Influence of Cu Seed Deposition Temperature on Electroplated Cu Texture Formation in Damascene Structures

1999 ◽  
Vol 562 ◽  
Author(s):  
Qing-Tang Jiang ◽  
Robert Mikkola ◽  
Richard Ortega ◽  
Volker Blaschke

ABSTRACTThe deposition temperature of the PVD Cu seed layer has a critical impact on the subsequent electroplated Cu film. Sheet resistance transformation of electroplated Cu on 50°C seeded Cu was more than twice faster than on 150°C seeded Cu. X-ray pole figure analysis on a 3 mm × 3 mm dense array of 0.35 μm Cu damascene lines at spacing of 0.4 μm revealed significant grain orientation differences between directions parallel and perpendicular to the Cu lines. It was observed that for both seed process temperatures, the (111) pole figures showed a sharper texture parallel to the trench direction than to the perpendicular direction. After annealing at 450°C for 30 minutes, a (511) secondary orientation emerged and the (111) texture along the trench direction became even sharper. Perpendicular to the trench, the (111) texture split into two peaks after anneal, exhibiting near fiber texture. The deviation of the two splits from normal was 2.4° tilt towards the trench wall indicating strong interaction between trench sidewall and electrodeposited Cu inside the trench. The (111) pole figure analysis also revealed a 20% contribution of Cu growth directly from the side walls. Although the pole figure pattern of 150°C seeded sample resembles that of the 50°C seeded sample, the (111) fiber texture of a 50°C seeded sample was always stronger and sharper.

1999 ◽  
Vol 564 ◽  
Author(s):  
Qing-Tang Jiang ◽  
Robert Mikkola ◽  
Richard Ortega ◽  
Volker Blaschke

AbstractThe deposition temperature of the PVD Cu seed layer has a critical impact on the subsequent electroplated Cu film. Sheet resistance transformation of electroplated Cu on 50°C seeded Cu was more than twice faster than on 150°C seeded Cu. X-ray pole figure analysis on a 3 mm × 3 mm dense array of 0.35 μm Cu damascene lines at spacing of 0.4 μm revealed significant grain orientation differences between directions parallel and perpendicular to the Cu lines. It was observed that for both seed process temperatures, the (II I) pole figures showed a sharper texture parallel to the trench direction than to the perpendicular direction. After annealing at 450°C for 30 minutes, a (511) secondary orientation emerged and the (111) texture along the trench direction became even sharper. Perpendicular to the trench, the (111) texture split into two peaks after anneal, exhibiting near fiber texture. The deviation of the two splits from normal was 2.40 tilt towards the trench wall indicating strong interaction between trench sidewall and electrodeposited Cu inside the trench. The (111) pole figure analysis also revealed a 20% contribution of Cu growth directly from the side walls. Although the pole figure pattern of 150°C seeded sample resembles that of the 50°C seeded sample, the (111) fiber texture of a 50°C seeded sample was always stronger and sharper.


1957 ◽  
Vol 1 ◽  
pp. 131-142
Author(s):  
Otto Renius

AbstractWork at the Detroit Arsenal has shown that techniques similar to those employed for the determination of pole figures of metals can be utilized for studying organic materials such a a stretched rubber latex. The rubber, when stretched, forms a preferred orientation pattern which is proportional in intensity to the degree of elongation, and which can be used to plot a pole figure.A Geiger-counter spectrometer was used to study samples of rubber stretched 600 to 1000 per cent. Using a transmission technique, the specimens were tilted to the impinging X-ray beam in five degree increments while rotating through 360 degrees to allow the measurement of the diffracted beam from the selected atomic planes at various angles within the specimen. The intensities of the diffracted beam at these angles were plotted on a stereographic net to form the pole figures of the (002) and (012) planes of the stretched rubber. The geometry of the sample arrangements permitted the outer portion of the pole figure to be plotted from alpha angle 0 degrees to alpha angle 45 degrees.


1989 ◽  
Vol 10 (2) ◽  
pp. 117-134 ◽  
Author(s):  
R. Shimizu ◽  
J. Harase ◽  
K. Ohta

In an attempt to investigate the accuracy of the vector method for crystal texture analysis, a comparison has been made between the inversion result of the pole figure made by X-ray studies using the VM and the inversion result of the pole figure made by ECP. A comparison has been made between the inversion by the pole figure generated by direct measurement of orientations by ECP and the actual orientation distribution (measured by ECP) displayed in the same mode. The materials studied were recrystallized Fe–3% Si and Fe–50% Ni. The main findings were:• In the mean intensities of each individual Box, the inversion results of pole figures made from orientations determined by ECP were in good agreement with the inversion from (100) pole figures made by X-ray or actual orientation distribution (made by ECP) displayed in the same mode as the vector method.• For Fe–3% Si, quite a good agreement was obtained between the results inverted from X-ray pole figure and the direct measurement by ECP for the intensity distribution of minor texture component along ζ angle. It was concluded from these investigations that the inversion of the pole figure by the vector method is accurate enough for most practical purposes.


2008 ◽  
Vol 23 (2) ◽  
pp. 87-91 ◽  
Author(s):  
Anthony C. Rizzie ◽  
Thomas R. Watkins ◽  
E. Andrew Payzant

A practical description of the mathematics required to implement the hexagonal grid and spiral trace pole figure data collection schemes is presented. Applying the concepts of stereographic and equal area projections with geometry, spreadsheets were created to calculate the angular settings of the goniometer. Using the generated settings, the hexagonal grid and spiral trace schemes were programmed into the existing X-ray software and employed to collect data for a sample of aluminum foil. The resulting (111) pole figures were similar to those collected with the conventional 5°χ×5°ϕ grid. The hexagonal grid has been shown by others to reduce the number of data points and time needed to complete a pole figure, while providing equal area sampling. Although not optimized, the spiral method was also investigated as another alternative to the 5°χ×5°ϕ grid.


1968 ◽  
Vol 12 ◽  
pp. 391-403 ◽  
Author(s):  
Hung-Chi Chao

AbstractThe texture of sheet metal Is best described, by means of pole figures, which are very expensive and time-consuming to prepare. About 8 to 12 hours of effort by a specially trained, and. highly skilled technician are needed to prepare each pole figure. Accordingly, pole figures are not used as extensively in research studies as they would, be if they could be obtained more easily.A method has been developed for automatically producing pole figures by printing results directly from a digital computer. This method does not require the use of additional plotting attachments and, is therefore less expensive and time consuming than other methods. With this method, any laboratory with access to a digital computer can produce pole figures automatically.X-ray diffraction intensities are recorded on punched tape or on punched cards and are fed into the digital computer. A computer program corrects X-ray data obtained, by either transmission or reflection X-ray techniques, maps the stereographic projection, and prints pole figures directly. The time required, to prepare an accurate pole figure is reduced from 8 to 12 hours to 20 minutes or less depending on the type of digital computer used.


1986 ◽  
Vol 30 ◽  
pp. 421-427
Author(s):  
J.J. Heizmann ◽  
C. Laruelle ◽  
A. Vadon

At the moment, the time required to obtain a pole figure is about three hours, and generally several pole figures are needed to make a texture analysis. Therefore the time and the cost of texture measurement are widely increased.Now, new position sensitive X-Ray detectors are appearing, which allow to record at one and the same time the whole 2Θ spectrum, i.e. the beams diffracted by several (hi k1l1) lattice planes. So with this kind of detector, it will be possible to get simultaneously several pole figures.


2009 ◽  
Vol 1156 ◽  
Author(s):  
Conal E. Murray ◽  
Paul R. Besser ◽  
Christian Witt ◽  
Jean L. Jordan-Sweet

AbstractGlancing-incidence X-ray diffraction (GIXRD) has been applied to the investigation of depth-dependent stress distributions within electroplated Cu films due to overlying capping layers. 0.65 μm thick Cu films plated on conventional barrier and seed layers received a CVD SiCxNyHz cap, an electrolessly-deposited CoWP layer, or a CoWP layer followed by a SiCxNyHz cap. GIXRD and conventional X-ray diffraction measurements revealed that strain gradients were created in Cu films possessing a SiCxNyHz cap, where a greater in-plane tensile stress was generated near the film / cap interface. The constraint imposed by the SiCxNyHz layer during cooling from the cap deposition temperature led to an increase in the in-plane stress of approximately 180 MPa from the value measured in the bulk Cu. However, Cu films possessing a CoWP cap without a SiCxNyHz layer did not exhibit depth-dependent stress distributions. Because the CoWP capping deposition temperature was much lower than that employed in SiCxNyHz deposition, the Cu experienced elastic deformation during the capping process. Cross-sectional transmission electron microscopy indicated that the top surface of the Cu films exhibited extrusions near grain boundaries for the samples undergoing the thermal excursion during SiCxNyHz deposition. The conformal nature of these caps confirmed that the morphological changes of the Cu film surface occurred prior to capping and are a consequence of the thermal excursions associated with cap deposition.


2005 ◽  
Vol 495-497 ◽  
pp. 1693-0 ◽  
Author(s):  
Vladimir Serebryany ◽  
Sergey F. Kurtasov ◽  
T.I. Savyolova

Influences of an X-ray experimental texture plan and of pole figure primary errors on the accuracy of the ODF coefficients are discussed. A modified harmonic method, which is a statistical method of ridge estimates, has been used for ODF restitution from pole figures. The plan of measurement of the pole figures was analysed on the basis of minimization of the primary experimental errors and the errors of the ODF restitution method. Principles of the optimum texture plan construction for hexagonal metals (magnesium example) and for the URD-6 texture diffractometer are considered.


2002 ◽  
Vol 721 ◽  
Author(s):  
K. Barmak ◽  
A. Gungor ◽  
A. D. Rollett ◽  
C. Cabral ◽  
J. M. E. Harper

AbstractAnnealing of dilute binary Cu(Ti), Cu(In), Cu(Al), Cu(Sn), Cu(Mg), Cu(Nb), Cu(B), Cu(Co) and Cu(Ag) alloy films resulted in the strongest <111> fiber texture for Cu(Ti) and the lowest resistivity for Cu(Ag). The behavior of the alloy films was compared and contrasted with that for a pure evaporated Cu film. Electron beam evaporated films with compositions in the range of 2.0-4.2 at% and thicknesses in the range of 420-560 nm were annealed at 400°C for 5 hours. Two different approaches were used to derive volume fractions of texture components, namely fiber plots and orientation distributions. It is argued that for polytextured films such as the copper alloys studied here, orientation distributions derived from pole figures provide the most reliable basis for quantitative characterization.


2000 ◽  
Vol 34 (1) ◽  
pp. 55-62 ◽  
Author(s):  
H. Siemes ◽  
C. A. Rosière ◽  
P. Hackspacher ◽  
W. Schäfer ◽  
E. Jansen

A rather simple empirical defocusing correction method is described which makes use of X-ray and neutron measurement of pole figure data from a sample with a weak preferred orientation.


Sign in / Sign up

Export Citation Format

Share Document