Breaking Invisible Specimens with Zero Force · PDF fileLoading through fabrication-induced...

of 20 /20
Breaking Invisible Specimens with Zero Force Roberto Ballarini University of Minnesota Collaborators: Arthur Heuer, Hal Kahn Sponsors: NSF, DARPA, ARO, NASA Strength and Fracture Standards at the Micro and Nano Scales 1/27/08

Embed Size (px)

Transcript of Breaking Invisible Specimens with Zero Force · PDF fileLoading through fabrication-induced...

  • Breaking Invisible Specimens with Zero Force

    Roberto BallariniUniversity of Minnesota

    Collaborators:

    Arthur Heuer, Hal Kahn

    Sponsors:

    NSF, DARPA, ARO, NASA

    Strength and FractureStandards at the Micro and

    Nano Scales1/27/08

  • Surface Micromachining

    Si wafersacrificial SiO2 (4 m)

    undoped polysilicon (5.7 m)masking SiO2 (1 m)

    Plasma etch

    Pd sputter (17 nm)

    B diffusion dopeHF release

    HF release

  • Two types of on-chip specimenshave been developed:

    Loading through electrostatic actuationLoading through fabrication-induced residual stress

  • 100 m

    100 m

    specimen

    electrostatic actuator notch

    notch

    500 m

    Fatigue Testing

    1438 pairs of comb fingers; 0.8 mN at 150 V

  • ADVANTAGES OF THIS ON-CHIP SPECIMEN

    No need for external loading device.Resonance loading can be used to study very high cycle fatigue.

    Uncracked ligament size of the same order as dimensions of typical MEMS components.

    Can adjust mean stress and alternating stress.

  • Stre

    ss

    Schematic Bend Strength Tests

    Monotonic Increasing Amplitude Fatigue

    1 Comp/Mono Fixed

    Fatigue/Mono High T Hold/Mono

    Time

    Stre

    ss

    0

    m cr

    10 min.= 6x106cycles

    TimeSt

    ress

    0

    crhold

    10 min.

    =(max -min )

    Time

    Stre

    ss

    0max = cr

    min

    m

    ~1 min.(6x105cycles)

    Time

    0

    cr

    min

    Time

    Stre

    ss

    0

    cr

  • Dynamic Fatigue of Polysilicon

    Bagdahn and Sharpe, Sensors and Actuators A, 2003

    1.0

    0.8

    0.6

    0.4

    (Pea

    k S

    tress

    /Mea

    n Te

    nsile

    Stre

    ngth

    )

    100 103 106 109 1012

    Cycles to Failure, N

    10 kHz 1 Hz 40 kHz 50, 200, 1000, 6000 Hz

    Nor

    mal

    ized

    Fat

    igue

    Stre

    ss

    Bagdahn and Sharpe

    No frequency dependence of fatigue life, only on total number of cycles

    Sharpe et al., Bagdahn and Sharpe, and

    Muhlstein et al.used the MUMPSfoundry

  • Dynamic Fatigue Resultslow-cycle fatigue

    1

    2

    3

    4

    5

    6

    -4 -3 -2 -1 0 1

    Load Ratio, RLow

    -Cyc

    le F

    atig

    ue S

    tren

    gth,

    m

    ax(G

    Pa)

    PolySi thickness Test Ambient3.5 m air (105 Pa)5.7 m air (105 Pa)5.7 m vacuum (10 Pa)

  • Monotonic Bend Strength after cycling with a fixed mean stress

    Time

    Stre

    ss

    0

    crit

    Time

    Stre

    ss

    0

    m crit

    a = /2

    Mon

    oton

    ic S

    tren

    gth,

    cr

    it(G

    Pa)

    Fatigue Amplitude, a (GPa)

    Mean Stress, m = -2.2 GPa

    2

    2.5

    3

    3.5

    4

    0 0.5 1 1.5 2

  • Monotonic Bend Strength after cycling with a fixed (low) amplitude

    2

    2.5

    3

    3.5

    4

    4.5

    -5 -4 -3 -2 -1 0 1 2 3

    Mean Stress, m (GPa)

    Mon

    oton

    ic S

    tren

    gth,

    cr

    it(G

    Pa)

    a = 1.0 GPa

    Time

    Stre

    ss

    0

    crit

    Time

    Stre

    ss

    0

    m crit

    10 min.= 6x106cycles

    a = /2

  • polysilicon

    indent

    2h = 500 m

    w=60 maA

    silicon substrate

    polysilicon

    SiO2 anchors

    B pre-crack

    Stress Intensity and Stress vs. Crack Length

    0.0

    0.2

    0.4

    0.6

    0.8

    1.0

    1.2

    1.0 5 9 13 17 21 25 29 33 37 41 45 49 53 57

    Crack Length [um]

    Str

    ess

    Inte

    nsi

    ty

    [MP

    a m

    1/2 ]

    0.05.010.015.020.025.030.035.040.045.050.0

    Str

    ess

    [MP

    a]

    Stress Intensity Stress*

    )(* FaK =

    * = residual /(1+4aV()/2h)

    PASSIVE DEVICE ASSOCIATEDWITH CONSTANT TENSION

    (Science 298, 1215-1218, 2002)

  • 5 um

    indent

    substrate

    pre-crack

    beam

    residual tensile stressbeam anchor (to substrate)

    crack tip

    1 um

    INDENTATION CRACK

    CWRU

  • 0

    0.2

    0.4

    0.6

    0.8

    1

    1.2

    1.4

    1.6

    0 5 10 15 20 25 30 35 40 45 50 55 60

    Crack Length [m]

    Stre

    ss In

    tens

    ity [M

    Pa m

    1/2 ]

    = no propagation = propagation to failure

    45 MPa

    56 MPa

    69 MPa

    FINE-GRAINED POLYSILICONFRACTURE TOUGHNESS

  • 0

    0.2

    0.4

    0.6

    0.8

    1

    1.2

    1.4

    1.6

    0 5 10 15 20 25 30 35 40 45 50 55 60

    Crack Length [m]

    Stre

    ss In

    tens

    ity [M

    Pa m

    1/2 ]

    = no propagation = propagation to failure

    69 MPa

    FINE-GRAINED SILICONSTATIC FATIGUE STUDY

    90% RH

    K between 0.62- 0.86 MPa-m1/2No growth in 30 days

    V< 3.9 x10-14 m/sSame results for eight multipoly specimens

  • Static Fatigue ExperimentNotched Tensile BeamsUndoped LPCVD Polysilicon

    deposited at 570C, annealed at 615C 318 MPa (Tensile)500 m

    20 m

    a

    20 m

    a

    a(m) max (GPa) broken5.2 2.3 0/146.2 2.8 0/147.2 3.4 0/148.2 3.9 11/149.2 4.5 14/1410.2 5.0 14/14

    After 200 hrs in 90% humidity no additional beams broke

  • 20 m

    Small Residual TensionSputtered Aluminum

  • Anchor

    Anchor Anchor

    Anchor

    Anchor Anchor

    200 m

    20 m 20 m

    10 m

    Large Residual TensionSilicon Nitride

    Residual CompressionColumnar Polysilicon

  • 0

    20

    40

    60

    80

    100

    120

    140

    160

    0 1 2 3 4 5

    1st test2nd test3rd test4th test

    with fibril

    1st test2nd test

    without fibril

    Displacement (m)

    Appl

    ied

    Vol

    tage

    (V)

    0

    20

    40

    60

    80

    100

    120

    0 5 10 15 20 25 30 35

    Strain (%)S

    tress

    (MP

    a)

    200 m

    5 mA

    B D

    C

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    with fibrilwithout fibril

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    with fibrilwithout fibril

    F

    E

    fixed fingers

    overlap

    movablefingers

    tether beams

    anchor

    anchoranchor anchor

    0

    20

    40

    60

    80

    100

    120

    140

    160

    0 1 2 3 4 5

    1st test2nd test3rd test4th test

    with fibril

    1st test2nd test

    without fibril

    Displacement (m)

    Appl

    ied

    Vol

    tage

    (V)

    0

    20

    40

    60

    80

    100

    120

    0 5 10 15 20 25 30 35

    Strain (%)S

    tress

    (MP

    a)

    200 m

    5 mA

    B D

    C

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    with fibrilwithout fibril

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    0

    10

    20

    30

    40

    50

    60

    0 1 2 3 4 5

    Displacement (m)

    Forc

    e (

    N)

    extra force needed to strain fibril

    with fibrilwithout fibril

    F

    E

    fixed fingers

    overlap

    movablefingers

    fixed fingers

    overlap

    movablefingers

    tether beams

    anchor

    anchoranchor anchor

  • Slide Number 1Surface MicromachiningSlide Number 3Fatigue TestingSlide Number 5Slide Number 6Slide Number 7Dynamic Fatigue of PolysiliconSlide Number 9Monotonic Bend Strengthafter cycling with a fixed mean stressMonotonic Bend Strengthafter cycling with a fixed (low) amplitudeSlide Number 12Slide Number 13Slide Number 14Slide Number 15Slide Number 16Slide Number 17Slide Number 18Slide Number 19Slide Number 20