DESCRIPTION

I. Overview

[0018] Referring to FIGS. 1 a - 1 g, accelerometer 100 may be etched as a solid state structure, for instance, out of a block of silicon. Accelerometer 100 may include a movable portion 300 and a fixed portion 400 , each including a plurality of electrodes 112 , 114 , 116 , 118 , 122 , 128 generally formed as high-aspect-ratio beams or capacitor plates, each oriented in planes perpendicular to the silicon wafer and parallel to the z-axis 130 of the wafer along which acceleration is to be measured. Electrodes 112 , 114 , 122 of fixed portion 400 may be interdigitated between electrodes 116 , 118 , 128 of movable portion 300 , with capacitance gaps 142 , 148 between. Resilient mounting 120 holds movable portion 300 in an equilibrium rest position relative to fixed portion 400 (as shown in FIGS. 1 d and 1 e ), and allows motion of movable portion 300 relative to fixed portion 400 along z-axis 130 in response to acceleration (as shown in FIGS. 1 a, 1 b, 1 c, 1 f and 1 g ). As movable electrodes 116 , 118 , 128 move relative to the fixed electrodes 112 , 114 , 122 , some of the opposed pairs of electrodes come into more-direct opposition to each other (that is, as the surface area of movable electrode 116 comes to be more directly opposed to fixed electrode 112 , 122 , and movable electrode 118 , 128 comes to be more directly opposed to fixed electrode 114 ), and the capacitance increases. Other pairs of opposed electrodes move out of opposition to each other, and the capacitance between these pairs decreases. These changes in capacitance can be measured, giving a measure of the displacement of movable portion 300 relative to fixed portion 400 , and thus a measure of the acceleration imposed on the accelerometer 100 as a whole.

[0019] At least some of the capacitor fingers may be divided into two separate conductors (e.g., 112 , 122 and 118 , 128 ), with an insulating layer 132 , 138 separating the two conductors of a single finger. For instance, electrodes 112 , 114 , 116 , 118 may be formed in silicon layers (typically doped to improve conductivity), and electrodes 122 , 128 may be formed in a metal layer laid atop the silicon structural members. Capacitance 142 , 148 may be measured between pairs of electrodes ( 122 , 116 and 114 , 128 ) that are not directly opposed to each other when the accelerometer is in its equilibrium state. Capacitance 142 , 148 may be measured between electrodes of different materials—for instance, capacitance may be measured between metal electrodes 122 , 128 and silicon electrodes 114 , 116 . Some electrodes 112 , 118 may be grounded 152 , 158 , or otherwise electrically connected to consume field lines 154 , 156 , to reduce the total capacitance 142 , 148 between electrodes 114 , 116 , 122 , 128 .

[0020] To a first-order approximation, the capacitance 142 , 148 between metal electrodes 122 , 128 and their opposed silicon electrodes 114 , 116 is at a maximum when the metal electrode 122 , 128 is most nearly centered on the face of the opposing silicon electrode 114 , 116 , because of the shapes of the fringing fields around the electrodes. (Because of the field lines that terminate in the grounded electrodes 112 , 118 , the maximal capacitance is actually achieved when the metal electrode 122 , 128 is somewhat below the mid-point of the opposing silicon electrode 114 , 116 .) Thus, capacitance 142 between electrodes 116 and 122 increases as movable portion 300 moves up within fixed portion 400 (the motion depicted in moving from FIG. 1 d to FIG. 1 b ), and capacitance 148 between electrodes 114 and 128 decreases as metal electrode 128 moves out from between electrodes 114 (the motion depicted in moving from FIG. 1 e to FIG. 1 c ). Similarly, capacitance 142 between electrodes 116 and 122 falls as movable portion 300 , carrying silicon electrode 116 , moves down between electrodes 122 (the motion depicted in moving from FIG. 1 d to FIG. 1 f ), and the capacitance 148 between electrodes 114 and 128 increases (the motion depicted in moving from FIG. 1 e to FIG. 1 g ).

[0021] Electronics may measure differential capacitance. For instance, because the 114 -to- 128 capacitance 148 and the 116 -to- 122 capacitance 142 change in opposite directions with motion, and electronics (discussed in section II.E, infra) may measure the 114 -to- 128 capacitance 148 less the 116 -to- 122 capacitance 142 . That difference will generally reflect the degree of deflection of movable portion 300 . From that difference, electronics may determine the amount of acceleration imposed on accelerometer 100 .

[0022] Isolation joint 160 electrically isolates electrodes 116 on the left half of FIG. 1 a (and FIGS. 1 b, 1 d, and 1 f ) from the electrodes 118 , 128 on the right half of FIG. 1 a (and FIGS. 1 c, 1 e and 1 g ).

[0023] Resilient mounting 120 may be a torsional flexure, or a beam that acts in torsion, that provides for movement of movable portion 300 through torsional rotation (arrows 162 ). Various portions of resilient mounting 120 may also serve as conductors to drive electrodes 112 , 114 , 116 , 118 , 122 , 128 .

II. Structure

[0024] A. Major Structural Elements

[0025] Referring to FIG. 2 , accelerometer 100 may include a center backbone 204 , movable outer frame 210 , electrodes 112 , 114 , 116 , 118 mounted to center backbone 204 and outer frame 210 , and torsional flexure 120 . Center backbone 204 and fixed electrodes 112 , 114 may be anchored to the silicon wafer to form fixed portion 400 . Resilient torsional flexure 120 may be anchored to the wafer at anchor point 206 and may be otherwise released from the floor of the wafer (by undercutting, as described below in connection with FIGS. 6 c and 6 g ). Outer frame 210 may carry movable electrodes 116 , 118 , to form movable portion 300 . Outer frame 210 and movable electrodes 116 , 118 are released from the floor of the wafer, so that motion tends to be greatest at the right-most end 212 of movable portion 300 . Center backbone 204 , outer frame 210 and torsional flexure 120 may all be formed from truss-structured silicon members. The walls of torsional flexure 120 may be thinner than the walls of center backbone 204 or outer frame 210 in order to increase flexibility of torsional flexure 120 .

[0026] Anchor point 206 may be relatively small, to provide strain relief, or to allow movable portion 300 and fixed portion 400 to curl together in a common mode of deformation when the manufacturing process or temperature variations cause curling or bending. In other embodiments, anchor point 206 may extend farther along the edge of torsional flexure 120 , to provide rotational stability for movable portion 300 about z-axis 130 ( FIG. 1 a ). Torsional flexure 120 may be compliant, primarily in torsion, permitting rotation of movable portion 300 about axis x′. The rotation angle may typically be less than 0.04 degrees; therefore, the motion of movable portion 300 may be predominantly along z-axis 130 and proportional to the distance from the torsional flexure 120 . Maximum deflection of movable portion 300 relative to fixed portion 400 along the z-axis 130 perpendicular to the wafer may be on the order of tens to hundreds of nanometers. Off-axis motion (within the plane of the wafer) may be confined to an order of magnitude less, by designing appropriate stiffening truss structures into movable portion 300 and fixed portion 400 .

[0027] The overall size of accelerometer 100 may be about 1 mm×1.5 mm. The mass of the movable structure may be about 10 ⁻⁸ kg. The inertial moment of movable portion 300 around torsional flexure 120 may be about 5 to 6×10 ⁻⁸ kg m ² .

[0028] The resonant frequency of movable portion 300 within fixed portion 400 , moving in the z-axis direction 130 , may be about 1.3 to 1.4 kHz, and is desirably about 1 kHz for sensing accelerations in the 1 to 10 g range. For higher g accelerometers or higher frequency response, higher resonances are generally preferred and can be accommodated through stiffer torsional flexures 120 . Movable portion 300 may move relative to fixed portion 400 in other vibrational modes, for instance, rotating around the z-axis 130 , or rotating side-to-side around the y-axis (where the left half of FIG. 1 a rotates up and the right half down, for instance). In an ideal accelerometer, movable portion 300 would only move only in the z-axis direction, giving a finite resonance frequency for resonance in the z-axis direction, and would resist motion in all other directions, giving an infinite resonance frequency for all other vibrational modes. Thus, higher resonant frequencies are generally preferable for these off-axis motions, and resonant frequencies of between 5 and 10 kHz may be achieved in actual devices. For a given amount of acceleration along z-axis 130 , fixed portion 400 may deflect by about 1/29 of the amount of deflection of movable portion 300 . Generally, smaller ratios of deflection are better than larger, to the degree permitted by other engineering constraints.

[0029] Referring again to FIG. 1 a, typically, the silicon structures may be 20 to 40 microns high (dimension 240 ). The silicon elements may typically be 2 microns wide (dimension 242 ).

[0030] Referring to FIGS. 3 and 4 , movable portion 300 and fixed portion 400 may include repeated patterns of silicon and metal structures. Two important patterns are shown in FIG. 3 and FIG. 4 . The primary structure of movable portion 300 and fixed portion 400 may be formed of silicon beams 112 , 114 , 116 , 118 , 332 , 338 , 362 , 364 , 432 , 438 . These beams may be etched from a single-crystal silicon wafer. Before the silicon is etched, the top surface of the silicon may be oxidized to form an insulating layer of silicon dioxide of 0.5 to 1 microns, and metal may be laid on top of the silicon in the pattern shown in stipple. When the silicon wafer is etched to form the physical structure, metal overlaying the etched portions of the silicon may be removed as well, so that the metal remains only atop the silicon beams. Isolation joints 160 , 360 , 336 , 436 , 462 electrically isolate different portions of the silicon structure from each other. Vias 324 , 334 , 337 , 423 , 434 , 437 connect the metal layer through the insulating oxide to the underlying silicon.

[0031] B. Movable Portion

[0032] Four voltage potentials (which will be designated potentials 310 , 312 , 314 and 316 ) may be applied to the various components. Capacitances (including changes in capacitance, differential capacitance, or changes in differential capacitance) between pairs of these potentials may then be measured to determine acceleration.

[0033] Referring to FIG. 3 , in conjunction with the left half of FIG. 1 a, and FIGS. 1 b, 1 d and 1 f, potential 316 may be applied to silicon electrode 116 . Electrode 116 may be electrically contiguous with silicon beam 322 . Beam 322 may be electrically connected through via 324 to metal 326 . Metal 326 , 328 may connect drive and measurement electronics (see discussion in section II.E, infra) to apply potential 316 to silicon electrode 116 . Because silicon electrode 112 and metal electrode 122 are part of fixed portion 400 , detailed discussion will be deferred until section II.C, infra, and discussion of FIG. 4 . The gap between fingers 112 , 116 and 114 , 118 may be about 3 microns.

[0034] Referring to FIG. 3 , in conjunction with the right half of FIG. 1 a, and FIGS. 1 c, 1 e and 1 g, ground potential 310 may be applied to silicon electrode 118 . Electrode 118 may be electrically contiguous with silicon beam 332 . Silicon beam may connect through via 334 to metal 335 , which crosses isolation joint 336 , and reconnects to silicon beam 338 through via 337 . (Vias 334 , 337 , isolation joint 336 and metal 335 may not be required by electrical considerations; via 334 and isolation joint 336 may serve to improve the match between the thermal expansion of leg 332 and the thermal expansion of leg 322 .) Drive and measurement electronics may apply ground potential 310 to silicon beam 338 . Metal electrode 128 may be connected through metal 342 to drive and measurement electronics, which may drive metal electrode 128 at potential 312 . Because silicon electrode 114 may be part of fixed portion 400 , detailed discussion will be deferred until section II.C and discussion of FIG. 4 .

[0035] Isolation joints 160 , 360 may electrically isolate portions of the silicon from each other. For instance, isolation joints 160 may isolate electrodes 116 (electrical potential 316 ) on the left half of FIG. 3 from electrodes 118 (ground potential 310 ) on the right half of FIG. 3 . Isolation joints 360 may isolate electrodes 116 (electrical potential 316 ) from silicon beams 338 (ground potential 310 ). Isolation joints 160 , 360 may be formed as follows. Slits or trenches may be etched into the wafer, in the locations that become isolation joints 160 , 360 . During the same thermal oxidation process that forms the oxide layer on top of the entire wafer to insulate metal layer 122 , 128 from the underlying silicon, silicon dioxide may be grow on the wafer to fill in the trenches. This growth may cause the two opposing faces of silicon dioxide to fuse to each other. Further, the growth of silicon dioxide around the circular ends of the trenches may provide a connection across the two sides of the isolation joint. Together, the fusing of opposing faces and growth across the ends of the trenches may provide sufficient structural integrity to provide mechanical support for electrodes 116 , 118 on silicon beams 362 , 364 .

[0036] Metal laid across the tops of isolation joints 160 , 360 is electrically insulated from the silicon on both sides of the isolation joint, but is electronically continuous across the top of the isolation joints.

[0037] Conducting vias 324 , 334 , 337 , 423 , 434 , 437 , etc. may be formed in the conventional manner. In FIG. 3 , they are shown as slightly wider beam region than the silicon beam regions immediately adjacent. Beams may be widened where vias are placed in order to keep the vias interior to the beam geometry.

[0038] C. Fixed Portion

[0039] Referring to FIG. 4 , in conjunction with the right half of FIG. 1 a, and FIGS. 1 c, 1 e and 1 g, potential 314 may be applied to silicon electrode 114 . Electrode 114 may be connected through beam 422 through via 423 to metal 424 , which may run over isolation joint 462 out to the edges of the device. Drive and measurement electronics may apply potential 314 to metal 424 .

[0040] Referring to FIG. 4 , in conjunction with the left half of FIG. 1 a, and FIGS. 1 b, 1 d and 1 f, ground potential 310 may be applied to silicon electrode 112 . Electrode 112 may be electrically contiguous with silicon beam 432 . Beam 432 may connect through via 434 to metal 435 . Metal 435 may cross isolation joint 436 , to via 437 , which may in turn connect metal 435 to silicon beam 438 . Drive and measurement electronics may apply ground potential 310 to beam 438 . (Vias 434 , 437 , isolation joint 436 and metal 435 may not be required by electrical considerations; via 434 and isolation joint 436 may improve the match between thermal expansion of leg 432 and thermal expansion of leg 422 .) Metal electrode 122 may be electrically connected to metal 424 across isolation joints 460 . Potential 314 may be applied to metal 424 as discussed above.

[0041] Isolation joint 460 separates silicon electrode 112 from silicon electrode 114 .

[0042] D. Mechanical Stops

[0043] Referring to FIGS. 5 a and 5 b, recall that movable portion 300 may be held by anchor 206 about 10 microns above a “floor” of the silicon substrate that remains after the etching process. Downward motion of movable portion 300 may be contained when the right edge 212 of frame 210 reaches this floor. Excessive upward movement of movable portion 300 may be contained by a stop 510 that extends in the opposite direction from the capacitive fingers of movable portion 300 , so that as movable portion 300 moves up, stop 510 moves down until it makes contact with the substrate floor. The length of stop 510 may be anywhere from one-fourth as long as the distance from anchor 206 to right edge 212 (allowing movable portion to move up four times as far as it can move down), to essentially the same length as the distance from anchor 206 to edge 212 (confining both ranges of motion roughly equally). It may be desirable that stop 510 have a low moment of inertia around anchor 206 , to reduce the attenuation of response of movable portion 300 in response to acceleration.

[0044] Stop 510 may end with fingers 512 that are interdigitated with fingers 514 mounted on fixed portion 400 . Similarly, stops 520 mounted on movable portion 300 may be interdigitated with fingers 522 mounted on fixed portion 400 . Stops 512 , 520 constrain rotation and translation of movable portion 300 in the plane of the wafer.

[0045] E. Drive and Measurement Electronics

[0046] An accelerometer as described above may have a sensitivity in the range of 10 to 15 fF/g (femtofarads per g of acceleration).

[0047] As discussed in section 1 , supra, accelerometer 100 may use a differential capacitor approach. In a differential capacitor arrangement, for an acceleration in one direction, capacitance increases between one pair of electrodes, and capacitance decreases between the other pair. For acceleration in the opposite direction, the changes in capacitance are reversed. Thus, the difference between the capacitances indicates the amount of acceleration. An ASIC (application-specific integrated circuit) converts the capacitance difference into a voltage that represents acceleration.

[0048] In one design, the ASIC places equal but opposite square wave voltages across the two capacitors and integrates the difference of the capacitor currents. The output of the integrator will be a voltage that is proportional to the difference in capacitance. This voltage is then amplified and low pass filtered to give the desired sensitivity and frequency response. A programmable voltage can be added or subtracted from this signal to provide for an offset adjustment. Additionally the gain of the capacitance-to-voltage conversion can be programmed to account for sensor performance distributions and different sensor designs.

[0049] Just as an electrical generator can function as a motor if the proper electrical current is applied to the generator outputs, so voltages can be applied to conductors 310 , 312 , 314 and 316 of accelerometer, to cause movable portion 300 to move relative to fixed portion 400 . By altering the carrier signals used to sense the capacitance difference it is possible to implement a self-test mechanism. In self-test mode, electrical signals 310 , 312 , 314 and 316 are driven so that movable portion 300 is displaced, to verify that the movable structure 300 can move and that the appropriate capacitance change results. An electrical force is always generated by voltages such as the carrier signals for sensing. However, under normal operation the carrier signals are balanced and no net force arises. By altering the carrier signals such that the RMS voltages are not the same on the two sides of the differential capacitor used for sensing, a net force results. The net force causes a relative motion between the fixed portion 400 and movable portion 300 . This is a standard self-test method used in most commercial accelerometer ASIC's.

[0050] ASIC's operating under this principle, as well as other techniques for translating a capacitance change into a voltage representing acceleration, are available from a number of universities and companies, including Kionix, Inc., Bosch GmbH, and MicroSensors, Inc. of Costa Mea, Calif.

III. Fabrication

[0051] The overall silicon structure may be manufactured using silicon fabrication technologies available from Kionix, Inc. of Ithaca, N.Y. This is a mature process that is well suited to mass production. The Kionix process is an all-dry process, and lithography steps are carried out on planar surfaces.

[0052] Referring to FIGS. 6 a - 6 g, accelerometer 100 may be fabricated using a plasma micromachining process. One such plasma micromachining process may use four masks and industry-standard silicon wafers. The first mask may define trenches that are etched into the silicon to form isolation joints. As shown in FIG. 6 a, these trenches may be filled with silicon dioxide 612 . Using the second mask, vias 620 may be defined and opened in field. Implants 622 may be made and aluminum 624 may be deposited. As shown in FIG. 6 b, using the third mask, metal 624 may be patterned to break electrical connections where necessary. The fourth and final mask may be used to define the structural beams. The profile of the structural beams may etched into the silicon using a production ICP silicon etcher, for example, a PlasmaTherm VLR 770 with ICP Bosch Etch & ICP Oxide Etch Chambers, resulting in the structures shown in FIG. 6 e. The sidewalls may be passivated 630 with a deposited layer of silicon dioxide. The oxide on the trench bottoms that surround the beams may be cleared using an anisotropic silicon dioxide etch, while the sidewall passivation 630 remains, yielding the configuration of FIG. 6 f.

[0053] Finally, as shown in FIGS. 6 c and 6 g, the silicon may be etched isotropically to release 640 the beams 642 from the substrate 644 . (Beam 642 may be any one of electrode fingers 116 , 118 , beams 332 , 338 , torsional flexure 120 , or any other portion of movable structure 300 . In some embodiments, fixed structure 400 may also be released from substrate 644 , and the truss structure of f) “Tails” 646 of oxide may extend below the silicon of beams 642 . These tails 646 may provide added stability in thermal expansion, because tails 646 may counter-balance any bending moment imparted by thermal expansion or contraction of the oxide 648 at the tops of the beams. In embodiments in which the release etch is carried out as a dry-etch process, stiction between adjacent structures or between structures and the substrate floor may be reduced to negligible levels, or so as to be non-existent.

[0054] Referring again to FIGS. 3 and 4 , in conjunction with FIGS. 6 c and 6 g, in some embodiments some of the fingers 112 , 114 , 116 , 118 may be omitted, to make easer the step of FIGS. 6 c and 6 g in which beams 114 , 118 , 332 , 338 , 642 are released from the substrate. On the other hand, preserving all of the electrode fingers increases the sensitivity of accelerometer 100 .

[0055] Typical beams 642 generated by the plasma micromachining process are 2 μm wide, 10 to 30 μm tall, and separated from the substrate by 15 μm. Structures that are larger than 10 μm wide generally do not release from substrate 644 during the isotropic etch. Such wide structures may provide the points where the movable or fixed structures anchor to the silicon substrate.

[0056] To form large structures on the order of millimeters, typically the beams are laid out in an open cellular structure, as shown in FIG. 2 . Such layouts aid in achieving higher oscillation Q's (a high Q-factor oscillator is one that oscillates consistently at the same frequency, a low Q oscillator may resonate a different frequencies depending on the impulse applied). A high oscillation Q may in turn improve precision in accelerometer 100 . Since the structures formed from the plasma micromachining process are predominantly composed of stress-free, single-crystal crystal silicon, which is a well-characterized and reproducible material, the performance of the structures may be predictable and reproducible.

[0057] Processes for forming accelerometer 100 , and isolation joints 160 , 360 , are discussed in U.S. Pat. No. 6,239,473, Adams et al., Trench Isolation Process for Microelectromechanical Devices, U.S. Pat. No. 5,719,073, Adams et al., Microstructures and Single-mask, Single-crystal Process for Fabrication Thereof, U.S. Pat. No. 5,846,849, Microstructure and Single mask, Single-crystal Process for Fabrication Thereof, U.S. Pat. No. 6,051,866, and S. G. Adams et. al., “Single-Crystal Silicon Gyroscope with Decoupled Drive and Sense,” in Micromachined Devices and Components V, Patrick J. French, Eric Peeters, Editors, Proceedings of SPIE Vol. 3876, 74-83(1999), K. A. Shaw, Z. L. Zhang, and N. C. Macdonald, “SCREAM I: A single mask, single-crystal silicon process for microelectromechanical structures,” Sensors and Actuators A, vol. 40, pp. 63-70, 1994, and Z. L. Zhang, N. C. MacDonald, “A rie process for submicron, silicon electromechanical structures,” J. Micromech. Microeng., v2, pp. 31-38, 1992, all of which are incorporated herein by reference.

IV. Alternative Embodiments

[0058] In another embodiment, electrodes 112 , 114 , 116 , 118 may be formed out of a thick polysilicon layer deposited on a silicon substrate wafer, with the silicon substrate serving only as a structural substrate or as shielding, and not playing an active role in sensing.

[0059] In another embodiment, electrodes 11 , 114 , 116 , 122 , 128 may be formed in multiple SOI (silicon-on-insulator) layers laid on the silicon substrate wafer.

[0060] In another embodiment, electrodes 112 , 114 , 116 , 118 , 122 , 128 may be formed in multiple metal layers laid on the silicon, with the silicon serving only as a structural substrate or as shielding, and not playing an active role in sensing.

[0061] Referring to FIG. 7 , torsional flexure 120 may be reconfigured to reduce motion in directions other than the z-axis 130 (up and down out of the paper) while preserving compliance for z-axis motion. For instance, torsional flexure 120 may configured in a more triangular shape, to maintain good torsional compliance - and freedom of movement of movable portion 300 along the z-axis—while maintaining good stiffness in other directions and resistance to other vibrational modes.

[0062] Referring again to FIG. 7 , additional structures 710 , 712 , 714 may be incorporated to maintain a relatively uniform density of structure, to improve the uniformity with which the fabrication steps operate. In alternative embodiments, the spacing between some elements may be increased, to ease the release step (see FIGS. 6 c and 6 g ).

[0063] The placement of electrodes 112 , 114 , 116 , 118 may be arranged to reduce temperature-sensitive curvature of the device that may result from fabrication.

[0064] The internal truss structure of movable portion 300 and fixed portion 400 may be configured to improve rigidity.

[0065] In the embodiment shown in FIG. 2 , movable portion 300 is arranged as a “diving board,” cantilevered toward a single side of anchor point 206 . In other embodiments, movable electrodes 116 , 118 may be arranged as a “teeter totter,” arranged bilaterally about a central resilient mounting. In other cases, electrodes 112 , 116 may be formed on one side, and electrodes 114 , 118 on the other. In other cases, movable electrodes 116 , 118 may be arranged on one side of anchor point 206 , and a dummy mass may extend from the other side of anchor point 206 . The two sides of the teeter totter will generally have different rotational moments about the resilient mounting, so that acceleration will induce rotation.

[0066] It should be understood that all dimensions, electrical values, geometrical aspects, fabrication technologies, etc. describe only some example embodiments as they may be preferred in 2000-01. As new fabrication technologies emerge, these values may change.

[0067] For the convenience of the reader, the above description has focused on a representative sample of all possible embodiments, a sample that teaches the principles of the invention and conveys the best mode contemplated for carrying it out. The description has not attempted to exhaustively enumerate all possible variations. Further undescribed alternative embodiments are possible. It will be appreciated that many of those undescribed embodiments are within the literal scope of the following claims, and others are equivalent.