Wilshaw, K. Mazzer "High-resolution scanning electron microscopy of dopants in p-i-n junctions with quantum wells" Inst. Copyright by the Institute of Physics, GB. Kronik, Y. Gruenther, Duncan G. Steel and Leopald Bayvel, Elsevier, Oxford, 5, 36, Katz, D. Faiman, B. Mishori, Y. Shapira, A. Isakina and M. Baksht, S. Solodky, M. Leibovitch, G. Bunin, Y. Rothschild, Y. Komem, A. Levakov, Y. Shapira, N.
Ashkenasy "Electronic and transport properties of reduced and oxidized nanocrystalline TiO2 films" Appl. Letters, 82, , Hallakoun, I. Toledo, J. Kaplun, G. Bunin, M. Leibovitch, Y. Shapira "Critical dimension improvement of plasma enhanced chemical vapor deposition silicon nitride thin films in GaAs devices" Mat. Rothschild, A.
Ashkenasy, Y. Komem "Surface photovoltage spectroscopy study of reduced and oxidized nanocrystalline TiO2 films" Surf. Gordon, Y. Khramtsov, T. Baksht, M. Letters, 83, , Yang, Y. Letters, 81, , Copyright by The American Institute of Physics. Kinrot, Y. Shapira "Surface electronic structure of P-InP using temperature-controlled surface photovoltage spectroscopy" Phys.
B, 65, , Jihua Yang, Y. Shapira " Alloy composition and electronic structure of Cd1-xZnxTe by surface photovoltage spectroscopy" J.
Copyright The American Institute of Physics. Shalish, Y.
Patent No. Shalish, C. Shapira, J. Salzman "Hall photovoltage deep-level spectroscopy of GaN films" Phys. B, 64, , Shapira "Photoinduced charge carriers at surfaces and interfaces of poly [2-methoxy 2 -ethyl-hexyloxy -1,4-phenylene vinylene] with Au and GaAs" Phys. Shapira, L. Burstein, J. Salzman "Surface states and surface oxide in GaN layers" J.
Shapira "Surface photovoltage of semiconductor structures: At the crossroads of physics, chemistry and electrical engineering" Surface and Interface Analysis, 31, , Ashkenasy, M. Rosenwaks, Y. Shapira "Characterization of quantum well structures using surface photovoltage spectroscopy" Mater. Shalish, L.
Kronik, G. Segal, Y. Shapira, S. Zamir, B. Meyler, J. Salzman "Grain boundary controlled transport in GaN layers" Phys. Eizenberg and J. Leibovitch, N. Rosenwaks, I.
Hallakoun, Y. Shapira "A new technology for thermodynamically stable contacts for binary wide bandgap semiconductors" U. Copyright by The American Vacuum Society. Shapira, U. Tisch, J. Salzman "Yellow luminescence and related deep levels in unintentionally doped GaN films" Phys. B 59, , Shapira, K. Barnham, J. Nelson, J. Shapira "Surface photovoltage phenomena: Theory, experiment and applications" Surface Science Reports, 37, , Kronik, B. Mishori, E. Shapira, W. Solar Energy Materials and Solar Cells , 51 1 , 21, Faiman, S.
Goren, S. Shtutina, B. Shapira, F. Pollak, G. Burnham, X. Faiman, A. Belu-Marian, Y. Kronik, N. Leibovitch, E. Gorer, G. Hodes "Surface states and photovoltaic effects in CdSe quantum dot films" J. Copyright by The Electrochemical Society Inc. Mishori, M. Pollak, D. Streit, M. Letters , 73, , Aphek, L. Kronik, M. Shapira "Quantitative assessment of the photosaturation technique" Surface Science , , , Shames, S. Shtutina, S. Belu-Marian, M. Manciu, A. Devenyi "Studies of C 60 thin films using surface photovoltage spectroscopy" Chem. Letters , , , Partee, J.
Shinar, Y. Lubianiker, I. Balberg "Surface photovoltage spectroscopy of porous silicon " Phys. B - Rapid Communications , 55, , Faiman, Y. Shapira "Studies of electron structure of C 60 thin films by surface photovoltage spectroscopy" Solid State Communications , , , Balberg, Y. Lubianiker, J. Shinar, J. Partee, L. Burstein, Y. Weisz, M. Gomez "The relation between phototransport and photoluminescence in porous silicon" J. Balberg, E.
Shapira "Following directly the effect of the various deep states on the phototransport properties of a-Si:H " J. Solids , , , Moons, D. Gal, J. Beier, G. Hodes, D. Cahen, L. Kronik, L. Burstein, B.
Shapira, D. Hariskos, H. Solar Energy Materials and Solar Cells , 43, 73, Fefer , L. Riedl "In-situ Monitoring of surface chemistry and charge transfer at semiconductor surfaces" Appl. Ashkenasy, L. Shapira, Y. Rosenwaks, M. Hanna, M. Leibovitch, P. Ram "Surface photovoltage spectroscopy of quantum wells and superlattices" Appl. Leibovitch, L.
Kronik, E. Fefer, L. Burstein, V. Korobov, Y. Shapira "Surface photovoltage spectroscopy of thin films" J. Hanson, A. Shapira and H. Golombek "Spacecraft with integrated array of solar cells and electronically scannable antenna" U. Filed and issued. An electrical method for determining the thickness of a dielectric utilizes capacitance measurements of MOS capacitors fabricated, for the test purposes, on the dielectric layer. In other electrical methods, the dielectric thickness can be determined without fabricating MOS test capacitors by charging the surface of the dielectric layer with a corona discharge and measuring the resulting surface potential of the charged dielectric layer with a Kelvin or a Monroe probe.
These techniques are discussed, for example, in U. Verkuil, and 6,,, to J. Lagowski et. Due to substantial leakage of current across the dielectric layer, via tunneling, the methods described in U. During a typical measuring time for these techniques, e. For ultra-thin dielectric layers, the leakage currents are orders of magnitude larger than the leakage currents for thicker dielectric layers, e. In general, the invention relates to an apparatus and method for producing non-contact electrical measurements of capacitance and thickness of ultra-thin dielectric layers on semiconductor substrates wafers.
The apparatus and method produces effective and accurate measurements of the dielectric layer thickness despite substantial leakage of current across the layer and no apriori knowledge of the relationship between the leakage current characteristics, i. Si3N4, and barium strontium titianate BST. The non-contact electrical technique can be used to record multiple, repeatable measurements of ultra-thin dielectric capacitance and thickness at the same location on the wafer under highly reproducible conditions.
In an aspect, the method of determining the thickness of a dielectric layer on a semiconductor wafer includes depositing an electric charge sufficient to cause a steady state condition in which charge current is equal to the leakage current; measuring the potential of the dielectric surface; and comparing the measured parameters to calibrated parameters to derive the dielectric layer thickness. In another aspect, the method of determining the thickness of a dielectric layer deposited on a semiconducting wafer includes depositing an ionic charge onto a surface of the dielectric layer deposited on the semiconducting wafer with an ionic current sufficient to cause a steady state condition; measuring, via a non-contact probe, a voltage decay on the dielectric surface as a function of time; and determining the thickness of the dielectric layer based upon the measured voltage decay.
The method can further include measuring the voltage decay after terminating the deposition of ionic charge. Embodiments of the invention can include one or more the following. The steady state condition results when the ionic current equals a leakage current flowing from the semiconducting wafer and across the dielectric layer. The step of determining the thickness of the dielectric layer includes determining the initial surface potential, V0, on the dielectric layer from the measured voltage decay.
The coefficients a and b in the linear expression are determined by a calibrating procedure. The calibrating procedure comprises recording a decay voltage on a plurality of semiconducting wafers each having a known dielectric layer thickness, and determining from each measured voltage decay an initial surface potential.
The steps of depositing a charge onto a surface of the dielectric layer, measuring the voltage, V0, and determining the thickness of the dielectric layer all occur in about 7 seconds or less. The method further includes determining the capacitance of the dielectric layer deposited on the semiconducting wafer. The voltage decay is measured after terminating the deposition of ionic charge. The steps of depositing ionic charge, measuring the voltage decay, and determining the thickness are performed on a measurement area smaller than a total surface area of the semiconducting wafer.
The method further includes depositing a precharging ionic charge on the dielectric layer. The method further includes performing the steps of depositing ionic charge, measuring voltage decay, and determining the dielectric thickness on a plurality of measurement sites on the dielectric layer. In ultra-thin dielectric layers, the leakage current mechanism may differ from the mechanism of leakage in thicker dielectrics layers. For non-SiO2 dielectrics, leakage current mechanisms, other than F-N tunneling or direct tunneling, such as Schottky emission or Frenkel-Poole emission can be dominant.
As a result, the exact tunneling current equation relating the layer thickness to electrical parameters such as the tunneling leakage current and the voltage drop across a dielectric layer may not be known apriori. Advantageously the method of this invention, unlike the method discussed in U. The apparatus and methodology can be used to determine dielectric layer thicknesses with a 0.
The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims. The methodology for measuring ultra-thin dielectric layer thicknesses includes depositing an electric charge using a corona current, JC, from a corona discharge source, onto a dielectric layer of a semiconductor wafer to produce a steady state condition in which the leakage current, JLEAK, equals JC; terminating the corona current; measuring electrical parameters of the wafer, such as the dielectric potential or the potential decay rate; and comparing the measured electrical parameters to calibrated parameters to derive the ultra-thin dielectric thickness.
Referring to FIG. An example of the computer controlled test system 10 can be found in U. Semiconducting wafer 5 includes a semiconductor substrate wafer 11 e. Semiconductor substrate wafer 11 is connected to the ground potential via a back-contact device 17 in electrical contact with grounded wafer chuck Test system 10 measures the oxide thickness by depositing electric charge onto a surface 15 of top dielectric layer 13a, and then monitoring the voltage decay due to current flow through this dielectric layer into or from semiconductor substrate wafer Computer 12 also calculates an initial surface potential, V0, i.
The system 10 includes a testing device that contains a charging station 14 and a charge measuring station 19 both of which are translatable, using a solenoid 20, relative to a chuck Wafer 5 is held, e. A back-contact device 17 protrudes through dielectric layer 13b making electrical contact to semiconductor substrate wafer 11, and connecting it to the ground potential. An example of a back-contact device capable of protruding through dielectric layer 13b can be found in U.
Chuck 18 is mounted on a moveable stage not shown in FIG. Charging station 14 and the charge measuring station 19 are spaced apart from each other on a mount 21 at a fixed distance X0, e. Solenoid 20 is used to translate charging station 14 and measuring station 19 by a distance X0 such that the measuring station is precisely above the wafer at the X1-position, i. Charging station 14 includes a corona discharge source 16 and a light source Corona discharge source 16 includes a corona charging wire 14a which receives a high voltage potential of either a positive or negative polarity as needed and a corona-confining electrode ring 14b, e.
In general, test system 10 deposits a corona ionic current on the oxide to reach, in less than about 10 seconds, a steady state condition in which the tunneling current, JLEAK, flowing across dielectric layer 13a is equal to corona ionic current, JC, without generating traps in the dielectric that result from large corona ion fluxes.
Test system 10 can deposit either a positive or negative charge on the surface. Typically, test system 10 positively charges the surface with a positive corona discharge because a negative corona discharge is more difficult to control with respect to charging uniformity. Non-uniform surface charging can produce non-uniformity in the initial voltage, V0, which in turn would generate errors in deriving the oxide thickness.
Charge measuring station 19 includes a light source 25 and a contact potential sensor 22, such as a Kelvin probe or a Monroe-type probe, which is used to measure the contact potential of dielectric layer 13a with respect to a reference electrode 30, i. Sensors of these types are described, respectively, in G. Reedyk and M. Perlman: Journal of the Electrochemical Society, Vol. An example of a commercially available device is the Isoprobe model by Monroe Electronics, Lyndonville, N. Typically, electrode 30 is separated from the top surface of the dielectric film by an air gap of about fraction of about a millimeter.
The oxide surface potential, V, measured with respect to a vibrating reference Kelvin or Monroe electrode is often referred to as contact potential difference. The dielectric surface potential, V, decays with time, t, after charging, i. Test system 10 includes light sources 23, 25, preferably green or blue light emitting diodes, to illuminate testing site 15a during charging light source 23 and during measuring light source 25 thereby reducing the value of VSB by collapsing the surface depletion region in the semiconductor.
Illuminating the dielectric surface is especially necessary in the case of positive charging of an oxide, e. This specific case, i. For p-type n-type silicon substrates, positive negative corona charge creates a very large depletion layer surface barrier, VSB, ranging from 10 to volts, which in dark would decay slowly after terminating corona charging. Without illumination, the potential drop across the silicon depletion region would dominate the non-contact potential measurements and thereby prevent reliable measurement of the oxide thickness. Typically, green or blue diodes are used that emit light at a wavelength of about nm to nm.
Shorter wavelengths of light are not desired since they can change trap occupation in the dielectric layer. In operation, computer 12 sends a signal to move chuck 18 into position below charging station Computer 12 presets these charging conditions, such as ionic current and charging duration, i. After depositing the charge onto testing site 15a of dielectric surface 15, computer 12 sends signals to turn the discharge source off, to turn on the charge measuring station's light source, to turn off the discharge station's light source, and to move the charge measuring station, via the solenoid, to position the contact potential sensor above the charged surface of the wafer, i.
This set of operations is done fast enough, e. Once measuring station 19 is above charged testing site 15a, the computer acquires the measured data of the surface potential, V, vs. The typical time period of acquiring the data is about 3 seconds. The computer can derive these values by fitting the measured voltage decay to a polynomial expression or other appropriate mathematical model.
After calculating the initial surface potential, the computer can determine an equivalent oxide thickness, EOT , by using the relationship:. The calibrating procedure includes measuring the surface potential and determining V0 for several different thicknesses of SiO2 layers on silicon e. Each of the calibrating measurements are conducted with the same predetermined corona flux that achieves the steady state condition without generating corona stress and dielectric traps. In general, the corona flux is predetermined by depositing a corona flux sufficient to cause the steady state condition but small enough not to generate corona stress and dielectric traps.
The latter can be ascertained by monitoring the stress-induced leakage current SILC. For calibration, each thickness of the SiO2 layers must be known e. The calibrating corona charging and V0 measurements, typically, are performed on the same surface location used to record the optical thickness measurement.
T using a linear regression. For positive corona charging of ultra-thin dielectric layers, e. Typically, Test system 10 can measure the surface potential with a precision of 1 mV and thereby determine the dielectric thickness with a precision of 0. This value of b is representative if the reference electrode used for the potential measurements is made of platinum. This coefficient can be scaled for different electrodes, e. It can also be scaled to account for different wafer doping concentrations between the calibrated wafers and the measured wafers.
Although illuminating the wafer, as described above, eliminates contributions from the depleted surface space charge layer, illuminating the wafer does not completely eliminate the inversion or accumulation space charge layer. Test system 10 follows the same protocol used in determining EOT's for the ultra-thin dielectric layers, except that the computer uses different values of coefficients a and b. The robustness results from the steady state condition and the strong dependence of the tunneling leakage current, JLEAK, on the electric field, F.