Catching Leaks on the Fly

The use of blister cards, a popular drug-packaging format in Europe for many years, is growing in the United States. This increased popularity results partly from the rising number of drugs that are moisture- or oxygen-sensitive. It also comes from the fact that blister cards can be printed with dosage times, thereby serving as compliance packages that help patients adhere to dosage regimens.

Hallie Forcinio

To ensure package integrity, standard operating procedures typically include a testing protocol. Traditionally, leak testing has relied on blue-dye, water-bath, drop, or burst tests. Unfortunately, these tests destroy the package and the product inside, which then must be disposed of in an audited process. For an expensive drug, the cost of product lost to testing quickly mounts to hundreds of dollars per hour. Water-bath and dye tests are relatively time-consuming and exacerbate this cost problem.

Many bad packages often are produced before a problem is identified. As a result, faster, nondestructive testing is replacing destructive methods in many situations. Nondestructive leak testing relies on various technologies, including high-voltage, pressure, vacuum, and imaging.

High-voltage testing

High-voltage electrical testing ensures incoming blister-packaging materials (i.e., formable bottom webs and lidstock material) are free of pinholes. In this test, the company that produces the blister material or lidstock mounts a tester near the end of its production line so the material passes between high-voltage and grounded electrodes before it is wound into rolls. Blister material or lidstock should insulate electrodes from each other. If a pinhole is present, an arc will discharge through the flaw to the ground. With sufficient voltage, the system can detect pinholes smaller than 1 μm ("PDC20" pinhole tester, Clinton Instrument Co., Villa Rica, GA, www.clintoninstrument.com). This level of resolution is more than adequate, because the threshold for detectable pinholes typically is greater than 50 μm.

Although this test ensures the quality of incoming blister-package materials, it still may be necessary to check finished blister cards to ensure that no damage occurred during the form–fill–seal process and to confirm seal integrity.

The "VeriPac 225/BLV" unit from Packaging Technologies and Inspection holds a blister card under vacuum.

Pressure and vacuum testing

Quality-control systems for finished blister packs rely on pressure or vacuum conditions. Most units are operated off-line and allow operators or quality-control personnel to check a periodic sampling of a run such as 8 or 10 packs every 30 min.

One vacuum method relies on the detection of a tracer gas that is present naturally or added by flushing the blister package just before it's sealed. Under vacuum conditions, the gas escapes through any pinholes or cracks that are present and is detected by an infrared sensor ("Pac Guard Model 400," leak tester, Mocon, Minneapolis, MN, www.mocon.com).

Some leak detectors look for pressure or vacuum decay to identify leaks. One dual-mode unit detects defects as small as 5 μm in a pressue or vacuum test cycle that lasts less than 30 s. An audible and visual reject alarm alerts the operator if the blister card fails. The system can store 100 different tests and record 5000 results. A two-way RS232 connection links the unit to a host system or network and enables uploading of test data and downloading of test parameters. Interchangeable test chambers accommodate various card sizes and shapes ("TME Solution-C Blister Card Chamber Test System," TMElectronics, Inc. , Boylston, MA, www.tmelectronics.com).

Many leak detectors rely solely on vacuum decay. If a leak is present when the vacuum is pulled, air leaves the flawed blister cavity and creates a concave surface in the lidstock. If the card is leak-free, the lidstock over each cavity domes slightly from the internal air pressure. There are various ways to measure these surface changes, including contact methods such as proximity sensors and load cells and noncontact imaging systems. Each has advantages and disadvantages.

The ideal blister-pack leak tester

Proximity sensors require complex tools to position them above each blister cavity. Hence, each blister format requires its own array. This requirement is an expensive proposition in tooling, operation, and validation if multiple blister formats must be inspected.

A load cell measures how much the lidstock pushes against it. Load cells can be difficult to adjust and validate because each cavity requires a corresponding cell. The technology also may be impractical for blister formats with cavities that are arranged extremely close together.

Imaging-based testing

Some imaging systems use a laser as a light source. These systems can be affected by variations in reflectivity caused by printing on the lidstock. The systems also require considerable precision because the center point of each blister must be identified to reveal changes in position.

Another camera-based system combines vacuum technology with indirect imaging. Key to this patented and patent-pending equipment is a flexible bladder positioned over the blister as it rests in the test fixture. The flexible membrane serves several purposes. During the test, it supports the seal and prevents deformation and peeling. It also increases the sensitivity of the pressure measurements. Finally, covering the lidstock not only eliminates problems the imaging system might have with printed or glossy surfaces, but also provides a clear indication of the flaw locations and identifies leaks as tiny as 3.5 μm.

After a test cycle of less than 10 s, results are presented in numerical, statistical, and image formats. The touch-screen operator interface and personal-computer–based controller also network with a printer and database to print and archive test results. Changeover is almost instantaneous and requires lifting out one fixture and setting a new one in its place. Once in position, the machine reads a notched identification code on the fixture and sets test parameters for the new blister card ("VeriPac 225/BLV," vacuum leak-testing system, Packaging Technologies and Inspection, LLC, Tuckahoe, NY, www.ptiusa.com).

To increase throughput, test fixtures can be designed to hold four blisters. In addition, the semiautomatic machine could be equipped with a shuttle system that would allow the operator to remove a tested card and replace it with an untested card while a test cycle is running. With suitable integration and material-handling hardware, an on-line model could perform 100% inspections. Moving leak testing on-line would require integration with the blister-packaging machine and related material handling. Most likely, this would take the form of a pick-and-place device that would move the blister cards to the leak tester and then back onto the line. In addition, the tester itself would probably require multiple stations to keep up with line speeds. This means on-line testing could cost significantly more than off-line testing, perhaps as much as 10 times more. Experts say it also would be harder to validate.

FYI

For sensitive, expensive drugs, however, the expense could be justifiable. At least one other alternative is available to drug makers not quite ready to make that investment: surface inspection of the formable web before filling and sealing. This imaging-based system potentially could eliminate some leakers by detecting visually imperceptible cracks, pinholes, and other defects as small as 10 μm at 3.75 m/s. Especially suitable for foil, the unit also can check nontransparent film and laminated paper. Multiple zone inspection prevents rejection of good blisters ("VisioScan," inspection system Uhlmann VisioTec, Towaco, NJ, www.uhlmann-visiotec.com).

Meanwhile, when implementing a leak-testing program, quality specifications for incoming materials must be considered. When looking at in-house quality control, a product's sensitivity and cost will determine whether leak testing is needed, whether it's off- or on-line, and how sensitive the equipment must be. Although it may technically be possible to locate defects measuring just a few micrometers, it's only necessary to do so if a hole that size will affect the product negatively. Other considerations in equipment selection include data-collection and networking capabilities, ease of calibration and validation, material limitations, tooling costs, cycle time, and changeover.

Hallie Forcinio is Pharmaceutical Technology's Packaging Forum editor, 4708 Morningside Drive, Cleveland, OH 44109, tel. 216.351.5824, fax 216.351.5684, editorhal@cs.com.