Asymmetric charge injection barrier yields high rectification in a rhenium(I)–organometallic compound

Abstract

Electrical rectification is an important electronic function in molecular electronics. Many organic molecules and metal complexes have been studied in this context, assisted by analysis of the related charge transport mechanisms. However, Re(I)–organometallic compounds, which hold many opportunities in electronic functions, have not been explored in electronic circuitry so far. To this end, the present study focuses on emulating electrical current rectification by π-stacked Re(I) organometallic compounds embedded between p-doped Si and indium tin oxide electrodes in a two-terminal junction configuration. Among the tested compounds, viz. [Re(CO)₄(PPh₃){κ¹-(N)-saccharinate}] (1) and [Re(CO)₃(κ²-phen){κ¹-(N)-saccharinate}] (2), the latter demonstrates a remarkable electrical current rectification ratio of ≈4 × 10³ at ±2.0 V at room temperature. The model device, composed of 2, demonstrates proficient alternating current (AC) to direct current (DC) conversion at a frequency of up to 1 kHz, tested in a half-wave rectifier configuration. Temperature-dependent experimental current–voltage analysis implies the primary role of activated long-distance hopping for the charge transport and the asymmetric charge injection barrier heights at both electrode interfaces for current rectification. The above results lay the groundwork for using diverse organometallic compounds as circuit elements in nanoelectronic devices for specific electronic functions.

---

1. Introduction

Molecular electronics (ME) aims to emulate functional building blocks of electronic circuits by molecules and their assemblies, such as self-assembled monolayers (SAMs) and oligomer films. The molecules are generally embedded between two electrodes, building molecular junctions (MJs), which, in some cases, can also be addressed by external stimuli.^1–10 These MJs can be fabricated differently by manipulating molecular functionalities, structures, compositions, orientations, deposition methods, and diverse top and bottom electrode materials and configurations.^11–14 Sitting at the crossroads of chemistry, physics, materials science, and engineering, ME provides a new platform for electronic device fabrication and for gaining a deep understanding of charge transport behavior on the molecular level.^15–20 Based on such an understanding, the rational design of ME devices – a critical aspect of ME – should be possible.^9,19,21–23 A particularly notable electronic function observed in diverse MJs is the electrical current rectification, i.e., efficient flow of electrical current in one direction while impeding it at the reverse bias polarity.^5,9,24–27 The original concept of electrical rectification in molecular electronics was theoretically introduced by Aviram and Ratner and involved a saturated σ bridge separating donor and acceptor moieties.²⁸ This concept was extended afterward so that currently, the asymmetric current–voltage (I–V) response in MJs, meanwhile realized experimentally, is attributed to various factors such as asymmetric electrode–molecule interfaces, different coupling strengths, disparate Fermi levels, specific frontier molecular orbitals, etc.^29–33

An up-and-coming class of materials in ME is organometallic compounds. They possess significant potential in terms of flexible molecular structures, metal-to-ligand coordination, charge-transport-promoting redox activity, and tunable positions of frontier molecular orbitals, which are crucial for long-range charge transport.^34–37 Various metal complexes, coordination polymers (CPs), and metal–organic frameworks (MOFs) have been extensively used to understand charge transport phenomena in two-terminal electronic devices.^38–41 SAM-based MJs and thick films of metal complexes have been well-explored in various electronic features, including resistors, rectifiers, energy conversion, memory, photodetectors, and memristors.^7,42–53 Self-assembled monolayer-based MJs and thick films of metal complexes have been well-explored in rectification that occurred both in tunneling and long-range hopping regimes.^54–58 However, the exploration and use of organometallic compounds for rectification remains limited,⁵⁹ except probably for metallocenes and metallocenyl dendrimers, which have become a key building block in molecular rectification.^60–64

In the present work, we explore a specific class of Re(I)–organometallic compounds, viz. metal carbonyls, as active circuit elements in molecular electronics. The third-row transition metal carbonyls are known to exhibit robustness, structural integrity, and stability under ambient conditions, and are thus easy to handle and utilize in a chemical reaction. We used two such rhenium(I) carbonyl compounds bearing a saccharinate and a phosphine/diimine ligand to understand electrical features in solid-state devices. The present work showcases the ability of the metal carbonyl compounds in efficient electrical current rectification in two-terminal (2T) MJs. These compounds offer several significant advantages, including the formation of diverse metallacycles and supramolecular architectures that exhibit intriguing photophysical and redox activities.⁶⁵ We provide two representative examples of such a supramolecular architecture relying on the specific structures of the parent compounds, 1 and 2 (Fig. 1a–d). Specifically, compound 2, with its robust π–π interactions between the adjacent saccharinate and phenanthroline units, demonstrates significant electrical current rectification, unlike compound 1, which exhibits symmetrical currents at both bias polarities. The potential of molecular and supramolecular design of Re(I)–organometallic compounds in ME is thus well proven, opening an avenue for further advances in this direction.

Fig. 1 Crystal structure of the Re(I)-organometallic compounds. (a) and (c) General view of (a) [Re(CO)₄(PPh₃){κ¹-(N)-sac}] (1) and (c) [Re(CO)₃(κ²-phen){κ¹-(N)-sac}] (2) showing 50% probability atomic displacement ellipsoids, with selected bond distances (Å) and angles (°): (a) Re(1)–P(1) 2.5028(6), Re(1)–N(1) 2.183(2), P(1)–Re(1)–N(1) 88.79(6); (c) Re(1)–N(1) 2.205(3), Re(1)–N(2) 2.170(3), Re(1)–N(3) 2.182(3), N(1)–Re(1)–N(2) 81.38(10), N(1)–Re(1)–N(3) 84.77(10), N(2)–Re(1)–N(3) 76.05(10). The packing structures of 1 and 2 (fragments) are shown in panels (b) and (d), respectively. Compound 2 features strong intermolecular π–π interactions. The packing structures have the same color code as used for the general view of the crystal structures. Color code: carbon (grey), nitrogen (blue), oxygen (red), sulfur (orange), and rhenium (green).

2. Results and discussion

2.1. Design, synthesis, characterization, and crystal structures

To specify the molecular characteristics providing the asymmetric current–voltage (I–V) response, two distinct Re(I)-organometallic compounds were designed, exhibiting either lesser (1) or larger (2) propensity towards intermolecular π–π stacking. The synthesis involved the reaction of a lightly stabilized mononuclear rhenium complex, [Re(CO)₄(NCMe){κ¹-(N)-sac}], with PPh₃ and 1,10-phenanthroline (phen), yielding [Re(CO)₄(PPh₃){κ¹-(N)-sac}] (1) and [Re(CO)₃(κ²-phen){κ¹-(N)-sac}] (2) in 51 and 77% yields, respectively (Scheme S1). The crystal structures of 1 (CCDC No. 2236790) and 2 (CCDC No. 2236791) and their solid-state packing in the unit cells are presented in Fig. 1a–d, respectively (also see Fig. S1 and Table S1 in the SI). Compound 1 comprises Re(I) coordinated by four carbonyls (CO), a triphenylphosphine (PPh₃), and a saccharinate (sac) ligand, whereas 2 consists of Re(I) coordinated by a sac, a phen, and three carbonyl ligands with the phen ligand chelating the metal center. In both compounds, carbonyl (CO) ligands are bound to Re(I) in a terminal fashion, and the three CO ligands in 2 adopt a facial arrangement around the metal center, occupying one face of the octahedron. The sac ligand binds Re(I) through the nitrogen in a κ¹-fashion, oriented cis to the PPh₃ ligand in 1 and cis to the phen ligand in 2. The Re(1)–P(1) [2.5028(6) Å] and Re(1)–N(1) [2.183(2) Å] bond distances in 1 are similar to those reported in the literature for related complexes.^66,67 For 2, the Re–N bond distances involving both the sac [Re(1)–N(1) 2.205(3) Å and Re(2)–N(4) 2.214(3) Å, for two independent molecules] and phen [Re(1)–N(2) 2.170(3) Å and Re(1)–N(3) 2.182(3) Å; Re(2)–N(5) 2.175(3) Å and Re(2)–N(6) 2.182(3) Å] ligands, are within the range reported for similar complexes.⁶⁸ Compound 1 has one molecule in the asymmetric unit; however, 2 has two independent molecules, one disordered dichloromethane and half a molecule of n-hexane that had been squeezed during a final refinement. According to the literature, the symmetry-independent molecules in noncentrosymmetric structures with Z′ = 2 often interact via motifs that are ‘inversion-favouring’.⁶⁹ Analyses of intermolecular contacts reveal strong π–π interactions between sac⋯sac and phen⋯phen in 2 (the range of the intermolecular C⋯C distances is 3.368–3.508 Å), while it is slightly weaker in 1 (the range of the intermolecular C⋯C distances is 3.400–3.535 Å). Such π–π interactions typically promote reduction of the HOMO–LUMO gap, which eventually could not only affect general photophysical properties of a material but also result in some striking electronic features.^70–72

2.2. Theoretical calculations

The theoretically obtained highest occupied molecular orbital (HOMO) energies for 1 (ca. −6.8 eV) and 2 (ca. −6.3 eV), along with the respective lowest unoccupied molecular orbital (LUMO) values for 1 (ca. −1.8 eV) and 2 (ca. −2.8 eV), are shown in Fig. S2a–d at the M06/Re(SDD(f)), 6-311++G**(rest) level of theory.^73–78 Accordingly, the HOMO–LUMO gap was estimated at 5.0 eV for 1 and 3.5 eV for 2. Notably, the HOMO primarily displays metal-centric d-orbital character in both compounds, while the LUMO is mostly comprised of ligand-centered π*-orbitals (Fig. S3), suggesting the potential for metal-to-ligand charge transfer. As discussed in previous reports, the shrinkage of the HOMO–LUMO gap leads to greater polarization and, thereby, to better charge transport. Therefore, the smaller HOMO–LUMO gap in 2 certainly indicates a more polarized situation as compared to 1, which is favourable for the desired long-distance hopping transport. Indeed, this is evident from the higher electric dipole moment of 10.21 D in 2 as compared to 7.39 D in 1. A molecular electrostatic potential plot (MESP) further corroborates this analysis and delineates augmented positive potential zones around oxygen atoms from the sac-ligand, as compared to that in 1 (Fig. S2e–h).

Additional information is provided by the spectroscopic data. The ¹H NMR spectrum of both Re(I) complexes displayed a series of multiplets in the aromatic region, attributed to the sac and phenyl/pyridyl protons (Fig. S4 and S5). Complementarily, the ³¹P{¹H} NMR spectrum of 1 showed a singlet at δ 9.9 ppm (Fig. S6), signifying the presence of phosphine. The FT-IR spectra of both compounds exhibit vibrational stretching frequencies at 1674 cm⁻¹ for 1 and 1680 cm⁻¹ for 2, attributed to the −C=O group in the sac ligand (Fig. S7). Additionally, features associated with the rhenium-bound carbonyl stretching mode, ν_CO, appeared between 2110–1915 cm⁻¹ for 1 and 2027–1910 cm⁻¹ for 2 (Fig. S7). In the electrospray mass spectrum of 1, a major peak corresponding to [M + Na]⁺ at m/z 766 was observed (Fig. S8a). For 2, the isotope envelopes correspond to [M + H]⁺ at m/z 634 (Fig. S8b).

2.3. Thermogravimetric analysis and photophysical properties

Thermogravimetric analysis (TGA) of the crystal samples displayed a noticeable weight loss at 307 °C for 1 and 390 °C for 2 (Fig. S9). This observation suggests high-temperature device processability and operational stability for both compounds. The higher decomposition temperature of 2 can be attributed to the stronger π–π interactions between the adjacent sac and phen units. Alternately, by looking at the first weight loss of 1 (3.34%; ca. 25 amu) and 2 (5.55%; ca. 35 amu), it seems that both molecules most likely first give off a carbonyl ligand (28 amu) during their decomposition process, and the decarbonylation of 2 would require more energy as compared to 1 since the rhenium center in 2 is electron-richer with respect to 1, thus making the Re–CO bonds relatively stronger.

The UV-vis absorption spectra of 1 and 2 recorded in acetonitrile showed ligand-based π → π* electronic transitions centered at λ_max = 268 nm (ε = 1.4 × 10³ M⁻¹ cm⁻¹) and at 274 nm (ε = 5.2 × 10³ M⁻¹ cm⁻¹), respectively (Fig. S10). The lowest energy transitions related to the metal-to-ligand charge transfer (MLCT, dπ(Re) → π*_L) appeared at λ_max = 322 nm (ε = 0.78 × 10³ M⁻¹ cm⁻¹) for 1 and at λ_max = 392 nm (ε = 0.74 × 10³ M⁻¹ cm⁻¹) for 2. Similar absorption spectra were reported for different Re(I) compounds earlier.⁷⁹ Interestingly, at an excitation wavelength, λ_ex = 392 nm, 2 showed remarkable luminescence (emission wavelength, λ_em = 580 nm, Fig. S11 and Table S2) associated with a triplet MLCT emissive state (³MLCT, (dπ(Re) → π*(N^N))) as observed for related Re(I) tricarbonyl diimine complexes.^80,81 In contrast, 1 was found to be non-luminescent.

2.4. Electrochemical properties

The electrochemical properties of Re(I) compounds were studied with cyclic voltammetry (CV) and differential pulse voltammetry (DPV). We used a 1 mM solution of the respective compound in dry CH₃CN with 0.1 M tetrabutylammonium hexafluorophosphate (Bu₄NPF₆) as a supporting electrolyte. Complex 1 showed a one-electron quasi-reversible reduction event at E_½ = −1.35 V vs. Ag/AgNO₃, related to ligand reduction forming [Re(CO)₄(PPh₃){κ¹-(N)-sac}]˙⁻, which is a well-known redox process.⁸² In addition, 1 showed a metal-center-based irreversible one-electron oxidation at E_p = +1.42 V vs. Ag/AgNO₃ (Fig. S12 and Table S3). No significant change was observed in the CV when the scan rate was varied between 5 to 500 mVs⁻¹. The CV and DPV curves of complex 2 were recorded at ambient conditions and in a glove box (O₂ concentration < 0.5 ppm). Compound 2 showed ligand-based quasi-reversible reduction at E_½ = −1.72 V together with a quasi-reversible one-electron oxidation process at E_p = +1.08 V vs. Ag/AgNO₃ under inert conditions, assigned to the Re(I)/Re(II) couple (d⁶ to d⁵ electronic state) (Fig. S13 and S14). At a more anodic potential, another oxidative feature at E_p = +1.53 V vs. Ag/AgNO₃ was observed. It can be ascribed to the oxidation of [Re(CO)₃(κ²-phen)(NCMe)]˙ radical species generated during the first oxidation process, owing to the substitution of saccharinate by a solvent molecule, a well-renowned electrochemical–chemical–electrochemical (ECE) process observed in related Re(I) tricarbonyl diimine complexes.⁸³ However, this mechanism remains debatable; instead of solvent substitution, few electrochemically generated [Re^II(CO)₃(κ²-phen){κ¹-(N)-sac}]⁺ species might undergo dimerization via a sac ligand coordinating to both rhenium metal centers by the N and O donor sites forming [Re^II(CO)₃(κ²-phen)]-sac-[Re^II(CO)₃(κ²-phen){κ¹-(N)-sac}] that ultimately disproportionates into [Re^I(CO)₃(κ²-phen)(NCMe)]⁺ and [Re^III(CO)₃(κ²-phen){κ¹-(N)-sac}{κ¹-(O)-sac}]²⁺ (Fig. S15). The chemically generated [Re^I(CO)₃(κ²-phen)(NCMe)]⁺ might be responsible for the second anodic feature. A similar electrochemical behavior of 2 was also observed in ambient conditions, with a higher anodic current. Presumably, more chemical species are generated at the interface, stimulating the current, which is expected since atmospheric dioxygen can catalyze the related chemical processes. Additional electrochemical data for 1 and 2 are presented in the SI (Fig. S16–S18 and Table S4).

2.5. Thin film preparation and characterization

Thin films of 1 and 2 were prepared on freshly cleaned and dried p⁺⁺–Si electrode surfaces (highly boron-doped Si, orientation 〈100〉, NANOSHEL, resistivity ∼0–100 Ω cm, a thickness of 430 μm) by spin-coating, followed by drying. The thickness of the molecular films was measured by an optical profilometer and estimated at 31 ± 2 nm for 2 (Fig. S19). To further characterize the chemical contents of the thin films, X-ray photoelectron spectroscopy (XPS) was used. The survey and range-specific XPS spectra are shown in Fig. S20–S23. The Re 4f spectra of 1 and 2 exhibit a single Re 4f_7/2,5/2 doublet at binding energies (BEs) of 41.2 and 43.53 eV for 1 and 41.1 and 43.38 eV for 2, respectively (Fig. S22), which is in good agreement with the previous reports.^84,85 The N 1s spectra of 1 and 2 are dominated by a strong peak at a binding energy of 401.8 eV assigned to the C−N and N−(C=O) bonds (Fig. S22). This peak is accompanied by a weaker signal at a BE of ∼399 eV, assigned to the organic matrix environment and bonding with the Re(I) metal ion.^85,86 Notably, a Si 2p signal associated with the substrate was found in the survey scans of both films, indicating possibly pinhole formation in the films (Fig. S23). UV-vis spectra of the films revealed a HOMO–LUMO gap of 3.67 and 2.36 eV for 1 and 2, respectively (Fig. S24), somewhat different from the theoretical values for the gas phase (see above). Furthermore, the FT-IR spectra of the films recorded both with and without the application of 3 V bias to the p⁺⁺–Si substrate do not show any significant change in the CO stretching frequencies for both 1 and 2, indicating the stability of the CO group on the application of external bias (Fig. S25). This information suggests that Re(I)–organometallic-based electronic devices can sustain such potential without losing any molecular components.

Field-emission scanning electron microscopy (FE-SEM) images of thin films of 1 and 2 prepared on freshly cleaned p⁺⁺–Si substrates exhibit homogenous thin films with smooth surfaces (Fig. S26a and b). Cross-sectional FE-SEM images were employed to determine the film thickness at four distinct locations (T₁–T₄), revealing average thicknesses of 26.91 ± 1.29 nm for compound 1 and 34.65 ± 2.42 nm for compound 2, corresponding to thick oligomeric layers (Fig. S26c–d and Table S5). Furthermore, high-resolution transmission electron microscopy (HR-TEM) images reveal the formation of circular particles for both films (Fig. S27).

2.6. Device fabrication and electrical characterization

Two-terminal MEs devices were prepared by placing thin films of 1 and 2 between the bottom p⁺⁺–Si electrode (substrate) and the top ITO electrode. The vertical configuration p⁺⁺–Si/1/ITO and p⁺⁺–Si/2/ITO MJs were thus formed, termed devices 1 and 2, respectively. The respective fabrication steps are shown in Fig. 2a (see section 15 in the SI for details). The films of 1 and 2 were prepared by spin-coating, as described in the previous section. The top electrodes of thickness 100 ± 5 nm were deposited by magnetron sputtering over a shadow mask. Their lateral dimensions defining the area of the molecular junctions were 1.4 × 0.05 cm².

Fig. 2 Fabrication and characterization of the two-terminal electronic device. (a) Schematic diagram of the p⁺⁺–Si/2/ITO molecular junctions and illustration of the respective fabrication steps. The side view of the molecular junction is sketched; (b) and (c) single-sweep I–V characteristics of device 2 in (b) linear and (c) semi-logarithmic scale; (d) dependence of the rectification ratio for this device on the applied bias; and (e) Arrhenius plots at different applied voltages, varying from −0.2 to −1.0 V.

An external bias was applied to the top electrode during the electrical measurements, while the bottom electrode was grounded. The I–V characteristics of device 2 in linear and semi-logarithmic scales are presented in Fig. 2b and c. Additional electrical measurement data, including those for device 1, are provided in the SI (Fig. S28 and S29). Note that the devices were tested in both single (from −V to +V) and double (from −V to +V and back) sweeping modes, with no noticeable difference in the current values, suggesting no charge trapping by defects or water adsorption, leading frequently to a hysteresis.

Additional data, including the double sweep I–V curves recorded in the different bias ranges (from ±1 to ±5 V) for device 2 in linear and semi-logarithmic scales, are presented in Fig. S30, along with the respective statistical conductance plot. In all cases, for device 2, more efficient charge transport in the negative bias region than the positive one was recorded, corresponding to the rectifier behavior. The efficiency of rectification can be quantified by the so-called electrical current rectification ratio, RR, defined by eqn (1),

A plot of RR as a function of applied potential is shown in Fig. 2d. Accordingly, the rectification starts at a “threshold voltage” (V_T) of ∼0.4 V, and increases progressively with increasing voltage. At 2 V, the RR value was found to be 4.3 (±0.5) × 10³, whereas at 5 V, it was 8.4 (±0.5) × 10³. To ensure the reliability of the results, 22 devices fabricated in different batches (1–3) were tested, and a yield of more than 70% was achieved (Tables S6 and S7). In addition, device 2 was tested over a period of 30 minutes to evaluate short-term stability, showing no significant RR degradation (Fig. S31a). Moreover, reversing the SMU bias connections resulted in ‘mirror’ rectification behavior, with the direction of higher current appropriately reversed (Fig. S32), further confirming the molecular origin of the observed rectification behaviour. Significantly, the I–V characteristics of device 1 showed a symmetrical I–V response at both single and double sweep modes (Fig. S28 and S29). At the same time, device 1 showed higher conductance (268 ± 25 μS) than device 2 (80 ± 3 μS). Significantly, an Ohmic response from the reference p⁺⁺–Si/ITO device containing no molecular film (Fig. S33) confirms that the rectifier behavior of device 2 has a molecular origin and is not because of the Schottky barrier formation between the electrodes. The activation energies (E_a) corresponding to the change transport in device 2 at several specific bias values were obtained from the Arrhenius plots drawn in a temperature range of 100–300 K (Fig. 2e). These plots show a quasi-linear character in the 100–180 and 220–300 K ranges, with different slopes. The respective E_a values are compiled in Table S8. These values are noticeably higher for the 220–300 K range (E_a ≈ 150–162 meV) compared to the E_a values recorded in a temperature range of 100–180 K (E_a ≈ 14.4–58.5 meV). This behavior is attributed to the thermally activated hopping transport mechanism for long-range charge transport, which is a well-established phenomenon in ME.^87–90

2.7. Electrical impedance spectra

The electrical impedance spectra of device 2 were measured in a frequency range of 10 Hz to 1 MHz with an AC amplitude of 150 mV while the DC bias was kept at 0 V. The respective Nyquist plot is presented in Fig. S34a. A plot of real impedance as a function of frequency and a plot of the phase angle (in degrees) as a function of frequency are shown in Fig. S34b and c. Furthermore, plots of real capacitance vs. frequency are shown in Fig. S34d and e, while the proposed equivalent circuit is displayed in Fig. S34f. The fits performed for this circuit agree well with the experimental data, with a goodness of fit of 2.6 × 10⁻³. The fitting parameters are compiled in Table S9.

2.8. AC to DC efficacy of the Re(I)–organometallic molecular junctions

The rectification ability of devices 1 and 2 was tested in the half-wave rectifier configuration (Fig. 3a). An AC signal, V_input = V_A sin (2πft), with variable frequency (f), was given as the input, and the output signal was measured across a 180 kΩ series resistance. The applied AC peak voltage amplitude (V_A) was kept at ±3 V, and a DC voltage (V_DC = √2 × V_A) was fixed at 2.12 V for the AC to DC conversion (Fig. S35). The series resistance was introduced into the circuit to prevent the electrical breakdown of the device. Other researchers used a similar approach for AC-based rectification studies with MJs.⁹¹

Fig. 3 Rectification performance (AC to DC conversion) of device 2. (a) Schematics of the half-wave rectifier based on the two-terminal molecular electronic devices and (b) the respective input and output curves at different frequencies (from 50 Hz to 10 kHz). The input signal was applied to the top ITO electrode. The output voltage was measured across a 180 kΩ load at the bottom electrode, as shown in Fig. 3a.

The recorded output voltage of device 2 ensures efficient half-wave rectification for AC to DC conversion in the frequency range of 50 Hz–10 kHz (Fig. 3b). At 10 kHz, the input AC signal passes through the device with nearly no AC to DC conversion (Fig. 3b). Significantly, the control AC to DC conversion measurements with device 1 showed no electrical current rectification ability (Fig. S36), emphasizing the performance of device 2, which is of a molecular origin.

2.9. Mechanism of rectification through the band diagram model

An energy band diagram model was designed to illustrate the probable rectification mechanism. To depict the energy level diagrams of the devices, the experimentally determined HOMO and LUMO values of the respective Re(I) compounds were used. The Fermi levels of the heavily boron-doped Si and ITO were set to −5.0 eV (E_F1) and −4.3 eV (E_F2), respectively.^92–94 Note that a weaker electronic coupling between the molecules and the bottom electrode compared to the top electrode was assumed, as the electrodes were connected to the organic film by non-covalent interaction. The energy level diagram of device 2 in the case of isolated electrodes shows that the HOMO (−5.73 eV) of 2 has the energy barriers of 0.73 and 1.43 eV with respect to p⁺ Si (E_F1) and ITO (E_F2), respectively (Fig. 4a). Under the electrical contact conditions (V = 0 V), a joint Fermi level, corresponding to the potential equilibrium, is established (Fig. 4b). The bottom p⁺⁺–Si electrode was grounded, and the top ITO electrode was subjected to the applied bias. When a bias of −2 V (this voltage is preferred for model design, as we have obtained a mean rectification ratio of 4.3 (±0.5) × 10³ at ±2 V) was applied to the junction, the Fermi level (E_F2) of ITO lifts and approaches the LUMO (−2.6 eV) and LUMO+1 (−2.46 eV) of compound 2, elevating charge conduction through the film (Fig. 4c). On the contrary, when a bias of +2 V was applied to the junction, the Fermi level of the ITO electrode lowers, leading to a high potential barrier for the charge carriers between the Fermi level of ITO and the frontier molecular orbitals of 2 (Fig. 4d). This barrier restricted the charge transport in the MJs and yielded an asymmetric current flow with a high rectification ratio. On the contrary, device 1 showed symmetrical I–V characteristics in response to biases of ±2 V. According to the band diagram model, at −2 V, the Fermi energy of the ITO electrode reaches the LUMO, thus, current flow occurs in the junction. Similarly, at +2 V, the HOMO of compound 1 remains close to the Fermi level of ITO, so charge conduction occurs in the device (Fig. S37).

Fig. 4 Proposed asymmetric charge transport mechanism for device 2. The energy level diagrams and illustration of the charge transport mechanism for device 2 (a) at electrically isolated conditions, (b) V = 0 V, (c) V = −2 V, and (d) V = +2 V. The voltage is applied to the top ITO electrode, while the bottom p⁺⁺–Si electrode is grounded.

3. Concluding remarks

A primary goal in the design of molecular rectifiers is achieving a high rectification ratio (RR > 10³) at possibly low bias while operating at high frequency (>1 kHz) for AC to DC signal conversion. To pursue this objective, we integrated two Re(I) organometallic compounds with different π-stacking abilities, electric dipole moments, and positions of the frontier molecular orbitals into MJs with boron-doped Si and ITO electrodes. The above differences turned out to be decisive in the context of rectification. Whereas the MJs of the compound with low π–π stacking tendency and lower dipole moment (device 1) showed no rectification ability, those of the compound with high stacking tendency and higher dipole moment (device 2) exhibited a distinct rectification behavior with an RR of 4.3 × 10³ at ±2 V and 8.4 × 10³ at ±5 V. The AC to DC conversion experiments performed with the latter device in the half-wave rectifier configuration confirmed its rectification ability and showed efficient rectification and the related AC to DC conversion in a broad frequency range, up to 1 kHz and probably even somewhat higher. A model of the charge transport mechanism behind the observed behavior was suggested, assisted by complementary experiments. While our Re(I) organometallic molecular diodes exhibit high electrical current rectification features, there remains room for improvement, particularly in achieving even higher RR values, broadening the frequency range, and enhancing stability. This suggests the need for further research and optimization in designing these molecular systems, alongside the DC and AC to DC electrical measurements. In a broader context, we hope that our results will promote the utilization of various organic, metal-complexes, organometallics, metal–organic frameworks (MOFs), and covalent organic frameworks (COFs) in electronic devices.

4. Experimental section

4.1. Materials

Dirhenium decacarbonyl [Re₂(CO)₁₀] was purchased from Strem Chemical Inc. and used without further purification. Triphenylphosphine (PPh₃, purity = 99%), 1,10-phenanthroline (phen, 99%), and saccharin (sacH, 98%) were purchased from Sigma-Aldrich and used as received. All reactions were performed under an atmosphere of dry nitrogen using standard Schlenk techniques unless otherwise stated. Reagent-grade solvents were dried by standard methods and freshly distilled before use.

4.2. Synthesis of [Re(CO)4(PPh3){κ¹-(N)-sac}] (1)

The precursor compounds, [Re(CO)₄(NCMe){κ¹-(N)-sac}] (25 mg, 0.048 mmol) and PPh₃ (13 mg, 0.050 mmol) were taken in 20 mL of benzene and stirred at room temperature for 72 h. The reaction was monitored by performing analytical TLC. The solvent was removed by using rotary evaporation, and the residue was chromatographed by preparative TLC on silica gel. Elution with cyclohexane/CH₂Cl₂ (2 : 3; V/V) developed three bands on the TLC plates. The first and third bands were unconsumed PPh₃ (2 mg) and [Re(CO)₄(NCMe){κ¹-(N)-sac}] (5 mg), respectively, while the second band gave [Re(CO)₄(PPh₃){κ₁-(N)-sac}] (1) (18 mg, 51%) isolated as colorless crystals after recrystallization from a toluene/CH₂Cl₂ mixture kept at 4 °C.

4.3. Synthesis of [Re(CO)₃(κ²-phen){κ¹-(N)-sac}] (2)

The precursor compound, [Re(CO)₄(NCMe){κ¹-(N)-sac}] (25 mg, 0.048 mmol) and 1,10-phenanthroline (9 mg, 0.050 mmol) were added to 20 mL of benzene solution, and the reaction mixture was stirred at room temperature for 72 h. The colorless reaction mixture turned yellow. The reaction was monitored by analytical chromatography. The solvent was removed under a vacuum, and the residue was separated by preparative TLC on silica gel. Elution with CH₂Cl₂ developed three bands on the TLC plates. The first and third bands were unconsumed [Re(CO)₄(NCMe){κ¹-(N)-sac}] (3 mg) and 1,10-phenanthroline (trace), respectively. The second band afforded [Re(CO)₃(κ²-phen){κ¹-(N)-sac}] (2) (24 mg, 77%) as yellow crystals after recrystallization from an n-hexane/CH₂Cl₂ mixture kept at 4 °C.

4.4. Crystal structure analysis

X-ray diffraction data were collected on a XtaLAB Synergy, Dualflex, HyPix diffractometer at 150 K using Mo-Kα radiation (λ = 0.71073 Å). Unit cell determination, data reduction, and absorption corrections were carried out using CrysAlisPro. The structures were solved using intrinsic phasing with the ShelXT structure solution program and refined by full-matrix least-squares based on F2 using ShelXL within the OLEX2 graphical user interface. Single crystals of 1 and 2 suitable for X-ray diffraction characterization were grown by slow diffusion of either toluene (for 1) or n-hexane (for 2) into a CH₂Cl₂ solution of each compound at 4 °C. The crystals were mounted on a nylon loop. All non-hydrogen atoms were refined anisotropically, and hydrogen atoms were included using a riding model. The asymmetric unit of 2 consists of two molecules of 2, a CH₂Cl₂ molecule, and a half molecule of n-hexane. The CH₂Cl₂ molecule is disordered over two sites with an occupancy of ca. 7 : 3, while the n-hexane molecule is masked as it could not be modeled due to bad geometry. The details of the data collection and structure refinement are given in Table S1. CCDC 2236790 and CCDC 2236791 contain the supplementary crystallographic data for 1 and 2, respectively.

4.5. Computational details

All calculations were performed within the Gaussian 09 quantum chemical package.⁸ Geometries were optimized in the solvent phase with the B3LYP functional, in conjunction with SDD basis and the MWB60 effective core potential on Re and Pople's 6-31G(d,p) double-ζ split valence basis set on the rest of the atoms. The hybrid B3LYP functional was chosen for optimization of 1 and 2 in the gas phase, with 6-31G** Pople basis set on all atoms except Re(I), which was treated with the SDD all-electron basis set and MWB60 ECP assigning 60 electrons in the core–shell. A medium cube was used to generate the density and the potential, which were then mapped together to plot the molecular electrostatic potential.

4.6. Instrumentation

The NMR spectra were recorded on a Bruker Avance III HD (400 MHz) instrument. All chemical shifts are reported in δ in ppm units and are referenced to the residual protons of the deuterated solvents (¹H) and external 85% H₃PO₄ (³¹P). Fourier-transform infrared (FT-IR) spectra were recorded on a Shimadzu FTIR Prestige 21 spectrophotometer. Elemental analyses were performed by the Microanalytical Laboratory of Wazed Miah Science Research Center, Jahangirnagar University. The thermal stability of the compounds was determined by performing TGA under a nitrogen atmosphere using the SDT-650 TGA system. The baseline-corrected UV-vis absorption spectra were recorded on a V-770 JASCO double-beam spectrophotometer. X-ray photoelectron spectra were recorded with a K-Alpha instrument (Thermo Scientific) equipped with a monochromatic Al X-ray source (Al Kα line: 1486 eV), under UHV conditions.

4.7. Electrochemical measurements in ambient and inert environments

Cyclic voltammograms (CV) and differential pulse voltammograms (DPV) of 1 and 2 were recorded under nitrogen-saturated dry acetonitrile with 0.1 M tetrabutylammonium hexafluorophosphate (Bu₄NPF₆) as a supporting electrolyte using a glassy carbon electrode (working electrode) with an electrode area of 0.0707 cm², platinum wire (counter electrode), and Ag/AgNO₃ (reference electrode). For the CV recording, 1 mM of the respective compound (1) was taken in an electrochemical cell, and the scan rates were varied from 5 to 500 mV s⁻¹ (Fig. S12). For 2, the CVs were recorded under two conditions, viz., in (i) ambient and (ii) inert environment (in a glove box, moisture and O₂ concentration < 0.5 ppm) (Fig. S13). The CVs recorded at the variable scan rates are shown in Fig. S14.

4.8. Thin film preparation and two-terminal device fabrication

Compounds 1 and 2 were independently deposited on quartz and p⁺⁺–Si (orientation 〈100〉, NANOSHEL, resistivity ∼0–100 Ω cm, thickness 430 μm) substrates. A solution of a respective compound, 10⁻⁵ M in anhydrous acetonitrile (dried with a molecular sieve), was spin-coated (at 4000 rpm for 40 s) on freshly cleaned p⁺⁺–Si quartz substrates. All films were dried at 70 °C in a vacuum oven for 10 minutes to remove the excess solvent. A top ITO contact of a thickness of 100 ± 5 nm was deposited on top of either 1 or 2 film using magnetron sputtering (45 minutes, 30 W, 15 SCCM, 20 rpm, 1.76 × 10⁻³ mbar) through a customer-designed shadow mask.

4.9. DC and AC-based electrical characterization

The electrical characterization (current–voltage) of the Re(I) organometallic molecular junctions was performed using a Keithley source meter (Model-2604B). Electrical impedance measurements were carried out using Gamry Instruments (Reference 600+). The half rectifier characteristics were measured by applying an AC signal to the circuit, using a Tektronix AFG 1022 arbitrary function generator (f_max = 25 MHz). The output and input sinusoidal signals were monitored using a Tektronix TBS 1000C series digital oscilloscope.

Author contributions

S. R.: experimental, data analysis, main draft writing, N. P.: experimental, data analysis, main draft writing, R. G.: electrochemical experiments, draft editing, R. R.: synthesis, P. S.: characterization, and revision, V. N. N.: XRD measurement, L. R.: DFT study, S. G.: draft writing, supervision, M. Z.: main draft editing, P. C. M.: conceptualization, main draft writing and editing, supervision, funding.

Conflicts of interest

The authors declare no competing financial interests.

Data availability

The data supporting this article have been included as part of the SI. Supplementary information: Experimental procedures; FT-IR data; ¹H, ³¹P{¹H} NMR; XPS data; crystal data; and structure refinement; TGA data; solid-state UV-vis data; electrochemical data recorded under intert conditions; additional I–V plots; activation energy calculation; AC-based electrical measurements; electrical impedance data; and computational details (PDF). See DOI: https://doi.org/10.1039/d5tc01614b.

CCDC 2236790 and 2236791 contain the supplementary crystallographic data for this paper.^95a,b

Acknowledgements

S. R. acknowledges the Department of Science and Technology, Indian National Science Academy (DST, INSA), Government of India, for the prestigious Indian Research Fellowship (ISRF, Grant No. INSA/CHM/2022427) to a short-term research stay at IIT Kanpur. N. P. is thankful to IIT Kanpur for providing an Institute Post-doctoral Fellowship (PDF-313). R. G. is thankful to IIT Kanpur for a senior research fellowship. P. C. M. acknowledges the Ministry of Education for the SCHEME FOR TRANSFORMATIONAL AND ADVANCED RESEARCH IN SCIENCES (STARS) (Grant No. 2023-0535), and the Anusandhan National Research Foundation, New Delhi (Grant No. CRG/2022/005325) for financial support. S. G. and S. R. acknowledge the Ministry of Education, Government of the People's Republic of Bangladesh, for the financial support (Grant No. PS20211624). The authors acknowledge IIT Kanpur for its infrastructure and equipment facilities. The authors are thankful to Dr A. Vilan and Prof. C. Frisbie for their critical comments that helped enrich the quality of the draft.

References

1. M. Gu, C. Lai, I. Ni, C. Wu and C. Chen, Angew. Chem., Int. Ed., 2023, 62, 202214963.
2. L. Li, C. R. Prindle, W. Shi, C. Nuckolls and L. Venkataraman, J. Am. Chem. Soc., 2023, 145, 18182–18204.
3. R. Gupta, J. A. Fereiro, A. Bayat, A. Pritam, M. Zharnikov and P. C. Mondal, Nat. Rev. Chem., 2023, 7, 106–122.
4. T. Wang, X. Zhang, Z. Wang, X. Zhu, J. Liu, X. Min, T. Cao and X. Fan, Polymers, 2019, 11, 1564.
5. D. A. Egger, F. Rissner, E. Zojer and G. Heimel, Adv. Mater., 2012, 24, 4403–4407.
6. Z. S. Yao, Z. Tang and J. Tao, Chem. Commun., 2020, 56, 2071–2086.
7. H. Atesci, V. Kaliginedi, J. A. Celis Gil, H. Ozawa, J. M. Thijssen, P. Broekmann, M. Haga and S. J. van der Molen, Nat. Nanotechnol., 2018, 13, 117–121.
8. P. C. Mondal, U. M. Tefashe and R. L. McCreery, J. Am. Chem. Soc., 2018, 140, 7239–7247.
9. A. Vilan, D. Aswal and D. Cahen, Chem. Rev., 2017, 117, 4248–4286.
10. D. I. Son, T. W. Kim, J. H. Shim, J. H. Jung, D. U. Lee, J. M. Lee, W. Il Park and W. K. Choi, Nano Lett., 2010, 10, 2441–2447.
11. G. D. Kong, J. Jang, S. Choi, G. Lim, I. S. Kim, T. Ohto, S. Maeda, H. Tada and H. J. Yoon, Small, 2024, 20, 2305997.
12. Y. Zhu, Y. Zhou, L. Ren, J. Ye, H. Wang, X. Liu, R. Huang, H. Liu, J. Liu, J. Shi, P. Gao and W. Hong, Angew. Chem., Int. Ed., 2023, 135, 202302693.
13. Q. Pi, D. Bi, D. Qiu, H. Wang, X. Cheng, Y. Feng, Q. Zhao and M. Zhou, J. Mater. Chem. C, 2021, 9, 8994–9000.
14. Z. Wang, H. Dong, T. Li, R. Hviid, Y. Zou, Z. Wei, X. Fu, E. Wang, Y. Zhen, K. Nørgaard, B. W. Laursen and W. Hu, Nat. Commun., 2015, 6, 7478.
15. R. Gupta, S. Bhandari, S. Kaya, K. P. Katin and P. C. Mondal, Nano Lett., 2023, 23, 10998–11005.
16. B. Capozzi, J. Xia, O. Adak, E. J. Dell, Z.-F. Liu, J. C. Taylor, J. B. Neaton, L. M. Campos and L. Venkataraman, Nat. Nanotechnol., 2015, 10, 522–527.
17. G. Koplovitz, G. Leitus, S. Ghosh, B. P. Bloom, S. Yochelis, D. Rotem, F. Vischio, M. Striccoli, E. Fanizza, R. Naaman, D. H. Waldeck, D. Porath and Y. Paltiel, Small, 2019, 15, 1804557.
18. J. Tang, Y. Wang, J. E. Klare, G. S. Tulevski, S. J. Wind and C. Nuckolls, Angew. Chem., Int. Ed., 2007, 46, 3892–3895.
19. L. Venkataraman, J. E. Klare, C. Nuckolls, M. S. Hybertsen and M. L. Steigerwald, Nature, 2006, 442, 904–907.
20. R. Kaur, B. Singh, V. Singh, M. Zharnikov and P. C. Mondal, Coord. Chem. Rev., 2006, 442, 904–907.
21. M. Kulke, D. M. Olson, J. Huang, D. M. Kramer and J. V. Vermaas, Small, 2023, 19, 2304013.
22. D. Rambabu, A. E. Lakraychi, J. Wang, L. Sieuw, D. Gupta, P. Apostol, G. Chanteux, T. Goossens, K. Robeyns and A. Vlad, J. Am. Chem. Soc., 2021, 143, 11641–11650.
23. N. Singh, A. Malik, P. Sethi and P. C. Mondal, Small, 2024, 20, 2403108.
24. K. S. Wimbush, W. F. Reus, W. G. van der Wiel, D. N. Reinhoudt, G. M. Whitesides, C. A. Nijhuis and A. H. Velders, Angew. Chem., Int. Ed., 2010, 49, 10176–10180.
25. C. Van Dyck and M. A. Ratner, Nano Lett., 2015, 15, 1577–1584.
26. D. D. James, A. Bayat, S. R. Smith, J.-C. Lacroix and R. L. McCreery, Nanoscale Horiz., 2017, 3, 45–52.
27. P. Song, S. Guerin, S. J. R. Tan, H. V. Annadata, X. Yu, M. Scully, Y. M. Han, M. Roemer, K. P. Loh, D. Thompson and C. A. Nijhuis, Adv. Mater., 2018, 30, 1706322.
28. A. Aviram and M. A. Ratner, Chem. Phys. Lett., 1974, 29, 277–283.
29. H. Jeong, D. Kim, D. Xiang and T. Lee, ACS Nano, 2017, 11, 6511–6548.
30. C. N. Verani, Dalton Trans., 2018, 47, 14153–14168.
31. J. E. Green, J. Wook Choi, A. Boukai, Y. Bunimovich, E. Johnston-Halperin, E. DeIonno, Y. Luo, B. A. Sheriff, K. Xu, Y. Shik Shin, H.-R. Tseng, J. F. Stoddart and J. R. Heath, Nature, 2007, 445, 414–417.
32. W. Peng, N. Chen, C. Wang, Y. Xie, S. Qiu, S. Li, L. Zhang and Y. Li, Angew. Chem., Int. Ed., 2023, 62, e202307733.
33. R. K. Parashar, P. Jash, M. Zharnikov and P. C. Mondal, Angew. Chem., Int. Ed., 2024, e202317413.
34. U. M. Tefashe, Q. Van Nguyen, F. Lafolet, J.-C. Lacroix and R. L. McCreery, J. Am. Chem. Soc., 2017, 139, 7436–7439.
35. S. Goswami, D. Deb, A. Tempez, M. Chaigneau, S. P. Rath, M. Lal, Ariando, R. S. Williams, S. Goswami and T. Venkatesan, Adv. Mater., 2020, 32, 2004370.
36. Y. Liu, R. Sakamoto, C. L. Ho, H. Nishihara and W. Y. Wong, J. Mater. Chem. C, 2019, 7, 9159–9166.
37. L. Yuan, R. Breuer, L. Jiang, M. Schmittel and C. A. Nijhuis, Nano Lett., 2015, 15, 5506–5512.
38. V. Malik, S. Maity and R. Chatterjee, Org. Electron., 2018, 56, 5–10.
39. J. H. Dou, L. Sun, Y. Ge, W. Li, C. H. Hendon, J. Li, S. Gul, J. Yano, E. A. Stach and M. Dincǎ, J. Am. Chem. Soc., 2017, 139, 13608–13611.
40. L. S. Xie, E. V. Alexandrov, G. Skorupskii, D. M. Proserpio and M. Dincă, Chem. Sci., 2019, 10, 8558–8565.
41. A. Prasoon, B. Dhara, D. Roy, S. Rana, S. Bhand and N. Ballav, Chem. Sci., 2019, 10, 10040–10047.
42. M. Supur, S. K. Saxena and R. L. McCreery, J. Am. Chem. Soc., 2020, 142, 11658–11662.
43. A. Schoedel, C. Scherb and T. Bein, Angew. Chem., Int. Ed., 2010, 49, 7225–7228.
44. S. Sinha, M. Sahad E, R. Mondal, S. Das, L. T. Manamel, P. Brandão, B. de Bruin, B. C. Das and N. D. Paul, J. Am. Chem. Soc., 2022, 144, 20442–20451.
45. P. Sachan and P. C. Mondal, J. Mater. Chem. C, 2022, 10, 14532–14541.
46. R. Sakamoto, S. Katagiri, H. Maeda and H. Nishihara, Coord. Chem. Rev., 2013, 257, 1493–1506.
47. K. Terada, K. Kanaizuka, V. M. Iyer, M. Sannodo, S. Saito, K. Kobayashi and M. Haga, Angew. Chem., Int. Ed., 2011, 123, 6411–6415.
48. V. Rubio-Giménez, M. Galbiati, J. Castells-Gil, N. Almora-Barrios, J. Navarro-Sánchez, G. Escorcia-Ariza, M. Mattera, T. Arnold, J. Rawle, S. Tatay, E. Coronado and C. Martí-Gastaldo, Adv. Mater., 2018, 30, 1704291.
49. C. Musumeci, G. Zappalà, N. Martsinovich, E. Orgiu, S. Schuster, S. Quici, M. Zharnikov, A. Troisi, A. Licciardello and P. Samorì, Adv. Mater., 2014, 26, 1688–1693.
50. J. Shang, G. Liu, H. Yang, X. Zhu, X. Chen, H. Tan, B. Hu, L. Pan, W. Xue and R. Li, Adv. Funct. Mater., 2014, 24, 2171–2179.
51. N. Tuccitto, V. Ferri, M. Cavazzini, S. Quici, G. Zhavnerko, A. Licciardello and M. A. Rampi, Nat. Mater., 2009, 8, 41–46.
52. H. J. Yoon, K. C. Liao, M. R. Lockett, S. W. Kwok, M. Baghbanzadeh and G. M. Whitesides, J. Am. Chem. Soc., 2014, 136, 17155–17162.
53. X. Yao, X. Sun, F. Lafolet and J.-C. Lacroix, Nano Lett., 2020, 20, 6899–6907.
54. S. Shmueli, M. Cohen Jungerman, P. Shekhter and Y. Selzer, J. Phy. Chem. Lett., 2024, 15, 10602–10608.
55. X. Song, X. Yu and W. Hu, J. Phys. Chem. C, 2020, 124, 24408–24419.
56. J. Komeda, K. Takada, H. Maeda, N. Fukui, T. Tsuji and H. Nishihara, Chem. – Eur. J., 2022, 28, e202201316.
57. L. Tian, A. Fan, X. Yu and W. Hu, CCS Chem., 2023, 5, 902–914.
58. L. Tian, X. Song, X. Yu and W. Hu, J. Phys. Chem. C, 2021, 125, 21614–21623.
59. F. Schwarz, G. Kastlunger, F. Lissel, C. Egler-Lucas, S. N. Semenov, K. Venkatesan, H. Berke, R. Stadler and E. Lörtscher, Nat. Nanotechnol., 2016, 11, 170–176.
60. D. Astruc, C. Ornelas and J. Ruiz, Acc. Chem. Res., 2008, 41, 841–856.
61. G. Roy, R. Gupta, S. Ranjan Sahoo, S. Saha, D. Asthana and P. C. Mondal, Coord. Chem. Rev., 2022, 473, 214816.
62. B. Lawson, P. Zahl, M. S. Hybertsen and M. Kamenetska, J. Am. Chem. Soc., 2022, 144, 6504–6515.
63. X. Chen, M. Roemer, L. Yuan, W. Du, D. Thompson, E. Barco and C. A. Nijhuis, Nat. Nanotechnol., 2017, 12, 797–803.
64. Y. Liu, E. Zojer and M. Zharnikov, ACS Appl. Mater. Interfaces, 2022, 14, 52499–52507.
65. D. Gupta, P. Rajakannu, B. Shankar, R. Shanmugam, F. Hussain, B. Sarkar and M. Sathiyendiran, Dalton Trans., 2011, 40, 5433.
66. Md. R. Moni, S. Ghosh, S. M. Mobin, D. A. Tocher, G. Hogarth, M. G. Richmond and S. E. Kabir, J. Organomet. Chem., 2018, 871, 167–177.
67. S. E. Kabir, F. Ahmed, S. Ghosh, M. R. Hassan, M. S. Islam, A. Sharmin, D. A. Tocher, D. T. Haworth, S. V. Lindeman, T. A. Siddiquee, D. W. Bennett and K. I. Hardcastle, J. Organomet. Chem., 2008, 693, 2657–2665.
68. S. E. Kabir, J. Alam, S. Ghosh, K. Kundu, G. Hogarth, D. A. Tocher, G. M. G. Hossain and H. W. Roesky, Dalton Trans., 2009, 4458.
69. R. Taylor, J. C. Cole and C. R. Groom, Cryst. Growth Des., 2016, 16, 2988–3001.
70. H. Tsubaki, A. Sekine, Y. Ohashi, K. Koike, H. Takeda and O. Ishitani, J. Am. Chem. Soc., 2005, 127, 15544–15555.
71. T. Morimoto, C. Nishiura, M. Tanaka, J. Rohacova, Y. Nakagawa, Y. Funada, K. Koike, Y. Yamamoto, S. Shishido, T. Kojima, T. Saeki, T. Ozeki and O. Ishitani, J. Am. Chem. Soc., 2013, 135, 13266–13269.
72. M. Dharmarwardana, B. M. Otten, M. M. Ghimire, B. S. Arimilli, C. M. Williams, S. Boateng, Z. Lu, G. T. McCandless, J. J. Gassensmith and M. A. Omary, PNAS, 2021, 118, e2106572118.
73. C. J. Cramer and D. G. Truhlar, Phys. Chem. Chem. Phys., 2009, 11, 10757.
74. Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241.
75. H. Tang, E. N. Brothers, C. A. Grapperhaus and M. B. Hall, ACS Catal., 2020, 10, 3778–3789.
76. D. Hernández-Valdés, R. Alberto and U. Jáuregui-Haza, RSC Adv., 2016, 6, 107127–107140.
77. D. W. Demoin, Y. Li, S. S. Jurisson and C. A. Deakyne, Comput. Theor. Chem., 2012, 997, 34–41.
78. A. W. Ehlers, M. Böhme, S. Dapprich, A. Gobbi, A. Höllwarth, V. Jonas, K. F. Köhler, R. Stegmann, A. Veldkamp and G. Frenking, Chem. Phys. Lett., 1993, 208, 111–114.
79. B. Shankar, S. Sahu, N. Deibel, D. Schweinfurth, B. Sarkar, P. Elumalai, D. Gupta, F. Hussain, G. Krishnamoorthy and M. Sathiyendiran, Inorg. Chem., 2014, 53, 922–930.
80. K. Schindler and F. Zobi, Chimia, 2021, 75, 837–844.
81. R. Baková, M. Chergui, C. Daniel, A. Vlček Jr. and S. Záliš, Coord. Chem. Rev., 2011, 255, 975–989.
82. C. J. Stanton, C. W. Machan, J. E. Vandezande, T. Jin, G. F. Majetich, H. F. Schaefer, C. P. Kubiak, G. Li and J. Agarwal, Inorg. Chem., 2016, 55, 3136–3144.
83. A. Carreño, E. Solís-Céspedes, C. Zúñiga, J. Nevermann, M. M. Rivera-Zaldívar, M. Gacitúa, A. Ramírez-Osorio, D. Páez-Hernández, R. Arratia-Pérez and J. A. Fuentes, Chem. Phys. Lett., 2019, 715, 231–238.
84. H. Hori, K. Koike, M. Ishizuka, K. Takeuchi, T. Ibusuki and O. Ishitani, J. Organomet. Chem., 1997, 530, 169–176.
85. E. M. Johnson, R. Haiges and S. C. Marinescu, ACS Appl. Mater. Interfaces, 2018, 10, 37919–37927.
86. K. Artyushkova, B. Kiefer, B. Halevi, A. Knop-Gericke, R. Schlogl and P. Atanassov, Chem. Commun., 2013, 49, 2539.
87. K. S. Kumar, R. R. Pasula, S. Lim and C. A. Nijhuis, Adv. Mater., 2016, 28, 1824–1830.
88. Q. van Nguyen and C. D. Frisbie, J. Am. Chem. Soc., 2021, 143, 2638–2643.
89. X. Yao, X. Sun, F. Lafolet and J.-C. Lacroix, Nano Lett., 2020, 20, 6899–6907.
90. Q. Van Nguyen, P. Martin, D. Frath, M. L. Della Rocca, F. Lafolet, S. Bellinck, P. Lafarge and J.-C. Lacroix, J. Am. Chem. Soc., 2018, 140, 10131–10134.
91. J. Trasobares, D. Vuillaume, D. Théron and N. Clément, Nat. Commun., 2016, 7, 1–9.
92. L. Chkoda, C. Heske, M. Sokolowski, E. Umbach, F. Steuber, J. Staudigel, M. Stößel and J. Simmerer, Synth. Met., 2000, 111–112, 315–319.
93. P.-E. Hellberg, S.-L. Zhang and C. S. Petersson, IEEE Electron Device Lett., 1997, 18, 456–458.
94. A. Novikov, Solid State Electron, 2010, 54, 8–13.
95. (a) S. Rajbangshia, N. Pal, R. Gupta, R. Rahman, P. Sachan, V. N. Nesterov, L. Roy, S. Ghosh, M. Zharnikov and P. C. Mondal, CCDC 2236790: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2f2kjm; (b) S. Rajbangshia, N. Pal, R. Gupta, R. Rahman, P. Sachan, V. N. Nesterov, L. Roy, S. Ghosh, M. Zharnikov and P. C. Mondal, CCDC 2236791: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2f2kkn.

---

Footnote

† These two authors equally contributed to this work.

---