Semi Report

8/6/2019 Semi Report

1/16

1 / 16

NANYANG POLYTECHNIC

Report

For

EG 3244

Semiconductor Technology

Thin Film Thickness Measurement Techniques

Written by:

Fu Lingdong (095319R)

Huang Yu (090020W)

Liu Jiannan (094250A)

Date of Submission: 17/06/2011


2/16

2 / 16

Content

1. Introduction

2. Optical techniques

2.1 Interference

2.2 Spectral Reflectance

2.3 Ellisometry

3. X-rays techniques

3.1 Abstract

3.2 X-ray Fluorescence analysis (XRF)

3.3 X-ray Reflectance spectrometry (XRR)

4. Thermal techniques

4.1 Rutherford Backscattering Spectrometry (RBS)

5. Applications

5.1 NanoCalc spectrometry systems

5.2 FilmTek 4000

6. Summary

7. References


3/16

3 / 16

1. IntroductionVery thin layers of different materials that are deposited on the surface of another material

(thin films) are extremely important to many technology-based industries. Thin films are

widely used, for example, to provide passivation, insulating layers between conductors,

diffusion barriers, and hardness coatings for scratch and wear resistance. The fabrication of

integrated circuits consists primarily of the deposition and selective removal of a series of

thin films.

Films typically used in thin-film applications range from a few atoms to half a millimeter.

(


4/16

4 / 16

measures the amount of light reflected from a thin film over a range of wavelengths, with

the incident light normal (perpendicular) to the sample surface. Ellipsometry is similar,

except that it measures reflectance at non-normal incidence and at two different

polarizations. In general, spectral reflectance is much simpler and less expensive than

ellipsometry, but it is restricted to measuring less complex structures.

2.2 Spectral ReflectanceSpectral reflectance measure a thin film's characteristics by reflecting light off the film and

analyzing the resulting reflectance spectrum over a range of wavelengths. Light reflected

from different interfaces of the film can be in- or out-of-phase so these reflections add or

subtract, depending upon the wavelength of the light and the film's thickness and index. The

result is intensity oscillations in the reflectance spectrum that are characteristic of the film.

To determine the film's thickness, the software calculates a theoretical reflectance spectrum

that matches as closely as possible to the measured spectrum. It begins with an initial guess

for what the reflectance spectrum should look like, based on the nominal film stack. This

includes information on the thickness and the refractive index of the different layers and the

substrate that make up the sample. The theoretical reflectance spectrum is then adjusted by

adjusting the film's properties until a best fit to the measured spectrum is found.

This metrology can be used on thin film thickness measurement, even for transparent films.2.2.1 Basicprinciples of spectralreflectance-Determining Film Properties from Spectral Reflectance

The amplitude and periodicity of the reflectance of a thin film is determined by the film's

thickness, optical constants, and other properties such as interface roughness. In cases

where there is more than one interface, it is not possible to solve for film properties in

closed form, nor is it possible to solve for n and k at each wavelength individually. In

practice, mathematical models are used that describe n and k over a range of wavelengths

using only a few adjustable parameters. A film's properties are determined by calculating

reflectance spectra based on trial values of thickness and the n and k model parameters,

and then adjusting these values until the calculated reflectance matches the measured

reflectance.


5/16

5 / 16

Figure2.1.1.1a Principle of spectral reflectance

2.2.2 Models for nand k

There are many models for describing n and k as a function of wavelength. When choosing

a model for a particular film, it is important that the model be able to accurately describe n

and k over the wavelength range of interest using as few parameters as possible. In general,

the optical constants of different classes of materials (e.g., dielectrics, semiconductors,

metals, and amorphous materials) vary quite differently with wavelength, and require

different models to describe them (see below) .Models for dielectrics (k=0) generally have

three parameters, while non-dielectrics generally have five or more parameters. Therefore,

as an example, to model the two-layer structure, a total of 18 adjustable parameters must

be considered in the solution.

Figure2.1.1.1b Different models for n and k


6/16

6 / 16

2.3 EllipsometryEllipsometry is an optical technique for the investigation of the dielectric properties

(complex refractive index or dielectric function) of thin films. It has applications in many

different fields, from semiconductor physics to microelectronics and biology, from basic

research to industrial applications. Ellipsometry is a very sensitive measurement technique

and provides unequalled capabilities for thin film metrology. As an optical technique,

spectroscopic ellipsometry is non-destructive and contactless.

Upon the analysis of the change of polarization of light, which is reflected off a sample,

ellipsometry can yield information about layers that are thinner than the wavelength of the

probing light itself, even down to a single atomic layer. Ellipsometry can probe the complex

refractive index or dielectric function tensor, which gives access to fundamental physical

parameters and is related to a variety of sample properties, including morphology, crystal

quality, chemical composition, or electrical conductivity. It is commonly used to characterize

film thickness for single layers or complex multilayer stacks ranging from a few angstroms or

tenths of a nanometer to several micrometers with an excellent accuracy.

2.3.1 Basicprinciples ofellipsometryEllipsometry measures the change of polarization upon reflection or transmission. Typically,

ellipsometry is done only in the reflection setup. The exact nature of the polarization changeis determined by the sample's properties (thickness, complex refractive index or dielectric

function tensor). Although optical techniques are inherently diffraction limited, ellipsometry

exploits phase information and the polarization state of light, and can achieve angstrom

resolution. In its simplest form, the technique is applicable to thin films with thickness less

than a nanometer to several micrometers. The sample must be composed of a small number

of discrete, well-defined layers that are optically homogeneous and isotropic. Violation of

these assumptions will invalidate the standard ellipsometric modelling procedure, and more

advanced variants of the technique must be applied.

2.3.2 Experimental details ofellipsometry-Experimental setup


7/16

7 / 16

F

e2 1 2 1Sc e

cse

of an e

some

ye e

ment.

Electromagnetic radiation isemitted by a light source and linearly polarized by a polarizer. t

can pass through an optional compensator (retarder, quarter wave plate and falls onto the

sample. After reflection the radiation passes a compensator (optional) and a second

polarizer, which iscalled an analyzer, and falls into the detector. nstead of the

compensatorssomeellipsometers use a phase-modulator in the path of the incident light

beam. Ellipsometry is a specular optical technique (the angle of incidenceequals the angle of

reflection). The incident and the reflected beam span theplane ofincidence. Light which is

polarized parallel to this plane is namedp-polarized(p-polarised). A polarization direction

perpendicular iscalled s-polarized(s-polarised), accordingly. The"s"iscontributed from the

German "senkrecht"(perpendicular).

- ata acquisition

Ellipsometry measures thecomple reflectance ratio, , of a system, which may be

parametrized by and . The polarization state of the light incident upon thesample may

be decomposed into an s and apcomponent (thescomponent is oscillating perpendicular

to the plane of incidence and parallel to thesamplesurface, and thepcomponent is

oscillating parallel to the plane of incidence). The amplitudes of thes andpcomponents,

after reflection and normalized to their initial value, are denoted byrs and rp, respectively.

Ellipsometry measures thecomple reflectance ratio, (a comple quantity), which is the

ratio ofrp over rs:


8/16

8 / 16

Thus, tan() is the amplitude ratio upon reflection, and is the phaseshift (difference).

(Note that the right hand side of theequation issimply another way to represent a comple

number.) Sinceellipsometry is measuring the ratio (or difference) of two values (rather than

the absolutevalue ofeither), it isvery robust, accurate, and reproducible. For instance, it is

relatively insensitive to scatter and fluctuations, and requires no standard sample or

reference beam.

- ata analysis

Ellipsometry is an indirect method, i.e. in general the measured and cannot be

converted directly into the optical constants of thesample. Normally, a model analysis must

be performed. Direct inversion of and is only possible in verysimplecases of isotropic,

homogeneous and infinitely thic films. n all other cases a layer model must beestablished,

which considers the optical constants (refractive inde or dielectric function tensor) and

thic ness parameters of all individual layers of thesample including thecorrect layer

sequence. Using an iterative procedure (least-squares minimization) unknown optical

constants and/or thickness parameters arevaried, and and values arecalculated using

theFresnel equations. Thecalculated and values which match thee perimental data

best provide the optical constants and thickness parameters of thesample.

3. X-rays t ! hniqu s

3.1 Abstract

A film thickness measuring method comprises thesteps of measuring reflectance of X-rays

on a film, e" tracting interference oscillations from the measured X-ray reflectance, and

Fourier transforming the interference oscillations to compute a film thicknessof the film, an

average reflectance being given by fitting the measured X-ray reflectance to an analysis


9/16

9 / 16

formula including a term of a product of a power function of an incident angle, which

expresses attenuation of reflectance on a smooth surface of the film and an exponent

function which expresses influence of roughness of the surface of the film, and a constant

term expressing a background added to the product; the interference oscillations being

given by using the measured X-ray reflectance and the average reflectance. The film

thickness measuring method can extract interference oscillations of a reflectance curve by a

method including arbitrariness and by a simple procedure. Generally, there are two types of

X-rays techniques being commonly used for thin film thickness measurement in

semiconductor industry, they are X-ray fluorescence analysis (XRF) and X-ray reflectance

spectrometry (XRR).

3.2 X-rayFluorescenceanalysis (XRF)

In X-ray fluorescence analysis, X-ray fluorescence emitted from a sample when X-rays are

irradiated onto the sample is measured, and based on an intensity of the X-ray fluorescence,

the amounts of elements contained in the sample are determined, and the amounts are

converted into a film thickness. In X-ray fluorescence analysis, amounts of elements alone

are measured, and a resultant film thickness is a converted film thickness and is not always

an actual film thickness. X-ray fluorescence analysis is not suitable for measuring multiple

layers, such as layers of Ti (titanium) and TiN (titanium nitride).

3.2.1 Basicprinciples of X-rayFluorescenceanalysis (XRF)Thin-film x-ray fluorescence deals with samples which ideally are infinitely thin according to

exciting radiation (primary) as well as characteristic (secondary) radiation. This means that

the characteristic radiation generated in the sample is ideally not diminished. In general the

loss of intensity of the characteristic radiation of an element transmitting a sample of the

thickness t can be described by the following expression:

dI = k( 1/sinJI 0 e- {[ ( Q/V)

P, pri1/sinJ+ ( Q/V)

P, L1/sin]] Vt}dt

where,

V=density of the sample in g/cm3


10/16

10 / 16

J= angle between primary beam and sample surface

]= angle between sample surface and secondary beam

PpriP,L = wavelengths of primary- and analyte-line respectively

QVP= mass absorption coefficient of the sample at a wavelength l in cm2/g

In case of a very thin sample the absorption of x-rays will be very small or approximately

zero. So the exponential term of the foregoing equation will become negligible, resulting in:

dI= k 1/sinJI 0 dt , or(I = k 1/sinJI 0(t .

For a constant sample area n (n = atom number of the analyte) can be regarded as

proportional to t.

(I = k 1/sinJI 0 (n

This means that in case of a thin sample film an increase in intensity I is linearly dependent

on the absolute number of analyte atoms. If primary beam intensity and the angle f are kept

constant, the expression simplifies to:

(I = k(n.

3.3 X-rayReflectance spectrometry (XRR)X-ray reflectance spectrometry (XRR) is a non-destructive and non-contact technique for

thickness determination between 2-200 nm with a precision of about 1-3 A. In addition tothickness determination, this technique is also employed for the determination of density

and roughness of films and also multi-layers with a high precision.

3.3.1 Basicprinciples of X-rayreflectance spectrometry (XRR)XRR method involves monitoring the intensity of the x-ray beam reflected by a sample at

grazing angles. A monochromatic x-ray beam of wavelength irradiates a sample at a

grazing angle and the reflected intensity at an angle 2 is recorded.


11/16

11 / 16

Figure3.3.1: /2-Scan: The condition of incident angle = (2)/2 == outgoing angle is

satisfied. The detector D rotates at twice the speed of the sample P. This arrangement is

sensitive only to the planes parallel to the surface of the sample. The beam makes an

incident angle with the surface of the sample P. The reflected intensity at angle of 2 is

measured. Both the rotation of the sample and the detector (2) are about the same axis

MP (perpendicular to the drawing). The sample is adjusted so that the rotation axis lies on

the sample surface. The Detector circle is fixed through the (programmable) detector slit

(PRS, programmable receiving slit). The anode focus, F of the tube lies on the detector circle.

By a detector, see figure3.3.1. This figure illustrates specular reflection where the condition

= 2/2 is satisfied. The mode of operation is therefore /2 mode which makes sure the

incident angle is always half of the angle of diffraction. The reflection at the surface and

interfaces is due to the different electron densities in the different layers (films), which

corresponds to different reflective indexes in the classical optics. For incident angles below

a critical angle c, total external reflection occurs. The critical angle for most materials is less

than 0.3. The density of the material is determined from the critical angle. Above c the

reflection from the different interfaces interference and give rise to interference fringes. The

period of the interference fringes and the fall in the intensity are related to the thickness

and the roughness of the layer (layers in case of multi-layers). The reflection can be analyzed

using the classical theory (Fresnel equation). The typical range for these measurements is

between 0 and 5 in .

3.3.2 Filmthickness calculationFor incident angles greater than c, (>c) the x-ray beam penetrates inside the film.

Reflection therefore occurs at the top and the bottom surfaces of the film. The interference


12/16

12 / 16

between the rays reflected from the top and the bottom of the film surfaces results in

interference fringes which does not depend on the frequency like in the caseof optical

spectroscopy but is angle dependent. Due to the low amplitude reflection coefficient (v,h

1/ sin2 Rv,h = |rv,h|1/ sin41/4) of interface between adjacent layers, contributions of

multiply reflected beams can be neglected. The m-th interference maximum for a path

difference = m, is located at

m = = 2dX,1(m) (1)

2d , with m N (2)

m2 m2(2/4d2)+2 (3)

= m2(2/4d2)+C2 (4)

If the substrate is optically denser than the film, a phase difference of occurs at the

reflection film / substrate interface and m is substituted with m+ . Employing equation (3)

and the difference between two neighboring maxima and minima, the thickness can be

determined and is given by

d

(5)

, for m C (6)

The thickness is often determined with a precision better than 1 A for measurements

exhibiting interference fringes in a bigger angular range.

4. Thermaltechniques4.1 RutherfordBackscattering Spectrometry (RBS)

Rutherford Backscattering Spectrometry (RBS) is an ion scattering technique that is used for

compositional thin film analysis. RBS is unique in that it allows quantification without the use

of reference standards. During an RBS measurement, high-energy (MeV) He++ ions are

directed onto a sample and the energy distribution and yield of the backscattered He++ ions

at a given angle is recorded. Since the backscattering cross section for each element is

known, it is possible to obtain quantitative depth profiles from the RBS spectra (for thin films

that are less than 1mm thick).

4.1.1 Basicprinciples of Rutherford Backscattering Spectrometry (RBS)


13/16

13 / 16

In RBS, where a beam of monoenergetic (in the MeV range) collimated alpha particles

impinge on a target and a fraction of the backscattered particles is analyzed with a detector

that is placed at a fixed scattering angle. The alpha particle penetrates and loses energy

along its trajectory due to interaction with bound and free electrons in the target. The

magnitude of the energy loss depends on the total distance that the incident particle has

traveled. It also depends on the density and composition of the target, as well as the

incident particles velocity. The relation between the energy loss of the backscattered

particles and the depth x is given by

E = [S]x;

where E is the energy difference between the particles scattered from the atoms at the

surface and those scattered from atoms at depth x, and S is the backscattering energy loss

factor. By measuring the energy difference E, the thickness of the sample can then be

determined.

4.1.2 A typicalapplication-Characterization ofaburied germaniumlayer

An example of an RBS spectrum of a buried germanium layer is given in Figure4.1.2. The

sample consists of a silicon substrate with a Ge/Si/ SiO2 layer stack on it. A 2 MeV helium

beam is directed at the sample and the energy of the backscattered helium ions is measured.

In Figure4.1.2, the surface energy positions of the elements oxygen, silicon and germaniumare indicated by dotted lines. The surface energy position indicates the energy of helium

ions back-scattered off an element in the outermost layer of the sample. The Ge peak (the

red peak) shifts towards lower energy (compared to the surface energy position) because of

the extra energy loss of the helium ions in the Si and SiO2 layers on top of the Ge layer. The

intensity of the Ge peak is linearly dependent on the amount of Ge. The Si and O signal start

at the surface energy position because the SiO2 forms the top layer. The width of the O and

the first Si signal (both blue) both represent the thickness of the SiO2 layer. At the lower

energy side of the first Si signal an additional peak can be observed. This is the Si signal of

the pure Si layer (yellow). The width of this peak gives the thickness of the pure Si layer. The

signal of the substrate is represented by a continuum instead of a peak, because the Si

substrate is infinitely thick for the helium beam (thicker than 2 m). A summary of the

results is given in Table4.1.2.


14/16

14 / 16

Figure4.1.2 Example of an RBS spectrum Table4.1.2 Thickness in atoms/cm2

with the x-axis representing the energy determined from the data shown in

scale and the y-axis the relative signal Figure4.1.2. By dividing the amount in

intensity atoms/cm2 by the density the thickness

is obtained

5. Applications ofthetechniques5.1 NanoCalc spectrometry systems

Thickness of optical layers can be measured quickly and easily with NanoCalc advanced

spectrometry systems. The entire system sets up in minutes and measurements can be

made by anyone with basic computer skills. The simple hardware and intuitive software

provides thin-film capability to a whole new group of users.

For semiconductor process films, NanoCalc measurement systems are routinely used to

measure the thickness of oxides, SiNx, photo resists and other semiconductor process films.

In addition to these single-layer applications, many two- and three-layer film measurements

are also possible.

For In-Situ measurements, a flexible optical probe assembly makes on-line and in-situ

thickness measurements easily possible. All that is required is optical access for

normal-incidence reflectance measurements. An example is online-measuring of removal

rate in nm per second while removing resist-layers from wafers.


15/16

15 / 16

Figure5.1 Application of optical measurement technique with NanoCalc

5.2 FilmTek 4000

FilmTek 4000 capability employs the patented DPSD (Differential Power Spectral Density)

technique. Spectroscopic reflection data are gathered at normal incidence and 70 degrees.

PSD processing results in two peaks in the Power Spectral Density domain. The ratio of their

positions is a function of the index of refraction of the film, and the angle of incidence of the

oblique measurement. This ratio is used to calculate the index. Once the index is known, the

thickness can be calculated from the optical thickness of the normal incident peak.

Figure5.2a Differential Power Spectral Figure5.2b Index Map @ 1550nm of a doped

Density analysis of an oxide film oxide measured with FilmTekTM 4000


16/16

16 / 16

6. SummaryFor thin film thickness metrology, precision is an important feature defining the potential

capability of a metrology tool, but it does not prove anything about its compatibility to

end-users` needs. Repeatability depends both on the metrology tool itself and on the size

and uniformity of the measured test sites. When evaluating stability, it is important that the

chosen measured object be much more stable that the metrology tool itself.

According to the different thin film thickness measurement techniques we discussed in the

report, we conclude that all these techniques can accurately achieve the result of the

thickness measurement base on their principles. In nowadays semiconductor industries,

optical techniques are more commonly applied in the thickness measurement field, because

optical techniques accurately measure the thickness of thin transparent and

semi-transparent layers by analyzing white light interference. Also, this method is

non-destructive and does not need any special sample preparation.

7. References http://en.wikipedia.org/wiki/Ellipsometry

http://dymek.com/Dymek-products-2-010.aspx

http://www.kirklau.com/proj/thinfilm.pdf http://www.mikropack.de/download/brochure_nanocalc.pdf

http://www.opticaltest.com/download/flatscan2010.pdf

http://www.sci-soft.com/FilmTek%204000.htm

http://ia.physik.rwth-aachen.de/methods/xray/www-xray-eng.pdf?menu_id=80&langu

age=german

http://resources.aims.ac.za/archive/2009/wafa.pdf

http://www.cnt.fraunhofer.de/en/ Images/0909_Produktblatt_R%C3%B6ntgendiffrakto

metrie_en_tcm101-30765.pdf

http://www.research.philips.com/technologies/matanalysis/downloads/6-rbstn.pdf

Semi Report

Documents

Transcript of Semi Report