throbber
A 4
`
`ELSEVIER
`
`Surface and Coatings Technology 93 ( i998 ) 12454250
`
`AYIIIHIIE
`a£'J477NE3
`
`/zzmmrr
`
`
`
`Asymmetric bipolar pulsed DC: the enabling technology for reactive PVD
`
`J. Sellers *
`
`EN], 4150 Freidrich Ln., Suite J, Austin. TX 78744, USA
`
`
`
`Abstract
`
`The use of reactive DC sputtering for the deposition of insulators from conductive targets has been limited by the intrinsic
`problem of target poisoning and the consequent arcing and process instabilities. The need to deposit high quality dielectric films
`rapidly is becoming more important as technology pushes forward. Asymmetric bipolar pulsed DC eliminates target poisoning
`through preferential sputtering, enabling existing PVD tools to produce the high—quality, low-defect dielectric films needed for
`next generation processes. Typical films being produced with asymmetric bipolar pulsed,DC from metallic targets include A1203,
`AIN, Si02, SiN, Ta2O,, DLC. TaN, TiN and ITO. The mechanisms of target poisoning and dielectric arcing are explained in this
`paper, and solutions are given. © 1998 Published by Elsevier Science S.A.
`
`Keywords: Sputter; Reactive; Magnetron; Thin film; PVD
`
`1. Introduction
`
`Asymmetric bipolar pulsed DC technology has been
`quickly proving its effectiveness
`in a wide variety
`of standard and reactive
`sputtering applications.
`Acceptance of this technology has been extremely rapid
`due to its elimination of many long-standing process
`limitations. Asymmetric bipolar pulsed DC was specifi-
`cally developed to optimize the deposition of insulating
`films from conductive targets with reactive sputtering.
`Further, it is important to understand that this technol-
`ogy did not evolve as an afterthought to control arcing:
`rather, it was conceived with a definite understanding
`of the electrical and physical process requirements. This
`understanding led to a unique solution in which the
`power source is given the dual properties of a current
`source and a voltage source, depending on which plasma
`constituent is being driven. Thus, both ions and electrons
`are driven in their optimal manner, and the plasma’s
`asymmetry of masses (ions versus electrons) is matched
`by the supply’s source dual characteristics (current
`source (forward sputter) versus voltage source (reverse
`bias). Films successfully reactively sputtered with this
`technology include A1203, Ta2O5, BST, PZT, Ta,O5,
`TaN, TiO2, TiN and ITO. Additional applications for
`
`the technique include etching/cleaning, CVD bias, and
`substrate bias for sputtered films.
`
`2. Target poisoning
`
`The key to the successful production of insulating
`films from metallic targets is the elimination of target
`poisoning. Poisoning is the build-up of insulating layers
`on the target surface. In simple metallic sputtering, the
`target (or cathode) is driven by the sputtering supply to
`a specific DC voltage based on the power, chamber
`pressure, magnetron design. etc. This voltage accelerates
`the argon ions into the target with sufficient kinetic
`energy to cause them to knock atoms from the target.
`The freed target atoms then condense on the substrate
`to form the desired film. The free atoms also deposit on
`the walls of the chamber and back on the target. In
`pure metallic sputtering, this redeposition on the target
`does not represent a problem, since the target and the
`redeposited atoms are the same material. But in reactive
`sputtering, the deposited film is a compound, and there-
`fore a different material
`to the target. For example,
`aluminium oxide, a ceramic, has properties which are
`quite dilferent from metallic aluminum. If the reactive
`film is an insulator, such as aluminum oxide, then the
`situation becomes intolerable.
`
`“ Corresponding author. Tel: +1 512 4622191; Fax: + I 512 2462941 I.
`
`When an insulator is deposited on the surface of the
`
`0257-8972/98/$19.00 © 1998 Published by Elsevier Science S.A. All rights reserved.
`Pl! S0257-8972(97)00403-9
`
`GILLETTE 1022
`
`GILLETTE 1022
`
`

`
`1245
`
`.1. Sellers /' Sur/(we and Coutings Tet‘/inology 98 ( I 998) I245-1 250
`
`target, a capacitor is formed (see Fig. l). The target acts
`as one conductor, the plasma acts as the other conduc-
`tor, and the insulating film forms the dielectric of the
`capacitor. DC current cannot flow through a capacitor.
`This results in two problems. First, current flow is ion
`current, and therefore, if no argon ions strike the area.
`then no target atoms can be freed and no sputtering can
`occur. Consequently, this area of the target is poisoned.
`Second,
`the parasitic capacitor may not have enough
`dielectric capability to charge all
`the way up to the
`applied voltage. If not. the breakdown of the insulator
`will cause a sudden release of charge carriers. forcing
`the local current density to increase into the are dis-
`charge region, which results in arcing with all its atten-
`dant particulate problems.
`Asymmetric bipolar pulsed DC is the optimum solu-
`tion to the target poisoning problem because it sets up
`conditions which cause the insulators on the target to
`be sputtered first and with a higher sputter yield than
`the base material (a mechanism called preferential sput~
`tering). This eliminates target poisoning by removing
`the cause.
`
`3. Preferential sputtering
`
`Preferential sputtering is accomplished by adding a
`reverse voltage bias pulse to the normal DC waveform
`(see Fig. 2). First,
`if typical sputtering runs at about
`
`-400 V. then the conditions are as shown. Argon ions
`accelerate towards the target at ——400 V, striking the
`target and sputtering the aluminum. However, the rede-
`posited insulating film on the target behaves in a very
`different manner. As the film forms, it tends to collect
`
`low-energy ions on its surface, charging the parasitic
`capacitor towards the applied voltage. As capacitor
`voltage climbs, the available energy with which the ions
`can strike the insulator is decreased by the voltage on
`the capacitor, thus reducing the likelihood of the film
`being sputtered off. Even worse, as the charge builds
`up,
`the ions are actually repelled by the electrostatic
`repulsion of the Ar* ions and the positive capacitor
`voltage. Next, the polarity is rapidly reversed to about
`+100 V (see Fig. 3), causing the plasma—facing surface
`of the dielectric film (parasitic capacitor) to be charged
`up to the opposite polarity (— 100 V). The magic occurs
`as the reverse pulse ends and the voltage returns to
`sputter mode (-400 V), as shown in Fig. 4. Since the
`plasma side of the parasitic capacitor is now charged to
`-100 V. when the target reaches —40OV the effective
`voltage on the plasma side of the parasitic cap is
`— 500 V. Thus, the argon ions are drawn by electrostatic
`attraction to the insulators, and strike with extra energy
`(500 V versus 400 V), thereby sputtering the insulators
`off the target first, eliminating target poisoning.
`The effectiveness of asymmetric bipolar pulsed DC is
`also dependent on pulse frequency. The charging pulses
`must occur frequently enough to prevent the build-up
`
`Dielectric breakdown
`(Micro-Arcing)
`
`Plasma
`
`
`W (dielectri /%C)
`
`tare
`
`
`
`A1403 (dielectric)j
`
`
`
`Fig. 1. Fundamental capacitor model.
`
`
`
`Time
`
`Fig. 2. Normal sputter mode.
`
`Vsurface =Vsputter
`
`
`
`

`
`J. Sellers / Surface and Coatings Technology 98 ( 1998) 12454250
`
`l247
`
`surface = '
`
`Vreversal
`
`
`73
`T0
`T1 T2
`Time
`
`Fig. 4. Return to sputter mode.
`
`be driven by a very still‘ current source, giving a constant
`current regardless of the voltage required, and allowing
`the full forward current to be re-established immediately.
`These opposite drive requirements reflect the asymmetriv
`cal nature of the plasma itself (i.e. electron mass versus
`ion mass).
`In actual use, asymmetric bipolar pulsed DC technol-
`ogy has demonstrated reactive deposition rates of
`60~l00% of the metallic deposition rate [3]. A typical
`example would be reactive TaN film (see Table I). This
`is a much higher rate than any competitive techno-
`logy. Bipolar dual cathode systems have difiiculty
`approaching 50% due to the lost duty cycle in the
`transitions from positive to negative sputter voltage.
`Also, they require two magnetron target assemblies (see
`
`Table l
`Rate and unifomtity comparison (TaN)
`
`
`
`Technique
`
`Rate
`
`Uniformity (%)
`
`of voltage on the parasitic capacitors from exceeding
`their breakdown [1]. Typical process frequencies are
`80-150 kHz. As this technology has been applied to
`materials other than true insulators, it has become clear
`that preferential sputtering is also effective for resistive
`films. Slightly higher frequencies are needed due to the
`tendency of the resistive film to self-discharge. TaN,
`TiN and ITO seem to work best at around 150-180 kHz.
`The value of the reverse bias voltage is critical. It must
`be chosen to be low enough not to generate backsputter—
`ing, and yet high enough to maintain preferential sput-
`tering. Experience on many different chambers has
`shown 75«lO0V to be effective and safe. Full bipolar
`pulses have enough voltage to sputter in both directions.
`Clearly, if there is enough voltage to sputter one way,
`then there is enough to sputter in the other direction
`also. Most magnetrons are asymmetrical in their voltage
`withstand, and will tolerate up to about + l50 V.
`
`4. Deposition rate
`
`It is equally critical to return to forward current (ion
`current) flow as quickly as possible after the reverse
`bias pulse in order to maximize the deposition rate. This
`drives the requirement for the forward sputter power to
`
`2
`1.06
`RF diode
`ll.2
`2.00
`RF magnetron
`Not long—term stable
`3.53
`DC magnetron
`
`
`3.48Pulse DC magnetron l.97
`
`

`
`1248
`
`J. Sellers / Surface and Coutings Teclmology 98 ( 1998) 1245-1250
`
`Fig. 5). Asymmetric bipolar pulsed DC technology was
`optimized to sputter insulators from a conductive target.
`
`5. Uniformity enhancement
`
`Asymmetric bipolar pulsed DC technology has proven
`to be particularly beneficial for the enhancement of film
`qualities. Film uniformity (deposition uniformity and
`film characteristics) has demonstrated benefits from the
`use of pulsed DC techniques. There are two mechanisms
`at work in these results. First, in the case of a complex
`film (the deposited film is a compound), the additional
`ionization of the pulsed plasma results in a hotter
`(greater electron temperature) and more chemically
`active plasma, which tends to improve the consistency
`of the film chemistry. This ionization enhancement
`comes from the high-frequency (edge rates) and the mid-
`frequency (pulse rate) components of the waveforms.
`The second mechanism is the eiTects of peak to average
`power ratio on the deposition pattern. The greater the
`power applied to a magnetron, the wider the erosion
`zone (race track). This occurs due to the opposing
`electrostatic and electromagnetic forces of the magnet-
`
`+400
`
`Amps
`
`ron and the plasma ions. The magnetron has a preferred
`location for the ions to strike due to the geometry of
`the intersection of the electrical and magnetic fields.
`This “sweet spot" is located at the juncture of the peak
`magnetic and electrical fields, at the point where E x B
`is a maximum (see Fig. 6). At low power density the
`ions will simply form a narrow ring at this “sweet spot”
`radius. resulting in a very narrow erosion zone. As the
`power is increased, the ions begin to crowd together in
`the “sweet spot”, but the electrostatic repulsion for the
`like-charged ions forces them to stay apart. Thus, the
`ions are forced into a widening pattern around the
`“sweet spot” radius (see Fig. 7). As the ions are forced
`apart, the additional energy required to generate ioniza-
`tion at the lower magnetic field intensity is compensated
`by an increase in plasma voltage. As the erosion zone
`(race track) becomes wider,
`the deposition pattern at
`the substrate also becomes more uniform. Another effect
`
`
`
`Full Bipolar
`
`"Sweet Spot": Max Point of EXB
`
`Fig. 6. "Sweet spot".
`
`
`
` — Racetrack
`
`Low Power
`
`High Power
`
`Fig. 5. Waveform comparison curves.
`
`Fig. 7. Uniformity.
`
`

`
`J. Sellers / Surface and Caating: Technology 98 (1998) 1245-1250
`
`1249
`
`Table 2
`Pulse DC etch results
`
`Substrate Contaminant Press (mT) Frequency (kHz) PW (us) Power (W)Run Approx. etch rate (nm 5‘ ‘)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1
`2
`3
`4
`5
`6
`
`Silicon
`Silicon
`Silicon
`Steel
`Steel
`Steel
`
`AL,(),
`ALZO-3
`AL,O3
`Light oils
`Heavy rust
`Heavy rust
`
`30
`70
`46
`33
`39
`75
`
`162
`162
`185
`215
`2l5
`63
`
`1.32
`l .3
`1.2
`1.01
`1.01
`l.6
`
`192
`193
`308
`385
`413
`l20
`
`0.02
`0.01
`0.056
`0.2
`0,2
`0.05
`
`processes. Work is presently in progress to harness the
`clear benefits of the technique, which include no RF, no
`RF match, direct DC—type control of the substrate bias
`energy and the rate improvements due to the increased
`bias duty factor. Several types of films, including DLC
`and B4C, have shown benefit from asymmetric bipolar
`pulsed DC bias.
`Additionally, the use of substrate bias can allow the
`elimination of some adhesion layers due to the enhance-
`ment of adhesion levels achieved by the use of high-
`energy bias during the growth of the sputtered films.
`The mechanisms of this adhesion enhancement are the
`dual effects of surface activation of the substrate through
`pre-etching and the additional energy available to the
`growing film from the substrate ion current. As the film
`begins to form, the early stages of formation are the
`most critical to the characteristics of the film, since the
`
`first layer will tend to set the form (crystallography) and
`adhesion for the entire film [2,4]. The addition of energy
`to this early film growth allows the atoms to achieve
`greater mobility and find the low-energy wells on the
`substrate surface where the adhesion will be maximized.
`
`Also, the high flux of energetic ions will tend to wash
`away any poorly adhered film atoms. These effects are
`also enhanced because of the increased ionization of the
`
`asymmetric bipolar pulsed DC. Strongly ionized film
`atoms can be implanted into the substrate due to the
`bias energy adding to the energy of the film atoms.
`Clearly, there is never a “free lunch”, so the negative
`efl"ect of using high—energy bias is a reduction in the
`overall deposition rate due to the resputtering of the
`growing film. The use of bias to generate crystalline
`films at low substrate temperatures is also progressing
`well.
`
`7. Conclusion
`
`Asymmetric bipolar pulsed DC technology is the
`future of DC plasma processing because of its ability to
`extend the range of DC processes and to broaden the
`use of sputtered film applications into entirely new
`markets.
`
`of the high power density ion crowding is an increase
`in the ionization of the sputtered material. This is caused
`by an increase in collisions between the freed target
`atoms and the incoming argon ions. As the density of
`freed target material and ions increases, the frequency
`of collisions between ions and molecules also increases.
`These effects are applicable to asymmetric bipolar pulsed
`DC simply due to the high peak to average power ratios
`attainable with the technology (see Fig. 8).
`
`6. Other applications
`
`6.1. Etching/cleaning
`
`Just as it can remove insulating coatings from conduc-
`tive sputter targets, asymmetric bipolar pulsed DC tech-
`nology can remove non-conductive contamination from
`steel tools, silicon wafers and other conductive materials
`(see Table 2).
`
`6.2. Bias/source
`
`The AC characteristics of the technology also allow
`it to be applied to substrate bias of insulating CVD
`
`Current
`
`llIII
`
`Peak Power = 4.8kW
`Av. Power = 1kW
`Peak — Av. Ratio = 4.8
`
`
` Voltage
`
`Fig. 8. Peak power.
`
`

`
`1250
`
`References
`
`J. Sellers / Surface and Coatings Technology 98 ( I998) 1245-1250
`
`_
`.
`‘
`_
`{1} S. Schiller. K. Goedicke. J. Reschke. V. Krrchofi, S. Schneider, F.
`Milde. Pulsed Magnetron DC Sputtering Technology, FhG FEP.
`ICMCTF. San Diego. CA. I993. PPr9-10-
`[2] 5- Fessman. 0. Kamschulte. J. Ebbemik. Improved control of TiN
`coating properties using cathodic arc evaporation with a pulsed
`bias by W. Olbrich_ Surf, Coat, Technol, 49 (1991) 258,
`
`[3] J. Schneider, W. Sproul, A. Lefkow, Birl, A. Matthews. U. Hui].
`‘
`Scaleable Process for Pulsed DC Magnetron Sputtering of Non-
`
`Conductive Oxides, 39th l996 SUC Tech. Cont". Proceedings,
`phma_ PA p_,6g_
`[4] Ammina.chmmica coatings deposited by reactive magnetron sput-
`tering. J.M. Schneider. WODA Sproul. Birl, Northwestern Univer-
`sity, Evanslon IL; J. Sellers. ENI-SW. Austin. TX and A. Mathews,
`RCSE.Hul1University. United Kingdom. 1997 Society of Vacuum
`Coaters, 40th Technical Conference Proceedings 1997.

This document is available on Docket Alarm but you must sign up to view it.


Or .

Accessing this document will incur an additional charge of $.

After purchase, you can access this document again without charge.

Accept $ Charge
throbber

Still Working On It

This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.

Give it another minute or two to complete, and then try the refresh button.

throbber

A few More Minutes ... Still Working

It can take up to 5 minutes for us to download a document if the court servers are running slowly.

Thank you for your continued patience.

This document could not be displayed.

We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.

Your account does not support viewing this document.

You need a Paid Account to view this document. Click here to change your account type.

Your account does not support viewing this document.

Set your membership status to view this document.

With a Docket Alarm membership, you'll get a whole lot more, including:

  • Up-to-date information for this case.
  • Email alerts whenever there is an update.
  • Full text search for other cases.
  • Get email alerts whenever a new case matches your search.

Become a Member

One Moment Please

The filing “” is large (MB) and is being downloaded.

Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.

Your document is on its way!

If you do not receive the document in five minutes, contact support at support@docketalarm.com.

Sealed Document

We are unable to display this document, it may be under a court ordered seal.

If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.


Access Government Site

We are redirecting you
to a mobile optimized page.





Document Unreadable or Corrupt

Refresh this Document
Go to the Docket

We are unable to display this document.

Refresh this Document
Go to the Docket