Title of Invention | A PROCESS FOR PREPARING A PIEZORESISTIVE PRESSURE SENSOR WITH ENHANCED SENSITIVITY |
---|---|
Abstract | The present invention relates to a process for preparing a piezoresistive pressure sensor with enhanced sensitivity comprising in combination - (a) metal lines(6)/contacts(1), (b) at least one piezoresistors(7), relatively thick and suitably doped to lead to an optimization between the confinement of current lines and piezoresistive coefficient diaphragm(8) and (d) oxide layer on silicon block / base surface(9). This invention also pertains to a process for preparing the above sensor which comprises forming a thick oxide layer on a p-type crystalline silicon wafer by a dry-wet-dry sequence of oxidation followed by lithography to expose the region for micromachining. The sample is micromachined at room temperature to around 20 to 50 µm, and then subjected to selective, controlled diffusion to form the piezoresistors. Then patterned metal contacts of Al or Ni are formed on the sample to produce the desired sensor, |
Full Text | This invention relates to a process for preparing piezoresistive pressure sensor with enhanced Sensitivity and prooooo for preparing the oome More particularly the present invention pertains to a Silicon MEMS (Micro Electro Mechanical System) piezoresistive pressure sensor with enhanced system and a process for preparing the same. In 1954 Charles S Smith observed piezoresistance effect in germanium and silicon, and a piezoresistance tensor had been determined experimentally for these materials. Since then attempts have been and are being made to apply this effect in the development of mechanical sensors to measure motion related parameters such as position, displacement velocity flow and force - related measured such as pressure, acceleration torque, etc Silicon has been material of choice for such sensors in view of its compatibility with IC technology Work is in progress for optimization of the design of such pressure sensors and a number of conventional simulation tools like "MEMSCAD COVENTORWARE" have been developed which can be applied to highly doped and thin piezoresistors. Thin piezoresistors are normally used for better confinement of the current lines. However, it has been found that piezoresistive coefficient decreases with increases in surface concentration. The sensitivity of the commercially available piezoresistive pressure sensors fabricated using the process as desc'nbed in the claim is lower by almost three times than the proposed scheme. Sensitivity can also be increased by formation of a very thin membrane (2-5µm) by anodic bonding technique but the process requires greater number of fabrication steps and hence is costlier. The process described here requires simple fabrication steps-photolithography, oxidation, and diffusion/ion implantation and at the same time yields higher sensitivity by the innovative concept of optimization between' the confinements of current line sand piezoresistive coefficient Also it is well known that the temperature influence of the heavy doped silicon piezoresistors is lesser compared to that of lightly doped piezoresistors. But it has been observed that the overall signal to noise ratio (considering temperature coefficient as noise) increases in case of the proposed design of pressure sensor since the change in temperature coefficient of piezoresistance is less than one and half times whereas the increase ins sensitivity is around three times. The present invention finds useful application in pressure sensors. The commercially available high-pressure sensors have reduced the thickness of the membrane to increase sensitivity but requires more complicated fabrication steps to withstand the high pressure. The proposed design aims at a cost effective sensor where the thickness of the membrane need not be reduced to achieve high sensitivity since the sensitivity is increased by an optimization process and thus requires simple fabrication steps The improvedfeatures of the present invention can be summarized as follows: •High signal to noise ratio in the high-pressure range. • Cost effective. • Applicable in high pressure range requiring simple fabrication steps. - 2 - The present invention aims at designing silicon MEMS (Micro Electro Mechanical System) piezoresistive pressure sensor having enhanced sensitivity to overcome the difficulties envisaged in pressure sensors hitherto known The sensors of this invention can perform the simulation studies for various piezoresistors designs unlike conventional simulation tools. With this analysis an optimization has been achieved between doping concentration and junction depth of the diffused strip forming the piezoresistors, which is capable of functioning as a pressure sensor within 0-1 bar range with enhanced sensitivity. The principal object of the present invention is to provide a silicon MEMS piezoresistive pressure sensor with enhanced sensitivity by forming a relatively thick and low "doped piezoresistors on bulk silicon leading to an optimization between the confinement of current lines and value of piezoresistive coefficient which is capable of overcoming the drawbacks encountered in the prior art sensors A further object of this invention is to provide an improved pressure sensor with marked sensitivity capable of functioning within 0-1 bar range. A still further object of this invention is to provide a process for preparing an improved piezoresistive pressure sensor with enhanced sensitivity. 3 The foregoing objects are achieved by the present invention, which is concerned with an improved piezoresistive pressure sensor with enhanced sensitivity comprising in combination the following' (a) metal lines (b) at least one piezoresistors, relatively thick and suitably doped on bulk Si layer (c) diaphragm and (d) oxide layer formed on silicon block/base surface This invention further provides a process for preparing an improved silicon MEMS piezoresistive pressure sensor which enhanced sensitivity as defined herein which comprises. [I] Selecting a p-type polished monocrystalline silicon wafer [II] Carrying out oxidation on both sides to form a thick oxide layer, [III] Opening of window(s) by photolithography technique using a suitable mask, [iv] Micromachining the window(s) to form diaphragm of desired thickness [v] Conducting selective, controlled diffusion/ion implantation preferably to form the piezoresistors and [vi] Forming patterned metal contacts along or across the piezoresistors where the initial p-type wafer has a doping concentration/resistivity between 1015 and 1016 atoms/cc and thickness in the range of 10-50 microns As noted earlier, an oxide layer is formed on bulk silicon and metal contacts are provided on the said oxide layer(s) after suitable etching. The front face of the silicon block is micromachined Slabs of resistance are provided by the bulk silicon below the said metal contacts for satisfactory functioning of the sensor The monocrystalline silicon used as the starting material is of p-type having a doping concentration/resistivity varying between 1015 and 1016 atoms per c c. and oxidation of the wafer is done in a dry-wet-dryisequence to form a thick oxide layer. The steps of dry oxidation are carried out at temperatures between 900°C-1000°C for around 15 minutes. Wet oxidation is carried out in presence of water vapour/moisture for 1-1 5 hours at room temperature or thereabout For exposing the desired region/location for micromachining, lithography is carried out at room temperature in 20%-40% aqueous solution of KOH, EDP (Ethylene Diamine Pyrocatechol) or TMAH (Tetra Methyl Ammonium Hydroxide) Micromachining1' of the wafer is done to around 50µm, followed by selective, controlled diffusion with boron at elevated temperatures of ~1000°-1100°C for around 4-5 hours to form the piezoresistors or/by ion implantation. Thereafter metallic aluminum or nickel contacts are formed along or across the piezoresistors, as may be desired. At contacts are formed on the sample by vacuum evaporation, followed by annealing at around 450°C for 45 seconds, and Ni-contacts are formed by electroplating/electro deposition in NiCI? or Ni(CN)2 in KCN bath The present invention has achieved a simplified piezoresistors design where current passing through the sample and the stress applied are along particular cubic axes of the crystal This has led to the assumption of average values of Young's Modulus and Poisson's ratio, neglecting their variations in the crystallographic directions. - 4- The theory used here for the mechanical properties assumes the diaphragm to be made of an isotropic material although the single crystal silicon diaphragms are actually crystallographically anisotropic and therefore have properties and parameters like Young's Modulus which vary with direction The analysis has been approximated by the average values of the mechanical properties due to a particular orientation of the diffused strip. In case of pure bending of plates (which mean that cross sections of the bar remain plane during bending and rotate only with respect to their neutral axis so as to be always normal to the deflection curve) the stress in the elemental lamina 'abed' at a distance 'Z' from the neutral surface are found from Hooke's (aw to be equal to where v is the Porsson's ratio, E is the Young's Modulus and 'D' is the flexure ngidity given by The normal stresses distributed over the lateral sides of the element can be reduced to couples, the magnitudes of which per unit length must equal the external moments M x and M y For uniformly loaded circular plates the bending moments can be proved to be From equations (a) - (e) we obtain where 'a' is the radius of the diaphragm,' r' is the radial distance from the center of the diaphragm. 'z' is the distance from the neutral line, 'q' is the pressure applied and 'h' is the thickness of the diaphragm. From the expressions of radial and tangential stresses, it is clear that they increase with distance from the neutral axis. Depending on the position and geometry of the piezoresistor, the average radial and tangential stress is given by 5 For thin and highly doped piezoresistors, current lines are mostly confined to the surface so that z - h/2 leading to the following expressions of stress: which are actually the conventional expressions used in the CAD tools[3]. Electrical Modeling One can divide the membrane into elemental stabs of thickness 't'. The resistance of the nth slab = Rn where pnl is the longitudinal resistivity and pnt is the transverse resistivity pnl is either the resistivity of the diffused strip or that of the bulk silicon depending on whether the slab is in the diffused silicon or the bulk silicon, where as pnt may include the resistivity of both the diffused region and the bulk silicon. 'A' is the contact area, I is the distance between the contacts, and V is the width of the contact' For simplicity a square shaped contact region has been assumed and hence The resistances of the slabs are in parallel and they increase with distance from the metal contacts') Simulation has been done considenng slices of silicon of 0.01-micron thickness. The total resistance R0of the membrane is calculated as whereN is the total number of slabs in which the membrane is divided Some of the elemental resistances are that of the diffused silicon and the rest are that of the bulk silicon. Thus the equivalent resistance of the membrane can be expressed as Now, on application of pressure, the resistance of each elemental slab changes according to the following equation: where (, the piezoresistive coefficient depends on temperature, concentration and band structure. (1 is the longitudinal piezoresistive coefficient and (t which relates changes in resistance in the longitudinal direction to stresses in the transverse direction is referred to as the transverse coefficient The values of (I and (t vary widely with crystallographic directions But in our analysis we have considered a uniaxial stress instead of the complete tensor due to a particular orientation of the diffused strip. The piezoresistive coefficients differ depending on the doping concentration of the slabs 6 (1 and (t are given in equations (f). The ± sign indicates whether the stress is elongative or compressive respectively. The equivalent resistance on application of pressure (R01) is formulated as: The magnitude of the resistance of the slabs on one side of the neutral axis decrease and on the other side it increases. Thus the simulated sensitivity takes into account the effect of both the compressive and eiongative stress unlike the conventional MEMS software. Without diffusion, the current lines are not confined to one portion of the neutral axis and hence a part of compressive stress is neutralized by eiongative stress. But with diffused layer the current lines are confined on one side of the neutral axis and hence the sensitivity is grater. However as the plezoresistive coefficients decrease with surface concentration, there is a scope of optimization between the doping concentration and the junction depth of the piezoresistor with a view to maximize sensitivity.. The invention will now be further illustrated by means of the drawings accompanying, this specification, in which Fig. 1 Shows the equivalent electrical circuit diagram of a piezoresistive pressure sensor of this invention and the distribution of current lines there through. Fig 2 Depicts distribution of current lines within the silicon membrane. Fig.3 Gives the top view of the piezoresistive pressure sensor of this invention. Fig.4 Illustrates the flow chart diagram showing different stages of fabrication of the invented piezoresistive pressure sensor Fig.5 Gives a cross sectional view of the fabricated sensor and Fig.6 Shows a set-up for measuring/determining the sensitivity of the specimen of the pressure sensor prepared according to the process of this invention In Fig.1, (1) denotes the metal contacts. (2) is bulk silicon, (3) shows slabs of resistance, and (4) is the micromabhined silicon surface. In Fig 2, (1), (2) and (4) have the same significance as above, and (5) shows distribution of current items In Fig 3, metal lines are shown by (6) and (7) are the piezoresistors Diaphragm location is shown by (8) and (9) is the oxide layer. From the top view of the sensor design it may be noted that 7 - I - to increase the sensitivity and to compensate for the temperature fluctuations, four piezoresistors have been incorporated in a Wheatstone bridge configuration i The flow chart diagram of Fig.4 not only shows different stages of production, but also gives the outline of the intermediate products obtained in each stage. The cross-sectioned view of the sensor in Fig.5 shows metal contacts (1), bulk silicon (2), micromachined silicon surface (4) and oxide layer on top (9). A measurement set-up of Fig 6 shows the locations of metal contacts (1), Sensor (10), phosphor bronze .. (11), digital meter for current measurement (12), vacuum pump (13), mercury manometer (14) and voltage source. These components are arranged on a suitable firm base (16), supporting stands (17) and fixing/holding means (18) The invention will be more particularly described with the help of the following example taking into account the design parameters for a single piezoresistors. which are given by way of illustration and not by way of limitation. Example Before undertaking actual fabrication, the following considerations were gone into Design: Design parameters for a single piezoresistor Thickness of the membrane To design a piezoresistive pressure sensor within 0-1 bar range, the thickness of the membrane is calculated so that the maximum stress do not exceed the fracture stress of silicon which is 3x109dyn/crri2. The stresses on the diaphragm are given by Metal contacts The positions of the metal contacts are dependent on the position of the maximum stress. For the chosen design parameters, the position of the maximum stress has been determined by using Coventorware MEMSCAD software. Piezoresistor parameters With higher doping concentration and reduced junction depth, the piezoresistive coefficient decreases but the confinement of current lines are better. Hence optimum values have -8- to be chosen with a view to maximize sensitivity. From the simulation results, the optimum values of junction depth and doping concentration are fixed at 5 µm and 1017 atoms /cc. The actual sequence of fabrication of piezoresistive sensors may be stated as under :- On a p-type microcrystalline silicon wafer of resistivity 1016 atoms/cc. oxidation is to be carried out in a dry-wet-dry sequence to form a thick oxide layer. Dry oxidation is conducted at an elevated temperature for 900°C-1000°C for around 15 minutes. Wet oxidation is effected in presence of moisture/water vapour at ambient temperature for 1-1.5 hours This is followed by lithography in accordance with conventional methods to expose the region for micromachining, which is conducted at room temperature to about 50µm. Controlled diffusion/ion implantation is done to form relatively thick and suitably doped piezoresistors on bulk silicon. Then Al-contacts are evaporated on the sample followed by annealing at 450°C for 45 sees Simulation results: Simulation has been performed with the following parameters :- [a] Substrate doping concentration - 1016 atoms/cc. [b] Thickness of the diaphragm - 50 µm [c] Radius of the diaphragm - 1500 µm [d] Length and width of the metal contacts - 200 µm each [e] Radial distance of the metal contacts from the centre of the diaphragm - 500 µm [f] Length of the piezoresistors - 1000 µm [g] Yong's Modulus of Silicon - 1.6 x 1012 syn/cm2 [h] Poisson's ratio of silicon - 0.33 [i] Area of the diaphragm - 9 sq mm 0] Length of the device - 1 cm. The simulation has been performed for various pressure differences. From the results, we can observe that for higher doping concentrations (>1018 atoms/cc) the sensitivity decreases with increasing junction depth. Apparently there should not be any striking difference between the sensitivities for varying junction depths in a highly doped diffused silicon because the current is more or less confined in the diffused region for all the above cases. This can also be proved from the simulation results But the anomaly lies in the fact that for greater junction depths the number of resistances in parallel increase and hence the overall resistance decrease Thus Ro for 5µm is 1.2 ohm and that for 3µm is 1.8 ohm and so on. Similarly on application of pressure Ro1 for 5 µm is less than that for 3 µm. This explains the difference in sensitivity. For comparatively lower doping concentrations ( Simulation to obtain the optimum values of doping concentration and junction depth for different thickness of the bulk silicon has also been performed. From the simulation results, we observe that 9 the sensitivity is greater for thinner diaphragm, which is quite obvious since the stress is inversely proportional to the thickness of the diaphragm (eqn.f). The optimum value of doping concentration is found to be 1017 atoms/cc for different thickness of diaphragms. The piezoresistive pressure sensors are subjected to testing with the help of a measurement set-up as illustrated in Fig 4 of the drawings. Experiment for verifying the activity of the pressure sensor fabricated according to this inventions conducted by producing differential pressures by placing the diaphragm on an insulator plate, which is connected to a vacuum pump by a rubber tube. The tube can be evacuated and held at any pressure between 760mm and and zero by means of a controlled leak. The pressure was measured [with a mercury manometer. All electrical measurements were made with a standard power supply and a Keithley meter set-up. The magnitude of current changes on application of stress and the sensitivity is calculated from the fractional change in current The agreement between the observed and predicted results is quite good, particularly in view of several possible sources of error in the diaphragm experiments, e.g in thickness measurements., in the degree of correspondence of the experimental clamping conditions to the idealized theoretical conditions, etc. It has also been observed that the sensitivity increases almost twice with the optimized values of doping concentration and junction depth. Apart from providing a convenient tool to analyze composite MEMS structure like porous Si-Si, diaphragms made in this manner are like to have applications in view of their marked sensitivity While the invention has been described in detail and with reference to specific embodiments thereof, it will be apparent to one skilled in the art that various changes and modifications can be made therein without deviating or departing from the spirit and scope of this invention. Thus the disclosure contained herein includes within its ambit the obvious equivalents and substitutes as well Having described the invention in detail with particular reference to the illustrative example given above and the accompanying drawings, it will now be more specifically defined by means of claims appended hereafter. 10 I claim- 1 A process for preparing an improved piezoresistive pressure sensor with enhanced sensitivity which comprises - [i] Selecting a suitably doped p-type polished monocrystalline silicon wafer [ 11] Carrying out oxidation on both sides to form a thick oxide layer. [III] Opening of wihdow(s) by photolithography technique, using a suitable mask [Iv] Micromachining the window(s) to form diaphragm of desired thickness [v] Conducting selective / ion implantation preferably controlled diffusion to form the piezoresistors and [vi] Forming patterned metal contacts along or across the piezoresistors, wherein the initial p-type monocrystalline wafer has a doping concentration/resistivity varying between 1015 and 1016 atoms per c c and thickness in the range of 10 to 50 micron. 2] A process as claimed in anyone of claim 1, wherein oxidation of the said wafer is done in a dry-wet-dry sequence to form a Thick oxide layer. 3] A process as claimed in anyone of claim 1 and 2, wherein the steps of dry oxidation is carried out at temperatures between 900°C and 1000°C for around 15 minutes. 4] A process as claimed in claims 1 and 2, wherein wet oxidation is carried out in presence of water vapour/moisture for 1 hour to 1 5 hours at room temperature or thereabout 5] A process as claimed in claims 1 to 4, wherein lithography is done to expose the region for micromachining which is carried out at room temperature in 20%-40% aqueous KOW, EDP or TMAH solution 6] A process as claimed in claim 5, wherein the wafer is micromachined to around 50µm 7] A process as claimed in claims 1 to 6, wherein said selective, controlled diffusion is carried out with boron at elevated temperatures of ~1000°C-1100°C for around 4-5 hours to form the piezoresistors or/by ton-implantation. 8] A process as claimed in claims 1-7, wherein metallic Al or Ni-contacts are formed along or across the said piezoresistors 9] A process as claimed in claims 1 and 8, wherein Al-contacts are formed by vacuum evaporation on the sample, followed by annealing at around 450°C for 45 seconds and Ni-contacts are formed by electroplating in NiCl2 or Ni(CN)2 in KCN bath. -11- 10] A process for preparing an improved piezoresistive pressure sensor with enhanced sensitivity, substantially as hereinbefore described and illustrated in the Example and accompanying flow sheet diagram Dated this 26th day of June 2003 - 12- The present invention relates to a process for preparing a piezoresistive pressure sensor with enhanced sensitivity comprising in combination - (a) metal lines(6)/contacts(1), (b) at least one piezoresistors(7), relatively thick and suitably doped to lead to an optimization between the confinement of current lines and piezoresistive coefficient diaphragm(8) and (d) oxide layer on silicon block / base surface(9). This invention also pertains to a process for preparing the above sensor which comprises forming a thick oxide layer on a p-type crystalline silicon wafer by a dry-wet-dry sequence of oxidation followed by lithography to expose the region for micromachining. The sample is micromachined at room temperature to around 20 to 50 µm, and then subjected to selective, controlled diffusion to form the piezoresistors. Then patterned metal contacts of Al or Ni are formed on the sample to produce the desired sensor, |
---|
00357-kol-2003 correspondence.pdf
00357-kol-2003 description(complete).pdf
357-kol-2003-granted-abstract.pdf
357-kol-2003-granted-assignment.pdf
357-kol-2003-granted-claims.pdf
357-kol-2003-granted-correspondence.pdf
357-kol-2003-granted-description (complete).pdf
357-kol-2003-granted-drawings.pdf
357-kol-2003-granted-examination report.pdf
357-kol-2003-granted-form 1.pdf
357-kol-2003-granted-form 18.pdf
357-kol-2003-granted-form 2.pdf
357-kol-2003-granted-form 3.pdf
357-kol-2003-granted-letter patent.pdf
357-kol-2003-granted-reply to examination report.pdf
357-kol-2003-granted-specification.pdf
Patent Number | 195162 | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Indian Patent Application Number | 357/KOL/2003 | |||||||||
PG Journal Number | 30/2009 | |||||||||
Publication Date | 24-Jul-2009 | |||||||||
Grant Date | 02-Sep-2005 | |||||||||
Date of Filing | 27-Jun-2003 | |||||||||
Name of Patentee | HIRANMAY SAHA | |||||||||
Applicant Address | IC DESIGN & FABRICATION CENTRE, DEPARTMENT OF ELECTRONICS & TELE-COMMUNICATION ENGINEERING, JADAVPUR UNIVERSITY, KOLKATA | |||||||||
Inventors:
|
||||||||||
PCT International Classification Number | G01L 9/06 | |||||||||
PCT International Application Number | N/A | |||||||||
PCT International Filing date | ||||||||||
PCT Conventions:
|