Article pubs.acs.org/Langmuir
Single-Layer Graphene as a Stable and Transparent Electrode for Nonaqueous Radical Annihilation Electrogenerated Chemiluminescence Teresa C. Cristarella,† Adam J. Chinderle,† Jingshu Hui,†,‡ and Joaquín Rodríguez-López*,† †
Department of Chemistry and ‡Department of Material Sciences and Engineering, University of Illinois at Urbana−Champaign, 600 South Matthews Ave., Urbana, Illinois 61801, United States S Supporting Information *
ABSTRACT: We explored the use of single-layer graphene (SLG) obtained by chemical vapor deposition, and transferred to a glass substrate, as a transparent electrode material for use in coupled electrochemical and spectroscopic experiments in nonaqueous media through electrogenerated chemiluminescence (ECL). SLG was used with classical ECL luminophores, rubrene and 9,10-diphenylanthracene, in an inert environment to generate stable electrochemical responses and measure light emission through it. As an electrode material, SLG displayed excellent stability during electrochemical potential stepping and voltammetry in a window that spanned at least from ca. −2.4 to +1.8 V versus SCE in acetonitrile and acetonitrile/benzene. Although the peak splitting between forward and reverse sweeps in voltammetry was larger in comparison to metal electrodes due to in-plane resistance, SLG displayed sufficiently facile electron transfer properties to yield stable voltammetric cycling and ECL. SLG electrodes patterned with poly tetrafluoroethylene permitted the stable generation of radical ions on an SLG microelectrode to be studied through scanning electrochemical microscopy in the generation/collection mode. SLG was able to stably collect radical ions produced by a 50 μm gold tip with up to 96% collection efficiency. The transparency of graphene was used to obtain accurate spectral responses in ECL. While inner filter effects are known to cause a shift in peak emission wavelength of spectroelectrochemical studies, the use of SLG electrodes with detection through the graphene window reduced apparent peak shifts by up to 10 nm in peak wavelength. This work introduces SLG as a virtually transparent, electrochemically active, and chemically stable platform for studying ECL in the radical annihilation mode, where large electrode polarizations could compromise the chemical stability of other existing transparent electrodes.
■
INTRODUCTION Single-layer graphene (SLG) has garnered much interest for use in electronics, biosensors, and electrochemistry due to its unique electronic behavior and optical transparency.1−7 SLG obtained through chemical vapor deposition (CVD) can be prepared to yield electrodes of a few cm2 and can be easily transferred to a variety of substrates, such as silicon and transparent materials such as glass. This method results in electrodes where the basal plane of SLG is exposed to species in solution.6,7 The use of this type of monolayer carbon, in contrast to multilayer solution-processed graphenes or graphene oxides,8−10 offers unparalleled optical transparence, ca. 98% in the visible range.11 While this property could be useful in spectro- and photoelectrochemical experiments, the © 2015 American Chemical Society
exploration of the performance and stability of this monolayer electrode when participating in electrochemical reactions, especially in nonaqueous media and in highly anodic and cathodic polarizations, is still incipient. With a smaller density of structural and chemical defects than exfoliated graphenes,7,12 experimentation with SLG could lead to a better understanding of the role of defects and grain boundaries in the reactivity of carbon materials7,12−17 Nonetheless, there is also a common misconception that SLG is not sufficiently active or stable in solution for its use as a substrate electrode. Here, we Received: December 30, 2014 Revised: March 11, 2015 Published: March 17, 2015 3999
DOI: 10.1021/la5050317 Langmuir 2015, 31, 3999−4007
Article
Langmuir
as a substrate both at steady state and with high collection efficiency. Previous work utilizing radical annihilation ECL has observed steady-state emission from rubrene (RU) by using SECM44 as well as emission from individual chemical events.45 This work was limited by the use of ITO since it can only produce stably radical cations. We offer SLG as an alternative transparent electrode for electrochemical studies coupled with optical measurements. Since graphene is stable at both oxidizing and reducing potentials it is an interesting material for SECM and SECM/ECL coupled studies.1,44,46,47 We show that SLG maintains stable behavior in a ∼4 V potential window producing stably both radical anions and radical cations. Although atomically thin, here we show that SLG’s electronic conductivity and optical transparency make it an ideal platform for spectroelectrochemical experiments.
demonstrate the suitability of SLG as a versatile platform for investigating coupled spectral and electrode processes in an electrolytic medium. This study opens new avenues into various fields such as the use of SLG in light display technology, energy storage, electrocatalysis, and photocatalysis among other relevant research fields. Here, we take advantage of the transparency of SLG, as well as its stability in nonaqueous electrochemistry experiments, for the generation of light through electrogenerated chemiluminescence (ECL) in the radical annihilation mode.17−34 ECL can be produced by two different modes: coreactant and annihilation.33,34 Coreactant ECL is typically conducted in aqueous solution and consists of a single potential step experiment where both the luminophore and a sacrificial coreactant are simultaneously reacted at the electrode to produce emission. Previous work has shown use of graphene as a support for [Ru(bpy)3]2+-impregnated thin films to probe coreactant ECL in an aqueous medium.35 In the present work, we use pristine SLG to produce ECL via the annihilation mechanism, a schematic of which is shown in Scheme 1. In this
■
EXPERIMENTAL SECTION
Chemicals. Rubrene (≥98%), electrochemical grade supporting electrolyte tetra-n-butylammonium hexafluorophosphate (TBA·PF6), and anhydrous acetonitrile (99.93%) were obtained from SigmaAldrich Co. (St. Louis, MO) and used as received. 9,10Diphenylanthracene (DPA, 99%) was purchased from Aldrich (Milwaukee, WI) and twice recrystallized with xylenes before use. Benzene was purified by passage through a bed of activated alumina immediately prior to use (Glass Countour, Laguna Beach, CA). Graphene Preparation. SLG was grown by chemical vapor deposition on Cu foil (25 μm, Alfa Aesar).19 Prior to growth in the CVD, the copper foil was treated with acetone (10 s), water (DI) (10 s), glacial acetic acid (Fischer Scientific, 10 min), water (DI) (10 s), acetone (10 s), and isopropyl alcohol (10 s) before growth. Hydrogen was flowed at 500 sccm, and argon was flowed at 1000 sccm for 5 min to remove any oxygen. The system was pumped to 40 mTorr of pressure, and then the flow of hydrogen was reduced to 50 sccm for the remainder of the growth. The temperature was ramped to 1000 °C, and the flow of argon was stopped before methane was introduced. SLG was grown under the flow of 100 sccm of CH4 for 25 min. The system returned to room temperature and pressure while 500 sccm of argon was flowed. A layer of poly(methyl methacrylate) (PMMA) was added for support during transfer. 495 PMMA A2 (MicroChem) was first spin-coated on top of the graphene at 3000 rpm for 60 s. Two layers of 950 PMMA A4 (MicroChem) were then spin-coated at 3000 rpm for 60 s. Copper etchant solution (Transene Company, CE-100) was used to remove the copper from beneath the graphene/PMMA. The graphene/PMMA was then transferred to water and washed three times to rinse away the etchant salts. The graphene/PMMA was then transferred to 0.1 M EDTA (Fischer, 99%) for 1 h to remove residual metal ions. The graphene/PMMA was then transferred to water and washed three times to rinse away any remaining chelating agent. The graphene/PMMA was then transferred onto a piece of glass slide that was half coated in gold to make a good electrical contact and blown with argon to dry. To remove the PMMA, the sample was placed in anisole for 3 h, then placed in a dichloromethane/acetone mixture (1:1) for 5 h, and finally placed in isopropyl alcohol for 2 h. The sample was then blown with argon to dry. To form SLG microelectrodes, as well as acting as a protective barrier between the SLG and the O-ring of the electrochemical cell, polytetrafluoroethylene (PTFE) was selectively deposited on the surface in accordance with an existing method.48 Briefly, a positive photoresist was patterned on the substrate and allowed to dry. A PTFE precursor was loaded onto the substrate via spin-coating. Once dry, the sample was shaken in acetone to remove the photoresist, leaving a pore in the PTFE where the photoresist was. The PTFE was cured in a furnace under an inert atmosphere at 350 °C for 20 min. Raman spectra of SLG samples were collected with a Nanophoton Raman-11 laser Raman microscope using a 532 nm laser. Atomic force microscopy (AFM) analysis in the tapping mode was performed on an Asylum Cypher AFM using silicon cantilevers coated in aluminum.
Scheme 1. Radical Annihilation ECL of a Generic Mediator, Using 9,10-Diphenylanthracene (DPA) as Examplea
a The last step shows emission of photon of energy hν, blue light in the case of DPA. Standard potentials are given versus a saturated calomel electrode (SCE).
work, we demonstrate that large area graphene electrodes can be polarized to yield an electrochemical window as required for the generation of the highly reactive radical ions in Scheme 1. Despite this large polarization, we show that SLG electrodes display stable electrochemical operation. As an allotrope of carbon, SLG is sufficiently conductive to behave as an electrode, similar to boron-doped diamond, glassy carbon, and highly ordered pyrolitic graphite (HOPG), but offers an enhanced transparency in comparison to these materials.16,35−38 Traditional ECL measurements are known to display shifts in their emission spectra when compared to photoluminescence (PL) spectra due to inner filter effects.29,31,39−41 Unlike PL, in ECL the light must pass through a more concentrated solution of the luminophore in order to achieve the same light intensity. The use of a transparent electrode would allow radical annihilation ECL to be measured on the backside of the electrode. Transparent metal oxide electrodes, such as indiumdoped tin oxide (ITO) and fluorine-doped tin oxide (FTO), are the current standards for transparent electrodes, but they are mainly useful for oxidation chemistry as they become unstable upon reduction in organic media.42,43 Recent works have explored thin films of carbon nanotubes and metal nanowires as transparent electrodes, but these materials quickly decrease in transmittance in order to increase conductivity.5,11 These materials absorb at least 20% of the visible wavelengths of light when they are at their ideal maximum resistance, whereas SLG absorbs only 2.3% of visible light.11 Using scanning electrochemical microscopy (SECM), we show that it is possible to produce radical ions using graphene 4000
DOI: 10.1021/la5050317 Langmuir 2015, 31, 3999−4007
Article
Langmuir
Figure 1. Experimental design for ECL spectroscopic measurements with (A) a metal electrode surrounded by an insulating sheath. (B) SLG electrode with the fiber-optic cable positioned such that the light must pass through the solution before collection, and (C) a SLG electrode with the light measured beneath the cell. The base of a typical cell used is shown in (D) where the area within the red dashed line is the only region exposed to the solution. In (B) and (C) the presence of SLG is represented by a star.
Figure 2. Characterization of single-layer graphene electrodes deposited on glass. (A) Typical Raman spectrum and (B) UV−vis spectrum in transmittance mode of SLG on glass. The sheet resistance of SLG was derived from conductivity data obtained using an Alessi Industries in-line four point probe setup by averaging 10 measurements over 10 different sites. UV−vis measurements were obtained using a Hewlett-Packard spectrophotometer in a transmittance configuration and using a blank glass slide for baseline acquisition. Electrochemical Measurements. Electrochemical experiments were performed in an inert environment using a CHI1242b portable bipotentiostat (CH Instruments, Austin, TX). Solutions were prepared in an argon atmosphere glovebox. Cyclic voltammograms (CVs) were obtained using a silver wire as a quasi-reference electrode (QRE) and a platinum wire as a counter electrode. All potentials are reported versus the QRE unless otherwise stated. The CVs showed at which standard potentials the radical cations (Eox) and anions (Ered) were generated. Chronoamperometry was used to step between Eox and Ered to produce radical annhilation ECL. SECM Measurements. A 50 μm radius gold electrode was prepared by sealing a gold wire (Goodfellow) in a 1.5 mm o.d. and 0.75 mm i.d. patch clamp glass capillary (World Precision Instruments) using a heated metal coil. Electrical connections were made to the gold wire filling the unsealed portion of the capillary with silver epoxy (Ted Pella) and inserting a long wire before curing the epoxy. The gold was exposed, and the electrode was sharpened to a tip by manual grinding. Figure S1 depicts the coalignment of an SECM tip over a small microfabricated SLG window used for radical ion collection experiments using SECM. The gold tip was biased to Ered and approached to the PTFE-coated region of the substrate in feedback mode, displaying a decrease in current as it approached the insulating PTFE (i.e.,
negative feedback). Then the tip was then scanned across the sample using the tip generation/substrate collection mode in order to find a graphene microelectrode. When a spot of exposed SLG was found, the tip was aligned above that spot and approached with the substrate biased at 0 V to collect the radical anion from the tip. Tip generation/ substrate collection experiments were performed where a radical ion is produced at the tip and then converted to its original oxidation state by the substrate. To calculate collection efficiency, a tip generation/ substrate collection CV was obtained with the tip directly above the exposed spot of graphene. ECL and Fluorescence Measurements. To produce transient ECL, the electrode potential was stepped between an oxidizing (Eox) and reducing potential (Ered) every 0.5 s. Spectroscopic data were collected with the different configurations depicted in Figure 1. Configuration A was used with the commercial polymer-encased metal electrodes (Au and Pt), where the electrode was suspended in solution such that the ECL could be collected by a fiber optic placed below the glass base. Light had to traverse several millimeters of solution to finally exit through the glass base. Configurations B and C correspond to measurements with a SLG electrode. Case B served as a control experiment to evaluate inner filter effects. Here, light generated via ECL close to the surface of the SLG electrode had to travel though several mm of solution to be collected at the fiber optic. In configuration C, ECL was collected as it directly went through the transparent SLG electrode with the fiber optic placed beneath the base. In all cases, the collected light on the fiber optic led to a CCD spectrometer (Avantes, AvaSpec-ULS3648TEC). Spectra were collected with the shutter open for 40 s and averaged 10 times. PL 4001
DOI: 10.1021/la5050317 Langmuir 2015, 31, 3999−4007
Article
Langmuir
−2.4 to +1.8 V versus a SCE. The difference between the E1/2 values is consistent with previously reported data for DPA, i.e., 3.4 V.52,53 On the basis of calculations via the Randles−Sevcik equation, and using the parameters in Figure 3, we estimate that the active area of the electrodes was equal to 0.32 cm2. This value was consistent with the measured geometric area of the electrode, 0.28 cm2, which was exposed to the solution of DPA. This comparison ensures that the totality of the SLG was electroactive and, as will be shown in the section pertaining to ECL, that the entire electrode surface generated conditions for light emission. Once properly assembled into the electrochemical cell, the SLG electrodes could be reused for several experiments under similar conditions as those in Figure 3, provided that the electrode is treated with care. For example, the sample of SLG used in Figure 3 was used in four experiments in an interval of a week before the cell was disassembled. Table 1 shows a consistently larger peak splitting between the forward and reverse sweeps for CVs collected on SLG
emission spectra were collected with dilute solutions of RU in benzene and DPA in acetonitrile using a Cary Eclipse fluorometer.
■
RESULTS AND DISCUSSION Characterization of SLG. Raman spectroscopy, UV−vis spectroscopy, and atomic force microscopy were used to assess the quality of SLG samples deposited on glass. The Raman spectrum in Figure 2A shows a small D band around 1370 cm−1, which indicates few defects on the graphene structure.49 The ratio of 2D band (around 1590 cm−1) to G band (2690 cm−1) as well as the width of the 2D band reflects the stacked layer number of graphene19,50 In this case, the small width of the 2D line, as well as a 2D band intensity at 2698 cm−1 more than twice as large with respect to the G band at 1581 cm−1, strongly suggest that the graphene was a single layer. UV−vis transmittance measurements in Figure 2B show additionally that our typical SLG sample on glass had a high transmittance of 97.4% at 550 nm, further supporting the presence of monolayer carbon.51 We carried out four-point-probe measurements to establish the sheet resistance of our SLG samples, which was equal to 575 ± 77 Ω/□ for 10 independent measurements on 10 sites. This value is consistent with the range of reported values for wet-transferred SLG grown by chemical vapor deposition.5 AFM imaging shown in Figure S2 shows a mostly smooth surface with several small (
