thin Organic Phosphonic Acid Layers on the Oxide of Aluminum

Article pubs.acs.org/Langmuir

In Situ Study of the Deposition of (Ultra)thin Organic Phosphonic Acid Layers on the Oxide of Aluminum Tom Hauffman,*,† Luk Van Lokeren,‡ Rudolph Willem,‡ Annick Hubin,† and Herman Terryn† †

Department of Electrochemical and Surface Engineering (SURF), Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium High Resolution NMR Centre (HNMR), Department of Materials and Chemistry (MACH), Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium

‡

S Supporting Information *

ABSTRACT: The interest in self-assembling monolayer deposition on various oxide substrate surfaces is steeply increasing in the last decades. Although many studies are being performed, literature does not come with a general insight in the adsorption of these layers on oxide surfaces. Also for the deposition of phosphonic acids on aluminum oxides, there is no global consensus. In this paper, we present an original in situ analysis in order to eludicate the real layer formation mechanism. First of all, the state of the phosphonic acid molecules was determined using DOSY NMR, making sure that no structures other than free molecules were present at the concentration used. With in situ atomic force microscopy and in situ visual ellipsometry, multilayers of phosphonic acids, showing 3D island growth, were determined. It was shown that using the variation of the in situ obtained roughness and bearing ratio, together with the equivalent thickness modeled by ellipsometry, the growth of the layers occurs in situ in three different stages. They consist of increasing number of islands growth, followed by filling up the gaps between islands. At last, within the adsorption time frame measured, the islands grow further in dimensions but not in numbers. This closely corresponds with the behavior of the octylphosphonic acid films analyzed by ex situ techniques.

■

INTRODUCTION The study of the adsorption of self-assembled organic monolayers (SAMs) on a wide variety of materials is a research field in full development because of the potential applications in various specialized domains: nanotechnology, biosensors, ideal conversion layers for corrosion protection, and so forth. Although in the last two decades a lot of research has been done, the adsorption of thiols on gold remains the best known system. Thiols bind through their sulfur group with gold atoms.1 Because of the ideality of the gold crystals, a very profound knowledge about this type of adsorption could be set up. It was observed that the thiolic layers do not immediately form a well-defined SAM, but that some residence time in the deposition environment is necessary to create the architectures. In the first moments of adsorption, it was observed with in situ scanning tunneling microscopy that a laying down phase is present.2 As more and more molecules adsorb on the surface, this phase transforms to a standing up one, creating the SAM. Furthermore, it has been observed with X-ray photoelectron spectroscopy (XPS) that such a behavior translates itself in the changes of the full width at half-maximum of the peaks coming from the adsorbed organic layer, due to changing chemical interactions.3 Although the thiol−gold system is very well-known and can be used in several applications (and more in particular in biosensors), it also exhibits some drawbacks. First of all, the usage © 2012 American Chemical Society

of gold implies a limited number of organic functionalities which can be used to form monolayers.4 Furthermore, gold is a noble, thus expensive, substrate and not suited for all industries. For these reasons, there has been a steep increase in interest in the adsorption of SAMs on engineering materials such as oxides on metals. Ultrathin film analysis of alkylsilanes, thiols, phosphonates on zirconium and titanium oxides, esters, hydroxymic acids, silanes, carboxylic acids on aluminum oxides, organophosphonates on NiTi oxide surfaces, and iron oxides can be found in literature.5−9 In this work, we will focus on the adsorption of phosphonic acids on aluminum oxides. Although some reports exist on the formation of SAMs of phosphonic acids on the oxide of iron,10 silicon,11,12 zirconium, and titanium,6 most of the work has been performed on aluminum oxide. Phosphonic acids bind with aluminum hydroxyl groups through an acid−base condensation elimination reaction.13 Maege et al. showed with FTIR that dodecyl- and octadecylphosphonic acids are able to form self-assembling monolayers from a number of solutions (ethanol, water, and THF) on technical aluminum plates, as long as the molecules were given several days to form the layer.14 This is very long compared to reported times of 24 h to form a Received: October 12, 2011 Revised: December 12, 2011 Published: January 13, 2012 3167

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

well-packed and dense layer.15,16 Thissen et al. showed with atomic force microscopy (AFM) and infrared spectroscopy that octadecylphosphonic acid can form a stable self-assembled monolayer on various aluminum oxide substrates, going from the native oxide to Physical Vapour Deposition (PVD) sputtered amorphous oxides over crystalline sapphire.15 This was confirmed for octadecylphosphonic and dodecylphosphonic acids from ethanol solutions using XPS and large area AFM scans.16 Although most authors claim that a chain length of at least 14 atoms17 is an important condition to form a SAM, others found SAMs with shorter molecules. Forget et al. showed with XPS and electrochemical techniques monolayers of pentanephosphonic acids on aluminum oxide substrates.18 Lewington et al. demonstrated the existence of disordered SAMs for decylphosphonic acids.19 Also for chains with 2, 6,20,21 and 821 carbon atoms, monolayers were detected with XPS and attenuated total reflection. Not only solvent involving deposition techniques can be used to create the SAMs structures. Monolayers of phenylphosphonic acids on amorphous alumina were formed by PVD.22 The XPS results were consistent with those for SAMs prepared through immersion of the aluminum oxide surfaces in acid solution. As outlined above, a lot of work has been performed with respect to the deposition of phosphonic acids on oxides. Although many complementary surface analysis techniques and different molecule−solvent systems have been used, no global image about the phosphonic adsorption can be retrieved from literature. Furthermore, in most cases, the blank substrate is poorly characterized. As hydroxyl groups play an important role in the anchoring of the phosphonic acid molecules, their presence on the surface should be controlled in terms of amount and distribution. Furthermore, the influence of the solvent is not recognized. Interactions between the solvent and the molecules used is possible, giving rise to structures in the solution that may affect the adsorption architectures. Finally, the adsorption of phosphonic acids on oxides is mostly studied ex situ. Ex situ analysis can have some major drawbacks with respect to in situ analysis. When taking a sample out of the solution, extra molecules can be withdrawn from the solution, giving rise to an extra Langmuir−Blodgett film.1,23 Moreover, rinsing and deposition of contamination may effect the analysis. In a previous study,24 we performed an ex situ study of the deposition of octylphosphonic acid from an aqueous solution on amorphous aluminum oxide. The oxide pretreatment was optimized. We combined direct information of the deposition structure by AFM with chemical information retrieved from XPS. The molecule’s choice was based on various articles stating that short lengthened molecules can form SAMs.16,18,25,26 In this paper, we present an in situ analysis in order to eludicate the real layer formation mechanism. Diffusion ordered NMR spectroscopy (DOSY NMR)27,28 is a powerful technique to investigate slowly diffusing species in solution, for instance, nanoclusters, functionalized nanoparticles, and micelles. Owing to the combination of high resolution in the spectral chemical shift domain with an additional resolution in the diffusion domain, DOSY NMR enables one to analyze complex mixtures. On the basis of the direct relationship between the diffusion coefficient and the molecular weight, DOSY NMR results in a virtual and molecular weight based separation of the mixture compounds in the diffusion domain and simultaneously provides detailed structural information in the spectral domain. Thus, as it is essential in the present investigation that the monolayers can grow from nonaggregated phosphonate ligands

in solution, we used DOSY NMR to find out below which critical aggregation concentration the phosphonate solutions should be used for this purpose. Subsequently, the in situ study was performed using two techniques. In situ AFM is employed to dynamically visualize the layer formation of the aluminum oxide surface. In order to collect in-depth information, visual ellipsometry is used. Here we can in a fast way (a few seconds per spectrum) obtain a data set which is sensitive to atomic layer changes. To our knowledge, this is the first report on the in situ analysis of the growth of ultrathin organic layers on pretreated oxide surfaces.

■

EXPERIMENTAL SECTION

The phosphonic acid employed was n-octylphosphonic acid from Alfa Aesar (purity 98%), used as received. N-Octylphosphonic acid (CH3(CH2)8POOH2) has a length of 1.3 nm and an adsorbed surface footprint of 0.27 nm2 (see the Supporting Information). The molecule was dissolved in Milli-Q water. The blank substrate was an ultrapure aluminum (99.99%) sheet received from Hydro Aluminum. The thickness of the sheets was 0.3 mm. The aluminum samples received the following pretreatment. First, the sample was immersed during 1 min in a 25 g/L aqueous solution of NaOH at 70 °C. Next, the sample was rinsed during 15 s in water, whereafter it was ultrasonically cleaned with water for 2 min. After drying, the degreased substrate was electropolished during 6 min, in an 80 vol % ethanol/20 vol % perchloric acid solution with a current density of 70 mA/cm2. The sample was then rinsed again for 15 s with water. After drying, the sample was galvanostatically anodized with a current density of 5 mA/cm2, with the voltage going up to 150 V. After rinsing it for 15 s with water, the sample was dried with compressed nitrogen. This is exactly the same pretreatment as used in the previous, ex situ, paper.24 All diffusion NMR measurements were performed at 300 K on a Bruker Avance DRX 250 NMR spectrometer equipped with a high gradient diffusion probe Diff30. The DOSY spectra were acquired with the ledbpgp2s pulse program from Bruker Topspin software adapted to include water presaturation. All spectra were recorded with 16 K time domain data points in t2 dimension and 32 t1 increments, with 16 transients each and a relaxation delay of 5 s. The experimental parameters for diffusion measurements were optimized according to a procedure described previously.29 All measurements were performed with a diffusion delay Δ of 50 or 200 ms and a gradient pulse length δ of 1000 or 500 μs. The gradient pulse strength G was varied in 32 linear steps from 2% of maximum gradient output (12 T/m) to a gradient level adjusted in order to ensure full signal attenuation. The spectra were processed as explained in ref 29. AFM measurements were carried out on a CP II instrument from VEECO. Commercially available cantilevers with a resonance frequency around 50 kHz were employed. The tips are made of silicon nitride, with the back side covered with a 15 nm Cr/60 nm Au layer. The opening angle of the tip is 30°, and the radius of curvature is less than 10 nm. AFM images consist of 256 pixels × 256 pixels. Due to the experimental conditions, it takes some time before the first AFM image can be acquired. Therefore, no data were obtained in the approximately first 5 min after putting the solvent with the molecules on the substrate. Data was treated using the Image Analysis 2.1 software. The in situ visual ellipsometric measurements were performed on a M200X, by J.A. Woollam Co. The ellipsometric data were analyzed using the CompleteEASE software version 4.41 developed by the J.A. Woollam Co.

■

RESULTS AND DISCUSSION DOSY−NMR Analysis. In this work, a series of octylphosphonic acid solutions in deuterated water (concentrations: 0.016, 0.052, 0.106, 0.128, and 0.153 M) were investigated by DOSY NMR in order to monitor possible aggregation phenomena such as micellization.30 The diffusion coefficients obtained

3168

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

Figure 1. Hydrodynamic radius and diffusion coefficient of octylphosphonic acid molecules at different concentrations in water.

found in the Supporting Information. It can be seen that the mean roughness varies between 0.82 and 1.18 nm, with a standard deviation of at maximum 0.26 nm. In a second stage, a blank substrate was submersed in an n-octylphosphonic acid containing solution. An AFM image

can be translated into a hydrodynamic particle size, that is, the size of a fictitious spherical particle diffusing identically, using the Stokes−Einstein equation:

D=

kBT c πηr h

(1)

with η being the viscosity of the solvent, rh the hydrodynamic radius of the particle, and c a parameter set to 6 for most nanoparticles. The diffusion coefficients of phosphonic acid obtained for the different solution concentrations, together with the different corresponding hydrodynamic radii, are displayed in Figure 1. It can be observed that, at low concentrations, rh is around 0.5 nm, characteristic for an individual phosphonic acid molecule free in solution, since its molecular length is 1.3 nm.24 Solutions with concentrations higher than 0.106 M exhibit supramolecular structures which are larger than the individual molecules, as evidenced by a lower diffusion coefficient and a corresponding larger hydrodynamic radius. Above this critical concentration, molecular aggregates, most probably micelles, are formed, and these can influence unfavorably the monolayer deposition. In this work, therefore, a concentration of 0.001 M, far below the critical aggregation concentration, will subsequently be used, thus ensuring that all molecules in solution are present as individual free molecules. In this way, only individual molecules will approach the surface and consequently determine the desired adsorption behavior. In Situ AFM Study. The adsorption of n-octylphosphonic acid on aluminum oxide was studied with in situ contact mode AFM. No effects due to tip−sample interactions could be observed. The authors are fully aware of the fact that lateral forces are exercised on the surface when scanning in contact mode. However, tapping mode AFM induced so much movement in the solvent that no representative AFM images could be obtained. As a function of time, AFM topographical images were obtained, measuring the same spot on the surface. First, in order to estimate the effect of the solvent on the oxide, the bare substrate is measured continuously for 2 h in pure water with contact mode AFM. The mean rms roughnesses for three samples during submersion together with the standard deviation can be

Figure 2. 3D plot of the adsorption of n-octylphosphonic acid on aluminum oxide using in situ contact mode AFM at 7 min of submersion.

taken at 7 min of immersion is shown in Figure 2. AFM images taken at other submersion times can be found in the Supporting Information. One can clearly see that the morphology changes from a roughened surface to a quite flat one. Based on the DOSY NMR data we can state that the structures are not due to organic molecule architectures in the solution but are formed as a result of the interaction of the molecules with the aluminum oxide surface. In order to quantify this behavior, the rms roughness of each image was calculated. The result, together with the roughnesses measured ex situ for the adsorption of the same molecule in a previous work,24 can be seen in Figure 3. One sees a fluctuating trend of the rms roughness as a function of adsorption time for 3169

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

Figure 4. Bearing ratio plot of the measured AFM surfaces as a function of adsorption time.

Figure 3. rms roughness plot as a function of submersion time of the adsorption of n-octylphosphonic acid on aluminum oxide using in situ contact mode (full curve) and ex situ (dashed curve) tapping mode AFM. Three roughness regions are shown. The dotted lines indicate the adsorption times of the 3D images shown.

However, AFM is only a relative technique. There is no insight provided in what is going on underneath the surface measured. Therefore, in situ visual ellipsometry was employed to explore the total growth character of the organic layer. Ex Situ Visual Ellipsometry Analysis of the Blank Substrate. The visual ellipsometry spectrum of a blank aluminum oxide substrate in air is measured at different angles: 70°, 75°, and 80°. Both phi and delta parameters are measured.31 The spectrum for a blank substrate can be seen in Figure 5. Visual ellipsometry data should be modeled to retrieve quantitative data. Therefore, an optical model is proposed which represents the system under investigation. The aluminum oxide was modeled as being an oxide of a certain thickness on top of an infinite thick metallic aluminum substrate. In between both layers, it was shown that there is an intermix layer.32 This intermix consists of an effective medium approximation layer (EMA) of aluminum oxide and aluminum (see also the Supporting Information). As most of the literature values for the optical constants of aluminum oxide are measured for the ideal crystalline sapphire form of the oxide, the optical constants of the samples used need to be fitted. The oxide is a nonadsorptive, transparent layer. Therefore, the optical constants can be modeled using a Cauchy dispersion relation:

the roughnesses calculated from the in situ AFM analysis. The in situ rms roughness can be divided in three parts. Part (A) represents an increasing trend, coming from valleys and hills visible on the AFM images. Part (B) shows a decreasing rms roughness as a function of adsorption time, and this is seen in the 3D figures which become flatter. In division (C), the rms roughness increases again, due to the upcoming islands in the AFM images. The variations in roughness throughout deposition time are at a maximum of 4 nm. This is considerably higher than the standard deviation observed for the deposition of the bare substrate in water, being 0.26 nm. It can be concluded that the variations observed in the n-octylphosphonic acid solution are due to the adsorption of organic structures on the surface. When comparing the ex situ measured roughnesses to the in situ ones, it can be seen that both show the same trend. However, the variations appear in a much shorter time scale in situ than ex situ: a decrease is ex situ observed after 120 min of adsorption, in situ this is after 50 min. In order to highlight better the changes in the roughness, the bearing ratio at 20% was calculated. The bearing ratio gives the height (relative to the lowest point in the AFM image) above which 20% of the surface lays. It is visualized in Figure 4. Here, also three regions (A), (B), and (C) can be observed, which correspond closely with the time domain of the rms roughness. In domain (A), the bearing ratio is constant and the rms roughness increases which points out that here the growth mechanism is mainly due to an increase in number of islands. It should be noted that, although domain A is mainly marked by a growth in the number of islands, there is a time interval between 40 and 60 min of submersion where an increase in bearing ratio is observed. Although it is small compared to the increase of the bearing ratio in domain C, here the growth mechanism exhibits also a small growth of the existing islands. In domain B, a decrease in roughness is observed. It can also be seen that here the bearing ratio value is lower than in zone A. Due to the filling of the volumes between the islands generated in zone A, the lowest point measured in AFM rises toward the highest point and thus the height differences become smaller. In domain (C), both the bearing ratio and the rms roughness increase, coming from the fact that the islands grow heigher.

B n(λ) = A + 2 λ

(2)

Here, n represents the refractive index as a function of the wavelength λ, and A and B are fit parameters. To check whether the optical constants are reproducible, a series of 40 blank aluminum oxide samples is measured and both the optical constants and thicknesses of all fits are averaged. One obtains an oxide thickness of 183 ± 29 nm, an EMA thickness of 1.3 ± 0.99 nm, an A coefficient of 1.633 ± 0.016, a B coefficient of 0.00763 ± 0.000315, and a percentage of aluminum oxide in the EMA layer of 36% ± 11%. The optical constants for the blank substrate are in good agreement for the ones mentioned in literature for an electropolished and barrier anodized sample in diammoniumtartrate: A = 1.65 and B = 0.007.32 Also the percent of aluminum oxide does agree with the literature value of 34%. In Situ Visual Ellipsometry Analysis. The blank substrate analyzed in the previous chapter was mounted in a homemade 3170

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

Figure 5. VISSE phi and delta data for the bare aluminum oxide substrate for different measurement angles.

liquid analysis cell, with an operation mode which is restricted to an angle of 70°. First the cell was filled with water. A spectrum between 300 and 900 nm wavelength was taken every 2.58 s. Every spectrum is separately fitted to obtain an insight in the dynamic adsorption behavior. At first, the new “pseudo” optical constants were determined to include solvent and cell effects. The influence of water itself on the aluminum oxide was determined. It can be seen for a series of a spectra that visually no difference in the spectra can be observed (Supporting Information). The sample was not removed from the cell, and a concentrated n-octylphosphonic acid solution was injected. This way a homogeneous solution was reached. The concentrated solution was made with a concentration well below 100 mM, so that no architectures are present in the solution. In order to find an appropriate model for the adsorption of the phosphonic acid layer, we started from the blank oxide model. The refractive index of the surroundings (being the water/n-octylphosphonic acid solution) is measured with a refractometer to be 1.333. When the adsorption was fitted with this model, it was shown that the value of the Cauchy optical constants of the oxide decreased and the oxide thickness increased. This means that, due to the injection of the phosphonic acid solution, the oxide grows with changing optical constants. This in contrast with the fact that oxide growth requires an external field. So, the growth should come from an adsorbing organic layer, which is also in agreement with the in situ AFM analysis. The optical constants of the adsorbed layer were obtained from literature. For the adsorption of n-octadecylphosphonic acid on silicon, a refractive index n of 1.61 is measured in the visible light spectrum.33 Although the refractive index of organic constituents depends on the molecular weight and architecture, it was shown that, for small variations of the chain length (or branching), the variation observed is less than 0.01.34 As this refractive index is lower than the one for the aluminum oxide, this explains why in the previous model (blank substrate only) the refractive index of the oxide drops. Therefore, the refractive index taken here is 1.61. The model used can be seen in Figure 6. The quality of the fit is represented by a mean-squared error of less than 20, which is generally

Figure 6. Optical model used to fit adsorption of the n-octylphosphonic acid molecules on the blank substrate from water.

considered for ellipsometry modeling to be a good fit31 (see the Supporting Information). At 0 min of submersion time, the thickness is 0 nm. After injection, the thickness rises very steeply to the value of around 1.25 nm. The standard deviation of the organic layer thicknesses fitted is situated between 0.109 and 0.112 nm. First of all, a continuous layer growth is observed. The thickness reached in the first 2 h (e.g., time measured with in situ AFM) is 1.9 nm. The thickness of the organic layer increases faster in the beginning of the submersion. This is in agreement with most literature regarding SAMs and ultrathin films.2,35 At first, on the blank substrate, molecules can rapidly adsorb on the surface. However, as more and more molecules adsorb, the growth of the layer is made more difficult because of the decrease of adsorption space. This implies that lateral diffusion and reorientation of the molecules is necessary which takes time. This growth is also visualized in the fitting result. A steep increase in thickness is observed in the first 10 min of submersion. Hereafter, a continuous growth is still observed, 3171

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

in situ. At first, the solution−organic molecule interactions were analyzed. It was shown that DOSY NMR enables one to find accurately in which concentration range the molecules diffuse as free monomeric molecules. No supramolecular structures were created prior to deposition. It was observed with in situ AFM and ellipsometry that multilayers exist on the surface, with a continuous increase of deposited molecules as a function of submersion time. The layers are not homogeneous in nature, but grow according to a Stranski−Krastanov mechanism. Three growth zones were shown within the adsorption time of 160 min: island growth, island coalescence, and new island growth. The structures do not exist of compact dense molecular structures. This behavior was also observed for the same system, analyzed ex situ.

but with a slower rate. It can also be seen that, after 1000 min, there is still an increase in organic layer thickness. Discussion in Situ AFM versus in Situ Ellipsometry. When comparing the visual ellipsometry data with the AFM data, it can be noticed that ellipsometry finds a thickness of 1.9 nm within 2 h of submersion. On the other hand, AFM shows that, even at early adsorption times (7 min), the bearing ratio of 20% lays above 16 nm. This means that at least 20% of the surface is 16 nm above the lowest point measured. This is not shown by the ellipsometry fitting. The reason for this difference lays within the model used for the ellipsometry data. Here it is assumed that the organic layer is a compact layer; hence, the ellipsometric model represents an equivalent compact layer. The ellipsometry data give the information that there are enough molecules present at 2 h of adsorption to create a densely well-packed organic layer. The AFM results show however that most of the surface is covered by higher islands. This is only possible if the structures observed are not well-packed and thus show some voids. In the AFM images, three domains are visible in the rms roughness plot versus deposition time (Figure 3) and the 3D AFM plots. In time zone (A), the islands are growing. In time zone (B), the islands coalesce, where after they start to grow again in zone (C). In this paper, it is additionally shown that the islands do not consist of a compact organic layer sequence. The combination of increasing number of molecules, measured by VISSE, together with the varying character of the islands on the surface, leads to the conclusion that the layer exhibits a 3D growth. Whether this growth is a Stranski−Krastanov one (Figure 7), as previously shown for the ex situ analysis of the

■

ASSOCIATED CONTENT

S Supporting Information *

(1) N-Octylphosphonic acid. (2) Mean rms roughness and standard deviation of three different blank samples during submersion in Milli-Q water. (3) 3D plot of the adsorption of n-octylphopshonic acid on aluminum oxide using in situ contact mode AFM at (b) 25, (c) 43, (d) 61, (e) 79, (f) 103, (g) 121, and (h) 139 min of submersion. (4) Optical model used to fit the blank substrate in air. (5) Overlay of 10 VISSE spectra recorded for the blank substrate submersed in water. (6) Visual ellipsometry data measured at 70° of the bare aluminum substrate and the aluminum substrate with phosphonic acid layer after 1000 min of continuous submersion. (7) Result of the ellipsometry fit for the thickness of the organic layer as a function of continuous adsorption time. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

■

ACKNOWLEDGMENTS Research funded by a Ph.D. grant of the Fonds Wetenschappelijk Onderzoek in Flanders (FWO).

■

REFERENCES

(1) Schreiber, F. Prog. Surf. Sci. 2000, 65, 151−256. (2) Poirier, G.; Pylant, E. Science 1996, 272, 1145−1148. (3) Zharnikov, M. J. Electron Spectrosc. Relat. Phenom. 2010, 178, 380−393. (4) Blajiev, O.; Ithurbide, A.; Hubin, A.; Van Haesendonck, C.; Terryn, H. Prog. Org. Coat. 2008, 63, 272−281. (5) van den Brand, J.; Blajiev, O.; Beentjes, P. C. J.; Terryn, H.; de Wit, J. H. W. Langmuir 2004, 20, 6308−6317. (6) Gao, W.; Dickinson, L.; Grozinger, C.; Morin, G.; Reven, L. Langmuir 1996, 12, 64296−6435. (7) Franquet, A.; Terryn, H.; Vereecken, J. Surf. Interface Anal. 2004, 36, 681−684. (8) Nuzzo, R. Langmuir 1995, 11, 813−824. (9) Alexander, M. R.; Payan, S.; Duc, T. M. Surf. Interface Anal. 1998, 26, 961−973. (10) Felhösi, I.; Kalman, E. Corros. Sci. 2005, 47, 695−708. (11) Gouzman, I.; Dubey, M.; Carolus, M.; Schwartz, J.; Bernasek, S. Surf. Sci. 2006, 600, 773−781. (12) Eschner, M.; Frenzel, R.; Simon, F.; Pleul, D.; Uhlmann, P.; Adelr, H.-J. Macromol. Symp. 2004, 210, 77−84. (13) Wapner, K.; Stratmann, M.; Grundmeier, G. Int. J. Adhes. Adhes. 2007, 28, 59−70.

Figure 7. Stranski−Krastanov growth mechanism where the different AFM regimes noticed in the in situ AFM analysis are indicated.

adsorption of n-octylphosphonic acid from an aqueous environment on aluminum oxides,24 cannot be proven with the techniques used in this work. Nor can contact AFM or in situ VISSE show here that a monolayer is present on the oxide before 3D growth occurs.

■

CONCLUSIONS In this work, the adsorption of n-octylphosphonic acid ultrathin films on aluminum oxides from aqueous solutions was studied 3172

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173

Langmuir

Article

(14) Maege, I.; Jaehne, E.; Henke, A.; Adler, H.; Bram, C.; Jung, C.; Stratmann, M. Prog. Org. Coat. 1998, 34, 1−12. (15) Thissen, P.; Valtiner, M.; Grundmeier, G. Langmuir 2009, 26, 156−164. (16) Hoque, E.; DeRose, J.; Hoffmann, P.; Mathieu, H. J.; Bhushan, B.; Cichomski, M. J. Chem. Phys. 2006, 124, 174710. (17) Koutsioubas, A.; Spiliopoulos, N.; Anastassopoulos, D.; Vradis, A.; Priftis, G. Surf. Interface Anal. 2009, 41, 897−903. (18) Forget, L.; Wilwers, F.; Delhalle, J.; Mekhalif, Z. Appl. Surf. Sci. 2003, 205, 44−55. (19) Lewington, T.; Alexander, M.; Thomspon, G.; McAlpine, E. Surf. Eng. 2002, 18, 228−232. (20) Adolphi, B.; Jähne, E.; Busch, G.; Cau, X. Anal. Bioanal. Chem. 2004, 3179, 646−652. (21) Van Alsten, J. Langmuir 1999, 15, 7605−7614. (22) Tsud, N.; Yoshitake, M. Surf. Sci. 2007, 601, 3060−3066. (23) Ulman, A. An introduction to ultrathin organic films: from Langmuir-Blodgett to self-assembly; Academic Press: Boston, 1991. (24) Hauffman, T.; Blajiev, O.; Snauwaert, J.; van Haesendonck, C.; Hubin, A.; Terryn, H. Langmuir 2008, 24, 13450−13456. (25) Hoque, E.; DeRose, J.; Kulik, G.; Hoffman, P.; Mathieu, H.; Bhusban, B. J. Phys. Chem. B 2006, 110, 10855−10861. (26) Luschtinetz, R.; Oliveira, A.; Frenzel, J.; Joswig, J.-O.; Seifert, G.; Duarte, H. Surf. Sci. 2008, 602, 1347−1359. (27) Johnson, C. J. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203− 256. (28) Morris, G. In Encyclopedia of Magnetic Resonance; Harris, R., Wasylishen, R., Eds.; John Wiley: Hoboken, New Jersey, 2003. (29) Denkova, P.; Van Lokeren, L.; Verbruggen, I.; R., W. J. Phys. Chem. B 2008, 112, 10935−10941. (30) Denkova, P.; Van Lokeren, L.; Willem, R. J. Phys. Chem. B 2009, 113, 6703−6709. (31) Co., J. W. Guide to using WVASE32 Spectoscopic ellipsometry data acquisition and analysis software; J.A. Woollam Co., Inc.: Lincoln, Nebraska, 2008. (32) Van Gils, S.; Melendres, C.; Terryn, H.; Stijns, E. Thin Solid Films 2004, 455−456, 742−746. (33) Nie, H.-Y.; Walzak, M.; McIntyre, N. Langmuir 2002, 18, 2955− 2958. (34) Candau, F.; Francois, J.; Benoit, H. Polymer 1974, 15, 626−630. (35) Barrena, E.; Ocal, C.; Salmeron, M. J. Chem. Phys. 1999, 111, 9797−9802.

3173

dx.doi.org/10.1021/la203988m | Langmuir 2012, 28, 3167−3173