UC Berkeley

The increasing global prevalence of dry-eye disease has spurred clinical and research interest in understanding this condition. Most cases of dry eye are attributed to increased tear-evaporation rates. The current paradigm holds that a dysfunctional tear-film lipid layer (TFLL) leads to an increased tear-evaporation rate from the exposed ocular surface, which causes elevated tear salinity (hyperosmolarity). Chronic tear hyperosmolarity then triggers an inflammatory cascade that leads to clinical dry-eye disease. However, a limited fundamental understanding of the role of the TFLL in dry eye impedes progress in developing care strategies. In an effort to advance knowledge in the field, this thesis focuses on tear evaporation in dry eye, on the structure and physical properties of tear lipid, and on the role of duplex-oil films in water-evaporation reduction.

First, a compartmental tear-dynamics model is presented that predicts the role of various tear flows on relevant tear parameters such as tear osmolarity and volume. The tears are compartmentalized based on physiology, and tear physics is described over an entire blink cycle. Coupled water- and salt-conservation equations govern the dynamics within each compartment. Tear-supply and tear-evaporation rates are varied to investigate tear behavior in normal and dry-eye conditions. Model predictions match clinical measurements over a wide range of tear-supply and tear-evaporation rates. The tear-evaporation rate is the strongest driver of tear osmolarity. Inclusion of osmotic water flow through the cornea and conjunctiva enables better matching to clinical data than previous models. Additionally, osmotic water flow is found to contribute significantly to tear flow, especially in dry-eye conditions.

Despite substantial knowledge of the chemical composition of the secreted lipid, called meibum, little is known of its structure or bulk properties. We investigate the physical and structural properties of collected bulk samples of human and bovine meibum utilizing rheology, calorimetry, and x-ray scattering. Steady and oscillatory shear rheology demonstrate both human and bovine meibum to be remarkably viscoelastic and shear-thinning even at elevated temperatures. Small- and wide-angle x-ray scattering (SAXS and WAXS) confirm the presence of structured crystalline domains at physiological temperature that dissipate with increasing temperature. The melting of crystalline structures near eye temperature measured by SAXS and differential scanning calorimetry (DSC) corresponds to a decrease in meibum viscosity and elasticity over the same temperature range. Meibum viscoelasticity persists even after crystal structures liquefy in SAXS experiments. Thus, our proposed structure for the TFLL at physiologic temperature is a highly viscoelastic, shear-thinning liquid suspension consisting of lipid lamellar-crystallite particulates immersed in a continuous liquid phase. This new paradigm for viewing the lipid layer contrasts with the current picture of several layers of stacked lipids.

Finally, we measure water-evaporation reduction by duplex-oil films at thicknesses from 100 nm and 100 µm. Water-evaporation rates are measured gravimetrically with a newly constructed apparatus under controlled gas-phase mass transfer and environmental conditions. White-light interferometry permits continuous visualization of the deposited oil layers to monitor film uniformity. Duplex-film spreading and dewetting are identified as key challenges to obtaining reliable water-evaporation reduction. Duplex-oil films of oxidized mineral oil, bovine meibum, and human meibum reduce evaporation by a proposed dissolution-diffusion mechanism. The data are fitted to dissolution-diffusion theory to yield the water permeability of the duplex-oil film, Dk, which is a material property of the oil. Measured film permeabilities of oxidized mineral oil agree with those reported in literature. Bovine and human meibum, however, do not reduce water-evaporation rates significantly at thicknesses near 100 nm. These findings appear to contradict in-vivo tear-evaporation measurements performed clinically. However, clinical evaporation measurements lack calibration and gas-phase mass-transfer characterization. Consequently, more experiments are needed to clear up this apparent contradiction and to discern the true role of the TFLL in evaporation reduction.