13652
2007, 111, 13652-13654 Published on Web 08/25/2007
Chemical Site Capacitance: Submolecular Measurements and a Model Roie Yerushalmi,*,† Milko E. van der Boom,‡ and Hagai Cohen§ Department of Electrical Engineering and Computer Sciences, UniVersity of California at Berkeley, Berkeley, California 94720, Department of Organic Chemistry, The Weizmann Institute of Science, 76100 RehoVot, Israel, and Department of Chemical Research Support, The Weizmann Institute of Science, 76100 RehoVot, Israel ReceiVed: July 23, 2007
The study of molecular electric properties is an intriguing, rapidly developing field in which technological and basic scientific challenges and developments are evolving. Nevertheless, understanding of the interplay of intermolecular interactions, substrate effects, and electrode contacts remains challenging. Here, we present noncontact chemically resolved electrical measurements (CREM) of halide-terminated molecular layers and a straightforward model for quantitative analysis of submolecular chemical site capacitance. We demonstrate that under low current densities, the main electronic effects can be accounted for by considering the (sub)molecular properties of the monolayers, whereas the excess potential due to charge injection can be described as site capacitance corresponding to chemically identifiable molecular sites.
Introduction The study of molecular electric properties is an intriguing, rapidly developing field in which technological as well as basic scientific challenges and developments are evolving. The first proposed molecular rectifier, introduced three decades ago by Ratner and Aviram, still stimulates enormous interest in the context of molecular electronics.1 Much progress has been made regarding advances in experimental and theoretical approaches in studying single molecules and monolayers.2-6 Nevertheless, gaps still remain in our basic understanding of the influence of monolayer design, intermolecular interactions, substrate effects, and electrode contacts.3,7 Understanding the relationships between molecular structure and electrical properties is crucial for providing predictive propensity models applicable for systems of high complexity. Notably, the study of molecular electronic and electric properties commonly involves electrode contacts that affect and contribute to the electronic structure of the studied system, thus introducing intrinsic complications in understanding the molecular-level phenomena.7 In addition to the experimental challenges, the electrical properties of molecules and molecular structures such as monolayers present an intrinsic duality with manifestation of both classical- and quantum mechanical-like characteristics.8 Here, we present direct capacitance measurement of functional groups in monolayers utilizing the noncontact chemically resolved electrical measurement (CREM) method for monolayers and present a straightforward model based on individual molecular properties for quantifying the main electronic effects. The model-based analyses are in good agreement with the * To whom correspondence should be addressed. E-mail: roie@ berkeley.edu. † University of California at Berkeley. ‡ Department of Organic Chemistry, The Weizmann Institute of Science. § Department of Chemical Research Support, The Weizmann Institute of Science.
10.1021/jp075795b CCC: $37.00
experimentally derived data and reveal valuable fundamental information regarding the monolayer response to charge injection at the submolecular level. Methods Sample Preparation. Halide-containing monolayers (1-4, Table 1) were prepared using well-established siloxane-based condensation chemistry and indium-tin-oxide (ITO) substrates. The substrates were treated with a dry pentane solution of the corresponding silane (1:50 v/v) at room temperature for 20 min under N2. The substrates were then thoroughly washed with copious amounts of dry solvents and dried at 115 °C for about 10 min. X-ray photoelectron spectroscopy (XPS) measurements yielded the expected elemental composition and molecular thickness. Preparation and characterization of the samples by complementary techniques is described elsewhere.9-11 Chemically Resolved Electrical Measurements (CREM). We utilized the recently introduced noncontact chemically resolved electrical measurement method (CREM).12 CREM is capable of directly probing the electrical properties at subsurface sites with no need for a physical (top) contact between the device setup, power supply, and voltmeter.12,13 In particular, it is now possible to resolve submolecular electrical data;13b hence, quantitative understanding of these experiments becomes of utmost importance. Data were collected with a customized Kratos AXIS-HS X-ray photoelectron spectrometer, using monochromatized Al(KR) radiation. Under electrical input signals, the kinetic energy of photoelectrons ejected from site i is given by12-14
Ek,i ) hν - EiB + eφi
(1)
where hν is the photon energy, EiB is the binding energy of the electron, and eφi is the electrostatic energy related to the local © 2007 American Chemical Society
Letters TABLE 1: Summary of Experimental and Calculated Results
J. Phys. Chem. C, Vol. 111, No. 37, 2007 13653 correlation functional.16 The correlation-consistent triple-ζ basis sets were used for molecules 2-4 (B3LYP/cc-pvTz), and the SDB-cc-pVTZ basis set was used for molecule 1 to account for iodine.17 Atomic charges were obtained by using the NPA atomic charge analysis on optimized structures.18,19 Computational analysis was performed using the Gaussian 03 package.20 Results and Discussion Previously, XPS line shifts were introduced as a tool for depth profiling capable of resolving depth-related charging for consecutive molecular layers. The observed spectral shifts were shown to correlate with the layer depth, where the charge density is modeled as a collection of parallel-plate capacitors. This approximation successfully accounts for relatively high current densities and film thicknesses of a few nanometers.21 However, at low current densities and, more importantly, at the submolecular level, local electronic effects may become more pronounced than the collective response of the monolayer. In particular, CREM can provide element-specific I-V measurements upon charge injection.12-14,21 The here introduced model is based on the following rational: In order to quantify the contribution of atomic and group capacitance to the potential buildup at low current and charge densities, we considered here atomic and functional group components of the molecular units as polarizable metallic spheres.22 Additionally, the local dielectric constant was evaluated using the atomic susceptibility, χi, or the atomic site polarizability, Ri, and the microscopic definition of the dielectric constant
ri ) 1 + 4πχi ) 1 + 4πRi
a
The Br value was used arbitrarily for normalization.
potential, φi, at the corresponding site. This expression allows one to derive the local electrostatic potential variations (∆Φ) of chemical moieties using XPS spectral line shifts at specific submolecular sites.13b,c The input electrical signal was supplied here via a distant electron flood gun (eFG), providing low kinetic energy electrons, below 1 eV, and low flux, below 100 nA cm-2. The electron current parameters were controlled via a filament current and two bias voltages.12 Sample current detection and sample biasing were provided on the back contact using a Keithley 487 picoammeter. The ∆Φ values (CREM data, Table 1) could be determined at an accuracy of ∼10 meV by applying a best-fit analysis to the full spectral envelopes of measured and reference spectra (>200 data points per CREM measurement).12 A previously presented experimental protocol, consisting of repeated reversibility checks, was utilized here to exclude irreversible changes such as chemical degradation.13b,15 The time scale for “damage-free” data collection is system-dependent and should be carefully determined. Several data sets were collected for each monolayer sample to evaluate the long time-scale damage process. Computational Methods. Geometry optimization, Hessian calculations, and atomic charge analysis were performed for molecules (1-4) in the gas phase. For computational convenience, Si-OH groups were used to model the substrate interface. We employed hybrid density functional (HDFT) techniques using Becke’s three-parameter hybrid exchange functional with the widely used Lee-Yang-Parr (LYP) gradient-corrected
(2)
The second part in eq 2 holds as long as the microscopic picture is a valid approximation.23 When the analogy of atoms as conducting spheres is used,8,22 the local potential buildup, ∆φi, evolving upon charging at site i is then given by
∆φi )
δqi 4πo(1 + 4πRi)ri
(3)
where o is the permittivity of free space and δqi is the excess partial charge of the corresponding atomic site or chemical functional group at each molecule, for the singly charged versus the neutral state. Charge distribution values were independent of our proposed model as these values were obtained by computational analyses of the molecular systems employing quantum mechanical techniques as described in the Computational Methods section. Various monolayers were studied in order to provide a comprehensive comparison of the model and experimental findings (Table 1). We present Figure 1, where all data sets from Table 1 are plotted with a single free parameter per monolayer sample, which accounts for the effective current density for each sample. The experimental results and calculated values using eq 3 show an overall good correlation (R2 ) 0.9) for the different molecules. The halide data for monolayers 1-4 (0, Figure 1) show significant potential variations, considerably larger than those observed for the silicon (O, Figure 1) and carbon (×, Figure 1) in the low charge injection density regime.15 Notably, the correlation quality indicates that the model-based analyses serve as a highly sensitive and informative tool. The various halogen atoms at different molecular sites are well described by the present model, pointing to (atomic-size)
13654 J. Phys. Chem. C, Vol. 111, No. 37, 2007
Letters discussions. This research was supported by the G.M.J. Schmidt Minerva Center, BMBF and the ISF, and by a Human Frontiers Science Program fellowship (R.Y.). M.E.vd.B. is the incumbent of the Dewey David Stone and Harry Levine career development chair and Head of the MJRG for Molecular Materials and Interface Design. References and Notes
Figure 1. Normalized CREM data of monolayers 1-4 and the calculated potential profiles obtained by eq 3.
capacitance characteristics that are resolvable by the CREM method. Consideration of the phenyl carbons in samples 1-3 as sp3 instead of conjugated sp2 carbons, utilizing the corresponding polarizabilities in eq 3, yields ∆φi values that markedly lower the correlation quality value (not shown). In fact, the computational analysis suggests a collective response of the phenyl ring electrons to the electric perturbation. The silicon data show relatively substantial deviations from an ideal correlation, which is attributed to the small magnitude of the experimental values and the oversimplified Si environment modeling excluding the substrate interface. Our combined experimental/computational results suggest that a significant portion of the electric response can be accounted for via sitespecific capacitance, described by chemically identifiable groups and indices. The measured electrical response of the various chemical sites can be mainly attributed to the molecular charge density distribution, derived from the molecular level calculations. Thus, the amount of charge accumulated at each atomic site and the potential buildup is dictated by the local properties and electronic structure. Our results are in line with previous theoretical and computational works where the quantum scaling behavior of atomic capacitance was found to strongly resemble that of macroscopic spherical conductors.8,24,25 Conclusions We have presented direct electrical measurements of the submolecular response to charge injection. Our proposed model successfully reproduces the main contributions of local potential buildup for a range of chemically different environments. The combined CREM and computational analysis of various halogenterminated monolayers reveals that at low current densities, the detailed monolayer response can be quantitatively described by molecularly recognizable components. Comparing the molecular sites with classical spherical capacitors perhaps presents a simplified view of the complex molecular system; yet, this model-based approach captures the overall behavior with comprehensible indices. The combined experimental and computational approach demonstrated here benefits from the elimination of electrode contacts by applying a source for nonwired currents. Further application of the methodology presented here is expected to provide a bridge between the chemical, electronic, and electric properties of complex molecular systems. Acknowledgment. The authors thank Dr. L. Kronik and Mr. A. Natan from the Weizmann Institute of Science for helpful
(1) Aviram, A.; Ratner, M. A. Chem. Phys. Lett. 1974, 29, 277. (2) Datta, S.; Tian, W. D.; Hong, S. H.; Reifenberger, R.; Henderson, J. I.; Kubiak, C. P. Phys. ReV. Lett. 1997, 79, 2530. (3) Houck, A. A.; Labaziewicz, J.; Chan, E. K.; Folk, J. A.; Chuang, I. L. Nano Lett. 2005, 5, 1685. (4) Metzger, R. M. Chem. ReV. 2003, 103, 3803. (5) (a) Collman, J. P.; Devaraji, N. K.; Decreau, R. A.; Yang, Y.; Yan, Y. L.; Ebina, W.; Eberspacher, T. A.; Chidsey, C. E. D. Science 2007, 315, 1565. (b) Nitzan, A.; Ratner, M. A. Science 2003, 300, 1384. (6) Piva, P. G.; DiLabio, G. A.; Pitters, J. L.; Zikovsky, J.; Rezeq, M.; Dogel, S.; Hofer, W. A.; Wolkow, R. A. Nature 2005, 435, 658. (7) Ratner, M. Nature 2005, 435, 575. (8) Ellenbogen, J. C. Phys. ReV. A 2006, 7, 034501. (9) (a) Maoz, R.; Sagiv, J. Langmuir 1987, 3, 1034. (b) Zngst, D. L.; Simmons, G. W. Langmuir 1991, 7, 2236. (c) Richter, A. G.; Yu, C. J.; Data, A.; Kmetko, J.; Dutta, P. Phys. ReV. E 2000, 6, 607. (10) (a) Yerushalmi, R.; Scherz, A.; van der Boom, M. E. J. Am. Chem. Soc. 2004, 126, 2700. (b) Filip-Granit, N.; Yerushalmi, R.; Brandis, A.; van der Boom, M. E.; Scherz, A. J. Phys. Chem. B 2005, 109, 6933. (11) (a) Lin, W.; Lin, W.; Wong, G. K.; Marks, T. J. J. Am. Chem. Soc. 1996, 118, 8034. (b) Li, D.-Q.; Swanson, B. I.; Robinson, J. M.; Hoffbauer, M. A. J. Am. Chem. Soc. 1993, 115, 6975. (12) Cohen, H. Appl. Phys. Lett. 2004, 85, 1271. (13) (a) Cohen, H.; Nogues, C.; Zon, I.; Lubomirsky, I. J. Appl. Phys. 2005, 97, 113701. (b) Filip-Granit, N.; van der Boom, M. E.; Yerushalmi, R.; Scherz, A.; Cohen, H. Nano Lett. 2006, 6, 2848. (c) Cohen, H.; Zenkina, O.V.; Shukla, A. D.; van der Boom, M. E. J. Phys. Chem. B 2006, 110, 1506. (14) Shabtai, K.; Rubinstein, I.; Cohen, S. R.; Cohen, H. J. Am. Chem. Soc. 2000, 122, 4959. (15) The trapped charge density under higher charging conditions was recently shown to be ∼0.02 e- per molecule; see Cohen, H.; Maoz, R.; Sagiv, J. Nano Lett. 2006, 6, 2462. (16) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 8, 899. (17) Martin, J. M. L.; Sundermann, A. J. Chem. Phys. 2001, 114, 3408. (18) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735. (19) Yerushalmi, R.; Scherz, A.; Baldridge, K. K. J. Am. Chem. Soc. 2004, 126, 5897. (20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004. (21) Doron-Mor, I.; Hatzor, A.; Vaskevich, A.; van der Boom-Moav, T.; Shanzer, A.; Rubinstein, I.; Cohen, H. Nature 2000, 406, 382. (22) Komorowski, L. Chem. Phys. Lett. 1987, 134, 536. (23) For example, see: Jackson, J. D. Classical Electrodynamics, 2nd ed.; Wiley & Sons: New York, 1999. We used microscopic polarizability of single atoms rather than bulk material; hence, the Clausius-Mossotti relation, usually used to relate the dielectric constant and polarizability, is not applicable here. We restricted the analysis to low currents and low charge density, as mentioned in ref 15. (24) Gazquez, J. L.; Ortiz, E. J. Chem. Phys. 1984, 8, 2741. (25) Perdew, J. P. Phys. ReV. B 1988, 37, 6175.
