JENDL photonuclear data file 2016

ABSTRACT

Nuclear reactions of gamma-rays with materials have attracted much attention in various application fields. To allow the use in those fields, JENDL photonuclear data file was updated from former JENDL/PD-2004 and was opened as JENDL/PD-2016 in 2017. The nuclear data were revised by adopting resonance analysis-like method for light nuclides and applying nuclear reaction models to the evaluations for heavier nuclides. In JENDL/PD-2016, standard and expanded versions are provided up to an incident energy of 140 MeV, covering 181 and 2681 nuclides, respectively. The evaluated results show better reproducibility to measured data, relative to JENDL/PD-2004. JENDL/PD-2016 is freely available from a website.

Graphical Abstract

KEYWORDS:

• Photonuclear data
• photo-induced reaction
• cross section
• gamma-ray
• JENDL/PD-2016

1. Introduction

Many experimental studies for photo-induced nuclear (photonuclear) reactions have used gamma-rays produced by methods with electron bremsstrahlung (e.g. [1]), annihilation in flight of positrons (e.g. [2]), and laser inverse-Compton scattering (e.g. [3]). They have investigated especially the form of giant dipole resonance (GDR) related to collective motion of neutrons and protons in a nucleus. These efforts result in providing photo-neutron and photo-fission cross sections for a variety of nuclides. These cross section data were collected, and the typical compilation was made by Dietrich and Berman [4].

Photonuclear data have drawn much attention from many application fields and subjects: shielding design of accelerator facilities [5], assessment and management of activation materials [6,7], impact assessment of photo-neutrons in a fusion reactor [8–10], estimation of transmutation of long-lived fission products [11], estimation of patient dose during radiotherapy [12,13], nuclear inspection, and safeguards by non-destructive assay for detecting fissile materials [14,15], radioisotope production for medical use [16], development of photo-induced neutron source [17], and nucleosynthesis in hydrostatic and explosive stages of stars [18,19].

The development of photonuclear data library became active around the world in the late 1990s. International Atomic Energy Agency (IAEA) held collaborative research projects for making photonuclear data library and released the product with 164 nuclides in 1999 [20,21]. Around the same time, LA150 developed in the USA was released and it contains the data of 12 nuclides up to an incident energy of 150 MeV [22]. Korean Atomic Energy Research Institute also published the photonuclear data library with 143 nuclides up to 140 MeV in 2000 [23]. The release of Japanese evaluated file, JENDL/PD-2004 [24], followed the above two libraries. These three libraries provided some of the data to develop the IAEA library. TENDL series have included comprehensive data up to 200 MeV for over 2600 nuclides based on nuclear reaction model calculations since 2008 [25]. ENDF/B-VIII.0 prepares the photonuclear data for 164 nuclides [26]. Updated collaborative research project hosted by IAEA from 2016 to 2019 developed photonuclear data library, IAEA/PD-2019, in which the nuclear data for 219 nuclides are included and maximum energy for some nuclides is extended up to 200 MeV [27].

JENDL/PD-2004 is the first version of photonuclear data file in JENDL series and it was released in 2004 [24]. In JENDL/PD-2004, the nuclear data for 68 nuclides are contained up to the gamma-ray energy of typically 140 MeV. The light and structural nuclides were mainly selected as well as ^235,238U and ²³⁷Np. Since the release of JENDL/PD-2004, improvement of cross sections and extension of nuclide range are required with expanding application fields for photo-induced reactions. Hence, new photonuclear data file of JENDL was developed [28] and released as JENDL/PD-2016 in December 2017.

This paper aims to explain the details on the development of JENDL/PD-2016. The evaluation methods of nuclear data and a data format are expressed in Section 2. Some evaluated results are shown in Section 3. Summary is presented in Section 4.

2. Evaluation methods and data format

First, the nomenclature related to photonuclear cross sections is defined in this section (see also [20,27]). Photo-absorption cross section is that gamma-rays are absorbed to nuclides. Photo-fission cross section is that gamma-rays trigger fission reaction on nuclides. Photo-neutron cross section represented by ($\gamma ,Sn$) is that reaction of nuclides with gamma-rays generates neutrons. In relation to the photo-neutron cross sections, the cross section of one-neutron emission reactions such as ($\gamma ,n$), ($\gamma ,np$) and ($\gamma ,n\alpha$) ones is collectively referred to as ($\gamma ,1nx$) reaction cross section. In the case of two-neutron emissions, it is expressed as ($\gamma ,2nx$). Photo-neutron yield cross section represented by ($\gamma ,xn$) is that neutron multiplicity is taken into account for each photo-neutron cross section.

JENDL/PD-2016 contains the nuclear data of photo-absorption cross sections, production cross sections of light particles, gamma-rays and residual nuclides, and energy-angle distributions (i.e. double differential cross sections with respect to solid angle and emission energy) of light particles and gamma-rays. If nuclides are fissile, the fission-related data (average number of total neutrons, prompt neutrons and delayed neutrons, photo-fission cross section, and energy-angle distribution of fission neutrons) are added. The list of used MF and MT in the ENDF-6 format [29] is summarized in Table 1. The maximum energy of the data is 140 MeV, which corresponds to the threshold of pion production. This upper limit is the same as JENDL/PD-2004 in which the upper limits exceeding 140 MeV are given for 19 nuclides (³He, ⁴⁰Ca, ⁶⁰Ni, ^{152,154–158,160}Gd, ¹⁸¹Ta, ^{196,198–202,204}Hg and ²⁰⁹Bi).

Table 1. List of MF and MT used in JENDL/PD-2016.

MF	MT		Description
	Non-fissile	Fissile
1	451	451	Descriptive data
1		452	Average number of total neutrons per fission
1		455	Average number of delayed neutrons per fission (if available in JENDL-4.0)
1		456	Average number of prompt neutrons per fission
3		3	Photo-absorption cross section
3	5	5	Total reaction (except photo-fission) cross section
3		18	Photo-fission cross section
4		18	Angular distributions of neutrons in photo-fission reaction
5		18	Energy distributions of neutrons in photo-fission reaction
5		455	Delayed neutron spectrum (if available in JENDL-4.0)
6	5	5	Energy-angle distributions of emitted particles, and
			residual production cross section
6		18	Energy-angle distributions of emitted neutrons in photo-fission reaction

Two versions are prepared in JENDL/PD-2016, in which the number of nuclides is different. One is the standard version, in which the nuclear data of 181 nuclides in and close to the $\beta$-stability line are prepared. This is for use to typical applications e.g. in estimating dose rate and photo-induced activation. The other is the expanded version aiming wider application fields regardless of conventional uses. The nuclear data of 2681 nuclides covering unstable nuclides far from the $\beta$-stability line are included. The nuclear data in the standard version are taken in the expanded version.

The nuclear data of ²H and ³He were adopted from JENDL/PD-2004. For ²H, MF3/MT3, MF3/MT28 and MF6/MT28 in the ENDF-6 format were adopted since ²H disintegrates into proton and neutron, regarding this process as the ($\gamma ,np$) reaction (MT28) with no residual nuclide. For ³He, the storage form of cross sections and energy-angle distributions for neutron, proton and deuteron productions was changed from JENDL/PD-2004, and consolidated into MF6/MT5. For the other nuclides, JENDL/PD-2004 contains the energy-angle distributions of emitted particles in MF6/MT201,203–207. In JENDL/PD-2016, these data are stored in MF6/MT5. The distributions of emitted gamma-ray are also given in MF6/MT5, except for ²H, ³He and light nuclides listed in Table 2.

Table 2. List of evaluated nuclides based on the resonance analysis-like method.

Element	Mass number	Element	Mass number
Li	6,7	Mg	24,25,26
Be	9	Al	27
B	10,11	Si	28,29,30
C	12,13	P	31
N	14,15	S	32,33,34,36
O	16,17,18	Cl	35,37
F	19	Ar	36,38,40
Ne	20,21,22	Ca	40,42,43,44,
Na	23		46,48

In the development of JENDL/PD-2016, three methods were applied for evaluation; (i) evaluation by resonance analysis-like method for light nuclides, (ii) evaluation by the nuclear reaction model code CCONE [30], and (iii) evaluation by the nuclear reaction model code ALICE-F [31]. Each method will be explained below.

2.1. Evaluation by resonance analysis-like method

Since light nuclides with mass number smaller than, e.g., 50 have sparse excited levels and result in wide level spacings, measured photo-neutron cross sections show resonance shapes. In the resonance regions of those nuclides a usual statistical model is not applicable. A method similar to the analysis of neutron resonance (namely resonance analysis-like method) is required to follow the shape of cross sections. In order to make systematic evaluations for light stable nuclides from Li to Ca, the following methodologies were applied, and are briefly explained in this paper. The details are found in Murata et al. [32].

When measured data of photo-absorption, photo-neutron and photo-neutron yield cross sections were available, the evaluations were made by resonance model and quasi-deuteron (QD) model [33,34]. The resonance cross section for photo-absorption was adopted with multi-level consideration [35] and given by

(1) ${\sigma }_{res}(E)=\sum _j\frac{2\pi }{k_j^2}\frac{2I_j+1}{2I_0+1}{B}_j\frac{{(E{\mathrm{\Gamma }}_j)}^2}{{(E_j^2-E^2)}^2+{(E{\mathrm{\Gamma }}_j)}^2},$(1)

where $E$ is the incident gamma-ray energy, $I_0$ is the spin of the target, subscript $j$ stands for the value in an excited state inferred from measured data and in the known excited states of ENSDF [36], $I_j$ is the spin, $B_j$ is the gamma-ray transition strength between the ground state and an excited state $j$, $E_j$, $k_j$, and ${\mathrm{\Gamma }}_j$ are the resonance energy in MeV, its wave number and resonance width in MeV, respectively.

On the other hand, when measured data were not available, a Lorentzian model with theoretical GDR parameters in RIPL-2 [37] was applied, or GDR structure was deduced by level information in ENSDF [36]. The photo-absorption cross sections (${\sigma }_{GDR}$) of the GDR expressed by the Lorentzian model are described by

(2) ${\sigma }_{GDR}(E)={\sigma }_0\sum _{i=1}^2\frac{a_i{(E{\mathrm{\Gamma }}_i)}^2}{{(E_i^2-E^2)}^2+{(E{\mathrm{\Gamma }}_i)}^2},$(2)

where ${\sigma }_0$ in mb is the normalization factor to meet the GDR sum rule [38], $a_1$ is the ratio of long to short axes for nucleus deformation and $a_2=1$, $E_i$ and ${\mathrm{\Gamma }}_i$ are the resonance energy and width in MeV, respectively, for collective motions along long and short axes. The photo-absorption cross sections (${\sigma }_{QD}$) by the QD process were calculated by

(3) ${\sigma }_{QD}(E)=L\frac{NZ}{A}{\sigma }_d(E)P_b(E)=R_{QD}\frac{NZ}{A}\frac{{(E-E_b)}^{3/2}}{E^3}{P}_b(E),$(3)

where $L$ is the Levinger parameter; $N$, $Z$, and $A$ are neutron, atomic, and mass numbers of target, respectively; and ${\sigma }_d(E)$ is the free deuteron photodisintegration cross section in mb given by

(4) ${\sigma }_d(E)=61.2\frac{{(E-E_b)}^{3/2}}{E^3},$(4)

$R_{QD}(=61.2L)$ is the normalization factor in mb MeV${ }^{3/2}$, $E_b$ is the binding energy of deuteron (2.223 MeV), and $P_b(E)$ is the Pauli blocking function [34]. The factor $R_{QD}$ was adjusted to reproduce measured data in the region where the QD process is dominant. The present GDR sum rule with the effect of the QD process is described by

(5) $\int \left({\sigma }_{GDR}+{\sigma }_{QD}\right)dE=C_s\frac{NZ}{A},$(5)

where $C_s$ is the normalization factor (typically $C_s=105$-110 mb MeV). The integration range is from the lowest energy to 140 MeV. This GDR sum rule was approximately imposed for the photo-absorption cross sections. After the determination of photo-absorption cross section, ($\gamma ,1nx$) and ($\gamma ,2nx$) reaction cross sections were evaluated using measured data. The cross sections of the remaining reactions were estimated by calculating branching ratios to the photo-absorption cross section by ALICE-F.

The decays to discrete levels by particle emissions are dominant in the low incident gamma-ray energy region. Those effects are expected to be seen in energy-angle distributions. However, since the information of discrete levels cannot be treated in ALICE-F, the energy-angle distributions for emitted light particles were calculated by CCONE. The normalized distributions were adopted for light nuclides.

The evaluated nuclides based on this framework are listed in Table 2, in which the nuclear data of ¹³C, ¹⁵N, ^17,18O, ^20–22Ne, ^32–34,36S, ^35,37Cl, ^36,38,40Ar, and ^42–44,46Ca were newly available in JENDL/PD-2016.

2.2. Evaluation by CCONE

The nuclear reaction model code CCONE [30] was used to obtain photonuclear data of nuclides listed in Table 3, in which most of nuclides are stable with $Z\ge 30$. CCONE comprises the Hauser-Feshbach statistical model and preequilibrium model. The incident gamma-ray energy region is from 1 MeV to the maximum energy. It should be noted that a simple photo-absorption cross section calculated by CCONE is compiled even in the energy region below a particle emission threshold (e.g. typically 5–10 MeV) where nuclear resonance fluorescence (NRF) is known to be prominent. In the present evaluation, the effect coming from NRF was not taken into account due to small cross sections, compared with particle emission ones.

Table 3. List of nuclides evaluated by CCONE.

Element	Mass number	Element	Mass number
Si	27	Cs	133,135,137
K	39,40,41	Pr	141
Sc	45	Sm	144,147,148,149,150,151,152,154
Zn	64,66,67,68,70	Gd	152,154,155,156,157,158,160
Ge	70,72,73,74,76	Tb	158,159
Sr	84,86,87,88,90	Ho	165
Zr	90,91,92,93,94,96	Ta	180$m$,181
Nb	93,94	W	180,182,183,184,186
Mo	92,94,95,96,97,98,100	Au	197
Tc	99	Hg	196,198,199,200,201,202,204
Pd	102,104,105,106,107,108,110	Pb	204,206,207,208
Ag	107,108,109	Bi	209
Cd	106,108,110,111,112,113,114,116	Th	232
Sn	112,114,115,116,117,118,119,120,122,124	U	233,234,235,236,238
Sb	121,123	Np	237
Te	120,122,123,124,125,126,128,130	Pu	238,239,241
I	127,129

In the statistical model calculations, information of excitation energy and gamma-ray decay branching ratios for discrete levels was taken from RIPL-3 [39]. Nuclear level density above the discrete levels was calculated by the composite form of constant temperature and Fermi-gas models [40], in which the Fermi-gas model was revisited by Mengoni and Nakajima [41]. Emission particles are considered to be neutron, proton, deuteron, triton, ³He and $\alpha$-particle. Emission probabilities of these particles were obtained by optical model calculations with potentials of Koning and Delaroche (KD03) [42] for neutrons and protons, Han et al. [43] for deuterons, folding model based on KD03 for tritons, Xu et al. [44] for ³He, and Avrigeanu and Avrigeanu [45] for $\alpha$-particles. Gamma-ray strength function is an important factor to estimate the photo-absorption (and photo-neutron) cross sections. In this evaluation, E1, M1, and E2 transitions were taken into account for the gamma-ray decays from continuum states. For the E1 transition, modified Lorentzian model (MLO1) [46] was utilized. The GDR parameters were adjusted so as to reproduce measured photo-absorption, photo-neutron, and photo-neutron yield cross sections. When measured data were not available, the systematics of the GDR parameters taken from RIPL-3 was employed. For M1 and E2 transitions, the formulations of Kopecky and Uhl [47] were applied.

The preequilibrium process was calculated with the two-component exciton model [48,49]. The initial particle-hole configuration was set to be $1p$-$0h$, which was assumed in the neutron shell. This configuration may be unphysical for photonuclear reactions, but it is validated based on the results that measured data of energy-angle distributions could be well reproduced rather than adopting more appropriate 1$p$-1$h$ and 2$p$-2$h$ configurations for the GDR and QD process, respectively.

The energy-angle distributions for emissions of the light particles and gamma-ray were calculated by the above models and were stored in MF6/MT5. Specially for light nuclides in Table 2, the normalized distributions were included only for light particles as explained in Section 2.1.

The average number of delayed neutrons per fission for ²³²Th, ^{233–236,238}U, ²³⁷Np, ^238,239,241Pu, and delayed neutron spectrum for ^{233–236,238}U, ²³⁷Np, ^239,241Pu were brought from the neutron nuclear data library, JENDL-4.0 [50]. The average number of prompt neutrons per fission (${\bar{\nu }}_p$) was estimated as the linear function with respect to gamma-ray energy below ($\gamma ,2n$) reaction threshold for ²³²Th, ^{233–236,238}U, ²³⁷Np, and ^238,239,241Pu [51,52]. Above the ($\gamma ,2n$) reaction threshold, an exponential function (Equation (6)(6) ${\bar{\nu }}_p(E)=\left\{\begin{array}{cc}{p}_0(E)& (E<E_0),\\ p_2+(a_2-p_2)(1-exp(-p_{1}E))& (E\ge E_0),\end{array}\right.$(6) [53,54]) was assumed and smoothly connected with the linear one at the threshold energy. For the other fissile nuclides, the following functions were applied:

$a_0=-0.00213A+0.649,$

$a_1=0.146A-32.8,$

$p_0(E)=a_{0}E+a_1,$

$p_1=a_0/(a_2-p_0(E_0)),$

$p_2=a_2-a_0/(p_{1}exp(-p_1{E}_0)),$

(6) ${\bar{\nu }}_p(E)=\left\{\begin{array}{cc}{p}_0(E)& (E<E_0),\\ p_2+(a_2-p_2)(1-exp(-p_{1}E))& (E\ge E_0),\end{array}\right.$(6)

where $a_0=0.1$ if $a_0\le 0.1$, $a_1=0.5$ if $a_1\le 0.5$, $a_2=12$, $E_0$ is the threshold energy of ($\gamma ,2n$) reaction, $a_0$ and $p_1$ are in MeV⁻¹, $E$ and $E_0$ are in MeV. The mass number dependences of $a_0$ and $a_1$ were derived by least square fits to the known parameters [51,52].

For the QD process, the same formalism as Equation (3)(3) ${\sigma }_{QD}(E)=L\frac{NZ}{A}{\sigma }_d(E)P_b(E)=R_{QD}\frac{NZ}{A}\frac{{(E-E_b)}^{3/2}}{E^3}{P}_b(E),$(3) was used, but the normalization factor $R_{QD}$, in which $L=6.5$ by default in CCONE, was modified to reproduce measured data. Note that $R_{QD}$ reduces to 120$E_b^{3/2}$ if adopting $L=6.5$.

2.3. Evaluation by ALICE-F

For nuclides up to ²⁶⁶Lr ($Z=103$) other than listed in Tables 2 and 3, the following calculation method was applied to deriving the nuclear data.

The photo-absorption cross section was calculated by the Lorentzian model of Equation (2)(2) ${\sigma }_{GDR}(E)={\sigma }_0\sum _{i=1}^2\frac{a_i{(E{\mathrm{\Gamma }}_i)}^2}{{(E_i^2-E^2)}^2+{(E{\mathrm{\Gamma }}_i)}^2},$(2) with GDR parameters in RIPL-3 and by the QD model. The GDR parameters were adjusted to match measured photo-absorption and photo-neutron cross sections. The QD model was the same as Equation (3)(3) ${\sigma }_{QD}(E)=L\frac{NZ}{A}{\sigma }_d(E)P_b(E)=R_{QD}\frac{NZ}{A}\frac{{(E-E_b)}^{3/2}}{E^3}{P}_b(E),$(3) , but $R_{QD}$ is modified as follows: $R_{QD}=159.2E_b^{3/2}$ for light nuclides, $R_{QD}=120E_b^{3/2}$ for nuclides with $Z\ge 83$, and $R_{QD}=(145-0.362Z)E_b^{3/2}$ for the other nuclides. The definition of the GDR sum rule is the same as Equation (5)(5) $\int \left({\sigma }_{GDR}+{\sigma }_{QD}\right)dE=C_s\frac{NZ}{A},$(5) , but the normalization factor $C_s$ was changed as follows: $C_s=110$ mb MeV for light nuclides, $C_s=115$ for nuclides with $20<Z<84$, and $C_s=115-2/3\cdot (Z-83)$ for nuclides with $Z\ge 84$. For the photo-absorption cross sections excluding photo-fission ones, the condition of GDR sum rule was approximately satisfied.

The ALICE-F code consists of a evaporation model [55] and a hybrid model for preequilibrium process. ALICE-F was employed to obtain ratios of particle-emission and residual production cross sections to the photo-absorption one, as well as energy-angle distributions of emitted particles calculated by the systematics of Kalbach [56]. As the emission particles, neutrons, protons, deuterons, tritons, ³He, $\alpha$-particles and gamma-rays were taken into account. The incident energy range is from a smallest threshold energy of particle emissions to the maximum energy, where the $Q$-values of photonuclear reactions were calculated with the atomic mass evaluated by Wapstra and Audi [57].

The calculation of photo-fission cross section was taken into account for nuclides with $Z\ge 66$ by the FISCAL2003 code [58].

3. Examples of evaluation results

In the present evaluation, the measured data were taken from EXFOR [59]. Some of these data were recommended to multiply a factor by cross sections published in the 1960s and 1970s, in order to correct the normalization of cross sections (e.g. [60]). In addition to this, cross sections were renormalized to be consistent with other more reliable or up-to-date measured data. The measured data, for which renormalization was made, are listed in Table 4.

Table 4. Normalization factors applied in the present evaluation.

Nuclide	Author	Factor
^64Zn	Carlos et al. [61]	0.75
^90Zr	Lepretre et al. [62]	0.88
^107Ag	Berman et al. [63]	1.28
^118Sn	Lepretre et al. [64]	0.9
^127I	Bergere et al. [65]	0.8
	Bramblett et al. [66]	1.14
^133Cs	Berman et al. [63]	1.09
^141Pr	Bramblett et al. [66]	1.04
^144,148,150,152,154Sm	Carlos et al. [67]	0.8
^159Tb	Bramblett et al. [68]	1.08
^165Ho	Berman et al. [63]	1.16
^181Ta	Bramblett et al. [69]	1.22
^186W	Berman et al. [63]	1.09
^197Au	Berman et al. [60]	1.07
^208Pb	Veyssiere et al. [70]	0.93
^206,207,208Pb	Harvey et al. [71]	1.22
^209Bi	Harvey et al. [71]	1.22
^232Th	Veyssière et al. [72]	1.26

Figure 1 shows the photo-neutron cross sections by ($\gamma ,n$) reaction for ²⁸Si in JENDL/PD-2016, compared with JENDL/PD-2004 and measured data [73–76]. Since the ($\gamma ,n$) reaction cross section is not normally included in many of photonuclear data files except for TENDL, the production cross section of residual nuclides is substituted for a comparison purpose. In the case of ²⁸Si, the ($\gamma ,n$) reaction cross section corresponds to the production cross section of ²⁷Si. Hereafter such substitution is made, when the reaction product coincides with the residual. The data of Pywell et al. [75], Caldwell et al. [76], and Goryachev et al. [73] were measured with neutron-detection techniques, by which the cross sections of ($\gamma ,1nx$) reaction can be obtained. The data of Webb et al. [74] were measured by activation technique and correspond to pure ($\gamma ,n$) reaction cross sections. The threshold energies of ($\gamma ,np$) and ($\gamma ,n\alpha$) reaction are 24.6 and 26.5 MeV, respectively. In the region between 17.2 and 24.6 MeV, ($\gamma ,n$) reaction is only responsible for neutron production. The data of Goryachev et al. [73] are larger than those of Pywell et al. [75] and Caldwell et al. [76] below 21 MeV, below which it seems that JENDL/PD-2004 reproduces the data of Goryachev et al. [73]. In JENDL/PD-2016, the data of Webb et al. [74] and Caldwell et al. [76] were adopted for evaluation and are well reproduced with resonance analysis-like method. Above 21 MeV, the present evaluation gives consistent results with the data of Caldwell et al. [76].

Figure 1. Comparison of the ($\gamma ,n$) reaction cross section for ²⁸Si in JENDL/PD-2016 with JENDL/PD-2004 and measured data [73–76]. The data of Webb et al. [74] are for ($\gamma ,n$) reaction, and the others are for ($\gamma ,1nx$) reactions.

Although the half-lives of generated ¹³N and ¹⁵O are short in the order of minutes, attention has been paid to the activation of air in electron linear accelerator (linac) facilities. From this point of view, the photonuclear cross sections of nitrogen, oxygen, and the other constituents of air were evaluated with the detailed analyses [32] in JENDL/PD-2016. Some evaluation examples of these nuclides were already presented in Refs [32,77]. Here, the neutron emission energy-angle distributions by photonuclear reaction of ¹⁴N at incident energy of 22.5 MeV and emitted angle of 90° are shown for JENDL/PD-2016 and JENDL/PD-2004, together with measured data [1] in Figure 2. The energy-angle distributions (MF6) for light particle emissions in JENDL/PD-2016 were adopted from the results calculated by CCONE as already mentioned. The neutrons following these distributions are expected to accompany the productions of ¹³N. It is found that the continuum component of the distribution in JENDL/PD-2016 is extremely small compared to heavier nuclides and that the direct transitions to discrete levels of ¹³N by neutron emissions from ¹⁴N is dominant at a low incident energy region, supplemented with the de-excitation neutrons from discrete levels of ¹³C generated by the ¹⁴N($\gamma ,p$) reaction. A prominent peak by de-excitation neutrons from ¹³C is seen around 6.5 MeV in the measured data. The present result unfortunately shows a tiny contribution of the de-excitation. JENDL/PD-2004 does not have any signature of the transitions from the discrete levels. As the incident energy goes up, the effect of direct transition to discrete levels becomes small. On the other hand, JENDL/PD-2004 represents a usual evaporation peak around 1 MeV. However, this peak does not exist in the measured data. Hence, the present data better explain the global feature of measured distribution than JENDL/PD-2004. Another feature can be seen in JENDL/PD-2004 which shows an abrupt drop above the neutron emission energy of 8 MeV. This drop comes from a data interpolation between the incident energies of 22 and 24 MeV given in MF6.

Figure 2. Comparison of the energy-angle distribution of emitted neutrons for ¹⁴N in JENDL/PD-2016 with JENDL/PD-2004 and measured data [1]. The distributions are illustrated at the incident gamma-ray energy of 22.5 MeV and neutron emission angle of 90°. The data of Sherman et al. [1] were derived with difference spectrum by subtraction of 21.5-MeV bremsstrahlung spectrum from 23.5-MeV one. The normalization of measured data was made on the basis of JENDL/PD-2016.

Manganese is contained in structural materials, although its amount is small. For comprehensive estimation of activation induced by photo-neutrons in an electron linac facility, the photonuclear data of Mn are required. Figure 3 compares the photo-neutron yield, ($\gamma ,n$) and ($\gamma ,2n$) reaction cross sections of ⁵⁵Mn in JENDL/PD-2016 with JENDL/PD-2004 and measured data [78–80]. The photo-neutron yield cross section below 20 MeV in JENDL/PD-2004 explains the data of Alvarez et al. [80]. In contrast, above 20 MeV it is unexpectedly large. The same situation is encountered for the ($\gamma ,n$) and ($\gamma ,2n$) reaction cross sections in JENDL/PD-2004. Such errors should be improved. The present photo-neutron yield cross section is consistent with the data of Alvarez et al. [80] above 18 MeV and Ishkhanov et al. [79]. The ($\gamma ,n$) and ($\gamma ,2n$) reaction cross sections in JENDL/PD-2016 are also reasonable.

Figure 3. Comparison of the photo-neutron yield, ($\gamma ,n$) and ($\gamma ,2n$) reaction cross sections for ⁵⁵Mn in JENDL/PD-2016 (solid, long-dashed and dot-dashed lines, respectively) with JENDL/PD-2004 (short-dashed, dotted and dot-dot-dashed lines, respectively), and measured data [78–80].

Copper is frequently used in a head of medical electron linac to generate neutrons. The photonuclear data of Cu is important for shielding design of electron linac facility. There were experiments to measure the angular distributions of neutron ambient dose equivalent rate (hereafter neutron dose rate) and ²⁷Al($n,p$) reaction rate at a 45 MeV electron linac of Hokkaido University [77,81,82]. In the experiments, neutrons were generated by a photonuclear reaction between bremsstrahlung photons and a natural Cu target (consisting of stable ^63,65Cu) bombarded with 18, 28, and 38 MeV-electrons. It was found that the results simulated with MCNP5 [83] and JENDL/PD-2004 were underestimated, compared with the measured rates in especially the 18 MeV-electron case [81]. In order to enhance the consistency with the measured rates, photo-neutron yield cross section larger than JENDL/PD-2004 was desired in the GDR region. The ($\gamma ,n$) reaction and photo-neutron yield cross sections of ⁶⁵Cu in JENDL/PD-2016 are shown with JENDL/PD-2004 and measured data [84–86] in Figure 4. The present evaluation gives larger cross sections than the data of Fultz et al. [84], especially in the incident energy region where the ($\gamma ,n$) reaction is dominant. The relative difference in ($\gamma ,n$) reaction cross sections reaches to about 40% between 14 and 17 MeV incident energies. The discrepancy might be bigger than an acceptable value (up to 30% in Table 4) for the correction of cross sections. In the MCNP5 simulation using JENDL/PD-2016, the angular distributions of neutron dose rate and ²⁷Al($n,p$) reaction rate increase and get closer to the measured ones in the 18 MeV-electron cases [77,82]. However, the resulting rates still remain underestimated. In Figure 4, Katz and Cameron obtained ($\gamma ,n$) reaction cross sections much larger than JENDL/PD-2016, whereas they pointed out that there was uncertain contribution of the ⁶⁴Cu production by the neutron capture reaction on ⁶³Cu in their activation measurement with a natural Cu target. Therefore, their data are considered to be overestimated. It should be mentioned that the present evaluation for ⁶³Cu also provides large cross sections for those reactions as is the case in ⁶⁵Cu [77].

Figure 4. Comparison of the ($\gamma ,n$) reaction and photo-neutron yield cross sections for ⁶⁵Cu in JENDL/PD-2016 (solid and long-dashed lines, respectively) with JENDL/PD-2004 (short-dashed and dotted lines, respectively) and measured data [84–86].

Tungsten is also used for neutron generation in a head of medical electron linac. The photonuclear data are requisite for the same purpose as Cu. JENDL/PD-2004 contains the data of only ^182,184,186W, for which the total natural abundance is 85.6%. In JENDL/PD-2016 all natural isotopes were prepared, and the nuclear data of natural W are available in, e.g. simulations. Figure 5 compares the photo-neutron cross section of natural W in JENDL/PD-2016 with JENDL/PD-2004 and measured data [87]. It should be noted that the data without the contributions of stable ^180,183W for JENDL/PD-2004 are drawn. The measured photo-neutron data include the contributions of ($\gamma ,n$), ($\gamma ,np$), ($\gamma ,n\alpha$), and ($\gamma ,2n$) reactions below the incident energy of 20 MeV. According to the CCONE calculations, the cross sections of ($\gamma ,np$) and ($\gamma ,n\alpha$) reactions are normally negligible for heavier nuclides such as W isotopes (smaller than 1 mb). In the evaluation of JENDL/PD-2016, the photo-neutron cross sections were evaluated for each isotope, using measured data of photo-absorption and photo-neutron yield cross sections. The peak cross section of JENDL/PD-2004 is clearly smaller than JENDL/PD-2016 and the measured data due to the lack of ^180,183W. This results in remarkable underestimation of amount of neutrons generated by tungsten sources.

Figure 5. Comparison of the photo-neutron cross section for ^natW in JENDL/PD-2016 (solid line) with JENDL/PD-2004 (dashed line) and measured data [87].

The ($\gamma ,n$) reaction cross section of ¹⁹⁷Au has been used as standard data for activation measurements. Many measurements with activation and neutron counting methods have been performed to obtain ($\gamma ,n$) and ($\gamma ,1nx$) reaction cross sections. First, the present ($\gamma ,n$) reaction cross section is depicted with JENDL/PD-2004 and the measured data [2,60,70,88–93] in Figure 6. Note that the contributions from one-neutron emission reactions other than ($\gamma ,n$) reaction are possibly incorporated in the data particularly measured by Fultz et al. [2], Veyssiere et al. [70], and Berman et al. [60]. However, the influence is likely to be small. JENDL/PD-2016 and JENDL/PD-2004 are consistent with the measured data within their uncertainties, although the most up-to-date data of Plaisir et al. [93] have remarkably small cross sections between 14 and 17 MeV. It should be noted that there is the difference between JENDL/PD-2016 and JENDL/PD-2004 around the threshold energy of ($\gamma ,n$) reaction (8.07 MeV). This is because in JENDL/PD-2016 the cross sections were provided every 0.5 MeV in an incident energy below 20 MeV and the cross section was set to be zero at and below 8.5 MeV for the ¹⁹⁶Au production.

Figure 6. Comparison of the ($\gamma ,n$) reaction cross section for ¹⁹⁷Au in JENDL/PD-2016 (solid line) with JENDL/PD-2004 (dashed line) and measured data [2,60,70,88–93].

Next, the comparisons of energy-angle and energy distributions of photo-neutrons emitted from ¹⁹⁷Au in JENDL/PD-2016 are made with JENDL/PD-2004 and measured data¹ [94,95] in Figure 7. In these comparisons, the incident energies of each measurement were taken from the end-point energies of electron bremsstrahlung. The energy-angle distributions in the incident energy of 14 MeV for JENDL/PD-2016 and JENDL/PD-2004 were derived at neutron emission angle of 90°. The energy distributions in the incident energy of 20 MeV were obtained by the summation of energy-angle distributions over neutron emission angles from 30° to 150° at 15° intervals. In the 14 MeV case, the present data give a consistent results with the measured data, although JENDL/PD-2004 shows abrupt drop around the neutron emission energy of 4 MeV. In the 20 MeV case, it is found that a non-statistical decay component seen above 4 MeV has an important contribution. A preequilibrium component in JENDL/PD-2016 reproduces the measured data in the region. In contrast, the distribution in JENDL/PD-2004 underestimates the measured data between 4.5 and 10 MeV, since the effect from preequilibrium process is weaker.

Figure 7. Comparison of the neutron emission energy-angle and energy distributions of ¹⁹⁷Au in JENDL/PD-2016 with JENDL/PD-2004 and measured data [94,95]. The energy-angle distribution (left axis) was measured with the bremsstrahlung end-point energy of 14 MeV at the neutron emission angle of 90°. The energy distribution integrated over neutron emission angles of 30° to 150° (right axis) was measured with bremsstrahlung end-point energy of 20 MeV. Both data are based on relative measurements, and are normalized to the integral value calculated by the present distributions over measured neutron emission energy ranges.

The photonuclear data of actinides such as ^235,238U and ²³⁹Pu are considered to be important for nuclear inspection and safeguards, in order to estimate properly the amount of fissile materials. Figure 8 shows the evaluated result of the average number of prompt neutrons per fission (${\bar{\nu }}_p$) for ²³⁸U compared to JENDL/PD-2004, a least-square fit by Caldwell et al. [51], and measured data [51,96,97]. The present data below ($\gamma ,2n$) reaction threshold are based on the least-square fit by Caldwell et al. [51] and start to leave the fit at 11.3 MeV, following the present formulation (Equation (6)(6) ${\bar{\nu }}_p(E)=\left\{\begin{array}{cc}{p}_0(E)& (E<E_0),\\ p_2+(a_2-p_2)(1-exp(-p_{1}E))& (E\ge E_0),\end{array}\right.$(6) ). In JENDL/PD-2004, ${\bar{\nu }}_p$ increases linearly below 26.2 MeV, although above the energy it is assumed to have a constant value (${\bar{\nu }}_p=5.26$). The data of Silano et al. [97] show the sharp decrease below 5.2 MeV. This may be attributed to small photo-fission cross sections.

Figure 8. Comparison of the average number of prompt neutrons per fission for ²³⁸U in JENDL/PD-2016 (solid line) with JENDL/PD-2004 (dashed line), a least-square fit by Caldwell et al. [51] (dotted line) and measured data [51,96,97].

Figure 9 compares the photo-neutron yield cross section of ²³⁵U in JENDL/PD-2016 with JENDL/PD-2004 and measured data of Caldwell et al. [98]. For fissile nuclides, the photo-neutron yield cross section stands for the total neutron production cross sections supplemented with the neutrons produced by fission reaction. The cross section of JENDL/PD-2016 is consistent with the measured data [98]. In contrast, JENDL/PD-2004 is larger than measured data especially around 14 MeV, around which the cross section has a peak.

Figure 9. Comparison of the photo-neutron yield cross section for ²³⁵U in JENDL/PD-2016 (solid line) with JENDL/PD-2004 (dashed line) and measured data [98].

The photo-fission cross section for ²³⁵U in JENDL/PD-2016 is depicted up to 140 MeV in Figure 10, together with JENDL/PD-2004 and measured data [98–107]. The evaluated result of JENDL/PD-2016 as well as JENDL/PD-2004 well reproduces the measured data even in the region above 40 MeV where the QD process is predominant. The difference between JENDL/PD-2016 and JENDL/PD-2004 is found below 7 MeV. The cross section of JENDL/PD-2016 is smaller than JENDL/PD-2004. However, it is still valid, compared with measured data in which relatively wide spread of cross sections can be seen around the critical energy.

Figure 10. Comparison of the photo-fission cross section for ²³⁵U in JENDL/PD-2016 (solid line) with JENDL/PD-2004 (dashed line) and measured data [98–107].

4. Summary

Photonuclear data file was developed in order to meet various and expanding needs. JENDL/PD-2016 was released in December 2017. The file consists of standard and expanded versions, having photonuclear data of 181 and 2681 nuclides, respectively. The maximum energy is fixed at 140 MeV. The evaluations were performed with the resonance analysis-like method and the nuclear reaction model codes CCONE and ALICE-F, using available measured data. The nuclear data cover almost all photo-induced reactions required for applications, that is, photo-absorption, photo-neutron (proton, deuteron, triton, ³He and $\alpha$-particle and gamma-rays) yield cross sections, photo-fission cross sections, the average number of total, prompt and delayed neutrons per fission, energy-angle distributions of emitted particles and gamma-rays, delayed neutron fission spectrum (if available from JENDL-4.0) and production cross sections of residual nuclides. The data revised from JENDL/PD-2004 result in better agreements with the measured ones. JENDL/PD-2016 is available from a website of Nuclear Data Center in Japan Atomic Energy Agency: https://wwwndc.jaea.go.jp/ftpnd/jendl/jendl-pd-2016.html.

Acknowledgement

The authors are grateful to Dr Toru Murata for his many years’ contributions to photonuclear data evaluation of light nuclides.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. All information on the measurements by Zatsepina et al. [94] and Lepestkin et al. [95] relies on EXFOR [59], because the details of their measurements and analyses are not available.