© Copyright 2002 American Chemical Society
OCTOBER 29, 2002 VOLUME 18, NUMBER 22
Letters “Internal pKa’s” in Polyelectrolyte Multilayers: Coupling Protons and Salt Hassan H. Rmaile and Joseph B. Schlenoff* Department of Chemistry and Biochemistry, Center for Materials Research and Technology (MARTECH), The Florida State University, Tallahassee, Florida 32306-4390 Received February 12, 2002. In Final Form: August 1, 2002 In situ UV-vis absorption spectroscopy is used to probe the extent of dissociation of a weak polycarboxylic acid incorporated within a polyelectrolyte multilayer. The effective acid dissociation constant is shown to depend on the nature of the multilayer. Salt counterion and proton populations within the multilayer are shown to be coupled, leading to a quantitative relationship between polyelectrolyte acidity and salt within these reluctant exchangers.
Ultrathin films of complexed polyelectrolytes, also termed “polyelectrolyte multilayers” (PEMUs), may be formed by multiple ion pairing interactions between charged polymers.1-3 To produce PEMUs, oppositely charged components, which may also comprise biopolymers3 or particulates,4 are adsorbed to a substrate in an alternating fashion. For combinations of strongly dissociating synthetic polymers, charges between polymers are generally well matched, excluding small “salt” ions from the bulk.5 However, multilayers have also been prepared using weakly dissociating polyelectrolytes, where pH controls ionization and therefore the internal charge balance.6-8 It has been observed that the apparent dissociation constant of a weak polyacid shifts when it is incorporated into a PEMU.6-9 For example, poly(acrylic
acid) became more acidic when part of a PEMU.6,7 Similar shifts in pKa have been observed in the solution phase precipitation of polyacids in the presence of polybases.10 In the present work, we measure, in situ, these apparent shifts in polyelectrolyte acidity and we show how they are coupled to total ionic strength (salt concentration). Multilayers were prepared on quartz plates using poly(1-[4-(3-carboxy-4-hydroxyphenylazo)benzenesulfonamido]1,2-ethanediyl), PAZO, a highly colored, weak polycarboxylic acid with a solution pKa of about 3.3, and poly(diallyldimethylammonium chloride), PDADMA, a strongly dissociated quaternary ammonium polycation. The visible absorption maximum, λmax, of PAZO depends on the protonation state of the carboxylate group. By use of concentrations in lieu of activities
(1) Decher, G. Science 1997, 277, 1232. (2) Bertrand, P.; Jonas, A.; Laschewsky, A.; Legras, R. Macromol. Rapid. Commun. 2000, 21, 319. (3) Lvov, Y. In Protein Architecture. Interfacing Molecular Assemblies and Immobilization Biotechnology; Lvov, Y., Mo¨hwald, H., Eds.; M. Dekker: New York, 2000. (4) Ostrander, J. W.; Mamedov, A. A.; Kotov, N. A. J. Am. Chem. Soc. 2001, 123, 1101. (5) Schlenoff, J. B.; Ly, H.; Li, M. J. Am. Chem. Soc. 1998, 120, 7626. (6) Mendelsohn, J. D.; Barrett, C. J.; Chan, V. V.; Pal, A. J.; Mayes, A. M.; Rubner, M. F. Langmuir 2000, 16, 5017. (7) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213. (8) Xie, A. F.; Granick, S. J. Am. Chem. Soc. 2001, 123, 3175. (9) Klitzing, R.; Mohwald, H. Langmuir 1995, 11, 3554.
Pol-H+aq h Pol-aq + H+aq Ka(aq) )
(1)
[Polaq-][Haq+] [Pol-Haq+]
where Pol- is a polyelectrolyte repeat unit. There is a ca. 12 nm difference in λmax between protonated and unprotonated forms, and the molar extinction coefficients of both forms are almost identical.
10.1021/la025624s CCC: $22.00 © 2002 American Chemical Society Published on Web 09/28/2002
8264
Langmuir, Vol. 18, No. 22, 2002
Letters
Figure 1. Absorption wavelength (λmax) relative to λmax for the fully protonated form, of PAZO in solution (triangles) and in multilayers made with PDADMA (40 layers, 120 nm thick, squares) or PNO4VP (40 layers, 100 nm, circles). All measurements were made in 0.1 M NaClaq. Solid lines are fits to eq 3 using respective pKa values of 3.3, 2.2, and 1.78. The dotted line represents the titration curve of a PAZO/PDADMA multilayer immersed in 0.5 M NaCl (pKa(m)app ) 2.55). Arrow indicates the span for ∆λD.
The fraction unprotonated in a particular sample, R, is provided by λmax according to
[Pol-]
∆λ R) ) - + ∆λ [Pol H ] + [Pol ] D
(2)
where ∆λ is the shift in λmax, with ∆λ ) 0 and ∆λ ) ∆λD corresponding to λmax for the fully protonated and unprotonated, respectively, form of the polymer. R, and therefore ∆λ, is related to pH according to
pH ) pKa + log
R 1-R
(3)
A pH vs ∆λ curve for 10-4 M PAZO in 0.1 M NaClaq is depicted in Figure 1. Correlation of the data to a fit (solid line) to eq 3 with pKa ) 3.3 shows reasonably uncomplicated behavior at this level of precision. When PAZO is combined with PDADMA to form a multilayer,11,12 the entire absorption spectrum shifts, as has been observed previously,13 but the difference between protonated and unprotonated forms, ∆λD, remains the same (about 12 nm). As seen in Figure 1, the titration curve is shifted by 1.1 pH units, with a good fit to eq 3 assuming the apparent pKa, pKa(m)app, inside the multilayer ) 2.2.14,15 Creation of a negatively charged unit leads to the formation of an energetically favorable polymer/ polymer ion pair. Protonation removes this interaction, (10) Kabanov, V. A.; Zezin, A. B. Sov. Sci. Rev., Sect. B 1982, 4, 207. (11) Multilayers were deposited on spinning polished silicon wafers using a robot (StratoSequence V, nanoStrata Inc.) as described previously (ref 12). Twenty layers of each polyelectrolyte were deposited, starting with the positive polymer, from 1 mM solutions of polymer buffered to pH 4.4 with 10 mM acetate. PDADMA and PAZO were deposited from 0.2 M NaCl(aq) and PNO4VP was deposited from ethanol. (12) Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153. (13) Lvov, Y.; Yamada, S.; Kunitake, T. Thin Solid Films 1997, 300, 107. (14) For very low pH, some distortion is seen due to the fact that the acid concentration approaches the salt concentration and thus contributes to the net ionic strength. (15) Unlike multilayers made with poly(acrylic acid) (see refs 6 and 16), multilayers remained stable down to a pH of 0.5. (16) Dubas, S. T.; Schlenoff, J. B. Macromolecules 2001, 34, 3736.
thus the polyacids in multilayers are more difficult to protonate (lower apparent pKa). Using the solution form of PAZO as a reference state, under the defined salt concentration, the free energy change due to ion pairing, or the “driving force” of multilayer formation, is estimated by ∆Go(0.1MNaCl) ) -2.3RT(pKa(aq) - pKa(m)app).10 For the conditions in Figure 1 (0.1 M salt) ∆Go(0.1MNaCl) ) -6.3 kJ mol-1. In situ titration curves were also determined for multilayers made from PAZO and poly(N-octyl-4-vinyl pyridinium iodide), PNO4VP, which is a more hydrophobic polyelectrolyte than PDADMA.17 The corresponding PNO4VP/PAZO multilayer was also more hydrophobic (it was found to contain 50% less water using FTIR18 measurements) and the ion pairs may be presumed to be more tightly bound in comparison with PDADMA/PAZO. The apparent pKa was likewise shifted to even lower values (as shown in Figure 1), with ∆Go(0.1MNaCl) ) -8.7 kJ mol-1. Dissociation of Pol-H+ within the multilayer is represented by
Pol-H+m + Pol+Cl-m h Pol+Pol-m + H+aq + Cl-aq (4a) where Pol+Pol-m represents an ion paired unit of opposite polyelectrolyte charges. The “m” subscripts refer to multilayer phase. The corresponding acid dissociation constant is
Ka(m) )
RPol+Polm-[Haq+][Claq-] RPol-Hm+RPol+Clm-
(4b)
Multilayer quantities have been written as fractions, R, existing in a particular form. Equation 4b shows that Cl- participates in the equilibrium. In all cases, the experiments have been performed with a defined salt concentration. The salt concentration has been chosen to be much greater than [H+aq], so that the ionic strength is essentially constant. Salt perturbs the overall balance of Pol+Cl-, Pol+Pol-, etc. A PEMU swells when immersed in a solution containing salt, according to the following19,20
Pol+Pol-m + Na+aq + Cl-aq h Pol+Cl-m + Pol-Na+m (5a) K1 )
RPol+Clm- RPol-Nam+ RPol+Polm- [Naaq+][Claq-]
(5b)
As more salt is added, more segments are charge compensated by salt counterions, and the population of polyelectrolyte ion pairs decreases. Since a deprotonated polymer segment is now less likely to become an energetically favorable polymer/polymer ion pair, the acidity of the polyacid decreases and there is a positive shift in the apparent pKa. When the salt concentration of a solution of PAZO at pH ) 3.3 (pH ) pKa(aq)) changes, λmax is constant at 352 nm, i.e., R is roughly constant (at about 0.5), an observation which does not contravene many decades of conventional wisdom concerning solution dissociation (only slight (17) Rmaile, H. H.; Schlenoff, J. B. Polym. Mater. Sci. Eng. 2001, 84, 678. (18) Farhat, T.; Yassin, G.; Dubas, S. T.; Schlenoff, J. B Langmuir 1999, 15, 6621. (19) Farhat, T. R.; Schlenoff, J. B. Langmuir 2001, 17, 1184. (20) Dubas, S. T.; Schlenoff, J. B. Langmuir 2001, 17, 7725.
Letters
Langmuir, Vol. 18, No. 22, 2002 8265
Figure 2. λmax vs salt concentration for the PAZO/PDADMA multilayer from Figure 1. pH was constant at 2.2 (squares, 2.2) pKa(m)app for 0.1 M NaCl) or 4.45 (triangles). The solid line is given by eqs 7 and 8 using ∆λD from Figure 1.
changes in Ka(aq) are expected due to activity coefficient variations). As shown in Figure 2, an increase in salt concentration for a multilayer immersed at pH ) pKa(m)app (for 0.1 M NaCl) leads to a shift in λmax toward values corresponding to dissociated polyelectrolyte. The polyacid within the multilayer effectively becomes less acidic, and the titration curve shifts toward higher pKa values as shown by the dotted line in Figure 1. Also shown is the same experiment with the PAZO/PDADMA multilayer poised at pH 4.45. No measurable shift in λmax is observed because the system is already at the flat (unresponsive), fully dissociated (R ) 1) end of the titration curve. The difference between solution phase and multilayer phase protonation of weak bases is that the latter is coupled to salt ions. When a proton enters the multilayer, it must (a) be accompanied by a chloride ion or (b) exchange with a sodium ion already present.21 The salt ion population within the PEMU is coupled to solution salt via eq 5. Combining eqs 4b and 5b +
[H ] )
Ka(m)K1RPol-H+m[Na+]aq RPol-Na+m
(6)
The fraction of acid protonated in the multilayer (1 - Rm) is RPol-H+m. Unprotonated forms exist as Pol-Na+ and Pol-Pol+, but in the limit of high swelling RPol-Pol+m f 0 and RPol-Na+m f Rm. The upper limit for [NaCl] is defined by the salt concentration at which all polymer/polymer contacts are dissociated (RPol-Pol+m ) 0) and the multilayer decomposes, which occurs at [NaCl] g 0.65 M. To relate wavelength to salt concentration, we make use of
λmax ) λo - ∆λD(Rm)
(7)
where ∆λD is the wavelength span between fully dissociated multilayer PAZO and fully protonated PAZO. Rearranging eq 6
Rm )
Ka(m)app -pH
10
+ Ka(m)app
(8)
where Ka(m)app is the apparent pKa for the multilayer. Ka(m)app ) Ka(m)K1[Na+]. As seen in Figure 2, the data adhere
reasonably well22 to theory for ∆λD ) 12.2 nm, λo ) 368.2, and Ka(m)K1 ) 0.034. The deviation of the experimental λmax vs [NaCl]aq curve from the theory at low salt concentration is due to the breakdown of the assumption that all unprotonated polymer exists as Pol-Na+(m) (the high-salt limit). Also, at the low salt limit one does not see the predicted full Pol-H+ dissociation because not all the polyelectrolyte ion pairs are able to formsallowing some Pol- to remain protonated. Segments may be kinetically trapped, due to extremely slow interdiffusion of polyelectrolyte at low salt concentration. This phenomenon has been observed previously with time-resolved in situ measurements of PEMU surface roughness.20 The self-consistent arguments and equations employed above may be used to address some important semantics questions concerning the “identity” of ions within PEMUs. Salt ions (or protons), forced into the multilayer under the influence of solution chemical potential, are associated with charged polymer segments. These salt ions may exchange with other charged solution species (a “reluctant” exchange mechanism19). There are no “extra” ions roaming around the PEMU bulk that are not balanced by polymer segments, as is possible in bulk aqueous solution.23 This is due to excluded volume restrictions. An as-made multilayer, containing no salt ions, is essentially 100% physically cross-linked through multiple polymer ion pairs. Salt ions are highly hydrated and incur strong polymer entropy penalties when they enter a multilayer.25 Polymer/ polymer ion pairs are broken to accommodate salt. If salt ions in balanced numbers were to enter the multilayer oblivious to polymer charge, they would not influence the state of protonation as in Figure 2 and eq 6. Additional interactions, such as enthalpic ones between hydrophobicsor entropic ones between multiply chargeds ions and PEMUs, shift the swelling equilibrium (eq 5) to the right. Due to the requirement of net charge neutrality within multilayers, or any reluctant exchanger, equilibria between charged species are coupled: a cation cannot enter the multilayer without an anion or creation of a negative charge on one of the polymers. Coupling between external salt and internal pKa is an intriguing phenomenon with implications beyond multilayers. The requirements for coupling are simply ion pairing between immobilized opposite charges, where one of the charges is also a weak acid/base conjugate. Any membrane, including a biological one, having paired immobilized charges could exhibit salt-controlled dissociation constants. Acknowledgment. This work was supported by a grant from the National Science Foundation (DMR-9727717). LA025624S +
H (21) The product K1Ka(m) gives the selectivity coefficient KNa + for the H+ ion exchange reaction Pol-H+aq + Na+aq ) Pol-Na+m + H+aq, KNa + ) RPol-Na+m [Haq+]/RPol-H+m[Naaq+]. (22) No dependence of pKa(m)app on the identity of the last layer was observed, in contrast to the results of Klitzing and Mo¨hwald (ref 9) and Xie and Granick (ref 8). PEMUs with 40 layers showed the same pH response as those with 41. This is probably because our PEMUs employed many more layers, so surface effects were diminished. (23) Ions within PEMUs are, of course, still free to move under concentration gradient. See refs 19 and 24. (24) Harris, J. J.; Bruening, M. L. Langmuir 2000, 16, 2006. (25) Schlenoff, J. B.; Dubas, S. T. Macromolecules 2001, 34, 592.
