Preventing Contamination During Sterile Powder Transfer
A chlorine dioxide sterilization cycle was developed for a novel split-valve aseptic powder transfer device.
Several techniques can be used to help ensure product sterility assurance during the transfer of pharmaceutical powders from one process stage to the next using an intermediate container (e.g., vessel or bag). One of these is split butterfly valve (SBV) technology.
The ChargePoint AseptiSafe® bio SBV traditionally uses hydrogen peroxide (H202) gas to bio-decontaminate and remove contamination on the two mating surfaces of the valve before the product is transferred. As shown in Figure 1, the two halves of the valve create a sealed chamber by partially docking the two disc faces, allowing a decontamination to take place in a closed environment. This prevents further contamination from taking place after the decontamination process is finished, unlike the traditional sporicidal spray and wipe approach, which is generally carried out in an unsealed environment therefore allowing contamination to occur.
The decontamination cycle time remains an important factor as ultimately only optimal performance will be considered in-line with the commercial and technical objectives of the production facility. Current H202 decontamination cycle times have been established at 4 minutes to 20 minutes. Chlorine dioxide (CD) gas has long been used as an alternative to H202 vapor for decontamination. Previous research, such as that of M. Sawyer et al. in 2014, lists the differences between the two types of bio-decontamination media in terms of a typical cycle, however SBV technology with CD gas had not been evaluated. For any technology to be adopted within a validated production process, qualification evidence should exist to prove the technology prior to live process validation. This article describes the qualification of chlorine dioxide gas to decontaminate the ChargePoint AseptiSafe® bio SBV. The qualification testing was carried out in partnership with ChlorDisys Solutions, who provide highly effective decontamination services utilising CD gas.
To provide optimal sterility assurance, the SBV must prove to bring about a 6 log (10-6 ) reduction in microbiological contamination on the surface areas of the device exposed to the decontamination cycle. Typically, this reduction is validated with the use of a biological indicator (BI) and subsequent controlled laboratory analysis of bacteria culture growth.
A cycle development and optimization study was performed to test the feasibility of CD gas to sterilize the mating surfaces to the required sterility assurance level and to perform such cycles in as short a time as possible. The study determined the compatibility of materials in contact with the CD gas, and also examined how pressure may or may not cause leaks in the system.
Cycle development and optimization methodology
All the testing was done with CD gas generated by ClorDiSys’ Minidox-M CD Gas Generator. CD gas was injected manually from the Minidox-M (see Figures 2 and 3) and allowed to flow for (n) seconds through the valve assembly. The outlet valve was then closed, after which the Minidox-M valve was closed. When the valves were closed, a high concentration of CD gas (100 mg/L) was trapped in the valve assembly.
Material and equipment. The following materials and equipment were used for the experiment:
Minidox-M CD gas generator
ChargePoint type: 100-mm BIO ACTIVE, Equipment No.: DE0024-02-2
CSI BSC Scrubber
PVC ¼” ball valve
3/8” Kynar tubing
Biological indicators (BI) G. stearothermophilus spores, Crosstex Code: TCDS-06, population ≥106 (1,000,000) , lot#: RU 53
Prepared culture color-change media (NAMSA Code: GMBCP-100, lot#: GM005213)
Incubator for incubating G. stearothermophilus spore strips
ATI C16 PortaSense II low-level chlorine dioxide sensor
Parameters and operation.
ClorDiSys approaches every gaseous decontamination effort as if it were to obtain a 6-log sporicidal reduction. Reducing spore counts is more difficult than reducing viruses and other organisms. Sporicidal reduction also requires humidity to soften the spore walls to enable the gas to penetrate and kill the microbes.
The CD decontamination process consists of the following steps:
- humidification to soften the spore walls
- introduction of CD gas to reach the desired concentration
- a dwell period (i.e., exposure to the gas) in which the gas sits for a period of time to obtain the desired kill level
- aeration to remove the gas.
The humidification range required to soften the spore walls is at least 60%, with an optimal level of 65–70%. Lower relative humidity levels require longer exposure times or more dosage. The exposure level and time that ClorDiSys targets to obtain a 6-log sporicidal reduction is 1 mg/L (360 ppm) at 2 hours. This equates to 720 ppm-hours of dosage. ClorDiSys has demonstrated 6-log sporicidal reduction as low as 450 ppm-hours. At the end of exposure, the gas is aerated until the concentration drops to 0.1 ppm, which is the 8-hour safety level as well as the odor threshold level.
In this application, the concentration levels of CD used were reduced compared to a typical cycle, which are generally carried out in large rooms, due to the small size of the chamber once the valves were closed.
A first set of experiments were performed with a fixed time gas flow. The system was connected as shown in Figure 3 with the 1/4” ball valve open to the scrubber and one BI placed in the valve assembly. The Minidox-M outlet valve was manually opened and gas flowed to the scrubber for 20 seconds at 20 LPM. The ¼” ball valve was then closed, followed by the Minidox-M outlet valve. The BI was then tested with multiple exposure times of 1, 2, 5, and 10 minutes. Each exposure time cycle was repeated three times for three cycles, for a total of 12 cycles with 12 BIs.
At the end of each exposure (1, 2, 5, and 10 minutes) the Minidox-M sample pump was energized and the ¼” ball valve was opened to blow the CD gas to the scrubber. This aeration time required to bring the CD gas to below detectable levels was 2 minutes.
A second set of experiments was performed with a fixed time gas flow using the same system connection and experiment times, but the gas flowed to the scrubber for 60 seconds at 20 LPM.
A third set of experiments was performed with an exposure time gas flow. The system was connected in the same way (see Figure 3) with the ¼” ball valve open to the scrubber and one BI placed in the valve assembly. The Minidox-M outlet valve was manually opened and gas flowed to the scrubber for 20 seconds at 20 LPM. The ¼” ball valve was not closed, however, and gas was allowed to flow for the entire exposure time. Exposure time started after 20 seconds of flow.
The BI was then tested with the same exposure times and number of cycles as in the first experiment. At the end of exposure (1, 2, 5, and 10 minutes) the Minidox-M sample pump was energized and the ¼” ball valve was opened to blow the CD gas to the scrubber. In this case, the aeration time required to bring the CD gas to below detectable levels was also 2 minutes.
A fourth set of experiments was performed with a fixed time gas flow and pressure build up in the valve assembly. For these cycles, the system was connected as shown in Figure 3 with the ¼” ball valve open to the scrubber and one BI placed in the valve assembly. The Minidox-M outlet valve was manually opened and gas flowed to the scrubber for 20 seconds at 20 LPM. The ¼” ball valve was then closed. The Minidox-M outlet valve was closed after pressure increased to 10 psi (68.9 KPa). The BI was then tested with the same exposure times and number of cycles as in the first experiment. At the end of exposure (1, 2, 5, and 10 minutes) the Minidox-M sample pump was energized and the ¼” ball valve was opened to blow the CD gas to the scrubber. Again, the aeration time required to bring the CD gas to below detectable levels was 2 minutes.
Cycle times for the four sets of cycles varied due to the exposure time and gas flow time. The shortest cycle was the 3.3 minutes and the longest was 13 minutes. The shortest successful cycle (defined as three runs with no positive BIs) was 7.3 minutes. Table II shows the cycle data BI results. In each run, only one BI was placed in the valve assembly. Room relative humidity during the cycles was 30–40%.
- * Total cycle time includes charge, exposure, and aeration to safe levels.
- + indicates a positive BI (incomplete kill or growth)
- - indicates a negative BI (complete kill or no growth)
- Green highlights indicate successful cycle times and approximate CD dosages.
In the pressure-hold test (the fourth set of tests), when the pressure within the assembly was raised, a low level chlorine dioxide gas sensor was used to measure any leakage on both sides of the valve assembly for the entire exposure time. No leaks were detected on either side.
No material effects (e.g., degradation/cracking) were noticed after all testing was performed. The valve underwent a minimum of 48 cycles at 100 mg/L CD gas concentration. The total dosage for all cycles was approximate 130,000 ppm-hrs. This total dosage is equivalent to 180 runs at a normal room concentration of 1mg/L.
The purpose of the test was to ensure that a 6 log (10-6) reduction in microbiological contamination was achieved on the exposed faces of the valve. This result was achieved on all 5- and 10-minute cycles. No biological growth was seen in any of the BIs in these cycles. The cycle times presented consistent bio decontamination at five minutes, which is faster and more consistent than was found in previous tests of H202 cycles.
During the leak rate analysis, in which the gassing space was pressurized and low level readings were taken above and below the discs, no levels of gas detection were found. Considering that CD is a pure gas and therefore inherently easier to detect than H202 vapor, the results proved the process to be a robust solution that can be safely used with both CD and H202 vapor.
In the future, plans call for testing the system with longer pipe runs to ensure that the fast aeration times can be maintained. Longer pipes would allow drug manufacturers to house the generator in one location within the facility and the valve in the process area without the concern of prolonged cycle times.
1. M. Sawyer et al., “Got Gas? Chlorine Dioxide or Vaporized Hydrogen Peroxide: Which one is right for you?”, Poster Presentation, Biosecurity Research Institute, Kansas State University, 2014, www.absaconference.org/pdf55/003Sawyer.pdf