Enrichment of methanol inside pNIPAM gels in the cononsolvency-induced collapse

Katja Nothdurfta, David H. Müllerb, Thorsten Brandsb, André Bardow*b and Walter Richtering*a
aInstitute of Physical Chemistry, RWTH Aachen University, Landoltweg 2, 52056 Aachen, Germany. E-mail: richtering@rwth-aachen.de
bInstitute of Technical Thermodynamics, RWTH Aachen University, Schinkelstr. 8, 52062 Aachen, Germany. E-mail: andre.bardow@ltt.rwth-aachen.de

Received 7th August 2019 , Accepted 29th September 2019

First published on 30th September 2019


Crosslinked poly-N-isopropylacrylamide (pNIPAM) gels adapt to their environment by a unique transition from a flexible, swollen macromolecular network to a collapsed particle. pNIPAM gels are swollen in both, pure water and pure methanol (MeOH). However, a drastic volume loss is observed in mixtures of water and methanol over a wide composition range. This effect is referred to as cononsolvency. Cononsolvency couples the volume phase transition to the transport of the cosolvent into the polymeric network. So far, the mechanisms underlying cononsolvency have not been fully elucidated. To obtain insights on cononsolvency, Raman microspectroscopy was applied to capture spatially resolved spectra distinguishing between the surroundings and the inside of the gel. Here, we used Indirect Hard Modelling (IHM) for the spectral analysis. Mass balancing allowed the calculation of the solvent composition inside the pNIPAM gel. The results show an increased methanol fraction inside the collapsed gel as compared to its surroundings. Furthermore, the sensitivity of the vibrational bands of methanol to its local hydrogen bonding environment allow to derive information about the molecular interactions. The methanol peak shifts measured inside the gel point towards donor-type hydrogen bonds between methanol and the peptide group of pNIPAM in the cononsolvency-induced collapse. The presented data should enhance our understanding of cononsolvency.


Introduction

Environmentally sensitive polymer gels are three-dimensional, crosslinked polymeric networks that are reversibly affected by changes in their surroundings e.g. temperature, pH or solvent composition.1,2 In particular, water-based systems are of great interest for variable applications,3 such as biomedicine,4 catalysis,5,6 drug delivery,7 sensors and photonics.8 The most prominent example of a stimuli-responsive gel is based on poly-N-isopropylacrylamide (pNIPAM) in combination with the crosslinker N,N′-methylenebis(acrylamide) (BIS).9 Besides the well-studied thermoresponsive behaviour, pNIPAM is sensitive to the solvent composition e.g. of water–alcohol mixtures:10,11 Either of the respective pure solvent swells the polymer network. However, a drastic, very fast volume loss is caused by changing the solvent mixture to a certain intermediate range.12 This volume change is similar to the volume phase transition induced by temperature. For water–methanol, a minimum in size is found around the most unfavourable mixture of about 20 mol% of methanol. This effect is referred to as cononsolvency. Cononsolvency is an intriguing phenomenon whose underlying processes are controversially discussed since the 1990s.10,13,14 For pNIPAM, cononsolvency behaviour was found in mixtures of water and methanol10,15 or other alcohols16–19 and in water–organic solvent19–24 mixtures. The cononsolvency phenomenon has been reported for other polymers as well,25–30 e.g. for poly-N,N-diethylacrylamide,27 polyvinylpyrrolidone,28 polyvinyl alcohol29 and copolymers of pNIPAM.30 The understanding of the cononsolvency is essential for the optimal design of new functional materials for applications such as catalysis, actuators, sensors or extraction and separation processes.31–35 An example is the system developed by Wang et al.32 who designed bilayer hydrogel actuators made of pNIPAM and poly(N-hydroxyethylacrylamide) with controllable actuation performance combining the temperature and solvent-dependent behaviour of the pNIPAM layer. Another application of cononsolvency has been demonstrated for double hydrophilic block copolymers composed of pNIPAM and b-poly(N-vinylimidazole) that exhibit an enhanced catalytic performance in esterolysis reactions within the cononsolvency-induced micellization.36 Moreover, greater insights about the collapse behaviour of polymers in cononsolvency mixtures will be of interest for the usage of pNIPAM-based brushes as a transfer system for nanoparticles that would be prone to aggregation otherwise.34 Another, theoretical study was presented by Li et al.35 concerning a system with polymer brushes forming cylindrical nanopores switchable by a cononsolvent.

There are various theories on the origin of cononsolvency, such as competitive hydrogen bonding,37–40 preferential adsorption,24,27,40–46 geometric frustration47,48 and strong water–cosolvent interactions49 including the formation of water–methanol clusters.50 Tanaka et al.37 proposed that the cooperativity of hydration is reduced by competitive hydrogen bonding of methanol and water molecules with the polymer. As a result, the polymer undergoes a minimum of the total coverage and collapses. Related to these ideas, the concept of preferential adsorption of the cosolvent is discussed frequently. Here, methanol is assumed to be enriched within the polymer network. According to Dalgicdir et al.48 applying computational calorimetry, methanol obstructs the formation of hydrogen bonds between water and the peptide group of pNIPAM causing a dehydration of the polymer. The importance of the amide proton to explain the cononsolvency behaviour is stressed by Scherzinger and Hofmann et al.51,52 They compared the dependence on the solvent composition of pNIPAM, bearing a secondary amide group, to poly-N,N-diethylacrylamide which is a tertiary amide and does not show cononsolvency. Walter et al.53 combined swelling experiments with molecular dynamic simulations. They proposed that the methanol molecules orient their methyl groups towards the bulk solvent generating an overall hydrophobic appearance to the water-rich bulk. Further studies by Mukherji et al.54 and Zhu et al.,55 using experimental data in combination with molecular dynamic simulations or mean-field approaches, suggested additional bridging of the cosolvent between polymer chains. In contrast, indirect mechanisms have been proposed without polymer–solvent interactions: in the work of Zhang and Wu,50 stoichiometric water–methanol complexes are formed through hydrogen bonds acting as a poor solvent for pNIPAM. Zuo et al.49 also recently supported the idea of strong water–cosolvent interactions exceeding the hydrogen bonding to the polymer chains in combination with preferential adsorption as origin of the cononsolvency. Here, results of neutron total scattering were combined with all-atom molecular dynamic simulations. Bischofberger et al.56 explained that hydrophobic hydration governs the phase behaviour of amphiphilic polymers. The hydration of hydrophobes is decreased by the kosmotropic effect of the cosolvent which strengthens the hydrogen-bonded water network.

Along with publications regarding the origin of the cononsolvency of pNIPAM in binary water–alcohol mixtures, several studies have investigated the difference in solvent composition within the pNIPAM gel compared to the surrounding fluid. As mentioned above, various theoretical studies and molecular dynamic simulations have suggested a preferential adsorption of alcohol by strong interactions with the pNIPAM chain in the collapsed region.40,43,44,48,53 Wang et al.42 performed high-resolution 1H MAS NMR experiments on pNIPAM microgels in pure water and in binary water–alcohol mixtures. Solely mixtures with low alcohol contents (2.5 or 5 mol%) were investigated. They determined preferential adsorption of alcohol molecules within the collapsed microgel network by quantification of the free and confined solvent species. Similar results were found by Mukae et al.57 and Hüther et al.17 performing indirect mass balancing experiments with macroscopic pNIPAM gels. Maeda et al.58 investigated linear pNIPAM in water–isopropanol mixtures by Raman microspectroscopy. They distinguished between polymer-rich domains of adsorbed chains and the solvent-rich matrix. Preferential adsorption of isopropanol in the polymer-rich regions was observed. The opposite behaviour was found for linear pNIPAM in water–DMSO mixtures, where the cosolvent is excluded from the collapsed, polymer-rich phase.21 Regarding water–methanol mixtures, Maeda et al.58 briefly stated that methanol was not enriched in the polymer-rich phase. Yang et al.59 proposed that methanol molecules interact solely with water molecules in the first solvation shell of the pNIPAM chains up to x(MeOH) < 0.75.

In our work, we explore the cononsolvency of pNIPAM gels in water–methanol mixtures over a wide composition range. We combine the experimental approach to mass balancing pursued by Hüther et al.17 with Raman microspectroscopic measurements. For the quantitative evaluation of the Raman spectra, we used the method of Indirect Hard Modelling (IHM).60 The spatial resolution of the Raman spectrometer in combination with a polymer gel allows a clear distinction between the polymer and its surroundings. We used the mass balancing method to determine the solvent composition within a macroscopic gel in comparison to the surrounding solution over a wide range of water–methanol mixtures. Subsequently, the pNIPAM–solvent interactions are elucidated by direct Raman microspectroscopy measurements inside the gel. Here, the sensitivity of the methanol peaks on the methanol molecules’ local hydrogen bonding environment is discussed.

Experimental

Materials

N-Isopropylacrylamide (NIPAM) was purchased from Acros Organics. The initiator ammonium persulfate (APS), the crosslinker N,N′-methylenebis(acrylamide) (BIS) and N,N,N′,N′-tetramethylethylenediamine (TEMED) were obtained from Merck. As mold for the synthesis 96 or 48 flat bottom Cellstar well plates from Greiner Bio-One GmbH were used (height 10.9 mm, diameter 6.39–6.96 mm and height 17.4 mm, diameter 11.05–11.56 mm, respectively). Methanol (Uvasol) were received from Merck Millipore and deionized water from Bernd Kraft.

Synthesis pNIPAM gels

Macroscopic pNIPAM gels were synthesized using a cooled molding approach. A solution of NIPAM, BIS (5 mol%) and TEMED (0.5 vol%) was prepared as well as APS was dissolved in water. Both solutions were degassed for 10 minutes. APS was added (1 mol%) to the monomer solution and mixed. Immediately afterwards, the solution was casted into a 96 and/or 48 well plate, sealed with a coverslip and kept in the fridge overnight to complete the polymerization. During the synthesis procedure, all solutions, well plates and coverslips were cooled with ice. Finally, the gels were removed from the well plate and washed in MilliQ water for 2 weeks.

Mass balancing experiments

The gels were dried for one week at atmospheric pressure and one week under vacuum. The weight of the individual, dried gels was determined with a micro balance. Then, a certain water–methanol mixture of known composition was added. The sample vials were closed and kept at 23 °C for about 7 days to achieve equilibrium. Subsequently, Raman microspectroscopy measurements were performed at two measurement points, one for the surrounding solution and a second one for the gel itself. Afterwards, the gels were removed from their vial and weighted. The solvent composition of the surrounding solution was evaluated by spectral analysis of the component areas using IHM as described below, whereas the methanol concentration inside the gel was calculated by mass balancing.

Raman microspectroscopy

Raman spectra were measured using an inverse confocal Raman microscope (inVia: Leica microscope, Renishaw, UK) equipped with a frequency-doubled Nd:YAG-laser of 100 mW at 532 nm excitation wavelength. For the mass balancing experiment, an air objective (20×) was used. The attached multichannel analyser consists of a spectrometer with a grating of 1800 l mm−1, a spectral resolution of about 1–2 cm−1 and a cooled CCD-camera. All measurements were performed at room temperature (23 °C). Two separate spectral regions (619–2287 cm−1 and 2505–3825 cm−1) were recorded. The fingerprint region around 1500 cm−1 was solely used to determine the position of the methanol–CO peak (1035.3 cm−1 in pure methanol). Here, the exposure time was set to 5 s but lowered to 2 s for >50 mol% methanol. The higher wavenumber region was evaluated for the mass balancing experiment as well as for the position of the methanol–CH peak (2836.5 cm−1 in pure methanol). The exposure time was set to 1 s but lowered to 0.5 s for the measurement of the pure methanol. In all cases, full laser power was used and 10 accumulations were averaged to improve the signal-to-noise ratio.

Indirect hard modelling (IHM)60–63

Spectra of pure water and pure methanol were measured to create pure component models. These spectra were modelled with peak-shaped pseudo-Voigt functions (see Fig. S1 in ESI). A spectral model for the glass signal was used to describe the background. All component models were combined to a mixture model. For each peak function in the mixture model, several parameters are available, namely peak position, full width at half maximum (FWHM) and Gaussian part of the pseudo-Voigt peak shape. These parameters are used in the mixture fit to account for nonlinear effects, as often seen for liquid mixtures, especially in case of water–methanol mixtures. For each peak function, parameters were assigned as fixed or flexible based on calibration data of binary water–methanol mixtures. Flexible parameters can be adjusted – as an additional degree of freedom – during optimization in the following spectral fitting step. Spectral fitting minimizes the residuals between the measured spectrum and the spectral mixture model in a least squares procedure. For each pure component in the mixture model, the intensities were integrated and weighted. Calibration spectra of binary mixtures of known composition were analysed to determine the calibration factor for converting the weight into the corresponding mole fraction.

To extract peak positions, only the spectral range of the relevant peak was evaluated. The remaining spectra was excluded from the evaluation and the mixture model was reduced accordingly. In addition to both solvent models, a pNIPAM model was included in the mixture model. For the methanol–CO peak, the spectral range of 975–1075 cm−1 and for the methanol–CH peak 2800–2860 cm−1 is considered.

Results and discussion

The swelling behaviour of the gels was evaluated in binary water–methanol mixtures between 0 and 70 mol% methanol. Fig. 1 shows the dependence of the swelling degree q on the methanol mole fraction. The swelling degree is defined as the weight of the gel equilibrated in the mixture divided by the dry mass.
image file: c9cp04383g-t1.tif
In water, the gel is fully swollen at ambient temperature. Upon addition of methanol, the polymer network collapses with a significant volume loss, also observable in the decreasing swelling degree. The smallest swelling degrees are reached for 15, 20 and 30 mol% methanol. Subsequently, a further increase in methanol fraction leads to re-swelling of the gel. The results are in agreement with previous experiments of pNIPAM microgels, macrogels or linear chains in water–methanol mixtures.11

image file: c9cp04383g-f1.tif
Fig. 1 Swelling degree (q = m(gel)/m(gel,dry)) of pNIPAM gels in dependence of the methanol mole fraction in water–methanol mixtures.

After equilibration of the gels in known water–methanol mixtures, Raman spectra were measured in the surrounding solution. The composition was evaluated by spectral analysis of the component areas using IHM.60 To this end, measurements were performed with binary mixtures of water and methanol of known compositions beforehand for calibration. The calibration yields a root mean square error for the mole fractions of 0.36% (see Fig. S2 in ESI). An exemplary fit of the mixture model to the measured spectra is shown for 20 mol% methanol in water including the corresponding residuals in the ESI (see Fig. S3). From the IHM analysis, we obtain the methanol mole fraction in the surrounding fluid in dependence of the overall methanol concentration. We convert the mole fractions into mass fractions ωsolMeOH. In combination with the weight mgel of the gel swollen in the respective mixture, the solvent composition inside the gel was calculated using the mass balance of methanol:64

image file: c9cp04383g-t2.tif
Here, m0i denote the known initial masses of the components in the system. The weight fraction (ω(MeOH)) is afterwards converted to the corresponding mole fraction (x(MeOH)).

Fig. 2 displays the comparison of the methanol concentration inside the gel to that in the surroundings in dependence of the overall concentration. An accumulation of the alcohol within the polymer network is found for solvent compositions that induce the gel collapse (15–30 mol% MeOH). The highest deviation (7.4 mol%) from the overall composition is found for the most unfavourable mixture of 20 mol% methanol. For the partly swollen states at 10, 50 and 70 mol% methanol, no difference from the overall or surrounding solvent composition is measured.


image file: c9cp04383g-f2.tif
Fig. 2 Methanol distribution inside the gel compared to its surroundings after equilibration in various water–methanol mixtures. The diagonal line represents the case of a uniform distribution of the solvent mixture within the whole system.

Comparable results were found for various aqueous mixtures inducing cononsolvency.17,42,58 Wang et al.42 compared the preferential adsorption of methanol and ethanol with pNIPAM in binary water–alcohol mixtures using 1H MAS NMR. Here, Wang et al. could separate confined and free alcohol species due to peak splitting. For comparatively low overall concentrations of 2.5 and 5.0 mol% of alcohol, they determined an increased mole fraction of confined binary solvents. They ascribed this effect to a preferential interaction of the polymer with the alcohols and observed a more pronounced effect for the more hydrophobic ethanol compared to methanol. For 2.5 and 5.0 mol% alcohol overall concentration, 5.7 and 7.4 mol% was found inside the gel in case of ethanol, while methanol only reached values of 3.3 and 5.8 mol%. Hüther et al.17,64 conducted mass balancing experiments using pNIPAM gels in combination with water–ethanol mixtures over the whole composition range. An increased fraction of the cosolvent inside the gel is suggested – more precisely, within the collapsed region of pNIPAM in water–ethanol mixtures (between 11 and 50 wt% of ethanol). The greatest difference (15.2 wt%) between the inside of the gel and the overall mixture was evaluated for an initial concentration of 28.9 wt% ethanol. Wang et al. concluded for low alcohol fractions that the enrichment of alcohol inside the gels increases with the hydrophobicity of the alcohol. This conclusion is confirmed by our data on water–methanol mixtures over a wider composition range in comparison to the results of Hüther et al.17 for water–ethanol. The accumulation of ethanol (15.2 wt%) found by Hüther et al. is higher than the accumulation of methanol found in our work (7.4 mol% methanol corresponding to 12.5 wt%).

Besides the preferential adsorption of alcohol species, other aqueous mixtures inducing cononsolvency show an accumulation of the cosolvent in the polymer network. Hüther et al.17 investigated water–acetone mixtures and found that acetone was less enriched than ethanol. Additionally, preferential interaction of acetone with pNIPAM was found using 1H HR-MAS NMR spectroscopy23 and molecular dynamic simulations.24 Zhu et al.55 recently used a local-bulk partitioning model to quantify the distribution of DMF molecules between the chain surface and the surrounding bulk solution to study pNIPAM microgels in water–DMF mixtures. The number of DMF molecules adsorbed to the chain, in particular the number of DMF bridges, is increased in the collapsed structure.

In addition, we tested whether Raman microspectroscopy can be applied to directly determine the solvent composition inside the gel. To this end, Raman was additionally measured inside the pNIPAM gels (Fig. S4 in ESI). However, the pNIPAM peaks in the spectra strongly overlap with the methanol peak in the CH-stretching region between 2850–3050 cm−1. Furthermore, non-linear effects appear, such as peak shifts of all components, changes in the Gaussian part of the pseudo-Voigt peak shape and width, complicating the evaluation of the peak area during the spectral analysis by IHM. Although, there is less overlap in the fingerprint region at low wavenumbers, no water signal is present to determine the water–methanol ratio. As no quantitative evaluation of the solvent composition can be achieved using the peak area, the sensitivity of the methanol peak positions to the hydrogen bonding environment is exploited.

For the binary mixture of water and methanol, it is well known that a significant shift occurs for the methanol CO- and CH-stretching band located at 1035.3 and 2836.5 cm−1 in pure methanol, respectively.65,66 We measured a reference solution consisting solely of the binary mixture of water and methanol (0.1 < x(MeOH) < 0.7) and found a linear dependence of the peak shifts on the methanol concentration over a wide composition range (Fig. 3 spectra dashed lines, Fig. 4 black symbols). The CO band experiences a red shift by 16.2 cm−1 upon decreasing the methanol fraction to 10 mol%, whereas the CH band experiences a blue shift by 8.5 cm−1. The opposite direction and difference in strength of the CO and CH shift were observed experimentally before and supported by ab initio calculations.67 Investigations focusing on the dependence of the methanol peak position on the methanol hydration were published by Dixit et al.65 In this context, a linear shift was observed in the intermediate composition range (0.7 > x(MeOH) > 0.25/0.15). In the outer regions (for x(MeOH) > 0.7, x(MeOH) < 0.25/0.15), a deviation from this linear dependence was found. Different hydrogen bonding configurations of methanol and water are proposed to explain the behaviour. These configurations are discussed in more detail further below.


image file: c9cp04383g-f3.tif
Fig. 3 Comparison of the CO-peak position (left) and the CH-stretching region (right) of reference binary solutions (dashed lines) and inside a pNIPAM gel (solid lines) at different water–methanol mixtures.

image file: c9cp04383g-f4.tif
Fig. 4 Methanol CO-peak shift (left) and CH-peak shift (right) extracted from spectra measured in a binary reference solution of solely water and methanol (black) and inside the gel (full red). The open red symbols represent the peak shifts corresponding to the methanol fraction obtained in the mass balancing experiment which were calculated using the linear fit (black line) of the binary reference solution.

Evaluating the same methanol peaks and their shifts in the ternary system from the spectra captured inside the pNIPAM gel, a different correlation with the methanol fraction is found (Fig. 3 spectra solid lines, Fig. 4 red full symbols): the peaks shift is strongly reduced inside the gel compared to the value expected for a uniform distribution of solvent molecules in the whole system, especially in the region of the cononsolvency-induced collapse. The peak positions found inside the gel are always closer to their original positions in pure methanol. The highest deviation of the peak shift inside the gel compared to the pure binary solvent mixture is again visible for the collapsed polymer region between 15 and 30 mol% methanol. If the peak shift would still be linearly correlated to the methanol fraction – as for the binary system, the measurements would indicate a higher amount of methanol inside the polymer network.

To correlate the aforementioned enrichment of methanol inside the gel with the sensitivity of the peak positions, the mole fraction of methanol calculated in the mass balancing experiment is transferred to the corresponding peak shift. Here, a linear fit of the binary reference solution is used to convert x(MeOH) in the gel to peak shifts (Fig. 4 red open symbols): in principle, the same trend is found that peak shifts are reduced inside the gel compared to the initial solution. However, the effect is significantly weaker. The peak shift caused by the enrichment of methanol inside the gel is much smaller than the peak shift actually measured. Thus, it is not possible to simply evaluate peak shifts to obtain the methanol concentration in the ternary system due to additional effects on the methanol peak positions. Most likely, methanol–polymer interactions affect the vibrational bands in addition to the solvent composition.

To elucidate the origin of the methanol peak positions inside pNIPAM gels, we compare our results to the literature by Gruenloh et al.68 and Dixit et al.65 on the binary mixture and molecular simulations. First, we describe the effect of various hydrogen-bonding options on the methanol peak position by looking at the binary solvent mixture.

Gruenloh et al. derived trends for the shift of the CH-peak position by evaluating well-known benzene–methanol clusters measured with resonant ion-dip infrared spectroscopy: when methanol acts as a hydrogen-bond donor (D), the wavenumber of the CH peak decreases. In contrast, the peak position increases to higher wavenumbers when methanol is an acceptor (A) of a hydrogen bond or is existent in an acceptor–acceptor–donor (AAD) configuration. The strength of this increase is similar for both configurations. This behaviour is explained by the compensating effects of the combination of acceptor and donor hydrogen bonding. In pure liquid methanol, the solvent molecules form chain-like clusters where each methanol molecule acts as single acceptor/single donor (AD). As aforementioned, Dixit et al. explained the shift of the CH- and CO–methanol-peak bands by the change in hydrogen-bonding environment. If only little water is added to pure methanol, water molecules coordinate at the end of the chain-like methanol clusters. Here, the chain-end methanol molecules act as hydrogen-bond acceptors. Upon further hydration, water breaks/shortens the methanol chains and, in addition, individually solvated methanol molecules are formed. Each methanol molecule can accept two and donate one hydrogen bond leading to an AAD configuration. These explanations align with Gruenloh's findings that A and AAD methanol molecules cause a blue shift of the CH band in the spectra.

These trends are applied to interpret the peak shifts inside a pNIPAM gel shown before. For facilitation of the discussion, we focus on the CH-peak position:

One factor decreasing the peak shift of the CH band is the increased methanol fraction inside the pNIPAM gels in the region of the cononsolvency-induced collapse. Here, the AAD hydrogen bonding is reduced compared to that in the binary solvent mixture. As mentioned in the discussion of Fig. 4, only a part of the peak shift is caused by this increased methanol fraction. The remaining difference in peak shift arises from interactions with the polymer. Various studies show that there is a strong preference of the pNIPAM chains to form hydrogen bonds.48,53,54 In pure water, hydrogen bonds predominantly exist between pNIPAM–CO and water. These bonds are broken upon the addition of methanol and are partly replaced by hydrogen bonds to methanol. As aforementioned, Dalgicdir et al.48 debated that the methanol molecules geometrically frustrate the ability of water to form hydrogen bonds with the pNIPAM peptide group. The total number of hydrogen bonds decreases in the cononsolvency-induced collapse, however, the number of pNIPAM–methanol hydrogen-bonds is enlarged. According to Mukherji et al.54 hydrogen bonds between the oxygen of the carbonyl group of pNIPAM and the hydrogen of methanol are expected to be most prominent. Here, methanol acts as a donor. Other hydrogen-bonding options involve the pNIPAM–NH group. In this case, methanol can act either act as an acceptor or donor. With respect to the trends established by Gruenloh et al.,68 donor-type hydrogen-bonding causes a decrease in wavenumber of the CH-peak position. Thus, to align with the decrease in CH-peak shift measured inside the pNIPAM gel, donor-type hydrogen-bonding between methanol and either the oxygen or nitrogen of the pNIPAM peptide group is indicated.

As previously mentioned, the effect of different hydrogen bonds on the CH-peak position is opposite compared to the effect of the CO peak position. Therefore, the red shift of the CO peak indicates donor-type hydrogen-bonds as well.

Conclusions

In the region of the cononsolvency-induced collapse, we find an increased fraction of methanol inside pNIPAM gels compared to the surrounding liquid. However, less methanol is observed in the pNIPAM gel compared to the same system with ethanol presented by Hüther et al.17 Thus, the accumulation of the alcohol species inside the polymer network in its collapsed region is more pronounced with increasing hydrophobicity of the alcohol moiety.

For more detailed information on molecular interactions, the sensitivity of the methanol vibrational bands to the local environment of methanol was exploited. More explicitly, the impact of acceptor/donor hydrogen bonding of methanol is taken into account. The binary water–methanol mixture exhibits a linear dependence of the methanol peak positions on the methanol fraction over a wide composition range. However, a deviation from this trend is found for the ternary system with water–methanol–pNIPAM which is ascribed to two effects: firstly, a small shift is caused by the different solvent composition inside the gel leading to a change in water–methanol hydrogen bonding. Secondly, a major part of the peak shift is due to interactions between methanol and pNIPAM. Based on literature finding, donor-type hydrogen-bonding can be responsible for a decrease in wavenumber of the methanol CH-peak position (and an increase of the methanol CO-peak position).65,68 Therefore, the presented results suggest that methanol predominantly donates its hydrogen for interactions with the oxygen of the carbonyl group of pNIPAM within the cononsolvency-induced collapse.

The obtained results regarding the cononsolvency behaviour, the enrichment of methanol inside the gel and the molecular interactions can provide further understanding of the performance of new functional materials. For applications in catalysis, drug delivery systems or in extraction and separation processes, information on the local structure and polarity within the gels are valuable.69,70 For example, the properties of the environment during catalysis impacts the catalytic activity and selectivity.71 Hydrophobic substrates and/or catalysts will prefer the increased hydrophobicity of the collapsed, methanol-enriched polymer network. Furthermore, Mukherji et al.31 recently pointed out the importance of a detailed understanding of the cononsolvency to create design principles for new materials. Here, molecular information about the local arrangement of the solvent molecules can support the setup or verification of simulations predicting e.g. properties and structures of pNIPAM-based systems in cononsolvency mixtures. In this respect, further detailed insights regarding the kinetics of the volume phase transition induced by cononsolvency should be pursued.12,72

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge financial support by the DFG (Deutsche Forschungsgemeinschaft) within the SFB 985 “Functional Microgels and Microgel Systems”. We thank Carsten Flake, Julia Thien, Hans-Jürgen Koß and Rico Keidel for helpful discussions.

Notes and references

  1. F. A. Plamper and W. Richtering, Acc. Chem. Res., 2017, 50, 131–140 CrossRef CAS PubMed.
  2. M. A. C. Stuart, W. T. S. Huck, J. Genzer, M. Müller, C. Ober, M. Stamm, G. B. Sukhorukov, I. Szleifer, V. V. Tsukruk, M. Urban, F. Winnik, S. Zauscher, I. Luzinov and S. Minko, Nat. Mater., 2010, 9, 101 CrossRef PubMed.
  3. M. Karg, A. Pich, T. Hellweg, T. Hoare, L. A. Lyon, J. J. Crassous, D. Suzuki, R. A. Gumerov, S. Schneider, I. I. Potemkin and W. Richtering, Langmuir, 2019, 35, 6231–6255 CrossRef CAS PubMed.
  4. M. A. Ward and T. K. Georgiou, Polymers, 2011, 3, 1215–1242 CrossRef CAS.
  5. S. Wiese, A. C. Spiess and W. Richtering, Angew. Chem., Int. Ed., 2013, 52, 576–579 CrossRef CAS PubMed.
  6. S. Wu, J. Dzubiella, J. Kaiser, M. Drechsler, X. Guo, M. Ballauff and Y. Lu, Angew. Chem., Int. Ed., 2012, 51, 2229–2233 CrossRef CAS PubMed.
  7. T. Hoare and R. Pelton, Langmuir, 2008, 24, 1005–1012 CrossRef CAS PubMed.
  8. A. K. Yetisen, H. Butt, L. R. Volpatti, I. Pavlichenko, M. Humar, S. J. J. Kwok, H. Koo, K. S. Kim, I. Naydenova, A. Khademhosseini, S. K. Hahn and S. H. Yun, Biotechnol. Adv., 2016, 34, 250–271 CrossRef CAS PubMed.
  9. R. Pelton, Adv. Colloid Interface Sci., 2000, 85, 1–33 CrossRef CAS PubMed.
  10. F. M. Winnik, H. Ringsdorf and J. Venzmer, Macromolecules, 1990, 23, 2415–2416 CrossRef CAS.
  11. C. Scherzinger, A. Schwarz, A. Bardow, K. Leonhard and W. Richtering, Curr. Opin. Colloid Interface Sci., 2014, 19, 84–94 CrossRef CAS.
  12. R. Keidel, A. Ghavami, D. M. Lugo, G. Lotze, O. Virtanen, P. Beumers, J. S. Pedersen, A. Bardow, R. G. Winkler and W. Richtering, Sci. Adv., 2018, 4, eaao7086 CrossRef PubMed.
  13. T. Amiya, Y. Hirokawa, Y. Hirose, Y. Li and T. Tanaka, J. Chem. Phys., 1987, 86, 2375–2379 CrossRef CAS.
  14. S. Hirotsu, J. Phys. Soc. Jpn., 1987, 56, 233–242 CrossRef CAS.
  15. C. K. Chee, B. J. Hunt, S. Rimmer, I. Soutar and L. Swanson, Soft Matter, 2011, 7, 1176–1184 RSC.
  16. H. M. Crowther and B. Vincent, Colloid Polym. Sci., 1998, 276, 46–51 CrossRef CAS.
  17. A. Hüther, X. Xu and G. Maurer, Fluid Phase Equilib., 2004, 219, 231–244 CrossRef.
  18. M. C. Arndt and G. Sadowski, Macromolecules, 2012, 45, 6686–6696 CrossRef CAS.
  19. R. O. R. Costa and R. F. S. Freitas, Polymer, 2002, 43, 5879–5885 CrossRef CAS.
  20. G. Dalkas, K. Pagonis and G. Bokias, Polymer, 2006, 47, 243–248 CrossRef CAS.
  21. H. Yamauchi and Y. Maeda, J. Phys. Chem. B, 2007, 111, 12964–12968 CrossRef CAS PubMed.
  22. P. W. Zhu and D. H. Napper, Macromol. Chem. Phys., 1999, 200, 1950–1955 CrossRef CAS.
  23. J. Wang, N. Wang, B. Liu, J. Bai, P. Gong, G. Ru and J. Feng, Phys. Chem. Chem. Phys., 2017, 19, 30097–30106 RSC.
  24. H. A. Pérez-Ramírez, C. Haro-Pérez, E. Vázquez-Contreras, J. Klapp, G. Bautista-Carbajal and G. Odriozola, Phys. Chem. Chem. Phys., 2019, 21, 5106–5116 RSC.
  25. H. Mahdavi, B. O. Mahdi Sadeghzadeh and N. T. Qazvini, J. Polym. Sci., Part B: Polym. Phys., 2009, 47, 455–462 CrossRef CAS.
  26. N. Orakdogen and O. Okay, Polymer, 2006, 47, 561–568 CrossRef CAS.
  27. B. Liu, J. Wang, G. Ru, C. Liu and J. Feng, Macromolecules, 2015, 48, 1126–1133 CrossRef CAS.
  28. M. Guettari, R. Gomati and A. Gharbi, J. Macromol. Sci., Part B: Phys., 2010, 49, 552–562 CrossRef CAS.
  29. N. Takahashi, T. Kanaya, K. Nishida and K. Kaji, Polymer, 2003, 44, 4075–4078 CrossRef CAS.
  30. Y. Maeda and M. Yamabe, Polymer, 2009, 50, 519–523 CrossRef CAS.
  31. D. Mukherji, M. D. Watson, S. Morsbach, M. Schmutz, M. Wagner, C. M. Marques and K. Kremer, Macromolecules, 2019, 52, 3471–3478 CrossRef CAS.
  32. X. Wang, H. Huang, H. Liu, F. Rehfeldt, X. Wang and K. Zhang, Macromol. Chem. Phys., 2019, 220, 1800562 CrossRef.
  33. S. Nian and L. Pu, J. Org. Chem., 2019, 84, 909–913 CrossRef CAS PubMed.
  34. Y. Yu, R. A. Lopez de la Cruz, B. D. Kieviet, H. Gojzewski, A. Pons, G. Julius Vancso and S. de Beer, Nanoscale, 2017, 9, 1670–1675 RSC.
  35. C.-W. Li, H. Merlitz, C.-X. Wu and J.-U. Sommer, Macromolecules, 2018, 51, 6238–6247 CrossRef CAS.
  36. Z. Ge, D. Xie, D. Chen, X. Jiang, Y. Zhang, H. Liu and S. Liu, Macromolecules, 2007, 40, 3538–3546 CrossRef CAS.
  37. F. Tanaka, T. Koga and F. M. Winnik, Phys. Rev. Lett., 2008, 101, 028302 CrossRef PubMed.
  38. H. Kojima, F. Tanaka, C. Scherzinger and W. Richtering, J. Polym. Sci., Part B: Polym. Phys., 2013, 51, 1100–1111 CrossRef CAS.
  39. J. Dudowicz, K. F. Freed and J. F. Douglas, J. Phys. Chem. B, 2016, 120, 5753–5758 CrossRef CAS PubMed.
  40. F. Tanaka, T. Koga, H. Kojima, N. Xue and F. M. Winnik, Macromolecules, 2011, 44, 2978–2989 CrossRef CAS.
  41. S. Bharadwaj and N. F. A. van der Vegt, Macromolecules, 2019, 52, 4131–4138 CrossRef CAS.
  42. N. Wang, G. Ru, L. Wang and J. Feng, Langmuir, 2009, 25, 5898–5902 CrossRef CAS PubMed.
  43. D. Mukherji, C. M. Marques and K. Kremer, Nat. Commun., 2014, 5, 4882 CrossRef PubMed.
  44. F. Rodríguez-Ropero, T. Hajari and N. F. A. van der Vegt, J. Phys. Chem. B, 2015, 119, 15780–15788 CrossRef PubMed.
  45. V. I. Michailova, D. B. Momekova, H. A. Velichkova, E. H. Ivanov, R. K. Kotsilkova, D. B. Karashanova, E. D. Mileva, I. V. Dimitrov and S. M. Rangelov, J. Phys. Chem. B, 2018, 122, 6072–6078 CrossRef CAS PubMed.
  46. F. M. Winnik, M. F. Ottaviani, S. H. Bossmann, M. Garcia-Garibay and N. J. Turro, Macromolecules, 1992, 25, 6007–6017 CrossRef CAS.
  47. A. Pica and G. Graziano, Phys. Chem. Chem. Phys., 2016, 18, 25601–25608 RSC.
  48. C. Dalgicdir, F. Rodríguez-Ropero and N. F. A. van der Vegt, J. Phys. Chem. B, 2017, 121, 7741–7748 CrossRef CAS PubMed.
  49. T. Zuo, C. Ma, G. Jiao, Z. Han, S. Xiao, H. Liang, L. Hong, D. Bowron, A. Soper, C. C. Han and H. Cheng, Macromolecules, 2019, 52, 457–464 CrossRef CAS.
  50. G. Zhang and C. Wu, J. Am. Chem. Soc., 2001, 123, 1376–1380 CrossRef CAS.
  51. C. Scherzinger, A. Balaceanu, C. H. Hofmann, A. Schwarz, K. Leonhard, A. Pich and W. Richtering, Polymer, 2015, 62, 50–59 CrossRef CAS.
  52. C. H. Hofmann, F. A. Plamper, C. Scherzinger, S. Hietala and W. Richtering, Macromolecules, 2013, 46, 523–532 CrossRef CAS.
  53. J. Walter, J. Sehrt, J. Vrabec and H. Hasse, J. Phys. Chem. B, 2012, 116, 5251–5259 CrossRef CAS PubMed.
  54. D. Mukherji, M. Wagner, M. D. Watson, S. Winzen, T. E. de Oliveira, C. M. Marques and K. Kremer, Soft Matter, 2016, 12, 7995–8003 RSC.
  55. P.-w. Zhu and L. Chen, Phys. Rev. E, 2019, 99, 022501 CrossRef CAS PubMed.
  56. I. Bischofberger, D. C. E. Calzolari, P. De Los Rios, I. Jelezarov and V. Trappe, Sci. Rep., 2014, 4, 4377 CrossRef CAS PubMed.
  57. K. Mukae, M. Sakurai, S. Sawamura, K. Makino, S. W. Kim, I. Ueda and K. Shirahama, J. Phys. Chem., 1993, 97, 737–741 CrossRef CAS.
  58. Y. Maeda, H. Yamamoto and I. Ikeda, Macromolecules, 2003, 36, 5055–5057 CrossRef CAS.
  59. M. Yang and K. Zhao, J. Polym. Sci., Part B: Polym. Phys., 2017, 55, 1227–1234 CrossRef CAS.
  60. F. Alsmeyer, H.-J. Koß and W. Marquardt, Appl. Spectrosc., 2004, 58, 975–985 CrossRef CAS PubMed.
  61. F. Alsmeyer and W. Marquardt, Appl. Spectrosc., 2004, 58, 986–994 CrossRef CAS PubMed.
  62. P. Beumers, D. Engel, T. Brands, H. J. Koß and A. Bardow, Chemom. Intell. Lab. Syst., 2018, 172, 1–9 CrossRef CAS.
  63. C. Peters, L. Wolff, S. Haase, J. Thien, T. Brands, H.-J. Koß and A. Bardow, Lab Chip, 2017, 17, 2768–2776 RSC.
  64. A. Hüther, PhD dissertation, Universität Kaiserslautern, Germany, 2000.
  65. S. Dixit, W. C. K. Poon, J. Crain, S. Dixit and W. C. K. Poon, J. Phys.: Condens. Matter, 2000, 12, L323 CrossRef CAS.
  66. Y. Numata, Y. Iida and H. Tanaka, J. Quant. Spectrosc. Radiat. Transfer, 2011, 112, 1043–1049 CrossRef CAS.
  67. C. A. Holden, S. S. Hunnicutt, R. Sánchez-Ponce, J. M. Craig and S. C. Rutan, Appl. Spectrosc., 2003, 57, 483–490 CrossRef CAS PubMed.
  68. C. J. Gruenloh, G. M. Florio, J. R. Carney, F. C. Hagemeister and T. S. Zwier, J. Phys. Chem. A, 1999, 103, 496–502 CrossRef CAS.
  69. A. Purohit, S. P. Centeno Benigno, W. Richtering, S. K. Wypysek and D. Wöll, Chem. Sci., 2019 10.1039/C9SC03373D.
  70. D. Parasuraman and M. J. Serpe, ACS Appl. Mater. Interfaces, 2011, 3, 4714–4721 CrossRef CAS PubMed.
  71. N. Mase, Y. Nakai, N. Ohara, H. Yoda, K. Takabe, F. Tanaka and C. F. Barbas, J. Am. Chem. Soc., 2006, 128, 734–735 CrossRef CAS PubMed.
  72. C. Scherzinger, O. Holderer, D. Richter and W. Richtering, Phys. Chem. Chem. Phys., 2012, 14, 2762–2768 RSC.

Footnotes

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9cp04383g
Additional supporting research data for this article may be accessed at no charge at https://hdl.handle.net/21.11102/c49c9328-b83a-11e9-9a63-e41f1366df48

This journal is © the Owner Societies 2019