Stimuli-responsive hybrid porous polymers based on acetals of polyvinyl alcohol and acrylic hydrogels

Accepted Manuscript Title: STIMULI-RESPONSIVE HYBRID POROUS POLYMERS BASED ON ACETALS OF POLYVINYL ALCOHOL AND ACRYLIC HYDROGELS Authors: Yu. Samchenko, O. Korotych, L. Kernosenko, S. Kryklia, O. Litsis, M. Skoryk, T. Poltoratska, N. Pasmurtseva PII: DOI: Reference:

S0927-7757(18)30092-X https://doi.org/10.1016/j.colsurfa.2018.02.015 COLSUA 22267

To appear in:

Colloids and Surfaces A: Physicochem. Eng. Aspects

Received date: Revised date: Accepted date:

15-12-2017 6-2-2018 8-2-2018

Please cite this article as: Samchenko Y, Korotych O, Kernosenko L, Kryklia S, Litsis O, Skoryk M, Poltoratska T, Pasmurtseva N, STIMULI-RESPONSIVE HYBRID POROUS POLYMERS BASED ON ACETALS OF POLYVINYL ALCOHOL AND ACRYLIC HYDROGELS, Colloids and Surfaces A: Physicochemical and Engineering Aspects (2010), https://doi.org/10.1016/j.colsurfa.2018.02.015 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

STIMULI-RESPONSIVE HYBRID POROUS POLYMERS BASED ON ACETALS OF POLYVINYL ALCOHOL AND ACRYLIC HYDROGELS Yu. Samchenko 1,a, O. Korotych T. Poltoratska a, and N. Pasmurtseva a

a, ,b, c

, L. Kernosenko a, S. Kryklia a, O. Litsis d, M. Skoryk e,

a

IP T

F. D. Ovcharenko Institute of Biocolloid Chemistry, Department of Functional Hydrogels 42 Akademika Vernadskogo Blvd., Kyiv-03142, Ukraine b University of Florida, J. Crayton Pruitt Family Department of Biomedical Engineering 1275 Center Dr., Gainesville, FL 32611, USA c University of Florida, Department of Chemical Engineering 1030 Center Dr. Gainesville, FL 32611, USA d Taras Shevchenko National University, Department of Chemistry 64/13, Volodymyrska St., Kyiv 01601, Ukraine e NanoMedTech LLC 68 Antonovycha St., Kyiv 03680, Ukraine 1

SC R

All correspondence and requests for materials should be addressed to Yu. Samchenko at [email protected]

ED

M

A

N

U

Graphical abstract

A

CC E

PT

Abstract Hybrid hydrogels have gained a lot of attention due to their unique properties which can be tailored for a variety of applications. In this paper, hybrid porous polymers based on sponge-like acetals of polyvinyl alcohol (polyvinyl formals) with functionalized pore structure by pH-sensitive or thermosensitive hydrogels have been synthesized. Synergistic improvement of hybrid hydrogel physicochemical properties (mechanical, swelling, and sorption characteristics) compared to the components from which they were constructed is demonstrated, as well as application of these materials for sorption and removal of model dyes (methylene blue and fluorescein) from aqueous solutions. The hybrid materials have the potential to be used as effective sorbents in numerous applications such as industrial wastewater treatment due to their improved mechanical properties, high-water retention, fast sorption, high sorption capacity, and low cost. List of abbreviations:

AA APS CPD ESD FL FTIR HH IR MB

acrylic acid ammonium persulfate, (NH4)2S2O8 critical point drying equilibrium swelling degree fluorescein Fourier transform infrared spectroscopy hybrid hydrogel infrared methylene blue

Page | 1

N,N′-methylenebisacrylamide National Academy of Sciences of Ukraine N-isopropylacrylamide poly(acrylic acid) poly(acrylic acid) hydrogel probability density function potassium persulfate, K2S2O8 polyvinyl alcohol polyvinyl formal standard error scanning electron microscopy N,N,N′,N′-tetramethylenediamine ultraviolet-visible

IP T

MBAAm NASU NIPAAm PAA PAAH PDF PPS PVA PVF SE SEM TEMED UV-vis

CONTENTS 1. 2.

Introduction Materials and Methods

SC R

Keywords: hybrid hydrogel; polyvinyl alcohol; acrylic acid; N-isopropylacrylamide; dye sorption; swelling and sorption kinetic.

3 4

Materials

2.2.

Synthesis of hybrid hydrogels and its components

2.3.

FTIR spectroscopy

2.4.

Pore size characterization

2.5.

Mechanical properties: tensile stress and elongation

5

2.6.

Equilibrium swelling degree and swelling kinetics

5

2.7.

Sorption capacity and kinetics

6

2.8.

Chemical structures and modeling

7

2.9.

Statistical data analysis

ED

M

A

N

U

2.1.

Results and Discussion

4.

Conclusions

4 4 5

7 8 13 14

A

CC E

References

PT

3.

4

Page | 2

INTRODUCTION Hydrogels, spatially cross-linked hydrophilic polymers consisting of more than 90 % water, are used successfully in different areas, predominantly for biomedical applications such as tissue engineering [1], plastic surgery [2], and targeted drug delivery [3] mostly due to their high biocompatibility. They are widely used for contact lenses, wound dressing, and transdermal drug delivery systems. Additionally, hydrogels are used in food and agriculture industries (water storage granules and for controlled release of chemicals). Advances in hydrogel application in the field of catalysis [4, 5], sensors [5], and actuators [5, 6] have been also achieved and the reader is referred to the review by Döring et al. for a full summary on this topic [7]. Recently polymer systems have also gained great attention as effective sorbents [8] for removal of metal cations [9], anions, dyes, radionuclides, and other pollutants from water. So far, rapid and convenient removal of chemical compounds from wastewater has been a challenge faced by scientists [10]. Usage of hydrogels for wastewater treatment has been limited due to their poor mechanical properties [11] such as low tensile strength and poor elasticity, brittleness, and fragility [12]. Mentioned limitations can be overcome through synthesis of so called hybrid hydrogels (HHs) which will have improved and tuned properties, and high surface area-to-volume ratio due to their small size (micro- and nanohydrogels) or high porosity. These systems can be designed based on interpenetrating polymeric networks of different natures and, consequently, with different hydrophilicity, mechanical properties, etc. HHs can also be synthesized by incorporation of inorganic filler into the polymer composition [8], for example, silver nanoparticles [13]. Another example of hybrid hydrogels can be magnetosensitive polymer system obtained by doping the hydrogel with magnetite particles [14]. Among HH systems, "smart" or stimuli-responsive HHs have recently gained a lot of interest due to their unique properties such as triggered uptake or release of solvent from the polymer network by slightly altering external conditions: temperature, pH, magnetic or electric [15] fields, or ultraviolet (UV) radiation [16]. These processes on a molecular level are accompanied by a volume phase transition between a swollen and a collapsed state of the gel, and lead to a change in properties of the macroscopic polymer network (dimensions, elasticity, turbidity, etc.). Thus, designed HHs will be able to accumulate benefits of initial hydrogel components while eliminating its drawbacks. As mentioned earlier, one practical application of HHs is their usage as sorbents. Given that sorption methods are inexpensive, they are widely used for removal of both organic and inorganic contaminants. Other methods for removal of contaminants can include sedimentation, chemical precipitation, flocculation, ion exchange, membrane filtration, electrochemical treatment, etc. [17, 18]. Technogenic pollutants and heavy metals from wastewater accompany mining, dyeing, leather tanning and many other processes must be removed to ensure safety and regulatory compliance. Metals in waste streams do not naturally degrade and are toxic to aquatic life, even at very low concentrations, and tend to accumulate in living organisms. Metals, which must be removed, include soluble and/or particulate heavy metals, such as lead, copper, chromium, iron, manganese, mercury, nickel, and zinc. Wastewater is also full of contaminants including bacteria, and different organic substances, for example, dyes, toxins. HHs with a well-developed highly porous, functionalized, and hydrophilic surface can be suitable candidates as sorbents for water purification owing to fast penetration of aqueous solutions and selective sorption of contaminants. Also, functionalization of HH pore space with different functional groups, e.g., with carboxyl groups, makes them promising materials for applications as sensors, transducers for analytical needs, accumulators for gas chromatography, etc. Currently, however, poor mechanical properties of hydrogels hinder their practical application [12] mainly due to their inability for long-term usage and multiple regeneration cycles. Thus, the paper is devoted to the design and synthesis of highly porous HHs which can be potentially used as sorbents. The main components of the HH systems are polyvinyl acetals and hydrogels based on acrylic acid or N-isopropylacrylamide. Polyvinyl acetals were chosen due to their excellent chemical resistance, high tensile strength, impact resistance, and elasticity; while the acrylic hydrogels – due to their hydrophilicity, pH-sensitivity, and thermosensitivity. Dependence of physical properties (mechanical, swelling and sorption characteristics) of obtained hybrid polymers on their composition under various conditions are studied from a perspective of their application for industrial wastewater treatment.

A

CC E

PT

ED

M

A

N

U

SC R

IP T

1.

Page | 3

2. 2.1.

MATERIALS AND METHODS Materials Acrylic acid (AA) (Merck, 99 %) was distilled under vacuum with the addition of 1 mL of concentrated sulfuric acid and purified by fractional crystallization. N-isopropylacrylamide (NIPAAm) (Sigma-Aldrich, 97 %) was recrystallized from hexane and dried under vacuum. N,N′-methylenebisacrylamide (MBAAm) (Merck, 98 %); ammonium persulfate (APS) (NH4)2S2O8 (Sigma, 98 %); potassium persulfate (PPS) K2S2O8 (Sigma, 98 %); N,N,N′,N′-tetramethylenediamine (TEMED) (Merck, 99 %); linear polyvinyl alcohol (PVA) (AppliChem GmbH, 98 %; 72 kDa); formaldehyde (LAB-SCAN, 37 %); concentrated sulfuric acid H2SO4; ethanol C2H5OH; Triton X-100 (AppliChem GmbH); laponite RD (Rockwood Ltd.; 771.45 Da1); methylene blue (MB) and fluorescein (FL) dyes were used as received without further purification. Distilled water was used as a solvent in all experiments. Synthesis of hybrid hydrogels and its components HHs and individual components (acetals of PVA and functional hydrogels) were synthesized as described below. PVA acetals. The acetalization of PVA was performed by condensing PVA with formaldehyde in the presence of a strong acid. The details of the synthesis of polyvinyl formals (PVFs) are discussed in our previous article [19]. The composition of PVA acetal used for further functionalization is given in Table 1.

SC R

IP T

2.2.

ED

M

A

N

U

Stimuli-responsive hydrogels. Hydrogels based on NIPAAm were synthesized by radical polymerization of monomer aqueous solutions at 10 ℃ and covalently crosslinked using bifunctional monomer MBAAm. For initiation of the polymerization, redox system based on APS and TEMED was used. The hydrogel synthesis was performed as follows. Argon was bubbled through a reaction mixture (aqueous solutions of monomers and cross-linking agent) prior to adding the redox system. After the addition of the redox system, the composition was mixed and transferred to a mold consisting of two parallel glass plates separated by 1 mm-thick spacers and kept for 1 h at room temperature (RT). Polymerization of AA was performed in an analogous way, but at a higher temperature (at 60 ℃) and using PPS as a thermal initiator. Approximately four hours later, the polymerized hydrogels were removed from molds and washed extensively in distilled water at RT to remove unreacted residues. Water was changed 1–3 times a day, and the washing process was controlled using a UV-vis spectrometer (SPECORD M40, Carl Zeiss). Gel disks were cut from swollen hydrogel films using a hole puncher (d = 10 mm), and dried to a constant mass at 25 ℃. The detailed composition of chemically and physically crosslinked hydrogels is listed in Table 2 and A1 (see SI: A).

A

CC E

PT

Hybrid hydrogels. Hybrid crosslinked polymers based on porous PVA acetals and stimuli-responsive hydrogels have been obtained through the swelling of PVA sponges in a gel-forming composition for 5 min. After being swollen, the sponges were placed in an argon atmosphere and kept there for 24 h at 40 ℃. Composites with partially filled pores have also been synthesized (Fig. 1). Partial pore filling was achieved by expelling a certain portion of the gel forming composition from a swollen sponge (e.g., swollen sponges were put under mechanical compression). In the case of cylindrically shaped sponges, the mentioned process of pore filling was carried out inside of a syringe with a suitable diameter by squeezing the required portion of composition gradually by a piston. Synthesized HHs with different degrees of gel-forming composition were designated as HH-100, HH75, and so on. The notation HH-75, for example, means that the HH based on PVF after mechanical compression contains 75 wt. % of hydrogel-forming composition comparing to completely swollen uncompressed system. 2.3.

FTIR spectroscopy The Fourier transform infrared (FTIR) spectral analysis of functional groups [20] of air-dried polymeric systems, monomers and laponite was carried out on a Spectrum BX FT-IR spectrometer (Perkin Elmer). The spectra of polymers, monomers and laponite were recorded using an attenuated total reflection technique (internal reflection spectroscopy) in the spectral range 550–4000 cm−1 with a resolution of 2 cm−1 and accumulations of 8 scans which were combined to average out random absorption artifacts. In the same spectral range, the transmittance spectrum of AA was recorded using a PIKE transmission cell for liquid 1

Molecular weight of laponite was theoretically calculated based on its following chemical structure: Na[(Si8Mg5.5Li0.3)O20(OH)4]. Page | 4

samples with KBr windows. 2.4.

Pore size characterization Structure of opened pores of polymer systems, and pore size distribution were determined from photos of PVF and HHs scanned on Epson Perfection V330 and analyzed using ImageJ software (version 1.51j). All samples were approximately 1 mm thick. Pore diameters 𝑑 (µm) calculated from pore areas 𝐴𝑝𝑜𝑟𝑒 (µm2) as 𝐴𝑝𝑜𝑟𝑒

𝑑 = 2√

(1)

𝜋

correlate with determined Feret’s diameters, and on average are a few tens of micrometers smaller than the latter one. For central tendency measurements, pore size distributions were fitted by a lognormal probability density function (PDF): 𝑓(𝑑) =

𝐴 √2𝜋𝑑𝜎

𝑒

−

(ln 𝑑−ln 𝑑𝑠 )2 2𝜎2

(2)

SC R

IP T

where 𝐴 is the area under the curve; 𝑑𝑠 – the scale parameter (and it is also the median of the distribution); 𝜎 – the shape parameter (and it is the standard deviation of the logarithm of the distribution). The arithmetic mean (𝑑𝑚𝑒𝑎𝑛 ), median (𝑑𝑚𝑒𝑑𝑖𝑎𝑛 ) and mode (𝑑𝑚𝑜𝑑𝑒 ) diameters of opened pores were calculated from parameters of lognormal function using the formulas below [21]: 2 𝑑𝑚𝑒𝑎𝑛 = 𝑑𝑠 𝑒 0.5𝜎 (3) 𝑑𝑚𝑒𝑑𝑖𝑎𝑛 = 𝑑𝑠 2

(4)

PT

ED

M

A

N

U

𝑑𝑚𝑜𝑑𝑒 = 𝑑𝑠 𝑒 −𝜎 (5) Given non-symmetrical nature of pore size distribution (skewness), mode diameter (𝑑𝑚𝑜𝑑𝑒 ) calculated according to the eq. 5 have been used for pore size characterization of synthesized PVFs and HHs. More detailed information about pore morphology and pore structure of PVF, acrylic hydrogels, and HHs were obtained based on microphotos taken on a high-resolution scanning electron microscope (HR-SEM) Tescan Mira 3 LMU and on SEM JSM-6060 LА (JEOL, Japan). For the microscopical study the dried polymer specimens were attached to standard stubs with adhesive conductive tapes and coated with Au/Pd (25 nm thick) using an ion-sputter coater (Gatan Pecs 682). Fields of interest were imaged using secondary electron detector with 10 kV accelerating voltage. For microscopical studies specimens dried in the air at RT, at critical points of CO2 using Tousimis Critical Point Dryer (Samdri-780A), and lyophilized in sublimation devise UZV-10 (Kharkiv, Ukraine) were compared. The critical point drying (CPD) and freeze-drying techniques were employed to avoid the severe deformation and collapse of structure, which occurs during air drying due to surface tension effects. The CPD was performed in CO2 using ethanol as an intermediate fluid. Mechanical properties: tensile stress and elongation Mechanical properties of synthesized polymer systems were characterized using a tensile testing machine (Zapadpribor LLC, R-50). Specimens which were swollen in distilled water at RT to the equilibrium state were placed in the testing machine and slowly extended at a rate of 20 mm/min until it fractures. Initial length of the specimen between clamps, 𝑙0 , was 25 mm, width – 10 mm, and thickness – approximately 1.5 mm. Tensile stress 𝜎 (Pa) was calculated as the ratio of the applied force 𝐹 (N) to the cross-sectional area of the sample 𝐴𝑐𝑠 (m2) as follows 𝐹 𝜎=𝐴 ∙ (6)

A

CC E

2.5.

𝑐𝑠

Elongation 𝜀 (%) was calculated using a formula below: 𝑙−𝑙 ∆𝑙 𝜀= 𝑙 𝑜=𝑙 , 𝑜

𝑜

(7)

where 𝑙 is a sample length (mm) deformed at a certain time point; 𝑙𝑜 is an initial sample length (mm). 2.6.

Equilibrium swelling degree and swelling kinetics Kinetics of HHs swelling and its components were studied by the gravimetric method [22] in distilled water and buffer solutions with pH 1.68 and 9.18. The kinetic data were plotted as 𝑄𝜏 − 𝜏 curves or as normalized curves (𝑄𝜏 ⁄𝑄∞ ∙ 100 % 𝑣𝑠. 𝜏), where 𝑄∞ – equilibrium swelling degree (ESD) (g/g); and 𝑄𝜏 – Page | 5

swelling degree (g/g) at time τ. The swelling rate 𝑟𝑄 of polymer systems is assumed to follow first-order kinetic model with the following equation used to describe the swelling process [23]: 𝑑𝑄 𝑟𝑄 = 𝑑𝜏𝜏 = 𝑘𝑄1 (𝑄∞ − 𝑄𝜏 ) (8) where 𝑘𝑄1 is the swelling rate constant for first-order model (min-1). After the integrating the equation the following expression is obtained: 𝑄 ln (1 − 𝑄 𝜏 ) = −𝑘𝑄1 𝜏 (9) ∞

N

U

SC R

IP T

which can be rearranged as 𝑄𝜏 = 𝑄∞ ∙ (1 − 𝑒 −𝑘𝑄1 𝜏 ) (10) Swelling rate constant 𝑘𝑄1 for synthesized PVF, hydrogels, and HHs can be further calculated as a slope of a tangent to a plot ln(1 − 𝑄𝜏 ⁄𝑄∞ ) versus τ. Additionally other models for describing swelling kinetics, such as second- and third-order models, models for phase-boundary controlled reactions (with contracting area or volume), and three-dimensional diffusion models based on Jander or Ginstling-Braunshtein equations were analyzed and compared to the first-order kinetic model (see SI: A, Table A2) [24]. Given that a previous history of swelling, particularly pH of solutions, affects the swelling kinetics [25], all samples were kept in distilled water for a few days and dried at room temperature previously to the kinetic studies. Since the kinetics of swelling is determined primarily by the diffusion of solvent into the hydrogel, it depends on the geometrical dimensions of a specimen, ceteris paribus. Considering this, all hydrogels for kinetic studies had a cylindrical shape with a diameter and height of 10.0 mm and 1.5 mm respectively; while PVA acetals and HHs were 10.0 mm in diameter and 25.0 mm tall2. The swelling rate and acceleration were calculated as a first and second derivatives of 𝑄𝜏 on 𝜏, respectively. Given the decrease of swelling rate over time due to the decrease in osmotic pressure inside the polymer system for comparison of different crosslinked polymers newly established time independent parameters (still dependent on dimensions) were defined and calculated as follows 𝑟𝑄𝑚𝑎𝑥 = 𝑘𝑄1 𝑄∞ (11) 2 𝑟𝑄𝑚𝑎𝑥

A

ln 2

𝜏1⁄ = 𝑘

𝑄1

(12)

ED

M

is the maximum rate of swelling at 𝜏 → 0 (g/(g·min)); 𝜏1⁄2 is the half-swelling or the time taken for the swelling degree of a hydrogel system to reach a half of its ESD value (min). Non-equilibrium 𝑄𝜏 and equilibrium swelling degree 𝑄∞ of samples were calculated using the equations below: 𝑚 −𝑚 𝑄𝜏 = 𝜏𝑚 0 (13) where

𝑄∞ =

𝑚−𝑚0 𝑚0

PT

0

(14)

Sorption capacity and kinetics It is well-known that diffusion of substances through a polymer mesh is determined by physical properties of the mesh and the interaction between the polymer and the substance. Sorption of organic dyes belonging to various classes (namely, MB and FL) was studied spectroscopically using a UV-vis spectrometer (SPECORD M40, Carl Zeiss). Briefly, a weighted specimen of a PVA acetal, hydrogel, or hybrid system based on PVF and stimuli-sensitive hydrogel was loaded inside a 10-mL syringe (Fig. 2). Then 5 mL of aqueous solution of an individual organic dye or its mixture was drawn into the syringe. After a predetermined amount of time, solution was squeezed out, and its absorbance was measured at 480 nm for FL, and at 670 nm for MB. Dye concentrations were calculated by dividing the absorbance value by the slope of the calibration curve3 [18]. Non-equilibrium 𝑆𝜏 and equilibrium sorption degree S∞ were calculated as [10]

A

2.7.

CC E

where 𝑚0 – mass of dry sample (g); 𝑚𝜏 – mass of swollen sample (g) at time τ (min); m – mass of swollen sample (g) at equilibrium at 𝜏 → ∞.

2

Dimensions are given for swollen samples in distilled water at RT. Slope values for FL and MB for a dye concentration up to 3·10-4 wt. % are 5465.7 ± 128.3 (0.05 wt. % NaOH aqueous solution) and 1760.9 ± 18.2, respectively. Page | 6 3

𝐴0 −𝐴𝜏

𝑆𝜏 =

𝐴0

S∞ =

,

𝐴0 −𝐴∞ 𝐴0

(15) ,

(16)

SC R

IP T

where 𝐴0 is the absorbance of dye initial solution, (a.u.); 𝐴𝜏 – the absorbance of dye solution after sorption (a.u.) collected at time τ (min); 𝐴∞ – the absorbance of dye solution (a.u.) at equilibrium at 𝜏 → ∞. The weight ratio between dye solutions and polymer systems was 10 : 1 unless stated otherwise. For interpretation of sorption data, many different kinetic models have been suggested. Among them, the widely used Langmuir kinetic model fits numerous cases and can be further simplified to first-order or second-order4 kinetic equations [9]. The sufficient and necessary conditions for such simplification of the Langmuir kinetics have been discussed in detail by Yu Liu and Liang Shen [26]. Azizian [27] has also done similar theoretical analysis of first- and second-order kinetic models. He, as well as Liu and Shen, also showed that the sorption rate constants for first-order and second-order mechanisms depend on initial concentration of solute. The first-order rate equation for sorption (𝑟𝑆 ) can be expressed as follows 𝑑𝑆 𝑟𝑆 = 𝑑𝜏𝜏 = 𝑘𝑆1 (𝑆∞ − 𝑆𝜏 ) (17) -1 where 𝑘𝑆1 is the sorption rate constant for the first-order mechanism (min ). Integration of the above equation yields 𝑆 ln (1 − 𝑆 𝜏 ) = −𝑘𝑆1 𝜏 (18) ∞

𝜏

∞

The eq. 20 can be rearranged as 𝜏 1 𝜏 = 2 + 𝑆𝜏

𝑘𝑆2 𝑆∞

𝑆∞

M

∞

A

N

U

For second-order mechanism the following expression can be applied: 𝑑𝑆 𝑟𝑆 = 𝑑𝜏𝜏 = 𝑘𝑆2 (𝑆∞ − 𝑆𝜏 )2 where 𝑘𝑆2 is the sorption rate constant for the second-order mechanism (min-1). Integration of the eq. 19 yields 1 1 − 𝑆 = 𝑘𝑆2 𝜏 𝑆 −𝑆

(19)

(20) (21)

25 ℃ 1000∙(𝐶0 −𝐶∞ )𝑉𝜌𝐻 2𝑂

𝑚0

=

1000∙𝑆∞ 𝐶0 𝑚0

PT

𝑆𝑐 =

ED

Thus, the linear plot of ln(1 − 𝑆𝜏 ⁄𝑆∞ ) or 𝜏⁄𝑆𝜏 versus time τ can be used to determine the sorption rate constants, and for the latter case – an equilibrium sorption degree. The equilibrium sorption capacity 𝑆𝑐 (µg/(g of dry gel)) for HH systems was calculated based on experimental data as (22)

Chemical structures and modeling MarvinSketch (version 17.4.3, 2017) was used for drawing and displaying of chemical structures and for chemical modeling; ChemAxon (http://www.chemaxon.com). For predicting the properties of model dyes the physicochemical plugins including pKa, hydrophilic-lipophilic balance (HLB), logP, logD, isoelectric point (IEP), dipole moment, and Van der Waals surface area (3D) have been computed and used to explain the sorption data. The calculations were performed with the default settings unless stated otherwise.

A

2.8.

CC E

where 𝐶0 is the initial concentration of a dye (wt. %); 𝐶∞ is the equilibrium concentration of a dye (wt. %) after sorption; V is the volume of dye solution used in the sorption experiment (mL); 𝜌𝐻252 𝑂℃ – water density at 25 ℃ (0.997 g/mL).

2.9.

Statistical data analysis The data is presented as an arithmetic mean ± standard error (SE). The SE was estimated as5

4

In literature, these models are very often referred as pseudo first-order and pseudo second-order kinetic models, but we are going to omit this vague terminology. 5 The systematic error was assumed to be negligible. Page | 7

2

𝑠

𝑆𝐸 = √(1.96√𝑁 𝑡0.05,𝑑𝑓 ) + (∆𝑦)2

(23)

where 𝑠 is the sample standard deviation; 𝑁 is the number of parallel experiments; 𝑡𝛼,𝑑𝑓 is the critical value of two-tailed t-distribution for α = 0.05 with defined degrees of freedom (𝑑𝑓 = 𝑁 − 1); ∆𝑦 – the error of indirect measurements, which was calculated using the numerical differentiation method [28]: 𝜕𝑓 ∆𝑦 = ∑𝑛𝑖=1 |𝜕𝑥 | ∆𝑥𝑖 (24) 𝑖

SC R

IP T

where 𝑓 is a function of directly measured variables 𝑥𝑖 ; ∆𝑥𝑖 is the error of direct measurements. Differences between groups were examined for statistical significance using two-tailed t-test for a significance level 𝛼 = 0.05. All calculations were performed using Microsoft Excel 2016, while OriginPro (OriginLab, Northampton, MA) was used for statistical analysis, nonlinear and linear fitting, plotting graphs, analyzing peaks, and graphical residual analysis. RESULTS AND DISCUSSION Polyvinyl formal (PVF) sponge, AA- and NIPAAm-based hydrogels, and hybrid porous polymer systems based on PVF and acrylic hydrogels as well as their components (monomers and laponite) were studied using FTIR-spectroscopy (Fig. 3, Fig. 4, and Fig. A1–A4 (see SI: A)) [29]. As we know based on analogy with the spectra of simple molecules and on the calculated characteristic frequency of various molecular groups, bands corresponding to characteristic vibrations of sp3-hybridized C– H bonds (methyl and methylene groups) are present in three regions: 2800–3000, 1300–1400 and below 700 cm−1, while alkene (sp2-hybridized) C–H stretching vibrations are observed above 3000 cm−1. Absorption bands related to skeleton vibrations are of low intensity and appear at 700–1100 cm−1 (stretching vibrations) and below 500 cm−1 (deformational vibrations). Absorption arising from alkane (sp3-hybridized) C–H stretching occurs in the region 2800–3000 cm-1 and corresponding bands can be found in IR spectra of all polymers and monomers, but AA (Fig. 3, Fig. 4, and SI: A “FTIR spectroscopy”). Molecules containing methyl and methylene groups according to theoretical calculations are supposed to have two distinct bands for each group, which can be observed at 2962 and 2872 cm-1, and at 2926 and 2853 cm-1, respectively [29]. The first of these bands results from the asymmetrical (as) stretching mode, while the second band arises from symmetrical (s) stretching. Bands assigned to sp3-hybridized methyl and methylene vibrations are listed in Table A3 (see SI: A). Methyl group has also two bending vibrations: the symmetrical one (δsCH3) occurs near 1375 cm-1, and the asymmetrical (δasCH3) — near 1450 cm-1 [29]. Configurations in which two methyl groups are attached to the same carbon atom (gem-dimethyl groups) exhibit distinctive absorption in the C–H bending region. Thus, the isopropyl group of NIPAAm unit shows a strong doublet in spectra of hydrogel systems and NIPAAm monomer (Fig. 3b-d) with the peaks of almost equal intensity at 1365–1368 and 1386–1388 cm-1. It is interesting to note that the bands in the IR spectrum of PVF at 1362, 1388, 1404, and 1430 cm-1 (Fig. 3a) are located in the characteristic region for the deformation frequencies of the methyl group [30]. This can suggest the presence of methyl groups in polymer structure, more probably due to residual acetate units. This can be further confirmed be a presence of a low-intensity band at 1735 cm-1, which we attributed to the stretching vibration of the carbonyl group (Fig. 3a, c). For methylene group four bending vibrations known as scissoring, rocking, wagging, and twisting are usually observed in the 1150–1470 cm-1 region and near 720 cm-1. The scissoring vibrations of methylene group in the spectrum of PVF can be assigned to a band at 1473 cm-1, while in spectra of other thermosensitive polymers and NIPAAm this band is overlapping with a band at 1460 cm-1, which was assigned to in-plane NH bending vibrations. Absorbance of NIPAAm monomer (Fig. 3d) due to the presence of monosubstituted alkene double bond (–CH=CH2) was assigned to a band at 1622 cm-1. The vinyl group also produces three closely spaced C–H stretching bands which can be found in the spectrum. Two of these bands result from symmetrical and asymmetrical stretching of the terminal C–H groups (=CH2) at 3102 and 2936 cm-1, and the third band — at

A

CC E

PT

ED

M

A

N

U

3.

Page | 8

A

CC E

PT

ED

M

A

N

U

SC R

IP T

3030 cm-1 (weak, not shown) due to the stretching of the remaining single C–H bond (=CH–R) (see SI: A, Fig. A2–A4). Alkene C–H bonds can also undergo bending either in the same plane as the C=C bond or perpendicular to it. Thus, the absorbance of NIPAAm at 1410 cm-1 can be assigned to a scissoring vibration of the terminal methylene (in-plane bending vibration of the vinyl groups). Bands at 918 and 992 cm-1 can be assigned to out-of-plane C–H bending vibration, while bands at 2714, 2789, 3074 and 3172 cm-1 were hard to attribute to any specific group. By comparison the spectrum of monomer with the spectra of hydrogels one can see the weak absorbance at 1414–1418 cm-1. Appearance of this band (in-plane δCH) may suggest the presence of unpolymerized NIPAAm units with relatively low concentration, while the absence of band at 1622 cm-1 (C=C stretching vibrations, which can be masked by its overlapping with stronger bands) and the lack of 1980 cm-1 vibration mode (overtone of out-of-plane δCH) may indicate the absence of alkene double bonds. But unfortunately, due to low intensity and overlapping of bands corresponding to double bond we could not prove or disprove its presence in synthesized polymer systems. Comparing IR spectra of hydrogels and NIPAAm (Fig. 3b and c) with the spectrum of PVF (Fig. 3a) one can also see they differ substantially in the region of amide fragment vibrations. As we know amides have two absorption bands called Amide I and Amide II. The first band is caused by a complex vibration of carbonyl group along with the C–N bond and C–C–O and C–N–C angles. Amide I absorption band appears at 1658, 1641, and 1647 cm-1 for NIPAAm monomer, hydrogel, and HH respectively. Another amide band presented in primary and secondary amides is related to the deformation vibrations of N–H bond. Secondary amides display the Amide II band in the region of 1536–1550 cm-1 (Fig. 3b, c, and d). Low frequency shift of the absorption bands (Amide I and II) in hydrogel systems can be a sign of its association via H-bond mechanism or cyclic fragment formation. Amide II band results from interaction between the N–H bending and the C–N stretching of the C–N–H group and split into two bands for NIPAAm-based hydrogel and HH due to the combination of vibrations of NIPAAm and MBAAm units. A second, weaker band near 1240–1246 cm-1 can be a result of coupling of the N–H bending and C–H stretching. The C–N stretching band of amides occurs near 1456–1460 cm-1 while a broad, medium band at 716 cm-1 in the spectrum of NIPAAm (Fig. 3d) can be assigned to out-of-plane N–H wagging [29]. Moreover, in IR spectra of secondary amides, which exist mainly in the trans configuration, the N–H stretching vibration of NIPAAm monomer are observed at 3299 cm-1 and at 3172 cm-1. The latter band can be caused by association of secondary amides. In hydrogel systems, the frequency decrease to 3275–3279 cm-1 and probably can be caused by the inter- and intramolecular hydrogel bonding formation. Another difference between PVF and NIPAAm hydrogel can be a broad characteristic absorption peak centered at 3426-3432 cm-1 in PVF spectrum which originates from the stretching vibration of the O–H groups (Fig. 3a, Fig. 4a). Broadening of the band was shown to be caused by hydrogen bonding. Absorbance at 1008 cm-1 can be attributed to C–O stretching vibration of C–OH groups or C–O–C groups in a six-membered ring coupled with the adjacent C–C stretching vibration. This can suggest the presence of unreacted polyvinyl alcohol units in PVF sponges. In liquid or solid state, carboxylic acid exists as dimers (cis-cis and trans-trans conformations) or “polymeric” structures due to strong hydrogen bonding (Fig. A2-A4 and Fig. A5a, b (see SI: A)) [31]. The exceptional strength of the hydrogen bonding is explained on the basis of the large contribution of the ionic resonance structure (see SI: A, Fig. A5c). Because of the strong bonding, intermolecular carboxylic acid dimers/“polymers” display very broad intense O–H stretching (centered around 3050–3076 cm-1, Fig. 4b, d; Fig. A2–A4 (see SI: A)) with the weaker C–H stretching bands being superimposed upon the broad O–H band (Fig. 4 and Table A3 (see SI: A)). Structural confinements in PVF and pH-sensitive hydrogel systems lead to decrease in the degree of association of hydroxyl groups and weaken the intermolecular hydrogen bonds. This cause the shift of O–H absorption band to 3396, 3398, or 3426 cm-1 in case of pH- sensitive HH, PAAH or PVF, respectively. For monomeric form of acrylic acid with nonhydrogen-bonded or “free” hydroxyl groups the corresponding band can be usually observed at 3550 cm-1. Strong band at 1732 cm-1 is associated with the C=O absorption of the monomeric form of acrylic acid, and it is split into a triplet with each frequency component been separated by about 16 cm-1 (Fig. 4d). Bands at 1698 and 1714 cm-1 correspond to C=O absorption of cis-cis and trans-trans conformations of AA, while two absorptions at 1614 and 1634 cm-1 are attributable to the C=C stretching vibrations. Additionally, absorption at 1434 and 1298 cm-1 can be explained by coupling between in-plane O–H bending and C–O stretching of dimers, and the broad band of medium intensity at 928 cm-1 due to O–H out-of-plane bending. As one can see in Fig. A2 (see SI: A) for pH-sensitive HHs the intensity of all bands increases with Page | 9

A

CC E

PT

ED

M

A

N

U

SC R

IP T

increasing the amount of incorporated acrylic hydrogel component. The corresponding bands represent the combination of AA and MBAAm vibrational modes, with bands at 1646 and 1654 cm-1 being assigned to Amide I of MBAAm (Fig. 4c). Thus, the detailed analysis of IR spectra of synthesized polymer systems revealed the successful incorporation of thermo- and pH-sensitive units to the composition of porous PVF as well as the presence of hydroxyl and probably acetate groups in PVF and HH systems. Obtained results are in good agreement with the predicted one. To examine the porous structure of polymer systems and to determine the effect of hydrogel incorporation on pore size and its distribution in hybrid hydrogels the further research have been performed. As can be seen in Fig. 5a synthesized crosslinked polymers based on PVF and stimuli-responsive acrylic hydrogels have a well-developed system of open pores with diameters up to 1 mm. The average pore diameter for dried PVA acetals and pH-sensitive HH-25 are 22.0 ± 1.6 µm and 22.3 ± 1.2 µm, respectively (Fig. 5c). The pore diameter increases slightly to 25.1 ± 1.9 µm for the HH-25 polymer system swollen in distilled water, and for the most part does not depend on pH. Thus, the modification of PVF with pH-responsive hydrogels as well as with other acrylic hydrogels allows functionalization of PVF surface without affecting the structure of pores of synthesized HHs. For more detailed microscopical study SEM have been employed. In Fig. 6 and Fig. A6 (see SI: A) pores with diameter approximately 10 and 500 µm can be observed. One can also notice that mostly the pore surface was modified by acrylic hydrogels in case of hybrid hydrogel systems with partially filled pores (Fig. 6c, d). The pores inside the walls almost did not get affected (Fig. 6), while the pores on the wall surface got filled with hydrogel composition (Fig. 7). The exception is HH-100 system (Fig. 6b). For this hybrid hydrogel the pores inside walls are completely filled with hydrogel composition what caused their collapse and distortion of pore system during drying. It worth mention that method of drying effects the observed pore structure (see SI: A, Fig. A7). Drying of sample in the air cause the severe deformation and collapse of macro- and microstructure due to surface tension effects. Drying at critical point of CO2 allows decreasing the deformation to some extent, with freezedrying being the most suitable method for studying of polymer pore structure. In case of PVF the difference between drying methods are insignificant, while for the hybrid hydrogels, and especially for acrylic hydrogels the drying method plays a crucial role. Comparison of polymer walls of PVF and PAAH shows that for latter system the pore walls are very thin (less than 1 µm), non-porous, and non-rigid. Pore size distribution for PAAH is more uniform with pore diameters approximately a few tens of micrometers (around 50 µm). In general modification of PVF with acrylic hydrogel decrease the size of pores inside the walls for all HH systems and also fill the larger pores with diameters approx. 500 µm. The synthesized polymer systems belong to porous materials with pore diameters from a few micrometers to a few hundred micrometers. Nanopores have not been found in synthesized systems. A considerable increase in the tensile strength of hybrid materials in comparison to PVF or hydrogels alone was also observed. Functionalization of PVF porous space with a pH-sensitive hydrogel based on poly(acrylic acid) (PAA) approximately doubles the tensile strength of the composite compared to the original unmodified sponge (Fig. 9a). In contrast to the hydrogel system, the HH is approximately an order of magnitude stronger (Fig. 9a). An alternative approach to interpenetrating polymeric meshes for enhancement of hydrogel mechanical properties is the incorporation of nanosized fillers to the hydrogel composition. As seen in Fig. 9, the tensile strength of the hydrogel based on polyNIPAAm (physically crosslinked) exceeds the strength of the AA-based hydrogel systems (crosslinked with MBAAm) fourfold, while elasticity increases approximately by two orders of magnitude. Consequently, synergetic improvement of hybrid material strength is also achieved when the physically cross-linked NIPAAm-based hydrogel is used for PVF pore space functionalization. As a result, HHs with an open hydrophilic pore system, and improved mechanical properties have been synthesized and can be potentially used in a variety of biomedical and technological applications (e.g., for the development of long lasting sorbent systems). Besides mechanical properties the application of hybrid hydrogels also depends on their swelling characteristics. For example, hydrogels with high degrees of swelling cannot withstand their own weight, show no mechanical integrity, and cannot be used for any practical application [32]. Thus, for determination of the swelling properties of synthesized systems, the equilibrium swelling degree and swelling kinetics of PVF, hydrogels, and HHs (with varying degrees of functionalization) have been studied in distilled water, and buffer solutions with рН 1.68 and 9.18 (Fig. 10, Fig. 11). Page | 10

A

CC E

PT

ED

M

A

N

U

SC R

IP T

As can be seen in Fig. 10a, ESD of PVF sponges in distilled water at room temperature is 19.5 ± 0.8 g/g, and its swelling degree is 1.5 times higher than for the poly(acrylic acid) hydrogel (PAAH). However, the ESD is almost equivalent to that of the hydrogel based on polyNIPAAm. The ESD of HHs increases with increasing amount of AA units to a certain concentration (HH-38), after which the swelling of hybrid systems starts to decrease (Fig. 10a). Extremum dependence of the ESD on degree of filling with a pHsensitive hydrogel can be explained by competition of two opposite factors. For hybrid systems with low degrees of filling (approximately 40 % and less) the number of pH-sensitive units decreases, causing an overall decrease in pH-sensitivity of HHs. On the other hand, HHs with degree of filling more than 40 % have a higher pH-sensitivity due to an increasing number of AA units. However, after pH-sensitive hydrogels become surrounded completely by PVF mesh, gel swelling is counterbalanced by the constraining force of the rigid mesh. In general, functionalization of PVF pore space with a pH-sensitive hydrogel (22–69 %) improves water sorption capacity of HH systems and increases their ESD in alkaline solution, while in acidic medium ESD of HHs decreases drastically compared to the initial unmodified PVF. The observed pH-sensitivity is explained by protonation-deprotonation of carboxyl groups around the 𝑝𝐾𝑎 6 of PAA [32]. As a result, the pHsensitive systems collapse in solutions with pH lower than 𝑝𝐾𝑎 due to hydrogen bond formation. On the other hand, in solutions with pH higher than 𝑝𝐾𝑎 , they tend to swell due to the carboxyl group deprotonation, which leads to an increase of both the osmotic pressure inside the polymer system and the electrostatic repulsion between carboxyl groups. As can be seen from the experimental data, the change in the swelling degree of pH-sensitive systems is the result of a delicate balance between the pH of the swelling medium, the pKa of the polyelectrolyte network, and various structural parameters of the crosslinked system [32]. Also, one should note that HHs partially filled with a pH-sensitive hydrogel have higher swelling degrees compared to the fully filled hybrid polymer (HH-100). Among them, HH-38 has the highest ESD, at pH 9.18, its swelling degree is the same as for PAAH, while at lower pH, HH-38 swells 1.5-2.0 times more than the corresponding hydrogel. In Fig. 10b a temperature dependence of the swelling degree for thermosensitive hydrogel systems is shown. A phase transition from a swollen to a collapsed state occurs around 35 ℃ for synthesized hybrid polymers and NIPAAm-based hydrogel, while the ESD of PVF remains temperature independent. The intensity of phase transition (difference between the swollen and collapsed states) of HHs does not depend on the number of incorporated thermosensitive units into HHs composition (within 42–100 % range). Approximately twofold weight change can be observed in case of HHs, whereas pure polyNIPAAm hydrogel system is characterized by an order of magnitude weight change7. Hence, the properties of unfunctionalized PVA acetals do not depend on pH, or temperature; and its dependence on external stimuli also confirms the successful incorporation of PAA or polyNIPAAm into the HH composition. Besides equilibrium swelling degree, swelling kinetics can also determine the application of hydrogel polymers. It is well-known that swelling kinetics varies from one hydrogel system to another, but even for one system the kinetic mechanism can depend on external conditions, such as temperature, pH or ionic strength of swelling medium. In paper [24], authors analyzed 16 different models for swelling kinetics and showed that a model for a phase-boundary controlled reaction (for contracting area) adequately describes the swelling of PAAH at a studied temperature range 25 – 45 ℃. To describe a swelling process of synthesized PVF, PAAH, and HHs, we chose the first-order model among seven tested kinetic mechanisms (see SI: A, Swelling kinetics and Model validation). The calculated kinetic parameters of swelling for synthesized systems are listed in Table A5–A7 (see SI: A). The assessment of the validity and quality of the linear fit for the first-order model was performed using a graphical residual analysis (see SI: A, Swelling kinetics and Model validation). The performed residual analysis revealed a non-random structure in residuals for PVF as well as for all HH systems (Fig. A15 and A16) and confirms the poor fit of experimental swelling data by the first-order model. In case of PAAH it was shown that the model describes swelling data adequately. Therefore, based on the results of graphical residual analysis, it was shown that among seven analyzed kinetic models only the first-order model reasonably describes the swelling process of PAAH at neutral and alkaline pH range. The differences in swelling kinetics between crosslinked polymer systems (see SI: A, Swelling kinetics According to the literature the 𝑝𝐾𝑎 of PAA is approximately 4.2. In our previous study, we showed that volume ESD (cm3/cm3) for thermosensitive hydrogel is approximately half of weight ESD (g/g) in a temperature range from 5 to 45℃. Page | 11 6 7

A

CC E

PT

ED

M

A

N

U

SC R

IP T

and Fig. 11) can be explained by the differences in polymer relaxation times, polymer-polymer and polymersolvent interactions, and structure of polymer systems. As can be seen in Fig. 11, pH-sensitive hydrogels as well as other hydrogel systems reach the ESD in at least 12 h, while unfilled PVF sorb more than 95 % of their maximum water content in just a few minutes and reach equilibrium in less than 1 h depending on pH. This can be explained by higher porosity and larger surface area of PVF in comparison to hydrogels. However, differences in hydrophilicity and stiffness of polymer mesh as well as pore space morphology of polymer systems at a microscale level can also affect swelling kinetic. Despite the difficulties in modelling the swelling process of HH systems, the functionalization of PVF pore space with pH-sensitive hydrogel not only improves the water sorption capacity of HH systems (Fig. 10a), but also improves the ability to selectively sorb positively charged species (e.g., positively charged dyes or cations [9]) due to electrostatic interactions. Moreover, pH-sensitivity of hybrid polymers can be further used to regenerate the sorbent by its switching from swollen to collapsed state by changing the pH of external medium. As it has been discussed earlier, sorption is the most commonly used technique for the treatment of industrial wastewaters. For its practical applications, it is necessary to know sorption characteristics of sorbents such as sorption rate and capacity under various process conditions. Thus, to evaluate the synthesized polymer systems as sorbents, the sorption of two model dyes of different natures (methylene blue and fluorescein) due to their ease of optical visualization was investigated. Kinetic models, namely, the first-order (eq. 18) and the second-order (eq. 21), for describing of sorption process of dyes from aqueous solutions have been also tested (Fig. A19). The sorption experiments were conducted in aqueous solutions of individual dyes or their mixture. For the latter experiment dyes mixture was prepared by combining the equal volumes of individual dye solutions of same normalized absorbance (1.865 cm-1) with the final concentration of MB and FL in dyes mixture (5.30± 0.06) ·10-4 wt. % and (1.71 ± 0.04) ·10-4 wt. %, respectively. Sorption of dyes from their mixture and the chemical structures of MB and FL are depicted in Fig. 12a. As can be seen, dye sorption depends on the dye type. For positively charged MB sorption degree reaches 98.0 ± 4.9 % in 3 min (after 2nd pump), while mostly neutral PVF sorbs only 3.2 ± 0.2 % of negatively charged FL8 (see SI: A, Fig. A23). The determined equilibrium sorption capacity for MB and FL is 72.0 ± 2.1 µg/g and 1.7 ± 1.1 µg/g, respectively. Using the second-order kinetic model (Fig. A19) the calculated equilibrium sorption degree and the sorption rate constant are, respectively, 98.23 ± 0.03 % and 0.4 ± 0.1 (min·%)-1 for MB, and 6.5 ± 0.2 % and 0.030 ± 0.007 (min·%)-1 for FL9. As a result, functionalization of PVA acetal with pH-sensitive hydrogel improves the sorption capacity of the HH system (Fig. 12b) and increases the sorption degree of FL from 3.2 ± 0.2 % to 95.1 ± 0.2 %, while the sorption capacity for MB (99.4 ± 0.2 %) remains almost the same within the error. The difference in the sorption process can be attributed mainly to the difference in the interaction between the dye molecules and polymer chains. Electrostatic interaction between tertiary amine and nelectrons of the oxygen atoms of PVF chain favors the sorption of MB, while hinder the FL sorption due to repulsion between deprotonated carboxyl and hydroxyl groups of the dye and n-electrons of the heteroatom in polymer chains (see SI: A, Chemical Modeling). To explain the increase of FL sorption by negatively charged HH matrix we need to take into consideration the structural changes of fluorescein molecules at different pH, which also affect UV-vis absorbance and fluorescence spectra (see SI: A, Fluorescein) [33, 34]. While carrying out the experiment we note, that the main absorption peak for dianion IV (see SI: A, Fig. A17–A18) in dyes mixture at 490 nm with a shoulder around 475 nm disappears and a new peak around 440 nm appears instead. As we know absorption at 440 nm is caused by FL cation I, which has the maximum absorbance at 437 nm (53,000 M-1·cm-1), with two additional peaks at 297 (7,1000 M-1·cm-1) and 250 nm (33,000 M-1·cm-1) [34]. The neutral species have maximum at 434 and 475 nm, but its contribution is negligible due to its low absorptivity: 11,000 M-1·cm-1 and 3,600 M-1·cm-1, respectively. Comparison of UV-vis spectrum of dye mixture after sorption with a spectrum of FL dye in pH buffer at 1.68 confirms the presence of cation I in aqueous solution. The conclusion was made based on the position of absorbance maximum and the shape of the peak [34, 35]. Thus, the sorption process of FL by negatively charged crosslinked polymer matrix can be explained by the protonation of dianion IV and its converting into cation I with further sorption of positively charged FL molecule by HH-25 due to electrostatic interactions. 8 9

𝑝𝐾𝑎 values for FL are 2.22, 4.34, and 6.68 [36]. The weight ratio between solutions and sorbents was 8.2 ± 0.2 g/g. Page | 12

U

SC R

IP T

In general sorption increases with increasing of initial dye concentration (see SI: A, Fig. A20) until it reaches saturation. As an example, increasing the weight ratio between dyes mixture and sorbents threefold (from 8.2 ± 0.2 g/g to 24.3 ± 0.5 g/g) results in increase of equilibrium sorption capacity for MB and FL to 251.7 ± 6.4 µg/g and 17.3 ± 3.3 µg/g for PVF, and to 256.7 ± 6.3 µg/g and 38.1 ± 2.3 µg/g for HH-25, respectively. Thus, functionalization of PVA acetal pore system with hydrogels allows modifying the surface chemistry while preserving its high permeability. As a result, HHs which combine the properties of polymeric systems they are composed of can be synthesized. Improvement of the properties can be seen in increasing of HH tensile strength and elasticity. In addition, the HH systems become pH-sensitive when PVA acetals are filled with hydrogels based on PAA or thermosensitive upon the filling with polyNIPAAm. Incorporation of stimuli-responsive hydrogel units allows changing swelling degree and sorption-desorption considerably via external stimuli. For рН-sensitive hybrid systems, ESD can be controlled by adjusting the pH of solution, while in the case of thermosensitive HHs, swelling degree can be controlled by changing the temperature. Both properties are of high importance for regeneration of a sorbent. Also, functionalization of PVF with acrylic hydrogels not only improves the sorption capacity of HH systems, but also makes sorbents more selective toward a targeted class of substances. Among synthesized polymer systems, pH-sensitive HHs, especially HH-38, have a potential for further functionalization, and can be used as sorbents, for example, for purification of industrial wastewater. However, wastewater composition, especially the composition of industrial wastewaters depends on numerous factors and it is much more complicated than the dye model systems studied here. Thus, the next step will be to study the sorption of other dye molecules and heavy metals for further validation of synthesized HHs for treatment of industrial wastewater. Additionally, a regeneration process for sorbents will be developed and tested. CONCLUSIONS In this work, the synthesis of highly porous hybrid hydrogels via functionalization of surface of PVA acetals with pH-sensitive or thermosensitive hydrogels was documented. Synergetic improvement of physicochemical properties such as tensile strength, elasticity, swelling and sorption parameters of hybrid hydrogels has been shown. A maximum synergy in hybrid material properties is achieved by partial pore filling of PVF with hydrogel and corresponds to HH-38 based on PAAH. Designed hybrid hydrogels can be potentially used as sorbents for water treatment, for example, to remove the impurities from industrial wastewater or in separation processes. Several kinetic models to describe swelling and sorption processes were analyzed and validated based on results of nonlinear and linear fitting, and graphical residual analysis. It was shown that among all tested models, the first-order model reasonably describes the swelling kinetics of PAAH at pH » 𝑝𝐾𝑎 , while at lower pH and for other polymer systems the first-order model as well as other models fit data poorly. For dye sorption, the second-order model describes the experimental data well, while the first-order model has a poor fit.

PT

ED

M

A

N

4.

CC E

Authors’ Agreement/Declaration. Authors warrant that the article is their original work, has not received prior publication, and is not under consideration for publication elsewhere. All authors have seen and approved the version of the manuscript being submitted.

A

Authors’ contributions. YuS and OK supervised this work, helped in the analysis and interpretation of data, and, together with SK, LK, and OL worked on writing and revising the manuscript. SK and TP carried out the synthesis of polymer systems and its characterization (mechanical properties and swelling behavior). NP performed sorption experiments. OL characterized samples using FTIR spectroscopy, while MS performed microscopical study using SEM. Funding. This research did not receive any specific grant from funding agencies in the public, commercial, or not-forprofit sectors. Conflict of Interest. The authors declare that there's no financial or personal interest or believes that could affect their objectivity. Acknowledgments. Authors thank Victor Marik and Olivia Lanier for participating in discussions, and providing language help, writing assistance and proofreading of the article; and Vitalii Starchenko as well as journal reviewers for reviewing the article. Also, authors acknowledge financial support from National Academy of Sciences of Ukraine (NASU) and Ernest Samas.

Page | 13

A

CC E

PT

ED

M

A

N

U

SC R

IP T

REFERENCES [1] D.E. Discher, D.J. Mooney, P.W. Zandstra, Growth factors, matrices, and forces combine and control stem cells, Science. 324 (2009) 1673–1677. doi:10.1126/science.1171643. [2] K.Y. Lee, D.J. Mooney, Hydrogels for tissue engineering, Chem. Rev. 101 (2001) 1869–1879. doi:10.1021/cr000108x. [3] Y. Qiu, K. Park, Environment-sensitive hydrogels for drug delivery, Adv. Drug Deliv. Rev. 64 (2012) 49–60. doi:10.1016/j.addr.2012.09.024. [4] F. Hapiot, S. Menuel, E. Monflier, Thermoresponsive hydrogels in catalysis, ACS Catal. 3 (2013) 1006–1010. doi:10.1021/cs400118c. [5] X. He, M. Aizenberg, O. Kuksenok, L.D. Zarzar, A. Shastri, A.C. Balazs, J. Aizenberg, Synthetic homeostatic materials with chemo-mechano-chemical self-regulation, Nature. 487 (2012) 214–218. doi:10.1038/nature11223. [6] X. Zhang, C.L. Pint, M.H. Lee, B.E. Schubert, A. Jamshidi, K. Takei, H. Ko, A. Gillies, R. Bardhan, J.J. Urban, M. Wu, R. Fearing, A. Javey, Optically- and thermally-responsive programmable materials based on carbon nanotube-hydrogel polymer composites, Nano Lett. 11 (2011) 3239–3244. doi:10.1021/nl201503e. [7] A. Döring, W. Birnbaum, D. Kuckling, Responsive hydrogels - structurally and dimensionally optimized smart frameworks for applications in catalysis, micro-system technology and material science., Chem. Soc. Rev. 42 (2013) 7391–420. doi:10.1039/c3cs60031a. [8] G. Jing, L. Wang, H. Yu, W.A. Amer, L. Zhang, Recent progress on study of hybrid hydrogels for water treatment, Colloids Surfaces A Physicochem. Eng. Asp. 416 (2013) 86–94. doi:10.1016/j.colsurfa.2012.09.043. [9] J. Chen, W. Zhang, X. Li, Adsorption of Cu(II) ion from aqueous solutions on hydrogel prepared from Konjac glucomannan, Polym. Bull. 73 (2016) 1965–1984. doi:10.1007/s00289-015-1588-9. [10] H. Guo, T. Jiao, Q. Zhang, W. Guo, Q. Peng, X. Yan, Preparation of graphene oxide-based hydrogels as efficient dye adsorbents for wastewater treatment., Nanoscale Res. Lett. 10 (2015) 931. doi:10.1186/s11671-015-0931-2. [11] P. Calvert, Hydrogels for soft machines, Adv. Mater. 21 (2009) 743–756. doi:10.1002/adma.200800534. [12] G.J. Lake, A.G. Thomas, The strength of highly elastic materials, Proc. R. Soc. A Math. Phys. Eng. Sci. 300 (1967) 108–119. doi:10.1098/rspa.1967.0160. [13] M.S. Vasylieva, M.L. Malysheva, Y.M. Samchenko, A.Y. Papuga, Hydrogel nanocomposites with silver nanoparticles, French-Ukrainian J. Chem. 1 (2013) 69–71. http://kyivtoulouse.univ.kiev.ua/journal/index.php/fruajc/article/view/42 (accessed August 3, 2017). [14] P. Rao, I.M.C. Lo, K. Yin, S.C.N. Tang, Removal of natural organic matter by cationic hydrogel with magnetic properties, J. Environ. Manage. 92 (2011) 1690–1695. doi:10.1016/j.jenvman.2011.01.028. [15] J. Yun, D. Jin, Y.S. Lee, H. Il Kim, Photocatalytic treatment of acidic waste water by electrospun composite nanofibers of pH-sensitive hydrogel and TiO2, Mater. Lett. 64 (2010) 2431–2434. doi:10.1016/j.matlet.2010.08.001. [16] Y.D. Liu, H.J. Choi, Recent progress in smart polymer composite particles in electric and magnetic fields, Polym. Int. 62 (2013) 147–151. doi:10.1002/pi.4441. [17] P. Malaviya, A. Singh, Physicochemical technologies for remediation of chromium-containing waters and wastewaters, Crit. Rev. Environ. Sci. Technol. 41 (2011) 1111–1172. doi:10.1080/10643380903392817. [18] Y. Samchenko, Z. Ulberg, O. Korotych, Multipurpose smart hydrogel systems, Adv. Colloid Interface Sci. 168 (2011). doi:10.1016/j.cis.2011.06.005. [19] S. Kryklia, Y. Samchenko, V. Konovalova, T. Poltoratska, N. Pasmurtseva, Z. Ulberg, Hybrid pH- and thermosensitive hydrogels based on polyvinylalcohol and acrylic monomers, Magisterium. Chem. Sci. 63 (2016) 20–28. [20] D.A. Skoog, D.M. West, J.F. Holler, S.R. Crouch, Fundamentals of analytical chemistry, 8th ed., Thomson-Brooks/Cole, 2004. [21] Lognormal Distribution, (n.d.). http://www.itl.nist.gov/div898/handbook/eda/section3/eda3669.htm (accessed May 22, 2017). Page | 14

[29] [30] [31] [32] [33]

[34] [35]

A

CC E

PT

[36]

IP T

[28]

SC R

[27]

U

[26]

N

[25]

A

[24]

M

[23]

O. Korotych, Y. Samchenko, I. Boldeskul, Z. Ulberg, N. Zholobak, L. Sukhodub, Nisopropylacrylamide-based fine-dispersed thermosensitive ferrogels obtained via in-situ technique, Mater. Sci. Eng. C. 33 (2013). doi:10.1016/j.msec.2012.11.017. Z.X. Zhao, Z. Li, Q.B. Xia, E. Bajalis, H.X. Xi, Y.S. Lin, Swelling/deswelling kinetics of PNIPAAm hydrogels synthesized by microwave irradiation, Chem. Eng. J. 142 (2008) 263–270. doi:10.1016/j.cej.2007.12.009. B. Adnadjevic, J. Jovanovic, Novel approach in investigation of the poly(acrylic acid) hydrogel swelling kinetics in water, J. Appl. Polym. Sci. 107 (2008) 3579–3587. doi:10.1002/app.27451. A. Prior-Cabanillas, I. Quijada-Garrido, G. Frutos, J.M. Barrales-Rienda, Influence of the swelling history on the swelling kinetics of stimuli-responsive poly[(N-isopropylacrylamide)-co-(methacrylic acid)] hydrogels, Polymer (Guildf). 46 (2005) 685–693. doi:10.1016/j.polymer.2004.11.088. Y. Liu, L. Shen, From Langmuir kinetics to first-and second-order rate equations for adsorption, Langmuir. 24 (2008) 11625–11630. doi:10.1021/la801839b. S. Azizian, Kinetic models of sorption: A theoretical analysis, J. Colloid Interface Sci. 276 (2004) 47– 52. doi:10.1016/j.jcis.2004.03.048. C. Ferregut, S. Nazarian, K. Vennalaganti, C.-C. Chang, V. Kreinovich, Fast error estimates for indirect measurements: applications to pavement engineering, Reliab. Comput. 2 (1996) 219–228. R.M. Silverstein, F.X. Webster, D.J. Kiemle, Spectrometric identification of organic compounds., John Wiley & Sons, 2005. E.R. Blout, R. Karplus, The infrared spectrum of polyvinyl alcohol, J. Am. Chem. Soc. 70 (1948) 862– 864. doi:10.1021/ja01182a504. S.W. Charles, F.C. Cullen, N.L. Owen, G.A. Williams, Infrared spectrum and rotational isomerism of acrylic acid, J. Mol. Struct. 157 (1987) 17–29. doi:10.1016/0022-2860(87)87079-5. L. Brannon-Peppas, N.A. Peppas, Equilibrium swelling behavior of pH-sensitive hydrogels, Chem. Eng. Sci. 46 (1991) 715–722. doi:10.1016/0009-2509(91)80177-Z. B. Wang, W. Gao, Y. Ma, D. Li, L. Wu, L. Bi, Enhanced sensitivity of color/emission switching of fluorescein film by incorporation of polyoxometalate using HCl and NH3 gases as in situ stimuli, RSC Adv. 5 (2015) 41814–41819. doi:10.1039/C5RA05097A. R. Sjöback, J. Nygren, M. Kubista, Absorption and fluorescence properties of fluorescein, Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 51 (1995) L7–L21. doi:10.1016/0584-8539(95)01421-P. V.R. Batistela, J. da Costa Cedran, H.P. Mois de Oliveira, I.S. Scarminio, L.T. Ueno, A. Eduardo da Hora Machado, N. Hioka, Protolytic fluorescein species evaluated using chemometry and DFT studies, Dye. Pigment. 86 (2010) 15–24. doi:10.1016/j.dyepig.2009.11.002. S.A. Smith, W.A. Pretorius, Spectrophotometric determination of pKa values for fluorescein using activity coefficient corrections, Water SA. 28 (2002) 395–402.

ED

[22]

Page | 15

ED

M

A

N

U

SC R

IP T

Fig. 1. Schematic illustration of hybrid hydrogels based on PVF and acrylic hydrogels with unfilled, filled, and partially filled pores (from left to right).

A

CC E

PT

Fig. 2. Illustration of dye sorption experiment: pumping an aqueous dye solution through a 10-mL syringe filled with a hybrid hydrogel.

Page | 16

IP T SC R U N

A

CC E

PT

ED

M

A

Fig. 3. Infrared spectra of PVF sponge (a), chemically crosslinked NIPAAm-based hydrogel (b), thermosensitive HH-42 (c), and NIPAAm (monomer) (d).

Page | 17

N

U

SC R

IP T

Fig. 4. Infrared spectra of PVF sponge (a), PAAH (b), pH-sensitive HH-100 (c), and AA (monomer) (d).

A

CC E

PT

ED

M

A

Fig. 5. Photos of air-dried synthesized crosslinked polymers (a), pore maps (b), and pore size distributions fitted by lognormal function (eq. 2) (c) for PVF (1) and pH-sensitive HH-25 (2). Scale bar – 1 mm.

Fig. 6. Microphotographs (SEM) of PVF (a) and pH-sensitive hybrid hydrogels at different magnifications: HH-100 (b), HH-62 (c), HH-36 (d). Specimens were dried at critical point of CO2.

Page | 18

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 7. Microphotographs (SEM) of PVF (a) and pH-sensitive HH-75 (b) at different magnifications. Specimens were air-dried.

Page | 19

IP T SC R U N A M ED PT

A

CC E

Fig. 8. Microphotographs (SEM) of PVF (a), PAAH (d), and pH-sensitive hybrid hydrogels at different magnifications: HH-100 (b) and HH-62 (c). Specimens were freeze-dried.

Page | 20

IP T SC R

CC E

PT

ED

M

A

N

U

Fig. 9. Stress–strain curves for hybrid hydrogels (1), PVF (2), and acrylic hydrogels (3). Hydrogel and HH100 are based on covalently crosslinked poly(acrylic acid) (a) or physically crosslinked polyNIPAAm (b). Plotted lines are for guiding the eye.

A

Fig. 10. Equilibrium swelling degree of polymer systems at different pH (a) or at different temperatures in distilled water (b): PVF (1), HH-42 (2), HH-81 (3), HH-100 (4), and physically crosslinked thermosensitive hydrogel based on polyNIPAAm (5). Plotted lines are for guiding the eye.

Page | 21

IP T SC R U N A M ED PT

A

CC E

Fig. 11. Swelling kinetics of pH-sensitive HH-38 (1), PVF (2), and poly(acrylic acid) hydrogel (3) in water (a), and in buffer solutions with pH 9.18 (b) and 1.68 (c); and the plot of −𝒍𝒏 (𝟏 − 𝑸𝒕 ⁄𝑸∞ ) versus time for poly(acrylic acid) hydrogel at different pH (d). Solid lines correspond to first-order kinetic model (eq. 9), while dash lines are for guiding the eye.

Page | 22

IP T SC R

A

CC E

PT

ED

M

A

N

U

Fig. 12. Kinetic curves of dye sorption from their mixture by PVF (a) or pH-sensitive HH-25 (b). Dash lines are for guiding the eye. Insert: chemical structures of dyes (a) and photo of dyes mixture before and dye mixture and HH-25 after sorption (from left to right) (b).

Page | 23

Table 1. Composition of PVF. Component

m, wt. %

PVA Triton X-100 Formalin Sulfuric acid Distilled water

9.1 0.3 3.5 3.2 83.9

Table 2. Composition of stimuli-responsive hydrogels. pH-sensitive hydrogel

IP T

Thermosensitive hydrogel m, wt. %

n, mol % Component

m, wt. %

n, mol %

monomer

NIPAAm

20.1

4.972

AA

9.9

4.131

crosslinker

Laponite

15.1

0.547

MBAAm

0.1

APS TEMED

2.5 2.0

0.308 0.484

PPS

Distilled water

60.3

93.689

Distilled water

redox system

34.7 55.3

0.019 3.855

91.995

A

CC E

PT

ED

M

A

N

U

solvent

SC R

Component

Page | 24