Toward an understanding of aqueous biphasic formation in polymer–polymer aqueous systems

Toward an understanding of aqueous biphasic formation in polymer–polymer aqueous systems

Polymer 83 (2016) 1e11 Contents lists available at ScienceDirect Polymer journal homepage: www.elsevier.com/locate/polymer Toward an understanding ...

2MB Sizes 0 Downloads 61 Views

Polymer 83 (2016) 1e11

Contents lists available at ScienceDirect

Polymer journal homepage: www.elsevier.com/locate/polymer

Toward an understanding of aqueous biphasic formation in polymerepolymer aqueous systems Rahmat Sadeghi*, Moora Maali Department of Chemistry, University of Kurdistan, Sanandaj, Iran

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 September 2015 Received in revised form 14 November 2015 Accepted 14 November 2015 Available online 1 December 2015

Aiming at gathering further evidence about the molecular-level mechanisms of the polymerepolymer aqueous biphasic systems (ABSs), systematic studies on the vaporeliquid and liquideliquid equilibria of several ternary aqueous solutions of two water soluble polymers capable or not of inducing phase separation were performed at 298.15 K. It was found that the hydrophilicity of the investigated polymers follows the order PEGs z PEGDMEs > PVP >> PPG400 and the aqueous solutions of PPG400 form ABS with PEG, PEGDME, and PVP however other investigated systems don't form ABS. The ability of PPG400polymer solutions to form ABS increases by increasing temperature and polymer molar mass. It was also found that, the constant water activity lines of the ternary systems which capable and not of inducing phase separation respectively have the concave (or plane) and convex slopes. Therefore, from the slopes of the isopiestic constant water activity lines one can predict the phase forming abilities of the corresponding systems. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Aqueous two-phase system Vaporeliquid equilibria Liquideliquid equilibria Isopiestic Constant water activity lines

1. Introduction Ternary aqueous solutions of two different water soluble polymers are separated into two aqueous phases above a certain concentration in which each phases will be enriched in one of the polymers. The aqueous nature of two coexisting phases along with their different properties makes the aqueous two-phase systems possible to use them for the separation and partitioning of different biomaterials such as enzymes, cells, proteins, organelles [1e3], etc. Although both polymers are water-soluble, their aqueous solutions separate into two coexisting phases with same components but with different concentrations. Thus aqueous biphasic formation is an unusual phenomenon. Although the aqueous biphasic system (ABS) formation is well documented, its molecular level mechanism isn't still clear. During the past decade, aiming to achieve a deeper understanding about the molecular-level mechanisms of ABS formation, the salting-in and -out effects of solute 1 in solute 2 aqueous media have been theoretical and experimentally studied for salt-polymer [4,5], salt-ionic liquid [6e12], ionic liquid-polymer [13e15], amino acids-ionic liquid [16,17], and carbohydrate-ionic liquid [17,18] systems. Although the polymerepolymer ABS are the oldest ABS in relative to the other types of ABS mentioned

* Corresponding author. E-mail addresses: [email protected], [email protected] (R. Sadeghi). http://dx.doi.org/10.1016/j.polymer.2015.11.032 0032-3861/© 2015 Elsevier Ltd. All rights reserved.

above, the experimental and theoretical efforts in order to the understanding of their mechanism are considerably fewer and there remains considerable uncertainty regarding to the details of the mechanism of the phase separation in these systems. According to the effects of salts on the solubilities of ionic liquids in aqueous solutions and also based on the 1H NMR and molecular dynamic evidence, it was found that [7,9,10] the salting-out effect is an entropically driven phenomena which is a consequence of the formation of hydration complexes and the increase of the surface tension of cavity formation, but the salting-in effect is resulting from the direct binding of the ions to the hydrophobic part of the ionic liquid. The activities of different components in a solution are important for the understanding of interactions in solutions. The water activity is determined by the formation of the bond between water and solutes. Thus if the mechanism of different salting effects involves solutes dehydration or soluteesolute interactions, the water activity would be expected to be strongly dependent on the different types of solutes. The isopiestic method, which is a simple experimental method for determination of solvent activity in solutions, is based on the phenomenon that different solutions connected through their vapor spaces, approach equilibrium by transferring solvent mass by distillation. The different solutions under isopiestic equilibrium have the uniform temperature, pressure and solvent chemical potential and therefore there are no concentration gradients in the same liquid phase. In the present

2

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

paper, in order to obtain further details about the mechanisms of polymer (1) e polymer (2) ABS formation as well as a correlation between the liquideliquid equilibria (LLE) and vaporeliquid equilibria (VLE) behavior of the polymerepolymer ABS, the isopiestic equilibrium measurements for several aqueous polymer (1) epolymer (2) solutions capable or not of inducing phase separation were performed at 298.15 K. In order to cover a range of hydrophilic and hydrophobic behaviors, the polymers polyethylene glycol 200 (PEG200), polyethylene glycol 400 (PEG400), polyethylene glycol 2000 (PEG2000), polyethylene glycol 4000 (PEG4000), polyethylene glycol 6000 (PEG6000), polyethylene glycol 10000 (PEG10000), polyethylene glycol dimethyl ether 250 (PEGDME250), polyethylene glycol dimethyl ether 500 (PEGDME500), polyethylene glycol dimethyl ether 2000 (PEGDME2000), polyvinyl pyrrolidone 10000 (PVP) and polypropylene glycol 400 (PPG400) were investigated. PEGs, PPGs, PEGDMEs and PVPs are the most important and nontoxic water soluble polymers which have been used extensively for the formation of different polymerepolymer and polymer-salt ABS, so that, these polymers have been used in the more than 95% of the polymer-based ABS. In order to study the effect of structure of the polymers, the polyethers PEG, PEGDME and PPG were selected. PEG and PEGDME have similar repeating units with different end groups (OH in the case of PEG and OCH3 in the case of PEGDME). PPG and PEG have similar end groups with different repeating units so that, PPG contains a side chain methyl group which hinders hydrogen bonding between water molecules and the ether oxygen atoms. Both of repeating unit and end groups for PVP are different with those for the investigated polyethers. 2. Experimental section 2.1. Materials PPG400 was obtained from Fluka. PEGs, PEGDMEs and PVP were obtained from Merck. The polymers were used as received and doubly distilled and deionized water was used in the preparation of the all solutions. The structures of the investigated polymers have been presented in Scheme 1.

2.2. Experimental procedures The solutions were prepared on a mass basis using an analytical balance (Sartorius CP124S) with an accuracy of ±1  107 kg. The binodal curves were determined in a glass vessel, volume 50 cm [3], using the cloud point titration method at atmospheric pressure, as previously described [19]. The vessel was provided with an external jacket in which water circulated at the certain temperature (±0.05 K) using a Julabo water thermostat. Briefly, in the glass vessel containing an aqueous polymer 1 solution of known concentration, an aqueous polymer 2 solution of known mass fraction was added drop wise or vice versa, until the appearance of a cloudy solution (biphasic region), then a known mass of water was added to make the mixture clear (monophasic region) again. All the additions occurred under constant stirring. This procedure was repeated to obtain sufficient data for the construction of a complete phase diagram. The details of isopiestic apparatus and the procedure used in this work have been described previously [4]. In this method, the vapor space of different solutions, which connected through the vapor space, is evacuated and the volatile component is transported through the vapor phase until the solutions reach equilibrium. At isopiestic equilibrium, the solvent chemical potentials in each of the solutions within the isopiestic apparatus are identical. Since the activity of solvent in one solution (standard solution) at different temperatures and concentrations is known, therefore the solvent activity for other solutions in the isopiestic apparatus can be known. The used isopiestic cells have 7 or 8 sample cups uniformly distributed at the bottom. Known masses of pure anhydrous NaCl (two cups), pure polymer 1 (one cup), pure polymer 2 (one cup), and polymer 1 þ polymer 2 mixture with a certain composition (two or three cups) were added to each isopiestic sample cup and the central cup was used as a water reservoir. These sample cups were then placed in the isopiestic apparatus, air was removed, and the apparatus was held in a constant-temperature bath for period 5e9 days (depending on the solutes concentration) at 298.15 K to within ±0.05 K. Water is transferred from the central cup into the other cups containing the dry samples through the vapor phase until the central cup is dried. The chamber containing the solutions

Scheme 1. The structures of the investigated polymers: a, PEG; b, PPG, c, PEGDME and d, PVP.

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

3

Fig. 1. a. Water activity data for water (w) þ Polymer (p) systems obtained from the VPO method at 308.15 K [21]: B, PEG200;  , PEGDME250; △, PEG400 (measured in this work); >, PEGDME500; ,, PEGDME2000; þ, PEG2000 (measured in this work); A, PEG4000 (measured in this work); :, PEG6000; -, PVP (measured in this work); C, PPG400. b. Water activity data for water (w) þ PPG400 (p) system at 298.15 K [21].

Water 0.0000

0.1000

Water 1.0000

0.0000

0.9000

0.2000

0.1000

0.8000

0.3000

1.0000

0.0000 0.2000

0.3000

0.4000

0.5000

0.6000

0.7000

0.8000

0.9000

0.3000

0.8000

0.1000

0.1000

0.4000

0.7000

0.2000

0.9000

0.5000

0.6000

0.3000

0.8000

0.0000

0.6000

0.5000

0.4000

0.7000

0.7000

0.4000

0.5000

0.6000

PPG400

0.8000

0.3000

0.6000

0.5000

0.9000

0.2000

0.7000

0.4000

1.0000

0.2000

0.9000

PEG600

1.0000

a

0.1000

1.0000

PPG400

0.0000

0.0000 0.1000

0.2000

0.3000

0.4000

0.5000

0.6000

0.7000

0.8000

PEGDME2000

1.0000

b

Fig. 2. Binodal curves for a. PPG-PEG600, and b. PPG-PEGDME2000 ABS: B, 298.15 K;  , 303.15 K; C, 308.15 K; >, 313.15 K;

is kept at isothermal conditions at specific temperature until no more change in the concentration of the same solutions is observed, thus, thermodynamic equilibrium is reached. After reaching isopiestic equilibrium, the manifold assembly was removed from the bath, and the sample cups were removed and weighed with an analytical balance with a precision of ±1  104 g. Then the mass fractions of each solution were calculated from the total mass of the cups after equilibration and the initial weight of the solutes. The water activities for the cups contained the standard NaCl þ water solutions were calculated from the relation reported in the literature [20]. The estimated uncertainty for the measured water activity was ±5  104. In this study, the vapor pressure osmometery (VPO) method was also used to obtain the water activities of the binary water þ polymer solutions at low polymer concentration range. The VPO measurements were carried out with an Osmomat K-7000

0.9000

:, 318.15.

(Knauer Inc.). The details and relevant procedure of the VPO method are similarly to that in our previous paper [21]. The osmometer has two thermistors which located in a cell where the gas phase is saturated with solvent vapor. The temperature of the cell is electronically controlled and maintained with a standard uncertainty of ±1  103 K. Firstly, one droplet of the pure water is located on each thermistor, and after 5 min of equilibration, the reading is adjusted to zero. The vapor pressure (and water activity) of any aqueous solution containing solutes is lower than that of the pure water. Hence, replacing one drop of pure water with one drop of a solution leads to a vapor pressure difference between the two droplets. This difference however is compensated as follows: some vapor of the pure water that saturates the gas phase condenses on the droplet of the solution until the vapor pressures are balanced. The increasing vapor pressure of the solution droplet leads to an increase of temperature. Once equilibrium is reached, a constant

4

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11 Water 0.0

1.0

0.1

0.9

0.2

0.8

0.3

0.7

0.4

0.6

0.5

0.5

0.6

0.4

0.7

Polymer

0.0

0.3 0.1

0.2

0.3

0.4

0.5

0.6

0.7

PPG400

Fig. 3. Binodal curves for PPG-Polymer ABS at 318.15 K: C, PEG400; B, PEG600; PEG2000; △, PEG4000; -, PEG6000; ,, PEGDME2000; A, PVP.

;,

measurement value is achieved. This particular DT between the thermistors is always proportional to the vapor pressure (and water activity) of the solution. From the obtained relation between the panel readings and the corresponding concentrations of the reference NaCl solutions and therefore their water activity, the water activities for the studied polymer solutions were determined. The uncertainty in the measurement of water activity was estimated to be better than ±2  104. 3. Results and discussion In the present work, extensive isopiestic equilibrium measurements at 298.15 K were carried out on several ternary aqueous polymer 1 þ polymer 2 solutions capable or not of ABS formation to achieve further understanding about underlying reasons for the formation of polymerepolymer ABS. The investigated systems were PEG200-PG2000, PEG200-PEG10000, PEGDME250PEGDME500, PEGDME250-PEGDME2000, PVP-PEG6000, PVPPEGDME2000, PEG400-PEGDME2000, PPG400-PEG200, PPG400PEG400, PPG400-PEG600, PPG400-PEG2000, PPG400-PEG4000, PPG400-PEG6000, PPG400-PEGDME250, PPG400-PEGDME500, PPG400-PEGDME2000 and PVP-PPG400. In order to study the liquideliquid equilibria behavior of the systems which are capable of inducing phase separation, the binodal curves were also obtained. In the Supplementary Material of this manuscript, all of the obtained experimental vaporeliquid and liquideliquid equilibria data have been presented. Fig. 1 shows the plots of water activity of the investigated binary polymer þ water solutions obtained from the VPO technique at 308.15 K [21] vs concentration of the polymer. As can be seen, in a certain concentration region (mass fractions 0 to 0.4) the plot of water activity, aw, against polymer mass fraction, wp, for PPG400 has a smaller negative slope relative to the other investigated polymers. In fact because of the greater proportion of hydrocarbon in its molecule, the hydrophobicity of PPG is larger than that of the other investigated polymers. The large and small negative slopes of the aw against wp respectively are attributed to the hydrophilic and hydrophobic hydration of the polymer. From Fig. 1b it can be concluded that for the binary PPG þ water solutions with wp < 0.8 and wp > 0.8 concentration range, respectively the hydrophobic and hydrophilic hydrations of the polymer are the important factors.21From the obtained water activity data we can conclude that the hydrophilicity of the investigated polymers follows the order PEG200 > PEGDME250 > PEG400 > PEGDME500

> (PEG2000 z PEGDME2000) > (PEG4000 z PEG6000) > PVP > PPG400. If we consider the value 1 for the hydrophilicity of PEG200, from the obtained water activity data we can obtain the values 0.994, 0.991, 0.988, 0.983, 0.982, 0.981, 0.980, 0974 and 0.880 respectively for the hydrophilicity of PEGDME250, PEG400, PEGDME500, PEGDME2000, PEG2000, PEG4000, PEG6000, PVP and PPG400. Among all the investigated systems, the aqueous solutions of PPG400 form aqueous two-phase systems with PEGs, PEGDME2000, and PVP. However, the other investigated aqueous systems (including PEGsePEGs, PEGs-PEGDMEs, PEGs-PVP, PEGDMEs-PVP, PPG400-PEGDME250 and PPG400-PEGDME500) don't form aqueous two-phase systems. In Fig. 2, the binodal curves of PPG-PEG600 and PPG-PEGDME2000 ABS have been shown at different temperatures. For all of the investigated systems in the whole concentration range, the biphasic region is expanded by increasing temperature. In other words, the required polymer concentration to form a two-phase system decreases by increasing temperatures. The preferentially hydration complexes formed between the more hydrophilic polymers (PEG, PEGDME or PVP) and water molecules cause the exclusion of these polymers from the near surface region of the more hydrophobic polymer (PPG) in solution. This exclusion increases by increasing the concentration of polymers and ultimately, the system could reach a state where, phase formation would become entropically favorable and the systems separate into a more hydrophilic polymer-rich phase and a PPG-rich phase over part of the composition space. ABS formation can only involve partial dehydration of solutes. However, aqueous solutions of two polymers which have same structure (PEGePEG, PEGDMEePEGDME and PEG-PEGDME) or same hydrophilic behavior (PVP-PEG, PVP-PEGDME) don't form ABS. The hydrophobicity of PPG increases with increasing temperature [21] and therefore with an increase in temperature, the required concentration of PEG, PEGDME or PVP to form a two-phase system decreases. The effect of type and molar mass of polymer on the LLE behavior of PPG-polymer ABS, has been shown in Fig. 3. Fig. 3 shows that, an increase in the polymer molar mass causes an expansion of the biphasic region of the studied PPG-polymer ABS in the whole polymers concentration range. The abilities of the investigated polymers for the ABS formation with PPG decrease in the order: (PEG6000 z PEG4000) > (PEG2000 z PVP) > PEG600 > (PEG400 z PEGDME2000). Although for the same polymer molar mass the hydrophilicity of the investigated polymers follow the order PEGs z PEGDMEs > PVP, their abilities for ABS formation with PPG decrease in the order PEGs > PVP > PEGDME2000 so that aqueous solutions containing PPG-PEGDME250 and PPG-PEGDME500 can't form aqueous twophase systems. This may be attributed to this fact that, because of their more hydrophobic end groups (OCH3), PEGDMEs (especially those with lower molar mass which at a same mass fraction have more end groups) can interact with PPG and therefore the ability of PEGDME to salting-out of the PPG molecules in aqueous solution is smaller than that of PEGs and PVP. Another factor that can influence the phase behavior of the investigated systems is polymer chain entanglement or intermolecular hydrogen-bonding. PPG400 is entangled by transient coupling via hydrogen-bonding of the chain ends [22], which reduces its hydration in aqueous solutions. Therefore in the case of PPG400-polymer aqueous solutions, entanglements of the PPG chains reduce the PPG-polymer interactions and therefore these systems separate into two aqueous phases. However, in the case of PEGs, PEGDMEs and PVPs, only for high molar mass, they are mainly present as a helix and for low molar mass, a hydrated polymer shows a conformation as an expanded random coil [23,24]. In fact, the physical state of these polymers has been

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

5

Fig. 4. a. Plot of weight fraction of PEG2000, wp1, against weight fraction of PEG200, wp2, for constant water activity curves of PEG2000 þ PEG200 þ H2O system at 298.15 K: C, 0.9325; B, 0.9534; △, 0.9809; :, 0.9875. b. Plot of weight fraction of PEG10000, wp1, against weight fraction of PEG200, wp2, for constant water activity curves of PEG10000 þ PEG200 þ H2O system at 298.15 K: C, 0.9325; B, 0.9534; △, 0.9809; :, 0.9875. c. Plot of weight fraction of PEGDME500, wp1, against weight fraction of PEGDME250, wp2, for constant water activity curves of PEGDME500 þ PEGDME250 þ H2O system at 298.15 K: C, 0.9695; B, 0.9788; △, 0.9844; :, 0.9881. d. Plot of weight fraction of PEGDME2000, wp1, against weight fraction of PEGDME250, wp2, for constant water activity curves of PEGDME2000 þ PEGDME250 þ H2O system at 298.15 K: C, 0.9695; B, 0.9788; △, 0.9844; :, 0.9881; A, 0.9911; …, calculated by the eq (1).

reported to depend on their molecular weight, where higher molecular weight polymers have been shown to give rise to more solid polymeric materials, due to the entanglements of the polymer chains [25], and the inter-polymer chain bonding formed by hydrogen bonding [26]. At low temperatures low-energy polar conformations exist in the polymers and lead to aqueous solubility (more hydrophilicity), while higher energy and less polar conformations (more hydrophobicity) are favored at higher temperatures [27]. Thus, the polymers become more hydrophobic with increasing temperature. On the other hand, the polymer solubility may also depend on the degree of hydration of the polymer chain in water. The hydrogen bond interactions between the water and polymer molecules lead to mutual stabilization of the water and polymers structure (negative contribution to the entropy of mixing) at low temperatures. At higher temperatures, hydration shells around the polymer are less developed and the entropy of mixing increases. These behaviors are in agreement with the expansion of the biphasic region of the studied PPG-polymer ABS by increasing the polymer molar mass/temperature. The PEGs and PEGDMEs are telechelic oligomers; capable of crosslinking the longer PVP macromolecules by Hbonding through both terminal hydroxyl groups and therefore these ternary aqueous systems are completely miscible.

The lines of constant water activity or vapor pressure for PEG200 þ PEG2000 þ H2O, PEG200 þ PEG10000 þ H2O, PEGDME250 þ PEGDME500 þ H2O and PEGDME250 þ PEGDME2000 þ H2O systems, in which two polymers have a similar repeating units, are plotted in Fig. 4. Fig. 5 shows the constant water activity lines for aqueous PEG-PVP, PEG- PEGDME and PEGDME-PVP systems, in which two different polymers have similar hydrophilicity. Figs. 6 and 7 respectively show the constant water activity lines of PPG400 þ PEGs (PEG200, PEG400, PEG2000 and PEG4000) þ H2O and PPG400 þ PVP or PEGDMEs (PEGDME250, PEGDME500 and PEGDME2000) þ H2O systems, in which two different polymers have different hydrophobic properties. In these figures, the results of the ZdanovskiieStokes-Robinson rule [28] for the semi-ideal behavior (following equation) have also been shown.

wp1 w0p1

þ

wp2 w0p2

¼1

aw ¼ consant; 0 

wp1 w0p1

 1 and 0 

wp2 w0p2

! 1 (1)

where wpi and wopi are the polymer i mass fractions in the ternary

6

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

Fig. 5. a. Plot of weight fraction of PEG6000, wp1, against weight fraction of PVP, wp2, for constant water activity curves of PEG6000 þ PVP þ H2O system at 298.15 K: C, 0.9701; B, 0.9819; △, 0.9908; :, 0.9959. b. Plot of weight fraction of PEG400, wp1, against weight fraction of PEGDME2000, wp2, for constant water activity curves of PEG400 þ PEGDME2000 þ H2O system at 298.15 K: C, 0.8674; B, 0.9524; △, 0.9722. c. Plot of weight fraction of PVP, wp1, against weight fraction of PEGDME2000, wp2, for constant water activity curves of PVP þ PEGDME2000 þ H2O system at 298.15 K: C, 0.8305; B, 0.9524; △, 0.9722; …, calculated by the eq (1).

and binary solutions of equal aw, respectively. According to the semi-ideal solvation theory, the similar relation was theoretically derived by Stokes and Robinson [29] for the mixed non-electrolyte aqueous solutions in the isopiestic equilibrium. The semi-ideality indicates that under constant water activity the constituent binary solutions mix ideally. This means that the soluteewater interactions in the ternary solutions are similar to those in the constituent binary solutions under isopiestic equilibrium. As can be seen, the constant water activity lines of polymer 1 þ polymer 2 þ H2O systems show the positive deviation from the eq (1) in the case of: a) polymers 1 and 2 have a similar repeating units (PEG2000 þ PEG200 þ H2O, PEG10000 þ PEG200 þ H2O, PEGDME500 þ PEGDME250 þ H2O and PEGDME2000 þ PEGDME250 þ H2O, as can be seen from Fig. 4) and b) polymers 1 and 2 have different repeating units with a similar hydrophobic property (PEG6000 þ PVP þ H2O, PEG400 þ PEGDME2000 þ H2O and PVP þ PEGDME2000 þ H2O, as can be seen from Fig. 5). These systems don't form aqueous twophase systems (they are not capable of inducing phase separation). However the ternary polymer 1 þ polymer 2 þ water systems in which two polymers have different repeating units with large differences between their hydrophobicity (PPG400 þ PEG200 þ H2O, PPG400 þ PEG400 þ H2O, PPG400 þ PEG2000 þ H2O, PPG4000 þ PEG4000 þ H2O, PPG400 þ PEGDME250 þ H2O, PPG400 þ PEGDME500 þ H2O, PPG400 þ PEGDME2000 þ H2O and PPG4000 þ PVP þ H2O, as can

be seen from Figs. 6 and 7) are capable of inducing ABS formation. The constant water activity lines of these systems in the low solutes concentration range (one-phase region and two-phase area near the binodal curve) show the negative deviation from the eq (1) and those in the two-phase area far from the binodal curve show the positive deviation from the eq (1). Therefore, it can be concluded that, similar to the other types of ABSs [4,12,15], there may be a correlation between the VLE behavior and the aqueous biphasic forming ability of the polymer 1 þ polymer 2 þ water solutions. The soluteewater interaction in a ternary aqueous semi-ideal system is same as those in the constituent binary aqueous solutions under isopiestic equilibrium and therefore they conform to the eq (1). For the systems with a positive (or negative) deviation from the linear isopiestic relation, the concentrations of polymers 1 and 2 in a ternary aqueous polymer 1 þ polymer 2 solution under isopiestic equilibrium with certain binary aqueous solutions of polymers 1 and 2 are larger (or smaller) than those one expect for the semi-ideal behavior (straight line). In the systems with positive deviations from the linear isopiestic relation, (which cannot form ABS) because of the polymer 1-polymer 2 interactions, the polymer 1-water interactions weaken in the presence of polymer 2 which in turn increase the free water molecules respecting to the semi-ideal behavior. Fig. 8 shows that, for the ternary aqueous PEGePEG or PEGDMEePEGDME systems, the positive deviations increase by increasing water activity and at same water activity, the positive deviations increase by increasing the differences between the

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

7

Fig. 6. a. Plot of weight fraction of PEG200, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEG200 þ H2O system at 298.15 K: C, 0.7958; B, 0.8830; △, 0.9148; :, 0.9559. b. Plot of weight fraction of PEG400, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEG400 þ H2O system at 298.15 K: C, 0.7794; B, 0.8704; △, 0.9335; :, 0.9559; A, 0.9714;  , Binodal. c. Plot of weight fraction of PEG2000, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEG2000 þ H2O system at 298.15 K: C, 0.9092; B, 0.9383; △, 0.9431; >, 0.9657; :, 0.9761; A, 0.8820;  , Binodal. d. Plot of weight fraction of PEG4000, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG4000 þ PEG4000 þ H2O system at 298.15 K: C, 0.9020; B, 0.9258; △, 0.9657; :, 0.9761; A, 0.9820;  , Binodal; …, calculated by the eq (1).

molar mass of polymers 1 and 2. However, in the case of ternary aqueous polymer 1 þ polymer 2 systems capable of inducing phase separation, the polymers 1 and 2 have different structures and because of the unfavorable polymer 1-polymer 2 interactions, the interaction of polymer 1 with water becomes more favorable in the presence of polymer 2. In other words, unfavorable polymer 1-polymer 2 interactions decrease the amount of free water molecules respecting to the semi-ideal behavior and therefore, these polymers exclude themselves from the vicinity of each other due to their preferential hydration. With rise up the polymers concentrations, this exclusion increases and the system becomes unstable entropically so that, the system could reach a state where, ABS formation becomes favorable [30] and PPG becomes salted-out by the other polymer. In fact, in the polymer e PPG ABS, the hydrophilic hydration cosphere of the polymer and hydrophobic hydration cosphere of PPG cause the repulsive force between two polymers which in turn cause the structural incompatibility and therefore, phase separation in these systems. The results show that the negative deviations from the linear isopiestic relation increase as the constant water activity curves are closed to the binodal curve (Figs. 6 and 7). The results illustrated in Fig. 8c show that at the same concentration of polymers the negative deviations from the linear isopiestic relation for the PPG þ PEGDME2000 system increase by

increasing temperature. As mentioned above, PPG-water interaction weakens by increasing temperature and therefore, at higher temperature the ability of PEGDME to increase the PPG-water interactions is larger than that in lower temperature. Therefore we can expect that the negative deviations from the linear isopiestic relation increase by increasing temperature. This is in agreement with our LLE observations that, the ability of PEGDME to salting-out of the PPG molecules in solution increases by increasing temperature which leads to a lower PEGDME concentration required to form a two-phase system (as can be seen from Fig. 2). The constant water activity lines of PEGDMEs (different molecular weight) þ PPG400 þ water systems have been shown in Fig. 8d. As can be seen, although polymers with higher molecular weight (more hydrophobic polymers) require the lower concentrations to form an ABS (Fig. 3), the negative deviations of constant water activity lines from the linear isopiestic relation don't vary significantly by increasing the molecular weight of PEGDME. Partial dehydration of the polymers occurs with phase separation and then the polymerewater interaction, which was strengthened by the other polymer, becomes weaker through ABS formation and therefore, in the biphasic area far from the binodal curve, the positive deviation from the linear isopiestic relation is observed. In this work, for the correlation of obtained water activity data, the segment-based local composition Wilson model [31] was

8

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

Fig. 7. a. Plot of weight fraction of PEGDME250, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEGDME250 þ H2O system at 298.15 K: C, 0.7043; B, 0.8606; △, 0.9308; :, 0.9508; A, 0.9646. b. Plot of weight fraction of PEGDME500, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEGDME500 þ H2O system at 298.15 K: C, 0.7746; B, 0.8606; △, 0.9148; :, 0.9557. c. Plot of weight fraction of PEGDME2000, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG400 þ PEGDME2000 þ H2O system at 298.15 K: C, 0.8485; B, 0.9508; △, 0.9646;  , Binodal. d. Plot of weight fraction of PVP, wp1, against weight fraction of PPG400, wp2, for constant water activity curves of PPG4000 þ PVP þ H2O system at 298.15 K: C, 0.8780; B, 0.9485; △, 0.9648;  , Binodal curve at 318.15 K; …, calculated by the eq (1).

considered. In this model, the activity coefficient of the water (w), gw, is considered as the sum of the combinatorial, ln gComb: , and the w residual, ln gRes: , contributions: w

ln gw ¼ ln gComb: þ ln gRes: w w

(2)

The FloryeHuggins and Wilson models were used for combinatorial and residual contributions, respectively. The FloryeHuggins relation, which accounts the excess entropy associated with random mixing, for the water activity coefficient in a polymer (1) þ polymer (2) þ water (w) system can be written as:

ln

gComb: w

  4 4 4 4 ¼ ln w þ 1  rw w þ 1 þ 2 xw rw r1 r2

(3)

In this equation

ri xi 4i ¼ rw xw þ r1 x1 þ r2 x2

(4)

where, xi and ri respectively are the mole fraction and the number of the segments of the component i. The Wilson model, which accounts the residual contribution, for the water activity coefficient in a polymer (1) þ polymer (2) þ water (w) system can be written as:

1  ln gWilson ¼ lnð41 H1w þ 42 H2w þ 4w Þ w Crw   Hw1  ð41 þ 42 H21 þ 4w Hw1 Þ þ 41 41 þ 42 H21 þ 4w Hw1   Hw2  ð41 H12 þ 42 þ 4w Hw2 Þ þ 42 41 H12 þ 42 þ 4w Hw2   1  ð41 H1w þ 42 H2w þ 4w Þ þ 4w 41 H1w þ 42 H2w þ 4w

(5)

  Eji Hji ¼ exp  CRT

(6)

Eji ¼ hji  hii

(7)

where C is the effective coordination number in the system which was fixed at 10. hji and hii are enthalpies of jei and iei interactions, respectively. In the present paper, the values of r ¼ 1 and (molar volume of polymer at 298.15 K/molar volume of water at 298.15 K) were considered for water and polymer, respectively. The specific volumes of PEG, PVP, PEGDME and PPG at 298.15 K are 0.844, 0.798,

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

w

9

w

p2 Fig. 8. a. Plot of wp1 for constant water activity curves of PEG2000 (1) þ PEG200 (2) þ H2O (solid lines) and PEG10000 (1) þ PEG200 (2) þ H2O (dashed lines) systems at 0 against w0 p1

p2

298.15 K: C, 0.9325; B, 0.9809. b. Plot of

wp1 w0p1

against

wp2 w0p2

for constant water activity curves of PEGDME500 (1) þ PEGDME250 (2) þ H2O (solid lines) and PEGDME2000

(1) þ PEGDME250 (2) þ H2O (dashed lines) systems at 298.15 K: C, 0.9695; B, 0.9881. c. Plot of (2) þ H2O system at: C, 298.15 K; B, 308.15 K;  , 318.15 K d. Plot of

wp1 w0p1

against

wp2 w0p2

wp1 w0p1

against

wp2 w0p2

for constant water activity curves of PEGDME2000 (1) þ PPG400

for constant water activity curves of PEGDME250 (1) þ PPG400 (2) þ H2O (solid lines) and

PEGDME2000 (1) þ PPG400 (2) þ H2O (dashed lines) systems at 298.15 K: C, 0.8508; B, 0.9646; …, calculated by the eq (1).

Table 1 Parameters of the Wilson model obtained from the correlation of water activity data of the investigated polymer (1) þ polymer (2) þ water (w) systems (given in Table 1 of the supporting information) at T ¼ 298.15 K along with the corresponding standard deviations (102s). System

E1w/ kJ.mol1

Ew1/ kJ.mol1

E2w/ kJ.mol1

Ew2/ kJ.mol1

E12/ kJ.mol1

E21/ kJ.mol1

102s

PPG (1) þ PEG (2) þ water (w) PEG (1) þ PEG (2) þ water (w) PPG (1) þ PEGDME (2) þ water (w) PEG (1) þ PEGDME (2) þ water (w) PEGDME (1) þ PEGDME (2) þ water (w) PVP (1) þ PEG (2) þ water (w) PVP (1) þ PEGDME (2) þ water (w) PVP (1) þ PPG (2) þ water (w)

18.7860 6.9047 18.7860 6.9047 11.9539 5.8215 5.8215 5.8215

11.2160 5.0428 11.2160 5.0428 8.2531 3.9557 3.9557 3.9557

6.9047 6.9047 11.9539 11.9539 11.9539 6.9047 11.9539 18.7860

5.0428 5.0428 8.2531 8.2531 8.2531 5.0428 8.2531 11.2160

5.8213 0.0000 3.0857 3.1361 0.0000 2.6131 0.4155 6.9965

5.3970 0.0000 3.3014 1.8370 0.0000 3.0258 0.7503 7.4628

0.82 0.45 1.42 1.96 0.43 0.35 1.71 0.97

0.905 and 0.920 cm3  g1, respectively. The segment-based local composition Wilson model was used for correlation of the experimental water activity data obtained from the isopiestic technique and the obtained parameters for the studied systems along with the corresponding standard deviations (102s) of the fit are presented in Table 1. According to the deviations given in Table 1, we conclude that the segment-based local composition Wilson model represents the experimental water activity data of the investigated polymer solutions, with good accuracy. In fact, the

used model in this work is a segment-based local composition model and these types of models, unlike the conventional models, have a predictive capability. This is because; the interaction parameters of the segment-based models are for the interaction between two segments not molecules so, they are quite independent of chain length, or degree of polymerization. Therefore, these models can cover a wide range of polymer molecular weights with a series of interaction parameters. As can be seen from Table 1, the calculated parameters are independent on the

10

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11

Table 2 The calculated values of DGmix for some ternary PPG400 þ PEG400 þ H2O (an example for the negative deviation from the linear isopiestic relation) and PEG200 þ PEG10000 þ H2O (an example for the positive deviation from the linear isopiestic relation) at T ¼ 298.15 K along with the corresponding deviations from the linear isopiestic relation. System

wp1

PPG400 (1) þ PEG400 (2) þ water (w)

PEG200 (1) þ PEG10000 (2) þ water (w)

a

0.1122 0.2030 0.1444 0.2672 0.1833 0.3468 0.1243 0.2580 0.1061 0.2129 0.0807 0.1394

wp2 0.1788 0.1138 0.2293 0.1483 0.2910 0.1925 0.3571 0.1733 0.3050 0.1430 0.2089 0.0920

wp1 w0p1

w

þ wp2 0  1

DGmix/kJ mol1

DGmixa/kJ mol1

0.438 0.393 0.597 0.551 0.830 0.810 1.039 0.909 0.825 0.709 0.525 0.429

2.705 3.105 1.254 1.564 1.566 2.069 0.214 0.347 0.037 0.104 0.043 0.029

p2

0.0474 0.0729 0.0666 0.0925 0.0492 0.0707 0.0443 0.0342 0.0562 0.0487 0.0652 0.0582

The calculated values of DGmix for the semi ideal behavior.

polymer molar mass and can also give a good representation of the ternary polymerepolymerewater solutions VLE behavior at different polymer molar mass. The Wilson relation for the molar Gibbs free energy of mixing, DGmix, of polymer 1 e polymer 2 aqueous solutions is:

DGmix ¼ x1 ln 41 þ x2 ln 42 þ xw ln 4w  C  ½x1 r1 lnð41 RT þ 42 H21 þ 4w Hw1 Þ þ x2 r2 lnð41 H12 þ 42 þ 4w Hw2 Þ þ xw rw lnð41 H1w þ 42 H2w þ 4w Þ (8) Using the eq (8) and obtained parameters given in Table 1, the molar Gibbs free energy of mixing for the investigated systems was determined. In Table 2, the calculated values of DGmix for some ternary solutions PEG10000 þ PEG200 þ H2O (an example for the positive deviation from the linear isopiestic relation) and PPG400 þ PEG400 þ H2O (an example for the negative deviation from the linear isopiestic relation) have been given. As can be seen, in the case of the negative deviations from the linear isopiestic relation, the values of DGmix for the real ternary solutions are less negative than those we expect in the case of semi-ideal behavior. These systems are less stable than those of the corresponding semi ideal solutions. Therefore, by increasing the solutes concentration, the system could reach a state where, for entropic reasons, phase formation would become favorable. On the other hand, for the ternary aqueous polymer 1 þ polymer 2 solutions, which have the positive deviations from the linear isopiestic relation, the calculated values of DGmix for the real ternary solutions are more negative than those we expect in the case of semi-ideal behavior. In fact these systems, because of the preferential polymer 1-polymer 2 interactions, are more stable than those of the corresponding semiideal solutions and therefore, these systems don't form ABS. This is in agreement with our VLE and LLE results.

4. Conclusions Isopiestic equilibrium measurements at 298.15 K on PEG200PG2000, PEG200-PEG10000, PEGDME250-PEGDME500, PEGDME250-PEGDME2000, PVP-PEG6000, PVP-PEGDME2000, PEG400-PEGDME2000, PPG400-PEG200, PPG400-PEG400, PPG400-PEG600, PPG400-PEG2000, PPG400-PEG4000, PPG400PEG6000, PPG400-PEGDME250, PPG400-PEGDME500, PPG400PEGDME2000 and PPG400-PVP aqueous solutions were carried out in order to gathering further experimental evidence about the molecular-level mechanisms of the polymerepolymer aqueous

biphasic formation. The results show that, in the case of aqueous solutions of PPG400 (the most hydrophobic polymer)-polymer which form aqueous biphasic systems, because of the repulsive force due to the structural incompatibility of the hydrophilic hydration cosphere of polymer and hydrophobic hydration cosphere of PPG, the interaction of each polymer with water becomes more favorable in the presence of the other polymer and therefore, these systems show the negative deviation from the linear isopiestic relation in the one-phase area or in the two-phase area near the binodal curve. For these systems, the negative deviation of constant water activity curves from the linear isopiestic relation increases by increasing temperature. The abilities of these PPG-polymer aqueous solutions to form the aqueous biphasic systems are promoted by increasing the temperature or polymer molar mass and decrease in the order (PEG6000 z PEG4000) > (PEG2000 z PVP) > PEG600 > (PEG400 z PEGDME2000). However, in the two-phase area far from the binodal curve, partial dehydration of polymers due to the formation of aqueous two-phase systems will cause the positive deviation from the linear isopiestic relation. Other investigated polymer 1 þ polymer 2 þ water systems cannot form aqueous two-phase systems. In these systems, because of the favorable interactions between polymers 1 and 2, the polymer 1 e water interactions become less favorable in the presence of polymer 2. Hence, the solutes concentrations in a ternary solution, which is in isopiestic equilibrium with certain binary polymer 1 þ water and polymer 2 þ water solutions are larger than those we expect in the case of semi-ideal solution, which means that they show the positive deviations from the linear isopiestic relation. Finally, it was found that, in the case of negative and positive deviations from the linear isopiestic relation, the calculated values of DGmix (by the segment-based local composition Wilson model) for real ternary solutions are less and more negative than those we expect in the case of semi-ideal behavior, respectively. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2015.11.032. References [1] P.A. Albertsson, Partitioning of Cell Particles and Macromolecules, third ed., Wiley-Interscience, New York, 1986. [2] H. Walter, D.E. Brooks, D. Fisher, Partitioning in Aqueous Two-phase Systems, Academic Press, New York, 1985. [3] J. Chen, G.X. Ma, D.Q. Li, HPCPC separation of proteins using polyethylene glycol-potassium phosphate aqueous two-phase, Prep. Biochem. Biotechnol. 29 (1999) 371e383. [4] R. Sadeghi, F. Jahani, Salting-in and salting-out of water soluble polymers in

R. Sadeghi, M. Maali / Polymer 83 (2016) 1e11 aqueous salt solutions, J. Phys. Chem. B 116 (2012) 5234e5241. [5] C.-L. Ren, W.-D. Tian, I. Szleifer, Y.-Q. Ma, Specific salt effects on poly(ethylene oxide) electrolyte solutions, Macromolecules 44 (2011) 1719e1727. [6] J.R. Trindade, Z.P. Visak, M. Blesic, I.M. Marrucho, J.A.P. Coutinho, J.N.C. Lopes, L.P.N. Rebelo, Salting-out effects in aqueous ionic liquid solutions: cloud-point temperature shifts, J. Phys. Chem. B 111 (2007) 4737e4741. [7] M.G. Freire, P.J. Carvalho, A.M.S. Silva, L.M.N.B.F. Santos, L.P.N. Rebelo, I.M. Marrucho, J.A.P. Coutinho, Ion specific effects on the mutual solubilities of water and hydrophobic ionic liquids, J. Phys. Chem. B 113 (2009) 202e211. [8] C.M.S.S. Neves, S.P.M. Ventura, M.G. Freire, I.M. Marrucho, J.A.P. Coutinho, Evaluation of cation influence on the formation and extraction capability of ionic-liquid-based aqueous biphasic systems, J. Phys. Chem. B 113 (2009) 5194e5199. [9] L.I.N. Tome, F.R. Varanda, M.G. Freire, I.M. Marrucho, J.A.P. Coutinho, Towards an understanding of the mutual solubilities of water and hydrophobic ionic liquids in the presence of salts: the anion effect, J. Phys. Chem. B 113 (2009) 2815e2825. [10] M.G. Freire, C.M.S.S. Neves, A.M.S. Silva, J.O.S. Santos, I.M. Marrucho, L.P.N. Rebelo, J.K. Shah, E.J. Maginn, J.A.P. Coutinho, H NMR and molecular dynamic evidence for an unexpected interaction on the origin of salting-in/ salting-out phenomena, J. Phys. Chem. B 114 (2010) 2004e2014. [11] S. Shahriari, C.M.S.S. Neves, M.G. Freire, J.A.P. Coutinho, Role of the Hofmeister series in the formation of ionic-liquid-based aqueous biphasic systems, J. Phys. Chem. B 116 (2012) 7252e7258. [12] R. Sadeghi, B. Mostafa, E. Parsi, Y. Shahebrahimi, Toward an understanding of the salting-out effects in aqueous ionic liquid solutions: vapor-liquid equilibria, liquid-liquid equilibria, volumetric, compressibility, and conductivity behavior, J. Phys. Chem. B 114 (2010) 16528e16541. [13] Z. Li, X. Liu, Y. Pei, J. Wang, M. He, Design of environmentally friendly ionic liquid aqueous two-phase systems for the efficient and high activity extraction of proteins, Green Chem. 14 (2012) 2941e2950. [14] M.G. Freire, J.F.B. Pereira, M. Francisco, H. Rodriguez, L.P.N. Rebelo, R.D. Rogers, J.A.P. Coutinho, Insight into the interactions that control the phase behaviour of new aqueous biphasic systems composed of polyethylene glycol polymers and ionic liquids, Chem. Eur. J. 18 (2012) 1831e1839. [15] T. Moradian, R. Sadeghi, Isopiestic investigations on the interactions of water soluble polymers with imidazolium-based ionic liquids in aqueous solutions, J. Phys. Chem. B 117 (2013) 7710e7717. [16] M. Domínguez-Perez, L.I.N. Tome, M.G. Freire, I.M. Marrucho, O. Cabeza, J.A.P. Coutinho, (Extraction of biomolecules using) aqueous biphasic systems formed by ionic liquids and aminoacids, Sep. Purif. Technol. 72 (2010) 85e91. [17] M.G. Freire, A.F.M. Claudio, J.M.M. Araujo, J.A.P. Coutinho, I.M. Marrucho, J.N.C. Lopes, L.P.N. Rebelo, Aqueous biphasic systems: a boost brought about

11

by using ionic liquids, Chem. Soc. Rev. 41 (2012) 4966e4995. [18] M.G. Freire, C.L.S. Louros, L.P.N. Rebelo, J.A.P. Coutinho, Aqueous biphasic systems composed of a water-stable ionic liquid þ carbohydrates and their applications, Green Chem. 13 (2011) 1536e1545. [19] M.T. Zafarani-Moattar, R. Sadeghi, Liquid-liquid equilibria of aqueous twophase systems containing polyethylene glycol and sodium dihydrogen phosphate or disodium hydrogen phosphate - experiment and correlation, Fluid Phase Equilib. 181 (1e2) (2001) 95e112. [20] E.C.W. Clarke, D.N. Glew, Evaluation of the thermodynamic functions for aqueous sodium chloride from equilibrium and calorimetric measurement below 154 ◦C, J. Phys. Chem. Ref. Data 14 (1985) 489e610. [21] R. Sadeghi, Y. Shahebrahimi, Vapor-liquid equilibria of aqueous polymer solutions from vapor pressure osmometry and isopiestic measurements, J. Chem. Eng. Data 56 (2011) 789e799. [22] C.M. Roland, M. Paluch, R. Casalini, Effects of the volume and temperature on the global and segmental dynamics in poly (propyleneglycol) and 1,4-polyisoprene, J. Polym. Sci. B Polym. Phys. 42 (2004) 4313e4319. [23] D. Eagland, N.J. Crowther, C.J. Butler, Interaction of poly(oxyethylene) with water as a function of molar mass, Polymer 34 (1993) 2804e2808. [24] E. Sabadini, E.M. Assano, T.D.Z. Atvars, Stoichiometry of poly(ethylene glycol) in water and benzene by excess volume, J. Appl. Polym. Sci. 63 (1997) 301e306. [25] R. Walkenhorst, J.C. Selser, G. Piet, Long-ranged relaxations in poly(ethylene oxide) melts: evidence for network behavior, J. Chem. Phys. 109 (1998) 11043e11050. [26] J.P. Coffman, C.A. Naumann, Molecular weight dependence of viscoelastic properties in two-dimensional physical polymer networks: amphiphilic lipopolymer monolayers at the airwater interface, Macromolecules 35 (2002) 1835e1839. [27] H. Hartounian, E. Floeter, E.W. Kaler, S.I. Sandler, Effect of temperature on the phase equilibrium of aqueous two-phase polymer systems, AIChE J. 39 (1993) 1976e1984. [28] A.B. Zdanovskii, Regularities in the property variations of mixed solutions, Tr. Solyanoi Lab. Akad. Nauk. SSSR 6 (1936) 5e70. [29] R.H. Stokes, R.A. Robinson, Interactions in aqueous nonelectrolyte solutions. I. Solute-solvent equilibria, J. Phys. Chem. 70 (1966) 2126e2131. [30] L.H.M. da Silva, W. Loh, Calorimetric investigation of the formation of aqueous two-phase systems in ternary mixtures of water, poly(ethylene oxide) and electrolytes (Or dextran), J. Phys. Chem. B 104 (2000) 10069e10073. [31] R. Sadeghi, Extension of the Wilson model to multicomponent polymer solutions: applications to polymer-polymer aqueous two-phase systems, J. Chem. Thermodyn. 37 (1) (2005) 55e60.