Chemiresistive Detection of Gaseous Hydrocarbons and Interrogation

both a measurable room-temperature conductivity (10−8 S/cm) and permanent ... a single Debye-type relaxation process with a single time constant (Fi...
0 downloads 0 Views 651KB Size
Subscriber access provided by ECU Libraries

Article

Chemiresistive Detection of Gaseous Hydrocarbons and Interrogation of Charge Transport in Cu[Ni(2,3-pyrazinedithiolate)] by Gas Adsorption 2

Michael L Aubrey, Matthew T. Kapelewski, Jonathan F Melville, Julia Oktawiec, Davide Presti, Laura Gagliardi, and Jeffrey R. Long J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 04 Mar 2019 Downloaded from http://pubs.acs.org on March 4, 2019

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 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 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.

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 11 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

Journal of the American Chemical Society

Chemiresistive Detection of Gaseous Hydrocarbons and Interrogation of Charge Transport in Cu[Ni(2,3pyrazinedithiolate)2] by Gas Adsorption Michael L. Aubrey,1 Matthew T. Kapelewski,1 Jonathan F. Melville,1, Julia Oktawiec,1 Davide Presti,2 Laura Gagliardi,2 Jeffrey R. Long1,3,4* 1Department of Chemistry, University of California, Berkeley, CA 94720, USA. 2Department of Chemistry, Minnesota Supercomputing Institute, and Chemical Theory Center, University of Minnesota, Minneapolis, MN 55455, United States 3Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. 4Department of Chemical and Biomolecular Engineering, University of California, Berkeley, CA 94720, USA.

ABSTRACT: The development of new chemiresistive materials for use in chemical sensors that operate near ambient conditions could potentially reduce the costs of implementation, encouraging their use in new areas. Conductive metal–organic frameworks represent one intriguing class of materials for further investigation in this area, given their vast structural diversity and the specificity of adsorbate interactions afforded by their crystallinity. Here, we re-examine the electronic conductivity of the desolvated and acetonitrile-solvated microporous framework Cu[Ni(pdt)2] (pdt2– = 2,3-pyrazinedithiolate), and find that the conductivity in the pristine material is 200-fold greater than in the solvated state, highlighting the sensitivity of sample conductivity to guest inclusion. Additionally, the desolvated material is demonstrated to selectively adsorb the gaseous hydrocarbons ethane, ethylene, acetylene, propane, propylene, and cis-2-butene at ambient temperature. Investigation of the effect of gas adsorption on conductivity using an in-situ measurement cell reveals a chemiresistive response for each adsorbate, and the change in conductivity with adsorbate pressure closely follows an empirical model identical in form to the Langmuir-Freundlich equation. The relative sensitivity of the framework to each adsorbate is, surprisingly, not correlated with binding strength. Instead, the differences in chemiresistive response between adsorbates are found to correlate strongly with gas phase specific heat capacity of the adsorbate. Nanoconfinement effects, manifest as a relative deviation from the expected chemiresistive response, may influence charge transport in the case of the largest adsorbate considered, cis-2-butene. Time-resolved conductance and adsorption measurements additionally show that the chemiresistive response of the sensor equilibrates on a shorter time scale than gas adsorption, suggesting that interparticle contacts limit conduction through the bulk material and that conductivity at the crystallite surfaces is most responsive to gas adsorption.

INTRODUCTION The quantification of gas mixture composition is a critical capability for any industry that requires atmospheric control or employs reactive gases, including the chemical refinement, semiconductor, and medical industries.1 Gas sensors are typically constructed from three basic components: a sensing material that undergoes a change in physical properties upon interaction with analyte, a complementary transducer, and the electronics required to measure the electrical response of the transducer.2 Sensors based on chemiresistivity—a change in electrical conductance upon interaction with a target compound—are particularly attractive, as they eliminate the need for an expensive transducer, such as a piezoelectric sensor, photometer, or spectrometer. However, suitable chemiresistive materials do not exist for many environments and gas compositions. The development of new chemiresistive gas sensors may therefore enhance control over existing processes and environments and promote greater adoption of sensors in industry, as a result of anticipated cost reductions. Some of the most versatile chemiresistive materials are composites of semiconducting metal oxides (e.g., TiO2 and SnO2).3-5 Sensor technologies based on such materials have

been used in many applications, including hydrogen–carbon monoxide and hydrocarbon-based gas mixtures. However, they remain fundamentally limited by the high operating temperatures required, low analyte selectivity in certain instances, and a lack of rational design methodologies for tailoring sensor materials to specific gas mixtures and operating conditions. One possible materials solution to the challenge of designing improved chemiresistive sensors would be to replace the metal oxide sensor component with a metal–organic framework—a porous, crystalline solid composed of metal nodes connected via organic linkers. The vast structural diversity and modular constitution of metal–organic frameworks are distinct advantages compared to metal oxides, and are properties largely responsible for the widespread interest in these materials for other applications, including gas storage and catalysis.6 For example, the hybrid nature of metal–organic frameworks may support the amalgamation of disparate properties that rarely coexist—such as crystallinity, microporosity, and conductivity—to yield new materials with synergistic properties.7-10 Recent advances in the chemistry of electronically conductive metal–organic frameworks11,12 have introduced the prospect of coupling charge transport and

ACS Paragon Plus Environment

Journal of the American Chemical Society 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 11

conductance of the material in response to pure component adsorption can be directly correlated to gas adsorption measurements collected on the same equipment under identical conditions. Accordingly, it is possible to determine accurate conductivity-composition profiles in order to elucidate how specific adsorbates modulate conductivity in this rare example of a porous conductor.

selective host guest interactions. Fundamentally, highly porous materials with

EXPERIMENTAL PROCEDURES General procedures and measurements. The synthesis of Na2[Ni(pdt)2] was carried out according to the literature procedure.31 Diffraction data were collected using a Bruker AXS D8 Advance diffractometer with the generator set at 40 kV and 40 mA, and samples were loaded onto zero-background sample holders. Elemental analyses were obtained from the Microanalytical Laboratory at the University of California, Berkeley. Modified Synthesis of Cu[Ni(pdt)2]. The synthesis of Cu[Ni(pdt)2] was adapted from a previous report.31 A flask was charged with 270 mL of acetonitrile and sparged with dry N2 for 30 min. Next, Na2[Ni(pdt)2] (0.630 g, 1.62 mmol) was added to the solvent and the solution was stirred. A separate flask was charged with 75 mL of acetonitrile, sparged with dry N2 for 30 min, and CuI (0.309 g, 1.62 mmol) was added. While under a positive atmosphere of N2, the CuI solution was slowly added to the solution of Na2[Ni(pdt)2] via a syringe, which resulted in the rapid precipitation of Cu[Ni(pdt)2] as a black solid. The solid was filtered on 0.22-µm nylon filter paper in air and activated overnight at 90 ºC under high vacuum on a Micromeritics ASAP 2420 instrument equipped with a turbo pump, yielding 0.453 g (69%) of the product as a dark red microcrystalline solid. The X-ray powder diffraction pattern for the activated material is shown in Figure S1 and the gas adsorption isotherm used to calculate the Langmuir surface area is provided in Figure S2. Elemental Analysis for Cu[Ni(pdt)2]: calculated., C 23.63%, H 0.99%, N 13.78%, S 31.54%; found, C 22.92%, H 1.05%, N 13.74%, S 30.43%. Gas Adsorption Measurements. Gas adsorption data were measured between 0 and 1.1 bar using a Micromeritics 3-Flex instrument. Samples of Cu[Ni(pdt)2] powder were loaded into a pre-weighed tube, and heated at 90 ºC overnight. The mass of the activated sample was used as the basis for the adsorption measurements. After an adsorption isotherm was measured, the sample was reactivated at 90 ºC before measuring a subsequent adsorption isotherm. Ac Impedance Measurements. In an Ar-filled glovebox, activated Cu[Ni(pdt)2] was pressed into a pellet inside a 2contact PEEK screw cell, as previously reported.33 The ac conductivity of Cu[Ni(pdt)2] was determined using a Bio-Logic VMP-3 multipotentiostat equipped with an impedance analyzer operating between the frequencies 1 MHz and 0.01 Hz at a sinus amplitude of 100 mV and dc bias voltage of 0 mV (Figure S3). The observed spectrum was ohmic with dc bias voltage (Figure S4), linear with sinus amplitude (Figure S5), time invariant (Figure S5), and matched closely to what would be expected for a single Debye-type relaxation process with a single time constant (Figures S3, S6–S8).34,35 Time Resolved Conductivity Measurements upon Gas Adsorption. Conductivity and gas adsorption data were measured simultaneously as a function of time using a custombuilt conductivity cell (Figure 2). Printed circuit board arrays were designed using the free computer-aided design software

Figure 1. (Top) A portion of the crystal structure of Cu[Ni(pdt)2] illustrating the one-dimensional pore structure normal to the page. Teal, green, yellow, blue, and grey spheres represent Cu, Ni, S, N, and C atoms, respectively; H atoms are omitted for clarity. (Bottom) Illustration of the redox-active Ni(pdt)2 subunit.

crystallographically-defined surfaces may also provide a unique testing ground for exploring how surface–adsorbate interactions modulate electronic transport—given that these surface interactions are expressed as changes in the bulk chemical structure. Further developing synthetic control over charge mobility in porous crystals may also promote the use of such materials in new technologies surrounding energy storage,13-15 electrocatalysis,16-20 and electronic devices.21-23 Previous efforts towards the synthesis and integration of metal–organic frameworks into sensor devices have been recently reviewed.24,25 Notably, there have been a number of breakthroughs with regard to chemiresistive applications of metal–organic frameworks in the presence of either redox active gases26 or adsorbed solvents and volatile organic guest species.27-30 For example, a chemiresistive sensor array constructed from two-dimensional triphenylene-based Cu2+ and Ni2+ frameworks has been shown to be capable of classifying a diverse range of volatile organic compounds based on chemical functional group.29 In this case, the effect of guest binding on conductivity was found to be rather complex and somewhat unpredictable, opening up an enticing opportunity to explore the effects of adsorption on conductivity in greater detail. Motivated by this study, we chose to investigate the previously reported three-dimensional conductive framework material Cu[Ni(pdt)2] (pdt2–=pyrazine-2,3-dithiolate; Figure 1) as a case study for chemiresistivity in metal–organic frameworks.31 This particular phase was an early example of a permanently porous, conductive metal–organic framework—in contrast to the isostructural compound Cu[Cu(pdt)2]32—and remains a rare case of a metal–organic framework exhibiting both a measurable room-temperature conductivity (10−8 S/cm) and permanent porosity. Herein, we examine the electrical conductance of Cu[Ni(pdt)2] as a function of absolute pressure in the presence of selected light hydrocarbons. Changes in the

2 ACS Paragon Plus Environment

Page 3 of 11 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

Journal of the American Chemical Society

package PBC123 and submitted to the rapid prototyping service Sunstone Circuits for

Figure 2. Schematic of the conductivity cell used for in situ conductance measurements on a gas adsorption analyzer. A pressed pellet of the polycrystalline Cu[Ni(pdt)2] sample was embedded between two copper rails represented by yellow strips, which sit on top of the green circuit board; the resistance of the completed circuit is then measured.

using a custom-built gas dosing manifold. Each dose of gas was equilibrated on the sample for over half an hour, until no further change was observed in both the pressure above the sample and in the resulting diffraction patterns. The average wavelength of measurement was 0.45212 Å. In between dosing C2H2 and dosing C2H4, the sample was activated at 120 ºC under vacuum for 1 h to remove all traces of the previously dosed gas from the material. The re-evacuated sample had a diffraction pattern that was identical to the initial evacuated sample. An extended description of the structure refinement can be found in the Supplementary Information. Computational Methods. The parallel version of the CRYSTAL17 suite of programs was used for all calculations.36,37 The structure of activated Cu[Ni(pdt)2], as well as the material containing either two C2H2 or C2H4 molecules per unit cell was optimized while keeping the lattice parameters fixed at their experimental values. This prevented the unit cell from becoming too compressed due to the overestimation of dispersive interactions. Frequency calculations were carried out for the optimized structures to confirm that they were in a local minimum. The original tetragonal symmetry was removed in order to define the spin on each metal center in the unit cell. The unrestricted PBE038-D3(BJ)38,39 hybrid density functional was employed, with the pob_TZVP basis-set,40 which consists of a Gaussian-type, all-electron triple-ζ basis set derived from the popular def2-TZVP,41 the former being devised especially for solid-state calculations. Grimme’s -D3 a posteriori correction scheme39 for dispersive interactions was coupled with the Becke-Johnson’s (BJ) damping function.42 A very fine DFT grid (keyword XXLGRID37) was utilized whereas, based on total energy convergence, a Pack-Monkhorst grid of 112 k-points (keyword SHRINK 6 6) within the first Brillouin zone was chosen. The tolerance thresholds on bielectronic Coulomb and exchange integrals (see keyword TOLINTEG), were set to 10-12 10-12 10-12 10-14 10-22. The DIIS acceleration algorithm43 was used throughout. Starting guesses of the density matrix for different spin multiplets were originated by assigning the spin to each metal center with the keyword ATOMSPIN, at cycle 0 of the self-consistent procedure (SCF). No locking of unpaired electrons—that is, no locking to a particular spin configuration guess—was imposed in the successive cycles of the SCF.

fabrication. The arrays consisted of sets of two via and two 1oz copper rails separated by a 0.152-µm gap printed onto an FR4 base. A segment of ½-inch stainless steel tubing was cut, deburred, and fitted with Swagelok ferrules. Individual boards were cut to fit within the ½-inch tubing and silver coated copper wires were soldered into place. The board was then threaded through the pre-cut tubing, and the bottom two-thirds of the tube was filled with a solventless high-vacuum epoxy (TorrSeal® or Hysol 1C) such that the board was fully exposed above the end of the tube (Figure 2). The gas-tight connection to the Micromeritics 3-Flex instrument consisted of a ½-inch union, a reducing union, and a two-way Swagelok ball valve for air-free sample transfers. The circuit board and adapter were transferred to an Ar-filled glovebox, and a thin pellet (~50 µm) of Cu[Ni(pdt)2] powder was formed between the two copper rails using a mechanical press. The circuit board was sealed in the measurement cell in the Ar-filled glovebox and the entire apparatus was then connected to the analysis port on the gas sorption analyzer. After multiple cycles between 1 bar and