1500
Langmuir 2009, 25, 1500-1508
Two-Dimensional Lamellar Phase of Poly(styrene sulfonate) Adsorbed onto an Oppositely Charged Lipid Monolayer Jens-Uwe Gu¨nther, Heiko Ahrens, and Christiane A. Helm* Institut fu¨r Physik, UniVersität Greifswald, Felix-Hausdorff-Strasse 6, D-17487 Greifswald, Germany ReceiVed September 11, 2008. ReVised Manuscript ReceiVed NoVember 7, 2008 Polystyrene sulfonate (PSS 77 kDa) adsorbed onto oppositely charged dioctadecyldimethylammonium bromide (DODA) monolayers at the air/water interface is investigated with X-ray reflectivity and grazing incidence diffraction. The alkyl tails of DODA in the condensed phase form an oblique lattice with large tilts and intermediate azimuth angle. On PSS adsorption, the alkyl tail structure is maintained; only the tilt angle changes. Bragg peaks caused by flatly adsorbed, aligned PSS chains are observed, when DODA is in the fluid and also when it is in the condensed phase. The two-dimensional lamellar phase is only found at intermediate PSS bulk concentrations (0.001-1 mmol/L). In this phase, the PSS coverage can be varied by a factor of 3, depending on DODA molecular area and polymer bulk concentration. Charge compensation in the lamellar phase is almost achieved at 1 mmol/L. At larger bulk concentrations, PSS adsorbs flatly yet without chain alignment. Presumably, a necessary condition for a two-dimensional lamellar phase is a pronounced electrostatic force which causes a large persistence length as well as repulsion between the aligned chains.
Introduction We study experimentally the adsorption of polyelectrolytes onto an oppositely charged surface. In the past decade, the formation of polyelectrolyte multilayers by sequential adsorption of positively and negatively charged polymers was studied extensively.1-3 Due to careful selection of the preparation conditions, the polyelectrolytes adsorb as stratified layers. The composition of the thin film along the surface normal can be controlled on the nanometer level. However, within the surface the lateral distribution of the molecules is little explored and considered random. We want to find the conditions under which adsorbed polyelectrolytes form a two-dimensional lamellar phase. This question is also interesting for lithographic purposes. Theoretical calculations predict thin adsorption layers for the case of stiff chains (which includes chain stiffening due to electrostatic monomer-monomer repulsion). The polymers essentially lie flat on the surface and can be considered as a two-dimensional layer. Furthermore, in the case of a large electrostatic screening length, a two-dimensional lamellar phase is predicted, where the polymer chains run parallel and do not overlap.4 Indeed, a lamellar phase has been observed for DNA, which is attributed to the large structural persistence length (g15 nm).5,6 For artificial polyelectrolytes with a persistence length of a few nanometers, only the flat adsorption layer has been observed, not a two-dimensional lamellar phase.7,8 We use a lipid monolayer at the air/water interface as a substrate. The advantage of this system is that the surface charge and the
lipid phase (fluid or condensed) can be controlled by adjusting the molecular area. Furthermore, since the substrate is mobile, the polyelectrolyte adjusts fast to a new surface charge.7 In this work, we use monolayers of the positively charged double-chain amphiphile DODA (dioctadecyldimethylammonium bromide) interacting with the strong polyanion PSS [poly(styrene sulfonate)]. However, adsorption of polyelectrolyte onto the lipid monolayer may affect the crystallographic lattice of the alkyl tails. The degree change depends on the exact nature of both the polyelectrolyte and the amphiphile.9-12 The condensed lipid phase as well as the two-dimensional lamellar phase of the PSS is investigated with X-ray diffraction and reflectivity. The concentration of the polyelectrolyte in solution is varied to find both the conditions for a two-dimensional lamellar phase and the changes occurring in the alkyl tail lattice. PSS provides strong contrast for X-ray methods due to the large electron density of the sulfonated styrene group. The analysis of grazing-incidence X-ray wide-angle diffraction caused by ordered alkyl tails is well established.13 Similarly, the analysis of grazing-incidence X-ray low-angle diffraction patterns from systems with lattice distances between 5 nm and 5 µm is very advanced (freeware IsGISAXS available at ESRF).14 Yet, we want to find if and how the adsorbed polyelectrolyte influences the lattice of the aligned alkyl tails. Since we have up to five first order peaks from different lattice planes, and no higher order peaks, we need to consider the X-ray methods in some detail.
Experimental and Analysis
* To whom correspondence should be addressed. E-mail: helm@ physik.uni-greifswald.de.
Lipids and Polymers. DODA is used as the lipid (Avanti, Birmingham, AL) and PSS (Polymer Standard Services, Mainz, Germany, 77 kDa, corresponding to N ) 370) as polymer. The PSS
(1) Decher, G. Science 1997, 277, 1232–1237. (2) Klitzing, R. v. Phys. Chem. Chem. Phys. 2006, 8, 5012–5033. (3) Scho¨nhoff, M. J. Phys.: Condens. Matter 2003, 15, R1781-R1808. (4) Netz, R. R.; Joanny, J.-F. Macromolecules 1999, 32, 9013–9025. (5) Symietz, C.; Schneider, M.; Brezesinski, G.; Mo¨hwald, H. Macromolecules 2004, 37, 3865–3873. (6) Clausen-Schaumann, H.; Gaub, H. E. Langmuir 1999, 15, 8246–8251. (7) Ahrens, H.; Baltes, H.; Schmitt, J.; Mo¨hwald, H.; Helm, C. A. Macromolecules 2001, 34, 4504–4512. (8) Meijere, K. d.; Brezesinski, G.; Mo¨hwald, H. Macromolecules 1997, 30, 2337–2342.
(9) Meijere, K. d.; Brezesinski, G.; Kjaer, K.; Mo¨hwald, H. Langmuir 1998, 14, 4204–4209. (10) Goubard, F.; Fichet, O.; Teyssie´, D.; Fontaine, P.; Goldmann, M. J. Colloid Interface Sci. 2007, 306, 82–88. (11) Engelking, J.; Menzel, H. Thin Solid Films 1998, 329, 90–95. (12) Taylor, D. J. F.; Thomas, R. K.; Li, P. X. Langmuir 2003, 19, 3712–3719. (13) Kaganer, V. M.; Mo¨hwald, H.; Dutta, P. ReV. Mod. Phys. 1999, 71, 779– 819. (14) Mu¨ller-Buschbaum, P.; Gutmann, J. S.; Stamm, M.; Cubitt, R.; Cunis, S.; Krosigk, G. v.; Gehrke, R.; Petry, W. Physica B 2000, 283, 53–59.
10.1021/la802987k CCC: $40.75 2009 American Chemical Society Published on Web 01/06/2009
PSS Adsorbed onto Oppositely Charged Monolayers
Langmuir, Vol. 25, No. 3, 2009 1501
concentration in the subphase (cPSS) is always given with respect to the monomer concentration. Monolayer Preparation and Isotherms. The water used for all experiments was Millipore purified and deionized (18.2 MΩ). Chloroform/methanol (3:1) solutions of the DODA are spread over the aqueous polyelectrolyte solutions in a Langmuir trough. To measure isotherms, the DODA molecular area ADODA is adjusted by a movable barrier, and the pressure is detected by a Wilhelmy balance. The temperature is kept at 20 °C, and the equilibrium time after monolayer spreading is at least 30 min. A whole experiment usually takes 12-24 h. X-ray Reflectivity. The X-ray reflectivity of a monolayer at the air/water interface is given as a function of the perpendicular scattering vector Qz ) 4π/λ sin R, with R being the incidence angle and λ the X-ray wavelength. In the kinematic approximation, the reflectivity (R) divided by the Fresnel reflectivity (RF) expected for an infinitely sharp interface is given by15
|
R(Qz) 1 ) RF(Qz) Fsub
exp(iQz · z) dz| ∫ dF(z) dz
2
(1)
The electron density of the semi-infinite subphase Fsub below the interface is constant (Fsub ) 0.334 e/Å3 for water). To derive the vertical electron density profile F(z), one has to use models. A successful strategy is to represent the monolayer as a stack of slabs, each with a constant density Fi and thickness li. The model density must be smeared at slab interfaces to account for the intrinsic vertical roughness or diffuseness of the interfaces. The roughness, σ, is on the order of 3-4 Å and stems mainly from thermally excited capillary waves on the water surface. The rather large value means that in the atomic model of F(z) it is unnecessary to use accurate charge densities of the atoms. As long as each atom contributes its proper charge Zi, the description will be adequate. The large value of σ is also the reason that the monolayer may be adequately represented by a slab model. To allow for multiple reflections, as well as absorption effects close to the critical angle, for the actual data fitting the exact Fresnel equations are used.16 For a lipid monolayer the electron density profile is described by a model consisting of two slabs, one for the tails and one for the headgroup. Since Ftail ≈ Fsub and Fhead > Fsub, the position of the first minimum gives the length17
1 3π ltail + lhead ) 2 2Qz,min
(2)
X-ray reflectivity measurements up to Qz ) 0.55 Å- 1 are performed with a home-built setup described elsewhere.18 Grazing-Incidence X-ray Diffraction (GID). According to the geometry of diffraction,13 the scattering vector Q is described in terms of a horizontal component, Qxy, with
2π Qxy ) √cos2 Ri + cos2 Rf - 2 cos Ri cos Rf cos 2θxy (3) λ and a vertical component, Qz, with
Qz )
2π (sin Ri + sin Rf) λ
(4)
Ri (or Rf) is the angle between the plane of the liquid surface and the incident (or the diffracted) beam, and 2θxy is the angle between the incident and diffracted beams projected onto the horizontal (xy) plane. To reduce the background of photons scattered from the subphase, the incidence angle Ri is set below the critical angle Rc for total external reflection: Ri ) 0.85Rc. This limits the penetration to 50-100 Å.19 (If Rf , θxy, eq 3 would transform into the widely used approximation for small molecules, Qxy ≈ 4π/λ sin θxy.) The (15) Pershan, P. S.; Als-Nielsen, J. Phys. ReV. Lett. 1984, 52, 759–762. (16) Parratt, L. G. Phys. ReV. 1954, 95, 359–369. (17) Kjaer, K. Physica B 1994, 198, 100–109. (18) Baltes, H.; Schwendler, M.; Helm, C. A.; Mo¨hwald, H. J. Colloid Interface Sci. 1996, 178, 135–143.
Qxy positions of the Bragg peaks yield the two-dimensional lattice repeat distances
dhk )
2π Qhk xy
(5)
which can be marked by two indices (h,k) to yield the unit cell. A monolayer has no periodicity in the z direction; therefore, the scattering is concentrated along rods in reciprocal space. The measured intensity in the z-direction is only given by the atomic form factor
Ihk(Qz) ∝ |Fmol(Qz)|2 at Qxy ) Qhk xy
(6)
Due to the large interfacial rms roughness, in the z-direction, pseudoatoms consisting of groups of atoms are justified. The alkyl tails are described as cylinders with constant electron density aligned vertical to the surface. The cylinder length l is the thickness of the ordered part of the monolayer (i.e., the thickness of aliphatic tails of fatty acids or alcohols). One obtains17
1 sin lQz 2 Fmol(Qz) ) Fhk(Qz) ) 2 at Qxy ) Qhk xy Qz
( )
(7)
The central maximum of |Fmol(Qz)|2 occurs at Qz ) 0 and the next with a substantially decreased intensity (i.e., 3% of the intensity found at the central maximum) at Qz ) 3π/l. Equation 7 neglects roughness. To allow for the structural and diffusive roughness of the scattering centers, the vertical density profile is convoluted with a Gauss function with width ∆l [Fmol(Qz) is always calculated at Qxy ) Qhk xy; for brevity, this is not repeated in the following equations]20
1 sin2 lQz 2 |Fmol(Qz)| ) 4 exp(-Qz2∆l2) Qz2 2
( )
(7a)
However, the backbone of adsorbed polymer chains is found in the surface plane. Therefore, the polymer chains are modeled as horizontal cylinders. Due to the circular cross-section, we obtain a slightly changed atomic form factor20
Fmol(Qz) ) 2π
RJ1(RQz) Qz
(8)
Cylinders with radius R are used to describe the polymer chains, J1 is a Bessel function of the first kind and of the first order. Again, the central maximum occurs at Qz ) 0, the second maximum at Qz ) 5.13/R. Compared to the first maximum, its intensity is reduced to 1.7%. Due to the low intensity of the second maximum, we expect to see only first-order maxima (always true for alkyl tails in a hexagonal lattice13). For simplicity, all peaks are approximated as Gaussian curves in the z-direction. If the first and second maxima occur with comparable intensities at the same Qxy but at different Qz values, two scattering centers separated by L along the surface normal may be the reason (obviously, (19) Als-Nielsen, J.; Mo¨hwald, H.In Handbook of Synchrotron Radiation; Ebashi, S., Rubenstein, E., Koch, M., Eds.; North Holland, 1989; Vol. 5, pp 1-53. (20) Ahrens, H.; Papastavrou, G.; Schmidt, M.; Helm, C. A. J. Phys. Chem. B 2004, 108, 7080–7091.
1502 Langmuir, Vol. 25, No. 3, 2009
Günther et al.
sin ψ0hk
tan ψ0hk tan ψhk ) ) cos t cos ψ0hk cos t
(10a)
Yet, more interesting is the reciprocal lattice vector
Qhk xy )
Figure 1. Cylinder model of alkyl tail lattice (left) and its Fourier transform (right). The polar tilt angle t is given with respect to the surface plane. The tilt direction deviates from the side length connecting adjacent lattice points by the azimuthal tilt angle ψ. ψ is defined relative to that side length where ψ is smallest. In the reciprocal space, the Bragg rods of the alkyl tails are found in the plane perpendicular to the alkyl tails (shaded). Also given is the plane parallel to the surface in real space (indicated by a distorted hexagon).
(
0 2π 2π cos ψhk cos t ez × ghk ) 0 Adiffr sin ψ0hk dhk
)
(11)
Here, Adiffr is the area of the unit cell (i.e., twice the area of 0 0 0 the triangle with Adiffr ) ghkdhk ) ghk d hk/cos t), d hk is the absolute 0 value of dhk , and ez is the unit vector normal to the water surface. Because of cos2 ψ cos2 t + sin2 ψ ) 1 - cos2 ψ sin2 t, the absolute hk value of Qxy is22
Qhk xy )
2π √1 - cos2 ψ0hk sin2 t dhk0
(11a)
Correspondingly, the absolute value of Qzhk is22 that assumption needs to be verified). In that case, the scattering centers can be approximated by two Gaussians20
|Fmol(Qz)|2 ∝
(
|∫
exp -
∞
-∞
[ (
dz exp(iQzz) B exp -
(z + L/2) 2w22
2
) ]|
2
)
(z - L/2)2 + 2w12
(
w22 exp(-Qz2w22) + 2Bw1w2 exp -
w12 + w22
(
)
2
)
Q2xy + Q2z )
×
cos(QzL) (9) with w as the width of the Gaussians and B as a parameter to describe the relative amplitude of the two scattering centers (subscripts 1 and 2 denote the first and second peak, respectively). The extension of each scattering center perpendicular to the surface is defined as the fwhm (full width half-maximum) of the Gaussians given by Wi ) (8 ln 2)1/2wi ) 2.3wi. For ordered alkyl tails, the vertical scattering vector component can provide information about the polar tilt angle t and the tilt azimuth ψ of the rodlike alkyl tails (cf. Figure 1). The azimuthal angle Ψ is given with respect to one side length, denominated a, where its absolute value is smallest.13,21 Decrease of the DODA molecular area on monolayer compression can be caused either by an increase of the alkyl tail density or by a decrease of the polar tilt angle. To clarify this issue, the local coordinate system with the z0 axis parallel to the alkyl tails is considered. In general, ψhk is the tilt azimuth with respect to the interplanar spacing vector dhk. In the local coordinate system with the z0-axis parallel to the alkyl tails and the x0y0-plane tilted by angle t with respect to the surface plane (i.e., perpendicular to the alkyl tails, cf. 0 0 Figure 1), the corresponding parameters are ψhk and dhk . Therefore, 0 0 the lattice vector ghk is perpendicular to dhk , and the angle between 0 0 ghk and tilt direction is π/2 - ψhk . Expressing the lattice vector in the laboratory coordinate system means stretching the component of the lattice vector in the local coordinate system parallel to the tilt direction (the tilt direction is set parallel to the x-axis):
(
g0hk ) g0hk
sin ψ0hk -cos
ψ0hk
)
(
w ghk ) g0hk
sin ψ0hk/cos t -cos ψ0hk
(
ghk
)
2π cos ψ0hk sin t dhk0
(12)
Summarizing, one obtains
) B2w12 exp(-Qz2w12) + Qz2
Qhk z )
)
)
sin ψhk (10) -cos ψhk
For ψhk, this equation also yields the coordinate transformation from the laboratory to the local system (21) Jacquemain, D.; Wolf, S. G.; Leveiller, F.; Deutsch, M.; Kjaer, K.; AlsNielsen, J.; Lahav, M.; Leiserowitz, L. Angew. Chem., Int. Ed. Engl. 1992, 31, 130–152.
4π2 (d 0)2
(13)
Note that eq 13 describes a circle with radius 2π/d 0. The full width at half-maximum (fwhm) of the horizontal alkyl tail peaks is related to the positional correlation length ξ. For an exponential decay of positional correlation as observed in liquid crystals (corresponding to a Lorentzian as peak profile) it is
ξ)
2 fwhm(Qxy)
(14)
The diffraction data are represented by a two-dimensional intensity distribution plot, showing the diffracted intensity I(Qxy,Qz) as a function of Qxy and Qz. The GID measurements are performed at the liquid surface diffractometer, at the beamline BW1 in HASYLAB, Hamburg, Germany.23 The scattered intensity is detected by a linear position sensitive detector (PSD) equipped with Soller slits. The achieved resolution in Qz at low angles is detailed in the Supporting Information. For each monolayer, at selected molecular areas ADODA, both the small and large angle diffraction are measured in sequential compression-expansion cycles.
Results The Phases of the Lipid. Figure 2 shows the DODA isotherm and the diffraction patterns. The subphase is an aqueous solution of 10-4 mol/L PSS. From the established knowledge about lipid monolayers,24 the isotherm can be understood as follows: The first weak pressure increase is attributed to a fluid phase with disordered alkyl tails. The onset of a flat part in the isotherm (marked π1) indicates the start of a first-order phase transition toward a condensed phase with ordered alkyl tails. The fluid phase coexists with µm-sized domains of lipids in the condensed phase. At the end of the transition region, the steep pressure increase indicates that the condensed phase dominates. The phase of the ordered alkyl tails is investigated with GID experiments at large angles (Qxy ) 1.15-1.55 Å-1). Three peaks are observed, one at Qz ≈ 0.1-0.3 Å-1 and two overlapping peaks at large Qz. Detailed data analysis shows that at low surface pressures one (22) Weidemann, G.; Brezesinski, G.; Vollhardt, D.; Mo¨hwald, H. Langmuir 1998, 14, 6485–6492. (23) Kuzmenko, I.; Rapaport, H.; Kjaer, K.; Als-Nielsen, J.; Weissbuch, I.; Lahav, M.; Leiserowitz, L. Chem. ReV. 2001, 101, 1659–1696. (24) Mo¨hwald, H. Annu. ReV. Phys. Chem. 1990, 41, 441–476.
PSS Adsorbed onto Oppositely Charged Monolayers
Langmuir, Vol. 25, No. 3, 2009 1503
Figure 2. X-ray grazing incidence measurements at the DODA molecular areas indicated in the isotherm, with cPSS ) 10-4 mol/L (with respect to the monomer concentration). (a) Measurement is taken in the fluid phase, and a low angle Grazing Incidence Diffraction (GID) peak caused by aligned PSS chains is found. (b and c) Measurements are taken at the end of the coexistence region: at wide-angle GID the peaks due to the ordered alkyl tails are shown. At low-angle GID two PSS peaks are observed, one attributed to the PSS chains below the fluid phase and the other to the PSS chains below the condensed phase. The latter is found at larger Qxy and with a structured rod. (d and e) Measurements are taken in the high-pressure region: the alkyl tails are ordered, and only the low-angle peaks due to PSS chains below the condensed phase remain. The background is subtracted from all GID measurements. Contour plots are calculated from least-squares fits to the models outlined in the text. Table 1. Parameters of the Alkyl Tail Lattice As Derived from GID Measurements on a Subphase of 10-4 mol/L PSS during Compression π (mN/m)
ADODA (Å2)
Adiffr (Å2)
ha (Å)
hb(Å)
hc (Å)
a (Å)
b(Å)
c (Å)
ψ (deg)
t (deg)
17.4 24.7 32.7 39.8
59.4 54.8 52.6 51.1
52.02 50.28 50.61 50.27
4.53 4.55 4.60 4.62
4.66 4.51 4.49 4.47
5.05 4.92 4.90 4.90
5.75 5.51 5.46 5.45
5.6 5.57 5.60 5.62
5.16 5.10 5.14 5.13
9.02 10.00 11.31 11.69
38.8 36.3 37.6 37.8
of the latter peaks occurs at Qz ≈ 0.77 Å-1, the other at Qz ≈ 0.95 Å-1. On compression, they shift to Qz ≈ 0.73 and 0.9 Å-1, respectively. These features characterize tilted alkyl tails ordered in a oblique lattice, with a decreasing tilt angle on compression (39°f37°) (cf. Table 1). On compression, the azimuthal angle ψ increases slightly (from 9° to 12°), indicating that the tilt direction moves slightly away from mainly NN (nearest-neighbor) direction toward NNN (next-nearest-neighbor) direction, but being far from arriving there. As the distorted shape of the peaks with low Qz already suggests (the Qxy position of maximum intensity decreases on increase of Qz), the azimuthal angle is broadly distributed.22 The distribution of ψ is about 15°-20°; it does not change on compression. To better understand the influence of PSS on the phases of the lipid monolayer, isotherms are measured with different PSS concentrations in the subphase (cf. Figure 3, bottom). For various experiments, the same monolayer is used; only the subphase is exchanged. The onset of the phase transition (called π1) from liquid to the condensed phase does depend strongly on the PSS concentration in the subphase. With 10-5 mol/L in the subphase, π1 is much lower than on pure water (with 10-6 mol/L, there are pronounced differences between the first and second compression; therefore, we do not show an isotherm. Probably, it takes a few hours of equilibration time to obtain a reproducible isotherm.) Upon increasing the polymer bulk concentration, π1 increases, too. At 3 × 10-4 mol/L, the value is the same as on pure water; at 10-3 mol/L, it levels off. X-ray diffraction measurements of the alkyl tail lattice are qualitatively similar (cf. Figure 3, top). Always, one finds three broad Bragg peaks that are distorted and overlapping. Since at least two of these peaks occur at high Qz values, obviously the alkyl tails are in an oblique lattice. If one compares the measurement on clean water to one on a 10-6 mol/L PSS solution,
one observes that the overlapping peaks at large Qz shift to significantly lower Qz and larger Qxy values. This indicates a lower tilt angle t accompanied by shorter lattice distances. The latter suggests a smaller area per lipid molecule. Upon increasing the PSS bulk concentration, the overlapping peaks move back to large Qz and smaller Qxy values. Quantitative analysis shows that Adiffr, the DODA area determined from diffraction measurements, decreases on addition of PSS at low concentrations and then increases until it levels off at a value very similar to the one of clean water; a similar phenomenon is observed for the tilt angle (cf. Figure 3, bottom). At low PSS bulk concentration, few PSS molecules are adsorbed, leading to a decreased repulsion (or increased attraction) between the DODA molecules within the monolayer. This feature is evidenced both by the fact that the condensed phase occurs at a lower surface pressure π1 and that the molecular area in the condensed phase is decreased. To characterize the changes of the alkyl tail lattice on compression and on addition of PSS, the lattice constants a0, b0, and c0 in the local coordinate system are calculated (cf. Tables 1 and 2 for the quantitative values deduced from the measurements shown in Figure 2). In Figure 4, a sketch of the obtained lattice in the laboratory coordinate system is given. The interplanar spacings dhk we measure correspond to the heights ha, hb, and hc. On transformation from the laboratory to the local coordinate system, the height ha0 is least affected, since it is almost perpendicular to the tilt direction. However, both hb and hc exhibit small angles relative to the tilt direction; therefore, hb0 and h0c are much shorter than their counterparts in the laboratory coordinate system. Thus, ha0 ends up as the largest interplanar distance in the local coordinate system (even though its counterpart ha is not always the longest in the laboratory coordinate system). The peak positions as deduced from the least-squares fits are given in Figure 4. The data are analyzed according to eq 13: as
1504 Langmuir, Vol. 25, No. 3, 2009
Günther et al.
Figure 3. (Top) X-ray diffraction data from the alkyl tail lattice at 30-33 mN/m on clean water and on aqueous solutions containing different PSS bulk concentrations cPSS. The cross at (Qxy,Qz) ) (1.35, 0.8) is given as guide to the eye, to better recognize changes of the diffraction pattern. (Bottom) Compression isotherms of DODA on pure water (bold, gray) and with various PSS concentrations. The inset shows the results from the GID measurements: Adiffr, the area per DODA, and t, the tilt angle. Additionally, the liquid/condensed phase transition pressure π1 as a function of cPSS is given. Table 2. The Same Parameters of the Alkyl Tail Lattice as in Table 1, Yet Now in the Local Coordinate System with the z° Axis Parallel to the Alkyl Chains π (mN/m)
ADODA (Å2)
0 Adiffr (Å2)
h0a (Å)
h0b (Å)
h0c (Å)
a0 (Å)
b0 (Å)
c0 (Å)
ψ0 (deg)
t (deg)
17.4 24.7 32.7 39.8
59.4 54.8 52.6 51.1
40.64 40.42 39.86 39.72
4.5 4.52 4.55 4.56
4.04 4.00 3.96 3.95
4.03 4.02 3.93 3.92
4.52 4.48 4.38 4.36
5.02 5.05 5.03 5.03
5.04 5.03 5.07 5.07
11.51 12.35 14.17 14.68
38.8 36.3 37.6 37.8
long as the interplanar spacing in the local coordinate system is constant, the peak position should lie on the circumference of a circle. This assumption holds for the two peaks observed at high Qz, which give the shorter interplanar spacings (hb0 and h0c ) in the local coordinate system. The data analysis shows that on compression a decrease of the tilt angle t occurs, accompanied by a small change of the tilt azimuth. Yet, the interplanar separations (hb0 and h0c ) decrease a little (less than 2%). However, the shift for the peaks at low Qz does not follow a circle circumference. On compression, the corresponding interplanar spacing increases (by about 2%). The overall effect is a slightly decreased area per alkyl tail, even in the local coordinate system. Figure 4 shows more than the results obtained from the measurement on 10-4 mol/L PSS bulk solution (cf. Figure 2); additionally, the results obtained from DODA on clean water as well as on PSS solutions with concentrations between 10-6 and 10-3 mol/L are given. The results obtained at 10-4 mol/L polymer bulk concentration are confirmed: For large Qz, all peak positions coincide with the respective circle circumferences (within 2%), the distribution of the data points indicates an impressive range of tilt angles. Peak positions at low Qz do not coincide with a circle circumference; on monolayer compression the interplanar spacing increases. We have to conclude that PSS adsorption
does not change the structure of the alkyl tail lattice. In the laboratory coordinate system, the lattice is oblique. The three different interplanar separations that are accessible to diffraction measurements (ha0, hb0, and h0c ) are each different and independent of the PSS concentration in the subphase. On monolayer compression, they all change by at most 2%. Always, hb0 and h0c decrease and ha0 increases, as is listed for the data measured on 10-4 mol/L PSS solution in Table 2. ha0 is the largest interplanar distance and is oriented almost perpendicular to the tilt direction. The increase of ha0 corresponds to a decrease of the side length a0, which is almost parallel to the tilt direction (note that the other two side lengths b0 and c0 are almost constant). Stretching of the unit cell in the tilt direction has been observed before and explained theoretically.13 For the monolayer on 10-4 mol/L PSS solution, in the laboratory coordinate system the DODA area Adiffr determined from GID is decreased on compression by 3-4%, which is partly due to a decrease of the tilt angle and partly to an increase of the lateral alkyl tail density 1/A0diffr (cf. Tables 1 and 2). The molecular area in the local coordinate system, A0diffr, is found to be between 39.7 and 40.7 Å2; i.e., the variation amounts to about 2.5%. Since the three Bragg peaks are distorted and overlapping, the accuracy
PSS Adsorbed onto Oppositely Charged Monolayers
Langmuir, Vol. 25, No. 3, 2009 1505
Figure 4. (Left) Peak positions for DODA alkyl tail lattice on compression. According to eq 13, the radius of the respective circles is inversely proportional to the square of the measured interplanar distance in the local coordinate system. (Right) Sketch of the lattice in the laboratory coordinate system. On transformation from the laboratory to the local coordinate system, the height ha is least affected, since it is almost perpendicular to the tilt direction (indicated by a gray line). However, both hb and hc exhibit small angles relative to the tilt direction; therefore, h0b (green) and h0c (red) are much shorter than hb and hc. Thus, h0a (blue) ends up as the largest interplanar distance in the local coordinate system (even though ha is not necessarily the largest height in the laboratory coordinate system). For the measurements shown on the left, the subphase is pure water (crosses) or PSS with different concentrations (in mol/L): 10-6 (up triangles), 10-5 (down triangles), 10-4 (squares), and 10-3 (circles). The arrows indicate changes of the peak position on monolayer compression.
of the lattice constants is not ideal. The errors are also evident in the scatter of Figure 4. Nevertheless, upon adding PSS to a solution, clear changes in the hydrophobic moiety occur. In Figure 3, the changes in the DODA area Adiffr measured by GID at a selected surface pressure (30-33 mN/m) are shown as a function of cPSS, the PSS bulk concentration. Adiffr varies between 47.8 and 53.7 Å2, i.e., by 11%. The large changes (about 10° for the tilt angle t) occurring as function of the subphase composition are unambiguous. The observed change of Adiffr on PSS adsorption at a selected surface pressure is basically due to a change in the tilt angle, A0diffr ) Adiffr cos t ) 39.9-40 Å2. The Phases of the Adsorbed Polyelectrolyte. Very surprising is the occurrence of in-plane peaks at small angles (cf. Figure 2). In the fluid lipid phase, an in-plane peak at Qxy ≈ 0.13 Å-1 is observed. This Bragg peak corresponds to an interplanar spacing of ≈48 Å. On compression, this peak shifts to larger Qxy values; i.e., the interplanar spacing shrinks to ≈39 Å. This peak is attributed to PSS chains in the lamellar phase.5,20 Aligned PSS chains adsorb beneath the monolayer, the PSS coverage and thus the separation between aligned chains depends on the surface charge offered by the fluid lipid monolayer. In the coexistence region of the lipid monolayer, PSS peaks at two distinct Qxy values are observed: one at low Qxy due to PSS chains adsorbed beneath lipids in the fluid phase and the other one at larger Qxy (Qxy ≈ 0.25-0.27 Å-1 corresponding to 25-23 Å). The intensity distribution of the peak at larger Qxy is different in the Qz direction; it has two maxima at different Qz values. When the monolayer is very compressed and the lateral pressure exceeds 25 mN/m, only the latter remains. Therefore, it can be attributed to PSS adsorbed to lipids in the condensed phase. We need to understand the second maximum at high Qz. The first idea is to claim that it is a second-order maximum, according to equation 8. Its position measures the diameter 2R of the scattering center, according to Qz ) 5.13/R. Given that Qz ≈ 0.3 Å- 1, one obtains 2R ≈ 34.2 Å. This value is too high for a flatly adsorbed PSS layer; the PSS chain diameter is 11-12 Å.7,25 Another parameter suggesting that we do not deal with a second-order maximum is the intensity; a strong decrease relative (25) Ahrens, H.; Fo¨rster, S.; Helm, C. A. Phys. ReV. Lett. 1998, 81, 4172– 4175.
Figure 5. Normalized X-ray reflectivity curves of the DODA monolayers at 30-33 mN/m on clean water and with different PSS bulk concentrations. For clarity, each curve is displaced by 0.4. Points symbolize measurements, and the lines are least mean square fits derived from the electron density profiles shown in the inset. The electron density profiles of the monolayer on clean water and on 10-5 mol/L PSS are indicated. At concentrations exceeding 10-4 mol/L, the PSS coverage increases only a little.
to the central maximum at Qz ) 0 is expected. However, the intensity of the second maximum is only decreased by 50%. Thus, the two maxima of the rod suggest that besides the aligned PSS chains there is a second scattering center. A possible explanation is a corrugated lipid/air interface. The wavelength of the corrugation is the same as the separation of the aligned PSS chains. In this case, the Qz position of the second peak in the rod is determined by the vertical separation L of the two scattering centers as described in eq 9. To verify this idea, the electron density profile of the monolayer is measured independently by X-ray reflectivity (cf. Figure 5). All measurements are performed with the monolayer in the condensed phase, at a surface pressure of 30-33 mN/m.
1506 Langmuir, Vol. 25, No. 3, 2009
Günther et al.
Figure 6. (Left) Side view of the DODA monolayer in the condensed phase, with adsorbed PSS chains. The PSS chains are described as horizontal cylinders. (Bottom left) Varying electron density Fdiff(x,z) at various heights z. As an example, Fdiff(x,z) is given along the lines indicated in the schematic, in the center of the adsorbed cylinders (straight line) or below the centers (dotted line). Note that the width D(z) of the aligned PSS chains depends on the height z. (Center) Typical profile of the average electron density, Fav(z), together with the Fourier transformation F˜ diff(z) at a peak position (i.e., Qxy ) Qhk) determined from model calculations (cf. eqs 8, 9a, 16, and 17). Parameters for the calculations are FPSS ) 0.49 e/Å3, FH2O ) 0.334 e/Å3, Falkyl tails ) 0.30 e/Å3, cylinder separation d ) 26 Å, cylinder diameter 2R ) 12 Å, and for the corrugation at the air/film interface, w1 ) 1.44 Å. Also shown is the length L, which separates the centers of lateral structure. (Right) The length L as determined from rod scans (open symbols), as well as the separation ltail + 1/2lhead as determined from reflectivity (full symbols) versus the DODA molecular area ADODA. Table 3. Parameters of the Aligned PSS Chains on a Subphase of 10-4 mol/L PSS during Compressiona π (mN/m) 3.8 17.4 24.7 32.7 39.8 a
ADODA (Å2) 59.4 54.8 52.6 51.1
W1 (Å)
W2 (Å)
3.48 4.16 4.54 4.55
9.06 5.30 5.62 6.16 7.17
B 5.9 4.05 3.77 3.87
L (Å)
D (Å)
ξ (Å)
ASS (Å)
19.3 19.5 20.5 21.8
50.3 24.9 24.3 23.5 23.3
30.7 46.6 47.6 59.4 68.0
128.7 63.8 62.1 60.2 59.6
Index 1 denotes the peak caused by the corrugated air/water interface, and index 2 indicates the peak caused by the aligned PSS chains.
On clean water, the reflectivity curves show a very weak maximum at low Qz, followed by a minimum and a second maximum. On addition of PSS to the subphase, both maxima gain contrast and the minimum shifts slightly to the left. This effect is clear at a bulk concentration of 10-5 mol/L. For monomer concentrations exceeding 10-4 mol/L, the reflectivity curves are very similar, suggesting that the amount of adsorbed PSS increases only slightly. The electron density profile of the monolayer is consistent with the classical model for lipids (one slab for the tails, one slab for the headgroup, which consists of PSS chains and the hydrophilic part of the DODA). If PSS is adsorbed, the PSS chains provide the maximum in the vertical electron density profile. In the reflectivity curves, the high electron density of the adsorbed PSS chains increase the intensity of the maxima.7 The flatly adsorbed PSS also explains the shift of the minimum to slightly lower Qz values, the shift is due to a thickness increase of the headgroup. The separation between the laterally structured parts of the thin film should correspond to L ) 1/2lhead + ltail. L is determined both from the reflectivity measurements (cf. eq 2) and from the fits to the rods (cf. eq 9), the good agreement can be seen in Figure 6. Clearly, the adsorbed PSS introduces a disturbance in the lipid monolayer that extends to the air/film interface. The disturbance does not affect the density nor the crystalline phase of the alkyl tails; it leads to a corrugation of the air/water interface, akin to one-dimensional frozen surface wave, with a wavelength identical to the separation of the aligned PSS chains. Note that the PSS alignment is stabilized beneath the condensed lipids even with the surface corrugation. When the lipids are in the fluid phase, the PSS correlation length ξ (determined according to eq 14 from the width of the in-plane peak) is less than one chain separation (cf. Table 3). In contrast, when the lipids are in the condensed phase, ξ amounts to two to three chain separations.
Therefore, we can develop a model of the density distribution of the adsorbed PSS beneath the lipids in the condensed phase (cf. scheme in Figure 6). Usually, in crystallography we consider the order of some deformed spheres closed-packed in vacuum. Here, it is different: there are PSS chains with high electron density embedded in water with lower electron density. The electron density Ftotal is the sum of a laterally homogeneous part, F(z) (known from X-ray reflectivity measurements) and its local deviation Fdiff(x,y,z) resulting in
Ftotal(x, y, z) ) F(z) + Fdiff(x, y, z) with ∫∫dx dy Fdiff(x, y, z) ) 0 (15) The lateral structure exhibits a periodicity of width d, consisting of a stripe of width D(z) and electron density Fmaterial embedded in the surroundings. The latter condition of eq 15 is met when
D(z)(Fmaterial - F(z)) + (d - D(z))(Fsurroundings - F(z)) ) 0 (16) D(z) is the lateral width of the structure at height z. The adsorbed polymers are assumed to be cylinders (cf. schematics in Figure 6). Accordingly, the maximum of D(z) is the chain diameter, 2R. Considering the circular cross-sections of the scattering centers,20 the Fourier transformation of the lateral structure is calculated
F˜ diff(z) ) ∫∫dx dy Fdiff(x, y, z) exp[i(Qxx + Qyy)] ) d(Fmaterial - Fsurroundings) πD(z) at Qxy ) Qxyhk (17) sin 2π d
(
)
An analogous formula can be derived for the description of the corrugation at the air/film interface. Finally, the molecular form factor is given by a Fourier transformation in the z-direction
PSS Adsorbed onto Oppositely Charged Monolayers
|Fmol(Qz)|2 ) |∫ dz F˜ diff(z) exp(iQzz)|2
Langmuir, Vol. 25, No. 3, 2009 1507
(9a)
Approximations for eq 17 and consideration of two scattering centers give eq 9.20 Consistent with the X-ray reflectivity measurements, the PSS chain diameter is set to 12 Å, and the resulting F˜ diff(z) is sketched in Figure 6 (center). The difference Fmaterial - Fsurroundings is small (∼0.15 e/Å3), yet F˜ diff(z)extends the main part of the chain diameter in the vertical direction, leading to a measurable Bragg peak. As figures 2 and 6 suggest and Table 3 quantifies, the out-of-plane widths of the maxima give a thickness of the aligned PSS chains between 9 (lipids in the fluid phase) and 5-7 Å (lipids in the condensed phase). This deviation may be attributed to the fact that the PSS cross-section is not circular. PSS in all-trans conformation has the cross-section of a three-armed star.26 If one considers F˜ diff(z) at the corrugated film/air interface, the situation is reversed. The difference between the electron density of the alkyl tails and the air is large (∼0.30 e/Å3), yet the vertical extension is small (3.5-4.5 Å). Indeed, the fits show that the amplitude of the scattering center modeling the surface corrugation exceeds the one modeling the aligned chains by a factor 4-6. The electron density profiles at PSS bulk concentrations of 10-3 and 3 × 10-3 mol/L cannot be distinguished, even though the former does show a corrugated film/air interface, while the latter does not. Obviously, the structural roughness due to the corrugation of the film/air interface is so small (w1 ) 1.5-2 Å, cf. Table 3) that it does not influence the film/air roughness, which at large surface pressures is mainly due to a diffuse roughness caused by capillary waves (σ ≈ 4 Å).27 Factors Influencing PSS Coverage. The intensity of the PSS peaks depends on the PSS bulk concentration. The PSS in-plane peaks with the highest intensity occur at cPSS ) 10-4 mol/L. At 10-3 mol/L, the peaks are very weak when the lipids are in the condensed phase and impossible to resolve when the lipids are in the fluid phase. At cPSS ) 3 × 10-3, we do not observe any small-angle GID peaks. However, with X-ray reflectivity we find almost the same electron density profiles with a PSS concentration of 10-3 or 3 × 10-3 mol/L, at both PSS concentrations there is a well defined flat adsorption layer. To find out more about the lateral density of the PSS chains, the area per PSS monomer ASS of the aligned chains is determined. ASS is the product of the separation d of the aligned PSS chains and the length of a styrene sulfonate monomer (ASS ) d · 2.56 Å), cf. Figure 7. To determine if the adsorbed PSS compensates the surface charge of the DODA monolayer, the DODA molecular area needs to be considered. In the fluid phase the area per molecule is calculated from the isotherm (i.e., identical to ADODA) and in the condensed phase from the diffraction measurements (i.e., identical to Adiffr). In the coexistence phase, molecular areas are assigned to the fluid and the condensed phase, which are both assumed to be constant.28 Only their relative proportion changes. For the fluid phase, the value measured at the onset of the phase transition is chosen28 [i.e., identical to ADODA(π1)], whereas for the condensed phase, the largest molecular area found by diffraction measurements is selected (i.e., maximum of Adiffr for a specific monolayer). Note that in the coexistence region, the molecular area from the isotherm deviates strongly of the molecular area occurring in the respective phases. At a given PSS concentration, the monomer area ASS decreases when the monolayer is compressed (cf. Figure 7). If we focus (26) Donath, E.; Walter, D.; Shilov, V. N.; Knippel, E.; Budde, A.; Lowack, K.; Helm, C. A.; Mo¨hwald, H. Langmuir 1997, 13, 5294–5305. (27) Schlossman, M. L.; Schwartz, D. K.; Pershan, P. S.; Kawamoto, E. H.; Kellog, G. J.; Lee, S. Proc. R. Soc. London 1991, 66, 1599–1602. (28) Helm, C. A.; Mo¨hwald, H. J. Phys. Chem. 1988, 92, 1262–1266.
Figure 7. (Top) The area of a SS monomer, ASS, deduced from the GID measurements, as function of the DODA molecular area determined from the isotherm. Different PSS concentrations cPSS in the subphase are used (10-6, 10-5, 10-4, and 10-3 mol/L, with respect to the monomer concentration). On monolayer compression, ASS decreases. The phase transition from the DODA fluid to the condensed phase is marked by the coexistence of two different values of ASS. Bottom: The ratio between SS monomers to DODA molecules, nss/nDODA, as described in the text.
on the measurement depicted in Figure 2, the largest value for ASS is 141 Å2 (for ADODA ) 114 Å2). When the monolayer is compressed, ASS beneath lipids in the fluid phase decreases to 85 Å2. In the coexistence region, two very different lipid areas occur. In the same manner, two different monomer areas Ass can be measured simultaneously, 85 Å2 (lowest value found beneath lipids in the fluid phase) and 69 Å2 (the highest value found beneath lipids in the condensed phase). On further compression, ASS decreases only slightly beneath the lipids in the condensed phase, to 62 Å2. Summarizing, while the DODA molecular area ADODA decreased along the isotherm by a factor 2.2, the monomer area ASS decreased a bit more, by a factor 2.3. On increasing the PSS bulk concentration from 10-6 to 10-4 mol/L, the area per monomer ASS decreases (cf. Figure 7). At a selected DODA molecular area in the fluid phase, for instance ADODA ) 100 Å2, ASS changes from 170 to 130 Å2. The same trend is observed in the DODA condensed phase. If the extreme cases are considered, depending on the PSS concentration and the DODA molecular area, the monomer area ASS in the twodimensional lamellar phase can be varied by a factor 3.2 (from 170 to 53 Å2). To investigate charge compensation, the ratio between SS monomers and DODA molecules (nss/nDODA) is considered. On decrease of the DODA molecular area, there is a slight increase of nss/nDODA. The ratio between SS monomers and DODA molecules, nss/nDODA, is smaller in the fluid than in the condensed phase (cf. Figure 7). Increasing the PSS bulk concentration from 10-6 to 10-3 mol/L increases nss/nDODA until it approaches 1. nss/nDODA ) 1 is equivalent to charge compensation, and it is almost achieved when the lipids are in the condensed phase and cPSS ) 10-3 mol/L.
Discussion We find a two-dimensional lamellar phase of PSS adsorbed onto the oppositely charged DODA monolayer, if the bulk monomer concentration is low enough (e10-3 mol/L). The
1508 Langmuir, Vol. 25, No. 3, 2009
coverage of the adsorbed PSS in the two-dimensional lamellar phase is so low that it does not compensate the DODA surface charge. In the framework of DLVO theory, the positive net charge of the system DODA/PSS is compensated by some negative ions in the solution (Br-, OH-, HCO3-). The Br- ions are provided by the DODA molecules, when DODA with bound Br is spread on the water surface. Obviously, the DODA molecules dissociate; otherwise, we would observe a Br- layer with X-ray reflectivity on clean water (cf. Figure 5). Considering the volume of the Langmuir trough, the Br- concentration is less than 10-9 mol/L and can be neglected. Due to CO2 dissolution from the air into the water, the pH is about 5, suggesting a OH- concentration of about 10-9 mol/L, which can be neglected, too. Among the anions, the concentration of the HCO3- ions is largest, about 10-5 mol/L (similar to the H+ concentration). This ion pair is created by the dissociation of H2CO3 formed after dissolution of airborne CO2 in water. The amount of dissolved CO2 is not controlled). Anyway, on clean water, the charge of the DODA monolayer is compensated mainly by HCO3- ions. The cation concentration in the solution is dominated by the Na+ ions dissociated from of the PSS chains. Depending on the monomer bulk concentration, it varies in the experiments between 10-6 and 10-3 mol/L. According to theory,4 a necessary condition for PSS chain alignment within a flat adsorption layer is a pronounced electrostatic force that causes a large persistence length as well as repulsion between the aligned chains. The highest ion concentration when a two-dimensional lamellar phase is observed corresponds to a cation concentration of 10-3 mol/L and a somewhat smaller anion concentration. Thus, a lower limit of the screening length of the electrostatic interaction (the Debye length) is obtained, 96 Å. Diluting the anion or cation concentration increases the Debye length. Whatever the experimental details, when a two-dimensional lamellar phase is found, the Debye length exceeds by far the separation of the PSS chains (between 21 and 70 Å). This fact is consistent with an electrostatic repulsion between the chains. Note that the smallest lattice distance observed, 21 Å, still exceeds the chain diameter, 12 Å,7,25 by a factor 1.7. Obviously, the aligned PSS chains do not touch each other. As long as the net charge of the DODA/PSS monolayer is not compensated, additional anions (mainly HCO3-) are necessary to compensate the monolayer charge. When the DODA monolayer charge is overcompensated, additional Na+ ions are necessary to compensate the excess PSS charge. These cations shield the surface charge; obviously, they also decrease the repulsion between the PSS chains. Apparently, the increased cation concentration at the surface is the end of the two-dimensional lamellar phase, even though a flat adsorption layer is still maintained. At this point it is also necessary to mention that PSS adsorbs under certain conditions onto hydrophobic surfaces.7,25 A short-ranged attractive hydrophobic force between the PSS chains is counterproductive for a two-dimensional lamellar phase, since touching chains would start to form bundles instead of aligning parallel to each other. In previous studies it was found that the adsorbed polyelectrolyte changes the phase of the alkyl tails.5,6,9,10 We find not only the same oblique lattice of the alkyl tails but even the various interplanar spacings in the local coordinate system at a selected pressure are independent of the PSS bulk concentration. However, there is clear evidence that the PSS concentration in solution causes a change in the SS monomer area and the separation of the aligned PSS chains. Yet, these changes in the adsorption layer do not influence the symmetry of the alkyl tail lattice, only the tilt angle. To understand that it is helpful to consider the extremely small headgroup of DODA. Most natural lipids have headgroups whose molecular area is identical with or slightly larger than the area of two nontilted alkyl tails, ∼40 Å2.13 Therefore, the available headgroup area strongly influences the area of the alkyl tails in the condensed phase. DODA is different.
Günther et al.
The headgroup area (the area of the ammonium group) is smaller than the area of two nontilted alkyl tails. Therefore, the alkyl tail lattice is strongly determined by the interaction of the tails only. Changes of the headgroup which have an effect on the alkyl tail lattice could be possibly induced if a monovalent ion binds specifically to the headgroup, affecting the molecular area of each ammonium group. PSS alone has no effect. An effect that we do observe is a decrease of the molecular area, caused by a decrease in the tilt angle of the alkyl tails at large surface pressures on the introduction of very small PSS concentrations such as 10-5 mol/L (cf. Figure 3). Concomitantly, the transition pressure π1 decreases. This finding is attributed to a decreased electrostatic repulsion between the DODA molecules, since the surface charge is reduced by the adsorbed PSS chains.29 However, when more PSS chains adsorb, the transition pressure increases again; simultaneously, the DODA area in the diffraction measures grows. Apparently, the adsorbed chains repel each other entropically and add a lateral repulsion that compensates the decreased electrostatic repulsion.
Conclusion PSS adsorbed onto oppositely charged DODA monolayers at the air/water interface shows a two-dimensional lamellar phase if the PSS bulk concentration is low, between 10-6 and 10-3 mol/L (with respect to the monomer concentration), as is demonstrated by combining X-ray reflectivity and low-angle and wide-angle grazing-incidence diffraction. On monolayer compression, both the DODA molecular area and the PSS monomer area are decreased by about a factor of 2. When the lipids undergo a phase transition from fluid to condensed, both the DODA molecular area and the PSS monomer area are subject to a sudden decrease. When fluid and condensed phase coexist, two different PSS chain separations are found, depending if PSS is adsorbed beneath DODA in the fluid or in the condensed phase. The DODA crystalline lattice in the condensed phase is oblique and not disturbed by the PSS adsorption. Within 2%, all interplanar distances in the local coordinate system remain unchanged; on monolayer compression, the tilt angle is reduced and the lateral alkyl tail density is increased. At low PSS concentrations (e10-4 mol/L), the molecular area in the condensed phase as well as the surface pressure indicating the fluid/condensed phase transition are decreased, an effect that is attributed to PSS shielding the electrostatic repulsion of the DODA molecules. Within the two-dimensional lamellar phase, the area per PSS monomer can be varied by a factor of 3, dependent on the PSS concentration and the area per DODA molecule. On increase of the PSS concentration, more PSS is adsorbed until the monolayer surface charge is compensated. The PSS chain separation always exceeds the chain diameter by at least a factor 1.7. If the PSS concentration is further increased, the two-dimensional lamellar phase within the flat adsorption layer disappears. Apparently, the necessary condition for chain alignment is a pronounced electrostatic force that causes a large persistence length as well as repulsion between the aligned chains. Acknowledgment. Financial support of the SFB TR 24 and the BMBF (FKZ 03Z2CK1) with the ZIK HIKE project is acknowledged. We thank HASYLAB at DESY, Hamburg, for beam time and for providing all necessary facilities. Supporting Information Available: A description of the Qz resolution for GID measurements. This material is available free of charge via the Internet at http://pubs.acs.org. LA802987K (29) Helm, C. A.; Laxhuber, L. A.; Lo¨sche, M.; Mo¨hwald, H. Colloid Polym. Sci. 1986, 264, 46–55.
