Study of the SiO2-to-Si3N4 etch selectivity mechanism in inductively
`coupled fluorocarbon plasmas and a comparison
`with the SiO2-to-Si mechanism
`M. Schaepkens, T. E. F. M. Standaert, N. R. Rueger, P. G. M. Sebel,a)
`and G. S. Oehrleinb)
`Department of Physics, University at Albany, State University of New York, Albany, New York 12222
`J. M. Cook
`Lam Research Corporation, Fremont, California 94538-6470
`共Received 21 July 1998; accepted 2 October 1998兲
`The mechanisms underlying selective etching of a SiO2 layer over a Si or Si3N4 underlayer, a
`process of vital importance to modern integrated circuit fabrication technology, has been studied.
`Selective etching of SiO2-to-Si3N4 in various inductively coupled fluorocarbon plasmas (CHF3 ,
`C2F6/C3F6 , and C3F6/H2) was performed, and the results compared to selective SiO2-to-Si etching.
`A fluorocarbon film is present on the surfaces of all investigated substrate materials during steady
`state etching conditions. A general trend is that the substrate etch rate is inversely proportional to the
`thickness of this fluorocarbon film. Oxide substrates are covered with a thin fluorocarbon film 共⬍1.5
`nm兲 during steady-state etching and at sufficiently high self-bias voltages, the oxide etch rates are
`found to be roughly independent of the feedgas chemistry. The fluorocarbon film thicknesses on
`silicon, on the other hand, are strongly dependent on the feedgas chemistry and range from ⬃2 to
`⬃7 nm in the investigated process regime. The fluorocarbon film thickness on nitride is found to be
`intermediate between the oxide and silicon cases. The fluorocarbon film thicknesses on nitride range
`from ⬃1 to ⬃4 nm and the etch rates appear to be dependent on the feedgas chemistry only for
`specific conditions. The differences in etching behavior of SiO2 , Si3N4 , and Si are suggested to be
`related to a substrate-specific ability to consume carbon during etching reactions. Carbon
`consumption affects the balance between fluorocarbon deposition and fluorocarbon etching, which
`controls the fluorocarbon steady-state thickness and ultimately the substrate etching. © 1999
`American Vacuum Society. 关S0734-2101共99兲03201-7兴
`
`I. INTRODUCTION
`
`Etching of via or contact holes into SiO2 to make electri-
`cal contact with an underlayer is an indispensable process in
`modern integrated circuit fabrication technology. High SiO2
`etch rate and selectivity of SiO2-to-Si are important require-
`ments for etch processes to be commercially viable in manu-
`facturing. Etch processes employing fluorocarbon discharges
`are typically able to meet these demands, as first reported by
`Heinecke1
`and an extensive number of
`subsequent
`studies.2–12
`It is believed that the primary mechanism for highly se-
`lective SiO2-to-Si etching using fluorocarbon plasmas is the
`selective formation of a relatively thick passivating film on
`the Si surface during steady-state etching conditions. The Si
`etch rate in that situation is limited by the arrival of atomic
`fluorine that needs to diffuse through the film to the Si sur-
`face, where it chemically reacts.13–16 For the same process
`conditions SiO2 surfaces stay clean of fluorocarbon material,
`and are etched directly through a mechanism of chemical
`sputtering.6,17,18
`The ability to achieve selective etching of SiO2 over
`Si3N4 is becoming an increasingly important requirement.
`
`a兲On leave from Eindhoven University of Technology.
`b兲Electronic mail: oehrlein@csc.albany.edu
`J. Vac. Sci. Technol. A 17„1…, Jan/Feb 1999
`
`26
`
`Silicon nitride is used as a passivating layer that protects
`circuits from mechanical and chemical attack, or as an etch
`stop layer, enabling the fabrication of certain damascene and
`self-aligned contact 共SAC兲 structures. Selective SiO2-to-
`Si3N4 etching has been demonstrated in several systems.19–24
`Correlations between the Si3N4 etch rate and the amount of
`fluorocarbon material present on the surface during etching
`suggest a SiO2-to-Si3N4 selectivity mechanism that is analo-
`gous to the SiO2-to-Si etching mechanism. A detailed com-
`parison between the two mechanisms, however, is lacking.
`This work summarizes results obtained in a study where
`SiO2 , Si3N4 and Si were processed in an inductively coupled
`plasma source fed with various fluorocarbon feedgas chem-
`istries (CHF3 , C2F6/C3F6 and C3F6/H2). Etch rates of SiO2 ,
`Si3N4 , and poly-Si samples and surface modifications of
`crystalline Si samples were measured using in situ ellipsom-
`etry. The surface chemistry of processed SiO2 and Si3N4
`samples was examined using postplasma in situ x-ray photo-
`electron spectroscopy 共XPS兲. The experimental results allow
`a direct comparison of SiO2 , Si3N4 , and Si etch mechanisms
`the SiO2-to-Si3N4 and
`and therefore a comparison of
`SiO2-to-Si etch selectivity mechanisms. From this compari-
`IP Bridge Exhibit 2226
`son it can be understood why certain feedgas chemistries that
`TSMC v. Godo Kaisha IP Bridge 1
`give SiO2-to-Si selectivity do not necessarily give SiO2-to-
`IPR2017-01843
`Si3N4 selectivity. However, it has been found that SiO2-to-
`0734-2101/99/17„1…/26/12/$15.00
`
`©1999 American Vacuum Society
`
`26
`
`

`

`Study of the SiO2-to-Si3N4 etch selectivity mechanism in inductively
`coupled fluorocarbon plasmas and a comparison
`with the SiO2-to-Si mechanism
`M. Schaepkens, T. E. F. M. Standaert, N. R. Rueger, P. G. M. Sebel,a)
`and G. S. Oehrleinb)
`Department of Physics, University at Albany, State University of New York, Albany, New York 12222
`J. M. Cook
`Lam Research Corporation, Fremont, California 94538-6470
`共Received 21 July 1998; accepted 2 October 1998兲
`
`The mechanisms underlying selective etching of a SiO2 layer over a Si or Si3N4 underlayer, a
`process of vital importance to modern integrated circuit fabrication technology, has been studied.
`Selective etching of SiO2-to-Si3N4 in various inductively coupled fluorocarbon plasmas (CHF3 ,
`C2F6/C3F6 , and C3F6/H2) was performed, and the results compared to selective SiO2-to-Si etching.
`A fluorocarbon film is present on the surfaces of all investigated substrate materials during steady
`state etching conditions. A general trend is that the substrate etch rate is inversely proportional to the
`thickness of this fluorocarbon film. Oxide substrates are covered with a thin fluorocarbon film 共⬍1.5
`nm兲 during steady-state etching and at sufficiently high self-bias voltages, the oxide etch rates are
`found to be roughly independent of the feedgas chemistry. The fluorocarbon film thicknesses on
`silicon, on the other hand, are strongly dependent on the feedgas chemistry and range from ⬃2 to
`⬃7 nm in the investigated process regime. The fluorocarbon film thickness on nitride is found to be
`intermediate between the oxide and silicon cases. The fluorocarbon film thicknesses on nitride range
`from ⬃1 to ⬃4 nm and the etch rates appear to be dependent on the feedgas chemistry only for
`specific conditions. The differences in etching behavior of SiO2 , Si3N4 , and Si are suggested to be
`related to a substrate-specific ability to consume carbon during etching reactions. Carbon
`consumption affects the balance between fluorocarbon deposition and fluorocarbon etching, which
`controls the fluorocarbon steady-state thickness and ultimately the substrate etching. © 1999
`American Vacuum Society. 关S0734-2101共99兲03201-7兴
`
`I. INTRODUCTION
`
`Etching of via or contact holes into SiO2 to make electri-
`cal contact with an underlayer is an indispensable process in
`modern integrated circuit fabrication technology. High SiO2
`etch rate and selectivity of SiO2-to-Si are important require-
`ments for etch processes to be commercially viable in manu-
`facturing. Etch processes employing fluorocarbon discharges
`are typically able to meet these demands, as first reported by
`Heinecke1
`and an extensive number of
`subsequent
`studies.2–12
`It is believed that the primary mechanism for highly se-
`lective SiO2-to-Si etching using fluorocarbon plasmas is the
`selective formation of a relatively thick passivating film on
`the Si surface during steady-state etching conditions. The Si
`etch rate in that situation is limited by the arrival of atomic
`fluorine that needs to diffuse through the film to the Si sur-
`face, where it chemically reacts.13–16 For the same process
`conditions SiO2 surfaces stay clean of fluorocarbon material,
`and are etched directly through a mechanism of chemical
`sputtering.6,17,18
`The ability to achieve selective etching of SiO2 over
`Si3N4 is becoming an increasingly important requirement.
`
`a兲On leave from Eindhoven University of Technology.
`b兲Electronic mail: oehrlein@csc.albany.edu
`
`Silicon nitride is used as a passivating layer that protects
`circuits from mechanical and chemical attack, or as an etch
`stop layer, enabling the fabrication of certain damascene and
`self-aligned contact 共SAC兲 structures. Selective SiO2-to-
`Si3N4 etching has been demonstrated in several systems.19–24
`Correlations between the Si3N4 etch rate and the amount of
`fluorocarbon material present on the surface during etching
`suggest a SiO2-to-Si3N4 selectivity mechanism that is analo-
`gous to the SiO2-to-Si etching mechanism. A detailed com-
`parison between the two mechanisms, however, is lacking.
`This work summarizes results obtained in a study where
`SiO2 , Si3N4 and Si were processed in an inductively coupled
`plasma source fed with various fluorocarbon feedgas chem-
`istries (CHF3 , C2F6/C3F6 and C3F6/H2). Etch rates of SiO2 ,
`Si3N4 , and poly-Si samples and surface modifications of
`crystalline Si samples were measured using in situ ellipsom-
`etry. The surface chemistry of processed SiO2 and Si3N4
`samples was examined using postplasma in situ x-ray photo-
`electron spectroscopy 共XPS兲. The experimental results allow
`a direct comparison of SiO2 , Si3N4 , and Si etch mechanisms
`the SiO2-to-Si3N4 and
`and therefore a comparison of
`SiO2-to-Si etch selectivity mechanisms. From this compari-
`son it can be understood why certain feedgas chemistries that
`give SiO2-to-Si selectivity do not necessarily give SiO2-to-
`Si3N4 selectivity. However, it has been found that SiO2-to-
`
`26
`
`J. Vac. Sci. Technol. A 17(cid:132)1(cid:133), Jan/Feb 1999
`
`0734-2101/99/17(cid:132)1(cid:133)/26/12/$15.00
`
`©1999 American Vacuum Society
`
`26
`
`

`

`27
`
`Schaepkens etal.: Study of the SiO2-to-Si3N4 etch selectivity mechanism
`
`27
`
`A variable frequency rf power supply 共500 kHz–40 MHz,
`0–300 W兲 is used to bias the wafer for etching experiments.
`The experiments reported in this work were all performed at
`3.4 MHz. In the process regime investigated, no significant
`influence of rf biasing on the ion current density measured
`with the Langmuir probe was observed. The ion energy and
`the ion flux to the surfaces can thus be varied independently.
`The process chamber is pumped using a 450 l/s turbomo-
`lecular pump backed by a roughing pumpstack, consisting of
`a roots blower and a mechanical pump. The process gases
`are admitted into the reactor through a gas inlet ring located
`just under the quartz window. The pressure is measured with
`a capacitance manometer. Pressure control is achieved by an
`automatic throttle valve in the pump line.
`In order to remove deposited fluorocarbon films from the
`walls of the process chamber, the chamber is cleaned with an
`O2 plasma after each experiment. The cleaning process is
`monitored by taking real time OES data. The absence of
`optical emission of fluorocarbon related gas phase species is
`a measure for the cleanliness of the chamber.
`The ICP apparatus is connected to a wafer handling clus-
`ter system. Processed samples were transported under UHV
`conditions to a Vacuum Generators ESCA Mk II surface
`analysis chamber
`for x-ray photoelectron spectroscopy
`共XPS兲 analysis using a polychromatic Mg K␣ x-ray source
`共1253.6 eV兲. Photoelectron spectra were obtained at photo-
`emission angles of 90° and 15° with respect to the sample
`surface.
`
`III. EXPERIMENTAL RESULTS
`
`A. Fluorocarbon deposition and etch rates
`
`In fluorocarbon plasma processing, deposition and etching
`occur simultaneously. Deposition and etching processes have
`different dependencies on the energy of the ions that bom-
`bard the surface on which they occur. By varying the ion
`bombardment energy one can therefore switch from fluoro-
`carbon deposition to etching of earlier deposited fluorocar-
`bon material or other substrate materials, such as SiO2 ,
`Si3N4 , or Si.
`At low self-bias voltages, i.e., low ion bombardment en-
`ergy, these processes produce net fluorocarbon deposition.
`The deposition characteristics are strongly dependent on the
`feedgas chemistry. Figure 2共a兲 shows the fluorocarbon depo-
`sition rates obtained as a function of feedgas (CHF3, C2F6 ,
`C3F6 , and C3F6/H2) in discharges at 6 mTorr operating pres-
`sure, 1400 W inductive power, and no rf bias power applied
`to the surface. The deposition rate in CHF3 and C2F6 is sig-
`nificantly lower than in C3F6 or C3F6/H2 . The highly poly-
`merizing character of C3F6 or C3F6/H2 discharges may be
`related to the size and chemical structure of the fluorocarbon
`parent molecules.2,24
`Figure 2共b兲 shows the refractive index of the deposited
`fluorocarbon films using the different gases. The refractive
`index of the fluorocarbon deposited using C2F6 and C3F6 are
`found to be lower than if CHF3 or C3F6/H2 is used. Crystal-
`line Si samples on which 150 nm fluorocarbon material was
`deposited were analyzed by XPS. It was found that the
`
`FIG. 1. Schematic outline of the experimental setup of the used ICP source.
`
`selectivity in an etching process also provides
`Si3N4
`SiO2-to-Si selectivity. More importantly, a trend in the etch
`rate behavior of the different materials has been identified
`allowing a general description of fluorocarbon plasma etch-
`ing to be formulated.
`
`II. EXPERIMENTAL SETUP
`The high-density plasma source used in this work is a
`radio-frequency inductively coupled plasma 共ICP兲 source of
`planar coil design. This plasma source has also been referred
`to in the literature as transformer coupled plasma 共TCP兲25
`and radio frequency induction 共RFI兲26 source. A schematic
`outline of the used ICP reactor is shown in Fig. 1. It is
`similar to the one described by Keller et al.27
`The apparatus consists of an ultrahigh vacuum 共UHV兲
`compatible processing chamber in which the plasma source
`and a wafer holding electrostatic chuck are located. The cen-
`ter part of the ICP source is a planar, 160-mm-diam induc-
`tion coil that is separated from the process chamber by a
`19.6-mm-thick, 230-mm-diam quartz window. The coil is
`powered through a matching network by a 13.56 MHz, 0–
`2000 W power supply.
`Wafers with a diameter of 125 mm can be placed on a
`bipolar electrostatic chuck during processing. The chuck is
`located at a distance of 7 cm downstream from the ICP
`source and allows the wafer to be rf biased and cooled during
`processing. A helium pressure of 5 Torr is applied to the
`backside of the wafer during the experiment to achieve good
`thermal conduction between the wafer and the chuck.28 The
`chuck is cooled by circulation of a cooling liquid.
`Samples placed at the center of a wafer can be monitored
`by an in situ He–Ne 共632.8 nm兲 rotating compensator ellip-
`someter 共RCE兲 in a polarizer-compensator-sample-analyzer
`共PCSA兲 configuration. Plasma diagnostics like a retractable
`Langmuir probe and optical emission spectroscopy 共OES兲
`can be used for plasma gas-phase characterization. With the
`retractable Langmuir probe it is possible to make a scan of
`the ion current density over 70% of the wafer at a distance of
`2 cm above the wafer surface.
`
`JVST A - Vacuum, Surfaces, and Films
`
`

`

`28
`
`Schaepkens etal.: Study of the SiO2-to-Si3N4 etch selectivity mechanism
`
`28
`
`FIG. 2. 共a兲 Deposition rates, 共b兲 refractive index, and 共c兲 fluorine-to-carbon
`ratio determined from XPS analysis of fluorocarbon material deposited at 6
`mTorr operating pressure in discharges fed with 40 sccm of CHF3 , C2F6 ,
`C3F6 , or C3F6/H2 共27%兲. The fluorine-to-carbon ratios determined under
`15° electron escape angle are slightly higher than the ratio determined under
`90°. The values in 共c兲 are averages of the two values.
`
`FIG. 3. Fluorocarbon etch rates as a function of self-bias voltage, measured
`with respect to ground, at 6 mTorr operating pressure in discharges fed with
`40 sccm of CHF3 , C2F6 , C3F6 , or C3F6/H2 共27%兲. The fluorocarbon sub-
`strate was deposited at 0 W rf bias power at the same process condition
`where the etch rates were determined. The plasma potential at these condi-
`tions typically varies between 20 and 30 V. This potential needs to be added
`to the self-bias voltage in order to estimate the actual average ion bombard-
`ment energy.
`
`fluorine-to-carbon ratio of the fluorocarbon material, see Fig.
`2共c兲, is inversely proportional to the refractive index. Fur-
`ther, the fluorine-to-carbon ratio of the deposited fluorocar-
`bon material is lower if hydrogen is present in the feedgas
`chemistry. This trend can be explained by the fluorine scav-
`enging effect of hydrogen in the gas phase,2 or more likely at
`these operating pressures fluorine reduction as a result of
`hydrogen–fluorine recombination at the reactor walls. Also,
`fluorine abstraction by hydrogen from a fluorocarbon surface
`could explain the observations.29
`The fluorocarbon films that deposit if no bias is applied to
`the substrate, can be etched off the thin film substrate at
`self-bias voltages above a certain threshold value. The etch
`rates of the fluorocarbon films deposited at unbiased condi-
`tions have been measured as a function of self-bias voltage
`by in situ ellipsometry in discharges of CHF3, C2F6 , C3F6 ,
`and C3F6/H2 at 6 mTorr operating pressure and 1400 W in-
`ductive power. The fluorocarbon etch rates are plotted as
`positive values in Fig. 3. Also included in Fig. 3 are the
`fluorocarbon deposition rates 共negative values兲 measured by
`in situ ellipsometry at self-bias voltages below the threshold
`for etching. It shows that the fluorocarbon deposition rate
`decreases as the self-bias voltage increases.
`The fluorocarbon etch rates are found to be strongly de-
`pendent on the feedgas chemistry. First, the fluorocarbon
`etch rate is relatively low at conditions where hydrogen is
`present in the feedgas mixture, e.g., compare CHF3 to C2F6
`processing and C3F6 to C3F6/H2 processing. This observation
`can be attributed to the fact that fluorine is a precursor for
`etching of fluorocarbon material.3 The presence of hydrogen
`in the feedgas chemistry namely results in 共a兲 a reduction of
`fluorine in the plasma gas phase 共observed when comparing
`optical emission spectra from 40 sccm CF4 and 40 sccm
`CHF3 plasmas at identical conditions兲, and 共b兲 a more fluo-
`
`J. Vac. Sci. Technol. A, Vol. 17, No. 1, Jan/Feb 1999
`
`rine depleted fluorocarbon substrate 共the fluorine-to-carbon
`ratio of fluorocarbon is reduced, see Fig. 2兲. A second obser-
`vation is that the higher the fluorocarbon deposition rate at 0
`W bias power, the lower the fluorocarbon etch rate under
`biased conditions, e.g., compare C2F6 to C3F6 processing and
`CHF3 to C3F6/H2 processing. Net fluorocarbon etching ap-
`parently benefits from a reduction in fluorocarbon deposi-
`tion.
`The position of the threshold voltage for net etching is
`consistent with the above observations.
`If hydrogen is
`present in the feedgas chemistry or if the fluorocarbon depo-
`sition rate at 0 W bias power is high, the threshold voltage is
`relatively high, and vice versa.
`
`B. SiO2 , Si3N4 , and Si etch rates
`At conditions where net fluorocarbon etching takes place,
`other substrate materials, such as SiO2 , Si3N4 , and Si, can
`also be etched. Figure 4 shows the SiO2 , Si3N4 , and Si etch
`rates measured on blanket samples by in situ ellipsometry as
`a function of self-bias voltage in discharges of CHF3, C2F6 ,
`C3F6 , and C3F6/H2 at 6 mTorr operating pressure and 1400
`W inductive power.
`At sufficiently high self-bias voltage, SiO2 etching of the
`blanket samples occurs at a relatively high rate, which is
`roughly independent of the feedgas chemistry. This is con-
`sistent with Langmuir probe measurements performed for the
`different discharges. The Langmuir probe data showed that
`the ion current density does not vary significantly with
`feedgas chemistry. Since SiO2 etching has been suggested to
`occur through a chemical sputtering mechanism in which the
`ion flux is the limiting factor,18 the SiO2 etch rate is expected
`to be relatively independent of the feedgas chemistry. 共Note:
`this is the case if the average composition of the fluorocar-
`
`

`

`29
`
`Schaepkens etal.: Study of the SiO2-to-Si3N4 etch selectivity mechanism
`
`29
`
`FIG. 4. Etch rates of SiO2 , Si3N4 , and Si substrates as a function of self-bias
`voltage at 6 mTorr operating pressure in discharges fed with 40 sccm of
`CHF3 , C2F6 , C3F6 , or C3F6/H2 共27%兲.
`
`bon ion flux that bombards the SiO2 surface does not change
`significantly with the various conditions. If the average com-
`⫹ to CF⫹, the
`position of the ion flux changes, e.g., from CF2
`average SiO2 sputter yield possibly changes as well.兲
`The blanket Si etch rates for the above conditions are
`significantly lower than the SiO2 etch rates. The etch rates of
`silicon show a strong dependence on feedgas chemistry. The
`dependence is similar to that observed for fluorocarbon etch-
`ing, suggesting that the effects due to hydrogen addition and
`fluorocarbon deposition rate ultimately are also responsible
`for the etch rate behavior of Si.
`Etching of Si3N4 is intermediate between SiO2 and Si,
`both in etch rate and dependence on the feedgas chemistry.
`For chemistries that result in a low fluorocarbon deposition
`rate, the Si3N4 etch rate is relatively independent of the
`feedgas, similar to SiO2 etching. For chemistries resulting in
`a high fluorocarbon deposition rate, the Si3N4 etch rate de-
`pendence is similar to that observed for Si and fluorocarbon
`etching. In other words, hydrogen addition only helps to sup-
`press Si3N4 etching at conditions where the fluorocarbon
`deposition rate is sufficiently high.
`
`C. Surface analysis: Steady-state fluorocarbon films
`
`The surface chemistry of processed SiO2 and Si3N4
`samples was investigated by XPS as a function feedgas
`chemistry in discharges at 6 and 20 mTorr operating pres-
`sure. The feedgas chemistries used were CHF3 , C2F6 , C3F6 ,
`and mixtures of C2F6/C3F6 and C3F6/H2 . At the same condi-
`tions the surface modifications of crystalline Si samples were
`measured by in situ ellipsometry.30
`Figure 5 shows high-resolution C 共1s兲 photoemission
`spectra obtained from SiO2 and Si3N4 surfaces etched at
`⫺100 V self-bias voltage at 6 mTorr operating pressure in
`
`JVST A - Vacuum, Surfaces, and Films
`
`FIG. 5. C共1s兲 spectra obtained by XPS surface analysis under 90° emission
`angle on SiO2 and Si3N4 samples processed in CHF3 共40 sccm兲, C2F6 共40
`sccm兲, C3F6 共40 sccm兲, and C3F6/H2 共20 sccm/15 sccm兲 discharges at 6
`mTorr and a self-bias voltage of ⫺100 V. The C 共1s兲 intensity is a measure
`for the amount of fluorocarbon material present on the surfaces during
`steady-state etching conditions.
`
`discharges fed with CHF3 , C2F6 , C3F6 , and C3F6 /H2 . The
`binding energies were corrected for charging of the dielectric
`substrate materials. The integrated intensity of the C 共1s兲
`spectrum, plotted in the upper left corner of each graph in
`Fig. 5, is a measure for the amount of fluorocarbon material
`present on the sample surface. It shows that the C 共1s兲 inten-
`sity from processed SiO2 surfaces is typically lower than the
`intensities measured on Si3N4 samples processed under the
`same process conditions, and compared to processed Si3N4
`samples relatively independent of the feedgas chemistry. The
`intensity of the C 共1s兲 spectrum from processed Si3N4 de-
`pends significantly on the feedgas chemistry. It clearly shows
`that the amount of fluorocarbon material on the Si3N4 surface
`shown in Fig. 5, is inversely proportional to the correspond-
`ing etch rate at ⫺100 V self-bias from Fig. 4.
`The C 共1s兲 intensity measured on the processed surfaces
`can be expressed as a fluorocarbon film thickness. This
`thickness can be calculated from photoemission intensities,
`due to the exponential decay of the XPS signal with depth. A
`method that uses angular resolved intensities from substrate
`elements is described by Rueger et al.18 Standaert et al.16
`describe a method of quantifying the fluorocarbon film thick-
`ness by comparing the C 共1s兲 intensity from a processed
`sample to the C 共1s兲 intensity of a semi-infinitely thick fluo-
`rocarbon film 共e.g., a fluorocarbon film deposited at 0 W兲. A
`comparison of the two methods showed that at low C 共1s兲
`intensities, both methods give very similar values for the
`fluorocarbon thickness. At high C 共1s兲 intensities, slightly
`
`

`

`30
`
`Schaepkens etal.: Study of the SiO2-to-Si3N4 etch selectivity mechanism
`
`30
`
`to be limited by the arrival of atomic fluorine at the Si sur-
`face. Since the fluorine needs to be transported through the
`relatively
`thick
`fluorocarbon
`film by
`a
`diffusion
`mechanism;13–16 the Si etch rates are suppressed. The fluo-
`rocarbon films on Si3N4 are of intermediate thickness com-
`pared to SiO2 and Si 共1.25–3.75 nm兲. For certain conditions
`a thick fluorocarbon film is present and the Si3N4 etch rate is
`suppressed, similarly to the Si etch rate. For other conditions
`the fluorocarbon films are relatively thin, and the Si3N4 etch
`behavior is similar to that of SiO2 .
`From Fig. 6 it is clear that SiO2-to-Si3N4 selectivity can
`be achieved only if a thick enough fluorocarbon film forms
`on the Si3N4 surface, while keeping the SiO2 surface clean.
`This is completely analogous to the SiO2-to-Si selectivity
`mechanism. However, Fig. 6 also shows that for conditions
`where Si is already covered with a relatively thick fluorocar-
`bon and the etch rate is suppressed, the fluorocarbon film on
`Si3N4 can be still thin enough for direct chemical sputtering
`to occur. This explains why a process condition that enables
`SiO2-to-Si selectivity does not necessarily allow for SiO2-to-
`Si3N4 selectivity, while all process conditions enabling
`SiO2-to-Si3N4 etch selectivity, automatically also provide
`SiO2-to-Si selectivity.
`
`D. Surface analysis: Reaction layers
`
`In the previous section it was suggested that in the case
`that a substrate is covered with a relatively thick steady-state
`fluorocarbon film, the etch rate is limited by the arrival of
`atomic fluorine which needs to diffuse towards the substrate
`surface. This suggestion is based on the observations by
`Oehrlein et al.1 who found that in the case of Si etching the
`thickness of a fluorinated silicon reaction layer located under
`the steady-state fluorocarbon film decreases if the thickness
`of the fluorocarbon overlayer increases. This observation
`rules in favor of the fluorocarbon film protecting the Si from
`fluorine attack, rather than that diffusion of etch products out
`through the fluorocarbon film would be the rate limiting step
`共which is also consistent with the etch rate being inversely
`proportional to the fluorocarbon film thickness兲.
`Based on the observations in Fig. 6, which suggest a gen-
`eral etch mechanism for SiO2 , Si3N4 , and Si in fluorocarbon
`plasmas, the presence of a fluorinated oxide or nitride reac-
`tion layer is to be expected. To verify the presence of these
`reaction layers, a detailed analysis of high resolution C 共1s兲,
`F 共1s兲, Si 共2p兲, and N共1s兲 or O共1s兲, was performed.
`The C 共1s兲 spectra from Fig. 5 can typically be decon-
`volved into four different Gaussian peaks corresponding to
`different chemical carbon bonds 共C–CFn , C–F, C–F2 , and
`C–F3), using a least-squares fitting routine and linear back-
`ground subtraction. In certain cases where the fluorocarbon
`films are relatively thin and the fluorocarbon–substrate inter-
`face is visible to XPS analysis, a C–C/C–Si peak needs to be
`included for obtaining a proper fit of the C 共1s兲 spectrum.
`The binding energies corresponding to the different carbon
`bonds are given in Table I. From the deconvolving of the C
`共1s兲 spectra fluorine-to-carbon ratios F/Cdec were calculated
`as follows:
`
`FIG. 6. Etch rates of SiO2 , Si3N4 , and Si samples plotted vs the thickness of
`the fluorocarbon film present on the surface during steady-state etching con-
`ditions. The varying parameters are feedgas chemistry 关CHF3 共40 sccm兲,
`C2F6 共40 sccm兲, C3F6 共40 sccm兲, and C3F6/H2 共20 sccm/15 sccm兲兴 and
`operating pressure 共6 and 20 mTorr兲. The rf bias power level corresponded
`to a self-bias voltage of ⫺100 V. It clearly shows that the thicker the fluo-
`rocarbon film, the lower the etch rate.
`
`higher values for the fluorocarbon film thickness are found
`with the latter method. The discrepancy between the two
`methods results from the fact that the error in the determina-
`tion of the thickness in both methods increases as the amount
`of fluorocarbon material on the surface increases. In this
`work the average of the thicknesses determined by the two
`methods is used to quantify the fluorocarbon film thickness.
`In Fig. 6 the etch rates of SiO2 and Si3N4 are plotted as a
`function of the fluorocarbon film thickness for all investi-
`gated chemistries (CHF3 , C2F6/C3F6, and C3F6/H2) and pres-
`sures 共6 and 20 mTorr兲 obtained at ⫺100 V self-bias. Data
`for silicon etched at the same conditions as SiO2 and Si3N4 is
`also included. The film thicknesses on silicon were deter-
`mined in real time by in situ He–Ne ellipsometry. In the
`conversion from ellipsometric angles ⌿ and ⌬ to fluorocar-
`bon film thickness, it is assumed that the steady-state fluoro-
`carbon film has the same refractive index as fluorocarbon
`deposited at the same process conditions without rf bias ap-
`plied. This assumption is supported by XPS analysis,16
`which indicates that the composition of the steady-state fluo-
`rocarbon films and passively deposited fluorocarbon material
`does not vary significantly.
`Figure 6 clearly shows a general trend for the investigated
`process regime; i.e., the etch rate of a thin film material
`decreases with increasing thickness of the steady-state fluo-
`rocarbon film on its surface. It can be seen that independent
`of the process conditions the SiO2 surfaces stay relatively
`clean of fluorocarbon material 共fluorocarbon film thick-
`ness⭐1.5 nm兲. The ion penetration depth in the investigated
`energy range is around 1 nm.31 The SiO2 surfaces can thus be
`etched directly by a mechanism of chemical sputtering,6,17,18
`which explains the relatively high etch rate. The Si surfaces
`are covered with a fluorocarbon film that is strongly depen-
`dent on the process conditions but typically relatively thick
`共2–7 nm兲. It is therefore unlikely that direct ion impact con-
`tributes to the Si etch rate. Instead, the etch rate is suggested
`
`J. Vac. Sci. Technol. A, Vol. 17, No. 1, Jan/Feb 1999
`
`

`

`31
`
`Schaepkens etal.: Study of the SiO2-to-Si3N4 etch selectivity mechanism
`
`31
`
`TABLE I. Listing of photoelectron binding energies.
`
`Element
`
`C 共1s兲
`
`F 共1s兲
`Si 共2p兲
`Si 共2p兲
`O 共1s兲
`N 共1s兲
`
`BE
`
`Specification
`
`Reference
`
`283.4 eV
`285.5 eV
`288.0 eV
`290.3 eV
`292.6 eV
`687.1 eV
`103.4 eV
`101.7 eV
`532.6 eV
`397.5 eV
`
`C–C/C–Si
`C–CFn
`C–F
`C–F2
`C–F3
`
`SiO2
`Si3N4
`SiO2
`Si3N4
`
`16
`
`32
`32
`33
`33
`33
`
`fluorocarbon film on the surface using the 90° take off angle
`primarily originate from fluorine bonded to carbon in the
`fluorocarbon film. If the fluorocarbon film thickness is small,
`however, the F共1s兲/C共1s兲 ratio is found to be higher than the
`F/Cdec ratio, resulting in a value for ⌬F/C that is larger than
`zero. This means that a significant part of the F 共1s兲 photo-
`electrons detected from samples with a relatively thin fluo-
`rocarbon film on the surface at 90° take off angle must origi-
`nate from fluorine species that are not bonded to carbon.
`The above data suggest that the fluorine that is not bonded
`to carbon is located underneath the steady-state fluorocarbon
`film, since it can be detected only if the path of the photo-
`electrons from the underlayer through the fluorocarbon film
`is significantly smaller than the electron escape depth 共e.g.,
`90° escape angle and thin fluorocarbon overlayers兲. The
`electron escape depth is typically in the order of a few na-
`nometers.
`The fluorine that is not bonded to carbon and that is lo-
`cated underneath the fluorocarbon film can be expected to
`form fluorinated SiO2 and Si3N4 reaction layers. In the case
`of Si etching, the films could be observed in the high reso-
`lution Si 共2p兲 XPS spectra. In the case of SiO2 and Si3N4
`etching, it is possible to obtain information on the reaction
`layers from substrate elements, Si 共2p兲 and N 共1s兲 or O 共1s兲.
`Indeed, the Si 共2p兲 spectra at 15° electron escape angles
`obtained from processed Si3N4 samples were found to be
`antisymmetric and have been deconvolved into two Gaussian
`peaks, as shown in Fig. 8共a兲. The low binding energy peak
`corresponds to Si 共2p兲 in bulk Si3N4 关BE⫽101.7 eV; full
`width at half maximum 共FWHM兲⫽2.3 eV兲.33 The high bind-
`ing energy peak is found to be shifted by 2.1⫾0.2 eV and
`has a FWHM of 2.7⫾0.1 eV. The shift towards a higher
`binding energy with respect to bulk Si3N4 can be explained
`by silicon bonding to electronegative fluorine. Similar