You are on page 1of 4

Modelling of Nanoparticle Diffusion in the Paranasal Sinuses

Qinjiang Ge, Kiao Inthavong, Jiyuan Tu

School of Aerospace Mechanical Manufacturing Engineering, RMIT University Bundoora, Australia
Abstract A nasal-sinus cavity model was created in order to determine if any nanoparticles (NPs) would deposit within the maxillary sinuses at a steady inhalation flow rate of 10L/min. The computational model was simulated in Fluent v12.1using the Brownian diffusion model. Airflow contour patterns and streamlines showed very low flow (<0.006% of the inhaled flow rate) were present in the maxillary sinuses. This suggests that any particles reaching the maxillary sinus would be caused by the particle diffusion rather than the airflow momentum. Deposition patterns showed that 98% of 1nm particles deposited early in the nasal cavity while for 10nm the deposition efficiency reduced to 29%. However a more evenly distributed deposition pattern was found for 10nm particles. It was found that only a small percentage of particles <0.5% deposited within the maxillary sinuses. Additionally the particle trajectory time is important for NP deposition studies as it gives an indication of the likelihood of deposition in different regions of the nasal cavity. For example the shorter residence time of 1nm means that deposition occurs nearly immediately and the deposition zone is restricted to the nasal cavity and further deposition downstream is unlikely. Keywords: nasalcavity, sinus, nanoparticle, deposition, CFD

the brain, while its absorption into the bloodstream via the highly vascularised mucosal walls can translocate to other body organs. Recently Xiong et al. [4] used Computational Fluid Dynamics (CFD) to simulate steady inhalation and exhalation flow through a computational model of a human nasal cavity and its paranasal sinuses. Their results showed there was very little flow through the opening of the maxillary sinus (ostium) and the maxillary sinus itself. With limited flows directed into the sinuses, micron sized particles are not expected to deposit in the sinuses. However the Brownian motion of ultrafine particles may in fact deposit within the ostium as well as inside the sinuses as well. To date nearly all CFD studies of respiration through the nasal cavity with flow distributions, resistance, and ultrafine particle deposition results, have neglected the paranasal sinuses. Therefore the aim of this study is to investigate the flow patterns in the nasal cavity, in the paranasal ostium and its corresponding sinus, and uptake of NPs with a focus on the ostium region that may cause occlusion and contributing to sinusitis. The uptake of the NPs within the nasal-sinus model and a ratio of sinus deposition to nasal deposition for different particle sizes are given. II. METHOD



With the exponential growth in the production of engineered nanoparticles for manufacturing, health and medical science, there is an increased need for health risk assessment due to increased daily exposure of nanoparticles (NPs). Toxicology studies of NPs have established the decreasing deposition fraction or efficiencies as the NP increases in size from 1-100nm [1, 2], which is primarily caused by the particle Brownian diffusion. The deposition of particles in the nasal cavity is a common phenomenon during respiration as it is a major route of inhalation and is the first line of defence for filtering airborne NPs from the human respiratory tract. For inhalation of micron sized particles, inertial deposition has been found to be the dominant deposition mechanism for particles with aerodynamic diameters greater than 2m [3] whereas submicron particles such as NPs of less than 100nm are dominated by diffusion. This means that where micron particles behave like tracer gas particles until its inertia causes deposition, NPs will diffuse evenly throughout the airway geometry. This causes a random dispersion and may lead to deposition in distal regions of the nasal cavity such as within the paranasal sinuses. It has been found that NP deposition in the olfactory region can lead to a rapid and efficient pathway to several regions of

A. Nasal cavity geometry creation CT scans of a healthy nose from a 51-year-old Asian male, mass were obtained. The CT scan was performed using a CTI Whole Body Scanner (Phillips/Brilliance) which extended from the superior frontal and ethmoid sinuses to the oropharynx. The outline of the model was segmented from the CT-scans using the software Mimics and meshed with ICEM-CFD software. An outline of the model is shown in Figure 1

Figure 1. Coronal slices of the airway showing the presence of the maxillary sinus.

The nasal-sinus cavity dimensions and its comparison with other models in the literature are given in Table 1. The meshing scheme used a hybrid mesh that included six-prismatic layers with an inner tetrahedral core. Grid independence based on velocity profiles at the outlet was performed and the final mesh size for the nasal-sinus model was 3.2million cells.
TABLE I. NASAL CAVITY GEOMETRY COMPARISONS Xiong [4] Current 9.1 9.7 6.6 7.3 290 *right nasal cavity only

Doorly et al. [5] Model 1 Model 2 Model 3 Overall length (cm) 10.5 10.6 11 Overall width (cm) Surface area (cm2) 106* 107* 109* Volume (cm3) 13.8 14.2 22.4

relative to atmospheric pressure that is caused by the movement of the diaphragm. Boundary conditions for the particles are set up as a circular particle release entrained in the flow field. Particles were released from 0.01m from the inlet to prevent any spurious data exiting the inlet upon immediate release. Furthermore, a particle was located at no less than 0.1mm away from the wall to eliminate artificial immediate deposition on the walls due to the stochastic nature of the Brownian motion model. A near wall interpolation scheme [1] is applied onto wall adjacent cells to allow the particle velocity to vary as it approaches zero rather than be a uniform velocity throughout the cell. III. RESULTS

B. Numerical modelling The commercial CFD code, Fluent v12.1 was used to solve the governing equations for fluid flow under steady flow rate of 10 L/min. The inspiratory flow rates for adults can range between 5-12L/min for light breathing and 12-40L/min for non-normal conditions such as during exertion and physical exercise. Up to a flow rate of 15L/min the flow regime in the respiratory airways has been determined as dominantly laminar, although traces of turbulent flow structures may exist [6]. In this study a laminar flow model is used to focus on the diffusion process of the NPs and because of the low flow rates. The steady-state continuity and momentum equations for the gas phase (air) in Cartesian tensor notation can be cast as:
g u ig = 0 xi

u ig g x j

ug j

u ig 1 p g = + x j xi x j


A. Brownian diffusion modelling validation The number of particles tracked was checked for statistical independence since modelling Brownian motion is inherently of a stochastic nature. Independence was achieved for 70000 particles, since an increase of particles to 100,000 particles yielded a difference of less than 1% in the deposition efficiency. Deposition of NPs in the range of 5-12nm particles was tested for a 90o bend pipe (Figure 2). The applicability of the Gaussian white noise Brownian diffusion model used within Fluent-v12.1 (BMv12)needs to be verified given that the same model in Fluent-v6.3 failed to predict nanoparticle deposition [1, 2]. The results show that the model in Fluentv6.3 under predicts the deposition of the NPs especially for 11nm. The results show up to two orders of magnitude below the experimental data. The results for Fluent-v12 show better agreement with the experimental data. It must be noted that the models for Brownian motion that are available in Fluent v6.3 and v12.1 is the same model based on a Gaussian white noise process.

Deposition Efficiency (%)

The equations were discretised with the QUICK scheme while the pressure-velocity coupling was resolved through the SIMPLE method. For a low volume fraction of dispersed phase (particles), the Lagrangian approach with one-way coupling is used, i.e. the airflow transports the particles, but the effect of particle movements on the flow is neglected. In this approach, the airflow field is first simulated, and then the trajectories of individual particles are tracked by integrating a force balance equation on the particle which can be written as:
duip = FD + Fg + FB + FL + FT dt

10 10 10 10 10 10

90o Bend Pipe





Wang (2002) FLUENT v6.3 FLUENT v12.1

5 7 9 11



Particle Diameter (nm)

Figure 2. Comparison of deposition efficiency results using Brownian models from Fluent 6.3, and Fluent 12.1 in (a) straight pipe 1L/min, (b) straight pipe 10L/min, and a (c) 90o bend pipe.

FD is the drag force per unit particle mass taking the form of Stokes' drag law [7], Fg is the gravity term, FB is the Brownian force [8], is Saffman's lift force due to shear, and FT is the thermophoretic force. Particle rebounding from the surfaces was ignored and particle deposition was determined when the distance between the particle centre and a surface was less than or equal to the particle radius. The particle tracking is then terminated. Inhalation through the nasal cavity is induced through a pressure difference between the nostril inlets (Pin = 0Pa) and the nasopharynx outlet (Pout) that is set to a negative pressure

B. Flow patterns and streamlines Contours of velocity magnitude at slice a-e (as defined in Figure 1) are shown below in Figure 3. Flow acceleration is found in slice-a with a peak velocity of 1.8m/s while pockets of low velocity are found at the top and bottom of the slice. As the flow travels downstream the peak velocity decreases and in

slice-c and slice-d, the peak velocity reaches only 1.3m/s. This is due to the airway passage expanding in cross-sectional area resulting in lower velocities. The contours show that the bulk flow regions occur mainly through the mid-height region and close to the nasal septum which separates the two cavities. The contour at slice-e shows a well mixed pattern which is caused by the airflow from the left and right sides of the nasal cavity merging together. Very low flows at < 0.1m/s are found in the sinus regions.

In both instances the streamlines initially accelerate near the nostril opening before passing mainly through the main nasal passage at mid-height. Some streamlines travel along the floor of the nasal cavity, while some reach the olfactory regions, but these streamlines are highlighted with blue colour, which denotes a low velocity 0.01m/s. Streamlines also pass into the maxillary sinus, squeezing through the narrow ostium (Figures 4c and 4d). These streamlines are very low velocity that recirculates inside the maxillary sinus. CFD analysis showed that the minimum ostium diameter is 3.03mm and 3.78mm, and the pressure difference between the ostium entrance and inside the maxillary sinus are 0.056Pa and 0.0026Pa for the left and right sides respectively. The mass flow rate through the left and right ostium is 11.4e-9 kg/s and 6.77e-9 kg/s which are < 0.006% of the total inhalation flow rate. This small percentage of flow is not conducive to particle deposition and that the only mode of deposition in the paranasal sinus will be due to the particle Brownian diffusion. C. Nanoparticle deposition The number of particles tracked was checked for statistical independence because the BM is inherently of a stochastic nature. Particle number independence was achieved for 70,000 particles, because an increase of particles to 100,000 particles yielded a difference of less than 1% in the deposition efficiency. Deposition patterns for 1nm and 10nm particles are shown in Figure 5 which shows that early deposition occurs for 1nm with nearly all particles depositing in the anterior half of the nasal cavity with a large proportion of the particles persisting for less than 0.022secs in the nasal cavity domain.

Figure 3. Velocity magnitude contours at 10L/min for Slices a-e. Colour scale is given in [m/s].

Flow streamlines were released from the left and right nostrils in order to trace the flow patterns (Figure 4).






Figure 5. NP deposition pattern in the nasal-sinus cavity for (a) 1nm resulting in 98% deposition and (b)10nm - resulting in 29.8% deposition. Particles are coloured by trajectory time within the nasal cavity before impacting onto the surfaces at 10L/min..

Figure 4. Streamlines passing through the nasal cavity that originates from the (a) left and (b) right nostrils at 10L/min. Colour scale represents the velocity magnitude in [m/s].

This deposition pattern is indicative of the highly diffusive nature of NPs that become as small as 1nm. The strength or influence of the Brownian diffusion decreases as the particle size increases from 1nm upwards, since the molecular collisions with the particle become less significant. This is evident in the deposition pattern of 10nm which shows more random and evenly distributed deposition sites. For 10nm particles, the time scale is 10x as great as that for 1nm which suggests that the particles are not as diffusive and are transported with the inhaled flow field for longer and hence the ability to travel deeper into the nasal cavity. This particle trajectory time is particularly important for NP deposition studies as it gives an indication of the likelihood of deposition in different regions of the nasal cavity. For example the shorter residence time of 1nm means that deposition occurs nearly immediately and the deposition zone is restricted to the nasal cavity and further deposition downstream is unlikely. This protects the sensitive lung airways from those NPs that exhibit dangerous properties for respiratory health. Conversely the ability to deposit particles in the middle regions of the nasal cavity or even deeper into the lung airways with high deposition, can be important for therapeutic drug delivery. Since the diffusion property of NPs provides maximum deposition for 1nm and decreases rapidly as the particle size increases, a number of different sized NPs were released from the nostril inlets to determine if any would deposit within the maxillary sinuses. The particles that reach the ostium is expected to deposit due to diffusion only and that the inertial momentum from the inhaled air has insignificant effect because of the low percentage of flow passing through the ostium. Figure 6a shows that in the right maxillary sinus, inclusive of the ostium, a small percentage of particles <0.04% are deposited.

able to pass through the ostium, but a larger percentage of particles <0.5% deposited within the ostium alone. This is mainly due to the curved geometry and longer ostium length providing a narrow tube passageway for the particles to diffuse onto. IV. CONCLUSION

A nasal cavity model including the sinuses was created in order to determine if any NPs would deposit within the maxillary sinuses given that these particles are transported through the nasal cavity mainly by diffusion. It was shown that 1nm particles deposited early in the nasal cavity with a deposition efficiency of 98%. As the particle increased in size to 10nm the diffusive nature of the NP decreased and the deposition efficiency reduced to 29%. However a more evenly distributed deposition pattern was found for 10nm particles. The maxillary ostium, which is the connecting passage from the main nasal cavity to the maxillary sinus, plays a role in restricting the NPs from entering the sinus. It was found that only a small percentage of particles <0.5% deposited within the maxillary sinuses. For example the shorter residence time of 1nm means that deposition occurs nearly immediately and the deposition zone is restricted to the nasal cavity and further deposition downstream is unlikely. ACKNOWLEDGMENT The financial support provided by the Australian Research Council (project ID LP0989452) and by the RMIT University through an Emerging Research Grant is gratefully acknowledged. REFERENCES
[1] Inthavong, K., K. Zhang, and J. Tu, Numerical modelling of nanoparticle deposition in the nasal cavity and the tracheobronchial airway. Computer Methods in Biomechanics and Biomedical Engineering, 2011. Longest, P.W. and J. Xi, Effectiveness of direct Lagrangian tracking models for simulating nanoparticle deposition in the upper airways. Aerosol Science and Technology, 2007. 41(4): p. 380-397. Inthavong, K., et al., A numerical study of spray particle deposition in a human nasal cavity. Aerosol Science Technology, 2006. 40(1034-1045). Xiong, G.X., et al., Computational fluid dynamics simulation of airflow in the normal nasal cavity and paranasal sinuses. American Journal of Rhinology, 2008. 22: p. 477-482. Doorly, D.J., D.J. Taylor, and R.C. Schroter, Mechanics of airflow in the human nasal airways. Respiratory Physiology & Neurobiology, 2008. 163(1-3): p. 100-110. Garcia, G.J.M., et al., Atrophic rhinitis: a CFD study of air conditioning in the nasal cavity. Journal of Applied Physiology, 2007. 103(3): p. 1082-1092. Ounis, H., G. Ahmadi, and M. J.B., Brownian diffusion of submicrometer particles in the viscous sublayer. J. Colloid and Interface Science, 1991. 143(1): p. 266-277. Li, A. and G. Ahmadi, Dispersion and deposition of spherical particles from point sources in a turbulent channel flow. Aerosol Sci.Technol., 1992. 16: p. 209-226.


[3] [4]




Figure 6. Frontal view showing the NP deposition in maxialliary ostium and sinus for the (a) right nasal cavity, and (b) the left nasal cavity at 10L/min. Different sized particles are coloured as follows: 1nm red circles; 5nm yellow diamonds; 10nm blue squares; 40nm black left triangles; 100nm pink right triangles.


It can also be seen that 1nm particles are captured within the narrow ostium and in fact don't make it through to the maxillary sinus. For the left maxillary sinus, no particles were