Electron Microscopy Testing of Nanostructures

Transcription

1 MEMS devices for In-Situ Electron Microscopy Testing of Nanostructures Horacio D. Espinosa Y. Zhu, C-H. C Ke,, N. Moldovan Acknowledgments: NSF-NIRT, NSF-NSEC, FAA, ONR

2 NEMS Characterization Technique - Component Constitutive Behavior (CNTs/NWs( CNTs/NWs) - Electro-Mechanical Characterization at the Device Level

3 Techniques for Testing Nanostructures AFM Tension AFM bending Salvetat et al. PRL (1999) Marszalek et al., PNAS (2) SEM Tension Stiff cantilever to Piezo-positioner TEM Tension stretch Compliant cantilever Yu et al., Science (2) Haque and Saif, Exp. Mechanics (22)

4 Limitations Lack direct or simultaneous real time measurement of load, deformation and imaging of atomic defects. Imaging is required for both load measurement and atomic defects identification. However, the required magnifications are quite different. Do not exhibit a single loading condition specially at large deformations. Are not suitable for in-situ TEM (highest atomic resolution). Cannot measure specimen electronic properties under a well-characterized loading condition.

5 MEMS-based Material Testing System Displacement Control Force Control Thermal Actuator NWs/CNTs Thin Films Zhu, Moldovan and Espinosa, Appl. Phys. Lett. 86, 1356 (25)

6 Electrostatic (Comb Drive) Actuator Force Control shuttle movable combs stationary combs folded beam width gap wing arm overlap F=Nε V 2 h/d Height (h): 2 µm Pairs of combs: 3/12 Folded beam length: 125 µm Folded beam width: 2.5 µm Folded beam height: 3.5 µm Width (w): 2 µm Gap (d): 2 µm Overlap (o): 15 µm Length (l): 3 µm

7 Thermal Actuator Displacement Control (a) Anchors Motion u A A (b) Poly-Si Beam Air Cross Section A-A h h air Inclined beams θ V Shuttle Si 3 N 4 Si Substrate h n l α T l sinθ FS 2 2NEA Al U = ; ψ = 12I ( 2 2 ψ sin θ + cos θ ) ψ :axial over bending stiffness ratio Pair of inclined beams: 5/1 Beam length: 3 µm Beam width: 8 µm Beam height: 3.5 µm Prorok, Zhu, Espinosa, Bazant & Yakobson, in Encyclopedia of Nanoscience and Nanotechnology (24), p. 555; Zhu, Corigliano & Espinosa, in press J. Micromech. Microeng., (26)

8 25 2 Beam Angle Selection - 1 Buckling analytical FEA-linear region.3.25 U/aDTl 15 1 FEA-nonlinear region Kl/2EA Displacement U sinθ = α Tl I sin θ + cos θ 2 Al θ ( ) -N (mn) P cr min.5 Buckling force θ ( ) θ ( ) Ktbl 2EA = sin θ + cos EA sin 2 2 Beam internal force N = α T 2 cos 2 Al θ 12I Stiffness 12I θ 2 Al 2 θ + cos 2 θ

9 Beam Angle Selection o beam angle 3 o beam angle 1 µm 1 µm 1 beam pairs 5 beam pairs

10 Lumped System Model X A X LS K A K S K LS F X S =X A -X LS X K K LS LS S X + X X S LS S = + K = X K A S X X A A S = F = 2Nα TEAsinθ To test a particular structure, K S and X S are given Small K ls Large X LS high load sensor resolution large X A high temperature increase K LS is an important design parameter

11 Thermal Actuator Lumped Model CNT Example Thermal Actuator Capacitive Load Sensor Au film Specimen X T X L K T K S K L F Governing Equations: X S = X T -X L K S X S = K L X L K T X T = F K S X S F= NEAα Tcosθ Thermal actuator stiffness: K A Ebh 2 Eb h 12Eb = (sin θ + cos θ ) + 3 l l l 3 MWCNT specimen stiffness: Capacitive load sensor stiffness: 24EI = l 3 2EbLh = 3 l L = 3 L L 18.6 N/m h = N/m For a MWCNT specimen (l s = 2 µm; F=1 µn, ε = 4%) X T =157.3 nm, X S =93.5nm and X L =53.8nm Zhu and Espinosa, PNAS, Vol. 12, 25 K S K ES A ESπD t = = =1.7 N/m l l L S S sb 3 sb Pair of inclined beams: 5/1 Beam length: 3 µm Beam width: 8 µm Beam height: 3.5 µm

12 Thermal Actuator Multiphysics Simulation anchors Temp Actuation Voltage: 1V In Vacuum u Heat sink beams 55 o C 66nm displacement (nm) no sink beams one pair of sink beams three pairs of sink beams six pairs of sink beams displacement (nm) current (ma) /R actuation voltage (V) temperature increase close to specimen ( C) actuation voltage (V) Zhu, Corigliano & Espinosa, in press J. Micromech. Microeng., (26)

13 Comparison between Experiments and Simulations resistivity (Ω µm) experiment 1 experiment 2 experiment current (ma) avg. temp. change (K) displacement (nm) experiment 1 experiment 2 experiment 3 FEA Damage current (ma) Highest temperature Highest temperature (a) 1 µm (b) 1 µm (c) 1 µm

14 Load Sensor R 1 R 2 R 12 d C 1 C 2 R 2 C 3 R 1 R 3 S 2 S 1 R S 2 C = C C N A 2 1 = ε d 2NεA C = d + ο( d 2 d V sense V = C o C ~ 2NεAV 2 C d 3 1 d ) o d d 1 + d Zhu, Moldovan and Espinosa, Appl. Phys. Lett. 86, 1356 (25)

15 Load Sensor and Signal Conditioning Charge sensing method: V sense d Eliminates parasitic capacitances Double chip Architecture Both chips on a custom-made PCB; Minimizes stray capacitance and electromagnetic interference. Device Chip Sensing IC Chip 1 cm C1 Cf capacitive load sensor C2 - + Vref synchronous demodulator LPF output voltage

16 Load Sensor Calibration - 1 (a) Actuation voltage = 2 V 2 nm (e) output voltage (V) (b) time (s) (c) Actuation voltage = 5 V d 2 nm V Vsense = C f C output voltage (V) 1 mv corresponds to 1 ff (d) time (s)

17 5 Load Sensor Calibration displacement (nm) capacitance change (ff) Stiffness = 11.8 N/m actuation voltage (V) actuation voltage (V) displacement (nm) capacitance change (ff) load (nn) Load = displacement stiffness Device Resolution Capacitance change:.5 ff Displacement: 1 nm Load: 12 nn

18 Other Configurations Compression/ Buckling 2 µm Electronic measurement of load and elongation 2 µm

19 In-situ TEM Devices: Nanostructure & Thin Film Load sensor Thermal actuator Backside window Heat sink beams Displacement markers

20 Fabrication Process of In-situ TEM Device

21 in-situ SEM Experiments

22 Nanoscale Polysilicon Film - 1 FIB machined

23 Nanoscale Polysilicon Film - 2 Displacement Markers Fracture surface Young s modulus = 155±5 GPa Fracture strength.72~1.42 GPa Zhu and Espinosa, PNAS, (25)

24 in-situ SEM/TEM Experiments of CNTs

25 In-situ TEM Setup 1.5 mm TEM holder platform 5 mm Φ = 3 mm Electric contacts 1 1 mm 5 mm Chip with MEMS devices In collaboration with Prof. Ivan Petrov, Center for Microanalysis of Materials, UIUC

26 Specimen Mounting on MEMS A B FIB Cut 5 µm EBID Pt nanoweld C D

27 Experimental Data CVD Grown MWCNTs Load [µm] Ion Irradiation (FIB-TEM) E-beam 1 (TEM) E-beam 2 (TEM) Low irradiation (SEM) Stress [GPa] Ion Irradiation (13 shells) E-beam 1 (5 shells) E-beam 2 (3 shells) Low irradiation (1 shell) E=315 GPa Displacement [nm] Strain [%] E-beam 1 and 2 tested in TEM operated at 2 kv (Telescopic failure) E-beam 1 (nanowelding( very close to shuttle edge, failure inside observation window), E-beam E 2 (nanowelding( ~1.5 µm m from edge, failure just outside observation window) Ion Beam irradiation in FIB with gallium ions (entire cross-sectional sectional failure) Low irradiation sample tested in SEM operated at kv (Telescopic failure inside observation window)

28 Constitutive Behavior and Failure Modes Stress [GPa] Ion Irradiation (13 shells) E-beam 1 (5 shells) E-beam 2 (3 shells) Low irradiation (1 shell) E=315 GPa Two Failure Modes Strain [%] 5 nm CVD Grown MWCNTs E-beam 1 and 2 tested in TEM operated at 2 kv (T) Ion Beam irradiation in FIB with gallium ions (CS) Low irradiation sample tested in SEM operated at 5 kv (T) Young s Modulus E=315 GPa 5 nm Nano diffraction T: telescopic ; CS: entire cross-section

29 a Telescopic Failure 1 b 2 c

30 Load Transfer Mechanism- Inter-Shell Bridging SWCNT bundles Kis et al., Nature Mat. (24) Vacancy or Interstitial threshold Energy is 86 kev DFT prediction of C atom in a fourfold coordinated interstitial Telling et al., Nature Mat. (23)

31 Discussion QM MM-MTB MTB Experiments Arc-grown r h Zhang et al., PRB, 71 (25) CVD and arc grown MWCNTs exhibit the same E and comparable failure stresses. For the first time, in-situ TEM experiments reveal two failure modes with increasing number of failing shells as the irradiation dose and type is varied. The lowest dose was obtained in in-situ SEM experiments where single shell failure was observed. Failure stress shows statistical behavior consistent with brittle failure and volume scaling. Large number of missing atoms, holes with radii of about 3-4 nm are required to explain the low strengths experimentally measured. Models based on van der Waals inter-shell interactions are insufficient to explain the multi-shell failure observed experimentally. Additional QM studies of mechanical and electronic states are needed.

32 In-situ TEM Testing of Carbon Nanotubes Before Straining low mag. mid. mag. high mag.

33 After Straining Walls disappeared?

34 High Resolution TEM

35 EELS Spectrum 6.E+4 5.E+4 4.E+4 3.E+4 2.E+4 1.E+4.E photon energy (ev)

36 Gold Nanowires HRTEM [11] [11] twin [111] nanotwin 2nm 2nm [11] growth direction Average diameter: 2±5 nm One grain through the thickness A large number of twins but no dislocations Dr. C. Li In collaboration with Dr. Hsien-Hau Wang, Materials Division, Argonne National Lab

37 Tensile Stress (GPa) Tensile Stress (GPa) Tensile Yield Stress (GPa) A B C 5nm 1nm 15nm 17.5nm MD Simulations of NWs [111] Nanowires Strain nm 1nm 15nm 17.5nm [11] Nanowires Strain Diameter (nm) <111> <11> Equilibrium Strain (%) [11] [111] 1 1 (A and B) Tensile stress responses for [111] and [11] axially oriented wires (as seen in Figure 1A and 1B, respectively) with representative radii from 5nm to 17.5nm. (C) Tensile yield stress as a function of diameter for all examined sizes. Both orientations show a drop in tensile yield stress with increasing diameter Diameter (nm) [111] [11]

38 Role of Surface Defects 2 E 2 2 = E 3 ε + ξε V 3 Step is created by a ½ <11> translation; perfect dislocation shift. Orientation <111> D=5 nm Strain Rate = 5x1 7 s -1 Change in Energy (ev)/atom σ y =7.35 GPa Change in Energy (ev)/atom σ 1 = Deformation Deformation

39 Acknowledgments: Y. Zhu, Northwestern University C-H. Ke, Northwestern University N. Moldovan, Northwestern University K-H. Kim, Northwestern University B. Peng, Northwestern University T. Belytschko, Northwestern University G. Schatz, Northwestern University C. Mirkin, Northwestern University Z. Bazant, Northwestern University D. Farkas, Virginia Tech. B. Hyde, NU N. Pugno, Politecnico di Torino P. Zapol, ANL O. Auciello, ANL I. Petrov, UIUC J. Mahon, UIUC M. Marshall, UIUC B. Prorok, Aurburn University National Science Foundation-NIRT, J. Larsen-Basse, K. Chong Federal Aviation Administration, J. Newcomb NSF-NSEC at NU for Integrated Nanopatterning and Detection Technology Office of Naval Research, L. Kabakoff