Process for monitoring the thickness of layers in a microelectronic device

Information

  • Patent Grant
  • 6605482
  • Patent Number
    6,605,482
  • Date Filed
    Thursday, October 11, 2001
    23 years ago
  • Date Issued
    Tuesday, August 12, 2003
    21 years ago
Abstract
A method of determining the thickness of a thickness of a first layer of material in a semiconductor device using a reflectometer, the first layer of material being disposed outwardly from a second layer of material, the first and second layer of material both including silicon. The method includes generating at least one predicted behavior curve associated with a depth profile of an interface between the first and second layer of material, the predicted behavior curve including at least one expected optical measurement, the depth profile associated with the interface being present at a particular theoretical depth. The method also includes emitting light onto a surface of the semiconductor device. The method further includes collecting at least one optical measurement from portions of the emitted light that are reflected by the semiconductor device. The method additionally includes comparing the at least one optical measurement to the predicted behavior curve and determining the approximate actual depth of the interface in response to the compared optical measurement.
Description




TECHNICAL FIELD OF THE INVENTION




This invention relates in general to semiconductor processing, and more particularly to a process for monitoring the thickness of layers in a semiconductor device.




BACKGROUND OF THE INVENTION




Metrology of layers of semiconductor material that are formed of the same material of which the immediate underlying layer of semiconductor material is also formed is difficult because both layers display many of the same or similar physical properties. For example, no contrast between n-doped silicon or p-doped silicon and intrinsic silicon can be observed by scanning electron microscopy, which is one of the more common ways to determine layer thickness.




Present calibration methods utilized in such circumstances include destructive techniques such as profilometry, which uses a physical probe to measure a height differential, and transmission electron microscopy, which transmits electrons directly through wafers. Such methods may lead to wafer contamination, physical destruction of wafers, non-uniformity in testing, and/or excessive allocations of cost and time. Thus, optical techniques for determining semiconductor layer thickness are preferable alternatives. However, current optical techniques, such as fourier transform infrared spectroscopy (FTIR) using, for example, a BioRad instrument, or reflectometry using a ThermaWave instrument, for example, measure a thickness change in one semiconductor layer overlying a similar semiconductor layer.




For example, FTIR may seek to measure the thickness of an epitaxial silicon layer formed over a silicon substrate using optical detection of a change in dopant ion concentration. One problem with using FTIR to measure the thickness of particular layers of epitaxial silicon is that smaller thicknesses of silicon epitaxy are likely to be beneath the detection limit of the FTIR technique.




SUMMARY OF THE INVENTION




In accordance with the present invention, a process for monitoring the thickness of layers in a microelectronic device is provided that substantially eliminates or reduces disadvantages and problems associated with previous developed systems and methods.




In one embodiment of the present invention, a method is presented for determining the thickness of a thickness of a first layer of material in a semiconductor device using a reflectometer, the first layer of material being disposed outwardly from a second layer of material, the first and second layer of material both including silicon. The method includes generating at least one predicted behavior curve associated with a depth profile of an interface between the first and second layer of material, the predicted behavior curve including at least one expected optical measurement, the depth profile associated with the interface being present at a particular theoretical depth. The method also includes emitting light onto a surface of the semiconductor device. The method further includes collecting at least one optical measurement from portions of the emitted light that are reflected by the semiconductor device. The method additionally includes comparing the at least one optical measurement to the predicted behavior curve and determining the approximate actual depth of the interface in response to the compared optical measurement.




In another embodiment of the present invention, a method is presented for monitoring a thickness of a first layer of material in a semiconductor device using a reflectometer, the first layer of material being disposed outwardly from a second layer of material, the first and second layer of material both comprising silicon.




One advantage of the present invention is that it presents an improved process for monitoring the thickness of semiconductor layers that addresses disadvantages of present monitoring processes. An additional advantage of various embodiments of the present invention is that the thickness of a semiconductor layer of material may be achieved without destroying or contaminating wafers of semiconductor material. A further advantage of various embodiments of the present invention is that a process is presented for monitoring the thickness of material that can be performed on-line with semiconductor device manufacture. Yet another advantage of various embodiments of the present invention is that a process is presented for monitoring the thickness of semiconductor layers and material that allows the thickness of one semiconductor layer to be measured even when such semiconductor layer is formed on a second semiconductor layer having the same or a similar material composition.











BRIEF DESCRIPTION OF THE DRAWINGS




For a more complete understanding of the present invention and its advantages, reference is now made to the following description taken in conjunction with the accompanying drawings in which:





FIGS. 1A through 1C

are schematic diagrams of a reflectometer and a cross-sectional view of a processing step in the creation of a semiconductor device illustrating one embodiment of the present invention;





FIG. 2

is a flow chart illustrating the operation of the reflectometer to measure the thickness of materials according to another embodiment of the present invention;





FIGS. 3A through 3C

are diagrams representing the optical characteristics of a semiconductor device at a particular point in semiconductor processing; and





FIG. 4

is a flowchart of a particular embodiment of the operation illustrated in FIG.


2


.











DETAILED DESCRIPTION OF THE INVENTION





FIGS. 1A-C

illustrate the use of a reflectometer such as ellipsometer


10


to monitor the thickness of layers of a semiconductor device


20


. In particular, the reflectometer in the described embodiments of

FIGS. 1A-1C

is used to detect the thickness of a un-doped layer of material that is formed over a doped layer of material when both layers of material have the same or similar compositions.




In

FIG. 1A

, Semiconductor device


20


includes an un-doped substrate layer


22


and a doped substrate layer


24


. Un-doped layer


22


is a monocrystalline silicon substrate. Doped layer


24


is an amorphous silicon layer amorphized because of damage to the crystalline structure of silicon caused by the implantation of either p- or n-type dopant ions. Un-doped layer


22


and doped layer


24


may have different optical characteristics, such as the magnitude of index of refraction, for example, because of such damage to the crystalline structure of doped layer


24


. Similarly, optical characteristics between un-doped layer


22


and doped layer


24


may vary because of the introduction of the dopant ions into the silicon comprising doped layer


24


. Although layers


22


and


24


are identified as being un-doped and doped layers of a silicon substrate, layers


22


and


24


may be any layers composed of the same or similar semiconductor material, such that the monitoring or detection of the thickness of doped layer


24


is desirable using ellipsometer


10


.





FIG. 1A

also illustrates ellipsometer


10


positioned over semiconductor device


20


such that light emitted by ellipsometer


10


may strike the surface of semiconductor device


20


and such that light reflected from semiconductor device


20


can be detected by ellipsometer


10


. Ellipsometer


10


is a spectroscopic ellipsometer; however ellipsometer


10


may be any suitable ellipsometer or other reflectometer including an ellipsometer using one or more wavelengths of light. Many varieties of ellipsometer


10


are well known in the semiconductor processing industry, and may include any suitable number of light emission sources, wave guides, polarized or unpolarized lenses, light sensors, and processing components.




In operation, ellipsometer


10


directs beams of light at the surface of doped layer


24


at a suitable angle and intensity such that some photons from the beams of light are absorbed by semiconductor device


20


and some photons are reflected at different depths within semiconductor device


20


for detection by the detectors of ellipsometer


10


. Ellipsometer


10


compares data perceived from the reflected photons in order to analyze differences at various depths of semiconductor device


20


. In particular, ellipsometer


10


may examine the intensity and phases of light detected by ellipsometer


10


for each of one or more wavelengths of light.




Ellipsometer


10


may compare the intensity and phase data for a particular wavelength of light to known characteristics and models for: a given semiconductor material, a specific semiconductor material processed in a particular manner, or a specific interface known to form between two layers of the same semiconductor material after particular semiconductor processing techniques have been utilized. With regard to semiconductor device


20


, models may be used that focus on changes in the optical properties of doped, amorphous silicon and un-doped, crystalline silicon. Ellipsometer


10


may compare data associated with the intensity and phase of light reflected by semiconductor device


20


to such models in order to determine the thickness of the amorphized silicon. A more detailed description of the operation of ellipsometer


10


and the modeling of such semiconductor material, processes, and interfaces is described more particularly with regard to the flow chart of FIG.


2


.





FIG. 1B

illustrates the anneal of doped layer


24


. The amorphous silicon comprising doped layer


24


is annealed using a thermal anneal process. Such an anneal heals the crystalline lattice structure of the silicon. Importantly, the optical properties, such as the index of refraction and absorption coefficient, for example, of annealed doped layer


24


may closely resemble that of un-doped layer


22


. However, the anneal of doped layer


24


generally does not result in the full anneal of doped layer


24


exactly at the interface between doped layer


24


and un-doped layer


22


. Thus, a defect layer


26


of un-annealed doped layer


24


may remain between the annealed portion of doped layer


24


and un-doped layer


22


. Alternatively, even if the anneal is relatively complete so as not to introduce optical changes caused by un-annealed silicon, the uneven interface between the crystalline structures of annealed doped layer


24


and un-doped layer


22


may provide a change in optical properties along such an interface. The presence of defect layer


26


and/or the uneven interface between layers


22


and


24


may be utilized by ellipsometer


10


to determine the thickness of annealed doped layer


24


using modeling of the optical properties of such defect layer


26


and/or such uneven interface. Such a determination may be made using the process described in reference to FIG.


2


.





FIG. 1C

illustrates the formation of a silicon epitaxial layer


28


outwardly from the annealed doped layer


24


. Silicon epitaxial layer


28


may be formed using a suitable chemical vapor deposition process. For example, a chemical vapor deposition process may be used at a pressure of 40 torr, at a temperature of 850 degrees Celsius, and utilizing a flow rate of 24 slm H


2


, 0.2 slm dichlorosilane, and 0.13 slm HCl. Formation of silicon epitaxial layer


28


may result in the formation of an interfacial layer


30


between the silicon epitaxial layer and the underlying doped layer


24


. Even if interfacial layer


30


is not formed, differences in the index of refraction and absorption coefficient may result between un-doped silicon epitaxial layer


28


and doped layer


24


. Either the presence of interfacial layer


30


and/or differences between doped and un-doped silicon may be used by ellipsometer


10


to determine the thickness of silicon epitaxial layer


28


using modeling of the optical properties of interfacial layer


30


and/or such differences between doped and un-doped silicon. Such a determination may be made using the process described in reference to FIG.


2


.





FIG. 2

is a flow chart illustrating the operation of the ellipsometer


10


in order to measure the thickness of semiconductor material in semiconductor device


20


.




In step


210


, differences in the optical constants embodied by the refractive index ‘n’ and the absorption or extinction coefficient ‘k’ that are determined along interfaces between two different materials. Such differences may occur because of difference in material composition, roughness along an interface, the presence of dopants, gradients in material composition or dopant concentration, defects or damage to one or both of the layers such as end-of-range damage, the formation of defect or interfacial layers, and/or other differences brought about by the processing of the two materials (collectively referred to hereafter as “process factors”). Changes in ‘n’ and ‘k’ may be determined based on such process factors using relationships between the optical behavior of materials and standard ellipsometric equations. Such relationships and ellipsometric equations are described in general in the book “Optics,” Volume 4, written by Arnold Summerfeld and published by the Academic Press in 1949.




In step


220


, a desired thickness of an overlying layer of material to be formed on a semiconductor device is determined.




In step


230


, once changes in ‘n’ and ‘k’ are determined across an interface of two materials based on process factors in a given semiconductor process, and once the desired thickness of the overlying material is determined, such changes can be modeled and a depth profile


40


of optical constants may be asserted for use by ellipsometer


10


. An example of depth profile


40


for a particular thickness of crystalline silicon over amorphous silicon is illustrated in FIG.


3


A. The crystalline silicon over amorphous silicon profile


40


represents epitaxial layer


28


and partially-annealed doped layer


24


over defect layer


26


as described in reference to FIG.


1


B.




In step


240


, ellipsometer


10


may use depth profile


40


to construct predicted behavior curves


42


and


44


representing, respectively, the predicted intensity and phase of light reflected from semiconductor device


20


at various wavelengths of light and measured by ellipsometer


10


. Such intensity and phase are conventionally represented by ellipsometer


10


as tangent(psi) and cosine(delta). Examples of both a predicted behavior tangent(psi) curve


42


and a predicted behavior cosine(delta) curve


44


are presented in

FIGS. 3B and 3C

that correspond to the predicted optical behavior of a device having an epitaxial silicon layer


28


deposited on a partially annealed layer of doped silicon


24


. In the example illustrated by

FIGS. 3B and 3C

, doped silicon layer


24


is only partially annealed such that defect layer


26


is present. Thus, curves


42


and


44


may be generated by ellipsometer


10


based on the particular depth profile


40


illustrated in FIG.


3


A.




Once ellipsometer


10


has generated curves


42


and


44


that are representative of the intensity and phase of light at various wavelengths, ellipsometer


10


may then utilize curves


42


and


44


during semiconductor processing in order to determine when or whether an overlying layer of material has reached the desired thickness. Thus, in step


245


, beams of light are emitted by ellipsometer


10


to strike the surface of a semiconductor device. In step


250


, ellipsometer


10


collects phase and intensity measurement data for different wavelengths of light that have been reflected from the semiconductor device during or after the formation of the overlying layer of material. In an alternative embodiment, data associated with only one wavelength of light may be utilized.




In step


260


, ellipsometer


10


performs analytic techniques such as iterative processing and curve-fitting in order to attempt to match the collected phase and intensity data to curves during the formation of the overlying layer of material.




In step


270


, ellipsometer


10


determines if the phase and intensity data closely match the optical data predicted by curves


42


and


44


as illustrated by the presence of dashed curves


46


and


48


in

FIGS. 3B and 3C

. If a match is identified, ellipsometer


10


indicates that the overlying layer of material is deposited to the correct thickness in step


280


. If a match is not identified, ellipsometer


10


indicates that the overlying layer of material is deposited to an incorrect thickness in step


290


.





FIG. 4

illustrates a flowchart of a particular embodiment of the process illustrated in the flowchart of FIG.


2


. More particularly,

FIG. 4

illustrates a method for forming or verifying the formation of silicon epitaxial layer


28


to a desired thickness over a partially annealed doped layer


24


such that defect layer


26


is present.




In step


410


, a particular depth profile


40


is received by ellipsometer


10


, the particular depth profile corresponding to the optical behavior at defect layer


26


given a desired total thickness of partially annealed doped layer


24


and silicon epitaxial layer


28


. In step


420


, particular curves


42


and


44


are generated to model the intensity and phase of light reflected by semiconductor device


20


given the particular depth profile


40


. In step


425


, light generated by ellipsometer


10


is directed toward the surface of semiconductor device


20


during or after formation of silicon epitaxial layer


28


. In step


430


, data corresponding to the intensity and phase of light reflected by semiconductor device


20


in step


425


is collected. In step


440


, the collected data is compared to the particular curves


42


and


44


. In step


450


, ellipsometer


10


determines if the collected data matches the particular curves


42


and


44


. If there is a match, ellipsometer


10


indicates that a desired thickness has been reached in step


460


. If there is no match between the collected data and curves


42


and


44


, ellipsometer


10


indicates that the desired thickness has not been achieved in step


470


.




In one embodiment, the processes described with reference to

FIGS. 2 through 4

may be utilized to determine the thickness of a third layer of material using an interface between a first and second layer of material. For example, the thickness of an epitaxial layer of silicon may be determined by determining the thickness of an underlying annealed amorphous silicon layer before the formation of the epitaxial layer. In such an example, the thickness of the underlying annealed amorphous silicon may be determined by examining an interface between the annealed amorphous silicon and an underlying substrate.




Such an interface may have distinctive optical properties caused by, for example, the less than complete anneal of the annealed amorphous silicon. After the thickness of the underlying annealed amorphous silicon is determined, the same interface may again be examined after the formation of the epitaxial layer to determine a cumulative depth of the annealed amorphous silicon and epitaxial layer combined. By subtracting the depth of annealed amorphous silicon from the cumulative depth, the approximate depth of the epitaxial layer may be easily determined.




Although the present invention has been described using several embodiments, various changes and modifications may be suggested to one skilled in the art after a review of this description. It is intended that the present invention encompass such changes and modifications as fall within the scope of the appended claims.



Claims
  • 1. A process for manufacturing a silicon integrated circuit product that includes a crystalline silicon substrate, a thin epitaxial silicon layer, and a thin silicon interface layer therebetween, comprising:a. constructing a depth profile of a target structure including: i a silicon substrate having a certain thickness, a known index of refraction and extinction coefficient; ii a silicon interface layer overlaying the silicon substrate, having an empirical depth, index of refraction and extinction coefficient; and iii an epitaxial silicon layer overlaying the interface layer, having a desired thickness and the known index of refraction and extinction coefficient; b. generating a first set of model behavior curves of the target structure, based on the depth profile, including: i an intensity change curve, also known as a tangent curve, also known as psi curve; and ii a phase change curve, also known as a cosine curve, also known as delta curve; c. forming a thin silicon epitaxial layer having the known index of refraction and extinction coefficient over the product substrate, the forming also including the forming of some interface layer; d. emitting light signals of certain wavelengths from an ellipsometer on the surface of the epitaxial layer at certain angles with respect to the surface of the epitaxial layer; e. generating reflected light signals from the surface of the epitaxial layer and the interface layer; f. collecting the reflected light signals with the ellipsometer; g. generating a second set of behavior curves from the reflected light signals collected by the ellipsometer including: i an intensity change curve, also known as a tangent curve, also known as psi curve; and ii a phase change curve, also known as a cosine curve, also known as delta curve; and h. ascertaining the thickness of the formed epitaxial layer by comparing the first set and the second set of the behavior curves and calculating the difference, if any, between the target thickness and the formed thickness.
Parent Case Info

This is a divisional application of application No. 09/711,125 filed Nov. 9, 2000, now abandoned.

US Referenced Citations (4)
Number Name Date Kind
5061072 Folkard et al. Oct 1991 A
5337150 Mumola Aug 1994 A
5666200 Drevillon et al. Sep 1997 A
5963327 He et al. Oct 1999 A
Non-Patent Literature Citations (2)
Entry
Warnick et al., “Ellipsometry as a Sensor Technology for the Control of Deposition Processes”, Decision and Control, Dec. 1998, Proceedings of the 37th IEEE Conference on, vol. 3, pp. 3162-3167.*
Jean-Louis et al., “SIMOX Layers and Interfaces Studies With A New Fast Multichannel Spectroscopic Ellipsometer”, IEEE SOS/SOI Technology Conference, Oct. 1990, pp. 152-153.