DE102008007588A1 - Barrier layer creating process for microstructured component involves preparing component in plasma reactor, plasma treatment, and supplying precursor and carrier gas - Google Patents

Abstract

The barrier layer creating process consists of preparing the component in a plasma reactor with steam doser and gas supply; plasma treating the component in the region to be coated with a plasma from inert carrier gas to activate it, then supplying a steam or gas precursor to the carrier gas by steam dosing, gas exchanging back to inert carrier gas, and repeating the process several times.

Description

The The invention relates to a method for producing a barrier layer on a microstructured component. Furthermore, the concerns Invention obtained by the method devices with a barrier layer.

State of the art and technological background

The Deposition of inorganic and organic barrier layers for Protection of components for microelectronics, in particular based on electronic and optoelectronic devices of polymers, was previously mainly for planar surfaces quite successful. In particular, components that are opposite the oxygen and water vapor content of the air are sensitive or which are used for exhaust control in internal combustion engines, protected by appropriate barrier layers.

In the protection of such devices, plasma assisted deposition (PECVD) plays an important role. Thus, the layer deposition on the basis of the starting gases silane and ammonia (for Si _x N _y ) or silane and O ₂ or N ₂ O (for SiO _x ) from an inductively coupled high-density plasma led to high-quality layers even at low process temperatures ( R. Wolf et al., Surface and Coatings Technology 2001, 142-144, 786 ).

Single layers, however, always contain defects which propagate as the layer grows further. The permeation or barrier properties of such layers can be improved to a limited extent by a greater layer thickness. Further improvements are possible through the deposition of multiple layers (see for example WO 2006 007313 A and US 2007 0049155 A1 ), in which the defect growth is interrupted by a generally organic polymeric intermediate layer and does not occur in the same place in the next barrier layer. However, it is difficult in a reactor chamber successively to deposit high-quality barrier layers (eg Si _x N _y ) alternating with typical polymeric decoupling layers based on polymers which differ greatly from the barrier layers in their elemental composition (eg CF _x polymers) , The reactor conditions and the plasma compositions change during polymer deposition such that the properties of the barrier layers deteriorate. Consequently, a second plasma reactor would have to be used for the deposition of such decoupling layers. This represents a significant increase in the cost, which still increases when the change between the reactors must be made in the absence of air, because z. B. the components to be protected are sensitive to oxygen.

A particular problem is the generation of barrier layers at structural edges (described, for example, in US Pat US 6,664,137 ). It is not enough to apply the barrier layers only on the front sides of the sensitive components (microstructured components), because then the erosion takes place by diffusion from the sides. However, commonly used plasma coating processes produce layers whose thickness and properties differ on horizontally and vertically oriented surfaces. In extreme cases, no layer is deposited on the vertical surfaces at all, especially when two vertical surfaces are located at a small distance. Here, the splitting of the process plays a crucial role.

Atomic Layer Deposition (ALD) is a process that satisfies such claims (described as ALE in) DE 195 81 483 T5 ; M. Leskela, M. Ritala: Thin Solid Films 409 (2002) 138 ; KE Elers, et al., Chemical Vapor Deposition 12 (2006) 13 ; Kunz et al., US 2005 0147919 ). The ALD allows, within certain limits, deposition in structures of high aspect ratio devices; it is possible to realize largely constant layer thicknesses and largely constant properties of the layers (for example electrical properties). This is achieved by the self-limitation of two alternating process steps. In a first process step, organometallic precursors are used, followed by a second process step using oxygen- or nitrogen-containing compounds. For example, Al ₂ O ₃ barrier layers can be prepared ( E. Langereis, et al. Applied Physics Letters 89 (2006) ). The process steps can also be performed plasma-assisted (plasma-assisted ALD: PALD or PEALD) ( PF Carcia et al .; Applied Physics Letters 89 (2006) ). In the previously known ALD or PALD processes arise inorganic crosslinked barrier layers.

ALD or PALD processes are therefore suitable for the protection of microstructured components with high cracking properties for barrier layers. The properties and thickness of the barrier layer can be influenced by a suitable choice of the two precursors and by the process temperature. However, the ALD or PALD method has the disadvantage that it can not produce organically crosslinked layers and thus also no layer sequences of organic and inorganic crosslinked layers in the quality required for a sufficient barrier effect in a reactor. Furthermore, the known ALD or PALD method have the disadvantage that the layer properties largely by the chemical reac tion of the two precursors are determined and can be changed only to a limited extent by the process parameters. Therefore, they can not be used as an effective method for the production of barrier layers.

invention solution

• (i) Bereitstellen des Bauelements in einem Plasmareaktor, der zusätzlich eine Dampfdosiervorrichtung und eine für schnelle Gaswechsel ausgestattete Gaszufuhr besitzt;
• (ii) Plasmabehandlung des Bauelements im zu beschichtenden Bereich mit einem aus einem inerten Trägergas erzeugten Plasma und dabei Aktivierung des zu beschichtenden Bereichs;
• (iii) Nach Beendigung der Plasmabehandlung gemäß Schritt (ii), Zusatz eines dampf- oder gasförmigen organischen Precursors zum Trägergas mittels der Dampfdosiervorrichtung, wobei sich der Precursor im aktivierten Bereich unter Ausbildung einer organischen Schicht abscheidet;
• (iv) Nach Beendigung der Zuführung des Precursors gemäß Schritt (iii), Gaswechsel hin zum inerten Trägergas ohne Zusatz des Precursors; und
• (v) n-faches Wiederholen der Schritte (ii) bis (iv) zur Erzeugung der Barriereschicht, wobei n = 2 bis 1.000 ist.

With the aid of the method according to the invention, one or more of the stated disadvantages of the prior art can be overcome. The method according to the invention for producing a barrier layer on a microstructured component is characterized in that the method comprises the following steps:

• (i) providing the device in a plasma reactor additionally having a vapor metering device and a fast gas exchange gas delivery;
• (ii) plasma treatment of the device in the region to be coated with a plasma generated from an inert carrier gas and thereby activating the region to be coated;
• (iii) After completion of the plasma treatment according to step (ii), adding a vapor or gaseous organic precursor to the carrier gas by means of the Dampfdosiervorrichtung, wherein the precursor deposited in the activated region to form an organic layer;
• (iv) after completion of the supply of the precursor according to step (iii), gas exchange to the inert carrier gas without addition of the precursor; and
• (v) repeating steps (ii) to (iv) n times to form the barrier layer, where n = 2 to 1,000.

Of the Among other things, the invention is based on the finding that, too in tight structures of microstructured devices a barrier layer deposited with largely constant thickness and constant properties can be if an alternating sequence of plasma activation and deposition of an organic precursor from the gas phase without simultaneous plasma treatment takes place. Requirement for the coating in tight structures is the steps (ii) and (iii) are self-limiting. This can reduce the duration of this Steps are chosen so that all face the activated structure to saturation or a monomolecular layer deposited to saturation is. In other words, a duration of steps (ii) and (iii) is preferably chosen so that for all protective surfaces of the device saturation is reached. The sequence comprises at least two cycles, i. H. The resulting barrier layer consists of at least two consecutive deposited partial layers together.

It It has now been found that a self-limiting process also applies to organically crosslinked barrier layers is possible. For the desired splitting is a prerequisite that the process is self-limiting. self-limitation means that the layer growth in individual steps of the Process is limited, d. H. growing one in step (iii) produced partial layer only up to a certain maximum layer thickness regardless of whether the precursor is still provided becomes. Self-limitation in the sense of the invention means about it In addition, in step (ii) activation via a minimum period of time not to a further increase the layer thickness leads. In other words, those found here So method uses an activation phase with an inert, per se non-film-forming carrier gas plasma and alternating to the addition of constituents of a layer-forming precursor in a second, plasma-free phase of deposition.

in the In contrast to the conventional ALD method, the Accumulation of the precursor on a plasma-activated substrate. Another, particularly advantageous difference is that of 2nd cycle on (n ≥ 2) in step (ii) of the plasma treatment simultaneously with the activation of a network of the previously attached Precursor molecules takes place. In dependence of the plasma parameters are the precursor molecules at a large structural preservation modified and the previously deposited layers with each other networked. That offers an additional possibility for controlling the layer properties.

From the already known plasma plug (see for example Siow, KS et al .; Plasma Proc. Polym. (2006) p. 392 , or US 20030129322 ), the method found differs in that the plasma process not only acts once on the substrate, but each newly applied layer is exposed to the plasma process, this not only the connection of the next layer, but also showed a plasmachemic remodeling of the applied layers , preferably the last applied, causes.

The inventive method is carried out in a known plasma reactor, which is additionally equipped with a Dampfdosiervorrichtung and equipped for fast gas exchange gas supply. The regulations of pressure and plasma power should correspond to the fast gas change. The control software should contain functions for the simple programming (recipe generation) of periodical processes. Preferably, the plasma reactor has an inductively coupled plasma source for generating a high density plasma. The inductively coupled plasma source (ICP) for generating a high density plasma can be mounted on a stainless steel UHV plasma reactor be done. The plasma source may use a planar triple spiral antenna to couple the RF power into the plasma through an alumina dielectric disk (see for example DE 195 48 657 ). For plasma excitation, an RF generator at a frequency of 13.56 MHz and a maximum power of 1000 W can be used. With such an ICP source, a planar plasma is generated, which is characterized by good spatial homogeneity. In the reactor chamber there is a sample table whose distance to the plasma can be varied. The components to be coated are therefore treatable both in the decaying and in the active plasma. Thus, the effect of the plasma on the location-dependent plasma composition is controllable. An RF power from a second generator may be supplied to the sample stage to control the energy of the ions striking the sample from the plasma.

in the Step (ii) of the process, a plasma treatment of the device takes place in the area to be coated with one of an inert carrier gas generated plasma. This results in an activation of the to be coated Area for the subsequent reaction with the precursor. The plasma treatment in step (ii) is preferably carried out via a duration in the range of 10 to 40 s. The carrier gas is in step (ii), preferably with a flow in the range of 20 to Fed to 100 sccm. An injected plasma power lies preferably in the range of 50 to 200 W. A gas pressure of the carrier gas is preferably in the range of 2 to 10 Pa.

For the inventive method particularly suitable Components exist for. B. of polymeric electronic semiconductors.

The Carrier gas is preferably argon or a mixture based of argon with nitrogen and / or hydrogen.

in the Step (iii) of the process, ie after completion of the plasma treatment from step (ii) becomes a vapor or gaseous organic Precursor added to the carrier gas by means of the Dampfdosiervorrichtung. The precursor separates in the activated area under training an organic (part) layer.

A Dosage of the precursors can be via an annular Gas inlet system, taking place from a central feed point Stretch approximately the same flow resistance lead to all gas outlet openings. So will Ensures that large outlets allow a quick gas exchange, but the Gas flows from the individual openings do not like when using uneven flow resistances differ and lead to a non-uniform deposition. For the precise dosage of liquid precursors a known liquid regulator (mass flow Controler (MFC)) are used.

Of the Precursor is preferably a nitrogen-containing or oxygen-containing organosilicon compound or a hydrocarbon compound. The nitrogen-containing organosilicon compound is preferred selected from the group consisting of bis (dimethylamino) dimethylsilane (BDMADMS), hexamethyldisilazane (HMDSN), tetramethyldisilazane (TMDSN), Dimethylaminotrimethylsilane (DMATMS) Tris (dimethylamino) methylsilane (TDMAMS) and dimethylbis (2,2-dimethylhydrazyl) silane (DMBDMHS). The oxygen-containing organosilicon compound is preferred selected from the group consisting of hexamethyldisiloxane (HMDSO), Tetraethoxysilanes (TEOS) and tetramethyldisiloxane (TMDSO). The hydrocarbon compound is preferably selected from the group consisting of Methane, ethane and ethyne.

The Feeding of the precursor in step (iii) is preferably carried out via a duration in the range of 1 to 40 s. The offset with the precursor Carrier gas is preferably in step (iii) with a Carrier gas flow in the range of 20 to 100 sccm and a Precursorgasfluss supplied in the range of 0.3 to 1.0 sccm. A gas pressure of the carrier gas mixed with the precursor is preferably in the range of 2 to 10 Pa.

in the Step (iv) of the method, ie after completion of the feed of the precursor in step (iii), a gas exchange takes place back to the inert carrier gas without addition of the precursor. The step (iv) can be divided into two sub-steps, namely a first Sub-step of the evacuation of the reactor chamber under application of a Vacuum and a second sub-step in which the reactor chamber with a purge gas (usually the carrier gas) is rinsed.

The Steps (ii) to (iv) become n times to form the barrier layer repeated (step (v)). A number n of cycles becomes n = 2 to 1,000, in particular n = 20 to 200 limited.

• (vi) Nach Beendigung des Schritts (v), Zusatz des Precursors zum Trägergas mittels der Dampfdosiervorrichtung und Plasmabehandlung des Bauelements im zu beschichtenden Bereich, wobei sich der Precursor im zu beschichtenden Bereich unter Ausbildung einer organisch/anorganischen Mischschicht abscheidet;
• (vii) m-faches Wiederholen der Schritte (ii) bis (vi) zur Erzeugung der Barriereschicht, wobei m = 2 bis 10 ist.

According to a further preferred variant of the method, the method includes the following additional steps:

• (vi) After completion of step (v), adding the precursor to the carrier gas by means of the vapor metering device and plasma treatment of the device in the region to be coated, wherein the precursor deposits in the region to be coated to form an organic / inorganic mixed layer;
• (vii) repeating steps (ii) to (vi) m times to create the barrier layer, where m = 2 to 10 is.

To this method variant thus follows the deposition of a first Multiple layer according to steps (ii) to (v) a step (vi), wherein the same precursor but with the plasma switched on is deposited. The conditions in step (vi) therefore correspond the conditions in a conventional PECVD method. The result is a dense, partially inorganic layer. To this Layer is a second multiple layer according to the Plotted steps (ii) to (v). Finally, a further layer under the conditions of step (vi), ie at switched on plasma, applied. A number m of cycles off Depositing the multilayers according to the steps (ii) to (v) and plasma deposition according to the step (vi) is limited to m = 2 to 10, in particular m = 2 to 5. It the desired barrier layer is created.

One Another aspect of the invention relates to components that after a the method of the invention described above have generated barrier layer. Such barrier layers have not been described in the prior art.

The Method will be described below in embodiments and explained in detail with reference to the accompanying drawings. Show it:

1 An illustration of a time sequence of the method according to the invention.

2 A growth rate of the barrier layer as a function of the precursor concentration.

3 A growth rate of the barrier layer as a function of plasma activation time.

Detailed description the invention

In The following investigations were inductively coupled Plasma (ICP) source used to generate a high density plasma, which was mounted on a stainless steel UHV plasma reactor. about a planar triple spiral antenna was the RF power through a Dielectric alumina disk coupled into the plasma. For the Plasma excitation became an RF generator at a frequency of 13.56 MHz and a maximum power of 1000W. With this ICP source was generated a plane plasma, which is characterized by good spatial homogeneity distinguished.

The Dosing of liquid precursors with for gas phase processes sufficient vapor pressure was via an annular Gas inlet system, which approximates distances from a central feed point same flow resistance to all gas outlet openings led. This ensured that big Outlets allow quick gas exchange, but the gas flows from the individual openings not, as with the use of unequal flow resistance, differ from each other and lead to non-uniform deposition. For the exact dosage of the precursor vapor was a Liquid Mass Flow Controler (MFC) is used.

in The embodiments were polyetheretherketone wafers (PEEK) by plasma-activated deposition using bis (dimethylamino) dimethylsilane (BDMADMS) successively coated with monolayers. Under the influence The cross-linking plasma effect results in organic polymer layers.

For the process became the low-pressure plasma reactor described above used. The PEEK wafer was at a distance of 18 cm from the coupling window the inductive plasma source and thus outside the active Plasmas arranged in a vacuum vessel. In the reactor Argon was admitted at a flow of 100 sccm (sccm = standard cubic centimeter per minute). The gas pressure became 5 Pa and the injected plasma power set to 50 W and regulated. In the following sub-step of the process (step (ii)) became the wafer with argon plasma treated for 15 s. After the plasma shutdown, 0.3 sccm BDMADMS added to the gas phase for 10 s. it then the entire gas flow was turned off and the residual gas is pumped off for a period of 30 s. After one Purge with argon (30 s, 100 sccm, 5 Pa) was the Cycle restarted with the power on of the plasma. The plasma crosslinked and activated the attached precursor layer. By the number of cycles passed became the thickness of the total layer set. With 60 cycles, 24 nm thickness was achieved.

The chronological sequence of the method according to the invention will be described below 1 explained in more detail. The process begins with the generation of an inert gas plasma. During this first phase, no layer-forming vapor (no precursor) flows through the reactor. In this phase, the surface is activated. In the example, this first phase lasts 15 s (➀ in 1 ). After switching off the plasma, a layer-forming precursor vapor is added to the carrier gas in a second phase for 15 seconds (➁ in 1 ). This reacts with the previously activated surface. In the scheme of 1 In the second phase, the states plasma: <off>, precursor flow: <on>, outlet valve: <up> and Ar (argon): <on> are realized. In a third phase, a purging process (vacuum) runs with the states plasma: <off>, precursor flow: <off>, outlet valve: <to> and Ar: <off>. The cycle concludes with a fourth phase, a purge process in which the states plasma: <off>, precursor flow: <off>, exhaust valve: <to> and Ar: <off> are realized. Then the cycle begins again with the first phase.

The Self-limitation of the involved processes of activation and attachment was proven in complementary experiments.

For this purpose, initially the conditions described above for the times of admission of BDMADMS were varied between 1 s and 40 s. Each new PEEK wafer was subjected to 60 cycles of precursor flow phases of different lengths (0 s, 1 s, 5 s, 10 s, 20 s and 40 s, respectively). The growth rates per cycle were determined from the layer thicknesses achieved. 2 shows a growth rate (nm / cycle) as a function of BDMADMS dose. How out 2 As can be seen, a growth rate of approximately 0.4 nm / cycle is achieved after 1 s of precursor flow. This rate does not change with a longer precursor flow time - which demonstrates the self-limitation of the addition of BDMADMS. Full coverage of activated surfaces with similar coverage is to be expected.

In analogously, in a second series of experiments, the duration of the Plasma action between 0 s and 40 s with constant precursor flow time varies from 10 s.

The result is in 3 (rate of growth (nm / cycle) versus plasma activation time) and shows that the plasma activation process is self-limiting as well.

Of the Mechanism of self-limitation itself is still poorly understood. The activation by metastable Ar atoms is conceivable. It can assume that the annealing ends when the energy the activated surface is no longer sufficient to bonds break up in the precursor molecule. The observed in a short time Self-limitation of the activation process is even more surprising and not yet understood, because the activating plasma causes at the same time a chemical change of the layers, presumably without the layers completely or partially removed can be. The layers become molecularly compartmentalized applied. Therefore, the process may differ from the PALD also as plasma-activated deposition of monomolecular monolayers (PAMM) be designated.

In another embodiment, the periodic plasma-activated deposition of molecular monolayers was combined with a pulsed PECVD method. First, an organic polymer layer was deposited in 60 cycles according to the procedure described above. Subsequently, a layer was deposited on this polymeric multilayer without interruption of the vacuum in the same plasma reactor by means of PECVD method of BDMADMS. To this was added a gas mixture of argon (22 sccm), NH ₃ (3 sccm) and BDMADMS (1 sccm). The gas pressure was adjusted to 3.5 Pa and the injected plasma power to 550 W and controlled. The plasma was pulsed: turned on for 1 second and turned off for 4 seconds. The effective treatment time, in which the plasma was switched on, was a total of 3 min. This plasma process resulted in a dense, partially inorganic layer of 185 nm thickness. The overall process described is repeated once. The result is a layer stack of four layers: alternating two soft polymeric multilayers and two hard layers deposited by PECVD. This barrier layer has a reduced gas permeation compared to a single PECVD layer of 2 × 185 nm thickness.

For the production of barrier layers can therefore be the PAMM method with a conventional low-temperature plasma deposition (PECVD) be combined. For example, from an organosilicon Precursor a first layer by conventional plasma-induced Deposition are formed. A second layer is formed, by repeating the same precursor only for a certain amount Time (typically for 1-100 s) in a similar manner or otherwise composed carrier gas plasma (pulsed plasma). A third layer is then using the described PAMM method educated. This forms layers with different properties: The first layer has essentially good barrier properties. The second and third layers contain larger ones Concentrations of cross-linked CH bonds and therefore can serve to block the defects of the first layer. The third Layer differs from the second layer by a different one Networking and greater splitting and can therefore be good in lower layer thicknesses and for coating complex, nonplanar, microstructured substrates use Find.

The Method has the advantage that the precursor can not be changed got to. Therefore, the change in reactor conditions is between the deposition phases low and it can successively layers the required quality are deposited.

For the alternating process management, a fast gas exchange should be guaranteed. This is achieved by minimizing the volume of the reactor chamber and by using precursor vapor buffer volumes just prior to the reactor inlet valve, but with the uniformity of the Schich must be secured by the system for the supply of precursors.

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - WO 2006007313 A [0004]
• US 20070049155 A1 [0004]
• - US 6664137 [0005]
• - DE 19581483 T5 [0006]
• US 20050147919 [0006]
• - US 20030129322 [0012]
• DE 19548657 [0013]

Cited non-patent literature

• R. Wolf et al., Surface and Coatings Technology 2001, 142-144, 786 [0003]
• - M. Leskela, M. Ritala: Thin Solid Films 409 (2002) 138 [0006]
• KE Elers, et al., Chemical Vapor Deposition 12 (2006) 13 [0006]
• E. Langereis, et al. Applied Physics Letters 89 (2006) [0006]
• - PF Carcia et al .; Applied Physics Letters 89 (2006) [0006]
• Siow, KS et al .; Plasma Proc. Polym. (2006) p. 392 [0012]

Claims (14)

A method of forming a barrier layer on a microstructured device, characterized in that the method comprises the steps of: (i) providing the device in a plasma reactor which additionally has a vapor deposition device and a fast gas exchange gas supply; (ii) plasma treatment of the device in the region to be coated with a plasma generated from an inert carrier gas and thereby activating the region to be coated; (iii) After completion of the plasma treatment according to step (ii), adding a vapor or gaseous organic precursor to the carrier gas by means of the Dampfdosiervorrichtung, wherein the precursor deposited in the activated region to form an organic layer; (iv) after completion of the supply of the precursor according to step (iii), gas exchange to the inert carrier gas without addition of the precursor; and (v) repeating steps (ii) through (iv) n times to form the barrier layer, where n = 2 to 1,000. The method of claim 1, wherein the carrier gas Argon or a mixture based on argon with nitrogen and / or Is hydrogen. The method of claim 1 or 2, wherein in the the precursor is a nitrogen-containing organosilicon compound is. The method of claim 4, wherein the nitrogen-containing organosilicon compound is selected from the group consisting of bis (dimethylamino) dimethylsilane (BDMADMS), hexamethyldisilazane (HMDSN), tetramethyldisilazane (TMDSN), dimethylaminotrimethylsilane (DMATMS) tris (dimethylamino) -methylsilane (TDMAMS) and dimethylbis (2,2-dimethylhydrazyl) silane (DMBDMHS). The method of claim 1 or 2, wherein in the the precursor is an oxygen-containing organosilicon compound is. The method of claim 5, wherein the oxygen-containing organosilicon compound is selected from the group consisting of hexamethyldisiloxane (HMDSO), tetraethoxysilanes (TEOS) and tetramethyldisiloxane (TMDSO). The method of claim 1 or 2, wherein in the the precursor is a hydrocarbon compound. Process according to claim 7, wherein the hydrocarbon compound is selected from the group consisting of methane, ethane and ethyne. Method according to one of the preceding claims, in which the plasma treatment in step (ii) via a duration in the range of ... to ... s takes place. Method according to one of the preceding claims, in which the supply of the precursor in the step (iii) over a duration in the range of ... to ... s. Method according to one of the preceding claims, in which the plasma reactor is an inductively coupled plasma source for generating a high density plasma. Method according to one of the preceding claims, in which the method the following additional Steps includes: (vi) after completion of step (v), Addition of the precursor to the carrier gas by means of the steam metering device and plasma treatment of the device in the area to be coated, wherein the precursor in the area to be coated under training an organic / inorganic mixed layer separates; (Vii) repeating steps (ii) to (vi) to produce the m Barrier layer, where m = 2 to 10. Method according to one of the preceding claims, in which a duration of steps (ii) and (iii) is chosen will that for all surfaces to be protected of the device saturation is achieved. Microstructured component with a barrier layer, by the method according to one of claims 1 to 13 was produced or can be produced.