stigateded by theptedGENT,”tions used toiumfollowedbotactice revealedocalizedorescencen length isall analysisy phase in

Journal of Colloid and Interface Science 262 (2003) 579–587www.elsevier.com/locate/jcis

Characterization of the solvation environment providedby dilute aqueous solutions of novel siloxane polysoaps using

the fluorescence probe pyrene

Siddharth Pandey,a,∗ Rebecca A. Redden,a Ashley E. Hendricks,a Kristin A. Fletcher,a

and Christopher P. Palmerb

a Department of Chemistry, New Mexico Institute of Mining and Technology, Socorro, NM 87801, USAb Department of Chemistry, University of Montana—Missoula, Missoula, MT 59812, USA

Received 30 July 2002; accepted 31 January 2003

Abstract

Solubilization environment afforded by several of the novel allyl glycidyl ether-modified methylhydrosiloxane polymers are inveusing a common polycyclic aromatic hydrocarbon fluorescence probe, pyrene. The backbone of the polymer has been modifiaddition of an alkyl chain of varying length (either C8, C12, or C18) and to differing degrees of substitution. The nomenclature adofor the purposes of these studies is as follows: “AGENT” represents the backbone polymer with no alkyl substitution, and “OA“DAGENT,” and “SAGENT” are substituted withn-octyl, n-dodecyl, andn-octadecyl, respectively. The percentage of alkyl substituis designated as 10, 15, and 20%. The pyrene polarity scale (defined as the ratio of the intensity of peak I to peak III) wadetermine the relative dipolarity of the cybotactic region provided by∼1 w/w% aqueous polymer solutions compared to 10 mM soddodecylsulfate (SDS) micellar solution. Results indicate that 10–15% DAGENT afforded the most hydrophobic solubilization site,by 15% OAGENT and 15% SAGENT. In addition, as the degree of alkyl substitution of DAGENT increased from 10 to 20%, the cyregion appeared to become more hydrophobic. Furthermore, a deeper investigation into the relative size of the solubilization sitthat all alkyl-substituted polymers promoted excimer formation at relatively low pyrene concentrations, indicating the possibility of lconcentration enhancement within the solvation pockets and/or compartmentalization of the solute molecules. The pyrene fluexcitation data strongly indicates ground-state heterogeneity that is most prominent in AGENT and decreases as the alkyl chaiincreased. This provides a relative sense of the size and shape of the solvation pockets afforded by each polymer solution. An overof the collected data indicated that these alkyl-substituted polymers may provide a more selective and efficient pseudostationarelectrokinetic chromatography with better solvation capacity for hydrophobic compounds compared to SDS. 2003 Elsevier Science (USA). All rights reserved.

Keywords: Polysoaps; Polysiloxane; Solvation environments; Polarity probe; Pyrene; Solvatochromic probe; Fluorescence; Excimer

oly-ue-m-ilarcanitybeen

re-

veti-

neticad-

ica-ts,

queoreandturend

1. Introduction

A large number of studies have been dedicated to pmers that can form hydrophobic microdomains in aqous solution [1–12]. Polymerized surfactants and ionic aphiphilic copolymers in many cases have properties simto those of conventional surfactant micelles in that theysolubilize compounds with otherwise low water solubil[13,14]. These polymers have seen application or have

* Corresponding author.E-mail addresses: pandey@nmt.edu (S. Pandey),

palmerc@selway.umt.edu (C.P. Palmer).

0021-9797/03/$ – see front matter 2003 Elsevier Science (USA). All rights rdoi:10.1016/S0021-9797(03)00135-8

proposed for application in drug delivery, detergency, oilcovery, and analytical chemistry [1–18].

In recent years, amphiphilic polymers of this type hareceived significant attention for application in analycal chemistry as pseudo-stationary phases in electrokichromatography [13,14]. The polymers provide severalvantages relative to conventional micelles for this appltion, including stability in the presence of organic solvenapplicability with mass spectrometric detection, and uniand varied chemical separation selectivity. In order to mfully realize the advantages of these materials for thisother applications, it is necessary to understand the naof the solvation environment afforded by the polymers, ahow this is affected by polymer structure.

eserved.

580 S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587

veryol-thet re-lec-e ofrtiesrobeenceecaygenis-

abil-vepo-on-ane

sur-ty asdi-facestad-owekes

tudylox-. Forally

ies.

E isTilox-edf the

tep-15%,

ess-etic

d ineseys,

lva-entdedupra)hro-it aentf theand

sedndT

ion;ti-

sedoly-fullyffer

ter-Inc.,waslassru-

rcis-enelassriateded

Solvatochromic fluorescence probes are known to beeffective in providing physicochemical properties of the subilizing site immediately surrounding the probe (i.e.,cybotactic region) [19–26]. Because the emission of lighquires the prior generation of an excited species, the moular environment can strongly influence the fluorescencmolecules. Any changes in the physicochemical propeof the cybotactic region surrounding the fluorescence pmay influence excitation and emission spectra, fluorescquantum yields, excited-state fluorescence intensity dparameters (e.g., lifetimes), and relaxation processes. Ineral, they can be interpreted in terms of dipolarity, microvcosity, presence of other solute(s), hydrogen-bondingity, etc. of the solubilizing medium. In this paper, we haeffectively utilized a common and popular fluorescencelarity probe, pyrene, to characterize the solvation envirment provided by dilute aqueous solutions of novel siloxpolysoaps.

Siloxanes have been widely used for many years asfactants because of their high surface energy and activiwell as their stability toward heat, chemicals, and UV raation. The specific surface free energy of siloxanes (surtension of 21 mN/m) is significantly lower than that of mohydrocarbons, which means not only that they will besorbed at hydrocarbon surfaces but also that siloxanes lthe surface tension of their solutions. This property masiloxanes very useful as surfactants [12].

The chemical structures of the polymers used in this sare presented in Fig. 1. The nomenclature of the polysianes used in the present studies is devised as followsionic siloxanes, the first three letters of the name gene

Fig. 1. Chemical structures of the polymers used in the present stud

-

r

r

refer to the starting material. In polymers used here, AGan allyl glycidyl ether modified methylhydro siloxane. Nstands for the ionic group that has been added to the sane, i.e.,N -methyl taurine. If an alkyl chain has been addto the siloxane backbone this appears at the beginning oname, e.g., O for C8, D for C12, and S for C18. The amounof the backbone that is modified with the alkyl chain is rresented as the total percent of the backbone (i.e., 10%,20%).

The siloxane polymers in Fig. 1 have been succfully employed as pseudo-stationary phases in electrokinchromatography, and their selectivity has been studiesome detail by chromatographic methods [27–29]. Thamphiphilic polymers can aggregate in a variety of waand that change in the aggregation will affect the sotion environment that they provide. The goal of the presstudy is to characterize the solvation environment afforby these polymers by fluorescence spectroscopy (vide sand, where possible, to correlate these results with the cmatographic results obtained previously. This will permmore complete characterization of the solvation environmprovided by the polymers, and a greater understanding oeffects of polymer structure on solvation environmentchemical separation selectivity.

2. Materials and methods

2.1. Chemicals and reagents

Synthetic procedure for obtaining the polymers uin the present studies, AGENT, OAGENT, DAGENT, aSAGENT, are reported in the literature [27–29]. (AGENrepresents the backbone polymer with no alkyl substitutOAGENT, DAGENT, and SAGENT are respectively substuted withn-octyl, n-dodecyl, andn-octadecyl.)

Sodium dodecylsulfate (SDS, 99.7%) was purchafrom Fisher Scientific and was used as received. All the pmers as well as SDS solutions were prepared by caredissolving the appropriate amount in pH 10 phosphate buand doubly deionized water, respectively, and filtering afwards. Pyrene (99%) was obtained from AccuStandard,and was used as received. The stock solution of pyreneprepared by dissolving in ethanol in precleaned amber gvials and stored at∼4 ◦C. Doubly distilled deionized watewas obtained from a Millipore, Milli-Q Academic water prification system.

2.2. Methods

Samples for spectroscopic studies were prepared exeing extreme care as follows: appropriate aliquots of pyrstock solutions were transferred into appropriate size gcentrifuge tubes and evaporated under argon. Appropamount of the desired polymer or SDS solutions were ad

S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587 581

al-t).ifu-samneab-

ett–the

PTIec-ci-ave-ctedounddata

s isall,

os-cur

es-ro-

fouroth-

o a

ar-micklyndForrityim-

ion,prob

by

ex-

-.0,rs asphy,

alantsedlary

%.0,m-slit

e 2NT,u-re-

ffer-mi-eachlu-

ntlowdentys-

n-o-d

olved%

to the tube, mixed thoroughly by a vortex mixture, andlowed to equilibrate for sufficient time (usually overnighSamples solutions were finally centrifuged for 1 h. Centrgation removed the suspended excess pyrene from theple solutions. Analytical concentrations of all final pyresample solutions were checked by measuring electronicsorbance of each sample solution using a Agilent HewlPackard 8453 Photo-Diode Array spectrophotometer inusual manner with a 1 cm2 quartz cuvette at 25◦C. Steady-state emission experiments were performed with aQuantaMaster Model C-60/2000 L-format scanning sptrofluorometer with a 75 W xenon arc lamp as the extation source and single-grating monochromators as wlength selection devices. All emission spectra were correfor emission monochromator response and were backgrsubtracted using appropriate blanks. All fluorescencewere measured in a 1 cm2 quartz cuvette at 25◦C.

3. Results and discussion

3.1. Pyrene II/IIII

A significant advantage of using fluorescent probethat the concentration of the probe used can be very smusually in µM and sub-µM range. This minimizes the psibility of any probe–probe interaction(s) that may oc[19–22].

Pyrene is one of the most widely used neutral fluorcence probes [19,23,24]. Pyrene, a typical polycyclic amatic hydrocarbon (see Scheme 1) is constituted offused benzene rings and contains no functional grouperwise. The pyrene solvent polarity scale is defined asII/IIIIemission intensity ratio, where band I corresponds tS1(v = 0) → S0(v = 0) transition and band III is a S1(v =0) → S0(v = 1) transition [30–37]. TheII/IIII emission in-tensity ratio increases with increasing solvent polarity. Kpovich and Blanchard rationalized pyrene’s solvatochrobehavior in terms of vibronic coupling between the weaallowed pyrene first electronically allowed singlet state athe strongly allowed second electronic singlet state [36].a detailed description of the origin of the pyrene polascale, readers are referred to references [30–35]. It isportant to mention here that after a detailed investigatStreet Jr. and Acree Jr. have reported on the numerouslems associated with the correct determination of theII/IIIIemission intensity ratio [37]. Among the problems listed

Scheme 1. Pyrene.

-

-

the authors are slit width effects, inner-filtering artifacts,citation wavelengths, and temperature control.

We have measured pyreneII/IIII values in∼1% w/wsolutions of AGENT, 15% OAGENT, 10% DAGENT, 15%DAGENT, 20% DAGENT, 30% DAGENT, and 15% SAGENT at different concentrations of pyrene (1.0, 2.5, 57.5, 10.0, and 25.0 µM) at 25◦C. In order to compare ouresults with reference to the use of the above polymerpseudo-stationary phases in electrokinetic chromatograwe have also measured pyreneII/IIII values for all pyreneconcentrations within 10 mM SDS solution at identicexperimental conditions. Micellar solutions of the surfactSDS were the first, and remain the most widely upseudo-stationary phase in micellar electrokinetic capilchromatography experiments.

Table 1 presents measured pyreneII/IIII emission in-tensity ratios within SDS, AGENT, 15% OAGENT, 15DAGENT, and 15% SAGENT at 1.0, 2.5, 5.0, 7.5, 10and 25.0 µM pyrene under ambient conditions. All saples were excited at 337 nm and excitation and emissionwidths were fixed at 2 nm and 1 nm, respectively. Figurshows 10-µM pyrene emission spectra in SDS, AGE15% OAGENT, 15% DAGENT, and 15% SAGENT soltions. A careful examination of the entries in Table 1veals several important observations. First, our pyreneII/IIIIvalues in each of the aforementioned solutions are dient from one another, strongly suggesting a differentcroenvironment sensed by the excited-state pyrene inof the polymer solutions. Second, in each polymer sotion the measured pyreneII/IIII values are not dependeon pyrene concentration. This is manifested through theobserved standard deviations for concentration-depenpyreneII/IIII values as well as a complete lack of any stematic trend in concentration-dependent pyreneII/IIII val-ues in any of the polymer solutions.

Our mean pyreneII/IIII values reveal that pyrene encouters an extremely dipolar environment in dilute AGENT slution (a highest pyreneII/IIII value of 1.94 was measurein AGENT). Previously reported pyreneII/IIII values from

Fig. 2. Normalized steady-state emission spectra of 10-µM pyrene dissin aqueous SDS, AGENT, 15% OAGENT, 15% DAGENT, and 15SAGENT under ambient conditions. Excitation wavelength is 337 nm.

582 S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587

)

at

Table 1PyreneII/IIII values observed in various aqueous polymer solutions under ambient conditions

[Pyrene] µM SDS (10 mM) AGENT OAGENT (15%) DAGENT (15%) SAGENT (15%

1.0 1.29 1.94 1.05 0.98 1.082.5 1.27 1.95 1.05 0.92 1.135.0 1.28 1.90 1.06 0.95 1.127.5 1.29 1.87 1.07 0.94 1.19

10.0 1.30 1.97 1.09 0.93 1.1425.0 1.28 2.01 1.06 0.98 1.19

Mean PyII/IIII 1.29 1.94 1.06 0.95 1.14(±Std. Dev.) (±0.01) (±0.05) (±0.02) (±0.03) (±0.04)

All polymer solutions are∼1% w/w in water. Uncertainties associated with pyreneII/IIII values are within±0.02 units. Band I and band III peaks are372–374 and 382–384 nm, respectively.λexcitation= 337 nm and the excitation and emission slit widths are at 2 and 1 nm, respectively.

ionsvery

em-of

ainter

theat

theare

T,eryithinodi

hy-catero-bone inex-out-gTmerro-T.olyost

ultted

en-d-

po-%nceex-

15%

s, orult-nelar-by)tion

lu-dblyne

of

by

m-,

inol-

olu-ob-

esom-

fer-ns.se-

pac-m-

theonthe,

up.that.930%yl-

our group in various neat solvents under identical conditsuggest the environment encountered by pyrene to besimilar to that found in water (a pyreneII/IIII value= 1.98)[38,39]. This result is easy to comprehend, as the chical structure of AGENT clearly indicates the absenceany long hydrocarbon chain. The allyl glycidyl ether chpresent in AGENT, on one hand, is fairly polar in characto provide the observed high pyreneII/IIII value, and on theother hand, is creating a solubilization site giving rise tohigher solubility of pyrene in AGENT in comparison to thin water (pyrene solubility in neat water is< 1 µM). In addi-tion, it may also suggest a solvation pocket afforded byaqueous AGENT solution in which the pyrene moietiessignificantly exposed to the bulk water.

Measured pyreneII/IIII values in alkyl-chain-modifieddilute aqueous polymer solutions (i.e., OAGENT, DAGENand SAGENT, all 15%), however, clearly indicate a vnonpolar environment encountered by pyrene probes wthese aqueous polymer solutions. These polymers are mfied by attaching nonpolar octyl, dodecyl, and octadecyldrocarbon chains (vide supra) and our results clearly indisolubilization sites for pyrene moieties which are well ptected from bulk water, suggesting the role of hydrocarchains present within the polymer molecular architecturcreating more hydrophobic solubilization sites. A carefulamination of the entries in Table 1 reveals a surprisingcome. The measured pyreneII/IIII values decrease in goinfrom 15% SAGENT to 15% OAGENT to 15% DAGENsolution. This indicates that for the three aqueous polysolutions, the most hydrophobic solubilization site is pvided by 15% DAGENT, and the least by 15% SAGENThis comes as a surprise, since one would expect the pmer modified by the longest carbon chain to provide the mhydrophobic solubilization site for pyrene. A similar reswas obtained in chromatographic studies, which indicathat the 15% SAGENT provided a more polar solvationvironment than 15% DAGENT [28]. Chromatographic stuies, however, indicated that 15% OAGENT was morelar than 15% SAGENT. The relatively high polarity of 15SAGENT observed by both approaches and the differein relative polarities between the two approaches may beplained by the relatively high cohesiveness observed for

-

-

SAGENT in chromatographic studies. This cohesivenesresistance to formation of a solvation cavity, would resin 15% SAGENT being relatively unlikely to form a sufficiently large solvation cavity for the relatively large pyreprobe. We would also like to remind readers that the dipoity of the pyrene solubilization site will also be affectedthe location of theother chain (i.e., the allyl glycidyl etherpresent on the polymer backbone with respect to the locaof the pyrene solute. Although the measured pyreneII/IIIIvalues are statistically different in the three polymer sotions (0.95 for 15% DAGENT, 1.06 for 15% OAGENT, an1.14 for 15% SAGENT), the differences are considerasmaller than that observed within AGENT solution (pyreII/IIII = 1.94). It is not inconceivable that the proximitythe solubilized pyrene to either –SO−

3 or –N– groups on theallyl glycidyl chain can influence the dipolarity reportedthe pyrene probe.

It is important to mention, at this point, that the coparison of our pyreneII/IIII values within 15% OAGENTDAGENT, and SAGENT solutions with those measured10 mM SDS under identical conditions clearly indicate a subilization environment encountered in these polymer stions which is much more nonpolar in nature than thatserved in 10 mM SDS (pyreneII/IIII = 1.29 within SDS).This low polarity solvation region relative to SDS micellwas also observed in chromatographic studies [28]. In cparison to 10 mM SDS, the lower measured pyreneII/IIIIvalues in the above polymer solutions indicate a difent solubilization site for polycyclic aromatic hydrocarboMore hydrophobic solvation pockets may indicate a morelective pseudo-stationary phase with better solvation caity for hydrophobic compounds, which should lead to iproved electrokinetic separations in some applications.

Next, we investigated the effect of the amount ofDAGENT backbone that is modified with dodecyl chainthe dipolarity of the solubilization site. Table 2 presentspyreneII/IIII values in DAGENT solutions that have 10%15%, and 20% of the chain modified with the dodecyl groA careful examination of the entries in Table 2 revealsthe pyreneII/IIII values decrease from 1.04 to 0.95 to 0as the amount of chain modification is increased from 1to 15% to 20%. DAGENT with a higher amount of dodec

S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587 583

ons

ds

sol-easein-lu-

is-dd-udyaraularer

so-ionn-

we-

WeT

e

bi-t

neval-

ncie

tionioniatedt de-ma-m-ize

re-

ndi-andgth,

ma-tion

ec-nden-ondef-en

onelo-Theivers arehereevi-s ab-ex-

malis tolter-

ex-that

theyim-areres-

entlytionxeshe

ay53].ssionThe

orng

est.uire

theperfor

,S,d

Table 2Pyrene II/IIII values observed in various aqueous DAGENT solutiunder ambient conditions

[Pyrene] µM 10% 15% 20%

1.0 1.02 0.98 0.912.5 1.02 0.92 0.895.0 1.03 0.95 0.927.5 1.03 0.94 0.94

10.0 1.06 0.93 0.9525.0 1.05 0.98 0.95

Mean PyII/IIII 1.04 0.95 0.93(±Std. Dev.) (±0.02) (±0.03) (±0.02)

All DAGENT solutions are∼1% w/w in water. Uncertainties associatewith pyreneII/IIII values are within±0.02 units. Band I and band III peakare at 372–374 and 382–384 nm, respectively.λexcitation= 337 nm and theexcitation and emission slit widths are at 2 and 1 nm, respectively.

modified backbone should present a more hydrophobicubilization site for the pyrene probe, and hence a decrin pyreneII/IIII values. It should be noted that we alsovestigated pyrene probe behavior in 30% DAGENT sotions. However, the observed pyreneII/IIII values in 30%DAGENT solution are relatively high and are not constent with the pyreneII/IIII trend observed in 10%, 15%, an20% DAGENT solutions. This polymer could not be stuied by chromatographic experiments because of the clonature of the solutions needed for chromatographic septions. We suspect there could be significant intermolecaggregation within dilute 30% DAGENT solutions. Furthstudies on this polymer were not conducted at this time.

We conclude that, in general, among all polymerlutions investigated, aqueous 15–20% DAGENT solutmay provide the most hydrophobic solubilization enviroment for polycyclic aromatic hydrocarbons. Further,have measured pyreneII/IIII using exactly the same experimental conditions in various neat solvents [40].found that the pyrene cybotactic region in 15% DAGENis similar to that inn-butyl ether and 1-octanol (pyrenII/IIII in n-butyl ether and 1-octanol is 0.92 (±0.01) and0.99 (±0.01), respectively). PyreneII/IIII in 15% OAGENTis similar to that in neat 1-octanoic acid (1.04 (±0.01)),m-xylene (1.10 (±0.01)), and o-xylene (1.11 (±0.01))suggesting a similar microenvironment for pyrene solulization. PyreneII/IIII in 15% SAGENT was closer to thameasured in neat toluene (1.15 (±0.01)), but much lowerthan that in butyl benzoate (1.26 (±0.01)). It is importantto mention at this point that a comparison of the pyreII/IIII values measured in the present studies to theues obtained by other groups shows some discrepa[30–37]. As mentioned earlier, the pyreneII/IIII values de-pend on several experimental variables such as excitawavelength, excitation/emission slit widths, self-absorptdue to pyrene concentration effects, efficiencies assocwith double-pass versus single-pass cell compartmensigns, procedures involved with data acquisition andnipulation, solvent and pyrene purity, and finally, the teperature of the pyrene solution [37]. In order to minim

-

s

the inconsistencies arising from one or more of the afomentioned reasons, we measured pyreneII/IIII values inall the solvents at more or less similar experimental cotions. All the data were acquired on the same instrumentpyrene concentration, temperature, excitation wavelenexcitation/emission slit widths, and data acquisition andnipulation strategy were identical for each pyrene soluinvestigated.

3.2. Pyrene IE/IM

An excimer is a dimer which is associated in an eltronic excited state and which is dissociative in its groustate [41]. The formation of a pyrene excimer requirescounter of an electronically excited pyrene with a secpyrene in its ground electronic state. According to this dinition, the two pyrenes must be sufficiently far apart whlight is absorbed, so that the excitation is localized onof them [42]. This excited pyrene, often referred to ascally excited pyrene, gives rise to monomer emission.observation of excimer emission indicates that a diffusencounter between pyrene has occurred. These excimecalled dynamic excimers. There are also instances wan excimer-like emission is observed, but there is nodence that the pyrenes are separated when the light isorbed [43]. This emission is due to ground-state dimercitation, and some people call it static excimers. In norcases, one way to achieve a distinction between the twoperform time-resolved fluorescence experiments. An anative, and perhaps easier, distinction between dynamiccimers and those formed from preassociated pyrenes isin the latter case the pyrenes are sufficiently close thatexhibit perturbed absorption and excitation spectra. It isportant to mention here that excitation spectral outcomesmuch more sensitive in nature for the same reasons fluocence is a more sensitive technique than absorbance.

The pyrene excimer fluorescence technique has recbeen modified to more directly measure the solubilizacapacities of micelles and of polymer–surfactant comple[44–46]. Within a molecularly organized assembly, if tlocal concentration of pyrene is high, an excited pyrene mbind to a ground state pyrene and form an excimer [47–The broad and structureless excimer fluorescence emipeak is generally observed between 440 and 500 nm.excimer-to-monomer fluorescence intensity ratio (IE/IM,where IM is the intensity of the first monomer band) fsolubilized pyrene is closely related to its distribution amomicelles or molecularly organized assemblies of interSince both dynamic and static excimer fluorescence reqdimerization during an excited state lifetime and inground state, respectively, a minimum of two pyrenessolubilization site (such as in a micelle) is requiredsolubilized pyrene to produce excimer emission.

Table 3 presents pyreneIE/IM measured for 1.0, 2.55.0, 7.5, 10.0, and 25.0 µM bulk pyrene in 10 mM SD∼1% w/w OAGENT, DAGENT, SAGENT (15% each), an

584 S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587

)

on

on

Table 3PyreneIE/IM measured in various aqueous polymer solutions at ambient conditions

[Pyrene] µM SDS (10 mM) AGENT OAGENT (15%) DAGENT (15%) SAGENT (15%

1.0 0.02 (±0.001) [476] 0.03 (±0.002) [465] 0.13 (±0.01) [473] 0.13 (±0.004) [474] 0.50 (±0.04) [477]2.5 0.04 (±0.002) [476] 0.08 (±0.01) [466] 0.29 (±0.02) [473] 0.30 (±0.03) [473] 0.66 (±0.04) [475]5.0 0.07 (±0.003) [476] 0.18 (±0.01) [468] 0.45 (±0.02) [475] 0.61 (±0.02) [475] 0.76 (±0.05) [474]7.5 0.11 (±0.002) [475] 0.16 (±0.003) [471] 0.40 (±0.01) [473] 0.51 (±0.05) [473] 0.52 (±0.02) [477]

10.0 0.02 (±0.001) [476] 0.14 (±0.003) [465] 0.47 (±0.02) [474] 0.20 (±0.01) [473] 0.90 (±0.05) [473]25.0 0.26 (±0.003) [478] 0.22 (±0.01) [444] 0.61 (±0.05) [472] 0.82 (±0.02) [475] 0.71 (±0.02) [477]

All polymer solutions are∼1% w/w in water. Monomer band (i.e., band I) maxima is at 372–374 nm.λexcitation= 337 nm and the excitation and emissislit widths are at 2 and 1 nm, respectively.

Table 4PyreneIE/IM measured in various aqueous DAGENT solutions at ambient conditions

[Pyrene] µM 10% 15% 20%

1.0 0.11 (±0.01) [472] 0.13 (±0.004) [474] 0.17 (±0.01) [469]2.5 0.23 (±0.02) [474] 0.30 (±0.03) [473] 0.02 (±0.001) [472]5.0 0.39 (±0.04) [474] 0.61 (±0.02) [475] 0.07 (±0.01) [472]7.5 0.43 (±0.04) [473] 0.51 (±0.05) [473] 0.54 (±0.03) [472]

10.0 0.44 (±0.01) [475] 0.20 (±0.01) [473] 0.74 (±0.02) [472]25.0 0.49 (±0.02) [475] 0.82 (±0.02) [475] 0.89 (±0.05) [473]

All polymer solutions are∼1% w/w in water. Monomer band (i.e., band I) maxima is at 372–374 nm.λexcitation= 337 nm and the excitation and emissislit widths are at 2 and 1 nm, respectively.

olved5%

areren-sity

-, it isne

ionseneell. It

tionctionres.ionons

rthate

-ed

ofmer

anck-u-t isle 4ttheTa-or-

ular

intra-tieslso

ob-ene(i.e.,

inT.s in

Fig. 3. Normalized steady-state emission spectra of 25-µM pyrene dissin aqueous SDS, AGENT, 15% OAGENT, 15% DAGENT, and 1SAGENT under ambient conditions. Excitation wavelength is 337 nm.

AGENT aqueous polymer solutions (a part of the datashown in Fig. 3). Table 3 also contains, in square patheses [ ], the wavelengths where the maximum intenof pyrene excimer is observed,λmax

E in nm. Before any discussion on the measured data in Table 3 is undertakenimportant to realize at this point that the efficiency of pyreexcimer formation in these dilute aqueous polymer solutwill depend on many factors [41–53]. Among others, pyrIE/IM will depend on external pyrene concentration as was on each polymers’ solubilization capacity for pyrenewill also depend on the nature of the pyrene solubilizasite, such as its dimensions and dynamics. A close inspeof the entries in Table 3 reveals the following salient featu(i) First and foremost, there is significant excimer formatwithin each aqueous polymer solution for all concentrati

of pyrene. (ii) The observedIE/IM in almost all the polymesolutions at each pyrene concentration is higher thanobserved in SDS. (iii) Except for AGENT, in most of thcases, the excimer maximaλmax

E = 475 (±2) nm, suggesting a similar equilibrium geometries of the excimers formin each case. A lowerλmax

E value observed for AGENTsuggests a slightly different overall equilibrium geometryexcimer than that observed in SDS and the other polysolutions. (iv) PyreneIE/IM values in 15% OAGENT, DA-GENT, and SAGENT solutions are significantly higher ththat observed in SDS or AGENT. The solubilization poets within 15% OAGENT, DAGENT, and SAGENT soltions clearly promote an excimer formation process thamore efficient in nature than that observed in SDS. Tabpresents the pyreneIE/IM values in DAGENT solutions thahave 10%, 15%, and 20% of the chains modified withdodecyl group. A careful examination of the entries inble 4 provides similar observations as above. It is imptant to mention here that in organic solvents intermolecexcimer formation is only observed at� mM pyrene con-centrations [41,42]. A better excimer formation efficiencythe polymer solutions even at very low pyrene concentions suggests the proximity of two or more pyrene moiewithin the solvation pocket. Solubilization environment afacilitates the excimer formation efficiency.

In order to obtain a better understanding of theseservations, we collected the excitation spectra of pyrwhen the emission was monitored at the monomer band∼373 nm)and at the excimer band (i.e.,∼475 nm). Fig-ure 4 shows the excitation spectra atλem = 373 nm (—) andλem = 475 nm (- - -) for 10 and 25 µM pyrene solutionsOAGENT, DAGENT, SAGENT (15% each), and AGENFor each pyrene concentration, the two excitation scan

S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587 585

r ambientitat

Fig. 4. Normalized excitation scans of 10- and 25-µM pyrene solutions in aqueous AGENT, 15% OAGENT, 15% DAGENT, and 15% SAGENT undeconditions. The solid lines represent excitation scans when emission is monitored at the first monomer peak and the dotted lines represent the excion scanswhen emission is monitored at the excimer peak.

cantug-and53]

thes to

ex-forave-eryhatene

SDS solution overlap each other and there is no signifidifference in the two scans (results not shown). This sgests formations of excimers only in the excited statethat no ground state heterogeneity may be present [41–This could be attributed to the relatively bigger size ofprobe-hosting SDS micelles where a pyrene moiety ha

.

diffuse to the other excited pyrene moiety to form thecimer within the micellar assembly. On the contrary,25 µM pyrene the excitation scans at the two emission wlengths are significantly different from each other in evpolymer solution investigated. It is interesting to note tthese differences become less significant for 10 µM pyr

586 S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587

tionA-fors aforten-omi-en-erthe

tructhe

thetiontiesim-outenttionter-

dedA-ter-morbouheseersprothat

lva-ent

eachtionly-by

erynstivengtingnte

nedsult

iningaverd

lles.eemthate to

d sotate.formerentobeForay

bic

ng-

le-

24.R.

rein.40,

rom

ew

ics,9.

andmer

and

wer

al

lo-47–

d. 2

m. 2

oid

pec-ew

lica-

977)

polymer solutions. The differences between the excitascans increase in going from SAGENT to DAGENT to OGENT (15% each) to AGENT. While the excitation scanthe excimer emission in SAGENT and DAGENT showsmall bathochromic shift relative to the excitation scanthe monomer emission and different excitation band insity ratios, these observations become much more prnent in OAGENT and are accompanied by band broading as well. In AGENT, the excitation scan for the excimemission is completely different than that measured formonomer emission. Here, the bands begin to lose their sture, become very broad, and are considerably shifted tolower energies. These results indicate the possibility ofpresence of ground state heterogeneity within the solvapockets, i.e., the interaction of two or more pyrene moiebefore the occurrence of the excitation event [43]. It isportant to mention here that we cannot completely rulethe possibility that the pyrene could be forming a fluoresccomplex with the polysoap and that the observed excitaspectra could be a combination of the pyrene–pyrene inactions and the pyrene–polysoap complex.

Due to the smaller size of the solvation pockets afforby AGENT and OAGENT, and to a lesser extent in DGENT and SAGENT, it is possible that ground-state inactions between one or more pyrene moieties becomesprobable. These results, in effect, provide us an idea athe average size of the solvation pockets afforded by tpolymers. The solvation pockets provided by the polymmay be more restricted in space and geometry than thatvided by the SDS micelles. These results also indicatemore than one pyrene moiety is solubilized within the sotion pocket and at the instance of the pyrene excitation evtwo or more of the pyrene moieties are close enough toother such as to manifest their interactions via the excitascans. All in all, the solubilization sites provided by the pomers solutions are distinctly different from that providedthe SDS micellar solution.

4. Summary

Whereas defining polarity on an absolute scale is vdifficult due to the fact that several different interactiomay be involved, one may wish to determine the relapolarity of the cybotactic region afforded by the solubilizimedia of interest with respect to the probe represenall possible interactions. The fluorescence results presehere provide complementary information to that obtaiby chromatographic measurements. The fluorescence reare more sensitive, and appear more reliable in determthe polarity of the solvation region. These studies hconfirmed that the alkyl-modified siloxane polymers affoa much more nonpolar environment than SDS miceApparently, the solvation pocket size and geometry sto be different than that in SDS as well in a sensewe observe locally enhanced concentration of pyren

-

et

-

,

d

s

an extent where solubilized pyrene moieties are locateclose to each other that they interact in the ground sAdditionally, the results have provided further evidenceseemingly anomalous changes in polarity when the polyis modified with longer alkyl chains or to a greater extwith hydrophobic groups. However, the use of a single prmay lead to results affected by more than one factor.example, the relatively large size of the pyrene probe mhave led to its being excluded from the most hydrophoregions of the more cohesive polymers.

References

[1] A. Laschewsky, Adv. Polym. Sci. 124 (1995) 1.[2] R. Parthasarathy, M.M. Labes, Langmuir 6 (1990) 542.[3] D. Cochin, F.C. De Schryver, A. Laschewsky, J. van Stam, La

muir 17 (2001) 2579.[4] Y. Morishima, T. Kobayashi, T. Furui, S. Nozakura, Macromo

cules 20 (1987) 1707.[5] D.Y. Chu, J.K. Thomas, Macromolecules 24 (1991) 2212.[6] W. Binana-Limbele, R. Zana, Macromolecules 20 (1987) 1331.[7] W. Binana-Limbele, R. Zana, Macromolecules 23 (1990) 2731.[8] D.Y. Chu, J.K. Thomas, Macromolecules 20 (1987) 2133.[9] C.L. McCormick, C.E. Hoyle, M.D. Clark, Polymer 33 (1992) 243.

[10] S. Yusa, M. Kamachi, Y. Morishima, Macromolecules 33 (2000) 12[11] A.C. Nieuwkerk, E.J.M. van Kan, A. Koudijs, A.T.M. Marcelis, E.J.

Sudholter, Langmuir 14 (1998) 5702.[12] J.G. Matisons, A. Provatas, Macromolecules 27 (1994) 3397.[13] C.P. Palmer, Electrophoresis 21 (2000) 4054, and references the[14] S.A. Shamsi, C.P. Palmer, I.M. Warner, Anal. Chem. A 73 (2001) 1

and references cited therein.[15] R.S. Farinato, P.L. Dubin (Eds.), Colloid–Polymer Interactions: F

Fundamentals to Practice, Wiley, New York, 1999.[16] C.E. Carraher Jr., Polymer Chemistry: An Introduction, Dekker, N

York, 1996.[17] D.F. Evans, H. Wennerstrom, The Colloidal Domain: Where Phys

Chemistry, Biology, Technology Meet, Wiley–VCH, New York, 199[18] K. Kawasaki, B. Lindman, H. Okabayashi (Eds.), Formation

Dynamics of Self-Organized Structures in Surfactants and PolySolutions, in: F. Kremer, G. Lagaly (Eds.), Progress in ColloidPolymer Science, Vol. 106, Springer-Verlag, New York, 1997.

[19] J. Georges, Spectrochim. Acta Rev. 13 (1990) 27.[20] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Klu

Academic/Plenum Press, New York, 1999, Chap. 6.[21] W.E. Acree Jr., in: R.A. Meyers (Ed.), Encyclopedia of Analytic

Chemistry, Wiley, Chichester, 2000, pp. 10,280–10,305.[22] F.V. Bright, S. Pandey, G.A. Baker, in: R.A. Meyers (Ed.), Encyc

pedia of Analytical Chemistry, Wiley, Chichester, 2000, pp. 10,410,472.

[23] W.E. Acree Jr., S.A. Tucker, J.C. Fetzer, Polycycl. Aromat. Comp(1991) 75.

[24] W.E. Acree Jr., S. Pandey, S.A. Tucker, Curr. Top. Solution Che(1997) 1.

[25] G.B. Behera, B.K. Mishra, P.K. Behera, M. Panda, Adv. CollInterface Sci. 82 (1999) 1.

[26] B. Valeur, in: S.G. Schulman (Ed.), Molecular Luminescence Stroscopy, Part 3, in: Chemical Analysis Series, Vol. 77, Wiley, NYork, 1993.

[27] D.S. Peterson, C.P. Palmer, J. Chromatogr. A 924 (2001) 103.[28] D.S. Peterson, C.P. Palmer, Electrophoresis 22 (2001) 3562.[29] D.S. Peterson, C.P. Palmer, J. Chromatogr. A, submitted for pub

tion.[30] K. Kalyanasundaram, J.K. Thomas, J. Am. Chem. Soc. 99 (1

2039.[31] K. Hara, W.R. Ware, Chem. Phys. 51 (1980) 61.

S. Pandey et al. / Journal of Colloid and Interface Science 262 (2003) 579–587 587

et,

nd

. 3

ce,

)

9)

.

89.30

86)

e,

es-

[32] D.C. Dong, M.A. Winnik, Can. J. Chem. 62 (1984) 2560.[33] W.E. Acree Jr., S.A. Tucker, L.E. Cretella, A.I. Zvaigzne, K.W. Stre

J.C. Fetzer, Appl. Spectrosc. 44 (1990) 951.[34] J.W. Carr, J.M. Harris, Anal. Chem. 58 (1986) 626.[35] S.H. Chen, V.L. McGuffin, Appl. Spectrosc. 48 (1994) 596.[36] D.S. Karpovich, G.J. Blanchard, J. Phys. Chem. 99 (1995) 3951.[37] K.W. Street Jr., W.E. Acree Jr., Analyst 111 (1986) 1197, a

references therein.[38] K.A. Fletcher, I.K. Storey, A.E. Hendricks, S. Pandey, Green Chem

(2001) 210.[39] K.A. Fletcher, S. Pandey, Appl. Spectrosc. 56 (2002) 266.[40] K.A. Fletcher, S. Pandey, unpublished results.[41] J.B. Birks, Photophysics of Aromatic Molecules, Wiley–Interscien

London, 1970, Chap. 7.[42] P.P. Infelta, Chem. Phys. Lett. 61 (1979) 88.[43] F.M. Winnik, Chem. Rev. 93 (1993) 587.

[44] J.-H. Kim, M.M. Domach, R.D. Tilton, Colloids Surf. A 150 (199953.

[45] J.-H. Kim, M.M. Domach, R.D. Tilton, J. Phys. Chem. B 103 (19910,582.

[46] J.-H. Kim, M.M. Domach, R.D. Tilton, Langmuir 16 (2000) 10,037[47] T. Fujii, E. Shimizu, Chem. Phys. Lett. 137 (1987) 448.[48] A. Jover, F. Meijide, E.R. Nunez, J.V. Tato, Langmuir 12 (1996) 17[49] M.R. Vigil, J. Bravo, T.D.Z. Atvars, J. Baselga, Macromolecules

(1997) 4871.[50] A.K. Dutta, T.N. Misra, A.J. Pal, Langmuir 12 (1996) 459.[51] T. Yamazaki, N. Tamai, I. Yamazaki, Chem. Phys. Lett. 124 (19

326.[52] D. Avnir, R. Busse, M. Ottolenghi, E. Wellner, K.A. Zachariass

J. Phys. Chem. 89 (1985) 3521.[53] T. Kitamura, Y. Takahashi, T. Yamanaka, K. Uchida, J. Lumin

cence 48/49 (1991) 373.