Contact Lens & Anterior Eye 27 (2004) 193208 www.elsevier.

com/locate/clae

Water properties of soft contact lens materials

Ioannis Tranoudis, Nathan Efron*
Eurolens Research, Department of Optometry, The University of Manchester, P.O. Box 88, Manchester M60 1QD, UK

Abstract The properties of water in soft contact lenses such as the water content, free-to-bound water ratio, and the extent to which soft lenses dehydrate during wear, are key determinants of their in eye performance and oxygen transmissibility characteristics. This study describes clinical and laboratory experiments that were conducted in order to examine the state of water in eight soft contact lenses manufactured from different materials. Specically, lenses made from the following eight materials (and nominal water contents) were used: HEMA/VP 40%, HEMA/VP 55%, HEMA/VP 70%, VP/MMA 55%, VP/MMA 70%, HEMA 40%, HEMA/MAA 55% and HEMA/MAA 70% [HEMA = 2hydroxy-ethyl methacrylate, VP = vinyl pyrrolidone, MMA = methyl methacrylate, MAA = methacrylic acid]. Differential scanning calorimetry (DSC) was used for measuring the free water content in the materials listed above. Some noticeable differences in water properties among soft contact lens materials having approximately the same water contents were revealed. Low water content materials exhibited a simple endotherm and all water melted around 0 8C. On the other hand, medium and high water content materials exhibited multiple melting endotherms, representing a broad range of interactions between water and the polymer. Low water content soft contact lenses have approximately the same amount of bound water as those with much higher water contents. Six subjects were then tted with the same lenses for one day. In vitro measurements of water content and oxygen transmissibility were taken at 35 8C, both before lens tting and after 6 h of lens wear. Water content and oxygen transmissibility were correlated with the water properties of the soft contact lens materials. The relative change in lens water content (%DWC) and relative change in lens oxygen transmissibility (%DDk/t) were calculated and correlated with the water properties of the eight soft contact lens materials. It was concluded that (a) oxygen transmissibility, free water content and free-to-bound water ratio are increased when the water content of a contact lens is increased and (b) water content, free water content and free-to-bound water ratio cannot be used for the prediction of soft contact lens dehydration in vivo. # 2004 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
Keywords: Water properties; Hydrogel contact lens materials; Differential scanning calorimetry; Free water content; Bound water content; Free-to-bound water ratio; Lens dehydration

1. Introduction The soft contact lens has a exible property mainly due to retentional water. A polymer usually has both hydrophilic and hydrophobic groups. The hydrated state is formed as a result of the interaction between hydrophilic and hydrophobic groups and water. A water molecule is liable to be bound to a polymer or trapped in a small space formed within the polymer. An hydrated soft contact lens is a typical example of a material carrying bound water, trapped in molecular spaces. In an hydrated soft contact lens of three dimensional network structure, its physical properties are governed by the state of the water within the polymer [1].
* Corresponding author. Tel.: +44 161 306 3886; fax: +44 870 831 6625. E-mail address: n.efron@manchester.ac.uk (N. Efron).

Alternative approaches to classifying and describing the state of water in hydrogels have been described (Table 1) [2]. This classication has been used by Pedley and Tighe [3] and others and is the most widely used system. Tightly bound water is usually associated with water molecules, which have direct hydrogen bonding with the polar groups of the polymer matrix, or water molecules that strongly interact with ionic residues of the polymer matrix. As a result, it is non-freezable under normal characterisation conditions. Free or bulk water refers to the water molecules, which do not interact at all with the polymer matrix and have hydrogen-bonding characteristic of pure water (freezing at 273 K). Loosely bound water is more vaguely dened and covers diverse classes of water that remain in liquid state below the normal freezing temperature. Melting temperatures vary depending on the amount of water present and the

1367-0484/$ see front matter # 2004 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clae.2004.08.003

194

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

Table 1 Classication of the state of water in water-swollen polymer systems Bound water Non-freezable Freezable at/or below 180 K Bound Primary bound Tightly bound Freezable Freezing between 180 and 273 K Interfacial Secondary bound Loosely bound Free water, freezable Freeze at 273 K Bulk Free bulk

state of bonding with the polymer matrix. Loosely bound water usually refers to water molecules in water-swollen polymer systems which are only loosely associated with polar groups through hydrogen bonding, but have higher hydrogen-bonding energies than that of pure water. Descriptions of bound water include tightly bound water and loosely bound water. There exists a great deal of confusion concerning nonfreezable and freezable bound water due to the limitation of measurements by some analytical techniques. The existence and quantication of free water, bound water, freezable water and non-freezable water have been determined through the use of differential scanning calorimetry (DSC), thermogravimetric analysis (TGA) and nuclear magnetic resonance (NMR) spectroscopy. Infra-red spectroscopy and chromatography have also been used to study the state of water. The water absorbance characteristics of hydrophilic polymers vary widely depending on many factors. The most widely recognised factors are: (1) nature of hydrophilic groups; (2) cross linking density of the polymer system; (3) degree of water saturation; (4) the medium of the waterswollen polymer system; (5) hydrophobic character and (6) temperature. These factors are further elaborated below. It is well known that polar groups in a polymer molecule interact more strongly than non-polar groups with water molecules, in the form of iondipole, dipoledipole or hydrogen bonding. Functional groups of different polarity bind with water in varying degrees, thus affecting the distribution of bound to free water. Ionic functional groups/ residues and strong acids bind water more strongly than other non-ionic polar groups, when compared on the basis of number of water molecules per functional group. It is also clear that amide groups, particularly lactam groups such as pyrrolidone, have stronger binding power than either hydroxy or ether groups. However, when comparing the binding capability of hydrophilic polymersexcept for HEMA and polyvinyl pyrrolidone (PVP)there is little information in the existing literature dening the amount of free water, or the ratio of bound to free water. The limited information that is available suggests that a functional group with stronger water binding capacity does not necessarily translate into a higher bound to free water ratio. Some polymers with functional groups of average water binding capacity, such as HEMA, have higher ratios of bound to free

water compared with those polymers containing functional groups of higher water-binding capability, such as the pyrrolidone group [2]. The presence of a cross-linking agent is the key to holding a hydrogel system together. Many hydrogel properties are affected by the number of crosslinkages (or cross-linking density). It is well documented that increasing the density of crosslinkages causes a decrease in water content [4]. At the same time, the absolute amount of bound water increases as the amount of free water decreases, due to the reduced mobility of water molecules in the more rigid structure [46]. Normally, at low levels of water content, all water in hydrogels exists in a tightly bound form. This is obvious in view of the available sites for hydrogen bonding and strong iondipole interaction. At a medium level of water content, loosely bound water exists in addition to tightly bound water, which already occupies the available sites for strong interaction. Free water begins to appear at higher levels of water content. The relationship between water content and tightly bound, loosely bound and free water is illustrated in Fig. 1. This relationship implies that for hydrogels derived from the same polymeric system, a low water content hydrogel with x% of water would have the same amount of tightly bound water as those with much higher water contents, such as those with y%, z% or v% of water. Similarly, a hydrogel with water content z%, would have the same amount of tightly and loosely bound water as a hydrogel with a much higher water content, such as that with v% of water. Most free and bound water measurements have been performed on polymers swollen in distilled water. Adding ionic salts or water-soluble polar organic compounds changes the distribution of bound and free water, favouring bound water by creating additional water binding sites. However, it does not necessarily change the saturated water content of the swollen polymer system. This is due to the reduced mobility of water molecules caused by the solvation of ions and polar species. For example, HEMA lenses have higher levels of tightly bound water in aqueous sodium chloride solution than when the lenses are placed in distilled water [7]. As hydrogen bonding is responsible to a large degree for the amount of tightly bound water, inter-chain and intermolecular hydrogen bonding reduce the total available sites for hydrogen bonding with water molecules, thus reducing the amount of bound water.

Fig. 1. The relationship between water content and tightly-bound, loosely bound and free water for hydrogels.

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

195

The amount of bound water is also affected by temperature. Increasing temperature reduces the amount of bound water due to many factors [2]. This phenomenon has been observed for many water-swollen systems, including cellulose (Ogiwara et al., 1969, cited in Bausch and Lomb) [2]. Lee et al. [8] obtained bulk gel conductivity data for HEMA. The activation energy for specic conduction was calculated. A plot of the activation energy versus percent of water in the gel, clearly indicated three different zones, showing three possible classes of water in the gels. These results were conrmed by thermal expansion measurements. The high water content gels (50%) demonstrated an extremely sharp volume change at 0 8C, indicating the presence of normal free water. Lower water content gels (20%) showed no anomalous change in thermal expansion, indicating that the water is bound. The medium water content gels exhibited intermediate behaviour. A semiquantitative analysis of the three classes of water using DSC studies demonstrated that the low water content gel (20%) consisted mainly of bound water, which exhibited no phase transitions over the range 15 8C to 24 8C. The high water content gels showed phase transitions near 0 8C. The medium water content gels show gradual changes in phase transition at temperatures near 0 8C. DSC was also used by Pedley and Tighe [3] in a study of a series of hydrogels, in an attempt to correlate water binding and transport properties. DSC and oxygen transport studies were carried out on a series of styrene-2-hydroxyethyl methacrylate copolymers. The transport of dissolved oxygen through those copolymers, which contained no free water, was found to be negligible in comparison to those in which both free and bound water was present. The free and bound water contents in peruorosulphonated membranes and sulphonated hydrocarbon membranes using DSC were estimated by Tasaka et al. [9]. They found that in the sulphonated hydrocarbon membranes the amount of free water decreased with increasing divinylbenzene content and decreasing water content. Mirejovsky et al. [10] examined the effect of absorbed substances on the properties of the water in various soft contact lens materials by exposing contact lenses [Hydron 04 (Allergan Optical, Irvine, CA, USA), B&L 70 (Bausch and Lomb, Rochester, NY, USA), Durasoft 3 (Wesley Jessen Inc., Chicago, IL, USA), Vistamarc and Acuvue (Vistakon, Jacksonville, FL, USA)] to an articial tear solution for various periods up to 14 days. They found that the only materials affected were the high water ionic lenses, which absorbed a large amount of protein, predominantly lysozyme. They evaluated the free water of contact lenses using DSC. In the Durasoft 3 lenses, the equilibrium water content dropped from 49% to 46% and the free water from 28% to 21%. Similar changes were seen in the Vistamarc lenses. After a 10-day exposure of the Acuvue lens to articial tears, the water content decreased from 53% to 47% and the amount of free water from 33% to 23%. The decrease

in the permeability of water seen with these materials was consistent with the decrease of the free water, i.e., the water able to participate in diffusion. Since the free water content determines the transport through hydrogels, it is anticipated that lens characteristics dependant upon this would be affected by the presence of proteins within the polymer matrix. Mirejovsky et al. [10]using Dk data obtained from manufacturersestimated that an absolute change of 10% in the amount of free water could lead to a decrease in oxygen permeability of as much as 7 Dk units. Mirejovsky et al. [11] tried to determine a set of properties for the water contained within soft contact lens materials with the aim of developing a model, which would predict their propensity to induce corneal desiccation staining. They postulated that materials containing a larger proportion free water would tend to induce corneal desiccation more readily than materials containing a larger proportion bound water. The water structure (as measured by DSC) and the permeabilities of water and glucose were determined for a series of commercial soft lenses. They noted lower levels of staining for a material with a lower glucose permeability and a larger amount of water melting below 0 8C than for a control lens, even though both materials were similar in water content and water permeability. These authors used only two contact lens materials to test their hypothesis, thus precluding the drawing of general conclusions. DSC has been also used in order to examine the state of water in human crystalline lenses [12,13]. Isothermal weight loss studies have been performed on soft contact lenses by Kwok [14]. He examined 35 lenses, with nominal water contents of 37.5% to 74%. Isothermal dehydration at 34 8C demonstrated two major phases, each with a different rate constant. The rst phase was assumed to be evaporation of a free water fraction. The second phase, with a higher activation energy, was apparently due to water ligand binding in the lens matrix. The estimated bound fraction at zero time was found to reach 20% of total lens water in low water content lenses. Kwoks results indicated that the effective amount of free water may be lower than previously assumed, especially in low water content soft contact lenses. The water structure in polymers [1517] and the dehydration process in soft contact lens materials [18,19] have been examined using NMR. For example, NMR relaxation data has been used as a predictor for on-eye soft contact lens dehydration [18]. Proton NMR relaxation times (T1 and T2), were determined for a series of contact lenses for which on eye dehydration data were also available. It was demonstrated that NMR relaxation times are dependant upon lens water content, but the dependence is not monotonic. T1 values varied between 100 and 800 ms, and T2 values varied between 6 and 85 ms for the lenses studied. The mobility of these protons varies by more than a factor of 10 for the lenses studied. A test for linear correlation between NMR relaxation rate, 1/T1 and relative

196

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

change in lens water mass, %Dmw gave r = 0.830 for all data, and r = 0.904 if one lens was excluded. Masters et al. [20] investigated the suitability of water proton spinlattice relaxation time T1 as a measure of corneal dysfunction of rabbit corneas. Proton NMR measurements were performed on excised rabbit corneas using the saturation recovery method to determine the proton spinlattice relaxation time T1. Both freshly excised and progressively swollen rabbit corneas were studied. The experimental results were consistent with a two-compartment model of bound water and free water. This model resulted in a linear correlation between 1/T1 (measured) and 1/corneal hydration. This correlation suggests that water proton T1 values can be used as an index of corneal dysfunction. Quinn et al. [21] and Smyth et al. [22] also conducted DSC and NMR studies on hydrated copolymers VP/MMA and HEMA. They concluded that although HEMA is less hydrophilic than VP/MMA, the relative fraction of bound water is signicantly higher. Castoro and Bettelheim [23] and Wang and Bettelheim [24] investigated the cortical and nuclear samples from rat lenses by DSC (free water content) and TGA (total water content). The aim of this experiment was to evaluate the free-tobound water ratio of eight different types of soft contact lens materials using the DSC technique and to examine the relationships that exist among the various water properties of soft contact lens materials.

unit contains a vinyl group that is capable of being polymerized. The following are a few examples of monomers that are commonly used in hydrogel contact lens materials:  2-Hydroxy-ethyl methacrylate(HEMA) monomer and its non-cross linked low molecular weight polymer are water-soluble. It is the primary monomer from which the rst commercial soft contact lens was made.  Methacrylic acid(MAA) is used to boost the water content in the hydrogel.  Methyl methacrylate(MMA) is the monomer unit that makes up the PMMA (polymethyl methacrylate) lens. It is sometimes used to lower water content or in order to improve hardness and strength in some soft contact lenses.  Vinyl pyrrolidone(VP) is an important monomer in addition to the methacrylates. Due to its hydrophilicity, it is commonly used to increase water content. In order to be able to compare lenses made from different materials, non-commercial contact lenses were used throughout this study. The purpose was to examine lenses with material composition of HEMA/VP, VP/MMA and HEMA/ MAA of low (40%), medium (55%) and high (70%) water content. Hence, the contact lens materials were formulated specically for these experiments by the same contact lens material company (Vista Optics, Cheshire, England). Some of the lenses are not commercially available. All lenses were manufactured (lathe-cut) by the same contact lens laboratory and had the same bicurve design with similar nominal parameters: TD (14.00 mm), BVP (3.00 D), BOZR (8.7 mm) and tc (0.12 mm). The fabricated lenses covered a wide range of contact lens groups according to UK and FDA classications. The water contents of the contact lenses are only approximate values. These materials are described in Table 2. Two lenses fabricated from each of these materials were used in these experiments. The measured parameters (BVP, BOZR, TD, tc and WC) of the lenses used in these experiments can be found in Table 3.

2. Methods 2.1. Lenses Notwithstanding the recent advent of silicone hydrogel materials, the majority of soft contact lenses available at the present time are manufactured from hydrogels. These are cross-linked hydrophilic (water-loving) polymers and can be made by polymerising suitable monomers with a crosslinking agent, or less commonly, by the post-treatment of non-cross linked hydrophilic polymers. The monomer is the building block for these polymers. In general, the monomer

Table 2 Soft contact lens materials used in this study Copolymer type (WC) HEMA/VP (40%) HEMA/VP (55%) HEMA/VP (70%) VP/MMA (55%) VP/MMA (70%) HEMA (40%) HEMA/MAA (55%) HEMA/MAA (70%)
a b

Trade name Vistagel Vistagel Vistagel Vistagel Vistagel Vistagel Vistagel Vistagel 42 55 75 60 75 38 55 70 HC H H

UK classication Filcon Filcon Filcon Filcon Filcon Filcon Filcon Filcon 3a 3a 3a 4a 4a 1a 1b 1b

FDA classication Group Group Group Group Group Group Group Group I II II II II I IV IV

Material composition weight (%) 90/10 70/30 45/55 70/30 75/25 100 98/2 96/4

Cross linker EGDMAa EGDMA EGDMA AMAb AMA EGDMA EGDMA EGDMA

R MA MA

Ethylene glycol dimethacrylate. Alkyl methacrylate.

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208 Table 3 Verication (at 20 8C) of lenses used for DSC Lens type Lens code Parameters (20 8C) BVP (D) HEMA/VP 40% HEMA/VP 55% HEMA/VP 70% VP/MMA 55% VP/MMA 70% HEMA 40% HEMA/MAA 55% HEMA/MAA 70% Mean S.D. A6 A7 B5 B6 C5 C3 E5 E6 F4 F5 G6 G7 H5 H6 I5 I6 3.01 2.51 2.63 2.17 2.85 3.55 2.57 2.87 3.18 2.61 2.84 2.96 2.97 3.25 3.00 3.21 2.89 0.34 tc (mm) 0.096 0.099 0.093 0.101 0.085 0.077 0.159 0.088 0.086 0.161 0.104 0.108 0.129 0.148 0.124 0.148 0.113 0.028 TD (mm) 13.80 13.90 13.80 13.90 13.70 13.90 14.30 14.00 13.90 14.10 14.00 13.90 14.50 14.10 14.50 14.30 14.04 0.24 BOZR (mm) 9.30 9.30 9.70 9.60 8.70 8.30 9.10 9.40 8.70 8.70 9.20 9.10 9.20 9.20 8.80 8.80 9.07 0.37

197

WC (%) 37 37 47 52 72 69 60 60 69 68 38 38 56 56 67 67

2.2. Differential scanning calorimetry The 910 DSC System (Du Pont Company, Wilmington, DE, USA) was used in this work; this is a plug-in module that can be used with any of the Du Pont thermal analysers. The system measures temperature and heat ow associated with material transitions, providing quantitative and qualitative data on endothermic (heat absorption) and exothermic (heat evolution) processes. 2.3. Experimental procedure Melting endotherms of water in soft contact lenses were determined using the DSC instrument described above. Lenses were rst equilibrated overnight in saline solution. Mirejovsky et al. [10] reported that before scanning, all samples were kept at 40 8C for 815 h. Mirejovsky et al. [11] found that 4 h of cooling were insufcient as compared with 824 h. Pedley and Tighe [3] did not report such a preparation of the samples. The lenses to be studied by DSC were lightly blotted with tissue to remove surface water and then hermetically sealed in aluminium pans (Fig. 2). The contact lens samples were then punched out with a cork borer to t into the aluminium sample pan. In order to achieve a high reproducibility of results for material prone to fast dehydration, the blotted samples were placed quickly onto a tared pre-weighed sample pan and lid, sealed, weighed and immediately placed in the DSC instrument. The aim of this procedure was to avoid water evaporation before scanning. Two samples from each material were analysed and the mean values of the two results were taken. Good reproducibility of results for the same lens (with samples cut in the centre and at the edge) and between lenses has been reported by Mirejovsky et al.

[11]. In separate experiments they also found that the results are not affected by the thickness/surface ratio of the lens. The samples were scanned from 40 8C to +30 8C with a heating rate of 5 8C/min. Mirejovsky et al. [10,11] scanned their samples from 40 8C to +10 8C with a heating rate of 5 8C/min. Pedley and Tighe [3] scanned their samples from 40 8C to +20 8C at a heating rate of 1.25 8C/min or 5 8C/ min. The area under the melting peak and the heat of fusion of pure water (79.72 cal/g = 340.6 J/g) were used to calculate the percentage of free water. The amount of bound water was obtained by subtracting the amount of free water from the total percent water content, whereby: Total water content% free water content% bound water content%:

Fig. 2. Process for encapsulating the contact lens sample.

198

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

All water amounts were expressed as percentages of the total weight of the hydrated polymer. The DSC experiment was conducted in a masked and randomised manner in order to avoid experimental bias. The following formula [25] was used to calculate the percentage of free water: C DHtr 1 106 m DHf

3. Results The endothermic DSC curves for the eight different contact lens materials used in this study (two samples per material) are shown in Figs. 318. In these gures, two peaks can be observed: one sharp peak (peak 1) at about 0 8C and another broad peak (peak 2) at about 10 8C, respectively [9]. According to Tasaka et al. [9], peak 1 corresponds to the free water in contact lenses. Peak 2 corresponds to the water with partially restricted movement due to the presence of xed charges; that is, loosely bound water. The area of each peak was estimated from the difference between the endotherm curve and the straight line drawn in the gure, using the computer program that controls the DSC experiments. Certainly, the absolute value of the area depends on how this line was drawn. However, any resultant errors will be minimal because the portion of bound water is much smaller than that of the total water. If it is assumed that the heat of fusion of pure water is 340.6 J/g (79.72 cal/g), the amount of free water of peaks 1 and 2 can be estimated.

where C = concentration of water (mg/g), DHtr = heat of transition (mJ), m = sample weight (mg), DHf = heat fusion of water (340.6 J/g) and DHtr = 60ABEDqs; where A = peak area (cm2), B = time base (min/cm), E = cell calibration coefcient (mW/mV (dimensionless)) and Dqs = Y-axis sensitivity (mV/cm). The amount (DHtr/m) can automatically be estimated using a specic utility of the computer program that controls the DSC measurements. To test the precision of measurement of this technique, ten measurements were made on samples taken from identical Acuvue lenses (Vistakon, Jacksonville, FL, USA). The means and standard deviations of these measurements were as follows: for free water content 30.15 3.80%; for bound water content 27.85 3.80%; and for free-to-bound water ratio 1.12 0.31. 2.4. Clinical study Two lenses from each of the eight soft contact lens groups were used in experiments concerning the stability of water content and oxygen transmissibility1. Six subjects were tted with lenses for one day. In vitro measurements of water content and oxygen transmissibility were undertaken at 35 8C before lens tting and after 6 h of lens wear. In order to compare the difference of the parameters both before and after wear, among the eight different materials that were used in this study, the relative change in lens parameter (%DP) was determined [26]. That is: %DP P P0 100 P

Fig. 3. Endothermic DSC curve for the HEMA/VP 40% (sample A6).

where P is the lens parameter before wear and P0 is the lens parameter after 6-h wear, giving the amount %DP a positive value. The relative change in lens water content (%DWC) and relative change in lens oxygen transmissibility (%DDk/t) were calculated. The results of this clinical study, which have been reported previously [27], were used to examine the relationships that exist among the water properties of soft contact lens materials.

Exponential terms and units of oxygen transmissibility (Dk/t) are 109 (cm/s) (mlO2/ml mmHg). In the exponential terms and the units for transmissibility will be omitted.
1

Fig. 4. Endothermic DSC curve for the HEMA/VP 40% (sample A7).

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

199

Fig. 5. Endothermic DSC curve for the HEMA/VP 55% (sample B5).

Fig. 8. Endothermic DSC curve for the HEMA/VP 70% (sample C5).

Fig. 6. Endothermic DSC curve for the HEMA/VP 55% (sample B6).

Fig. 9. Endothermic DSC curve for the VP/MMA 55% (sample E5).

Fig. 7. Endothermic DSC curve for the HEMA/VP 70% (sample C3).

Fig. 10. Endothermic DSC curve for the VP/MMA 55% (sample E6).

200

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

Fig. 11. Endothermic DSC curve for the VP/MMA 70% (sample F4).

Fig. 14. Endothermic DSC curve for the HEMA 40% (sample G7).

Fig. 12. Endothermic DSC curve for the VP/MMA 70% (sample F5).

Fig. 15. Endothermic DSC curve for the HEMA/MAA 55% (sample H5).

Fig. 13. Endothermic DSC curve for the HEMA 40% (sample G6).

Fig. 16. Endothermic DSC curve for the HEMA/MAA 55% (sample H6).

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

201

Fig. 17. Endothermic DSC curve for the HEMA/MAA 70% (sample I5).

Fig. 19. Equilibrium water content and the relative proportions of free and bound water for the eight soft contact lens materials.

Fig. 18. Endothermic DSC curve for the HEMA/MAA 70% (sample I6).

These amounts, calculated from the area of the peaks in Figs. 318, are shown in Table 4. Two samples per material were used and the calculated mean values are presented. The amount of bound water is estimated from the difference between the total amount of water measured using the Atago CL-1 soft contact lens refractometer and the amount of free water. The last column in Table 4 gives the

estimated mean values of the free-to-bound water ratio for each of the eight contact lens materials. It is not possible to test statistically whether the differences in the water properties among the eight different soft contact lens materials are signicant, due to the fact that only two measurements were obtained for each material. Considering the standard deviations of the measured values obtained during the testing of the precision of this technique (S.D. of free water content = 3.80%), it can be assumed that there are real differences in the values of the water properties tested among the eight materials used in this study. Fig. 19 shows the equilibrium water content and the relative proportions of free and bound water for each type of lens material. As total water content increases, the free water content also increases. The amount of bound water content remains almost constant for all the materials except the VP/ MMA 55%. This nding is consistent with the theory presented earlier, whereby low water content soft contact lenses would be expected to have the same amount of bound water as lenses with much higher water contents.

Table 4 Total, free and bound water content (WC; %) and free-to-bound water ratio of the eight different soft contact lens materials Materials Mean WC (%) HEMA/VP 40% HEMA/VP 55% HEMA/VP 70% VP/MMA 55% VP/MMA 70% HEMA 40% HEMA/MAA 55% HEMA/MAA 70% 37.00 49.50 70.50 60.00 68.50 38.00 56.00 67.00 Free WC (%) 13.38 21.31 43.73 21.79 41.56 16.78 26.96 41.58 Bound WC (%) 23.62 28.19 26.77 38.21 26.94 21.22 29.04 25.42 Free-to-bound WC (%) 0.565 0.765 1.635 0.575 1.540 0.795 0.930 1.640

202

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

Fig. 20. Relationship between free-to-bound water ratio vs. free water content.

3.1. Free-to-bound water ratio versus free WC The free-to-bound water ratio of the materials measured correlated signicantly with their free water content (Fig. 20; R2 = 0.9155, p = 0.0002). The linear regression (R = 0.9568) is positive, indicating that by increasing free water content, the free-to-bound water ratio is increased.

3.2. Free WC versus WC The free water content of the materials correlated signicantly with their water content (Fig. 21; R2 = 0.8575, p = 0.0010). The linear regression (R = 0.9260) is positive, which indicates that increasing water content will result in an increase in free water content.

Fig. 21. Relationship between free water content vs. water content.

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

203

Fig. 22. Relationship between free-to-bound water ratio vs. water content.

3.3. Free-to-bound water ratio versus WC The free-to-bound water ratio of the materials correlated signicantly with their water content (Fig. 22; R2 = 0.6366, p = 0.0176). The linear regression (R = 0.7979) is positive, demonstrating that by increasing water content, the free-tobound water ratio is increased.

3.4. Free WC versus Dk/t The free water content of the materials correlated signicantly with their Dk/t (Fig. 23; R2 = 0.7685, p = 0.0043). The linear regression (R = 0.8766) is positive, indicating that increasing the free water content will increase Dk/t.

Fig. 23. Relationship between free water content vs. oxygen transmissibility.

204

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

Fig. 24. Relationship between free-to-bound water ratio vs. oxygen transmissibility.

3.5. Free-to-bound water ratio versus Dk/t The free-to-bound water ratio of the materials correlated signicantly with their Dk/t (Fig. 24; R2 = 0.5999, p = 0.0240). The linear regression (R = 0.7745) is positive, demonstrating that by increasing the free-to-bound water ratio, Dk/t will be increased. 3.6. %DWC versus free WC Correlating the relative change in water content following a 6-h wear period of the materials measured to their free water content failed to reveal a signicant relationship (R2 = 0.0608, p = 0.5560). 3.7. %DWC versus free-to-bound water ratio Correlating the relative change in water content following a 6-h wear period of the materials measured to their freeto-bound water ratio failed to reveal a signicant relationship (R2 = 0.0260, p = 0.7030). 3.8. %DDk/t versus free WC Correlating the relative change in Dk/t following a 6-h wear period of the materials used to their free water content failed to reveal a signicant relationship (R2 = 0.4209, p = 0.0818). 3.9. %DDk/t versus free-to-bound water ratio Correlating the relative change in Dk/t following a 6-h wear period of the materials measured to their free-to-

bound water ratio failed to reveal a signicant relationship (R2 = 0.2742, p = 0.1829).

4. Discussion The calculation of the relative amounts of free and bound water is approximate, since exact heats of melting are required for the calculation of the amount of free water from the peak area and the measured heat/g of a wet sample. It can be expected that, in polymers showing multiple endotherms, a portion of the water melting below 0 8C would have a lower heat of fusion than pure water. It is clear that the use of the heat of fusion of pure water (an upper limit) for the calculation of the amounts of free water would lead to a slight overestimation of the amount of bound water [11]. In spite of these shortcomings, Fig. 19 reveals some noticeable differences among soft contact lens materials. The HEMA 40% material contains more free water than the HEMA/VP 40%. In medium water content materials the HEMA/MAA 55% has the highest amount of water melting at 0 8C. HEMA/VP 55% and VP/MMA 55% have about the same amount of free water content. In high water content materials the HEMA/VP 70% has the highest amount of free water. The VP/MMA 70% has about the same amount of free water as the HEMA/MAA 70%. However, the differences among materials are even more pronounced when the patterns of melting endotherms (Figs. 318) are simply compared. A close agreement between the two runs for each material can be observed. The terminology of free, partially bound and bound water has been used freely. The different categories of water can be

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

205

designated on the basis of three parameters: (1) the average number of hydrogen bonds/molecule of water that are formed, (2) the length of hydrogen bonds and (3) the angle of hydrogen bonds in water. In free liquid water every molecule is hydrogen bonded, on average to two or three nearest neighbours. In the hexagonal ice structure, each water molecule is hydrogen bonded to four nearest neighbours. Presumably the length of the H-bond in free liquid water is somewhat longer, thus weaker [12]. Another way of looking at water structure is considering that the bond angles in ice are tetrahedral, while in liquid (free) water the exibility of hydrogen bonds increases. This results in both bonding and distortions. The distortion provides the absorption of energy in the fusion process without decreasing the actual number of hydrogen bonds (breaking the hydrogen bonds). The DSC technique is rapid and accurate and measures only the free water content, as the bound water does not freeze when the temperature of the polymer is lowered below the freezing point of water. The bound water forms hydrogen bonds with the polymers atoms, rather than with other water molecules, as would be necessary for the water to freeze. Free water, however, can bond with other water molecules to form ice crystals when the temperature is lowered sufciently. The most important nding of this experiment is that differences in water/polymer interactions among various soft contact lens materials can be identied and analysed using a relatively simple technique. This work demonstrates the usefulness of DSC measurements. The melting endotherms are reproducible and indirectly give insights into the extent of interactions of water with a polymer matrix. As has been outlined earlierwater, which does not participate in transport, is water which interacts strongly with the polymer. In the DSC experiment, this water corresponds to bound water. Inspection of Table 4 shows that the amounts of bound water varied from 21% to 38 %, depending upon the type of polymer. This implies that in some materials, a portion of the water melting below 0 8C will not participate in water transport. The results of the present study are in general agreement with reported values [2,3,5,10,11] for similar types of polymers (see Table 5). Table 5 presents data of the state of water in soft contact lenses and hydrogels from other researchers as well as literature from manufacturers. Total, free and bound water contents have been expressed as percentages in order to be able to compare the values of the water properties presented in this table. Direct comparisons between the present data and data quoted from different references in Table 5 cannot be made due to the fact that (1) lenses or hydrogels with different compositions are presented; (2) the test conditions are different (temperature range, heating rate); (3) the sample preparation may be different and (4) in the report of Bausch and Lomb [2], the method of obtaining the presented data is not specied. There is obviously a distinction between the more complex of the descriptions of water states shown in Table 1 and the experimental results under consideration, because

the former are theoretical concepts, which do not necessarily correspond to the results of any particular experimental technique. Similarly, the distinction that different experimental techniques make between bound and free water will be affected to varying extents by the presence of intermediate states of water [3]. The results of this study are consistent with the idea of a continuum of water states between the primary statewhich is hydrogen bonded to functional groups in the polymer and the state whereby water is unaffected by its polymer environment. In the latter case, water crystallises and remelts in a manner indistinguishable from that of pure water. There is thus a continuum of water states whose crystallisation is somewhat affected by the environment and takes place more slowly. The freezing behaviour of water-swollen polymers has been shown by a number of workers to be anomalous, in that only a part of the available water freezes even when cooled to very low temperatures. The use of DSC enables a quantitative determination of the relative amounts of free and bound water to be made. On the basis of this information, coupled with the availability of hydrogen bonding sites in each monomer unit and in light of the fact that the bound water has remained unaffected (i.e. does not freeze) on cooling to the temperatures of liquid nitrogen it seems reasonable to identify bound water with water that is directly associated by hydrogen bonding with the polymer. The use of DSC as described here offers a valuable method for the study of the states of water in polymers used for soft contact lens materials. In agreement with the literature, low water content materials exhibited a simple endotherm and all water melted around 0 8C. Medium and high water content materials exhibited multiple-melting endotherms, representing a broad range of interactions between the water and the polymer, similar to that reported by Pedley and Tighe [3] and Mirejovsky et al. [10,11]. A portion of the water, which maintained the properties of the free water melted at 0 8C, while some proportion of the water, which weakly interacted with the polymer, had already melted at a temperature below 0 8C. Brennan and Efron [28] suggested that the way in which water is contained within the polymer matrix of the lensin other words the free-to-bound water ratiomay govern the extent of dehydration in that the free water component may leave the lens more readily than the bound component. This hypothesis was tested in the present study. Preliminary correlations among free-to-bound water ratio, free water content and water content conrmed the theory that the higher water content lenses contain higher amounts of free water and that the free-to-bound water ratio is increased by increasing the free water content [2,3]. In addition, positive signicant correlations were revealed among oxygen transmissibility, free water content and free-to-bound water ratio. The attempt to correlate relative change in water content following a 6-h wear period and relative change in oxygen

206

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208

Table 5 Water properties of soft contact lenses and hydrogels Reference Pedley and Tighea [3] Trade name MSS:ACM:Sa Hydrogel composition WC (%) Total 50.00:50.00:0.00 49.50:49.50:1.00 48.75:48.75:2.50 48.00:48.00:4.00 47.50:47.50:5.00 47.00:47.00:6.00 46.25:46.25:7.50 45.00:45.00:10.00 43.75:43.75:12.50 42.50:42.50:15.00 39.6 33.5 31.3 29.1 29.2 28.1 33.1 32.1 34.7 33.8 34 48 52 53 67 35 40 48 50 52 52 53 53 63 64 67 71 72 System 1, HEMA contains 0.2% MAA, 0.16% diethylene glycol methacrylate and 0.01% EGDMA 18 25 30 35 38.5 40 45 50 22 25 30 35 38.5 40 45 50 9 17 30 14.5 18 40.8 56.7 74.1 77.5 37 49.5 70.5 60 68.5 38 Free 12.4 6.8 4.4 2.1 1.1 0.7 0.7 0.1 0.4 0.3 11 27 32 31 42 9 21 32 28 33 28 37 35 38 41 49 51 57 0 0 3.6 10.5 14.2 17.2 24.3 31 0 0 0 0.4 19.9 6 13.5 21.5 0 0 3.6 0 0 0.9 23.2 54.8 58.9 13.38 21.31 43.73 21.79 41.56 16.78 Bound 27.2 26.7 26.9 27 28.1 27.4 32.4 32 34.3 33.5 23 21 30 22 25 26 19 16 22 19 24 16 18 25 23 18 20 15 18 25 26.4 24.5 24.3 22.8 20.7 19 22 25 30 34.6 36.6 34 31.5 28.5 9 17 26.4 14.5 18 39.9 33.5 19.3 18.6 23.62 28.19 26.77 38.21 26.94 21.22 0.456 0.255 0.163 0.078 0.039 0.026 0.022 0.003 0.012 0.009 0.478 1.286 1.067 1.409 1.680 0.346 1.105 2.000 1.273 1.737 1.167 2.313 1.944 1.520 1.783 2.722 2.550 3.800 0 0 0.136 0.429 0.584 0.754 1.174 1.632 0 0 0 0.012 0.052 0.176 0.429 0.754 0 0 0.136 0 0 0.023 0.693 2.839 3.116 0.565 0.765 1.635 0.575 1.540 0.795 Free-to-bound water ratio

Mirejovsky et al.b [10]

Zero 4 Durasoft 3 Acuvue Vistamarc B&L 70 Zero 4 CSI Durasoft 3 Hydrocurve II Hydrosoft Softcon EW Acuvue Vistamarc SoftPerm 67 H 67 B&L 70 Durasoft 4 Permaex

Mirejovsky et al.c [11]

Bausch and Lombd [2]

System 2, HEMA (same as system 1) contains 1 mole TEGDMA

System 3 based on polymacon lens using HEMA VP/MMA containing 70.66% VP 29.12% MMA 0.026% EGDMA 0.19% AMA

Present study

Vistagel Vistagel Vistagel Vistagel Vistagel Vistagel

42 HC 55H 75 H 60 75 38 R

HEMA/VP 40% HEMA/VP 55% HEMA/VP 70% VP/MMA 55% VP/MMA 70% HEMA 40%

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208 Table 5 ( Continued ) Reference Trade name Hydrogel composition WC (%) Total Vistagel 55MA Vistagel 70MA
a b

207

Free-to-bound water ratio Free 29.04 25.42 Bound 0.930 1.640

HEMA/MAA 55% HEMA/MAA 70%

56 67

26.96 41.58

Molar composition MAA:ACM:S (methacrylic acid:acrylamide:styrene). Figures extrapolated from Fig. 1 of Mirejovsky et al. [10]. c Figures extrapolated from Fig. 1 of Mirejovsky et al. [11]. d HEMA (2-hydroxy-ethyl methacrylate), MAA (methacrylic acid), EGDMA (ethylene glycol dimethacrylate), TEGDMA (tetra ethylene glycol dimethacrylate), VP (vinyl pyrrolidone), MMA (methyl methacrylate), AMA (alkyl methacrylate).

transmissibility following a 6-h wear period with the free water content and free-to-bound water ratio failed to reveal signicant correlations. The ndings of this study conrmed the results obtained by Mirejovsky et al. [10] and Pedley and Tighe [3], whereby lenses of high water content seem to have a higher free water content as compared to lenses with low water contents. Larsen et al. [18] correlated in vivo lens dehydration to water mobility using proton NMR relaxation. Fatt [29] proposed a different approach to predict in vivo lens dehydration based upon swelling pressure and the calculation of water transport coefcients in hydrogels. In both cases, substantial differences in water properties were found between high water content and low water content hydrogels. Mirejovsky et al. [11] criticised both papers by noticing that it is not clear whether these approaches are sensitive enough to predict in vivo desiccation induced by materials with similar water contents but slightly different compositions. In order to answer this question, Mirejovsky et al. [11] compared an unspecied experimental material B to the Acuvue lens (Vistakon, Jacksonville, FL, USA). DSC measurements indicated a larger proportion of free water in the Acuvue lens than in the material B lens. The authors then tted these 2 materials to 14 subjects and corneal staining with uorescein was monitored immediately before and after a 6-h wear period. At the end of the study, corneal staining was present in 10 eyes wearing Acuvue and in 3 eyes wearing material B lenses. Although using a binomial distribution, the two-tailed probability of this occurring by change is low (p = 0.047), the above experiment cannot be used to draw general rules about the aetiology of soft contact lens dehydration. Larsen et al. [18] demonstrated that proton NMR T1 values can be used to predict on eye lens dehydration behaviour. NMR results have often been discussed in terms of fractions of water in various states, such as bound and free. Such a discussion uses a site model, normally involving two or three sites. For two sites, the relaxation time depends on the fractions of free water (Xf) and bound water (Xb) in the sample according to:     1 Xf Xb T1 T1f T1b where T1f and T1b are relaxation times for the free and bound water, respectively. Since only one parameter (the relaxation

time) is measured and three are required in the above equation, the problem is underestimated. Furthermore, there are also assumptions concerning the relaxation mechanism implicit in the site model, and these have recently been seriously challenged by further NMR studies [16,17]. Presumably, this is why Larsen et al. [18] did not consider estimating fractions of bound and free water from their data. Larsen et al. [18] also admitted that the NMR signal measured by them arises from both water and exchangeable polymer protons hence the inclusion of a polymer contribution may be required for the NMR results to correlate with the on-eye dehydration data, so that a parameter that characterises only lens water may not correlate with dehydration data. The DSC technique as applied here provides a rapid and accurate method for measuring free water in polymers used for soft contact lens materials, by virtue of its high constant calorimetric sensitivity, superior baseline performance and high precision and accuracy of calorimetric measurements. The procedure is sensitive, specic and requires only milligram quantities of contact lens material. Using the DSC method revealed some noticeable differences among soft contact lens materials with approximately the same or different water contents. As a general rule, it can be concluded that by increasing the total water content, the bound water remains constant and the free water increases. Finally, by increasing the water content of a material, oxygen transmissibility, free water content and free-tobound water ratio are also increased and the free water content and free-to-bound water ratio cannot be used for the prediction of dehydration. Research must be continued to further understand the in vitro and in vivo behaviour of hydrogel contact lens materials. Designing an optimal hydrogel contact lens material requires the achievement of a balance between hydrophilic and mechanical properties. By using unique chemistries, it ought to be possible to achieve an optimal balance of these properties without compromising the comfort of the patient.

Acknowledgement Dr. Tranoudis was supported by a grant from the State Scholarships Foundation, Republic of Greece.

208

I. Tranoudis, N. Efron / Contact Lens & Anterior Eye 27 (2004) 193208 [14] Kwok S. Thermogravimetric quantication of water ligand binding in soft contact lenses. Optom Vis Sci 1988;65:188S. [15] Amano H, Kawai K, Takahashi K, Hayano S, Nagaoka S, Sogami M, et al. Comparative H-NMR studies on the water structures in soft contact lens and protein gel. J Jpn Contact Lens Soc 1984;26:23744. [16] Roorda WE, de Bleyser J, Junginger HE, Leyte JC. Nuclear magnetic relaxation of water in hydrogels. Biomaterials 1990;11:1723. [17] Yamada-Nosaka A, Tanzawa H. H-NMR studies on water in methacrylate hydrogels. J Appl Polym Sci 1991;43:116570. [18] Larsen DW, Huff JW, Holden BA. Proton NMR relaxation in hydrogel contact lenses: correlation with in vivo lens dehydration data. Curr Eye Res 1990;9:697706. [19] Pescosolido N, Lupelli L, Brosio E, Delni M, Aureli T. Preliminary study on the dehydration of hydrogel contact lenses by NMR at low resolution. Contact Lens J 1990;18:1015. [20] Masters BR, Subramanian VH, Chance B. Rabbit cornea stromal hydration measured with NMR spectroscopy. Curr Eye Res 1983;2:31721. [21] Quinn FX, Kampff E, Smyth G, McBriety VJ. Water in hydrogels 1. A study of water in poly(N-vinyl-2-pyrrolidone/methylmethacrylate) copolymer. Macromolecules 1988;21:31918. [22] Smyth G, Quinn FX, McBrierty VJ. Water in hydrogels 2. A study of water in poly(hydroxyethyl methacrylate). Macromolecules 1988;21:3198204. [23] Castoro JA, Bettelheim FA. Distribution of the total and non-freezable water in rat lenses. Exp Eye Res 1986;43:18591. [24] Wang X, Bettelheim FA. Distribution of total and non-freezable water contents of galactosemic rat lenses. Curr Eye Res 1988;7:7716. [25] Baker KF, Cattiaux J. Thermal analysis high sensitivity determination of clustered water in polyethylene by differential scanning calorimetry. Paris: Du Pont Company; 1977. [26] Brennan NA, Efron N, Truong VT, Watkins RD. Denitions for hydration changes of hydrogel lenses. Ophthalmic Physiol Opt 1986;6:3338. [27] Tranoudis I, Efron N. Parameter stability of soft contact lenses made from different materials. Contact Lens Ant Eye 2004;27:11531. [28] Brennan NA, Efron N. Hydrogel lens dehydration: a material-dependent phenomenon? Contact Lens Forum 1987;12:289. [29] Fatt I. A predictive model for dehydration of a hydrogel contact lens in the eye. J Br Contact Lens Assoc 1989;12:1531.

References
[1] Kanome S. Fundamental chemistry and physical properties of polymer materials. In: Iwata S, editor. MeniconToyos 30th anniversary special compilation of research reports. Nagoya: MeniconToyo Contact Lens Co. Ltd.; 1982. p. 10938 (Chapter 6). [2] Bausch and Lomb. Free and bound water in water swollen polymer systems and their correlation with lens water dehydration. Rochester: Bausch and Lomb; 1994. [3] Pedley DG, Tighe BJ. Water binding properties of hydrogel polymers for reverse osmosis and related applications. Br Polym J 1979;11:130 6. [4] Sung YK, Gregonis DE, John MS, Andrade JD. Thermal and pulse NMR analysis of water in poly(2-hydroxyethyl methacrylate). J Appl Polym Sci 1981;26:371928. [5] Hatakeyema T, Yamouchi A, Hatakeyema H. Studies on bound water in poly(vinyl alcohol) hydrogel by DSC and FT-NMR. Eur Polym J 1984;20:614. [6] Collett JH, Spillane DEM, Pywell EJ. Infrared attenuated total reectance (ATR) in the characterisation of water structure within polyHEMA hydrogels. ACS Polym Reprints 1987;28:1412. [7] Quinn FX, McBrierty VJ, Wilson AC, Friends GD. Water in hydrogels. 3-Poly hydroxyethyl methacrylate/saline solution systems. Macromolecules 1990;23:457681. [8] Lee HB, John MS, Andrade JD. Nature of water in synthetic hydrogels. J Colliod Interface Sci 1975;51:22531. [9] Tasaka M, Suzuki S, Ogawa Y, Kamaya M. Freezing and nonfreezing water in charged membranes. J Membr Sci 1988;38:17583. [10] Mirejovsky D, Patel AS, Rodriquez DD. Effect of proteins on water and transport properties of various hydrogel contact lens materials. Curr Eye Res 1991;10:18796. [11] Mirejovsky D, Patel AS, Young G. Water properties of hydrogel contact lens materials: a possible predictive model for corneal desiccation staining. Biomaterials 1993;14:10808. [12] Bettelheim FA, Christian S, Lee LK. Differential scanning calorimetric measurements on human lenses. Curr Eye Res 1983;2:8038. [13] Bettelheim FA, Castoro JA, White O, Chylack Jr LT. Topographic correspondence between total and non-freezable water content and the appearance of cataract in human lenses. Curr Eye Res 1986;5:92532.