Experimental data for electrochemical impedance characterization of Li-ion batteries under varying state of charge, current amplitude and rest time

Abstract

State of Health (SOH) monitoring is critical for the efficient and reliable operation of lithium-ion batteries, particularly in electric vehicle applications. Electrochemical Impedance Spectroscopy (EIS) is a widely used non-destructive technique for estimating the SOH of batteries, as it provides detailed insights into the battery’s internal condition across a range of frequencies. In this study, datasets collected from four cylindrical lithium-ion batteries manufactured by Molicel of model INR-21,700-P42A under various operating conditions were presented. Each cell was tested under a range of operating conditions, like four State of Charge (SOC) levels: 80%, 60%, 40%, and 20%; five current amplitudes: 30 mA, 50 mA, 100 mA, 500 mA, and 1 A; and three rest times: 30 min, 1 hour, and 2 h. The experiments were conducted using an Arbin battery cycler, and EIS tests were performed in galvanostatic mode using a Gamry Interface 5000P Potentiostat. The experiment resulted in 60 unique test conditions per battery (4 SOC levels × 5 current amplitudes × 3 rest times). The entire experimental procedure was repeated for four battery cells, resulting in a total of 240 impedance data. The resulting dataset consists of frequency, real and imaginary components of impedance data. This data is used to generate Nyquist plots. This plot provides information about different parameters like Ohmic resistance, solid electrolyte interface capacitance and resistance, charge transfer resistance, double layer capacitance and ionic diffusion. These parameters are sensitive to battery aging and can be used to assess the battery’s SOH.

Specifications Table

Subject	Engineering & Materials science
Specific subject area	Electrochemical Impedance Spectroscopy of Li-ion Batteries
Type of data	Raw Data
Data collection	Pilot testing - This test was conducted initially to select appropriate AC amplitudes for main EIS experiment. The tests were conducted on the single cylindrical cell made by Molicel of type INR-21,700-P42A, labelled MCLE3231. The test was conducted over a frequency range of 0.1 Hz to 20 kHz using different current amplitudes 1 mA, 2 mA, 3 mA, 4 mA, 5 mA, 10 mA, 15 mA, 20 mA, 25 mA, 30 mA, 50 mA, 60 mA, 70 mA, 80 mA, 90 mA, 100 mA, 500 mA, and 1 A. Variability testing - The data was collected on four cylindrical batteries made by Molicel of type INR-21,700-P42A labelled as MCLI01, MCLI02, MCLI03, MCLI04. The experiments were performed at different SOC levels (80 %, 60 %, 40 %, and 20 %), various current amplitudes (30 mA, 50 mA, 100 mA, 500 mA, and 1 A), and rest times (30 min, 1 hour, and 2 h). All tests were conducted over a frequency range of 0.01 Hz to 10 kHz. The experiments were conducted using two instruments Arbin battery cycler and Gamry 5000P potentiostat. The Arbin battery cycler instrument (LBT21084, Arbin Instruments, United States of America) [1] includes 16 independently controlled channels each with a voltage range of 0–5 V and a current range of ±10 A. The cells to be tested are placed on battery holder and connected to selected channel in Arbin Cycler. The EIS experiment was performed by integrating a Gamry Interface 5000P Potentiostat with the Arbin cycler [4]. This device is designed for single cell testing and it supports two modes of operation namely: Galvanostatic and Potentiostatic. In this study, all the experiments were performed at Galvanostatic mode, where a sinusoidal current of certain amplitude is applied and measuring the corresponding output voltage. Gamry device supports currents up to 5 A and a maximum voltage of ±6 V. The Arbin battery cycler and Gamry device is controlled using MITS Pro software provided by Arbin. The data files (CSV) can be imported by selecting the desired test channels and choosing the export test option from the menu bar in the Data Watcher window in MITS Pro Software.
Data source location	Institution: University of Windsor City: Windsor, ON Country: Canada Latitude and longitude for collected data: 42.3043 N, 83.0660 W
Data accessibility	Repository name: Experimental Data for Electrochemical Impedance Characterization of Li-ion Batteries Under Varying State of Charge, Current Amplitude and Rest Time Data identification number: 10.17632/dcbrc4zmc5.2 Direct URL to data: https://data.mendeley.com/datasets/dcbrc4zmc5/2
Related research article	B. Balasubramanian, P. Pillai, A.H.Sakr, B. Balasingam, “Quantification of Drift in Battery Impedance Spectroscopy Due to State of Charge, Current Amplitude and Rest Time”, in IEEE Transactions on Transportation Electrification (conditionally accepted), 2026 [2].

1. Value of the Data

• •This dataset can be used to train machine learning models that automate the interpretation of EIS data and evaluate the SOH of Li-ion batteries. This data provides a complete dataset to create and validate models/analysis, across varying operating conditions [5].
• •The experiment used standardized instruments (Arbin LBT21084 battery cycler and Gamry 5000P Potentiostat), making the dataset valuable for academic researchers using similar hardware setups to directly replicate or extend the experimental procedures for their study.
• •This dataset can be useful for both academic researchers and industrial engineers to gain a deeper understanding of the performance of the battery cell (INR-21,700-P42A) at different SOC levels, rest time and current amplitudes.
• •Battery equivalent circuit model (ECM) based approaches can use the dataset to extract battery parameters like ohmic resistance, charge transfer resistance and solid electrolyte interface resistance.
• •This dataset helps researchers to better understand and analyze the uncertainties in EIS due to variations in SOC levels, rest periods, and current amplitudes.

2. Background

Lithium-ion batteries (LiBs) are vital to modern energy storage, powering consumer electronics, electric vehicles (EVs), and grid systems due to their lightweight, high-energy density, low self-discharge, and improved cost. Despite recent advancements, characterizing and understanding the internal electrochemical processes of LIBs remains a significant challenge. EIS is an effective method for evaluating the internal condition of LiBs across a range of frequencies and aging states. The impedance spectrum obtained through EIS is influenced by several factors like SOC, rest time and current amplitude. Therefore, understanding the effects of these factors on EIS measurements is crucial for real-world applications. This dataset was created to investigate the influence of SOC, rest time, and current amplitude on the electrochemical behaviour of LiBs. In this experiment, EIS is performed by applying a sinusoidal current and measuring the corresponding output voltage. EIS generates impedance spectra that reveal internal battery parameters such as ohmic resistance, charge transfer resistance and solid electrolyte resistance through a Nyquist plot. Monitoring changes in these internal parameters over time helps to assess the battery’s SoH.

3. Data Description

This section details the organization of data from two experimental studies: pilot testing and variability testing. Pilot testing was conducted as a preliminary step to select the current amplitude for the subsequent main EIS experiment. Following this, variability testing was performed to analyse the effects of different operating condition like state of charge (SOC), rest time, and current amplitude on impedance measurements. The dataset collected from pilot testing and variability testing are arranged in two folders to keep the collected data structured and easy to navigate. The arrangement and structure of these folders are discussed in the following subsections.

3.1. Pilot testing data description

The data collected during pilot testing are stored as CSV (Comma-Separated Values) files inside a folder named “Factor_CurrentAmplitude”. This folder contains a total of 18 CSV files, each corresponding to a different current amplitude. The files are named using the format “BatteryName_CurrentAmplitude”. For example, the filename “MCLE3231_1A.csv” refers to data collected from the battery labelled MCLE3231 under an applied current amplitude of 1 A.

Each CSV file consists of three columns:

• •The first column represents the frequency of the current excitation signal, ranging from 0.1 Hz to 20 kHz.
• •The second column shows the real part of the impedance.
• •The third column contains the imaginary part of the impedance.

Additionally, a MATLAB script named “demo01_CurrentAmplitude.m” located in the same “Factor_CurrentAmplitude” folder is used to generate Nyquist plots for all current amplitudes. Fig. 1 shows the folder structure including the CSV and MATLAB files organized in the pilot testing.

Fig. 1.

Organization of the pilot testing dataset. Data is organized by the battery number and the current amplitude at which the EIS test is performed.

3.2. Variability testing data description

The data collected during the variability testing are in CSV format and are stored in a folder named “Factor_Resttime.” This folder contains 12 CSV files corresponding to four batteries (MCLI01, MCLI02, MCLI03, and MCLI04) tested across three rest times: 0.5 hr, 1 hr, and 2 hrs. Each file is named using the format “BatteryName_Resttime.csv.” For example, ``MCLI01_0.5hr.csv'' refers to data collected from battery MCLI01 after a rest duration of 0.5 h. In addition to the CSV files, the folder includes three MATLAB scripts used to generate Nyquist plots based on different test factors: rest time, SOC, and current amplitude. The file demo01_resttime.m is used to analyze Nyquist plots for different rest times, demo02_SOC.m is used to examine the plots across varying SOC levels, and demo03_currentAmplitude.m analyzes the plots based on different current amplitudes. Fig. 2 illustrates the folder structure and organization of the CSV and MATLAB files used in the variability testing.

Fig. 2.

Organization of the variability testing dataset. Data is organized by the battery number and the rest time after which the EIS test is performed.

The EIS Measurements CSV files are organized in 17 columns and 1221 rows. Table 1 describes some important columns of the CSV files.

Table 1.

Description of columns in the recorded EIS data files.

Column Name	Column Number	Description
Test_Time	2	Time at which the test was conducted
Step_ID	3	Identifier for the test step
Cycle_ID	5	Identifier for the cycle number(charge/discharge)
EIS_Test_ID	6	Identifier for the Electrochemical Impedance Spectroscopy (EIS) test
EIS_Data_Point	7	Index of data point in the EIS measurement
Frequency	8	Pulse Frequency (Hz)
Zmod	9	Magnitude of impedance
Zphz	10	Phase angle of impedance
Zreal	11	Real part of impedance
Zimg	12	Imaginary part of impedance
OCV	13	Open Circuit Voltage measured in the battery
AC_Amp_RMS	14	Root Mean Square (RMS) value of the AC current amplitude used in the measurement

Similar to the column information, the row data in CSV files also plays an important role in identifying the SOC levels and current amplitudes of EIS measurements. The SOC levels can be determined from the EIS_TestID column, where values range from 0 to 19. For example, when EIS_TestID = 0, it corresponds to SOC of 80 %. A detailed mapping of SOC levels for different current amplitudes based on EIS_TestID are provided in Table 2. The current amplitudes are identified using the AC_Amp_RMS column. For instance, a value of 0.03 A indicates the current amplitude used during the EIS measurement. All CSV files within the Factor_Resttime folder follow the similar structure as described in Table 2.

Table 2.

Description of rows in the recorded EIS data files.

Row Number	EIS_TestID	SOC	Current	Description
1				Header information
2–62	0	80 %	0.03 A	EIS measurement carried at 80 % SOC at current amplitude 0.03 A
63–123	1	80 %	0.05 A	EIS measurement carried at 80 % SOC at current amplitude 0.05A
124–184	2	80 %	0.1 A	EIS measurement carried at 80 % SOC at current amplitude 0.1 A
185–245	3	80 %	0.5 A	EIS measurement carried at 80 % SOC at current amplitude 0.5 A
246–306	4	80 %	1 A	EIS measurement carried at 80 % SOC at current amplitude 1 A
307–367	5	60 %	0.03 A	EIS measurement carried at 60 % SOC at current amplitude 0.03 A
368–428	6	60 %	0.05 A	EIS measurement carried at 60 % SOC at current amplitude 0.05 A
429–489	7	60 %	0.1 A	EIS measurement carried at 60 % SOC at current amplitude 0.1A
490–550	8	60 %	0.5 A	EIS measurement carried at 60 % SOC at current amplitude 0.5 A
551–611	9	60 %	1 A	EIS measurement carried at 60 % SOC at current amplitude 1 A
612–672	10	40 %	0.03 A	EIS measurement carried at 40 % SOC at current amplitude 0.03 A
673–733	11	40 %	0.05 A	EIS measurement carried at 40 % SOC at current amplitude 0.05 A
734–794	12	40 %	0.1 A	EIS measurement carried at 40 % SOC at current amplitude 0.1 A
795–855	13	40 %	0.5 A	EIS measurement carried at 40 % SOC at current amplitude 0.5 A
856–916	14	40 %	1 A	EIS measurement carried at 40 % SOC at current amplitude 1 A
917–977	15	20 %	0.03 A	EIS measurement carried at 20 % SOC at current amplitude 0.03 A
978–1038	16	20 %	0.05 A	EIS measurement carried at 20 % SOC at current amplitude 0.05 A
1039–1099	17	20 %	0.1 A	EIS measurement carried at 20 % SOC at current amplitude 0.1 A
1100–1160	18	20 %	0.5 A	EIS measurement carried at 20 % SOC at current amplitude 0.5 A
1161–1221	19	20 %	1 A	EIS measurement carried at 20 % SOC at current amplitude 1 A

4. Experimental Design, Materials and Methods

This section describes about the experimental setup, low-rate OCV-SOC test, pilot testing, variability testing and Nyquist plot for different rest time, SOC levels and current amplitudes.

4.1. Experimental setup

The dataset was obtained from four Molicel lithium-ion battery cells, specifically of type INR-21,700-P42A. The capacity of the batteries obtained using low rate OCV test are described in Section B. Low Rate OCV test, and the charge and discharge capacity of each battery are listed in Table 3. The cell to be tested is placed on a battery holder, and a channel is selected from the Arbin battery cycler. The Arbin battery cycler (Model LBT21084, manufactured by Arbin Instruments, United States of America) [1], supports 16 independently controlled channels. Each channel can operate within a voltage range of 0 to 5 V and a current range of ±10 A was used in this study. EIS measurements were carried out by integrating a Gamry Interface 5000P Potentiostat connected to the Arbin cycler. This system supports testing currents up to 5 A and voltages up to ±6 V. This device is designed specifically for single-cell testing. The device supports both galvanostatic and potentiostatic modes. The entire experiment was carried out in galvanostatic mode in this study. In galvanostatic mode, a small AC amplitude is applied over a range of frequencies from 0.01 Hz to 10 kHz to the battery and its corresponding voltage response is measured. In the experiment, the voltage cable (V–, green) and current cable (I–, black) from selected Arbin battery cycler channel are connected to the negative terminal of the cell and the voltage cable (V+, white) and current cable (I+, red) from the selected Arbin battery cycler channel are connected to the positive terminal of the cell placed in battery holder. The Arbin battery cycler and Gamry device are controlled using MITSPRO software installed on a PC provided by Arbin.

Table 3.

Charge capacity and discharge capacity for Molicel batteries.

Battery Label	Discharge Capacity $Q_d$ (mAh)	Charge Capacity $Q_c$(mAh)
MCLI01	4054.9	4057.2
MCLI02	4045.4	4045.3
MCLI03	4059.3	4059.4
MCLI04	4058.4	4061.6

4.2. Low rate OCV test

The charge and discharge capacity of each cell is computed to precisely estimate the SOC of battery for the experiment. To find the charge and discharge capacity of the battery, the battery is fully charged using a constant current and constant voltage (CC-CV) charging topology. Once the battery is fully charged, it is allowed to rest for 1 hour, and the low-rate OCV characterization test [3] is initiated by slowly discharging the battery with a constant current of $i_d$ = C/32 rate until the terminal voltage drops to the minimum threshold voltage of the battery $OC{V}_{min}=2.5V$. The total time taken to reach the cutoff voltage is referred to as the discharge time $T_d$. In the next step, the battery is slowly charged using a current $i_c$= C/32 rate until the terminal voltage reaches $OC{V}_{max}=4.2V$. The total charging time is denoted as $T_c$. From the data recorded during the low rate OCV-SOC characterization test, the charge capacity $Q_c$ and discharge capacity$Q_d$ are computed from the data as follows:

$$Q_c=i_c{T}_c$$ (1)

$$Q_d=i_d{T}_d$$ (2)

The charge and discharge capacities of each battery computed from (1) and (2) are listed in Table 3.

4.3. Pilot testing

Pilot testing was conducted as a preliminary step before the main experimental study, with the objective of selecting a suitable current amplitude for main EIS measurements. The tests were conducted on the single cylindrical cell made by Molicel of type INR-21,700-P42A. The cell was labelled as MCLE3231, and the test was conducted over a frequency range of 0.1 Hz to 20 kHz. The experiment was initiated by fully charging the battery using CC-CV charging topology followed by a series of EIS measurements by varying the AC current amplitude across a range of values: 1 mA, 2 mA, 3 mA, 4 mA, 5 mA, 10 mA, 15 mA, 20 mA, 25 mA, 30 mA, 50 mA, 60 mA, 70 mA, 80 mA, 90 mA, 100 mA, 500 mA, and 1 A.

Fig. 3 shows the Nyquist plot generated for all chosen current amplitudes. Each current amplitude is represented by a unique combination of marker style and color tone to enhance visual distinction in the Nyquist plot. Current amplitudes (1–5 mA) appear in orange to red tones, Current amplitudes (10–25 mA) are shown in green to yellow tones, while (30 mA to 1 A) are represented in varying blue tones. Specific markers assigned to each current amplitude are as follows: ``o'' for 1A, ``+” for 500 mA, ``*''for 100 mA, ``.'' for 90 mA, ``x'' for 80 mA, ``_'' for 70 mA, ``|'' for 60 mA, ``s'' for 50 mA, ``D'' for 30 mA, ``^'' for 25 ma, ``v'' for 20 mA, ``>'' for 15 mA, ``<'' for 10 mA, ``p'' for 5 mA, ``h'' for 4 mA, ``H'' for 3 mA, ``d'' for 2 mA, ``P'' for 1 mA.

Fig. 3.

Nyquist Plot for different AC amplitudes collected during the pilot testing for the purpose of selecting the current amplitude for the main experiments.

From Fig. 3, it was observed that a high level of noise was observed at the following low current amplitudes: 1 mA, 2 mA, 3 mA, 4 mA, 5 mA, 10 mA, 15 mA,20 mA, 25 mA and 30 mA. The impedance spectra from 50 mA onward appeared to be less noisy in the visual inspection. Therefore, the current amplitudes of 30 mA, 50 mA, 100 mA, 500 mA, and 1 A were selected for the main EIS experiments. The MATLAB script used to generate the Nyquist plots shown in Fig. 3 is provided in the file named demo01_CurrentAmplitude.m, located in the folder Factor_CurrentAmplitude.

4.4. Variability testing

The battery is fully charged (SOC = 100 %) using CC-CV charging topology, and it is further discharged by 20 % to perform the EIS test. The battery capacity varies for each cell (refer to Table 3); the specific capacities are used to precisely discharge the battery by 20 %. To decrease the battery SOC by 20 %, a constant current at a C/2 rate is applied for 24 min. For instance, the MCLI02 cell has a discharge capacity of 4.0454 Ah, a current of I = 2.0227A at C/2 (4.0454/2) rate is applied for 24 min (time=4.0454×0.2/2.0227). After each discharge step, the cell was allowed to rest for a specified duration before impedance measurements were taken. Initially, a rest time was set to 30 min. EIS measurements were performed at various alternating current amplitudes, starting with 30 mA. The same measurement process was then repeated for current amplitudes of 50 mA, 100 mA, 500 mA, and 1 A, while keeping the rest time 30 min. The battery is again discharged to 20 %, followed by EIS measurements. This sequence of discharging, resting, and EIS measurement was repeated until the battery’s SOC was decreased to 20 %. EIS tests are conducted in the fixed order of SOC levels 80 %, 60 %, 40 %, and 20 %. After completing the full set of measurements at a 30-minute rest time, the entire procedure was repeated for two additional rest durations, 1 hour and 2 h. The tests were continuous, and no aging or recalibration of the chosen batteries was executed between these repeated measurements. This approach is repeated for the other Molicel batteries. As a result, the dataset includes measurements at multiple SOC levels, rest times, and current amplitudes for each battery. The data set comprises 60 test conditions for a single cell (4 SOC levels, 5 current amplitudes, 3 rest times) and a total of 240 experiments for all 4 batteries. The total number of EIS measurements performed on all 4 Molicel batteries are listed in Table 4.

Table 4.

Summary of EIS measurement on batteries.

EIS Test Parameter	EIS Test Break point conditions	Total number of Test conditions
State of Charge (SOC)	80 % 60 % 40 % 20 %	4
Rest time	0.5hr, 1hr, 2hrs	3
Current Amplitude	30 mA, 50 mA, 100 mA, 500 mA, 1A	5
Number of operating points across single cell for EIS measurement		60 (4 SOC*3 rest times *5 current amplitudes)
Total number of experiments performed across all 4 batteries		240

4.5. Nyquist plots for different rest time, SOC and current amplitudes

This section describes about the Nyquist plots generated under varying conditions of rest time, SOC and current amplitudes.

4.6. Nyquist plots at different rest times

The Nyquist plot for the MCLI04 cell grouped by rest time 30 min, 1hr and 2 hrs are shown in subplots (a), (b) and (c) of Fig. 4. The distinction across current amplitudes is shown using varying marker symbols, i.e., “o” for 30 mA, “+” for 50 mA, “*” for 100 mA, “□” for 500 mA, and “x” for 1 A. Similarly, the distinction across SoC is shown using varying colors, i.e., purple tone for 80 %, blue tone for 60 %, green tone for 40 % and yellow tones for 20 %.

Fig. 4.

Nyquist plot for MCLI04 cell. Colors specify SOC and line markers specify current amplitude.

The MATLAB script demo01_resttime.m located in the Factor_resttime folder is used to generate the Nyquist plots shown in Fig. 4. To generate plots for different batteries the variable Batt can be modified in the MATLAB script. Setting Batt = 1 corresponds to Battery 1 (MCLI01), Batt = 2 to Battery 2 (MCLI02), Batt = 3 to Battery 3 (MCLI03), and Batt = 4 to Battery 4 (MCLI04).

4.7. Nyquist plots at different SOC

The Nyquist plot for MCLI04 cell recorded at four SOC levels 20 %, 40 %, 60 %, and 80 % are represented in subplots (a),(b),(c) and (d) of Fig. 5. The distinction across current amplitudes is shown using varying marker symbols, i.e., “o” for 30 mA, “+” for 50 mA, “*” for 100 mA, “□” for 500 mA, and “x” for 1 A. Similarly, the distinction across SoC is shown using varying colors, i.e., purple tone for 80 %, blue tone for 60 %, green tone for 40 % and yellow tones for 20 %.

Fig. 5.

Nyquist plot at different SOC levels for MCLI04 cell. Colors specify different rest times, and line markers specify current amplitudes.

The MATLAB script demo02_SOC.m located in the Factor_resttime folder is used to generate the Nyquist plots shown in Fig. 5. To generate plots for different batteries the variable Batt in can be modified in the MATLAB script. Setting Batt = 1 corresponds to Battery 1 (MCLI01), Batt = 2 to Battery 2 (MCLI02), Batt = 3 to Battery 3 (MCLI03), and Batt = 4 to Battery 4 (MCLI04).

4.8. Nyquist plots at different current amplitudes

The Nyquist plots for the MCLI04 cell, grouped by current amplitudes of 30 mA, 50 mA, 100 mA, 500 mA, and 1 A, are shown in subplots (a), (b), (c), (d), and (e) of Fig. 6. The distinction across SoC is shown using varying marker symbols, i.e., “o” for 80 %, “+” for 60 %, “*” for 40 %, “□” for 20 %. Similarly, the distinction across rest time is shown using varying colors i.e., green tone for 0.5 h, purple tone for 1 hour, and orange tone for 2 h.

Fig. 6.

Nyquist plot at different AC amplitudes for MCLI04 cell. Colors specify different rest times, and line markers specify SOC.

The MATLAB script demo03_CurrentAmplitude.m located in the Factor_resttime folder is used to generate the Nyquist plots shown in Fig. 6. To generate plots for different batteries, the variable Batt can be modified in the MATLAB script. Setting Batt = 1 corresponds to Battery 1 (MCLI01), Batt = 2 to Battery 2 (MCLI02), Batt = 3 to Battery 3 (MCLI03), and Batt = 4 to Battery 4 (MCLI04).

Limitations

The experiment was conducted at room temperature (air-conditioned room; average temperature was in the range of 20–28 °C). Other than this, no strict temperature enforcement was done on the battery cells.

Ethics Statement

The authors have read and follow the ethical requirements for publication in Data in Brief and confirm that the current work does not involve human subjects, animal experiments, or any data collected from social media platforms.

CRediT authorship contribution statement

Banuselvasaraswathy Balasubramanian: Investigation, Data curation, Software, Writing – original draft. Prarthana Pillai: Software, Writing – review & editing. Balakumar Balasingam: Supervision.

Data Availability

• Mendeley DataExperimental Data for Electrochemical Impedance Characterization of Li-ion Batteries Under Varying State of Charge, Current Amplitude and Rest Time (Original data).