Studies on physicochemical properties of pure and iron substituted chromium niobates, Cr1−xFexNbO4 (x = 0, 0.2, 0.4, 0.6)

Studies on physicochemical properties of pure and iron substituted chromium niobates, Cr1−xFexNbO4 (x = 0, 0.2, 0.4, 0.6)

Materials Science and Engineering B 217 (2017) 63–73 Contents lists available at ScienceDirect Materials Science and Engineering B journal homepage:...

2MB Sizes 0 Downloads 1 Views

Materials Science and Engineering B 217 (2017) 63–73

Contents lists available at ScienceDirect

Materials Science and Engineering B journal homepage: www.elsevier.com/locate/mseb

Studies on physicochemical properties of pure and iron substituted chromium niobates, Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) A. Sree Rama Murthy a, K.I. Gnanasekar a, R. Govindaraj b, V. Jayaraman a,⇑, A.M. Umarji c a

Materials Chemistry and Metal Fuel Cycle Group, Indira Gandhi Centre for Atomic Research, Kalpakkam 603102, India Materials Science Group, Indira Gandhi Centre for Atomic Research, Kalpakkam 603102, India c Materials Research Centre, Indian Institute of Science, Bangalore 560012, India b

a r t i c l e

i n f o

Article history: Received 20 September 2016 Received in revised form 28 November 2016 Accepted 12 December 2016 Available online 25 January 2017 Keywords: Chromium niobate Mössbauer spectroscopy X-ray photoelectron spectroscopy Impedance

a b s t r a c t Pristine and iron substituted chromium niobates (Cr1xFexNbO4 with x = 0, 0.2, 0.4 and 0.6) are prepared by solid-state synthesis and phase characterised by X-ray diffraction. Microstructure is determined using scanning electron microscope and micro-chemical analysis is performed by energy dispersive X-ray analysis (EDX). The current-voltage characteristics are studied in the temperature range of 423–723 K. The electrical conductivity of sintered pellets is measured by impedance spectroscopy. Temperature dependent magnetization studies are performed using vibrating sample magnetometer (VSM) and room temperature Bohr magneton number is calculated from the magnetic susceptibility data. The conductivity is passed through a minimum at x = 0.2 in Cr1xFexNbO4 for x = 0–0.6. X-ray photoelectron spectroscopic studies revealed the surface non-stoichiometry in these compositions. Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction Chromium niobate (CrNbO4) is a rutile structured compound with random distribution of chromium and niobium ions [1,2]. Rutile structure is an open framework of hexagonally close packed oxygen ions with octahedrally coordinated metal ions. The metal ions are edge-shared infinitely in [0 0 1] direction of the tetragonal unit cell (P42/mnm) (Fig. 1). The variations in metal-metal distances and overlap of d orbitals and M-O p⁄ interactions, resulted the differences in electrical properties of metal oxides with rutile structure [3]. Ternary rutile phases of Cr0.5xNb0.5+xO2 (x = 0.5, 0, 0.1, 0.17, 0.19, 0.21) were reported by Marinder et al., showing the variations in ratio of lattice parameters (c/a) and unit cell volumes [4]. Different groups studied further structural and electrical properties of CrNbO4 [5–7]. Balamurugan et al. studied the ammonia, ethanol and LPG sensing characteristics of CrNbO4 [8]. Earlier Khazai et al., studied magnetic and electronic properties of Fe1xCrxNbO4 with x = 0.0, 0.1 for their photo-response behaviour [9]. Gas sensing characteristics of CrNbO4 and FeNbO4 were explored in author’s laboratory [10–12]. Recently, Meilin et al., studied the photocatalytic activity of CrNbO4 for hydrogen production [13]. In radioactive spent fuel storage facilities of nuclear industry, radiolysis results in production of hydrogen. Further, monitoring of

⇑ Corresponding author. E-mail address: [email protected] (V. Jayaraman). http://dx.doi.org/10.1016/j.mseb.2016.12.001 0921-5107/Ó 2016 Elsevier B.V. All rights reserved.

steam leak into argon cover gas of sodium cooled fast breeder reactors at its inception is necessary to avoid any obnoxious effects. In the development of wide dynamic range hydrogen sensing material, we try to understand the systematic variation in the properties of CrNbO4 on substitution by different iron concentrations, with a view of employ these compositions (Cr1xFexNbO4) in analyte specific sensing applications in nuclear industries. In this context, we studied the electrical, magnetic, and X-ray photoelectron spectroscopic properties of Cr1xFexNbO4 compositions with x = 0, 0.2, 0.4, and 0.6 in this current work.

2. Materials and methods 2.1. Preparation The powdered samples of Cr1xFexNbO4 (x = 0, 0.2, 0.4, and 0.6) were prepared by high temperature solid-state reaction method. Stoichiometric amounts of Cr2O3, Fe2O3, and Nb2O5 (>99.9% pure, M/s Alfa Aesar) were mixed thoroughly using agate mortar and pestle for about 2 h. The resultant mixtures were compacted to 16 mm diameter and 3 mm thick pellets by uniaxial pressing to about 20 MPa pressure. The pellets were heated to a temperature of 1373 K for 24 h in air at a heating rate of 5 K min1 in a muffle furnace (M/s Carbolite, UK) with intermittent grindings and pelletization.

64

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

(i)

Fig. 3. Changes in (a) ratio of lattice parameters (c/a) and (b) unit cell volume with iron concentration of Cr1xFexNbO4 (x = 0–0.6) powders.

(ii)

Fig. 1. (a) Unit cell and (b) extended structure of CrNbO4.

(X’Pert Pro, Netherlands) with Cu Ka radiation. The lattice parameters were estimated using linear least squares fit algorithm. The powders were pressed into 8 mm pellets, the morphological features were recorded by scanning electron microscope (M/s Philips, Model # XL 30) and a semi-quantitative analysis of the elemental composition was performed by energy dispersive X-ray analysis (EDX).

2.2. Phase characterization

(ii)

(a)

(110)

(a)

Intensity (a.u)

(110)

The heat-treated samples were finely ground and powder X-ray diffraction data was recorded on M/s Pan Analytical diffractometer

(b)

(b)

(c)

(c)

(d)(d) 28.0

(202)

(101)

(301), (112)

(220)

(111)

(a) (a)

Intensity (a.u)

(210)

(211)

(101)

(i)

(310)

27.5

2θ (degrees)

(002)

27.0

(200)

(110)

26.5

(b) (b)

(c) (c)

(d)(e)

20

30

40

50

60

70

80

2θ (degrees) Fig. 2. (i) Powder XRD patterns of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4, (ii) Zoom-in view of (1 1 0) peak indicating the change in 2h value with varying iron concentration.

65

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

Approximately 300 mg of a given composition was mixed with 500 lL of 2 wt% poly ethylene glycol (PEG) in water and compressed to 8 mm diameter (d) and 2 mm thick (t) pellets. The PEG was removed by heating to 773 K for 5 h at a heating rate of 2 K min1, and the pellets were sintered at 1273 K for 10 h at a heating rate of 5 K min1. The flat surfaces of these pellets were gold coated by DC magnetron sputtering unit (M/s Hind High Vacuum, India) and were loaded into a home-built glass chamber (Fig. SF1 of supplementary section), sandwiched between gold electrodes for electrical conductance measurements. This conductivity cell was housed inside a homemade nichrome furnace and the temperature of the sample was regulated by PID controller within ±1 K. The current–voltage characteristics were studied in air ambience using precision source/measure unit of M/s Keysight technologies, # B2912A. The temperature of the sample was varied from 423 to 723 K and at each temperature, a voltage cycle from 20 V to +20 V and then from +20 to 20 V was performed with a step increment of 323 mV (corresponding to 250 data points). Further, impedance was measured in the frequency range of 100 Hz to 1 MHz in air using a frequency response analyzer (model SI 1255 of Solatron, M/s Schlumberger, UK) coupled with an electrochemical interface (model 1286 of Solatron, M/s Schlumberger, UK).

2.4. Magnetic studies

Table 1 Average grain size and semi-quantitative analysis of Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) by SEM and EDX studies. Composition

CrNbO4 Cr0.8Fe0.2NbO4 Cr0.6Fe0.4NbO4 Cr0.4Fe0.6NbO4

Average grain size ±200 (nm)

466 541 577 893

Atomic ratio of Cr to Fe Theoretical

Experimental

– 4 1.5 0.67

– 4.2 ± 1.03 1.62 ± 0.23 0.53 ± 0.10

0.03 (a) 0.02

Current (A)

2.3. Electrical conductance characteristics

0.01

(d) (c)

0.00

(b)

-0.01 -0.02 -0.03 -20

Zero field cooled (ZFC) and field cooled (FC) magnetization measurements were made with a vibrating sample magnetometer (Cryogenic Inc, UK) operating at 20.4 Hz. Sintered polycrystalline

-15

-10

-5

0

5

10

15

20

Voltage (V) Fig. 5. Typical current-voltage characteristics of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4 at 623 K.

Fig. 4. SEM images of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4.

66

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

pressed pellets were shaped as rectangular bars of approximate dimensions 5 mm  3.5 mm  2.5 mm and the field was applied along the longest dimension to reduce demagnetization effects. Temperature and magnetic fields were considered accurate to 0.25% and 0.5% respectively as provided by the manufacturer. Magnetic susceptibility measurements were carried out from liquid helium (4 K) to room temperature (298 K) at a field of 500 Oe. Wissel based Mössbauer spectrometer was operated in constant acceleration mode and in transmission geometry. 57Co dispersed in Rh matrix with a specific activity of 50 mCi was used as a source for measurements. Each Mössbauer spectrum has been acquired in 1024 channels and fitted to Lorentzian line shapes of line width Ci using a non-linear least squares program to obtain hyperfine parameters such as isomer shift di, quadrupole splitting (2ei) and magnetic hyperfine fields (Bihf) experienced by relative fractions fi of distinct 57Fe absorber atoms. The velocity calibration was carried out using 10 lm thick iron foil at 300 K. The values of isomer shifts presented in this study are given with respect to that of a-Fe absorber. Along with the iron substituted chromium niobate samples, FeNbO4 spectrum was recorded for comparison.

The increase in lattice volume may be due to about 4% higher ionic radius of VI coordinated high spin Fe3+ (64.5 pm) over that of Cr3+ (61.5 pm) (ionic radius of low spin Fe3+ is 55 pm) [17]. The details of finding the spin of iron in these compositions are discussed in Section 3.3. The SEM images of Cr1xFexNbO4 (x = 0, 0.2, 0.4 and 0.6) are shown in Fig. 4(a) to (d). The corresponding EDX spectra are presented in Fig. SF2 of supplementary section. The average grain size and the atomic ratio of chromium to iron in these compositions analysed at different regions of the samples are presented in Table 1. The average grain size of the samples was found to be increasing with iron content and varies between ±200 nm about their respective mean value. The grain growth may be because of the movement of the grain boundaries to minimize the interfacial energies in higher amount of iron containing compounds. The semi-quantitative measurements of atomic ratios were in good agreement with their theoretical values.

2.5. X-ray photoelectron spectroscopic studies

3.2.1. Current–voltage (I-V) characteristics Fig. 5 shows a typical current response to the varying applied voltage across Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) pellets of unit thickness at 623 K. Linear I-V profiles were obtained for all the compositions. The hysteresis was observed during voltage cycling,

I1 ðEi Þ N1 k1 r1 a I2 ðEi Þ N2 k2 r2

4

10

(a)

(i)

(b) Resistance (Ω)

Surface characterization of these samples was carried out by X-ray photoelectron spectroscopy (M/s SPECS, Germany). Sample pellets were analysed under a vacuum level of 1010 mbar. Monochromatic Al Ka (1486.6 eV) radiation was used for excitation and hemi-spherical Phoibos 150 mm analyzer connected to 1D – delay line detector was operated with 50 eV pass energy in fixed analyzer transmission mode to record the kinetic energies of the photo electrons. The spectrometer work function was adjusted to binding energy of Au 4f7/2 electron, centred at 84.0 eV. Charge correction for specimen was made by operating electron gun at 1.0 eV with an emission current of 2.5 lA. Survey scans were recorded between 1000 and 4 eV range with a step size of one eV, whereas selected regions were recorded with an energy step of 0.05 eV. Shirley background subtraction procedure was adopted for processing all the recorded spectra [14] using CASAXPS software (Version 2.3.16 Pre-rel 1.4). The backgroundsubtracted spectra were fitted to Voigt Amplitude peak shape using Peak FitÒ software (Version 4). Eq. (1) gives quantification of different elements in an XPS sample from peak intensities.

3.2. Electrical characterization

(c) (d)

3

10

(e) (f) (g) 2

ð1Þ

10

-20

-15

-10

-5

0

5

10

15

20

Voltage (V)

where N1, N2 are concentrations, k1, k2 are the attenuation lengths, and r1, r2 are the photoionization cross sections of elements 1 and 2 respectively [15,16].

6

10

(ii)

3. Results and discussion 3.1. Powder X-ray diffraction and morphological studies The powder XRD patterns of Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) are shown in Fig. 2(i). All the compositions form tetragonal phases with P42/mnm space group (JCPDS card # 34-0366) which is in agreement with Khazai et al. [9]. With increasing iron content, diffraction pattern shifted towards the lesser 2h value side. The systematic decrease in 2h value of (1 1 0) plane is shown in Fig. 2(ii), which is indicative of increase in inter-planar spacing upon iron substitution. Lattice parameters were estimated using a linear least square fit algorithm. The ratio of lattice parameters c=a and volume of unit cell (Vunit cell) were found to be increasing with increasing iron concentration (Fig. 3(a) and (b)).

Resistance (Ω)

(a) (b)

5

10

(c) (d)

4

10

(e) (f) (g)

3

10

-20

-15

-10

-5

0 5 Voltage (V)

10

15

20

Fig. 6. Typical resistance-voltage characteristics of (i) CrNbO4 and (ii) Cr0.8Fe0.2NbO4, at (a) 423 K, (b) 473 K, (c) 523 K, (d) 573 K, (e) 623 K, (f) 673 K and (g) 723 K.

67

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

-12k

-100k

(i) (a) Z" (Ω)

Z" (Ω)

(ii)

-80k

-9k

-6k (b)

(a)

-60k -40k

-3k

(b)

(d)

-20k

(c)

(c) (d)

0 0

0

4k

8k

12k 16k Z' (Ω)

20k

0

24k

80k

120k

160k

200k

Z' (Ω)

-30k

-4k

(iii)

(iv) -3k

(a)

-20k

(a) -2k

Z" (Ω)

Z" (Ω)

40k

-10k (b)

-1k

(c) 0

(b) (d)

(c)

(d) 0

15k

30k

45k Z' (Ω)

60k

75k

2k

4k

6k

8k

Z' (Ω)

Fig. 7. Typical Nyquist plots of (i) CrNbO4 and (ii) Cr0.8Fe0.2NbO4, (iii) Cr0.6Fe0.4NbO4 and (iv) Cr0.4Fe0.6NbO4 at temperatures (a) 423 K, (b) 473 K, (c) 523 K, and (d) 573 K.

CPE1 R0

CPE1

CPE2

R1

R2

R0

R1

EC 1

EC 2

Fig. 8. Equivalent circuits used for fitting complex impedance data; EC1 for single depressed semicircle and EC2 for two overlapping depressed semicircles.

Table 2 Activation energies for electrical conduction derived from Arrhenius plot for Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) pellets.

-2.5

-1

log (σ (S cm ))

-3.0 -3.5 -4.0 -4.5

(a)

-5.0

(d) (c)

(b)

-5.5 1.4

1.6

1.8

2.0

2.2

2.4

-1

1000/T (K ) Fig. 9. Arrhenius behaviour of electrical conductivity of (a) CrNbO4 and (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4 samples in the temperature range 423–723 K.

Composition

Eact ± 0.002 (eV)

Range of conductivity (Scm1) in the temperature interval of 423–648 K

CrNbO4 Cr0.8Fe0.2NbO4 Cr0.6Fe0.4NbO4 Cr0.4Fe0.6NbO4

0.159 0.187 0.183 0.194

104.68 105.45 104.97 104.87

to to to to

103.06 103.52 103.11 102.88

probably due to thermal heating of the sample. Among the compositions, CrNbO4 exhibited higher conductance and Cr0.8Fe0.2NbO4 possessed the lowest for a given voltage. Figs. SF3 (i) to SF3 (iv) of supplementary section show the current response for varying voltage across Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) pellets in the temperature range of 423–723 K at an interval of 25 K. All the compositions at all the temperatures of study resemble ohmic behaviour. With increase in temperature, increase in conductance was observed in the entire applied voltage range. To confirm this linear relation between current and voltage, resistance was plotted against varying voltage. The R-V characteristics

68

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

of CrNbO4, Cr0.8Fe0.2NbO4 are presented in Fig. 6(i) and (ii), and for the rest of the two compositions are presented in Fig. SF4 ((i) and (ii)) of supplementary section. These profiles show the non-linear relation between current and voltage in the temperature range of 423–723 K.

All the compositions Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) exhibited non-linear resistance-voltage relationship in the entire temperature range of studies. Further, CrNbO4 exhibited high conductance, and with iron substitution, the conductance dropped initially and then increased from Cr0.6Fe0.4NbO4. This is possibly due to the

40k

20µ

1/χm(g Oe/emu)

χm(emu/(g Oe))

30k θ = ~ -33.9 K

200µ 20k 100µ

10k

0

dχ m/dT(emu/(g Oe K))

(a1)

TN

300µ

50

100 150 200 Temperature (K)

250

TN = ~ 8.2 K 0 -10µ -20µ -30µ

0 0

(a2)

10µ

4

300

6

8

10 12 14 Temperature (K)

16

18

20

60k

200µ

40k 150µ 100µ

20k

50µ 0 100 150 200 Temperature (K)

TN = ~ 6.9 K 0

-10µ

4

300

(c1)

TN

200µ

250

30k

160µ

80µ

10k

1/χm(g Oe/emu)

20k

120µ

40µ

dχm/dT (emu/(g Oe K))

50

50

100 150 200 Temperature (K)

250

TN= ~ 6.8 K

-10µ

0 300

dχm/dT (emu/(g Oe K))

χm(emu/(g Oe))

20µ

1/χm(g Oe/emu)

20k

250

8

12 16 20 Temperature (K)

24

28

(d2)

40µ

100 150 200 Temperature (K)

28



40k

50

24

(c2)

4

(d1)

60µ

0

12 16 20 Temperature (K)

0

300

60k TN

80µ

8

10µ

0 0

(b2)

10µ

-20µ

0 0

χm(emu/(g Oe))

dχm/dT (emu/(g Oe K))

(b1)

TN

1/χm(g Oe/emu)

χm(emu/(g Oe))

250µ

2µ TN = ~ 15.4 K 1µ 0 -1µ 10 12 14 16 18 20 22 24 26 28 30 Temperature (K)

Fig. 10. Zero field cooled magnetization curves showing the variation in mass susceptibility and inverse mass susceptibility with temperature for (a1) CrNbO4, (b1) Cr0.8Fe0.2NbO4, (c1) Cr0.6Fe0.4NbO4 and (d1) Cr0.4Fe0.6NbO4 and Neel temperatures were estimated from the first derivative magnetization curves of (a2) CrNbO4, (b2) Cr0.8Fe0.2NbO4, (c2) Cr0.6Fe0.4NbO4 and (d2) Cr0.4Fe0.6NbO4.

69

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

Table 3 Neel temperature and h derived from field cooled magnetization curves; theoretically calculated and experimentally obtained magnetic moment per formula unit of Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6, 1) in Bohr magnetons. (LS: low spin, HS: high spin, f.u: formula unit.) Composition

TN (K) (approx.)

lcalc eff (HS) (lB per f.u)

lcalc eff (LS) (lB per f.u)

lexp (lB per f.u) ±0.05 eff

CrNbO4 Cr0.8Fe0.2NbO4 Cr0.6Fe0.4NbO4 Cr0.4Fe0.6NbO4 FeNbO4

8.2 6.9 6.8 15.4 48 and 250

3.87 4.28 4.69 5.10 5.91

3.87 3.44 3.02 2.59 1.73

3.80 3.13 4.05 3.40 5.81

Table 4 Results of Mössbauer spectral analysis of Cr1xFexNbO4 (x = 0.2, 0.4, 0.6 and 1) showing the variations in line widths, isomer shift, quadrupole splitting and relative area. Line width (mm/s)

Isomer shift (mm/s)

Quad splitting (mm/s)

Relative area

FeNbO4 Cr0.8Fe0.2NbO4

0.74 ± 0.02 0.47 + 0.06 0.34 + 0.1 0.44 + 0.1 0.42 + 0.14 0.32 + 0.07 0.40 + 0.08

0.42 ± 0.01 0.4 + 0.02 0.44 + 0.02 0.42 + 0.02 0.42 + 0.02 0.41 + 0.04 0.38 + 0.01

0 0.77 + 0.08 0.19 + 0.05 0.82 + 0.08 0.21 + 0.04 0.82 + 0.06 0.34 + 0.04

100% 65% 35% 47% 53% 30% 70%

Cr0.6Fe0.4NbO4

-1

Cr0.4Fe0.6NbO4

Quadrapole splitting, <Δ> (mm s )

Sample

1.00

Relative transmission

0.98

(a)

1.00 0.96

(b)

0.92 1.00 0.98

0.6

0.5

0.0

(c) 0

0.96

20

40

60

80

% Cr

1.00 0.96

Fig. 12. Change in average quadrupole splitting with varying chromium concentration in Cr1xFexNbO4 compositions (x = 0.2, 0.4, 0.6, 1).

(d)

0.92 -6

-4

-2

0

2

4

6

velocity (mm/sec) Fig. 11. Mössbauer spectra of (a) Cr0.8Fe0.2NbO4, (b) Cr0.6Fe0.4NbO4, (c) Cr0.4Fe0.6NbO4 and (d) FeNbO4 at room temperature.

compensation of majority charge carriers (holes) in CrNbO4 by electrons on iron substitution. 3.2.2. Complex impedance measurements The complex impedance data obtained for the compositions Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) were fitted with ZviewÒ (version 2.8). Typical Nyquist plots of CrNbO4, Cr0.8Fe0.2NbO4, Cr0.6Fe0.4NbO4, and Cr0.4Fe0.6NbO4 at different temperatures from 423 to 573 K are shown in Fig. 7(i) to (iv) respectively. The patterns are depressed and/or truncated semicircles. At lower temperatures, the data were fitted with two overlapping depressed semi-circles. The data at higher temperatures (above 473 K) could not be resolvable into two semi-circles and were fitted with a single parallel combination of resistor (R) and constant phase element (CPE). The impedance of CPE is given by ZCPE = [T (jx)/]1, where T is a constant and is designated as CPE-T, and / is the phase angle and is designated as CPE-P. The equivalent circuits used for single

and double depressed semi-circular fits are shown in Fig. 8 as EC1 and EC2 respectively. The estimated values of impedance obtained from complex non-linear least squared fitting for all the compositions at 473 K and 573 K are presented in Table ST1 of supplementary section. The conductivity of these samples in the temperature range shows Arrhenius behaviour as shown in Fig. 9. The activation energies for the conduction process in these samples were deduced from the Eq. (2) and presented in Table 2.

r ¼ r0 expðEact =kTÞ

ð2Þ

From the Arrhenius plot, it is evident that Cr0.8Fe0.2NbO4 exhibited the lowest conductivity, followed by Cr0.6Fe0.4NbO4. CrNbO4 and Cr0.4Fe0.6NbO4 exhibit higher conductivity and show a crossover around 500 K. The low activation energies indicate electronic conduction in all the compositions. These results are in accordance with the current-voltage characteristics presented in the previous section. 3.3. Magnetic studies 3.3.1. Vibrating sample magnetometer studies The variation in mass susceptibility (v) and inverse mass susceptibility (1/v) of Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) samples with

70

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

temperature (T) are shown in Fig. 10(a1) to (d1). CrNbO4 and iron substituted chromium niobate compositions exhibit a characteristic antiferromagnetic transition at low temperatures. Inverse susceptibility data of CrNbO4 follows Curie-Weiss equation and the paramagnetic Curie temperature, h and curie constant, C were calculated by extrapolating the linear fit data in the temperature range of 50 to 300 K, to zero susceptibility. The paramagnetic Curie temperature, h of CrNbO4 was found to be 33.8 K with a molar Curie constant, CM = 2.02. Christensen et al. reported a spin-glass transition at 9.3 K with a h value of 31 K and molar Curie constant, CM = 1.85 [1]. In other compositions 1/v versus T (temperature) followed a non-linear behaviour. The first derivatives of magnetization curves of Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) with respect to temperature are shown in Fig. 10(a2) to (d2). All these compositions exhibit antiferro to paramagnetic transition at lower temperatures. Figs. SF5 (a) and SF5 (b) in the supplementary section show the temperature dependent mass susceptibility and its first derivative for FeNbO4. It is seen that, FeNbO4 exhibited two antiferromagnetic transitions around 48 K and 250 K respectively. Neel temperatures were estimated from dvm/dT vs T profiles and the results are presented in Table 3.   Further, effective spin only magnetic moment lcalc per foreff mula unit (f.u) at 293 K was calculated using the Eq. (3). The observed magnetic behaviour with iron substitution was understood by simplistic calculations of expected change in moment due to substitution of Fe in Cr site. The change in magnetic moment (in Bohr magneton) was calculated using Eq. (4), by considering both low spin (LS) and high spin (HS) states of Fe3+. The theoretically calculated and experimentally obtained results are presented in Table 4.

C Th qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1  xÞ l2Cr þ x l2Fe

ð4Þ

2p3/2

572

576

580

584

588

(b)

Counts / (a.u)

2p3/2

2p1/2

568

3.3.2. Mössbauer studies Fig. 11 shows the Mössbauer spectra of iron containing samples. Iron is found to be present in +3 oxidation state corresponding to an isomer shift of about 0.42 mm/s. In FeNbO4 the value of effective magnetic moment at 300 K is deduced to be 5.81 lB based on VSM results. Corroboration of these results suggests the presence of Fe3+ in high spin state. The resultant valence charge distribution in a half-filled 3d subshell of Fe3+ (t2g""" eg"") is expected to result in a small quadrupole splitting. With increasing Cr3+ (t2g""") concentration at the cost of Fe3+, results in a decrease in the effective magnetic moment due to decrease in average number of unpaired electrons per atom in the lattice. This is expected to result in a tetragonal distortion leading to higher value of the quadrupole splitting. Hence the quadrupole splitting in chromium containing Cr1xFexNbO4 compositions is also indicative of the existence of spin-orbit coupling interactions [18,19], which is consistent with the results corresponding to magnetization measurements. With decrease in iron content, peak broadening was observed with decreased quadrupole splitting parameter. The variation in weighted average quadrupole splitting with decreasing chromium

(a)

Counts / (a.u)

592

2p1/2

572

B.E / (eV)

576 580 584 Binding Energy / (eV)

588

592

(d)

(c) 2p3/2

2p3/2

Counts (a.u)

lcalc eff ¼

ð3Þ

Counts (a.u)



  The calculated spin only moment lcalc of CrNbO4 is 3.87 lB. eff The experimentally determined magnetic moment is 3.80 ± 0.05 lB is in agreement with the literature reported value of 3.85 ± 0.02 lB [1]. With iron substitution the measured magnetic moments are less than the weighted high spin (HS) or weak field spin-only moment calculated from the Eq. (4). This may be attributed to the spin-orbit interactions involved. Further, it can be noted that the weighted low spin (LS) or strong field spin-only moment in Cr0.6Fe0.4NbO4 and Cr0.4Fe0.6NbO4 are lower than the measured values. This indicates the presence of high spin Fe3+ in these compositions with varying spin-orbit interactions. This is also supportive to the increase in unit cell volume with iron substitution as discussed in Section 3.1.

2p1/2

570

575

580

585

Binding energy (eV)

590

595

570

2p1/2

575

580 585 Binding energy (eV)

590

Fig. 13. Typical XPS spectra of Cr2p components of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4.

71

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

3p3/2

(b)

(a)

Counts (a.u)

3p3/2

355

360

365

370 375 Binding energy (eV)

380

Counts(a.u)

3p1/2

385

3p1/2

390

355

360

365

370 375 380 Binding energy (eV)

385

(c) 3p3/2

Counts (a.u)

360

365

370

375

380

Counts (a.u)

3p1/2

355

(d)

3p3/2

385

390

3p1/2

355

360

365

Binding energy (eV)

370 375 380 Binding energy (eV)

385

390

Fig. 14. Typical XPS spectra of Nb3p components of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4.

(a)

2p 1/2 2+

Fe (s) 3+

Fe (s)

(b)

2p3/2

Counts (a.u)

Counts (a.u)

2p 3 /2

2+

Fe (s) 3+

2p1/2 2+

Fe (s)

3+

Fe (s)

Fe (s)

2+

Fe (s)

705

710

715

720

725

730

735

704

708

712

716

Binding energy (eV)

720

724

728

732

Binding energy (eV)

Counts (a.u)

2p3/2

(c)

2p1/2 3+

2+ Fe (s) Fe (s) 2+

Fe (s)

705

710

715 720 725 Binding energy (eV)

730

3+

Fe (s)

735

Fig. 15. Typical XPS spectra of Fe2p components of (a) Cr0.8Fe0.2NbO4, (b) Cr0.6Fe0.4NbO4 and (c) Cr0.4Fe0.6NbO4.

concentration is shown in Fig. 12. The linear variation in relative area of Mössbauer spectra with increase in chromium concentration in Cr1xFexNbO4 is presented in Fig. SF6 (supplementary section). The variations in quadrupole splitting, line width, isomer shift and relative area of de-convoluted peaks with increase in chromium concentration are presented in Table 4.

3.4. X-ray photoelectron spectroscopic studies The high resolution selected area scans of Cr2p, Nb3p, Fe2p and O1s of Cr1xFexNbO4 (x = 0, 0.2, 0.4 and 0.6) are shown in Figs. 13– 16 respectively. In all the compositions, the peaks corresponding to Cr2p are asymmetric with a shoulder appearing at higher binding

72

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

(b)

Lattice oxygen 1

Counts (a.u)

Counts (a.u)

(a)

Lattice Oxygen

Lattice oxygen 2 Hydroxyl oxygen

Hydroxyl Oxygen

526

528

530 532 534 Binding Energy (eV)

536

538

526

(c)

528

530 532 Binding Energy / (eV)

534

(d)

Counts (a.u)

Counts (a.u)

Lattice oxygen (1)

Lattice oxygen (2)

Lattice oxygen 1

Hydroxyl oxygen

Lattice oxygen 2

526

528

530 532 534 Binding energy (eV)

536

524

526

528

530

532

534

536

538

Binding energy (eV)

Fig. 16. Typical XPS spectra of O1s components of (a) CrNbO4, (b) Cr0.8Fe0.2NbO4, (c) Cr0.6Fe0.4NbO4 and (d) Cr0.4Fe0.6NbO4.

Table 5 Relative percentages (±5%) of different oxidation states of individual elements and ratio of relative concentrations of higher valence chromium and lower valence iron in Cr1xFexNbO4 (x = 0, 0.2, 0.4, 0.6) samples. % Cr3+

% Cr4+d or Cr6d

Fe3+

Fe2+

Olat1

Olat2

(Cr4+d or Cr6d):Fe2+

CrNbO4 Cr0.8Fe0.2NbO4 Cr0.6Fe0.4NbO4 Cr0.4Fe0.6NbO4

63 70 81 78

37 30 19 22

– 79 86 75

– 21 14 25

100 88 73 92

– 12 27 8

– 11.32 4.59 1.50

2+

6+δ

1.2

0.8

4+δ

log([Cr

energy. Niobium 3p3/2 and 3p1/2 components are symmetric and are corresponding to a single oxidation state of +5. Iron exhibits an asymmetric pattern with a shoulder at lower binding energy spectrum along with the characteristic satellites of Fe2+ and Fe3+ oxidation states. XPS studies of CrNbO4 [10] were considered for comparing all other spectra. The binding energies of all the components are presented in Table ST2 (supplementary section). The little variations observed in metal ion binding energies, with respect to pristine compound may be due to the distortion in oxide ion octahedra around the central metal ion, due to iron substitution. In O1s spectrum of iron containing samples, lattice oxygen was fitted with two components, probably corresponding to two different metal ion environments around the O2 ions. The distorted MO6 octahedra due to iron substitution resulting in two different M-O bond lengths with different electron binding energies of bonded oxygen atoms. The different valence states of individual elements were quantified by considering total area of the respective peak components (e.g. for Cr2p, the total area of the peak components 2p3/2 and 2p1/2 are considered). Iron valences were quantified by ignoring the satellite peaks. The distribution of different oxidation state components of individual elements present in Cr1xFexNbO4 samples are presented in Table 5. The ratio of higher valence chromium to that of lower valence iron is decreasing with increasing iron content.

or Cr ] / [Fe ])

Sample

0.4

0.0 0.2

0.3

0.4

0.5

0.6

x Fig. 17. Variation of higher valence chromium species relative to lower valence iron species with increase in iron concentration in Cr1xFexNbO4 (x = 0.2, 0.4, and 0.6) samples.

The co-existence of Cr4+ and Fe2+ in FeCr2O4 spinels, is indicative of the plausible electron exchange between iron and chromium on thermodynamical grounds [20–22]. The probable Kröger-Vink notation for iron-substituted compounds is given in Eq. (5).

A. Sree Rama Murthy et al. / Materials Science and Engineering B 217 (2017) 63–73

ð1  xÞCr Cr þ NbNb þ 4OO þ xFeFe y ! ð1  x  yÞCr Cr þ þ NbNb þ ð4  zÞOO þ V 000 3 Cr z þ ðx  zÞFeFe þ zFe0Fe þ V O 2

73

diffraction measurements. Also, acknowledge Mr. P. Rajasekar from Indian Institute of Science, Bangalore for recording the scanning electron microscopic images of the samples.

yCrcr

ð5Þ

The variation in concentration of higher valence chromium with respect to lower valence iron is shown in Fig. 17. The concentration of hole donating species is decreasing exponentially with increase in iron concentration. These changes in acceptor/donor concentrations are manifesting in the electrical conductivity in Cr1xFexNbO4 with increasing iron content (Section 3.2.1). The reason for this behaviour is not clearly known presently. 4. Conclusions Chromium niobate and iron-substituted chromium niobates were prepared by solid-state method. Powder X-ray diffraction analysis showed that up to 60% of iron substitution, Cr1xFexNbO4 samples crystallized in P42/mnm space group. Microanalysis shows the homogeneous distribution of iron, chromium and niobium in the respective samples. Surface morphology studies revealed sub-micron level grain growth in all the samples. Non-linear current-voltage characteristics were revealed in resistance–voltage space. Total conductivity and activation energies were deduced from the complex impedance spectra recorded in the temperature range of 423–723 K. Initially the conductivity decreased due to Fe substitution up to 20% in Cr1xFexNbO4 and exhibited an increase in the case of 40, 60% iron substituted samples. Magnetization studies showed an anti-ferromagnetic transition in all the compositions. Increase in quadrupole splitting in 57Fe Mössbauer spectra was observed with decreasing chromium content. Room temperature Bohr magneton number and 57Fe Mössbauer spectra revealed the presence of Fe3+ in higher oxidation state. Further, metal and oxygen ions valences were deduced from X-ray photoelectron spectroscopic studies. The increase in unit cell volume upon iron substitution and variations in electrical properties could be helpful in the development of wide dynamic range hydrogen sensors, in contrast to limited sensing range of the existing semiconducting metal oxide sensors. Extensive studies on the analyte sensing characteristics of different compositions may help identifying suitable material for a given gaseous molecule. Such a study results in building an electronic nose for a given application by choosing an array of differentially selective compositions towards a given set of analytes. Acknowledgements Authors acknowledge Dr. K. Vinod, and Mr. R. Raja Madhavan from Indira Gandhi Centre for Atomic Research, Kalpakkam for carrying out the vibrating sample magnetometer and powder X-ray

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.mseb.2016.12. 001. References [1] A.N. Christensen, T. Johanssen, B. Lebech, J. Phys. C: Solid State Phys. 9 (1976) 2601. [2] Y.E. Smirnov, I.A. Zvereva, Russ. J. Gen. Chem. 73 (2001) 165–168. [3] D.B. Rogers, R.D. Shannon, A.W. Sleight, J.L. Gillson, Inorg. Chem. 8 (1969) 841– 849. [4] B.-O. Marinder, E. Dorm, M. Seleborg, Acta Chem. Scand. 16 (1962) 4. [5] Y.-J. Hsiao, Y.-H. Chang, T.-H. Fang, Y.-S. Chang, Y.-L. Chai, J. Alloy. Compd. 421 (2006) 240–246. [6] J.C. Pérez Flores, F. García-Alvarado, Solid State Sci. 11 (2009) 207–213. [7] S. Santhosh, N. Lakshminarasimhan, Ceram. Int. 40 (2014) 12129–12137. [8] C. Balamurugan, G. Bhuvanalogini, A. Subramania, Sens. Actuators, B 168 (2012) 165–171. [9] B. Khazai, R. Kershaw, K. Dwight, A. Wold, J. Solid State Chem. 39 (1981) 395– 400. [10] A. Sree Rama Murthy, K.I. Gnanasekar, V. Jayaraman, A.M. Umarji, T. Gnanasekaran, IEEE Sens. J. 15 (2015). [11] A.S.R. Murthy, K.I. Gnanasekar, V. Jayaraman, T. Gnanasekaran, in: J.P. Raina, M. Khalid, Z.C. Alex, S.K. Sudheer, S.A.V.S. Murthy, P.V. Krishna, I.S. Rao (Eds.), International Conference on Sensors and Related Networks, Allied Publishers Pvt Ltd, India, Vellore, India, 2009, pp. 69–72. [12] K.I. Gnanasekar, V. Jayaraman, E. Prabhu, T. Gnanasekaran, G. Periaswami, Sens. Actuators, B 55 (1999) 170–174. [13] M. Lv, S. Ni, Z. Wang, T. Cao, X. Xu, Int. J. Hydrogen Energy 41 (2016) 1550– 1558. [14] J. Végh, Surf. Sci. 577 (2005) 220–228. [15] C.J. Powell, P.E. Larson, Appl. Surf. Sci. 1 (1978) 186–201. [16] J.J. Yeh, I. Lindau, At. Data Nucl. Data Tables 32 (1985) 1–155. [17] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, 2nd ed., Elsevier, India, 2005. [18] P. Gütlich, R. Link, A. Trautwein, Mössbauer Spectroscopy and Transition Metal Chemistry, Springer, Berlin Heidelberg, 2013. [19] P. Day, Electronic Structure and Magnetism of Inorganic Compounds, Chemical Society, 1974. [20] H.E. Evans, A.T. Donaldson, T.C. Gilmour, Oxidation of Metals, 52 (1999) 379– 402. [21] S.E. Ziemniak, L.M. Anovitz, R.A. Castelli, W.D. Porter, J. Chem. Thermodyn. 39 (2007) 1474–1492. [22] A. Col, V. Parry, C. Pascal, Corros. Sci. (2016).