Self-Assembled Organic Radicals on Au(111) Surfaces: A Combined

3, I-50019 Sesto Fiorentino, Italy, Department of Chemical and Biosystems Science ... The scientific appeal of thin films of magnetic materials is uph...
0 downloads 0 Views 656KB Size
Download PDF
Langmuir 2007, 23, 2389-2397

2389

Self-Assembled Organic Radicals on Au(111) Surfaces: A Combined ToF-SIMS, STM, and ESR Study Matteo Mannini,‡ Lorenzo Sorace,‡ Lapo Gorini,‡ Federica M. Piras,† Andrea Caneschi,‡ Agnese Magnani,† Stefano Menichetti,§ and Dante Gatteschi*,‡ Department of Chemistry, INSTM RU, UniVersity of Florence, Via della Lastruccia n. 3, I-50019 Sesto Fiorentino, Italy, Department of Chemical and Biosystems Science and Technology and UniVersity Centre of Colle di Val d’Elsa, UniVersity of Siena, Via A. Moro 2, I-53100 Siena, Italy, and Department of Organic Chemistry, UniVersity of Florence, Via della Lastruccia n. 13, I-50019 Sesto Fiorentino, Italy ReceiVed July 13, 2006. In Final Form: NoVember 20, 2006 Electron spin resonance (ESR), time-of-flight secondary ion mass spectrometry (ToF-SIMS), and scanning tunneling microscopy (STM) have been used in parallel to characterize the deposition on gold surface of a series of nitronyl nitroxide radicals. These compounds have been specifically synthesized with methyl-thio linking groups suitable to interact with the gold surface to form self-assembled monolayers (SAMs), which can be considered relevant in the research for molecular-based spintronics devices, as suggested in recent papers. The degree of the expected ordering on the surface of these SAMs has been tuned by varying the chemical structure of synthesized radicals. ToF-SIMS has been used to support the evidence of the occurrence of the deposition process. STM has shown the different qualities of the obtained SAMs, with the degree of local order increasing as the degree of freedom of the molecules on the surface is decreased. Finally, ESR has confirmed that the deposition process does not affect the paramagnetic characteristics of radicals and that it affords a complete single-layered coverage of the surface. Further, the absence of angular dependence in the spectra indicates that the small regions of local ordering do not give rise to a long-range order and suggests a quite large mobility of the radical on the surface, probably due to the weak interaction with gold provided by the methyl-thio linking group.

Introduction The scientific appeal of thin films of magnetic materials is upheld by their great technological interest. Studies in this field are mainly related to the development of new kinds of data storage materials and to the miniaturization of data storage devices. Bottom-up procedures involved in these miniaturization processes suggest that the ultimate solution for the scaling-down is related to the use of single molecules as functional objects. Focusing on magnetic devices, we can consider as a starting point of this kind of study the pioneering work on ultrathin magnetic films presented by Pomerantz and co-workers in the late 1970s,1 who investigated the magnetism of one monolayer of molecules deposited by a Langmuir-Blodgett (LB) film technique. This was the first attempt to explore one of the limits of magnetic data storage devices controlling the thickness of the active material down to the molecular scale.2 Three decades after that study, various solutions permit control of the deposition of complex functional molecules in single layers with regular bidimensional organization. For chemists, the most attractive and widely used of such techniques is the selfassembling monolayer (SAM) one,3 which enables one to obtain a bidimensional lattice of molecules. These molecules are characterized by functional groups suitable to bind to a specific surface and to enable lateral intermolecular interactions. Moreover, specific substituents can be added to such molecules in * Corresponding author. Tel.: +390554573285; fax: +390554573372; e-mail: [email protected]. ‡ Department of Chemistry, University of Florence. † University of Siena. § Department of Organic Chemistry, University of Florence. (1) Pomerantz, M.; Pollak, R. A. Chem. Phys. Lett. 1975, 31, 602. (2) Talham, D. R. Chem. ReV. 2004, 104, 5479. (3) Ulman, A. Ultrathin Organic Films; Academic Press, Inc.: San Diego, CA, 1991.

order to obtain complete or partially functionalized monolayers of selected molecules. Our group is interested in the study and application of the molecular magnetism of innovative compounds4 and their potential application in the field of advanced nanodevices for magnetic data storage.5 This goal can, in principle, be achieved by using so-called single-molecule magnets (SMMs), molecules that show magnetic hysteresis at low temperatures. In order to develop nanodevices, it is necessary to elaborate techniques to obtain SAMs of SMMs and locally address these molecules using the specific evolution of scanning probe microscopes (SPMs)6 to access single molecular magnetism. To correctly implement suitable strategies for SAM formation and characterization, it is convenient to start from a simple system. Here we explore the possibility of employing the SAM technique to deposit a series of nitronyl nitroxide radicals (hereafter referred to as NitRs) obtaining a single layer of locally ordered paramagnetic centers. NitRs are a class of paramagnetic compounds7 widely studied both for their intrinsically interesting magnetic properties8,9 and (4) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets; Oxford University Press: Oxford, 2006. (5) Cornia, A.; Fabretti, A. C.; Zobbi, L.; Caneschi, A.; Gatteschi, D.; Mannini, M.; Sessoli, R. Struct. Bonding 2006, 122, 133. (6) (a) Martin Y.; Wickramasinghe, H. K. Appl. Phys. Lett. 1987, 50, 1455. (b) Rugar, D.; Budakian, R.; Mamin, H. J.; Chui, B. W. Nature 2004, 430, 29. (c) Oral, A.; Bending, S. J.; Henini, M. Appl. Phys. Lett. 1996, 69, 1325. (d) Zhao, A. D.; Li, Q. X.; Chen, L.; Xiang, H. J.; Wang, W. H.; Pan, S.; Wang, B.; Xiao, X. D.; Yang, J. L.; Hou, J. G.; Zhu, Q. S. Science 2005, 309, 1542. (e) Wiesendanger, R.; Guntherodt, H. J.; Guntherodt, G.; Gambino, R. J.; Ruf, R. Phys. ReV. Lett. 1990, 65, 247. (f) Manassen, Y.; Hamers, R. J.; Demuth, J. E.; Castellano, A. J., Jr. Phys. ReV. Lett. 1989, 62, 2531. (g) Durkan, C.; Welland, M. E. Appl. Phys. Lett. 2002, 80, 458. (h) Durkan, C. Contemp. Phys. 2004, 45, 1. (7) Osiecki, J. H.; Ullman, E. F. J. Am. Chem. Soc. 1968, 90, 1078. (8) Tamura, M.; Nakazawa, Y.; Shiomi, D.; Nozawa, Y.; Hosokoshi, M.; Ishikawa, M.; Kinoshita, M. Chem. Phys. Lett. 1991, 186, 401. (9) Chiarelli, R.; Novak, M. A.; Rassat, A.; Tholence, J. L. Nature 1993, 363, 147.

10.1021/la062028f CCC: $37.00 © 2007 American Chemical Society Published on Web 01/31/2007

2390 Langmuir, Vol. 23, No. 5, 2007

Mannini et al. Scheme 1. Synthesis of NitRs 1, 2, and 3

as building blocks for assembling complex magnetic structures in which they act as chemical and magnetic linkers between transition metals10 to yield magnetic polymers including singlechain magnets (SCMs).11 Moreover, recent works suggest that NitR function can be used to increase the spin polarization of the current in molecular wires connecting a network of gold nanoparticles.12 This may ultimately lead to the development of interesting devices for spintronics applications, that is, devices whose operation depends on both the charge and the spin of the conduction electrons.13 Indeed, the weak spin-orbit coupling and hyperfine interactions characterizing organic molecular materials result in an increasein the spin coherence, which is considered a fundamental requisite for obtaining molecular spintronics devices.14 A previous work focused on deposing NitRs with mono- and multilayer control through the LB technique has been presented by Turek and coworkers.15 In that work, the NitR moieties were close to the support surface and thus difficult to be accessed with a tip of an (10) Caneschi, A.; Gatteschi, D.; Sessoli, R.; Rey, P. Acc. Chem. Res. 1989, 22, 392. (11) Caneschi, A.; Gatteschi, D.; Lalioti, N.; Sangregorio, C.; Sessoli, R.; Venturi, G.; Vindigni, A.; Rettori, A.; Pini, M. G.; Novak, M. A. Angew. Chem., Int. Ed. Engl. 2001, 40, 1760. (12) (a) Minamoto, M.; Matsushita, M. M.; Sugawara, T. Polyhedron 2005, 24, 2263. (b) Mitsutaka, O.; Yasutaka, K.; Takeshi, T.; Kizashi, Y. Polyhedron 2005, 24, 2330. (c) Minamoto, M.; Matsushita, M. M.; Sugawara, T.; Nickels, P.; Komiyama, S. Proceedings of the 10th International Conference on Molecule Based Magnets, Victoria, CA, August 2006.

SPM. Another drawback is that LB films lack rigid bidimensional regularity. Sugawara’s group communicated an attempt to deposit NitRs through the SAMs technique, but all the presented characteristics do not fit the monolayer formation suggested by the authors.16 We present here a complete characterization of the deposited monolayers of the NitR radicals, including chemical characterization through the time-of-flight secondary ion mass spectrometry (ToF-SIMS) technique, morphological characterization through scanning tunneling microscopy (STM) measurement, and magnetic characterization through electron spin resonance (ESR) spectroscopy. In particular, we investigate how it is possible to deposit a series of NitRs by modifying the chemical structure of the linker group and we evaluate the effect of this variation on the assembling process and on the magnetic properties. With this paper we would like to provide a new link between molecular magnetism and surface science, and a starting point toward studies (13) (a) Wolf, S. A.; Awschalom, D. D.; Buhrman, R. A.; Daughton, J. M.; von Molna`r, S.; Roukes, M. L.; Chtchelkanova, A. Y.; Treger D. M. Science 2001, 294, 1488. (b) Tsukagoshi, K.; Alphenaar, B. W. Ago, H. Nature 1999, 401, 572. (c) Prinz, G. A. Science 1998, 282, 1660. (d) Ouyang, M.; Awschalom, D. D. Science 2003, 301, 1074. (e) Sharma, P. Science 2005, 307, 531. (14) Rocha, A. R.; Garcia-Suarez, V. M.; Bailey, S. W; Lambert, C. J.; Ferrer, J.; Sanvito, S. Nat. Mater. 2005, 4, 335. (15) Gallani, J. L.; Le Moigne, J.; Oswald, L.; Bernard, M.; Turek, P. Langmuir 2001, 17, 1104. (16) Matsushita, M. M.; Ozaki, N.; Sugawara, T.; Nakamura, F.; Hara, M. Chem. Lett. 2002, 596.

ToF-SIMS, STM, and ESR Analysis of NitR SAMs.

Langmuir, Vol. 23, No. 5, 2007 2391

Scheme 2. Characteristic Fragments Detected for All Three NitRs in ToF-SIMS Positive Spectra.

of magnetic properties of nitronyl nitroxide-based materials deposited on surfaces, such as, for example, SCMs. Experimental Section Synthesis. Radical p-phenyl-S-methyl-nitronyl-nitroxide (1) was prepared by standard methods, as described elsewhere.17 Detailed description of step by step synthesis of p-biphenyl-S-methyl-nitronylnitroxide (2) and p-benzyl-S-methyl-nitronyl-nitroxide (3) is reported in the Supporting Information. Preparation of the Samples. Physical and chemical characterizations were performed on the uniformly prepared samples; deposition was carried out by soaking flame-annealed Au(111) substrates (a film of 150 nm evaporated on mica, annealed with a hydrogen flame) in 1-2 mM solutions of NitRs in CH2Cl2. The slides, of approximate dimensions 3 × 6 mm2, were incubated in the dark at 60 °C overnight and then rinsed with anhydrous CH2Cl2 and dried under nitrogen flux. Drop-casted samples were obtained by evaporating 10-50 µL of 2 mM solutions of NitRs directly on the same kind of gold substrates. Chemical Characterization. ToF-SIMS analysis was carried out with a TRIFT III time-of-flight secondary ion mass spectrometer (Physical Electronics, Chanhassen, MN) equipped with a 69Ga+ liquid-metal primary ion source. Positive and negative ion spectra were acquired with a pulsed, bunched 15 keV primary ion beam at 600 pA by rastering the ion beam over a 100 × 100 µm sample area. The primary ion dose was kept below 1012 ions/cm2 to maintain static SIMS conditions. Positive mass spectra were calibrated to CH3+ (15.023 m/z), C2H3+ (27.023 m/z), C7H7+ (91.005 m/z), and AuC2H4+ (224.998 m/z); negative data were calibrated to CH- (13.008 m/z), C2H- (25.008 m/z), and AuS- (228.938 m/z). The mass resolution (m/∆m) at C2H3+ (27.023 m/z) and C2H- (25.008 m/z) was typically above 6000 and 5000, respectively. Three samples and three points per sample were analyzed for each sample type. Physical Characterization. STM measurements were performed with a P47-Pro system (NT-MDT, Zelenograd, Moscow, Russia) equipped with a customized low-current STM head and Pt/Ir 90/10 mechanically-etched tips prepared immediately before use (a positive bias voltage of 600-700 mV was applied; a constant current mode with a set point within 20 and 40 pA was used). All STM measurements were carried out at room temperature, under N2 atmosphere. (17) Caneschi, A.; David, L.; Ferraro, F.; Gatteschi, D.; Fabretti, A. C. Inorg. Chim. Acta 1995, 235, 159.

In order to support our hypothesis about the most stable configuration for each NitR in vacuum, we performed a raw molecular modeling calculation by using the commercial package HyperChem and applying a semiempirical CNDO method with a geometrical optimization based on the Polak-Ribiere conjugate-gradient algorithm.18 ESR measurements were carried out on an X-band Bruker Elexsys E500 spectrometer equipped with a ER 4131VT liquid nitrogen cryostat for variable temperature studies and a ER4122SHQE cavity to enhance the sensitivity of the instruments (with a nominal weak pitch sensitivity of 3000:1). Nonetheless, the obtained S/N ratio was quite low, and this, coupled with the presence of a relevant instrumental background signal that had to be subtracted from the raw ESR signals, made the analysis of the spectra difficult. The sample was mounted on a sample holder for single crystals connected to a goniometer for orientation-dependent studies. Samples were prepared according to the procedure described above. Solution spectra of 1, 2, and 3 were obtained on 1 mM solutions in CH2Cl2.

Results and Discussion Synthesis. The advantage of using molecular synthetic techniques in preparing functional materials lies in the possibility of tuning molecular properties and molecular assembling through a chemical modification. In principle, this can be easily accomplished by choosing appropriate reagents in a catalog. However, when modifications have to be introduced into specifically functionalized molecules, a fine chemistry work is needed. This concept is clearly evidenced in the synthetic work that is necessary for introducing a simple CH2 or a phenyl group in the desired position to the well-known NitR 1. The self-assembling procedure requires the use of specific linker groups to enable chemisorption of the molecules on the surface. Quite obviously, one of the crucial issues is finding a linker chemically compatible with the rest of the molecule. This is the factor that led us to using a methyl-thio group19 and discarding the introduction of a thiolic function, commonly used in deposition on gold.3 Indeed, a redox competition takes place (18) HyperChem Molecular Modelling System, release 7.5 for Windows; Hypercube, Inc.: Gainesville, FL, 2003 (19) (a) Troughton, E. B.; Bain, C. D.; Whitesides, G. M.; Nuzzo, R. G.; Hill, M.; Allara, D. L.; Porter, M. D. Langmuir 1988, 4, 365. (b) Lavrich, D. J.; Wetterer, S. M.; Bernasek, S. L.; Scoles, G. J. Phys. Chem. B 1998, 102, 3456.

2392 Langmuir, Vol. 23, No. 5, 2007

Figure 1. Positive ToF-SIMS spectra of SAMs of NitRs 1, 2, and 3 on gold. The mass regions between (a) 100 and 200 m/z, (b) 200 and 400 m/z, and (c) 400 and 800 m/z are presented. In panels a and b, the intensity of the spectra of 2 and 3 is amplified 10 times.

between the NitR radical and the thiol group, but not with the methyl-thio group. The formation of SAMs based on the use of the latter functional group is described in the literature.19a In this case, the binding energy is reduced with respect to the use of thiols.19b The key step in the synthesis of NitRs is the formation of the aldehydes (Scheme 1). While, for the preparation of NitR 1, the aldehyde was a commercial one, for the synthesis of NitRs 2 and 3, the respective aldehydes were prepared by using wellknown reaction procedures. Afterward, the condensation reactions of aldehydes with the dihydroxylamine 6 and the oxidation of the obtained precursors produced NitRs 1, 2, 3, respectively, by using both classical oxidation methods7,20 and a novel procedure described in the literature.21 ToF-SIMS. ToF-SIMS is a powerful surface characterization technique to determine the composition and structure of molecules on surfaces through spectral analysis.22 The use of static SIMS for characterizing alkanethiol SAMs23 and aromatic thiol SAMs24 has been reported in the literature. Studies on the characterization of bulk NitRs by mass spectrometry techniques have recently (20) (a) Ullman, E. F.; Call, L.; Osiecki, J. H. J. Org. Chem. 1970, 35, 3623. (b) Ullman, E. F.; Osiecki, J. H.; Boocock, D. G. B.; Darcy, R. J. Am. Chem. Soc. 1972, 94, 7049. (21) Gorini, L.; Menichetti, S.; Caneschi, A. Synlett 2006, 6, 948. (22) ToF-SIMS: Surface Analysis by Mass Spectrometry; Vickerman, J. C., Briggs, D., Eds.; Surface Spectra IMPublications: Chichester, U.K., 2001. (23) (a) Tarlov, M. J.; Newman, J. G.; Langmuir 1992, 8, 1398. (b) Graham, D. J.; Price, D. D.; Ratner, B. D. Langmuir 2002, 18, 1518. (c) Graham, D. J.; Ratner, B. D. Langmuir 2002, 18, 5861. (24) (a) Tao, Y.-T.; Wu, C.-C.; Eu, J.-Y.; Lin, W.-L. Langmuir 1997, 13, 4018. (b) Cyganik, P.; Vandeweert, E.; Postawa, Z.; Bastiaansen, J.; Vervaecke, F.; Lievens, P.; Silverans, R. E.; Winograd, N. J. Phys. Chem. B 2005, 109, 5085. (c) Auditore, A.; Tuccitto, N.; Quici, S.; Marzanni, G.; Puntoriero, F.; Campagna, S.; Licciardello, A. Appl. Surf. Sci. 2004, 314, 231.

Mannini et al.

been published.25-27 Matrix-assisted desorption/ionization timeof-flight mass spectrometry (MALDI ToF-MS),25 electron impact (EI),26 and electrospray ionization mass spectrometry (ESI-MS)27 have been applied to the characterization of bulk NitRs. However, no ToF-SIMS studies of NitRs chemisorbed on surfaces have been reported to our knowledge, even if this technique can give access to relevant pieces of information on such systems. Positive ToF-SIMS spectra acquired on SAMs of NitRs on gold are much more informative than negative ones, in agreement with the results obtained on similar NitRs analyzed by MALDIToF,25 EI-MS,26 and ESI-MS.27 In the mass region between 100 and 200 m/z (Figure 1a), significant differences are observed in the three spectra. The spectrum of 2 is dominated by the Au+ peak at 197 m/z, while the spectrum of 3 also shows two intense peaks at 158 and 117 m/z assigned to the molecular ion fragments C7H14N2O2+ and C7H5N2+, respectively.27 In the spectrum of 1, intense peaks at 191, 150, and 151 m/z attributed to C10H11N2S+, C7H6N2O2+, and C7H7N2S+, respectively, are seen. Most interesting is the region at higher masses (200-800 m/z, Figure 1b,c), exhibiting similar fragmentation patterns of molecular ions and molecular ion fragments shifted by the right difference in molecular weight of the NitR analyzed. The characteristic peaks detected in the positive ToF-SIMS spectra of SAMs of 1, 2, and 3 on gold and their assignments are reported in Scheme 2. The detection of the biprotonated molecular ion [M+2H]+ in all three NitR spectra is consistent with literature data on bulk NitRs27 and confirm the absence of a cleavage of a C-S bond during the self-assembling process.28 It has to be noted that, in the spectra of 1 and 3, this peak is weak, while in the spectrum of 2, it is the most intense peak in the molecular ion mass region. Differences are observed in the relative intensity of the peaks occurring in the high mass region (>400 m/z, Figure 1c), with the spectrum of 1 showing more intense peaks at higher masses. In detail, dimer fragment and Au-dimer fragment cluster ions are detected in the spectrum of 1, indicating the formation of peroxy dimers in the SIMS experiment. The corresponding peaks are not detected in the spectra of 2 and 3; however, Au-molecular fragment cluster ions and molecular cluster ions were detected for these derivatives. Details of the peak assignments are reported in Table 1. The differences observed in the three NitR spectra can be attributed to different packing of the formed monolayers, as indicated by STM measurements. Indeed, studies of ToF-SIMS spectral fragmentation patterns23 during the assembly of thioalkane SAMs have shown an enhancement of longer hydrocarbon fragments at short assembling times, while, for longer times (i.e., fully packed, crystalline SAM), an enhancement of shorter hydrocarbon fragments is detected. In addition, long assembly times correspond to molecular ion clusters emitted from the gold surface.23 In agreement with this observation, the ToF-SIMS spectra of the three NitRs investigated show an increase in the relative intensity of larger mass fragments as the degree of ordering of SAMs decreases (NitR 1 < NitR 2 < NitR 3). STM. STM measurements confirmed large differences among the three deposited NitRs and assessed the deposition of a (25) Stroh, C.; Mayor, M.; von Ha¨nisch, C. Tetrahedron Lett. 2004, 45, 9623. (26) Ordina, J.; Garı´n, J.; Boulle, C.; Cirujeda, J.; Ju¨rgens, O., Veciana, J. Rapid Commun. Mass Spectrom. 1997, 2, 1103. (27) Smith, C. D.; Bartley, J. P.; Bottle, S. E. J. Mass Spectrom. 2002, 37, 897. (28) Zhong, C.-J.; Brush, R. C.; Anderegg, R. C.; Porter, M. D. Langmuir 1999, 15, 518. (29) Rampi, M. A.; Whitesides, G. M. Chem. Phys. 2002, 281, 373.

ToF-SIMS, STM, and ESR Analysis of NitR SAMs.

Langmuir, Vol. 23, No. 5, 2007 2393

Figure 2. In-air STM image (5.3 × 5.3 nm2) of the NitR 1 SAM on a gold surface: (a) unfiltered scan, the dark lines indicate the suggested primitive cell; (b) filtered with fast Fourier transform; (c) 3D representation of NitR 1; (d) the (2 × 2) suggested packing model. Table 1. Characteristic Peaks (m/z) in the Positive ToF-SIMS Spectra of SAMs of 1, 2, and 3 on Gold peak (m/z) 1

2

3

ion assignment

30 42 56 58 70 91 103

30 42 56 58 70 91 103

30 42 56 58 70 91 103 117

CH4N+ C2H4N+ C3H6N+ C3H8N+ C4H8N+ C7H7+ C7H5N+ C7H5N2+ C7H6N2O2+ C7H7N2S+ C7H14N2O2+ C10H11N2S+ Au+ [M-CH3-2O-S+3H]+ [M-CH3-O-S+3H]+ [M-2O+2H]+ [M-O+2H]+ [M+2H]+ [M+O]+ or [(M-CH3-O-S+2H)+Ph]+ [M+2O]+ or [(M-CH3-2O+2H)+Ph]+ [(M-CH3-2O-S+2H)+Au]+ [(M-CH3-O-S+2H)+Au]+ [(M-CH3-2O+2H)+Au]+ [(M-CH3-O+2H)+Au]+ [(M-CH3-S+2H)2+H]+ [(M-CH3+2H)2+H]+ [(M-CH3-O-S+3H)2+Au]+

150 151 158 191 197 203 219 249 265 281

197 279 295 325 341 357

507 523 467 531 633

197 217 233 263 279 295 309 325 413 429 445 461

molecular monolayer. SAMs obtained with 1 appear to be very hard to characterize by STM, and, in the experimental conditions we used, only a few small areas of the analyzed samples showed

a bidimensional ordering (Figure 2). In one of these structured regions, 0.6 × 0.6 nm2 features, which are in agreement with the size of NitR 1, are clearly visible. These dimensions suggest a (2 × 2) organization over the Au (111) surface, as presented in the model shown in Figure 2c, in which molecules should be linked perpendicularly to the surface. STM room-temperature data of the SAM obtained with 2 show a non-regular organized system (Figure 3). The presence of biphenyl groups instead of phenyl ones should enable stronger stacking interactions between neighboring NitRs, supplying to the molecules stronger lateral interactions. We can only roughly estimate the presence of 1.6 × 1.0 nm2 objects, in agreement with the presence of strongly tilted NitR 2 molecules, but no order was observed. This may be due either to the decreased conduction through longer molecules making difficult to find the correct imaging conditions,29 or to an augmented rotational freedom around the long molecular axis. Moreover, NitRs 1 and 2 can bind the gold surface assuming two different orientations, in analogy to what was observed in the literature for phenyl-containing thiols.24a This is responsible for the low order we noticed in the SAMs. The methyl group, which is close to the sulfur atom in both cases, appears to hinder a regular packing, as suggested by a raw calculation with a molecular modeling package. Thanks to the -CH2- spacer group between the sulfur atom and the aromatic ring, NitR 3 provides more ordered SAMs. In fact, this simple chemical modification reduces the degree of freedom of the molecules on the surface.24a Moreover, by introducing the CH2 spacer group, we stabilized the steric hindrance of the methyl group present in each system. This

2394 Langmuir, Vol. 23, No. 5, 2007

Figure 3. In-air STM image (50 × 50 nm2) of the NitR 2 SAM on a gold surface: (a) unfiltered scan; (b) 3D representation of NitR 2.

modification stabilizes a conformation of the molecule in which the methyl group is in a position closer to the methylenic spacer. This allows an easier link of the sulfur atom to the gold surface, as confirmed by molecular modeling calculations.18 In agreement with this hypothesis, STM images of NitR 3 SAMs present a stronger bidimensional ordering on relatively large areas (Figure 4). The typical signature of a regular SAMs is clearly distinguishable: dark spots ∼5 nm wide in the images correspond to the “pinholes”; these are a defect of the SAM.30 Moreover, at large scan area, it is also possible to notice the bidimensional lattice multidomain features of the monolayer characterized by features of 0.8 × 0.7 nm2, that can be modeled in agreement with the size of the deposited molecules, considering each molecule as being tilted approximately by 60° with respect to the surface plane and assuming a (3 × 2) adlayer structure with respect to the gold surface. Concluding, STM measurements in air show a different degree of ordering of SAMs of NitRs and evidence how local measurement differs from averaged measurement over the entire sample. ESR. X-Band ESR spectroscopy has been used to verify the effect of the deposition procedure on the magnetic properties of different NitRs and to get information about the dynamics and the degree of organization of the deposited molecules, in order to complement the information obtained by STM and ToF-SIMS techniques. Indeed, ESR acts as a local probe and it is, in principle, (30) Bucher, J. P.; Santesson, L.; Kern, K. Langmuir 1994, 10, 979.

Mannini et al.

an ideal technique to obtain such kind of information on organized monolayers, which can be derived from the g- and A- tensor anisotropy as well as from the analysis of the temperature dependence of the spectrum. However, the main problem that strongly limits the number of studies published in the literature up to date is the very small amount of material that is probed when a single monolayer is investigated, which is often close to the sensitivity limit of ESR.15,31 A typical ESR spectrum of a SAM of 1 obtained at room temperature is reported in Figure 5a, bottom. The detection of an ESR signal indicates the presence of a paramagnetic species on the gold surface. We note here that the spectral appearance does not appreciably change with time, thus indicating that the organization of the SAM is stable with time. The general features of the ESR spectra are essentially the same for the three radicals so that in the following we will just discuss the spectra of one derivative. Analysis of more subtle peculiarities could, in principle, give access to important pieces of information on the differences among motional properties and spin-spin interactions in different radicals. However, a suitable simulation program for quantitatively reproducing the dynamic effects in species containing more than one nuclei coupled to one unpaired electron has been reported in the literature only recently.32 As a consequence, we will restrict ourselves to a qualitative analysis of the system. A more detailed and quantitative analysis of the differences among the spectra of different radicals is currently in progress. The hyperfine structure of the signal, structured in five lines with an approximate spacing of 7.5 G and centered around g ) 2.009, indicates without any doubt that the deposited species is a NitR with the unpaired electron being delocalized on the two equivalent 14N (I ) 1) nuclei. The first piece of information can be gained by comparing the SAM spectrum with that obtained on samples prepared by drop casting and successive evaporation of the solvent: in this case, the hyperfine structure is lost due to the strong intermolecular exchange interactions, typical for a concentrated spin system (see Figure 5b). This indicates that the intermolecular interactions in the radical SAMs are reduced with respect to the sample obtained by drop casting, suggesting that the distance between each radical is increased, and the coverage is reduced. Furthermore, it rules out the possibility that the ESR spectra observed for SAMs are due to a simple evaporation of the solution on the surface with solid radical precipitation. Instead, these data clearly indicate that a real chemisorption process occurs. The comparison of the SAM spectrum to that of a NitR solution containing a known and comparable number of spins allowed us to obtain a rough estimate of the number of paramagnetic centers on the surface (∼5 × 1013 spin). By assuming a surface occupation for each radical of 6 × 6 Å2 and considering a surface of the gold slide of about 20 mm2, an approximately complete coverage of the surface is estimated. Notwithstanding the inherent limitations of quantitative ESR measurements and the approximations involved (neglecting the roughness of the gold surface and neglecting possible SAM defects), this result suggests a quite homogeneous coverage of the surface and confirms the presence of only one molecular layer. Further, it might be considered as an independent confirmation of the intermolecular distances estimated by STM. In this respect, it has to be considered that, with these distances, the calculated dipolar contribution would be large enough to hamper the detection of a resolved hyperfine structure. Indeed, for a pair of nitronyl nitroxides at a distance (31) Ruthstein, S.; Artzi, R.; Goldfarb, D.; Naaman, R. Phys. Chem. Chem. Phys. 2005, 7, 524.

ToF-SIMS, STM, and ESR Analysis of NitR SAMs.

Langmuir, Vol. 23, No. 5, 2007 2395

Figure 4. In-air STM image of the NitR 3 SAM on a gold surface: (a) 120 × 120 nm2 sized unfiltered scan; (b) 35 × 35 nm2 unfiltered scan; (c) digital zoom 8 × 8 nm2 of previous image, the dark lines indicate the primitive cell suggested; (d) the (3 × 2) suggested packing model; (e) two 3D representations of NitR 3.

of 6 Å, setting half a spin on the midpoint of each N-O bond,33 would result in a dipolar field of about 40 G. However, this value refers to a fixed configuration of neighboring molecules and does not take into account the dynamic of the paramagnetic groups, which is expected to produce an averaging of the anisotropic interactions. Indeed, the global spectral appearance suggests a quite large mobility of the radical on the surface. This is not completely unexpected because the methyl-thio linking group provides a weaker interaction with gold than that of thiol.19 A comparison of the SAM room-temperature spectrum with the temperature dependence of the solution spectra (Figure 5a) suggests that the motion of the radicals on the gold surface occurs on a time scale that is comparable to that of a viscous solution (around 180 K, just above the melting point of CH2Cl2). This indicates a slow dynamic of the radical molecules on the surface and suggests that molecular motion occurs through large oscillations that result in quasi-isotropic spectra. This interpretation is strengthened by the analysis of the temperature dependence of the SAM spectra, reported in Figure 6. At 150 K and at lower temperatures, the lines broaden so much that the spectrum collapses to a single broad transition. (32) Polimeno, A.; Zerbetto, M.; Franco, L.; Maggini, M.; Corvaja, C. J. Am. Chem. Soc. 2006, 128, 4734. (33) Benelli, C.; Caneschi, A.; Gatteschi, D.; Pardi, L.; Rey, P. Inorg. Chem. 1989, 28, 3230.

Upon heating back to room temperature, the original spectrum is restored. This behavior is a signature of the freezing of the radical motion occurring between 180 and 150 K, resulting in a non-averaging of the anisotropic interactions and then in a broadening of the lines, which leads to the appearance of a single broad transition. The reversibility of the phenomenon indicates the stability of the SAMs with respect to thermal cycling and thus rules out the idea that the spectral changes were due to a disruption of the SAMs with consequent formation of a powderlike phase. Finally, in Figure 7, the variation of the spectrum as a function of the angle of the applied field with the gold surface is reported. The virtually complete absence of orientation dependence of the spectra is evident. At first glance, this result might appear to be in contradiction with STM observation, which indicated the presence of ordering within the monolayer. However, this is not unexpected if we consider that the qualitative analysis of the spectra suggested a quasi-isotropic motion of the radicals and that this motion occurs on a time scale that is slow for ESR spectroscopy but is fast and thus undetectable for STM. We suggest that this spectral behavior might result from the sum of the contribution of regions with slow, quasi-isotropic dynamics of the radicals and with different relative orientations. These can easily form as a consequence of differently oriented domains of the gold substrate and the roughness of the polycrystalline surface.3

2396 Langmuir, Vol. 23, No. 5, 2007

Mannini et al.

Figure 7. Angular dependence of the room-temperature ESR spectrum of a SAM of 1 deposited on gold; the angle indicated in the figure is that made by the magnetic field with the normal to the gold slide.

paramagnetic molecules with an approximately complete coverage of the surface. The spectra show an intermediate behavior between that of a fluid solution and the one observed in the solid state, thus suggesting a quite large mobility of the radical on the surface. Finally, the polycrystalline nature of the gold surface increases the variability in possible SAMs orientations and thus results in the absence of the orientation dependence of the spectrum.

Conclusions

Figure 5. (a) Temperature dependence of the ESR solution spectra of 1 compared to the room-temperature ESR spectrum of its SAM. (b) Comparison of a frozen-solution ESR spectrum of 1 with that of a drop-cast sample.

Figure 6. Temperature and time evolution of the ESR spectrum of a SAM of 1 deposited on gold. The magnetic field is oriented along the normal to the gold slide.

We can then conclude that the ESR analysis prove unambiguously that the paramagnetic character of the molecules is retained after deposition on gold, and the SAMs are really made up of NitRs. ESR spectra confirm the formation of a single layer of (34) Messina, P.; Mannini, M.; Caneschi, A.; Gatteschi, D.; Sorace, L.; Sigalotti, P., Sandrin, C.; Pittana, P.; Manassen, Y. J. Appl. Phys., in press (doi:10.1063/ 1.2434832). (35) (a) Balatsky, A. V.; Manassen, Y.; Salem, R. Philos. Mag. B 2002, 82, 1291. (b) Sleator, T.; Hahn, E. L.; Hilbert, C.; Clarke, J. Phys. ReV. Lett. 1985, 55, 1742.

We presented a complete study of a series of paramagnetic purely organic molecules organized on Au(111) in a bidimensional array through the self-assembling of a monolayer technique. The ToF-SIMS technique demonstrated the occurrence of the chemisorption process on the surface. Further, it indicated that this process occurs by keeping intact the chemical structure of the organic radicals and supported the ordering evaluation estimated through the STM. The latter technique indicated the formation of the monolayers and showed how the structures of the molecules can be tuned in order to achieve a real ordering on the surface. This was accomplished by replacing the phenyl group with a benzyl group that improves the order in the assembling step by stabilizing a conformation more favorable for grafting. On the other hand, the increase of π-π stacking interaction through substitution of the phenyl group with a biphenyl one does not lead to a significant improvement of the ordering because the degrees of freedom of the molecule are increased. Finally, ESR spectra evidenced that paramagnetic behavior was maintained after the deposition procedure and allowed us to estimate a complete coverage of the surface by NitR derivatives. Further, the absence of angular dependence in the spectra indicated that the small regions of local ordering observed by STM do not sum to give a large-scale order, due to the intrinsic properties of the gold substrate. The spectra also suggest a quite large mobility of the radical on the surface, probably due to the weak interaction with gold provided by the methyl-thio linking group. To our knowledge, there are no previous papers concerning the STM imaging of NitR monolayers. We note here that the electron spin noise (ESN)-STM6f technique has been recently used to characterize radicals on surfaces prepared via drop casting and dip-and-wet techniques.6g,h,34 In all previous cases, no bidimensional ordering is observed because this is a prerogative of carefully evaporated systems and SAMs. Therefore our approach poses the challenge of the detection of an ESN-STM signal of ordered spin lattice on the surface. This technique, also

ToF-SIMS, STM, and ESR Analysis of NitR SAMs.

known as ESR-STM, is based on the detection of the fluctuation induced by a magnetic field of the electronic spins of a single molecule localized by STM.6g However, as no external energy is supplied to the system, we prefer to refer to this technique as a noise spectroscopy.35 On the basis of the results discussed in this paper, the organized monolayers of NitR derivatives are supposed to be good candidates for ESN-STM characterization, which is currently in progress. This will allow exploring the spin properties of single molecule through this technique. This analysis could provide important insights before potential applications in molecularbased spintronics devices can be developed. Acknowledgment. Authors gratefully acknowledge Dr. F. Buatier de Mongeot for his assistance in preliminary STM

Langmuir, Vol. 23, No. 5, 2007 2397

investigations and Dr. P. Messina for useful discussion concerning possible investigation through ESN-STM. The financial support was provided by EU through NoE MAGMANet (NMP3-CT2005-515767), and QueMolNa (MRTN-CT-2003-504880), by Ente Cassa di Risparmio di Firenze, and by Italian MIUR through FIRB and PRIN projects. F.M.P. thanks the MIUR for financialsupport through the program “Incentivazione alla mobilita` di studiosi stranieri e italiani residenti all’estero”. Supporting Information Available: Detailed description of each step of the synthesis of 2 and 3. This material is available free of charge via the Internet at http://pubs.acs.org. LA062028F