Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide

G Model

ARTICLE IN PRESS

SNA-111691; No. of Pages 11

Sensors and Actuators A xxx (xxxx) xxx

Contents lists available at ScienceDirect

Sensors and Actuators A: Physical journal homepage: www.elsevier.com/locate/sna

Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide Zaid B. Jildeh a,b,c , Patrick Kirchner a , Jan Oberländer b , Farnoosh Vahidpour b , Patrick H. Wagner c , Michael J. Schöning b,∗ a

Imagine Engineering GmbH, Walter-Gropius-Str. 30, 50126 Bergheim, Germany Institute of Nano- and Biotechnologies (INB), FH Aachen, Campus Jülich, Heinrich-Mußmann-Str. 1, 52428 Jülich, Germany c Laboratory for Soft Matter and Biophysics, KU Leuven, Celestijnenlaan 200 D, 3001 Leuven, Belgium b

a r t i c l e

i n f o

Article history: Received 1 April 2019 Received in revised form 30 September 2019 Accepted 22 October 2019 Available online xxx Keywords: Sterilization process Finite-element method Hydrogen peroxide Condensation Sterility test Aseptic filling machines

a b s t r a c t Within the present work a sterilization process by a heated gas mixture that contains hydrogen peroxide (H2 O2 ) is validated by experiments and numerical modeling techniques. The operational parameters that affect the sterilization efficacy are described alongside the two modes of sterilization: gaseous and condensed H2 O2 . Measurements with a previously developed H2 O2 gas sensor are carried out to validate the applied H2 O2 gas concentration during sterilization. We performed microbiological tests at different H2 O2 gas concentrations by applying an end-point method to carrier strips, which contain different inoculation loads of Geobacillus stearothermophilus spores. The analysis of the sterilization process of a pharmaceutical glass vial is performed by numerical modeling. The numerical model combines heat- and advection-diffusion mass transfer with vapor–pressure equations to predict the location of condensate formation and the concentration of H2 O2 at the packaging surfaces by changing the gas temperature. For a sterilization process of 0.7 s, a H2 O2 gas concentration above 4% v/v is required to reach a logcount reduction above six. The numerical results showed the location of H2 O2 condensate formation, which decreases with increasing sterilant-gas temperature. The model can be transferred to different gas nozzle- and packaging geometries to assure the absence of H2 O2 residues. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Hydrogen peroxide (H2 O2 ) is a widely used disinfectant and sterilizing agent in healthcare and pharmaceutical industries [1,2]. Due to its strong and fast microbicidal properties, it is employed in aseptic filling machines for the sterilization of packaging surfaces prior to product filling [3]. An advantage of using H2 O2 over other chemical sterilants such as ethylene oxide are the environmentalfriendly, non-toxic and odorless end-products of water (H2 O) and oxygen [1,4]. H2 O2 tends to decompose when it comes in contact with organic and inorganic substances, which include some packaging materials and the cell membrane of microorganisms [5–8]. The decomposition of H2 O2 follows a long reaction chain that produces free radical species [8,9]. These molecular species have a high reactivity that interfere with the cell membrane, DNA and protein molecules,

∗ Corresponding author. E-mail address: [email protected] (M.J. Schöning).

leading to the inactivation of microorganisms [1,9]. Recent analysis by transmission-electron microscopy (TEM) of Bacillus atrophaeus spores points to the destruction of the cell wall and the loss of cell matter as a possible cause of inactivation [10]. A common method of H2 O2 application is in the gas phase. A typically used food-grade H2 O2 solution consists of a miscible mixture of 65% w/w (weight fraction) H2 O and 35% w/w H2 O2 with traces of stabilizing agents and impurities. The package sterilization using H2 O2 gas follows a three-step process, which corresponds also to the three physical components of a sterilization apparatus [3,11]. These are:

1. The conditioning or pre-treatment of the packaging material. Here, a stream of heated sterile air is directed at the packaging surfaces to increase the surface and bulk temperatures of the package by convection to avoid strong condensation of H2 O2 gas in the following step. 2. The surface sterilization of packages by a heated gas mixture of air, H2 O2 and H2 O. This gas mixture is generated from an evaporation unit that is similar in function to a heat exchanger.

https://doi.org/10.1016/j.sna.2019.111691 0924-4247/© 2019 Elsevier B.V. All rights reserved.

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11 2

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

A detailed description of an exemplary design is given in Refs. [12–15]. The volumetric concentration of H2 O2 in the gas mixture depends on the process window and the operational parameters of the filling machine. In this case, a volumetric concentration up to 8% v/v (volumetric fraction) for a short sterilization time window below 1 s may be required to reach up to a six-log count reduction rate for test microorganisms [10,16], as required by standards and regulations [1,17,18]. 3. The package aeration or post-treatment with a stream of heated sterile air is done prior to product filling to reduce the residual H2 O2 concentration on the packaging surface. The heated air aids in the evaporation of possible condensate and in the removal of H2 O2 from the package by air substitution [19]. By controlling the operational properties of the first and the second step, two sterilization principles can be achieved: gas-phase sterilization and sterilization by (micro-)condensation [20–22]. These operational parameters include the applied sterilant-gas temperature, exposure time, volumetric flow rates, humidity level and H2 O2 concentration [1,3,4]. A gas-phase sterilization implies that the temperature on the package surface is higher than the dewpoint temperature of the H2 O2 -containing gas at the operational pressure of the sterilization system. Alternatively, sterilization by (micro-)condensation occurs when: i. the temperature of the gas mixture is lower than its dew-point temperature, ii. the H2 O2 - and H2 O content in the atmosphere reach saturation level, iii. the temperature of contact surfaces is lower than the dew-point temperature of the gas mixture. Any of these points can lead to condensation that is a function of the surrounding temperature and pressure. Klemm et al. observed that the evaporation of an aqueous H2 O2 solution mixed with a carrier stream of nitrogen gas occurs at a lower temperature than the dew-point temperature of a gas mixture containing only H2 O2 and H2 O [23]. Therefore, the partial pressure of the carrier-gas stream has an effect on the dew-point temperature of the H2 O2 -H2 O gas mixture. Due to the fact that H2 O2 has a lower vapor pressure than H2 O, it tends to condense first [24,25]. This produces a thin film of concentrated H2 O2 that may have an even stronger microbicidal effect than gas-phase sterilization [19,21]. Eschlbeck et al. studied the inactivation kinetics of test microorganisms on the surface of different carrier materials corresponding to packaging surfaces [26]. They found that H2 O2 sterilization, in which condensation is allowed to take place, results in higher inactivation kinetics of spores on a hydrophobic surface. This can be explained by the fact that the spores act as condensation nuclei for the H2 O2 condensation [26]. Consequently, depending on the characteristics of the packaging surface, such as its wettability [3], tuning of the sterilization-system parameters is required to regulate the prevailing sterilization principle and ensure product safety. The design of the sterilization process and the setting off its operational parameters depend on the thermal and chemical compatibility of the packaging material. For instance, exposure to heat can cause degradation of typical packaging polymers leading to a change in their chemical or mechanical properties [27]. Hence, setting up the operational condition of the sterilization system to achieve a specific sterilization principle involves a wide range of validation tests. Furthermore, sterilization by (micro)condensation forms a condensate film that affects the operational parameters of the post-treatment step. As a result, the overall design of the sterilization process aims not only to inactivate all present microorganisms, but also to maintain a residual H2 O2 concentration below 0.5 ppm (according to the US Code of Federal

Regulations: 21 CFR 178.1005 (c)). Consequently, to determine the optimal operational conditions of the sterilization process, one requires the quantification of residual H2 O2 on the package surface. One method to approach this is to use numerical analysis methods, in particular in situations where data acquisition by sensing elements is not possible due to dimensions of the package. For example, Spanu and Vignali used numerical models to optimize a pouch-package sterilization process using vaporized H2 O2 in terms of nozzle location and cost reduction of H2 O2 consumption [28]. Moreover, Fisher and Caputo used a numerical technique to analyze the distribution of H2 O2 in a barrier-based isolator decontamination system and assist in determining the regions with the lowest H2 O2 concentration [29]. In this work, we propose a surface-sterilization process for pharmaceutical vials that combines the three described steps of H2 O2 sterilization in a single system. To validate the sterilization efficacy of the developed process, we combined measurements of the H2 O2 gas concentration during the sterilization step with microbiological challenge tests at varying dosing volumes of a 35% w/wH2 O2 solution. The determination of the gaseous H2 O2 concentration is carried out by using a calorimetric H2 O2 gas sensor that was previously developed as described in Refs. [16,30]. The microbiological tests are performed with commercial test strips of Geobacillus stearothermophilus spores (DSMZ 5934, gke GmbH, Waldems, Germany) as the recommended challenge microorganism due to their resistance to H2 O2 and heat [4,18,31–33]. To determine the efficacy of the sterilization process, results from the microbiological tests are analyzed using a combination of an endpoint method as described in Ref. [34,35] and the calculation of most-probable number (MPN) method as described in Refs. [36,37] to determine the log-count reduction (LCR). This value depicts a logarithmic representation of the decrease in the number of spores from an original culture value and is used as an indication of the sterilization efficacy [38,39]. The analysis of the sterilization process that occurs in the small pharmaceutical vial is challenging due to the relatively small size of the package. Therefore, we describe a numerical model based on a finite-element method (FEM) to analyze the sterilization process. The numerical model combines non-isothermal turbulent flow, temperature-dependent diffusive mass transfer, heat transfer in solids, and empirical equations for the vapor pressure of H2 O2 -H2 O mixtures. This model is capable of estimating the distribution of H2 O2 gas in a package, the theoretical condensation profile of H2 O2 and the mass fraction of each phase (liquid or gas) at varied modeling conditions. These conditions include: ambient parameters (temperature or humidity), geometric designs (package and gas-delivery nozzle) or inlet-flow conditions of the sterilizing gas (volumetric concentrations). For the current example, we focus our analysis on the effect of applying different sterilantgas temperatures. Throughout the validation and the numerical modeling, we consider a sterilization process with an exposure time of 0.7 s to H2 O2 gas throughout the aseptic filling line. This time reflects the continuous development in the field of packaging industry. This trend, which is toward a higher machine throughput, means the development of machines such as “Siedel wet aseptic line” with a production capacity of 40,000 units/hour (uph); “OPTIMA Pharma VFVM” with 30,000 uph; “Tetra Pak A3/Speed” with 24,000 uph; and “Romaco MicroMaxx Aseptic injectable powder filling machine” with 24,000 uph.

2. Experimental set-up The two leading factors that affect the sterilization efficacy of H2 O2 are the exposure time and the concentration of H2 O2 in the gas mixture [18]. Here, we defined an exposure time of 0.7 s for

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

3

Fig. 2. Developed calorimetric gas sensor used for determining the H2 O2 gas concentration in the gas mixture. The Image shows the differential setup of active (with a manganese (IV) oxide catalytic surface) and passive temperature element developed on a polyimide substrate. The solder-contact points and the circuit paths are chemically passivated by a silicone coating (RTV 118, Momentive Performance Materials GmbH, Leverkusen, Germany).

Fig. 1. Scheme of the developed H2 O2 gas-sterilization system. A detailed description of the design of the installed evaporation unit can be found in Ref. [14,15].

the sterilization step. To test the packaging-sterilization device, we have validated the H2 O2 gas concentration in the sterilization step at different volumetric flow rates of H2 O2 solution. Hence, we compared the concentration of H2 O2 gas in the outlet-gas mixture with the analytically derived values. If a difference is observed between the two values, catalytic or thermal decomposition of H2 O2 takes place on the surfaces of the evaporation unit. For more details on the design and development procedure of the utilized evaporation unit, we refer to Ref. [15]. In the second experiment, we determined the correlation between the volumetric flow rate of H2 O2 solution and the sterilization efficacy of the system. For this purpose, we applied microbiological challenge tests based on the end-point method using spore-test strips of G. stearothermophilus. Fig. 1 presents a realistic model of an industrial sterilization apparatus that is undergoing development. For both tests, the calorimetric H2 O2 gas sensor and the spore-test strips of G. stearothermophilus are attached to a carrier on the transport system. The inflow of H2 O2 solution is set by the dosing pump and the inlet valve located at the H2 O2 dosing unit above the evaporation unit. By using a calibrated H2 O2 gas sensor, the H2 O2 gas concentrations for the experiment are set to 0, 2.3, 4.2, 5.8, 7.0 and 7.5% v/v, respectively. The calorimetric H2 O2 gas sensor, shown in Fig. 2 and described in Ref. [30], is used for the fast detection and quantification of the H2 O2 gas concentration in the outlet-gas stream of the evaporation unit. The sensor consists of two platinum-meander structures that act as the temperature-sensitive elements, where a change in gas temperature results in a detectable change of the electrical resistance. One of the elements (passive) serves as a temperature reference for the H2 O2 -containing gas. The other sensing element (active) is coated with a catalytic material (e.g., manganese (IV) oxide) that induces an exothermic, catalytic decomposition reaction of H2 O2 . The temperature difference between the active and passive element is proportional to the concentration of H2 O2 in the gas mixture. For more information about the working principle of the sensor, we refer to Refs. [40,41].

To test the sterilization efficacy for the various applied concentrations of H2 O2 gas, the end-point method is carried out. In this method, polyethylenterephthalat (PET) strips are inoculated with 104 , 105 and 106 colony-forming units (CFU) of G. stearothermophilus spores. The LCR value is then determined by applying the MPN method as explained in Ref. [36]. To achieve statistically relevant results for the sterilization efficacy, at each H2 O2 gas concentration and each microbiological stage we repeat the exposure of the microbiological test strip to H2 O2 gas seven times. Therefore, each spore strip is placed in a vial that contains a nutrient medium with a pH-sensitive color indicator (gke GmbH, Waldems, Germany) and incubated for seven days after the exposure to the H2 O2 gas. The incubation temperature is maintained by a laboratory-incubation unit (Labnet International, USA) at the recommended temperature of 55 ◦ C. A change in the color of the nutrient medium from purple to yellow represents a change of the pH value in the solution that indicates in turn cell metabolism and viability [27]. 3. Numerical model A pharmaceutical glass vial with a geometry based on a SCHOTT 6R vial (Pharmaceutical Systems Schott AG, Germany) is presented in Fig. 3. The vial has a vertical axis of symmetry that is used to simplify the geometry from a three- to a two-dimensional model. This reduces the required computational resources and allows using an advanced turbulent-flow physics and finer mesh elements around the regions of interest, especially the inner surfaces of the vial. To sterilize both the inner- and outer-vial surface, a guiding plate fixed on the distribution nozzle is introduced to direct the gas flow. To simplify the operation of the sterilization system and to decrease the number of components of the sterilization system, we combined the pre-treatment, the gas treatment with H2 O2 and the post-treatment steps into a single treatment step. The goal of this design is to approximate a system that uses a single heat exchanger (evaporation unit) for all treatment steps. The exemplary modeling scenario follows an isochronic sterilization process of a total time of 2.1 s: The three steps of pre-treatment, gas treatment (sterilization) and post-treatment take 0.7 s each as illustrated in Fig. 4. For our model, we defined a fluid inflow to the heat exchanger of 2.5 m3 /h dry air (for each step) and 6 ml/min of aqueous H2 O2 solution (gas-treatment step) resulting in a H2 O2 gas concentration of approximately 3.5% v/v. For the numerical model, the material properties of the resultant gas mixture were adopted from Ref. [15]. Due to the inner-vial volume of 7.5 ml, we estimated that this inflow is sufficient for a simultaneous sterilization (batch treatment) of 10 vials using a pipe-manifold system to distribute the gas mixture. Point (A) at 1.05 s in Fig. 4 represents the time section in the mid of the treatment process with H2 O2 gas, which is used to analyze the numerical results.

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

4

Fig. 3. (a) Three-dimensional geometry of the glass vial showing the line of axisymmetry and (b) cross-sectional view through the geometry. The glass vial is shown in cyan and the gas nozzle of the sterilization system as well as the guiding plate to direct the gas flow is shown in green. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

has the same initial temperature as the outlet gas. To approximate the heat losses to the atmosphere, the bottom of the glass vial and the outer surfaces of the guiding plate (refer to Fig. 3) have a constant convective heat-transfer coefficient of 25 W/(m2 ·K) with an external temperature of 20 ◦ C. 3.2. Single-phase turbulent flow The momentum equation in a transient simulation is mathematically represented by the Reynolds-averaged Navier-Stokes (RANS) equation, shown in Eq. (2) [42].





Fig. 4. Overview of the sterilization-system operation showing the pre-treatment, sterilization and post-treatment process, respectively. Point (A) indicates the moment for the analysis of the numerical results.

The physics of the numerical model includes turbulent fluid flow coupled with heat-transfer physics and mass transfer by diffusion. To predict possible condensation of the gas mixture and the concentration of H2 O2 , vapor–pressure equations of H2 O2 and H2 O vapors are used. As described in Section 1, temperature and pressure are the main parameters that affect the dominant sterilization principle (gas phase, or condensation or a mixture thereof). Since the treatment process takes place at atmospheric pressure of 1013.25 hPa, we are interested especially in the effect of gas temperature on the formation of condensation. 3.1. Heat transfer in solids and fluids The conjugate heat transfer in the numerical model is described by the transient heat equation, given in Eq. (1) [42]. Cp

∂T + Cp u · ∇ T + ∇ · (−∇ T ) = Qt ∂t

(1)

Here  is the density of the material in kg/m3 , Cp is the heat capacity at constant pressure of the material in J/(kg·K), T is the temperature at the Kelvin scale, t is the time in seconds, u is the velocity field of the fluid in m/s,  is the thermal conductivity in W/(m·K) and Qt is the net rate of heat transfer in W/m3 . In the numerical model, we assume a constant temperature of 20 ◦ C for the surrounding atmosphere. Additionally, the gas nozzle



2 ∂u + u · ∇ u = ∇ · −pI + (∇ u + (∇ u)T ) − (∇ · u)I 3 ∂t

(2)

Here p is the pressure of the fluid in Pa, I is an identity matrix and  is the dynamic viscosity of the fluid in Pa·s. Eq. (2) is solved taking into account the continuity equation, shown in Eq. (3) [42].

∂ + ∇ · (u) = 0 ∂t

(3)

Eq. (3) states that a change in the fluid density through, for example, a change of temperature is accompanied by an inversely proportional change in flow velocity. This equation maintains the conservation of mass in the numerical model. 3.3. Mass transfer by diffusion Eq. (4) represents the transient form of the mass-transfer equation [42]. This equation combines Fick’s first law with the bulk-mass transfer (fluid flow).

∂ci + ∇ · (−Di ∇ ci ) + u · ∇ ci = ri ∂t

(4)

Here ci is the molar volume of the gas-mixture components in mol/m3 , Di represents the diffusion coefficient of a component in m2 /s and ri is the reaction- or generation rate for a component in mol/(m3 ·s). The gas-mixture components in this model consist of air, H2 O2 and H2 O. Due to the small molar concentration of H2 O2 and H2 O in the gas mixture, a binary diffusion coefficient is introduced to the model. The Chapman-Enskog diffusion equation and Lennard-Jones potential data are used to estimate the non-isothermal diffusion coefficients of H2 O2 - and H2 O vapor in air, as shown in Fig. 5. For details on the analytical derivation, we refer to Ref. [43].

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

5

Fig. 5. Calculated non-isothermal diffusion coefficients of H2 O2 and H2 O gases in air at 1 atm pressure.

3.4. Vapor–liquid equilibrium Condensation takes place when the temperature of the surface is lower than the dew-point temperature of the gas mixture at a given concentration. In other words, condensation happens when the concentration of H2 O2 and H2 O at a surface reaches saturation level at the defined surface temperature. Accordingly, the dew-point depends on the present molar ratio or partial pressure of the gas mixture’s constituents. An estimation of the molar ratio and partial pressure of the H2 O2 -H2 O system can be made from the numerical model by tracking the distribution of the molar volumes of the mixture constituents. The dew-point temperature, however, is estimated from vapor–pressure equations of the H2 O2 -H2 O binary system at a defined molar concentration (partial pressure). For a pure H2 O2 -H2 O binary system, the relation between the temperature and mass fraction (w) of H2 O2 in the H2 O2 -H2 O binary system is represented in a T–w vapor–liquid equilibrium (VLE) diagram. By adding air, the molar ratio and partial pressure of H2 O2 -H2 O vapor mixture changes and that affects the dew-point temperature. For instance, a 3.4% v/v H2 O2 -containing gas mixture has a partial pressure of 155 hPa and a corresponding dew-point temperature of about 75 ◦ C at 1 atm. The VLE T–w of the described gas mixture is shown in Fig. 6(a). The effect of changing the partial pressure (concentration) of H2 O2 gas in the gas mixture can be described in a T–p VLE diagram, as shown in Fig. 6(b). The black arrow in Fig. 6(a) indicates the direction of cooling for a gas mixture that contains 3.4% v/v H2 O2 . The vapor–pressure equations and the method for creating the VLE diagrams of H2 O2 -H2 O are described in Refs. [15,25]. To determine the fraction of liquid and gas phases at an arbitrary point (X) in Fig. 6(a), the lever rule is applied as described in Eqs. (5) and (6), respectively. wL =

wX − wA wB − wA

(5)

wG =

wB − wX wB − wA

(6)

Here wL is the liquid-phase fraction and wG is the gas-phase fraction of H2 O2 ; wA , wB and wX are the mass fraction of points A, B and X, respectively. The resultant liquid and gas phase during cooling is shown in Fig. 7. When the bulk-gas temperature reaches a value below the dew-point curve (red colored) of 75 ◦ C, condensation starts to form. Below the dew-point temperature of approximately 74 ◦ C, a small portion of the gas mixture starts to condense. Using Fig. 7, we determine that the vapor from the original 35% w/wH2 O2 gas mixture consists of approximately 10% liquid and 90% gas. Also, from Fig. 6(a) we can see that the formed con-

Fig. 6. Calculated vapor–liquid equilibrium (VLE) diagrams for a H2 O2 -H2 O system: (a) T–w VLE diagram of a H2 O2 -H2 O system in a gas mixture containing air with a H2 O2 gas concentration of 3.4% v/v. The black arrow shows the direction of gas cooling and the dew- and bubble-point curves (red and blue) around the initial H2 O2 concentration of 35% w/w H2 O2 solution, (b) a T–p VLE diagram of a H2 O2 -H2 O system. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 7. Calculated phase-fraction diagram of a 3.4% v/v H2 O2 -containing gas mixture as a function of temperature, generated from Fig. 6(a) using the lever rule. The black arrow indicates the direction of gas cooling. (For interpretation of the references to color in the text, the reader is referred to the web version of this article.)

densate contains 75% w/w H2 O2 solution, while the gas phase has a H2 O2 concentration of 30% w/w. 4. Results and discussion 4.1. Sensor measurements Using the sterilization system described in Fig. 1, experiments are carried out to monitor the outlet H2 O2 gas concentration at continuous and pulsed dosing operation. For these experiments, the temperature of the gas mixture is regulated to a value of 240 ◦ C.

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11 6

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

port back to the loading chamber. This experiment is repeated for four times, corresponding to four sensor signals with 20 s break between each measurement. The recorded temperature value is lower than the regulated gas temperature at the outlet of the evaporation unit of 240 ◦ C. This is due to the mixing of the hot H2 O2 gas with the surrounding air and due to the short process time that prevents the sensor from reaching a steady-state signal. Furthermore, one can notice an increase in the peak temperature of both the active and reference element following each repeated experiment at the same H2 O2 gas concentration. Even with a stop time of 0.7 s under the nozzle of the evaporation unit, the sensor and the carrier retain some of the heat. Therefore, the following sensor measurement occurs at a higher initial temperature. This results in a slight increase in the peak of the sensor signal. Nonetheless, the increase in the peak temperatures is constant for the reference- and active elements of the calorimetric sensor. These peak-temperature difference, shown in Fig. 8(c), is used to monitor the H2 O2 gas sterilization at cyclic operation and to calibrate the sensor as described in Ref. [44]. The observed negative signal change before each peak is related to the difference in the reaction times of the active and passive sensor. The presence of a catalytic material on the active sensor surface causes a slightly slower rate of temperature increase as compared to the reference sensor (due to the additional thermal mass), which can also be seen in Fig. 8(d). 4.2. Sterility tests

Fig. 8. (a) Recorded H2 O2 gas concentration (data points) in the outlet-gas stream of the evaporation unit plotted alongside with the theoretical values (black line). (b) Sensor-active- and passive temperature data during cyclic measurement of 0.7 s shown in red and blue curves, respectively. (c) Temperature difference (active − passive) indicating an increase related to the increase in the inflow of H2 O2 solution. (d) Zoomed-in section of one of the measurement peaks showing the reaction of sensor elements. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 8 shows the recorded H2 O2 concentration in the outlet-gas mixture of the sterilization system using a pre-calibrated H2 O2 gas sensor. The measurements in Fig. 8(a) are compared with theoretical H2 O2 gas concentrations from mass-balance calculation. Here, the recorded results correlate with the analytically determined values and the operation of the sterilization system is validated. The variation between the two values at high inflow rates of H2 O2 solution can be related to multiple reasons including a variation in the dosing system or thermal decomposition at the surfaces of the evaporation unit, since a higher thermal power is required at a higher volumetric flow rate of H2 O2 solution. Fig. 8(b) indicates the sensor measurements during cyclic operation. With reference to Fig. 1, the transport system brings the carrier with the H2 O2 gas sensor (similarly: spore-test strips) from the loading chamber to the defined stop location below the outlet nozzle of the evaporation unit. To ensure that no condensation occurs on the surface of the sensor during the experiment, the temperature of the sensor is maintained above the dew-point temperature of the H2 O2 -containing gas by actively operating the sensor (Ohmic heating from the measuring current). The sensor measurement is carried out for an exposure time of 0.7 s, followed by trans-

To determine the sterilization efficacy of the H2 O2 gas at different gas concentrations, the microbiological tests are performed at varied inflows of H2 O2 solution. The volumetric inflow of air is kept at a constant value of 2.5 m3 /h and the outlet-gas temperature of the evaporation unit is regulated to a temperature of 240 ◦ C. For these tests, we prepared 6 sets of the three spore-suspension solutions of G. stearothermophilus (DSMZ 2934) to inoculate 104 , 105 and 106 CFU per carrier strip. Each set is repeated 7 times to achieve statistical significance. These biological test strips are exposed to a standard test protocol with a H2 O2 gas concentration of 0, 2.3, 4.2, 5.8, 7.0, 7.5% v/v and an exposure time of 0.7 s. The exposure time is set by controlling the stop time of the transport system below the gas nozzle of the evaporation unit and corresponds to the time used for the sensor tests. After the treatment of the spore-test strips with the sterilization system, each test strip is transformed to a vial containing a nutrient medium mixed with a pH-sensitive color indicator. These vials are then placed in an incubator for a period of one week at an incubation temperature of 55 ◦ C. Initially, the original color of the medium is purple as indicated in Fig. 9(a). After the exposure with H2 O2 gas and upon successful inactivation of G. stearothermophilus spores and other microorganisms on the test strip, the color remains unchanged. However, if viable microorganisms are present, the solution’s color changes to yellow indicating the growth and proliferation of the biological material as shown in Fig. 9b). At the end of the seventh day, we determined the sterility of each test sample by noting the color change of the incubation vial at the defined inflow of H2 O2 solution. Following a visual inspection, we defined a numeric value of 1 for nutrient-medium vials that show an observable color change (non-sterile sample) and a value of 0 otherwise. Table 1 summarizes the results of the microbiological tests carried out at different H2 O2 gas concentrations and spore loads. The number of non-sterile test samples out of a total of 7 samples for each spore load and test scenario is depicted in the table. A H2 O2 gas concentration of 0% v/v serves as a control for the experiment, which showed a state of non-sterility in all cases in the absence of H2 O2 in the outlet-gas stream. Increasing the H2 O2 gas concen-

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model

ARTICLE IN PRESS

SNA-111691; No. of Pages 11

Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

7

Table 1 Results of the microbiological tests using stripes loaded with spores of G. stearothermophilus exposed to H2 O2 gas at varying concentration. The digits under the “spore inoculation per test carrier” present the number of non-sterile samples out of the total of 7 test samples. Most probable number- and log-count reduction values are calculated following the instructions stated in Refs. [36,37]. The confidence interval of the calculations is described by a minimum (min.) and a maximum (max.) confidence value. H2 O2 gas concentration (% v/v)

0 2.3 4.2 5.8 7.0 7.5

Spore inoculation per test carrier (CFU)

Most-probable number

Log-count reduction

104

105

106

Mean

Min. conf.

Max. conf.

Mean

Min. conf.

Max. conf.

7 0 0 0 0 0

7 3 0 0 0 0

7 7 1 2 0 0

>194.59 5.26 0.30 0.14 <> <>

94.64 1.96 0.07 0.02 – –

– 14.10 1.20 0.98 0.91 0.91

<3.89 5.46 6.70 7.04 >7.07 >7.07

– 5.03 6.10 6.19 6.22 6.22

4.20 5.88 7.31 7.89 – –

Fig. 9. (a) Original vial of the nutrient medium with its purple color. (b) Vials of two different inoculated strips treated with 4.2% v/v H2 O2 concentration; left: color change of the solution for a treated strip inoculated with 105 CFU that indicates the presence of viable microorganisms and right: nutrient-medium vial containing a treated strip inoculated with 106 CFU that indicates total inactivation of microorganisms. Image was taken after one week of incubation at a temperature of 55 ◦ C. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

tration increased the chance of sterility of the samples. For the present system, a H2 O2 gas concentration above 7% v/v results in complete sterile samples (no color change). Based on the sterility of the test-strip sample presented in Table 1, the MPN method is calculated with a 95% confidence. Hence, with the mean, minimum and maximum confidence of the MPN, the LCR value are calculated and noted in the table. A detailed description of the calculation method and a tool to calculate the MPN value are provided in Refs. [36,37]. Here, the critical H2 O2 gas concentration required to reach a LCR value above 6 for an exposure time of 0.7 s lies between 2.3 and 4.2% v/v. Therefore, for the numerical model, we have analyzed the sterilization process at a H2 O2 gas concentration of 3.4% v/v. 4.3. Numerical models The analysis of the numerical results is carried out at a time point of 1.05 s. With reference to Fig. 4, this process time at point (A) presents the midway through the sterilization step. At a H2 O2 gas concentration of 3.4% v/v, we solved the numerical model at different gas-mixture temperatures from 120 ◦ C up to 270 ◦ C with a 30 ◦ C increment. Fig. 10(a) shows the velocity profile of a 120 ◦ C gas stream. Here, the guiding plate (green) helps in directing the gas to the outer surface of the glass vial. This allows for the sterilization of the inneras well as the outer-vial surfaces. The velocity is the highest near the exit of the gas nozzle, which is followed by a decrease in velocity due to the expansion of gases. Due to the low density of the heated gas mixture and the volumetric inflow of air and H2 O2 solution, mass transfer occurs

dominantly by turbulent flow. This enhances the mass transfer by turbulent mixing, which is faster than diffusion. Moreover, the gas molecules have a small collision diameter and due to their molecular size, they are able to diffuse into all corners of the packaging material such as corner folds, micro-cracks or crevices caused by any non-uniformity of the surface, which can harbor microorganisms [33]. Additionally, the H2 O2 gas molecules have a higher kinetic energy than in the liquid phase that leads to numerous random collisions with other neighboring molecules and with the cell walls of the present microorganisms. The collisions with the cell membrane of microorganisms lead to a decomposition reaction and the formation of reactive radical molecules that are responsible for cell inactivation [45]. At the process time of 1.05 s, the H2 O2 gas concentration is evenly distributed within the vial and equal to the outlet of the nozzle (3.4% v/v) as indicated by the uniform color in Fig. 10(b). This represents the outflow-gas concentration from the nozzle. From the concentration profile and Fig. 6(b), we are able to estimate the total partial pressure of H2 O2 and H2 O in the gas mixture as presented in Fig. 10(c). Here, the partial pressure reaches a value of about 155 hPa inside the glass vial. At the outer surface, the diffusion of the vapors and the mixing with atmospheric air decreases the molar concentration and the partial pressure of a gas mixture of H2 O2 and H2 O. With reference to the surface temperature of the vial (shown in Fig. 10(d)), it is possible to predict the formation of H2 O2 condensation. Therefore, one can define a state indicator for the three distinct regions on the vial surfaces. The first region indicates complete condensation of the gas. This occurs when the temperature of the surface is lower than the bubble-point temperature of the gas mixture. At this state the concentration of H2 O2 is equal to the original concentration in the gas phase (35% w/w). The second region is when no condensation forms, where the surface temperature is higher than the dew-point temperature of the gas mixture. The final region occurs when the surface temperature is between the bubbleand dew-point temperature. With reference to Fig. 6(b), at a partial gas pressure of 155 hPa, the bubble- and dew-point temperatures are 59.9 ◦ C and 75.7 ◦ C, respectively. To determine the liquid- and gas-phase fraction of the mixed condition and the H2 O2 concentration at each phase, information derived from the VLE diagram in Fig. 7 is required. Using the described method, we are able to estimate condensation location for the H2 O2 -containing gas mixture and the properties (concentration and phase fraction) of the condensate at different outlet-gas temperatures. Table 2 shows various plots at different inlet-gas temperatures from 120 ◦ C up to 270 ◦ C with 30 ◦ C increments. The presented subplots from left to right are: vial-surface temperature, state indicator (gas, condensation, or a mix state), phase fraction with 0 for a gas phase and 1 for a condensation, and the concentration of aqueous H2 O2 in the condensate. The description of Table 2 from left to right is as follows:

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11 8

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

Fig. 10. Results of the numerical model at a temperature of 120 ◦ C showing (a) the velocity streamline of the fluid flow, (b) the distribution of the H2 O2 gas concentration, (c) the partial pressure of H2 O2 -H2 O, and (d) the temperature distribution. (For interpretation of the references to color in the text, the reader is referred to the web version of this article.) Table 2 Results of the numerical model at a process time of 1.05 s (Fig. 4, point (A)). From left to right: inlet temperature of H2 O2 -containing gas mixture, vial-surface temperature in ◦ C, state indicator (liquid in blue, gas in red or a mix state in green), phase fraction (0: gas and 1: liquid or condensation), and the concentration of H2 O2 in the condensation (a value of 0 indicates the absence of a condensation).

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

i. Surface temperature: An increase in the inlet-gas temperature increases the surface temperature of the vial. The inner vial surface near to the inlet-gas nozzle (bottom part) has the highest recorded temperature due to a higher velocity magnitude of the gas mixture (s. Fig. 10(a)). This region represents the leastprobable location for a condensation to occur. Due to the short time period of 1.05 s and the low thermal conductivity of glass (about 1.4 W/(m·K)), the surface temperature reaches a maximum of 131 ◦ C, while the bulk temperature of the glass vial remains at approximately 20 ◦ C. ii. State indicator: The first region in blue represents a complete condensation of the gas mixture. The second region in red indicates the presence of only a gas phase (no condensation). The green region shows possible locations, where a part of the gas mixture condenses. Moreover, one can observe that with a higher inlet-gas temperature, a state indicator is plotted on the outer-vial surface. This occurs due to an increase in the outletgas velocity caused by an increase in bulk-gas temperature and subsequent decrease in the fluid density (s. Eq. (3)). Additionally, due to the mixing of the H2 O2 -containing gas with the atmospheric air, the H2 O2 gas concentration at the outer-vial surface is lower than at the inner surface as shown in Fig. 10(b). A gas mixture with a lower H2 O2 concentration has a lower partial pressure. Fig. 6(b) states that a lower surface temperature (below the dew-point temperature at stated partial pressure) is required to cause the formation of condensation. Since this is not the case, a gas-phase indicator (red color) is shown. iii. Phase fraction: The liquid fraction of the mix state during condensation is presented as a mass fraction relative to the total mass of the gas mixture. A complete condensation means a liquid-mass fraction of 1 and a value of 0 when no condensation forms. When the temperature of the surface is between the bubble- and dew-point temperature, a concentration gradient is observed. The liquid fraction near to the dew-point temperature of the H2 O2 -containing gas is the lowest and by a decrease of temperature (toward the bubble-point temperature), more condensation forms. iv. Aqueous H2 O2 concentration in the condensation: Analogous to the phase-fraction plot, at a gas state, no condensation forms and therefore the concentration of aqueous H2 O2 is 0. For a liquid phase, the mass fraction of aqueous H2 O2 is at a lowest value of 35% w/w near to the bubble-point temperature of the gas mixture. With an increase in vial-surface temperature, H2 O2 condenses before H2 O and a higher concentration of H2 O2 solution can be noted. The limitation of the 2D-axisymmetric model is defined by the modeled geometry since the results are constant along the circumference of the vial. In other words, what is presented here are theoretical results when the location of the nozzle is perfectly central, without considering the location of the microorganisms that are sporadically presented on a surface. Nevertheless, we deduce from the surface temperature of the vial shown in Table 2 that even if the gas parameters are constant inside a packaging material, a temperature gradient is formed. The lower part of the vial (closer to the gas nozzle) receives more thermal energy than the upper corners. This results in the formation of a non-uniform condensation film. Nonetheless, the numerical model and the presented results provide a starting point for the experiments to help in controlling the prevailing sterilization process. 4.4. Sterilization by gas-phase vs. (micro-)condensation We briefly discuss in this section the differences between gasphase sterilization and sterilization by (micro-)condensation:

9

i. In gas-phase sterilization, a higher kill rate per unit volume of H2 O2 solution can be achieved, since the H2 O2 concentration in the gas phase remains constant across the surface under treatment. In turn, this decreases the consumption of H2 O2 and the required thermal energy of the evaporation unit. Additionally, recent studies show that gas-phase sterilization is not affected by the type or the roughness of either packaging material or the present microorganisms [26]. This guarantees a homogeneous sterilization efficacy of the package surface. ii. The efficacy of a (micro-)condensation sterilization is dependent on the hydrophobic property of the surface of the packaging material and the microorganisms present on it [26]. For instance, liquid-surface tension can prevent the complete wetting of a surface that results in a non-uniform sterilization efficacy. This occurs in addition to the condensation nuclei that can be sharp points on packaging surfaces (due to surface roughness), the distribution of microorganisms on the surface or due to the presence of dust particles. iii. From an energy point-of-view, the formation of a condensate means that more thermal energy is required in the final posttreatment step to vaporize the condensate and to reduce the H2 O2 surface residual. Meanwhile, in gas-phase sterilization, even if a higher initial surface temperature of a package is required, lower energy consumption is expected in the posttreatment step. Consequently, the H2 O2 residual levels are lower and easier to handle with lower net-energy consumption. 5. Conclusions The sterilization of a package surface prior to product filling is a necessary step in pharmaceutical-, food- and beverage industries. A typical sterilization process follows a pre-treatment step with heated air, a gas-treatment step with a heated H2 O2 -containing gas and a post-treatment process with heated air. By altering the operational conditions of the first two steps of the sterilization process, a gas-phase sterilization or a sterilization by (micro-)condensation can be achieved. A sterilization system was developed that combines the three steps in a single system and was used to test a sterilization procedure for packages. Validation of the system was carried out using a H2 O2 gas sensor and commercial spore test strips of Geobacillus stearothermophilus on a carrier surface. The results showed that an increase in H2 O2 gas concentration increases the chance of sterility and that a H2 O2 gas concentration of at least 4.2% v/v is required to reach a LCR value of more than six in a 0.7 s sterilization process. A numerical model of the sterilization process was developed to predict the dominant mode of sterilization due to changes of process parameters such as the gas-inlet temperature. The model estimates the location where condensation occurs, the ratio of the condensed vapor and its H2 O2 condensation. The geometry of a pharmaceutical vial and a simplified three-step sterilization process were introduced during the development of the numerical model. Conflicts of interest The authors declare no conflicts of interest. Acknowledgments We thank A. Bromm from Imagine Engineering GmbH for the support in the construction and installation of the sterilization system. M. Langenberg from Imagine Engineering GmbH is acknowledged for setting-up the transport system. J. Arreola from FH Aachen is acknowledged for his support in the preparation of biological test strips. This work was financially supported by

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11 10

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

the European Union and by the State of North Rhine-Westphalia within the operational program EFRE.NRW 2014–2020, project “EfficientSterile”. References [1] W.A. Rutala, D.J. Weber, Healthcare Infection Control Practices Advisory Committee, Guideline for Disinfection and Sterilization in Healthcare Facilities, Tech. Rep., Centers for Disease Control and Prevention, USA, 2008. [2] G.E. McDonnell, The use of hydrogen peroxide for disinfection and sterilization applications, in: Patai’s Chemistry of Functional Groups, John Wiley & Sons, Ltd., Chichester, England, UK, 2009. [3] B.V. Bockelmann, I.A.V. Bockelmann, Long-Life Products: Heat-Treated, Aseptically Packed: A Guide to Quality, B. von Bockelmann, Âkarp, Sweden, 1998. [4] T. Sandle, Sterility, Sterilisation and Sterility Assurance for Pharmaceuticals: Technology, Validation and Current Regulations, Vol. 32 of Woodhead Publishing Series in Biomedicine, Woodhead Publishing Limited, Oxford, UK, 2013. [5] W.C. Schumb, Stabilization of concentrated solutions of hydrogen peroxide, Ind. Eng. Chem. 49 (10) (1957) 1759–1762. [6] C.N. Satterfield, T. Stein, Decomposition of hydrogen peroxide vapor on relatively inert surfaces, Ind. Eng. Chem. 49 (7) (1957) 1173–1180. [7] P.L. Garwig, Heterogeneous Decomposition of H2 O2 by Inorganic Catalysts: Technical Report AFRPL-TR-66-136, Tech. Rep., Air Force Rocket Propulsion Laboratory and Chemical Research and Development Center, 1966. [8] B.G. Petri, R.J. Watts, A.L. Teel, S.G. Huling, R.A. Brown, Fundamentals of ISCO using hydrogen peroxide, in: R.L. Siegrist, M. Crimi, T.J. Simpkin (Eds.), In Situ Chemical Oxidation for Groundwater Remediation, Vol. 3 of SERDP/ESTCP Environmental Remediation Technology, Springer, New York, NY, USA, 2011, pp. 33–88. [9] A.P. Fraise, J.-Y. Maillard, S.A. Sattar (Eds.), Principles and Practice of Disinfection, Preservation, and Sterilization, 5th ed., John Wiley & Sons, Ltd., Chichester, England, UK, 2012. [10] J. Oberländer, M. Mayer, A. Greeff, M. Keusgen, M.J. Schöning, Spore-based biosensor to monitor the microbicidal efficacy of gaseous hydrogen peroxide sterilization processes, Biosens. Bioelectron. 104 (2018) 87–94. [11] C. Weiler, H. Geissler, H.-W. Mainz, S. Boonkaew, Method and device for sterilizing containers, WO Patent 2016/096472 A1 (2015). [12] M. Reiter, Verfahren und Vorrichtung zum Entkeimung von Verpackungsmaterial, insbesondere von Verpackungsbehältern, DE Patent 3540161 A1 (1987). [13] H. Müller, G. Deimel, Method of an apparatus for sterilizing packaging material, especially container-type packages, US Patent 4742667 (1988). [14] K. Baltes, P. Kirchner, A. Meckbach, Z.B. Jildeh, A. Bromm, Verdampfer mit mäandrierendem Strömungskanalsystem, DE Patent 102017215200 A1 (2019). [15] Z.B. Jildeh, P. Kirchner, K. Baltes, P.H. Wagner, M.J. Schöning, Development of an in-line evaporation unit for the production of gas mixtures containing hydrogen peroxide – numerical modeling and experimental results, Int. J. Heat Mass Transf. 143 (2019), 118519 (1–15). [16] P. Kirchner, J. Oberländer, H.-P. Suso, G. Rysstad, M. Keusgen, M.J. Schöning, Monitoring the microbicidal effectiveness of gaseous hydrogen peroxide in sterilization processes by means of a calorimetric gas sensor, Food Control 31 (2) (2013) 530–538. [17] Food and Drug Adminstration, Guideline for Industry: Sterile Drug Products Produced by Aseptic Processing – Current Good Manufacturing Practice, Tech. Rep., U.S. Department of Health and Human Services, Office of Regulatory Affairs (ORA), Rockville, MD, USA, 2003. [18] J.P. Agalloco, J.E. Akers (Eds.), Advanced Aseptic Processing Technology, Drugs and the Pharmaceutical Sciences, Informa Healthcare, Boca Raton, FL, USA, 2010. [19] K. Pruß, S. Stirtzel, U. Kulozik, Influence of the surface temperature of packaging specimens on the inactivation of Bacillus spores by means of gaseous H2 O2 , J. Appl. Microbiol. 112 (3) (2012) 493–501. [20] G. Cerny, Testing of aseptic machines for efficiency of sterilization of packaging materials by means of hydrogen peroxide, Packag. Technol. Sci. 5 (2) (1992) 77–81. ˇ cek, K. Syslová, J. Václavík, D. Pavlík, J. Cerven ˇ ´ M. Kuzma, y, [21] P. Kaˇcer, J. Svrˇ Vapor phase hydrogen peroxide – method for decontamination of surfaces and working areas from organic pollutants, in: Dr.T. Puzyn (Ed.), Organic Pollutants Ten Years After the Stockholm Convention – Environmental and Analytical Update, InTech, 2012, pp. 399–430. [22] D. Watling, C. Ryle, M. Parks, M. Christopher, Theoretical analysis of the condensation of hydrogen peroxide gas and water vapour as used in surface decontamination, PDA J. Pharm. Sci. Technol. 56 (6) (2002) 291–299. [23] E. Klemm, G. Mathivanan, T. Schwarz, S. Schirrmeister, Evaporation of hydrogen peroxide with a microstructured falling film, Chem. Eng. Process. Process Intensif. 50 (10) (2011) 1010–1016. [24] G. Scatchard, G.M. Kavanagh, L.B. Ticknor, Vapor–liquid equilibrium: VIII. Hydrogen peroxide–water mixtures, J. Am. Chem. Soc. 74 (15) (1952) 3715–3720. [25] S.L. Manatt, M.R.R. Manatt, On the analyses of mixture vapor pressure data: the hydrogen peroxide/water system and its excess thermodynamic functions, Chemistry 10 (24) (2004) 6540–6557.

[26] E. Eschlbeck, C. Seeburger, U. Kulozik, Influence of spore and carrier material surface hydrophobicity on decontamination efficacy with condensing hydrogen peroxide vapor, J. Appl. Microbiol. 124 (5) (2018) 1071–1081. [27] W.J. Rogers, Sterilisation of Polymer Healthcare Products, Rapra Technology Limited, Shawbury, UK, 2005. [28] S. Spanu, G. Vignali, Modelling and multi-objective optimisation of the VHP pouch packaging sterilisation process, Int. J. Food Eng. 12 (8) (2016) 739–752. [29] J. Fisher, R.A. Caputo, Comparing and contrasting barrier isolator decontamination systems, Pharm. Technol. 28 (11) (2004) 68–82. [30] F. Vahidpour, J. Oberländer, M.J. Schöning, Flexible calorimetric gas sensors for detection of a broad concentration range of gaseous hydrogen peroxide: a step forward to online monitoring of food-package sterilization processes, Phys. Status Solidi A 215 (15) (2018), 1800044 (1–7). [31] N.A. Klapes, D. Vesley, Vapor-phase hydrogen peroxide as a surface decontaminant and sterilant, Appl. Environ. Microbiol. 56 (2) (1990) 503–506. [32] J.V. Rogers, C.L.K. Sabourin, Y.W. Choi, W.R. Richter, D.C. Rudnicki, K.B. Riggs, M.L. Taylor, J. Chang, Decontamination assessment of Bacillus anthracis, Bacillus subtilis, and Geobacillus stearothermophilus spores on indoor surfaces using a hydrogen peroxide gas generator, J. Appl. Microbiol. 99 (4) (2005) 739–748. [33] B.K.L. Unger-Bimczok, T. Kosian, V. Kottke, C. Hertel, J. Rauschnabel, Hydrogen peroxide vapor penetration into small cavities during low-temperature decontamination cycles, J. Pharm. Innov. 6 (1) (2011) 32–46. [34] Mechanical Engineering Industry Association e.V, Code of Practice: Filling Machines of VDMA Hygiene Class V: Testing the Effectiveness of Packaging Sterilization Devices, No. 6/2002, 2008. [35] Mechanical Engineering Industry Association e.V, Code of Practice: Testing Filling Machines of VDMA Hygiene Class V (aseptic plants): Sterilizing the Sterile Zone in a Machine Interior, No. 8/2003, 2014. [36] G. Moruzzi, W.E. Garthright, J.D. Floros, Aseptic packaging machine pre-sterilisation and package sterilisation: statistical aspects of microbiological validation, Food Control 11 (2000) 57–66. [37] R. Blodgett, Bam Appendix 2: Most Probable Number From Serial Dilutions, 2017 https://www.fda.gov/food/laboratory-methods-food/bam-appendix-2most-probable-number-serial-dilutions. [38] International Organization for Standardization, EN ISO 11139, Sterilization of Health Care Products – Vocabulary of Terms Used in Sterilization and Related Equipment and Process Standards, 2018 https://www.iso.org/obp/ui/ #iso:std:iso:11139:ed-1:v1:en. [39] A.G. Mosely, Sterility assurance level (sal): the term and its definition continues to cause confusion in the industry, Pharm. Microbiol. Forum Newsl. 14 (5) (2002) 2–15. [40] J. Oberländer, P. Kirchner, M. Keusgen, M.J. Schöning, Strategies in developing thin-film sensors for monitoring aseptic food processes: theoretical considerations and investigations of passivation materials, Electrochim. Acta 183 (2015) 130–136. [41] Z.B. Jildeh, J. Oberländer, P. Kirchner, P.H. Wagner, M.J. Schöning, Thermocatalytic behavior of manganese (IV) oxide as nanoporous material on the dissociation of a gas mixture containing hydrogen peroxide, Nanomaterials 8 (4) (2018), 262 (1–16). [42] A. Datta, V. Rakesh, An Introduction to Modeling of Transport Processes: Applications to Biomedical Systems, Cambridge Texts in Biomedical Engineering, Cambridge University Press, Cambridge, UK, 2010. [43] Z.B. Jildeh, P. Kirchner, J. Oberländer, A. Kremers, T. Wagner, P.H. Wagner, M.J. Schöning, FEM-based modeling of a calorimetric gas sensor for hydrogen peroxide monitoring, Phys. Status Solidi A 214 (9) (2017) 1600912 (1–9). [44] M.J. Schöning, P. Kirchner, J. Oberländer, Method for Establishing Gas Sensor in Gas Phase for Detecting Concentration of Specific Gas, Involves Evaluating Measured Value of Concentration of Gas With Aid of Calibration Performed in Gas Sensor, and Transmitting Evaluated Result, DE 102012221436 A1, 2014. [45] E. Linley, S.P. Denyer, G. McDonnell, C. Simons, J.-Y. Maillard, Use of hydrogen peroxide as a biocide: new consideration of its mechanisms of biocidal action, J. Antimicrob. Chemother. 67 (7) (2012) 1589–1596.

Biographies Zaid B. Jildeh received his M.Sc. degree in Energy Engineering at the Aachen University of Applied Sciences in 2014. Later he worked as a research assistant at IMAGINE Engineering GmbH (2014–2019). Afterwards, he joined Koch Pac-Systeme GmbH as a test and process engineer. Jildeh is also involved in a cooperative doctoral program between the Institute for Nano- and Biotechnologies of Aachen University of Applied Sciences and Catholic University of Leuven. His main research activity involves the devlopment of an in-line hydrogen peroxide gas sterilization unit for aseptic filling machines. Patrick Kirchner studied physical engineering at Aachen University of Applied Sciences (2004–2008). Within his following research project at the Institute of Nanoand Biotechnologies of the same University, Mr. Kirchner developed a novel sensor system for monitoring sterilization processes (2008–2012) and earned his doctoral degree in pharmacy at Philipps University of Marburg (2013). At the end of his PhD, he worked for Elopak focusing on R&D topics in the field of aseptic filling technologies (2012–2013). In 2013, Mr. Kirchner joined Imagine Engineering, a technology service provider in hygienic engineering and process technology. At present, he is

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691

G Model SNA-111691; No. of Pages 11

ARTICLE IN PRESS Z.B. Jildeh, P. Kirchner, J. Oberländer et al. / Sensors and Actuators A xxx (xxxx) xxx

head of technology and simulation. His field of activities primarily includes projects concerning aseptic processing and sealing technologies. Jan Oberländer received his B.Eng. (2010) and M.Sc. (2012) in Biomedical Engineering from Aachen University of Applied Sciences. In 2018 he received the PhD degree from Philipps-University Marburg. His current research interest is focused on sensor-based solutions to monitor gaseous sterilization processes. Farnoosh Vahidpour finished her B.Sc. in 2005 and M.Sc. of atomic and molecular physics in 2009 in Tehran, Iran. In 2009, she moved to Belgium and followed courses at the Catholic University of Leuven. She started in 2011 her doctoral research in material sciences at Hasselt University with the focus on microfabrication of diamond microelectrode arrays and flexible electrodes for biomedical applications. Since 2017, she holds a postdoctoral position at the Aachen university of Applied Sciences. Her main topics involve microfabrication techniques, flexible materials, biosensors and various characterization methods for material surface studies. Patrick H. Wagner obtained a Ph.D. in physics in 1994 at Technical University Darmstadt and did postdoctoral research on the magneto-transport properties of

11

perovskites at KU Leuven between 1995 and 2001. In 2001 he became a professor for experimental biophysics at Hasselt University and in 2014 he returned to KU Leuven as a full professor for the physics of biofunctional surfaces. P. Wagner received several grants and distinctions, was president of the Belgian Physical Society (2006–2007), and serves currently as an editor-in-chief of the Elsevier journal Physics in Medicine. Michael J. Schöning received his diploma degree in electrical engineering (1989) and his PhD in the field of semiconductor-based microsensors for the detection of ions in liquids (1993), both from the Karlsruhe University of Technology. In 1989, he joined the Institute of Radiochemistry at the Research Centre Karlsruhe. Since 1993, he has been with the Institute of Thin Films and Interfaces (now, Peter Grünberg Institute, PGI-8) at the Research Centre Jülich, and since 1999 he was appointed as full professor at Aachen University of Applied Sciences, Campus Jülich. Since 2006, he serves as a director of the Institute of Nano- and Biotechnologies (INB) at the Aachen University of Applied Sciences. His main research subjects concern siliconbased chemical and biological sensors, thin-film technologies, solid-state physics, microsystem and nano(bio-)technology.

Please cite this article as: Z.B. Jildeh, P. Kirchner, J. Oberländer et al., Development of a package-sterilization process for aseptic filling machines: A numerical approach and validation for surface treatment with hydrogen peroxide, Sens. Actuators A: Phys., https://doi.org/10.1016/j.sna.2019.111691