View PDF Version
 
DOI: 10.1039/D5CC03912F (Communication) Chem. Commun., 2025, Advance Article

Magnetic field-enhanced transformation of biochemical energy into motion of enzyme-modified graphene monolayers

Ruchao Gaoab, Sabrina Bichonc, Sébastien Gounelc, Gerardo Salinasb, Lin Zhang*a, Nicolas Mano*c and Alexander Kuhn*ab
aEngineering Research Center for Nanomaterials, Henan University, 475000, Kaifeng, China. E-mail: lin.zhang@henu.edu.cn
bUniversity of Bordeaux, CNRS, Bordeaux INP, ISM UMR 5255, 33600, Pessac, France. E-mail: kuhn@enscbp.fr
cCentre de Recherche Paul Pascal, University of Bordeaux, CNRS, UMR 5031, Pessac, France. E-mail: nicolas.mano@crpp.cnrs.fr

Received 11th July 2025 , Accepted 14th August 2025

First published on 16th August 2025

Abstract

This contribution studies enzymatically driven graphene monolayer swimmers for the direct transformation of biochemical energy into motion. The swimmers are elaborated via bipolar electrodeposition of gold and asymmetric immobilization of glucose oxidase and bilirubin oxidase on small pieces of a graphene monolayer. Most importantly, we demonstrate that the speed of the free-standing hybrid objects can be enhanced by an external magnetic field, allowing also controlled rotation, despite the absence of ferromagnetic construction elements. These results illustrate the potential of 2D materials for designing, in synergy with magnetic fields, original enzyme-based energy conversion devices.

With the rapid consumption of global energy and growing environmental problems, the development of sustainable energy conversion technologies has become one of the central focuses of scientific research.1 Energy conversion strategies based on biological systems have attracted extensive attention due to their environmental compatibility and renewability.2 For instance, some microbes or enzymes, capable of converting biomass into commodity chemicals or fuels, demonstrate significant potential for generating green and clean energy.3,4 Enzyme-powered micro/nano swimmers,5 which convert chemical energy directly into mechanical energy, have become a new direction in the field of energy conversion due to the use of bio-sourced fuel. Such enzymatically powered systems typically undergo redox or hydrolysis reactions, catalysed by single or multiple enzymes (such as bilirubin oxidase (BOD),6 glucose oxidase (Gox),7 urease8), generating local ion concentration gradients and self-electrophoresis to drive autonomous movement. Compared with traditional chemical and bubble propulsion systems, enzymatic swimmers offer distinct advantages. Their fuels (such as glucose, oxygen or urea) are widely available and of natural origin, while the reaction products are environmentally friendly.9 Those advantages make them compatible with the requirements of sustainable energy conversion devices. Thus, the development and optimisation of such devices is an ongoing challenge in current research, whether it concerns their design, the improvement of their speed or a better directional control. In addition to the often reported micro- and nanoswimmers,10 macro-scale enzymatic swimmers have also been studied, using e.g. conductive substrates like carbon fibers,5a,11 or PPy films,12 modified with enzymes, leading to interesting self-propulsion behaviour.

In this general context, two-dimensional (2D) materials offer significant potential for developing enzymatic swimmer systems due to their excellent mechanical properties, high intrinsic conductivity, and atomic thickness.13 To achieve directed motion, the symmetry of such materials needs to be broken and one straightforward approach, among many others,14 is based on bipolar electrochemistry (BPE).15 Recently, this versatile technique has been employed to deposit with high spatial control various metals and polymer composites at specific sites of monolayer graphene,16 and it therefore can be advantageously used for designing novel graphene-based bioelectrochemical swimmers.

In this study, we developed an enzymatic graphene swimmer by generating first ultrathin gold layers at both extremities of a piece of monolayer graphene via bipolar electrodeposition. Then, glucose oxidase and bilirubin oxidase were immobilized on these oppositely located Au-modified areas (Fig. 1A). These bi-enzymatic hybrid 2D objects exhibit intriguing dynamic features in the presence of glucose and oxygen. Most importantly, we show for the first time that by applying an external magnetic field, their behavior can be precisely controlled, allowing well-defined rotational motion, combined with an improved energy conversion efficiency.


image file: d5cc03912f-f1.tif
Fig. 1 (A) Illustration of the mediated enzymatic graphene swimmer. Trajectory of an enzymatic graphene swimmer (BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 1[thin space (1/6-em)]:[thin space (1/6-em)]1) moving at the air/water interface (B) in the absence of an external magnetic field, (C) in the presence of a magnetic field (200 mT) with the north or (D) south pole facing up. The solution consists of an O2 saturated 0.1 M PBS, 20 mM glucose solution at pH 7 and 37 °C.

The graphene monolayer sheets are prepared via a wet-transfer method.17 Au layers are deposited at both extremities of the graphene sheets by bipolar electrodeposition. As shown in Scheme S1, the graphene sheet is first positioned at the air/water interface of an Au salt solution between two feeder electrodes. When a sufficiently high electric field is applied to the electrolyte, an Au layer is deposited at the cathodically polarized extremity of the graphene sheet. Then the sheet is rotated by 180° and the same electric field is applied to deposit Au at the opposite extremity. Fig. S1A displays the optical image of a graphene sheet after Au deposition. A distinct color contrast is observed at the two ends of the monolayer, attributed to the Au deposition, while the middle part of the sheet remains transparent, indicating no metal deposition. Fig. S1B and C show the corresponding EDS mapping of the two Au-modified ends, confirming the successful site-specific metal deposition. The bipolar electrodeposition of Au layers on specific areas of graphene provides a better contact between the graphene surface and the enzyme-containing redox polymer. In addition, the extension of the modified area is controllable by adjusting the applied electric field, a feature which can only be achieved by BPE.15b,c

The enzymatic graphene swimmer was prepared by immobilizing two enzymes (Gox and BOD) at the opposite extremities of an Au-modified graphene sheet using Os-based redox polymer hydrogels.18 The enzyme-hydrogel mixtures were site-specifically loaded onto the Au-modified areas via drop-casting (see SI). In the presence of both, glucose and oxygen, the spatially separated enzymatic reactions establish a short-circuited biofuel cell. Gox catalyzes the oxidation of glucose at the anodic side, generating electrons (e−, red arrow in Fig. 1A) which move through the graphene monolayer to be used for the oxygen reduction reaction mediated by BOD at the cathodic side.

Simultaneously, the glucose oxidation at the anodic side continuously generates protons, while the oxygen reduction at the cathode consumes protons. Consequently, protons are asymmetrically distributed along the graphene layer and develop a local electric field. This self-generated field causes an electroosmotic fluid flow by proton migration in the electric double layer of the swimmer. This induces a force (FE, orange arrow, Fig. 1A) experienced by the graphene swimmer, propelling it in the direction opposite to the proton flux. To validate this self-electrophoresis mechanism, the dynamic behaviour of the enzymatic graphene swimmer was systematically investigated under varying conditions. A first control experiment was carried out in an oxygen-free solution. As in this case no electron transfer is possible, consequently no significant motion is observed besides a small random drift (Fig. S2 and Video S1A). In the next set of experiments, an oxygen-saturated 0.1 M phosphate buffer solution (pH 7.0), maintained at 37 °C via a water bath and containing 20 mM glucose as fuel, has been employed. As shown in Video S1B, and also illustrated by the trajectory displayed in Fig. 1B, the graphene swimmer exhibits a linear propulsion in the direction of the Gox-modified edge with an average speed of 2.7 cm s−1.

Most interestingly, when an external magnetic field (Fig. 1, blue arrow) is imposed on this system, the induced Lorentz force (FL, green arrow in Fig. 1), in synergy with FE, results in a complete change of the swimmer's trajectory, despite the absence of magnetic elements in its architecture. We analyzed in more detail the generated circular motion in the first 5 seconds. With the northpole facing up (Fig. 1C and Video S1C), a clockwise rotation is observed with an average speed of 2.6 cm s−1. In contrast, with the southpole up (Fig. 1D and Video S1D), counter clockwise rotation with an average speed of 2.7 cm s−1 is recorded. Notably, the linear component of the motion is always in the direction of the Gox-modified extremity. The presence of an external magnetic field also creates a magnetohydrodynamic (MHD) flow around the two enzyme spots. This convection can also accelerate the reaction,19 and thus the energy conversion efficiency. An enhanced linear speed is observed at short times in the presence of the magnetic field with the north pole facing up (Fig. S3A), and a similar increase is detected for the south pole-up configuration (not shown). This improvement can be attributed to the generation of a local MHD flow, facilitating the transport of substrate molecules towards both enzyme spots in the early moments of the experiment when the swimmer starts to move. Once swimmer motion is established, this additional MHD-related mass transport has no longer a significant impact on the overall reaction rate and thus the speed.

To provide a more quantitative analysis of the influence of the magnetic field on the dynamic behaviour of the swimmers, the angular displacement and angular speed under different conditions (no external magnetic field, north pole up, and south pole up) were analysed using trajectory data extracted from Videos S1A–C. Fig. 2 compares the initial angular speed of the three experiments after placing the swimmer at the air/solution interface. In the absence of the external magnetic field, it exhibits only a neglectable angular speed, due to either a drift or a slight asymmetry in the shape of the graphene sheet. When the magnetic field is applied, the angular speed is 193° s−1 and 185° s−1 for the north pole and south pole-up configuration, respectively.


image file: d5cc03912f-f2.tif
Fig. 2 Initial maximum of angular speed of enzymatic graphene swimmers without magnetic field (grey), with northpole (purple) or southpole up (orange). Error bars correspond to the precision of the speed measurement.

The linear and angular speeds of the graphene swimmer not only depend on the magnetic field, but are also directly correlated with the efficiency of both redox reactions involving Gox and BOD, respectively. A critical parameter influencing the efficiency of the redox reactions is the enzyme loading. All the above discussed experiments were conducted with the same enzyme loading ratio (BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 1[thin space (1/6-em)]:[thin space (1/6-em)]1). However, the speed of the enzymatic graphene swimmer may be limited by the conversion of oxygen, due to its low solubility, and the associated consumption of protons by BOD. To address this aspect, the loading ratio of BOD and Gox was changed to 2[thin space (1/6-em)]:[thin space (1/6-em)]1. This enhanced BOD loading enables a faster consumption of the protons and electrons generated by Gox. Consequently, swimmers with a doubled BOD loading exhibit an average speed of 3.7 cm s−1, a 1.4 times enhancement compared to the loading ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]1 under otherwise identical conditions (Fig. 3A, trajectory extracted from Video S2A). Optimization of the enzyme loading represents therefore an effective strategy to improve propulsion performance and thus we further investigated in more detail the motion and angular speed as a function of this parameter. Fig. S3 illustrates the effects of the magnetic field and varying enzyme loadings. First, in the case of the same loading of BOD and Gox (black and green curves in Fig. S3A), the speed of both swimmers gradually decreases and the magnetic field doesn’t seem to have a major impact on the linear speed component, except for the very beginning of the experiment (vide supra). In strong contrast to this, the angular speed component (black and green curves in Fig. S3B) is very sensitive to the presence of the magnetic field. In the absence of the magnetic field, the angular speed is close to zero with minor fluctuations. However, when the magnetic field is applied, the angular speed is significantly higher. Nevertheless, the angular speed tends to decrease gradually over time, which is consistent with the trend of the linear motion. Two reasons might be at the origin of this behaviour: (1) initially, the redox polymer contains dissolved oxygen, so when the swimmer is placed at the air/solution interface, the enzymes encapsulated in the redox polymer will consume both, the stored oxygen and oxygen from the solution. This simultaneous consumption of oxygen, provided from two sources, results in a temporarily higher speed. After the oxygen, locally stored in the hydrogel, is consumed, oxygen from the solution needs to penetrate the gel to reach the enzyme, and this transport limitation leads to a gradual decrease of linear and angular speed. (2) The swimmers move at an initially quite high speed and this creates rather important shear forces which gradually remove parts of the enzyme-containing hydrogel as it is facing downwards and is therefore directly exposed to the strong local hydrodynamics. Thus, less and less fuel can be converted per time, inducing a decrease in speed. Unfortunately, this decrease of activity prevents using the same swimmer several times. However, we have studied multiple independent swimmers and observed a very similar dynamic behavior, indicating a good reproducibility of the protocols. Related to this aspect, improvements of the life-time might be achieved by protecting the enzyme layers with membranes.


image file: d5cc03912f-f3.tif
Fig. 3 (A) Trajectory of an enzymatic graphene swimmer (BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1) moving at the air/water interface of a O2 saturated 0.1 M PBS, 20 mM glucose solution, at pH 7 and 37 °C, in the presence of a magnetic field with the north pole up. (B) Trajectory of an enzymatic graphene swimmer (BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1) moving at the air/water interface of a O2 saturated 0.1 M PBS, 40 mM glucose solution, at pH 5 and 37 °C, in the presence of a magnetic field with the north pole up. (C) Maximum of the initial angular speed of enzymatic graphene swimmers with the north pole-up configuration with different loading ratios between BOD and Gox, glucose concentrations, and for different pH values. (Green[thin space (1/6-em)]:[thin space (1/6-em)]BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 1[thin space (1/6-em)]:[thin space (1/6-em)]1, 20 mM glucose, pH 7. Blue[thin space (1/6-em)]:[thin space (1/6-em)]BOD[thin space (1/6-em)]:Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1, 20 mM glucose, pH 7. Red[thin space (1/6-em)]:[thin space (1/6-em)]BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1, 40 mM glucose, pH 5.) Error bars correspond to the precision of the speed measurement.

When the BOD loading is doubled with respect to the amount of immobilized Gox, linear and angular speed show a considerable increase (blue curves in Fig. S3). This indicates that oxygen conversion is the rate limiting process and directly impacts the propulsion efficiency (vide supra). Also in this case the speed decreases as a function of time, and it is reasonable to assume that this is due to a gradual stripping of the redox hydrogel exposed to strong hydrodynamic forces.

To further enhance the speed of the graphene swimmers, the composition of the fuel solution can also be optimized. Based on swimmers with the BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1 loading, the glucose concentration has been increased from 20 mM to 40 mM, and the solution pH was adjusted to 5. Under these conditions, the swimmer shows a significantly improved average linear speed of 4.8 cm s−1 (see Fig. 3B and Video S2B). Obviously, the higher glucose concentration can provide more substrate and thus promote the catalytic activity of Gox, but most importantly, a pH of 5 is close to the optimal pH of Gox and BOD, which enhances the activity of both enzymes, thus strengthening the proton gradient and the resulting electrophoretic force. The weight of the swimmer is also an important parameter to achieve high speed. The weight of a graphene swimmer is around 30 μg, which represents a unique advantage of using a graphene monolayer as a support.

Besides the linear speed component, the average angular speed of the enzymatic graphene swimmers is also affected by the enzyme loading and the composition of the fuel solution. Fig. 3C compares swimmers with different enzyme loading ratios and after fuel solution optimization. The green bar represents the initial condition (BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 1[thin space (1/6-em)]:[thin space (1/6-em)]1) with an average angular speed of 193° s−1. Doubling the BOD loading (blue bar, BOD[thin space (1/6-em)]:[thin space (1/6-em)]Gox = 2[thin space (1/6-em)]:[thin space (1/6-em)]1) improves the average angular speed to 286° s−1, while further fuel solution optimization (red bar, 40 mM glucose, pH 5) allows achieving a remarkable average angular speed of 383° s−1, representing a 100% enhancement compared to the initial condition. The increased BOD loading and the optimization of the composition of the fuel solution amplify both, the proton flux between the two enzyme patches and the electron flow through the graphene sheet. This leads to an enhanced FE and, most importantly, in the presence of the external magnetic field, also to an increase in FL. To estimate the power output of such a device, electrochemical measurements were conducted with a classic biofuel cell of analogue composition (Scheme S2). The maximum current density flowing between the bioanode and biocathode reaches 0.35 mA cm−2 and a potential difference of 0.32 V (Fig. S4). The corresponding initial power output is 0.11 mA cm−2 and decreases over time, reaching a smaller but rather stationary value at longer times (Fig. S5A). This is in good agreement with the temporal evolution of the swimmer speed (Fig. S5B and Video S2B). In addition, the conversion efficiency from electrical to mechanical energy was estimated to be at least 2.4% (see details in the SI).

In conclusion, an enzymatic swimmer has been successfully developed through controlled immobilization of two enzymes on a graphene monolayer. The spatial separation of Gox/BOD patches enables the generation of a proton gradient, triggering autonomous motion with an extremely high speed, reaching up to 6 cm s−1 after optimization. Most importantly, for the first time we show that the trajectory and speed of such an enzymatic swimmer can be controlled and enhanced by an external magnetic field, allowing for an optimized energy transfer efficiency. This is also made possible, because graphene, as a typical 2D material, offers distinct advantages, including excellent conductivity, flexibility and a high surface area per weight. These proof-of-concept results bridge energy conversion with macroscopic robotics, providing an interesting alternative for biomimetic soft robotic design and targeted delivery applications.

This work has been supported by the National Natural Science Foundation of China (No. 22272044), Natural Science Foundation of Henan Province (232300421079), the Henan center for Outstanding Overseas Scientist (GZS2024004) and the French National Research Agency (ANR) in the frame of the project COOOL (ANR-24-CE07-1874-02). We thank Dr Sara Grecchi for helping us with some experiments.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Supplementary information contains detailed description of the methodology, additional results and supporting videos. See DOI: https://doi.org/10.1039/d5cc03912f

Notes and references

  1. (a) V. K. Vaithyanathan, B. Goyette and R. Rajagopal, Environ. Challenges, 2023, 11, 100700 CrossRef CAS ; (b) A. Hommes, H. J. Heeres and J. Yue, ChemCatChem, 2019, 11, 4671–4708 CrossRef CAS .
  2. X. Arqué, T. Patiño and S. Sánchez, Chem. Sci., 2022, 13, 9128–9146 RSC .
  3. C. Beans, Proc. Natl. Acad. Sci. U. S. A., 2025, 122, e2426630122 CrossRef CAS .
  4. J. M. Le and K. L. Bren, ACS Energy Lett., 2019, 4, 2168–2180 CrossRef CAS .
  5. (a) D. Pantarotto, W. R. Browne and B. L. Feringa, Chem. Commun., 2008, 1533–1535 RSC ; (b) M. Mathesh, J. Sun and D. A. Wilson, J. Mater. Chem. B, 2020, 8, 7319–7334 RSC ; (c) T. Patiño, N. Feiner-Gracia, X. Arqué, A. Miguel-López, A. Jannasch, T. Stumpp, E. Schäffer, L. Albertazzi and S. Sánchez, J. Am. Chem. Soc., 2018, 140, 7896–7903 CrossRef .
  6. S. Arnaboldi, G. Salinas, A. Karajić, P. Garrigue, T. Benincori, G. Bonetti, R. Cirilli, S. Bichon, S. Gounel, N. Mano and A. Kuhn, Nat. Chem., 2021, 13, 1241–1247 CrossRef CAS PubMed .
  7. L. K. E. A. Abdelmohsen, M. Nijemeisland, G. M. Pawar, G.-J. A. Janssen, R. J. M. Nolte, J. C. M. van Hest and D. A. Wilson, ACS Nano, 2016, 10, 2652–2660 CrossRef CAS PubMed .
  8. S. Tang, F. Zhang, H. Gong, F. Wei, J. Zhuang, E. Karshalev, B. Esteban-Fernández de Ávila, C. Huang, Z. Zhou, Z. Li, L. Yin, H. Dong, R. H. Fang, X. Zhang, L. Zhang and J. Wang, Sci. Robot., 2020, 5, eaba6137 CrossRef PubMed .
  9. (a) S. El Ichi, A. Zebda, A. Laaroussi, N. Reverdy-Bruas, D. Chaussy, M. Naceur Belgacem, P. Cinquin and D. K. Martin, Chem. Commun., 2014, 50, 14535–14538 RSC ; (b) H.-B. Noh, M. Halappa Naveen, Y.-J. Choi, E. S. Choe and Y.-B. Shim, Chem. Commun., 2015, 51, 6659–6662 RSC ; (c) T. Zhang, Science, 2020, 367, 1305–1306 CrossRef CAS PubMed .
  10. (a) C. Gao, C. Zhou, Z. Lin, M. Yang and Q. He, ACS Nano, 2019, 13, 12758–12766 CrossRef CAS PubMed ; (b) C. Zhou, C. Gao, Z. Lin, D. Wang, Y. Li, Y. Yuan, B. Zhu and Q. He, Langmuir, 2020, 36, 7039–7045 CrossRef CAS PubMed .
  11. N. Mano and A. Heller, J. Am. Chem. Soc., 2005, 127, 11574–11575 CrossRef CAS PubMed .
  12. S. Arnaboldi, G. Salinas, S. Bichon, S. Gounel, N. Mano and A. Kuhn, Nat. Commun., 2023, 14, 6390 CrossRef CAS PubMed .
  13. (a) A.-I. Lazăr, K. Aghasoleimani, A. Semertsidou, J. Vyas, A.-L. Rosca, D. Ficai and A. Ficai, Nanomaterials, 2023, 13, 1092 CrossRef PubMed ; (b) Z. Li, S. Li, Y. Xu and N. Tang, Chem. Commun., 2023, 59, 6286–6300 RSC ; (c) K. Cao, S. Feng, Y. Han, L. Gao, T. Hue Ly, Z. Xu and Y. Lu, Nat. Commun., 2020, 11, 284 CrossRef CAS PubMed .
  14. U. Choudhury, L. Soler, J. G. Gibbs, S. Sanchez and P. Fischer, Chem. Commun., 2015, 51, 8660–8663 RSC .
  15. (a) N. Shida, Y. Zhou and S. Inagi, Acc. Chem. Res., 2019, 52, 2598–2608 CrossRef CAS PubMed ; (b) G. Loget, D. Zigah, L. Bouffier, N. Sojic and A. Kuhn, Acc. Chem. Res., 2013, 46, 2513–2523 CrossRef CAS PubMed ; (c) L. Zuccaro, A. Kuhn, M. Konuma, H. K. Yu, K. Kern and K. Balasubramanian, ChemElectroChem, 2016, 3, 372–377 CrossRef CAS .
  16. (a) R. Gao, M. Beladi-Mousavi, G. Salinas, L. Zhang and A. Kuhn, Chem. Phys. Chem., 2024, 25, e202400257 CrossRef CAS PubMed ; (b) R. Gao, S. M. Beladi-Mousavi, G. Salinas, P. Garrigue, L. Zhang and A. Kuhn, Nano Lett., 2023, 23, 8180–8185 CrossRef CAS PubMed .
  17. S. Krause, E. Ploetz, J. Bohlen, P. Schüler, R. Yaadav, F. Selbach, F. Steiner, I. Kamińska and P. Tinnefeld, ACS Nano, 2021, 15, 6430–6438 CrossRef CAS PubMed .
  18. (a) F. Mao, N. Mano and A. Heller, J. Am. Chem. Soc., 2003, 125, 4951–4957 CrossRef CAS PubMed ; (b) F. Conzuelo, N. Marković, A. Ruff and W. Schuhmann, Angew. Chem., Int. Ed., 2018, 57, 13681–13685 CrossRef CAS PubMed .
  19. (a) G. Salinas, T. Safarik, M. Meneghello, S. Bichon, S. Gounel, N. Mano and A. Kuhn, Chem. – Eur. J., 2025, 31, e202403329 CrossRef CAS PubMed ; (b) G. Salinas, K. Tieriekhov, P. Garrigue, N. Sojic, L. Bouffier and A. Kuhn, J. Am. Chem. Soc., 2021, 143, 12708–12714 CrossRef CAS PubMed .
This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.