Molecularly smooth cellulose surfaces for adhesion studies

Cellulose is deposited on silicon wafer surfaces via spin coating from a solution of cellulose in dimethylacetamide (+7% lithium chloride). The resulting cellulose layers were analyzed by ellipsometry, AFM, FTIR, ICP-MS, X-ray reflectivity, and contact angle measurements. For cellulose concentrations below 0.07 wt% the wafer surfaces are covered with a network of cellulose fibrils. For concentrations between 0.07 and 0.5 wt%, closed films with thicknesses between 1.5 and approximately 10 nm are obtained. These films are molecularly smooth (rms roughness<2 nm). Higher concentrations result in thicker films with significantly rougher surfaces (rms roughness>2 nm). The cellulose layers were used to investigate cellulose/cellulose adhesion and their modification by polyelectrolytes. To this end the sticking behavior of cellulose beads was analyzed. It is demonstrated that the sticking of the beads depends on the type, amount, and adsorption symmetry of adsorbed polyelectrolyte. Low, incomplete polyelectrolyte coverage always enhances sticking, whereas for high coverage the symmetry of the polyelectrolyte coating is very important. In this case, sticking (adhesion) is enhanced if only one surface is covered with polyelectrolyte prior to contact. If both surfaces were fully covered with polyelectrolytes before contact, sticking (adhesion) is decreased.     

Spectroscopic Ellipsometry Characterisation and Estimation of the Hamaker Constant of Cellulose

Calculations of Hamaker constants using Lifshitz theory require the availability of accurate dielectric data, especially in the visible-ultraviolet region. We present spectroscopic ellipsometry data on well-defined cellulose films of a limited thickness range (100–140 layers) deposited on an oxidised and hydrophobised silicon substrate. The spectral data, representing measurements from a perpendicular orientation to the fibre deposition direction, was used for estimates of the necessary spectral parameters, i.e. the oscillator strengths and characteristic frequencies in the UV-range. Our calculations show that cellulose has a relatively low Hamaker constant in air (58 zJ) and water (8.0 zJ). The implications for the surface energy estimates of cellulose and colloidal interactions between cellulose and various types of fillers and coating colours are indicated.     

AFM of adsorbed polyelectrolytes on cellulose I surfaces spin-coated on silicon wafers

Thin layers of cellulose I nanocrystals were spin-coated onto silicon wafers to give a flat model cellulose surface. A mild heat treatment was required to stabilize the cellulose layer. Interactions of this surface with polyelectrolyte layers and multilayers were probed by atomic force microscopy in water and dilute salt solutions. Deflection–distance curves for standard silicon nitride tips were measured for silicon, cellulose-coated silicon, and for polyelectrolytes adsorbed on the cellulose surface. Transfer of polymer to the tip was checked by running deflection–distance curves against clean silicon. Deflection–distance curves were relatively insensitive to adsorbed polyelectrolyte, but salt addition caused transfer of cationic polyelectrolyte to the tip, and swelling of the polyelectrolyte multilayers.     

Assembly of multilayer films from polyelectrolytes containing weak and strong acid moieties

Multilayer films were assembled from a copolymer containing both weakly and strongly charged pendant groups, poly(4-styrenesulfonic acid-co-maleic acid) (PSSMA), deposited in alternation with poly(allylamine hydrochloride) (PAH). The strongly charged groups (styrene sulfonate, SS) are expected to form electrostatic linkages (to enhance film stability), while the weakly charged groups (maleic acid, MA) can alter multilayer film properties because they are responsive to external pH changes. In this study, we varied several assembly conditions such as pH, SS/MA ratio in PSSMA, and the ionic strength of the polyelectrolyte solutions. The multilayer films were also treated by immersion into pH 2 and 11 solutions after assembly. Quartz crystal microgravimetry and UV-visible spectrophotometry showed that the thickness of PSSMA/PAH multilayers decreases with increasing assembly pH regardless of whether salt was present in the polyelectrolyte solutions. When no salt was added, the multilayers are thinner, smoother, and grow less regularly. Atomic force microscopy images indicate that the presence of salt in polyelectrolyte solutions results in rougher surface morphologies, and this effect is especially significant in multilayers assembled at pH 2 and pH 11. When both polyelectrolytes are adsorbed at conditions where they are highly charged, salt was necessary to promote regular multilayer growth. Fourier transform infrared spectroscopy studies show that the carboxylic acids in the multilayers are essentially ionized when assembled from different pHs in 0.5 M sodium chloride solutions, whereas some carboxylic acids remain protonated in the multilayers assembled from solutions with no added salt. This resulted in different pH stability regimes when the multilayers were exposed to different pH solutions, post assembly.     

Characteristics of polyelectrolyte multilayers: effect of PEI anchoring layer and posttreatment after deposition

The influence of a first (anchoring) layer and film treatment on the structure and properties of polyelectrolyte multilayer (PEM) films obtained from polyallylamine hydrochloride (PAH) and polysodium 4-styrenesulfonate (PSS) was studied. Branched polyethyleneimine (PEI) was used as an anchoring layer. The film thickness was measured by ellipsometry. Complementary X-ray reflectometry and AFM experiments were performed to study the change in the interfacial roughness. We found that the thickness of the PEM films increased linearly with the number of layers and depended on the presence of an anchoring PEI layer. Thicker films were obtained for multilayers having PEI as the first layer comparing to films having the same number of layers but consisting of PAH/PSS only. We investigated the wettability of PEM surfaces using direct image analysis of the shape of sessile water drops. Periodic oscillations in contact angle were observed. PAH-terminated films were more hydrophobic than films with PSS as the outermost layer. The effect of long time conditioning of PEM films in solutions of various pH's or salt (NaCl) concentrations was also examined. Salt or base solutions induced modification in wetting properties of the polyelectrolyte multilayers but had a negligible effect on the film thickness.     

Layer-by-layer electrostatic self-assembly of single-wall carbon nanotube polyelectrolytes

We have used anionic and cationic single-wall carbon nanotube polyelectrolytes (SWNT-PEs), prepared by the noncovalent adsorption of ionic naphthalene or pyrene derivatives on nanotube sidewalls, for the layer-by-layer self-assembly to prepare multilayers from carbon nanotubes with polycations, such as poly(diallyldimethylammonium) or poly(allylamine hydrochloride) (PDADMA or PAH, respectively), and polyanions (poly(styrenesulfonate), PSS). This is a general and powerful technique for the fabrication of thin carbon nanotube films of arbitrary composition and architecture and allows also an easy preparation of all-SWNT (SWNT/SWNT) multilayers. The multilayers were characterized with vis-near-IR spectroscopy, X-ray photoelectron spectroscopy (XPS), surface plasmon resonance (SPR) measurements, atomic force microscopy (AFM), and imaging ellipsometry. The charge compensation in multilayers is mainly intrinsic, which shows the electrostatic nature of the self-assembly process. The multilayer growth is linear after the initial layers, and in SWNT/polyelectrolyte films it can be greatly accelerated by increasing the ionic strength in the SWNT solution. However, SWNT/SWNT multilayers are much more inert to the effect of added electrolyte. In SWNT/SWNT multilayers, the adsorption results in the deposition of 1-3 theoretical nanotube monolayers per adsorbed layer, whereas the nominal SWNT layer thickness is 2-3 times higher in SWNT/polyelectrolyte films prepared with added electrolyte. AFM images show that the multilayers contain a random network of nanotube bundles lying on the surface. Flexible polyelectrolytes (e.g., PDADMA, PSS) probably surround the nanotubes and bind them together. On macroscopic scale, the surface roughness of the multilayers depends on the components and increases with the film thickness.     

Assembly of multilayered films using well-defined, end-labeled poly(acrylic acid): influence of molecular weight on exponential growth in a synthetic weak polyelectrolyte system

We report on the influence of polyanion molecular weight on the growth and structure of multilayered thin films fabricated from poly(allylamine) (PAH) and well-defined, end-labeled poly(acrylic acid) (PAA) synthesized by atom transfer radical polymerization. We observed striking differences in the growth of PAH/PAA films fabricated using well-defined PAA compared to films fabricated using higher molecular weight, commercially available PAA. Past studies demonstrate that the thicknesses of PAH/PAA films increase as linear functions of the number of PAH and PAA layers deposited over a broad range of pH (e.g., from pH 2.5 to 4.5). We observed the thicknesses of films fabricated using solutions of PAH and PAA adjusted to pH 7.5 and 3.5, respectively, to increase in a nonlinear manner. Films fabricated using well-defined, low molecular weight samples of PAA under these conditions increased in thickness exponentially. Experiments using samples of PAA having substantially non-overlapping molecular weight distributions demonstrated a clear relationship between the molecular weight of PAA and rates of film growth. We also used confocal microscopy, in combination with fluorescently end-labeled samples of PAA, to characterize the location of PAA in these assemblies. The results of these experiments, when combined, support the general conclusion that PAA is able to penetrate or diffuse into these films over large distances during assembly. The mechanism of growth for these films thus appears similar to that recently reported for the exponential growth of films fabricated using a variety of biologically relevant polyelectrolytes. The use of living/controlled methods of polymerization to synthesize well-defined samples of PAA facilitates an interpretation of these differences in film behavior as arising largely from differences in polymer molecular weight and polydispersity. This work provides insight into the assembly and structure of a well-studied weak polyelectrolyte film system and illustrates the potential of living/controlled methods of polymerization to contribute to the characterization and understanding of the physical properties of these ionically cross-linked materials.     

Adsorption of polyelectrolyte multilayers and complexes on silica and cellulose surfaces studied by QCM-D

The adsorption of polyelectrolyte (PE) multilayers and complexes, obtained from both high- and low-charge polyelectrolytes, was studied on silica and on cellulose model surfaces by quartz crystal microbalance with dissipation (QCM-D). The film properties acquired with the different strategies were compared. When polyelectrolytes were added on an oppositely charged surface in sequence to form multilayers both the change in frequency and dissipation increased. The changes in frequency and dissipation were clearly higher if low-charge PEs were used in the multilayer formation. The substrate, silica or cellulose, did not affect the adsorption behaviour of low-charge PEs and only minor differences were seen in the adsorbed amounts and changes in dissipation of high-charge PEs between SiO₂ and cellulose. The complexes formed by low-charge PEs had higher changes in frequency and dissipation at low ionic strength on both surfaces, while the complexes formed from high-charge polyelectrolytes adsorbed more at high salt concentration. The complexes of low-charge polyelectrolytes adsorbed more on silica, while the complexes formed by high-charge PEs formed thicker layers on cellulose. The charge ratio had a significant effect on the adsorption and the highest changes in frequency and dissipation were obtained in the anionic/cationic charge ratio of 0.5–0.6. Generally, the multilayers and complexes formed by low-charge polyacrylamides adsorbed highly and formed rather thick layers on both surfaces, unlike the high-charge PEs which formed thin layers using either one of the addition techniques.     

Langmuir-Blodgett films of cellulose nanocrystals: preparation and characterization

The goal of this work is the preparation of monolayers of cellulose I nanocrystals providing flat crystalline cellulose surfaces. Suspensions of cellulose nanocrystals were prepared by hydrolyzing ramie and tunicin fibers with sulfuric acid. Due to surface grafted sulfate groups, the negatively charged, rod-like cellulose nanocrystals were found to form stable layers at the air-water interface in the presence of a cationic amphiphilic molecule such as dioctadecyldimethylammonium (DODA) used in this work. These layers were formed at different cellulose-DODA weight ratios, compressed and analyzed by tensiometry, ellipsometry and Brewster angle microscopy. At low cellulose concentrations the layers are discontinuous, becoming dense and homogeneous upon reaching a critical weight ratio, which depends on the aspect ratio of the cellulose nanocrystals. After transfer onto silicon wafers, the surface composition and morphology as well as the thickness of the films were examined by X-ray photoelectron spectroscopy, ellipsometry and atomic force microscopy. The results indicate that they are monolayer films, well structured, relatively smooth and pure. These films offer a crystalline and easily reproducible model cellulose surface.     

Adsorption of polyelectrolyte-nanoparticle systems on silica: influence of ionic strength

In this work the effect of ionic strength on the adsorption behavior of cationic polyelectrolyte (acrylamide-acrylamidopropyltrimethylammonium chloride) and negatively charged silica particles has been studied by means of ellipsometry. The adsorption of the polyelectrolyte was observed to increase with increasing salt concentration, a behavior typical for polyelectrolytes with a screening-reduced solvency and a nonelectrostatic affinity for the surface. A similar dependence on the ionic strength was observed when studying the electrolyte effect on the nanoparticle adsorption to the preadsorbed polyelectrolyte film, suggesting that the polyelectrolyte surface conformations largely govern the binding capacity of the particles to the surface.     

Study of film structure and adsorption kinetics of polyelectrolyte multilayer films: effect of pH and polymer concentration

The alternate adsorption of polycation poly(allylamine hydrochloride)(PAH) and the sodium salt of the polymeric dye poly(1-[ p-(3'-carboxy-4'-hydroxyphenylazo)benzenesulfonamido]-1,2-ethandiyl)(PCBS) on quartz crystals coated with silica was studied to understand the structural properties and adsorption kinetics of these films using a combination of quartz crystal microbalance with dissipation monitoring (QCM-D), absorbance, and ellipsometry measurements. In-situ deposition of the polycation PAH on QCM crystals was monitored, followed by rinsing with water and then deposition of the polyanion PCBS. The effects of polymer concentration and pH on film structure, composition and adsorption kinetics were probed. The polymers were adsorbed at neutral pH conditions and at elevated pH conditions where PAH was essentially uncharged to obtain much thicker films. The change in the resonant frequency, Deltaf, of the QCM-D showed a linear decrease with the number of bilayers, a finding consistent with absorbance and ellipsometric thickness measurements which showed linear growth of film thickness. By using the Delta f ratios of PCBS to PAH, the molar ratios of repeat units of PCBS to PAH in the bilayer films as determined by QCM-D were approximately 1:1 at polyelectrolyte concentrations 5-10 mM repeat unit, indicating complete dissociation of the ionic groups. The frequency and dissipation data from the QCM-D experiments were analyzed with the Voigt model to estimate the thickness of the hydrated films which were then compared with thicknesses of dry films measured by ellipsometry. This led to estimates of the water content of the films to be approximately 45 wt %. In addition to the QCM-D, some films were also characterized by a QCM which measures only the first harmonic without dissipation monitoring. For the deposition conditions studied, the deposited mass values measured by the QCM's first harmonic were similar to the results obtained using higher harmonics from QCM-D, indicating that the self-assembled polyelectrolyte films were rigid.     

Internal structure of polyelectrolyte multilayers probed via neutral impact collision ion scattering spectroscopy

Layering in polyelectrolyte multilayer films has been studied by neutral impact collision ion scattering spectroscopy. The method affords a direct look at vertical ordering within these films at the nanometer scale. By labeling certain polyelectrolyte layers with heavy atom (Ru) probes, sufficient contrast has been obtained to visualize and quantify the distribution of these labeled polyelectrolytes throughout the film. The results indicate that the materials under investigation here produce linear film growth with very limited layer interpenetration. The interdigitation length between neighboring layers within the film is measured as 3.9 nm, which is approximately 1.4 times the thickness of an individual polycation/polyanion pair and is slightly less than the measured air/film roughness (4.7 nm). Detailed analysis shows that under the conditions used and at the depths probed in this study, the observed layer thickness is not significantly broadened by either instrumental or stochastic factors.     

Human serum albumin self-assembly on weak polyelectrolyte multilayer films structurally modified by pH changes

Adsorption of proteins onto film surfaces built up layer by layer from oppositely charged polyelectrolytes is a complex phenomenon, governed by electrostatic forces, hydrogen bonds, and hydrophobic interactions. The amounts of the interacting charges, however, both in polyelectrolytes and in proteins adsorbed on such films are a function of the pH of the solution. In addition, the number and the accessibility of free charges in proteins depend on the secondary structure of the protein. The subtle interplay of all these factors determines the adsorption of the proteins onto the polyelectrolyte film surfaces. We investigated the effect of these parameters for polyelectrolyte films built up from weak "protein-like" polyelectrolytes (i.e., polypeptides), poly(L-lysine) (PLL), and poly(glutamic acid) (PGA) and for the adsorption of human serum albumin (HSA) onto these films in the pH range 3.0-10.5. It was found that the buildup of the polyelectrolyte films is not a simple function of the pure charges of the individual polyelectrolytes, as estimated from their respective pKa values. The adsorption of HSA onto (PLL/PGA)n films depended strongly on the polyelectrolyte terminating the film. For PLL-terminated polyelectrolyte films, at low pH, repulsion, as expected, is limiting the adsorption of HSA (having net positive charge below pH 4.6) since PLL is also positively charged here. At high pH values, an unexpected HSA uptake was found on the PGA-ending films, even when both PGA and HSA were negatively charged. It is suggested that the higher surface rugosity and the decrease of the alpha-helix content at basic pH values (making accessible certain charged groups of the protein for interactions with the polyelectrolyte film) could explain this behavior.     

Equilibrium aspects of polycation adsorption on silica surface: how the adsorbed layer responds to changes in bulk solution

Adsorption of cationic high molecular weight polyacrylamides (CPAM) (M(w) is about 800 kDa) with different fractions of cationic units tau = 0.09 and tau = 0.018 onto silica surface was studied over a wide range of pH (4-9) and KCl concentration (c(s) = 10(-3)-10(-1) M) by in-situ null ellipsometry. We discuss how the adsorbed layer depends on the bulk conditions as well as kinetically responds to changes in solution conditions. The adsorbed amount Gamma of CPAM increases with pH for all studied electrolyte concentrations until a plateau Gamma is reached at pH > 6. At low pH we observed an increase in adsorbed amount with electrolyte concentration. At high pH there is no remarkable influence of added salt on the values of the adsorbed amount. The thickness of adsorbed polymer layers, obtained by ellipsometry, increases with electrolyte concentration and decreases with pH. At low c(s) and high pH the polyelectrolyte adsorbs in a flat conformation. An overcompensation of the surface charge (charge reversal) by the adsorbed polyelectrolyte is observed at high c(s) and low pH. To reveal the reversibility of the polyelectrolyte adsorption with respect to the adsorbed amount and layer thickness, parameters such as polyelectrolyte concentration (c(p)), c(s), and pH were changed during the experiment. Generally, similar adsorbed layer properties were obtained independent of whether adsorption was obtained directly to initially bare surface or by changing pH, c(s), or the concentration of polyelectrolyte solution in the presence of a preadsorbed layer, provided that the coverage of the preadsorbed layer was low. Once a steady state of the measured parameters (Gamma, d) was reached, experimental conditions were restored to the original values and corresponding changes in Gamma and adsorbed layer thickness were recorded. For initially low surface coverage it was impossible to restore the layer properties, and in this case we always ended up with higher coverage than the initial values. For initial high surface coverage it was usually possible to restore the initial layer properties. Thus, we concluded that polyelectrolyte appears only partially reversible to changes in the solution conditions due the slow rearrangement process within the adsorbed layer.     

Non-electrostatic building of biomimetic cellulose-xyloglucan multilayers

Layer-by-layer assembly was used to build thin films, consisting of multiple layers alternating cellulose nanocrystals and xyloglucan, benefiting from the strong non-electrostatic cellulose-xyloglucan interaction. Data from atomic force microscopy and neutron reflectivity showed that these well-defined films exhibited a thickness increasing linearly with the number of layers, without increase in surface roughness. These "green" nanocomposite films, reminiscent of plant cell wall, are composed of a regular stack of single layers of cellulose nanocrystals separated by very thin xyloglucan spacers. Such architecture differs from the one formed by cellulose/polycations multilayers, where the cellulose phase itself consists of a double layer.     

Influence of the polyelectrolyte molecular weight on exponentially growing multilayer films in the linear regime

Alternated deposition of polyanions and polycations on a charged solid substrate leads to the buildup of polyelectrolyte multilayer (PEM) films. Two types of PEM films were reported in the literature: films whose thickness increases linearly and films whose thickness increases exponentially with the number of deposition steps. However, it was recently found that, for exponentially growing films, the exponential increase of the film thickness takes place only during the initially deposited pairs of layers and is then followed by a linear increase. In this study, we investigate the growth process of hyaluronic acid/poly(L-lysine) (HA/PLL) and poly(L-glutamic acid)/poly(allylamine) (PGA/PAH) films, two films whose growth is initially exponential, when the growth process enters the linear regime. We focus, in particular, on the influence of the molecular weight (Mw) of the polyelectrolytes. For both systems, we find that the film thickness increment per polyanion/polycation deposition step in the linear growth regime is fairly independent of the molecular weights of the polyelectrolytes. We also find that when the (HA/PLL)n films are constructed with low molecular weight PLL, these chains can diffuse into the entire film during each buildup cycle, even for very thick films, whereas the PLL diffusion of high molecular weight chains is restricted to the upper part of the film. Our results lead to refinement of the buildup mechanism model, introduced previously for the exponentially growing films, which is based on the existence of three zones over the entire film thickness. The mechanism no longer needs all the "in" and "out" diffusing polyanions or polycations to be involved in the buildup process to explain the linear growth regime but merely relies on the interaction between the polyelectrolytes with an upper zone of the film. This zone is constituted of polyanion/polycation complexes which are "loosely bound" and rich in the polyelectrolyte deposited during the former deposition step.     

How electrolyte and polyelectrolyte affect the adsorption of the anionic surfactant SDS onto the surface of a cellulose thin film and the structure of the cellulose film. 1. Hydrophobic cellulose

The nature of hydrophobic thin cellulose films, formed by Langmuir-Blodgett (LB) deposition on silica, has been studied using neutron reflectivity (NR). The impact of electrolyte and a polyelectrolyte, poly(dimethyldiallylammonium chloride) (polydmdaac), on the adsorption of the anionic surfactant sodium dodecyl sulfate (SDS) onto the surface of the hydrophobic cellulose film and upon the structure of the cellulose film has been investigated. The results show how a combination of polyelectrolytes and electrolyte can be used to manipulate surfactant adsorption onto hydrophobic cellulose surfaces and modify the structure of the cellulose film by swelling and penetration. The results illustrate how polyelectrolytes can be used to reverse adsorption and swelling of cellulose films which are not reversible simply by dilution in solvent.     

Determination of electrokinetic charge with a particle-charge detector, and its relationship to the total charge

The main features of polyelectrolyte titrations with end-point indication by means of a particle-charge detector (PCD) were investigated. Because of their well-defined character, the charges of different synthetic polyelectrolytes and of latex samples with different functional groups were measured at different pH and ionic strength. The results show that PCD is a valuable tool for detecting effective or dissociated counterion charge without additional model assumptions. For negatively charged samples with exclusively strong acid functional groups, an excellent agreement was obtained between cation-exchange capacity and the charge measured by PCD over a wide pH range. For samples with additional carboxyl groups, the PCD charge was significantly lower than the total charge calculated from cation-exchange results. It can be concluded that counterion immobilization by a Stern layer-type arrangement is responsible for this effect.     

Increased layer interdiffusion in polyelectrolyte films upon annealing in water and aqueous salt solutions

As-deposited films of multilayered polyelectrolytes are considered to be non-equilibrium structures. Due to the strong attraction between oppositely charged polyions, polyelectrolyte interdiffusion is thought to be suppressed during the adsorption process. Equilibration is promoted by a decrease of the electrostatic attraction between polyion pairs. We have used neutral impact collision ion scattering spectroscopy to investigate the influence of polyelectrolyte multilayer annealing in water and aqueous 1 M NaCl solutions at different temperatures (20 and 70 degrees C) on the increase in interpenetration of a single polyelectrolyte layer throughout the whole film. The multilayers were composed of poly(4-vinylpyridinium) and poly(4-styrenesulfonate). Contrast between neighboring layers was established by labelling the layer in question with the heavy atom ruthenium. It is found that both temperature and salt increase layer interpenetration, whereas salt has a stronger influence than temperature. From numerical simulations polyelectrolyte diffusion coefficients were evaluated for the different annealing conditions. The influence of temperature and salt on the equilibration of the film is interpreted in terms of increased screening of polyion charges and binding of small counterions to polyion monomeric units.     

Effect of microfibrillated cellulose and fines on the drainage of kraft pulp suspension and paper strength

Different types of microfibrillated cellulose (MFC) and fines suspensions were produced, characterized, and then added to a papermaking pulp suspension. High and medium molar mass cationic polyelectrolytes were used as fixatives. The drainage behavior of the pulp suspensions with additives were evaluated against the strength properties of hand sheets made thereof. The effects of salt concentration, pH, fixative type, dosage and type of fibrillar material on drainage were examined. All the MFC and fines samples produced had clearly different properties due to their dissimilar production methods, and they also introduced specific responses on the measured drainage and paper strength. Generally, the addition of MFC decreased the drainage rate of pulp suspension and increased the strength of paper. However, it was shown that by optimum selection of materials and process conditions an enhancement of the strength properties could be achieved without simultaneously deteriorating the drainage.