Green Chemistry

PAPER

Cite this: Green Chem., 2017, 19,1327

Received 9th January 2017,Accepted 16th January 2017

DOI: 10.1039/c7gc00084g

rsc.li/greenchem

Phosphine-free avenue to Co2P nanoparticleencapsulated N,P co-doped CNTs: a novelnon-enzymatic glucose sensor and an efficientelectrocatalyst for oxygen evolution reaction†‡

Debanjan Das,* Abhinaba Das, Meera Reghunath and Karuna Kar Nanda*

A novel one-step, one-pot strategy to synthesize Co2P encapsulated N,P dual doped carbon nanotubes

(Co2P/NPCNTs) is developed via a g-C3N4 intermediated approach. The method uses readily available, in-

expensive precursors without involving toxic phosphine gas (PH3) during the phosphidation process.

Moreover, the CNTs are synthesized and doped in situ without the aid of any external carbon additives.

The morphology and structure were characterized by field-emission scanning electron microscopy

(FESEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), Raman spectroscopy and X-ray

photoelectron spectroscopy (XPS). The as-prepared Co2P/NPCNTs were then for the first time demon-

strated to be an electrochemical sensor for the enzymeless detection of glucose. The Co2P/NPCNT-

modified electrode exhibits a quick response and good linearity (R2 = 0.997) up to 7 mM with a high sen-

sitivity and a low detection limit (LOD) of 338.8 μA mM−1 cm−2 and 0.88 μM, respectively, towards

glucose, selectively in the presence of various endogenous interfering agents such as D-sucrose,

D-mannose, L-cysteine, urea etc. with a good long-term stability. When applied as an electrocatalyst for

the oxygen evolution reaction (OER), Co2P/NPCNT delivers a current density of 10 mA cm−2 at an over-

potential of 370 mV with a Tafel slope of only 53 mV dec−1, outperforming even the state-of-the-art

RuO2 while maintaining excellent long-term durability.

Introduction

Diabetes mellitus is rapidly growing to be a global epidemicwith over 62 million affected worldwide. The situation is par-ticularly grave in the context of India, which leads the worldwith over 31.7 million people suffering from diabetes, result-ing in around one million related deaths annually.1,2 Thediagnosis and management of the disease requires analysis ofblood sugar levels ranging from one to several times a day,which calls for the development of highly sensitive andefficient glucose sensors. Presently, the global market is domi-nated by sensors based on glucose oxidase (GOD)-assisted(enzymatic) electrochemical oxidation. Though enzyme basedglucose sensors provide the advantage of high sensitivity andselectivity, they suffer from various drawbacks such as (i) atedious and complex process of enzyme immobilization,

(ii) high cost of the enzymes used, (iii) instability due to thedenaturation of the enzymes and (iv) low conductivity due toindirect electron transfer.3,4 Thus the development of nonenzy-matic glucose sensors based on direct electrocatalytic oxi-dation has received increasing attention due to their direct-electron-transfer-shuttle-free detection.5,6

Conventional non-enzymatic glucose sensors have mainlyconcentrated on the use of precious metals (Pt, Au, Pd)7–9 ortransition metals (Cu, Co, Ni) and their oxides.5,6,10 However,owing to their prohibitive cost, noble metals and their alloysare not suited for large scale applications. Despite beingcheaper, the widespread use of metal oxide based sensors isimpeded mainly because of the poor electronic conductivity.To alleviate this problem, one strategy is to load the metal/metal-oxides onto conducting substrates such as carbon nano-tubes, graphene, porous carbon etc. to result in a largerexposed surface area and faster electron transfer. Another strat-egy may be the development of materials such as metal chalco-genides, phosphides etc. which possess intrinsically higherelectrical conductivity than metal oxides.

On the other hand, the oxygen evolution reaction (OER) liesat the heart of various clean energy technologies such as fuel-cells, metal–air batteries and alkaline water electrolyzers.11–16

†M. R. was an Indian Academy of Sciences summer fellow at MRC, IISc.‡Electronic supplementary information (ESI) available. See DOI: 10.1039/c7gc00084g

Materials Research Centre, Indian Institute of Science, Bangalore-560012, India.

E-mail: das@mrc.iisc.ernet.in, nanda@mrc.iisc.ernet.in

This journal is © The Royal Society of Chemistry 2017 Green Chem., 2017, 19, 1327–1335 | 1327

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

.

View Article OnlineView Journal | View Issue

However, the high energy barrier of O–H bond breakingfollowed by O–O bond formation results in sluggish reactionkinetics, which is a major bottleneck in the practical appli-cation of the OER-centric sustainable technologies on a largescale.17 An efficient electrocatalyst is thus required to effec-tively couple the multiple electron and proton transfer pro-cesses for O2 evolution at low overpotentials. Though ruthe-nium (Ru) and iridium (Ir) based materials form the state-of-the-art OER electrocatalysts,18 their high cost and scarcityseverely limit their practical applications. Thus a tremendouseffort is now being devoted to the development of highly activeand stable low-cost OER electrocatalysts wherein Co-based cat-alysts are found to perform extremely well.19–22

Recently, dicobalt phosphide (Co2P) has received increasedattention due to their catalytic and magnetic properties as wellas being a promising electrode material.23–25 However, most ofthe existing protocols to synthesize Co2P nanoparticles involvethe use of toxic alkylphosphine, various surfactants andrigorous air-free conditions, which limit their large scaleapplication.24–26 A one-step process to synthesize Co2P usinginexpensive and non-toxic precursors is thus highly desirable.Furthermore, packing Co2P nanoparticles within carbon layerswould provide a dual advantage of a conducting support andprotection against harsh environments: this strategy was mani-fested in many forms and was found to be highly effective fora variety of electrocatalysts.27–30 However, very limited successhas been achieved in this direction, which included multiplesynthesis steps involving an externally incorporated carbonsupport.31 Unfortunately, the synthesis of carbon catalyst sup-ports such as fullerenes, carbon nanotubes, carbon nanofibresetc. is heavily reliant on fossil fuels (e.g. phenol, pitch,methane etc.) using various energy intensive techniques suchas electric-arc discharge, chemical vapour deposition and laserablation, which are environmentally harmful and also raisethe production cost of these materials. Even worse are somesynthesis procedures which use toxic reagents which lead toenvironmental pollution.32

A classic example is Hummer’s method for the synthesis ofgraphene oxide (GO), which requires a strong oxidant (KMnO4)and concentrated H2SO4, raising environmental concerns.Moreover, to the best of our knowledge, Co2P has never beenexplored towards electrochemical glucose detection. Besides,Co2P has been rarely studied as an electrocatalyst for OER.33,34

Mindful of these facts, herein we have developed a one-stepstrategy to synthesize phase-pure Co2P nanoparticle encapsu-lated N,P-dual-doped CNTs, which were formed and dopedin situ during the course of the reaction. This “two birds withone stone” approach integrates the high intrinsic conductivityof Co2P

35 with the conducting N,P-dual-doped CNT support.The as-prepared Co2P/NPCNTs were then for the first timedemonstrated to be an efficient non-enzymatic sensor for theelectrochemical detection of glucose. When applied as an OERelectrocatalyst, Co2P/NPCNTs outperforms even the state-of-the-art RuO2 by reaching a current density of 10 mA cm−2 at anoverpotential of 370 mV with a small Tafel slope of 53 mV dec−1

in 1 M KOH.

ExperimentalReagents

Cobalt nitrate hexahydrate, melamine, D-glucose, urea, NaCl,Na2SO4, sucrose and mannose were purchased from SD FineChemicals Ltd. Triphenylphosphine, 5 wt% Nafion solutionand L-cysteine were acquired from Sigma Aldrich. Milliporewater from a Merck Millipore system was used throughout theexperiment.

Synthesis of materials

In a typical procedure for Co2P/NPCNT synthesis (Scheme 1),250 mg Co(NO3)2·6H2O, 500 mg triphenylphosphine and 1.5 gmelamine were ground together using a mortar and pestle for10 min. The process was repeated twice to ensure that theprecursors were well-mixed together. The mixture was thenloaded in a fused silica crucible and placed at the centre of ahorizontal tube furnace and was heated up to 850 °C at aramping rate of 4.7 °C min−1 under 80 sccm N2 flow andmaintained for 1 h. After being cooled to room temperaturenaturally, the samples were washed thoroughly with waterseveral times over and dried at 80 °C under vacuum. The yieldof this synthesis procedure was 7.2% (162 mg in a typical syn-thesis). Co/NCNT was synthesized following the exact sameprocedure used for Co2P/NPCNT synthesis except for the useof triphenylphosphine.

Characterization

The diffraction data were collected at room temperature withthe 2θ scan range between 10 and 90° using an X-ray diffrac-tometer (XRD, PANalytical) equipped with Cu Kα radiation(1.54 Å). The scanning electron microscopy (SEM) images wereobtained using a FESEM FEI Inspect 50. TEM characterizationwas carried out using a JEOL JEM-2100F at an acceleratingvoltage of 80 kV. X-ray photoelectron spectroscopy (XPS) wasperformed for the elemental analysis on an ESCALAB 250(Thermo Electron) with a monochromatic Al Kα (1486.6 eV)source. The binding energy of all the elements was calibratedby placing the principal C 1s peak at 284.6 eV. Photoluminesence(PL) and Raman spectra of the samples were recorded on aWITec system with 355 nm and 532 nm excitation wavelengths

Scheme 1 Schematic illustration of the synthesis procedure of Co2P/NPCNTs.

Paper Green Chemistry

1328 | Green Chem., 2017, 19, 1327–1335 This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

respectively. Thermogravimetric analysis (TGA) of the sampleswas carried out using a PerkinElmer model TGA Q50 V20.13Build 39.

Electrochemical measurements

The electrochemical tests were carried out on a conventionalthree-electrode cell by using a CHI750E workstation. A com-mercial glassy carbon electrode (GCE, 3 mm diameter,0.07 cm2) served as the working electrode while an Ag/AgClelectrode and a Pt wire served as the reference and counterelectrodes respectively. The GCE was sequentially polishedwith 1.0, 0.5 and 0.05 μm alumina powder to obtain a mirror-like finish which was then cleaned with Millipore water andethanol under sonication. The electrode preparation wascarried out as follows: 2 mg of catalyst was added to 500 μLethanol followed by 20 μL of a 5 wt% Nafion© 117 solution.The as-prepared mixture was sonicated for 30 min. 15 μLof the catalyst ink so obtained was drop-cast onto theGCE and air-dried, which resulted in a catalyst loading of∼0.75 mg cm−2.

All the experiments related to glucose sensing were carriedout in a 0.1 M aqueous KOH solution. Cyclic voltammetry (CV)measurements were carried out within a potential window of0.0 to 0.6 V vs. Ag/AgCl. Chronoamperometry (CA) was carriedout at a constant applied potential of 0.55 V vs. Ag/AgCl withsuccessive addition of a 1 mM glucose solution. All the experi-ments pertaining to OER were carried out in a 1 M KOH solu-tion. All the experiments were carried out at room temperature(∼28 °C).

Saliva collection protocol

(a) Rinse the mouth thoroughly with water and wait for5 minutes.

(b) Swallowing was minimized to hold saliva in the mouth.(c) A sterile dental cotton sponge was placed in the mouth

and chewed on until it was soaked with saliva.(d) Directly deposit the wet sponge from the mouth into the

syringe to minimize the risk of contamination.(e) Insert the plunger to squeeze the saliva through a

0.2 μm PVDF membrane (Sigma Aldrich) into sterilized centri-fuge tubes.

(f ) The obtained samples were centrifuged at 3800 rpm for5 min. The supernatant was collected for further studies.

Results and discussion

A simple one-step, one-pot strategy (Scheme 1) was employedto synthesize Co2P capped N,P dual-doped CNTs (Co2P/CNT)from cobalt acetylacetonate, triphenylphosphine and mela-mine as cobalt, phosphorus, and nitrogen & carbon sources,respectively. Melamine acted as the source of carbon and nitro-gen dopants, as well as directing the growth of CNTs.Triphenylphosphine acts as a phosphorus source for not onlythe synthesis of phase pure Co2P but also for phosphorusdoping in the CNTs.

Fig. 1(a) shows the X-ray diffraction (XRD) pattern of Co2P/NPCNTs obtained at 850 °C along with the g-C3N4 intermedi-ate obtained at 550 °C. A sharp, prominent (002) peak corres-ponding to g-C3N4 can be seen at 27.5°. When the sample washeated to 850 °C, a broad peak centred at 26.1° correspondingto the graphitic carbon peak of (002) along with peaks at 40.6,43.2, 44.1, 48.2, 51.6 and 56.1° that can be assigned to the(121), (220), (211), (031), (002) and (320) planes of Co2P (JCPDSno. 32-0306), respectively, can be seen. This points to the trans-formation of the g-C3N4 intermediate into Co2P/NPCNTs.

Additionally, a blue luminescence is observed at 463.5 nm,which is characteristic of g-C3N4

36 but disappears when thesample is heated to 850 °C (Fig. 1b), further pointing out theconversion of Co3O4/g-C3N4 into Co2P/NPCNT.

Based on the above results, a plausible mechanism for thesynthesis of Co2P/NPCNTs is proposed. Co(NO3)2·6H2O isknown to dissociate above 200 °C to yield Co3O4 under a N2

atmosphere with the release of water vapor, HNO3, HNOx andother volatile species.37

Meanwhile, melamine begins to undergo thermal conden-sation beyond 300 °C to finally yield g-C3N4 in a temperaturerange of 520–600 °C.36 The SEM image shown in Fig. S1‡reveals a thick sheet like morphology of Co3O4 modifiedg-C3N4 obtained at 550 °C. The polymeric g-C3N4 then decom-poses at 700 °C into nitrogen and cyano fragments38 whichcould serve as the source of nitrogen and carbon towardsN-doped CNT formation. It is in the presence of this reducingatmosphere that the preformed Co3O4 may be reduced to met-allic Co, where in the presence PPh3 they underwent anUllmann-type reaction to yield dicobalt phosphide (4Co +2PPh3 = 2Co2P + 3Ph-Ph) nanoparticles.39 This process pro-vides a significant route to avoid the use of pyrophoric andexpensive alkylphosphine or NaH2PO2, the traditional phos-phorus source for the synthesis of metal phosphides, both ofwhich release toxic phosphine (PH3) at elevated temperatures.

To emphasize the importance of melamine and the role ofthe g-C3N4 intermediate, two additional experiments were per-formed, (i) D-glucose was used as a source of carbon instead ofmelamine while keeping the other parameters intact. SinceD-glucose cannot condense to produce g-C3N4 large, irregularbead-like carbon structures modified with mixed phase Co2P–

Fig. 1 (a) XRD pattern of Co2P/NPCNT and the g-C3N4 precursorobtained at 550 °C with their respective digital micrographs; (b) photo-luminescence (PL) spectra of Co2P/NPCNT in comparison with theg-C3N4 precursor (inset: the respective optical micrographs).

Green Chemistry Paper

This journal is © The Royal Society of Chemistry 2017 Green Chem., 2017, 19, 1327–1335 | 1329

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

CoP were obtained upon heating it to 850 °C (Fig. S2‡). (ii)Melamine and Co (NO3)2·6H2O were heated to 850 °C withouttriphenylphosphine, wherein Co nanoparticle encapsulatedCNTs were obtained (Fig. S3‡). From these control experi-ments, it can be concluded that (i) the formation of the g-C3N4

intermediate is fundamental to the formation of definite CNTswhile triphenylphosphine acts as the source of phosphorus forCo2P formation as well as P-doping in the CNT framework and(ii) Co3O4 obtained via decomposition of Co (NO3)2·6H2O atlower temperature (∼200 °C) indeed gets reduced to metallicCo at higher temperatures during the course of the reaction.

Raman spectroscopy is a powerful tool for the characteriz-ation of carbon based materials. The Raman spectrumrecorded for Co2P/NPCNTs shows two distinct peaks at 1356and 1587 cm−1 corresponding to the typical D and G bands,respectively (Fig. 2a). The D band originates from the hybri-dized vibrational modes associated with the CNT edges, whilethe G band is attributed to the vibrations and tangential oscil-lations of the sp2-hybridized carbon atoms in CNT.40 A broad2D band centered around 2706 cm−1, a characteristic of gra-phitic carbon materials, is observed, which is in accordancewith the XRD results. In addition to the ID/IG ratio of Co2P/NPCNTs being 0.89, a downshift of the G band compared topristine graphitic carbon provides evidence of significantstructural defects or microstructural rearrangement of atomsin the CNT framework due to heteroatom doping.Thermogravimetric analysis (TGA) (Fig. 2b) estimates the Co2Ploading in Co2P/NPCNT to be around 38.6% while the rest ismade of N,P dual-doped CNTs.

SEM and TEM were conducted to analyse the micro-morphology and the finer structural details of Co2P/NPCNTs.Orderly oriented, vertically-aligned Co2P/NPCNTs wereobserved over a large area (Fig. 3a). A Gaussian fitting of theparticle size distribution (averaged over 80 particles) as shownin the inset of Fig. 3a revealed the mean diameter of the Co2Ptips to be 55 nm. The diameter of the NPCNTs range from 80to 250 nm and are more than a micron long. The stark contrastbetween the Co2P nanoparticle and the NPCNT encapsulatingit points to the extremely thin nature of the CNT walls with ahighly wrinkled surface (Fig. 3b).

These attributes of a high aspect ratio along with a rough-ened surface provides a large surface area to Co2P/NPCNTswhich is greatly beneficial to catalysis, sensing and a host of

other applications. TEM analysis (Fig. 3c) reveals a typicalbamboo-like compartmentalization in the NPCNTs, a typicalmorphological feature associated with N-doping in the carbonframework.41 The spot-patterns making up the rings in theSAED pattern point to the polycrystalline nature of Co2P/NPCNTs (inset of Fig. 3c). The high-resolution TEM (HRTEM)image clearly shows the lattice spacings of the Co2P (121)plane (Fig. 3d).

XPS studies were carried out to study the surface chemistryof the as-prepared Co2P/NPCNT nanostructures. The XPSsurvey scan of Co2P/NPCNTs confirms the presence of Co, C, Nand P (Fig. S4‡). The high resolution C 1s spectra (Fig. 4a) canbe deconvoluted into four contributions centred at 284.7 eV

Fig. 2 (a) Raman spectrum of Co2P/NPCNT and (b) TGA curves ofCo2P/CNT carried out under N2 (black) and air (red).

Fig. 3 (a & b) Low- and high-magnification SEM images of Co2P/NPCNTs. (c) TEM image (inset: SAED pattern) and (d) HRTEM image ofCo2P/NPCNT.

Fig. 4 High resolution XPS of (a) carbon, (b) nitrogen, (c) phosphorusand (d) cobalt in Co2P/NPCNT.

Paper Green Chemistry

1330 | Green Chem., 2017, 19, 1327–1335 This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

(sp2-hybridized graphitic C), 285.7 eV (N-sp2 C), 287.3 eV(N-sp3 C) and 289 eV (π → π*).42,43 Deconvolution of the N 1sspectrum reveals the presence of pyridinic-N, pyrolic-N andgraphitic-N centered at 398.3 eV, 399.5 eV and 401.4 eV,respectively (Fig. 4b).15 These features point to the presence ofcarbons that contain N atoms in the carbonaceous framework,which play a significant role in the sensing and catalyticactivity of carbon based materials.41,44 With regard to the P 2pcore level spectrum, a sharp doublet is observed at 129.6 eVand 130.5 eV that can be attributed to the P 2p3/2 and P 2p1/2respectively, besides a broad peak that can be deconvolutedinto two subpeaks located at 133.7 eV and 134.5 eV that areascribed to P–C and P–O bonds respectively (Fig. 4c).45

The high resolution scan of Co 2p electrons yielded twomain peaks positioned at 781.2 eV (Co 2p3/2) and 796.4 eV(Co 2p1/2) along with their respective satellite peaks at 786.7 eVand 802.3 eV (Fig. 4d). The Co 2p3/2 peak at 781.2 eV and theP 2p3/2 peak at 129.6 eV can be ascribed to the typical bindingenergies for Co 2p and P 2p contributions in Co2P.

31 It isworth noting that the Co 2p binding energy of Co2P/NPCNT(781.2 eV) has positively shifted from that of metallic Co(778.1–778.2 eV) while that of P 2p shifts negatively to 129.6 eVin comparison with elemental P (130.2 eV), suggesting thetransfer of electron density from Co to P.46

The Co2P/NPCNTs were then investigated as a novel non-enzymatic glucose sensor and their electrocatalytic activitytowards glucose oxidation was monitored by cyclic voltamme-try (CV). Fig. 5a shows the CV curves of bare GCE in compari-son with Co/NCNT and Co2P/NPCNT-modified GCE in 0.1 M

KOH at a scan rate of 50 mV s−1. Bare GCE does not show anyredox feature in the potential range under consideration(0.0–0.6 V vs. Ag/AgCl), although both Co/NCNT and Co2P/NPCNT-modified GCE exhibit a well-defined oxidation andreduction peak in the aforesaid potential range, suggestingtheir electrocatalytic activity in the region.

However, Co/NCNT (as well as bare GCE) does not show anysignificant response to the addition of glucose (Fig. S5a &S5b‡), hence Co2P/NPCT was further explored towards glucosesensing. The CVs recorded for Co2P/NPCNT-modified GCE inN2 and O2 saturated 0.1 M KOH are very similar, suggestingthat Co2P/NPCNT is not much affected by oxygen (Fig. 5b).Fig. 5c shows the CV curves of Co2P/NPCNT with successiveaddition of glucose from 0 to 3 mM at a scan rate of 50 mV s−1

in 0.1 M KOH solution where the current increases in a clearconcentration-dependent manner.

Fig. 5d demonstrates the effect of scan rate on the oxidationof glucose by Co2P/NPCNT. The oxidation current increaseslinearly with scan rate in the sweep range of 2–100 mV s−1

(inset of Fig. 5d) that can be fitted to the equation I (μA) =152.02 + 4.79ν (ν is the scan rate) having a correlation coeffi-cient of 0.992, which illustrates that the electrochemical kine-tics of glucose oxidation by Co2P/NPCNT is controlled bysurface adsorption of the glucose molecule.47 The electro-chemical process occurring on the surface of Co2P in alkalinemedium is not clear; however a tentative mechanism forglucose oxidation by Co2P/NPCNT can be derived in the lineproposed for Co3O4 by Chen and co-workers.10 It is believedthat Coδ+ in Co2P is easily oxidized to Co(III) (Coδ+ + nOH + H2O= CoOOH + ne−) under a positive bias, a feature clearly noticedin Fig. 5a & b wherein no glucose was added. However, as soonas glucose is injected into the reaction system, electrooxidationof glucose takes place by the CoOOH/Co4+ couple in an alka-line medium (eqn (1) & (2)):48

CoOOHþ OH� ¼ Co4þ þH2Oþ e� ð1Þ

Glucoseþ 2OH� ¼ GluconolactoneþH2Oþ 2e�: ð2ÞThis subtle change can be observed in Fig. 5c, wherein

there is a distinct positive shift in the anodic peak positionwith the addition of glucose compared to the curve (black)without any glucose that points to the above discussedmechanism.

Fig. 6a shows the amperometric response curve of Co2P/NPCNT in 0.1 M KOH solution at a potential of 0.55 V withsuccessive addition of 1 mM glucose. To ensure a homo-geneous glucose concentration, the glucose solution wasadded to the electrolyte under continuous stirring. A distinctincrease in the current is observed upon addition of glucosewhich gradually plateaus off at higher concentrations ofglucose due to the accumulation of intermediate species uponthe surface of the electrode.47 Additionally, the inset showsthat the sensor takes no more than 6 s to achieve a steady-statecurrent on the addition of glucose, indicating a rapid responseof the Co2P/NPCNT sensor towards glucose. Fig. 6b shows thecorresponding calibration curve of a Co2P/NPCNT-based

Fig. 5 (a) CV curves of bare GCE, Co2P/CNT-modified GCE and Co/NCNT-modified GCE in 0.1 M KOH at a scan rate of 50 mV s−1; (b) CV ofbare GCE (black) modified with Co2P/NPCNT in O2 saturated (blue) andN2 saturated (red) 0.1 M KOH, at 50 mV s−1; (c) CV curves of Co2P/NPCNT-GCE at different glucose concentrations; (d) CV curves of Co2P/NPCNT-GCE at a scan rate of 2–100 mV s−1 in 0.1 M KOH with aglucose concentration of 0.5 mM (inset: a plot of oxidative peak currentvs. scan rate).

Green Chemistry Paper

This journal is © The Royal Society of Chemistry 2017 Green Chem., 2017, 19, 1327–1335 | 1331

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

sensor derived from the amperometric response. A well-defined linear region is observed in the concentration rangeup to 7 mM. Beyond this concentration, the response currentstarts to level off due to the accumulation of intermediatespecies on the electrode surface. Fortuitously, this linear range(up to 7 mM) is suitable for the detection of blood glucoselevels (4.4–6.6 mM).47 The fitting equation is I (mA) = 0.224 +0.02372 C (mM) with a correlation coefficient of 0.997. TheCo2P/NPCNT sensor shows a high sensitivity of 338.8 μA mM−1

cm−2 and a limit of detection (LOD) of 0.88 μM. The sensitivityis better than or comparable to the very recently developedCo-based electrochemical glucose sensors such as Zn-dopedCo3O4 film on FTO (192 μA mM−1 cm−2),49 nanoporousCo3O4 nanowires (300 μA mM−1 cm−2),50 3D N-Co-CNT@NG(9.05 μA μM−1 cm−2),51 Co3O4 nanoflowers/3D grapheneoxide hydrogel (492.8 μA mM−1 cm−2)52 and CoP nanorods(116.8 μA mM−1 cm−2).48

A major challenge for any glucose sensor is the selectivedetection of glucose while avoiding various endogenous inter-fering species. To analyse the anti-interference activity of Co2P/NPCNT, its selectivity towards glucose was tested in the pres-ence of various potential interfering reagents via chrono-amperometry. As shown in Fig. 6c, addition of 1 mM glucoseresults in a quick and significant increase in current, whereasthe addition of even 2 mM of other saccharides, namely,sucrose and mannose, fails to produce any appreciableincrease in the current. Urea is commonly present in bloodserum as well as urine whose exclusion in the presence ofglucose is desirable for practical applications.

Even adding twice the glucose concentration (2 mM), ureafailed to generate any increase in the current. Interference

arising due to amino acids and inorganic salts was analysedusing L-cysteine and Na2SO4 respectively as representatives;while L-cysteine showed a negligible increase in current, nosuch change was observed after the addition of Na2SO4. Inaddition, chloride poisoning of the sensor was monitored byintroducing NaCl which showed no significant response evenat a concentration of 2 mM, underlining the excellent resist-ance to poisoning of the Co2P/NPCNT electrode. Finally, whenglucose was successively added again after carrying out theanti-interference tests, the increase in current was quick andpronounced, suggesting that the sensor remains active towardsglucose even after coming into contact with a variety of poten-tially interfering species.

The stability of the Co2P/NPCNT sensor towards continuousglucose monitoring was analysed by chronoamperometry at+0.55 V in the presence of 1 mM glucose. As shown in Fig. 6dthere is only a slight attenuation of 6.7% in the originalcurrent after over 90 minutes of operation, suggesting con-siderable stability of Co2P/NPCNT towards glucose. We alsotested the long term stability of the sensor in view of practicalapplications. The Co2P/NPCNT electrode was stored in 0.1 MKOH solution at 4 °C when not in use. The response of thatsensor towards 1 mM glucose in 0.1 M KOH solution was ana-lysed daily for a week. A slight attenuation, of 4.4%, of theinitial response current was observed at the end of the week(Fig. S6‡). The SEM and XRD analyses of the sample after thestability test (Fig. S7a & b‡) reveal that the Co2P capped CNTmorphology was preserved along with structural integrity,which is a testimony of the sensor’s robustness. Additionally,concerns may arise regarding the use of cobalt for the purposeof glucose sensing, which is considered to be highly toxic. Toevaluate this issue, we carried out chronoamperometry onboth of our samples, i.e., Co/NCNT and Co2P/NPCNTs, for12 hours at +0.55 V (vs. Ag/AgCl) in 0.1 M KOH to evaluatewhether any cobalt has leached out into the solution duringthat period of time. Subsequently, ICP-MS studies carried outin the test solution revealed cobalt concentrations of only 8.3and 7.1 ppb for Co/NCNT and Co2P/NPCNT respectively. Sinceour sensor shows a small response time of only 6 s for areading, the amount of cobalt that may be released during theperiod would be practically irrelevant. This quality of oursensor may be attributed to the protective cover provided bythe N,P dual doped CNTs.

The reproducibility of the Co2P/NPCNT modified GCEsensor was investigated by monitoring the response currentfrom 10 different Co2P/NPCNT-modified GCE towards fivedifferent glucose concentrations in 0.1 M KOH. The maximumrelative standard deviation (RSD, %) was exhibited for 0.1 mMglucose of only 4.17% (Fig. S8‡), demonstrating good reprodu-cibility of the sensor.

Diabetics need to monitor the blood glucose level on adaily basis wherein they are repetitively subjected to theanxiety and pain associated with frequent finger pricking. Anon-invasive method of glucose level monitoring would begreatly beneficial to alleviate this situation. Since a close corre-lation exists between saliva and blood glucose levels, the moni-

Fig. 6 (a) Amperometric response of Co2P/NPCNT to glucose at anapplied potential of +0.55 V; (b) fitting of the amperometric response;(c) anti-interference test of Co2P/NPCNT in the presence of variousinterfering species (1: glucose; 2: D-sucrose; 3: D-mannose; 4: urea; 5:L-cysteine; 6: Na2SO4; 7: NaCl) with a glucose concentration of 1 mMand the interferent concentration of 2 mM; (d) stability of Co2P/NPCNTin the presence of 1 mM glucose at +0.55 V. All the measurements werecarried out in 0.1 M KOH.

Paper Green Chemistry

1332 | Green Chem., 2017, 19, 1327–1335 This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

toring of salivary glucose levels may provide a painless avenueto diabetes control.53 To test the feasibility of our sensor underpractical (physiological) conditions, a Co2P/NPCNT-modifiedGCE sensor was applied to detect glucose in human saliva.The CV curves (Fig. 7a) show a regular increase in the oxi-dation current with the successive addition of glucose to adiluted saliva sample (10%), the corresponding calibrationcurve (Fig. 7b) shows a linear increase in the anodic currentwith increasing glucose concentration ranging from 1 to6 mM. Thus, the sensor has potential to be used for clinicalapplications.

Next, the electrocatalytic activity of Co2P/NPCNT towardsOER was evaluated in 1 M KOH solution (pH = 14). Fig. 8(a)shows the linear sweep voltammetry (LSV) curves of varioussamples obtained at a scan rate of 5 mV s−1 under a rotation of2000 rpm. As expected, a bare GCE electrode does not exhibitany activity towards OER. The performance of Co2P/NPCNT farsurpasses that of Co/NCNT, which is well-studied as anefficient OER catalyst.54 More interestingly, it even betters theactivity of RuO2, which is considered to be the state-of-the-artcatalyst for OER besides comparing favourably to variousrecently reported OER catalysts (Table 1).

Co2P/NPCNT exhibits a small onset potential of 270 mVwhile requiring an overpotential of only 370 mV to reach acurrent density of 10 mA cm−2, a value expected to be deliveredby a 10% efficient solar water-splitting device.55 In compari-son, Co/NCNT and RuO2 require an overpotential of 416 mVand 400 mV respectively to do the same. The kinetics of thecatalysis is evaluated using the Tafel plots (Fig. 8b). Co2P/NPCNT presents the smallest Tafel slope of 53 mV dec−1 fol-lowed by RuO2 and Co/NCNT with a Tafel slope of 74 and91 mV dec−1. A smaller Tafel slope is indicative of faster elec-tron transport which is duly reflected in the OER activity of thecatalyst in the same order. The electrochemically active surfacearea of Co2P/NPCNT and Co/NCNT was estimated by measur-ing the capacitance of the double-layer at the solid–liquidinterface by means of cyclic voltammetry56 (Fig. S9‡). Becauseof their similar morphologies, Cdl can be considered a reason-

Fig. 7 (a) CV curves Co2P/NPCNT-modified GCE in a human salivasample (10%) at different glucose concentrations in 0.1 M KOH; (b) thecorresponding calibration curve at +0.55 V.

Fig. 8 (a) OER polarization curves of different catalysts in 1 M KOH at ascan rate of 5 mV s−1; (b) the corresponding Tafel plots; (c) chargingcurrent density difference (ΔJ = Ja − Jc) plotted against scan rate toyield the double-layer capacitance (Cdl) of Co2P/NPCNT and Co/NCNT;(d) accelerated degradation tests (ADT) of Co2P/NPCNTs carried out at100 mV s−1 (inset: chronoamperometric response of Co2P/NPCNT at anoverpotential of 370 mV).

Table 1 Comparison of the OER activity of Co2P/NPCNT with some of the recently developed electrocatalysts

Catalyst Electrolyteη10 mA cm−2

(mV)Tafel slope(mV dec−1) Ref.

Au/mCo3O4 0.1 M KOH 440 46 57Ni–Co based catalyst 1 M KOH 390 75 58Co3O4 honeycomb like structures 0.1 M KOH 450 89 59CoP hollow polyhedra 1 M KOH 400 59 60Fe/Fe3C@N-graphitic layer 0.1 M KOH 580 — 15Ni Co2.7(OH)x 1 M KOH 350 65 61Mn3O4/CoSe2 0.1 M KOH 450 49 62NiSx film 1 M KOH 408 41 63FeNC sheets/NiO 0.1 M KOH 390 76 64CoNC sheets/NiO 0.1 M KOH 410 80 64Zn-Doped CoSe2 1 M KOH 356 88 65Co3S4/CNT 0.1 M KOH 432 70 66Co0.5Fe0.5S@N-MC 1 M KOH 410 159 67Co/NCNT 1 M KOH 416 91 This workCo2P/NPCNT 1 M KOH 370 53 This work

Green Chemistry Paper

This journal is © The Royal Society of Chemistry 2017 Green Chem., 2017, 19, 1327–1335 | 1333

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

able parameter to represent their active surface area. Fig. 8cshows the Cdl calculated for Co2P/NPCNT and Co/NCNT to be18.7 mF cm−2 and 10.8 mF cm−2, which correspond to theirrespective OER activity.

Stability of the catalyst is of prime importance for any appli-cation. Accelerated degradation tests (ADT) were carried out ata scan rate of 100 mV s−1 for 1000 cycles (Fig. 8d). No notablepositive shift in overpotential to reach the current density of10 mA cm−2 was observed after 1000 continuous cycles whilethe maximum current density reached at 1.85 V was reducedby only 8.6% (initial: 117.4 mA cm−2, after 1000 cycles:107.1 mA cm−2). The chronoamperometric response of Co2P/NPCNT was carried out at a fixed overpotential of 370 mV tofurther evaluate the stability of the catalyst (inset of Fig. 8d),where only a slight attenuation of 13.1% in the anodic currentwas observed after 10 h of operation.

The excellent performance of Co2P/NPCNTs towards OERcan be attributed to the following factors: (i) strong interactionbetween the Co2P nanoparticles and the encapsulating N,P co-doped CNTs leading to an enhanced electron transfer capa-bility, (ii) an improved conducting pathway provided by theCNTs and (iii) long-term structural stability of Co2P main-tained by CNT protection.

Conclusions

In summary, a simple one-step strategy was developed to syn-thesize phase-pure Co2P encapsulated N,P dual-doped CNTs.The method uses inexpensive and readily available precursorswithout involving the use of toxic phosphine (PH3) gas duringany stage of the reaction process. Co2P/NPCNTs were then forthe first time analysed as a non-enzymatic sensor towards theelectrochemical detection of glucose. Additionally, whenapplied as an anode for water oxidation, the OER activity ofCo2P/NPCNTs surpassed even the state-of-the-art RuO2 over anextended period of operation. Overall, this work provides analternative avenue to design metal phosphides integrated withheteroatom doped CNTs with improved properties which mayfind applications across diverse fields such as water-splitting,metal–air batteries, supercapacitors, biomolecule sensing etc.

Acknowledgements

D. D. is grateful to Dr Saswati Santra (a Dr D. S. Kothari post-doctoral fellow at MRC, IISc) for insightful suggestions andhelping with the manuscript preparation. Mr Subrata Senapatiis acknowledged for acquiring the PL and Raman spectra ofthe samples and Dr. Ramananda Chakrabarti for ICP-MSstudies. Mr Barun Kumar Barman kindly acquired the SEMimages. D. D. acknowledges the IISc for a researchscholarship. A. D. acknowledges a scholarship throughDST-KVPY. M. R. acknowledges the Indian Academy ofSciences (IAS) for a summer research fellowship. The ChemicalSciences Division, IISc provided the TEM facility. This work is

financially supported by the Council for Scientific andIndustrial Research (CSIR)-India. Most importantly, we arethankful to the nameless reviewers whose time and effortshave resulted in greatly improving the manuscript.

Notes and references

1 S. A. Kaveeshwar and J. Cornwall, Australas. Med. J., 2014, 7,45–48.

2 K. Deshpande, A. Jain, R. Sharma, S. Prashar and R. Jain,J. Indian Soc. Periodontol., 2010, 14, 207–212.

3 C. Li, B. Jiang, Z. Wang, Y. Li, M. S. A. Hossain, J. H. Kim,T. Takei, J. Henzie, Ö. Dag, Y. Bando and Y. Yamauchi,Angew. Chem., Int. Ed., 2016, 55, 12746–12750.

4 Y. Su, B. Luo and J. Z. Zhang, Anal. Chem., 2016, 88, 1617–1624.

5 C. Dong, H. Zhong, T. Kou, J. Frenzel, G. Eggeler andZ. Zhang, ACS Appl. Mater. Interfaces, 2015, 7, 20215–20223.

6 Z. Li, Y. Chen, Y. Xin and Z. Zhang, Sci. Rep., 2015, 5,16115.

7 X. Niu, M. Lan, C. Chen and H. Zhao, Talanta, 2012, 99,1062–1067.

8 J.-S. Ye, C.-W. Chen and C.-L. Lee, Sens. Actuators, B, 2015,208, 569–574.

9 C. Li, H. Wang and Y. Yamauchi, Chem. – Eur. J., 2013, 19,2242–2246.

10 X.-C. Dong, H. Xu, X.-W. Wang, Y.-X. Huang, M. B. Chan-Park, H. Zhang, L.-H. Wang, W. Huang and P. Chen, ACSNano, 2012, 6, 3206–3213.

11 J. Tang, S. Wu, T. Wang, H. Gong, H. Zhang, S. M. Alshehri,T. Ahamad, H. Zhou and Y. Yamauchi, ACS Appl. Mater.Interfaces, 2016, 8, 2796–2804.

12 X.-Y. Yu, Y. Feng, B. Guan, X. W. Lou and U. Paik, EnergyEnviron. Sci., 2016, 9, 1246–1250.

13 S. Park, Y. Shao, J. Liu and Y. Wang, Energy Environ. Sci.,2012, 5, 9331–9344.

14 B. Jiang, C. Li, V. Malgras, Y. Bando and Y. Yamauchi,Chem. Commun., 2016, 52, 1186–1189.

15 B. K. Barman and K. K. Nanda, Green Chem., 2016, 18, 427–432.

16 B. Jiang, C. Li, V. Malgras, M. Imura, S. Tominaka andY. Yamauchi, Chem. Sci., 2016, 7, 1575–1581.

17 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher,Q. Mi, E. A. Santori and N. S. Lewis, Chem. Rev., 2010, 110,6446–6473.

18 T. Reier, M. Oezaslan and P. Strasser, ACS Catal., 2012, 2,1765–1772.

19 B. Bayatsarmadi, Y. Zheng, Y. Tang, M. Jaroniec andS.-Z. Qiao, Small, 2016, 12, 3703–3711.

20 B. Bayatsarmadi, Y. Zheng, C. S. Casari, V. Russo andS.-Z. Qiao, Chem. Commun., 2016, 52, 11947–11950.

21 J. Duan, S. Chen, A. Vasileff and S. Z. Qiao, ACS Nano,2016, 10, 8738–8745.

22 A. Vasileff, S. Chen and S. Z. Qiao, Nanoscale Horiz., 2016,1, 41–44.

Paper Green Chemistry

1334 | Green Chem., 2017, 19, 1327–1335 This journal is © The Royal Society of Chemistry 2017

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online

23 V. V. T. Doan-Nguyen, S. Zhang, E. B. Trigg, R. Agarwal,J. Li, D. Su, K. I. Winey and C. B. Murray, ACS Nano, 2015,9, 8108–8115.

24 A. Lu, Y. Chen, H. Li, A. Dowd, M. B. Cortie, Q. Xie, H. Guo,Q. Qi and D.-L. Peng, Int. J. Hydrogen Energy, 2014, 39,18919–18928.

25 X. Chen, M. Cheng, D. Chen and R. Wang, ACS Appl. Mater.Interfaces, 2016, 8, 3892–3900.

26 J. F. Callejas, C. G. Read, E. J. Popczun, J. M. McEnaneyand R. E. Schaak, Chem. Mater., 2015, 27, 3769–3774.

27 H.-F. Wang, C. Tang, X. Zhu and Q. Zhang, J. Mater. Chem.A, 2016, 4, 3379–3385.

28 C. Tang, H.-S. Wang, H.-F. Wang, Q. Zhang, G.-L. Tian,J.-Q. Nie and F. Wei, Adv. Mater., 2015, 27, 4516–4522.

29 C.-C. Hou, S. Cao, W.-F. Fu and Y. Chen, ACS Appl. Mater.Interfaces, 2015, 7, 28412–28419.

30 Y. Bai, H. Zhang, Y. Feng, L. Fang and Y. Wang, J. Mater.Chem. A, 2016, 4, 9072–9079.

31 M. Zhuang, X. Ou, Y. Dou, L. Zhang, Q. Zhang, R. Wu,Y. Ding, M. Shao and Z. Luo, Nano Lett., 2016, 16, 4691–4698.

32 J. Deng, M. Li and Y. Wang, Green Chem., 2016, 18, 4824–4854.

33 X. Li, Y. Fang, F. Li, M. Tian, X. Long, J. Jin and J. Ma,J. Mater. Chem. A, 2016, 4, 15501–15510.

34 D. Das and K. K. Nanda, Nano Energy, 2016, 30, 303–311.35 S. T. Oyama, T. Gott, H. Zhao and Y.-K. Lee, Catal. Today,

2009, 143, 94–107.36 D. Das, S. L. Shinde and K. K. Nanda, ACS Appl. Mater.

Interfaces, 2016, 8, 2181–2186.37 C. Ehrhardt, M. Gjikaj and W. Brockner, Thermochim. Acta,

2005, 432, 36–40.38 J. Liu, H. Wang and M. Antonietti, Chem. Soc. Rev., 2016,

45, 2308–2326.39 Y. Qing, T. Kaibin, L. Qianrong, Y. Hao, W. Chunrui and

Q. Yitai, Nanotechnology, 2004, 15, 918.40 X. Zou, X. Huang, A. Goswami, R. Silva, B. R. Sathe,

E. Mikmeková and T. Asefa, Angew. Chem., 2014, 126, 4461–4465.

41 K. Gong, F. Du, Z. Xia, M. Durstock and L. Dai, Science,2009, 323, 760–764.

42 Y.-P. Zhu, Y. Liu, Y.-P. Liu, T.-Z. Ren, G.-H. Du, T. Chen andZ.-Y. Yuan, J. Mater. Chem. A, 2015, 3, 11725–11729.

43 Y.-P. Zhu, X. Xu, H. Su, Y.-P. Liu, T. Chen and Z.-Y. Yuan,ACS Appl. Mater. Interfaces, 2015, 7, 28369–28376.

44 S. Feng, M. C. dos Santos, B. R. Carvalho, R. Lv, Q. Li,K. Fujisawa, A. L. Elías, Y. Lei, N. Perea-López, M. Endo,M. Pan, M. A. Pimenta and M. Terrones, Sci. Adv., 2016, 2,e1600322.

45 H. Yan, C. Tian, L. Wang, A. Wu, M. Meng, L. Zhao andH. Fu, Angew. Chem., Int. Ed., 2015, 54, 6325–6329.

46 P. Jiang, Q. Liu, C. Ge, W. Cui, Z. Pu, A. M. Asiri andX. Sun, J. Mater. Chem. A, 2014, 2, 14634–14640.

47 C. Kong, L. Tang, X. Zhang, S. Sun, S. Yang, X. Song andZ. Yang, J. Mater. Chem. A, 2014, 2, 7306–7312.

48 Q.-Q. Sun, M. Wang, S.-J. Bao, Y. C. Wang and S. Gu,Analyst, 2016, 141, 256–260.

49 M. Chowdhury, F. Cummings, M. Kebede and V. Fester,Electroanalysis, 2016, DOI: 10.1002/elan.201600440.

50 L. Kang, D. He, L. Bie and P. Jiang, Sens. Actuators, B, 2015,220, 888–894.

51 J. Balamurugan, T. D. Thanh, G. Karthikeyan, N. H. Kimand J. H. Lee, Biosens. Bioelectron., DOI: 10.1016/j.bios.2016.09.077.

52 L. T. Hoa, J. S. Chung and S. H. Hur, Sens. Actuators, B,2016, 223, 76–82.

53 W. Zhang, Y. Du and M. L. Wang, Sens. Biosensing Res.,2015, 4, 23–29.

54 A. Aijaz, J. Masa, C. Rösler, W. Xia, P. Weide, A. J. R. Botz,R. A. Fischer, W. Schuhmann and M. Muhler, Angew.Chem., Int. Ed., 2016, 55, 4087–4091.

55 T. Y. Ma, S. Dai, M. Jaroniec and S. Z. Qiao, Angew. Chem.,Int. Ed., 2014, 53, 7281–7285.

56 S. Trasatti and O. A. Petrii, J. Electroanal. Chem., 1992, 327,353–376.

57 X. Lu, Y. H. Ng and C. Zhao, ChemSusChem, 2014, 7, 82–86.58 H. Ju, Z. Li and Y. Xu, Mater. Res. Bull., 2015, 64, 171–174.59 L. Li, T. Tian, J. Jiang and L. Ai, J. Power Sources, 2015, 294,

103–111.60 M. Liu and J. Li, ACS Appl. Mater. Interfaces, 2016, 8, 2158–

2165.61 J. Nai, H. Yin, T. You, L. Zheng, J. Zhang, P. Wang, Z. Jin,

Y. Tian, J. Liu, Z. Tang and L. Guo, Adv. Energy Mater.,2015, 5, 1401880–1401886.

62 M.-R. Gao, Y.-F. Xu, J. Jiang, Y.-R. Zheng and S.-H. Yu,J. Am. Chem. Soc., 2012, 134, 2930–2933.

63 H. Li, Y. Shao, Y. Su, Y. Gao and X. Wang, Chem. Mater.,2016, 28, 1155–1164.

64 J. Wang, K. Li, H.-x. Zhong, D. Xu, Z.-l. Wang, Z. Jiang,Z.-j. Wu and X.-b. Zhang, Angew. Chem., Int. Ed., 2015, 54,10530–10534.

65 Q. Dong, Q. Wang, Z. Dai, H. Qiu and X. Dong, ACS Appl.Mater. Interfaces, 2016, 8, 26902–26907.

66 H. Wang, Z. Li, G. Li, F. Peng and H. Yu, Catal. Today,2015, 245, 74–78.

67 M. Shen, C. Ruan, Y. Chen, C. Jiang, K. Ai and L. Lu, ACSAppl. Mater. Interfaces, 2015, 7, 1207–1218.

Green Chemistry Paper

This journal is © The Royal Society of Chemistry 2017 Green Chem., 2017, 19, 1327–1335 | 1335

Publ

ishe

d on

16

Janu

ary

2017

. Dow

nloa

ded

by I

ndia

n In

stitu

te o

f Sc

ienc

e on

12/

04/2

017

05:4

1:05

. View Article Online