semi合集-English.pdf - 第7367页
SEMI MF1392-1103 © SEMI 2003 11 11.4.2.3 Bring the mercury column(s) into con tact with the surface of the wafe r. 11.4.2.4 Set the bias to a nominal value of 0 V, a nd read the capacitance, C , and the phase angl e, θ ,…
SEMI MF1392-1103 © SEMI 2003 10
10.5.2 Connect the cables to one of the precision
capacitors. Measure and record the capacitance in pF to
three significant figures. Disconnect the capacitor.
Repeat for the other precision capacitors.
10.5.3 If the measured capacitance values are not
within 1% of the known values, make necessary
adjustments consistent with the appropriate instrument
instruction manuals to bring the instrument to within
specifications before proceeding with measurements of
test specimens.
10.6 Voltmeter Verification
10.6.1 Measure the precision voltage source at five or
more voltages within the range from 0 to ± 200 V,
inclusive to verify that the digital voltmeter is within
specification over this range.
10.6.2 If any measured voltage value is not within
0.5% of the known value, make necessary adjustments
consistent with the appropriate instrument instruction
manuals to bring the instrument to within specifications
before proceeding with measurements of test
specimens.
11 Procedure
11.1 Refer to Related Information 2 for suggested data
sheet formats for recording the data if the data
collection and calculations are carried out manually or
off-line.
NOTE 14: The following procedures are given in sufficient
detail for manual data collection and calculations to be carried
out. However, it is strongly recommended that both data
collection and analysis be carried out using computer con-
trolled equipment, with data storage and display capabilities.
In such cases, the procedures and algorithms employed must
be equivalent to those given in this test method.
11.2 If not known, determine the conductivity type and
surface orientation of the test wafers in accordance with
SEMI MF42 and SEMI MF26, respectively.
11.3 Estimate the reverse bias voltage range over
which the measurements are to be made based on the
curve in Figure 1, an estimate of the value for the
dopant density of the test specimen, and the range of
depth over which the profile is desired. Do not exceed
200 V or 80% of the breakdown voltage, whichever is
lower.
11.4 Determination of the Series Resistance of the
Diode Circuit — Determine the series resistance of the
diode circuit in one of the following ways:
11.4.1 Measurement of Forward Resistance — If the
capacitance bridge or meter measures capacitance only,
measure the diode forward resistance, R, in ohms, at 1
V forward bias as follows:
11.4.1.1 Connect the curve tracer to the mercury probe
column and to the return contact of the probe fixture.
11.4.1.2 Place the wafer to be tested onto the mercury
probe fixture set up in the configuration used to
determine C
comp
(see Section 10.3) in such a way that
the mercury column(s) will contact the polished or
epitaxial surface of the wafer. If the back-surface-
return-contact configuration is used, make a suitable
return contact to the substrate or back surface of the
wafer.
11.4.1.3 Bring the mercury column(s) into contact with
the surface of the wafer.
11.4.1.4 Measure and record as I
1
the current through
the diode at 0.9 V forward bias, in mA, to two
significant figures.
11.4.1.5 Measure and record as I
2
the current through
the diode at 1.1 V forward bias, in mA, to two
significant figures.
11.4.1.6 Calculate the forward resistance, R, in kΩ, as
follows:
12
2.0
II
R
−
=
(6)
where:
I
1
= current at 0.9 V forward bias, mA, and
I
2
= current at 1.1 V forward bias, mA.
11.4.1.7 If the forward resistance is 1 kΩ or less,
proceed to Section 11.5. If the forward resistance
exceeds 1 kΩ, improve the return contact, and repeat
Section 11.4.1.
NOTE 15: The diode forward resistance at 1 V, R,
determined in this way is a measure of the total series
resistance of the test circuit that includes the bulk, cable, and
return contact resistances.
11.4.2 Direct Determination of Equivalent Series
Resistance — For capacitance meters or bridges
capable of measurement of phase angle, conductance,
or total impedance, determine the equivalent series
resistance directly as follows:
11.4.2.1 Place the wafer to be tested onto the mercury
probe fixture set up in the configuration used to
determine C
comp
(see Section 10.3) in such a way that
the mercury column(s) are just above but do not contact
the polished or epitaxial surface of the wafer. If the
back-surface-return-contact configuration is used, make
a suitable return contact to the substrate or back surface
of the wafer.
11.4.2.2 Zero the meter in accordance with the manu-
facturer' s instructions.
SEMI MF1392-1103 © SEMI 2003 11
11.4.2.3 Bring the mercury column(s) into contact with
the surface of the wafer.
11.4.2.4 Set the bias to a nominal value of 0 V, and
read the capacitance, C, and the phase angle,
θ
, or the
conductance, G
m
.
11.4.2.5 If the phase angle and capacitance were
measured, calculate the series resistance, R
s
, in kΩ, as
follows:
θπ
tan2
10
3
Cf
R
s
=
(7)
where:
θ
= measured phase angle, degrees,
C = measured capacitance, pF, and
f = measurement frequency, MHz.
11.4.2.6 If the capacitance and conductance were
measured, calculate the series resistance, R
s
, in kΩ, and
the phase angle,
θ
, in degrees, as follows:
232
)102(
−
×+
=
CfG
G
R
m
m
s
π
(8)
and
−=
−
CRf
s
π
θ
2
10
tan
3
1
(9)
where:
G
m
= measured conductance, S,
f = measurement frequency, MHz, and
C = measured capacitance, pF.
11.4.2.7 If the series resistance is 1 kΩ or less and the
phase angle is between −87° and −90° , proceed to
Section 11.5. If either of these conditions is not met,
improve the return contact, and repeat Section 11.4.2.
If improvement of the return contact decreases the
series resistance but does not bring the phase angle
within the desired range, qualify the Schottky contact in
accordance with Section 11.5; when the Schottky
contact is satisfactory, repeat Section 11.4.2 to verify
that the phase angle is within the desired range.
11.5 Qualification of Schottky Contact — To qualify
the Schottky contact, determine the reverse current
characteristics of the mercury probe contact as follows:
11.5.1 If a curve tracer was used to determine the diode
forward resistance, do not disconnect it. If the series
resistance was measured directly, connect a curve tracer
or other apparatus for monitoring the current-voltage
characteristics of the mercury probe contact (see
Section 7.7). Apply a reverse bias voltage of about 1 V
to the mercury column. Warning: Avoid physical
contact with the probe fixture when bias is applied.
11.5.2 Measure and record this voltage as V
1
, and
measure the current that exists at this voltage.
Calculate the current density at this value of reverse
bias voltage, J
r1
, in mA·cm
−2
, as follows:
eff
r
r
A
I
J
1
1
= (10)
where:
I
r1
= current, mA, at the reverse bias voltage V
1
, and
A
eff
= mercury probe contact area, cm
2
, see Section
10.3.4.
11.5.3 Increase the magnitude of the reverse voltage at
intervals until the maximum reverse bias voltage that is
to be applied during the test (see Section 11.3) is
reached. Measure each current and calculate the
current density, J
r
, at each value of voltage. In
addition, calculate the rate of increase of the reverse
current density with voltage, in mA·V
−1
·cm
−2
, as
follows:
ii
riir
r
VV
JJ
V
J
−
−
=
∆
∆
+
+
1
)1(
(11)
where:
∆
J
r
/
∆
V
= rate of increase of the reverse current
density with voltage,
J
ri
= current density at voltage, V
i
, and
J
r(i+1)
= current density at voltage, V
i+ 1
.
11.5.4 Also observe whether or not the reverse current
density is stable with time.
11.5.5 If the reverse current density, J
r
, equals or
exceeds 3 mA/cm
2
at any voltage up to the maximum
value applied, first determine if this is due to carrier
density variations in the structure. In this case, reduce
the maximum applied reverse bias voltage to be used in
the test to the highest value for which the reverse
current density is less than 3 mA·cm
−2
.
NOTE 16: If the depletion depth extends to a region with a
rapidly increasing doping density, the breakdown voltage may
be significantly lower than estimated from the expected net
carrier density. For example, if the test specimen consists of a
lightly doped epitaxial layer on a heavily doped substrate and
if the profile extends deeper into the structure than the flat
region of the layer, the breakdown voltage would be lower
than that estimated from the expected net carrier density in the
flat region.
11.5.6 Otherwise, if J
r
> 3 mA·cm
−2
, or if ∆J
r
/∆V > 0.3
mA·V
−1
·cm
−2
, or if the reverse current is unstable in
time, treat the wafer surface with an acceptable
SEMI MF1392-1103 © SEMI 2003 12
chemical process (see Related Information 1), and
repeat the procedure beginning with Section 11.4.
11.5.7 When the Schottky contact is satisfactory,
disconnect the curve tracer or other apparatus for
monitoring the current-voltage characteristics of the
mercury probe contact.
NOTE 17: To ensure the greatest possible accuracy of the
measurement, monitor the attributes of the Schottky contact
during the collection of data (Section 11.6) and discontinue
data collection if any of the conditions in Section 11.5.6 are
violated.
11.6 Collection of Data — Measure a series of
capacitance-voltage pairs from which the net carrier
density profile can be calculated.
11.6.1 Disconnect the curve tracer from the mercury
probe fixture if one was used to determine the diode
current density.
11.6.2 Connect the capacitance bridge or meter to the
mercury probe fixture in accordance with the manu
-
facturer' s instructions and zero the capacitance bridge
or meter, if required.
11.6.3 Bring the mercury column(s) into contact with
the test wafer surface, and apply a nominal 1 V reverse
bias between the mercury probe contact and the return
contact (Warning—see Section 11.5.1).
11.6.4 Read and record the capacitance and applied
voltage, each to three or more significant figures, as C
1
and V
1
, respectively. Use a data table format
appropriate to the calculation method to be selected. In
all cases, record the voltages as positive numbers even
though reverse biases are involved.
11.6.5 Adjust the bias voltage to obtain a new value of
capacitance that is up to 5% lower than the previous
value. Read and record the capacitance and applied
voltage, each to three or more significant figures, as C
2
and V
2
, respectively (Warning—see Section 11.5.1).
11.6.6 Repeat Section 11.6.5, adjusting the voltage for
a decrease in capacitance such that the ratio of C
i+1
to
C
i
is approximately equal to C
2
/C
1
at each step, until the
maximum applied reverse bias voltage (see Section
11.3) is reached. Discontinue the measurement before
the maximum applied reverse bias voltage is reached if
the capacitance values start to increase with increased
reverse bias. Obtain a minimum of n capacitance-
voltage pairs, where n is sufficient to calculate at least
five values of net carrier density, N
i
, and depth, W
i
(see
Note 9).
11.7 Shut-Down Cycle — When the measurement
sequence is complete, reduce the reverse bias voltage to
0 V, disengage the mercury column(s) from the wafer
surface, and remove the test wafer from the probe
fixture.
12 Calculations
12.1 Subtract the compensation capacitance, C
comp
,
from each measured capacitance, C
′
i
, to obtain the
corrected capacitance, C
′
i
:
compii
CCC −=
′
(12)
Record these values of C
′
i
.
12.2 Calculate the net carrier density profile by the
Incremental Method (see Section 12.3) or by the Curve-
Fitting Method (see Section 12.4).
12.3 Incremental Method:
9
12.3.1 Calculate the average depletion depth
corresponding to each interval as follows:
12.3.1.1 Calculate and record the dimensionless
quantity S
i
for each value of i from 1 to n
−
k as follows:
′
′
+
+
=
+
+
ki
i
i
ki
i
C
C
V
V
S
ln
6.0
6.0
ln
(13)
where:
V
i
= i
th
recorded voltage, V,
V
i+k
= (i + k)
th
recorded voltage, V,
C
′
i
= i
th
corrected capacitance, pF,
C
′
i+k
= (i + k)
th
corrected capacitance, pF,
k = whole number such that C
i+k
is between 80 and
85% of C
i
, and
n = number of measured capacitance-voltage pairs.
12.3.1.2 Calculate and record the depth, W
i
′
, in µm,
corresponding to each corrected capacitance, C
i
′
, as
follows:
i
eff
i
C
A
W
′
=
′
10359 (14)
where:
A
eff
= mercury probe contact area, cm
2
, as determined
in 10.4.4, and
C
′
i
= the i
th
corrected capacitance, pF.
9 Niehaus, W. C., van Gelder, W., Jones, T. O., and Langer, P.,
“Variations of a Basic Capacitance-Voltage Technique for
Determination of Impurity Profiles in Semiconductors,”
Semiconductor Device Processing , NBS Spec. Publ. 337, (U.S.
National Institute of Standards and Technology, Washington, DC,
1970) pp. 266–268.