained from the procedure just described is listed in Table IV and graphed in Fig. 6.
To obtain values toward lower energies, we applied the same procedure to the low-energy measurement.However, this procedure starts at comparably small energies, at which the level density is small and the statistical method is at the limits of its applicability.For comparison, the result is also shown in Fig. 6.The cross section values are uncertain and fluctuate strongly.In particular, there are two steep decreases with a fluctuation to higher values in between.Besides, the values do not continue much further to low energies compared with the values from the high-energy measurement.This demonstrates the limits of the described methods and is not considered reliable for applications.Scattering and absorption cross sections can also be deduced from the experiments with quasimonoenergetic photon beams at HIγ S. The advantage is that only a bunch of states in an energy window of typically 200 keV is excited.These states are not affected by feeding and the corresponding inelastic transitions can clearly be identified, if they are intense enough.This allows in principle the determination of the cross sections without the application of model-dependent cascade simulations.There may be, however, also weak unresolved transitions in the spectrum, which requires a treatment of the quasicontinuum as well.The disadvantage of these experiments is the necessary unfolding of the detected energy distribution of the incident photon beam based on simulations, and the need of additional data for the calibration of the photon flux, which is usually performed on the basis of cross sections of individual transitions known from earlier FIG.6. Photoabsorption cross sections of 66 Zn resulting from the present (γ , γ ) experiments at E e = 7.5 MeV (blue triangles) and E e = 13.4MeV (red circles), from (γ , n) data for nat Zn (green squares) taken from Ref. [86], from calculations using the TALYS code as given in the TENDL-2019 library (black solid curve), and from the TLO with parameters given in the text (black dashed curve).experiments with bremsstrahlung.In previous experiments, we compared average branching ratios b 0 (E ) from experiments at both γ ELBE and HIγ S, and found a good agreement, for example for the nuclides 136 Ba [60] and 128,134 Xe [37].Moreover, the cross section in 136 Ba [60] deduced from experiments at γ ELBE is consistent with that in the neighbor 138 Ba obtained at HIγ S [84].A good agreement of the cross sections deduced from the described analysis of γ ELBE data with those deduced from HIγ S data was also achieved in a recent study of 87 Rb [85].We plan the determination of absorption cross sections in 66 Zn also on the basis of the present HIγ S data.For this purpose, the integrated cross sections in Table I can be used to deduce the photon flux in the selected beam-energy windows.This complex analysis is the subject of a forthcoming paper.

IV. DISCUSSION
The photoabsorption cross section resulting from the present experiments is compared with the cross section for the (γ , n) reaction [86] in Fig. 6.In addition, the TLO with the parameters just mentioned and the photoabsorption cross section given in the latest TALYS-based evaluated nuclear data library (TENDL-2019) [87] are displayed as well.Therein, the standard Lorentzian (Brink-Axel model) [2,13] was used for the (γ , γ ) reaction [88].The TENDL cross section is composed of the cross sections of various reaction channels shown in Fig. 7.The present (γ , γ ) cross section shows pronounced extra strength above the TLO in the energy region from about 6 MeV to S n , which is attributed to the PDR.The experimental cross section is well approximated by the TENDL cross section because of its smooth behavior.This differs considerably from the strongly fluctuating cross section of the N = 28 nuclide 54 Fe [46], which is compared with the one of 66 Zn in Fig. 8.Those fluctuations in 54 Fe are caused by prominent excitations manifesting in intense isolated peaks 54 Fe(γ,γ') 66 Zn(γ,γ') FIG. 8. Photoabsorption cross sections of 66 Zn from the present experiments (red circles) and of 54 Fe taken from Ref. [46] (blue squares).
in the spectrum of 54 Fe.In 66 Zn, there are fewer such intense peaks but a greater number of peaks with average intensity.There are 128 peaks between 3.4 and 11.0 MeV in 66 Zn (cf.Table I) compared to about 100 between 5.1 and 13.2 MeV in 54 Fe.The higher level density combined with fewer prominent peaks in the open-shell nuclide 66 Zn leads to a smoother curve for the photoabsorption cross section compared to that in the N = 28 nuclide 54 Fe.
Photon strength functions can also be determined in lightion induced reactions, applying the so-called Oslo method [89].While the strength functions deduced from (γ , γ ) data contain exclusively ground-state transitions from J = 1 states, the strength functions obtained from light-ion induced reactions comprise a large number of transitions linking many states of various spins up to about J = 10, which may cause different characteristics of the strength functions.The nearest neighbor of 66 Zn, for which such data are available, is 64 Ni.In Fig. 9, the photon strength function f deduced from the present cross section data for 66 Zn is compared with the one obtained from (p, p γ ) experiments for 64 Ni [90].The f of 66 Zn crosses the one of 64 Ni in the range from about 6 to 7 MeV and is nearly parallel, but greater by about 50% at higher energies, still lying inside the extreme limits of the f of 64 Ni.The latter continues toward small γ -ray energies, including transitions between close-lying states at high excitation energy, and displays the beginning upbend below about 3.5 MeV, which cannot be identified from conventional singles (γ , γ ) data (see Sec.I and Ref. [25]).
The relation between E 1 and M1 strengths in 66 Zn can be analyzed only for the intense transitions from states with parity assignments given in Table I.The summed integrated scattering cross section of all definite and tentative 1 − states in ) 66 Zn(γ,γ') 64 Ni(p,p'γ) FIG. 9. Photon strength functions obtained from the present 66 Zn(γ , γ ) data (red circles) and from 64 Ni(p, p γ ) data taken from Ref. [90] (blue squares and lines).found for 54 Fe [46] or other N = 28 isotones [39][40][41]45], this shows that going from closed-shell to open-shell nuclides results in an increasing portion of E 1 strength and a decreasing portion of M1 strength.
Reduced transition strengths B(M1) were deduced assuming branching ratios of b 0 = 1 for the ground-state transitions, because transitions from J π = 1 + states to the 2 + 1 state, which are expected to be the most intense branching transitions [84], were not identified.The experimental B(M1, 1 + → 0 + 1 ) values are presented in Fig. 10.simulated spectrum of photons scattered from the target by atomic interactions was subtracted.The remaining spectrum contains a continuum part in addition to the resolved peaks, which was included in the determination of the photoabsorption cross section.An assignment of inelastic transitions to particular levels and, thus, the determination of branching ratios was, in general, not possible.Therefore, simulations of statistical γ -ray cascades were performed to estimate intensities of branching transitions.These were subtracted from the experimental intensity distribution and the intensities of ground-state transitions were corrected on average for their branching ratios.In this way, a continuous photoabsorption cross section was derived for the energy range from about 5 MeV up to the neutron threshold at 11.1 MeV, which can be combined with the (γ , p) and (γ , n) cross sections toward higher energies.
The absorption cross section of 66 Zn displays an extra strength on top of the tail of a Lorentz function for the GDR in the range between 6 and 11 MeV that can be considered as the PDR.The shape of the PDR is smooth compared to the one in the nuclide 54 Fe, which is characterized by several peaks caused by intense resolved transitions typical for the PDR in nuclides at closed-shells.The smooth PDR is well approximated by cross sections calculated in the statistical model as given in the TENDL library.The strength function deduced from the present 66 Zn data is similar to the one obtained from Oslo data for the neighboring isotone 64 Ni.This shows, first, that the two methods produce roughly consistent results and, second, that the neighboring isotones have a similar structure.
The data indicate that the M1 strength in 66 Zn amounts to about 5% of the total strength in the resolved transitions.This represents a considerable reduction of M1 strength compared to about 45% in 54 Fe and other nuclides around the shell closure at N = 28.This means that the nuclear structure at closed shells induces strong M1, 1 + → 0 + 1 transitions, whereas the M1 contribution to the strength function is small in open-shell nuclei.
The experimental B(M1) values were compared with predictions of shell-model calculations.The calculations yield in general greater B(M1) values compared with the experimental ones.There are, therefore, more calculated than observed transitions with B(M1) > 0.005μ 2 N in the energy region under consideration.The experimental and calculated accumulative strengths differ by a factor of about 2, but reveal similar characteristics, a step caused by large values between about 4 and 5 MeV, continued by a flat plateau up to 7 MeV in the calculation and 8 MeV in the experiment.Calculations in model spaces without proton excitations over the Z = 28 gap are not capable of reproducing the experimental values, in particular not above about 5 MeV.This indicates the importance of those excitations for the generation of large B(M1) values in 66 Zn.

FIG. 1 .
FIG. 1. Part of a spectrum of photons scattered from 66 Zn combined with 11 B, measured during the irradiation with bremsstrahlung produced by electrons of an energy of E kin e = 13.4MeV.This spectrum is the sum of the spectra measured with the two detectors placed at 127 • relative to the beam at γ ELBE.

FIG. 2 .
FIG.2.Absolute photon flux on the target deduced from intensities of known transitions in11 B for the measurements at E e = 7.5 MeV (squares) and E e = 13.4MeV (circles) at γ ELBE.The curves represent the calculated flux described in the text.

FIG. 4 .
FIG.4.Spectrum of the two detectors at 127 • , corrected for detector response (blue), and simulated spectrum of photons scattered from the target to the detectors by atomic processes (black).

TABLE I
The uncertainty of this and the other quantities in the table is given in parentheses in units of the last digit.This energy value was deduced from the γ -ray energy measured at 127 • including a recoil and Doppler-shift correction.
[67]citation energy.bRatio of the intensities measured at angles of 90 • and 127 • .The expected values for an elastic dipole transition (spin sequence 0-1-0) and for an elastic quadrupole transition (spin sequence 0-2-0) are 0.74 and 2.15, respectively.cAzimuthalasymmetryA= (I γ H − I γ V )/(I γ H + I γ V ) of the intensities I γ H and I γ V measured with the detectors placed in a horizontal and vertical plane, respectively.A negative asymmetry indicates E 1 radiation and a positive asymmetry indicates M1/E 2 radiation.dSpinandparitydeduced from angular correlation and azimuthal asymmetry, respectively, of the ground-state transition.eEnergy-integratedscatteringcrosssection.Below an excitation energy of 7.0 MeV the value was deduced from the measurement at 7.5 MeV electron energy, otherwise the value was deduced from the measurement at 13.4 MeV.fThis state is assumed to be a known 2 + state[67].gThisstate is assumed to be a known 1 + state[67].h Up to about 40% of the intensity of the ground-state transition may belong to a single-escape peak.

TABLE II .
Comparison of level widths and spin and parity assignments deduced from the present measurements with those from previous work.

TABLE III .
Assumed inelastic transitions feeding the 2 + 1 state.
a Energy of the elastic transition to the ground state.b Energy of the inelastic transition to the 2 + 1 state.c Ratio of the intensities of the inelastic and elastic transitions.

TABLE IV .
Photoabsorption cross section deduced from the present (γ , γ ) experiment at E e = 13.4MeV.

TABLE IV
Table I amounts to I s (1 − ) = 5019(94) eV b and that of 1 + states amounts to I s (1 + ) = 278(20) eV b.The contribution of M1 strength is hence about 5% of the total dipole strength from the resolved states.Compared to the 45% contribution