Langmuir 2006, 22, 6249-6255
6249
Phenylphosphonic Acid Functionalization of Indium Tin Oxide: Surface Chemistry and Work Functions Sharon E. Koh,† Krystal D. McDonald, David H. Holt, and Charles S. Dulcey* Center for Bio/Molecular Science and Engineering, Code 6900, NaVal Research Laboratory, Washington, D.C. 20375
John A. Chaney‡ and Pehr E. Pehrsson Chemistry DiVision, Code 6174, NaVal Research Laboratory, Washington, D.C. 20375 ReceiVed August 31, 2005. In Final Form: March 14, 2006 The work function of indium tin oxide (ITO) substrates was modified with phosphonic acid molecular films. The ITO surfaces were treated prior to functionalization with a base cleaning procedure. The film growth and coverage were quantified by contact angle goniometry and XPS. Film orientation was determined by reflection/absorption infrared spectroscopy using ITO-on-Cr substrates. The absolute work functions of nitrophenyl- and cyanophenylphosphonic acid films in ITO were determined by Kelvin probe measurement to be 5.60 and 5.77 eV, respectively.
Introduction There is increasing interest in the use of self-assembly to control interfaces in semiconductor devices.1,2 Interfacial properties such as surface energy, charge, and polarity can be manipulated through self-assembly to enhance device performance. Much research on self-assembled monolayers (SAMs) has focused on alkanethiols and organosilanes,3 which can be used to functionalize noble metals and hydroxyl-terminated surfaces, respectively. A third class of self-assembly, molecular acids on metal oxides4 is less well characterized than thiols and organosilanes but is very promising for the modification of metal oxide surfaces and is the subject of this study. Charge injection at the organic-inorganic interface is a key issue in organic diodes, transistors, and photovoltaics.5 Surface treatments can be used to alter charge injection at such interfaces. This may occur as a result of lowering or raising the work function of cathodes or anodes, respectively, or by enhancing cohesion and thus lowering interfacial series resistance. Robust methods to control these interfaces for electron-hole balance is a critical issue in organic light-emitting diodes (OLEDs).6 Indium tin oxide (ITO) is presently the conductive material of choice for OLED anodes as well as in sensors and solar cells.7 The work function * To whom correspondence should be addressed. E-mail: csd@ cbmse.nrl.navy.mil. † GeoCenters Inc. Present address: Chemistry Department, Northwestern University. ‡ NRC Postdoctoral Associate. Present address: Department of Chemical Engineering, University of Louisville. (1) Ashkenasy, G.; Cahen, D.; Cohen, R.; Shanzer, A.; Vilan, A. Acc. Chem. Res. 2002, 35, 121-128. (2) Vilan, A.; Cahen, D. Trends Biotech. 2002, 20, 22-29. (3) (a) Ulman, A. Chem. ReV. 1996, 96, 1533-1554. (b) Ulman, A. An Introduction to Ultrathin Organic Films: From Langmuir-Blodgett to SelfAssembly; Academic Press: New York, 1991. (4) (a) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45-52. (b) Van Alsten, J. G. Langmuir 1999, 15, 7605-7614. (c) Folkers, J. P.; Gorman, C. B.; Laibinis, P. E.; Buchholz, S.; Whitesides G. M. Langmuir 1995, 11, 813-824. (d) Lewington, T. A.; Alexander M. R.; Thompson, G. E.; McAlpine, E. Surf. Eng. 2002, 18, 228-232. (5) (a) Scott, J. C. J. Vac. Sci. Technol., A 2003, 21, 521-531. (b) Cui, J.; Huang, Q. L.; Veinot, J. C. G.; Yan, H.; Wang, Q. W.; Hutchison, G. R.; Richter, A. G.; Evemenenko, E.; Dutta, P.; Marks, T. J. Langmuir 2002, 18, 9958-9970. (6) Friend, R. H.; Gymer, R. W.; Holmes, A. B.; Burroughes, J. H.; Marks, R. N.; Taliani, C.; Bradley, D. D. C.; Dos Santos, D. A.; Bredas, J. L.; Longlund, M.; Salaneck, W. R. Nature 1999, 397, 121-128.
and charge injection efficiency of ITO have been shown to be sensitive to surface chemical treatments including oxidation,8 acid-base treatments,9 and self-assembly.10-14 Most studies of self-assembly for OLEDs have investigated electrical performance but generally have not reported the detailed characterization of the surface modification process and resulting films. For example, Appleyard and Willis10 modified ITO with 2-chloroethane phosphonic acid, reducing the threshold voltage of an OLED made with this material by 4 V and allowing the replacement of problematic Mg anodes with Al. No surface characterization or work function measurements of the ITO surface were reported. Nuesch et al.11 modified ITO cathodes with 4-nitrobenzoic acid, causing an indirectly estimated work function shift of 0.5 eV and improving their device performance. The relative lack of surface characterization is primarily due to the complexity of the ITO surface, which can be difficult to modify and characterize reproducibly. The ITO surface exhibits different physical and chemical properties depending on the fabrication method, and it possesses high surface roughness and heterogeneity across a given sample. A few detailed surface analytical studies of ITO have looked at surface cleaning12 and self-assembly.13,14 Markovich and Mandler13 examined the modification of ITO by alkylsilanes for sensor applications, and Yan et al.14 studied the modification of (7) Hartnagel, H. L.; Dawer, A. L.; Jain A. K.; Jagadish, C. Semiconducting Transparent Thin Films; Institute of Physics Publishing: Bristol, England, 1995. (8) Milliron, D. J.; Hill, I. G.; Shen, C.; Kahn, A.; Schwartz, J. J. Appl. Phys. 2000, 87, 572-576. (9) (a) Nuesch, F.; Kamaris, K.; Zuppiroli, L. Chem. Phys. Lett. 1998, 283, 194-200. (b) Nuesch, F.; Rothberg, L. J.; Forsythe, E. W.; Le, T. E.; Gao, Y. Appl. Phys. Lett. 1999, 74, 880-882. (c) Nuesch, F.; Forsythe, E. W.; Le, T. E.; Gao, Y.; Rothberg, L. J. J. Appl. Phys. 2000, 87, 7973-7980. (10) (a) Appleyard, S. J. F.; Willlis M. R. Opt. Mater. 1998, 9, 120-124. (b) Appleyard, S. J. F.; Day, S. R.; Pickford R. D.; Willis, M. R. J. Mater. Chem. 2000, 10, 169-173. (c) Hatton, R. A.; Day, S. R.; Chester, M. A.; Willis, M. R. J. Mater. Chem. 2003, 13, 38-43. (11) (a) Nuesch, F.; Rotzinger, F.; Si-Ahmed, L.; Zuppiroli, L. Chem. Phys. Lett. 1998, 288, 861-867. (b) Zuppiroli, L.; Si-Ahmed, L.; Kamaras, K.; Nuesch, M. N.; Bussac, M. N.; Ades, D.; Siove, A.; Moons E.; Gratzel, M.; Eur. Phys. J. B 1999 11, 505-512. (12) Donley, C.; Dunphy, D.; Paine, D.; Carter, C.; Nebesny K.; Lee, P,; Alloway, D.; Armstrong N. R. Langmuir 2002, 18, 450-457. (13) Markovich, I.; Mandler, D. J. Electroanal. Chem. 2001, 500, 453-460. (14) Yan, C.; Zharnikov, M.; Goldhauser, A.; Grunze, M. Langmuir 2000, 16, 6206-6215.
10.1021/la052379e CCC: $33.50 © 2006 American Chemical Society Published on Web 06/06/2006
6250 Langmuir, Vol. 22, No. 14, 2006
Figure 1. Structures of the phosphonic and carboxylic acid molecules used in this work: (a) nitrobenzoic acid, (b) cyanobenzoic acid, (c) benzoic acid, (d) nitrophenylphosphonic acid, (e) cyanophenylphosphonic acid, (f) undecanoic acid, and (g) octadecylphosphonic acid.
ITO by alkanethiols and carboxylic acids. As noted, there have been some studies of phosphonic acid modification of ITO for devices,10,11 but there has been no systematic study of the functionalization process. In this work, we treated indium tin oxide with phenylphosphonic acid derivatives possessing large dipole moments, concentrating on the characterization of films of nitrophenylphosphonic acid. Nitrophenylphosphonic acid possesses a nitro group where the N 1s XPS peak is well suited to monitor coverage and the nitro infrared bands can be used to establish orientation. The objectives of the study were to (1) optimize an ITO cleaning protocol; (2) characterize the bonding, growth, coverage, and structure of the resulting films; and (3) determine the resulting surface work functions. The work functions of ITO modified by nitrophenyl- and cyanophenylphosphonic acids were measured by Kelvin probe, and the results were related to the organic film structure. Experimental Section Substrates and Cleaning. The ITO (30 Ω/0 sheet resistance) was obtained from Applied Films and cleaned by a sequence of wet chemical steps. Organic solvents examined for cleaning effectiveness were spectroscopic-grade acetone, acetonitrile, chloroform, and trichloroethylene (TCE). Each solvent treatment was carried out for 10 min in an ultrasonic bath. On the basis of the results of the individual solvent treatments, a multistep cleaning sequence was developed to achieve both the minimum water contact angle and minimum XPS C(1s) signal, indicative of a clean surface. The sequence adopted was trichloroethylene followed by acetone and then methanol and finally immersion in 10% KOH/isopropyl alcohol (IPA) at 50 °C for 10 min. A deionized H2O rinse was performed after each solvent step. The KOH/IPA solution was allowed to stir on a hotplate at 50 °C for ∼30 min prior to use. Deionized H2O used in contact angle measurements and in substrate cleaning was >18 MΩ‚cm deionized H2O from a Barnstead Nanopure filtration system. Reagents. 4-Nitrobenzoic acid, 4-cyanobenzoic acid, benzoic acid, undecanoic acid, 4-nitro-phenyl phosphonic (NPPA), and phenylphosphonic acid were used as received from Aldrich Chemical Co. to prepare self-assembling monolayer treatment solutions. Octadecylphosphonic acid was synthesized according to the Michaelis-Arbuzov rearrangement.15 4-Cyanophenyl phosphonic acid (CPPA) was synthesized using a dialkylarenephosphonate synthetic method with a palladium catalyst. The purity of the synthesized materials was confirmed by proton NMR. The molecular structures are shown in Figure 1. SAM Formation. Self-assembling monolayer (SAM) treatment solutions were prepared by dissolving self-assembling molecules in (15) Bhattacharya, A. K.; Thyagarajan, G. Chem. ReV. 1981, 81, 415-430.
Koh et al. an organic solvent; 4-nitrobenzoic acid, 4-cyanobenzoic acid, and benzoic acid were dissolved in chloroform; and undecanoic acid, 4-nitro-phenyl-phosphonic acid, 4-cyano-phenyl-phosphonic acid, and octadecylphosphonic acid were dissolved in tetrahydrofuran (THF). Solutions were prepared in ∼1 mM concentrations. The cleaned substrates were immersed in the prepared self-assembling monolayer (SAM) treatment solution for ∼15 h. The film-treated substrates were rinsed with the appropriate solvent (chloroform or THF) and then blown dry with a stream of dry N2. Contact Angles. Contact angles were measured by the sessile drop method16 using a Rame-Hart model 100 contact angle goniometer under ambient conditions in a class 1000 cleanroom. Several 15 µL drops of >18 MΩ‚cm deionized H2O were dispensed by a micropipettor, and multiple contact angle measurements were taken with ∼3° precision. Each contact angle value reported here is the average of at least three measurements per substrate from multiple substrates. The reproducibility of the contact angle measurements for different substrates treated under the same conditions was (5°. X-ray Photoelectron Spectroscopy. X-ray photoelectron spectroscopy experiments were performed on a Surface Science Instruments SSX-100 surface analysis system with a monochromatic Al KR source (hν ) 1486.6 eV), a hemispherical analyzer, and a position-sensitive detector. The pressure in the analysis chamber during measurements was typically between 1 and 5 × 10-9 Torr. The photoelectron takeoff angle was 55° as measured from the surface normal. Angle-dependent measurements were performed by making measurements with takeoff angles of ∼ 0, 45, and 90°. Survey scans were taken to identify major elements present on the surface. Chemical state information and surface concentrations were derived from highresolution scans of individual elements measured with an instrument pass energy of 50 or 100 eV (∼0.05 and 0.10 eV resolution, respectively). Absolute binding energies were determined by referencing the main carbon peak to 285.0 eV. Surface atomic compositions were determined using standard XPS cross sections, and atomic percent values have a minimum detection limit of ∼0.1% and an accuracy and precision estimated to be (10% of the calculated value or better. Infrared Spectroscopy. Infrared spectra were taken on a Nicolet Magna 750 Fourier transform infrared (FTIR) spectrometer using a KBr beam splitter and a mercury cadmium telluride (MCT) detector. Reflection/absorption measurements were made with p-polarized light at an angle of incidence of 80° using a SpectraTech FT80 reflection accessory. The reflective substrates were made by coating a glass slide with 2000 Å of chromium for adhesion and reflectivity and then with 500 Å of ITO. The sample compartment was purged overnight with nitrogen prior to measurement, and the signal was averaged for 2000 scans at a resolution of 2 cm-1. Dipole Moments and Work Functions. Dipole moments to correlate with work functions were calculated by an ab initio calculation employing a STO3-21G basis set using Hyperchem software. The work functions of the modified ITO samples were determined using a vibrating Kelvin probe (KP-6500 McAllister Tech.). The various ITO samples were introduced into an ultrahigh vacuum (UHV) system with a base pressure of 2 × 10-10 Torr. All samples, including clean, unfunctionalized surfaces, were annealed at 50 °C for ∼10 to 15 min prior to measurement to drive off any atmospheric adsorbates and to establish thermal equilibrium. The use of this technique for ITO preparation has been described previously in greater detail.17 In the Kelvin probe technique,18 the probe surface is positioned in proximity and parallel to the grounded sample to form a parallel plate capacitor, and a small (3 to 4 V) backing potential is applied between the two surfaces. The probe vibrates back and forth with respect to the sample, changing the capacitance and inducing a current. When the backing potential equals the contact potential difference (16) Zisman, W. A. Contact Angle, Wettability, and Adhesion. Advances in Chemistry Series. American Chemical Society: Washington, DC, 1964; Vol. 43. (17) Chaney, J. A.; Pehrsson, P. E. Appl. Surf. Sci. 2001, 180, 214-226. (18) Baikie, I.; Petermann, U.; Lagel, B. Surf. Sci. 1999, 435, 249-253.
Functionalization of Indium Tin Oxide
Langmuir, Vol. 22, No. 14, 2006 6251 Table 1. ITO Surface Characteristics (A) Surface Carbon on ITO after Single-Step Solvent Cleaning
solvent
as delivered
CH3Cl
CH3CN
(CH3)2CO
CH2CCl2
%C
42 ( 4
30
28
28
20
(B) Wettability and XPS of ITO Surface during the Multistep Wet Cleaning Procedure treatment steps as delivered TCE TCE/Ace TCE/Ace/MeOH TCE/Ace/MeOH...KOH/IPA O2 plasma
θw
%C
In/Sn
81 ( 3 32° 23° 15° 40% as measured by XPS (Table 1). Deconvolution of the C 1s region shows components at 285.0, 286.4, 287.8, and 289.4 eV, consistent with carbon in alkylic, alcoholic, carbonylic, and carboxylic oxidation states22 (with relative concentrations of 75, 10, 5, and 10%, respectively). A varying amount
Figure 2. Angle-resolved XPS of as-delivered ITO at near-grazing, intermediate, and near-normal photoelectron takeoff angles for the O 1s peak.
O/(In+Sn) 3.0 ( 0.3 2.9 2.8 2.7 2.4 2.7
of nitrogen N 1s at ∼400 eV (up to 0.4%) nitrogen is also observed, which does not interfere with the measurement of nitro nitrogen at 406 eV but can potentially interfere with the measurement of cyano nitrogen.
ITO Surface Preparation We explored wet cleaning processes for ITO because plasma cleaning itself can alter the ITO surface stoichometry as well as change the work function.23 The as-delivered ITO samples were very hydrophobic with contact angles near 80°. Angle-resolved XPS (Figure 2) indicated the presence of an oxygen-containing contaminant (632 eV) concentrated at the surface (low takeoff angles). Organic degreasing solvents acetone, acetonitrile, chloroform, and trichloroethylene were evaluated first to remove the carbon contamination presumed to be the source of the hydrophobicity. Surface carbon was measured by XPS after each treatment, and the results are tabulated in Table 1. Trichloroethylene (TCE) was found to be the most effective single solvent, removing over 50% of surface carbon. A multistep cleaning procedure was developed because any single solvent left a significant amount of C on the surface. After TCE, a progression toward more aqueous-like solvents was used. The adopted sequence was trichloroethylene/acetone/methanol/ KOH in IPA. The KOH treatment was intended to etch the ITO surface mildly, lifting off remaining organic and inorganic surface contaminants. AFM did not reveal any significant disruption of the surface caused by the base cleaning. The rms surface roughness of ITO actually decreased slightly from 7 nm before cleaning to 5 nm after cleaning. The multistep cleaning procedure reduced the contact angle to below 7° and dropped the surface carbon concentration from ∼42 to
