Micrometer-Scale Magnetic-Resonance-Coupled Radio-Frequency Identification and Transceivers for Wireless Sensors in Cells

We report the design, analysis, and characterization of a three-inductor radio-frequency identification (RFID) and transceiver system for potential applications in individual cell tracking and monitoring. The RFID diameter is $22\mu \mathrm{m}$ and can be naturally internalized by living cells. Using magnetic resonance coupling, the system shows resonance shifts when the RFID is present and also when the RFID loading capacitance changes. It operates at 60 GHz with a high signal magnitude up to $-50\mathrm{dB}$ and a sensitivity of 0.2. This miniaturized RFID with a high signal magnitude is a promising step toward continuous, real-time monitoring of activities at cellular levels.

Figure 1

Schematic illustrating the instrumentation for wireless communication with RFIDs in cells. (a) RFID-tagged cells are flown through a PDMS microfluidic channel and interrogated by the TRX. The signal is probed through GSG pads connected to the TRX by a CPW TL. (b) Enlarged schematic of the sensor, signal conditioner, and the three-inductor RFID and TRX wireless front end.

Figure 2

Models of the coupling among the multiple-loop RFID and TRX as a three-port network $\mathbf{Z}$. The self-impedance terms are considered as an inductor in series with a resistor. The mutual impedance terms represent the mutual inductance. The load impedances are capacitors that cancel out the TRX2 and RFID inductors at the same resonance frequency.

Figure 3

Evolution of the RFID and TRX resonances in reflection coefficients $S_{11}$ with RFID presence and its loading capacitance changes. With $100\%$ $C_3$, the RFID resonates at the same frequency with TRX2. (a),(c),(e) Three-loop data from analytical model, ansys hfss 3D electromagnetic (EM) simulation, and experimental measurement. (b),(d),(f) Corresponding two-loop data.

Figure 4

The simulation setup (a) and simulation results (b) of the RFID and TRX fabricated using the TSMC 40-nm general process. The inductor and capacitor models are extracted separately from PeakView EMD™ and TSMC vendor models. The simulation results have similar resonance frequencies with the models, HFSS simulations, and measurements in Fig. 3.

Figure 5

(a) Measurement structures for RFID and TRX and deembedding structures. (b),(c) Measurement structure and results for the three-loop RFID and TRX connected with a $300\text{−}\mu \mathrm{m}$-long CPW TL attached.

Figure 6

Time-sequence bright-field microscopy images of the cellular uptake and interaction of the house-fabricated RFID. The cells used are mouse melanoma cells. The complete video is included in the Supplemental Material [40].

Figure 7

The measured resonance response of the house-fabricated RFID. The inset shows the tilted top-down scanning-electron-microscopy image of a RFID on the substrate.

Figure 8

Characterization of the sensor system for optimal operating frequency. The modeled signal magnitude and sensitivity are plotted from up to 60 GHz, with an ideal loading capacitor for simplicity.

Figure 9

HFSS-simulated RFID and TRX signal sensitivity $\sigma$ and nonlinearity $\nu$ dependence with respect to the normalized detection range $R/D$. ${\sigma }_L$ and ${\nu }_L$ are calculated with the lower split resonance, while ${\sigma }_A$ and ${\nu }_A$ with the overall resonance trough. (a) Sensitivity $\sigma$ and nonlinearity $\nu$ trends with respect to $R/D$. (b) Magnetic field distribution at the $x\text{−}z$ cross section through the center of the TRX1 and TRX2. This cross section is indicated in the inset. The TRX1 and TRX2 are at the bottom of the figure with their wires outlined. The field intensity diminishes above around unit $R/D$. (c)–(f) Resonance evolutions as the RFID is gradually elevated away from the TRX. $R/D$ increases from 0.1 to 1.2.

Figure 10

HFSS-simulated RFID and TRX signal spectra when the RFID is placed at $7.5\mu \mathrm{m}$ above the TRX for different lateral positions. The RFID starts from the center (a), moves to be right within TRX2 (b), right within TRX1 (c), and finally moves almost out of TRX (d).