The aim of this Project is the development of novel surface structuring and improved thin film deposition techniques to be applied to small-scale production of advanced gas sensors. Since investment into research to develop new ideas is too expensive for a small firm, the four Research Partners of the consortium will do the time consuming part of the R&D by solving the scientific problems, thereby promoting the Industrial Partner to an advanced technical level appropriate for commercial production of the new generation of sensors. In the short term, the benefit to the Company will be a new product of high added value leading to an increasing market share, while the Research Partners will be trained for effective use of their equipment and knowledge in product oriented competitive research. In the long term, the expected benefits are new workplaces and a new generation of graduates, PhD students and junior research workers who are acquainted with product oriented R&D and motivated to contribute to the development of a modern industrial environment in the Partner Countries.
Artificial sensors are vital elements of both domestic life and industry. Although sensors of a great variety of types have become well established in industries, agriculture, medicine and many other areas, the development of new sensing techniques and elements proceeds at an unprecedented rate. During recent years sensor markets in industrialized countries have been growing at an average rate in excess of 10% per year, and there has been a correspondingly high level of investment in sensor research and development [P.T. Moseley and A.J. Crocker: Sensor Materials, IOP Publishing Ltd, Bristol & Philadelphia, 1996].
This activity is driven by various factors. On the one hand there is a growing concern for the protection of the environment, for the improvement of fuel economy and for the enhancement of safety and security. On the other hand the industrial environment has reached the level necessary for backing the production of highly sophisticated sensors. Electronic signal processing which provides readable information and the possibility of automatic control has been developed well in advance of the sensors that convert physical and chemical parameters to be sensed into a form in which it can be processed, controlled and stored.
The protection of the environment is a key factor in modernization of the industry in the NATO Partner Countries in general, and in Hungary and Romania in particular. As a consequence, there is an increasing demand for sensors and sensing technologies appearing as a purchasing power on the domestic sensor market. What is more important, entering the globally expanding field of sensor R&D and production is a real challenge for NATO Partner Countries, since the existing strong R&D potential can be exploited for product oriented prospective activities and the fabrication of these products containing high added value (mainly by flexible small or medium sized enterprises like MicroVacuum, the Industrial Partner in this Proposal) will be a driving force of modernisation.
Integrated Optical (IO) sensors
Among the large number of sensor structures which are already used
or under investigation for environmental sensing applications, monolithic
Integrated Optical (IO) sensors are one of the most promising, since
they are relatively cheap though extremely sensitive. The key element in
IO sensors is the planar waveguide consisting of a thin dielectric layer
on a substrate. The incoupled light is guided in this layer by total internal
reflection. The field of the guided wave penetrates into the adlayer formed
by adsorption (or binding) of molecules from the ambient as an evanescent
wave and it is the corresponding evanescent field that senses changes in
the optical properties (notably: refractive index distribution) of this
adlayer.
Since thin film waveguides are not self supporting, the coupling of light into (or out of) the waveguide is not trivial. One straightforward solution is the use of optical gratings on the surface of the thin film as a surface relief grating. The waveguide and the coupler(s) form the grating coupler (waveguide) sensor chip.
The optical grating-coupler waveguide sensor effects are well documented in scientific literature. Recent activities have been reviewed by W. Lukosz and R.E. Kunz, in Sensors & Actuators B29 (1995) p. 37 and B11 (1993) p. 167, respectively. The measurement of changes in the effective refractive indices (N) of the guided modes provides a very effective means for sensing and quantifying the presence of a particular analyte at the waveguide surface (schematic representation of the sensing process is presented in two animated gif files ). Changes in thickness of the adlayer can be detected with less than 0.1 nm resolution. Minimum detectable changes in surface coverage expressed in terms of mass are of the order of some ng/cm2 [K. Tiefenthaler in Advances in Biosensors, Vol. 2, pp. 261-289, (1992)].
The key advantage of IO sensor chip fabrication is that all components can be integrated within one monolithic module using a single technology of few compatible processes. Planar waveguide technology started in the 60's to meet the needs of the semiconductor industry to integrate optical components for semiconductor applications, and matured due to progress in photolithography, thin film processing, miniaturised light sources (lasers), and photodetectors. The parallelism and interaction between the development in microelectronics and integrated optics strongly suggests that monolithic Integrated Optical sensor technology could emerge as successful as the production technology of monolithic integrated circuits.
For reading out the optical information stored in the sensor chip a rather complicated instrumentation is necessary. Commercially available instruments representing the state of art in sensor sensitivity in the range of 1-10 pg/mm2 appeared in the scientific equipment market by the end of the 80's. All these instruments - reviewed recently by J.J. Ramsden in J. of Molecular Recognition vol. 10., 109-120 (1997) - use optical thin film sensors.
Among these, the BIOS-I made by ASI AG (Zürich) uses input grating-coupler sensor chips ASI 2400 µV that are manufactured by MicroVacuum Ltd, one of the Partners in the present Project. The instrument (i.e. BIOS-I) measures the changes in the effective refractive indices (N) of the guided modes by measuring the angle between the resonance angle of incidence at which mode excitation occurs and the angle of autocollimation at which the incident laser beam (usually a low power He-Ne) is retro-reflected. To tilt the sensor chip into a position where coupling resonance occurs, a high accuracy goniometer with stepping motor positioning is necessary. This instrument works well in laboratories: The sensing method is label free and surface interactions like affinity reactions, immuno-reactions, protein adsorption etc. can be studied in real time with a resolution of several pg/mm2. With sensors ASI 2400 µV, not only the thickness and the refractive index of the adsorbed submonolayer material but also its surface concentration can be determined as a function of time.
In its present form, BIOS-I is a rather complicated, therefore quite expensive machine, that can hardly be used for on-field monitoring. When using sensor chips with two gratings, one for incoupling the laserlight into and the other for outcoupling it from the waveguide, the goniometer could be replaced by a position sensitive detector resulting in a compact, handy and considerably less expensive instrument. Although several publications demonstrate that such instruments are feasible, no commercial version exists on the market yet.
The reason for the lack of such instruments is that there is no industrial-scale technology for (mass) production of sensor chips with more than one grating yet. Optical grating couplers can be fabricated by:
So, there is a need for a flexible, yet cost effective fabrication method for producing sensor chips with minimum two optical grating couplers.
If in- and out-coupling on the same sensor chip (possessing two grating couplers) is realized, one problem still remains to be solved. A sensor having only one sensing pad can detect only one material via one discrete reaction. On the one hand, this is disadvantageous because when changing the sample and/or the reagent on the sensor surface, a new sensor or at least a cleaned sensor surface is needed. On the other hand, most environmental monitoring tasks (waste water, exhaust gases etc.) and biological, clinical screening need multiple sensing pads with differently functionalized surface areas to respond differently to the same sample. These multiple sensing pads (as an artificial nose) should interact selectively with the complex mixture of different molecules of the sample medium.
Integrated optics provides high sensitivity but can not guarantee selectivity or specificity. In biochemical sensors chemically selective coatings containing receptor molecules that specifically or selectively bind certain components of the material to be analysed ensure this selectivity [W. Lukosz, Sensors and Actuators, B29, (1995) 37-50]. The microtitre plates used in analytical laboratories use hundreds and thousands of separate sensing pads or wells. One goal of the R&D proposed in this project is to transfer this concept to environmental monitoring applications by developing selective coatings for production of IO sensor chips with multiple sensing pads for complex sensing applications.
Surface Acoustic Wave (SAW) sensors
Sensors based on piezoelectric effect form another attractive family
of devices. Surface Acoustic Wave (SAW) devices in particular are
most promising for chemical applications because of their intrinsically
high sensitivity, flexibility for detection of many different species and
low cost. A typical SAW sensor consists of a delay line made on a piezoelectric
substrate and a membrane deposited along the acoustic path as the sensor
element. The interaction of the membrane with the ambient (e.g. via the
formation of an adsorbed layer) causes a phase shift on the wave propagation
that can be detected by simple signal processing techniques. When the delay
line is inserted in an oscillator loop, the presence and the concentration
of the analyte to be measured is related to the frequency variation of
the oscillator. In conventional SAW devices the electronics is separated
from the delay line resulting in relative large volume and higher costs.
The main goal of SAW sensors design is the integration of the acoustoelectric
structure and the associated circuitry on the same substrate. Since the
substrates used in electronics are not piezoelectric, an essential first
step of integration of the acoustoelectric structure requires deposition
and processing techniques which are fully compatible with the well established
substate materials and process technologies of the electronics industry.
Since in environmental monitoring applications (air pollution, exhaust gases) the presence of many different gas components should be detected, there is a demand for more and more sophisticated sensors that are intelligent enough to analyse gas mixtures and measure (the concentration of) each principal component separately. In order to meet these requirements the (new generation) SAW sensors should possess a set of membranes, each of different sensitivity to the same gas mixture. Since the sensitivity is controlled by the composition of the respective membrane, for fabricating such multiple membrane matrix structure prototypes a deposition technique ensuring easy though controllable tuning of film properties is necessary.
The corollary of the above analysis is that for the fabrication of multipurpose IO & SAW sensors
Cost-benefit analysis, potential economic impact
The development of the double grating-coupler sensor chip would open
a new market in the biology namely in clinical application. Up to now this
field is intact because the presently available single grating, single
compound sensors can not be used in clinical (hospital) environment, as
they are not selective enough, and their cross selectivity can not be eliminated.
With the development of double grating, multiple compound couplers the
technology and knowledge will be available to produce multiple grating,
multiple compound sensors, so called "multi-plates", which are highly accepted
for clinical use. The same is true for environmental sensor development.
Presently the market is limited because the measuring equipment uses precise
and delicate rotating mechanics, which is not suitable for on site measurements.
The cross selectivity of the presently available single grating, single
compound sensor limits its application in real environmental circumstances.
The successful development of the double grating, multiple compound sensor
will open this market for the application of this high accuracy, reliable
measuring method. The estimated price of a measuring equipment with no
moving part is in the range of 100000 DEM and the price of the sensor chip
is in the range of 100 DM/pc.
Competitors in high accuracy measuring techniques in the biology, environmental field are selling thousands of equipment and sensors annually, and we hope we can get our share from this market as our concept will result cheaper equipment and sensors with higher accuracy, better selectivity and moderate price, suitable as a disposable sensor.
For more detailed description click on the abbreviation of each participating institution or visit their Web pages.
Partner Country Groups:
Research Group on Laser Physics of the Hungarian Academy of Sciences,
(RGLP)
H-6720 Szeged, Dóm tér 9., Hungary.
Postal: H-6701 Szeged, P.O. Box 406, Hungary
Telephone: +36.62.454273, Fax: +36.62.425854
Home Page: http://www.jate.u-szeged.hu/jate/sci/physics/laser/
Partner Country Project Director:
Dr. Tamás Szörényi, PhD, CSc., senior research fellow,
Head of Laser Microfabrication Laboratory,
Telephone: +36.62.454274,
E-mail: T.Szorenyi@physx.u-szeged.hu
Key personnel:
Dr. Zsolt Geretovszky; PhD, Magyary fellow
Telephone:+36.62.454274,
E-mail: gero@physx.u-szeged.hu
Dr. Péter Heszler, PhD, CSc, senior research fellow,
Telephone: +36.62.454421,
E-mail: heszler@physx.u-szeged.hu
Dr. Zoltán Kántor, PhD, research fellow
Telephone:+36.62.454274,
E-mail: zkantor@physx.u-szeged.hu
Dr. Zsolt Tóth, PhD, research fellow
Telephone: +36.62.454421,
E-mail: ztoth@physx.u-szeged.hu
MicroVacuum Ltd., (MVLTD)
Project partner and also main user of research results,
H-1147 Budapest, Kerékgyártó u. 10., Hungary
Telephone: +36.1.2521991, Fax: +36.1.2217996
Home Page: http://www.microvacuum.com
Project Co-Director:
Dr. István Szendrô, managing director
Telephone: +36.1.2521991
E-mail: MicroVacuum@compuserve.com
Key personnel:
Dr. Katalin Erdélyi, technical director
Telephone: +36.1.2521991
E-mail: MicroVacuum@compuserve.com
Németné Margit Sallai, chemical engineer
Telephone: +36.1.2521991
E-mail: MicroVacuum@compuserve.com
Katalin Fischer, electrical engineer
Telephone: +36.1.2521991
E-mail: MicroVacuum@compuserve.com
National Institute for Lasers, Plasma and Radiation Physics, (NILPRP)
Postal: P.O.Box: MG-16 RO 76900, Bucharest V, Romania
Telephone: +40.1.7806925, Fax: +40.1.4231791/4231650
Home Page: N.A.
Project Co-Director:
Dr. Maria Dinescu, PhD, senior research fellow
Telephone: +40.1.7806925, ext. 1920
E-mail: dinescum@ifin.nipne.ro
Key personnel:
Dr. Dan Dumitras, PhD, senior research fellow,
Telephone: +40.1.7806925, ext. 1275
E-mail: dumitras@roifa.ifa.ro
Raluca Dinu, PhD student,
Telephone: +40.1.7806925, ext. 1920
E-mail: dinur@roifa.ifa.ro
Daniela Ghica, PhD student,
Telephone: +40.1.7806925, ext. 1839
E-mail: dghica@roifa.ifa.ro
Florin Ciobanu, PhD student,
Telephone: +40.1.7803725, ext. 129
E-mail: fciobanu@roifa.ifa.ro
NATO Country Groups:
Université de Mons-Hainaut, Département des Matériaux
et Procédés, Laboratoire de Physique de l'Etat Solide, (UMH-LPES)
Avenue Maistriau, 23,
B-7000 Mons, Belgium
Telephone: +32.65.373420, Fax: +32.65.373427
Home Page: http://www.umh.ac.be
NATO Country Project Director:
Professor Lucien Diego Laude, Head of Department
Telephone: +32.65.373420,
E-mail: lucien.laude@umh.ac.be
Key personnel:
Dr. Alain Jadin, PhD, research associate,
Telephone: +32.65.373425,
E-mail: alain.jadin@umh.ac.be
Francoise Hanus, assistant professor
Telephone: +32.65.373425,
E-mail: francoise.hanus@umh.ac.be
Konstantin Kolev, research associate
Telephone: +32.65.373424,
E-mail: konstantin.kolev@umh.ac.be
CNR Istituto di Acustica "O. M. Corbino", (IDAC)
Area della Ricerca Tor Vergata, Via Fosso del Cavaliere 100,
I-00133 Roma, Italy
Telephone: +39.6.49934050, Fax: +39.6.20660061
Home Page: http://www.idac.rm.cnr.it
Project Co-Director:
Dr. Patrizio Verardi, PhD, Head of the Thin Film Laboratory
Telephone: +39.6.49934050,
E-mail: verardi@idac.rm.cnr.it
Key personnel :
Dr. Floriana Craciun, PhD, senior research associate, Professor of University
"La Sapienza",
Telephone: +39.6.49934024,
E-mail: floriana@idac.rm.cnr.it
Dr. Enrico Verona, PhD, Director of the Institute of Acoustics,
Telephone: +39.6.49934481,
E-mail: verona@idac.rm.cnr.it
The Project is funded by NATO under its Science for Peace sub-programme.
The objectives of this proposal are:
Apart from such conventional products like pressure sensors and those used in cars there is no industrial-scale production of chemical sensors for environmental monitoring applications in general either in Hungary or in Romania. The design and fabrication of state-of-the-art gas sensors is definitely a segment, where product oriented R&D is essential and beneficial in order to promote small and madium size companies to the technical level appropriate for commercial production of sensors.
In sensor fabrication both the materials and processes are continuously
evolving. In opposition to the well-established technologies of microelectronics,
there is room and need for novel approaches. In this phase laser-based
technologies offer a promising alternative since they are clean, versatile
and flexible, with a lateral resolution that meets the standard requirements
of production.
1. Production of thin films
Out of the numerous thin film deposition methods available, Pulsed
Laser Deposition is the most effective laboratory technique, since
it is simple but extremely versatile, and the capital cost is relatively
low. Virtually any material, from pure elements to multicomponent compounds
can be deposited, the stoichiometry of the target material can faithfully
be reproduced in the film, in situ deposition of oxide, nitride, carbide
materials is straightforward in reactive atmospheres. The decoupling of
the vacuum hardware and the evaporation power source (the laser) makes
the technique so flexible that it is easily adaptable to different operation
modes without the constraints imposed by the use of of internally powered
evaporation sources. It can also be operated in conjunction with with other
types of evaporation sources in a hybrid approach [Pulsed Laser Deposition of
Thin Films, Eds.: D. B. Chrisey and G. K. Hubler, John Wiley & Sons,
Inc., 1994].
We propose PLD for producing films and continuously improving their properties in continuous feedback from analyses in the preparatory phase, in order to establish reliable technologies for fabrication of thin films of appropriate sensing and piezoelectric properties to be applied as active layers in both IO and SAW sensor devices.
1.1. Thin film coatings of specific sensing properties to be used
in IO sensors
The thin films to be applied as a sensing layer in multipurpose optical
waveguide sensors should be highly sensitive and selective, and should
be of high optical quality possessing appropriate waveguiding properties,
with low optical loss. The thickness of surface sensor layers should be
in the order of 10 nm, while for volume sensing thicker, porous films are
necessary.
Dense thin films of zirconia (ZrO2), tin oxide (SnO2), indium-tin oxide (ITO), and zinc oxide (ZnO) will be deposited by (Reactive) Pulsed Laser Deposition. The effect of PLD conditions on film structure and properties will be investigated using electron microscopy, optical spectroscopy, surface analytical (XPS) and mechanical techniques (AFM). After having the knowledge of firm control of film properties, layers of appropriate thicknesses will be deposited onto ASI 2400 µV sensor chips. The optical characteristics of these prototype single compound single grating-coupler IO sensor devices will be measured in the laboratory of MicroVacuum Ltd. in Budapest, and the (O2, SO2, NOx, H2S, CO, CO2) gas sensing properties in the Surface Acoustic Wave Laboratory of the Istituto di Acustica "O.M. Corbino".
There is also a need for thin film materials which are highly stable even in harsh environments (waste water, chimney exhaust gases, etc.) Therefore the work will be extended to the fabrication of stable, non leaching, non degrading layers for reference sensing application in waveguide sensors.
Porous materials having high surface to volume ratio and thus high sensitivity to interface reactions offer an alternative. Using SOL-GEL technique for film fabrication various molecules can easily be incorporated into the matrix. Based on the experience of MicroVacuum in handling SOL-GEL materials, attempts for incorporating various analyte-sensitive reagents into the micro-porous support matrix and subsequent characterization of the resulting doped SOL-GEL films form also an objective of the work.
1.2. Piezoelectric and metal thin films and thin film structures
to be used in SAW sensors
One of the critical parameters that determine the quality of a SAW
sensor is the acoustoelectrical conversion efficiency, the value of which
is controlled by the quality of the piezoelectric thin film. Therefore,
one essential first step in improving SAW sensor characteristics is the
improvement of the piezoelectric properties of the deposited films. The
results obtained by the Romanian and Italian Partners on ZnO, AlN and PZT
deposition by PLD make us confident of the possibility of producing films
that are superior in piezoelectric characteristics to those obtained using
conventional deposition techniques.
In the first year the optimal conditions for depositing 2-3 µm thick crystalline, c-axis oriented ZnO, AlN, and PZT films which are homogeneous over 2-3 cm2 area and are practically droplet-free, should be determined. A parametric study on the influence of laser wavelength, reactive gas pressure, incident laser fluence, substrate temperature on the film properties is planned together with testing of new structures and materials to be included in the sensors as active layers. Two years must be enough for establishing the technology of fabrication of highly conductive TiN thin films, starting from a Ti target in nitrogen reactive atmosphere, and realisation of Si/TiN/ZnO/TiN, Si/TiN/AlN/TiN and Si/TiN/PZT/TiN multilayer structures. Feed-back from Fourier Transform Infrared Spectroscopy, spectroellipsometry, X-ray diffraction, electron spectroscopy and electron microscopy will ensure effective work. Electrical conductivity measurements and characterization of the TiN/Si, TiN/ZnO, TiN/AlN, TiN/PZT interfaces (by XPS depth profiling and electron microscopy) will decide on the applicability of the multilayer structures in sensors.
The sensing element of a SAW sensor is a thin (metal) membrane between the transducers. For the analysis of gas mixtures a device should possess multiple sensing elements, that are sensitive to the same gas mixture but respond differently to the components. In this case the composition and the concentration of the gas mixture can be determined by electronic processing of the signals obtained from each single element. The response of a membrane can be tuned by the modification of its composition. PLD is again the favourite for the optimisation study since the composition of the PtNi & PdNi alloy thin films can conveniently be tuned by using either two or more elemental targets or one target containing both metals in the desired percentage. The response of the membranes of different composition will be tested on single element SAW sensor structures exposed to gas mixtures of known gas concentration and composition in Rome. This feed-back will yield the informations necessary for the optimization of the deposition parameters.
2. Production of gratings
While the production of thin films will be the result of a painstaking
optimisation process, adopting a known technique, the fabrication of gratings
onto thin film waveguides needs novel approaches, basic scientific achievements.
The main problem to be solved is that the SiO2-TiO2 film used as waveguide in the present generation sensors ASI 2400 µV is transparent throughout the whole visible range, therefore only lasers emitting in the UV (excimers, possessing quite poor beam characteristics, or frequency tripled and quadrupled YAGs with better beam characteristics but less output energy) can be considered for direct ablation when sticking to the well established technology. It is the reason why photochemical processing of the sol-gel precursor and resist materials are considered as alternatives.
Attempts at definition of gratings directly into the hard waveguide film material will be made by combining laser interferometric techniques with laser ablation, ie. selective removal of the material from areas illuminated by high intensity laser light, using special optical arrangements for generating interference patterns [Zs. Bor: Novel pumping arrangement for tunable single picosecond pulse generation with a N2 laser pumped distributed feedback dye laser, Opt. Commun. 29 (1979) 103; Zs. Bor, B. Rácz, G. Szabó, A. Müller and H.P. Dorn: Picosecond pulse generation by distributed feedback dye laser, Helv. Phys. Acta 56 (1983) 383; H.M.Phillips, D.L. Callahan, R. Sauerbrey, G. Szabó and Zs. Bor: Sub-l00 nm lines produced by direct laser ablation in polyimide, Appl. Phys. Lett. 58 (1991) 2761]. Concomitantly, the feasibility of fabrication of gratings by patterned heating of the SOL-GEL (waveguide) material during solidification, by UV laser curing of the precursor (UV laser-induced condensation) or by pattern definition in photosensitised SOL-GEL material will also be studied, using similar optical arrangements.
High resolution (sub 0.2 µm) photolithography using dry-resist materials is a state of art research in semiconductor industry. We consider some dry resist materials as potential candidates for dielectric waveguide fabrication. The initiative to produce high resolution pattern (grating) into a dry resist material waveguide is feasible. The study should include both optical (laser interferometric pattern definition) and materials science aspects (reliable and cost effective deposition techniques).
Since the know-how of grating fabrication is a prerequisite for the development of multiple compound dual grating coupler IO sensors, this sub-project should be finished by the end of the second year.
3. Production of devices
Having the know-how of producing the thin film components of both types
of sensors and the gratings for of the IO sensor chips by the end of the
second year, the task of the last year will be the fabrication of IO sensor
chips with multiple grating coupler pairs, each covered with different
sensing layers, and SAW sensor devices with a matrix of membranes, both
appropriate for quantitative analysis of gas mixtures. Another significant
output of the project will be the comparison of performance and an estimation
of development & production costs of IO and SAW sensors developed for
the same sensing applications and tested under identical conditions.
Note : The R&D is oriented primarily to the development of gas sensors, but since PLD allows for fabrication of c-axis oriented piezoelectric films, which are key components in sensor structures operating in liquid environment as well, the feasibility of fabrication of sensor structures for liquid environments will also be tested.
The result of the project is a new family of Integrated Optical and Surface Acoustic Wave sensors for clinical and environmental monitoring applications. The Industrial Partner, MicroVacuum Ltd. will start manufacturing as soon as the technology is available, after the completion of the project, at latest.
The estimated price of the sensor chips is in the range of 100 DM/pc, while that of the measuring equipment lies in the range of 100000 DEM. Competitors in high accuracy measuring techniques in the biology and environmental field are selling thousands of equipment and sensors annually. It is realistic to get a fair share from this market, as our concept will result cheaper equipment and sensors with higher accuracy, better selectivity and moderate price, suitable as a disposable sensor.
The scientific results obtained during the R&D work will be published in leading archival journals. As usual, the support of NATO will be acknowledged in all papers. We plan to make the achievements visible to the scientific and technical community via the HomePage of the Project, as well.
The visibility of this NATO initiative is even now fairly good in the media of the Partner Countries. After receiving the notification on being selected for the preparation of a Project Plan, a press release was issued in the local newspaper, in Szeged, Hungary. This practice will definitely be continued and broadened if the Proposal will be accepted.