Chemically Active, Porous 3D-Printed Thermoplastic Composites

Subscriber access provided by Gothenburg University Library

Article

Chemically-Active, Porous 3D-Printed Thermoplastic Composites Kent Evans, Zachary C Kennedy, Bruce W. Arey, Josef F Christ, Herbert Todd Schaef, Satish K Nune, and Rebecca L Erikson ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b17565 • Publication Date (Web): 31 Jan 2018 Downloaded from http://pubs.acs.org on January 31, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Chemically-Active, Porous 3D-Printed Thermoplastic Composites Kent A. Evans,† Zachary C. Kennedy, † Bruce W. Arey, ‡ Josef F. Christ, † Herbert T. Schaef,§ Satish K. Nune, ‡* and Rebecca L. Erikson†* †

National Security Directorate, ‡Energy and Environment Directorate, and §Physical and

Computational Sciences Directorate, Pacific Northwest National Laboratory, Richland, WA 99352, USA KEYWORDS: 3D printing, metal-organic framework, fused-deposition modeling, polymer composite, flexible, porous, extrusion

ABSTRACT

Metal-organic frameworks (MOFs) exhibit exceptional properties and are widely investigated because of their structural and functional versatility relevant to catalysis, separations, and sensing applications. However, their commercial or large-scale application is often limited by their powder forms which make integration into devices challenging. Here, we report the production of MOF-thermoplastic polymer composites in well-defined and customizable forms and with complex internal structural features accessed via a standard 3D printer. MOFs (Zeolitic imidazolate framework; ZIF-8) were incorporated homogeneously into both poly(lactic acid) (PLA) and thermoplastic polyurethane (TPU) matrices at high loadings (up to 50% by mass),

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 33

extruded into filaments, and utilized for on-demand access to 3D structures by fused-deposition modeling. Printed, rigid PLA-MOF composites display large surface area ((SA)avg = 531 m2 g-1) and hierarchical pore features, whereas flexible TPU-MOF composites achieve high surface area ((SAavg = 706 m2 g-1) by employing a simple method developed to expose obstructed micropores post-printing. Critically, embedded particles in the plastic matrices retain their ability to participate in chemical interactions characteristic of the parent framework. The fabrication strategies were extended to other MOFs and illustrate the potential of 3D printing to create unique porous and high surface area chemically-active structures.

INTRODUCTION A multitude of metal organic frameworks (MOFs) have been developed and used broadly throughout the fields of chemistry, materials science, and engineering. The ability to fine tune the crystalline structure of a MOF through judicious selection of both the organic linker moieties and metal ion centers allows the modulation of catalytic activities, pore sizes for molecular storage and separations, and luminescent properties for sensing applications.1-3 One outstanding hurdle to the commercial and industrial application of MOFs is their powder form as pure crystalline materials.4,5 For incorporation into devices, ideally MOF crystallites could be integrated into a composite platform to improve the mechanical properties yet still retain their desired chemical reactivity characteristics by preservation of their crystallinity, porosity, and surface area. To this end, MOFs have been incorporated into polymers and cast into films or formed into “mixed-matrix membranes” (MMM), grown in situ on the surface of other materials, deposited onto surfaces using a layer-by-layer strategy, electrospun into mats, or inkjet-printed onto surfaces.6-10 In some cases, the growth of a MOF layer directly on a pre-formed substrate


2

Page 3 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


may be adequate. However, issues related to the chemical compatibility of the underlying structure and the MOF synthetic procedure and the potential for poor adhesion of the MOF to the pre-fabricated structure motivate the development of novel methods to disperse pre-grown MOFs entirely throughout a composite structure. 3D printing is a disruptive technology poised to fundamentally augment small-scale manufacturing capabilities in numerous industries spanning from automotive and aerospace, healthcare, to nuclear energy.11-14 Currently, a major limitation across 3D printing technologies is the versatility of feedstock materials.15 For example, typical 3D printing thermoplastics, used primarily for their mechanical and structural properties, have little intrinsic functionality to act as sensors, conduct electricity, or perform catalytic functions, which significantly limits their application. As such, there is an impetus behind the development of printable feedstocks with intrinsic chemical and physical functionality in addition to mechanical stability. While a number of fillers and additives have been explored to address many of the functional shortcomings of 3D printing materials,16-18 surprisingly little has been done to apply the functionalities intrinsic to MOFs (catalytic sites, defined pore structures for chemical separation and storage, etc.) in this space. The ability to fine tune the properties of MOFs through their structures make them ideal candidates as additives for developing multifunctional 3D printing feedstocks. While a number of MOF composites have been proposed, there have been few efforts to design printable MOFinfused materials.19-24 Specifically, protocols to coat the surface of previously 3D-printed objects with MOFs for dye removal have been introduced.25, 26 And another account reports fabrication of MOF composites using acrylonitrile butadiene styrene (ABS) as the base polymer with MOF loadings up to 10 wt. %, although framework degradation and composite inhomogeneity was observed.27 In addition, 3D-printed monolithic contactors containing MOF-74(Ni) and UTSA-


3


Page 4 of 33

16(Co) MOFs were recently fabricated by extrusion of paste precursor in a pressurized syringe.28 3D printing techniques have been also used to print reactors and disposable pressure vessels for the synthesis of MOFs.29,30 In this work, we aimed to apply straightforward and generalizable approaches to achieve thermoplastic-MOF composites, utilize the resulting composites as feedstocks for FDM printing, and maintain valuable MOF structure and chemical activity in the printed objects. PLA was selected as an initial matrix polymer for our investigations printing MOF composites. The thermoplastic properties of PLA are well understood,31 it is simple to print with on a standard benchtop fused deposition modeling (FDM) printer by both experts and non-specialists, and is cost effective, making a 3D-printed PLA/MOF composite poised for wide-spread adoption in numerous applications. In addition, we targeted the production of TPU/MOF composites. TPUbased 3D printing, although less established than PLA or ABS matrices for FDM, offers a significant advantage of high-flexibility that if conferred to a MOF composite could be useful in device fabrication. Semiflex, a particular TPU that demonstrates substantial flexibility by sustaining up to 600% elongation while simultaneously retaining structural stability with a shore hardness of 98A, was used in these studies. ZIF-8 was selected as the model MOF component due to its proven effectiveness in catalysis, chemical separations and sensing, high-stability (thermal and chemical),32 and for its commercial availability in large quantities.33-35 ZIF-8 possesses tetrahedrally coordinated zinc atoms bridged by imidazolate linkers, is readily synthesized by a variety of methods (for example: solvothermally, hydrothermally, or simply at room temperature in aqueous or organic solution) in a range of crystallite sizes (down to the nanoscale).36,37


4

Page 5 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Herein, we demonstrate the production of MOF composites with large surface area, in customizable dimensions, as a result of high metal-organic framework particle loadings in both rigid poly(lactic acid) (PLA) and flexible Semiflex polymer matrices using a benchtop 3D printer. Composite preparation by solution blending and subsequent casting into films prior to extrusion into printing feedstocks was critical to achieve a homogeneous and well-adhered MOF dispersion even at loadings as high as 50% by mass. Our formulation method yields composites with mechanical properties suitable for 3D printing free-standing materials with complex structures. In the rigid PLA matrix, a hierarchical porosity resulting from a combination of preserved MOF crystallinity, micropores, mesopores, and giant voids in the composites was observed at 40% MOF loadings. With more elastic Semiflex, large voids are absent and micropores are blocked at similarly high MOF filler ratios. Access to MOF micropores in Semiflex composites was instead achieved by a procedure where sacrificial fluoropolymer was doped in, maintained throughout printing, then selectively removed. The use of sacrificial polymeric additives to induce porosity is a relatively unexplored area generally in 3D printing. The reported printed composites remain open to gaseous (CO2) and liquid guests (solvents). The fabrication strategies were also extended successfully to another class of MOF (UiO-66) to illustrate the potential generality of this method. The approach we developed offers access to a new class of 3D printing materials that retain valuable framework and functional characteristics while also imparting new structural properties from the polymeric foundation. EXPERIMENTAL Materials and General Methods. ZIF-8 powder (Basolite Z1200) and poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP, pellets, PN 427187, melt index = 4–10 g/10 min @ 230 °C) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Natural poly(lactic) acid


5


Page 6 of 33

(PLA) pellets were obtained from Push Plastic (Springdale, AR, USA) and Semiflex (as 1.75 mm diameter filament), a thermoplastic polyurethane (TPU) polyester elastomer with shore hardness of 98A, from Ninjatek (Manheim, PA, USA). Zirconium 1,4-dicarboxybenzene MOF (UiO-66) powder was purchased from Strem Chemicals (Newburyport, MA, USA). Solvents used in these studies were of reagent grade at a minimum and were obtained by commercial suppliers (typically Sigma-Aldrich or Fisher Scientific (USA)). All materials were used as received. A Branson 2510 was used for bath sonication and a Hielscher UP50H operating at 60% power for ultrasonic horn processing. Preparation of PLA/ZIF-8 Composites. A PLA stock solution (0.05 g mL-1) was prepared by dissolving PLA pellets (7.5 g) in CHCl3 (150 mL) while stirring at room temperature. A typical composite with a total mass of 2 g and a ZIF-8 loading of 40% by mass was achieved first by dispersing ZIF-8 (800 mg) in ethyl acetate (10 mL) by bath sonication for 15 min. The PLA stock solution (24 mL) was then added to the MOF solution, briefly stirred (~1 min), and the mixture was sonicated by ultrasonic horn for 20 min. The resulting mixture was cast onto a Teflon dish, most of the solvents were removed by placing the dish on a hot plate at 75 °C, then the composite was fully dried at 75 °C for ≥ 16 h in a vacuum oven under low vacuum to obtain a thick film. Preparation of Semiflex/ZIF-8 Composites. A Semiflex stock solution (0.1 g mL-1) was prepared by dissolving Semiflex filament (5.0 g) in DMF (50 mL) while stirring at 55 °C. Upon complete dissolution, the stock solution was stored at room temperature. A typical composite with a total mass of 2 g and ZIF-8 loading of 40% by mass was achieved first by dispersing ZIF8 (800 mg) in acetone (15 mL) by bath sonication for 15 min. The Semiflex stock solution (12 mL) was then added to the MOF solution, blended, acetone removed by rotary evaporation, cast,


6

Page 7 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


then dried as with the PLA composite. For a composite with composition (50:40:10)Semiflex/ZIF-8/PVDF-HFP and a total mass of 2 g, ZIF-8 (800 mg) was first dispersed in acetone (15 mL) by bath sonication for 10 min. A PVDF-HFP stock solution (0.1 g mL-1 in DMF, 2 mL) was added to the MOF solution and blended by bath sonication for 10 min. Lastly, a Semiflex solution (0.1 g mL-1 in DMF, 10 mL) was added to the ZIF-8/PVDF-HFP mixture, stirred by gently shaking the mixture (~ 1 min), then blended by ultrasonic horn for 10 min. The composite was cast and dried as above. Filament Extrusion and 3D Printing. The composite films were cut into discrete sections and loaded into the heating barrel of a custom extruder (Figure 3b). The material was allowed approximately two minutes to reach the targeted barrel temperature of 185 °C. The material was then pressed through a 1.56 mm die to form the feedstock filaments. Using this simple extrusion method, filament diameter for a (60:40)-PLA/ZIF-8 feedstock was 1.67 ± 0.05 mm (n = 21 measurements, along a ~1 m section) and filament diameter for a (40:40:20)-Semiflex/ZIF8/PVDF-HFP was 1.64 ± 0.05 mm (n = 21, along a ~0.8 m section of filament). For all subsequent experiments reported, the feedstock was then printed through a 400 µm diameter brass nozzle (E3D, Oxfordshire, UK) mounted onto a MakerBot Replicator 2X printer pre-heated to 200 °C with a bed temperature of 60 °C. Filament feedstock could also be printed through hardened steel nozzles (E3D) (Figure S1) using the same printing conditions. Solvent exchange for MOF activation and post-processing treatment of ternary Semiflex/ZIF-8/PVDF-HFP composites. Extruded or printed samples were submerged in a solution of CH3OH (min. 48 h). The solution was then removed and the sample was “activated” by drying in a low vacuum oven at 75 °C (min. 3 h) prior to further analysis. Post-printing treatment of Semiflex/ZIF-8/PVDF-HFP composites was performed by submerging the material


7


Page 8 of 33

in boiling acetone for 10 min, removing from the acetone solution, washing the material with deionized water (3x), and drying the sample in a low vacuum oven as with the activation procedure. Characterization procedures. Nitrogen physisorption isotherms were measured on a Quadrasorb SI (Quantachrome Instruments). Prior to measurement, samples were degassed under vacuum at 105 °C for ≥ 18 h. Measurements were performed at -196 °C with 3 min pressure equilibration times. Surface area values were determined using the BET method. Appropriate pressure points (min. 7 points) used to calculate the BET surface area were selected by satisfying the criteria originally outlined by Rouquerol and discussed in detail in relation to MOF-polymer composites previously.38 The water vapor adsorption isotherms were obtained using a water vapor adsorption analyzer (VTI-SA+, TA Instruments). Typically, a sample was regenerated at 105 °C with dry N2 flow for 12 h before the measurements. The relative humidity was achieved by controlling the ratio of the flow rates of the moisture stream out from the humidity generator using N2 as the carrier gas. Pure CO2 isotherms were collected at 25 °C and up to 1 bar using a Quantachrome Autosorb iQ2 gas-adsorption instrument. Thermogravimetric analyses (TGA) were run on a Netzsch TG 209 F3 in Al2O3 pans. The samples were initially held at 40 °C for 20 min then heated to 800 °C at a ramp rate of 5 °C min-1 in an air atmosphere. The PXRD was characterized on a D8 Discover XRD unit equipped with a rotating Cu anode (1.54 Å), göbel mirror, 0.5 mm collimator, and 0.5 mm pin hole. A GADDS® area detector system positioned at 28.0° 2θ with a measured distance from the sample of 15 cm was used to capture diffraction images. Collection of individual XRD tracings required 200 seconds with power settings of 45 kV and 200 mA. Initially, images were processed with Bruker-AXS GADDS® software before importing into JADE® XRD software to obtain peak positions (° 2theta) and intensities. Helium


8

Page 9 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ion microscopy (HeIM) was performed using a Zeiss Helium Ion Orion Plus. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) data was acquired on a FEI Helios 660 Dual Beam and EDAX Team EDS software with an Apollo detector. Secondary electron (SE) and backscatter electron (BSE) images along with elemental maps were obtained in plane view and on cross-sections of printed samples. Samples for cross-sectional analysis were prepared by standard metallographic specimen preparation techniques where the printed materials were placed vertically into mounts, sandwiched between two glass slides, then polished with diamond abrasive. Elemental analysis of the bulk F content in printed samples was performed by the ion selective electrode method at Robertson Microlit Laboratories (Ledgewood, NJ, USA). RESULTS AND DISCUSSION 3D printing of rigid PLA/MOF compositesSolution casting was the first step in the preparation of all MOF composite filaments in this study, a technique previously proven effective for generation of MMM with high-particulate loadings.39 PLA/ZIF-8 composites were obtained by suspending commercial ZIF-8 (Basolite Z1200) in ethyl acetate (EtOAc) by sonication, followed by addition of a dilute PLA solution (0.05 g/mL in CHCl3) to the MOF suspension. After additional sonication, the mixtures were cast and dried. The resulting thick films were cut, fed into the heated (185 °C) barrel of a custom-built extruder, and pressed through a die to yield filament composites with diameters of ca. 1750 µm. Filaments were prepared with up to 40% ZIF-8 loading by mass with no flaking or powdering of the MOF upon visual inspection suggesting good adhesion to the PLA matrix (further examined in a subsequent section with helium ion microscopy (HeIM)). By comparison, the direct addition of ZIF-8 powder into polymer solutions resulted in large aggregation of ZIF-8 and clumping within the


9


Page 10 of 33

polymer matrix regardless of subsequent extended sonication treatments. Structural analysis of the (60:40)-PLA/ZIF-8 filament feedstock (1750 µm diameter) by PXRD confirmed that the ZIF-8 framework retained a very high crystallinity after the processing steps (Figure 1a).

Figure 1. (a) Powder X-ray diffraction (PXRD) data for (60:40)-PLA/ZIF-8 as an extruded 1750 µm (diameter) filament (top) and the simulated ZIF-8 PXRD pattern (bottom) generated from single crystal data. (b) TGA data recorded in air of PLA, (60:40)-PLA/ZIF-8 as a 400 µm printed strand before and after activation by CH3OH treatment and evacuation. (c) N2 adsorption isotherms recorded at -196 °C on 400 µm strands (as printed) and after activation with corresponding desorption isotherms denoted with open markers. The PLA filaments were then used to FDM-print materials, in the form of 400 µm cylindrical strands as a standardized geometry, for further characterization of the physical characteristics of the embedded MOFs in the composites. TGA measurements recorded in air on printed (60:40)-


10

Page 11 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


PLA/ZIF-8 (Figure 1b, dashed pink trace) featured a significant mass loss (10.4%) from 100-200 °C, expected to be a result of EtOAc and CHCl3 guest solvent molecules retained in the pores even after extensive drying. It is established that ZIF-8 can retain trapped solvent molecules in its pores at temperatures exceeding the boiling points of the guest molecules.32 To test this hypothesis, a printed ZIF-8 composite was “activated” by soaking the sample in CH3OH for 48 hours to facilitate exchange of these solvents with easier to remove CH3OH molecules, then evacuated. PLA is completely insoluble in CH3OH, therefore no morphological changes were observed in the test object. The TGA trace for the activated (60:40)-PLA/ZIF-8 composite (Figure 1b, solid blue trace) significantly reduced the mass loss observed, relative to the as printed sample, to 3.0 % (from 100-200 °C) which indicated successful exchange. In addition, we observed a remarkable decrease in the thermal stability (in air) of the PLA matrix from a decomposition onset of ~293 °C in pure PLA to ~200 °C for both as-printed and activated composites. Although, it should be noted that mass loss from remaining strongly bound CH3OH guest molecules present within the MOF after solvent exchange may overlap and convolute the precise determination of PLA decomposition onset. The thermal stability of PLA in air was found to correlate with the concentration of ZIF-8, where increased ZIF-8 significantly reduced the stability (Figure S2). ZIF-8 particles are proposed to act as heterogeneous catalysts for the hydrolysis of PLA and therefore reduce the thermal stability as has been observed previously with melt-blended PLA/ZIF-8 composites at low ZIF-8 loadings and in PLA/ZnO nanoparticle composites.40,41 As the extrusion and printing processes only expose the composites to elevated temperatures for short time intervals (maximum 3-5 minutes at 185 °C during extrusion; and ~30 seconds to a few minutes up to 200 °C), the decomposition of PLA is not expected to occur to an appreciable extent when printing. Importantly, the thermogravimetric


11


Page 12 of 33

behavior of the composite suggests the MOF pores remain accessible to guest molecules as validated by their ability to undergo solvent exchange. The surface areas (SAs) of extruded and printed ZIF-8/PLA composites, an important metric to further assess the gas accessibility to the MOFs and their presentation within the composites, were determined using the Brunauer-Emmett-Teller (BET) method from nitrogen adsorption experiments at -196 °C. Extruded filament (d = 1750 µm) with 40% ZIF-8 loading and no other post-processing yielded a very low BET surface area (

Chemically Active, Porous 3D-Printed Thermoplastic Composites

Recommend Documents