Langmuir monolayers of non-ionic polymers: equilibrium or metastability? Case study of PEO and its PPO-PEO diblock copolymers.

Journal of Colloid and Interface Science 449 (2015) 494–505

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Langmuir monolayers of non-ionic polymers: Equilibrium or metastability? Case study of PEO and its PPO–PEO diblock copolymers Louise Deschênes a,⇑, François Saint-Germain a, Johannes Lyklema b a b

Food Research and Development Centre, 3600 Casavant Blvd West, Saint-Hyacinthe, QC J2S 8E3, Canada Laboratory for Physical Chemistry and Colloid Science, Wageningen University, Dreijenplein 6, 6703 HB Wageningen, Netherlands

g r a p h i c a l a b s t r a c t Apparent PEO losses

%Δ: ∼22%, 15 min up

%Δ: ∼22%, 15 min up %Δ: ∼30%, 2h up

%Δ: ∼50%, 2h up

a r t i c l e

i n f o

Article history: Received 1 December 2014 Accepted 26 January 2015 Available online 7 February 2015 Keywords: PEO PPO–PEO Hysteresis Equilibrium Langmuir monolayers Deborah number Self-similarity

a b s t r a c t Stability and reorganization in Langmuir ﬁlms of PEO in PEO hom*opolymers and PPO–PEO block copolymers were investigated using ﬁlm balance measurements. The apparent fractional losses of EO segments transferred into the subphase resulting from successive compression–expansion cycles have been estimated. The apparent loss is mainly Cmax, Mn and time-dependent. At surface concentrations C 6 0.32 mg/m2, PEO ﬁlms are in equilibrium. For 0.32 6 C 6 0.7 mg/m2, the losses remain modest. Further compression leads to densiﬁcation of the monolayer, requiring the interplay of thermodynamics and kinetic factors In the plateau regime, the loss is higher and constant for 1 6 Cmax 6 2 mg/m2 upon maintaining the achieved surface area for 15 min. Similar losses were obtained for PEO hom*opolymers of high Mn and PPO353–PEO2295. It suggests that the PEO remains anchored in a metastable state at the air–water interface at surface concentration well above the onset of the plateau. Additional losses are incurred for PEO hom*opolymers for monolayers kept compressed in the plateau for 2 h. For the interpretation of these phenomena a combination of elements from self-consistent ﬁeld theory and scaling is desirable with as a trend an increasing contribution of the latter with increasing surface concentration. Crown Copyright Ó 2015 Published by Elsevier Inc. All rights reserved.

1. Introduction This paper deals with deposited, or Langmuir, monolayers of uncharged polymers, emphasizing poly(ethylene oxide), PEO. Several reasons can be advanced to motivate this choice. In the ﬁrst place PEO is an interesting polymer. For the ethylene oxide monomer water is a good solvent, nevertheless its polymers are surface-active, the more so the higher the degree of polymerization is. Hence, surface pressures p can be measured as a function of the molecular area, A, so that information for two-dimensional

equations of state can be obtained. Surface pressures of up to about 10 mN/m can be easily reached at room temperature provided the molecular mass M is not too low. If by compression in a Langmuir trough it is tried to increase the pressure too much, the PEO layer starts to collapse: PEO will be expelled from the surface and transported to the subphase (water). The pressure of the collapse shows both Mn and temperature-dependency [1]. In a previous paper [2] we have extensively studied the thermodynamic properties of p(A) curves as functions of the temperature T and molecular mass. Thermodynamic analysis enabled us to obtain the adsorption energy and entropy per EO segment. For the former 1.2 kT was found, a slightly positive value, implying that the driving force for the adsorption has an entropic origin. For many other polymers

495

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

values of the same order of magnitude were found, some of them positive, others negative. Apparently, the driving force for attachment is in an interesting way related to the hydration changes involved in transferring segments from the bulk to the surface. Another option is the possibility of obtaining from the temperature coefﬁcient of the surface pressure the (differential) surface excess entropy Sa Bð@S=@AÞT where S is the total excess entropy of the layer. This surface excess entropy contains information on the number of degrees of freedom and it therefore offers a means of quantiﬁcation the conﬁguration of the deposited polymer layers. In a following paper, we hope to return to that. Still another promising approach was carrying out similar studies with PEO–PPO (polypropylene oxide) block copolymers. The propylene oxide monomer is poorly soluble in water, and the polymer dissolves not at all. However, Langmuir monolayers can be made of PPO and it is not surprising that these can withstand much higher surface pressures, about 15 mN/m, also depending on M. In [2] we paid much attention to PEO–PPO block copolymers which essentially consist of a water-soluble and a water-insoluble moiety. The PPO blocks act as anchors for the PEO, so that the properties of PEO’s can be studied at much higher surface pressures than in the absence of the hydrophobic block. The properties of tethered PEO loops in such systems can be compared with those attached to less strongly bound trains of hom*opolymers. In the initial (low concentration) part of the p(A) isotherm the molecules at the interface can spread completely, leading to dumbbells consisting of a 2D PEO and a chemically bound 2D PPO islet to form a double pancake in the case of the PPO–PEO block copolymer. Upon compression p will rise till 2D space ﬁlling is complete, followed by expulsion of EO segments into the subphase. For a soluble polymer like PEO the collapse can imply complete desorption, but for PPO– PEO copolymers the collapse remains restricted to the PEO part at surface pressures 6 15 mN/m. This phenomenon results in the appearance of a plateau in the isotherms. The simultaneous rapid increase of Sa witnesses the increased space available for the PEO moieties. Not much is known about the timescale of the various phenomena involved. Consequently, the validity and pertinence of using Langmuir ﬁlms isotherms to quantify thermodynamics properties of polymeric monolayers remains an open issue. All of this referred to essentially static properties. We had to ascertain that equilibrium properties were measured, otherwise it would have been impossible to extract thermodynamic quantities or, for that matter, the time t did not play a role. However, additional interesting information on the monolayer properties can be obtained by evaluating the dynamics of processes occurring upon changing the conditions. Variables that can be studied include rates of equilibration processes, and the dynamic response of monolayers to imposed compression – expansion cycles. Otherwise stated, we shall include hysteresis phenomena. Some studies report the observation of hysteresis in PPO and polystyrene (PS)-based PEO block copolymers [3–8]. The general trend is a shift of the curves towards higher surface concentrations. For both PEO–PPO–PEO and PS–PEO block copolymers, the loss between compression and expansion was dependent on the maximal surface pressures achieved and on the structure of the copolymers. The interpretation of those hystereses is not straightforward as the PEO–PPO–PEO contained only short PEO moieties (Neo 6 103) [5] and as PS–PEO forms aggregates at the air/water interface [9,10]. Surprisingly, information is desperately missing about hysteresis of PEO hom*opolymers. The nature of what is going on in the plateau regime of the hom*opolymer and that of the PEO-based surface active block copolymers remains obscure. In practice, this question is mostly simply approached on the basis of dissolution of (parts of) the polymer. In fact, the kinetics of this particular regime of PEO chains is not well understood and had never

received the attention it deserves. Improving this situation will be the main theme of the present paper. The paper is essentially experimental, although we shall venture giving suggestions for interpretation at the molecular level. This paper is meant to be a tribute to a Festschrift dedicated to Darsh T. Wasan. One of us (JL) remembers vivid discussions with him on a related dynamic phenomenon, namely the drainage of free liquid ﬁlms stabilized by concentrated surfactants, leading to the transient, stepwise formation of stratiﬁed ﬁlms, i.e. ﬁlms consisting of multiple duplex surfactant layers. In reﬂected light such ﬁlms are visible as ﬁlms with different shades of grey. They may be looked upon as smectic 2D liquid crystals. The oldest references to stratiﬁed liquid ﬁlms date back to Johonnott [11] and Perrin [12]. We observed the phenomenon for ﬁlms stabilized by LiDS or NaDS [13]. More recently, Sethumadhavan, Nikolov and Wasan extended these studies to ﬁlms containing micelles and/or colloidal particles, see for instance [14], where older references can be found. This digression illustrates the richness of physical phenomena observable with dynamic measurements. 2. Materials and methods 2.1. Materials The PEO hom*opolymers used were obtained from Polymer Laboratories (PL UK, now part of Varian) and the PPO–PEO diblock copolymers were supplied by Polymer Source Inc. (Montreal, Canada). Their main physical characteristics are presented in Table 1. The water used for the subphase (resistivity of 18.2 MX cm) was puriﬁed with an Easypure II system (Barnstead International, Dubuque, IA, USA). HPLC grade chloroform was purchased from Sigma–Aldrich, (Oakville Ontario, Canada). Polymers and solvent were used as received. 2.2. Film balance measurements The Langmuir ﬁlms p(A) isotherms are obtained by ﬁlm balance measurement. The experimental procedure starts with the deposition of a known amount of polymer, dissolved in chloroform on the water surface in a Langmuir trough, let the solvent evaporate and measure the surface pressure p as a function of the molecular area A, which can be subjected to compression and/or expansion. The experiments were carried out in a KSV2000 system (KSV Instruments Ltd., Helsinki, Finland) with an effective trough surface area of 150 530 mm2 and a trough volume of 950 ml. The system was installed on an anti-vibration table (System 63-541, TMC, MA, USA). IUPAC recommendations [15] were heeded and the Trurnit’s deposition method [16] was applied. The polymer solutions

Table 1 Characteristics of polymers used.

Polymer

Neo

Npo

Mn (g/mol)

PIa

PEO86 PEO136 PEO177 PEO256 PEO431 PEO1200 PEO2261 PEO5966 PEO19200 PPO259 PPO95PEO409 PPO353PEO2295

86 136 177 256 431 1201 2261 5966 19,205 0 409 2295

0 0 0 0 0 0 0 0 0 259 95 353

3780 5990 7800 11,260 18,970 52,850 99,500 262,500 845,000 15,000 23,500 121,500

1.03 1.05 1.04 1.07 1.07 1.06 1.04 1.06 1.11 1.16 1.17 1.25

PI: polydispersity index.

496

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

(otherwise stated, concentrations of about 1 mg/ml) were deposited on a water subphase using gas-tight syringes (Hamilton Co., Reno, USA). The solvent ﬂush injection technique was used for higher accuracy. Upon deposition, ﬁfteen minutes were allowed for solvent evaporation. Otherwise stated, p(A) isotherms were recorded under symmetric compression applied at a rate of 5 mm/min. The system was thermostated (variation ±0.2 K) for isotherms collected at temperature other than 295 K. The isotherms reported at 295 were collected at room temperature with a variation of ±0.7 K. The surface pressure was measured with platinum Wilhelmy plates. Otherwise stated the barrier speed was 5 mm/min. The effect of water evaporation has been veriﬁed and was shown to induce a variation less than 0.08 mN/m at 295 K for a pure air/water interface over a period of 24 h. Pressure drops at constant areas have been monitored. Hysteresis incurred upon compression–expansion experiments have been studied out for critical surface concentrations of interest for PEO, for its hom*opolymers and block copolymers. The effects of waiting time, molecular weight, limiting surface pressures and concentrations and temperature have been explored in surface pressure ranges corresponding to various PEO regimes. Compression– decompression experiments carried out at various parts of the isotherm relating to different properties of the monolayer and hence to different time behaviour.

2.3. Quantiﬁcation of the apparent fractional of loss from successive compression–expansion cycles Terminologically, the term hysteresis refers herein to the noncoincidence of a compression and its successive expansion. The impact of the compression–expansion cycles were quantiﬁed in terms of % of apparent loss deﬁned as follows:

ACi Apparent loss ð%Þ ¼ 100 1 AC1

ð1Þ

where AC1 is a value of molecular area of the ﬁrst compression isotherm and ACi the molecular area at a successive compression (second compression (C2), third compression (C3), etc.) for a similar surface pressure. The term apparent is added because the observation is purely phenomenological, independent of the origin of the changes of molecular area at speciﬁc surface pressures. We have veriﬁed the applicability of the equation using molecular areas extrapolated to zero surface pressure (A0) and for molecular areas at speciﬁc low surface pressures (2 and 4 mN/m). For the systems investigated, the results were similar, regardless the surface pressure of reference (0, 2 or 4 mN/m). The quantiﬁcation of hystereses using the differences in limiting molecular areas A0 obtained from the ﬁrst compression and the subsequent expansion have already been applied in the case of hysteresis of PS–PEO systems [3]. The quantiﬁcation of apparent losses in terms of Eq. (1) is advantageous in describing the effects of the various parameters such as the waiting time between a ﬁrst compression–expansion cycles and a second compression and allows for direct comparison between hom*opolymers of different Mn. The apparent loss of PEO has been determined for block copolymers as well. In this speciﬁc case, the % of loss corresponding to that of the PEO has been calculated using the surface areas of the PEO segment only. Brieﬂy, this surface area is obtained by subtracting the contribution of the PPO based on its surface concentration obtained from the isotherm of a PPO hom*opolymer at the surface pressure of interest. This is possible from the fact that both PPO and PEO segments behave independently at the air–water interface for the surface pressures of interest [2].

2.4. Convention in reporting subsequent compression–expansion cycles condition Henceforth, in order to streamline conditions for reporting and discussion, we shall use the following convention: Ci = compression, i = 1, 2, 3. 1 refers to ﬁrst compression, 2 to second compression, etc. Ei = expansion, i = 1, 2, 3. 1 refers to ﬁrst expansion, 2 to second expansion, etc. The sequences of waiting time will be reported as follows: tC1/E1–tE1/C2–tC2/E2 where tC1/E1 is the waiting time between C1 and E1, tE1/C2 the waiting time between the E1 and C2, and so on.

3. Results and discussion One of the underlying issues raised in this study is the problem of the relation between the surface pressure, the composition (amount adsorbed and segment distribution) of the monolayer and its equilibrium. In classical isotherms of spread monolayers, only two variables are usually considered, viz. the surface pressure p and the amount adsorbed at the interface, say C (or the corresponding surface molecular area A). Obviously these two variables are not enough to fully describe the equilibrium segment distribution, let alone under dynamic conditions. For extremely dilute (ideal) 2D monolayers of low Mw surfactants the Gibbs adsorption equation leads to the 2D equivalent of the 3D equation of state. Extensions to account for lateral interactions and excluded volume effects, (like the 2D Van der Waals equation, virial expansions, etc.,) can to some extent account for these deviation from ideality, but all of these are remote from polymeric adsorbates because the connectivity between segments is not accounted for. For the description of polymers attached to surfaces or at water–air interfaces, basically two ways are open, viz. statistical thermodynamics approaches, with the self-consistent ﬁeld (SCF) theories by Fleer et al. [17–19] and with the polymer scaling theory based on the de Gennes and Alexander-de Gennes approaches [20,21]. Both theories have their advantages and limitations. Statistical thermodynamics are based on a molecular picture. They require the formulation of a partition function which is maximized as a function of the segment density distribution u(z) where z is the distance from the surface. Several assumptions have to be made to keep the number of conﬁgurations within bounds. In the self-consistent Scheutjens–Fleer theory this is achieved by invoking a lattice model [18]. This is a restrictive assumption but has the advantage that molecular parameters like the segment size, the pair interaction energies and the solvent quality are explicitly accounted for. The Alexander–de Gennes theory is rather phenomenological. There is no need of a molecular picture and macroscopic power laws can be derived. In this respect, it is akin to thermodynamics. Scaling theory rather more easily describes dynamic features than SCF theory. Another advantage is its universal validity. A disadvantage is that testing of the various power laws requires data over several decades. For a ﬁrst interpretation of our observations it is conducive to exploit results from both approaches. SCF theory is explicit in that it recognizes segments in different lattice layers: the ﬁrst is the lattice layer adjacent to the air the next one is layer 2, etc. Trains, have all their segments in the proximal layer 1, loops start and end in layer 2 and tails start in layer 2 but end far from the ﬁrst layer [22]. The surface pressure is mainly determined by the trains, to a lesser extent by the segments in the second layer, even less in the third layer, etc. The decay with z is quadratic with a rate that depends on the quality of the solvent. The poorer the solvent, the slower the decay. This theory is only valid for low surface concentrations. The de Gennes theory does not discriminate

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

between subsequent layers, the extending segments in the subphase being described by a self-similar density proﬁle [23,24]. However, according to [25], de Gennes also concluded the dominance of the ﬁrst layer(s) in determining p and in [25] (their Fig. 1a) many experiments are given to conﬁrm this theoretical prediction. This collective information can act as a ﬁrst step of interpretation of our experiments, at least at low p. The next point is whether it is also allowed to invoke the above rules for out-of-equilibrium situations. As far as thermodynamics are concerned we may state that the question of the applicability of these rules and laws remains valid under non-equilibrium conditions depends on the Deborah number (De B tprocess/tmeasurement) [26]. For very high De the system is out of equilibrium but the change is so slow that it is not measured during the time of the measurement. For very low De the system is continually at equilibrium and no changes are observed. However, in the intermediate range the changes are measurable and this is exactly the range that we are observing in our compression–expansion cycles. Speciﬁcally, our Fig. 5 indicates relaxation processes in the minute range, which can be compatible with both the relatively slow detachment or expulsion of trains and with relatively slow polymer dynamics. Relatively fast processes like conformational changes in the last segments of tails and relatively slow processes like evaporation are not visible in the p(t) behaviour. 3.1. Typical isotherms of PEO hom*opolymers Typical p(C) isotherms of PEO hom*opolymers of different molecular weights are presented in Fig. 1. The Mns chosen cover a large range, including masses smaller and larger than that of the critical entanglement mass (Mc) of PEO (for PEO, Mc = 5.87 kg/mol [27]). This ﬁgure illustrates the increasing surface activity of PEO’s with increasing degree of polymerization. Monomeric ethylene oxide is hardly surface active, but polymers are. The surface pressure of the onset of the plateau is Mn-dependent. This is already known and has been reported previously [1,2]. It is an important fact to take into consideration in determining the parameters of hysteresis analysis. Up to 0.15 mg/m, the polymer is in a dilute regime (insert, Fig. 1). This is in agreement with Kato and Kawaguchi [28] who estimated the end of the dilute regime at C = 0.1 mg/m2. At surface concentrations 0.15 6 C 6 0.32 mg/m2, the p(C) isotherms are virtually superimposing. In terms of the de Gennes theory, the behaviour observed in that regime of concentration corresponds to that of a semidilute regime, described by a scaling-independency with regard to the molar mass of the

Fig. 1. p(C) isotherms of PEO hom*opolymers at 295 K. Mn in kg/mol indicated in the graph. Dashed lines: identiﬁcation of the surface pressures selected as p0 in the p(t) curves presented in Fig. 2.

497

polymer [2,21]. That also means that equilibrium was reached upon deposition of quantities under the experimental conditions used. Our value for the end of the semidilute regime is intermediate amongst those already reported for PEO (0.24 mg/m2 in Ref. [29] and 0.4 mg/m2 in Ref. [28]). Independently, experimental evidence from various techniques has been reported supporting that PEO Langmuir ﬁlms behave as almost purely bi-dimensional (monomer thick) for C < 0.4 mg/m2 and that segments of the PEO chains are transferred to the subphase only above that surface concentrations [30–34]. Upon compression beyond that critical surface concentration, segments of the polymer chains are expelled from the interface, creating loops and tails [35]. Per PEO molecule maximally two tails can develop. SCF theory predicts that loops contribute most to the adsorbed mass. The formation of loops could favour the development of entanglements which are expected to give rise to hysteresis upon expansion if the time of relaxation and reptation are slow. The point is that the thickening of the layer beyond that of the trains and short loops does hardly affect the surface pressure, hence the evolution towards a plateau. The equilibrium of PEO Langmuir ﬁlms was investigated under different conditions selected in order to isolate the effects of various parameters on the stability of the PEO chains at the air–water interface. A ﬁrst set of experiments were carried out at ﬁxed surface areas. It was followed by a more extensive examination of the effect of the variables on hysteresis from compression–expansion cycles. In the latter case, the variables included waiting times between compression–expansion cycles, molecular weight, limits in terms of surface concentration and temperature.

3.2. Evaluating stability at constant area The stability of PEO hom*opolymers of 262.5, 18.97 and 7.8 kg/mol has been investigated at constant area. The drop in surface pressure was monitored at p values corresponding to key surface concentration regimes of this polymer. The selected p correspond to the end of the semidilute regime (4.5 mN/m) and to a surface pressure slightly higher (5.5 mN/m), respectively. These surface pressures are indicated by the horizontal dotted lines in Fig. 1. This choice is based on reported evidences, according to which the transfer of EO segments into the subphase starts to take place beyond the end of the semidilute regime, leading to the formation of loops and tails [30,35,36]. This ﬁrst p could be thus considered as a transition point from one regime to another. At a surface pressure of 5.5 mN/m, PEO is in a concentrated regime in which p increases linearly as a function of A (or C). The estimation of the isothermal compressibility of the ﬁlm (jT ¼ ð1=AÞð@A=@ pÞT ¼ ð@lnA=@ pÞT , data not shown) conﬁrmed that the lowest compressibility is observed in the range of 4.5– 6.5 mN/m, with a value of 0.092 (mN/m)1. It corresponds to the inverse of the static elastic modulus es, which is the dilational modulus ~eK without any dissipative contribution [37]. Our experimental value of es (11 mN/m) for the concentrated regime at 295 K is similar to the results of Akentiev and Noskov [35]. For a PEO of 100 kg/mole these authors observed a maximal surface elasticity 12 mN/m in the same range of surface pressure at 293 K. The surface pressure and concentration at which the concentrated regime ends is Mn-dependent. When the chosen surface pressures were reached, the compression was stopped, the surface area kept constant and the surface pressure recorded as a function of time. Typical results are presented in Fig. 2. The results of triplicates indicate that the pressure drop was not signiﬁcant over a period of 30 min. Therefore, up to 5.5 mN/m at room temperature, the PEO chains can be considered in relative equilibrium with the subphase within the considered period of time.

498

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

relaxation times, the ﬁrst one involving segments transport of the proximal layer. These experiments carried out at constant areas allowed for identifying some Mn-dependent instability at pcollapse and for pointing out the capability of the monolayer to re-equilibrate over time upon overload. However, they are limited in supplying insight into the reversibility of the conformational changes occurring in the polymer chain and in quantifying the proportions of the chains involved in these processes. For that, compression–expansion cycles and their hystereses are more powerful. 3.3. Compression–expansion experiments

Fig. 2. p(t) curves at 295 K for PEO hom*opolymers of different Mn. For each polymer, the surface pressure p at t0 is indicated on the isotherms of Fig. 1 (dashed lines). The values of Mn (kg/mol) are indicated for each curve in the graph.

A third key surface pressure for PEO monolayers it that of the collapse, for which a plateau of saturation is observed (Fig. 1). In our previous work, it was already observed that for low molecular weight PEO hom*opolymers there is a signiﬁcant drop of p as a function of time upon maintaining the surface area corresponding to the onset of the plateau for low molecular weight chains [2]. There also is some inﬂuence of the chain length: after 30 min, the observed average drops of pressure were 0.03 ± 0.01, 0.18 ± 0.02 and 0.29 ± 0.03 mN/m for PEO of 262.5, 18.97 and 7.8 kg/mol, respectively. In typical compression isotherms of PEO, beyond the plateau, the p(C) proﬁle is linear, showing a very slight increase of surface pressure with increasing surface concentration (Fig. 1 and Fig. 3A). In the present study, the stability of the monolayer of the PEO chain of 262.5 kg/mol has been challenged at a ﬁxed surface area over a longer period of time (19 h) for an apparent surface concentration beyond those achievable in usual isotherm experiments. For these experiments (Fig. 3B), large quantities of molecules were deposited onto the interface. Before compression, the surface concentration was already high (2 mg/m2). The initial surface pressure was similar to that observed for C = 2 mg/m2 in Fig. 3A. Then, the monolayer was compressed to an apparent surface concentration ten times larger (Fig. 3B). The compression increased the surface pressure to 10.3 mN/m. This value is signiﬁcantly higher than that of pcollapse. After compression and keeping the attained surface area constant, it was observed that the surface pressure remained unchanged for 1 h, than dropped slowly to gradually stabilize at 10.0 mN/m. It is worth mentioning that the drop of pressure observed is much larger than can be accounted for by the effect of water evaporation which was estimated to be less than 0.1 mN/m over a period of 24 h at room temperature with our set-up. This experiment illustrates nicely our point that the ﬁrst lattice layer(s) dominate(s) the surface pressure. In the plateaus of the isotherms of Fig. 1 and Fig. 3A that layer is virtually saturated. Only by extraordinary means can an increase of about 0.3 mN/m be reached for PEO hom*opolymer Langmuir ﬁlms. However, that situation is not at equilibrium; it slowly relaxes to the ordinary values of pcollapse, conﬁrming the surface pressure of the plateau is the maximal surface pressure of equilibrium for the PEO hom*opolymer. An interesting question is what the reason is for the circa 1 h time lag before the surface pressure drops and returns to its equilibrium value. The question deserves further investigation. We suggest a succession of at least two processes of different

3.3.1. Effect of maximal surface pressure Very scarce information remains available in the literature concerning hysteresis of compression of PEO hom*opolymers. Hysteresis has been reported for a PEO (2 kg/mol) for a target p of 2 mN/m (pcollapse 4 mN/m) [38]. However the chain is very short (almost an oligomer) and its Langmuir ﬁlm reportedly instable. For longer chains, as a starting point, there are recent data reported for a PEO hom*opolymer of 52.85 kg/mol. The hysteresis was small [2] for a waiting time sequence of 0–2, that is to say 0 min up (between C1 and E1) and 2 min down (between E1 and C2). In that particular case, Cmax was ﬁxed at 0.70 mg/m2 (almost at plateau (Ccollapse = 0.78 mg/m2)). Applying Eq. (1) on the data reported in [2] (procedure described in Section 2.3), the apparent loss can be quantiﬁed using the values of the ﬁrst and the second compression (C1 and C2, respectively). A value of 4% of loss is obtained, indicating minor and/or reversible changes before the plateau at this time scale. These results support those obtained at constant area (Fig. 2) and suggest that for p 6 pcollape, the Langmuir ﬁlm of PEO remains close to equilibrium. Either very few segments are transferred into the subphase, or their mobility and transition from loops back to trains is relatively fast (low De). In the present work, more attention has been given to the plateau regime. A ﬁrst set of experiments have been carried out with a PEO hom*opolymer of a higher Mn to ensure a strong adsorption efﬁciency. For PEO with molecular weights P 85 kg/mol, pcollape is not Mn-dependent and the saturation is maximized [1]. The waiting time sequence was ﬁxed at 15–15, (15 min between C1 and E1 and 15 min between E1 and C2). The only variable was the surface area achievable upon the ﬁrst compression, which was determined by a target surface pressure. In contrast to the observation in the lower concentration regimes, compression at C P Ccollapse results in substantial hysteresis for the ﬁrst cycle (Fig. 4). In that regime, compressing at higher apparent surface concentrations leads to a major increase of the hysteresis (difference between C1 and E1). For Cmax P 0.90 mg/m2, there is almost no difference between E1 and C2 and E2 and C3. This piece of information could suggest that if hystereses are provoked by entanglements, a maximal amount of entanglements have been attained at such surface concentrations. However, at that point, it is not possible to decipher between entanglements preventing re-adsorption or effective loss of PEO chains. The completion of a second cycle resulted in additional losses. 3.3.2. Effect of waiting time The inﬂuence of the waiting time has been examined for a PEO hom*opolymer of 99.5 kg/mol containing a number of monomers close to that of PEO of the longest PPO–PEO block copolymer selected in this study (details in Table 1). The maximal limit of compression was ﬁxed in terms of surface concentration. Over all the experiments, the variation of the target Cmax was 1.40 ± 0.06 mg/m2. Such apparent surface concentration brings the system far beyond the onset of the collapse (Fig. 5A). It can be seen that applying a ﬁxed waiting time up of 15 min between

499

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

9.9 mN/m

π monitored at constant area

compression

Deposition of 150 μl

Deposition of 15 μl π -Γ isotherm

Γ0 = 2.1 mg/m2 isotherm Γapparent after compression = 19.45 mg/m2

Fig. 3. p(C) isotherm of PEO 262.5 kg/mol at 295 K (A) and p(t) curves obtained from large deposits of the same PEO hom*opolymer (blue, 0.5 h; green, 19 h) (B). (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

Fig. 4. p(A) compression–expansion cycles for PEO 845 kg/mol at 298 K. The apparent surface concentrations reached upon compression are indicated in the insert (mg/m2).

the ﬁrst compression (C1) and subsequent expansion (E), and varying the pause down from 15 to 125 min between the E and C2 has only a minor effect on the molecular areas of C2, which are switched to identical (within the given boundaries) lower values

regardless the waiting time down. Within the experimental variation, similar apparent losses of 22 ± 3% were obtained between C1 and C2 under these conditions. A priori, there are two options for the underlying process: (1) the formation of entanglements of looping segments limiting re-adsorption and/or (2) the loss of molecules into the subphase. Assuming that the surface pressure is mainly dependent on the surface concentration of the train segments, the fact that higher molecular areas are obtained for C2 in comparison to those of E1 indicates that compressed segments can return to the interface upon relaxation. Therefore, the option of entanglements cannot be excluded, neither a combination of both options. Since the isotherm of C2 is not dependent on the waiting time (down) between E and C2, the re-adsorption is apparently completed during the expansion and this process is relatively rapid (minutes). Applying a longer waiting time up between C1 and E (120 min instead of 15 min) provokes a more pronounced hysteresis in the isotherms for which the apparent loss of molecules for C1–C2 is now 52%. We infer as an indication that a waiting time of 15 min is not sufﬁcient for chains to relax completely in the densely packed plateau regime. An additional interesting piece of information is that for the extensive compression of the PEO monolayer far beyond the surface pressure of the plateau, after a waiting time of 120 min at a

loss = 22% loss = 52%

C1n C2n

Fig. 5. (A) Effect of time of waiting time after compression and expansion of the p(C) isotherms of PEO 99.5 kg/mol at 295 K. The two numbers in the insert are the waiting times up and down, respectively. In this speciﬁc case, the compression rate was 10 mm/min. (B) Isotherm with a waiting time sequence of 120–15 for the original isotherm (light blue) and the same isotherm normalized by taking into account the apparent loss (green). (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

500

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

high surface concentration, no difference was observed between E1 and C2. A normalized isotherm was calculated taking into account the apparent loss between C1 and C2 (Fig. 5B). E and C2 of the normalized isotherm (labelled En and C2n in Fig. 5B) perfectly superimposed C1 of the original isotherm. It implies that up to the plateau, the increase in surface pressure upon compression depends only on the surface concentration in the proximal layer regardless the number of entanglements formed during the previous compression. Our overall impression is that, with regard to changes in the surface pressure, two processes with different timescales are distinguishable. The shorter one has a time scale of a few minutes and might be related to processes taking place inside or close to the ﬁrst layer. The long-time process, with a scale of hours, must be attributed to layers beyond that and involve slow steps like disentanglements and/or reptation. Regarding disentanglements, a complete relaxation of chains cannot be expected in within the time of experiments at least not for high Mn. The time of relaxation of a polymer in 2D is predicted to scales proportionally to N3log N [39]. A priori the observed phenomena cannot be explained by complete disentangling of the polymer chains upon expansion since for long chains the process is expected to take days. However, this information does not sufﬁce to accurately predict the relaxation time of 2D monolayers. We refrain from generalizing these results. Generally speaking the literature points to controversies regarding the formation of entanglement in 2D. For instance, the results of Sato et al. on shear viscosity of Langmuir ﬁlms of poly(vinyl octanal acetal) (PVO) suggest that little entanglement is present in the spread monolayer [40]. More recently, these conclusions have been supported by the work of Srivastava and Basu who ruled out the option of entanglements and reptation in dense layers of conﬁned polymers [41]. Their conclusions are in contradiction with the ﬁndings of Maestro et al. who found evidences of entanglements and reptation in the case of poly(ter-butyl-acrylate) (PTBA) monolayers [25]. It is also worth pointing out that some evidences have been reported indicating that the critical mass of entanglement (Mc) in conﬁned polymers is higher than in the bulk [42]. 3.3.3. Effects of molecular weight and temperature The effect of molecular weight was examined by comparing compression–expansion cycles of PEO hom*opolymers of 11.26 and 845 kg/mol which present signiﬁcantly different pcollapse (Fig. 1). In both cases, compression–expansion cycles have been

ﬁrst carried out at 298 K with a maximal limit of compression ﬁxed close to the onset of the plateau. The same experiment has been carried out at 283 K (Fig. 6A). In the latter case, the surface concentrations reached values far beyond the plateau. This inﬂuence of temperature was expected from the positive dpcollapse/dT value for PEO [1,2]. This also brought the system at surface concentrations far beyond the plateau. Regarding the effect of compression–expansion cycles, comparable general trends are observed for both Mn: the hysteresis and apparent loss are strongly temperature, and consequently, C-dependent in agreement with the results of Fig. 4. The effect of Mn is observable in all conditions. In the case of the shortest chain, few entanglements are expected, the critical mass of entanglements (Mc) for PEO being 5.87 kg/mol in the bulk [27]. The trend for shifting to higher pcollapse upon increasing the temperature means that in this direction the subphase the PEO segments are less prone to move into the subphase because it becomes a poorer solvent for the PEO. The poorer solvency of the subphase for PEO is also reﬂected in the material losses upon cycles. Using Eq. (1) we computed these (Fig. 6B). An increase of temperature from 283 to 298 K reduced the losses by 60% and 50% for the lower and the higher Mn, respectively. From this basis, it appears that (1) the number of entanglements (Mn-dependent) plays a major role in the apparent loss at a ﬁxed temperature and (2) the effect of temperature on the stability of the ﬁlm at the air–water interface does not follow a strong Mn-dependency, if any. It is also remarkable that for PEO 845 kg/mol, the curves of E1 and C2 and those of E2 and C3 are not superimposing at 283 K as it was observed at 298 K (Fig. 4). This is additional evidence supporting the fact that an entropically-driven reorganization takes place in the plateau, even if the surface pressure is constant. For the PEO of higher Mn, the effect of temperature was examined in the plateau regime over a larger range of temperatures (Fig. 7). The interest of that ﬁgure is showing that for C P 1 mg/m2, the losses for C1–C2 and C1–C3 are constant regardless the apparent surface concentration attained in the plateau for 283 K 6 T 6 293 K. The system appears to be frozen. The situation changes at 298 K. At that temperature there is a dramatic increase of losses which are almost similar for C1–C2 and C1–C3. This trend is interpreted by a switch in the interplay of inﬂuencing factors. It is suggested that at 298 K, the impact of the reduction of solubility is overcome by an increase of kinetic, temperature-dependent parameters promoting diffusion, reptation and other entropic processes.

Fig. 6. (A) C(p) isotherms of compression–relaxation cycles of PEO 11.26 and 845 kg/mol at 283 and 298 K. pmax were ﬁxed at 8.8 and 9.9 mN/m for PEO 11.26 and PEO 845 kg/mol, respectively and the waiting time sequence was 15–15–15–15. (B) Corresponding apparent losses for C1–C2 (blue) and C1–C3 (green). (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

Fig. 7. Apparent loss (%) of PEO 845 kg/mol as a function of Cmax reached upon the ﬁrst compression at different temperatures after two complete compression– expansion cycles followed by a third compression C3. The results for C1–C2 and C1– C3 are identiﬁed by full and empty symbols, respectively. Colors were used to distinguish between the temperatures. The waiting time sequence was 15–15–15– 15 in all cases. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

These results illustrate that the effect of temperature is multifold. Temperature affects the solubility of PEO in water and at the interface. A static effect is that in the range of temperature of interest, the solubility in water decreases with increasing temperature [43]. This effect can explain the positive dpcollapse/dT factor. For the same reason, it impedes the deﬁnition of the applicability limits for compression–expansion experiments because it redeﬁnes both pcollapse and surface concentrations of transition, involving both a static and a dynamic phenomenon. The dynamic aspect is that the impact of hysteresis is more pronounced at low than at high temperature. As all processes are being somewhat slower at 283 K it is expected that more time is needed to reach equilibrium. However, in the present case the fact that at lower temperature the proximate layer is more compressed may also play a role because loss of polymer from this region may also contribute. At any rate, for further study is appears relevant to distinguish between static and dynamic contributions.

501

Fig. 8. Typical p(C) isotherms of PPO95–PEO409 and (B) PPO353–PEO2295 block copolymers (blue and green curves, respectively). (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

by EO segments. This trend has been conﬁrmed in [2] where it was found to become more prominent at higher surface pressures. Now the intention is to address the dynamics of the PEO chain in such a system. Typical isotherms of the selected PPO–PEO diblock copolymers are presented in Fig. 8. In order to be able to judge the relevance of our system to examine the dynamics of PEO moiety, tests have been carried out to evaluate the stability of PPO at air–water interfaces. In a recent study, we have demonstrated that the adsorption of PPO is not affected by temperature in the range of 283–298 K [2]. However, as compared to PEO compression of PPO results in higher collapse surface pressures. In terms of SCF theory this is a direct consequence of the higher Gibbs adsorption energies per segment. The stronger bonding of trains must also make PPO monolayers less sensitive to hysteresis. To check this expectation, hysteresis experiments of a PPO hom*opolymer of N = 259 (15 kg/mol) have been carried out ﬁxing pmax at 15 and 30 mN/m (Fig. 9), which are above the surface pressures that can be attained with PEO. As expected, the hysteresis is very small for pmax of 15 mN/m. However, substantial hysteresis was observed from ﬁxing pmax beyond that of the collapse of the monolayer, even if the waiting time at pmax was 0 min. The loss of molecules at the air

4. PPO–PEO diblock copolymers The stronger surface activity of PPO as compared to PEO, gives rise to the option of studying in more details the behaviour and the stability of PEO at surface pressures corresponding to the plateau by using PPO–PEO diblock copolymers. In those systems the PPO block acts as a stable anchor (at least up to 15 mN/m), preventing the loss of PEO chains into the subphase. The obvious advantage is that we have now in hand a system allowing for isolating the contribution of entanglements and conformational change on the compression hysteresis in PEO monolayers. The thermodynamics of this material have already been described in [1]. PPO is water insoluble and also known to be surface active. The surface pressure of the semidilute regime of the PPO hom*opolymer is 11 mN/m (determination based on the end of linearity of the log p(log C) curves in [2]). In the PPO–PEO system, both blocks spread and are in good solvent conditions at the air– water interface. The PEO block remains anchored at the air–water interface upon compression by the hydrophobic PPO moiety. With increasing surface concentration of the spread monolayer, the expectation is that the train layer will be enriched in PPO whereas the second and further lattice layers are more likely to be enriched

Fig. 9. C(p) isotherms of compression–relaxation cycles of PPO 15 kg/mol at 295 K for one complete compression–relaxation cycle followed by a second compression. Waiting times sequence of 0–15 and pmax of 15 and 30 mN/m (full and dashed curve, respectively).

502

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

Fig. 10. Compression–expansion curves for 3 cycles applied on (A) PPO95–PEO409 and (B) PPO353–PEO2295 with pmax ﬁxed at 6 mN/m a waiting time sequence of 15–15–15–15 obtained at 283 and 298 K (black and red curves, respectively). (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

Fig. 11. (A) Typical compression–expansion cycles for PPO353–PEO2295, pmax = 10.5 mN/m, waiting time of 15 min up and down. Similar volumes of 20 lL of a solution of 1 mg/ml were used for the deposition. For clarity, the results at only three temperatures are presented. (B) Surface concentration obtained after compressing PPO353–PEO2295 ﬁlms at 6 and 10.5 mN/m (green triangles and blue squares, respectively). Insert: zoom to better show the inﬂuence of temperature on C at 6 mN/m. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

water-interface was estimated to be 3% and 23% for pmax of 15 and 30 mN/m, respectively. A ﬁrst set of results for the examination of the dynamics of PPO–PEO at the air–water interface is presented in Fig. 10 for the two diblock copolymers selected for a maximal compression ﬁxed at 6 mN/m and 3 successive compression–expansion cycles. There is an obvious Mn-dependent effect of temperature on the isotherms. For similar values of concentrations, the longest polymer adsorbs at lower surface pressures with increasing temperature. The apparent loss between C1 and C2 and between C1 and C3 were estimated for both polymers at various temperatures at 6 mN/m. As for the PEO hom*opolymers, at a surface pressure below the plateau, these losses remain minor. Apparent losses 0.4 mg/m2, the inﬂuence of the temperature goes in a reverse direction as compared with surface concentrations below 0.4 mg/m2. Between 288 and 303 K, a linear decrease in the surface concentrations attained is observed when the limit of surface pressure was ﬁxed at

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

503

Fig. 12. A. Successive compression–expansion cycles for PPO353–PEO2295 at 283 K. Waiting time 15 min up and down. B. Original p(A) compression–expansion curves presented in A (red) supplemented with curves recalculated taking into account the apparent weight loss as estimated after two complete compression–expansion cycles. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

10.5 mN/m with an exception for the values obtained at 283 K for pmax = 10.5 mN/m which are signiﬁcantly higher. However, in the latter case, the PEO chains are compressed beyond the end of the plateau (Fig. 11A). This abrupt increase in surface concentration beyond the end of the plateau is due to the change in the PPO/ PEO proportion at the interface. The question that immediately comes to mind concerns the impact of the formation of such brushes on the hysteresis and the dynamics of the PEO anchored chains. This issue is considered in Fig. 12. In this illustration, the isotherms of three successive compression–expansion cycles are shown into details for the lowest temperature (283 K). As previously mentioned, at this temperature, a compression at a ﬁxed pmax of 10.5 mN/m results in compressing the monolayer of the block copolymer beyond the linear PEO plateau (Fig. 12A). In Fig. 12B, the isotherms presented in Fig. 12A have been recalculated taking into account the apparent losses estimated applying Eq. (1) (in this particular case, the apparent loss on the third compression is 28%). It allows us to demonstrate that the shape of the isotherm remains unchanged up to a surface pressure of 8.5 mN/m, which is the same as what was observed for the PEO hom*opolymer of 99.5 kg/mol (Fig. 5B). These results show that the densely entangled layer formed upon successive compression–expansion cycles behaves exactly the same as the originally spread ﬁlm. These results support the insight that up to 8.5 mN/m only the train segments are involved in determining the increase of surface pressure. Beyond that surface concentration, the compression of the loops present in the subphase inﬂuences to the increase of the surface pressure of the Langmuir ﬁlm. The slope of the plateau increases upon successive compression. However, its length (DA) is decreasing from C1, C2 and C3, indicating that the number of EO monomers in train conformation is decreasing accordingly. Apparent losses of PEO for the block copolymer PPO353–PEO2295 and those of PEO 845 kg/mole are presented in Fig. 13. This ﬁgure includes the values obtained at all temperatures for both molecules except the ones obtained for hom*opolymer at 298 K at very high surface concentration (loss of almost 60–80%, Fig. 7). Otherwise, in all other cases, the onset of saturation in the losses coincides with the surface concentrations at the onset of the plateau and the losses are similar regardless is the chain is anchored or not. Such high losses have not been obtained for the copolymer. Experiments for PPO353–PEO2295 with a waiting time of 2 h up resulted in the same values observed upon the application of a second compression (data not shown). The apparent losses were 27% at

Fig. 13. Apparent loss of PEO upon successive compressions for PPO353–PEO2295 (circles) and PEO hom*opolymer 845 kg/mole (crosses) as a function of Cmax. Losses were estimated for C1–C2 and C1–C3 (blue and green symbols, respectively). The lines were only added as guides for the eye. Waiting time sequence: 15–15–15–15. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

293 K and 31% at 298 K. These ﬁndings support that for the hom*opolymer, apparent losses higher than 30% could be associated with effective dissolution and loss of complete chains into the subphase. This is an important observation because it infers that in the course of a single compression, most of the apparent losses are due to entanglements. It is remarkable that the apparent losses for the PEO hom*opolymer upon successive compression–relaxation cycles are similar to those obtained for the block copolymer. These experimental results are strong evidence that under the applied conditions, most of the PEO chains of the hom*opolymer are still all present at the air–water interface. The plateau of saturation of apparent loss strongly suggests that from pcollapse, the regime of concentration of the EO segments is changing. A schematic summary of the main ﬁndings for PEO is presented in Fig. 14. 5. Conclusions The purpose of the present study was to obtain insight into static and dynamic properties of spread (Langmuir) monolayers of poly(ethylene oxide), PEO, hom*opolymers. Notwithstanding the fact that PEO is a familiar polymer, that also is scientiﬁcally

504

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

Transfer of segments

22%

10%

4% 0.32 mg/m²

Slower concentraon of train coverage Densiﬁcaon of the loops, increase of entanglements

Purely 2D behavior

Mobility restricons and metastability

Stabilty

Entropic processes become more important from 298 K

Major increase of train coverage

Time-dependent reorganizaon associated with relaxaon of submerged segments (hours)

Fig. 14. Transitions and dynamics in PEO Langmuir ﬁlms at a ﬁrst compression.

interesting, only a limited number of publications have so far been devoted to the stability of its thin ﬁlms at ﬂuid interfaces. As a ﬁrst aim we investigated such monolayers more systematically, mostly through surface pressure measurements in a Langmuir trough varying the molecular mass and the temperature. In addition, experiments have been carried out for diblock copolymers with PPO, which acts as an anchor. Considering that for polymeric monolayers time effects play an essential role, we also systematically studied this parameter, mostly by allowing various waiting times between compression and expansion, and expansion and recompression. The abundance of graphs in the paper illustrates that the behaviour of deposited layers of PEO and PEO–PPO block copolymers is a many-sided topic, determined by a variety of variables all with their auxiliary conditions. The phenomena can be roughly divided into two classes, static and dynamic ones. The former category applies to the properties of layers at equilibrium: surface pressure, surface concentration, density distribution of segments over the layer, etc. The latter involves the extent and rate of response of the systems to changes in condition. Often this category involves the time, but important issues include the changes in the layers upon changes in pressure, in particular the possibility of molecules to disappear from the layer upon strong and long compression. Our experimental parameter is the surface pressure p, which is a useful, but not omnipotent variable. Some phenomena are under some conditions directly reﬂected in p, others only indirectly or not at all. Moreover, often it is not possible to isolate the action of external variables; for example, the temperature does not only rates because its effect on viscosity; it also inﬂuences the solubility of PEO in the bulk and at the surface, probably in different ways. This multifaceted behaviour guarantees experimental and interpretational challenges and we would not be surprised if our data would be further extended by coming studies. Fig. 14 can serve as a graphic to overview the various ranges of properties we have found. Brieﬂy, several regimes can be distinguished, which varied in terms of static and kinetic factors. It has been found that the ﬁlm is in a dilute regime up to about 0.2 mg/m2 because below such a surface concentration p(A) is Mn-dependent. Upon compression, it is followed by a Mn-independency observed for 0.2 6 C 6 0.32 mg/m2. Such surface concentration for the end of the semidilute regime is in agreement with previous studies on the rheology of PEO spread ﬁlms [28]. Up to that point, the system is stable and reversible upon compression and re-expansion. For

0.32 6 C 6 0.7 mg/m2, only minor hystereses develop and the apparent losses remain modest (610%). In that range of surface concentration, it appeared that loops are formed in the water subphase, but remain relatively loose with few entanglements. Further compression induces the transfer of additional PEO segments into the subphase (12%) to reach a value of 22% at the onset of the plateau. In the regime just preceding the plateau, the surface pressure increase is very small although the apparent loss is major. It is associated to a minor change in the train surface coverage, but to a substantial reorganization and densiﬁcation of the loops. In the plateau regime, the loss is higher and constant (22%) for 1 6 Cmax 6 2 mg/m2 in maintaining the achieved surface area for 15 min. Similar losses were obtained for PEO hom*opolymers of high Mn and PPO353–PEO2295. It strongly suggests that the hystereses should be mainly accounted for by entanglements and chain reorganization at high surface pressure. It also infers that under usual isotherm experiments, PEO remains anchored in a metastable state at the air–water interface at surface concentration well above the onset of the plateau. Additional losses are incurred for PEO hom*opolymers for monolayers kept compressed in the plateau for 2 h or the temperature raised to 298 K in the hom*opolymer. The maximal level of entanglement obtained in the block copolymer was 30%. For the hom*opolymer, apparent losses higher than this amount could be interpreted in term of effective dissolution and loss of whole chains into the subphase. In both PEO hom*opolymers and PPO–PEO diblock copolymer the maximum amount of entanglements was obtained from a second compression at surface concentration beyond the plateau or by applying a longer time of pause between the ﬁrst compression and the subsequent expansion. Overall, the surface pressure of the plateau appears to be a real surface pressure of equilibrium. However, this information does not sufﬁce to establish the equilibrium as kinetically-driven reorganization takes place in the looping segments. Because of the densiﬁcation of the monolayer in the plateau, this reorganization is slow. Evidence have also been found indicating a keyentropic changes in PEO behaviour at the air–water interface with a toggling point at surface concentration of 0.4 mg/m2. This intriguing ﬁnding will be explored into more details in a next paper dedicated to the entropy changes in PEO Langmuir ﬁlms. For the theoretical interpretation of the dependencies it is a ﬁrst prerequisite to establish whether the system is at equilibrium. Our studies helped to distinguish between three categories: systems in absolute equilibrium, systems in a metastable state and systems

L. Deschênes et al. / Journal of Colloid and Interface Science 449 (2015) 494–505

that were not at all equilibrated. We showed that these distinctions are rather differences of the Deborah number (De). Under various conditions relaxation phenomena were found that consisted of two processes with different De. When these two numbers differ markedly it is perfectly possible to consider the situation where the faster relaxing process is completed but the next one is slow to come as metastable. On such monolayers thermodynamics may be applied. In fact, this conclusion validates a posteriori our application of the Clapeyron equation in 2D [2] and hence also validates the obtained value for the enthalpy of adsorption of a PEO segment. As to the interpretation in terms of physicochemical model theories, we have not done that systematically but have given many ad hoc suggestions, either stemming from SCF [17,18] or from scaling theories [21,24]. The trend is that the former is suitable in the ﬁrst parts of the isotherm, where the surface pressure is dominated by the train layer, less so in the adjoining layers and not at all by segments remote from the surface. The theory is also explicit that it predicts a quadratic solvent quality-dependent decay with distance from the surface. At higher surface concentrations the validity of this model becomes questionable and recourse must be sought in scaling laws describing compact quasi-3D polymer layers. The results of the present study support that in the plateau regime there is reorganization in the adsorbed PEO chain, evolving towards a redistribution of segments in a self-similar layer (SSL). The SSL model in PEO monolayers is supported by evidence from surface dynamic elasticity measurements [35]. As compared to the surface rheology measurements, our approach has the advantage of an easy way in quantifying the number of segments involved in such reorganization. Acknowledgments The authors are grateful to Claude Danis for technical assistance and his contribution in carrying out several of the isotherms included in the present study. This work was supported by Agriculture and Agri-Food Canada, Grant #1480. References [1] D.J. Kuzmenka, S. Granick, Macromolecules 21 (1988) 779. [2] L. Deschênes, J. Lykelma, C. Danis, F. Saint-Germain, Advances in Colloid and Interface Science, in press, http://dx.doi.org/10.1016/j.cis.2014.11.002. [3] A.M. Gonçalves Da Silva, E.J.M. Filipe, J.M.R. D’Oliveira, J.M.G. Martinho, Langmuir 12 (1996) 6547. [4] M.C. Fauré, P. Bassereau, M.A. Carignano, I. Szleifer, Y. Gallot, D. Andelman, Eur. Phys. J. B 3 (1998) 365.

505

[5] B. Rippner Blomqvist, T. Wärnheim, P.M. Claesson, Langmuir 21 (2005) 6373. [6] X. Wang, G. Wen, C. Huang, Z. Wang, Y. Shi, RSC Adv. 4 (2014) 49219. [7] A.M. Goncalves da Silva, A.L.S. Gamboa, J.M.G. Martinho, Langmuir 14 (1998) 5327. [8] H.D. Bijsterbosch, V.O. De Haan, A.W. De Graaf, M. Mellema, F.A.M. Leermakers, M.A. Cohen Stuart, A.A. Van Well, Langmuir 11 (1995) 4467. [9] J.K. Cox, K. Yu, B. Constantine, A. Eisenberg, R.B. Lennox, Langmuir 15 (1999) 7714. [10] L. Deschenes, M. Bousmina, A.M. Ritcey, Langmuir 24 (2008) 3699. [11] E.S. Johonnott Jun, Philos. Mag. Series 6 (11) (1906) 746. [12] J. Perrin, Ann. Phys. X (1918) 160. [13] H.G. Bruil, J. Lyklema, Nature 233 (1971) 19. [14] G.N. Sethumadhavan, A. Nikolov, D. Wasan, Langmuir 17 (2001) 2059. [15] L. Ter-Minassian-Saraga, C.o.C.a.S.C.I. Catalysis, Pure Appl. Chem. 57 (1985) 621–632. [16] H.J. Trurnit, J. Colloid Sci. 15 (1960) 1. [17] G.J. Fleer, M.A. Cohen Stuart, J.M.H.M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman et Hall, London, 1993. [18] J.M.H.M. Scheutjens, G.J. Fleer, J. Phys. Chem. 83 (1979) 1619. [19] G.J. Fleer, F.A.M. Leermakers, Curr. Opin. Colloid Interface Sci. 2 (1997) 308. [20] S. Alexander, J. Phys. (Paris) 38 (1977) 983. [21] P.G. de Gennes, Scaling Concepts in Polymer Physics, Cornell University Press, Ithaca, NY, 1979. [22] J.M.H.M. Scheutjens, G.J. Fleer, J. Phys. Chem. 84 (1980) 178. [23] M. Aubouy, O. Guiselin, E. Raphaël, Macromolecules 29 (1996) 7261. [24] P.-G. De Gennes, C.R. Hebd Seances Acad. Sci. Ser. II 294 (1982) 1317. [25] A. Maestro, H.M. Hilles, F. Ortega, R.G. Rubio, D. Langevin, F. Monroy, Soft Matter 6 (2010) 4407. [26] M. Reiner, Phys. Today 17 (1964) 62. [27] L.J. Fetters, D.J. Lohse, R.H. Colby, in: J.E. Mark (Ed.), Physical Properties of Polymers Handbook, second ed., Springer, New York, 2007, p. 447. [28] S. Kato, M. Kawaguchi, J. Colloid Interface Sci. 384 (2012) 87. [29] D. Langevin, F. Monroy, Curr. Opin. Colloid Interface Sci. 15 (2010) 283. [30] S.K. Gissing, R.W. Richards, B.R. Rochford, Colloids Surf. A 86 (1994) 171. [31] J.A. Henderson, J. Penfold, R.K. Thomas, J.R. Lu, Macromolecules 26 (1993) 4591. [32] M. Kawaguchi, M. Tohyama, Y. Mutoh, A. Takahashi, Langmuir 4 (1988) 407. [33] J.R. Lu, T.J. Su, R.K. Thomas, J. Penfold, R.W. Richards, Polymer 37 (1996) 109. [34] B.B. Sauer, H. Yu, M. Yazdanian, G. Zograﬁ, W.K. Mahn, Macromolecules 22 (1989) 2332. [35] A.V. Akentiev, B.A. Noskov, Colloid J. 64 (2002) 129. [36] S.K. Peace, R.W. Richards, The man, in: B.J. Gabrys (Ed.), Applications of Neutrons Scattering for Soft Condensed Matter, Gordon and Breach Science Publishers, Amsterdam, 2000, pp. 163–204 (Chapter 6). [37] F. Monroy, F. Ortega, R.G. Rubio, Plasmas, Fluids, Related Interdiscipl. Top. 58 (1998) 7629. [38] T.J. Joncheray, K.M. Denoncourt, C. Mathieu, M.A.R. Meieir, U.S. Schubert, R.S. Duran, Langmuir 22 (2006) 9264. [39] P.G. de Gennes, in: J. Charvolin, J.F. Joanny, J. Zinn-Justin, (Eds.), Motion of Polymer Near a Solid Surface, in Liquids at Interfaces, Elsevier, Les Houches, 1988, p. 311. [40] N. Sato, S. Ito, M. Yamamoto, Macromolecules 31 (1998) 2673. [41] S. Srivastava, J.K. Basu, J. Chem. Phys. 130 (2009). [42] H. Meyer, T. Kreer, A. Cavallo, J.P. Wittmer, J. Baschnagel, Eur. Phys. J.: Special Top. 141 (2007) 167. [43] C.L. Ren, R.J. Nap, I. Szleifer, J. Phys. Chem. B 112 (2008) 16238.

Langmuir monolayers of non-ionic polymers: equilibrium or metastability? Case study of PEO and its PPO-PEO diblock copolymers. - PDF Download Free (2024)

References