WO2012164546A2

WO2012164546A2 - Microheater-based in situ self-cleaning dynamic stencil lithography

Info

Publication number: WO2012164546A2
Application number: PCT/IB2012/052806
Authority: WO
Inventors: Shenqi XIE; Andreea Veronica SAVU; Juergen Brugger
Original assignee: Ecole Polytechnique Federale De Lausanne (Epfl)
Priority date: 2011-06-03
Filing date: 2012-06-04
Publication date: 2012-12-06
Also published as: WO2012164546A3

Abstract

The invention concerns a method and a device wherein a stencil is provided with an integrated microhotplate on the stencil membrane to prevent aperture clogging, whereby the stencil can be locally heated up in order to minimize material condensation, remove clogging and control the resolution of the deposited material by deforming the membrane towards or away from a substrate.

Description

MICROHEATER-BASED IN SITU SELF-CLEANING DYNAMIC

STENCIL LITHOGRAPHY

CROSS-RELATED APPLICATION

The present application claims priority to the application EP1 1 168719.0 filed on June 3, 201 1 in the name of the same Applicant, the content of which is incorporated by reference in its entirety in the present application. FIELD OF THE INVENTION

The present invention concerns the field of nanostencil lithography with evaporated material being deposited on a substrate. More specifically, the present invention concerns a novel concept for nanostencil lithography with means aimed at eliminating or at least at reducing aperture clogging during material deposition.

BACKGROUND OF THE INVENTION

Stencil lithography (SL) is a shadow mask based technique which allows parallel, resistless, micro- and nanopatterning of material through apertures in a membrane (stencil) onto a substrate. The stencil can be used as a shadow mask to define nanometric structures on a large variety of substrates in different processes, e.g. thin-film deposition, plasma etching or ion implantation. As the most conventional application of SL, thin-film deposition has emerged as a reliable micro-/nano- patterning process in both static [see reference 1] and dynamic mode [see reference 2]. Sub-100 nm patterns have been achieved by static SL (the stencil is clamped to the substrate without relative motion). Various materials like metals, oxides, magnetic materials and organic molecules etc. have been successfully deposited through stencils.

Besides the advantage of having a variety of materials which can be deposited, SL offers unique capabilities in patterning on flexible substrates and on three- dimensional (3D) topography [see reference 3 and 4]. Due to the simple fabrication process and duplication capability, low cost and high volume production could also be realized by industrial manufacturing.

This technique has already shown a promising future to become one of the alternatives as the next generation lithography. Applications in CMOS circuits, MEMS/NEMS, SAMs and organic molecules have already been demonstrated. However, there are still few main technical challenges preventing further development. One of them concerns the deposition of materials on the sidewalls of the apertures, leading to the aperture shrinking or even clogging. This results in the size reduction of the deposited patterns and limits the useful life time of the stencil.

In the case of static SL, clogging doesn't have such a big impact on the first deposition, and the life time of the stencil can be extended by cleaning the membrane after deposition or by pre-coating it with self-assembled monolayers (which has the effect of slowing down the clogging rate). However, in the case of (quasi) dynamic SL (stencil has relative motion to the substrate during or in between depositions) usually more material is deposited in situ before the stencil can be taken out and cleaned (e.g. patterning long lines or several consecutive materials). The total amount of deposited material is integrated over all the evaporations, leading to clogging and eventually to no pattern on the substrate.

Another important challenge is blurring, i.e. the enlargement of the transferred pattern on the substrate. It highly depends on the setup geometry (mainly the gap between the stencil and the substrate) as well as the surface energy of the substrate.

Figure 1 a illustrates the clogging of stencil apertures during metal deposition. This limits the useful life time of the stencil in only one pump-down if precise duplication of the patterns is required [see reference 5]. It becomes an even more severe issue in the dynamic mode, where the continuous motion of the clogged aperture translates to a gradually narrowed pattern. Various approaches have been shown to extend the life time of the stencil. The reusability of nanostencils after Al deposition has been proved by using selective wet etching in order to clean the membrane [see reference 6]. The nanostencils can be reused several times after each cleaning process. However, the thickness of the total deposited material is still limited when the thickness is comparable to the dimension of the stencil nanoapertures.

A different method uses pre-coated self-assembled monolayer on the stencil membrane before deposition [see reference 7]. While this showed a reduced adhesion of gold inside the apertures, the clogging still remained an issue in the thick gold layer building up on top of the membrane. Neither of the aforementioned methods can totally prevent clogging as materials still accumulate on the membranes, leading to the inevitable size reduction of the apertures.

In physical vapor deposition (PVD), material condensation on the substrate is a dynamic process. The condensation of the material on the surface is the difference between the incident material flux and re-evaporated material flux on the surface [see reference 8]. As the re-evaporated material flux depends mainly on the substrate temperature, condensation would be reduced and eventually eliminated with certain substrate temperature. Therefore, one of the solutions to keep the stencil membrane clean is to raise the membrane temperature in order to minimize the material condensation.

US patent 5,742,065 relates to a heater for a membrane mask in an electron-beam lithography system. The aim of this patent is to provide a method for reducing or eliminating contaminants from the surface of a membrane mask used in an e-beam or in an ion-beam lithography system.

In this patent and the technology used, there is a recognition that in the absence of heating a contamination has two detrimental effects on the beam in a projection e- beam lithography system. First the contamination can charge the beam as it impinges on the mask and this charging can cause the beam to be deflected into undesired directions and uncontrollable areas. Secondly, the contamination can grow sufficiently thick to be imaged by the beam.

Therefore, to overcome these undesired effects, the solution is to dope at least one surface of the silicon membrane mask with boron to lower the electrical resistance of the mask and applying a voltage between opposite surfaces of the membrane mask. The voltage generates an electric field that heats the membrane mask. Typically, the experimental values have shown that the disclosed heating method and process raise the temperature of the bulk silicon to approximately 200°C in vacuum.

Japanese patent application JP 62-272529 relates to an electron beam exposing device maintaining the temperature constant in an aperture for shaping an electron beam to eliminate the variation in a shaped beam size due to a variation in an electron beam emitting time. To this effect, an aperture has a rectangular opening at the center, and comprises a linear heater therein along the outer periphery of the opening. The temperature to be raised by the heater in the aperture is set by considering the temperature rising by the emission of an electron beam, the heating of the heater is suppressed to low during a period that the aperture is emitted by the beam, the aperture is heated by the heater during a period that the aperture is not emitted by the beam to maintain the temperature of the aperture during the operation constant, thereby compensating the shape size of the opening constantly. In this prior art, although they are heating horizontally the stencil, they are heating the bulk material, therefore not being able to reach as high temperatures. The aim here is too keep the shape of the aperture fixed for the electron beam to pass through by keeping the stencil at a constant temperature. US Patent 5,428,203 relates to an electron beam exposing apparatus permitting the temperature of mask to be constant without affecting an electron beam. The whole stencil is being heated and the use the apertures to shape an electron beam, thus the stencil is heated by the electron beam in an uncontrollable way, which results in uncontrollable apertures deformation. All three above mentioned prior art publications are used for low temperatures, such as 200°C, to keep a constant temperature or to avoid electron beam scattering. They cannot go to higher temperatures or more local heating. This would require a large electrical current passing close to their apertures, which would induce a large magnetic field. This magnetic field would deflect the electron beam, therefore working against their purpose. Other prior art publications include the following documents: US 2002/1 18027, US 2002/187433, US 5529862, US 5834142, US 5428293, JP 2004202351 , EP 0297506, WO 9975649.

SUMMARY OF THE INVENTION

An aim of the present invention is to provide an improved method and device for lithography.

More specifically, the present invention concerns a novel concept for nanostencil lithography with an integrated microhotplate/microheater on the stencil membrane as means aimed at eliminating or at least at reducing aperture clogging during material deposition.

The method implies local heating, higher temperatures than disclosed in the prior art, for material re-evaporation.

The stencil can be locally heated up for example by integrated coils above room temperature, up to more than 800 °C. The high temperature reduces or even eliminates material accumulation on the membrane. FEM simulation has been carried out to predict the temperature distribution across the heated membrane, which agrees well with the experimental results in ambient conditions. To demonstrate the process, for example Aluminum depositions were performed in static mode simultaneously through heated and non-heated stencils in an electron- beam evaporator. No clogging was found on the heated membrane whereas a clear clogging of the aperture was observed on the non-heated one. The clogging becomes an even more severe issue in a dynamic mode where the stencil can move relatively to the substrate during evaporation. The pattern achieved on the substrate follows the stencil trajectory through the material deposited through the moving aperture. The pattern becomes narrower and narrower until eventually it disappears, due to a complete clogging of the aperture.

Therefore, the method according to the present invention could significantly extend the life time of the stencil by the means of preventing clogging of apertures, especially for thick deposited layers and in the dynamic mode. The present invention concerns a method for nanostencil lithography of evaporated material on a substrate with a stencil membrane with apertures wherein said stencil is provided with heating means on the stencil membrane to prevent aperture clogging, whereby the stencil can be locally heated up in order to minimize material condensation and remove clogging.

In an embodiment the local heating may deform the membrane towards or away from said substrate to modify a gap between said membrane and said substrate thereby influencing the resolution of the deposited structure on the substrate.

In an embodiment the stencil may be heated up to 800°C. In an embodiment the heating means are electrodes powered by a DC source.

In an embodiment a constant voltage may be applied during the evaporation process and the current may be monitored. In an embodiment the temperature may be maintained constant during the evaporation process by increase of the power of the source.

In an embodiment the lithography may made in static mode where the stencil is fixed with respect to the substrate or in dynamic mode where the stencil moves with respect to the substrate.

The invention also concerns a method of fabricating a stencil for use in one of the preceding method, said method comprising the following steps:

-) providing a stencil substrate;

-) coating said substrate with a first low stress film material,

-) forming resistive heating coils with electrically conductive materials;

-) depositing a second low stress film material,

-) forming contact pads and stencil aperture;

-) forming a free standing membrane.

In an embodiment the stencil substrate is a Si substrate.

In an embodiment the low stress film material may be a LPCVD SiN having a thickness of about 200nm.

In an embodiment the heating coils may be made of a layer of 5nm of Ta and a layer of about 200nm of Pt.

In an embodiment the heating coils may be made by lithography and lift off process. In an embodiment the contact pads and stencil aperture may be made by dry etching followed by a DRIE etching to open the membrane window. In an embodiment the free standing membrane may be released by a wet etching.

The invention also concerns a stencil fabricated by the method as defined herein.

The invention also further concerns a device using the method as defined herein with at least a stencil as defined herein.

According to the invention, a stencil for nanostencil lithography of evaporated material on a substrate, may comprise a stencil membrane with apertures wherein said stencil membrane further comprises heating means to prevent aperture clogging, whereby the stencil can be locally heated up in order to minimize material condensation and remove clogging.

In an embodiment the membrane may be deformed by the heating towards or away from said substrate to modify a gap between said membrane and said substrate thereby influencing the resolution of the deposited structure on the substrate.

In an embodiment the stencil may comprise at least a Si substrate with two layers of low stress thin film material and Pt resisting coils as heating means and a Ta layer as an adhesion layer for the coils.

In an embodiment the thin film material may have a thickness of about 200nm.

In an embodiment the Pt coil may have a thickness of about 200nm. In an embodiment the Ta layer may have a thickness of about 5nm.

In an embodiment the stencil as defined herein may comprise a shadow mask to localize the metal deposition on the heated membrane and extend the life-time of the heated stencil.

DETAILED DESCRIPTION OF THE INVENTION

In the present application, in a first embodiment, we describe a novel concept for eliminating clogging by locally heating up the stencil during metal deposition, minimizing thus materials' accumulation on the membrane.

In another embodiment, a function of the heated membrane is its deformation towards or away from the substrate, tuned by the coil design of the microhotplate used for heating to form a bi-morph structure. The gap between the stencil and the substrate may for example be greatly reduced by this thermal actuation of the membrane, leading to an improvement of the resolution of the deposited structures on the substrate.

Further features and embodiments of the invention will be apparent from the following description and from the drawings which show

Figure 1 a illustrates the clogging of stencil apertures during metal deposition;

Figure 1 b illustrates the prevention of clogging and the minimization of the gap by using an integrated microhotplate on the stencil membrane;

Figures 2(a) to 2(f) illustrate fabrication steps of the heated stencil according to the present invention

2(a) Deposition of low stress thin film material;

2(b) Definition of resistive heater;

2(c) Deposition of second low stress thin film material for passivation,

2(d) Patterning of contact pads and stencil apertures;

2(e) Definition of backside windows and

2(f) Releasing of membrane.

Figures 3(a) to 3(d) illustrate optical micrographs of

3(a) full wafer heated stencil,

3(b) single membrane with integrated microhotplate with

3(c) zoom-in and

3(d) SEM image of the apertures in between the coils;

Figure 4(a) illustrates an optical image of a glowing Pt coil in ambient conditions and figure 4(b) illustrates the temperature distribution of the same structure by FEM simulation;

Figures 5(a) and (b) illustrates Optical images with illumination from the back side of the microscope of the heated stencil according to the present invention;

Figures 6(a) to (f) illustrate SEM images of apertures on the membrane coated with 120 nm Al and of corresponding patterns on a Si substrate;

Figure 7 illustrates (a) Localized deposition on heated membrane, (b) Graph showing the average temperature of the heated membrane vs. the thickness of evaporated Al. The localized deposition extends the life-time of the heated stencil

The heated membranes are fabricated by integrating microhotplates as resistive heaters [see reference 9] with stencils. The resistive heater is integrated on a stress- free membrane of LPCVD SiN and is thermally isolated from the surrounding bulk material in order to achieve a high local temperature. The heating coils are made of electrically conducting materials, such as for example Pt with Ta serving as an adhesion layer. Ta has the advantage over other adhesion layers, e.g. Ti, to be more compatible with the second deposition of LPCVD SiN, providing more reliable and robust membranes [see reference 8]. Stencil apertures 6 are located in between the coils where the temperature necessary for unclogging the apertures is reached.

The fabrication process of the heated stencils started from a Si substrate wafer 1 coated with low stress thin film material 2 (for example a 200 nm LPCVD SiN) on both sides (Figure 2a).

The resistive heating coils 3 are made of electrically conducting materials (see Figure 2b), such as a 5nm Ta/200nm Pt defined by lithography and lift off process for example. A second layer of the low stress thin film material 4 (for example 300 nm LPCVD SiN) is then deposited on both sides of the wafer of Figure 2(b) to serve as the passivation layer (see Figure 2c).

Contact pads 5 and stencil apertures 6 were formed simultaneously, for example by SiN dry etching (see Figure 2d), followed by an etching process on the back side to open the membrane window 7 (see Figure 2e).

The free-standing membrane was released finally in a controllable wet etching process (see Figure 2f).

Figures 3(a) to 3(d) shows optical images of a full wafer heated stencil 10 (Figure 3(a)), a single membrane with integrated microheater 1 (Figure 3(b)), the coils 2 (Figure 3(c)) and a close-up SEM image of the apertures 6 (Figure 3(d)), respectively. The flatness of the membrane after fabrication was measured with an optical profilometer. No obvious deformation due to the induced stress from the metal coils has been observed.

The measured temperature is an average value over the heated area, with the temperature distribution on the membrane being not uniform. The heated stencil was tested with an increasing power until its glowing point (for example 600 °C for Pt [see reference 10]) in order to make the temperature variation visible.

Figure 4a shows the glowing coil 3 in ambient conditions. The brightest area in the centre of the coil 3 corresponds to the part with the highest temperature, which must be higher than the measured average value (650 °C). The optical image validates the Pt thin film visible glows starting at around 600 °C [see reference 9].

A FEM model was created to simulate the temperature gradient on the membrane 10 (see Figure 4b). Boundary conditions fix room temperature on the border of the membrane 10, which confines the heat generated by the resistive heater 1 1 , on the heated stencil. The result agrees very well with the optical image. Due the special layout of the electrode, a stronger thermal coupling effect must happen in the middle of the membrane 10, which decreases the temperature gradient in that area. The relatively uniform temperature distribution in the center of the membrane 10 provides a stable thermal environment for the stencil apertures 6.

To prove the concept, for example, the heated stencil 10 was placed on a substrate for metal deposition in an e-beam evaporator. The integrated electrodes were powered by a DC power source. A constant voltage was applied and the current through the heater was monitored during the whole evaporation process. The membrane 10 was heated in vacuum before the evaporation to achieve a stable temperature. Due to the thermal conduction of the metal condensing on the surface of the membrane 10, the temperature would be continuously decreasing if the power is kept constant during the evaporation. Therefore, a gradually increased input power was applied in order to maintain the variation of the temperature in time as small as possible. The optimal evaporation rate is a balance between a maximum required for minimizing condensation of the material on the heated membrane 10, and a minimum set by the requirement of having enough material patterned through apertures onto the substrate.

Figures 5(a) and (b) show the backside illuminated optical images of the heatable stencil 10 before (Figure 5(a)) and after (Figure 5(b)) the deposition of a layer for example of 120 nm Al. The transparent part in the center of the stencil 10 after deposition indicates very little Al in that area. This part also corresponds to the hottest area during evaporation.

The apertures 6 in the middle of the stencil on the non-heated membrane 10 and the heated one are compared from both sides of the membrane 10, as well as the corresponding patterns on the substrate, as shown in Figures 6(a) to 6(f). By comparing the front side images (figures 6(b) and 6(e)), a clear 300 nm shrinkage of the aperture 6 was observed on the non-heated membrane whereas no clogging was found on the heated one. A very short selective etching was performed on the substrate to enhance the contrast of the patterned structures under scanning electron microscope. The transferred pattern through the heated aperture 6 (Figure 6f) shows a flatter surface on the top of the main structure than the pattern achieved from the non-heated aperture (Figure 6c). This result concurs that there is no clogging on the heated aperture 6. Further studies show that the effectiveness of heating the stencil is reduced for patterning thermal conducting material, as the evaporated material increases the thermal conductance of the membrane 10 to the surrounding bulk frame. In order to maintain a constant membrane 10 temperature, the input power has to compensate for the decrease in thermal resistance to the bulk frame. For example, 200 nm of evaporated Al the power necessary to maintain a temperature effective for unclogging is large enough to break the membrane made of SiN. Therefore, the metal thickness determines now the life-time of the heated membrane. To extend the life-time of a heated stencil 10, an extra aligned shadow mask 15 is introduced to localize the metal deposition on the heated membrane 10, as shown in figure 7a where the mask 15 is placed in the incident metal flux 16.

Experimental results verified that the heated stencil's life-time with thermal confinement increases significantly comparing to the case without thermal confinement, as shown in figure 7b. The thermal control could also be achieved by other means, for example, by perforation on the membrane around the heating area to reduce thermal conduction or by freestanding the central membrane through cantilever-like structures for limiting the thermal conductive path. In summary, according to the present invention a novel stencil concept with an integrated microhotplate on the stencil membrane to prevent aperture clogging is disclosed. The stencil can be locally heated up by the integrated resistive heater, or other equivalent means, in order to minimize material condensation. To prove the concept, Al deposition through the heated and non-heated stencils was performed. No clogging was found on the heated membrane whereas a clear shrinkage of the aperture was observed on the non-heated one. This method could significantly extend the life time of the stencil, especially for thick layers and in the dynamic mode.

The examples and embodiments given in the present application are of course only for illustrative purposes and should not be considered in a limiting fashion. Other variants using equivalent means are of course possible as well without imparting from the spirit a scope of the present invention. In particular, other equivalent materials and dimensions may be envisaged in the frame of the present invention. For example the invention could use other means than the disclosed Pt coils to heat the stencil. Also, any material may be unclogged with the present method and the temperature may be chosen according to the circumstances and may be anything above room temperature.

REFERENCES (all incorporated by reference in the present application)

[1] M. A. F. van den Boogaart, et al., "Corrugated membranes for improved pattern definition with micro/nanostencil lithography", Sensors and Actuators A, vol. 130-

131 , p. 568-574, 2006.

[2] V. Savu, et al., "Dynamic stencil lithography on full wafer scale", Journal of

Vacuum Science and Technology B, 26(6), 2008.

[3] K. Sidler ef a/., Sensors and Actuators A 162, 155 (2010).

[4] D. S. Engstrom et al., Nano Lett. 11 , 1568 (201 1 )

[5] M. Lishchynska, et al., "Predicting Mask Distortion, Clogging and Pattern Transfer for Stencil Lithography", Microelectronic Engineering, 84 (2007) 42-53.

[6] O. Vazquez-Mena, et al., "Reusability of Nanostencils for the Patterning of

Aluminum Nanostructures by Selective Wet Etching", Microelectronic Engineering, 85 (2008) 1237-1240.

[7] M. Kolbel, et al., "Shadow-Mask Evaporation through Monolayer-Modified

Nanostencils", Nano Letters, 2002, 2 (12), 1339-1343.

[8] J. E. Mahan, "Physical Vapor Deposition of Thin Films", John Wiley and Sons,

2000.

[9] D. Briand, et al., "Design and fabrication of high-temperature micro-hotplates for drop-coated gas sensors", Sensors and Actuators B, vol. 68, p. 223-233, 2000.

[10] R.M. Tiggelaar, "Silicon-based microreactors for high-temperature heterogeneous partial oxidation reactions", Ph.D. dissertation, Univ. of Twente, Enschede, The Netherlands, 2004.

Claims

1. A method for nanostencil lithography of evaporated material on a substrate with a stencil membrane with apertures wherein said stencil is provided with heating means on the stencil membrane to prevent aperture clogging, whereby the stencil can be locally heated up in order to minimize material condensation and remove clogging.

2. The method as defined in claim 1 , wherein the local heating deforms the membrane towards or away from said substrate to modify a gap between said membrane and said substrate thereby influencing the resolution of the deposited structure on the substrate.

3. The method as defined in claim 1 or 2, wherein the stencil is heated up to 800°C.

4. The method as defined in one of the preceding claims, wherein the heating means are electrodes powered by a DC source.

5. The method as defined in claim 4, wherein a constant voltage is applied during the evaporation process and the current is monitored.

6. The method as defined in claim 4 or 5, wherein the temperature is maintained constant during the evaporation process by increase of the power of the source.

7. The method as defined in one of the preceding claims, wherein the lithography is made in static mode where the stencil is fixed with respect to the substrate or in dynamic mode where the stencil moves with respect to the substrate.

8. A method of fabricating a stencil for use in one of the preceding method, said method comprising the following steps:

-) providing a stencil substrate;

-) coating said substrate with a first low stress film material,

-) forming resistive heating coils with electrically conductive materials;

-) depositing a second low stress film material,

-) forming contact pads and stencil aperture;

-) forming a free standing membrane.

9. The method as defined in claim 8, wherein the stencil substrate is a Si substrate.

10. The method as defined in claim 8 or 9, wherein the low stress film material is a LPCVD SiN having a thickness of at least 200nm.

1 1 . The method as defined in one of claims 8 to 10, wherein the heating coils are made of a layer of 5nm of Ta and a layer of 200nm of Pt.

12. The method as defined in one of claims 8 to 1 1 , wherein the heating coils are made by lithography and lift off process.

13. The method as defined in one of claims 8 to 12, wherein the contact pads and stencil aperture are made by dry etching followed by a DRIE etching to open the membrane window.

14. The method as defined in one of claims 8 to 13, wherein the free standing membrane is released by a wet etching.

15. A stencil fabricated by the method as defined in one of claims 8 to 14.

16. A device using the method as defined in one of claim 1 to 7 with at least a stencil as defined in claim 15.

17. A stencil for nanostencil lithography of evaporated material on a substrate, comprising a stencil membrane (10) with apertures (6) wherein said stencil membrane further comprises heating means (3) to prevent aperture clogging, whereby the stencil can be locally heated up in order to minimize material condensation and remove clogging.

18. The stencil as defined in claim 17, whereby the membrane is deformed by the heating towards or away from said substrate to modify a gap between said membrane and said substrate thereby influencing the resolution of the deposited structure on the substrate.

19. The stencil as defined in claim 17 or 18, wherein it comprises at least a Si substrate with two layers of low stress thin film material and Pt resisting coils as heating means and a Ta layer as an adhesion layer for the coils.

20. The stencil as defined in claim 19, wherein the thin film material has a thickness of at least 200nm.

21 . The stencil as defined in claim 19 or 20, wherein the Pt coil has a thickness of about 200nm.

22. The stencil as defined in one of claims 19 to 21 , wherein the Ta layer has thickness of about 5nm.

23. The stencil as defined in one of claims 17 to 22, wherein it comprises a shadow mask (15) to localize the metal deposition on the heated membrane (10) and extend the life-time of the heated stencil (10).