From π to σ: Enhanced Charge Transport in Iodine-Substituted Benzene Junctions

Abstract

Understanding how nonbonded atoms contribute to charge transport offers a pathway to molecular conduction mechanisms beyond conventional π-delocalization. Iodine-substituted benzenes provide a simple and structurally well-defined platform in which halogen–metal anchoring and potential σ-type interactions can be modulated solely by substituent number and topology. Here we investigate p-, m-, and o-diiodobenzene, 1,2,4,5-tetraiodobenzene, and hexaiodobenzene to elucidate how substitution patterns regulate the balance between π-dominated and σ-involved transport in single-molecule junctions. Break-junction measurements reveal conventional π-HOMO transport in the para isomer, destructive quantum interference in the meta isomer, and the absence of stable junction formation in the ortho isomer, where adjacent iodine atoms create a very short and geometrically constrained effective junction length. The tetraiodo derivative shows only modest conductance enhancement, indicating that partial substitution does not generate a continuous σ-framework. In contrast, hexaiodobenzene exhibits a single, narrow, and contact-insensitive conductance peak and positive thermopower, and the corresponding I–V analysis yields a HOMO level only ∼0.9 eV from the Fermi level. Together with molecular-orbital calculations, these results lead us to conclude that σ-involved HOMO-mediated transport emerges only when a complete peripheral iodine ring is present. By establishing a substituent-number-controlled transition from π-dominated to σ-involved transport in a chemically simple series, this work provides a concise design principle for accessing nonbonded σ-delocalized channels in aromatic molecular junctions.

1. Introduction

Understanding and controlling charge transport through single molecules is a central challenge in molecular electronics, offering prospects for the realization of ultrasmall electronic devices with tailored quantum functionalities. − Molecular junctions, in which an individual molecule bridges two metal electrodes, provide a powerful platform to probe fundamental charge transport phenomena at the single molecule scale. Over the past two decades, research in this field has primarily focused on π-conjugated molecules, where delocalized π-orbitals facilitate efficient charge transport and underpin a variety of quantum interference − and thermoelectric − effects at the nanoscale. Although these studies have deepened our knowledge of π-mediated conduction, relying exclusively on π-delocalized systems constrains the structural and functional diversity available for molecular design.

σ-Delocalized systems formed through through-space interactions among nonbonded atoms − represent a complementary and less explored class of molecular conductors. In this work, the term σ-delocalization does not refer to covalent σ-conjugation along bonded backbones, as in alkanes or oligosilanes, nor to σ-bonds within an aromatic ring. Instead, it specifically denotes a σ-type orbital manifold arising from through-space overlap among nonbonded peripheral atoms, such as the in-plane lone-pair orbitals of iodine substituents. In fully substituted halogenated benzenes, dense peripheral substitution enables adjacent iodine atoms to interact electronically without direct covalent bonding, forming a σ-type orbital manifold distributed around the molecular periphery (Figure a). Importantly, charge transport is governed not by the formation of σ bonds themselves but by the transport-relevant frontier orbital within this σ-type manifold; its phase characteristics, which are critical for electronic coupling, are illustrated schematically in Figure b. Because these orbitals are spatially extended and can couple to metal electrodes through multiple equivalent anchoring sites, they provide an alternative pathway for charge transport that differs fundamentally from conventional π-mediated conduction.

1.

Schematic illustration of σ-involved charge transport in iodine-substituted benzene junctions. (a) Formation of a σ-type orbital manifold via through-space overlap of nonbonded iodine lone-pair orbitals in hexaiodobenzene (C₆I₆). Dense peripheral iodine substitution enables electronic interactions among adjacent iodine atoms without formation of covalent I–I bonds, giving rise to a delocalized σ-type orbital manifold around the molecular periphery. (b) Transport-relevant frontier orbital corresponding to the HOMO within this σ-type manifold. Charge transport is governed by this frontier orbital, which exhibits antibonding character, rather than by σ-bond formation itself. For comparison, the π-type HOMO of p-C₆H₄I₂ is shown as an inset. (c) Chemical structures of the iodine-substituted benzene molecules investigated in this study.

In this work, we investigate single-molecule transport through iodine-substituted benzenes (Figure c) to elucidate the cooperative effects of anchoring and σ-delocalization. In this study, σ-delocalization refers to through-space interactions among nonbonded peripheral iodine atoms, distinct from σ-conjugation in covalently bonded backbones such as alkanes or oligosilanes. , Compared with the previously studied selenium-substituted systems, ,, the iodine-substituted systems form well-defined metal–molecule contacts through Au–halogen bonding , and serve as model systems for elucidating the interplay between anchoring effects and σ-type charge transport, owing to their single binding motif and the relatively simple junction structures and electronic states. This framework allows us to identify substituent-controlled transitions between π-dominated and σ-involved transport pathways across a structurally simple molecular series. However, the electron-withdrawing nature of iodine substitution shifts the HOMO energy further from the electrode Fermi level (E _F) in π-delocalized systems, thereby reducing both conductance and thermopower. Systematic break junction (BJ) measurements on para- and meta-diiodobenzene and hexaiodobenzene reveal a clear structure–property relationship: the para isomer shows reproducible conductance plateaus and finite thermopower, the meta isomer exhibits strongly suppressed conductance due to destructive quantum interference. − The tetraiodo derivative shows only modest conductance enhancement. In contrast, the hexa-substituted derivative achieves enhanced conductance via the σ-involved system. Notably, only hexaiodobenzene exhibits a single, narrow conductance peak that is relatively insensitive to the specific Au–I binding site. This behavior suggests that, owing to the high symmetry of C₆I₆ and the presence of multiple equivalent Au–I binding sites, the conductance distribution is relatively insensitive to the specific iodine atoms involved in electrode contact, compared with lower-symmetry derivatives.

Together with theoretical analysis, these findings establish iodine substitution as a fundamental model for a unique molecular design principle that leverages both robust metal–halogen contacts and σ-mediated charge transport. Beyond expanding the scope of molecular junction architectures, our results demonstrate that σ-systems can be strategically harnessed to complement π-systems, thereby advancing the fundamental understanding and practical design of molecular conductors.

2. Methods

2.1. Materials

1,4-Diiodobenzene (p-C₆H₄I₂, Tokyo Chemical Industry, purity >98%),1,3-diiodobenzene (m-C₆H₄I₂, Tokyo Chemical Industry, purity >99%), 1,2-diiodobenzene (o-C₆H₄I₂, Tokyo Chemical Industry, purity >98%), 1,2,4,5-tetraiodobenzene (1,2,4,5-C₆H₂I₄, BLD Pharmatech, purity >95%), and 1,2,3,4,5,6-hexaiodobenzene (C₆I₆, Sigma-Aldrich, L162558) were purchased from commercial sources and used without further purification. A gold (Au) wire (Nilaco, 0.3 mm diameter, 99.95% purity) was mechanically cut to fabricate STM tips. Au(111) substrates were prepared by thermal evaporation of Au onto freshly cleaved mica, followed by annealing at 500 °C for 2 h. The Au(111) substrates were flame annealed and cleaned before use. For sample preparation, Au substrates were immersed in 1 mg/mL DMF solution of each molecule for at least 1 h. After immersion, the substrates were dried under an inert gas flow.

2.2. Conductance Measurements

The conductance measurements were performed by a commercially available ambient scanning tunnelling microscopy (STM) (MS-10 and Nanoscope V, Bruker) with a signal access module III (Bruker), an external piezo driver (M-2141, MES-TEK), and a data-acquisition-device with LabVIEW2016 (NI PXI-4461, National Instruments). In BJ method, an Au-tip was repeatedly moved in and out of contact with the Au(111) substrate in the presence of the molecules. After the breakage of the Au point contact, a small gap with nanosized electrodes was prepared between the Au tip and the Au substrate. A surface-deposited molecule can be trapped into the nanogap and a single-molecule junction (SMJ) can be fabricated using the BJ technique. The left panels of Figure show 2D conductance versus stretching distance histograms of the conductance traces of the SMJs, in which electronic currents of the SMJs are repeatedly measured during the stretching process of the junctions by applying a fixed bias voltage between the Au sample electrodes and the Au-STM tip (Figure S1). In the 2D histograms, the displacement was defined as zero at the point where the conductance dropped below 0.01 G ₀ for m- and o-C₆H₄I₂, and below 0.5 G ₀ for the other molecules, and all conductance traces were overlaid accordingly. As a result, the statistical distributions in the 2D histograms reflect the most probable conductance values and effective junction elongation of the SMJs. The 1D conductance histograms in Figure were constructed from the same data set used in the 2D histograms.

2.

Two-dimensional (2D) histograms of conductance vs junction-stretching distance traces and 1D conductance histograms for (a) p-C₆H₄I₂, (b) m-C₆H₄I₂, (c) o-C₆H₄I₂, (d) 1,2,4,5-C₆H₂I₄, and (e) C₆I₆. The bias voltage was 0.4 V for o-C₆H₄I₂, 0.25 V for m-C₆H₄I₂, and 0.1 V for the others. G ₀ denotes the conductance quantum (G ₀ = 2e ²/h). A linear bin size of 0.01 nm and a logarithmic bin size (Δlog­(G/G ₀)) of 0.01 are used. The histograms were constructed from 3000 measurements for o-C₆H₄I₂ and 10 000 measurements for the others. In the 1D conductance histograms, the distributions were fitted with a Gaussian function, except for o-C₆H₄I₂, and each Gaussian function is represented by blue lines. (f) Plots of the observed peak conductance values. The y-axis is shown on a logarithmic scale (left) and a linear scale (right). The error bars represent the fwhm.

2.3. Thermopower Measurements

In the thermopower measurements, the temperature of the STM tip was maintained at room temperature while that of the substrate was controlled by a Peltier device (CP0.8–31–06L, Laird Technologies) using a temperature controller (Model 331, Lake Shore Cryotronics). The substrate temperature was monitored with a resistive thermometer. Meanwhile, SMJs were made by the BJ method at a constant bias voltage of 5 mV. After forming the SMJ, the current–voltage (I–V) curve of the junction was measured by varying the tip bias voltage from 5 mV to – 5 mV at a constant electrode spacing, while applying a constant temperature difference to the junction (Figure S2). The thermoelectric voltage was determined as the bias voltage shift at zero current in the I–V curve with and without a temperature gradient in the molecular junction. Details of the experimental setup are described in previous reports. ,

2.4. Current–Voltage Characteristic Measurements

The I–V curves of the SMJs^26–28 were obtained according to the following procedure. An Au STM tip was brought into contact with an Au substrate until the formation of an Au point contact was confirmed by a conductance greater than 1 G ₀. The tip was then withdrawn at a speed of approximately 10 nm s^–1 to break the point contact, resulting in a nanoscale gap between the Au electrodes and enabling the stochastic formation of a SMJ. The current passing through the molecular junction was monitored at a fixed bias voltage of 100 mV. After the conductance decreased below 1 G ₀, the tip position was fixed and an I–V curve was recorded by sweeping the bias voltage from 100 to 2000 mV, then from 100 to −2000 mV, and finally back to 100 mV at a constant tip–sample separation. The SMJ was subsequently broken by retracting the tip from the substrate. To capture the structural variations of the junction, the making/breaking cycle was repeated so that a new junction configuration was formed prior to each I–V measurement. ,

2.5. DFT Calculations

The optimized geometries were obtained by DFT calculations at the B3LYP/6–31+G­(d) for C and H atoms and the B3LYP/SDD level for I atoms using the Gaussian 16 program (Figures ,).

4.

Calculated molecular orbital levels and spatial distributions for p-C₆H₄I₂ and C₆I₆. The HOMO of p-C₆H₄I₂ is dominated by π-orbitals, whereas that of C₆I₆ is dominated by σ-orbitals. Gray, white, and purple spheres represent C, H, and I atoms, respectively.

5.

Calculated HOMO levels of benzene with increasing iodine substitution. Red and black lines represent orbital levels with π- and σ-character, respectively. Spatial distributions are shown for the σ-character orbitals. As the degree of substitution increases, interactions among the peripheral iodine atoms lead to splitting of σ-type orbitals into bonding and antibonding states, as indicated by the green arrows. A similar upward shift of HOMO energy levels, attributed to σ-delocalization among adjacent selenium atoms, is experimentally observed in the series of selenium-substituted benzenes.

3. Results and Discussion

3.1. Single-Molecule Conductance of Iodine-Substituted Benzenes

Guided by the conceptual framework illustrated in Figure , we examine how the number and topology of iodine substituents regulate the balance between π-dominated and σ-involved transport pathways in single-molecule junctions. We first investigated the charge transport properties of iodine-substituted benzenes using the break junction (BJ) technique under ambient conditions (Figure and Figure S1). Figure a shows a representative two-dimensional (2D) plot of conductance versus stretching distance, together with a one-dimensional (1D) conductance histogram for para-diiodobenzene (p-C₆H₄I₂), both exhibiting a distinct distribution centered at 0.9 × 10^–4 G ₀ (log G/G ₀ ≈ – 4.0 ± 0.8). We note that no irreversible conductance changes or abrupt switching events were observed under the low-bias conditions employed in this study, indicating that Au–I coordination remains stable and that Au–C bond formation does not occur during the measurements. Although p-C₆H₄I₂ is the shortest molecule among those investigated, its conductance remains on the order of 10^–4 G ₀ rather than reaching higher values expected for such a short tunneling length. This behavior reflects the balance between the short molecular length, which favors higher conductance, and the Au–I molecule–electrode coupling, which is weaker than that of covalent anchoring groups such as Au–S and thus limits the overall conductance. The present value is somewhat reduced compared with the previous measurement for 1,4-diiodobenzene in solution (3.6 × 10^–4 G ₀). Nevertheless, it is still in reasonable agreement, which may partly reflect the different experimental environments (air vs solution), thereby confirming that the Au–I interaction enables stable junction formation. In contrast, meta-diiodobenzene (m-C₆H₄I₂) displays a significantly reduced conductance of ∼ 10^–6 G ₀ with log G/G ₀ ∼ −6.0 ± 0.4 (Figure b), which is over 2 orders of magnitude lower than that of the para isomer. Such pronounced suppression is consistent with destructive quantum interference (DQI) in π-orbital-mediated transport, a phenomenon well established for meta-connected π systems. − For the ortho-diiodobenzene (o-C₆H₄I₂), no junction formation is confirmed (Figure c). The geometric arrangement of the substituents reduces the effective molecular span between the iodine atoms, which may lead to ill-defined contacts and hinder the reliable formation of stable junctions and reproducible conductance. For the tetraiodobenzene (1,2,4,5-C₆H₂I₄), a slight increase in conductance to 1.5 × 10^–4 G ₀ (log G/G ₀ ≈ – 3.8 ± 0.8) is observed compared with p-C₆H₄I₂. (Figure d). Strikingly, the hexaiodobenzene (C₆I₆) junctions exhibit a higher conductance peak at 2.6 × 10^–4 G ₀ (log G/G ₀ ≈ – 3.6 ± 0.5, Figure e). It should be noted that the displacement axis in the conductance–displacement histograms represents an effective junction elongation, which includes contributions from electrode snapback and atomic rearrangements and does not correspond directly to the geometric molecular length. Within the width of the conductance distributions, a systematic upward trend is observed from p-C₆H₄I₂ to 1,2,4,5-C₆H₂I₄ and C₆I₆ (Figure f), suggesting that full peripheral iodine substitution is associated with enhanced charge transport. This enhancement cannot be rationalized solely by the contact geometry in π-orbital-mediated transport but rather suggests the emergence of a transport pathway consistent with σ-type orbital involvement. These results establish a clear structure–conductance relationship: para connectivity enables finite π-type conduction, meta connectivity enforces strong DQI, and full substitution yields conductance behavior consistent with σ-involved charge transport with enhanced conductance. Notably, the C₆I₆ junctions display a single conductance peak, similar to what was previously observed for hexa-selenium-substituted benzene. This behavior suggests that, in σ-involved systems, the overall conductance is likely to be relatively insensitive to which iodine sites are in contact with the Au electrodes, in contrast to π-mediated transport, where the conductance strongly depends on the specific contact geometry (Figure a,b).

3.2. Thermoelectric Properties

To further probe the nature of the conducting orbitals, we performed thermopower measurements ,,, of p-C₆H₄I₂. The distribution of thermoelectric voltages as a function of temperature difference yielded a positive thermopower of +4.0 ± 3.2 μV K^–1 (Figure a). The positive sign indicates that hole transport dominates, consistent with conduction through the HOMO, and confirms that the transport pathway in p-C₆H₄I₂ is HOMO-mediated. In addition, thermopower measurements of C₆I₆ revealed thermopower of +5.0 ± 1.2 μV K^–1 (Figure b,c and Figure S2). The thermopower values obtained for p-C₆H₄I₂ and C₆I₆ overlap within experimental uncertainty. However, the thermopower distribution of C₆I₆ is narrower and more symmetric, indicating a more well-defined transport channel. As thermopower measurements primarily reflect the sign and energetic proximity of the dominant transport orbital, these results are consistent with HOMO-mediated transport but do not, by themselves, distinguish between π- and σ-type orbitals. Taking the expanded experimental data set as a whole, we therefore conclude that C₆I₆ exhibits σ-involved HOMO-mediated transport, with this assignment supported by the combined evidence from substituent-dependent conductance trends, geometric considerations, and the molecular orbital analysis described below.

3.

(a) Distribution of thermopower measured at a temperature difference (ΔT) of 19.4 K for p-C₆H₄I₂, based on 1000 measurements. Thermopower S was calculated as S = −V _th/ΔT, where V _th and ΔT are the thermoelectric voltage and temperature difference across the molecular junction, respectively. Because the junction formation probability of p-C₆H₄I₂ is relatively low under a temperature gradient, data collection was limited, and the thermopower was determined from thermoelectric voltage measurements at a single ΔT. The distribution of S was fitted with a Gaussian function (blue line), yielding a peak value of 4.0 ± 3.2 μV K^–1. (b) Distributions of V _th for C₆I₆ measured at different ΔT values, each based on 5000 measurements (left). Peak values were obtained by fitting the distributions with Gaussian functions. Plot of peak V _th values vs ΔT for C₆I₆ (right). (c) Distribution of thermopower for C₆I₆ at ΔT = 25.5 K, based on 5000 measurements. The S value was calculated as S = −V _th/ΔT, with a peak value of 5.0 ± 1.2 μV K^–1.

3.3. Molecular Orbital Analysis

Density functional theory (DFT) calculations (Figures and ) provide deeper insight into the origin of the observed conductance trends. For p-C₆H₄I₂, the HOMO is a π-type orbital on the benzene ring, whereas in C₆I₆ it transforms into a σ-type orbital distributed over the peripheral iodine atoms (Figure ). Within the single-level tunneling model, charge transport is governed by the frontier orbital energy ε relative to the electrode Fermi level and by the electronic coupling to electrodes Γ. At the zero bias limit, the transmission τ at E _F is expressed as τ = Γ²/(ε ² + Γ²). With increasing iodine substitution, the electron-withdrawing effect pushes the π-orbital levels to lower energies (larger |ε|), thereby suppressing π-mediated transport. At the same time, interactions among the iodine lone pairs generate σ-type orbitals, among which the frontier orbital approaches E _F, thereby reducing ε for σ-mediated transport and enhancing their contribution (Figure ). In C₆I₆, dense packing of the substituents and steric congestion maximize this effect: ε for the σ-orbital becomes particularly small (see the upper right corner of Figure ), while Γ remains robust owing to well-defined Au–I contacts. As a result, the transmission at E _F is enhanced, consistent with the experimentally observed conductance of 2.6 × 10^–4 G ₀. Thus, the overall conductance reflects a quantitative trade-off between diminished π-transport and emergent σ-transport, with the high conductance of C₆I₆ providing evidence for an increased contribution from σ-type orbitals in iodine-substituted molecular junctions. Here, the gas-phase DFT calculations are used solely to illustrate qualitative trends in orbital character and energy ordering, and are not intended to describe quantitative level alignment in the junction, which may be significantly modified by molecule–electrode interactions.

3.4. Molecular Orbital Level

To directly evaluate the energy alignment of the dominant transport orbital (HOMO) of C₆I₆, we measured the current–voltage (I–V) characteristics of the C₆I₆ junctions and analyzed them within the framework of the single-level tunneling model. Figure a shows a 2D histogram of the measured I–V curves. Each I–V curve was fitted using the following eq (Figure b), which describes the current through a single-molecule junction, where Γ denotes the metal–molecule electronic coupling and ε represents the energy of the molecular level relative to the electrode Fermi level. ,

$$I(V)=\frac{2e}{h}\mathrm{\Gamma }[{\tan }^{-1}\left(\frac{eV/2-\epsilon }{\mathrm{\Gamma }}\right)+{\tan }^{-1}\left(\frac{eV/2+\epsilon }{\mathrm{\Gamma }}\right)]$$ 1

6.

(a) 2D map of the I–V characteristics of C₆I₆ junctions, constructed from 20 000 experimentally obtained I–V curves. The bin sizes along the x- and y-axes are 20 mV and 1 nA, respectively. (b) Example of the curve fitting. The solid and dotted lines represent the experimental and fitted results, respectively, with the following parameters: ε ≈ 0.96 eV and Γ ≈ 9 meV. (c and d) 1D histograms of ε and Γ, respectively, obtained from curve fitting. The solid and dotted lines represent the fitted results and each Gaussian component, respectively, with the following peak values: ε ≈ 0.93 and 1.15 eV and Γ ≈ 16 meV (Figure S3).

The model is used here as a convenient phenomenological description of off-resonant transport; although it does not capture the full multiorbital complexity, it provides a consistent framework to extract effective ε and Γ values. In this study, the single-level tunneling model is employed solely as a phenomenological framework to extract effective level alignment parameters and is not intended to describe transport mechanisms involving quantum interference; accordingly, it is not applied to meta-connected molecules in this study. To determine the ε and Γ values corresponding to the junctions formed in the conductance measurements at 0.1 V (log G/G ₀ = – 3.6 ± 0.5, Figure e), the I–V curves exhibiting conductance within this range were selected from the full set of measured curves to construct the histograms. As shown in Figure c,d, the 1D distributions of the fitted parameters yields ε ≈ 0.93 eV and Γ ≈ 16 meV. Another peak at ε ≈ 1.15 eV can be attributed to junction configurations with ill-defined metal–molecule contacts, likely induced by the application of high bias voltages (up to ± 2 V) in the I–V measurements.

Using the gold work function of 5.3 eV, the corresponding HOMO energy of the C₆I₆ junction is estimated to be – 6.23 eV, which lies slightly closer to the Fermi level than the calculated gas-phase HOMO energy of – 6.28 eV (Figure ). Although orbital energies from DFT calculations on the isolated molecule cannot be directly compared with experimental level alignment, this qualitative proximity supports the internal consistency of our analysis without implying quantitative agreement.

Within the single level tunneling model, thermopower S of a single-molecule junction is described by the following equation.

$$S=\frac{{\pi }^2{k_{\mathrm{B}}}^{2}T}{3e}{\frac{\partial \ln \tau (E)}{\partial E}|}_{E=E_{\mathrm{F}}}=\frac{2{\pi }^2{k_{\mathrm{B}}}^{2}T}{3e}\frac{\epsilon }{{\epsilon }^2+{\mathrm{\Gamma }}^2}$$ 2

Using the experimentally determined values of ε and Γ yields S = 16 μV K^–1 of the C₆I₆ junction, which is close to the upper end of the experimentally observed thermopower distribution (Figure c). Thermopower is highly sensitive to experimental factors, particularly the determination of the local temperature and the temperature difference across the junction. Because thermoelectric voltage is directly proportional to ΔT, such uncertainties can contribute to noticeable deviations between the calculated and measured values. Consequently, the calculated thermopower (16 μV K^–1) shows only modest agreement with the measured value (∼5 μV K^–1).

We note that a full microscopic identification of the transport channels would ultimately benefit from NEGF–DFT calculations of the transmission functions. However, for Au–I junctions, such calculations require extensive sampling of metal–molecule binding geometries, which is computationally demanding and beyond the scope of the present experimentally focused study. Instead, the present analysis relies on a self-consistent interpretation of experimentally accessible observables, including substituent-dependent conductance trends, junction stability, thermopower sign, and level alignment extracted from current–voltage characteristics. Within these experimentally grounded constraints, the observed transport behavior of C₆I₆ is most consistently explained by an increased contribution from σ-involved orbitals.

3.5. Interplay of Anchoring and σ-Delocalization

The formation of σ-delocalized orbitals among peripheral iodine atoms is consistent with previous UPS studies of polyiodobenzenes, which revealed pronounced through-space I···I interactions. An important aspect of iodine substitution is its dual role as both an anchoring group to electrodes and a σ-involved contributor. Previous studies have demonstrated that Au–I halogen bonding is stronger than Au–NH₂ coordination, which accounts for the robust junction formation of p-C₆H₄I₂, 1,2,4,5-C₆H₂I₄, and C₆I₆. In C₆I₆, the cooperative action of Au–I contacts together with σ-type orbital contributions offers a plausible explanation for its enhanced conductance relative to the para isomer. This dual role of iodine substitutionstabilizing the metal–molecule contacts while enabling transport behavior consistent with σ-involved pathwayshighlights its potential as a versatile design element for single-molecule electronics. Together with the thermopower and I–V analysis, these conductance trends reinforce our conclusion that C₆I₆ supports σ-involved transport only when a complete peripheral iodine ring is present. We emphasize that the pronounced σ-channel transport discussed here is observed most clearly in fully substituted benzene derivatives such as C₆I₆, and that extending this behavior to more general molecular systems will require careful molecular design and further experimental validation.

4. Conclusions

This study demonstrates how the number and topology of nonbonded halogen substituents regulate the balance between π-dominated and σ-involved transport pathways in iodine-substituted benzene junctions. Break-junction measurements across five systematically varied molecules reveal a clear structure–transport relationship: para-diiodobenzene exhibits conventional π-HOMO transport, the meta isomer shows destructive quantum interference, and the ortho isomer fails to form stable junctions because the adjacent iodine substituents yield a very short and geometrically constrained effective junction length that prevents the molecule from adopting a reproducible bridging configuration. The tetraiodo derivative displays only modest conductance enhancement, indicating that partial iodine substitution does not generate a continuous σ-framework. In contrast, hexaiodobenzene exhibits a single, narrow, and contact-insensitive conductance peak accompanied by positive thermopower, signaling robust HOMO-mediated transport.

While the absolute values of conductance and thermopower overlap within experimental distributions, the combined substituent-dependent trends, junction stability, and level-alignment analysis consistently indicate an increased contribution from σ-type orbitals in C₆I₆. Combined with molecular-orbital calculations and level alignment extracted from I–V analysis, these observations lead us to conclude that C₆I₆ exhibits σ-involved transport pathways that emerge only when a full peripheral ring of nonbonded iodine atoms is present. Together, these findings identify a minimal structural requirement, namely a continuous halogen periphery, for accessing transport behavior consistent with σ-involved pathways in aromatic systems. Although NEGF–DFT transport calculations were beyond the scope of this experimentally focused study, the clear substituent-dependent trends revealed here provide a solid foundation for future theoretical work. Such calculations, particularly those that sample multiple Au–I binding geometries, will be valuable for quantifying the σ-involved contributions identified experimentally. By revealing a substituent-number-controlled transition from π-dominated to σ-involved transport in a chemically simple series, this work offers a concise and general design principle for engineering nonbonded σ-delocalized channels that complement traditional π-based molecular conductors. Future transport calculations that explicitly evaluate transmission eigenchannels for realistic Au–I junction geometries will be valuable for further quantifying the σ-involved contributions identified experimentally in this work.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.5c24444.

• Representative conductance traces, conductance histograms, time course of thermoelectric voltage measurements, and additional I–V analysis (PDF)

The authors declare no competing financial interest.