Langmuir 1995,11, 2713-2718
2713
Formation of Polypeptide-Dye Multilayers by an Electrostatic Self-AssemblyTechnique Thomas M. Cooper,* Angela L. Campbell, and Robert L. Crane Materials Directorate, Wright Laboratory, WL I MLPJ 3005 P Street, Ste. 1, Wright-Patterson Air Force Base, Ohio 45433 Received February 6, 1995. I n Final Form: April 26, 1 9 9 9 To demonstrate the feasibility of preparing ordered multilayers composed of charged macromolecule/dye bilayers, we used an electrostatic self-assembly technique to prepare polypeptide/dye multilayers. We immersed previously protonated silanized glass slides in aqueous solutions (1mM) of two representative anionic dyes: rod-shaped congo red or plate-shaped copper phthalocyaninetetrasulfonic acid sodium salt and cationic polypeptide poly(L-lysine)solutions (15mM monomer). Films resulting from up to 100 dipping cycles were characterized by UV-vis absorption spectra, ellipsometry, CD, and FT IR. FT IR spectra of film material embedded in KBr pellets showed an amide I band at 1654 cm-l, suggestingthin film a-helical conformation. In both systems, the dye absorbance increased with the number of layers. In the copper phthalocyanine system, the absorption spectrum was a combination of phthalocyanine monomer and dimer contributions, with no evidence of higher aggregates. The congo red film dye absorption maximum was a function of the number of bilayers, suggestingcongo red resided in a polar, hydrophilic environment that became less polar with increasing bilayer number. From ellipsometry measurements, the bilayer thickness in both films was determined to be 20 A. The phthalocyanineQ band showed negative-induced CD, while the congo red nn* band exhibited positive-induced CD. The CD measurements gave evidence for ordered dye-polypeptide complexesin films greater than 100-Athickness. The current work suggested a bilayer structure where monomeric congo red or dimeric copper phthalocyanine tetrasulfonate were randomly distributed among excess uolv(K)monomer bindings sites, giving clear films with no scattering centers.
Introduction Thin films have been placed on optical surfaces by a variety of techniques, including spin coating,l dip coating,2 the Langmuir-Blodgett technique? covalent attachment of dyes to a silica surface via zirconium phosphate ~hemistry,4,~ or silicon self-assembly technology.6 Oriented polypeptide thin films have been prepared by direct polymerization onto a ~ u r f a c e .Recently,B-12 ~ multilayers of oppositely charged polymers have been described, with potential electrooptical applications. An early patent describes the preparation of inorganic multilayers on charged surfaces, yielding thin films with varying interference reflection colors.l3 Suspensions of highly charged colloidal particles form crystalline arrays with narrow bandwidth Bragg reflection bands. l4 Previously, we prepared films based on the interaction of oppositely charged polypeptides poly(L-glutamic acid) and poly(^lysine).15 Abstract published in Advance ACSAbstracts, June 15,1995. (1)Levinson, W. A.;Arnold, A.; Dehodgins, 0.Polym. Eng. Sci. 1993, 33,980. (2)Scriven, L. E. Mat. Res. SOC. Symp. 1988,3,717. (3)Fuchs, H.;Ohst, H.; Prass, W. Adu. Muter. 1991,3,10. (4)Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S. Chem. Mater. 1991,3,699. (5) Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey, C. E. D. Science 1991,254,1485. (6)Kakkar, A. K.; Yitzchaik, S.; Roscoe, S. B.; Kubota, F.; Allan, D. S.; Marks, T. J.; Lin, W.; Wong, G. K. Langmuir 1993,9,388. (7)Whitesell, J. K.; Chang, H. K. Science 1993,261,73. (8)Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992,210, 831. (9)Schmitt, J.; Grunewald, T.; Decher, G.; Pershan, P.; Kjaer, K.; Loscge, M. Macromolecules 1993,26,7058. (10)Cheung, J.Ph.D. Thesis, Massachusetts Institute of Technology, 1993. (11)Kleinfeld, E. R.; Ferguson, G. S. Science 1994,265,370. (12)Enriquez, E.; Samulski, E. T. MRS Symp. Proc. 1992,225,423. (13)Iler, R. K. U.S. Patent 3 485 658,1969. (14)Kamenetzky, E. A.; Magliocco,J. G.; Panzer, H. P. Science 1994, 263,207. (15)Cooper, T. M.; Campbell, A. L.; Noffsinger, C.; Gunther-Greer, J.; Crane, R. L.; Adams, W. W. MRS Symp. Proc. 1994,351, 239. @
In this paper, we describe multilayers prepared by electrostaticself-assemblyof the cationic polypeptide poly(L-lysine) (poly(K)) and the anionic dyes congo red (CR) and copper phthalocyanine tetrasulfonic acid (CF'TA).The dyes we chose emphasize molecular shape variation, have documented ability to bind to polypeptides and proteins, and exhibit characteristics similar to dyes used for electrooptical applications. CR is a rod-shaped dichroic molecule which can be aligned into noncentrosymmetric, second harmonic generating films when brushed in one direction.16 The dye undergoes magnetic-field-induced alignment in the presence of aqueous lyotropic nematic liquid crystals17and orients in stretched poly(viny1alcohol) films.lg CR has been shown to bind to proteins, including albuminlg and insulin.20 Phthalocyanines are plateshaped molecules with unique excited states useful for optical limiting or switching applications.21 Chiral phthalocyanines have helical superstructure in the discotic mesophase.22 As a n example of protein binding, copper phthalocyanine trisulfonate gel binds to hemoglobin23and cytochrome c oxidase a p ~ p r o t e i n .In ~ both ~ types of films, film thickness and optical density uniformly increased with the number of processing steps. We also observed induced circular dichroism and linear dichroism in both types of films suggesting the formation of ordered dyepolypeptide complexes. (16)Shimizu, Y.; Kotani, M. Mol. C y s t . Liq. C y s t . 1990,183,341. (17)Kuzma, M. R.;Skarda, V.; Labes, M. M. J.Chem. Phys. 1984, 81,2925. (18)Frackowiak, K.; Fiksinski, K.; Pienkowska, H. Photobiochem. -Phntnhinnhua .______ _ r __. _l-.R -l -R-l , -2 , 2-1-. (19)Tayyab, S.;Qasim, M. A. Med. Sci. Res. 1990,18, 413. (20)Turnell, W.G.; Finch, J. T. J. Mol. Bid. 1992,227,1205. (21)Mansour, K.;Alvarez, D.; Perry, K. J.;Choong, I.; Marder, S. R.; Perry, J. W. SPIE O g .Biol. Optoelect. 1993,1853,132. (22)Van Nostrum, C.; Bosman, A. W.; Gelinck, G. H.; Picken, S. J.; Schouten, P. G.; Warman, J. M.; Schouten, A.; Nolte, R. J. M. J.Chem. SOC.,Chem. Commun. 1993,1120. (23)Baba, Y.; Kawano, Y. Chem. Lett. 1994,181. (24)Ostropolska, L.; Przywarska-Boniecka, H.; Swirska, H. Polyhedron 1988,7,2667.
This article not subject to U.S. Copyright. Published 1995 by the American Chemical Society
Cooper et al.
2714 Langmuir, Vol. 11, No. 7, 1995
e
--+-
20 Bilayers &40 Biluyysrs --st 60 Bilayers --t 80 Bilayers
1I
10'
1@
40, e
2
2
450 500 5 5 0 600
l
650 700
750 800
Wavelength(nm)
Copper phthalocyanine tetrasulfonic acid(CPTA)
Figure 2. Absorption spectra of (pOly(K),CPTA)nmultilayer thin films as a function of the number of bilayers. --A1 -8- 2
Bilayer Bilayers --t 5 Bilayers -e10 Bilayers -5I5 Bilayers
a
8
0.11
0.08 0.06 0.00
400
Alcian Blue(AB)
450
500
550
600
Wavelength(nm)
650
700
Figure 3. Absorption spectra of (poly(K),CR), films as a function of the number of bilayers.
Congo Red(CR) Figure 1. Chemical formulas and abbreviations for dyes used
in this study.
Materials and Methods Materials. Structures and abbreviations for poly(L-lysine)
hydrogen bromide salt, DP = 2300 (Sigma), dyes, and other materials (Aldrich Chemical) are shown in Figure 1. Preparation of SilanizedGlass Slides. Glass slides were cleaned with a boiling solution of 560 mL of concentrated HzS04 and 240 mL of 30% H2Oz for 5 min while stirring. The slides were dipped in a bath of milli-Q water, rinsed with a stream of milli-Q water, and dried with Nz. The glass slides were then cleaned in a solution of 750 mL of milli-Q water, 225 mL of 30% H202, and 225 mL of concentrated NH4OH/&O at room temperature for 2-3 days. The slides were then dipped in a bath of milli-Q water, rinsed with MeOH, and dried with Nz. The reverse of the slide was swabbed with 12% HF solution and immediately rinsed. A room-temperature solution of 784 mL of 95% ethanol and 16 mL of N-[3-(trimethoxysilyl)propyl]ethylenediamine sat for 5 min to ensure silanol formation. The slides were then silanized for 10 min while stirring. The slides were dipped in an ethanol bath, rinsed with ethanol, dried with Nz, and cured for 10 min in a vacuum oven at 110 "C. Thin Film Preparation Procedure. All thin films were prepared in a certified class 100 clean room. Dye (1mg/mL)and polypeptide (2 mg/mL) solutions in milli-Q water were used in the preparation of the slides. All silanized slides were treated with 0.01 N HCl solution for 2 min to protonate the aminogroups present on the surface. The slides were then dipped into and rinsed with milli-Q water and dried with N2. The slides were dipped into the anionic electrolyte solution, then dipped in a milli-Q water bath, rinsed with milli-Q water, dried with Nz, then dipped into the cationicelectrolyte,dippedin a milli-Qwater
bath, rinsed with milli-Q water, and dried with N2. All slides were reproducibly dipped by vertical insertion into the solution. All rinsing and dryingsteps were performed in the same manner during film preparation. The cycle was repeated to obtain the desired number of bilayers (bilayer = material resulting from one anion dip + one cation dip). Nomenclaturefor the multilayer is (x,y),, where x is the cation, y is the anion, and n is the number of bilayers. A special type of multilayer was (Na+,CPTA),or (AB,Cl-),, where films were prepared from dipping into the dye solution, followed by dip into the milli-Q water bath. Although not a true multilayer, the same nomenclature was used for convenience. Absorption Spectra. Absorption spectra and linear dichroism measurements of the thin films on a silanized glass substrate were collectedwith a Perkin-Elmer Lambda 9 spectrophotometer. CircularDichroism. CD spectra were collectedusing a Jasco Model 720 spectropolarimeter. Ellipsometry. Ellipsometry measurements were performed with a Rudolph Research Model S2000 thin film ellipsometer. Infrared Spectra. FT IR spectra were obtained using a Perkin-Elmer Model 1725 infrared spectrometer.
Results and Discussion We collected absorption spectra (Figures 2,3, and 8), spectroscopic ellipsometry (Figures 4A,B and 5A,B), and CD spectra (Figures 6 and 7) of our thin films as a function of the number of bilayers. FT IR spectra of film material removed from the glass and embedded in KBr pellet were also collected (Table 1). We also collected CD spectra of polypeptide-dye complexes in neutral aqueous solution (Table 1). Table 2 contains the measurement of material deposition for varying combinations of dye and surface characteristics. Linear dichroism spectra of selected films (Figures 9 and 10) were also obtained. For all uncomplexed polypeptides, the solution CD spectra (Table 1)in the backbone region (180-260 nm) were characteristic of a coil conformation. Both poly(K)dye complexes retained the solution coil conformation, although the molar ellipticity decreased, suggesting a
Formation of Polypeptide-Dye Multilayers
Langmuir, Vol. 11, No. 7, 1995 2715
+1 Bilayer -a-
Oi
hM
20 Bilayers 40 Bilayers
-E3- 60 Bilayers -B-- 80 Bilayers
5
100 Bilayer
~
0
I
-5
.10
B
c)
8e 0 . 7 & 0.6
g
500
0.5
.B
550
600
650
700
750
800
Wavelength(nm)
u 0.4 2 0.3
Figure 6. CD spectra of (poly(K),CFTA),thin films as a function of the number of bilayers.
0.2
OD
0.1
-
0
-
0
8 6
M
Figure 4. Real (A) and imaginary (B) refractive index components at 632 nm calculated from ellipsometric data for
4E 4
(poly(K),CP!t'A), films.
v
3
.I
2
*
A
. I
.e I
0 -2 -4 300 350 400 450 5 0 0 5 5 0 600 6 5 0 700
Wavelength(nm)
Figure 7. CD spectra of (poly(K),CR),thin films as a function of the number of bilayers.
B
0.5 D
I
0.05
o !
Figure 5. Real (A) and imaginary (B) refractive index components at 523 nm calculated from ellipsometric data for (poly(K),CR), films.
smaller coil radius of gyration as dye negative charges neutralized poly(K) positive charges. FT IR spectra of KBr pellets of multilayer material exhibited amide I bands characteristic of a-helical conformation (Table 1). The CD and FT IR data suggested a surface-induced coil to helix transition during bilayer deposition. The electrostatic assembly process involves a n ionexchange and adsorption processz6 whereby ion A in solution diffuses through the boundary layer and replaces ion B bound to the solid phase: Aaq
+ film-B - B,, + fi1m.A
We investigated the adsorption of one dye layer on silanized slides as a function of dip time and concentration. Silanized slides were dipped into a CPTA solution, and the amount of material deposited onto the slide was (25)Veith, J. A.; Sposito, G. Soil. Sei. SOC.Am. J. 1977, 41, 697.
1
1
1
10
100
Number of Bilayers Figure 8. Mole fraction monomer as a function of the number of bilayers for phthalocyanine-containingfilms.
measured from the absorbance at 680 nm. A series of slides were immersed into dye solutions (concentrations 0.1, 0.5, 1.0, and 1.5 mg/mL) for 2 min. There was no statistically significant variation in material deposited (absorbance = 0.0442 f 0.0015) with increasing dye concentration. A series of slides were dipped into a 1mg/ mL solution for times varying from 1to 300 min. For all dip times, there was no statistically significant variation in absorbance (absorbance = 0.0455 f 0.0015). The data suggested rapid deposition of material reaching equilibrium following formation of one layer. Multiple CPTA dippinglwash cycles did increase the amount of material adsorbed onto the slide a t a rate 2.4 x lov4au per dip cycle (Table 2). An alternative multilayer formation mechanism involves preferential, nonspecific adsorption of the dye of one charge with associated charged counterions, onto the surface. We tested this mechanism by preparing films
Cooper et al.
2716 Langmuir, Vol. 11, No. 7, 1995 Table 1. Aqueous Solution CD and Thin Film FT IR Data band system"
CD Ib
CD I1
amide IC
poly(K) poly(K),CR poly(K),CPTA
202 (-19 200p 197 (-10 600F 202 (-17OO)f
218 (5500) 218 (5500) 218 (1300)
1655 1654 1653
a Abbreviations are poly(K), poly(L-lysine); CPTA, copper phthalocyanine tetrasulfonic acid; CR, congo red. CD spectra of complex in neutral deionized water. Amide I frequency (cm-l) obtained from FT IR spectra of thin film sample in KBr pellet. d Wavelength of maximum CD in nm. Quantity in parentheses is molar ellipticity in deg cm2 dmol-I. e Solution CD spectra were obtained from monomer molar ratio 0.69.0 poly(K),CR complex. fSolution CD spectra were obtained from monomer molar ratio 0.251.0 poly(K),CPTA complex.
Table 2. Dye Charge and Surface Characteristic Effects on Material Deposition materiala
LL4blLW
(Na+,CPTA), (AB,Cl-)n (AB,CPTA), (poly(K),CPTA),, unsilanized slide (poly(K), CPTA),, silanized slide
2.4(0.2) 1.2(0.2) llO(20)
116) 79(5)
a Dye adsorbed onto silanized slide by repeated dipping. Slope (standard deviation) x lo4 (au(680 nm)/bilayer).
0.20
Odegrees 0.164t
/
/ ,
I
\
r . . \
-------
0.044 0.00
4 400
1 450
500
550
600
650
700
Wavelength(nm)
Figure 9. Linear dichroism measurements ofthe (poly(K),CR)zo film. Film orientation axis is parallel to the dipping direction.
500
550
600
650
700
7SO
800
Wavelength(nm)
Figure 10. Linear dichroism measurements on the (poly(K),CPTA)loo film. Film orientation axis is perpendicular to the dipping direction.
(Na+,CPTA),, (AB,Cl-),, or (AB,CPTA),. The rate of absorbance increase per bilayer was significantly greater in the (AB,CPTA), combination than in either (Na+,CPTA), or (AB,Cl-), (Table 2). There was a slightly larger buildup in multilayers (Na+,CPTA), compared to (AEi,Cl-),. The silanized slide had positively charged amino groups which inhibited adsorption of AB relative to CPTA. Multilayer buildup required oppositely charged dyes and was not due to nonspecific adsorption of a dye of one charge plus counterions.
To demonstrate the necessity of silanizing the surface prior to film deposition, we prepared a film series (poly(K),CPTA), on silanized and unsilanized slides. The absorbance buildup per bilayer was 8 times larger in films deposited on the silanized slide than on the unsilanized slide (Table 21, demonstrating the influence of the surface characteristics on the amount of material on the slide. We determined the reversibility of the layer deposition process by preparing a (poly(K),CPTA)l5film and immersing it into 1 M NaCl solution for 10 min. If the material were weakly bound to the surface, the salt ions would displace polypeptide and dye from the surface. Comparison of UV-vis spectra before and after immersion showed a small decrease inA(680 nm) from 0.14 to 0.12, showing the free energy of binding of (poly(K),CPTA)15 was larger than the free energy of poly(K) and CPTA in solution. Beer's law described the absorption spectra (Figures 2 and 31,
A =~cnt
(1)
where A is the absorbance, n is the number of bilayers, 6 is the dye extinction coefficient, c is the concentration of dye within the film, and t is the average thickness per bilayer. For the (poly(K),CPTA),films, the slope of a plot ofA(630 nm) vs n was 0.0086 a a i l a y e r ( r = 0.9989), and for the (poly(K),CR),films, a plot of A(500 nm) vs n was 0.0069 a a i l a y e r (r = 0.9865). The linearity of the plot suggested uniform buildup of material per dipping cycle. Spectroscopic ellipsometry (Figures 4A,B and 5A,B) measured the variation of the real and imaginary index components as a function of bilayer number. We measured A, the phase difference between the parallel and perpendicular reflection components, and tan VI, the ratio of the magnitude of parallel and perpendicular reflection components. tan Y and cos A were measured for each film from 300 to 750 nm. The model used for data analysis of light (wavelength A) in air with an index of refraction n1 = 1and angle ofincidence 41 = 70" interacts with a uniform thin film (thickness d , index nz = 112 - ikz, refraction angle 4 2 ) placed on a substrate (index n3 = n3 - ik3, refraction angle 43). For determination of n3, separate measurements were performed on film-free slides.26-28The filmfree index n3 was determined to be 1.416-0.216i and was independent of wavelength within experimental error. The imaginary component of n3 most likely resulted from the presence of organic silanizing agent on the glass. The wavelength region within the absorption band gave the best signal-to-noise ratio for ellipsometry measurements. We guessed a thickness per bilayer ranging from 10 to 40 A and solved the ellipsometric equations for n2 and k2 (Figures 4A,B and 5A,B). With a large number of bilayers, kz had small variation with guess film thickness. The behavior of n2 and k2 with increasing bilayer number contained information about the film composition, uniformity, and roughness. A nonvarying index with increasing bilayer number would suggest uniform bilayer thickness and composition. A real component approaching unity would suggest high surface roughness. The ellipsometry results for the (poly(K),CPTA),films showed high variation in calculated values of n2 up to 10 bilayers (Figure 4A). The real component then increased with bilayer number, reaching a maximum around 50 bilayers and decreasing thereafter. The data suggested nonuniform (26) Saxena, A. N. J. Opt. SOC.A m . 1965,55, 1061. (27) Tompkins, H. G. A User's Guide to Ellipsometry; Academic: Boston, 1993. (28) Azzam, R. M. A.; Bashara, N. M. Ellipsometry and Polarized Light; North-Holland: Amsterdam, 1987.
Formation of Polypeptide-Dye Multilayers
Langmuir, Vol. 11, No. 7,1995 2717
film buildup for less than 10bilayers, followed by a steady change in composition beginning to level off at around 50 bilayers. The decreasing real component in films greater than 50 bilayers most likely resulted from increasing surface roughness from a large number of dipping cycles. The imaginary component (Figure 4B) showed a steady increase, beginning to level off at k2 0.6 a t 70 bilayers, suggesting uniform deposition of dye per bilayer. The absorbance can be related in terms ofthe imaginary refractive index of refraction according to the expressionz8
-
k2 =
2.303ccil 4n
(2)
Combining eqs 1and 2, we obtained a simple expression for the film thickness
nt =
2.303A.A 47Ck2
(3)
For optically dense multilayers, kz becomes independent of the estimated film thicknessz7and eq 3 was used to estimate the film thickness. We estimated dye adsorption per unit area from measured film thickness, absorption spectra, and published extinction coefficients. Assuming a n extinction coefficient for copper phthalocyanine 4630 nm) = 70 000 M-l cm-1,29for a 100 bilayer film, A(630 nm) = 1and k2 0.6. From Beer's law and the ellipsometric measurement of kz,t$e 100-bilayerfilm$hickness was determined to be 1930 A or average 19 A per bilayer. The molar concentration of dye in the film was calculated to be 0.74 M correspondigg to a dye number density = 4.5 x 1020/ cm3. For a 19-A x l-cm x l-cm bilayer, the number of dye molecules/cm2/bilayerwas 8.6 x 1013or 1.43 x mol of dye/cm2/bilayer. An area density of 1014/cm2cprresponds to a n area of 100A2/moleculeor diameter of 11A/molecule. A recent study of Langmuir-Blodgett films of copper phthalocyanine derivatives30 gives a core surface area of approximately 200 Az per molecule. From compression isotherm data, the two-$imensional gas-to-crystal transition point occurs a t 70 A2 per molecule, corresponding to a n edge-on molecular arrangement. The.similarity with Langmuir-Blodgett data suggested the copper phthalocyanine molecular plane had a n edge-on orientation with respect to the surface. The calculated real refractive index component of the (poly(K),CR), films showed high fluctuation up to five bilayers, becoming uniform thereafter (Figure 5A). Similar behavior was observed in the imaginary component (Figure 5B), decreasing to a constant value kz 0.5 for greater than five bilayers. The results suggested nonuniform material deposition for less than five bilayers and becoming uniform with more than five bilayers. For a (poly(K),CR)lsfilm (A(523 nm) = 0.17, 6 = 30 000 M-l cm-I,17 kz 0.51, the estimated av5rage film thickness was 326 A, corresponding to 22 A per bilayer. The calculated number density of the dye in (poly(K),CR)ls was 1.7 M or 1.1x 1021/cm3. For a 22-A x l-cm x l-cm bilayer, the number of dyes/cm2/bilayerwas calculated to be 2 x 1014or 3.3 x mol/cm2/bilayer. Our bilayer thicknesses (-20 A) wereocomparable to those reported for pol er-polymer (23 A/bilayer)8v9and polymer-silicate (32 ilayer)" systems. The calculated density per unit area paralleled literature results. The density of reactive groups per unit area in silanized glass varies from 8 x 1014 to 8 x 10I6 molecules/cm2.31Approximately 10% of the available amino groups actually
-
-
E
,
(29) Blagrove, R. J.; Gruen, L. C. Aust. J . Chem. 1972,25, 2553. (30)Albouy, P. A. J . Phys. Chem. 1994,98,8543. (31) Ishida, H. Polym. Compos. 1984, 5, 101.
bonded to the dye. Literature measurement of the density of the groups on a silanized surface includes 3 x 1014/cm2 for zirconium phosphate multilayer^,^ 1.02 x 1014/cm2for fidlerene-modifiedsilanized glass surface,32and 1.1x 1013/ cm2 for cytochrome c modified silanized glass surface.33 The aggregation of CPTA in solution and thin filmsz9 causes changes in the Q-band absorption spectrum. The apparent extinction coefficient has monomer ( x ) and dimer (1 - x) contributions
Mole fraction monomer was used as a probe of film structure. We prepared four types of related films: (AB,CPTA),, composed of oppositely charged dye molecules, (Na+,CPTA),, (AB,Cl-),, and (poly(K),CPTA),. Using published monomer and dimer extinction coefficients, mole fraction monomer was measured as a function of the number of dipping cycles (Figure 8). The highest fraction monomer ( x GZ 0.35) was observed in films containing only CPTA. The lowest mole fraction monomer ( x 0.075) appeared in the (AB,CPTA), and (AB,Cl-), films, with intermediate values of x = 0.25 in the (poly(K),CPTA), films. In all films, the proportion dimer increased with the bilayer number, the highest dimer formation occurred in films containingtwo oppositely charged phthalocyaninelike species or just a positively charged species, intermediate dimer formation in polypeptide-dye films, and minimum dimer formation in films containing just negatively charged CPTA. We observed no evidence of higher aggregates29in any of the films. CD spectra of (poly(K),CPTA),films (Figure6) exhibited a negative band associated with the phthalocyanine Q band. There was no induced dichroism in the first 10 bilayers, with increasing dichroism thereafter. The dimerized species resided in a chiral environment formed from electrostatic association with poly(K). Positive Q-band CD with a band shape mirroring the absorption spectrum has been observed in the helical superstructurecontaining liquid-crystalline phase of a chiral phthalocyanine.22 Similarly, the (poly(K),CPTA),film CD spectrum containing a single negative band mirroring the dimer absorption spectrum implying the films contained optically active dimeric superstructures was observed. With increasing bilayer number, the (poly(K),CR), absorption band shifted from 498 to 506 nm as n increased from 1to 20. We fit the data to the empirical equation
,Imm(nm)= 498
+ 11.7(1 - e
-0.0642h-1)
1
(5)
The CR nn* transition energy probed the chromophore environment. Most of the transition energy shift occurred in the first 10 bilayers. There was no direct interaction of the initial layers with glass, as mechanically brushed CR thin films onglass absorb a t 510 nm.34 Coincidentally, for large n ,A converged to 510 nm. A plot of published solvent effect data vs solvent pa1-ameterEd30)~~,~~ revealed CR in high-polarity solvents has a n absorption band at -490 nm, red shifting to -520 nm in low-polarity solvents. CR ligated to uncharged, nonpolar, protein-binding sites has a n absorption maximum in the range 522-540 nm.19,37,38 CR bound to poly(K)in neutral aqueous solution (32) Chen, K.;Caldwell,W. B.; Mirkin,C. A. J . A m . Chem. SOC.1993, 115, 1193. (33) Stayton, P. S.; Olinger, J. M.;Jiang, M.; Bohn, P. W.; Sligar, S. G. J . A m . Chem. SOC.1992,114, 9298. (34) Shimizu, Y.; Kotani, M. Opt. Commun. 1989, 74, 190. (35) Mera, S. L.; Davies, J. D. Histochem. J . 1984, 16, 195. (36) Kosower,E. M. An Introduction to Physical Organic Chemistry; Wiley: New York, 1968. (37) Benditt, E. P.; Eriksen, N.; Berglund, C. Proc. Natl. h a d . Sci. U S A . 1970, 66, 1044.
Cooper et al.
2718 Langmuir, Vol. 11, No. 7, 1995 has a n absorption maximum of 486-500 nm with the charged side-chain polypeptide in a coil c o n f ~ r m a t i o n , ~ ~ shifting to 540 nm at pH 11, with the polypeptide being a neutral side-chain a-helix. Our thin film absorption spectra suggested CR resided in a polar, hydrophilic environment that became less polar with increasing layer number. As the FT IR data demonstrated the film polypeptide conformation to be a-helical (Table 11, the position of the absorption maximum most likely resulted from electrostatic interaction between the CR sulfonate negative charges and protonated €-aminogroups of poly(K). CD spectra of (poly(K),CR),films showedinduced optical activity that increased with bilayer number. No dichroism was observed in the first five bilayers, then increased steadily thereafter. There was a positive band (-347 nm) associated with an absorption band a t the same wavelength. We also observed positive (-470-nm) and negative (-590-nm) bands associated with the 500-nm JCJC*transition. Measurement of optical activity has been performed on CR complexed with p o l y ( ~ - l y s i n e ) . ~The ~ - ~neutral ~ solution poly(K)/CR complex exhibits a negative band a t -365 nm, a weak positive band a t -435 nm, and a n intense negative band a t -520 nm.39 In our thin films, the observed positive CD band had a n ellipticity maximum (470nm) 3 standard deviations below the mean observed = 525 f 17 for a series of protein-CR complexes:45,A nm. The positive ellipticity maximum was blue-shifted from the film absorption maximum. The CD spectrum blue shift suggested the presence of two CR fractions in the film. The first was a low mole fraction (x < 0.1) chiral CR fraction tightly bound to the charged a-helical (Table 1)+amino poly(K) side chains, distinct from either the protein-CR complex45or the neutral solution poly(K1-CR complex.39 The second was a nonchiral, loosely bound fraction ( x > 0.9) with a n absorption maximum around 500 nm. Preliminary microscopic observations under crossed polarizers showed birefringent regions. We present examples of the dye orientation measurement in (poly(K),CR)zo films (Figure 9) and (poly(K),CPTA)loofilms (Figure 10). Slight orientation was observed in both systems, with a n orientation factor S = 0.08 in the (poly(K),CR)zo film and S = -0.03 in the (poly(K),CPTA)loo film. The parallel-to-molecular axis CR XJC* transition dipole tended to orient parallel to the dipping direction. The CPTA in-plane Q-band transition dipole tended to orient perpendicular to the dipping direction. The orientation driving force may have been slide-dipping or shear induced by the milli-Q water rinse. Current work in our laboratory aims to understand the dye orientation mechanism. The multilayer deposition technique has analogies to the biomineralization process. The mechanism of crystal formation is epitaxial growth of the mineral phase upon an organic matrix c o m p ~ s i t e . ~Mollusk ~ , ~ ’ organic matrices
comprise many macromolecules, including polypeptides rich in aspartate and glutamic acids and acidic sulfated polysaccharides as well as an insoluble fraction containing hydrophobic amino acids. Two biomineralization processes have been described: biologically-induced mineralization, where there is an interaction between biologicallyproduced metabolite end products, and organic matrixmediated biomineralization, where minerals are grown onto a preformed organic structural frameworka4*In the films described in this work, the silanized glass served as a n organic matrix for ordered growth of multilayers. Polyions (“metabolites”) adhered to the surface in a planned sequence, analogous to biologically-induced mineralization. Film thickness and properties will potentially be influenced by the type of silanizing agent used, electrolyte concentration, ionic strength, immersion time, slide insertion speed, and rinsing and drying conditions. Our current electrolyte concentration conditions (-1 mM dye, 15 mM lysine monomer) minimized dye aggregation. Film morphology will be influenced by the molar ratio of opposite-charged species. For example, mixed solutions of polyb-glutamate)/poly(K) had a coil conformation when monomeric molar excess of one species was present but had increased @-sheetcontent when the monomer molar ratio approached unity. Films formed from electrolyte solutions containing equimolar amounts of poly(Lglutamate)/poly(K) tended to be cloudy and have high surface r o u g h n e s ~ . lIn ~ the ~~~ current work, the electrolyte solution poly(K)monomer to dye molar ratio was >>1.The current work suggested a bilayer structure where monomeric CR or dimeric CPTA was randomly distributed among excess poly(K)monomer binding sites, giving clear films with no scattering centers.
(38)Klunk, W. E.; Pettegrew, J. W.; Abraham, D. J . J.Histochem. Cytochem. 1989,37,1293. (39)Yamamoto, H.; Nakazawa, A.; Hayakawa, T. J. Polym. Sci: Polym. Lett. Ed. 1983,21, 131. (40)Hatano, M.; Yoneyama, M.; Sato, Y.; Kawamura, Y. Biopolymers 1973,12,2423. (41)Sato, Y.; Woody, R. W. Biopolymers 1980,19,2021. (42)Yamamoto, H.; Nakazawa, A. Bull. Chem. SOC.Jpn. 1983,56, 2535. (43)Yamamoto, H.Makromol. Chem. 1983,184,1479. (44)Bouvier, M.;Brown, G. R. Biochim. Biophys. Acta 1989,991, 303. (45)Edwards, R.;Woody, R. Biochemistry 1979,18,5197-204. (46)Crenshaw, M.A. In Mechanisms of Normal Biological Mineralization of Calcium Carbonates; Nancollas, G. H., Ed.; SpringerVerlag: New York, 1982;pp 243-257. (47)Weiner, S.;Traub, W. Phil. Trans. R. SOC.London, Ser. B 1984, 304, 425.
Acknowledgment. We thank Ms. Donna Brandelik (Science Applications International Corp.) for assistance in preparing silanized glass, Mr. George Orbits (Science Applications International Corp.) for construction of dipping apparatus, and Dr. Mike Rubner and Mr. Jeff Baur (Massachusetts Institute of Technology) and Dr. Ashock Sabata (ARMCO Steel Corp.)for useful discussions.
Conclusions In this paper, we have demonstrated the preparation of polypeptide-dye multilayers by a n electrostatic selfassembly technique. The results demonstrate the feasibility of multilayer preparation based on interaction between a macromolecule and a n oppositply-charged low molecular weight species. Films of 20-A thickness per bilayer were prepared and showed a uniform increase in thickness and optical density. Evidence for induced ordering of the dye by the polypeptide matrix was obtained from circular and linear dichroism measurements. The linear dichroism data suggested flow-induced dye orientation during the dipping process. This work and literature results suggest the electrostatic assembly technique is useful for preparing multilayers composed of small molecules, macromolecules/small molecule hybrids, and macromolecules.
LA950086N (48)Lowenstam, H.A.; Weiner, S. In Biomineralization and Biological Metal Accumulation; Westbroek, P., de Jong, E. W., Eds.; D. Reidel: New York, 1983;pp 191-203. (49)Cooper, T.M.Unpublished observations.