The use of chemical actinometry for the evaluation of the light absorption efficiency in scattering photopolymerizable miniemulsions

Photochemical &Photobiological Sciences

PAPER

Cite this: DOI: 10.1039/c4pp00323c

Received 22nd August 2014,Accepted 9th October 2014

DOI: 10.1039/c4pp00323c

www.rsc.org/pps

The use of chemical actinometry for theevaluation of the light absorption efficiency inscattering photopolymerizable miniemulsions†

Marta Penconi,a Emeline Lobry,b Florent Jasinski,b Abraham Chemtob,b

Céline Croutxé-Barghorn,b Adrien Criqui,c André M. Braund and Esther Oliveros*a

Oil-in-water miniemulsions containing a mixture of monomers as the dispersed organic phase have been

shown recently to be promising media for the development of photoinitiated polymerization processes.

Albeit a crucial factor for a successful application, the efficiency of light absorption by the photoinitiator

in these highly scattering systems is difficult to evaluate. In this work, a well-characterized water insoluble

chemical actinometer (DFIS) replaced the oil-soluble photoinitiator, and was used as a probe and a model

for UV light absorption in miniemulsions of variable droplet sizes and organic phase compositions (i.e. at

different levels of scattered light). In the first step, the photon flux absorbed by the actinometer was deter-

mined in model miniemulsions based on an inert solvent (ethyl acetate), at a low oil phase content

(3.0–6.0 wt%). For these low to moderately scattering systems, the photon flux absorbed by the actino-

meter in the miniemulsions was comparable to that in a homogeneous solution of ethyl acetate. In the

second step, the absorbed photon flux was investigated in photopolymerizable miniemulsions (a mixture

of acrylate monomers as oil phase). Surprisingly, in spite of much higher scattering coefficients than

those found for ethyl acetate based miniemulsions of otherwise the same composition, the photon flux

absorbed by the actinometer in photopolymerizable miniemulsions showed only a small decreasing

trend. Such a result may be considered favorable for the further development of applications of photo-

polymerizations in miniemulsions.

1. Introduction

In recent years, increasing attention has been devoted to hetero-geneous free radical polymerization for the production ofaqueous dispersions of polymer nanoparticles.1 Such latexesare important raw materials used in a variety of industrial pro-cesses, to protect or finish metals and wood, or they are usedas binders for pigments or fillers. They are prepared bypolymerization of monomers emulsified in water. Emulsionpolymerization processes are generally initiated thermally

through the decomposition of a radical initiator (peroxide, azocompounds) triggered by a temperature increase.

The photoinitiated polymerization is a viable and emergingalternative to the thermal initiation and offers several advan-tages, such as lower reaction temperatures, higher polymeriz-ation rates due to a faster initiation and a better controlon radical generation by tuning the characteristics of thelight source and the irradiation time. Although photo-polymerization is essentially exploited as UV-curing tech-nology to produce cross-linked films, numerous investigationsin the field of linear polymer preparation reported photo-polymerizations of monomer emulsions,2–6 microemulsions7–12

and miniemulsions.13–19

Recently, acrylate miniemulsions, with droplet sizesbetween 50 and 500 nm, have been explored as highly suitablemedia for the preparation of polymer latexes using UV light.To obtain a miniemulsion, a surfactant and a costabilizer areemployed in order to stabilize the monomer nanodroplets inthe aqueous continuous phase. In particular, the costabilizer,i.e. a water-insoluble low-molecular-weight compound, isneeded in order to prevent Ostwald ripening and guaranteethe metastability of the system.20–22 Moreover, a high shear

†Electronic supplementary information (ESI) available. See DOI: 10.1039/c4pp00323c

aLaboratoire des Interactions Moléculaires et Réactivité Chimique et Photochimique

(IMRCP), UMR CNRS/UPS 5623, Université Toulouse III (Paul Sabatier), 118, route

de Narbonne, F-31062 Toulouse cédex 9, France.

E-mail: [email protected], [email protected] of Photochemistry and Macromolecular Engineering, ENSCMu,

University of Haute Alsace, 3 rue Alfred Werner 68093, Mulhouse Cedex, FrancecMäder Research MADER GROUP, 130 rue de la Mer Rouge, 68200 Mulhouse, FrancedEngler-Bunte-Institute, Karlsruhe Institute of Technology (KIT), 76131 Karlsruhe,

Germany

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article OnlineView Journal

force, provided for instance by ultrasounds, is required to formmonomer nanodroplets. In principle, the high number ofnanodroplets generated makes them the main locus of nuclea-tion; thus polymer formation occurs in principle inside thedroplets.21,23 The size and composition of the droplets play anessential role in determining the optical properties of theminiemulsion. Despite their nanometric size range, miniemul-sions are not transparent due to their polydispersity, and thelight may be both absorbed and scattered by the medium.

Conventional spectroscopic techniques can only providethe attenuation (or extinction) coefficient, that is the sum ofabsorption and scattering coefficients. Recently, the opticalproperties of diluted miniemulsions of acrylate monomershave been investigated by Jasinski et al.24 using different spec-trophotometric methods based on an approach alreadyapplied by Sun and Bolton.25 Results showed that when thedroplet size decreased, the absorption coefficient remainedconstant and close to that measured in a solution of mono-mers. In contrast, the scattering coefficient strongly decreased,leading to a faster photopolymerization reaction due to thebetter light penetration into the heterogeneous medium. Inthe case of highly concentrated miniemulsions, multiple scat-tering cannot be avoided and the determination of the absorp-tion and single scattering coefficients is a more complexissue.26 The depth of light penetration within the miniemul-sion is of primary importance for the photopolymerizationrate, as the generation of primary radicals occurs uponirradiation of a photoinitiator system. The most commonradical photoinitiators have a benzoyl-based structure andundergo homolytic α-cleavage (Norrish I cleavage) from theexcited triplet state.27–29 In order to optimize the photochemi-cal initiation step, the photon flux absorbed by the photo-initiator should therefore be maximized.30

The issue in determining the efficiency of photon absorp-tion in heterogeneous systems has been largely investigated inthe field of photocatalysis in TiO2 suspensions aiming at thecalculation of quantum efficiencies.25,31–35 The main problemis related to the estimation of the radiation distribution fieldin these systems, due to the simultaneous occurrence of lightscattering and absorption by the photocatalyst particles.However, the determination of the absorption and scatteringcoefficients is not sufficient to evaluate the absorbed photonflux, since multiple scattered photons may be subsequentlyabsorbed giving an additional contribution. Similar to thecase of TiO2 suspensions, both absorption and scattering areresponsible for light attenuation in monomer miniemulsions.However, in contrast to TiO2 photocatalysis, the photochemi-cally active species (photoinitiator) is consumed duringpolymerization and evaluation of the absorbed photon flux insuch a system is extremely complex.

In order to improve the understanding of the photonabsorption process within the core of the monomer droplets,we replaced the oil-soluble photoinitiator by a chemical actino-meter that can be easily dissolved into the organic dispersedphase. A chemical actinometer undergoes a well-characterizedphotochemical reaction and allows the estimation of the inci-

dent and absorbed photon flux.36 The main advantage withrespect to physical actinometers (radiometers) lies in the exactreproducibility of experimental conditions using the sameexperimental arrangement, i.e. the same reaction vessel (a cell,a photochemical reactor), equipment (lenses, filters) andsolvent or solvating system, for sample and reference chemicalactinometry. To our knowledge, this study reports the firstexample of actinometry carried out by dissolving the chosenchemical actinometer ((E)-[1-(2,5-dimethyl-3-furyl)ethylidene]-(isopropylidene)-succinic anhydride, DFIS) in the organic partof a dispersed system. The absorbed photon flux has beenevaluated on an optical bench under monochromatic exci-tation. Actinometry has been performed in miniemulsionswith variable droplet sizes and organic phase contents, i.e. atdifferent levels of scattered light, to better understand theeffect of scattering on the efficiency of photon absorption. Inthe first part, actinometric experiments have been performedin a model type miniemulsion, where the organic phase con-sisted of the actinometer dissolved in an inert solvent (ethylacetate). This system allowed working under low scatteringconditions and represented a first and simple approach. In thesecond part, studies have been carried out in highly scatteringphotopolymerizable miniemulsions where the oil phase con-sisted of the actinometer dissolved in a mixture of acrylatemonomers.

2 Materials and methods2.1 Chemicals

(E)-[1-(2,5-Dimethyl-3-furyl)ethylidene](isopropylidene)-succinicanhydride (DFIS, >95%) was supplied by Extrasynthese,France, and used without further purification. Note that DFISis commonly known as Aberchrome 540, the commercialname of the compound produced by Aberchromics Ltd, butsince 2000 Aberchrome 540 is no longer available from thiscompany. Ethyl acetate (EtAc, >99.7%) and technical grademonomers, methyl methacrylate (MMA, 99%), butyl acrylate(BA, >99%) and acrylic acid (AA, 99%), were obtained fromSigma-Aldrich and used as received. Sodium dodecyl sulfate(SDS, >99.0%, Aldrich) and octadecyl acrylate (OA, 97%,Aldrich) were used as the surfactant and costabilizer, respecti-vely. All miniemulsions were prepared with ultrapure water.

2.2 Miniemulsion preparation

Miniemulsions containing a monomer mixture (MM) as theorganic phase (φorg) were prepared as follows. The organicphase was obtained by mixing the three acrylate monomersaccording to the ratio MMA/BA/AA (49.5/49.5/1 wt%). The cost-abilizer OA (4 wt/wtφorg

%) was added to the mixture of mono-mers. The aqueous phase consisted of an aqueous SDSsolution. The concentration of SDS (0.08–1.5 wt/wtφorg

%) wasvaried to obtain droplet sizes ranging from 47 nm to 150 nm.The organic and aqueous phases were mixed using a magneticstirrer at 700 rpm for at least 10 min to form a coarse emul-sion. The high homogenization energy input required to form

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

a miniemulsion was provided by sonication using a BransonSonifier 450 (450 W L−1) for 5 min under magnetic stirring.Miniemulsions, where the organic phase consisted of EtAc,were prepared according to the same procedure. All mini-emulsions were prepared with an organic phase content of30 wtφorg

% (weight of MM or EtAc with respect to the totalweight). When lower organic phase content is mentioned,miniemulsions were obtained by subsequent dilution of the30 wtφorg

% concentrated miniemulsion. For the actinometricexperiments in miniemulsions, the organic phase was pre-pared by dissolving the actinometer DFIS in MM or EtAc.

2.3 Miniemulsion characterization

The droplet size (z-average diameter) was determined bydynamic light scattering (DLS) using a Zetasizer Nano ZS,Malvern Instruments. In the prepared miniemulsions, thehigh concentration of droplets gives rise to a multiple scatter-ing effect of the incident light that affects the reliability of theDLS measurement. To avoid this problem, samples were pre-pared by diluting miniemulsions 125 times with ultrapurewater. Further dilution did not affect the size of the droplet,indicating that the observed values of the diameters are practi-cally independent on the droplet concentration. Measure-ments were conducted within 10 minutes of samplepreparation, although no significant variation of the dropletsize was observed over 2 hours. UV/Visible spectrophotometricmeasurements were carried out using a HP 8453 diode arraysingle beam spectrophotometer equipped with a HP 89090APeltier temperature controller. The spectra were recorded at22 °C in quartz cells of 1 cm, 0.5 cm or 0.01 cm optical pathlengths, using water, EtAc or the monomer mixture MM as areference.

2.4 Optical properties of heterogeneous systems

In an homogeneous isotropic medium containing an absorb-ing species (Ac), incident photons can be either transmitted orabsorbed, and the sum of the photon fluxes absorbed (Pa,λ)and transmitted (Pt,λ) is equal to the incident photon flux (P0,λ)at wavelength λ. Photon fluxes are usually expressed in einsteins−1 or in einstein L−1 s−1.

P0;λ ¼ Pa;λ þ Pt;λ ð1ÞSpectrophotometric analysis carried out by determining the

transmitted photon flux (Pt,λ) provides an estimation of thevalue of the transmittance (Tλ = Pt,λ/P0,λ), and the absorbance(Aλ) may be calculated according to the Beer–Lambert law,37

Aλ ¼ log1Tλ

¼ K λ‘ ¼ ½Ac�ελ‘ ð2Þ

with ℓ: optical path length [cm], Kλ: absorption coefficient[cm−1], [Ac]: the concentration [mol L−1] of the absorbingspecies Ac and ελ: the molar absorption coefficient of Ac[L mol−1 cm−1]. Note that if spectroscopic cells of 1 cm opticalpathlength are used, Aλ = Kλ.

The absorbed photon flux is thus related to the incidentphoton flux using eqn (3).

Pa;λ ¼ P0;λð1� 10�Aλ Þ ¼ P0;λð1� 10�K λ‘Þ ð3ÞIn heterogeneous systems, such as miniemulsions, light

attenuation (extinction) arises not only from absorption ofphotons but also from scattering effects. In this case,

P0;λ ¼ Pa;λ þ Ps;λ þ Pt;λ ð4Þwhere Ps,λ is the scattered photon flux.

Under these conditions, the Beer–Lambert law isexpressed as

Dλ ¼ log1Tλ

¼ Eλ‘ ¼ Ndσλ‘ ¼ K λ‘þ Sλ‘ ð5Þ

where Dλ is the extinction (or attenuance),37 Eλ and Sλ are theextinction and scattering coefficients [cm−1], respectively;Nd [m−3] is the number density of the droplets and σλ [m

2] isthe overall cross-section.

The sum of Pa,λ and Ps,λ may be expressed as a functionof P0,λ using eqn (6), similar to eqn (3), where the absorbanceAλ has been substituted by the extinction (attenuance) Dλ.

Pa;λ þ Ps;λ ¼ P0;λð1� 10�DλÞ ð6ÞUnder these conditions, UV/Vis transmission spectro-

photometry yields the value of Eλ (= Sλ + Kλ, eqn (5)) andseparate determination of Kλ and Sλ is not possible with thistechnique. Spectroscopic methods using an integrating sphereto collect reflected and/or transmitted light have been pro-posed for studying the optical properties of heterogeneoussystems and for determining absorption and scattering co-efficients in these media.24–26,38

In this work, we have chosen a different approach for theevaluation of the absorption efficiency and the absorbedphoton flux has been determined by chemical actinometry.36

The spectra herein presented were recorded by collectinglight transmitted both in homogeneous and heterogeneousmedia.

2.5 Chemical actinometry

2.5.1 Choice of the chemical actinometer. (E)-1-(2,5-Dimethyl-3-furyl)ethylidene(isopropylidene)-succinic anhy-dride (DFIS) is a photochromic fulgide compound. It exhibitsP-type photochromism, i.e. the light-induced reaction betweenthe two isomers having different absorption spectra is photo-chemically but not thermally reversible. Irradiation of DFIS inthe UV spectral region (310–370 nm) induces an electrocyclicreaction leading to ring-closure and formation of the 5,6-dicarb-oxylic acid anhydride of 7,7a-dihydro-2,4,7,7a-pentamethyl-benzo[b]furan (7,7a-DHBF), a compound that absorbs in thevisible spectral region. The back reaction to the open-formspecies (DFIS) is promoted by visible radiation (Scheme 1).

DFIS has been widely used as an actinometer in the UV(310–370 nm) and visible (436–546 nm) ranges since the1980s.36,39–43 It has to be pointed out that photoisomerization

Photochemical & Photobiological Sciences Paper

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

of DFIS (E-isomer) to the Z-form competes with photocycliza-tion to 7,7a-DHBF. Therefore, the reuse of the actinometer inthe UV region after reconversion of 7,7a-DHBF to DFIS is notrecommended due to the possible underestimation of the inci-dent photons.44,45

DFIS has been chosen as an actinometer (Ac) in this studyas it meets some important requirements. It absorbs in theUV-A region, matching the spectra of some of the mostcommon photoinitiators used for polymerization; its highsolubility in organic solvents allows solubilisation exclusivelyin the monomer droplets. Moreover, the photochemical reac-tion can be easily followed by spectrophotometry in the visiblespectral region where absorption is only due to the photo-product (7,7a-DHBF). Finally, since DFIS is a well-known actino-meter, the photochemical reaction and the actinometricprocedure have been extensively discussed in the literature.

DFIS (Ac) has been used in this study to estimate theabsorbed photon flux (PAc,λ) in miniemulsions under differentexperimental conditions. By definition, PAc,λ is the number ofmoles of photons (Np,Ac,λ, [einstein]) absorbed per unit of timeby Ac at the wavelength of irradiation λ.

PAc;λ ¼ dNp;Ac;λ

dtð7Þ

The value of PAc,λ can be calculated from the rate of thephotochemical reaction of Ac (dnAc/dt, the number of moles ofAc transformed or of the product formed per unit of time)according to the definition of the quantum yield of a photo-chemical reaction:

ΦAc;λ ¼ dnAc=dtdNp;Ac;λ =dt

ð8Þ

In homogeneous systems, the experimental determinationof dnAc/dt was carried out by spectrophotometry following thevariation of the absorbance (or of the absorption coefficient) of7,7a-DHBF at the wavelength of its absorption maximum inthe visible λ′ (Aλ′ or Kλ′) during irradiation. Combining eqn (2),(7) and (8), PAc,λ can be written as

PAc;λ ¼ dAλ′dt

� VΦAc;λ � ελ′ � ‘ ¼

dK λ′

dt� VΦAc;λ � ελ′ ð9Þ

with V: irradiated volume ([L]); note that λ is the wavelength ofirradiation (UV spectral region) and λ′ is the wavelength of ana-lysis (visible spectral region).

In actinometric experiments in miniemulsions, evenwithout 7,7a-DHBF, the initial value of the extinction coeffi-cient at λ′ (E0,λ′) may be greater than zero due to non-negligiblescattering by the miniemulsion (E0,λ′ = SME,λ′). In this case, thevalues of the absorption coefficient Kλ′ versus irradiation timewere calculated by subtracting E0,λ′ from the values of theextinction coefficient Eλ′ obtained by transmittance measure-ments (eqn (5)). Therefore, in the case of scattering miniemul-sions, the photon flux absorbed by the actinometer PAc,ME,λ atthe irradiation wavelength was calculated as

PAc;ME;λ ¼ dK λ′

dt� VΦAc;λ � ελ′ ¼

dðEλ′ � E0;λ′Þdt

� VΦAc;λ � ελ′ ð10Þ

It should be noted that in homogeneous solution where noscattering occurs, the absorbed photon flux calculated by acti-nometry, i.e. from the number of 7,7a-DHBF molecules formed(eqn (9)), is equal to the absorbed photon flux as defined byeqn (3), i.e. PAc,λ = Pa,λ, since Pa,λ is directly related to the valueof the absorbance at λ. This is not the case in scatteringsystems. In fact, in miniemulsions, scattering at λ (UV spectralregion) is much larger than at λ′ (visible spectral region). Evenif Kλ could be precisely calculated from the measured value ofEλ after subtraction of the scattering by the miniemulsion(SME,λ), the value of Pa,ME,λ, calculated using eqn (3), would notaccount reliably for multiple scattering by the miniemulsiondroplets and potential reabsorption by DFIS, especially in rela-tively concentrated miniemulsions. Therefore, PAc,ME,λ calcu-lated from the number of 7,7a-DHBF molecules formed shouldbe a much better approximation of the effective photon fluxabsorbed by DFIS at λ.

2.5.2 Experimental procedure for actinometric measure-ments. The experimental equipment used to perform actino-metry consisted of a xenon–Hg lamp HBO1000 (MüllerElektronik Optik, Germany) equipped with a water filter to cutoff the IR radiation and a monochromator (JobinYvon SPEXISA Instruments, 6 nm band width) to select the irradiationwavelength. 367 nm was chosen for the actinometric experi-ments, since DFIS has a relatively high molar absorptioncoefficient at this wavelength and self-initiated polymerizationby the acrylate monomers, known to occur below 300 nm,19

could be avoided. A gray filter having 25% transmittance wasused to reduce the radiant power of the lamp. A thermopile(Laser Instrumentation, model 154) placed after the spectro-scopic cell was used for monitoring the incident radiantpower. A value of approximately 0.90 mW was measured at367 nm. All the measurements were performed in the darkand at controlled temperature.

In a typical actinometric experiment, a known volume ofthe sample (DFIS dissolved in a homogeneous system or in theminiemulsion droplets) was placed in a quartz cell of 1 cmoptical pathlength and irradiated at 367 nm under stirring fora given period of time. UV/Visible absorption or extinction

Scheme 1 Structure and reversible photochromic reaction of (E)-1-(2,5-dimethyl-3-furyl)ethylidene(isopropylidene)-succinic anhydride(DFIS) to 5,6-dicarboxylic acid anhydride of 7,7a-dihydro-2,4,7,7a-pentamethylbenzo[b]furan (7,7a-DHBF).

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

spectra were recorded after different irradiation periods andthe absorbance (Aλ′), or the extinction (Dλ′) in the case of scat-tering media, of the coloured solution was followed at thewavelength of the maximum λ′ of the 7,7a-DHBF absorptionband in the visible region.

For actinometric measurements in heterogeneous systems,miniemulsions were prepared with MM or EtAc as the organicphase, at different SDS and organic phase contents (section2.2). Although high organic phase contents are desirable forphotopolymerization processes, in this first approach to theissue, a diluted miniemulsion represents a simpler case studywhere the lower concentration of droplets leads to a moretransparent medium and allows the use of spectrophotometryto follow the photoreaction. Therefore, each miniemulsion wasdiluted with water by factors of 5, 7 and 10, thus the resultingorganic phase contents were 6.0%, 4.3% and 3.0%, respecti-vely. Upon dilution, the amounts of surfactant and costabilizerwith respect to the organic phase were kept constant. Sinceminiemulsions were diluted after sonication, it can beassumed that the size of the droplets remained unchangedand the only effect of the dilution was to lower the concen-tration of droplets in the miniemulsion.24

For each type of miniemulsion, the photon flux absorbedby the actinometer (PAc,ME,λ) was determined and comparedwith the value obtained in the analogous homogeneoussystem (MM or EtAc, PAc,λ) having the same DFIS concen-tration. After repeated measurements (at least three) on thesame system, the experimental error on the absorbed photonflux was evaluated to be approx. 10% and 15% in the case ofthe EtAc and MM based miniemulsions, respectively.

3 Results and discussion3.1 Physical characteristics and optical properties of EtAcbased miniemulsions

In the first approach, the efficiency of light absorption wasstudied for a model-type miniemulsion, where the organicphase consisted of ethyl acetate (EtAc). This latter inert solventwas chosen for actinometric measurements because (i) thephotophysical properties and the photochemical reaction ofthe chosen chemical actinometer DFIS (Scheme 1, section2.5.1) were investigated in detail in this solvent;45 (ii) mini-emulsions based on EtAc could be easily prepared and turnedout to be stable over several hours; (iii) these model miniemul-sions have physical characteristics close to those of the mini-emulsions based on the mixture of acrylate monomers (MM)under the same conditions of preparation. Although theoptical properties of the two systems are different (section 3.3),the strongly reduced light scattering of the model mini-emulsions made them particularly well-suited for our study.

EtAc based miniemulsions were first characterized in theabsence of DFIS. Four different miniemulsions were prepared,having 30 wtφorg

% and 0.08, 0.15, 0.5 and 1.5 wt/wtφorg% SDS

content leading to droplet average diameters of 150 nm,100 nm, 73 nm and 47 nm, respectively. As expected, droplet

sizes could be tuned by varying the ratio SDS/φorg.22 The as-

prepared miniemulsions (30 wtφorg%) were highly scattering.

For all samples the values of the extinction coefficient at themaximum of the band of the closed form of the actinometer(7,7a-DHBF, Scheme 1) in the visible region were higher than2. Spectrophotometric analysis of the actinometric reactionunder these conditions would require an optical pathlengthseveral orders of magnitude shorter than 1 cm and thereforean extremely high concentration of DFIS. To overcome thisdrawback, miniemulsions were diluted as mentioned insection 2.5.2.

The composition, droplet size and extinction coefficient at367 nm of the EtAc based miniemulsions investigated in thiswork are reported in Table 1. Each sample has been namedwith the use of a letter indicating the increasing size: A for thesmallest droplet size (47 nm), B and C for droplet sizes of73 nm and 100 nm, respectively, and D for the largestdroplet size (150 nm), and a number (6, 4 or 3) indicating thepercent amount of the organic phase (6.0, 4.3 or 3.0 wtφorg

%).In the absence of DFIS, there was no species absorbingat 367 nm and the value of the extinction coefficient (in therange between 0.10 and 3.06) represents the scatteringcoefficient.

Light transmission over the entire UV and visible range wasstrongly dependent on the size of the droplets and also on theorganic phase amount. In Fig. 1A, the effect of droplet size onthe extinction spectra is highlighted for miniemulsions having6.0 wtφorg

%. By increasing the size of the droplet, miniemul-sions became less and less transparent. Since aqueous solu-tions of SDS do not absorb in the range between 200 and800 nm and EtAc absorbs only below 260 nm, the increasingextinction above 260 nm can only be attributed to the higheramount of photons scattered by bigger droplets with respect tothe smaller ones.

In Fig. 1B, the effect of the amount of organic phase on theextinction spectra is shown for miniemulsions of 100 nmdroplet size (C3, C4 and C6). According to eqn (5) and exclud-ing the contribution due to absorption, the extinction coeffi-

Table 1 Characteristics of EtAc based miniemulsions: composition inweight percent (SDS and organic phase contents), droplet average dia-meter (d ) and extinction coefficient at 367 nm (E367, measured inspectroscopic cells of ℓ = 1 cm or 0.5 cm)a

% SDS/φorg d (nm) Miniemulsion % φorg (EtAc) E367 (cm−1)

1.5 47 A3 3.0 0.10A4 4.3 0.15A6 6.0 0.22

0.5 73 B3 3.0 0.35B4 4.3 0.48B6 6.0 0.70

0.15 100 C3 3.0 0.98C4 4.3 1.40C6 6.0 1.92

0.08 150 D3 3.0 1.50D4 4.3 2.18D6 6.0 3.06

a Experimental errors ≈15% calculated from repeated measurements.

Photochemical & Photobiological Sciences Paper

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

cient was proportional to the number density of droplets Nd,thus, upon dilution of the miniemulsion, a decrease of Eλ wasobserved in the spectra.

3.2 Determination of the photon flux absorbed by a chemicalactinometer in EtAc based miniemulsions

3.2.1 Reference actinometric experiments in EtAc: thehomogeneous system. Reference experiments were performedusing the chosen chemical actinometer (DFIS, section 2.5.1) ina homogeneous solution of EtAc. The wavelength of themaximum absorption of DFIS in EtAc is 340 nm (ε340 = 6270 Lmol−1 cm−1), while the closed form (7,7a-DHBF) has anabsorption maximum in the visible region at 492 nm (ε492 =8850 L mol−1 cm−1).45 The quantum yield of DFIS photocycli-zation (ΦAc,λ) in EtAc is 0.18 in the interval 310–375 nm.36

Fig. 2 shows the spectral evolution of a DFIS solution inEtAc upon irradiation at 367 nm. The absorption band of 7,7a-DHBF in the visible region (maximum at 492 nm) grew andthe colour of the solution changed from yellow to red. Underconditions of total absorption of light by DFIS (A367 > 2 for ℓ =

1 cm, i.e. PAc,367 > 99%, eqn (3)), a straight line was obtainedby plotting K492 as a function of the irradiation time, as shownin Fig. 3A, and its slope represents the rate of the photochemi-cal reaction (dnAc/dt = (dK492/dt )(V/ε492), eqn (8) and (9),section 2.5.1). When A367 was significantly lower than 2, thelower concentration of DFIS led to a higher fraction ofthe actinometer being converted to 7,7a-DHBF (and to theZ-isomer) and a deviation from the linear trend was observeddue to partial absorption of photons by the latter compounds(Fig. 3B). Under these conditions, the variation of K492 versusirradiation time was interpolated with a polynomial function(third degree) and the slope of the tangent to this function attime zero was considered as the initial rate of the photochemi-cal reaction.

3.2.2 Actinometric experiments in EtAc basedminiemulsions. Fig. 4A shows the comparison of the extinc-tion spectra of a miniemulsion in the absence and in the pres-ence of DFIS dissolved in the EtAc droplets. As expected, thespectral region above 400 nm was not influenced by the pres-ence of DFIS, since there was no absorption in the visible

Fig. 1 Extinction spectra of EtAc based miniemulsions: (A) comparisonof miniemulsions containing 6.0 wtφorg

% but different droplet diameters(□ 150 nm, 100 nm, ○ 73 nm and ▲ 47 nm); (B) comparison of mini-emulsions having 100 nm droplet diameter but different organic phasecontents ( 6.0 wtφorg

%, ○ 4.3 wtφorg% and ▲ 3.0 wtφorg

%) (ℓ = 1 cm).

Fig. 2 Spectral evolution of a solution of DFIS in EtAc under irradiationat 367 nm (ℓ = 1 cm); vertical arrows indicate the direction of the evol-ution (increase of the absorption of the closed form 7,7a-DHBF centredat 492 nm); [DFIS]0 = 2.4 × 10−4 mol L−1.

Fig. 3 Evolution of the absorption coefficient at 492 nm (K492) of a solution of DFIS in EtAc as a function of irradiation time at 367 nm: (A) [DFIS]0 =2.2 × 10−3 mol L−1 (conditions of total absorption of light: K0,367 > 2, ℓ = 1 cm), data fitted with a linear function (solid line); (B) [DFIS]0 = 2.4 × 10−4

mol L−1 (K0,367 = 0.77, ℓ = 1 cm), data interpolated using a polynomial function (solid line) and tangent at time zero to this function (dotted line).

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

region by the open-form of the photochromic compound. Itcan be observed that in the same spectral region the extinctiondue to the scattering remained unchanged after introductionof DFIS. Moreover and within experimental error, the averagesize of the droplets was not affected by the presence of DFIS inthe miniemulsion. Within the range of concentrations used inthis study, the addition of DFIS did not cause any change inthe spectral and colloidal properties of miniemulsions. Theabsorption spectra of DFIS in homogeneous solution and inminiemulsions recorded with the same DFIS concentration areshown in Fig. 4B. The spectrum of DFIS in the miniemulsionwas obtained by subtracting the scattering contribution due tothe medium. The spectra matched extremely well and neitherspectral shift nor change in absorbance was observed, confirm-ing that DFIS was localized within the EtAc droplets.

Fig. 5 shows the comparison of the extinction spectra of asolution of DFIS irradiated at 367 nm for 240 seconds inhomogeneous solution and in a miniemulsion, after subtract-ing the spectrum of the DFIS-free miniemulsion from thelatter. As observed for the non-irradiated systems (Fig. 4B), thespectra are also well matched in the case of the DFIS photopro-duct (7,7a-DHBF). Irradiation of the actinometer in homo-geneous solution and in the miniemulsion under the sameconditions led to an equivalent photochemical conversion ofDFIS and the quantum yield of the photochemical reactionin the miniemulsion can be assumed to be equal to thatin homogeneous solution (0.18). Therefore, the literature valueof the product ФAc,367·ε492 was used for the calculation ofthe absorbed photon flux in EtAc as well as in EtAc basedminiemulsions (eqn (9) and (10)).

The data obtained by actinometry in EtAc miniemulsionsare listed in Table 2. As discussed in section 3.1, the as-pre-

pared miniemulsions of different droplet sizes were diluted toobtain 6.0, 4.3 and 3.0 wtφorg

%. Under these conditions,although the local concentration of the actinometer in the oildroplets remained constant, the nominal concentration ofDFIS in the miniemulsions varied with the organic phasecontent. Therefore, for comparison purposes, actinometricmeasurements were carried out in parallel in homogeneoussolutions of EtAc having the same DFIS concentration. Theratio of the photon flux absorbed in the miniemulsion to that

absorbed in homogeneous solution R ¼ PAc;ME;367

PAc;367

� �was calcu-

lated in each case. The results show that, within experimentalerror (10%), no effect of the variable miniemulsion compo-sition could be observed on the value of R (Table 2): for all thesamples it was nearly 1.0 and the absorbed photon flux in theminiemulsion was comparable to that in homogeneous solu-tion. This finding was unexpected if we consider that, as aresult of the increasing droplet size (from miniemulsions A toD) and the organic phase content (from 3.0% to 6.0%), thescattering coefficient increased and miniemulsions becameless and less transparent (Table 1). Surprisingly, these differ-ences in the optical properties did not seem to affect theefficiency of light absorption by DFIS in the core of the organic(EtAc) droplets, with respect to homogeneous solution. Never-theless, this result is in line with recent findings by Jasinskiet al. concerning the optical properties of a series of dilutedMM based miniemulsions with the droplet size ranging from40 nm to 300 nm in average diameter.24 Using a combinationof spectrophotometric methods, these authors demonstratedthat a droplet size increase caused a significant and progress-ive increase of the scattering coefficient (Sλ), but withoutaffecting the absorption coefficient (Kλ). The latter remainedconstant and comparable to that of a solution of MM at thesame concentration in an organic solvent.

Fig. 4 (A) Extinction spectra of an EtAc miniemulsion containing DFIS(bold line) and DFIS-free (solid line); (B) extinction spectra of DFIS in anEtAc miniemulsion after subtracting the spectrum of the EtAc miniemul-sion (solid line), compared to the spectrum of DFIS in EtAc (bold line);[DFIS] = 2.6 × 10−4 mol L−1; ℓ = 1 cm; miniemulsion composition: B6(Table 1).

Fig. 5 Extinction spectra after irradiation of DFIS at 367 nm in EtAc(bold line), in an EtAc miniemulsion before (dashed line) and after (solidline) subtracting the EtAc miniemulsion spectrum (7,7a-DHBF absorptionband centred at 492 nm); [DFIS]0 = 2.6 × 10−4 mol L−1; ℓ = 1 cm; mini-emulsion composition: B6 (Table 1); irradiation time: 240 seconds.

Photochemical & Photobiological Sciences Paper

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

Hence, it may be concluded that, similarly, the absorptionproperties of DFIS dissolved in miniemulsions based on EtAcwere not significantly affected by scattering under the experi-mental conditions used, the absorbed photon flux being closeto that of DFIS in pure EtAC. It should be noted that, for the12 miniemulsions investigated, the ratio of the absorptioncoefficient of DFIS at 367 nm in EtAc at a given concentrationto the miniemulsion extinction coefficient at the same wave-length (K(DFIS)EtAc,367/E(ME)367, Tables 1 and 2, respectively)varied in the range from 3.9 for the miniemulsion A (the smal-lest droplet size) to a minimum of 0.26 for miniemulsion D(the largest droplet size), respectively, under the experimentalconditions used. Therefore, in spite of an important contri-bution of scattering in miniemulsion D, the DFIS absorptionefficiency was not significantly affected.

3.3 Physical characteristics and optical properties ofminiemulsions based on a mixture of monomers

EtAc based miniemulsions represent a simple model system toinvestigate the effect of the scattering on the efficiency of lightabsorption in the core of the miniemulsion droplets. In orderto evaluate the absorbed photon flux in photopolymerizablesystems, a step forward in the study was taken by replacingEtAc by the MM mixture. In Table 3, the composition, dropletsize and extinction coefficient at 367 nm of the MM based

miniemulsions are reported. Miniemulsions contained 0.15and 1.5 wt/wtφorg

% SDS and they were diluted in order toobtain 3.0 and 6.0 wtφorg

%. The average diameters of the MMdroplets were 54 nm for the smallest one (S) at 1.5 wt/wtφorg

%SDS and 140 nm for the largest one (L) at 0.15 wt/wtφorg

% SDS.Miniemulsions based on EtAc and MM having otherwise

the same composition showed similar colloidal properties;only slightly higher droplet average diameters were observedfor the latter from DLS measurements (Tables 1 and 3). Never-theless, the optical properties of the two types of miniemul-sions were definitely different, in particular light scattering inthe MM based miniemulsions was quite strong even at highdilution (i.e. at low organic phase content). An example is pre-sented in Fig. 6 for miniemulsions containing 0.5 wt/wtφorg

%SDS and 0.6 wtφorg

%.

Table 2 Photon flux absorbed by DFIS at 367 nm (PAc,367) calculated from actinometric experiments performed in EtAc homogeneous solutionsand in EtAc based miniemulsions (PAc,ME,367) (see Table 1 for the composition of miniemulsions and droplet average diameters)a

[DFIS](10−4 mol L−1)

K(DFIS)367in EtAc

PAc,367 in EtAc(10−9 einstein s−1) Miniemulsion

PAc,ME,367 in EtAcminiemulsion(10−9 einstein s−1) R ¼ PAc;ME;367

PAc;367

1.28 0.39 2.3 A3 2.2 1.0B3 2.5 1.1C3 2.5 1.1D3 2.5 1.1

1.83 0.59 2.7 A4 2.7 1.0B4 3.1 1.1C4 2.5 0.9D4 2.5 0.9

2.56 0.84 3.2 A6 3.1 1.0B6 3.1 1.0C6 3.0 0.9D6 2.9 0.9

a Experimental errors ≈10% calculated from repeated measurements (average values are listed).

Table 3 Characteristics of miniemulsions based on a mixture of mono-mers (MM): composition in weight percent (SDS and organic phase con-tents), droplet average diameter (d ) and extinction coefficient at 367 nm(E367, measured in spectroscopic cells of ℓ = 0.01 cm)a

% SDS/φorg d (nm) Miniemulsion %φorg (MM) E367 (cm−1)

0.15 140 L3 3.0 57L6 6.0 115

1.5 54 S3 3.0 21S6 6.0 41

a Experimental errors ≈15% calculated from repeated measurements.

Fig. 6 Comparison of extinction spectra of miniemulsions based onMM (solid line) and on EtAc (dotted line) as an organic phase (0.5 wt/wtφorg

% SDS and 0.6 wtφorg%); ℓ = 1 cm; d = 85 nm and 73 nm,

respectively.

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

3.4 Determination of the photon flux absorbed by a chemicalactinometer in miniemulsions based on a mixture ofmonomers

3.4.1 Reference actinometric experiments on the mixtureof monomers: the homogeneous system. The opticalproperties of DFIS in the homogeneous solution of MMwere first investigated. Fig. 7 shows the comparison of theabsorption spectra of a DFIS solution in EtAc and in MM,before (A) and after (B) irradiation at 367 nm. A slight shift ofthe spectrum was observed and the maxima of the absorptionbands for the open and closed forms of the actinometer inMM were found at 342 nm and 494 nm, respectively. In orderto estimate the photon flux absorbed by DFIS dissolvedin MM, knowledge of the product (ΦAc,367·ε494)MM value wasrequired (eqn (9)). This value was estimated from actinometricexperiments carried out in EtAc and MM homogeneous solu-tions under the same experimental conditions. The actino-metric experiment in EtAc allowed the determination ofthe incident photon flux (P0,367) using eqn (3) and (9). Then,actinometry was performed in MM, and the product(ΦAc,367·ε494)MM was calculated using the same equationsrearranged to eqn (11).

ðΦAc;367 �ε494ÞMM ¼ 1ð1� 10�A367Þ �

dA494dt

� V‘� 1P0;367

ð11Þ

The experiments were repeated four times, giving meanvalues of P0,367 = (3.6 ± 0.2) × 10−9 einstein s−1 and(ΦAc,367·ε494)MM = (1520 ± 11) L mol−1 cm−1 in MM comparedto 1590 L mol−1 cm−1 in EtAc.45 A separate determination ofthe quantum yield and of the molar absorption coefficient wasnot necessary within the scope of this study. The photon flux

absorbed in the homogeneous solution of MM was evaluatedusing eqn (12) (modified eqn (9)):

PAc;367 ¼ dK494

dt� VðΦAc;367 �ε494ÞMM

ð12Þ

3.4.2 Actinometric experiments in miniemulsions basedon a mixture of monomers. Monomer miniemulsions arehighly scattering media due to a broader droplet size distri-bution, in the UV as well as in the visible spectral ranges, andlight penetration is strongly hampered even at low organicphase contents. Thus, in order to estimate the photochemicalconversion of DFIS to 7,7a-DHBF by spectrophotometry, a sub-sequent dilution was required. The irradiated miniemulsionwas diluted by a factor of 3 using an aqueous solution of SDS(200 g L−1). This led to the formation of a transparent pseudo-microemulsion (PM) (Fig. S1†). The compositions of the PMsobtained from the different miniemulsions are listed inTable S1.†

As an example, the absorption spectra of DFIS (2.1 × 10−4

mol L−1) in a homogeneous solution of MM and in a PMobtained from the MM miniemulsion L6, before and afterirradiation at 367 nm, are shown in Fig. 8. A bathochromicshift was observed for DFIS dissolved in PM and the newmaxima of the absorption bands for the open (DFIS) andclosed forms (7,7a-DHBF) were 348 nm and 512 nm, respect-ively. Moreover, a decrease of the absorption coefficient couldbe noticed. In the PM, the high amount of SDS added led toreorganization of the interfacial region that represents a newmicroenvironment for the DFIS molecules. The changed sol-vation shell resulted in a change of the optical properties ofDFIS. The value of the molar absorption coefficient at 512 nmin PM was required for the spectrophotometric determinationof the number of reacted molecules and, thus, of the absorbedphoton flux. Assuming that, upon irradiation, the quantumyield of the DFIS photochemical reaction in the MM organic

Fig. 8 (A) Absorption spectra of DFIS dissolved in MM (dotted line) andin the corresponding pseudo-microemulsion (PM, solid line); (B) absorp-tion spectra of the same solutions after irradiation of DFIS at 367 nm(note the difference in the 7,7a-DHBF absorption band in the visiblespectral range); [DFIS]0 = 2.1 × 10−4 mol L−1; ℓ = 1 cm; PM composition(wt%): 11.81% SDS, 1.77% (φorg + OA), 86.42% water.

Fig. 7 (A) Absorption spectra of DFIS in EtAc (solid line) and in themixture of monomers (MM) (dotted line); (B) absorption spectra of thesame solutions after irradiation at 367 nm for 300 seconds (the 7,7a-DHBF absorption band in the visible range); [DFIS]0 = 2.5 × 10−4 mol L−1;ℓ = 1 cm.

Photochemical & Photobiological Sciences Paper

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

droplets was the same as in the homogeneous solution of MM,the change in the molar absorption coefficient of DFIS at512 nm in PM could be taken into account simply by calculat-ing the product (ΦAc,367·ε512)PM using the following equation:

ðΦAc;367 �ε512ÞPM ¼ K512;PM

K494;MM� ðΦAc;367 �ε494ÞMM ð13Þ

where K512,PM and K494,MM are the absorption coefficients atthe maximum absorption of 7,7a-DHBF in PM and in thehomogeneous solution of MM, respectively.

Experimentally, a solution of DFIS (9.77 × 10−4 mol L−1) inMM was prepared and irradiated at 367 nm in a 1 cm quartzcell for a long enough time period, so that a sufficiently highphotochemical conversion of DFIS to 7,7a-DHBF was obtained.The PM was prepared by adding known amounts of costabili-zer OA and of an aqueous solution of SDS to the irradiatedMM miniemulsion, and the absorption spectrum wasrecorded. It was compared with that of the irradiated MM solu-tion after dilution with MM to obtain the same actinometerconcentration as in PM. This method allowed the determi-nation of the absorption coefficients in the two media at thesame concentration of 7,7a-DHBF, since the two samples wereobtained from the same irradiated solution. The valueobtained for the product (ΦAc,367·ε512)PM was 1368 ± 25 Lmol−1 cm−1 and was used to calculate the absorbed photonflux in MM based miniemulsions by applying eqn (14)(modified eqn (10)), where the moles of the reacted DFIS weredetermined from the absorption spectrum of PM taking intoaccount the dilution factor.

PAc;ME;367 ¼ dK512;PM

dt� VðΦAc;367 � ε512ÞPM ð14Þ

The results from actinometry in miniemulsions based on amixture of monomers are listed in Table 4. They show a

decrease of the ratioPAc;ME;367

PAc;367(0.7–0.9, Table 4) when com-

pared to EtAc based miniemulsions (0.9–1.1, Table 2).However, the larger experimental errors in the case of MMbased miniemulsions, due to the necessity of forming pseudo-microemulsions for the spectrophotometric analysis, have tobe taken into account. Therefore, the decreasing trendobserved may be considered significant only for the highestscattering systems, i.e. when the average diameter of the dro-

plets is the largest (140 nm, L3 and L6, Table 4), the effectbeing more pronounced for L6 (6 wtφorg

%) than for L3 (3wtφorg

%). In any case, considering that the ratio K(DFIS)MM,367/E(ME)367 is only 0.02–0.05 in MM based miniemulsions due toconsiderable scattering (compared to 0.26–3.9 in EtAc basedsystems, section 3.2.2), the effect of the highly increased scat-tering on the photon flux absorbed by DFIS in the formersystems is again surprisingly small. This result confirms thatan increase of the scattering coefficient of the system has onlya negligible to small effect on the absorption properties of theactinometer dissolved in the model (EtAc based), as well as inphotopolymerizable (MM based) miniemulsions. Withinexperimental error, the photon flux absorbed by the oil solubleactinometer in miniemulsions (either moderately or highlyscattering) remains comparable to that of the actinometer dis-solved in the homogeneous organic medium that constitutesthe dispersed phase of the miniemulsion.

4 Conclusion

In recent years, free radical photoinitiated polymerization inoil-in-water miniemulsions containing mixtures of monomershas been proposed as an advantageous method for the prepa-ration of aqueous dispersions of polymer nanoparticles(latexes). In this context, the objective of our study was toevaluate to what extent the efficiency of light absorption insidethe dispersed organic phase (oil droplets) was affected by theoptical properties of the miniemulsions that are known to bestrongly dependent on the droplet size and on the chemicalnature of the oil phase. To this aim, a well-characterizedchemical actinometer (DFIS) insoluble in water, and thus dis-solved in the oil droplets, was used as a probe and a model forlight absorption in the UV spectral range. The absorbedphoton flux could be determined by following the photochemi-cal conversion of DFIS to its photoproduct 7,7a-DHBF by spec-trophotometry in the visible spectral range, where scattering ismuch lower than in the UV region.

In the first approach, actinometry was performed in modelminiemulsions based on an inert solvent (EtAc), at a low oilphase content (ranging from 3.0 to 6.0 wt%). For these low tomoderately scattering systems, the value of the miniemulsionscattering coefficient at the irradiation wavelength (367 nm)did not exceed 4 times the value of the absorption coefficient

Table 4 Photon flux absorbed by DFIS at 367 nm (PAc,367) calculated from actinometric experiments performed in the homogeneous solution ofthe monomer mixture (MM) and in MM based miniemulsions (see Table 3 for the composition of miniemulsions)a

[DFIS] (10−4

mol L−1)K(DFIS)367in MM

PAc,367 in MM(10−9 einstein s−1) Miniemulsion

PAc,ME,367 (MM basedminiemulsion) (10−9 einstein s−1)

PAc;ME;367

PAc;367

3.1 1.06 3.4 S3 3.1 0.9L3 2.7 0.8

6.1 1.93 3.5 S6 3.3 0.9L6 2.6 0.7

a Experimental errors ≈15% calculated from repeated measurements (average values are listed).

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

of the actinometer in EtAc at the same concentration. In thesecases, within experimental error, the photon flux absorbed bythe actinometer in the miniemulsion was comparable to thatin the homogeneous solution of EtAc. Although surprising atfirst sight, this result is in line with a recent finding that, inthe case of diluted miniemulsions, a droplet size increasecaused a progressive increase of the scattering coefficient, butwithout affecting the absorption coefficient of the miniemul-sion itself.

In the second part of this work, the light absorptionefficiency was investigated in photopolymerizable miniemul-sions where the oil phase consisted of the actinometer dis-solved in a mixture of acrylate monomers (MM). These systemsshowed much higher scattering coefficients, in the UV as wellas in the visible spectral ranges, than EtAc miniemulsions ofotherwise the same composition. In these cases, the valueof the miniemulsion scattering coefficient at the irradiationwavelength (367 nm) was 20–50 times higher than that ofthe absorption coefficient of the actinometer in MM at thesame concentration. Nevertheless, a rather small decreasingtrend for the photon flux absorbed by the actinometerin the MM based miniemulsions compared to MMhomogeneous solutions was observed (10%–30%). The effectwas more pronounced for the highest scattering systems, i.e.miniemulsions with the largest average diameter of droplets(140 nm).

Therefore, an increase of the scattering coefficient ofthe system, even considerable such as in photopolymerizable(MM based) miniemulsions, did not affect to a large extentthe efficiency of light absorption by the actinometer. Thisresult may be related to the fact that, although the increasingscattering contribution increases extinction and reducesthe light penetration depth, photons can be scattered severaltimes by the oil droplets and be subsequently absorbed bythe actinometer, becoming efficient for the actinometer photo-reaction. This effect may compensate, at least partially,the consequences of the considerably shorter optical path-length. As the same would apply to any absorbing speciessuch as a polymerization photoinitiator, this result may beconsidered favourable for the further development of appli-cations of photopolymerizations in miniemulsions. Never-theless, polymerization processes require the use ofconcentrated miniemulsions (30 wt% of the organicphase relative to the total weight). In these systems, the occur-rence of multiple scattering due to the density of dropletsmuch higher than in diluted miniemulsions may give rise toa more significant loss of photons. The effective extent ofthis loss and its potential impact on the rate of photo-polymerization in concentrated miniemulsions remain tobe investigated. Although the present study was limited todiluted miniemulsions, this new approach, based on the useof an actinometric method, was useful for a better understand-ing of the correlation between the optical properties of photo-polymerizable dispersed systems and the efficiency ofthe photoinitiation step in the presence of an oil-solublephotoinitiator.

Acknowledgements

The authors gratefully acknowledge the French Agence Natio-nale de Recherche (ANR), Programme Chimie Durable-Indus-tries-Innovation (CDII, ANR-2012-CDII-006-02) for funding.

References

1 B. Charleux and F. Ganachaud, Polymerizations in aqueousdispersed media, in Macromolecular Engineering, ed. K. Maty-jaszewski, Y. Gnanou and L. Leibler, Wiley-VCH VerlagGmbH, Weinheim, 2007, vol. 1, pp. 605–642.

2 N. J. Turro, M.–F. Chow, C.-J. Chung and C.-H. Tung, Mag-netic field and magnetic isotope effects on photoinducedemulsion polymerization, J. Am. Chem. Soc., 1983, 105,1572–1577.

3 X. Hu, J. Zhang and W. Yan, Preparation of transparentpolystyrene nano-latexes by an UV-induced routine emul-sion polymerization, Polymer, 2009, 50, 141–147.

4 L. Liu and W. Yang, Photoinitiated, inverse emulsionpolymerization of acrylamide: some mechanistic andkinetic aspects, J. Polym. Sci., Part A: Polym. Chem., 2004,42, 846–852.

5 S. Mah, D. Lee, D. Koo and S. Kwon, Photo-induced emul-sion polymerization of vinyl acetate in the presence of poly(oxyethylene) nonyl phenyl ether, a non-ionic emulsifier(II), J. Appl. Polym. Sci., 2002, 86, 2153–2158.

6 S. E. Shim, Y. Shin, J. W. Jun, K. Lee, H. Jung and S. Choe,Living-free-radical emulsion photopolymerization ofmethyl methacrylate by a surface active iniferter (surinifer-ter), Macromolecules, 2003, 36, 7994–8000.

7 I. Capek, Photopolymerization of butyl acrylate microemul-sion. Effect of reaction conditions and additives on fates ofdesorbed radicals, Polym. J., 1996, 28, 400–406.

8 G. David, F. Ozer, B. C. Simionescu, H. Zareie andE. Piskin, Microemulsion photopolymerization of metha-crylates stabilized with sodium dodecyl sulphate and poly(N-acetylethylenimine) macromonomers, Eur. Polym. J.,2002, 38, 73–78.

9 L. Wang, X. Liu and Y. Li, Synthesis and evaluation of asurface-active photoinitiator for microemulsion polymeriz-ation, Macromolecules, 1998, 31, 3446–3453.

10 K. Jain, J. Klier and A. B. Scranton, Photopolymerization ofbutyl acrylate-in-water microemulsions: polymer molecularweight and end-groups, Polymer, 2005, 46, 11273–11278.

11 P.-L. Kuo, N. J. Turro, C.-M. Tseng, M. S. El-Aasser andJ. W. Vanderhoff, Photoinitiated polymerization of styrenein microemulsions, Macromolecules, 1987, 20, 1216–1221.

12 L. Wang, X. Liu and Y. Li, Microemulsion polymerizationof styrene using surface-active peresters as photoinitiators,Langmuir, 1998, 14, 6879–6885.

13 J. Tonnar, E. Pouget, P. Lacroix-Desmazes and B. Boutevin,Synthesis of poly(vinyl acetate)-block-poly(dimethylsilox-ane)-block-poly(vinyl acetate) copolymers by iodine transferphotopolymerization in miniemulsion, Macromol. Symp.,2009, 281, 20–30.

Photochemical & Photobiological Sciences Paper

This journal is © The Royal Society of Chemistry and Owner Societies 2014 Photochem. Photobiol. Sci.

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

14 A. V. Fuchs and G. D. Will, Photo-initiated miniemulsionpolymerization as a route to the synthesis of gold nano-particle encapsulated latexes, Polymer, 2010, 51, 2119–2124.

15 I. Capek, Photopolymerization of alkyl(meth)acrylates andpolyoxyethylene macromonomers in fine emulsions, Eur.Polym. J., 2000, 36, 255–263.

16 J. Dou, Q. Zhang, L. Jian and J. Gu, Magnetic nanoparticlesencapsulated latexes prepared with photo-initiated mini-emulsion polymerization, Colloid Polym. Sci., 2010, 288,1751–1756.

17 A. Chemtob, B. Kunstler, C. Croutxé-Barghorn andS. Fouchard, Photoinduced miniemulsion polymerization,Colloid Polym. Sci., 2010, 288, 579–587.

18 P. A. Hoijemberg, A. Chemtob, C. Croutxé-Barghorn, J. Polyand A. M. Braun, Radical photopolymerization in mini-emulsions. Fundamental investigations and technicaldevelopment, Macromolecules, 2011, 44, 8727–8738.

19 P. A. Hoijemberg, A. Chemtob and C. Croutxé-Barghorn,Two routes towards photoinitiator-free photo-polymerization in miniemulsion: acrylate self-initiationand photoactive surfactant, Macromol. Chem. Phys., 2011,212, 2417–2422.

20 I. Capek and C.-S. Chern, Radical polymerization in directmini-emulsion systems, Adv. Polym. Sci., 2001, 155, 101–165.

21 J. M. Asua, Miniemulsion polymerization, Prog. Polym. Sci.,2002, 27, 1283–1346.

22 K. Landfester, N. Bechthold, F. Tiarks and M. Antonietti,Formulation and stability mechanisms of polymerizableminiemulsions, Macromolecules, 1999, 32, 5222–5228.

23 J. M. Asua, Challenges for industrialization of miniemul-sionpolymerization, Prog. Polym. Sci., 2014, 39, 1797–1826.

24 F. Jasinski, E. Lobry, A. Chemtob, C. Croutxé-Barghorn andA. Criqui, Photopolymerizable monomer miniemulsions:why does droplet size matter?, Macromol. Chem. Phys.,2013, 214, 1669–1676.

25 L. Sun and J. R. Bolton, Determination of the quantumyield for the photochemical generation of hydroxylradicals in TiO2 suspensions, J. Phys. Chem., 1996, 100,4127–4134.

26 E. Lobry, F. Jasinski, M. Penconi, A. Chemtob, C. Ley,C. Croutxé-Barghorn, E. Oliveros, A. M. Braun andA. Criqui, Absorption and scattering in concentratedmonomer miniemulsions: static and dynamic investi-gations, Macromol. Chem. Phys., 2014, 215, 1201–1211.

27 F. D. Lewis, R. T. Lauterbach, H.-G. Heine, W. Hartmannand H. Rudolphz, Photochemical α-cleavage of benzoinderivatives. Polar transition states for free-radical for-mation, J. Am. Chem. Soc., 1975, 97, 1519–1525.

28 N. K. Shrestha, E. J. Yagi, Y. Takatori, A. Kawai, Y. Kajii,K. Shibuya and K. Obi, Photochemical α-cleavage reactionof benzoin and its derivatives, J. Photochem. Photobiol. A,1998, 116, 179–185.

29 M. Lipson and N. J. Turro, Picosecond investigation of theeffect of solvent on the photochemistry of benzoin, J. Photo-chem. Photobiol. A, 1996, 99, 93–96.

30 C. Autran, J. C. de la Cal and J. M. Asua, (Mini)emulsionpolymerization kinetics using oil-soluble initiators, Macro-molecules, 2007, 40, 6233–6238.

31 C. Wang, J. Rabani, D. W. Bahnemann andJ. K. Dohrmann, Photonic efficiency and quantum yield offormaldehyde formation from methanol in the presence ofvarious TiO2 photocatalysts, J. Photochem. Photobiol. A,2002, 148, 169–176.

32 N. Serpone and A. Salinaro, Terminology, relative photonicefficiencies and quantum yields in heterogeneous photo-catalysis. Part I: Suggested protocol, Pure Appl. Chem., 1999,71, 303–320.

33 A. Salinaro, N. Serpone, A. Emeline, V. Ryabchuk andH. Hidaka, Terminology, relative photonic efficiencies andquantum yields in heterogeneous photocatalysis. Part II:Experimental determination of quantum yields, Pure Appl.Chem., 1999, 71, 321–335.

34 R. J. Brandi, M. A. Citroni, O. M. Alfano and A. E. Cassano,Absolute quantum yields in photocatalytic slurry reactors,Chem. Eng. Sci., 2003, 58, 979–985.

35 M. L. Satuf, R. J. Brandi, A. E. Cassano and O. M. Alfano,Experimental method to evaluate the optical properties ofaqueous titanium dioxide suspensions, Ind. Eng. Chem.Res., 2005, 44, 6643–6649.

36 H. J. Kuhn, S. E. Braslavsky and R. Schmidt, Chemical acti-nometry. IUPAC technical report, Pure Appl. Chem., 2004,76, 2105–2146.

37 S. E. Braslavsky, et al., Glossary of terms used in photo-chemistry 3rd edition (IUPAC recommendations 2006),Pure Appl. Chem., 2007, 79, 293–465 and references therein.

38 M. I. Cabrera, O. M. Alfano and A. E. Cassano, Absorptionand scattering coefficients of titanium dioxide particulatesuspensions in water, J. Phys. Chem., 1996, 100, 20043–20050.

39 A. M. Braun, M.-T. Maurette and E. Oliveros, PhotochemicalTechnology, Wiley, Chichester, 1991, ch. 2.

40 H. G. Heller, Development of photochromic compoundsfor use in optical information stores, Chem. Ind., 1978,193.

41 H. G. Heller and J. R. Langan, Photochromic heterocyclicfulgides. Part 3. The use of (E)-α-(2,5-dimethyl-3-furylethyl-idene)(isopropylidene)succinic anhydride as a simple con-venient chemical actinometer, J. Chem. Soc., Perkin Trans.2, 1981, 341–343.

42 H. G. Heller and J. R. Langan, A new reusable chemicalactinometer, EPA Newsl., 1981, 71–73.

43 Y. Yokoyama, H. Hayata, H. Ito and Y. Kurita, Photochro-mism of a furylfulgide, 2-[1-(2,5-dimethyl-3-furyl)ethyli-dene]-3-isopropylidenesuccinic anhydride in solvents andpolymer films, Bull. Chem. Soc. Jpn., 1990, 63, 1607–1610.

44 P. Boule and J. F. Pilichowski, Comments about the use ofAberchromeTM 540 in chemical actinometry, J. Photochem.Photobiol. A, 1993, 71, 51–53.

45 E. Uhlmann and G. Gauglitz, New aspects in the photo-kinetics of Aberchrome 540, J. Photochem. Photobiol. A,1996, 98, 45–49.

Paper Photochemical & Photobiological Sciences

Photochem. Photobiol. Sci. This journal is © The Royal Society of Chemistry and Owner Societies 2014

Publ

ishe

d on

10

Oct

ober

201

4. D

ownl

oade

d by

Uni

vers

ite d

e H

aute

Als

ace

on 1

1/11

/201

4 17

:49:

00.

View Article Online

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.