Rectangular Transition Metal-rTCNQ Organic Frameworks Enabling Polysulfide Anchoring and Fast Electrocatalytic Activity in Li-Sulfur Batteries: A Density Functional Theory Perspective

Two-dimensional metal-organic frameworks (MOFs) have shown great development po-tential in the field of lithium-sulfur (Li-S) batteries. In this theoretical research work, we propose a novel 3d transition metals (TM)-embedded rectangular tetracyanoquinodimethane (TM-rTCNQ) as a potential high-performance sulfur host. The calculated results show that all TM-rTCNQ structures have excellent structural stability and metallic properties. Through exploring different adsorption patterns, we discovered that TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) monolayers possess moderate adsorption strength for all polysulfide species, which is mainly due to the existence of the TM-N4 active center in these frame systems. Especially for the non-synthesized V-rCTNQ, the theoretical calculation fully predicts that the material has the most suitable adsorption strength for polysul-fides, excellent charging-discharging reaction and Li-ion diffusion performance. Additionally, Mn-rTCNQ, which has been synthesized experimentally, is also suitable for further experimental con-firmation. These findings not only provide novel MOFs for promoting the commercialization of Li-S batteries, but also provide unique insights for fully understanding their catalytic reaction mecha-nism.


Structure, Stability and Electronic Properties of TM-rTCNQ and S 8 /LiPSs
The schematic diagram of the 2 × 1 × 1 supercell of all 3d family TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu and Zn) structures is shown in Figure 1a, and black/red rectangular box areas represent the TM-rTCNQ unit cell and TCNQ molecule, respect-tively. Among them, each 3d TM atom is combined with four N atoms in TCNQ clusters. The detailed structural information is listed in Table S1. As observed in Figure S1, all at-oms lie in the same plane of TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co and Zn) mon-olayers. Some structural distortion of TM-N 4 center is observed for Ni-rTCNQ and Cu-rTCNQ, which is also consistent with related research works [51]. In addition, all the S 8 /Li 2 S x (x = 1, 2, 4, 6, 8) species are fully optimized ( Figure 1b) and are highly consistent with the previously reported theoretical calculations [16,55,56].
Additionally, the stability of these catalysts is evaluated by the energy difference (E di f f ) between adsorption energies (E a ) and the average cohesive energies (E coh ). Generally, systems with E di f f = E a − E coh < 0 represent that TM atoms can be atomically dis-persed on the TM-rTCNQ monolayers through the coordination effect. As shown in Fig-ure 1c, the E a (TM) of TM-rTCNQ is within the range of −4.19-−9.95 eV, which indicates the strong interaction between TM atoms and the N 4 active center. The negative values of E di f f reflect that all TM-rTCNQ configurations have sufficient structural stability to meet the possibility of experimental preparation. In addition, excellent electrical conductivity of the sulfur host materials is an important factor for practical Li-S battery applications. We focused on the conductivity of TM-rTCNQ and calculated the density of states (DOS) of all TM-rTCNQ unit cells, as shown in Figure 1d and Figure S2. The TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni and Cu) structures, except for Zn-rTCNQ, all exhibit metallic prop-erties and are expected to improve the electrochemical performance of Li-S cells. ure 1c, the Ea(TM) of TM-rTCNQ is within the range of −4.19-−9.95 eV, which indic the strong interaction between TM atoms and the N4 active center. The negative valu reflect that all TM-rTCNQ configurations have sufficient structural stability to m the possibility of experimental preparation. In addition, excellent electrical conduct of the sulfur host materials is an important factor for practical Li-S battery applicati We focused on the conductivity of TM-rTCNQ and calculated the density of states (D of all TM-rTCNQ unit cells, as shown in Figure 1d and S2. The TM-rTCNQ (TM = Sc V, Cr, Mn, Fe, Co, Ni and Cu) structures, except for Zn-rTCNQ, all exhibit metallic p erties and are expected to improve the electrochemical performance of Li-S cells.

Adsorption Performance of TM-rTCNQ for S8/LiPSs
All well known, 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME) electro molecules can yield a competitive trapping to soluble Li2S4, Li2S6 and Li2S8. Therefore adsorption energies of Li2S4 (Li2S6, Li2S8) anchored on TM-rTCNQ substrates were calculated. As displayed in the Figure S3, the energy range is from −0.88-−0.77 eV mainly depends on Li-O bond interactions, which is highly in accordance with prev works [3]. An optimal host material should trap the soluble LiPSs more strongly than DOL/DME solvent molecules to restrain the shuttle effect.
Herein, the adsorption energies of all TM-rTCNQ for polysulfides were calcul and shown in Figures 2b, S4 and S5. We considered two adsorption patterns (Z1 and As shown in Figure 2a, Z1 mode involves trapping the LiPSs in the center of TM-N4 c

Adsorption Performance of TM-rTCNQ for S 8 /LiPSs
All well known, 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME) electrolyte molecules can yield a competitive trapping to soluble Li 2 S 4 , Li 2 S 6 and Li 2 S 8 . Therefore, the adsorption energies of Li 2 S 4 (Li 2 S 6 , Li 2 S 8 ) anchored on TM-rTCNQ substrates were first calculated. As displayed in the Figure S3, the energy range is from −0.88-−0.77 eV and mainly depends on Li-O bond interactions, which is highly in accordance with previous works [3]. An optimal host material should trap the soluble LiPSs more strongly than the DOL/DME solvent molecules to restrain the shuttle effect.
Herein, the adsorption energies of all TM-rTCNQ for polysulfides were calculated and shown in Figure 2b, Figures S4 and S5. We considered two adsorption patterns (Z1 and Z2). As shown in Figure 2a, Z1 mode involves trapping the LiPSs in the center of TM-N 4 coor-dination network, and Z2 mode refers to adsorbing the LiPSs with two TM atoms of the TM-rTCNQ supercell. It can be seen that Sc-rTCNQ and Ti-rTCNQ exhibit significant binding and structural distortion for S 8 and soluble LiPSs through Z2 mode due to the strong interaction between two TMs and S atoms. Moreover, our results also show obvi-ous structural changes for the Ni-rTCNQ/Li 2 S(Z1) and Cu-rTCNQ/Li 2 S(Z1) adsorption systems. Therefore, TM-rTCNQ (TM = Sc, Ti, Ni and Cu) monolayers are ruled out since the structural integrity of the host is essential to the subsequent reactions. For the remain-ing TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) structures, the adsorption energies of the most stable configurations for S 8 /LiPSs on TM-rTCNQ are summarized in the Figure 2b. Obviously, these structures can effectively inhibit the shuttle effect problem and V-rTCNQ possesses the most outstanding adsorption performance shown in Figure 2c.
systems. Therefore, TM-rTCNQ (TM = Sc, Ti, Ni and Cu) monolayers are ruled out since the structural integrity of the host is essential to the subsequent reactions. For the remain ing TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) structures, the adsorption energies of the most stable configurations for S8/LiPSs on TM-rTCNQ are summarized in the Figure 2b Obviously, these structures can effectively inhibit the shuttle effect problem and V rTCNQ possesses the most outstanding adsorption performance shown in Figure 2c.
Finally, the ratio of van der Waals (RvdW) effect is shown in Figure 2d to further ana lyze the physical or chemical interactions between TM-rTCNQ and polysulfides. Accord ing to Figure 2d, one can see that the effect of chemical binding is relatively increased along with the progression of the lithiation process. For example, the weight of vdW in teraction contributing to the adsorption energy is 83.33% for the Mn-rTCNQ/S8 system whereas for Mn-rTCNQ/Li2S, this ratio decreases to 6.84%. In addition, the vdW chemica interactions play a dominant role in the early transition metal decorated V-rTCNQ, with RvdW in the range of 7.06-48.18%, leading to an impressive anchoring effect on S8/LiPSs Among them, the RvdW of V-rTCNQ/Li2S6 and V-rTCNQ/Li2S8 adsorption systems is lower than that of V-rTCNQ/Li2S4, which is significantly related to the extremely strong adsorp tion energy of V-rTCNQ for Li2S6 and Li2S8 clusters.  Finally, the ratio of van der Waals (R vdW ) effect is shown in Figure 2d to further analyze the physical or chemical interactions between TM-rTCNQ and polysulfides. According to Figure 2d, one can see that the effect of chemical binding is relatively increased along with the progression of the lithiation process. For example, the weight of vdW in-teraction contributing to the adsorption energy is 83.33% for the Mn-rTCNQ/S 8 system, whereas for Mn-rTCNQ/Li 2 S, this ratio decreases to 6.84%. In addition, the vdW chemical interactions play a dominant role in the early transition metal decorated V-rTCNQ, with R vdW in the range of 7.06-48.18%, leading to an impressive anchoring effect on S 8 /LiPSs. Among them, the R vdW of V-rTCNQ/Li 2 S 6 and V-rTCNQ/Li 2 S 8 adsorption systems is lower than that of V-rTCNQ/Li 2 S 4 , which is significantly related to the extremely strong adsorp-tion energy of V-rTCNQ for Li 2 S 6 and Li 2 S 8 clusters.

Electronic Structure Analysis
In order to further reveal the interactions of TM-rTCNQ configurations anchoring S 8 /LiPSs, the charge density differences (∆ρ) and DOS were calculated. The ∆ρ and charge transfer amount of all adsorption systems of TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) for S 8  bond with the adjacent S atoms of S 8 /LiPSs, as shown in Figure 3a. Moreover, Li 2 S, Li 2 S 2 and Li 2 S 4 donate electrons to the V-rTCNQ host, whereas the high order Li 2 S 6 , Li 2 S 8 and S 8 act as acceptors with the increase in S content.
bonds with the N atoms on the V-rTCNQ surface, whereas the V atoms mainly bond wit the adjacent S atoms of S8/LiPSs, as shown in Figure 3a. Moreover, Li2S, Li2S2 and Li2S donate electrons to the V-rTCNQ host, whereas the high order Li2S6, Li2S8 and S8 act a acceptors with the increase in S content.
Additionally, taking V-rTCNQ and Mn-rTCNQ as examples, the DOS of the adsorp tion systems were calculated and are plotted in Figure 3b. The metallic features of V rTCNQ and Mn-rTCNQ are well maintained during the whole process, ensuring fast elec tron transportation during the redox process of LiPSs. In addition, the tiny contributio of S8 to the electronic states near the Fermi level is observed, manifesting the physica adsorption features. As Li ratio increases in the LiPSs, the s orbitals of lithium and p orbit als of N atoms have significant hybridization, verifying the formation of Li-N chem-ica bonds.

Charging-Discharging Catalytic and Li-ion Diffusion Capability
In addition to the anchoring effect for S8/LiPSs, we also systematically investigate the catalytic performance of TM-rTCNQ for the whole charging-discharging process o LiSBs by simulation of the oxidation and reduction reactions of elemental sulfur, as we as the kinetic rate of Li-ion diffusion [2,12,57]. In this regard, the energy barriers of thre dynamic processes were calculated based on DFT methods, and the summarized result are shown in Figure 4 and S7, as well as in Table S2 and S3. The Gibbs free energy ladde diagram of the sulfur reduction reactions (SRRs) process (discharging process) in Figur 4a shows that the reaction process from S8 to *Li2S8 is basically exothermic, except for th V-rTCNQ conversion of *Li2S6 to *Li2S4, whereas the conversion process from *Li2S2 t *Li2S is all endothermic. The rate-limiting step (RLS) of the whole catalytic reaction pro cess is the *Li2S2 → *Li2S, due to the maximum Gibbs free energy change (ΔG6). As dis played in Figure 4a, ΔG6 values for TM-rTCNQ (V, Cr, Mn, Fe and Co) are 3.15 eV, 3.2 eV, 3.36 eV, 3.50 eV and 3.68 eV respectively, and V-rTCNQ exhibits the smallest ΔG (3.1 eV) for RLS.
The first-step reaction of the charging process (Li2S dissociation: *Li2S → *LiS + Li + e − ) and the transition states of the Li-ion diffusion were also calculated using the CI-NE method. The results show that V-rTCNQ and Co-rTCNQ have the lowest Li2S decompo

Charging-Discharging Catalytic and Li-ion Diffusion Capability
In addition to the anchoring effect for S 8 /LiPSs, we also systematically investigated the catalytic performance of TM-rTCNQ for the whole charging-discharging process of LiSBs by simulation of the oxidation and reduction reactions of elemental sulfur, as well as the kinetic rate of Li-ion diffusion [2,12,57]. In this regard, the energy barriers of three dynamic processes were calculated based on DFT methods, and the summarized results are shown in Figure 4 and Figure S7, as well as in Tables S2 and S3. The Gibbs free energy ladder diagram of the sulfur reduction reactions (SRRs) process (discharging process) in Figure 4a shows that the reaction process from S 8 to *Li 2 S 8 is basically exothermic, except for the V-rTCNQ conversion of *Li 2 S 6 to *Li 2 S 4 , whereas the conversion process from *Li 2 S 2 to *Li 2 S is all endothermic. The rate-limiting step (RLS) of the whole catalytic reaction process is the *Li 2 S 2 → *Li 2 S, due to the maximum Gibbs free energy change (∆G6). As displayed in Figure 4a, ∆G6 values for TM-rTCNQ (V, Cr, Mn, Fe and Co) are 3.15 eV, 3.28 eV, 3.36 eV, 3.50 eV and 3.68 eV respectively, and V-rTCNQ exhibits the smallest ∆G (3.15 eV) for RLS.
The first-step reaction of the charging process (Li 2 S dissociation: *Li 2 S → *LiS + Li + + e − ) and the transition states of the Li-ion diffusion were also calculated using the CI-NEB method. The results show that V-rTCNQ and Co-rTCNQ have the lowest Li 2 S decomposition and Li-ion diffusion barriers, respectively. The migration pathways of V-rTCNQ are plotted in the right side of Figure 4b, c (0 is the initial-state position and 4 is the final-state position), and those of other structures are shown in Figure S7. It is evi-dent that the dissociated single Li-ion moves far away from the center TM atoms to the N atoms of adjacent N 4 centers, accompanied by the fracture of the Li−S bonds. As shown in Table 1, our calculated results predict that un-synthesized V-rTCNQ and synthesized Mn-rTCNQ can be applied as sulfur hosts with excellent charging-discharging performance, and their RLS and Li 2 S dissociation energy barriers are comparable with, or even lower than, those of SACs in the recent literature. The Li-ion diffusion behaviors of TM-rTCNQ (TM = V, Mn, Fe and Co) are also comparable with some other types of 2D material (the Li diffusion barrier is 0.66 eV for β 12 -borophene [58], 0.38 eV for an InP 3 monolayer [59], 0.33 eV for graphene [60], 0.43 eV for C 3 N [61], 0.35 eV for TiPc [12], etc.).

AIMD Simulation Results
In particular, long-term stability is a basic and vital element for applica material in LiSBs. Theoretically, AIMD simulations are further performed thermal stability of TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) substrates an systems. As representatives, TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) mono adsorption systems of V-rTCNQ/Li2S4 (Li2S6, Li2S8) were chosen to implemen ulations with the NVT ensemble at 500 K for 10 ps. As displayed in Figure  geometries of the TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) monolayers are w when the temperature increases to 500 K. In addition, AIMD simulations we out to check the overall thermal stability of the long-chain LiPSs (Li2S8, Li  All in all, considering that V-rTCNQ has the most suitable adsorption strength for S 8 /LiPSs clusters among all TM-rTCNQ, the lowest charging-discharging reaction energy barrier and acceptable Li-ion diffusion, it exhibits very promising value for LiSBs. Although the binding strength of Mn-rTCNQ to polysulfides is not the strongest, the en-ergy barrier of its discharging reaction is significantly lower than that of Fe-rTCNQ and Co-rTCNQ, and charging reaction performance is also batter than that of Cr-rTCNQ and is comparable with that of Fe-rTCNQ and Co-rTCNQ. What's more, the Li-ion diffusion energy barrier of Mn-rTCNQ is even lower than that of V-rTCNQ and Cr-rTCNQ, and it is very close to that of the best Co-rTCNQ. More importantly, this material has been pre-pared in experiments, thus also shows great application potential.

AIMD Simulation Results
In particular, long-term stability is a basic and vital element for application as a host material in LiSBs. Theoretically, AIMD simulations are further performed to ensure the thermal stability of TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) substrates and adsorption systems. As representatives, TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) monolayers and the adsorption systems of V-rTCNQ/Li 2 S 4 (Li 2 S 6 , Li 2 S 8 ) were chosen to implement AIMD sim-ulations with the NVT ensemble at 500 K for 10 ps. As displayed in Figure 5a and Figure S8, the geometries of the TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) monolayers are well preserved when the temperature increases to 500 K. In addition, AIMD simulations were also carried out to check the overall thermal stability of the long-chain LiPSs (Li 2 S 8 , Li 2 S 6 and Li 2 S 4 )-terminated V-rTCNQ systems. As detailed in Figure 5b-d, the results reveal that the sys-tems with adsorbed long-chain LiPSs and their hosts have no significant bond breakage and structural distortion. Our AIMD results show that the TM-rTCNQ monolayers and the adsorption systems of V-rTCNQ for soluble Li 2 S 4 (Li 2 S 6 , Li 2 S 8 ) possess excellent ther-mal stability. Hence, V-rTCNQ can be further prepared and has great application poten-tial as a substrate that can greatly improve LiSBs performance. Last, but not least, synthe-sized Mn-rTCNQ is also extremely ready to be used in the experimental exploration of LiSBs.

Computational Details
In our work, all spin-polarized first-principles calculations were completed

Computational Details
In our work, all spin-polarized first-principles calculations were completed by the Vienna ab-initio Simulation Package (VASP) based on the density functional theory (DFT) [63]. The ion-electron interactions were described by the projector augmented wave (PAW) method [64] and the general gradient approximation (GGA) [65,66] in the Perdew-Burke-Ernzerhof (PBE) [67] form was used. The DFT-D3 [68] semi-empirical correction scheme of Gimme was adopted to treat the van der Waals (vdW) interaction between S 8 /LiPSs and substrates. S 8 and Li 2 S x (x = 1, 2, 4, 6, 8) clusters were placed in a box of 20 × 20 × 20 Å 3 to fully relax. A supercell, consisting of 2 × 1× 1 two-dimensional (2D) TM-rTCNQ monolayers, was utilized to anchor S 8 /LiPSs, and the vacuum layer was set as 20 Å to eliminate the interactions between two periodic units. The simulation was run with a cutoff energy of 520 eV throughout the computations. These settings ensure convergence of the total energies and force within 10 -5 eV and 0.02 eV/Å, respectively. The energy bar-riers of Li 2 S decomposition and Li-ion diffusion were calculated using the climbing image nudged elastic band (CI-NEB) method [69,70]. Bader-charge analysis [71] was conducted to simulate the amount of charge transfer between TM-rTCNQ and S 8 /LiPSs. Finally, ab-initio molecular dynamics (AIMD) simulations with a total time of 10 ps (5000 steps and step size set as 2 fs) were conducted to evaluate the thermal stability of TM-rTCNQ sub-strates and the adsorption systems of V-rTCNQ/Li 2 S 4 (Li 2 S 6 and Li 2 S 8 ).

Computational Methods
Structural stability is one of the important factors for theoretically predicting the fea-sibility of experimental preparation of catalytic materials. It is understood that the TM-rTCNQ formed by the combination of TM and TNCQ can be fully judged through the adsorption energy (E a ), cohesion energy (E c ) and the difference (E diff ) between them [72,73]. Their detailed definitions are as follows: where E TM−rTCNQ , E support and E single-TM are the total energies of TM-rTCNQ, the supports and isolated single transition metal (TM) atoms (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu and Zn), respectively.µ TM are the energies of TM in their stable bulk phase. When E diff (TM) < 0, TM-rTCNQ can be considered as an experimental preparation and won't easily agglomerate into clusters. In order to accurately evaluate the anchoring ability of TM-rTCNQ structures to S 8 /LiPSs, the formula of adsorption energy (E ads ) is defined as [6,22]: where E TM−rTCNQ+S 8 /LiPSs , E TM−rTCNQ and E S 8 /LiPSs are the total energies of the adsorption systems of TM-rTCNQ substrates with E S 8 /LiPSs clusters, TM-rTCNQ monolayers and isolated S 8 /LiPSs molecules, respectively. According to the above definition, more nega-tive E ads corresponds to stronger adsorption interaction. The charge density difference (∆ρ) is used to visualize the electron transfer between TM-rTCNQ and S 8 /LiPSs, and its calculation formula is as follows [12,33]: where ρ TM−rTCNQ+S 8 /LiPSs , ρ TM−rTCNQ and ρ S 8 /LiPSs are the charge density of adsorption systems, substrate materials and polysulfides, respectively. To quantitatively analyze the degree of vdW and chemical interactions to determine the nature of the substrates' adsorption of polysulfides, the contribution ratio of vdW (R vdW ) interaction is calculated by the following formula [2,33]: where E with-vdW ads and E without-vdW ads stand for the adsorption energies with or without vdW correction.
The Gibbs free energy (∆G) for above each elementary step was calculated as follows [5,12,14,62]: Herein, ∆E DFT is directly obtained by electronic energy difference of each elemen-tary reaction. ∆E ZPE and T∆S stand for zero-point energy and entropy change at 298.15 K, respectively. n and U are the number of electrons transferred and the applied voltage. For the battery systems, the ∆E ZPE and T∆S can be ignored in the calculation [11,12,62], therefore, when U = 0 V, ∆G ≈ ∆E DFT .

Conclusions
In our work, we systematically evaluated the potential application of all 3d family TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu and Zn) configurations in LiSBs. Among them, TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) structures not only have excellent conductivity and structural stability, but also can produce appropriate adsorption strength for S 8 /LiPSs to inhibit shuttle effects. V-rTCNQ is predicted to have excellent charging-discharging catalytic performance because of its lower RLS barrier (3.15 eV) and Li 2 S dissociation barrier (0.88 eV) than many other 2D materials, as well as having a con-siderably low Li-ion diffusion barrier (0.41eV). AIMD simulation results reveal that the TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) substrates and the adsorption system for soluble LiPSs onto V-rTCNQ are highly evaluated for their good thermodynamic stability at high temperatures. Additionally, the synthesized Mn-rTCNQ is also very worthy of further experimental confirmation and exploration. However, we also realize that obtaining ex-cellent substrate materials from a variety of constructed SACs is a relatively complicated process, which also indicates that future theoretical work needs to explore more rapid methods for comprehensively predicting the performance of LiSBs, such as high-through-put screening or machine learning. Nevertheless, our theoretical research still proposes a new candidate for exploring novel MOF catalysts with outstanding LiSBs catalytic perfor-mance and has great guiding significance for further experimental work.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/molecules28052389/s1, Figure S1: The fully optimized unit cell of TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu and Zn) monolayers; Figure S2: The DOS diagrams of TM-rTCNQ (TM = Sc, Ti, Cr, Mn, Fe, Co, Ni, Cu and Zn) unit cell, including the total DOS (TDOS), the projected DOS of d orbital of TM(TM_d), p orbital of C and N (C_p and N_p), as well as s orbital of H (H_s). Fermi-level is at the position of gray dotted line; Figure S3: The stable adsorption configurations and energies of 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME) solvent molecules for soluble polysulfides (Li 2 S 4 , Li 2 S 6 , and Li 2 S 8 ); Figure S4: The most stable adsorption configurations and energies of TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni and Cu) for S 8 /LiPSs clusters in the Z1 adsorption pattern; Figure S5: The most stable adsorption configurations and energies of TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni and Cu) for S 8 /LiPSs clusters in the Z2 adsorption pattern; Figure S6: The charge density differences and the charge transfer amount of the adsorption systems of TM-rTCNQ (TM = Cr, Mn, Fe and Co) for S 8 /LiPSs; Figure S7: The Li 2 S decomposition (a) and Li + diffusion (b) pathways of TM-rTCNQ (TM = Cr, Mn, Fe and Co) structures; Figure S8: The AIMD simulation results of TM-rTCNQ (TM = Cr, Mn, Fe and Co) substrates at 500 K and 10 ps; Table S1: The lattice constants and the magnetic moments (Mag) of the unit cell of TM-rTCNQ (TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu and Zn) monolayers; Table S2: The relative energies of Li 2 S dissociation pathways on TM-rTCNQ (TM = V, Cr, Mn, Fe and Co) monolayers;