Claudio Zanelli(1)†, Dushyanth Giridhar(1), Manish Keswani(2), Nagaya Okada(3), Jyhwei Hsu(4), Petrie Yam(1) (1)Onda Corporation, (2)University of Arizona, (3)Honda Electronics, (4)Suss MicroTec † Correspondence: et@ondacorp.com

Acoustic Characterization of Two Megasonic Devices for Photomask Cleaning

RESULTS & DISCUSSION

METHODS

CONCLUSIONS Although the frequency and generator power settings are equivalent, the acoustic performance of the nozzle and cone transducers is significantly different. Clearly frequency and electrical power alone are not the only determinants of acoustic performance and the subsequent cleaning activity. For the same input power, the direct field pressure output from the nozzle transducer is approximately 10 times greater than that from the cone transducer, certainly because of its smaller footprint. It is also observed that in both designs the pressure from cavitation is about two orders of magnitude lower than the direct pressures. So, what physical mechanism from each megasonic transducer is cleaning? The results presented here indicate a significantly higher level of direct field pressure than stable and transient cavitation pressure. However, the test conditions may not fully represent actual cleaning processes leaving this still an open question. Schlieren imaging highlights the complex behavior of the sound waves propagating between the mask, transducer, and water surface. For instance, it shows the nozzle transducer jet angle affects the interference pattern. Imaging the cone transducer indicates that the incident wave propagates at an offset angle from transducer. The resultant sound field reveals a complex pattern from multiple reflections, yielding a “scrubbing” mechanism at both 1 and 3 MHz.

Direct Field and Cavitation Pressure

Stable Cavitation (PS) Direct Field (P0) Cavitation vs. Power

AIMS III Automated Scanning Tank

EMDS-USB Motion Controller

MCT-2000 Cavitation Meter

Computer

Driving Electronics

f0: 970 kHz P0: 41 kPa PS: 28 kPa PT: 14 kPa

Harmonic f = 1.940 MHz

Sub-Harmonic f = 485 kHz

Ultra-Harmonic f = 1.455 MHz

Aco

ustic

Pre

ssur

e A

mpl

itude

[kP

a]

Frequency [MHz]

Scanning Setup

1 MHz Xdcr

θ2

θ1

Z Y X

Hydrophone / Mask Sensor

Absorbing Liner

Fundamental f = 970 kHz

INTRODUCTION Wet photomask cleaning relies on megasonic agitation to enhance the process, but there are many challenges to reliably control performance in terms of particle removal efficiency (PRE) and damage. Typically all that is specified about the acoustic output is the driving frequency and the electric power delivered to a transducer. Although there are several studies on particle removal and pattern damage at 1 MHz, there is very little known about the cavitation performance at higher frequencies such as 2-4 MHz. Two transducer configurations are acoustically evaluated here to better understand how the acoustic waves interact with the substrate which is the basis to optimizing cleaning performance.

Transducers

Sensors

3 MHz Nozzle (Honda: W-357-3MP)

3 MHz Cone (Honda: W-357-3MQB-SKC)

Needle Hydrophone (Onda: HNA-0400)

Mask Sensor Array (Onda)

Computer

Digital Camera

Schlieren Imaging Setup

OptiSon

Xdcr

Quartz Mask

Liner

Laser

transducer

Driving Electronics

2D Sound Pressure Fields

Hydrophone in a Free Field

Mask Sensor in a Standing

Wave Field

3W Setting, 2.48W Measured 12W Setting, 8.95W Measured

Incident wave propagates at an off-angle from transducer; standing waves pattern consistent at 3 MHz

Incident wave disturbed by reflected wave from both top and bottom surface of quartz; some waves transmit through the quartz mask

The resultant sound field reveal a complex pattern from multiple reflections; similar scaled pattern from both 1 and 3 MHz.

Schlieren Imaging (Videos: HERE)

Cone 1 MHz

15 mm

Cone 3 MHz

10 mm

10 mm

Nozzle 3 MHz

10 deg 12 mm

Nozzle 3 MHz 0 deg

12 & 28 mm

10 mm

10 mm

10 mm

Incident wave is launched

Reflections create interference pattern, even

at an angle of 10 deg

Interference pattern at 12 mm distance from transducer

Interference pattern at 28 mm, indicating a

larger footprint

Nozzle

Mask Scanning Away from

Transducer

Hydrophone in a Free Field

Mask Sensor

Needle Hydrophone

3 MHz Transducer

Mask Sensor

Needle Hydrophone

3 MHz Transducer

Nozzle: Pressure vs. Position from Transducer

Cone

Hydrophone Scanning

Along Mask Surface

• Direct field pressure is ≥ 10X higher than cavitation pressure for both nozzle and cone

• 3 MHz nozzle yields ~3X higher stable cavitation pressure than 3 MHz cone at generator powers of 1 and 2 W

• Low levels of transient cavitation pressure under all conditions

Acoustic Pressure

vs.

Generator

Power

Nozzle & Cone: Varying Generator Power

Standing waves observed from both sensors; reduction in P0 as mask moved

away from transducer

• Pressure reduced by 8X at 30 mm from beam

• Water surface height affects P0

• Minimal level of cavitation pressure.

San Jose, CA September 12-14, 2016