On the Origin of Mesoscale Structures in Aqueous Solutions ofTertiary Butyl Alcohol: The Mystery ResolvedMarian Sedlak* and Dmytro Rak

Institute of Experimental Physics, Slovak Academy of Sciences, Watsonova 47, 040 01 Kosice, Slovakia

*S Supporting Information

ABSTRACT: We have performed a detailed experimental study on aqueous solutions of tertiary butyl alcohol which were asubject of long-standing controversies regarding the puzzling presence of virtually infinitely stable large-scale structures in suchsolutions occurring at length scales exceeding appreciably dimensions of individual molecules, referred to also as mesoscalestructures. A combination of static and dynamic light scattering yielding information on solution structure and dynamics and gaschromatography coupled with mass spectrometry yielding information on chemical composition was used. We show that tertiarybutyl alcohol clearly exhibiting such structures upon mixing with water does not contain any propylene oxide, which waspreviously considered as a source of these structures (an impurity expected to be present in all commercial samples of TBA).More importantly, we show that no mesoscale structures are generated upon addition of propylene oxide to aqueous solutions ofTBA. The ternary system TBA/water/propylene oxide exhibits homogeneous mixing of the components on mesoscales. Weshow that the source of the mesoscale structures is a mesophase separation of appreciably more hydrophobic compounds thanpropylene oxide. These substances are explicitly analytically identified as well as their disappearance upon filtering out themesoscale structures by nanopore filtration. We clearly show which substances are disappearing upon filtration and which are not.This enables us to estimate with rather high probability the chemical composition of the mesoscale structures. Visualization oflarge-scale structures via nanoparticle tracking analysis is also presented. Video capturing the mesoscale particles as well as theirBrownian motion can be found in the Supporting Information.

1. INTRODUCTION

Aqueous solutions of tertiary butyl alcohol (TBA), the last froma series of alcohols still freely miscible with water, have been asubject of interest, discussion, as well as controversies regardingthe observation of anomalously high light scattering, indicatingthe presence of some sort of large structures present in suchsolutions. The first observation of the anomalously high lightscattering was reported in 1972 by Vuks and Shurupova1 andinterpreted as a “phase transition” between clathrate-like moreordered structure and less ordered molecular structure. Beerand Jolly2 reported that observations similar to those found inref 1 were dependent on the sample history as well as purity.Euliss and Sorensen3 noted enhanced inhomogeneities uponcooling and attributed them possibly to clathrates formed byTBA, water, and some help gas. Bender and Pecora4 did notreport any inhomogeneities in aqueous TBA solutions but did

report inhomogeneities in aqueous solutions of 2-butoxyetha-nol.5

Independent of the forgotten TBA story, years later, a shortcommunication by Georgalis et al.6 appeared reporting theobservation of a slow dynamic mode in dynamic light scattering(DLS) experiments on three different types of electrolytes inaqueous solutions. This mode was interpreted as being due tosub-micrometer size clusters of ions. In 2006, a series of threeextensive works by Sedlak7−9 showed that mesoscale structuresappreciably exceeding dimensions of individual molecules arenot a privilege of one or a few special system(s) but ratherpresent in whole classes of systems. It was shown quitesurprisingly that the majority of solutions and mixtures as usedin everyday life and research practice do possess long-lived

Received: January 27, 2014Published: February 21, 2014

Article

pubs.acs.org/JPCB

© 2014 American Chemical Society 2726 dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−2737

large-scale inhomogeneities. Around 100 different solute−solvent pairs were investigated in a detailed study7−9 with theaim to classify systems with respect to the capability offormation of the large-scale structures as well as to try to shedsome light on the mechanism of their formation. The presenceand intensity of the large-scale structures was correlated withproperties of constituent molecules and ions such as theircharge, dipole moment, protic vs aprotic character, etc. It wasfound9 that electrolytes of both inorganic and organic originexhibit large-scale structures in aqueous solutions (in all cases)and in organic solvents (selectively). Solutions of non-electrolytes including mixtures of liquids exhibit large-scalestructures in aqueous solutions (in all cases) and in organicsolvents (selectively). Nonaqueous mixtures did not exhibitlarge-scale organization in the case of nonpolar and weaklypolar components. Since anomalously high light scattering anda slow dynamic mode in dynamic light scattering can havevarious origins, a detailed light scattering analysis wasperformed7 with a conclusion that the source of the slowmode scattering is real objects (not fluctuations) and that theseare discrete objects (not bicontinuous phases with largecorrelation lengths). They possess smaller or larger poly-dispersities, in certain cases with sizes spanning up to a decade(radii from ∼30 to ∼300 nm). A nice agreement between sizedistributions obtained from static light scattering and nano-particle tracking analysis (analysis of particle dynamics frommicroscopic video images) was published recently.10 Averagesizes as well as size distributions are usually concentration-dependent.7 A detailed study was also devoted to the kinetics ofthe structure formation as well as to its long-time stability.8 Thelarge-scale structures develop upon mixing the components ofthe solution or mixture on time scales varying from minutes toweeks, depending on the concrete system. Importantly, a trulyhomogeneous mixture is obtained at the beginning, and onlyaf terward, a buildup of large structures begins. The long-timestability of large-scale structures was studied in detail over timeintervals ranging up to 15 months. While in some systems thestructures appeared virtually infinitely stable, a very slowceasing was observed in others (weeks to months). While thecharacteristics are, at least qualitatively, similar in varioussystems, it is not clear whether the mechanism of the formationof mesoscale structures is the same in all systems or whether wedeal with more than one mechanism. The presence of largestructures was confirmed in works from several groups, namely,in aqueous solutions of several common well-solublecompounds (urea, α-cyclodextrin, ethanol),11,12 in aqueoussolutions of amino acids,13 in mixtures of water with severalorganic liquids,14 and in THF/water mixtures.11,12,15

Several interpretations of the origin of these structures werehypothesized in the literature. It was reported9 that mesoscalestructures appear in solvents capable of 3D hydrogen-bondnetworking and therefore hypothesized that the effect may berelated to solute−solute and solute−solvent hydrogen bonding.They were interpreted also as nanobubbles,11,12 whilesubsequently the nanobubble nature of these structures wasdismissed.10,14 While the work of Habich et al.14 was focusedmainly on verification of the eventual nanobubble nature ofsuch structures, it was noted that the scattering signal issensitive to the purity of the substances and thus, in principle,may originate from some kind of mesophase separation ofminority components in the mixtures. A similar conclusion wasdrawn from LS and SANS data on THF/water mixtures where

THF contained hydrophobic butylhydroxytoluene antioxidantas an additive in various concentrations.15

Coming back to the TBA story, Subramanian andAnisimov16 reported recently on resolving the mystery ofaqueous solutions of tertiary butyl alcohol. The mesoscalestructures were reported to originate as a consequence of thepresence of propylene oxide in solution (“an impurity expectedto be present in all commercial samples of TBA”16,17). Thisconclusion was made on the basis of the observation that theslow dynamic mode in DLS disappeared irreversibly afterintensive filtration through nanoporous filters and was thenregenerated upon addition of trace amounts of propylene oxide.No effect was reported when other substances were added, suchas 10−4 mole fraction isopropanol, 10−4 mole fractionisobutanol, and 4 × 10−5 mole fraction cyclohexane.16 Similarly,no effect was seen upon adding PO to isopropanol aqueoussolution. It was concluded that the formation of mesoscalestructures requires specific interactions of the major compo-nents (TBA, water in this case) with minor components (as POin this case). As authors correctly note in their work,16 manyopen questions arise, such as what is the internal structure ofsuch structures, what controls their size, are they thermody-namically stable or kinetically arrested, and can results from thissystem be generalized to other systems?The aim of this work was the following: (i) to perform a

much more detailed light scattering investigation in a muchwider variety of experimental conditions and mixturecompositions including static light scattering not performedso far on these systems, (ii) to perform high precision analyticalmeasurements via GC-MS (gas chromatography coupled withmass spectrometry) and to correlate them with structural anddynamical data from static and dynamic light scattering, and(iii) according to results from GCMS to try to purify thecomponents and to repeat measurements on pure (orsignificantly purer) major components. We wanted amongothers to answer the question whether the structures presentprior to filtration are the same as structures generated by POaddition. In principle, these might be two different types ofstructures. Our approach proved to be useful. We show that nomesoscale structures are generated upon addition of propyleneoxide to aqueous solutions of TBA. The ternary system TBA/water/propylene oxide exhibits homogeneous mixing of thecomponents on mesoscales. We show that the source of themesoscale structures is a mesophase separation of appreciablymore hydrophobic compounds than propylene oxide. Thesesubstances are explicitly analytically identified as well as theirdisappearance upon filtering out the mesoscale structures bynanopore filtration. We clearly show which substances aredisappearing upon filtration and which are not. This enables usto estimate with rather high probability the chemicalcomposition of the mesoscale structures. We also explain thereason why addition of PO misleadingly mimics the generationof mesoscale structures.

2. EXPERIMENTMaterials. TBA (tert-butyl alcohol) was from Sigma-Aldrich

(99.7%). PO (propylene oxide) was also from Sigma-Aldrich(99.5%). Both compounds were additionally purified withemphasis on removing mainly hydrophobic impurities using thefollowing procedure. 30% water solution of TBA was stirredwith activated charcoal (DARCO, Sigma-Aldrich) overnight at70 °C. Then, TBA was distilled off, dried using a waterabsorbing molecular sieve (Merck, Darmstadt), and distilled

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372727

once again. The same method was used to purify the propyleneoxide (PO). In this case, due to the low boiling point of PO, theaqueous solution with activated charcoal was stirred at roomtemperature. Water was purified by reversed osmosis andactivated carbon TOC reduction, freshly double-distilled in aquartz apparatus, and subsequently deionized by analyticalgrade mixed-bed ion exchange resins (Bio-Rad, Richmond,CA). We have used also water from commercial apparatusELGA Purelab Ultra Analytic (Elga, United Kingdom). Nodependence of results on water source was observed. Theresistivity of water was always above 15 MΩ cm.Static Light Scattering (SLS). SLS measurements were

made using a 40 mW HeNe laser, model 25LHP928 (CVIMelles Griot, Albuquerque, NM) with 632.8 nm verticallypolarized beam. No change of data with laser power wasobserved in the range 1−40 mW. A laboratory madegoniometer with an angular range from 30 to 135° was usedto collect data for both static and dynamic light scatteringexperiments. Scattering intensities were measured by photoncounting.Scattering intensities were normalized using doubly distilled

and filtered benzene as a standard and expressed as I/IB ratios,where IB is benzene total scattering. Temperature was set with aprecision of ±0.03 °C. Great attention was paid to the purity ofsamples. Scattering cells were thoroughly cleaned from dust. Allsolvents were filtered through nanoporous filters because of thesame reason. Experiments on blank samples (nonstructuredmixtures or pure liquids) showed a complete dust removal.Angular dependencies of scattering intensity were usuallymeasured several times on one sample and subsequentlyaveraged.Dynamic Light Scattering (DLS). An ALV5000E

correlator with a fast correlation board option (ALV, Langen,Germany) was used for photon correlation measurements.Characteristic decay times of dynamic modes τi and theirrelative amplitudes Ai(τi) were evaluated through the momentsof distribution functions of decay times A(τ) obtained by fittingcorrelation curves using CONTIN18 and GENDIST19,20

programs as

∫ τ τ= τ∞

−g t A( ) ( )e dt(1)

0

/(1)

Diffusion coefficients were calculated as Di = (1/τi)q−2, where q

is the scattering vector defined as q = (4πn/λ0) sin(θ/2), with nbeing the solution refractive index, λ0 the laser wavelength, andθ the scattering angle. Two diffusive modes were detected.They were characterized by diffusion coefficients Df, Ds andamplitudes Af, As (subscripts f and s refer to faster and slower,respectively). Correlation curves at various angles wererecorded concurrently with integral scattering intensities I(θ)(solution scattering) and IB(θ) (scattering of a benzenestandard). Normalized scattering amplitudes of the twodynamic modes Af(θ) and As(θ) were calculated as

θθ θ

θ θ=

+A

I IA A

( )( )/ ( )

1 ( )/ ( )sB

f s (2)

θθ θ

θ θ=

+A

I IA A

( )( )/ ( )

1 ( )/ ( )fB

s f (3)

assuming that I(θ)/IB(θ) = Af(θ) + As(θ). Dimensionless ratiosAs(θ)/Af(θ) and Af(θ)/As(θ) were taken from DLS spectra ofrelaxation times.

GC-MS (Gas Chromatography Coupled with MassSpectrometry). Samples were analyzed on an Agilent 7890AGC system with an Agilent 5975 C Mass Selective Detector(Agilent, Palo Alto, CA) and nonpolar HP-5 ms column (30 m× 0.25 mm × 0.25 μm). Two methods were used: directinjection and the HS-SPME method (headspace solid-phasemicroextraction). Direct injection of 1 μL of sample wasrealized with a pulse of 140 kPa, 24 s, at 280 °C and followedby analysis with the following temperature program: 50 °C − 2min − 2 °C/min − 300 °C − 10 min. Helium was used as acarrier gas (99.998%; flow rate: 0.9 mL/min; SIAD, Bergamo,Italy). In the case of the HS-SPME method, 3 mL aliquots ofsolutions in 20 mL glass vials with a silicone septum weresampled using a DVB/CAR/PDMS (divinylbenzene/carboxen/polydimethylsiloxane) extraction fiber (SUPELCO, Bellefonte,USA) for 1 h at ambient temperature. Extraction fiber was thenplaced in the headspace. Desorption occurred in the injectionport at 280 °C for 5 min; the flow rate was 2 mL/min for 12 sand then 0.9 mL/min for the rest of the analysis. Separatedcompounds were identified on the basis of their mass spectra inthe range 29−520 m/z (m being the mass and z the charge),comparison with the NIST08 spectra library, and taking intoaccount other facts and factors such as expected elution orderand literature data on boiling points.

Nanoparticle Tracking Analysis (NTA). NTA was carriedout with an LM10B Nanoparticle characterization system fromNano Sight (Amesbury, U.K.) with a trinocular microscope andan LM12 viewing unit with a 60mW laser working at λ = 405nm. Video sequences were recorded via CCD camera operatingat 30 frames per second (fps) and evaluated via theNANOSIGHT NTA 2.3 Analytical Software Suite. BlankNTA experiments with pure water or pure TBA wereperformed to exclude a possible contamination of water orviewing unit with parasitic scatterers. Perfectly clean darkimages (without any traces of particles) were obtained on pureliquids.

3. RESULTS AND DISCUSSIONFigure 1 shows typical examples of light scattering data onaqueous solutions of TBA. Autocorrelation functions fromdynamic light scattering clearly exhibit two modes. Both modesare diffusive in nature, since the characteristic frequencies ofboth exhibit q2 dependences, where q is the scattering vector(not shown). The fast mode corresponds to the “ordinarymolecular scattering”, more exactly to the concentration(mixture composition) fluctuations relaxing by diffusion, andthe slower mode is due to the presence of large structuresappreciably exceeding molecular dimensions (also referred to asmesoscale structures). While both fast and slow diffusive modesare clearly seen at 330 g/kg due to comparable amplitudes, theslow diffusive mode is dominating at 150 g/kg, since itsamplitude is incomparably higher than the fast mode amplitude.The calculated apparent hydrodynamic radii are Rhapp = 66 nm(cTBA = 150 g/kg) and 53 nm (cTBA = 330 g/kg), respectively.These values were obtained via the Stokes−Einstein formulausing viscosities of particular TBA/H2O mixtures. Correspond-ing angular dependencies of the slow mode amplitude fromstatic light scattering are shown in Figure 1B. Radii of gyrationobtained from Guinier approximation are Rg = 60 nm (cTBA =150 g/kg) and 45 nm (cTBA = 330 g/kg), respectively. Thewhole concentration diagram is shown in Figure 2. Evidently,the ability to form mesoscale structures is concentration-dependent with a maximum somewhere between 100 and 200

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372728

g/kg that corresponds to TBA molar fractions of 0.026−0.057.The effect is ceasing at high concentrations and is practicallyundetectable for TBA concentrations of ∼700−1000 g/kg.Upon going to low concentrations, the effect is still relatively

strongly present, the concentration-normalized intensity is stillrelatively high, but the scattering signal in terms of its absolutevalue is becoming weak and hence measurements are becomingmore and more difficult and data less reliable.Long-term stability of mesoscale structures was tested by

repeated measurements on selected samples over intervals ofseveral weeks or even months with a conclusion that thesestructures were stable. This is in agreement with the previouswork on aqueous TBA solutions16 as well as on aqueoussolutions of other organic compounds.8 A distinct feature ofaqueous TBA solutions (not found in some other types ofsolutes investigated in our laboratory in the past) was thereported influence of temperature: the effect of mesoscalestructures was more pronounced at lower temperatures than athigher temperatures.16 In order to obtain a closer insight intothis, we have performed detailed static and dynamic lightscattering measurements upon temperature cycling (Figure 3).

Temperature changes appear to be fully reversible without anyhysteresis. Scattering signal from mesoscale structures becomesstronger upon cooling and weakens upon heating (Figure 3A).This can be due to an increase in the number of mesoscalestructures and/or their size, and/or their scattering contrastwith respect to the rest of the solution (“matrix”). Figure 3Bclearly shows that the size does change reversibly with

Figure 1. (A) Typical example of bimodal autocorrelation functionsfrom dynamic light scattering experiments on aqueous TBA solutions.Red curve: concentration 150 g/kg (molar fraction 0.041). Greencurve: concentration 330 g/kg (molar fraction 0.107). While both fastand slow diffusive modes are clearly seen at 330 g/kg, the slowdiffusive mode is dominating at 150 g/kg, since its amplitude isincomparably higher than the fast mode amplitude. Scattering angle θ= 90°. (B) Static light scattering angular dependencies of the slowmode amplitude. Corresponding radii of gyration Rg = 60 nm (cTBA =150 g/kg) and Rg = 45 nm (cTBA = 330 g/kg), respectively. T = 25 °C.

Figure 2. Concentration-normalized scattering intensity frommesoscale structures in aqueous solutions of TBA. Scattering angle θ= 45°. T = 7 °C.

Figure 3. (A) Temperature dependence of scattering intensity frommesoscale structures. Scattering angle θ = 45°. Blue symbolscorrespond to gradual cooling and red symbols to gradual heating,respectively. (B) Temperature dependence of the ratio Dsp/Ds, whereDsp is the slow diffusion coefficient predicted for temperature Tprovided that no change occurs in the sample upon changingtemperature except for changes of solution viscosity η. T = 25 °C istaken as a reference temperature so that Dsp/Ds(25 °C) = 1. Thedecrease of Dsp/Ds with T is due to the decrease of the apparenthydrodynamic radius Rh,app of mesoscale structures. Scattering angle θ= 45°.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372729

temperature, namely, that the size increases upon cooling anddecreases upon heating. The ratio Dsp/Ds is shown in Figure3B, where Dsp is the slow diffusion coefficient expected attemperature T provided that no change occurs in the sampleupon changing temperature except for changes of solutionviscosity η. Ds is the actually measured slow diffusion coefficientat temperature T. T = 25 °C is taken as a reference temperatureso that Dsp/Ds(25 °C) = 1. The ratio Dsp/Ds can be thusinterpreted as the ratio of apparent hydrodynamic radii ofmesoscale structures Rhapp,T/Rhapp,25°C. A similar result wasobtained at a TBA concentration of 330 g/kg. However, sincethe scattering from mesoscale structures is much weaker here,almost a complete disappearance of signal was observed above40 °C. Its reappearance was observed upon cooling back. Thedisappearance of the slow mode signal does not necessarilymean a complete disintegration of structures but can mean thattheir size and/or scattering contrast get so small that the signaldrops below the detectability limit.The next focus in this work was on the verification of the

reported claim16 that the source of the mesoscale structures inaqueous TBA solutions is propylene oxide (PO), an impurityassumed16 to be present in all commercial TBA products,including those of p.a. quality. This claim was based onexperiments which showed that mesoscale structures could beirreversibly eliminated by intensive filtration through mem-brane filters with nanopores and subsequently regenerated byPO addition.16 Our expected goal was not only to repeat andverify these findings in our laboratory but also to try to answerthe question of whether the structures present in solution priorto filtration and those generated by PO addition are thestructures of the same origin, simply because they do not haveto be.Figure 4 shows results of an experiment where 150 g/kg TBA

solution was first measured without filtration (just thecomponents of the mixture, i.e., water and TBA, wereseparately filtered prior to mixingadded via nanoporousfilters directly to the scattering cell and afterward gently mixed).Subsequently, the mesoscale structures were completelyeliminated via filtration of the mixture and the scattering was

afterward solely due to “molecular scattering” (fast mode).Then, propylene oxide was added to the sample via a syringefilter such that the resulting concentration of added propyleneoxide in the sample (in solution) was cPO = 0.9 mass %. AddedPO indeed generated scattering signal similar to the signal priorto filtration. What is, however, evident in this experiment is thatthe generated intensity of the slow mode after addition of PO isstill approximately 3 times smaller than the former intensity inthe TBA/water mixture prior to filtration in spite of the factthat the concentration of added PO is incomparably higherthan what can be expected in TBA of p.a. quality withdetermined 99.7% purity. The 0.9% concentration of PO inTBA solution corresponds to the 6% concentration of PO inpure TBA. It is of course absolutely unrealistic that the TBAwould contain such a concentration of PO, since all impuritiesaltogether constitute 0.3%. This was the first observationleading to questioning the origin of mesoscale structures inaqueous solutions of TBA as due to the action of PO.Subsequently, the influence of PO was investigated in moredetail in a whole variety of concentrations and mixturecompositions.Figure 5 shows a summary of results from experiment where

aqueous TBA solution (c = 330g/kg) was first cold-filtered at 7

°C such that the mesoscale structures were eliminated (As = 0),and subsequently cold PO (T = 7 °C) was gradually added.Work at lower temperature was chosen, since the effect is morepronounced at lower temperatures (Figure 3). After each POaddition, the sample was maintained and monitored at 7 °C fora couple of days (typically 2−6) while the intensity was stillchanging (vertical arrows). Sample compositions are specifiedin Table 1.Figure 6 shows angular dependencies of scattering from

large-scale structures generated by PO addition. The scenario isusually such that intensity goes down after addition of PO andthe angular dependence becomes less steep which means thatthe structures get smaller after addition of PO. Then, structuresgrow over several days and consequently angular dependenciesbecome steeper. The inset in Figure 5 shows in detail data atvery low PO concentrations and documents that no effect is

Figure 4. Angular dependencies of the overall light scattering intensity(fast mode “molecular scattering” plus slow mode scattering originatedfrom mesoscale structures): (blue ○) TBA solution 150 g/kg prior tofiltration; (blue ●) the same sample after filtration; (red □) thefiltered sample after addition of filtered dust-free propylene oxide. Theresulting concentration of added propylene oxide in the sample c = 0.9mass %. The whole experiment was done at 7 °C, including cold POadded. Scattering intensity is normalized to benzene scattering IB.

Figure 5. Intensity of scattering from mesoscale structures in aqueoussolutions of TBA (c = 330 g/kg) as a function of the concentration ofadded propylene oxide. TBA solution was first cold-filtered at 7 °Csuch that As = 0, and subsequently, cold PO (T = 7 °C) was added.After each PO addition, the sample was maintained and monitored at7 °C for a couple of days (typically 2−6) while the intensity was stillchanging (vertical arrows). Sample compositions are specified in Table1. Scattering angle θ = 45°.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372730

seen up to cPO = 1% and that it is necessary to increase the POconcentration to cPO ≈ 4% to reproduce the intensity found inunfiltered TBA solution. Concentrations of cPO = 1 and 4% insolution correspond to cPO = 3 and 12% in pure TBA. This is ofcourse in strong contrast to the 99.7% purity of the TBA used.Results from Figure 5 are thus in agreement with conclusionsdrawn from Figure 4. The effect of the formation of mesoscalestructures reaches a maximum for cPO = 60−70% and graduallyceases at high cPO.Figure 7 shows results of an experiment where the TBA/PO

mixture with the composition 42.5/57.5 was investigated. Thismixture shows no evidence of any mesoscale structuring. Onlythe fast mode (“ordinary molecular scattering”) is present.Then, small amounts of filtered water were gradually addeddirectly to the scattering cell and scattering was monitored. Assmall addition of water as that leading to the concentration ofwater in the mixture cH2O = 0.012% resulted in the formation ofmesoscale structures. Mesoscale structures then grew as afunction of added water content (overall increase of scatteringaccompanied with gradually steeper and steeper angulardependencies).

Table 2 shows radii of gyration of mesoscale structurescalculated in Guinier approximation as a function of the

concentration of added water. It is evident from Figure 7 thatalso the polydispersity increases. Hence, the radii of gyrationlisted in Table 2 are rather some upper limit estimates of theparticular size distributions. Figure 8 shows the kinetics of themesoscale structure formationreal time monitoring via lightscattering. The TBA/PO mixture was measured first withoutaddition of water where only one mode (“ordinary molecularscattering”) is seen in DLS data. Then, one drop of filtered coldwater (7 °C) was added to the mixture in the scattering cell, thecell was shortly gently shaken to homogenize the mixture, andmeasurement continued. A gradual increase of intensity due tothe growth of large scattering objects (mesoscale structures) isevident. An accompanied gradual increase of intensityfluctuations is also evident, as expected.Figure 9 shows a ternary diagram of the TBA/water/PO

system summarizing mixture compositions that have beeninvestigated in this work. The height of the vertical barscorresponds to scattering As from mesoscale structuresmeasured at an angle of 45°, normalized to benzene scattering.The overall dependence of As on the mixture composition overthe whole ternary diagram will be discussed later in this paper.We would like to stress at this point that, in all cases (all

Table 1. Mixture Compositionsa

cTBA (%) cH2O (%) cPO (%)

1 32.87 67.13 0.002 32.74 66.85 0.403 32.61 66.59 0.804 31.47 64.26 4.265 30.19 61.64 8.186 27.90 59.98 15.127 25.62 52.30 22.088 22.75 46.44 30.829 17.65 36.04 46.3210 11.11 22.67 66.2211 3.92 7.99 88.0912 1.95 3.99 94.0613 0.47 0.97 98.56

aData refer to the experiment from Figure 5.

Figure 6. Angular dependencies of scattering from mesoscalestructures in aqueous solutions of TBA (c = 330 g/kg) as a functionof the concentration of added propylene oxide. Data refer to theexperiment described in Figure 5. The same color of the symbolsmeans measurements almost at the same time. An open symbol meansmeasurement shortly before addition of PO, and a closed symbolmeans measurement shortly after addition of PO. The wholeexperiment was done at 7 °C, including cold PO added.

Figure 7. Angular dependencies of scattering from TBA/PO mixtureupon gradual addition of water. The initial composition of the mixturewas 57.5 mass % TBA and 42.5 mass % PO. Then, water was graduallyadded to this mixture. Concentration of added water in the mixture(mass %): (black ○) 0, (red ○) 0.012, (green ○) 2.37, (blue ○) 3.97,(teal ○) 5.52, (olive ○) 9.87, and (magenta ○) 19.72. The wholeexperiment was performed at 7 °C. Scattering intensity is normalizedto benzene scattering IB.

Table 2. Radii of Gyration of Mesoscale Structures in TBA/PO Mixtures as a Function of the Concentration of AddedWatera

cH2O (%) cPO (%) cTBA (%) Rg (nm)

0.00 42.53 57.470.012 42.53 57.46 70.32.37 41.52 56.10 85.13.97 40.84 55.19 1395.52 40.18 54.30 1449.87 38.33 51.80 17319.72 34.14 46.13 185

aData refer to experiment from Figure 7.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372731

compositions) where mesoscale structures were found, thescattering signal from mesoscale structures could be finallyirreversibly eliminated by filtration via fine nanoporous filters,including mixtures with huge concentrations of PO on theorder of tens of percent. This finding is in disagreement withthe interpretation that mesoscale structures in aqueous TBAsolutions are due to the presence of PO as an impuritycontained in TBA16 and that the irreversible disappearance ofthese structures is due to the removal of PO segregated in thesestructures.16 It can be easily verified by weighing the filterbefore and after filtration that only a very small amount offiltered material is retained on the filter and hence practically allPO is passing through the filter to the filtered solution whenfiltering mixtures with high PO content. Evidently themesoscale structures disappear not due to the removal of PO.

This finding together with other above-mentioned facts leadus to the next step, namely, to correlating light scattering datawith analytical measurements by GC-MS. Figure 10 shows a

GC-MS elugram of pure TBA. First of all, it should bementioned that no propylene oxide was detected in our TBA.Main components contained in our TBA are listed in the figurecaption together with numbers of corresponding peaks in theelugram. Because a nonpolar column was used in the GC-MSexperiment, more hydrophilic compounds eluted at shortertimes while more hydrophobic compounds eluted at longertimes. It should be taken into account that sizes (areas) ofpeaks corresponding to particular compounds are not directlyproportional to their concentrations, since these areas depend(aside from compound concentrations) also on otherparameters.Figure 11 shows results of GC-MS analysis on aqueous TBA

solution (c = 150 g/kg) prior to filtration and after filtration,respectively.It is clearly evident which compounds are eliminated by

filtration and which are not. More detailed data are listed inTable 3, but it is clear that dodecane, tetradecane, hexadecane,and octadecane, i.e., extremely hydrophobic compounds, arepractically completely eliminated (corresponding peaks de-creased 1000 times or more). On the other hand, compoundsless hydrophobic like 2-methoxy-2-methylpropane, 1-(1,1-dimethylethoxy)-2,2-dimethylpropane, or methylhexanol arenot eliminated from solution by filtration. The third group ofcompounds numbered as 10−14, i.e., silans and siloxans,originate from the leaking of the GC column and extractionfiber when exposed to water vapors and are not coming fromthe sample itself. They are as expected practically the samewhether filtered or unfiltered sample is analyzed. Lightscattering data for the filtered and unfiltered samples fromFigure 11 are in the Supporting Information. The filteredsample shows no presence of mesoscale structures. It is possibleto conclude with very high confidence that mesoscale structuresare composed mainly of the highly hydrophobic, relatively longhydrocarbons dodecane, tetradecane, hexadecane, and octade-cane, while less hydrophobic compounds are not involved (orvery marginally) in mesoscale structures. Propylene oxidebelongs to weakly hydrophobic compounds; it is miscible with

Figure 8. Real-time monitoring of the origination of mesoscalestructures. TBA/PO mixture with mass composition 57.5% TBA and42.5% PO was measured at a temperature of 7 °C (scattering signalfrom t = 0 to t = 5 min). Only one mode (“ordinary molecularscattering”) was seen in DLS data. Then, one drop of filtered coldwater (7 °C) was added to the mixture directly to the scattering cell,the mixture was shortly gently shaken to homogenize it, andmeasurement continued immediately. A gradual increase of intensityas well as increase of fluctuations due to the growth of large scatteringobjects (mesoscale structures) is evident. Scattering angle θ = 45°.Scattering intensity is normalized to benzene scattering IB.

Figure 9. Ternary diagram of the TBA/water/PO system at T = 7 °C.The height of the vertical bars corresponds to scattering As frommesoscale structures measured at an angle of 45°. Commercial TBAand PO both of p.a. grade were used without special purification (used“as is”).

Figure 10. GC-MS (gas chromatography coupled with massspectrometry) elugram. Direct injection of TBA. Injection temperature280 °C. (2) 2-Methoxy-2-methylpropane; (3) 1-(1,1-dimethylethoxy)-2,2-dimethylpropane; (4) methylhexanol; (5) dodecane; (6) tetrade-cane; (7) hexadecane; (8) octadecane; (9) eicosane.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372732

water up to 40%. This supports our conclusions from lightscattering experiments that PO is not a source of mesoscalestructures in TBA/water mixtures. Another explanatory argu-ment comes from Figure 12 which shows a GC-MS elugram of

propylene oxide itself. It is clear that our PO contains exactlysuch long hydrophobic hydrocarbons which were identified asmesoscale structure makers in the case of the TBA/watermixture (Figure 11). These are the compounds numbered inFigure 12 as (6) hexadecane, (7) octadecane, (8) eicosane, (9)pentacosane, and (10) squalene. By adding propylene oxide tothe TBA/water mixture, the concentration of these “mesoscalestructure makers” increases, but since their concentration in POis low, one needs to add extremely high concentrations of POto regenerate the mesoscale structures after filtration. That iswhat was exactly observed in the light scattering experimentsdescribed above. It is actually the action of “mesoscale structuremakers”, not PO itself, that leads to the regeneration of theeffect after filtration. Other compounds that can be identified inthe GC-MS elugram of PO in Figure 12 are hydrophilic (1−5)and partly hydrophilic/partly hydrophobic compounds (11−14). These are expected to not be principally involved.

Figure 11. GC-MS (gas chromatography coupled with massspectrometry) elugrams. (A) TBA aqueous solution, c = 150 g/kg,before filtration (red) and after filtration (green). (B) Zoomed plotshowing peaks with small amplitudes in detail. Elugrams were obtainedby the HS-SPME method (headspace solid-phase microextraction).Desorption occurred in the injection port at 280 °C for 5 min. (1)TBA; (2) 2-methoxy-2-methylpropane; (3) 1-(1,1-dimethylethoxy)-2,2-dimethylpropane; (4) methylhexanol; (5) dodecane; (6) tetrade-cane; (7) hexadecane; (8) octadecane; (10) dimethylsilandiol; (11)hexa-methylcyclotetrasiloxane; (12) octamethylcyclopentasiloxane;(13) decamethylcyclopenta-siloxane; (14) dodecamethylcyclohexasi-loxane. Further parameters are given in Table 3.

Table 3. Summary of HS-SPME GC-MS Data (Headspace Solid-Phase Microextraction Gas Chromatography-MassSpectrometry)a

abundance (a.u.)

substance tr (min) unfiltered filtered content decrease, times

1 tert-butyl alcohol 1.571 1167463229 1247473634 12 2-methoxy-2-methylpropane 2.651 109786601 109113720 1.083 1-(1,1-dimethylethoxy)-2,2-dimethylpropane 5.071 7163730 7046643 1.094 methylhexanol 5.510 25969673 28566900 0.975 dodecane 15.309 44585954 50471 9436 tetradecane 20.719 53964376 29291 19697 hexadecane 25.559 25226268 10567 25518 octadecane 29.922 7438535 4137 192110 dimethylsilandiol 2.651 33449397 28902715 1.2411 hexamethylcyclotrisiloxan 4.680 12575989 10438659 1.2912 octamethylcyclotetrasiloxane 9.416 4615266 4194355 1.1813 decamethylcyclopentasiloxane 14.057 6176944 4885203 1.3514 dodecamethylcyclohexasiloxane 18.868 2811742 2563775 1.53

aFiltered and unfiltered aqueous solution of TBA, cTBA = 150 g/kg. tr is the retention time. Content decrease is calculated from the ratio ofabundances from unfiltered and filtered solutions, respectively.

Figure 12. GC-MS (gas chromatography coupled with massspectrometry) elugram. Direct injection of propylene oxide. Injectiontemperature 280 °C. (1) 2,4-Dimethylfuran; (2) 2-ethyl-3-methyl-1,3-dioxolane; (3) 1,1′-oxybis(2-propanol); (4) 2,2′-oxybis(1-propanol);(5) tripropylene glycol; (6) hexadecane; (7) octadecane; (8) eicosane;(9) pentacosane; (10) squalene; (11) dodecyl acrylate; (12) diisobutylphthalate; (13) bis(2-ethylhexyl) adipate; (14) bis(2-ethylhexyl)phthalate.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372733

To test the conclusions given above, we have decided topurify TBA and propylene oxide with special emphasis onremoving highly hydrophobic compounds such as longhydrocarbons. Details of purification are in the experimentalpart. Figure 13 shows a summary of experiments similar to

those presented in Figure 5 which were now repeated withpurified components (both TBA and PO purified). As can beseen in plot A, the effect is almost eliminated. A carefulinspection of very small values of scattering from mesoscalestructures As in the zoomed plot B indicates that the shape ofthe dependence is practically identical to that in plot A; thekinetics is very similar, and just the values are very small (2orders of magnitude smaller). We can therefore conclude thatwe deal here with some residual effect which is the same innature, just dramatically eliminated.Figure 14 shows the results of an experiment where purified

TBA was doped with known amounts of octadecane andafterward such TBA was mixed with water at a concentration ofcTBA = 150 g/kg. The presence of mesoscale structures isevident, and light scattering data are semiquantitatively similarto data on nonpurified TBA mixed with water. In terms ofscattering intensity from mesoscale structures, the closestmatch to data on unpurified TBA mixed with water is obtainedat 0.0053% octadecane concentration. It is well possible thatvarious hydrocarbons may have various abilities to form

mesoscale structure; nevertheless, the 0.0053% concentrationis quite a realistic number that can be expected in TBA of p.a.quality with 99.7% purity and the relative composition ofminority compounds is given by Figure 10. This is in contrastto unrealistic concentrations of propylene oxide necessary togenerate the effect of mesoscale structures.Figure 15 shows direct visualization of the long-lived

mesoscale structures by a NTA (nanoparticle tracking analysis)apparatus. Structures are clearly visualized by optical micro-scope as individual bright spots directly in the approximately 80μm wide laser beam (volume approximately 80 × 100 × 20μm3).It should be stressed that these structures/particles are not

being directly “imaged”. The particles act as point scattererswhose dimensions are below the Abbe limit, only above whichcan structural information and shape be resolved by opticalmicroscopy. Some spots are larger and some smaller, which canbe due to either different sizes of particles or different positionsin the beam. Video sequence capturing of the Brownian motionof the mesoscale particles can be found in the SupportingInformation. This method21−26 permits one to further track theBrownian pathways of individual particles over a suitable periodof time (typically several seconds), compute their diffusioncoefficients based on their mean square displacements,subsequently compute their hydrodynamic radii via theStokes−Einstein formula, and finally yield size distributions.Determination of unknown particle concentrations is alsopossible. A more detailed NTA analysis will be the subject of anupcoming separate experimental work. Current NTA dataconfirm our conclusions from the light scattering work, namely,(i) that these are real objects, not fluctuations, and (ii) thatthese are discrete objects (not bicontinuous phases with largecorrelation lengths).

4. CONCLUSIONSIt can be concluded that binary mixtures of TBA/water andternary mixtures of TBA/PO/water are homogeneous on largelength scales appreciably exceeding molecular dimensions anddo not possess mesoscale structures. Such structures form fromsubstantially more hydrophobic compounds that were exactlyanalytically identified by GC-MS and were found to be mostly

Figure 13. (A) Intensity of scattering from mesoscale structures inaqueous solutions of TBA (c = 330 g/kg) as a function of theconcentration of added propylene oxide. TBA solution was first cold-filtered at 7 °C such that As = 0, and subsequently, cold PO (T = 7 °C)was added. After each PO addition, the sample was maintained andmonitored at 7 °C for a couple of days (typically 2−6) while theintensity was still changing (vertical arrows). Red symbols correspondto mixtures prepared with untreated compounds (adapted from Figure5). Green symbols correspond to mixtures prepared with purifiedcompounds, as described in the Experiment section. (B) Zoomed plotshowing small values in detail. Scattering angle θ = 45°.

Figure 14. Scattering from mesoscale structures in aqueous solutionsof TBA (c = 150 g/kg). TBA was purified with emphasis on removal ofhydrophobic compounds, as described in the Experiment section, andafterward, a known amount of octadecane was added to TBA such thatthe resulting concentration of octadecane in TBA was (blue ○)0.00053%, (green ○) 0.0053%, and (red ○) 0.053%. Afterward, theoctadecane-doped TBA was mixed with water. Temperature T = 7 °C.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372734

long hydrocarbons such as dodecane (C12H26) and longer ones.These hydrophobes present in both TBA and PO as minoritycomponents segregate upon mixing with water into relativelywell-defined and long-term stable mesoscale structures withsizes on the order of ∼100 nm. Less hydrophobic minoritycomponents (typically containing oxygen or hydroxyl groups)were found to be also present in both TBA and PO but are notinvolved in the mesoscale structures. Propylene oxide itselfbelongs to this category. The observation of the generation ofmesoscale structures by addition of PO into TBA/watermixtures led to a conclusion that PO is a source of mesoscalestructures in TBA/water mixtures.16,17 In fact, the hydrophobescontained in PO are the source of mesoscale structuresdeveloping upon PO addition.Another experiment confirming the nature of observed

mesoscale structuring is shown in Figure 16. Following thediagram in Figure 9, we have chosen the mixture compositionwhere the effect reaches its maximum (cTBA = 18.3%, cPO =15.5%, cH2O = 66.2%). Scattering intensity is as high as AS ∼ 900at this composition. Then, the mixture was filtered via finefilters to completely eliminate AS and the sample wasimmediately monitored via light scattering.At the beginning, the overall intensity was equal to Af

(monomodal correlation curve, just “ordinary molecularscattering”), and then, a gradual reappearing of the slowmode (mesoscale structures) was monitored online. Theintensity grew relatively quickly within the first 30 min, andthen, a slow increase (another order of magnitude) wasobserved over the next 2 days. No increase was observedafterward. The final intensity was As = 12, i.e., 75 times lowerthan the initial intensity As ∼ 900. It can be easily verified by

weighing the filter before and after filtration that only a verysmall amount of filtered material is retained on the filter andhence practically all PO is passing through the filter to thefiltered solution in this experiment. We have thus demonstratedthat PO is not the source of mesoscale structures even in thatpart of the diagram which represents the maximum effect ofmesoscale structuring. Regarding the explanation of the partialreappearance of the structures upon filtration, two scenarios canbe envisaged. The first would be based on an assumption that

Figure 15. Microscopic images from the NTA (nanoparticle tracking analysis) experiment on an aqueous solution of TBA (c = 150 g/kg). TBA wasfirst purified, subsequently doped with a known amount of octadecane (0.0053% octadecane in TBA), and finally mixed with water (see Figure 14).Two still images from a video sequence are shown where the captured area is 80 × 100 μm2 while the focal depth is approximately 20 μm. Below areindividual Brownian motion tracks of several selected particles as monitored and evaluated by the NTA software. The experiment was performed atroom temperature. The whole video is available in the Supporting Information.

Figure 16. Online monitoring of a gradual reappearing of the slowmode (mesoscale structures) after filtration by a fine filter. The TBA/PO/H2O mixture composition corresponds to that with the maximumeffect of formation of mesoscale structures (cTBA = 18.3%, cPO = 15.5%,cH2O = 66.2%). I/IB is the benzene-normalized overall scattering

intensity. Fast mode (“ordinary molecular scattering”) intensity Af =0.22. Scattering angle θ = 45°.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372735

not all hydrophobes are initially incorporated in mesoscalestructures and some portion is still molecularly dissolved. Uponremoval of the mesoscale structures, the quasi-equilibrium isdistorted and molecularly dissolved hydrophobes segregate intonew mesoscale structures. The second scenario would be basedon an assumption that the structures present in the mixtureprior to filtration are partially retained on the filter and partiallybroken into small fragments that pass the filter and that couldcoalesce later into larger structures. However, we do not seeany indication of broken fragments very shortly after filtrationand therefore the first scenario seems to be more probable.Regarding the dependence of the degree of large-scale

structure formation on mixture composition, it can beconcluded as follows. In binary mixtures of TBA and water(Figure 2), a maximum can be observed around 150 g/kg, whilethe effect gradually ceases for TBA concentrations larger than500 g/kg. When TBA concentration is high, hydrophobiccompounds can be sufficiently solubilized such that they can bemolecularly dissolved. Upon decreasing TBA content (increas-ing water content) in the mixture, hydrophobes become moreand more segregated into large-scale structures. Going withTBA concentration further down (below 150 g/kg), thescattering from these structures become eventually weaker,since the decrease of TBA concentration means also a decreaseof the absolute value of the concentration of hydrophobes inthe mixture. Gradual addition of water into the TBA/POmixture (Figure 7) decreases the ability of the mixture tosolubilize hydrophobes molecularly, and therefore, these aremore and more segregated. When the water contentapproaches 70%, the effect reaches its maximum (Figure 9).Further increases of water content go at the expense of theconcentrations of TBA and PO and hence also at the expenseof the concentration of hydrophobes as a “building material” formesoscale structures. A more detailed discussion on the wholeternary diagram shown in Figure 9 is not much possible, sincethis diagram implicitly reflects not only different mixturecompositions in terms of TBA/PO/water but also differenttypes and quantities of hydrophobes contained in TBA and PO,respectively. The role of TBA and/or PO in the stabilization ofmesoscale structures cannot be therefore assessed from Figure9. A separate detailed study is necessary for this purpose.The effect investigated in this work can be characterized as a

mesoscopic solubility. Hydrophobic compounds are neithermacroscopically phase separated nor molecularly dissolved inTBA/water mixtures. While it is clear that water is the source ofthe segregation and hydrophobes are “the building material”,the role of the main solute (TBA in our case) in thestabilization of discrete mesoscale structures will be investigatedin detail in the near future. Results obtained on the givensystem have much broader implications, since, as described inthe Introduction, practically all aqueous mixtures and solutionsof organic molecules do exhibit large-scale structures withsimilar characteristics.7,9,11−17 It is well possible that in otheraqueous systems (aqueous solutions or mixtures of otherorganic molecules), less hydrophobic compounds may also beinvolved, since this may depend on the ability of the mainsolute component (as TBA in this work) to solubilize minoritycomponents present as impurities. Since TBA is a good solventfor hydrophobes, only very hydrophobic compounds (longhydrocarbons C12H26−C30H62) segregate upon mixing TBAwith water. Compounds with a lower ability to solubilizehydrophobes may yield mesoscale structures upon mixing withwater, while far less hydrophobic minority compounds

(impurities) could also be involved in the formation of thesestructures.In summary, on the basis of the results of this and our

previous work, we can conclude that aqueous mixtures oforganic compounds as used in everyday life and researchpractice do possess long-lived large-scale inhomogeneities in aquite substantial manner, even research-grade compounds ofp.a. quality and even after special in-lab purification procedures.Under conditions strongly favoring mesoscale segregation, veryspecial and sophisticated procedures would be needed tocompletely eliminate the effect by producing compounds ofvirtually absolute purity. Under conditions strongly favoringmesoscale segregation, a partial reappearance of mesoscalestructures can be seen even after complete elimination of theeffect by nanopore filtration (Figure 16). Many questionsremain open, especially those related to the stabilizationmechanism of these mesoscale structures preventing macro-scopic phase separation. Another related issue is the presence ofsimilar structures in solutions of ionic species (organic andinorganic salts)6−10 and polyions27−31 which may or may nothave similar origin.

■ ASSOCIATED CONTENT

*S Supporting InformationFigure showing the light scattering characterization of theunfiltered sample from Figure 11 and video sequence capturingthe Brownian motion of mesoscale particles in aqueous solutionof TBA. This material is available free of charge via the Internetat http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Author*E-mail: marsed@saske.sk.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

Support of the Slovak Research and Development Agency(Grant No. 048610) and Scientific Grant Agency VEGA (GrantNo. 2/0182/14) is acknowledged. This work was realizedwithin the frame of the project “Build up of the infrastructure ofthe Centre of excellence for progressive materials with nano-and sub-micrometer structure, Project No. 26220120035”,which is supported by the Operational Program “Research andDevelopment” of the Slovak republic financed throughEuropean Regional Development Fund. The authors are alsothankful to P. Bartak from Palacky University Olomouc foracquiring and interpreting GCMS elugrams.

■ REFERENCES(1) Vuks, M. F.; Shurupova, L. V. The Scattering of Light and PhaseTransition in Solutions of Tertiary Butyl Alcohol in Water. Opt.Commun. 1972, 5, 277−278.(2) Beer, C. W., Jr.; Jolly, D. J. Comments on “The Scattering ofLight and Phase Transition in Solutions of Tertiary Butyl Alcohol inWater. Opt. Commun. 1974, 11, 150−151.(3) Euliss, G. W.; Sorensen, C. M. Dynamic Light Scattering Studiesof Concentration Fluctuations in Aqueous t-butyl Alcohol Solutions. J.Chem. Phys. 1984, 80, 4767−4773.(4) Bender, T. M.; Pecora, R. A Dynamic Light Scattering Study ofthe tert-Butyl Alcohol-Water System. J. Phys. Chem. 1986, 90, 1700−1706.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372736

(5) Bender, M. T.; Pecora, R. Dynamic Light Scattering Measure-ments of Mutual Diffusion Coefficients of Water-Rich 2-butoxyethanolWater Systems. J. Phys. Chem. 1988, 92, 1675−1677.(6) Georgalis, Y.; Kierzek, A. M.; Saenger, W. Cluster Formation inAqueous Electrolyte Solutions Observed by Dynamic Light Scattering.J. Phys. Chem. B 2000, 104, 3405−3406.(7) Sedlak, M. Large-Scale Supramolecular Structure in Solutions ofLow Molar Mass Compounds and Mixtures of Liquids: I. LightScattering Characterization. J. Phys. Chem. B 2006, 110, 4329−4338.(8) Sedlak, M. Large-Scale Supramolecular Structure in Solutions ofLow Molar Mass Compounds and Mixtures of Liquids: II. Kinetics ofthe Formation and Long-Time Stability. J. Phys. Chem. B 2006, 110,4339−4345.(9) Sedlak, M. Large-Scale Supramolecular Structure in Solutions ofLow Molar Mass Compounds and Mixtures of Liquids. III. Correlationwith Molecular Properties and Interactions. J. Phys. Chem. B 2006, 110,13976−13984.(10) Sedlak, M.; Rak, D. Large-Scale Inhomogeneities in Solutions ofLow Molar Mass Compounds and Mixtures of Liquids: Supra-molecular Structures or Nanobubbles? J. Phys. Chem. B 2013, 117,2495−2504.(11) Jin, F.; Ye, J.; Hong, L.; Lam, H.; Wu, C. Slow Relaxation Modein Mixtures of Water and Organic Molecules: SupramolecularStructures or Nanobubbles? J. Phys. Chem. B 2007, 111, 2255−2261.(12) Jin, F.; Li, J.; Ye, X.; Wu, C. Effects of pH and Ionic Strength onthe Stability of Nanobubbles in Aqueous Solutions of Alpha-Cyclodextrin. J. Phys. Chem. B 2007, 111, 11745−11749.(13) Hagmeyer, D.; Ruesing, J.; Fenske, T.; Klein, H. W.; Schmuck,C.; Schrader, W.; da Piedade, M. E. M.; Epple, M. Direct ExperimentalObservation of the Aggregation of Alpha-Amino Acids into 100−200nm Clusters in Aqueous Solution. RSC Adv. 2012, 2, 4690−4696.(14) Habich, A.; Ducker, W.; Dunstan, D. E.; Zhang, X. Do StableNanobubbles Exist in Mixtures of Organic Solvents and Water? J. Phys.Chem. B 2010, 114, 6962−6967.(15) Li, Z. Y.; Cheng, H.; Li, J. Y.; Hao, J. K.; Zhang, L.; Hammouda,B.; Han, C. C. Large-Scale Structures in Tetrahydrofuran - WaterMixture with a Trace Amount of Antioxidant Butylhydroxytoluene(BHT). J. Phys. Chem. B 2011, 115, 7887−7895.(16) Subramanian, D.; Anisimov, M. A. Resolving the Mystery ofAqueous Solutions of Tertiary Butyl Alcohol. J. Phys. Chem. B 2011,115, 9179−9183.(17) Subramanian, D.; Ivanov, D. A.; Yudin, I. K.; Anisimov, M. A.;Sengers, J. V. Mesoscale Inhomogeneities in Aqueous Solutions of 3-methylpyridne and Tertiary Butyl Alcohol. J. Chem. Eng. Data 2011,56, 1238−1248.(18) Provencher, S. W. A Constrained Regularization Method forInverting Data Represented by Linear Algebraic or DifferentialEquations. Comput. Phys. Commun. 1982, 27, 213−219.(19) Jakes, J. Testing of the Constrained Regularization Method ofInverting Laplace Transform on Simulated Very Wide QuasielasticLight-Scattering Auto-Correlation Functions. Czech. J. Phys. 1988, 38,1305−1316.(20) Stepanek, P. Data Analysis in Dynamic Light Scattering.Dynamic Light Scattering. The Method and Some Applications;Clarendon: Oxford, U.K., 1993; Chapter 4, pp 177−240.(21) Carr, B.; Hole, P.; Malloy, A.; Nelson, P.; Wright, M.; Smith, J.Applications of Nanoparticle Tracking Analysis in NanoparticleResearch − a Mini Review. Eur. J. Parenter. Pharm. Sci. 2009, 14,45−50.(22) Syvitski, J. P. M. Principles, Methods, and Application of ParticleSize Analysis; Cambridge University Press: Cambridge, U.K., 1991.(23) Walker, J. G. Improved Nano-Particle Tracking Analysis. Meas.Sci. Technol. 2012, 23, 065605.(24) Filipe, V.; Hawe, A.; Jiskoot, W. Critical Evaluation ofNanoparticle Tracking Analysis (NTA) by NanoSight for theMeasurement of Nanoparticles and Protein Aggregates. Pharm. Res.2010, 27, 796−810.(25) Malloy, A.; Carr, B. Nanoparticle Tracking Analysis - The Halo(TM) system. Part. Part. Syst. Charact. 2006, 23, 197−204.

(26) Saveyn, H.; De Baets, B.; Thas, O.; Hole, P.; Smith, J.; Van derMeeren, P. Accurate Particle Size Distribution Determination byNanoparticle Tracking Analysis Based on 2-D Brownian DynamicsSimulation. J. Colloid Interface Sci. 2010, 352, 593−600.(27) Lin, S. C.; Lee, W. I.; Schurr, J. M. Brownian Motion of HiglyCharged Poly(L-Lysine). Effects of Salt and Polyion Concentration.Biopolymers 1978, 17, 1041−1064.(28) Schmitz, K. S. Macroions in Solution and Colloidal Suspension;VCH Publishers: New York, 1993.(29) Sedlak, M. The Ionic Strength Dependence of the Structure andDynamics of Polyelectrolyte Solutions as Seen by Light Scattering:The Slow Mode Dilemma. J. Chem. Phys. 1996, 105, 10123−10133.(30) Sedlak, M. What Can Be Seen by Static and Dynamic LightScattering in Polyelectrolyte Solutions and Mixtures? Langmuir 1999,15, 4045−4051.(31) Sedlak, M. Structure and Dynamics of Polyelectrolyte Solutionsby Light Scattering. Physical Chemistry of Polyelectrolytes; MarcelDekker: New York, 2001; Chapter 1, pp 1−58.

The Journal of Physical Chemistry B Article

dx.doi.org/10.1021/jp500953m | J. Phys. Chem. B 2014, 118, 2726−27372737