Bamboo Boundary Research Articles

Experimental study of electromigration at bamboo grain boundaries with a new test structure using the single-crystal aluminum interconnection

Electromigration (EM) voids in the bamboo structure interconnect were observed by a new test structure with single-crystal aluminum leads. The new test structure consists of a single grain connected to single-crystal aluminum leads formed by lateral-solid phase epitaxial growth (L-SPE). The grain was formed by suppressing L-SPE of the single-crystal aluminum leads. Since the void nucleation sites were confined to the grain boundaries, the voids were easily located and observed. In addition, the crystal orientation of single-crystal aluminum leads could be controlled by L-SPE, and so the analysis could be performed more accurately than using traditional test structures that have a series of grains with random crystal orientation. The accelerated EM test was carried out under ideal conditions similar to that of real devices, because the temperature gradients around the test site of the bamboo grain boundaries were negligible. In our preliminary experiment, a void was observed in the grain, located next to the positive voltage lead. This seems to be contradictory to general understandings, we think this is because of the grain boundary configuration difference and/or EM induced vacancy fluxes difference.

Further experiments on the bamboo boundary internal friction peaks in high-purity aluminium

The bamboo boundary internal friction (BB) peak observed in high-purity aluminium was shown to be connected with the existence of dislocation substructures nearby the bamboo boundary. It is shown that the BB peak is reduced in specimens quenched from high temperatures and changes when the quenched specimen is aged at various temperatures, and disappears when aged at lower temperatures. TEM observation made after internal friction measurement shows that the configuration of the dislocation substructures changes after quenching and aging treatments. The configurations responsible for the BB peak and for the complementary peak appearing at a much higher temperature (HT peak) are determined by comparing the results of internal friction measurement and TEM observation. The condition of appearance of the BB peak predicted by the four-parameter mechanical model is confirmed.

Mechanism of the bamboo boundary internal friction peak

At the beginning of the 1980s, an internal friction peak situated below the optimum temperature of the grain boundary peak of fine grained specimens (the Kê peak) was observed in high-purity aluminium consisting of bamboo crystals. This discovery has given rise to much controversy, and the opinion has been expressed that this peak is only the Kê peak shifted toward lower temperature, and is not a new peak. However, our systematic study shows that the behaviour of this peak is quite different from that of the Kê peak. Recently, the bamboo boundary internal friction peak and the Kê peak were observed separately on one internal friction curve in sheet specimens consisting of mixed grains. This shows that the bamboo boundary peak is a new peak. TEM observations showed that dislocation substructures exist near the bamboo boundaries. Accordingly, we propose that such substructures intersecting and interacting with the bamboo boundaries are dragged along with the bamboo boundary during the viscous sliding of the boundary, so that the sliding process is limited with the appearance of an internal friction peak. A four-parameter mechanical model is suggested for describing the relaxation behavior of the bamboo boundary peak. The mathematical formulations and the experimental conditions for the appearance of the bamboo boundary peak derived according to this model agree with experiments.

The bamboo boundary peak in an AlZr alloy

An inverted torsion pendulum is used to study the high temperature internal friction of Al-0.073 wt% Zr alloy. Two internal friction peaks well separated from each other, the bamboo boundary peak and the conventional grain boundary peak, are observed in the original internal friction curve when the size of some grains is larger than the diameter of the specimen. The height of the bamboo boundary peak increases while that of the grain boundary peak decreases with increasing number of bamboo-like grains. The bamboo boundary peak in the completely bamboo-structured specimen decreases linearly in height with increasing size of bamboo grains, until the peak is completely absent in the single crystal specimen. The peak temperature of the bamboo boundary peak and the associated activation energy are 690 K (fp = 0.29 Hz) and 3.20 eV when bamboo grains coexist with fine grains, and 635 K (fp = 0.31 Hz) and 2.74 eV when the complete bamboo structure is formed, respectively. Combining the electron microscopic observation of the second phase particles in boundaries, it is suggested that the bamboo boundary peak is originated by the stress relaxation of the particle-bearing boundary.

Electromigration induced transgranular slit failures in near bamboo Al and Al-2% Cu thin-film interconnects

We report the morphology of electromigration induced transgranular slit-like voids in Al and Al-2% Cu thin-film interconnects with near bamboo microstructures. The grain structure surrounding fatal voids was determined from images produced by scanning electron and focused ion beam microscopy. Fatal voids were seen to traverse grains near to, but apart from, bamboo boundaries or grain triple points, and often appear to emanate from small wedge-shaped voids at the interconnect edge. The small volume slit voids are consistent with an observed decrease in fatal void volume with reduced linewidth/grain size ratio. However, the transgranular void pathway represents a new failure mechanism responsible for reduced reliability in the narrow interconnects which are more relevant for advanced integrated circuit technologies. The conditions for stress-induced transgranular slit voiding are briefly discussed.

Experimental evidence on the existence of the bamboo boundary internal friction peak

Experimental evidence on the existence of the bamboo boundary internal friction peak

On the origin of the macrocrystalline internal-friction peak (the bamboo boundary peak)

The different responses to annealing temperature by the finegrained grain-boundary internal-friction peak (Kě peak) and the macrocrystalline peak (the bamboo-boundary peak) are re-confirmed. Critical analysis is made concerning the constraints limiting the relaxation process across the boundaries. Relevant experimental results are presented in answering the arguments raised by Iwasaki in a recent paper. Die unterschiedlichen Responses gegenüber der Ausheiltemperatur des feinkörnigen Korngrenzenpeaks der inneren Reibung (Kě-Peak) und des makrokristallinen Peaks (der Bambusgrenzepeak) werden erneut bestätigt. Eine kritische Analyse wird bezüglich der Bedingungen durchgeführt, die die Relaxationsprozesse über die Grenzen begrenzen. Relevante experimentelle Ergebnisse werden als Antwort gegenüber Argumenten von Iwasaki in einer neueren Arbeit vorgelegt.

Bamboo boundary internal-friction peak in 99.9999% aluminium and the effect of cold work on the peak

An internal friction peak is observed in zone-melt 99.9999% aluminium wire specimens with bamboo crystals. The height of the peak (Q) is found to be proportional to the number of bamboo boundaries N contained in the specimen with the straight line passing through the origin. The activation energy associated with this bamboo boundary peak is found to be 1.44 eV. These results confirm the previous findings in the case of 99.999% aluminium. A systematic study is made on the effect of cold deformation on the bamboo boundary peak. Results show that the internal friction peak previously observed in cold-worked 99.9999% aluminium by some authors and claimed to be due to lattice dislocations is actually a composite peak consisting of PH, PB, and PL, where PB is the bamboo boundary peak, PH and PL are two peaks originated from cold deformation and are situated at higher and lower temperatures than that of PB. Conditions for the appearance of PH and PL are discussed. In zonengezogenem 99,9999%-Aluminiumdrahtproben mit Bambuskristallen wird ein Maximum der inneren Reibung beobachtet. Die Peakhohe (Q) ist proportional zur Zahl der Bambusgrenzen in der Probe, deren gerade Linie durch den Ursprung gehen. Die mit diesem Bambusgrenzenpeak verbundene Aktivierungsenergie betragt 1,44 eV. Diese Ergebnisse bestatigen fruhere Ergebnisse im Falle des 99,999%-Aluminium. Der Einflus einer Kaltdeformation auf den Bambusgrenzepeak wird systematisch untersucht. Die Ergebnisse zeigen, das das Maximum der inneren Reibung, das fruher in gewalzten 99,9999%-Aluminium von einigen Autoren beobachtet und Gitterversetzungen zugeschrieben wird, tatsachlich ein zusammengesetztes Maximum ist, das aus PH, PB und PL besteht, wobei PB der Bambusgrenzenpeak ist und PH und PL zwei Maxima, die von der Kaltdeformation herruhren und bei hoheren und tieferen Temperaturen als die von PB auftreten. Die Bedingungen fur das Auftreten von PH und PL werden diskutiert.

On the bamboo boundary internal friction peak in 99.9999 aluminium

An internal friction peak has been observed in bamboo crystals of 99.9999wt% (6N) Al. The height of the peak was found to be proportional to the number of bamboo boundaries contained in the specimen. Analysis of the relaxation parameters of this peak shows that this is a new peak associated with the presence of bamboo boundaries.

Contribution of Bamboo Boundaries to the Internal Friction Peak in Macrocrystalline High-Purity Aluminium

en