Measuring and testing – Vibration – By mechanical waves
Reexamination Certificate
2000-08-25
2002-04-09
Williams, Hezron (Department: 2856)
Measuring and testing
Vibration
By mechanical waves
C073S024010, C216S059000
Reexamination Certificate
active
06367329
ABSTRACT:
The present invention relates generally to plasma systems and, more particularly, to a method for determining the endpoint of a plasma etch process.
BACKGROUND OF THE INVENTION
In many processes for microelectronic device fabrication, a pattern defined by a lithographic technique is transferred through a layer of material formed on the surface of a substrate. Typically, the pattern is transferred by etching using a plasma. The term plasma, as used in this disclosure, refers to a partially ionized gas consisting of positively and negatively charged molecular species, as well as neutrals.
Plasma etching processes are typically performed in an apparatus such as a plasma reactor. Plasma reactors generally include a reaction chamber, a plasma generating system, a wafer holder and handling system and a gas delivery system (i.e., inlet, exhaust, and flow control). The term reaction chamber, as used in this disclosure, refers to the area within a plasma reactor where ionized gases physically and/or chemically interact with a material layer formed on the surface of a substrate.
A cross-sectional view of an example of a plasma reactor, called a parallel plate reactor
10
is shown in FIG.
1
. Parallel plate reactor
10
includes two electrodes
11
,
12
positioned parallel to each other in a reaction chamber
14
. Substrates
15
with lithographically defined patterns (not shown) formed thereon are placed on the surface
12
a
of electrode
12
. In a typical etching process using a plasma reactor such as a parallel plate reactor
10
, gases are mixed and introduced into the reaction chamber
14
. The mixed gases flow between electrodes
11
,
12
. An electric field applied between electrodes
11
,
12
ionizes the gases and forms a plasma
13
. The plasma
13
then etches the layer of material (not shown) formed on the surface of substrates
15
and transfers the lithographically defined pattern therethrough.
A problem associated with plasma etching processes is a difficulty in determining when the etch step has been completed. This difficulty occurs because plasma techniques are typically timed processes, based on predetermined etch rates. The predetermined etch rates are identified by performing a calibration step. Since the exact conditions (i.e., pressure, gas flow, electric field) used during the calibration step are typically not duplicated for the etch step, timed processes are inaccurate and only provide an estimate as to when the plasma etch process is completed.
In order to avoid the use of timed processes for determining the endpoint of an etch step, diagnostic techniques have been developed which analyze the plasma in the reaction chamber. One such technique, called optical emission spectroscopy, monitors the intensity of the optical emission in the plasma. The intensity of the optical emission is related to the concentration of molecular species in the plasma. The completion of the etch process is determined when a change in the intensity of the optical emission is observed. A change in the intensity of the optical emission is observed when the concentration of molecular species in the plasma changes as a result of etching through the top layer and into the underlying substrate. Optical emission techniques require the reaction chamber to be equipped with an optical port for monitoring the optical emission of the plasma. Optical ports are not universally available in production environments, which limits the use of optical emission techniques for plasma etch endpoint detection.
Other diagnostic techniques such as laser interferometry, ellipsometry and mass spectrometry have been utilized in laboratory environments to identify the endpoint of an etch process. However, these techniques are both expensive and difficult to implement. In addition, optical ports are required to monitor the plasma etch process using laser interferometry and/or ellipsometry. While optical ports are not required to perform mass spectrometry, the detectors used for such techniques are placed in the reaction chamber and are often corroded by the etchants used for etching the material layers, limiting the ability of the detectors to accurately detect the completion of the etch step.
Accordingly, techniques useful for determining when an etch step is complete and which do not rely on the optical emissions of the plasma or direct optical access to the substrate in the reaction chamber and which are not corroded by the chemical gases used for such processes, are sought.
U.S. Pat. No. 5,877,407 assigned to the assignee of the present invention describes a method for determining the endpoint of a plasma etch process using acoustic cells. For the purpose of this description, the endpoint of the plasma etch process refers to when a first material layer formed on the surface of a substrate is etched through its thickness to its interface with an underlying material layer. The acoustic cell is configured to have a transmitter and a receiver located at opposite ends of a conduit. The transmitter and the receiver are acoustically matched transducers, which preferably operate in the kilohertz range, such as a transducer of lead-zirconate-titanate crystal. At least a portion of a gas stream from a reaction chamber of a plasma reactor flows through the acoustic cell during the plasma etch process with the pressure at which the gas stream flows in the acoustic cell preferably at least about 10 torr.
As the gas stream from the reaction chamber flows in the acoustic cell, acoustic signals are periodically transmitted from the transmitter to the receiver and the velocity of such acoustic signals is determined. The acoustic signals are periodically transmitted at intervals of at least about 20 hz (hertz), for the duration of the etch process. The acoustic signals, when transmitted, preferably travel a distance less than about 6 inches in the acoustic cell and are transmitted at a frequency within the range of about 50 kilohertz to about 500 kilohertz.
According to U.S. Pat. No. 5,877,407, the velocity of an acoustic signal is related to the average molecular weight of the gas stream according to the expression
v
s
=
γ
RT
M
(
1
)
where v
s
is the velocity of the acoustic signals, R is the universal gas constant (8.3143 J/mol K), T is the temperature in degrees Kelvin, M is the average molecular weight of the gas, and &ggr; is the ratio of the average specific heat at constant pressure to the average specific heat at constant volume (C
p
/C
v
). Thus, at constant temperatures, the velocity of an acoustic signal changes as the average molecular weight of the gas changes. For example, if the average molecular weight of the gas decreases, the velocity of acoustic signals transmitted through the gas increases.
As the first material layer formed on the surface of a substrate is etched, the average molecular weight of the gas flowing in the acoustic cell does not vary significantly. Thus, from equation (1), the velocity of the acoustic signals transmitted through the gas exhausted from the reaction chamber as the first material layer is etched approximates a constant value. For the purpose of this description, the acoustic signals transmitted through the gas exhausted from the reaction chamber when the first material layer formed on the surface of the substrate is etched, have a first velocity (or reference velocity).
When the first material layer is etched through its thickness to its interface with the underlying material layer, the average molecular weight of the gas flowing into the acoustic cell changes. The change in the average molecular weight of the gas changes the speed at which the acoustic signals are transmitted through the gas in the acoustic cell. Thus, the first velocity determined for the acoustic signals transmitted through the gas as the first material layer is etched changes to a second velocity associated with reaching the interface of the underlying material layer. The second velocity for the acoustic signals differs from the first velocity by more than 1%. The etch process endpoi
Cadet Gardy
Reitman Edward Alois
Agere Systems Guardian Corp.
Beusee, Brownlee, Bowdoin & Wolter, P.A.
Beusse, Esq. James H.
Saint-Surin Jacques M
Williams Hezron
LandOfFree
Acoustic time of flight and acoustic resonance methods for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Acoustic time of flight and acoustic resonance methods for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Acoustic time of flight and acoustic resonance methods for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2898842