6. Procedure
6.1 Manual Ampule Filling and Sealing:
6.1.1 Apparatus - Devices used for manual filling of ampules include glass pipettes as well as other types of commercially available hand-operated, mechanical, liquid-dispensing devices.
6.1.2 Storage of Bulk Material - Bulk gasoline and similar liquid hydrocarbon materials must be adequately sealed and stored to prevent loss of volatile components prior to ampulization. Refrigerated storage in sealed metal drums, barrels, or amber glass containers is recommended.
6.1.3 Compatibility of Materials/Sources of Contamination:
6.1.3.1 Materials that come in contact with the bulk RM and its vapors during dispensing must be compatible with the gasoline or hydrocarbon material. Glass pipettes are recommended. Plastic or rubber materials containing phthalates or other types of plasticizers must be avoided.
6.1.3.2 Any part of the dispensing device that comes in contact with the material, including glass pipettes, hand dispensers, and any necessary connection hardware, must be cleaned prior to packaging a different material. Recommended cleaning procedures involve soaking parts in soapy water, rinsing with clean water, followed by methanol or other suitable solvent, followed by drying under a stream of clean nitrogen.
6.1.4 Assessment of Material to Be Ampulized:
6.1.4.1 Volatility - Prior to packaging, materials containing highly volatile components must be cooled sufficiently to minimize volatile losses during ampulization. Failure to sufficiently cool the material also may result in difficulty in obtaining effective ampule sealing. The material must not be cooled to temperatures below which the composition of the RM would be affected (for example, producing precipitation or solidification). Gasoline may be cooled to -20°C without incurring compositional changes. The bulk material must be kept cold during the filling process.
6.1.4.2 Reactivity - Consideration should be given to the chemical reactivity of the RM being packaged. Gasoline samples containing olefins and diolefins should be packaged under an inert atmosphere blanket of nitrogen, argon, or other suitable gas. Ampules should be flushed with inert gas immediately prior to dispensing of the gasoline. Use of amber glass ampules will minimize photo-oxidation.
6.1.4.3 Odors - Odorous materials such as gasoline should be packaged in a well-ventilated area. The bulk material should be kept adequately sealed during the ampulization process to minimize loss of volatiles.
6.1.4.4 Flammability - Ampule sealing requires use of a flame hot enough to melt glass. Care must be taken in ampulization of highly flammable materials since ampule contents could ignite. Ampules must be kept cold through the sealing step. However, care should be taken to avoid, as much as possible, condensation of water inside the ampule. Ampulization is best carried out when the room humidity is low.
6.1.5 Sampling of the Bulk Container:
6.1.5.1 After bringing the bulk container temperature down to the working temperature, withdraw a minimum of three samples from each bulk container, using clean manual pipettes. Immediately dispense the material into crimp top chromatography vials, seal, and label. These samples will be designated as representative of the bulk material and will be used to establish reference values for the homogeneity testing.
6.1.5.2 Some vial closures are not suitable for hydrocarbon analyses, such as uncoated silicone rubber. Only TFE-fluorocarbon-coated closures should be used. In addition, the vials should be analyzed as soon as is practical, since no crimped vial is completely leak free.
6.1.6 Adjusting Dispensing Volume:
6.1.6.1 Typically, it is more important to provide a minimum dispensed volume in the ampule rather than to provide an accurately determined volume of RM. The minimum dispensing volume for packaging the RM must be known ahead of time.
6.1.6.2 Introduce an inert atmosphere into the ampule by purging the ampule for a few seconds with nitrogen or other inert gas immediately prior to filling. A disposable glass dropper connected to a gas source using rubber tubing provides a convenient way of purging the ampule.
6.1.6.3 If using graduated pipettes, introduce a sufficient volume of material to the ampule to meet the minimum dispensing volume requirements for packaging the RM. Note that the final dispensed volume at room temperature will be affected by the bulk material temperature at the time of dispensing. Therefore, for consistent volume dispensing, the temperature of the bulk material must be known and must be kept constant during the entire dispensing process.
6.1.6.4 If other types of nongraduated, manual, filling devices are being used, they must be calibrated. Using Class A glassware or pipettes, measure into an ampule a volume of room temperature water equal to the volume of RM to be dispensed. Mark the level on the ampule.
6.1.6.5 Make adjustments to the manual dispensing device until 50 consecutive ampules are consistently filled to the predetermined mark on the ampule.
6.1.6.6 Once volume dispensing adjustments have been completed, begin filling ampules from the bulk supply, keeping the filled ampules cold by placing them immediately into a container that is at a temperature of approximately -20°C. This may be achieved by using crushed dry ice.
6.1.6.7 The ampules should be sealed as soon as possible after filling to avoid loss of volatile components. If ampules are being manually sealed, a two person operation, in which one person dispenses the material and a second person seals the ampules, is suggested.
6.1.6.8 Periodically inspect filled and sealed ampules to ensure that the fill volume is maintained throughout the packaging run.
6.1.7 Ampule Sealing:
6.1.7.1 Ampules may be flame-sealed by hand, using a suitable torch. The flame used must be hot enough to quickly soften the neck of the ampule. Propane/air or natural gas/air flames are sufficient for most applications. Hydrogen/oxygen flames may be required for sealing large, thick-walled glass ampules.
6.1.7.2 The ampule should be kept cold through the sealing process.
6.1.7.3 To facilitate sealing, the torch should be mounted on a stand on a bench top such that both hands can be free to perform the actual sealing process.
6.1.7.4 Wearing gloves, hold the ampule by the bottom in one hand and by the neck tip in the other. Alternatively, large tongs may be used to hold the neck in order to minimize the risk. The ampule neck is placed into the flame, constantly rotating to ensure uniform heating of the glass. Focus the flame midway between the open end and the breakmark on the ampule. The flame should never contact the contents of the ampule or the direct open end of the ampule. (Warning - For safety reasons, if the material being ampulized is flammable, it is recommended that the volume being ampulized be minimized.)
6.1.7.5 After several seconds of exposure to the flame, the glass should begin to soften. Once the glass is softened, the neck may be pulled away from the ampule while still in the flame, until the glass draws down and forms a seal. The top of the ampule should be polished by rotating it in the flame until a smooth seal is obtained.
6.1.7.6 The seal should be inspected and should be smooth and free of any carbon deposits. The thickness of the seal should be comparable to the thickness of the ampule wall. If the RM contains volatile materials, sealing may be difficult if the material is not sufficiently cooled. Some problems encountered include thin seals or bubbles, formed due to pressure from volatile components, and carbon formation at the seal. These problems usually can be eliminated by further cooling of the ampules. For gasoline, a temperature of -20°C usually is sufficient to avoid such difficulties.
6.2 Automated Ampule Filling and Sealing:
6.2.1 Apparatus - Various commercial devices are available for automated filling and sealing of ampules. These devices typically consist of a pump, liquid transfer lines, dispensing or dosing needles, sealing torch, and an ampule conveying mechanism.
6.2.2 Storage of Bulk Material - Bulk gasoline and similar liquid hydrocarbon materials must be adequately sealed and stored to prevent loss of volatile components prior to ampulization. Refrigerated storage in sealed metal drums, barrels, or amber glass containers is recommended.
6.2.3 Compatibility of Materials/Sources of Contamination:
6.2.3.1 Tubing used to transfer the bulk material to automated dispensing devices must be compatible with the gasoline or hydrocarbon material. Plastic or rubber tubing containing phthalates or other types of plasticizers must be avoided. TFE-fluorocarbon tubing is highly recommended. To avoid cross-contamination, the transfer tubing must be replaced after every RM.
6.2.3.2 Any part of the dispensing apparatus that comes in contact with the RM, including pumps, dosing needles, and connection hardware, must be cleaned prior to packaging a different material. Recommended cleaning procedures involve soaking removable parts in soapy water, followed by rinsing with water, followed by methanol or other suitable solvent, followed by drying under a stream of clean nitrogen.
6.2.4 Assessment of Material to Be Ampulized:
6.2.4.1 Volatility - Prior to packaging, materials containing highly volatile components must be cooled sufficiently to minimize volatile losses during ampulization. Failure to sufficiently cool the material also may result in difficulty in obtaining effective ampule sealing. The material must not be cooled to temperatures below which the composition of the RM would be affected (for example, producing precipitation or solidification). Gasoline may be cooled to -20°C without incurring compositional changes. The bulk material must be kept cold during the filling process. Provisions should be made to maintain constant temperature throughout the transfer tubing lines. As the bulk container is emptied and the headspace is correspondingly increased, differential vaporization can change the bulk concentration, enriching the high boilers (less volatile components) and depleting the low boilers (more volatile components). Care must therefore be taken to complete the ampulization process as expeditiously as is reasonably possible.
6.2.4.2 Reactivity - Consideration should be given to the chemical reactivity of the RM being packaged. Gasoline samples containing olefins and diolefins should be packaged under an inert atmosphere blanket of nitrogen, argon, or other suitable gas. Ampules should be flushed with inert gas immediately prior to dispensing of the gasoline. Use of amber glass ampules will minimize photo-oxidation.
6.2.4.3 Odors - Odorous materials such as gasoline should be packaged in a well-ventilated area. The bulk material should be kept adequately sealed during the ampulization process to minimize loss of volatiles.
6.2.4.4 Flammability - Ampule sealing requires use of a flame hot enough to melt glass. Care must be taken in ampulization of highly flammable materials or ampule contents could ignite. Dosing needle dripping of flammable materials may result in ignition.
6.2.5 Sampling of the Bulk Container:
6.2.5.1 After bringing the bulk container temperature down to the working temperature, withdraw a minimum of three samples from each bulk container, using clean manual pipettes. Immediately dispense the material into crimp top chromatography vials, seal, and label. These samples will be designated as representative of the bulk material and will be used to establish reference values for the homogeneity testing.
6.2.5.2 Place the end of the pump transfer line tubing into the bulk container and secure it. A vent line should be connected to the bulk container to avoid transfer problems due to vacuum buildup during the pumping process and to minimize odors.
6.2.6 Adjusting Filling/Sealing Parameters:
6.2.6.1 Automated filling and sealing machines typically have the capability of automatically introducing a purge gas into the ampule prior to filling. Make sure that the appropriate inert gas supply is in place and that a sufficient volume of gas will completely flush the ampule container prior to filling (at least 1.5 ampule volumes).
6.2.6.2 Adjust the dosing needle (fill heads) dispensing height so that filling occurs from the bottom of the ampule up.
6.2.6.3 Set the fill/seal rate at a sufficient speed such that the temperature of the dispensed RM does not increase by more than 5°C over the bulk temperature during the filling process.
6.2.6.4 Adjust the liquid dispensing volume in accordance with the filling device instruction manual. Typically, it is more important to provide a minimum dispensed volume into the ampule rather than to provide an accurately determined volume of RM. The minimum dispensing volume for packaging the RM must be known ahead of time.
6.2.6.5 Dispensed volume will be affected by the temperature of the bulk material. Therefore, for consistent volume dispensing, the temperature of the bulk material must be known and must be kept constant during the entire dispensing process. The temperature must be specified for volume measurements.
6.2.6.6 Using Class A glassware or pipettes, verify the consistency of the dispensed volume by measuring into an ampule a volume of room temperature water equal to the volume of RM to be dispensed. Mark the level on the ampule.
6.2.6.7 Make adjustments to the dispensing device until ten consecutive ampules are consistently filled to the predetermined mark on the ampule.
6.2.6.8 The flame used for sealing must be hot enough to quickly soften the neck of the ampule. Propane/air or natural gas/air flames are sufficient for most applications. Hydrogen/oxygen flames may be required for sealing large, thick-walled, glass ampules.
6.2.6.9 The seal should be inspected and should be smooth and free of any carbon deposits. The thickness of the seal should be comparable to the thickness of the ampule wall. If the RM contains volatile materials, sealing may be difficult if the material is not sufficiently cooled. Some problems encountered include thin seals or bubbles, formed due to pressure from volatile components, and carbon formation at the seal. These problems usually can be eliminated by further cooling of the ampules. For gasoline, a temperature of -20°C usually is sufficient to avoid such difficulties.
6.2.6.10 Ampules should only be filled to a maximum of 80 % capacity since the dispensing volume at -20°C will expand once the material warms to room temperature.
6.2.6.11 Once volume dispensing and sealing adjustments have been completed, begin filling ampules from the bulk supply.
6.3 Ampule Labeling:
6.3.1 Ampules should be sequentially labeled in order to assess packaging homogeneity. If sequential labeling is not possible, the lot must be physically fractionated. This involves sequestering fractions of the batch (for example, thirds, quarters, and so forth) immediately after they are filled and sealed. Should packaging homogeneity testing show unacceptable variability across the entire lot, it may be possible to salvage specific fractions of the lot.
6.3.2 Labels may be applied to the ampules using automated labeling equipment or manually. Information that should be placed on the label is listed in 7.2.
6.3.3 If ampules cannot be sequentially labeled, the ampules collected from each fraction must be kept sequestered until the packaging homogeneity has been determined to be acceptable (if applicable).
6.4 Packaging Homogeneity (Optional):
6.4.1 Any ampulization process can introduce variation into the finished product as a result of a number of variables mentioned earlier. Determination of packaging homogeneity provides an added level of confidence in the consistency of the ampulization process. ISO Guides 30, 31, and 35 offer guidance on criteria and procedures for determining homogeneity. If packaging homogeneity is chosen, the following steps can be followed.
6.4.2 Homogeneity Sampling:
6.4.2.1 The samples should be taken at regions where physical differences are expected to occur. Random sampling should only be adopted when cases of physical differences are unknown or believed to be absent (see ISO/REMCO N280).
6.4.2.2 After ampulization is complete, at least nine samples (or 7 % of the lot, whichever is the greater) are to be selected from the lot: three from among the first ampules filled, three from the middle of the process, and three from among the last ampules filled. These ampules are to be used for determining packaging homogeneity for the lot.
6.4.2.3 Preliminary Test for Homogeneity (Optional) (see ISO/REMCO N280) The preliminary assessment of the homogeneity of a prospective RM can be performed after homogenization as an integral part of the preparation process. The number of samples taken and the corresponding replicate determinations should be such that the appropriate statistical test would be capable of detecting the possible existence of inhomogeneity at a predetermined level.
NOTE 1 - Practice E826 gives one detailed procedure for determining homogeneity of bulk material. This practice is specialized to the case of testing homogeneity of metals, in either solid or powdered form. For most RM certification programs, an appropriate preliminary test for homogeneity can be obtained by applying a straightforward adaptation of Practice E826.
6.4.3 Homogeneity Testing:
6.4.3.1 Homogeneity across the lot is established through replicate analysis of the homogeneity samples. The measurement system used must be calibrated with sufficient standards and blanks in the target range to meet acceptability criteria.
6.4.3.2 A sample of the RM obtained from the bulk material is used as the reference for determining homogeneity. This bulk sample must be analyzed for the analyte or analytes of interest, using a calibration protocol that ensures that the specified accuracy and precision specifications are achieved.
6.4.3.3 Homogeneity samples are analyzed using the appropriate analytical protocol and the data compiled. ISO Guide 35 offers statistical procedures for evaluating packaging homogeneity. The grand mean for an analyte, group of analytes, or value is calculated from the mean for all measurements. The grand mean must lie within an acceptable percentage of the TV or consensus value. The means of the measurements determinations for each set of three or more samples must lie within an acceptable percentage of the grand mean.
6.4.3.4 The pooled standard deviation (sp) of the early, middle, and late samples is calculated as:
where:
nE, nM, andnL = the respective number of ampules analyzed from the early, middle, and late portion of the packaging run, and
sE, sM, andsL = the respective standard deviations.
The pooled standard deviation must be within acceptable limits (x %) of the grand mean.
6.4.3.5 In addition, the grand mean for each analyte, group of analytes, or value must fall within acceptable limits (x %) of the TV or consensus value.
6.4.3.6 An example of homogeneity testing may be found in Appendix X1.
6.5 Stability Testing:
6.5.1 Ampules from each packaging lot must be retained for determining stability and shelf life of the RM. Ampules for stability testing must be stored under the same conditions as the ampulized product. A randomly selected sample should be analyzed on the six month, the one year and the two year anniversaries of the storage commencement. The mean measurement for the sample must fall within acceptable limits (x %) of the TV or consensus value of the RM.
6.5.2 Alternative aggravated stability testing (see Practice D6362) can be used to predict instability of an analyte, group of analytes, or value, but should not to be used in place of a routine stability testing program. The test is based on an assumption that the rate of chemical decomposition doubles for every 10°C increase in temperature. Aggravated stability testing is accomplished by storing the RM at a temperature higher than the normal storage temperature in accordance with the following equation:
where:
t = stability time at normal storage temperature, days,
t'= stability time at aggravated (higher) temperature, days, and
ΔT = difference in temperature (°C) between normal and aggravated storage temperatures.
Example: An 11.25 day aggravated stability test at 64°C is theoretically equivalent to a two year (720 day) stability time at 4°C
.6.6 Multilaboratory Studies to Establish Acceptance Criteria - In most cases, RMs will have been characterized through formal evaluation studies involving interlaboratory cross-check programs or round-robin studies. Acceptance criteria for the method analytes will be determined using these data and should be updated with data from additional studies as such studies become available.