Letter pubs.acs.org/macroletters
pH-Triggered Release from Polyamide Microcapsules Prepared by Interfacial Polymerization of a Simple Diester Monomer Hsuan-Chin Wang,† Joshua M. Grolman,‡,§ Aoon Rizvi,† Grant S. Hisao,† Chad M. Rienstra,† and Steven C. Zimmerman*,† †
Department of Chemistry, ‡Beckman Institute for Advanced Science and Technology, and §Department of Materials Science and Engineering, University of Illinois at Urbana−Champaign, Urbana, Illinois 61801, United States S Supporting Information *
ABSTRACT: The majority of current pH-triggered release systems is designed to respond to either low or high pH. Encapsulants based on polyampholytes are an example of materials that can respond to both acidic and basic pH. However, polyampholyte-based encapsulants generally possess a low loading capacity and have difficulty retaining their small-molecule cargo. The current work utilizes interfacial polymerization between polyamines and a pyromellitic diester diacid chloride to form high capacity “liquid core−shell” polyamide microcapsules that are stable in a dry or nonpolar environment but undergo steady, controlled release at pH 7.4 and accelerated release at pH 5 and pH 10. The rate of release can be tuned by adjusting the amine cross-linker feed ratio, which varies the degree of cross-linking in the polymer shell. The thin-shell microcapsule exhibited suitable barrier properties and tunable dual acid/base-triggered release, with applications in a wide range of pH environments.
T
small-molecule dye at neutral pH and accelerated release at both acidic and basic pH. To prepare robust, high capacity microcapsules, we utilized interfacial microencapsulation, a demonstrated technique extensively investigated and, indeed, regularly employed on an industrial production level.21,22 Analogous in concept to the classic “nylon-rope trick”,23 interfacial polyamide microencapsulation takes advantage of the rapid polymerization of a hydrophobic polyacid chloride with hydrophilic polyamines at the interface of immiscible phases to ensure the full encapsulation of the emulsified O/W or W/O droplet template. In the current work, the hydrophobic phase was an emulsified toluene droplet containing pyromellitic diester diacid chloride (PDDC, 1) as the hydrophobic monomer. With an electron-withdrawing ester group ortho to each of the two acyl groups, 1 is expected to produce polyamides that are more susceptible to hydrolysis than those prepared from terephthaloyl chloride (TC), which is an ester-free analogue commonly used in the preparation of polyamide films,24 fibers, and encapsulating shells.25 We speculated that the polyamide shell synthesized from 1 would more readily undergo hydrolytic degradation at the carbonyl groups to form anionic carboxylate groups that lead to polymer swelling by electrostatic repulsion and the dissolution of cleaved polymer fragments. The hydrophilic continuous phase consisted of an aqueous solution of two polyamines, N,N′,N″-tris(2-aminoethyl)-1,3,5-triazine2,4,6-triamine (2) and diethylenetriamine (DETA, 3).
hin-shell encapsulation is attractive because the large coreto-shell volume ratio enables delivery of a large payload with minimal investment in material for the surrounding shell. However, for many applications, a high carrying capacity is insufficient by itself; an efficient, stimuli-responsive release mechanism is required to achieve both the desired release rate and a high payload delivery. Not surprisingly, pH is a popular and useful release trigger for thin-shell capsules given the ubiquity of water with specific pH levels in various application environments ranging from biomedical to agricultural to those in consumer products and industrial processes.1−3 Most current pH-sensitive triggered release capsular systems utilize either low or high pH to initiate efficient payload release.4−11 Polyampholytes are one type of pH-responsive material that can respond to both high and low pH by virtue of their polymeric framework containing both cationic and anionic groups. Stimuli-responsive polyampholytes have been used in gels, films, and thin-shell capsules,12−18 the latter typically formed around a template or at a liquid−liquid interface. However, as is common for polyelectrolyte assemblies, preparation of stable, liquid-core-thin-shell ampholyte capsules generally requires multiple steps, typically layer by layer deposition, core template dissolution, shell cross-linking for structural stabilization, or cargo encapsulation by postfabrication loading. Furthermore, although dual acid/base-triggered release of small molecules from polyampholyte hydrogels and thin films has been reported, analogous release from liquid-core microcapsules, to the best of our knowledge, has yet to be demonstrated.19,20 In this work, we fabricated in one step a high capacity polymeric microcapsule that remained stable in a nonpolar environment but demonstrated sustained release of a © XXXX American Chemical Society
Received: December 22, 2016 Accepted: February 27, 2017
321
DOI: 10.1021/acsmacrolett.6b00968 ACS Macro Lett. 2017, 6, 321−325
Letter
ACS Macro Letters Scheme 1. Encapsulation by Interfacial Polymerization and Release by pH Triggers
upon drying. These observations suggest that the two polyamines work synergistically to produce microcapsules with enhanced barrier properties. The core solvent weight percent was experimentally determined by shredding weighed capsules with sonication and collecting the polymer fragments by vacuum filtration. The obtained mass difference between full capsules and dried capsule fragments was used in the following equation to calculate the core solvent weight percent
A slight modification of a procedure reported by Fréchet and co-workers was used for capsule preparation.8 As shown in Scheme 1, a saturated solution of 1 in toluene was emulsified with a 0.4% w/w aqueous poly(vinyl alcohol) solution with magnetic stirring to produce a suspension of oil-in-water emulsion droplets, whose shapes and sizes determine the final capsule dimensions. Dropwise addition of an aqueous solution of triazine 2 and DETA initiated a polycondensation reaction at the water−oil interface to form a cross-linked polyamide shell. Maturation for 60 min afforded a suspension of microcapsules that were vacuum filtered, rinsed with DI water and acetone, and allowed to dry in air to give a white, free-flowing powder. The selection of the polyamine monomers, triazine 2, and DETA 3, was based on results from screening experiments. Various water-soluble polyamines (SI Figure S1) were reacted with 1 to form microcapsules, and the quality of the resulting capsules was preliminarily evaluated based on core solvent retention and capsule dispersibility. Initial polymerization attempts with commercially available DETA produced soft, aggregated microcapsules with low core solvent retention (∼30%), despite the fact that DETA is known to polymerize interfacially with the closely related TC to form polyamide microcapsules with good barrier properties.25 This reported stability of the DETA/TC microcapsules was confirmed in this study (data not shown). A plausible hypothesis to explain the observed differences in barrier and mechanical properties is a reduced level of cross-linking of DETA as a result of steric hindrance from the ester groups in the ortho-position. Although this possible reduction in cross-linking might weaken the barrier properties while leaving some amines unreacted, we sought to utilize this limitation to impart additional pH responsiveness. Thus, the unreacted amines might render the polymer shell polycationic in low pH environments, thereby inducing swelling and cargo release. To compensate for the reduction in barrier properties, triazine cross-linker 2, which by itself was found to react with 1 to form microcapsules with good core solvent retention (∼70%), was used in combination with DETA at an ca. 1:1 molar ratio. Interestingly, the resulting mixture of triazine 2 and DETA afforded robust PDDC microcapsules with solvent retention up to 95% and as a dispersible free-flowing powder
Core weight % mass of dry capsule − mass of dried polymer = mass of dry capsules × 100%
In most cases, the core content was roughly 95% of the total capsule weight. It is worth noting that after 18 months of storage at room temperature in a plastic-capped vial the weight percent of the toluene core was still around 95% in the microcapsules, suggesting high storage stability in the dry state. Detailed characterization of capsule morphology, size, and shell thickness was performed by scanning electron microscopy (SEM). The shell thickness was determined by cutting capsules with a razor blade and imaging the cross-section. As shown in Figure 1, the typical capsule diameters were between 200 and 350 μm, with shell thicknesses in the range of 0.8−1.5 μm. The dimensional variations found in the microcapsules reflected the size range of emulsified droplet templates. The inner and outer surfaces also differed in their roughness, with the outer surface smoother than the inner surface, which were lined with small pillar-like features. The rugged inner surface is characteristic of interfacially polymerized microcapsules and is thought to be evidence for the inward, self-limiting diffusion of the amine monomers toward the hydrophobic core center.26 Solid-state attenuated total reflectance Fourier transform infrared (ATR FT-IR) spectroscopy confirmed the presence of amide (characteristic amide I at 1634 cm−1, amide II at ∼1557 cm−1 likely overlapping with the C−N band from the triazine) and the inclusion of triazine moieties (813 cm−1, ring out of plane deformation)27 in the polymer shell (SI Figure S8a). The peak near 1716 cm−1 is assigned to the carbonyl stretch from 322
DOI: 10.1021/acsmacrolett.6b00968 ACS Macro Lett. 2017, 6, 321−325
Letter
ACS Macro Letters
Figure 1. SEM images of (a) 3:1 triazine 2:DETA, (b) cross-section of 3:1 triazine 2:DETA capsule shell, (c) 1:1 triazine 2:DETA capsules, and (d) cross-section of 1:1 triazine 2:DETA capsule shell. The scale bar is 500 μm.
remaining ester or imide groups, the latter potentially forming by intramolecular cyclization. Although differentiation of these two possible functional groups is difficult, it was possible to experimentally confirm that both the ester and imide groups in small-molecule analogues readily hydrolyze at high pH and should lead to the anticipated pH response (SI Figures S4a, S5a, and S6a). IR analyses of capsules after exposure to different pH buffers also indicated disappearance of the 1716 cm−1 peak in basic pH (SI Figure S8b). The barrier properties of the microcapsules were tested through standard dye release experiments. The hydrophobic UV dye coumarin 1 served as a model active, being readily dissolved in the organic phase along with the PDDC monomer prior to capsule synthesis. Microcapsules were prepared from two different feed ratios of triazine 2 and DETA in the aqueous phase to examine the effect of increasing the concentration of triazine, which functioned as a cross-linking agent. The coumarin 1-loaded capsules were incubated in toluene in the presence and absence of an aqueous layer containing varying pH buffers to assess the rate of dye release under different conditions. As can be seen in Figures 2a and 2b, in the absence of water the capsules showed no noticeable release of dye in toluene, regardless of the feed ratios of the aqueous monomers. In contrast, all samples that were in contact with aqueous buffers at pH 5, 7.4, and 10 released dye into the toluene phase over time. Increasing the encapsulated dye content by 10-fold appeared to have a minimal effect on the general release patterns (SI Figure S2). Clearly, the observed release behavior indicates that exposure to an aqueous environment triggered the release of the encapsulated dye, with the rate of release influenced by the pH of the aqueous phase. The observed barrier properties suggest that the PDDC microcpasule could be stably stored as a dry or hydrophobic formulation and with release triggered in a wet environment such as that encountered in an agricultural setting.28,29 It is interesting to note that the release behavior at pH 7.4 showed a consistent feature for both 1:1 and 3:1 triazine 2:DETA microcapsules: a roughly linear, sustained release profile, albeit at a significantly reduced rate in comparison to high and low pH. Because hydrolysis is expected to be minimal at neutral pH, it is possible the release occurs by water-
Figure 2. Release of coumarin 1 dye under different solvent conditions for PDDC capsules prepared with (a) 3:1 triazine 2:DETA and (b) 1:1 triazine 2:DETA. (c) Control experiments: 1:1 triazine 2:DETA with TC. Uncertainty represents standard deviation of three measurements.
mediated plasticization30 of the polyamide shell. We hypothesize that the combined effect of enhanced mobility within the plasticized polymer and the slow dissolution of the hydrophobic coumarin 1 in the polar polymer shell contributed to the steady zero-order release of the dye.31 The increase in rate of release from pH 7.4 to pH 10 can be understood as a result of base-accelerated hydrolysis of the ester or imide groups to form carboxylate groups that cause polymer swelling and enhanced solvent exchange. Interestingly, the rate of release at pH 5 was higher than that observed at pH 7.4 and comparable to that at pH 10, an effect that was clearly absent in the control capsules made with TC as the hydrophobic monomer (Figure 2c). To investigate whether the observed release trends were controlled by ester or imide hydrolysis, three structurally analogous small molecules (bisimide 4, tetraamide 5, and diester diamide 6, shown in Figure 3) were dissolved in D2O at pD 5, 7, and 10 and their degradation kinetics monitored by 1H NMR. All three molecules were significantly more stable at pD 5 than at pD 7 and pD 10 (SI Figures S4c, S5c, and S6c). 323
DOI: 10.1021/acsmacrolett.6b00968 ACS Macro Lett. 2017, 6, 321−325
Letter
ACS Macro Letters
fixed amount of residual amine groups. Furthermore, an additional piece of evidence to support the “residual amine hypothesis” came from solid-state 13C NMR analysis that (SI Figure S7) is consistent with the presence of unreacted DETA, which has a unique methylene signal at around 53 ppm in the free amine form, but which shifts upfield to 48 ppm when converted to an aromatic amide. Furthermore, it should be noted that, at pH 5, the diethylaniline group of coumarin 1 dye can become protonated, potentially increasing the dye’s solubility and diffusivity in the polar polymer shell. The role of the dye structure in the mechanism and rate of release is not known and will be examined in future studies. As seen in Figure 2, the capsule release behavior also depended on the amine monomer feed ratio. Increasing the proportion of the trifunctional triazine monomer slowed the release of dye across the three pH conditions tested, presumably as a result of an increased cross-linking density giving a denser and less permeable polymeric shell. This possibility was examined indirectly by measuring the mechanical properties of two batches of capsules prepared with a 1:1 or 3:1 triazine 2 to DETA (3) ratio. As seen in Figure 4,
Figure 3. Small-molecule analogues to simulate potential microcapsule shell polymer structures.
Indeed, enhanced hydrolysis was not observed at pD 5, suggesting a different triggered release mechanism is likely dominating in the capsule-triggered release. These data are consistent with operation of the originally designed mechanism in which protonation of residual, unreacted amine groups increased swelling as a result of the increased polyionic character. Compared to TC, the structurally hindered PDDC potentially prevented the amine groups from completely reacting with the acid chloride groups, effectively rendering the capsule shell “cationic” when exposed to low pH. Additional evidence for the stability of the polymer capsules at pH 5 can be gleaned from the degrees of swelling in capsules at different pHs. Microcapsules were sprinkled on a thin film of polyurethane adhesive and cured for 24 h. The hardened adhesive patches with surface-lodged microcapsules were incubated in pH buffers and imaged under an optical microscope to determine the effect of pH on their diameter (SI Figure S3). The capsule diameter swelling ratios were calculated by dividing the average diameter measured in buffer by the average diameter of the dry capsules (see Table 1). At Table 1. Swelling Ratios of 1:1 Triazine 2:DETA Capsules in Different Buffers at 1 and 4 Days pH
day 1
day 4
5 7.3 10.3
1.19 ± 0.04 1.110 ± 0.002 1.31 ± 0.09
1.19 ± 0.04 1.13 ± 1 × 10−5 1.7 ± 0.9a
Figure 4. Stiffness measurements of 3:1 and 1:1 triazine 2:DETA capsules. Uncertainty represents standard deviation of three measurements.
a
Most capsules detached from the adhesive; ratio calculated by averaging 41 capsules on surface of glue from 4 days and comparing to average of 21 capsules from 0 days. Uncertainty values reflected standard deviations of at least 10 diameter measurements.
nanoindentation indicated that an increase in triazine concentration led to an increase in capsule stiffness. These stiffness data provide indirect evidence for increased crosslinking in capsules made with a higher triazine cross-linker concentration. In turn, these data show that the permeability and mechanical strength of the capsules can be easily tuned by varying the feed ratio of the triazine and DETA monomers. In summary, we have developed a polyamide microcapsule using interfacial polymerization of a diester diacid chloride with DETA and a triazine trisamine to impart high loading capacity and dual acid/base-responsive properties. The preparation is sufficiently simple that it should be suitable to a larger-scale production level. To adjust the release profile, shell permeability was varied by altering the amine monomer feed ratio, demonstrating a simple tuning mechanism. The ability to retain a volatile core over a long period in a dry or hydrophobic state and to achieve different rates of small-molecule release over days when exposed to different pH suggests that this versatile pH-responsive microcapsule might be useful for applications such as fragrance release on moist skin surfaces,
pH 5, a constant swelling ratio of 1.19 was observed at 1 and 4 days, whereas at pH 7.3, the swelling ratio increased slightly from 1.11 at 1 day to 1.13 at 4 days. The largest change in swelling ratio occurred at pH 10.3, starting at 1.31 at 1 day and reaching 1.7 at 4 days. The differences in swelling ratios could be attributed to increased hydrolytic cleavage of polymer chains at higher pH that decreased cross-linking and allowed the capsules to expand with increased solvent intake. Consistent with this notion, the trend in swelling ratios correlated with the relative stabilities observed in small-molecule degradation studies previously discussed. It is interesting to note that the constant swelling at pH 5 suggested the capsules responded to the pH 5 aqueous buffer in a consistent way, which fits the behavior predicted for a constant overall degree of protonation at a constant pH for a 324
DOI: 10.1021/acsmacrolett.6b00968 ACS Macro Lett. 2017, 6, 321−325
Letter
ACS Macro Letters
(17) Kozlovskaya, V.; Sukhishvili, S. A. Macromolecules 2006, 39, 6191. (18) Kudaibergenov, S. E.; Nuraje, N.; Khutoryanskiy, V. V. Soft Matter 2012, 8, 9302. (19) Ng, L.-T.; Ng, K.-S. Radiat. Phys. Chem. 2008, 77, 192. (20) Barcellona, M. N.; Johnson, N.; Bernards, M. T. Langmuir 2015, 31, 13402. (21) Andrade, B.; Song, Z.; Li, J.; Zimmerman, S. C.; Cheng, J.; Moore, J. S.; Harris, K.; Katz, J. S. ACS Appl. Mater. Interfaces 2015, 7, 6359. (22) Fessi, H.; Briancon, S. p.; Puel, F. ß. In Microencapsulation; Informa Healthcare: 2005; p 149. (23) Morgan, P. W.; Kwolek, S. L. J. Chem. Educ. 1959, 36, 182. (24) Takahashi, Y.; Iijima, M.; Oishi, Y.; Kakimoto, M.; Imai, Y. Macromolecules 1991, 24, 3543. (25) Mathiowitz, E.; Cohen, M. D. J. Membr. Sci. 1989, 40, 27. (26) Mathiowitz, E.; Cohen, M. D. J. Membr. Sci. 1989, 40, 1. (27) Larkin, P. J.; Makowski, M. P.; Colthup, N. B.; Flood, L. A. Vib. Spectrosc. 1998, 17, 53. (28) Beestman, G. B. In Pesticide Formulation and Adjuvant Technology; Foy, C. L., Pritchard, D. W., Eds.; CRC Press: 1996; p 43. (29) Markus, A.; Linder, C. In Microencapsulation: Methods and Industrial Applications, 2nd ed.; Benita, S., Ed.; CRC Press: 2005; Vol. 18, p 55. (30) Ellis, T. S. J. Appl. Polym. Sci. 1988, 36, 451. (31) Guzman, G.; Es-haghi, S. S.; Nugay, T.; Cakmak, M. Adv. Healthcare Mater. 2017, 6, 1600775.
alkaline laundry process, or delivery of pest control agents in agricultural settings. The use of the single encapsulation/release system for multiple environments is a key advantage of the chemistry described herein.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmacrolett.6b00968. Experimental and synthetic details and 1H NMR spectra for degradation kinetics studies (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Chad M. Rienstra: 0000-0002-9912-5596 Steven C. Zimmerman: 0000-0002-5333-3437 Author Contributions
All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We would like to thank Prof. Martin Burke for the use of their optical microscope. We also want to thank the Beckman ITG facilities for their characterization instruments and helpful discussions with the instrument specialists. This work was supported by a grant from the Dow Chemical Company.
■
REFERENCES
(1) Binauld, S.; Stenzel, M. H. Chem. Commun. 2013, 49, 2082. (2) Esser-Kahn, A. P.; Odom, S. A.; Sottos, N. R.; White, S. R.; Moore, J. S. Macromolecules 2011, 44, 5539. (3) Wang, H.-C.; Zhang, Y.; Possanza, C. M.; Zimmerman, S. C.; Cheng, J.; Moore, J. S.; Harris, K.; Katz, J. S. ACS Appl. Mater. Interfaces 2015, 7, 6369. (4) Okahata, Y.; Noguchi, H.; Seki, T. Macromolecules 1987, 20, 15. (5) Sukhorukov, G. B.; Antipov, A. A.; Voigt, A.; Donath, E.; Möhwald, H. Macromol. Rapid Commun. 2001, 22, 44. (6) De Geest, B. G.; Van Camp, W.; Du Prez, F. E.; De Smedt, S. C.; Demeester, J.; Hennink, W. E. Macromol. Rapid Commun. 2008, 29, 1111. (7) Esser-Kahn, A. P.; Sottos, N. R.; White, S. R.; Moore, J. S. J. Am. Chem. Soc. 2010, 132, 10266. (8) Broaders, K. E.; Pastine, S. J.; Grandhe, S.; Frechet, J. M. J. Chem. Commun. 2011, 47, 665. (9) Abbaspourrad, A.; Datta, S. S.; Weitz, D. A. Langmuir 2013, 29, 12697. (10) Ejima, H.; Richardson, J. J.; Liang, K.; Best, J. P.; van Koeverden, M. P.; Such, G. K.; Cui, J.; Caruso, F. Science 2013, 341, 154. (11) Grolman, J. M.; Inci, B.; Moore, J. S. ACS Macro Lett. 2015, 4, 441. (12) Baker, J. P.; Stephens, D. R.; Blanch, H. W.; Prausnitz, J. M. Macromolecules 1992, 25, 1955. (13) Kashiwabara, M.; Fujimoto, K.; Kawaguchi, H. Colloid Polym. Sci. 1995, 273, 339. (14) Das, M.; Kumacheva, E. Colloid Polym. Sci. 2006, 284, 1073. (15) Georgiou, T. K.; Patrickios, C. S. Biomacromolecules 2008, 9, 574. (16) Walter, H.; Harrats, C.; Müller-Buschbaum, P.; Jérôme, R.; Stamm, M. Langmuir 1999, 15, 1260. 325
DOI: 10.1021/acsmacrolett.6b00968 ACS Macro Lett. 2017, 6, 321−325
