Download pdf - Dissolution of a percolation cluster on UV-cured polymeric films: a fluorescence study

The mainhown thatentangleded that whenfficients,

Journal of Colloid and Interface Science 277 (2004) 359–365www.elsevier.com/locate/jcis

The effect of annealing temperature on latex film dissolution

Saziye Ugur, Önder Pekcan∗

Istanbul Technical University, Department of Physics, 34469 Maslak, Istanbul, Turkey

Received 5 March 2004; accepted 23 April 2004

Available online 21 July 2004

Abstract

The steady-state fluorescence technique (SSF) was used to study the dissolution of polystyrene (PS) latex films in toluene.objective of this study was to explore the effects of annealing temperature on the dissolution behavior of the PS films. It has been sdissolution coefficients of PS films are highly affected by the annealing temperatures. It was found that, while PS films dissolved asclusters when annealed at low temperatures, they dissolved as individual chains at high annealing temperatures. The results showthe annealing temperature increases, the dissolution takes place in the disentanglement regime. The decrease in dissolution coeDd,can be explained as a shift of dissolution from the entangled clusters regime to the disentanglement regime. The measuredDd values werefound to be in between 4.2 and 1.0× 10−10 cm2 s−1 for an annealing temperature range of 110–220◦C. 2004 Elsevier Inc. All rights reserved.

Keywords:Latex film; Fluorescence; Dissolution; Annealing temperature

ly-d of

sncessrane

odis-assyllyer.r

theis-nt-ingan-

byffu-the

on-nalo thefu-inmerif-ticisso-

yflex-sle-f the

atedyerles.

hanb-ringnd-

1. Introduction

Polymer dissolution is an important phenomenon in pomer science and engineering. For example, in the fielcontrolled drug release, zeroth-order drug release systemhave been designed[1] by rendering the polymer dissolutiophenomenon as the controlling step in the release proPolymer dissolution also has some applications in membscience[2], the semiconductor industry[3], microlithogra-phy [4], and packaging[5].

The dissolution of a polymer in a solvent involves twtransport processes, namely, solvent diffusion and chainentanglement. When an uncross-linked, amorphous, glpolymer is brought into contact with a thermodynamicacompatible solvent, the latter diffuses into the polymA gel-like layer is formed adjacent to the solvent–polymeinterface due to the plasticization of the polymer bysolvent. After an induction time, the polymer starts to dsolve. Polymer dissolution from the bulk is very differefrom, and more complicated than, small molecule dissolution. Polymers require an induction time before startto dissolve, while nonpolymeric materials dissolve inst

* Corresponding author.E-mail address:[email protected] (Ö. Pekcan).

0021-9797/$ – see front matter 2004 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2004.04.040

.

taneously. Polymer dissolution can be controlled eitherthe disentanglement of the polymer chains or by the dision of the chains through a boundary layer adjacent tosolvent–polymer interface. However, the dissolution of npolymeric materials is generally controlled by the extermass transfer resistance through a liquid layer adjacent tsolid–liquid interface and can be explained by simple difsion laws[6]. However, polymeric glass dissolves mainlythree different stages: (a) solvent penetration, (b) polyrelaxation and formation of a moving boundary, and (c) dfusion of polymer chains into solvent reservoir. A schemarepresentation of these three sequential steps for the dlution of a polymer glass is presented inFig. 1. In the firststage, the penetration distance of solvent molecules mainldepends on free volume, which in turn depends on theibility of the chains, backbone, and side groups, as well athe thermal history of the polymer. These first solvent mocules act as plasticizers, and as a result these regions ofilm start to swell. In the second stage, the gel layer is creby the relaxing polymer chains. This moving transition lais composed of both polymer chains and solvent molecuIf the solvent–polymer interactions are more dominant tthe polymer–polymer interactions, maximum swelling is otained. This is the case when a good solvent is used dudissolution of a polymer glass. Here an advancing bou

http://www.elsevier.com/locate/jcis

360 S. Ugur, Ö. Pekcan / Journal of Colloid and Interface Science 277 (2004) 359–365

ne-in

akes

s thentedthessstir-du

var-ightof

-lvenexg of

oef-lesneueionution

-in

orasonmer

utioningi-

ari-cu-

,eal-was

oflms

toentdis-d outn Pho-gdif-nd

ia aoly-tted

n-atecon-of

94)the)cediza-

is-rops

were

-sthem-

g atse

influ-that

ht.lled

ci-

o-the

(a)

(b)

(c)

Fig. 1. Cartoon presentation of the dissolution of film: (a) solvent petration, (b) formation of gel layer, (c) propagation of gel layer and chadesorption.

ary is formed. In the last stage, chain disentanglement tplace, and then chains separatefrom the bulk and diffuse intothe solvent, while the advancing boundary moves acrospolymeric glass. The rate of dissolution can be represeby a velocity of solvent penetration, which determinesvelocity of the gel front penetrating into the polymeric glasubstance. The velocity of dissolution increases with therer frequency and decreases with increasing chain lengthto increasing entanglement of the polymer chains.

Solvent penetration in polymers has been studied byious techniques. The most traditional ones are the wemeasurements and the monitoring of the redistributionisotopic tracers in the polymer[7]. The electron spin resonance (ESR) technique is used to investigate nonsopenetration into poly(methyl methacrylate) (PMMA) latparticles[8]. The ESR method, based on the scavenginradicals produced by high-energyγ -irradiation of PMMAby oxygen, was used for the measurement of diffusion cficient in PMMA [9]. Penetration of naphthalene molecuinto PMMA latex particles stabilized by polyisobutyle(PIB) was studied by a time-resolved fluorescence techniqbelow Tg [10]. Fluorescent quenching and depolarizatmethods have been used for penetration and dissolstudies in solid polymers[11–13]. An in situ fluorescencequenching experiment in conjunction with laser interferometry was used to investigate dissolution of PMMA filmvarious solvents[14]. A real-time nondestructive method fmonitoring small-molecule diffusion in polymer films wdeveloped[15,16], which is basically based on the detectiof excited fluorescent molecules desorbing from a polyfilm into a solution in which the film is placed[17,18]. The insitu steady-state fluorescence (SSF) findings on dissolof latex film and polymer glasses using real-time monitorof fluorescent probes[19–22]desorbing from these materals were reported.

The polymer dissolution process can be affected by vous parameters, including solvent quality, polymer mole

e

t

lar weight, solvent thermodynamic compatibility, agitationand temperature. In this work, the effect of the anning temperature on the latex film dissolution processstudied using the SSF method by real-time monitoringthe pyrene (P) intensity change. P-labeled PS latex fiwere annealed at elevated temperatures ranging from 110220◦C for 10 min. Toluene was used as a dissolution agand in situ SSF experiments were performed to monitorsolution processes. Dissolution experiments were carrieby illuminating the toluene reservoir and the increase iintensity,IP, was observed using a fluorescent spectroptometer. Dissolution coefficients,Dd, were measured usincurves of P intensity versus time by employing a case Ifusion model.Dd values were found in between 4.25 a1.0× 10−10 cm2 s−1.

2. Experimental

Fluorescent polystyrene (PS) latex was produced vsurfactant-free emulsion polymerization process. The pmerization was performed batchwise using a thermostareactor equipped with a condenser, thermocouple, mechaical stirring paddle, and nitrogen inlet. The agitation rwas 400 RPM and the polymerization temperature wastrolled at 70◦C. Water (80 g), styrene (4.8 g), and 0.012 gfluorescent 1-pyrenylmethyl methacrylate (PolyFluor 3were first mixed in the polymerization reactor and whentemperature was constant (at 70◦C), the KPS initiator (0.2 gdissolved in a small amount of water (3 ml), was introduin order to induce styrene polymerization. The polymertion was conducted for 17 h.

Seven different latex films were prepared from the dpersion of PS particles by placing the same number of don glass plates with a size of 0.8 × 2.5 cm2 and allowingwater to evaporate at room temperature. Then samplesseparately annealed aboveTg of PS, 105◦C, for 10 min attemperaturesTan ranging from 110 to 220◦C. The temper-ature was maintained within±2◦C during annealing. Samples were weighted before and after the dissolution procesis completed, to determine the film thickness, which iscritical parameter for modeling of the dissolution phenoenon.

In situ dissolution experiments were performed usinPerkin–Elmer LS-50 spectrofluorimeter. All measuremenwere made at the 90◦ position and the slit widths werkept at 8 mm. Dissolution experiments were performeda 1× 1 cm quartz cell which was placed in the spectroorimeter. The fluorescence emission was monitored sofilm samples were not illuminated by the excitation ligFilm samples were placed at one side of a quartz cell fiwith toluene and the cell was then illuminated with 346-nmexcitation light. The positions of the latex films and the extation and emission light are presented inFig. 2. The pyrenefluorescence intensity,IP, was monitored during the disslution process at 394 nm using the “time drive” mode of

S. Ugur, Ö. Pekcan / Journal of Colloid and Interface Science 277 (2004) 359–365 361

ively.

usly

delsandfu-enthisthe

ce.ed tol

omerderf theec-g

pedr

desl is

set

i-of a

des-l,

thely-atesd-

lene ofion.thepre-th

r the

s

, and

ythat

orb-ssthegsid-n

Fig. 2. Dissolution cell in LS-50 Perkin–Elmer spectrofluorimeter.I0 andIPare the excitation and emission intensities at 346 and 394 nm, respect

spectrofluorimeter. Emission of P was recorded continuoat 394 nm as a function of time.

3. Case I and case II diffusion

Various mechanism and various mathematical mohave been considered for the polymer dissolution. TuQuano[23] proposed a model that includes polymer difsion in a liquid layer adjacent to the polymer and movemof the liquid–polymer boundary. The key parameter for tmodel was the polymer disassociation rate, defined asrate at which polymer chains desorb from the gel interfaLee and Peppas[24] extended this model for films to thpolymer dissolution rate, where gel thickness was founbe proportional to(time)1/2. A relaxation-controlled modewas proposed by Brochard and de Gennes[25], where, aftera swelling gel layer was formed, desorption of polymer frthe swollen bulk was governedby the relaxation rate of thpolymer stress. This rate was found to be of the same oof magnitude as the reptation time. The dependencies oradius of gyration and the reptation time on polymer molular weight and concentration were studied, using a scalinlaw [26] based on the reptation model.

In this paper, we employed a simpler model develoby Enscore et al.[27] to interpret the results of polymeswelling and dissolution experiments. This model incluCase I and Case II diffusion kinetics. The Case I modethe solution of a unidirectional diffusion equation for aof boundary conditions that is cited by Cranck[28]. For aconstant diffusion coefficient,D, and fixed boundary condtions, the sorption and desorption transport in and outthin slab are given by the following relation

Mt

M∞= 1− 8

π2

∞∑n=0

1

(2n + 1)2 exp

(−(2n + 1)2Dπ2t

d2

).

(1)

Here,Mt represents the amount of materials absorbed ororbed at timet , M∞ is the equilibrium amount of materiaandd is the thickness of the slab.

The case II transport mechanism is characterized byfollowing steps. As the solvent molecules enter into the pomer film, a sharp advancing boundary forms and separthe glassy part of the film from the swollen gel. This bounary moves into the film at a constant velocity. The swolgel behind the advancing front is always at a uniform statswelling and its thickness stay constant during dissolutSolvent penetration (a), formation (b), propagation (c) ofgel layer, and desorption (c) of the polymer chains aresented inFig. 1. Now, consider a cross-section of a film withicknessd , undergoing case II diffusion, whereL is the po-sition of the advancing sorption front,C0 is the equilibriumpenetrant concentration, andk0 (mg/cm2 min) is defined asthe case II relaxation constant. The kinetic expression fosorption in the film slab of an areaA is given by

(2)dMt

dt= k0A.

The amount of penetrant,Mt, absorbed in timet will be

(3)Mt = C0A(d − L).

After Eq. (3) is substituted intoEq. (2) the following rela-tionship is obtained:

(4)dL

dt= − k0

C0.

It can be seen that the relaxation front, positioned atL,moves toward the origin with a constant velocity,k0/C0. Thealgebraic relation forL, as a function of timet , is describedby Eq. (5):

(5)L = d − k0

C0t .

SinceMt = k0At andM∞ = C0Ad , the following relation-ship is obtained:

(6)Mt

M∞= k0

C0dt.

4. Results and discussion

Plots of P intensity,IP, versus dissolution time for filmannealed at different temperatures are shown inFig. 3. It isseen that all curves present a sudden jump at early timesthen increase slowly at later times. The curves inFig. 3seemto follow a case I (Fickian) diffusion model except at earltimes. In processing the dissolution data, it is assumedIP is proportional to the number of P-labeled chains desing from the gel layer of the PMMA film. Since the thickneof the gel layer stays constant during the dissolution ofpolymer film,Eq. (1)turns to be the solution of the movinslab problem. Here, since the solvent reservoir is conered to be infinitely large (seeFig. 2), the above assumptio


sthe a

at

ur-ngecanition

-rns

t thein

-

f the-

s. Atrti-as ationme

ms.nce

g

-

o-

ormledar. Atat thex-and

Fig. 3. Pyrene emission intensities,IP, versus dissolution time for the filmannealed at various temperatures. Numbers on each curve representnealing temperature.

can be used to employEq. (1). The logarithmic form ofEq. (1)is written forn = 0, with Ad = Ddπ

2/(d − L)2 andBd = ln(8/π2), as follows:

(7)ln

(1− IP

IPm

)= Bd − Adt .

Here, IPm presents the number of P-labeled chainsequilibrium,Dd is the dissolution coefficient, and (d − L) isthe initial thickness of the gel layer, which stay constant ding dissolution. (d −L) values can be estimated by knowithe film thickness,d , and the final dissolution time. If onknows the gel formation time, then the dissolution speedbe calculated, and can be used to estimate the initial posof the gel front,L. Figs. 4a, 4b, and 4cpresent dissolutioncurves for 130, 150, and 200◦C, which are digitized for numerical treatment according toEq. (7). When these lineacurves inFigs. 4a, 4b, and 4care compared to computatiousingEq. (7), dissolution coefficients,Dd, of P-labeled poly-mer molecules are obtained. Here it has to be noted thaearly portions of the dissolution curves are not includedthe fitting procedure to determine theDd values. The producedDd values are listed inTable 1and are plotted inFig. 5against annealing temperature. Here the higher values odissolution coefficient,Dd, for the low annealing temperatures are most probably belongs to the small PS chainlow annealing temperatures film still owns individual pacles and one may expect dissolution of these particleswhole. Some small chains may also contribute to dissoluprocess. At higher annealing temperatures however, polychains relax and form mechanically strong continuous filThese films dissolve much more slowly due to the resistaof the polymer film to the solvent molecules; as a resultDdvalues are found to be much smaller.

To make this point more clear, the SEM micrographs[29]of these films are presented inFig. 6 at elevated annealintemperatures. InFig. 6a, film annealed at 120◦C, no par-

n-

r

Fig. 4. The logarithmic plots of the curves inFig. 3 and their fits according to Eq. (7) for the films annealed at (a) 130, (b )150, and (c) 200◦C,respectively. The slopes of the straight lines produce curve dissolution cefficients,Dd.

Table 1Experimentally produced parameters

Tan(◦C)

d

(µm)Dd

(cm2 s−1) × 10−10k0

(mgcm−2 min−1)×10−2

110 21.9 4.25 –130 22.3 2.14 0.76150 23.8 1.15 0.24170 17.0 1.33 0.91180 18.1 0.95 0.22200 22.8 1.48 0.41220 22.8 1.35 0.07

Note. Tan: annealing temperature;d : film thickness;Dd: dissolution coeffi-cient;k0: relaxation constant.

ticle deformation is observed; close-packed particles fa powder film that includes many voids. In film anneaat 130◦C (Fig. 6b), interparticle voids start to disappeand polymeric material occupies the interparticle voidshigh annealing temperatures the healing process startsparticle–particle junction[30–32]and polymer chains relaacross the junction surface (seeFig. 6c). Above this temperature (Fig. 6d), particle boundaries completely disappear,


ong,onlingeen

seer

les.dis-

-

nt.tionres.s; aster-maysug-oef-sys-eal-earssur--mer.theu),

n-ious-s,lventids,ver,n-ring

consequently, the latex film becomes a mechanically strtransparent film. The effect ofthese stages on the dissolutiprocess can be understood as follows. At a low anneatemperature, since the film contains many voids (as sfrom Fig. 6a and 6b), solvent molecules enter into thevoids without any resistance by the film. As a result, polymfilm dissolves as entangled clusters, i.e., individual particAt this stage of dissolution, however, some small chains

Fig. 5. The plot of the dissolution coefficients,Dd, against annealing temperatures.

solve immediately, presenting a high diffusion coefficieHere one may argue the shortcomings of the dissolumodel for the film samples annealed at lower temperatuIn other words, these samples are not quite homogeneoua result, the dissolution model can partially be used to inpret the data. In other words, some parts of these filmshave local homogeneities that can be treated with thegested dissolution model and the produced dissolution cficients can be accepted as the local parameters for thetem under consideration. In films annealed at higher anning temperatures, the particle–particle interface disappand polymer chains completely relax across the junctionface, which makes the film void-free; as a result, during dissolution, disentanglement takes place between the polychains and the polymer film dissolves much more slowly

To support these findings, the maximum value ofemission intensity of P at infinite time (i.e., at the plateaIPm is plotted versus annealing temperatures inFig. 7. Asseen fromFig. 7, IPm has substantially increased with icreasing annealing temperature, which supports our prevargument about the behavior ofDd coefficients versus annealing temperatures. Film annealed at low temperatureIPm, presents small values. It appears that as the somolecules enter into the film, due to the interparticle vomost of the PS film dissolve as individual clusters. Howein films annealed at high temperatures, PS chains are disetangled and desorb into the solution homogeneously du

(a) (b)

(c) (d)

Fig. 6. Scanning electron micrographs(SEM) of latex films annealed for 10 min at (a) 120, (b) 130, (c) 150, and (d) 170◦C (adopted from Ref.[29]).


a) at

-d

eates-

c

nterede-e o

, and

an-ss.

clus-oly-cessf thean-oef-

J.

.ert,

Fig. 7. The plot of the maximum pyrene intensity,IPm, versus annealingtemperatures,Tan.

(a)

(b)

Fig. 8. Cartoon presentation of the dissolution of PS films annealed (low and (b) at high temperatures.

the dissolution, yielding highIPm values. This picture is depicted inFigs. 8a and 8b, where dissolution of films annealeat low and high temperatures is shown.

Curves inFig. 3at very short times are plotted inFigs. 9a,9b, and 9caccording toEq. (6). Here it is assumed that thnumber of toluene molecules entering into the latex filmearly times is proportional to the number of PS chains dorbing from the swollen gel. NowEq. (6)becomes

(8)IP

IPm= k0

C0Lt.

Fitting Eq. (8)to the data presented inFigs. 9a, 9b, and 9produced thek0 parameters, which are listed inTable 1. Ithas to be noted that the observedk0 values are independeof the annealing temperature for latex film formation. Hone may state that polymer relaxation in latex films is inpendent of the annealing temperature at the early staddissolution.

f

Fig. 9. Comparison of early time region of the data inFigs. 4a–4cwiththe computations usingEq. (8). Relaxation constants,k0, were obtainedfrom the slopes of the plots for the films annealed at (a) 130, (b) 150(c) 200◦C.

5. Conclusion

In conclusion, this work has presented the effect ofnealing temperature on polymer film dissolution proceLow annealing temperatures cause the films dissolve asters. However, films annealed at high temperatures pmer chains dissolve individually as disentanglement protake place due to the increase in mechanical strength ofilms. These films dissolved much slower than the filmsnealed at low temperatures and yield lower dissolution cficients,Dd.

References

[1] P. Colombo, P.L. Catellani, N.A. Peppas, L. Maggi, U. Conte, Int.Pharm. 88 (1992) 99.

[2] C.S. Tsay, A.J. McHugh, J. Polym. Sci. Phys. Ed. 28 (1990) 1327[3] T.F. Yeh, A. Reiser, R.R. Dammel, G. Pawlowski, H. Roesch

Macromolecules 26 (1993) 3862.


ro-em.ing-

.

.on,

37.

cal

2.

,

pl.

92)

.

3.859.93..03)

[4] M.J. O’Brein, D.S. Soane, in: D.W. Hess, K.F. Jensen (Eds.), Micelectronics Processing: Chemical Engineering Aspects, in: Am. ChSoc. Adv. Chem. Ser., vol. 221, American Chemical Society, Washton, DC, 1989, pp. 325–376.

[5] L. Wielgolinkski, Polym. Prepr. Am. Chem. Soc. Div. PolymChem. 32 (2) (1991) 15.

[6] J. Crank, The Mathematics of Diffusion, Clarendon, Oxford, 1975[7] J. Crank, G.S. Park, Diffusion in Polymers, Academic Press, Lond

1968.[8] Z. Veksli, W.G. Willer, J. Polym. Sci. 54 (1976) 299.[9] Y. Kaptan, Ö. Pekcan, O. Güven, J. Appl. Polym. Sci. 37 (1989) 25

[10] Ö. Pekcan, J. Appl. Polym. Sci. 49 (1993) 151.[11] J.E. Guillet, in: M.A. Winnik (Ed.), Photophysical and Photochemi

Tools in Polymer Science, Reidel, Dordrecht, 1986.[12] T. Nivaggioli, F. Wang, M.A. Winnik, J. Phys. Chem. 96 (1992) 746[13] D. Pascal, J. Duhamel, Y.Wang, M.A. Winnik, D.H. Napper,

R. Gilbert, Polymer 34 (1993) 1134.[14] W. Limm, G.D. Dimnik, D. Stanton, M.A. Winnik, B.A. Smith

J. Appl. Polym. Sci. 35 (1988) 2099.[15] L. Lu, R.G. Weiss, Macromolecules 27 (1) (1994) 219.

[16] V.V. Kronganz, W.F. Mooney, J.W. Palmer, J.J. Patricia, J. ApPolym. Sci. 56 (9) (1995) 1077.

[17] V.V. Kronganz, R.M. Yohannan, Polymer 31 (9) (1990) 1130.[18] Z. He, G.S. Hammond, R.G. Weiss, Macromolecules 25 (1) (19

501.[19] Ö. Pekcan, ¸S. Ugur, Polymer 43 (6) (2002) 1937–1941.[20] Ö. Pekcan, ¸S. Ugur, J. Coating Technol. 75944 (2003) 55–63.[21] S. Ugur, Ö. Pekcan, J. Appl. Polym. Sci. 77 (5) (2000) 1087–1095[22] S. Ugur, Ö. Pekcan, Polymer 44 (2003) 2041–2047.[23] Y.O. Tu, A.C. Quano, IBM J. Res. Dev. 21 (1977) 131.[24] P.I. Lee, N.A. Peppas, J. Control. Release 6 (1987) 207.[25] F. Brochard, P.G. de Gennes, PhysicoChem. Hydrodyn. 4 (1983) 31[26] J.S. Papanu, D.S. Soane, A.T. Bell, J. Appl. Polym. Sci. 38 (1989)[27] D.J. Enscore, H.B. Hopfenberg, V.T. Stannett, Polym. 18 (1977) 7[28] J. Crank, The Mathematics of Diffusion, Clarendon, Oxford, 1975[29] S. Ugur, A. Elaissari, Ö. Pekcan, J. Colloid Interface Sci. 263 (20

674–683.[30] M. Canpolat, Ö. Pekcan, Polymer 36 (1995) 2025.[31] Ö. Pekcan, Trends Polym. Sci. 2 (1994) 236.[32] Ö. Pekcan, M. Canpolat, A. Göçmen, Polymer 34 (1993) 2025.