### Create a StudySoup account

#### Be part of our community, it's free to join!

Already have a StudySoup account? Login here

# TOP FOURIER OPTICS ECE 510

PSU

GPA 3.78

### View Full Document

## 6

## 0

## Popular in Course

## Popular in ELECTRICAL AND COMPUTER ENGINEERING

This 303 page Class Notes was uploaded by Miss Chadrick Doyle on Tuesday September 1, 2015. The Class Notes belongs to ECE 510 at Portland State University taught by Staff in Fall. Since its upload, it has received 6 views. For similar materials see /class/168243/ece-510-portland-state-university in ELECTRICAL AND COMPUTER ENGINEERING at Portland State University.

## Reviews for TOP FOURIER OPTICS

### What is Karma?

#### Karma is the currency of StudySoup.

#### You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!

Date Created: 09/01/15

Between Failure Reliability Yield and IC Layout ECE 510 DeVice Reliability Spring 2003 Lecture 3 Yield Enhancement through Layout changes Malgorzata Chrzanowska Jeske Electrical and Computer Engineering Portland State University 510 RetiaJIiLigv 7 Lecture 1 5M9 1 Layout Techniques for Yield enhancement 30 510 Reliallt39lhy r Letrm 3 5M9 2 Yield Enhancement at the Design stage Adding redundancy increases the chip size Modifications at the design stage can decrease the critical area without increasing the chip area Decreasing the critical area increases the yield 511 Relialliblgv r Lmure 3 5M9 3 Effect of Decrease in Critical Area 38 Percentage of Yield Improvement Wage Olineld 5 Reliability 7 LA crure 3 5M9 4 Critical Areas Critical Area must be averaged over all defect sizes x and defect types 139 Averaging over defect diameters x pox 7P0 xxdx Cntlcal Area Calculation To predict yield critical area or pof need to be calculated Various methods Geometrybased for calculating Alix polygon expansion techniques etc Monte Carlo approach for calculating pofl x simulated circles representing defects of different sizes randomly placed in the layout the fraction of defects I and diameter X that cause fault estimation for poflx WWW rum 5MB Yield Enhancement E eCt39Ve Y39e39d Adding redundancy increases yield Three levels for yleld enhancement It also increases the chip area less chips on wafer Even With higher yield a number of good chips Architecture redundancy could be smaller Yield might not be a good measure for circuits D 39 d 39 th 39t39 1 Wimr dund m eSIgH ecreasing e on 10 area Effective Yield takes into account increased chip placement routing compaction a 17 Amp no redundancy Process decreasrng defect densrty Ym Ym W 5117 mm rLeaum is 51m 7 m7 mmm 714mm 3 SM 5 Effective Yield um Euciwicld imld Yield 33 Modification of the compaction algorithm for critical area 0 z A a 10 reduction Rmm quot Reduction in Critical area about 8 The maximum value of effective yield the optimal redundancy 33 y r 510 mm mm 5mg szvmtumw mm mm Critical Area Reduction Two Approaches to Yield enhancement during Compaction stage Local modification in the layout as a postcompaction step Short circuit type faults Original Layout Modi ed layout 38 510 Rummy 7 Led n 5M2 M Effect of Uniform Spacing Probability of failure vs Defect size a now a i m m De ant aim uniamm r in u m Dam size Mm Short circuit faulm Open Circuit faulm as 39n 739 5w 1mm 7 mm a sum 12 Open circuit faults Modi ed for the same defect density Modi ed for higher short circuit defect density 517 Ralmbttny rLecmre 3 Slide 13 Effect of Wider Wires Probability of failure vs Defect size K On39y39x 1 f mm 7 r ml 7 m r 1 m v m 15 In Defeat am moms Dam Him mime 381 Short circuit faulm Open circuit faulm 5w Rwawa r Leann a Slide m Wire Spacing Routing Improvement Better yield by reducing the critical area Better electrical behavior by changed capacitance 39 v VA 7 mummy rumquot 01 5111215 New Material Implications 0 Lower dielectric 7 Reduces total capacitance 7 Doesn t change crossground proportions 0 Copper metallization 7 Reduces RC delay 7 Avoids electromigration 7 Thinner deposition reduces cross cap 0 Multi layers of routing 7 Relative routing pitch may increase 7 Room for shie 39 xv 510 Rabatth r mum a Slide 15 Routlng Techniques Increasing Wire spacing 7 smaller cross coupling capacitance 7 higher ground capacitance 7 smaller probability for shorts 7 better process uniformity Wire Widening 7 smaller resistance 7 smaller probability for opens Wire shielding 7 lower cross coupling noise Via strengtheni 7 smaller resistance 7 higher reliability x W 7 5mm Wire Spacing Techniques Need for smart jogging Work on all layers simultaneously Move around Wires an vias Don t change relative position of Wires and vias Maintain full hierarchy Handle all routing styles Handle full chip striping multiprocessing Support dn ttouch signals PG clocks Support net speci cations clearance Width Support pairWise net proximity w m mmm 4mm 2 511mm Wire Spacing and Shielding Prerouting speci cation 7 ConvenienL handled by router 7 Robust but conservative 7 May consume big area Postrouting speci cation 7 Area ef cient shield only Where needed and free space 7 Ease task of router 7 suf cient shielding is not guaranteed m mmm 45ng at SM 1 9 Opportunities for Via Strengthening Add cut holes Where possible 7Wire Widening may needlargermore vias 7 Non square via cells Increase metalvia overhang 7 Non uniform overhang m7 mmm 714mm 3 SM 2n Wire spacing example Before Spacing 39 A er SpaCing A1quot 510 Reimbilhy me 3 Slide 21 Wire shielding 7l 1 V 139 l DD l ampi T T Substmte Mi mummy me Slide 22 Performance yield enhancement 0 Reducing cross coupling capacitance 7 Increase clock speed 7 Reduce power consumption 7 Smaller capacitive load enables smaller devices 40 510 Reliabiliw me 3 Slide 23 Yield Enhancement 3 2 Wan In Poly tuning and Paly Moved 2h Swing Umauhad SiFiX Implementation of larger PolyContact extends without modifying Diffusion mask 510 mummy ham 3 Slide 24 Yield and Reliabilty Vias In dualdamascene Cu interconnect l30nm processes Vias between metal layers in the interconnect stack Speci c to dualdamascene copper metal stacks Some vias appeared to be correctly imaged amp formed were not connected or presented such a resistance that they had a maj or electrical impact on the circuit Vias It was discovered that those effects could appear during accelerated life testing in chips that had shown neither initial defects nor to the concern of reliability engineers infant mortality w r u Hzoelmtm MW 3 5M 25 m mum Lm 1 W 26 Failure mechanisms In Vias m i mum mm tn 17mmquot andstrcssmmuihmmmwamn m MSW Mammal salt mw F El Three mechanisms Wilmi vniag 1 1 wa one or more small voids in Cu that forms Via r 39 it ll 15 1 due to electromigmtion l WWW l mm mm mm man ms 5 2 due to thermalstress migration 7 WWW the void can join migmte to the bottom of the via i all 5351322le l W W Minimumquot and open the connection between the Via and the l n i M W barrier over the underlying metal layer am ms m quoti plum Y T nmmaign f 5m minimal um W 3 thermal stresses are so great that the via metal physically tears away from the barrier forming a void quotquot 39 quot quot mil in m um 40 i at m m n my mu at m m m m Reliabilm mm 3 5M 27 W 28 as in Cu process The thermal cycle lies within the range of temperatures used in wafer processing Failure can also occur in accelerated life tests and normal operation on some chips The mechanism beoomes a major issue It occurs primarily in upperlayers m 340mm 44va u of the metal stack where copper traces are widest 5m 29 Current Coping Strategies Design solutions do not place isolated vias under wide areas of metal by separating a wide metal run into several narrower strips in the immediate area of a via by placing multiple redundant vias under wider chunks of coppe it has been incorporated into some design rules m1 mm Jamquot 1 5m cm Addition of Redundant Vias I Extm via or contact could be introduced to partner a single nonredundant via I This is particularly true of routing Where many tmcks have suf cient space available for an extm via I This extmvia enables a single via failure to be tolemted This assumes that via failure is an isolated mndom event that does not impact the partnering via m Rumba 7mm a 5m 1 Redundant Vias 0n track lay 0n r 1 added via track layer 2 added via Offtrack added via immommy 7141va 5m 32 Redundant Vias The via addition is done suc ias can be restricted to be place on only one of the conductor layers e other conductor layer eg metal3 and the via layer are modi ed assuming no extra overlap of the via is required This restricted modi cation might be used Where the cost of modifying more than two masks was not cost effective 51221233 Redundant Vias The via placement can also be restricted to placement on either of the existing conductor layers on track placement On track placement minimizes the addition of extra conductor material Important if intralevel shorts are predicted to cause a high level of faults m mmm 4mm 2 5m u Redundant Vias The placement strategy Which obtains the highest number of additional vias attempts to place vias on track rst If this cannot be achieved attempts to place the via off the existing metal tracks beside the existing via off track my mmm 4mm dz 51221235 Wire Displacement Wire displacemean can be used to reduce the critical area of a layout ie they can reduce the probability that a defect falling on the chip Will cause a fault Fault maps indicate regions of a layout Which are most sensitive to defects m7 mmm 49mm 3 51221235 Wire Displacement Original Layout Fault Map Critical Areas l Wire displacement The amount of critical areaquot is reduced at the displacement site 0 smaller defects darkest critical area can no longer cause a fault at the displacement site U L 0 larger defects are less likely to cause a fault Fault Map Critical Areas Displaced Track 41 l 510 Reimbilhy Lmunz 3 554 37 511 llzb39alziligz Lacmrz 3 554 38 Critical Area Reduction in Routing LaYOUt compacuon Movestretchshrink objects while preserving The ElTectiveness of enhancements during compaction topology depends on the previously obtained layout Movestretchshrink to Two approaches to yield enhancement in routing stage 7 Fix DRC Violations migrate X bad layout Local modi cations of layout as a postrouting step Optimize Objective mc on Area Modi cation of the routing algorithm for yield Yield enhancement DelayCrosstalk etc 510 Reliabiliw LEL DHE 3 39 510 Rtll39abiliw Lamrz 3 4n References we umasseduec crmyxel o h Cadmoe mwnal s 38 1 Korm mwnal Yxeld SmmcalModelmg andEnhanmmLTechmqueg mapAw a dk dy Ls des pdf oluums LawaLmlModx caums for Paformanoe and Yxeldquot vlsxcadsuda uD c99rUT0R1ALmaanm 40 Km wusm Faxluresplague Banancmem 1c processes EETxmes August 27 2002 mp WWW eeJmes canstayOEGZOUZOSZ SOOZ 41 GamiA Allan YxeldRehabxhtyFnhancanmtusmgAutOmatedMnorLayout Mom cancms quot Unwasxty ofEdmb Edmburgq EH9 3JLUK mp waseeedac YFPEYEhtml amman andI Korm deeon enabmty 311mm Free We 1996IEEEIrLcemzmaml 5y 05m 71 Defeame alermce m Vmsyxnem pp 67775 November 19 43 CadmceTuwnal Solutions 5th e quayouLquot vlsxcad cs ucla eduDAC99TUTORIAL s between Yxeld and R quot F 96 m mum rum 3 5mm ECE 510 OCE BDDs and Their Applications Overview 0 Connection between variable ordering and BDD size 0 The size of BDDs for symmetric functions 0 Exponential lower bounds o Exact minimization is NPcomplete o The fanin based heuristic to find good static variable ordering for BDDs representing outputs of logic networks Dynamic reordering based on variable swap Reordering algorithms window permutation sifting block sifring symmetric sifting April 9 2000 ECE 510 OCE BDDS and Their Applicanons Variable Ordering for Fa1b1 azbz V i I i U D a1ltb1lta3lt b2 alltalltb1lt b2 April 9 2000 ECE 510 OCE BDDS and Their Applicaiions BDDs for Symmetric Functions 0 The value depends only on the number of 1 s in the input vector 0 For nvariable functions there are n1 possible numbers of 1 s and so 2n1 functions 0 Cofactors of symmetric functions themselves are symmetric functions 0 Ik the function is cofactored k times there are no more than n1k different cofactors o The BDD size is limited by nn1k On2 April 9 2000 ECE 510 OCE BDDS and Their Applicaiions Exponential Lower Bounds o The dependence of the BDD size on variable ordering is very strong It would be good if for each function at least one ordering gave a small BDD However this is not true BDD representations shares the fatal property of all other representations the representations for nearly all functions take exponential space The reason is that there are 2 functions over n variables This number is so large that it is impossible to have polynomial size representation for all but a very small percent of the functions 1r April 9 2000 ECE 510 OCE BDDS and Their Applicalions Exact Minimization of BDDs 0 Because the size of BDDs depends on variable order algorithms for constructing good orders are of great practical importance 0 It can be shown that the test whether the given BDD has the minimum possible size is an NPcomplete problem 0 Hence it is an NPcomplete problem to construct the optimal oder for a given BDD and efficient algorithms cannot be expected April 9 2000 ECE 510 OCE BDDS and Their Applicalions Exact Minimization Historical Picture 0 in the original paper introducing BDDs Randy Bryant stated without proof that the construction of an optimal order is NPcomplete STani KHamaguchi SYijima proved the weaker case for shared BDDs BBoig and IWegener proved the theorem for BDDs with exactly one root April 9 2000 ECE 510 OCE BDDS an Optimizing Variable Ordering Depending on the initial representation Boolean formula net list SOP etc the problem of finding a good stat39c order may have i Often i can be used if i ll ction be already April 9 2000 ECE 510 OCE BDDS and Their Applicalions Fanin Heuristic for Static Ordering l for the node of the combinational boolean network is 0 if it is an output node or Zlurriagtltwi TFil wi 3 otherwise 0 Heuristic transform the network to have one output perform a depthfirst search of the network starting from the outputs and when there are two or more predecessors prefer the one with the largest TFI the earlier the variable appears in a predecessor s variable list the earlier it is put in the global list April 9 2000 ECE 510 OCE BDDS and Their Applicaiions April 9 2000 ECE 510 OCE BDDS and Their Applicaiions Motivation for Dynamic Reordering Static methods are too problem specific for example the fanin heuristic exploit the topology of the circuit 0 The methods work heuristically Static ordering even if good is not ideal In order to succeed in many cases we need to change the variable ordering as the computation proceeds April 9 2000 ECE 510 OCE BDDS and Their Applicaiions Variable Swap is a Local Operation April 9 2000 ECE 510 OCE BDDS and Their Applicaiions Algorithms for Dynamic Reordering Window permutation Sifting algorithm 0 Block sifting Symmetric sifting Blockrestricted sifting April 9 2000 ECE 510 OCE BDDS and Their Applicaiions Quantitative Comparison April 9 2000 ECE 510 OCE BDDS and Their Applicaiions ECE 510 OCE BDDs and Their Applications Lecture 7 1 BDDs for Symmetric Functions i 2 Constraint Satisfaction Problems April 18 2000 Alan Mishchenko Overview BDDs for symmetric functions Characteristic functions of sets and sets of sets 0 Building BDD for Tuples k out of nquot Constraint satisfaction problems 0 Example Nqueen Problem Binate Covering Problem 0 Solution of BCP using Shortest Path on BDDs April 18 2000 ECE 510 OCE BDDS and Their Applicaiions BDDs for Symmetric Functions 0 The value of a symmetric function depends only on the number of 1 s in the input vector 0 For nvar functions there are n1 possible numbers of 1 s and so 2n1 different functions 0 Cofactors of symmetric functions are symmetric functions 0 If the function is cofactored k times there are no more than n1k different cofactors o The BDD size is limited by nn1k On2 April 18 2000 ECE 510 OCE BDDS and Their Applicalions Characteristic Function of a Set Function F fin gt 1 defines a subset of minterms of 39 on which it is l Given a binary encoding of a set of elements of a subset of this set is a boolean function which is 1 for minterms encoding the subset and ii for other minterms April 18 2000 ECE 510 OCE BDDS and Their Applicalions Example 7 3 GiVen the set 30 31 32 33 34 p5 and its encoding P0 T 31 10 33 7 DE 1 0 find characteristic function of subset p0 p2 p3 and represent the subset using BDD o Ti Define a function over the encoding variables 2 such that it is equal to i for minterms representing subset p0 p2 p3 pp0 p2 p3 J V 1 IJ 51 April 18 2000 ECE 510 OCE BDDS and Their Applicalions BDD Representation of the Characteristic Function April 18 2000 ECE 510 OCE BDDS and Their Applicalions Char Function of Set of Subsets 0 Let the set S p1 p2 pn contain n elements Assuming positional notation the elements of the set are encoded using one variable per element f Suppose n 5 and the subset is p1 p3 p4 This subset can be represented by the characteristic function pan 393 p4 xlxz x3x4x5 Example For the same set with n 5 the set of all subsets containing exactly one element is represented by the characteristic function p1 xlxz x3 x4 x5 xl x2x3 x4 x5 xl xz x3x4 x5 xl xz x3 x4x5 xl xz x3 x4 x5 April 18 2000 ECE 510 OCE BDDS and Their Applicaiions Char Function of Tuples k out of nquot Given the set S of n elements build a BDD for the characteristic function of the set of subjects containing exactly k out n elements of set S o The brute force approach results in creating the sum of nknk cubes each of which represents a characteristic function of a subset o The intelligent approach uses a recursive BDD procedure to build the char function April 18 2000 ECE 510 OCE BDDS and Their Applicanons Building Char Function of Tuple Set bdd Tupies int k int n if k lt 0 II n lt k return bddfalse if k 0 ampamp n 0 return bddtrue bdd F0 Tupies k n 1 bdd F1 Tupies k 1 n 1 bdd Res bddite bddithvarn F1 F0 return Res April 18 2000 ECE 510 OCE BDDS and Their Applicaiions BDD for Tupe 2 3 JL April 18 2000 ECE 510 OCE BDDS and Their Applicanons Constraint Satisfaction Problems 0 The parameters of SAT problems are represented using binary variables and the requirements for a solution are expressed as boolean formulas over the binary variables 0 When all these formulas are multiplied we get a BDD whose path lead to constant 1 for those assignments that satisfy the problem 0 A SAT problem is solved by analyzing this BDD for example by counting the number of paths or finding the shortest path April 18 2000 ECE 510 OCE BDDS and Their Applicaiions NQueen Problem 0 Is it possible to place n queens on a n x n chess board so that no queen can be captured by another queen If yes how many different placements are possible 0 Let us encode the presenceabsence of a queen in each cell by a binary variable altogether we need n x n binary variables 0 Now the requirements of not capturing can be expressed in terms of these varables April 18 2000 ECE 510 OCE BDDS and Their Applicaiions Encoding the Constraints 0 There is no more than one queen in each row each column and each diagonal Xij gt HXik lleltn k j Xij gt Hij ISmSn m I etc There is a queen in each row Xi1 Xi2 XiN After multiplying the constraint we get the BDD representing all solutions a for 8 queens there 2450 nodes and 92 solutions S Minato Calculation of Unate Cube Set Algebra Using ZBDDs DAC 94 httpwwwsigdaacmorgArchivesProceedingArchives April 1 8 2000 ECE 5 10 OCE BDDS and Their Applicaiions 13 Binate Covering Problem BCP 0 Given a covering table and the objective function min2jwjxj BCP is formulated as follows 0 Find a subset S of columns of minimum cost according to the objective function such that for every row fj either 1 Eli fij1 amp H e S or 2 Eli fij0 amp H e S This problem arises in incompletelyspecified FSM state minimization technology mapping decomposition of functions with DCs BDD minimization etc BCP is reduced to the problem of finding the shortest path on the BDD representing the product of all the constraints April 18 2000 ECE 510 OCE BDDS and Their Applicanons Solution 1 X1IX2 Solution 2 X3IX4 April 18 2000 ECE 510 OCE BDDS and Their Applicaiions BCP in State Minimization of ISFSMs 0 Covering constraints c1c11c1c2c5c2 c3c5 c6c7c8 c1 c2c4 c6c10c1c4c3 c7c9c12 C3 C8C9C11C4C10 0 Closure constraints C2C1C2IC11C2IC1C4C3I C2C6 C3C4C4IC1C4IC1C6C11C6 C1C4 C7C2C6C8C2C6C8C3C9C9IC4 0 Solution c1c4c5c91 April 18 2000 ECE 510 OCE BDDS and Their Applicaiions 1 11 lIll lrersalH SETH Limit117 hith Nanotechnology and Biosensors Dr Shalini Prasad Electrical and Computer Engineering Biomedical Microdevices and Nanotechnology Laboratory sprasadpdxedu httpwwwece dxedu rasads ECE 510 Lecture 3 Sprung 20 8 S Prg ggs Course Outline 1 Introduction to BioMEMS and Nanobiotechnology 2 Silicon Microfabrication 3 Soft Fabrication Techniques 4Poymer Materials and Microfluidics 5 Sensor Principles 6Detection and Measurement Methods 7 Drug delivery systems 8 MicroTotal Analysis Systems uTAS Emerging Applications 9 Nanotechnology Applications ECE 510 Lecture 13 Spring 2008 S Prg goyl SIIde2 s SS Salitern ian Sensors o A sensor converts one form of energy to another and in so doing detects and conveys information about some physical chemical or biological phenomena 0 More specifically a sensor is a transducer that converts the measurand a quantity or a parameter into a signal that carries information ECE 510 Lecture 13 Spring 2008 5 Prgffg SS Ideal Sensor 0 Continuous operation without effecting the measurand 0 Appropriate sensitivity and selectivity 0 Fast and predictable response 0 Reversible behavior 0 High signal to noise ratio 0 Compact 0 Immunity to environment 0 Easy to calibrate ECE 510 Lecture 13 Spring 2008 S P331355 Issues with Sensors Sensors are not ideal when designed as a standalone device or as part of uTAS detection scheme they may affect the measurand Example Biomolecule samples guided through micro channels must be separated immobilized for measurement purposes as a part of a continuous flow process Mechanical alteration of samples may occur ECE 510 Lecture 13 Spring 2008 S Prg ggisa Categories 0 Biosensor 0 Smart sensor 0 Passive or active 0 Sensor array 0 Multisensor o Multifunction ECE 510 Lecture 13 Spring 2008 S PrcaafnggsS giggles Sensor Categories Biosensors measure or indicates living things and their products Analyzers measure the constituents of a substance and may have one or more biosensors Control Systems regulate the value of the measurand to some specific value typically as a part of the feedback loop ECE 510 Lecture 13 Spring 2008 S Prg ggisa Sensor Categories Smart Sensors usually integrated with support electronics including a microcontroller and is programmed to accomplish diverse tasks as digital signal processing compensation of static errors selfcalibration and testing Passive Sensors These generate an electrical signal in response to an external stimulus No additional power source is needed Examples Thermocouples and Piezoelectric sensors ECE 510 Lecture 13 Spring 2008 S PrgLsaoyl Slide8 lites 83 Saliterman Sensor Categories Active Sensors Requires external power source or excitation energy to modulate its output signal SensorArray consists of several similarly functioning sensor elements Multisensor Device consists of several sensor sensor elements With different functions Example DNA microarray chip that consists of several thousand strands of oligonucleotide fra ments or probes and thousands of different pro es can be grouped together for detection of fluorescent label RNA fragments in sample ECE 510 Lecture 13 Spring 2008 S PrgLsaoyl Slide9 lites 83 Saliterman Sensor Categories Multifunction sensor is a single device that can perform different functions under different conditions Some sensors have integrated signal conditioning including amplification offset balancing and temperature compensation ECE 510 Lecture 13 Spring 2008 S P3323133 333510 Sensor Specifications The transfer function relates the ideal or theoretical output of the sensor to the measurand and may be linear or non Hnear An ideal sensor output would follow the actual value of the measurand without deviation Error is specified in terms of the input value ECE 510 Lecture 13 Spring 2008 S P3323133 Specifications 0 Transfer function lrs uulvui S s umr In W mi mm mum um ipcci ed mumy A mm H 1 MTme mum me FS quotquotm39 D siimulus i l39 V I 5 i Ira B ECE 510 i mei n m Lecture 13 Spring 2008 3 Prints Sign Static Characteristics 0 Static characteristic The relationship between the sensor output signal and the measurand S a 17s Where S is the electrical signal produced by the sensor a is the intercept or output signal at zero input signal and b is the slope or sensitiva In the case of a simple linear relation the slope of this function is known as sensitivity ECE 510 Lecture 13 Spring 2008 S Prguszg slidez13 s s saineiman Sensor Characteristics 0 Dynamic range Range of measurand values Full scale output 0 Linearity 0 Dynamic characteristics Warmup time Frequency response 0 Other specifications ECE 510 Lecture 13 Spring 2008 S Hajng S it Dynamic Specifications Dynamic Range A sensor responds to some minimal and maximal stimulus range of measurand values Full Scale Output FSO is the maximal output signal over the dynamic range of the stimulus values or the difference in output between the highest and lowest stimulus ECE 510 Lecture 13 Spring 2008 S P3323133 Dynamic Specifications Linearity The closeness of the sensor s calibration curve to a straight line for a linear transfer function expressed as percent of the F80 which is the maximum deviation of any calibration point from the corresponding point on the specified straight line For nonlinear transfer functions this deviation may be referred to a inaccuracy ECE 510 Lecture 13 Spring 2008 S P3323133 333516 Dynamic Specifications Hysteresis The maximum difference in output at any measured value within the specified range when the value is approached first with an increasing and then with a decreasing stimulus Dynamic Characteristics Output following the continuous change to the stimulus without lag time The dynamic characteristic of a sensor is an expression of the time dependency of the response and the dynamic error is the difference between the reported value and the actual value of the stimulus at the moment of sensing ECE 510 Lecture 13 Spring 2008 S P3323133 Hysteresis o Hysteresis Hysteresls lt lam bnlnncc F5 slimlllus ECE 510 Lecture 13 Spring 2008 3 W325 331318 Dynamic Specifications Warmup Time is the time required after application of excitation energy before a sensor can respond to its specifications Frequency Response specifies how fast a sensor can respond to a change in the input stimulus ECE 510 Lecture 13 Spring 2008 S P3323133 333519 m quot n quot 39 i H quot 5 r Tzlzi 1 quot 3 1 7 r wan L 39i 39 v39 39 w lt mu 39 39 39 39 gt1 i 39 s 39 39v r K n v q E q i i f r r L H l i v a a 3 gm mmquot a39r v v 4 I A I r q I I 9 s 2 1A F 1 u 2 V F Egg 5 E V r13i 391 5 quot V i 39 39 39 1 quot L y 1 quotIquot quot h x M quoti n 39 j 39i f quot w 392 3 5 a 12 5 3 39 f i n I 1 3 v f f quot v 1 9 3 x a quot 39 quotI quot 0 39 393 quot f 39 it 39 in 22 Ei RI 1 Hint 39 39 29 39 quot 305 as x 42m I x 39 no 393 a V s x x x 3 M 39 x it x A Other Specifications ture 13 Spring 2008 S Prasad Slide20 Couriesy 88 Saliterman Other Specifications ECE 510 Lecture 13 Spring 2008 3 Prasad Slide21 Couriesy SS Saiiterman Basic Sensors 0 Thermal 0 Radiation 0 Mechanical 0 Flow 0 Magnetic 0 Optical o Piezoelectric and Saw 0 Electrochemical Sensors may be grouped based on the energy domain of their primary ECE 510 Lecture 13 Spring 2008 S P331355 Thermal Sensors 0 Thermal sensors measure temperature heat flow and thermal conductivity BioMEMS devices offer both a means for measuring temperature and a reason to do so LOC devices require controlled temperatures for chemical reactions to take place Such sensors may be incorporated into the device to measure the temperature and compensate according to the requirement of the reaction ECE 510 Lecture 13 Spring 2008 S Prasad slidezzs Thermal Sensors Application of temperature control Laser Diode These devices are used in detection schemes their calibration and use depends on the stability of operation To achieve a stable laser diode output it is necessary to regulate its temperature This is accomplished by mounting the laser diode with a thermistor and thermoelectric module that may heat or cool the laser diode according to a preset level ECE 510 Lecture 13 Spring 2008 S P3323133 333524 Types 0 Platinum resistor Linear stable reproducible Material property dependency on temperature 0 Thermistor Made from semiconductors o Thermocouples o Thermodiode and Thermotransistor ECE 510 Lecture 13 Spring 2008 S P331355 Thermocouple o Potentiometric devices fabricated by the joining of two different metals forming a sensing junction Based on the thermoelectric Seebeck effect in which a temperature difference in a conductor or semiconductor creates an electric voltage memn MetalB A Va5AT A Where A V is the electrical voltage a is the Seebeck coefficient expressed in voltsK and Tm AT is the temperature difference T5 7 ECE 510 Lecture 13 Spring 2008 8 Frags 33326 Thermodiode and Thermotransistor 0 When a pn diode is operated in a constant current 0 circuit the forward voltage Vow is directly proportional to the absolute temperature PTA T V kBT 111 q i i 1 Thermodiode Themiotlnnslstor I T 5 Where A kb is the Bolztnan constanL om T is temperature q is the charge on an electron I is the operating current and 13 is the saturation cmrent ECE 510 Lecture 13 Spring 2008 S Prasad Slide27 Couitesv s s Saiiteiman Radiation Sources Radiation may be charged particulate radiation that includes fast electrons and heavy charged particles and it may be uncharged radiation that includes electromagnetic radiation and neutrons ECE 510 Lecture 13 Spring 2008 S P3323133 333528 Radiation Sources 0 Fast electrons Beta particles 0 Heavy charged particles Alpha particles 0 Electromagnetic radiation Xrays Gamma rays 0 Neutrons Slow and fast ECE 510 Lecture 13 Spring 2008 S P331355 533529 Radioactive Decay o A radioisotope source is defined by its activity or radioactive decay CW 7 7 N If A Mm Where N is the number of radtoacnve nuclei and it is the radioisotope decay constant 111 Ziltalflife o The Curie Ci is defined as disintegrationssecond The SI equivalent is the Becquerel Bq 1Bq2703x10 Ci ECE 510 Lecture 13 Spring 2008 5 Prgfzg Slidezao s s samevmen Specific Activity 0 The specific activity is defined as the activity per unit mass ofthe radioisotope sample in a pure sample u i activity AN A41 bpec1t1c actmty E mass NM i39A r M Where M is the molecular weight of the sample and AV is Avogadro s number 602 x 1023 nucleimole ECE 510 Lecture 13 Spring 2008 S Prcaafnggss 531331 Energy 0 The electron volt eV is the kinetic energy gained by an electron by its acceleration through a potential difference of 1 volt 16V 1 60310491 and 1116241x103ey Wirerc l is the St unit ofenet gy the Joule ant His 10quot Joule o The energy of an X ray or gammaray photon is E in Wire ii is Planck s constant 6626 x10 and l39 is frequency ECE 510 Lecture 13 Spring 2008 5 Pram Slidest 5 ss Salnerman Radiation Detection 0 Pulse mode Individual quantum detection GeigerMuller Tubes 0 Current mode Detector response is slow relative to events Averaging occurs 0 Mean square voltage mode Mixed radiation environments ECE 510 Lecture 13 Spring 2008 S P331355 Detection Principles Pulse mode each individual quantum of radiation that interacts with the detector is recorded The time integral of each burst of current or total charge is recorded This method is not suitable for very high event rates when the time between events is too short to carry out analysis ECE 510 Lecture 13 Spring 2008 S P3323133 333534 Detection Principles Current mode The detector response time is typically long compared with the average time between events so that a time average occurs and is recorded Mean square voltage mode is useful in mixed radiation environments when charges produced by two are more radiations are present Here the derived signal is proportional to the square of the charge per event ECE 510 Lecture 13 Spring 2008 S P3323133 Scintillation Detectors 0 When ionizing radiation strikes certain materials scintillation light is produced Fluorescence is the prompt emission of visible radiation from a substance following its excitation by some means Undesirable emissions include phosphorescence and delayed uorescence ECE 510 Lecture 13 Spring 2008 S P331355 533536 Scintillation Detectors Detection of scintillation li ht output requires use of photomultiplier tube PM or photodiode PMT s are highly sensitive detectors with a single scintillation pulse giving rise to 1071010 electrons The goal IIS to convert a few hundred photons into a usable sngna Inside a photomultiplier tube a photosensitive layer called the photocathode is coupled to the electron multiplier structure The photocathode serves to convert incident light into low energy electrons These photoelectrons are multiplied within the PMT and provide an output signal ECE 510 Lecture 13 Spring 2008 S Prg goyl Slide37 s 83 Saliterman Scintillation Counters mm mm minim Xi39ay V 39 V Fimloracirnn 7 and Q V m K k V Dnl i imtmltwa yummucimu a nah E in timme N i winnings IYHZGHOH regmquot elecumm r v 39 W 5 V2 Scintillation Lighl 7 region phalanx r V Grid V3 mm mm mm mm H nim m mm 391 mu m pimmm innit my ECE 510 Lecture 13 Spring 2008 S Prasad Slide38 Couiiesv s s Saiiiennan Solid State Detection 0 Conventional photodiodes convert photons directly into electronhole charge carriers Srinlilluliun phalanx Depleted i regiun ZOO 500 11m n type layer ECE 510 Lecture 13 Spring 2008 S Prc zg slide39 s s Saiiterma v Solid State Detection o Avalanche diodes have internal gain through higher electric fields and increased number of charge carriere 39 Ligiu Significant eh pair quot generation I Multipiyn n i ECE 510 Lecture 13 Spring 2008 S P331355 533540 Mechanical Sensors 0 Construction 0 Detection Methods 3 Cantilever beam Electrical b Bridge structure Magnetic 0 Diagram or Optical membrane Acoustic on due u a C l ECE 510 Lecture 13 Spring 2008 5 Prgfnggss 353341 Cantilever Beam 0 The displacement X of the beam is related to the applied force and length of the beam AXLFX 01 E IfmAX km is the spring constant win Where Equot is Young39s modulus Iquot is the second moment of inertia F is the force or point load and 1 is the length ECE 510 Lecture 13 Spring 2008 S Prgusad may 55 Salmman Bridge Structure 0 The sinusoidal solution for displacement X of a bridge structure is F 1T2 AX Asin E y nd FCIMMI the buckling force Where A is a constant Equot is Young39s modulus M is the second moment of inertia 1 is the force and is the length All materials exhibit physical properties that are similar in both the micro and macroscales This is true for other material so long as the average grain size of the material is much smaller than the smallest feature size of the fabricated structure ECE 510 SPrasad Slide43 Lecture 13 Spring 2008 Flow Sensor Methods Monitoring the measurement of flow in microfluidic devices 0 Mechanical turbines cantilever 0 Thermal o Electromagnetic probes o Ultrasonic Doppler 0 Piezoelectric ECE 510 Lecture 13 Spring 2008 S P331355 533544 Thermal Flow Sensor 0 The heat transferred per unit time form a resistive wire heater to a moving liquid is monitored with a thermocouple TC Heater TC Flow quotIi ihi l r i Hii ECE 510 Lecture 13 Spring 2008 S Pr gsg k igjms Flow Rates 0 In a steady state the mass flow rate can be determlned Q 5 m if Cm Where 0 is the mass ow rate 02 7 r1 Ph is the heat transfeued per unit time cm is me speci c heat capacity of the uid and 112 are empemnue o The volumetric flow rate is calculated as follows Where QV is the volmneuic ow rate QM is the mass ow rate and pm is the density 55 Salmman ECE 510 Lecture 13 Spring 2008 S Prim slidez46 Magnetic 0 Galvanic Hall devices 0 Conductometric magnetoresistive o Voltaic magnetodiode and magnetotransistor 0 Acoustic SAW devices SQUle ECE 510 Lecture 13 Spring 2008 S Prg gss 533547 Hall Effect Sensor 0 The theoretical Hal voltage is expressed as follows Magumv mm 7 7 LB 7 LB Immu IH we 6 RH d H Where 77 7 H K n is the carrier density eistlte eleczrouic charge 1 165x10quotC C IV is the mutant B 15 the llngHCTlC flu r 15th slab thickness and RH is the Hall Coef cient small in metals 1gtlt104 uniC and highei in semiconducmrs 39 n a s ab of material of length I thickness d When a current X Is passed 0 and a perpendicular magnetic flux density B1 is applied a voltage VH B l r 1 appears across the slab perpendicular to IX and 1 v ring 2008 S Prasad Slide48 WW 3 slaw ECE 510 Lecture 13 Sp Optical Fibers 0 An optical ber consists of a solid cylindrical core of transparent material surrounded by a cladding of similar material but of lower refractive index than the core Confinement Angie Fiber Optic Wave Guide Cladding Ligm Hay Core lntemal Reliectlon ECE 510 Lecture 13 Spring 2008 S Proausad S e new gt s Salnennaii Refractive Index and Snell s Law 0 The refractive index is the ratio of the speed of light in a vacuum to the speed of light in the medium Cmcuum cznareria I o Snell s law defines the relationship between incident and refracted light measured as an angle from a perpendicular to the surface 1 sin I 171 sin R ECE 510 Lecture 13 Spring 2008 S P331355 533550 Critical Angle w Refraction can not take place when the angle of incidence is too large or greater than the critical angle For air refractive index of 1 and glass refractive index of 15 the critical angle is defined as Q arcsin 111 nj 418 ECE 510 Lecture 13 Spring 2008 S P331355 Acceptance Angle o The acceptance angle is the angle over which light rays entering the fiber will be guided along its core Light Ray Outside of Acceptance Angle Fiber Optic Wave Guide Acceptance Angle Internally Reflected Light Flay EcE 510 Lecture 13 Spring 2008 3 Prg g ss Salmman Numeric Aperture 0 Numeric Aperture NA is the light gathering ability of a fiber The material NA relates to the refractive indices of the core and cladding NA n12gt sin6 nosinec Where 120 is the core index 111 is the cladding index and 639 is half the acceptance angle and Sr is the confinement angle ECE 510 Lecture 13 Spring 2008 S P331355 Modes 0 Light propagates through the core in a series of wave fronts or modes mauve a tn ul Om m HtghOmsy D s emon tndex M We E W J i 7 Mummo n Step Yndax 7me ME StnaterMode Slaw ndax Mutlimnde Gla e masx ECE 510 Lecture 13 Spring 2008 s Prgs ad slide54 cu m as Sah39ennan Optical Chemical Sensors 0 Optical chemical sensors are usually con gured as transducers With transductions steps of electrical opticalchemicalopticalelectrical conversion r almnml pawtr Electrical Itsml c Tmlmincar ionucalelcuriczl hrugmn39on m mi G Mombrm Mulligan mm mummy mm 0pm mama mm quotSurinan Also known as Optode Chemical toaolmamewm optical transduction for chemical ECE 510 detection Lecture13prlng 2008 SP ggb 331355 s Summary 0 A sensor is a transducer that converts the measurand a quantity or a parameter into a signal that carries information The transfer function relates the ideal or theoretical output of a sensor to the measurand and may be linear or nonlinear 0 An ideal sensor output would follow the actual value of the measurand without deviation In practice a sensor will deviate from the transfer function Sensors reviewed today included thermal radiation mechanical flow magnetic and optical ECE 510 Lecture 13 Spring 2008 S P331355 533556 1 113113131 393 STATE LJINIKillilib lT r Biomicroelectrical mechanical systems BioMEMS ancl Nanobiotechnology Dr Shalini Prasad Electrical and Computer Engineering Biomedical Microdevices and Nanotechnology Laboratory sprasadpdxedu httpwwwece pdxed uprasagls ECE 510 Prasad Slide1 C ooooo sy SS Sali ttttt n Course Outline 1 Introduction to BioMEMS and Nanobiotechnology 2 Silicon Microfabrication 3 Soft Fabrication Techniques 4Poymer Materials and Microfluidics 5 Sensor Principles 6Detection and Measurement Methods 7 Drug delivery systems 8 MicroTotal Analysis Systems uTAS I Emerging Applications 9 Nanotechnology Applications E 510 EC S Prasad Slide2 Courtesy SS Saliterman Sensors o A sensor converts one form of energy to another and in so doing detects and conveys information about some physical chemical or biological phenomena 0 More specifically a sensor is a transducer that converts the measurand a quantity or a parameter into a signal that carries information ECE 510 S Prasadr VSlide3 Ideal Sensor 0 Continuous operation without effecting the measurand 0 Appropriate sensitivity and selectivity 0 Fast and predictable response 0 Reversible behavior 0 High signal to noise ratio 0 Compact 0 Immunity to environment 0 Easy to calibrate ECE 510 S Prasad Slide4 Caunesv s s Sslilemisn Issues with Sensors Sensors are not ideal when designed as a standalone device or as part of uTAS detection scheme they may affect the measurand Example Biomolecule samples guided through micro channels must be separated immobilized for measurement purposes as a part ofa continuous flow process Mechanical alteration of samples may occur ECE 510 S Prasad Slide5 C ooooo 5y SS Sali ttttt n Categories 0 Biosensor 0 Smart sensor 0 Passive or active 0 Sensor array 0 Multisensor o Multifunction ECE 510 S Prasad Slide6 an m s s Saiileim Sensor Categories Biosensors measure or indicates living things and their products Analyzers measure the constituents of a substance and may have one or more biosensors Control Systems regulate the value of the measurand to some specific value typically as a part of the feedback loop ECE 510 S Prasad Slide7 C ooooo 5y SS Sali ttttt n Sensor Categories Smart Sensors usually integrated with support electronics including a microcontroller and is programmed to accomplish diverse tasks as digital signal processing compensation of static errors selfcalibration and testing Passive Sensors These generate an electrical signal in response to an external stimulus No additional power source is needed Examples Thermocouples and Piezoelectric sensors ECE 510 S Prasad Slide8 Courtesy SS Saliterman Sensor Categories Active Sensors Requires external power source or excitation energy to modulate its output signal SensorArray consists of several similarly functioning sensor elements Multisensor Device consists of several sensor sensor elements With different functions Example DNA microarray chip that consists of several thousand strands of oligonucleotide fragments orr probes and thousands of different probes can be grouped together for detection of fluorscent label RNA fragments in sample ECE 510 S Prasad Slide9 Courtesy SS Saliterman Sensor Categories Multifunction sensor is a single device that can perform different functions under different conditions Some sensors have integrated signal conditioning including amplification offset balancing and temperature compensation ECE 510 S Prasad Slide10 C ooooo sy SS Sali ttttt n Sensor Specifications The transfer function relates the ideal or theoretical output of the sensor to the measurand and may be linear or non Hnear An ideal sensor output would follow the actual value of the measurand without deviation Erroris specified in terms of the input value ECE 510 S Prasad Slide11 C ooooo sy SS Sali ttttt n Specifications 0 Transfer function tr uuwui S 5 mm n W n 7 mi mm FilnL lInn mu m speci ed muncy A mm JH39 1 ml mm Fmon tum siimulus F3 sllmulus D d quotJ EP i 39 I mien 1607 ECE 510 Static Characteristics 0 Static characteristic The relationship between the sensor output signal and the measurand S a 115 Where S is the electrical signal produced by the sensor a is the intercept or output signal at zero input signal and b is the slope orsemifiviry In the case of a simple linear relation the slope of this function is known as sensitivity ECE 510 S Prasad Slide13 cuunesv as Saiitemian Sensor Characteristics 0 Dynamic range Range of measurand values Full scale output 0 Linearity 0 Dynamic characteristics Warmup time Frequency response 0 Other specifications ECE 510 s Prasad A Slide14 Dynamic Specifications Dynamic Range A sensor responds to some minimal and maximal stimulus range of measurand values Full Scale Output FSO is the maximal output signal over the dynamic range of the stimulus values or the difference in output between the highest and lowest stimulus ECE 510 S Prasad Slide15 C ooooo sy SS Sali ttttt n Dynamic Specifications Linearity The closeness of the sensors calibration curve to a straight line for a linear transfer function expressed as percent of the F80 which is the maximum deviation of any calibration point from the corresponding point on the specified straight line For nonlinear transfer functions this deviation may be referred to a inaccuracy ECE 510 S Prasad Slide16 C ooooo sy SS Sali ttttt n Dynamic Specifications Hysteresis The maximum difference in output at any measured value within the specified range when the value is approached first with an increasing and then with a decreasing stimulus Dynamic Characteristics Output following the continuous change to the stimulus without lag time The dynamic characteristic of a sensor is an expression of the time dependency of the response and the dynamic error is the difference between the reported value and the actual value of the stimulus at the moment of sensing ECE 510 S Prasad Slide17 C ooooo sy SS Sali ttttt n Hysteresis o Hysteresis V a E E I m gt I Y zcmbnlnncc F5 slimlllus ECE 510 S Prasadr VSlide18 Dynamic Specifications Warmup Time is the time required after application of excitation energy before a sensor can respond to its specifications Frequency Response specifies how fast a sensor can respond to a change in the input stimulus ECE 510 S Prasad Slide19 C ooooo sy SS Sali ttttt n 3 3 6E ECE 510 ons Cnher pecmca values Excitation energy is an active transducer Interchangeability is the maximum possible error of measurement when individual sensor device is changed Within the measurement appli cation for anolher of the same type the electrical signal voltage or current needed for to operate uulmdt opetational conditions Limit of deterrion is the lowesa value of the measurand thnt can be reel is the ability of n sensor to reproduce output readi easurand value is applied to it consecu tively under the same conditions and in the same direction Resolution it the smallest increment in the output given as FSO e Repeatability nr raprmlllcibt lr tjv ngs when the same m a percent of Respome time is the amount of time required tor the output to rise to a speci ed percentage generally 90 of its nal value as a result of a Step change in the measttrand S Prasad Slide o 9 mm 55 Salhmm 10 11 12 13 14 ECE 510 Other Specifications Jaipur impedance is the eleeLrie impedance that must be matched bemeen the sensor and its electronic circuit Saturation is the point at which ihere ia 110 further output signal in spite Jf a rising stimulus Selectivity is the suppressimi of environmental interference e g temp eraiure Sensitivity shift is the change in slope of the calibration cane Zem meammnd aarpur is the natput to he at zero when an measuranri is applied Zena shi Jr di l is the change in the mere measured nutput under speci c eenditiens such as temperature change lungterm statage aging ere S Prasad Slide21 Courtesy SS Saliterman Basic Sensors 0 Thermal 0 Radiation 0 Mechanical 0 Flow 0 Magnetic 0 Optical o Piezoelectric and Saw 0 Electrochemical Sensors may be grouped based on the energy domain of their primary inpu ECE 510 S Prasad Slide22 Mews Salim Thermal Sensors 9 Thermal sensors measure temperature heat flow and thermal conductivity BioMEMS devices offer both a means for measuring temperature and a reason to do so LOC devices require controlled temperatures for chemical reactions to take place Such sensors may be incorporated into the device to measure the temperature and compensate according to the requirement of the reaction 8 Prasad Slide23 Cmmesv s s Salilevman ECE 510 Thermal Sensors Application of temperature control Laser Diode These devices are used in detection schemes their calibration and use depends on the stability of operation To achieve a stable laser diode output it is necessary to regulate its temperature This is accomplished by mounting the laser diode with a thermistor and thermoelectric module that may heat or cool the laser diode according to a preset level ECE 510 S Prasad Slide24 C ooooo sy SS Sali ttttt n Types 0 Platinum resistor Linear stable reproducible Material property dependency on temperature 0 Thermistor Made from semiconductors o Thermocouples o Thermodiode and Thermotransistor ECE 510 S Prasad Slide25 Caunesv s s Sslilemisn Thermocouple o Potentiometric devices fabricated by the joining of two different metals forming a sensing junction Based on the thermoelectric Seebeck effect in which a temperature difference in a conductor or semiconductor creates an electric voltage memm MetalB A Va5AT AV When A V is the electrical voltage a is the Seebeck coefficient expressed in voltsK and Tm AT is the temperature difference T5 7 s Salitemian ECE 510 S Prasad Slide26 came 5 Thermodiode and Thermotransistor 0 When a pn diode is operated in a constant current 0 circuit the forward voltage Vow is directly proportional to the absolute temperature PTA T k T 773 Tt d d Th 1 W Van q 111 lumn m e smtoldmhor I t EH Where lrb is the Bolzman constanh T is temperature q is the charge on an electron I is the operating current and h 13 15 the saturatron current V ECE 510 s Prasad Slide27 CUUHEW 95 Sahtevmn Radiation Sources Radiation may be charged particulate radiation that includes fast electrons and heavy charged particles and it may be uncharged radiation that includes electromagnetic radiation and neutrons ECE 510 S Prasad Slide28 C ooooo sy SS Sali ttttt n Radiation Sources 0 Fast electrons Beta particles 0 Heavy charged particles Alpha particles 0 Electromagnetic radiation Xrays Gamma rays 0 Neutrons Slow and fast ECE 510 S Prasad Slide29 Caunesv s s Sslilemisn Radioactive Decay o A radioisotope source is defined by its activity or radioactive decay W 7 eA39V If I my Where N IS the number of radioactive nuclei and it is the radioisotope decay cousiaut in Zilialflife o The Curie Ci is defined as disintegrationssecond The SI equivalent is the Becquerel Bq qu2703gtlt1039 Ci ECE 510 s Pras ad Slide30 Wu 99 as samemn Specific Activity 0 The specific activity is defined as the activity per unit mass ofthe radioisotope sample in a pure sample activity AN AAV mass NM AV M Speci c atmin E Where M is the molecular weright of the sample and AV is Avogadro s1111111ber602 x 10quot3 nucleimole ECE 510 S Pmsad Slide31 2mm 5 s aiiieiman Energy 0 The electron volt eV is the kinetic energy gained by an electron by its acceleration through a potential difference of 1 volt leVl 60240191 and 1136241x103ev Where i is the SI unit ofenergy the Joule and His 10quot Joule o The energy of an X ray or gammaray photon is E ln39 Where ll is Planck s constant 6626 X1034 and 2 is frequency ECE 510 s Pras ad Slide32 Cuu esy 55 Salnerman Radiation Detection 0 Pulse mode Individual quantum detection GeigerMuller Tubes 0 Current mode Detector response is slow relative to events Averaging occurs 0 Mean square voltage mode Mixed radiation environments ECE 510 S Prasad Slide33 Caunesv s s Sslilemisn Detection Principles Pulse mode each individual quantum of radiation that interacts with the detector is recorded The time integral of each burst of current or total charge is recorded This method is not suitable for very high event rates when the time between events is too short to carry out analysis ECE 510 S Prasad Slide34 C ooooo 5y SS Sali ttttt n Detection Principles Current mode The detector response time is typically long compared with the average time between events so that a time average occurs and is recorded Mean square voltage mode is useful in mixed radiation environments when charges produced by two are more radiations are present Here the derived signal is proportional to the square of the charge per event ECE 510 S Prasad Slide35 C ooooo sy SS Sali ttttt n Scintillation Detectors 0 When ionizing radiation strikes certain materials scintillation light is produced Fluorescence is the prompt emission of visible radiation from a substance following its excitation by some means Undesirable emissions include phosphorescence and delayed uorescence ECE 510 S Prasad Slide36 Caunesv s s Sslilermsn Scintillation Detectors Detection of scintillation light output requires use of photomultiplier tube PMT or photodiode PMT s are highly sensitive detectors with a single scintillation pulse giving rise to 1071010 electrons The goal is to convert a few hundred photons into a usable sngna Inside a photomultiplier tube a photosensitive layer called the photocathode is coupled to the electron multiplier structure The photocathode serves to convert incident light into low energy electrons These photoelectrons are multiplied within the PMT and provide an output signal ECE 510 S Prasad Slide37 Courtesy SS Saliterman Scintillation Counters mm mm Semtlanmnt 1m mm W mm W in W m i WW mm vw mm mutthiw H Wm m minimalmm 391 mm m pimmm ECE 510 Fhmowaumn Imiizalian elemmm Ligzl pholmu X ray vGridV Xegas lt Drift reginn Um v2 Sziptillntion mgmn gnaw PM lube UV cmilivc S Prasad Slide3938 Cuuneay ss Sahtevman Solid State Detection 0 Conventional photodiodes convert photons directly into electronhole charge carriers Srinlilluliun phUIUHS Depleted i Iegiun 200 500 um n type layer ECE 510 S Prasad Slide39 Solid State Detection o Avalanche diodes have internal gain through higher electric fields and increased number of charge carriers Q eh pair 39 generation Slgnificant gain 1 ECE 510 s Prasad Slide40 Mechanical Sensors 0 Construction 0 Detection Methods a Cantilever beam Electrical b Bridge structure Magnetic 0 Diagram or Optical membrane Acoustic rzmdnm 2 3 20m ECE 510 S Prasad Slide41 mm s s Salilevman Cantilever Beam 0 The displacement X of the beam is related to the applied force and length of the beam 13 7 AximFx 01 Erik qu km is the spring constant Where Equot is Young39s modulus Iquot is the second moment of inettia F is the force or point load and is the length cal 39 04 l ECE 510 S Prasad Slide42 cm esv s s Sslilemisn Bridge Structure 0 The sinusoidal solution for displacement X of a bridge structure is K2 AX Asin 112 the buckling force F EMEHA and F Critical Where A is a constant Equot is Young39s modulus Iquot is the second moment of inertia 1 is the force and I is the length All materials exhibit physical properties that are similar in both the micro acroscales Th39 is true for other material so long as the average grain size of the material is much smaller than the smallest feature size of the fabricated structure S Prasad Slide43 cm esv s s Sslilemisn Flow Sensor Methods Monitoring the measurement of flow in microfluidic devices 0 Mechanical turbines cantilever 0 Thermal o Electromagnetic probes o Ultrasonic Doppler 0 Piezoelectric ECE 510 S Prasad Slide44 Caunesv s s Sslllemlsn Thermal Flow Sensor 0 The heat transferred per unit time form a resistive wire heater to a moving liquid is monitored with a thermocouple TC Heater TC C C quotIiiih lsi Ii39iyi ECE 510 S Prasad Slide45 s Flow Rates 0 In a steady state the mass flow rate can be determlned g l m Where Q is the mass ow rate a is the heat transfen39ed pel unit time cm is me speci c heat capacity ofthe uid and III are temperature 0 The volumetric flow rate is calculated as follows 0 7 dV 7 Q V i i i Where Q is the volmnem39c ow rate Q is the mass ow rate and pm is the density ECE 510 S Prasad Slide46 esv s e Sslllemlsn Magnetic 0 Galvanic Hall devices 0 Conductometric magnetoresistive o Voltaic magnetodiode and magnetotransistor 0 Acoustic SAW devices SQUle ECE 510 S Prasad Slide47 Caunesv s a Sslllemlsn Hall Effect Sensor 0 The theoretical Hall voltage is expressed as follows Miglmlu mu illB A R 1 3 quot 77 112 d H d a Where 7 is the tamer density e is the electronic chmge l 165 x10 C I is he cuu39em B is the magnetic flux 039 ls the slab tlnckness and RH is the Hall Coef cient small in metals 1 x 104 cuf 391 and liigliu in semicmulucms When a currenth is passed down a slab of material of length I thickness d and a perpendicular magnetic flux density BI is applied a voltage VH appears across the slab perpendicular to and B1 7 r ECE 510 S 5351 Optical Fibers 0 An optical ber consists of a solid cylindrical core of transparent material surrounded b a cladding of similar material but of lower refractive index than the core Con nemem Angle Flber Optic Wave Gulde Cladding Ligm Ray Chm lntemal Reflecticn em w n m ECE 510 S Prasad Slide49 Refractive Index and Snell s Law 0 The refractive index is the ratio of the speed of light in a vacuum to the speed of light in the medium C vacuum 11 2 1 CIImreria o Snell s law defines the relationship between incident and refracted light measured as an angle from a perpendicular to the surface 171 sin I 171 sin R ECE 510 S Prasad Slide5o Caunesv s s Sslilevmsn Critical Angle c Refraction can not take place when the angle of incidence is too large or greater than the critical angle For air refractive index of 1 and glass refractive index of 15 the critical angle is defined as 9C 2 arcsinn1 nj 2 418 ECE 510 S Prasad Slide51 Caunesv s 3 Sslilemisn Acceptance Angle o The acceptance angle is the angle over which light rays entering the fiber will be guided along its core Light Ray Outside of Acceptance Angle Fiber Optic Wave Guide Acceptance Angle Internally Reflected Light Flay ECE 510 Numeric Aperture 0 Numeric Aperture NA is the light gathering ability of a fiber The material NA relates to the refractive indices of the core and cladding NA n12gt sin6 nosin c Where no is the core index 111 is the cladding index and L9 is half the acceptance angle and t is the confinement angle ECE 510 S Prasad Slide53 Caunesv s s Settlemer Modes 0 Light propagates through the core in a series of wave fronts or modes Rm we tndex P on mam Output HtghOmsv D s won Puke Puts Made y a a 3142 in Mummu n Step Yndax 7me ME StncterMode smp ndax Muttimndv Glade masx Srmg 200E ECE 510 S Prasad Slide54 c nurtes as Sahtemxan Optical Chemical Sensors 0 Optical chemical sensors are usually con gured as transducers With transductions steps of electrical opticalchemicalopticalelectrical conversion r Electrical pam EIccm39ml mi 3 Transducer onucalelcctriczl hrugm on m miquot 3 Membrm lnlellrgenl cm mummy 1mm 0pm mama Mm quotSmzn39unxur Also known as Optode Chemical toaolsaemHarmer1997 tical transduction for chemical op I ECE 510 detection S Prcaaiag Slide55 s Salilemian S u m mary o A sensor is a transducer that converts the measurand a quantity or a parameter into a signal that carries information The transfer function relates the ideal or theoretical output of a sensor to the measurand and may be linear or nonlinear 0 An ideal sensor output would follow the actual value of the measurand without deviation In practice a sensor will deviate from the transfer function Sensors reviewed today included thermal radiation mechanical flow magnetic and optical ECE 510 S Prasad Slide56 Caunesv s s Sslilemisn 1339 quotHana w 1 39w 5TH LJn39i1mTamp Biomicroelectrical mechanical systems BioMEMS and Nanobiotechnoloclv Dr Shalini Prasad Electrical and Computer Engineering Biomedical Microdevices and Nanotechnology Laboratory sprasadpdxedu httpwwwecedeeduprasads Lect re12 S Prasad Slide1 Courtesy39 SS Salilerman ECE 510 Course Outline 1 Introduction to BioMEMS and Nanobiotechnology 2 Silicon Microfabrication 3 Soft Fabrication Techniques 4Poymer Materials and Microfluidics 5 Sensor Principles 6 Detection and Measurement Methods 7 Drug delivery systems 8 MicroTotal Analysis Systems uTAS Emerging Applications 9 Nanotechnology Applications ECE 510 Lecture12 s Prasad Slide2 Courtesy39 SS Saliterman ECE 510 What is Microfluidics S Prasad Caumesy SE Saliterman Lecture12 Slide3 ICGS 4 Slide 88 Saliterman S Prasad Lecture12 Courtesy wwwuvmkmm ws 222222 Mvavv wwwvvvvvv 9 u 33 ug muhu icroflui ics Dev M ECE 510 CommerCIal Aspects The micro uidics markets for life science applications should reach M750 in 2004 and is expected to reach B2 in 2010 Today challenges of microfuidics suppliers are to reach the specifications of the life science companies in the in Vitro diagiostic drug discovery drug delivery marketsi Miuulluiuiv Technologies 8 Millions EOEsm 1997 my my mum 2mm mm mm zuua sPrasad Slums Simvman enmesy 55 Labona Chip Device Configuration ECE 510 S Prasad Slide6 Courtesy SS Saliterman Lecture12 ECE 510 Lecture12 S Prasad Courtesy SS Saliterman Slide 7 3 1 1 1 1 1 3 w xknzntntntntw Microfluidic Components Fabrication of uFluidic Devices 0 The science of micro uidics includes specialized fabrication techniques physical properties of uids and materials on the micro scale and electrokinetic effects Micro uidic devices serve to manipulate a uid and its constituents sample ECE 510 Lecture12 s Prasad Slide8 Courtesy SS Sallterman Flow in Microfluidic Devices ECE 510 Lecture12 S Prasad Slide9 Courtesy 88 Saliterman Transport Process Considerations ECE 510 Lecture12 S Prasad Slide10 Courtesy 88 Saliterman Transport Processes 0 Small scale of micro uidics results in changes to uid behavior Since typically only area where uid is processed needs to be miniaturized 0 For long term processes such as passively or actively obtaining samples the entire device needs to be miniaturized ECE 510 Lecture12 s Prcasad lind 11 ounesy SS Salter Classifications of Transport Processes Three categories based on the characteristic critical dimension of the system LC 39 Macroscale Systems LC gt 200 microns Microscale Systems 100 nm lt LClt 200 microns Nanoscale Systems LClt 100 nm ECE 510 Lecture12 S Prcaised gs Silidenz 0quot 85y al erman Approaches in Analyzing Fluid Flow 0 Eulerian Method Uses field concepts and considers properties of pressure density and velocity as functions of time and space 39 It is possible to understand ow in terms of What happens at the fixed points in space as uid passes those points 0 Lagrangian Method Individual particles are followed as they move over time ECE 510 Lecture12 S Prcasad gs S tlide13 0quot 85y al erman Factors Influencing Fluid Flow 0 F law may be in uenced by 0 Kinematic Properties velocity acceleration 0 Transport Properties viscosity thermal conductivity diffusivity 0 Thermodynamic Properties surface tension vapor pressure surface accommodation coefficients ECE 510 Lecture12 s Prcasad lino n4 ounesy SS Salter Courtesy 88 Saliterman Size Comparison of Devices Lecture12 ECE 510 Courtesy 88 Saliterman ECE 510 Lecture12 Relevant Detection Concentrations Sample Volume 39 Effect of reduced volume on the number of analyte targets available for study 0 Smaller volume implies less analyte is present it is necessary to determine how much sample is required to adequately perform an analysis given the level of the detector ability 0 If the sample volume is too small it may not contain the agent of interest ECE 510 Lecture12 s Prcasad slig17 ounesy SS Salter zii1lz v v zgt 4553555555231 quot39 iiiii 739 ECE 510 Determination of Sample Vdume Lecture1 2 3 Prasad Slide18 Courtesy 88 Saliterman Sample Requirements for Detection ECE 510 Lecture12 S Prasad Slide19 Courtesy 88 Saliterman What is a Fluid 0 A fluid is a substance that deforms continuously under the application of shear tangential stress of any magnitude 0 This includes gases and liquids o Newtonian Force Force L L V I I 39 I I I I I l 1 II I I II I I I I ECE 510 Lecture12 5 Prasfid S lideizo Louvesv S Saieiman Newtonian Fluid 39 A uid is said to be Newtonian if the shear stress shear force area uid contact is directly proportional to the rate of strain dudy 39 We assume uids and gases to be Newtonian 39 The Viscosity of a uid is an additional property that must be considered ECE 510 Lecture12 S Prasad Slide21 C 39 man ounesy SS Salter Viscosity 39 If the uid is placed between two parallel plates and a shearing stress is applied to the upper plate the uid deforms continuously 39 However uid in contact with the upper plate moves with the plate velocity While the uid in contact with the lower plate has zero velocity at nOSlip condition ECE 510 Lecture12 s Prcasad suggzz ounesy SS Salter Viscosity Determination 32m y 1155 1 a 39 quotM 39 3 y y 39 vvvvvvvvv 39 ECE 510 Lecture12 s Prasad Slide23 Courtesy SS Saliterman Nature of Flow 39 Flow can be considered as being laminar transitional or turbulent and is dependent on the uid density and Viscosity characteristic velocity geometry of the channel and Whether or not the ow is past an object 0 For most simple micro uidic channels in which the Width and height are less than 1 mm and because the velocities are less than a cmsec the ow will be laminar ECE 510 Lecture12 S Prcasad lide24 ounesy SS Salter Relevant Factors Affecting Flow ECE 510 Lecture12 s Prasad Slide25 Courtesy 88 Saliterman Courtesy SS Saliterman Slide ECE 510 Lecture12 S Prasad 26 MAvmmmmmmmmmx A mu mmmmu ww mmmmmmm f a x a Flow Dependency Reynolds Number 39 Flow patterns are functions of Reynolds Number 39 It is a measure of the ratio between inertial and Viscous forces in a particular ow 0 This is a dimensionless quantity ECE 510 Lecture12 S Prcasad lide27 ounesy SS Salter ECE 510 Lecture12 s Prasad Slide28 Courtesy 88 Saliterman Reynolds Number Example 0 In a channel carrying water viscosity of 103 kgs m density of 103 kgm3 in a channel with diameter of 10 pm at a velocity of 1 mms the Reynolds number is 0 Contrast this to a channel with diameter of 100 um and fluid velocity of 10 ms where the Reynolds number is ECE 510 Lecture12 s Prfszgss slums Surface Area to Volume ECE 510 Lecture12 s Prasad Slide30 Courtesy 88 Saliterman Surface Area to Volume 0 As the SAV increases processes such as capillary electrophoresis becomes more efficient due to easier removal of heat and transport due to electrokinetic ow decreases because of rapid diffusion of macromolecules ECE 510 Lecture12 s Prcasad slig31 ounesy SS Salter Diffusion m ix A 15 w 39 ti mg ECE 510 Lecture12 s Prasad Slide32 Courtesy SS Saliterman Diffusion Example Hemoglobin 0 Since distance varies to the square power diffusion becomes very important on the microscale o For example Hgb D1O7cmzs1 takes 106 sec to diffuse 1 cm in water but only 1 sec to diffuse 10 pm 0 Therefore in a 1 cm wide tube diffusion of hemoglobin is not usually an important consideration but in a microchannel 10 um wide the distance traveled due to diffusion becomes important ECE 510 Lecture12 s Prasnad V stigma Design Considerations 0 The optimal size domain for microfluidic channel cross sections is somewhere between 10 pm and 100 pm 0 At smaller dimensions detection is too difficult and at greater dimensions unaided mixing is too slow 0 Therefore the typical cross section diameter will be 2 X 103 mmZ The flow range will be 1 to 20 nLsec c When diluting an assay component the two flows must be controlled within 1 or pLsec range ECE 510 Lecture12 8 Frags S S 321334 Microfluidics Flow Summary 0 Microfluidics o Microfluidic devices components and LOC 0 Silicon and polymer materials and fabrication 0 Volumes required for detection of analytes o Fluid dynamics based on the continuum assumption Newtonian fluids Reynolds number and laminar flow Fluid kinematics including field representation of flow and the velocity field Eowiseuille flow including streamlines and Poiseuille s 0 Surface area to volume 0 Diffusion ECE 510 Lecture12 s PraSSd W slidess Electrokinetic Phenomena 39 Harnessing electrokinetic phenomena in micro uidic devices for moving uid and 39 particles including proteins cells bacteria and Viruses is essential for MTAS and other LOC applications ECE 510 Lecture12 S Prcasad gs S tlide36 0quot 85y al erman Electrokinetic Phenomena o Electraosmosis Fluid movement relative to a stationary charged or conducting surface through application of an electric field 0 Electrophoresis e pre ence of an electric eld the particle can be induced to move relative to the stationary or movrng liquid a Streaming potential Occurs when an aqueous ion containing solution is forced to flow through a capillary or microchannel under an applied hydrostatic pressure in the absence of an applied electric eld An eectroviscous effect occurs or resistant to flow 0 Dielectrophoresis Movement of dielectric particles in a spatially nonuniform electric eld 3 3 5m 51o Lemme s Prasad Slide37 my 55 Formation of Electrical Double Layer 0 Application of electric field induces the formation of EDL The EDL consists of the charged surface and the compact liquid layer containing the immobile balancing charges and a a i ase liquid layer of mobile ions 39 From the compact layer to the electrically neutral bulk liquid the net charge density gradually reduces to zero ECE 510 Lecture12 S Prcaised gs S tlide38 0quot 85y al erman Formation of Electrical Double Layer 39 The compact layer is several angstroms thick 0 The shear plane is the boundary between the compact layer and the diffuse layer Where the liquid velocity is zero ECE 510 Lecture12 S Prcaised gs S tlide39 0quot 85y al erman Electric Double Layer EDL ear Plane 0 c0i0ns countereions Solid Surface Diffuse layer gt Compact layer mobile ions immobile ions ECE 510 Lecture12 s Prasaq Slide40 Electric Double Layer Potential PS shear plane 4 Potential measurable WK 0 f 3 5 k Compact Layer Diffuse Layer ECE 510 Lecture12 3 P Z gse 5333 ECE 510 OCE BDDs and Their Applications Lecture 11 FSM Equivalence Checking and FSM State Minimization May 2 2000 Alan Mishcheriko Overview 0 Equivalenoe for FSMs and FSM states 0 Product machine PM 0 Solving the problem of FSM equivalenoe e Derive transition and output relation of the PM 7 Perform reachability on the PM and derive the state equivalence relation 7 Transform the initial FSM s transition and output relations 0 Compatible projection operator MAyZ mun ECE 5m OCE EDDs andThztrAfphcaonns 2 FSM Equivalence 0 De nition Two state machines are equivalent if starting from their reset states for any sequence of input vectors they produce identical sequences of output vectors MAyZ mun ECE 5m OCE EDDs andThztrAfphcaonns z FSM State Equivalence 0 De nition Two states s1 and s2 ofan FSM are equivalent if for any sequence of input vectors the FSM starting from state s1 produces the same sequence of output vectors as the FSM starting from state s2 MAyZ mun m5 5m OCE EDDs mmmmm m Product Machine PM 0 Given FSM 1 O S B A with k inputs n slats m output the product machine is 1 01 5x5 52 72 the product machine has k inputs 2n states 1 output Deriving Representation of PM 0 Given transition relations and output functions of component machines T1 is1n1 2L is1 and T2 is2n2 l1k is2 those of the PM can be computed as follows TDM i151 T1i51n1 ampT2 i1521n2 14M LS Hk 7M L51 7 L52 where is s1 and s2 are the sets of the current state variables for the component machines and sis the union of these sets similar for n ma 5m OCE EDDs and Tim Amlwaonns MAyZ mun FSM Equivalence Checking 0 Find the transition relations and output functions of M1 and M2 Find the transition relation and output function oft e PM 0 Perform reachability for the PM while checking its output 0 If the output of the product machine is 1 for all reachable states M1 and M2 are equivalent othenNise generate an error trace It is possible to define equivalence relative to any subsetof inputs and outputs of the FSM MAyZ mun m5 5m OCE EDDs mmmmm m Equivalence Checking Formulas 0 Property expresses equivalence of M1 and M2 in states s1 and s2 which constitute state s of PM Ps V llamas o Machines M1 and M2 are equivalent iff VSARS gt Ps 1 where ARs is the set of reachable states of the PM and Ps is the property that expresses equivalence of M1 and M2 in the product state s 0 Alternatively M1 and M2 are notequivalent iff 35ARSamp PS39 0 W312 mun m5 5m OCE EDDs mmmmm m 2 Reachability Analysis Procedure bool Veri properQUsingReachabilimnahsm FSMquot pM bdd Property bdd IniSbate FinddeCube0ergtNBiECSVars0 bdd R c ed Inismm From IniSlate NewMAXITERNUM int NIter 0 do bdd To bddappexergtTratheIhombddopjhdAIICSVars C d bdd Check New Nlmr gtgt Property if Check bddlrue return false From New NIter 1 Reached Reached New Nlmr 1 while New Nlmr bddfalse remrh true Mayz2nnn EcEsmocE BDDsdehurAg mmns 9 FSM State Minimization 0 Find the transition relations and the out ut functions of M 39 the transition relation and output function of the PM created by two identical instances of M 0 Compute the state equivalence relation describing the sets of all equivalent state pairs 0 Compute the equivalence class characterization relation by selecting a representative state from each class of equivalence states 0 Compute the transformed transition relation and the transformed output relat39on MAyZZEIEIEI EcEsmocE EDDsdehz Amlwaonns Product Machine PM 0 Given FSM 1 O S B A with k inputs n stats m output the product machine is 1 01 5x5 52 72 the product machine has k inputs 2n states 1 output EquivalenceDistinquishabiIity Relations State equivalence relation is a boolean function Esl 52 which is true for codes 51 and s2 iff the corresponding states are equivalent State distinquishability relation is a boolean function Dsl 52 which is true for codes 51 and s2 iff the corresponding states are not equivalent MAyZ mun m5 5m OCE EDDs mmmmm m Properties of Equivalence Relation o Equivalence relation is reflexive symmetric e and tranSItIv se the equivalence classes are 0 Suppo 00011110 m5 5m OCE EDDs and Tim Amlwaonns MAyZ mun Computing Equivalence Relation Esl 52 can be computed using the following procedure iterated until Ejs Ej1S E05 Vil M405 51415 Ejs amp Vi3n T irsm amp Ejn where Ejn Rs gtnEJs and Rs gtn is the variable replacement operator m5 5m OCE EDDs and Tim Amlwaonns MAyZ mun Computing Distinquishability Relation Dsl 52 can be computed using the following procedure iterated until Djs 01415 D05 3H lpMiS 01415 Djs 3i3nl T irsm amp Djn where Djn Rs gtnDJs and Rs gtn is the variable replacement operator MAyZ mun m5 5m OCE EDDs mmmmm m Deriving Esl 52 from Dsl 52 o ARs1 is the set of reachable states AR51 352 E5152 or AR51 352 D5152 o The equivalence relation is derived as follows E5152 D51Sz39 amp M51 amp A9252 0 Similarly for the distinquishability relation D5152 E51Sz39 amp M51 amp A9252 MAyZ mun m5 5m OCE EDDs andThurAfphcaonns Equivalence Class Characterization Relation o Equivalence class characterization relation E selects exactly one representative from each equivalence class of states defined by Esl 52 sl 52 is a boolean function that is one for the codes 51 and s2 iff the state corresponding to 51 represents the state corresponding to 52 sl 52 CProjection Esl 52 x0 where x0 is the reset state expressed using variables 51 MAyZ mun m5 5m OCE EDDs andThurAfphcaonns Reduced State Set Transition and Output Relations of StateMinimum FSM o The state set of the minimumstate FSM Qsl 352 E61 52 0 Transition relation of the stateminimum FSM Tminllslln1 352m Ti52n2 amp 26152 amp n1n2 0 Output relation of the stateminimum FSM 0mmi1510 352 005213 amp 26152 MAyZ mun m5 5m OCE EDDs andThurAfphcaonns Compatible Projection Operator 0 Given an equivalence relation Egtlt1 X2 0139quotgtlt 01m a 01 the compatible projection is a boolean function F09 X2 Xuxz I X1 X2 E gtlt2 SELgtlt1 where SELgtlt1 is a selection function that uniquely selects one representative from each equivalence class Mayzzuuu EcEsmocE anmmmmmmm 19 Pseudocode of Compatible Projection function CProjection E on i on 1 return E ifE0retum0 ifE 1retumoc y1 is the top variable in on if 1 0 Gel xl39 else if aw 0quotoc1 x1 Y 3x1 Eul retum ocl amp CProjection EEl ocul 39oq39 amp CProjection Eur 0cE1 Mayzzuuu EcEsmocE anmmmmmmm 2n Complete Source Code for CProjection mum m m Inlays oIF wllh my lo um Illerzl m r n mum n m m mam bdeexlntKFy hlenl 1 cur bdeexlntKFy Illlenl ddlme bddhlse u bddlme m um domzm when mm does em hukudm Inrrezdymzde mun Md Dun2n bdLexslt vascum ulAmn m mm BhoyullonPoxEolF mud m mm sum cpmumnr mam mm m m Inan mml mum mm a luaum m bddhlse lhe lop gum mg lhe mun m 21h relum Rem a m nol m zxxe ni Maj2 mun m5 5m OCE EDDs mmmmm m MAyZ mun Homework A Study of Random FSMs 0 Generate transition relations of random FSMs with N states and K transitions in each state 0 Perform reachability analysis using the generated transition relations and determine the number of reachable states and the number of iterations in the FSM traversal 0 Assume N 10000 K 1210 Draw a graph visualizing the number of reachable states the number of iterations and the time needed to complete the reachability analysis as a function of K EcEsmocE EDDsdehurAfphcaonns 22 Design Problems Design a plausible process ow to fabricate the structure represented in the next View graph Follow the ideas of Silicon fabrication process ow You do not have to include any quantitative process parameters times temperaturesdoses etc Your answer should be given in terms of a series of sketches of the structure after each major process step Brie y explain your reasoning for each step and the order in which the process was selected ECE 510 Lectures Spring 2007 Design Problem 1Micromachined Platform A miniaturized sensor for continuous lactate measurement in saliva was developed and tested The sensor was built using silicon microfabrication technologies The size of the chip is 55mmx64mmgtlt07mm and features a working a counter PtTi and an Iridium reference electrode The chip has a cavity whose oor is perforated by fine pores The cavity contains the enzyme lactate oxidase LOD which is immobilised in an agarose gel Prior to the amperometric detection of the reaction product hydrogen peroxide at the working electrode the analyte lactate has to pass the pores to reach the cavity with the lactate oxidase by diffusion ECE 510 Lecture5 Spring 2007 0 The chips were fabricated using double sided polished silicon on insulator SOI wafers 0 The thicknesses of the layers of siliconsilicon dioxide and bulk silicon of the SOI wafer are 12 um300 nm and 675 Mm respectively ECE 510 Lecture5 Spring 2007 Cross Sectional View of the Device P1 in Titanium at um Irldlum counter 5 reference Elemde llactate electrode aquot at an 39 x H w 39 N 30 nm Silicon Wnrk ng POIYester g aling foil dIOXIde electmde PlatinumTitanium 200 nm Silcion Oxy Nitride ECE 510 Lecture5 Spring 2007 Design Problem 2 3D Micro fabricated Platform 0 Design the process ow for a micro uidic system for manipulation and separation of micronsized particles based on the combined use of negative dielectrophoresis DEP and hydrodynamic forces 0 A 3D microelectrode structure so called paired electrode array is constructed face to face on the top and bottom sides of the microchannel and driven with high frequency AC voltage to generate dielectrophoretic gates Depending on the relative strengths of the two forces particles such as polystyrene beads or cells carrying by a laminar ow can either penetrate the gate or settle there ECE 510 Lecture5 Spring 2007 Cross Sectional View of the Device gt PyreX Glass of 4 in Width 75 micron height and 800 micron Width 7oao00aao00aao00aao00aao00aao00aao00abobbbbobbbbobvhoobbbba 39 39 66666666666666666666666666666666666666666666666660 tb0000000000000000000000000 65660 p ZSDHUIOO nn1 CrAu Silicon 1pm Si Outlet depth of waferwidth 600 AC Fi ld FIOW micron ECE 510 Lectures Spring 2007 Design Problem 3 Planar micro fabricated platform 2 Neuronal cell networks have been constructed on planar microelectrode arrays MEAs from dissociated hippocampal pyramidal neurons 0 Micropatterning in 2D and a photoresist liftoff method were used to selectively identify metallic locations for localization of neurons ECE 510 Lecture5 Spring 2007 Design Problem 3 J55555555553mama 55 5555555555553555555555555v5555vv 54455 quot 55555555555A 5av 555 555555555 545555 39393939 3355 515555515 55r555 5555555 5555 55555 55555 kkki39 IMAMquot 39 W 555552555555 55555 H 39 egg 1 55555 W a 55555 W w 55555 W 55555 55 55555 55 55555 5 55555 55555 55555 x 55555 5 55555555 55555 539 u 55 5555 555 555555 55 555 555555 55555 55555555 5 55 39555555quot 5 93 5 65 5 3815 5 5 39 395395395395395 93 z 95 35 55555 55 5555 5 55555 5555 13 55555 5 55 5515 lt5 5555555539555 55 55555 5 55555 m 5 55555 2555555525 quot 555v555 WM 55555 595555555 555555 5555555 555 s 5555555 1 quot 15 55 5 555555 5555 g5 5 5 5555 55555 395 933 95 395395395395395395 quot WM 5555 quot WM 5555 quotquotquotquot39quot quot WM mauvw quot3 quot 3999 35 5 5 3 v quot 55555 5555 5 55 55555 5555 3939 39 55555 555555 55555 552555555 55555 555155 55555 5555555 n x m no m m 55555 5555555 55555 5555555 5 555 5555555 5 55 55555555 55 5555555 5 933 33 39 5 5x5 33 lt 395 933 3 539 55555 555 5 5 55555 5555 5 55555 555 5 55555 555 5 55555 555 5 55555 555 5 5 3999 33 539 55555 555 5 5 5353 may 515 555395 5 525 39 M 55 H 5 1 5 5 5155 5555555 5 5555555555555555 555555555555555 2 55535155 g z 5 5555 5 5555555 55553335555355 5555555555555555555555555 5555 555 5 39 3953 525 5555555 53 1515153355555 5 5 5 5 55 5 555 5555555555555 5555555 5555555 5555 5 5 55555555555555555555 55555555555555555 5555 55555555555555555 5 555 555555 I 395 5555555555555 39539 55555253a5 15355555555535 5 5 15 555555555 55555 55555 5555555555555555555 5 t new a a IIIIIIIIII 55555553 mum r 5555555555555955555555H 555552555555553r5555 55555 5 55 555 55555 5555555 Etttttstttttttttttt 5555 55 55 5 935255 55 5554525 525225 V 55523 515 g 555555555555555 5 55555555555 5 555555555555 5 5555 3 0 55 5 5 5m v5mwm wg 555 231 3 159155 35 553655515 55 W I a 3 5 5 539 15 5 a 35552555555555 3amp3 35555555 3555334 A 93 3239 W 52535 5 555 355552 5 3 55 5555555 555 55 55 555 quot5 5 333 33924 5555 5555555555555555555555 555 55 N 5555 39 39 39 39 39 quot 5 5 55 5555 5555 55395555555 55535555553553555555z5z555555z5555555555553235333335555 5555555555555555555555555555555555551 555355 5555555555555a5555555555555555 55a555555555555555 5 5555555555555555555555555555555555555555555555555 5555 A 555 55555555555555555555555555555553zgagggg yw5 55 5 r 55 5z5a5555353315555555gr5555grggggggggggggggg www 55 35 353353 33 3 2 3 539 5 fv u 5 55555555555555555555555555555 55 at 5555555555555 5 5 3 a5aav w aavac 55555525 5 1 55 5555555555555555mm we 3 3935555 5 3 3 3 3 3 r 3 r awawwwwwwxwww TZ 3553933133313333 3 T 355 5 39 a 5 5555555555 555555555555555555555555555555555555555555555555 55 5 55 5 5555555555515555555555525555555555555555555555WW 555 395 55555 3 53535525255552513353552525 5155 535555555152535355 5 5 5 5 5 5 515 5 55555a545a55555555555555555555 5555555555555 aam dwlma 5 5045 5 5 a z a 3quotquot 96Mwuaggg55 55555555555 55 5 55555555555555555555555555555555555555555555555555m5555555555555555555555555555555555555555555555 5555 555 m 5 5 55 5 5 3 w 5 w 555 55555 3955555355555555555555 55 5 5555555555555555555555555555555555555555555555555555555555555 555 5 w 5 5 N m 555 5 5 t 395t 555555553555555555 5 5555555555555555555555555555555 5 5 5t5555555555555555555555555555555555555555555555555555555555555 51555z5z5z555 55z55z5555 5555555 5555555 55 gt quotquot 33955 5 5 5 5 5 5 5 5 5 5555555555555555555wwwwwwww5 5 539 55 5 5 5555355 N b optical Imcrcgraph 0f m1cr0electr0des 0n the device Scale bar 200 mm c SEM micrograph of a set of microelectrcdes Scale bar 25 um ECE 510 Lectures Spring 2007 Cross Sectional View of the DeV1oe Platinum Black 1 micron OxideNitride Oxide 1 km mm W p n n mm quotW SiliconSilicon Dioxide Titanium 5 lOnn1 Gold 1 micron 9250 300 nm ECE 510 Lectures Spring 2007 1 11 lIll lrersalH SETH Li 1 SITE Nanotechnology and Biosensors Dr Shalini Prasad Electrical and Computer Engineering Biomedical Microdevices and Nanotechnology Laboratory sprasadpdxedu httpwwweceEdxedu rasads Le t g ECE 510 c ure1 Sprln 20 8 s Prcasadssgiliden Course Outline 1 Introduction to BioMEMS and Nanobiotechnology 2 Silicon Microfabrication 3 Soft Fabrication Techniques 4Poymer Materials and Microfluidics 5 Sensor Principles 6 Detection and Measurement Methods 7 Drug delivery systems 8 MicroTotal Analysis Systems uTAS Emerging Applications 9 Nanotechnology Applications ECE 510 Lecture 12 Spring 2008 S Prasad Slide2 Courtesy39 SS Saliterman Electrokinetic Phenomena 39 Harnessing electrokinetic phenomena in micro uidic devices for moving uid and particles including proteins cells bacteria and Viruses is essential for MTAS and other LOC applications ECE 510 Lecture 12 Spring 2008 sPrCas1adgS S tlide3 Electrokinetic Phenomena o Electraosmosis Fluid movement relative to a stationary charged or conducting surface through application of an electric field 0 Electrophoresis e pre ence of an electric eld the particle can be induced to move relative to the stationary or movrng liquid a Streaming potential Occurs when an aqueous ion containing solution is forced to flow through a capillary or microchannel under an applied hydrostatic pressure in the absence of an applied electric eld An eectroviscous effect occurs or resistant to flow 0 Dielectrophoresis Movement of dielectric particles in a spatially nonuniform electric eld 3 3 ECE510 Lecture 12 Spring 2008 5 Prgfad Slide4 My 55 stemquot Formation of Electrical Double Layer 0 Application of electric field induces the formation of EDL The EDL consists of the charged surface and the compact liquid layer containing the immobile balancing charges and a a i ase liquid layer of mobile ions 39 From the compact layer to the electrically neutral bulk liquid the net charge density gradually reduces to zero ECE 510 Lecture 12 Spring 2008 sPrCals1adgS Silide Formation of Electrical Double Layer 39 The compact layer is several angstroms thick 0 The shear plane is the boundary between the compact layer and the diffuse layer Where the liquid velocity is zero ECE 510 Lecture 12 Spring 2008 sPrCas1adgS Sindee Electric Double Layer EDL ear Liquid Plane I O c0i0ns countereions Solid Surface Diffuse layer gt Compact layer mobile ions immobile ions ECE 510 Lecture 12 Spring 2008 s Prasgq S iide Electric Double Layer Potential PS shear plane 4 Potential measurable WK 0 f 3 5 k Compact Layer Diffuse Layer ECE 510 Lecture 12 Spring 2008 S Prg s ggs EIectroosmotic Flow 0 When an electric field is applied the excess counter ions in the diffuse layer move under the applied electric force this is called electro osmosis 39 Electro osmotic Flow occurs when the moving ions drag the surrounding uid with them due to the Viscous effect creating bulk ow ECE 510 Lecture 12 Spring 2008 s Prcasad Iide9 ounesy SS Salter EIectro osmotic Flow EOF gt i Electric Double Layer Negativer charged wall ECE 510 Lecture 12 Spring 2008 S Prasgd lideno Relevance of EIectroOsmosis in nFluidic Devices 0 Many solidliquid and liquidliquid interfaces have electrostatic charge and consequently an electric field near the interface 0 These interfacial electrokinetic phenomena are important to micro uidic processes ECE 510 Lecture 12 Spring 2008 S Prcalsfid S S lidem Calculations Assumptions 0 Uniform zeta potential 0 Electric double layer is thin compared to the channel dimensions 0 Electrically insulated channel walls 0 Low Reynolds numbers 0 Parallel flow at inlets and outlets 0 Uniform fluid properties 0 Constant viscosity and electrical permittivity ECE 510 Lecture 12 Spring 2008 S Prgsgds Q idez12 Poisson Equation 0 The Poisson equation describes the electrical field potential in a dielectric medium Where V is the operator quotdelquot from F 391 JX 3f391OJ VJIE av a Li39 psi is the electrical eld potential V p rho is the free charge density 39391113 or coulomb per Cubic meter 5 epsilonrelative is the dielectric constant of the medium and 5D epsilonnought is the peimittivity of a vacuum 85 gtlt 103912 Film or F ar39adav per meter Note that when there is no free charge p O and V211 0 the Laplace aquarium ECE 510 Lecture 12 Spring 2008 S Prfxsnad W Snlidez13 Origin of Surface Charge t Different affinities for ions of different signs to two phases The distribution of anions and cations between two immiscible phases such as oil and water Preferential adsorption of certain ions from an electrolyte solution onto a solid surface or Preferential dissolution of ions from a crystal lattice 2 Ionization of surface groups 3 Charged crystal surfaces ECE 510 Lecture 12 Spring 2008 S Prgsgds gment Boltzmann Equation 0 Both glass and polymer micro uidic devices tend to have negatively charged surfaces 39 The electrostatic charges on the solid surface attract counter ions in the liquid 39 The concentration of these counter ions is higher near the surface than in the bulk uid further away from the surface ECE 510 Lecture 12 Spring 2008 S Prcasad Iide15 ounesy SS Salter Boltzmann Equation 0 The Boltzmann equation is used to describe the ion and potential distributions of the diffuse layer 0 The diffuse layer thickness is dependant on the bulk ionic concentration and the electrical properties of the liquid 0 The potential at the shear plane is called Zeta Potential and can be measured experimentally 0 It is the approximation of the surface potential ECE 510 Lecture 12 Spring 2008 sPrCals1adgS Sillidene Boltzmann Equation The distribution of ions near a charged surface can be expressed by the Boltzmann equation 7 X izieu I ll 7 11 exp ka where n 15 the ionic ntunber concentration of type 139 ion at a given position 71 is the ionic number concentration of type 139 ion in nitely away from the charged stuface is the value of the ionic Valence e is the fundamental charge of an electron 1602 x 103919 C L psi is the electiical eld potential V at a given position kb is the Boltzmann constant 138 x 103923 JK and 139 is the absolute temperature in Kelvin ECE 510 Lecture 12 Spring 2008 S Prasyd iidez17 Poisson Boltzmann Equation 0 The PoissonBoltzmann equation is used to describe the ion and potential distributions in the diffuse layer zeal V ili 7 smh 1le 7 Zzen0 Vt r 39lquot a A more general form of the equation considering charge density and inclusive of the bulk ionic concentration and valence of type i ions is e 2211 exp Vl u I I V r F We L urm ECE 510 Lecture 12 Spring 2008 S Prgsgds Q idez1s Debye Huckel Parameter o By defining the DebyeHucke parameter 3 228277 k H 39 39appa 130ka 11 psi ka and the nondimensional electric potential we may rewrite the PoissonBoltzmann equation as Vz lf m2 sinh 11 SPrasad lideng Lecture 12 Spring 2008 ECE 510 DebyeHuckel Parameter 0 The DebyeHuckel Parameter is independent of the solid surface properties and is determined by the liquid properties of the ionic valence and the bulk ionic concentration ECE 510 Lecture 12 Spring 2008 S Prcasfid S S tlide20 EDL o The characteristic thickness of the EDL is defined as ws 1 Shea plane lt Potential measurable W01 Jg Cumpm Laycr Dif lse Layer ECE 510 Lecture 12 Spring 2008 s Pr sad slide21 a 00 r1257 53 Samvmah EDL o The bulk ionic concentration may be expressed in terms of molarity x M E L 1000L3NA IOOONAM In 11101 and the characteristic thickness of the EDL is 12 l 5750ka 7 7 r 22 e I1 ECE 510 Lecture 12 Spring 2008 s PrgamszgS slidezzz s Saiievmaii Thickness of Diffuse Layer Temperature T 298K Dielectric constant 51 785 Pennittivity of a vacuum 230 885X103912F1n Faradmeter Charge of an electron e 1602X103919C Boltzmann s constant kl 1 3 8 l X 10 23 JK Avogadro s number N A 6022 X 102311101 12 1 E E Tl 33904 x1040 m K 22 6quot m ECE 510 Lecture 12 Spring 2008 s PrgamszgS slidezza s Sauemn Example 0 Consider different molarity solutions of a symmetric 22 11 solution like KCl or NaCl o At 10396 M the ionic concentration is considered practically water The thickness of the diffuse layer may be larger than a micron for pure water and pure organic liquids M 1llt nm 106 104 102 ECE 510 Lecture 12 Spring 2008 s Prf szd idezu EOF Velocitv o In a slit microchannel formed between two parallel plates where the width is much larger than the height of the channel the electroosmotic flow velocity may be determined by the HelmholtzSmoiuchowski equation EOg 39 H 39 7 1m Where rm is the average electroosmotic ow velocity 1115 E is the applied electrical eld Vim s epsilonrelative 18 the dielecnic constant of the medium a epsilonnought is the permittivity of a vacuum 885 x 1039 Fm Q zeta is the zeta potential at the shear plane V and ti nut is the dynamic Viscosity kg1111 ECE 510 Lecture 12 Spring 2008 s Prasad slidezzs Comtesv ss sainevman Difference Between Electrophoresis and EOF Electrophoresis is the effect by which charged species in a uid are moved by an electric field relative to the uid molecules 0 The charged species accelerates until the electrical force is equal to the frictional force 0 In EOF the charged wall is stationary and uid moves under an applied electric field In electrophoresis the liquid is stationary While the particle moves under the in uence of the electric charge ECE 510 Lecture 12 Spring 2008 S Prcasad Iig 26 ounesy SS Salter Electrophoresis 0 Charge distribution around an eiectrophoretic particle ECE 510 Lecture 12 Spring 2008 s Prasad A slide27 WW 5 Electrophoretic Velocity o A particle s electrophoretic velocity may also be calculated by the HelmholtzSmoluchowski equation erz vpp is the particle s electrophoretic Velocity1ns E2 is the applied electrical field W111 5 epsilonrelative is the dielectric constant of the medium 50 epsilonnought is the permittivity of a vacuum 885 x 1039 F111 C zeta is the zeta potential at the shear plane V and p 11111 is the dynamic Viscosity kg111st ECE 510 Lecture 12 Spring 2008 S Pr sadr Slidezzs a a nesv ss Salnman Electrophoretic Motility o Electrophoretic motility is defined as the electrophoretic velocity per unit of applied electrical field strength characterizing how fast a particle moves in an electrical eld ECE 510 Lecture 12 Spring 2008 s Prasad A slide29 WW 5 5mmquot Henry 3 Function 0 A more general formulation for electrophoretic velocity is Henry s equation 3 EI EOEZCP 939quot 3 lL where v f H0 fm is Henry39s rncfion o Henry s function approaches 1 for small H61 and 32 for large ma ECE 510 Lecture 12 Spring 2008 5 Prefde W ll d e w Applications 0 It is commonly done to separate molecules based on size 0 When performed in microchannels the process is referred to as capillary electrophoresis 39 It is generally used in DNA separation ECE 510 Lecture 12 Spring 2008 S Prasad Slide31 Courtesy SS Saliterman Streaming Potential 0 Forcing an aqueous ion containing solution through a microchannel under applied hydrostatic force induces a resistance to ow known as electroviscous effect 0 As uids is pushed through the channel the counterions in the diffuse layer the mobile part of the EDL are also pushed along creating a streaming current 0 A potential difference builds up along the channel wall known as streaming potential ECE 510 Lecture 12 Spring 2008 S Prasad Slide32 Courtesy SS Saliterman Streaming Potential 0 This streaming potential acts to drive the counter ions in the diffuse layer in an opposite direction generating an electrical current called the conduction current 0 This drag on the ow makes the uid seem more Viscous because the predicted ow is greater than What is actually observed ECE 510 Lecture 12 Spring 2008 S Prcasad 19533 ounesy SS Salter Streamina Potential 0 Illustration of the flowinduced electrokinetic eld in a microchannel 4 9900999 gt Conduction GD 9 Q G ouble Layer Ions Current I e 9 9 9 9 G ztreammlg I x Flow Directiongt 9 men 9 3 Q 9 G 9 Q Streaming potential 9 E 9 a I 39 I Q Steady State 1 O ie I IC ECE 510 Lecture 12 Spring 2008 3 PrgamszgS slidezu s Samvmah Dielectrophoresis o The physical phenomenon whereby dielectric particles in response to a spatially nonuniform electric field experience a net force directed toward locations with increasing or decreasing field intensity according to the physical properties of particles and medium In the first case the force is called positive dielectrophoresis pDEP while in the second case it is called negative dielectrophoresis nDEP If the particle is charged it also experiences a coulombic force superimposed on the dielectrophoretic force due to the nonuniform field ECE 510 Lecture 12 Spring 2008 s PrgamszgS slidezas s Sallerniaii ElectroWetting Theoretically it is possible to create droplets from samples and reagents divide transport and merge them for purposes of dilution and reaction isolate them for detection and finally dispose of unwanted waste products 0 Surface tensions is a property of the liquid and is dependant on the temperature and the other uid it is in contact with 0 At the interface between a liquid and a gas or two immiscible liquids forces develop in the liquid surface that causes the surface to behave as if a membrane was stretched over it ECE 510 Lecture 12 Spring 2008 S Prasad Slide36 Courtesy SS Saliterman Electrowetting 0 The phenomena of electrowetting and electro capillary are that an externally added electrostatic charge modifies the surface tension or capillary forces at the uidsurface interface ECE 510 Lecture 12 Spring 2008 S Prcasad 19537 ounesy SS Salter Wetti ng o Young s equation after Thomas Young who first proposed it in 1805 describes the simple balance of force between the liquidsolid liquidvapor and solidvapor interfacial surface energies of a droplet on a solid surface leG OSGLW SL NlSG where am gamma liquidgas is he ligmdgas 11115139facialte11s10111 nSL iganum sohdhqtud 13 the solldhqmd111Ie1fac1nl161151011 Aqueous mm mm gamma solidgas is the solidgas inler rciahmsion and Dielectric layer K in in L39 3 metal is the comet angle Electrode Suballets ECE 510 Lecture 12 Spring 2008 S Prams 3538 Electrowetting i Create mm mm mm U N npul qu s m quot1 m m ix Non wetlmg gm ms undevnupmenhal l V l 39 I n 2 CM Reversble J Iquot 3 M6193 Welling r undere ectri potenlu Menquot gt L Dulvulmoium n M WW L A w H 3331 Electrowetting Digital Micro uidic Circuit inquot at 9 Iquot A ECE 510 Lecture 12 Spring 2008 S Prism Slide39 was 55 Sahverman Electrowetting o The effect of a potential Von the contact angle is then determined by the following coseV cosea iii3i 1L6 where 9 theta is the contact angle 60 thetanought is the equilibrium contact angle at V 0 V is the electric potential across the interface V 5 epsilon the dielectric constant of the dielectric layer 50 epsilon is the pennittivity of a vacuum 885 V 1039 Bin and I is its thickness in ECE 510 Lecture 12 Spring 2008 S PrisquotSS SaSnydezdo Microvalves 0 Passive Valves Check Valves Directional like a diode Smart polymers external stimuli Stop Valves r Surface modifications of hydrophobicityhydrophilicity for immobilization of fluid and materials ECE 510 Lecture 12 Spring 2008 S Prg gs gig Passive Valve 0 Hydrogel check valve V V H a Valve leaflets gt PH 5mm Anchors7 gt Non pH sensitive strips Basic Solu onsC Expanding and closing the valve and Acidic Solutionsd Contacting and opening the valve as i Re 3 Fa d a ECE 510 Lecture 12 Spring 2008 S PrgmgSS 333342 Active Microvalves 39 Active valves may be analog or digital and have an initial state that is normally open normally closed or bistable 0 An analog proportional valve has a ow that is a function of the valve opening 39 To accomplish this digitally there must either be a digital to analog converter or use of pulse Width modulation ECE 510 Lecture 12 Spring 2008 S Prcasfid S S tlide43 Active Valve Types 0 Pneumatic o Thermopneumatic o Thermomechanical o Piezoelectric o Electrostatic o Electromagnetic o Electrochemical o Capillary force ECE 510 Lecture 12 Spring 2008 s Prasad A slidez44 WW 5 elm Electrostatic Valves 0 Electrostatic valves are based on the attractive force between two oppositely charged plates til V3 ad 2 FijrlhoAlFi i l i lJ A is the overlapping plate area It is the distance between plates where Ii is an insulator layer thickness V is the applied voltage 3 epsilonrelative is the relative dielectric coef cient of the medium 5 epsiloninsulator is the relative dielectric coef cient of the insulator and a epsilonnought is the pennittivity of a vacuum rm Hm Lecture 12 Spring 2008 3 Presa Slidez45 unesv ss Saittermaii ECE 510 Micromixers 0 Passive mixers have no moving parts but instead rely on diffusion and geometry of the device 0 Active mixing increases the interfacial area between fluids and can be accomplished by piezoelectric devices electrokinetic mixers chaotic convection ECE 510 Lecture 12 Spring 2008 s Prfszd SQ gimme M ixer Tmixer and Ymixer Flude Fluid2 1qu12 11 a c Fluld1 9mm 1 Fmid1 on 0 on oquot g p I on on o on n time gt ECE 510 Lecture 12 Spring 2008 S Prasgd Slidez47 Serpentine Mixers Serpentine mixers r Flurd 2 W Fluid 1 a Fluid 1 b Q Fluid 2 Q ECE 510 Lecture 12 Spring 2008 s Prasad A slidez4s WW 5 gammaquot Summary 0 Electraosmosis Fluid movement relativeto a stationarycharged or conducting surface through application of an electric field 0 Electrophoresis In the presence of an electric field the particle can be induced to move relative to the stationary or moVIng Ilquld a Streaming potential Occurs when an a ueous ion containin solution is forced to flow through a capi lary or microchanne underan applied hydrostatic pressure in the absence of an applied electric eld An electroviscoLs effect occurs or resistant to flow I Dielectrophoresis Movement ofdielectric particles in a spatially nonuniform electric field ECE 510 Lecture 12 Spring 2008 S Prga sgdS Slidez49 s mamquot Summary 9 The Poisson equation describes the electrical field potential in a dielectric medium 0 The Boltzmann equation describes the distribution of ions near a charged surface 0 The PoissonBoltzmann equation is used to describe the ion and potential distributions in the diffuse layer 0 The DebyeHuoke parameter is used to define the characteristic thickness of the diffuse layer 0 The HelmholtzSmoluchowski equation is used for both electroosmotic flow velocity and electrophoretic velocity determination ECE 510 Lecture 12 Spring 2008 s Pr a ggS slidezso s sailemian Summary o Young39s equation describes the simple balance of force between the liquidsolid liquidvapor and solid vapor interfacial surface energies of a droplet on a solid surface 0 Microvalves micromixers ECE 510 Lecture 12 Spring 2008 S Prg gs i ezm 1 11 lIll lrersalH SETH Li 1 SITE Nanotechnology and Biosensors Dr Shalini Prasad Electrical and Computer Engineering Biomedical Microdevices and Nanotechnology Laboratory sprasadpdxedu httpwwwecepdxeduprasads ECE 510 Lecture 5 Spring 2008 s Prasad Slide1 Course Outline 1 Introduction to BioMEMS and Nanobiotechnology 2Siicon Microfabrication 3 Soft Fabrication Techniques 4 Polymer Materials and Microfluidics 5 Sensor Principles 6 Detection and Measurement Methods 7 Drug delivery systems 8 MicroTotal Analysis Systems uTAS Emerging Applications 9 Nanotechnology Applications ECE 510 Lecture 5 Spring 2008 s Prasad Slide2 3D Structures with Sacrificial 3D Structures with Sacrificial Layers Polysilicon is the main structural element deposited as amorphous silicon annealed at 58090 Disadvantage is that LPCVD polysilicon films are thin and there are dimensional uncertainties without crystallographic control ECE 510 Lecture 5 Spring 2008 s Prasad Slide4 3D Structures and Sacrificial Layers ECE 510 Lecture 5Spring 2008 S Prasgd Slide5 Structural Elements Common structural elements include polysillcon polyimide silicon nitride and tungsten The polysilicon and its sacrificial layer silicon dioxide can be applied by LPCVD Silicon dioxide can be etched away with hydrofluoric acid l lF solution without etching the polysilicon Polyimide can be used with aluminum as the sacrificial layer the latter being dissolyable with acidbased etchants Silicon nitride is both a good structural material and electrical insulator Polysilicon can be used as the sacrificial layer in which case KOl l and EDP can be used as the etchants Tungsten can be applied by CVD over silicon dioxide and again llF is a suitable etchant to remove the silicon dioxide sacrificial layer Steven S Salfrerman MD FACP Lecture 5Spring 2008 8 Prasad Slide6 Stiction and Release 39 Wet release processes cause stiction or sticking of suspended structures to the substrate 39 Stiction can be minimized by using critical point drying 0 The critical point process occurs at a specific temperature pressure and density such that there is no apparent difference between the liquid and gas state ECE 510 Lecture 5 Spring 2008 s Prasad Slide7 Stiction and Critical Point Drying pmysmmn yearn Spacer made 5 Rinse water ECE 510 Lecture 5 Spring 2008 s Prafad Slige Silicon on Insulator SOI Single layer epilayer is used and surface features up to 100 micron are obtainable More reliable and reproducible sensors can be produced using 01 01 is achieved by anodic bonding SiOXideSi Increases the density of deVices ECE 510 Lecture 5 Spring 2008 s Prasad Slide9 Stress and Strain Properties of thin film is different from bulk with smaller grains and increased surface area to volume ratio Stress is defined microscopically as the force per unit area acting on the surface of a differential volume element Stress perpendicular to the surface is called normal and those acting along the sides of a surface are shear Strain is the differential deformation of a solid in response to a force ECE 510 Lecture 5 Spring 2008 s Prasad Slide10 Deep Reactive Ion Etching DRIE High aspect ratio micro structures using plasma sources Inductively coupled plasma and electroncyclotron resonance Etch rate is diffusion limited and decreases with increasing aspect ratio Inductively coupled plasma is generated by a helical resonator combined with an electrostatic shield to produce E fields that are circumferential in response to the aXial RF magnetic field Electroncyclotron resonance uses a microwave source that interacts with magnetic fields Produces high density plasma at low pressure and low energy plasma ECE 510 Lecture 5 Spring 2008 s Prasad Slide11 Deep Reactive Ion Etching DRIE High aspect ratio etching obtained by Bosch Process 0 Glow discharge processes create polymeric species by chemical crosslinking The deposition of the material is slower than its removal and etching is stopped or passivated by this layer 0 The Bosch Process alternates etching and passivation allowing deeply etched trenches Microscopically Visualized as a scalloped side wall ECE 510 Lecture 5 Spring 2008 s Prasad Slide12 Deep Reactive Ion Etching DRIE Mask DRIE using Bosch 7 ncrx 7 Process The process M i I 3 cycles between SFB gas DeP SquotP e etching and 04F8 polymer deposition I in N m ECE 510 Lecture 5spring 2008 s Prasad Slide13 Applications in BioMEMS Nanowires for molding to make nanocapillaries Doherty et al 2003 Micromirrors used for laser beam scanning in optical coherence tomography Xie etal 2003 0 Etch silicon and aluminum membranes for piezoelectric aluminum nitride thin film ultrasonic transducers Valbin and Sevely 2001 Piezoelecctric micro array jet for drug delivery Yuan et al 2003 ECE 510 Lecture 5 Spring 2008 s Prasad Slide14 LIGA LIGA an acronym from German words for lithography electroplating and molding is a micromachining technology By using the penetrating power of xrays from a synchrotron LIGA allows the fabrication of structures which have vertical dimensions from hundreds of microns to millimeters and horizontal dimensions which can be as small as microns These are 3D microstructures defined by 2D lithographic patterns ECE 510 Lecture 5 Spring 2008 s Prasad Slide15 UVLIGA UVLIGA is an approach where a proximity UV aligner is used with a thick resist in place of the synchrotron xray exposure step After the lithography electrodeposition and planarization are used to produce metal microparts or a metal replication tool This approach has the advantage of achieving high aspect ratio patterns using standard lithography processing tools instead of a synchrotron However modifications andor improvements of available resists along with modified processing techniques are required to successfully achieve a high quality resisttoparts process ECE 510 Lecture 5 Spring 2008 s Prasad Slide16 ECE 510 LIGAUVLIGA Hyml ultrnxl nllrl1d V 139 39r Lecture 5Spring 2008 S Prasad Slide17 Electroplating Provides thicker layers of metal structures greater than 100 s of microns The substrate is placed into a solution of desired metal ions and serves as the negative cathode The positive anode is typically made of the desired material to be deposited commonly nickel When the current is applied ions are reduced at the substrate surface become insoluble andare deposited on to the substrate surface ECE 510 Lecture 5 Spring 2008 s Prasad Slide18 Substrate Bonding Silicon Direct Bonding Si iicon to Silicon Siiicen en Insulator SDI u Anedic Bending Siiicnn to Giaee Pyrex 7740 in intermediate Adhesive Layers in Lasers ECE 510 Lecture 5 Spring 2008 s Prasad Slide19 Silicon Bonding 0 Bonding Si to Insulator allows the creation of enclosed 3D channels for micro scale uid transport 39 Si to Sisurfaces rendered hydrophilic and brought tightly together attracted by Van der Waals forces and subsequent high temperature annealing ECE 510 Lecture 5 Spring 2008 s Prasad Slide20 Anodic Bonding Ca mdc Depletion quotquot MOMquot layer 39 Hm plan I Lf liaie E et a Hard and So MI cmmachming for BFDMEMS 2004 ECE 510 Lecture 5Spring 2008 8 Erased 7 ilte21 Anodic Bonding 0 Si to glass by compression and heating to 3OOC 4OOC Then applying lOOOV of DC current across the composite with glass as the negative cathode and Si as the positive anode 0 Adhesive residues low bonding strengthheating of entire parts during joining and poor long term stabilities are disadvantages of conventional techniques 0 Use of laser allows small area bonding in sub micro regime ECE 510 Lecture 5 Spring 2008 s Prasad Slide22 Applications of Anodic Bonding 0 Packaging of sensor Chips Najafi 2003 0 Sealing pump Chambers Ridgeway et a1 2002 0 Adjoining pumps to REI etched Si Channels for bidirectional micro uidic regulation Chang et a1 2002 ECE 510 Lecture 5 Spring 2008 s Prasad Slide23 Summary Important thinfilms inoiude o Thermai Siiioon Oxide Silicon Dioxide Si zi Poiysilioon Siiioon Nitride Siam Pnoonhosiiioate GIBBS iF SG Metal fiiins egg tungsten and aluminum 0 Thin fiims may be produced by Therinoi oxidation 0 PiiniC El Vapor Deposition o Chemioat Vapor Depoaition o Epitaxiat Deposition ECE 510 Lecture 5 Spring 2008 s Prosad Slide24 Summary Ion irripiahtatiori may be used to improve eiectrical conductivity or to control etching characteristics Wet buik micromachining isotropic and anisotropic etching a 3D structure and sacrificiai layers in Deep Reactive Ion Etching BRIE a LIGA and UV LIGA Substrate bonding in Silicon direct bonding in Anodic bonding ECE 510 Lecture 5 Spring 2008 s Prasad Slide25 ECE 510 OCE BDDs and Their Applications Lecture 8 Image Computation and Reachability Analysis April 20 2000 Alan Mishchenko Overview 1 Relations as the most fundamental representation of discrete phenomena Image computation for relations Representing FSMs using relations Reachability analysis as an exploration of the state space of FSMs using the transition relation Applications of reachability analysis Computing the transitive closure of relations Equivalence checking of FSMs April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Boolean Functions and Relations 0 Function is a mapping gt 0 Function with DCs is a mapping quot 0 Relation is a mapping gt Example F X1IX2 YIIYZIY3 1 n April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Example 3 000 B3 001 010 011 100 101 110 111 April 20 2000 ELE 510 OCE BDDS and Their Applicanons HHHHOOOOJ Definition of Image Given the relation if 7 n gt Bm the set of all vertices In B is the input domain while the set of all vertices in Bm is the output domain The domain of the relation is the subset of the input domain for which the relation is defined The range of the relation is the subset of the output domain which under some inputs may be the value of the relation Fi Given subset X of the relation s domain Xe Bquot the image of X wrt the relation R is the subset Y of the range Ye Bm composed of values the relation can take if its input values belong to X imRXu ECE 5 10 OCE BDDS and Their Applicaiions April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Image Computation Given the relation X Y and the set of input assignments X it is possible to compute the set of assignments Y such that X and Y satisfy relation X Y Y X Y X Y is the forward image of the set X in the relation X Y Similarly it is possible to compute the backward image X of the set Y April 20 2000 ECE 510 OCE BDDS and Their Applicalions Example Given the relation Rand the set X 2 001 110 the image is Y imRX 61 11 3000 I B3 001 April 20 2000 ECE 510 OCE BDDS and Their Applicalions Reducing Relations to Functions 0 Boolean relation over variables X1X2Xn ym can be represented as a on which is 1 for a given minterm iff this minterm represents related assignments of variables X1X2Xn and YIIYZIIYm39 Example continued x1 x3 39I39I HHHHOOOO HOHOHOI IO HHOOOOOOL HHHOHHHOL 1 other OHHHHHHHH FSM Transition Relation o v is I O S 8 k Suppose r is the number of inputs I rn is the number of states S and r is the number of outputs O 8kis and kkis are vectors of nextstate and output functions 11 1 39 1 is a boolean function 01r 01m 01m 01 such that i s n 1 iff state n can be reached in exactly one transition from state s when input i is applied April 20 2000 ECE 510 OCE BDDS and Their Applicaiions FSM Output Relation a 39 is a boolean function 01r 01m 01n 01 such that i x o 1 iff output 0 can be produced when the FSM is in state x and input i is applied April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Deriving Transition and Output Relations from NSOutput Functions o If the FSM is given as I O S 5 k where 5kis and kkis are vectors of nextstate and output functions it is possible to compute the transition and output relations as follows T isn quot nk quot 5kis O isO 0k vkis Given the transition and output relations it is possible to derive the next state and output functions 5kis 1 T isn 1 nk X at k ECE 5 10 OCE BDDS and Their Applicaiions Reachability Analysis for FSMs o Reachability analysis is the use of image computation to derive the set of all reachable states by traversing the STG of the FSM starting from the set of initialreset states 0 The set of reachable states can be used to simplify the initial transition and output relations to re encode the FSM to check whether a property is true for these states in equivalence checking in symbolic model checking April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Computing Reachable State Set Reached ResetStates do ReachedBefore Reached Reached 211m Reached while ReachedBefore Reached the state sets in this procedure are given by characteristic functions represented using BDDs April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Simplifying Transition and Output Relations Using Reachable State Set 0 Suppose RX is the reachable state set then the transition and output relations can be rewritten T isn TiSn Rs Rn O iso Oiso Rs April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Definition of FSM Equivalence ll State machines are equivalent if starting from the reset states for any sequence of input vectors they respond by producing the same sequences of output vectors April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Equivalence Checking 0 Construct the transition relation of machines M1 and M2 0 Find the quot transition relation P 0 Perform state traversal of the product machine while checking its output 0 If the output of the product machine is 1 for all reachable states the machines M1 and M2 are equivalent 0 It is possible to define equivalence relative to any subset of inputs and outputs of the FSM April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Product Machine 0 Given x I O S 8 k the is I 01 5x5 82 l2 3 inputs 2 states 7 output a 01 M1 April 20 2000 ECE 510 OCE BDDS and Their Applicaiions Equivalence Checking Formulas o Machines M1 and M2 are equivalent iff I le r iley where xy is the characteristic function of the set of reachable states of product machine and ixy is the output relation of product machine 0 Alternatively machines M1 and M2 are no equivalent iff le 39 iIXIy April 20 2000 ECE 510 OCE BDDS and Their Applicaiions ECE 510 OCE BDDs and Their Applications Overview 0 The concept of functional decomposition 0 Two uses of BDDs for decomposition as a computation engine to implement algorithms as a representation that helps finding decompositions Two ways to direct decomposition using BDDs bound set on top LaiPedramVardhula DAC 93 free set on top StanionSechen DAC 95 other approaches Disjoint and nondisjoint decomposition Implicit support minimization May 16 2000 ECE 510 OCE BDDS and Their Applicaiions Applications of Decomposition o MultiIevel FPGA synthesis 0 VLSI design 0 Finite state machine design 0 Machine learning and data mining May 16 2000 ECE 510 OCE BDDS and Their Applicaiions TwoLevel Curtis Decomposition i i 3a F 7 li B Bound Set XEBwF F Free Set if B n F Q this is disjoint decomposition if B m F 4 Q this is nondisjoint decomposition May 16 2000 ECE 510 OCE BDDS and Their Applicaiions Column Multiplicity Beund setzm b Incompatibility Free Set 00 01 11 1o Graphr A Cioi May 16 2000 ECE 510 OCE BDDS and Their Applica ons MultiLevel Curtis Decomposition Twolevel decomposition is iteratively applied to new functions Hi and Eli until smaller functions at and Hi are created that are not further decomposable One of the possible cost functions is Decomposed Function Cardinality It is the total cost of all blocks where the cost of a binary block with n inputs and m outputs is m May 16 2000 ECE 510 OCE BDDS and Their Applica ons Typical Decomposition Algorithm 0 Find a set of partitions quot of input variables I 2 1 into bound set variables Bi and free set variables iii 0 For each parti on find decomposition All 1 such that the column multipliCIty is minimal and compute DFC 0 Repeat the process for all partitions until the decomposition with minimum DFC is found May 16 2000 ECE 510 OCE BDDS and Their Applicaiions Uses of BDDs for Decomposition 0 Whatever is the decomposition algorithm BDDs can be used to store data and perform computation using cubes partitions etc Alternatively the algorithm may exploit the BDD structure of the function F to direct the decomposition in the bound set selection column multiplicity computation and deriving the decomposed functions E and H Finally the process of decomposition may be made fully implicit May 16 2000 ECE 510 OCE BDDS and Their Applicaiions BDDBased Decomposition 0 Bound set on top i 0 Free set on top quot dominators Recursive decompositio M iii 0 Implicit decomposition Bidecomposition using 1 0 and EXOR A quot May 16 2000 ECE 510 OC39E BDDS andTheirApplicil ons Bound Set on Top Function G Gg1ga Ag1 g3 Bg1ga 9 g2 quot May 16 2000 ECE 510 OC39E BDDS andTheu Appllcil ons Bound Set on Top Function H H gl 7 r H HZgOIgIIQI gogld gOIgl g May 16 2000 ECE 510 OC39E BDDS and Their Applications Bound Set on Top Algorithm Reorder variables in BDD for F and check column multiplicity for each bound set For the bound set with the smallest column multiplicity perform decomposition derive functions and ii Iteratively repeat the process for functions and i2 typically only El This algorithms can be modi ed to work for nondisjoint decompositions but does not work with DCs May 16 2000 ECE 510 OC39E BDDS and Their Applications May 16 2000 Free Set on Topquot Algorithm 0 Find good variable order 0 Derive implicit representation of all feasible cuts on the BDD representing F Use a cost function to find the best bound set and perform decomposition Repeat the process for functions and F This algorithms is faster than bound set on topquot but it does not work for nondisjoint decompositions and with DCs May 16 2000 ECE 510 OCE BDDS and Their Applicaiions NonDisjoint Decomposition o Nondisjoint decomposition can be reduced to disjoint decomposition by adding variables i Free Set L Disjoint decomposition can be generated by introducing variables 3 22a instead of o In terms of the Karnaugh map it is equivalent to introducing two variables instead of one in such a way that clc2 is a don t care set May 16 2000 ECE 510 OCE BDDS and Their Applicaiions NonDisjoint Decomposition Example AABABBBA There is no disjoint decomposition with any bound set there is nondisjoint decomposition w 1th bound set E ECE 510 OC39E BDDS and Their Applications Relation Between Decomposition Types 0 Decomposition of completely specified functions is the simplest case 0 Decomposition of incompletely specified functions is closely related to nondisjoint decomposition o Decomposition of relations is the generalization of functional decomposition o Decomposition of multivalued relations is the most general decomposition considered so far May 16 2000 ECE 510 OC39E BDDS and Their Applications Implicit Support Minimization 0 An incompletely specified function over n variables Fri Jun B can be represented by an interval T 3 g in o The problem of support minimization is to find a function it jig depending on the minimum number of variables from 39 i39na such that 7 S quotf S Lquot 0 Lemma Variable can be removed from the l support of D May 16 2000 ECE 510 OCE BDDS and Their Applicaiions PseudoCode for Fin bdd FindSupportsbdd L bdd U ifL0 U1 returnl if L U 1 return 0 check cache for results xi is the top variable in L and U solve subproblems P0 FindSupports 39 L P1 FindSupports in Ugtlti M i 39ts Lgtlti Ugtlti bdd P xi P1 P0 insert into cache return P May 16 2000 ECE 510 OCE BDDS and Their Applicaiions Source Code for F1 bdd FindSupports const bdd 39 const bdd U if 39 ddf v bddtrue return bddtrue if I 1139 bddtrue return bddfalse check cache for results int TopVarL 39 L intTopVarU TopVarL lt TopVarU TopVarL TopVar 39 39 quot TopVar I39uccr393xki39 39 0 US FindSupports L v gXVarsTopVar 1 I30 insert into cache return 3 May 16 2000 ECE 510 OCE BDDs and Their Applications Lecture 11 FSM Equivalence Checking and FSM State Minimization May 2 2000 Alan Mishchenko Overview Equivalence for FSMs and FSM states Product machine PM Solving the problem of FSM equivalence Derive transition and output relation of the PM Perform reachability on the PM and verify property Output 1 Generate an error trace if the equivalence check has failed Solving the problem of FSM state minimization Derive transition and output relation of the PM Perform reachability on the PM and derive the state equivalence relation Transform the initial FSM s transition and output relations Compatible projection operator May 2 2000 ECE 510 OCE BDDS and Their Applicaiions FSM Equivalence Jim Two state machines are equivalent if starting from their reset states for any sequence of input vectors they produce identical sequences of output vectors ECE 5 10 OCE BDDS and Their Applicaiions FSM State Equivalence Two states 31 and of an FSM are equivalent if for any sequence of input vectors the FSM starting from state produces the same sequence of output vectors as the FSM starting from state ECE 5 10 OCE BDDS and Their Applicaiions Product Machine PM 0 Given FEW I O 5 with llt inputs n states an output the product machine is 1 51 5 28 the product machine has 4 inputs 2n states l output M ay 2 2000 quot E BDDS and Their Applications Deriving Representation of PM 0 Given transition relations and output functions of component machines T1 isln1 tZk is1 and T2 isZn2 k1k is2 those of the PM can be computed as follows A PMA r 1 n where is 51 and 52 are th sets of the current state variables for the component machines and sis the union of these sets similar for nfi May 2 2000 ECE 510 OCE BDDS and Their Applicaiions FSM Equivalence Checking 0 Find the transition relations and output functions of M1 and M2 Find the transition relation and output function of the PM 0 Perform reachability for the PM while checking its output 0 If the output of the product machine is 1 for all reachable states M1 and M2 are equivalent otherwise generate an error trace It is possible to define equivalence relative to any subset of inputs and outputs of the FSM May 2 2000 ECE 510 OCE BDDS and Their Applicaiions Equivalence Checking Formulas Property expresses equivalence of WI and 3amp2 in states 51 and 52 which constitute state of PM r l s the set of reachable states of x is the property that expresses equivalence of and M2 in the product state 0 Alternatively and IV are not equivalent iff itquot quotl x39a 39 n Reachability Analysis Procedure bool Veri Pro ert Usin Reachabili Anal sis FSM pM bdd Property bdd InitState FinddeCube 0 pM gtNBits CSVars 0 bdd Reached InitState From InitState NewMAX1TERNUM int NIter 0 do bdd To bddappexpM gtTransRelFrombddopandAllCSVars To bddreplace To pNS4CS New NIter To Reached bdd Check New NIter gtgt Property if Check bddtrue return false From New NIter Reached Reached New NIter while New NIter bddfalse return true guy 00 ECE 510 OCE BDDS and Their Applications FSM State Minimization Find the transition relations and the output functions of M Find the transition relation and output function of the PM created by two identical instances of M Compute the state equivalence relation describing the sets of all equivalent state pairs Compute the equivalence class characterization relation by selecting a representative state from each class of equivalence states Compute the transformed transition relation and the transformed output relation May 2 2000 ECE 510 OCE BDDS and Product Machine PM 0 Given FEM I O 5 2v with llt inputs n states m output the product machine is 1 ill SXS 28 the product machine has 4 inputs 2n states 1 output May 2 2000 ECE 510 OCE BDDS and Their Applicaiions EquivalenceDistinquishability Relations State equivalence relation is a boolean function x which is true for codes and iff the corresponding states are equivalent State distinquishabilitv relation is a boolean function which is true for codes and iff the corresponding states are not equivalent May 2 2000 ECE 510 OCE BDDS and Their Applicaiions Properties of Equivalence Relation o Equivalence relation is re up ECE 510 OCE BDDS and Their Applica ons Computing Equivalence Relation 212i can be computed using the ng procedure iterated until where J p M and is the variable replacement operator ECE 510 OCE BDDS and Their Applica ons Computing Distinquishability Relation can be computed using the following procedure iterated until May 2 2000 ECE 510 OCE BDDS and Their Applicaiions o The equival tion is derived as follows a l 39 quot 21 lity relation r f 2 May 2 2000 ECE 510 OCE BDDS and Their Applicaiions Equivalence Class Characterization Relation selects exactly one representative from each equivalence class of states defined by as 52 nal 32 is a boolean function that is one for the codes 51 and 52 iff the state corresponding to 51 represents the state corresponding to 52 sl 2 2 2ij titinl E031 32 x0 where x0 is the reset state expressed using variables 51 2 2000 ECE 510 OCE BDDS and Their Applicaiions Reduced State Set Transition and Output Relations of StateMinimum FSM o The state set of the minimumstate FSM I 92 34511 52 Transition relation of the stateminimum FSM Tminarslini i 17quot TKLSZFHJ 39 51252 Eml zi Output relation of the stateminimum FSM 39 l Ominis1o 2 mis t EJljsjfsz r ECE 5 10 OCE BDDS and Their Applicaiions

### BOOM! Enjoy Your Free Notes!

We've added these Notes to your profile, click here to view them now.

### You're already Subscribed!

Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'

## Why people love StudySoup

#### "There's no way I would have passed my Organic Chemistry class this semester without the notes and study guides I got from StudySoup."

#### "Selling my MCAT study guides and notes has been a great source of side revenue while I'm in school. Some months I'm making over $500! Plus, it makes me happy knowing that I'm helping future med students with their MCAT."

#### "I was shooting for a perfect 4.0 GPA this semester. Having StudySoup as a study aid was critical to helping me achieve my goal...and I nailed it!"

#### "It's a great way for students to improve their educational experience and it seemed like a product that everybody wants, so all the people participating are winning."

### Refund Policy

#### STUDYSOUP CANCELLATION POLICY

All subscriptions to StudySoup are paid in full at the time of subscribing. To change your credit card information or to cancel your subscription, go to "Edit Settings". All credit card information will be available there. If you should decide to cancel your subscription, it will continue to be valid until the next payment period, as all payments for the current period were made in advance. For special circumstances, please email support@studysoup.com

#### STUDYSOUP REFUND POLICY

StudySoup has more than 1 million course-specific study resources to help students study smarter. If you’re having trouble finding what you’re looking for, our customer support team can help you find what you need! Feel free to contact them here: support@studysoup.com

Recurring Subscriptions: If you have canceled your recurring subscription on the day of renewal and have not downloaded any documents, you may request a refund by submitting an email to support@studysoup.com

Satisfaction Guarantee: If you’re not satisfied with your subscription, you can contact us for further help. Contact must be made within 3 business days of your subscription purchase and your refund request will be subject for review.

Please Note: Refunds can never be provided more than 30 days after the initial purchase date regardless of your activity on the site.