Magnetic Hydrogel Micromachines with Active Release of Antibacterial Agent for Biofilm Eradication

Bacterial biofilms are widely found in nature, and they can be a major factor in many local chronic infections inside the human body. To date, effective elimination of biofilms is intractable for conventional antibiotic therapies due to the existence of bacteria‐protecting extracellular polymeric substances. For biofilm eradication in tiny tubular structures, magnetic hydrogel micromachines (MHM) are designed that are capable of performing mechanical disruption of biofilms while releasing antibacterial agents in a controllable fashion. Two modes of motion controlled by external magnetic fields, namely planar rotation and wobbling modes, have been investigated to drive the micromachines to interact and destruct biofilms. In addition, the presented micromachines are composed of thermosensitive PNIPAm hydrogel that can load with the antibacterial agent (H2O2) and release them upon heating. The Fe3O4 nanoparticles embedded in the hydrogel matrix act as catalysts for the Fenton reaction, generating bacteria‐killing free radicals. The synergetic biofilm elimination effect has been experimentally demonstrated in well plates and small tubes. The presented MHMs may provide a new design to treat biofilm‐related infections in narrow and confined spaces.


Introduction
Bacterial biofilms, the three-dimensional assembly of microorganisms and extracellular polymeric substances (EPS), are highly related to infectious diseases. [1,2]Both biotic (e.g., teeth) or abiotic (e.g., implants) surfaces can be colonized by pathogenic biofilms, where the latter may be more predisposed to microbial Bacterial biofilms are widely found in nature, and they can be a major factor in many local chronic infections inside the human body.To date, effective elimination of biofilms is intractable for conventional antibiotic therapies due to the existence of bacteria-protecting extracellular polymeric substances.For biofilm eradication in tiny tubular structures, magnetic hydrogel micromachines (MHM) are designed that are capable of performing mechanical disruption of biofilms while releasing antibacterial agents in a controllable fashion.Two modes of motion controlled by external magnetic fields, namely planar rotation and wobbling modes, have been investigated to drive the micromachines to interact and destruct biofilms.In addition, the presented micromachines are composed of thermosensitive PNIPAm hydrogel that can load with the antibacterial agent (H 2 O 2 ) and release them upon heating.The Fe 3 O 4 nanoparticles embedded in the hydrogel matrix act as catalysts for the Fenton reaction, generating bacteria-killing free radicals.The synergetic biofilm elimination effect has been experimentally demonstrated in well plates and small tubes.The presented MHMs may provide a new design to treat biofilm-related infections in narrow and confined spaces.biofilm eradication. [19,20]For example, light-driven micromotors with photocatalytic properties are capable of destroying biofilm structures through their movement, meanwhile catalyzing hydrogen peroxide (H 2 O 2 ) to generate free radicals (i.e., hydroxyl radicals) and kill bacterial cells. [21,22]Liquid metal droplets develop sharp edges under a rotating magnetic field, which enables the droplets to physically break the biofilm matrix and rupture the bacterial cells. [23,24]Other examples include automatic aqua sperm micromotors, [25] peptide/Pt Janus microrobots, [26] and self-propelled catalytic microrobots. [27,28][31][32][33][34] Iron oxide nanoparticles can be magnetically actuated to form collectives (namely microswarm [12,[35][36][37][38] ), which mechanically disrupt biofilms while contributing to the Fenton reaction.The iron oxide nanoparticles do not have the intrinsic bactericidal ability; however, the free radicals generated by the Fenton reaction can effectively kill bacteria cells. [39]In addition, the catalytic effect of iron oxide nanoparticles embedded in agar gel [32] or graphene oxide [40] has also been demonstrated.In summary, two factors are essential for the antibiofilm MNRs to achieve synergetic biofilm eradication: one is the mechanical force derived from the motion of the robot, and the other is an antimicrobial agent, which usually needs to be preinjected into the biofilm colony region.
Herein, we introduce a magnetic hydrogel micromachine (MHM) capable of carrying and releasing antimicrobial agents for targeted biofilm eradication.The proposed micromachine is composed of PNIPAm hydrogel with incorporated Fe 3 O 4 nanoparticles and NdFeB microparticles.PNIPAm hydrogel has been widely investigated in controlled release due to its unique nature: when PNIPAm hydrogel is heated above its lower critical solution temperature (LCST) around 32 °C, it turns to be hydrophobic and tends to expel the carried liquid content. [41]esides, hydrogels are soft, wet, and biocompatible, which makes them great candidates for biomedical applications. [42]We first prepare a magnetic PNIPAm hydrogel composite, then swell the hydrogel with H 2 O 2 solution.After the MHM reaches the target location, we apply an alternating magnetic field (AMF) to heat the magnetic hydrogel through the magnetothermal effect. [43]Consequently, the carried H 2 O 2 solution is released from the MHM (Figure 1a).The polymeric matrix of PNIPAm hydrogel is permeable to H 2 O 2 , thus the Fenton reaction occurs in the hydrogel with embedded Fe 3 O 4 nanoparticles acting as a nanocatalyst (Figure 1b). [32]With the above principles, we design the biofilm eradication experiments as follows: first, the MHM loaded with H 2 O 2 moves to the biofilm location with a wobbling motion (Figure 1c).The magnetothermal effect then heats the MHM to release the H 2 O 2 .After that, the MHM keeps rotating to mechanically disrupt the EPS matrix while carrying out the catalytic reaction (Figure 1d).During this process, bacterial cells are killed by the hydroxyl radicals produced by the Fenton reaction. [44]We present two motion modes of the MHM, planar rotation and wobbling.Finite element simulations are applied to investigate the fluid field profile and diffusion caused by two motions of the MHM.We characterize the catalytic performance of MHMs in static and rotating modes, and then characterize the active release performance of MHMs.We demonstrate that the presented MHMs are capable of removing lab-cultured E. coli and B. cereus biofilms and killing bacterial cells with the presence of H 2 O 2 .Finally, biofilm eradication using MHMs loaded with H 2 O 2 in tube models is demonstrated.

Fabrication of Magnetic Hydrogel Micromachines
We first prepare a precursor of magnetic PNIPAm hydrogel through a simple blending method [45] (Figure S1, Supporting Information).Two types of magnetic particles are mixed into the precursor, where soft-magnetic Fe 3 O 4 nanoparticles with intrinsic peroxidase-like properties endow catalytic activity for the magnetic hydrogel, [32] and hard-magnetic NdFeB microparticles with high residual magnetization endow programmed locomotion for the magnetic hydrogel. [46]It is worth mentioning that NdFeB is toxic to the human body, hence a silica coating is needed for in vivo applications of NdFeB microparticles. [13,47]anoclay is mixed to increase the viscosity of the precursor and prevent undesired aggregation of the magnetic particles. [48]he precursor is photopolymerized in capillary tubes to obtain magnetic hydrogel of different diameters and arbitrary lengths (Figure S2a, Supporting Information).A cross-section SEM image of obtained magnetic hydrogel (Figure S2b, Supporting Information) reveals that the magnetic particles have been incorporated into the hydrogel matrix.The magnetization of the hydrogel is programmed using a template-assisted magnetization method [49] and further verified by a magneto-optical sensor (Figure S3, Supporting Information).

Motion of the Magnetic Hydrogel Micromachines
Since the MHMs incorporate hard-magnetic NdFeB particles, their motion can be precisely controlled by external magnetic fields.The magnetic force exerted on the micromachine is expressed as F ¼ ∇B⋅M, τ ¼ M Â B, [50] where F is the magnetic body force, τ is the generated body torque, B is the magnetic flux density, and M is the magnetization of the micromachine.In our case, the magnetization of the micromachine is consistent with its length direction.Without constraints, the micromachines rotate their body to align the magnetization direction under the applied magnetic field.We present two representative motions of the MHMs: The first mode, planar rotation, is adapted to remove biofilm on flat terrain (Figure 2a).The applied magnetic field vector rotates in a plane, causing the MHM rotates in the same plane (Figure 2b).The rotating frequency of the micromachine in PBS solution is consistent with the applied magnetic field (Figure 2c).The rotating magnetic field can be generated either by a Helmholtz coil setup or by a ball magnet bonded to a motor, the latter setup produces a gradient field that drags the micromachine to the magnet.To better understand the flow field and fluid diffusion induced by the wobbling MHM, we conduct finite element simulations.It is found by simulation that the fluid pressure caused by the rotating MHM ranges from À57.7 Pa (near the rotation center) to 1.08 Pa (about 15 mm away from the micromachine, Figure S4, Supporting Information), such pressure (1.08 Pa) is insufficient to fully remove established biofilms. [51]Still, the rotation motion of the magnetic micromachines is beneficial to improve liquid convection.High-speed rotation of MHM generates fast flow, further resulting in rapid diffusion of released H 2 O 2 (Figure 2d).
The other mode, namely wobbling motion, [11,40,52,53] can drive the rod-like micromachines to be propelled forward or backward in tubes.Such motion is generated by a conical magnetic field, where the terminal point of the field vector varies along a circle (as shown in Figure 2e). [35,36]The conical magnetic field is expressed as where H y is the DC component along the principal axis of the cone, and H r is the rotational component of the conical magnetic field.Subsequently, the micromachine rotates around the y-axis in response to the conical magnetic field.Due to the existence of gravitational force (acting in the minus z-direction), the frictional force between the MHM and the lower half tube wall is greater than that of the upper half, as illustrated in Figure 2f.Note that the density of the MHMs is much greater than that of water due to the incorporated magnetic particles.The asymmetric frictional force drives the micromachine to move along the tube (Figure 2g, Movie S1, Supporting Information).We use the MHM in the dish (which is also premagnetized along its length direction) to indicate the applied magnetic field direction.Moreover, the MHM in the dish shows a walking motion under conical magnetic fields, where both ends take turns hitting the ground, and the friction drives the MHM in the dish moving toward the positive x-direction (Movie S1, Supporting Information).We can treat the dish as a portion of a tube with a very large diameter, and the frictional force between the MHM and the tube generates only at the lower half surface of the tube.In this way, the MHM in the dish also indicates the total frictional force direction of the MHM in the tube.Because the motion of MHM in the tube is constrained by the tube wall, the MHM finally moves along the tube while scratching the tube's inner surface periodically.We employ such motion of MHMs to disrupt biofilms attached to the inner surface of the tube.We find that the moving direction of the MHM can be altered by changing the rotation axis of the conical magnetic field, and the translational velocity of the MHM in the tube increases with the rotating frequency of the applied magnetic field (Figure 2h).The translational velocity of MHMs in the tube is also related to the angle between the tube and the principal axis of the conical magnetic field (Figure S5, Supporting Information).We model the wobbling motion of the micromachine as a rotating rod in a cylindrical space filled with liquid, and the diffusion of the released H 2 O 2 solution is modeled as transport of diluted species.Simulation results show a stable flow field after 5 s of actuation, where the maximum flow rate appears near the micromachine's wobbling trajectory.Also, the fast diffusion of H 2 O 2 in 5 s has been shown (Figure 2i).Although many studies have demonstrated the locomotion of soft-magnetic micromachines under uniform magnetic fields, our experimental results indicate that Fe 3 O 4 -embedded MHMs without NdFeB content can hardly show wobbling motion under weak magnetic fields (Figure S6, Supporting Information).

Catalytic and Release Performance
It has been verified that Fe 3 O 4 nanoparticles act as peroxidasemimetics which can be used as a catalyst for the Fenton reaction. [30]However, the catalytic performance of Fe 3 O 4 nanoparticles embedded in a hydrogel matrix has yet to be adequately studied.Here we test the catalytic performance of the MHMs in H 2 O 2 solution with a TMB (3,3 0 ,5,5 0 -tetramethylbenzidine) colorimetric experiment.In a TMB/H 2 O 2 system, the decomposition of H 2 O 2 generates hydroxyl radicals (•OH), which further oxidates TMB into blue oxTMB and turns a transparent solution into a blue one. [54]We measure the absorption spectrum of the TMB/H 2 O 2 solution after reacting with the MHMs for a given time period, it is found that the 3 mg rotating hydrogel experiment group exhibits higher catalytic activity than the other groups (Figure 3a and S7a, Supporting Information); the 3 mg static hydrogel group shows a peak absorbance slightly higher than the 1 mg static hydrogel group, but much lower than the 3 mg rotating hydrogel group.Such a result highlights the importance of liquid convection caused by micromachine rotation for enhanced catalytic performance.We further measured the absorbance change of the TMB/H 2 O 2 solution catalyzed by rotating MHM or static MHM.Rotating MHM shows higher catalytic activity compared to the static MHM in the 10 min of the experiment (Figure 3b).It is noted that even at a relatively low rotating speed, rotating MHMs still exhibit better performance than static ones.The catalytic activity of MHMs is also characterized by a methyl violet degradation assay (Figures S7b and S8a, Supporting Information) and a DCFH-DA assay (Figure S8b, Supporting Information).We design a controlled TMB colorimetric experiment to verify the drug loading and release ability of the MHMs.As shown in Figure 3c, three groups of TMB solution are prepared.We swell the MHMs in H 2 O 2 solution to load the H 2 O 2 , then, the micromachines are immersed into the prepared TMB solutions.For the control group where no H 2 O 2 has been loaded, the solution remains transparent after 5 min of reaction (Figure 3c).
For the passive release group, hardly any blue color can be observed because of slow diffusion while there is no observable liquid convection.For the active release group, the micromachine is first heated by an induction heating system to release the H 2 O 2 and then subjected to a rotating magnetic field to stir the solution.The MHMs can be remotely heated to above 39.3 °C by the magnetothermal effect (Figure 3d).The resulting blue solution suggests more hydroxyl radicals (•OH) have been produced by the catalytic activity of the MHM.The absorbance at 652 nm (peak absorption) is also measured for the three groups (Figure 3e). Figure 3f shows the mass change of the MHMs during the active release experiment.After 15 min of swelling, the total mass of the micromachines and the carried solution is %1.9 times compared to the as-prepared state.After releasing the solution, the total mass becomes %1.15 times of the as-prepared state, indicating 75% cargo capacity of the MHMs.

Biofilm Eradication in the Well Plate
To investigate the synergetic biofilm eradication performance of the MHMs, we incubate gram-negative E. coli or gram-positive B. cereus biofilms on 24-well cell culture plates and then treat them with the MHMs.At the beginning of the experiment, the bottom of the well is covered with biofilm, which made the bacterial cells resistant to external environments (Figure 4a).The MHM is added to the bacterial solution, and a rotating magnetic field is applied to actuate the MHM.It is observed that the biofilm in a circular area on the MHM's rotating trajectory has been scraped away from the bottom.The mechanical strength of bacterial biofilms lies on the magnitude of %1 kPa, [55] and our experimental result suggests that the force exerted by the rotating MHM on the biofilm has exceeded its critical strength, hence the biofilm structure is destroyed by the rotating MHM.The MHM then moves along a rectangular path controlled by the external magnetic field, and a clear area on the track can be seen (Figure 4a, Movie S2, Supporting Information).After 3 min of actuation, the bacterial solution becomes turbid with floating EPS, and a clear well can be seen after the well has been rinsed with deionized water.The remaining biofilm structures on the wells are characterized by volumetric fluorescent imaging (Figure 4b).Compared to the control group, the H 2 O 2 -treated group shows lower fluorescent density, suggesting a deactivation effect of H 2 O 2 on the bacterial cells.Static MHM cannot remove biofilm nor kill bacteria; however, both the rotating MHM-treated group and the rotating MHM/H 2 O 2 -treated group demonstrate a great effect on removing the biofilm structures on the well plate.To further find out the efficacy of bacteria killing, we take out some bacterial solution from each well after being treated by the MHMs and culture them for one other day.It is found that the solution of the H 2 O 2 -treated group and of the rotating MHM/H 2 O 2 -treated group show much lower absorbance at 600 nm (OD600) compared with the other three groups (Figure 4c), indicating a good bacteria-killing effect of H 2 O 2 .This result is further confirmed by a plate count method (Figure S9, Supporting Information).Yet, H 2 O 2 treatment alone cannot remove the established biofilm structure, which may lead to the regrowth of the bacterial cells. [56]The efficacy of biofilm removal is characterized by a crystal violet staining assay.After staining, the absorbance at 590 nm is measured as representative of the total biomass in the wells.Both the rotating MHM-treated group and the rotating MHM/H 2 O 2 -treated group show much lower absorbance than the other three groups, which suggests that most of the biofilm has been removed from the wells.In summary, MHM/H 2 O 2 shows a synergetic effect on both chemically killing the bacterial cells and mechanically removing the biofilm.

Biofilm Eradication in Tube Models
Next, we attempt to examine the ability of MHMs in eliminating biofilms in tube models.Two tube models are prepared with biofilms adhered to the inner surface.The MHMs are first immersed in a H 2 O 2 solution to reach a swelled state.We estimate that the catalytic reaction would continuously occur during loading, so the MHMs should be kept immersed in the H 2 O 2 solution before deploying to the biofilm location to minimize the loss of H 2 O 2 .We use a conical magnetic field to drive the MHM to move in the curved tube with a wobbling motion.At the beginning of the experiment, the cloudy biofilm clogs the curved tube.The MHM moves along the tube while wobbling, destroying the biofilm on the way (Figure 5a, Movie S3, Supporting Information).It is observed that little bubbles occur around the wobbling MHM, which can be interpreted as the wobbling motion speeds up the liquid exchange between the hydrogel and the solution, and the catalytic effect of the MHM contributes to the generation of oxygen.The AMF is applied when the MHM moves to the other side of the tube.It is worth noting that most of the visible biofilm has been scratched away from the inner surface of the tube on the first pass-by of the MHM.To completely remove the biofilm, the MHM is controlled to move between the two ends and finally retrieved using a permanent magnet.A clean tube can be seen after rinsing with deionized water.For the second biofilm eradication experiment, biofilm has been established at the end of a microcentrifuge tube (Figure 5b).The rotating gradient magnetic field is used to actuate the MHM to rotate where close to the tube end.Driven by the magnetic field, the micromachine breaks up the densely deposited biofilm and reaches the very end of the tube (Figure 5b and Movie S4, Supporting Information).The efficacy of biofilm removal in tubes is characterized by a crystal violet staining test, and the result shows that tubes treated with MHM contain much lower biomass compared to the control group (Figure 5c).Then, live/dead fluorescent microscopy is performed to access the deactivation effect of the MHM/H 2 O 2 treatment.The active release MHM kills 49.68% AE 12.07% of bacterial cells after treatment, while the passive release MHM kills only 8.38% AE 1.73% of bacterial cells within the same time period (Figure 5d,e and S10, Supporting Information).The result suggests that MHMs with active release ability possess higher efficiency in deactivating bacterial cells within a short time period.

Conclusion
In this article, we introduced MHMs with active release ability for targeted biofilm eradication.The micromachine consists of magnetic PNIPAm hydrogel, which can absorb H 2 O 2 solution and release them upon triggering.Two modes of motion, planar rotation and wobbling, have been proposed to drive the MHMs to disrupt biofilm in different terrains.The rotational motion promotes the liquid convection around the micromachine and further enhances the catalytic performance of the MHMs to generate bacteria-killing free radicals.The synergetic effect of MHMs on both mechanically disrupting biofilm structures and chemically deactivating bacterial cells was verified in well plates.Finally, biofilm eradication using MHMs was demonstrated in curved tubes and microcentrifuge tubes.The on-demand release ability of MHMs eliminates the need to inject large amounts of antibacterial agents into the biofilm site, which makes the MHMs of great potential to remove bacterial biofilms in hard-to-reach locations such as water pipes or medical implants.The limitation of the proposed MHMs is that they can hardly operate in complicated environments such as deep narrow crevices or liquid/gas interfaces.Improvement could be made in terms of material composition, control system, and robot design to achieve better performance of biofilm eradication.

Experimental Section
Preparation of the Magnetic Hydrogel Micromachines: First, deionized water (1 mL), monomer (N-isopropylacrylamide, 0.14 g), crosslinking agent (N,N 0 -methylenebisacrylamide, 1 mg), nanoclay (Laponite XLG, 20 mg), photoinitiator (modified 2,4,6-trimethylbenzoyldiphenyl phosphine oxide, [57] 5 mg), and magnetic particles (0.1 g of Fe 3 O 4 and 0.1 g of NdFeB) were mixed and stirred thoroughly.The mixture was then transferred to a syringe and injected into a polytetrafluoroethylene capillary tube (inner diameter, 300 μm to 1.5 mm).The capillary tube loaded with hydrogel precursor was put into an ice water bath and irradiated under 365 nm UV light for 1 h.After that, the tube was adhered to a mold and magnetized (magnetic field strength, 1.7 T).Finally, the tube was cut to release the MHMs of desired lengths.
Magnetic Actuation Systems: Two different setups were used to actuate the MHMs.A three-axis Helmholtz coil system was used to generate the conical magnetic field for the wobbling motion (maximum magnetic field strength, 10 mT).A permanent magnet ball mounted on a motor was used to generate the rotating gradient magnetic field (maximum magnetic field strength, %100 mT).For the wobbling motion shown in Figure 2, conical magnetic fields of 5 mT in strength, 1-5 Hz in rotating frequency, and an angle of 80°between the magnetic field vector and the y-axis (principal axis) were applied.The translational speed of the MHM was measured with an angle of 45°between the tube and the y-axis.
Numerical Simulation: Simulations were conducted with COMSOL Multiphysics 5.5 software.2D circular geometry was used for the planar rotation in wells, and 3D cylindrical geometry was used for the wobbling motion in tubes.Rotation domain with dynamic mesh was used to simulate the rotation motion of the MHMs.The diffusion of H 2 O 2 was modeled using the transport of diluted species coupled to the flow field, and the initial concentration was set to be Gaussian centered at the centroid of the micromachine.The liquid medium was assumed to be water, and the walls were defined as nonslipping boundaries.
Catalytic Performance Tests: TMB chromogenic assay: First, an acetic acid buffer (pH 4) with 1% H 2 O 2 v/v was prepared as solution A, 3,3 0 ,5,5 0 -tetramethylbenzidine (TMB, 0.2 g) was dissolved in dimethyl sulfoxide (10 mL) as solution B. Then, 5 μL of solution B and 1 mL of solution A were mixed together with the MHM.The mixture was held still or put under a rotating magnetic field (%10 mT, 3 Hz) for 3 min.The corresponding absorption spectrum was recorded using a spectrophotometer (HITACHI U-2910).After that, 5 mL of the above TMB solution was added to a 35 mm culture dish, along with 1 mg of the MHM.A rotating magnetic field was then applied to actuate the micromachine.The corresponding absorbance at 652 nm was measured after every minute.
Active Release of H 2 O 2 : MHM (5 mg) was first swelled in 30% H 2 O 2 solution for 15 min.The micromachine was then put into a TMB solution (1 mL, 0.1 μg mL À1 , pH 4).A commercial induction heating system was used to heat the micromachine to trigger the release of H 2 O 2 .The TMB solution along with the MHM was put into the center of the induction coil and heated for 5 min.Then, a rotating magnetic field (3 Hz) was applied.An infrared camera connected to a cellphone was used to capture the infrared images.
Biofilm Culture: First, E. coli or B. cereus was dispersed in LB (Luria-Bertani) liquid medium.The bacterial solution was shacked for two days (170 RPM) at 37 °C to reach a density of %10 7 CFU mL À1 .After that, the bacterial solution was added to a 24-well culture plate (Thermal Fisher Nunc, 500 μL well À1 ).Then, another 1500 μL of LB liquid medium was added to each well.The bacterial cells in 24-well plate were then cultured at 37 °C for two days to form biofilm at the bottom of the wells.
Biofilm Eradication in 24-Well Plates: The upper layer of bacterial solution was removed, and 500 μL of PBS solution (0.1 M) was added to each well.Then, 3 mg of the MHM and 500 μL of 1% H 2 O 2 solution were added to each well.The pH of the obtained solution was 6.8.Subsequently, a rotating magnetic field (5 Hz) was applied to trigger the motion of the magnetic hydrogel.After 3 min of treatment, the magnetic hydrogel was retrieved using a permanent magnet.
Live/dead fluorescence staining was conducted to access the biofilm eradication efficacy.First, a solution of DI water (1 mL), SYTO-9 (3 μL, Thermo Fisher), and propidium iodide (6 μL) was prepared.Then, 200 μL of the above solution was added to each well and stained for 30 min.Finally, volumetric fluorescence microscopy was performed.For each well, three random areas (1390 Â 988 Â 40 μm 3 ) were scanned, and a total of nine volumetric fluorescent images were obtained for each treatment group.
To evaluate the bacterial-killing efficacy, 100 μL of the bacterial solution was taken out from each well after the treatment and added to a 96 well plate with 1 mL of LB medium in each well.After being cultured at 37 °C for 24 h, the optical density at 600 nm (OD600) of the bacterial solution was measured.
To evaluate the biofilm removal efficacy, crystal violet staining was conducted on the 24-well plate after the biofilm eradication treatment.First, the bacterial solution was removed from the wells, and 200 μL of methanol was added to each well.After 15 min, the liquid was removed from the wells, then the wells were washed with PBS.Next, 200 μL of 1% w/v crystal violet solution was added to each well.After being stained for 5 min, the liquid was removed, then the wells were washed with DI water for 3 times.Finally, 200 μL of 33% v/v acetic acid was added to dissolve the stain.After 30 min, absorbance at 590 nm was measured.
Biofilm Eradication in Tube Models: E. coli and B. cereus cultured for 3 days were collected through centrifuging (2 min at 5000 rpm).The bacteria were transferred to silicon tubes or microcentrifuge tubes with PBS solution (0.1 M) and further cultured for 1 day at 37 °C.The pH of the bacterial solution was 7.1.Before the biofilm eradication experiments, MHMs were swelled in 30% H 2 O 2 solution for 15 min.For the curved tube, the biofilm eradication experiments were conducted with the three-axis Helmholtz coil setup.For the microcentrifuge tube, the biofilm eradication experiments were conducted with the ball magnet setup.In each experiment, the MHM was first navigated to the biofilm site, then subjected to an AMF to trigger the release of H 2 O 2 .After 5 min of treatment, crystal violet staining assay was performed for the curved tubes to evaluate the efficacy of biofilm removal with the same method as aforementioned.Live/dead fluorescent microscopy was performed to evaluate the bacteria-killing efficacy of the bacterial solution in the tubes.Percentage of dead bacteria was calculated as the ratio of red to the green area on the fluorescent images.
Statistical Analyses: All biofilm eradication experiments have been repeated at least three times.Error bars denote mean AE standard deviation.Unpaired Student's t-test was performed to assess the statistical significance with the GraphPad Prism 9 software.

Figure 1 .
Figure 1.Schematic diagram of biofilm eradication using magnetic hydrogel micromachines.a) The magnetic hydrogel micromachines (MHMs) can absorb H 2 O 2 solution and actively release the cargo when heated above the lower critical solution temperature (LCST).b) The incorporated Fe 3 O 4 nanoparticles endow MHM with intrinsic peroxidase-like activity.When immersed in the H 2 O 2 solution, the MHMs act as catalysis for the Fenton reaction, which generates bacteria-killing free radicals.c) Driven by the external magnetic field, the MHM moves to the biofilm location with a wobbling motion.d) The synergetic antibiofilm effect of the MHMs: the micromachines can disrupt the biofilm mechanically and kill bacteria cells by catalyzing the released H 2 O 2 solution.

Figure 2 .
Figure 2. The motion of the magnetic hydrogel micromachines.a) Schematic illustration of the planar rotation.b) The applied magnetic field for the planar rotation.c) Relationship between the rotating frequency of the micromachine and the rotating frequency of the applied field.d) Simulation of the flow field and diffusion caused by the rotation of the MHM after 5 s.The concentration field shown in the figures has been normalized.e) Schematic illustration of the wobbling motion.f ) Schematic illustration of the asymmetric frictional force on the MHM during the wobbling motion.g)The moving direction of the wobbling MHM depends on the direction of the total frictional force and the tube, while the direction of the total frictional force depends on the applied magnetic field.h) Relationship between the translational velocity of the micromachine and the rotating frequency of the applied field.The translational velocity is measured under a 45°angle between the tube and the principal angle of the conical magnetic field.i) Simulation of the flow field and diffusion caused by the wobbling motion of the MHM.

Figure 3 .
Figure 3. Catalytic and release performance of the magnetic hydrogel micromachines.a) The absorption spectrum of TMB/H 2 O 2 solution after reaction with the MHMs.b) Absorbance change of the TMB/H 2 O 2 solution after a given time interval.Inset: snapshots for two experimental groups at 0, 3, and 5 min.c) A comparison between active and passive release.The active release group is heated by the magnetothermal effect and subjected to a rotating magnetic field, the passive release group is reacted without heating or magnetic field, and the control group is reacted with MHM not loaded with H 2 O 2 .d) Thermal image of the MHM before and during heating.e) The chromogenic reaction was quantitively characterized by measuring the absorbance of the solution at 652 nm.f ) Mass change of the MHMs after swelling and releasing the H 2 O 2 solution.

Figure 4 .
Figure 4. Biofilm eradication in the well plate.a) Snapshots of the biofilm eradication process in a well.i: initial biofilm.ii-v: the MHM is deployed into the well and a rotating magnetic field is applied to actuate the micromachine to rotate while moving along a rectangular path.A clean area can be seen on the trajectory of the micromachine.vi: The culture well is rinsed with deionized water.White arrows denote the location of the MHM.b) Volumetric fluorescent images of the biofilm of different treatment groups.Scale bar denotes 500 μm.c) The efficacy of killing the planktonic bacteria.d) The efficacy of removing the biofilm.*P < 0.05.**P < 0.01.***P < 0.001.ns: not significant.

Figure 5 .
Figure 5. Biofilm eradication in tube models.a) Biofilm eradication using MHM in a curved tube.Blue dashed lines denote the location of biofilm, and white arrows denote the moving direction of the MHMs.b) Biofilm eradication using MHM in a microcentrifuge tube.The biofilm is densely deposited at the end of the microcentrifuge tube.c) Total biomass in tubes before and after the MHM treatment.d) Live/dead assay for bacterial cells after passive or active release treatment.e) Fluorescent images for bacterial cells after passive or active release treatment.*P < 0.05.**P < 0.01.***P < 0.001.