An Investigation of Ionization Technology for Cleaning Cabin Air in a Business Jet ^†

Abstract

This paper describes an experimental investigation on the spread of a virus in a business jet cabin and the potential of ionization to reduce the pathogen load. In contrast to priorly investigated recirculation air cleaning, ionization can act directly in the cabin by introducing ions into the supply air. Tests were performed by emitting a surrogate virus through a breathing head in a business jet mock-up. The results allow for the conclusion that ionization technology, along with increased airflow, is a well-suited tool to sanitize cabins. Additionally, the effect of ionization on particles was investigated where it became obvious that the presence of particles reduces the ion level; however, the presence of ions hardly impact particles.

1. Introduction

This paper presents an experimental investigation on the spread of a virus in a business jet cabin. It furthermore evaluates the potential of an ionization system to reduce the pathogen load in the cabin air, on the table surfaces, and the carpet.

Since the COVID-19 pandemic much more focus has been placed on hygiene in aircraft cabins. Nevertheless, investigations and descriptions of cross-infection in aircraft cabins were initiated even before the outbreak of the SARS-CoV-2 pandemic. Gupta et al. [1] simulated droplet dispersion in a twin-aisle aircraft cabin section using CFD (Computational Fluid Dynamics). When an influenza-infected passenger breathes, coughs, or speaks, he thereby emit an aerosol in the cabin. Yihuan et al. [2] used a similar approach for a single-aisle cabin. Furthermore, Gupta et al. [3] demonstrated the reduction in infection risk by wearing a mask. In a literature review, Stover and Weiss [4] stated that the risk of cross-infection is 6% for the two rows in front and behind an infected passenger. Outside this area, the risk is still 2%. Crew movements, as well as contact during boarding/de-boarding and at the gate, are discussed as reasons. As one of the first documented cross-infections of SARS-CoV-2 in an aircraft, Eldin et al. [5] reported that a likely cross-infection of a passenger occurred on 24 February 2020 on an Air France flight.

In previous work by the authors [6], different air cleaning alternatives (Plasma and UV-C) to state-of-the-art HEPA filtration were investigated for airliner cabins. However, none of these showed an advantage over the already used HEPA filtration. While medical tests and norms [7] typically ask for a reduction of at least one log level (i.e., factor of 10), only about 50% of the cabin air is recirculated. Hence, even a perfect cleaning can only ensure a 50% reduction by recirculation air cleaning. One alternative, to provide more direct cleaning, could be the injection of ions directly into the cabin. Such ionization systems promise to inactivate pathogens by the reaction between the air ions and the pathogen’s cell surface [8]. However, a functional experiment in an aircraft cabin has not yet been performed for such a system.

While several studies (e.g., [1,2,3,6,9,10]) have investigated particle spread and derived infection risk in airliner cabins, there is, to the authors’ knowledge, no similar investigation for the business jet cabin. In the frame of the Clean Sky 2 Airframe project, this research gap was filled by investigating the effect of ionization on pathogen inactivation in a business jet cabin demonstrator.

2. Method

Figure 1 shows the business jet cabin demonstrator and its instrumentation. The cabin linings are made of translucent polycarbonate, as former projects required the visibility of the insulation. The demonstrator has a diameter of 2.25 m, a floor area of 3.6 × 1.69 m², and a cabin center height of 1.84 m. Test equipment is summarized in

Table 1. Test equipment.

Equipment	Device Information
Breathing head with virus emission	Breathing machine (9 breaths/min, 2 L/breath) pump module MT149 (InfoTech GmbH, Holzminden, Germany), connected to the head and to an aerosol nebulizer AGK 2000 (Palas GmbH, Karlsruhe, Germany)
Aerosol concentration	Light scattering aerosol spectrometers Fidas-Frog (Palas GmbH, Karlsruhe, Germany), measurement range 0.18–20 μm
Ion concentration	Ionometer IM 806 (Umweltanalytik Holbach GmbH, Wadern, Germany), measurement range 0–40 million ions/cm3
Virus air sampling	Air germ collector MBASS30V3 (Umweltanalytik Holbach GmbH, Wadern, Germany), collecting 1 m3 of air at a rate of 50 L/min (sampling interval 20 min)
Supply airflow rate	Thermal anemometer, Schmidt SS20.500 (Schmidt Technology GmbH, St. Georgen / Schwarzwald, Germany), measurement range 0.06–35 m/s, ±3%
Airflow velocity (dynamic pressure)	Differential pressure transducer, Ahlborn FDA602S2K (Ahlborn Mess- und Regelungstechnik GmbH, Holzkirchen, Germany), measurement range ±250 Pa, resolution 0.1 Pa

. A patented combination of a breathing machine, an aerosolizer, and a Sheffield head [11] simulates an infectious passenger. Two heated dummies are placed in the cabin next to and opposite the breathing head. They emit 75 W each and have a human-like temperature distribution. Aerosol and ion concentration are continuously monitored.

A Phi6 phage suspension is used as the viral aerosol. These phages are not pathogenic to humans and have properties similar to SARS-CoV-2, making them a suitable surrogate. Samples of air for surrogate viruses are taken opposite the breathing head at 2 m distance. The seating area table is sampled by wiping a defined area of 5 × 5 cm². To investigate the effect of ions on viruses on carpet, five drops of viral solution are initially placed on a separate piece of cabin carpet and placed on the cabin floor. After test phase 1 and phase 2 (

Table 2. Test conduct.

Phase	Time in min	Tasks
Phase 1	0	Start ventilation Start breathing head with virus surrogate aerosol emission Ionization inactive Place carpet sample 1 Cover table sample 2
	70	Activate air sample 1
	90	Retrieve air sample 1 gel filter Take table wipe sample 1 Retrieve carpet sample 1
Phase 2	90	Activate ionizer Uncover table sample location 2 Place carpet sample 2
	160	Activate air sample 2
	180	Retrieve air sample 2 gel filter Take table wipe sample 2 Retrieve carpet sample 2
Phase 3 (1 test)	180	Switch off breathing head and aerosol production Ionization active
	215	Air ions stabilization reached

), the respective carpet pieces are retrieved, and the virus drops are brought into solution. The number of collected active viruses in the sample (virulence) is determined in the laboratory using the plaque assay test [12].

The used ionizer is a prototype and is integrated in the supply air duct of the cabin demonstrator’s ventilation (Figure 2). Two electrodes supplied with power generate air ions inside the duct. The airflow rate is measured upstream the ionizer.

The test consists of two test phases (Table 2). In phase 1, the ventilation is activated without the ionization device being active, but surrogate virus is emitted through the breathing head. A first carpet sample is placed on the cabin floor. One half of the table wipe sample location gets covered to avoid virus buildup on this side in phase 1, while the other remains uncovered. For the last 20 min an air sample is drawn opposite the breathing head followed by taking a wipe sample of the uncovered table and retrieving the carpet sample. In phase 2, a new carpet sample is placed on the floor, the second half of the table wipe location is uncovered, and the ionization device is activated for the next 90 min. Again, an air sample is drawn in the last 20 min followed by taking a table wipe sample and retrieving the carpet sample. This procedure allows for the comparison of virulence with and without the ionizer being active. As the surrogate virus solution lots potentially have different concentrations and even vary within a days’ time, all comparisons are made on a percentage level compared to phase 1. In one of the tests, a third phase was added to assess the air ions level in the cabin when no particles are emitted.

Two cabin ventilation air renewal times were tested with virus measurements: 7.0 ± 0.1 min and 3.5 ± 0.1 min. Additionally, the ion levels during aerosol emission were measured for 9.1 ± 0.1 min cabin air renewal time.

In a pretest, the breathing head was characterized. As thermal anemometers showed a thermal inertia that was too high to capture the cycle of inhalation and exhalation, a pitot tube was used to measure the dynamic pressure. Particles were drawn into the aerosol spectrometer through a tube. The measurement was subsequently moved away from the mouth opening (Figure 3). The cabin ventilation was set to a cabin air renewal time of 7.0 min for the characterization of the breathing head.

Without airflow, the pressure transducer showed fluctuations of 0.2 Pa; therefore, this was defined as the minimum threshold for a valid measurement (corresponding to approx. 0.6 m/s). Due to the laboratory site elevation of 694 m, the air density considered is 1.1 kg/m³. The flow velocity is computed by the Bernoulli equation for dynamic pressure (1):

$$v=\sqrt{{\displaystyle \frac{2·p_{dyn}}{1.1\frac{\mathrm{k}\mathrm{g}}{\mathrm{m}³}}}}$$

(1)

3. Results

3.1. Breathing Head Characterization Pretest

The flow velocities were calculated using Equation (1). Per breathing cycle only three measurements (blue rectangles in Figure 4) were above the quality threshold of 0.2 Pa of dynamic pressure in the mouth opening during exhalation. During inhalation, no pressures higher than 0.2 Pa (absolute) were detected. Hence, breathing out results in directed airflow, while breathing in results in undirected aspiration from the sides. We attempted to approximate the measurements by a sinus function (black line) depicted in Equation (2). Further downstream (>0.5 m) the breathing head, no velocity could be detected any more.

$$v=\left\{\begin{array}{cc}sin\left({\displaystyle \frac{9}{60}}·\pi ·t\right)·1.3{\displaystyle \frac{\mathrm{m}}{\mathrm{s}}}& breathing out\\ below limit,<0.6{\displaystyle \frac{\mathrm{m}}{\mathrm{s}}}& breathing in\end{array}\right.$$

(2)

The particle concentration, however, appeared to decline with increased distance from the breathing head (Figure 5). The vertical bars depict the measurement’s standard deviation. In the first meter downstream the breathing head, a major decline in particles is found. As the virus air sampling is performed 2 m away from the breathing head, it is not considered to be majorly impacted by the gradient of the particle emissions.

3.2. Air Ion Levels

Figure 6 (top) shows the measured 30 min average ion levels in the air inlet and in the cabin. The vertical bars depict the standard deviation of the measurements. In test phase 1 (with aerosol emission and without ionization), the background concentration is measured to be between 10² and 10³ ions/cm³. When switching on the ionizer (phase 2), the levels of air ions increase for all three cabin air renewal times. A clear trend of reduced inlet air ion concentration with increasing cabin air renewal time (decreasing flow rate) is obvious (black arrow). At the cabin measurement position, only the 3.5 min and 7.0 min tests show major increases in levels compared to the background concentration, while in the 9.1 min case the ion concentration remains in the order of magnitude of the background concentration.

The measured aerosol concentrations in the 3.5 min case were 4557 particles/cm³ in phase 1, 5508 particles/cm³ in phase 2, and dropped to 91 particles/cm³ in phase 3. Leaving the ionizer activated, it becomes obvious that the air ion concentration measured in the cabin increases, while it remains constant at the air inlet (green arrow). Hence, the aerosolized particles apparently consume parts of the ions. Activating the ionizer, however, hardly impacts the particle concentration.

The measured air ion concentrations prove that the ion emission does not behave like the “classical” source dilution system. For example, if, instead of ions, a tracer gas was released into the supply air flow, an increase in flow rate (reduction in cabin air renewal time) would lead to higher dilution and thus lower measured concentrations. With the ions exactly the opposite is found; increased flow rates lead to increased measured air ions concentrations.

A second deviation from the classical source dilution system is that if a tracer is injected into the supply air, the concentration in the cabin and the supply air should be similar at steady state. However, the air ions show approx. one order of magnitude (~factor 10) lower concentrations in the cabin than in the air inlet.

Hence, the measurements suggest that the air ion concentration tends to decline, and thus proves to be unstable. Instantaneously after their production, ions start to recombine and the duration of air travel from the emission to the measurement location is crucial for the level of air ions measured. During pretesting, the ion measurement device was once held into the duct branch directly behind the ionizer and the measurement read millions of ions per cm³. The air ion level very much depends on distance and, thus, the travel time between the source (ionizer) and the location the air ions are supposed to be active (Figure 6, bottom). If they travel too long, the system becomes obsolete.

3.3. Virulence Reduction

The virulence reduction is determined by Equation (3)

$$reduction={\displaystyle \frac{sample1-sample2}{sample1}}$$

(3)

Thus, a positive reduction means that the virulence in sample 2 is lower than in sample 1; negative means an increase in virulence. Samples of the surrogate virus were taken in the 3.5 min and 7.0 min cabin air renewal time cases.

Table 3. Virulence reduction test results.

Location	7.0 min	3.5 min
Air sample	30%	67%
Table sample	-	−22% */−47%
Carpet sample	−112%	99%

* below threshold for valid detectability.

summarizes the results. The air samples show a higher reduction when decreasing the cabin air renewal time (increasing airflow), and thus increasing the air ion levels in the cabin. For the 7 min renewal case, the table sample unfortunately got contaminated and therefore is not reported here. In the 3.5 min case, the −22% reduction detected was below the formal level of virulence detectability; therefore, this sample was redone and revealed a −47% reduction. Hence, the ventilation and air ions were not able to reduce the virulence at the table. The carpet sample showed an increase in virulence in the 7 min test, while a decrease of 99% is obvious in the 3.5 min test. Air ion levels were not measured at the floor level, so it cannot be stated whether some ions may have reached the carpet and acted there. One difference between the table and carpet is that the table was initially clean and received its surrogate virus load during the test, while for the carpet, five drips of the viral solution, that was further used in the breathing head, were introduced, and thus the carpet was pre-charged.

4. Discussion and Conclusions

A test of ionization technology’s ability to sanitize the business jet cabin was performed in combination with increased airflow. This study shows that ionization allows the cabin to reach positive virulence reduction in the breathing air when the ionizer is installed in such a way that the elapsed time to cabin outlet is low.

As measurement instrumentation and personnel capacity is always limited in such test campaigns, further details that would be useful could not be measured in the test conduct. Ideally, each sampling location should have been accompanied by a measurement of the local ion level, especially as the decline of the ions proved to be very quick. As such data are not retrievable from the measurement campaign, the current results can indicate that the system works, but still lack the data to be able to generate a performance map of the ionization technology. This test series’ statistical relevance would increase by repetition using the same conditions and procedures. For the table sample, this was tested once as the first sample was below the threshold of valid detectability and showed a variation from −22% to −47% virus reduction, thus a 25% difference.

Despite these possible test conduct optimizations, the conclusion is drawn that the ionization system achieves sanitation of the cabin air if the travel time of the ions is low. Future research should focus on the potential to scale this system to airliner cabins, and especially investigate the achievable ion concentration. In an airliner cabin, the cabin fresh air renewal time is in the order of 5 min; however, ducts are potentially longer due to the larger size of the aircraft.

The decline rate in air ions per second should be investigated to obtain the half-life time for the ion recombination to support the system design. Preliminary analyses using duct diameters and flow rates in the business jet demonstrator ventilation combined with the measurement of the elapsed time between injected fog entering through the air outlet and reaching the ion measurement position allow for a first estimate that 5 s of travel should not be exceeded for the ions.

Furthermore, ionization technology should be investigated for its capability to inactivate other pathogens, such as non-encapsulated viruses, bacteria, and fungi.

5. Patents

The breathing head with aerosol emission is patented: EP 4 194 837 A1: Aerosolgenerator und Verfahren zur Abgabe eines Aerosols, published 14 June 2023.