Solvent-driven electron trapping and mass transport in reduced graphites to access perfect graphene

Introduction

Graphite intercalation compounds (GICs) involving alkali metals as interlayer guests1,2,3 represent a well-established class of compounds and have found practical applications, for example, as reducing agents (KC8) in organic synthesis4,5,6 or as lithium-ion batteries7,8. More recently, we9,10,11,12,13 and others14,15,16 have demonstrated that GICs are very suitable precursors for the covalent functionalization of graphene. In typical covalent functionalization sequences, the negatively charged graphene layers first act as reductants for electrophiles such as diazonium17 or iodonium12 compounds, alkyl iodides or protons18, which are subsequently attacked by the intermediately generated organic radicals or H-atoms to give arylated11, alkylated or hydrogenated10graphene. This wet-chemical functionalization concept is facilitated by the fact that due to Coulomb repulsion the negatively charged graphenide layers within the solid GICs can be dispersed in suitable organic solvents13,15,17. However, in this respect two fundamental issues have been overlooked so far: The first question is whether all negative charges of the graphenide intermediates can be controlled or even completely removed in such redox reactions19. Only the complete oxidation is expected to avoid reactions with moisture and oxygen during workup leading to side products with undesired and additional oxygen- and hydrogen functionalities. We have already reported on such side reactions in the field of reductive carbon nanotube chemistry20 and will show below that these take indeed place also with the corresponding graphenides. The second goal is if the wet-chemical exfoliation of GICs into dispersed graphenide sheets can be used for the bulk production of defect-free graphene. Also this latter question is directly associated with the possibility of a controlled removal of all negative charges from the intrinsic air-sensitive graphenide intermediates. To the best of our knowledge, there are neither scalable methods nor any simple liquid oxidizing agents reported in literature, which are able to quantitatively discharge any type of graphenide solutions without a simultaneous alteration of the carbon framework.

We report here on a significant discovery, which gives answer to these questions and solves the associated challenges. Moreover, this discovery reveals a very fundamental physical–chemical phenomenon, namely a quantitative solvent reduction induced and electrostatically driven mass transport of K+ ions from the GIC into the liquid. The fundamental finding is that the treatment of GICs with benzonitrile (PhCN), leads to a quantitative discharging of the individual graphenide sheets upon the formation of the coloured radical anion PhCN˙−, which is easy to monitor quantitatively, the accompanying exhaustive and Coulomb force-driven migration of the potassium counterions from GICs into the surrounding benzonitrile phase, the suppression of any reactions of dispersed graphenides with moisture and air that is shown to take place when no treatment with benzonitrile is provided and the successful generation of defect-free single-layer graphene suspended on silica substrates. This latter discovery represents a rather mild, scalable, and inexpensive method for the wet-chemical graphene production.

Results

Solvent-driven oxidation of GICs quantified by UV/Vis spectroscopy

Benzonitrile is mainly used as highly polar organic solvent, while the interaction with potassium GICs especially in the field of graphene research has not been fully evaluated yet. It has a comparatively low reduction potential of −2.74 V versus Ag in N,N-dimethylformamide (DMF). Generally, typical strong oxidizing agents show much higher reduction potentials, for instance tetracyanoquinodimethane with −0.2 V versus Ag in DMF21,22,23. Electropositive metals such as elemental potassium can be dissolved in PhCN, which is accompanied by the formation of K+ cations and PhCN˙− radical anions24,25,26. The red solution of PhCN˙− displays a broad absorption pattern with a peak at∼390 nm and another feature of bands at 500 nm. For the determination of the extinction coefficient of PhCN˙−, which has not been reported in literature so far, we performed a dilution series of potassium in benzonitrile. The linear correlation between amount of potassium and measured absorption reveals quantitative reactions (Fig. 1b). According to Lambert-Beer, the extinction coefficient of PhCN˙−was determined to be ɛ390=4,000 (±50) L mol−1 cm−1 (Supplementary Figure 1). This value is in the regime of literature estimates, but has never been exactly determined yet24. With this information at hand we set out to investigate the interaction between PhCN and GICs involving potassium as guest ions. For this purpose, we conducted a precise dilution series of pure potassium and various stages of GICs with different K:C ratios, namely 1:0 (pure K), 1:8, 1:16, 1:24 and 1:48. Therefore, 480 mg (40 mmol carbon) spherical graphite SGN18 and 195 mg (5 mmol) potassium were heated to 200 °C in a glass vial. The formation of the final stage I intercalation compound was verified by in situ Raman spectroscopy under inert conditions (Fig. 2) as well as X-ray diffraction (XRD) analysis (Supplementary Figure 2). After the complete formation of the first stage K GIC, the powder was allowed to cool to ambient temperature. For the synthesis of KC8 in PhCN (Fig. 1), the intercalation compounds were dissolved in absolute benzonitrile (0.15 mg ml−1) by brief ultrasonication (5 min, 20 kJ, 1 s pulse) under argon atmosphere. The resulting formation of PhCN˙− was investigated by UV/Vis absorption spectroscopy in sealed cuvettes to exclude any air exposure (Fig. 1c,d).

Figure 1: Charge quantification in GICs using UV/Vis spectroscopy.

(a) Reaction scheme for the quantitative electron transfer from various GICs to PhCN leading to dissolved K+ ions and the red coloured radical anion PhCN·−. (b) Photograph of a sealed vial containing KC8 in a concentration of 5.0 × 10−4 M in PhCN·−. (c) Absorption spectra under inert conditions of the dilution series of four different stages of GICs dispersed in PhCN. The resulting absorption profiles correlate to the benzonitrile radical anion PhCN·−. The dashed line indicates the wavelength used for the determination of ɛ390.. The exact concentration values—colour coded—are given in Supplementary Table 1. (d) Determination of the extinction coefficient by correlation of the potassium concentration versus the extinction at 390 nm. The black curve represents the dilution series of pure potassium K(s) in PhCN.