- Email: [email protected]
High-strain shape memory polymers as practical dry adhesives
International Journal of Adhesion and Adhesives 81 (2018) 74–78
Contents lists available at ScienceDirect
International Journal of Adhesion and Adhesives journal homepage: www.elsevier.com/locate/ijadhadh
High-strain shape memory polymers as practical dry adhesives Jeffrey Eisenhaure, Seok Kim
⁎
T
Department of Mechanical Science and Engineering, University of Illinois, 1206 W Green St, Urbana, IL 61801, United States
A B S T R A C T The adhesive benefits of incorporating shape memory polymers (SMPs) in various dry adhesive designs have been well documented in recent years. Chemically-crosslinked SMPs can possess remarkable shape fixity and recovery ability, chemical and thermal stability, a sharp glass transition, and other desirable qualities. However, low failure strain and a relatively rigid rubbery state limit their performance as a dry adhesive component in many practical applications. This work explores the adhesive benefits of using an epoxy-based SMP which has been modified to possess a softer rubbery state and a greatly increased maximum strain before failure. One of the modified SMPs forms the active adhesive surface of a prototype consumer level device.
1. Introduction Substantial research in the area of dry adhesives has been conducted in recent years, yielding many significant discoveries and producing a wide variety of strategies for improving adhesive performance [1]. Despite the steady progress, notable limitations remain in the current state of the art examples which have thus far limited their use. The use of shape memory polymers (SMPs) as a functional material in dry adhesives has been previously explored [2–7], where its dynamic rigidity and ability to passively recover its shape were found to provide enhanced adhesive strength, reduce the required preload to bond, and enable very easy removal when desired (good reversibility). Whether as a structural layer or as a direct-contact adhesive, the current body of work primarily utilizes epoxy-based SMPs which possess many desirable characteristics including excellent shape fixity and recovery, thermo-mechanical stability, and ease of processing. Less desirable are their tendencies to remain stiff in their rubbery state relative to most pressure sensitive adhesives (PSAs), and their susceptibility to tearing under moderate strains. A soft rubbery state is desirable to facilitate adhesive bonding by reducing the required preload, while large recoverable strains are desirable to enable greater flexibility in the adhesives’ structural design and to improve general durability. A recent study has demonstrated a successful modification of an oftused epoxy SMP formula, wherein the authors used a heavier (longer) epoxy monomer and greatly increased the concentration of the curing, or crosslinking, agent [8]. The resulting SMP was found to be both softer in its rubbery state and substantially more stretchable, while retaining its superlative shape memory properties. The differences in the SMP's properties are directly attributed to a decrease in crosslinking
⁎
Corresponding author. E-mail address: [email protected] (S. Kim).
https://doi.org/10.1016/j.ijadhadh.2017.11.008 Accepted 10 November 2017 Available online 21 November 2017 0143-7496/ © 2017 Elsevier Ltd. All rights reserved.
density, which allows the polymer chains greater freedom of movement in the polymer's rubbery state. The chains therefore may rearrange more readily when subjected to stress, aligning in the direction of the stress more completely and with less resistance. The present work seeks to investigate the effects of these modifications to the epoxy SMP formula regarding dry adhesive performance, and to produce a prototype practical dry adhesive using similar formulations described herein. The effect of increased curing agent and increased monomer chain length are investigated independently through the production of several SMP formulations. These SMPs are tested with regard to their dry adhesive performance, and to confirm that they possess excellent mechanical and shape memory properties on par with those previously reported, for example in Ref. [8]. 2. Materials and methods 2.1. SMP materials and sample fabrication The SMPs used in this work were created from a mixture of epoxy monomer and curing agent Jeffamine D230 (poly(propylene glycol)bis (2-aminopropyl) ether; Huntsman), hereafter referred to as Jeffamine. In general, the epoxy monomer was comprised of a mixture of two chemically similar products with different average molecular weights: EPON 826 obtained from Momentive (molecular weight, Mn ~ 362 g mol−1), and Poly(Bisphenol A-co-epichlorohydrin), glycidyl endcapped obtained from Sigma-Aldrich (molecular weight, Mn ~ 1075 g mol−1), hereafter referred to as E1075. Intermediate epoxy monomer weights, determined on a molar basis, were created by mixing EPON 826 and E1075 in ratios provided in Table 1. The
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Table 1 Component mixing proportions for the high-strain SMPs. Formulation
Epoxy Mn (g/mol)
Molar proportion
Weight proportion
EPON 826
E1075
Jeffamine
EPON 826
E1075
Jeffamine
E362-J100 E362-J075 E362-J060 E362-J050
361.9
1.000
–
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
– – – –
0.636 0.478 0.381 0.318
E448-J100 E448-J075 E448-J060 E448-J050
447.6
0.899
0.101
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
0.334 0.334 0.334 0.334
0.707 0.531 0.423 0.353
E533-J100 E533-J075 E533-J060 E533-J050
533.3
0.760
0.240
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
0.940 0.940 0.940 0.940
0.837 0.629 0.501 0.418
regulating system, increasing pressure by ~ 70 kPa min−1 until interfacial failure was observed to have initiated.
proportion of Jeffamine used for each SMP formulation is likewise included in Table 1, on both molar and weight bases. Formulation names are derived from the average molecular weight of the epoxy monomer mixture, followed by the ratio of Jeffamine to epoxy monomer mixture. E1075 and EPON 826 were combined in a glass container, heated to 120 °C in order to completely melt the E1075, then thoroughly mixed by manual stirring while still at 120 °C. Room temperature Jeffamine was then stirred into the epoxy mixture. Test samples for dynamic mechanical analysis (DMA) and adhesive blister testing were created by pouring the resulting mixture onto glass plates and curing for several minutes at 100 °C until the precursor becomes viscous enough to spread into a uniform and thin sheet. Curing was completed in an oven at 100 °C for 90 min. Samples were cut from the resulting sheets using a laser cutter, then removed from the glass at elevated temperature. The formula for NGDE2 has been previously described [9], while in this work the curing procedure and sample preparation are the same as described above for the high-strain formulations.
2.4. Practical adhesive demonstration Nichrome wire with 0.62 mm diameter was obtained from Sigma Aldrich. 3D printed parts were fabricated using an EOS Formiga P100 selective laser sintering machine, using PA 2200 polyamide powder. The power source is a 7.4 V lithium polymer battery. 3. Results and discussion Plots of storage modulus versus temperature for each of the SMP formulations listed in Table 1 are provided in Fig. 1. The data indicate the effects of increasing Jeffamine concentration are very similar to those reported in Ref. [8], decreasing the glass transition temperature (Tg) and rubbery state storage modulus (Er) while maintaining a similar glassy state storage modulus. Comparing the E448-J050 and E448-J100 formulations, for example, shows a decrease in Er by a factor of eight from 16 MPa to 2 MPa by doubling the Jeffamine concentration. These observations indicate that the expected reduction in crosslinking density was achieved. The flat rubbery and glassy plateaus, together with the narrow Tg band, suggest the presence of strong shape memory performance in each of the twelve formulations. Similarly to Ref. [8], changing the proportion of Jeffamine has a strong effect on the material's Tg, particularly as the concentration becomes large. This is clearly evident in E362 and E448 series tests where for example E362-J050 and E362-J060 are nearly the same, whereas E362-J060, E362-J075, and J362-J100 are substantially different, though significant and nearly uniform changes to Tg are observed between each variant of the E533 series of SMP. Varying the average molecular weight of the epoxy base has little effect on the shape and character of the modulus curves, except a noticeable reduction in rubbery state modulus for the higherweight formulations. Among those formulations with the highest Jeffamine concentration, E362-J100, E448-J100, and E533-J100, the storage modulus decreases with increasing epoxy base molecular weight from 4 MPa to 2 MPa and 1.7 MPa, respectively. The shape memory performance of E533-J100 was tested, with results depicted in Fig. 2a. Both fixity and recovery factors of > 99% were found for a 6% peak strain, showing that even the formulation with the lowest crosslinking density shows excellent shape memory properties. Recovery at larger strains was tested using E448-J100. The first ten cycles of ~ 100% strain with temperature held constant at 50 °C shown in Fig. 2b. The first cycle appears to show some irrecoverable strain, which is partially due to the slower, more viscous response of the SMP at relatively low temperature not allowing complete recovery during
2.2. DMA testing A Q800 DMA from TA Instruments (New Castle, DE) was used to perform all SMP mechanical analysis, failure strain measurements, and SMP fixity and recovery factor measurements. Laser-cut samples of rectangular cross section approximately 1.75 mm wide and 0.5 mm thick were used. Storage modulus data is reported at 1 Hz excitation, with data collection at discrete temperature intervals where the chamber was allowed to equilibrate for four minutes. Failure strain data was collected in uniaxial tension by ramping tensile force on each sample at a rate of 1 N min−1 until failure. Engineering stress and strain are reported, based on the test sample's initial length and cross sectional area. Shape fixity and recovery factors were determined according to the definitions provided in Ref. [10] for a single shape fixing and recovery cycle. 2.3. Blister adhesion testing Thin (~ 0.50 mm) circular samples of several SMP formulations were bonded to form a seal over the 1 cm diameter circular glass opening of a pressure chamber by first heating the glass surface to 75 °C, followed by light pressure to ensure SMP-glass contact and maintaining 75 °C for three minutes. Temperature was controlled by an AGPtek® universal digital PID temperature controller. Tests were performed after allowing the sample to cool to the desired temperature. The flexibility of the thin SMP membranes allowed them to self-conform to the glass adherend when heated, obviating the need for preload during the bonding process. Pressure was applied through a manual 75
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Fig. 1. The SMP storage modulus as a function of temperature for formulations with base epoxy average Mn of 362 g/mol (a), 448 g/mol (b), and 533 g/mol (c).
the cycle time. Some slippage or deformation of material at the tensile grips as they sink into the SMP during its first cycle may additionally account for some apparent irrecoverable strain. Stress versus strain curves for uniaxial tension tests of E448-J100 are shown in Fig. 2c for three temperatures corresponding to the upper end of its Tg transition (50 °C, 60 °C, and 70 °C). In agreement with previous studies, the strain at failure or break, ɛb, is greater when deforming the SMP within its glass transition and show similar characteristic shapes as those provided in Ref. [8]. Data for ɛb was collected for each of the three temperatures for E363-J100, E448-J100, and E533-J100 SMP formulations, and compared against the previouslydeveloped NGDE2 formulation which has seen significant use in dry adhesive studies [2–7]. Each of these formulations has a similar Tg in terms of both onset and width of transition, and therefore direct comparison of results at each temperature is reasonably fair. Fig. 2d shows the data, presented as the average of five tests for each formulation at each temperature. Error bars indicate a 95% confidence interval for the true mean based on Student's t-test of the available sample data. The results reinforce the observation that ɛb increases when tested at temperatures within each SMP's glass transition. An enhancement in ɛb is also clearly apparent for all three newer formulations over the more highly-crosslinked NGDE2. The role of increased Jeffamine with respect to ɛb is evident through a comparison of NGDE2 with those of the other three formulations of Fig. 2d. NGDE2 is expected to have greatly increased crosslink density compared with each of the other formulations, due in small part to the
Fig. 2. A shape memory cycle of E533-J100 (a). Cyclic strain and recovery of E448-J100 at 50 °C (b). Typical uniaxial stress-strain curves at selected temperatures for E448-J100 (c). Failure strain results for four SMP formulations (d). Error bars represent 95% Student's t-test confidence intervals.
shorter average chain length of its precursor epoxy monomers [9], but more significantly due to the disparity in Jeffamine. Comparing the measured average ɛb for each at 50 °C (NGDE2, 58%: E362-J100, 240%), 60 °C (NGDE2, 30%: E362-J100, 89%), and 70 °C (NGDE2, 17%: E362-J100, 49%) shows a three to four-fold increase. The effect of the epoxy base average molecular weight on ɛb is directly comparable across the three -J100 series formulations, and clearly indicates that a longer base monomer provides an enhancement, particularly when each SMP is in its rubbery state. The trend is less clear at 50 °C, given the surprisingly large measured ɛb of the E362J100 formulation which may be attributable to the fact that the glass transitions of each formulation are similar but not identical, and thus 50 °C may happen to be nearer to the optimal ɛb temperature of E362J100 than that of the other formulations. At 70 °C the difference in ɛb between the three newer formulations is substantial at 49%, 65%, and 133% for E362-J100, E448-J100, and E533-J100 respectively. It is worthwhile to compare results for our E448-J100 to those of the EPON1 formula of Ref. [8], since they are identical in all respects except that we achieve the target average molecular weight by mixing epoxy bases of dissimilar weights. The comparison indicates a reduced ɛb at 50 °C and 70 °C for E448-J100, but a similar result at 60 °C. The differences 76
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Fig. 3. The temperature-controlled blister test apparatus is shown (a). A cross section schematic of the blister substrate and pressure chamber (b). Blister test pressure results (c).
Fig. 4. The steps to fabricate a practical SMP adhesive are depicted (a). The completed adhesive unit is rendered usable by attaching a hook, which fits together with the attachment piece (b). Using this system, a loaded backpack is suspended by the SMP adhesive hook assembly on a glass door (c), a curved plastic refrigerator door (d), a wooden door (e), and a powdercoated metal door (f).
enhanced adhesive strength, as the SMP's temperature is reduced through its glass transition. This trend has been previously observed and attributed primarily to the increase in storage modulus and corresponding inhibition of crack propagation [2,4]. The failure pressure is also influenced by the SMP formula, with greater monomer weight yielding enhanced adhesion. We speculate that this result is due to enhanced energy dissipation of the higher-weight formulation, particularly at lower temperatures. This theory is reinforced by observing that the measured loss modulus increases significantly with increased
may be attributable to dissimilar testing equipment and procedures, or may suggest a difference in the polymer's structure and behavior due to the bimodal distribution of monomer length in E448-J100. The blister test apparatus shown in Fig. 3a, b was constructed to quantitatively measure the relative adhesive performance of the highstrain SMP formulations including E363-J100, E448-J100, and E533J100. Tests were conducted with each SMP in its glassy state (25 °C), rubbery state (75 °C), and at two intermediate temperatures (Fig. 3c). The results clearly show an increase in failure pressure, indicating 77
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Improved adhesive performance may be the result of better conformation during the bonding process, owing to the reduced rubbery storage modulus. Reduced crosslink density may also enhance dissipative losses within the material. An internally-heated dry adhesive system was developed using one of the new SMP formulations, demonstrating strong adhesive performance on a variety of chemically dissimilar materials. Continued refinement of this system, including its materials and construction, is sure to improve performance further and may produce a class of effective consumer products.
monomer weight (231 MPa, 299 MPa, and 534 MPa for E362-J100, E448-J100 and E533-J100 respectively at their peaks ~ 25 °C). The maximum pressure the test apparatus can measure is 410 kPa, which was achieved for 25 °C tests of both E448-J100 and E533-J100 specimens. A prototype practical dry adhesive system was fabricated as shown in Fig. 4a, comprised of a block of E448-J100 with an embedded nichrome wire heating element secured to a 3D-printed attachment piece which allows for a firm connection to a separate hook. The adhesive is bonded by first briefly heating the SMP with an attached power source (battery) and pressing it by hand against an adherend for several seconds as it cools. The hook then slides into place. It is shown supporting a backpack, loaded with approximately 3 kg of items, on multiple adherends with varying surface characteristics. The adhesive may be easily removed by peeling, once the SMP is re-heated to its rubbery state. As a dry adhesive relying primarily on van der Waals adhesive forces, rather than a flowing glue layer, it is inherently reusable. When stored at room temperature, the glassy SMP is rigid and non-tacky, additionally improving its resistance to fouling from the environment.
References [1] Eisenhaure J, Kim S. A review of the state of dry adhesives: biomimetic structures and the alternative designs they inspire. Micromachines 2017;8:125. [2] Eisenhaure JD, Xie T, Varghese S, Kim S. Microstructured shape memory polymer surfaces with reversible dry adhesion. ACS Appl Mater Interfaces 2013;5:7714–7. [3] Eisenhaure J, Kim S. An internally heated shape memory polymer dry adhesive. Polymers 2014;6:2274–86. [4] Eisenhaure JD, Rhee SI, Al-Okaily AM, Carlson A, Ferreira PM, Kim S. The use of shape memory polymers for MEMS assembly. J Microelectromech Syst 2016;25:69–77. [5] Eisenhaure J, Kim S. Laser-driven shape memory effect for transfer printing combining parallelism with individual object control. Adv Mater Technol 2016;1. [1600098–n/a]. [6] Kim S, Sitti M, Xie T, Xiao X. Reversible dry micro-fibrillar adhesives with thermally controllable adhesion. Soft Matter 2009;5:3689–93. [7] Xie T, Xiao X. Self-peeling reversible dry adhesive system. Chem Mater 2008;20:2866–8. [8] Zheng N, Fang G, Cao Z, Zhao Q, Xie T. High strain epoxy shape memory polymer. Polym Chem 2015;6:3046–53. [9] Xie T, Rousseau IA. Facile tailoring of thermal transition temperatures of epoxy shape memory polymers. Polymer 2009;50:1852–6. [10] Xie T. Recent advances in polymer shape memory. Polymer 2011;52:4985–5000.
4. Conclusions Modifying a previously reported high-strain SMP by mixing epoxy monomers of differing molecular weights was successful, yielding similar high-strain performance and allowing for an additional degree of control over the final product's properties by varying the effective monomer weight. A reduction in crosslink density, both by increasing concentrations of Jeffamine and by increasing the monomer weight, enhanced the resulting SMP's failure strain and adhesive properties.
78
Contents lists available at ScienceDirect
International Journal of Adhesion and Adhesives journal homepage: www.elsevier.com/locate/ijadhadh
High-strain shape memory polymers as practical dry adhesives Jeffrey Eisenhaure, Seok Kim
⁎
T
Department of Mechanical Science and Engineering, University of Illinois, 1206 W Green St, Urbana, IL 61801, United States
A B S T R A C T The adhesive benefits of incorporating shape memory polymers (SMPs) in various dry adhesive designs have been well documented in recent years. Chemically-crosslinked SMPs can possess remarkable shape fixity and recovery ability, chemical and thermal stability, a sharp glass transition, and other desirable qualities. However, low failure strain and a relatively rigid rubbery state limit their performance as a dry adhesive component in many practical applications. This work explores the adhesive benefits of using an epoxy-based SMP which has been modified to possess a softer rubbery state and a greatly increased maximum strain before failure. One of the modified SMPs forms the active adhesive surface of a prototype consumer level device.
1. Introduction Substantial research in the area of dry adhesives has been conducted in recent years, yielding many significant discoveries and producing a wide variety of strategies for improving adhesive performance [1]. Despite the steady progress, notable limitations remain in the current state of the art examples which have thus far limited their use. The use of shape memory polymers (SMPs) as a functional material in dry adhesives has been previously explored [2–7], where its dynamic rigidity and ability to passively recover its shape were found to provide enhanced adhesive strength, reduce the required preload to bond, and enable very easy removal when desired (good reversibility). Whether as a structural layer or as a direct-contact adhesive, the current body of work primarily utilizes epoxy-based SMPs which possess many desirable characteristics including excellent shape fixity and recovery, thermo-mechanical stability, and ease of processing. Less desirable are their tendencies to remain stiff in their rubbery state relative to most pressure sensitive adhesives (PSAs), and their susceptibility to tearing under moderate strains. A soft rubbery state is desirable to facilitate adhesive bonding by reducing the required preload, while large recoverable strains are desirable to enable greater flexibility in the adhesives’ structural design and to improve general durability. A recent study has demonstrated a successful modification of an oftused epoxy SMP formula, wherein the authors used a heavier (longer) epoxy monomer and greatly increased the concentration of the curing, or crosslinking, agent [8]. The resulting SMP was found to be both softer in its rubbery state and substantially more stretchable, while retaining its superlative shape memory properties. The differences in the SMP's properties are directly attributed to a decrease in crosslinking
⁎
Corresponding author. E-mail address: [email protected] (S. Kim).
https://doi.org/10.1016/j.ijadhadh.2017.11.008 Accepted 10 November 2017 Available online 21 November 2017 0143-7496/ © 2017 Elsevier Ltd. All rights reserved.
density, which allows the polymer chains greater freedom of movement in the polymer's rubbery state. The chains therefore may rearrange more readily when subjected to stress, aligning in the direction of the stress more completely and with less resistance. The present work seeks to investigate the effects of these modifications to the epoxy SMP formula regarding dry adhesive performance, and to produce a prototype practical dry adhesive using similar formulations described herein. The effect of increased curing agent and increased monomer chain length are investigated independently through the production of several SMP formulations. These SMPs are tested with regard to their dry adhesive performance, and to confirm that they possess excellent mechanical and shape memory properties on par with those previously reported, for example in Ref. [8]. 2. Materials and methods 2.1. SMP materials and sample fabrication The SMPs used in this work were created from a mixture of epoxy monomer and curing agent Jeffamine D230 (poly(propylene glycol)bis (2-aminopropyl) ether; Huntsman), hereafter referred to as Jeffamine. In general, the epoxy monomer was comprised of a mixture of two chemically similar products with different average molecular weights: EPON 826 obtained from Momentive (molecular weight, Mn ~ 362 g mol−1), and Poly(Bisphenol A-co-epichlorohydrin), glycidyl endcapped obtained from Sigma-Aldrich (molecular weight, Mn ~ 1075 g mol−1), hereafter referred to as E1075. Intermediate epoxy monomer weights, determined on a molar basis, were created by mixing EPON 826 and E1075 in ratios provided in Table 1. The
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Table 1 Component mixing proportions for the high-strain SMPs. Formulation
Epoxy Mn (g/mol)
Molar proportion
Weight proportion
EPON 826
E1075
Jeffamine
EPON 826
E1075
Jeffamine
E362-J100 E362-J075 E362-J060 E362-J050
361.9
1.000
–
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
– – – –
0.636 0.478 0.381 0.318
E448-J100 E448-J075 E448-J060 E448-J050
447.6
0.899
0.101
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
0.334 0.334 0.334 0.334
0.707 0.531 0.423 0.353
E533-J100 E533-J075 E533-J060 E533-J050
533.3
0.760
0.240
1.000 0.752 0.599 0.500
1.000 1.000 1.000 1.000
0.940 0.940 0.940 0.940
0.837 0.629 0.501 0.418
regulating system, increasing pressure by ~ 70 kPa min−1 until interfacial failure was observed to have initiated.
proportion of Jeffamine used for each SMP formulation is likewise included in Table 1, on both molar and weight bases. Formulation names are derived from the average molecular weight of the epoxy monomer mixture, followed by the ratio of Jeffamine to epoxy monomer mixture. E1075 and EPON 826 were combined in a glass container, heated to 120 °C in order to completely melt the E1075, then thoroughly mixed by manual stirring while still at 120 °C. Room temperature Jeffamine was then stirred into the epoxy mixture. Test samples for dynamic mechanical analysis (DMA) and adhesive blister testing were created by pouring the resulting mixture onto glass plates and curing for several minutes at 100 °C until the precursor becomes viscous enough to spread into a uniform and thin sheet. Curing was completed in an oven at 100 °C for 90 min. Samples were cut from the resulting sheets using a laser cutter, then removed from the glass at elevated temperature. The formula for NGDE2 has been previously described [9], while in this work the curing procedure and sample preparation are the same as described above for the high-strain formulations.
2.4. Practical adhesive demonstration Nichrome wire with 0.62 mm diameter was obtained from Sigma Aldrich. 3D printed parts were fabricated using an EOS Formiga P100 selective laser sintering machine, using PA 2200 polyamide powder. The power source is a 7.4 V lithium polymer battery. 3. Results and discussion Plots of storage modulus versus temperature for each of the SMP formulations listed in Table 1 are provided in Fig. 1. The data indicate the effects of increasing Jeffamine concentration are very similar to those reported in Ref. [8], decreasing the glass transition temperature (Tg) and rubbery state storage modulus (Er) while maintaining a similar glassy state storage modulus. Comparing the E448-J050 and E448-J100 formulations, for example, shows a decrease in Er by a factor of eight from 16 MPa to 2 MPa by doubling the Jeffamine concentration. These observations indicate that the expected reduction in crosslinking density was achieved. The flat rubbery and glassy plateaus, together with the narrow Tg band, suggest the presence of strong shape memory performance in each of the twelve formulations. Similarly to Ref. [8], changing the proportion of Jeffamine has a strong effect on the material's Tg, particularly as the concentration becomes large. This is clearly evident in E362 and E448 series tests where for example E362-J050 and E362-J060 are nearly the same, whereas E362-J060, E362-J075, and J362-J100 are substantially different, though significant and nearly uniform changes to Tg are observed between each variant of the E533 series of SMP. Varying the average molecular weight of the epoxy base has little effect on the shape and character of the modulus curves, except a noticeable reduction in rubbery state modulus for the higherweight formulations. Among those formulations with the highest Jeffamine concentration, E362-J100, E448-J100, and E533-J100, the storage modulus decreases with increasing epoxy base molecular weight from 4 MPa to 2 MPa and 1.7 MPa, respectively. The shape memory performance of E533-J100 was tested, with results depicted in Fig. 2a. Both fixity and recovery factors of > 99% were found for a 6% peak strain, showing that even the formulation with the lowest crosslinking density shows excellent shape memory properties. Recovery at larger strains was tested using E448-J100. The first ten cycles of ~ 100% strain with temperature held constant at 50 °C shown in Fig. 2b. The first cycle appears to show some irrecoverable strain, which is partially due to the slower, more viscous response of the SMP at relatively low temperature not allowing complete recovery during
2.2. DMA testing A Q800 DMA from TA Instruments (New Castle, DE) was used to perform all SMP mechanical analysis, failure strain measurements, and SMP fixity and recovery factor measurements. Laser-cut samples of rectangular cross section approximately 1.75 mm wide and 0.5 mm thick were used. Storage modulus data is reported at 1 Hz excitation, with data collection at discrete temperature intervals where the chamber was allowed to equilibrate for four minutes. Failure strain data was collected in uniaxial tension by ramping tensile force on each sample at a rate of 1 N min−1 until failure. Engineering stress and strain are reported, based on the test sample's initial length and cross sectional area. Shape fixity and recovery factors were determined according to the definitions provided in Ref. [10] for a single shape fixing and recovery cycle. 2.3. Blister adhesion testing Thin (~ 0.50 mm) circular samples of several SMP formulations were bonded to form a seal over the 1 cm diameter circular glass opening of a pressure chamber by first heating the glass surface to 75 °C, followed by light pressure to ensure SMP-glass contact and maintaining 75 °C for three minutes. Temperature was controlled by an AGPtek® universal digital PID temperature controller. Tests were performed after allowing the sample to cool to the desired temperature. The flexibility of the thin SMP membranes allowed them to self-conform to the glass adherend when heated, obviating the need for preload during the bonding process. Pressure was applied through a manual 75
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Fig. 1. The SMP storage modulus as a function of temperature for formulations with base epoxy average Mn of 362 g/mol (a), 448 g/mol (b), and 533 g/mol (c).
the cycle time. Some slippage or deformation of material at the tensile grips as they sink into the SMP during its first cycle may additionally account for some apparent irrecoverable strain. Stress versus strain curves for uniaxial tension tests of E448-J100 are shown in Fig. 2c for three temperatures corresponding to the upper end of its Tg transition (50 °C, 60 °C, and 70 °C). In agreement with previous studies, the strain at failure or break, ɛb, is greater when deforming the SMP within its glass transition and show similar characteristic shapes as those provided in Ref. [8]. Data for ɛb was collected for each of the three temperatures for E363-J100, E448-J100, and E533-J100 SMP formulations, and compared against the previouslydeveloped NGDE2 formulation which has seen significant use in dry adhesive studies [2–7]. Each of these formulations has a similar Tg in terms of both onset and width of transition, and therefore direct comparison of results at each temperature is reasonably fair. Fig. 2d shows the data, presented as the average of five tests for each formulation at each temperature. Error bars indicate a 95% confidence interval for the true mean based on Student's t-test of the available sample data. The results reinforce the observation that ɛb increases when tested at temperatures within each SMP's glass transition. An enhancement in ɛb is also clearly apparent for all three newer formulations over the more highly-crosslinked NGDE2. The role of increased Jeffamine with respect to ɛb is evident through a comparison of NGDE2 with those of the other three formulations of Fig. 2d. NGDE2 is expected to have greatly increased crosslink density compared with each of the other formulations, due in small part to the
Fig. 2. A shape memory cycle of E533-J100 (a). Cyclic strain and recovery of E448-J100 at 50 °C (b). Typical uniaxial stress-strain curves at selected temperatures for E448-J100 (c). Failure strain results for four SMP formulations (d). Error bars represent 95% Student's t-test confidence intervals.
shorter average chain length of its precursor epoxy monomers [9], but more significantly due to the disparity in Jeffamine. Comparing the measured average ɛb for each at 50 °C (NGDE2, 58%: E362-J100, 240%), 60 °C (NGDE2, 30%: E362-J100, 89%), and 70 °C (NGDE2, 17%: E362-J100, 49%) shows a three to four-fold increase. The effect of the epoxy base average molecular weight on ɛb is directly comparable across the three -J100 series formulations, and clearly indicates that a longer base monomer provides an enhancement, particularly when each SMP is in its rubbery state. The trend is less clear at 50 °C, given the surprisingly large measured ɛb of the E362J100 formulation which may be attributable to the fact that the glass transitions of each formulation are similar but not identical, and thus 50 °C may happen to be nearer to the optimal ɛb temperature of E362J100 than that of the other formulations. At 70 °C the difference in ɛb between the three newer formulations is substantial at 49%, 65%, and 133% for E362-J100, E448-J100, and E533-J100 respectively. It is worthwhile to compare results for our E448-J100 to those of the EPON1 formula of Ref. [8], since they are identical in all respects except that we achieve the target average molecular weight by mixing epoxy bases of dissimilar weights. The comparison indicates a reduced ɛb at 50 °C and 70 °C for E448-J100, but a similar result at 60 °C. The differences 76
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Fig. 3. The temperature-controlled blister test apparatus is shown (a). A cross section schematic of the blister substrate and pressure chamber (b). Blister test pressure results (c).
Fig. 4. The steps to fabricate a practical SMP adhesive are depicted (a). The completed adhesive unit is rendered usable by attaching a hook, which fits together with the attachment piece (b). Using this system, a loaded backpack is suspended by the SMP adhesive hook assembly on a glass door (c), a curved plastic refrigerator door (d), a wooden door (e), and a powdercoated metal door (f).
enhanced adhesive strength, as the SMP's temperature is reduced through its glass transition. This trend has been previously observed and attributed primarily to the increase in storage modulus and corresponding inhibition of crack propagation [2,4]. The failure pressure is also influenced by the SMP formula, with greater monomer weight yielding enhanced adhesion. We speculate that this result is due to enhanced energy dissipation of the higher-weight formulation, particularly at lower temperatures. This theory is reinforced by observing that the measured loss modulus increases significantly with increased
may be attributable to dissimilar testing equipment and procedures, or may suggest a difference in the polymer's structure and behavior due to the bimodal distribution of monomer length in E448-J100. The blister test apparatus shown in Fig. 3a, b was constructed to quantitatively measure the relative adhesive performance of the highstrain SMP formulations including E363-J100, E448-J100, and E533J100. Tests were conducted with each SMP in its glassy state (25 °C), rubbery state (75 °C), and at two intermediate temperatures (Fig. 3c). The results clearly show an increase in failure pressure, indicating 77
International Journal of Adhesion and Adhesives 81 (2018) 74–78
J. Eisenhaure, S. Kim
Improved adhesive performance may be the result of better conformation during the bonding process, owing to the reduced rubbery storage modulus. Reduced crosslink density may also enhance dissipative losses within the material. An internally-heated dry adhesive system was developed using one of the new SMP formulations, demonstrating strong adhesive performance on a variety of chemically dissimilar materials. Continued refinement of this system, including its materials and construction, is sure to improve performance further and may produce a class of effective consumer products.
monomer weight (231 MPa, 299 MPa, and 534 MPa for E362-J100, E448-J100 and E533-J100 respectively at their peaks ~ 25 °C). The maximum pressure the test apparatus can measure is 410 kPa, which was achieved for 25 °C tests of both E448-J100 and E533-J100 specimens. A prototype practical dry adhesive system was fabricated as shown in Fig. 4a, comprised of a block of E448-J100 with an embedded nichrome wire heating element secured to a 3D-printed attachment piece which allows for a firm connection to a separate hook. The adhesive is bonded by first briefly heating the SMP with an attached power source (battery) and pressing it by hand against an adherend for several seconds as it cools. The hook then slides into place. It is shown supporting a backpack, loaded with approximately 3 kg of items, on multiple adherends with varying surface characteristics. The adhesive may be easily removed by peeling, once the SMP is re-heated to its rubbery state. As a dry adhesive relying primarily on van der Waals adhesive forces, rather than a flowing glue layer, it is inherently reusable. When stored at room temperature, the glassy SMP is rigid and non-tacky, additionally improving its resistance to fouling from the environment.
References [1] Eisenhaure J, Kim S. A review of the state of dry adhesives: biomimetic structures and the alternative designs they inspire. Micromachines 2017;8:125. [2] Eisenhaure JD, Xie T, Varghese S, Kim S. Microstructured shape memory polymer surfaces with reversible dry adhesion. ACS Appl Mater Interfaces 2013;5:7714–7. [3] Eisenhaure J, Kim S. An internally heated shape memory polymer dry adhesive. Polymers 2014;6:2274–86. [4] Eisenhaure JD, Rhee SI, Al-Okaily AM, Carlson A, Ferreira PM, Kim S. The use of shape memory polymers for MEMS assembly. J Microelectromech Syst 2016;25:69–77. [5] Eisenhaure J, Kim S. Laser-driven shape memory effect for transfer printing combining parallelism with individual object control. Adv Mater Technol 2016;1. [1600098–n/a]. [6] Kim S, Sitti M, Xie T, Xiao X. Reversible dry micro-fibrillar adhesives with thermally controllable adhesion. Soft Matter 2009;5:3689–93. [7] Xie T, Xiao X. Self-peeling reversible dry adhesive system. Chem Mater 2008;20:2866–8. [8] Zheng N, Fang G, Cao Z, Zhao Q, Xie T. High strain epoxy shape memory polymer. Polym Chem 2015;6:3046–53. [9] Xie T, Rousseau IA. Facile tailoring of thermal transition temperatures of epoxy shape memory polymers. Polymer 2009;50:1852–6. [10] Xie T. Recent advances in polymer shape memory. Polymer 2011;52:4985–5000.
4. Conclusions Modifying a previously reported high-strain SMP by mixing epoxy monomers of differing molecular weights was successful, yielding similar high-strain performance and allowing for an additional degree of control over the final product's properties by varying the effective monomer weight. A reduction in crosslink density, both by increasing concentrations of Jeffamine and by increasing the monomer weight, enhanced the resulting SMP's failure strain and adhesive properties.
78