1 Erratum to: Appl Phys B (2012) 107:221–228 DOI 10.1007/s00340-012-4896-x

In the continuation of our research [1], we discovered an error in the implementation of the Quantum Sutton-Chen (Q-SC) [25] potential used to define the nickel surface. This error overestimated the many-body attractive force between the nickel atoms, causing the nickel to remain in solid phase when it should have been liquid. When coupled with the ab initio-derived Morse pairwise potential between the argon and nickel atoms, the corrected Q-SC model produces a larger thermal accommodation coefficient compared to the value in ref. [1]. The corrected values are summarized in the table below (Table 1).

Table 1 Revised accommodation coefficients for Ni/Ar system

Figure 1 shows the probability that a gas molecule hits the surface a specified number of times during a scattering event. The ab initio-derived Morse potential predicts that the gas molecule will scatter almost directly, while the erroneously deep LJ 6–12 potential predicted by the Lorentz Berthelot rules results in a more prolonged interaction.

Fig. 1
figure 1

Probability of an argon molecule undergoing a specified number of hits before scattering from a nickel surface

Figure 2 shows that the probability densities of scattered trajectories found using kernel density estimation. The distributions for the normal and tangential scattered velocities corresponding to the ab initio-derived potential are closer together compared to the curves shown in [1], since the liquid state of the nickel droplet results in a larger α t . Two-temperature Maxwell–Boltzmann distributions are in good agreement with the corresponding probability densities, but for clarity, the corresponding Maxwell–Boltzmann distributions are only shown for 1,250 and 2,500 K, representative of the partial and complete accommodation scenarios predicted by the ab initio- and Lorentz–Berthelot-derived potentials, respectively.

Fig. 2
figure 2

Scattered molecular speeds obtained from molecular dynamics

The main conclusion of the original paper that the Lorentz–Berthelot combining rules produce a gas/surface potential that overestimates the accommodation coefficient is unaffected by this error.