Ultrasonics Sonochemistry 24 (2015) 238–246

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry

journal homepage: www.elsevier .com/ locate/ul tson

High-intensity ultrasonication as a way to prepare graphene/amorphousiron oxyhydroxide hybrid electrode with high capacity in lithium battery

http://dx.doi.org/10.1016/j.ultsonch.2014.12.0011350-4177/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +34 957218637; fax: +34 957218621.E-mail address: iq2alror@uco.es (R. Alcántara).

José R. González a, Rosa Menéndez b, Ricardo Alcántara a,⇑, Francisco Nacimiento a, José L. Tirado a,Ekaterina Zhecheva c, Radostina Stoyanova c

a Universidad de Córdoba, Campus de Rabanales, Edificio C3, Laboratorio de Química Inorgánica, 14071 Córdoba, Spainb Instituto Nacional del Carbón (INCAR-CSIC), Apdo. 73, 33080 Oviedo, Spainc Institute of General and Inorganic Chemistry, Bulgarian Academy of Science, 1113 Sofia, Bulgaria

a r t i c l e i n f o

Article history:Received 29 September 2014Received in revised form 24 November 2014Accepted 2 December 2014Available online 9 December 2014

Keywords:Lithium batteriesMössbauer spectroscopyElectron paramagnetic resonanceGrapheneComposite

a b s t r a c t

The preparation of graphene/iron oxyhydroxide hybrid electrode material with very homogeneous distri-bution and close contact of graphene and amorphous iron oxyhydroxide nanoparticles has been achievedby using high-intensity ultrasonication. Due to the negative charge of the graphene surface, iron ions areattracted toward the surface of dispersed graphene, according to the zeta potential measurements. Theanchoring of the FeO(OH) particles to the graphene layers has been revealed by using mainly TEM, XPSand EPR. TEM observations show that the size of the iron oxide particles is about 4 nm. The ultrasonica-tion treatment is the key parameter to achieve small particle size in these graphene/iron oxyhydroxidehybrid materials. The electrochemical behavior of composite graphene/amorphous iron oxyhydroxideprepared by using high-intensity ultrasonication is outstanding in terms of gravimetric capacity andcycling stability, particularly when metallic foam is used as both the substrate and current collector.The XRD-amorphous character of iron oxyhydroxide in the hybrid electrode material and the smallparticle size contribute to achieve the improved electrochemical performance.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

Modern society claims for the development of batteries withmore power and energy density, longer life, and lower cost andenvironmental impact. A strategy to achieve these goals is toreplace the negative electrode based on graphite by alternativematerials.

Microstructured composites of graphite/metal oxide have exten-sively studied for the negative electrode of lithium ion batteries [1].Nanostructured composites of graphene/metal oxide [2–4] andother forms of carbon/metal oxide composites [5–7] are also beingexplored. The hybrid character (carbonaceous material/non-carbo-naceous material) of these composite electrode materials canenhance the resulting electrochemical properties. Thus, for exam-ple, the graphene phase can act like a carbonaceous network thatprovides an electronic conduction pathway, prevents the agglomer-ation of the metal oxide particles and buffers the volume changes. Inaddition, graphene can exhibit higher gravimetric capacity thangraphite to store lithium. In addition, taking into account the low

toxicity and large abundance of iron, one of the most interestinggraphene/metal oxide composites is graphene/iron oxide [8–13].In all papers cited above, the iron oxide phase was microcrystallineor nanocrystalline to X-ray diffraction (XRD). However, the use ofiron oxide particles with very small size and amorphous charactermay improve the electrochemical behavior in comparison with lar-ger and crystalline particles. Thus, the use of infrared irradiation toobtain graphene/amorphous FeO(OH) was recently reported [14].

On the other hand, previous works found that sonochemistry isa very promising method to prepare electrode materials for batter-ies [15,16], particularly when high dispersion, small size and amor-phous character of the particles are advantageous.

Having all this in mind, we decided to explore the use the sono-chemical method to prepare graphene/iron oxide electrode materi-als. In this paper we report a new and easy way to obtain graphene/amorphous iron oxyhydroxide hybrid nanostructure in which theiron phase is XRD-amorphous. In order to prepare this materialwe report the use of high intensity ultrasonication for first time.The structure and electrochemical properties of hybrid materialsare examined by means of powder XRD diffraction, scanning elec-tron microscopy (SEM) and transmission electron microscopy(TEM), particle size distribution, 57Fe Mössbauer spectroscopy,

J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246 239

electron paramagnetic resonance (EPR), thermogravimetric analy-sis (TGA), galvanostatic measurements and impedance spectros-copy. Here we reveal the interactions between the graphenelayers and the iron oxyhydroxide nanoparticles using EPR for thefirst time. These interactions are correlated with very high elec-trode capacities.

5

10

15

20

V

2. Experimental

2.1. Preparation of graphene and graphene/iron oxyhydroxide

Reduced graphene oxide (RGO) was prepared as follows. Firstly,graphene oxide was obtained from graphite and using the Brodiemethod [17]. Secondly, graphene oxide was exfoliated and reducedat 700 �C under nitrogen atmosphere.

Composite material containing reduced graphene oxide andiron oxyhydroxide (RGO/Fe) was prepared using the ultrasonica-tion method. For this purpose, 2.5 g of Fe(NO3)3�9H2O and33.3 mg of RGO were dispersed in 250 ml of ethanol and ultrasoni-cated for 1 h under high intensity radiation using a Sonics VCX750sonicator with Ti-horn. The ultrasound irradiation formed a homo-geneous suspension and promoted the anchoring of ultrafineFeO(OH) particles to graphene layers. It is known that ultrasoundirradiation can exfoliate stacked graphene layers and favors forma-tion of stable colloidal dispersions under mild conditions [18]. Theamount of iron added to the RGO suspension in ethanol (10,370%wt. as referred to the RGO mass) is in large excess as referred tothe amount of iron that is effectively incorporated in the resultingRGO/iron oxide composite. In fact, only a small fraction of theadded iron ions are trapped by RGO, as discussed below. For com-parative purposes, an RGO/Fe composite was prepared by adding asmaller amount of iron (8% wt.) to the dispersion. The samples pre-pared with smaller and larger iron amount will be henceforthreferred to as RGO/Fe1 and RGO/Fe2, respectively. The resultantsuspensions were collected by centrifugation, washed with etha-nol, recentrifugated, and dried under vacuum at 120 �C overnight.Also for comparative purposes, a sample was prepared followingthe same method although ultrasonication was not applied.

In addition, graphene-free iron oxyhydroxide was also preparedby following a similar method except for the absence of graphene.For this sample, heating at 80 �C for 2 h was used to partially evap-orate the solvent and to precipitate the solid, due to the fact that inabsence of external heating precipitation was not observed. Afterformation of the precipitate, it was separated by ultracentrifuga-tion, washed with ethanol, re-ultracentrifugated, and finally driedunder vacuum at 120 �C overnight.

In order to evaluate the interactions between dispersed graph-ene particles and iron ions, zeta potential was measured using aZPALS Brookhaven instrument and the Smoluchovski equationwas used. The same instrument was used to obtain particle sizedistribution by the dynamic light scattering method.

-2 0 2 4 6 8 10 12 14-25

-20

-15

-10

-5

0

Zeta

pot

entia

l, m

Fe wt., %

Fig. 1. Zeta potential as a function of relative amount of iron (percentage of ironmass/RGO mass).

2.2. Materials characterization

X-ray diffraction (XRD) experiments were performed in a Sie-mens D5000 instrument with CuKa radiation. Mössbauer spectrawere recorded at room temperature using a Wissel instrumentand calibration was performed by using a foil of a-Fe.

The EPR spectra were recorded as the first derivative of theabsorption signal on an ERS-220/Q spectrometer within the tem-perature range of 90–400 K. The g factors relative to a Mn2+/ZnSstandard were determined. The signal intensity was establishedby double integration of the experimental EPR spectrum. To facili-tate the analysis of the EPR spectra of the hybrid materials, EPRstandards for bulk FeOOH and a-Fe2O3 was used.

Thermogravimetric analysis (TGA) experiments were per-formed in an air atmosphere, at 5 �C/min of heating rate, using aShimadzu instrument and alumina pans.

For scanning electron microscopy (SEM) a JSM6300 instrumentwith electron probe microanalyzer (EPMA) based on an energy dis-persive spectrometer (EDS) was used. Transmission ElectronMicroscopy (TEM) micrographs were alternatively obtained byusing a JEM2010 equipped with EPMA, or a JEOL1400 instrument.

2.3. Electrochemical characterization

The active material of the working (positive) electrode wasRGO/Fe (80% wt.) mixed with binder (PVDF, 10% wt.) and carbonblack (10% wt.). The mixture was deposited on a Cu foil current col-lector supplied by Goodfellow. Alternatively, nickel (or copper)foam was used instead of copper foil. The electrodes were driedunder vacuum at 120 �C prior to the electrochemical tests, whichwere carried out in cells mounted in an Ar-filled glove box. Thecounter electrode was a lithium disk. The electrolyte was 1 M LiPF6

in EC:DEC (1:1) solvents mixture. Galvanostatic cycling experi-ments and impedance spectra were obtained with a VMP (Biologic)instrument.

3. Results and discussion

3.1. Materials characterization

The zeta-potential of RGO particles dispersed in ethanol byultrasonication was ca. �20 mV (Fig. 1). After adding 8–14% wt.iron ions (from Fe(NO3)3�9H2O) to the RGO dispersion, the zetapotential was shifted to values in the 10–15 mV range. Theseresults evidence that the negatively charged surface of the RGOnanoparticles attracts the iron(III) ions and then these ions areanchored to the surface of RGO. Thus, this procedure allows an inti-mate mixture between graphene layers and iron ions. The high-intensity ultrasound irradiation contributes to enhance the disper-sion and mixture homogeneity of RGO particles and iron ions.

The presence of crystalline phases and sample compositionwere studied by using XRD (Fig. 2) and TGA (Fig. 3). Fig. 2A showsthe XRD patterns for pristine RGO, RGO/Fe2 (prepared with andwithout ultrasonication) and RGO/Fe2 recuperated after the TGAexperiment (annealing in air atmosphere up to 600 �C). The XRDpattern of RGO (Fig. 2Aa) exhibits a broadened peak centered atca. 25�/2h that is assigned to the (002) Bragg reflection of graphitic

(300

)(2

14)(116

)

(024

)

(113

)(110

)(104

)

(b) RGO/Fe2

2θ, ° (Cu Kα)

(c) Fe2O3 from (b) after TGA at 600°C

(a) RGO

(A)

10 20 30 40 50 60 70 80

10 20 30 40 50 60 70

(b) Fe2O3 from (a) after 600°C in air

2θ, ° (Cu Kα)

(B)

(a) FeO(OH)

Fig. 2. XRD patterns for several samples. (A) (a) reduced graphene oxide (RGO), (b)composite RGO/Fe2 obtained with ultrasonication and (c) the same sample than (b)but after thermal analysis up to 600 �C in air atmosphere. In (c), the Braggreflections of hematite (Fe2O3, XRD file number 33-0664) are indexed. (B) Iron oxideobtained following the same procedure by in the absence of graphene, before (a)and after (b) annealing in air atmosphere at 600 �C.

0 100 200 300 400 500 600

-100

-80

-60

-40

-20

0

20

Mas

s lo

sss,

%

Temperature, °C

(a) RGO (b) RGO/Fe1 (c) RGO/Fe2

Fig. 3. TGA for (a) iron-free RGO, (b) RGO/Fe1 and (c) RGO/Fe2.

-4 -2 0 2 4

0.994

0.995

0.996

0.997

0.998

0.999

1.000

1.001

Rel

ativ

e in

tens

ity

Velocity, mm s-1

(B)

-4 -2 0 2 40.965

0.970

0.975

0.980

0.985

0.990

0.995

1.000

1.005

Rel

ativ

e in

tens

ity

Velocity, mm s-1

(A)

Fig. 4. 57Fe Mössbauer spectrum for (A) nanocrystalline b-FeO(OH) and (B) RGO/Fe2.

240 J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246

carbon, as it is typically found in the literature for reduced graph-ene oxide [19]. Apparently, this (002) reflections is even morebroadened and less intense for RGO/Fe2 (Fig. 2Ab) than for thepristine RGO. After ultrasonication and further exfoliation of thegraphene particles, the presence of FeO(OH) may contribute toavoid the stacking of graphene layers. The peak with very lowintensity at about 42�/2h can be ascribed to the (100) reflection.The XRD pattern of the hybrid RGO/Fe2 nanostructure does notexhibit reflections ascribable to iron compounds. On the otherhand, the amount of carbon in the composite can be determinedusing TGA (Fig. 3). According to the TGA results, the pristine RGOsample (Fig. 3a) is partially oxidized at 200 �C (ca. 8% mass gainand formation of graphene oxide), and then is oxidized (CO2 gasformation) at about 530 �C (Fig. 3a). Apparently, the presence ofiron oxyhydroxide decreases the temperature of graphene oxida-tion, as a result of the oxidation of carbon by Fe(III). The TGAresults show a total mass loss of 71% wt. for RGO/Fe1 and 30%wt. for RGO/Fe2. The oxidation of carbon to carbon dioxide takesplace at about 320–450 �C (ca. 52% mass loss for RGO/Fe-1 andca. 17% for RGO/Fe-2). The small initial mass loss up to 200 �C isascribed to solvent (ethanol) and water evaporation (2–6% massloss). The estimated iron oxide content for RGO/Fe1 and RGO/Fe2is ca. 25% and 66%, respectively. Consequently, only a relativelysmall amount of the iron ions that were initially added to the dis-persion of graphene nanosheets in ethanol become trapped in the

294 292 290 288 286 284 282 2800

20000

40000

60000

80000

Inte

nsity

Binding energy, eV

C-C / C=C

C-OO-C=O

(A)C1s

540 538 536 534 532 530 528 526

20000

30000

40000

50000

60000

70000

80000

C=O(B)

O1s

Inte

nsity

Binding energy, eV

Fe-O

C-O

Fe-O

735 730 725 720 715 710 705

40000

45000

50000

55000

60000

65000

70000Fe3+

Fe2+

(C)Fe2p

Inte

nsity

Binding energy, eV

2p1/2 2p3/2

satellite

satelliteFe3+

Fe2+

Fig. 5. XPS results and fitting for RGO/Fe2 sample in the regions (A) C1s, (B) O1s and(C) Fe3p.

Fig. 6. SEM micrograph of RGO/Fe2.

Fig. 7. TEM micrographs of RGO/Fe2.

J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246 241

resulting composite material RGO/Fe. The XRD pattern of the sam-ple recuperated after the TGA experiment (Fig. 2Ac) shows reflec-tions of hematite (hexagonal Fe2O3, XRD file number 33-0664).From the broadening of the hematite reflections and applying theScherrer equation, the resulting grain size of the hematite phasecalcined at 600 �C is Lc = 26 nm. On the other hand, the XRD ofthe sample obtained by using an equivalent experimental proce-dure but in the absence of graphene (Fig. 2Ba) shows the broad-ened XRD reflections of b-FeO(OH) (akaganeite, XRD file number34-1266), and the average crystallite size is Lc = 6.9 nm. It is worthnoting that in order to precipitate FeO(OH) in the absence of RGO itwas necessary to evaporate the ethanol solvent. After annealingakaganeite b-FeO(OH) at 600 �C in air atmosphere, hematite

Fe2O3 is formed (Fig. 2Bb). The transformation of akaganeite tohematite contributes to the initial mass loss of RGO/Fe at about200 �C in the TGA experiments.

The iron atoms were studied by using the selective Mössbauereffect. The room temperature 57Fe Mössbauer spectrum of theobtained akaganeite b-FeO(OH) is shown in Fig. 4A. According tothe literature this compound is paramagnetic at room temperatureand antiferromagnetic at lower temperatures (Neel temperature atabout 110–120 K) [20]. Two doublets have been used to fit thespectrum. For the first doublet (62.6% of relative contribution),the resulting isomer shift was d = 0.376(9) mm/s, and the quadru-pole splitting was D = 0.61(2) mm/s. For the second doublet (37.4%of relative contribution), the parameters were d = 0.38(2) mm/sand D = 1.06(4) mm/s. These results agree well with the literatureand trivalent state of iron atoms placed at two different sites of theakaganeite lattice [21,22]. According to the quadrupole splittingvalue, the second doublet corresponds to a more distorted octahe-dral site. The Mössbauer spectrum of the composite sampleRGOFe2 (Fig. 4A) agrees well with the presence of FeOOH. In addi-tion, the Mössbauer spectrum of RGOFe2 prepared without apply-ing ultrasonication showed no significant differences (not shown).

242 J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246

To further investigate the composition of the RGO/Fe2 hybridnanostructure, XPS was used (Fig. 5) and the spectra were inter-preted according to the literature [23–25]. The C1s region can bedeconvoluted into three components (Fig. 5A). The most intensepeak is placed at 285.2 eV and is ascribed to C–C and C@C. The peakat 286.8 eV is ascribed to C–O. The peak placed at 289.1 eV is due toO–C@O. In the O1s region (Fig. 5B), the spectrum was fitted withfour peaks. The peak at lower binding energy (530 eV) is ascribedto O@ ions bonded to iron ions [25]. The two peaks placed at about532.3 eV are most probably due to oxygen bonded to both carbonand iron simultaneously, indicating that FeO(OH) particles areanchored to the surface of RGO. The peak centered at higher energy(533.8 eV) is ascribed to oxygen in C–O. The XPS results in theregion of Fe2p are shown in Fig. 5C. The spin–orbit coupling splitsthe spectra in 2p1/2 and 2p3/2 regions. Both Fe(II) and Fe(III) ionsare detected. The satellites peaks indicate that Fe(II) ions are inhigh spin configuration and, consequently, the presence of Fe3O4

can be discarded [29]. Summarizing, the XPS results evidence thatthe surface of the iron nanoparticles is non-stoichiometric ironoxide, being in good agreement with the presence of XRD-amor-phous FeO(OH) [26,27]. The chemical interactions with the car-bon–oxygen functional groups in the graphene layers stabilizethe non-stoichiometric iron oxide particles and anchor them tothe graphene layer.

Fig. 8. TEM micrographs and corresp

The morphology and microstructure of the RGO/Fe particleswere examined by using SEM (Fig. 6) and TEM (Figs. 7 and 8).Fig. 6 shows flat or corrugated graphene layers while iron oxyhy-droxide particles are not observed. The uniform iron distributionwas confirmed by using EPMA coupled to SEM and TEM. TEMimages and EDS spectra (Figs. 7 and 8) show that the graphenesheets and the iron oxyhydroxide nanoparticles are homogeneouslydistributed and intimately mixed at nanoscale. The average size ofthe primary iron oxyhydroxide particles is about 4 nm. A fewagglomerates of about 80 nm in diameter were rarely observed.Fringes were not observed in the iron oxyhydroxide particles. Theseobservations are in good agreement with the XRD-amorphous char-acter of the iron oxyhydroxide nanoparticles dispersed with RGO.Iron oxide nanoparticles are firmly anchored on the graphene sur-face and embedded in the network that forms the graphene layers.The thickness of the stacked graphene layers is ca. 4 nm.

In Fig. 9, the particle size distribution reveals that the main dif-ference between the samples prepared with and without ultrason-ication is smaller sizes for the ultrasound-assisted sample. For thesample RGO/Fe2 prepared under ultrasonication (Fig. 9A), most ofthe particles have diameter smaller than 100 nm. In contrast, thesample prepared by using the same experimental procedure exceptthat ultrasonication was not applied (Fig. 9B) exhibits particlediameters in the range between 400 and 500 nm.

onding EDS spectra of RGO/Fe2.

0 100 200 300 400 500 600 700 800 900 10000

20

40

60

80

100 (B)

num

ber

particle diameter, nm

0 100 200 300 400 500 600 700 800 900 10000

20

40

60

80

100

num

ber

particle diameter, nm

(A)

Fig. 9. Particle size distribution for RGO/Fe2 sample prepared with (A) and without(B) ultrasonication.

500400300200

α-Fe2O3

RGO/Fe2

RGO/Fe1

dP /

dB, a

.u.

B, mT

RGO

FeOOH

(A)295K

100

500400300200100

α-Fe2O3

RGO/Fe2

RGO/Fe1

RGO

FeOOH

(B)78K

dP/d

B, a

.u.

B, mT

Fig. 10. EPR spectra of reduced graphene oxide (RGO), hybrids RGO/Fe1 and RGO/Fe2 registered at 295 (A) and 78 K (B). For the sake of comparison, the EPR spectra ofFeOOH and a-Fe2O3 are also given.

J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246 243

To analyze the structure of the hybrid materials, EPR spectrawere obtained. Fig. 10 compares the EPR spectra of individualRGO and FeOOH compounds with their hybrid analogs. As onecan see, individual RGO does not display any EPR signal, which isindicative of high quality nanocarbons [28–30]. On the contrary,the EPR spectrum of FeOOH consists of a broad symmetric linelocated near g = 2 at room temperature. After cooling from 295 to78 K, the resonance position is slightly moved to the lower mag-netic field. This EPR behavior is consistent with the literature datafor the magnetic structure of FeOOH [31]. For the sake of compar-ison, we also provide the EPR spectra of a-Fe2O3. a-Fe2O3 is antifer-romagnetic below the Morin temperature (TM � 260 K) andweakly ferromagnetic above TM. As a result, at T > 260 K the EPRspectrum of a-Fe2O3 consisted of a nearly symmetric broad signalcentered at g � 2.0, while on cooling the EPR profile changed dras-tically, concomitant with the loss in the signal intensity and thestrong broadening (Fig. 10).

In comparison with the individual components, the EPR spectraof RGO/Fe1 and RGO/Fe2 are different. The EPR spectrum of RGO/Fe2 is dominated by a broad asymmetric line, whose resonanceposition is strongly moved to the lower magnetic field with thelowering of the registration temperature. The strong tempera-ture-induced effect on the resonance line is an indication for theoccurrence of superparamagnetic particles in RGO/Fe2. This is con-sistent with the TEM observation on the average size of the pri-mary FeOOH particles (about 4 nm). The EPR spectrum of FGO/Fe1 displays at 78 K a narrow line with g = 2.1. The assignment ofthis line is not an easy task. However, the lack of any EPR spectra

of individual RGO allows assigning the EPR signal to isolated octa-hedrally coordinated Fe3+ ions. This is a first experimental evidencefor the appearance of anchored iron species on a graphenesubstrate.

3.2. Electrochemical behavior

Our goal here is not to study the mechanism of the reactionbetween lithium and graphene, but to compare the electrochemi-cal behavior of RGO before and after forming composite RGO/Feunder ultrasonication treatment.

In Fig. 11, the capacity values are shown as a function of cyclenumber in lithium cells. Initially, it is observed that the RGO elec-trode material has poor coulombic efficiency (37% in the first cycle)due to the irreversible lithium consumption for SEI formation, asusually found in the literature. The large specific surface area ofRGO (SBET = 660 m2/g) enhances the irreversible processes. The ini-tial reversible capacity of bare RGO (Fig. 11a) is about 700 mAh/g.However, the capacity decreases markedly during the first fewcycles, reaching a value of ca. 300 mAh/g. However, further cycling,evidences an excellent stability. These capacity values are in goodagreement with previous studies reported by other authors [2,3].Thus, the capacity of graphene-based electrode materials reported

0 50 100 150 2000

200

400

600

800

1000

1200

1400

Spec

ific

capa

city

, mA

hg-1

Cycle number

(a) RGO (b) RGO/Fe1 (c) RGO/Fe2 (d) RGO-Fe2 (no ultrasounds)

Fig. 11. Capacity as a function of cycle number in lithium cells for (a) reducedgraphene oxide, (b) reduced graphene oxide with small amount of iron andultrasonication, (c) reduced graphene oxide with large amount of iron andultrasonication and (d) reduced graphene oxide with large amount of iron and noultrasonication. Potential range: 0.01–3.0 V. Normalized current: 0.1 A/g.

0.0 0.5 1.0 1.5 2.0 2.5 3.0-2000

-1000

0

dQ/d

V

Potential, V

Cycle number123

(C)

0.0 0.5 1.0 1.5 2.0 2.5 3.0-20000

-10000

0

Cycle number123

dQ/d

V

Potential, V

(B)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

-4000

-2000

0

dQ/d

V

Potential, V

Cycle number123

(A)

Fig. 12. Derivative plots (dQ/dV vs. voltage) for the first three cycles of (A) RGO, (B)ultrasonicated RGO/Fe2 and (C) non-ultrasonicated RGO/Fe2.

244 J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246

in the literature oscillate in a wide range between 1400 and200 mAh/g [2,3]. Few-layered graphene with low specific surfacewas synthesized from graphene oxide using microwave-assistedexfoliation by Petnikova et al. and the resulting capacity in lithiumcell was about 400 mAh/g [32].

Graphene-free particles of nanocyrstalline FeO(OH) exhibit verypoor capacity retention in lithium test cell (Supplementary data,Fig. S1), in good agreement with the literature [14,33,34]. Thecapacity fade is due to the low conductivity and marked volumechanges during the charge–discharge process. Also in agreementwith the results reported in the literature, the reversible capacityof bare Fe2O3 is usually between 400 and 600 mAh/g, and capacityretention upon cycling is also poor [5].

The obtained RGO/Fe electrode materials (Fig. 12B and C) showhigher gravimetric capacity (around 400–500 mAh/g after the firstfew cycles) than bare RGO and also better coulombic efficiency(50–58% in the first cycle). The higher gravimetric capacity maybe related to the very small particle size and the amorphous char-acter of the FeO(OH) particles, and to synergistic effects betweenFeO(OH) particles and the graphene layer. The graphene nano-sheets form a matrix which may prevent the agglomeration ofthe anchored FeO(OH) nanoparticles, improving the electrical con-tact and consequently the cycling stability as compared with bareFeO(OH). The ultrasound irradiation can result in exfoliation of thegraphene sheets, smaller particle size of graphene/FeOOH and for-mation of more sites available for accommodation of lithium. Thus,the sample of RGO/Fe2 prepared using the same method but ultra-sonication was not applied (Fig. 10d) shows lower capacity values(about 400 mAh/g after 20 cycles and 363 mAh/g after 30 cycles)and poorer capacity retention.

The derivative plots (dQ/dV vs. voltage) obtained from galvano-static cycling are shown in Fig. 12. For RGO (Fig. 12A), irreversiblepeaks are observed at ca. 1.5 and ca. 0.8 V in the first discharge pro-cess. Most of the reversible capacity in the charge process takesplace in the region between 0.9 and 2.0 V which is observed likea broadened and low-intensity region. In all the derivative plots(Fig. 12), an intense peak at about 0.8–1.0 V is observed in the firstdischarge process of all samples, being more intense for the ultr-asonicated RGO/Fe2 sample. This peak is due to the SEI formationand the conversion reaction of iron oxyhydroxide. The conversionreaction of FeO(OH) can be written as follows [33]:

FeOðOHÞ þ 3Liþ þ 3e� ! Feþ LiOHþ Li2O ð1Þ

Thus, for FeO(OH) and Fe2O3 the theoretical capacity values are908 and 1007 mAh/g, respectively. However, this capacity can beexceeded by SEI formation and non-faradic processes (charge sep-aration at the solid electrode/liquid electrolyte interface) [35–37].The small particle size of FeO(OH) and the surface of exfoliatedRGO can enhance the intensity of the peak at about 0.8–1.0 V forRGO/Fe2. In the second and third discharge process, there is stilla low-intensity peak at about 0.9 V for RGO/Fe samples but it isnot observed for bare RGO. Consequently, the reduction of ironoxide is reversible. The reversible oxidation process can be writtenlike follows [33]:

2Feþ 3Li2O ¼ Fe2O3 þ 6Liþ þ 6e� ð2Þ

0 200 400 600 800 1000 1200 1400 1600 18000.0

0.5

1.0

1.5

2.0

2.5

3.0Vo

ltage

,V

Capacity,mAhg-1

cycle 1 cycle 2 cycle 3

(A)

0 5 10 15 20 250

200

400

600

800

1000

1200

1400

1600

1800

Spec

ific

capa

city

, mA

hg-1

Cycle Number

(B)

Fig. 13. Electrochemical behavior in lithium cell of RGO/Fe2 deposited on nickel-foam current collector. (A) Voltage–capacity plot. (B) Capacity plotted as a functionof cycle number. Potential range: 0.01–3.0 V. Normalized current: 0.1 A/g.

0 10 20 30 40 50 60 70 80 90 1000

200

400

600

800

1000

1200

1400

1600

1800

Spec

ific

capa

city

, mA

h g-1

Cycle number

0.1 A/g 0.2 A/g 0.4 A/g 0.8 A/g 1 A/g

relaxation period for 12 h

Fig. 14. Capacity vs. cycle number in lithium cell of RGO/Fe2 deposited on copper-foam current collector and using variable current density (0.1–1 A/g). Potentialrange: 0.01–3.0 V.

J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246 245

During the charge process, narrow peaks are not observed in thederivative curves, being indicative of the amorphous character ofthe electrode material.

The ac impedance spectra (Nyquist plot) of RGO/Fe2 (Supple-mentary data, Fig. S2) show a depressed semicircle at high andintermediate frequencies (surface layer and charge transfer) anda sloped line at low frequencies (Warburg-type processes). The sizeof the semicircle is small, in good agreement with low charge

transfer resistance and good electrochemical behavior. Our resultsare in excellent agreement with the results reported by Sun et al.which found better electrical conductivity for the composite incomparison with bare FeOOH [14].

To further improve the electrochemical performance, the activeelectrode material was deposited on metal-foam (Figs. 13 and 14).In we compare Figs. 11 and 13, it is observed that the resultingcapacity values in lithium cells are higher and capacity retentionupon cycling is improved after using nickel-foam. Thus, the disper-sion of the active material in the 3D current collector can be usedto improve the electrochemistry, particularly for microbatteries.Besides, the good response to variable kinetics of RGO/Fe2 withcopper-foam is revealed in Fig. 14, and after 94 cycles, the specificcapacity is 562 mAh/g (higher than graphite).

4. Conclusions

We report a method to prepare graphene/amorphous FeOOHnanostructured composite using high intensity ultrasonication.The anchoring of the FeO(OH) particles to the graphene layershas been revealed by using mainly TEM, XPS and EPR. Accordingto the EPR results, for RGO/Fe1 (25% mass of iron oxide), isolatedoctahedrally coordinated Fe3+ species anchored on the graphenesubstrate are observed, while FeOOH nanoparticles contributemainly to RGO/Fe2 (26% mass of iron oxide). The resulting elec-trode material exhibits a capacity significantly higher than graph-ite. The improvement of the electrochemical behavior can beattributed to the amorphous character of the iron oxide, the verysmall particle size and the excellent mutual dispersion of the twophases which are achieved by the ultrasonication treatment. Inaddition, by using nickel-foam as substrate and current collectorof the graphene/amorphous iron oxide composites, the reversiblecapacity reached 1200 mAh/g.

Acknowledgments

The authors are indebted to MEC (MAT2011-22753) and Juntade Andalucía (FQM288), and they appreciate the collaboration ofJ.I. Corredor (SCAI, UCO) to obtain the XPS results.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.ultsonch.2014.12.001.

References

[1] R. Alcántara, P. Lavela, G.F. Ortiz, J.L. Tirado, R. Stoyanova, E. Zhecheva, J.M.Jiménez-Mateos, Modification of petroleum coke for lithium-ion batteries byheat-treatment with iron oxide, J. Electrochem. Soc. 151 (2004) A2113–A2119.

[2] S. Kuo, W. Liu, C. Kuo, N. Wu, H. Wu, Lithium storage in reduced grapheneoxides, J. Power Sources 244 (2013) 552–556.

[3] D.A.C. Brownson, D.K. Kampouris, C.E. Banks, An overview of graphene inenergy production and storage applications, J. Power Sources 196 (2011)4873–4885.

[4] J. Ye, J. Zhang, F. Wang, Q. Su, G. Du, One-pot synthesis of Fe2O3/graphene andits lithium-storage performance, Electrochim. Acta 113 (2013) 212–217.

[5] N.K. Chaudhari, M.S. Kim, T.S. Bae, J.S. Yu, Hematite (a-Fe2O3) nanoparticles onvulcan carbon as an ultrahigh capacity anode material in lithium ion battery,Electrochim. Acta 114 (2013) 60–67.

[6] Z. Zeng, H. Zhao, J. Wang, P. Lva, T. Zhang, Q. Xia, Nanostructured Fe3O4@C asanode material for lithium-ion batteries, J. Power Sources 248 (2014) 15–21.

[7] C. Lei, F. Han, Q. Sun, W.C. Li, A.H. Lu, Confined nanospace pyrolysis for thefabrication of coaxial Fe3O4@ C hollow particles with a penetratedmesochannel as a superior anode for Li-ion batteries, Chem. Eur. J. 20 (2014)139–145.

[8] G. Zhou, D.-W. Wang, F. Li, L. Zhang, N. Li, Z.-S. Wu, L. Wen, G.Q. Lu, H.-M.Cheng, Graphene-wrapped Fe3O4 anode material with improved reversiblecapacity and cyclic stability for lithium ion batteries, Chem. Mater. 22 (2010)5306–5313.

246 J.R. González et al. / Ultrasonics Sonochemistry 24 (2015) 238–246

[9] J. Kan, Y. Wang, Large and fast reversible Li-ion storages in Fe2O3–graphenesheet-on-sheet sandwich-like nanocomposites, Sci. Rep. 3 (2013) 3502–3512.

[10] J. Luo, J. Liu, Z. Zeng, C. Fan Ng, L. Ma, H. Zhang, J. Lin, Z. Shen, H.J. Fan, Three-dimensional graphene foam supported Fe3O4 lithium battery anodes with longcycle life and high rate capability, Nano Lett. 13 (2013) 6136–6143.

[11] D. Wei, J. Liang, Y. Zhu, J. Zhang, X. Li, K. Zhang, Z. Yuan, Y. Qian, Facileformation of graphene-encapsulated a-Fe2O3 nanorice as enhanced anodematerials for lithium storage, Electrochim. Acta 114 (2013) 779–784.

[12] G. Chen, R. Rodriguez, L. Fei, Y. Xua, S. Deng, S. Smirnov, H. Luo, A facilehydrothermal route to iron(III) oxide with conductive additives as compositeanode for lithium ion batteries, J. Power Sources 259 (2014) 227–232.

[13] X. Xia, W. Lei, Q. Hao, W. Wang, Y. Sun, X. Wang, One-pot synthesis andelectrochemical properties of nitrogen-doped graphene decorated with M(OH)x

(M = FeO, Ni, Co) nanoparticles, Electrochim. Acta 113 (2013) 117–126.[14] Y. Sun, X. Hu, W. Luo, H. Xu, C. Hu, Y. Huang, Synthesis of amorphous FeOOH/

reduced graphene oxide composite by infrared irradiation and its superiorlithium storage performance, ACS Appl. Mater. Interfaces 5 (2013) 10145–10150.

[15] F. Nacimiento, R. Alcántara, U.G. Nwokeke, J.R. González, J.L. Tirado,Nanocrystalline CoSn2-carbon composite electrode prepared by usingsonochemistry, Ultrason. Sonochem. 19 (2012) 352–357.

[16] R. Abu-Much, A. Gedanken, Sonochemical synthesis under a magnetic field:fabrication of nickel and cobalt particles and variation of their physicalproperties, Chem. Eur. J. 14 (2008) 10115–10122.

[17] C. Botas, P. Álvarez, P. Blanco, M. Granda, C. Blanco, R. Santamaría, L.J.Romasanta, R. Verdejo, M.A. López-Manchado, R. Menéndez, Graphenematerials with different structures prepared from the same graphite by theHummers and Brodie methods, Carbon 65 (2013) 156–164.

[18] G. Cravotto, P. Cintas, Sonication-assisted fabrication and post-syntheticmodifications of graphene-like materials, Chem. Eur. J. 16 (2010) 5246–5259.

[19] A. Ramadoss, S.J. Kim, Improved activity of a graphene–TiO2 hybrid electrodein an electrochemical supercapacitor, Carbon 63 (2013) 434–445.

[20] R.J. Pollard, C.M. Cardille, D.G. Lewis, L.J. Brown, Characterization of FeOOHpolymorphs and ferrihydrite using low-temperature, applied-field, Mössbauerspectroscopy, Clay Miner. 27 (1992) 57–71.

[21] M. Zic, M. Ristic, S. Music, Microstructural changes in particles detected duringthe transformation from b-FeOOH to a-Fe2O3 in dense aqueous suspensions, J.Alloys Compd. 464 (2008) 81–88.

[22] Y.X. Yang, M.L. Liu, H. Zhu, Y.R. Chen, G.J. Mu, X.N. Liu, Y.Q. Jia, Preparation,characterization, magnetic property, and Mössbauer spectra of the b-FeOOHnanoparticles modified by nonionic surfactant, J. Magn. Magn. Mater. 320(2008) L132–L136.

[23] D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R.D. Piner, S.Stankovich, I. Jung, D.A. Field, C.A. Ventrice Jr., R.S. Ruoff, Chemical analysis

of graphene oxide films after heat and chemical treatments by X-rayphotoelectron and Micro-Raman spectroscopy, Carbon 47 (2009) 145–152.

[24] A. Abouimrane, O.C. Compton, K. Amine, S.T. Nguyen, Non-annealed graphenepaper as a binder-free anode for lithium-ion batteries, J. Phys. Chem. C 114(2010) 12800–12804.

[25] A.P. Grosvenor, B.A. Kobe, M.C. Biesinger, N.S. McIntyre, Investigation ofmultiplet splitting of Fe 2p XPS spectra and bonding in iron compounds, Surf.Interface Anal. 36 (2004) 1564–1574.

[26] T. Fujii, F.M.F. de Groot, G.A. Sawatzky, F.C. Voogt, T. Hibma, K. Okada, In situXPS analysis of various iron oxide films grown by NO2-assisted molecular-beam epitaxy, Phys. Rev. B 59 (1999) 3195–3202.

[27] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, Handbook of X-rayPhotoelectron Spectroscopy, Perkin-Elmer, Eden Prairie, MN, 1992.

[28] C.V. Pham, M. Krueger, M. Eck, S. Weber, E. Erdem, Comparative electronparamagnetic resonance investigation of reduced graphene oxide and carbonnanotubes with different chemical functionalities for quantum dotattachment, Appl. Phys. Lett. 104 (2014) 132102, http://dx.doi.org/10.1063/1.4870297.

[29] R.R. Nair, M. Sepioni, I.-L. Tsai, O. Lehtinen, J. Keinonen, A.V. Krasheninnikov, T.Thomson, A.K. Geim, I.V. Grigorieva, Spin-half paramagnetism in grapheneinduced by point defects, Nat. Phys. 8 (2012) 199–202.

[30] M. Sepioni, R.R. Nair, S. Rablen, J. Narayanan, F. Tuna, R. Winpenny, A.K. Geim,I.V. Grigorieva, Limits on intrinsic magnetism in graphene, Phys. Rev. Lett. 105(2010) 207205.

[31] K.E. Garcia, A.L. Morales, C.A. Barrero, C.E. Arroyave, J.M. Greneche, Magneticand crystal structure refinement in akaganeite nanoparticle, Physica B 354(2004) 187–190.

[32] S. Petnikota, N.K. Rotte, V.V.S.S. Srikanth, B.S.R. Kota, M.V. Reddy, K.P. Loh,B.V.R. Chowdari, Electrochemical studies of few-layered graphene as an anodematerial for Li ion batteries, J. Solid State Electrochem. 18 (2014) 941–949.

[33] X.M. Lou, X.Z. Wu, Y.X. Zhang, Goethite nanorods as anode electrode materialsfor rechargeable Li-ion batteries, Electrochem. Commun. 11 (2009) 1696–1699.

[34] T. Tabuchi, Y. Katayama, T. Nukuda, Z. Ogumi, Surface reaction of b-FeOOH filmnegative electrode for lithium-ion cells, J. Power Sources 191 (2009) 636–639.

[35] M. López, G.F. Ortiz, P. Lavela, R. Alcántara, J.L. Tirado, Improved energy storagesolution based on hybrid oxide materials, ACS Sustainable Chem. Eng. 1 (2013)46–56.

[36] M.V. Reddy, T. Yu, C.H. Sow, Z.X. Shen, C.T. Lim, G.V.S. Rao, B.V.R. Chowdari,Alpha-Fe2O3 nanoflakes as an anode material for Li-ion batteries, Adv. Funct.Mater. 17 (2007) 2792–2799.

[37] J. Jamnik, J. Maier, Nanocrystallinity effects in lithium battery materials –aspects of nano-ionics. Part IV, Phys. Chem. Chem. Phys. 5 (2003) 5215–5220.