Measuring headspace gas concentrations and residual moisture content

Technological advances and regulatory documents, such as Annex 1 of the EU Guideline for Good Manufacturing Practice, have encouraged the pharmaceutical industry to install various new technologies into their production lines.

Headspace analysis
Gas concentrations, such as oxygen or moisture, can be reliably detected in freeze‑drying vials by guiding a wavelength modulated low‑power laser beam (class 1) through the headspace. We have improved this process to enable up to 600 vials/min to be tested, with only a few laser based absorption test heads mounted on a continuous motion rotor. Head Space Analysis (HSA) technology offers enhanced detection sensitivity and sensor head stability, and can provide accurate measurements of the smallest oxygen levels or, by analysing moisture concentrations, provide information regarding the remaining absolute pressure in the headspace. The benefit of this is a final quality concept that can be used as a release test — all products are being tested and individual faulty products are reliably detected. This also fulfills the Annex 1 requirements of the EC’s Guide to Good Manufacturing Practice (February 2008), which states: “Containers sealed under vacuum should be tested for maintenance of that vacuum after an appropriate, predetermined period”. The technology can also be used as a leak detector — by allowing the vials a certain time period to change headspace conditions, leak sizes of less than 1 μm can be reliably detected.

As well as being used to improve the quality of lyophilised products, the technology can also be used as an effective PAT tool. One of our clients, for example, a pharmaceutical company based in France, has received authorisation to use this testing technology instead of placing their capping process into a Grade A air supply until the cap has been crimped, as requested by Annex 1 of the EU Guideline to Good Manufacturing Practice.

NIR inspection
Our NIR inspection system for determining moisture content in freeze‑dried products uses NIR diffuse reflection spectroscopy and is capable of handling up to 3000 vials/min. The vial containing the residual material of the lyophilisation process is exposed to a broadband light source that contains a high amount of NIR radiation in the range 1000–2200 nm. This radiation interacts with the lyophilised product and reflects the portion of radiation that has not been absorbed from the material. The lost energy along the NIR wavelength range provides a specific spectral response that enables individual chemical and physical properties to be identified. One of the properties that can be extracted is a signal of approximately 1920 nm, which relates to a specific absorption of water molecules. This signal scales directly to the relative water content in the lyophilisation cake.

The NIR diffuse reflection is collected and correlated with a predefined model of water absorption. The calibration model is created using model values created by varying known water concentrations. In turn, this model is used to predict the residual water content in unknown samples by simply recording and evaluating the NIR spectrum taken from vials on a production machine to achieve the appropriate moisture value in percent residual humidity.

The freeze‑drying process is well known for producing unidentified, out-of-specification products when it comes to moisture content — even under strict parameter control. The above system, however, can be a useful tool for providing a final quality check.

PAT
From a PAT point of view, the system demonstrates successful implementation of direct process understanding — using scientific methods to implement a system that ensures the quality of the complete production process for freeze‑dried products.

The described NIR function, provided by Uhlmann‑VisioTec (Germany), can be integrated as a single sensor into a headspace analysing system to provide the user with a final quality testing approach where gas concentration information is received from the headspace and residual moisture or meltback information is received from the lyophilisation cake. Vials that don’t meet certain criteria are automatically removed from the production line while good vials can continue on to cosmetic inspection and labelling.

Based on a contribution by Gerhard Schramm, Sales Director on Head Space Analysis Technology at Wilco AG Switzerland, and Dr Heino Prinz, Head of PAT and Consulting at Uhlmann‑VisioTec Germany.