ORIGINAL ARTICLE

Controlling the morphological and redox properties of the CuTCNQcatalyst through solvent engineering

Zakir Hussain1& Ruchika Ojha2 & Lisandra L. Martin2

& Alan M. Bond2& Rajesh Ramanathan1

& Vipul Bansal1

Received: 9 January 2019 /Accepted: 21 February 2019 /Published online: 12 March 2019# Qatar University and Springer Nature Switzerland AG 2019

AbstractThe solution-based synthesis of the coordination polymer, CuTCNQ (TCNQ = 7,7,8,8-tetracyanoquinodimethane), which existsin two phases, has predominately used acetonitrile (MeCN) as the solvent. However, our knowledge on the growth and propertiesof CuTCNQ in other solvents remains limited. In this work, the synthesis of CuTCNQ on Cu foil in two protic (MeOH, EtOH) aswell as four aprotic (MeCN, DMSO, DMF, THF) solvents has allowed us to obtain new insights into the important role of thereaction medium in the spontaneous crystallization of CuTCNQ in discrete morphologies and phases. A new electrochemicalmethod for phase identification also has been developed to support this study. Findings reveal that (i) the solvents with higherdielectric constants favor CuTCNQ crystallization; (ii) irrespective of the solvent, use of high temperature (60 °C vs. 25 °C inconventional synthesis) promotes CuTCNQ crystallization and facilitate conversion of phase I to phase II; (iii) phase I CuTCNQpossess enhanced redox catalysis (ferricyanide reduction by thiosulfate) performance over phase II CuTCNQ; and (iv) theamount of catalyst is not necessarily the most important factor for driving catalytic reactions, and other factors, such as,morphology, redox characteristics and solvent in which the CuTCNQ is synthesized may dictate the overall catalytic perfor-mance. These findings emphasize the importance of understanding the influence of parameters, such as, solvent and temperaturein CuTCNQ synthesis as a means of providing materials with improved catalytic activity.

Keywords Catalysis . Metal-organic semiconductor . CuTCNQ . Charge-transfer complex .Morphological control

1 Introduction

The ambipolarity, superconductivity, and a myriad of interest-ing properties of charge transfer (CT) complexes based on7,7,8,8-tetracyanoquinodimethane (TCNQ) have seen signif-icant interest in exploring these materials for new applications[1–3]. As such, a number of strategies based on electrochemical

[4–14], vapor deposition [6, 14–20], and wet chemical routes[21–31] have been employed to fabricate these materials. Alarge body of work in this area has focused on MTCNQ [9,25], as the strong electron withdrawing ability of neutralTCNQ (TCNQ0) from a metal (M0) allows a wide breadth ofMTCNQ (M+TCNQ−) materials to be fabricated with relativeease [8, 9, 25]. The unique electronic properties of these metal-organic CT complexes have allowed them to be used in fieldemission and molecular electronics [9, 32]. Our group and othershave been able to recently expand the applicability of these ma-terials beyond the field of electronics [7, 16, 17, 21–24, 26–31,33–38]. Some examples include, nanoarrays of CuTCNQ,AgTCNQ, CuTCNQF4 (2,3,5,6-Tetrafluoro-7,7,8,8-tetracyanoquinodimethane), AgTCNQF4, CuTCNQ/AgTCNQhybrids, CuTCNQ/noble metal hybrids, and AgTCNQ/noblemetal hybrids as excellent photocatalysts [21, 24–28, 31, 39];CuTCNQF4/Ag hybrids as redox catalysts [23]; heterojunctionsof CuTCNQ/ZnO as a humidity sensor [40]; alkali-TCNQnanoarrays as gas sensors [16, 17]; AgTCNQ nanowires as anantibacterial material [22]; Fe(TCNQ)2, Mn(TCNQ)2, andCo(TCNQ)2 as electrocatalysts for oxygen and hydrogen

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s42247-019-00026-8) contains supplementarymaterial, which is available to authorized users.

* Rajesh RamanathanRajesh.Ramanathan@rmit.edu.au

* Vipul BansalVipul.Bansal@rmit.edu.au

1 Ian Potter NanoBioSensing Facility, NanoBiotechnology ResearchLaboratory, School of Science, RMIT University, GPO Box 2476,Melbourne, VIC 3001, Australia

2 School of Chemistry, Monash University, 17 Rainforest Walk,Clayton, VIC 3800, Australia

Emergent Materials (2019) 2:35–44https://doi.org/10.1007/s42247-019-00026-8

evolution reactions [33, 34, 38, 41], and CuO/TCNQ andCo(OH)2/TCNQ nanoarrays on copper foams as electrocatalysts[42–44]. These wide-ranging possibilities using MTCNQ mate-rials have seen renewed interest to improve their fabricationstrategies.

Since the initial report on MTCNQ materials by Melbyet al. [8], much of the effort in the synthesis of these materialshas focused on solution-based or vapor deposition strategies.Of these, the former is preferred as it only requires simpleimmersion of a metal foil in an acetonitrile (MeCN) solutioncontaining TCNQ0 [9, 25]. The ease of reduction (strong elec-tron withdrawing ability) of TCNQ facilitates its spontaneousreaction with many metals, resulting in the formation ofMTCNQ on the surface of metal foil [9, 25]. Although thisstrategy is simple and robust, it poses certain limitations incontrolling the aspect ratio, the overall morphology and thephase of the resulting MTCNQ structures. Since these factorscan significantly influence the properties of MTCNQ, there isa strong motivation to explore alternative fabrication strate-gies. Specifically, in the context of CuTCNQ, our group re-cently developed two independent approaches to control theaspect ratios of the resulting CuTCNQ microrods. In the firststrategy, we employed a bi-solvent approach where the addi-tion of small quantities of water to MeCN allowed dynamiccontrol over the reaction kinetics in favor of CuTCNQ crys-tallization, leading tomicrorods of ≈ 100 μm in length, insteadof 5–10 μm achieved through conventional synthesis [31]. Inanother approach, we could overcome the equilibrium barrierusing a recycling strategy, where the replenishment of thereaction solution with fresh TCNQ0

(MeCN) over multiple cy-cles favored CuTCNQ crystallization, allowing the formationof ≈ 70 μm long microrods [30]. In most studies, a polaraprotic solvent, typically MeCN, has remained the preferredreaction medium for CuTCNQ growth. Notably, CuTCNQ isknown to exist in two distinct phases (phase I and phase II) [9,45, 46]. During its synthesis in MeCN, phase I is formedunder ambient conditions, while phase II is typically observedat elevated temperatures [45]. Given that the choice of thesolvent and the synthesis temperature can have a large influ-ence on the reaction kinetics and reaction equilibrium [47, 48],it is important to understand the influence of these parameterson the morphology and the phase of CuTCNQ.

In line with the above considerations, this work is focusedon the investigation of the crystallization and growth ofCuTCNQ in a range of polar solvents at 25 and 60 °C. Toobtain a detailed understanding of the role of solvent andtemperature on CuTCNQ crystallization, we also synthesizedphase-pure CuTCNQ using well-established methods [45].These materials served as reference standards for the charac-terization of CuTCNQ formed in this study (Fig. S1,Supporting Information). The study of CuTCNQ crystalliza-tion in a range of solvents provides an in-depth understandingof factors involved in modulating the crystallization kinetics.

The outcomes reveal that the CuTCNQ crystallization kinet-ics, their morphological characteristics and phases are signif-icantly influenced by (i) the nature of the solvent (protic vs.aprotic, dielectric constant and dipole moment) and (ii) thereaction temperature. Importantly, it was discovered that themorphology of CuTCNQ has a major influence on the cata-lytic activity. However, the catalytic efficiency of morpholog-ically distinct CuTCNQ materials was independent of theamount of CuTCNQ crystallized.

2 Materials and methods

Materials Copper foil (99% pure), potassium ferricyanide, cop-per (I) iodide (Cu2I2), DMSO, DMF, ethanol, methanol, THF,sodium thiosulfate, and copper sulfate were purchased fromSigma Aldrich, Australia. 7,7,8,8-tetracyanoquinodimethane(TCNQ) was purchased from Chem Supply, Australia. MeCNwas purchased from Ajax Fine Chem, Australia. All chemicalswere used as received. Deionized water was dispensed from aMillipore Milli-Q Ultrapure Water system fitted with Organex-QCartridge filters. To remove surface oxides, before the growthof CuTCNQ, the copper foil was treated with dilute nitric acid,washed with deionized water, and dried with a flow of nitrogenimmediately prior to use.

Synthesis of CuTCNQ in different solvents A 1.0 × 1.0 ×0.50 cm3 piece of Cu foil was placed in contact with 5 mMTCNQ dissolved in the solvent of interest. The reaction be-tween Cu foil and TCNQ in each solvent was allowed toproceed for 24 h at 25 °C or 60 °C. Subsequently, the Cu foilswere washed with the relevant solvent to remove anyunreacted TCNQ from the surface. These substrates were usedfor materials characterization and catalysis.

Synthesis of pristine phase-pure CuTCNQ reference standardsPure phase I and II CuTCNQ were synthesized using a liter-ature method [45]. Phase I CuTCNQ was formed by a redoxreaction between TCNQ (hot solution dissolved in MeCN)with Cu2I2 [45]. This reaction resulted in the formation ofphase I CuTCNQ, which was precipitated, then washed withMeCN. Phase II CuTCNQwas obtained by prolonged stirringof a suspension containing the phase I CuTCNQ in MeCN atroom temperature [45].

Materials characterization The Cu foils containing CuTCNQgrown in different solvents were characterized by a range ofmethods. Instrumental techniques used for this characteriza-tion are as follows: a FEI Verios 460L FE-SEM instrumentoperated at an accelerating voltage of 10 kV; EDX analysiswith a FEI Verios 460L FE-SEM instrument coupled withEDX Oxford X-Max Silicon Drift Detector; FTIR spectrosco-py with a Perkin–Elmer D100 spectrophotometer in

36 emergent mater. (2019) 2:35–44

attenuated total reflectance mode with a resolution of 4 cm−1;Raman spectroscopy with a Perkin–Elmer Raman Station 200Fat an excitation wavelength of 785 nm and 100 μm spot size.

Electrochemical solid-solid state interconversions for phaseidentification A BASi Epsilon-EC electrochemical worksta-tion was used for voltammetric experiments. A glassy carbon(GC) working electrode was first prepared by polishing it with0.3 μm alumina (Buehler, Lake Bluff, IL, USA) on a cleanpolishing cloth (Buehler, USA), followed by washing severaltimes with deionized water under sonication for 5 min. TheGC electrode was finally washedwith acetone and dried undera N2 stream before modification with the CuTCNQ sample.The CuTCNQ samples grown on Cu foil were collected byscraping from the Cu foil surface using a spatula. TheseCuTCNQ samples, as well as phase I and II reference stan-dards were loaded onto the GC (0.00786 cm2) macro-discelectrode. Pt wire was used as a counter electrode and Ag/AgCl (3.0 M KCl) was used as a reference electrode in a 3-electrode cell. An aqueous 0.10 M CuSO4.5H2O solution wasused as a solvent and supporting electrolyte medium. Thepresence of Cu2+ in the aqueous electrolyte minimizes thedissolution of the solid from the CuTCNQ-modified electrodeupon repetitive cycling of the potential [49]. Given that theCu2+ is reduced at potentials more negative than 0.200 V, theelectrochemistry of CuTCNQ adhered to a GC electrode incontact with 0.10 M CuSO4 was studied over the potentialrange of 0.250 to 0.600 V. Samples for electrochemical studieswere prepared freshly as prolonged immersion of theCuTCNQ-loaded GC electrode in aqueous Cu2+ electrolytesolution facilitated the conversion of phase I to phase II. Allexperiments were performed in triplicate.

Catalytic reaction The ferricyanide reduction reaction was usedas a model pseudo-first order reaction to test the catalytic effi-ciency of CuTCNQ samples grown in different solvents andtemperatures. A 30 mL solution was used for the catalyticreaction in the presence of visible light. The reaction contained1.0 mM ferricyanide, 0.10 M thiosulfate, and CuTCNQ cata-lysts grown on Cu foil (1.0 × 1.0 × 0.50 cm3). Care was takento maintain the temperature of the reaction at 25 ± 1 °C understirring at 1200 rpm using a magnetic stirring bar. The catalyticconversation of ferricyanide to ferrocyanide wasmonitored as afunction of time using UV-vis absorbance spectroscopy (Cary50 Bio spectrophotometer).

3 Results and discussions

In MeCN, the synthesis of CuTCNQ proceeded via a oneelectron transfer reaction between Cu0(metal) foil andTCNQ0

(MeCN) resulting in the spontaneous crystallization ofphase I CuTCNQmicrorods on the surface of the Cu foil (24 h

at 25 °C). In accordance with the previous reports, thesemicrorods showed faceted corners with 1–4 μm diameters(Fig. 1a and Fig. S2, supporting information) [9, 25]. If thesynthesis of CuTCNQ in MeCN is performed at lower tem-perature (8 °C for 24 h), the reaction significantly slows down,but does not markedly alter the morphology of the CuTCNQcrystals (Fig. S3, supporting information). Conversely, if theCuTCNQ crystallization is performed at higher temperatures(24 h at 60 °C), an obvious change in morphology is observed(Fig. 1b and Fig. S4, supporting information). Two discreteCuTCNQ populations are noted. The predominant one ex-hibits larger structures which results from the merger of thesmaller microrods (minor population) formed at ambient tem-perature. This increase in the diameter to 50–100 μm in themajor population is a characteristic of phase II CuTCNQ [45].Based on these observations, further studies were focused oncomparing the CuTCNQ crystallization at 25 °C and 60 °C.

To understand the influence of the solvent on CuTCNQcrystallization and growth, we chose four aprotic (acetonitrile,dimethyl sulfoxide, dimethylformamide, and tetrahydrofuran)and two protic (ethanol and methanol) solvents (Table 1).When the Cu0(metal) foil was exposed to TCNQ0

(solution) dis-solved in different solvents, it was visibly obvious that thenature of the solvent plays an important role in CuTCNQcrystallization (Fig. S5, supporting information). With the ex-ception of THF, all aprotic solvents crystallized CuTCNQ onthe Cu foil surface (foil turned bluish-black). In contrast, thecolor of the Cu foil did not change when the protic solventswere used. Based on the properties of the solvents which in-duced CuTCNQ crystallization, we suggest that the dielectricconstant and/or dipole moment of the solvent has a majorinfluence on the spontaneous reaction between Cu andTCNQ. For example, the CuTCNQ crystallization was ob-served in solvents with high dipole moments and dielectricconstants, such as DMSO (3.96 D/47.24 Fm-1), MeCN (3.92D/36.64 Fm-1), and DMF (3.82 D/36.7 Fm-1). In contrast,THF with a low dipole moment and dielectric constant (1.75D/7.52 Fm-1) failed to crystallize CuTCNQ. This suggests thatpolar aprotic solvents with lower dielectric constants may nothave sufficient driving potential to crystallize CuTCNQ. Thisobservation supports our previous finding that introduction ofsmall amounts of water, a high dielectric medium (1.85D/80.1 Fm-1) inMeCN significantly promotes CuTCNQ crys-tallization [31]. Since only DMSO, MeCN and DMF showedvisual evidence of CuTCNQ crystallization, further studieswere confined to these solvents.

Figure 2 and Figs. S6–S9, supporting information, showthe morphological characteristics of CuTCNQ grown usingTCNQ0 dissolved in DMSO or DMF at 25 °C or 60 °C. At25 °C, when the reaction was carried out in DMSO, cylindri-cal rod-like structures appear perpendicular to the surface ofthe Cu foil. Although these morphological characteristics aresimilar to those seen in MeCN (Fig. 1a), the packing density

emergent mater. (2019) 2:35–44 37

of the rods is significantly higher (Fig. 2a and Fig. S6,supporting information). In the case of DMF, at 25 °C, arod-shaped morphology is observed (Fig. 2c and Fig. S8,supporting information). However, in contrast to the discreterod-like morphology observed in MeCN (Fig. 1a), the rods inDMF are fused. This results in small cube-like appearance ofthese structures, more apparent at lower magnifications. Thesefused rods are of 2–4 μm in diameter and 6–10 μm long. It isnoted that the microrods typically observed in MeCN underambient conditions are 0.5–2 μm in diameter and less than10μm in length [25–31]. If the reaction is carried out at 60 °C,in DMSO, the morphology changes from rod-like to very thinplate-like structures with an edge thickness of less than100 nm (Fig. 2b and Fig. S7, supporting information). InDMF, two distinct morphologies, viz., a minor population ofrod-like structures, dominated by 60–80 μm long and 15–20 μm wide platelets, is observed (Fig. 2d and Fig. S9,supporting information). These morphologies distinctly differfrom those observed in MeCN during conventional synthesis(Fig. 1). A plausible explanation for this morphologicalchange may be attributed to solvent-dependent surface ten-sion. It is well known that during epitaxial growth, the mor-phology of the resultant particles depends on parameters suchas, temperature, atomic flux, and the surface tension of themedium. In particular, an increase in the surface tension typ-ically leads to the formation of smoother surfaces [50, 51]. In

this work, the surface tension increases in the order MeCN <DMF <DMSO. This implies that the surface tension plays animportant role in creating smoother and thinner CuTCNQcrystals in DMSO.

Energy dispersive X-ray (EDX) spectral analyses of theCuTCNQ crystals formed in different solvents show charac-teristic energy lines corresponding to C Kα (0.277 eV) and NKα (0.392 keV) [25, 26, 30, 31] that are assigned to the pres-ence of TCNQ− in CuTCNQ (Fig. S10, supportinginformation). In addition, the expected energy lines corre-sponding to Cu Kα and Cu Kβ are observed [25, 26, 30,31]. The absence of an O Kα signal confirms minimal oxida-tion of Cu in all cases. An elemental EDXmap establishes thatthe growth of CuTCNQ remains uniform under all experimen-tal conditions (Fig. S11, supporting information).

The chemical identity of the CuTCNQ crystals was furtherassessed using Raman and Fourier-transform infrared (FTIR)spectroscopy (Fig. 3). In particular, Raman spectroscopy candifferentiate between neutral TCNQ0 and the radical TCNQ−

anion in CuTCNQ [22, 26–31]. Raman spectra (Fig. 3a) ob-tained from CuTCNQ grown under different conditions werebackground corrected using an in-house developedsmoothing-free algorithm [52]. All CuTCNQ samplescontained TCNQ−, as evident from the red shift in theRaman bands assigned to the C−CN wing stretch(1380 cm−1 in TCNQ− vs. 1450 cm−1 in TCNQ0) and theC≡N stretch (2200 cm−1 in TCNQ− vs. 2225 cm−1 inTCNQ0) [22, 26–31]. Additional bands corresponding to theC=CH bending (1200 cm−1) and C=C ring stretching(1600 cm−1) confirm the presence of the TCNQ− anion.These shifts in the Raman bands occur because TCNQ coor-dinates with Cu through its nitrile group [22, 26–31]. Theabsence of TCNQ0 bands proves that TCNQ0 from solutiondoes not directly crystallize either on the surface of the Cu foilor CuTCNQ.

The phase of the CuTCNQ cannot be distinguished byRaman spectroscopy alone, as evident from the Raman spec-tra of phase-pure CuTCNQ in Fig. S1f (supporting informa-tion). Thus, the phase was assessed using FTIR spectroscopy

baFig. 1 SEM images of CuTCNQgrown on Cu foil in MeCN for24 h at (a) 25 °C or (b) 60 °C

Table 1 Properties of the polar solvents used in this study at 20 °C

Solvent Dipole moment(μ-Debye) [42]

Dielectric constant(F/m) [42]

Acetonitrile (aprotic) 3.92 36.64

Dimethyl sulfoxide (aprotic) 3.96 47.24

Dimethylformamide (aprotic) 3.82 36.7

Tetrahydrofuran (aprotic) 1.75 7.52

Ethanol (protic) 1.69 24.3

Methanol (protic) 1.7 33

38 emergent mater. (2019) 2:35–44

(Fig. 3b and c and Fig. S1g and h, supporting information). AllFTIR spectra including the phase-pure CuTCNQ (Fig. S1g)showed the δ(C–H) bending vibrations at ca. 823 cm−1, asexpected from TCNQ− in CuTCNQ, negating the possibilityof TCNQ0 or TCNQ2− (Fig. 3b) [22, 26–31]. However, theband at ca. 2200–2215 cm−1 can be used to identify the phase,such that phases I and II show bands at ca. 2200 and ca.2215 cm−1, respectively (Fig. S1h, supporting information)[39, 45]. An additional band at ca. 2170 cm−1 is noted in allthe samples; however, it is more prominent in phase II. FTIRspectral analysis reveals that irrespective of the solvent, the

CuTCNQ grown at 25 °C shows characteristics of a phase Imaterial (Fig. 3c). However, if grown at 60 °C, predominantlyphase II material with a minor residual phase I is produced(Fig. 3c), consistent with the morphological characteristics ofthese materials described earlier (Fig. 1). This implies thatduring the synthesis of CuTCNQ at 60 °C, CuTCNQ firstcrystalizes as phase I and converts to the thermodynamicallystable phase II.

The observation of the solvent-dependent, temperature-induced phase change mimics the behavior previously de-duced from cyclic voltammetry, where conversion of the

2150 2200 2250

stnuoC

.u.a/

Wavenumber /cm-1

(vii)

(vi)

(v)

(ii)

(i)

(iv)

(iii)

750 800 850

stnuoC

.u.a/

Wavenumber /cm-1

(vii)

(vi)

(v)

(ii)

(i)

(iv)

(iii)

1100 1700 2300

stnuoC

.u.a/

Wavenumber /cm-1

(vii)

(vi)

(v)

(ii)

(i)

(iv)

(iii)

a b cFig. 3 Characterization ofCuTCNQ by (a) Raman and (band c) FTIR spectroscopy; (i)shows the spectrum from pristineTCNQ0, while (ii–vii) correspondto CuTCNQ fabricated in (ii)MeCN: 25 °C; (iii) MeCN: 60 °C;(iv) DMSO: 25 °C; (v) DMSO:60 °C; (vi) DMF: 25 °C; and (vii)DMF: 60 °C

a b

c d

Fig. 2 SEM images of CuTCNQgrown on Cu foil in (a, b) DMSOor (c, d) DMF at (a, c) 25 °C or (b,d) 60 °C for 24 h

emergent mater. (2019) 2:35–44 39

kinetically-favored phase I to the thermodynamically-favoredphase II occurs even under ambient conditions [10].Electrochemistry may therefore offer a superior approach todistinguish the two phases of CuTCNQ. On this basis, wehave developed a method using electrochemical solid-solidstate interconversion for precise identification of the phasesof CuTCNQ. The underlying concept exploits the fact thatmicroparticles of TCNQ0 adhered to a glassy carbon (GC)electrode in contact with aqueous 0.1 M CuSO4 can be re-duced to form CuTCNQ via reaction (1) [10].

Cu2þ aqð Þ þ TCNQ0sð Þ þ 2e−⇌CuTCNQ sð Þ ð1Þ

Under voltammetric conditions, the kinetically-favoredphase I CuTCNQ is first formed electrochemically in the ini-tial cycles of potential but is then transformed to thethermodynamically-stable phase II on extensive, repetitive cy-cling of the potential. Therefore, in this voltammetric methodof phase identification, if the material under investigation isphase I CuTCNQ, making the initial scan direction positive,starting from 0.400 V vs. Ag/AgCl, followed by repetitivecycling of the potential, over the range of 0.600 to 0.250 V,leads to conversion to phase II, in aqueous solution containingCu2+(aq). Alternatively, if a phase II material is adhered to aGC electrode, insignificant changes in its voltammetry on re-petitive cycling is expected. If both phases co-exist, then thevoltammetric response is expected to contain features of bothphases.

Experimentally, this concept is validated in Fig. 4, usingreference standards pristine phase I (Fig. 4a), phase II (Fig.4b), and a mixture of these phases (Fig. 4c). The synthesis ofthe pure materials is elaborated in the experimental section[10, 45]. For pristine phase I CuTCNQ, an oxidation peak at479 mV, the reverse of Eq. (1) and a companion reductionpeak at 285 mV (Eq. 1) are observed (Fig. 4a). The largeseparation of ~ 194 mV in the peak potentials, the shape andother characteristics are consistent with an electrochemicallyirreversible solid-solid interconversion that occurs by a nucle-ation growth process, as described in detail in the literature[10]. Thus, phase I is easier to oxidize than phase II CuTCNQwhere both these processes occur via an overall two-electronoxidation process giving Cu2+(aqueous) and TCNQ0

(solid) [10].Multi-cycling of the potential of this phase I material progres-sively causes solid-solid conversion of phase I to phase IICuTCNQ. This conversion is clearly evident after the 10thand subsequent cycles, when a distinct, alternative process,associated with the oxidation of phase II emerges at morepositive potential, at the expense of phase I. In contrast, thecyclic voltammetry for pristine phase II CuTCNQ gives aninitial oxidation peak directly at a more positive potential of493 mV (Fig. 4b) than that observed for phase I. For thisphase, multi-cycling of the potential does not lead to a signif-icant change in the peak potential. Next, if a mixture

containing pristine phase I and II is analyzed by cyclic volt-ammetry (Fig. 4c), two distinct oxidation peak potentials areobserved, and multi-cycling of potential resulted in an in-crease in the current magnitude of the more positive potentialpeak corresponding to phase II at the expense of the phase Ipeak. This pattern of behavior is as expected for a sample thatcontained a mixture of both phases. It should be noted that forphase analysis of CuTCNQ, the first cycle is not used due toits unique properties and complexities in solid-solid state elec-trochemical interconversion, restricted to the first cycle [9,10].

a

b

c

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-15

-10

-5

0

5

10

15

Phase II

E / V vs. Ag/AgCl

I /

A

Phase I

0.3 0.5

479 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-2

-1

0

1

2

E /V vs. Ag/AgCl

I /A

Phase II

0.3 0.5

493 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-4

-2

0

2

4

Phase II

E / V vs. Ag/AgCl

I /

APhase I

0.3 0.5

492 mV462 mV

Fig. 4 Cyclic voltammetry of the reference standards of CuTCNQ withpre-determined phases, adhered to a GC electrode and placed in contactwith 0.1 M aqueous CuSO4: (a) pristine phase I; (b) pristine phase II and(c) a mixture of phase I and phase II. The insets show the 2nd and 3rdcycles of potential, whereas the main figure shows multiple cycles ofpotential. The initial potential is 0.400 V vs. Ag/AgCl. The initial scandirection is positive. The initial switching potential is 0.600 V vs. Ag/AgCl. The second switching potential is 0.250 V vs. Ag/AgCl. The scanrate is 20 mV·s−1

40 emergent mater. (2019) 2:35–44

The samples synthesized in different solvents were ana-lyzed by the electrochemical method validated above(Fig. 5). For CuTCNQ synthesized at 25 °C in MeCN (Fig.5a), DMSO (Fig. 5b), and DMF (Fig. 5c), multi-cycling clear-ly shows the conversion of phase I to phase II via progressivephase transition. This unambiguously confirms that theCuTCNQ synthesized at 25 °C only contained phase I. Forthe CuTCNQ synthesized at 60 °C, voltammetry is cycle-independent in case of MeCN (Fig. 5d) and DMSO (Fig.5e), confirming the formation of phase II, predominately inthese solvents. The synthesis in DMF, however, reveals amixture of the two phases (Fig. 5f), which we previously

assigned as phase II based on FTIR spectroscopy.Comparison of our electrochemical characterization tool forphase identification with conventional FTIR spectroscopy(Fig. 3c) reveals that where FTIR is not be sufficiently dis-criminative to resolve admixtures of different phases, electro-chemistry offers enhanced capabilities.

To obtain further mechanistic insights into the influence ofsolvent on CuTCNQ crystallization, we performed UV–visabsorbance studies and probed the consumption of TCNQ0

during CuTCNQ synthesis (Fig. S12, supportinginformation). The absorption characteristics of a chemical spe-cies is strongly influenced by the surrounding medium. Thus,

0.3 0.5

485 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Phase II

E / V vs. Ag/ AgCl

I /

A

Phase I

0.3 0.5

483 mV

0.3 0.5

488 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Phase II

E / V vs. Ag/AgCl

I /

A

Phase I

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2

Phase II

E / V vs. Ag/AgCl

I /

A

Phase I

a

c

b

0.3 0.5

515 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

E / V vs. Ag/ AgCl

I /

A

Phase II

0.3 0.5

490 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Phase II

E / V vs. Ag/AgCl

I /

A

0.3 0.5

516 mV487 mV

0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

-6

-4

-2

0

2

4

Phase IIPhase I

E / V vs. Ag/AgCl

I /

A

d

f

e

Fig. 5 Cyclic voltammetry of CuTCNQ solids fabricated under differentconditions. (a) MeCN: 25 °C; (b) DMSO: 25 °C; (c) DMF: 25 °C; (d)MeCN: 60 °C; (e) DMSO: 60 °C; and (f) DMF: 60 °C. The insets showthe 2nd and 3rd cycles of potential, whereas the main figure shows

multiple cycles of potential. The initial potential is 0.400 V vs. Ag/AgCl. The initial scan direction is positive. The initial switchingpotential is 0.600 V vs. Ag/AgCl. The second switching potential is0.250 V vs. Ag/AgCl. The scan rate is 20 mV·s−1

emergent mater. (2019) 2:35–44 41

to compare the solvent-dependent CuTCNQ crystallization,we calculated the extinction coefficients (ε) of TCNQ0 in eachsolvent (Fig. 6a). It is important to note that the λmax value forTCNQ0 is slightly solvent-dependent (396, 398, and 404 nminMeCN, DMSO and DMF, respectively). Using the ε valuesat λmax, the amount of TCNQ0 consumed during each reactioncould be calculated. This provides us with the reaction kineticsof CuTCNQ crystallization relevant to each solvent (Fig. 6b).The analysis of the UV-vis spectra reveals that CuTCNQ crys-tallization is promoted at higher temperatures, and the solventplays a large influence on the degree of crystallization. It isalso noted that at 25 °C, the degree of CuTCNQ crystallizationis lowest in DMSO.

To assess the catalytic ability of CuTCNQ materials pre-pared as described above, we employed a model pseudo-firstorder redox reaction involving the reduction of ferricyanide toferrocyanide in the presence of excess thiosulfate ions asshown in Eq. (2) [25, 26, 30, 31].

2 Fe CNð Þ6� �3−

aqð Þ þ 2 S4O3½ �2− aqð Þ⇌2 Fe CNð Þ6� �4−

aqð Þ

þ S4O6½ �2− aqð Þ ð2Þ

This process is readily monitored by absorbance spectros-copy by measuring the change in λmax at 420 nm (ferricya-nide). Although the amount of CuTCNQ catalyst was notconstant, the physical size of the Cu foil on which the activecatalyst was grown was kept consistent (1.0 cm × 1.0 cm) forall catalytic studies. Figure 7 and Fig. S13 (supporting infor-mation) reveal that all samples exhibit an ability to reduceferricyanide, albeit with different efficiencies. Firstly, the cat-alysts prepared in the conventional solvent MeCN show sig-nificantly less activity than those prepared in DMSO andDMF. Secondly, in all cases, the catalytic performance ofCuTCNQ grown at 60 °C is lower than those grown at25 °C. For example, the time taken for the reaction to go tocompletion using DMSO and DMF catalysts prepared at60 °C (predominantly phase II) took 18 and 31 min, respec-tively. In contrast, the catalysts grown at 25 °C (predominantlyphase I) in the same solvents showed superior performance,taking 10 and 21 min, respectively. This corresponds to animprovement of 44% and 32% in catalysts prepared underambient conditions over those grown at 60 °C in DMSO andDMF, respectively. Overall, this suggests that phase ICuTCNQ displays better catalytic performance than phaseII. This is not surprising as the conductivity of phase I isconsiderably superior to that of phase II [9, 25, 39], which isexpected to result in better electron transfer ability in the caseof the former. While it is difficult to calculate the number ofactive sites in case of each catalyst, based on the amount ofCuTCNQ crystallized on the surface of the Cu foil, theseresults suggest that CuTCNQ grown in DMSO (25 °C) pro-vides the best catalytic performance.

The plots of the ln of the ratios of the reactant at time t (At)and that at initial time point (A0) vs. reaction time allowed usto calculate the apparent reaction rate, kapp from the slope ofthe linear fits (Table 2) [9, 25, 39]. Interestingly, CuTCNQcatalysts show an induction time (or lag time) before the re-action starts to display pseudo-first order kinetics. It is widely

a b

MeCN

MeCN

DMSO

DMSO

DMF

DMF

0 25 50 75 100

TCNQ0 consumption /%

tnevlossisehtnyS

25 °C

25 °C

25 °C

60 °C

60 °C

60 °C

300 400 500

0

1x104

2x104

3x104

4x104

5x104

6x104

7x104

8x104

9x104

1x105

lo

m.L/

-1.c

m-1

Wavelength /nm

MeCN

DMSO

DMF

Fig. 6 (a) The absorption spectraof TCNQ0 in designated solvents,along with (b) the calculated % ofTCNQ0 consumed duringCuTCNQ crystallization underdesignated conditions. 100%corresponds to 5 mM TCNQ0

0 10 20 30 40 50 60

-4

-3

-2

-1

0

ln(A

t/A0)

Time /min

MeCN 25 °C

MeCN 60 °C

DMF 25 °C

DMF 60 °C

DMSO 25 °C

DMSO 60 °C

Fig. 7 Time–dependent catalytic activity of CuTCNQ catalystsfabricated under designated conditions, plotted as ln(At/A0) vs. time,where At and A0 correspond to the absorbance of ferricyanide at time tand at t = 0, respectively

42 emergent mater. (2019) 2:35–44

believed that the occurrence of an induction time is due to thefaster reactivity of the electron donor with the dissolved/adsorbed oxygen than with the key reactant [31]. We havepreviously established that in the case of CuTCNQ preparedinMeCN, the dissolved oxygen adsorbed on the surface of theCuTCNQ led to an induction time during catalysis [31].Interestingly, this induction time can be tuned through synthe-sizing CuTCNQ in different solvents. The induction time fol-lows the solvent order MeCN > DMF > DMSO. In all sol-vents, the induction time for phase II CuTCNQ is significantlylarger than that for phase I, revealing that kinetically-favoredphase I CuTCNQ is a better redox catalyst.

4 Conclusions

In summary, this work provides new insights into the impor-tant role of the reaction medium in the spontaneous crystalli-zation of CuTCNQ on the surface of Cu foil. Changing thesolvent can have important implications on the morphologicalcharacteristics of CuTCNQ. While polar aprotic solvents typ-ically have the ability to crystalize CuTCNQ, we note thatsolvents with higher dielectric constant can favorably driveCuTCNQ crystallization. Further, irrespective of the solvent,the elevated temperatures not only promote CuTCNQ crystal-lization, the conversion of phase I to phase II is simultaneous-ly favored. To extend the phase identification resolution capa-bilities available with FTIR, we have also introduced a robustand more sensitive technique based on electrochemical solid-solid state interconversion. While the amount of catalyst usedin a reaction is typically considered important for driving areaction, we observe that the catalyst with the lowest amountof CuTCNQ (DMSO/25 °C) outperformed all other materialswith least induction time. These results outline the importanceof understanding the influence of simple synthesis parameterssuch as solvent and temperature that can induce morphologi-cal and phase changes during CuTCNQ crystallization.Combining the outstanding redox capabilities of the newCuTCNQ materials reported here with optically-tunable two-

dimensional (2D) materials [53–61] will offer opportunitiesfor developing new photo-redox catalysts.

Acknowledgments V.B. thanks the Australian Research Council (ARC)for a Future Fellowship (FT140101285). V.B., R.R., L.L.M., and A.M.B.acknowledge the ARC for funding this project though an ARCDiscovery(DP170103477) Grant. The authors are appreciative of the generous sup-port of the Ian Potter Foundation towards establishing the Sir Ian PotterNanoBioSensing Facility at RMIT University. R.R. acknowledges RMITUniversity for a Vice Chancellor Fellowship. Authors acknowledge thesupport from the RMITMicroscopy andMicroanalysis Facility (RMMF)for technical assistance and providing access to characterization facilities.

Funding sources V.B.: Australian Research Council (ARC), FutureFellowship (FT140101285).

V.B., R.R., L.L.M., A.M.B.: ARC, Discovery Grant (DP170103477).

Author contributions The manuscript was written through contributionsof all authors. All authors have given approval to the final version of themanuscript.

Compliance with ethical standards

Conflict of interest statement On behalf of all authors, the correspond-ing author states that there is no conflict of interest.

References

1. L.B. Coleman, M.J. Cohen, D.J. Sandman, F.G. Yamagishi, A.F.Garito, A.J. Heeger, Solid State Commun. 12, 1125–1132 (1973)

2. J. Ferraris, D.O. Cowan, V. Walatka, J.H. Perlstein, J. Am. Chem.Soc. 95, 948–949 (1973)

3. K.P. Goetz, D. Vermeulen,M.E. Payne, C. Kloc, L.E. McNeil, O.D.Jurchescu, J. Mater. Chem. C. 2, 3065–3076 (2014)

4. T.H. Le, A. Nafady, N.T. Vo, R.W. Elliott, T.A. Hudson, R. Robson,et al., Inorg. Chem. 53, 3230–3242 (2014)

5. H. Liu, Q. Zhao, Y. Li, Y. Liu, F. Lu, J. Zhuang, et al., J. Am. Chem.Soc. 127, 1120–1121 (2005)

6. S.-G. Liu, Y.-Q. Liu, P.-J. Wu, D.-B. Zhu, Chem. Mater. 8, 2779–2787 (1996)

7. M. Mahajan, S.K. Bhargava, A.P. O’Mullane, Electrochim. Acta101, 186–195 (2013)

8. L.R. Melby, R.J. Harder, W.R. Hertler, W. Mahler, R.E. Benson,W.E. Mochel, J. Am. Chem. Soc. 84, 3374–3387 (1962)

9. A. Nafady, A.P. O’Mullane, A.M. Bond, Coordin. Chem. Rev. 268,101–142 (2014)

10. A.K. Neufeld, I. Madsen, A.M. Bond, C.F. Hogan, Chem. Mater.15, 3573–3585 (2003)

11. E. Nossol, A.B.S. Nossol, S.-X. Guo, J. Zhang, X.-Y. Fang, A.J.G.Zarbin, et al., J. Mater. Chem. C 2, 870–878 (2014)

12. A.P. O'Mullane, A.K. Neufeld, A.M. Bond, Anal. Chem. 77, 5447–5452 (2005)

13. L. Shields, J. Chem. Soc. Faraday Trans. 81, 1–9 (1985)14. E.B. Vickers, I.D. Giles, J.S. Miller, Chem. Mater. 17, 1667–1672

(2005)15. M.J. Capitán, J. Álvarez, C. Navío, R.Miranda, J. Phys. 28, 185002

(2016)16. R. Ramanathan, A.E. Kandjani, S. Walia, S. Balendhran, S.K.

Bhargava, K. Kalantar-zadeh, et al., RSC Adv. 3, 17654–17658(2013)

Table 2 Catalytic performance of CuTCNQ catalysts under designatedconditions

Sample kapp [min−1] Induction

time [min]Time taken for 95%reaction completion[min]

MeCN 25 0.06 16 41

MeCN 60 0.03 25 56

DMF 25 0.15 17 21

DMF 60 0.09 25 31

DMSO 25 0.30 3 10

DMSO 60 0.16 9 18

emergent mater. (2019) 2:35–44 43

17. R. Ramanathan, S. Walia, A.E. Kandjani, S. Balendran, M.Mohammadtaheri, S.K. Bhargava, et al., Langmuir 31, 1581–1587 (2014)

18. K. Xiao, I.N. Ivanov, A.A. Puretzky, Z. Liu, D.B. Geohegan, Adv.Mater. 18, 2184–2188 (2006)

19. K. Xiao, J. Tao, Z. Pan, A.A. Puretzky, I.N. Ivanov, S.J. Pennycook,et al., Angew. Chem. Int. Ed. 46, 2650–2654 (2007)

20. K. Xiao, J. Tao, A.A. Puretzky, I.N. Ivanov, S.T. Retterer, S.J.Pennycook, et al., Adv. Funct. Mater. 18, 3043–3048 (2008)

21. A. Pearson, A.P. O’Mullane, ChemPlusChem 78, 1343–1348(2013)

22. Z.M. Davoudi, A.E. Kandjani, A.I. Bhatt, I.L. Kyratzis, A.P.O'Mullane, V. Bansal, Adv. Funct. Mater. 24, 1047–1053 (2014)

23. D.D. La, R. Ramanathan, D. Kumar, T. Ahmed, S. Walia, K.J.Berean, et al., ChemistrySelect 2, 9962–9969 (2017)

24. D.D. La, R. Ramanathan, A. Rananaware, V. Bansal, S.V. Bhosale,RSC Adv. 6, 33931–33936 (2016)

25. M.Mohammadtaheri, R. Ramanathan, V. Bansal, Catal. Today 278,319–329 (2016)

26. M. Mohammadtaheri, R. Ramanathan, S. Walia, T. Ahmed, P.Weerathunge, S.R. Anderson, et al., Appl. Mater. Today 13, 107–115 (2018)

27. A. Pearson, A.P. O’Mullane, V. Bansal, S.K. Bhargava, Inorg.Chem. 50, 1705–1712 (2011)

28. A. Pearson, A.P. O’Mullane, S.K. Bhargava, V. Bansal, Inorg.Chem. 51, 8791–8801 (2012)

29. A. Pearson, R. Ramanathan, A.P. O'Mullane, V. Bansal, Adv.Funct. Mater. 24, 7570–7579 (2014)

30. R. Ramanathan, A. Pearson, S. Walia, A.E. Kandjani, M.Mohammadtaheri, M. Bhaskaran, et al., Appl. Mater. Today 10,12–17 (2018)

31. M.C. Siu, S.R. Anderson, M. Mohammadtaheri, T. Ahmed, S.Walia, R. Ramanathan, et al., Adv. Mater. Interfaces 1700097, 4(2017)

32. W. Kaim, M. Moscherosch, Coordin. Chem. Rev. 129, 157–193(1994)

33. X. Guo, S. Zhu, R.M. Kong, X. Zhang, F. Qu, ACS Sustain. Chem.Eng. 6, 1545–1549 (2018)

34. L. Liu, Q. Wu, H. Guo, L. Li, M. Liu, D. Li, et al., Mater. Lett. 230,53–56 (2018)

35. K. Miyao, A. Funabiki, K. Takahashi, T. Mochida, M. Uruichi,New J. Chem. 38, 739–743 (2014)

36. A. Pearson, V. Bansal, A.P. O’Mullane, Electrochim. Acta 114,189–197 (2013)

37. M. Shafiei, F. Hoshyargar, J. Lipton-Duffin, C. Piloto, N. Motta,A.P. O’Mullane, J. Phys, Chem. C. 119, 22208–22216 (2015)

38. M. Xie, X. Xiong, L. Yang, X. Shi, A.M. Asiri, X. Sun, Chem.Commun. 54, 2300–2303 (2018)

39. M. Mahajan, S.K. Bhargava, A.P. O’Mullane, RSC Adv. 3, 4440–4446 (2013)

40. K.Wang, X. Qian, L. Zhang, Y. Li, H. Liu, ACS Appl. Mater. Inter.5, 5825–5831 (2013)

41. Y. Wei, X. Ren, H. Ma, X. Sun, Y. Zhang, X. Kuang, et al., Chem.Euro. J. 24, 2075–2079 (2018)

42. X. Ren, X. Ji, Y. Wei, D. Wu, Y. Zhang, M. Ma, et al., Chem.Commun. 54, 1425–1428 (2018)

43. D.Wu, Y. Wei, X. Ren, X. Ji, Y. Liu, X. Guo, et al., Adv. Mater. 30,1705366 (2018)

44. X. Zhu, X. Shi, A.M. Asiri, Y. Luo, X. Sun, Inorg. Chem. Front. 5,1188–1192 (2018)

45. R.A. Heintz, H. Zhao, X. Ouyang, G. Grandinetti, J. Cowen, K.R.Dunbar, Inorg. Chem. 38, 144–156 (1998)

46. Y. Liu, Z. Ji, H. Li, W. Hu, Y. Liu, D. Zhu, Nano Res. 2, 630–637(2009)

47. R. Ramanathan, A.P. O’Mullane, R.Y. Parikh, P.M. Smooker, S.K.Bhargava, V. Bansal, Langmuir 27, 714–719 (2010)

48. V. Bansal, A. Syed, S.K. Bhargava, A. Ahmad, M. Sastry,Langmuir 23, 4993–4998 (2007)

49. A. Nafady, A.M. Bond, V. Qu, L.L. Martin, J. Solid StateElectrochem. 17, 1609–1620 (2013)

50. R.J. Phaneuf, H.-C. Kan, E.D. Williams, in Low dimensional struc-tures prepared by epitaxial growth or regrowth on patternedsubstrates, ed. by K. Eberl, P. M. Petroff, P. Demeester. Evolutionof Si surface nanostructure under growth conditions (SpringerNetherlands, Dordrecht, 1995), pp. 185–195

51. Y. Yamashita, T. Suzuki, T.Mukai, J. Chem. Soc. Chem. Commun.,1184–1185 (1987)

52. A.E. Kandjani, M.J. Griffin, R. Ramanathan, S.J. Ippolito, S.K.Bhargava, V. Bansal, J. Raman Spectrosc. 44, 608–621 (2013)

53. S. Walia, Y. Sabri, A. Taimur, M.R. Field, R. Ramanathan, A.Arash, et al., 2D Mater. 015025, 4 (2017)

54. M.N. Karim, M. Singh, P. Weerathunge, P. Bian, R. Zheng, C.Dekiwadia, et al., ACS Appl. Nano Mater. 1, 1694–1704 (2018)

55. M. Singh, P. Weerathunge, P.D. Liyanage, E. Mayes, R.Ramanathan, V. Bansal, Langmuir 33, 10006–10015 (2017)

56. A.R. Rezk, S. Walia, R. Ramanathan, H. Nili, J.Z. Ou, V. Bansal,et al., Adv. Opt. Mater. 3, 888–894 (2015)

57. M. Singh, E. Della Gaspera, T. Ahmed, S.Walia, R. Ramanathan, J.van Embden, et al., 2D Mater. 4, 025110 (2017)

58. S. Walia, S. Balendhran, T. Ahmed, M. Singh, C. El-Badawi, M.D.Brennan, et al., Adv. Mater. 1700152, 29 (2017)

59. M. Singh, D. Jampaiah, A.E. Kandjani, Y.M. Sabri, E. DellaGaspera, P. Reineck, et al., Nanoscale 10, 6039–6050 (2018)

60. T. Ahmed, S. Balendhran, M.N. Karim, E.L. Mayes, M.R. Field, R.Ramanathan, et al., npj 2D Mater. Appl. 1, 18 (2017)

61. S. Kuriakose, T. Ahmed, P. Taylor, Y. Zhu, M.J.S. Spencer, S.Balendhran, et al., 2D Mater. 015009, 6 (2018)

44 emergent mater. (2019) 2:35–44