Stability and interface properties of thin cellulose ester films adsorbed from acetone and ethyl acetate solutions

Journal of Colloid and Interface Science 332 (2009) 477–483

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Stability and interface properties of thin cellulose ester films adsorbed from acetone and ethyl acetate solutions Jorge Amim Jr. a , Priscila M. Kosaka a , Denise F.S. Petri a,∗ , Francisco C.B. Maia b , Paulo B. Miranda b a b

Instituto de Quimica, Universidade de São Paulo, Av. Prof. Lineu Prestes 748, 05508-900 São Paulo, Brazil Instituto de Física, Universidade de São Paulo, São Carlos, Brazil

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 6 November 2008 Accepted 22 December 2008 Available online 25 December 2008

Stability and interface properties of cellulose acetate propionate (CAP) and cellulose acetate butyrate (CAB) films adsorbed from acetone or ethyl acetate onto Si wafers have been investigated by means of contact angle measurements and atomic force microscopy (AFM). Surface energy (γ Stotal ) values determined for CAP adsorbed from acetone are larger than those from ethyl acetate. In the case of CAB films adsorbed from ethyl acetate and acetone were similar. Dewetting was observed by AFM only for CAP films prepared from ethyl acetate. Positive values of effective Hamaker constant ( A eff ) were found only for CAP prepared from ethyl acetate, corroborating with dewetting phenomena observed by AFM. On the contrary, negative values of A eff were determined for CAP and CAB prepared from acetone and for CAB prepared from ethyl acetate, corroborating with experimental observations. Sum frequency generation (SFG) vibrational spectra indicated that CAP and CAB films prepared from ethyl acetate present more alkyl groups oriented perpendicularly to the polymer–air interface than those films prepared from acetone. Such preferential orientation corroborates with macroscopic contact angle measurements. Moreover, SFG spectra showed that acetone binds strongly to Si wafers, creating a new surface for CAP and CAB films. © 2008 Elsevier Inc. All rights reserved.

Keywords: Cellulose ester Films Dewetting Atomic force microscopy Sum frequency generation

1. Introduction Thin polymeric films are largely used in many technological applications, as for instance, as coatings, paint, and sensors. In such situations, stable, continuous and homogeneous layers are required. Polymer surface energy and interface tension between substrate and polymeric film drive the stability of thin polymeric films. If the work of cohesion is larger than the work of adhesion, thin polymer film becomes unstable, resulting in the dewetting phenomena [1,2]. However, the solvents used for film preparation might also affect structure, wettability and stability of thin polymeric films. Special attention has been devoted to the effect of solvent used for spin-coating polymer films on the resulting surface properties [3–9]. Not only the evaporation rate of the solvent plays a very import role on the spin-coated [3–9] or cast film [10] uniformity, but also solvent quality and the balance between substrate–polymer interaction and substrate–solvent interaction [7–9,11]. Nevertheless, there are scarce studies on the effect of solvent on the stability and surface properties of ultrathin (d < 5 nm) adsorbed polymer films.

*

Corresponding author. Fax: +55 11 3815 5579. E-mail address: [email protected] (D.F.S. Petri).

0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.12.057

©

2008 Elsevier Inc. All rights reserved.

Cellulose esters are frequently applied as binders, additives, film formers or modifiers in automotive, wood, plastic and leather coatings applications [12]. Cellulose acetate propionate (CAP) and cellulose acetate butyrate (CAB) find place in the coating industry because they reduce dry time, improve flow and leveling, control viscosity and gloss, are stable carriers for metallic pigments, improve UV stability and reduce plasticizer migration, among many other benefits [12]. Moreover, cellulose esters films also proved to be efficient supports for adsorption of proteins [13,14] and cell growth [15]. In this work, surface properties and stability of CAP and CAB ultrathin films adsorbed from acetone or ethyl acetate solutions onto Si wafers, when they are heated above their glass transition temperature, has been investigated. Acetone and ethyl acetate are good solvents for CAP and CAB [9]. Surface properties have been determined for adsorbed CAP and CAB films by means of contact angle measurements. Film stability was predicted with basis on thermodynamic parameters and compared with atomic force microscopy (AFM) analyses. Sum frequency generation (SFG) vibrational spectroscopy, a sensitive tool for interfacial analyses, was used to gain insight about molecular orientation at polymer/substrate and solvent/substrate interfaces.

478

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

Table 1 ¯ w ), glass transition Cellulose esters characteristics: weight-average molar mass ( M temperature (T g ), melting point (T m ), degree of substitution for acetate (DSac), propionate (DSpr) and butyrate (DSbu), refractive index (n) for wavelength of 632.8 nm. The degree of esterification is the ratio of ester groups to glucose residues. Data provided by the producer. Sample

¯w M (g mol−1 )

Tg (◦ C)

Tm (◦ C)

DSac

DSpr

DSbu

n632.8

CAP CAB

25,000 30,000

142 101

184 135

0.2 0.2

2.3 –

– 2.5

1.475 1.475

Test liquids

γLV (mJ/m2 )

Water Formamide Diiodomethane

72.8 58.0 50.8

d γLV

γLVp

21.8 39.0 50.8

51.0 19.0 0

(mJ/m2 )

(mJ/m2 )

(2.550 ± 0.2) nm. After that, the mean thickness of adsorbed cellulose ester layers was determined in air by means of ellipsometry, considering the nominal refractive index indicated in Table 1.

2. Experimental 2.1. Materials Silicon wafers (100) purchased from Silicon Quest (CA, USA) with native oxide layer approximately 2 nm thick were used as substrates. They were cut in small pieces of (1.0 × 1.0) cm2 , rinsed in a standard manner [16], dried under a stream of N2 and characterized prior to use. Cellulose acetate propionate and cellulose acetate butyrate (powders free of plasticizer) were kindly supplied by Eastman Chemical Co., Brazil. Analytical grade acetone and ethyl acetate were used to prepare the solutions at the polymer concentration of 5 mg/mL. The samples characteristics are shown in Table 1. The degree of crystallinity of cellulose acetate propionate (CAP) and cellulose acetate butyrate (CAB) powders were calculated using X-ray diffractograms as reported elsewhere [17], they amounted to 42 and 38%, respectively. 2.2. Methods 2.2.1. Preparation of cellulose ester films Silicon wafers were dipped into homogeneous solutions of CAP and CAB prepared in acetone or ethyl acetate at 5 mg mL−1 at (24 ± 1) ◦ C. After 21 h they were removed from the solutions, washed with pure acetone or ethyl acetate and dried under a stream of N2 . The samples were annealed at 170 ◦ C under reduced pressure (60 mm Hg) during 4, 15, 60 and 168 h. 2.2.2. Ellipsometry Ellipsometric measurements were performed in the air using a vertical computer-controlled DRE-EL02 ellipsometer (Ratzeburg, Germany). The angle of incidence was set at 70.0◦ and the wavelength, λ, of the He–Ne laser was 632.8 nm. For the data interpretation, a multilayer model composed by the substrate, the unknown layer and the surrounding medium should be used. Then the thickness (dx ) and refractive index (nx ) of the unknown layer can be calculated from the ellipsometric angles,  and Ψ , using the fundamental ellipsometric equation and iterative calculations with Jones matrices [18]: e i  tan Ψ = R p / R s = f (nx , dx , λ, φ),

Table 2 Surface tension values reported in literature [22] for the different test liquids, which were used for the determination of the surface energy of cellulose esters.

(1)

where R p and R s are the overall reflection coefficients for the parallel and perpendicular waves. They are a function of the angle of incidence φ , the wavelength λ of the radiation and of the refractive index and the thickness of each layer of the model, nx , dx . From the ellipsometric angles,  and Ψ , and a multilayer model composed by silicon, silicon dioxide, polysaccharide layer and air it is possible to determine only the thickness of the polysaccharide layer, dpoly . First of all, the thickness of the SiO2 layers was determined in air, considering the refractive index for Si as n˜ = 3.8 − i0.018 [19] and its thickness as an infinite one, for the surrounding medium (air) the refractive index was considered as 1.00. Because the native SiO2 layer is very thin, its refractive index was set as 1.462 [19] and just the thickness was calculated. The mean thickness of the native SiO2 layer was amounted to

2.2.3. Contact angle measurements Contact angle measurements were performed at (24 ± 1) ◦ C in a home-built apparatus [20]. In order to investigate the effect of solvent used on the surface energy of adsorbed film, measurements were performed for CAP or CAB films adsorbed from ethyl acetate or acetone solutions (5 mg/mL). The surface energy of the polymers was determined indirectly using contact angle measurements performed either with diiodomethane >99.5% (purely dispersive nature) and water as described elsewhere [20–23]. The surface energy parameters for the different test liquids are presented in Table 2. Sessile drops of 8 μL were used for the θDiiodomethane and θWater . At least three samples of the same composition were analyzed for each test liquid. 2.2.4. Atomic force microscopy (AFM) Atomic force microscopy (AFM) measurements were performed in a PICO SPM-LE (Molecular Imaging) microscope in the intermittent contact in air at room temperature, using silicon cantilevers with resonance frequency close to 300 kHz. Scan areas varying from (5 × 5) μm2 to (1 × 1) μm2 with a resolution of 512 × 512 pixels were obtained. Image processing and the determination of the root mean square (rms) were performed using the Pico Scan software. Mean roughness value (rms) is calculated as the root mean square value of all peak-to-valley distances measured over the scanned area after surface flattening, where the software subtracts linear fit from data. At least two films of the same composition were analyzed at different areas of the surface before and after annealing. 2.2.5. Sum frequency generation (SFG) vibrational spectra Sum frequency generation (SFG) [24] vibrational spectra were obtained with a commercial SFG spectrometer (Ekspla, Lithuania) equipped with a pulsed Nd3+ :YAG laser at 1064 nm (28 ps pulse duration, 20 Hz repetition rate) and harmonic unit generating second and third harmonics (532 and 355 nm, respectively). The first is the visible beam that excites the sample (pulse energy ∼950 μJ). The third harmonic and fundamental beams pump an optical parametric amplifier with a difference-frequency stage that generates an infrared (IR) beam tunable from 1000 to 4000 cm−1 (pulse energy ∼30–150 μJ), which overlaps the visible beam on the sample. The incidence angles and approximate spot sizes on the sample are 51◦ , 500 μm and 60◦ , 1000 μm for the IR and visible beams, respectively. The sum-frequency signal as a function of IR frequency is collected by a photomultiplier after spatial and spectral filtering. For each scan, data are collected with 100 shots/data point in 3 cm−1 increments typically during 8 min. SFG spectra were obtained with ssp polarization combination, where s and p stand for polarized light perpendicular and parallel to the surface, respectively, and the first s, second s, and last p represent the polarizations of SFG, visible, and infrared light, respectively. CAP and CAB films adsorbed onto quartz (SiO2 ) slides either from acetone or ethyl acetate solutions at 5 mg/mL and (24 ± 1) ◦ C. After 21 h slides were removed from the solutions, washed with

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

479

Fig. 1. AFM images (1 × 1) μm2 obtained in air for CAP and CAB adsorbed from acetone or ethyl acetate solutions (5 mg/mL) onto Si wafers: (a) CAP (acetone) without thermal treatment ( Z = 3 nm), (b) CAP (acetone) after 15 h of annealing ( Z = 3.5 nm), (c) CAP (acetone) after 60 h of annealing ( Z = 3 nm), (d) CAP (acetone) after 168 h of annealing ( Z = 2 nm), (e) CAP (ethyl acetate) without thermal treatment ( Z = 2 nm), (f) CAP (ethyl acetate) after 15 h of annealing ( Z = 9.5 nm), (g) CAP (ethyl acetate) after 60 h of annealing ( Z = 11 nm), (h) CAP (acetone) after 168 h of annealing ( Z = 11 nm), (i) CAB (acetone) without thermal treatment ( Z = 2 nm), (j) CAB (acetone) after 15 h of annealing ( Z = 3 nm), (k) CAB (acetone) after 60 h of annealing ( Z = 2 nm), (l) CAB (acetone) after 168 h of annealing ( Z = 2 nm), (m) CAB (ethyl acetate) without thermal treatment ( Z = 4 nm), (n) CAB (ethyl acetate) after 15 h of annealing ( Z = 5 nm), (o) CAB (ethyl acetate) after 60 h of annealing ( Z = 2 nm), (p) CAB (ethyl acetate) after 168 h of annealing ( Z = 2 nm).

pure acetone or ethyl acetate and dried under a stream of N2 . After that spectra were obtained in air. For in situ measurements at (24 ± 1) ◦ C, quartz slide was arranged inside the fluid cell, which was filled up either with acetone or ethyl acetate. 3. Results and discussion The mean thickness values determined in air for CAP and CAB films obtained from acetone amounted to (2.150 ± 0.9) nm and (1.950 ± 0.3) nm, respectively, and from ethyl acetate they were (1.350 ± 0.3) nm and (0.950 ± 0.2) nm, respectively. Considering that a cellulose monolayer is ∼0.7 nm thick [25]. Adsorbed layers stemming from acetone solution were thicker than those from ethyl acetate. The same tendency was observed for spin-coated CAP and CAB films when acetone or ethyl acetate was used as solvents [9]. Two factors might explain this behavior: (i) acetone is more volatile than ethyl acetate, leading to thicker films [3–9] and (ii) the more favorable acid–base interaction for SiO2 /acetone

(+18RT) than for SiO2 /ethyl acetate (+24RT) [26], so that acetone evaporation is not complete, resulting in thicker polymer films. Figs. 1a to 1d show AFM topographic images obtained for CAP films adsorbed onto silicon wafers from acetone just after preparation (Fig. 1a) and after annealing (Figs. 1b–1d). These films are very stable even after 168 h annealing (Fig. 1d); no significant changes in mean thickness or roughness values could be observed. CAP films adsorbed from ethyl acetate solutions were very smooth (rms = 0.13 nm, Fig. 1e). However, upon annealing dewetting structural features appeared on the surface. After 15 h of annealing (Fig. 1f) rims could already be observed and rms values increased to 0.31 nm. Upon further annealing rims appeared more frequently and rms values increased (Figs. 1g and 1h). CAB films adsorbed from acetone or ethyl acetate solutions were very smooth, as shown in Figs. 1i and 1m, respectively. Regardless the solvent used for adsorption of CAB onto Si wafers, films were very smooth and stable, as evidenced by AFM topographic images (Figs. 1j to 1p), even after one week annealing.

480

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

Table 3 Advancing contact angle measurements for CAP and CAB adsorbed from acetone or ethyl acetate solutions (5 mg/mL) onto Si wafers using drops of diiodomethane (θDiiodomethane ) and water (θWater ) as test liquids. Dispersive (γ d ) and polar (γ p ) components values of the surface tension determined by harmonic model (hm, Eq. (4)) and geometric model (gm, Eq. (5)). Effective Hamaker constant A eff is the difference between A poly/poly and A poly/substrate , which was calculated as 6.5 × 10−20 J (see text for details). Sample

γd

γp

γ total

(mJ/m2 )

(mJ/m2 )

(mJ/m2 )

A poly/poly × 10−20 (J)

A eff × 10−20 (J)

30 ± 1

34.9 ± 0.3 (hm) 33.2 ± 0.3 (gm)

34.350 ± 0.3 (hm) 33.050 ± 0.3 (gm)

69.250 ± 0.4 (hm) 66.250 ± 0.3 (hm)

6.7 ± 0.1 (hm) 6.4 ± 0.1 (gm)

+0.250 ± 0.1 (hm) −0.1 ± 0.1 (gm)

45 ± 2

48 ± 1

37.8 ± 0.4 (hm) 37.0 ± 0.4 (gm)

24.550 ± 0.2 (hm) 20.5 ± 0.2 (gm)

62.3 ± 0.4 (hm) 57.5 ± 0.4 (gm)

7.350 ± 0.1 (hm) 7.150 ± 0.1 (gm)

+0.850 ± 0.1 (hm) +0.650 ± 0.1 (gm)

56 ±2

42 ±2

θDiiodomethane

θWater

CAP (Acetone)

52 ± 1

CAP (Ethyl acetate) CAB

(◦ )

(◦ )

(Acetone) 51 ±1

CAB (Ethyl acetate)

50 ±1

32.450 ± 0.3 (hm)

29.550 ± 0.3 (hm)

61.950 ± 0.4 (hm)

6.350 ± 0.1 (hm)

−0.250 ± 0.1 (hm)

30.950 ± 0.3 (gm)

27.650 ± 0.3 (gm)

58.550 ± 0.4 (gm)

6.050 ± 0.1 (gm)

−0.550 ± 0.1 (gm)

34.950 ± 0.3 (hm) 33.750 ± 0.3 (gm)

24.350 ± 0.2 (hm) 21.050 ± 0.2 (gm)

59.250 ± 0.4 (hm) 54.750 ± 0.4 (gm)

6.750 ± 0.1 (hm) 6.550 ± 0.1 (gm)

+0.250 ± 0.1 (hm) 0 (gm)

p

The stability of polymeric films depends on polymer and substrate surface energies. The surface tension γ of a material can p be considered as the sum of the dispersive (γ Sd ) and polar (γ S ) d components [27]. From γ S it is possible to calculate the Hamaker constant for the polymer/polymer attraction ( A poly/poly ) [28]:

polar components of γLV , respectively, and γ S is the polar component of solid surface tension. Combining the work of adhesion at solid–liquid interface [20, 38] with Young’s equation one obtains the geometric mean equation [39,40], which also allows the estimation of γ S components:

γ Sd 24π d2 = A poly/poly ,

d d γLV (1 + cos θ) = 2 γLV γS

(2)

where d (d ∼ 1.6 Å) is the distance between two atoms [21,29]. The difference between Hamaker constant values for the polymer/polymer attraction ( A poly/poly ) and polymer/substrate interaction ( A poly/substrate ) yields the effective Hamaker constant ( A eff ). If A eff is positive, fluctuations in film thickness are expected, and after a characteristic time, the film will rupture [2,30]. Negative values for A eff indicate long range apolar van der Waals repulsion, which promotes film stability and wetting [31]. According to Lifshitz–van der Waals theory, Hamaker constant for a material 1 interacting with material 2 across medium 3 ( A poly/substrate ) can be calculated by [32–34]: A poly/substrate 3



= kT 4

ε1 − ε3 ε1 + ε3





ε2 − ε3 ε2 + ε3



 (n21 − n23 )(n22 − n23 )    + √  , 8 2 n21 + n23 n22 + n23 ( n21 + n23 + n22 + n23 ) 3hν

(3)

where k is the Boltzmann constant, T is temperature, ε is the dielectric permittivity, n is the refractive index, h is Planck constant, ν the mean ionization frequency of the material (typically this is ν ∼ 3 × 1015 Hz) [21,35]. The first term represents the Keesom plus Debey energy and plays an important role for forces in water, since water molecules have a strong dipole moment. However, generally, the second term dominates. In this work, T was set to 298.15 K and media 1, 2 and 3 corresponded to cellulose ester (ε1 = 6.00 and n1 = 1.48) [36], SiO2 (upper native layer of Si wafer) (ε2 = 3.82 and n2 = 1.46) [21], and air (ε3 = 1.005 and n3 = 1.00), respectively. One should notice that ε value assigned to cellulose ester is that available in the literature [36] for cellulose. Substituting all these numerical values in Eq. (3), A poly/substrate was calculated as 6.5 × 10−20 J. A poly/poly also can be calculated from Eq. (2), if γ Sd value is known. Equation (4) is a combination of Young’s equation [37] and harmonic mean equation and enables one to calculate γ Sd value [38,39]:



γLV (1 + cos θ) = 4



d d γLV γS γLVp γ Sp , + d γLVp + γ Sp γLV + γ Sd

(4)

where γLV is surface tension of a liquid, which was used for the p d and γLV are the dispersive and contact angle (θ ) measurement, γLV



1/2

 p p 1/2

+ γLV γ S .

(5)

Therefore, if contact angles measurements are performed with p d and γLV onto polydrops of at least two liquids of known γLV

mer surface, it is possible to calculate the dispersive γ Sd and polar γ Sp components of polymer surface. Such equations are defined for smooth, rigid, chemically homogeneous, insoluble and non-reactive polymer surfaces. Figs. 1a, 1e, 1i and 1m evidenced smooth CAP and CAB films adsorbed onto Si wafers either form acetone or ethyl acetate solutions, thus contact angle θ measurements were performed with drops of water and diiodomethane, as presented in Table 3. Comparing θ values obtained for water (highest polarity) drops on CAP films adsorbed from acetone (30◦ ) and ethyl acetate (48◦ ), one notices a difference of 18◦ , which indicates that in the former surfaces are more hydrophilic than in the latter. The same behavior was also observed for CAB films, with difference of 8◦ . θ values obtained for diiodomethane (apolar) drops on CAP or CAB films were smaller when the solvent used was ethyl acetate, indicating that ethyl acetate induced the formation of more hydrophobic films. d Substituting experimental θDiiodomethane values, γLV and γLV for diiodomethane in Eqs. (4) and (5), which correspond to harmonic and geometric models, respectively, the dispersive γ Sd component of CAP and CAB surface could be calculated, as shown in Table 3. Substituting γ Sd values in Eq. (2), A poly/poly were calculated (Table 3). A poly/poly are larger for CAP and CAB films adsorbed from ethyl acetate than from acetone. A eff values were also calculated, as shown in Table 3. Regardless the model used for calculation of dispersive γ Sd component, positive A eff value indicative of film instability was observed for CAP films adsorbed from ethyl acetate. In fact, AFM images in Fig. 1 evidenced dewetting only for annealed CAP films adsorbed from ethyl acetate. These findings are very important because they reveal that thin films stability depends not only on the polymer surface and substrate properties but also on the solvent used for film preparation. Moreover, γ Sd and A poly/poly values calculated with harmonic model were larger than those determined with geometric model, indicating that prediction of film stability might be model dependent. Regardless the model used for calculation of surface energy γ Stotal values determined for CAP adsorbed from acetone are larger p than those from ethyl acetate, mainly because γ S contribution for films from acetone is much larger than that for films from ethyl acetate. The same tendency was observed for CAB films. How-

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

481

Fig. 2. SFG spectra obtained in air for dried CAP films adsorbed from (a) acetone and (b) ethyl acetate solutions.

Fig. 3. SFG spectra obtained in air for dried CAB films adsorbed from (a) acetone and (b) ethyl acetate solutions onto quartz slides.

ever, γ Stotal values determined for CAB films are smaller than those found for CAP films adsorbed from the same solvent. A similar behavior was recently reported for a series of cellulose esters [23]. The main reason for this effect is surface energy decrease with the increase of alkyl ester group size, which weakens van der Waals interactions. SFG is a second-order nonlinear optical process, where electricdipole in a bulk with inversion symmetry is forbidden, but it is allowed at interfaces because inversion symmetry is broken [24, 41–43]. Therefore, it is sensitive to the interfacial structure between two centrosymmetric media. Moreover, input/output polarization dependence of the spectra provides information about average orientation of molecules at surfaces. SFG vibrational spectra were obtained in order to understand the role played by solvent in the polymer film molecular orientation and surface energy. Fig. 2 shows SFG spectra obtained in air for CAP adsorbed from acetone (Fig. 2a) and ethyl acetate (Fig. 2b). Although the intensities in the latter are stronger than in the former, both spectra presented one band at ∼2882 cm−1 and another one at 2950 cm−1 , which were assigned to the CH3 symmetric stretch (CH3(s) ) and to the Fermi resonance between the symmetric-stretch fundamental and an overtone of a bending mode of CH3 (CH3(F ) ), respectively. Such characteristics bands were also observed for other polymeric interfaces [44–48] and systems formed by small molecules [49–53], where methyl groups are oriented perpendicularly to the surface. The larger intensities observed for CAP adsorbed from ethyl acetate solutions (Fig. 2b) than those for CAP films formed from

acetone solution (Fig. 2a) indicated that in average a larger number of methyl groups were oriented perpendicularly to the surface when ethyl acetate was used as solvent. Such preferential orientation corroborates with macroscopic contact angle measurements, which showed that CAP films are more hydrophobic when adsorbing from ethyl acetate (Table 3). SFG spectra obtained in air for CAB adsorbed from acetone and ethyl acetate, Figs. 3a and 3b, respectively, were less intensive than those obtained for CAP. Fig. 3a shows weak intensity bands with the following assignments: 2878 cm−1 CH3(s) and 2945 cm−1 CH3(F ) . CAB films prepared from ethyl acetate (Fig. 3b) showed two intensive bands at 2879 and 2942 cm−1 , which were assigned to CH3(s) and CH3(F ) , respectively. Similarly to CAP films, the spectra obtained for CAB indicated that ethyl acetate induces more efficiently the orientation of methyl groups perpendicularly to the surface than acetone, explaining the more hydrophobic nature of CAB films adsorbed from ethyl acetate (Table 3). Interestingly, spectra obtained for CAP and CAB showed no bands in the carbonyl typical spectral range from 1600 to 1800 cm−1 (not shown), indicating random orientation of carbonyl groups or absence of carbonyl group oriented perpendicularly to the surface. In a three component system, where solvent and polymer might compete for the substrate, the characteristics of resulting adsorbed layer depends on the affinity between polymer and solvent, polymer and substrate and solvent and substrate. In order to gain insight about the affinity between solvent and substrate SFG spectra were obtained for acetone and ethyl acetate onto quartz slides

482

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

contact angle measurements, providing information at molecular level and giving support for macroscopic data. Acknowledgments The authors acknowledge CNPq and FAPESP for financial support and Eastman do Brasil for supplying cellulose ester samples. Supplementary material Supplementary material for this article may be found on Science Direct, in the online version. Please visit DOI: 10.1016/j.jcis.2008.12.057.

References

Fig. 4. SFG spectra obtained in situ for acetone and ethyl acetate onto quartz slides.

(Fig. 4). In the case of acetone, an intense band at ∼2923 cm−1 was assigned to the symmetric stretch vibration mode of methyl group of acetone (CHacetone ) [53]. In the case of ethyl acetate the 3(s) intensities were too weak to make any assignment. These findings indicate strong affinity between acetone and SiO2 surface and weak affinity between ethyl acetate and SiO2 surface and corroborate with literature [26] reports. The strong adsorption of acetone onto SiO2 creates in situ a new layer between CAP or CAB and SiO2 , which stabilizes the thin polymeric film avoiding dewetting phenomena. On the other hand, the adsorption of ethyl acetate onto SiO2 seems to be weak, and polymer layers face SiO2 directly. In that case, film stability is controlled by polymer/polymer attraction ( A poly/poly ) and polymer/substrate interaction ( A poly/substrate ). SFG spectra obtained in situ for acetone onto SiO2 (Fig. 4) evidenced preferential orientation of methyl groups perpendicular to the surface. After in situ measurements, SiO2 slides were dried under a stream of N2 and new SFG spectra were recorded (Supplementary material-1). A band at ∼2955 cm−1 , assigned to methyl groups, became weaker, indicating that even after drying acetone cannot be totally removed from SiO2 surface. Another indication that methyl groups are predominantly oriented perpendicularly to the surface was confirmed by contact angle measurements with water drops, which amounted to (40 ± 2)◦ . For comparison contact angle for water onto clean SiO2 is typically 5◦ [20]. Therefore, these findings prove that acetone molecules bind strongly to SiO2 surfaces. 4. Conclusions Many technological processes apply polymer solutions for the production of polymer films. The present work has shown that the solvent used for polymer films formation might play a crucial role on interface properties and polymer film stability. Acetone binds strongly to SiO2 surfaces due to acid–base interactions, creating a new surface for CAP or CAB, which avoids dewetting. The ability of this new surface to induce alkyl groups’ orientation to the polymer–air interface is limited, leading to more hydrophilic surfaces. Ethyl acetate was able to induce perpendicular orientation of CAP or CAB alkyl groups to the polymer–air interface. This preferential orientation led to hydrophobic surfaces, but film stability was driven by polymer surface energy and interface tension between substrate and polymeric film. From the experimental point of view, SFG proved to be an excellent complement for AFM and

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]

G. Reiter, Phys. Rev. Lett. 68 (1992) 75. G. Reiter, Langmuir 9 (1993) 1344. C.B. Walsh, E.I. Franses, Thin Solid Films 429 (2003) 71. K.E. Strawhecker, S.K. Kumar, J.F. Douglas, A. Karim, Macromolecules 34 (2001) 4669. P. Müller-Buschbaum, J.S. Gutmann, M. Wolkenhauer, J. Kraus, M. Stamm, D. Smilgies, W. Petry, Macromolecules 34 (2001) 1369. L. Cui, Y. Ding, X. Li, Z. Wang, Y. Han, Thin Solid Films 515 (2006) 2038. Y.-C. Lin, M. Müler, K. Binder, J. Chem. Phys. 121 (2004) 3816. D.F.S. Petri, J. Braz. Chem. Soc. 13 (2002) 695. J. Amim Jr., P.M. Kosaka, D.F.S. Petri, Cellulose 15 (2008) 527. X. Wang, H. Ni, D. Xue, X. Wang, R. Feng, H. Wang, J. Colloid Interface Sci. 321 (2008) 373. K.P. Cheung, R. Grover, Y. Wang, C. Gurkovich, G. Wang, J. Scheinbeim, Appl. Phys. Lett. 87 (2005) 214103. K.J. Edgar, C.M. Buchanam, J.S. Debenham, P.A. Rundquist, B.D. Seiler, M.C. Shelton, D. Tindall, Prog. Polym. Sci. 26 (2001) 1605. P.M. Kosaka, M.C. Salvadori, Y. Kawano, D.F.S. Petri, Cellulose 12 (2005) 351. P.M. Kosaka, Y. Kawano, O.A. El Seoud, D.F.S. Petri, Langmuir 23 (2007) 12167. G.S. Gomes, A.T. Almeida, P.M. Kosaka, S.O. Rogero, A.S. Cruz, T.I. Ikeda, D.F.S. Petri, Mater. Res. 10 (2007) 469. D.F.S. Petri, S.W. Choi, G. Wenz, M. Bruns, H. Beyer, T. Schimmel, Polymer 40 (1999) 1593. P.M. Kosaka, Y. Kawano, H.M. Petri, M.C. Fantini, D.F.S. Petri, J. Appl. Polym. Sci. 103 (2007) 402. R.M.A. Azzam, N.M. Bashara, Ellipsometry and Polarized Light, North Holland, Amsterdam, 1987. E.D. Palik, Handbook of Optical Constants of Solids, Academic Press, London, 1985. W.A. Adamson, Physical Chemistry of Surfaces, fifth ed., Wiley, Toronto, 1990. H.-J. Butt, K. Graf, M. Kappl, Physics and Chemistry of Interfaces, Wiley–VCH, Weinheim, 2003. B. Bouali, F. Ganachaud, J.-P. Chapel, C. Pichot, P. Lanteri, J. Colloid Interface Sci. 208 (1998) 81. P.M. Kosaka, Y. Kawano, D.F.S. Petri, J. Colloid Interface Sci. 316 (2007) 671. A.G. Lambert, P.B. Davies, D.J. Nelvandt, Appl. Spectrosc. Rev. 40 (2005) 103. G. Wiegang, T. Jaworek, G. Wegner, E. Sackmann, Langmuir 13 (1997) 3563. M.K. Chaudhury, Mater. Sci. Eng. R 16 (1996) 97. F.M. Fowkes, Ind. Eng. Chem. 56 (1964) 40. J. Mahanty, B.W. Ninham, Dispersion Forces, Academic Press, New York, 1976. M. Eriksson, S.M. Notley, L. Wågberg, Biomacromolecules 8 (2007) 912. G. Reiter, Phys. Rev. Lett. 87 (2001) 166103. A. Sharma, R. Khanna, Phys. Rev. Lett. 81 (1998) 3463. E.M. Lifshitz, Sov. Phys. JETP (Engl. Transl.) 2 (1956) 73. I.E. Dzyaloshinskii, E.M. Lifshitz, L.P. Pitaevskii, Adv. Phys. 10 (1961) 165. J.N. Israelachvili, Quart. Ver. Biophys. 6 (1974) 341. F. Brochard-Wyart, C. Redon, C. Sykes, C. R. Acad. Sci. Paris, Ser. II 13 (1992) 314. M. Holmberg, J. Berg, S. Stemme, L. Ödberg, J. Rasmusson, P. Claesson, J. Colloid Interface Sci. 186 (1997) 369. T. Young, Philos. Trans. R. Soc. London 95 (1805) 65. S. Wu, Polymer Interface and Adhesion, Dekker, New York, 1982. R.J. Good, L.A. Girifalco, J. Phys. Chem. 64 (1960) 561. D.K. Owens, R.D. Wendt, J. Appl. Polym. Sci. 13 (1969) 1741. D. Zhang, Y.R. Shen, G.A. Somorjai, Chem. Phys. Lett. 281 (1997) 394. D.H. Gracias, D. Zhang, L. Lianos, W. Ibach, Y.R. Hen, G.A. Somorjai, Chem. Phys. 245 (1999) 277. X. Wei, X. Zhuang, S. Hong, T. Goto, Y.R. Shen, Phys. Rev. Lett. 82 (1999) 4256. Q. Shi, S. Ye, C. Kristalyn, Y. Su, Z. Jiang, Z. Chen, Langmuir 24 (2008) 7939.

J. Amim Jr. et al. / Journal of Colloid and Interface Science 332 (2009) 477–483

[45] A. Opdahl, T.S. Koffas, E. Amitay-Sadovsky, J. Kim, G.A. Somerjai, J. Phys. Condens. Matter 16 (2004) R659. [46] C. Chen, J. Wang, S.E. Woodcock, Z. Chen, Langmuir 18 (2002) 1302. [47] H. Rangwalla, A.D. Schwab, B. Yurdumakan, D.G. Yablon, M.S. Yeganeh, A. Dhinojwala, Langmuir 20 (2004) 8625. [48] J. Kim, T.S. Koffas, C.C. Lawrence, G.A. Somorjai, Langmuir 20 (2004) 4640. [49] R.A. MacPhail, H.L. Strauss, R.G. Snyder, C.A. Elliger, J. Phys. Chem. 88 (1984) 334.

483

[50] R. Lu, W. Gan, B.-h. Wu, Z. Zhang, Y. Guo, H.-f. Wang, J. Phys. Chem. B 109 (2005) 14118. [51] R. Lu, W. Gan, B.-h. Wu, H. Chen, H.-f. Wang, J. Phys. Chem. B 108 (2004) 7297. [52] Y. Yu, K. Lin, X. Zhoun, H. Wang, S. Liu, X. Ma, J. Phys. Chem. C 111 (2007) 8971. [53] H. Chen, W. Gan, B. Wu, D. Wu, Z. Zhang, H. Wang, Chem. Phys. Lett. 408 (2005) 284.