Semiconductor device manufacturing: process – Chemical etching – Vapor phase etching
Reexamination Certificate
2002-06-14
2003-12-30
Lebentritt, Michael S. (Department: 2824)
Semiconductor device manufacturing: process
Chemical etching
Vapor phase etching
C438S950000, C438S952000, C438S737000, C438S003000
Reexamination Certificate
active
06670280
ABSTRACT:
BACKGROUND OF THE INVENTION
The invention relates to microstructuring ferroelectric materials, more especially but not exclusively to methods of etching ferroelectric materials and to methods of frustrating etching in ferroelectric materials.
Ferroelectric materials are of great interest to the optics, telecommunications and laser communities. Some of these materials, particularly lithium niobate, have large electro-optic and non-linear optical coefficients, and hence have extensive application in piezoelectric, acousto-optic, pyroelectric and photorefractive devices. This remarkable range of electrical and optical phenomena has been widely exploited in a number of technological applications.
However, many applications depend upon the ability to micro-machine ferroelectric crystals by the selective removal of material. These range from relief grating patterning, where sub-micron periods are required, to bulk micro-machining for improved modulator performance, fabrication of photonic band gap devices, or use in micro-electro-mechanical systems and micro opto-electro-mechanical systems. Conventional techniques for microstructuring lithium niobate include direct laser machining (pulsed laser ablation), and a range of etching procedures. The technique of wet-etching, in which combinations of acids are used to remove surface material from crystals, is extremely hard to localize without attendant undercutting, and compromise of feature integrity. Etching also requires additional precision mask-making or photolithographic steps, and this has cost and time implications.
It has been shown that in iron-doped lithium niobate (Fe:LiNbO
3
), wet-etching with a solution of HF and HNO
3
acids can be resisted, or frustrated, by illuminating the surface of the Fe:LiNbO
3
during etching with continuous wave (CW) 488 nm visible radiation [1, 2]. It has been shown that the CW 488 nm radiation can reduce or completely inhibit the etching behavior of the material. This etch frustration effect can be used to structure the surface of the material by selective exposure of parts of the surface during etching.
The precise surface science behind this etch frustration effect is not known. It has however been established from experiments that the effect does not exist with undoped LiNbO
3
. It appears only to function with Fe:LiNbO
3
. A degree of latency in the effect with a time constant of a few hours has also been observed [2]. In these observations, the Fe:LiNbO
3
was immersed in etchant for a period of five hours and exposed to the illumination only for a one hour period at different stages over the five hours. These experiments show that the etch frustration effect is not triggered until illumination is started, but continues for some hours after illumination has been removed. The effect is believed to arise from absorption of 488 nm radiation by the Fe dopant atoms, which generates mobile charge carriers in the form of electrons. The visible light makes charge move to the crystal surface. The presence of this charge rich surface layer disables the surface etch chemistry while the Fe:LiNbO
3
surface is immersed in the etchant acid solution of HF and HNO
3
.
A further etch frustration effect has also been reported in which Fe:LiNbO
3
is irradiated by pulsed 248 nm ultraviolet (UV) radiation and subsequently etched (see ref [2] FIG.
4
(
b
) and text at page 2793, right hand column).
Although the etch frustration effect in Fe:LiNbO
3
is of interest for holographic storage applications, the presence of Fe is undesirable in most if not all other applications, since Fe increases the photorefractive response by providing a source of extrinsic charge (electrons usually). Light excites the Fe via the Fe
2+
→Fe
3+
+e reaction. The electrons liberated then move by diffusion and drift, and a charge pattern is set up. This makes the crystal host highly affected by light, which is desirable for holographic storage, but undesirable for any devices that exploit electro-optic, piezoelectric, or nonlinear responses, for example.
An etch enhancement effect has been reported for pure LiNbO
3
and Ti:LiNbO
3
[3]. In this work, pulsed UV radiation at 248 nm is used in a first step to selectively irradiate the +z surface of the LiNbO
3
. The sample is then etched in a second step using an etchant solution of HF and HNO
3
. The etchant preferentially attacks the irradiated areas. The method is said to be crystal-cut insensitive, with fabrication of surface relief gratings being reported on both x-cut and z-cut LiNbO
3
. However, a limitation of this effect is that the etch enhancement only appears to allow shallow structures to be etched, of the order of 50-100 nm in depth. The effect is thus not capable of forming general micro-structuring, for example it is not capable of creating mesa structures, trenches, or ridge waveguides on the micron scale.
LiNbO
3
and other ferroelectrics etch very differently on the +z face compared to the −z face. The −z face etches relatively rapidly and deeply (0.6 to 0.8 &mgr;m per hour at room temperature for LiNbO
3
in HF and HNO
3
) while, as mentioned above for LiNbO
3
, the +z face only etches to nanometer depths. Therefore, etching of the −z face can produce far more useful structures than etching of the +z face.
The difference in etching behavior for the two faces may be attributable to a variety of causes. For example, the bonds between the metal-oxygen ions are shorter on the +z face, so are likely to be more stable and less reactive than the longer bonds on the −z face. Therefore the +z face will resist etching so that only shallow features can be formed by the +z face etch enhancement technique.
This differential etching has been utilized in a reported method of surface structuring [4]. A ferroelectric wafer is poled to form a distribution of +z and −z areas corresponding to the desired surface structure. The poled wafer is then etched, and the −z areas are removed by the etchant to a greater depth than the +z areas. The poling process requires the application of a photoresist mask to the wafer. Moreover, the wafer needs to be ‘pre-poled’ before poling to initially arrange all the domains in the same orientation.
SUMMARY OF THE INVENTION
A first aspect of the present invention is directed to a method of structuring a surface of a sample of ferroelectric material, the method comprising:
a) taking a sample of ferroelectric material having a −z face which is to be etched;
b) illuminating the −z face with ultraviolet light to define illuminated and unilluminated parts of the surface; and
c) immersing the −z face in an etchant to selectively remove the unilluminated parts of the −z face at a greater rate than the illuminated parts.
The method allows the frustration of wet-etching techniques to be used to apply high quality micron and sub-micron scale patterns and structures to the surfaces of samples of ferroelectric materials, without the requirement for the illumination to be done during the wet-etching process or that Fe be present in the material. This flexibility is advantageous because it allows for the illumination and etching steps to performed at different times and/or at different locations, to suit the requirements of the user. It arises from the fact that the frustration effect is retained by the ferroelectric material after the illumination ceases, so that the wet-etching does not need to be performed contemporaneously with the illumination. However, it will be understood that the etching may also be carried out during the ultraviolet illumination in other embodiments.
The method may be carried out with ferroelectric material that is not doped with iron, thus allowing widespread application of the method. Specifically, the ferroelectric material may contain less than 0.01 wt % of iron (or lower values such as 0.008, 0.006, 0.004, 0.002, 0.001, or 0.0001 wt %), whereas inclusion of at least
Brown Paul
Eason Robert William
Mailis Sakellaris
Finnegan Henderson Farabow Garrett & Dunner L.L.P.
Lebentritt Michael S.
University of Southampton
LandOfFree
Methods of microstructuring ferroelectric materials does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods of microstructuring ferroelectric materials, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods of microstructuring ferroelectric materials will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3108732