Preparation of hydrogels via ultrasonic polymerization

Ultrasonics Sonochemistry 17 (2010) 326–332

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultsonch

Preparation of hydrogels via ultrasonic polymerization Peter Cass *, Warren Knower, Eliana Pereeia, Natalie P. Holmes, Tim Hughes CSIRO Molecular and Health Technologies, Bayview Avenue, Clayton 3168 VIC, Australia

a r t i c l e

i n f o

Article history: Received 16 March 2009 Received in revised form 12 August 2009 Accepted 19 August 2009 Available online 23 August 2009 Keywords: Ultrasound Sonochemistry Hydrogel Polymerization

a b s t r a c t Several acrylic hydrogels were prepared via ultrasonic polymerization of water soluble monomers and macromonomers. Ultrasound was used to create initiating radicals in viscous aqueous monomer solutions using the additives glycerol, sorbitol or glucose in an open system at 37 °C. The water soluble additives were essential for the hydrogel production, glycerol being the most effective. Hydrogels were prepared from the monomers 2-hydroxyethyl methacrylate, poly(ethylene glycol) dimethacrylate, dextran methacrylate, acrylic acid/ethylene glycol dimethacrylate and acrylamide/bis-acrylamide. For example a 5% w/w solution of dextran methacrylate formed a hydrogel in 6.5 min in a 70% w/w solution of glycerol in water at 37° C with 20 kHz ultrasound, 56 W cm2. The ultrasonic polymerization method described here has a wide range of applications such a biomaterial synthesis where initiators are not desired. Crown Copyright Ó 2009 Published by Elsevier B.V. All rights reserved.

1. Introduction Ultrasonic radiation is used in a vast range of applications. More commonly this includes imaging (industrial and medical) [1,2], physiotherapy and welding of metals and plastics. High frequencies are commonly used for imaging, whereas low frequencies (less than 500 kHz) provide strong physical and chemical interaction with materials. The physical effects of low frequency (high power) ultrasound include high shear rates, free radical production in solution and heat generation. This has been utilized for fragmentation and dissolution of solids, biological techniques, the preparation of nanomaterials, various synthetic sonochemical reactions, thermal applications such as plastic welding and environment remediation [3]. Ultrasound has also been investigated for many polymer based applications. These include polymer synthesis through the generation of free radicals, activation of free radical initiators, degradation of polymers in solution by high shear rates and physical mixing of heterogeneous emulsion/suspension polymerization systems [4]. The first report of ultrasonic polymerization without the use of a free radical initiator was by Lindstrom and Lamm in 1950 [5]. They successfully polymerized acrylonitrile in water. This followed the work of Grabar and Prudhomme [6] who demonstrated that free radicals can be produced in water by low frequency ultrasound in the absence of oxygen and without the addition of conventional free radical initiators. This is believed to occur via a process known as cavitation. Under ultrasonic irradiation, bubbles of gases in solu* Corresponding author. Tel.: +61 3 9545 2428; fax: +61 3 9545 2446. E-mail address: [email protected] (P. Cass).

tion rapidly expand and collapse as a result of rapidly oscillating shock waves. The rapid collapse can produce localized high pressure and temperatures estimated to be approximately 5000 K [7] and result in the scission of solvent molecules to form free radicals. This is known to occur in both aqueous and organic solvents. Ultrasound has been investigated for the initiation of many different solution and bulk polymerization reactions. These include polymerization of styrene [8–11], acrylonitrile [5], acrylamide [12] and acrylic acid [13]. For emulsion polymerization, ultrasound can be used for both dispersion of monomer droplets and for generation of free radicals. Monomers commonly investigated include styrene [14–18], methyl methacrylate [8,19–26], butyl acrylate [17,18,25,27] and vinyl acetate [27]. A similar polymerization system to emulsion polymerization uses supercritical carbon dioxide as the dispersant phase [28,29]. This provides an advantageous method for isolating the polymer after the reaction. Block copolymers can also be produced by ultrasound. When high molecular weight polymer chains break in an ultrasonic field, macroradicals are produced. They are then capable of either initiating solution polymerization [30–32] or recombining with different macroradicals [33]. This may also be extended by preparing polymer with labile peroxide midsections of the polymer which degrade more easily [34]. The commercial application of ultrasound to produce polymers has been limited due to the high power demands and low conversion of monomer. Most research has focused around industrially significant monomers such as alkyl methacrylates and styrene. There has been relatively little information on the polymerization of water soluble monomers by radicals generated by ultrasound and to our knowledge this has not been reported for the production

1350-4177/$ - see front matter Crown Copyright Ó 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ultsonch.2009.08.008

327

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332

of hydrogels. Our current work investigates the polymerization of water soluble acrylic monomers by ultrasound with the inclusion of viscosity enhancing additives. The use of different monomers and solvent compositions and their effect on reaction kinetics are explored. We propose that this technology may find use in the biomedical materials field and may be suitable for the industrial preparation of hydrogels. This work is currently the subject of a provisional patent application [35].

US horn

solvent / monomer

thermometer

cooling water

sonifier

2. Methods

magnetic stirrers

2.1. Materials All chemicals used in these experiments were supplied by Sigma–Aldrich. High purity anhydrous glycerol, 2-isocyanatoethyl methacrylate and dibutyltindilaurate were used as supplied. Monomers 2-hydroxyethyl methacrylate (HEMA) and acrylic acid (AA) were purified by distillation under vacuum prior to use. Acrylamide (Am) was purified by recrystalisation. Monomers poly(ethylene glycol) dimethacrylate (PEG-DMA) (Mn 4344) and ethylene glycol dimethacrylate (EGDMA) were cleared of inhibitor by passing through a column of inhibitor removal resin. Dimethylsulfoxide was dried over activated 4 Å molecular sieves for 5 days. Bis-acrylamide (Bis-Am), dextran (Mn 10,000) and poly(HEMA) (Mn 20,000) were used without purification. 2.2. Dextran methacrylate (Dex-MA) synthesis Dextran methacrylate was prepared by a similar method to that used by Thermes et al. [36]. Dextran was dried over phosphorus pentoxide in vacuum for 5 days. 10.06 g of dry dextran (Mn 10,000) was added to a round bottom flask and 60 ml of dry dimethylsulfoxide was added to dissolve the dextran. 3.14 g of 2isocyanatoethyl methacrylate was added to the flask followed by 0.2 ml of 20% w/v dibutyltindilaurate in toluene. The flask was fitted with a drying tube and placed in an oil bath at 70 °C for 20 h. The solution was then cooled to room temperature and precipitated into ethanol to isolate the product. The polymer was then filtered and dried at room temperature under high vacuum. The sample was analyzed by proton nuclear magnetic resonance spectroscopy (1H NMR) using a 400 MHz Bruker NMR spectrometer. Following synthesis, the methacrylate loading on dextran was determined to be 20.0% with respect to repeat units by comparing the relative area of the vinylic proton at 5.6 ppm with an alcohol proton at 4.6 ppm. 2.3. Standard ultrasound polymerization conditions Fifteen grams of the polymerization solvent and monomer were preheated to 35 °C and suspended in an empty cooling bath (Fig. 1). The ultrasonic horn (Branson 250, probe diameter 1.8 mm) and thermometer were lowered 10 mm into the mixture. On commencement of sonication the cooling water temperature was quickly adjusted with ice to ensure the polymerization solvent remained at 37 °C. This temperature was selected to simulate typical body temperatures for potential in vivo use. The polymerization/curing reaction was terminated by pouring the liquid (or gel) into 200 ml of distilled water and rapidly stirring for 30 min. If any gel had formed it was important to ensure that large particles were broken while stirring to aid in the extraction of additives and unreacted monomer. The polymer was then either filtered or the supernatant decanted followed by refreshing the distilled water and stirring for an additional 30 min. For gelled products this extraction procedure was repeated three additional times in warm

hotplate/ stirrer Fig. 1. Schematic diagram of the ultrasound apparatus and setup used for the polymerization of vinyl monomers and the formation of hydrogels.

water (60 °C). The gel was then dried under vacuum for 8 h at 60 °C and the conversion was determined gravimetrically. 2.4. Preparation of Dex-MA hydrogel using a thermal free radical initiator This hydrogel was prepared as a comparison to the ultrasonically produced hydrogel. Prior to use all solvents were warmed and purged with high purity nitrogen for 30 min to remove oxygen. An initiator solution was prepared from 7.5 mg of 4,40 -azobis(4cyanopentanoic acid) and 4.5 mg of sodium hydrogen carbonate in 1 g of water. This solution was then injected with gas tight syringes into a vial containing of 0.75 g Dex-MA, 3.27 g of water and 9.97 g of glycerol. The vial was then heated at 70 °C for 2 h. The extractable materials of the resultant hydrogel were then exchanged for water using the same procedure specified in Section 2.3. 2.5. Analysis techniques Fourier Transform Infrared (FTIR) analysis was performed by pressing a dry potassium bromide disk containing 0.1 mg of poly (HEMA). The sample was measured on a Perkin–Elmer Spectrum 100 infrared spectrometer. Viscosities were measured with an Ubbelohde-type glass capillary tube viscometer with a Schott Geräte automatic measuring unit (model AVS 350) in a water bath at 37 °C. The tube was a Schott Geräte type 526 20/II. Replicate measurements were performed on 20 ml samples. Swelling ratio tests were performed on the Dex-MA hydrogels as a relative measure of the degree of cross linking. The swelling ratio based on mass (Qm) was calculated by dividing the gel mass after swelling (Ms) by the dried gel mass (Md). Measurements were performed in duplicate. Scanning electron microscopy (SEM) studies were performed on the Dex-MA hydrogels using a Phillips XL30 Field Emission Scanning Electron Microscope (FESEM) at 1 kV with a Polaron LT7400 cryo-preparation system for the imaging of frozen hydrated samples. Imaging was performed at 190 °C. 3. Results and discussion A 10% w/w HEMA aqueous solution was used in the preliminary gel formation experiments. Fifteen grams of this solution was sonicated (20 kHz, 40 W cm2) for 20 min. However, only trace

328

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332

quantities of polymer were produced. In order to increase the polymerization rate by increasing the number of propagating free radicals, the cavitation intensity needed to be maximized. This was achieved by increasing the viscosity of the solution by substituting part of the water with glycerol. It is well known that the intensity of the cavitation and hence the population of free radicals is proportional to the square root of viscosity [37]. Within minutes of sonicating a 10% solution of HEMA in 70% glycerol/water a white flocculated polymer resulted. This was confirmed to be poly(HEMA) by comparing the infrared spectrum of the ultrasound product with a commercial poly(HEMA) and the starting material (Fig. 2). 3.1. Monomers Several common monomer and macromonomer systems were chosen for this work. The selection criteria was based on the monomer water solubility and the inclusion of polymerizable vinyl groups. The monomers included HEMA, PEG-DMA, Dex-MA, AA/ EGDMA and Am/Bis-Am. Dextran methacrylate was chosen as it represented a multifunction macromonomer that will have a rapid gelation rate. The monomers were added as a 10% w/w solution in 70% w/w glycerol and sonicated at 37 °C for between 5 and 30 min. The resultant polymeric hydrogels were purified by leaching the unreacted monomer and glycerol from the hydrogel in distilled water. The hydrogel was then dehydrated and the extent of the reaction was determined gravimetrically. An indicative test for hydrogel formation was performed by rapidly stirring the mixture in hot distilled water. Lack of solubility was evidence of the formation of a covalently crosslinked hydrogel. The reaction conditions and conversions of the monomers can be seen in Table 1. Two macromonomers, PEG-DMA and Dex-MA were chosen for their potential to be used in biomedical applications. Dex-MA gelation was much more rapid than PEG-DMA. This was the result of the higher concentration of methacrylate groups in the Dex-MA system. This is consistent with the low conversion of PEG-DMA (7% w/w). The hydrogel produced from Dex-MA was a continuous and relatively clear gel compared with PEG-DMA which formed a continuous opaque gel. Its lack of clarity is most likely the result of its sparing solubility in 70% w/w glycerol/water.

AA/EGDMA and Am/Bis-Am were polymerized to form continuous clear hydrogels in approximately 25 and 5.5 min, respectively (see Table 1.). The monomer components EGDMA and BisAm were added as approximately 20% w/w of the total monomer concentration to promote branching and eventually crosslinking. Such an addition was not necessary to promote crosslinking of HEMA. Gelation of HEMA without a crosslinker most likely occurred by a radical addition process via an intermediate generated by proton abstraction. Such a reaction has been observed by Hill et al. for HEMA polymerization by c irradiation [38]. Their crosslinking leading to gel formation was attributed to radical formation on a methylene group in a HEMA side chain. Another explanation for the crosslinking results from trace quantities of EGDMA, a known impurity of HEMA, which is difficult to remove due to its similar boiling point [39]. Our white flocculated water insoluble poly(HEMA) is likely to be the result of either of these processes combined with the poor solubility of the polymer in the 70% w/w glycerol/ water solution. When HEMA was polymerized by ultrasound in distilled water alone a water soluble poly(HEMA) was produced (see Table 1.). This is indicative of either a linear polymer, a polymer with very low crosslink density or an ultrasonically degraded polymer. 3.2. Solvent studies The composition of the glycerol/water polymerization solvent was varied for the ultrasonic polymerization of HEMA to determine optimum operating concentrations. As can be seen in Fig. 3, above 50% glycerol the reaction conversion at fixed time increases almost linearly. The maximum conversion occurs at 95% glycerol in water. At 100% glycerol there is a small decrease in activity as a result of intermittent cavitation due to power levels not being sufficient for a highly viscous solution. This may also be reduced by the vapor pressure of the solvent decreasing dramatically. The increasing, conversion above 50% is likely to be explained by the increasing solution viscosity leading to intensified cavitation as well as reduced radical termination reactions. Fig. 3 shows that the increase in conversion (constant polymerization time) coincides with an increase in the square root of the solution viscosity at a comparable

% Transmittance

HEMA

pHEMA prepared by ultrasound

Commercial pHEMA

3980

3480

2980

2480

1980

1480

980

480

-1

Wavelength (cm ) Fig. 2. FTIR spectra of HEMA and poly(HEMA) produced by ultrasound in glycerol/water compared with commercial poly(HEMA).

329

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332 Table 1 Preparation of several hydrogels by ultrasound, T = 37 °C, ultrasonic power = 56 W cm2. Monomer mass (g)

Solvent

80% AA/20% EGDMA 87% Am/13% BisAm HEMA PEG-DMA Dex-MA HEMA

1.5 1.5 1.5 3 0.75 1.5

13.5 g 13.5 g 13.5 g 12.0 g 14.2 g 13.5 g

70% w/w 70% w/w 70% w/w 70% w/w 70% w/w water

glycerol glycerol glycerol glycerol glycerol

Sonication time (min)

% Conversion

30 min (gelled at 25 min) 5.5 min (gel time) 20 min 15 min 6.5 min (gel time) 60 min

45 35 40 7 72 68*

No hydrogel produced. Result determined from the evaporation of unreacted monomer and solvent.

% Conversion

*

Monomer mixture

80

18

70

16

60

14 12

50

10 40

8

30

6

20

4

10

2

0 0

20

40

60

80

0 100

% Glycerol Fig. 3. Conversion of HEMA and the square root of viscosity as a function of% glycerol in water, [HEMA] = 4.0% w/w, solution mass = 15.0 g, ultrasonic power = 56 W cm2, t = 10 min, T = 37 °C.

enhancing additives decrease the solubility of the formed polymer/hydrogel enabling it to resist depolymerization. This is supported by the observation of PEG-DMA having decreased solubility as the glycerol component of the solvent is increased while still being effectively polymerized. Therefore if any hydrogel were produced by the polymerization of HEMA in water it is probable that it was simultaneously depolymerized to a limiting molecular weight. This highlights that the role of the solvent additive is likely to be dual purpose, for increasing viscosity and reducing polymer solubility. However depolymerization using our current polymerization system with glycerol was still achievable simply by increasing the ultrasonic power. For example, the Dex-MA hydrogel which was polymerized in 6.5 min with 56 W cm2 of ultrasound was subsequently depolymerized to become a clear liquid in 20 s with 80 W cm2 of ultrasound.

3.3. Reaction progress and hydrogel properties The progress of the gelling polymerization reaction was investigated for HEMA and Dex-MA in glycerol/water. After an apparent induction period the HEMA and multifunctional Dex-MA quickly formed a gelled polymer (Fig. 4). The induction period may be a result of either autocatalytic polymerization process, chain extension prior to gelation or the ultrasound degassing oxygen from the sys-

80

60

% Conversion

glycerol/water composition. In addition to the effect of the viscosity, the reaction rate is also likely to increase as the concentration of glycerol increases in the sonication medium. Aliphatic alcohols such as glycerol are known to increase the population of radicals in solution as secondary alkyl radicals are formed [40,41]. These have greater stability than water derived hydroxyl radicals. Preparation of hydrogel by sonification was also investigated in several aqueous solutions. These include glycerol/water, sorbitol/ water and glucose/water. The intended purpose of the solute was to increase the viscosity and hence the cavitation intensity of the system. Dex-MA was polymerized with the three solvent systems of comparable viscosity. The results are displayed in Table 2. The gel times and percent conversions were similar for sorbitol/water and glucose/water, ranging from 7 to 7.5 min and 37% to 43% conversion respectively. However the gel time and conversion for glycerol/water was marginally faster and higher. The polymerization of Dex-MA was also conducted in distilled water. After continuous sonication for 60 min no hydrogel resulted. With prolonged sonication we expected that some hydrogel would result even if it were much slower. This is based on the fact that HEMA (Table 1.) could be polymerized by ultrasound in water alone. However this was not the case for Dex-MA. It is likely that by increasing the additive concentration in water that the depolymerization effect by ultrasound is significantly reduced. It has been reported that the better the solvent is, the more extended a polymer chain will be, and hence the more likely it is to shear [42]. Therefore our viscosity

(a) Dex-MA

40

(b) HEMA

20

0 0

5

10

15

20

25

30

35

40

Time (min) Fig. 4. Conversion profile for the ultrasonic gelling polymerization with time of (a) Dex-MA, T = 37 °C, [glycerol] = 70% w/w, [monomer] = 5% w/w, ultrasonic power = 56 W cm2, (b) HEMA, T = 37 °C, [glycerol] = 70% w/w, [monomer] = 10% w/w, ultrasonic power = 56 Wcm2.

Table 2 Preparation of Dex-MA hydrogels by ultrasound in various solvents, monomer mass = 1.5 g, T = 37 °C, ultrasonic power = 42 W cm2. Solvent

Solvent viscosity (centistokes)

Sonication time (min)

% Conversion

13.5 g 13.5 g 13.5 g 13.5 g

9.2 9.6 9.0 0.7

6.5 min (gel time) 7 min (gel time) 7 min (gel time) 60 min

55 43 37 0

70% w/w glycerol 55% w/w sorbitol 56% w/w glucose water

330

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332

-1

1730 cm

-1

1635 cm

Absorbance

0 min

4.5 min

5.0 min

15.0 min

thermal initiator

2000

1850

1700

1550

1400

1250

1100

950

800

650

-1

Wavelength (cm ) Fig. 5. Infrared Spectra of extracted and dried Dex-MA hydrogels with increasing polymerization time, polymerization conditions; T = 37 °C, [glycerol] = 70% w/w, [monomer] = 5% w/w, ultrasonic power = 56 W cm82, compared to a thermal free radical initiation system.

tem. The gelation reaction is very fast for Dex-MA. From the onset of hydrogel formation it takes less than 30 s to achieve 75% mass conversion. This has been attributed to the multifunctional macromonomer. On average each chain contains 13 pendant methacrylate polymerizable groups. Therefore to achieve 75% mass conversion only 6% methacrylate conversion is necessary. The hydrogel will however continue to cure after this point and increase in strength. The conversion of HEMA occurs much less rapidly than the Dex-MA as the HEMA predominantly has a polymerization functionality of two. Interestingly the conversion increase is linear which may indicate a zero order reaction with re-

spect to monomer concentration. We would expect that as the monomer supply is depleted the rate of conversion would decrease as the ultrasound is providing a constant supply of short lived radicals. If this decrease in rate is being offset then the population of propagating species may be increasing. In addition it was observed in a separate experiment that Dex-MA polymerization continued after the sonication ceased. For example, if the Dex-MA reaction was not terminated when the ultrasound ceased, the conversion would continue by up to 20% w/w (from approximately 60% conversion) over a 30 min period. We cannot attribute this to the Trommsdorff effect [43] as it also can occur if sonication has been

0.9 0.8

45

0.7

40

Qm, Swelling Ratio

A1635 / A1730

0.6 0.5 0.4 0.3 0.2

35 30 25 20 15 10

0.1

5

0 0

4.5

5

15

Sonication Time (min)

thermal radical initiator 1

Fig. 6. Relative intensity of infrared adsorption bands at 1635 cm (conjugated C@C stretch) and 1730 cm1 (C@O stretch) of Dex-MA hydrogels with increasing polymerization time, compared to a thermal free radical initiation system, results derived from Fig. 5.

0 4.5

5.0

15.0

Sonication Time (min)

thermal radical initiator

Fig. 7. Swelling ratio, Qm, of extracted Dex-MA hydrogels with increasing polymerization time, compared to a thermal free radical initiation system.

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332

331

Fig. 8. SEM images of extracted and freeze dried Dex-MA hydrogels with increasing polymerization time, compared to a thermal free radical initiation system. Imaging was performed on a Phillips XL30 Field Emission Scanning Electron Microscope (FESEM) at 1 kV with a Polaron LT7400 cryo-preparation system.

ceased immediately prior to gelation (<10% conversion). This would indicate that in addition to the direct sonochemical polymerization, sonication of glycerol produces an initiating species which subsequently decays. Investigation of this is currently in progress. During the Dex-MA polymerization, hydrogel samples were intermittently isolated and leached of glycerol and non-crosslinked polymer. The samples were then analyzed for methacrylate conversion (FTIR), swelling ratio, gel conversion (as previously detailed in Fig. 4.) and microstructure (SEM). The analysis was compared to a hydrogel prepared using a thermal free radical initiator. Evidence of hydrogel conversion being the result of covalent bonding between macromonomer chains can be obtained by observing a depletion of vinyl groups by FTIR spectroscopy. Fig. 5. shows the spectra of the extracted and dried and Dex-MA hydrogel at various stages of polymerization. The band at 1635 cm1 is characteristic for a conjugated C@C stretch, which decreases as the polymerization time increases. The decrease is compared to an internal standard band at 1730 cm1, which is characteristic of C@O stretching. This relative vinyl concentration decrease can be clearly seen in a plot of A1635 cm1/A1730 cm1 with increasing polymerization time (Fig. 6.). A large decrease in the concentration of vinyl groups can be seen after 4.5 min polymerization. The decrease from 5 to 15 min is more gradual. This corresponds well with the small increase in conversion after an initial rapid increase between 4 and 5 min (Fig. 4). These results also compare well with an increase in the swelling ratio of the hydrogel samples (Fig. 7). As the conversion increases so does the interchain crosslinking reactions which produces a polymer network with greater crosslink density and therefore less swelling potential. Once again after 5 min of sonication only a small decrease in the swelling ratio can be observed. Differences however do exist when comparing the swelling ratio of the thermally initiated hydrogel with the relative decrease in vinyl groups. This relative concentration of vinyl groups is similar to the ultrasonically produced hydrogels but its’ swelling ratio is much higher. This can be explained by differences between the microstructure of the thermally initiated hydrogel and the ultrasonic hydrogel. Fig. 8 shows that as the

ultrasonic polymerization time increases, the initial 6 lm diameter pores at 4.5 min polymerization rapidly begin to fill in with a fine secondary structure after 5 min. This secondary structure reveals 3 lm diameter pores which subsequently thickens to give a more uniform structured hydrogel which can be observed after 15 min polymerization. The initial uniformity is likely to be influenced by the viscous solution stabilized micro bubbles generated by ultrasound. Such uniformity can not be observed for the thermally initiated hydrogel. Its loose and irregular structure explains its higher swelling ratio at a similar conversion of vinyl groups. 4. Conclusion Ultrasound was found to be an effective method for the polymerization of water soluble vinyl monomers and for the production of hydrogels. This occurs rapidly in the absence of a chemical initiator. The water soluble additive was essential for the preparation of hydrogel. We propose that its effect is attributed to the enhanced viscosity increasing the free radical production as well as reducing the polymer solubility and hence retarding depolymerization. The most effective additive for the preparation of hydrogels was glycerol. Such a technique may find application in the field of biomaterial synthesis to avoid problems associated with cytotoxic initiators. In combination with techniques such as high intensity focused ultrasound, the polymerization method described here may also allow the formation of hydrogels in vivo. References [1] Y.P. Zheng, R.G. Maev, I.Y. Solodov, Can. J. Phys. 77 (1999) 927. [2] C.J. Harvey, J.M. Pilcher, R.J. Eckersley, M.J.K. Blomley, D.O. Cosgrove, Clin. Radiol. 57 (2002) 157. [3] T.J. Mason, Prog. Biophys. Mol. Biol. 93 (2007) 166. [4] J.M.J. Paulusse, R.P. Sijbesma, J. Polym. Sci. Polym. Chem. 44 (2006) 5445. [5] O. Lindstrom, O. Lamm, J. Phys. Colloid Chem. 55 (1951) 1139. [6] P. Grabar, R.O. Prudhomme, Compt. Rend. 226 (1948) 1821. [7] K.S. Suslick, D.A. Hammerton, R.E. Cline Jr., J. Am. Chem. Soc. 108 (1986) 5641. [8] P. Kruus, J.A.G. Lawrie, M.L. O’Neill, Ultrasonics 26 (1988) 352. [9] Y. Kojima, S. Koda, H. Nomura, Ultrason. Sonochem. 8 (2001) 75. [10] Y. Kojima, S. Koda, H. Nomura, S. Kawaguchi, Ultrason. Sonochem. 8 (2001) 81. [11] D. Kobayashi, H. Matsumoto, C. Kuroda, Chem. Eng. J. 135 (2008) 43. [12] V.A. Henglein, Macromol. Chem. 14 (1954) 15. [13] N. Xiuyuan, H. Yuefang, L. Bailin, X. Xi, Eur. Polym. J. 37 (2001) 201.

332 [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

P. Cass et al. / Ultrasonics Sonochemistry 17 (2010) 326–332 S. Biggs, F. Grieser, Macromolecules 28 (1995) 4877. S.K. Ooi, S. Biggs, Ultrason. Sonochem. 7 (2000) 125. J. Zhang, Y. Cao, Y. He, J. Appl. Polym. Sci. 94 (2004) 763. N. Yin, K. Chen, W. Kang, Ultrason. Sonochem. 13 (2006) 345. Y. He, Y. Cao, Y. Fan, J. Appl. Polym. Sci. 107 (2008) 2022. H.C.J. Chou, J.O. Stoffer, J. Appl. Polym. Chem. 72 (1999) 797. G.J. Price, A.A. Clifton, Polymer 37 (1996) 3971. G.J. Price, L. Garland, J. Comina, M. Davis, D.J. Snell, P.J. West, Res. Chem. Intermed. 30 (2004) 807. Y. He, Y. Cao, Y. Liu, J. Polym. Sci. Polym. Phys. 43 (2005) 2617. M.A. Bradley, S.W. Prescott, H.A.S. Schoonbrood, K. Landfester, F. Grieser, Macromolecules 38 (2005) 6346. Y. Liao, Q. Wang, H. Xia, X. Xu, S.M. Baxter, R.V. Slone, S. Wu, G. Swift, D.G. Westmoreland, J. Polym. Sci. Polym. Chem. 39 (2001) 3356. H. Xia, Q. Wang, G. Qiu, Chem. Mater. 15 (2003) 3879. C. Wnag, Q. Wnag, G.H. Qiu, Acta Polym. Sin. 9 (2006) 1038. G. Cooper, F. Grieser, S. Biggs, J. Colloid. Interface Sci. 184 (1996) 52. R. Wang, H.M. Cheung, J. Supercrit. Fluids 33 (2005) 269. M.W.A. Kuijpers, D. van Eck, M.F. Kemmere, J.T.F. Keurentjes, Science 298 (2002) 1969.

[30] M. Choi, K. Lee, J. Lee, H. Kim, Macromol. Symp. 249 (2007) 350. [31] M. Bartsch, G. Schmidt-Naake, Macromol. Chem. Phys. 207 (2006) 209. [32] M. Degirmenci, H. Catalgil-Giz, Y. Yagci, J. Polym. Sci. Polym. Chem. 42 (2004) 534. [33] A.H. Lebovitz, M.K. Gray, A.C. Chen, J.M. Torkelson, Polymer 44 (2003) 2823. [34] M.V. Encina, E. Lissi, M. Sarasusa, L. Gargallo, D. Radic, J. Polym. Sci. Polym. Lett. 18 (1980) 757. [35] P. Cass, T. Hughes, W. Knower, Aust. Patent Appl. 2008905787 (2008). [36] F. Thermes, J. Grove, A. Rozier, B. Plazonnet, A. Constancis, C. Bunel, J.-P. Vairon, Pharm. Res. 9 (1992) 1563. [37] J.L. Luche, Ultrason. Sonochem. 3 (1996) S215. [38] D.J.T. Hill, J.H. O’Donnell, P.J. Pomery, G. Saadat, Radiat. Phys. Chem. 48 (1996) 605. [39] M. Feng, M.V. Sefton, J. Biomater. Sci. Polym. Edn. 11 (2000) 537. [40] M. Ashokkumar, F. Grieser, Rev. Chem. Eng. 15 (1999) 41. [41] C.M. Krishna, T. Kondo, P. Riesz, J. Phys. Chem. 93 (1989) 5166. [42] M.S. Doulah, J. Appl. Polym. Sci. 22 (1978) 1735. [43] E. Trommsdorff, H. Kohle, P. Lagally, Macromol. Chem. 1 (1948) 169.