A polythiophene derivative bearing TEMPO as a cathode material for rechargeable batteries

European Polymer Journal 47 (2011) 2283–2294

Contents lists available at SciVerse ScienceDirect

European Polymer Journal

journal homepage: www.elsevier .com/locate /europol j

A polythiophene derivative bearing TEMPO as a cathode materialfor rechargeable batteries

M. Aydın a,1, B. Esat b,⇑, Ç. Kılıç c,2, M.E. Köse d,3, A. Ata e,4, F. Yılmaz a,5

a Department of Chemistry, Gebze Institute of Technology, 41400 Gebze, Kocaeli, Turkeyb Department of Chemistry, Fatih University, 34500 Buyukcekmece, Istanbul, Turkeyc Department of Physics, Gebze Institute of Technology, 41400 Gebze, Kocaeli, Turkeyd Department of Chemistry and Biochemistry, North Dakota State University, Fargo, ND 58108-6050, USAe Department of Materials Science and Engineering, Gebze Institute of Technology, 41400 Gebze, Kocaeli, Turkey

a r t i c l e i n f o

Article history:Received 8 March 2011Received in revised form 18 July 2011Accepted 3 September 2011Available online 12 September 2011

Keywords:PolythiopheneTEMPOElectrodeRechargeableBattery

0014-3057/$ - see front matter � 2011 Elsevier Ltddoi:10.1016/j.eurpolymj.2011.09.002

⇑ Corresponding author. Tel.: +90 (212) 866 33 0034 02.

E-mail addresses: [email protected] (M. Aydın(B. Esat), [email protected] (Ç. Kılıç), Muham(M.E. Köse), [email protected] (A. Ata), fyilmaz@gyt

1 Tel.: +90 (262) 605 31 00; fax: +90 (262) 605 312 Tel.: +90 (262) 605 13 27; fax: +90 (262) 653 843 Tel.: +1 701 231 8694; fax: +1 701 231 8831.4 Tel.: +90 (262) 605 11 76; fax: +90 (262) 653 845 Tel.: +90 (262) 605 31 33; fax: +90 (262) 605 31

a b s t r a c t

A polythiophene derivative bearing TEMPO radical was synthesized by oxidative chemicalpolymerization of its monomer. The polymer had a high spin density (2.05 � 1021 spins/gof polymer). CV studies of the polymer showed that the electrochemical redox reaction ofthe TEMPO radicals were completely reversible. We demonstrated, for the first time, con-struction and charge/discharge characteristics of an organic radical battery utilizing aTEMPO bearing polythiophene based cathode material. The battery had an initial specificdischarge capacity of 79 A h/kg (87% of the theoretical capacity) and an average outputvoltage of 3.6 V. The specific energy capacity initially discharged was 268 W h/kg.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

The growing public demand for the use of portable de-vices such as laptop computers, cell phones, and PDAsnecessitated the fabrication of new secondary batterieswith improved properties such as very fast rechargeability,flexibility, smaller size, light-weight, and environmentalbenignity together with high power and energy capacities.Li-ion batteries are the current choice of power source inportable devices due to increased cycle- and calendar-life,superior charge retention, improved reliability, high spe-

. All rights reserved.

; fax: +90 (212) 866

), [email protected]@ndsu.edu

e.edu.tr (F. Yılmaz).05.90.

90.05.

cific power (W/kg), and specific energy (W h/kg). Currentlithium-ion batteries use metallophosphates, lithium spi-nel materials, and metal-oxides as cathodes where lithiuminsertion–extraction was achieved by an intercalationmechanism within the crystal structure of active material[1]. Carbon based materials are commercially used as an-ode materials due to their flat charge–discharge plateauand excellent cycling ability. Moreover, graphitic carbonanode is safer than metallic lithium which is very reactiveunder ambient conditions. The specific capacity of graphiteanode is 372 A h/kg, whereas a typical LiMn2O4 cathodehas a reversible capacity of 120 A h/kg. Lithium metal itselfhas a specific capacity of 3860 A h/kg. However, its use insecondary batteries is avoided due to safety reasons. Manygroups have focused their research on developing new andimproved anode and cathode materials as well as on newelectrolytes that can be used in secondary batteries. Specif-ically, the studies aim to exploit novel materials to improvedischarge capacity of cathodes as it is the bottle-neck forachieving light weight batteries. Particular attention is ori-ented toward organic cathodes due to their low weight,

2284 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

low cost of production, ease of synthesis, processability,and molding ability as well as their environmental com-patibility [2]. Conducting polymers such as polyacetylene,polythiophene, and polyanilene have been investigated aspotential electrode materials [3]. Their use in rechargeablebatteries has been hampered because they exhibit lowdoping levels which lower their capacities and slow elec-trode kinetics which limit charge–discharge rates of thebatteries. As an alternative, redox polymers carrying func-tional groups which can be reversibly oxidized or reducedon their backbone may be used. One example of such func-tional groups is nitroxide radicals. Among nitroxides, TEM-PO (2,2,6,6-tetramethylpiperidine-1-oxyl) is known to be avery robust radical and therefore has been chosen as anelectro-active pendant functional group to be used in re-dox polymers. TEMPO radical has two redox couples, asshown in Scheme 1, which makes it even more attractivesince this gives it a propensity to be used both as an n-type(anode) material and as a p-type (cathode) material. Thefirst attempt on the use of an organic radical as an elec-trode-active group for charge storage in a lithium batterywas made by Nakahara’s group in 2002 [4]. They demon-strated the use of poly(2,2,6,6-tetramethylpiperidinyloxy-4-yl methacrylate) (PTMA), a polymer with pendantTEMPO radicals, as the organic polyradical cathode-activematerial in a rechargeable battery. Following their explora-tion, more work on polymers carrying nitroxide radicalswas published by Nishide et al. [5–8]. Organic polymerswhich have TEMPO radical attached to the aliphatic ornonconjugated polymer backbone were shown to exhibitexcellent properties such as rapid electron transfer, highcharge capacity, transparent film formation, and an outputvoltage of 3.6 V against a lithium anode. Some of theirmost striking features are the efficient electron hoppingwithin the polymer film allowing fast electron transfer rateand quantitative redox reactions occurring under a con-stant electronic potential, in contrast to those of the p-con-jugated conducting polymers which have small chargestorage capacity and the fluctuating voltage due to thelow doping levels and the potential shifts due to dopingprocess [9]. The theoretical charge capacity of pure PTMAis 112 A h/kg [10] compared to over 140 A h/kg for transi-tion metal based materials. In order to increase the energy

Scheme 1. The two redox cou

density of radical batteries, it is necessary to increase thenumber of radical centers present per gram of polymerand/or to increase the cell voltage by using radicals withmore positive oxidation potentials. Several groups havepursued this challenging target and reported new nitroxidebearing polymers. In their efforts, nitroxides were attachedmostly to polymers with nonconducting type backboneand, in only few examples, to polymers with conductivebackbone such as polyacetylene [11–24]. Polymers substi-tuted with nitroxides other than TEMPO have also beensynthesized such as poly(p-tert-butylaminoxy-styrene)[25] , polyacetylene, polynorbornene, or cellulose bearing2,2,5,5-tetramethyl-1-pyrrolidinyloxy moieties [23,24,26,27] or polyoxirane substituted with 2,2,5,5-tetramethyl-3-pyrroline-N-oxyls [18]. However, the theoretical chargecapacities of these new nitroxide polymers did not signifi-cantly exceed that of PTMA. In fact, the highest value(135 A h/kg) reached by poly(2,2,6,6-tetramethylpiperidi-nyloxy-4-yl vinyl ether) [13] is still lower than that ofthe currently used LiCoO2 (about 140 A h/kg). The oxida-tion potential of other nitroxides attached to the polymerscited above was practically similar to that of TEMPO(0.88 V vs. NHE).

Polythiophene (PT) and its derivatives have been usedas both cathode- and anode-active materials in batteries.Several rechargeable batteries with p-doped polythio-phene cathodes have exhibited specific charge capacitiesof 25–100 A h/kg. Specific energy capacities of 50–325W h/kg have been reported [28]. A specific charge capacityof about 90 A h/kg has been reported for a cell employingpoly(3-methylthiophene) (P3MT) as the cathode materialagainst an aluminum anode in a propylene carbonate(PC)/ethylene carbonate (EC) mixture containing 1 M LiBF4

as an electrolyte [29]. Also, composites containing P3MT, acarbon based conducting agent and a polymeric binderhave been prepared and used as cathode material againstLi metal in rechargeable cells. One study showed theeffects of different binders on cell capacity in such cells[30]. In this study, discharge capacities were reported tobe 25, 50, and about 80 A h/kg for cells utilizing compositecathodes with PVdF [poly(vinylidine difluoride)], SEBS[poly(styrene-ethylene-butylene-styrene)], and SSEBS [sul-fonated poly(styrene-ethylene-butylene-styrene)] binders,

ples of TEMPO radical.

M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294 2285

respectively. The doping of the P3MT polymer occurredaround 3.8 V whereas the dedoping was observed at 2.8 V.

PT derivatives carrying stable free radicals have previ-ously been synthesized and tested for their electric andmagnetic properties [31]. However, they have never beenused as an electrode material. Attempts of polymerizationof thiophene monomers with t-butyl nitroxide, nitronylnitroxide, or verdazyl groups directly attached to the 3-po-sition of thiophene ring have not been successful. This hasbeen attributed to the steric hindrance at the 2-positionimposed by a large substituent at 3-position [31]. How-ever, a PT with 2,6-di-tert-butylphenoxyl group attachedto 3-position of the thiophene ring has been successfullysynthesized and magnetically characterized [31].

In this study, we report the synthesis of a first PT homo-polymer with pendant TEMPO electro-active species andits utilization as a cathode-active material in a recharge-able battery.

2. Experimental

All reagents were used as obtained unless stated other-wise. Celgard 2400 separators were kindly donated by Cel-gard Co. The carbon conducting agent (KS6L) used inmaking of cathode composite material was donated byTimcal Ltd.

2016 size steel coin cell casings including a circularplastic gasket was purchased from MTI Corp.

The monomer and the polymer were synthesized chem-ically by using the procedures given below. They werecharacterized via elemental analysis by using Thermo Sci-entific Flash EA 2000 Series – Organic Elemental Analyzer,via IR-ATR spectroscopy using Perkin Elmer BXII FT-IRspectrometer equipped with a diamond ATR kit, and viaESR spectroscopy using Bruker EMX series spectrometerdesigned for measurements in the X-band (9.5 GHz). Massspectrometric measurements were recorded on a Micro-mass Ultima API spectrometer.

ESR spin count studies on polymer were performed on acarefully weighed 2 mg of solid polymer sample. The ESRsignal at about 3360 Gauss (G; 1 G = 10�4 Tesla) was dou-ble integrated and spin concentration was determined bycomparing it to double integration of the ESR signal ob-tained from 1 mg sample of 2-hydroxy TEMPO referencematerial.

Cyclic voltammetry (CV) studies were performed on CHInstruments CHI 842B electrochemical analyzer by dissolv-ing the monomer in electrolyte or by attaching the solidpolymer and composite sample onto a Pt working elec-trode (0.12 cm2). A 0.1 M acetonitrile solution of tetrabu-tylammonium perchlorate (Bu4NClO4) as electrolyte, a Ptwire counter electrode, an Ag/AgCl reference electrodesubmersed in saturated solution of Bu4NCl in acetonitrilewere used. A scan rate of 0.05 V/s was used for the TEMPOreference and the monomer. A slower scan rate of 0.005 V/s were used for the polymer and composite samples. Theelectrolyte solution was degassed and the experimentswere conducted under argon.

The thermal stability was determined via thermogravi-metric analysis (TGA) by using Perkin Elmer Instruments

STA 6000 model analyzer. The TGA thermograms were re-corded for 5 mg powder samples at a heating rate of 10 �C/min in the temperature range of 30–800 �C under nitrogenatmosphere.

The electrical conductivity of the polymer was studiedin the range of 20–80 �C. The samples used were in theform of circular pellets of 10 mm diameter and 0.3 mmthickness. The pellets were sandwiched between gold elec-trodes and the conductivities were measured by usingNovocontrol dielectric impedance analyzer in the fre-quency range 1–3 MHz. The temperature was controlledwith a Novocool Cryosystem. The dielectric data (e0 ande00) were collected during heating as a function offrequency.

The coin cell was analyzed by using MTI 8 Channel Bat-tery Analyzer (0.1–10 mA, up to 5 V). It was subjected toconstant current charge and discharge cycles between 2.5and 3.8 V. All specific energy and charge capacities re-ported are on per kilogram of the active polymeric materialbasis.

2.1. Synthesis of the monomer [2,2,6,6-tetramethylpiperidin-1-oxyl-4-yl 2-(thiophen-3- yl)acetate-3ThAcTempEst]

3ThAcTempEst monomer was synthesized according tothe reported procedure [32,33]. 3-thiopheneacetic acid(1.42 g, 10 mmol) was added to a solution of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride(EDCI�HCl) (2.87 g, 15 mmol) and 4-dimethylamino-pyri-dine (DMAP) (0.68 g, 5.5 mmol) in 70 ml dry THF at roomtemperature under nitrogen. 4-hydroxy-TEMPO (1.72 g,10 mmol) was added to the solution, and the resulting mix-ture was stirred at room temperature for 2 days. The pre-cipitate was filtered and the reaction mixture was washedwith water (80 mL) four times, and then with brine solu-tion. Finally, the organic layer was dried over anhydrousMgSO4. After filtration, the solvent was evaporated to afforda crude monomer. Crude product was purified by columnchromatography (silica gel, gradient elution: 5/1 ? 7/3hexane/ethyl acetate eluent) to yield a dark orange solid(75% yield). M.p. 53–65 �C. FT-IR (ATR, cm�1): 3102 (thio-phene ring C–H stretch), 2990, 2964, 2940, 2872, 1732(C@0 stretch), 1364(N–O stretch), 1340, 1316, 1262, 1172,798. Elemental analysis: C15H22NO3S� (theoretical); C:60.78; H: 7.48; N: 4.73; S: 10.82; C15H22NO3S� (found); C:60.50; H: 7.30; N: 3.48; S:10.65. Mass spectrometry: m/z = 298.15 [M+2], (theoretical m/z = 296.41). ESR (CHCl3

solution): Three peaks with aN = 14.4 G.

2.2. Synthesis of the polymer [Poly 2,2,6,6-tetramethylpiperi-din-1-oxyl-4-yl 2-(thiophen-3-yl)acetate-P3ThAcTempEst]

Polymer was synthesized according to the reported pro-cedure [34,35]. A solution of 2.01 g (12.4 mmol) of anhy-drous FeCl3 in 11.5 mL of CH3NO2 was added dropwiseover 20 min to a solution of 0.75 g (3.04 mmol) of 3ThAc-TempEst in 35 mL of CCl4 at �10 �C. The mixture was stir-red at 0 �C under a flux of dry nitrogen for 100 min. Themolar ratio of the oxidant to monomer was 4:1. Theblue-black polymer suspension obtained was poured intoa large excess of methanol (1 L) to precipitate P3ThAcTem-

2286 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

pEst, which turned orange. The precipitate was collectedby filtration, purified by repeated washing with freshmethanol (using a soxhlet apparatus) and with deionizedwater to remove the residual FeCl3 and the oligomers toyield the pure polymer (40%). FT-IR (ATR, cm�1): 3543,3514, 3084, 3000, 2976, 2946, 2912, 1731 (C@O), 1621,1475, 1394, 1361 (N–O), 1315, 1174, 1161, 706. UV(in THF) of the soluble part of the polymer: kMax = 412 nm.ESR: 2.06 � 1021 spins/g. Elemental analysis: C15H22NO3S�

(theoretical); C: 61.20; H: 6.85; N: 4.76; S: 10.89;C15H22NO3S� (found); C: 60.75; H: 6.81; N: 3.68; S:10.61.

2.3. Preparation of the composite cathode material

60 mg P3ThTempEst polymer, 20 mg PVdF, and 120 mgcarbon conducting agent (Timcal Ltd., KS6L) were mixedwith N-methylpyrrolidone (NMP) for about 45 min in aball-mill. The resulting slurry was evaporated to drynessin vacuum oven at 40 �C. The composite obtained waspressed into a circular pellet of about 10 mm diameterand 0.2 mm thickness.

2.4. Cell construction

A 2016 size coin cell was prepared by sandwiching apolypropylene separator (Celgard 2400) between the cath-ode material and Li metal anode (dimensions 6 � 6 � 0.35mm) in a glove-box under dry He gas. An electrolyte solu-tion containing 1 M lithium phosphorus hexafluoride(LiPF6) in 1/1 (v/v) mixture of ethylene carbonate (EC)and diethyl carbonate (DEC) was added. The cell wassealed and taken out of the glove-box for determinationof its charge capacity.

3. Results and discussion

3.1. Monomer synthesis and characterization

Monomer, 2,2,6,6-tetramethylpiperidin-1-oxyl-4-yl 2-(thiophen-3-yl)acetate (3ThAcTempEst), was synthesizedfrom the reaction between 3-thiophenyl acetic acidand 4-hyroxy-2,2,6,6-tetramethylpiperidin-1-oxyl radical(4-hydroxy TEMPO) in the presence of EDCI�HCl and DMAP(Scheme 2) with a high yield (75%) as proven by FT-IRspectroscopy, mass spectrometry, and elemental analysis(refer to the Section 2 and the Supplementary data formore information). Its successful synthesis was evidencedby the presence of a carbonyl stretching peak at 1732

Scheme 2. Synthesis of 3Th

cm�1 which is higher than that of the parent carboxylicacid (1682 cm�1), the absence of the broad O–H stretchingpeak in 2500–3500 cm�1 region which was observable inthe spectrum of the parent acid in the FT-IR spectrum,and the presence of an m/z = 298.15 (M+2 peak) in themass spectrum. The cyclic voltammogram of this monomerin 0.1 M Bu4NClO4 acetonitrile solution showed a nearlycompletely reversible and quite rapid redox behavior at0.5 V (DEpp = 0.07 V) with respect to Ag/AgCl referenceelectrode (Fig. 1). The ESR spectrum of CHCl3 solutionshows a triplet which is typical of a TEMPO nitroxide rad-ical with nitrogen hyperfine splitting constant (aN) of14.4 G (g = 2.00632) (Fig. 2).

3.2. Polymer synthesis and characterization

The attempt to electropolymerize the monomer failedpossibly due to presence of nitroxide radical moiety whichprobably has a lower oxidation potential than the thio-phene ring for this particular molecule.

After the failure of our electropolymerization attempt,chemical polymerization of the monomer with FeCl3 inCCl4–CH3NO2 solvent mixture at near 0 �C was performed(Scheme 3). The use of such a solvent system which solubi-lizes the oxidant (FeCl3) yielded an orange colored polymerwith about 40% yield after continuous washing with meth-anol in a soxhlet apparatus to get rid of excess FeCl3, mono-mer and soluble oligomers. The effects of concentration ofthe monomer, reaction temperature and time have alsobeen investigated. In general, decreased spin concentrationon the polymer and polymer yield have been observedwhen the monomer concentration is above 0.1 M, reactiontemperature is high (around room temperature) andpolymerization reaction time is long (more than 2 h). Thepolymerization mechanism is believed to involve a radi-cal-cation or cation mechanism [36,37] because the grow-ing polymer chain cannot be neutral under these stronglyoxidizing conditions. Quite reasonable yields of polymerwith high spin density have been obtained when monomerconcentration is around 0.5–0.65 M, the reaction tempera-ture is kept near 0 �C and the reaction time is relativelyshort (2 h). Under these conditions, reactions of the TEMPOester side chain with FeCl3 which may lead to ester hydro-lysis or removal of the nitroxide group oxygen atom andalso with the radicals possibly formed on the growingpolymer chain must have been avoided, to at least someextent, leading to an appreciable yield of the desired poly-mer with high spin density.

The polymer had almost no solubility in tetrahydrofu-ran (THF). Only a small portion of the polymer which could

AcTempEst monomer.

1.2 1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6-8

-6

-4

-2

0

2

4

6

Cur

rent

(μA

)

Voltage (V)

Fig. 1. CV study of 3ThAcTempEst. Scan rate = 0.05 V/s, 1 M Bu4NClO4 in acetonitrile as electrolyte, Ag/AgCl reference electrode.

3300 3320 3340 3360 3380 3400 3420-15000

-10000

-5000

0

5000

10000

15000

Inte

nsity

Gauss

AN= 14.361 G.

Fig. 2. X-band ESR spectrum of 3ThAcTempEst monomer in CHCl3.

Scheme 3. Synthesis of P3ThAcTempEst polymer.

M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294 2287

be solubilized in this solvent was analyzed by gel perme-ation chromatography (GPC), and UV spectroscopy. UVspectrum showed an absorption maximum in THF at412 nm which was absent in the monomer indicating thepresence of a p-conjugated structure in the polymer back-bone. The molecular weight of this soluble part was quitehigh (Mn = 356000, Mw = 468200 with PDI = 1.315). Aneven higher molecular weight may be expected for theTHF insoluble portion of the polymer.

The polymer ESR spectrum (in powder form) was alsorecorded to show a broad peak centering around 3367 G(g = 2.01308) at 9.49 GHz frequency (Fig. 3) which is typi-cally expected for an amorphous solid nitroxide radicalsample. Double integration of this peak and comparisonwith that of a TEMPO radical reference sample of carefullymeasured weight revealed the spin concentration as2.05 � 1021 spins/g of polymer which nearly correspondsto 1 spin per repeating unit.

3200 3250 3300 3350 3400 3450 3500 3550

-2000

-1500

-1000

-500

0

500

1000

1500

2000

Inte

nsity

Gauss

Fig. 3. X-band ESR spectrum of P3ThAcTempEst.

2288 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

TGA analysis of the polymer proved that it is thermallystable up to about 250 �C (Fig. 4). Variable temperatureconductivity measurements between 20 and 80 �C on poly-mer sample were performed by using an impedance spec-trometer (Fig. 5). This study demonstrated that the DCconductivity was below 10�13 S/cm and showed very littledependence on temperature. This implies that the polymeris in the undoped original state and it does not undergo anappreciable structural change up to 80 �C. The low conduc-tivity may be attributed to the absence of an extended con-jugation length over polymer chain due to formation of aconsiderable amount of head-to-head (H–H) couplingsamong monomer units in the polymer backbone duringrandom chemical polymerization. CV studies revealed thatthe polymer is oxidized around 1.06 V (DEpp = 0.027 V)(Fig. 6). The polymer had a smaller separation betweenits anodic oxidation and reduction peaks compared to that

0 100 200 30020

30

40

50

60

70

80

90

100

110

(%) W

eigh

t

Temper

Fig. 4. TGA analysis of

of poly(2,2,6,6-tetramethylpiperidinyloxy methacrylate)(PTMA) (ca. 0.146 V) [4,5]. The smaller gaps between thereduction and oxidation peaks generally imply larger elec-trode reaction rates, which suggest that P3ThTempEstpolymer will exert high power rates during charge/dis-charge processes of battery under constant battery processconditions. The CV studies at various scan rates indicatedthat the peak current is proportional to the scan rate, sug-gesting a fast electron transfer process in the polymer film.Poly[ethyl 2-(3-thiophenyl) acetate] was also prepared andanalyzed by CV for comparison to make sure that the oxi-dation peak observed around 1.06 V is due to the oxidationof the TEMPO moiety rather than being due to oxidation ofthiophene rings in polymer backbone. This polymer did notshow any redox peak in this potential range (see the Sup-plementary data section), meaning that the redox behaviorobserved is that of the TEMPO nitroxide moiety only.

400 500 600 700

ature ( oC)

P3ThAcTempEst.

10-1 100 101 102 103 104 105 106 107

Frequency [Hz]

10-1

410

-12

10-1

010

-810

-610

-4

Con

duct

ivity

' [S/

cm]

Temp. [°C]=19.561 AC Volt [Vrms]=1.00 Temp. [°C]=40.094 AC Volt [Vrms]=1.00 Temp. [°C]=60.253 AC Volt [Vrms]=1.00 Temp. [°C]=80.154 AC Volt [Vrms]=1.00

10-1

100

101

102

Perm

ittiv

ity''

Fig. 5. Impedance spectroscopy of P3ThAcTempEst.

M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294 2289

3.3. The composite cathode material and its characterization

A cathode composite material was prepared by ball-milling a mixture containing 30% polymer, 60% KS6Lgraphite (conductive agent), and 10% PVdF (binder) byweight in a very small amount of NMP. The slurry obtainedwas dried in vacuum at 40 �C. It was then pressed into acircular pellet (10 mm diameter, 0.25 mm thickness). CVstudy of this composite cathode material was performed(Fig. 7) under the same conditions with CV study of thepolymer, showing a reversible oxidation at about 1.03 Vwith respect to Ag/AgCl reference electrode. The conduc-tivity of these pellets was about 9 S/cm as determined byfour-probe technique. A small correction was also includedin this reported conductivity value to compensate for anyerrors that may be induced by the small sample diameter.

Scanning electron microscopy (SEM) studies were per-formed on the polymer, the composite cathode materialand KS6L graphite (Fig. 8). As can seen in Fig. 8(a), polymerpowder appears to be a collection of aggregated polymerchains with particle size smaller than 5 lm. KS6L carbonconducting agent obtained from Timcal Corp. seems to becomposed of graphite chips with particle size in the rangeof 1.5–6 lm (Fig. 8(c)). PVdF polymer is in the form ofgranules much smaller than 1 lm (Fig. 8(d)). The SEM pic-tures of the composite material clearly indicate that poly-mer was spread over KS6 graphite chips in a layeredfashion and the composite has a particle size of about 2–5 lm (Fig. 8(b)).

3.4. Cell construction and characterization

A R2016 size button cell was assembled in a glove-boxunder dry He gas atmosphere by using this cathode, Cel-gard 2400 separator and Li sheet anode immersed in anelectrolyte solution comprised of 1 M LiPF6 in 1:1 mixtureof EC and DEC.

This cell was subjected to constant current charge anddischarge cycles between 2.5 and 3.8 V at charge and dis-charge rates of 0.25 and 0.05 mA, respectively. During thecharging process, the radical polymer (NO�) in the cathodeis oxidized to the oxoammonium cation (N+@O). Duringthe discharging process, the nitroxide radical is regener-ated by reduction of the oxoammonium cation (Schemes1 and 4). A nearly constant voltage plateau between 3.70and 3.55 V was observed during discharge of this cell asdemonstrated in Fig. 9. The initial specific charge and dis-charge capacities were 85.0 and 79.2 A h/kg, respectively,but they slowly decreased to 78.5 and 74.0 A h/kg afterthe first 10 cycles, and then to 63.3 and 61.1 A h/kg atthe end of 50 cycles (corresponding to an efficiency ofnearly 96%). An unrecoverable discharge capacity of about11 A h/kg was observable below 3.00 V which may beattributed to an irreversible reduction of the nitroxide rad-ical to its aminoxy anion (N–O�) under these conditions.The initial specific charge and discharge energy capacitiesobtained were 306 and 268 W h/kg after the first 10 cycles(Fig. 10). These values dropped down to 228 and 213 W h/kg after 50 charge/discharge cycles. The cell charge and

-12

-10

-8

-6

-4

-2

0

2

4

6

8

Cur

rent

(μΑ

)

Voltage V

Scan Rate= 50 m/s(A)

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

1.2 1.1 1.0 0.9 0.8 0.7 0.6

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

Cur

rent

(μ A

)

Voltage (V)

Scan Rate= 5 mV/s(B)

Fig. 6. CV study of P3ThAcTempEst at: (A) Scan rate = 50 mV/s, and (B) scan rate = 5 mV/s, 1 M Bu4NClO4 in acetonitrile as electrolyte, Ag/AgCl referenceelectrode.

2290 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

discharge capacities were measured to be 0.65 and0.628 mA h, respectively at the end of 50 cycles. The aver-age discharge voltage was 3.6 V. The cell charge and dis-charge time at these charge and discharge rates wereabout 2.5 and 13 h, respectively.

The same cell was also subjected to fifty-five constantcurrent discharge cycles at equal charge and discharge ratesof 0.5 mA (Fig. 11) between 3.0 and 3.8 V. The dischargecapacities were 49.9 and 45.8 at the end of second and fiftyfifth cycles (each with 99% efficiency). The discharge capac-ity decrease was only 8.2%. The discharge specific energywas 177 W h/kg (96% efficiency) at the end of second cycleand 161 W h/kg (94% efficiency) at the end of fifty fifth cy-cle. The charge and discharge at this rate both took 56 minwhich means that 1C rate is about 0.5 mA for this particularcell (where 1C is indicating the rate in mA which com-pletely charges or discharges the cell in 1 h).

Another set of sixty-five charge and discharge cycleswas performed on the same cell at equal charge and dis-charge rates of 1.0 mA (Fig. 12) between 3.0 and 3.8 V.The discharge capacities were measured to be 38.1 and34.2 at the end of second and sixty fifth cycles (each with99% efficiency). The discharge capacity decrease was only10.2%. The discharge specific energy was 132 W h/kg(98% energy efficiency) at the second cycle and 118 W h/kg (90% efficiency) at the fifty fifth cycle). The charge anddischarge at this rate both took 23 min.

The drop in the specific capacity and specific energycapacity may be attributed to structural changes occurringin the cathode material or the partial dissolution of thepolymer after extensive cycling. The specific dischargecapacity after the first cycle at 0.05 mA discharge ratewas 79.2 A h/kg which corresponds to about 87% of thetheoretical value (theoretical specific capacity was calcu-

Cur

rent

(μA

)

-0

-0

-0

-0

0

0

0

1.

.20

.15

.10

.05

.00

.05

.10

.3 1.22 1.1

Voltage

1.0

e (V))

00.9 0..8

Fig. 7. CV study of the cathode material containing 30% P3ThAcTempEst, 60% KS6L graphite, 10% PVdF. Scan rate = 5 mV/s, 1 M Bu4NClO4 in acetonitrile aselectrolyte, Ag/AgCl reference electrode.

Fig. 8. SEM pictures of (a) P3ThAcTempEst polymer, (b) cathode composite material, (c) KS6L graphite and (d) PVdF.

M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294 2291

lated to be 91 A h/kg) [10]. The difference may be due toincomplete oxidation of TEMPO nitroxide groups caused

by limited access of electrolyte solution to inner regionsof bulk of the composite cathode material.

Scheme 4. Charging and discharging processes of the cell constructed.

0 20 40 60 80

2400

2600

2800

3000

3200

3400

3600

3800

Cycle 50Cycle 30

Cycle 3

Specific Discharge Capacity (Ah/kg)

Cycle 3 Cycle 10 Cycle 20 Cycle 30 Cycle 40 Cycle 50

Vol

tage

(m

V)

Fig. 9. Voltage vs. constant current specific discharge capacity at 0.05 mA discharge rate.

0 10 20 30 40 50

200

220

240

260

280

300

320

Spe

cific

Ene

rgy

(Wh/

kg)

Cycle ID

Charge

Discharge

Fig. 10. Specific energy at 0.25 mA charge rate and 0.05 mA discharge rate.

2292 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

0 10 20 30 40 50

2900

3000

3100

3200

3300

3400

3500

3600

3700

3800

Cycle 30

Cycle 2

Vol

tage

(m

V)

Specific Capacity/ Ah/kg

Cycle 2 Cycle 10 Cycle 20 Cycle 30 Cycle 40 Cycle 50

Cycle 55

Fig. 11. Voltage vs. constant current specific discharge capacity at 0.5 mA discharge rate.

-5 0 5 10 15 20 25 30 35 402900

3000

3100

3200

3300

3400

3500

3600

3700

Cycle 1

Cycle 65

Vol

tage

(m

V)

Specific Discharge Capacity (mAh/g)

Cycle 1Cycle 20Cycle 40Cycle 65

Fig. 12. Voltage vs. constant current specific discharge capacity at 1.0 mA discharge rate.

M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294 2293

4. Conclusions

The first example of use of nitroxide bearing polythio-phene as an electroactive cathode material in a recharge-able Li battery has been demonstrated successfully. Thispolymer had 79 A h/kg initial specific charge capacity ofnearly 87% of its theoretical value, and endured chargeand discharge cycles between 2.5 and 3.8 V with onlyabout 23% loss in capacity after 50 cycles. The capacity losswas much smaller (about 8%) and the cell was more stablewhen the same cell was subjected to charge and dischargecycles between 3.0 and 3.8 V at 1C rate (about 0.5 mA)after fifty-five cycles. These results show that polythio-

phene bearing TEMPO radicals can be used as a cathodematerial against Li anode in a rechargeable battery withan output voltage of 3.6 V.

Acknowledgments

This work was financially supported by TUBITAK (TheScientific and Technological Research Council of Turkey)under the Project number 108T596 and by Fatih University(BAP 50020801-2). Graphite samples were supplied byTimcal Ltd. We would like to thank Assoc. Prof. Yusuf Yerlifor his support in ESR studies.

2294 M. Aydın et al. / European Polymer Journal 47 (2011) 2283–2294

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at doi:10.1016/j.eurpolymj.2011.09.002.

References

[1] Enrico S, Claudio N. New nanomaterials for light weight lithiumbatteries. Anal Chim Acta 2006;568:57–64.

[2] Oyama N, Tatsuma T, Sato T, Sotomura T. Dimercaptan-polyanilinecomposite electrodes for lithium batteries with high energy density.Nature 1995;374:196.

[3] Novak P, Müller K, Santhanam KSV, Haas O. Electrochemically activepolymers for rechargeable batteries. Chem Rev 1997;97:207–81.

[4] Nakahara K, Iwasa S, Satoh M, Morioka Y, Iriyama J, Suguro M, et al.Chem Phys Lett 2002;359:351.

[5] Suga T, Konishi H, Nishide H. Photocrosslinked nitroxide polymercathode-active materials for application in an organic-based paperbattery. Chem Commun 2007:1730–2.

[6] Suga T, Yoshimura K, Nishide H. Nitroxide-substituted polyether as anew material for batteries. Macromol Symp 2006;416:245–6.

[7] Suga T, Pu YJ, Oyaizu K, Nishide H. Electron-transfer kinetics ofnitroxide radicals as an electrode-active material. Bull Chem Soc Jpn2004;77:2203.

[8] Nishide H, Suga T. Organic radical battery. Electrochem Soc Interface2005;14:32.

[9] (a) MacInnes D, Druy MA, Nigrey PJ et al. Organic batteries:reversible n- and p-type electrochemical doping of polyacetylene,(CH)x. J Chem Soc Chem Commun 1981;317.;(b) Carlberg JC, Inganas O. Poly(3,4-ethylenedioxythiophene) aselectrode material in electrochemical capacitors. J Electrochem Soc1997;144(4):L61–4.

[10] Charge capacity [A h kg�1] = nF/3600; n = moles of nitroxide groupsper kilogram of material, F = Faraday constant (96,485.34 C/mol).

[11] Gallani JL, Bonomme L, Drury A, Blau WJ. Photosensitive magnetismof radicals coupled with carbon nanotubes. Organic Electronics2003;4:15–20.

[12] Takahashi Y, Hayashi N, Oyaizu K, Honda K, Nishide H. Totallyorganic polymer-based electrochromic cell using TEMPO-substituted polynorbornene as a counter electrode-active material.Polym J 2008;40(8):763–7.

[13] Suga T, Konishi H, Nishide H. Photocrosslinked nitroxide polymercathode-active materials for application in an organic-based paperbattery. Chem Commun 2007;17:1730–2.

[14] Qu J, Katsumata T, Satoh M, Wada J, Masuda T. Poly(7-oxanorbornenes) carrying 2,2,6,6-tetramethylpiperidine-1-oxy(TEMPO) radicals: synthesis and charge/discharge properties.Polymer 2009;50(2):391–6.

[15] Koshika K, Sano N, Oyaizu K, Nishide H. An ultrafast chargeablepolymer electrode based on the combination of nitroxide radical andaqueous electrolyte. Chem Commun 2009;7:836–8.

[16] Suguro M, Iwasa S, Kusachi Y, Morioka Y, Nakahara K. Cationicpolymerization of poly(vinyl ether) bearing a tempo radical: a newcathode-active material for organic radical batteries. MacromolRapid Commun 2007;28(18–19):1929–33.

[17] Suguro M, Iwasa S, Nakahara K. Fabrication of a practical andpolymer-rich organic radical polymer electrode and its ratedependence. Macromol Rapid Commun 2008;29(20):1635–9.

[18] Suga T, Yoshimura K, Nishide H. Nitroxide-substituted polyether as anew material for batteries. Macromol Symp 2006;245–246:416–22.

[19] Oyaizu K, Suga T, Yoshimura K, Nishide H. Synthesis andcharacterization of radical-bearing polyethers as an electrode-active material for organic secondary batteries. Macromolecules2008;41:6646–52.

[20] Zhang X, Li H, Li L, Lu G, Zhang S, Gu L. Polyallene with pendantnitroxyl radicals. Polymer 2008;49(16):3393–8.

[21] Qu J, Katsumata T, Satoh M, Wada J, Igarashi J, Mizoguchi K, et al.Synthesis and charge/discharge properties of polyacetylenescarrying 2,2,6,6-tetramethyl-1-piperidinoxy radicals. Chem Eur J2007;13(28):7965–73.

[22] Qu J, Fujii T, Katsumata T, Suzuki Y, Shiotsuki M, Sanda F, et al.Helical polyacetylenes carrying 2,2,6,6-tetramethyl-1-piperidinyloxy and 2,2,5,5-tetramethyl-1-pyrrolidinyloxy moieties:their synthesis, properties, and function. Polym Sci A Polym Chem2007;45(23):5431–45.

[23] Qu J, Khan FZ, Satoh M, Wada J, Hayashi H, Mizoguchi K, et al.Synthesis and charge/discharge properties of cellulose derivativescarrying free radicals. Polymer 2008;49(6):1490–6.

[24] Qu J, Morita R, Satoh M, Wada J, Terakura F, Mizoguchi K, et al.Synthesis and properties of dna complexes containing 2,2,6,6-tetramethyl-1-piperidinoxy (TEMPO) moieties as organic radicalbattery materials. Chem Eur J 2008;14(11):3250–9.

[25] Suga T, Pu Y-J, Kasatori S, Nishide H. Cathode- and anode-activepoly(nitroxylstyrene)s for rechargeable batteries: p- and n-typeredox switching via substituent effects. Macromolecules2007;40:3167–73.

[26] Qu J, Katsumata T, Satoh M, Wada J, Masuda T. Synthesis andproperties of polyacetylene and polynorbornene derivatives carrying2,2,5,5-tetramethyl-1-pyrrolidinyloxy moieties. Macromolecules2007;40:3136–44.

[27] Nesvadba P, Bugnon L, Maire P, Novak P. Synthesis of a novelspirobisnitroxide polymer and its evaluation in an organic radicalbattery. Chem Mater 2010;22:783–8.

[28] Skotheim TA, Reynolds R. Conjugated polymers processing andapplications. New York: CRC Press; 2007 [p. 9–8].

[29] Nagatomo T, Omoto O. Poly(3-methylthiophene) – a stable cathode-active material for secondary batteries. J Electrochem Soc1988;135:2124–8.

[30] Ju Y-W, Jung H-R, Lee W-J. Electrochemical characteristics of poly(3-methyl thiophene)/sulfonated-SEBS composite electrode forpolymer battery. Synth Met 2007;157:823–6.

[31] Miyasaka M, Yamazaki T, Tsuchida E, Nishide H. Magnetic andelectrical properties of poly(3-radical-substituted thiophene)s.Polyhedron 2001;20:1157–62.

[32] Opsteen JA, Van Hest JCM. Modular synthesis of block copolymersvia cycloaddition of terminal azide and alkyne functionalizedpolymers. Chem Commun 2005:57–9.

[33] Buzas A, Egnell C, Freon PCR. Hebd Séances Acad Sci (Paris)1962;255:945–7.

[34] Kim B, Chen L, Gong J, Osada Y. Titration behavior and spectraltransitions of water-soluble polythiophene carboxylic acids.Macromolecules 1999;32:3964–9.

[35] Lanzi M, Paganin L, Costa-Bizzarri P, Della-Casa C, Fraleoni A. Facilesynthesis of soluble multifunctional polyalkylthiophenes. MacromolRapid Commun 2002;23:630–3.

[36] Andersson MR, Selse D, Berggren M, Jarvinen H, Hjertberg T, InganasO, et al. Regioselective polymerization of 3-(4-octylphenyl)thiophene with FeC13. Macromolecules 1994;27:6503–6.

[37] Olinga T, Francois B. Kinetics of polymerization of thiophene byFeC13 in chloroform and acetonitrile. Synth Met 1995;69:297–8.