Achieving High-Performance Molecular Rectification through Fast Screening Alkanethiol Carboxylate-Metal Complexes Electroactive Units

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLES19 May 2022Achieving High-Performance Molecular Rectification through Fast Screening Alkanethiol Carboxylate-Metal Complexes Electroactive Units Lixian Tian†, Aiqing Fan†, Xi Yu and Wenping Hu Lixian Tian† Tianjin Key Laboratory of Molecular Optoelectronic Science, School of Science, Tianjin University, Tianjin 300072 †L. Tian and A. Fan contributed equally to this work.Google Scholar More articles by this author , Aiqing Fan† Tianjin Key Laboratory of Molecular Optoelectronic Science, School of Science, Tianjin University, Tianjin 300072 †L. Tian and A. Fan contributed equally to this work.Google Scholar More articles by this author , Xi Yu *Corresponding author: E-mail Address: [email protected] Tianjin Key Laboratory of Molecular Optoelectronic Science, School of Science, Tianjin University, Tianjin 300072 Google Scholar More articles by this author and Wenping Hu Tianjin Key Laboratory of Molecular Optoelectronic Science, School of Science, Tianjin University, Tianjin 300072 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.022.202201890 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Achieving high performance in molecular scale diode devices remains a formidable challenge due to the complexity of the charge transport process and the difficulty in device structure modulation. Herein, we demonstrated a low-cost and fast screening strategy to search for proper electroactive units and construct high performance self-assembled monolayer (SAM)-based molecular diodes. The strategy is based on off-the-shelf carboxylate-terminated alkane thiols and simple carboxylate-metal ion complexes, structures denoted as Au-S-(CH2)n-1COO−Mm+ (Cn+Mm+), where n = 11, 12, 13, 14, 16, 18, and Mm+=Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+, as the library of functional SAM layers on gold substrate. Combing the fast screening measurements using a eutectic indium-gallium alloy top contact (EGaIn), C18+Ca2+ and C18+Zn2+ structures were found to afford record high rectification ratio (RR) >700 at ±1.5 V. Theoretical analysis based on a single level tunneling model shows that the C18+Ca2+ and C18+Zn2+ devices possessed an optimized combination of asymmetric voltage division, energy offset, and coupling of carboxylate-metal complexes with the electrode. This newly developed method represents a general strategy for fast, inexpensive, and effective exploration of the functional metal complex chemical space, and can largely accelerate the development of practical high performance molecular diode devices. Download figure Download PowerPoint Introduction The purpose of molecular electronics is to construct circuits using single or molecular assemblies, and to prepare functional molecular electronic devices with electrical properties determined by the chemical and supramolecular properties of molecules.1–9 Over the past few decades, a variety of functional molecular devices have been theoretically designed and experimentally prepared, such as molecular switches,10–23 molecular wires,24–27 negative differential resistance,28–30 molecular transistors,31,32 sensors,33–35 and molecular diodes.36–40 The diode is a device that conducts electricity under a forward bias and blocks under a reverse bias, and is one of the fundamental elements in the circuit. Since Aviram and Ratner first proposed the concept of a molecular diode composed of donor-bridge-acceptor (D-bridge-A) molecule in 1974,41 many molecular diodes devices have been studied39,42–49 as summarized by Metzger.43 Since the 2010s, molecular diodes based on asymmetric single electronic states40,50–52 (Figure 1a) have emerged and been extensively developed, mainly by Whitesides and Nijhuis et al.,53,54 and importantly, a higher rectification ratio (RR) was achieved. This type of molecular rectifier generally utilizes electroactive moieties, such as ferrocene,50,51,54,55 bipyridyl,40,56 various polycyclic-aromatic-hydrocarbons,57 and so on, to provide active electronic states and inert alkanes as spacers to position the electronic state symmetrically in the junction. The underlying working principle of the rectification device has been appropriately demonstrated by theoretical studies using either ab initio calculations based on density functional theory (DFT) and Green’s function,58,59 or single level transmission models.60–62 As depicted by the single level model, in its simplest physical picture, the current across the junction is given by the Landauer–Büttiker formula,63 namely, I = 2 e h ∫ d E [ f L ( E ) − f R ( E ) ] T r ( E ) (1)where e is the electron charge and h is Planck’s constant. f L and f L are the Fermi functions of the left and right electrodes, respectively, given by f L , R ( E , V ) = 1 exp ( E − e V − μ L , R k B T ) + 1 (2)where μ L , R is the chemical potential of the left and right electrode. T r ( E ) is the transmission function, which is related to the coupling strength Γ L , Γ R , the energy offset (or energy barrier) ε , and the voltage division factor or asymmetric factor α , which dictates the shift of electronic state of the single electronic state in the junction due to the polarization of the state by the electric field in the junction with applied bias and is the key for rectification, as shown in following eq. 3 T r ( E , V , Γ L Γ R ) = Γ L Γ R ( E − ε − α e V ) 2 + [ Γ ( Γ L , Γ R ) 2 ] 2 (3) Figure 1 | Schematic diagram of the asymmetric molecular junction (a) and I-V characteristics based on the single level model (b). Here, we show the case where the energy of the electronic state is higher than the Fermi level of the electrode and its position is closer to the top electrode. The energy state can enter the bias window at the positive bias, and the current rapidly increases. While at a negative bias, the energy level is outside the bias window so the current is low. (c) demonstrates the dependence of the rectification direction on the energy offset (or energy barrier) ε, and the voltage division factor or voltage division factor α. When the energy of the electronic state is higher than the Fermi level of the electrode, and α is less than 0.5 (i.e., closer to the top electrode), the rectification direction is defined as positive. On the contrary, when the energy of electronic state is lower than the Fermi level of the electrode, an opposite dependence of the rectification direction on α can be expected (for details, see the Supporting Information Figure S21 and ref 61). (d) Schematic illustration of the molecular junction based on the metal-organic coordinated monolayer of Au-Cn+Mm+-GaOx/EGaIn, where n = 11, 12, 13, 14, 16, 18, Mm+ = Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+. Download figure Download PowerPoint With this model, rectification is realized when the transmission peak enters the bias window under one bias over the other polarity under the condition of the position of the electronic state in the junction, so that the polarization and voltage division factor α in the transmission function become asymmetric, as shown in Figure 1b. The direction of the rectification depends on the joined effect ε and α , as shown in Figure 1c (for details, see the Supporting Information and ref 61). Over the past decade, by various electroactive groups,40,57,64,65 asymetricity,51 and molecule connection and orientation,55 the rectification performance of single state-based devices has increased from several tens to 103,50,51,64–68 and even reached 105,54 comparable to the practical inorganic p–n junction molecular diodes. Furthermore, recent studies have showed that the RR can be modulated by the molecular mix,69–72 the extent of external applied voltage,73 internal orbital gating,74 Fermi level pinning,56 and embedded dipole.75 However, there remains a large chemical space to be explored for inexpensive, reliable, and high performance molecular rectification. In our recent model study, we extensively studied the performance of the molecular rectifier depending on the coupling strength Γ, energy barrier ε , and asymmetric factor α (Figure 1), and it was shown that the performance optimization of the molecular diode needed delicate control of all three parameters.61,62 The theoretical studies imply that, in principle, there is a vast space for rectification performance modulation defined by chemical and device structures. In addition, the interface between the top electrode, mostly the GaIn/GaOx, and the molecule is ill-defined due to the unknown composition and structure of GaOx,62,76–78 which further blurred the design path. Therefore, performance optimization of the molecular diode through trial and error by synthetic chemistry and device construction is inevitably a highly time consuming and resource-intensive process, which largely hinders the development of the molecular diode. Therefore, a method that can significantly increase the trial and error efficiency and decrease the cost of exploring the chemical space of functional units for molecular rectification junctions is highly desired. In this work, by adopting the idea of in-situ metal ion complexes for molecular junctions79,80 we have developed a fast screening strategy to efficiently explore the junction structure space to achieve high performance molecular rectification. Off-the-shelf carboxylic acid-terminated n-mercaptoalkanoic acid (n = 11, 12, 13, 14, 16, 18) self-assembled monolayers (SAMs) coordinated with various metal ions (Mm+=Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+) were used as a library of functional units with different coupling strength Γ, energy barrier ε , and asymmetric factor α . Combing through the fast screening tests using an EGaIn top electrode, Ca2+ and Zn2+ were filtered out as good electroactive groups for rectification devices, of which the RR can reach record highs up to ∼750 at ±1.5 V in conjunction with a C18 linker. Experimental Methods Materials and reagents Silicon (100) wafers with a 300 nm SiO2 layer were purchased from China Electronics Technology Group Corporation (Tianjin, China) 46th Research Institute (CETC 46). Au (99.999%) and Cr (99.9%) pellets were purchased from Zhongnuo Advanced Material (Beijing, China) Technology Co., Ltd. The n-mercaptoalkanoic acid HSCnO2H (n = 11, 12, 16, C11, C12, C16, 98%, 99%, 99%, respectively) were purchased from Sigma-Aldrich (Shanghai, China). 14-Mercaptotetradecanoic acid (C14, 95%) was purchased from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). 13 and 18-Mercaptooctadecanoic acid (C13, 95%; C18, 98%) were purchased from Shanghai Synchem Technolgy Co., Ltd. (Shanghai, China). Copper(II) perchlorate hexahydrate (Cu(ClO4)2·6H2O, 98%) and zinc perchlorate hexahydrate (Zn(ClO4)2·6H2O, 99.995%) were purchased from Sigma-Aldrich (Shanghai, China). Calcium chloride anhydrous (CaCl2, 99.99%), manganese(II) chloride (MnCl2, 99.99%), iron(II) chloride (FeCl2, 99.5%), iron(III) chloride (FeCl3, ≥99.9%), cobalt(II) chloride hexahydrate (CoCl2·6H2O, 99.99%), nickel chloride hexahydrate (NiCl2·6H2O, 99.9%) and anhydrous ethanol (high-performance liquid chromatography/HPLC) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All reagents were used as provided without further purification. Preparation of the SAMs of alkanethiol carboxylate-metal complexes A silicon (100) wafer with a native SiO2 layer was thermally evaporated with 5 nm chromium and a 60 nm gold at rates of 0.6 and 1.2 Å/s under a pressure of 10−6 mbar. Then the gold substrates were cut into ∼1.0 cm × 1.0 cm squares. The roughness of the gold surface is ∼1 nm,81 larger than that of template-stripped gold (0.4 nm).54 SAMs of n-mercaptoalkanoic acid on Au were prepared as reported previously.62 The freshly evaporated gold substrate was incubated for 16–18 h in 10 mL of 20 μM mercaptoalkanoic acid solution in degassed absolute ethanol followed by the addition of 2 μL of 1 M HCl at room temperature and N2 atmosphere. Then the sample was removed from solution, rinsed with ethanol to remove physiosorbed molecules, and then dried under a N2 stream. For the preparation of carboxylate-metal complex SAMs (COOM), the samples were then immersed in 5 mM ethanolic solutions containing different inorganic salts (CaCl2, MnCl2, FeCl2, FeCl3, CoCl2·6H2O, NiCl2·6H2O, Cu(ClO4)2·6H2O, and Zn(ClO4)2·6H2O) for 30 min and subsequently rinsed with ethanol and dried with N2. Results and Discussion The structural characterization of C11+M m+ SAMs We first prepared and characterized the 11-mercaptoundecanoic acid (C11-COOH) SAM on a vacuum-deposited gold surface, and various metal ions were then coordinated to the SAM surface (C11+Mm+; Mm+ = Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+). Ellipsometry measurements found a slightly increased thickness (2∼3 Å) of the SAM after metal ion binding, compared to 11-mercaptoalkanoate SAM (10.4 ± 1 Å), as shown in Figure 2a, due to the difference in refractive index between the organic film and metal ion. Figure 2 | The thickness (a) and IRRA spectra (b) of 11-mercaptoundecanoic acid SAM on Au before and after incubation with Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+. In panel (b), the left shading area covers the peaks of the stretching modes of the methylene group (CH2). The dotted line indicates that the peak position did not change after metal ion chelation, implying the ordered SAM was well maintained. The right shading area covers the peak of the carboxylate (COO−) formed after cheating with metal ions. Download figure Download PowerPoint Surface reflective infrared spectroscopy was performed to further characterize the order and coordination information between metal ions and 11-mercaptoundecanoic acid SAM. Figure 2b is the spectra of 11-mercaptoundecanoic acid SAM before and after metal ion adsorption. The peak position and band assignments are given in Supporting Information Table S1. The carboxylic acid group peaks at 1741 and 1714 cm−1, attributed to the asymmetric non-hydrogen-bonded and hydrogen-bonded carbonyl group stretching vibration,62,82–85 respectively, disappeared after metal ion adsorption. Meanwhile, peaks at ∼1548 and ∼1446 cm−1 appear, as shown in the red shaded part in Figure 2b, which can be assigned to the asymmetric (νa(COO−)) and symmetric (νs(COO−)) stretching vibrations of carboxylate,62 respectively, implying the formation of metal carboxylate complexes, in agreement with previous reports.83–87 The –COOH peak completely disappeared (1741, 1714 cm−1) in the metal complex-terminated C11 SAM (C11+Mm+) as a carboxylate group appeared (1548.6, 1446.4 cm−1), indicating that the carboxylic acid functional group is fully deprotonated and coordinated completely to the metal ion. Moreover, the peaks at 2921 and 2850 cm−1, which are assigned to CH2 asymmetric (νa(CH2)) and symmetric stretching modes (νs(CH2))88 of the alkane group in the SAM, respectively, indicate good packing order of the SAM.82,88–90 After adsorption of the metal ion solution, we found that the νa(CH2) stretching peak did not change compared with C11 SAM (Figure 2b, as shown by the dotted line and Supporting Information Table S1). This implies that the SAM kept its packing order after metal ion chelation, which secured the well-defined structure of the metal ion in the junction. X-ray photoelectron spectroscopy (XPS) was performed to characterize the elemental composition of the metal-complexed SAM ( Supporting Information Figure S1). We observed the formation of the metal complex at the SAM surface with an O:M ratio of 4:1, except for Ca2+ and Zn2+ (2:1), see Supporting Information Figure S2 and Table S2. Also, the absence of counter ion Cl− in the complexed SAM ( Supporting Information Figure S6a) also indicates the formation of the metal ion carboxylate complex, see the Supporting Information for details. As discussed above, due to the carboxylic acid functional group being fully coordinated to the metal ion for the carboxylate complex SAM, the difference in O:M ratio between (Ca/Zn) and other metal ions obtained by XPS should be attributed to the difference in coordination type. The coordination type of the Ca2+- and Zn2+-carboxylate complex is ionic-chelating bidentate, and the other coordination type of carboxylate complexes is bridging monodentate, see the Supporting Information for more detail. Rectification behavior defined by the metal ion complexes We further constructed the molecular junction (Figure 1d) by applying EGaIn on top of the C11+Mm+ SAM as the top electrode. EGaIn has been well recognized as a stable and harmless soft contact for SAM,91–93 it is convenient to fabricate and operate, and affords quick and reproducible high yield working junctions, allowing statistically significant data to be collected.91,94–97 Our recent work shows that carboxylate alkali metal complexes can help to increase RR, compared with a COOH-terminated SAM due to the carboxylic acid (COOH)–GaOx/GaIn interface.62 The C11+Mm+ SAM junctions of eight types of metal complexes were measured, and the I-V response and rectification performance was summarized in Figure 3. The yield of working junction is >88% ( Supporting Information Table S3). We define the RR (|RR|≥1) as the current density J ratio at positive and negative bias voltages, where positive RR+ = +|J(+V)|/|J(−V)| and negative RR− = −|J(−V)|/|J(+V)| at ±1.5 V or ±1 V, and the value and positive/negative sign indicate the magnitude and polarity of the rectification. Figure 3 | Transport behavior and rectification performance of the 11-mercaptoalkanoate-metal complex monolayer junction (C11+Mm+). (a) The averaged log |J|-V traces of the C11+Mm+ junctions. The J-V traces were shifted vertically for a clearer view, while keeping the same scale. The vertical bar represents one order of magnitude. (b) RR of the junctions. RR of Cu2+ and Fe2+ was calculated at ±1.0 V because they were easy to get short above +1.0 V. Download figure Download PowerPoint The averaged log J-V traces of the C11-COOH and C11+Mm+ junctions are shown in Figure 3a and Supporting Information Figure S4. Supporting Information Figure S3 provides a statistical heat map of J-V plots. It can be seen that various C11+Mm+ junctions exhibited different rectification behaviors (Figure 3). Cu2+, Mn2+, Co2+, and Ni2+ junctions have very low positive RR (<5), while Fe2+ and Fe3+ exhibit lightly negative RR (−2 and −13), respectively. In contrast, the Zn2+ and Ca2+ junctions afford much higher positive RR (more than 40), see Figure 3b and Supporting Information Figure S5. The rectification directions of Ca2+- and Zn2+-complexed SAM junctions along with all the C11+Mm+ junctions of positive RR are the same as C11-COOH, C11+Na+, and C11+K+ SAM junctions,62,98 which is opposite to that for Fc-based junctions.64,99 Considering the rectification polarity and special asymmetricity (α < 0.5) of the junction defined by the alkane molecule spacer, it can be implied that the lowest unoccupied molecular orbital (LUMO) energy level in C11-COOH and complexed SAM junction, that is, energy level higher than the Fermi energy of the electrode pair, dominates the charge transport. When a positive bias is applied, the energy level of the metal ion atomic orbital enters the bias window, causing the current to increase rapidly (resonant tunneling), thereby causing rectification (as shown in Figure 1b). On the contrary, an electronic state of energy lower than the Fermi level of the electrode, which is often the highest occupied molecular orbital (HOMO), should dominate the charge transport process in Fe2+ and Fe3+ junctions, and is responsible for the negative RR. For Cu2+, Mn2+, Co2+, Fe2+, and Ni2+, it is possible that the dominating electronic state is very close to the Fermi level of the electrode so that the rectification becomes less significant, as shown in our model study.61 The above arguments can be further supported by the ultraviolet photoelectron spectroscopy (UPS) measurement on the SAMs ( Supporting Information Figures S12 and S13), which shows that Ca2+ and Zn2+ have the highest HOMO barrier offsets, which implies that LUMO-dominated transport is more favorable, while Fe2+, Fe3+, and Cu2+ have the lowest HOMO barrier offsets, preferring HOMO-dominated transport and low rectification. On the other hand, the above arguments can be partially supported by the DFT calculations on the electronic structures of carboxylate-metal ion complexes in the gold metal junction,100 where Mn2+, Ni2+, and Fe2+ were found to possess energy states, both LUMO and HOMO, close to the Fermi level of the gold electrodes, accounting for their low rectification behavior. It is also important to point out that we believe the coordination type (detailed analysis in the Supporting Information) in carboxylate-metal complexes is not the key to the rectification behavior ( Supporting Information Figures S18 and S19). As shown in Table 1, Zn and Ca have larger surface concentrations with metal/oxygen ratios of 1:2, with electrostatic interaction and chelating bidentate. In contrast, the surface concentrations of other metal complexes are half that of the Zn and Ca complex, that is, a metal/oxygen ratio of 1:4, and the coordination type is bridging monodentate, which surely reduces the amount of metal complexes, and so the number of electronic active channels in the junction. However, the low concentration of the metal, other than Zn and Ca, on the surface will only reduce the overall current, but not the RR. As we have discussed before, the RR of the single state-based rectification device is jointly affected by the voltage division factor α , the coupling Γ, and the energy offset ε of the electronic state with respect to the electrode. Considering that the voltage division factor α is mainly determined by the spatial position of the electronic state in the junction, we can assume that α is the same for all C11+Mm+ junctions. Therefore, the high performance of Ca2+ and Zn2+ should be attributed to the appropriate energy offset and coupling with the top GaOx/GaIn electrode. Table 1 | Summary of the Characterization Results of C11+Mm+ (Mm+ = Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, and Zn2+) SAMs on Gold, Including Elemental Ratios of Oxygen and Metal by XPS, HOMO-Fermi Energy Difference (Energy Barrier) by UPS, Symmetric and Asymmetric Stretching Peak and the Difference (Δ Value) of COO− Groups by IRRAS, and Coordination Type Deduced from XPS and IR Results SAM O:Metala HOMO-Fermi barrier (eV)b νa(COO−) (cm−1) νs(COO−) (cm−1) Δ = νa(COO−) − νs(COO−) (cm−1) Coordination type C11+Ca2+ 1.8 2.14 1548.6 1446.2 102.4 Ionic C11+Mn2+ 5.2 2.14 1548.6 1446.4 102.2 Bridging monodentate C11+Fe3+ 3.4 2.30 1548.6 1446.4 102.2 Bridging monodentate C11+Fe2+ 3.9 2.29 1548.6 1446.4 102.2 Bridging monodentate C11+Co2+ 4.0 2.22 1548.6 1446.2 102.4 Bridging monodentate C11+Ni2+ 5.1 2.22 1546.7 1463.8 82.9 Bridging monodentate C11+Cu2+ 4.5 2.03 1548.6 1444.5 104.1 Bridging monodentate C11+Zn2+ 3.6 2.12 1550.6 1463.8 86.8 Chelating bidentate aObtained from XPS results. bObtained from UPS results. We would like to mention there is still no way, as far as we know, to obtain the actual electronic structure of the junction. The electronic structure obtained by UPS of the SAMs will necessarily change when the GaIn top electrode is applied, by the renormalization effect101 and possible charge transfer between metal ion and top electrode. Therefore, we have tried to prevent over-interpretation of the UPS. On the other hand, DFT calculations102 are not sophisticated enough for the ill-defined GaOx structure. Therefore, a complete interpretation of the electronic structure and control of charge transport in these junctions is still possible. However, we can see that the advantage of our strategy, as mentioned before, is that we can find good candidate electroactive units for high performance molecular rectifications by bypassing all the difficulties in designing and controlling the energy barrier and coupling at the interface, by a fast and low-cost screening strategy without detailed electronic structure information for this ill-defined system. Asymmetry can increase the RR as high as 750 at ±1.5 V in the C18+Ca2+/Zn2+ junction Next, we try to further increase the performance of the molecular diode by introducing more asymmetry on the voltage division factor α . This was done by applying a longer linker between bottom electrode and carboxylate-metal, that is, we constructed the SAM of Cn+Ca2+/Zn2+ where n = 11, 12, 13, 14, 16, 18. The SAM thickness and structures were extensively characterized by ellipsometry, infrared reflection–absorption spectroscopy (IRRAS), and XPS ( Supporting Information Figures S7–S9). IRRAS shows that C18-COOH SAM is more ordered than C11-COOH due to the lower position of CH2 antisymmetric stretching (νa(CH2)) (2917.9 vs 2919.8) ( Supporting Information Table S1). Figure 4 shows the J-V traces and rectification performance of the Cn+Ca2+/Zn2+ junctions ( Supporting Information Figure S10 provides the heat maps of all the J-V traces). We can find that the off-state current density J at negative bias decreases gradually as the chain length increases, and remains almost constant at positive bias, which leads to an increase of RR (Figures 4a and 4b and Supporting Information Figure S11). Meanwhile, it can also be found that there is a sharp increase in J value at on-state voltage around 1 V (arrow in Figures 4a and 4b), which can be recognized as the sign that the available electronic state enters the bias window. While at the negative bias range, the electronic state is always outside the bias window. At the same time, the RR+ increases as the voltage increases (Figures 4c and 4d). We summarized the RR of Cn+Ca2+/Zn2+ at ±1.5 V as shown in Figures 4e and 4f. As the number of carbon atoms n increases from 11 to 18, the RR increases from 45 to 573 and 756 for Ca2+ and Zn2+, respectively. Table 2 summarizes the RR larger than 50 in single level molecular junction fabricated with the EGaIn top electrode, and the C18+Zn2+ junction in this work is one of the best molecular diodes. Figure 4 | Heat maps and average log |J|-V curves (white) (a, b), RR(V) (c, d) and RR at ±1.5 V (e, f) for n-mercaptoalkanoate-Ca2+/Zn2+ complex monolayer junctions (Cn+Ca2+/Zn2+, n = 11, 12, 13, 14, 16, 18). The error bars represent one standard deviation. Download figure Download PowerPoint Table 2 | Summary of RR Values (>50) Based on Monolayer Molecular Junctions Using EGaIn as the Top Electrode Bottom Electrode SAM Structure Voltage (V) RRa References AuTS HSC9Fc 1.0 −1.7 103 AuTS HSC11Fc 1.0 −3.1 103 AuTS HSC13Fc 1.0 −1.7 103 AgTS HSC9Fc 1.0 −120 66, 103 AgTS HSC11Fc 1.0 ∼−100 53, 55, 64, 66, 99, 103–106 AgTS HSC13Fc 1.0 −81 51, 103 AgTS HSC10NHC=OFc 1.0 −90 55 AgTS HSC11FcC2 1.0 −77.4 51 AgTS HSC11Fc2 1.0 −500 53, 99, 105 AgTS HSC11Fc=Fcα 0.875 −1100 50 AgTS HSC11Fc-Fc 1.125 −610 50 AgTS HSC15Fc-C≡C-Fc 1.0 −1000 54 AgTS HSC15Fc-C≡C-Fc 2.0 −794 54 PtTS HSC15Fc-C≡C-Fc 1.0 −398 54 PtTS HSC15Fc-C≡C-Fc 3.0 −6×105 54 AuTS HSC15Fc-C≡C-Fc 1.0 −251 54 AuTS HSC15Fc-C≡C-Fc 3.0 −1 × 104 54 AgTS HSC11BIPY 1.0 +85 40 AuTS HSC11BIPY 1.0 +40 40 AuTS HSC11BIPY-MCl2(M=Mn, Fe, Co) 1.0 ∼+80 79 AuTS HSC11S-BTTFBTTF 1.5 +83 65 PtTS HSC11S-BTTFBTTF 2.5 −912 65 AgTS HSC11 PAH 0.741.0 ∼−170∼−150 57 Note: TS, template stripping; BTTF, benzotetrathiafulvalene; PAH, polycyclic-aromatic-hydrocarbon. aRR+ = +|J(+V)|/|J(−V)| and RR− = −|J(−V)|/|J(+V)|. We also studied the effect of the linker on other metal ions carboxylate complex (shown in Supporting Information Figures S14–S17) and observed a similar but smaller increase in RR in the C18 complex, except for Fe2+, Fe3+, and Cu2+, where the RR increased from 4 to 19, 3.2 to 131, 3.5 to 28 for Mn2+, Co2+, and Ni2+, respectively ( Supporting Information Figure S16b). The mechanism of rectification We try to further model the performance of the Cn+Ca2+/Zn2+ using the single level tunneling model.62,81 Details of the model can be found in the Supporting Information, where we describe the main idea