Measurements of characteristic transients of planar electrostatic probes in

cold plasmas
J. P. Booth, N. St. J. Braithwaite, A. Goodyear, and P. Barroy

Citation: Rev. Sci. Instrum. 71, 2722 (2000); doi: 10.1063/1.1150681

View online: http://dx.doi.org/10.1063/1.1150681
View Table of Contents: http://rsi.aip.org/resource/1/RSINAK/v71/i7
Published by the AIP Publishing LLC.

Additional information on Rev. Sci. Instrum.

Journal Homepage: http://rsi.aip.org
Journal Information: http://rsi.aip.org/about/about_the_journal
Top downloads: http://rsi.aip.org/features/most_downloaded
Information for Authors: http://rsi.aip.org/authors

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
REVIEW OF SCIENTIFIC INSTRUMENTS VOLUME 71, NUMBER 7 JULY 2000

Measurements of characteristic transients of planar electrostatic

probes in cold plasmas
J. P. Booth,a) N. St. J. Braithwaite, A. Goodyear,b) and P. Barroy
The Open University, Oxford Research Unit, Boars Hill, Oxford OX1 5HR, United Kingdom
共Received 29 December 1999; accepted for publication 15 March 2000兲
This article describes how to extract accurate information about a plasma from a capacitively
coupled planar probe that is biased using pulsed radio-frequency excitation. The conditions
necessary to observe correct saturation of the probe current are investigated, particularly the use of
correct geometry and biasing for the guard ring. With these precautions the probe is an effective
diagnostic for electron tail temperature at energies beyond those probed by conventional cylindrical
probes. The dynamic response of the probe is investigated using conventional sweep voltages and
shows the onset of displacement current and inertial effects associated with ions and electrons. In
addition the effect of insulating films on the probe surface is examined, showing how the probe
continues to operate even when it is coated. Characteristic changes caused by the presence of an
insulating film give information about its electrical properties and its thickness. © 2000 American
Institute of Physics. 关S0034-6748共00兲01707-X兴

I. INTRODUCTION probes. For cylindrical probes the ion current is some func-
tion of probe potential and must be modeled.
Electrostatic probe techniques are well established for Planar probes also have their disadvantages. Guarded
determining charged particle densities and electron energy planar configurations necessarily involve a relatively large
distribution functions 共EEDFs兲 in cold plasmas. The vast ma- area so they cannot be operated close to the plasma potential,
jority of measurements are made with small 共Ⰶ 1 mm diam兲
as the large current that is drawn causes severe perturbation
cylindrical probes, that are suitable for probing the EEDF in
of the plasma under investigation. This precludes measure-
the low energy region without, in principle, excessively per-
turbing the system under investigation. ment of the EEDF in the lower energy region.
Larger area planar probes have seen much less usage, Another disadvantage 共and possibly the reason why pla-
despite the fact that they have certain advantages over cylin- nar probes have been so infrequently used兲 is that it is diffi-
drical probes. They are particularly suitable for measuring cult to keep the large probe surface clean by the technique of
the current–voltage (I – V) characteristic in the positive ion intense electron current heating used routinely for smaller
current region. If careful attention is paid to assuring that the cylindrical probes. However, a novel technique has been
sheath in front of the probe is truly planar 共using a guard ring devised1,2 that allows planar probe I – V characteristics to be
to suppress edge effects兲, then the positive ion flux to the determined even with contaminated probe surfaces. The
surface, ⌫ i , is independent of the probe potential 共at least for probe is connected via a capacitance 共that can even include
applied voltages that do not significantly perturb the plasma兲. an insulating film on the probe itself兲; the characteristic is
In this case, when the applied voltage is sufficiently negative
determined through slow transient currents that charge and
to repel virtually all electrons, the net current collected satu-
discharge the capacitance. Prior to the measurement the
rates at a value equal to e⌫ i . For laboratory discharges, a
planar probe located in one of the confining surfaces there- probe is biased negatively with respect to its steady floating
fore gives a direct measurement of the ion flux to that sur- potential by the application of a short burst of rf potential.
face. In contrast, when using cylindrical probes this param- The nonlinearity of the sheath charges the capacitance—this
eter 共which is important for monitoring and controlling is the well known dc bias effect.3,4 As a result the rf applied
processing plasmas兲 can be deduced from the characteristics to the probe charges any capacitance in series with it. At the
only by the use of one of several contradictory models. end of the rf burst, the probe potential returns to its original
A further advantage of planar probes is that the I – V floating potential as the capacitance discharges, initially
characteristic in the electron retardation region gives a good through the arrival of positive charge from the plasma. Con-
measurement of the EEDF in the important high-energy tail sequently, only that part of the I – V characteristic more
region. This is because the positive ion current, which must negative than the floating potential is accessible. This rf burst
be subtracted, is constant in the case of guarded planar technique essentially autosweeps the bias from ion saturation
to floating potential.
a兲
Permanent address: Laboratoire Spectrométrie Physique, Université Joseph This article describes how to extract the most accurate
Fourier, Grenoble.
b兲
Author to whom correspondence should be addressed; electronic mail: information about the plasma from the observed signals and
A.Goodyear@open.ac.uk reports in more detail the limits of the technique. The condi-

0034-6748/2000/71(7)/2722/6/$17.00 2722 © 2000 American Institute of Physics

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
Rev. Sci. Instrum., Vol. 71, No. 7, July 2000 Probes in cold plasmas 2723

The normal operation of the probe involves the applica-

tion of a pulse modulated rf wave form. Following the ter-
mination of a pulse of rf the equivalent electrical circuit is
that shown in Fig. 1共b兲. The technique was originally devel-
oped to measure ion fluxes in capacitively coupled rf plasma
in a fluorocarbon gas.2 The films produced by this plasma
were relatively thin 共⬃10–100 nm兲, so that the film capaci-
tance C f was large, and the effective capacitance was deter-
mined by the external capacitor C x . Further refinements con-
cerning insulating films are discussed in Sec. III.
After the end of the rf burst used to bias the probe,
positive ion current flows to the probe surface and discharges
C x 共and C f 兲. The impedance of the rf generator is small 关in
Fig. 1共b兲 V B ⰆV C 兴 so that the current flowing can be deduced
simply from the observed voltage by
I 共 t 兲 ⫽C x dV C /dt, 共1兲
where V C is the voltage measured at point C in Fig. 1共a兲. The
I – V characteristic can then be obtained by plotting I against
V C . Differentiated signals are inherently noisy. An alterna-
FIG. 1. 共a兲 The measurement circuit; 共b兲 the equivalent electrical circuit
after the end of a rf burst. tive way to determine the current is also shown in Fig. 1.
Here the current returning to the rf generator side of C x is
deduced from the voltage V B , across the resistor R, provided
tions necessary to observe correct saturation of the probe
the generator impedance is small. The head-to-tail diodes are
current are investigated, particularly the use of correct geom-
necessary to pass the high-amplitude rf signal during the rf
etry and biasing for the guard ring. In addition the effect of
biasing step. The value of R 共⬃1 k⍀兲 is chosen so as to give
insulating films on the probe surface is examined, showing
an adequate signal 共⬃10–100 mV兲 without causing appre-
how the probe continues to operate even when it is coated.
ciable commutation of the diodes. This current measurement
Characteristic changes caused by the presence of an insulat-
technique gives significantly cleaner measurements than
ing film give information about its electrical properties and
those obtained by differentiating the voltage V C .
its thickness.
During the discharging of the rf self bias on the external
Experiments are reported in which the dc current to a
capacitance, the probe potential rises, coming closer to that
directly coupled uncoated probe is measured. This allows the
of the plasma. In this region although the electron flux is
validity of the capacitively coupled technique to be tested.
retarded by the probe potential, a flux of negative charge
Furthermore, it allows a systematic study of time-dependent
reaches the surface, offsetting a significant fraction of the
effects associated with sweep rates, including inertial effects
positive ion flux. Eventually the probe potential settles back
associated with ions and electrons. The onset of these effects
to the normal dc floating potential. At this point the retarded
defines the fastest voltage sweep rate and, consequently, the
electron flux just balances the steady positive ion flux. The
smallest capacitance or thickest film that should be used with
effective ‘‘tail’’ temperature of the negative charge distribu-
the capacitively coupled technique. The capacitively coupled
tion can be obtained from the I – V characteristic by fitting
technique is shown to be a superior way of measuring the
the electron current to that from a Maxwellian distribution.
I – V characteristics of a planar probe, as it is far simpler to
In general a satisfactory fitting is achieved to an equation of
implement and gives a significantly better signal-to-noise ra-
the form
tio when measuring small currents 共␮ A range兲 at bias volt-
ages up to several tens of volts. This allows the probe to be I⫽I 0 兵 1⫺exp关 e 共 V surf共 t 兲 ⫺V 0 兲 /kT 兴 其 , 共2兲
an effective diagnostic for electron tail temperature at ener-
gies beyond those probed by conventional probes. in which V 0 is the steady floating potential and V surf(t) is the
Finally, the probe method is shown to be useful in mea- potential on the probe surface. Temperature measurements
suring electron temperature in discharges, even through an- are reported in Sec. V.
odized surfaces and from behind a thin 共hundreds of nanom- In using this method, it is important to ensure that the
eters兲 glass coating. I – V data are free from the perturbations arising from inad-
equate matching of the guard ring 共Sec. IV兲 and current leak-
ing through voltage probes or parasitic capacitances.
II. THE CAPACITIVELY COUPLED rf-BIASED PROBE
When the ion current is relatively small, the current
TECHNIQUE
flowing through the voltage probe must be taken into ac-
The experimental configuration is shown in Fig. 1共a兲; the count. This effect can be minimized by using a high imped-
probe is normally a disk of a few millimeters in diameter ance oscilloscope probe. Alternatively the voltage probe can
with a concentric and coplanar guard ring that is indepen- be dispensed with entirely and the probe potential deter-
dently biased to the same potential as the probe. The guard mined from the integral of the probe current divided by the
ring is further discussed in Sec. IV. value of the external capacitance.

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
2724 Rev. Sci. Instrum., Vol. 71, No. 7, July 2000 Booth et al.

Parasitic capacitances arise on both sides of the external

capacitor 关see Fig. 1共b兲兴, owing to connecting cables. Both
contribute undesirable additional loads on the rf generator
and the use of short cables is good practice. In addition,
current through C p1 and C p2 is not detected by the resistive
current sensor, so the conditions ( ␻ C p1 ) ⫺1 ⰇR and C x
ⰇC p2 are necessary for that configuration.

III. THE EFFECT OF A DIELECTRIC FILM ON THE

PROBE SURFACE

Now let us consider what happens when there is a

thicker dielectric film on the probe, so that C f becomes com-
parable with, or smaller than, C x . The high frequency rf
signal used to bias the probe will be unaffected, provided the
effective capacitance, given by FIG. 2. The effect of an insulating layer on the probe: C2F6 RIE plasma, 50
mTorr 100 W, C x ⫽10 nF.
C fCx
C eff⫽ 共3兲
C f ⫹C x
For a film with dielectric constant ⑀ r and area A, it follows
does not become so small that it is comparable to the sheath that its thickness is given by

冉 冊
capacitance C sh . In this case, the potential of the film sur-
face at the end of the rf burst and the initial current to the ⌬V clean ␧ 0␧ rA
d⫽ ⫺1 . 共6兲
probe are just what they would be in the absence of the film. ⌬V dirty Cx
However, the dielectric film leads to a smaller effective ca-
Thus the film thickness can be deduced provided the dielec-
pacitance which therefore discharges more quickly. Further-
tric constant of the film is known. For the data shown in Fig.
more, the voltage observed externally V C is no longer equal
2, a value of about 5 nF is deduced for C f , indicating a film
to the voltage on the probe surface V surf but is related to it by
thickness of about 400 nm 共assuming the film has a dielectric
C f ⫹C x constant of 2.2, which is equal to that of polytetrafluoro eth-
V surf⫽V C . 共4兲 ylene兲. That the thickness of the film was of this order of
Cf
magnitude was evident also from the appearance of interfer-
The correct voltage scale for the I – V characteristic can only ence colors when viewed in white light.
be deduced if either the film is so thin that it can be ignored, A further complication occurs when the film acts as a
or if the film capacitance can be estimated. The thicker the leaky capacitance owing to incomplete coverage or conduc-
film, the less charge is built up by the rf burst in establishing tance of the film. In this case the potential across the film
a given surface potential, and the faster the surface returns to will decay with an RC time constant. Reliable measurements
its steady floating potential after the termination of the rf can still be made, provided a sufficiently small value of C x is
burst. Ultimately, very thick films on the probe will lead to chosen such that the potential drop across C f remains negli-
very fast probe voltage sweep rates. The thickest film that gible 关see Eq. 共4兲兴.
can be tolerated on the probe surface is therefore defined by
the maximum tolerable voltage sweep rate. This will be fur-
IV. PROBE CURRENT SATURATION AND EDGE
ther discussed in Sec. V. EFFECTS
Figure 2 shows the voltage and current signals observed
in a C2F6 plasma, both with a clean probe, and after 4 h of Planar probe I – V characteristics can only be correctly
plasma operation, by which time a significant film has been analyzed if the sheath in front of the probe is planar. Curva-
deposited. As expected, the initial dc bias that is observed ture of the boundary between the sheath and the plasma leads
externally for a fixed amount of rf bias is smaller when the to the effective area being dependent on probe potential. As
probe is contaminated, and the probe discharges faster. How- a result, the ion current collected at large negative bias does
ever, the initial current 共when the probe is in ion saturation兲 not achieve the constant value expected of a planar sheath.
is seen to be unchanged by the presence of the film. For probes of a diameter less than or comparable with the
The film capacitance can be estimated from the decrease sheath thickness, edge effects must be suppressed by includ-
in the observed initial bias. First define ⌬V clean as the exter- ing a concentric and coplanar guard ring that is indepen-
nal potential difference induced by the rf burst on a clean dently biased to the same potential as the probe throughout
probe and ⌬V dirty as that when the probe is coated. Assuming the measurement. In practice, the guard ring is biased by
that the film is not so thick as to affect the initial rf self bias applying the same rf burst that is used to bias the probe via a
of the film surface separate capacitor C g . The value of C g must be finely ad-
justed 共using a bank of switchable capacitors兲 so that its dis-
⌬V clean C f ⫹C x charge rate closely follows that of the probe. A clear indica-
⫽ . 共5兲
⌬V dirty Cf tion that the same rf burst has been applied to probe and

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
Rev. Sci. Instrum., Vol. 71, No. 7, July 2000 Probes in cold plasmas 2725

FIG. 3. I – V curves in the edge plasma of a 50 W RIE discharge in argon:

共a兲 7 mm diam probe with 24 mm diam guard ring, grounded and actively
matched to the probe 共400 mTorr兲, located in the remote edge plasma. 共b兲
Actively matched 24 mm guard ring and 7 mm probe embedded in the
counterelectrode at low 共50 mTorr兲 and high 共400 mTorr兲 pressure. FIG. 4. Experimental layout for the determination of probe current with
arbitrary voltage wave forms, incorporating active screen and noncontact
current sensing.
guard is when the initial rf self biases of the two are identi-
cal. To achieve this great care is necessary to match the
loading introduced by connecting cables to each. ally by the cable supplying the guard ring. Nevertheless, it
The guard ring must extend the probe sheath sideways at was still necessary to subtract the residual displacement cur-
least as far as the maximum sheath thickness, which is about rent associated with connections and cabling. It was also
one Debye length for every square root of normalized poten- necessary to enclose the whole measurement assembly in a
tial (eV/kT). Failure to achieve this design limit is evident shielded box and to filter out the high frequency component
in the I – V plot as an incomplete saturation of the ion current of the signal 共⬎1 MHz兲 on all but the most rapid sweep
at large bias. Figure 3共a兲 shows the effect of using only a experiments. In this way a resolution of ⫾0.1 ␮A was
grounded surface to suppress edge effects, compared with achieved.
one that is carefully matched in potential. The most appropriate wave form to use for comparison
Correct adjustment of the guard ring biasing is a delicate with the capacitively coupled probe technique described
task and is essential to obtaining reliable information about above is a negatively shifted triangle varying between float-
the EEDF. When it has been assured that edge effects have ing potential 共close to the grounded reactor wall potential兲
been properly suppressed, it is then reasonable to associate and a few tens of volts more negative. Any time-dependent
nonsaturating characteristics with the presence of higher en- phenomena which distort the I – V curve will lead to hyster-
ergy electrons. Figure 3共b兲 shows data taken with a probe esis in the current wave form.
and guard ring embedded in the counter electrode of a reac- Experiments were carried out in a 100 W RIE type
tive ion etching 共RIE兲 system. The curve at 50 mTorr shows plasma in 50 mTorr Ar, with the 共clean兲 probe situated at the
that there is a very high energy tail, probably due to second- radial edge of the plasma. A wide range of voltage sweep
ary electrons emitted from the rf powered electrode.5 Section times was used from 5 ms to 0.5 ␮s with a ramp between 0
VI gives further instances of electron temperature measure- and ⫺20 V. Figure 5共a兲 shows the current observed when the
ments. probe was biased by a slow triangle wave 共5 ms sweep time兲.
In this case no hysteresis is observed.
In contrast, Fig. 5共b兲 shows the results when the voltage
V. PROBE POTENTIAL SWEEP RATE EFFECTS
is swept in 5 ␮s. Marked hysteresis is observed here, and can
As noted in Sec. III, the thicker the 共insulating兲 film on be attributed to two causes. First, displacement current must
the probe surface the smaller the value of C eff is, and the be supplied to charge the sheath capacitance. This changes
faster the surface potential then changes with respect to the sign as the voltage sweep direction changes. For a sheath
plasma following the rf pulse. This can ultimately lead to with a potential drop of 20 V the sheath thickness can be
distortion of the I – V characteristic when the probe sweeps estimated from Child’s law to be about 1 mm, giving a
itself too rapidly from ion saturation back to floating poten- sheath capacitance of about 1 pF cm⫺2. The voltage slew rate
tial. is 4⫻106 V s⫺1, giving displacement currents of
To investigate the effect of different probe voltage ⫾4 ␮A cm⫺2, entirely consistent with the step observed as
sweep rates, triangular voltage wave forms were applied di- the sweep changes direction 共of order 10 ␮A cm⫺2兲. Second,
rectly to a clean probe in an argon plasma. This avoids the the ion density profile will not remain in equilibrium when
complication of film coatings and permits the probe potential the sweep rate is too fast. This results in a drop in the posi-
to be swept in an arbitrary fashion. Figure 4 shows the ex- tive ion current as the voltage is swept from a negative po-
perimental arrangement. tential toward floating potential 共collapsing sheath兲, and an
The current was measured with a Tektronix A6302 increase in the positive ion current as the sheath expands.
probe with an AM503B amplifier. This unit combines a These effects will appear when the voltage sweep time be-
transformer-coupled measurement of the faster components comes comparable to the inverse of the ion–plasma fre-
of current with a Hall sensor for slower and dc components. quency (M ␧ 0 /ne 2 ) 1/2. This criterion suggests that sweeps
To minimize the capacitance to ground the wire from the should last longer than 1 ␮s or so for Ar⫹ at 1015 m⫺3.
measurement point to the probe surface was shielded coaxi- Figure 5共c兲 shows the results obtained with even shorter

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
2726 Rev. Sci. Instrum., Vol. 71, No. 7, July 2000 Booth et al.

FIG. 7. Ion flux 共a兲 and temperature of the tail of the electron distribution
共b兲 for a range of conditions in an asymmetrical parallel plate argon plasma.

ing Eq. 共2兲. For the lowest values of C x used, the effect of
the parasitic capacitance C p2 is seen in an apparent reduction
of the saturation current. Although this can be corrected for
if the value of C p2 is known, proper account must be taken
of this effect before displacement currents and ion dynamics
become important issues.

VI. ELECTRON TEMPERATURE MEASUREMENTS

Figure 7 shows a map of the power–pressure parameter
space for a RIE type rf discharge in argon. Ion flux is shown
in Fig. 7共a兲 and the tail electron temperature is shown in Fig.
7共b兲; both were determined from the rf biased probe, using
Eq. 共2兲. The electron temperature compares favorably with
that deduced from data obtained with a cylindrical Langmuir
FIG. 5. Current data with triangular voltage wave forms applied to a probe probe. It is important to note that the present method is par-
共with guard ring兲 in a 100 W discharge in 50 mTorr of argon: 共a兲 5 ms ramp; ticularly sensitive to electrons energetic enough to reach
共b兲 5 ␮s ramp; 共c兲 0.5 ␮s ramp. floating surfaces; this is also the group responsible for inelas-
tic processes. For cylindrical Langmuir probes, the accuracy
sweep times 共0.5 ␮s兲 In this case, the discontinuity as the of data for this group of electrons is always compromised by
voltage sweep changes direction induces a plasma sheath os- the need to model the ion current as a function of probe
cillation, with a frequency of about 25 MHz,6,7 leading to potential, severely limiting the sensitivity to electrons with
severe distortion of the I – V characteristic. energies beyond that corresponding with the floating poten-
Returning to the pulsed bias scheme, the voltage slew tial 共typically more than 10 eV兲. When correctly guarded, the
rate increases when smaller values of external capacitance rf biased probe starts at this ‘‘floating potential’’ energy and
C x are used 关see Eq. 共1兲兴. A further source of infidelity in the remains sensitive to higher energies limited only by the reso-
measured current then arises from the effects of parasitic lution of the current sensor.
capacitance 关C p2 in Fig. 1共b兲兴. Figure 6 shows the results The utility of the rf biased method is demonstrated by
obtained in an argon RIE type discharge at 400 mTorr and the data in Fig. 8. This shows an I – V curve for a glass
50 W. The voltage sweep times varied from 10 ms (C x coated probe installed in an earthed surface and exposed to
⫽3.8 nF) to 0.5 ms (C x ⫽0.22 nF). The same electron tem- the same argon plasma as above. The coating of silica glass,
perature was obtained for all curves (T e ⫽1.7⫾0.1 eV), us-

FIG. 8. I – V curve for a coated planar probe in a 100 mTorr 25 W argon

FIG. 6. Probe I – V characteristics obtained with different values of the plasma. The probe was coated with 400 nm of SiO2 by Applied Vision Ltd.,
blocking capacitor C x . UK.

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions
Rev. Sci. Instrum., Vol. 71, No. 7, July 2000 Probes in cold plasmas 2727

400 nm thick, had been deposited over the probe and guard tance. The effect of insulating films on the probe surface has
ring ex situ, using a sputter deposition process.8 Note that the been examined showing that the probe continues to operate
voltage scale is in terms of the potential on the surface ex- even when it is coated. Characteristic changes caused by the
posed to the plasma—see Sec. III. The parameters from Fig. presence of an insulating film give information about its
8 are well fitted by a Maxwellian electron tail population and electrical properties and its thickness.
agree with those in Fig. 7.
ACKNOWLEDGMENT
DISCUSSION
This work was carried out while J.P.B. was a visiting
This article has described a pulsed rf bias technique for a fellow at the Open University, funded by the EPSRC, UK
planar probe and how to extract accurate information with it 共Grant No. GR/L78864兲.
about a plasma. The conditions necessary to observe correct
saturation of the probe current have been reported, justifying 1
N. St. J. Braithwaite and J. P. Booth, Patent No. FR 96/01451, France
proper geometry and biasing of the guard ring. The probe has 共1995兲.
2
N. St. J. Braithwaite, J. P. Booth, and G. Cunge, Plasma Sources Sci.
been shown to be an effective diagnostic for electron tail
Technol. 5, 677 共1996兲.
temperature at energies beyond those probed by conventional 3
K. Kohler, J. W. Coburn, D. E. Horne, E. Kay, and J. H. Keller, J. Appl.
probes. The dynamic response of the probe using conven- Phys. 57, 59 共1985兲.
tional sweep voltages showed the onset of inertial effects
4
A. Boschi and F. Magistrelli, Nuovo Cimento 29, 487 共1963兲.
5
S. G. Ingram and N. St. J. Braithwaite, J. Phys. D 21, 1469 共1988兲.
associated with ions and electrons, and displacement current 6
F. W. Crawford, J. Appl. Phys. 34, 1897 共1963兲.
due to the sheath capacitance. In practice though, the rf pulse 7
M. Klick, J. Appl. Phys. 79, 3445 共1996兲.
8
method was limited to longer time scales by stray capaci- Coating supplied by Applied Vision, Coalville, UK.

Downloaded 23 Jul 2013 to 128.118.88.48. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://rsi.aip.org/about/rights_and_permissions