Quantifying the Shape Memory Performance of a Three-Dimensional-Printed Biobased Polyester/Cellulose Composite Material

Abstract

A biobased composite material with heat-triggered shape memory ability was successfully formulated for three-dimensional (3D) printing. It was produced from cellulose nanocrystals and cellulose micro-powder particles within a bioderived thermally cured polyester matrix based on glycerol, citric acid, and sebacic acid. The effect of curing duration on the material's shape memory behavior was quantified by using two thermo-mechanical approaches to measure recovery: (1) displacement in three-point bending and (2) angular recovery from a beam bent at 90° in a single cantilever setup. Extending curing duration increased the material's glass-transition temperature from −26°C after 6 h to 13°C after 72 h of curing. Fourier-transform infrared spectroscopy confirmed the associated progressive conversion of functional groups consistent with polyester formation. Slow recovery rates and low levels of shape recovery (22–70%) were found for samples cured less than 24 h. Those results also indicated a high dependence on the measurement approach. In contrast, samples cured for 48 and 72 h exhibited faster recovery rates, a significantly higher recovery percentage (90–100%) and were less sensitive to the measurement approach. Results demonstrated that once a sufficient curing threshold was achieved, additional curing time could be used to tune the material glass-transition temperature and create heat-triggered 3D-printed products.

Introduction

Since its inception, four-dimensional (4D) printing has generated a significant amount of interest and enthusiasm in the scientific community.^1,2 Four-dimensional printing focuses on creating dynamic parts with new functionalities within a printed object, such as the ability to change shape over time with the application of a triggering stimulus.^2–6 By combining three-dimensional (3D) printing techniques with smart or active materials, complex objects can be created with intentional responses to stimuli and additional functionality, including dynamic shape changes.^7–9 A range of smart materials already exist, including metal alloys and ceramics; however, in terms of sustainability, biodegradability, cost-effectiveness, lightness, and biocompatibility, polymers are considered promising.¹⁰ Smart or stimuli-responsive polymers can be triggered by several external factors, including temperature,¹¹ humidity,¹² pH, light, and electric and magnetic fields.^13,14 Shape memory polymers (SMPs) are capable of storing a temporary shape and then recovering back to a preprogrammed permanent shape when triggered by an external stimulus.¹⁵ The SMP mechanisms are often regarded as being composed of a minimum of two distinct structural phases. One phase is sensitive to the stimulus and is sometimes referred to as the “switching phase” or “switching segments” (with a stimulus-related transition used to store the temporary shape).¹⁶ The nonsensitive phase is responsible for the recovery from the temporary to the permanent shape. It is described as “netpoints” within an interpenetrating network structure that stores the memory of the permanent shape. These netpoints can be covalent crosslinks in thermosets or crystals in thermoplastics. In the case of the crystals, their melting point must be substantially higher than the shape transition temperatures.^17,18 This network structure allows locking of latent strain energy into the temporary shape that can then be released on exposure to the chosen stimulus (often via stress relaxation of polymer chains back to a higher entropy conformation).^10,19

In the case of heat-triggered stimulus, the transition from a temporary shape to permanent shape occurs at a specific temperature called a transition temperature (T_trans). The most common transitions in polymers are usually a melting temperature (T_m) or a glass transition temperature (T_g).²⁰ Thermally responsive SMPs are first manufactured in their permanent shape. They are then deformed into a temporary shape by applying an external load. This temporary shape is fixed by reversible modification of the chemical or physical structure of the SMP; in the case of thermally responsive SMPs, the material is cooled below T_trans while under strain. The permanent shape is then recovered by applying heat to release the internal constraint.^2,21,22

The quantification of the shape memory ability for thermally responsive SMPs is generally carried out by using thermomechanical cyclic tests. The “fixity ratio” (describes the ability to fix the mechanical deformation) and the “shape recovery ratio” (describes how much the permanent shape is recovered after being loaded, memorized, and then triggered) are determined in these tests.^23,24 The SMPs can find application in a range of areas but remain widely used in biomedical fields such as biodegradable sutures,²⁵ stents,²⁶ and actuators.^10,27

The drive toward sustainability and the need to replace petroleum resources have encouraged recent developments of biobased polymers and composites.²⁸ Biobased polyesters such as poly(glycerol-citrate) have been studied due to their biodegradable and biocompatible characteristics. Halpern et al. described the esterification of glycerol and citric acid for the fabrication of a thermoset polymer used in drug delivery systems.²⁹ They reported that the reaction time, molar ratio, and temperature were important factors in the degree of crosslinking. The reaction of glycerol with sebacic acid produces the thermoset elastomer poly(glycerol-sebacate) (PGS) developed for biomedical and tissue engineering-type applications for cardiac muscle, nerve, cartilage, and retina.^30,31 Cai and Liu reported excellent shape memory behavior for a PGS elastomer describing the 3D network acting as a fixed phase, whereas the amorphous phase provides a reversible switching phase.³²

Cellulose nanocrystals (CNC) have previously been added to thermoplastic matrices to improve shape memory performance,^33,34 and nanoparticles such as CNC are often used to create shear thinning pastes with suitable rheological properties for “direct ink writing”-type 3D printing.^35,36 The 3D printing with thermoset materials typically involves photocuring the formulation during and/or after the printing process.^36–39 The requirements for photocuring place restrictions on formulation design and create difficulties in achieving 100% biobased content in a 3D printable formulation. The shape memory behavior of the CNC composite adapted for 3D printing techniques with a completely biobased thermoset matrix has not yet been explored. A thermoset system based on a combination of citric acid and sebacic acid provides scope to adjust properties of the material by varying the proportion of these monomers while achieving a high biobased content in the material.

This study investigates the influence of postprinting curing duration on shape memory performance of a new biobased 3D-printable composite formulation. Cellulose is used as a rheology modifier during the 3D printing process. In addition, it provides some structural stability to the 3D printing process, whereas the glycerol–citrate–sebacate polyester matrix polymerizes during the postprinting thermal curing process.

Materials and Methods

Paste preparation

The CNC in powder form were purchased from Celluforce™, Canada. Citric acid crystals and glycerol (99.7% USP grade) were purchased from Pure Ingredients Ltd, New Zealand. Sebacic acid granules (99%), Sigmacell microcrystalline cellulose powder (20 type), and analysis-grade ethanol were purchased from Sigma-Aldrich Pty. Ltd, New Zealand. Using an approach adapted from Dorris and Grey,⁴⁰ a thixotropic gel was prepared by dispersing CNC at 5 wt% into a solution of 50 wt% glycerol in distilled water with the subsequent evaporation of water in an oven (80°C, 48 h) used to produce a concentrated CNC-glycerol gel (77 wt% glycerol 14 wt% water, 9 wt% CNC).

Ball-milled citric acid powder (52 g) and ball-milled sebacic acid (52 g) were dry mixed in a beaker with Sigmacell cellulose powder (27 g) before the concentrated CNC-glycerol gel (41 g) was added with further mixing. This formulation corresponds to a molar ratio for the polyester monomers (citric acid: sebacic acid: glycerol) of ∼0.78: 0.74: 1. The beaker of mixed reagents was placed in a sand bath at 160°C, and manual mixing was continued as the organic acids melted for 10 min before the formulation was cooled to room temperature. Ethanol (17 g) was added with further mixing and homogenized (3x1 min, 8000 rpm) by using an Ultra-Turrax^® (IKA, Germany) high shear mixer.

The resulting paste had thixotropic behavior that is suitable for 3D printing with the syringe extrusion process. At the printing stage, this paste has a calculated composition of 28 wt% citric acid, 28 wt% sebacic acid, 18 wt% glycerol, 2 wt% CNC, 15 wt% Sigmacell cellulose powder, and 9 wt% ethanol.

Paste printing

Three-dimensional printing was carried out on a PrintrBot Model 1403 (PrintrBot, Inc.) 3D printer fitted with a syringe printing attachment (Paste & Food Extruder kit). The printer was controlled by using Simplify3D software. The paste was loaded into a 60 mL syringe equipped with a nozzle tip (tapered polypropylene, inner diameter = 1.523 mm; Jensen Global) and maintained at 35°C. The nozzle tip height was adjusted with a G-code offset in the Z-axis when multilayer printing was performed to allow space for extrusion and to prevent damage to the previous layer.

Two simple sample geometries were printed: single strands (70 × 2.50 × 0.70 mm) used for initial dynamic mechanical thermal analysis (DMTA) characterization and large beams (70 × 5.50 × 2 mm) composed of two layers of three adjacent strands printed along the beam length (parallel and anti-parallel) for shape recovery experiments. Single-strand samples were printed at 500 mm/min, and the large beams were printed at 1000 mm/min. Samples were printed onto a polytetrafluoroethylene-coated fiberglass sheet and cured for a series of different durations (6, 16, 24, 48, and 72 h). The samples were placed in an oven at 104°C to evaporate water and drive the polycondensation reaction. After 24 h of curing, a metallic plate was placed on top of the samples to reduce any cure-related warping. An outline skirt surrounding the part was printed before the main beam to help prime the extruder and establish a smooth flow of material. These skirt outlines were used for Fourier transform infrared (FTIR) analysis.

FTIR spectroscopy

FTIR spectroscopy (Bruker Tensor 27, Germany) was used in Attenuated Total Reflectance (ATR, Bruker single bounce diamond cell) mode to monitor chemical variation in the printed material after the set amounts of curing time. All data were recorded at room temperature, in a spectral range of 4000–400 cm⁻¹ at 4 cm⁻¹ resolution with 16 sample and background scans. The spectra were processed with Bruker OPUS software. Spectra were normalized to the highest peak, and a baseline correction was applied. Each curing duration was tested in triplicate, and the average spectra were reported.

Dynamic mechanical thermal analysis

DMTA was carried out to determine the glass transition temperature (T_g) of the formulation after different curing durations. The tests were performed by a dynamic mechanical thermal analyzer (RSA-G2 Solids Analyzer; TA Instruments) using a dual cantilever geometry at a fixed frequency of 1 Hz and a strain of 0.1% from −20°C to 50°C with 5°C/min ramp. An adjustment in the temperature range was required for the samples oven-cured for 6 h that were tested from −50°C to 0°C. The T_g was then determined as the peak of Tan δ. Liquid nitrogen was used for the sub-ambient region. Samples used were single-strand beams, and each curing time was tested in triplicate. The glass transition was determined with TRIOS software v4.3.1.

Shape recovery measurements

Two approaches to measuring shape recovery performance were used: a first one measuring recovery from three-point bending (R(%)_3PB) and a method measuring the angular recovery (R(%)_ang). A method close to the approach of Raasch et al. was developed for measurements in the three-point bending geometry.⁴¹ A span length of 40 mm was used for the three-point bending tests by utilizing a dynamic mechanical thermal analyzer (RSA-G2 Solids Analyzer; TA Instruments). After positioning the dedicated sample across the supports, the temperature was ramped up to T_g+20°C and held for 10 min to homogenize the temperature within the specimen (Step 1). The upper crosshead was then displaced 10 mm downward at a rate of 8.33E-3 mm/s to deform the sample (Step 2). The temperature chamber was then cooled to T_g−20°C and held for 10 min to allow the sample to cool and maintain the temporary shape (Step 3). The instrument load was removed from the center of the beam, and the chamber was finally reheated up to T_g+20°C at a rate of 2°C/min to induce the shape recovery effect (Step 4). An single-lens reflex (SLR) camera (Canon EOS 80D) was used to take photos of the deformed sample (after Step 3) and after the reheating (Step 4). A millimeter-scaled paper reference was placed behind the span to determine the x_deformed and the x_recovered (Fig. 1 and Supplementary Fig. S1) and calculate the recovery percentage R(%)_3PB given by Equation (1),⁴¹ where x_deformed is the imposed deformation of the specimen in mm and x_recovered is the recovered deformation in mm. Triplicate samples were tested for each curing duration investigated.

FIG. 1.

Diagram of the recovery percentage measurement.

R {(%)}_{3 P B} = \frac{x_{d e f o r m e d} - x_{r e c o v e r e d}}{x_{d e f o r m e d}} \times 100

(1)

The angular recovery percentage was measured by following a modified method based on that reported by Wu et al.⁴² and used to calculate an angular recovery rate. The large beam samples were mounted as a single cantilever on a metal support. A protractor was placed behind the support to measure the angular recovery as a function of time. The experiment was carried out in a temperature-controlled cabinet (Excal 10013-HA by Climats, Bordeaux, France). The temperature was increased to T_g+20°C and held for 10 min; then, the sample was manually deformed until reaching a 90° angle. The temperature was then decreased to T_g−20°C at a rate of 2°C/min under strain to fix the temporary deformation. The sample was subsequently reheated to T_g+20°C at a rate of 2°C/min to measure the angular recovery % and recovery rate (against gravity). Triplicate samples for each cure time were tested. Due to equipment capabilities, the 6 h cured samples with low T_g values were not tested. Recovery was recorded in situ with a waterproof-encased GoPro Hero 6, and the evolution of the angle was measured in 30-s intervals by using Acrobat Premiere Pro software. The angular recovery percentages were calculated with Equation (2). The recovery rate, defined as the recovery percentage per minute, was obtained with the derivative curve of the recovery as a function of time. $R {(%)}_{a n g} = \frac{r e c o v e r y a n g l e}{90} \times 100$ (2)

Results and Discussion

Three-dimensional printing and curing observations

As the curing duration increased, a color change was observed as a yellowing of the sample. Yellowing has been previously reported with CNC embedded in thermoplastic matrix composites and was primarily attributed to sulfate groups introduced during CNC production by acid hydrolysis.⁴³ Reactions related to sebacic acid (or impurities with this reagent) may also cause color change via oxidation, as curing was conducted under air (reactions under argon are common in PGS literature^30,31).

FTIR spectroscopy

FTIR spectroscopy was performed in triplicate at each curing duration investigated. Ester formation as a function of curing duration was evaluated by monitoring the carbonyl signal for each sample (Fig. 2a). The range 1790–1610 cm⁻¹ (Fig. 2b) containing C = O stretching signals for both ester and carboxylic acid functional groups was investigated. The signal intensity around 1690 cm⁻¹ (attributed to the carboxylic acid) decreased whereas the peak at 1730 cm⁻¹ (attributed to the ester bond^32,44) increased in relative intensity as the curing duration was increased. As observed in Figure 2b, with the increase of the curing duration a shoulder appears at 1730 cm⁻¹. The ester formation during the curing process is also confirmed by the increase in relative intensity of the signal at 1165 cm⁻¹ attributed to ester C-O stretching (Fig. 2c). The intensity of the signal at 1165 cm⁻¹ became larger than the signal intensity at 1188 cm⁻¹ (carboxylic acid) for the samples cured for at least 16 h, indicating the important role of the curing time for the ester linkage formation.

FIG. 2.

Normalized Fourier transform infrared spectra demonstrating the effect of increasing oven-cure durations; (a) 3900–400 cm −1, (b) C = O stretching region, (c) C-O stretching region.

Determination of the glass transition temperature (T_g) and storage modulus

The DMTA results demonstrated that by extending the curing time from 6 to 72 h, the glass transition temperature gradually increased (Table 1 and Supplementary Fig. S2). A similar trend of increasing T_g with extended cure time was also observed by differential scanning calorimetry (Supplementary Fig. S3). When the material was cured for 6 h, the DMTA-determined T_g was observed at −26°C, rising to −1°C for 24 h and reaching a maximum of 13°C after 72 h of curing. As the curing was extended, a reduction in the rate of T_g increase was noticed and was attributed to the chemical conversion of the available reaction sites as the material polymerizes and crosslinks. This behavior is typical of thermoset curing and has previously been reported in the literature.^45,46

Table 1.

Glass Transition Temperature (T_g) and Storage Modulus (E′) at 23°C for Respective Cure Duration (h) (n = 3 ± SD)

Curing duration (h)	6	16	24	48	72
T_g (°C)	−26.0 ± 0.5	−7.5 ± 1.0	−2.0 ± 0.3	7 ± 0.01	13 ± 0.7
E′ (MPa)	39 ± 32	58 ± 32	57 ± 18	129 ± 14	133 ± 28

In this case, a higher T_g can be attributed to the reaction of hydroxyl groups (on glycerol and potentially cellulose surfaces) with diacid and triacid monomers, leading to polymerization, larger molecular weights in the matrix, and crosslinking into a network structure with the associated loss of mobility within the material. The specimens produced with longer cure durations were significantly stiffer, as indicated by higher storage modulus (Table 1).

Comparing recovery percentages obtained from three-point bending and angular recovery methods

Using the specific T_g for each sample and the method described in the Shape Recovery Measurements section, the recovery ability was tested. The recovery percentage was determined with photographs for the three-point bending method (Supplementary Fig. S1) and measured from video by using the angular recovery method.

The recovery percentage values obtained were dependent on the test method (Fig. 3). The three-point bending method (Fig. 3, diamonds) provided a higher shape recovery percentage values than the angular method (Fig. 3, circles), especially when samples of low cure duration were tested. The recovery percentage was 66% for 16 h-cured material with the three-point bending method, whereas the value dropped to 24% when the angular method was used. This difference is also observed after 24 h curing, with the recovery percentage being 88% as measured by the three-point bending method and 54% when the angular method is used. This difference could be attributed to the way the specimens were constrained, folded, and strained during the two different tests. The angular test deformed the material to a higher level of strain. The larger strain applied is more likely to induce an irreversible plastic deformation, with the samples produced through shorter curing durations appearing more vulnerable to plastic deformation. When samples were cured for 48 and 72 h, the recovery percentage reached a similarly high value regardless of which of the two testing methods were used. This could be attributed to improved crosslinking of the thermoset network, resulting in reduced plastic deformation at higher strains and more reversible viscoelastic behavior after longer cure times. In contrast, samples of low cure duration appeared to yield and/or undergo irreversible viscous flow.

FIG. 3.

Recovery percentage as function of curing time duration. Three-point bending method (diamonds) and angular method (circles) (n = 3 ± SD).

Recovery rate

Results using the angular method indicated that both delays before recovery initiation and the speed of recovery were highly dependent on the curing duration (Fig. 4a).

FIG. 4.

(a) Angular recovery as function of time and (b) derivative of the recovery rate as a function of time.

The angular recovery result shows a plateau over 90% for the 48 and 72 h, in contrast with 50% and 20% for the 24 and 16 h cure, respectively. The samples cured for 16 and 24 h indicated delayed recovery, as clearly seen when plotting the derivative of the recovery rate as a function of time (Fig. 4b). The times to reach maximum recovery rate were found at, respectively, ∼660, 630, 1000, and 1250 s for the 72, 48, 24, and 16 h cure. The samples cured for 48 and 72 h exhibited faster recovery with higher maximum recovery rates. High recovery percentage and rates are likely related to more extensive covalent crosslinking expected in samples with cure durations exceeding 48 h.

Those findings suggest losses through viscous flow caused by a lack of “netpoints,” that is, crosslinks below a 48 h cure. The similar rate of recovery characteristics (i.e., amplitude, time to achieve maximum rate) for 48 and 72 h cure suggested that despite a difference in glass transition temperature, a critical threshold of “netpoints” was created, allowing for a quasi-full recovery. Beyond the 48 h curing time, little improvement appeared to be gained from increasing the quantity of “netpoints” with further curing. Scanning electron microscopy was performed to assess potential differences (Supplementary Fig. S4). Matrix domains were observed surrounding the cellulose particles that appeared more obvious past 24 h of curing duration.

Similar results have been previously reported by Yakacki et al. in their work. They highlighted the dependency of the recovery rate on both the T_g and the crosslinking density for the tert-butyl acrylate and poly(ethylene glycol) dimethacrylate network systems studied. By varying the weight percent of crosslinker in their formulations, they demonstrated that increasing the crosslinking density resulted in SMPs, which achieved faster recovery.⁴⁷ Shape recovery times for SMPs vary with polymer chemistry and testing conditions from a few seconds to tens of minutes.^39,48,49 In the literature, recovery of SMPs is often measured or demonstrated at one specific temperature, usually chosen to be 10–30°C above the polymers T_g, and this often leads to rapid initiation of the transition.^48,49 In contrast, the experiment in Figure 4a involved the gradual heating of the sample at a rate of 2°C/min from a starting temperature 20°C below the materials T_g, leading to a slower recovery than a sample directly placed in a chamber at 20°C. Recovery rate plotted against chamber temperature showed that for the samples cured for 48 and 72 h the maximum recovery rates occur a few degrees above the samples T_g (Supplementary Fig. S5). These findings are consistent with Ge et al., who demonstrated very slow and partial recovery in their photocured-based SMP material when heated at a temperature below T_g. However, much faster recovery occurred when the material was heated above its T_g.³⁹

Further discussion and printing of the materials into more complex structures

To demonstrate the printing and shape memory behavior on a more complex structure, a tree-like design was printed (Fig. 5a). The part was cured for 48 h into the permanent shape (Fig. 5b), deformed into a curved temporary shape, and held in this shape during cooling at −20°C. This Figure 5c represents the transition from the temporary shape to the permanent shape under heating with an incandescent bulb. (A video of this shape memory transition is included as Supplemental Video S1)

FIG. 5.

(a) Paste printing of basic tree-like design, (b) after 48 h curing at 104°C, and (c) heat-driven shape recovery effect on the tree-like design.

A citric acid, sebacic acid, and glycerol thermoset system provides scope for tuning mechanical properties and glass-transition-related trigger temperature for specific shape memory applications. As demonstrated in this investigation, cure duration can be used to adjust T_g once sufficient crosslinking has formed for good shape recovery performance. Altering the molar ratio between sebacic acid and citric acid provides further scope to adjust material performance. With each citric acid molecule providing more functional groups compared with the bifunctional sebacic acid, it facilitates additional crosslinking during polyester formation. Formulations with lower sebacic/citric molar ratios cure into stiffer materials with higher T_g values, and this will be explored in the future. Cellulose addition (both nanoparticles and microparticles) was primarily used both to provide suitable formulation rheology for “printability” (using this model of printer) and to provide shape stability postprinting and during the oven cure process. Three-dimensional printing thermoset formulations and light-curing with radical polymerization is a common approach.^3,37,50 In this work, an alternative approach based on extrusion of matrix precursors and then using a postprinting thermal curing reaction to fix the permanent shape was demonstrated. Importantly, the two approaches (thermal curing compared with photocuring) place different constraints and limitations on the design of formulations and process. In this case, the formulation could be designed to be composed of biobased and readily bio-derived reagents, without the need for the addition of photo-initiators or designing the formulation around mitigating issues such as oxygen inhibition of free-radical crosslinking reactions.⁵⁰ However, cellulose additives were needed to control rheology and maintain the shape after printing and during the curing reaction.

Conclusion

Four-dimensional printed cellulose composites displaying memory shape behavior on heat stimulus were successfully synthesized. The shape memory ability was quantified and indicated up to 98% recovery. The dependence of this material's shape memory behavior was correlated to the curing duration that influenced the crosslinking density and the T_g of the composites. The T_g of the thermoset increased from −26°C to 13°C as the curing duration increased from 6 to 72 h. The crosslinking reaction was studied by FTIR spectroscopy, confirming that the presence of carboxyl groups decreased as the curing duration was increased with a corresponding increase in the intensity of bands associated with ester linkages. The shape recovery percentages were quantified by using a three-point bending geometry and an alternative single-cantilever approach where beams were bent to a 90° angle as the programmed temporary shape. The ability of the material to recover increased with increased curing duration.

Three-point bending recovery percentage increased from 37% to a maximum of 98% when the curing time was increased from 6 to 72 h. Samples cured for less than 48 h had much lower recovery percentages when tested after bending through a 90° angle, indicating low recovery ability at higher strain levels in these inadequately crosslinked materials. Little difference was found between samples cured at 48 and 72 h in terms of recovery percentage and recovery rate with their recovery percentages in the 90–100% range regardless of which of the two testing approaches were used. However, there was a substantial increase in glass transition temperature between samples cured at 48 and 72 h, which could enable the targeting of specific trigger temperatures for future applications requiring smart biobased materials with shape memory behaviors.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

The authors would like to acknowledge funding from the New Zealand National Science Challenge: Science for Technological Innovation—Kia kotahi mai—Te Ao Pūtaiao me Te Ao Hangarau as part of the Spearhead Project: “Additive manufacturing and 3D and/or 4D printing of bio-composites” Grant number 2019_55_CRS 4D printing, and the Scion Strategic Science Investment Fund, Dr. Angelique Greene, Dr. Kelly Wade, Mr. Robin Parr, and Mrs. Beatrix Theobald.

Supplementary Material

References

Tibbits

4D printing: Multi-material shape change. Archit Des, 2014; 84:116–121.

Momeni

, M. Mehdi

Hassani

.N S, Liu

, et al. A review of 4D printing. Mater Des, 2017; 122:42–79.

Bakarich

, Gorkin

, Panhuis

MIH

, et al. 4D printing with mechanically robust, thermally actuating hydrogels. Macromol Rapid Commun, 2015; 36:1211–1217.

, Qi

, Dunn

. Active materials by four-dimension printing. Appl Phys Lett, 2013; 103:131901.

Pei

4D Printing: Dawn of an emerging technology cycle. Assembly Autom, 2014; 34:310–314.

Ren

, Li

, Song

, et al. Bioinspired fiber-regulated composite with tunable permanent shape and shape memory properties via 3d magnetic printing. Compos B Eng, 2019; 164:458–466.

Lui

, Sow

, Tan

, et al. 4D printing and stimuli-responsive materials in biomedical aspects. Acta Biomater, 2019; 92:19–36.

Khoo

, Teoh

JEM

, Liu

, et al. 3D printing of smart materials: A review on recent progresses in 4D printing. Virtual Phys Prototyp, 2015; 10:103–122.

Tamay

, Dursun Usal

, Alagoz

, et al. 3D and 4d printing of polymers for tissue engineering applications. Front Bioeng Biotechnol, 2019; 7:164.

10.

Liu

, Qin

, Mather

. Review of progress in shape-memory polymers. J Mater Chem, 2007; 17:1543–1558.

11.

Liu

, Gall

, Dunn

, et al. Thermomechanics of shape memory polymer nanocomposites. Mech Mater, 2004; 36:929–940.

12.

Huang

, Yang

, An

, et al. Water-driven programmable polyurethane shape memory polymer: Demonstration and mechanism. Appl Phys Lett, 2005; 86:114105.

13.

Leng

, Lan

, Liu

, et al. Shape-memory polymers and their composites: Stimulus methods and applications. Prog Mater Sci, 2011; 56:1077–1135.

14.

Wang

, Lu

, Shi

, et al. A thermomechanical model of multi-shape memory effect for amorphous polymer with tunable segment compositions. Compos B Eng, 2019; 160:298–305.

15.

Behl

, Lendlein

. Shape-memory polymers. Mater Today, 2007; 10:20–28.

16.

Behl

, Lendlein

. Actively moving polymers. Soft Matter, 2007; 3:58–67.

17.

Hager

, Bode

, Weber

, et al. Shape memory polymers: Past, present and future developments. Prog Polym Sci, 2015; 49:3–33.

18.

Pretsch

Review on the functional determinants and durability of shape memory polymers. Polymers, 2010; 2:120–158.

19.

Andreas Lendlein

SK.

Shape-memory effect. Angew Chem Int Ed, 2002; 41:2034–2057.

20.

. Thermal Responsive Shape Memory Polymers for Biomedical

Applications

. 2011. https://escholarship.umassmed.edu/ortho_pp/96. Accessed August 16, 2020.

21.

Lendlein

, Gould

OEC

. Reprogrammable recovery and actuation behaviour of shape-memory polymers. Nat Rev Mater, 2019; 4:116–133.

22.

Auad

, Contos

, Nutt

, et al. Characterization of nanocellulose- reinforced shape memory polyurethanes. Polym Int, 2008; 57:651–659.

23.

Wagermaier

, Kratz

, Heuchel

, et al. Characterization methods for shape-memory polymers. In: Shape-Memory Polymers. Switzerland: Springer Nature; 2009, pp. 97–145.

24.

Chatterjee

, Dey

, Nando

, et al. Thermo-responsive shape memory polymer blends based on alpha olefin and ethylene propylene diene rubber. Polymer (Guildf), 2015; 78:180–192.

25.

Lendlein

, Langer

. Biodegradable, elastic shape-memory polymers for potential biomedical applications. Science, 2002; 296:1673–1676.

26.

Gall

, Yakacki

, Liu

, et al. Thermomechanics of the shape memory effect in polymers for biomedical applications. J Biomed Mater Res A, 2005; 73:339–348.

27.

Maitland

, Metzger

, Schumann

, et al. Photothermal properties of shape memory polymer micro-actuators for treating stroke. Lasers Surg Med, 2002; 30:1–11.

28.

Badouard

, Traon

, Denoual

, et al. Exploring mechanical properties of fully compostable flax reinforced composite filaments for 3D printing applications. Ind Crops Prod, 2019; 135:246–250.

29.

Halpern

, Urbanski

, Weinstock

, et al. A biodegradable thermoset polymer made by esterification of citric acid and glycerol. J Biomed Mater Res A, 2014; 102:1467–1477.

30.

Loh

, Abdul Karim

, Owh

. Poly(glycerol sebacate) biomaterial: Synthesis and biomedical applications. J Mater Chem B, 2015; 3:7641–7652.

31.

Rai

, Tallawi

, Grigore

, et al. Synthesis, properties and biomedical applications of poly(glycerol sebacate) (PGS): A review. Prog Polym Sci, 2012; 37:1051–1078.

32.

Cai

, Liu

. Shape-memory effect of poly (glycerol–sebacate) elastomer. Mater Lett, 2008; 62:2171–2173.

33.

Saralegi

, Gonzalez

, Valea

, et al. The role of cellulose nanocrystals in the improvement of the shape-memory properties of castor oil-based segmented thermoplastic polyurethanes. Compos Sci Technol, 2014; 92:27–33.

34.

Cao

, Liu

, Zou

, et al. Using cellulose nanocrystals as sustainable additive to enhance mechanical and shape memory properties of PLA/ENR thermoplastic vulcanizates. Carbohydr Polym, 2020; 230:115618.

35.

, Dunn

, Zhang

, et al. Direct ink write (DIW) 3D printed cellulose nanocrystal aerogel structures. Sci Rep, 2017; 7:8018.

36.

Siqueira

, Kokkinis

, Libanori

, et al. Cellulose nanocrystal inks for 3D printing of textured cellular architectures. Adv Funct Mater, 2017; 27:1604619.

37.

Yeh

Y-C

, Ouyang

, Highley

, et al. Norbornene-modified poly(glycerol sebacate) as a photocurable and biodegradable elastomer. Polym Chem, 2017; 8:5091–5099.

38.

Kretzschmar

, Lipponen

, Klar

, et al. Mechanical properties of ultraviolet-assisted paste extrusion and postextrusion ultraviolet-curing of three-dimensional printed biocomposites. 3D Print Addit Manuf, 2019; 6:127–137.

39.

, Sakhaei

, Lee

, et al. Multimaterial 4D printing with tailorable shape memory polymers. Sci Rep, 2016; 6:31110.

40.

Dorris

, Gray

. Gelation of cellulose nanocrystal suspensions in glycerol. Cellulose, 2012; 19:687–694.

41.

Raasch

, Ivey

, Aldrich

, et al. Characterization of polyurethane shape memory polymer processed by material extrusion additive manufacturing. Addit Manuf, 2015; 8:132–141.

42.

, Liu

, Leng

Study on shape recovery speed of SMP, SMP composite and SMP foam. Paper presented at: Proceedings of SPIE—The International Society for Optical Engineering 2008; San Diego, CA.

43.

Sapkota

, Natterodt

, Shirole

, et al. Fabrication and properties of polyethylene/cellulose nanocrystal composites. Macromol Mater Eng, 2017; 302:1600300.

44.

Meng

, Yoo

, Li

. Physicochemical structural changes of cellulosic substrates during enzymatic saccharificatio. J Appl Biotechnol Bioeng, 2016; 1:87–94.

45.

, Vatanparast

, Vuorimaa

, et al. Curing kinetics and glass-transition temperature of hexamethylene diisocyanate-based polyurethane. J Polym Sci B Polym Phys, 2000; 38:2213–2220.

46.

Wisanrakkit

, Gillham

. The glass transition temperature (Tg) as an index of chemical conversion for a hight-Tg amine/epoxy system: Chemical and diffusion-controlled reaction kinetics. J Appl Polym Sci, 1990; 41:2885–2929.

47.

Yakacki

, Shandas

, Lanning

, et al. Unconstrained recovery characterization of shape-memory polymer networks for cardiovascular applications. Biomaterials, 2007; 28:2255–2263.

48.

Alteheld

, Feng

, Kelch

, et al. Biodegradable, amorphous copolyester-urethane networks having shape-memory properties. Angew Chem Int Ed Engl, 2005; 44:1188–1192.

49.

Shumaker

, McClung

AJW

, Baur

. Synthesis of high temperature polyaspartimide-urea based shape memory polymers. Polymer (Guildf), 2012; 53:4637–4642.

50.

Gladman

, Matsumoto

, Nuzzo

, et al. Biomimetic 4D printing. Nat Mater, 2016; 15:413–418.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

33.41 MB

0.42 MB

0.07 MB

0.03 MB

0.76 MB

0.09 MB

Quantifying the Shape Memory Performance of a Three-Dimensional-Printed Biobased Polyester/Cellulose Composite Material

Abstract

Introduction

Materials and Methods

Paste preparation

Paste printing

FTIR spectroscopy

Dynamic mechanical thermal analysis

Shape recovery measurements

Results and Discussion

Three-dimensional printing and curing observations

FTIR spectroscopy

Determination of the glass transition temperature (Tg) and storage modulus

Comparing recovery percentages obtained from three-point bending and angular recovery methods

Recovery rate

Further discussion and printing of the materials into more complex structures

Conclusion

Footnotes

Author Disclosure Statement

Funding Information

Supplementary Material

References

Supplementary Material

Determination of the glass transition temperature (T_g) and storage modulus