Scalable piezoelectrochemical energy harvesting via Li-ion battery-based DEECs with laser-patterned electrodes

Summary

Piezoelectrochemical energy harvesting converts mechanical compression into electrical output through stress-coupled electrochemistry in lithium-ion systems. Here, we study lithium-ion battery-based distributed embedded energy converters (DEECs) as unit cells in DEEC-Tec metamaterials and examine how interconnection topology and electrode microarchitecture affect performance under 10 MPa loading. For commercial pouch cells, harvested energy increased from 153 to 395 μJ/cycle in series and from 101 to 204 μJ/cycle in parallel as the number of units increased from one to four, while a hybrid series-parallel network delivered the best performance (501 μJ/cycle; 16.73 nW/Ah). For laboratory-fabricated cells, laser-patterned electrodes improved harvested energy and capacity-normalized output by 84.5% despite an 18%–19% capacity loss. Electrochemical impedance spectroscopy showed 89.8% lower solid-electrolyte interphase resistance, 18.1% lower charge-transfer resistance, and 77.7% lower total resistance. These results show that topology and electrode microstructure can jointly improve scalable low-frequency piezoelectrochemical energy harvesting.

Graphical abstract

Highlights

• •Multi-DEEC arrays successfully scale piezoelectrochemical energy harvesting
• •Hybrid series-parallel wiring maximizes network power and energy output
• •Laser patterning of battery electrodes lowers total impedance by 77.7%
• •Laser-patterned cells boost capacity-normalized harvested energy by 84.5%

Energy engineering; Energy systems; Energy application

Introduction

The increasing demand for autonomous, sustainable, and scalable energy solutions has accelerated research in mechanical energy harvesting technologies capable of operating efficiently in low-frequency environments such as ocean waves, structural vibrations, and biomechanical movements.^1,2,3 While conventional piezoelectric harvesters have demonstrated effectiveness in mid- to high-frequency applications, they suffer from several limitations, including low energy density, limited flexibility, and reduced performance under sub-Hz dynamic conditions.^4,5 Consequently, these constraints make them suboptimal for distributed energy harvesting in compliant systems. In this context, the piezoelectrochemical (PEC) effect has emerged as a promising alternative that enables the direct coupling of mechanical deformation and electrochemical response, wherein applied stress induces changes in the equilibrium potential of ion-intercalation electrodes, commonly referred to as PEC materials (e.g., battery electrodes).^5,6,7,8

The PEC effect in lithium-ion systems arises from the intrinsic coupling between mechanical stress and electrochemical potential. This stress-voltage coupling, quantitatively described in the context of electrode mechanics by Cannarella and Arnold and later expanded through both theoretical modeling and experimental investigations by other researchers,^6,9,10 provides the fundamental basis for PEC energy conversion. Mechanical stress alters the lithium chemical potential, which in turn shifts the electrochemical potential of the active materials, thereby generating a voltage signal without requiring polarization or resonance.^5,6,11 Compared to conventional piezoelectric systems, PEC-based systems operate efficiently at ultra-low frequencies, offer higher energy density, and seamless integration with energy storage functionality, making them particularly attractive for soft and wearable systems as well as marine environments where oscillations are slow and variable.^5,6,9 Despite their potential, commercial lithium-ion batteries are not inherently designed for stress-to-voltage conversion.^12,13 In conventional electrodes, inherently random crystal orientations dissipate the stress-induced effect, while relatively high internal resistance, and long tortuous ion transport pathways reduce the rate of lithium redistribution under applied stress, thereby suppressing the overall PEC response.^14,15 Enhancing PEC energy harvesting performance involves optimization of various interrelated parameters, including internal resistance, electrode architecture, and ion transport properties.^9,16,17,18

Among various modification techniques available, laser microstructuring using ultrafast or femtosecond (fs) lasers has emerged as a highly controllable and scalable method for electrode surface engineering.^{19,20,21,22,23} Fs-laser ablation can precisely sculpt microgrooves, pores, or channels on electrode surfaces, thereby increasing the effective surface area, improving electrolyte wetting, facilitating ionic/electronic transport, and promoting more uniform stress distribution.^19,21,23,24 Recent studies have shown that fs-laser-patterned (LP) electrodes exhibit significantly lower charge-transfer resistance and enhanced ion diffusion kinetics, owing to improved electrolyte accessibility and reduced interfacial polarization.^19,25,26 These modifications also mitigate local stress accumulation during repeated lithiation and delithiation, leading to more uniform lithiation, reduced mechanical degradation, and improved cycling stability. Such properties are particularly beneficial for PEC energy harvesting, where efficient mechanical-electrochemical coupling is crucial.^27,28,29

Beyond electrode surface engineering, understanding how mechanical stress couples with electrochemical thermodynamics and transport is central to PEC energy harvesting. Prior studies on stress-dependent electrochemical potential in Li-ion systems have shown that external compression can shift electrode chemical potentials, and drive measurable voltage responses (V_R) even in the absence of net electrochemical cycling.^18,30 These stress-induced electrochemical responses are highly sensitive to electrode microstructure, interfacial resistance, and transport pathways, suggesting that architectural control—such as fs-laser patterning—can play a decisive role in amplifying PEC transduction efficiency. Establishing this structure-stress-electrochemistry linkage is therefore essential for translating material-level PEC effects into practical energy-harvesting devices. However, the integration of fs-laser microstructuring into PEC-enabled systems remains largely unexplored. Further investigation is needed to understand how laser-induced electrode architectures influence charge transport, local stress evolution, and voltage modulation under mechanical deformation.

While most prior studies on PEC energy harvesting have focused on isolated PEC devices or small-scale laboratory prototypes, real-world applications demand scalable architectures capable of distributed energy conversion. Addressing this need, distributed embedded energy converter (DEEC) technology has recently emerged as a transformative framework. A DEEC is a device (Li-ion battery in this case) that serves as both an energy transducer and a structural unit.^31,32 When many such DEECs are interconnected, they form DEEC-Tec metamaterials (Figure 1), which are engineered materials capable of converting mechanical energy throughout their entire volume rather than at discrete points. These metamaterials represent a paradigm shift in energy harvesting, offering mechanical adaptability and system-level resilience. This paradigm of distributed energy conversion introduces several key advantages: (1) modularity, enabling custom geometries and design flexibility; (2) fault tolerance, where local failure does not compromise system functionality; and (3) scalability, from laboratory-scale systems to full-scale marine devices.

Figure 1.

Schematic for sample PEC DEEC-Tec metamaterial

The DEEC-to-DEEC-Tec progression culminates in the concept of wave energy converters (WECs) built from DEEC-Tec metamaterials. Unlike conventional WECs, which are typically large, expensive, mechanically complex, and centralized in their energy conversion, DEEC-Tec-based systems are lightweight, flexible, and inherently distributed, enabling low-maintenance and reconfigurable deployment in diverse oceanic conditions.^31,32,33,34 Additionally, their ability to function efficiently under low-frequency, low-amplitude oscillations makes them particularly suited for the marine environment, where conventional turbines or oscillating water columns struggle to maintain consistent performance.

In this work, we investigate the PEC performance of commercial non-patterned (CNP) and laboratory-fabricated laser-patterned (LP) and non-patterned (NP) DEECs (Figure 2; see the “Preparation of Commercial DEECs and Laboratory-Fabricated DEECs” section in the STAR Methods for details) under mechanical compression, using a custom-built test setup (Figure 3; see the “Custom-built test setup for PEC testing” section in the STAR Methods for details), focusing on the PEC performance enhancement enabled by innovations in electrode patterning, and multi-DEEC configurations (Figure 4; see the “Multi DEEC configuration” section in the STAR Methods for details). First, we analyze the harvested energy and power output of commercial DEECs arranged in single to multi-DEEC configurations (Figure 5; see the “PEC performance testing configurations” section in the STAR Methods for details), which provide insights into energy scaling. Next, we explore the impact of multi-DEEC in series, parallel, and hybrid configurations, focusing on the effects of interconnection on overall performance. Additionally, we examine laboratory-fabricated modified DEECs, specifically assessing the role of laser-patterning of the electrodes, which reduces internal resistance and improve ion transport, on the PEC coupling and energy conversion efficiency. The results of this study not only demonstrate the potential of PEC-enabled DEEC-Tec metamaterials as scalable, robust, and efficient wave energy harvesters, but also offer valuable insights into the development of next-generation distributed energy harvesting systems with superior energy conversion characteristics via electrode architectural engineering.

Figure 2.

Laser patterning of Li-ion DEEC electrodes and device assembly

(A) Schematic of cathode and anode patterned with 200 μm pitch micro-grooves (30 mm × 65 mm active area); corresponding SEM top-view and cross-section images confirm groove formation (scale bars: 200 μm for top-view images and 50 μm for cross-section images).

(B) Layer stack of the laboratory-fabricated pouch cell and photograph of the NP/LP DEEC.

Figure 3.

Schematic of the custom-built experimental setup for PEC tests

Figure 4.

Schematic of the stacked multi-working-DEEC configuration with compression platens

Figure 5.

DEECs configurations

(A) Single and (B) multiple.

Results and discussion

CNP DEECs

The PEC performance of commercial non-patterned DEECs (CNP DEECs) was systematically examined to evaluate their intrinsic energy-harvesting behavior and scalability under mechanical compression. First, we investigate how electrical configuration, spanning single-DEEC to multi-DEEC series, parallel, and hybrid arrangements, affects voltage generation, harvested energy, and specific power output (see the “Quantification and statistical analysis” section in the STAR Methods for details). The results reveal how interconnected DEEC units collectively influence the efficiency and output characteristics of DEEC-Tec metamaterials, providing insights into their suitability for scalable and distributed energy conversion systems.

Series-parallel scaling of CNP DEECs: 1–4 units

The voltage response V_R across a 1 Ω resistor during PEC cycling at 100 μHz under 10 MPa compressive stress (Figure 6; see the “Multi DEEC configuration” section in the STAR Methods for details) was measured to assess the energy harvesting performance of one to four CNP DEECs connected in both parallel and series configurations under 10 MPa compressive stress (Figure 7). Voltage profiles for parallel (Figure 7A) and series (Figure 7B) configurations display distinct behaviors reflective of their electrical arrangements.

Figure 6.

Representative compressive stress-time profiles used during PEC testing under 10 MPa

(A) Laboratory-fabricated DEECs at 10 μHz and (B) commercial non-patterned (CNP) DEECs at 100 μHz.

Figure 7.

Voltage response over time for 1–4 CNP DEECs during PEC cycling

(A and B) Under 10 MPa compressive stress at 100 μHz cycling frequency in (A) parallel (B) series configurations.

In parallel configurations (Figures 7A and 8A), the voltage amplitude remains relatively stable with an increase from approximately 0.17 mV for a single DEEC (W1,R1-P) to 0.21 mV for four DEECs (W4P,R4P). While an ideal parallel network would maintain identical voltage regardless of unit count, the observed nonlinear increase reflects non-idealities such as contact resistance, interconnect losses, and cell-to-cell variations. Notably, the voltage decay in the parallel configuration is slower, indicating sustained charge delivery over a longer duration, which is advantageous for ultra-low-frequency and extended-time energy-harvesting applications.

Figure 8.

Voltage response under compression during the first PEC cycle for 1 to 4 CNP DEECs

(A and B) In (A) parallel (B) series configurations; shaded area represents harvested charge (since R = 1 Ω, according to Ohm’s law, the voltage V_R (y axis) is equal to the current I_R).

In contrast, series configurations (Figures 7B and 8B) exhibit a near-linear increase in peak voltage, rising from 0.25 mV for single DEEC (W1,R1-S) to 0.57 mV for four DEECs (W4S,R4S). The peak voltage does not reach the ideal 4-fold increase due to cumulative internal resistances, voltage drops across interconnections, and imperfect synchronization of individual DEEC responses under mechanical loading. These losses become increasingly significant as more units are added in series.

The harvested energy per cycle, calculated from the integrated voltage response, demonstrates a clear scaling trend for both configurations (Figure 9). In the parallel arrangement (Figure 9A), energy increases from approximately 101 μJ/cycle for a single DEEC to 204 μJ/cycle for four DEECs, effectively doubling the energy harvested. The series configuration (Figure 9B) shows an even more pronounced enhancement, with energy rising from 153 μJ/cycle for single DEEC to 395 μJ/cycle for four DEECs—an increase of approximately 158%. This confirms the scalable nature of the DEEC-Tec metamaterial, where multiple DEECs operating in concert significantly enhance the total harvested energy.

Figure 9.

Comparison of harvested energy and specific power output

(A and B) Under 10 MPa compressive stress at 100 μHz cycling frequency for 1–4 CNP DEECs in (A) parallel (B) series configurations.

The specific power output exhibits a general decreasing trend as the number of DEECs increases in both parallel and series configurations, primarily due to increased internal resistance and associated system losses, including interconnection contact resistance and parasitic losses arising from cell-to-cell variations as additional units are integrated into the circuit. In the parallel configuration, the specific power output decreases from 13.46 nW/Ah for single DEEC to 6.81 nW/Ah for four DEECs, a reduction of approximately 49.4% (Figure 9A). Similarly, in the series configuration (Figure 9B), it decreases by about 35.4%, from 20.42 nW/Ah (1 DEEC) to 13.18 nW/Ah (4 DEECs). Collectively, these results highlight a fundamental trade-off: series connections favor higher instantaneous voltage and energy per cycle, whereas parallel connections provide more sustained voltage output, making them particularly attractive for low-frequency or long-duration PEC energy-harvesting scenarios.

Series-parallel-hybrid performance of CNP DEEC-Tec: 4 units

The performance of four CNP DEECs under 10 MPa compressive stress was evaluated in three electrical configurations: series (W4S,R4S-V2), parallel (W4P,R4P-V2), and hybrid series-parallel (W2S2P,R2S2P). To isolate intrinsic PEC responses and eliminate cycling effects caused by PEC-induced aging and degradation, a new, unused set of DEECs was employed for all configurations. To avoid confusion with the DEECs used in the previous section, the DEECs in series and parallel configurations are denoted as version 2 (V2). The V2 DEECs were obtained from the same production batch and have the same nominal capacity as the previous set, ensuring consistent electrochemical characteristics across all tests.

Voltage responses (V_R) over four consecutive PEC cycles are presented in Figure 10, demonstrating the repeatability of the PEC performance across multiple loading cycles. Figure 11 shows the detailed voltage profile of the first compression cycle, which was used to calculate the harvested energy and specific power output. Figure 12 compares the energy and power results derived from these analyses.

Figure 10.

Voltage response over time for 4 CNP DEECs during PEC cycling under 10 MPa compressive stress at 100 μHz cycling frequency in series, parallel, and hybrid configurations

Figure 11.

Voltage response under compression during the first PEC cycle for 4 CNP DEECs; shaded area represents harvested charge (since R = 1 Ω, according to Ohm’s law, the voltage V_R (y axis) is equal to the current I_R)

Figure 12.

Comparison of harvested energy and specific power output under 10 MPa compressive stress at 100 μHz cycling frequency for 4 CNP DEECs in series, parallel, and hybrid configurations

Figure 10 shows the voltage response V_R over four consecutive PEC cycles for the 4 CNP DEECs under 10 MPa compressive stress. The 4 DEECs in series configuration exhibits peak voltages of approximately 0.519 mV, while the hybrid and parallel configurations show peak voltages of 0.398 and 0.301 mV, respectively. The hybrid configuration’s peak voltage is 23.3% lower than the series but 32.2% higher than the parallel configuration, demonstrating an intermediate voltage level and stable cyclic behavior.

Harvested energy per cycle using 4 DEECs in hybrid configuration exhibits a gain over both series and parallel configurations: the hybrid configuration delivers 501 μJ/cycle, which is 3.5% higher than the series configuration and 23.5% higher than the parallel configuration (Figure 12). This reflects the hybrid configuration’s ability to combine series-induced voltage enhancement with parallel-enabled sustained energy delivery.

Additionally, hybrid configuration resulted in superior power output performance over both series and parallel configurations. A maximum specific power output of 16.73 nW/Ah is observed for the hybrid system, exceeding the series and parallel configurations by 3.8% and 23.7%, respectively (Figure 12). This improvement highlights the hybrid arrangement’s effectiveness in balancing power density and energy harvesting efficiency, demonstrating a compromise between voltage and current scaling that makes it a promising solution for scalable energy harvesting in DEEC-Tec metamaterials.

A preliminary system-level volumetric estimate for the stacked DEEC-Tec architecture (a), including the inactive volume associated with intermediate compression platens, is provided in the supplementary information (Equations S9–S13); this analysis further emphasizes that the present multi-cell demonstration should be regarded as a proof-of-concept rather than an optimized practical composite.

Laboratory-fabricated DEECs

To elucidate the effect of electrode architecture on PEC performance, two laboratory-fabricated DEECs were examined: a non-patterned (NP) DEEC and a LP DEEC modified using an fs-laser to create surface microchannels on both electrodes (i.e., Lithium cobalt oxide cathode and graphite anode). Both DEECs shared identical materials, geometry, and fabrication procedures, differing only in electrode surface morphology (Figure 2; see the “Preparation of Commercial DEECs and Laboratory-Fabricated DEECs” section in the STAR Methods for details). The following analysis compares their electrochemical impedance, capacity loss, and energy-harvesting behavior to evaluate the influence of laser patterning on interfacial kinetics and PEC performance.

Electrochemical impedance spectroscopy analysis on laboratory-fabricated DEECs

In this study, the electrochemical impedance of NP DEEC at 20% state of charge (SOC) in its pristine state is compared with that of LP DEEC at 20% SOCs to ensure a fair comparison at the same SOC. Figure 13 presents the Nyquist plots for these samples. The fitted resistive values with their relative standard error are given in Table 1.

Figure 13.

Nyquist plots of NP DEEC and LP DEEC at 20% SOC with corresponding fitted curves, showing the reduced impedance response of the laser-patterned DEEC

Table 1.

The fitted resistive values with their relative standard error for LP and NP DEECs at 20% SOC

Resistive parameters	LP DEEC @20% SOC		NP DEEC @20% SOC
	PEC-0	Rel. std. error (%)	PEC-0	Rel. std. error (%)
Rs (Ω)	0.411	5.30	0.242	0.930
RSEI (Ω)	0.992	2.61	9.737	0.545
Rct (Ω)	1.123	1.23	1.371	5.195
Total (Ω)	2.526	–	11.35	–

Relative to the NP DEEC at 20% SOC, the LP DEEC at the same SOC exhibits substantially lower interfacial resistances. The solid–electrolyte interphase resistance (R_SEI) decreases from 9.737 to 0.992 Ω, corresponding to a reduction of approximately 89.8%, and the charge-transfer resistance (R_ct) decreases from 1.371 to 1.123 Ω, corresponding to a reduction of approximately 18.1%, demonstrating a strong suppression of interfacial resistance due to laser patterning.

The relative standard errors for R_SEI and R_ct are low for both DEECs, confirming excellent fitting robustness. The solution/ohmic resistance (R_s) shows only modest variation, increasing from 0.242 Ω for the NP DEEC to 0.411 Ω for the LP DEEC, corresponding to an increase of approximately 69.9%, which likely arises from contact or electrolyte-path differences rather than interfacial chemistry. When the resistive components are aggregated, the total impedance (R_s + R_SEI + R_ct) decreases dramatically from 11.350 Ω for the NP DEEC to 2.526 Ω for the LP DEEC, corresponding to overall reduction of approximately 77.7%. These results indicate that the observed impedance reduction is governed primarily by laser patterning, thereby reducing energy losses during charge transfer and enabling more efficient PEC energy harvesting.

Capacity loss after laser patterning: NP vs. LP DEECs

Figure 14 compares charge-discharge capacities of the NP DEEC and LP DEEC to quantify the capacity loss associated with laser ablation of active material. The NP DEEC delivers 32.80 mAh (charge) and 32.37 mAh (discharge) in the first cycle, whereas the LP DEEC provides 26.45 mAh (charge) and 26.39 mAh (discharge), respectively. Thus, patterning reduces capacity by 18.5% on discharge (LP retains 81.5% of NP) and 19.4% on charge (LP retains 80.6%).

Figure 14.

Comparison of discharge and charge capacities between NP DEEC and LP DEEC

This reduction is consistent with the intentional removal of active material to create laser-grooved channels, which lowers areal mass loading and, consequently, the delivered capacity. Notably, the charge and discharge capacity values for the LP DEEC remain closely matched (i.e., high coulombic efficiency, 99.77%), indicating that patterning primarily trades capacity for improved transport/kinetics (as corroborated by the large decreases in R_SEI and R_ct in the electrochemical impedance spectroscopy (EIS) results), which is advantageous for PEC energy-harvesting performance even though the storage capacity per device declines.

Effect of laser patterning on energy harvesting performance

The PEC performance of the laboratory-fabricated DEECs was assessed for the LP DEEC and the NP DEEC under 10 MPa at 10 μHz using the single-DEEC configuration (Figures 5A and 6A; see the “PEC performance testing configurations” section in the STAR Methods for details). V_R over four consecutive cycles are shown in Figure 15, with the first compression half-cycle highlighted in Figure 16. Figure 17 summarizes the harvested energy and specific power output obtained at 10 μHz for NP and LP DEECs using single-DEEC configuration. For completeness, the comparison involving the CNP DEEC is presented in supplementary information (Figure S3), where the comparison basis is described in greater detail.

Figure 15.

Voltage response over time for LP DEEC and NP DEEC during PEC cycling under 10 MPa compressive stress at 10 μHz cycling frequency using the single DEEC configuration

Figure 16.

Voltage response under compression during the first PEC cycle of LP DEEC and NP DEEC; shaded area represents charge (since R = 1 Ω, according to Ohm’s law, the voltage V_R (y axis) is equal to the current I_R)

Figure 17.

Comparison of harvested energy and specific power output under 10 MPa compressive stress at 10 μHz cycling frequency for LP DEEC and NP DEEC

Across the four cycles, the LP DEEC exhibits a slightly higher peak voltage than the NP DEEC with V_p = 0.0388 mV for LP and 0.0354 mV for NP, corresponding to a 9.6% increase. A more pronounced difference is observed in the first-cycle compression response, where the integrated area under V_R(t) during compression for the first cycle increases from 0.00763 C for NP to 0.01284 C for LP, representing a 68.4% increase (Figure 16). This larger integrated response indicates enhanced charge harvesting in the LP device.

A similar trend is observed in the harvested energy output. The LP DEEC harvests 0.5 μJ/cycle, an +84.5% increase over NP DEEC (0.27 μJ/cycle) (Figure 17). When normalized by capacity and compression time, the LP DEEC also exhibits higher specific power output than the NP DEEC, increasing from 16.47 nW/Ah for NP to 37.67 nW/Ah for LP, which corresponds to a 128.7% enhancement (Figure 17). These results demonstrate that laser patterning enhances PEC performance by improving the stress-induced electrochemical response of the laboratory-fabricated DEECs.

This enhancement is attributed to the microchannel features introduced by fs-laser patterning, which improve electrolyte wetting and facilitate Li⁺ transport across the electrode interface. Together with the reduced R_SEI and R_ct identified from impedance analysis, these effects promote a stronger stress-induced electrochemical response and thereby increase the harvested energy. The results therefore establish laser patterning as an effective electrode-design strategy for enhancing PEC energy harvesting in laboratory-fabricated Li-ion battery-based DEECs.

Taken together, the present results establish a coherent picture of PEC energy harvesting in Li-ion battery-based DEECs under compressive loading. We investigated PEC performance of Li-ion battery-based DEECs, including CNP and laboratory fabricated NP and fs-LP DEECs under 10 MPa compressive loading, spanning scalability from single DEECs to four-unit arrays configured in series, parallel, and hybrid topologies. The PEC performance for both DEEC variants was quantified using per-cycle harvested energy and capacity-normalized specific power at 100 and 10 μHz, respectively. In addition, EIS was conducted on LP and NP DEECs to elucidate the effects of electrode microstructuring on interfacial charge-transfer kinetics and ion-transport-related resistive processes.

The scalability study shows that under 10 MPa compression, harvested energy scales with the number of CNP DEECs. For four CNP DEECs, the hybrid series-parallel configuration achieved 501 μJ/cycle and 16.73 nW/Ah, outperforming series (484 μJ; −3.5%) and parallel (406 μJ; −23.5%), confirming the hybrid network as the best compromise between voltage and current scaling for DEEC-Tec metamaterials.

Laser patterning improves per-capacity performance. Despite an ∼18%–19% decrease in charge/discharge capacity (LP vs. NP), LP electrodes increased harvested energy by 84.5% at 10 μHz. The capacity-normalized specific power also increased from 16.47 to 37.67 nW/Ah, representing a 128.7% enhancement. This confirms that fs-laser-induced microstructuring is an effective strategy for improving the PEC response of DEECs.

The electrochemical origins of improvement are supported by EIS fitting using an LCO/graphite equivalent-circuit model showed that, at matched 20% SOC, laser patterning reduced R_SEI by ∼89.8% and R_ct by ∼18.1% relative to the NP DEEC, while lowering total impedance (R_s+R_SEI+R_ct) by ∼77.7% and enabling larger charge transfer during compression.

Beyond optimizing groove geometry and stacking mechanics, combining fs-laser electrode patterning with materials exhibiting intrinsically high electrochemo-mechanical coupling (PEC coupling) offers a clear pathway to further amplify PEC performance. Future work includes co-designing electrode chemistry, LP micro-architecture, and compliant stack mechanics, while managing parasitic degradation processes such as SEI growth, gas evolution, and irreversible mechanical damage through binders and surface coating strategies. In parallel, series and parallel cell configurations can be strategically tailored to favor either high power density or longer-duration energy harvesting, depending on application needs. As schematically illustrated in Figure 18, in practical DEEC-Tec systems, especially for wave energy harvesting, externally applied multiaxial deformation may be transmitted to the embedded DEEC network as localized uniaxial compressive and transverse bending loads, which can then drive electrochemical output. While the present study focuses on controlled uniaxial compression as a first-order experimental framework, future studies should examine more complex load-transfer behavior in densely packed DEEC arrays, including the roles of interconnects, separators, and structural compliance, potentially through finite element analysis. Together, these approaches position LP, high-coupling DEEC-Tec metamaterials as a compelling platform for next-generation low-frequency mechanical energy harvesting.

Figure 18.

Schematic illustration of a practical DEEC-Tec concept in which multiaxial wave-induced deformation is translated into localized uniaxial compressive and transverse bending loads on flexible DEEC arrays, generating electrochemical output

Limitations of the study

This study demonstrates proof-of-concept PEC energy harvesting in DEEC-Tec metamaterials, but several limitations should be noted. Mechanical testing was restricted to controlled uniaxial compression at fixed loading conditions, whereas practical wave- or structure-based environments may involve more complex, multiaxial, and time-varying stresses. The stacked multi-DEEC architecture is not yet volume-optimized, and inactive spacer/platen volume reduces practical energy density; for wave-energy deployment, encapsulation against seawater exposure would likely add further inactive volume and should be considered in future system-level designs. Laser patterning improved harvested energy and reduced impedance, but it also removed active material and led to a measurable capacity loss. In addition, the processing step introduces extra fabrication cost, so techno-economic analysis will be needed to assess scalability and implementation. Finally, the study focused on LCO/graphite pouch-cell chemistries at selected states of charge, so the findings may not directly translate to other chemistries, formats, or long-term durability conditions.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, SeungYeon Kang (seung_yeon.kang@uconn.edu).

Materials availability

This study did not generate new materials.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This work was supported by the US Department of Energy’s Water Power Technologies Office (WPTO) through the Innovating Distributed Embedded Energy Prize (InDEEP), a three-phase American-Made competition administered by National Renewable Energy Laboratory (NREL). The authors also acknowledge the University of Connecticut for providing laboratory facilities and technical support. The views expressed in this article are those of the authors and do not necessarily reflect those of DOE, WPTO, NREL, or the US Government.

Author contributions

Conceptualization, M.A.A. and S.K.; methodology, M.A.A., N.K.K., and S.K.; investigation, M.A.A. and A.K.; writing – original draft, M.A.A.; writing – review and editing, M.A.A., A.K., N.K.K., and S.K.; resources, supervision, and funding acquisition, S.K.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Commercial Li-ion pouch cell (CNP DEEC)	PowerStream	PGEB064050
LCO cathode laminate	NEI Corporation	ANOMYTE® BE-20E
Graphite anode laminate	NEI Corporation	NANOMYTE® BE-200E
Polypropylene separator	Xiamen Tob New Energy Technology Co., Ltd.	T-PP-20∗76
1 M LiPF6 in EC/DMC/DEC electrolyte		T-EY-LiPF6-1M-1:1:1
Aluminum-laminated pouch film		T-ALF -DNP-113
Aluminum tabs		T-Al-0.1∗4∗61∗5∗2
Nickel tabs		T-Ni-0.1∗4∗61∗5∗2
Other
Femtosecond laser system	Spectra Physics- Solstice ACE	ASCEND-40T-SOL, SN: 33347
SEM	FEI	FP 2032/14, Quanta FEG 250 SN: QFE252/D9240
NEWARE battery cycler	NEWARE	MHW-200-110V, SN: M2301-366390
Gamry potentiostat	Gamry Instruments	Interface 5000E, SN: 20584
PAR potentiostat	Princeton Applied Research	VersaSTAT 3, SN: 21152013
Custom compression setup	This paper	N/A
Software and algorithms
ZSimpWin	AMETEK	Version 3.60, SN: 22167792

Experimental model and study participant details

Omitted as our study does not involve biological models.

Method details

Preparation of commercial DEECs and laboratory-fabricated DEECs

In the initial stage, PEC cyclic testing was conducted on commercial 1500-mAh DEECs (PGEB064050) comprising non-patterned LCO/graphite electrodes with a polypropylene (PP) separator (hereafter CNP DEECs). These DEECs were obtained from PowerStream and used as received, except that the strapping tape on edges was removed to allow more uniform stress distribution across the cell during testing. These DEECs were tested to evaluate their PEC performance under mechanical stress and provide a baseline reference against modified, laboratory-fabricated DEECs’ performance in the subsequent phase. PEC tests were performed in single-DEEC and multi-DEEC configurations under a compressive stress of 10 MPa at 100 μHz.

The second stage of the study involved the laboratory fabrication of two DEEC variants using commercial electrode laminates: laser-patterned (LP) DEEC and non-patterned (NP) DEEC. This approach enabled a controlled investigation of the role of electrode microstructuring on PEC performance. Single-side coated LCO cathodes (2.0 mAh/cm², ∼80 μm thick) and graphite anodes (2.4 mAh/cm², ∼85 μm thick), sourced from NEI Corporation, were cut to 65 mm × 30 mm using a precision die and then assigned to either the LP or NP condition. For LP-DEEC, both anode and cathode electrodes were patterned with parallel microgrooves at a 200-μm pitch via fs-laser (Spectra Physics- Solstice ACE) ablation guided by G-code, as illustrated in Figure 2. The laser system operated at a repetition rate of 10 kHz, with an average power of 1 W, and the beam was focused onto the electrode surface through a 10× objective lens. Patterning was performed at a scan rate of 3000 μm/s, producing well-defined microgrooves. Top-view and cross-sectional morphologies of the patterned electrodes were characterized using an FEI Quanta FEG 250 scanning electron microscope (SEM) operated at an accelerating voltage of 20 kV. SEM confirmed groove widths of approximately 50–55 μm and depths of 35–40 μm. Cross-sectional SEM images further confirmed that, under the selected laser processing conditions, the ablation was confined to the active material layer and did not penetrate the underlying current collector (Figure 2A). The use of fs-laser patterning is also expected to minimize thermal damage because the ultrashort pulse duration limits heat diffusion and reduces the heat-affected zone.^19,21,35 This microstructuring increases electrode surface area that is anticipated to enhance ion transport that in turn improves PEC coupling. For the NP DEECs, the electrodes were used without any laser modification, serving as a baseline for comparison.

Both DEEC types (i.e., LIBs with LP and NP electrodes) were assembled in pouch-cell configuration with a 25-μm polypropylene separator (Xiamen Tob New Energy Technology Co., Ltd.) cut to 70 mm × 35 mm. Aluminum and nickel tabs (Xiamen Tob New Energy Technology Co., Ltd.) were spot-welded to the LCO cathode and graphite anode current collectors, respectively. The cell stack consisted of one cathode, one anode, and the separator, secured with tape. Following electrolyte (Xiamen Tob New Energy Technology Co., Ltd.) injection (800 μL of 1 M LiPF₆ in EC:DMC:DEC = 1:1:1 v/v/v) in an argon atmosphere, cells were vacuum-sealed in aluminum-laminated film (Xiamen Tob New Energy Technology Co., Ltd.) (Figure 2B). To ensure structural integrity and electrode wetting, the sealed pouch cells were subjected to stacking pressure using C-clamps and sheet metal and left in this state overnight. Laser patterning parameters, representative SEM images of the LP-electrodes, and pouch cell assembly procedures are detailed in Figure 2. Individual DEEC underwent formation cycling at 0.1C (defined with respect to the nominal cell capacity) within a 2.75–4.2 V window prior to PEC cyclic tests. PEC tests were performed in single-DEEC configuration under a compressive stress of 10 MPa at a cycling frequency of 10 μHz.

Sample conditioning

Prior to PEC performance testing, each cell was conditioned to a state of charge (SOC) corresponding to the maximum stress sensitivity. We analyzed both the coupling factor (k = dε/dQ) and the differential capacity (dQ/dV) for the commercial non-patterned (CNP) DEECs. Given the low capacity of the laboratory-fabricated (NP and LP) DEECs, only dQ/dV analysis was performed to identify SOC regions associated with pronounced electrochemical transitions, which are expected to be favorable for PEC harvesting due to increased charge-transfer activity at relatively small voltage variations. As similar SOC-optimization and coupling-factor analyses have already been reported in detail in our previous publication,⁵ these plots are not repeated here to avoid redundancy. The SOC determined from these analyses was used in subsequent PEC cycling tests. Based on these analyses, the selected SOCs were 20% (∼3.63 V) for CNP-DEECs, 11% (∼3.73 V) for LP-DEECs, and 20% (∼3.77 V) for NP-DEECs. Subsequently, the DEECs were charged to their respective SOCs, and electrochemical impedance spectroscopy (EIS) measurements were performed at these SOCs to assess kinetic parameters prior to PEC cycling. All PEC performance tests were carried out using a custom-built test setup.

Electrochemical characterization

Galvanostatic charge–discharge tests were performed on a NEWARE battery cycler at a rate of 0.1C. The charging and discharging rate were 0.1C. The EIS of two-electrode DEEC was measured using the Princeton Applied Research (PAR) potentiostat over the frequency range 1 mHz to 100 kHz with amplitude of 5 mV. All electrochemical measurements were performed at room temperature (25 ± 1°C).

The impedance spectra were fitted in ZSimpWin software. To interpret the impedance data, the spectra were fitted with an equivalent circuit model that is typically used for LCO/graphite-type full cells: L–R_s–(R_SEI ∥ CPE1)–(R_ct∥ CPE2)–W, where L is the inductance, R_s the solution/ohmic resistance, R_SEI the solid–electrolyte interphase resistance, R_ct the charge-transfer resistance, CPEs are constant phase elements, and W is the Warburg diffusion element.^18,36,37

Custom-built test setup for PEC testing

The compression testing of DEECs were performed using a custom-built test setup (Figure 3) designed to apply controlled mechanical stress while simultaneously measuring electrochemical performance. The setup consists of an electric hydraulic pump, hydraulic jack, load cells, and displacement sensors, integrated with a potentiostat (Gamry Interface 5000E) for real-time voltage and current measurements. Upon activation of the hydraulic pump, the piston of the hydraulic jack moves upward, applying compressive stress to the DEEC against the adjustable plate. The applied load is measured by the load cell, and displacement is monitored using a displacement sensor, which tracks the change in thickness of the DEEC during charging, discharging, and PEC cycling. The data collected from the test is stored and analyzed in real-time via the local data acquisition (DAQ) system. All PEC performance tests were performed at room temperature (25 ± 1°C).

PEC performance testing configurations

Two primary configurations are utilized in this study to evaluate the PEC performance: single-DEEC and multi-DEEC. These configurations are designed to analyze the performance of a single working DEEC unit (single-DEEC) and multiple working DEEC units operating together (multi-DEEC). Detailed descriptions of each configuration are provided below.

Single DEEC configuration

A working DEEC, subjected to PEC cycling (i.e., periodic compression and decompression at a defined frequency), was electrically connected in parallel with a reference DEEC of identical SOC, which served as an electrochemical load and was not exposed to any mechanical deformation. A shunt resistor (1 Ω) was placed on either the positive or negative terminal to measure the voltage drop across the resistor. A potentiostat is used to record the voltage across the resistor. This configuration is denoted as W1,R1, where “W” represents the working DEEC, “R” represents the reference DEEC, and “1” indicates the number of DEECs used on both the working and reference sides in this configuration (Figure 5A). To distinguish the series and parallel testing batches, we used the suffix “-S” for series configurations (W1,R1-S) and “-P” for parallel configurations (W1,R1-P). This configuration is used for both commercial and laboratory-fabricated DEECs.

Multi DEEC configuration

In this configuration, multiple working DEECs (2, 3, or 4) are interconnected in series, parallel, or a hybrid arrangement (series-parallel) with an equal number of reference DEECs connected in a similar manner. A shunt resistor (1 Ω) was placed on either the positive or negative terminals to measure the voltage drop or current flow through the circuit. The configurations are denoted as W4S, R4S for 4 DEECs in series, W4P, R4P for 4 DEECs in parallel, and W2S2P, R2S2P for a hybrid configuration with 2 DEECs in series and 2 in parallel (Figure 5B). These configurations were applied to the commercial DEECs during compression studies to evaluate the impact of interconnection schemes on harvested energy. The series, parallel, and hybrid arrangements form DEEC-Tec metamaterials, enabling the exploration of scalable energy conversion and the study of energy harvesting from interconnected DEECs. During compression testing of multiple DEECs, compression platens are inserted between the working DEECs to ensure uniform load distribution, with the DEECs stacked vertically, as illustrated in Figure 4.

Figure 6 presents the representative mechanical loading profiles used during PEC measurements for laboratory-fabricated DEECs and commercial DEECs. In both cases, cyclic compressive stress was applied up to 10 MPa, with the laboratory-fabricated DEECs tested at 10 μHz and the commercial DEECs at 100 μHz. In each test, the applied stress reached the target value within 1 s, corresponding to a ramp rate of 10 MPa/s. A compressive stress of 10 MPa was selected as the standard testing condition to provide a fixed and consistent basis for evaluating PEC performance under cyclic compression. This loading condition is also consistent with our previous PEC testing methodology.¹⁸

Quantification and statistical analysis

Quantifying PEC output: Harvested energy and specific power

In this study, the performance of DEECs under mechanical compression was quantitatively assessed using two key parameters: harvested energy and specific power output. The harvested energy reflects the total energy generated per compression cycle (Equation 1), while the specific power output provides a normalized measure of power relative to the DEEC’s capacity (Equation 2), enabling performance comparison across devices of different sizes.

These parameters were computed as follows:

Harvested energy, $E_t=\frac{V_p}{R}{\int }_0^t{V}_R\left(t\right)dt$

$$or,E_t=\frac{V_p}{R}\times Areaunderthe{V}_{R}vs.tplot$$ (Equation 1)

$$\text{Specific}\text{power}\text{output},P_s=\frac{E_t}{C\times t}$$ (Equation 2)

Where.

• •E_t is the total harvested energy per cycle (J),
• •V_R(t)is the instantaneous voltage across the resistor (V),
• •V_p is the peak voltage (maximum value of V_R) (V),
• •R is the resistance of the external resistor (Ω),
• •t is the duration of compression (s),
• •Cis the total capacity of the working DEEC(s) (Ah).

Physically, the harvested energy represents the net electrical energy released during a PEC compression event due to the stress-induced electrochemical potential difference between the working and reference DEECs. This potential difference drives current through the external resistor, and the electrical output is determined by integrating the measured circuit response over the compression interval. For practical comparison among DEEC configurations, a simplified expression is used under the assumptions of approximately constant internal resistance during the analysis window and a peak-voltage-based representation of the working-cell response. Further details, including the arrangement of the DEECs for energy harvesting (Figure S1), the different states of the PEC cycle (Figure S2) and the derivation of the calculations for harvested energy and specific power output are provided in Supplementary Information as Equations S6 and S7.

Published: September 18, 2026