2494
Langmuir 2008, 24, 2494-2500
Effects of Changes in the Interparticle Separation Induced by Alkanethiols on the Surface Plasmon Band and Other Properties of Nanocrystalline Gold Films Ved Varun Agrawal, Neenu Varghese, G. U. Kulkarni, and C. N. R. Rao* Chemistry and Physics of Material Unit and DST Unit on Nanoscience, Jawaharlal Nehru Centre for AdVanced Scientific Research, Jakkur P.O., Bangalore 560064, India ReceiVed October 18, 2007. In Final Form: December 11, 2007 Effects of changing the interparticle separation on the surface plasmon bands of ultrathin films of gold nanoparticles have been investigated by examining the interaction of alkanethiols of varying chain length on nanocrystalline gold films generated at the organic-aqueous interface. Adsorption of alkanethiols causes blue-shifts of the surface plasmon adsorption band, the magnitude of the shift being proportional to the chain length. The disordered nanocrystals thus created (λmax, 530 m) are in equilibrium with the ordered nanocrystals in the film (λmax, 700 m) as indicated by an isosbestic point around 600 nm. Long chain thiols disintegrate or disorder the gold films more effectively, as demonstrated by the increased population of the thiol-capped gold nanocrystals in solution. The rate of interaction of the thiols with the film decreases with the decreasing chain length. The effect of an alkanethiol on the spectrum of the gold film is specific, in that the effects with long and short chains are reversible. The changes in the plasmon band of gold due to interparticle separation can be satisfactorily modeled on the basis of the Maxwell-Garnett formalism. Spectroscopic studies, augmented by calorimetric measurements, suggest that the interaction of alkanethiols involves two steps, the first step being the exothermic gold film-thiol interaction and the second step includes the endothermic disordering process followed by further thiol capping of isolated gold particles.
Introduction Mesoscalar assemblies of metal nanocrystals are of considerable interest since both short and long range interactions coexist in these systems and compete to give rise to a host of interesting structural, optical, electrical and magnetic, and other properties.1-8 A number of studies have aimed to explore their potential in technological applications.9-13 A nanocrystal assembly, analogous to an atomic lattice, derives its properties from the nature of the nanocrystal (also called a superatom14), from its geometric arrangement, and from the nature of the material in the intervening space. In this sense, it is a unique combination of zero and two dimensionalities. Specifically in the case of a simple twodimensional hexagonal assembly, the main variables are the nanocrystal diameter, the interparticle separation, and the nature of the spacer molecule as well as the domain size. There are * Corresponding author. E-mail:
[email protected]. (1) Zhong Lin, W. AdV. Mater. 1998, 10, 13-30. (2) Rao, C. N. R.; Thomas, P. J.; Kulkarni, G. U. Nanocrystals: Synthesis, Properties and Applications; Springer-Verlag: Berlin, 2007. (3) Rechberger Hohenau, A.; Leitner, A.; Krenn, J. R.; Lamprecht, B.; Aussenegg, L. Opt. Commun. 2003, 220, 137. (4) Kreibig, U.; Vollmer, M. Optical Properties Of Metal Clusters; SpringerVerlag: Berlin, 1995. (5) Pelka, J. B.; Brust, M.; Gierlowski, P.; Paszkowicz, W.; Schell, N. Appl. Phys. Lett. 2006, 89, 063110. (6) Terrill, R. H.; Postlethwaite, T. A.; Chen, C.-h.; Poon, C.-D.; Terzis, A.; Chen, A.; Hutchison, J. E.; Clark, M. R.; Wignall, G. J. Am. Chem. Soc. 1995, 117, 12537-12548. (7) Wuelfing, W. P.; Green, S. J.; Pietron, J. J.; Cliffel, D. E.; Murray, R. W. J. Am. Chem. Soc. 2000, 122, 11465-11472. (8) Pileni, M. P.; Lalatonne, Y.; Ingert, D.; Lisiecki, I.; Courty, A. Faraday Discuss. 2004, 125, 251. (9) Haes, A. J.; Zou, S.; Schatz, G. C.; VanDuyne, R. P. J. Phys. Chem. B 2004, 108, 109-116. (10) Joseph, Y.; Besnard, I.; Rosenberger, M.; Guse, B.; Nothofer, H. G.; Wessels, J. M.; Wild, U.; Knop-Gericke, A.; Su, D.; Schlogl, R.; Yasuda, A.; Vossmeyer, T. J. Phys. Chem. B 2003, 107, 7406-7413. (11) Krasteva, N.; Besnard, I.; Guse, B.; Bauer, R. E.; Mullen, K.; Yasuda, A.; Vossmeyer, T. Nano Lett. 2002, 2, 551-555. (12) Lu, Y.; Liu, G. L.; Lee, L. P. Nano Lett. 2005, 5, 5-9. (13) Yu, A.; Liang, Z.; Cho, J.; Caruso, F. Nano Lett. 2003, 3, 1203-1207. (14) Roduner, E. Chem. Soc. ReV. 2006, 35, 583.
several reports on the synthesis of well-defined two-dimensional assemblies formed by suitably derivatizing the nanocrystals,15 by assembling at the air-water interface,16,17 or by growing nanocrystals in predefined locations.18,19 These methods have enabled a greater degree of control on the structural and physical properties of nanocrystal assemblies. Optical properties of metal nanocrystal assemblies find immense applications ranging from bioanalysis to glass industry. The nature of the ligating molecule or of the medium and the proximity of particles within an assembly can influence the localized optical excitations and cause measurable shifts in the optical spectra.20-24 For example, Mulvaney and co-workers25 have monitored the localized surface plasmon band from an assembly of gold nanoparticles coated with SiO2 that caused red shifts with decreasing shell thickness. The near-field coupling interaction between surface plasmon modes of neighboring metal nanoparticles have been investigated by Sih and Wolf26 in the case of a thin film of oligothiophene-linked gold nanoparticles. They find that a higher dielectric constant of the medium leads to weaker coupling between the particles. The surface plasmon resonance band of gold nanoparticles assemblies embedded in glass and quartz matrices has been exploited to obtain a range (15) Collier, C. P.; Vossmeyer, T.; Heath, J. R. Annu. ReV. Phys. Chem. 1998, 49, 371. (16) Heath, J. R.; Knobler, C. M.; Leff, D. V. J. Phys. Chem. B 1997, 101, 189-197. (17) Kim, B.; Tripp, S. L.; Wei, A. J. Am. Chem. Soc. 2001, 123, 7955-7956. (18) Chandrasekharan, N.; Kamat, P. V. Nano Lett. 2001, 1, 67-70. (19) Haynes, C. L.; McFarland, A. D.; Zhao, L.; Van Duyne, R. P.; Schatz, G. C.; Gunnarsson, L.; Prikulis, J.; Kasemo, B.; Kall, M. J. Phys. Chem. B 2003, 107, 7337-7342. (20) Markel, V. A. J. Mod. Opt. 1993, 40, 2281-2291. (21) Zhao, L.; Kelly, K. L.; Schatz, G. C. J. Phys. Chem. B 2003, 107, 73437350. (22) Okamoto, T.; Yamaguchi, I. J. Phys. Chem. B 2003, 107, 10321-10324. (23) Meli, M. V.; Lennox, R. B. J. Phys. Chem. C 2007, 111, 3658-3664. (24) Liz-Marzan, L. M. Langmuir 2006, 22, 32-41. (25) Ung, T.; Liz-Marzan, L. M.; Mulvaney, P. J. Phys. Chem. B 2001, 105, 3441-3452. (26) Sih, B. C.; Wolf, M. O. J. Phys. Chem. B 2006, 110, 22298-22301.
10.1021/la703237m CCC: $40.75 © 2008 American Chemical Society Published on Web 02/16/2008
Interparticle Separation Induced by Alkanethiols
of transmitting colors.23,27 A nanoscale optical biosensor based on localized surface plasmon resonance properties of noble metal nanoparticles has been realized.9 Metal nanocrystal assemblies have been employed as solid substrates for surface enhanced Raman scattering (SERS) of adsorbed molecules,28,29 and there are studies relating SERS activity to the surface plasmon resonance behavior.30 We have been interested in investigating changes in the UVvisible spectra of metal nanocrystal assemblies caused by a variation in the interparticle distance. For this purpose, we have prepared nanocrystalline gold films at the liquid-liquid interface.31 Unlike the literature procedures that involve at least two steps, this is a simple one-step synthesis, wherein an organometallic compound dissolved in an organic layer reacts with a reducing agent present in the aqueous layer at the interface. Careful studies have demonstrated that the material formed at the interface is an ultrathin nanocrystalline film consisting of closely packed metal nanocrystals, a few nanometers in diameter. Increasing the temperature to 75 °C increases the mean diameter to 15 nm.32 The film at the interface is essentially free-standing and can be readily taken out of the interface and deposited on a solid substrate of choice.31-33 In the present study, we have investigated how the surface plasmon band responds to the changing interparticle separation between the gold nanocrystals when the film interacts with alkanethiols of different chain lengths, CH3(CH2)nSH. The surface plasmon of gold serves as a good diagnostic tool to monitor such changes. The study not only provides an insight into the collective behavior of the nanocrystals in the film and its dependence on the interparticle distance but also throws light on the nature of interaction of alkanethiols with metal nanoparticle assemblies. Thus, we find that the electronic interaction between the nanoparticles decreases linearly with the increase in the separation between the nanoparticles as determined by the length of the alkane chain. Experimental Details An ultrathin nanocrystalline gold film was prepared at the toluenewater interface using the procedure described elsewhere.33 Au(PPh3)Cl was used as the organometallic resource. In a typical preparation, 10 mL of 1.66 mM solution of the metal precursor in toluene formed the top organic layer, with 16 mL of 6.25 mM aqueous NaOH solution forming the bottom aqueous layer. The reducing agent, tetrakis(hydroxymethyl)phosphonium chloride (THPC, Fluka) was gently added to the bottom layer (330 µL of 50 mM) to initiate the reaction at the interface.34 Nanocrystalline films were obtained at room temperature (28 °C) as well as at 75 ( 1 °C after 3 h of the reaction. Water used in the experiments was double distilled using a quartz apparatus. For adsorption and calorimetric studies, alkanethiols procured from Aldrich (purity, 95%) were used. Transmission electron microscope (TEM) measurements were carried out using a JEOL 3010 operating at 300 kV. UV-visible absorption spectra of the films were recorded using a Perkin-Elmer Lambda 900 spectrometer. For alkanethiol adsorption studies, the films on glass slides were introduced into a thiol solution present (27) De, G.; Rao, C. N. R. J. Phys. Chem. B 2003, 107, 13597-13600. (28) Baia, L.; Baia, M.; Popp, J.; Astilean, S. J. Phys. Chem. B 2006, 110, 23982-23986. (29) Baker, G. A.; Moore, D. S. Anal. Bioanal. Chem. 2005, 382, 1751-1770. (30) Alvarez-Puebla, R.; Cui, B.; Bravo-Vasquez, J. P.; Veres, T.; Fenniri, H. J. Phys. Chem. C 2007, 111, 6720-6723. (31) Rao, C. N. R.; Kulkarni, G. U.; Agrawal, V. V.; Gautam, U. K.; Ghosh, M.; Tumkurkar, U. J. Colloid Interface Sci. 2005, 289, 305-318. (32) Agrawal, V. V.; Kulkarni, G. U.; Rao, C. N. R. J. Phys. Chem. B 2005, 109, 7300-7305. (33) Rao, C. N. R.; Kulkarni, G. U.; Thomas, P. J.; Agrawal, V. V.; Saravanan, P. J. Phys. Chem. B 2003, 107, 7391-7395. (34) Duff, D. G.; Baiker, A.; Gameson, I.; Edwards, P. P. Langmuir 1993, 9, 2310-2317.
Langmuir, Vol. 24, No. 6, 2008 2495
Figure 1. Optical absorption spectra of (a) an Au nanocrystalline film prepared at the liquid-liquid interface and deposited on a glass substrate, (b) a sol obtained by sonicating the film in toluene for 3 min, (c) a sol obtained by derivatizing the nanocrystals from the film using hexadecanethiol in toluene. The corresponding TEM images are shown. in a cuvette. The reference cuvette contained a clean glass slide immersed in the thiol solution. Isothermal titration calorimetric (ITC) experiments were performed using a Microcal VP-ITC instrument at 30 °C. The instrument consists of two identical cells, one sample cell, and one reference cell. In a typical experiment, the sample cell of the microcalorimeter was filled with 0.9 mL of 6.25 mM solution of NaOH and 37.5 µL of 50 mM solution of THPC. A 0.85 mL of 1.5 mM solution of Au(PPh3)Cl in toluene was added to the cell, and the solution was kept undisturbed for 23 h for the formation of the nanocrystalline gold film. The reference cell was filled with 0.9 mL of distilled water and 0.85 mL of toluene, and both cells were maintained at the same temperature. A syringe of 280µL capacity was filled with 1.73 M solution of the alkanethiol. A 5µL amount of a 1.73 M solution of alkanethiol taken in the syringe was injected in equal intervals of 5 min to the mixture (film) in the sample cell, and the heat absorbed for each injection was measured. Control experiments were carried out by the addition of the same concentration of alkanethiol to 0.9 mL of distilled water and 0.85 mL of toluene taken in the sample cell and subtracted with the experimental data to remove the dilution effects of alkanethiol.
Results and Discussion Figure 1 shows how the surface plasmon band of a nanocrystalline gold film changes on perturbation through sonication or interaction with an alkanethiol. Against each spectrum, a TEM image is shown to demonstrate the nature of the nanostructure present. The spectrum corresponding to a nanocrystalline film obtained at the liquid-liquid interface at room temperature exhibits a broad band centered at 700 nm, with a weak shoulder at ∼540 nm. While the absorption band around 540 nm is characteristic of surface plasmons from uncoupled nanocrystals,35,36 the higher wavelength band at 700 nm is a typical response from an assembly of interacting nanocrystals.4,37 As evident from the corresponding TEM image (Figure 1), the nanocrystalline film consists of close-packed assembly of particles with ∼1 nm spacing due to the organic shell. At such close distances, one would expect electronic coupling between the neighboring particles to be significant. Furthermore, the number of interacting (35) Creighton, J. A.; Eadon, D. G. J. Chem. Soc., Faraday Trans. 1991, 87, 3881. (36) Granqvist, C. G.; Hunderi, O. Phys. ReV. B 1977, 16, 3513. (37) Gunnarsson, L.; Rindzevicius, T.; Prikulis, J.; Kasemo, B.; Ka¨ll, M.; Zou, S.; Schatz, G. C. J. Phys. Chem. B 2005, 109, 1079.
2496 Langmuir, Vol. 24, No. 6, 2008
Agrawal et al.
canethiol (100 mM in toluene) for increasing time periods. The surface plasmon band of the pristine film undergoes a gradual blue-shift from 697 to 560 nm, with an accompanying decrease in the absorption intensity. The projected length of a hexadecanethiol molecule being 2.2 nm, the average interparticle distance is expected to increase upon thiol adsorption, from ∼1 nm (see top TEM image (a) in Figure 1) up to 4.4 nm. The increasing interparticle distance reduces the effective electronic coupling between the nanocrystals, thereby causing a blue-shift of the surface plasmon band. There is an isosbestic point38 at 600 nm indicating the presence of an equilibrium between two species with characteristic absorption maxima at 690 and 560 nm. The equilibrium is clearly between the thiol-covered film and the thiol-capped nanocrystals formed by disordering of the film. Figure 2b shows a plot of the relative intensity at 700 nm as a function of time, for three different concentrations of the thiol (1, 10, and 100 mM), to demonstrate how the intensity drops quite rapidly during the first 10 min when the nanocrystalline film interacts with a 100 mM thiol solution, exhibiting only a negligible decrease in intensity after longer periods. For lower concentrations of the thiol, the initial drop is less steep (particularly for 1 mM), followed by a gradual decrease over 100-150 min. It appears that the nanocrystal surface gets rapidly covered by the thiol molecules initially, further adsorption becoming gradual as additional surface sites become difficult to access. This is reminiscent of the two-stage adsorption process known in the case of self-assembled monolayers of alkanethiols on plane metal surfaces.39,40 It is known that molecules crowd the surface during the initial process, followed by a slow rearrangement to a wellordered monolayer. Alkanethiol adsorption on metal nanoparticle surfaces has been treated as a modified first-order process in earlier studies.41 Thus, the time variation of the optical absorbance (A) is described by, Figure 2. (a) Absorption spectra of a Au nanocrystalline film immersed in a toluene solution of hexadecanethiol (100 mM) recorded at intervals of 4 min at 15 °C. (b) Variation in the absorption intensity at 700 nm normalized with respect to the initial intensity, for three different concentrations of the thiol. Initial part (I) shows a linear variation (dashed) while the latter part (II) is fitted to eq 1 (see Results and Discussion).
nanocrystals may be very large, as the assembly can extend as in the present case, to several hundred microns. This results in collective absorption by the nanocrystals but red-shifted by as much as 160 nm from the position of the normal surface plasmon band of gold nanocrystals (see spectrum a). The collective absorption is affected when the nanocrystalline film is subjected to ultrasonication as can be seen in spectrum ‘b’. There is a rise in absorption at 540 nm followed by an ill-defined broad feature. The TEM image shown alongside depicts fragments of the extended assembly, comprising smaller assemblies of a few tens of particles, with a similar interparticle spacing (1 nm). The absence of a clear red-shift and the broad nature of the spectrum (b) arise from the reduction in the domain size of the interacting particles. In the limiting case of an organosol formed by derivatizing the nanocrystals in the film with hexadecanethiol, we obtain a well-defined band around 530 nm (spectrum c), typical of uncoupled colloidal gold nanocrystals. Accordingly, the TEM image shows isolated nanocrystals.32 The observed contrast between the surface plasmon band of the nanocrystalline film and isolated nanocrystals led us to systematically investigate the effect of alkanethiol adsorption on the nanocrystalline film. For this purpose, we chose nanocrystalline films of gold prepared at 75 °C with a mean particle diameter of 15 nm. Figure 2a shows the absorption spectra from the nanocrystalline film following the interaction with hexade-
dA ) -(k)C1/2∆Ar(1 - ∆Ar) exp(-b∆Ar) dt
(1)
where C is the thiol concentration, k the apparent rate constant, and b an empirical parameter. The relative absorbance is, ∆Ar ) (A(t) - A(∞))/(A(0) - A(∞)). This equation is not applicable to the initial part of the curve where the drop in intensity is rapid but is well suited to deal with the gradual change. The former, when fitted to a linear behavior, yields rate constants of 0.01, 0.02, and 0.06 mol s-1, respectively, for 1, 10, and 100 mM thiol concentrations. Satisfactory fittings have been obtained for the rest of the data using eq 1, with apparent rate constants (k) of 0.79, 0.54, and 0.16 dm3/2 s-1 mol-1/2 for 1, 10, and 100 mM, thiol concentrations, respectively. The apparent rate constants are concentration dependent (C1/2 dependence). The concentration independent rate constant is found to be ∼2 dm3 mol-1 s-1. Figure 3 depicts the effect of temperature on the interaction of hexadecanethiol with the gold nanocrystals in the film. The thiol adsorption process is much slower at 15 °C, the time taken being ∼150 min compared to 120 min at 25 °C. It is even slower at lower temperatures (10 °C and 5 °C) as seen from the plot of the intensity values at 700 nm against time (Figure 3a). The data in the figure were fitted with eq 1 to determine the values of the apparent rate constant, k. Figure 3b shows a plot of ln(k) vs 1/T, (38) Lim, I.-I. S.; Mayue, M. M.; Luo, J.; Zhong, C.-J. J. Phys. Chem. B 2005, 109, 2578. (39) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321-335. (40) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. ReV. 2005, 105, 1103-1170. (41) Bellino, M. G.; Calvo, E. J.; Gordillo, G. Phys. Chem. Chem. Phys. 2004, 6, 424.
Interparticle Separation Induced by Alkanethiols
Figure 3. (a) Variation in the absorption intensity at 700 nm (normalized with respect to the initial intensity) at 5 °C, 10 °C, 15 °C, and 25 °C. Initial part (I) shows a linear variation (dashed) while the latter part (II) is fitted to eq 1 (see Results and Discussion). (b) Arrhenius plot where the slope corresponds to an activation energy of 42 kJ/mol.
where the fit to the data is linear, corresponding to an Arrhenius behavior with an activation energy of 40 kJ/mol. Considering the activation energy of 29 kJ/mol for adsorption of alkanethiol on gold42 and 0.84 kJ/mol per methylene unit of the alkane chain,39 we estimate the total activation energy for hexadecanethiol adsorption on the gold surface to be 42 kJ/mol, which is close to the value obtained in the present experiment. The interaction of alkanethiols of different chain lengths with the nanocrystalline Au films has been examined in some detail. Figure 4 gives a plot of the absorbance values at 700 nm normalized with respect to the initial value against the time period that a film is immersed into a thiol solution (100 mM). The relative absorbance decreases in all the cases, more so at longer chain lengths. Thus, the chain length of the thiol has a remarkable effect on the rate of adsorption, the higher the chain length the faster being the thiol adsorption.43 The data in Figure 4a were fitted with eq 1 in order to determine the apparent rate constants for the different alkanethiols. The rate constant increases linearly with chain length as shown in Figure 4b. Thus, the longer chain alkanethiols are more effective in separating the nanocrystals within the film, causing increasing larger blue-shifts of the plasmon band. Thus, the interaction between the gold nanocrystals increases linearly with decrease in chain length, akin to the observation made on alkanethiol adsorption on gold surfaces.44 (42) Lavrich, D. J.; Wetterer, S. M.; Bernasek, S. L.; Scoles, G. J. Phys. Chem. B 1998, 102, 3456-3465. (43) Ulman, A. Chem. ReV. 1996, 96, 1533. (44) Schreiber, F. Prog. Surf. Sci. 2000, 65, 151.
Langmuir, Vol. 24, No. 6, 2008 2497
Figure 4. (a) Time variation in the absorption intensity at 700 nm normalized with respect to the initial intensity, for alkanethiols of different chain lengths (100 mM, 25 °C). The curves are fitted to eq 1 (see Results and Discussion). (b) A plot of the apparent rate constant (k) versus the number of carbon atoms in the alkanethiol, n.
Figure 4a suggests that after a sufficient time interval, the absorbance of the nanocrystalline film tapers off, indicating saturation coverage of the thiol molecule. The time interval at which the saturation is reached clearly depends on the rate of thiol adsorption or the length of the alkanethiol molecule, while the relative absorbance at saturation is related to the effective change in the interparticle distance. This is also evident from the position of the absorption maximum at saturation coverage (Figure 5). While the band is positioned at ∼700 nm for the as-prepared film, it is blue-shifted after thiol adsorption, the shift increasing with the chain length. Figure 5b shows how the position of the absorption band varies with the alkane chain length or interparticle distance. The shift increases rapidly with chain length at relatively small chain lengths (up to C8) and becomes somewhat gradual for longer chain lengths. The spectral changes shown in Figure 5 can be explained based on the Maxwell-Garnett formalism.45,46 Assuming close packing, the optical absorption by a granular film depends on its packing fraction,
f ) 0.741
(r +r s)
3
(2)
(45) Garnett, J. C. M. Philos. Trans. R. Soc. London Ser. A 1904, 203, 385. (46) Abeles, B.; Sheng, P.; Coutts, M. D.; Arie, Y. AdV. Phys. 1975, 24, 407-461.
2498 Langmuir, Vol. 24, No. 6, 2008
Agrawal et al.
Figure 6. A Schematic showing the process of thiol adsorption and increasing disorder in the nanocrystalline film.
Figure 5. (a) Absorption spectra obtained after treating the Au nanocrystalline films with alkanethiols of different chainlengths (n ) number of carbon atoms in the alkane chain). In each case, the saturation adsorption was marked by no further changes in absorbance with time. (b) Absorption maximum versus the number of carbon atoms in the alkane chain. The values calculated using MaxwellGarnett formalism (circles) are shown along with the experimental values (squares). The latter may be fitted to an exponential behavior with a characteristic decay length of 2 nm.
r being the radius of the grain (nanoparticle) and s the interparticle separation. In our nanocrystalline films, the size of the nanocrystals (r) is almost constant (∼15 nm), as all the films were prepared under similar conditions. The spectral changes take place primarily due to the changes in the value of s (length of the adsorbed thiol molecule as well as coverage). By inserting suitable values for f, we have calculated absorption intensities using,
ζ)
(
4πnMGh λ
D(ω) - M MG ) M 1 + 3f D(ω)(1 - f) + M(f + 2)
(3)
)
(4)
Here ζ is the absorption intensity, nMG is the real part of the effective complex index of refraction λ the wavelength, h is the film thickness, M the dielectric constant of the matrix, D Drude dielectric function, and MG the Maxwell-Garnett dielectric function. The calculated absorption maxima are plotted against alkane chain length in Figure 5b. We see that the calculated values show a monotonic variation and are similar to the experimental values at shorter chain lengths but stand underestimated for hexadecane (f, 0.51) and octadecanethiols (f, 0.48). This is understandable since the Maxwell-Garnett formalism holds well for films with a higher volume fraction of nanocrystals (f ∼ 0.67) or for shorter lengths of the spacer. We have not taken
into account possible interdigitation of molecules, relevant specially for longer thiols;47 furthermore, the longer chains tend to curl up. It is important to understand the processes responsible for the changes in the surface plasmon bands of the Au nanocrystalline films occurring on addition of alkanethiols. The primary process involves the interaction of the alkanethiol with the nanocrystalline Au film, causing disorder in the film due to the capping of individual nanocrystals by the thiol molecules (see the schematic diagram in Figure 6). Such disorder causes an increase in the separation between nanocrystals relative to that present in the pristine film. The separation between the particles is determined by the length of alkane chain. Thus, the blue-shift of the surface plasmon band increases with the increasing chain length because of the increased separation between the nanocrystals. In other words, the interaction between the particles decreases with the increase in chain length as revealed by the linear plot in Figure 5b. The effect of chain length is also seen in terms of the rate of disordering or disintegration of the nanocrystalline film, where the longer chain thiols are more effective, resulting in a higher rate of film disintegration (Figure 4). The extreme case of disordered nanocrystalline film is one where the nanocrystals present in the film get fully separated and go into the solution, giving rise to isolated thiol-capped nanocrystals. There is a small concentration of such nanocrystals in solution as indicated by the pink-colored solutions as well as by the surface plasmon band of the organosol around 530 nm. The concentration of such nanocrystals in the organosol increases with the chain length of thiol, since long chain thiols such as hexadecanethiol are more effective in causing disorder and disintegrating the film. Considering that stage 1 in Figure 6 is essentially an exothermic process while stage 2 involves an endothermic disordering process, possibly followed by an exothermic thiol binding process, we have carried out calorimetric experiments on the interaction of nanocrystalline gold films formed at the liquid-liquid interface with alkanethiols. Addition of small amounts of an alkanethiol (47) Thomas, P. J.; Kulkarni, G. U.; Rao, C. N. R. J. Phys. Chem. B 2000, 104, 8138-8144.
Interparticle Separation Induced by Alkanethiols
Langmuir, Vol. 24, No. 6, 2008 2499
Figure 7. Heat content change per mole of the alkanethiols obtained by isothermal titration calorimetric (ITC) experiments versus the molar ratio of alkanethiol (a) C4 thiol, (b) C12 thiol to Au(PPh3)Cl. Solid lines are a guide to the eye. Figure 9. Absorption spectra of (a) Au nanocrystalline film obtained after successive dipping in (b) hexadecanethiol, (c) hexanethiol, and (d) hexadecanethiol solutions.
Figure 8. Dependence of the molar ratio of the alkanethiol to Au(PPh3)Cl at which the peak appears due to maximum endothermicity on the number of carbon atoms in the alkanethiol, n.
first gives rise to an exothermic reaction. With increase in the thiol concentration, the exothermicity decreases and the process become endothermic. This results in a peak in a plot of the heat change versus the molar ratio of the thiol to the Au precursor. Figure 7 shows typical data obtained from the isothermal titration calorimetric experiments48 with two alkanethiols. The peak appears at a lower molar ratio of the surfactant when the alkane chain is longer. The position of the peak relates to the effectiveness of the thiol in disintegrating or disordering the Au film. We have plotted the concentration at which the peak appears against the chain length or the number of carbon atoms in the alkane chain, n, in Figure 8. We see a near linear relationship with the peak appearing at lower concentrations with the increase in chain length. These observations are consistent with the changes in the plasmon absorption band described in Figures 4 and 5 and the processes described in Figure 6. The first exothermic process is due to the binding of alkanethiols to the gold nanoparticles in the film (stage 1). The endothermic process corresponds to the disordering or disintegration of the film in stage 2. Note that stage 2 is more complex. After the disordering or disintegration of the film, isolated nanocrystals get created which are then capped by the thiol molecules present in the solution. The last step is exothermic. After the endothermic peak, there is indeed an exothermic feature (Figure 7). In order to demonstrate the alkanethiol (interparticle distance) specific interaction between gold nanocrystals, the surface (48) Ladbury, J. E.; Chowdhry, B. Z. Chem. Biol. 1996, 3, 791-801.
plasmon features have been recorded after interaction with hexanethiol and hexadecanethiol alternately. Figure 9 shows the spectra of a film sequentially treated with 100 mM of hexadecanethiol and hexanethiol, till saturation. The band of the asprepared film is blue-shifted by 140 nm after dipping in hexadecanethiol and is red-shifted by as much as 120 nm after dipping in hexanethiol. The latter thiol being shorter in length brings the particles closer, causing a red-shift. This observation confirms the occurrence of interdiffusion and exchange of alkanethiols of different chain lengths, known in the case of self-assembled monolayers.44 This process cannot, however, be repeated indefinitely, as the spectral intensity diminishes with each dip (see Figure 9) as the nanocrystals from the film migrate into the solution.
Conclusions The present study of the effect of adsorption of alkanethiols of different chain lengths on the surface plasmon band of gold nanocrystalline films prepared at the liquid-liquid interface has yielded several interesting results. The as-prepared films consist of gold nanocrystals close-packed in an assembly with nearly 1 nm spacing. The plasmon band of as-prepared films occurs around 700 nm because of electronic coupling between the nanocrystals. Upon ultrasonication, fragmented assemblies are formed as demonstrated by TEM, with the band becoming broad, extending from 550 to 700 nm, in contrast to a well-defined band at 540 nm from an organosol of thiol-capped nanocrystals. Electronic coupling between the nanocrystals in the assembly can be varied by interacting the film with alkanethiols of different chain lengths. While thiols with longer chains cause large blue-shifts (down to 560 nm in the case of octadecanethiol) indicative of diminishing coupling, smaller thiols show marginal shifts. For a given alkanethiol, the diminishing absorbance at 700 nm and the increasing in intensity at 530 nm results in an isosbestic point around 600 nm, which is an indication of a chemical equilibrium between the ordered nanocrystalline film and the disordered structure involving nanocrystals separated by alkanethiols. Alkanethiols with longer chains disintegrate or disorder the gold films more effectively and at lower concentrations as indicated from the spectral data as well as calorimetric data. The rate of interaction of the thiols as measured by the time-dependent band
2500 Langmuir, Vol. 24, No. 6, 2008
shifts also depends on the chain length, the longer chain length rendering the process faster. Apparent rate constants can be estimated considering the thiol adsorption on nanoparticle surface to be a modified first-order process. The activation energy of the process (40 kJ/mol for hexadecanethiol) is consistent with the adsorption energy. The spectral changes due to changes in the interparticle separation upon thiol interaction can be modeled
Agrawal et al.
using the Maxwell-Garnett formalism. The calculated values of absorbance agree well with the experimental trend for shorter chain lengths but are somewhat overestimated for longer chain lengths. The latter is attributed to the curling of the long chains adsorbed on nanocrystals packed in an assembly. LA703237M
