The Experts below are selected from a list of 18333 Experts worldwide ranked by ideXlab platform
Mitsumasa Koyanagi - One of the best experts on this subject based on the ideXlab platform.
-
deteriorated device characteristics in 3d lsi caused by distorted Silicon Lattice
IEEE Transactions on Electron Devices, 2014Co-Authors: Murugesan Mariappan, Yasuhiko Imai, Shigeru Kimura, Takafumi Fukushima, J C Bea, Hisashi Kino, Kangwook Lee, Tetsu Tanaka, Mitsumasa KoyanagiAbstract:Silicon-Lattice distortion in the 50- μm-thick stacked large scale integrated circuit (LSI) chip over Cu-Sn μ-bumps was studied by synchrotron-assisted micro-X-ray diffraction. The top and bottom surfaces of the upper chip experienced 0.25% and 0.1% tensile strain (equivalent to 450 and 200 MPa of tensile stress), respectively. Si [004] plane showed a maximum tilt value of +0.45° and -0.25°, respectively, over the μ-bump and in the bump-space region. Raman spectroscopy revealed that upper stacked chip experienced ~ 1000 MPa of tensile stress and ~ -200 MPa of compressive stress, respectively, over the μ-bump and bump-space regions. Distorted Si-Lattice in 3D-LSIs caused 4% and 12% change in ON-current characteristic for n- and p-MOSFET devices, respectively.
-
revisiting the Silicon Lattice in the high density 3d lsis in the perspective of device reliability
International Electron Devices Meeting, 2013Co-Authors: M Murugesan, Takafumi Fukushima, J C Bea, Kangwook Lee, Tetsu Tanaka, Mitsumasa KoyanagiAbstract:The dependence of device reliability on the Lattice perfectness of the active Silicon in the high-density 3D-LSIs containing through-Silicon via (TSV) and micro-bump (μ-bump) is extensively investigated using hard X rays at SPring8. The reciprocal Lattice space (RLS) data revealed that the Si-Lattice structure is highly deteriorated owing to the thermo-mechanical (TM) stress exerted by Cu-TSVs and CuSn μ-bumps, and the local mechanical (LM) stress caused by local deformation. The TM stress caused by 20 μm-width Cu-TSV at 300 °C has introduced (i) ~3 degrees of Lattice-tilt (mis-orientation) and (ii) ~8.3 % reduction in Lattice space (d) values for Si(004) Lattice planes in the 3D-LSI chip. This d change has caused a maximum strain of -0.96 %, which corresponds to -1300 MPa of compressive stress. After the curing, the locally deformed upper thin LSI die with 30 μm thickness witnessed as high as 4.9 % increase in d value, and the Lattice tilt amount to 0.65 degree. More importantly, the lower 300 μm-thick active/passive interposer has also experienced the Lattice tilt and the change in d to the magnitude of around 0.2 degree and 0.4 %, respectively. We have also observed a degradation in the retention time for the stacked memory chip with a decrease in the chip thickness. The median retention time in the 30 μm-thick DRAM-chip was reduced to one-half the retention period for the 100 μm-thick DRAM chip. We explain this phenomenon by deteriorated Young's modulus values and distorted Lattice structures in the ultra-thin LSI Si chip. We were able to minimize the TM stress in the active Si to one-third from that of the initial value by sandwiching an organic stress-absorbing polymer between the dielectric layer and the Ta barrier layer, and the polymer is stable up to 400°C.
P Becker - One of the best experts on this subject based on the ideXlab platform.
-
the Silicon Lattice parameter an invariant quantity of nature
Metrologia, 1998Co-Authors: J Martin, U Kuetgens, J Stumpel, P BeckerAbstract:The Lattice parameter of Silicon plays an important role in the determination of the Avogadro constant and the fine-structure constant. Today, three values of the d220 Silicon Lattice spacing are available, measured at the Physikalisch-Technische Bundesanstalt (PTB, Germany), the Istituto di Metrologia G. Colonnetti (IMGC, Italy) and the National Research Laboratory of Metrology (NRLM, Japan) and based on the metre scale. Using the PTB Lattice spacing comparator, the spacings of the different materials were measured and compared with one another in order to check the possibility of combining the results to form a common best value. The characterization of the relevant impurities and crystal defects and their number densities are discussed and corrections applied for their influences on the interatomic distances. The NRLM data are in disagreement with those of the PTB and the IMGC, from which a Lattice parameter for a hypothetical, ideal, perfect Si crystal was derived with a relative uncertainty of about .
P R Wilshaw - One of the best experts on this subject based on the ideXlab platform.
-
the effect of impurity induced Lattice strain and fermi level position on low temperature oxygen diffusion in Silicon
Journal of Applied Physics, 2011Co-Authors: Zhidan Zeng, John D Murphy, Robert J Falster, Deren Yang, P R WilshawAbstract:Oxygen diffusion in Silicon is known to be affected by high concentrations of impurities, although the mechanism underpinning this is poorly understood. We have studied oxygen transport in Czochralski Silicon by analyzing data on the locking of dislocations by oxygen as a function of time and temperature. In this paper, we present new data from crystals grown to contain high levels of germanium and arsenic. We analyze these new data, together with our previous data for Silicon with a high boron concentration, to further the understanding of the mechanism by which high impurity concentrations affect oxygen transport at temperatures at which the oxygen dimer dominates transport (up to 550 °C). Our results show that a high level of boron doping (∼3 × 1018cm−3) enhances the effective diffusivity of oxygen by a factor of ∼8 to ∼25 relative to low doped material with the same oxygen concentration. High levels of germanium doping (∼8 × 1019cm−3) and arsenic doping (∼2 × 1019cm−3) can both have a slight retardation effect on oxygen transport. The magnitude of the reduction measured is less than a factor of ∼4 in the heavily germanium doped specimens and less than a factor of ∼5 in the heavily arsenic doped specimens, and in most cases is significantly less than this. Germanium doping introduces considerable strain into the Silicon Lattice without affecting the Fermi level position, so data from these samples show that Lattice strain affects oxygen dimer transport. The arsenic and boron doping levels in the materials studied give rise to Lattice strain with a smaller magnitude and opposite sign to that in the germanium doped samples. It is therefore suggested that the Fermi level position also affects the transport of oxygen dimers.
R Feidenhansl - One of the best experts on this subject based on the ideXlab platform.
-
structure determination of the indium induced si 111 4 1 reconstruction by surface x ray diffraction
Physical Review B, 1999Co-Authors: Oliver Bunk, G Falkenberg, J H Zeysing, L Lottermoser, R L Johnson, M Nielsen, F Bergrasmussen, Jeff Baker, R FeidenhanslAbstract:A detailed structural model for the indium-induced $\mathrm{Si}(111)\ensuremath{-}(4\ifmmode\times\else\texttimes\fi{}1)$ surface reconstruction has been determined by analyzing an extensive set of x-ray-diffraction data recorded with monochromatic $(\ensuremath{\Elzxh}\ensuremath{\omega}=9.1\mathrm{keV})$ synchrotron radiation. The reconstruction is quasi-one-dimensional. The main features in the structure are chains of Silicon atoms alternating with zigzag chains of indium atoms on top of an essentially unperturbed Silicon Lattice. The indium coverage corresponds to one monolayer. The structural model consistently explains all previously published experimental data.
S D Kevan - One of the best experts on this subject based on the ideXlab platform.
-
indium square root 7 x square root 3 on si 111 a nearly free electron metal in two dimensions
Physical Review Letters, 2003Co-Authors: Eli Rotenberg, H Koh, K Rossnagel, Han Woong Yeom, J Schafer, B Krenzer, M Rocha, S D KevanAbstract:We present measurements of the Fermi surface and underlying band structure of a single layer of indium on Si(111) with $\sqrt{7}\ifmmode\times\else\texttimes\fi{}\sqrt{3}$ periodicity. Electrons from both indium valence electrons and Silicon dangling bonds contribute to a nearly free, two-dimensional metal on a pseudo-4-fold Lattice, which is almost completely decoupled at the Fermi level from the underlying hexagonal Silicon Lattice. The mean free path inferred from our data is quite long, suggesting the system might be a suitable model for studying the ground state of two-dimensional metals.