- Email: [email protected]
Void volume variations in contact lens polymers
Contact Lens & Anterior Eye 34 (2011) 2–6
Contents lists available at ScienceDirect
Contact Lens & Anterior Eye journal homepage: www.elsevier.com/locate/clae
Void volume variations in contact lens polymers P. Sane a,∗ , F. Tuomisto a , J.M. Holopainen b a b
Department of Applied Physics, Aalto University, Otakaari 1M, 02150 Espoo, Finland Helsinki Eye Lab, Department of Ophthalmology, University of Helsinki, Finland
a r t i c l e
i n f o
Keywords: Contact lens Positron annihilation spectroscopy Free volume pockets/voids Filling of voids
a b s t r a c t Aim: In this study, void size and free volume properties in different contact lens materials have been investigated in their hydrated state using positron annihilation lifetime spectroscopy (PALS). Methods: PALS is used to characterize the void size distributions inside the lens materials. Three different types of contact lenses were used (Balafilcon A, Hilafilcon B and Polymacon). Results: Measurements on different contact lenses reveal significant differences between the materials, up to ∼100% difference in void volume was observed between Hilafilcon B and Balafilcon A, the latter having larger voids. As oxygen diffusion is strongly correlated with the void sizes, the results are in good agreement with the usage recommendations of the specific lens types (daily disposable lenses or 1 month continuous use lenses). The void sizes in monthly lenses (Balafilcon A) were found to decrease 25% under artificial aqueous tear (albumin–water solution) exposure in 4 weeks leading to a significant decrease in the oxygen permeation rate through the contact lens. Yet, the voids were still significantly larger than in disposable or semi-disposable lenses. Conclusions: We have showed that PALS is a viable method to probe the microstructure of biotechnologically relevant polymers and can be used to quantify the void properties in different types of contact lenses. Usage recommendations correlate well with measured void sizes and the median void size decreases during the incubation of albumin solution as a function of time. We anticipate the use of PALS for any polymer-based intracorneal/intraocular device in which diffusivity plays a crucial role. © 2010 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
1. Introduction The local free volume of a polymer structure (void), corresponding to the unoccupied regions accessible to segmental motions, is an important parameter in the overall diffusion of atoms and molecules, either in gaseous or solid states, through the polymer. Positron annihilation lifetime spectroscopy (PALS) has been widely used to characterize atomic scale defects in semiconductors and metals and is also routinely used to study void distributions in polymer materials and in characterizing structural parameters such as the glass transition temperatures [1]. Lately we have employed this method to study also biological materials, such as lipid bilayers [2] as well as intact crystalline intraocular lenses [3]. Diffusion of molecules is largely determined by the presence and volume of voids in any material, and hence PALS can give direct information of the diffusion properties of the material in question. Oxygen diffusion is one of the most important factors in contact lens (CL) materials. Typically the permeation of oxygen through the CL is much less efficient in soft disposable CLs compared to
∗ Corresponding author. Tel.: +358 947023145. E-mail address: petri.sane@tkk.fi (P. Sane).
harder CLs designed for daily long term use. In an earlier study by Singh et al. [4] where PALS and measurements of gas diffusion rates were compared, it was shown that oxygen diffusion through CLs was strongly dependent on the void sizes. This effect was explained by the fact that gas permeability is the product of solubility and diffusion coefficient, causing the permeability to change with free volume. In another PALS study on CLs, Deepa and Ranganathaiah [5] reported that calcification of CLs reduced the void volume of these lenses, as the calcium deposits filled the voids. The focus of that work was, however, on the optical properties of the CLs before and after calcification. Both of these studies show that PALS is applicable to study ready-touse CLs (instead of representative polymer samples) of different types. Different types of CLs have different specifications on how long these should be used continuously. This is partly dependent in the oxygen permeability of each CL polymer. Furthermore, the contamination of CL by proteins, ions, and impurities present in the tear fluid will eventually reduce the diffusion properties of the CL by filling the voids within the polymer. In this work we have studied the void size distributions (and hence the diffusion characteristics) of different types of CLs. We also exposed different types of CLs to aqueous albumin solution and followed changes in the void volume as a function of time.
1367-0484/$ – see front matter © 2010 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clae.2010.06.008
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
2. Positron annihilation lifetime spectroscopy in polymers When an energetic positron from a radioactive source (22 Na in most cases) enters molecular media, it thermalizes rapidly after interacting with the surrounding molecules through inelastic collisions. A fraction of positrons forms a bound state with spin parallel electrons (ortho-positronium, o-Ps) that preferentially localizes in the free volume pockets in the material. In vacuum, o-Ps has a relatively long lifetime of 142 ns, compared to 125 ps for p-Ps (bound state of a positron and an electron with opposite spins). In a medium, the o-Ps prefers to undergo so-called pick-off annihilation with an electron of opposite spin during collision with molecules in the cavity wall in which it is localized. The pick-off process reduces the lifetime of o-Ps down to a few nanoseconds. The smaller the cavity size, the higher the frequency of collisions and the shorter the o-Ps lifetime, hence the o-Ps lifetime provides information of the size of the free volume pockets (i.e., voids). A quantitative semi-empirical relation has been established by Tao and Eldrup [6] (work on positronium trapping to voids in polymers) correlating the o-Ps lifetime ( polymer ) with the radius (R) of a spherical trap: polymer (ns) =
1 R 1 1− + sin 2 R+C 2
2R −1 R+C
,
(1)
where C is a semi-empirical constant of 0.166 nm [7]. The model applies directly to spherical voids and as such it is not directly applicable to characterize accurately the void sizes in CLs, as the specific form and distribution of forms of the voids is not known. However, the Tao–Eldrup model can be used to provide an orderof-magnitude estimate and means to compare the void sizes in polymers. 3. Experimental methods 3.1. Instrumentation The PALS measurements were performed using a sandwichgeometry of the sample material and the positron source. The positron/positronium lifetime is measured with a pair of scintillation detectors, one detecting the birth of a positron and the other its annihilation. When the positron source (22 Na) decays, it produces simultaneously one positron and one ∼1.27 MeV ␥quantum. When this quantum denoting the birth of the positron is detected, the pulse acts as a START-signal for the equipment. When the positron annihilates in the sample, as a free positron, p-Ps or o-Ps (pick-off) it produces two ␥-quanta of about 511 keV each, either of which act as the STOP signal. The time difference of START- and STOP-pulses gives the positron lifetime () in the material. In this study the annihilation spectrum was measured with a normal digital lifetime setup, consisting of two scintillation detectors with large scintillation heads producing a broad time resolution (full-width at half-maximum, FWHM, of the Gaussian resolution function ∼550 ps for the setup for preliminary measurements, ∼320 ps/260 ps for the further studies described later on) but with high efficiency. A broad time resolution is not optimal for lifetime separation, but as the measurements were planned to be performed in the hydrated state, the short duration of measurement was an important factor: the lenses were placed inside a sealed plastic bag, and the liquid (borate-buffered saline from the lens containers) drained slowly from the lenses to the bottom of the bag due to the horizontal geometry of the measurement setup. The setup is characterized in more detail in Ref. [8]. The lifetime spectra were analyzed first with PALSfit [9] to analyze the properties of the time resolution of the setup and then analyzed with MELT [10] in order to minimize possible fitting problems due to continuous void size distributions, the mathematical
3
basis of the aforementioned tools are presented in the references. Statistics of ∼2 to 3 million events were collected in each annihilation spectrum and the analysis was performed using a free fit with no source corrections resulting to 3–4 lifetime components depending on the sample and whether the water-lifetime ( 3 ) could be separated from the polymer lifetime ( 4 ), the two shortest components representing the annihilation of free positrons and p-Ps ( 1 ) as well as a ∼1 ns long o-Ps component ( 2 ). Goodness of the fits was evaluated with the 2 -output of the analysis software and similar range of values (1.1–1.2) was observed in all cases. 3.2. Contact lens materials and their measurement Preliminary experiments were performed with three different types of lenses, all from the same manufacturer, (Bausch & Lomb, B&L). The purpose of these experiments was to characterize the differences of lifetimes in different types of CLs. The studied lenses were: B&L Daily Disposable (DD, 59% water), B&L Soflens® (SL, 38%), and B&L Purevision® (PV, 36%). The lenses were chosen to have profound differences in their usage recommendations, ranging from the short term use (DD) to lenses that can be used continuously for a month (PV). The B&L Soflens is a CL that can be reused but is not recommended for continuous use. The base materials in these CLs are polymer-based Hilafilcon B (DD), Polymacon (SL) and a silicone–hydrogel based Balafilcon A (PV). We measured contact lenses in their “in-use” hydrated state, containing reasonable amounts of water (36–59%) and hence some of the o-Ps always annihilated in water. Because the lifetime components of some of the studied polymers were quite near the o-Ps lifetime in water, the measured lifetime component was in most cases an intensity weighted sum of water and polymer lifetimes. This effect is more significant in the samples measured with a PALS setup with broader time resolution. Due to this and other factors (e.g., chemical interactions of o-Ps with the material), the measured o-Ps lifetimes are, as usual in polymers, not discrete, but a distribution. Hence the radius R obtained by using the Tao–Eldrup model (Eq. (1)) should be interpreted as the mean radius of the free volume pockets. A layer consisting of at least five contact lenses was placed on both sides of the small (2 × 2 × 0.003 mm) positron source, surrounding the small source completely and thus the vast majority of positrons annihilated in the lenses based on the estimate of positron range in the CL polymers. The positron sources (approximately 0.5 MBq each) were manufactured by injecting 22 NaCl saline inside a thin aluminium foil package and the package was dried to prevent radioactive contamination when handling the source package; additionally, each source package was used only once for enhanced radiation safety. The sample sandwich was placed inside a hermetically sealed MinigripTM bag and moisturised with the borate (Hilafilcon B, Balafilcon A) or phosphate (Polymacon) buffered saline from the lens packages, keeping the lenses moist during the measurement and preventing instantaneous drying. The drying of the lenses was observed to happen within 12–24 h after placement inside the bag as the saline slowly drained to the bottom of the bag. The PALS measurements were performed in normal room temperature of 20 ◦ C and the duration of each measurement was maximally 3–4 h for each annihilation spectrum (the total measurement time of each sample sandwich was at most 12 h). For studies on the possible filling of the voids during the submersion time in artificial tears (water–bovine serum albumin (Sigma–Aldrich) solution), Balafilcon A (PV) CLs were used based on the results of the preliminary studies: the voids in the Balafilcon A (PV) CLs were large enough to produce an o-Ps lifetime component that could be separated from the o-Ps annihilating in water with high certainty. All CLs
4
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Table 1 PALS results of different CL polymers, measured with measurement setup with time resolution of ∼550 ps. Separation of H2 O and polymer lifetime was not possible for the first two materials. Lens type
3 (ns)
I3 (%)
4 (ns)
I4 (%)
Hilafilcon B (DD) Polymacon (SL) Balafilcon A (PV)
2.12 ± 0.05 2.13 ± 0.04 1.80 ± 0.05 (H2 O)
15.4 ± 0.6 13.9 ± 0.2 10.6 ± 0.5
– – 3.40 ± 0.06
9.5 ± 0.5
Table 2 Calculated void sizes and void volume in each CL type based on Eq. (1). Calculations for Hilafilcon B (DD), Polymacon (SL) are irrelevant an sich as the measured lifetime is a sum of the lifetime in CL polymer and water due to poor time resolution (∼550 ps) of the setup, however upper limit estimates for the results for void sizes are presented in brackets based on the estimate that the polymer dependant lifetime component should be at maximum 2.5 ns not to be separated from water with our setup. Based on additional measurements on Hilafilcon B (DD)-lenses, the actual polymer lifetime cannot be far from the resulting 2.12 ns as the separation of 3 (water) and 4 was not possible even with very high time resolution (260 ps). Sample
o-Ps lifetime (ns)
d (nm)
VVoid (nm3 )
Hilafilcon B (DD), low res Polymacon (SL) Balafilcon A (PV), measured
2.12 ± 0.05 2.13 ± 0.03 3.40 ± 0.06
0.59 ± 0.02 (0.66 ± 0.02) 0.59 ± 0.02 (0.66 ± 0.02) 0.78 ± 0.06
0.11 ± 0.01 (0.15 ± 0.01) 0.11 ± 0.01 (0.15 ± 0.01) 0.25 ± 0.02
were stored in their original packaging immersed in a solution of borate-buffered saline prior to experiments. 4. Results The lifetime results obtained from the three types of CLs are presented in Table 1. Values for 1 and 2 (typically ∼0.2 ns and 0.4–1.0 ns with intensities ∼25% and ∼50%, respectively) are not presented as they provide no relevant information on the void size distributions in the CLs. All lifetime results are based on the average value of the distribution calculated by MELT and subsequently averaged over three separate measurements. In the Hilafilcon B (DD) and Polymacon (SL)-lenses the separation of o-Ps lifetimes in water and polymer itself cannot be made with the time resolution of the used experimental setup. The measured lifetime for Polymacon (SL) is different from ∼1.9 ns by Deepa and Ranganathaiah [5], but it should be noted that in that work the measurement geometry and details were not described preventing direct comparison. In the Balafilcon A (PV)-lens the polymer lifetime is significantly longer than the water-lifetime and thus it was possible to separate the lifetime components. In addition, the o-Ps lifetimes in the Balafilcon A (PV)-lens are much longer than in the other studied lenses. We emphasize that the intensity data cannot be taken as a measure of the density of the voids due to the nonlinear positronium formation and pick-off annihilation processes [11], and hence we refrain from making estimates on the total free volume in the CLs. Based on the experience on PALS, one can estimate that the corresponding polymer lifetime component ( 4 ) for Hilafilcon B (DD) and Polymacon (SL)-lenses are at most ∼2.5 ns as a component longer than that would be separable from the o-Ps lifetime in water with the used equipment. Table 2 presents the estimated void diameters (based on Eq. (1)) and volumes in three kinds of lenses. As the lifetime separation was not possible in the Hilafilcon B (DD) and Polymacon (SL) lenses, we present both the estimate based on the measured lifetime component ( 3 ) and the upper limit estimate for an assumed lifetime of 2.5 ns (in parentheses in the table). We performed lifetime experiments also with a setup with a narrow time resolution on the Hilafilcon B (DD) CLs, in order to find whether a longer polymer-specific lifetime component ( 4 ) could be separated similarly as in the Balafilcon A (PV)-lenses. Interestingly, we obtained 2.02 ± 0.03 ns for an average of three spectra – i.e., even with very narrow time resolution (FWHM 260 ps) the separation of 3 and 4 was not possible. This indicates that the o-Ps lifetime in these CL polymers is not much different from that in pure water (1.8 ns), even though still definitely longer than the lifetime in pure water. This could be due to the fact that Hilafilcon B
(DD) lenses have very high water content and thus most of the free voids are already filled with water and the fraction of open voids in the polymer is limited. This would result in a low yield (i.e., relative intensity) for 4 which would make the separation of 3 and 4 harder. However, the difference in the void sizes in the Balafilcon A (PV) is about 2-times larger by volume compared to the other lenses. In order to study the possible filling of the voids during CL use, Balafilcon A (PV) CLs were submerged into albumin solution (12.5 mg/ml in water). The submerged samples were stored in a normal office room at room temperature in normal lighting conditions, resulting to approximately 12/12 h of lighting and darkness. The sample container lid was kept closed and the container was shaken on a daily basis to prevent precipitation of the proteins in the solution. The lenses (5 + 5 lenses in each case) were removed from the solution with tweezers and measured after 2–5 min, 1–4 weeks of submersion in the protein solution. The results from these measurements are shown in Fig. 1. The annihilation spectra from the ‘2–5 min’ and ‘3 weeks’ submersion periods are plotted together in Fig. 2. The difference in the long lifetime-region (>5 ns) is clearly visible in the spectra demonstrating that the difference in the fitted lifetime components represents a real physical difference. The aging test was also performed on Hilafilcon B (DD) CLs, where a batch of 5 + 5 lenses was submerged into albumin solution for 10 days. However no significant changes were observed
Fig. 1. Measured lifetimes and corresponding void diameters in Balafilcon A (PV) CLs submerged into 12.5 mg/ml aqueous albumin solution. The intensity of the o-Ps lifetime component was in the 10–15% range in all measurements. The first data points (without albumin immersion and after 2–5 min incubation in albumin solution) coincide, they are plotted with a small horizontal shift for visual purposes. Overall void volume change between, i.e., reference point and 3 weeks of submersion is ∼25%.
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Fig. 2. Positron lifetime spectra (background reduced) from Balafilcon A (PV) CLs submerged for 2–5 min (crosses) and 3 weeks (circles) in albumin solution (12.5 mg/ml). The difference in the lifetimes is clearly visible as longer lifetimes are more abundant in the ‘2–5 min’ spectrum compared to the data from ‘3 weeks’. The lifetime components i are found from the spectrum by fitting slopes to the spectrum, such slope corresponds to a specific decay rate of positron/positronium and the lifetime component is defined as an inverse of the decay rate.
in the o-Ps lifetimes after submersion (2.06 ± 0.05 ns) compared to the reference (2.02 ± 0.03 ns). As the Hilafilcon B (DD)-lenses are disposable single-use lenses (usage time <24 h), performing more extensive aging experiments seemed unnecessary. 5. Discussion Based on the positron experiments, it seems that the void size distributions in the studied contact lenses correlate well with the usage recommendations: the smaller the average void size, the shorter is the recommended time of use. However, it remains unresolved whether the differences in void volumes between, e.g., Hilafilcon B (DD) and Balafilcon A (PV) CLs are significant to the overall oxygen diffusion through the lens, even though the difference in the void sizes is clear. In the work of Singh et al. [4], oxygen diffusion and o-Ps lifetimes in different CL polymers were observed to exhibit a clear linear correlation. O2 -permeability1 values in the study at 35 ◦ C were measured from 32.8, corresponding to an o-Ps lifetime of 2.36 ns, to 109.0 for lifetime of 2.80 ns. If one uses Eq. (1) to estimate the void sizes for these two values, resulting in volumes of 0.13 nm3 and 0.17 nm3 , respectively, one can estimate that a decrease of ∼25% in the void volume reduces the O2 permeability by a factor of 3. Hence it seems likely that the reduction of the o-Ps lifetime and the void sizes in the aging experiment induces a reduction in oxygen diffusion through the lens. Other factors than the free volume available in the voids need be taken into account when comparing different lenses. The permeability of other solvents, i.e., water (discussed more thoroughly in Ref. [12]) plays an important role in the overall diffusion through the lens. If the water solubility in the lens material is high, water can flow through the CL and fill the voids, decreasing the possibility of gas diffusion though the lens. As all the CLs were stored in aqueous environment prior to the experiments, it is unlikely that the effect of excess water molecules (filling the voids from the base hydration level) would be a significant source of error. Another possible
1
Gas permeability P
cm3 (O2 (STP) cm3 polymer·mm Hg
cm2 s
× 10−11 .
5
uncertainty of the data interpretation is the fact the experimental conditions were not completely identical to the usage environment where the other side of the lens is constantly exposed to air and thus the hydration level may be different from that in our positron measurements. However, as the experimental conditions were kept as close to identical as possible for all the different lenses, we consider the relative behaviour of the lenses to be correctly interpreted. We found that the albumin in the solution either fills the voids or causes decrease in the void size in the silicone–hydrogel type Balafilcon A CL (PV) over a time period of weeks but such effect was not observed in polymer-based Hilafilcon B CL (DD). This decrease of void size most likely reduces the oxygen diffusion through the lens. Based on the present data, it is not possible to state whether the voids are actually filled with denatured proteins or is the observed effect due to a change in the polymer’s structural dynamics (the protein chains could change the packing ratios of the polymer chains and cause an increase of the material density of the structure when albumin permeates the polymer structure, reducing the void size). The effect of change of environment from borate-buffered saline in lens packages to the albumin–water solution remains an open question. However as the difference in lifetimes is small between the reference sample (where such change of environment did not occur) and samples incubated 2–5 min and 1 week it would suggest that the effect is quite insignificant on the microstructural level of the CL. In any case, the conclusion remains: the average size of the free volume pockets in the polymer structure decreases when subjected to usage-mimicking conditions. This is likely to result in a lower oxygen diffusion rate through the lens. Further studies are planned to see whether the decrease in the void volume continues after the time period of ∼3 to 4 weeks, where a hint of stabilization is observed. The reason why no such effect was observed in Hilafilcon B (DD) CLs can be due to many factors, one of course being the larger hydration level of the lenses keeping the voids filled with water during the incubation, other maybe the sensibility limitation of the method. The effect of the change on lens environment from borate-buffered storage liquid to water–albumin solution should be also taken into account, yet based on the comparison of reference Balafilcon A (PV)-lens measured in environment of buffering solution and the identical CL-sample incubated 2–5 min in water/albumin solution the difference seems to be negligible. However long-term effects could exist but whether such effects are relevant in the void size decrease during the incubation remains unknown. More interestingly, when comparing the data from disposable Hilafilcon B and monthly Balafilcon A lenses, it is clear that even after a 4 weeks of submersion in albumin solution the void properties in Balafilcon A (PV) CLs are still superior to disposable lenses, indicating that disposable CLs are inferior regarding to oxygen diffusion properties and proper monthly lenses would seem to provide extensive oxygen diffusion throughout the usage period. This is an important notion due to the large difference in price between daily disposable and continuous use lenses. Hilafilcon B (DD) and Polymacon (SL) CLs users might be tempted to over-extend the use of these lenses, or instead of monthly lenses use cheaper disposable or semi-disposable CLs as monthly CLs. In conclusion this study shows three important aspects. Firstly, PALS is a viable method to study and provide quantitative analysis of the void sizes in actual CLs. Secondly, submersion of the CLs into artificial tear fluid decreases the void sizes in Balafilcon A (PV) lenses. This decrease in the free volume can be quantified. PALS seem to offer a sensitive and simple method to assess the free volume properties and thus also oxygen diffusion properties of contact lenses as well as in other polymer-based materials. Thirdly, PALS can be used to probe the microstructure of any polymer material in which diffusion of molecules is crucial for its function. Especially the method should be very valuable for corneal inlays and onlays.
6
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Acknowledgment This work was supported by Academy of Finland (P.S., F.T.). References [1] Jean YC, Mallon PE, Schrader DM. Principles and applications of positron & positronium chemistry. World Scientific; 2003. [2] Sane P, Salonen E, Falck E, Repakova J, Tuomisto F, Holopainen JM, et al. Probing biomembranes with positrons. Journal of Physical Chemistry B 2009;113:1810–2. [3] Sane P, Tuomisto F, Wiedmer Susanne K, Nyman T, Vattulainen I, Holopainen Juha M. Temperature-induced structural transition in situ in porcine lenschanges observed in void size distribution. Biochimica et Biophysica Acta (BBA) – Biomembranes 2010;1798(5):958–65. [4] Singh JJ, Eftekhari A, Upchurch BT, Burns KS. An investigation on microstructural characteristics of contact lens polymers. NASA technical paper; 1990. p. 3034.
[5] Deepa MV, Ranganathaiah C. Influence of spoliation in poly(2-hydroxyethylmethacrylate) soft contact lens on its free volume and optical transparency. Journal of Materials Science: Materials in Medicine 2007. [6] Eldrup M, Lightbody D, Sherwood JN. The temperature dependence of positron lifetimes in solid pivalic acid. Chemistry and Physics 1981;63:51–8. [7] Nakanishi H. In: Sharma SC, editor. Proceedings for the international conference on positron annihilation in fluids, Arlington, TX. Singapore: World Scientific Publishing; 1987. [8] Sane P, Kilpeläinen S, Tuomisto F. 4-Channel digital positron lifetime spectrometer for studying biological samples. Materials Science Forum 2009;607:254–6. [9] Olsen JV, Kirkegaard P, Pedersen NJ, Eldrup M. Palsfit: a new program for the evaluation of positron lifetime spectra. Physica Status Solidi C 2007;4:4004–6. [10] Shukla A, Peter M, Hoffman L. Analysis of positron lifetime spectra using quantified maximum entropy and a general linear filter. Nuclear Instruments and Methods in Physics Research A 1993;335:310–7. [11] Mogensen OE. Positron annihilation in chemistry. Heidelberg: Springer-Verlag; 1995. [12] Fornasiero F, Krull F, Prausnitz JM, Radke CJ. Steady-state diffusion of water through soft-contact-lens materials. Biomaterials 2005;26(28):5704–16.
Contents lists available at ScienceDirect
Contact Lens & Anterior Eye journal homepage: www.elsevier.com/locate/clae
Void volume variations in contact lens polymers P. Sane a,∗ , F. Tuomisto a , J.M. Holopainen b a b
Department of Applied Physics, Aalto University, Otakaari 1M, 02150 Espoo, Finland Helsinki Eye Lab, Department of Ophthalmology, University of Helsinki, Finland
a r t i c l e
i n f o
Keywords: Contact lens Positron annihilation spectroscopy Free volume pockets/voids Filling of voids
a b s t r a c t Aim: In this study, void size and free volume properties in different contact lens materials have been investigated in their hydrated state using positron annihilation lifetime spectroscopy (PALS). Methods: PALS is used to characterize the void size distributions inside the lens materials. Three different types of contact lenses were used (Balafilcon A, Hilafilcon B and Polymacon). Results: Measurements on different contact lenses reveal significant differences between the materials, up to ∼100% difference in void volume was observed between Hilafilcon B and Balafilcon A, the latter having larger voids. As oxygen diffusion is strongly correlated with the void sizes, the results are in good agreement with the usage recommendations of the specific lens types (daily disposable lenses or 1 month continuous use lenses). The void sizes in monthly lenses (Balafilcon A) were found to decrease 25% under artificial aqueous tear (albumin–water solution) exposure in 4 weeks leading to a significant decrease in the oxygen permeation rate through the contact lens. Yet, the voids were still significantly larger than in disposable or semi-disposable lenses. Conclusions: We have showed that PALS is a viable method to probe the microstructure of biotechnologically relevant polymers and can be used to quantify the void properties in different types of contact lenses. Usage recommendations correlate well with measured void sizes and the median void size decreases during the incubation of albumin solution as a function of time. We anticipate the use of PALS for any polymer-based intracorneal/intraocular device in which diffusivity plays a crucial role. © 2010 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
1. Introduction The local free volume of a polymer structure (void), corresponding to the unoccupied regions accessible to segmental motions, is an important parameter in the overall diffusion of atoms and molecules, either in gaseous or solid states, through the polymer. Positron annihilation lifetime spectroscopy (PALS) has been widely used to characterize atomic scale defects in semiconductors and metals and is also routinely used to study void distributions in polymer materials and in characterizing structural parameters such as the glass transition temperatures [1]. Lately we have employed this method to study also biological materials, such as lipid bilayers [2] as well as intact crystalline intraocular lenses [3]. Diffusion of molecules is largely determined by the presence and volume of voids in any material, and hence PALS can give direct information of the diffusion properties of the material in question. Oxygen diffusion is one of the most important factors in contact lens (CL) materials. Typically the permeation of oxygen through the CL is much less efficient in soft disposable CLs compared to
∗ Corresponding author. Tel.: +358 947023145. E-mail address: petri.sane@tkk.fi (P. Sane).
harder CLs designed for daily long term use. In an earlier study by Singh et al. [4] where PALS and measurements of gas diffusion rates were compared, it was shown that oxygen diffusion through CLs was strongly dependent on the void sizes. This effect was explained by the fact that gas permeability is the product of solubility and diffusion coefficient, causing the permeability to change with free volume. In another PALS study on CLs, Deepa and Ranganathaiah [5] reported that calcification of CLs reduced the void volume of these lenses, as the calcium deposits filled the voids. The focus of that work was, however, on the optical properties of the CLs before and after calcification. Both of these studies show that PALS is applicable to study ready-touse CLs (instead of representative polymer samples) of different types. Different types of CLs have different specifications on how long these should be used continuously. This is partly dependent in the oxygen permeability of each CL polymer. Furthermore, the contamination of CL by proteins, ions, and impurities present in the tear fluid will eventually reduce the diffusion properties of the CL by filling the voids within the polymer. In this work we have studied the void size distributions (and hence the diffusion characteristics) of different types of CLs. We also exposed different types of CLs to aqueous albumin solution and followed changes in the void volume as a function of time.
1367-0484/$ – see front matter © 2010 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clae.2010.06.008
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
2. Positron annihilation lifetime spectroscopy in polymers When an energetic positron from a radioactive source (22 Na in most cases) enters molecular media, it thermalizes rapidly after interacting with the surrounding molecules through inelastic collisions. A fraction of positrons forms a bound state with spin parallel electrons (ortho-positronium, o-Ps) that preferentially localizes in the free volume pockets in the material. In vacuum, o-Ps has a relatively long lifetime of 142 ns, compared to 125 ps for p-Ps (bound state of a positron and an electron with opposite spins). In a medium, the o-Ps prefers to undergo so-called pick-off annihilation with an electron of opposite spin during collision with molecules in the cavity wall in which it is localized. The pick-off process reduces the lifetime of o-Ps down to a few nanoseconds. The smaller the cavity size, the higher the frequency of collisions and the shorter the o-Ps lifetime, hence the o-Ps lifetime provides information of the size of the free volume pockets (i.e., voids). A quantitative semi-empirical relation has been established by Tao and Eldrup [6] (work on positronium trapping to voids in polymers) correlating the o-Ps lifetime ( polymer ) with the radius (R) of a spherical trap: polymer (ns) =
1 R 1 1− + sin 2 R+C 2
2R −1 R+C
,
(1)
where C is a semi-empirical constant of 0.166 nm [7]. The model applies directly to spherical voids and as such it is not directly applicable to characterize accurately the void sizes in CLs, as the specific form and distribution of forms of the voids is not known. However, the Tao–Eldrup model can be used to provide an orderof-magnitude estimate and means to compare the void sizes in polymers. 3. Experimental methods 3.1. Instrumentation The PALS measurements were performed using a sandwichgeometry of the sample material and the positron source. The positron/positronium lifetime is measured with a pair of scintillation detectors, one detecting the birth of a positron and the other its annihilation. When the positron source (22 Na) decays, it produces simultaneously one positron and one ∼1.27 MeV ␥quantum. When this quantum denoting the birth of the positron is detected, the pulse acts as a START-signal for the equipment. When the positron annihilates in the sample, as a free positron, p-Ps or o-Ps (pick-off) it produces two ␥-quanta of about 511 keV each, either of which act as the STOP signal. The time difference of START- and STOP-pulses gives the positron lifetime () in the material. In this study the annihilation spectrum was measured with a normal digital lifetime setup, consisting of two scintillation detectors with large scintillation heads producing a broad time resolution (full-width at half-maximum, FWHM, of the Gaussian resolution function ∼550 ps for the setup for preliminary measurements, ∼320 ps/260 ps for the further studies described later on) but with high efficiency. A broad time resolution is not optimal for lifetime separation, but as the measurements were planned to be performed in the hydrated state, the short duration of measurement was an important factor: the lenses were placed inside a sealed plastic bag, and the liquid (borate-buffered saline from the lens containers) drained slowly from the lenses to the bottom of the bag due to the horizontal geometry of the measurement setup. The setup is characterized in more detail in Ref. [8]. The lifetime spectra were analyzed first with PALSfit [9] to analyze the properties of the time resolution of the setup and then analyzed with MELT [10] in order to minimize possible fitting problems due to continuous void size distributions, the mathematical
3
basis of the aforementioned tools are presented in the references. Statistics of ∼2 to 3 million events were collected in each annihilation spectrum and the analysis was performed using a free fit with no source corrections resulting to 3–4 lifetime components depending on the sample and whether the water-lifetime ( 3 ) could be separated from the polymer lifetime ( 4 ), the two shortest components representing the annihilation of free positrons and p-Ps ( 1 ) as well as a ∼1 ns long o-Ps component ( 2 ). Goodness of the fits was evaluated with the 2 -output of the analysis software and similar range of values (1.1–1.2) was observed in all cases. 3.2. Contact lens materials and their measurement Preliminary experiments were performed with three different types of lenses, all from the same manufacturer, (Bausch & Lomb, B&L). The purpose of these experiments was to characterize the differences of lifetimes in different types of CLs. The studied lenses were: B&L Daily Disposable (DD, 59% water), B&L Soflens® (SL, 38%), and B&L Purevision® (PV, 36%). The lenses were chosen to have profound differences in their usage recommendations, ranging from the short term use (DD) to lenses that can be used continuously for a month (PV). The B&L Soflens is a CL that can be reused but is not recommended for continuous use. The base materials in these CLs are polymer-based Hilafilcon B (DD), Polymacon (SL) and a silicone–hydrogel based Balafilcon A (PV). We measured contact lenses in their “in-use” hydrated state, containing reasonable amounts of water (36–59%) and hence some of the o-Ps always annihilated in water. Because the lifetime components of some of the studied polymers were quite near the o-Ps lifetime in water, the measured lifetime component was in most cases an intensity weighted sum of water and polymer lifetimes. This effect is more significant in the samples measured with a PALS setup with broader time resolution. Due to this and other factors (e.g., chemical interactions of o-Ps with the material), the measured o-Ps lifetimes are, as usual in polymers, not discrete, but a distribution. Hence the radius R obtained by using the Tao–Eldrup model (Eq. (1)) should be interpreted as the mean radius of the free volume pockets. A layer consisting of at least five contact lenses was placed on both sides of the small (2 × 2 × 0.003 mm) positron source, surrounding the small source completely and thus the vast majority of positrons annihilated in the lenses based on the estimate of positron range in the CL polymers. The positron sources (approximately 0.5 MBq each) were manufactured by injecting 22 NaCl saline inside a thin aluminium foil package and the package was dried to prevent radioactive contamination when handling the source package; additionally, each source package was used only once for enhanced radiation safety. The sample sandwich was placed inside a hermetically sealed MinigripTM bag and moisturised with the borate (Hilafilcon B, Balafilcon A) or phosphate (Polymacon) buffered saline from the lens packages, keeping the lenses moist during the measurement and preventing instantaneous drying. The drying of the lenses was observed to happen within 12–24 h after placement inside the bag as the saline slowly drained to the bottom of the bag. The PALS measurements were performed in normal room temperature of 20 ◦ C and the duration of each measurement was maximally 3–4 h for each annihilation spectrum (the total measurement time of each sample sandwich was at most 12 h). For studies on the possible filling of the voids during the submersion time in artificial tears (water–bovine serum albumin (Sigma–Aldrich) solution), Balafilcon A (PV) CLs were used based on the results of the preliminary studies: the voids in the Balafilcon A (PV) CLs were large enough to produce an o-Ps lifetime component that could be separated from the o-Ps annihilating in water with high certainty. All CLs
4
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Table 1 PALS results of different CL polymers, measured with measurement setup with time resolution of ∼550 ps. Separation of H2 O and polymer lifetime was not possible for the first two materials. Lens type
3 (ns)
I3 (%)
4 (ns)
I4 (%)
Hilafilcon B (DD) Polymacon (SL) Balafilcon A (PV)
2.12 ± 0.05 2.13 ± 0.04 1.80 ± 0.05 (H2 O)
15.4 ± 0.6 13.9 ± 0.2 10.6 ± 0.5
– – 3.40 ± 0.06
9.5 ± 0.5
Table 2 Calculated void sizes and void volume in each CL type based on Eq. (1). Calculations for Hilafilcon B (DD), Polymacon (SL) are irrelevant an sich as the measured lifetime is a sum of the lifetime in CL polymer and water due to poor time resolution (∼550 ps) of the setup, however upper limit estimates for the results for void sizes are presented in brackets based on the estimate that the polymer dependant lifetime component should be at maximum 2.5 ns not to be separated from water with our setup. Based on additional measurements on Hilafilcon B (DD)-lenses, the actual polymer lifetime cannot be far from the resulting 2.12 ns as the separation of 3 (water) and 4 was not possible even with very high time resolution (260 ps). Sample
o-Ps lifetime (ns)
d (nm)
VVoid (nm3 )
Hilafilcon B (DD), low res Polymacon (SL) Balafilcon A (PV), measured
2.12 ± 0.05 2.13 ± 0.03 3.40 ± 0.06
0.59 ± 0.02 (0.66 ± 0.02) 0.59 ± 0.02 (0.66 ± 0.02) 0.78 ± 0.06
0.11 ± 0.01 (0.15 ± 0.01) 0.11 ± 0.01 (0.15 ± 0.01) 0.25 ± 0.02
were stored in their original packaging immersed in a solution of borate-buffered saline prior to experiments. 4. Results The lifetime results obtained from the three types of CLs are presented in Table 1. Values for 1 and 2 (typically ∼0.2 ns and 0.4–1.0 ns with intensities ∼25% and ∼50%, respectively) are not presented as they provide no relevant information on the void size distributions in the CLs. All lifetime results are based on the average value of the distribution calculated by MELT and subsequently averaged over three separate measurements. In the Hilafilcon B (DD) and Polymacon (SL)-lenses the separation of o-Ps lifetimes in water and polymer itself cannot be made with the time resolution of the used experimental setup. The measured lifetime for Polymacon (SL) is different from ∼1.9 ns by Deepa and Ranganathaiah [5], but it should be noted that in that work the measurement geometry and details were not described preventing direct comparison. In the Balafilcon A (PV)-lens the polymer lifetime is significantly longer than the water-lifetime and thus it was possible to separate the lifetime components. In addition, the o-Ps lifetimes in the Balafilcon A (PV)-lens are much longer than in the other studied lenses. We emphasize that the intensity data cannot be taken as a measure of the density of the voids due to the nonlinear positronium formation and pick-off annihilation processes [11], and hence we refrain from making estimates on the total free volume in the CLs. Based on the experience on PALS, one can estimate that the corresponding polymer lifetime component ( 4 ) for Hilafilcon B (DD) and Polymacon (SL)-lenses are at most ∼2.5 ns as a component longer than that would be separable from the o-Ps lifetime in water with the used equipment. Table 2 presents the estimated void diameters (based on Eq. (1)) and volumes in three kinds of lenses. As the lifetime separation was not possible in the Hilafilcon B (DD) and Polymacon (SL) lenses, we present both the estimate based on the measured lifetime component ( 3 ) and the upper limit estimate for an assumed lifetime of 2.5 ns (in parentheses in the table). We performed lifetime experiments also with a setup with a narrow time resolution on the Hilafilcon B (DD) CLs, in order to find whether a longer polymer-specific lifetime component ( 4 ) could be separated similarly as in the Balafilcon A (PV)-lenses. Interestingly, we obtained 2.02 ± 0.03 ns for an average of three spectra – i.e., even with very narrow time resolution (FWHM 260 ps) the separation of 3 and 4 was not possible. This indicates that the o-Ps lifetime in these CL polymers is not much different from that in pure water (1.8 ns), even though still definitely longer than the lifetime in pure water. This could be due to the fact that Hilafilcon B
(DD) lenses have very high water content and thus most of the free voids are already filled with water and the fraction of open voids in the polymer is limited. This would result in a low yield (i.e., relative intensity) for 4 which would make the separation of 3 and 4 harder. However, the difference in the void sizes in the Balafilcon A (PV) is about 2-times larger by volume compared to the other lenses. In order to study the possible filling of the voids during CL use, Balafilcon A (PV) CLs were submerged into albumin solution (12.5 mg/ml in water). The submerged samples were stored in a normal office room at room temperature in normal lighting conditions, resulting to approximately 12/12 h of lighting and darkness. The sample container lid was kept closed and the container was shaken on a daily basis to prevent precipitation of the proteins in the solution. The lenses (5 + 5 lenses in each case) were removed from the solution with tweezers and measured after 2–5 min, 1–4 weeks of submersion in the protein solution. The results from these measurements are shown in Fig. 1. The annihilation spectra from the ‘2–5 min’ and ‘3 weeks’ submersion periods are plotted together in Fig. 2. The difference in the long lifetime-region (>5 ns) is clearly visible in the spectra demonstrating that the difference in the fitted lifetime components represents a real physical difference. The aging test was also performed on Hilafilcon B (DD) CLs, where a batch of 5 + 5 lenses was submerged into albumin solution for 10 days. However no significant changes were observed
Fig. 1. Measured lifetimes and corresponding void diameters in Balafilcon A (PV) CLs submerged into 12.5 mg/ml aqueous albumin solution. The intensity of the o-Ps lifetime component was in the 10–15% range in all measurements. The first data points (without albumin immersion and after 2–5 min incubation in albumin solution) coincide, they are plotted with a small horizontal shift for visual purposes. Overall void volume change between, i.e., reference point and 3 weeks of submersion is ∼25%.
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Fig. 2. Positron lifetime spectra (background reduced) from Balafilcon A (PV) CLs submerged for 2–5 min (crosses) and 3 weeks (circles) in albumin solution (12.5 mg/ml). The difference in the lifetimes is clearly visible as longer lifetimes are more abundant in the ‘2–5 min’ spectrum compared to the data from ‘3 weeks’. The lifetime components i are found from the spectrum by fitting slopes to the spectrum, such slope corresponds to a specific decay rate of positron/positronium and the lifetime component is defined as an inverse of the decay rate.
in the o-Ps lifetimes after submersion (2.06 ± 0.05 ns) compared to the reference (2.02 ± 0.03 ns). As the Hilafilcon B (DD)-lenses are disposable single-use lenses (usage time <24 h), performing more extensive aging experiments seemed unnecessary. 5. Discussion Based on the positron experiments, it seems that the void size distributions in the studied contact lenses correlate well with the usage recommendations: the smaller the average void size, the shorter is the recommended time of use. However, it remains unresolved whether the differences in void volumes between, e.g., Hilafilcon B (DD) and Balafilcon A (PV) CLs are significant to the overall oxygen diffusion through the lens, even though the difference in the void sizes is clear. In the work of Singh et al. [4], oxygen diffusion and o-Ps lifetimes in different CL polymers were observed to exhibit a clear linear correlation. O2 -permeability1 values in the study at 35 ◦ C were measured from 32.8, corresponding to an o-Ps lifetime of 2.36 ns, to 109.0 for lifetime of 2.80 ns. If one uses Eq. (1) to estimate the void sizes for these two values, resulting in volumes of 0.13 nm3 and 0.17 nm3 , respectively, one can estimate that a decrease of ∼25% in the void volume reduces the O2 permeability by a factor of 3. Hence it seems likely that the reduction of the o-Ps lifetime and the void sizes in the aging experiment induces a reduction in oxygen diffusion through the lens. Other factors than the free volume available in the voids need be taken into account when comparing different lenses. The permeability of other solvents, i.e., water (discussed more thoroughly in Ref. [12]) plays an important role in the overall diffusion through the lens. If the water solubility in the lens material is high, water can flow through the CL and fill the voids, decreasing the possibility of gas diffusion though the lens. As all the CLs were stored in aqueous environment prior to the experiments, it is unlikely that the effect of excess water molecules (filling the voids from the base hydration level) would be a significant source of error. Another possible
1
Gas permeability P
cm3 (O2 (STP) cm3 polymer·mm Hg
cm2 s
× 10−11 .
5
uncertainty of the data interpretation is the fact the experimental conditions were not completely identical to the usage environment where the other side of the lens is constantly exposed to air and thus the hydration level may be different from that in our positron measurements. However, as the experimental conditions were kept as close to identical as possible for all the different lenses, we consider the relative behaviour of the lenses to be correctly interpreted. We found that the albumin in the solution either fills the voids or causes decrease in the void size in the silicone–hydrogel type Balafilcon A CL (PV) over a time period of weeks but such effect was not observed in polymer-based Hilafilcon B CL (DD). This decrease of void size most likely reduces the oxygen diffusion through the lens. Based on the present data, it is not possible to state whether the voids are actually filled with denatured proteins or is the observed effect due to a change in the polymer’s structural dynamics (the protein chains could change the packing ratios of the polymer chains and cause an increase of the material density of the structure when albumin permeates the polymer structure, reducing the void size). The effect of change of environment from borate-buffered saline in lens packages to the albumin–water solution remains an open question. However as the difference in lifetimes is small between the reference sample (where such change of environment did not occur) and samples incubated 2–5 min and 1 week it would suggest that the effect is quite insignificant on the microstructural level of the CL. In any case, the conclusion remains: the average size of the free volume pockets in the polymer structure decreases when subjected to usage-mimicking conditions. This is likely to result in a lower oxygen diffusion rate through the lens. Further studies are planned to see whether the decrease in the void volume continues after the time period of ∼3 to 4 weeks, where a hint of stabilization is observed. The reason why no such effect was observed in Hilafilcon B (DD) CLs can be due to many factors, one of course being the larger hydration level of the lenses keeping the voids filled with water during the incubation, other maybe the sensibility limitation of the method. The effect of the change on lens environment from borate-buffered storage liquid to water–albumin solution should be also taken into account, yet based on the comparison of reference Balafilcon A (PV)-lens measured in environment of buffering solution and the identical CL-sample incubated 2–5 min in water/albumin solution the difference seems to be negligible. However long-term effects could exist but whether such effects are relevant in the void size decrease during the incubation remains unknown. More interestingly, when comparing the data from disposable Hilafilcon B and monthly Balafilcon A lenses, it is clear that even after a 4 weeks of submersion in albumin solution the void properties in Balafilcon A (PV) CLs are still superior to disposable lenses, indicating that disposable CLs are inferior regarding to oxygen diffusion properties and proper monthly lenses would seem to provide extensive oxygen diffusion throughout the usage period. This is an important notion due to the large difference in price between daily disposable and continuous use lenses. Hilafilcon B (DD) and Polymacon (SL) CLs users might be tempted to over-extend the use of these lenses, or instead of monthly lenses use cheaper disposable or semi-disposable CLs as monthly CLs. In conclusion this study shows three important aspects. Firstly, PALS is a viable method to study and provide quantitative analysis of the void sizes in actual CLs. Secondly, submersion of the CLs into artificial tear fluid decreases the void sizes in Balafilcon A (PV) lenses. This decrease in the free volume can be quantified. PALS seem to offer a sensitive and simple method to assess the free volume properties and thus also oxygen diffusion properties of contact lenses as well as in other polymer-based materials. Thirdly, PALS can be used to probe the microstructure of any polymer material in which diffusion of molecules is crucial for its function. Especially the method should be very valuable for corneal inlays and onlays.
6
P. Sane et al. / Contact Lens & Anterior Eye 34 (2011) 2–6
Acknowledgment This work was supported by Academy of Finland (P.S., F.T.). References [1] Jean YC, Mallon PE, Schrader DM. Principles and applications of positron & positronium chemistry. World Scientific; 2003. [2] Sane P, Salonen E, Falck E, Repakova J, Tuomisto F, Holopainen JM, et al. Probing biomembranes with positrons. Journal of Physical Chemistry B 2009;113:1810–2. [3] Sane P, Tuomisto F, Wiedmer Susanne K, Nyman T, Vattulainen I, Holopainen Juha M. Temperature-induced structural transition in situ in porcine lenschanges observed in void size distribution. Biochimica et Biophysica Acta (BBA) – Biomembranes 2010;1798(5):958–65. [4] Singh JJ, Eftekhari A, Upchurch BT, Burns KS. An investigation on microstructural characteristics of contact lens polymers. NASA technical paper; 1990. p. 3034.
[5] Deepa MV, Ranganathaiah C. Influence of spoliation in poly(2-hydroxyethylmethacrylate) soft contact lens on its free volume and optical transparency. Journal of Materials Science: Materials in Medicine 2007. [6] Eldrup M, Lightbody D, Sherwood JN. The temperature dependence of positron lifetimes in solid pivalic acid. Chemistry and Physics 1981;63:51–8. [7] Nakanishi H. In: Sharma SC, editor. Proceedings for the international conference on positron annihilation in fluids, Arlington, TX. Singapore: World Scientific Publishing; 1987. [8] Sane P, Kilpeläinen S, Tuomisto F. 4-Channel digital positron lifetime spectrometer for studying biological samples. Materials Science Forum 2009;607:254–6. [9] Olsen JV, Kirkegaard P, Pedersen NJ, Eldrup M. Palsfit: a new program for the evaluation of positron lifetime spectra. Physica Status Solidi C 2007;4:4004–6. [10] Shukla A, Peter M, Hoffman L. Analysis of positron lifetime spectra using quantified maximum entropy and a general linear filter. Nuclear Instruments and Methods in Physics Research A 1993;335:310–7. [11] Mogensen OE. Positron annihilation in chemistry. Heidelberg: Springer-Verlag; 1995. [12] Fornasiero F, Krull F, Prausnitz JM, Radke CJ. Steady-state diffusion of water through soft-contact-lens materials. Biomaterials 2005;26(28):5704–16.