---
title: MXene-Dispersed Polymer Trilayer Actuators
url: https://www.emergentmind.com/topics/mxene-dispersed-polymer-trilayer-actuators-mptas
type: topic
---

# MXene-Dispersed Polymer Trilayer Actuators

Searching arXiv for the specified paper and closely related MXene/polymer actuator modeling work to ground the article in current literature.
arxiv.search code:
{"query":"id:2509.00019 OR \"Photothermomechanicaly Efficient, Low-Cost, High-Cycle-Life, Hybrid MXene-Polymer Actuators\"","max_results":5,"sort_by":"submittedDate","sort_order":"descending"}
arxiv.search code:
{"query":"id:2210.08689 OR \"Strongly Modulated Exfoliation and Functionalization of MXene with Rational Designed Groups in Polymer: A Theoretical Study\"","max_results":5,"sort_by":"submittedDate","sort_order":"descending"}
arxiv.search code:
{"query":"id:2506.14840 OR id:2305.07655","max_results":10,"sort_by":"submittedDate","sort_order":"descending"}
MXene-dispersed polymer trilayer actuators (MPTAs) are a class of **UV-driven photothermomechanical soft film actuators** in which **Ti\(_3\)C\(_2\)T\(_x\)** MXene is physically dispersed within an elastomeric top layer of a trilaminate composed of a photothermally expanding elastomer, a plastic adhesive interlayer, and a thermohygroscopically contracting paper layer. Under **365 nm UVA irradiation**, uniform heating generates a strain mismatch: the MXene-elastomer expands, the paper contracts as adsorbed moisture is driven off, and the intermediate plastic layer transmits and stabilizes the mismatch, producing reversible bending. The design was introduced to address high operational temperatures, limited mechanical output relative to incident radiation, inadequate durability, and cost barriers in prior photothermomechanical polymer actuators [2509.00019].

## 1. Device class and trilayer architecture

MPTAs are defined by a **true trilayer** architecture rather than a coated bimorph or standalone MXene film. The three constituent layers are a **photothermally expanding top layer**, a **rigid/interfacial adhesive intermediate layer**, and a **thermohygroscopically contracting bottom layer**. In the reported implementation, the top layer is a MXene-dispersed UV-curable elastomer based on **Resione F80 Elastic 3D Printer Resin Black** with a noted high content of **acrylated aliphatic urethane**; the intermediate layer is **Esun Hard-Tough Resin** with high **urethane acrylate** content; and the bottom layer is a **75 GSM paper substrate** [2509.00019].

The layer functions are distinct. The top MXene-elastomer layer acts simultaneously as **photothermal absorber** and **thermally expanding active layer**. The plastic middle layer provides **interfacial adhesion**, mechanical integrity, and some rigidity; the authors report that this resin penetrates paper better than the elastomer and reduces delamination. The paper bottom layer provides **thermally induced hygroscopic contraction**, primarily along the axis orthogonal to fiber alignment. This contraction opposes the elastomer’s positive thermal strain and amplifies curvature.

A concise summary of the architecture is given below.

| Layer | Material | Reported role |
|---|---|---|
| Top | MXene-dispersed elastomer | Photothermal absorber; thermally expanding active layer |
| Intermediate | Plastic photopolymer | Interfacial adhesion; mechanical integrity; rigidity |
| Bottom | 75 GSM paper | Thermohygroscopically contracting layer |

MXene is incorporated **only in the elastomeric top layer**. The paper explicitly distinguishes this from MXene coatings or standalone MXene films and states that placing MXene elsewhere tends to reduce interlayer adhesion and increase delamination risk. SEM/EDS showed **homogeneous dispersion** of titanium-based nanosheets in the top layer, **no obvious agglomeration**, and a sieved MXene sheet area distribution of about **10 to 300 \(\mu\text{m}^2\)**.

The reported thicknesses further emphasize the laminate character. SEM cross sections gave approximately **70 \(\mu\)m** for the MXene-elastomer top layer, **50 \(\mu\)m** for the intermediate plastic layer, and **100 \(\mu\)m** for the paper. In the optimized dimensions section, the elastomer thickness was **0.03 mm** without MXene and **0.07 mm** after MXene addition, while the optimized plastic and paper thicknesses were **0.05 mm** and **0.1 mm**, respectively.

## 2. Constituent materials and MXene-specific features

The MXene used in the reported MPTA platform is commercially purchased **Ti\(_3\)C\(_2\)T\(_x\)** from Nanochemazone, with **99%** purity and reported thickness **80–100 nm** [2509.00019]. The work did **not** synthesize MXene from a MAX precursor in-house; instead, the powder was sieved, mixed into elastomer, mechanically ground, degassed, spin-coated, and UV-cured in situ.

FT-IR analysis identified surface-termination-related features consistent with terminated **Ti\(_3\)C\(_2\)T\(_x\)**: **C–F** at about **1100 cm\(^{-1}\)**, **Ti–F** in **750–700 cm\(^{-1}\)**, **Ti–O** in **650–550 cm\(^{-1}\)**, **Ti–C** in about **450–350 cm\(^{-1}\)**, O–H bands at about **3335 cm\(^{-1}\)** and **1395 cm\(^{-1}\)**, and water adsorbed on MXene at about **1640 cm\(^{-1}\)**. These observations were interpreted as consistent with fluorine-, oxygen-, hydroxyl-, and water-associated surface species.

The reported MXene loading series in the elastomer was **0.0, 0.5, 1.0, 2.5, 5.0, 7.5, and 10.0%**, with **2.5%** identified as the optimal practical composition. The rationale given for MXene use is multi-part: **very high theoretical photothermal conversion efficiency** approaching **100%**, strong UV absorption relevant to **365 nm UVA**, **low mid-IR emissivity**, and favorable behavior when dispersed within a polymer rather than used as a pure MXene film, for which the paper notes photothermal efficiencies can be as low as **20%**.

The paper also argues that MPTA performance is not simply a consequence of black coloration. Under otherwise similar conditions at **2.5% filler loading**, a **2.5% MXene** trilayer reached **63.2 \(^\circ\)C**, about **4 \(^\circ\)C** higher than a **2.5% graphite** control [2509.00019]. This is presented as evidence that MXene provides additional photothermal effectiveness beyond optical darkening.

A broader materials-design perspective is supplied by a first-principles study of polymer–MXene interfaces, which is not an actuator paper but is relevant to MPTAs as an interface-selection reference. That work concludes that **bare Ti\(_3\)C\(_2\)** is highly reactive and can cleave or decompose monomers, whereas **O-, F-, and mixed FO-terminated Ti\(_3\)C\(_2\)** interact with monomers mainly through weak van der Waals forces; **Ti\(_3\)C\(_2\)(OH)\(_2\)** can provide stronger binding but may remain chemically active [2210.08689]. This suggests that surface termination is a primary variable in polymer-compatible MXene dispersion, even though the MPTA paper itself does not undertake atomistic interface engineering.

## 3. Fabrication workflow and photothermomechanical operation

Fabrication was designed to be simple and low-cost. MXene particulates were first passed through a **200-mesh screen** with **75 \(\mu\)m aperture**, then added directly to the UV-curable elastomer at the desired concentration and ground using an **agate mortar and pestle** to obtain homogeneous dispersion. The suspension was **vacuum degassed for 10 min**, dispensed onto a **glass wafer**, and **spin-coated at 1,000 rpm for 30 s**. UV curing of the top layer used **365 nm UV light** for **30 s**, leveraging photoinitiators already present in the elastomer. The trilayer laminate was then formed with the paper and plastic photopolymer, peeled from the glass wafer with a **rigid rod** in a direction **orthogonal to the fiber alignment**, and cut into actuator strips; a common strip width for actuation tests was **3 mm** [2509.00019].

The reported paper-fiber orientation is not incidental. The strip axis was placed **orthogonal to the paper fiber alignment**, because this maximizes contraction-induced bending. Fabrication of an MPTA on a **10 cm glass wafer** took about **2 hours**. No added liquid solvent was used for MXene dispersion; the filler was mixed directly into the UV-curable resin.

The photothermal mechanism is described in terms of absorbed radiant power and thermal transients. The paper gives
$$
P = P_{\text{in}(1-10^{-A}) \tag{1}
$$
with absorbance defined by Beer–Lambert’s law,
$$
A = -\log_{10}(P_{\text{out}/P_{\text{in}) \tag{2}
$$
and models heating and cooling as
$$
T = T_{\max} + (T_0 - T_{\max}) e^{-(G/mC)t} \tag{3}
$$
and
$$
T = T_0 + (T_{\max} - T_0) e^{-(F/mC)t} \tag{4}
$$
where \(G\) and \(F\) are heating and cooling coefficients. The authors state that both coefficients, and hence film temperature, vary linearly with MXene concentration.

The optimized **2.5% MXene** composition was identified through spectroscopic and thermomechanical considerations: strong UV absorbance in the irradiation band, substantial visible absorbance from **400–700 nm**, low fluorescence emission, and low reflectance. The authors infer that at **\(\ge 2.5\%\)** MXene, a larger fraction of incident radiation is converted to heat rather than lost by reflection or fluorescence. They also suggest that visible fluorescence emitted by the plastic layer may be reabsorbed and converted thermally when sufficient MXene is present.

The bending mechanism is described by a temperature-to-curvature model adapted from Cai et al.:
$$
\xi \propto x\left[(d_e-a_c)\Delta T + B_e \Delta C_e - B_c \Delta C_c \right] / k \tag{5}
$$
where \(\xi\) is curvature, \(k\) is the plastic photopolymer layer stiffness coefficient, \(d_e\) and \(a_c\) are thermal expansion coefficients of the expanding and contracting layers, and \(B_e\), \(B_c\), \(\Delta C_e\), and \(\Delta C_c\) describe hygroscopic response. The paper emphasizes the material-property contrast enabling this mechanism: elastomer CTE **100–1,000 \(\times 10^{-6}\ \text{K}^{-1}\)**, paper CTE **4–16 \(\times 10^{-6}\ \text{K}^{-1}\)**, and paper CHE up to **0.1 \(C^{-1}\)**.

## 4. Quantitative performance and mechanical behavior

The headline metrics reported for the MPTA platform are a **shed power demand** of
$$
0.1\ \text{mW cm}^{-2}\,^\circ\text{C}^{-1},
$$
a **bending capacity per irradiation power per time** of
$$
0.1^\circ\ \text{mW}^{-1}\text{cm}^2\text{s}^{-1},
$$
and **fatigueless operation of at least 1,000 cycles** [2509.00019]. These are the principal normalized figures used by the paper to position MPTAs among photothermomechanical actuators.

Under **365 nm UV**, **25 mW cm\(^{-2}\)**, with **500 s on / 500 s off**, temperature rose from a maximum of **62 \(^\circ\)C** for **0% MXene** to **80 \(^\circ\)C** for **10% MXene**. Intermediate values included **66.9 \(^\circ\)C** for **2.5% MXene**, **65.2 \(^\circ\)C** for **5%**, and **74.1 \(^\circ\)C** for **7.5%**. Despite the higher temperatures at larger loadings, **2.5%** was selected as the cost–performance optimum.

For **3 mm wide, 2.5% MPTA** strips under **5.4 mW cm\(^{-2}\)** and a **20 s UV on / 40 s off** cycle, lengths from **10 mm** to **65 mm** were tested. Peak temperatures ranged from **39 \(^\circ\)C** to **45 \(^\circ\)C**, with the highest value at **10 mm** length. Displacement increased approximately linearly with length, from about **0.1 mm** at **10 mm** to **3.5 mm** at **65 mm**.

Mechanical output was quantified for the kirigami-inspired flower configuration over **100 cycles**. The maximum force averaged **7.9 mN**. A previous larger analog of **29 mm** length was also noted to show **10 mN** maximum force and **7.9 mN** average force output. From force–displacement analysis, vertical stiffness was **11.4 N m\(^{-1}\)**. Tensile testing of trilayer samples gave an average elastic modulus of
$$
2.93 \pm 0.47\ \text{GPa}.
$$
The paper interprets this relatively high modulus, compared with very soft hydrogels, as supporting durability and useful force output.

Cycle life is one of the central claims. The **kirigami-inspired flower structure** was cycled for **1,000 cycles** under **5.4 mW cm\(^{-2}\)** with **20 s light / 40 s dark**, for a total duration greater than **60,000 s**. Microscopy and visual inspection showed **no delamination** after **1,000 cycles**. In the first **100 cycles**, average peak temperature was **49.0 \(^\circ\)C**, average minimum temperature **35.0 \(^\circ\)C**, and per-cycle temperature swing about **14–15 \(^\circ\)C**. Measured vertical displacement between \(t=0\) and \(t=20\,\text{s}\) was **0.78 mm**. After about **10,200 s**—approximately **170 cycles**—displacement stabilized at about **0.8 mm per cycle**.

Under a long steady-state test of **5,000 s illumination**, temperature reached **66.3 \(^\circ\)C**, stable actuation displacement after about **130 s** was **0.94 mm**, peak temperature was **70.2 \(^\circ\)C**, and peak displacement **0.91 mm**. The paper interprets this stable long-duration response as evidence of robust trilayer bonding.

## 5. Device-scale demonstrations

Three application-scale demonstrations were reported: a **kirigami-inspired flower**, a **parallel manipulator**, and a **soft gripper** [2509.00019]. These demonstrations are significant because they move MPTAs beyond isolated strip tests and show that the trilayer concept can be assembled into coordinated soft robotic mechanisms.

The **kirigami-inspired flower structure** was built from four **2 \(\times\) 29 mm** strips of **2.5% MPTA**, stacked and bonded at the midpoint into an **eight-petaled upside-down flower** configuration. Its role in the paper is primarily as an endurance demonstrator: repeatable cyclic actuation, stable displacement over long times, robust bonding, and measurable force output. Thermography at the **34th cycle** showed the center temperature rising from **25.9 \(^\circ\)C** to **48.0 \(^\circ\)C**.

The **parallel manipulator** used three **10 \(\times\) 20 mm** **2.5% MPTA** legs supporting either a hexagonal paper platform with **10 mm radius** or a rectangular **12 \(\times\) 20 mm** platform. With all three supports irradiated, the platform lowered by **0.96 mm** after **100 s** and recovered after **100 s** without irradiation. With only the left support illuminated, the platform tilted **5\(^\circ\)** leftward after **100 s** and recovered in the dark. The paper explicitly relates this mode of actuation to distributed, independently addressable support actuators for **parallel mechanisms**, **reconfigurable surfaces**, **shape-programmable platforms**, and potentially large actuator arrays.

The **soft gripper** used two **10 \(\times\) 20 mm** **2.5% MPTAs** attached to triple-stacked **20 \(\times\) 3 mm** neodymium coin magnets, each stack providing **47.7 mT**. The arm was a 3D-printed TPU origami-inspired spring structure. Vertical steering was provided by a **59.0 mT electromagnet**, and manual horizontal steering used a **200 mT magnet**. In the reported operating sequence, the arm was aligned in **11 s**, UV was turned on at **\(t = 12\ \text{s}\)**, gripping force became sufficient by about **80 s** to lift a marshmallow, the object was moved above a muffin in **30 s**, UV was turned off at **\(t = 125\ \text{s}\)**, and the object was released and laid flat after **60 s**, with final placement at **\(t = 197\ \text{s}\)**. The paper presents this as evidence of untethered local actuation at the end effector and compatibility with gentle manipulation.

A systems-level point made in connection with the gripper is the spectral choice of UVA. The authors note that visible light can remain available for machine vision while IR can be reserved for other functions, reducing spectral interference in mixed-function robotic environments.

## 6. Design rules, limitations, and connections to related research

Several design rules are stated explicitly. The first is **optimal MXene loading**: **2.5% MXene** in the elastomer is identified as the best practical composition. Below this level, absorbance and photothermal conversion are weaker; above it, temperature gains become marginal relative to increased cost, fragility, and microstructural granularity [2509.00019]. The second rule is **dispersion quality**: the reported performance is attributed in part to homogeneous, unagglomerated MXene dispersion. The third is **layer placement**: MXene should be placed **only in the elastomer layer** to minimize interfacial weakening and delamination. The fourth is **fiber orientation**: the actuator long axis should be **orthogonal to paper fiber alignment**. The fifth is a **stiffness trade-off**: lower stiffness favors easier bending, whereas higher stiffness supports greater force output.

The paper also defines a specific limitations set. **UV exposure concerns** are acknowledged directly: UV and visible light can degrade polymers, and long-term UV exposure can harm biological tissue. **Actuation speed is moderate rather than ultrafast**; practical demonstrations operate on **20–100 s** timescales, and the gripper requires tens of seconds for full grip and release. **Mechanical output characterization remains incomplete**, with future work suggested for more comprehensive force, work, and efficiency evaluation. Additional limits include **sample variance and fabrication variability**, the probable **environmental sensitivity** associated with paper moisture exchange, and trade-offs at higher MXene loading.

Several broader research connections sharpen the interpretation of MPTAs without changing their reported metrics. The theoretical study of MXene–polymer interfaces already noted implies that terminated **Ti\(_3\)C\(_2\)T\(_x\)** surfaces can support non-destructive polymer integration and tunable interfacial charge transfer, whereas bare **Ti\(_3\)C\(_2\)** is too reactive for controlled polymer incorporation [2210.08689]. A plausible implication is that future MPTA variants could treat surface termination and polymer functional group chemistry as co-optimized variables rather than regarding MXene simply as a photothermal filler.

A separate MXene–hydrogel study shows that dispersed **Ti\(_3\)C\(_2\)T\(_x\)** can act as a **polymerization microreactor**, producing a **heterogeneous single covalent network nanocomposite hydrogel** with **hyperbranched domains embedded in a highly entangled matrix**, tensile strength up to **2.4 MPa**, toughness **75.2 kJ m\(^{-2}\)**, and hysteresis **2.9%** at **200% strain** [2506.14840]. That work is not an MPTA paper, but it suggests a route toward more fatigue-resistant passive or interlayer materials in future trilayer actuators.

For modeling, a PEDOT-based **pseudo-trilayer** actuator study offers a continuum framework in which mobile ions and a deformable porous solid are coupled through electrostatics, swelling, and beam bending [2305.07655]. The paper is not about MXene or photothermal operation, but its treatment of **through-thickness functional asymmetry**, **near-electrode steep profiles**, and beam-resultant quantities such as tip displacement and blocking force suggests a transferable formal structure for future MPTA constitutive modeling. This suggests that, beyond their present experimental formulation, MPTAs can be situated within a broader family of layered or pseudo-layered soft actuators in which localized transduction, asymmetrical strain generation, and laminated mechanical stabilization are the governing principles.

Source: https://www.emergentmind.com/topics/mxene-dispersed-polymer-trilayer-actuators-mptas