Sematech 3D Interconnect Metrology. 3D Magnetic Field Imaging Applied to a 2-Die Through-Silicon-Via Device

Size: px

Start display at page:

Download "Sematech 3D Interconnect Metrology. 3D Magnetic Field Imaging Applied to a 2-Die Through-Silicon-Via Device"

Leslie Hodges
6 years ago
Views:

1 Sematech 3D Interconnect Metrology 3D Magnetic Field Imaging Applied to a 2-Die Through-Silicon-Via Device Antonio Orozco R&D Manager/Scientist Neocera, LLC Fred Wellstood Professor Center for Nanophysics Department of Physics University of Maryland, College Park, MD July 11,

2 Overview Motivation. Magnetic Field Imaging Principles. Advances towards 3D Imaging. 2-Die TSV example. Acknowledgments and Q&A. 2

Motivation Various 3D Structures: Flip-Chip assemblies, 3D

Courtesy: Philippe Perdu, CNES Courtesy: Ingrid DeWolf, IMEC

metallization layers Increased number of opaque layers

Courtesy: Philippe Perdu, CNES New 3D interconnect approaches

3 Motivation Various 3D Structures: Flip-Chip assemblies, 3D cubes, Stacked dies (bonding or TSV). Courtesy: Philippe Perdu, CNES Courtesy: Ingrid DeWolf, IMEC NASA/Neocera Buried interconnections Multiple dies, metallization layers Increased number of opaque layers Interconnects integral part of defect localization. Courtesy: Philippe Perdu, CNES New 3D interconnect approaches complicates the FI/FA landscape. New technologies are required for: Isolate failure location in 3D Help development of new 3D technologies 3

representing the magnetic field component(s) intensity and direction.

4 Magnetic Field Imaging (I) Magnetic Field Imaging: Scanning technique using magnetic sensors to map magnetic field distributions over an area of interest resulting in a false color image representing the magnetic field component(s) intensity and direction. Sensor at different positions in magnetic field of a wire B Z 0 B-field vs. position X 3-D and 2-D representation of BZ for a wire 4

5 Magnetic Field Imaging (II) Circuit of interest in DUT is powered up (we use AC lock-in technique to filter out any magnetic field other than the one generated by the current in the DUT). DUT is raster scanned under the sensor. Magnetic field is recorded at regular intervals resulting in 2D map of magnetic field distribution. Magnetic field Bz B(x,y) Magnetic sensor Magnetic field image Current 5

Magnetic Field Imaging (IV) Magnetic filed data is used to determine the current density distribution that generated it by solving the magnetic field inverse problem.

6 Magnetic Field Imaging (IV) Magnetic filed data is used to determine the current density distribution that generated it by solving the magnetic field inverse problem. The end result is a map of the current density distribution in the DUT. B(x,y) FFT + constraints (*) Magnetic field image (*) Assumes no vertical current (all current is planar) 2D in nature. There is no unique sol. for the general 3D problem. J(x,y) Current density image 6

7 Magnetic Field Imaging (III) Requirements: High magnetic field sensitivity (detect weak currents deeply buried in the device) High resolution Accomplished using 2 sensor: Superconducting Quantum Interference Device (SQUID) to provide high field sensitivity Giant Magnetoresistance (GMR) spin valve device to provide submicron resolution. 7

8 Magnetic Field Imaging (III) SQUID GMR 8

8 Advances towards 3D imaging Magnetic field

Layering Information CAD layout 3D information is

9 8 Advances towards 3D imaging Magnetic field image Edge I Current density 3F Y y (mm) 6 4 C B D A Edge IV Magnetic Field Inversion 4F 2 Edge III G F E Defect 2F Edge II x X (mm) Layering Information CAD layout 3D information is embedded in data Need to develop capability to extract it 9

10 Advances towards 3D imaging Magnetic imaging Not restricted by packaging/device layers. Non destructive, non invasive, contact-less Maps currents deeply buried in the DUT. Natural candidate for 3D problems Need develop vertical/3d resolution. How Developing/applying innovations in magnetic field microscopy in 3 (inter-related) areas: Software/Algorithm Development. Extend conventional magnetic inversion. Develop novel algorithms. Use information about the sample. New Scanning Modes Develop new scanning techniques. Data Acquisition Improvements. Sensor Development. Lower noise, decrease scanning distance, detect different components. Sensor combination (SQUID, GMR, MTJ). 10

TSV diameter ~5microns, barrel height ~25 microns.

11 2-Die TSV example 2-Die Through-Silicon-Via device (Sematech) : Full bottom Si wafer carrier (750 microns). Top die thinned to 25 microns. 2 Metallization layers (400nm). TSV diameter ~5microns, barrel height ~25 microns. Passivation layer on top metallization (unknown thickness). Device showed multiple open circuit failures. Selected TSV via chain serpentine for scanning (red box). CAD Optical Image 11

12 2-Die TSV example Optical/current density and CAD/current density overlays. I = 330 A. CAD/current Optical/current Zoom in Note: missing segments (bottom) U shape at horizontal connecting segments. No short locations are visible. 12

13 2-Die TSV example Scan: 98um x 77um. Step size dx=dy= 250nm. I = 330 A Top segment Bottom segment CAD Current density U shape Current deviation Deformation of metal structures... or top/bottom die misalignment 13

14 2-Die TSV example High resolution scan: 98um x 77um. Step size dx=dy= 250nm. I = 330 A IR image High current IR/current overlay Low current 14

15 2-Die TSV example Forward simulation based on CAD using FEM. Also shifted top layer. CAD Matching top/bottom layers Solve region Top layer shift x=3 m y=-6 m 15

16 2-Die TSV example Magnetic field and current density comparison. Scan Simulation no shift Simulation shift 16

17 2-Die TSV example Magnetic field comparison. Scan Simulation no shift Simulation shift Cross-section magnetic field profile Adjusted for shift = scan = sim. no shift = sim. shift 17

18 Magnetic field comparison. 2-Die TSV example Top Bottom Top layer, scan Bottom layer, scan Top layer sim. shift Bottom layer, sim. shift Magnetic field simulation allows us to determine distance to sensor ~ 2 m and the relative shift to be x=3 m, y=6 m. Magnetic field analysis shows that the separation between metallization layers (top/bottom) is ~25 m. 18

19 Summary Magnetic Field Imaging detected a misalignment of about x=3 m, y=6 m on the top die. We computed 3D simulations of the structure with and without top layer shifted. For the proper shift values in the simulation, we get a perfect match between simulated and experimental data. Magnetic field analysis allowed us to determine the distance from the top metallization to the sensor to be ~2 m. This gives ~1 m for the passivation layer. Magnetic field intensity analysis allowed us also to estimate the separation between metallization layers to be ~25 m, in good agreement with the intended height of the TSV. Magnetic Field Imaging have the potential to be an effective tool at providing 3D current mapping and 3D Fault Isolation capability for complex 3D interconnected devices. 19

20 Andy Rudack, Sematech. H. Kwon, A. Jeffers, UMD Legal Stuff: Acknowledgments and Q&A J. Gaudestad, C. Rowlett, S. Garrahan, F. Cawthorne, V. Talanov, Neocera Supported by the Intelligence Advanced Research Projects Activity (IARPA) via Air Force Research Laboratory (AFRL) contract number FA C-7101 (*). (*) The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon. Disclaimer: The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, AFRL, or the U.S. 20 Government.

450mm and Moore s Law Advanced Packaging Challenges and the Impact of 3D

450mm and Moore s Law Advanced Packaging Challenges and the Impact of 3D Doug Anberg VP, Technical Marketing Ultratech SOKUDO Lithography Breakfast Forum July 10, 2013 Agenda Next Generation Technology