Ultra High Thermal Conductivity Nanowire Filled Polymer Composites And Interfaces

Transcription

1 Ultra High Thermal Conductivity Nanowire Filled Polymer Composites And Interfaces Indira Seshadri Graduate Research Assistant Department of Materials Science and Engineering Rensselaer Polytechnic Institute, Troy, NY, USA Advisors : Dr. Ganpati Ramanath and Dr. Theo Borca-Tasciuc Collaborators : N. Balachander, R. J. Mehta, G. Esquenazi, L. Schadler, P. Keblinski Sponsors: IBM Gift Grant, Rensselaer Nanotechnology Center, NSF CMMI Rensselaer Nanotechnology Center Research Symposium Wednesday, November 6, 2013

2 2 Thermal Interfaces Effective interface heat transfer Device efficiency Total thermal conductance - G t t G c,1 G b G c,2 G t = 1 G b + 1G c,1 + 1G c,2 1 G b Thermal conductivity k Contact thermal conductance G c Low elastic modulus E, compliant relieves thermal mismatch = k t

3 3 Current thermal interface materials Commercial high k thermal interface materials Greases Phase change materials Elastomers Adhesives k = 2-5 Wm -1 K -1 (Filler loading φ vol. %) High φ Loss of compliance Polymer Nanocomposites filled with Carbon fillers CNT, graphene, graphite k < 2 Wm -1 K -1 Metal nanofillers flakes, particles and fibers of Ag, Cu, Ni

4 4 Our approach Nanostructured networks for high k Nanostructured networks Connected high thermal conductivity pathways through percolation at low filler loading φ Metallic nanowire filled polymers High aspect ratio (= length/diameter) metal nanowires. Percolation filler fraction φ percolation α 1/ aspect ratio. Low φ percolation High k + Low E (compliant)

5 5 Ultrathin ( < 10 nm diameter) flexible gold nanowires Nanowires: 6-8 nm diameter, µm length Aspect ratio ~ 100 Percolation at φ = 1/100 = 1 vol. % nanowire loading Bent, looped nanowires Cold welded junctions* *Y. Lu et. al., Nat. Nanotechnol. 5, (2010).

6 6 Nanowire polymer composite preparation Polymer matrix : Crosslinked (cured) polydimethylsiloxane Compliant low elastic modulus (E < 5 MPa) silicone Chemically inert

7 7 Gold nanowire-filled PDMS composite Before mixing PDMS Heavily intertwined nanowire networks Minimal change in nanowire diameter/length after polymer dispersion After mixing PDMS

8 8 Low aspect ratio gold nanorods Synthesized for understanding effect of nanofiller shape 50 nm diameter, nm length nanorods + nanoparticles No connectivity or flexibility

9 Thermal conductivity - Au NW-PDMS Theoretical models k PDMSAu decreasing Aligned connected fillers Upper bound Predicted φ percolation Best match! Random well connected fillers ~ 0 R K k PDMSAu = 5 Wm -1 K -1 ~ 32x k PDMS Non linear increase, φ > φ percolaion = 1 vol. % Theoretical modeling indicates a percolating, well connected random nanowire network 9 Random poorly connected fillers Finite R K

10 10 Extraordinary enhancement in k! η : Enhancement in k NW Nanowires NR - Nanorods η PDMSAuNW > 6 η graphene, CNT > 10 η Ag, η Au-NR High aspect ratio ( Low percolation threshold) Not sufficient for high k! Shahil & Balandin, Nano Lett. 12, (2012). Razeeb et.al, 59 th Electronic Components and Technology Conference, (2009). Biercuk et.al, Appl. Phys. Lett. 80, (2002). Connectivity (small R K ) is critical!! 10 nm diameter Cold welded contacts in Au nanowires

11 11 Au-PDMS electrical conductivity Remains insulating ~ ohms resistance Short breaks in nanowire network Heat propagates through polymer,current cannot.

12 12 Au Nanowire PDMS Mechanical properties E Au = 70 GPa E PDMS ~ 3 MPa Near constancy in E & H Measured values Theoretical values for flexible nanowires Cold welded Metal nanowire network 32x k enhancement without affecting compliance

13 13 Nanocomposite metal thermal contact conductance (G c ) Thermal contact conductance G c ~30% of interface G t G t = 1 G b + 1G c,1 + 1G c,2 1 P ~ few kpa BOTH G c and k for G t G c => Phonon mismatch Conformance, Contact pressure G c,1 G c,2 G b t Cu substrates Apply uncured nanocomposite Cure and measure G c

14 14 Au-PDMS / Cu contact thermal conductance G c G c k AuPDMS Thermal contact conductance G c with thermal conductivity k AuPDMS Sharp change near φ percolation ~ 1 vol. % Pure PDMS G c ~ 140 kwm -2 K -1 (R c = 7 mm 2 KW -1 ) 4 % Au PDMS G c ~ 1.5 kwm -2 K -1 (R c = 670 mm 2 KW -1 )

15 Au/PDMS Cu G c theoretical modeling Partial wetting model G c = 2k1k k + k A A real nomin al R a Solid contact model G c = 0.95 m 2k1k2 P 1.25 R a k 1 + k 2 E Experiment G c matches wetting model for φ AuPDMS < 1 % solid contact model for φ AuPDMS > 1 % No wetting beyond nanofiller percolation threshold? 15

16 Au-PDMS pre-cure flow characterization Pre-cure AuPDMS η Complex viscosity M s Storage (Elastic) modulus M L Loss (Viscous) modulus 1 η = ω M S M L Modulii cross over M S > M L, for φ AuPDMS > 1% x viscosity increase Solid like behavior PRIOR to curing (φ AuPDMS > 1% ) Nanowire percolation + welding induces pre-cure gelation Inhibits flow, Low G c 16

17 17 Further work : Increase G c through controlled welding Microwave induced welding of > 30 nm diameter silver nanowires 100 nm Unwelded junction 200 nm Microwave-welded junctions Wires weld in solution Possible to weld in polymer during cure Allow pre-cure flow High k + High G c

18 18 Summary Demonstrated novel nanowire network polymer nanocomposite Ultra high k without affecting compliance Key understanding : High interconnectivity + percolation Large k enhancements at small φ! Cold welding sub-10-nm diameter, high aspect ratio High k ~5 Wm -1 K -1, at φ < 5 % low E - compliant ~32x k enhancement > poorly connected nanorods, nanowires, CNT, graphene. Nanowire connectivity induces gelation inhibits flow and conformal interfaces, low interface conductance G c Next steps : Controlled welding to increase k without affecting flow.

19 Acknowledgements Ramanath Group Hafez Fard, Zepu Wang, Kamyar Pashayi RPI Micro-nano fabrication laboratory IBM Gift Grant, Rensselaer Nanotechnology Center NSF CMMI Thank you! Questions? 19