Langmuir 1993,9, 315S3165
3159
Verification of the Internal Reflection Spectroscopy Adsorption Density Equation by Fourier Transform Infrared Spectroscopy Analysis of Transferred Langmuir-Blodgett Films W. H. Jang and J. D. Miller' Department of Metallurgical Engineering, University of Utah, Salt Lake City, Utah 84112 Received April 12,1993. In Final Form: September 10,199P Recent research has demonstrated the utility of the internal reflection spectroscopy (IRS) adsorption density equation for in situ study of surfactant adsorption phenomena. The IRS adsorption density equation has been examined by several researchers and in one form is r = (AIN - €Cbde)/1000€(2ddd), where I' is the adsorption density of adsorbed species, AIN is the absorbance per reflection, t is t i e integrated molar absorptivity, cb is the bulk concentrationof the surfactant, de is the effective thickness, and d, is the depth of penetration. Previously, the validity of the adsorption density equation had only been confirmed from solution depletion measurements using high surface area powders. Now a more direct examination of the IRS adsorption density equation has been completed in our laboratory using a Langmuir-Blodgett (LB) FTIRIIRS technique. Experimental results with LB films of 18C fatty acids transferred to the internal reflection element provide further confiiation of the validity of the adsorption density equation and thus clearly show that the adsorption density equation can be used, with confidence, for in situ and ex situ studies of surfactant adsorption phenomena.
Introduction Internal reflection spectroscopy (IRS),alternatively referred to as attenuated total reflection (ATR), is a method for recording the spectrum of species which are in intimate contact with an internal reflection element (IRE).The technique was first proposed by Harrickl and independently by Fahrenfort2 in the 1960s. Reviews of IRS have been given by several researcher^.^^ A combination of multiple internal reflection with Fourier transform infrared spectroscopy (FTIR)gives an improved signallnoise ratio over more conventionalIR sampling and allows for the study of surfactant adsorption on single crystal IRE surfaces even a t submonolayer coverage. The adsorption of surfactants is important in many areas of technology including suspension stabilization, detergency, tribology, and flotation.&'l In recent years, FTIR/ IRS has been used with reactive internal reflection elements (IRES)to determine in situ and in real time the adsorption density of surfactant a t solid-liquid interfaces using the IRS adsorption density equation.12-16 Sperline et al.12calculated the adsorption density of cetylpyridinium chloride a t the surface of a reactive zinc selenide IRE for
both aqueous and organic solutions. The results were validated, with a relative standard deviation f25%, by comparison with the r values determined by solution depletion experiments with high surface area powders. Subsequently, researchers at the University of Utah have developed the use of reactive internal reflection elements to study surfactant adsorption and measure adsorption densities using the IRS adsorption densityequation. Many different internal reflection elements have been used including &03,13@-18 CaF2,13-17KCl,'@and CaC03.20In this way adsorption isotherms were constructed and in several instancesthe validity of the IRSadsorption density equation was confirmed by solution depletion measurements using high surface area powders. Figure 1compares in situ FTIR/IRS adsorption densities calculated from the IRS adsorption density equation14with those determined by solution depletion using an ex-situ radiotracer technique.21 Based on these results the IRS adsorption density equation has been used to analyze protein adsorption phenomena.22 Now, a more direct examination of the IRS adsorption density equation has been completed in our laboratory usinga Langm+Blodgett (LB)23-26 FTIR/IRStechnique.
Abetractpublishedin Advance ACSAbstracts, October 15,1993. (1)Harrick, N. J. J. Phys. Chem. 1960,64,1110-1114. (2)Fahrenfort, J. Spectrochim. Acta 1961,17,698-709. (3)Haller, G.L.; Rice, R. W.; Wan, Z. C. Catal. Rev.-Sci. Eng. 1976,
Saatry, K. V.,Fueretenau, M. C. Eds.; SME/AIME: Littleton, CO, 1989;
13,259-284.
(4)Hansen, W. N. In Advances in Electrochemistry and Electrochemical Engineering;Muller, R. H., E&.; Wiley-Interscience: New York, 1973:DD 1-60. ( 5 )Strojek, J. W.; Mielczarski, J.; Nowak, P. Adu. Colloid Interface Sci. 1983,19,309-327. (6)Mirabella, F. M. Internal Reflection Spectroscopy: Review and Supplement; Harrick, N. J., Ed.; Harrick Scientific Corp.: Ossining, NY, 1985. (7) Mirabella, F. M. Appl. Spectrosc. 1990,8,20-30. (8)McIntyre, R. T. In Fatty Acids in Industry; Johnson, R. W., Fritz, E., Eds.; Dekker: New York, 1989; pp 351-375. (9)Rakoff, P.; Nidock, J. J. In Fatty Acids in Industry; Johnson, R. W., Fritz, E., Eds.;Dekker: New York, 1989; pp 431-450. (10)Hiemenz, P. C. Principles of Colloid and Surface Chemistry, 2nd ed.; Dekker: New York, 1986. (11)Fuerstenau, M. C.;Miller, J. D.; Kuhn, M. C. Chemistry of Flotation; SME/AIME: New York. (12)Sperline, R. P.;Muralidharan, S.; Freiser, H. Langmuir 1987,3, 198-202.
(13)Miller, J. D.; Kellar, J. J. In Challenges in Mineral Processing;
pp 109-129. (14)Kellar, J. J.; Cross, W. M.; Miller, J. D. Appl. Spectrosc. 1989,43, 1456-1459. (15)Kellar, J. J.; Young, C. A,; Knutaon, K.; Miller, J. D. J. Colloid Interface Sci. 1991,144,381-389. (16)Miller, J. D.; Kellar, J. J.; Cross, W. M. In Advances in Coal and Mineral Processing Using Flotation; Chandar, S., Ed.; SME/AIME, Littleton, CO, 1990,pp 33-44. (17)Kellar, J. J.; Cross, W. M.; Miller, J. D. Sep. Sci. Technol. 1990, 25,2133-2156. (18)Cross, W. M.; Kellar, J. J.; Miller, J. D. Prepr. Pap.-Znt. Miner. Process. Congr., 17th 1991,319-338. (19)Yalamanchili, M. R.;Kellar, J. J.; Miller, J. D. Prepr. Pap.-Int. Miner. Process. Congr., 17th 1991,131-142. (20)Young, C. A.; Miller, J. D. Prepr., SME Annu. Meet. Reno, Feb. 15-18, 1993. (21)Hu, J. S.;Misra, M.; Miller, J. D. Znt. J. Miner. Process. 1986, 57-72. (22)Fu, F. N.;Fuller, M. P.; Sigh, B. R. Appl. Spectrosc. 1993,47, 98-102. 1935,57,1007-1022. (23)Blodgett, K. B. J.Am. Chem. SOC.
0743-7'46319312409-3159$04.00/0 0 1993 American Chemical Society
Jang and Miller
3160 Langmuir, Vol. 9, No. 11, 1993
-
5
0 '
h
.4 t -d E
"
In-situ FTIMRS, Kellar et al. 0 Ex-situ, Hu et al. 0
1
I 0
1 2 3 4 Equilibrium Oleate Concentration (xl0'M)
5
Figure 1. Comparison of the adsorption of oleate by fluorite using ex situ solution depletion(radiotracer)l'and in situ (FTW IRS)21techniques.
In this paper, the adsorption densities of transferred LB monolayer films of stearic and oleic acids were calculated from FTIR/IRs spectra using the IR8 adsorption density equation and comparedwithadsorption densitiesobtained from the measured LB surface pressurearea isotherms. Molecular orientation in the monolayer after transfer was examined by polarized light FTIR/IRS experiments and a detail discussion of these results will be presented in another publicati~n.~' Experimental Section Materials and Equipment. Stearic acid (+99% pure) obtained from John-Matthey and oleic acid (+99% pure) obtained from Sigma Chemical were used without further purification. Chloroformand carbon tetrachloride(99.99%pure) were used as received from EM Science. The LB f i i deposition system was a Lauda Langmuir f i i balance manufactured by SYBRON/Brinkmann. FTIR experimentswere conducted with a dry-air purged Bio-Ftad Digilab Division FTS-40FTIR spectrometer equipped with a liquid-nitrogen-cooled wide-band mercury cadmium telluride (MCT) detector. All spectra were obtained at a resolution of 8 cm-l. Deuterium oxide (+99.9% pure) was purchased from Cambridge Isotope Laboratoriesand was used as received. Milli-Qwater (+UMa) was used in all experiments. Calcium fluoride single-crystal parallelepiped internal reflection elements (60X 10 X 2 mm,75O acute angle) were purchased from Optovac, Inc. Calcium fluoride windows obtained from Spectra Tech, Inc., and germanium windows obtainedfrom Harrick Scientificwere used for molar absorptivity measurements. LB Film Preparation. Monolayer spreadingsolutionswere produced by dissolving stearic acid or oleic acid in chloroform; the concentrations of stearic and oleic acids were 7.6 and 4.8 mM, respectively. The monolayers were spread from the chloroform solutions on the surface of high purity water (+18 Mil) with a large enough area to achieve complete spreading. Chloroform was permitted to evaporate for 30 min before compression of the Langmuir f i i . Surface pressurearea isothermswere measured using the Langmuir film balance at 23 OC.
For preparation of a monolayer, the LangmwBlodgett techniquep28 was applied. A calcium fluoride single crystal (50 x 10 x 2 mm,75O acute angle) was used as the substrate. The substratewas dipped into the subphase (water)before spreading the monolayer solution. After chloroform wm permitted to evaporatefor 30 min, the substrate was withdrawn at a speed of 3 mm/min with the surface pressure held at the desired level. (24) Blodgett, K. B.; Langmuir, I. Phys. Rev. 1937,51,964-982. (25) Gaines, G. L. Inaoluble Monolayers at Liquid-Gas Interfaces; Interscience Publishers: New York, 1966. (26) Roberts, G. G. Langmuir-Blodgett Elms; Plenum Press: New York, 1990. (27) Jang, W. H.;Cross, W. M.; Miller, J. D., Department of Metallurgical Engineering, University of Utah, unpublished results.
Internal Reflection Element (IRE)
Figure 2. Schematicof light undergoing multipletotalinternal reflections in an IRE (0 is the incident angle of light for internal
reflection).
This transferspeed is in the range of typicalspeedsfor transferring the initial monolayer onto a prepared substrate,2Bfrom 10 rm/s to about 1 "/a, and therefore shouldensurethe uniform transfer of the monolayer to the substrate. While the monolayers were transferred onto the calcium fluoride surface, all the transfer curves were recorded so as to ensure that each monolayer was
transferred at a constant pressure. The variation in the surface pressure of the transfer was within fl mN/m. For multilayer deposition experiments,the Y-typedepositionmethod was Transmission Spectroscopy. Liquid transmission measurements using a demountable liquid transmission cell with a spacer and calciumfluoridewindowswere conductedtodetermine molar absorptivitiesof stearicor oleicacid in carbon tetrachloride and sodium oleate in DzO. The path length of the sample was evaluatedmore accurately using an interferencefringe counting techniquW rather thanby the specifiedthicknesa of the spacer. Germanium windows were used only for the molar absorptivity measurement of pure oleic acid. Even with the thinnest spacer available (15pm),the absorbance of pure oleic acid at the desired frequencieswas toohigh to be measured. Therefore,transmiesion measurements of pure oleic acid were carried out after squeezing the oleic acid between the windows without a spacer in order to obtain band absorbances which obeyed the Beer-Lambert law. The path lengthof the pure oleic acid samplewas found by using an interference fringe counting technique or by applying the Beer-Lambert law in somecases. Allthe spectrawere the results of 512 or 1024 coadded scans with the same IR beam size. FTIR/IRS. Due to the nature of organic contaminants in suchsurfacechemistry studies,the followingcleaningprocedures were applied to clean the IRE: Firat, the IRE was rinsed with acetone/water/methanol/water. Second, the IREwas placed in a Tegalplasma chemistryreactor and subjected to oxygenplaama for 30-40 min. In somecases, the surface of the IRE was polished with alumina powders (0.5or 0.05 pm) and washed with highpurity water to create a fresh surface. An internalreflectioncellwith a variable angleholder (Harrick Scientific) was wed. The entrance face of the parallelepiped calcium fluoride IRE was cut at a 75O angle and the variable angle holder was set at 4 2 O giving an incident angle of 52.3O at each internal reflection. The background spectrum for all IRS measurements was the single-beam spectrum of the clean IRE. Sufficientcare was taken to reposition the IREin its holder after the LB f i was transferred. All spectra were recorded using 1024 coadded scans with the same IR beam size. Absorbance spectra of the LB f i i of stearic and oleic acids were ratioed against the background spectra of a clean calcium fluoride IRE.
FTIR/IRS Adsorption Density Equation Electric Field Amplitudes. IRSinvolves the internal reflection of light through an IRE. Figure 2 is a schematic representation of a ray of light undergoing multiple internal reflections in an IRE. For light to be totally internally reflected the following conditions must be satisfied (28) G f l i h , P. R.; de Haseth, J. A. Fourier Transform Infrared Spectroscopy; John Wiley & Sons: New York 1986; pp 350-363. (29)FT04-036DemountableLiquid CeUKitInstructions,SpectraTech, hC.
IRS Adsorption Density Equation
Langmuir, Vol. 9, No. 11, 1993 3161
n1> n2
e > ec,
(1)
ec = sin-' nZ1
(2) where nl is the refractive index of IRE, n2 is the refractive index of sample, n21= n2/n1,0is the incident angle of light for internal reflection, and 8, is the critical angle. As the light is totally reflected at the IREIsample interface, an evanescent wave passes from the IRE and decays exponentially into the sample. The propagating distance is typically a few micrometers or less away from the IRE surface. This wave consists of the three electric field vectors which can interact with the sample. The electric field amplitudes are given by30 Ey =
2 COS e (1- n212)1/2
(3)
E, =
2 cos e (sin%- n2:)1/2 (4) (1- n2:)1/2[(~ + nZl2)sin2e - n 2 1 2 ~ 1 / 2
E, =
2 COS 8 sin 8 (1- n2:)''2[(1 + nZl2)sin2e - n 2 1 2 ~ 1 / 2 (5)
For perpendicularly polarized light (transverse electric wave), Eo2 = Ey2,and for parallel polarized light (transverse magnetic wave), Eo2 = Ex2+ EZ2. Depth of Penetration and Effective Thickness. Harrickso arbitrarily defined the depth of penetration, d,, to be the distance required for the electric field amplitude to fall to l / e of its value at the surface
d~
h - 2?m1(sin2e - nZ12
1/2
(6)
where h is the wavelength of the light. Harrickm also defined an effective thickness, de, as the thickness of material which gives the same absorbance in transmission spectra at normal incidence as that found from IRS
n21E,2d, de = (7) 2 COS e Accordingly, the effective thicknesses for perpendicular (TE wave) and parallel (TM wave) polarization are, respectively' - 2n2,d, cos2 0 de, 1 - n21 2nzld, cos O(2 sin2 0 - n,:) (9) dell = (1- n2,2)[(1+ n,?) sin2 e - n2:1 Thus, the average effective thickness of the unpolarized radiation is approximated bye
de = d
+d 2 Absorbance per Reflection. Tompkins31 derived an equation that described the absorbance per reflection for multiple IRS when a thin film is adjacent the IRE -=A n2'E,2 e
s," C ( z ) exp(-")
dz (11) d, where A is the integrated absorbance (cm-l) due to internal reflections, N is the number of internal reflections, t is the
N
case
(30)Harrick, N. J. InternalRejlection Spectroscopy, 2nd ed.; Harrick Scientific Corp.: Oesining, NY,1979. (31)Tompkins, H. G.Appl. Spectrosc. 1974,223, 335-341.
integrated molar absorptivity (L/cm2mol),z is the distance from the IRE surface (cm), and C ( z ) is the concentration as a function of distance from the IRE surface (mol/L). The number of internal reflections, N , can be found for parallelepiped geometry from the relationship 1 N = -cot e (12) h where 1 is the length of the IRE with adsorbed surfactant and h is the thickness of an IRE. Molar Absorptivity. The molar absorptivity, e, can be evaluated from transmission experiments and using the Beer-Lambert equation
A = ebC (13) where b is the thickness of sample (cm) and C is the surfactant concentration (mol/L). The thickness of the sample, b, can be estimated using the following equation and an interference fringe counting procedure2812Q
where n is the refractive index of sample, f1 and f2 are the wavenumbers where counting starts and ends (cm-l), and k is the number of fringes. Adsorption Density Equation. By assuming a concentration step profile for species adsorbed or in intimate contact with the IRE, Le., c ( z ) = Ci + c b for 0 < z < t, and C ( z ) = c b for t < z < infinity and substituting d , and de into eq 11,together with the approximation that the thickness of the coated layer, t , is much less than d , (exp(-2t/d,) = 1- 2t/d,) the absorption per reflection, AIN, is
A / N = tcbd, + t(2dddp)(Cit) (15) Substituting Cit = O l OOr and rearranging eq 15,the FTIR/ IRS adsorption density equation similar to the one proposed by Sperline et a1.12 can be obtained A / N - &,de r = 1000 e (2d$d,) where r is the adsorption density of adsorbed species (mol/ cm2). In some cases, the absorbance contribution due to transmission through adsorbed surfactant at the IR entrance and exit faces of the IRE might be considered and this correction will be discussed elsewhere.32 Generally, this contribution, if present, amounts to only a few percent of the calculated adsorption density and can frequently be ignored.33 It is clear that this relative contribution to the total absorbance decreases with an increasing number of internal reflections. To calculate the adsorption densities for deposited LB films, eq 16 should be modified as follows: The concentration of the LB film, C ( z ) ,is a step function with C ( z ) = Ci for 0 < z < t and C ( z ) = 0 elsewhere, i.e., c b = 0. Therefore, for the LB films, the FTIR/IRS adsorption density equation is
r =1000AIN e(2dJdP)
(17)
On the other hand the area/molecule can be directly determined for LB films found at the aidwater interface (32)Free, M.; Jang, W. H.; Miller, J. D., Department of Metallurgical Engineering, University of Utah, unpublished results. (33)Ulman, A. Fourier Transform Infrared Spectroscopy in Colloid andznterjace Science; Scheuing, D. R., Ed.; ACS Symposium Series 477; American Chemical Society: Washington, DC, 1990;pp 144-159.
Jang and Miller
3162 Langmuir, Vol. 9, No. 11,1993
(a) Transmission Spechum of Stearic Acid in CC4 Solution
"
\ \
m IO
0
0
IO
20 30 40 50 60 Area per Molecule (A*/tvIolecule)
(b) IRS Speceum of Transferred LB Film One Monolayer of Stearic Acid on CaF2
70
Figure 3. Surface pressure-specific area isotherms for stearic acid and oleic acid. during Langmuir experiments. Thus, the adsorption density of LB monolayer films can be monitored quantitatively from the surface pressure-area isotherm and a direct measurement of the adsorption density for the transferred monolayer can be made without introducing uncertain assumptions. It should be noted that, for good transfers, the orientation packing and characteristics of molecules on the substrate are expected to be similar to those on the water ~ u b p h a s e . ~ Although ~7~~ some researchers challenge this it is expected that only small variations in parking density occur due to transfer.3M8 These effects can be minimized by careful control of experimental conditions. Therefore, the adsorption densities found from the surface pressurearea isotherms for LB films can be compared with the adsorption densities calculated using the FTIR/IRS adsorption density equation, eq 17.
( c ) IRS Spectrum of Transferred LB Film
Five Monolayers of Stearic Acid on CaF2
I
I
I
I
I
1750
1700
1650
1600
1550
I
1500
I
1450
I
1400
Wavenumbers (cm-1)
Figure 4. Comparison of the FTIR/IRS spectra of transferred LB f i i of stearicacid (1and 5 monolayers)with the transmission spectrum of stearic acid in CCL solution.
than saturated C-C bonds and upon compression the attraction between the double bond and water molecules causes the parking area of oleic molecules to reflect amore expanded state. Results and Discussion Characteristic of the Transferred LB Films. After examination of the r-A isotherms, the spread monolayers Surface Pressure-Area Isotherms. Figure 3 shows of stearic acid were found to be in liquid condensed and the surface pressure-specific area ( F A ) isotherms for superliquid states39at the surface pressures of transfer, stearic and oleic acids at the aidwater interface as obtained whereas the monolayers of oleic acid were in the liquid using the Langmuir film balance. These isotherms were expanded state. Figure 4 compares the typical transmisfound to be in excellent agreement with the isotherms sion spectrum of stearic acid in CC4 and the IRS spectra reported in the l i t e r a t ~ r e .As ~ ~indicated ~~~ by a rapid of transferred LB films of stearic acid. Of particular decrease in surface pressure, the monolayer of stearic acid importance is that the peak at 1710 cm-' (free acid C 4 ruptured at approximately 20 A2/molecule. For oleic acid, stretch in the transmission spectrum) is not evident for the trend was quite different. The surface pressure the transferred monolayer of stearic acid at the surface of increased up to 31 mN/m a t about 27 A2/molecule. As the calcium fluoride IRE. Rather, a new absorption band further compression occurred the surface pressue did not at 1550 cm-' has appeared. It seems clear that the change appreciably. It appears that due to unsaturation, transferred stearic acid molecules have reacted with oleic acid has a higher effective molecular cross-sectional calcium sites a t the surface of the calcium fluoride IRE, area, the liquid expanded state exists over a wider range resulting in a chemisorbed monolayer. As shown in Figure of molecular areas, and oleic acid has a higher compressibility than that observed for stearic acid. G a i n e ~ ~ ~5, the same result was obtained for the transferred LB monolayer of oleic acid. This single band in the vicinity explained the expanded behavior of the oleic acid monoof 1550 cm-l was confirmed by previous IR experiments layer in terms of the effect of unsaturation on packing on various calcium bearing minerals including fluoride considerations. Based on packing in a crystal, oleic acid (CaF2)+5calcite (CaC03),4land apatite (Ca&(PO&, where has an effective area of 22.5 A2/molecule for compact X = F, OH, et^.).^^ The free acid C=O stretching band packing and stearic acid has 20.5 A2/molecule.40It is also at 1701 cm-l reappeared for multilayer deposition, as shown known that the unsaturated C = C bond is more hydrophilic in Figure 4c, indicating that only the fatty acid head group (34)Riegler, J. E.Reu. Sci. Znstrum. 1988,59 (lo),2220-2224. of the first monolayer reacts with the surface calcium sites (35)Maoz, R.; Sagiv, J. J. Colloid Interface Sci. 1984,100, 465-496. of the calcium fluoride IRE. (36)Kamata, T.; Kato, A.; Umemura, J.;Takenaka,T. Langmuir 1987, Molar Absorptivity by Transmission Experiments. 3,1150-1154. (37)Kimura, F.;Umemura, J.; Takenaka, T. Langmuir 1986,2,96Use of the FTIR/IRSadsorption density equation requires 101. that the molar absorptivity, e, be determined a priori. Thus (38)Howarth, V. A,; Petty, M. C.; Davies, G. H.; Yarwood, J. Thin the molar absorptivities for oleic and stearic acids as Solid F i l m 1988,160,483-489. (39)Lundquist, M.In Progress in the Chemistry of Fats and Other Liquids; Holman, R. T., Ed.; Pergamon: New York, 1978;Vol. 16,pp 101-124. (401Wells, H.L.;Zografi, G.; Scrimgeour, C. M.; Gunstone, F. D. Adu. Chem. Ser. 1975,No. 144,135-151.
(41)Young, C. A.; Miller, J. D. To be submitted for publication in Colloids Surf, (42)Gong, W .Q.;Parentich, A.; Little, L. H.; Warren, L. J. Langmuir 1992,8,118-124.
IRS Adsorption Density Equation
Langmuir, Vol. 9,No. 11,1993 3163
(a) Transmission Spechum of Oleic Acid in CC4 Solution
/\
I
g8
(b) Zzsmission Spectrum of Pure Oleic
0 0.0000
0.0002 0.0004 bC (cm.M)
0.0006
Figure7. Effect of solventson absorbance(CHstretchingregion, Table I) of oleic acid and ita sodium salt revealed by the BeerLambert plot from transmission spectra (CaF2 windows).
I
(c) IRS Spectrum of Transferred LB Film One Monolayer of Oleic Acid on CaF2
1800 1750
1700
1650
1600 1550
1500
1450
1400
Table I. Molar Absorptivity Data integrated range concentration (M) (cm-l) stearic acid/CC4 1.05 X lo-" to 1.26 X 1W2 3000-2800 oleic acid/CC& 5.76 X 10-4 to 5.05 X le23032-2802 sodium oleate/DZO 1.80 X 103 to 2.40 X 1W2 3030-2820 linoleic a~id/CC4~9 3.00 X 10-4 to 5.00 X 103 3025-2825
t
(L/(cm2 mol)) 67 600 43 300 42 600 21 800
Wavenumbers (cm-1)
Figure5. Comparison of the FTIR/IRSspectrumof a transferred LB monolayer film of oleic acid with the transmission spectra of pure oleic acid and oleic acid in CC& solution. I
h
E
.
0.60000
0 Stearic Acid
0.00002
0.00004
0.00006
0.00008
bC (cmM)
Figure 6. Effect of unsaturation on absorbance (CH stretching region, Table I) as revealed by the Beer-Lambert plot from transmiasion spectra (CaF2 windows).
revealed by -CH stretching bands were determined by transmission experiments. Figure 6 shows the change in the integrated absorbance with respect to path length and concentration in carbon tetrachloride for stearic and oleic acids using calcium fluoride windows. As shown in Figure 6, the molar absorptivity ( 6 = A/bC) of oleic acid as determined by Beer-Lambert law was found to be smaller than that of stearic acid, which is probably due to the decrease in the number of methylene groups and the effect of the single C = C bond on the neighboring methylene groups. For example, the molar absorptivity of linoleic acid (18C fatty acid with two C = C double bonds) was found to be even lower, ca. 21,800 L/(cm2 mol).43 The effect of solvent polarity on the molar absorptivities of oleic acid and sodium oleate was also examined in order to use the IRS adsorption density equation with confidence. As shown in Figure 7 and Table I, the effect of solvent polarity on the absorbances of oleic acid and its (43)Free, M.; Miller, J. D., Department of Metallurgical Engineering, University of Utah, unpublished data.
sodium salt was not appreciable. The integrated molar absorptivities shown in Table I were determined by a h e a r regression method with the bC term of the Beer-Lambert law as an independent variable and a zero intercept. The coefficient of determination values (R2)are very close to 1 PO.99) and indicate that the Beer-Lambert law is valid under these circumstances. As further confirmationof the molar absorptivity values, it was desired to determine the molar absorptivity of pure oleic acid which was expected to be similar to those values determined from solution transmission measurements. However, even with the thinnest spacer available (15 gm), absorbance of pure oleic acid at the CH2 asymmetric and symmetric stretching frequencies was too strong to allow analysis with the Beer-Lambert equation. Therefore, transmission experiments were made without a spacer between the windows and the windows were squeezed to create a thin film suitable for analysis with the BeerLambert equation. The thickness of the pure oleic acid in this thin film was determined by two different approaches: (1) application of the Beer-Lambert law at the - C = C H stretching frequency and (2) the use of an interference fringe counting procedure. In the case of the Beer-Lambert approach the measurements were made for two different windows, calcium fluoride and germanium. The Beer-Lambert law was applied to the -C=CH band at 3005 cm-l, the absorbance of which is small enough even with different spacers to allow a Beer-Lambert type analysis of pure oleic acid. Figure 8 plots the absorbance (peak height at 3005 cm-l) of oleic acid with respect to bC. These bC values correspond to concentration of oleic acid in carbon tetrachloride, from 2.3 X to 3.97 M (pure oleic acid). cm M corresponds For example the bC value of 9.93 X to the concentration of pure oleic acid, 3.97 M, multiplied by the spacer thickness of 2.5 X cm. The data were found to obey the Beer-Lambert law, as shown in Figure 8 for CaFz windows. This curve was used as a calibration curve for calculating the thickness of pure oleic acid films, in the absence of a spacer, created by squeezing the windows together. By use of a linear regression method, the peak height molar absorptivity of oleic acid at 3005
Jang and Miller
3164 Langmuir, Vol. 9,No. 11, 1993
Maximum Packing Density
2
0.4
Y
i
/
4z 6
1
0
e
" I
I
.-3
E 4
0"C
.-
E* P
:: 0
30
Surface Pressure (mN/m)
Figure 8. Calibration curve constructed from transmission measurementsof pure oleic acid with spacers (CaF2 windows). This relationship was used to determine the thickness (path length) of pure oleic acid films created by squeezing pure oleic acid between the windows without a spacer. Table 11. Comparison of Integrated Molar Absorptivity Values for Pure Oleic Acid A
approach to determine path length ( b ) Beer-Lambert law at 3005 cm-1 w/CaFz Beer-Lambert law at 3005 cm-1 w/Ge interference fringe counting
20
10
bC (cm.M)
c
b (cm) 3.0 X lo"'
(3032-2802 cm-1) 45.9
(L/(cm2 mol)) 38 500
2.4 X lo-'
39.0
40 100
2.5 X lo"'
39.0
39 600
cm-l was estimated to be 57.9 L/(cm mol). With this peak height, the molar absorptivity value, and the spectrum of pure oleic acid, the Beer-Lambert equation allowed for the calculation of the path length for the pure oleic acid film (b = peak height at 3005 cm-'/(57.9 X 3.97). From the calculated path length (b), the concentration of pure oleicacid, 3.97 M, and the measured integrated absorbance ( A ) ,the integrated molar absorptivity (0was determined for the CH- stretching frequencies, again according to the Beer-Lambert law. Thus, the thickness of pure oleic acid films was calculated from FTIR transmission spectra with CaF2 and Ge windows using this calibration curve. The calculated thickness values and the corresponding integrated molar absorptivities are listed in Table 11. For the interference fringe counting procedure, germanium windows were used to obtain better optical contrast. The refractive index of germanium is larger than that of oleic acid by about 2.5 units, while the difference in refractive indices between calcium fluoride and oleic acid is less than 0.1 unit. Accordingly, the reflectivity3Ocan be improved by more than 600 times with germanium windows. In both the reflectivity calculation and the measurement of the interference fringe spacing, 1.5 was used as the index of refraction for oleic acid at the CHstretching frequencies.3~- Table I1compares the results of our integrated molar absorptivity determinations. The three values are in very good agreement. It should be noted that pure oleic acid has a slightly smaller molar absorptivity (ca. 40 00OL/ (cm2mol)) when comparedwith oleic acid in carbon tetrachloride solution (43300 L/(cm2 mol)) or sodium oleate in D2O (42600L/(cm2mol)). The variation would appear to be due to the interaction between the solvent molecules and hydrocarbon chain of oleic acid. These results are in good agreement with the results reported by Kellar et for a slightly different frequency range. (44) Hall, A. C. J. Phys. Chem. 1970, 74,2742-2746. (46) Haller, G. L.; Rice, R. W. J. Phys. Chem. 1970, 74, 4386-4393. (46) Frey, S.; Ta", L. K.Biophys. J. 1991,60, 922-930.
Figure 9. Adsorption density vs surface pressure for LB monolayer of stearic acid determined from the F A isotherm and by the FTIR/IRS adsorptiondensity equationaftertransfer of the LB film to the surface of a CaFz IRE. m
OLEIC ACID
N
E %c 3 8
0
4
2
8
61.
.I2/ 0
0
I Flj
Packing Density
Maxi"
E
.
,
.
10 20 SuIface Pressure (mN/m)
30
Figure 10. Adsorption density vs surface pressure for LB monolayer of oleic acid determined from the P A isotherm and by the FTIR/IRS adsorption density equation after transfer of the LB film to the surface of a CaFz IRE.
Adsorption Density of LB Films. The adsorption density was calculated both from the F A isotherm and from the FTIR/IRS adsorption density equation. In the case of the F A isotherm the adsorption density, I', was calculated from eq 18 (18) where NA is Avogadro's number (6.023X and a is the area/molecule estimated from the P A isotherm at a given surface pressure. After the LB film was deposited at a specified pressure, the adsorption density was calculated from FTIR/IRS adsorption density equation according to eq. 17.
r = lo00AIN € (2d$d,)
(17)
Since the transferred LB fiis are in contact with air rather than water, the molar absorptivities for stearic and oleic acids in the nonpolar solvent, carbon tetrachloride solution were used for the adsorption density calculation. Figures 9 and 10compareboth seta of adsorption density values as a function of the surface pressure for transfer. The adsorption density values determined from FTIR/ IRS spectra are in good agreement with the ones calculated from the surface pressurearea isotherms. For the transferred LB monolayer film of stearic acid, quite good agreement between the adsorption density values evaluated using the F T W I R S adsorption density equation and the values calculated from the n-A isotherm
IRS Adsorption Density Equation was obtained, as can be seen from Figure 9. Based on crystal packing, the calculated adsorption density for a closely packed monolayer of stearic acid is 8.1 X W0mol/ cm2. The transferred stearic acid monolayer at a surface pressure of 27 mN/m (the superliquid state) gave an adsorption density value of 7.4 X 10-lomol/cm2. The FTIR/IRS spectra of the transferred LB monolayer of oleic acid gave adsorption density values close to the ones obtained from the F A isotherm. An adsorption density of 7.4 X mol/cm2 is expected for crystal packing of oleic acid. As shown in Figure 10,an adsorption density of 6.1 X mol/cm2 was obtained for the transferred monolayer filmof oleic acid at 25 mN/m which, when compared to 7.4 X lo-"-'mo1/cm2,indicates that the oleic acid could not be compressed to the more compact state but remained in the liquid expanded state.
Conclusion By comparison of calculated FTIR/IRS adsorption density values from T-A isotherms, prior to transfer, it was clearly shown for both stearic acid and oleic acid at
Langmuir, Vol. 9, No. 11,1993 3166 a calcium fluoride IRE that the FTIR/IRS adsorption density equation is valid and can be used,with confidence, for quantitative analysis of surfactant adsorption phenomena. In addition, the single band in the vicinity of 1550 cm-l, the characteristic peak of chemisorbed carboxylate at surface calcium sites was found from the FTWIRSspectra of LB monolayer films of stearic acid at the CaFz surface. The eame result was found for LB monolayer f i i of oleic acid. The free acid C 4 stretching band reappeared for LB multilayer deposition, indicating that only the fatty acid head group of the first monolayer reacts with the surface calcium sites of the calcium fluoride IRE.
Acknowledgment. This work waa supported by the Office of DOE Basic Sciences Grant No. DE-FG-0393ER14316. The authors expreas their appreciation to J. J. Kellar for many valuable discussions. Thanks is also extended to J. D. Andrade and researchers in Bioengineering who made the Langmuir film balance available for this research.
