Abstract
The presence of water strongly influences structure, dynamics and properties of ion-containing soft matter. Yet, the hydration of such matter is not well understood. Here, we show through a large study of monovalent π-conjugated polyelectrolytes that their reversible hydration, up to several water molecules per ion pair, occurs chiefly at the interface between the ion clusters and the hydrophobic matrix without disrupting ion packing. This establishes the appropriate model to be surface hydration, not the often-assumed internal hydration of the ion clusters. Through detailed analysis of desorption energies and O–H vibrational frequencies, together with OPLS4 and DFT calculations, we have elucidated key binding motifs of the sorbed water. Type-I water, which desorbs below 50 °C, corresponds to hydrogen-bonded water clusters constituting secondary hydration. Type-II water, which typically desorbs over 50–150 °C, corresponds to water bound to the anion under the influence of a proximal cation, or to a cation‒anion pair, at the cluster surface. This constitutes primary hydration. Type-III water, which irreversibly desorbs beyond 150 °C, corresponds to water kinetically trapped between ions. Its amount varies strongly with processing and heat treatment. As a consequence, hygroscopicity—which is the water sorption capacity per ion pair—depends not only on the ions, but also their cluster morphology.
Similar content being viewed by others
Introduction
Sorbed moisture is ubiquitously present in ion-containing soft matter. Polyelectrolytes and ionomers—which both comprise polymer backbones tethered with fixed ion groups counterbalanced by free ions—typically show an ambient molar water sorption capacity of the order of their ion density1,2. Here, fixed ions refer to those that cannot be exchanged from the polymer, unlike free ions. Polymer electrolytes, which are polymers infused with salts, also show similarly large amounts of sorbed moisture3,4,5. Amongst these materials, π-conjugated polyelectrolytes have attracted the most interest6,7,8,9,10,11. Their highly tailorable semiconducting polymer backbones open numerous applications in bioelectronics12, batteries13,14, and semiconductor devices15,16. However, sorbed moisture can complicate these applications. It has been implicated in photo-doping17, operando interfacial doping/dedoping18,19,20, and charge-carrier trapping21,22,23. Thus, it is crucial to understand the hydration of these materials.
At first sight, the hydration of conventional polyelectrolytes appears well understood. Their constituent ions assemble into hydrophilic clusters inside a hydrophobic matrix24,25, which is the case also for conjugated polyelectrolytes26,27. The ‘standard’ view is that the first sorbed water molecules bind to ions in the interior of these ion clusters, especially the cations, due to their large negative enthalpy of hydration, to give solvent-shared ion pairs. This model, first popularized by Zundel and co-workers, may be denoted 'bulk hydration' or ‘internal hydration’28,29. The subsequently sorbed water molecules progressively generate solvent-separated ion pairs, and then a concentrated ionic solution30. This final stage is not in doubt for water-soluble polyelectrolytes, and for strongly water-swellable perfluorosulfonate ionomers, such as Nafion and Nafion-M31,32,33,34.
The degree of hydration can be expressed by the hydration number nw, which is the mole ratio of water molecules to ion formula units. In the case of monovalent polyelectrolytes, which have singly-charged cations balanced by single-charged anions, this is given by the mole ratio of water molecules to ion pairs. In the Zundel model, internal hydration is taken to occur from the onset of hydration, that is from the lowest nw values. Pioneering infrared studies in the 1960s‒70 s found that the frequency of the water O–H stretching vibration (νOH) and nw depend on both the cation and anion28,29. These features suggest a strong interaction between sorbed water and both ions from the onset of hydration. Subsequent thermal studies in the 1980s–90s apparently supported this interpretation35,36,37. The first sorbed water molecules are non-freezable, which suggests ‘bound water’ tightly held between the ions. Only when nw exceeds a threshold, which is about seven for poly(styrenesulfonate sodium) (PSSNa) and Nafion-Na, does the ‘excess’ water molecules become freezable or ‘free’.
Yet, a number of puzzling features have been uncovered recently. Neutron reflectometry of polyelectrolyte multilayer films found the first sorbed water molecules to fill voids to give so-called “void water”, while the latter sorbed molecules give “swelling water”38,39. Unless the interior of the ion clusters comprises significant voids, this would appear incompatible with internal hydration. Dielectric relaxation spectroscopy of Nafion-Na found the first sorbed water molecules to be more not less orientationally mobile than the latter sorbed molecules40. Again, this feature is incongruous with the internal hydration model.
Nevertheless, internal hydration clearly occurs in ionic liquids41,42,43. At a low nw ≲ 0.2, water is dispersed throughout the ionic liquid as solitary molecules bridging adjacent anions by hydrogen bonding. For hydrophobic ionic liquids, such as 1-hexyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide, hydration is limited to nw ≲ 0.3 even at saturation with water44,45,46. For hydrophilic ionic liquids, on the other hand, such as 1-butyl-3-methylimidazolium perchlorate, nw can be as large as several water molecules per ion pair for a relative humidity (RH) of 65‒70% at room temperature, giving solvent-shared and solvent-separated ion pairs47,48,49,50,51.
Whether monovalent polyelectrolytes hydrated up to the moderately hydrated regime of several water molecules per ion pair also show similar interior hydration is unclear. This question is scientifically and technologically important, because it relates to hygroscopicity control and hydration structure of the material in ambient air (vide infra). We became interested in this question because of its relevance to several phenomena we were studying, including small ion clusters acting as universal electron donors16, stability of ultrahigh-workfunction hole-doped polymers15,52, and efficiency of the nitrene photocrosslinking mechanism53,54,55.
In this work, we report an extensive study of monovalent conjugated polyelectrolytes leading to the establishment of the surface hydration model as the dominant one for monovalent polyelectrolytes up to the moderately hydrated regime. We focussed on monovalent systems because of their chemical simplicity and technological relevance. We surveyed ions across the Hofmeister series42,56,57: cations in the set of {Li+, Na+, K+, Cs+, TMA+, TEA+, TPP+; –NMe3+}, where TMA+ is tetramethylammonium, TEA+ is tetraethylammonium, TPP+ is tetraphenylphosphonium, and –NMe3+ is tethered trimethylammonium; and anions in the set of {TfO–, TFSI–, BArF–; –CH2SO3–, –CH2SO2N–SO2CF3, –CH2SO2N–SO2CF2CF3}, where TfO– is triflate (CF3SO3–); TFSI– is bis(trifluoromethanesulfonyl)imide ((CF3SO2)2N–); BArF– is B(C6H3(m-CF3)2)4–, and –CH2SO2N–SO2RF, where RF ∈ {CF3, CF2CF3}, is tethered perfluoroalkylsulfonylimidosulfonyl (denoted ‘RFSIS’ here). Most of these polyelectrolytes were synthesized in our laboratories. We shall first discuss the experimental evaluation of hygroscopicity and water desorption energies, then theoretical insights gained from OPLS4 molecular force-field simulations and DFT quantum chemical calculations, and finally the synthesis of all results to establish new fundamental insights into the initial hydration structure of monovalent polyelectrolytes.
Results and discussion
Thermogravimetry (TG)
TG is the simplest technique to probe water desorption. But it suffers from poor resolution. For example, a set of thermograms for mTFF-SO3-Na powder, a key model of sulfonate conjugated polyelectrolytes, is shown in Fig. 1a. The chemical structure of this polymer is shown in S/N 3, Supplementary Table 1, and computed ion cluster morphology in Supplementary Fig. 1, which resembles those of other conjugated polyelectrolytes26,27. The sample was scanned from room temperature to 150 °C in flowing nitrogen, at a temperature ramp (dT/dt) of between 1 and 10 °C min–1, cooled to room temperature (22 °C) in nitrogen, exposed to ambient air for 1 h (RH, 65–70%), re-scanned at a different dT/dt, and the cycle repeated a few more times. All the thermograms show that water—confirmed by tandem mass spectrometry—desorbs continually over this temperature range, without splitting into distinct steps. However, the derivative thermograms, i.e., –dm/dT vs T, where m is normalized mass, clearly reveal two regimes, one with divergent behaviour that occurs below 60 °C, and the other with convergent behaviour above this temperature (Fig. 1b). This suggests two fractions of water are involved. The fraction that desorbs over the higher temperature range exhibits a dm/dT characteristic determined by temperature alone, while that at the lower range a more complex behaviour. Nevertheless, similar mass loss was recorded in each run except the first, indicating that hygroscopicity is well-behaved, except the first run when the sample was not fully equilibrated in the ambient.
To resolve these two regimes, we devised a simple ‘segmented TG’ protocol. An isothermal segment is applied to desorb the lower water fraction, then a temperature ramp is applied to desorb the higher fraction(s) in sequence, and a final isothermal segment to fully desorb the highest fraction at a temperature below the polymer degradation temperature. The following conditions provide a good balance between resolution power and time:
-
a.
jump to initial temperature Ti of 35 °C and hold for 10 min;
-
b.
heat at 5 or 10 °C min–1 to a final temperature Tf (230 °C or lower) and hold for 10 min.
The resulting segmented thermogram for the mTFF-SO3-Na powder now exhibits three water loss steps (Fig. 1c). The first one is characterized by nw of 1.2 H2O per ion pair; second one, 1.0 H2O/ion pair, with extrapolated onset and end temperatures of 45 °C and 135 °C, respectively; third one, 0.6 H2O/ion pair, with extrapolated onset of 180 °C. These data are compiled at S/N 3 in Supplementary Table 1, together with sample preparation notes. After the first scan, the sample was cooled in flowing nitrogen, equilibrated in the ambient and re-scanned. The second thermogram turns out similar with nw of 0.9 and 1.0 H2O/ion pair for the first two steps, respectively, but nw < 0.02 for the third step. This indicates irreversible elimination of the binding sites for the highest water fraction.
This three-step desorption behaviour is also found in other polyelectrolytes. For example, when the hydrophilic sulfonate moiety of the polymer is substituted by the hydrophobic pentafluoroethylsulfonylimidosulfonyl moiety, the steps are similar albeit smaller in size (Fig. 1d; see also Supplementary Table 1: S/N 18).
In this way, we have studied 29 conjugated polyelectrolytes, including charge-doped forms of some of them, together with two non-conjugated polyelectrolytes as references (S/N 1 PSSNa, and S/N2 PVB-NMe3-Cl). The ‘universal’ desorption characteristics observed suggest a simple labelling scheme. The water fraction that desorbs below 50 °C into dry nitrogen is labelled type-I, and those requiring heating between 50–150 °C and between 150–230 °C, type-II and type-III water, respectively (Fig. 2a).
Type-I water
This type of water desorbs readily into dry nitrogen, and resorbs readily from humid ambient. Thus, it is partially lost during the TG setup. To quantify its amount accurately, we performed quartz-crystal-microbalance (QCM) measurements of films cast on the microbalance mounted in a humidity chamber. In this way, we found for PSSNa that nw for type-I water is 2.2, 4.9 and 7.3 H2O/ion pair, at a water activity aw (equivalently, RH) of 0.58, 0.85 and 0.975, respectively, all at 22 °C (Supplementary Table 1: S/N 1). This agrees well with literature58. Likewise, mTFF-SO3-Na gives type-I nw of 2.7, 5.2 and 7.3 for the same set of aw values. In fact, nw shows the same scaling behaviour with aw for all polyelectrolytes studied here. It roughly doubles as aw increases from 0.58 to 0.85, and triples as aw increases to 0.975 (Fig. 2b). Thus, the sorption isotherms are concave upwards34,59,60,61,62. The Flory–Huggins fit gives an apparent polymer–water interaction parameter of 1.0 (±0.1).
This simple result suggests that a simple scale can be set up to quantify the hygroscopicity for type-I water. If we choose PSSNa to define an index value of ‘1’ and a hypothetical non-hygroscopic material ‘zero’ on this scale, conjugated polyelectrolytes span a wide range, from 0.25 to more than 1. Hygroscopicity decreases with decreasing M+ charge density28,29,34,59,63, e.g. for TFB-CF3SIS-M: Li+ > Na+ > Cs+ > TMA+. Similarly, it decreases with decreasing X– charge density, e.g. for mTFF-X-Na: –CH2SO3– > –CH2SO2N–SO2CF3 ≈ –CH2SO2N–SO2CF2CF3.
Type-II water
This type of water desorbs into nominally ‘dry’ nitrogen only at elevated temperatures. It can desorb at room temperature into vacuum, provided the vacuum is better than about 10–5 mbar. Thus, type-II water can resorb from dry nitrogen at room temperature, and persist at aw of nominal zero. As aw decreases to this zero, the total nw falls to the type-II value, which sets the background on which the type-I water sorption isotherm builds. Indeed neutron reflectometry found sorbed water in dry polyelectrolyte multilayer films38,64, and X-ray photoelectron spectroscopy found water even under ultrahigh vacuum conditions53. This phenomenon can be described by the Langmuir–Henry dual-mode model65 and the Guggenheim–Anderson–De Boer model66. The amount of type-II water sorbed appears to vary with the type-I hygroscopicity index (Fig. 2b). It also decreases with decreasing cation and anion charge density: Li+ > Na+ > Cs+ ≈ TMA+ ≈ TEA+ ≈ TPP+ (Supplementary Table 1: S/N 9‒14; 16 cf 17), and: –CH2SO3– > –CH2SO2N–SO2CF3 ≈ –CH2SO2N–SO2CF2CF3 (S/N 6 cf 10, 15; 3 cf 16, 18).
A surprising result is that polyelectrolytes with large hydrophobic cation and anion can still host a sizeable amount of type-II water. Examples include tethered trimethylammonium polyelectrolytes with X– ∈ {TfO–, TFSI– and BArF–} (S/N 22–26, 29, 30) and tethered RFSIS polyelectrolytes with M+ ∈ {TMA+, TEA+ and TPP+} (S/N 12‒14, 17). Treating these materials at high temperature, however, suppresses the resorption of type-II water, in some cases, for the largest ions, completely (Fig. 2c). Thus, a fraction of the binding sites for type-II water can be irreversibly eliminated by baking, which is required to turn these polyelectrolytes non-hygroscopic.
Type-III water
This type of water typically shows only irreversible desorption (Fig. 2c). Its amount becomes large when water is present during processing (S/N 3, 5, 7, 8). For example, when TFB-SO3-Na is precipitated from dimethyl sulfoxide solution using water instead of chloroform, the nw for type-III water jumps from 0.12 to 0.66 H2O/ion pair (S/N 5 cf 6). Thus, in most of the work here, we use diethyl ether as the precipitating non-solvent. Even so, tethered trimethylammonium polyelectrolytes can still host large amounts of type-III water (S/N 22–26, 29, 30), whose binding sites can be fully eliminated by heating.
Overall, the results show that polyelectrolytes generally exhibit a diversity of behavior in hydration and a larger hygroscopicity than even hydrophilic ionic liquids. In the moderately hydrated regime, they host at least three distinct sorbed water species. Clear, the sorption sites are associated with the ion clusters, since the polymer backbones themselves do not sorb much water. For example, both poly(3-hexylthiophene) and tetradecyl-substituted poly(bithiophene-alt-thienothiophene) sorb only ca. 0.005 H2O/thiophene ring in the ambient17. Some of these sites, however, can be eliminated by heat treatment.
Desorption activation energy analysis
To identify the water binding sites, we first estimate the water binding energy by desorption kinetics analysis67. Figure 3 presents the results for the mTFF-SO3-Na sample. We evaluated the surviving water fraction α as a function of time t in dry nitrogen at temperature T of 25, 100 and 200 °C to study type-I, -II and -III water, respectively (Fig. 3). To evaluate the apparent kinetic order n, we fitted the data to Arrhenius kinetics: \(-\frac{d\alpha }{{dt}}=A\exp (-\frac{{E}_{{{{{\mathrm{a}}}}}}}{{{{\mbox{k}}}}_{{{{{\mathrm{B}}}}}}T}){\alpha }^{n}\), where A is the pre-exponential factor, \({E}_{{{{{\mathrm{a}}}}}}\) is the activation energy, and kB is the Boltzmann constant. The ln (‒dα/dt) vs ln (α) plots turn out to be linear, giving n ≈ 1.8 for type-I water (Fig. 3a), and 1.0 for both type-II and -III water (Fig. 3b, c). The pseudo second-order kinetics indicates that the desorption rate of type-I water decreases as α decreases, suggesting interaction between the water molecules. The unimolecular kinetics, on the other hand, indicates that the desorption rates of type-II and -III water are constant, suggesting the water molecules bind to quasi-independent sites.
To evaluate A and Ea for type-II and -III water, we thus fitted their segmented TG data to first-order kinetics: \({{{{\mathrm{ln}}}}}\big(-\frac{d\left({{{{\mathrm{ln}}}}}\alpha \right)}{dt}\big)={{{{\mathrm{ln}}}}}A-\frac{{E}_{{{{{\mathrm{a}}}}}}}{{{{\mbox{k}}}}_{{{{{\mathrm{B}}}}}}T}\). This gives Ea of 0.18 and 0.72 eV, respectively. The Ea and A parameters are compiled in Supplementary Table 2. These parameters characterize the desorption–temperature profile. A steeper step gives a larger Ea together with a larger A. Such a correlation is reminiscent of the enthalpy–entropy compensation68. But it may also receive a contribution here from irreversibility. Assuming diffusion through the sample is not rate-limiting, Ea should reflect the water binding energy69,70.
The results reveal that Ea values for reversible desorptions are generally lower than for irreversible ones. For reversible type-II water, the Ea values fall in the range of 0.15‒0.2 eV for tethered sulfonate polyelectrolytes (Supplementary Table 2: S/N 1–5) and 0.2‒0.3 eV for tethered RFSIS polyelectrolytes (S/N 6–9), largely independent of the cation. They are ca. 0.2 eV for tethered trimethylammonium polyelectrolytes (S/N 15, 19). For irreversible water, the Ea values can be considerably larger. For irreversible type-II water, the Ea values are 0.3‒0.4 eV for the usual polyelectrolytes (S/N 11–14, 16, 17) and 0.55‒0.65 eV for those with large hydrophobic ion pairs (S/N 10 and 18). For type-III water, the Ea values lie between 0.35 eV and 0.95 eV. Given the hydrogen-bond (H-bond) binding energy in a water dimer is only 0.15 eV71,72, the results suggest that reversible type-II water is bound with similar or marginally larger strengths, but irreversible water is bound much more strongly, with a strength approaching those of multivalent and transition metal cations73.
Water binding motifs: OPLS4 simulations
To survey the possible binding motifs, we employ the OPLS4 all-atom molecular mechanics force field. This has been well-parametrized for a wide chemical space74,75. It has also been found to have acceptable accuracy for electrostatic and H-bond simulations (Supplementary Table 3). We computed various hydrated ion clusters given by the formula (2 M+ –A‒(CH2)7‒A–)8 (H2O)n, where A– ∈ {SO3–, C2F5SIS–}, M+ ∈ {Na+, TMA+}, and n ∈ {24, 104}, corresponding to nw of 1.5 and 6.5, respectively, with the ‒(CH2)7‒ tether imposing realistic steric constraints to simulate the ion cluster ‘surface’.
Molecular snapshots of the thermally-equilibrated clusters reveal three essential features. First, the cation–anion framework is not disrupted by hydration at the level of one to two water molecules per ion pair. The water molecules do not insert into the ion clusters, but bind to ions at their surfaces. Even at a moderately high nw of 6.5, the cation‒anion framework is still substantially retained, but water intrusion splits the ion cluster into smaller aggregates. See Fig. 4, and Supplementary Video 1, which provides 360°-view of the hydrated clusters. [This tendency, however, may have been amplified by overestimation of water-ion binding energies in OPLS4.] Second, the distribution of water is inhomogeneous. At nw of 1.5, most of the water molecules bind to the ion cluster directly in what we term as “primary hydration”. At nw of 6.5, many of the water molecules bind to other water molecules, in what we term as “secondary hydration”.
Third, the primary hydration water can bind in several motifs via: (i) H-bond to the anion, which we denote α binding (‘α’ for anion); (ii) H-bond between two adjacent anions, β binding (‘β’ for bridging); (iii) electrostatic bond to the cation, χ binding; (iv) dual bond to both cation and anion, δ binding; or (v) combination bond to one cation and two anions, i.e., δ + α binding. This contrasts with the ionic liquid picture, where the water molecules participate primarily in β binding which then elaborates into H-bonded chains and clusters in the bulk of the ionic liquid47,48,49,50.
Thus, OPLS4 indicates a preference for surface rather than internal hydration of the ion clusters in polyelectrolytes. This is attributed to the excess free energy and volume present at such surfaces. Because the sorbed water cannot attain the 3D order necessary for freezing, it becomes non-freezable. Thus, non-freezability alone does not imply strong binding.
Water binding energies: DFT computations
We employ DFT/CAM-B3LYP/6-31 + + G(d,p) to compute the water‒site binding energies and associated vibrational frequencies. This level of DFT theory reproduces the H-bond better than OPLS4. See Supplementary Table 3, and also Supplementary Tables 1-1, 1-2 and 2 in ref. 52. Due to computational costs, however, the calculations are limited to small ion multiplet models. We studied (M+ X–)r (H2O)p, where M+ ∈ {Li+, Na+, K+, TMA+}, X– ∈ {RSO3–, TfO–, (CH3SO2)(CF3SO2)N–}, and r ∈ {1, 2, 4}.
Figure 5 shows selected geometry-optimized structures, while Supplementary Figs. 2, 3 and 4 show the structures for Na+, Li+ and K+, and TMA+, respectively, totalling sixty. The computed water binding energies ΔUo are compiled in Supplementary Table 4. Such values should be regarded as upper limits because the optimal geometry for the gas phase may be frustrated in the condensed phase. Moreover, because small ion multiplets expose cations more strongly, the calculations yield more occurrences of χ, δ and δ + α binding.
The results reveal that δ binding is strong. The ΔUo values are a few tenths of an eV, varying strongly with cation. Representative δ motifs are shown in Fig. 5 (for ion pairs: II b; IV b & d; and ion octets: II i & j; VI i & j; and VII f). The binding energies decrease by 30% as r increases from 1 to 4 for sulfonate clusters with M+ ∈ {Li+, Na+, K+, TMA+}. This decrease stems from an averaging out of the electrostatic field gradient at the binding site, which is a manifestation of the local Madelung potential effect76,77. Since realistic ion cluster sizes are generally larger than four26,27, realistic δ binding energies are estimated to lie in the range of 0.35–0.55 eV.
The α binding energies are smaller. But they vary little with the anions studied. ΔUo is ca. 0.3 eV for sulfonate, and 0.25 eV for both TfO– and (CH3SO2)(CF3SO2)N–. Typical α motifs are given in Fig. 5 (II a) and Supplementary Figs. 2–4 (I a, I e–g, IV a, VI a and IX a). β Binding gives similar ΔUo values as α binding, not twice as large as might have been naively expected. Typical motifs are shown in V g, V i and V j. Finally, χ binding varies strongly with cation: 0.45 eV for Li+, 0.65 eV for Na+, and 0.15 eV for TMA+. Typical motifs are shown in I b, III c, V h and VII g.
The electrostatic interaction with cation can be modelled as a water–cation dipole interaction. Such an effect persists beyond physical contact of the water molecule with the cation. To show this, we employ PM3 to compute ΔUo for δ binding in a variety of cation‒anion pairs (Fig. 6), and find: ΔUo = A + B * d–2, where d is the water oxygen-to-cation distance, A is the limiting binding energy (0.25 eV for CH3SO3–, and 0.20 eV for both TfO– and TFSI–), and B is an empirical constant (2.5 eV Å2). The form of this equation is characteristic of the dipole–ion interaction. It predicts the attraction to remain sizeable (≳0.1 eV) for distances up to 5 Å. This implies a proximal cation not in direct contact with the water molecule can still contribute to electrostatic bonding that reinforces its α binding.
In summary, the computed results here provide the basis to assign binding motifs to the measured Ea values. Values in the range of 0.15–0.3 eV are consistent with α and β bindings, possibly reinforced with electrostatic bonding to a proximal small cation or to a contacting large cation. Values in the range of 0.3–0.6 eV are consistent with χ, δ and δ + α bindings; but values larger than 0.6 eV must involve caging effects. Thus, reversible type-II water is chiefly H-bonded to anions, but irreversible type-II water involves strong contribution also from the cation, while type-III water involves additional caging effects.
Hydrogen-bond character: computed frequencies
To confirm these assignments, we compute the νOH wavenumbers and compare them with the observed wavenumbers (Supplementary Table 4)78. The νOH vibration is highly sensitive to binding motif of the water molecule, in particular, the amount of H-bond donor character (D) and H-bond acceptor character (A) at the O–H group of the motif79,80,81. To extract association insights, we first generate a correlation chart between νOH wavenumber and H-bond character, based on literature calculations of water clusters (Fig. 7)79. The chart shows that νOH of water shifts systematically with A and D character, as is well known79,80,82,83,84.
Then we added to this chart our calculated results for the hydrated ion multiplets. The augmented chart shows that νOH of a water molecule changes little between H-bonding (as donor) to another water or to sulfonate (black cf red bands). This has been confirmed by measurements in ionic liquids44,46. An additional electrostatic bonding to M+, however, behaves like adding weak A character to the water molecule, redshifting its νOH band. The amount of redshift varies little with the size of M+. Thus, α binding under the influence of a proximal small cation produces a similar spectroscopic signature as δ binding, corresponding to DA with weak A character. Similarly, β binding under the influence of a proximal small cation produces a similar spectroscopic signature as δ + α binding, corresponding to DDA with weak A character. In this way, the observed νOH wavenumber can be interpreted to support or exclude certain binding motifs.
Hydrogen-bond character: experiments
To realize this analysis for mTFF-SO3-Na, we collected in situ FTIR spectra of its film in the following sequence: ambient air (where nw ≈ 5) → nitrogen (nw ≈ 1) → after 230°C-bake, in nitrogen (nw ≈ 1) → after short exposure in ambient air (nw ≈ 2.5). The spectra in ambient air contain both type-I and -II water, while that in nitrogen only type-II water. These spectra and their difference (Fig. 8a, b, respectively) show that hydration intensifies the νOH envelope at 3460 cm–1, and redshifts the SO3 bands, both νs (symmetric) and νa (anti-symmetric), as expected85,86. The νOH envelope is considerably blue-shifted from bulk water87.
We modelled this envelope with a sum-of-Gaussians: \(I={\sum }_{{{{{{\rm{i}}}}}}}{I}_{{{{{{\rm{i}}}}}}}{{\exp }}\left(-\tfrac{{\left(v-{v}_{{{{{{\rm{i}}}}}}}\right)}^{2}}{2{\sigma }_{{{{{{\rm{i}}}}}}}^{2}}\right)\), where νi is the wavenumber and σi is the standard deviation, of the i-th component (Fig. 8c). The component basis set employed corresponds approximately to the following H-bond character (in cm–1): (i) 3200 – DAA; (ii) 3300 – DAA, and DA with strong A; (iii) 3420 – DAA, DDAA, and DA with weak A; (iv) 3520 – DDA, and D; and (v) 3630 – DD, and free OH; where “weak A” captures electrostatic interaction with M+, and “free OH” refers to that in the condensed phase, not vacuum. The 3630-cm−1 component is significantly narrower than the others (σi, 38 vs 60 cm−1), so it corresponds indeed to free OH.
Type-I water yields very weak intensity in the 3200-cm−1 component (spectra i and iii). This indicates that DAA character is nearly absent, which then also rules out other extensively H-bonded water species. Thus, type-I water comprises small H-bonded water chains and clusters with DA, DD, D and/or A character, together with free OH. Type-II water yields strong intensities in the 3420; 3520 and 3630-cm−1 components (spectrum ii). This indicates that DA (weak A) character, and possibly D, is present together with free OH, which is consistent with α binding under the influence of proximal cation and/or δ binding. We can rule out any significant β binding, or δ + α binding, as these show two sharp bands over the 3525–3675 cm−1 spectral region44,46. Since Ea analysis has ruled out δ binding for mTFF-SO3-Na, the dominant motif for its type-II water is α binding under the influence of proximal cation.
To estimate population size, we collected in situ Raman spectra. Raman cross sections vary much less strongly with H-bond character than infrared cross sections88,89. Thus, Raman spectroscopy is more suitable for quantification. The observed Raman spectral components are broadly consistent with the infrared components (Fig. 8d). Taking into account the approximate Raman cross sections—equal for all H-bonded OH and halved for free OH88,89—we deduce:
-
a.
ratio of type-I to type-II water is roughly five, which agrees with QCM measurements;
-
b.
type-I water has a bonded OH-to-free OH ratio of roughly 2:1, which confirms small H-bonded water chains or clusters90,91,92;
-
c.
type-II water has a bonded OH-to-free OH ratio of roughly 1:1, which agrees with the assigned α binding.
Summary analysis of a hydrophobic polyelectrolyte family
We applied a similar analysis to the TFB-CF3SIS-M family of polyelectrolytes. Their hygroscopicities are lower than mTFF-SO3-Na, and decrease with increasing size of M+ ∈ {Li+, Na+, Cs+, TMA+, TEA+} (Fig. 9a). Type-II water (red dotted line) is dominated by three components at 3350, 3500 and 3630 cm−1, largely independent of M+ (Fig. 9b). This indicates predominant α binding with D and/or DA (weak A) character, together with free OH. H-bonding of water to CF3SIS results in a smaller redshift of its νOH by 50–100 cm−1 compared with sulfonate (Supplementary Table 4; see also refs. 44,46). On the other hand, type-I water is dominated by components at 3520 and 3630 cm–1, which indicates D character and higher free OH content than mTFF-SO3-Na. Thus, the type-I water in TFB-CF3SIS-M occurs in even smaller H-bonded clusters. Its desorption shifts νs SO2 bands (1060 and 1200 cm−1) towards lower wavenumbers for M+ ∈ {Li+, Na+} but higher wavenumbers for {TMA+, TEA+}. Again, type-I water influences anion vibrations, mediated perhaps by type-II water, similar to the case of sulfonate polyelectrolytes.
In summary, the reversible type-II water in both hydrophilic and hydrophobic polyelectrolyte models here comprises water molecules that are chiefly H-bonded to the anion with weak electrostatic bonding to the cation. On the other hand, type-I water comprises small H-bonded chains and clusters attached to type-II water. Thus, the pronounced effect of cations on hygroscopicity is attributed to proximal effect, and influence via anion, packing and cluster morphology, rather than direct bonding to water.
Generalization: internal vs surface hydration models
The internal hydration and surface hydration models are illustrated schematically in Fig. 10a, b, respectively, to underscore fundamental differences in their nature and binding of water. Evidences from desorption activation energy analysis and νOH vibration analysis support the surface hydration model for monovalent polyelectrolytes in the moderate hydration regime of up to several water molecules per ion pair. Reversible type-II water is substantially held by α binding to the anion at the ion cluster surface, reinforced by electrostatic bonding to proximal small cation or to contacting large cation, while keeping the ionic framework substantially intact. This is undoubtedly driven by electrostatic stabilization of the ion cluster core. Despite its greater strength, δ binding to anion‒small cation pair appears to be generally suppressed, even at the ion cluster surface, despite OPLS4 predictions. This suggests recession of the small cation away from the surface and other local geometry effects that frustrate dual binding. The model explains the existence of “void water” and its orientational mobility. At higher hydration levels, water may intrude into the ion cluster to ‘carve out’ new surfaces. Type-I water is characterized by small H-bonded chains or clusters, not an extensively H-bonded network. On the other hand, irreversible type-II water is likely held by δ binding, possibly in a weak cage, also at the surface of the ion cluster. This becomes particularly prominent in polyelectrolytes with large hydrophobic cations. Type-III water is held by δ binding, often in a strong cage, likely in the interior of the ion cluster, due to trapping of molecular water during material deposition.
A key consequence of surface hydration of the ion clusters is that hygroscopicity, and hence its control, depends not only on hydrophilicity of the ion pairs, but also morphology of the ion clusters. The latter is determined by relative sizes of the ions, polymer backbone structure, ion attachment site, ion density, etc. There is evidence for this from the variation in hygroscopicity with different polymer backbone structures for the same ion pair (Supplementary Table 1). Another key consequence is that the hydrated complexes can be described as small H-bonded water clusters attached to the local ion packing. The water substantially retains its own properties—for example, ionization energetics, chemical reactivity—albeit perturbed by the interaction. This picture has enabled the influence of sorbed moisture on several phenomena to be successfully modelled, including: dehydration-induced electron transfer from small ion clusters16, hydration-induced chemical trapping of holes from ultrahigh-workfunction polymer semiconductors15,52, and hydration-induced interference of nitrene photocrosslinking mechanism54,55.
Finally, we summarize here the hygroscopicity trends for charge-doped polymers. Charge-doping introduces mobile carriers on the polymer backbone, which are balanced by counter-ions, whether free or tethered. Similar to their undoped counterparts, charge-doped polyelectrolytes also show both type-I and -II water, with similar hygroscopicity (Supplementary Table 1: S/N 4 cf 3; 19 cf 18; 21 cf 20; 27 cf 26). But charge-doped polymers that are counterbalanced by particularly large hydrophobic ions can show considerably lower hygroscopicity (S/N 31 and 32) than those counterbalanced by presently available tethered ions. This suggests further reduction in the hygroscopicity of charge-doped polyelectrolytes should be possible with better design of the tethered counterion.
Methods
Materials
Polyelectrolytes were synthesized and ion-exchanged following well-established protocols in our laboratory15,26,27. PSSNa and PVBTMA-Cl were commercial products used as received. Where required, films or solutions were doped following well-established protocols in our laboratory, as indicated in Supplementary Table 115,26,27. All polyelectrolytes were dried at 80 °C overnight in a dynamic vacuum (10−2 mbar), and stored in a nitrogen glovebox.
Thermogravimetry and quartz crystal microbalance
Thermogravimetry was performed on Discovery TGA 1 (TA Instruments). Samples in powder form (4‒6 mg) were loaded in the ambient (22 °C, 65‒70% RH) onto the sample pan which was then moved into the nitrogen-purged furnace under computer control. Typical scan parameters: isothermal hold at 35 oC for 10 min, temperature ramp at 5 or 10 oC min−1 to 230 oC, isothermal hold for 10 min, temperature ramp at –5 oC min−1 to 35 oC, all in flowing nitrogen (>99.999% purity). The samples were typically exposed to the ambient for 1‒2 h before repeating the scan. Quartz crystal microbalance measurements were performed on QCM200 Quartz Crystal Microbalance (Stanford Research Systems) at room temperature with a 5 MHz AT-cut quartz crystal. Films were spin-cast onto the quartz crystal and mounted inside a gas flow cell. Resonant frequencies were measured in dry nitrogen (fo) and in controlled-humidity nitrogen (f). The mass of water sorbed by the film (Δm) was computed by the Sauerbrey equation: Δf = –Cf × Δm, where Cf is the crystal sensitivity factor (56.6 Hz µg–1 cm2) and Δf is the change in frequency given by f–fo. The mass of film was estimated from its thickness, measured by profilometry, and density, measured by flotation method.
FTIR and Raman spectroscopies
FTIR spectra were collected on a nitrogen-purged Nicolet 8700 FTIR spectrometer operated in the temperature-stabilized cleanroom to provide measurement stability of better than 1 mAU in the transmission mode. Films were cast onto intrinsic silicon substrates and mounted inside an optical access flow cell fitted with KBr windows. Unpolarised Raman spectra were collected using a Renishaw Raman microscope in the back-scattered geometry. Films were cast onto intrinsic silicon substrates and mounted on a Linkam hot stage inside an optical access flow cell fitted with a quartz window. Ar ion laser (514 nm) was focused through a long working distance objective lens (numerical aperture 0.55, 50×) for excitation.
Molecular mechanics and quantum chemical calculations
OPLS4 calculations were performed in MacroModel (Schrödinger LLC). The hydrated ion clusters were generated haphazardly and equilibrated for 4 ns at 300 K, with background dielectric constant set to 1. Several snapshots were taken from the production run to check that the results are representative. DFT calculations were performed at the CAM-B3LYP/6-31 + + G(d,p) level in Gaussian 09. The parent ion multiplet (M+ X−)r was first generated and geometry optimized. Then the desired number of water molecules were brought haphazardly to its surface to give a hydrated seed configuration, which was geometry re-optimized to give the local-minimum (M+ X−)r (H2O)p hydrated complex. Vibrational frequencies were computed. Then the water molecules were removed, and the geometry of the remaining (M+ X−)r re-optimized to give the corresponding dry complex. The adiabatic water detachment internal energy ΔU0 was evaluated as the difference in internal energies between the dry and hydrated complexes at 0 K, corrected for zero-point energy. Multiple hydrated seed configurations were generated to investigate the diversity of hydrated complexes.
Data availability
Source data for all figures are available from the corresponding author.
References
Nalwa, H. S. (ed.) Handbook of Polyelectrolytes and Their Applications. (American Scientific Publishers, 2002).
Rembaum, A. & Sélégny, E. (eds.) Polyelectrolytes and their Applications. (Springer, 2012).
Quatarone, E. & Mustarelli, P. Electrolytes for solid-state lithium rechargeable batteries: recent advances and perspectives. Chem. Soc. Rev. 40, 2525–2540 (2011).
Choudhury, S. et al. Solid-state polymer electrolytes for high-performance lithium metal batteries. Nat. Commun. 10, 4398 (2019).
Baskoro, F., Wong, H. Q. & Yen, H. J. Strategic structural design of a gel polymer electrolyte toward a high efficiency lithium-ion battery. ACS Appl. Energy Mater. 2, 3937–3971 (2019).
Shi, W. et al. Synthesis of novel triphenylamine-based conjugated polyelectrolytes and their application as hole-transport layers in polymeric light-emitting diodes. J. Mater. Chem. 16, 2387–2394 (2006).
Jin, Y., Bazan, G. C., Heeger, A. J., Kim, J. Y. & Lee, K. Improved electron injection in polymer light-emitting diodes using anionic conjugated polyelectrolyte. Appl. Phys. Lett. 93, 346 (2008).
Hoven, C. V., Garcia, A., Bazan, G. C. & Nguyen, T. Q. Recent applications of conjugated polyelectrolytes in optoelectronic devices. Adv. Mater. 20, 3793–3810 (2008).
Huang, F., Wu, H. & Cao, Y. Water/ alcohol soluble conjugated polymers as highly efficient electron transporting/ injection layer in optoelctronic layers. Chem. Soc. Rev. 39, 2500–2521 (2010).
Liu, B. & Bazan, G. C. (eds.) Conjugated Polyelectrolytes: Fundamentals and Applications. (Wiley-VCH, Weinheim, 2013).
Lee, J. H. et al. Radical cation-anion coupling-induced work function tunability in anionic conjugated polyelectrolytes. Adv. Energy Mater. 5, 1501292 (2015).
Groenendaal, L. B., Jonas, F., Freitag, D., Pielartzik, H. & Reynolds, J. R. Poly(3,4-ethylenedioxythiophene) and Its derivatives: past, present, and future. Adv. Mater. 12, 481–494 (2000).
Gracia, R. & Mecerreyes, D. Polymers with redox properties: materials for batteries, biosensors and more. Polym. Chem. 4, 2206–2214 (2013).
Gaikwad, A. M., Arias, A. C. & Steingart, D. A. Recent progress on printed flexible batteries: mechanical challenges, printing technologies, and future prospects. Energy Technol. 3, 305–328 (2015).
Tang, C. G. et al. Doped polymer semiconductors with ultrahigh and ultralow work functions for ohmic contacts. Nature 539, 536–540 (2016).
Tang, C. G. et al. Multivalent anions as universal latent electron donors. Nature 573, 519–525 (2019).
Zhuo, J. M. et al. Direct spectroscopic evidence for a photodoping mechanism in polythiophene and poly(bithiophene-alt-thienothiophene) organic semiconductor thin films involving oxygen and sorbed moisture. Adv. Mater. 21, 4747–4752 (2009).
Chia, P. J. et al. Injection-induced de-doping in a conducting polymer during device operation: asymmetry in thoe hole injection and extraction rates. Adv. Mater. 19, 4202–4207 (2007).
Belaineh, D. et al. A high-performance p-doped conducting polymer blend based on sulfonated polyalkoxythiophene and poly (4-hydroxystyrene). Chem. Mater. 26, 4724–4730 (2014).
Zhao, C. et al. Bias-induced electrochemical electron doping of organic semiconductor contacts. Adv. Mater. Interfaces 16, 1900607 (2019).
Nicolai, H. T. et al. Unification of trap-limited electron transport in semiconducting polymers. Nat. Mater. 11, 882–887 (2012).
Nikolka, M. et al. High operational an environmental stability of high-mobility conjugated polymer field-effect transistors through the use of molecular additives. Nat. Mater. 16, 356–362 (2017).
Kotadiya, N. B., Mondal, A., Blom, P. W. M., Andrienko, D. & Wetzelaer, G. A. H. A window to trap-free charge transport in organic semiconductor thin films. Nat. Mater. 18, 1182–1185 (2019).
Eisenberg, A. & King, M. Supermolecular Structure and Glass Transitions. (Academic Press, 1977).
Hsu, W. Y. & Gierke, T. D. Ion transport and clustering in Nafion perfluorinated membranes. J. Membr. Sci. 13, 307–326 (1983).
Ang, M. C. Y. et al. Bulk ion-clustering and surface ion-layering effects on work function of self-compensated charge-doped polymer semiconductors. Mater. Horiz. 7, 1073–1082 (2020).
Ang, M. C. Y. et al. Spectator cation size effect on the work function and stability of self-compensated hole-doped polymers. J. Mater. Chem. C 8, 124–131 (2020).
Zundel, G. Hydration and Intermolecular Interaction - Ir Investigations of Polyelectrolyte Membranes. (Academic Press, 1969).
Zundel, G. Hydrate structures, intermolecular interactions and proton conducting mechanism in polyelectrolyte membranes - infrared results. J. Membr. Sci. 11, 249–274 (1982).
van der Vegt, N. F. A. et al. Water-mediated ion pairing: occurrence and relevance. Chem. Rev. 116, 7626–7641 (2016).
Laporta, M., Pegoraro, M. & Zanderighi, L. Perfluorosulfonated membrane (Nafion): FT-IR study of the state of water with increasing humidity. Phys. Chem. Chem. Phys. 1, 4619–4628 (1999).
Mauritz, K. A. & Moore, R. B. State of understanding of Nafion. Chem. Rev. 104, 4535–4586 (2004).
Garrido, L. et al. Influence of the water content on the diffusion coefficients of Li+ and water across naphthalenic based copolyimide cation-exchange membranes. J. Phys. Chem. B 116, 11754–11766 (2012).
Peng, J., Lou, K., Goenaga, G. & Zawodzinski, T. Transport properties of perfluorosulfonate membranes ion exchanged with cations. ACS Appl. Mater. Interfaces 10, 38418–38430 (2018).
Hatakeyama, T., Nakamura, K., Yoshida, H. & Hatakeyama, H. Phase transition on the water-sodium poly(styrenesulfonate) system. Thermochim. Acta 88, 223–228 (1985).
Yoshida, H. & Miura, Y. Behavior of water in perfluorinated ionomer membranes containing various monovalent cations. J. Membrane Sci. 68, 1–10 (1992).
Hatakeyama, T., Hatakeyama, H. & Nakamura, K. Non-freezing water content of mono- and divalent cation salts of polyelectrolyte-water systems studied by DSC. Thermochim. Acta 253, 137–148 (1995).
Dodoo, S., Steitz, R., Laschewsky, A., Von & Klitzing, R. Effect of ionic strength and type of ions on the structure of water swollen polyelectrolyte multilayers. Phys. Chem. Chem. Phys. 13, 10318–10325 (2011).
Koehler, R., Steitz, R. & von Klitzing, R. About different types of water in swollen polyelectrolyte multilayers. Adv. Colloid Interf. Sci. 207, 325–331 (2014).
Ensing, W., Hunger, J., Ottosson, N. & Bakker, H. J. On the orientational mobility of water molecules in proton and sodium terminated Nafion membranes. J. Phys. Chem. C 117, 12930–12935 (2013).
Huddleston, J. G. et al. Characterization and comparison of hydrophilic and hydrophobic room temperature ionic liquids incorporating the imidazolium cation. Green Chem. 3, 156–164 (2001).
Zhou, T. et al. An overview of mutual solubility of ionic liquids and water predicted by COSMO-RS. Ind. Eng. Chem. Res. 51, 6256–6264 (2012).
Królikowska, M. Solid + liquid) and (liquid + liquid) phase equilibria of (IL + water) binary systems. The influence of the ionic liquid structure on mutual solubility. Fluid Phase Equilibria 361, 273–281 (2014).
Cammarata, L., Kazarian, S. G., Salter, P. A. & Welton, T. Molecular states of water in room temperature ionic liquids. Phys. Chem. Chem. Phys. 3, 5192–5200 (2001).
Masaki, T., Nishikawa, K. & Shirota, H. Microscopic study of ionic liquid-H2O systems: alkyl-group dependence of1-alkyl-3-methylimidazolium cation. J. Phys. Chem. B 114, 6323–6331 (2010).
Mukherjee, K., Palchowdhury, S. & Maroncelli, M. OH stretching and libration bands of solitary water in ionic liquids and dipolar solvents share a single dependence on solvent polarity. J. Phys. Chem. B 126, 4584–4598 (2022).
Schröder, C., Rudas, T., Neumayr, G., Benkner, S. & Steinhauser, O. On the collective network of ionic liquid/water mixtures. I. Orientational structure. J. Chem. Phys. 127, 234503 (2007).
Moreno, M., Castiglione, F., Mele, A., Pasqui, C. & Raos, G. Interaction of water with the model ionic liquid [bmim][BF4]: molecular dynamics simulations and comparison with NMR data. J. Phys. Chem. B 112, 7826–7836 (2008).
Zhong, X., Fan, Z., Liu, Z. & Cao, D. Local structure evolution and its connection to thermodynamic and transport properties of 1-butyl-3-methylimidazolium tetrafluoroborate and water mixtures by molecular dynamics simulations. J. Phys. Chem. B 116, 3249–3263 (2012).
Niazi, A. A., Rabideau, B. D. & Ismail, A. E. Effects of water concentration on the structural and diffusion properties of imidazolium-based ionic liquid−water mixtures. J. Phys. Chem. B 117, 1378–1388 (2013).
Spieβ, A., Brandt, P., Schmitz, A. & Janiak, C. Water sorption by ionic liquids: evidence of a diffusion-controlled sorption process derived from the case study of [BMIm][OAc]. J. Mol. Liq. 348, 118023 (2022).
Koh, Q. M. et al. Overcoming the water oxidative limit for ultra-high-workfunction hole-doped polymers. Nat. Commun. 12, 3345 (2021).
Khong, S. H. et al. General photo-patterning and post-deposition modification of polyelectrolyte thin films via efficient ionic bis(fluorinated phenyl azide) photo-crosslinkers. Adv. Funct. Mater. 17, 2490–2499 (2007).
Teo, D. W. Y. et al. Nearly 100% photocrosslinking efficiency in ultrahigh work function hole-doped conjugated polymers using bis(fluorophenylazide) additives. ACS Appl. Mater. Interfaces 11, 48103–48112 (2019).
Teo, D. W. Y., Jamal, Z., Seah, Q. J., Png, R. Q. & Chua, L. L. General bis(fluorophenyl azide) photo-crosslinkers for conjugated and non-conjugated polyelectrolytes. J. Mater. Chem. C 8, 253–261 (2020).
Mazzini, V. & Craig, V. S. J. What is the fundamental ion-specific series for anions and cations? Ion specificity in standard partial molar volumes of electrolytes and electrostriction in water and non-aqueous solvents. Chem. Sci. 8, 7052–7065 (2017).
Gibb, B. C. Hofmeister’s curse. Nat. Chem. 11, 963–965 (2019).
Toribio, F., Bellat, J. P., Nguyen, P. H. & Dupont, M. Adsorption of water vapor by poly(styrenesulfonic acid), sodium salt: isothermal and isobaric adsorption equilibria. J. Colloid Interface Sci. 280, 315–321 (2004).
Pushpa, K. K., Nandan, D. & Iyer, R. M. Thermodynamics of water sorption by perfluorosulphonate (Nafion-117) and polystyrene-divinylbenzene sulphonate (Dowex 50W) ion-exchange resins at 298±1 K. J. Chem. Soc. Faraday Trans. 1 84, 2047–2056 (1988).
Hirai, Y. & Nakajima, T. Sorption behavior of water wapor into polyelectrolyte complex of poly(acrylic acid)/poly(4-vinylpyridine). J. Appl. Polym. Sci. 37, 2275–2281 (1989).
Köhler, R. et al. Neutron reflectometry study of swelling of polyelectrolyte multilayers in water vapors: influence of charge density of the polycation. Langmuir 25, 11576–11585 (2009).
Ferrell, W. H., Kushner, D. I. & Hickner, M. A. Investigation of polymer–solvent interactions in poly(styrene sulfonate) thin films. J. Polym. Sci. B Polym. Phys. 55, 1365–1372 (2017).
Ortony, J. H. et al. Thermophysical properties of conjugated polyelectrolytes. Adv. Mater. 20, 298–302 (2008).
Ivanova, O. et al. Immobile light water and proton-deuterium exchange in polyelectrolyte multilayers. Macromolecules 41, 7179–7185 (2008).
Okuzaki, H., Kondo, T. & Kunugi, T. Characteristics of water in polypyrrole films. Polymer 40, 995–1000 (1999).
Preda, F. M. et al. Investigation of water diffusion mechanisms in relation to polymer relaxations in polyamides. Macromolecules 48, 5730–5741 (2015).
Vyazovkin, S. et al. ICTAC Kinetics Committee recommendations for collecting experimental thermal analysis data for kinetic computations. Thermochim. Acta 590, 1–23 (2014).
Zheng, J. M., Qiu, J., Madeira, L. M. & Mendes, A. Polymer structure and the compensation effect of the diffusion pre-exponential factor and activation energy of a permeating solute. J. Phys. Chem. 111, 2828–2835 (2007).
van Krevelen, D. W. Properties of Polymers: Their Correlation with Chemical Structure; Their Numerical Estimation and Prediction from Additive Group Contributions. (Elsevier, 1990).
Brandrup, J., Immergut, E. H. & Grulke, E. A. (eds.) Polymer Handbook, 4th edn. (Wiley, 2003).
Rusdi, M. & Moroi, Y. Evaporation rate measurement of water and liquid 1-alkanols across air–liquid interface using thermogravimetry. Bull. Chem. Soc. Jpn. 76, 919–926 (2002).
Kathmann, S. M., Palmer, B. J., Schenter, G. K. & Garrett, B. C. Activation energies and potentials of mean force for water cluster evaporation. J. Chem. Phys. 128, 064306 (2008).
Rosi, M. & Bauschlicher, C. W. The binding energies of one and two water molecules to the first transition-row metal positive ions. J. Chem. Phys. 90, 7264–7272 (1989).
Harder, E. et al. OPLS3: a force field providing broad coverage of drug-like small molecules and proteins. J. Chem. Theory Comput. 12, 281–296 (2016).
Lu, C. et al. OPLS4: improving force field accuracy on challenging regimes of chemical space. J. Chem. Theory Comput. 17, 4291–4300 (2021).
Chia, P. J. et al. Direct evidence for the role of the Madelung potential in determining the work function of doped organic semiconductors. Phys. Rev. Lett. 102, 0966021–0966024 (2009).
Png, R. Q. et al. Madelung and Hubbard interactions in polaron band model of doped organic semiconductors. Nat. Commun. 7, 11948 (2016).
Ben-Amotz, D. Hydration-shell vibrational spectroscopy. J. Am. Chem. Soc. 141, 10569–10580 (2019).
Ohno, K., Okimura, M., Akai, N. & Katsumoto, Y. The effect of cooperative hydrogen bonding on the OH stretching-band shift for water clusters studied by matrix-isolation infrared spectroscopy and density functional theory. Phys. Chem. Chem. Phys. 7, 3005–3014 (2005).
Lenz, A., Ojamäe, L. & Theoretical, I. R. spectra for water clusters (H2O)n (n = 6-22, 28, 30) and identification of spectral contributions from different H-bond conformations in gaseous and liquid water. J. Phys. Chem. A 110, 13388–13393 (2006).
Ojha, D., Karhan, K. & Kühne, T. D. On the hydrogen bond strength and vibrational spectroscopy of liquid water. Sci. Rep. 8, 16888 (2018).
Huisken, F., Kaloudis, M. & Kulcke, A. Infrared spectroscopy of small size-selected water clusters. J. Chem. Phys. 104, 17–25 (1996).
Watanabe, T., Ebata, T., Tanabe, S. & Mikami, N. Size-selected vibrational spectra of phenol-(H2O)n (n=1-4) clusters observed by IR-UV double resonance and stimulated Raman-UV double resonance spectroscopies. J. Chem. Phys. 105, 408–419 (1996).
Paul, J. B., Collier, C. P., Saykally, R. J., Scherer, J. J. & O’Keefe, A. Direct measurement of water cluster concentrations by infrared cavity ringdown laser absorption spectroscopy. J. Phys. Chem. A 101, 5211–5214 (1997).
Cable, K. M., Mauritz, K. A. & Moore, R. B. Effects of hydrophilic and hydrophobic counterions on the Coulombic interactions in perfluorosulfonate ionomers. J. Polym. Sci. B Polym. Phys. 33, 1065–1072 (1995).
Blanchard, R. M. & Nuzzo, R. G. An infrared study of the effects of hydration on cation-loaded Nafion thin films. J. Polym. Sci. B Polym. Phys. 38, 1512–1520 (2000).
Raichlin, Y., Millo, A. & Katzir, A. Investigations of the structure of water using mid-IR fiberoptic evanescent wave spectroscopy. Phys. Rev. Lett. 93, 1857031–1857034 (2004).
Cybulski, H. & Sadlej, J. On the calculations of the vibrational Raman spectra of small water clusters. Chem. Phys. 342, 163–172 (2007).
Perera, P. N. et al. Observation of water dangling OH bonds around dissolved nonpolar groups. Proc. Natl Acad. Sci. USA 106, 12230–12234 (2009).
Higuchi, A. & Iijima, T. D. s.c. investigation of the states of water in poly(vinyl alcohol-co-itaconic acid membranes. Polymer 26, 1833–1837 (1985).
Kusanagi, H. & Yukawa, S. Fourier transform infra-red spectroscopic studies of water molecules sorbed in solid polymers. Polymer 35, 5637–5640 (1994).
Zimm, B. H. & Lundberg, J. L. Sorption of vapors by high polymers. J. Phys. Chem. 60, 425–428 (1956).
Acknowledgements
This research is partially supported by the National Research Foundation, Prime Minister’s Office, Singapore under its Competitive Research Programme (CRP Award No. NRF-CRP24-2020-0006: A-0008375-00-00, A-0008375-01-00). The conjugated polyelectrolytes used in this report were designed and synthesized over few years by the combined effort of members of our chemistry team, including Qiu-Jing SEAH and Venu KEERTHI.
Author information
Authors and Affiliations
Contributions
C.G.T., M.N., M.M.S., M.C.Y.A., K.K.C. and Q.M.K. conducted the experiments. C.G.T., M.N. and R.Q.P. performed the quantum chemical calculations. R.Q.P. performed the OPLS4 calculations. M.C. and Y.P.F. performed the AMBER calculations. K.K.C. and M.C.Y.A. contributed to materials development. R.Q.P., L.L.C. and P.K.H.H. developed the research methodology and framework. L.L.C. directed materials development and characterization; R.Q.P. directed research project; P.K.H.H. directed theory support. L.L.C., R.Q.P. and P.K.H.H. wrote the manuscript. All authors discussed the experiments and results.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Tang, C.G., Syafiqah, M.N., Koh, QM. et al. Water binding and hygroscopicity in π-conjugated polyelectrolytes. Nat Commun 14, 3978 (2023). https://doi.org/10.1038/s41467-023-39215-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-023-39215-9
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.