Experimental measurements and computational modeling of aerosol deposition in the Carleton-Civic standardized human nasal cavity

Abstract

Aerosol deposition in the novel, “Carleton-Civic” standardized geometry of the human nasal cavity was studied both numerically and experimentally. Inhalation flow rates varied from 30 to 90 L/min in the experiments, and aerosol droplets had diameters ranging from 1.71 to 9.14 μm (impaction parameters ranging from 123.3 to 2527.6 μm L/min). For the numerical simulations, both the RANS/EIM (Reynolds averaged Navier–Stokes equations for the gas phase and eddy-interaction random walk models for the particulate phase) and large eddy simulations were used. The mechanism of aerosol deposition in the standardized nasal cavity was dominated by inertial impaction. Deposition data from the standardized nasal cavity transected cited in vitro data based on individual subjects. The data also correlated very well with cited in vivo measurements but generally showed less aerosol deposition for a given value of the impaction parameter. Regional deposition characteristics within the nasal passages were also investigated both experimentally and numerically and new trends of regional deposition versus impaction parameter are discussed. These trends provide new insight into the general deposition behaviour of various sized aerosols within the human nasal cavity.

Introduction

Aerosol deposition in the human nasal cavity has practical significance ranging from nasal drug delivery (congestion medication, migraine medication, and even vaccination, as pointed out by Mathison, Nagilla, & Kompella, 1998; Illum, 2000) to filtration of particulate matter from the atmosphere during normal respiration (Morgan & Monticello, 1990; Subramaniam, Asgharian, Freijer, Miller, & Satish Anjilvel, 2003; Swift, 1981). The highly complex geometry of the human nasal cavity not only makes the research of airflow patterns in the nasal cavity difficult, but also raises significant challenges in clinical research in many related fields including nasal surgery (e.g. Adamson, 1987; Minn et al., 2002; Murray, El-Hakim, Ahsan, & Nunez, 2003; O’Neill & Tolley, 1998; Pang, 2005; Saunders, Jones, & Kabala, 1999; Zhao et al., 2006), nasal toxicology (e.g. Bennett & Zeman, 2005; Kimbell, 2006; Kimbell, Subramaniam, Gross, Schlosser, & Morgan, 2001; Overton, Kimbell, & Miller, 2001; Roberts et al., 2006; Schroeter et al., 2008; Tian, Inthavong, & Tu, 2007; Yu, Zhang, & Lessman, 1998), risk assessment (e.g. Rhomberg & Lewandowski, 2006; Roush, 1979; Wang et al., 2003), disease diagnosis (e.g. Davis & Eccles, 2004; Kramer & Wray, 2000; Lenders & Pirsig, 1990; Raj, Stableforth, & Morgan, 2000; Suzina, Hamzah, & Samsudin, 2003), and olfaction (e.g. Churchill, Shackelford, Georgi, & Black, 2004; Damm et al., 2002; Zhao, Scherer, Hajiloo, & Dalton, 2004). Despite this significance and the increased scientific interest in the past decades, the physics of the aerosol deposition in the nasal cavity is not completely understood.

Aerosol deposition in human nasal cavity has been studied both experimentally and numerically. Experimental measurements of the deposition efficiency in the airway passages have been performed in vitro (using replicas) and in vivo. The principle of in vivo determination of nasal particle deposition is measurement of the aerosol concentration before it enters the nasal cavity and after it leaves the mouth (Cheng, K. H. et al., 1996; Cheng et al. (1996), Heyder and Rudolf, (1977); Hounam, Black, & Walsh, 1969; Kesavanathan & Swift, 1998; Pattle, 1961; Stahlhofen, Rudolf, & James, 1989; Swift & Strong, 1996). Cheng (2003) summarized and analyzed all in vivo data in human volunteers by Landahl and Black (1947), Pattle (1961), Hounam et al. (1969), Giacomelli-Maltoni, Melandri, Prodis, and Tarroni (1972), Heyder and Rudolf (1977) and concluded that the deposition data can be reasonably correlated using the following equation

$$\mathit{DF}=1-\exp (-a^{\prime }{d}_a^{2}Q)$$

where DF is the total deposition fraction, d_a is the aerodynamic diameter in μm (defined as the square root of the particle to water density ratio multiplied by the particle diameter, (ρ_p/ρ_w)^1/2d_p), Q is the inhalation flow rate (L/min), and a′ is an experimentally determined constant (μm⁻² L⁻¹ min).

In vitro experiments using artificial nasal cavities (nasal casts or replicas of nasal airways) have been used extensively as a complement to difficult and expensive in vivo studies for decades (see Cheng et al., 2001; Gradon & Podgorski, 1992; Itoh, Smaldone, Swift, & Wagner, 1985; Janssens et al., 2001; Kelly, Asgharian, Kimbell, & Wong, 2004; Strong & Swift, 1987; Zwartz & Guilmette, 2001). In general, in vitro experiments have advantages of better control of experimental conditions and repeatability, and avoidance of any potential adverse affects on human test subjects.

Numerical investigations of the flow and/or the aerosol deposition in the nasal tract have been performed by Tang et al. (2004), Schroeter, Kimbell, and Asgharian (2006), Inthavong et al. (2006), Tian et al. (2007), Liu, Matida, Gu, and Johnson (2007), Shi, Kleinstreuer, and Zhang (2007) and Shanley, Zamankhan, Ahmadi, Hopke, and Cheng (2008). Yu et al. (1998) performed simulations of particle diffusion in a human upper respiratory system and Sarangapani and Wexler (1999) performed a numerical analysis using an extrathoracic model (i.e., nasal to larynx).

To the authors’ knowledge, all in vitro experiments and numerical investigations that have appeared in the literature have used the geometry of isolated subjects, and could be expected to lack some universality. The present in vitro and numerical deposition investigations are performed in a standardized human nasal geometry (median-average of 60 nasal airways from 30 subjects), referred to as the Carleton-Civic standardized nasal geometry (Liu, Johnson, Matida, Kherani, & Marsan, 2009).

As reported by Bilgen, Arbour, and Turik (1974) and Tang et al. (2004), the inhalation and exhalation flow rate during normal breathing, as occurs during relaxation, rest or sleep, is usually 7.5–12 L/min (or 125–200 mL/s) and the main flow features are periodic and laminar flow. Periodically, such as during exertion, sickness (asthma, nasal polyp, heavy vibrissae, sneezing, etc.) and physical exercise, the inhalation and exhalation flow rates can reach approximately 12–38 L/min (200–625 mL/s) and the main patterns of flow include constricted flows, obstructed flows, extra mucus flows, and turbulent flows (Tang et al., 2004). Even higher flow rates, 150 L/min (2500 mL/s), can be reached during extreme and forced inhalation conditions (Hooper, 2001). In the present work, inhalation flow rates of 30, 60, and 90 L/min were considered, which would be representative of conditions during nasal drug delivery (elevated inhalation rates compared to normal breathing). Some additional simulations were run at 15 and 45 L/min flow rates to cover a broader range of conditions.