Radiant energy – Inspection of solids or liquids by charged particles – Methods
Reexamination Certificate
2000-10-12
2004-02-10
Lee, John R. (Department: 2881)
Radiant energy
Inspection of solids or liquids by charged particles
Methods
C250S310000, C250S311000
Reexamination Certificate
active
06690009
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a method of determining the charge carrier concentration in a doped specimen, notably a semiconductor. The charge carrier concentration for semiconductors is to be understood to mean the electron concentration in the case of n-type semiconductors as well as the hole concentration in the case of p-type semiconductors. Generally speaking, the dope atom concentration in the material follows directly from such a charge carrier concentration.
The invention also relates to an electron beam apparatus for carrying out such a method.
The distribution of dope atoms is of major importance for correct operation of semiconductor elements. The concentration thereof determines the electrical properties of the doped material and the exact distribution thereof co-determines the correct operation of the semiconductor elements. The manufacture of such semiconductor elements involves various process steps which may influence the distribution of the dope atoms, for example thermal treatments. Therefore, it is desirable to have a method of measuring the distribution of the dope atoms so as to check the effect of the process steps; this is desirable not only during the development of the semiconductor elements, but also during critical process steps in their manufacture.
The following problem will be further illustrated on the basis of dope atom distributions in Si (silicon) doped with B (borium), so in a p-type semiconductor material. The present generation of Si semiconductors is so small that a technique such as “secondary ion mass spectroscopy”, being renowned for its favorable detection limit, cannot offer the desired lateral resolution. An alternative technique, wherein a Scanning Electron Microscope (SEM) is used to analyze the energy of secondary electrons obtained by bombardment of a specimen (substrate or wafer) by means of electrons and to measure the charge carrier concentration on the basis of the plasmon frequency, is known from JP-A-63/271949. Because the plasmon energy is of the same order of magnitude as that of the secondary electrons, according to the cited publication measurement is performed on the derivative of the energy spectrum, this leads to inadequate accuracy.
SUMMARY OF THE INVENTION
It is an object of the invention to provide a method of accurately determining the charge carrier concentration, notably in the case of semiconductors of very small dimensions. This object is achieved according to the invention in that the method of the kind set forth is characterized in that the electron beam emanating from an electron source is made to interact with the specimen, after which an energy spectrum of the electrons is derived from this beam by means of a spectrometer, said spectrum being analyzed so as to derive therefrom the plasmon frequency in the specimen and the charge carrier concentration of the doped material being derived therefrom. In other words, measurement is performed on the primary electrons. The detection efficiency of primary electrons is many times higher than that of secondary electrons, because they fill a much smaller solid angle. Furthermore, analysis of the spectrum of secondary electrons is significantly more difficult, because the plasmon peaks occurring appear against a strong signal background; the energy of at least the first plasmon peak is of the same order of magnitude as the energy of the secondary electrons. Furthermore, it is difficult to separate secondary electrons of an energy near that of the primary electrons from the primary electrons; this necessitates the use of complex and hence expensive equipment.
The energy spectrum is preferably determined by means of an “Electron Energy Loss Spectroscopy” (EELS) technique. In as far as the electrons do not lose energy in the specimen, they produce a so-called “zero loss peak” in the energy spectrum, whereas the electrons which interact with the specimen and hence lose energy generally exhibit one or more plasmon peaks in the energy spectrum.
For a plasmon peak the maximum energy loss E
max
derived from the measured energy spectrum can be expressed by the relation:
E
max
=[(
E
p
)
2
−(&Dgr;
E
p
/2)
2
]
½
wherein E
p
represents the plasmon energy and &Dgr;E
p
represents the width of the plasmon peak. Therefrom, the plasmon energy E
p
can be determined and, because E
p
equals h/2&pgr;.&ohgr;
p
, also the plasmon frequency &ohgr;
p
. The following relation holds approximately for the plasmon frequency &ohgr;
p
:
n
=
m
ϵ
o
e
2
ω
p
2
wherein m is the mass of an electron, e is the charge thereof, ∈
o
is the dielectric constant and n is the charge carrier concentration.
Because in the example involving borium-doped silicon the electron concentration n in the specimen consists mainly of the difference between the silicon valence electron concentration n
v
and the borium hole concentration n
g
, the value n obtained should be corrected by way of the value of n
v
. Moreover, for example, because in the case of phosphor-doped silicon the electron concentration n in the specimen consists mainly of the sum of the silicon valence electron concentration n
v
and the phosphor conductance electron concentration n
g
, the value n obtained should be corrected by way of the value of n
v
. These corrections can be determined by calculation or, as is much more accurate, by calibration. Calibration is then performed by applying the method according to the invention to a non-doped specimen or a non-doped part of a doped specimen.
It has been found in practice that the measured plasmon frequency is also dependent on the thickness of the specimen. Therefore, according to the invention the plasmon frequency &ohgr;
p
is corrected for the thickness of the specimen. When the (P)EELS spectrum is measured from the “zero loss” peak to the plasmon peak, a measure of the thickness can be calculated from the ratio of the two peak heights. Using this thickness, the plasmon frequency can be transformed to a thickness-independent value. It has been found in practice that a mainly linear relationship exists between the plasmon frequency and the thickness of the specimen for B-doped Si: &ohgr;
p
=a.thickness+b(n
g
). Therein, “a” is independent of the B concentration n
g
within the measuring accuracy, but “b” is dependent thereon. The previously determined thickness of the specimen can then be used to determine the value of “b” from the measured plasmon frequency, and hence also the plasmon frequency for a thickness “zero” or the plasmon frequency for a standard thickness of the specimen.
In order to derive an energy spectrum of the primary electrons, interacting with the atoms in the specimen or not, notably an “Electron Energy Loss Spectroscopy” (EELS) technique is used. Use is preferably made of “Parallel Electron Energy Loss Spectroscopy” (PEELS) wherein the output signals of the spectrometer used are simultaneously read out and hence the zero loss peak and the plasmon peak or peaks are measured simultaneously. It is a drawback of the serial reading out of the output channels of the spectrometer that, when variations in time (for example, of the current or voltage) occur in, for example, the spectrometer, the measured positions of essentially the same plasmon peak may differ.
Even though it is possible to determine the energy spectrum of the primary electron beam after it has entered into interaction with the specimen during reflection (where the beam is incident on the sample at an acute angle), the electron beam preferably is made to interact with the specimen in transmission. According to the invention the plasmon frequency, therefore, is determined in a “Transmission Electron Microscope” (TEM). When the electron beam is focused on the specimen in a TEM, the charge carrier concentration is determined at that area. The spot size is dependent on the degree of (de)focusing. In the case of a comparatively large spot, the mean charge carrier concentration is determined. However, notably for sem
Koninklijke Philips Electronics , N.V.
Lee John R.
Quash Anthony
LandOfFree
Method of determining the charge carrier concentration in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of determining the charge carrier concentration in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of determining the charge carrier concentration in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3316834