Micro and Nano fabrication platform
Micro- and nano-fabrication is at the heart of research and development of the Lab’s activities: magnetotransport measurements, in the DNA of the Laboratoire Albert Fert, magnetic imaging, spintronic, superconducting tunnel junctions, etc…
The various instruments available are described briefly below, with a more detailed focus on microfabrication using synchrotron-LIGA radiation, developed and run by Fayçal Bouamrane (IR CNRS).
In addition to the specific instruments listed below, the clean room shared with other Thales RT units and groups is equipped with the usual tools for lithography. Various other tools are available (after training), such as optical microscopy, profilometry, interferometry, reactive plasma etching, metal deposition by evaporation, etc…
UV Lithography
— Coordinators: S. Collin, S. Mesoraca
Two mask aligners (MJB4 and MJB3 from SUSS MicroTec) are available in our cleanroom for high-precision UV photolithography (> 0.5 µm) on standard-size wafers/substrates as well as irregular shapes and various thicknesses.
Laser lithography
— Coordinator: S. Collin
The laboratory is equipped with direct-write laser lithography for device manufacturing. With a resolution of around 1 µm, this widely-used equipment can be used to produce a large variety of devices (no physical mask is required), while remaining highly ergonomic and open to all trained users.
Smartprint lithography
— Coordinators: J. Trastoy, N. Reyren, J. Briatico
In mid-2019, we acquired the first version of a UV projection lithography instrument (contactless, in other words) from the French start-up SmartForce, the Smart Print. The principle is simple: a modified projector (the projection lens is replaced by a microscope lens) projects “fullHD” (blue) images of photolithographic masks onto a surface area of a few mm² (depending on the lens chosen). When the pattern exceeds the size of a field, the machine can connect various fields, albeit with limitations in terms of the precision of the connections (1-2 µm). Given the enormous success of the machine, a second has been acquired (delivered in March 2022), the SP-UV Microlight3D with a better sample platform (improving the accuracy of field splices (< 200 nm), and using a more powerful UV diode source and enabling resins sensitive to the i, h and g lines to be exposed. The “real” lateral resolution is 1 to 2 µm depending on the geometry.
Electronic lithography
— Coordinators: S. Collin, S. Mesoraca
Our laboratory is equipped with a Raith PIONEER Two: an electron beam lithography (EBL) system integrated with a scanning electron microscope (SEM) with a maximum accelerating voltage of 30 kV. It is also equipped with a very high-precision laser interferometer-controlled stage and backscattered in-lens secondary electron detectors (AsB).
Ion bombardment etching (IBE)
— Coordinator: S. Mesoraca
We have a PLASSYS MU600S ion etching system. It operates with Ar ions accelerated between 200 and 700 eV for samples up to 4 inches in diameter with a planetary substrate holder, tilted (0-90°) and cooled (> 2°C). Thanks to the associated mass spectroscopy (SIMS, Hidden Analytical), it is possible to stop etching in specific layers as thin as a few nm. The presence of an in-situ sputtering gun enables a layer to be deposited on the freshly etched area.
Microfabrication with synchrotron radiation – LIGA
— Coordinator: F. Bouamrane
The production of microstructures sometimes requires the ability to combine micrometric spatial resolutions and structure heights of several tens or even hundreds of micrometres. Conventional technologies (electronic writing, laser lithography, UV, etc.) do not allow the following properties to be obtained simultaneously: large depth of field, vertical walls, large accessible heights (> a few 10 µm to a few 100 µm), high aspect ratio (>10). The LIGA process (German acronym for the three successive stages: lithography, electroforming and moulding), using X-rays from synchrotron radiation, benefits from two unique properties: an extremely intense source and very low divergence, which will be used to achieve this unique micro-structuring.
What we have: An X-ray lithography line has been installed on SOLEIL’s Metrology beamline since 2009. It is used for the first stage of the LIGA, X-ray lithography. This beamline, which has undergone several upgrades since it was first installed, offers the unique possibility of working on the same equipment with a white beam (all energies from 3 to 45 keV) and working in monochromatic mode (choice of a specific energy between 3 and 45 keV). For this purpose, it has a cooled beryllium double window with a thickness of 200 µm. Thanks to the high photon flux, the white beam mode enables thick resins (up to mm) to be exposed in a short time (around one hour), but lateral resolutions are limited to a few µm. In monochromatic beam mode, depending on the energy selected, grey-scale lithography or sub-micrometre resolution can be achieved. However, the thicknesses of the resins to be exposed are thinner (a few tens of µm maximum), and the photon flux is lower, requiring relatively long exposure times (around ten hours depending on the sensitivity of the resins). The beam sizes currently available are 65 mm wide (horizontally), which with the scanner’s scanning mode allows vertical exposure over a height of 100 mm, i.e. a sample with a diameter of 3 inches as standard, or even 4 inches under certain conditions.
The lithography station will be upgraded between 2018 and 2020 in order to modernize it (vacuum, precision displacement and thermalization of the sample, etc.) and to be able to pilot the entire lithography procedure completely autonomously using the SOLEIL control-command protocols, while ensuring optimum safety for the experiments.
Expertise: LIGA technology requires the development and manufacture of specific masks for X-ray lithography. These masks are not commercial products and must be custom-made.
A LIGA mask must meet the following specifications:
- made of a rigid material, a membrane, which is flat, with controlled roughness (ideally, a Ra of a few hundred nm) and with two parallel faces over an area of several cm2. In addition, the material must be transparent and not present any danger or toxicity to users. These are materials with a low atomic number (typically based on carbon, silicon, etc.). The type of material and its specific characteristics (thickness, rigidity, roughness, etc.) will be chosen according to the production conditions: energy and intensity of the X-rays, resolution required, etc.
- this mask will be covered with patterns that absorb the X-rays used, while having variable geometries with centimetric to micrometric dimensions and themselves structured with details that are sometimes micrometric or even sub-micrometric. These patterns will be made of materials with a high atomic number (platinum, gold, lead, etc.) and will not present any particular risks for their use. Their thickness will be chosen to allow sufficient absorption of X-rays to ensure adequate contrast between the transparent and absorbing areas, so that the patterns can be reproduced accurately in all their detail. For the LIGA process, it is mainly the resolution of these patterns that will ensure the resolution of the objects produced. Typically, with a white beam at SOLEIL, the thickness of gold required varies between 15 and 25 µm, depending on the results required. The technology of choice for producing these patterns is the electro-growth of materials such as gold, after a UV lithography procedure in a 20 to 30 µm thick resin. The resolution (smallest pattern size) of a standard mask is 5 µm. Typically, we produce LIGA masks based on graphite membranes 300 µm thick or glass membranes 20-50 µm thick (i.e. optically transparent for multi-level alignment).
The thicknesses of the resins used, and this is a specific feature of the LIGA, enable lithography to be carried out to a depth of a few dozen to a few hundred micrometers, or even mm, which also requires special know-how to prepare them.
There are three processes:
- a conventional spin-coating process using commercial resins, but limited to thicknesses of a few tens of micrometres for some resins (epoxy SU-8, BPN, AZ15nXT, etc.).
- a process of casting resin in solution onto a heated bench, of which can be precisely adjusted (flatness and horizontality corrected with a bubble level to an accuracy of around 0.1 mm/m) to ensure even distribution of the deposit on the substrate. This process enables resin thicknesses of a few tens to a few hundreds of micrometres to be achieved, but requires resin solutions with a viscosity suited to the targeted thicknesses (SU-8 epoxy, BPN, AZ15nXT, commercial PMMA solution, etc.).
- finally, a casting process which consists of mixing the components of a polymer, in this case PMMA (poly-methyl methacrylate), and casting it directly onto a substrate which has been previously prepared to receive this layer in order to locate the deposit and roughly define its thickness (set of wedges, etc.). The polymerisation reaction can be chosen so that it takes place at room temperature or, on the contrary, triggered and thermally controlled in an oven depending on the properties sought in the final polymer (cross-linking or not of the polymer chains, internal mechanical stresses, quality of adhesion, etc.). A polishing operation may be necessary to ensure a uniform thickness of resin on the substrate. A dedicated polishing bench and a specific polishing head are used for plane and parallel grinding. This process can achieve resin thicknesses ranging from a few tens of micrometres to a few millimetres, but is essentially suited to PMMA, which is LIGA’s resin of choice.
Where we want to go from here: while seeking to maintain lateral resolutions of a few hundred nanometres, the aim would be to take advantage of a broadband low-energy X-ray photon source (around 2-3.5 keV), i.e. to achieve a photon flux that is one or even 2 orders of magnitude greater than the photon flux from a standard silicon crystal monochromator (111). This would make it possible to significantly reduce exposure times and to consider manufacturing microstructures with an exposure time of around one hour, thus moving from a study process to an optimised process for the micro-fabrication of a larger number of pieces.
The fabrication of high-resolution LIGA masks consisting of a membrane transparent to low-energy X-rays (diamond, Si3N4-SiO2 or SiC) with a maximum thickness of a few µm and apertures of a few cm2. These membranes must be strong enough to support the gold structures that make up the absorbing patterns and perfectly flat to ensure the desired resolution. These masks are not commercially available and require the use of precision lithography to produce the structures and gold electro-growth of a few µm. Collaborations as part of the Renatech network (CNRS/Thales lab-LAAS, D. Bourrier) were carried out from 2014 to 2018 using a UV masker and then a UV stepper and demonstrated the feasibility and interest of this process on Si3N4-SiO2. In the future, the masks will have to be structured using electronic writing (resolution better than 100 nm).
Research and development projects: LIGA’s activities in recent years have focused on two major topics: research into dielectric metamaterials for Terahertz (THz) and technological development in X-ray lithography at SOLEIL.
* Imaging and spectroscopy in the THz range has great potential for applications in the medical and security fields. To this end, we have exploited the unique possibilities offered by metamaterials (artificial materials obtained by structuring) in order to obtain electromagnetic properties that do not exist in their natural state. During the TeraMetaDiel project (ANR, 2013-2016), we had exploited the possibility of generating in a dielectric a resonant permeability and effective permittivity. The choice of material had been SrTiO3 (STO). This experience was put to good use during the TeraCerNuT project (MITI, 2020 and 2021) and then the DisPoNT project (ASTRID, 2022-25), and the choice then fell on TiO2, a material with less absorption in the THz range. Micro-fabrication using synchrotron radiation is particularly suitable given the constraints imposed (tolerance, positioning accuracy, geometry). PMMA moulds obtained using LIGA technology with micrometric resolution (photo left) can be used for micro-moulding using ceramic powder (IRCER). ICP (Induced Coupled Plasma) processing of quartz substrates, in collaboration with TRT, was used to carry out powder sintering tests using the SPS (Spark Plasma Sintering) method and conventional sintering (IRCER).
* Interest in low-energy X-rays (2-3.5 keV) would benefit from significant development, but this field currently suffers from a lack of technological possibilities for selecting an energy with sufficient photon flux. Conventional VUV gratings and single crystals (Bragg reflection) are inefficient at these energies. The creation of a high-performance array would be a first building block for development in this field, particularly using synchrotron radiation. The two major difficulties are antinomy: the dimensions of the grating depend on the illumination conditions (a few cm on synchrotron radiation) and the desired resolutions require nanometric dimensions of the structures on millimeter fields. Moreover, the thermomechanical stability of these optics also need to be taken into account, given the measurement performance targets. For this collaboration with SOLEIL (P. Mercère) started in 2018, the idea was to launch the development of a monochromator made up of a broadband diffractive multilayer grating on the METROLOGIE beamline while also being able to have an application in X-ray lithography (duplication of masks, etc.). A series of grating tests was carried out to see whether it would be possible to meet these requirements using e-beam lithography (TRT 2019 and 2020 collaboration), followed by a cobalt growth test using sputtering and then lift-off to check feasibility (S. Collin). Following these promising initial results for a grating at 2000 lines/mm with PMMA (Photo on the left, field connection zone), we decided to push the limits of electronic lithography for gratings with resolutions of 3000 and 4000 lines/mm on millimetre-scale surfaces.
