Comparison of the Effectiveness of Four Different
`Crosslinking Agents with Hyaluronic Acid Hydrogel
`Films for Tissue-Culture Applications
`
`M. N. Collins, C. Birkinshaw
`
`Department of Materials Science and Technology, University of Limerick, Limerick, Ireland
`
`Received 23 October 2006; accepted 13 December 2006
`DOI 10.1002/app.25993
`Published online 5 March 2007 in Wiley InterScience (www.interscience.wiley.com).
`
`ABSTRACT: The effectiveness of four different reagents,
`glutaraldehyde (GTA), 1-ethyl-3-(3-dimethylaminopropyl)
`carbodiimide (EDC), poly(ethyelene glycol) diglycidyl
`ether (EX 810), and divinyl sulfone (DVS) as crosslinkers
`for cast hyaluronic acid (HA) films has been evaluated.
`Films were prepared by casting from solution and exposed
`to solutions of the crosslinkers in acetone–water solution.
`Swelling in water and in phosphate buffered saline (PBS)
`was then used to assess the effectiveness of the cross-
`linkers. GTA-crosslinked films were found to be of low
`stability compared with those treated with EDC, EX 810,
`and DVS. Results suggest that instability in GTA-cross-
`linked materials arises in part from residual acid catalyst.
`The effects of polymer molecular weight are not uniform.
`
`With GTA-crosslinked film produced from higher molecu-
`lar weight HA swells more, and this is attributed to
`reduced diffusion of the crosslinker, but with EDC, the
`opposite effect is observed, implying some additional mo-
`lecular weight dependent mechanism. Differential scan-
`ning calorimetry and dynamic mechanical thermal analysis
`results suggest that there are no significant structural dif-
`ference between the gels for each crosslinker system and
`only the crosslink density and moisture content alters the
`transitions. Ó 2007 Wiley Periodicals, Inc. J Appl Polym Sci 104:
`3183–3191, 2007
`
`Key words: hydrogel; biological applications of polymers;
`biomaterials; mechanical properties
`
`INTRODUCTION
`
`Hyaluronan, also known as hyaluronic acid (HA), an
`abundant nonsulfated glycosaminoglycan (GAG)
`component of synovial fluid and extracellular matri-
`ces,1,2 is an attractive building block for new biocom-
`patible and biodegradable polymers with possible
`applications in drug delivery,3 tissue engineering,4–6
`and visco supplementation.7 As a polysaccharide of
`the extra cellular matrix (ECM), it plays a multitask
`role, having many structural, rheological, physiologi-
`cal, and biological functions in the body. It is a linear
`and anionic polymer consisting of two modified sug-
`ars, glucuronic acid and N-acetylglucosamine, with
`the molecular structure:-[D-glucuronic acid
`(1-b-3) N-acetyl-D-glucosamine (1-b-4)]n.
`Hyaluronan is synthesized by many types of cells
`in the body and extruded into the extracellular space
`where it interacts with the other constituents of the
`ECM to create the supportive and protective struc-
`ture around the cells. It is present as a constituent in
`all body fluids and tissues and is found in higher
`concentrations in the vitreous humor of the eye and
`
`Correspondence to: C. Birkinshaw (colin.birkinshaw@ul.ie).
`
`Journal of Applied Polymer Science, Vol. 104, 3183–3191 (2007)
`VVC 2007 Wiley Periodicals, Inc.
`
`the synovial fluid in the joints. Commercial HA is
`usually obtained from rooster comb, although full
`details of its preparation are not always provided.
`The biomedical application of HA is hindered by
`its short residence time and lack of mechanical integ-
`rity in an aqueous environment and these drawbacks
`must be addressed in order to realize its potential.
`In this article, we compare the effects of four differ-
`ent chemical crosslinkers to protract the material’s
`degradation and dissolution and thereby improve
`mechanical stability. Crosslinking is the most com-
`mon modification of hyaluronan to form a hydrogel
`and a number of mechanisms have been reported in
`the literature.8–11 The functional groups, which are
`mainly responsible for crosslinking of HA molecules
`are the hydroxyl and carboxyl groups. Hydroxyl
`groups may be crosslinked via an ether linkage and
`carboxyl groups via an ester linkage. If desired, the
`HA may be chemically modified prior to crosslink-
`ing to form other chemically reactive groups. Thus,
`for example, HA may be treated with acid or base
`such that it will undergo at least partial deacetaliza-
`tion, resulting in the presence of free amino groups.
`It is said that amino groups may be crosslinked via
`an amide (C(O)NH); imino (N¼¼CH) or
`secondary amine (NHCH) bond. An imino
`linkage can be converted into an amine linkage in
`the presence of a reducing agent.
`
`ALL 2070
`PROLLENIUM V. ALLERGAN
`IPR2019-01505 et al.
`
`

`

`3184
`
`COLLINS AND BIRKINSHAW
`
`formation of
`the crosslinking occurs through the initial
`(2)
`Scheme 1 With water soluble carbodimides (WSC)
`O-acylisourea (3) on the polysaccharide, through reaction with neighboring carboxyl groups (1) an anhydride (4) is
`formed, and this anhydride then reacts with nearby hydroxyls to give both inter- and intramolecular crosslinks.
`
`Reaction has been accomplished under acidic, neu-
`tral,
`and
`alkaline
`conditions
`using
`carbodi-
`mides,10,12–14 hydrazides,14,15 aldehydes,8 sulfides,16
`and polyfunctional epoxides.9,17–20 Autocrosslink-
`ing21,22 and photocrosslinking23–25 have also been
`reported. With carbodimides, the crosslinking occurs
`through the initial formation of anhydride on the
`polysaccharide, through reaction with neighboring
`carboxyl groups, and this anhydride then reacts with
`nearby hydroxyls to give both inter- and intramolec-
`ular crosslinks. It was postulated that crosslinking
`took place via ester groups. Scheme 1 gives some
`details of these reactions. A reported10 modification
`of
`this reaction introduces L-lysine methyl ester,
`which offers the opportunity to form higher stability
`amide crosslinks. Dialdehydes are believed to cross-
`link through formation of acetal or hemiacetal
`groups on neighboring chains, with kinetic and
`spectroscopic evidence indicating a prevalence of
`the hemiacetal. Glutaraldehyde (GTA) is believed
`to form either a hemiacetal or an ether link with
`HA under acidic conditions26 as shown in Scheme 2.
`With divinyl
`sulfone (DVS),
`the crosslinking
`occurs via the hydroxyl groups forming an ether
`bond as shown in Scheme 3. The epoxy group of
`EX 810 is known to react with COOH and the
`OH functional groups, therefore forming ester
`and ether bonds, respectively, and is shown in
`Scheme 4.
`Reactions may be carried out using heterogeneous
`or homogeneous methods. Heterogeneous reactions
`are carried out on solid HA, cast in the form of films
`or membranes, in which case diffusion rates maybe
`
`at least as important as chemical kinetics, whereas
`homogeneous reactions are carried out using HA
`solutions. The former method has the advantage of
`allowing shaping of a product before crosslinking,
`whereas the latter method offers the advantage of
`better control of the chemistry with greater product
`homogeneity.
`the repeatability of
`Experience has shown that
`reported experimental procedures and their out-
`comes is low, perhaps because of differences in the
`sources and preparation procedures of the HA and
`the HA solutions. Therefore, an important considera-
`tion in the work reported here was to compare,
`under a set of standard conditions, the performance
`of four important crosslinkers using well character-
`ized HA from identified sources.
`The first part of the study established a method
`for the controlled dissolution of HA and film casting,
`and then utilized the crosslinking reagents to give
`materials with increased stability with time. Control
`of dissolution procedure is considered to be impor-
`tant because dissolution and solution degradation
`may be concurrent processes. Complete dissolution
`is also important to maximize intermolecular cross-
`linking and reduce wasteful intramolecular reactions.
`Previously, crosslinking HA films by immersion in a
`crosslinking mixture has only been reported for
`GTA8 and carbodiimide.10
`The relationship between crosslink density and
`solvent swelling is described by the well-known
`Flory–Rhener equation, which demonstrates that
`increasing crosslinker effectiveness will be shown
`by a reduced volumetric swelling. For this work,
`
`Scheme 2 The OH (1) group on the hyaluronic acid reacts under acidic conditions with glutaraldehyde (2) to give hemi-
`acetal or ether crosslinks (3).
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`COMPARISON OF THE EFFECTIVENESS OF FOUR REAGENTS
`
`3185
`
`Scheme 3 The OH (1) group on the hyaluronic acid
`reacts under alkaline conditions with divinyl sulfone (2) to
`give sulfonyl bisethyl crosslinks (3).
`
`the swelling ratio (SR) was calculated via the
`equation:
`
`Swelling ratio ¼ Ws
`Wd
`
`where Ws is the weight of the sample at equilibrium
`at each temperature and Wd is the weight of the
`dried sample.
`Huglin et al. defines the equilibrium water content
`according to the equation:
`
`EWC ¼ ðSR 1Þ
`
`SR
`
`bound
`‘‘secondary
`or
`2. Intermediate water
`water:’’ other water molecules that interact with
`polymer molecules are referred to as intermedi-
`ate water. This kind of freezing water has a
`melting point <08C.
`3. Free water: water molecules that do not take
`part in hydrogen bonding with polymer mole-
`cules are called free water because of
`their
`greater degree of mobility in comparison with
`other water molecules. Free water is freezing
`water showing a melting point at 08C.28
`
`An initial set of survey experiments compared the
`effects of the different crosslinkers and reaction envi-
`ronments, and the results obtained were then used
`in designing the experiments to determine the reac-
`tion conditions for further experiments.
`
`EXPERIMENTAL
`
`Materials
`
`The sodium salt of HA with an average molecular weight
`of 2.14  106 was supplied by Clear Solutions (New York,
`NY) as dry powder. This material is prepared in high
`yield from streptococcus bacteria by fermenting the bac-
`teria under anaerobic conditions in CO2 enriched growth
`medium.29 HA powders of average molecular weight
`1.2  106, 8.5  105, and 1.4  105 were purchased from
`Bioiberica (Barcelona, Spain). This material is obtained
`from rooster comb. HA has been mainly extracted from
`rooster combs for many years and many papers have
`been published on clinical application of HA from this
`source.
`1-Ethyl-3-(3-dimethylaminopropyl)
`carbodii-
`mide (EDC), GTA, poly(ethylene glycol) diglycidyl
`ether, and DVS were purchased from Lancaster (UK).
`
`Preparation of HA-crosslinked films
`
`Solutions were prepared by sieving HA particles
`into double-distilled water to expose the maximum
`
`Therefore, the water content can be expressed as
`follows27;
`ð%Þ ¼ Wt ðWf
`
`Wb
`
`þ Wfb
`
`Þ ¼ Wt ðQendo
`
`Þ  100
`
`=Qf
`
`where Wb is the amount of bound water (%); Wf and
`Wfb are the amounts of free and freezing bound
`water, respectively, and Wt is the EWC (%).
`The three types of water found in gels are defined
`as follows:
`
`1. Bound water: this term refers to the water mole-
`cules that are bound to polymer molecules
`through hydrogen bounds. This kind of water
`shows no endothermic peak in the temperature
`range: 70 to 08C.
`
`Scheme 4 The OH (1) group on the hyaluronic acid reacts with the epoxy group of the poly(ethyelene glycol) diglycidyl
`ether (2) to give ether crosslinks (3).
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`3186
`
`COLLINS AND BIRKINSHAW
`
`area for solvent interaction. This was followed by
`agitation, to minimize shear stress, in a shaking bath
`at 258C for up to 102 h and was found to give repro-
`ducible solutions of uniform viscosity. The viscosity
`time profiles of the solutions were then obtained
`using an Ubbelohde viscometer. Samples were fully
`dissolved after 24 h and the molecular properties of
`all the solutions were evaluated using size exclusion
`chromatography (SEC). Films were then prepared by
`casting a 1 wt % aqueous solution of each HA type
`onto a clean petri dish, followed by drying at 258C
`under vacuum for 120 h. The volume of solution used
`determined the thickness of the resulting film and this
`was adjusted to give a thickness of  0.2 mm.
`For the initial survey experiments, 10  10  0.2
`¼ 7.6  105 Da) film
`mm3 samples of cast HA (MW
`were weighed and placed in 10 mL of acetone–water
`solution (80:20 by volume) containing 0.01M HCl
`and varying mole ratios of 1-EDC, GTA, ethylene
`glycol diglycidyl
`ether
`(Denacol EX-810),
`and
`divinylsulphone (DVS). The acetone prevents the
`dissolution of the HA film into the reaction solution,
`and all the reaction vessels were sealed to prevent
`evaporation of the acetone. The acetone concentra-
`tion of 80 vol % was selected after experimentation
`had shown that lower concentrations resulted in ex-
`cessive water swelling of the cast HA film. Cross-
`linker molarity and mole ratio of crosslinker to HA
`were based on previously published data.8 A high
`crosslinker concentration is used to drive the diffu-
`sion process. The crosslinking reaction was allowed
`to proceed at room temperature for 24 h unless oth-
`erwise specified. With EDC and GTA, crosslinking is
`favored by acidic conditions and 0.01M HCl was
`used as pH adjuster and catalyst,8 while DVS
`requires alkaline conditions and in this case, 0.01M
`NaOH was used to adjust pH.
`
`Characterization of films
`
`On the basis of the initial observations, further series
`of crosslinked films were cast, using film thicknesses
`up to 0.2 mm, crosslinking times between 24 and 72 h,
`and crosslinking temperatures of 4 and 208C. The
`crosslinker:HA mole ratio used was 2 : 1 to 4 : 1 and
`the medium was 80 : 20 acetone:water. These films
`were washed in distilled water for 1 h and then
`dried overnight in a vacuum oven at room tempera-
`ture. They were then swollen in distilled water and
`in phosphate buffered saline (PBS) and the water
`content of the gel and the SR were measured.
`DSC analysis was carried out on a TA Instruments
`DSC 10 differential scanning calorimeter. The ther-
`mal analysis profiles were of dried hydrogel samples
`and swollen samples. The temperature was incre-
`ased from room temperature to 3008C at a rate of
`108C/min under a nitrogen atmosphere. Differential
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`scanning calorimetry (DSC) (TA Instruments 10) was
`also employed to examine the state of water in the
`swollen hydrogels with different water contents.
`Samples sealed in aluminum pans were cooled to
`208C and then heated to 208C at a heating rate of
`58C/min under 60 cc/min of nitrogen gas flow.
`Using peak areas, normalized for sample mass, the
`endotherm associated with water loss was obtained
`and compared with the theoretical value for water.
`The fraction of free water in the total water was then
`calculated using eq. (3) as the ratio of the endother-
`mic peak area for water-swollen hydrogels to the
`melting endothermic heat of fusion (334 J/g) for
`pure water. Bound water because of hydrogen bond-
`ing was expressed as the difference between the total
`water and the free water.
`The dynamic mechanical thermal analysis (DMTA)
`of the materials in the form of films (0.7-mm thick)
`was carried out with a Polymer Laboratories DMTA
`MK-1 apparatus, operating in the parallel plate
`mode. The scans were performed on samples main-
`tained under room conditions, at a frequency of 1 Hz,
`temperature range of 30 to 1008C, and a heating
`1.
`rate of 48C min
`
`RESULTS AND DISCUSSION
`
`Preparation of HA films
`
`The relative flow times for the HA solutions (Fig. 1)
`are consistent with the molecular weights indicated
`by the suppliers. All the viscosity-time curves have a
`similar shape with the samples becoming fully sol-
`vated between 24 and 40 h. After this time, the solu-
`tion viscosities are seen to fall and this is taken to
`indicate the onset of polymer degradation through
`hydrolysis. As a result of
`these observations, all
`films used for crosslinking were cast after 24 h of
`dissolution and thus in the fully solvated and chain-
`disentangled state. The molecular properties of each
`
`Figure 1 Dissolution time profile of 1 wt % hyaluronic
`acid (HA) solutions.
`
`

`

`COMPARISON OF THE EFFECTIVENESS OF FOUR REAGENTS
`
`3187
`
`TABLE I
`Molecular Properties of Hyaluronic Acid After 24 h Dissolution, Measurements were in PBS
`
`Sample
`2.25  106
`1.20  106
`0.85  106
`0.14  106
`
`Mw/Mn
`
`1.08
`2.14
`3.27
`1.38
`
`Intrinsic
`viscosity
`(dL/g)
`
`24.61
`20.13
`14.80
`3.26
`
`Radius of
`hydration
`(nm)
`
`94.84
`66.54
`51.68
`18.31
`
`Radius of
`gyration
`(nm)
`
`162.61
`114.33
`97.01
`26.94
`
`Mark–Houwink
`(a)
`
`0.61
`0.91
`0.95
`1.09
`
`Mark–Houwink
`(log K)
`2.49
`4.23
`4.44
`9.884
`
`sample is summarized in Table I and the values
`given were calculated using the Viscotek Omnisec
`software, version 4.2.
`the survey
`Table II summarizes the results of
`experiments and details the swelling ratios of the
`films after they had been submerged in the solution
`of crosslinker in 80% acetone and 20% water for 24 h.
`The results show that the molarity of the crosslinker
`solution has only small effect indicating that film
`surface area is the critical issue. With both DVS and
`GTA, the need for acidification to optimize crosslink-
`ing is apparent.
`
`Swelling studies
`
`In Figures 2 and 3, the swelling in water and PBS of
`films, crosslinked under optimum conditions,
`is
`compared and in both media all three crosslinkers
`give an initial volume swelling of between 2.0 and
`2.5, indicating similar crosslink densities. Consider-
`ing the behavior in distilled water, both the DVS
`
`and EDC-crosslinked materials shrink with time and
`this is taken to indicate continuing crosslinking
`action, suggesting that the crosslinker diffusion into
`the film is faster than the crosslinker reaction with
`the constituent polymer.
`GTA-crosslinked materials are clearly very sensi-
`tive to the nature of the swelling medium and it is
`thought that the lower swelling and greater stability
`in buffer arises from neutralization of residual acid
`in the film. With these materials, progressive swel-
`ling occurs in water and this is thought to arise from
`network scission. On the basis of Tomihata and Ika-
`da’s proposal that
`the crosslinks in this case are
`hemiacetals,8 as shown in Scheme 2, it is reasonable
`to presume that
`these are undergoing hydrolyitic
`scission with time, catalyzed by residual acidity, and
`then the gel SR increases as the crosslink density
`falls. Taken overall, these results suggest that the
`crosslinking reagents remain reactive in the gel and
`that swelling with buffer has a stabilizing effect
`through neutralization.
`
`Crosslinking Conditions for HA Films ( 0.1-mm Thick) Mw 5 8.5 3 105 Da
`
`TABLE II
`
`Crosslinker
`
`Mole Ratio
`crosslinker/polymer
`
`[crosslinker] M
`
`pH
`
`Equilibrium swelling
`ratio after 24 h in 80%
`acetone/20% water mixtures
`
`DVS
`DVS
`GTA
`GTA
`GTA
`GTA
`GTA
`GTA
`EDC
`EDC
`EDC
`EDC
`EDC
`EDC
`EDC
`EX 810
`EX 810
`EX 810
`EX 810
`EX 810
`EX 810
`
`02:01
`04:01
`02:01
`02:01
`02:01
`02:01
`02:01
`04:01
`01:01
`02:01
`02:01
`03:01
`04:01
`04:01
`04:01
`04:01
`04:01
`04:01
`04:01
`02:01
`02:01
`
`0.26
`0.57
`0.19
`0.19
`0.21
`0.22
`0.22
`0.40
`0.14
`0.22
`0.30
`0.26
`0.46
`0.46
`0.58
`0.28
`0.32
`0.34
`0.36
`0.17
`0.22
`
`12
`12
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`7
`7
`
`1.53
`1.37
`1.28
`1.28
`1.27
`1.26
`1.26
`1.20
`1.67
`1.47
`1.33
`1.57
`1.28
`1.28
`1.31
`1.17
`1.25
`1.21
`1.21
`1.33
`1.20
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`3188
`
`COLLINS AND BIRKINSHAW
`
`Figure 2 Swelling ratios in distilled water at 228C (hyal-
`uronic acid (Mw 0.14  106 Da) to crosslinker is 1 : 5 for all
`
`samples).
`
`Figure 4 Crosslinked hyaluronic acid films (Mw 2.25
` 106 Da) swollen in distilled water pH 6.2 at 258C, mono-
`mer crosslinker mole ratio 1 : 4.
`
`Degradation of films
`
`Figures 4–6 show the results of longer time scale
`swelling experiments with HA of different
`initial
`molecular weights. HA molecular weight can be
`expected to influence gel character in two ways. Dif-
`fusion of reactants can be expected to occur faster in
`lower molecular weight polymer
`because
`of
`increased chain end free volume, thereby increasing
`crosslink density, but counteracting this will be the
`more rapid macronetwork formation with higher
`molecular weight chains. With the lower molecular
`weight HA, the gels reached their maximum SR at
`24 h and after this time start to degrade, presumably
`through both crosslink hydrolysis and main-chain
`scission. At 5 days, water contents reach 98% and
`gels started to break up.
`The two crosslinkers EDC and GTA were chosen
`for further swelling experiments at 378C on thicker
`films (0.2 mm). Figures 7 and 8 show the swelling
`characteristics of films produced by EDC and GTA
`crosslinking, respectively, and the time scale of over
`3 weeks demonstrates their long term stability. As
`expected, the SR decreases slightly with increased
`
`reflecting the consequent
`crosslinker mole ratio,
`increase in crosslink density; however, this effect is
`small confirming that the surface area is the domi-
`nating factor. When compared with the GTA-cross-
`linked materials, EDC-crosslinked films seem to be
`more stable and the degradation appears to occur
`via the main chain as opposed to crosslink scission.
`However, the GTA-crosslinked films show a large
`initial increase in volume before a reduction in swel-
`ling occurs and it is possible that this initial swelling
`is due to crosslink scission and the consecutive
`reduction in swelling can be attributed to main-chain
`degradation. This behavior is consistent with the
`presumed mode of crosslinking of these reagents, as
`shown in Schemes 1 and 2, in that hemiacetal cross-
`links, obtained by reaction with GTA, can be
`expected to be of low stability compared with the
`anhydride derived crosslinks arising with EDC. It is
`likely that the small increase in SR apparent in Fig-
`ure 4 with the GTA-crosslinked material reflects the
`early stages of network breakdown. With both GTA
`and EDC crosslinker, molarity has the expected
`small effect, again indicating that surface area is the
`dominating factor.
`
`Figure 3 Swelling ratios in PBS at 228C (hyaluronic acid
`(Mw 0.14  106 Da) to crosslinker is 1 : 5 for all samples).
`
`Figure 5 Crosslinked hyaluronic acid films (Mw 1.20
` 106 Da) swollen in distilled water pH 6.2 at 258C, mono-
`mer crosslinker mole ratio 1 : 4.
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`COMPARISON OF THE EFFECTIVENESS OF FOUR REAGENTS
`
`3189
`
`Figure 6 Crosslinked hyaluronic acid films (Mw 0.14
` 106 Da) swollen in distilled water pH 6.2 at 258C, mono-
`mer crosslinker mole ratio 1 : 4.
`
`Figure 8 Swelling of HA films (8.5  105 Da) in distilled
`water at 378C when crosslinked with EDC.
`
`Thermal and mechanical characterization of films
`
`The thermal behavior of the HA-hydrogel film was
`investigated by DSC analysis (Figs. 9 and 10). All the
`samples showed the presence of a broad endother-
`mic peak below 858C (Table III) and this is thought
`to be associated with the loss of moisture remaining
`after the drying procedure. In addition, significant
`sharp exothermic peaks were observed for each sam-
`ple, which probably represent decomposition. The
`DSC profiles show that after HA was crosslinked,
`the endothermic and exothermic peaks shifted, indi-
`cating an altered structure. The relatively low endo-
`therm peak temperature obtained with the EDC-
`crosslinked material may reflect lower water reten-
`tion through the drying period.
`Figure 10 shows the DSC thermograms of the fully
`swollen hydrogels. The endothermic peak maximum
`of the swollen hydrogel ranges between 1.17 and
`2.58C. The fraction of free water in total water is
`approximately calculated as the ratio of the endo-
`thermic peak area for water-swollen hydrogel
`to
`melting endothermic heat of fusion (334 J/g) for
`
`pure water as described in Mansor and Malcolm’s
`work.30 The bound water is expressed as the diff-
`erence between the total and free waters. The EWC
`value and the free and bound water contents were
`calculated and are listed in Table IV. The EWC was
`calculated at pH 7 in water at 258C. Clearly, samples
`with the highest crosslink density have the lowest
`EWC and free water content indicating a more com-
`pact structure.
`The dynamic mechanical analysis results shown in
`Figure 11 indicate that there is little change in the
`storage modulus between the initial temperature and
`158C. However, between 20 and 408C, the storage
`modulus drops by over two orders of magnitude.
`This abrupt change in physical properties is associ-
`ated with the onset of short-range molecular motions
`at the glass transition. Because of the crosslinking,
`the modulus beyond the glass transition does not
`decline to near zero and the materials still exhibit
`load bearing characteristics even 50–758C
`useful
`above Tg. The logarithmic plot shows that, for HA
`crosslinked systems, a direct relationship is found
`between post-Tg modulus
`and the degree of
`
`Figure 7 Swelling of HA films (8.5  105 Da) in distilled
`water at 378C when crosslinked with GTA.
`
`Figure 9 The thermal behavior of HA films crosslinked
`and uncrosslinked prior to swelling.
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`3190
`
`COLLINS AND BIRKINSHAW
`
`TABLE IV
`Water Content of the Gels Used in the Degradation
`Study (Fig. 6)
`
`Sample
`
`EWC (%)
`
`Bound water Wb (%)
`
`Wf
`
`þ Wfb (%)
`
`EDC
`DVS
`EX810
`GTA
`
`81.2
`70.2
`82.0
`70.9
`
`23.27
`15.61
`22.04
`20.79
`
`57.93
`54.57
`60.00
`50.14
`
`All samples were swollen at pH 6.2 in water at 258C.
`
`Figure 11 Storage modulus versus temperature of swol-
`len hyaluronic acid films.
`
`Figure 12 Tan d versus temperature of highly swollen
`hyaluronic acid films. The Tg is similar for all samples.
`
`CONCLUSIONS
`
`When HA is cast in a fully solvated and chain-disen-
`tangled state, films can be crosslinked with GTA,
`1-EDC, poly(ethyelene glycol) diglycidyl ether (EX
`810), and DVS by immersion in solutions of the
`crosslinker in 80% acetone–water mixtures. Swelling
`studies showed that both the molecular weight of
`the films and the crosslinker type influenced the
`crosslink density of the films. Diffusion of the cross-
`linker through the films played a significant role.
`GTA-crosslinked films were found to be of low sta-
`bility compared with those treated with EDC, EX
`810, and DVS. Results suggest
`that
`instability in
`GTA-crosslinked materials arises in part from resid-
`ual acid catalyst. When compared with the GTA-
`
`Figure 10 Crosslink density effect on the endotherm.
`
`crosslinking, i.e., the EDC-crosslinked HA had low-
`est crosslink density and therefore the lowest post Tg
`modulus. Interestingly, the GTA-modified samples
`seem to soften at a lower temperature than the rest
`of the films and this may explain their high swelling
`ratios at room temperature (Fig. 2).
`At low temperatures, leading up to the glass tran-
`sition, tan d is well below 0.1 (Fig. 12). The rapid
`rise in the tan d curve coincides with the rapid
`decline in the storage modulus confirming that this
`is the glass transition. For all the HA-crosslinked
`systems, the tan d values above Tg are relatively low.
`Interestingly, sharp peaks are observed for
`the
`highly swollen samples. The magnitude and broad-
`ness of the peak seem to be related to changes in the
`degree of swelling. GTA-crosslinked materials had a
`broadest peak and a SR of 3.5 while the sharpest
`peak was obtained with the highest SR of 13.3 for
`EDC-crosslinked materials. EX 810 crosslinked mate-
`rials and DVS crosslinked materials had swelling
`ratios of 7.7 and 6.5, respectively.
`For clarity purposes, a linear scale was used for all
`storage modulus data, however there was no occur-
`rence of secondary thermal transitions in the films
`and no inflections or peaks were noticed, confirms
`that the structure of the films is similar.
`
`TABLE III
`DSC Data for Unmodified and Modified Hyaluronic Acid
`Films Prior to Swelling
`
`Sample
`
`HA unmodified
`EDC
`DVS
`EX810
`GTA
`
`Endothermic
`peak (8C)
`
`Degradation temperature
`(8C)
`Second peak
`
`First peak
`
`76.1
`74.9
`83.1
`83.6
`83.1
`
`233.3
`232.2
`231.1
`229.9
`231.0
`
`234.3
`236.8
`236.5
`234.4
`236.1
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`

`

`COMPARISON OF THE EFFECTIVENESS OF FOUR REAGENTS
`
`3191
`
`crosslinked materials, EDC-crosslinked films seem to
`be more stable and the degradation appears to occur
`via the main chain as oppose to crosslink scission.
`The GTA-crosslinked films show a large initial
`increase in volume before a reduction in swelling
`occurs and it is possible that this initial swelling is
`due to crosslink scission and the following reduction
`in swelling can be attributed to main-chain degrada-
`tion. The effects of polymer molecular weight are
`not uniform. The thermal and structural properties
`of the films were further studied by DSC and DMTA
`and it was found that films were structurally similar
`and had almost
`identical
`thermal behavior. The
`main difference in the films appears to be due to
`crosslink density variations.
`
`References
`
`1. Weissmanand, B.; Meyer, K. J Am Chem Soc 1954, 76, 1753.
`2. Laurent, T. C. The Chemistry, Biology and Medical Applica-
`tions of Hyaluronan and its Derivatives; Portland Press:
`London, 1998.
`3. Palumbo, F.S.; Pitarresi, G.; Mandracchia, D.; Tripodoand, G.;
`Giammona, G. Carbohydr Polym 2006, 66, 379.
`4. Ibrahim, S.; Joddar, B.; Crapsand, M.; Ramamurthi, A. Bioma-
`terials 2007, 28, 825.
`5. Ji, Y.; Ghosh, K.; Shu, X. Z.; Li, B.; Sokolov, J. C.; Prestwich, G.
`D.; Clark, R. A. F.; Rafailovich, M. H. Biomaterials 2006, 27, 3782.
`6. Takagi, A.; Yamashita, N.; Yoshioka, T.; Takaishi, Y.; Naka-
`nishi, K.; Takemura, S.; Maeda, A.; Saito, K.; Takakuraand, Y.;
`Hashida, M. J Controlled Release 2006, 115, 134.
`7. Fernandez Lopez, J. C.; Ruano-Ravina, A. Osteoarthritis Carti-
`lage 2006, 14, 1306.
`8. Tomihata, K.; Ikada, Y. J Polym Sci Part A: Polym Chem 1997,
`35, 3553.
`9. Tomihata, K.; Ikada, Y. Biomaterials 1997, 18, 189.
`
`10. Tomihata, K.; Ikada, Y. J Biomed Mater Res 1997, 37, 243.
`11. Prestwich, G. D. The Science of Hyaluronan Today. Available
`at http://www.glycoforum.gr.jp/index.htm 2001.
`12. Hamilton, R.; Fox, E. M.; Acharya, R. A.; Walts, A.
`(to Genzyme Corporation, Boston, MA). U.S. Pat. 4,937,270
`(1990).
`13. Burns, J. W.; Cox, S.; Walts, A. (to Genzyme Corporation,
`Cambridge, MA). U.S. Pat. 5,017,229 (1991).
`14. Prestwich, G. D.; Marecak, D. M.; Marecek, J. F.; Vercruysse,
`K. P.; Ziebell, M. R. J Controlled Release 1998, 53, 93.
`15. Luo, Y.; Kirker, K. R.; Prestwich, G. D. J Controlled Release
`2000, 69, 469.
`16. Balazs, E. A.; Leshchiner, A.; Leshchiner A.; Band, P. (to Bio-
`matrix, Ridgefield, NJ). U.S. Pat. 4,713,448 (1987).
`17. Malson, T.; Lindqvist, B. (to Pharmacia AB, Uppsala, Sweden).
`U.S. Pat. 4,716,154 (1987).
`18. Zhao, X. B.; Fraser, J. E.; Alexander, C.; Lockett, C.; White, B. J.
`J Mater Sci: Mater Med 2002, 13, 11.
`19. Simkovic, I.; Hricovini, M.; Soltes, L.; Mendichi, R.; Cosentino,
`C. Carbohydr Polym 2000, 41, 9.
`20. Sakurai, K.; Ueno Y.; Okuyama, T. (to Seikagaku Kogyo Co.,
`Japan). U.S. Pat. 4,716,224 (1987).
`21. Mensitieri, M.; Ambrosio, L.; Nicolais, L. J Mater Sci: Mater
`Med 1996, 7, 695.
`22. Sikkink, C. J. J. M.; de Man, B.; Bleichrodt, R. P.; van Goor, H.
`J Surg Res 2006, 136, 255.
`23. Matsuda, T.; Moghaddam, M. J.; Sakurai, K. (to Seikagaku
`Kogyo Kabushiki Kaisha, Tokyo, Japan). U.S. Pat. 5,462,976
`(1995).
`24. Hubbell, J. A.; Pathak, C. P.; Sawhney, A. S.; Desai, N. P.;
`Hill-West, J. L. (to Board of Regents, The University of Texas
`System, Austin, TX). U.S. Pat. 5,567,435 (1996).
`25. Yui, N.; Okano T.; Sakurai, Y. J Controlled Release 1993, 26,
`141.
`26. Zhao, X. (to A-Life Limited, GB). U.S. Pat. 6,703,444 (2004).
`27. Huglin, M. B.; Liu, Y.; Velada, J. L. Polymer 1997, 38, 5785.
`28. Shin, M.-S.; Kim, S.; Park, S. J.; Lee, Y. H. J Appl Polym Sci
`2002, 86, 2498.
`29. Bracke, J.; Thacker, K. (to Diagnostic, Inc., Minneapolis, MN).
`U.S. Pat. 4,517,295 (1985).
`30. Mansor, B.; Malcolm, B. H. Polym Int 1994, 33, 273.
`
`Journal of Applied Polymer Science DOI 10.1002/app
`
`