The detection of rare cells in complex media like blood is crucial for biomedical research and clinical diagnostics. Current methods often involve cumbersome sample preparation. Micro-Hall detectors (µHDs) offer a promising alternative due to their high sensitivity and ability to operate in complex media without extensive sample preparation. However, limitations such as low throughput, clogging susceptibility, and incompatibility with commercial CMOS foundry processing hinder their clinical translation. This research addresses these challenges by developing CMOS-compatible graphene Hall sensors integrated with PDMS microfluidics. The use of graphene offers advantages due to its high carrier mobility and ease of integration with CMOS chips. The integration of microfluidics allows for controlled flow of the sample over the sensors. The passivation layer protects the graphene from biofouling and ensures long-term stability in whole blood. The researchers aim to improve the throughput and robustness of µHDs while maintaining their high sensitivity, paving the way for clinical applications.

Existing methods for rare cell detection include immunomagnetic labeling followed by detection using various techniques such as microfluidic magnetic separation, giant magnetoresistance (GMR) sensors, magnetic susceptometry, nuclear magnetic resonance (NMR), and Hall effect sensors. Each method has its own advantages and limitations. µHDs have shown promise due to their high sensitivity and linear response to magnetic fields, especially compared to NMR which suffers from low signal-to-noise ratio at cell-sized volumes. GMR sensors, while sensitive, operate within a narrow dynamic range. µHDs can handle the larger fields needed for full MNP magnetization, enhancing the signal. However, their serial interrogation of cells limits throughput. To overcome this, the integration of large arrays of µHDs with on-chip CMOS electronics is proposed. Prior work has integrated magnetic sensors with CMOS, but challenges remain in terms of sensitivity and biocompatibility. Two-dimensional materials (2DMs), particularly graphene, offer a solution due to their high carrier mobility and ease of integration. Graphene has shown superior performance compared to traditional semiconductor materials in Hall sensing applications. However, challenges remain in ensuring the long-term stability of the graphene sensors in biologically complex fluids. This paper addresses these limitations by combining graphene Hall sensors, microfluidics, and CMOS technology.

The researchers developed a fabrication strategy combining graphene Hall sensors (µGSs) with microelectronics and microfluidics. Graphene was grown via chemical vapor deposition (CVD) and transferred onto a silicon chip with prefabricated gold electrodes. The µGS array was patterned using photolithography and oxygen plasma etching. A passivation layer of HSQ and silicon nitride (Si3N4) was added to protect the graphene from biofluids. PDMS microfluidic channels were then bonded to the chip. The graphene quality was characterized using atomic force microscopy (AFM) and Raman spectroscopy. The electron mobility of the µGSs was measured using the direct transconductance method. Magnetic sensing performance was evaluated in dry conditions and in BSA and whole blood using magnetic agarose beads as a model system. The minimum detectable magnetic field was determined using an FFT spectrum analyzer. The stability of the system was assessed over extended periods of operation in whole blood.

The CVD-grown graphene was confirmed to be monolayer and exhibited high quality based on AFM and Raman spectroscopy results. The average electron mobility of the passivated µGSs was 4600 ± 300 cm²V⁻¹s⁻¹, a significant improvement over silicon sensors, though reduced by ~30% after passivation. In dry conditions, the µGS demonstrated a linear relationship between Hall voltage and magnetic field strength, with an absolute sensitivity of 175 mVT⁻¹ and current-related sensitivity of 175 VA⁻¹T⁻¹. The minimum detectable magnetic field was 1.35 µTHz⁻⁰·⁵ at 3 kHz. The system demonstrated stable operation for up to 39 hours in human whole blood. The ability to detect magnetic agarose beads, a model for cells, in BSA and whole blood demonstrated the successful integration of the graphene Hall sensors, microfluidics, and passivation layer for use in complex biofluids.

The results demonstrate the successful fabrication and integration of high-performance, stable graphene Hall sensors with microfluidics for magnetic sensing in whole blood. The use of graphene and a robust passivation strategy allowed for achieving high sensitivity and long-term stability in a biologically relevant environment. The ability to detect magnetic beads in whole blood shows promise for future applications in rare cell detection, eliminating the need for extensive sample preparation and simplifying the workflow. The linear response of the sensors over a wide range of magnetic fields and the tunability of the sensor response through back-gate voltage control provide further advantages for diverse applications. The achieved performance surpasses many existing Hall sensors, showcasing the potential of this integrated approach.

This study successfully demonstrated a highly stable and sensitive platform for magnetic sensing in whole blood using integrated graphene Hall sensors and microfluidics. The platform achieved performance comparable to or exceeding other high-performing Hall sensors, showcasing the potential for applications in clinical diagnostics, particularly for detecting rare cells. Future work will focus on integrating these sensors into a custom CMOS chip for parallel detection and improved throughput, ultimately facilitating real-world clinical applications.

The current study utilizes magnetic agarose beads as a model system for rare cell detection. Further validation is needed using actual rare cells to confirm the platform's effectiveness in a more clinically relevant setting. The long-term stability was tested up to 39 hours; longer-term studies are necessary to fully assess the durability of the system for continuous operation. Although the passivation layer enhances stability, the effect of prolonged exposure to various biological fluids on sensor performance needs further investigation. While the current device demonstrated proof-of-concept functionality, further optimization and miniaturization may be needed for practical clinical implementation.