A superparamagnetic nanocomposite obtained by dispersing superparamagnetic magnetite nanoparticles in the epoxy SU-8 is used to fabricate microstructures by photolithography. The dispersion of the nanoparticles and the level of agglomerations are analyzed by optical microscopy, TEM (transmission electron microscope), SAXS (small-angle X-ray scattering), XDC (X-ray disc centrifuge) and XRD (X-ray diffraction). Two different phosphate-based dispersing agents are compared. In order to obtain a high-quality nanocomposite, the influence of particle concentration 1-10 vol.% (4-32 wt.%) on composite fabrication steps such as spin coating and UV exposure are systematically analyzed. Features with narrow widths (down to 1.3 m) are obtained for composites with 5 vol.% particle concentration. Mechanical, magnetic and wetting properties of the nanocomposites are characterized. These nanocomposites exhibit superparamagnetic properties with a saturation magnetization up to 27.9 kA m −1 for10 vol.%. All nanocomposites show no differences in surface polarity with respect to pure SU-8, and exhibit a moderate hydrophobic behavior (advancing dynamic contact angles approximately 81 • ). Microcantilevers with particle concentrations of 0-5 vol.% were successfully fabricated and were used to determine the dynamic Young's modulus of the composite. A slight increase of the Young's modulus with increased particle concentration from 4.1 GPa (pure SU-8) up to 5.1 GPa (for 5 vol.%) was observed.

We present a photocurable polymer composite with superparamagnetic characteristics for the fabrication of microcantilevers. Uniform distribution and low particle agglomeration (<50 nm) in the photocurable polymer matrix SU-8 are achieved by using superparamagnetic nanoparticles with a surfactant. Particles and composite are characterized by a transmission electron microscope, UV-VIS spectrometer and magnetic measurements. The composite contains 5 vol.% (18 wt.%) of Fe3O4 nanoparticles with diameters of 12.1 ± 3.5 nm. The composite exhibits a magnetization saturation of 13.2 kA m−1. Superparamagnetic composite microcantilevers with typical dimension of 2 µm × 14 µm × 80–300 µm are successfully fabricated by two conventional photolithography steps and a sacrificial layer etch. Exposure doses of 10 000 mJ cm−2 must be applied for microcantilever thicknesses of 1.8 µm due to the high UV absorption of the particles in the composite. The magnetic polymer cantilevers are successfully actuated in resonance in air with an amplitude of 29 nm. An off-chip coil is used to generate a magnetic field to actuate the cantilevers.

Heat dissipation from three-dimensional (3D) chip stacks can cause large thermal gradients due to the accumulation of dissipated heat and thermal interfaces from each integrated die. To reduce the overall thermal resistance and thereby the thermal gradients, this publication will provide an overview of several studies on the formation of sequential thermal underfills that result in percolation and quasi-areal thermal contacts between the filler particles in the composite material. The quasi-areal contacts are formed from nanoparticles self-assembled by capillary bridging, so-called necks. Thermal conductivities of up to 2.5 W/m K and 2.8 W/m K were demonstrated experimentally for the percolating and the neck-based underfills, respectively. This is a substantial improvement with respect to a state-of-the-art capillary thermal underfill (0.7 W/m K). Critical parameters in the formation of sequential thermal underfills will be discussed, such as the material choice and refinement, as well as the characteristics and limitations of the individual process steps. Guidelines are provided on dry versus wet filling of filler particles, the optimal bimodal nanosuspension formulation and matrix material feed, and the over-pressure cure to mitigate voids in the underfill during backfilling. Finally, the sequential filling process is successfully applied on microprocessor demonstrator modules, without any detectable sign of degradation after 1500 thermal cycles, as well as to a two-die chip stack. The morphology and performance of the novel underfills are further discussed, ranging from particle arrangements in the filler particle bed, to cracks formed in the necks. The thermal and mechanical performance is benchmarked with respect to the capillary thermal and mechanical underfills. Finally, the thermal improvements within a chip stack are discussed. An 8 - or 16-die chip stack can dissipate 46% and 65% more power with the optimized neck-based thermal underfill than with a state-of-the-artcapillary thermal underfill.

No abstract

Thermal underfills are crucial to support integration density scaling of future integrated circuit packages. Therefore, a sequential process using hierarchical self-assembly of micro- and nanoparticles is proposed to achieve percolating thermal underfills with enhanced particle contacts. The three main process steps hereby are assembly of filler particles by centrifugation, formation of nanoparticle necks by capillary bridging, and the backfilling of the porous structure with an unfilled capillary adhesive. Numerical simulations predicting trajectories and distributions of micron-sized particles dispensed into a rotating disk are presented. The trajectories exhibit a strong dependence on the particle size; thus in the case of polydisperse filler particles nonuniform particle beds may result. An efficient centrifugal disk design with spiral-like guiding structures is experimentally validated. Defect-free, percolating particle beds in confined space with fill fractions of 46 vol-% to 66 vol-%, i.e., close to the theoretical limit, are also presented. The self-assembly of nanoparticles, forming enhanced thermal contacts between the percolating filler particles, is discussed. Two consecutive evaporation patterns during the capillary bridging process were identified: 1) dendritic network growth and 2) collapse of capillary bridges. The concave neck topology could only be achieved at temperatures below the boiling point. An optimal evaporation temperature of 60°C with respect to in-plane uniformity and neck shape was identified. Existing thermal gradients normal to the cavity surface resulted in strongly asymmetric neck formation in the cavity. Hence, uniform heating in an oven is the preferred method to initiate evaporation. Two types of bimodal dielectric necks are demonstrated. Polystyrene acts as the adhesive between thermally conductive alumina particles to form mechanically stable dielectric necks after an annealing step at 140°C. Interstitial and core-shell necks are presented. Finally, a benchmark study was performed to compare the effective thermal conductivity of the percolating thermal underfill with and without necks with state-of-the-art capillary underfills. A close to fivefold improvement could be obtained for diamond filler particles with silver necks (3.8 W/m-K).

The current feed capability of typical flip chip electrical interconnects is constrained by the solder alloy, as it is more susceptible to electromigration than the copper used for the pads and wires. Hence, interconnects formed by copper only mitigate the electromigration risk and/or allow to increase the current limit of the all-copper interconnect. In this work, two methods to form all-copper flip chip interconnects at an annealing temperature of 250 °C are presented. The interconnects in the contact region between Cu pillars and Cu pads with a pitch down to 150 μm are formed by Cu nanoparticle self-assembly and sintering. In the first method, the entire gap between a Cu pillar chip and a substrate was filled with a Cu nano-suspension. The formation of capillary bridges during the evaporation of the dispersant directed the self-assembly of the nanoparticles towards the contact region between Cu pillars and Cu pad. In the second method, the Cu pillar chip was dipped into a film of the Cu nano-suspension, followed by a transfer, placement and release with a die bonder onto pads on a substrate. The annealing of the Cu nanoparticles is performed in both cases in a reducing formic acid atmosphere. The first method was more susceptible to the formation of shorts between pillars, whereas the second method resulted in electrical functional chip to substrate assemblies. Interconnects with a mean electrical resistance of 26 ± 3 mΩ and a shear strength ranging from 4.6 to 12.3 MPa were achieved. The sintered Cu nanoparticles bridged gaps up to 10 μm between copper pillars and pads, demonstrating the potential to apply the joint also on non-planar substrates. Nevertheless, imperfections such as voids and cracks are still present in the joints and need further process development, to improve the quality and process robustness further

No abstract

The reliability of microelectronic devices is largely dependent on electronic packaging, which includes heat removal. The appropriate packaging design therefore necessitates precise knowledge of the relevant material properties, including thermal resistance and thermal conductivity. Thin materials and high conductivity layers make their thermal characterization challenging. A steady state measurement technique is presented and evaluated with the purpose to characterize samples with a thermal resistance below 100 mm(2) K/W. It is based on the heat flow meter bar approach made up by two copper blocks and relies exclusively on temperature measurements from thermocouples. The importance of thermocouple calibration is emphasized in order to obtain accurate temperature readings. An in depth error analysis, based on Gaussian error propagation, is carried out. An error sensitivity analysis highlights the importance of the precise knowledge of the thermal interface materials required for the measurements. Reference measurements on Mo samples reveal a measurement uncertainty in the range of 5% and most accurate measurements are obtained at high heat fluxes. Measurement techniques for homogeneous bulk samples, layered materials, and protruding cavity samples are discussed. Ultimately, a comprehensive overview of a steady state thermal characterization technique is provided, evaluating the accuracy of sample measurements with thermal resistances well below state of the art setups. Accurate characterization of materials used in heat removal applications, such as electronic packaging, will enable more efficient designs and ultimately contribute to energy savings.