Keywords

1 Introduction

Soft actuators in the robotics field have recently attracted attention for their low-cost in fabrication, adaptability, and deformability [1, 2]. Compared with their rigid counterparts, soft robotic devices show a range of notable improvements in safe interaction, complex motions through dynamic shape change and resilience to unstructured environments [3, 4]. Inspired by nature, it is also possible to define a new paradigm in which a robot should be able to evolve selectively in a specific scenario, adapting its morphology according to environmental stimuli and biodegrade at the end of its life cycle [5]. Moreover, an autonomous system should be able to embody chemical or electrical energy sources directly within its materials and structures, rather than requiring separate battery packs [6].

In this framework, bioinspiration represents a role model in the design of multifunctional tools to emulate the highly sophisticated and interconnected systems of natural organisms, such as animals and plants. For instance, the Geraniaceae seeds (e.g., Pelargonium and Erodium genus’) promote their germination exploiting the deformable properties of the bilayer structures, in which bending and twisting occur through exploiting the expansion of an active tissue coupled with an inert material [7, 8].

Multi-layer actuators represent a simple solution as a testing platform for multifunctional materials. Several actuators for soft robotics are demonstrated in literature, such as grasping, lifting, locomotion and soil-exploration [8,9,10]. The inclusion of photothermal properties in thermo-sensitive structures permits to power the actuator from the environmental illumination as a wireless system, without requiring any battery onboard.

Inspired by the bilayer and tissue structure of Geraniaceae seeds, we present a bending photothermal actuator, composed by biodegradable polymers: Polycaprolactone-PluronicF127-Lignin (PPL) blend [11,12,13] printed using Fused Deposition Modeling (FDM) on a cellulose acetate (CA) substrate [14]. We combined the thermal expansion of polycaprolactone (PCL), with the photothermal properties of lignin [15], to induce a local expansion of tissues and, consequently, a local photothermal strain. Pluronic F127 is mainly used as surfactant to better homogenize the blend and to increase the printability properties. The actuator shows a reversible bending behavior when it is subjected to 300 mW/cm2of simulated solar irradiance, in which the change in curvature is Δκ = 24.3 ± 1.3 m−1 (25.34% of relative variation). The bending actuator provides a static moment of M = 80.2 ± 5.3 µN · m.

2 Materials and Methods

2.1 Materials and Chemicals

PCL (Number average molar mass 80 kg/mol) was purchased from ThermoFisher Scientific (Massachusetts, USA). Alkaline lignin was purchased from Tokyo Chemical Industry (Japan). Pluronic F127 and acetic acid (CH3COOH, 99.8%) were purchased from Merck Millipore (Massachusetts, USA). CA 25 µm thick film was purchased from Goodfellow GmbH (Germany).

2.2 Methods

We proceed to describe chronologically the procedures used for the realization and characterization of the photothermal actuator. The methods are presented in the order they were performed.

Preparation of PPL Blend.

The PPL blend was prepared using solvent casting method, mixing 6.5 g of PCL, 0.5 g of Pluronic F127, 0.5 g or 2 g of alkaline lignin (7% w/w and 17% w/w with respect the total weight of the blend) with 10 ml of acetic acid. The solution was continuously stirred at 200rpm with a temperature of T = 80 ℃ for 48 h. Next, the solvent was removed via evaporation, keeping the temperature constant at T = 150 ℃ for 60 min.

4D Printing of the Bending Actuator.

We extruded the solid and homogeneous blend with FDM printer (3DBioplotter, EnvisionTEC) using the following printing conditions: nozzle internal diameter 0.3 mm, temperature 150 ℃, needle offset 0.24 mm, pressure 6 bar, speed 22 mm/s, pre-flow 0.05 s, post-flow −0.05 s. Hence, we fabricated the actuator depositing aligned fibers of PPL with distance between strands 1.5 mm on a CA film. Finally, the CA-PPL sheet is cut using CO2 laser cutter (Beamo, FLUX) with relative power 20 (defined by Beam Studio GUI) and cutting speed 8 mm/s.

Plasma Activation.

To promote adhesion between the two polymers, we activated the CA surface using air plasma (Tergeo, Pie Scientific). The plasma time was set to 60 s, power setpoint to 75 W, gas setpoint 30 sccm (cm3/min) and base vacuum to 0.50 mbar.

Photothermal Characterization.

FDM printed PPL samples (1 cm × 1 cm × 300 µm) were irradiated under solar spectrum SciSun 300 Sun Simulator (Sciencetech, Canada) at different powers (100–300 mW/cm2). Samples were suspended in air using Teflon tweezers to avoid any thermal diffusion in the sample holder. Power densities of the simulated sunlight were verified using an RS PRO solar energy meter (RS, UK). The variation of temperature due to photothermal effect was carried out with an IR thermal camera (A700, FLIR Systems, USA). The experiments were performed in ambient air (T = 25 ℃, 30% RH).

Curvature Evaluation.

The actuator was exposed to 300 mW/cm2 of simulated solar irradiance (SciSun 300 Sun Simulator Sciencetech, Canada), while simultaneously video recording the variation of radius (Logitech Brio Stream, Logitech). To ensure complete thermalstability, the samples were exposed for 5 min. To evaluate the variation of curvature, all the video-data are post-processed using ImageJ software [16]. The experiments were performed in ambient air (T = 25 ℃, 30% RH measured using RS PRO RS-325A Digital Hygrometer). The evaluation of curvature variation due to hygroscopic effect was performed in climatic test chamber (CTC256, Memmert GmbH), fixing the temperature T = 30 ℃ and changing RH from 30% to 90%.

Static Moment Test.

For the moment evaluation, force is measured with 10 g sensitive load cell (Futek LSB200, USA), calibrated using samples with a weight force equal to 0.59 mN. Solar irradiance was selected using the Sun Simulator SciSun 300 (Sciencetech, Canada). The measurement consists in putting in contact with a suspended load cell the CA-PPL bilayer sample (1 cm × 2 cm,width and length, respectively) and subsequently the actuator is irradiated cyclically (period 15 min with 50% duty cycle) from 0 to 300 mW/cm2. The experiments were performed in ambient air (T = 25 ℃, 30% RH measured using RS PRO RS-325A Digital Hygrometer. According to the experimental setup, the moment was evaluated assuming that M = 1/2 F · L.

Photography.

Photographs were taken using a digital camera (D7100, Nikon, Japan).

Data Analysis.

The normality of data distribution was tested with the Shapiro–Wilk test; normally-distributed data were analyzed with ANOVA followed by LSD post hoc with Bonferroni correction and expressed as average ± standard error. Non-normally distributed data were analyzed with the Kruskal–Wallis test followed by pairwise Wilcox post hoc test with Holm correction and expressed as median ±95% confidence interval. Each experiment has been performed in triplicate (n = 3), if not differently indicated. The data were analyzed and plotted in Origin (Version 2019b, 32bit).

3 Results

3.1 Evaluation of the Photothermal Conversion Efficiency of PPL Blend

We first verified the photothermal properties of PPL blend, comparing the photothermal conversion efficiency of two different samples dispersing in the PCL polymer matrix respectively 7% and 17% w/w of alkaline lignin.

We tested the robustness of the effect changing irradiance with 12.5 min illumination cycle with 50% of duty cycle. Figure 1 shows an increase of temperature due to photothermal effect, when the specimens were subjected to 100 mW/cm2 of simulated solar irradiance. Figure 1A shows that PCL did not present any photothermal properties, while PPL samples with 7% and 17% w/w of lignin reach in both cases a temperature peak of TM = 42.5 ± 0.7 ℃ in τ = 87.6 ± 2.5 s.

Since the samples with different amount of lignin did not show a statistically significant variation (paired t-test p* > 0.05), we selected the PPL solution with lower dispersion grade, because an increase of lignin dispersant in PCL polymeric matrix will make the blend more brittle and it will reduce the processability and the printability [17].

Fig. 1.
figure 1

Photothermal properties of PPL blend at different concentrations of lignin. A) Time behavior of photothermal effect (12.5 min illumination cycle with 50% of duty cycle) under 100 mW/cm2 of simulated solar irradiance input. B) Exponential fit of the temperature variation (R2 = 0.98) used for the evaluation of the photothermal conversion efficiency in 7% w/w lignin sample.

Full size image

Consider the thermal kinetic of a solid body that exchanges heat with the environmental air in natural convection regime, it was possible to evaluate the photothermal conversion efficiency (\(\eta \)) [18, 19] from the exponential fit of the temperature dynamics (Fig. 1B):

$$T\left(t\right)={T}_{0}+\Delta T \left(1-{e}^{-t/\tau }\right)$$
$$\eta = \frac{\Delta T}{\tau } \frac{\rho {c}_{p}h}{{J}_{0}\left(1-{10}^{-\alpha }\right)}$$

where T0 is the initial environmental temperature, ΔT is the temperature variation from the initial to the saturation value, \(\tau \) is the characteristic rising time, ρ the density, cp the specific heat capacity, h the thickness of the sample, J0 the simulated solar irradiance, \(\alpha \) the absorption coefficient.

To evaluate \(\eta \), we assumed that the impinging light is completely absorbed by the PPL (\(\alpha =+\infty \)) and we consider in first approximation that the density and the specific heat capacity are mainly governed by PCL properties (ρ = 1120 kg m−3, cp = 2.2 J g−1 K−1 [20]). Hence, knowing that the sample thickness is h = 0.30 ± 0.01 mm, T, T0 = 25 ℃, ΔT = 17.5 ℃, τ = 87.6 s, the experimental PCE is equal to η = 13.5%.

3.2 Fabrication of the Photothermal Actuator

We proceed with the manufacturing of the bending actuator, FDM printing on a plasma activated cellulose acetate substrate the PPL blend (Fig. 2A). When the actuator is irradiated, the photothermal element will determine a local increase of temperature in the actuator. Subsequently, the PPL layer will anisotropic expand along the printing direction (Fig. 2B) since it is mainly composed by a high thermal expansion coefficient PCL polymer (16 × 10–5 ℃−1 [20]). When the sample is no longer irradiated, the reciprocal effect occurs on the PPL layer and the structure will recover its initial shape (Fig. 2C).

Fig. 2.
figure 2

Sketch and working principle of the photothermal bending actuator. A) Representation of the fabrication process to realize the PPL-CA actuator. In particular, the PPL blend is extruded on a CA substrate with aligned tracks. B) Working principle of the actuator during solar illumination. When the sample is irradiated, there will be an increase of temperature in the whole actuator. Then, the PPL region will anisotropic expand due to the thermal expansion properties of PCL. C) Working principle of the actuator without illumination. Due to the decrease of temperature, the PPL layer shows anisotropic contraction, recovering its initial curvature. D) 3D printed honeycomb sample. Scalebar is 1 cm. E) Two different samples of soft actuators. Scalebar is 2 cm.

Full size image

Figure 2D shows images of FDM printed PPL sample with 7% alkaline lignin. Due to the complete homogenization of lignin in the acetic acid solution, no aggregation was visible in the PPL. We further verified the 3D printing capabilities of the photothermal blend printing a squared structure (1 cm × 1 cm × 3 mm) with honeycomb features (hexagonal side 2.5 mm). Printing planes shows perfect adhesion (Fig. 2D), highlighting the possibility to fabricate complex 3D structures with biodegradable properties and photothermal features. Figure 2E shows two different samples of bending soft actuators fabricated via FDM printing, where the PPL fiber-like structures present a trapezoidal shape, with a maximum thickness of hF = 0.37 ± 0.07 mm and a minimum distance between strands of dF = 1.2 ± 0.2mm. We investigate the effective bending deformation of the structure due to photothermal expansion of the PPL.

3.3 Kinematics and Static of the Photothermal Actuator

In Fig. 3 is reported the curvature of the actuator exposed to 300 mW/cm2 of simulated solar irradiance. The initial curvature in the idle condition was equal to κ0 = 83.3 ± 2.3 m−1 (Fig. 3A). After the illumination, the actuator reaches the thermal equilibrium in 30 ± 2.5 s with a curvature κ1 = 59.0 ± 3.4 m−1 (Fig. 3B). Coming back to the idle state, the curvature reaches the initial value κ2 = 83.3 ± 1.6 m−1 in 30.0 ± 2.5 s (tested on n = 5 different samples). Therefore, the actuator shows a reversible bending behavior mediated by the photothermal effect, showing 25.34% in change of curvature. Considering that cellulose acetate is a hygroscopic material [21], we verified the effective hygroscopic behavior of the actuator measuring the variation of curvature when it is subjected to variation of relative humidity from 30% to 90%. The actuator showed 1% of curvature variation, value that is one order of magnitude lower with respect to the curvature variation induced by the photothermal effect. Hence, we have observed that, in a first approximation, the actuator is predominantly powered by the thermal expansion mechanism.

Fig. 3.
figure 3

Kinematics of photothermal actuator. A) The actuator in its resting position. The initial curvature is related to residual stressed due to fabrication process. Scalebar is 1 cm. B) The actuator subjected to 300 mW/cm2 of simulated solar irradiance. Scalebar is 1 cm.

Full size image

To understand the effective performance of CA-PPL, we performed the static evaluation of the beam moment. Figure 4 shows the time behavior of the bending moment when the actuator is subjected to 300 mW/cm2. The moment reaches a peak of M = 80.2 ± 5.3 µN · m (tested on 5 different trials). As in the curvature case, the moment reaches its saturation value in 30.0 ± 2.5 s. However, the main limitation of the system is evident after 5 cycles of 1 min illumination under an external load (constrained reaction of the load cell), since the PPL layer starts to redistribute the internal stress through a deformation of the entire body, due to the low melting point of the PCL component (TM = 60 ℃ [20]) in the PPL blend.

Fig. 4.
figure 4

Moment of the photothermal actuator as function of time exposed to 300 mW/cm2 of simulated solar irradiance.

Full size image

4 Conclusions

We presented a photothermal bilayer composed by biodegradable polymers, for the creation of a sunlight-driven bending actuator. Thanks to the synergic combination of photothermal properties of lignin and the thermal expansion of polycaprolactone, a bending actuator can be manufactured through versatile 3D printing, involving commonly available materials in the market (e.g., PCL and Pluronic F127) and in nature (e.g., cellulose-based and lignin). Moreover, we verified that the photothermal PPL blend can be 3D printed via FDM method. The actuator shows reversible deformation, featuring a 30 ± 2.5 s response time, associated with a ~25% change in curvature and a moment of 80.2 ± 5.3 µN · m, when samples are exposed to 300 mW/cm2 irradiation. However, the requirement of high solar irradiance implies that when the actuator is subjected to an external load the PPL blend begins to deform plastically after 5 illumination cycles, due to the melting of the polymer. To overcome this problem, a numerical analysis on the thermo-mechanical effects will be necessary to design properly a reliable actuator for environmental applications (i.e., grasping, lifting, locomotion and soil-exploration). A biodegradable and photothermal actuator offers an original and concrete possibility to operate in different scenarios where no contamination of the environment is required, with systems driven by wireless and renewable energy sources.