High frequency transducer for ultrasonic bonding

xxiii, 271 leaves : ill. (some col.) ; 30 cm.PolyU Library Call No.: [THS] LG51 .H577P AP 2001 OrStudy of the fabrication, material properties, and resonance characteristics of lead zirconate titanate (PZT)/epoxy 1-3 piezocomposite rings and investigation of the design, fabrication, and evaluation of ultrasonic transducers for high frequency wire bonding in microelectronic packaging based on these composite rings were presented and discussed in this thesis. Investigation of the properties of the constituent materials - PZT and epoxy - was performed. Conclusions and suggestions for future work were also included. Original contributions reported in this research included: 1) Hard PZT piezoceramic ring, PKI804, and epoxy, Araldite LY5210/HY2954, were selected as the active and passive phases in the 1-3 structure, respectively. Their material properties were measured so as to cover more parameters for later use in the modeling of composites and transducers. These constituent materials were fabricated into ring-shaped homogeneous 1-3 composites with PZT volume fractions Ø ranging from 0.82 to 0.94 and with a small epoxy width of 77 um. The technique for making composite rings was devised. 2) An existing parallel model was applied to predict the material properties of the composite rings as a function of Ø. This model was extended to include the prediction of dielectric loss tangent tan δe and mechanical quality factor Qm. Good agreement was obtained between the theoretical and experimental results in the new range of Ø. It was revealed that not all the parameters pertinent to the transducers could be optimized at the same time, so trade-offs must be made. An optimization of composite properties against different values of Ø for transducer design was discussed. 3) The resonance characteristics of the composite rings were investigated by classifying the resonance modes into two main categories, namely the longitudinal-thickness mode fH and lateral mode fL₁ and fL₂ resonances of the individual PZT elements inside the rings as well as the radial mode fR, wall-thickness mode fW, and stopband fS₁ and fS₂ resonances of the whole rings. fH was observed to increase linearly with the decrease in element height while fL₁ and fL₂ remained constant. When the height and width of the elements became comparable, coupling of fH with fL₁ and fL₂ occurred. The observed fH, fL₁, and fL₂ coincided with those calculated by the mode coupling theory. fR and fW, were almost independent of the ring thickness but increased as Ø increased. fS1 and fS2 were undetectable. A guide of operating fH in the rings without causing mode coupling was presented to optimize the composite structure for transducer design.4) Physical models for designing piezoelectric drivers and ultrasonic horns were consolidated based on the one-dimensional elastic wave theory. A computer program for solving these design models was complied in order to reduce the design effort. The conceptual design of a 138 kHz composite (Ø = 0.89) transducer was generated together with a PZT (PKI804) transducer of similar structure. These transducers were about 50 % shorter than the commercial designs for facilitating high-speed bonding and comprised minimal number of components for assuring good quality of manufacturing. 5) The natural frequencies and vibration mode shapes of both transducers were computed using finite element analysis (FEA). The computed working frequency and displacement distributions were in good agreement with those deduced by the elastic wave theory. The non-axial modes in the composite transducer were significantly weaker. Their natures were mainly governed by the radial and wall-thickness activities of the piezoelectric rings. A reliable design procedure was proposed to use the physical models for exploring a design possibility and to apply the FEA for detailed behavior study and/or design optimization. 6) The technique used to build a high quality ultrasonic bonding transducer was disclosed, whereby the design prototypes were fabricated. These transducers were evaluated by measuring their electrical and vibrational characteristics as well as by process-testing their bonding capabilities. The measurements were compared to the computational (FEA) and theoretical results with satisfactory agreement. Due to the low mode coupling behavior of the 1-3 composite, the non-axial and many other spurious resonances in the composite transducer were suppressed, retaining only the axial mode resonances. The purity of axial vibration excitation in the composite transducer was 9 % higher than its PZT counterparts. Moreover, Qm of the composite transducer was 2.4 times smaller because of damping in the epoxy matrix, which resulted in a rise and a fall time of about 50 % shorter. Furthermore, the composite transducer was easier to assemble owing to its relatively lower stiffness. In addition, it possessed compatible characteristics under high power driving. The process-test results showed that the use of the composite transducer on an ultrasonic wire bonder had widened its operating window, thereby promoting the fine-pitch and high-speed capabilities in terms of higher bonding yield. A number of patents and publications (please refer to the list of publications) were produced during the course of this work, further elucidating the originality and practical application of the present study.Department of Applied PhysicsPh.D., Dept. of Applied Physics, The Hong Kong Polytechnic University, 200