Fast Electrical Impedance Spectroscopy for Characterization and Counting - Q&A

March 23, 2021 by Meng Li

This blog post summarizes the webinar that Fluigent and Zurich Instruments presented together on our solution for fast electric impedance spectroscopy (EIS) in microfluidics. The system consisted of an HF2LI Lock-in Amplifier from Zurich Instruments and the LineUp series microfluidic flow controllers from Fluigent. We covered the working principle of EIS, introduced how to set up the measurement system, and provided two demos in detecting microbeads and water-in-oil droplets. We hope this webinar was insightful not only to researchers in microfluidics, but also in electrochemistry, materials science, and food science. You can also take a look at our impedance-related applications.

In case you missed the live event, you can watch the video recording here:

[Webinar Replay] Fast Electrical Impedance Spectroscopy for Characterization and Counting | FLUIGENT

Due to the limited time available during the webinar, Bruno and I only provided brief answers to the questions raised by the audience. We answer them here in full detail.

Can you control instruments from Zurich instruments using the Python API?

Python is one of the five supported application programming languages (others being C, MATLAB®, LabVIEW™ and .NET), available in all Zurich Instruments lock-in amplifiers and impedance analyzers. And besides exhaustive documentation in our LabOne programming manual, we also include common examples in the Python package (called as 'zhinst'), which you can download from either our website, pypi, or running 'pip install --upgrade zhinst' in your Python compiler. To quickly learn the Python API, you can also make good use of the command log in LabOne. Simply start the LabOne GUI, and see how the change you just make is translated into the latest syntax in the command log.

How many readings per second did we see in the demo? What is the fastest rate to readout?

The second demo shows a flow velocity at tens of droplets per second. Theoretically, with the HF2LI lock-in amplifier, the detection speed is limited by the low-pass filter bandwidth (maximum at ~200 kHz) as well as the data transfer rate (maximum at ~460 kSa/s). In practice, however, the speed will also be limited by other factors, including the pressure stability of setup (droplet itself, as well as the chip), the noise from the environment, and so on. In literature, a high throughput at a few thousand per second has already been reported using the HF2LI [1].

What happens when more than 2 particles pass the sensing electrodes at the same time?

If two identical particles pass the sensing electrode pair at exactly the same time, then you would see a stronger deviation (peak or trough) from the baseline. This may lead to a misconception that the signal is from a larger particle. To find the truth, one can possibly use a microtubular chip, which allows EIS measurement at different angles [2]. That is why particle alignment in the microfluidic channel is extremely important, which you can achieve with Fluigent's pressure controller. On the other hand, if the two particles are different, or if they pass at a slightly different time, then EIS can usually distinguish them thanks to the high sensitivity of HF2LI.

How would you reduce the noise of the baseline?

To suppress the noise, as introduced in this webinar, the differential current measurement scheme is the key. This technique subtracts the background from the noisy flowing medium, such that we will only detect signals from the particles. Effectively, it helps to extend the dynamic range of the HF2LI. You can also lower the low-pass filter bandwidth to reject the noise, but do make sure it is at least faster than the particle flow rate, in order to not miss any important flow events.

What is the maximum speed of cells that can still yield a correct measurement of the intracellular high-frequency components?

This question is similar to Question 2. The maximal throughput (a few thousand cells per second [1]) is limited by various factors. And in most cases, not from the HF2LI. This is because the HF2LI has a sufficiently wide low-pass filter bandwidth and data transfer rate to ensure fast and weak microfluidic events to be captured. For intracellular high-frequency components typically higher than 1 MHz, a bandwidth at 100 kHz (or lower) can help to reach a good signal-to-noise ratio, while not sacrificing too much temporal resolution.

What is the maximum sample rate of the HF2LI?

The data transfer rate (as known as 'readout rate'), which describes how fast the measured data are transferred to a PC and becomes visible to you in the end, has a maximum of 460 kSa/s. If you are running simultaneous multi-frequency EIS, please refer to the HF2 user manual or the table below for the maximal rate available.

Active demodulators Maximum readout rate
1 460 kSamples/s
2 – 3 230 kSamples/s
4 – 6 115 kSamples/s
7 – 8 57 kSamples/s

How do you avoid the creation of air bubbles in the microfluidic channel?

If you manage to set all the tubings and reservoirs correctly, you will likely see no or little air bubbles in microfluidic devices.

Can you give more information about the EIS chip company?

The EIS chip used in this webinar is from BEC. If you are interested in the chip details, please contact us.



[1] Spencer, D., & Morgan, H. (2020). High-speed single-cell dielectric spectroscopy. ACS sensors5(2), 423-430.

[2] Weiz, S. M., Medina‐Sánchez, M., & Schmidt, O. G. (2018). Microsystems for Single‐Cell analysis. Advanced Biosystems,2(2), 1700193.