Microscopic Structure of Crystalline Langmuir

due to the OH bridging groups.3 ... X-ray reflectivity and GID were measured at the Troika II ..... continuous line is the reflectivity calculated fro...
1 downloads 0 Views 565KB Size
Download PDF
Langmuir 2005, 21, 11213-11219

11213

Microscopic Structure of Crystalline Langmuir Monolayers of Hydroxystearic Acids by X-ray Reflectivity and GID: OH Group Position and Dimensionality Effect Luigi Cristofolini,*,†,‡ Marco P. Fontana,†,‡ Carla Boga,§ and Oleg Konovalov| INFM and Physics Department, University of Parma, Parco Area delle Scienze 7/A, I-43100 Parma, Italy, Centro Ricerca e Sviluppo SOFT, Physics Department, University of Rome I, Department of Organic Chemistry “A. Mangini”, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy, and European Synchrotron Radiation Facility, 6 rue Jules Horowitz, F-38043 Grenoble, France Received May 31, 2005. In Final Form: September 5, 2005 Hydroxystearic acid (HSA) molecules at the air-water interface present an interesting bicompetitive adsorption between primary and secondary hydrophilic groups on either end of an alkyl chain, which, depending on the position of the second hydrophilic group, may lead to a sharp transition from an expanded phase to a crystalline condensed morphology as surface pressure is increased. Here we report a set of measurements on a series of hydroxystearic acids in which the position of the secondary competing hydrophilic group position is varied along the whole extent of the alkyl chain from position 2 (i.e., close to the primary hydrophilic group) to positions 7, 9 and 12, the latter being the compounds mostly studied in the literature. We show here direct microscopic evidence, obtained by synchrotron radiation reflectometry and grazing incidence diffraction, that the position of the secondary hydrophilic group not only strongly influences the phase diagram as determined by compression isotherms and ellipsometry but also induces different crystallization patterns in the 2D system of the Langmuir monolayer. In particular, we report for the first time the existence of a turning point in the effects of the hydroxyl position on the monolayers structure at 7-HSA.

Introduction Insoluble monolayers of fatty acids containing two hydrophilic moieties have been studied by several groups in recent years.1-4 Such molecules present interesting complex behavior at the air-water interface because of the competition between the adsorption of the primary and secondary hydrophilic groups;2 furthermore, the presence of an additional hydroxyl group might lead to hydrogen bonding between adjacent molecules, thus producing different packing arrangements and interactions at the air-water interface. In the studies present in the literature,1,2,4 the isotherms of the hydroxystearic acids (HSA) show a plateau region that is commonly understood to be due to the coexistence of two distinct phases: at large areas per molecule, with the surface pressure below the plateau value, the molecules lie approximately flat with both of the hydrophilic groups on the water surface, whereas at a smaller area per molecule, with the surface pressure higher than the plateau, the molecules assume a more or less vertical conformation,2 although precise structural determination is available only for hydroxystearic acids with the second hydrophilic group well separated from the polar head.4 Indirect structural determination by in-situ polarized FT-IR external-reflection spectra of 12-hydroxystearic acid * Corresponding author. E-mail: [email protected]. † University of Parma. ‡ University of Rome I. § University of Bologna. | European Synchrotron Radiation Facility. (1) Sakai, H.; Umemura, J. Langmuir 1998, 14, 6249-6255. (2) Yim, K. S.; Rahaii, B.; Fuller, G. G. Langmuir 2002, 18, 65976601. (3) Meurk, A. Tribol. Lett. 2000, 8, 161-169. (4) Vollhardt, D.; Siegel, S.; Cadenhead, D. A. Langmuir 2004, 20, 7670-7677.

Figure 1. Langmuir isotherms for the different hydroxystearic acids (indicated in the legend) used in the present work. Conditions: 30 µL of a 0.33 mg/mL chloroform solution, T ) 21°C. (Top inset) Time evolution of film thickness, probed by ellipsometry during the compression along the Π-A plateau, for 9-HSA. (Bottom inset) Structure of 2-HSA. Positions of the second hydroxyl group in 7-HSA, 9-HSA, and 12-HSA are indicated by the arrows..

(12-HSA) suggests that the molecular orientation angles vary at different points on the isotherm: at the final stage of the plateau region of the compression isotherm (Figure 1), the orientation angle of the hydrocarbon chain was found to lie about 55° from the surface normal; however, upon monolayer compression, the angle decreased to about 28° in the solid-phase region.1 This finding suggests that some phase transition might take place in the solid phase. Additional interest is due to the fact that hydroxystearic acids are widespread in nature and in industrial products.

10.1021/la0514213 CCC: $30.25 © 2005 American Chemical Society Published on Web 10/13/2005

11214

Langmuir, Vol. 21, No. 24, 2005

It is apparent that in most cases hydroxystearic acids act on the rheological properties and the structure of the systems with which they interact and that a determining role is played by the OH group on the alkyl tail and the ensuing degrees of motional and structural freedom of the molecules. It is therefore of some interest to study, in a simple system, the effect of OH position along the alkyl tail on the structure and morphology of HSA assemblies, such as monolayers at the air-water interface. In addition, nanotribology measurements by atomic force microscopy on glass surfaces covered by 2-HSA and 12-HSA pointed out the importance of the position of the OH group in determining the rheological properties of hydroxystearic acid confined to the surfaces. Two different mechanical regimes are found, depending on the position of the OH group in the alkyl chain: whereas 2-HSA exhibits a static-dynamic transition (steady sliding replaced by a stick-slip regime as the velocity is decreased), 12-HSA has no such transition. This is attributed to the different configurations adopted by the molecules due to the OH bridging groups.3 In this work, we wish to address the specific point of the relative roles of hydrophobic and hydrophilic equilibriums on the molecular scale in determining the aggregation and morphology of molecules at the air-water interface in an effort to understand the generalized mechanical effects that seem to be associated with the presence of n-HSA. The functionalized n-HSA molecules studied here seem particularly interesting from this point of view because the molecules can be engineered so as to position the hydroxyl group any distance from the interface, thus varying practically continuously the relative hydrophilicity of the normally hydrophobic alkyl chain. In this article, we shall concentrate on the roomtemperature behavior, addressing the issue of whether the position of the OH group along the chain, leading to the different isotherms observed, also influences the crystal morphology and the crystallographic structure of the monolayer. We do this by X-ray reflectometry and grazing incidence diffraction (GID), using synchrotron radiation, and by ellipsometric imaging. From these measurements, we obtain information on the formation, structure, and molecular orientation of the microscopic 2D crystallites that form in the monolayer at sufficiently high pressures and for specific positions of the OH group along the alkyl chain. This information is then related to data from the compression isotherms and from nullellipsometric measurements. An important result we have obtained is the identification of a critical role of 7-HSA that seems to separate two different regimes on the basis of the effects of the OH group on the structure and morphology of the monolayer. The overall information we obtain allows us to propose a comprehensive interpretation of the role of competing hydrophilic interactions on the structure and morphology of the Langmuir monolayer of this family of fatty acids at the air-water interface. Experimental Section 2-Hydroxystearic acid (2-HSA) and 12-hydroxystearic acid (12HSA) were purchased from Sigma-Aldrich. 7-Hydroxystearic acid (7-HSA) and 9-hydroxystearic acid (9-HSA) were prepared according to the literature,5 with modifications. The corresponding structures are sketched in the bottom inset of Figure 1. For the ellipsometric measurements, we used a small trough made by us (84 × 305 mm2 surface), which was placed directly in the laser light path and was positioned through micrometric controls. The ellipsometer was aligned to operate in the vertical (5) Bergstrom, S.; Aulin-Erdtman, G.; Rolander, B.; Stenhagen, E.; Osting, S. Acta Chem. Scand. 1952, 6, 1157.

Cristofolini et al. plane, at an incidence angle of Φ ) 48° from the normal to the water surface, on an antivibration table with a He-Ne laser source (λ ) 632.8 nm). The same computer was used to drive the electronics controlling the Langmuir trough and to perform the ellipsometric measurement. Null-ellipsometric imaging measurements have been performed using the same setup while the light was collected with an M-PLAN APO 20× (Mitutoyo) objective and a CCD camera detector; images were captured with a frame grabber. In all of the experiments, the water surface was routinely characterized prior to film dispersion to ensure purity, by a complete compression cycle while measuring the surface pressure, to detect the presence of any contaminant. For the ellipsometric measurement, angles ∆ and Ψ were measured on the pure water surface (i.e., before film dispersion) to provide the final alignment and quantification of capillary waves, thus providing an accurate reference point. A standard inversion scheme, based on Fresnel formulas, was applied to convert ellipsometrically measured angles ∆ and Ψ to film thickness. For this, the refractive index is needed and was directly measured for all of the hydroxystearic acids from ellipsometric measurements on thick multilayers, obtaining the value n ) 1.51(1). We note that this value is sensibly larger than that tabulated6 for stearic acid (n ) 1.4299) probably because of the different, more dense packing of the HSA molecules with respect to the normal stearic acid, which is possibly related to the existence of extra hydrogen bonding between the additional OH groups. X-ray reflectivity and GID were measured at the Troika II (ID10B) beamline of ESRF operating at wavelength λ ) 1.542 Å. A Langmuir trough made by the ESRF workshops, with a single moving barrier (maximum surface 418 × 170 mm2) was mounted on an active antivibration support and was surrounded by a helium atmosphere to reduce diffuse scattering from air. Before all of the measurements on Langmuir films, the X-ray reflectivity from the bare water surface was measured to check the alignment and to provide a reference. The reflectivity data were subsequently analyzed both with our own software and with the program PARRATT,7 which calculates theoretical reflectivity curves according to the so-called Parratt recursive approach.8,9 In this method, one calculates the transmission and reflection Fresnel coefficients at each interface, starting from the approximation of independent smooth layers. To reproduce the reflectivity curve of a real multilayer, a structured model of independent layers, each with uniform electron density, has to be provided. The roughness of real layers is accounted for using the Ne´vot-Croce approximation,10 which implies a scaling of the reflectivity at the interface between layers a and b by the factor exp(-2kakbσ2), with σ being the roughness rms value and ka kb representing the z component of the wave vector for layers a and b, respectively. The incident beam had vertical size of 0.1 mm, and for the GID measurements, we selected an angle of incidence Ri ) 0.12° (i.e., 75% of the critical angle for the air-water interface at this wavelength), resulting in a 5 cm beam footprint. The scattered beam was spatially filtered by means of Soller slits, which give an angular resolution of 0.08° in the horizontal plane yielding the intrinsic width in the momentum space of fwhm ) 0.0056 Å-1, almost constant in our measurement range. The intensity of the scattered X-ray beam was subsequently recorded by means of a linear position-sensitive detector. The incident beam was attenuated to keep the total count below 5000 counts per second, thus avoiding pileup and detector saturation. Because we are not interested in absolute intensities but only in position and shapes, data have not been corrected for polarization effects. In the analysis of GID data, the isotropic 2D powder average can be safely assumed because the typical crystallite sizes, as observed also by ellipsometric imaging, are much smaller than the impinging beam footprint, that is, 5 cm long. (6) CRC Handbook of Chemistry and Physics, 79th ed.; Lide, D. R., Ed.; CRC Press: New York, 1998. (7) Braun, C. Parrat32 Software for Reflectivity; HMI: Berlin, 1999. (8) Parratt, L. G. Phys. Rev. 1954, 95, 359-369. (9) Daillant, J.; Gibaud, A. X-ray and Neutron Reflectivity: Principles and Applications; Springer: Berlin, 1999. (10) Nevot, L.; Croce, P. Rev. Phys. Appl. 1980, 15, 761-779.

OH Group Position and Dimensionality Effect

Figure 2. Ellipsometric imaging pictures of the different crystalline morphologies, all recorded at the end of the plateau in the isotherm, for the different HSAs. The length scale (same for all of the pictures) is indicated in the bottom right panel.

Results and Discussion a. Langmuir Isotherms, Ellipsometry, and Imaging. In Figure 1, we report the isotherms for the different HSAs used in the present work, all recorded at T ) 21 °C with the same aliquot of 0.33 mg/mL chloroform solution spread on the surface of the Langmuir trough. We clearly note that different positions of the OH group influence the extent and the pressure of the plateau in the isotherm. The isotherms of 9- and 12-HSA are consistent with those present in the literature, and 7-HSA also shows similar behavior. On the contrary, 2-HSA shows very different behavior, characterizing a remarkably robust film, with a steep onset of surface pressure, no sign of collapse up to 66 mN/m, and a final collapse pressure very close to (but below) the bare water surface tension. We double checked this result both by varying the dispersed aliquot and by checking the Wilhelmy balance calibration. Null-ellipsometry measurements performed on all of the Langmuir monolayers during compression invariably

Langmuir, Vol. 21, No. 24, 2005 11215

show an increase in film thickness as pressure is increased, as indicated by the continuous increase of the angle ∆, which is directly related to film thickness via the refractive index. The values for the final thickness of the compressed monolayers agree with those more precisely determined by X-ray reflectivity (vide infra). However, as shown in the inset of Figure 1, the time evolution of film thickness along the plateau shows discrete jumps between “high” and “low” film thickness. This indicates that the film is laterally inhomogeneous (i.e., it consists of dense patches separated by areas of lower density). This is also confirmed by the ellipsometric imaging data of Figure 2. Moreover, it is interesting that when the plateau coexistence range is well defined its temperature dependence in the isotherms follows the 3D behavior of the van der Waals fluid very closely (data not shown). This particular aspect deserves further investigation. The first two pictures of Figure 2 confirm the morphology previously observed,4 with the characteristic acicular morphology, whereas the other pictures reveal a new, unexpected morphology featuring round islands for 7-HSA and large uniform patches for 2-HSA. b. X-ray Reflectivity. As expected, very little structure is seen in the low-pressure reflectivity curves of the HSA (data not shown). On the contrary, at high pressure (Π ) 25 mN/m) 2-HSA shows a well-defined pattern that could be fitted (quality of the fit χ2 ) 1.7) by the model depicted in Figure 3, in which, coming from the water side (electron density ) 0.33 e- Å-3) we have a somewhat extended polar head region of thickness 9.36 Å (electron density ) 0.42 e- Å-3), followed by a shorter alkyl chain region that is only 15.6 Å in thickness (electron density ) 0.31 e- Å-3). We note that it is not possible to distinguish, in the electron density profile, a separate region with the side OH group. The overall film thickness (25.0(1) Å) is in good agreement with the results reported in the literature11-13 for Langmuir films of Cd or Ba-stearate by X-ray reflectivity (25.15(5) Å and 25.3(4) Å, respectively). We note, however, that in 2-HSA the polar head region is more extended; conversely, the alkyl chain region is shorter, most probably because of the presence of the additional OH group in the 2 position.

Figure 3. X-ray reflectivity pattern for 2-HSA at Π ) 25 mN/m. The error bars are the experimental data, and the continuous line between the points is the reflectivity calculated from the model illustrated in the inset.

11216

Langmuir, Vol. 21, No. 24, 2005

Figure 4. X-ray reflectivity pattern for 12-HSA acid at Π ) 15 mN/m. The error bars are the experimental data, and the continuous line is the reflectivity calculated from the threeslab model illustrated in the inset.

If we refer to the commonly accepted rule14,15 predicting that in the crystal structures of most saturated hydrocarbons the CH2-CH2 spacing projected onto the chain axis is d ) 1.265 (0.010)Å and that the final CH3 group counts for 9/8 of such value, we are led to the conclusion that in Langmuir monolayers of 2-HSA 6 carbon atoms are submerged and only 12 carbon atoms are left out of the water surface, which is reasonable given the very strong hydrophilic nature of the polar head with 2 OH groups. On the contrary, the reflectivity curve from 12-HSA at Π ) 15 mN/m shows a less-structured pattern, which in principle could be fitted with a simple two-slab model of overall thickness of 25.16 Å. However, the quality of this fit is not satisfying, particularly in the higher exchanged momentum region. This prompted the formulation of a more complex model containing three layers. The best fit of the experimental data was thus obtained (χ2 ) 1.29) with the model depicted in Figure 4, which implies a total thickness of 23.35 Å. This value, compared with that of 2-HSA previously discussed (25.0 Å), implies a tilting angle from the water surface normal of θ ) 20.9°, which is in excellent agreement with the GID determination as discussed in the following section. Looking at our model in more detail, we find first a region of polar heads extending for 3.88 Å. This value is in good agreement with what is commonly found for other fatty acid polar headgroups (e.g., in arachidic acid15 at Π ) 15.9 mN/m, t ) 3.88 Å). Next, we find a slab of alkyl chains extending for 8.85 Å and characterized by an electron density comparable to that of pure water and finally a region of anomalously high electron density (thickness 10.62 Å) that could be related to the presence of the OH group in the 12 position. Again, if we take the commonly accepted rule14,15 for the extension of the alkyl (11) Langmuir Blodgett Films; Roberts, G., Ed.; Plenum Press: New York, 1990; p 139. (12) Srivastava, V. K.; Ram Verma, A. Solid State Comm. 1966, 4, 367-371. (13) Matsuda, A.; Sugi, M.; Fukuj, T.; Izima, S.; Miyahara, M.; Otsubo, Y. J. Appl. Phys. 1977, 48, 771-774. (14) Tanford, C. The Hydrophobic Effect: Formation of Micelles and Biological Membranes; Wiley: New York, 1973. (15) Kjaer, K.; Als-Nielsen, J.; Helm, C. A.; Tippman-Krayer, P.; Mohwald, H. J. Phys. Chem. 1989, 93, 3200-3206.

Cristofolini et al.

Figure 5. (Main panel) Grazing incidence diffraction intensity map (counts are represented as gray intensity) from a Langmuir layer of 12-HSA at Π ) 7 mN/m as measured on the ID10B beamline. The letters A, B, and C mark the three diffraction peaks discussed in the text. The graph in the top panel is the GID curve (i.e., the projection of the 2D map on the Qxy axis). A small artifact can be seen at Qxy ) 1.466Å-1, which appears as extra intensity at a single particular value of Qxy. The graph on the right panel is the Qz dependence (rod) of the three peaks.

chain as a function of the number of CH2 units, we are led to the conclusion that only seven CH2 units are left unaffected by the OH group in the 12 position. We also tried to improve the model by the addition of a fourth layer, which could in principle account for the remaining part of the alkyl chain above the second OH group. However, the quality of the fit did not improve significantly upon this addition, which is related to the fact that there is little structure present in our reflectivity data, so we discarded this model. Comparing these data with those obtained for 2-HSA, we note that the submerged part contains only the polar head: thus for 12-HSA essentially all of the carbons are above the water surface. Furthermore, the electron density profile is qualitatively different in 12-HSA and in 2-HSA; namely, in 12-HSA the maximum electron density is found in the topmost layer. There is a minimum in the electron density apparently at the position of the second OH group and an anomalous increase in electron density for the remaining CH2 units. Comparing the profiles for 2-HSA and 12-HSA, respectively, we note that the average electron density is higher for 12-HSA (∼0.4 vs ∼0.35 e- Å-3). Considering that the first seven carbons in 12-HSA have essentially the same electron density as water, we can infer from this result that there is a kink in the alkyl chain at the 12 position, which is consistent with the average tilt angle of θ ) 20.9° that we have measured and which would yield a more compact alkyl chain structure for n > 12, and hence a higher electron density. c. Grazing Incidence Diffraction. 12-Hydroxystearic Acid. In Figure 5, we show a typical 2D GID pattern obtained for 12-HSA at Π ) 7 mN/m. We note that for all of the HSAs measured diffraction peaks from 2D ordered structures were found only when the surface pressure was higher than the plateau pressure. In particular, the minimum surface pressure required to observe diffraction peaks was 16 mN/m for 7-HSA, 12 mN/m for 9-HSA, and 7 mN/m for 12-HSA. This is in contrast to the situation encountered in normal fatty acids, where GID peaks are measured virtually at any pressure (e.g., in arachidic acid albeit with reduced intensity when the area per molecule is very large15). As a matter of fact, fatty acids show molecular self-aggrega-

OH Group Position and Dimensionality Effect

Langmuir, Vol. 21, No. 24, 2005 11217

Table 1. In-Plane and Out-of-Plane Exchanged Momenta and Corresponding Widths of the Three GID Peaks for 12-HSA at Increasing Surface Pressure, Π ) 7, 10, and 15 mN/m, as Indicated in the First Column Qxy (Å-1)

fwhm Qxy (Å-1)

Qz (Å-1)

7 mN/m A 7 mN/m B 7 mN/m C

1.363(1) 1.476(2) 1.574(1)

0.006(1) 0.014(4) 0.007(2)

0.20(4) 0.52(6) 0.31(6)

10 mN/m A 10 mN/m B 10 mN/m C

1.363(1) 1.481(2) 1.574 (2)

0.006(1) 0.024(4) 0.011(1)

0.22(3) 0.48(4) 0.29(4)

15 mN/m A 15 mN/m B 15mN/m C

1.371(1) 1.485(1) 1.573(1)

0.010(1) 0.020(1) 0.019(2)

0.15(3) 0.45(5) 0.32(16)

tion, and this gives GID Bragg peaks even at zero pressure. On the contrary, phospholipids do not show self-aggregation, and Bragg peaks are present only at pressures above the plateau. Clearly in the case of HSA, the presence of side groups prevents them from aggregating. This indirectly confirms our model for the molecular conformation on the water surface, which assumes that at low pressure the second OH group resides on the water surface. The GID curve of 12-HSA acid (top panel of Figure 5) consists of three peaks that could be fitted by a Lorentzian shape (peaks A-C). The peak positions and widths are reported in Table 1 also as a function of the applied pressure. The widths of peaks A and C appear to be resolution-limited, implying crystalline coherence extending at large distances, at least over 1000 Å, whereas peak B appears to be broader because of a geometrical distortion of the experimental geometry. We have repeated the measurements after increasing the surface pressure (Π ) 7, 10, and 15 mN/m), and the results are summarized in Table 1. No major differences are found between 7 and 10 mN/m. On the contrary, at 15 mN/m we observe the onset of collapse and 3D crystallization, as indicated by the periodic oscillations present in the rods of peaks B and C at this pressure. This is accompanied by a measurable reduction of the out-ofplane components of peaks A and B, indicative of a reduction of the tilt angle under increased lateral pressure. For the collapsed phase, from the period of the oscillation of the rods (dQ ) 0.143 (4) Å-1) we extract a value for the d spacing of d ) 44(1)Å that is slightly smaller than the spacing expected for a bilayer from the reflectivity data (46.7 Å), perhaps indicating some form of interdigitation of the alkyl chains in the 3D collapsed phase. The lattice parameters and molecular tilting angles obtained from the GID data are shown in Table 2, where we report the parameters of both the primitive, oblique cell and of an equivalent body-centered, pseudo-rectangular cell, which is useful for the following comparison with the crystal structures of the other compounds. The angle γPRIM refers to the angle between a and b in the primitive cell, and γRECT is the distortion of the pseudorectangular cell (which would be truly rectangular if γRECT ) 90°). Tilt angle θ refers to the molecular inclination

Figure 6. (Main panel) Grazing incidence diffraction intensity map (counts are represented as gray intensity) from a Langmuir layer of 9-HSA at Π ) 12 mN/m as measured on the ID10B beamline. The graph in the top panel is the GID curve (i.e., the integrated projection of the 2D map on the Qxy). The right panel shows the Bragg rods (i.e., Qz dependence of the GID intensity integrated in the Qxy range around the peak).

with respect to the vertical, whereas the azimuth ψ is the angle between aPRIM and the projection of the tilted molecule, measured clockwise. Our results for the values of lattice parameters are consistent with those of ref 4 even if the data are not taken at the same temperature and pressures; note that in ref 4 lattice parameters a and b are exchanged. Furthermore, we observe a small increase in γPRIM and a decrease in tilt angle θ as pressure is increased, in agreement with the cited reference.4 It is also interesting that tilt angle θ directly measured by GID, in agreement with that from X-ray reflectivity, is smaller (θ ) 20°) than that deduced by in-situ polarized FT-IR external-reflection spectra (θ ) 28°) in the same region of the isotherm1. Because FT-IR analysis is based on the CH2 modes, it seems reasonable to conclude that the molecules are not straight rods and different parts experience different orientations. 9-Hydroxystearic Acid. In Figure 6, we report the GID pattern of 9-HSA at Π ) 12 mN/m (i.e., at a pressure just above its plateau). The pattern consists of two peaks only, as indicated in Figure 6, indicating rectangular symmetry. In this case, we have also repeated the scan three times, with a 54-min interval between each scan, to detect any time evolution in the GID pattern. We find that whereas the position of peak A, at Qxy ) 1.465(1) Å-1, is timeindependent within the experimental accuracy, peak B shows a temporal evolution, especially in the out-of-plane component, as shown in Table 3. Note in particular the narrowing of the half width of both peak positions and a correlated increase in the out-of-plane position. This implies that in time the system evolves toward a better-

Table 2. Lattice Parameters and Tilting Angles for 12-HSA as a Function of Applied Pressurea

Π ) 7 mN/m Π ) 10 mN/m Π ) 15 mN/m

aPRIM (Å)

bPRIM (Å)

γPRIM (deg)

area (Å2)

aRECT (Å)

bRECT (Å)

γRECT (deg)

tilt θ (deg)

azimuth ψ (deg)

5.00(1) 5.00(1) 5.00(1)

4.62(1) 4.61(1) 4.61(1)

112.8 112.9 113.3

21.3(1) 21.3(1) 21.7(1)

5.33(1) 5.32(1) 5.29(1)

8.01(1) 8.01(1) 8.03(1)

85.0 84.8 84.9

19.8 18.0 17.2

43.1 37.4 46.0

a Results for the primitive (oblique) cell are shown, together with the equivalent body-centered, pseudo-rectangular cell. Tilt angle θ refers to the molecular inclination with respect to the vertical, whereas azimuth ψ is the angle between aPRIM and the projection of the tilted molecule, measured clockwise.

11218

Langmuir, Vol. 21, No. 24, 2005

Cristofolini et al.

Table 3. Position and Width of the GID Peaks for 9-HSA at Constant Surface Pressure as a Function of Time

first scan B second scan B third scan B

Qxy (Å-1)

fwhm Qxy (Å-1)

Qz (Å-1)

fwhm Qz (Å-1)

1.487(3) 1.488(1) 1.489(1)

0.017(6) 0.008(2) 0.010(2)

0.32(6) 0.38(4) 0.38(4)

0.36(16) 0.27(10) 0.27(9)

defined crystalline structure, in agreement with the images shown in Figure 2. The twofold splitting, with both of the peak maxima at Qz > 0, implies rectangular symmetry with molecular tilting toward the next nearest neighbors (NNN), in agreement with the data in the literature4 in which the rectangular symmetry deduced from the GID pattern seems to be distorted if one takes the lattice parameters reported in Table 1 of the paper. Given the 2:1 rule for the ratio of the Qz component of the nondegenerate to degenerate peak, we identify peak A with (1, 1) and (1, -1) and peak B with (0, 2). We therefore extract the lattice parameters as summarized in Table 4. We note that the area per molecule is stable with time and is comparable to the value from the isotherm at this pressure and slightly smaller than for the 12-HSA. On the contrary, from the reported increase of the position of the out-of-plane component (Table 3), we calculate an increase in the tilt angle upon annealing of the film, reaching the equilibrium value of θ ) 14.3°. If we further increase the surface pressure, 3D crystallization occurs, with Debye rings appearing as a signature (data not shown). 7-Hydroxystearic and 2-Hydroxystearic Acids. The GID intensity from 7-HSA at Π ) 16 mN/m is shown in Figure 7. It consists of a single peak at position Qxy ) 1.4578(4)Å-1 (quality factor for the fit χ2 ) 1.1), implying the packing of molecules in hexagonal symmetry. To compare with the other HSA crystalline structures, we chose to represent the cell with an equivalent centered rectangular lattice with a side-length ratio of b/a ) x3. The lattice parameters are then a ) 4.977Å and b ) 8.620 Å. The peak width fwhm ) 0.007(1) Å-1 is almost resolution-limited, indicating large crystalline domains. Hexagonal symmetry correlates nicely with the round shape of the crystallites observed in the ellipsometric imaging of 7-HSA, in contrast to the acicular shape observed for 9-HSA and 12-HSA (Figure 2). It is remarkable, however, that this single peak in the GID pattern is characterized by a measurably nonzero out-of-plane component, as shown in the inset of Figure 7, where the dots represent the measured integrated intensity and the continuous line is a fit of the peak with a Lorentzian shape whose maximum is located at Qz ) 0.25(7) Å-1. Assuming a rigid rodlike molecule, this value should correspond to a tilt angle of θ ) 9.7°, which, however, cannot be ascribed to the undistorted hexagonal

Figure 7. (Main frame) Grazing incidence diffraction intensity map (counts are represented as gray intensity) from a Langmuir layer of 7-HSA at Π ) 16 mN/m as measured on the ID10B beamline. (Top graph) GID curve (i.e., projection of the 2D map on the Qxy axis). (Right panel) Bragg rod (i.e., Qz dependence of the GID intensity integrated in the Qxy range around the peak, between 1.42 and 1.47 Å-1). The continuous line is the fit of the peak. (See the text for details.)

phase or to the so-called rotator phase,16 which should have maximum intensity at Qz ) 0. Such an apparently self-contradictory result could be explained by assuming that the 7-HSA molecule is not a rigid rod but presents a kink in the correspondence of the second OH group. This would allow truly hexagonal packing of the centers of the molecules, which would interact via the submerged polar heads, together with the presence of a tilt in the alkyl chains pointing out of the water surface. To test this hypothesis, additional measurements are needed (e.g., by neutron reflectivity with partially deuterated HSA). Finally, the GID pattern from 2-HSA, measured at Π ) 30 mN/m, as shown in Figure 8, is similar to that of 7-HSA, consisting of a single peak implying hexagonal symmetry, but with a much broader width (position Qxy ) 1.480(3)Å-1, width fwhm ) 0.05(1) Å-1), the hexagonal cell is equivalent to a centered rectangular lattice with lattice parameters of a ) 4.90(1)Å and b ) 8.49(1) Å. From the large value of the width of the peak, we can estimate the coherence length to be only about 30 lattice parameters in each direction, in contrast to the much more ordered 7-HSA lattice. As it was for 7-HSA, it is remarkable that this single peak is characterized by a measurably nonzero out-ofplane component, as shown in the inset of Figure 8. The peak maximum is located at Qz ) 0.29(1)Å-1. One could assume a rigid rodlike molecule with a tilt angle of 11°;

Table 4. Structure Parameters Extracted from GID Data for 9-HSA at Constant Surface Pressure a ) 4π/ (4Qxy(11))2 - (Qxy(02))2 (Å)

b ) 4π/Qxy(02) (Å)

area per molecule (Å2)

tilt angle θ ) a tan(Q02z/Q02xy)

4.98(1)

8.44(1)

21.0(1)

14.3°

x

Table 5. Summary of the Results for the Different 2D Crystalline Structures Found for the Different HSAs Studied molecule

plateau Π (mN/m)

symmetry

aRECT (Å)

bRECT (Å)

area (Å2)

cell angle γ (deg)

molecular tilt θ (deg)

2-HSA 7-HSA 9-HSA 12-HSA

? 14 11 6.5

HEX HEX RECT OBL

4.90(1) 4.98(1) 4.98(1) 5.33(1)

8.49(1) 8.62(1) 8.44(1) 8.01(1)

20.8(1) 21.4(1) 21.0(1) 21.3(1)

90 90 90 85.0

11 9.7 14.3 19.8

OH Group Position and Dimensionality Effect

Figure 8. (Main frame) Grazing incidence diffraction intensity map (counts are represented as gray intensity) from a Langmuir layer of 2-HSA at T ) 6.5 °C and Π ) 30 mN/m as measured on the ID10B beamline. (Top graph) GID curve (i.e., projection of the 2D map on the Qxy axis). (Right panel) Bragg rod (i.e., Qz dependence of the GID intensity integrated in the Qxy range around the peak, between 1.42 and 1.52 Å-1).

however, the possibility of a kink near the water surface should also be considered. Conclusions We have measured the crystalline structure of different HSA monolayers at the air-water interface, as summarized in Table 5, and have demonstrated a specific correlation between the position of the secondary OH group and the crystalline symmetry of the monolayer, which decreases with decreasing separation between the two competing hydrophilic groups, reaching hexagonal symmetry for 7-HSA. These data, together with some discrepancies in the data present in the literature, indicate that (i) the secondary OH groups might form a hydrogen bond network (16) Kenn, R. M.; Biihm, C.; Bib, A. M.; Peterson, I. R.; Mohwald, H.; Als-Nielsen, J.; Kjaer, K. J. Phys. Chem. 1991, 95, 2092-2097.

Langmuir, Vol. 21, No. 24, 2005 11219

even when the molecules are confined as Langmuir monolayers and (ii) the molecules are not rigid rods but rather might present a kink in correspondence of the second OH group. Both conclusions are clearly confirmed by the observed behavior of the electron density profiles obtained by reflectometry. Furthermore, the change in symmetry at 7-HSA, with 2-HSA and 7-HSA both showing hexagonal packing, indicates very clearly the importance of the competing interactions: 2-HSA and 7-HSA show a highly symmetric hexagonal phase; thereafter, the symmetry changes, accompanied by a change in the molecular tilt, leading to similar behavior for the 9-HSA and 12HSA members of this family. Thus, it is around n ) 7 (further data are necessary to pinpoint the position more precisely) that the competition between hydrophobic and hydrophilic interactions equilibrates: For small values of n, hydrophilic interaction will dominate, and the behavior will tend to that of pure HSA. For higher values of n, the OH will be too far from the water interface; therefore, hydrogen bonding or hydrophobic interactions will become more important, and 2D crystallization will take place. The symmetry lowering as the secondary OH group is placed further away from the interface can be interpreted as being partially due to entropic effects. In fact, if the alkyl chain is tilted at the position of the secondary OH group, then the remaining alkyl chain will have from 11 (n ) 7) to 6 carbons. It is reasonable to assume that the longer residual chains will be merely flexible and hence sample more orientational conformations, leading to the higher hexagonal symmetry of 7-HSA monolayers; conversely, the more rigid short residual chains of 12-HSA lead to tilted, more compact molecular arrangements of lower symmetry. We feel that our data and the analysis presented here put on a more quantitative, microscopic footing ideas about the correlation between the position of the secondary OH group and the general phenomenology of these interesting systems, with particular reference to the role of morphology on several space scales in the mechanical and viscoelastic behavior. Clearly our data should be supplemented by further studies, particularly of the lower n members of this interesting family of fatty acids. LA0514213