Accueil • M@CSEE

M@CSEE

Microéléctronique Composants Systèmes efficacité énergétique

Electrical characterization of microelectronic components: Low frequency noise

The electrical characterization of microelectronic components has been a research specialty of the laboratory for many years, in particular the characterization of low frequency noise. After having participated in the European project RF2THZ-SiSOC in a consortium labelled by the CATRENE organization, the team has been involved until 2021 in the European project TARANTO co-financed by an ECSEL JU H2020 program and by the Ministry of Economy, Finance and Industry. In this project our team is in charge of the characterization and the modeling of the Low Frequency Noise in Si/SiGe:C heterojunction bipolar transistors associated to two BiCMOS technologies 130 nm and 55 nm developed by STMicrolectronics and whose frequency performances exceed 300 GHz. The main objectives are compact modeling of low frequency noise including reliability studies under electrical stress and irradiation. We are also able to perform temperature characterizations from 100 K to 400 K The main results that we can highlight are the following:

The complete characterization of the noise measured in input (SIb), high impedance mounting, with the classical SPICE modeling leading us to extremely low KB values: best result published to date 7 10-11 µm². The same modeling was conducted under the statistical aspect considering the noise dispersion in 1/f on a whole plate and the presence of generation-recombination components and RTS noise
The complete study of the noise measured in output (SIc), low impedance assembly, with an associated SPICE model
The confirmation that the 1/f noise measured at the input was the image of the spontaneous fluctuations of the base current and, more innovative, that the 1/f noise measured at the output was only related to the fluctuations of the collector current
Measurements of the input LF noise at very low temperatures up to 100 K, which allowed us to extract two trap activation energies: 116 and 185 meV
The best X and gamma-ray robustness of the BiCMOS 55 nm HBTs compared to the BiCMOS 130 nm ones [RADECS 2019]

Energy Harvesting for sensors application

The Energy Harvesting team mainly focuses on 2 topics briefly described below:

Indoor PV: The development of the Internet of Things (IoT) or Wireless Sensor Networks (WSNs) applications is growing significantly. Reaching the energetically autonomy of the associated sensors remains a challenge. On one hand, a lot is done to reduce their power consumption, and on the other hand, a growing community of researchers works on improving the technologies able to harvest enough power from the nearby environment to supply electrical energy to such devices. One of the most used technologies, in indoor condition, is the photovoltaic technology. Since Indoor light, usually composed of several different light sources (artificial and natural, direct and reflection) conditions, has no standards yet, establishing the level of potential harvestable energy in those indoor conditions is still very challenging. In this context, we have developed several tools based on very low-cost commercial photodiodes and on supervised machine learning to estimate the potential harvestable energy in real environments. There are several challenges:
1. The luminosity in lux is not a reliable quantification of the incident harvestable power. Indeed, FIG 1A shows the spectra obtained @ 1000 lux for natural light through a window, a compact fluorescent light (CFL), and a light-emitting diode (LED). For the same lux value, it gives 3 very different levels of irradiance.
2. The indoor light is a time-varying mixture of multiple natural and artificial direct, reflective, and scattered sources. It can be seen in the video of the FIG 1B, where the spectra are captured over the day.
3. For real application, we still have to deal with non-ideal devices connected to a non-ideal electrical storage device via a non-ideal power management integrated circuit (PMIC) devices. It means that the electrical energy received by the final consumer device is technology dependent and is far from the standard theoretical Shockley- Queisser (SQ) limit model predictions (Shockley and Queisser, 1961). As an example, in FIG 1C is shown the theoretical efficiency that should be obtained for a monojunction solar cell for different light monosources and the real measured efficiency of some commercial solar cells. This gap between theory and reality should be considered in the models.
In this context, we are developing 2 kinds of objects:
1. Simple and instrumented Indoor PV energy harvester prototypes to supply enough energy to commercial medium power (10 mW) consumption sensor or electronic devices.
2. Methods and models to estimate the potential harvestable energy in any location of an indoor environment, based on classical spectrometers or based on extremely low-cost systems of light sensors (FIG 3) associated with a machine learning system. This later method allows to deploy on a greater scale to scan the light energy harvesting of a building in many places at the time, on long period of study, without assistance.
Apart from designing and manufacturing energy harvesters, our main goal is to optimize the tools and models that allow sizing properly, with the appropriate technology, any Indoor PV harvester in any location. Thanks to the instrumented PV harvester prototypes we have developed, we can compare the estimation of harvestable energy from our model and the real harvested energy. This result, published in Solar Energy [1] has been obtained using ‘simple’ spectrometers while the same results obtained with the very low-cost light sensor associated to machine learning algorithms (see FIG.5 for the concept) can be seen in the manuscript of B. Politi [2] who had defended its PhD tesis in March 2021. References: [1] B. Politi et al., “Practical PV energy harvesting under real indoor lighting conditions,” Sol. Energy, vol. 224, pp. 3–9, Aug. 2021, doi: 10.1016/j.solener.2021.05.084. [2] B. Politi, “Systèmes d’analyse de la récupération d’énergie lumineuse en intérieur pour l’alimentation d’objets connectés.,” https://tel.archives-ouvertes.fr/tel-03414940, Mar. 2021. + une video (fichier mp4 à disposition) si possible à mettre en ligne
Thermophotovoltaic conversion: Harvesting heat energy using photovoltaics is a field of research in the midst of a revival. Thermophotovoltaic (TPV) conversion consists in converting radiant heat energy from hot sources (at 600 °C and more) into electrical energy using photovoltaic cells. The thermal energy may come from combustion, nuclear reactions (for spatial applications), waste heat sources (high-temperature industrial processes) and concentrated solar radiation using an intermediate radiation absorber. Thermal energy storage can be used for overcoming the issue of intermittent supply of wind and direct solar generated electricity.

In this context, the Institute and its partners are involved in multiple actions: Ongoing:

Design, fabrication and characterization of low-bandgap (< 0.36 eV) thermophotovoltaic cells, in the frame of a project funded by the ANR (LOW-GAP-TPV, 11/2021-10/2025) involving the group NanoMIR of the Institute, the group TNR of Institut Pprime, and the group OR2T of CEMHTI.
Leading contributor of the project-team TREE on thermal radiation to electrical energy converters, gathering 5 French laboratories collaborating in this field.
Guest editor of a Special Issue on thermophotovoltaics in Solar Energy Materials and Solar Cells (Elsevier): the articles are under review and accepted manuscripts will be available online soon.
Contribution to the organisation of the 13th World Conference on Thermophotovoltaic Generation of Electricity (TPV-13).

Recent:

design, fabrication and implementation of a near-field thermophotovoltaic (NF-TPV) converter, in the frame of a project funded by the ANR (DEMO-NFR-TPV, 11/2016-10/2020) involving the group NanoMIR of the Institute and the group MiNT of the Centre for Energy and Thermal Science of Lyon.
contribution to the development of other NF-TPV converters, in the frame of a continuous collaboration with the University of Utah, USA (Radiative Energy Transfer Laboratory).
contribution to proposing a new concept of solid-state thermal-to-electrical energy converter, called thermionic-enhanced near-field thermophovoltaic converter, in the frame of a collaboration with the Instituto de Energia Solar, Universidad Politécnica de Madrid, Spain (group SyNC).
organisation of a Symposium “Advances in thermophotovoltaics: materials, devices and systems” at the E-MRS Spring meeting 2021.

For further information on this topic:

chapter on TPV energy conversion in a recent book on Ultra-high temperature thermal energy storage, transfer and conversion.
article on the principles and prospects of thermophotovoltaics (in French).
article with results showing near-field TPV conversion with record 14-20% efficiency and 0.75 W cm-2 electrical power density.
article revisiting the thermal behavior of TPV devices.
article demonstrating excellent performances of InSb PV cells designed for NF-TPV converters.

III-Sb multijunction solar cells:

Currently, “at 1 sun”, the best efficiencies of 3- to 6-junction solar cells are about 38-39 % whereas they reach 44 to 47% under concentrated light [1]. The number of junctions tends to increase over the years, always improving the solar spectrum harvesting. Fig. 1.a. presents the optimum bandgaps as a function of the number of subcells. We can observe that the fabrication of multi-junction with 4-junction and more requires materials with a wide range of bandgap values and particularly the specific value of 0.5 eV. As we can see on the cartography of the bandgap energies as a function of the lattice constant (Fig. 1.b.), it is difficult to find materials among all III-V and II-VI SC that are able to fill this requirement and especially if we would like to stay lattice-matched to standard wafers such as Ge, GaAs and InP. Our work aims at exploring a new strategy based on III-Sb alloys lattice-matched to GaSb. In fact, quaternary alloys matched to GaSb can address this concern (Fig. 2.) [2]. The 0.5 eV bandgap can be reached with the GaInAsSb alloy. Medium bandgaps from 0.726 to 1.64 eV can be achieved with GaSb and AlGaAsSb alloys and higher bandgaps with a II-VI quaternary matched to GaSb. We can also notice that the AlInAsSb alloy allows to cover a broad bandgaps range but its growth is not easy due to miscibility issues. We are able to design, fabricate and characterize III-Sb solar cells optimized (see Fig. 3.) [3]. In the first step, the cell architecture is optimized with the use of our in-house developed solvers. Based on those optimizations, the structure is grown by MBE (Molecular Beam Epitaxy) in strong collaboration with the NanoMIR team (IES). Then, we proceed with the cell fabrication in the cleanroom with dedicated equipments for antimonides. The as-fabricated solar cells are characterized and analyzed with numerical simulations. Latest results:

Record efficiency of GaSb single-junction solar cell (7.2 % “at 1 sun”) [3].
Elaboration of AlGaAsSb single-junction solar cells [2].
Elaboration of III-Sb tandem cells [4].
Development of a material library of III-Sb quaternary alloys for numerical simulations [2].
Development of a numerical 1D solver to simulate the single and multi-junction solar cells [4]. read more…
Development of a pseudo-3D solver to simulate the solar cells behavior under concentrated light.

References: [1] M.A. Green et al. « Solar cell efficiency tables (Version 53) », PIP 2019; 27:3-12. [2] Parola S, Vauthelin A, Martinez F, Tournet J, El Husseini J, Kret J, Quesnel E, Rouillard Y, Tournié E, Cuminal Y, “Investigation of antimonide-based semiconductors for high-efficiency multi-junction solar cells”, Proceedings of 7th World Conference on Photovoltaic Energy Conversion (WCPEC), 0937-0942 (2018). download link [3] Parola S, Vauthelin A, Tournet J, El Husseini J, Martinez F, Rouillard Y, Tournié E, Cuminal Y, “Improved efficiency of GaSb solar cells using an Al0.50Ga0.50As0.04Sb0.96 window layer”, Solar Energy Materials and Solar Cells 200 (2019) 110042. download link [4] Kret J, Tournet J, Parola S, Martinez F, Chemisana D, Morin R, de la Mata M, Fernandez-Delgado N, Khan AA, Molina SI, Rouillard Y, Tournié E, Cuminal Y, “Investigation of AlInAsSb/GaSb tandem cells – A first step towards GaSb-based multi-junction solar cells”, Solar Energy Materials and Solar Cells 219 (2021) 110795. download link

Silicon solar cells:

The work is developed on this topic aims to contribute to reduce the cost of the photovoltaic kW/h by using:

the reduction of the cost of existing PV technologies by adapting new manufacturing processes (dry texturization, electrodeposition …) to industrial development,
the adaptation of cells and materials for an application under solar concentration.

The M@CSEE team is collaborating for many years through CIFRE contracts and PhD thesis with cell manufacturers and equipment manufacturers. The research topics have in common the will to contribute to the development of industrial processes for the manufacture of silicon photovoltaic cells, in particular for an application under low-concentration solar flux (20x).

1. Solar cells based on multicrystalline silicon:

Over the last few years, we have participated in several projects with the following objectives:

The development of an industrial process for boron doping of multi-crystalline silicon,
The development of a new process for dry texturing of cells,
The development of front contacts by electrodeposition.

The developed skills in this topic are related to the modeling of industrial processes, the fine characterization and associated modeling of electrical contacts, the modeling of electrical transport in crystalline, microcrystalline and amorphous semiconductors.

2. Solar cells based on nano-crystalline silicon (nc-Si):

The work carried out in this field concerns:

the fabrication of micro/nanocrystalline silicon thin films (by PECVD and ink printing): (picture)
the modeling the PECVD growth of micro/nanocrystalline silicon: (picture)
the characterization and modeling of electrical transport in nanocrystalline materials: (picture)