JP2005504439A - Method for etching a pattern in an etching body using plasma - Google Patents

Abstract

The present invention relates to a method for etching a pattern in an etching body (19) using plasma (15), for example, a method for etching a notch precisely defined in the lateral direction by an etching mask in a silicon body. In this case, a high-frequency power pulsed and modulated at a low frequency is input-coupled to the etching body (19) at least occasionally using a high-frequency alternating voltage, and the intensity of the plasma (15) is a function of time. Modulate.

Description

[0001]
The present invention relates to a method for etching a pattern in an etching body using plasma as described in the main claim, for example, a method for etching a notch precisely defined in the lateral direction in a silicon body.
[0002]
Conventional technology
For example, in the plasma etching process in which a notch precisely defined in the lateral direction is etched into a silicon body according to a format such as DE 42 41 045 C1, there is a problem that pocket stability (Taschenstabilite) is insufficient. That is, the required etching profile deviation occurs, especially at the interface between the etching body and the dielectric interface, for example the interface between silicon and the underlying silicon oxide. Arise.
[0003]
For this reason, DE 199 57 169 A1 already describes a so-called double pulse technique, in which a high frequency modulated carrier signal with a high pulse peak output is applied to a substrate electrode in an etching chamber in an inductively coupled plasma apparatus. By adding it, such unwanted pocket formation is suppressed, and at the same time, a wide process window for the plasma etching process is obtained. For example, by doing so, sufficient pocket stability is achieved at an etching pattern aspect ratio of 5: 1 to 10: 1, and a predetermined tolerance to overetching is obtained. Nevertheless, pocket formation cannot be completely suppressed even in this process if the aspect ratio of the trench to be formed is higher or if the overetching time is higher.
[0004]
DE 199 33 842 A1 also proposes pulsing an inductively coupled plasma source, whereby negative ions generated during the plasma discharge pause are generated in a dielectric etching base or pattern in a pattern with a high aspect ratio. This contributes to discharging the positive charge of the resist. A significant problem when pulsing the ICP plasma source (ICP = “inductively coupled plasma”) in this way is that high reflected power is generated in the corresponding high-frequency generator. The reason is that during plasma discharge firing, there are unclear conditions in the plasma that make it very difficult to match the input coupled high frequency power to the plasma impedance during the transition period. . For this reason, the ignition of the plasma discharge is in a state of transition from the electrically capacitively coupled mode to the inductively coupled mode, which causes impedance mismatching and, consequently, reflected power.
[0005]
In order to overcome this problem, DE 199 27 806 A1 according to DE 199 27 806 A1 provides an excitation voltage during the transition phase via a feedback circuit in the form of a Meissner oscillator with a plasma source as frequency determining element and a high frequency generator as an amplifier in the feedback path It has been proposed to free up the frequency. However, the disadvantage of this method is that it may generate frequencies outside the frequency range permitted for industrial facilities, which necessitates shielding.
[0006]
Furthermore, an unpublished application DE 100 51 831.1 already proposes an apparatus and a method for etching a substrate using inductively coupled plasma. According to this, a static or time-varying magnetic field is placed between the substrate and the ICP source, and this magnetic field uses at least two upper and lower magnet coils, and currents flow in opposite directions with respect to these coils. It is generated by flowing.
[0007]
It is an object of the present invention to improve pocket stability in a method of forming an etched body by etching a pattern, particularly when the etched pattern has a high aspect ratio and a long overetch time.
[0008]
Advantages of the invention
Unlike the prior art, according to the method according to the invention, for example when etching silicon, etc.₂Pocket stability is particularly enhanced, especially when a buried dielectric etch stop layer such as a layer is reached, and the tolerance to overetching is increased.
[0009]
The dependent claims contain advantageous embodiments of the invention.
[0010]
It is particularly advantageous to modulate or pulse the plasma intensity so that the plasma discharge remains in the inductively coupled mode without being extinguished during the "discharge pause", i.e. it is only needed to maintain a minimal discharge during this period. A certain amount of high frequency power is supplied to the plasma source or inductively coupled plasma. Since these plasmas are not extinguished during the discharge pause or pulse pause, it is avoided that a high reflected power is generated each time the plasma is next increased to the maximum intensity. The reason for this is that the starting time phase of the electrically capacitively coupled plasma discharge is sufficiently avoided and can be started immediately in the inductively coupled time phase of the plasma discharge.
[0011]
It is furthermore advantageous in this case to correlate or synchronize the high-frequency power input coupled to the substrate electrode, which is performed according to the double pulse technique described in DE 199 57 169 A1, in time with the modulation of the pulse intensity. It is.
[0012]
A further advantage in this connection is that during the discharge cessation, the plasma, which is dominated by positively charged ions and electrons, transitions to a so-called “bipolar” plasma consisting of positively and negatively charged ions. That is, during the so-called “afterglow after-glow” time phase, free electrons are taken in by recombination with positively charged ions or by capture by neutral particles. The generation of negative ions by electron capture is the dominant reaction in this case due to the numerical loss of the neutral particles surrounding the electrons. For this reason, in a “normal” plasma, the number of negative charge carriers with a mass corresponding to several times the mass of the proton is equal to the number of positive charge carriers with a mass corresponding to several times the mass of the proton. The number of these negative and positive charge carriers is approximately equal during this time phase, while it is smaller than the number by an order of 3-4. Since the proportion of free electrons is smaller than that of ions and the causal relationship between the charge carrier mass and the charge carrier mobility that is unbalanced in the plasma disappears, the plasma potential is about 0 V from a positive value in the region of several tens of volts. Approaches the value, so that both positive and negative potentials can reach the etching body to be processed, for example a silicon wafer, in the same way, so that optimal charge balancing is achieved even with a high aspect ratio .
[0013]
Inductively coupled plasmas cannot be maintained without electrons, but as electron density decreases, positive and negative charge carriers become more equivalent and work better for neutralization of disturbing charges. . Particularly advantageous in the process according to the invention is to make the electron density as small as possible, or to keep the electron density as small as possible when carrying out the process according to the invention.
[0014]
In addition, the modulation of the plasma intensity as a function of time can be performed advantageously by, for example, periodically changing the high-frequency power input coupled from a suitable coil generator to the plasma, for example, with time, as an alternative. Or in addition to this, it can also be carried out by periodically changing the magnetic field strength of the magnetic field acting on the plasma, such as the magnetic field of a device of the form DE 100 51 831.1.
[0015]
Drawing
Next, the present invention will be described in detail with reference to the drawings.
[0016]
FIG. 1 is a basic configuration diagram of a plasma etching apparatus for carrying out a method according to the present invention. FIG. 2 is a diagram for explaining a first embodiment relating to temporal modulation of plasma intensity, and this modulation is synchronized with a high-frequency pulse that is input-coupled to a substrate electrode and is modulated at a low frequency. . FIG. 3 is a diagram for explaining the structure of the high-frequency power according to FIG. FIG. 4 is a diagram for explaining a second embodiment relating to modulation of plasma intensity and synchronization with high-frequency power input-coupled to the substrate electrode. FIG. 5 is a diagram for explaining a third embodiment during a pulse pause that is also clocked at a low frequency. FIG. 6 is a diagram for explaining a second embodiment which is also clock-controlled at a low frequency.
[0017]
Example
FIG. 1 shows a plasma etching apparatus 5 known from DE 100 51 831.1, by which an anisotropic plasma etching process is performed in silicon in order to produce a trench, for example in the form DE 42 41 045 C1. Executed. More specifically, an etching chamber 10, a substrate electrode 18, and a substrate 19 disposed thereon, such as a silicon wafer, are provided. Further, the substrate electrode 18 is electrically connected to the second match box 21 for impedance matching and the substrate power generator 22.
[0018]
Coils 11 surrounding the etching chamber 10 are provided in the upper region of the etching chamber 10, and these coils are connected to a coil generator 13 through a first match box 12 for impedance matching. The high frequency power 11 is input and coupled to the etching chamber 10 by the coil 11 via the coil generator 13 and the first match box 12 described above, whereby an inductively coupled plasma 15 is formed there. Further referring to FIG. 1, in order to supply or exhaust the process gas, for example to alternately supply or exhaust the etching gas and the inert gas, the etching chamber 10 has a gas supply pipe 14 in its upper region and a lower part thereof. A gas exhaust pipe 20 is provided in the region.
[0019]
Furthermore, the etching chamber 10 is surrounded by two magnetic field coils 16 between the inductively coupled plasma 15 and the substrate electrode 18, so that two suitable spacers 17 are incorporated in the sidewalls of the etching chamber 10, and these spacers. Contains the coil 16. Refer to DE 100 51 831.1 for the detailed structure of the plasma etching apparatus 5.
[0020]
A device known from DE 199 27 806 A1 or preferably from DE 199 33 842 A1 is provided for modulating the intensity of the plasma 15 as a function of time using the coil generator 13 and the first match box 12. This device is integrated into the first matchbox 12 or the coil generator 13, for example as described therein.
[0021]
Further, the substrate power generator 22, the second match box 21, and the substrate electrode 18 input and couples the high frequency pulsed and modulated at a low frequency to the substrate 19. This is described in DE 199 57 169 A1.
[0022]
FIG. 3 shows high-frequency power converted into high-frequency pulses and modulated at a low frequency. In this case, pulse packets 30 and clocked pulse pauses 31 which are clocked periodically and alternately at low frequencies are coupled to the substrate electrode 18, which has a frequency of, for example, 1 Hz to 500 Hz, preferably 10 Hz. With a frequency of ˜250 Hz, for example 100 Hz, with a duty cycle of 20% to 80%, preferably 50%, the average power is preferably 5 W to 20 W, for example 10 W. In this case, the low-frequency clock-controlled pulse packet 30 shown in FIG. 3 is composed of periodically alternating rows of high-frequency clock-controlled pulses 32 and high-frequency clock-controlled pulse pauses 33. The frequency is preferably between 10 kHz and 500 kHz, such as 100 kHz, and the duty cycle is preferably between 2% and 20%, such as 5%. In this case, the average power input-coupled to the substrate electrode 18 is, for example, in the vicinity of 5 W to 40 W in terms of time average, and is, for example, 20 W during the high frequency clock controlled pause 32.
[0023]
In addition, as shown in FIG. 3, each high frequency clocked individual pulse 32 comprises a high frequency carrier signal, which has a frequency of, for example, 13.56 MHz, preferably 100 W to 1 kW of high frequency power, for example 400 W. Has high frequency power. For further details on FIG. 3, reference is made to patent application DE 199 57 169 A1.
[0024]
Of particular note regarding the signal shape at the substrate electrode 18 according to FIG. 3 is that a sufficiently long pulse pause 31 generated by the low frequency clock control is observed, and the dielectric boundary in the trench to be etched. It is possible to discharge in the region of the layer during this one rest period. Such slow pulsing reduces the process stability itself, which results in a narrow process window, but the lowest possible pulse to pause ratio (duty cycle), eg 1:10 or 1 Due to the additional high frequency modulation of the high frequency carrier signal 34 with a ratio of: 20, a very high substrate electrode voltage is simultaneously applied to the substrate electrode 18 at a low current, resulting in a significantly wider and more tolerant process window. . In this case, the current-voltage relationship is controlled by the duty cycle, and as a result, the apparent resistance value of the plasma 15 viewed from the substrate electrode 18 is controlled.
[0025]
FIG. 2 shows a first embodiment of the method according to the invention, in which high-frequency pulsed and low-frequency modulated high-frequency power is synchronized with the modulation of the plasma intensity at the substrate electrode 18. As a result, the minimum power plasma excitation (there is almost no plasma), that is, the first minimum plasma intensity value 41 overlaps the pulse pause 31 controlled by the low frequency clock. As a result, the discharge is intensified during the pulse pause 31 controlled by the low-frequency clock in the formed trench. The reason is that not only self-discharge of the trench occurs, but also negative ions are increased and attracted to the pattern base, and are neutralized by combining with positive charges existing there.
[0026]
The on / off ratio of 1: 1 written in FIG. 2, that is, the ratio between the period of the first plasma intensity maximum value 40 and the period of the first plasma intensity minimum value 41 is merely an example. Rather, it is preferable for the reason of plasma excitation efficiency to excite the plasma 15 as long as possible and to extinguish it for as short a time as possible, that is, the excitation or on / off ratio of the plasma intensity is significantly smaller than 1: 1. It is advantageous to set the ratio. The purpose is to avoid that the pulse peak power required in the high frequency power input coupled to the plasma 15 becomes enormous. For example, for an average power of the coil 11 of 3 kW to 5 kW, a pulse peak power of 6 kW to 10 kW is required to obtain a desired time average value when the on / off ratio is 1: 1.
[0027]
As further shown in FIG. 2, the period of the first plasma intensity maximum 40 followed by the first plasma intensity minimum 41 is a low frequency clocked pulse packet 30 followed by a low frequency clock control. Equal to the period of the pulse pause 31 made. In addition, the intensity of the plasma 15 in the first plasma intensity minimum value 41 is selected to be low enough that the plasma 15 barely extinguishes within this plasma intensity minimum value.
[0028]
FIG. 4 shows a second embodiment relating to the synchronization of the plasma intensity modulation and the high-frequency power in the substrate electrode 18 that has been converted into a high-frequency pulse and modulated at a low frequency. In this case, for the purpose of maintaining as high an on / off ratio as possible, for example, as shown, two or more high frequency clocked pulses 32 each in the substrate electrode 18 are formed into the second plasma. Surrounded by an intensity maximum 40 ', the plasma 15 is switched to the "Low" mode during the subsequent high frequency clocked pulse pause 33, i.e. the plasma intensity reaches a second plasma intensity minimum 41'. . This plasma intensity minimum value 41 'is selected so low that the plasma 15 barely extinguishes during this period. Thus, always two or more high-frequency clocked pulses 32 fall into the second plasma intensity maximum 40 ', and then again the high-frequency clocked pulse pause 33 becomes the second plasma intensity. It overlaps with the minimum intensity value 41 '.
[0029]
Unlike the embodiment shown in FIG. 2, the advantage of the embodiment shown in FIG. 4 is that, in the case of FIG. 4, in the trench formed during a relatively long on-time in a relatively low frequency modulation of the plasma intensity. It means that only a little charge is accumulated. This allows a relatively small amount of charge to accumulate in the trench during this period each time only a relatively few high frequency clocked pulses 32, for example at most 20, overlap the second plasma intensity maximum 40 '. After that, the discharge again occurs during the subsequent second pulse minimum value 41.
[0030]
Moreover, the maximum negative ion density in the plasma 15 occurs only when the second plasma intensity maximum value 40 'disappears or during a short period immediately after that, that is, the maximum discharge action is the second plasma intensity minimum value 41. Only achieved for a short period during and after the transition to ′. Thereafter, only a significantly reduced negative ion density in the plasma 15 is obtained, which means that a plasma intensity minimum 41 'as short as possible and a correspondingly high discharge efficiency are obtained. Moreover, in the case of the embodiment according to FIG. 4, it is not a good idea to make the on / off ratio much smaller than 1: 1 when plasma is excited or when high frequency power is input to the plasma 15. This is because otherwise the pulse peak power required for plasma generation would be uneconomically high.
[0031]
Specifically, the high frequency power input coupled to the inductively coupled plasma 15 is now 3 kW to 5 kW, and the frequency of the high frequency clock controlled pulse 32 and the subsequent high frequency clock controlled pulse pause 33 is, for example, a duty of 5%. The average high frequency power that is in the vicinity of 100 kHz during the cycle and that is input coupled to the substrate electrode 18 is, for example, in the vicinity of 20 W in the low frequency clock controlled pulse packet 30. FIG. 4 shows only the modulation of the plasma intensity in only one pulse packet 30 under the low-frequency clock control shown in FIG. Following this low-frequency clocked pulse packet 30 is a low-frequency clocked pulse pause 31. During this period, the intensity of the plasma 15 is also simultaneously increased by the first plasma shown in FIG. It is reduced to a minimum intensity value 41 and remains there during the low frequency clocked pulse pause 31. This is depicted in full form in FIG.
[0032]
FIG. 5 shows a third embodiment for temporally synchronizing the intensity of the plasma 15 with the high frequency power that is pulsed at a high frequency, modulated at a low frequency, and input coupled to the substrate electrode 18. In this case, unlike FIG. 6, the time modulation of the intensity of the plasma 15 as shown in FIG. 4 is maintained even during the pulse pause 31 controlled by the low frequency clock.
[0033]
Therefore, during the relatively long low frequency clock controlled pulse pause 31, the plasma 15 repeatedly rises and falls, so that the plasma extinction phase is continually repeated with increasing negative ion density associated therewith. . Here, the expression “relatively long” means that it is relatively long with respect to the decay period of the negative ion density in the plasma 15 after the plasma is extinguished or after the transition of the plasma intensity to the second minimum plasma intensity 41 ′. It is. Therefore, at the beginning of the pulse pause 31, plasma extinction occurs only once with the increase in negative ion density, and then not only rapidly attenuates compared to the period of the low frequency clock controlled pulse pause 31. This kind of phase is repeatedly applied to discharge the trench formed in the etching body, so that the low frequency clock controlled pulse pause 31 can be used more efficiently. In this way, the low frequency clock controlled pulse pause 31 is not only for the self-discharge of the trench, but also periodically gives a negative ion density peak between the low frequency clock controlled pulse pause 31. This accelerates the discharge process. In addition, this also alleviates the problem of “duty cycle” being too low for plasma generation. The reason is that a “duty cycle” better than 1: 1 can be easily realized by separating plasma generation and low frequency modulation of high frequency power at the substrate electrode 18.
[0034]
For the purpose of setting and stabilizing the intensity of the plasma 15 at the minimum intensity values 41 and 41 ′ close to the extinction of the plasma 15, a transition is made from the induction mode to the electrically capacitively coupled mode when the plasma 15 is likely to extinguish. It is utilized that the electric power reflected to the coil generator 13 increases dramatically when it seems to be. Immediately adjusting and increasing the forward power of the coil generator 13 prevents this condition and keeps it at the limit of the inductive coupling mode of operation, thus increasing the power of the coil generator 13 in the plasma. The electron density in 15 is again increased to the value of the stable operating state.
[0035]
Here, the forward power P of the coil generator 13_Forward Is the power P reflected to the coil generator 13 according to the following equation at the plasma minimum values 41 and 41 '._Reflected Combined with:
P_Forward = P_Soll + V · P_Reflected
Here, V is an amplification coefficient of the adjusting circuit, and V >> 1 is preferably applied. The case of V = 1 corresponds to “Load” control that is common in the latest high-frequency generators, that is, when the forward power is generated, the difference between the forward power and the reverse power, that is, the plasma 15 is actually applied to the plasma 15. Control is performed so that the high frequency power to be input and coupled is kept constant in accordance with a predetermined target value. Such conventional “Load” control situations are often insufficient for stabilization of the plasma 15 at critical mode boundaries. This is because the effective high frequency power that is input coupled to the plasma 15 does not increase as needed to stop the extinguishing plasma 15. Therefore, it is advantageous here to set the reflection coefficient V to a value larger than 1, for example, a value of 5 to 10. In this case, the set target value (P_Soll), A value as close as possible to the value required for the limit operation of the plasma 15, that is, a slightly higher intensity than the extinction of the plasma 15, or a value smaller than that value is set.
[0036]
Note that the pulse strategy described above, that is, the modulation of the plasma intensity and the high frequency power input coupled to the substrate electrode 18 as a function of time is used in the process of DE 42 41 045 C1 even during the deposit cycle. And used during the etching cycle. However, it is usually sufficient to limit to the etching cycle, because the risk of pocket formation exists only during the etching cycle. In this case, the entire generator output can be used during the deposit cycle. In addition, it is advantageous to completely cut off the input coupling from the high frequency power to the substrate electrode 19 during the deposition cycle.
[0037]
A very simple modulation of the intensity of the plasma 15 can be performed by using an inductively coupled plasma source with a magnetic coil device, which is described in DE 100 51 831.1 and depicted in FIG. . In this case, at least two magnetic field coils 16 are arranged between the ICP source, that is, the inductively coupled plasma 15 and the substrate 19, the upper magnetic field coil 16 faces the ICP source, and the lower magnetic field coil is placed on the substrate 19. These magnetic field coils are opposed to each other, and generally different magnitudes of currents having opposite polarities are passed through them, and as a result, magnetic fields of different strengths in opposite directions are generated.
[0038]
More specifically, in this case, the upper magnetic field coil 16 facing the ICP source is set to a magnetic field strength required for optimal plasma generation, while the lower magnetic field coil 16 facing the substrate 19 has a magnetic field in the opposite direction. The strength of this magnetic field is set to the strength required for optimal etching homogeneity, that is, the strength required to optimally distribute the energy applied onto the substrate surface.
[0039]
The use of the field coil 16 firstly allows the plasma 15 to be held at a lower excitation density and electron density than in the absence of this coil, especially in the limiting situation of a plasma that is barely extinguished. This is due to the fact that the “lifetime” of the electrons present in the plasma 15 is extended by the magnetic field generated by reducing the conversion loss in the plasma source region, so that the desired bipolar plasma is very well achieved. Is maintained at the minimum value of plasma intensity 41, 41 'with the minimum free electron density.
[0040]
Now, in order to realize the modulation of the intensity of the plasma 15, in addition to the high frequency power input coupled to the plasma 15 from the coil generator 13 through the coil 11 in the plasma apparatus 5 shown in FIG. Alternatively, alternatively, the magnetic field generated by the magnetic field coil 16 can also be utilized in the chamber 10. Thus, for example, firstly the coil current in the magnetic field coil 16 is used to set the first plasma intensity 40, which corresponds to the target current value of the process according to DE 100 51 831.1, ie 10A is set for the magnetic field coil 16, and 7A is set for the lower magnetic field coil 16 having the opposite polarity.
[0041]
Thereafter, these currents are reduced in order to switch to the minimum plasma intensity values 41, 41 ', for example this is done so that both currents in the magnetic field coil 16 are returned to zero or timed. However, as an alternative to this, an intermediate value may be set. For example, 3A can be set for the upper magnetic field coil 16, and 2A can be set for the lower magnetic field coil 16.
[0042]
Therefore, in the simplest embodiment, both coil currents are switched between the upper extreme value and the lower extreme value at the same time, thereby having the same effect as returning the power of the coil generator 13. As a result, that is, when the magnetic field coil current is weakened, the plasma density disappears and reaches the plasma intensity minimum value 41, 41 ', and at that time, a high negative ion density is obtained due to recombination of electrons and neutral gas particles for a short period. Occur. However, it should be noted here that the magnetic field coil current cannot be modulated as rapidly as the high frequency power input coupled to the plasma 15. In particular, due to the inductance of the magnetic field coil 16, only clock frequencies lower than 10 kHz are possible. On the other hand, changing the direct current is significantly easier and less problematic than changing the high frequency alternating voltage using the coil generator 13.
[0043]
Particularly suitable for preventing reflected power is not to change the current in the field coil 16 immediately, but to have a finite rate of change, which of course is a coil generator. This also applies to the change of the high-frequency power of the machine 13, in which case impedance matching using the first match box 12 can generally be performed more easily by changing the power as slowly as possible.
[0044]
Such a finite change rate can be achieved very simply by modulation of the magnet coil current. This is because in this case, only the DC voltage or DC current needs to be superimposed on the modulation voltage having a finite side edge steepness.
[0045]
Thus, for example in the previously described embodiment
U₁(T) = U₀₁・ Sin (ωt) U₂(T) =-U₀₂・ Sin (ωt)
An alternating voltage or alternating current that varies in accordance with is used in both field coils 16. In this case, the AC voltage or AC current in both magnetic field coils 16 is out of phase at any point in time, and U₀₁And U₀₂Represents the voltage amplitude or current amplitude in one of the two magnetic field coils 16, respectively.
[0046]
Alternatively, an operation using a rectified AC voltage or AC current is also possible, in which case the term sin (ωt) is replaced by an absolute value abs (sin (ωt)), respectively.
[0047]
Furthermore, as already mentioned, it is often advantageous not to return the magnet coil current to zero. The coil voltage or coil current flowing through both field coils 16 is, for example:
U₁(T) = U_{offset, 1}+ U₀₁・ Abs (sin (ωt))
U₂(T) =-U_{offset, 2}-U₀₂・ Abs (sin (ωt))
In this case, offset current or offset voltage U_{offset, 1}Or U_{offset, 2}Each is selected so that the so-called “beaking effect” is still effectively suppressed in the peripheral region of the substrate 19 and thus a homogeneous etching result is still obtained over the entire substrate surface.
[0048]
Furthermore, the frequency ω according to the above equation is relatively small, that is, for example 10 Hz to 50 Hz, and the speed of the first match box 12 used for impedance matching is fast enough to follow such modulation of the plasma intensity. If there is a gap, it is possible to completely prevent the reflected power from entering the coil generator 13 in this way, and it is possible to remarkably suppress the formation of unwanted pockets. What is important here is that the density of the plasma 15 is modulated, a time phase in which negative ion density is advantageously increased by this modulation can be used, and the discharge of a trench having a high aspect ratio is caused by this time phase. It is to be obtained.
[Brief description of the drawings]
[Figure 1]
1 is a basic configuration diagram of a plasma etching apparatus for carrying out a method according to the present invention.
[Figure 2]
It is a figure explaining the 1st Example regarding the time modulation of plasma intensity.
[Fig. 3]
It is a figure explaining the structure of the high frequency electric power by FIG.
[Fig. 4]
It is a figure explaining the 2nd Example regarding the synchronism with the modulation | alteration of plasma intensity | strength and the high frequency electric power input-coupled to a substrate electrode.
[Figure 5]
It is also a figure explaining the 3rd Example in the pulse pause also clock-controlled by a low frequency.
[Fig. 6]
It is also a figure explaining the 2nd Example clock-controlled by low frequency.

Claims (15)

In a method for etching a pattern in an etching body using plasma, for example, in a method for etching a notch precisely defined in a lateral direction by an etching mask in a silicon body,
Using high frequency alternating voltage, the etching body is coupled with high frequency power that is at least occasionally high frequency pulsed and modulated at low frequency,
Characterized by modulating the intensity of the plasma (15) as a function of time,
A method of etching a pattern in an etching body. The intensity of the plasma (15) is at least occasionally, for example periodically, between a maximum value (40, 40 ') held over the first period and a minimum value (41, 41') held over the second period. The method according to claim 1, wherein the modulation (4, 41 ') is selected such that the plasma (15) does not extinguish during the second period. 3. The method according to claim 2, wherein the minimum value (41, 41 ') is selected to be small enough that the plasma (15) is barely extinguished. 2. The intensity of the plasma (15) is modulated at least occasionally, e.g. periodically, so that the plasma takes a maximum value (40, 40 ') over a first period and is reduced to zero over a second period. Method. Periodically maintaining the intensity of the plasma (15), between the maximum value (40, 40 ') held over the first period and the minimum value (41, 41') held over the second period. The method according to claim 1, wherein the minimum value (41, 41) is selected such that the plasma does not extinguish during the second period. During the first time interval, the intensity of the plasma (15) is sustained and periodically, the maximum value (40, 40 ') held for the first period and the minimum value (41 for the second period) , 41 '), for example, rectangular pulse modulation, and selecting the minimum value (41, 41') so that the plasma (15) does not extinguish during the second period;
During the second time interval, the intensity of the plasma (15) is reduced to zero or to an intensity (41) lower than the maximum value (40, 40 ') over a third period, for example, the plasma (15) has become scarce. Reduce to the strength not to extinguish,
6. A method according to any one of claims 1-5. The method according to claim 6, wherein the first time interval and the second time interval are repeated alternately, for example directly and successively. The high-frequency power pulsed and modulated at low frequency is correlated with the modulated plasma intensity, and the plasma is generated when the pulse packet (30) clocked at low frequency is coupled to the etching body (19). The method according to claim 1, wherein an intensity maximum value (40) of (15) is generated. The high frequency power pulsed and modulated at a low frequency is correlated with the modulated plasma intensity, and the plasma (15) when the low frequency clocked pulse pause (31) appears in the etching body (19). The method according to claim 1, wherein a minimum value of intensity is generated. A pulse packet (30) clocked at low frequency has a plurality of alternating pulses (32) and pulse pauses (33) clocked at high frequency, the period of the pulse (32) and the pulse The first pulse-to-pause ratio is defined by the period of the pause (33), and the intensity of the plasma (15) takes the maximum value (40 ') and the intensity of the plasma (15) reaches the minimum value (41). The second pulse-to-pause ratio is defined by the second period taking ′), wherein the second pulse-to-pause ratio is greater than the first pulse-to-pause ratio. The method described in the paragraph. The repetition frequency of the high frequency clock controlled pulse (32) and the high frequency clock controlled pulse pause (33) of the high frequency power is the maximum value (40, 40 ') and the minimum value (41, 41') of the plasma (15). 11. A method according to any one of claims 1 to 10, wherein the method is greater than the repetition frequency. The repetition frequency of the low frequency clocked pulse packet (30) and the low frequency clocked pulse pause (31) is at least approximately equal to the repetition frequency of the first and second periods. Item 12. The method according to any one of Items 1 to 11. The repetition frequency of the low frequency clocked pulse packet (30) and the low frequency clocked pulse pause (31) is at least approximately equal to the repetition frequency of the first period and the second period, or 13. A method according to any one of claims 1 to 12, which is smaller. Modulation of the intensity of the plasma (15), the high frequency power input coupled to the plasma (15), for example, high frequency power that periodically changes with time, and / or the magnetic field strength of the magnetic field acting on the plasma, for example, periodically changes with time The method according to claim 1, wherein the method is performed according to a magnetic field strength to be applied. The method according to claim 1, wherein an etching body (19) is connected to a substrate electrode (18) and the high-frequency power is supplied to the substrate electrode.