Cross sections for the electron-impact excitations ${\tilde{A}}^1{B}_1$ and ${\tilde{B}}^1{A}_1$ of ${\mathrm{H}}_2\mathrm{O}$ determined by high-energy electron scattering

Understanding the role of inelastic electron scattering in water is of fundamental importance in various fields ranging from atmospheric chemistry to radiation biology. The lack of accurate excitation cross sections for water results in a large uncertainty, for example, in the track structure simulation for modeling radiation damage to DNA. The large differences of the integral cross sections (ICSs) for the optical-allowed excitations ${\tilde{A}}^1{B}_1$ and ${\tilde{B}}^1{A}_1$ of ${\mathrm{H}}_2\mathrm{O}$ among experiments and theories have been maintaining for decades. To resolve this issue, by combining the ab initio calculation with the electron correlation being taken into account accurately, we compare the ICSs determined by high-energy electron scattering with the existing experiments at low and intermediate energies and theories, where the present experiment eliminated the errors from the spectral deconvolution in low-energy measurements. Our work provides a recommendation for the ICSs of the two excitations from thresholds up to several keV, and suggests the existing selected ICSs for ${\mathrm{H}}_2\mathrm{O}$ be revised for accurate modeling radiation effects in biological matter and describing the transport properties of electrons in aqueous systems.

Figure 1

A typical EELS of ${\mathrm{H}}_2\mathrm{O}$ at an incident electron energy of 1500 eV and a scattering angle of $2.0^{\circ }$. Solid lines are the fitted curves (see text for details).

Figure 2

The GOSs for (a) ${\tilde{A}}^1{B}_1$ and (b) ${\tilde{B}}^1{A}_1$ of ${\mathrm{H}}_2\mathrm{O}$ compared with different modeling calculations (see text for details).

Figure 3

The ICSs for (a) ${\tilde{A}}^1{B}_1$ and (b) ${\tilde{B}}^1{A}_1$ of ${\mathrm{H}}_2\mathrm{O}$ compared with the present calculations (see text for details).

Figure 4

The GOSs for (a) ${\tilde{A}}^1{B}_1$ and (b) ${\tilde{B}}^1{A}_1$, in comparison to different experimental results and calculations (see text for details).

Figure 5

The OOSs for (a) ${\tilde{A}}^1{B}_1$ and (b) ${\tilde{B}}^1{A}_1$, along with previous measurements and calculations. Experiments: 1 present, 2 Thorn et al. [23], 3 Lassettre et al. [25], 4 Chan et al. [56], 5 Yoshino et al. [68], 6 Lee and Suto [69], 7 Laufer and McNesby [70], 8 Harrison et al. [65], 9 Watanabe and Zelikoff [71]. Theories: 10 present, 11 Durante et al. [32], 12 Bhanuprakash et al. [37], 13 Phillips and Buenker [66], 14 Theodorakopoulos et al. [72], 15 Buenker and Peyerimhoff [75], 16 Nicolaides et al. [76], 17 Wood [77], 18 Williams and Langhoff [74], 19 Rauk and Barriel [73], and 20 Yeager et al. [67]. Shadow areas show error bar ranges of the present two OOSs.

Figure 6

The ICSs for (a, b) ${\tilde{A}}^1{B}_1$ and (c, d) ${\tilde{B}}^1{A}_1$, along with previous measurements and calculations. Experiments [(a) and (c)]: red solid line present (BE-scaled ICSs), black dashed line Matsui et al. ($\mathrm{BE}f$-scaled ICSs) [18], filled inverted triangle Lassettre et al. [25], filled square Sophia University [23], open circle Matsui et al. [18], filled rhombus the combined data from Sophia and Flinders Universities [23], filled triangle Flinders University [23], open rhombus Ralphs et al. [19]. Theories [(b) and (d)]: green solid line present (MRD-CI), orange dashed line Durante et al. (RPA) [32], black solid line Kutcher et al. (ELF) [9, 78], blue solid line Ralphs et al. (SMC) [19], magenta solid line Rescigno et al. (complex Kohn) [29], olive solid line Gil et al. (complex Kohn) [34], violet solid line Gorfinkiel et al. ($R$-matrix, ANA) [30], violet dotted line Gorfinkiel et al. ($R$-matrix, FNA) [30], and dark yellow solid line Morgan et al. ($R$-matrix) [31].