PHYSICAL REVIEW B 101, 026403 (2020) Comment on “Spin-dependent electron transmission model for chiral molecules in mesoscopic devices” Ron Naaman 1 , * and David H. Waldeck 2 , † 1 Department of Chemical and Biological Physics, Weizmann Institute of Science, Rehovot 76100 Israel 2 Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, USA (Received 12 July 2019; revised manuscript received 11 August 2019; published 17 January 2020) In a model presented in Yang et al. [Phys. Rev. B 99, 024418 (2019)] it was stated that in the linear regime spin selectivity cannot be measured when using two contacts configuration, and that the results observed when studying the chiral-induced spin-selectivity effect are not consistent with this statement. Here we claim that the results cited clearly indicate nonlinearity and therefore the spin selectivity observed is valid observation. DOI: 10.1103/PhysRevB.101.026403 The paper published by Yang et al. [1] models spin transmission through chiral molecules in mesoscopic devices. Based on their model, they claim that spin selectivity in elec- tron transport through chiral molecules, in the linear regime, cannot be measured by using a two-terminal device, unless a spin-flip process occurs in the molecule. Their simplified, two-terminal model assumes that charge is injected from a source electrode, transits through a chiral molecule and a ferromagnet, and is collected at a drain electrode. In this treatment, the ferromagnet transmits a given spin and reflects the other, but there is no dissipation in the ferromagnet. While the conclusions drawn by the authors may be consistent with the simplified model, the model itself is not realistic enough to account for experiments. Theoretical models for the chiral-induced spin-selectivity (CISS) effect, in two contact-spin measurements, exist in the literature already, and the conditions for observing spin polar- ization have been discussed in detail. As an example, consider the work by Matityahu et al. [2] which states: “When the helix is connected to two one-dimensional single-mode leads, time- reversal symmetry prevents spin polarization of the outgoing electrons. One possible way to retrieve such a polarization is to allow leakage of electrons from the helix to the environ- ment, via additional outgoing leads.” In other words, dephas- ing acts to create asymmetry in the transmission amplitude for spin up versus spin down, and it breaks Onsager’s reciprocity relation. For example, Sánchez and Büttiker [3] showed how asymmetry arises for magnetoconductance in a two-terminal device. The combination of interactions with a bath and the large electric fields at interfaces (typical of CISS experiments) can result in the observed asymmetry. In addition, we note that spin-selective backscattering, as an explanation for the spin selectivity, was also discussed previously [4] and even used to analyze for the extent of spin flipping in experiments [5]. To summarize, two-terminal models have been discussed before, * ron.naaman@weizmann.ac.il † dave@pitt.edu and it was shown that CISS can be observed if dissipation or a combination of nonlinearity and dissipation are included. The origin of the nonlinearity, to which we refer, is im- portant to clarify. The simplified model used by Yang et al. presents the linear approximation for the conduction, but it does not relate to the actual parameters characterizing the CISS measurements and could prove misleading to some readers. For charge moving through a system which is smaller in dimension than the screening length, the transport does not depend linearly on the field applied [6,7]. Because the chiral molecules studied in all the works cited in Ref. [1] are on the scale of a few nanometers, upon applying an electric potential the typical field is on the order of 10 8 V/m. Consequently, the electronic states in the molecules “mix,” and the electric field has two contributions: mixing of zeroth-order states by the Stark effect and driving current via the potential drop, conduction. For an example of a model-based treatment, see the recent work by Michaeli [8]. The nonlinearity of the conduction is readily apparent in experiments with two contacts that have already been published. For example, Fig. 1(a) presents the current versus potential curves that were measured in a magnetic conducting probe atomic force microscope (AFM) configuration, and Fig. 1(b) shows the corresponding plot of the conductance versus the applied potential. Figure 1(c) shows the spin po- larization as a function of the applied potential, which is extracted from the measurements shown in Fig. 1(a). Note that these data are obtained from “two contact experiments” that have been presented in figures 2 and 3 of a paper [9], referred to as Ref. [6] by Yang et al. [1]. The nonlinear response is apparent both in the current dependence on the voltage [Fig. 1(a)], as well as in the other curves. To illustrate the nonlinearity more clearly, Fig. 1(d) shows a plot of the data from Fig. 1(a) on a semilog graph. This plot reveals the exponential growth of the current at low voltage and the deviation of the currents from each other at higher voltages. The spin polarization changes dramatically at low potentials; it is basically zero at very low fields and increases as the electric field approaches a maximum of ∼5 × 10 8 V/m. This observation, which is apparent in most current vs voltage 2469-9950/2020/101(2)/026403(2) 026403-1 ©2020 American Physical Society