APPLICATION OF MICROMACHINING TECHNIQUES FOR FABRY-PEROT CAVITY BASED MICROSENSORS
by
YOUNGMIN KIM, B.S.E.E., M.S.E.E.
Copyright
by
Youngmin Kim
1995
Approved By
Dissertation Committee:
_________________________________
Dean P. Neikirk
_________________________________
Ilene J. Busch-Vishniac
_________________________________
Alvin B. Buckman
_________________________________
Ashley J. Welch
_________________________________
Sanjay K. Banerjee
DISSERTATION
Presented to the Faculty of the Graduate School of
The University of Texas at Austin
in Partial Fulfillment
of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
THE UNIVERSITY OF TEXAS AT AUSTIN
August 1995
Supervisor: Dean P. Neikirk
Abstract
A micromachined Fabry-Perot cavity based sensor has been studied. The study consists of a discussion on a new fabrication technique for the sensor and a manufacturability design study of the sensor. The new fabrication technique allows a Fabry-Perot cavity sensor to be monolithically fabricated without using a wafer bonding process. Using this technique high manufacturing yield and better performance should be possible. The Fabry-Perot cavity sensor, combined with a multimode optical fiber, was also used to measure differential pressure. Measured optical response of the cavity has been compared with the simulated response that takes into account the averaging effect caused by the shape of the deflected mirror.
In addition to the development of a new fabrication technique, a design
study for manufacture of Fabry-Perot cavity sensors has been performed where
thickness variation of layers in the cavity exists. An analytic method has
been developed which can efficiently calculate the variation of optical
response due to thickness variation of layers. From the calculation, uncertainty
in cavity gap due to process-induced variations has been obtained. This
uncertainty bounds the accuracy of manufactured sensors. As another error
source, fitting function-induced errors were also included in the calculation
for the accuracy of the sensors. To find an optimum design, accuracy contour
maps have been generated over the design space, i.e., initial cavity gap
and mechanical travel of a moving mirror. Through this study it is shown
that there exists an optimum design which gives high yield with a specific
level of performance.
List of Tables x
List of Figures xi
Chapter 2 Optical characteristics of Fabry-Perot cavity sensors 6
2.1 Wave propagation through multiple layers 6
2.2 Optical response of a Fabry-Perot cavity pressure sensor 12
2.3 Summary 18
Chapter 3 Design for manufacture of Fabry-Perot cavity sensors 19
3.1 Impact of thickness variations on optical response 20
3.2 Design methodology and case study for manufacture 24
3.2.1 Design Methodology 24
3.2.2 Example Designs 30
3.3 Summary 43
Chapter 4 Deflection of multiple thin film diaphragms 44
4.1 Mechanical properties of multiple film stacks 44
4.2 Deflection of multiple film stacks 48
4.3 Design issues for mechanical compliance 53
4.4 Summary 55
Chapter 5 Fabrication of micromachined Fabry-Perot cavity pressure sensors 56
5.1 Overview of Micromachining techniques 56
5.1.1 Bulk micromachining techniques 58
5.1.2 Surface micromachining techniques 63
5.2 Process for Fabry-Perot cavity sensor 68
5.3 Summary 82
6.1 Transmittance measurements in free space 83
6.2 Pressure measurement in free space 87
6.3 Pressure measurements using optical fiber interconnect 94
6.4 Summary 101
Appendix A Deflection of a Plate105
Appendix B Photolithography 107
Appendix C Wafer cleaning techniques 109
Bibliography 111
VITA 120
List of Tables
Table 5.1
Mechanical and electrical properties of a single crystal silicon
(from ref. [1, 2]) . 57
List of Figures
Figure 2.1
Schematic diagram of a multilayer Fabry-Perot sensor. 8
Figure 2.2
Refractive indices of dielectric materials commonly used in silicon process (from ref. [2]). 9
Figure 2.3
Schematic view and transmittance of a Fabry-Perot cavity with a polysilicon spacer. Both top and bottom dielectric film stacks consist of two 1000 Å silicon nitride layers cladding a 1400 Å silicon dioxide layer. 11
Figure 2.4
Cross section of a Fabry-Perot cavity with a silicon diaphragm. 13
Figure 2.5
Refractive index of gold versus wavelength (from ref. [2]). 14
Figure 2.6
Calculated reflectance of a Fabry-Perot cavity sensor with metal (Au) mirrors as a function of cavity gap. 15
Figure 2.7
Schematic cross sectional view of a Fabry-Perot cavity sensor with dielectric mirrors. 16
Figure 2.8
Calculated reflectance of the sensor versus cavity gap, assuming an illumination wavelength of 700 nm. 17
Figure 3.1
The first derivatives of the reflectance of a Fabry-Perot cavity with metal (Au) mirrors, as shown in Figure 2.5. The vertical axis represents the first derivative of the reflectance of the cavity with respect to the thickness of each layer. 22
Figure 3.2
The first derivatives of the reflectance of a Fabry-Perot cavity with dielectric films, as shown in Figure 2.6. For simplicity, only the four layers with the largest derivatives are shown. 23
Figure 3.3
Reflectance and process-induced response variations of a Fabry-Perot cavity with Au mirrors for a single wavelength (lambda = 700 nm). Solid line: reflectance; dotted line: bound on gap uncertainty Dgproc . 31
Figure 3.4
Reflectance and process-induced response variations of a Fabry-Perot cavity with dielectric mirrors for a single wavelength (lambda = 700 nm). Solid line: reflectance; dotted line: bound on gap uncertainty Dgproc . 32
Figure 3.5
Contour map over design space (i.e. initial gap and mechanical travel) when only process-induced thickness variations are considered. The numbers on the contour lines represent the accuracy of corresponding designs. 34
Figure 3.6
Accuracy contour map, including only linear fitting errors, for single wavelength detection. Numbers on lines represent the accuracy of the contour. 35
Figure 3.7
Accuracy contour map, including both mirror variations and linear fitting errors, for single wavelength detection. 36
Figure 3.8
Response curve and process-induced variations of a Fabry-Perot cavity with metal mirrors for dual wavelengths (560nm and 700nm). 38
Figure 3.9
Accuracy contour maps including only process-induced errors for two branches with best performance. 40
Figure 3.10
Accuracy contour maps including linear fitting errors for two branches with best performance. 41
Figure 3.11
Accuracy contour maps including both process-induced errors and linear fitting errors for two branches with best performance. 42
Figure 4.1
Residual stress measurement of deposited films using substrate curvature method. A composite film consists of silicon nitride (2000 Å) and silicon dioxide (1500 Å). 47
Figure 4.2
Dependency of a diaphragm deflection at center on residual stress of the diaphragm. 51
Figure 4.3
Normalized deflection of diaphragm as a function of residual stress. (a) without residual stress; (b) with residual stress ( ). Numbers on the contour lines represent normalized deflected height to deflection at center (origin in maps). 53
Figure 5.1
Cross sectional view of an anisotropically etched silicon substrate. Etch mask film could be either silicon dioxide or silicon nitride. 59
Figure 5.2
Basic procedure of surface micromachining technique. 65
Figure 5.3
A buckled silicon dioxide diaphragm prepared with low temperature LPCVD. (a) SEM photograph of the top view ; (b) cross section. 67
Figure 5.4
Schematic process flow for monolithically integrated Fabry-Perot pressure sensor. 70
Figure 5.5
Anisotropically etched (100) silicon substrate with a circular pattern. u is undercut obtained using equation (5.2). 74
Figure 5.6
SEM photograph of a Fabry-Perot cavity suffering from sticking. The top diaphragm collapsed onto the bottom diaphragm and the substrate. The interference pattern indicates bending of the top diaphragm. 78
Figure 5.7
SEM photograph of a cross section of a complete Fabry-Perot cavity pressure sensor. Note that the deposited films for the top mirror have excellent step coverage and a flat surface after release from the substrate. 79
Figure 5.8
Process flow chart for micromachined Fabry-Perot pressure sensor. 80
Figure 6.1
Transmittance measurement setup in free space. 84
Figure 6.2
Transmittance of a Fabry-Perot cavity with dielectric films mirrors and an air gap. Dotted line: measurement; solid line: simulation. 86
Figure 6.3
Schematic view of a device holder for transmittance measurement. The holder allows measurement of transmitted light while applying pressure to a Fabry-Perot cavity. 88
Figure 6.4
Transmittance of the Fabry-Perot cavity as a function of applied differential pressure. Solid line: simulation; dotted line: measurement. 90
Figure 6.5
Deflection of a square diaphragm without built-in stress. The deflection is normalized to the peak value. 91
Figure 6.6
Normalized deflection contour map of the bottom diaphragm with residual stress of 0.3 GPa. 93
Figure 6.7
Experimental setup for reflectance measurement with an optical fiber interconnect. 95
Figure 6.8
View of a Fabry-Perot cavity coupled to an optical fiber. (a) photograph of top view of the cavity; (b) cross section of the cavity coupled to an optical fiber. 97
Figure 6.9
Measurement of reflectance and transmittance of the Fabry-Perot cavity coupled to an optical fiber. Black dot: Transmittance; open dot: Reflectance. 98
Figure 6.10
A model for a Fabry-Perot cavity coupled to an optical fiber. A deflection of the bottom diaphragm causes changes in both d1 and d2, simultaneously. 99
Figure 6.11
Reflectance of a Fabry-Perot sensor as a function of applied pressure. Solid line: simulation ; dotted line: measurement. 100