Friction and Lubrication of Eye/Lens/Lid Interface: The Effect of Lubricant and Contact Lens Material

Irritation of the eye in the case of contact lens users may imply dry eye syndrome (DES), commonly treated by eye drops. The paper focuses on measuring the friction coefficient with the simultaneous observation of lubricating film formation of the simulated eye/lens/lid interface. The experiments are performed in a pin-on-plate setup under the linear reciprocal motion. The contact is lubricated by twelve different solutions as the potential ingredients for eye drops, focusing on the role of mucin, buffer type, and molecular weight (MW) of hyaluronic acid (HA). Two commercially available contact lenses are studied, while overall lens performance is evaluated, among others. The results showed that the lens material, the presence of mucin, and the type of buffer have a decisive influence on the contact behavior. Specifically, it is concluded that the selection of eye drops should reflect the type of lens the wearers use and the conditions of their tear film. The appropriate ingredient for Acuvue Oasys and Biofinity contact lenses is suggested based on the score of the biotribological performance of the lubricant, taking both friction and lubrication into account.


Introduction
Millions of contact lens wearers worldwide suffer from the discomfort associated with lens usage.Specifically, it is reported that approximately 50% of contact lens users have some difficulties with dry eye syndrome (DES) [1].DES was specified by the Tear Film and Ocular Surface Society as a "multifactorial disease of the ocular surface characterized by a loss of homeostasis of the tear film and accompanied by 118 Page 2 of 18 ocular symptoms, in which tear film instability and hyperosmolarity, ocular surface inflammation and damage, and neurosensory abnormalities play etiological roles" [2].The disease has several outcomes, such as deterioration of visual sense, tear film instability, and feeling of eye discomfort.The primary treatment is in the application of eye drops.The market offers a variety of products; however, the understanding of tribological processes in the eye/lid or eye-lens contact is yet to be understood entirely.Laboratory in vitro testing may provide a good insight into the topic while investigating friction and lubrication mechanisms should be of primary interest [3].It is expected that the essential treatment role is in the presence of HA, the importance of which was already clearly proven not only in eye lubrication [3] but also in other biological environments, such as joints cartilage [4,5], fascia tissue [6], catheters [7,8] or hip and knee joint replacements [9,10].
One of the pilot investigations was carried out by Nairn and Jiang [11].The authors developed a pin-on-disk tribometer to investigate the lubrication and friction of contact lenses.Attention was paid to the applied experimental conditions to adequately mimic eye blinking.Therefore, the experiments were performed at high speeds under low loads.Both sides of contact lenses were studied while sliding against different disk materials (i.e., PMMA, polyhydroxy-ethyl-methacrylate, and polycarbonate), while the contact was lubricated by various ophthalmic solutions.It was concluded that, in most cases, the contact operates under a boundary lubrication regime.In that case, the thin film does not provide sufficient separation of rubbing surfaces, which leads to a higher friction level.Friction coefficients reported in the study varied between 0.05 and 0.65.As these values are quite high with significant variance, it is evident that the contact behavior is very sensitive regarding the applied material and lubricant.This work put the basis for most of the upcoming references.Reciprocating tests performed in the pin-on-plate arrangement were introduced by Ngai et al. [12].The authors placed the lens on the silicone eye model, which was stationary.The opposing glass plate motioned while the contact zone was lubricated using saline solution.The focus of the study was on the difference in friction between the conventional hydrogel and silicone hydrogel lenses.The results in friction were found to be statistically insignificant.
The investigation was followed by Rennie et al. [13], who used a reciprocating pin-on-sphere micro-tribometer for friction analysis of hydrogel lenses.The pin was represented by a borosilicate sphere, sliding against the lens mounted on the curved polymer base, which underwent reciprocating motion.The experimental setup was limited by contact pressure, which was about an order larger than the physiological level.The authors additionally developed a numerical predictive model which included all the friction-influencing factors: the hydrogel viscoelastic nature, viscous shearing of solution used for packaging, and shear at the interface of the glass pin and the lens.The measured friction coefficient was less than 0.1, while the predicted film thickness ranged between 1 and 30 nm for the given conditions.Concerning the surface roughness, it was confirmed that the boundary lubrication regime occurs.An advanced numerical model focused on the prediction of the lubrication performance of the contact lens over the blinking period was presented by Dunn et al. [14].At the highest considered speed (100 mm/s), the calculated lens deflection was only 0.5%.The analysis showed a hydrodynamic effect between the surface of the contact lens and the lid wiper.However, a boundary lubrication regime is expected between the lens base and corneal epithelium.It is assumed that this fact, together with slower motions than was the considered maximum, contribute to the lower comfort of lens wearers.
Roba et al. [15] aimed to introduce a biologically valid experimental methodology considering physiological conditions for friction monitoring of contact lenses.The authors tested a variety of commercial contact lenses, focusing on the composition of the lens-counterface material, solution of the lubricant, applied load, and velocity.Measurements were carried out in a similar configuration as in the previous study.The lens were mounted on a convex polymer, while the disk was used as the counter body.The best combination of experimental conditions resulted in low sliding speed, glass coated by mucin, and the application of a solution containing lysosome and serum.Sterner et al. [16] also adopted the same experimental configuration and correlated friction measurements with the clinical comfort of soft contact lenses for the first time.A modification was in implementing a fluorescent optical method for contact zone observation.
The importance of mucin pronounced above [15] was later presented by An et al. [17].Friction measurements were conducted using atomic force microscopy, investigating the interaction of PMMA surfaces coated by the two types of mucin.Although the paper rather fits the scope of biochemistry, the important finding and implication for the present investigation are that both the examined mucins positively impacted frictional behavior.The findings were subsequently extended by Sterner et al. [18].The study showed that PVP represents another outstanding boundary lubricant.The friction reduction was also reported for HA [18,19].Furthermore, the impact of HA on tear film thickness was later examined by Kaya et al. [20].The tests with sixteen adult subjects showed that a single HA dose might increase tear film for up to 30 min.Therefore, the role of lubricant constituents and their interaction should be of interest of tribological investigations.
Most of the sources above dealt with concentrated (pin-on-disk, pin-on-plate, pin-on-sphere) experimental configurations, which do not completely comply with the real ball-in-cup eye/lens model.Therefore, Mabuchi et al. developed a pendulum simulator for measuring friction coefficient of contact lenses.The simulator employs a high sliding speed (90 mm/s) achieved due to a short period of oscillation.The friction between the lens and plastic hemisphere was measured in saline and HA solutions [21].Friction levels being 0.035 and 0.055 were reported, while the lowest friction was measured for pure saline solutions.In the subsequent study [22], the authors analyzed several types of soft contact lenses, finding that friction increases with increasing HA concentration.The higher friction is attributed to increasing viscosity, while the authors point out that such a phenomenon may have a negative impact on the ocular surface.
The above literature review introduces the problem of DES.As the main source of discomfort and eye irritation is in rubbing between the eye/lens/lid surfaces, it is evident that the tribological analyses may bring essential new findings toward more effective treatment.Various model configurations and experimental methods may be applied, but there are still a lack of knowledge, especially on the interaction of eye drop constituents and their role in eye lubricity.Therefore, the present study aims at a comprehensive rheological and tribological assessment of the frictional behavior of contact lenses.Attention is aimed at the effect of buffer solution, the composition of lubricant, and the type of contact lens.Specifically, two different contact lenses are studied, and twelve different lubricants are applied.

Materials and Methods
The main focus of this investigation in terms of the lubricants applied was on the MW of hyaluronan derivatives, the buffer solution, and the presence of mucin.Three fluorescently marked derivatives of hyaluronan (HA-FA) with MW from 66 to 562 kDa and degree of substitution (DS) from 5.5 to 6.3% (Contipro, Dolní Dobrouč, Czech Republic [23]) were tested.The HA-FA was dissolved in two different buffers, Borate and HEPES.The concentration of HA-FA was 0.3 wt% in these solutions.Following the implication from the literature review, the solutions with and without the presence of porcine stomach mucin (M1778, Type III, bound sialic acid 0.5-1.5%,partially purified powder, Sigma-Aldrich, St. Louis, Missouri, USA) were tested to clarify the mucin effect on the tribological performance.Therefore, 12 different lubricants were applied in total, see Table 1.Total amount of lubricant per friction test was 25 ml.The viscosity of the solutions was evaluated by a rotational rheometer Discovery HR 30 (TA Instruments, New Castle, Delaware, USA), in cone-plate configuration.Both the cone (60 mm in diameter, 1° cone angle) and the plate were made from stainless steel.The tests were realized under a temperature of a human body 37 °C.The viscosity was measured in the range of shear rates from 1 to 5000 s −1 .
Measurement of friction coefficient with a simultaneous record of the contact zone was performed using a custom lab-developed pin-on-plate simulator.The same simulator was previously successfully adopted when investigating cartilage friction, while the results were validated against a commercial tribometer [25][26][27].The pin is represented Table 1 Overview of the applied test lubricants a Dynamic viscosity at a shear rate of 33.9 s −1 [24] for calculating the mucoadhesive index b Mucoadhesive index is calculated as [3]: Δ(%) = [η mix − (η muc + η HAxxx )]/(η muc + η HAxxx ) × 100; where η mix is the specific dynamic viscosity of the lubricant, η muc is the specific dynamic viscosity of mucin, and η HAxxx is the specific dynamic viscosity of respective hyaluronan by a polydimethylsiloxane (PDMS) ball (ELASTOSIL LR 3003) with the lens placed at the top.The pin holder is stationary, while the carriage with the optical glass B270 plate carries a reciprocating motion.The normal force is applied through the pin and a strain gauge measures the frictional force.The friction coefficient is then evaluated as friction and normal force ratio.The scheme of the test device, including the optical observation system, is displayed in Fig. 1.Optical glass is chosen to allow direct in situ contact observation.Hyaluronan was fluorescently labeled with 5-aminofluorescein (FA, 201626, ~ 95%, Sigma-Aldrich, St. Louis, Missouri, USA) using a covalent bond.The experiment record is then evaluated using software, providing the average fluorescent emission intensity of the contact area is evaluated.Due to the linear character of fluorescence emission, the higher intensity indicates a thicker lubricating film [28].Therefore, the average fluorescent intensity determines the capability of the specific lubricant to create a protecting layer.In the later section, the intensity is plotted against the friction coefficient, while the ideal lubricant should provide minimum friction with sufficient and stable lubricating film preventing the surfaces from mutual contact (avoidance of mechanical irritation).The fluorescent intensity was evaluated continuously during each experiment with a frame rate of 100 frames per second, while the whole contact zone (approx.5 mm in diameter) was considered.In contrast to some other studies implementing fluorescent method for observation of lubrication mechanisms in joint replacements [10,29], the images were visually not as interesting.Due to relatively low HA concentration and high sliding speed, the lubricant was uniformly distributed in the contact zone with no significant clusters of HA molecules.Therefore, the images are not presented in the manuscript since they differ only in intensity levels.
The investigation focused on comparing two commercially available contact lenses; Acuvue Oasys (wearable for 14 days; Johnson & Johnson, New Brunswick, New Jersey, USA) and Biofinity (wearable for 30 days, CooperVision, San Ramon, California, USA).Both lenses are considered soft lenses made of silicone hydrogel, having a 14 mm diameter and a curvature radius of 8.4 and 8.6 mm, respectively.Acuvue Oasys contains 38% of water and 62% of bulk material, providing a low stiffness (elastic modulus of 0.73 MPa [30]).The lens consists of Senofilcon A, which is called to be a silicone hydrogel material of the second generation.This material does not depend on the surface treatments to maintain wetting (advanced contact angle 35.4° [31]); however, it has an internal wetting agent in the form of PVP.Biofinity contact lens is made from Comfilcon A (thirdgeneration silicone hydrogel), a naturally wettable polymer which also does not require surface treatment for enhanced wettability.These lenses do not contain any additional wetting agent and exhibit an even lower contact angle (29.6°).Thus, the Comfilcon contains more water (48%) and 52% of the bulk phase and maintains a low elastic modulus of 0.75 MPa.
The measurements were realized under the temperature of the body.The stroke length was 20 mm, and the frequency of reciprocating motion was 2 Hz.Therefore, the sliding speed was 80 mm/s.The contact zone was loaded by a normal force of 0.3 N, resulting in an approximate maximum contact pressure of 83 kPa for both tested lenses [32].The contact pressure level was determined based on the size of the actual contact area recorded by the camera under a given load.The length of one friction test was 20 s, while the first 1.5 s of data were cut due to the experiment running in.Concerning the lubricants described in Table 1, each test was carried out three times, always with a fresh solution dose and a new lens, while a single test was based on five repetitions in a row with 10-min lasting rehydration phase with unloaded contact.Rehydration is an important phase enabling the relaxing and recovery of soft-matter materials, such as cartilage [4,5,26] or hydrogels [33,34].Therefore, 15 measurements were carried out with each lubricant and both lenses, resulting in 360 tests in total.
Thus, the resulting dataset is significant, and the way of data curation had to be carefully considered.It should be noted that each test ( 5 2 is designed to illustrate the process of experiment design and data evaluation.

Results
Initially, the dynamic viscosity of all the tested lubricants was evaluated.The results of the three HA-FA derivatives with and without mucin solved in Borate and HEPES buffer are displayed in Fig. 2. As is expected, the viscosity increased with increasing HA MW.Further increase in viscosity is attributed to the addition of mucin due to mucoadhesive bonding.The effect of elevated viscosity is more pronounced for HEPES buffer.As can be seen from the graphs, HA-FA66 and HA-FA562 show a slight viscosity decrease at lower shear rates (up to 10 s −1 ), while the viscosity is almost constant for the rest shear rate range for the Borate buffer.However, it must be noted that at very low-shear rates, the measurement artifacts may influence the results.Concerning the HEPES buffer, a shear-thinning effect may be observed for HA-FA337 and HA-FA562 at higher shear rates.In contrast, HA-FA66 exhibits a viscosity increase for shear rates higher than 1000 s −1 ; however, this increase is attributed to the measurement artifact.
The resulting friction coefficients and fluorescent intensities directly related to the film thickness, including standard deviations (evaluated based on 15 tests in total, carried out in three experiments, each containing five rubbing tests in series) for Acuvue Oasys contact lenses are presented in Figs. 3, and 4. Concerning frictional behavior, the mucin addition caused a friction decrease for all HA-FA derivatives in both buffers except for HA-FA337 dissolved in HEPES (Fig. 3).There is no clear effect of data stabilization, while mucin led to lowering of standards deviation for some derivatives (HA-FA66 in Borate and HA-FA562 in HEPES), while the effect of the opposite for HA-FA562 in Borate and no effect was observed for the rest solutions.In general, it can be assumed that lubricants without mucin exhibited lower friction in HEPES, while in the case of lubricants with mucin, the effect was negative (HA-FA66) or negligible (HA-FA337 and HA-FA562).The experiments for Boratebased lubricants show a very good correlation between friction coefficient and lubricant film thickness.As can be seen, lower friction is always accompanied by increased fluorescent intensity for all HA-FA derivatives (Fig. 3 left vs Fig. 4 left).For the HEPES buffer, such behavior was observed only for HA-FA337, while for the rest two solutions, higher friction is in hand with thicker lubricating film.Therefore, it may be concluded that within the range of the performed experiments, mucin has a clear positive impact   and HA-FA337 (without mucin) in HEPES buffer.These two lubricants also exhibited the thickest lubricating layer, being around 600 and 720 counts of intensity.
Respective results of friction coefficient and film intensity for Biofinity contact lens are presented in Figs. 5 and  6.As is shown in Fig. 5 (left), the results are completely different than those for Acuvue Oasys.In the case of Biofinity, mucin in the Borate buffer caused an increase in friction for HA-FA66 and HA-FA337 and a slight decrease for HA-FA562.As in the previous case, lower friction is always associated with a thicker lubricating layer (Fig. 6 left).Some differences in lens behavior may also be observed for the HEPES buffer, where mucin led to lower friction for HA-FA337 and higher friction for HA-FA66 and HA-FA562.As for Acuvue Oasys, there is no clear correlation between friction and film thickness in the HEPES buffer.
In general, the friction of Biofinity lenses was slightly higher compared to Acuvue Oasys, while the minimum measured friction coefficient was around 0.11 (HA-FA66) and 0.13 (HA-FA337), both in Buffer solution without the presence of mucin.Interesting results in terms of film intensity were observed for the HEPES buffer.In that case, mucin caused a significant drop in fluorescent intensity independently of the HA-FA derivative MW (Fig. 6 right).
The results were subsequently plotted into color maps depicting the correlation between fluorescent intensity and friction coefficient.Each color map in the following section represents the area of results considering the 15 individual friction tests.Basically, it is expected that the higher intensity of fluorescence (i.e., the thicker film protecting the eye surface) and the lower friction coefficient indicate the optimal combination of properties for eye lens wearers.Figure 7 summarizes the results for the Acuvue Oasys contact lens.As can be seen, for the Borate buffer and mucin-free solutions, the range of friction coefficients is relatively wide, while film thickness is not as variable.Specifically, it can be seen that HA-FA562 covers the smallest area, indicating very stable behavior (coefficient of friction is between 0.13 and 0.2).In contrast, HA-FA66 showed a range of friction coefficient between 0.1 and 0.33 (Fig. 7 top left).As presented above, the addition of mucin led to stabilization of friction, specifically for HA-FA66, where the range is negligible; however, film intensity varies considerably (Fig. 7 top right).Concerning the HEPES buffer, most of the measured intensities are between 300 and 900, while mucin had a certain impact on data stabilization for HA-FA562.
The same color maps are further provided for Biofinity contact lenses (Fig. 8).In this case, the ranges of friction coefficients are not as substantial as were for Acuvue Oasys, independently of the MW of HA-FA derivative and the buffer used.Furthermore, the fluorescent intensities are generally larger.Specifically, it can be seen that HA-FA66 and HA-FA337 without mucin formed a relatively thick lubricating film together with low values of friction in the Borate buffer (Fig. 8 top left).The addition of mucin had a somewhat negative impact increasing friction for both derivatives.HA-FA562 exhibited the opposite behavior.When focusing on the lubricants in the HEPES buffer, a substantial impact of mucin on the stabilization of the measured parameters was found, as is shown in Fig. 8 bottom right.However, despite data stabilization, the film intensities are the lowest from all the four testes options, indicating formation of thin film, which can be associated with a risk of eye irritation.
The authors admit that the above graphs provide a lot of data, while it is a bit complicated to define which type of derivative, with or without mucin and which buffer offers the best properties for contact lens users.Therefore, the parameter called "Performance" is proposed, while the resulting representing number describing each type of lubricant is given as the ratio of average fluorescent intensity and average friction coefficient.Assuming the fact that thicker film protects the surfaces and the low friction ensures smooth motion of rubbing surfaces, the larger number represents the better combination of results.Figures 9 and 10 show that the maximum achieved performance for both types of contact is between 120,000 and 134,000.Therefore, it cannot be concluded either the first or the second contact lens is better.However, the performance parameter enables us to find important conclusions regarding lubricant suitability.As can be seen in Fig. 9, the HA-FA337 derivative provides very good performance for Acuvue Oasys in HEPES (either with or without mucin) and in the Borate (with mucin) buffer.Nevertheless, from a general point of view, it is suggested that Acuvue Oasys contact lenses will better operate for the drops containing mucin while the MW nor the buffer type plays a significant role (applies for five of the best six lubricants).For Biofinity contact lenses, different conclusions may be derived based on the performance values.It seems that mucin is not as important, and MW also has a limited impact, while the choice of the buffer solution is crucial.Nevertheless, it must be emphasized again that the obtained data apply to the interaction of the lubricants with the contact lenses, while the eye (without the lens) may behave differently.Five of the best six lubricants are dosed derivatives dissolved in the Borate buffer.In particular, all the derivatives without mucin in the Borate are in the top four tested lubricants.An interesting fact is that for both contact lenses, the best performance was achieved for HA-FA337 without mucin in a different buffer.

The Effect of Buffer, Mucin, and MW of HA-FA
The main difference in used buffers was their ability to compensate for the HA charges.HA is a negatively charged polyelectrolyte with COO-anions present at physiological conditions and in used buffers because the pKa of HA is around 2.8-3.1 [35,36] and pH in the Borate and Page 11 of 18 118 HEPES buffer were 6.9 and 7.5, respectively.The Borate buffer contained many counterions (Na+ concentration of 0.0711 mol/l) that could effectively compensate for the HA charges (COO-concentration of 0.0075 mol/l).On the other hand, the HEPES buffer is zwitterion without the ability to screen HA charges.The influence of buffers on the HA charge was confirmed by rheology (see Fig. 2).HA of each MW had significantly higher viscosity in the HEPES than the Borate buffer.Viscosity reflects the hydrodynamic size, which is increased considerably for polyelectrolytes at saltpoor conditions due to chain straightening and increased persistent length caused by charge repulsion [37,38].This behavior was observed in the HEPES buffer.The charges are screened, and uncharged dynamics and conformation are adopted at salt-rich conditions, as was observed in the Borate buffer.
The charge of HA influenced its interaction with contact lenses and mucin.HA with uncompensated charges interacted mainly via electrostatic interactions and also by H-bonding.It led to roughly similar fluorescent intensity for both contact lenses.HA with screened charges in the Borate buffer showed approximately two times lower intensity than charged HA in the HEPES when interacting with Acuvue Oasys lenses.The reason could be ineffective H-bonding with PVP brushes at salt-rich conditions and attractive electrostatic interactions at salt-poor conditions.A distinct situation was observed for Biofinity lenses.Significantly stronger fluorescence intensity was observed in the Borate than in the HEPES buffer.Different functional groups on the lens surface led to effective hydrogen bonding with HA in the Borate buffer without repulsing negative charges, which occurred in the HEPES buffer.
Generally, the effect of these changes was reflected in the friction coefficient.In the case of Borate buffer, the friction coefficient was lower for Biofinity than Acuvue Oasys.The friction coefficients in HEPES were very similar for both lenses except for middle MW, which showed the lowest friction coefficient in the case of Acuvue Oasys and the highest in the case of Biofinity.Such an influence of the specific MW will be further investigated.The highest MW kept roughly the same frictional coefficient inertly on buffer or lens type.It could correlate with larger space occupied by one interacting molecule on the lens's surface and high connectivity of longer and more entangled molecules.The friction coefficient of the lowest MW correlated well with the fluorescence intensity, i.e., the thickness of the HA layer on the lens's surface.It should be noted that simple comparative tests with pure buffers were carried out prior to HA derivatives testing, while the results achieved were identical.
Therefore, the influence of the buffer itself is eliminated, and the presented findings are thus directly implied by the interaction of HA (+ mucin) with buffers and lenses.
The present results showed that MW influences mostly the frictional processes for the Borate buffer, while the film thickness was not affected substantially.Recent investigations showed the poor ability of HA to adsorb onto both hydrophobic and hydrophobic rubbing surfaces [39], including contact lenses [40].Accounting for the limited adsorption performance together with the lower viscosity of Borate buffer, the formation of a thick lubricating film may be hardly expected.Therefore, the higher friction coefficient for low MW in the case of Acuvue Oasys is attributed to contact of surface asperities, as these lenses have higher surface roughness than Biofinity, where such a phenomenon was not observed.Moreover, Roba et al. [15] found that the lenses containing the PVP wetting agent experienced a significant friction increase in contact with the hydrophilic glass surface.Therefore, it is assumed that higher HA with higher MW forms a thicker and stable lubricating film, mitigating the interaction of PVP with glass, thus lowering the coefficient of friction.
Adding mucin led to a very complex situation of interacting ternary system.Previous studies showed that mucin could lower friction between both artificial and biological surfaces [41][42][43].Its benefit comes from good adsorption ability, preventing mutual contact of the surfaces [44].Mucin is a glycoprotein with high molecular weight and various functional groups, including hydroxyl, carboxyl, thiol, and sulfate moieties enabling diverse interactions, such as electrostatic, Van der Walls, hydrophobic, hydrogen bonding, and disulfide bridges [45][46][47].It seems mucin interacted with HA in the Borate buffer via hydrogen bonding, forming a hybrid system with better adhesion to PVP on the Acuvue Oasys surface, reflecting the larger thickness and lower friction coefficient.However, the same HA-mucin interaction had a different impact in the case of Biofinity lenses.HA alone adhered very well to these lenses.Then, the interactions with mucin and the creation of the hybrid structures worsened fluorescence intensity and increased friction coefficient.HA and mucin competed for the surface at the charged state in the HEPES buffer.Mucin adsorbed more preferentially onto the Biofinity surface than HA, resulting in a significant drop in the fluorescence intensity.Strong interactions of charged HA to PVP onto the Acuvue Oasys lenses ensured that only slight differences in the fluorescence intensity were observed after mucin addition.Good HA interactions with PVP and hydrogen bond formation were suggested in water in salt-poor conditions [48].
118 Page 12 of 18 Therefore, only minor changes in the frictional coefficient were observed for Acuvue Oasys in the HEPES buffer.Mucin was able to lubricate the lens's surface in the HEPES buffer.Therefore, the drop in fluorescence intensity was not followed by an increase in friction coefficient after adding mucin to the HA solution in the case of Biofinity lenses.
The importance of buffer solution was further revealed when performing the experiments with HEPES buffer.It is expected that there is an interaction of the buffer with the PVP agent, creating a hydration lubrication layer [49].As is presented in Figs.7 and 8, the use of HEPES buffer generally supported the formation of thicker film for Acuvue Oasys contact lenses (Fig. 7), while the effect can be partially attributed to higher lubricant viscosity.However, in the case of Biofinity lenses (Fig. 8), the film is considerably thinner, while the thinnest films were observed for a combination of HEPES and mucin.In contrast, the measurements with these lubricants experienced the best repeatability (the smallest dispersion of the points in Fig. 8).Limited ability of film formation in this specific case is attributed to the influence of pH and ionic strength of the buffer, introduced by Lee et al. [50].Therefore, it is suggested that mucin exhibits improved mutual interaction with the PVP wetting agent, which positively affects the lubricant film for the Acuvue Oasys lens.

The Effect of Contact Lens Material
As pointed out above, both lenses are made of different polymers [51].Acuvue Oasys contains the following components: PDMS, DMA, HEMA, TEGDMA, siloxane and PVP as the wetting agent.Biofinity lenses contain M3U, FMM, TAIC, IBM, HBMA, and NMNVA without additive of specific wetting agent (all abbreviations are stated in the Appendix).Although the lenses differ in materials, both are characterized as hydrophilic with nonionic charge [52].PVP wetting agent, in the case of Acuvue Oasys, is supposed to create a brush-like hydration layer at the surface of the contact lenses [53].Such polymer brushes may considerably lower friction coefficient and improve lubrication performance in boundary lubrication regime, as was shown by Lee et al. [54] and Muller et al. [55], who examined the role of PLL-g-PEG polymer brushes on rubbing surfaces.The mechanism responsible for the positive effect of polymer brushes is based on lowering the adhesive force, improving pressure resistance and forming low-shear fluid interlayer [56].
The results imply that HEPES puffer (without mucin) is a better solvent for Acuvue Oasys contact lenses.In contrast to Borate buffer, the friction is lower for each tested MW and the film thickness is higher (Figs. 3 and 4).Image data obtained by fluorescent microscopy revealed a bit limited adsorption HA expressed by an increase of film thickness between individual tests in the series.The adsorption resistance is attributed to the role of PVP [53,57] in combination with the impact of polymer brushes [58], which may positively influence tribological properties but may suppress the adsorption behavior.
The interaction of specific contact lenses with individual lubricants is expressed by the performance number (Figs. 9 and 10).This parameter was designed as it is expected that the thicker film, together with a lower friction coefficient, will result in comfortable use of contact lenses without eye/lid irritation.For the Acuvue Oasys lens, the results clearly showed that Borate buffer in combination with the absence of mucin leads to the worst tribological performance (Fig. 9).Mucin is thus essentially important for Borate buffer and even for HEPES buffer considering the lowest and the highest investigated HA MW.Keeping an eye on the Borate buffer, it seems that lower HA MW is beneficial, as this lubricant showed the second highest performance number.HA-FA66 also showed very good stability of friction data, as is shown in the top right image of Fig. 7. To sum up, HA-FA66 in the Borate buffer with mucin might represent an effective ingredient for eye drops used by the contact lens wearers; however, the long-term effect may not be sufficient due to a lower mucoadhesive index reported in Table 1.The best performance was achieved for HA-FA337 in HEPES buffer without mucin.This lubricant exhibits the lowest friction, thickest lubricating film, and very stable behavior illustrated in Fig. 7. Accounting that the mucoadhesive index of this lubricant is more than double compared to HA-FA66 with mucin in Borate, HA-FA337 in HEPES buffer represents the best solution for wearers of Acuvue Oasys contact lenses.
When the Biofinity contact lens was investigated, the achieved results were different in most aspects.A significant change was observed especially in film formation, as the film thickness continuously increased between individual tests, revealing a clear effect of HA adsorption onto surfaces.This effect was evident, especially for the Borate buffer, as is illustrated by the vertical dispersion of the points in the top left graph of Fig. 8. Adding the mucin led to data stabilization (top right graph in Fig. 8).It should be noted that film thickness increase was not observed during the rubbing test.Instead, sudden jumps were observed, occurring during a rest phase between the measurements when the contact was unloaded and the lens waded in the lubricant.This effect is therefore assigned to the rehydration effect of the Biofinity contact lens, during which HA adsorbs onto the surface of the lens and is partially absorbed by the lens matrix.The absorption phenomenon is considered especially due to the limited impact of increased film thickness on the friction coefficient.The enhanced absorption ability of Biofinity lenses may also be attributed to the higher water content.It is expected that part of the water is squeezed out from the matrix during loading and during a rest phase, the material soaks more HA.This statement is supported by the enhanced HA sorbing ability of the Biofinity lens, observed by Yamasaki et al. [40].
Concerning the Biofinity performance results, a combination of HEPES and mucin resulted in the worst performance (Fig. 10).The explanation may be found in the rapid fluorescent intensity decrease presented in the bottom right image in Fig. 8.When focusing on mucin interaction with Borate buffer, there is a clear positive tendency of increasing HA of MW-FA derivatives -the performance number increases with increasing MW.In general, three of the best four performances are achieved for mucin-free lubricants in the Borate buffer.As in the case of Acuvue Oasys, the number one performance is reported for the HA-FA337 derivative.Therefore, it may be summed up based on the extensive experimental analysis that middle MW (368 kDa) seems to be the best solution for preparing the ingredient for eye drops.

Limitations of the Study
It should be admitted that the investigation suffers from some limits.Firstly, the eye and the lid were modeled with artificial materials, and the contact was not conformal (pinon-plate).As the results were considerably supported by the optical in situ observations, the use of a transparent counterface is necessary.The experimental configuration is also the main reason why the obtained friction coefficients are relatively high compared to some of the previous studies.Nevertheless, it is suggested that the specific friction values should not be decisive parameters.The study provides a comprehensive comparison of data considering various influencing factors, while it is believed that the data are transferable to the clinical practice.Nevertheless, to support our data and to pursue research in this area, the new simulator adopting a fully conformal ball-in-cup eye model will be designed in the near future, while the present study represents a fundamental pilot investigation.
Secondly, the applied experimental conditions could comply completely with the physiological situation, as the device has some limitations in terms of speed and load.The contact pressure in the eye/lid interface is approximately between 0.3 and 7 kPa [59].It was possible to reach the pressure of 83 kPa using the adopted simulator and eye/lid model.Such pressure may be considered significantly higher compared to the natural eye; however, the contact pressure at the level of tens of kPa is still very reasonable for testing and much lower compared to some other tribological studies.
The assumption is that under a physiological state (lower pressure), the friction would be generally lower, while it is not expected that the change in pressure down to units of kPa would substantially affect the found phenomena.Concerning the applied speed, it is reported that the average eye blink speed is 134 ± 4 mm/s for closing and 26 ± 2 mm/s for the opening phase [60].As the simulator does not allow to change the speed for both reciprocating phases, the applied speed of 80 mm/s represents the average closing/opening blink speed.
Furthermore, as few datasets showed relatively large data dispersion, more measurement repetitions would be beneficial.However, it must be taken into account that the net time per experiment, including the experiment preparation, mixing of the fresh lubricant and data evaluation, lasts a couple of hours.Assuming the overall number of measurements performed and the costs per each lubricant dose, the three independent experiments (each composed of 5 tests in a sequence with a rehydration period) represent a sufficient dataset.As the market provides a variety of contact lenses, the selection of two representatives might also be considered insufficient.However, the explanation is similar to the number of tests performed.It was simply not possible in the timeframe of the project to examine more than two contact lenses.Acuvue Oasys and Biofinity were selected as these are very popular among the population and are made of different materials, which was found to be essential.
The study could also be extended, including some advanced analyses such as spectroscopy or ellipsometry toward revealing the mechanisms of adsorption of the molecules.Anyway, the greatest limitation is that surfaces for such analyses must be dry, while the contact lens undergoes irreversible damage associated with significant changes in its properties when dried.Partial explanations could come from the analysis of the optical glass; nevertheless, its surface is considered less important, assuming the lens-glass contact pair.Moreover, it is well known that the analyses above have some limitations when transparent surfaces are studied due to light reflection.Therefore, the detailed mechanisms of the adsorption of molecules could not be presented within the scope of the present study.Assuming the limitations of the optical observations, some artifacts related to the application of fluorescent optical methods could theoretically play a role.However, attention was paid to carrying out the experiments under the same light conditions.The whole simulator was completely covered by the thick dark-black fabric to prevent any potential influence by the ambient light.The tubes with the lubricant were prepared in the dark room and packed in aluminum foil to avoid any degradation of the fluorescent markers before the experiments.In addition, the background image was always taken prior to the measurement and subtracted from the recorded images to eliminate the noise signal of the optical setup.
The authors admit that the experimental investigations should be accompanied by numerical modeling and vice versa, if possible.However, it must be emphasized that the investigated contact (hydrogel on glass) belongs to the groups of very soft contacts, which are pretty challenging for modeling compared to rigid bodies.Moreover, the motion frequency was quite high, together with a relatively short stroke length.In addition, it must be considered that the interaction of the HA and mucin molecules with contact surfaces is nearly impossible to model.Therefore, the present study focused exclusively on experimental testing, paying attention to providing a comprehensive dataset.Nevertheless, the motivation to create a numerical model of the eye/ lens/lid interface, including the interaction of biological molecules with rubbing surfaces, is definitely a point for experts focusing on creating predictive models.
Finally, as explained in the Materials and Methods section, the complete dataset could not be presented in the manuscript due to the excessive data amount (288 independent charts).Anyway, two additional figures (see Figs. A and B in the Appendix section) are provided below to illustrate the evaluation process for Biofinity contact lens considering the HA-FA337 (Borate) lubricant exhibiting the best performance and HA-FA337 + mucin (HEPES) with the worst performance according to Fig. 10.The same approach was applied for all the tested combinations (lenses and lubricants).Such a data processing scheme was designed after detailed data analysis from various perspectives and is considered the most robust way to provide relevant conclusions.

Conclusion
The present investigation aimed at measuring the friction coefficient of contact lenses sliding against a flat optical glass specimen.The uniqueness of this investigation is mainly in the measurement of friction carried out simultaneously with in situ contact observation.The significant findings are listed as follows: -The contact lens material has a particular impact on lubricant film formation and the friction coefficient of hyaluronate solutions.Therefore, the treatment effect of eye drops for DES will differ for wearers of different contact lenses.-Buffer solution and the presence of mucin considerably influence the tribological performance.-Concerning the individual tested solutions, the best performance was observed for HA-FA337 + HEPES for Acuvue Oasys and HA-FA337 + Borate for Biofinity.
As the base hyaluronan is the same, there is an apparent link between the buffer type, the water content, and the presence of the wetting agent.-In general, it was found that for Acuvue Oasys, the presence of mucin is essential toward improved tribological performance, while neither the buffer type nor the MW of HA-FA derivatives plays a critical role.-In contrast, the Biofinity contact lenses achieved the best performance for the Borate-based lubricants, while the addition of mucin had a rather negative impact.
Future studies should reflect very recent improvements performed by Mabuchi et al. [21] and Iwashita et al. [22], who adopted a conformal ball-in-cup configuration for contact lens investigation for the first time.Implementation of optical methods together with precise friction measurement considering more realistic contact conditions could potentially become a new standard for testing drops for DES treatment.Funding Open access publishing supported by the National Technical Library in Prague.Technology Agency of the Czech Republic, FW01010060.

Fig. 1
Fig. 1 3D computer model of the test rig (left).The detail of the experimental configuration (right) steps in a sequence) results in four different diagrams, i.e., (i) original record of friction data -line chart with 5 lines; (ii) original record of fluorescent intensity -line chart with 5 lines; (iii) averaged friction coefficients including standard deviation -bar chart with 5 bars; and (iv) averaged fluorescent intensities including standard deviation -bar chart with 5 bars.Furthermore, it is noted that all the measurements were performed three times on different days and were performed for two different contact lenses and twelve different lubricants.To sum up, the experimental work resulted in 4 × 3 × 2 × 12 = 288 original charts + charts presented in the manuscript (Figs. 2, 3, 4, 5, 6, 7, 8, 9, 10).It is evident that the whole dataset cannot be presented in a manuscript.Therefore, thanks to the satisfying repeatability and qualitatively stable behavior of friction and fluorescent intensity for most of the tested combinations, the following section provides the averaged values (out of 15 tests, see Figs. 3, 4, 5, 6, 7, 8) with corresponding standard deviations.The Appendix section further contains Fig.A and Fig. B, illustrating the evaluation process.Since the experimental methodology is quite comprehensive, Table

Fig. 2
Fig.2The dependence of dynamic viscosity on the shear rate of the tested lubricants

Fig.
Fig. Fluorescent intensity (dimensionless film thickness) including standard deviation for Acuvue Oasys contact lenses for hyaluronan derivatives in Borate (left) and HEPES (right) buffer

Fig.
Fig. Fluorescent intensity (dimensionless film thickness) including standard deviation for Biofinity contact lenses for hyaluronan derivatives in Borate (left) and HEPES (right) buffer

Fig. 7
Fig. 7 Correlation of friction coefficient and fluorescent intensity (dimensionless film thickness) for Acuvue Oasys contact lenses for hyaluronan derivatives in Borate (top) and HEPES (bottom) buffer without (left) and with (right) mucin

Fig. 8
Fig. 8 Correlation of friction coefficient and fluorescent intensity (dimensionless film thickness) for Biofinity contact lenses for hyaluronan derivatives in Borate (top) and HEPES (bottom) buffer without (left) and with (right) mucin

Fig. 9
Fig. 9 Dimensionless performance (ratio of fluorescent intensity and film thickness) of Acuvue Oasys contact lenses for all the test lubricants

Fig. 10 10 118
Fig. 10 Dimensionless performance (ratio of fluorescent intensity and film thickness) of Biofinity contact lenses for all the test lubricants

Fig. 11
Fig. 11 Illustration of data evaluation algorithm for HA-FA337 (Borate) -the best overall performance