Neutron Reflectivity Study of Alkylated Azacrown Ether at the Air

Jul 20, 2009 - Mile End Road, London E1 4NS, U.K., ‡ISIS neutron facility, Science and Technology Facilities Council,. Rutherford Appleton Laborator...
0 downloads 0 Views 1MB Size
pubs.acs.org/Langmuir © 2009 American Chemical Society

Neutron Reflectivity Study of Alkylated Azacrown Ether at the Air-Liquid and the Liquid-Liquid Interfaces A. Zarbakhsh,*,† J. R. P. Webster,‡ and K. Wojciechowski*,§ †

School of Biological & Chemical Sciences, Queen Mary, University of London, Walter Basent Building, Mile End Road, London E1 4NS, U.K., ‡ISIS neutron facility, Science and Technology Facilities Council, Rutherford Appleton Laboratory, Harwell Science and Innovation Campus, Didcot OX11 0QX, U.K., and § Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland Received April 27, 2009. Revised Manuscript Received June 11, 2009 We report the neutron reflectometry study of partially deuterated di-hexadecyl-diaza-18-crown-6 ether (d-ACE-16) at the air-water and the oil-water interfaces. At the air-water interface, the thickness of the monolayer is smaller than that for a fully stretched d-ACE-16 molecule, suggesting a tilt of the alkyl chains with respect to the normal. At the oilwater interface, the same molecules were found to form a more diffuse layer distribution stretching across both sides of the interface. On the oil side, the molecules are densely packed within a thickness of 17 A˚, the hydrophilic part of the molecule with the azacrown ether ring being immersed in the adjacent aqueous side of the interface. The latter consists of a thick 38 A˚ layer comprising staggered, loosely adsorbed d-ACE-16 molecules. With increasing spread amount, the adsorbed layer density increases at the oil side until saturation at ca. 2.25  10-6 mol m-2, above which the layer collapses.

Introduction Transport of chemical species between any two phases inevitably involves passing an interface between them. In extraction, the transport is usually facilitated by some complexing agent (carrier), which transforms the species to be transported into a form more soluble in the receiving phase. When either the carrier or its complex with the transported species is surface active, the main transport pathway may involve transient adsorption at the interface.1,2 Crown and azacrown ethers have been widely used as carriers in numerous extraction-based techniques since their discovery in the late 1960s.3,4 Buffle et al. used the alkylated azacrown ethers in conjunction with fatty acids for transporting heavy metal ions (Cu(II), Cd(II), Pb(II)) against their concentration gradient in so-called permeation liquid membrane (PLM) devices.5 A typical PLM device consists of a hydrophobic membrane separating two aqueous compartments between which the transport of metal ions takes place. The process involves thus two extraction steps, one at each aqueous-membrane interface. The membrane could be either unsupported (bulk organic phase) or supported in the pores of a thin inert polymer support. In either case, the membrane consists of a solution of the carrier (e.g., a mixture of azacrown ether and fatty acid) in a nonpolar solvent. Due to experimental difficulties, the mechanistic studies of interfacial processes in PLM must be performed in model systems by replacing the membrane-solution interface with the airwater or organic-water interface. The exact role played by the *To whom correspondence should be addressed. E-mail addresses: a. [email protected]; [email protected]. (1) Chamupathi, V. G.; Freiser, H. Langmuir 1988, 4, 49–51. (2) Chamupathi, V. G.; Freiser, H. Inorg. Chem. 1989, 28, 1658–1660. (3) Pedersen, C. J. J. Am. Chem. Soc. 1967, 89, 7017–7036. (4) Gokel, G. W.; Leevy, W. M.; Weber, M. E. Chem. Rev. 2004, 104, 2723–2750. (5) Buffle, J.; Parthasarathy, N.; Djane, N. K.; Matthiasson, L. Permeation Liquid Membranes for Field Analysis and Speciation of Trace Compounds in Waters. In In Situ Monitoring of Aquatic Systems: Chemical Analysis and Speciation; Buffle, J., Horvai, G., Eds.; IUPAC Series on Analytical and Physical Chemistry of Environmental Systems; Wiley: Chichester, U.K., 2000; Vol. 6, pp 407-493.

Langmuir 2009, 25(19), 11569–11575

azacrown ether and fatty acid in the transport of Cu(II) ions through the PLM varies at different stages of transport. While fatty acid plays a major role in the diffusive transport of Cu(II) inside the membrane phase,6 the azacrown ether seems to govern the interfacial processes. Thanks to their amphiphilic nature, the alkylated azacrown ethers (ACEs) strongly adsorb at the liquidliquid interface.7 The dynamic interfacial tension and surface rheology studies at toluene-water interface in the absence and presence of Cu(II) and other metal ions showed that the ACE molecules form a densely packed monolayer, which undergoes some structural changes during metal ion complexation.8 The interfacial tension data were fitted to a reorientation model,9 which assumes the coexistence of two different orientations of adsorbed molecules in the monolayer, depending on the total surface concentration. Interestingly, in the more closely packed state, the area per ACE molecule is smaller than that expected from the size of the headgroup of ACE. This observation was rationalized by assuming that the molecules in the adsorbed layer either are stacked or even form a multilayer. Sole interfacial tension, however, does not provide sufficient information on the detailed structure of the adsorbed layer (except for an averaged area per molecule); especially it caries no information on the lateral structure. The latter is available through the reflectivity measurements, for example, X-ray (XR), neutron (NR), or optical (ellipsometry). For example, the group of Ringsdorf studied the monolayers of substituted tetraazacrown ether (cyclam)10 and hexaazacrown ether (hexacyclen)11 using ellipsometry and XR, respectively. (6) Wojciechowski, K.; Kucharek, M.; Buffle, J. J. Membr. Sci. 2008, 314, 152– 162. (7) Wojciechowski, K.; Buffle, J.; Miller, R. Colloids Surf. A. 2005, 261, 49–55. (8) Wojciechowski, K.; Buffle, J.; Miller, R. Colloids Surf. A 2007, 298, 63–71. (9) Fainerman, V. B.; Miller, R.; Wuestneck, R.; Makievski, A. V. J. Phys. Chem. 1996, 100, 7669–7675. (10) Ducharme, D.; Salesse, C.; Leblanc, R. M.; Meller, P.; Mertesdorf, C.; Ringsdorf, H. Langmuir 1993, 9, 2145–2150. (11) Idziak, S. H. J.; Maliszewskyj, N. C.; Vaughan, G. B. M.; Heiney, P. A.; Mertesdorf, C.; Ringsdorf, H.; McCauley, J. P. Jr.; Smith, A. B.III J. Chem. Soc., Chem. Commun. 1992, 98–99.

Published on Web 07/20/2009

DOI: 10.1021/la901485w

11569

Article

Zarbakhsh et al.

The Langmuir films formed by macrocyclic amphiphiles attract attention mostly due to their potential applications in mimicking biological systems,12 in chemical sensing,13-15 and in molecular electronics.16 M€obius and Zaitsev extensively investigated the Langmuir monolayers of water-insoluble crown etherbased dyes.17,18 Tschierske et al. studied Langmuir monolayers of crown ether-appended rod-shaped amphiphiles.19 ACEs were also shown to form Langmuir monolayers via spreading on the water-air interface, but due to some degree of solubility in water, these monolayers were not stable.20 The decane-substituted diazacrown ether derivative (ACE-10), as used in original PLM devices, did not even allow for recording the surface-pressure isotherm. It was also proved by dynamic surface tension and ellipsometry that this derivative dissolves very quickly in the aqueous subphase. Elongation of the alkyl chain length up to hexadecane (ACE-16) improved the situation and allowed for recording the isotherm.20 Nevertheless, still some material was being lost, as a result of which the spread amount could not be considered equal to the adsorbed amount. The surface-pressure isotherms obtained in this way are thus not fully quantitative. For a proper interpretation, one needs to know an independently measured adsorbed amount of ACE in the Langmuir monolayer. In this paper, we describe measurements of adsorbed amount of partially deuterated ACE-16 azacrown ether by using neutron reflectivity (NR), a technique that is very well established in the studies of adsorption at the aqueous-air interface.21-23 In addition to the actual adsorbed amount being measured, the NR provides density profile along the interfacial region, which sets a very good basis for drawing a molecular picture of ACE in the Langmuir monolayer. The structure of the ACE layer adsorbed at the liquid-liquid interface is studied using NR by taking advantage of a thin film entrapment technique of Zarbakhsh et al. described in detail earlier.24 So far the technique has been successfully used to study a bare hexadecane-water interface25 or a hexadecane--water interface with an adsorbed hexadecylphosphorylcholine monolayer,26 and this report is the first describing its use for the macrocyclic complexing agents.

Materials The deuterated azacrown ether (d-ACE-16), N,N0 -di(hexadecyl-d33)-4,13-diaza-18-crown-6 ether (C44H24D66N2O4) with molecular weight of 777.58 g mol-1, was synthesized by (12) Lednev, I. K.; Petty, M. C. Adv. Mater. 1996, 8, 615. (13) Kele, P.; Orbulescu, J.; Calhoun, T. L.; Gawley, R. E.; Leblanc, R. M. Langmuir 2002, 18, 8523. (14) Liu, Y.; Gong, A.; Liu, M.; Xi, F. New J. Chem. 2001, 25, 970. (15) Moore, A. J.; Goldenberg, L. M.; Bryce, M. R.; Petty, M. C.; Moloney, J.; Howard, J. A. K.; Joyce, M. J.; Port, S. N. J. Org. Chem. 2000, 65, 8269. (16) Akutagawa, T.; Kakiuchi, K.; Hasegawa, T.; Nakamura, T.; Christensen, C. A.; Becher, J. Langmuir 2004, 20, 4187. (17) Sergeeva, T. I.; Zaitsev, S. Y.; Tsarkova, M. S.; Gromov, S. P.; Vedernikov, A. I.; Kapichnikova, M. S.; Alfimov, M. V.; Druzhinina, T. S.; Mobius, D. J. Colloid Interface Sci. 2003, 265, 77. (18) Zaitsev, S. Y.; Sergeeva, T. I.; Baryshnikova, E. A.; Gromov, S. P.; Fedorova, O. A.; Alfimov, M. V.; Hacke, S.; Mobius, D. Colloids Surf. A 2002, 198-200, 473. (19) Plehnert, R.; Schroeter, J. A.; Tschierske, C. Langmuir 1998, 14, 5245. (20) Wojciechowski, K.; Grigoriev, D.; Ferdani, R.; Gokel, G. W. Langmuir 2006, 22, 8409–8415. (21) Valkovska, D.; Wilkinson, K. M.; Campbell, R. A.; Bain, C. D.; Wat, R.; Eastoe, J. Langmuir 2003, 19, 5960–5962. (22) Thomas, R. K. Annu. Rev. Phys. Chem. 2004, 55, 391–426. (23) Penfold, J.; Tucker, I.; Thomas, R. K.; Taylor, D. J. F.; Zhang, J.; Zhang, X. L. Langmuir 2007, 23, 3690–3698. (24) Bowers, J.; Zarbakhsh, A.; Webster, J. R. P.; Hutchings, L. R.; Richards, R. W. Langmuir 2001, 17, 140–145. (25) Zarbakhsh, A.; Bowers, J.; Webster, J. R. P. Langmuir 2005, 21, 11596– 11598. (26) Zarbakhsh, A.; Querol, A.; Bowers, J.; Yaseen, M.; Lu, J. R.; Webster, J. R. P. Langmuir 2005, 21, 11704–11709.

11570 DOI: 10.1021/la901485w

BDG Synthesis.27 Null-reflecting water (water with a neutron refractive index equal to that of air) was prepared by mixing approximately 8% (by volume) D2O with ultrapure H2O (18 MΩ cm-1) processed in an Elgastat water purification unit. Hexadecane-h34 was purchased from Aldrich (99%) and hexadecane-d34 was obtained from Cambridge Isotope Laboratories (>98 atom % D). Mixtures of protonated and deuterated solvents were prepared as appropriate for the required neutron refractive index. All solutions and isotopic mixtures were prepared by mass. The hexadecane was purified according to the method of Lunkenheimer and Goebel28 by passing the hexadecane as received through a column packed with alumina seven times.

Experimental Methods Neutron Reflectivity. Neutrons exhibit many optical properties analogous to those of electromagnetic radiation, and the classical laws of optics are applicable. Analogously to optical reflectometry, specular neutron reflectometry is used to determine the (neutron) refractive index profile n(z) in the interfacial region (z is the coordinate normal to the interface). In a medium, the neutron refractive index, n, is commonly written as n ≈ 1 (λ2/(2π))Nb, where λ is the neutron wavelength, N is the number density of nuclei, and b is the average bound coherent scattering length of a nucleus. The multiple Nb is the scattering length density and is simply related to the volume fraction composition of a medium. The sign and magnitude of b for neutrons, in contrast to X-rays, varies in an irregular fashion with atomic number.29 There is also variation of b among isotopes of individual elements, which helps in determining structures of complex samples using the so-called “contrast variation by isotopic substitution” technique. In order to fully utilize the unique contrast variation aspect of neutrons in resolving the details of the interface, deuterated ligands are essential for this work, hence the necessities for the custom synthesis and deuteration of the macrocyclic ligand. Air-Water Interface. In order to resolve the structural conformation of azacrown ether adsorbed layers at the air-water interface, we used a surface-pressure isotherm measurement simultaneously with neutron reflectivity. A spread d-ACE-16 film was prepared using a standard NIMA Langmuir trough. The average area of the trough could be varied in the range 90540 cm2 by two computer-controlled moveable barriers. Prior to use, the trough was cleaned in a standard way. The trough was then filled with water, and the surface was swept and aspirated before the deposition of the azacrown ether film. A solution of 0.5 mg mL-1 of d-ACE-16 was prepared in chloroform and used for spreading of d-ACE-16 films on water. The solvent was allowed to evaporate for about 15 min. All measurements were conducted at T = 298 ( 0.5 K. The trough was situated on a vibration-isolation table covered with a Perspex lid, with two neutron-transparent quartz windows, to prevent any perturbation of the surface. The height of the water surface was initially adjusted to within (1 μm of the same absolute position using a Micro-Epsilon optical displacement sensor prior to neutron alignments runs. This ensures that a constant scale factor is applied to the normalization of all raw reflectivity data. The contrast scheme employed for the air -water experiment was d-ACE-16 with air-null-reflecting water (contrast 1) and the second contrast was air-D2O (contrast 2). A constant pressure strategy was used during the reflectivity measurements. Neutron reflectivity spectra were measured using the time-offlight reflectometer SURF at the pulsed neutron source at the ISIS Facility, Rutherford Appleton Laboratory, Oxfordshire, U.K. In a reflection experiment, the reflectivity, R(Q), is measured as a function of momentum transfer of the neutrons normal to the (27) BDG Synthesis, New Zealand, www.bdg.co.nz. (28) Goebel, A.; Lunkenheimer, K. Langmuir 1997, 13, 369–372. (29) Bacon, G. E.; Noakes, G. R. Neutron Physics; Taylor & Francis: London, 1969.

Langmuir 2009, 25(19), 11569–11575

Zarbakhsh et al.

Article

interface, Q = 4π sin θ/λ, where θ is the grazing angle of incidence of neutrons with wavelength λ. A polychromatic beam of neutrons with wavelengths in the range 0.5 < λ < 6.5 A˚ was reflected from the air-water interface at an incidence angle θ = 1.5 and detected using a single 3He detector. The analysis of neutron reflectivity data for the air-water interface involved determining the Nb(z) from the measured reflectivity spectrum R(Q) by the optical matrix method30,31 (exact) based on a layer model (each layer with a characteristic thickness, composition, and interlayer roughness). Oil-Water Interface. The alkylated azacrown ether in real devices plays an active role at the interface between two immiscible phases. Analysis of the neutron reflectivity data for the adsorbed layer of ACE at the hexadecane-aqueous interface will provide further insight into its conformation at the buried oilwater interface. The deutrated ligand was again used with both the aqueous subphase and the oil (hexadecane) top phase, contrastmatched to silicon (contrast 3). The second contrast, the silicon contrast-matched oil, was used with D2O forming the aqueous bulk phase (contrast 4). The experimental procedure developed by us recently for the study of amphiphiles at a fluid-fluid interface32 was deployed. This technique involves forming a thin, uniform oil layer on an oleophilic polished (111) surface of a silicon block by the spin coating technique. In this experiment, a water subphase was placed in the trough forming a meniscus, and its surface was swept. Before assembly of the cell, the level of the water surface was lowered, by syringing through greaseless plug valves, until it was just lower than the O-ring seal. A known spread amount of dACE-16 was deposited on the bulk aqueous surface from a chloroform solution. A sufficient time was allowed for the solvent to evaporate. Oil was initially spread on the oleophilic silicon surface and spun on the block until distinct Newton rings are observed (approximately 20 s at 2500 rpm). This procedure produces a thin (∼2000-5000 A˚) and uniform oil layer on the substrate’s surface. The thin oil layer was then frozen and maintained in this state using a thermoelectric cooler. The oil layer was finally trapped between the silicon substrate and bulk of the second fluid subphase. Care was taken to ensure that no air was trapped in the cell. The thermoelectric cooler was turned off, and the oil was annealed to T = 298 ( 0.5 K for melting. The neutron reflectivity spectra were then measured using the time-of-flight reflectometer SURF at the pulsed neutron source at the ISIS Facility, Rutherford Appleton Laboratory, Oxfordshire, U.K. For a sufficiently thick oil film (approximately a few micrometers), the reflectivity, R, measured is given by R ¼ Rsil -oil þ

ARoil -water ð1 -Rsil -oil Þ2 1 -ARsil -oil Roil -water

ð1Þ

Rsil-oil, the reflectivity from the silicon-hexadecane interface, and Roil-water, the reflectivity from the hexadecane-aqueous solution interface, are calculated using the optical-matrix method. The attenuation factor, A, for the loss of intensity upon the beam crossing the oil film twice is   -2χðλÞdoil A ¼ exp ð2Þ sin θoil where doil is the thickness of the oil film and θoil is the incidence angle of the neutron beam in the oil phase. The linear absorption coefficient χ(λ), which embraces all loss processes and is assumed to be dominated by the oil layer, was determined from transmission measurements. A comprehensive description of the data analysis is given elsewhere.33 Reflectivity measurements were (30) Born, M.; Wolf, E. Principles of Optics; Pergamon: Oxford, U.K., 1980. (31) Heavens, O. S. Optical Properties of Thin Films; Butterworth: London, 1955. (32) Zarbakhsh, A.; Webster, J. R. P; Eames, J. Langmuir 2009, 25, 3953. (33) Zarbakhsh, A.; Querol, A.; Bowers, J.; Webster, J. R. P. Faraday Discuss. 2005, 129, 155–167.

Langmuir 2009, 25(19), 11569–11575

Figure 1. Reflectivity data for contrast 1 (air-null-reflecting water) interface for a solution of d-ACE-16 measured at 1.5. The solid lines correspond to a single layer model with film thickness 21.5 ( 0.5 A˚. The data are shifted up by a multiple of 10 (10, 100, etc.) for the purpose of clarity. conducted on the reflectometer SURF by employing nominal grazing incidence angles of 0.3, 0.4, and 1.4. The actual angles of incidence are determined by performing detector angle scans in reflection geometry once the sample had been aligned using a series of height and detector angle scans. The incidence angle was varied by use of a supermirror in order to maintain horizontal sample geometry. The collimating slit settings are varied with incidence angle in order to measure all reflectivities with nearconstant angular resolution, δθ/θ ≈ 8%, and to ensure underillumination of the sample. The measured reflectivity profiles are normalized relative to the incidence beam monitor spectrum and corrected for detector efficiency. The data are subsequently corrected for wavelength-dependent transmission through the silicon substrate. Normalization by the monitor spectrum also accounts for the supermirror efficiency. The data for each incidence angle are truncated at an appropriate wavelength cutoff, which was determined by the angle of incidence of the neutron beam at the supermirror. The reflectivity data are placed on an absolute scale by obtaining scale factors for each measured angle by calibration against the silicon-D2O reflectivity, which in turn was normalized using total reflection. In the presented data, only the information-rich reflectivity data at the highest angles are shown. However, model fitting has been performed for all angles of incidence. The value of χ(λ) was determined from transmission through the samples of hexadecane with the different H/D composition was measured in quartz spectrophotometric cuvettes (2 mm optical path, Hellma), and these measurements allowed χ to be determined for the oil phase using the BeerLambert law: T = exp(-χl), where l is the neutron path length. The transmission is wavelength-dependent; hence all the corrections for the oil-water measurements are carried out in λ. The reflectivity data for the oil interface are therefore shown as a function of λ. The measured reflectivity spectra R(Q) were analyzed using the optical matrix method, as described in the air-water section.

Results Air-Water. Structural conformations of the d-ACE-16 monolayer spread at the air-water interface were studied simultaneously using neutron reflectometry and surface-pressure isotherm (surface pressure/area) measurement using a standard NIMA trough. The reflectivity data for the contrast 1 (air/ contrast-matched air aqueous solution) and the contrast 2 (air/ D2O) are shown in Figures 1 and 2, respectively. The reflectivity data were collected for a series of points along the surface pressure DOI: 10.1021/la901485w

11571

Article

Zarbakhsh et al.

Figure 2. Reflectivity data for contrast 2 (air-D2O) interface for a solution of d-ACE-16 measured at 1.5. The solid lines correspond to a single layer model with film thickness 21.5 ( 0.5 A˚. The data are shifted up by a multiple of 10 (10, 100, etc.) for the purpose of clarity.

(π) isotherm curve, from 11 to 30 mN m-1 (Figure 3). A constant pressure was maintained during each reflectivity profile measured by the automatic adjustment of the Langmuir trough area (which is larger than the neutron beam footprint). It can be shown26 that the R þ¥ adsorbed amount can be written simply as Γn ¼ -¥ NbðzÞ dz with Νb(z) the film’s scattering length density. Thus, in null-reflecting water, for a slab with constant scattering length density, Nbslab, and thickness, dslab, the area per molecule, Apm, is defined by P i bi Apm ¼ ð3Þ dslab Nbslab The numerator in eq 3 is the molecular scattering length: P the coherent molecular scattering length for d-ACE-16 is ibi = 685.1 fm. The contrast 1 data (Figure 1) is fitted to a single layer model of 21.5 ( 0.5 A˚ with scattering length density values, Nb, changing from 2.63  10-6 for the smallest π value of 11 mN m-1 to 3.22  10-6 A-2 for the largest π value of 30 mN m-1. The surface pressure as a function of area per molecule calculated using eq 3 for the contrast 1 is shown in Figure 3a. This data indicates that the adsorbed amount increases (area per molecule decreases) as π increases. However, the experimentally determined values for the area per molecule are larger than these estimated from the area of the trough. This variation, being linear in nature, is shown graphically in Figure 3b. The volume fractions of the layer increase from 0.50 for the lowest π to 0.61 for the highest π. The contrast 2 profiles (Figure 2) were fitted to a single-layer model (21.5 ( 0.5 A˚) with interlayer roughness of 2 A˚. The fits are shown by the solid lines in Figure 2. The Nb of this layer is below that for D2O (6.35  10-6 A˚-2) and that for pure d-ACE-16 (5.30  10-6 A˚-2). The scattering length density profiles for the fits are shown in Figure 4. The scattering length density profiles indicate the ligand residing mainly at the surface of water and the scattering length density of the layer increases as the π increases. As the pressure increases, either the ligand or D2O content of this layer increases. Oil-Water. In this experiment, again two contrast regimes were deployed. In contrast 3, both the oil and the aqueous subphase scattering length density are matched to the silicon. In contrast 4, the oil scattering length density is matched to the 11572 DOI: 10.1021/la901485w

Figure 3. Analyses of reflectivity data measured for contrast 1: (a) π vs area per molecule for d-ACE-16 calculated from the neutron reflectivity data; (b) the area per molecule calculated from the area of the trough and the spread amount as a function of the area per molecule obtained from the fits to the neutron profiles.

Figure 4. Scattering length density profiles used to model the contrast 2 (for d-ACE-16 at the air-D2O) corresponding to the solid lines in Figure 2, from 11 (the lowest curve) to 30 mN m-1 (the highest).

silicon, but D2O was chosen as the aqueous subphase. The beam transmission through the top oil phase is a function of the neutron wavelength (i.e., energy). As a result, all the data corrections were carried out in wavelength, and the data are shown in Langmuir 2009, 25(19), 11569–11575

Zarbakhsh et al.

Article

Figure 5. Reflectivity data measured at 1.4 for contrast 3 (Sihexadecane scattering length density matched to the Si-aqueous solution with scattering length density matched to Si), for a series of spread amounts for the d-ACE-16. The solid lines correspond to a single-layer model with film thickness 29.0 ( 2.0 A˚. The data are shifted up by a multiple of 10 for the purpose of clarity.

Figure 7. Reflectivity data for contrast 4 (Si-hexadecane scattering length density matched to the Si-D2O) for a series of d-ACE16 spread amounts measured at 1.4. The solid lines correspond to a two-layer model (17 A˚ at the hexadecane side of the interface and 38 A˚ at the aqueous solution side of the interface). The data are shifted up by a multiple of 10 for the purpose of clarity.

Figure 6. The fitted adsorbed amount deduced from contrast 3 for the d-ACE-16 at the oil-water interface is plotted as a function of spread amount. The dashed line shows the ideal theoretical dependence (slope 1).

Figure 8. Scattering length density profiles used to model the contrast 4 data corresponding to the solid lines in Figure 7.

wavelength as well. The angle of incidence was 1.400 ( 0.001. The incident angle was aligned prior to each measurement. The reflectivity data for contrast 3 is shown in Figure 5. The normalized reflectivity data shows a steady increase in the slope from the lowest spread amount (Γ=1.5  10-6 mol m-2) and then a large reduction in the slope for the highest Γ. This change in slope indicates an increase of material at the interface followed by depletion as the highest spread amount is approached. The contrast 3 data are fitted to a single layer model with a layer thickness of 29 ( 2 A˚. The fits are shown by the solid lines in Figure 5. The fitted adsorbed amount deduced from this contrast is plotted as a function of spread amount and is shown in Figure 6. This again supports the conclusion made above regarding data prior to any fitting, namely, that the adsorbed amount at the oilwater interface increases reaching a maximum at the spread amount ∼2.25  10-6 mol m-2 and then falls as the spread amount increases further. The layer thickness of 29.0 ( 2.0 A˚ obtained at the oil-water interface is slightly larger than 21.5 ( 0.5 A˚ at the air-water interface. This suggests that at the oilwater interface the hydrophobic segments of the azacrown ether are now solvated and the interface is slightly broader due to a Langmuir 2009, 25(19), 11569–11575

reduction in surface tension compared with that at the air-water interface. The contrast 4 (oil contrast matched Si/D2O) reflectivity profiles are shown in Figure 7. The data were found to be best represented by a two-layer model consisting of a 17 A˚ layer on the oil side of the interface and a layer of 38 A˚ on the aqueous side of the interface. The fits to data are shown by the solid lines in Figure 7. The scattering length density profiles for these fits are shown in Figure 8. The results for the spread amounts 1.50  10-6 to 2.25  10-6 mol m-2 suggest a very dense adsorbed layer at the oil side of the interface with small penetration into the aqueous side of the interface (layer 2). As the spread amount increases, the composition on the oil side of the interface remains almost constant, while the d-ACE-16 content of layer 2 increases. At the value of spread amount of 6.01  10-6 mol m-2, the azacrown ether content on the oil side falls drastically, and its content on the aqueous side increases. This continues until the highest spread amount value of 1.35  10-5 mol m-2, when the molecules are also depleted on the aqueous side (scattering increases toward that of D2O), indicative of the desorption from the interface into the solution. These agree well with results obtained from the contrast 3 data and the adsorbed amount deduced from that contrast. DOI: 10.1021/la901485w

11573

Article

Zarbakhsh et al.

Discussion Air-Water. The combined contrasts 1 and 2 results for the Langmuir films of d-ACE-16 at the air-water interface are in good general agreement with those previously reported for the protonated azacrown ether (ACE-16).20 The overall shape of the two isotherms is very similar, with no clear phase transition below π = 30 mN m-1. In the surface-pressure isotherm for the protonated derivative supported by BAM images, compressing the monolayer above this value resulted in significant increase of the hysteresis in the isotherm upon expansion.20 We argued previously that this hysteresis is due to enhanced dissolution of the monolayer, and to avoid additional artifacts, in this study the compression was stopped after reaching π=30 mN m-1. It should be stressed, however, that even below this threshold surface pressure some dissolution of the monolayer material was observed by taking use of dynamic surface tension and ellipsometry.20 Unfortunately, in the absence of an independent measure of adsorbed amount, this loss of material could not be taken into account quantitatively. As a result, the surface-pressure isotherm measured uniquely with the Langmuir balance (as previously for the protonated ACE-16) is shifted horizontally by an unknown value to the left, that is, toward the smaller values of area per molecule. Simultaneous measurement of surface-pressure isotherm and adsorbed amount by combining the Langmuir trough and neutron reflectivity techniques enables one to properly correct for this bias. In this context, the difference between the area per molecule calculated from the NR data and the one calculated from the area of the trough (Figure 3b) is not surprising. The slope of the dependence of the adsorbed amount from NR vs that from the trough geometry exceeds 1, suggesting that the dissolution is enhanced with increasing compression. At low compression, both areas per molecule are almost identical, while close to the maximum value of surface pressure, π = 30 mN m-1, they deviate significantly. In perfect agreement with previous results, the solubility of the film of azacrown ether increases with compression of the monolayer. By taking into account this correction, the phase transition observed at π=32 mN m-1 previously at area per molecule of ca. 70 A˚2, in reality would take place at ca. 85 A˚2 or even higher (Figure 3a). Extrapolating the results of Figure 3b to even higher compressions, the area per molecule in the monolayer during its collapse would be much closer to 40 A˚2, instead of unrealistic 20 A˚2 observed previously20 without the correction for monolayer loss. Simple geometric considerations allow for estimating the minimum area per d-ACE-16 molecule as being equal to the area of the azacrown ring, assuming that the molecule is adsorbed with the azacrown ring parallel to the interface plane, and the alkyl chains perpendicular to the latter. In such geometry, the area per molecule is independent of the length of alkyl chains and cannot be smaller than twice the area per single alkyl chain, that is, ca. 40 A˚2. The present estimations, despite being much more realistic, are still biased by the fact that the measurement of the NR curve at each point on the isotherm requires about 1 h, during which the monolayer dissolution continues. Since the Langmuir trough operated in a constant surface pressure mode, the area per molecule defined by the compressing bars was continuously adjusted during the NR runs. This is signaled in Figure 3b with vertical error bars, which indicate the extent of the area reduction that was necessary to maintain constant surface pressure. The fact that these error bars increase with an increase of the surface pressure further confirms our observation that the monolayer dissolution increases with increasing the d-ACE-16 film compression. The NR results clearly confirm that d-ACE-16 forms monolayers with the alkyl chains extending in the direction normal to the interface. All the NR curves acquired along the surfacepressure isotherm could be reasonably fitted with a unique value 11574 DOI: 10.1021/la901485w

for the monolayer thickness (21.5 ( 0.5 A˚). An increase of the volume fraction with increase of the surface pressure is consistent with increasing surface coverage in the monolayer. The monolayer thickness obtained from a simultaneous fit of the contrast 1 and 2 data (21.5 ( 0.5 A˚) is consistent with that expected for the d-ACE-16 molecules laying flat on the surface with the azacrown ring parallel to the surface plane and the alkyl chains perpendicular to it. The monolayer thickness is slightly smaller than a sum of the thickness of the azacrown ether ring and that of fully stretched all-trans hexadecane chains (25.7 A˚). The alkyl chains in the monolayer at the air-water interface are thus on average inclined with respect to the normal by ca. 33, assuming that the azacrown ether ring is lying flat on the water surface. Oil-Water. Adsorption of an azacrown ether (ACE-10) at water-oil interfaces was previously studied with help of interfacial tension measurements.7 The results were interpreted in the context of a two-state adsorption model (“reorientation model”), assuming that at low surface coverage (low surface pressures) the molecules are adsorbed in a relaxed state, with area per molecule of 322 A˚2, while at high surface pressure the adsorbed molecules rearrange to take only 43 A˚2/molecule. Since the molecules in this compressed state are probably adsorbed with the azacrown ether parallel to the surface plane, increasing length of the alkyl chains would not affect the area per molecule in this high surface coverage orientation. The results of interfacial tension investigations for ACE-16 at the toluene-water interface indeed confirm that the area per molecule in the high surface coverage state is the same as that reported in ref 20 for ACE-10. On the other hand, in the expanded state, one dACE-16 molecule requires 525 A˚2, due to larger space taken by the hexadecane chains lying flat on the surface. In this low surface coverage state, the adsorbed layer is very thin, and the monolayer might not be detectable using NR. For this reason, all the reflectivity curves discussed in this paper refer to the high surface coverage situation. Even for the lowest spread amount (Γ=8.0  10-7 mol m-2), the high surface coverage orientation is predominant (Figure 9). Given the high concentration of azacrown ether in PLM membranes (usually >10-2 M, corresponding to the fully saturated monolayer, based on the interfacial tension results for ACE-107), this high surface coverage orientation also dominates at the PLM membrane-solution interface. On the other hand, the presence of low surface coverage orientation of d-ACE-16, forming a rather thin layer, cannot easily be modeled. Despite the fact that interfacial tension does not provide any direct information on the arrangement of molecules in the direction normal to the interface, the low value of area per molecule in the compressed state was assumed previously as an indication of formation of a multilayer of azacrown ether at the oil-water interface. This hypothesis is fully supported by the present NR data for adsorption of d-ACE-16 at the hexadecane-water interface. The data for contrast 4 clearly shows two distinct layers of azacrown ether at both sides of the interface. The organic side of the interface is more compact (17 A˚), with the thickness smaller than the one for a fully extended d-ACE-16 molecule (25.7 A˚), suggesting that the hydrophilic part of the molecule (azacrown ether part) is probably immersed in the aqueous phase. The scattering length profile for this layer at low spread amounts is very close to that of dACE-16 (5.30  10-6 A˚-2), suggesting a rather high surface coverage. The aqueous part of the adsorbed layer is much more diffuse, in agreement with several other studies at liquid-liquid interfaces.26,34,35 The overall thickness of this layer (38 A˚) exceeds (34) Lu, J. R.; Thomas, R. K.; Binks, B. P.; Fletcher, P. D. I.; Penfold, J. J. Phys. Chem. 1995, 99, 4113–4123. (35) Lu, J. R.; Fragneto, G.; Thomas, R. K.; Binks, B. P.; Fletcher, P. D. I.; Penfold, J. Colloids Surf. A. 1998, 135, 277–281.

Langmuir 2009, 25(19), 11569–11575

Zarbakhsh et al.

Article

and the low surface coverage (flat) orientation prevails, giving rise to no significant NR signal.

Conclusions

Figure 9. Surface coverage for two orientations of ACE-10 calculated from the interfacial tension data at the toluene-water interface from ref 7.

that of a single d-ACE-16 molecule, and the scattering length densities for this layer in Figure 8 are approaching the value of D2O (6.35  10-6 A˚-2). The layer may thus be composed of staggered (or loosely aggregated) molecules weakly bound to the interface. Initially, increase of the spread amount of d-ACE-16 results in densification of the oil side of the monolayer. However, when the latter reaches saturation, the molecules are pushed toward the aqueous part of the interface, resulting in a decrease of its scattering length density (part of the D2O molecules are replaced by d-ACE-16 with lower scattering length density). With further increase of the spread amount, azacrown ether is expelled from the interface and dissolves into the bulk aqueous phase. This phenomenon can be regarded as an interfacial equivalent of the spread monolayers collapse observed at the water surface. The same behavior is apparent from the analysis of the contrast 4 data (Figure 6). Once the Langmuir monolayer is disrupted and its material gets dissolved into the adjacent hexadecane and aqueous phases, its concentration is so low that the resulting Gibbs monolayer (in equilibrium with the bulk solutions) is very loose,

Langmuir 2009, 25(19), 11569–11575

Spread monolayers of partially deuterated N,N0 -di(hexadecyl-d33)4,13-diaza-18-crown-6 ether (d-ACE-16) were analyzed using neutron reflectivity at air-water and hexadecane-water interfaces. By combination of the surface-pressure-area isotherm with neutron reflectivity measurements, the extent of the d-ACE-16 monolayer dissolution at the air-water interface, observed qualitatively previously for the protonated azacrown ether ACE-16, was quantitatively addressed. The neutron data provided the true area per molecule at each point of the surface-pressure isotherm. The corrected isotherm provides much more realistic values for the area per molecule in the isotherm. The thickness of the monolayer is slightly smaller than that expected for a fully stretched d-ACE-16 molecule, implicating a perpendicular to the surface orientation of the azacrown ring and a slight tilt of the alkyl chains with respect to the normal (ca. 33). At the oil-water interface the same molecules form a more diffuse layer distributed at both sides of the interface. On the oil side, the molecules are densely packed within a thickness of 17 A˚, possibly with the hydrophilic part of the molecule and the azacrown ether ring being immersed in the adjacent aqueous side of the interface. The latter consists of a thick 38 A˚ dilute layer comprising staggered, loosely adsorbed (or aggregated) d-ACE16 molecules. With increasing spread amount, the monolayer density increases at the oil side until saturation at ca. 1.2  10-6 mol m-2, above which the monolayer material is expelled to the aqueous side of the interface and ultimately dissolved into the adjacent bulk phases. Acknowledgment. This work was financially supported by the Polish Ministry of Science and Higher Education (Grant No. N N204 236934). The authors wish to thank the STFC for allocation of beamtime at ISIS and for provision of consumables and subsistence.

DOI: 10.1021/la901485w

11575