Magnetocaloric Materials not only for cooling applications

Ekkes Brück

Fundamental Aspects of Materials and Energy,

Radiation Radionuclides Reactors, TNW,

Delft University of Technology, The Netherlands

Introduction

Magnetic cooling

Giant magnetocaloric effect

power generation

Review: E. Brück, Magnetic refrigeration near room temperature, Handbook of magnetic materials Vol 17 chapt. 4 (2007) ed. K.H.J. Buschow

Cooling techniques

Physical principle Application

Expansion ideal gas Stirling-cooler Claude-turbine

Joule-Thomson effect Liquefactor (Linde)

Peltier effect Electronics Infrared visors

Evaporation House-hold refrigerator Cool box

Adiabatic demagnetization Low-temperature physics

Introduction

Magnetic cooling: Debye and Giauque 1926

Nobel prize 194961g Gd2(SO4)3·8H2O, ∆B=0.8T, 1.5K →0.25K

Introduction

Introduction

Magnetic refrigeration is based on magnetocaloric effect (MCE)

∆Smax important for cooling capacity

∆T important for heat flow

( , )f

i

B

m

B

MS T B dB

T

∂ ∆ ∆ = − ∂ ∫

,

( , )f

i

B

p BB

T MT T B dB

C T

∂ ∆ ∆ = − ∂ ∫

4. Pulse field technique

3. Direct measurement of change of T

BTT

BTMBTMBTS

i ii

iiiim ∆

−−=∆∆ ∑

+

++

1

11 ),(),(),(

1. Magnetic measurements

'

0'),(

),( dBT

M

BTC

TBTT

B

Bad ∫

∂∂−=∆

'

0'

'' )0,(),(),( dT

T

TCBTCBTS

T

m ∫−=∆

2. Specific-heat measurements

Determination of MCE

Gd5(SixGe1-x)4 1997

MnAs1-xSbx 2001

La(Fe1-xSix)13 2001MnFeP1-xAsx 2002

Ni0.5Mn0.5-xSnx 2005

Giant MCE

First-order field-induced

magneto-structural transition

Pulse field technique

An adiabatic M(H) curve will intersect the isothermal

curves obtained at higher temperatures

Pulse field magnet allows fast magnetic measurements

Thermocouple enables to measure the temperature

more accurate before and after the pulse taking place

T = 77 – 400 KB = 0 – 20 Tesla

150

kJ c

apac

itor

Transition-metal compounds

High abundance (low price)

Intermediate magnetic moment (moderate MC effect)

Frequently Curie temperatures exceeding RT

Strong coupling to lattice (Simultaneous magnetic and structural transitions or metamagnetism)

Fe2P related materials

Hexagonal Fe2P type of structure

Bacmann, JMMM 1994

Space group:

P62m

Mn 3g sites

Fe 3f sites

P/As 1b&2c sites

_

Temperature dependence of magnetization

Step-liketransition

first order

but very littlehysteresis

0 100 200 300 4000

50

100

150

T (K)

M (

Am

2 /kg)

B=1 T

(1) (2)(3)

Replacing As by Si

(1) virgin effect, (2)heating, (3) subsequent cooling

MnFeP0.50Si0.50

very largehysteresis

Magnetic-entropy change

yet largeMCE

0 1 2 3 4 5 60

20

40

60

80

100

120

140

270 276 282 288 294 3000

10

20

30

40

50

60

70

80

90

B (T)

M (A

m2 /k

g)

269 K

303 K∆∆ ∆∆T

= 2 K

1 T 2 T 3 T 4 T 5 T

MnFeP.8Ge.2

T (K)

- ∆∆ ∆∆S

m (

J/kg

K)

Replace As by Ge

very largehysteresis

nice field induced transition

yet largeMCE

Concentration dependence of TCfor Ge and As

0.2 0.4 0.6150

200

250

300

350

400

450

500

550

600

MnFe(P1-x,Gex) MnFe(P1-x,Asx)

Cur

ie te

mpe

ratu

re (

K)

Concentration x

Almost linear concentration dependence.

Entropy change; effect of different element substitutions.

250 275 300 3250

5

10

15

20

25

30

35

Gd metal Mn

1.1Fe

.9P

.5As

.5

MnFeP.45

As.55

MnFeP.5As

.3Si

.2

- ∆∆ ∆∆S

m(J

/K-k

g)

T(K)

∆∆∆∆B: 0 - 2 T

More Mnincreases moment and MCE

Si produces sharper transition

285 290 295 300 305 310 3150

1

2

3

4

5∆∆∆∆B = 1.45 T

MnFeP0.45

As0.55MnFeP

0.47As

0.53

Mn1.1

Fe0.9

P0.47

As0.53

∆∆ ∆∆ Tad

(K

)

T (K)

Direct measurements MSU

Adiabatic temperature-change

Sample dependence need for careful preparation

220 240 260 280 300 320 340

0.3

0.4

0.5

0.6

Cp/T

(J/

mol

K2 )

T (K)

B = 0 T B = 0.5 T B = 1.0 T B = 1.5 T B = 2.0 T

Mn1.2

Fe0.8

P0.75

Ge0.25

(bulk)

specific heat in field

Sample dependence need for careful characterization

220 240 260 280 300 320 340

0.4

0.6

Cp/

T (

J/m

olK

2 )

T (K)

B = 0 T B = 0.5 T B = 1.0 T B = 1.5 T B = 2.0 T

Mn1.24

Fe0.76

P0.75

Ge0.25

(bulk)

1120oC-cdw-1120oC-20h-quenched

Improved Sample preparation

Melt-spinning

+ Ar gas pressure ∼ 1 atm.

surface speed of the wheel v = 40m/s+

ribbons were annealed for ± 10 min.+

Mn2-xFexP0.75Ge0.25 (x = 0.70, 0.76, 0.78, 0.80)+

Hexagonal Fe2P-type structure

Experimental results

Small thermal hyteresis, Tc = 288 K

Large MCE observed at low operation field

Experimental results

Magnetocaloric effect is directly observed

τ = 100 ms (f = 10 Hz): field sweep rate of 300 Tesla/sec. → adiabatic condition

Experimental results

∆Tad (H) is constructed via the crossing points of the adiabatic curve with the set of isothermal curves

For x = 0.80: ∆Tad ≈ 3 K/Tesla

Experimental results

Tc decreases with increasing the Mn/Fe ratio

Mn2-xFexP0.75Ge0.25 compounds:

Large MCE+

Small thermal hysteresis+

Large range of working temperature+

Specific heat measurements

Stot = Slat + Sm + Se

Magnetic field induces a shift of transition temperature ~ 4 K/T

MnO

'

0' ),(

),( dBT

M

BTC

TBTT

B

Bad ∫

∂∂−=∆∆

Adiabatic temperature change

Field driven 1st order magnetoelastictransition 150 K < Tc < 450 K .

MnFe(P,As) hexagonal above magnetic transition

hexagonal below.

Hardly any volume change( <0.1 %) but change of c/a.

Summary MnFe(P,As,Si,Ge)

dBT

MTdTcTdSdQ p ∂

∂+==

Heat input → temperature changemagnetization change

2222

−==dt

dB

R

SNRIWelect

Magnetocaloric power-generation

Edison's machine from 1892 directly employing the heat from the coal fire.Interesting design with very low efficiency.

Increased efficiency with regenerator.From numerical modeling 75% Carnot-efficiencyderived.

Modern design without moving parts.

Kirol & Mills JAP

0

25

20

15

10

5

0

0 20 40 60 80

T (C0)

∆S (

J/K

·kg)

Increased T span with active magnetic regenerator containing different materials with tailored TC

High efficiency with locally only small ∆T but working in series over large T span and thus larger power output.

Magnetocaloricenergy converter

Other heat

Possible scenario to increase efficiency of solar cells

ε≈20%400K in300K outε≈25%

Conclusions I

• First order magnetic transition common to giant MCE!

• Structural transition may cause extra hysteresis.

• Control of hysteresis very important but possible.

• Evaluation of entropy change needs care.

• Fe and Mn based systems with much lower materials costs.

• Relevant T range covered by MnFe(P,As,Si,Ge).

• Sample preparation simplest for MnFe(P,As) with As replaced by other element.

Conclusions II

1. Pulse field magnet provides a good approach for directly monitoring the MCE

2. ∆Tad is calculated by comparing the M(H) curves obtained in isothermal and adiabatic process

3. Mn2-xFexP0.75Ge0.25 ribbons exhibit excellent magnetocaloric properties

4. MnFe(P,As,Ge,Si) compounds can be used as magnetocaloric medium working at high frequencies

Thank you for your attention !