Effects of network structures on the tensile toughness of copper-catalyzed azide-alkyne cycloaddition (CuAAC)-based photopolymers

Abstract

In the present study, the photo-initiated copper-catalyzed azide-alkyne cycloaddition (CuAAC) polymerization was utilized to form structurally diverse glassy polymer networks. Systematic alterations in the monomer backbone rigidity (e.g., cyclic or aliphatic groups with a different length of backbone) and the reactive functional group density (e.g., tetra-, tri-, di-, and mono-functional azide and alkyne monomers) were used to provide readily tailorable network structures with crosslink densities (estimated from the rubbery modulus) varying by a factor of over 20. All eight of the resultant networks exhibited glass transition temperatures (T_g) between 50 and 80 °C with tensile toughness ranging from 28 to 61 MJ m⁻³. A nearly linear dependence of yield stress and elongation at break (broadly defined as strength and ductility, respectively) on the T_g and rubbery modulus was established in these triazole networks. When a flexible di-alkyne monomer (5 carbon spacing between alkynes) was incorporated in a network composed of a tri-alkyne and di-azide monomer, the elongation at break was improved from 166 to 300 %, while the yield stress was reduced from 36 to 23 MPa. Additionally, the polymer ductility was also varied by incorporating mono-functional azides as chain ends in the network - replacing a sterically hindered stiff mono-azide with a more flexible mono-azide increased the elongation at break from 24 to 185 % and the tensile toughness from 6 to 28 MJ m⁻³.

Graphical Abstract

INTRODUCTION

Photopolymerization is a versatile technique relevant in a wide array of practical applications such as coatings, adhesives, dental composites, and 3D printed parts.^1–4 Conventional photopolymers used in these applications are predominantly based on free radical chain-growth polymerizations of multi-functional (meth)acrylate monomers and oligomers, which upon incident exposure of light, generate densely crosslinked glassy networks with high mechanical strength.^4–6 However, the high degree of permanent crosslinking in these networks often induces brittleness and reduced toughness,^5,7 and this attribute can potentially lead to premature mechanical failure. With the emergence of additive manufacturing technologies, a search for readily photopolymerizable, low viscosity resins which can be nearly instantly transformed into glassy polymers with high toughness and tailorable material properties has gained substantial interest in the photopolymer community.⁸ Although, there exist many tough, glassy polymers such as linear polycarbonates,^9,10 polyurethane copolymer networks,¹¹ and ring-opening metathesis polymer (ROMP)-based networks/composites,^12–14 their lack of in situ placement and polymerization limits their implementation in many applications.

Several strategies for enhancing mechanical performance, particularly, yielding behavior, tensile or fracture toughness, and impact strength of crosslinked (meth)acrylate or epoxy-based glassy networks have been suggested by various researchers. As one approach, the incorporation of micro- or nano-sized high modulus fillers (e.g., inorganic particles,¹⁵ platelets,¹⁶ and fibres¹⁷) or the addition of prepolymer blends (e.g., linear/block polymers,^18–20 elastomers,²¹ and core-shell nanoparticles^22,23) in epoxy or acrylate-based polymer networks enhanced toughness via crack deflection, cavitation, or debonding of the additives. Alternatively, modifications of the chemistry in methacrylate-based glassy networks to tailor the overall network structures or induce stress relaxation have been proposed to enhance toughness. For instance, thiol-ene/yne step-growth monomers^24,25 were incorporated in methacrylate resins to regulate network homogeneity via delayed gelation, and this approach resulted in more compliant networks with improved toughness. Additionally, the presence of dynamic covalent chemistries such as irreversible addition-fragmentation chain transfer (AFCT) agents (e.g., allyl sulfone or vinyl sulfone ester)^26–28 in the resin matrix or thiol-thioester exchange (TTE)²⁹ moieties at the surface of inorganic nanoparticles improved toughness of glassy crosslinked methacrylate networks and composites via a stress relaxation mechanism.

In recent years, step-growth polymerizations based on “click” chemistries have been employed to form photopolymer networks due to the highly efficient nature of the “click” reactions that are orthogonal to various other functionalities.^30,31 This approach allows the network to form ideally without significant side products or non-stoichiometric consumption of the polymerizable functional groups. In addition, contrary to chain-growth polymers, the step-growth mechanism generates more uniform polymer network structures and allows a greater tailorability in monomer backbone functionalities and reactive/reacted functional group densities. This consequently leads to more precise control over polymer network morphologies and allows systematic investigation of the polymer structure-property relationships.

In addition, several thiol-ene³² based photopolymers utilizing a photo-“click” reaction have been developed which feature high ductility and toughness enabled by semi-crystallinity of the polymer chains,^33,34 or combined effects of chain mobility, degree of crosslinking, and backbone functionality.^35–37 As another type of photo-“click” polymer, the copper-catalyzed azide-alkyne cycloaddition (CuAAC)^38–40 photopolymerization has been shown to yield glassy polymer networks with enhanced mechanical properties^41–43 and high glass transition temperatures^44,45 resulting from relatively low viscosity resins, potentially due to the formation of triazole structures in high concentration as the product of the CuAAC reaction. In our previous study,⁴³ we demonstrated superior tensile toughness and high ductility achieved in glassy photopolymer networks made via photoinitiated CuAAC polymerizations. Surprisingly, these triazole-based networks exhibited multiple shape memory recovery transitions after large-scale tensile deformation in the glassy state, and these polymers also maintained excellent mechanical properties (e.g., Young’s modulus, yield stress, tensile toughness, and elongation) over multiple deformation-recovery cycles. From this study,⁴³ it was suggested that in spite of the highly crosslinked structure and glassy state of the network, the triazole moieties in the network could actively improve the tensile toughness when the network had a sufficient number of freely rotatable bonds between the network junctions.

Inspired by the previous work,^43,44,46 the effect of structural variations in the azide and alkyne monomers on the tensile behavior of the glassy polymer networks formed via CuAAC photopolymerizations was further explored in the present work. Various multi-functional azide and alkyne monomers having different numbers of reactive functional groups ranging from mono-functional to tetra-functional monomers were used to form diverse glassy photopolymer networks with varying degrees of crosslinking. Systematic variations in the monomer backbone rigidity and the number of reactive functional groups in the monomers were shown to improve the tensile toughness and ductility of the polymer networks while maintaining a relatively high glass transition temperature.

MATERIALS AND METHODS

Materials.

1,3-bis(2-isocyanatopropan-2-yl)benzene, 6-chloro-1-hexanol, dibutyltin dilaurate, 4,4-methylenebis(cyclohexyl isocyanate), 1,9-nonanediol, pentaerythritol, phloroglucinol, propargyl bromide, sodium azide, 1,1,1-tris(hydroxymethyl)propane, tetrabutylammonium iodide (TBAI), copper(II) chloride, N,N,N′,N′,N′′-pentamethyldiethylenetriamine (PMDETA), 2,2-dimethoxy-2-phenylacetophenone (DMPA), tetrahydrofuran, and acetonitrile were purchased from Sigma Aldrich and used without further purification. 6-Chloro-1-hexyne, 2,2,4,4-tetramethyl-1,3-cyclobutanediol, potassium carbonate, potassium hydroxide, sodium hydroxide, sodium sulfate, dimethylformamide, dimethyl sulfoxide, and methanol were purchased from Fisher Scientific and used without further purification. Two azide monomers, 1-azidoadamantane (Stiff-Az₁) and azidomethyl pivalate (Flexible-Az₁), and one alkyne monomer, 1,8-nonadiyne (5C-Ak₂), were purchased from Sigma Aldrich and used without further purification. Two azide monomers, bis(6-azidohexyl) (1,3-phenylenebis(propane-2,2-diyl))dicarbamate (Bz-Az₂) and bis(6-azidohexyl) (methylenebis(cyclohexane-4,1-diyl))dicarbamate (Ch-Az₂), and four alkyne monomers, 1-(prop-2-yn-1-yloxy)-2,2-bis((prop-2-yn-1-yloxy)methyl)butane (Al-Ak₃), 1,3,5-tris(hex-5-yn-1-yloxy)benzene (Ar-Ak₃), 3-(3-(prop-2-yn-1-yloxy)-2,2-bis((prop-2-yn-1-yloxy)methyl)propoxy)prop-1-yne (Al-Ak₄), and 1,1,3,3-tetramethyl-2,4-bis(prop-2-yn-1-yloxy)cyclobutane (CB-Ak₂), were synthesized according to previously reported procedures.^44–46 The diyne monomer, 1,9-bis(prop-2-yn-1-yloxy)nonane (11C-Ak₂), was newly synthesized based on the methods employed with the other alkyne monomers. Detailed synthetic procedures and NMR structural analysis of all synthesized monomers are provided in the Supporting Information. All azides were synthesized according to the azide safety rules and handled with appropriate precaution when working with monomers, resins, and polymers in small quantities.^30,47 It should be noted that all crosslinked networks investigated in this study were composed of a stoichiometric ratio of azide groups and alkyne groups. For each polymer network, the coding is given by the acronym of X-Az_y + X-Ak_y, where X = the backbone structure in the multi-functional monomer or the side group in the mono-functional monomer, Az = azide monomer, Ak = alkyne monomer, and y = monomer functionality. When more than one azide or alkyne monomers were used to formulate the networks, acronyms of X₁-Az_y1 + X₂-Az_y2 + X₃-Ak_y3 for the use of two azide monomers and X₁-Az_y1 + X₂-Ak_y2 + X₃-Ak_y3 for the use of two alkyne monomers were utilized to describe the networks.

Sample preparation.

Each CuAAC monomer mixture was composed of a stoichiometric molar ratio of 1 : 1 azide groups : alkyne groups with 1 mole % CuCl₂[PMDETA] and 2 mole % DMPA per azide or alkyne functional group. Scheme 1A shows a generalization of the components used to prepare the networks. Various structures of di-functional alkynes or mono-functional azides were incorporated in a multi-functional resin mixture composed of di-azides and tri-alkynes to systematically reduce the crosslink density of the polymer network. Specifically, a molar ratio per functional group of 1 : 0.5 : 0.5 with di-azide : tri-alkyne : di-alkyne was utilized to form tri/di-functional monomer mixtures, and a molar ratio per functional group of 0.63 : 0.37 : 1 with di-azide : mono-azide : tri-alkyne was used to form mono/poly-functional monomer mixtures. A minimal amount of methanol was first added to homogenize the resin mixture and to dissolve solid monomers. Then the solution mixture was filtered through a 0.45 μm PTFE syringe filter to remove any dust, and the methanol was evaporated in vacuo prior to photopolymerization. Each resin mixture was placed between two glass slides using a 0.25 mm spacer and photopolymerized using 10 mW cm⁻² of UV irradiation (λ_max = 366 nm) for 5 min at ambient temperature followed by postcuring in an oven at 70–80 °C for 24–48 hours. After post-cure, the thin polymer film (0.25 mm in thickness) was cut into a bar at elevated temperatures at around T_g for dynamic mechanical thermal analysis (DMTA) and a dogbone for tensile testing. Two consecutive temperature cycles from 0 to 150 °C were performed on each post-cured DMTA specimen. The close overlay of the tan δ curves between the two thermal cycles for each formulation indicated that all networks contained minimal residual (methanol) solvent and had reached quantitative conversion or the topological limit prior to tensile testing (see Figures S9, S15, and S26–S27 for multi-functional, di-functional, and mono-functional formulations, respectively). To erase any thermal history and physical aging behavior of the polymers,⁴³ each specimen was heated above the T_g for 1 minute followed by 3 minutes of cooling at ambient temperature prior to mechanical testing.

Scheme 1.

(A) A reaction diagram for the photo-initiated CuAAC polymerization to form a triazole-based polymer using DMPA as a photoinitiator and CuCl₂[PMDETA] as a copper catalyst. (B) A schematic illustration of three different CuAAC resin formulations composed of structurally various multi-functional azide and alkyne monomers to alter the network structure and crosslink density of the polymer.

¹H-NMR and ¹³C-NMR spectroscopy.

The spectra were recorded on a Bruker 400 MHz NMR spectrometer. Proton chemical shifts are expressed in parts per million (δ). The δ scale was referenced to deuterated solvents, as indicated for the respective measurements.

Tensile testing.

An MTS Exceed E42 universal testing machine was used with a 500 N load cell to give the room temperature engineering stress-strain curve, the Young’s modulus (determined from the initial linear elastic region of the stress-strain curve), the yield stress (the stress at the maximum), the elongation at break, and the toughness (as measured from the area under the stress-strain curve). Dogbone samples (similar to the ASTM dogbone die D638-V⁴⁸ except for a 15 mm gage length used in the study with a 3.15 mm width and a 0.25 mm thickness) were cut at elevated temperatures above T_g. The specimens were clamped in the grip areas and tested under uniaxial tensile loading at a crosshead rate of 0.75 mm min⁻¹ except for Figure S20 where a crosshead rate of 7.5 mm min⁻¹ was also employed on di-functional monomer systems to compare the effect of a crosshead rate on the tensile testing behavior. The stress was calculated from the applied force divided by the original cross-sectional area of the gage section, while the strain was determined from the ratio of the crosshead displacement to the gage length. With regard to the potential effect of instrument compliance on the tensile measurements, no extensometers were required in this testing due to the thin form of the specimens and the fact that all specimens were tested under identical conditions.

Dynamic mechanical thermal analysis.

A TA Instruments Q800 DMA was utilized in tension mode at a frequency of 1 Hz, using a heating rate of 3 °C min⁻¹ from a minimum temperature of 0 °C to a maximum temperature of 150 °C, to yield the storage (E’), loss moduli (E”), and the tan δ. The glass transition temperature (T_g) was taken to be the temperature at the peak of the tan δ curve. The DMTA specimens were approximately 0.25 mm in thickness, 3 mm in width, and 6 mm in length between grips.

RESULTS AND DISCUSSION

As illustrated in Scheme 1, three different network structures were investigated which are briefly described as follows: (1) highly crosslinked networks composed of a molar ratio per functional groups of 1 : 1 for di-functional azide : tri- or tetra-functional alkyne monomers, (2) less crosslinked networks containing a molar ratio per functional groups of 1 : 0.5 : 0.5 for di-azide : tri-alkyne : di-alkyne monomers, and (3) less densely crosslinked networks composed of a molar ratio per functional group of 0.63 : 0.37 : 1 for di-azide : mono-azide : tri-alkyne monomers. In addition to varying the numbers of reactive functional groups in both azide and alkyne monomers as described above, several monomers contained varying chemical backbone structures (e.g., aliphatic hydrocarbon in comparison with cyclic or aromatic functional groups in the monomer backbone) to study their effect on the tensile properties. These systematic variations in both monomer backbone structure and the number of reactive functional groups in the monomer yielded tailorable network structures (Scheme 1B) with variable crosslink density, which subsequently influences the glass transition temperature and tensile properties including modulus, yield stress, tensile toughness, and elongation at break.

In order to compare the effect of monomer structures on tensile properties, multi-functional azide and alkyne monomers having differing monomer functionality and backbone structure were synthesized (see monomers in Figure 1A). Specifically, three different alkyne monomers, an aliphatic tri-alkyne (Al-Ak₃), aromatic tri-alkyne (Ar-Ak₃), and aliphatic tetra-alkyne (Al-Ak₄), were synthesized and copolymerized with the same di-azide monomer (Bz-Az₂), and their properties were studied. As a contrast, two different azide monomers, a benzene di-azide (Bz-Az₂) and cyclohexane di-azide (Ch-Az₂) were synthesized and copolymerized with the same tri-alkyne monomer (Al-Ak₃). From the DMTA data in Figure 1B, an increase in reactive functionality in the alkyne monomer from tri-functional (Al-Ak₃) to tetra-functional (Al-Ak₄) resulted in a more highly crosslinked network as shown from an increase in rubbery modulus⁴⁹ and an approximately 10 °C increase in glass transition temperature (T_g) from 65 to 75 °C (see in Figure S8 and Table S1). In addition, when compared with the Bz-Az₂ + Al-Ak₃ network, the more highly crosslinked Bz-Az₂ + Al-Ak₄ network led to an increase in yield stress from 36 to 43 MPa and a drastic decrease in ductility was observed with a two-fold reduction in elongation at break from 160 to 80 %, resulting in a lowering of the tensile toughness from 53 to 30 MJ m⁻³ (Figure 1C and Table S1).

Figure 1.

(A) Structures of two different azide monomers (Bz-Az₂, Ch-Az₂) and three different alkyne monomers (Ar-Ak₃, Al-Ak₃, Al-Ak₄). The four crosslinked polymers (denoted as Bz-Az₂ + Al-Ak₃, Bz-Az₂ + Ar-Ak₃, Bz-Az₂ + Al-Ak₄, and Ch-Az₂ + Al-Ak₃) were composed of a molar ratio per functional group of 1 : 1 : 0.01 : 0.02 (azide from di-azide : alkyne from tri-alkyne : CuCl₂[PMDETA] : DMPA). Each mixture was irradiated at room temperature using 10 mW cm⁻² of UV light (λ_max = 366 nm) for 5 min followed by post cure at 70–80 °C for 24–48 hours. (B) Thermo-mechanical properties of four photopolymer systems as measured by DMTA (for storage modulus, see Figure S8). It should be noted that DMTA data for Bz-Az₂ + Al-Ak₃ and Ch-Az₂ + Al-Ak₃ was previously published (Baranek et al., 2016)⁴⁴ and is repeated here for comparative purposes. (C) Tensile testing of the four photopolymer systems at a crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) as measured via the MTS instrument (for replicates, see Figure S10–S13). Each specimen was heated above the T_g for 1 minute followed by 3 minutes of cooling at ambient temperature prior to mechanical testing. It should be noted that tensile testing data of Bz-Az₂ + Al-Ak₃ was previously published (Song and Baranek et al., 2018)⁴³ and is repeated here for comparative purposes.

In addition to altering the crosslink density, changes in the rigidity of the monomer backbone structure also influences the tensile properties (Figure 1). For instance, the Ch-Az₂ azide monomer composed of a stiff di-cyclohexane core yielded a network with higher T_g and yield stress, and a lower elongation at break, when compared with the more flexible mono-benzene core Bz-Az₂ azide monomer (Figure 1B and 1C, Table S1). Similarly, the network containing the Ar-Ak₃ alkyne monomer, which had a longer aliphatic hydrocarbon backbone, exhibited a slightly more ductile behavior with lower yield stress and higher elongation at break when compared with that for the Al-Ak₃ alkyne monomer (Figure 1C), while exhibiting a very similar T_g (Figure 1B) and a slightly lower crosslink density (Table S1). From these comparisons of networks formed from structurally different azide and alkyne monomers, it is concluded that yield stress, ductility (determined by elongation at break values), and glass transition temperature of the polymer network were tunable based on monomer structural rigidity and crosslink density of the network structures, as expected. Despite the large variations in both glass transition temperature and elongation at break among the systems shown in Figure 1, all of the networks yielded both high tensile toughness between 30 and 60 MJ m⁻³ and high glass transition temperature significantly above ambient ranging from 60 to 80 °C (Table S1).

In an attempt to improve further the ductility of the glassy polymer networks, di-functional alkyne monomers (see Figure 2A) were incorporated in a network composed of Bz-Az₂ di-azide and Al-Ak₃ tri-alkyne monomers to reduce the crosslink density. Specifically, 50 mol% of the alkyne functional groups provided by Al-Ak₃ were replaced by the alkyne groups of one of three di-functional alkynes (CB-Ak₂, 5C-Ak₂, or 11C-Ak₂) to form a stoichiometric resin mixture with Bz-Az₂. These three di-functional alkyne monomers either contained a stiff cyclobutane core (CB-Ak₂) or a more flexible hydrocarbon chain (5C-Ak₂ or 11C-Ak₂). As expected, adding di-functional monomers lowered the crosslink density as evidenced by a two-fold decrease in rubbery modulus⁴⁹ of all three di-functional formulations (Figure S14 and Table S2). As illustrated in Figure 2B, the incorporation of a di-functional alkyne monomer produced a slight reduction in the glass transition temperature (T_g) depending on the stiffness of the backbone. For instance, a 16 °C decrease in T_g was achieved by adding 50 mol% of the flexible 11C-Ak₂ monomer, while a negligible change in T_g arose when adding 50 mol% of CB-Ak₂ monomer which is attributed to the presence of a stiff and sterically hindered cyclobutane backbone enhancing the molecular chain rigidity. Despite the reduced crosslink density, all three networks containing a di-functional monomer (CB-Ak₂, 5C-Ak₂, or 11C-Ak₂) exhibited a T_g of at least 30 °C above ambient with glassy storage moduli of ~1.7 GPa at 25 °C (Table S2).

Figure 2.

(A) Structures of di-functional azide monomer (Bz-Az₂), tri-functional alkyne monomer (Al-Ak₃), and three different di-functional alkyne monomers (CB-Ak₂, 5C-Ak₂, 11C-Ak₂). The three crosslinked polymers (denoted as Bz-Az₂ + Al-Ak₃ + CB-Ak₂, Bz-Az₂ + Al-Ak₃ + 5C-Ak₂, and Bz-Az₂ + Al-Ak₃ + 11C-Ak₂) were composed of a molar ratio per functional group of 1 : 0.5 : 0.5 : 0.01 : 0.02 (azide from di-azide : alkyne from tri-alkyne : alkyne from di-alkyne : CuCl₂[PMDETA] : DMPA) and compared with the control network (denoted as Bz-Az₂ + Al-Ak₃) composed of a molar ratio per functional group of 1 : 1 : 0.01 : 0.02 (azide from di-azide : alkyne from tri-alkyne : CuCl₂[PMDETA] : DMPA). Each mixture was irradiated at room temperature using 10 mW cm⁻² of UV light (λ_max = 366 nm) for 5 min followed by post cure at 70–80 °C for 24–48 hours. (B) Thermo-mechanical properties of four photopolymer systems as measured by DMTA (for storage modulus, see Figure S14). (C) Tensile testing of the four photopolymer systems at a crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) as measured via the MTS instrument (for replicates, see Figure S16–S19). Each specimen was heated above the T_g for 1 minute followed by 3 minutes of cooling at ambient temperature prior to mechanical testing.

Tensile testing was performed on networks containing one of these three di-functional alkyne monomers at ambient temperature and compared with the mechanical behavior of the un-modified and more highly crosslinked Bz-Az₂ + Al-Ak₃ network (Figure 2C and Table S2). As seen from Figure 2C, the incorporation of a di-functional alkyne monomer resulted in a drastic decrease in both the yield stress and Young’s modulus along with a much larger elongation at break. For instance, when 50 mol% (per functional group) of flexible 11C-Ak₂ monomer was added to the Bz-Az₂ + Al-Ak₃ network, the Young’s modulus and yield stress were decreased from 1030 to 470 MPa and 36 to 16 MPa, respectively, while the elongation at break value was greatly improved from 166 to 300 % (Table S2). The use of a stiff cyclobutane backbone (CB-Ak₂), in place of a flexible backbone of 11C-Ak₂ monomer, also resulted in a similar trend - a moderate reduction in Young’s modulus and yield stress to 850 and 33 MPa, respectively, and a slight improvement of the elongation at break value to 200 % - when compared to the Bz-Az₂ + Al-Ak₃ control network. Despite the higher elongation at break value achieved in the di-functional monomer based networks when compared to the Bz-Az₂ + Al-Ak₃ control network, tensile toughness was relatively similar between all comparisons (e.g., toughness ranging between 44 and 54 MJ m⁻³ for the three networks containing di-functional monomers as opposed to 48 MJ m⁻³ for the Bz-Az₂ + Al-Ak₃ network). This data suggests that although ductility was greatly improved by an addition of di-functional adducts, the effect on tensile toughness was minor due to the much lower yield stress in the networks containing di-functional alkyne monomers.

It should be noted that testing of specimens at a crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) required a long measurement time of up to 1 hour for the networks which can be elongated up to 300 % prior to failure. We previously⁴³ have observed that the Bz-Az₂ + Al-Ak₃ network exhibits a rapid physical aging process at ambient temperature, which induces brittleness in a relatively short time scale. Therefore, one would reasonably expect that a structurally similar glassy network with a lower T_g could exhibit an accelerated physical aging behavior at ambient temperature (viz., the Bz-Az₂ + Al-Ak₃ + 11C-Ak₂ network with a T_g of 49 °C compared to the Bz-Az₂ + Al-Ak₃ network with a T_g of 65 °C). To exclude potential time-dependent physical aging effects during the tensile testing measurement, a higher crosshead rate of 7.5 mm min⁻¹ (a strain rate of 50 % min⁻¹) was also employed (see Figure S20 for stress-strain curve), and the effect of a strain rate on tensile mechanical properties of the networks containing di-alkyne monomers was further explored (Table S2). When tested at the different crosshead rates of 0.75 and 7.5 mm min⁻¹ (strain rates of 5 and 50 % min⁻¹), these networks presented typical time-dependent viscoelastic behavior.⁵⁰ For instance, for the Bz-Az₂ + Al-Ak₃ + 5C-Ak₂ network, the yield stress and Young’s modulus were enhanced from 23 to 42 MPa and 590 to 990 MPa, respectively, and the elongation at break was reduced from 300 to 230 % when the crosshead rate was increased from 0.75 to 7.5 mm min⁻¹ (Table S2 and Figures 2C and S20). Similar behavior was shown for networks containing the 11C-Ak₂ or CB-Ak₂ monomers. Despite the effect of crosshead speed on properties, the general behavior of decreased modulus and yield stress and raised elongation at break for the networks containing the di-alkyne monomers was also observed at a crosshead rate of 7.5 mm min⁻¹; however, the toughness was also higher than for the un-modified Bz-Az₂ + Al-Ak₃ network, perhaps due to a reduction in physical aging during the tensile testing experiment (Table S2).

These examples demonstrate that the ductility of the CuAAC polymer network is significantly improved by the addition of a relatively short-chain di-alkyne monomers. Aside from lowering the crosslink density, these di-functional monomers effectively produce a network composed of triazole rings not only at the network junctions but also within the polymer backbones which may, in part, contribute to enhanced ductility and toughness in these networks.

In addition to forming less highly crosslinked networks by incorporating high concentrations of di-functional monomers as analyzed in Figure 2, mono-functional azide monomers of similar molecular weights but differing stiffness (namely 1-azidoadamantane or Stiff-Az₁ and azidomethyl pivalate or Flexible-Az₁) were incorporated into the Bz-Az₂ + Al-Ak₃ network (see monomers in Figure 3A). It is important to note that these mono-functional monomers cannot form part of the polymer network strand but simply terminate the strand and form a non-stress bearing, dangling chain coupled through a triazole linkage. Specifically, networks were prepared with a stoichiometric ratio of azide and alkyne groups in which 37 mol% of the azide groups originated from the mono-functional azide monomer and 63 mol% of the azide groups originated from the di-functional azide monomer (Bz-Az₂) while the alkyne groups were supplied by the tri-functional alkyne monomer (Al-Ak₃). Based on the Flory-Stockmayer equation,^51–57 the molar concentration of 37 % per functional group was specifically chosen to form a more loosely crosslinked network, since for this composition, the theoretical gel point conversion is approximately 90 % in contrast to gel point conversions ranging from 58% to 82% for the other previous networks (see the end of SI for more information).

Figure 3.

(A) Structures of the di-functional azide monomer (Bz-Az₂), tri-functional alkyne monomer (Al-Ak₃), and two different mono-functional azide monomers (Stiff-Az₁, Flexible-Az₁). The two crosslinked polymers (denoted as Bz-Az₂ + Stiff-Az₁ + Al-Ak₃ and Bz-Az₂ + Flexible-Az₁ + Al-Ak₃) were composed of a molar ratio per functional group of 0.63 : 0.37 : 1 : 0.01 : 0.02 (azide from di-azide : azide from mono-azide : alkyne from tri-alkyne : CuCl₂[PMDETA] : DMPA) and compared with the control network (denoted as Bz-Az₂ + Al-Ak₃) composed of a molar ratio per functional group of 1 : 1 : 0.01 : 0.02 (azide from di-azide : alkyne from tri-alkyne : CuCl₂[PMDETA] : DMPA). Each mixture was irradiated at room temperature using 10 mW cm⁻² of UV light (λ_max = 366 nm) for 5 min followed by post cure at 70–80 °C for 24–48 hours. (B) Tensile testing of three photopolymer systems at a crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) as measured via the MTS instrument (for replicates, see Figure S28–S30). Each specimen was heated above T_g for 1 minute followed by 3 minutes of cooling at ambient temperature prior to mechanical testing. (C) Thermo-mechanical properties of the three photopolymer systems as measured by DMTA.

As expected, incorporating a mono-functional azide monomer into the networks drastically decreased the crosslink density and thus the rubbery modulus⁴⁹ from 4 to below 0.4 MPa (see the DMTA data shown in Figures 3C and 3D and Table S3). Interestingly, despite the more loosely crosslinked nature of these networks, both Stiff-Az₁ and Flexible-Az₁ formulations exhibited glassy storage moduli of 1 GPa at ambient temperature and glass transition temperatures (T_gs) above 50 °C. For the network containing the Stiff-Az₁ monomer, the factors which counteract the effect of the reduced crosslink density on T_g are the increased concentration of triazole rings and the stiffening effect of the bulky adamantane ring, whereas the network containing the Flexible-Az₁ has only the former counteracting factor, resulting in a lower T_g.

As shown in Figure 3B, the addition of the two different mono-functional azides (Stiff-Az₁ or Flexible-Az₁) resulted in noticeably different stress-strain curves. For example, incorporation of the Flexible-Az₁ into the Bz-Az₂ + Al-Ak₃ network resulted in a yield stress of approximately half that of the other two networks (Table S3), perhaps due to the reduction in crosslink density and the flexible nature of the methylene pivalate group. However, incorporation of Stiff-Az₁ into the Bz-Az₂ + Al-Ak₃ network had little effect on the yield stress, perhaps due to the counteracting effects of the reduced crosslink density and increased rigidity of the adamantane group. Also, the network with the Stiff-Az₁ resulted in very low toughness of 6 MJ m⁻³ and an elongation at break of only 24 %. In contrast, incorporation of Flexible-Az₁ yielded ductile behavior with considerably higher elongation at break over 180 % and a four times higher toughness of 28 MJ m⁻³ which is closer to that found with the pure Bz-Az₂ + Al-Ak₃ network (140 % and 45 MJ m⁻³ - see Table S3). However, it should be noted that when the concentration of mono-azide monomers was decreased from 37 to 26 mol % per azide functional group (see Table S4 for tensile testing and DMTA data), the structural effects associated with the differences from Stiff-Az₁ to Flexible-Az₁ were also reduced. For instance, both networks containing 26 mol % of either Stiff-Az₁ or Flexible-Az₁ exhibited similar rubbery moduli⁴⁹ of 2 MPa, elongation at break around 100 %, and toughness of 30 MJ m⁻³. These values were slightly lower than that of the Bz-Az₂ + Al-Ak₃ network (Table S4). Thus, in summary, it appears that incoporation of mono-azide monomers which introduce dangling chains into the network without increasing the network strand length tends to reduce the elongation at break, but that this effect is partially countered by a flexible dangling network chain (as with Flexible-Az₁). As a result, the tensile toughness of polymers containing the mono-azide monomers is reduced compared with the Bz-Az₂ + Al-Ak₃ network, and this behavior may be accentuated by the concomitant decrease in the concentration of di-functional azide monomers (Bz-Az₂) which contain the toughening effect of carbamate groups in a relatively long and flexible backbone.

The effect of the flexibility of the network structure on mechanical strength and ductility of the various CuAAC networks discussed above is summarized in Figure 4, in which yield stress, elongation at break, and tensile toughness are plotted as a function of glass transition temperature (T_g). Figure 4A presents the comparison of three networks varied by changing the alkyne monomer functionality (from tetra- to tri-, and di-functional alkyne monomers) without affecting the backbone structure. Figure 4B shows the comparison of three networks with different backbone rigidity (from a stiff cyclic group to a flexible aliphatic chain in the monomer backbones) without changing the monomer functionality. It should be noted that the tensile testing results in Figure 4B showing the correlation between toughness and elongation at break are not very convincing, and this behavior may be partly due to the physical aging effect discussed above. When the data obtained using a higher crosshead rate of 7.5 mm min⁻¹ are used, a stronger correlation is observed (see Figure S21). These near-linear relationships show that yield stress, elongation at break, and tensile toughness are readily manipulated by systematic modification of the monomer structures.

Figure 4.

Correlation of T_g and the tensile testing properties - yield stress (red squares), elongation to break (blue circles) and toughness (green triangles) - for networks with varying backbone stiffness and crosslink density. (A) The three different network structures, Bz-Az₂ + Al-Ak₃ (denoted as Tri-Alkyne) and Bz-Az₂ +Al-Ak₄ (denoted as Tetra-Alkyne) whose structures are shown in Figure 1, and Bz-Az₂ + Al-Ak₃ + 5C-Ak₂ (denoted as Di-Alkyne) whose structure is shown in Figure 2 were compared to investigate alkyne functionality variations in their network with Bz-Az₂. A crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) was utilized for tensile testing (see Figure 1C and 2C for the stress-strain curves). (B) The three different networks, Bz-Az₂ + Al-Ak₃ + CB-Ak₂ (denoted as cyclic CB-Ak₂), Bz-Az₂ + Al-Ak₃ + 5C-Ak₂, (denoted as aliphatic 5C-Ak₂), and Bz-Az₂ + Al-Ak₃ + 11C-Ak₂ (denoted as aliphatic 11C-Ak₂) whose structures are shown in Figure 2 were compared for the effect of backbone structure variations in their network with Bz-Az₂ and Al-Ak₃. A crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) was utilized for tensile testing (see Figure 2C for the stress-strain curves).

Extending the effect of alkyne monomer structures on properties shown above in Figure 4, the yield stress dependence on network mobility (viz. T_g) and the dependence of elongation at break on crosslink density (viz. rubbery modulus)⁴⁹ were further investigated for all of the CuAAC networks presented in Figures 1, 2, and 3, and the results are summarized in Figure 5. Common trends include that an increase in yield stress with increasing T_g and a decrease in elongation at break with increasing rubbery modulus were observed in all of the CuAAC networks except for the two mono-functional azide systems. For the latter two networks, the elongation at break differed tenfold while the crosslink density was only 30% different. Furthermore, the elongation at break values were even lower than other networks with higher crosslink density (e.g., for the case of di-functional systems). This behavior is possibly caused by the fact that the length of the stress-bearing network strands are not significantly changed by addition on these mono-functional monomers since they only become dangling chains in the network.

Figure 5.

Correlation of the DMTA T_g with the tensile testing results of all systems in Figure 1, 2, and 3. Four multi-functional systems (Bz-Az₂ + Al-Ak₃, Bz-Az₂ + Al-Ak₄, Bz-Az₂ + Ar-Ak₃, and Ch-Az₂ + Al-Ak₃ - see Figure 1 for structures), three di-functional system (Bz-Az₂ + Al-Ak₃ + CB-Ak₂, Bz-Az₂ + Al-Ak₃ + 5C-Ak₂, and Bz-Az₂ + Al-Ak₃ + 11C-Ak₂ - see Figure 2 for structures), and two mono-functional systems (Bz-Az₂ + Stiff-Az₁ + Al-Ak₃ and Bz-Az₂ + Flexible-Az₁ + Al-Ak₃ - see Figure 3 for structures) were compared. A crosshead rate of 0.75 mm min⁻¹ (a strain rate of 5 % min⁻¹) was utilized for tensile testing. (A) A plot of yield stress as a function of glass transition temperature. (B) A plot of elongation at break as a function of rubbery modulus. Data points with asterisk (*) indicate the average of two measurements were used without an error bar.

CONCLUSIONS

In this investigation, we have explored the effect of structural variations in multi-functional azide and alkyne monomers on the tensile properties of the resultant networks formed via CuAAC bulk photopolymerizations. Tailorable network structures were readily obtained by selection of azide and alkyne monomers with various backbones and through alterations in the number of reactive azide or alkyne functional groups of the monomers. Decreasing the backbone rigidity and crosslink density of the triazole-based photopolymer networks were found to gradually lower stiffness and strength (measured by Young’s modulus and yield stress) while considerably enhancing ductility (measured by elongation at break) and toughness. Due to the unique attribute of forming triazoles as a product of the CuAAC reaction, a T_g well above ambient was maintained even when mono- or di-functional monomers were introduced to form a less crosslinked network. The di-functional (alkyne) monomers become part of the polymer backbone and decreased crosslink density by increasing the network strand length, but also increased the triazole concentration, whereas the mono-functional (azide) monomers only become part of the polymer structure as dangling chains while also decreasing crosslink density. Thus, this work highlights the design of highly ductile photopolymer networks whose properties are tunable by crosslink density and chemical backbone structure changes.