3.1.2. [(rac)-2]ClO4 and [(R)-2]ClO4

Electrocrystallization of donor **2** provided crystalline materials only for the racemic and (*R*) enantiomer, with two different phases of 1:1 and 2:1 stoichiometry for the latter. However, the crystals of the [(*R*)-**2**]2ClO4 phase were not of sufficient quality to allow an accurate description of the structure, yet the packing of the donors is similar to the one in [(*R*)-**1**]2ClO4 (*vide supra*). [(*rac*)-**2**]ClO4 crystallized in the triclinic system, *P*–1 space group, with one independent donor and one anion in the asymmetric unit (Figure 4a), while [(*R*)-**2**]ClO4, which is isostructural to the racemic counterpart, crystallized in the non-centrosymmetric space group *P*1 of the triclinic system, with two donors and two anions in the asymmetric unit (Figure S3).

**Figure 4.** (**a**) Molecular structure of [(*rac*)-**2**]ClO4 along with the atom numbering scheme. Only the (*R*) enantiomer of the donor is shown; (**b**) Packing diagram of [(*rac*)-**2**]ClO4, with an emphasis on the intradimer S···S short contacts (3.37 and 3.47 Å, blue dotted lines).

According to the 1:1 stoichiometry the donors should bear a +1 charge, in agreement with the central C=C and internal C–S bond lengths (Table S1). In both structures the ethyl substituents are located in the equatorial position, although the donors arrange in head-to-tail dimers as in [(*rac*)-**1**]ClO4 where the methyl substituent is axial, with very short intradimer S···S distances of 3.37–3.47 Å for [(*rac*)-**2**]ClO4 (Figure 4b) and 3.35–3.37 Å for [(*R*)-**2**]ClO4 (Figure S4). This strong dimerization should lead, very likely, to rather poor electron transport properties (*vide infra*). As in the case of donor **1**, a complex set of hydrogen bond interactions between the oxygen atoms of the anion and the hydrogen atoms of the donors can be disclosed (Figure S5 for [(*rac*)-**2**]ClO4 and Figure S6 for [(*R*)-**2**]ClO4), emphasizing once again the template role of the anion in the structural disposition of the donors.

#### *3.2. Single Crystal Conductivity Measurements*

In the (**1**)2ClO4 series only the crystals of the enantiopure phases, namely [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4, were of suitable dimensions for two- and four-contact single-crystal resistivity measurements, although great care had to be taken because of the fragility of the crystalline plates. The measured room temperature conductivity values range between 5–10 S cm−<sup>1</sup> depending on the quality of the sample and of the contacts. However, the temperature dependence of the resistivity suggests metal-like conductivity in the high-temperature range (partially masked in the two points measurements), followed upon cooling by a localized regime with a very low activation energy of 29–47 meV (340–540 K) (Figure 5a). The thermoelectric power measurements are also indicative of metallic behavior in the high-temperature regime (Figure 5b) when considering the very small positive values decreasing towards zero upon cooling.

**Figure 5.** (**a**) Temperature dependence of the electrical resistivity ρ for single crystals of [(*S*)-**1**]2ClO4 (blue curves) and [(*R*)-**1**]2ClO4 (green curves) measured using either four in-line contacts (two lower curves, full symbols) or two contacts (two upper curves, empty symbols). The red lines are the fit to the activation law ρ = ρ<sup>0</sup> exp(Ea/T) in the 80–130 K temperature range. The inset shows the same resistivity data on a linear scale in order to emphasize the high temperature conducting regime in the three lower curves. (**b**) Temperature dependence of the thermoelectric power for a single crystal of [(*S*)-**1**]2ClO4 (blue curve) and [(*R*)-**1**]2ClO4 (green curve).

The 2:1 stoichiometry and β-type packing of the donors in [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4 are clearly in favor of high electrical conductivity in the organic layer, as suggested by the single crystal electrical resistivity measurements. The situation is drastically different in the case of (**2**)ClO4 salts. Suitable crystals for temperature-dependent resistivity measurements could be obtained only for [(*rac*)-**2**]ClO4. As expected when considering the strong dimerization of the donors in the crystal structure, the room temperature value of the resistivity of 10<sup>7</sup> Ω cm and activation energy Ea = 5100 K clearly indicate a very poor semiconducting, almost insulating, behavior of this material (Figure 6 and Figure S7). Moreover, a value of <sup>ρ</sup> = 2 <sup>×</sup> 108 <sup>Ω</sup> cm room temperature resistivity could be measured on small crystals of [(*R)*-**2**]ClO4, which are isostructural with the racemic form.

**Figure 6.** Temperature dependence of the electrical resistivity, ρ, plotted as log ρ versus 1000/T for a single crystal of [(*rac*)-**2**]ClO4 measured using two contacts. The red line is the linear fit to the data giving the activation energy.

Since the two racemic salts [(*rac*)-**1**]ClO4 and [(*rac*)-**2**]ClO4 are structurally very similar it can be inferred that the former is also a poor semiconductor.

#### *3.3. Band Structure Calculations*

### 3.3.1. [(*rac*)-**1**]ClO4 and [(*rac*)-**2**]ClO4

Although the samples of [(*rac*)-**1**]ClO4 did not allow single-crystal resistivity measurements, unlike those of [(*rac*)-**2**]ClO4, we performed band structure calculations on both racemic salts for comparison purpose. The donor lattice of the [(*rac*)-**1**]ClO4 salt is shown in Figure 7a, highlighting the presence of dimers. Although there are quite short contacts associated with all interactions (Table 4), the very good σ-type overlap associated with the interaction I should make this interaction largely dominant. Because of the head-to-tail overlap mode the S···S contacts along *b*- and *c*- are long. However, note that whereas the methyl substituents practically cut any interaction along *c*, the donors face each other through the non-substituted side of the molecule along *b*, providing a better situation for possible inter-dimer interaction. The S···S contacts shorter than 3.9 Å as well as the associated βHOMO-HOMO values are reported in Table 4. Interaction I is almost two orders of magnitude larger than the other interactions, so that this salt must be considered as made of chains, along the *a*-direction, of very stable (Me-EDT-TTF2) <sup>2</sup><sup>+</sup> dimers.

**Figure 7.** (**a**) Donor lattice of the [(*rac*)-**1**]ClO4 salt where the chains and the main intermolecular interactions are labeled; (**b**) Extended Hückel band structure for the donor lattice of [(*rac*)-**1**]ClO4. The dashed line refers to the highest occupied level and Γ = (0, 0, 0), X = (*a*\*/2, 0, 0), Y = (0, *b*\*/2, 0), M = (*a*\*/2, *b*\*/2, 0) and Z = (0, 0, *c*\*/2).

**Table 4.** Intermolecular S···S contacts shorter than 3.9 Å and absolute values of the βHOMO-HOMO interaction energies (eV) [[29]] for the different donor···donor interactions in the [(*rac*)-**1**]ClO4 salt.


The calculated band structure is shown in Figure 7b where an indirect bandgap (from Z to X) of 0.67 eV clearly separates the two HOMO bands (i.e., only the lower band, built from the bonding combination of the two HOMOs, Ψb, is filled). Note that whereas the dispersion of the lower band is weak, that of the upper one is quite considerable and of the same order along the directions *a* and *b*. Thus the effective mass of the electron carriers should be considerably larger than that of the hole carriers and may confer a considerably 2D character to the activated conductivity. The larger dispersion associated with the upper band is a consequence of the antibonding nature of the upper Ψab orbital of the dimer for two reasons: (i) because of the inclination of the molecules with respect to the *a* axis the inter-dimer overlap along this direction is better, and (ii) because of the intra-dimer antibonding character the Ψab orbital somewhat hybridizes with molecular σ levels so as to shift the wave function

towards the outer region of the dimer, thus slightly decreasing the antibonding nature of Ψab and somewhat favoring the inter-dimer overlap along *b*.

The calculated band structure for [(*rac*)-**2**]ClO4 (see Figure S8) leads to an equivalent description and an indirect bandgap of 0.8 eV from Γ to X, in excellent agreement with our conductivity measurements.

### 3.3.2. [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4

The donor layers of the [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4 salts contain two different donors and six different intermolecular interactions (see Figure 8). The layer can be described as a series of parallel chains of AB dimers (interaction I) running along the (*a* + *b*)-direction. Every donor is implicated in two different interactions along the chain and four inter-chain interactions. To understand the electronic structure and transport properties of these salts we need to have a hint on the strength of the HOMO···HOMO interactions which may be done by looking at the absolute value of the so-called HOMO···HOMO interaction energy, βHOMO-HOMO, [29] associated with each interaction. The calculated values for the six interactions in the two different layers of the two pure enantiomeric salts are reported in Table 5.

**Figure 8.** (**a**) Donor layer of the [(*R*)-**1**]2ClO4 salt where the different donors and intermolecular interactions are labeled; (**b**) Top view of the intra-dimer interaction I; (**c**) Top view of the inter-dimer interaction II.

**Table 5.** Absolute values of the βHOMO-HOMO interaction energies (eV) for the different donor···donor interactions in the [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4 salts.


The results for both enantiomers are practically identical. The more salient observation is that the interactions along the chain (I and II) are strong whereas the inter-chain interactions (III to VI) are practically one order of magnitude weaker. This should confer a clear one-dimensional (1D) character to the system. Note that the two intra-chain interactions, despite originating from different overlap modes (Figure 8b,c), are associated with similar interaction energy values, showing that a better orientation of the HOMOs (electronic effect) may perfectly compensate for a larger set of short contacts (metric effect). Because of this fact, the chains along the diagonal (*a* + *b*)-direction are quite uniform as far as the HOMO···HOMO interactions are concerned. Note that the lateral interactions although weaker are by no means negligible so that they should provide a substantial coupling between these quite uniform chains.

The calculated band structure and Fermi surface of the [(*R*)-**1**]2ClO4 salt are shown in Figure 9 (those for [(*S*)-**1**]2ClO4 are practically identical as shown in Figure S9). The diagram in Figure 9a contains four HOMO bands because there are four donors per repeat unit. Because of the stoichiometry, there are two holes in these bands so that two partially filled and very dispersive bands are generated, in agreement with the metallic character of these salts. The calculated Fermi surface may appear complex at a first sight although really it is not. The repeat unit of the layer contains two identical AB pairs so that the calculation could have been performed using a centered rectangular lattice with just two molecules in the unit cell. In that case, the calculated Fermi surface would simply be the red part of Figure 9b (and the area of the Brillouin zone would be twice larger). Using the crystallographic repeat unit, which is twice larger, two identical Fermi surfaces are folded into the new, smaller Brillouin zone shown in Figure 9b. The warped red lines of Figure 9b are perpendicular to the (*a* + *b*)-direction as it must be if the system is built from chains of HOMOs along the (*a* + *b*)-direction. The warping of the red lines is substantial because of the above noted lateral interactions; in fact the two lines almost touch at the S point so that it is possible that small structural changes brought about by thermal contraction or pressure could maybe close the lines and lead to a pseudo-elliptic closed Fermi surface (i.e., a 2D metal). Note that the red Fermi surface is almost identical to those reported for the (DM-EDT-TTF)XF6 (X = P, As, Sb) salts [5,6], where the donor possesses two stereogenic centers, as well as for (**1**)2PF6 [4], in which, as in the present salts, the donor has a single stereogenic center, thus denoting a very similar organization of the layers in all these salts. We direct the reader to these references for further discussion of the electronic structure of these layers. However, we note that even if some parts of the red Fermi surface of Figure 9b are nested, such nesting is incomplete. Consequently, if metal to insulator low-temperature transitions occur in these salts, as for some of the above-mentioned ones, it is not expected that they may originate from the charge or spin density wave instabilities so frequent in low dimensional systems but from electronic or structural localization.

**Figure 9.** (**a**) Extended Hückel band structure and (**b**) Fermi surface for the donor layers of [(*R*)-**1**]2ClO4. The dashed line refers to the calculated Fermi level and Γ = (0, 0), X = (*a*\*/2, 0), Y = (0, *b*\*/2), M = (*a*\*/2, *b*\*/2) and S = (*a*\*/2, −*b*\*/2).

#### **4. Conclusions**

Electrocrystallization of the chiral EDT-TTF derivatives **1** and **2** provided a complete series of radical cation salts for the former and racemic and enantiopure (*R*) salts for the latter, with the same perchlorate anion. Interestingly, unlike the series with the PF6 − anion previously described [4], where the racemic and enantiopure salts were isostructural, here, the perchlorate ion discriminates between the racemic and enantiopure forms (Table 6).


**Table 6.** Comparison between the radical cation salts of donors **1** and **2** with the ClO4 – and PF6 – anions.

In the [(*rac*)-**1**]ClO4 salt, which should be a poor semiconductor with an indirect gap, the donors are fully oxidized and arrange in centrosymmetric dimers. In sharp contrast, the donors form a β-type packing in the mixed-valence enantiopure salts [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4 salts, which show metal-like behavior in the high-temperature regime, in agreement with extended Hückel band structure calculations. From a structural, electron transport properties and electronic structure point of view, these enantiopure salts resemble the (**1**)2PF6 and (DM-EDT-TTF)2XF6 (X = As, Sb) series, but not to the enantiopure (DM-EDT-TTF)2ClO4 compounds [17], pointing out that the variation of the number of stereogenic centers and the use of racemic or enantiopure forms in combination with various anions is a simple mean of reaching a large collection of chiral molecular conductors with original electronic structures. Indeed, both donors **1** and DM-EDT-TTF provided mixed-valence salts of 2:1 stoichiometry with the ClO4 – anion, showing metal-like conductivity, yet their crystal structures are drastically different, i.e., monoclinic space group for [(*R*)-**1**]2ClO4 and [(*S*)-**1**]2ClO4 and enantiomorphic hexagonal space groups for [(*R*,*R*)-DM-EDT-TTF]2ClO4 and [(*S*,*S*)-DM-EDT-TTF]2ClO4 [17]. This is a clear consequence of the different number of stereogenic centers and symmetry of the donor, impacting the intermolecular hydrogen bonding interactions with the anion and the overlap interactions between the donors. Further work in these families of materials will be devoted to conductivity measurements under pressure and under a magnetic field, and to the use of larger octahedral or tetrahedral anions.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4352/10/11/1069/s1, Figure S1: Packing diagram of [(*rac*)-**1**]ClO4 in the *ac* plane with an emphasis on the short lateral S···S contacts, Figure S2: Solid state structure of [(*rac*)-**1**]ClO4 with an emphasis on the C–H···O short contacts, Figure S3: Molecular structure of [(*R*)-**2**]ClO4 along with the atom numbering scheme, Figure S4: Packing diagram of [(*R*)-**2**]ClO4, with an emphasis on the intradimer S···S short contacts, Figure S5: Solid state structure of [(*rac*)-**2**]ClO4 with an emphasis on the C–H···O short contacts, Figure S6: Solid state structure of [(*R*)-**2**]ClO4 with an emphasis on the C–H···O short contacts, Figure S7: Temperature dependence of the electrical resistivity ρ for a single crystal of [(*rac*)-**2**]ClO4 measured using two contacts, Figure S8: Extended Hückel band structure for the donor lattice of [(*rac*)-**2**]ClO4, Figure S9: Extended Hückel band structure (**a**) and Fermi surface (**b**) for the donor layers of [(*S*)-**1**]2ClO4, Table S1: Selected C=C and C–S bond lengths for [(*rac*)-**2**]ClO4 and [(*R*)-**2**]ClO4.

**Author Contributions:** N.A. conceived and designed the experiments; N.M. synthesized and characterized the materials; N.V. performed the chiral HPLC separation of the precursors; P.A.-S., E.B.L. and M.A. investigated the electron transport properties; E.C. undertook the theoretical study; N.A. and E.C. wrote and/or reviewed the manuscript with contributions from all authors. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was partially funded in France by the National Agency for Research (ANR), Project 15-CE29-0006-01 ChiraMolCo, in Spain by the Spanish MICIU through Grant PGC2018-096955-B-C44 and the Severo Ochoa FUNFUTURE (CEX2019-000917-S) Excellence Center distinction, as well as by Generalitat de Catalunya (2017SGR1506) and in Portugal by FCT under contracts UIDB/04349/2020 and LISBOA-01-0145-FEDER-029666.

**Acknowledgments:** This work was supported in France by the CNRS and the University of Angers. The collaboration between the Portuguese and French team members was also supported by a FCT–French Ministry of Foreign Affairs bilateral action FCT/PHC-PESSOA 2020-21 (Project 44647UB).

**Conflicts of Interest:** The authors declare no conflict of interest.
