Mechanisms during FAST 4: Oxide ceramics · f *= stress intensification factor G = grain size n =...

47
March 24, 2011 | FAST School | O. Guillon | 1 Mechanisms during FAST 4: Oxide ceramics Olivier Guillon

Transcript of Mechanisms during FAST 4: Oxide ceramics · f *= stress intensification factor G = grain size n =...

  • March 24, 2011 | FAST School | O. Guillon | 1

    Mechanisms during FAST 4:

    Oxide ceramics

    Olivier Guillon

  • March 24, 2011 | FAST School | O. Guillon | 2

    Outline

    Methodology to answer this question:

    FAST vs. Hot Pressing

    • Identification of main densification mechanism

    • Evaluation of microstructure

    Effect of heating rate

    Effect of electric field/current

    Lower temperatures, higher densities, smaller grain sizes

    …compared to free sintering!

    Question: What really happens during FAST?

  • March 24, 2011 | FAST School | O. Guillon | 3

    Goal

    Transition from the „black box“ era to the intelligent „tool box“ approach

    When you know how a tool works, you can get more out of it!

    Based on experimental evidence

    ?

    !

  • March 24, 2011 | FAST School | O. Guillon | 4

    FAST vs. Hot Pressing

    FAST (HP 25/1, FCT Systeme) Measuring system:

    • Pyrometer (P) from 450°C (standard)

    • Thermocouple (T) from

    room temperature (optional)

    HP (HPW 150, FCT Systeme) Measuring system:

    • Thermocouple (T)

    Temperature calibration

    By melting copper powder

    (T)

    Sample

    HP FAST

    (P)

    (T)

    Graphite

    felt

  • March 24, 2011 | FAST School | O. Guillon | 5

    a - alumina (Al2O3)

    Insulator (dielectric)

    Model material for sintering studies

    Purity: 99.99 %

    Average particle size: 150 nm

    (TMDAR, Taimei Chem., JP)

    Purity: 99.80 %

    Average particle size: 700 nm (CT 3000 SG, ALMATIS, USA)

    Theoretical density: 3.986 gcm-3

    Powder filling

    Ø 20mm pressing tool

    Initial compaction

    (50 MPa, 3 min)

    Load adjustment

    15-50 MPa

    Start of the sintering experiments

    Heating rate: 10 Kmin-1

    Max. temperature: 1100-1250 °C

    Atmosphere: Vacuum

    Dwell time: up to 2 h J. Langer, M. Hoffmann, O. Guillon, Acta Materialia (2009)

  • March 24, 2011 | FAST School | O. Guillon | 6

    Sintering behavior of alumina

    Curves for FAST and HP show similar trends, for both particle sizes

    Densification starts earlier for FAST but same final densities

    Differences in relative density at the beginning of dwell time:

    150 nm: Drrel = 0.076 DT ≈ 25 K

    700 nm: Drrel = 0.026 DT ≈ 05 K

    Rela

    tive d

    ensity

    Time

    Dwell time Dwell time

    Time

    Re

    lative

    den

    sity

  • March 24, 2011 | FAST School | O. Guillon | 7

    Analysis of the sintering mechanism

    r = relative density pa = applied pressure

    H = numerical constant T = temperature

    f = stress intensification factor* G = grain size

    n = stress exponent m = grain size exponent

    0

    lnh

    hz

    n

    am

    zp

    kTG

    HD

    dt

    d

    dt

    df

    r

    r

    1

    True strain:

    1

    01

    11

    r

    rf* after Montes et al., Comp. Mat. Sci. (2006)

    ρ0 green density

    or Helle et al., Acta Metall. (1985)

    0

    0

    ²

    1

    rrr

    rf

    Mechanism Stress

    exp. n

    Grain size

    exp. m

    Lattice diffusion 1 2

    Grain boundary

    diffusion

    1 3

    Viscous flow 1 0

    Grain boundary

    sliding

    1 or 2 1

    Plastic

    deformation

    ≥3 0

    M.N. Rahaman, Ceramic Processing and Sintering (2003)

  • March 24, 2011 | FAST School | O. Guillon | 8

    Grain size measurement

    Sample cross-section

    Homogeneous grain size

    in the whole sample for all

    investigated densities

    Grain size analysis

    (linear intercept method on

    SEM micrographs)

    h

  • March 24, 2011 | FAST School | O. Guillon | 9

    Grain size measurement

    Marked grain growth from ρrel > 0.95

    Identical sintering trajectories for FAST and HP

    Reduced grain growth in comparison to free sintering

    R. Zuo, E. Aulbach, J. Rödel

    Acta Mater. (2003)

  • March 24, 2011 | FAST School | O. Guillon | 10

    Analysis of the sintering mechanism

    r = relative density pa = applied pressure

    H = numerical constant T = temperature

    f = stress intensification factor* G = grain size

    n = stress exponent m = grain size exponent

    n

    amp

    GkT

    HD

    dt

    df

    r

    r

    11

    T & G

    const.

    T & (fpa)

    const.

    Rela

    tive d

    ensity

    Time

    Dwell time

  • March 24, 2011 | FAST School | O. Guillon | 11

    Analysis of the sintering mechanism

    r = relative density pa = applied pressure

    H = numerical constant T = temperature

    f = stress intensification factor* G = grain size

    n = stress exponent m = grain size exponent

    n

    amp

    GkT

    HD

    dt

    df

    r

    r

    11

    T & G

    const.

    T & (fpa)

    const.

    n ≈ 1

    for FAST and HP

    m ≈ 3

    for FAST and HP

    Densification is controlled

    by grain boundary

    diffusion

    Rela

    tive d

    ensity

    Time

    Dwell time

  • March 24, 2011 | FAST School | O. Guillon | 12

    Calculation of the activation energy

    RT

    Q

    eDD0

    Similar value for the activation energy:

    Grain boundary diffusion

    HP Q = 430 ± 50 kJ/mol

    FAST Q = 420 ± 35 kJ/mol

    Wang & Raj,

    J. Am. Ceram. Soc. (1990) Q = 440 ± 45 kJ/mol

    D0 = diffusion coefficient T = absolute temperature

    D0 = pre-exponential factor R = gas constant

    Q = activation energy

    Rela

    tive d

    ensity

    Time

    Dwell time

  • March 24, 2011 | FAST School | O. Guillon | 13

    Results for other materials

    Comparative studies FAST / HP:

    Cubic zirconia (8YSZ), ionic conductor: same densification mechanism and sintering trajectory

    Zinc oxide (ZnO), semi-conductor: same densification mechanism (grain size very sensitive to temperature variations)

    Tetragonal zirconia (TZ3Y): same densification mechanism and sintering trajectory

    AlCuFeB quasi-crystals: same densification mechanism

    L. Ramond, G. Bernard-Granger, A. Addad, C. Guizard, Acta Mater (2010)

    G. Bernard-Granger, A. Addad, G. Fantozzi, G. Bonnefont, C. Guizard, D. Vernat, Acta Mater. (2010)

    J. Langer, M. Hoffmann, O. Guillon, J. Am. Ceram. Soc. (2011)

    J. Langer, M. Hoffmann, O. Guillon, J. Am. Ceram. Soc. (2011)

  • March 24, 2011 | FAST School | O. Guillon | 14

    Results for other materials

    K. Morita, B.N. Kim, H. Yoshida, K. Hiraga,

    Scripta Mater. (2010)

    MgAl2O4 spinel

    SPS

    with grain size

    Similar mechanisms as for hot pressing

    Low stress regime: diffusion

    High stress regime: climb-controlled dislocation creep

    Densification maps for MgO

    R. Chaim, M. Margulis

    Mat.Sci. Eng. A (2005)

  • March 24, 2011 | FAST School | O. Guillon | 15

    Investigation of the initial sintering stage

    Thermal conductivity proportional to the contact area between particles

    Same values for HP und FAST samples above 65% density

    Effect of the temperature overshoot at the beginning of the FAST process?

    Th

    erm

    al conductivity

    Relative density

    Laser Flash

    measurements

    J. Langer, M. Hoffmann, O. Guillon,

    Acta Materialia (2009)

    Alumina

  • March 24, 2011 | FAST School | O. Guillon | 16

    Temperature profile during FAST process

    Overheating up to 130 °C at the

    beginning of the heating process

    (pyrometer control only above ~400°C;

    input power fixed)

    Effects on neck formation and neck

    growth (especially for nano-powders)

  • March 24, 2011 | FAST School | O. Guillon | 17

    Effect of the pressing tool and

    temperature control

    0 1 2 3 4 5 6 7 80.50

    0.55

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    0.95

    1.00

    1) FASTPyrometer

    2) FASTThermocouple

    3) HPThermocouple

    4) HPFAST-Tool

    TMDAR

    50 MPaR

    ela

    tive

    De

    nsity,

    rre

    l

    Time, 103*t [sec]

    Begin of

    Dwell Time

    Tmax

    = 1200 °C

    Density difference depends on:

    • the pressing tool used

    • the temperature control

    Alumina

    FASTThermocouple:

    without overshoot

  • Effect of temperature control

    Dramatic effect of transient overheating on densification kinetics

    for sensitive materials like ZnO (20 nm particle size)

    J. Langer, M. Hoffmann, O. Guillon, J. Am. Ceram. Soc. (2011)

    0 1 2 3 40.45

    0.50

    0.55

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    0.95

    1.00

    FAST

    HP

    NG20

    750 °C

    50 MPa

    Re

    lative

    De

    nsity,

    rre

    l

    Time, 103*t [sec]

    Starting point

    Dwell time

    (a)

    0.0 0.5 1.0 1.5

    FASTPyro

    FASTThermo

    HP

    Starting point

    Dwell time

    (b)

    NG20

    550 °C

    50 MPa

  • Consequences on properties

    400 450 500 550 600 650 700 750

    103

    2x103

    Re

    sis

    tivity,

    re

    l [

    cm

    ]

    FAST rrel

    = 0.74

    HP rrel

    = 0.75

    NA90

    50 MPa / 750 °C

    Temperature, T [°C]

    Slightly lower resistivity for FAST samples, but same order of magnitude

    Correlates with the larger interparticle contact area shown by

    Young‘s modulus measurements (due to initial temperature overshoot)

    ZnO 90 nm

  • March 24, 2011 | FAST School | O. Guillon | 20

    Outline

    Methodology to answer this question:

    FAST vs. Hot Pressing

    • Identification of main densification mechanism

    • Evaluation of microstructure

    Effect of heating rate

    Effect of electric field/current

    Lower temperatures, higher densities, smaller grain sizes

    …compared to free sintering!

    Question: What really happens during FAST?

  • March 24, 2011 | FAST School | O. Guillon | 21

    Effect of heating rate: alumina

    -800 -600 -400 -200 0 200 400 600

    0.5

    0.6

    0.7

    0.8

    0.9

    1.0

    Rel

    ativ

    e D

    ensi

    ty

    Time (s)

    35 K/min

    50 K/min

    100 K/min

    150 K/min

    Isothermal step

    Same final density is reached

    Is there a change in the densification mechanism?

    O. Guillon & J. Langer,

    J Mater Sci, 2010

    1200°C

  • March 24, 2011 | FAST School | O. Guillon | 22

    Effect of heating rate: alumina

    Slope independent of the heating rate

    Same results for 8YSZ

    Heating rate

    [K/min]

    Slope

    [10-3 K-1]

    10 1.7 ± 0.2

    35 1.6 ± 0.2

    50 1.7 ± 0.2

    100 1.5 ± 0.2

    150 1.4 ± 0.3

  • March 24, 2011 | FAST School | O. Guillon | 23

    Master Sintering Curve approach

    n

    amp

    kTG

    HD

    dt

    df

    r

    r

    1

    RT

    QDD exp

    0

    dtRT

    Q

    Td

    pHD

    kG

    nn

    a

    m

    exp

    1

    0

    rfr

    ),(exp1

    00 0

    TtdtRT

    Q

    Td

    G

    pHD

    kt

    n

    m

    n

    a

    r

    rf

    r

    r

    43

    3

    G

    D

    G

    D

    kTdt

    d bbVV

    r

    r

    r

    rrr

    r

    0

    )(3

    )(exp

    1))(,(

    00

    mt G

    D

    kdt

    RT

    Q

    TtTt

    A unique MSC ρ=f(Θ) can be obtained if:

    • only one diffusion mechanism is dominant during sintering

    • the microstructure is function only of density

    Free sintering:

    FAST/HP:

    with

    Su &Johnson, J Am Ceram Soc (1996)

  • March 24, 2011 | FAST School | O. Guillon | 24

    -16 -15 -14 -13 -12 -11 -10

    0.5

    0.6

    0.7

    0.8

    0.9

    1.0

    Rel

    ativ

    e D

    ensi

    ty

    Log()

    200 250 300 350 400 450 500

    0.10

    0.15

    0.20

    0.25

    0.30

    0.35

    0.40

    0.45

    0.50

    Res

    idua

    l sum

    of s

    quar

    es

    Activation energy (kJ/mol)

    MSC: alumina

    Master Sintering Curve obtained for the whole sintering cycle

    Apparent activation energy of 290 kJ/mol

    No change in the densification behavior

  • March 24, 2011 | FAST School | O. Guillon | 25

    Activation energy values

    M. Aminzare, F. Golestani-fard, O. Guillon, M. Mazaheri, H.R. Rezaie,

    Materials Science & Engineering A, 2010

    Free sintering of alumina (same powder)

    Literature: Q = 400-1100 kJ/mol (!)

    Dry pressing: 700 20 kJ/mol

    Pressure filtration: 605 15 kJ/mol

    Interplay between different diffusion mechanisms

    when whole sintering curve taken into account

    In the density range 70-85%:

    450 kJ/mol for both sample types

    Dry pressing

    Pressure filtration

  • March 24, 2011 | FAST School | O. Guillon | 26

    High(er) heating rates

    F. Meng, Z. Fu, J. Zhang, H. Wang, W. Wang, Y. Wang, Q. Zhang,

    J Am Ceram Soc (2007)

    Exothermic reaction to produce heat (SHS)

    Applied pressure: 60-120 MPa

    Heating rate of 1600°C/min

    Dense alumina (99%)

    in a few minutes

  • March 24, 2011 | FAST School | O. Guillon | 27

    High(er) heating rates

    J. Zhang, F. Meng, R. Todd, Z. Fu, Scripta Mater. (2010)

    Same grain size and density, but different resistance to mechanical abrasion

    (1)

    (3)

  • March 24, 2011 | FAST School | O. Guillon | 28

    High(er) heating rates

    Grain boundaries in non-equilibrium:

    Diffuse, open structure as opposed to

    relaxed boundaries?

    Higher diffusion coefficient and thicker GB?

    Also to be seen in other materials?

    (1)

    (4)

    As-sintered

    Annealed at 1500°C

  • March 24, 2011 | FAST School | O. Guillon | 29

    High heating rates: ZnO

    S. Schwarz, O. Guillon

    20 nm particles

    50 MPa

    Higher heating rates improve sinterability

  • March 24, 2011 | FAST School | O. Guillon | 30

    High heating rates: ZnO

    50°C/min 100°C/min (ρ= 67%)

    S. Schwarz, A. Thron, K. van Benthem, O. Guillon

    Curved GB Faceted GB

    (ρ= 76%)

  • March 24, 2011 | FAST School | O. Guillon | 31

    Outline

    Methodology to answer this question:

    FAST vs. Hot Pressing

    • Identification of main densification mechanism

    • Evaluation of microstructure

    Effect of heating rate

    Effect of electric field/current

    Lower temperatures, higher densities, smaller grain sizes

    …compared to free sintering!

    Question: What really happens during FAST?

  • March 24, 2011 | FAST School | O. Guillon | 32

    Carbon contamination?

    Secondary Ion Mass Spectroscopy (SIMS)

    S

    Cross section x

    y

    h

    TM TR

    Carbon hardly diffuses into the specimen (constant C-signal at depth of ~2 µm)

    Oxygen vacancies are responsible for specimen darkening

    No influence of carbon on electrical conductivity / sintering behavior

    J. Langer, M. Hoffmann, O. Guillon,

    J. Am. Ceram. Soc. (2011) 8YSZ

  • XPS measurements on pure ZnO

    • Identical emission spectra for FAST und HP

    • Only ZnO could be detected

    • No trace of carbon

  • Resistivity changes in pure ZnO

    Sample

    Al2O3-felt

    Sample resistance

    T = 750 °C 2.5x102 Ω

    Graphite tool resistance

    T = 750 °C 2.0x10-3 Ω

    100 200 300 400 500 600 700

    0.6

    0.7

    0.8

    0.9

    1.0

    Relative density

    Re

    lative

    de

    nsity,

    rre

    l

    Temperature, T [°C]

    a)

    103

    104

    105

    106

    Resistance

    NA90

    SPS / 750 °C /

    50 MPa / 5 min

    Re

    sis

    tan

    ce

    , R

    [

    ]

    Vanmeensel et al. J. Mater Sci. (2008)

    Semi-conductor behavior

  • Electrical boundary conditions

    sample

    Al2O3-discs

    pa direction of current

    0 1 2 30.55

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    0.95

    1.00

    (a) Standard

    (b) Electrically insulated

    FAST

    NA90

    750 °C

    50 MPa

    Re

    lative

    de

    nsity,

    rre

    l

    Time, 103*t [sec]

    0 1000 2000 3000

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    0.95

    1.00

    Starting point

    dwell time

    750 °C

    ZnO

    J. Langer, M. Hoffmann, O. Guillon,

    J. Am. Ceram. Soc. (2011)

    No effect of electrical boundary conditions

  • Sintering trajectory of pure ZnO

    No significant effect of the electric field / current

    Effect of the temperature overshoot

    (electrically insulated)

    (b)

    z

    x

    rrel = 0.97 2 µm

    (a) rrel = 0.97

    z

    x

    2 µm

    Standard FAST

    Insulated FAST

  • March 24, 2011 | FAST School | O. Guillon | 37

    Behavior of doped ZnO

    Sample resistance:

    T = 750 °C ~ 5 Ω

    T. Misawa, N. Shikatani, Y. Kawakami, T. Enjoji, Y. Ohtsu, H. Fujita

    J Mater Sci (2009)

    Standard SPS

    Insulated SPS

    Magnetic

    current

    probe

    Error due to positioning

  • March 24, 2011 | FAST School | O. Guillon | 38

    Current flow through the sample

    M. Herrmann, B. Weise, K. Sempf, A. Bales, J. Raethel, I. Schulz

    Workshop IFAM Dresden (2006)

    Resistivity of sample material [Ωm]

    Graphite tool

    I sam

    ple/I

    tota

    l

  • March 24, 2011 | FAST School | O. Guillon | 39

    Electrically conductive composite materials

    K. Vanmeensel, A. Laptev, O. Van der Biest, J. Vleugels

    Acta Mat (2007)

    Electrical conductivity of a composite material depends on:

    - volume fraction of conductive and insulating phases (incl. porosity)

    - temperature

    Polder-Van Santen mixture rule:

    m matrix

    p particles

    Similar equation for thermal conductivity

    with

    V*m volume fraction of

    matrix phase in a

    partially sintered compact

    *

    *

  • March 24, 2011 | FAST School | O. Guillon | 40

    Electrical conductivity

    90 vol.% TiN (grey) - ZrO2 (white)

    pores (black)

    Dense ZrO2-TiN composites

    K. Vanmeensel, A. Laptev, O. Van der Biest, J. Vleugels

    J Eur Ceram Soc (2007)

  • March 24, 2011 | FAST School | O. Guillon | 41

    Percolation threshold

    K. Vanmeensel, A. Laptev, O. Van der Biest, J. Vleugels

    Acta Mater (2007)

    Percolation: a continuous path for current is created

  • March 24, 2011 | FAST School | O. Guillon | 42

    Effect on densification

    ZrO2-TiN (60/40)

    Transition from insulator to conductor-like behavior during sintering

  • March 24, 2011 | FAST School | O. Guillon | 43

    Effect of electric field on surface diffusion

    Before densification takes place

    Neck growth kinetics estimated from I-V curves

    (without additional Joule heating)

    No effect of electric field

    (limited to 10 V/cm)

    Specific Surface Area measurements confirm this result

    (identical with and without electric field)

    TZ-3Y @ 1050°C

    Sample thickness:

    2 mm

    M. Cologna, R. Raj

    J. Am .Ceram. Soc. (2010)

  • March 24, 2011 | FAST School | O. Guillon | 44

    Flash sintering

    M. Cologna, B. Rahkova, R. Raj

    J. Am .Ceram. Soc. (2010)

    TZ-3Y

    DC-Field But why then:

    Proposed explanation:

    Joule heating at grain boundaries (several hundreds of °C)

  • March 24, 2011 | FAST School | O. Guillon | 45

    400 600 800 10000.50

    0.55

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    Rela

    tive D

    ensity

    Temperature

    0V

    40V/cm; max. 0.02 A/cm²

    40V/cm; max. ~1.5 A/cm²

    40V/cm at 1150°C; max ~6 A/cm²

    1200 20 40 60 80 100 1200.50

    0.55

    0.60

    0.65

    0.70

    0.75

    0.80

    0.85

    0.90

    Isothermal time [min]

    Flash sintering: an effect of current

    0 2 4 6 8 10 12 14 16 18

    4.35

    4.40

    4.45

    4.50

    4.55

    4.60

    4.65

    4.70

    4.75

    8YSZ flashed at 1150°C with 40V/cm; max. 5A

    current switch off

    induces a

    temperature drop

    of ~ 450 °C

    Ab

    so

    lute

    De

    ns

    ity

    [g

    /cm

    ³ ]

    Isothermal time [min]

    R. Baraki, S. Schwarz, O. Guillon

    8YSZ

    AC Field from room temperature

  • March 24, 2011 | FAST School | O. Guillon | 46

    Comparison Flash Sintering / FAST

    FAST:

    limited voltage, applied from the beginning

    electric equipotential affected by the conducting pressing tool

    Flash sintering conditions are not expected in standard FAST

    S. Schwarz,

    O. Guillon

    ZrO2 (sinter-forging)

    DVsample =10 V

    electrode

    ZrO2 (in graphite tool)

    DVsample =2.68 V

  • March 24, 2011 | FAST School | O. Guillon | 47

    Conclusions

    Identification of sintering mechanism requires „clean“ experimental

    conditions to avoid misunterpretations

    For „insulating“ oxide ceramics:

    • Same densification mechanism for HP and FAST

    (not only based on activation energy considerations)

    • Identical grain growth behavior

    • No effect of electrical field during sintering in FAST

    Transient heating affects subsequent densification

    Densification behavior may depend on the heating rate (or not: MSC)

    Large electric current (not field) may lead to additional phenomena

    (Flash sintering)