Classifications

• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32009—Arrangements for generation of plasma specially adapted for examination or treatment of objects, e.g. plasma sources
  • H01J37/32082—Radio frequency generated discharge
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32009—Arrangements for generation of plasma specially adapted for examination or treatment of objects, e.g. plasma sources
  • H01J37/32082—Radio frequency generated discharge
  • H01J37/32137—Radio frequency generated discharge controlling of the discharge by modulation of energy
  • H01J37/32146—Amplitude modulation, includes pulsing
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32009—Arrangements for generation of plasma specially adapted for examination or treatment of objects, e.g. plasma sources
  • H01J37/32082—Radio frequency generated discharge
  • H01J37/32137—Radio frequency generated discharge controlling of the discharge by modulation of energy
  • H01J37/32155—Frequency modulation
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32431—Constructional details of the reactor
  • H01J37/3244—Gas supply means
  • H01J37/32449—Gas control, e.g. control of the gas flow
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J2237/00—Discharge tubes exposing object to beam, e.g. for analysis treatment, etching, imaging
  • H01J2237/32—Processing objects by plasma generation
  • H01J2237/33—Processing objects by plasma generation characterised by the type of processing
  • H01J2237/334—Etching
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32009—Arrangements for generation of plasma specially adapted for examination or treatment of objects, e.g. plasma sources
  • H01J37/32082—Radio frequency generated discharge
  • H01J37/32174—Circuits specially adapted for controlling the RF discharge

Definitions

• Plasma processing systems have long been employed to process substrates (e.g., wafers or flat panels or LCD panels) to form integrated circuits or other electronic products.
• Popular plasma processing systems may include capacitiveiy coupled plasma processing systems (CCP) or inductively coupled plasma processing systems (ICP), among others.
• CCP capacitiveiy coupled plasma processing systems
• ICP inductively coupled plasma processing systems
• the source RF signal may be pulsed (e.g., on and off) in order to obtain a plasma that has the normal ion to neutral flux ratio during one phase of the pulse cycle ⁇ e.g., the pulse on phase) and a plasma with lower ion to neutral flux ratio dur ing another phase of the pulse cycle (e.g., during the pulse off phase), it is known that source RF signal may be pulsed synchronously with bias RF signal,
• Fig. 1 shows, in accordance with one or more embodiments of the invention, an example combination pulsing scheme where the input gas (such as reactant gas and/or inert gas) and the source RF signal are both pulsed, albeit at different pulsing frequencies.
• the input gas such as reactant gas and/or inert gas
• FIG. 2 shows, in accordance with one or more embodiments of the invention, another example combination pulsing scheme.
• FIG. 3 shows, in accordance with one or more embodiments of the invention, yet another example combination poising scheme.
• Fig. 4 sho ws, in accordance with one or more embodiments of the invention, other possible combinations for the combination pulsing scheme.
• FIG. 5 shows, in accordance with one or more embodiments of the in vention, the steps for performing combination pulsing.
• Fig. 6 shows, in accordance with one or more embodiments of the invention, die steps for performing gas pulsing.
• FIGs. 7A and 7B illustrate, in accordance with embodiments of the in vention, different example variations of the gas pulsing scheme discussed in connection with Fig. 6.
• FIG. 8 sho ws, in accordance with an embodimen t of the invention, conceptual MMP etching cycles for the siiicon etching example, with each cycle involving at least an MMP preparation phase and an MMP reactive phase.
• Fig. JO shows, in accordance with an embodiment of the in vention, a method for performing MMP etching in a production 1CP chamber.
• Embodiments of the invention related to a combination pulsing scheme that pulses the input gas (e.g., reactant gases and/or inert gases) using a first pulsing
• the input gas e.g., reactant gases and/or inert gases
• a complete gas pulse cycle is 1 second in this example.
• the gas pulsing duty cycle is 70%
• the gas may be on for 70% of the -second gas pulsing period and off for 30% of the 1 -second gas pulsing period. Since the source RF signal pulsing rate is 100 Hz, a complete RF signal pulsing period is 10 ms.
• the RF pulsing dut cycle is 40%
• the RF on-phase when the 13.56 MHz signal is on
• the RF oil phase when the 13.56 MHz signal is off
• the inductive source RF signal may be pulsed with two different frequencies while the gas is pulsed at its own gas pulsing frequency.
• the aforementioned 13,56 MHz RF signal may be pulsed not only at frequency fl of 100 H but amy also be pulsed with a different, higher frequency during the on-phase of frequency fl .
• the RF pulsing duty cycle is 40% of the fl pulse
• the on-phase off I is 40% of 10ms or 4ms.
• the RF signal may also be pulsed at a different, higher frequency of f2 (such as at 400 Hz).
• Embodiments of the invention contemplate that the gas pulses and R F pulses may be synchronous (i.e., with matching Leading edge and/or lowering edge of the pulse signals) or may be asynchronous.
• the duty cycle may be constant or may vary in a manner that is independent of the other pulsing frequency or hi a manner that is dependent on the other poising frequency.
• frequency chirping may be employed.
• the RF signal may change its fundamental frequency in a periodic or non- periodic manner so that during a phase or a portion of a phase of any of the pulsing periods (e.g., any of the RF signal, or gas pulsing periods), a different frequency (e.g., 60 MHz versus 13.56 MHz) m be employed.
• the gas ulsing frequency may be changed with time in a periodic or non-periodic manner if desired.
• the aforementioned gas and source RF pulsing may be combined with one or more pulsing or variation of another parameter (such as p ulsing of the bias RF signal, pulsing of the DC bias to the electrode, pulsing of the multiple RF frequencies at different pulsing frequencies, changing the phase of any of the parameters, etc.)
• another parameter such as p ulsing of the bias RF signal, pulsing of the DC bias to the electrode, pulsing of the multiple RF frequencies at different pulsing frequencies, changing the phase of any of the parameters, etc.
• Fig. 1 sho ws, in accordance with an embodiment of the invention, an exam le combinationinstallmg scheme where the input gas (such as reactant gas and/or inert gas) and the source RF signal are both pulsed, albeit at different pulsing frequencies, in the example of Fig. 1 , the input gas 1 2 is pulsed at a gas pulsing rate (defined as T where 3 ⁇ 4 > is the period of the gas pulse) of about 2 seconds/pulse or 2 M Hz.
• a gas pulsing rate defined as T where 3 ⁇ 4 > is the period of the gas pulse
• the TCP source RF signal 104 of 13.56 MHz is pulsed at a RF pulsing rate (defined as where 3 ⁇ 4, is the period of the RF pulsing).
• the RF signal is on (such as the 13.56 MHz RF ssgnal) during the time period 120 and the RF signal is off during the time period 122.
• Each of the gas pulsing rate and the RF pulsing rate may have its own duty cycle (defined as the pulse on-time divided by the total pulsing period). There are no requirements that the duty cycle has to be 50% for any of tlie pulse signals, and the duty cycle may vary as needed for a particular process,
• the gas pulsing and the RF signal pulsing are at the same duty cycle, in another embodiment the gas pulsing and the RF signal pulsing are at independently controllable (and may be different) duty cycles to maximize granular control.
• the leading and/or trailing edges of the gas pulsing signal and the RF pulsing signal may be synchronous.
• the leading and/or trailing edges of die gas pulsing signal and the RF pulsing signal may be asynchronous.
• the gas input 202 is pulsed at its own gas pulsing frequency.
• the source RF signal 204 may be pulsed with two different frequencies while the gas is pulsed at its own gas pulsing frequency (defined as 1/T gt ,, where T ; , ; , is the period of the gas pulse).
• the RF signal may be pulsed not only at frequency fl (defined as 1/Tn from the figure) but may also he pulsed with & different, higher frequency during the on-phase of fl pulsing.
• the RF signal may be pulsed, at a different pulsing frequency f2 (defined as l/T ' o from the figure).
• the gas input 302 is pulsed at its own gas pulsing frequency.
• the source RF signal 304 may be pulsed with three different frequencies while the gas is pulsed at its own gas pulsing frequency.
• the RF signal may be pulsed not onl at frequency fl (defined as 1/T-n from the figure) but may also be pulsed with a different, higher frequency during the on-phase of fl pulsing.
• the RF signal may he pulsed at a different pulsing frequency £2 (defined as ⁇ ⁇ ⁇ from the figure.
• the RF signal may he pulsed at a different pulsing frequency f3 (defined as ⁇ ⁇ from the figure).
• the duty cycle is shown to be constant in the examples of Figs. 1-3, the duty cycle may also vary, in a periodic or non-periodic manner and independently or dependency on the phases of one of the pulsing signals (whether gas pulsing signal, RF pulsing signal or otherwise). Further, the change in the duty cycle may be synchronous or asynchronous with respect to phase of any one of the pulsing signals (whether gas pulsing signal, RF pulsing signal, or otherwise).
• the doty cycle of the RF pulsing is advantageously set to be one value during the on-phase of the gas pulse (e.g., 154 in Fig. I), and the duty cy cle of the RF pulsing is set to be another differeiit value during the off-phase of the gas pul se (e.g., 156 of Fig. 1).
• the duty cycle of the RF pulsing is advantageously set to be one value during the on-phase of the gas pulse (e.g., 154 in Fig.
• the duty cycle of the RF pulsing is set to be a lower value during the off-phase of the gas pulse (e.g., 156 of Fig. 3 ).
• this RF pulsing duty cycle embodiment wherein the duty cy cle is higher during the on pha se of the gas pul sing and lower during the off phase of the gas pulsing is advantageous for some etches.
• this F pulsing duty cycle variance wherein the duty cycle is lower during the on phase of the gas pulsing and higher during the off phase of the gas pulsing is advantageous for some etches.
• the duty cycle is other than 100% during the time when the signal is pulsed (Le., pulsing and "always on" are two different concepts).
• frequency chirping may be employed with any of the pulsing signals (whether gas pulsing signal, RF pulsing signal, or otherwise).
• Frequency chirping is described in greater detail in connection with the RF pulsing signal in Fig. 4 below.
• the gas is pulsed such that during the gas pulsing on phase, reactant gasies) and inert gas(es) (such as Argon, Helium, Xenon, Krypton, Neon, etc) are as specified by the recipe.
• reactant gasies such as Argon, Helium, Xenon, Krypton, Neon, etc
• inert gas(es) such as Argon, Helium, Xenon, Krypton, Neon, etc
• the gas is pulsed such that during the gas pulsing on phase, reactant gasies) and inert gas(es) (such as Argon, Helium, Xenon, Krypton, Neon, etc) are as specified by the recipe.
• inert gas(es) such as Argon, Helium, Xenon, Krypton, Neon, etc
• tire reactant gas(es) is removed and replaced by inert gas(es) during the gas pulsing off phase.
• the percentage of inert gas(es) to total gas(es) flowed into the chamber may vary from about X% to about 100%, wherein X is the percentage of inert gasies) to total gas flow that is
• the percentage of inert gasies) to total gas(es) flowed into the chamber may vary from about 1,1 X to about 100%, wherein is the percentage of inert gas(es) to total gas flow that is employed during the gas pulsing on phase, in a preferred embodiment, the percentage of inert gas(es) to total gasies) flowed into the chamber may vary .from about 1.5 X to about 100%, wherein X is the percentage of inert gas(es) to total gas flow that is employed during the gas pulsing on phase.
• the gas pulsing rate is limited at the high end (upper frequency limit) by the residence time of the gas in the chamber.
• This residence time concept is one that is known, to one skilled in the art and varies from chamber design to chamber design. For example, residence time typically ranges in the tens of milliseconds for a capacitively coupled chamber, in another example, residence time typically ranges in the tens of milliseconds to hundreds of milliseconds for an inductively coupled chamber,
• the gas pulsing period may range from 10 milliseconds to 50 seconds, more preferably from 50 milliseconds to about 10 seconds and preferably from about 500 milliseconds to about 5 seconds.
• the RF pulsing frequency is limited at the upper end by the frequency of the RF sigaa! (e.g., 13.56 MHz would establish the upper limit for the RF pulsing frequency if the RF frequency is 13.56 MHz).
• Fig. 4 sho ws, in accordance with one or more embodiments of the invention, other possible combinations- in Fig. 4, another signal 406 (such as bias RF or any other periodic parameter) may be pulsed along wit gas pulsing signal 402 and source RF pulsing signal 404 (pulsed as shown with 430 and 432).
• the pulsing of signal 406 may be made synchronous or asynchronous with any other signals in the system.
• another signal 408 (such as DC bias or temperature or pressure or any other non-periodic parameter) may ' be pulsed along with gas pulsing signal 402 and source RF pulsing signal 404.
• the pulsing of signal 408 may be made synchronous or asynchronous with any other signals in the system.
• another signal 410 may be chirped and pulsed along with gas pulsing signal 402,
• the frequency of signal 410 may vary depending on the phase of signal 410 or another signal (such as the gas pulsing signal) or in response to a control signal from the tool control computer.
• reference 422 points to a region of higher frequency than die frequency associated with reference number 420.
• Fig. 5 shows, in accordance with an embodiment of the invention, the steps for performing combination pulsing.
• the steps of Fig. 5 may be executed via software under control of one or more computers., for examp le.
• the software may be stored in a computer readable medium, including a nan -transitory computer readable medium in one or more embodiments.
• step 502 a substrate is provided in a plasma processing chamber.
• step 504 the substrate is processed while pulsing both the RF source and the input gas.
• step 506 Optional pulsing of one or more other signals (such as RF bias or another signal) is shown i step 506.
• the frequency, duty cycle, gas percentages, etc, may optionally ' be varied while poising the RF source and the input gas.
• the gas is pulsed such that there are at least two phases per cycle, with cycles repeating periodically.
• the other parameters including the RF source signal, ma be left impulsed.
• the reactant gas which may comprise multiple different etching and/or polymer-forming gases
• inert gas such as one or more of Argon, Helium, Xenon, Krypton, Neon, etc
• the reactant gas to inert gas ratio is at a second ratio different from the first ratio, if the ratio of reactant gas flow to total gas flow into the chamber is reduced (i .e., the ratio of inert gas to total gas flo into the chamber is increased) during the second phase, the chamber contains a higher percentage of the inert gas during the second phase than in the first phase.
• an ton-dominant plasma results wherein the plasma ion flux is formed primarily with inert gas to perform the etching.
• the ratio is changed not by adding any reactant (such as eiehant or polymer- farming) gases into the chamber but by reducing the reactant gases flow rate such that the flow percentage of inert gas to reactant gas increases.
• the chamber pressure would inherently reduce during the second phase.
• the ratio of reactant gas(es) to inert gas(es) may he changed by increasing the inert gas(es) flow into the chamber while keeping the reactant gas(es) flow into the chamber either constant or by reducing tlie reactant gas(es) flow (but not by increasing the reactant gases flow into the chamber),
• the flow of inert gas is increased to offset the reduction in the flow of reactant gas.
• the chamber pressure remains substantially the same during the first and second phases.
• the flow of inert gas is increased but is insufficient to folly offset the reduction in the flow of reactant gas.
• the chamber pressure is reduced during the second phase.
• the flow of inert gas is increased more than sufficient to offset the reduction in the flow of reactant gas. in this embodiment, the chamber pressure is increased during the second phase,
• ie percentage of inert gas(es) to total gas(es) flowed into the chamber may vary from about X% to about 100%, wherein X is the percentage of inert gas(es) to total gas flow that is present when the plasma chamber is stabilized for processing or the percentage of inert gasfes) to total gas flow that is present during the first phase.
• the percentage of inert gas(es) to total gas(es) flowed into the chamber may vary from about 1.1. X to about 100%.
• the percentage of inert gas(es) to total gas(es) flowed into the chamber may vary from about 1.5 X to about 100% during the second phase.
• the gas pulsing rate is limited at the high end (upper frequency limit) by the residence time of the gas in the chamber.
• residence time typically ranges in the tens of milliseconds for a capacitively coupled chamber.
• residence time typically ranges in the tens of milliseconds to hundreds of milliseconds for an inductively coupled chamber.
• the gas pulsing period may range from 10 milliseconds to 50 seconds, more preferably from 50 milliseconds to about 10 seconds and preferably from about 500 milliseconds to about 5 seconds.
• the inert gas added during the second phase of the periodic pulsing m y he the same inert gas or a different inert gas with different chemical composition and/or different constituent gases.
• die duty cycle of the gas pulsing rate may vary from 1 % to 99%.
• the gas pulsing rate may be chirped, i.e., may change, during processing.
• the gas pulsing may be done with a 5-second gas pulsing period with a 40% duty cycle and then switched to a 9-second gas pulsing period with either the same 40% duty cycle or a different duty cycle.
• the chirping may be done periodically in
• a chirping frequency such as 20 second chirping frequency wherein the gas pulsing frequency may he changed every 20 seconds.
• Fi g. 6 shows, in accordance with one or more embodiments of the in vention, the steps for performing gas pulsing.
• the steps of Fig. 6 may be executed via software under control of one or more computers, for example.
• the software may be stored in a computer readable medium , incl uding a no -trans itory computer readable medium in one or more embodiments.
• step 602 a substrate is provided in a plasma processing chamber.
• a plasma is generated in the chamber and stabi lized with a baseline ratio of inert gas flow to reactant gas flow.
• step 606 the ratio of inert gas flow to reactant gas flow is increased in one phase of the gas pulsing without increasing the reactant gas flow into the chamber.
• step 60S the ratio of inert gas flow to reactant gas flow is decreased, relative to the ratio of inert gas flow to reactant gas flow of step 606, in another phase of the gas pulsing without increasing the reactant gas flow into the chamber.
• the ratio of inert gas flo to reactant gas flow in step 608 may be the substantially the same as the ratio of inert gas flow to reactant gas flow of step 604 (stabilize plasma step) or may be higher or lower than the ratio of inert gas flow to reactant gas flow of stabilize step 604.
• step 61.0 the substrate is processed while the gas is pulsed by having the aforementioned, inert-to-reactant flow ratio fluctuates periodically with the ratios of steps 606 and 608.
• Figs, 7A and 7B illustrate, in accordance with embodiments of the invention, different example variations of the gas pulsing scheme discussed in connection with. Fig. 6.
• cases A, C, D, and E represents the various ratio of inert gas to reactant gas.
• the ratio of inert gas (I) to reactant gas (R) is 3:7, for example.
• the ratio of inert gas to reactant gas is 8: 1, for example.
• die ratio of inert gas to reactant gas is 1:9, for example, in case D, the gas flow into the chamber is essentially all inert While example ratio values are given, the exact values of the ratios are only illustrative; the important point is that these cases all have different ratios relative to one another.
• an example pulsing 702 may be ADAD in a preferred embodiment where the gas pulse may fluctuate -periodically between case A and case ⁇ of Fig. 7A and repeat.
• Another example pulsing 704 may be ABABAB/ADAD/ABABAB/ADAD where the gas pulse may fluctuate periodically between case A and case B of Fig. 7 A, then between cases A and D of Fig. 7A ⁇ and then back to cases A and B of Fig. 7A and repeat.
• Another example pulsing 70 ⁇ may be ABABAB/ACAC/ABABAB/ACAC where the gas pulse may fluctuate periodically between case A and case B of Fig. 7A, then between cases A and D of Fig, 7 A, and then back to cases A and B of Fig. 7A and repeat.
• Another example pulsing 70S may be ABABAB/CDCD/ABABAB/CDCD where the gas pulse may fluctuate periodically between case A and case B of Fig. 7A, then between cases C and D of Fig, 7 A, and then back to cases A and B of Fig. 7 A and repeat.
• Another example pulsing 710 may be
• ABABAB/CDCD/ADAD ABABAB/CDCD/ADAD where the gas pulse may fluctuate periodically between case A and case B of Fig. 7 A, then between cases C and D of Fig. 7 A, then between cases A and D of Fig, 7 A and then back to cases A and B of Fig. 7 A and repeat.
• Other examples may include 4 phases such as ABAB/CDCD/ADA.D/ACAC and repeat.
• the complex pulsing is highly advantageous for processes involving, for example, in situ etch-then-clean or multi-step etches., etc.
• the RF bias signal when the gas is pulsed to a high inert gas percentage or 100% or near 100% inert gas percentage in one phase of the gas pulsing cyele > the RF bias signal is pulsed high.
• the RF bias signal is pulsed low or zero, in various embodiments, the pulsing frequency of the RF bias signal may be the same or different compared to the pulsing frequency of die gas pulsing.
• the duty cycle of the RF bias signal may be the same or different compared to the duty cycle of the gas pulsing. Chirping may be employed with one or both of the RF bias signal pulsing and the gas pulsing if desired.
• the pulsing frequency, the number of pulses, the duty cycle, etc. may be varied kept constant throughout the etch or ma vary periodically or non-periodically as required.
• embodiments of the invention provide another control knob that can widen the process window for etch processes. Since many current plasma chambers are already provided with pulsing valves or pulsing mass flow controllers, the implementation of gas-pulsing in accordance with Figs. 6- 7A/7B and the discussion herein may be achieved without requiring expensive hardware retrofi tting. Further, if RF pulsing is desired in conjunction with gas pulsing, many current plasma chambers are already provided with pulse-capable RF power supplies. Accordingly, the achievement of a wider process window via gas/RF power pulsing may be obtained without requiring expensive hardware retrofitting.
• ALE atomic layer etch
• mixed mode pulsing (MMP) etching whereby the etching involves repeating a multi-step sequence, each sequence involving at least an MMP preparation (MMPP) phase and an MMP reacti ve (MMPR) phase.
• the mixed mode pulsing is configured to more fully separate torn and neutral radicals temporally (i.e., in time) in situ in a production inductively coupled plasma (ICP, also known as TCP or transformer coupled plasma in some instances) chamber or in a capac vely coupled plasma (CCP) chamber.
• ICP production inductively coupled plasma
• TCP capac vely coupled plasma
• the MMP etching is practiced in a production inductively coupled plasma (ICP) chamber to accomplish, for example, atomic layer etching (ALE) or very precise etching of the type that typically requires the use of another chamber (such as a beam-type chamber) in the prior art.
• ICP production inductively coupled plasma
• inventive MM etching allows such atomic layer etching (ALE) or precise layer- y- layer etching in the production ICP chamber substantially improves the overall throughput since there is no need to transfer the substrate from the production chamber into another chamber for such ALE or precise layer-by-layer etching.
• the inventive MMP etching also eliminates the need for specialized ALE or layer-by-layer etching equipment, thereby reducing manufacturing cost.
• MMP etching is also employed in a production ICP chamber to accomplish high selectivity etching, as will be discussed later herein.
• an ICP chamber which, construction is well known, involves the use of at least one RF -powered inductive coil for inductively coupling, through a dielectric window, RF energy to a plasma cloud formed from reactant and other gases.
• the plasma cloud is disposed below the dielectric window but above a substrate for etching the substrate.
• the substrate itself is disposed on a work piece holder, typically an ES ' chuck for example.
• the work piece holder may also be supplied with its own RF signal(s), if desired.
• RF energ provided to the work piece holder is known as bias power.
• ICP chambers are commonly employed for producing substrates in today's IC (integrated circuit) fabrication facilities and are suitable for high throughput.
• the MMP preparation phase involves using plasma to generate radicals (also known as neutrals) trom eactant gases.
• radicals also known as neutrals
• bias power is applied to the substrate work piece holder in. one embodiment.
• the elimination or minimal usage of bias power is critical for reducing the influence of ions during the
• the reactant gas may be chlorine (CI 2 for example.
• other reactant gases may be for example C & F y or CH X F V (where x and y are integers), CH3CI, Hz, BCL 3 ⁇ 4 O2. or other commonly used reactant gases for etching substrates.
• a plasma is formed from the reactant gas and allowed to adsorb into exposed top layer of the silicon substrate.
• the MMP preparation phase is timed to allow the adsorption to penetrate at least one atomic layer of silicon in one embodiment and multiple atomic layers of silicon in another embodiment if a more aggressive etch is desired,
• Parameters of the chambers are optimized to increase the speed of adsorption without unduly removing the adsorbed SiCl layer in die MMP preparation phase.
• the inductive coil RF frequency may be different during the MMP preparation phase relative to the MMP reactive phase to promote adsorption i one or more
• the substrate or the substrate surface may be heated (or cooled) during the MMP preparation phase.
• the inductive coil RF power may be pulsed on and off (either symmetrically or non-symmetrically with, respect to the duration of the on and off cycles) to reduce ion energy and/or to promote adsorption
• inductive coil RF signal(s) may be chirped with different RF frequencies during a single MMP preparation phase.
• the chamber gap between the electrodes may be set larger during the MMP preparation phase relative to the MMP reactive phase in order to lower the ion energy level, reduce self-bias, and/or reduce the influence of ions.
• the parameters may be adj usted so that the ion energy is below the level required to etch the adsorbed SiCl layer in one or more
• chamber pressure may be kept high (e.g., above 40 tnT in one example etch.) during the MMP preparation phase to reduce the ion energy in one or more embodiments.
• some non-reactive gas such as argon may be allowed during the MMP preparation phase.
• non-reactive gas flow during the MMP preparation phase if allowed, is set to be lower than the amount of non-reactive gas flow that occurs during the MMP reactive phase.
• the same non-reactive gas may be employed in both the MMP preparation phase and the MMP reactive phase or different non-reactive gases may be empioyed.
• the MMP preparation phase involves onl reacti ve gases (such as chlorine) and no non-reacti ve gases (such as argon) is empioyed during the MMP preparation phase.
• different reactive gases may be employed simultaneously during a single MMP preparation phase.
• different reactive gases may be flowed in sequential order into the chamber during the MMP preparation phase. This may be advantageous for etching binary or other compounds.
• the chamber may be flushed with a non-reactive gas (such as argon) in between the flowing of different reactive gases during the MMP preparation phase.
• ALE etches where a single atomic layer etching is desired or where etching of a small number of atomic layers is desired, it is preferable that no bias power is applied during the MMP preparation phase. In applications where a higher throughput is desired while maintaming precision, a small amount of bias power (relative to the bias power applied during the MMP reactive phase) may be applied during the MMP
• this bias power may be kept constant during the MMP preparation phase or may be pulsed (either
• non-reactive gases such as inert gases
• argon may he employed as the non-reactive gas during the MM P reactive phase.
• the non-reactive gas(es) may be Xe, He, Ne or clusters of any of the above.
• the ion energy of the Ar+ ions (which is generated from non-reactive gases in the absence of reactant gases) is above the threshold required to etch the adsorbed SiCl layer but desired to be below the threshold required to etch the non-adsorbed Si substrate below.
• the ion energy window may be between 50eV and 70eV for etching silicon in one embod ment.
• Another aspect of the self-limiting feature of one embodiment of the MMP etching is the length of time of the MMP reactive phase to ensure that only some or all of the adsorbed SiCl layer is removed and the underlying Si material is not etched.
• Another aspect of the self-limiting feature of one embodiment of the MMP etching is the length of i me of the MMP preparation phase.
• the bias power is turned on during the MMP reactive phase (in contrast, the bias power is preferably completely off during the MMP preparation phase or is turned on to a level lower than the bias power level in the MMP reactive phase to help ensure that the ion energy remains below the threshold fo ion- induced etching of the adsorbed layer).
• Other parameters of the chamber may be optimized to promote the directional etching of the adsorbed SiCl layer by the plasma that is formed from the non-reactive gas.
• the chamber pressure may be reduced in the MMP reactive phase (relative to the higher chamber pressure of the MMP preparation phase) in order to reduce the number of collisions, thereby reducing the angle distribution of the ions and resulting in a raore directtonai etch.
• the bias power may be pulsed on and off multiple times during a single MMP reactive phase.
• the RF inducti ve coil power may be pulsed on and off multiple times during a single MMP reactive phase.
• both the bias power and the RF inductive coil power may be pulsed multiple times, either synchronously or asynchronously relative to one another, during a single MMP reactive phase.
• the inductive coil RF .frequency may be different (such as higher to increase the ion energy distribution function) during the MMP reactive phase relati ve to the MMP preparation phase, in an example, the MMP reactive phase may employ 60 MHz for inductive coil RF signal while the MMP preparation phase may employ 13.56 MHz for the inductive coil RF signal during the MMP reactive phase.
• the bias RF and/or the inductive coil RF may be chirped with different RF frequencies during a single MMP reactive phase.
• a tailored bias waveform may be employed during the MMP reactive phase to reduce the ion energy.
• a taiiored bias waveform is an RF bias signal having a waveform tailored or shaped (e.g., clipped or modified) in order to optimize or regulate the ion energy).
• the MMP preparation phase and the following MMP reactive phase form a cycle, which cycle may be repeated a number of times unti l etching is deemed completed.
• art MMP transitio phase may (but not required in all cases) be interposed in between the MMP preparation phase and the MMP reactive phase to, for example, facilitate more complete removal of the reactant gas(es) and/or to stabilize and/or prepare the chamber for the MMP reactive phase.
• another transition, phase may be employed in between the MMP transition phase of a preceding cycle and an MMP preparation phase to stabilize and/or prepare the chambe for the MMP preparation phase, i one or more embodiments.
• an. MMP transition phase may be employed (but not required in all cases) in between the MMP preparation phase and the MMP reactive phase to he!p prepare the chamber for the MMP reactive phase (such as to ensure that all reactive gases are removed or to stabilize the chamber in one embodiment),
• the MMP preparation phase may be betwee about 0. 1 second to about 5 seconds, more preferably from 0.2 second to about 1 second.
• the MMP reactive phase may be between about 0.01 second to about 5 seconds, more preferably tYom 0.05 second to about 1 second.
• the switching rate may be around lHz. This is a
• the MMP reactive phase may be timed or may be terminated responsive to chamber monitoring (using for example optical emission spectroscopy techniques), in one or more embodiments, the reactive etching during the MMP reactive phase is allowed to etch only a single atomic layer (ALE).
• the adsorption may be controlled such that the adsorbed layer is around one atomic layer thick, hi one or more embodiments, the reactive etchmg during the MMP reactive phase is allowed to proceed to etch thorough multiple atomic layers of the adsorbed substrate surface.
• parameters of the chamber may be adjusted such that there is a bulk MMP reacti e etch, followed by more precise but slower monolayer MMP reactive etch during a single MMP reactive phase.
• MMP etching is employed to improve
• the reactant gas may be chosen (for use during the MMP preparation phase) such that the reactant gas forms a plasma that favors adsorption into one material over the other material.
• the gas chosen may be adsorbed onto both materials but favors the formation of volatile compounds on one material o ver the other material.
• tlie gas chosen may cause deposition more on one material than on another material.
• the gas chosen ma decrease the bonding strength at the surface of one material to a greater extent than the decrease in bonding strength at the surface of another material. Addi tionally or
• the ion energy during the MMP reactive phase may be chosen to more aggressively etch one material over another material.
• selectivity etching is etching polystlicoa but not oxide.
• the reactant gas may be chosen to be Ch during the MMP preparation phase, which does not tend to etch oxide based on chemistry considerations alone, and the ion energy threshold during the MMP reactive phase may be 70eV for poly silicon and 80eV for oxide, for example.
• Fig. 8 shows, in accordance with an embodiment of the invention, conceptual MMP etching cycles (showing species density versus time) for the silicon etching example, with each cycle involving at least an MMP preparation phase and an MMP reactive phase.
• an MMP etching cycle 802 involves at least an M preparation phase 804 and an MMP reactive phase 806. Chamber and gas conditions for each of MMP preparation phase 804 and MMP reactive phase 806 are discussed above.
• radicals and ions are separated in time, with a high amount of radicals and substantially no ions during the MMP preparation phase 804 and high amount of ions and substantially no radicals during the MMP reactive phase 806.
• FIG. 9 shows, in accordance with an embodiment of the invention, other conceptual MMP etching cycles where some ions exist in the MMP preparation phase 904.
• lorn may be present as an unintended side-effect of plasma generation but is kepi below (by manipulating chamber parameters) the threshold ion energy level necessary to etch the adsorbed SiCI surface during the MMP preparation phase 904.
• Ions may also be intentionally introduced by employing some small amount of bias power to promote implantation as discussed earlier. Nevertheless, the ion energy is kept below the
• threshold ion energ level necessary to etch the adsorbed surface during the MMP preparation phase.
• MMP reactive phase 906 reaetant gas is excluded from the chamber and preferably substantially no reaeiants are present in the chamber during the MMP reactive phase 906.
• Chamber and gas condi tions for each of MMP preparation phase 904 and MMP reactive phase 906 are discussed above.
• an MMP transition phase may be interposed between MMP preparation phase 904 and MMP reactive phase 906 if desired.
• another MMP transition phase may be interposed between preceding MMP reactive phase 906 and the MMP preparation phase 90S of the next MMP cycle.
• Fig, 10 shows, in accordance with an embodiment of the invention, method for performing MMP etching in a production ICP chamber.
• a substra te is provided in the production ICP chamber to prepare for the in situ MMP etch.
• the substrate may have been disposed in the chamber for some time and other processing steps (such as bulk, etch) may have already taken place prior to the MMP etching.
• the chamber is configured to operate in the MMP preparation phase.
• reactant gas is allowed to adsorbed into the substrate surface with the assistance of plasma.
• the depth of adsorption is controlled to form one aspect of the self-limiting etch (to he performed during a subsequent MMP reactive phase).
• Other alternative or additional chamber conditions for the MMP preparation phase are discussed above.
• the chamber is configured to etch the substrate in the MMP reactive phase.
• reactant gas is excluded from the chamber and the bias power is increased (or turned o») to promote plasma-assisted removal of the adsorbed layer(s) using a plasma formed from inert gas(s).
• the ion energy during the MMP reactive phase is set to be higher than the level necessary to etch tlie adsorbed layer hut lower than the level necessary to etch the non-adsorbed layer underneath, thereby essentiall self-limit the etch.
• Other alternative or additional chamber conditions for the MMP reactive phase is discussed above.
• the MMP cycle including at least the MMP preparation phase and the MMP reactive phase is repeated (1012) until the MMP etch is deemed (1006) completed (1008).
• embodiments of the MMP etch are highly suitable for ALE etch or precise etches (such as etches for fabrication 3-D logic or memory devices or M AM) or high selectivity etches. Furthermore, embodiments of the invention reduce substrate damage and result in a flat etch front.
• the self-limiting nature and/or high selecti vity of the MMP etch helps reduce structural damage to layer(s) or structure(s) that should not be etched, hi some cases, the self-limiting nature of the MMP etch helps improve etch precision and/or etch profile and/or may reduce tlie need for Qveretching.
• MMP etch has been disclosed vising an ICP chamber example, MMP etching may be performed in a capactttvely coupled plasma (CCP) chamber if desired.
• CCP capactttvely coupled plasma
• the supplied higher RF frequency may be considered the source RF and the supplied lower RF frequenc may be considered the bias RF irrespective whether these RF signals are provided to only one plate of the chamber or split up among the plates of the chamber.
• the pulsing techniques discussed in the figures may be combined in any combination to suit the requirement of a particular pr ocess.
• the duty cycle variance may be practiced with techniques discussed with any one (or part of any one or a combination of multiple ones) of the figures.
• the frequency chirping may be practiced with techniques discussed with any one (or part of any one or a combination of multiple ones) of the figures and/or with duty cycle variance.
• inert gas substitution may be practiced with techniques discussed with any one (or part of any one or a combination of multiple ones) of the fi gures and/or with duty cycle variance and/or with frequency chirping.

Landscapes

• Physics & Mathematics (AREA)
• Engineering & Computer Science (AREA)
• Plasma & Fusion (AREA)
• Chemical & Material Sciences (AREA)
• Analytical Chemistry (AREA)
• Drying Of Semiconductors (AREA)
• Plasma Technology (AREA)
• Internal Circuitry In Semiconductor Integrated Circuit Devices (AREA)

Priority Applications (4)

Applications Claiming Priority (4)

Publications (2)

Family

ID=48694019

Family Applications (1)

Country Status (7)

Cited By (1)

Families Citing this family (466)

Family Cites Families (37)

• 2012
  • 2012-07-16 US US13/550,548 patent/US8883028B2/en active Active
  • 2012-12-17 SG SG11201403634TA patent/SG11201403634TA/en unknown
  • 2012-12-17 CN CN201610451176.8A patent/CN105895490B/zh active Active
  • 2012-12-17 JP JP2014549582A patent/JP6276704B2/ja active Active
  • 2012-12-17 WO PCT/IB2012/057385 patent/WO2013098702A2/en not_active Ceased
  • 2012-12-17 SG SG10201608686TA patent/SG10201608686TA/en unknown
  • 2012-12-17 KR KR1020147021178A patent/KR102062930B1/ko active Active
  • 2012-12-17 CN CN201280065464.XA patent/CN104040021B/zh active Active
  • 2012-12-25 TW TW106142833A patent/TWI654680B/zh active
  • 2012-12-25 TW TW101149781A patent/TWI612576B/zh active
• 2014
  • 2014-10-09 US US14/510,866 patent/US9425025B2/en active Active
• 2016
  • 2016-07-13 US US15/209,682 patent/US10121639B2/en active Active

Also Published As

Similar Documents

Legal Events