Langmuir 1993,9, 3567-3573
3667
Atomic Force Microscopy Characterization of Poly(amino acid) Langmuir-Blodgett Films N. Auduc, A. Ringenbach, I. Stevenson, Y. Jugnet, and Tran Minh Duc* Centre ESCA de Nanoanalyse et de Technologie de Surface (CENATS), Universitg Claude Bernard, 43, bld du 11 novembre 1918, F-69622 Villeurbanne Cedex, France Received April 1,1993. I n Final Form: September 23,199P Atomic force microscopy (AFM) has allowed us to obtain molecular resolved images of Langmub Blodgett (LB) monolayers of docosanoic acid (DA) deposited on a glass substrate, monolayers of poly(8-benzyl-L-aspartate) (PBA) and poly(i-alanine) (PAL) deposited on a mica substrate, and bilayers of PBA and PAL on graphite. Evidence of sample damage by the AF’Mtip has been shown. Studies in the periodicity and conformation of the molecules on each type of substrate (hydrophobicor hydrophilic) have thus become possible leading to the measurement of the interchain distance for PBA and PAL. have been used to determine the orientation of the chains Introduction and their head group attachment including infrared**Oand Ultrathin ordered organic films with a thickness of a RamanlOJl spectroscopies. LEED (low energy electron few nanometers to about a tenth of a micrometer such as diffraction)12and X-ray diffracti~n’~ techniques have been LB films show considerable technological potential as a successfully used to obtain structural information of LB novel class of materials. These films, which have been films. More useful but difficult to obtain are images. studied for two decades,’ find applications in the field of Because of the small size (typically 0.2-3 nm) of the coatinga,optics, electronics,biology, and polymerizati~n.~*~ molecular units making up the films, the only technique It is well-known that fatty acid molecules are ideally available was TEM (transmission electron microscopy).14 suited to form LB films’ which are two-dimensional More recently, scanning probe microscopy, mainly ordered arrays of amphiphilic molecules that can form at scanning tunnelingl”17 and atomic fordR1Sm microscopies an air-water interface and can be transferred to a variety (STM and AFM), appeared promising for the elucidation of solid surfaces. However, LB films of polymers are of the structure of LB films at the molecular level. These advantageous to use since they possess a better mechanical techniques have the advantage, for the study of fragile stability than fatty acids of low molecular weight due to organicthin films,of producing surface images with atomic the strong intra- and intermolecular interactions. A new resolution. Images of controlled modifications of polymers class of polymeric materials which undergo self-organi(LB monolayers of poly(octadecy1 acrylate) and polyzation to transferable monomolecular layers at the air(methylmethacrylate) on graphite) have been successfully water interface consist of rodlike molecules with covalently obtained with STM and AFMaZ1However, AFM appears attached flexible side chainse4Supramolecular structures to be of more general use since it no longer necessitates of these nonamphiphilic macromolecules have been deconducting tip and samples as STM does. In AF’M the signed and investigated for their optical and electronic probing tip is attached to a cantilever which is deflected proper tie^.^ Hairy rodlike polymers include poly(amino in response to the forces between the probing tip and the acid)swhich have an a-helical conformation. The polymer sample. The interatomic force F between the tip and the backbones tend to be more hydrophilic than the side chains sample’isproportional to kdz where dz is the variation of although some poly(amino acid)s possess hydrophilic z (height scale giving topography) and k the cantilever groups in their side chains (it is the case for PBA). It has force constant. The main problem in imaging soft organic been shown6 that the rigid rodlike nature of the a-helix or biological materials with AFM is the relatively high favors the horizontal alignment of the molecules at the force (which may be destructive) exerted by the cantilever surface of the water and that the side chains of adjacent molecules will interpenetrate and form mainly hydro(8)Akamatsu, N.; Domen, K.; Hirose, C.; Onishi, T.; Shimiizu, H.; phobic contacts. Another way to improve the stability Masutani, K. Chem. Phys. Lett. 1991,181 (2,3)175178. (9) Baler, R.E.;Zisman, W. A. Macromolecules 1970,3,7*79. and homogeneity of polymer LB films is to use spacers (IO)Swalen, J. D. Thin Solid Films 1987,152,151. such as hydrocarbon chains which give the side chains (11)Vandevyver,M.;Ruaudel-Teixier,A.;Brehamet,L.;Lutz,M.Thin better fle~ibility.~ Solid F i l m 1983,99,41-44. (12)Vogel, V.; Wtill, C. J. Chem. Phys. 1986,84 (9),5200-5204. In recent years, a number of techniques have been (13)Parry, D. A. D.; Elliott, A. J. Mol. Biol. 1967,25,1-13. applied to the problem of determining the structure, (14)Shen, Y.; Luk Andrew, C.; Wright, C.; Williams, 0. J. Thin Solid orientation, and packing of LB films. Common methods F i l m 1990,186,147-154.
(15)Binnig,G.;Quate,C. F.; Gerber,C. Phys. Rev. Lett. 1986,56,930.
* To whom correspondence should be addressed.
e Abstract published
in Advance ACS Abstracts, November 1,
1993. (1)Roberta, G.Langmuir-Blodgett films; Plenum Press, New York, 1990. (2)Barraud, A. Thin Solid F i l m 1983,99,317. (3)Barraud, A. J. Chim. Phys. 1986,82 (6), 683. (4)Schaub,M.; Mathauer, K.; Schwiegk, 5.; Albouy, P. A.; Wenz, G.; Wegner, G. Thin Solid F i l m 1992,210/211,347-400. (5)Schwiegk,S.;Vahlenkamp,T.;Wegner,G.; Xu,Y. Thin Solid F i l m 1992,210/211,6-8. (6)Malcolm, B. R. Bog. Surf. Membr. Sci. 1973,7, 183-229. (7)Bubeck, C.Polym. J. 1991,23 (5),603-611.
(16)Lang, C . A.; H6rber, J. K. H.; HHnech, T. W.; Heckl, W. M.;
Mdhwald, H. J. Vac. Sci. Technol., 1988,A6 (2),368-370. (17)Fuchs, M. Phys. Scr. 1988,38,264-268. (18)Meyer,E.;Howald,L.;Ovemey,R.M.;Hehlmann,H.;F”er, J.; Ghtherodt, H. J.; Wagner, T.; Schier, H.; Roth, S.Nature (London) 1991,349,398-400. (19)Bourdieu, L.;Silberzan, P.; Chatenay, D. Phys. Rev. Lett., 1991, 67 (16),2029-2032. (20)Weisenhom,A.L.;Egger,M.;Ohneeorge,F.;Gould,S.A.C.;Heyn, S. P.;Hansma, H. G.; Sinsheimer,R. L.; Gaub, H. E.; Hamma, P. K. Langmuir 1991,7 (I), 8-12. (21)Albrecht, T. R.;Dovek, M. M.; Lang, C. A,;Grtitter,P.; Quate,C. F.; Kuan, S.W. J.; Frank, C. W.;Pease, R. F. W. J.Appl. Phys. 1988,64 (3),1178-1184.
0743-7463/93/2409-3567~04.o0/00 1993 American Chemical Society
3568 Langmuir, Vol. 9, No. 12, 1993
tip onto the sample. Perssonn has shown with theoretical calculations that in order to avoid large substrate deformationswhen studyingsoft biologicalmaterials, one should operate with a forceF N. Unfortunately, the lowest forces used for imaging by AFM are still in the 10-8-W7 N range in air and le9N or less in Fatty acid monolayers, depending on the type of salt added, are oriented almost perpendicular to the surface or tilted with an angle of about 30" from normal to the surface.1° The AFM tip may scan the end methyl groups or methylene groups of the tilted alkyl chain. Poly(amino acid) monolayers lie onto the substrate parallel to the dipping directi0n.2~The PBA and PAL studied should have an a-helix conformation as long as the right solvent is u ~ e d . ~The S AFM tip may scan one side of these helices. Polypeptide deposited in drops onto graphite substrates (poly(y-benzyl-L-glutamate))gives structures which have Difbeen imaged by scanning tunneling micro~cope.~~ ferent conformations (a-helix, /3 sheet, random coil) have been observed as a function of the solvent used.28 In this paper we show that AFM allows us to describe, down to the molecular scale, LB monolayer of a fatty acid (docosanoic acid, C22H402) deposited on a glass microscope slide and monolayers and bilayers of polypeptides (of general formula -(NH-CHR-CO),- where R = CH3 for PAL and R = CH2COOCH2-CsHs for PBA) deposited on mica and graphite substrates, respectively.
Auduc et at.
-
Experimental Section The fatty acid (DA) was purchased from Aldrich Chemical Co. (purity >99%) and poly(amin0 acid)s used (PBA and PAL) were obtained from Sigma Chemical Co. Degrees of polymerization of PBA and PAL are 351 and 347, respectively. Both these poly(amino acid)s were dissolved in chloroform/dichloroacetic acid (99:l) and the fatty acid was dissolved in chloroform. They are spread onto the water surface of the LB trough (purified aqueous subphase, Milli Q water, resistivity 18 MQ/cm) via a microsyringe. Prior to use, mica and graphite substrates were cleaned by simply peeling off the top layers using adhesive tape and glass substrates were cleaned in a sulfochromic mixture. In the Langmuir-Blodgett experimental setup, these substrates are clipped on a dippingarm and lie perpendicular to the compression direction. The preparation of the LB films was performed using a KSV5000 Langmuir-Blodgett trough in a clean glovebox (under nitrogen flow and a t constant temperature) isolated from surrounding vibrations by an antivibration table. Isotherms (curves representing the variation of ?r (mN/m) the surface pressure as a function of MmA (A2)the mean molecular area) were recordedfor eachsampleto determine the best target surface pressure. Filmswere compressed to close packed two-dimensional solid conditions and transferred onto the different substrates (glass,mica, graphite). Acceptable transfer ratios were 1.0 f 0.1. All the LB experiments have been carried out at a temperature T = 24 f 1"C. A two degreevariation in the temperature changes considerablythe transfer ratio for a given target surface pressure. The average compression barrier speed used was 5 mm/mn. We have noticed that the dipping speed in the case of polypeptides must be greater than 2 mm/mn in order to obtain a good transfer. Maintaining a given target surface pressure, it was verified that variations of the barrier position as a function of time between
Figure 1. AFM imagesobtained for docosanoicacid (DA). Threedimensional molecular resolution AFM image of a DA LB monolayer (4 X 4 nm2)transferred onto a glass substrate studied with a cantilever spring constant k = 0.06 N/m and a high repulsive force. the isotherm recording and the dipping experiment were small to ensure a good monolayer stability at the water surface before transferring onto the substrates. Different types of LB deposita can be obtained: X, Y, or Z type. The Y type is the most common and can be obtained with hydrophilicand hydrophobic substrates. A monolayer is transferred a t each substrate movement and multilayers are consequently in a head-to-head or tail-to-tail configuration. The Z type is obtained only with hydrophilic substrates when each monolayeris transferred during an upstroke, and the X type is obtained only with hydrophobic substrates when each monolayer is transferred during a downstroke. Thus Z and X types are composed of multilayers in the tail-to-head and head-to-tail configurations, respectively. AFM measurements were carried out with a Nanoecope 11 (Digital Instruments) a t room temperature in air and in the contact mode. All the tips used for this study were made of a microfabricated SiSNd (Digital Instruments). The AFM and the LB films were allowed to come to a thermally stable state in order to minimize drift while taking images. Moat of the images shown in this paper have been recorded in the constant height mode and only two (Figures 3b and 8) in the constant force mode. Low pass filtering was only used for large scale images (Figure 3) and filtering in Fourier space was applied in the case of atomic scale images.
Results and Discussion First it was decided to characterize with AFM the LB monolayer of a well-known and simple fatty acid molecule: docosanoic acid (DA) to make sure that the quality of our LB films was satisfactory. The obtained isotherm of docosanoic acid has the same appearance as the one found in the literature.29 From this isotherm, one can deduce the best target surface pressure for transfer onto a glass substrate. Extrapolation of the solid-state domain gives the value 0.2 nm2per molecule a t a surface pressure of 30 mN/m. The docosanoic acid willthus be transferred onto a glass substrate when the monolayer is compressed up to this target surface pressure which is maintained constant. The DA multilayer deposits are of the Y type. Transfer ratios around 1were obtained. An AFM image of a DA LB monolayer reveals (Figure 1)a regular and homogeneous structure where a centered hexagonal pattern can be matched to the molecular (22) Penwon, B. N. J. Chem. Phys. Lett. 1987,141 (41,366-368. (23)Ha"a, H. G.; Gould, S. A. C.; Hansma, P. K.; Gaub, H. E.; distribution. It seems that each bright spot can be Longo, M. L.; Zasadzinski, J. A. N. Langmuir 1991,7,1051-1054. attributed to either the end methyl group of a docosanoic (24) Embs,F.;Funhoff,D.;Laschewisky,A.;Licht,U.;Ohst,H.;Prass, acid molecule if the chains are perpendicular to the W.; Ringadorf,H.; Wegner,G.;Wehrmann,R. Ado. Mater. 1991,3,25-31. (25) Gabrielli, G.;Puggelli, M. MonolayersMembr., Symp.1974,346substrate or an ethylene group if the chains are tilted with 361. regards to the substrate. These bright spots match a (26) Malcolm, B. R. Biopolymers 1970,9,911-922.
(27) McMaeter, T. J.; Cam,H.; Milea, M. J.; Cairns, P.; Morris, V. J. J. Vac. Sci. Technol. 1990, A8 (l),648-651. (28) Zubay, G. L. Biochemistry; Addison Wesley Reading,MA, 1983.
(29) Bettarini,S.;Bonosi,F.; Gabrielli,G.; Martini,G. Langmuir 1991, 7,1082-1087.
Langmuir, Vol. 9, No.12, 1993 3569
ATF of PoZy(amino acid) Films
-1.0
nN
I
0.5 nN
0.0 nN
Figure 2. AFM image of a tip-damaged DA LB monolayer (k = 0.06 N/m and high repulsive force).
0-
0
I
100
200
300
A0
400
nm
DAMAGED PAL MULTILAYER BY THE AFM TIP
centered hexagonal pattern disposed relatively regularly on the surface. The length of one side of the hexagon should correspond to the distance between the end of two behenic acid chains and is found to be equal to 0.45 f 0.05 nm. This value is confirmed by other AFM results30and is consistent with the molecular area of behenic acid in a packed monolayer (20A2at 30 mN/m). Although the aim of obtaining an AFM image of a well-known fatty acid LB monolayer had been achieved, soon a problem of tip/ sample damage became apparent as illustrated at large scale in Figure 2. This image was obtained using a cantilever spring constant of 0.06 N/m but using a very high repulsive force. Similar well-known effects will be observed latter on more complex systems such as poly(amino acid) (PBA and PAL) LB films. Typical values of forces used in our experiments were in the range of lo4 N. Hence we observed surface degradation at large and small scales. Images shown in Figure 3 illustrate the cantilever-tipdamage which occurred when studying PAL and PBA monolayers since in addition these images have been recorded with a high cantilever spring constant value (0.58 N/m). Figure 3a represents a tip-damaged PAL LB monolayer deposited on mica. Structures of about 50 nm wide became evident. With further scanning, these structures were better defined and increased in width and thickness forming oriented "bundles" on the surface. The "bundles" seen are definitely not individual polymer molecules;the distances are much toolarge. The bundles thus correspond to aggregates of molecules. The existence of superstructure and aggregates has been reported in a variety of polymer systems,3lespeciallythose of biological importance, probably due to a tip-sample interaction. A periodicity can be noticed in this image and the bundles lie in such a way that the main axis is perpendicular to the scan direction whatever the scanning direction. Figure 3b shows a hole which has been created by scanning the surface of a PBA monolayer transferred on mica. The depth of this hole is approximately 1.6 nm. This small area has been scanned continuously for a few minutes. Then the scanning size was increased to 700 X 700 nm2 and the image of this region was captured after one scan (seeFigure 3c). A damaged squared area showingdistinct aggregates and holes appears in the center of Figure 3c and corresponds to the previous scanned zone (shown in Figure 3b). The conclusion of this study is that since our LB films are composed of polymer chains which do not adhere well to the surface of the substrate, it is important to try to reduce the damage created by the tip. Hence to (30)Fiol, C.; Alexandre, S.;Delpire, N.; Valleton, J. M.; Paris,E.Thin Solid Film 1992,215,8043. (31) On Man Leung; Goh, M.C. Science 1992,255,64-66.
I
DAMAGED PBA L B MONOLAYER BY THE TIP 7 1 . 4 nN
O.? nN
I' I -/
0 . 0 nN
800
51,
-'SO0
0 /
nm
DAMAGED PBA L B MONOLAYER BY THE TIP
Figure 3. Sample/AFM tip interactions recorded with a cantilever spring constant k = 0.58 N/m: (a, top) threedimensional AFM image of a damaged PAL LB monolayer
deposited on a mica substrate (500 X 500 nm2);(b,middle) threedimensional AFM image of a damaged PBA LB monolayer deposited on a mica substrate (170 X 170 nm2) which shows a hole made by the AFM tip after a 15 X 15 nm2 area scan; (c, bottom) previous region (170 X 170nm2)has beenscanned several times, a zoom-out of 700 X 700 nm2 was captured after one scan.
study such samples, it is necessary to use smooth analysis conditions. Similar persistent deformation beha6or on a polystyrene film3' and on a cadmium arachidate monolayer23by an AFM tip has already been reported. When using AFM, the contact force of the cantilever on the film
Auduc et al.
3570 Langmuir, Vol. 9, No. 12,1993
Figure 5. Schematic representation of PAL or PBA a-helices lying parallel to the substrate.
Figure 4. AFM top view images (k = 0.38 N/m) of the substrates: (a, top) mica (5 X 5 nm2);(b, bottom) graphite (5 X 5 nm2).
surfacehas to be minimized. Thus a very low engagement force has to be used and the interacting repulsive force has to be as low as possible just before the jumping point. The value of the cantilever force constant k could influence the interaction with the surface. The smaller k is, the less likely is the damage to the LB monolayer, but the tip jumps off the sample surface relatively easily. Thus a compromise of choice between different cantilever force constants (0.06, 0.12, 0.38, and 0.58 N/m) was made, although a study of several force constants was found in the literature32where it was concluded that k = 0.38 N/m should be the best to study LB monolayers. AFM investigations have also been made on LB films with a liquid cell because this technique is meant to decrease the interacting forces between the tip and the ample.^^^^^ In aqueous medium we have checked the periodicity of the mica substrate at atomic scale and then AFM images of PBA LB multilayersdeposited on mica have been recorded. Unfortunately some damage done by the AFM tip has also been noticed. The AFM images of mica and graphite substrates have been obtained and are shown in parts a and b of Figure 4 respectively. The structure is regular for both substrates; the periodicity found on mica is 0.53 f 0.04 nm and the (32) Peltonen,J. P. K.; He Pingsheng; h n h o l m , J. B.J. Am. Chem. SW. 1992,114,7637-7642. (33) Weisenhom, A. L.; Maivald, P.; Butt,H. J.; Hamma, P. K. Phys. Reu. B 1992,45 (191,1122611232.
one found on graphite is 0.25 f 0.02 nm. The mica and graphite have been chosen as substrates for their opposite hydrophilicity properties. An amorphous glass substrate has been used for DA since this substrate appears to be commonly used in the literature for this compound and thus is a good standard. Being aware of the sample/tip interaction and being sure of mastering the LB experiment since good results have been obtained for DA, it was then possibleto attempt the AFM characterization of LB PBA and PAL monoand bilayers which show a rather different behavior to DA. Firstly, the poly(amino acid) isotherms show a characteristic pattern which also gives a relevant mean molecular area when the extrapolation to a zero surface pressure is made. At a given surface pressure which is characteristic of the polypeptide, a transition can be seen on the isotherm in the form of a plateau if the chains are long or just an inflection if the side chains are shortu Two areas can be observed: the first one corresponds to the formation of a monolayer and the second one to the formation of a bilayer. In theory, the monolayer is composed of ordered a helices, condensed asgroups parallel to the water surface which stay parallel to the substrate during transfer as shown in Figure 5, and undergoes a regular collapse when compression increases to form a bilayer which is represented by a plateau or an inflection on the isotherm. Secondly, the PBA and PAL transfer ratios show, as expected for many complex molecules and polymers,' that multilayers deposited on a hydrophilic substrate (such as mica) are of the Z type and multilayers deposited on a hydrophobic substrate (such as graphite) are of the Y type. The best results regarding the reproducibility of the transfer ratio (TR) are obtained when the substrate is hydrophilic. In this case, transfer ratios of 1.00 f 0.05 and 0.2 f 0.1are obtained for upstroke and downstroke, respectively. In the case of hydrophobic substrate the TR is 0.79 f 0.25 for downstroke and 0.94 f 0.12 for upstroke. The experimental isotherm of PBA shows a plateau at around 10 mN/m as can be seen in Figure 6 which is (34) Malcolm, B. R. Applied Chemistry at protein interfaces; 1973, Chapter 17, Air-water interfam Symp. Washington, pp 338-359.
ATF of Poly(amino acid) Films
Langmuir, Vol. 9, No. 12,1993 3571
poly-P-benzyl-L-aspartate n
0
0 0
10
20
30
40
50
60
70
Figure 6. PBA isotherm obtained experimentally (C = 0.288 mgmL-1).
Figure 7. Three-dimensionalAFM image (17 X 17 nm2) of a PBA monolayer deposited on a mica substrate where an aggregation phenomenon has taken place (k = 0.38 N/m) in constant height mode.
characteristic of the a helix conformation.35 Good correlation can be found with the reference isotherm^.^^^^ A target surface pressure of 7 mN/m was thus chosen to carry out the transfer. Transfer ratios around 1 were obtained. AFM investigations at molecular scale have been led on PBA LB monolayers and bilayers. PBA LB monolayers on mica AFM images are reported in Figures 7 and 8 corresponding to two different samples. Figure 7 shows separated chains which alternate with aggregates of two chains. The differentappearances of images 7 and 8clearly evidence tip perturbation which aligns the chains along the scanning direction. For both images the periodicity parallel and perpendicular to the chain (see Figure 5) are reported in Table I. Image 8, nonfiltered, is shown in Figure 8b. The averaged periodicities found along the chain for Figures 7 and 8 are respectively 0.59 f 0.10 and 0.54 f 0.05 nm and correlate with the literature value of the a-helix turn which is given equal to 0.54 nm.% It confirms the hypothesis that the polypeptide chains is packed in the a-helixconformation. Figure 9 shows a PBA LB bilayer on graphite. The interchain distance measured (35) Baglioni, P.; Dei, L.; Gaabrielli, G.; Innocenti,F. M.; Nimlai, A. Colloid Polym. Sci. 1988,266, 783-792. (36) Lehninger, f i n c i p e de Biochimie; Springer-Verlag: New York, 1982.
Figure 8. (a, top) Three-dimensionalAFM image of a PBA LB monolayer deposited onto a mica substrate obtained with k = 0.58N/m and constant force mode. (b, bottom)Image nonfiltered.
is equal to 0.62 f 0.05 nm. The periodicity along these chains has also been measured and is equal to 0.5 f 0.1 nm; hence the uncertainty contains the a helix turn value% (see Table I). The interchain distance value measured for PBA on mica and which is equal to 1.1f 0.2 nm when using what we called previously the best conditions for AFM measurements is comparable to X-ray diffraction measurem e n t obtained ~ ~ ~ on a 10-monolayer deposit which gives a mean value of 1.3 f 0.1 nm. Indeed, the X-ray diffraction technique measures a periodicity averaged in three dimensions while our AFM result is along one direction and is probably modified by tip-sample interaction. Similar X-ray diffraction measurementSl3 on poly(@-benzyl-Lglutamate) (PBG) cast films have shown that the closest chain separation was 1.4 nm (for PBG, R = (CH2)r COOCH2-C& which means that PBG has only one more CH2 group than PBA). In the case of PBA which displays a shorter side chain length, we expect and have obtained a shorter interchain distance than PBG. The isotherm of PAL (see Figure 10) shows only an inflection at around 22 mN/m because the side chains (R = CH3) are short. The target surface pressure giving the (37) Brunel, M. Private communication.
3572 Langmuir, Vol. 9, No. 12,1993
Auduc et al.
Figure 9. Three-dimensionalAFM image of a PBA LB bilayer deposited on a graphite substrate (k = 0.38 N/m). 40
h
Figure 11. Three-dimensionalAFM image (5 X 5 nm2)of a PAL LB monolayer deposited on a mica substrate (It = 0.58 N/m).
1
30
E \ E
20
W
c
10
0 0
10
20
30
40
50
Mma(I2)
Figure 10. PAL isotherm obtained experimentally (C = 0.22 mg.mL-l). Table 1. Periodicity Measurements Using AFM of the PBA and PAL LB Films Deposited on Mica and Grabhite figure
4a 4b 1 7 8 9 11 12
periodicity periodicity across the system (nm) chains (nm)dl mica 0.53 f 0.04 graphite 0.25 f 0.02 1 DNglass 0.45 f 0.05 1 PBA/mica 1.1f 0.2 1PBA/mica 0.8 f 0.1 2 PBNgraphite 0.62 f 0.05 1PAL/mica 0.7 f 0.1 2 PAL/graphite 0.7 f 0.1
periodicity along the chain (nm)dl Figure 12. Three-dimensionalAFM image (5 X 5 nm2)of a PAL LB bilayer deposited on a graphite substrate (k = 0.38 N/m). 0.59 f 0.10 0.54 f 0.05 0.5 f 0.1 0.5 f 0.1 0.5 i 0.1
optimum transfer ratio and thus chosen to carry out the transfer onto graphite and mica substrates is approximately 13-14 mN/m. This value has already been used in the literature.% AFM investigations at molecular scale have also been done on PAL LB films. Figure 11 shows a PAL LB monolayer on a mica substrate. The interchain periodicity found is approximately 0.7 f 0.1 nm. Figure 12 corresponds to the AFM image of a PAL bilayer deposited on graphite. The interchain distance measured is 0.7 f 0.1 nm. These values can be correlated to the 0.85 nm one found at the air/water interface.6*34An a-helix conformation is also expected for PAL6*%and the a-helix turn (38) Cornell, D.G.J. Colloid Interface Sci. 1979,70,167-180.
along the chain direction of the PAL chains deposited on mica and graphite is equal to 0.5 f 0.1 nm (see Figures 11 and 12). All the measured periodicities by AFM are gathered in Table I and illustrate the ease with which AFM can give good correlated values of interchain distancesfor two poly(amino acid)s which have different side chain lengths deposited by the LB technique on two different substrates (hydrophilic and hydrophobic). PBA LB films seem to be less damaged by the AFM tip than PAL LB films. This could be explained firstly by the differenceof side chain lengthsand/or by the difference of functional groups: PBA long side chains are more entangled than PAL short ones and/or PBA functional side chain groups involve more intermolecularforces than the CH3 PAL side groups. Secondly, due to the higher hydrophobicity of the side chains of PBA (due to bulky benzyl groups) relative to PAL, we expect that, at least for the first layer deposited on an hydrophobic substrate, PBA
ATF of Poly(amino acid) F i l m
will be less mobile than PAL. Both these effects can explain why PBA is more resistant to tip deformationthan PAL. When looking at the inter- (dl)and intrachain (dit) distances, we notice a constant dll value whatever the substrate (mica or graphite) and whatever the side chain length (PBA or PAL). The 0.5-nm value is indeed very close to the mica periodicity or twice the graphite one, so we cannot exclude an epitaxial effect on the substrate. However the dl value changes with the hydrophilicity of the substrate at least as far as PBA with long side chaina is concerned since modifications of dl are not so obvious with PAL. In fact the repartition and the orientation of the backbone on the substrate can depend on two effects: An epitaxial effect which would lead to a preferential orientation of the chains depending on the substrate periodicities. In this case the dl would be constant whatever the chains on a given substrate. However PAL and PBA have the same a-helix turn according to the literature so this cannot allow us to conclude on a possible epitaxial effect. Furthermore, as previously noticed, the mica periodicity is twice that of graphite, Le., too close to be distinguished. A hydrophilicity effect which would lead to attraction or repulsion of the side chains between themselves and the substrate. Note however that we compare one monolayer on mica and two monolayers on graphite. The dependence of dl on the type of substrate can be due to a preferential orientation of PBA side chains. The topmost layer is hydrophobicin both cases (one layer of PBA on mica and two layers of PBA on graphite since the type of deposita are Z and Y, respectively) although certainly slightly different according to the substrate. This leads to the conclusions that even if a possible epitaxial effect cannot
Langmuir, Vol. 9, No. 12, 1993 3673
be excluded,the hydrophilicity of the substrate determines the type of deposition and that the studied poly(amino acid)s (PAL and PBA) seem to have a not distinct amphiphilic behavior.
Conclusion The results shown in this paper have emphasized the ability of local probe microscopy such as AFM to i"ediately give a great deal of structural information (periodicity, conformation)on fatty acid and polypeptide LB films at the nanometer scale. AFM has allowed us to observe a close packed centered hexagonal structure for a docosanoic acid LB monolayer deposited on a glass substrate and a parallel chain distribution in the case of the PBA and PAL LB monolayer and bilayer deposited respectively on mica and graphite. The measured distance between two docosanoic acid chains is 0.45 f 0.05 nm. The PBA interchain distance d l is 1.1 f 0.2 nm and the PAL distance is 0.7 f 0.1 nm. This difference is attributed to the side chain length although at present no precise model can be designed. The measured periodicities by AFM along the chains of PBA and PAL on mica and on graphite are independent of the substrate and are equal to 0.53 f 0.05 nm (average on the five images), which is in agreement with the a helix turn value. The interchain distance in the z direction which would correspond to the monolayer thickness will be evaluated in the near future by angle resolved XPS and IR spectroscopy at grazing incidence.
Acknowledgment. We wish to thank J. C. Duclot for his skillful technical assistance. This research received financial support from a BRITE-EURAM project and we also wish to thank our BRITE-EURAM partners for fruitful discussions.