Self-aggregation and phase separation of a styryl dye in monolayer at the liquid–air interface and in Langmuir–Blodgett films

The styryl dye 4-[2-[6-(dioctylamino)-2-naphthalenyl]ethenyl]-1-(3-sulfopropyl)-pyridinium inner salt, (Di-8-ANEPPS), was incorporated in mixed Langmuir, (L), monolayers with the cationic surfactant, octadecyltrimethylammonium bromide, (OTMA).

The stability of the monolayers was found to vary with the surface pressure and the amount of the diluent in the mixtures.

The organization of the dye at the interface, deduced from surface pressure–area, πA, and surface potential–area, ΔVA, isotherms, is compared with the spectroscopic data in Langmuir and Langmuir–Blodgett, (LB), films.

UV–visible absorption spectra of the pure dye and of the mixtures in L and LB monolayers show the formation of aggregates or domains of Di-8-ANEPPS within the matrix at high amphiphile content.

On the contrary, successful transfer of the monomeric form of the dye was achieved using small molar fraction of the cationic matrix.


The investigation of new amphiphilic fluorescent molecules has gained attention due to their interesting electric and optical properties.1,2

In this regard the application of molecular assembly techniques has proven to be an excellent alternative to casting and evaporation methods in preparation of ultra-thin films.3–5

In particular, Langmuir, (L), monolayers have been widely used as mimetic systems to investigate the self-organization and the nature of the intermolecular interactions, either at the air–water interface or in films transferred onto solid supports by the Langmuir–Blodgett, (LB), technique.3–7

The present paper reports the spectroscopic characterization of a styryl dye, (Di-8-ANEPPS), embedded in a diluent matrix of a cationic surfactant, (OTMA), in L and LB films.

Like other styryl dyes,8 Di-8-ANEPPS molecules exhibit a strong tendency to aggregate in monolayer and in multilayer systems.

Moreover, these dye molecules are known to respond to changes in the surrounding medium by modifying their electronic properties.9–11

The film-forming features of the mixed monolayers were evaluated by means of surface pressure–area, πA, and surface potential–area, ΔVA, isotherms.

The spectroscopic behaviour of pure and mixed films at the subphase–air interface as well as in LB films was investigated and compared with the electronic spectra of Di-8-ANEPPS in solution.



The dye 4-[2-[6-(dioctylamino)-2-naphthalenyl]ethenyl]-1-(3-sulfopropyl)-pyridinium inner salt, Di-8-ANEPPS, was obtained from Molecular Probes (Eugene, OR) and octadecyltrimethylammonium bromide, OTMA (purity 99.5%), was purchased from Fluka.

The solvents chloroform and methanol were of spectroscopic grade (Aldrich).

Amphiphiles and solvents were used as purchased, without further purification.

Water was obtained from a Milli-RO coupled with a Milli-Q set-up (Millipore), 18.2 MΩ cm resistivity and pH 5.6 at 20 °C.

An aqueous buffer (pH 3.70), prepared using citric acid anhydrous (Fluka, purity 99.5%) and di-sodium hydrogen phosphate dihydrate (Fluka, purity 99.0%), was used as subphase.

Quartz slides (Hellma) were used as solid supports; they were cleaned with chromic acid overnight, then rinsed with water and immersed in an ultrasonic bath with CHCl3 for 10 min.


πA isotherms were obtained with a Lauda Filmwaage FW2 (Lauda, Germany) by discontinuous compression; the compression rate was about 8 Å2 molecule−1 min−1; three π values were recorded for each surface area with a time interval of 30 s between the measurements.

Thirty minutes were allowed for solvent evaporation and monolayer equilibration at the interface prior to compression.

All the spreading isotherms shown in this paper are the average of at least two curves and were obtained at 20 °C using a Haake thermostat with a water circulation bath.

ΔVA measurements were obtained with the method of the ionising electrode by using 241Am electrodes with an apparatus assembled in the Department of Chemistry (Florence) and previously described.12,13

The accuracy of π is ±0.1 mN m−1, of A is ±0.5 Å2 molecule−1 and of ΔV is ±10 mV.

Monolayer relaxation studies, at constant π = 15 mN m−1, were performed using a KSV 3000 film balance (KSV Instruments).

The compression rate of the two movable barriers was 3 mm min−1.

After relaxation, Langmuir films of Di-8-ANEPPS/OTMA mixtures were transferred onto quartz slides with an upstroke rate of 12 mm min−1 and a downstroke rate of 1–2 mm min−1.

The transfer ratio obtained12 is 1.0 ± 0.1 in all cases.

UV–visible absorption spectra were obtained using a Perkin-Elmer Lambda 900 spectrophotometer.

The absorption spectra at the liquid–air interface, at constant areas, were obtained using a Y-shaped fibre optic system connected to the light source and to the photodetector.

The light beam passes through the monolayer and it is reflected via a ‘hard coated’ mirror, placed just below the aqueous surface.14

The fluorescent spectra of the LB films were measured with a Perkin-Elmer LS-50B luminescence spectrometer, using a homemade slide holder especially designed to orient the substrates at 30° and at 60° in the excitation and in the emission mode, respectively.

Semi-empirical calculations were run using the software HyperChem 5.1 (HyperCube, USA)15.

Results and discussion

Surface pressure–area and surface potential–area isotherms

Fig. 1 shows πA and ΔVA isotherms for the monolayers of Di-8-ANEPPS and OTMA and their mixture on citrate buffer (pH 3.70).

The use of buffer solution, instead of pure water subphase, is required in order to avoid the complete dissolution of the lipidic matrix into the bulk phase.

In Table 1 we report the main parameters extracted from the isotherms, i.e. the limiting molecular area A0, the collapse surface pressure πc, the maximum surface compressibility modulus C−1s and the maximum surface potential value, ΔVmax.

A0 is determined extrapolating the linear portion of the isotherm to zero surface pressure, πc corresponds to the maximum change of slope of the πA isotherms and C−1s is defined as:16 The surface pressure–area profile of the dye monolayer on buffer solution shows a liquid-expanded behaviour, as confirmed by the low surface compressibility modulus value (C−1s = 44 mN m−1), with a limiting molecular area of 51 Å2 molecule−1 and a collapse surface pressure of 39 mN m−1 (see Table 1).

Surface potential values follow a linear increase with increasing surface density of the monolayer with a maximum ΔV = 509 mV.

From the analysis of the spreading isotherm it is possible to estimate the orientation of Di-8-ANEPPS molecules at the interface.

The experimental surface potential, ΔVexp, for uncharged or un-ionisable substances, is correlated to the vertical component of the dipole moment, μ, of the amphiphile forming the monolayer17,18 by the relation: where is the intrinsic dipole moment of the molecule, μ is its vertical component and ϑ is the angle between the dipole moment vector and the normal to the interface, A is the molecular area and ε and ε0 are the relative dielectric constant and the permittivity of the free space, respectively.

The dipole moment, , of a molecule can be calculated with respect to the centre of mass of the molecule using the following equation:where ZA is the charge of the nuclear core, RA is the distance between the origin and nucleus A and ri is the distance between the origin and electron i.

On the basis of previous works on analogous compounds in solution19 we considered two possible conformations of Di-8-ANEPPS (see Fig. 2) differing only for the position of the polar head group.

We computed the dipole moment for both conformations obtaining || = 33.9 D in case A, and || = 27.2 D in case B where the polar head group is rotated out-of-plane.

In both calculations the dipole moment vector is directed along the S–N+ bond, in Fig. 2 we also report the inertial axes and the dipole moment vectors computed with respect to the centre of mass of the molecule.

Application of eqn. (2) to the dye monolayer at maximum packing provided an angle ϑ = 89.5° between and the normal to the interface for conformation A. In the case of conformation B, we obtained ϑ = 89.3°.

Fig. 2 shows the molecules oriented according to these tilt angles: for conformation A the entire molecule would lies almost parallel to the water–air interface whereas for conformation B the polar group of the dye resides on the aqueous surface whereas the chromophoric moiety and the alkyl chains point towards the air phase.

Conformation A was excluded on the basis of the molecular areas at maximum packing observed experimentally (see Table 1): in fact for conformation A depicted in Fig. 2 we would obtain a cross section of 360 Å2, a value exceedingly larger than the experimentally observed A0.

On the contrary, the cross-section obtained for case B is 70 Å2 a value of the same order of magnitude of the experimental one, therefore we selected the latter conformation as more likely to occur at water–air interface.

This choice is also supported by NMR20 and polarized fluorescence studies21,22 of styryl dyes in a lipid membranes, where a perpendicular orientation of the molecule with respect to the interfacial plane is observed.

In the case of Di-8-ANEPPS/OTMA mixtures, πA profiles indicate monolayers in a liquid-expanded phase, as confirmed by the low surface compressibility modulus values (40–50 mN m−1).

Moreover, ΔVA curves are shifted towards surface potential values larger than in the case of the pure dye although the profile of the isotherms remains very similar.

We obtain additional information on the interaction of the dye with the lipidic matrix analysing the variation of mean molecular areas and of surface potential values at constant surface pressure as a function of the molar fraction of the dye.

The data are reported in Fig. 3a and in Fig. 3b, respectively.

The mean molecular areas and ΔV values are higher than those calculated for ideal mixed monolayers except in the case of the equimolar ratio.

In the case of mean molecular areas, the displacement is higher at lower surface pressures, while the surface potential data follow the opposite trend.

These results indicate that the Di-8-ANEPPS and OTMA molecules give non-ideal mixtures with repulsive interactions with the exception of the equimolar mixture where an ideal miscibility or a complete immiscibility is inferred.

Unfortunately, the difference in πc between the two components is too small to draw any exhaustive conclusion on this point.

We checked the stability of mixed Di-8-ANEPPS/OTMA Langmuir films by monolayer relaxation studies (see Fig. 4) recording the area loss at constant surface pressure.

The picture shows the results obtained at π = 15 mN m−1 for Di-8-ANEPPS/OTMA mixtures of different composition.

We observed that the relative decrease in At/A0 is more pronounced in the presence of OTMA molecules.

The analysis of the results indicates a partial desorption/dissolution process23,24 of the monolayer into the subphase, further studies are in progress to define the mechanism of this phenomenon.

Absorption spectra at the liquid–air interface

We acquired the absorption spectra of the dye molecules at the subphase–air interface in order to evaluate the influence of the lipidic matrix on the electronic properties of the dye.

As illustrated in Fig. 5a, the spectrum of the pure Di-8-ANEPPS, recorded immediately following the deposition of the dye on the surface (105 Å2 molecule−1), is similar to the solution spectrum in methanol.

The low energy band centred at 500 nm in solution, corresponding to the π–π* transition over the entire chromophore, is shifted to 484 nm at the subphase–air interface, while the band at 322 nm arising from the isolated transition on the pyridinium and naphthyl moieties19,25–27 is slightly visible.

It is interesting to note that, 30 min after spreading, the low energy band completely disappears and a new band centred at 348 nm emerges.

The latter transition could be ascribed to the formation of H-aggregates at the interface, as previously reported by other authors for similar compounds.19,27

After 60 min no more changes were observed in the absorption spectrum; the monolayer was further compressed to higher dye surface density and absorption spectra again acquired.

Surprisingly no changes were evidenced, either in the position of the bands or in the maximum absorbance values.

This indicates that the dye remains in the aggregated domains as the packing of the monolayer increases; the constancy of absorbance is ascribed to the relaxation phenomena at the liquid–air interface, previously described, that causes an overall loss of the chromophore from the monolayer towards the bulk phase.23,24

In the case of the mixed monolayers, the time required for the π–π* transition to vanish increases slightly with increasing molar fraction of Di-8-ANEPPS.

Fig. 5b shows typical absorption spectra acquired after the equilibration time at low surface density for some of the Di-8-ANEPPS/OTMA mixed monolayers examined; the corresponding absorption maxima are reported in Table 2.

In the same table we also report the ratio between absorbance maxima of the monomer (AbsM; λ = 465 nm) and aggregate (AbsA; λ = 348 nm); the data show that in the case of the 4∶1 mixture the ratio is in favour of the monomer form.

For XD8 ≤ 0.5 the absorption spectra show always a band between 320 and 370 nm, which could be ascribed to the convolution of the monomer and aggregate transitions.

On the contrary, in the case of 4∶1 Di-8-ANEPPS/OTMA monolayers, no aggregate formation is evidenced in the spectrum and the low energy band is shifted to lower wavelengths (see Table 2) with respect to the monomeric form in methanol.

Interestingly, the monomer absorption band in the spectrum recorded immediately after monolayer spreading was centred around λ = 465 nm, a value similar to the pure dye monolayer and only slightly smaller than λmax in methanol.

We recall that the monomer absorption in chloroform was observed at λ = 550 nm.28

The shift in the monomer band position might be correlated to a more polar microenvironment sensed by the chromophore in the mixture;10,11,29 this would imply a deeper penetration of the chromophoric part in the water sublayer at the interface.

UV–visible absorption and fluorescence emission spectroscopy of the LB films

We successfully transferred monolayers of the 1∶4, 1∶1 and 4∶1 mixtures onto quartz slides in the Y configuration: Figs. 6a and 6b describes the UV–visible absorption and fluorescence emission spectra of 1 LB monolayer of the selected mixtures.

Absorption spectra are similar to those obtained at the liquid–air interface.

In particular, for XD8 ≤ 0.5 the absorption spectra show a band centred at about 350 nm that we already ascribed to the transition of H-aggregate molecules.

The fluorescence emission spectrum shows a maximum at λem = 424 nm.

Thus, based on the electronic spectra recorded for the LB films, we deduce that the aggregate structure formed in Langmuir monolayers is transferred unaltered onto the substrate.

This statement is supported also by fluorescence emission experiments.

In the case of Di-8-ANEPPS in solution a fluorescence maximum was found at λem = 636 nm in chloroform28 and at λem = 610 nm in methanol.

Fluorescence emission maximum is found at λem = 434 nm in the case of a single LB layer for the mixtures at low Di-8-ANEPPS content (XD8 ≤ 0.5), such displacement in fluorescence maxima is assigned to emission from the H-aggregate27,30 in agreement with the previous deductions.

On the contrary, in the case of the 4∶1 Di-8-ANEPPS/OTMA LB film, the absorption band shifts to 450 nm, and the fluorescence spectrum shows a maximum at λem = 627 nm (see Table 2) confirming the presence of Di-8-ANEPPS monomer.

This is also clearly evidenced by the computed ratio of the absorbance maxima reported in Table 2.

The results obtained for the LB films are in good agreement with the L monolayer with the exception of the 1∶4 mixture where a larger uncertainty is expected due to the extremely low Di-8-ANEPPS content.

Moreover, comparison with a solvatochromic characterization10,11,29 of Di-8-ANEPPS previously performed indicates that the monomer resides in a rather polar environment also when transferred in the LB film.

The absorbance of the electronic spectra recorded for the multilayer systems does increase linearly with the number of layers (data not reported): the outcome is in accordance with the absorbance behaviour at the subphase–air interface.

The above results allowed to propose a hypothesis on the association of the dye in the mixed monolayers.

In the case of XDi-8-ANEPPS ≤ 0.5, OTMA molecules screen the high negative charge density of the dye, allowing the formation of domains of H-aggregates of Di-8-ANEPPS at the interface.

Such domains can be transferred on solid supports without any modification in the structure of the aggregate.

In contrast for XDi-8-ANEPPS > 0.5 the electrostatic repulsion between the dye amphiphiles predominates and hence Di-8-ANEPPS molecules are dispersed in the OTMA matrix mainly in a monomeric form.


Surface characterization of the styryl dye Di-8-ANEPPS is achieved by surface pressure–area and surface potential–area isotherms on buffer subphase as a function of increasing molar fraction of a diluent matrix, OTMA.

The data indicate the presence of repulsive interactions between the two components, except in the case of the equimolar ratio where a complete immiscibility is supposed.

The absorption spectra of the pure dye and of the mixtures at the liquid–air interface show the formation of domains of aggregates within the amphiphilic matrix at low dye content.

In particular, in the case of XDi-8-ANEPPS ≤ 0.5, no monomer species are detected in L and in LB films neither by absorbance spectroscopy nor by fluorescence emission.

Interestingly, a significant population of Di-8-ANEPPS monomers appears in the absorption and emission electronic spectra of the 4∶1 Di-8-ANEPPS/OTMA monolayer, both at water–air interface and in LB film.

In this case, we also observed that the chromophoric part of the dye molecule resides in a rather polar environment.

Since the understanding of the energy migration and transfer in such monolayer and LB systems is essential for the rational design and fabrication of chemical devices based on vectorial energy and electron transport, these results on the aggregation behaviour in an organic matrix appear relevant for the fabrication of microelectronic devices.