The Mg-Ti-H system: an accident of nature
Andrea Baldi
[email protected]
MH2008, June 27 2008, Reykjavik (Iceland)
(Ir)reversibility Hydrogenation of a Zr1.5Mn alloy
Why Mg-Ti? Metal-Hydride Batteries
Hydrogen Sensors Slaman et al., Sens. & Act. B 2007
Material
Smart coatings
Hydrogen capacity (wt% H)
Mg80Sc20
6.67
Mg80Ti20
6.53
Mg80V20
6.38
NiMH
~1.4
Baldi et al., Int. J. Hydrogen Energy 2008
H2
Resistivity
Niessen et al., Electrochem. Solid State Lett. (2005)
Mg0.7Ti0.3Hx ρ ∝ logT 10
100
MgH2 = insulator TiH2 = metal
Temperature (K)
Borsa et al., Phys. Rev. B (2007)
Borsa et al., Appl. Phys. Lett. (2006)
The immiscible alloy Mg-Ti alloy should not exist! Immiscible alloy: ΔHmix > 0 1600
βTi
L
Temperature (°C)
1400 1200 1000 ~865°C 800
αTi
~651°C
600
Mg 400 0
Ti
10
20
30
40
50
at.% Mg
60
70
80
90
100
Mg
The immiscible alloy Non-equilibrium processes
Metastable Phases Cu
Equilibrium at RT
Fe FCC+BCC
Liquid quenching Thermal evaporation
FCC
Sputtering Sputtering at LN Mechanical alloying Ma, Prog. Mater. Sci. (2005)
BCC
The immiscible alloy Non-equilibrium processes
Mg
Ti
Magnetron sputtering
Metastable Phases
The immiscible alloy Non-equilibrium processes
Metastable Phases
Room T Substrate
Why is it then so stable?! Mg
Ti
Magnetron sputtering
MgyTi1-y thin films + Pd
Microstructure X-Ray Diffraction Mg0.7Ti0.3
Ti
hcp-(002) as-deposited hydrogenated de-hydrogenated
38 2θ (°)
Intensity (a.u.)
Mg-Ti gradient: hcp-(002) peak
39
fcc-(111)
37 Mg-Ti
36 35
32
33
34 35 2θ (°)
36
37
34 0
20
40 60 at.% Mg
Mg 80 100
“Conventional diffraction methods are of limited sensitivity to short-range inhomogeneities or ordering.” Ma, Prog. Mater. Sci. (2005)
Microstructure X-Ray Diffraction Random 39
Mg-Ti gradient: hcp-(002) peak
Ordered
Ti
2θ (°)
38 37 Mg-Ti
36 35 34 0
20
40 60 at.% Mg
Mg 80 100
Mg “Conventional diffraction Ti methods are of limited sensitivity to short-range inhomogeneities or ordering.” Ma, Prog. Mater. Sci. (2005)
Coherency: multilayer Cu/Co Co Cu
Co Cu Co Cu
N times
Substrate
I = IN(ICu+ICo+ICuCo)
Michaelsen, Phil. Mag. (1995)
Coherency: multilayer Mg/Ti Simulation Coherent Interface
Experiment
1 x [Ti(20nm)/Mg(40nm)]
Intensity
5 x [Ti(4nm)/Mg(8nm)] 10 x [Ti(2nm)/Mg(4nm)]
Intensity
2 x [Ti(10nm)/Mg(20nm)]
20 x [Ti(1nm)/Mg(2nm)] 40 x [Ti(0.5nm)/Mg(1nm)] 30 32 34 36 38 40 42 44
30 32 34 36 38 40 42 44
2θ (°)
2θ (°)
Microstructure Quenching: spinodal decomposition
AgNi
CuAg
Wikipedia
He et al., Phys. Rev. Lett. (2001); Phys. Rev. Lett. (2002)
Microstructure Quenching: spinodal decomposition
AgNi
Chemical Short-Range Order (CSRO) 1
0
-1
Ordered (Segregated)
Random
Ordered (Antiferromagnetic)
CuAg
He et al., Phys. Rev. Lett. (2001); Phys. Rev. Lett. (2002)
Short-range order EXAFS (Extended X-ray Absorption Fine Structure)
For a AyB1-y alloy He et al., Phys. Rev. B (2001)
Short-range order 0.9 CSRO
0.8 0.7 0.6 0.5
Titanium K-edge
0.4
EXAFS As deposited
0.3
After 1 cycle
0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 y = at.% Mg Baldi et al., submitted to Phys. Rev. B
The accident: molar volumes Molar Volume (cm3/mol) 14
MgH2 Mg
Mg
Ti
TiH2
10
H2 pressure Vmolar(Mg) ≈ Vmolar(TiH2)
TiH2
The accident: molar volumes Molar Volume (cm3/mol) 14
MgH2 Mg
Mg
Ti
TiH2
TiH2
MgH2
MgH2
10 Molar Volume (cm3/mol) Mg 14 10
Mg-V disproportionates! V González et al., in preparation
V
VH2
Conclusions
Material
Hydrogen capacity (wt% H)
Mg80Sc20
6.67
Mg80Ti20
6.53
Mg80V20
6.38
NiMH
~1.4
Fundamental Physics Resistivity
Applications
ρ ∝ logT 10
Non-trivial microstructure
100
Temperature (K)
Accident of nature... MgH2 Mg
Mg
Ti
TiH2
TiH2
Acknowledgments
Robin Gremaud
Petra de Jongh
Marta González
Bernard Dam
Kees Balde
Ronald Griessen
Ad v.d. Eerden
Herman Schreuders
Jan Rector
Google