An experimental and theoretical investigation of the rheological properties and degradation of mucin solutions ( or why saliva becomes watery when removed from your mouth

The use of biological fluids such as saliva and cervical mucus as diagnostics for measurements of health status is becoming increasingly popular in the fields of biology and medecine, particularly given the non-invasiveness and ease of obtaining such fluids [39, 78]. In general, these biological fluids are polymeric, and as a result tend to be viscoelastic. However, as a result of protease and enzymatic activity, these fluids are often unstable and can degrade with time [23, 65]. This was observed in the case of saliva by Aggazzotti nearly a century ago [1]. Therefore, in order to reliably quantify their rheological properties for diagnostic purposes, it is essential to understand how their microstructure affects the bulk rheological behaviours observed under testing conditions. We develop two models to simulate the behaviour of saliva during simple elongational flow and account for the decrease in viscoelasticity with time. The first model considerd is the FENEP model of a fluid, which is particularly suitable for describing the rheology of dilute polymer solutions (Newtonian solvents containing small amounts of dissolved polymer) as a result of its ability to capture nonlinear effects arising from the finite extensibility of the polymer chains. In extensional flows, these polymer solutions exhibit dramatically different behaviour from the corresponding Newtonian solvents alone, notably through the creation of persistent filaments when stretched. By using the technique of capillary thinning to study the dynamics of the thinning process of these filaments, the transient extensional rheology of the fluid can be characterized. We show that under conditions of uniaxial elongational flow, a composite analytic solution can be developed to predict the time evolution of the radius of the filament. Furthermore we derive an analytic expression for the finite time to breakup of the fluid filaments. This breakup time agrees very well with results obtained from full numerical simulations, and both numerics and theory predict an increase in the time to breakup as the finite extensibility parameter b, related to the molecular weight of the polymer, is increased. As b -+ oo, the results converge to an asymptotic result for the breakup time which shows that the breakup time grows as tbrea~k ln(Mw), where Mw is the molecular weight of the dilute polymer solution. We then consider the importance of the network properties of saliva that arise due to entanglements of the polymer chains. In order to account for this, we combine the FENE-P model with the Rolie-Poly model developed by Graham et al [45, 50] to obtain the Rolie-Poly-FENE-P model. We show that this model 3 is better able to accurately predict the extensional behaviour of both polyethylene oxide (PEO) solutions and saliva based on actual properties of these materials. This model cannot capture the sudden filament breakup observed in young saliva samples, however, which motivates the incorporation of a mechanism for network junction association or 'stickiness', as has been done by [71, 74, 40, 25] amongst others in biological networks. We draw largely off of the work for Tripathi et al [67] who modeled the rheology of hydrophobically modified ethoxylate-urethane (HEUR) polymer solutions as associating networks in order to develop an analogous model for saliva. We show that this model can reproduce the asymptotic 'middle elastic time' exponential radius decay described by Entov and Hinch [22], the dynamics upon which CaBER experimental interpretation of the system relaxation time AH is based. We also show that incorporation of a stickiness parameter allows for good agreement between the model and experimental CaBER data for saliva samples at various ages. Thesis Supervisor: Gareth H. McKinley Title: Professor, Mechanical Engineering