Chapter 4: Porous Polymers by Ice Templating – Ice Templating and Freeze-Drying for Porous Materials and Their Applications

4
Porous Polymers by Ice Templating

4.1 Introduction

Porous materials, including polymers, are used in a very wide range of applications, including separation, catalysis, energy storage, tissue scaffolds, to name a few [14]. The pore sizes may be categorized as micropores (<2 nm), mesopores (2–50 nm), and macropores (>50 nm) [1]. The pores may be arranged either in order or at random. The pore shapes may be designed and synthesized for shape‐controlled molecular separation or catalysis [58]. A porous material containing pores of different sizes (particularly in different categories) is called a hierarchically porous material [1].

Mesoporous polymers and macroporous polymers are usually prepared by templating techniques [912]. The templates may be classified as hard (e.g. rigid objects such as colloids, particles, pre‐formed porous solid structures) or soft (usually the self‐assembled structures of surfactants/block copolymers or liquid drops) [11, 12]. Emulsion templating, where the liquid droplets are used as templates, has been an effective route to producing macroporous polymers and other materials [1315]. Gas foaming or supercritical fluid foaming are also efficient in fabricating highly interconnected macroporous polymers [9, 10]. Soft templating (mainly, micelles, microemulsions) [11, 12] and self‐assembly of block copolymers and associated processing techniques are widely known to generate mesoporous and/or microporous materials [16].

Microporous polymers, with pore sizes <2 nm, can be directly synthesized, with the potential to control pore shape and pore size. This has been a highly intensive research area, particularly due to the extremely high surface area, pore surface functionality, tailored pore shape for great potential in gas storage, gas separation, catalysis and biomedical applications [28]. Well‐known examples of microporous polymers include hypercrosslinked polymers [2, 17], polymers of intrinsic microporosity (PIMs) [3, 4], conjugated microporous polymers [5, 18], covalent organic frameworks (COFs) [57], and porous organic cages [6, 8]. Among them, COFs and porous organic cages are crystalline polymers, which is the basis for X‐ray diffraction and neutron diffraction studies to elucidate the pore structure and pore arrangement. It is also relatively easier to carry out simulation studies on these crystalline polymers [68].

A hierarchically porous polymer may be synthesized by employing dual or multiple templates or combining direct synthesis and templating [1, 6, 11]. Different templating techniques may be combined to tune porosity and pore morphologies. In this chapter, we introduce and focus on another templating technique, namely, ice templating and/or freeze‐drying, for the preparation of porous polymers. For the ice‐templating method, water (or other solvents) in a solution, suspension, or emulsion is frozen to form ice crystals. The ice crystals, acting as templates, are then removed (usually by freeze‐drying) to produce a porous structure (Figure 4.1) [10, 1921].

Figure 4.1 Schematic representation of an ice‐templating process for the preparation of a porous material.

Source: Qian and Zhang 2011 [10]. Reprinted with permission from John Wiley & Sons.

Porous polymers may be prepared by freeze‐drying solutions, suspensions, gels, and highly crosslinked polymer. When producing porous structures, ice templating is essential for aqueous solutions and suspensions where the solutes or particles may be excluded from the growing ice crystals and concentrated between ice crystals. For the weakly crosslinked gels, the flexible and loosely linked polymer chains may be partially pushed away by ice crystals, leading to the formation of ice‐templated structures. Indeed, the structures of the freeze‐dried gel and hydrated gel are usually different even if the freeze‐drying process is carefully performed [22]. However, for highly crosslinked polymers or framework materials, the rigid structure may prevent the templating effect caused by ice crystal growth. The freeze‐drying of such materials is mainly to maintain the highly porous structure (the sublimation of solid ice exerts much lower interface stress than drying by liquid phase evaporation). There are also some new developments in material preparation facilitated by the freezing process. For example, the freezing‐induced concentration of solutes can lead to surfactant self‐assembly and formation of mesoporous silica [23] or self‐assembly of peptides for peptide nanostructures [24]. Because these types of structures do not result from ice templating, they are not included in the discussion below.

4.2 Porous Polymers by Freeze‐drying of Solutions and Suspensions

4.2.1 Polymer Sponges

The sponge structure usually indicates a highly interconnected 3D macroporous structure. The distribution of pores in a sponge is usually random although local ordering may be present. Aqueous solutions or suspensions are mostly used for the preparation of polymer sponges. In brief, the suspensions/solutions contained in plastic vials (e.g. polystyrene, polypropylene, Teflon) or glass vials are placed in a cold environment such as a freezer or a cold room [2528], shelf of a freeze‐dryer with varying freezing rates [29, 30], or simply immersing in a cold liquid (e.g. liquid nitrogen) [2831]. The growth and orientation of ice crystals are not controlled but the local alignment is usually observed due to the presence of local temperature gradient.

Porous polymers can be readily prepared from aqueous solutions of hydrophilic polymers such as cellulose, chitosan, poly(vinyl alcohol) (PVA), etc [25, 28, 32]. However, these porous polymers are mechanically very week, shrink when exposed to moisture or in air, and rapidly dissolve when in contact with water or aqueous medium. This severely restricts their applications.

Cellulose nanofibers can be prepared by mechanical shearing and disintegrating of cellulose fibers with the widths in the range of 10–100 nm. Alternatively, bacterial cellulose nanofibers may be synthesized by disintegrating cellulose via the use of microorganism. The diameters are in the range of <100 nm but consist of much finer fibrils (2–4 nm). Cellulose nanocrystals are sometimes used as well, exhibiting crystalline defect‐free rod‐like structures. These nanocelluloses can be chemically modified to offer a variety of surface functionalities [33].

Cellulose nanofibrous foams or aerogel can be readily prepared by the ice‐templating technique from their nanosuspensions [34, 35]. The nanocellulose foams can be silyated to enhance surface hydrophobicity and hence the performance in the selective removal of oil from water [36, 37]. They are also widely used to enhance the mechanical stability of other foams by including cellulose nanofibers as additives in the formulations [38].

Great efforts in this area have been made in the preparation of porous scaffolds with useful biological properties. Silk fibroin is a natural fibrous protein with desirable properties such as high strength, elasticity, and haemocompatibility [39]. In addition to the use as a luxurious texture material, it has also been used as surgical suture thread and as scaffold for tissue engineering [26, 39]. Ice templating has been an effective method for the preparation of 3D silk fibroin scaffolds [29] and silk composite foams/gels, for example, poly(N‐isopropylacrylamide)/silk [40] and lactose‐silk fibroin conjugate sponge [26].

Biocompatible sponges have been prepared from collagen, collagen biowastes (extracted from animal skin waste) [41], and collagen‐glycosaminoglycan [29]. Wheat gluten was dispersed in water and then heated to 90 °C. The heated dispersion was homogenized to a 300% foam which was frozen and freeze‐dried. The foam could be chemically crosslinked by glutaraldehyde or thermally polymerized, exhibiting antimicrobial properties [28]. A solution of egg white powders was processed to generate a macroporous sponge that was chemically crosslinked to be used as scaffold for tissue engineering [27]. A 3D macroporous scaffold of poly(3,4‐ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) could be formed by ice templating [30]. Owing to its electrical conducting property and biocompatibility, this scaffold was used for the electrical control of protein conformation and for supporting cell growth [30].

Ice templating has been used to prepare highly interconnected macroporous materials that are advantageous for tissue engineering and liquid phase separation. However, due to the difficulty in forming nanosized ice crystals, the formation of ice‐templated materials with nanopores or mesopores has been highly challenging. Recently, the mixture solvents of tertiary‐butanol (TBA) and water was employed to produce porous polysaccharides (pectin, starch, and alginic acid) with both macropores and mesopores [42]. The macropores were formed by ice templates while the mesopores were attributed to the eutectic points of TBA–water. Figure 4.2 shows the phase diagram of the binary TBA–water system, exhibiting two eutectic compositions at ≈23 and 90 wt% TBA. For the porous polysaccharides prepared from the TBA–water mixture solvents (with different TBA compositions), the pore volumes measured by the Barrett–Joyner–Halenda (BJH) method (gas sorption) were the largest around the eutectic points (the maxima of 2.34 and 1.8 cm3 g−1 for the two eutectic points, respectively) [42].

Figure 4.2 The phase diagram of tert‐Butanol (TBA)‐water (compound α is a TBA hydrate and L = liquid) shows two eutectic points.

Source: Borisova et al. 2015 [42]. Reprinted with permission from John Wiley & Sons.

4.2.2 Aligned Porous Polymers by Directional Freezing

The significant development in ice‐templated porous materials in the last decade has been the development of directional freezing process for the preparation of aligned porous or layered materials [43, 44]. This process is also widely known as ‘unidirectional freezing’ or ‘ice segregation induced self‐assembly (ISISA)’ [21]. It has been employed to generate a wide range of aligned porous and composite materials with anisotropic properties [1921]. In principle, the directional freezing process is very simple, where a temperature gradient is applied across a liquid phase sample. The higher the temperature gradient, the faster the freezing rate and the smaller the ice‐templated pore sizes. The aligned pore structure is a result of directed growth of the ice crystals along the temperature gradient and the exclusion of molecules/particles from the ice front and concentration of molecules/particles between the ice crystals. This is demonstrated by the directional freezing of gold nanoparticle suspensions and is schematically represented in Figure 4.3 [45]. In practice, a vial containing solutions or suspensions can be dipped into liquid nitrogen (or another cold liquid) at a controlled rate. Alternatively, a vial or a mould can be placed on a cold finger, a cold plate, or sandwiched between two plates with a set temperature gradient. It is also possible to fabricate the substrates with surface patterns by the dipping method or via a motor‐driven freezing stage (where a substrate or glass slide with a spread of liquid film can be placed and moved between two temperature‐controlled metal plates/stages). The directional freezing technique has been applied to solutions or suspensions with water, organic solvents, and even compressed CO2.

Figure 4.3 (a) The scheme shows the directional freezing process. (b) An optical microscopic image shows the directional freezing of aqueous gold nanoparticle suspension. The white stripes are the orientated ice crystals while the red lines are excluded gold nanoparticles from ice crystals and concentrated between ice crystals. The blue dashed line indicates the interface of the frozen pattern and the liquid suspension.

Source: Zhang and Cooper 2007 [45]. Reprinted with permission from John Wiley & Sons.

4.2.2.1 Water‐based Systems

Aqueous PVA solutions were directionally frozen to fabricate the aligned fish‐bone like structure. It was demonstrated that the freezing rate could be varied to tune the aligned spacings of the frozen PVA solution with a computer‐controlled freezing stage as observed by optical microscopy [43]. By varying the molecular weight and hydrolysis degree of PVA, different freezing rates could be applied to tailor the morphologies of PVA, which could then be used for the controlled release of ciprofloxacin [46].

Another widely investigated hydrophilic polymer is chitin or chitosan. Chitin nanowhisker suspension was mixed with PVA and then processed to produce the aligned porous structures. The content of PVA could significantly influence the inter‐lamellar spaces (from 180 to 30 μm) from a PVA concentration of 0–2.0 wt% with an initial chitin nanowhisker concentration of 0.8 wt% [47]. 2.4 w/v% chitosan (low molecular weight) in 1 v/v% aqueous glacial acetic acid solution was directly frozen on a copper cold finger and then freeze‐dried. The aligned porous chitosan was used to direct neurite growth [48]. Recently, aligned porous chitosan with shape‐memory has been generated [49]. The diameters of the aligned channels could be tuned by varying the freezing temperature while the thickness of the channel wall could be controlled by chitosan concentration. When compressed, the channels squeezed the water out; they were deformed but not damaged and the shape could change back when the water was re‐absorbed into the channel. A variety of nanostructures including nanoparticles and nanowires could be readily incorporated and absorbed onto the channel wall mostly via electrostatic interaction. These structures could then be used to generate interesting macroscale assemblies [49].

Some other examples include cellulose and collagen [50, 51]. Cellulose was dissolved in aqueous NaOH/urea (5 wt%) and the solution was frozen from the bottom of the mode. Porous cellulose scaffolds with aligned columnar and open porosity were fabricated [50]. A wedge‐based system was used to produce a range of collagen scaffolds with unidirectional pores. The collagen suspension (0.7 w/v%) was prepared by incubating collagen fibrils overnight at 4 °C in 0.25 M acetic acid (pH 2.7) and homogenized on ice. The wedge‐based system consisted of an anodized aluminium wedge and a polyurethane wedge with thermal conductivities of 205 and 0.03 W (m K)−1, respectively. The freezing medium (liquid nitrogen) was in contact only with the polyurethane wedge. In addition to the large vertical temperature gradient, the wedge shape induced a small horizontal temperature gradient, likely facilitating local nucleation and growth of laterally directed ice crystals. This horizontal temperature gradient could result in small differences in height between adjacent upward growing ice crystals, thus stabilizing the upward growth of ice crystals by blocking inclined ice crystal growth [51].

4.2.2.2 Organic Solvent‐based Systems

The directional freezing process has also been applied to a range of organic solutions, mainly to prepare porous biodegradable polymers such as poly(ε‐caprolactone) (PCL) and poly(lactic‐co‐glycolic acid) (PLGA). For example, aligned porous PCL was prepared by directional freezing of 10 wt% PCL solution in dichloroethane [43]. Aligned porous PCL/zein composites were prepared from chloroform‐ethanol or chloroform‐acetic acid glacial solution [52]. Zein is a prolamin‐rich and alcohol‐soluble protein, available in corn at 2.5–10% dry basis.

For the preparation of porous PLGA, 1,4‐dioxane seems to be the most commonly used solvent, probably because of its high melting point (∼10 °C) for easy freeze‐drying. Porous poly(L‐lactic acid) (PLA) monoliths were prepared from 1,4‐dioxane solution [53]. The presence of small amounts of water in the solution could lead to the formation of small pits on the PLA structure. Creating pores in the wall of porous PLA could enhance mass transport for various applications. For this purpose, poly(ethylene glycol) (PEG) of different molecular weights (600, 2000, 4000, 6000 Da) and different ratios to PLA (50/50, 30/70, 10/90) were mixed with PLA. After freeze‐drying, PEG was leached out using ethanol. This approach allowed the formation of pores in the PLA wall. The pore sizes could be tuned by varying PEG molecular weight and the PEG/PLA ratio [54]. In order to enhance the interconnectivity of the aligned channels in porous PLGA scaffolds, an improved thermal‐induced phase separation by adding the second solvent (chloroform, benzene, ethanol or water) into the dioxane solution was employed [55]. The mould containing the solution was first cooled from bottom to top to −20 °C and then further treated in liquid nitrogen. Water is a benign solvent and seemed to be most effective in creating pores in the wall (with water content 3–5 w/w%). This was attributed to the crystallization of the second solvent after the initial crystal growth of dioxane [55].

Blaker et al. combined directional freezing and ice microspheres to prepare highly porous PLA scaffolds with the pore surface lined by bacterial cellulose nanowhiskers (BCNW, to provide hydrophilicity) [56]. The ice microspheres were formed by freezing a water mist generated from an ultrasonic fogger in liquid nitrogen. The frozen spheres were sieved and those in the diameter range of 100–500 μm were used as sacrificial templates. Ice spheres and BCNW were suspended in the PLA–chloroform solution at −25 °C. Owing to the hydrophilic interaction, BCNW migrated from the non‐polar solution to the polar surface of ice microspheres and formed a network coating the spheres. After freeze‐drying to remove both frozen chloroform and ice microspheres, the porous scaffold with BCNW lining the pore surface could be formed [56].

Polyacrylonitrile (PAN) is a semi‐conducting polymer and has been widely used to produce fibres and carbon materials. PAN is poorly soluble in water but can be readily dissolved in polar solvents such as dimethyl sulfoxide (DMSO) and dimethylformamide (DMF). Owing to its higher melting point, PAN–DMSO solution was directly frozen and freeze‐dried to produce aligned porous PAN [57]. In an effort to avoid using the freeze‐drying procedure, it was possible to remove the frozen DMSO by solvent exchange with water. The temperature was maintained at 0 °C using the ice/water mixture. Aligned porous PAN could be successfully formed by this approach [58].

A home‐made mould consisting of two plates (one steel plate and one glass plate) was developed to generate 2D temperature gradients for the preparation of aligned porous PLGA and a range of other polymers [59]. A degassed solution (formed with solvents of high melting points) was sealed between the two plates. The mould was then vertically placed into a water reservoir with different temperatures. The vertical temperature gradient was formed due to the relatively low reservoir temperature while the horizontal temperature gradient resulted from the different thermal conductivities of the steel and glass plates. Similarly to the wedge‐based system [51], the 2D temperature gradients facilitated the generation of polymer membranes with vertically aligned pores [59].

4.2.2.3 Compressed CO2 Solution

Supercritical fluids, with the temperature above the critical temperature and the pressure above the critical pressure, exhibit liquid‐like density and gas‐like diffusivity, and the properties can be tuned readily by varying the pressure and temperature, particularly near the critical region. Owing to its mild critical parameters, non‐toxicity, and non‐flammability, supercritical CO2 is the mostly investigated and used supercritical fluid. Supercritical CO2 foaming is an effective route to preparing highly interconnected porous scaffolds [60]. Compressed CO2, where the temperature is lower than the critical temperature (∼31 °C) but the pressure is higher than the critical pressure (∼72 bar), exhibits higher density and strong solvation power. It may be used as a reaction medium or solvating phase such as in the foaming process [6062]. In the foaming process, the polymers are softened and expanded by high pressure CO2. Releasing CO2 by depressurization leaves interconnected but disordered pores behind.

It is possible to freeze a CO2 solution or suspension and produce an aligned porous material. However, most of the polymers have very low solubility in CO2. Its low viscosity and low density make it difficult to be used as a medium for suspensions as well. Sugar acetates, including β‐1,2,3,4,6‐pentaacetyl‐D‐galactose (BGAL), α‐1,2,3,4,6‐pentaacetyl‐D‐glucose (AGLU), and β‐1,2,3,4,6‐pentaacetyl‐D‐glucose (BGLU), can be dissolved in compressed CO2 in sufficient solubility. For example, the BGAL solution in liquid CO2 (12 wt%) at 72 bar/21 °C in a stainless steel column was directionally frozen in liquid nitrogen [63]. Unlike aqueous or organic solutions, a freeze‐drying step was not required. Solid CO2 was readily sublimed when the stainless steel column warmed up at room temperature with the valve open. A dry monolith with aligned porous structure was successfully produced (Figure 4.4). Other CO2‐soluble or suspendable additives may be added into the solution to make porous composite materials, as demonstrated by a hydrophobic dye Oil Red O [63].

Figure 4.4 Aligned porous BGAL prepared by directional freezing of its solution in compressed CO2. The red arrow indicates the freezing direction.

Source: Zhang et al. 2005 [63]. Reprinted with permission from American Chemical Society.

Directional freezing of CO2 solution to fabricate aligned porous materials is a highly attractive method, because no organic solvent is required, there is no issue of solvent residual, and the energy‐intensive freeze‐drying step can be avoided. Although it is limited by the low solubility of most polymers in compressed CO2, with sugar acetate dissolved in CO2, it may pave the way for directional freezing of a variety of CO2 suspensions.

4.2.3 Nanofibrous Polymers

Porous polymers are usually formed when using ice templating as the preparation method. It has been noticed that porous polymers with nanofibrous network may be formed when the concentration is reduced [64]. Qian et al. investigated the freezing of highly diluted aqueous polymer solutions, with concentrations in the range of 0.02–0.1 wt% depending on the type and molecular weight of the polymers. Nanofibres can be successfully prepared from PVA, sodium carboxymethyl cellulose (SCMC), dextran, and sodium alginate (Figure 4.5) [65]. The diameters of the nanofibres fall in the range of 100–600 nm. In spite of the diluted solution, it is relatively easy to prepare a large quantity of nanofibres by this approach because the bulky solution can always be frozen in a vessel with a large surface to facilitate freeze‐drying. This is favourable compared to the electrospinning method that is the common method to prepare polymer nanofibres. The resulting hydrophilic polymer nanofibres may be further used as templates to produce hollow titania microtubes or iron oxide nanofibres [65].

Figure 4.5 An example of polymer nanofibers (SCMC, Mw = 250 K) prepared by freeze‐drying the dilute aqueous solution (0.02 wt%). Scale bar 5 μm, inset scale bar, 600 nm.

Source: Qian et al. 2009 [65]. Reprinted with permission from Royal Society of Chemistry.

This approach was also used to prepare chitosan nanofibres that were preferred to porous chitosan for the controlled release of proteins (bovine serum albumin (BSA)) and small molecules (Rhodamine B as a model). The release profile could be tuned by varying the morphologies of porous chitosan [32]. In another development, lignin nanofibres were produced continuously by a custom‐built set‐up (where aqueous solution could spread onto a rotating metal drum whose temperature was controlled by liquid nitrogen; the frozen fibres were removed by a blade and freeze‐dried) and then carbonized to make carbon nanofibres as electrode materials [66]. However, so far, this method has not been found to be very effective in the preparation of hydrophobic nanofibres by directional freezing of organic solutions.

Porous polymers with nanofibre networks may be also formed directly by freeze‐drying some hydrogels. Hydrogels are highly interconnected 3D porous networks that contain a high percentage of water. Hydrogels with entangled nanofibres may be formed. However, the control on the size of the nanofibres is usually limited. By pH‐initiated self‐assembly of perylene‐based molecules, driven by hydrophobic interaction and π–π stacking, well‐defined nanofibrous network can be formed, with diameters around 25 nm [67, 68]. After a proper washing procedure with water/acetone/cyclohexane and freeze‐drying, dry monoliths with nanofibrous networks could be produced [67, 68].

4.2.4 Combining Ice Templating and Other Templating Methods

In addition to its templating impact, freezing is an effective way to lock in the fluid structure. This has been used to combine ice templating and emulsion templating for the preparation of porous polymers and composite materials. For emulsion templating, the monomers in the continuous phase are polymerized and crosslinked to lock in the emulsion structure and the removal of solvents from the droplet phase and the continuous phase produces emulsion‐templated structures [14]. By freezing an emulsion (even with limited stability) both phases in the emulsion are frozen to lock in the structure. Similarly, removing the solvents from both internal phase and continuous phase by freeze‐drying generates a dry and highly porous structure [69]. Because the volume percentage of droplets in an emulsion can be varied in a wide range, this can be translated into systemic control of porosity and pore morphologies in the produced materials [25]. Figure 4.6 shows the evolution of pore structures in ice‐templated and emulsion‐templated SCMC (Mw = 90 K). Cyclohexane was emulsified into aqueous SCMC solution with sodium dodecyl sulphate (SDS) as the surfactant. The volume ratios of cyclohexane to the total volume of the emulsion were varied at 0, 20, 40, 60, and 75 v/v%. It can be clearly seen from Figure 4.6 that the ice‐templated structure is obtained from the emulsion with 0% cyclohexane (basically aqueous solution). Then, the number of cellular emulsion‐templated pores increases with the increasing ratio of the internal droplet phase, until a highly interconnected porous structure is formed with the 75 v/v% emulsion. Accordingly, the bulk densities of the SCMC materials decrease from 0.11 to 0.031 g cm−3 and the pore volumes (as measured by Hg intrusion porosimetry) increases from 7.75 to 29.67 cm3 g−1 [25].

Figure 4.6 The evolution of pore structures by emulsion templating and ice templating based on porous sodium carboxymethyl cellulose. The emulsions are prepared with different volume percentages of internal phase. (a) 0%, (b) 20% v/v. The white circle indicates one of the emulsion‐templated pores. (c) 40%, (d) 60%, (e) 75% v/v.

Source: Qian et al. 2009 [25]. Reprinted with permission from Royal Society of Chemistry.

An emulsion is usually formed from an organic phase and an aqueous phase, e.g. the oil‐in‐water (O/W) emulsion. The difference in the densities of organic solvent and aqueous phase can be explored by centrifugation. For example, when an O/W emulsion is formed with cyclohexane, the lighter oil droplets can move upward while those in the heavier aqueous phase mainly remain at the lower part when the emulsion is centrifuged. Thus, a distribution of oil droplets is formed with more droplets at the top and few droplets at the bottom. This emulsion can then be rapidly frozen and freeze‐dried to produce an emulsion‐templated gradient porous material [70]. This was demonstrated with the polymer (PVA and polyacrylamide) and polymer/silica composites [70]. For the O/W emulsion, hydrophobic compounds may be dissolved in the organic droplet phase. Freeze‐drying of such emulsions can produce porous polymers with the in situ formation of organic nanoparticles [69]. This method has been investigated for the production of aqueous nanodispersion of poorly water‐soluble drug nanoparticles [71].

In addition to liquid emulsions, a gas‐in‐liquid foam may be also freeze‐dried to generate porous scaffolds. This has been demonstrated with polysaccharide‐based scaffolds (chitosan, alginate, hyaluronic acid tetrabutylammonium) [72]. The foam was generated inside a glass reactor with the introduction of an inert gas (argon). The foam was frozen immediately in liquid nitrogen and then freeze‐dried. The diameters of the voids were around 200–300 μm [72]. In another study, ice microspheres were used as additional templates to fabricate highly porous PLA scaffold [56]. Both approaches produced porous materials with large macropores, which can be highly useful for tissue engineering.

4.3 Hydrogels and Crosslinked Porous Polymers

The porous polymers produced by freeze‐drying solutions, suspensions, and emulsions usually exhibit weak mechanical properties and significant shrinkage or dissolution in the presence of the solvent. This can considerably limit their uses in some applications. This problem may be addressed by crosslinking the polymer before, during or after freeze‐drying. The crosslinking processes may be performed either physically or chemically. Hydrogels are crosslinked polymers containing a large amount of water, which may be freeze‐dried to produce crosslinked porous polymer. A freezing process is always involved for the hydrogels discussed in this chapter.

4.3.1 Hydrogels By a Freeze–thaw Process

This process is widely known for the preparation of physically crosslinked PVA hydrogels. PVA hydrogels are widely used in pharmaceutical and biomedical applications. Compared to chemical crosslinking, the freeze–thaw process avoids the use of toxic solvents and reagents. During the freeze–thaw process, aqueous PVA solution is frozen for a certain period and then allowed to thaw to form aqueous solution again. Multiple freeze–thaw cycles are usually employed to produce stable PVA hydrogels [73]. The preparation of PVA hydrogels is attributed to the formation of small crystallites, which connect the amorphous PVA chains, as illustrated in Figure 4.7 [74]. The PVA crystallites may be described as a layered structure held together by hydrogen bonding and van der Waals forces. Folded PVA chains can lead to the formation of crystallites (small ordered region) distributed in an amorphous matrix [7375]. It is believed that the concentration of PVA during freezing facilitates the crystallization while the crystallization may be enhanced or weakened in the thawed state. The increased motion of PVA chains or sections may initially facilitate the crystallization but may subsequently break down the crystallites. It was observed that the size of the crystallites increased first and subsequently decreased [73].

Figure 4.7 Schematic representation of a PVA gel structure during freeze–thaw cycling/aging/imaging. Starting with a fresh solution (a), primary crystallites appear after the first freeze–thaw cycle (b), and these may be enhanced by secondary crystallites after cycling/aging (d). Removal of the frozen water for microscopic imaging collapses the amorphous chains, creating either a network of rounded pores (c), or a network of fibrils (e).

Source: Willcox et al. 1999 [74]. Reprinted with permission from John Wiley & Sons.

The factors that influence the freeze–thaw PVA hydrogels include the properties of PVA (molecular weight, degree of hydrolysis, and concentration) and processing parameters (the number of freeze–thaw cycles, freeze temperature/time, thawing temperature/time). The properties of PVA hydrogels, e.g. mechanical stability, elasticity, and strength under external stresses, are considerably influenced by the size, number, and distribution of PVA crystallites [73, 75]. The degree of crystallinity is in the order of 2–6% for freshly prepared gels [75]. The average size of the crystallites is in the region of 7 nm and the distance between the crystallites is approximately in the range of 15–20 nm [73].

Owing to the importance of the crystallites to PVA hydrogels, the crystallinity of PVA hydrogels has been investigated by various techniques [22, 7376]. These techniques include density measurement [73], infrared spectroscopy (the intensity of the peak at 1141 cm−1 relating to the degree of crystallinity) [73]; cryogenic transmission electronic microscopy (cryoTEM) [74]; differential scanning calorimetry (DSC, the crystallinity degree is expressed as the ratio of the heat of crystallization of PVA hydrogel and 100% crystalline PVA, ΔHHc, with ΔHc = 150 J g−1) [7376]; wide angel X‐ray diffraction (the crystallinity degree is described as the ratio of the peak area of crystalline aggregates at 18–21° to the sum area of the peaks of crystalline aggregates and amorphous swollen PVA) [22, 75]; and solid state 1H NMR (based on the fraction of rigid protons in PVA hydrogel, measured for the fractions of protons that relax during the first 20 μs) [75]. Based on the comparison study of freeze–thaw gels and aged gels (no freezing involved, gelation at room temperature), even with similar degree of crystallinity, the compressive modulus of the freeze–thaw gel was found to be higher. This led to the suggestion that phase separation during freeze–thaw cycles contributed to the gels with higher modulus [22]. This may be also linked to the effect of PVA concentration during freezing on the formation of crystallites (e.g. size, distribution).

PVA is produced by hydrolysis of poly(vinyl acetate) (PVAc) but 100% hydrolysis is very difficult to achieve. Therefore, the PVA is always a copolymer with PVAc. PVA has a simple linear polymeric chain with pendant hydroxyl groups. This can facilitate the formation of hydrogen bonding between PVA chains, which limits the solubility in water for PVA with high degree of hydrolysis. For PVA with lower degree of hydrolysis, residual hydrophobic groups can weaken or disrupt the interaction between PVA chains. This leads to higher solubility in water for PVA with low degree of hydrolysis [73]. In the studies of freeze–thaw hydrogels, PVAs with the degree of hydrolysis >98% and more often >99% are used. Usually, there is a minimum of PVA molecular weight required for the formation of hydrogels. Different molecular weights have been utilized, including, 115 K, 89–98 K, and 64 K, 36 K [22, 75, 76]. For example, PVA solutions with concentrations of 7, 10, and 15 wt% were frozen at −20 ° for 8 h and thawed at +25 °C for 4 h for different number of cycles [76]. Some general trends were observed. Increasing the number of freeze–thaw cycles can reinforce the PVA hydrogels. The higher PVA concentration leads to initially higher crystallinity and stability upon swelling. An increase in PVA molecular weight results in crystallites of higher lamellar thickness and broader crystallite size distribution [76].

Water‐miscible organic solvents have been added to aqueous PVA solution in an effort to generate hydrogels with improved mechanical stability or transparency. Among them, DMSO has been mostly investigated [73]. Preparation of PVA blend/composite hydrogels has also been investigated because they can provide additional useful properties. Some of the examples include pH‐sensitive PVA‐PAA (poly(acrylic acid)) hydrogel [77], PEG‐modification of PVA hydrogels to reduce protein adsorption [78], PVA‐chitosan hydrogel [79], PVA‐alginate hydrogel further crosslinked by Ca2+ [80], PVA‐Salecan (a water‐soluble extracellular glucan) hydrogel [81], and reinforced PVA hydrogels by cellulose nanocrystals [82]. Recent developments in the freeze–thaw hydrogels include self‐healing PVA hydrogel (attributed to hydrogen bonding between PVA chains at the cutting interface) [83], strong hemi‐cellulose‐based hydrogel [84], and different types of polysaccharide hydrogels [85].

4.3.2 Macroporous Cryogels

The macroporous cryogels are produced by polymerization or crosslinking reactions at sub‐zero temperatures, hence the name ‘cryogels’. The cryogels can be formed from monomers or polymers, in water (the focus of this section) or organic solvent, exhibiting a highly interconnected porous structure [8688]. For the cryogels to form, the solutions are usually cooled down to between −5 and −20 °C. The frozen (or sometimes called ‘semi‐frozen’) samples are stored in a cold environment to allow the polymerization to complete, sometimes facilitated by UV irradiation. It is understood that the polymerization can be very slow at low temperatures. However, the freezing process concentrates the monomers and initiators. The resulting higher concentrations can enhance the polymerization, which may balance out the negative effect of low temperature on the reaction rate. The accompanying depression effect on the melting point due to concentration leads to the non‐frozen monomer‐rich phase segregated within the frozen matrix, which is why such systems may be termed as ‘semi‐frozen’ and is the basis for the polymerization at sub‐zero temperatures.

The most often used system for cryogels is the redox‐initiated free radical polymerization with potassium persulfate (KPS)/ammonium persulfate (APS) and tetramethyl ethylenediamine (TMEDA). They are usually used for hydrophilic monomers such as acrylamide (AM), N‐isopropylacrylamide (NIPAM) or 2‐hydroxyethylmethacrylate (HEMA) [89]. For example, PAM cryogel with N, N′‐methylene‐bis(acrylamide) (MBAM) as a crosslinker was formed after freezing the solutions with APS/TMEDA and kept at –12 °C for 16 h [90]. PAM cryogel beads were prepared by emulsifying AM/MBAM/APS/TMEDA into n‐hexane using Span‐80 as the surfactant at 0–5 °C. The emulsions were then cooled to different temperatures (−12, −15, −18 °C) and kept at the respective low temperatures for 24 h before stopping the reaction and removing n‐hexane and un‐reacted reagents [91]. Similarly, PNIPAM cryogel was prepared by freezing NIPAM/MBAM/APS/TMEDA and then stored at −12 °C for 12 h [92].

In addition to the concentrations of monomers, the crosslinkers and the reaction temperatures are important parameters that can affect the physical properties of the cryogels. Specific to cryogels, the freezing temperature and probably more importantly the cooling rate can influence the size of ice crystals and hence the pore size of the cryogel. The fast freezing rate leads to a higher number of ice nucleation and smaller size of ice crystals [1921, 86]. To form a cryogel, the rate of polymerization must be slower than the rate of freezing. Otherwise, the polymer network may be formed before the formation of ice crystals is completed. Owing to the low temperature involved in preparing cryogels, a longer cryopolymerization time is usually required [86, 88]. However, a UV irradiated polymerization may be adopted to facilitate the polymerization. Such a reaction system contains a photoinitiator and the polymerization is initiated under UV irradiation. The UV frozen polymerization is usually applied to acrylate monomers or crosslinkers [86]. For example, PAM, PNIPAM, and polyHEMA cryogels were synthesized using poly(ethylene glycol) diacrylate (PEGDA, Mw ∼ 575) and H2O2 as the catalyst. The solutions in a mould (4 mm thick) were frozen in a freezer (−20 °C) and then polymerized in a thermostat using a UV curing instrument (power 400 W) for up to 10 min [93]. A similar approach was employed for the in situ entrapment of urease in PNIPAM cryogel [94]. However, for acrylate monomers (rather than PEGDA as crosslinker in ref [186), their poor solubility in water signifies that a polar organic solvent is better for the reaction system. 1,4‐dixoane has been often used to produce polyacrylate cryogels [9597].

In addition to the use of monomer solutions, cryogels may be also prepared by crosslinking polymer molecules under the freezing conditions. This has been widely known for the preparation of PVA and chitosan cryogels using glutaraldehyde as the crosslinker. However, glutaraldehyde is cytotoxic and it is not easy to completely remove it from the cryogel by washing. Since cryogels have been used widely in biomedical applications and the separation of biomolecules [86, 98], biocompatible crosslinkers such as genipin and oxidized dextran or the crosslinking mechanism via 1‐ethyl‐3‐3‐dimethylaminopropylcarbodiimide hydrochloride (EDC) may be employed [86]. The recently developed cryogels provide biocompatibility and functionality for bio‐separation or tissue engineering applications, for example, silk fibroin cryogels (fibroin with TMEDA as catalyst, ethylene glycol diglycidyl ether as crosslinker at −18 °C for one day) [99], transparent hyaluronate cryogel (disaccharides as additives, 2‐(N‐morpholino)‐ethanesulfonic acid as crosslinker, storage at −20 °C overnight) [100], and starPEG‐Heparin cryogels (via the EDC/N‐hydroxysulfosuccinimid (sulfo‐NHS) crosslinking chemistry at −20 °C overnight) [101].

Porosity and pore morphology are important properties of porous materials. For hydrogels and particularly supermacroporous cryogels described here, the difference in porosity and pore morphology should be noticed for the hydrogels and their dry structures. It is quite convenient to observe pore morphology by scanning electronic microscopy (SEM) and measure the porosity by Hg intrusion porosimetry. However, both techniques require the use of dry materials. The change in pore structure and porosity is unavoidable when drying the hydrogels, even by freeze‐drying or the time‐consuming supercritical fluid drying. These changes can be significant. As hydrogels are used in a hydrated state, the characterization of hydrogels is highly important. Cryogels are macroporous materials with pore diameters in the range of 1–300 µm and pore wall thickness of a few micrometers [89]. The pore sizes in this range can be observed by optical microscopy [22]. In order to observe the contrast between polymer network and the contained aqueous medium clearly, confocal laser scanning microscopy (CLSM) is often used. The auto‐reflection mode may be used to image the hydrogels, based on the difference in reflection from the polymer network and the aqueous medium [102]. Fluorescent dyes may be trapped in the hydrogels, either in the aqueous medium (e.g. Nile Red in water) [102] or by staining the polymer networks with different dyes (e.g. fluorescein isothiocyanate (FITC) or Rhodamine B) [103]. Other techniques include X‐ray micro‐computed microscopy (μCT), environmental SEM, and cryo‐SEM [86]. The porosity may be estimated from the 3D pore structures built from these characterization methods. It may be also determined by squeezing the excess water from a hydrated hydrogel and then calculating the percentage of the squeezed water to the swollen gel [86]. This approach, however, may be only suitable for rather stable hydrogels.

The state of water in a hydrated hydrogel is important because it can affect the water removal during a freeze‐drying process or the secondary/tertiary structure of adsorbed proteins. Gun'ko et al. have divided the water into five states: free unbound, weakly bound, strongly bound, weakly associated, and strongly associated. The states of water may be distinguished by various techniques such as 1H NMR, DSC and thermogravimetric analysis (TGA) [89].

4.3.3 Aligned Porous Materials By Frozen Polymerization

The chemical process discussed in this section is quite similar as that described in Section 4.3.2, i.e. a crosslinking or polymerization or both occur in the frozen state, providing a crosslinked porous structure. However, the focus is different. In Section 4.3.2, the focus is on the cryogels using water as solvent and the highly interconnected cellular porous structures. This section emphasizes the materials with aligned porous structures, using either water or organic solvents. After the frozen polymerization, the frozen samples are allowed to thaw and dry under vacuum (instead of freeze‐drying) for their intended applications.

Frozen polymerization has been used to prepare crosslinked aligned porous structures with good mechanical stability from monomer solutions instead of the commonly used polymer solutions [95, 96]. Because the frozen solutions cannot be thermally polymerized and the polymerization rate by free radical polymerization may be too slow, initiation and polymerization by UV irradiation would be a better choice. A photoinitiator and suitable monomers should be used for frozen polymerization. Most of photopolymerizations involve the use of acrylate monomers. For example, tetraethyleneglycol dimetharylate (TEGDMA) was dissolved in 1,4‐dioxane, cyclohexane, or camphene (by heating to 60 °C) with 1 wt% 2,2‐dimethoxy‐2‐phenylacetophenone (DMAP) as a UV initiator. After directionally freezing in liquid nitrogen, the frozen samples were placed on ice or dry ice under a UV lamp for 2 h with the samples being turned every 30 min to allow complete polymerization. The warmed samples were then dried in a vacuum oven (no need for freeze‐drying) overnight at room temperature [95]. Figure 4.8 shows the porous structure of polyTEGDMA from the solutions with 1 : 5 (v/v) TEGDMA to the relative solvent (the same mass ratio for camphene because it is solid at room temperature and it is easier to measure by mass). The Young modulus of these aligned porous materials was up to 9 MPa while the freeze‐dried porous polymers are normally in the range of 0.003–0.0175 MPa [95]. The same methodology has been used to prepare crosslinked polyHEMA, pH‐sensitive poly[2‐(dimethylamino) ethyl methacrylate] (PDMAEMA) [96], poly(butyl methacrylate‐co‐ethylene glycol dimethacrylate) (dioxane as solvent) [97], poly(urethane diacrylate) (dehydrated 1,4‐dioxane as solvent, diphenyl(2,4,6‐methylbenzoyl)phosphine oxide (TPO) as initiator) [104], and polyHEMA (using t‐butyl alcohol as solvent by γ irradiation) [64]. Okaji et al. utilized a binary solvent system (1,4‐dioxane and t‐butanol) to prepare a unique multi‐hollow‐core honeycomb structure [105]. Diurethane dimethacrylate was the monomer whilst Irgacure 184 was selected as the photoinitiator. For a 1.2 mm thick frozen sample, the UV irradiation was performed for 3 min using a mercury xenon lamp at 365 nm at light intensity of 35 mW cm−2. It was proposed that 1,4‐dioxane was firstly frozen to form orientated crystals, excluding both the monomer and t‐butanol between them. The crystallization of t‐butanol was initiated within the monomer‐rich phase to form aligned needle crystals. This resulted in the formation of aligned porous materials with needle‐like porosity in the wall [105].

Figure 4.8 Aligned porous polyTEGDMA prepared by UV‐irradiated frozen polymerization from camphene solution (a,b), cyclohexane solution (c), and dioxane solution (d).

Source: Barrow et al. 2012 [95]. Reprinted with permission from Royal Society of Chemistry.

Similarly to the preparation of cryogels, cryopolymerization to form aligned porous hydrogels can be achieved by employing a directional freezing process instead of freezing in a freezer. For example, aligned porous polyHEMA hydrogels were prepared by directional freezing of aqueous HEMA/KPS/TMEDA solution and then allowed to polymerize at −15 °C for 48 h [106]. PEG hydrogels were generated from aqueous solutions of PEGDA (Mw = 200, 700, 2000) containing APS and TMEDA by cryopolymerization at −15 °C for 12 h [107] and were further used to generate PEG/PNIPAM dual‐network with temperature‐sensitive volume change [103]. A similar procedure was used to prepare aligned porous poly(HEMA‐co‐AM) hydrogels [108].

Other types of chemical crosslinkers may be also used to prepare cryogels. For example, glutaraldehyde was used to form aligned porous agarose‐gelatin cryogels. The solutions were contained in 5 ml plastic syringe and placed in a cryostat at −12 °C for 16 h [109]. In another study, dextran was oxidized using sodium periodate (1–2 h at room temperature). The oxidized dextran was then dissolved in water and mixed with chitosan solution (in acetic acid water) at 0 °C. After freezing the mix solution completely at different temperatures (liquid nitrogen −196 °C, freezing ethanol −110 °C, freezing acetic ether −80 °C), the frozen samples were left in a freezer pre‐set at −8 °C for 3 days to allow the cryopolymerization to complete [110]. A crosslinker, sulfur monochloride (S2Cl2), was added to butyl rubber‐cyclohexane solution. The resulting solutions were frozen in a freezer or cryostat at different temperatures (−2, −7, −10, −18 °C) and maintained at these temperatures for 1–3 days to complete the cryopolymerization. This method produced tough organogels [111].

4.3.4 Post‐freeze‐drying Crosslinking

Instead of frozen polymerization, it is possible to crosslink the freeze‐dried materials to improve their mechanical stability and their stability in liquid medium. Thermal treatment may be employed directly to crosslink the freeze‐dried scaffolds. Biopolymeric scaffolds such as collagen‐glycosaminoglycan (dehydrothermal crosslinking at 140 °C under vacuum for 24 h) [29] and wheat gluten hybrids (thermal treatment at 130 °C) have been processed in this way. However, it may be necessary to include a crosslinker or pre‐polymer in the formulation to allow thermal crosslinking. For example, 3‐glycidoxypropyltrimethoxysilane (GOPS) was added to an aqueous dispersion of PEDOT:PSS. After freeze‐drying, the scaffolds were heated to 140 °C for 1 h to complete the crosslinking reaction [30]. In another study, poly(glycerol sebacate) (PGS) pre‐polymer and PLA were dissolved in 1,4‐dioxane. The freeze‐dried scaffolds were placed in a vacuum oven for 24 h at 150 °C for the curing of PGS pre‐polymer [112].

Alternatively, the freeze‐dried scaffolds can be crosslinked in a liquid medium. For this approach, the selection of crosslinker, solvent, and concentration is critical for successful crosslinking without damaging the pore structure. A non‐solvent to the scaffold may be used. It is also possible to use a very high concentration of a crosslinker in the solution. The crosslinking reaction can be completed before the dissolution occurs or the damage is done. For example, highly water‐soluble porous PVA beads were crosslinked in toluene using toluene 2,4‐diisocyanate as crosslinker and triethylamine as catalysts. The reaction was refluxed for 3 days. The longer reaction time may be attributed to poor solvation of PVA by toluene [25]. For biocompatible scaffold, the crosslinking via EDC in a polar solvent has been used very often. For example, egg white scaffolds were crosslinked with EDC and NHS in 4‐morpholineethanesulfonic acid (MES) buffer in 70% ethanol at 4 °C overnight [27]. Similarly, hyaluronic acid (HA) and alginate scaffold can be crosslinked by EDC chemistry, in a mixing solvent of acetonitrile/ethyl acetate (3 : 2) and acetone/water (9 : 1), respectively. The chitosan scaffold can be crosslinked by genipin (1 w/v%) in ethanol/water (8 : 2) [72].

There are some other options as well. Lactose‐silk fibroin sponges were soaked in a series of aqueous methanol solution to induce crystallization and water insolubility. The size of the sponge was very small during this treatment [26]. The SCMC nanofibres prepared by freeze‐drying were immersed in saturated aqueous FeCl3 solution. The crosslinking of carboxylic acid groups by Fe3+ and the low solubility of SCMC in the saturated solution allowed the intact SCMC nanofibres to be crosslinked in aqueous solution. The calcination of Fe3+‐crosslinked SCMC nanofibres produced Fe2O3 nanofibres, confirming that Fe3+ crosslinking did happen [65].

4.4 Applications

4.4.1 Biocompatibility and Tissue Engineering

The characteristics of highly interconnected sponges and hydrogels mimic the feature of an extracellular matrix (CEM). It is not surprising that these materials have been widely evaluated and used to support cell growth for tissue engineering. The ability to use different materials, to tune pore morphology and porosity as well as to modify the surface chemistry by different preparation approaches, provides a solid platform to generate suitable scaffolds for different applications.

When a scaffold is in contact with cells and tissues, it is important to make sure there is no unacceptable degree of harm to the body and that the scaffold can perform its intended function [113]. Therefore, it is a rather routine assessment of biocompatibility of porous materials or scaffolds if they are to be used for biomedical applications. Biocompatibility, at its basic level, indicates non‐cytotoxicity towards the targeted cells or tissues. It is now widely accepted that biocompatibility involves two principle levels [114]. The first one is ‘biosafety’, covering cytotoxicity and the complex mutagenesis and carcinogenesis. The second level is about ‘biofunctionality’, where cell adhesion, cell spreading, and cell biosynthetic function are assessed [114]. It should be pointed out that there are different biocompatibility assessment criteria for different parts of the tissues or long‐term medical devices [113]. The cytotoxicity assessment is focused on cell damage, cell growth, and cellular metabolism, using methods such as 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyl tetrazolium bromide (MTT) assay, DNA synthesis, and membrane integrity tests [114].

For a scaffold, in addition to porosity and pore morphology, other properties such as mechanical stability, transparency, and degradability are important parameters [40, 41, 100]. For example, PNIPAM/silk hybrid hydrogels are evaluated by mechanical stability tests and enzymatic degradation by protease XIV [40].

Many studies are carried out to assess the biocompatibility of 3D porous scaffolds or hydrogels. The agarose‐gelatin cryogels were evaluated by MTT test and by culturing with fibroblast (Cos‐70). Good cell adhesion and cell proliferation were observed [109]. Human foreskin fibroblast cells were found to easily migrate within silk fibroin scaffolds and reach scaffold periphery within 28 days of culture. Cell attachment and alignment of actin filaments within the scaffold were also confirmed by CLSM [39]. The collagen‐glycosaminoglycan scaffold was evaluated with human dermal fibroblasts and epidermal keratinocytes. For the potential to be used as engineered skin substitutes, the resulting structures were assessed for surface hydration and mitochondrial metabolism [29].

Laminin‐containing gelatin cryogels (glutaraldehyde as crosslinker) and dextran cryogels (with dextran macromer) were generated by cryopolymerization and then assessed for neural tissue regeneration [115]. The optimal pore size range was 80–100 μm. The scaffold seeded with the stem cells from human umbilical cord blood could generate a network of neurons and glia after differentiation. It was possible to transplant the scaffold into the brain tissue as demonstrated by rat models [115]. The transparent HA cryogels were formed by additives of mono/disaccharides and were still transparent after removing the additives. The transparence facilitates the observation of 3D growth of T47D breast cancer cells with immobilized growth factors by confocal microscopy [100]. Star‐shaped PEG was crosslinked with heparin via the EDC/sulfo‐NHS chemistry by cryopolymerization. The gels contained a porosity of up to 93% and interconnected large pores with the size in the range of 30–180 μm. As demonstrated with human umbilical vein endothelial cells, the scaffolds showed cell attachment and even spreading of the cells within the porous scaffold [101].

The scaffolds can also be fabricated from biowastes or easily available biosources. For example, egg white macroporous scaffolds were prepared and found to support active metabolism, proliferation, and migration of human dermal fibroblasts. The in vivo evaluation by subcutaneous implantation in mice showed negligible immune reaction and efficient cell and tissue ingrowth [27]. The collagen sponges made from bovine hide trimming pieces were found to be viable towards 293 T cells [41]. For the scaffolds made from semi‐conducting polymer PEDOT:PSS, in addition to supporting the growth of mouse fibroblasts (3T3‐L1), protein conformation, cell adhesion, and pro‐angiogenic capability could be tuned electrically by the volt change of +1 to −1 V [30].

Hydrogels with aligned pores can also be used as support for cell growth, just like the sponges or highly interconnected macroporous gels. More importantly, the aligned pores can act as guides to direct cell growth or differentiation. For example, porous PLGA scaffolds with aligned microtubule structures could guide rabbit aortic smooth muscle cells to grow better along the microtubule direction [55]. Porous chitosan was prepared with a highly regular and aligned lamellar architecture, with uniform ridges of controlled height and width on the lamellar surface. The ridges in the 3D aligned porous chitosan could direct DRG neurite growth, as shown in Figure 4.9a [48]. A lamella with the parallel ridges could be isolated from the 3D chitosan, which acted as a 2D substrate to clearly direct the DRG neuron growth (Figure 4.9b) [48].

Figure 4.9 Cell growth in the presence of aligned structures. (a) Confocal microscopy image of the aligned growth of DRG neurite in a 3D laminin‐coated porous chitosan; (b) Fluorescent image of the DRG neurite seeded on an isolated lamella from the 3D chitosan. ; (c) SEM image showing the growth of mMSCs parallel to ridges of the aligned PLGA‐chitosan substrate. ; (d) SEM image showing aligned porous PCL microspheres; (e) SEM image showing aligned porous PS microspheres; and phase contrast photomicrographs showing the growth of mESCs after 72 h culture in the presence of (f) PS microspheres; and (g) PCL microspheres.

Source: Zhang et al. 2008 [117]. Reprinted from Elsevier.

Source: Reprinted from Ref. [116]

Source: Reprinted from Ref. [48]

It is possible to directly fabricate the 2D aligned patterns by directional freezing of thin liquid phase spreading on a glass slide via a motor‐controlled freezing stage and followed by freeze‐drying [116]. All the aligned patterns of PLGA (prepared from PLGA‐dioxane solution), chitosan containing PLGA nanospheres and silica (prepared from silica suspension) could be used to direct the growth of mouse mesenchymal stem cells (mMSCs). The fluorescent microscopic imaging confirmed the cell growth along the aligned ridges. The SEM image in Figure 4.9c shows that the mMSCs are aligned, parallel to the ridges. The spindle‐shaped cells are seen on the ridges or with the extended filopodia contacting the ridges for the PLGA nanosphere‐containing chitosan patterns [116].

Aligned porous PCL microspheres and polystyrene (PS) microspheres can be prepared by directional freezing of O/W emulsions where PCL or PS is dissolved in the oil droplet phase [117]. The freeze‐dried composites can be readily dissolved in water and the aligned porous microspheres are collected by centrifugation, as shown in Figure 4.9d and e. The microsphere suspension was examined by culturing with mouse embryonic stem cells (mESCs). It was found that the mESCs could grow after 72 h with the microspheres suspended in the culture medium. PS microspheres were not binding to the cell clusters (Figure 4.9f) while PCL microspheres were shown to bind to one or more mESC clusters (Figure 4.9g). There was no cytotoxicity observed for both PS and PCL particles for 7 days of culturing [117].

The freeze‐drying approach is also effective for encapsulation of enzymes [94], bacteria [118], and cells [119]. The processing is straightforward. The biological entities are suspended in aqueous medium containing monomers or polymers. The resulting suspension can be either cryo‐polymerized [94] or directly freeze‐dried [119]. In the latter case, cryoprotectants may be added to help maintain the biological activity. The choice of encapsulation materials is highly important. These materials should be biocompatible and the formed matrix surrounding the biological entities should be porous and semi‐permeable. That allows the diffusion of oxygen, nutrients, and biomolecules, and also for the removal of wastes [120]. Some natural polymers have been widely used as encapsulation materials, including alginate [118, 119], agarose, chitosan, and HA, and proteins including collagen, gelatin, fibrin, and silk fibroin [120]. Other polymers may be also used, e.g. PVA [118] and PNIPAM [94].

4.4.2 Controlled Drug Release

Hydrogels and porous polymers have been widely used for drug delivery. For the ice‐templated structures, the capability to tune pore morphology by varying freezing conditions and solution compositions may translate to better control in drug delivery. PVA scaffolds have been prepared with tailored porosity and morphology by changing concentration, molecular weight, and freezing rates [46]. The tests on drug release are always performed in aqueous medium such as simulated body fluids (SBF) or phosphate buffer solutions. Apart from instant drug release by dissolving the matrix, water insolubility and stability in water are pre‐requisites for the ice‐templated structures to be used in drug release. Low molecular weight PVAs are readily soluble in water but high molecular weight PVAs with high degree of hydrolysis are dissolved in water only at high temperatures. This is why porous PVA (Mw 72 K, 98% hydrolysis) has been used as scaffold for the diffusional release of ciprofloxacin. The release could be tuned from minutes to days by using PVA scaffolds with different structures and drug loadings [46]. Physically crosslinked PVA hydrogels by the freeze–thaw approach have also been used for drug delivery. For example, the release of flurbiprofen from PVA hydrogels was investigated. The addition of sodium alginate or Pluronic L‐62 could enhance the gel strength but decrease the drug release [121]. Freeze–thaw PVA hydrogel could be formed onto a PVA/PAA substrate, which could increase the gel strength. Such a system could be used for localized delivery of theophylline [122].

Another widely used polymer is chitosan. Chitosan is soluble in acidic water but not in water with neutral pH. Therefore, porous chitosan can be prepared by freeze‐drying the acidic solution. Hydrophilic drugs can be easily included in the solution (or similarly suspending fine hydrophobic particles in the solution) and then incorporated in the porous matrix after freeze‐drying. In a study with chitosan by freeze‐drying, Rhodamine B and BSA were encapsulated in porous chitosan. The morphologies of freeze‐dried chitosan changed from interconnected porous structure to nanofibres by varying chitosan concentrations (from 1 to 0.02 wt%). Slow release was observed for porous chitosan while a much faster release was found for nanofibrous chitosan for both Rhodamine B and BSA [33]. However, to encapsulate water‐soluble proteins in a degradable hydrophobic porous polymer matrix, an emulsion system instead of aqueous solution may be used. Protein‐encapsulated microspheres are usually prepared by this method. By freeze‐drying a water‐in‐oil‐in‐water (W/O/W) double emulsion, protein‐containing PLGA microspheres can be formed in a porous hydrophilic matrix [123]. The release of proteins (e.g. BSA) may be adjusted directly from hydrophobic microspheres with different porosity or entrapping the microspheres within a hydrophilic matrix.

Owing to the nature of the interconnected macropore structures, a burst release and then significantly slowing‐down release are the features of release from hydrogels. In terms of therapeutic treatment, it is desirable that the plasma drug concentration should reach the therapeutic concentration region quickly and then maintain in that region. This may be achieved by a fast initial release followed by a near linear release to complement the drugs being metabolized or removed by the body. It is apparent that hydrogels alone are difficult to achieve this target but it may be realized by incorporating a slow release vehicle into the polymer matrix. As such, a model drug, curcumin was loaded into mesoporous silica microspheres that were then suspended in aqueous curcumin‐chitosan solution. The suspensions were then freeze‐dried to produce curcumin‐containing porous composites [124]. Figure 4.10a shows a network of nanofibres containing silica microspheres. The silica microspheres are either attached to the surface or partly embedded into the fibres. When curcumin was loaded into both chitosan and the mesoporous silica microspheres, a two‐stage release profile can be achieved, a fast initial release followed by near linear release (Figure 4.10b). Clearly, by varying the loading of silica microspheres in the composites, the release profile can be tuned accordingly. The other factors that may be further investigated include porosity of the polymer, drug loading level, and other physically/chemically crosslinked biopolymers.

Figure 4.10 Dual‐controlled release of curcumin from porous silica/chitosan composite. (a) SEM image showing the network of chitosan fibers with embedded silica microspheres (with equal mass of silica and chitosan); (b) The release profiles of curcumin from the composites with different loadings of curcumin‐loaded silica microspheres (50% and 91%, based on chitosan).

Source: Ahmed et al. 2012 [124]. Reprinted with permission from Royal Society of Chemistry.

4.4.3 Encapsulation

Freeze‐drying and spray freeze‐drying are effective methods to encapsulate oil, flavours, and proteins in a hydrophilic matrix, towards improving stability and easy transportation. For a detail description on this topic, interesting readers are referred to Chapter 3. The encapsulation efficiency and the stability and activity of the compounds or particles after reconstitution in water are important parameters. For example, chitosan (a polymer with positive charges) was found to improve retention and redispersibility of a flavour oil in a freeze‐dried emulsion containing a negatively charged emulsifier [125]. Maltodextrin was added as a bulk agent. Under optimal conditions, ∼95% retention levels were obtained and the redispersed droplets were only slightly bigger than the ones before freeze‐drying [125].

One interesting development is to prepare hydrophobic organic nanoparticles [69, 71] or microspheres [117] within porous hydrophilic polymers. PVA has been frequently used as the hydrophilic polymer. The freeze‐dried composites can be readily dissolved in water to release the microspheres [117] or produce stable aqueous nanoparticle dispersions [69, 71]. The hydrophobic organic particles can be extended to poorly water‐soluble drugs. The emulsion‐freeze‐drying approach has been highly effective to address poor water solubility and prepare size‐controlled drug nanoparticles [69, 71, 126, 127]. This concept is described in more detail in Chapter 8.

Alternatively, the freeze‐dried scaffolds can be used directly to absorb oil or organic solutions. Once dissolved in water, the oil phase localized in the pores could be released into water, forming an emulsion or dispersion stabilized by the dissolved polymer molecules. An example is the use of porous PVA containing surfactant SDS prepared by freeze‐drying aqueous solutions or O/W emulsion for the instant formation of emulsions [128]. After soaking porous PVA in cyclohexane for the defined times, the soaked PVA can be directly placed back into water and an O/W emulsion is formed immediately simply by shaking. By controlling the soaking time of cyclohexane in PVA, O/W emulsions with varying percentages of droplet phase can be formed, with the droplet sizes changing in the range of 8–100 μm (the peak size as measured by MasterSizer). This approach has been demonstrated with a variety of oils including soy oil, mineral oil and perfluorocarbons [128]. When the porous PVA is soaked in organic solutions, by evaporating the solvent afterwards, the hydrophobic solute can form nanoparticles in situ within the porous PVA and other polymers. Similarly, dissolution of the polymer in water releases the organic nanoparticles and stable aqueous nanoparticle dispersions are formed [129]. This has proved to be highly effective for the preparation of aqueous poorly water‐soluble drug nanoparticle dispersions [129, 130].

4.4.4 Water Treatment

In water treatment, the removal of heavy metal ions, dyes, and oil from aqueous phase is an important research area with significant practical importance. Molecular imprinted polymer (MIP) hydrogels are efficient for the removal of heavy metal ions [87]. Chitosan nanofibres can be used to remove Cu2+ from aqueous solution at a capacity of 2.57 mmol g−1 [33]. A sponge with a nanofibrous structure prepared by freeze‐drying of cyanobacterial anionic megamolecules shows selective removal of In3+ over Sn4+ at concentrations below 50 mM, which may by significant in recycling indium metal [131].

Owing to their highly interconnected macroporosity, ice‐templated sponges may be more suitable for high capacity oil absorption. However, a hydrophobic environment or hydrophobic surface functionality is required for effective absorption of oils. For example, due to its internal hydrophobic environment, macroporous cyclodextrin materials prepared by freeze‐drying aqueous cyclodextrin solutions showed remarkable absorption of organic solvent including 1,4‐dioxane and ethanol, with the maximum capacities over 20 ml g−1 for soybean oil [132]. A polymeric sponge was prepared by cryo‐crosslinking of polyethyleneimine using 1,4‐butanediol diglycidyl ether as crosslinker at −15 °C for 24 h. The thawed and washed gel was incubated in tetrahydrofuran (THF) for 24 h and dried in a vacuum oven at 55 °C. The dry gel was modified with different acid chlorides (valeroyl chloride (C4), nonanoyl chloride (C8), and palmitoyl chloride (C17)) in dry chloroform at room temperature under Ar. This produced omniphilic sponges that could effectively absorb hexane under water [133]. The C4‐modified sponge could absorb 12 times its weight of water or hexane whilst the C17‐modified sponge could absorb 15 times its weight of hexane or five times its weight in water [133].

Porous sponges made of cellulose nanofibres have recently drawn much attention in removing oil from water [36, 37]. This may be attributed to their highly accessible surface area, extremely high porosity (or ultralightweight), and the interconnected network structure leading to good flexibility. However, because cellulose nanofibres are hydrophilic, surface modification to enhance hydrophobicity and hence oil absorption is necessary. For example, methyltrimethoxysilane (MTMS) was partially hydrolyzed in water at pH 4 with HCl and mixed with aqueous cellulose nanofibre suspension. Freeze‐drying then led to the porous scaffolds with different level of silylation [36]. These sponges proved to be highly efficient in removing dodecane from water with excellent selectivity and recyclability. Depending on the density of the oil liquids, the absorption capacities could be up to 100 times their own weight [36]. In a recent study, bacterial cellulose nanofibres were modified by trimethylsilylation with trimethylchlorosilane and triethylamine in dichloromethane under reflux. The freeze‐dried scaffolds exhibited low density (<6.77 mg cm−3), high surface area (>169 m2 g−1) and high porosity (∼99.6%). As illustrated in Figure 4.11a, the porous scaffold shows high absorption capacity for a range of oils or organic solvents (86–185 g/g, depending on oil density). The equilibrium absorption time varied, depending on the viscosity of the oil phase. It took about 20 s for paraffin oil and plant oil whilst it took only 12 s for an organic solvent to achieve the absorption equilibrium. Owing to the high flexibility, the scaffold showed remarkable reusability (no obvious change in absorption capacity after 10 cycles (Figure 4.11b). This was completed by compressing the absorbed solvent out of the scaffold, rinsing with t‐butanol three times, and freeze‐dried again for the next test [37].

Figure 4.11 Absorption of oils/organic solvents by trimethylsilylated cellulose nanofibrous scaffolds. (a) Graph of the mass‐based absorption capacity for different solvents with different densities. The upper dashed line represents the theoretical volume‐based absorption capacity based on total porosity of the scaffold while the lower dashed line is for 80% of the capacity. It can be seen that most of organic phases can reach ∼90% of theoretical volume capacity. (b) The reusability tests for diesel oil over 10 rinsing–absorption cycles.

Source: Sai et al. 2015 [37]. Reprinted with permission from American Chemical Society.

4.4.5 Liquid Chromatography and Separation

The highly interconnected macroporous cryogels are well suited for bioseparation [98]. There are various reports where the cryogels are reported to be immobilized with biomolecules for cell capture and separation [98]. Vertically orientated porous poly(vinylidene fluoride) (PVDF) membrane was fabricated by ice templating and used for separation of yeasts and lactobacilli based on the size‐exclusion properties. The PVDF membrane prepared from 20 wt% solution (pore sizes ≤ 4.6 μm) could achieve almost 100% separation of yeasts (elliptical cells with diameter 4–6 μm) from lactobacilli (spherical cells with an average diameter of 0.7 μm) [57]. Recently, a solvent crystallization and polymer migration approach was used to construct PVDF membrane with nanopores (much smaller than the usual ice‐templated pores). Such porous membranes demonstrated high water permeation flux, stable flux after fouling, superior mechanical properties and better abrasion resistance, compared to the membranes prepared by the conventional phase‐separation process [134].

Cryogels have also been used as monolithic columns for liquid chromatography because they are able to provide high flow rates at low operating pressures and the macropores can be highly beneficial for separation of large biomolecules and cells [98, 135]. Aligned porous materials have been evaluated as monolithic columns for chromatographic separation with a view to offer low back pressure owing to the aligned microchannel structure. For example, aligned porous silica monoliths in fused‐silica capillaries were prepared from TM‐50 silica colloids and used for normal phase separation (back pressure 9 bar at a flow rate of 0.1 ml min−1) and reversed phase separation (C18‐modification, back pressure 17 bar at a flow rate of 0.1 ml min−1) [136]. A highly crosslinked polyacrylate column was fabricated by UV frozen polymerization [95]. The column was used for the separation of a mixture containing five compounds (uracil, caffeine, phenol, ethylbenzene, phenylene) using 50/50 v/v acetonitrile‐water as the mobile phase. Four well‐resolved peaks were observed, with phenol and caffeine peak overlapping due to the lack of hydroxyl groups on the monolith. The back pressure was 59 bar at a flow rate of 1 ml min−1 [95]. PEG dimethacrylate and poly(butylmethacrylate‐co‐ethylene dimethacrylate) monoliths were also assessed for the separation of protein mixtures under the reverse phase condition and hydrophilic interaction chromatography (HIC) condition, respectively. The proteins could be separated but the performance was not better than that seen in polymer monoliths prepared by convention approaches [97]. This was attributed to the fact that the polymer monolith was not fully attached to the cycloolefin copolymer (COC) tubing even when the COC was pre‐treated with a 1 : 1 (wt%) stock solution of methyl methacrylate (MMA) and ethylene glycol diacrylate (EDA) containing benzophenone as photoinitiator [97]. The full attachment of the monolith to the column wall is critical because that will remove voids near the wall. The presence of voids can lead to heterogeneous flow rates and partial mixing. Better attachment of the polymer monolith to the column also provides improved mechanical stability. Indeed, when preparing the silica monolith column, the fused‐silica capillary was pre‐treated by soaking in 1 M NaOH at 40 °C for 3 h and then flushed with water and acetone. This could increase the density of hydroxyl groups on the wall surface and enhance the interaction with silica colloids [136].

4.4.6 Other Applications

High porosity of ice‐templated polymers can be explored for other applications. PolyHEMA hydrogels with nanowire structures were prepared by cryopolymerization. Ag nanoparticles were then formed on polyHEMA nanowires by simply soaking in AgNO3 solution and subsequent chemical reduction. The aligned nanowire structure provided high surface area and enhanced mass transport to and from the active surface. This was demonstrated by conversion of o‐nitroaniline to 1,2‐benzenediamine with excellent catalytic activity (apparent rate constant up to 0.165 min−1) and reusability (by washing and immersing in water and then used again for 10 cycles) [64].

The high macroporosity of chitosan was utilized to upload an enzyme (glucose dehydrogenase) towards enzyme‐catalyzed flow‐through electrodes [31]. Aqueous enzyme‐chitosan solution was simply pipetted onto a pre‐chilled electrode at −20 °C and kept frozen at this temperature for 1 h. The electrode could then be freeze‐dried for evaluation. The chitosan scaffold prepared this way performed 3.25 times better than the air‐dried chitosan film on electrode, in terms of power density. This could be attributed to greater surface area and enhanced mass transport [31].

The ice‐templating technique may be also used to produce closed‐cell structures that are useful as insulation materials. The solution of wheat gluten in water at a concentration of 14 wt% was processed to produce a homogeneous porous structure with a high content of closed cells [137]. 20 wt% glycerol was added to plasticize the porous material. Owing to the close porosity, the wheat gluten materials were investigated for their thermal conductivity and combustion properties [138]. Both unplasticized and glycerol‐modified materials showed a significantly long time to ignition, a lower effective heat of combustion and a higher char content by cone calorimetry measurement. However, the plasticized materials showed better properties overall. The flammability test by the UL 94 method also demonstrated no dripping from the unplasticized structure [138].

4.5 Summary

Ice templating is a highly effective route to fabricating porous polymers and cryogels. This technique may be applied to solutions (including water, organic solvent, and compressed CO2), suspensions, and emulsions. All the processes involve a freezing step. The frozen samples may be directly freeze‐dried, or cryo‐polymerized and then freeze‐dried, to produce dry porous polymers. The directly freeze‐dried porous polymers are usually mechanically weak and soluble either in water or the originally used solvent. A post‐freeze‐drying crosslinking step may be performed to enhance the stability of the freeze‐dried materials. Alternatively, the frozen sample may undergo a polymerization or crosslinking reaction in the frozen state. Afterwards, the frozen sample is allowed to thaw to generate cryogels. In addition to chemical crosslinking approaches, a physically crosslinking method, multiple cycles of freeze–thaw, can also be used to produce hydrogels (mainly PVA hydrogel).

The porous scaffolds and cryogels by the ice‐templating method have found wide range of applications, particularly as scaffolds for tissue engineering, controlled release, water treatment, separation, and catalysis. For the majority of these applications, crosslinked porous structures or hydrogels are required. For some of the applications, the mechanical stability is an important parameter. There are other applications that utilize the fast dissolution property of the freeze‐dried polymers in water. These include the applications in the instant formation of emulsions and aqueous organic nanoparticle dispersions. The future development in this area will focus on the fabrication of new scaffolds with superior biocompatibility, green credentials, desirable surface functionalities, and required mechanical stability, and immobilization of biomolecules or nanostructures for biomedical applications and separations.

References

  1. 1 Sun, M., Huang, S., Chen, L. et al. (2016). Applications of hierarchically structured porous materials from energy storage and conversion, catalysis, photocatalysis, adsorption, separation, and sensing to biomedicine. Chem. Soc. Rev. 45: 3479–3563.
  2. 2 Tan, L. and Tan, B. (2017). Hypercrosslinked porous polymer materials: design, synthesis, and applications. Chem. Soc. Rev. 46: 3322–3356.
  3. 3 McKeown, N.B. and Budd, P.M. (2006). Polymers of intrinsic microporosity (PIMs): organic materials for membrane separations, heterogeneous catalysis and hydrogen storage. Chem. Soc. Rev. 35: 675–683.
  4. 4 Ramimoghadam, D., Gray, E.M., and Webb, C.J. (2016). Review of polymers of intrinsic microporosity for hydrogen storage applications. Int. J. Hydrogen Energy 41: 16944–16965.
  5. 5 Díaz, U. and Corma, A. (2016). Ordered covalent organic frameworks, COFs and PAFs. From preparation to application. Coord. Chem. Rev. 311: 85–124.
  6. 6 Slate, A.G. and Cooper, A.I. (2015). Functional‐led design of new porous materials. Science 348: 989–999.
  7. 7 Diercks, C.S. and Yaghi, O.M. (2017). The atom, the molecule, and the covalent organic framework. Science 355: 924–931.
  8. 8 Zhang, G. and Mastalerz, M. (2014). Organic cage compounds – from shape‐persistency to function. Chem. Soc. Rev. 42: 1934–1947.
  9. 9 Jaroniec, M. and Schüth, F. (2008). Preface to the special issue: templated materials. Chem. Mater. 20: 599–600.
  10. 10 Qian, L. and Zhang, H. (2011). Porogen incorporation and phase inversion. In: Porous Polymers (ed. M.S. Silverstein, N.R. Cameron and M.A. Hillmyer), 79–117. Hoboken: John Wiley & Sons, Inc.
  11. 11 Petkovich, N.D. and Stein, A. (2013). Controlling macro‐ and mesostructures with hierarchical porosity through combined hard and soft templating. Chem. Soc. Rev. 42: 3721–3739.
  12. 12 Liu, Y., Goebl, J., and Yin, Y. (2013). Templated synthesis of nanostructured materials. Chem. Soc. Rev. 42: 2610–2653.
  13. 13 Silverstein, M.S. (2014). PolyHIPEs: recent advances in emulsion‐templated porous polymers. Prog. Polym. Sci 39: 199–234.
  14. 14 Zhang, H. and Cooper, A.I. (2005). Synthesis and applications of emulsion‐templated porous materials. Soft Matter 1: 107–113.
  15. 15 Zhang, H. and Cooper, A.I. (2005). New approaches to the synthesis of macroporous metals. J. Mater. Chem.15: 2157–2159.
  16. 16 Wu, D., Xu, F., Sun, B. et al. (2012). Design and preparation of porous polymers. Chem. Rev. 112: 3959–4015.
  17. 17 Tsyurupa, M.P. and Davankov, V.A. (2006). Porous structure of hypercrosslinked polystyrene: state‐of‐the‐art mini‐review. React. Funct. Polym. 66: 768–779.
  18. 18 Xu, Y., Jin, S., Xu, H. et al. (2013). Conjugated microporous polymers: design, synthesis and application. Chem. Soc. Rev. 42: 8012–8031.
  19. 19 Qian, L. and Zhang, H. (2011). Controlled freezing and freeze drying: a versatile route for porous and micro‐/nano‐structured materials. J. Chem. Technol. Biotechnol. 86: 172–184.
  20. 20 Deville, S. (2013). Ice‐templating, freeze casting: beyond materials processing. J. Mater. Res. 28: 2202–2219.
  21. 21 Gutiérrez, M.C., Ferrer, M.L., and del Monte, F. (2008). Ice‐templated materials: sophisticated structures exhibiting enhanced functionalities obtained after unidirectional freezing and ice‐segregation‐induced self‐assembly. Chem. Mater. 20: 634–648.
  22. 22 Hollow, J.L., Lowman, A.M., and Palmese, G.R. (2013). The role of crystallization and phase separation in the formation of physically cross‐linked PVA hydrogels. Soft Matter 9: 826–833.
  23. 23 Dhainaut, J., Piana, G., Deville, S. et al. (2014). Freezing‐induced ordering of block copolymer micelles. Chem. Commun. 50: 12572–12574.
  24. 24 Scanlon, S., Aggeli, A., Boden, N. et al. (2009). Organization of self‐assembling peptide nanostructures into macroscopically ordered lamella‐like layers by ice crystallization. Soft Matter 5: 1237–1246.
  25. 25 Qian, L., Ahmed, A., Foster, A. et al. (2009). Systematic tuning of pore morphologies and pore volumes in macroporous materials by freezing. J. Mater. Chem. 19: 5212–5219.
  26. 26 Gotoh, Y., Ishizuka, Y., Matsuura, T., and Niimi, S. (2011). Spheroid formation and expression of liver‐specific functions of human hepatocellular carcinoma‐derived FLC‐4 cells cultured in lactose‐silk fibroin conjugate sponges. Biomacromolecules 12: 1532–1539.
  27. 27 Jalili‐Firoozinezhad, S., Rajabi‐Zeleti, S., Mohammadi, P. et al. (2015). Facile fabrication of egg white macroporous sponges for tissue engineering. Adv. Healthc. Mater. 4: 2281–2290.
  28. 28 Wu, Q., Yu, S., Kollert, M. et al. (2016). Highly absorbing antimicrobial biofoams based on wheat gluten and its biohybrids. ACS Sustain. Chem. Eng. 4: 2395–2404.
  29. 29 Lloyd, C., Besse, J., and Boyce, S. (2015). Controlled‐rate freezing to regulate the structure of collagen‐glycosaminoglycan scaffolds in engineered skin substitute. J. Biomed. Mater. Res. Part B 103: 832–840.
  30. 30 Wan, A.M., Inal, S., Williams, T. et al. (2015). 3D conducting polymer platforms for electrical control of protein conformation and cellular functions. J. Mater. Chem. B 3: 5040–5048.
  31. 31 Clooney, M.J., Lau, C., Windmeisser, M. et al. (2008). Design of chitosan gel pore structure: towards enzyme catalysed flow‐through electrodes. J. Mater. Chem. 18: 667–674.
  32. 32 Qian, L. and Zhang, H. (2010). Green synthesis of chitosan‐based nanofibers and their applications. Green Chem. 12: 1207–1214.
  33. 33 Habibi, Y. (2014). Key advances in the chemical modification of nanocelluloses. Chem. Soc. Rev. 43: 1519–1542.
  34. 34 Martoia, F., Cochereau, T., Dumont, P.J.J. et al. (2016). Cellulose nanofibril foams: links between ice‐templating conditions, microstructures and mechanical properties. Mater. Design 104: 376–391.
  35. 35 Pääkkö, M., Vapaavuori, J., Silvennoinen, R. et al. (2008). Long and entangled native cellulose I nanofibers allow flexible aerogels and hierarchically porous templates for functionalities. Soft Matter 4: 2492–2499.
  36. 36 Zhang, Z., Sèbe, G., Rentsch, D. et al. (2014). Ultralightweight and flexible silyated nanocellulose sponges for the selective removal of oil from water. Chem. Mater. 26: 2659–2668.
  37. 37 Sai, H., Fu, R., Xing, L. et al. (2015). Surface modification of bacterial cellulose aerogels' web‐like skeleton for oil/water separation. ACS Appl. Mater. Interfaces 7: 7373–7381.
  38. 38 Svagan, A.J., Jensen, P., Dvinskikh, S.V. et al. (2010). Towards tailored hierarchical structures in cellulose nanocomposite biofoams prepared by freezing/freeze‐drying. J. Mater. Chem. 20: 6646–6654.
  39. 39 Mandal, B.B. and Kundu, S.C. (2009). Cell proliferation and migration in silk fibroin 3D scaffolds. Biomaterials 30: 2956–2965.
  40. 40 Gil, E.S., Park, S., Tien, L.W. et al. (2010). Mechanically robust, rapidly actuating, and biologically functionalized macroporous poly(N‐isopropylacrylamide)/silk hybrid hydrogels. Langmuir 26: 15614–15624.
  41. 41 Ashokkumar, M., Chipara, A.C., Narayanan, N.T. et al. (2016). Three‐dimensional porous sponges from collagen biowastes. ACS Appl. Mater. Interfaces 8: 14836–14844.
  42. 42 Borisova, A., De Bruyn, M., Budarin, V.L. et al. (2015). A sustainable freeze‐drying route to porous polysaccharides with tailored hierarchical meso‐ and macroporosity. Macromol. Rapid Commun. 36: 774–779.
  43. 43 Zhang, H., Hussain, I., Brust, M. et al. (2005). Aligned two‐ and three‐dimensional structures by directional freezing of polymers and nanoparticles. Nat. Mater. 4: 787–793.
  44. 44 Deville, S., Saiz, E., Nalla, R.K., and Tomsia, A.P. (2006). Freezing as a path to build complex composites. Science 311: 515–518.
  45. 45 Zhang, H. and Cooper, A.I. (2007). Aligned porous structures by directional freezing. Adv. Mater. 19: 1529–1533.
  46. 46 Gutiérrez, M.C., García‐Carvajal, Z.Y., Jobbágy, M. et al. (2007). Poly(vinyl alcohol) scaffolds with tailored morphologies for drug delivery and controlled release. Adv. Funct. Mater. 17: 3505–3513.
  47. 47 Zhou, Y., Fu, S., Pu, Y. et al. (2014). Preparation of aligned porous chitin nanowhisker foams by directional freeze–casting technique. Carbohydr. Polym. 112: 277–283.
  48. 48 Riblett, B.W., Francis, N.L., Wheatley, M.A., and Wegst, U.G.K. (2012). Ice‐templated scaffolds with microridged pores direct DRG neurite growth. Adv. Funct. Mater. 22: 4920–4923.
  49. 49 Gao, H., Lu, Y., Mao, L. et al. (2014). A shape‐memory scaffold for macroscale assembly of functional nanoscale building blocks. Mater. Horiz. 1: 69–73.
  50. 50 Flauder, S., Heinze, T., and Muller, F.A. (2014). Cellulose scaffolds with an aligned and open porosity fabricated via ice‐templating. Cellulose 21: 97–103.
  51. 51 Pot, M.W., Faraj, K.A., Adawy, A. et al. (2015). Versatile wedge‐based system for the construction of unidirectional collagen scaffolds by directional freezing: practical and theoretical considerations. ACS Appl. Mater. Interfaces 7: 8495–8505.
  52. 52 Fereshteh, Z., Fathi, M., Bagri, A., and Boccaccini, A.R. (2016). Preparation and characterization of aligned porous PCL/Zein scaffolds as drug delivery systems via improved unidirectional freeze‐drying method. Mater. Sci. Eng., C 68: 613–622.
  53. 53 Kim, J., Taki, K., Nagamine, S., and Ohshima, M. (2008). Preparation of poly(l‐lactic acid) honeycomb monolith structure by unidirectional freezing and freeze‐drying. Chem. Eng. Sci. 63: 3858–3863.
  54. 54 Kim, J., Taki, K., Nagamine, S., and Ohshima, M. (2009). Preparation of porous poly(l‐lactic acid) honeycomb monolith structure by phase separation and unidirectional freezing. Langmuir 25: 5304–5312.
  55. 55 Shen, H., Niu, Y., Hu, X. et al. (2015). A biomimetic 3D microtubule‐orientated poly(lactide‐co‐glycolide) scaffold with interconnected pores for tissue engineering. J. Mater. Chem. B 3: 4417–4425.
  56. 56 Blaker, J.J., Lee, K.‐Y., Mantalaris, A., and Bismarck, A. (2010). Ice‐microsphere templating to produce highly porous nanocomposite PLA matrix scaffolds with pore selectively lined by bacterial cellulose nano‐whiskers. Compos. Sci. Technol. 70: 1879–1888.
  57. 57 Roberts, A.D., Wang, S., Li, X., and Zhang, H. (2014). Hierarchical porous nitrogen‐rich carbon monoliths via ice‐templating: high capacity and high‐rate performance as lithium‐ion battery anode materials. J. Mater. Chem. A 2: 17787–17796.
  58. 58 Roberts, A.D., Lee, J.M., Wong, S.Y. et al. (2017). Nitrogen‐rich activated carbon monoliths via ice‐templating with high CO2 and H2 adsorption capacities. J. Mater. Chem. A. 5: 2811–2820.
  59. 59 Liang, H., Ji, K., Zha, L. et al. (2016). Polymer membranes with vertically orientated pores constructed by 2D freezing at ambient temperature. ACS Appl. Mater. Interfaces 8: 14174–14181.
  60. 60 Bhamidipati, M., Scurto, A.M., and Detamore, M.S. (2013). The future of carbon dioxide for polymer processing in tissue engineering. Tissue Eng. Part B 19: 221–232.
  61. 61 Zhang, H. and Cooper, A.I. (2003). Compressed fluid sedimentation polymerization. Macromolecules 36: 5061–5064.
  62. 62 Diaz‐Gomez, L., Yang, F., Jansen, J.A. et al. (2016). Low viscosity‐PLGA scaffolds by compressed CO2 foaming for growth factor delivery. RSC Adv. 6: 70510–70519.
  63. 63 Zhang, H., Long, J., and Cooper, A.I. (2005). Aligned porous materials by directional freezing of solution in liquid CO2. J. Am. Chem. Soc. 127: 13482–13483.
  64. 64 Mao, Q., Shi, S., and Wang, H. (2015). Biomimetic nanowire structured hydrogels as highly active and recyclable catalyst carriers. ACS Sustainable Chem. Eng. 3: 1915–1924.
  65. 65 Qian, L., Willneff, E., and Zhang, H. (2009). A novel route to polymeric sub‐micron fibers and their use for inorganic structures. Chem. Commum. 3946–3948.
  66. 66 Spender, J., Demers, A.L., Xie, X. et al. (2012). Method for production of polymer and carbon nanofibers from water‐soluble polymers. Nano Lett. 12: 3857–3860.
  67. 67 Liu, X., Roberts, A.D., Ahmed, A. et al. (2015). Carbon nanofibers by pyrolysis of self‐assembled perylene diimide derivative gels as supercapacitor electrode materials. J. Mater. Chem. A 3: 15513–15522.
  68. 68 Liu, X., Ahmed, A., Wang, Z., and Zhang, H. (2015). Nanofibrous microspheres via emulsion gelation and carbonization. Chem. Commun. 51: 16864–16867.
  69. 69 Zhang, H., Wang, D., Butler, R. et al. (2008). Formation and enhanced biocidal activity of water‐dispersible organic nanoparticles. Nat. Nanotechnol. 3: 506–511.
  70. 70 Ahmed, A., Smith, J., and Zhang, H. (2011). Gradient porous materials by emulsion centrifugation. Chem. Commun. 47: 11754–11756.
  71. 71 Wais, U., Jackson, A.W., He, T., and Zhang, H. (2016). Nanoformulation and encapsulation approaches for poorly water‐soluble drug nanoparticles. Nanoscale 8: 1746–1769.
  72. 72 Barbetta, A., Carrino, A., Costantini, M., and Dentini, M. (2010). Polysaccharide based scaffolds obtained by freezing the external phase of gas‐in‐liquid foams. Soft Matter 6: 5213–5224.
  73. 73 Hassan, C.M. and Peppas, N.A. (2000). Structure and applications of poly(vinyl alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing methods. Adv. Polym. Sci. 153: 37–65.
  74. 74 Willcox, P.J., Howie, D.W., Schmidt‐Rohr, K. et al. (1999). Microstructure of poly(vinyl alcohol) hydrogels produced by freeze/thaw cycling. J. Polym. Sci., Part B: Polym. Phys. 37: 3438–3454.
  75. 75 Ricciardi, R., Auriemma, F., Gaillet, C. et al. (2004). Investigation of the crystallinity of freeze/thaw poly(vinyl alcohol) hydrogels by different techniques. Macromolecules 37: 9510–9516.
  76. 76 Hassan, C.M. and Peppas, N.A. (2000). Structure and morphology of freeze/thawed PVA hydrogels. Macromolecules 33: 2472–2479.
  77. 77 McGann, M.J., Higginbotham, C.L., Geever, L.M., and Nugent, M.J.D. (2009). The synthesis of novel pH‐sensitive poly(vinyl alcohol) composite hydrogels using a freeze/thaw process for biomedical applications. Int. J. Pharm. 372: 154–161.
  78. 78 Llanos, G.R. and Sefton, M.V. (1991). Immobilization of poly(ethylene glycol) onto a poly(vinyl alcohol) hydrogel. 1. Synthesis and characterization. Macromolecules 24: 6065–6072.
  79. 79 Mathews, D.T., Birney, Y.A., Cahill, P.A., and McGuinness, G.B. (2008). Mechanical and morphological characteristics of poly(vinyl alcohol)/chitosan hydrogels. J. Appl. Polym. Sci. 109: 1129–1137.
  80. 80 Xi, L., Jiang, M., Dong, X. et al. (2012). Controlled mechanical and swelling properties of poly(vinyl alcohol)/sodium alginate blend hydrogels prepared by freeze–thaw followed by Ca2+ crosslinking. J. Appl. Polym. Sci. 124: 823–831.
  81. 81 Qi, X., Hu, X., Wei, W. et al. (2015). Investigation of Salecan/poly(vinyl alcohol) hydrogels prepared by freeze/thaw method. Carbohydr. Polym. 118: 60–69.
  82. 82 Abitbol, T., Johnstone, T., Quinn, T.M., and Gray, D.G. (2011). Reinforcement with cellulose nanocrystals of poly(vinyl alcohol) hydrogels prepared by cyclic freezing and thawing. Soft Matter 7: 2373–2379.
  83. 83 Zhang, H., Xia, H., and Zhao, Y. (2012). Poly(vinyl alcohol) hydrogel can autonomously self‐heal. ACS Macro Lett. 1: 1233–1236.
  84. 84 Guan, Y., Bian, J., Peng, F. et al. (2014). High strength of hemicelluloses based hydrogels by freeze/thaw technique. Carbohydr. Polym. 101: 272–280.
  85. 85 Zhang, H., Zhang, F., and Wu, J. (2013). Physically crosslinked hydrogels from polysaccharides prepared by freeze–thaw technique. React. Funct. Polym. 73: 923–928.
  86. 86 Henderson, T.M.A., Ladewig, K., Haylock, D.N. et al. (2013). Cryogels for biomedical applications. J. Mater. Chem. B 1: 2682–2695.
  87. 87 Choudhury, S., Connolly, D., and White, B. (2015). Supermacroporous polyHIPE and cryogel monolithic materials as stationary phases in separation science: a review. Anal. Methods 7: 6967–6982.
  88. 88 Kirsebom, H. and Mattiasson, B. (2011). Cryostructuration as a tool for preparing highly porous polymer materials. Polym. Chem. 2: 1059–1062.
  89. 89 Gun'ko, V.M., Savina, I.N., and Mikhalovsky, S.V. (2013). Cryogels: Morphological, structure and adsorption characterization. Adv. Colloid Interface Sci. 187‐188: 1–46.
  90. 90 Plieva, F.M., Karlsson, M., Aguilar, M. et al. (2005). Pore structure in supermacroporous polyacrylamide based cryogels. Soft Matter 1: 303–309.
  91. 91 Zhan, X., Lu, D., Lin, D., and Yao, S. (2013). Preparation and characterization of supermacroporous polyacrylamide cryogel beads for biotechnological application. J. Appl. Polym. Sci. 130: 3082–3089.
  92. 92 Srivastava, A., Jain, E., and Kumar, A. (2007). The physical characterization of supermacroporous poly(N‐isopropylacrylamide) cryogel: mechanical strength and swelling/de‐swelling kinetics. Mater. Sci. Eng., A 464: 93–100.
  93. 93 Petrov, P., Petrova, E., and Tsvetanov, C.B. (2009). UV‐assisted synthesis of super‐macroporous polymer hydrogels. Polymer 50: 1118–1123.
  94. 94 Petrov, P., Pavlova, S., Tsvetanov, C.B. et al. (2011). In situ entrapment of urease in cryogels of poly(N‐isopropylacrylamide): an effective strategy for noncovalent immobilization of enzymes. J. Appl. Polym. Sci. 122: 1742–1748.
  95. 95 Barrow, M., Eltmimi, A., Ahmed, A. et al. (2012). Frozen polymerization for aligned porous structures with enhanced mechanical stability, conductivity, and as stationary phase for HPLC. J. Mater. Chem. 22: 11615–11620.
  96. 96 Barrow, M. and Zhang, H. (2013). Aligned porous stimuli‐responsive hydrogels via directional freezing and frozen UV initiated polymerization. Soft Matter 9: 2723–2729.
  97. 97 Arrua, R.D. and Hilder, E.F. (2015). Highly ordered monolithic structures by directional freezing and UV‐initiated cryopolymerisation: evaluation as stationary phases in high performance liquid chromatography. RSC Adv. 5: 71131–71138.
  98. 98 Ertürk, G. and Mattiasson, B. (2014). Cryogels‐versatile tools in bioseparation. J. Chromatogr. A 1357: 24–35.
  99. 99 Ak, F., Oztoprak, Z., Karakutuk, I., and Okay, O. (2013). Macroporous silk fibroin cryogels. Biomacromolecules 14: 719–727.
  100. 100 Tam, R.Y., Fisher, S.A., Baker, A.E.G., and Shoichet, M.S. (2016). Transparent porous polysaccharide cryogels provide biochemically defined, biomimetic matrices for tunable 3D cell culture. Chem. Mater. 28: 3762–3770.
  101. 101 Welzel, P.B., Grimmer, M., Renneberg, C. et al. (2012). Macroporous starPEG‐heparin cryogels. Biomacromolecules 13: 2349–2358.
  102. 102 Zhang, H., Hardy, G.C., Khimyak, Y.Z. et al. (2004). Synthesis of hierarchically porous silica and metal oxide beads using emulsion‐templated polymer scaffolds. Chem. Mater. 16: 4245–4256.
  103. 103 Wu, J., Lin, Y., and Sun, J. (2012). Anisotropic volume change of poly(N‐isopropylacrylamide)‐based hydrogels with an aligned dual‐network microstructure. J. Mater. Chem. 22: 17449–17451.
  104. 104 Okaji, R., Taki, K., Nagamine, S., and Ohshima, M. (2012). Preparation of porous honeycomb monolith from UV‐curable monomer/dioxane solution via unidirectional freezing and UV irradiation. J. Appl. Polym. Sci. 125: 2874–2882.
  105. 105 Okaji, R., Taki, K., Nagamine, S., and Ohshima, M. (2013). Preparation of a unique, multihollow‐core honeycomb structure via the unidirectional freezing of a binary solvent system. J. App. Polym. Sci. 130: 526–534.
  106. 106 Zhu, J., Wang, J., Lin, Q. et al. (2013). Anisotropic tough poly(2‐hydroxyethyl methacrylate) hydrogels fabricated by directional freezing redox polymerization. J. Mater. Chem. B 1: 978–986.
  107. 107 Wu, J., Zhao, Q., Sun, J., and Zhou, Q. (2012). Preparation of poly(ethylene glycol) aligned porous cryogels using a unidirectional freezing technique. Soft Matter 8: 3620–3626.
  108. 108 Zhao, D., Zhu, J., Zhu, Z. et al. (2014). Anisotropic hierarchical porous hydrogels with unique water loss/absorption and mechanical properties. RSC Adv. 4: 30308–30314.
  109. 109 Tripathi, A., Kathuria, N., and Kumar, A. (2009). Elastic and macroporous agarose–gelatin cryogels with isotropic and anisotropic porosity for tissue engineering. J. Biomed. Mater. Res. 90A: 680–694.
  110. 110 Wu, J., Zhao, Q., Liang, C., and Xie, T. (2013). Enzymatically degradable oxidized dextran‐chitosan hydrogels with an anisotropic aligned porous structure. Soft Matter 9: 11136–11142.
  111. 111 Dogu, S. and Okay, O. (2008). Tough organogels based on polyisobutylene with aligned porous structures. Polymer 49: 4626–4634.
  112. 112 Frydrych, M. and Chen, B. (2013). Large three‐dimensional poly(glycerol sebacate)‐based scaffolds – a freeze‐drying preparation approach. J. Mater. Chem. B 1: 6650–6661.
  113. 113 Williams, D.F. (2008). On the mechanisms of biocompabitility. Biomaterials 29: 2941–2953.
  114. 114 Kirkpatrick, C.J., Bittinger, F., Wagner, M. et al. (1998). Current trends in biocompatibility testing. Proc. Inst. Mech. Eng. H 212: 75–84.
  115. 115 Jurga, M., Dainiak, M.B., Sarnowska, A. et al. (2011). The performance of laminin‐containing cryogel scaffolds in neural tissue regeneration. Biomaterials 32: 3423–3434.
  116. 116 Qian, L., Ahmed, A., Glennon‐Alty, L. et al. (2015). Patterned substrates fabricated by a controlled freezing approach and biocompatibility evaluation by stem cells. Mater. Sci. Eng., C 49: 390–399.
  117. 117 Zhang, H., Edgar, D., Murray, P. et al. (2008). Synthesis of porous microparticles with aligned porosity. Adv. Funct. Mater. 18: 222–228.
  118. 118 Gutiérrez, M.C., García‐Carvajal, Z.Y., Jobbágy, M. et al. (2007). Hydrogel scaffolds with immobilized bacteria for 3D cultures. Chem. Mater. 19: 1968–1973.
  119. 119 Christoph, S., Kwiatoszynski, J., Coradin, T., and Fernandes, F.M. (2016). Cellularized cellular solids via freeze‐casting. Macromol. Biosci. 16: 182–187.
  120. 120 Gasperini, L., Mano, J.F., and Reis, R.L. (2014). Natural polymers for the microencapsulation of cells. J. R. Soc. Interface 11, 20140817.
  121. 121 Takamura, A., Ishii, F., and Hidaka, H. (1992). Drug release from poly(vinyl alcohol) gel prepared by freeze‐thaw procedure. J. Controlled Release 20: 21–28.
  122. 122 Nugent, M.J.D. and Higginbotham, C.L. (2007). Preparation of a novel freeze thawed poly(vinyl alcohol) composite hydrogel for drug delivery applications. Eur. J. Pharm. Biopharm. 67: 377–386.
  123. 123 Qian, L. and Zhang, H. (2013). One‐step synthesis of protein‐encapsulated microspheres in a porous scaffold by freeze‐drying double emulsions and tuneable protein release. Chem. Commun. 49: 8833–8835.
  124. 124 Ahmed, A., Hearn, J., Abdelmagid, W., and Zhang, H. (2012). Dual‐tuned drug release by nanofibrous scaffolds of chitosan and mesoporous silica microspheres. J. Mater. Chem. 22: 25027–25035.
  125. 125 Kaasgaard, T. and Keller, D. (2010). Chitosan coating improves retention and redispersibility of freeze‐dried flavour oil emulsions. J. Agric. Food Chem. 58: 2446–2454.
  126. 126 Grant, N. and Zhang, H. (2011). Poorly water‐soluble drug nanoparticles via an emulsion‐freeze‐drying approach. J. Colloid Interface Sci. 356: 573–578.
  127. 127 Wais, U., Jackson, A.W., Zuo, Y. et al. (2016). Drug nanoparticles by emulsion‐freeze‐drying via the employment of branched block copolymer nanoparticles. J. Controlled Release 222: 141–150.
  128. 128 Qian, L. and Zhang, H. (2010). Direct formation of emulsions using water‐soluble porous polymers as sacrificial scaffolds. J. Chem. Technol. Biotechnol. 85: 1508–1514.
  129. 129 Qian, L., Ahmed, A., and Zhang, H. (2011). Formation of organic nanoparticles by solvent evaporation within porous polymeric materials. Chem. Commun. 47: 10001–10003.
  130. 130 Roberts, A.D. and Zhang, H. (2013). Poorly water‐soluble drug nanoparticles via solvent evaporation in water‐soluble porous polymers. Int. J. Pharm. 447: 241–250.
  131. 131 Okajima, M.K., Nakamura, M., Ogawa, T. et al. (2012). Spongy hydrogels of cyanobacterial polyanions mediate energy‐saving electrolytic metal‐refinement. Ind. Eng. Chem. Res. 51: 8704–8707.
  132. 132 Marui, Y., Kikuzawa, A., Kida, T., and Akashi, M. (2010). Unique organogel formation with macroporous materials constructed by the freeze‐drying of aqueous cyclodextrin solutions. Langmuir 26: 11441–11445.
  133. 133 Chatterjee, S., Gupta, S.S., and Kumaraswamy, G. (2016). Omniphilic polymeric sponges by ice templating. Chem. Mater. 28: 1823–1831.
  134. 134 Wang, B., Ji, J., and Li, K. (2016). Crystal nuclei templated nanostructured membranes prepared by solvent crystallization and polymer migration. Nat. Commun. 7: 12804.
  135. 135 Hayes, R., Ahmed, A., Edge, T., and Zhang, H. (2014). Core–shell particles: Preparation, fundamentals and applications in high performance liquid chromatography. J. Chromatogr. A 1357: 36–52.
  136. 136 Ahmed, A., Myers, P., and Zhang, H. (2012). Preparation of aligned porous silica monolithic capillary columns and their evaluation for HPLC. Anal. Methods 4: 3942–3947.
  137. 137 Blomfeldt, T.O.J., Kuktaite, R., Johansson, E., and Hedenqvist, M.S. (2011). Mechanical properties and network structure of wheat gluten foams. Biomacromolecules 12: 1707–1715.
  138. 138 Blomfeldt, T.O.J., Nilsson, F., Holgate, T. et al. (2012). Thermal conductivity and combustion properties of wheat gluten foams. ACS Appl. Mater. Interfaces 4: 1629–1635.