PLASMA ELECTRODES

Cross-Reference to Related Applications

[0001] This patent application claims priority to U.S. Patent Application No. 14/198,429 entitled ELECTRICALLY- AND CHEMICALLY-ACTIVE ADLAYERS FOR PLASMA ELECTRODES, filed on March 5, 2014, which is incorporated herein by reference in its entirety.

Background

[0002] Plasma energy is sometimes used to facilitate or activate chemical reactions for thin film deposition. When applied to a gaseous precursor, plasma energy can generate neutral radicals and/or ions, transforming the more stable precursor into an activated species. The energetic activation barrier for a subsequent film forming reaction of the activated species would be expected to be lower than for the precursor. In turn, the deposition reaction should proceed with little need for thermal energy. Eliminating thermal reactions can be advantageous when the reaction temperature can activate undesired alternative reaction paths or alter the mechanical or physical properties of the substrate or other materials formed thereon.

[0003] Mass transfer limitations that can cause surface thickness non-uniformity are typically absent in kinetically-limited chemical vapor deposition (CVD) processes (e.g., some atomic layer deposition (ALD) processes). In an ALD process, a substrate is sequentially exposed to different precursor compounds, or precursors. These precursors chemically adsorb, or chemisorb, via chemical reaction with the substrate surface, to form metastable chemisorbed species. Subsequent exposure of these species to a different precursor leads to a surface reaction causing that precursor to become chemisorbed to the surface and incorporated therein, building up the film. In between exposures, non-chemisorbed amounts of the precursors are removed from the surface of the substrate (e.g., by evacuation and/or displacement via purging with a non-reactive gas) so that no more than chemisorbed amounts of one particular precursor is present on the surface during exposure to another precursor. This layer-by-layer exposure creates highly uniform films.

[0004] For example, some metal oxide ALD processes use organometallic compounds that may suffer from unwanted decomposition reactions when exposed to elevated temperatures. The metal atoms in these compounds are often stabilized with ligands bound to the metal atom which are broken upon adsorption of the precursor molecule to the surface.

[0005] In thermal ALD, surface energy activates all of the bond breaking and forming reactions during chemisorption for all of the precursors used to form the film. Because it is often problematic to vary the substrate temperature between subsequent exposure steps, it can be challenging to find a combination of precursors that will thermally decompose to form chemisorbed species stable enough to remain on the surface and active enough to remain reactive. That is, if the first precursor is too surface-stable, it may block chemisorption and/or reaction of the second precursor; if it is not stable enough, it may decompose.

[0006] In plasma enhanced ALD, the plasma supplies the energy to form a reactive intermediate. This intermediate can then react with a chemisorbed precursor to form a layer of film. While offering some advantages relative to thermal ALD, plasma ALD can be problematic, as the identity and concentration of reactive species formed within the plasma can be difficult to control. This can affect the chemical composition and/or the performance properties of the deposited layers.

Brief Description of the Drawings

[0007] FIG. 1 is a schematic view of an embodiment of a method of forming a thin film;

[0008] FIG. 2 is a model reaction scheme illustrating chemical activity between a precursor gas and a chemically active electrode surface according to an embodiment of the present disclosure;

[0009] FIG. 3 is a fragmentary perspective view of an embodiment of an electrode;

[0010] FIG. 4 is a schematic fragmentary view taken at location 4 of the embodiment shown in FIG. 3;

[0011] FIG. 5 is a schematic cross-sectional view of an embodiment of a roll-to- roll deposition system;

[0012] FIG. 6 is a graphical relationship between a duration of a plasma exposure time and deposition rates of T1O ₂ films formed according to embodiments of the present disclosure; [0013] FIG. 7 is a graphical relationship between the thickness of ΤΊΟ ₂ films deposited according to embodiments of the present disclosure and the water vapor transmission rate for those films;

[0014] FIG. 8 is a graphical relationship of the average plasma power in watts (W) as a function of the identity of coating materials on embodiments of chemically active cathodes;

[0015] FIG. 9 is a schematic, fragmentary cross-sectional view of an embodiment of a thin film deposition system; and

[0016] FIG. 10 is a schematic, fragmentary cross-sectional view of another embodiment of a thin film deposition system.

Detailed Description of Preferred Embodiments

[0017] FIG. 1 is a schematic view of an embodiment of a chemical vapor deposition (CVD) process 100 of forming a thin film 102 from two precursors, Pi and P ₂ . In some embodiments, process 100 may occur via atomic layer deposition (ALD). In FIG. 1 , thin film 102 is formed by a reaction between a chemisorbed amount of a first precursor 103 (shown in a chemisorbed state as indicated by an asterisk, Pi ^* ) on substrate 104 and radicals formed by plasma activation of a second precursor 106 (shown in a radical form as indicated by a dot, e.g., Ρ ₂ ·) that is chemically sensitive with Pi ^* . In the embodiment shown in FIG. 1 , a process gas including P ₂ and an optional diluent gas 108 (labeled I) mixed therewith is provided to a plasma generator 1 10 to generate a plasma 1 12. In the context of substrate deposition and processing, process gases refer to those gases to which substrate 104 is exposed before, during, and after thin film deposition. In the embodiment shown in FIG. 1 , process gases include first precursor gas 103, second precursor gas 106, and diluent gas 108. Diluent gas 108 is essentially non-reactive or inert with either precursor. In the embodiment shown in FIG. 1 , second precursor gas 106 and diluent gas 108 are both plasma sensitive process gases that may be ionized by a plasma generation field to form plasma 1 12. Example plasma sensitive process gases include, but are not limited to, argon (Ar), helium (He), oxygen (O ₂ ), carbon dioxide (CO ₂ ), carbon monoxide (CO), nitrogen (N ₂ ), nitrogen monoxide (NO), nitrous oxide (N ₂ O), ammonia (NH ₃ ), and nitrogen dioxide (NO ₂ ).

[0018] During use, plasma generator 1 10 activates gas molecules by applying an electric, magnetic, microwave, and/or radio frequency field to ionize at least some of the process gas, generating plasma 1 12. Activation of the process gas by the field creates free electrons within plasma 1 12 which are accelerated by the field. When these electrons collide with a precursor or diluent gas molecule, they may activate the molecule to generate various excited and/or ground state radicals, ions, and other species within plasma 1 12.

[0019] As shown in FIG. 1 , plasma generator 1 10 includes at least one pair of electrically conductive electrodes (shown at 1 14 and 1 16) that act as plasma generation surfaces by creating the discharge that activates the mixture of precursor and optional diluent gases. The electrically conductive electrodes are electrically insulated or isolated from one another and have different electrical potentials with respect to one another during operation to create a plasma discharge within a plasma generation region where plasma 1 12 is ignited and sustained. In some embodiments, one electrode may act as a cathode (e.g., plasma generation surface 1 14) that supplies electrons to the plasma, while the other may act as an anode (e.g., plasma generation surface 1 16) that drains electrons from the plasma. In some embodiments electrodes 1 14, 1 16 may include standalone electrode elements. For example, embodiments of the cathodes and anodes disclosed herein may be included on respective detachably mounted cathode and anode elements. In some embodiments electrodes 1 14, 1 16 may be included on another structure proximal to the plasma generation region. For example, a reaction vessel wall and/or a dark space shield may act as an anode relative to a cathode that is electrically insulated from the vessel wall and/or dark space shield. In some embodiments, one of the pair of electrodes need not be exposed to the gases in the plasma generation region. For example, an antenna proximal to the plasma generation region but separated therefrom by an insulating material (e.g., quartz) may act as an electrode.

[0020] Electrodes 1 14 and 1 16 are shown electrically coupled to a plasma controller 1 18 in the embodiment shown in FIG. 1 . Plasma controller 1 18 controls the distribution of electrical power from a power supply to electrodes 1 14 and 1 16. In some embodiments, plasma controller 1 18 may control power supplied to electrodes 1 14 and 1 16 so that both electrodes "float" and exhibit electrical potentials different than an earth potential. In some embodiments, plasma controller 1 18 may control power supplied to one of electrodes 1 14 or 1 16 while the other of electrodes 1 14 or 1 16 is earthed. [0021] Primarily plasma 1 12 is formed in an area where a field is applied by plasma generator 1 10 when a plasma sensitive process gas is supplied. In some embodiments, the total and/or partial gas pressure may affect the shape, density, and size of plasma 1 12. In some embodiments, the process gas may be held at a low pressure during plasma generation. For example, total gas pressure in a processing zone containing plasma generator 1 10 may be between about 0.1 Torr and 5 Torr, or in some other embodiments, between about 1 .2 Torr and about 2 Torr, or in still other embodiments, 0.5 Torr or less. Thus, the size, shape, and position of plasma 1 12 as shown in FIG. 1 are shown caricatured for purposes of illustration. Generally, a plasma region enveloping the electrical, magnetic, and/or radio frequency discharge of electrodes 1 14 and 1 16, and thus containing plasma 1 12, is positioned in the vicinity of a complementary pair of electrodes.

[0022] In the embodiment shown in FIG. 1 , substrate 104 is exposed, at step 120, to a mixture of radicals and stable precursor and diluent gases. The lifetime for some radicals formed in plasma 1 12 may be different from, and in some cases greater than, a lifetime for the free electrons and ions present in plasma 1 12. Some radicals generated in plasma 1 12 may be transported out of plasma 1 12 via a gas stream, which is referred to herein as remote or indirect plasma generation of radicals.

[0023] At least some of the radicals formed from second precursor 106 (Ρ ₂ ·) react (shown at step 122) with the chemisorbed first precursor 103 (Pi ^* ). Given a sufficiently long exposure to Ρ ₂ ·, eventually a majority of the chemisorbed first precursor 103 present on substrate 104 reacts (as indicated at 124), to form thin film 102.

[0024] Electrode 1 14 includes, on a surface exposed to the plasma generation environment, a chemically active adiayer 1 15, shown in FIG. 1 with a thickness greatly exaggerated for clarity. Adiayer 1 15 exhibits surface property characteristics that cause the electrode surface to be chemically active with at least one of the gases present within the plasma generation region. For example, adiayer 1 15 is substantially conserved while exposed to plasma 1 12. That is, a majority, and in some embodiments, nearly all, of the amount of adiayer 1 15 present on electrode 1 14 before plasma ignition remains on the surface during plasma operation. In some embodiments, a service life for chemically active adiayer 1 15, or a time during which adlayer 1 15 chemically active with at least one of the gases present within the plasma generation region (which may be substantially the same as a service life for electrode 1 14 in some embodiments), may have a duration of about an hour. In some other embodiments, an adlayer service life may have a duration of between an hour and 10 hours; or between 10 and 100 hours; or 100 and 1000 hours; or more than 1000 hours.

[0025] Because adlayer 1 15 is substantially conserved on electrode 1 14 (e.g., preferably), the chemically active adlayer is not expected to be removed via ion sputtering during plasma processing by more than an incidental amount. Nevertheless, it is anticipated that a small amount of adlayer 1 15 may be sacrificed over the life of electrode 1 14. These losses are expected to be minor, but may still be detectable as contaminants in the deposited films. In some embodiments, small amounts of optional, sacrificial dopant elements may be included along with the chemically active material. It is not believed that the dopants will contribute, at a substantial level, to the chemical activity of the surface, as the dopants are expected to be preferentially removed from the surface. In some of such embodiments, adlayer 1 15 may include, in an amount of about 15% or less, a p-type dopant such as one or more of fluorine (F), antimony (Sb), or aluminum (Al), or in some embodiments consisting essentially of F, Sb, or Al.

[0026] In some embodiments, adlayer 1 15 may include a material that is catalytically active with respect to a precursor gas present in the plasma generator during operation. Generally, a material is catalytically active for a chemical reaction if the material accelerates the rate of the reaction without being substantially consumed by the reaction. Non-limiting examples of catalytic materials include some transition metals, including one or more of palladium (Pd), platinum (Pt), cobalt (Co), iridium (Ir), or nickel (Ni), or in some embodiments consisting essentially of Pd, Pt, Co, Ir, or Ni, though catalytically active materials are not limited to metals.

[0027] However, though adlayer 1 15 is substantially conserved and chemically active, in some embodiments it need not behave catalytically. In some embodiments, adlayer 1 15 may include a material stabilized, during plasma generation, into an electrically conductive state with which one or more gases present in the plasma may be chemically active. For example, an oxygen plasma may be chemically active with an electrically conductive metal oxide or a mixture of metal oxides or included in forming the adlayer. One non-limiting example adlayer 1 15 includes a mixture of indium and tin oxides (e.g., ITO) stabilized in a partially reduced state during plasma operation. Another non-limiting example adlayer 1 15 consists essentially of (ln ₂ O3)i _-x (SnO)x stabilized under plasma conditions.

[0028] FIG. 2 is an embodiment of a reaction model 200 illustrating chemical activity 202 between an oxygen (O2) precursor gas and a metal (M _x ) included in a cathode surface (e.g., adlayer 1 15 shown in FIG. 1 used as a cathode), and electron transport 204 within the electrode surface, as described in more detail below. FIG. 2 is merely a simplified schematic model and is not meant to illustrate the exact mechanism of any particular reaction scheme. Generally, as shown in FIG. 2, two stable oxygen radicals (Ο·) are formed along with an electron (e ^" ), though it will be appreciated that a variety of oxygen ions and excited and/or ground state radicals may also be formed at the electrode surface within the highly energetic plasma environment.

[0029] In some embodiments, the chemically active adlayer may be capable of catalytically transforming some of the second precursor gas into radicals. Applying reaction scheme 200 to such embodiments, M _x catalytically cycles between oxidation states to dissociate oxygen gas adsorbed to the electrode surface, forming oxygen radicals. At step 202, O2 chemisorbs to the metal and temporarily oxidizes it, as the pair of electrons in an oxygen atom might oxidize platinum (Pt) from Pt ⁺² to Pt ⁺⁴ . Electrons supplied to the electrode surface material by the plasma power source reduce the metal at step 204, releasing the oxygen radicals and the electron as shown at step 202 in FIG. 2.

[0030] A surface that is chemically active and substantially conserved during plasma generation need not be limited to catalyst materials or to materials exhibiting catalytic activity. In some embodiments, the electrode surface may be chemically active with one of the gases in the plasma environment as a consequence of the electrical environment created by the plasma and/or plasma generation. Some partially reduced metal oxides (that is, a suboxide of the metal) are believed to be stabilized under some plasma generation conditions. For example, indium-tin-oxide (ITO, or (ln ₂ O3)i _-x (SnO2)x) is believed to be reduced to (ln ₂ O3)i _-x (SnO)x in the presence of the radio frequency used in some capacitively coupled plasmas. In turn, plasma generated radicals may maintain and/or promote the electrical conductivity of the electrode. [0031] Applying model reaction scheme 200 to such embodiments, a metal cation Μχ within a metal oxide held in a partially reduced state (in this example, M _x ^+a ) may become temporarily oxidized into an oxidation state M _x ^{+(a+1 )} by adsorption of an oxygen atom to the electrode surface at step 202. Electrons believed to be traveling through oxygen vacancies within the partially reduced metal oxide lattice may cause metal-oxygen bonds to break at step 204. The oxygen atoms freed may leave the surface as oxygen radicals and/or ions. Naturally, steps 202 and 204 may happen concurrently, so that the oxidation of one metal atom is balanced or nearly balanced by the reduction of a different metal atom elsewhere in the partially reduced metal oxide material.

[0032] Regardless of whether the chemical activity occurs catalytically or not, the concentration of metastable radicals formed from the precursor gases exposed to the plasma is believed to be altered by the chemically active adlayer, which is expected to affect the throughput rate for substrate 104 through the reaction space or vessel.

[0033] FIG. 3 is a fragmentary perspective view of an embodiment of an electrode 300. Electrode 300 includes an electrode body 302 that extends into or faces into a plasma generation region when installed in a plasma generator. Electrode body 302 is typically formed from electrically conductive material, like a metal. In some examples, electrode emissive material may be located on a support (which may be conductive or nonconductive) (not shown) to provide mechanical strength. The embodiment illustrated in FIG. 3 is a simple electrode fin that is electrically coupled to a power supply included in a plasma controller 306. In some embodiments, electrode body 302 may be made from one or more of aluminum (Al) or copper (Cu). In some embodiments, coating a readily oxidized metal like Al or Cu with another metal may protect the underlying electrode from corrosion. In some embodiments, an emissive electrode body 302 including Al and/or Cu may be coated with one or more of Pd, Pt, Co, Ir, Ni, or ITO.

[0034] In the embodiment shown in FIG. 3, chemically active adlayer 304, with a thickness greatly exaggerated for clarity, forms a contiguous layer or coating over a portion of electrode body 302. In some embodiments, adlayer 304 may be present on electrode body 302 in an amount sufficient to be present on a majority of the portion of electrode body 302 active for generating a plasma (e.g., on a majority of an electrode exposed to a plasma generation region) throughout its service life. In some embodiments, adiayer 304 may be present on electrode body 302 in a thickness range of between about 5 nanometers (nm) to 500 nm depending on the type of adiayer material. In some embodiments, adiayer 304 may be present on electrode body 302 in spaced apart islands or domains. In some embodiments, adiayer 304 may alloy with an outer surface of electrode body 302, so that the added material becomes chemically incorporated into electrode body 302.

[0035] Regardless of how adiayer is applied to electrode body 302, adiayer 304 is non-native to electrode body 302. That is, adiayer 304 introduces at least one other metallic or metal element to electrode body 302. Thus, adiayer 304 is different from a native ionic compound of the electrode body formed by inserting a non-metal into an anionic lattice position via oxidation of the electrode metal. Even if adiayer 304 becomes alloyed with electrode body 302 or only decorate portions the surface of electrode body 302 as discontinuous islands, the added metal may potentially aid in forming an electrically conductive and chemically active adiayer 304.

[0036] Adiayer 304 is sufficiently electrically conductive to a permit plasma to be generated nearby. In some embodiments, chemically active adiayer 304 may be formed on an electrically emissive discharge or field generation portion of the electrode.

[0037] In some embodiments, adiayer 304 may be bonded to electrode body 302 with an adhesion layer. FIG. 4 is a schematic fragmentary view of a cross-section of the embodiment of electrode 300 shown in FIG. 3 taken at location 4. As with FIG. 3, the size and appearance of the layers shown in FIG. 4 are exaggerated for clarity. In the embodiment shown in FIG. 4, adhesion layer 402 is sandwiched between adiayer 304 and electrode body 302 to mechanically and/or chemically bond the adiayer and body together. Using an adhesion layer 402 may prevent delamination of adiayer 304 from electrode body 302 in response to thermal or other stresses during operation. Non-limiting examples of adhesion layer 402 materials may include one or more of titanium (Ti), titanium nitride (TiN), tungsten (W), or chromium (Cr), or in some embodiments consisting essentially of Ti, TiN, W, or Cr.

[0038] Embodiments of method 100 and electrode 300 may be employed in embodiments of thin film deposition systems. For example, FIG. 5 is a schematic cross-sectional view of an embodiment of a roll-to-roll deposition system 500 used to perform tests, the results of which are discussed with reference to FIGS. 6, 7, and 8. System 500 is consistent with the systems described in US Pat. No. 8,202,366 B2 of Dickey and Barrow and especially with the system described in FIG. 5 of US Pat. Appl. Pub. 2010/0143710 A1 of Dickey and Barrow.

[0039] System 500 includes a reaction space or vessel 502 that includes a pair of precursor zones 504 supplied with a first precursor (Pi) and, in between those zones 504, a centrally positioned precursor zone 506 supplied with a second precursor (P ₂ ). With reference to FIG. 5 herein, a "loop-mode" configuration wraps a flexible substrate 508, such as a web of plastic film or metal foil, into an endless band (loop), which includes a single path that performs up to two ALD cycles on each revolution as the substrate moves, starting from the leftmost position, from the central precursor zone 506, up into the first precursor zone 504, back down into the central precursor zone 506, down into the second precursor zone 504, and then up to finish back in the central precursor zone 506. As the substrate web 508 travels between zones 504 and 506 it passes through slit valves 512, which are just slots in divider plates 514 that separate the different zones. In this configuration the substrate web 508 can be passed repeatedly through the precursor and isolation zones in a closed loop. (This system is referred to herein as a "roll-to-roH" deposition system, even though the loop substrate configuration used for experimental purposes does not involve transporting the substrate from a feed roll to an uptake roll). In the loop configuration illustrated in FIG. 5, a full traverse of the loop path results in two ALD deposition cycles when two plasma generators 516, 534 are employed in central precursor zone 506. The substrate band is circulated along this loop path X number of times to attain 2X ALD cycles.

[0040] System 500 was modified to perform only a single ALD cycle on each revolution of the substrate, as only a single plasma generator was employed and Pi was only supplied to one of precursor zones 504. Plasma generator 516 in FIG. 5 includes a cathode 518 coated by evaporation with different chemically active materials that are described below. Anode 520 (electrically coupled with vessel 502) is shown grounded to earth at 524, while cathode 518 is electrically isolated from vessel 502 by a glass insulator. Cathode 518 is powered by a plasma controller 532 which controls the electrical potential of cathode 518. In turn, a plasma region is generated within a vicinity of substrate 508 and plasma generator 516. While only a single plasma generator was used in the experiments described herein, system 500 includes an optional second plasma generator 526 which may be used to generate a second plasma region within a vicinity of substrate 508 and plasma generator 526. [0041] The experimental test results described in more detail below with reference to FIGS. 6, 7, and 8 were collected under conditions at which ΤΊΟ2 films (refractive index approximately 2.3) were deposited on flexible substrates (e.g., Melinex ^® ST-504 PET substrates sold by DuPont Teijin Films). The substrate was pretreated in-situ with an oxygen plasma (1 .2 Torr O2, 400 mA plasma current, and 1 lap of travel) prior to deposition. The reactor temperature was about 60 degrees Celsius. The first precursor (P^ was vapor supplied from a titanium tetrachloride (TiCU) source at room temperature (i.e., about 21 degrees Celsius) and supplied to the lower precursor zone 504 through a 1 mm orifice. The second precursor (P ₂ ) was a mixture of carbon dioxide (CO2) and nitrogen (N ₂ ) gases supplied at a total supply pressure of about 1 .4 Torr. The partial pressure of CO ₂ was about 0.3 Torr with the balance being nitrogen. Plasma controller 532 includes a DC power supply (specifically a model MDX 500 power supply manufactured by Advanced Energy Industries, Inc. of Fort Collins, Colorado) operated to maintain a constant current of about 140 mA during film deposition.

[0042] The cathode was an aluminum (Al) square that had a surface area of approximately 20 square inches (in ² ) (about 129 square centimeters (cm ² ) onto which one of the following adlayer materials was coated prior to testing: Pd, Pt, Co, Ir, Ni, and ITO. The cathodes were prepared by first evaporating approximately 20 nanometers (nm) of a Ti adhesion layer onto the Al electrode body. Second, approximately 200 nm of the desired chemically active adlayer material was evaporated onto the adhesion layer. Third, the cathode was installed in the reactor. The aluminum anode was uncoated in these experiments.

[0043] FIGS. 6, 7, and 8 illustrate experimental data collected using the modified version of system 500 under the conditions described herein. FIG. 6 is a graphical relationship 600 between the experimental duration of the plasma exposure time in milliseconds (ms), which is directly a function of the band travel speed, and the experimentally determined deposition rate (in Angstroms/complete revolution of the band) of T1O2 films formed according to embodiments of the present disclosure. Relationship 600 illustrates the deposition rate data for each of the different electrode surface materials and for an uncoated control electrode as a collection of marked curves shown in FIG. 6. As the travel speed decreases (shown as an increasing duration of the CO2/N2 plasma exposure time) the deposition rate plateaus, regardless of the kind of electrode employed. This behavior is believed to be the result of increasing chemisorption and decreasing availability of open surface adsorption sites for the first precursor, indicating the onset of saturation in ALD growth at about 1000 ms. As shown in relationship 600, the performance of the adlayer materials appears to vary with the duration of plasma exposure. For example, Pt and Pd exhibit changes in relative activity before and after the transition to ALD. This may be indicative of the kind of species formed by the different materials. Accelerating the band speed, so that the substrate travels out of the plasma but is still exposed to radicals generated therein, may promote film deposition. In contrast, a slower band speed extends the exposure of the chemisorbed species to the plasma. This may extend the exposure to ions within the plasma, which may degrade or scavenge chemisorbed precursor from the substrate surface, potentially appearing to lower the deposition rate and shift the transition region to ALD to slower speeds, as shown for Pd.

[0044] FIG. 6 also shows experimental data demonstrating that each of the adlayer-including cathodes outperforms, in terms of film deposition rate, the control uncoated cathode regardless of the extent to which the substrate is covered with chemisorbed first precursor. Each coated cathode leads to deposition of T1O2 at a faster rate relative to the rate at which T1O2 is deposited using an uncoated control cathode.

[0045] It is believed that the resulting films may have superior barrier performance properties (e.g., as water vapor or optical barriers, among other barrier properties). FIG. 7 is a graphical relationship 700 between the thickness of experimental T1O2 films deposited as described herein and the water vapor transmission rate (WVTR) for those films. The WVTR of the coated substrates was measured using a water vapor transmission analyzer (WVTA) model 7001 manufactured by Illinois Instruments, Inc. of Johnsburg, III., USA. The TiO2-coated PET substrates were clamped in a diffusion chamber of the model 7001 WVTA, which measures WVTR by subjecting the coated substrates to test and carrier gases that attempt to permeate the sample. The 7001 WVTA conforms to ISO 15105-2 and uses a modified ASTM standard that conforms to ISO 15106-3. WVTA measurements were conducted at 37.8 degrees Celsius with a relative humidity of 90 %. The 7001 WVTA has a lower sensitivity limit of 0.003 g/m ² /day, which is illustrated as a dotted line on FIG. 7. Relationship 700 compares films prepared using an example Pd adlayer-including cathode (WVTR data illustrated with square- shaped markers in FIG. 7) and an uncoated cathode (WVTR data illustrated with diamond-shaped markers in FIG. 7), and demonstrates that, for a given thickness, the films exhibit comparable performance. Because the adlayer-including cathodes deposit film faster than the uncoated cathodes, it is expected that substrates may be coated with higher throughput without compromising water vapor barrier performance.

[0046] It is contemplated that many of the chemically active materials described herein as being added to electrode surfaces will have different work functions or secondary electron emission characteristics from the underlying electrode body and/or adhesion layer materials. These differences may alter the plasma impedance. For example, higher impedance may allow a higher operating voltage for the plasma, potentially preventing plasma breakdown at that voltage. Lower impedance may allow higher plasma current for a given power. In some embodiments, an Al or an Al-containing conductive adlayer applied to a Co, Ni, Pd, Pt, or other electrode emissive body material may lower the plasma impedance of the electrode.

[0047] In the tests described above, the plasma current was maintained at about 140 mA during film deposition. Average plasma operating power readings were recorded during each experimental test. FIG. 8 is a graphical relationship 800 of the average plasma power (in watts) as a function of the adlayer material identity. As a baseline, the uncoated control cathode typically operated at about 66 W. Each of the adiayers tested increased the plasma power consumption by between about 10% and 16%, indicating an increase in plasma impedance.

[0048] FIG. 9 is a schematic, fragmentary cross-sectional view of an embodiment of a thin film deposition system 900 for coating, in a process space or vessel 902, a flexible substrate 904 (shown in profile in FIG. 9) with a flexible layer of material via ALD. In the embodiment shown in FIG. 9, thin film deposition system 900 includes first, second, and third precursor zones 906, 908, and 910, respectively, that are supplied with two different precursor gases Pi and P ₂ (shown as Precursor 1 Zone and

Precursor 2 Zone in FIG. 9). Because the precursor gases may be sensitive or chemically reactive with one another in some embodiments, isolation zones (not shown) may separate adjacent precursor zones from one another as described in US Pat. No. 8,202,366 B2 of Dickey and Barrow. In the embodiment shown in FIG. 9, the process environment is defined and bounded by process vessel 902, which separates the process environment from the external ambient environment. Vessel 902 may comprise a pressure vessel and/or a vacuum vessel. In the embodiment shown in FIG. 9, vessel 902 fluidly communicates with a pump/suction source (not shown) so that the process environment may be adjustably controlled to and maintained at a subatmospheric pressure.

[0049] The process environment may also be subdivided into two or more subchambers. In the embodiment shown in FIG. 9, precursor zones 906, 908, and 910 are separated from one another by zone dividers 912. In some other embodiments, additional zone dividers may separate the precursor zones and any isolation zones from one or more antechambers, such as a pre-deposition subchamber and/or a post-deposition subchamber (not shown). A series of passageways 913 through zone dividers 912 are spaced apart along a general direction of travel of substrate 904, and are arranged and configured so that substrate 904 traverses a pathway passing back and forth between first, second, and third precursor zones 906, 908, and 910, multiple times, in some embodiments passing through an isolation zone in between each passage through a particular precursor zone. Turning guides 914 direct substrate 904 back in an opposite direction so that substrate 904 is transported back and forth between the precursor zones (e.g., Precursor 1 Zone → Precursor 2 Zone → Precursor 1 Zone → Precursor 2 Zone→ ...). While the embodiment shown in FIG. 9 includes three precursor zones, it will be appreciated that some embodiments may include four or more precursor zones, and that some other embodiments may have fewer than three precursor zones.

[0050] In the embodiment shown in FIG. 9, plasma regions 916 are generated by a plurality of plasma generators 917 positioned within third precursor zone 910, though it will be appreciated that plasma generators 917 may be located in any of the precursor zones 906, 908, and 910. During use, plasma generators 917 generate plasma regions 916 primarily in an area where a field is applied by plasma generator 917 when a plasma sensitive process gas is supplied within third precursor zone 910. In some embodiments, plasma region 916 may extend toward the pathway so that the plasma gas contacts one or both surfaces of flexible substrate 904, permitting plasma generation to occur proximal to the exposed surface of substrate 904. It will be appreciated that the shape and size of plasma region 916 may be affected by plasma generator power conditions, total and partial pressure conditions near plasma generators 917, and/or sizes and locations of plasma generators 917. Thus, in some embodiments, two or more plasma generators 917 may contribute to and be enveloped by a common plasma region 916, even though the plasma regions 916 depicted in FIG. 9 appear discrete and separate.

[0051] Each of the plasma generators 917 shown in FIG. 9 includes a first electrode 918 having a non-native electrically conductive electrode surface (e.g., including an embodiment of an adlayer like those described herein) that exhibits property characteristics that cause the electrode surface to be substantially conserved and chemically active with at least one of the gases present within the plasma generation region during plasma generation, and a second, complementary electrode 919. First and second electrodes 918 and 919 are electrically insulated from one another by a space separating the electrodes at plasma region 916 and by suitable electrical insulators elsewhere (not shown). The plurality of plasma generators 917 shown in FIG. 9 are powered by a common power supply 920, though in some embodiments groups and/or individual plasma generators 917 may be powered by separate power supplies so that corresponding plasma regions 916 may be separately controlled (e.g., to provide zone control of plasma regions 916). In some embodiments, power supply 920 may include a DC power supply for generating a glow discharge plasma. In some other embodiments power supply 920 may include a radio frequency (RF) generator and a capacitance matching network for generating a capacitively coupled plasma.

[0052] In the embodiment shown in FIG. 9, all first electrodes 918 are in electrical communication with one another and with a first pole of power supply 920 via a first electrode bus 922, so that all first electrodes 918 may have the same electrical potential, within an acceptable tolerance. Likewise, all second electrodes 919 are in electrical communication with one another and with a different pole of power supply 920 via a second electrode bus 924 in the embodiment shown in FIG. 9. For example, in embodiments where power supply 920 includes a DC power supply, connecting first electrodes 918 with a negative pole of power supply 920 and second electrodes 919 with a positive pole of power supply 920 will cause first electrodes 918 to act as cathodes and second electrodes 919 to act as anodes. [0053] FIG. 10 is a schematic, fragmentary cross-sectional illustration of another embodiment of a thin film deposition system 1000. In FIG. 10, the reference numbers having the same final two digits as FIG. 9, as preceded by the number "10," identify analogous structures. For example, reference number 1002 in FIG. 10 identifies a process space or vessel analogous to process space or vessel 902 in FIG. 9. In the embodiment shown in FIG. 10, a plasma region 1016 is generated from a second precursor using single remote plasma generator 1017. Radicals formed within plasma 1016 are distributed via a radical delivery conduit 1032 to outlets 1034 within precursor zone 1010, where they interact with chemisorbed precursor Pi ^* on substrate 1004.

[0054] It will be obvious to those having skill in the art that many changes may be made to the details of the above-described embodiments without departing from the underlying principles of the invention. The scope of the present invention should, therefore, be determined only by the following claims.