WO2011009681A1

WO2011009681A1 - Plasma enhanced deposition method, semiconductor device, and deposition device

Info

Publication number: WO2011009681A1
Application number: PCT/EP2010/058740
Authority: WO
Inventors: Robert Seguin; Peter Engelhart; Bernd Hintze; Wilhelmus Mathijs Marie Kessels; Gijs Dingemans; Mauritius Cornelius Maria Van De Sanden
Original assignee: Q-Cells Se
Priority date: 2009-07-24
Filing date: 2010-06-21
Publication date: 2011-01-27
Also published as: DE102009026249B4; DE102009026249A1

Abstract

The invention relates to a method for plasma enhanced deposition of a material layer (5), a semiconductor device and a deposition device. During the plasma enhanced deposition of the material layer (5), a first precursor (7) and a second precursor (8) are introduced in vapor form into a reaction chamber (1 ) and participate in a chemical deposition reaction, by which the material layer (5) is deposited onto a surface region (4) of a substrate (3) placed inside the reaction chamber (1 ), wherein the introduction (12) of the first precursor (7) and / or a movement (22) of the substrate (3) is performed such that the concentration of the first precursor (7) in a surrounding space (6) above the surface region (4) is varied in a pulse shape, and in the surrounding space (6) at least partly a chemical gas phase reaction of the first precursor (7) with a reactant takes place, and wherein the second precursor (8) is activated for a chemical reaction with the first precursor (7) by a plasma discharge.

Description

Title:

Plasma enhanced deposition method, semiconductor device, and deposition device

Description:

The invention relates to a plasma enhanced deposition method, a

semiconductor device, and a deposition device.

The efficiency of solar cells can be reduced significantly due to the

recombination of charge carriers at the solar cell surface, which are then not available anymore for electricity generation. Therefore, in order to raise the solar cell efficiency, the recombination at the solar cell surface has to be reduced. Usually, this is achieved by so called surface passivation.

One possibility for surface passivation consists of applying a dielectric passivation onto the solar cell surface. Such a layer may for example be formed out of SiO2, SiNx, A12O3 or SiC. Above all, aluminum oxide (A12O3) has proven itself as a promising candidate for the application in industrially produced solar cells. To its positive properties belongs, besides a good passivation effect due to a very high negative surface charge density, also a high stability, for example against a subsequent temperature treatment in form of a so called firing step, which is necessary for burning in metallic electrodes of the solar cell using screen printing paste for industrial production.

A method for time and cost effective deposition of material layers is the plasma enhanced chemical vapor deposition (PECVD). Therein, precursors are continuously introduced into a reaction chamber, in which a substrate is placed. There, the precursors react with each other under the influence of a plasma discharge in the gas phase and on the surface of the substrate, such that eventually a material layer in form of a thin layer is deposited on the substrate surface. Because this deposition method takes place continuously, and the reaction is supported by the energy of the plasma, very high deposition rates and therefore also high layer growth rates are achievable. However, the material layers produced in this way do not meet the high quality requirements in particular for A12O3 passivation layers, which is why PECVD processes are usually not utilized therefor.

The method, with which high quality material layers have been produced until now, in particular A12O3 passivation layers on solar cell surfaces, is the so called atomic layer deposition (ALD). The ALD process can also be performed plasma enhanced. In this process, the precursors, for example

trimethylaluminum (TMA) and oxygen for the forming of A12O3, are

sequentially introduced into a reaction chamber, in which the substrate, for example a solar cell wafer, is placed. Unlike in the case of the PECVD process, the chemical reaction takes place not already in the gas phase as well, but exclusively on the surface of the substrate. Furthermore, the chemical reaction in the ALD process takes place in two sub-reactions on the substrate surface. Only when such a separation is possible, the ALD process, having a strict separation of the precursors in the gas phase, may work.

During the plasma enhanced ALD, one reactant is activated by the plasma energy, such that the surface reaction also takes place, whereby in each deposition cycle a monolayer of the material layer is made, just as during not plasma enhanced, thermal ALD processes. The material layer deposited in this way is very homogeneous. Furthermore, its layer thickness is almost digitally adjustable by way of the number of cycles, and its passivation properties are very good also at solar cell level, especially when compared to material layers produced with a PECVD process. The ALD process is, however, inherently a very slow deposition process, because the material layer is deposited monolayer after monolayer. The economic feasibility of this technology for industrial production is therefore questionable, despite the very good layer properties. A further deposition method is suggested in WO 2005104634 A2 and utilized in an embodiment described therein for the deposition of an A12O3 layer. It also relates to a plasma enhanced deposition process, which in the embodiment disclosed therein takes place in a self limiting process similar to the ALD process, and can therefore also obtain very slow layer growth rates. Unlike in the ALD process, one of the precursors, namely the precursor consisting of a metal or semiconductor compound, is introduced into the reaction chamber continuously and reacts in the gas phase with a further precursor, which may for example be oxygen, which is activated by the plasma discharge. However, due to the continuous introduction of the expensive metal precursor, this process is very cost-intensive.

It is therefore an objective of the invention to provide for a deposition method, a semiconductor device and a deposition device, for realizing a high quality material layer in a time and cost effective manner.

The object is solved according to the invention by a plasma enhanced deposition method with the features of claim 1 , by a semiconductor device with the features of claim 17, and by a deposition device with the features of claim 18. Advantageous embodiments of the invention are described in the sub-claims.

The invention is based on the idea that a gas phase reaction leads to an accelerated layer deposition, compared to a layer deposition process solely based on a self limiting chemical reaction on the substrate surface.

Furthermore, it is based on the consideration of providing the first precursor to a surrounding space above a surface region of the substrate in a pulse-shaped concentration variation, whereby the surrounding space may be a reaction space relevant to the material layer deposition onto the surface region, as will be explained in the following. Herein, the second precursor activated by plasma discharge may either also be varied in its concentration in the surrounding space in a pulse shape, or it may be added thereto continuously. - A -

Due to a pulse-shaped introduction of the first precursor, it may be utilized more efficiently.

The chemical deposition reaction is the entirety of sub-reactions, which begin in the gas phase, whereby they take place at least partially in the surrounding space of the substrate surface, and end in the solid phase in form of the material layer on the substrate surface. The deposition reaction may possible also comprise chemical reactions on the substrate surface, for example a conditioning of the substrate surface by plasma activated reactants. Herein, the participation of the two precursors in the chemical deposition reaction means that also other precursors may participate therein, preferably however, only the first and the second precursor participate therein.

The surface region may be substantially the entire substrate surface. The surrounding space above the surface region is the space, which surrounds the surface region adjacently inside the reaction chamber, and in which reaction products form due to gas phase reactions, which as part of the entire deposition reaction lead to the deposition of the solid thin layer on the surface region. In other words, the surrounding space is the space region relevant to the deposition onto the corresponding surface region.

The previous and following mentioning of the pulse shaped concentration variation, as well as any pulse shaped progress of temporally changing variables and parameters, relate not only to the case of a substantially rectangular progression, but also includes temporally repeating pulse shaped variations. Since strictly rectangular progressions are physically not possible, the slopes of such pulses inevitably have a more or less steep progression. The pulse shaped variation of the concentration of a precursor in the surrounding space may for example be achieve by way of a pulse shaped streaming of the precursor into the surrounding space.

A material layer produced in this manner can also have further advantages besides the already mentioned time and cost efficient production, for example a better layer quality. In particular, a passivation layer produced with this method to some extent provides for a better passivation quality for a wafer solar cell, than passivation layers produced by ALD processes, which leads to a lower surface recombination rate on the solar cell surface passivated in this way, and therefore to solar cells with a higher efficiency.

In silicon wafers passivated on both sides by way of such deposited A12O3 passivation layers, effective charge carrier lifetimes in the region of about 10 to 20 milliseconds can be achieved with minority charge carrier densities of about 10¹³ cm^"3 to about 3 x 10¹³ cm^"3, injected into the semiconductor by a suitable light exposure. Hereby, the measured lifetime depends, besides the recombination properties on the surface, also on the quality and the doping of the semiconductor material. The previously mentioned value ranges apply in particular to n-type silicon wafers with a specific electrical resistance of about 3 Ohm cm, which has been produced in a floating zone process (so called FZ material).

The previously used expression of concentration relates in all the herein described precursors and reactants in general on an amount of substance or a particle number per volume. As the second precursor, for example, molecular oxygen may be used. However, other substances or substance compounds that may be activated by plasma are thinkable. This includes substances and substance compounds, which are also active without a plasma discharge, which activity level is, however, raised by plasma discharge.

The deposition device at least comprising a reaction chamber, a transport device, a plurality of introduction units, such as introduction nozzles, a control device, and a plasma discharge device, is preferably designed as an inline device. It can therefore be integrated into a solar cell production line. Herein, by way of the transport device, which is controlled by the control device, the substrate is moved along the substrate movement direction through the reaction chamber from a start section to an end section, being preferably an entrance and an exit of the reaction chamber, respectively. Nevertheless, the previously described embodiments of the plasma enhanced deposition method may also be performed with simpler deposition devices. In particular, the pulse-shaped variation of the precursor concentration in the surrounding space above the surface region may be achieved exclusively by way of a time-dependent actuation of a single or multiple introduction units. A movement or transportation of the substrate by way of a transport device could then be omitted. The plasma discharge device, which produces the plasma discharge for activation of the second precursor, can use a method of energy supply suitable therefor. Usually, however, an actuation by a high frequency voltage or by high-frequency currents is preferred for this purpose, although a plasma generation by way of supplying an electromagnetic wave is also possible.

In a preferred embodiment of the method, the concentration of the first precursor in the surrounding space above the surface region is reduced in a pulse shape, such that the deposition of the material layer is substantially prevented. Hereby, the concentration of the first precursor is reduced such that, for example, it falls below a concentration threshold value necessary for maintaining the chemical deposition reaction. Usually, the concentration is reduced to zero, such that there is substantially no precursor presence in the surrounding space anymore. In a preferred embodiment, the second precursor is activated for the chemical gas phase reaction with the first precursor by the plasma discharge. In this case, the second precursor functions as a reactant for the chemical gas phase reaction with the first precursor. According to preferred embodiments, as reactant for the chemical gas phase reaction with the first precursor, a third precursor is introduced in vapor form into the reaction chamber. This third precursor can also be activated for the chemical gas phase reaction with the first precursor by a plasma discharge, for example generated by the plasma discharge device or by a further plasma discharge device independent thereon. Preferably, however, as the third precursor a substance or a component is used, which is reactive even without the help of a plasma discharge. For example, water (H2O) or ozone (03) may be used instead of molecular oxygen (02).

Preferably, the second and / or the third precursor are introduced such that their concentration in the surrounding space above the surface region is varied in a pulse shape. The varying of the concentration of the second and / or the third precursor in this as well as in the following embodiment may additionally or solely be controlled with the help of a movement of the substrate in the reaction chamber.

In an advantageous embodiment, the second and the third precursor are introduced such that their concentration in the surrounding space above the surface region is varied alternately in a pulse shape. In other words, in the surrounding space the second and the third precursor are present in a temporally alternating sequence, whereby the respective other precursor is present in the surrounding space preferably only as a residue or substantially not at all. Preferably, before the introduction of the second or the third precursor, the surrounding space is freed from the respective other precursor, for example by way of flushing the first precursor or a substantially inert gas through the surrounding space. In a further advantageous embodiment, the second precursor is introduced such that its concentration in the surrounding space above the surface region is substantially constant. With a second precursor such constantly present, the reactivity may be controlled by temporally varying the activation of the second precursor by way of controlling the plasma discharge.

According to a preferred embodiment, the plasma discharge for activating the second precursor is confined to a plasma region, having a distance to the surface region of the substrate. It is therefore an indirect plasma, which does not extend to the surface of the substrate. This way, one may accomplish that while the activation of the second precursor for the chemical reaction with the first precursor takes place, the surface region of the substrate is not under the direct influence of the utilized plasma discharge. Alternatively, the plasma discharge utilized for the activation of the second precursor may extend to the surface region of the substrate. In this case, one speaks of a direct plasma.

Preferably, the plasma discharge is controlled such that the surface region is at least temporarily conditioned. The conditioning of the surface region relates to an interaction between particles of the substrate surface and the plasma positioned above it, for influencing physical and / or chemical properties of the produced material layer. Such a conditioning of the surface region takes place preferably before each pulse-shaped appearance of the first precursor in the surrounding space. Herein, an indirect plasma may be continued to be utilized for the activation of the second precursor, the plasma having a distance to the surface region of the substrate, while for the conditioning of the surface region in-between the pulses of the second precursor it may be switched to a direct plasma. Alternatively, the same direct plasma may be used for the activation of the second precursor, and simultaneously for the conditioning of the surface region.

According to a preferred embodiment, an inert gas is introduced into the reaction chamber such that it reaches the surrounding space above the surface region continuously or in a pulse-shape. Preferably, the surrounding space is flushed by the inert gas in a pulsed manner to free it from the precursors present there. However, at least the reduction of the concentration of one or multiple of the precursors may be achieved in a pulsed manner by way of the inert gas, such that the deposition of the material layer is substantially prevented. Alternatively, the inert gas may be just a buffer or carrier gas, utilized for controlling the progression of the chemical gas phase reaction, or for measured application of one or multiple of the precursors. Preferably, the first precursor comprises a metal or semiconductor compound, preferably trimethylaluminum (Al(CH3)3) or aluminum chloride (AIC13).

However, other precursors suitable for the thin layer deposition may be utilized, in particular organometallic precursors. Further examples of the first precursor may be trimethylamine alane (TMMA ((CH3)3NAIH3)) or aluminum acetylacetonate (Al(C5H7O2)n, wherein preferably n=3, this material is also called in short Al(acac)3). As possible candidates for the second precursor for example oxygen (02) or nitrous oxide (N20) may be considered.

According to an advantageous embodiment, a wafer solar cell is provided as substrate, and as material layer a passivation layer for surface passivation of the solar cell is produced. Preferably, the introduction of the second and / or the third precursor and / or the movement of the substrate is performed such that the concentration of the second and / or the third precursor in a surrounding space above the surface region is varied in a pulse shape. Like the varying of the concentration of the first precursor, the pulse-shaped concentration variation of the second and / or the third precursor in the surrounding space above the surface region is preferably controlled solely by way of controlling the introduction of the respective precursor into the reaction chamber, or solely by way of movement of the substrate through a spatially varying concentration profile inside the reaction chamber.

In a preferred embodiment, the pulse-shaped variation of the concentration of the first precursor in the surrounding space above the surface region takes place with a pulse duration of between about 0.5 seconds to about 2 seconds, preferably with a pulse duration of about 1 second. These values may be achieved with a purely temporal control of the introduction unit by way of introducing the first precursor in a pulse shape with a pulse duration of about 5 milliseconds to about 50 milliseconds, preferably with a pulse duration of about 20 milliseconds. These values may thus for example be valve opening times. AU pulse duration values may for example be measured as full width at half maximum.

According to a preferred embodiment, the pulse-shaped variation of the concentration of the first precursor in the surrounding space above the surface region takes place with a pulse spacing of between about 0.1 seconds and about 5 seconds, preferably with a pulse spacing of about 3.5 seconds. These are preferably the spacings between two pulse peak values. The pulse durations and the pulse spacings optimal for the material layer production may among others be dependent on the reactor geometry, but also on valve switching times of the introduction unit. The previously mentioned values apply in particular but not solely for single wafer reactor chambers, which are comparatively small. In production-scale deposition devices having large volume reaction chambers, both the pulse durations and the pulse spacings may be chosen to be significantly longer.

In a preferred embodiment, during each pulse of the pulse-shaped variation of the concentration of the first precursor, a material sub-layer of the material layer with a layer thickness of between about 1 angstrom and about 50 angstrom is produced, preferably between about 2 angstrom and about 5 angstrom, more preferred with a layer thickness of about 3.5 angstrom.

In a preferred embodiment of the deposition device, the plasma discharge is placed in front of and behind at least one of the introduction units when viewed in the direction of the substrate movement direction. Herein, the plasma discharge may spatially be positioned above the introduction unit, thus on a side of the introduction unit facing away from the substrate. In such a case, it is an indirect plasma, which has no direct influence in form of a conditioning of the solar cell surface.

In an advantageous embodiment, the plasma discharge device produces a plasma discharge space, which spans or encompasses a plurality of introduction units. Thus, in the latter case, the introduction units are placed inside the plasma discharge space.

According to a preferred embodiment, the plasma discharge device comprises a plurality of plasma discharge units, which when viewed along the substrate transport direction each comprise plasma discharge spaces separated from each other. In this embodiment, the plasma discharge spaces can each have different plasma parameters. This way, there is a greater flexibility in the application of the deposition device. A further reason for greater flexibility for example compared to a PEALD deposition device designed as an inline device, is that in PEALD deposition devices the layer thickness of the deposited material layer is set, once the device length (and therefore also the number of deposition cycles) is determined during the device conception. In the present embodiment, there is the possibility to deposit material layers of different thickness by way of variation of process parameters (for example TMA amount or plasma intensity).

Using the deposition method and the deposition device according to one of the previously described or following embodiments, besides material layers made of A12O3 with very good layer qualities, among others, material layers made of the following materials may be deposited, when the precursors are chosen appropriately:

AIxOyNz, aluminum oxynitride or aluminum nitride, whereby as the precursor aluminum precursors as well as N2, H2, NH3, N2O and / or 02 may be utilized. Such material layers may for example be utilized for anti-reflective and / or passivation applications.

TiO2, whereby as precursor titanium tetrachloride (TiCW), tetraisopropyl titanate (TIPT, Ti(OC3H7)4) and / or tetraethoxy titanate (TEOT, Ti(C2H5)4 ) may be utilized. TiO2 material layers are suitable for example as anti- reflective coating. Tantalum oxide (Ta2O5), whereby as precursor tantalum pentaethoxide (Ta(OC2H5)5 or Ta(OCH3)5) may be utilized. Tantalum oxide material layers are suitable for example as corrosion protection layers. SiO2, whereby as precursor tetraethoxysilane (TEOS, Si(OC2H5)4),

hexamethyldisiloxane (HMDSO) and / or tetramethyldisiloxane (TMDSO) may be utilized. SiO2 material layers are suitable for various purposes, such as for example surface passivation. SiN, whereby as precursor hexamethyldisilazane (HMDSN) and / or

hexamethylcyclotrisilazane (HMCTSZN) may be utilized. SiN material layers are suitable for example for passivation and / or anti-reflective applications.

Layer systems of different materials may be deposited by for example changing the precursors and / or their combinations in a deposition process taking place continuously. One example therefor are layer systems, which comprise A12O3 material layers and AIxOyNz material layers. As a further example, a layer comprising A12O3 material layers and TiO2 material layers may be mentioned. In the following, exemplary embodiments of the invention are described with reference to the accompanying drawings. Herein:

Fig. 1 shows a schematic cross section view of a deposition device;

Fig. 2a and 2b show time diagrams for known PEALD deposition methods; Fig. 3 a time diagram for a known PECVD deposition method;

Fig. 4a to 4i time diagrams for deposition methods according to preferred embodiments;

Fig. 5 an arrangement for an inline deposition method according to one embodiment; and

Fig. 6 an arrangement for an inline deposition method according to a further embodiment. Fig. 1 shows an arrangement of a substrate 3 in a reaction chamber 1 of the deposition device 10 in a schematic cross section view. The substrate 3 is positioned on a substrate holder 2, with the help of which the substrate 3 may for example be heated up or cooled down for an optimal deposition

temperature. Above a surface region 4, which in the present case spreads over the entire surface of the substrate in form of a solar cell 3, a surrounding space 6 is extended, which is indicated by a dashed line. As indicated in Fig. 1 , a first precursor 7 and a second precursor 8 are in a vapor mixture in the reaction chamber 1. By way of a chemical deposition reaction, in which the two precursors 7, 8 take part, the material layer 5 is deposited on the surface region 4 of the substrate 3.

Herein, only that spatial region in the reaction chamber 1 shall be regarded as a surrounding space 6 above the surface region 4, where the precursors present in the spatial region and the processes taking place in it, in particular chemical gas phase reactions, have a direct influence on the surface region 4 lying underneath. In particular, the processes and precursors in the surrounding space 6 as well as their temporal succession are significant for the deposition and, if applicable, for the conditioning of the material layer 5 of the surface region 4 by way of a plasma, which is not shown herein.

The deposition device 10 comprises further an introduction unit 9, for example furnished as an injection nozzle. By way of the introduction unit, the first precursor 7 and the second precursor 8 and, if applicable, further precursors, reactants and / or inert substances are introduced into the reaction chamber 1. In alternative embodiments, further introduction units 9 may be provided for, through each of which different substances may be introduced into the reaction chamber 1. Preferably even multiple introduction units 9 are used for introducing one of the precursors into the reaction chamber 1.

In the following Fig. 2a, 2b, 3 and 4, the temporal actuations of the

introduction unit 9 for the introduction of the various gases into the reaction chamber as well as the energy supply to that plasma discharge are visualized by way of timing diagrams. In the following, reference is made to a high frequency actuation for plasma generation, although also other types of energy supply may be utilized. Although in the following it is assumed that the following figures show the temporal actuation of the introduction units 9, the diagrams shown therein may also be schematic depictions of the

concentrations of the corresponding gases in the surrounding space 6 above the surface region 4. Depending on the spatial positioning of the introduction units 9 and the substrate 3 in the reaction chamber 1 , a time delay between the introduction of a substance into the reaction chamber 1 (or the introduction actuation) and a corresponding change in the concentration of this substance in the surrounding space 6 has to be taken into account.

In all the now following timing diagrams, the following is depicted: The first (topmost) line shows a temporal introduction unit actuation 11 for the introduction of an inert gas, the second line shows a temporal introduction unit actuation 12 for the first precursor 7, the third line shows a temporal introduction unit actuation 13 for the second precursor 8, and the fourth line shows the temporal high frequency actuation 14 for plasma generation. In the timing diagrams, it is assumed that each actuation takes place digitally. In other words, the actuation is either activated or deactivated. A deposition comprises in general a multitude of deposition cycles, in order to obtain a sufficient layer thickness. Thus, the time diagrams depicted herein would continue periodically. In the activated state, the corresponding introduction unit will introduce the corresponding gas or the corresponding precursor with a predetermined pressure, while the high frequency source is turned on for plasma generation in the active state. In reality, each actuation will have a rise and a fall time, which may be more or less short, depending on the design of the mechanical or electronic components utilized herein.

The Fig. 2a, 2b and 3 depict the situation for the known prior art. While the Fig. 2a and 2b relate to two different plasma enhanced atomic layer deposition methods (PEALD methods), a time diagram for a plasma enhanced chemical vapor deposition method (PECVD method) is shown in Fig. 3. During the known PEALD method according to Fig. 2a, the two precursors 7 and 8 are introduced into the reaction chamber 1 in continuous alternation, whereby a plasma is ignited in the second precursor by way of the high frequency actuation 14, which activates it for a chemical reaction. Between phases of introducing the two precursors 7 and 8, the reaction chamber 1 is flushed by an inert gas, which is indicated by the corresponding introduction unit actuation 11. This takes place in order for the two precursors 7 and 8 not to be present in gas phase in the reaction chamber 1 , but instead react with each other exclusively on the surface, ideally in a monolayer, such that no gas phase reaction may take place.

The timing diagram shown in Fig. 2b differs from the one in Fig. 2a only by that the second precursor is introduced continuously into the reaction chamber 1. Because the high frequency actuation 14 for the plasma generation continues to take place in a pulse shape, the plasma generation and therefore the activation of the second precursor 8 takes place in a pulse shape as well. In this case, it is substantially not harmful that the two precursors 7 and 8 are temporarily inside the reaction chamber 1 simultaneously, since a chemical gas phase reaction cannot take place in this case either.

Fig. 3 shows the corresponding temporary progress in the PECVD process. As initially described, this is a continuous deposition process, whereby the actuations 12 and 13 of introduction units 9 for the two precursors 7 and 8 and also the high frequency actuation 14 take place simultaneously and

continuously. Therefore, the first precursor 7 and a second precursor 8 activated by plasma discharge coexist inside the reaction chamber 1.

Accordingly, the introduction unit actuation 11 for the inert gas is deactivated.

The Fig. 4a to 4i show in diagram form the temporal progress of the actuations 11 , 12, 13 for the introduction units 9 for introducing of the inert gas, the first precursor 7, and the second precursor 8, as well as for the high frequency actuation 14 in exemplary embodiments of a plasma enhanced deposition process. While in the cases shown in Fig. 4a to 4c the actuation 13 for the introduction of the second precursor 8 is activated continuously, and therefore the concentration of the second precursor 8 in the surrounding space 6 above the surface region 4 is or is held constant, supplying the second precursor 8 in the embodiments according to Fig. 4d to 4f takes place in a pulse shape. In all embodiments, however, the introduction of the first precursor 7 is taking place solely in a pulse shape. Herein it is advantageous that the pulse duration for the actuation 12 is substantially shorter that the pulse spacing between two pulses, although these two parameters appear to be of the same length at least in the Fig. 4a to 4c. Preferably, the pulse duration is shorter than the pulse spacing by about two orders of magnitude.

According to Fig. 4a, a feeding-in of the first precursor 7 and a feeding-in of the inert gas take place alternately, whereby the feeding-in of the second precursor 8 and its activation by way of plasma discharge takes place continuously. In contrast to this case, in the embodiment according to Fig. 4b, the inert gas is introduced continuously, while in the embodiment according to Fig. 4c it is avoided altogether. In the latter case, the continuously introduced second precursor takes over the purging function.

In the embodiment of the plasma enhanced deposition process according to Fig. 4f, the first precursor 7 and the inert gas are introduced alternately, like in the case according to Fig. 4a. However, unlike in the embodiment according to Fig. 4a, also the second precursor 8 is introduced substantially

simultaneously with the first precursor 7 in a pulse shape and also a plasma discharge is generated by way of a pulse-shaped high frequency actuation 14. The embodiments according to Fig. 4d and 4e differ by that the pulse-shaped introduction of the first precursor 7 takes place with half the repetition frequency compared to the pulse-shaped introduction of the second precursor 8. Furthermore, the embodiments according to Fig. 4d and 4e differ by way of a differing high frequency actuation 14 for the plasma generation, which is pulse-shaped in the former case and substantially continuously in the latter case.

The Fig. 4a to 4f comprise each the same four lines, which is why in the further figures the reference numerals are omitted. In contrast, a further temporal actuation is depicted in the Fig. 4g, 4h, and 4i by way of an additional line. It relates to a temporal introduction actuation 15 for the introduction of a third precursor. It temporarily replaces the second precursor 8 in the chemical gas phase reaction with the first precursor 7. While the second precursor 8 has to be first activated by way of the plasma discharge for the chemical reaction with the first precursor 7, the third precursor is preferably active on its own and does not need a plasma therefor. This may for example be ozone or water, which may replace the molecular oxygen as the second precursor 8. In the embodiments according to Fig. 4g to 4i, the introduction of the second precursor 8 takes place in alternation with the introduction of the third precursor. The high frequency actuation 14 corresponds to the introduction unit actuation 13 for the second precursor 8, and takes place preferably substantially synchronous to it. The flushing or purging of the surrounding space 6 above the surface region 4 or of the entire reaction chamber 1 takes place in a temporal spacing or a time window between the alternating introduction of the second and the third precursor.

The embodiments illustrated in Fig. 4a to 4i differ in a frequency, with which the introduction unit actuation 12 for the first precursor 7 takes place. While the introduction unit actuation 12 according to Fig. 4h takes place

simultaneously with the actuation 13 for the second precursor 8, it takes place according to Fig. 4g with double the repetition frequency. Finally, the repetition frequencies of the introduction unit actuations 12 and 13 for the first and the second precursor 7, 8 in the embodiment according to Fig. 4h and 4i are the same. However, the two introduction unit actuations 12 and 13 are displaced against each other by half a period length, and thus the feeding-in of the first precursor 7 and the second precursor 8 take place alternately. For the manufacture of material layers made of A12O3, which serve as passivation layers on wafer solar cells, are in particular suitable

trimethylaluminum (TMA) as the first precursor 7, and molecular oxygen (02), which is activated by way of a plasma discharge, as the second precursor 8. As the inert gas, argon (Ar) is well suitable. A very long charge carrier lifetime and therefore a good surface passivation is achieved in this case for example by way of a continuous plasma discharge, with a pulse duration of the introduction of TMA of about 20 milliseconds and a time interval between introduction pulses of the TMA or a pulse spacing of about 3.5 seconds (s).

The temperature of the substrate 3 should in this case be about 200 ⁰C, while the 02 gas is introduced with a gas flow of about 50 standard cubic centimetre per minute (seem), and the Ar gas with a gas flow of about 20 seem. The plasma frequency, that is the frequency of the high frequency actuation for the plasma discharge, is preferably 13.56 MHz, whereby the plasma has a plasma power of about 150 Watts. Finally, the pressure in the reaction chamber 1 should preferably have a value of about 150 millitorr. These parameters result in a growth rate for the thickness of the material layer 5 of about 3.5 angstrom (A) per TMA pulse or per deposition cycle, which corresponds to a layer deposition rate of about 1 angstrom per second (A/s). Changing the TMA pulse separation to values of 2.5 s, 1.5 s, and 0.5 s, results in growth rates of the thickness of the material layer 5 of 3,75 A, 4,5 A and 5 A as well as corresponding layer deposition rates of 1 ,5 A/s, 3 A/s and 10A/s. This is when neglecting the very short TMA pulse duration. The growth per pulse values are different, because the material sub-layers 51 , which

constitute the material layer 5, grow with differing densities. In the present deposition process, the material layer thickness per deposition cycle, or per pulse of the first precursor, may be adjusted by way of the amount of the first precursor introduced per deposition cycle. For the passivation of the solar cell surface, preferably a suitable material layer with a layer thickness of at least 5 nanometres is deposited. Depending on the concrete material properties, the layer thickness is preferably well above this value. If the material layer additionally or exclusively takes over the function of a backside mirror of the solar cell, a layer thickness of about 100

nanometres or more are advantageous.

Fig. 5 shows a schematic cross section view of an arrangement inside a reaction chamber 1 according to an inline embodiment of the deposition device 10. The deposition device 10 comprises multiple introduction units 9 distributed side by side equidistantly along a substrate movement direction 22. They are

enveloped by a plasma discharge space 21 , which is generated by a plasma discharge device (not shown in Fig. 5). Below the introduction units 9, a substrate 3 is positioned, which is moved by a transport device (also not shown) along a substrate movement direction 22, preferably with a constant transport speed.

At the lower end of Fig. 5, a schematic and, in comparison to the dimensions of the substrate 3, extremely expanded depiction of the material layer 5 is shown, in order to illustrate its growth along the substrate movement direction 22. The horizontal and vertical auxiliary dashed lines drawn in Fig. 5 make it clear that the material layer 5 is composed of material sub-layers 51 , each of which are generated due to the first precursor 7 introduced by a corresponding introduction unit 9. Thus, the layer thickness of the material layer 5 grows substantially in steps along the substrate movement direction 22.

In case of a purely temporal control of a single introduction unit 9 for the introduction of the first precursor 7, while the substrate 3 is held stationary, a material sub-layer 51 would correspond to the deposition result after a deposition cycle.

Finally, the Fig. 6 shows an arrangement in a reaction chamber 1 according to a further embodiment of the deposition device 10. Herein, multiple substrates 3 are arranged consecutively on a substrate holder 2. As in the case shown in Fig. 5, the substrates 6 are moved through the reaction chamber 1 along the transport direction 22 by a transport device not shown in Fig. 6. For example, the substrate holder 2 may be a transport belt, which is moved by spools positioned outside of the reaction chamber 1.

Above the substrate holder 2 with the substrates 3, an introduction unit 9 is positioned, with which help the first precursor 7 (dashed arrows) and the second precursor 8 (solid arrows) are introduced in the direction of the substrate 3, alternately along the transport direction 22. An energy source 25 shown in Fig. 6 may both serve for supplying the introduction unit 9 as well as for supplying the plasma discharge device for the generation of the plasma discharge in the plasma discharge spaces 21 , which are positioned along the transport direction 22. Therefore, the substrates 3, while moving along the transport direction 22, traverse alternately regions, in which exclusively the first precursor 7 is present, and plasma discharge spaces 21 , wherein the second precursor 8 is present. The plasma discharge spaces 21 distributed along the substrate transport direction 22 can be generated by one or by multiple plasma discharge units.

Reference numerals:

1 reaction chamber

2 substrate holder

3 substrate

4 surface region

5 material layer

51 material sub-layer

6 surrounding space above the surface region

7 first precursor

8 second precursor

9 introduction unit(s)

10 deposition device 11 temporal introduction unit actuation for inert gas

12 temporal introduction unit actuation for first precursor

13 temporal introduction unit actuation for second precursor

14 temporal introduction unit actuation for plasma discharge

15 temporal introduction unit actuation for third precursor

21 plasma discharge space

22 substrate movement direction

25 energy source

Claims

Claims:

1. Plasma enhanced deposition of a material layer (5), wherein a first

precursor (7) and a second precursor (8) are introduced in vapor form into a reaction chamber (1 ) and participate in a chemical deposition reaction, by which the material layer (5) is deposited onto a surface region (4) of a substrate (3) placed inside the reaction chamber (1 ), wherein the introduction (12) of the first precursor (7) and / or a movement (22) of the substrate (3) is performed such that the concentration of the first precursor (7) in a surrounding space (6) above the surface region (4) is varied in a pulse shape, and in the surrounding space (6) at least partly a chemical gas phase reaction of the first precursor (7) with a reactant takes place, and wherein the second precursor (8) is activated for a chemical reaction with the first precursor (7) by a plasma discharge.

2. Plasma enhanced deposition according to claim 1 , characterized by that the concentration of the first precursor (7) in the surrounding space (6) above the surface region (4) is reduced in a pulse shape such that the deposition of the material layer (5) is substantially prevented.

3. Plasma enhanced deposition according to claim 1 or 2, characterized by that the second precursor (8) is activated for the chemical gas phase reaction with the first precursor (7) by the plasma discharge.

4. Plasma enhanced deposition according to one of the claims 1 to 3,

characterized by that, as reactant for the chemical gas phase reaction with the first precursor (7), a third precursor is introduced in vapor form into the reaction chamber (1 ).

5. Plasma enhanced deposition according to one of the previous claims, characterized by that the second and / or the third precursor are introduced such that their concentration in the surrounding space (6) above the surface region (4) is varied in a pulse shape.

6. Plasma enhanced deposition according to claim 5, characterized by that the second and the third precursor are introduced such that their concentration in the surrounding space (6) above the surface region (4) is varied alternately in a pulse shape.

7. Plasma enhanced deposition according to one of the claims 1 to 4,

characterized by that the second precursor (8) is introduced such that its concentration in the surrounding space (6) above the surface region (4) is substantially constant.

8. Plasma enhanced deposition according to one of the previous claims, characterized by that the plasma discharge fort activating the second precursor (8) is confined to a plasma region (21 ), having a distance to the surface region (4) of the substrate.

9. Plasma enhanced deposition according to one of the claims 1 to 7,

characterized by that the plasma discharge is controlled such that the surface region (4) is at least temporarily conditioned.

10. Plasma enhanced deposition according to one of the previous claims, characterized by that an inert gas is introduced into the reaction chamber (1 ) such that it reaches the surrounding space (6) above the surface region (4) continuously of in a pulse-shape.

11. Plasma enhanced deposition according to one of the previous claims, characterized by that the first precursor (7) comprises a metal or semiconductor compound, preferably trimethylaluminum (Al(CH3)3) or aluminum chloride (AIC13).

12. Plasma enhanced deposition according to one of the previous claims, characterized by that as substrate (3), a wafer solar cell is provided, and as material layer (5) a passivation layer for surface passivation of the solar cell is produced.

13. Plasma enhanced deposition according to one of the previous claims, characterized by that the introduction of the second (8) and / or the third precursor and / or the movement of the substrate (3) is performed such that the concentration of the second (8) and / or the third precursor in a surrounding space (6) above the surface region (4) is varied in a pulse shape.

14. Plasma enhanced deposition according to one of the previous claims, characterized by that the pulse-shaped variation of the concentration of the first precursor (7) in the surrounding space (6) above the surface region (4) takes place with a pulse duration of between about 0.5 second to about 2 seconds, preferably with a pulse duration of about 1 second.

15. Plasma enhanced deposition according to one of the previous claims, characterized by that the pulse-shaped variation of the concentration of the first precursor (7) in the surrounding space (6) above the surface region (4) takes place with a pulse spacing of between about 0.1 seconds and about 5 seconds, preferably with a pulse spacing of about 3.5 seconds.

16. Plasma enhanced deposition according to one of the previous claims, characterized by that during each pulse of the pulse-shaped variation of the concentration of the first precursor (7), a material sub-layer (51 ) of the material layer (5) with a layer thickness of between about 1 angstrom and about 50 angstrom is produced, preferably between about 2 angstrom and about 5 angstrom, more preferred with a layer thickness of about 3.5 angstrom.

17. Semiconductor device, in particular wafer solar cell, with a material layer (5), which is deposited by a plasma enhanced deposition method according to one of the previous claims.

18. Deposition device (10) for plasma enhanced deposition of a material layer (5), comprising:

- a reaction chamber (1 ) extending along a substrate movement

direction (22), in which a first precursor (7) and a second precursor (8) can be introduced in a vapor form, in order to participate in a chemical deposition reaction, by which the material layer (5) is deposited on a surface region (4) of a substrate (3) placed inside the reaction chamber

(1 );

- a transport device for transporting the substrate along the substrate movement direction (22) from a start section to an and section of the reaction chamber (1 );

- a plurality of introduction units (9) for introducing the first precursor (7) into the reaction chamber (1 ), whereby the introduction units (9) are placed distributed along the substrate movement direction (22);

- a control device for controlling the introduction units (9) and the transport device such that the concentration of the first precursor (7) in a surrounding space (6) above the surface region (4) is varied in a pulse shape; and

- a plasma discharge device, which is designed to activate the second precursor (8) for a chemical reaction with the first precursor (7) using a plasma discharge.

19. Deposition device (10) according to claim 18, characterized by that the plasma discharge is placed in front of and behind at least one of the introduction units (9) when viewed in the direction of the substrate movement direction (22).

20. Deposition device (10) according to claim 18 or 19, characterized by that the plasma discharge device produces a plasma discharge space (21 ), which spans or encompasses a plurality of introduction units (9).

21. Deposition device (10) according to one of the claims 18 to 20,

characterized by that the plasma discharge device comprises a plurality of plasma discharge units, which when viewed along the substrate transport direction (22) each comprise plasma discharge spaces separated from each other.