Hydrophobin: fluorosurfactant-like properties without fluorine

Soft Matter

PAPER

aVTT – Technical Research Centre of Finland,

E-mail: roberto.milani@vtt.; markus.linderbCenter for Nano Science and Technology@P

Giovanni Pascoli 70/3, I-20133 Milano, ItalcNFMLab-DCMIC “Giulio Natta”, Politecni

Milano, Italy. E-mail: pierangelo.metrangolodNational Nanotechnology Laboratory (NNL)

U.O.S. Lecce, via per Arnesano, I-73100, ItaeUniversita del Salento, Dipartimento di M

Collegio Fiorini Campus extraurbano, via pfAalto University, School of Chemical Techn

00076 AALTO, Espoo, Finland

† Electronic supplementary informationisoelectric point; tensiometric measuremdescription and supplementary images oillustrating formation of the interfaciauorous droplets in aqueous environmen

‡ Dedicated to Prof. Maurizio Prato for hi

Cite this: Soft Matter, 2013, 9, 6505

Received 26th March 2013Accepted 14th May 2013

DOI: 10.1039/c3sm51262b

www.rsc.org/softmatter

This journal is ª The Royal Society of

Hydrophobin: fluorosurfactant-like properties withoutfluorine†‡

Roberto Milani,*ab Evanthia Monogioudi,a Michele Baldrighi,c Gabriella Cavallo,bc

Valentina Arima,d Lucia Marra,d Alessandra Zizzari,e Ross Rinaldi,de Markus Linder,*af

Giuseppe Resnati*bc and Pierangelo Metrangolo*abc

The stabilization of fluorous oil droplets in aqueous environment is a critical issue in the preparation of

emulsified systems for biomedical applications and in emulsion polymerization technology, due to the

extreme immiscibility of aqueous and fluorous phases. We present here a detailed study on the

behavior of the hydrophobin HFBI, i.e. a small natural protein endowed with exceptional surface

activity, at the interface between aqueous and fluorous phases. HFBI behaves as an efficient and

sustainable surfactant at remarkably low concentrations and forms a strong and elastic film at the

interface between the two phases. We also show proof-of-concept experiments on the use of HFBI as a

surfactant in fluorous oil/water emulsified systems and in microfluidic circuits.

Introduction

The term uorous has been coined to describe the unique phasebehavior of highly uorinated molecules, molecular fragments,materials, and media. Fluorous chemistry has been widelyadopted by the chemical community as an effective alternativeto traditional synthetic procedures, catalytic systems, andseparation technologies,1 thanks to the outstanding propertiesof peruorinated compounds. In fact, uorous compounds areusually highly dense, nonpolar liquids endowed with excep-tional chemical and biological inertness, and high thermalstability.2 Furthermore they are amphiphobic, i.e. at the sametime hydrophobic and oleophobic, and they are usuallyimmiscible with both water and organic solvents at roomtemperature, as well as with ionic liquids. This behavior,

Tietotie 2, Espoo, FI-02044 VTT, Finland.

@vtt.

oliMi, Istituto Italiano di Tecnologia, via

y

co di Milano, Via Mancinelli 7, I-20131

@polimi.it; [email protected]

, CNR – Institute of Nanoscience (NANO),

ly

atematica e Fisica “E. De Giorgi”, ex

er Arnesano, 73100, Lecce, Italy

ology, Kemistintie 1, P.O. Box 16100, FI-

(ESI) available: Measurement of HFBIents at varying pH values; detailed

f the microuidic circuit; short moviel HFBI lm and non-coalescence oft. See DOI: 10.1039/c3sm51262b

s 60th birthday.

Chemistry 2013

commonly known as “uorophobic effect”,3 results from thetendency of peruoroalkyl chains to segregate in order to avoidunfavorable interactions of uorine atoms with other elements,sometimes giving rise to unique interfacial structures,4 and hasbeen exploited in conjunction with biological materials inseveral biotechnology applications.5

The stabilization of uorous oil droplets in aqueous envi-ronment is a critical issue in the preparation of emulsiedsystems for biomedical applications and in emulsion polymer-ization technology. For this purpose, peruorosulfonic acids(PFSAs) and peruorocarboxylic acids (PFCAs), and their saltshave been largely used as surfactants.2c,6 However, hugeconcerns have been raised around their use, sincemany of themare persistent organic pollutants (POPs).7 In particular, per-uorooctanesulfonic acid (PFOS) and peruorooctanoic acid(PFOA) are of greater concern as they bioaccumulate in the foodchain. Although there are insufficient data available on acutetoxicity in humans to draw conclusions, numerous reportsallude to their toxicity.8 This has drawn considerable interestfrom governments and regulatory agencies all over the worldand agreements and regulations have been issued in order tolimit the production of some of these uorosurfactants.9

As a consequence of this, the leading uoropolymer compa-nies worldwide are actively searching more sustainable surfac-tants, generally containing either shorter peruorinated groupsor “weak” degradable points like methylene, methyne or ethergroups.10 Examples include, among others, ammonium 4,8-dioxa-3H-peruorononanoate (ADONA),11 peruoropolyethercarboxylic salts,12 lowmolecular weightuorinated sulfonates inwhich the uorinated carbon chains are interrupted either bymethylene (–CH2) units or by an ether (e.g., –O–) linkage,13 uo-rosurfactants containing shorter peruorinated end groups like

Soft Matter, 2013, 9, 6505–6514 | 6505

Soft Matter Paper

CF3–O or (CF3)2N linked to a linear alkyl sulfonate,14 or basedon bolaamphiphilic poly(uorooxetane) having short per-uoroalkyl chains.15 In this context, the possible use of naturallyoccurring surfactants looks extremely appealing.

Literature examples on the use of natural surfactants asstabilizers of peruorocarbon/water dispersions mainlyinvolve phospholipids, polyglycerol esters, and albumin, forapplications such as injectable oxygen carriers,16 pulmonarydrug delivery,17 ultrasonic and MRI imaging,18 and, to a lesserextent, emulsion polymerization of uorinated monomers.19

Interestingly, it was shown that saturation of gaseous phaseswith peruorocarbon vapors can accelerate the adsorptiondynamics of phospholipids at the interface with aqueoussolutions.20 In this scenario, keen to nd a safer and effectivesurfactant, we decided to study the behavior at the aqueous–uorous interface of naturally occurring surfactants knownas hydrophobins (HFBs),21 which are a family of small(�10 kDa), highly surface-active, lm-forming proteinsproduced by lamentous fungi. In nature they play a role ascoating or protective agents, adhesion promoters, and surfacemodiers.

It was observed that these proteins have a structure typical ofamphiphilic, Janus-like molecules.22 One continuous portion oftheir exposed surface is composed exclusively of amino acidswith lipophilic side chains and is commonly designated as thehydrophobic patch,23 while the remaining protein surface typi-cally exposes hydrophilic side chains. Moreover, all HFBsfeature a pattern of eight Cys residues in their primary struc-ture, giving rise to four disulde bridges in the protein core24

which are responsible for the remarkable structural stability ofthese proteins. Indeed, HFBs are able to withstand tempera-tures close to the boiling point of water without incurring indenaturation.25

These features have attracted considerable interest in theuse of HFBs as lm-forming surface active agents at interfaceswith aqueous phases. Over time, HFBs have been applied asfoaming26 and antifouling agents,27 adhesion promoters,28 fattyoil/water emulsion stabilizers and enhancers,29 as well as inpersonal care and biomedical applications.30 A few studieshave also shown that, thanks to their hydrophobic patch, HFBspossess the capability to assemble into lms on solid uori-nated surfaces from aqueous solutions, in spite of theomnirepellency of such surfaces.25a,29a,31 In one recent case,HFB-stabilized peruoroalkane/water emulsions have beenused as templates for the growth of mineral shells by Bokerand coworkers.32 With the idea that HFBs are natural non-toxiccompounds that may not only be well-suited to biomedicalapplications,33 but which may also represent a more sustain-able alternative to synthetic uorosurfactants commonlyemployed e.g. in emulsion polymerization or in microuidicsystems, we decided to investigate systematically the behaviorof hydrophobins at the interface between aqueous and uo-rous phases.

In this article, we report the behavior of the hydrophobinHFBI at the interface between aqueous phases and several u-orous uids such as peruorooctyl bromide (PFOB), the per-uoropolyether Galden� SV90 (GSV90), and Fluorinert� FC-70

6506 | Soft Matter, 2013, 9, 6505–6514

(tris(peruoropentylamine), FC-70). We study the shearmechanical properties of the lm formed thereby, and presentproof-of-concept experiments on the application of HFBI as adroplet stabilizer for uorous-oil-in-water emulsions and inmicrouidic systems.

ExperimentalMaterials

HFBI (molecular weight 7.5 kDa) was produced usingrecombinant strains of T. reesei, puried by RP-HPLC asdescribed previously34 and lyophilized before use. Galden�SV90 (GSV90) is a low molecular weight peruoropolyether uid(90 �C boiling point) produced by Solvay Solexis. Peruorooctylbromide (1-bromoheptadecauorooctane, PFOB), Fluorinert�FC-70 (tris(peruoropentylamine), FC-70) and Zonyl FSN-100were purchased from Sigma Aldrich.

Interfacial tension measurement

Tensiometric measurements were performed on a CAM 200(KSV Instruments Ltd.) and operated with the soware CAM2008 supplied with the instrument. All measurements wereperformed by injecting a 15 ml droplet of uorous liquid into900 ml of a freshly prepared HFBI solution in mQ water or in20 mM aqueous buffer (citrate, pH 4; sodium acetate, pH 5.5;phosphate buffer, pH 7; boric acid, pH 10) in a quartz cuvette.The interfacial energy was derived from the tting of dropletshape by the Young–Laplace method.

Interfacial shear rheology

Measurements were performed on a stress-controlled rheom-eter (AR-G2, TA Instruments, UK) equipped with a Pt–Ir duNouy ring (13 mm diameter), amed prior to use. Beforestarting each experiment the inertia of the instrument waschecked, and the response of the instrument was mappedwithin the experimental settings in order to certify a linearresponse. A 10 hours time sweep measurement was rst per-formed (strain 0.05%, frequency 0.1 Hz), followed by afrequency sweep (strain 0.05%, frequency range 0.01–10 Hz), astrain sweep (strain range 10�3 to 1%, frequency 0.1 Hz) andnally an additional time sweep step of 16 hours. All experi-ments were performed at 25 �C, and the system was allowed toequilibrate for 1 min before each measurement. GSV90 (33 g)was poured in a 65 mm diameter round glass vessel, and thering was positioned 0.5 cm below the liquid surface. 100 ml of afreshly prepared 1.00 mg ml�1 solution of HFBI in mQ waterwere then spread on the surface using a 10 mL glass syringe.The surface was nally covered with 30 ml of mQ water beforeraising the ring to the interface. The series of experiments wasalways started approximately 5 minutes aer spreading theHFBI solution on the uorinated phase.

Fluorocarbon emulsions

Emulsions of GSV90 and PFOB (both 5% v/v) in mQ water wereprepared simply by treatment of the phase-separated mixtureswith a homogenizer for 10 minutes (Heidolph DIAX 900 at

This journal is ª The Royal Society of Chemistry 2013

Paper Soft Matter

power 4, Sigma-Aldrich), at atmospheric pressure and in asingle cycle, testing different HFBI concentrations (0.01, 0.10and 1.00 mg ml�1). Similar emulsions were prepared by tipultrasonication for 10 minutes at 30% amplitude on a Vibra-Cell VCX750 ultrasonicator (750 W, Sonics & MaterialsInc., USA).

The particle size distribution of emulsions was measured bylaser diffraction (Beckman Coulter LS230, Brea, CA). Themeasurements were performed setting the dispersing uid(water) refractive index as 1.33, and the sample refractive indexas 1.280 and 1.305 for GSV90 and PFOB, respectively. As anindicative parameter of the emulsifying activity of the proteins,the volume-weighed geometric mean particle diameter wasdetermined from the particle size distribution of fresh andstored emulsions. Two measurements were taken for eachbatch. For the storage stability study, samples were analyzedaer vortexing from each emulsion at 0, 2, 5, 10, 15 and 30 days.Samples were stored at 4 �C. Emulsion imaging was performedon an Olympus BX40 optical microscope (Olympus, CenterValley, Pennsylvania, USA). Small emulsion amounts wereplaced directly onto a glass microscope slide and viewed under10–50� magnication.

Droplet microuidics

The microreactor employed for droplet microuidics was aglass-based microdevice with overall dimensions of (2.5 � 6.0)cm. A B270 glass sheet with photoresist and chrome layers(TELIC, USA) was used as a substrate. Uncoated glass waschosen for its chemical inertness, hydrophilicity, and wellestablished fabrication methods. The device is schematicallyillustrated in Fig. 8a, and a detailed description of the circuitand its fabrication can be found in the ESI.† Droplets weregenerated from ow focusing (FF) geometry, rst proposed byAnna et al.35 and Dreyfus et al.36 In general, in this congu-ration the droplet size is determined by the ow rates of thecontinuous (aqueous) and dispersed (uorous) phases and bytheir ratio, in addition to channel geometries and viscosities ofthe two phases. The ow rate of each uid was controlled byan independent pump (KD Scientic, Model 200 Series). Theresults reported here were obtained at a water/FC-70 ow ratioof 20. A speed imaging system (NIKON mod. DS-5MC camera)with 8 fps acquisition rate was used. Several experiments wereperformed to demonstrate the pH dependence of HFBI effi-ciency as a droplet stabilizer and to compare it with a commonuorosurfactant, Zonyl FSN-100. The aqueous phase, injectedfrom the side channels of the ow focusing device (inlet 1 ofFig. 8a), was a HFBI solution at different pH values (7.0, 10.3and 13.0 by addition of NaOH), while the oil phase (FC-70) wasinjected from the main channel (inlet 2 of Fig. 8a). HFBI andZonyl FSN-100 were solubilized in the aqueous phase at aconcentration of 0.10 mg ml�1. Aer owing into the micro-uidic circuit, droplets were analyzed in a collector chamber.The apparent sizes of the droplets were extracted by imageprocessing (NIS_Elements BR soware); the errors werecalculated as standard deviations over 30 droplets in thecollector.


Results and discussionSurface activity and interface lm formation

The surface activity of HFBI at aqueous–uorous liquid inter-faces was studied by the pendant drop technique. Single drop-lets of the three different uorinated oils were injected intoaqueous HFBI solutions at different protein concentrations,and the evolution of the interface energy over time was moni-tored (Fig. 1). The interface tension decreased by about15–20 mN m�1 during 1 hour, with respect to the originaltension values in the absence of the protein (59 mNm�1 for FC-70, 51mNm�1 for GSV90 and 50mNm�1 for PFOB). Most of theinterface tension drop occurred in the rst few seconds orminutes aer droplet injection, with faster kinetics at higherprotein concentrations. At 0.01mgml�1, inections were clearlyobserved in the interface tension curves within the rst 5–10minutes of the experiment, although this effect was no longervisible at higher HFBI concentrations. The interface tensionsubsequently kept decreasing at a much slower rate, withoutreaching a stable value within the timeframe of the experiment.Although it is assumed that a stable state could be reached at alater time, excessive deformation of the droplets eventuallyprevented a correct droplet shape tting.37

All these observations may suggest that the formation of theprotein lm around the uorous oil droplets occurs by twomain mechanisms: (i) adsorption of the protein to the inter-face, which is a fast process essentially dominated by diffusionand therefore highly dependent on the protein concentration;(ii) subsequent rearrangement of the adsorbed protein lm,associated with protein–protein interaction and leading to acloser packing of the overall structure.38 The effect of rear-rangement could be observed at low protein concentration(0.01 mg ml�1) by the presence of inections in the chartsshowing the time evolution of the interfacial tension. For HFBIconcentrations of 0.10 mg ml�1 or higher, interface saturationoccurred quickly and therefore the structural reorganization ofthe protein lm did not have major effects on the interfacetension. These results are in reasonable accordance with thosereported by Boker and coworkers32 for a different hydrophobinat water–peruoroalkane interfaces, and expand on their workby showing that an essentially similar interface energy reduc-tion is observed for peruoroalkanes and other highly uori-nated oils.

Analogous measurements were performed for GSV90 drop-lets in buffered solutions at pH 4, 5.5, 7, and 10 (HFBIconcentration 0.10 mg ml�1), showing results similar to thosepresented above. No signicant variations were observed at pH5.5, i.e. close to the isoelectric point of HFBI (calculated 5.7,experimental 6.0 from zeta potential measurement, see ESI†).

Remarkably, the HFBI lm assembled at the aqueous–uo-rous interface can be easily visualized by simply drawing back asmall portion of the uorous oil into the syringe needle (seeFig. 2, top droplet), leading to the formation of wrinkles on thedroplet surface. Already at 0.10 mg ml�1 protein concentration,this effect could be observed aer mere seconds from dropletinjection, while aging in the scale of minutes was necessary for0.01 mg ml�1 concentration.

Soft Matter, 2013, 9, 6505–6514 | 6507

Fig. 1 Plot of the interface tension as a function of time between droplets of different fluorous oils (Fluorinert� FC-70, Galden� SV90 and perfluorooctyl bromide)and aqueous solutions of HFBI at varying concentrations (0.01, 0.10, and 1.00 mg ml�1).

Fig. 2 The formation of an interface HFBI film prevents the coalescence of twoGSV90 droplets in aqueous HFBI solution at 0.10 mg ml�1 protein concentration(droplet size: �15 ml).

Soft Matter Paper

The protein lm appeared to be relatively strong and elastic,effectively preventing the coalescence of GSV90 droplets inaqueous environment even under moderate pressure (see Fig. 2and short movie available in the ESI†).

Fig. 3 Surface shear rheology as a function of time for a HFBI film formed at theinterface between GSV90 and water.

Rheology of the interface lm

The formation of a proper interfacial lm is a clearly dis-tinguishing feature and a potential advantage of the use ofHFBs with respect to other surfactants. For example, it waspreviously observed that the surface elasticity of HFB lms atthe air–water interface is markedly superior to that commonlyobserved for other proteins, leading to a considerableenhancement in bubble stability.39 This behavior would behighly desirable e.g. in the stabilization of uorous gas bubblesfor use as contrast agents in ultrasound imaging, where bubbleor droplet stability is one critical issue.18a We aimed therefore atobtaining a quantitative description of the mechanical featuresof the lm formed at the water–GSV90 interface by interfacialshear rheology in the du Nouy ring conguration (see Fig. 3).

The elastic shear modulus G0 increased steeply during therst 15 minutes of the experiment and reached a pseudo-

6508 | Soft Matter, 2013, 9, 6505–6514

plateau within about 90 minutes, rising very slowly during thefollowing nine hours. Consistent with the results reportedabove from tensiometric measurements, this suggests a rst,relatively fast protein lm formation until saturation of theinterface, followed by a slow, continued build-up andstrengthening, possibly accompanied by lm thickening. Thenal value obtained for the elastic shear modulus G0 was around0.5 Nm�1, with variations within 10% for experiment replicates.The viscous shear modulus G00, while constantly about an orderof magnitude lower than G0, decreased in 2 hours to a value ofabout 0.03 N m�1 and remained substantially constant there-aer. The damping factor tan d decreased somewhat exponen-tially to a value of 0.05 in the same time, thus testifying theessentially elastic and gel-like nature of the lm. These valuesare in good accordance with those reported for HFBI lms at theair–water interface,39 and are about two orders of magnitudehigher than those reported at air–water and various oil–waterinterfaces for phospholipid monolayers40 and proteins such asBSA, insulin, lysozyme,41 b-lactoglobulin42 and b-casein,43 and


Fig. 5 Creaming of fluorous oil emulsions (5% v/v oil in 0.10 mg ml�1 aqueousHFBI solution). Images were taken (a) immediately after vortexing, and after (b)5 min, (c) 45 min, and (d) 12 h.

Paper Soft Matter

still signicantly higher than those recently published for silkbroin.44 The small spikes observed in the curves appeared inseemingly random patterns throughout experiment replicates,and were likely due to the sticky protein lm being occasionally‘locked’ in place, or to the formation of air bubbles that origi-nated from clathrates in the uorous phase.

The frequency and strain sweep measurements are repor-ted in Fig. 4. There appeared to be little dependence of themoduli on these parameters within the investigated regions,conrming that the time dependency experiments were per-formed in the region of linear mechanical behavior of theprotein lm. The low dependence of both elastic andviscous moduli on frequency and strain, together with themuch higher value of G0 in comparison to G00, again showedthat the interfacial lm had essentially elastic propertiestypical of a gel-like structure, which could withstand shearstrains at least up to 1% without incurring in breakdown ofthe network established by the physical interactions betweenthe proteins.

Fluorous oil-in-water emulsions

The emulsifying capabilities of HFBI were tested for suspen-sions of GSV90 and PFOB in aqueous HFBI solutions, preparedby homogenizer treatment and tip ultrasonication at varyingprotein concentrations (0.01, 0.10, and 1.00 mg ml�1). Thedimensional stability of the emulsions was studied over onemonth by laser diffraction and optical microscopy. In theabsence of the protein, stable emulsions could not be formed byeither procedure. Immediately aer emulsion preparation, nostriking differences could be observed for different uorinatedoils or protein concentrations. However, creaming occurredshortly aer, leading to the formation of a milky layer at thebottom of the vial and leaving an opalescent aqueous phaseabove. The time necessary for creaming varied from a few

Fig. 4 Frequency (a) and strain (b) dependence of the shear moduli and loss facto


minutes to a few hours and increased for higher proteinconcentration and lower oil density (see Fig. 5), althoughcomplete sedimentation of the uorous droplets at the bottomof the vial, i.e. obtainment of a clear aqueous phase, generallyrequired hours or days. However, emulsions could be immedi-ately and repeatedly reformed upon simple manual shaking ormild vortexing.

At 0.01 mg ml�1 HFBI concentration, the stability of theemulsions proved to be rather low regardless of the specicuorous oil and preparation method used. In these cases, thesmall amount of protein present only allowed the coating oflarge droplets (around 100 mm in size), which were prone tocoalescence.

The highest protein concentration of 1.00 mg ml�1 insteadyielded occulated emulsions, as evidenced by contrast phasemicroscopy (see Fig. 6a). In these samples, individual dropletsof few- and sub-micron diameter aggregated into clusters ofvarying sizes. This peculiar result was likely due to adhesionbetween HFBI lms formed around different oil droplets, andmediated by excess protein still present in solution, i.e. deple-tion occulation.45 This hypothesis is consistent with the

r for HFBI films formed at the interface between GSV90 and water.

Soft Matter, 2013, 9, 6505–6514 | 6509

Fig. 6 Microscopic images of PFOB emulsions obtained by the sonication method, recorded at 5 days aging and containing (a) 1.00 mgml�1 HFBI, or (b) 0.10 mgml�1

HFBI in the aqueous phase.

Soft Matter Paper

observation that the creamed layer could be easily resuspendedby applying moderate shear forces, and in line with the gel-ication previously observed in oil emulsions for hydrophobinconcentrations as high as 1% in weight.38

At 0.10 mg ml�1 HFBI concentration, the emulsions did notocculate (Fig. 6b) and were dimensionally stable over a periodof at least 30 days, as it could be seen from laser diffraction datain Fig. 7a. The average diameter of the droplets measured bylaser diffraction appeared to critically depend on the emulsi-cation method, as homogenizer treatment yieldedmean dropletsizes of about 30 and 20 mm for GSV90 and PFOB, respectively,while the corresponding values were around 10 mm uponpreparation by ultrasonication.

Freshly prepared emulsions generally displayed dual distri-butions of droplet sizes (see e.g. Fig. 7b for a GSV90 emulsionprepared by a homogenizer). One rst, narrower population ofdroplet sizes was centered at lower values and was essentiallyconserved for at least one month; the second populationspanned a much broader range of sizes and was centered at

Fig. 7 (a) Mean droplet size as a function of time for 5% v/v fluorous oil emulsiosonicator; (b) droplet size distributions at different aging times for a 5% v/v GSV90

6510 | Soft Matter, 2013, 9, 6505–6514

substantially higher values, but quickly disappeared withsample aging. This second distribution is assumed to be due tothe incorporation of air bubbles during emulsion preparation,which eventually collapsed and disappeared, as supported bythe fact that no macroscopic phase separation was observed inthe samples. For this reason, the apparent average droplet sizeis in general signicantly higher for freshly prepared emulsions,and the actual size distribution of uorous oil droplets in thesesamples is correctly described by the lower-size (10–30 mm),narrow-shaped population, which remained essentially stableover a period of at least one month.

This remarkable stabilization of uorocarbon-in-wateremulsions brought forth by HFBs can only in part be attributedto the lowering of interfacial tension. Indeed, the energyreduction is somewhat modest when compared to that of uo-rinated surfactants and even phospholipids, which can takethe interface tension to values well below 10 mN m�1 in manycases.16d This means that the strong tendency of HFBs tosegregate and self-assemble into sturdy interfacial lms is likely

ns in 0.10 mg ml�1 HFBI aqueous solution, prepared either by a homogenizer oremulsion in 0.10 mg ml�1 HFBI aqueous solution prepared by a homogenizer.


Paper Soft Matter

to be key in emulsion stabilization, and represents both acommon characteristic of these proteins with uorosurfactants,and a distinctive asset over other natural surfactants. Phos-pholipid-stabilized emulsions with high stability and submi-cron droplet size have been reported;16d however, those weregenerally prepared by more complex procedures and equipment(e.g. high pressure/high shear homogenizers) than the onespresented here.

Fluorous droplet stabilization in microuidic circuits

The microuidic circuit schematized in Fig. 8a was used as analternative test platform for the stabilization of uorous oildroplets by HFBI. Several experiments were performed usingwater as a continuous phase and FC-70 as a dispersed phase. Weaimed particularly at optimizing the conditions for dropletgeneration, avoiding the formation of protein lms at theinternal walls of the device. Therefore, we gradually increasedthe pH of the aqueous continuous phase above the proteinisoelectric point (calculated 5.7, experimental 6.0 from zetapotential measurement, see ESI†). FC-70 droplet generation inwater was studied at three different pH values (7.0, 10.3 and13.0), both with and without the protein; it should be remarkedhere that HFBI lms have been previously reported to be stableat pH 13.46 The performance of HFBI was compared to that of anequivalent amount in weight of the commercial uo-rosurfactant Zonyl FSN-100. The results of these studies aresummarized in Fig. 8b–n and in S4 and S5 of ESI.†

Fig. 8b–d show FC-70 droplets generated in the absence ofany surfactant and collected in the chamber of the microuidic

Fig. 8 (a) Scheme of the microfluidic device. Details about the widths w1 and w2

serpentine and the width w4 and the length L4 of the coalescence chamber are repoaddition at (b) pH 7.0, (c) pH 10.3, (d) pH 13.0; for HFBI (0.10 mg ml�1) in H2O/FC-70FC-70 at (j) pH 7.0, (k) pH 10.3, (l–n) pH 13.0. The scale bar in the figures is 350 mm


circuit. At pH 7.0 and 10.3 the droplets generated were irregularin size and prone to sticking to the channel glass walls (Fig. S4ain the ESI†). At pH 13.0 the produced droplets were sphericalwith a constant radius and did not stick to the channel walls(Fig. S4b in the ESI†). However, in the absence of surfactant, FC-70 droplets eventually collapsed into macroscopic uorousphases both inside the microchannels and in the collectorchamber, regardless of the pH (Fig. 8b–d).

Fig. 8e–h show images of the collector chamber when FC-70droplets were formed in the presence of HFBI at a concentrationof 0.10 mg ml�1 in the aqueous stream at different pH values.Spherical droplets of constant diameter were produced at everystudied pH (Fig. S4c and d in the ESI†). At pH 7.0 and 10.3,droplet coalescence was observed throughout the circuit (seeFig. S5a and b in the ESI†) and inside the collector, which wasoen coated by a thin lm of protein (Fig. S5c in the ESI†). At pH13.0, droplets did not coalesce either in the serpentine circuit(see Fig. S5d–f in the ESI†) or in the collector chamber, and noprotein lm was formed on the circuit walls. Droplets in thecollector appeared spherical with an average diameter d1 of164 � 2 mm (Fig. 8h and i).

A comparison with the commercially available uo-rosurfactant Zonyl FSN-100 is shown in Fig. 8j–n. The dropletsgenerated at pH 7.0 and 10.3 did not coalesce in the circuit butwere highly irregular in size (Fig. S4e in the ESI†). Working atpH 13.0 yielded results comparable to those obtained withHFBI, as non-coalescing uorous droplets with reproduciblesizes were obtained (Fig. S4f in the ESI†). Interestingly, insidethe collector chamber a relevant portion of Zonyl-stabilizeddroplets had a less regular spherical shape than those produced

of the microchannels at the nozzle level, the width w3 and the length L3 of therted in the ESI.† Optical images of the collector for H2O/FC-70 without surfactantat (e) pH 7.0, (f) pH 10.3, (g–i) pH 13.0; for Zonyl FSN-100 (0.10 mg ml�1) in H2O/.

Soft Matter, 2013, 9, 6505–6514 | 6511

Soft Matter Paper

with HFBI (cf. Fig. 8i and n, with a1¼ 156� 3 mmand a2¼ 144�6 mm). The better shape regularity of the protein-stabilizeddroplets may be due to the superior mechanical properties ofthe HFBI lm, which is less prone to deformation under thestress applied by nearby droplets.

Relatively high pH conditions seem therefore important inorder to prevent theprotein fromsticking to the channelwalls, aswell as to produce non-coalescing droplets of reproducible sizefor Zonyl-stabilized systems. However, this feature may berelated to the optimization of the surface chemistry conditions atthe channel–water stream interface. In fact, at pH 13 the glasssurface is expected to be negatively charged (silica has two pKa

values of 4.9 and 8.5 according to Ong et al.).47 Also the protein atpH 13.0 is negatively charged, and electrostatic repulsionsuccessfully prevents the formation of a protein layer on thecircuit walls. In the case of Zonyl FSN-100, whose molecularformula is F(CF2CF2)x(CH2CH2O)yH, the pKa of the terminalhydroxyl group can be estimated to be close to that of ethox-yethanol (14.8), so the effect of pH is more difficult to explain.Silica deprotonation might play a role also here, as it wasdemonstrated that the interface tension between the channelwalls and the continuous liquid phase plays a crucial roletowards droplet stabilization in microuidic devices.48 On theother hand, ionic strength did not appear to have a signicantinuence, since using a pH 7 aqueous saline solution (NaCl 14.5mgml�1) as a continuous phase in the presence of Zonyl FSN-100yielded similar results to those observed in the absence of salt.

Even if some limitations were clear concerning the pH of thecontinuous aqueous phase, further chemical engineering of themicrochannel walls may signicantly broaden the usable pHrange, e.g. by functionalization with charged species or moietieshaving a pKa close to the isoelectric point of HFBI.

Conclusions

We have shown that the hydrophobin HFBI is an efficient andsustainable surfactant for the stabilization of uorous oil drop-lets in aqueous environments, effectively reducing the interfacialenergy and forming an exceptionally robust and elastic lmbetween the two liquid phases. Therefore HFBs can be regardedas effective stabilizers for aqueous–uorous liquid biphasicsystems, able to form stronger interfacial lms than othernatural surfactants, and performing at least comparably tocommon uorosurfactants but being exempt from the sustain-ability concerns that some of these compounds have raised.

In particular, uorous oil-in-water emulsions with over 1month stability were prepared by employing relatively simpleemulsication techniques and very low protein concentrations.These studies may pave the way to the effective use of HFBs inbiomedical applications as well as in emulsion polymerizationof uoropolymers, upon optimization of droplet size, e.g.through careful control of emulsication conditions. Further-more, we illustrated the use of HFBI for the stabilization ofuorous oil droplets in microuidic circuits employing water asthe continuous phase. These systems can potentially beemployed in analytical and synthetic uorous chemistry, as wellas in the preparation of e.g. uoropolymer particles or capsules

6512 | Soft Matter, 2013, 9, 6505–6514

with precise size control. Although the droplet stabilizationperformance of HFBI is comparable to that of more classicaluorosurfactants, at least four potential advantages can be seenfor using the protein: (i) HFBI is a natural, sustainable, and non-toxic compound; (ii) HFBI grants a better control over the shapeof the uorous droplets; (iii) HFBI is not soluble in the uorousphase, and therefore the environment within the uorousmicrodroplets is unaffected by the surfactant; (iv) the proteinlm offers barrier properties that may prevent dropletcontamination. Several applications in uorous chemistry ofthe microuidic circuits described herein are under currentinvestigation in our laboratories.

Acknowledgements

The authors gratefully thank Riitta Suihkonen (VTT) for tech-nical support. Funding from the Academy of Finland (project“Halosense”, decision 141558, and project “BioHal”, decision260565) and from the EU (project “ROC”, grant agreement n.213803) is also acknowledged.

Notes and references

1 (a) W. Zhang, Chem. Rev., 2004, 104, 2531–2556; (b)J. A. Gladysz, in Handbook of Green Chemistry, ed. P.Anastas and R. H. Crabtree, Wiley/VCH, Weinheim,Germany, 2004, vol. 1, pp. 17–38; (c) W. Zhang, Chem. Rev.,2009, 109, 749–795; (d) J.-M. Vincent, Top. Curr. Chem.,2012, 308, 153–174.

2 (a) B. E. Smart, Characteristics of C–F Systems, inOrganouorine Chemistry: Principles and CommercialApplications, ed. R. E. Banks, B. E. Smart and J. C. Tatlow,Plenum Press, New York, USA, 1994, pp. 57–88; (b)M. P. Kra, Applications of Fluorous Compounds inMaterials Chemistry, in Handbook of Fluorous Chemistry,ed. J. A. Gladysz, D. P. Curran and I. T. Horvath, Wiley-VCH, Weinheim, Germany, 2004, p. 478; (c) M. P. Kraand J. G. Riess, Chem. Rev., 2009, 109, 1714–1792.

3 C. J. Wilson, D. A. Wilson, A. E. Feiring and V. Percec,J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2498–2508,and references therein.

4 M. P. Kra, Acc. Chem. Res., 2012, 45, 514–524.5 (a) D. P. Curran, Science, 2008, 321, 1645–1646; (b)M. Cametti, B. Crousse, P. Metrangolo, R. Milani andG. Resnati, Chem. Soc. Rev., 2012, 41, 31–42.

6 (a) E. Kissa, in Fluorinated Surfactants and Repellents, MarcelDekker, Inc., New York (USA), 2nd edn, 2001; (b) C. K. Taylor,Fluorinated surfactants in practice, in Design and selection ofperformance surfactants: Annual surfactants review, ed. D.Karsa, John Wiley & Sons, New York (USA), 1999, pp. 271–316; (c) K. Skrabania, A. Laschewsky, H. von Berlepsch andC. Bottcher, Langmuir, 1999, 25, 7594–7606; (d) J. G. Riess,Curr. Opin. Colloid Interface Sci., 2001, 14, 294–304.

7 (a) H. J. Lehmler, Chemosphere, 2005, 58, 1471–1496; (b)H. Hori, E. Hayakawa, H. Einaga, S. Kutsuna, K. Koike,T. Ibusuki, H. Kiatagawa and R. Arakawa, Environ. Sci.Technol., 2004, 38, 6118–6124.


Paper Soft Matter

8 (a) S. Bull, K. Foxall, PFOS + PFOAToxicological Overview, Reportfrom the Health Protection Agency, 2009; (b) M. E. Andersen,J. L. Butenhoff, S.-C. Chang, D. G. Farrar, G. L. Kennedy Jr,C. Lau, G. W. Olsen, J. Seed and B. Wallace, Toxicol. Sci., 2008,102, 3–14; (c) J. Rubarth, A. Dreyer, N. Guse, J. W. Einax andR. Ebinghaus, Environ. Chem., 2011, 8, 419–428.

9 (a) OECD report on “Hazard Assessment of PFOS and Its Salts”,http://www.oecd.org/env/chemicalsafetyandbiosafety/riskmanagementonstallationsandchemicals/peruorooctanesulfonatepfosandrelatedchemicalproducts.htm; (b) US EPA2010/2015 PFOA Stewardship Program website, http://www.epa.gov/opptintr/pfoa/pubs/stewardship/index.html; (c) USEPA Peruorooctanoic Acid (PFOA) and Fluorinated Telo-mers -Related Programs and Studies, http://www.epa.gov/oppt/pfoa/pubs/related.html.

10 (a) US EPA Industry Progressing in Voluntary Effort to ReduceToxicChemicals/Companies have successfully developedover150 alternatives to long-chain peruorinated chemicals,http://yosemite.epa.gov/opa/admpress.nsf/6427a6b7538955c585257359003f0230/335d29e8d5e40864852579a0005f3760!OpenDocument; (b) A. Zaggia and B. Ameduri, Curr. Opin.Colloid Interface Sci., 2012, 17, 188–195.

11 S. C. Gordon, Regul. Toxicol. Pharmacol., 2011, 59, 64–80.12 G. Marchionni and V. Tortelli, PCT WO/2010/003931.13 S. Peng and M.-H. Hung, J. Fluorine Chem., 2012, 133, 77–85.14 M. Peschka, N. Fichtner, W. Hierse, P. Kirsch,

E. Montenegro, M. Seidel, R. D. Wilken and T. P. Knepper,Chemosphere, 2008, 72, 1534–1540.

15 C. M. Kausch, Y. Kim, V. M. Russell, R. E. Medsker andR. R. Thomas, Langmuir, 2003, 19, 7182–7187.

16 (a) C. I. Castro and J. C. Briceno, Artif. Organs, 2010, 34, 622–634; (b) M. P. Kra, A. Chittofrati and J. G. Riess, Curr. Opin.Colloid Interface Sci., 2003, 8, 251–258; (c) J. G. Riess,J. Fluorine Chem., 2002, 114, 119–126; (d) J. G. Riess, Chem.Rev., 2001, 101, 2797–2919.

17 (a) D. K. Sarker, Curr. Drug Delivery, 2005, 2, 297–310; (b)M. P. Kra, Adv. Drug Delivery Rev., 2001, 47, 209–228.

18 (a) R. Dıaz-Lopez, N. Tsapis and E. Fattal, Pharm. Res., 2010,27, 1–16; (b) J. M. Janjic and E. T. Ahrens, WIRES NanomedNanobi., 2009, 1, 492–501, and references therein.

19 A. Misra and M. W. Urban, Macromolecules, 2009, 42, 7828–7835.

20 P. N. Nguyen, T. T. Trinh Dang, G. Waton, T. Vandamme andM. P. Kra, ChemPhysChem, 2011, 12, 2646–2652.

21 M. B. Linder, Curr. Opin. Colloid Interface Sci., 2009, 14, 356–363.

22 A. Walther and A. H. E. Muller, SoMatter, 2008, 4, 663–668.23 A. H. Y. Kwan, R. D. Wineeld, M. Sunde, J. M. Matthews,

R. G. Haverkamp, M. D. Templeton and J. P. Mackay, Proc.Natl. Acad. Sci. U. S. A., 2006, 103, 3621–3626.

24 M. B. Linder, G. R. Szilvay, T. Nakari-Setala andM. E. Penttila, FEMS Microbiol. Rev., 2005, 29, 877–896.

25 (a) S. Askolin, M. Linder, K. Scholtmeijer, M. Tenkanen,M. Penttila, M. L. de Vocht and H. A. B. Wosten,Biomacromolecules, 2006, 7, 1295–1301; (b) T. Subkowski,M. Karos and T. Subkowski, J. Biotechnol., 2007, 131, S212–S213.


26 A. R. Cox, D. L. Aldred and A. B. Russell, Food Hydrocolloids,2009, 23, 366–376.

27 K. Scholtmeijer, J. G. Wessels and H. A. Wosten, Appl.Microbiol. Biotechnol., 2001, 56, 1–8.

28 (a) K. Scholtmeijer, M. I. Janssen, B. Gerssen, M. L. de Vocht,B. M. van Leeuwen, T. G. van Kooten, H. A. Wosten andJ. G. Wessels, Appl. Environ. Microbiol., 2002, 68, 1367–1373; (b) M. Linder, G. R. Szilvay, T. Nakari-Setala,H. Soderlund and M. Penttila, Protein Sci., 2002, 11, 2257–2266.

29 (a) S. O. Lumsdon, J. Green and B. Stieglitz, Colloids Surf.,B, 2005, 44, 172–178; (b) W. Wohlleben, T. Subkowski,C. Bollschweiler, B. von Vacano, Y. Liu, W. Schrepp andU. Baus, Eur. Biophys. J. Biophys. Lett., 2010, 39, 457–468.

30 R. Misra, J. Li, G. C. Cannon and S. E. Morgan,Biomacromolecules, 2006, 7, 1463–1470.

31 (a) K. Scholtmeijer, M. I. Janssen, M. B. M. van Leeuwen,T. G. van Kooten, H. Hektor and H. A. B. Wosten, Bio-Med.Mater. Eng., 2004, 14, 447–454; (b) X. Wang,H. P. Permentier, R. Rink, J. A. W. Kruijtzer,R. M. J. Liskamp, H. A. B. Wosten, B. Poolman andG. T. Robillard, Biophys. J., 2004, 87, 1919–1928; (c)M. L. de Vocht, K. Scholtmeijer, E. W. van der Vegte,O. M. H. de Vries, N. Sonveaux, H. A. B. Wosten,J.-M. Ruysschaert, G. Hadziioannou, J. G. H. Wessels andG. T. Robillard, Biophys. J., 1998, 74, 2059–2068.

32 A. Schulz, B. M. Liebeck, D. John, A. Heiss, T. Subkowski andA. Boker, J. Mater. Chem., 2011, 21, 9731–9736.

33 (a) R. Wang, Y.-L. Yang, M. Qin, L.-K. Wang, L. Yu, B. Shao,M.-Q. Qiao, C. Wang and X.-Z. Feng, Chem. Mater., 2007, 19,3227–3231; (b) L. M. Bimbo, M. Sarparanta, E. Makila,T. Laaksonen, P. Laaksonen, J. Salonen, M. B. Linder,J. Hirvonen, A. J. Airaksinen and H. A. Santos, Nanoscale,2012, 4, 3184–3192.

34 M. Linder, K. Selber, T. Nakari-Setala, M. Qiao, M. Kula andM. Penttila, Biomacromolecules, 2001, 2, 511–517.

35 S. L. Anna, N. Bontoux and H. A. Stone, Appl. Phys. Lett.,2003, 82, 364–366.

36 R. Dreyfus, P. Tabeling and H. Willaime, Phys. Rev. Lett.,2003, 90, 144505.

37 G. R. Szilvay, A. Paananen, K. Laurikainen, E. Vuorimaa,H. Lemmetyinen, J. Peltonen and M. B. Linder,Biochemistry, 2007, 46, 2345–2354.

38 M. Reger, T. Sekine, T. Okamoto and H. Hoffmann, SoMatter, 2011, 7, 8248–8257.

39 A. R. Cox, F. Cagnol, A. B. Russell and M. J. Izzard, Langmuir,2007, 23, 7995–8002.

40 (a) S. A. Roberts, I. W. Kellaway, K. M. G. Taylor,B. Warburton and K. Peters, J. Pharmaceut., 2005, 300, 48–55; (b) J. M. Rodriguez Patino, A. L. Caro, M. R. RodriguezNino, A. R. Mackie, A. P. Gunning and V. J. Morris, FoodChem., 2007, 102, 532–541.

41 S. G. Baldursdottir, M. S. Fullerton, S. Hougaard Nielsen andL. Jorgensen, Colloids Surf., B, 2010, 79, 41–46.

42 J.-M. Jung, D. Z. Gunes and R. Mezzenga, Langmuir, 2010, 26,15366–15375.

Soft Matter, 2013, 9, 6505–6514 | 6513

Soft Matter Paper

43 A. R. Mackie, A. P. Gunning, M. J. Ridout, P. J. Wilde andV. J. Morris, Langmuir, 2001, 17, 6593–6598.

44 L. Wang, H. Xie, X. Qiao, A. Goffin, T. Hodgkinson, X. Yuan,K. Sun and G. G. Fuller, Langmuir, 2012, 28, 459–467.

45 E. Dickinson, Curr. Opin. Colloid Interface Sci., 1998, 3,633–638.

6514 | Soft Matter, 2013, 9, 6505–6514

46 Z.-X. Zhao, H.-C. Wang, X. Qin, X.-S. Wang, M.-Q. Qiao,J. Anzai and Q. Chen, Colloids Surf., B, 2009, 71, 102–106.

47 S. W. Ong, X. L. Zhao and K. B. Eisenthal, Chem. Phys. Lett.,1992, 191, 327–335.

48 L. L. Shui, A. van den Berg and J. C. T. Eijkel, Lab Chip, 2009,9, 795.


Hydrophobin: fluorosurfactant-like properties without fluorine

Documents