Design and Fabrication of Si Microchip Inlet for Mass Spectrometry

Handling of liquid samples in mass spectrometry is particularly challenging with failure modes such as solvent interference, compromised sensitivity and vacuum integrity in the mass spectrometer. This underscores the critical importance of an effective sample introduction method capable of seamlessly converting a liquid sample into its vapour phase while preventing solvent interference and avoiding selective loss or discrimination among the volatiles in the mixture.

Among the state-of-the-art sample introduction methods, the micro-perforated membrane chip (MPMC), fabricated using a silicon-on-insulator (SOI) wafer, is a promising solution. The MPMC acts as an interface between a liquid-phase environment and an electrochemical mass spectrometer (EC-MS) vacuum. It mainly transfers dissolved volatiles from the liquid phase to the vapour phase by evaporation, using an inert carrier gas to guide the analytes through a pressure-matching microcapillary into the mass spectrometer. Ideally, this setup ensures that all analytes are effectively delivered into the mass spectrometer, resulting in high sensitivity and fast response time. Despite these advantages, the MPMC has several limitations: (i) low-pressure differential tolerance across the membrane under operating conditions, (ii) poor detection efficiency for low-volatility species due to limited evaporation, and (iii) the fixed thickness of the device and the buried oxide layers in the SOI wafer limit flexibility in adjusting the membrane thickness. These challenges motivate the fabrication of the nano-perforated membrane chip (NPMC), which addresses these issues while retaining the MPMC’s compact form factor and base functionality.

The NPMC was fabricated at a wafer scale using advanced microfabrication technologies in monocrystalline Si wafers. Its design features a suspended nano-perforated membrane (NPM) on one side and a gas channel system on the other side of the wafer. Two reliable and reproducible strategies for NPM formation and the challenges encountered were demonstrated. These strategies involved nanopore etching using either the DREM (Deposit, Remove, Etch Many times) or ORE (Oxidize, Remove, Etch) processes, sidewall protection combined with thermal oxidation and PECVD oxide or rapid thermal oxidation, and finally, isotropic etching of silicon. Limitations of using C4F8 as passivation, such as its inconsistent sidewall protection on a wafer scale and lack of reproducibility, were also presented. Alumina was chosen as the mask material for nanopore etching due to its high selectivity, and a systematic parameter optimization for inductively coupled plasma (ICP) alumina etching was carried out to improve the resist etch selectivity while achieving a nearly vertical etch profile. The fabrication process and key challenges in gas channel system formation were thoroughly addressed.

The sensitivity, response time, and pressure difference tolerance of the NPMC were evaluated using an EC-MS under two scenarios through breath and water-drop tests. The NPMC exhibited a pressure tolerance of 1 bar, which is lower than the theoretical value but significantly higher than the MPMC's 0.2 bar tolerance. Discrepancies observed between the theoretical and experimental capillary burst pressure (CBP) values of NPMC were investigated. Furthermore, the study demonstrated a significant improvement in the detection limit of low-volatility species, such as ethanol, by almost 2 orders of magnitude by operating the NPMC at 30 mbar instead at 1 bar, ultimately leading to the carrier gas-free operation of the NPMC, typically with water vapour as an inherent carrier gas.

The feasibility of carrier gas-free NPMC for in-line dissolved gas analysis, particularly in a 50 L bioreactor tank, was explored using in-line mass spectrometry (IMS), a significantly simplified and compact version of EC-MS. The performance of IMS-NPMC was benchmarked against conventional instruments by two experiments to harness the potential of this technology, assessing whether it could detect all changes captured by traditional instruments and further identify any previously undetected changes. These experiments included monitoring DVG (i) in bulk water under different sparging conditions and (ii) during the alcohol fermentation process using Saccharomyces cerevisiae. The IMS-NPMC system captured dynamic fluctuations in DVG concentrations caused by evolving chemical conditions, providing detailed insights into trends and anomalies in the process. These findings highlighted the IMSNPMC’s versatility and real-time data acquisition capabilities for multiple DVGs analysis. This established the IMS-NPMC as a potentially transformative technology for in-line process monitoring in bioreactors and beyond.

The study concludes with recommendations for future work, including optimizing NPM geometry to increase its CBP, surface functionalization to withstand corrosive environments (highly acidic or base), and refining the fabrication process by simplifying the gas channel system. These advancements will further solidify the IMS-NPMC’s role in broader industrial use cases where real-time analysis is crucial for process optimization.