Budapest University of Technology and Economics Faculty of Electrical Engineering and Informatics
CMOS compatible capacitive humidity sensor
Ph.D. Thesis Booklet
Author: László Juhász
Advisor: Prof. Dr. János Mizsei, D.Sc.
Department of Electron Devices
Budapest, 2013.
1 Introduction
Humidity is an important factor to consider in certain industrial and agri- cultural processes, in the preservation of objects of our art heritage and in other numerous areas of our lives. In many cases, it is important to measure and control air humidity. There are a lot of different operating principles and transduction techniques to measure quantities that are related to it.
The various areas of applications demand different sensors with different specifications give room to research and development of new layers, struc- tures and devices. Besides the improvement of precision, stability, response time and life-time there’s strong emphasis on the reduction of size, power consumption and price too.
The adsorption and absorption of water vapor may result in many different physical phenomena. It has influence on the dielectric properties of certain materials which is, for example the principle of operation in case of capacitive humidity sensors. The electric (DC) conduction may be altered due to a change in ionic conduction. The adsorption of water on semiconductor surfaces may alter the band-bending (barrier effect) altering both AC and DC conductance. In MEMS1devices it may increase the mass of movable parts or may change the mechanical stress in thin films of membranes. In SAW2 devices the phase velocity of the surface acoustic waves may be altered by surface-bonded water molecules.
As the sensing layer is an essential part of the mentioned devices it is subject to active research. Wide range of (inorganic) insulator and semi- conductor materials, organic polymers, composites made from these and nanostuctures are being investigated for this purpose. In many cases, porous materials are used. Currently, the most widely used group of materials are polymers. In capacitive sensors poly-imide is the most frequently used as sensing layer. The life-time of these sensors is limited due to fact that their structure is permanently altered by moisture. Porous metal oxides having large surface-to-volume ratio are advantageous for sensors with high sensitivity. One of these materials is alumina produced by anodic oxida- tion of aluminum under anodic bias. Its properties can be controlled by the anodizing parameters and the prepared layer remains stable at high temperatures.
1Micro-Electro-Mechanical System
2Surface Acoustic Wave
2 Aims and objectives
The state of the art in the case of humidity sensors is the realization in microsystem form. According to this demand my goal was to work out thin film forming processes, layer-structures and a sensor device that are compatible with CMOS3 integrated circuits’ high volume manufacturing processes.
Literature studies have shown that porous alumina is a promising can- didate for the purpose of a highly sensitive and stable integrated sensing layer. The use of this layer suggests capacitive layer-structure that fits well into the concept of integrated CMOS electronics and could result in a more stable behavior due to the very low temperature dependence of the dielectric constant on temperature.
My objectives required the development of a process that could be used to form porous aluminasensing layerfrom the aluminum thin film metalization of classic CMOS technology. Besides reliable thin film forming methods, capacitive, CMOS compatiblelayer-structureswere needed. These had to be sensitive enough to be measured with integrated electronics despite of the given small available sensing area (maximum few mm2). Based on such sensing layers and layer-structures, I aimed at the realization of a sensor devicewith heater and thermometer functions integrated.
3 Applied tools and inspection methodologies
The infrastructural background of my work was provided by the Semicon- ductor Laboratory of Department of Electron Devices of BME, which was located in the Building V2. The laboratory was moved to the Building Q where it was completely rebuilt recently.
The processing technology I worked out for the realization of the hu- midity sensor device consists of 23 individual complex work phases. For the implementation I used all the equipment available in the laboratory (photolithography and wet processing toolsets, tube furnaces for the high temperature processes, vacuum evaporator, RH sputtering equipment, dic- ing and die-bonding equipment, wire-bonder) and my new experimental setup used for the special anodic oxidation step. In certain parts of my work
3Complementary Metal–Oxide–Semiconductor
I cooperated with other laboratories to broaden the range of available tools.
The fellows of MFA4 Microtechnology Department helped with evapora- tion, sputtering and photomask manufacturing, while the fellows of MISEC Laboratory at HWU5made available the infrastructure of their MEMS lab- oratory for layer-structure preparation (photolithography, electron-beam evaporation, electroplating).
I examined the prepared layers with several methods: mechanical pro- filometry for thickness measurement, optical microscopy for inspection of layer homogeneity and features realized with photolithography, scanning electron microscopy (SEM) for the investigation of the pores and porosity of the alumina and atomic force microscopy (AFM) for the determination of surface roughness.
The inspection of the layer-structures and the humidity sensor device mainly targeted the registration of sensitivity characteristics (capacitance as a function of relative humidity). For this purpose calibrated control of relative humidity was necessary. During my work both fixed-point (with saturated salt solutions) method and climate chambers were used. The mea- surement of the sensors’ capacitance (and parallel resistance) was performed with an L-C meter and later with a precision LCR meter. TheCp–Rpequiva- lent circuit and the measurement frequency was selected after preliminary investigations. In climate chambers, it was possible to study the effect of am- bient temperature on the characteristics. The sensor device with integrated heating and thermometer enabled the analysis of the effect of increased chip temperature and heated regeneration.
4 New scientific achievements
During my work, I worked out new CMOS compatible thin film forming processes and layer-structures for the purpose of integrated humidity sen- sors. Based on these, I realized a new CMOS compatible humidity sensor device. The results are presented on the following pages in the form of 3 theses and 7 sub-theses.
4Hungarian Academy of Sciences, Research Centre for Natural Sciences, Institute of Techni- cal Physics and Materials Science
5Heriot-Watt University, Edinburgh, UK
Thesis 1. CMOS compatible thin film forming processes for integrated humidity sensors
I worked out new processes whereby nanoporous thin film alumina can be created and shaped with CMOS compatible silicon wafer processes from deposited aluminum thin films for the purpose of integrated sensor device [J1], [C1–C4].
A key factor of the integration is the shaping of the created layer in order to fit to the layout of the integrated electronics. I used two different approaches:
the first was based on classical photolithography and wet chemical etching steps, while the second uses selective anodic oxidation (using masking) of the aluminum and a subsequent selective etching step to remove the remaining metal layer. In the first case an alumina etching process, in the second case an appropriate masking method was needed, both compatible with CMOS processing technology.
Anodic oxidation followed by etching of alumina
Selective anodic oxidation followed by (selective) etching of remaining aluminum
Si SiO2 Al Porous Al2O3 Photoresist
Figure 1: Two different approaches used for the shaping of alumina layers
Sub-thesis 1.1 I worked out a new process whereby nanoporous thin film alumina can be created on the entire surface of an oxidized silicon wafer and the film can be shaped with wet chemical etching. During the porous layer formation the electrochemically oxidized aluminum itself serves as electrical contact without the aid of additional conductive layers, using a band of phororesist for its protection along the wafer perimeter.
Sub-thesis 1.2 I worked out a new process whereby selective anodic oxidation is used for creating nanoporous thin film alumina in the desired shape. During the layer formation the photoresist protected parts of the aluminum serves as contact for the electrochemical oxidation of the exposed parts. The remaining aluminum can be removed in a subsequent selective etching step without further photolithographic steps. With the help of scanning electron micrographs, I’ve demonstrated that the selective etchant causes increase of the pore diameter and porosity of the alumina during the aluminum etching. Such change of the layer properties can improve the sensitivity of layer-structures and sensors devices based on it.
Figure 2 demonstrates the efficient protection of the photoresist film. The increase of the pore diameter and porosity can be seen on Figure 3.
200 nm
Figure 2: Three SEM micrographs displaying different parts of a single sample (InLens SE- detector, 5 kV). Pictures from left to right: area not protected by photoresist, area affected by under-oxidation and area protected by photoresist (no pores present).
50 nm
Figure 3: Effect of selective aluminum etchant on porous alumina layer. SEM micrographs (InLens SE-detector, 5 kV). The reference sample (left) and the sample with alumina exposed to the aluminum etchant during etching after selective anodic oxidation (right).
Thesis 2. CMOS compatible layer-structures for humidity sensors Based on the thin film forming processes of the Thesis 1, I worked out capacitive, parallel-plate humidity sensitive layer-structures using the mentioned nanoporous alumina thin films as sensing layer. I have demonstrated with measurements that the layer-structures are sensitive in 0–100% relative humidity (RH) range, their char- acteristics are non-linear and the obtained average sensitivity reaches 15 pF/RH%
(3,8 pF/RH% on average below 65% RH and 35,7 pF/RH% above it). This sensitiv- ity is one order of magnitude higher than the values found in product catalogs of discrete, off-the-shelf capacitive humidity sensors [J1, J2], [C1–C3].
Sub-thesis 2.1 Based on wet chemical etching, I realized a new layer-structure consisting of ultra-thin palladium/palladium oxide, porous alumina, silicon dioxide and heavily doped silicon. The structure can be realized with CMOS processing and additional steps, enabling the construction of integrated humidity sensors. I have demonstrated that the agglomeration of the ultra-thin electrode during its annealing plays an important role in becoming a humidity permeable layer.
Figure 4 displays the layers of the layer-structure described in Sub-thesis 2.1 separated from each other for a better overview.
a) b) c)
Si SiO2 Al Porous Al2O3 Pd over porous Al2O3
Figure 4: Layer-structure with ultra-thin upper electrode: a) top view and cross-section of the structure along the dashed line; b) functionally distinct parts drawn separated; c) effective sensing part of the structure. The size of the chip is 2×2 mm, the sensing area is 1 mm2. The layer thicknesses are not drawn to scale.
Sub-thesis 2.2 Based on selective anodic oxidation, I realized a new layer- structure consisting of a grid-shaped metalization, porous alumina and titanium.
The structure can be realized on the passivated surface of a silicon wafer already containing the CMOS electronics, also enabling the realization of integrated sensors.
The layer-structure described in Sub-thesis 2.2 can be seen on Figure 5. Its characteristics are shown on Figure 6.
a) b) c)
Si SiO2 Ti Al Porous Al2O3 Cu Au
Figure 5: Layer-structure with grid-shaped upper electrode: a) top view and cross-section of the structure along the dashed line; b) functionally distinct parts drawn separated; c) effective sensing part of the structure. The size of the chip is 2, 3×2, 3 mm, the sensing area is 1 mm2. The layer thicknesses are not drawn to scale.
0 10 20 30 40 50 60 70 80 90 100
0 500 1000 1500
Relative humidity [%]
Capacitance[pF]
Figure 6: Capacitance–RH characteristics measured with fixed-point method at 25°C. The dotted line displays the tendency of the distribution of measured points.
Thesis 3. CMOS compatible humidity sensor device
Based on the new processes and the results of the investigations, I realized a new humidity sensor device that can be integrated with CMOS compatible processing.
The sensor built from the presented layer-structure with palladium/palladium oxide can be heated and its temperature can be measured using a p-n junction formed below the capacitance. The device is suitable for packaging with methods used in volume production [J1–J3], [C1, C2].
The new humidity sensor device can be seen on Figure 7.
a) b) 2 mm c)
Figure 7: Bond-plan and photo of the humidity sensor device with integrated heater- thermometer diode (left). Sensor device with optimized layout for packaging (right).
Sub-thesis 3.1 I have demonstrated that the p-n junction is suitable for the reproducible heated regeneration of the device providing a method for the offset com- pensation of the capacitance. Due to the construction of the device the regeneration can be performed in situ.
Sub-thesis 3.2 I have demonstrated that the p-n junction is suitable for keeping the temperature of the device constant, and the hysteresis decreases with increasing temperature.
Sub-thesis 3.3 I have demonstrated with measurements that the typical response time of my sensors is less than 6s, confirming the competitiveness with widely-used polymer-based, commercially available humidity sensors (having typical response time of 8sor more)
Figure 8. displays sensor characteristics at different chip-temperatures. Ca- pacitance values at 0% RH confirm the reproducibility of the heated regen- eration. Decrease of the hysteresis (as described in Sub-thesis 3.2) can be seen on Figure 9 (characteristics are normalized for better comparison). The measurement of the sensor’s response time is shown on Figure 10.
0 10 20 30 40 50 60 70 80 90 100 150
200 250
Relative humidity [%]
Capacitance[pF]
25°C 40°C 50°C
Figure 8: Capacitance–humidity characteristics of a sensor sample with integrated heater- thermometer diode at different chip-temperatures while ambient temperature was set to 25°C
0 10 20 30 40 50 60 70 80 90 100
0 20 40 60 80 100
Relative humidity [%]
Capacitancechange[%]
25°C 29°C 35°C
Figure 9: Sensitivity characteristics and its hysteresis measured on a single sensor sample at different chip-temperatures while ambient temperature was set to 25°C
0 5 10 15 20 25 30 0
20 40 60 80 100
63%
τ
Time [s]
Finalvalue[%]
20 40 60 80
Relativehumidity[%]
Figure 10: The sensor’s response (RH%) as function of time calculated from the sensor’s characteristics and the registered capacitance changes during response time measurement
5 Application of results
The new processes and layer-structures are CMOS compatible and may serve as basis for sensors in volume production. The realized sensor device containing integrated heater-thermometer and its version optimized for industrial packaging process are further steps taken in this direction. The demonstrated sensitivity and the favorable dynamic behavior, the effect of heated regeneration and continuous heating are confirming the benefits of the concept and its practical applications. The research and development work was done in the frame of the successful PATENT FP6 EU, BelAmI German-Hungarian bilateral and the SE2A ENIAC EU projects contributing to their achievements. In the last phase, a read-out circuitry and demonstra- tion device was built up on printed circuit board level at our department. The elaboration of the integrated electronics’ concept has begun in preparation for a possible CMOS-MEMS demonstrator.
Related publications
Journal papers
[J1] V. Timár-Horváth, L. Juhász, A. Vass-Várnai, and G. Perlaky. Us- age of Porous Al2O3Layers for RH Sensing. Microsystem Technologies, 14(7):1081–1086, 2008.
[J2] L. Juhász and J. Mizsei. Humidity sensor structures with thin film porous alumina for on-chip integration. Thin Solid Films, 517(22):6198–
6201, 2009.
[J3] L. Juhász and J. Mizsei. A simple humidity sensor with thin film porous alumina and integrated heating. Procedia Engineering, 5:701–704, 2010.
Conference proceedings
[C1] V. Timár-Horváth, L. Juhász, A. Vass-Várnai, and G. Perlaky. Usage of Porous Al2O3Layers for RH Sensing. In K. Chakrabarty, B. Courtois, E. Jung, V. Kempe, and R. Laur, editors,Proceedings of the Symposium on Design, Test, Integration and Packaging of MEMS/MOEMS (DTIP’07), pages 372–376, Stresa, Italy, 25-27 April, 2007. EDA Publishing.
[C2] L. Juhász, A. Vass-Várnai, Cs. Dominkovics, and V. Timár-Horváth.
Porous Al2O3Layers for Capacitive RH Sensors. In K. Hadjiivanov, V. Valtchev, S. Mintova, and G. Vayssilov, editors, Advanced Micro- and Mesoporous Materials, volume 1 ofTopics in Chemistry and Material Science, pages 209–220, Sofia, 2008. Heron Press.
[C3] L. Juhász, A. Vass-Várnai, V. Timár-Horváth, M. P. Y. Desmulliez, and R. S. Dhariwal. Porous Alumina Based Capacitive MEMS RH Sensor. In Victor M. Bright, Tarik Bourouina, Bernard Courtois, Marc Desmulliez, Jean Michel Karam, and Gou-Jen Wang, editors,Collection of Papers Presented at the Symposium on Design, Test, Integration and Packaging of MEMS/MOEMS (DTIP’08), pages 381–385, Nice, France, 9-11 April, 2008. EDA Publishing.
[C4] L. Juhász, L. Oláh, and J. Mizsei. Patterning of Porous Alumina for Integrated Humidity Sensors. In Tarik Bourouina, Bernard Courtois, Reza Ghodssi, Jean Michel Karam, Aurelio Soma, and Hsiharng Yang, editors,Collection of Papers Presented at the Symposium on Design, Test, Integration and Packaging of MEMS/MOEMS (DTIP’09), pages 219–222, Rome, Italy, 1-3 April, 2009. EDA Publishing.
Further publications
[N1] P. Sági, B. Plesz, L. Juhász, and V. Timár-Horváth. Design of the Heliotrex Sun-Tracking System. In Péter Kiss, Ádám Székely, and Bálint Németh, editors, IYCE 2007. International Youth Conference on Energetics, pages 1–6, Paper 409., Budapest, Hungary, 31 May-2 June, 2007.
[N2] V. Timár-Horváth, B. Plesz, L. Juhász, and J. Mizsei. Education of Microelectronics Technology at the Department of Electron Devices of Budapest University of Technology and Economics. InProceedings of The 7th European Workshop on Microelectronics Education (EWME’08), pages 102–103, Budapest, Hungary, 28-30 May, 2008. EDA Publishing.
[N3] B. Plesz, L. Juhász, and J. Mizsei. Feasibility Study of a CMOS- Compatible Integrated Solar Photovoltaic Cell Array. In Bernard Courtois, Jean Michel Karam, Ryutaro Maeda, Pascal Nouet, Peter Schneider, and Hsiharng Yang, editors,Collection of Papers Presented at the Symposium on Design, Test, Integration and Packaging of MEMS/- MOEMS (DTIP’10), pages 403–406, Sevilla, Spain, 5-7 May 2010. EDA Publishing.
[N4] V. Horváth-Timár, J. Mizsei, and L. Juhász. A new concept in solar cell related education at the Department of Electron Devices of the Budapest University of Technology and Economics. In V. Benda, editor, Proceedings of the 5th International Workshop on Teaching in Photovoltaics (IWTPV’10), pages 30–33, Prague, Czech Republic, 25-26 March 2010.
[N5] Székely Vladimír, Kollár Ern˝o, Somlay Gergely, Szabó Péter Gábor, Juhász László, Rencz Márta, Vass-Várnai András. Statikus TIM teszter tervezése. Híradástechnika, LXVI(1):37–46, 2011.
