SESSION 20 Sensors and Actuators for Health and Autonomy.pdf

1、ISSCC 2025SESSION 20Sensors and Actuators for Health and Autonomy20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference1 of 25A 3.5x3.5mm21.47mW/Ch 16-Channel MSS-CMOS HeterogeneousMulti-Modal Gas Sensor

2、 Chip StackKotaro Naruse1,Naru Kato1,Takuma Matsumori1,Jun Shiomi1,Yoshihiro Midoh1,Tetsuya Hirose1,Gaku Imamura1,2,Genki Yoshikawa2,3,Constantine Sideris4,Noriyuki Miura11Osaka University,Suita,Japan2National Institute for Materials Science,Tsukuba,Japan3University of Tsukuba,Tsukuba,Japan4Universi

3、ty of Southern California,Los Angeles,CA20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference2 of 25BackgroundIntegrated with Wearable DevicesIngestiblePillsSmart TagsFreshnessManagementIn-BodyDiagnosis

4、Breath Analysis Multi-modal gas sensors enable a wide array of applicationsMiniaturizing and reducing power consumption are required for daily life useMEMS-based gas sensors provide both low power and high sensitivity20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Senso

5、r Chip Stack 2025 IEEE International Solid-State Circuits Conference3 of 25Membrane-Type Surface-Stress Sensor(MSS)Gas MoleculesPiezoresistorsStressMSSWheatstone Bridge PiezoresistorsStressStressR1R2MembraneHeavily DopedPiezoresistorReceptor Detect gas from the stress transmitted to piezoresistors*G

6、.Yoshikawa et al.,Sensors,Nov.2012,pp.15873-15887xyOpposite Polarity20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference4 of 25Main Contributions:Maximize MSS Potential Miniaturized implementation by s

7、tacking MSS on CMOS readoutConventional implementation is large due to 2D wiring of MSS and readout Low power with no amp designConventional implementation is high power due to readout by universal ampMSS(0.5mm/Ch)Readout10mmConventionalImplementationhttps:/mss- 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-C

8、MOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference5 of 25Main Contributions:Maximize MSS Potential Miniaturized implementation by stacking MSS on CMOS readoutConventional implementation is large due to 2D wiring of MSS and readout Low power wi

9、th no amp designConventional implementation is high power due to readout by universal ampMSS(0.5mm/ch)Readout10mmConventionalImplementationhttps:/mss- 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference6 of

10、25MSS-CMOS Multi-Modal Gas Sensor Chip StackSOI SubstrateGas Flow Cavity0.5mm/ChCMOSReadoutMicro BumpMSS ArrayReceptorUltrasonic Bonding Minimized implementation size,low-noise readout of MSS output DC operation and no vibration in MSS,so high density possible20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel M

11、SS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference7 of 25Main Contributions:Maximize MSS Potential Miniaturized implementation by stacking MSS on CMOS readoutConventional implementation is large due to 2D wiring of MSS and readout Low powe

12、r with no amp designConventional implementation is high power due to readout by universal ampMSS(0.5mm/ch)Readout10mmConventionalImplementationhttps:/mss- 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference8

13、 of 25CMOS Readout Circuit Specialized for MSSDV/I ConverterfSENSfREFIREFVDDSBGRISENSVPMSSVIVOVNCMOSChannel1Shared AcrossAll Channels20-bit Counter20-bit CounterIBIASrf(=fSENS/fREF)ConstantCurrentGeneratorOSCx2480 A10 A Differential MSS output for common-mode noise/offset cancelling High sensitivity

14、 with noise suppression by frequency ratio rf20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference9 of 2510 ACMOS Readout Circuit Specialized for MSSfSENSfREFIREFVDDSBGRISENSChannel1Shared AcrossAll Cha

15、nnels20-bit Counter20-bit CounterIBIASrf(=fSENS/fREF)ConstantCurrentGeneratorOSCx2 Differential MSS output for common-mode noise/offset cancelling High sensitivity with noise suppression by frequency ratio rfDV/I ConverterCMOSVPMSSVIVOVN480 A20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogene

16、ous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference10 of 25480 ACMOS Readout Circuit Specialized for MSSfSENSfREFIREFVDDSBGRShared AcrossAll Channels20-bit Counter20-bit Counterrf(=fSENS/fREF)ConstantCurrentGeneratorOSCx2 Differential MSS output for common-m

17、ode noise/offset cancelling High sensitivity with noise suppression by frequency ratio rfCMOSVPMSSVIVOVNISENSChannel1DV/I ConverterIBIAS10 A20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference11 of 25I

18、BIAS10 A480 ACMOS Readout Circuit Specialized for MSSVDDSBGRShared AcrossAll ChannelsConstantCurrentGenerator Differential MSS output for common-mode noise/offset cancelling High sensitivity with noise suppression by frequency ratio rfCMOSDV/I ConverterVPMSSVIVOVNISENSIREFOSCx220-bit Counter20-bit C

19、ounterfSENSfREFrf(=fSENS/fREF)Channel120.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference12 of 25CMOS Readout Circuit Specialized for MSSDV/I ConverterfSENSfREFIREFVDDSBGRISENSVPMSSVIVOVNCMOSChannel1S

20、hared AcrossAll Channels20-bit Counter20-bit CounterIBIASrf(=fSENS/fREF)ConstantCurrentGeneratorOSCx2480 A10 A Differential MSS output for common-mode noise/offset cancelling High sensitivity with noise suppression by frequency ratio rfMSSCounterOSC60%DV/ILow Powerw/o Amp and ADCTotal:1.47mW/Ch20.1:

21、A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference13 of 25Implemented 4x4 Channels Gas Sensor3.5mm4x4 Channels CMOS Readout Array0.5mmVPVNDV/I ConverterCounterChannelVOVIOscillator0.9mmMSS Transducer Arra

22、y0.3mmMSS-CMOS Multi-Modal Gas Sensor Chip StackReceptors180nm Process1245Ch:36 Scalable to any number of channelsR4R3R1R24x4 Channels AluminumPatternIonImplantation20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Cir

23、cuits Conference14 of 25-2002040801.71.81.9Supply Voltage VDD VariationTemperature T Variation0.10.010.0010.00010.10.010.0010.0001T CVDDVReadout Circuit Performance Evaluation(1/2)Typical Condition:T=20,VDD=1.8V,VCM=800mV,VOFFSET=0VVP=VCM+V/2+VOFFSET,VN=VCM-V/2 Stable response to temperature and sup

24、ply voltage variationsDue to the readout by taking a ratio of similarly varying frequencies Frequency Ratio Shift,rfDifferential Voltage,V mV10.11010.11020.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Confer

25、ence15 of 25Readout Circuit Performance Evaluation(2/2)400500600700800900-50-2502550VOFFSETVariationVCMVariationDifferential Voltage,V mV0.10.010.0010.00010.10.010.0010.000110.11010.110VCM mVVOFFSET mVFrequency Ratio Shift,rfTypical Condition:T=20,VDD=1.8V,VCM=800mV,VOFFSET=0VVP=VCM+V/2+VOFFSET,VN=V

26、CM-V/2 Stable to common-mode voltage/offset voltage variationsDue to the differential output from the MSS 20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference16 of 25Setup for Gas Sensing TestGas Flow

27、ControllerGas SensorFPGAChamber20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference17 of 25Sensor Response to Ammonia(NH3)10ms/1sSampling3.33.313.323.333.123.133.143.15Supply NoisefSENSMHzfREFMHz500ppm

28、 NH3Noise Canceling by fREFN20.9450.9470.9490.9510.953050100150200Timesrf=fSENS/fREF500ppm NH3050100150200Times Suppress common-mode supply noise by frequency ratio rfCV 50%20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-S

29、tate Circuits Conference18 of 25Sensitivity for NH3DetectionFrequency Ratio Shift,rf0.0010.00010.01Concentration of NH3ppm3(Noise)10001001010.1T=25CAverage for 50s NH3Introduction3200rf TimesNH3Detection1000ppm0.9580.9620.9660.9708001600240050025010050Baseline(100%N2)Times rfx10-43=0.000279-60601002

30、00300 Limit of Detection(LoD)for NH3determined to be 0.1ppmLoD20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference19 of 25Temperature CBaseline rf-20020406080Before CompensationAfter Compensation1.210.

31、80.61/100 dependenceCompensation for Temperature Variations Compensated by parallel operating temperature sensor Using MSS-independent fREFlinear temperature dependenceTemperature CfREFMHz-20020406080Temperature SensingCapability3.323.283.263.33.2420.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Hete

32、rogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference20 of 25Multi-Modal Sensing of Several Gases(1/2)Different receptor channels have different unique responsesTolueneEthanolAcetaldehydeDimethyl disulfiden-heptane060012001800 06001200180000.010.02060012

33、00180000.010.02Ch2(P4MS)Ch5(P4MS)Frequency Ratio Shift,rfTimesCh1(PPPO)Same Receptor,Same ResponseCh4(PPPO)Ch3(CAB)Ch6(CAB)N2TargetN220.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference21 of 25Multi-Mo

34、dal Sensing of Several Gases(2/2)Different gases have different receptor responsesBy using these different responses,multi-dimensional analysis is possibleTimes00.010.02TolueneFrequency Ratio Shift,rf00.0040.008060012001800Ethanol060012001800Frequency Ratio Shift,rfTimesCh1(PPPO)Ch2(P4MS)Ch3(CAB)N2N

35、2N2N220.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference22 of 25Multi-Modal Sensing of Scent Samples(1/2)Different receptor channels have different unique responses00.010.020.0300.010.020.030306090120

36、 150180210 240270300330 360Frequency Ratio Shift,rfCh1Ch2TimesN2/Sample Switch Every 1minrepeatableCinnamonNutmegOreganoRosemaryGarlicLemonPeppermintAtlas cedar20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits

37、 Conference23 of 25Multi-Modal Sensing of Scent Samples(2/2)Successfully clustered smell samples by PCAExtracted features:response intensity,intensity ratio,decay slope-2-1012-4-20246PC2:13.1%PC1:81.7%CinnamonLemonPeppermintNutmegAtlascedarRosemaryGarlicOregano20.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel

38、MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference24 of 25Comparison Among ISSCC SOTA Gas SensorsTransducer OutputSensor Area mm2ChannelSensor TypeSupply Voltage VPower mW/ch(100%Active)Measured GasLoD ppmChip Size mm2/chProcess nm180This

39、 WorkMSSDC DifferentialVoltage3.5 x 3.5161.81.470.770.1(NH3)6 Chemicals,8 Scent Samples100ISSCC202SH-SAWOscillation2.4 x 2.061.89.00.800.02Toluene90ISSCC147ResistivePolymerDC Single-EndVoltage3.3 x 3.280.5Not Disclosed1.31Not DisclosedBreath SampleReadout Circuit Areamm2/ch0.0459Not DisclosedNot Dis

40、closed250ISSCC121Beam ResonatorOscillationNot Disclosed13.31.35Not Disclosed7.6Ethanol0.072620.1:A 3.5x3.5mm2 1.47mW/Ch 16-Channel MSS-CMOS Heterogeneous Multi-Modal Gas Sensor Chip Stack 2025 IEEE International Solid-State Circuits Conference25 of 25Conclusion MSS-CMOS Multi-Modal Gas Sensor Chip S

41、tackMiniaturized implementation by stacking MSS on CMOSLow power and high sensitivity with MSS specialized readoutScalable design for easy extension of sensor arrayDemonstrate multi-modal sensing with different receptors Future ProspectExpand sensor array for real-time breath analysis 20.2:A 67WChan

42、nel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference1/4320.2 A 67W/Channel,0.13nW/synapse/b Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning Dexuan Huo

43、*1,Yu-Hsien Lin*2,P K Shihabudeen2,Jilin Zhang1,Tao Li 1,Chi-Rong Chou2,Zhihua Wang1,Kea-Tiong Tang2,Hong Chen1*Equally Credited Authors(ECAs)1Tsinghua University,Beijing,China 2National Tsing Hua University,Hsinchu,Taiwan20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of

44、Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference2/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement Results Conclusions20.2:A 67WChannel,0.13nWsynapseb N

45、ose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference3/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement Results C

46、onclusions20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference4/43Background The framework of E-nose system The application of E-nose systemGas recognition algorithm:Food safet

47、yHealth carePublic safetyEnvironment monitor-Artificial neural network(BPNN,SNN,)-Statistical Method(SVM,KNN,PCA,)20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference5/432023 S

48、ymposium on VLSI Technology and CircuitsBackgroundGas recognitionDigital/AnalogKey FeaturesMethodProcessPowerISSCC20141KNN180nm1.05mWBioCAS20162CRBM90nm1.68mWTCASII20183KNN180nm2.60mWTCASI20184SNN65nm0.5uWNat.Mach.20205SNN14nm156uWVLSI20236SNN28nm40uW1 K.-T.Tang et al.,24.5 A 0.5 V 1.27 mW nose-on-a

49、-chip for rapid diagnosis of ventilator-associated pneumonia,ISSCC 2014.2 T.-I.Chou et al.,Design of a 0.5 v 1.68 mW nose-on-a-chip for rapid screen of chronic obstructive pulmonary disease,BioCAS,2016.3 T.-I.Chou et al.,A 1-V 2.6-mW Environmental Compensated Fully Integrated Nose-on-a-Chip,TCAS II,

50、2018.4 P.-C.Huang and J.M.Rabaey,A bio-inspired analog gas sensing front end,TCAS-I,2017.5 N.Imam and T.A.Cleland,Rapid online learning and robust recall in a neuromorphic olfactory circuit,Nature Machine Intelligence,2020.6 D.Huo et al.,ANP-G:A 28nm 1.04pJ/SOP Sub-mm2 Asynchronous Olfactory Process

51、or Enabling Few-shot Class-incremental On-chip Learning,VLSI 2023.SNN vs.Other methods Spiking Neural Network in E-noseLow power consumption:Binary spike local in space and time20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 20

52、25 IEEE International Solid-State Circuits Conference6/432023 Symposium on VLSI Technology and CircuitsMotivation Challenge 1:Different conditions lead to serious accuracy decrease E-nose in exhaled breath disease diagnosis20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of

53、 Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference7/432023 Symposium on VLSI Technology and CircuitsMotivation Solution 1:On-chip incremental learningSlide 7Retraining the network rapidly and locally to overcome accuracy reduction 20.2:A 67WChannel,0.

54、13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference8/432023 Symposium on VLSI Technology and CircuitsMotivation Challenge 2:Power consumption of portable E-nosesSlide 82 K.-T.Tang et al.,A low-po

55、wer electronic nose signal-processing chip for a portable artificial olfaction system,TBioCAS 2011.4 T.-I.Chou et al.,A 1-V 2.6-mW Environmental Compensated Fully Integrated Nose-on-a-Chip,TCAS II,2018.6 D.Huo et al.,ANP-G:A 28nm 1.04pJ/SOP Sub-mm2 Asynchronous Olfactory Processor Enabling Few-shot

56、Class-incremental On-chip Learning,VLSI 2023.8 N.Imam and T.A.Cleland,Rapid online learning and robust recall in a neuromorphic olfactory circuit,Nature Machine Intelligence,2020.20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning

57、2025 IEEE International Solid-State Circuits Conference9/432023 Symposium on VLSI Technology and CircuitsMotivation Solution 2:ADC-free Front-end circuitSlide 9ADC-free,1bit/channel20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learnin

58、g 2025 IEEE International Solid-State Circuits Conference10/432023 Symposium on VLSI Technology and CircuitsMotivationSlide 10 Solution 3:Clock-free design20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International

59、Solid-State Circuits Conference11/432023 Symposium on VLSI Technology and CircuitsMotivation Solution 3:Clock-free designSlide 11Event-driven,asynchronous logic20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE Internati

60、onal Solid-State Circuits Conference12/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement Results Conclusions20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-

61、chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference13/432023 Symposium on VLSI Technology and CircuitsGas sensor with sensing materialSlide 13On-chip sensor100um100um20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incr

62、emental Learning 2025 IEEE International Solid-State Circuits Conference14/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement Results Conclusions20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-inv

63、asive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference15/432023 Symposium on VLSI Technology and CircuitsADC-free Front-end circuitSlide 15 ADC-free design 16-channel 1 bit pulse signal output20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Ch

64、ip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference16/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement results Conclusions

65、20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference17/432023 Symposium on VLSI Technology and CircuitsAsynchronous olfactory processor-Algorithm Bio-inspired network Two-layer

66、 SNN Local unsupervised STDP Lateral inhibition Competition between glomerulus WTA(Winner-take-all)effect20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference18/432023 Symposium

67、 on VLSI Technology and Circuits-Exponential STDPSlide 18 Learning Rules:Classifier:-Metric learning-instead of output layer-Hamming distance Asynchronous olfactory processor-Algorithm Similarity=LTPLTDtwLTP20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with O

68、n-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference19/432023 Symposium on VLSI Technology and Circuits-Five-shot learning schemeSlide 19 Results:FTs:Flammable and toxic gas dataset 9-class 24-channel 225 samples 5 samples/concentration levelAsynchronous olfactory proc

69、essor-Algorithm 20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference20/432023 Symposium on VLSI Technology and CircuitsAsynchronous olfactory processor-HardwareExc.Timestamp Ge

70、n.Spike GeneratorExcitatory IF neuron576AERExc.Syn.Weight MemoryExc.Wgt.UpdateInhibitory LIF neuronInh.Spike History Buf.Excitatory Layer(EL)Inhibitory Layer(IL)Exc.Spike History BufferExc.Syn.Addr.GeneratorSkip Ctrl.EncoderDecoderLeaky Cur.LUTInh.Syn.Addr.Generator Inh.Syn.Wgt.Mem.Skip Ctrl.Inh.Wgt

71、.UpdateAsynchronous Olfactory Processor Pulse16Pulse01Pulse03Pulse02ResultPreprocessing ModuleN-class Readout Circuit20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference21/4320

72、23 Symposium on VLSI Technology and CircuitsSlide 21Click-based asynchronous pipeline Multi-stage pipeline chainTwo-phase handshake protocolReq_inAck_inAck_outReq_outfireAsynchronous olfactory processor-Hardware20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases wi

73、th On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference22/432023 Symposium on VLSI Technology and CircuitsSlide 22 Asynchronous LIF neuron:Asynchronous event-driven feature：spike=event Leaky current calculation Membrane potential cal.Determine the gen.of spikeLeakage

74、current calculation occurs only when the spike arrives!Asynchronous olfactory processor-Hardware20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference23/432023 Symposium on VLSI

75、Technology and CircuitsHardware overviewSlide 23Fire1Fire2Control path delayData path delayT(Data path delay)T(Control path delay)Ensure that data processing is completed before the Fire signal arrives!20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chi

76、p Incremental Learning 2025 IEEE International Solid-State Circuits Conference24/432023 Symposium on VLSI Technology and Circuits Self-adaptive synapse weight update skipping(SWUS)mechanism Decoupling for synapses Update weights in parallelAsynchronous olfactory processor-SWUS Skip unnecessary weigh

77、t update process Without changing any synapse weights20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference25/432023 Symposium on VLSI Technology and CircuitsSlide 25 LIF neuron

78、spikes The spike is an event Self-adaptive synapse weight update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conferenc

79、e26/432023 Symposium on VLSI Technology and CircuitsSlide 26 Record the spike time Distribute the synaptic address Transmit the spike information Self-adaptive synapse weight update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-in

80、vasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference27/432023 Symposium on VLSI Technology and CircuitsSlide 27 Update weights according to the spike time information Multi-segment LUTs reduces traversal cost Self-adaptive synapse wei

81、ght update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference28/432023 Symposium on VLSI Technology and CircuitsSl

82、ide 28 Check if the update value changes.If no change,then generate enable signal If changes,then go back the step(3)Self-adaptive synapse weight update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases

83、with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference29/432023 Symposium on VLSI Technology and CircuitsSlide 29 In the N consecutive events(spikes),no update weight change for this synapse.Then,the synaptic weight is no need for updating!Self-adaptive synapse wei

84、ght update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference30/432023 Symposium on VLSI Technology and CircuitsSl

85、ide 30.()().()(k).()(m)N=3N=3N=3SilentSilentSilent11111 Self-adaptive synapse weight update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS Setting skipping threshold N20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learnin

86、g 2025 IEEE International Solid-State Circuits Conference31/432023 Symposium on VLSI Technology and CircuitsSlide 31 Without accuracy degradation Saving unnecessary training time Skipping rate is 82%!Self-adaptive synapse weight update skipping(SWUS)mechanismAsynchronous olfactory processor-SWUS20.2

87、:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference32/432023 Symposium on VLSI Technology and CircuitsSlide 32 Trained synapse weight low-width storage(TWLS)methodAsynchronous olf

88、actory processor-TWLSGlom.1Glom.2Glom.16Lateral inhibition-w-w-w-w Over 95%trained synapse weights reach the minimum or maximum values20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits

89、Conference33/432023 Symposium on VLSI Technology and CircuitsSlide 33 Trained synapse weight low-width storage(TWLS)methodAsynchronous olfactory processor-TWLSTraining power(mW)NoneWith TWLSWith SWUSFull-load sim.result on FPGA-37.2%10Both5-48.8%-72.3%20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip

90、for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference34/43Outline Background and motivation Gas sensor array with sensing material ADC-free front-end circuit Asynchronous olfactory processor Measurement results Conclusions20.

91、2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference35/432023 Symposium on VLSI Technology and CircuitsMeasurement Results-BenchmarkSlide 35 Error rate of the output pulse concern

92、ing sensor resistance variation 20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference36/432023 Symposium on VLSI Technology and CircuitsMeasurement Results-BenchmarkSlide 3620.2

93、:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference37/432023 Symposium on VLSI Technology and CircuitsMeasurement Results-BenchmarkSlide 3792 38#Positive samples#Negative samples#

94、Training samples#Test samples104 130 On-chip learning 97.7%accuracy 98.9%TPR 12.5%accuracy improvement20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference38/432023 Symposium on

95、 VLSI Technology and CircuitsMeasurement Results-ComparisonSlide 3820.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference39/432023 Symposium on VLSI Technology and CircuitsMeasur

96、ement Results-ComparisonSlide 3920.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference40/432023 Symposium on VLSI Technology and CircuitsMeasurement Results SummarySlide 4016-cha

97、nnel gas sensor array and analog front-end circuit20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference41/432023 Symposium on VLSI Technology and CircuitsMeasurement Results Sum

98、marySlide 41Asynchronous olfactory processor20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference42/432023 Symposium on VLSI Technology and CircuitsMeasurement Results SummarySl

99、ide 42 SIP Package20.2:A 67WChannel,0.13nWsynapseb Nose-on-a-Chip for Non-invasive Diagnosis of Diseases with On-chip Incremental Learning 2025 IEEE International Solid-State Circuits Conference43/432023 Symposium on VLSI Technology and CircuitsConclusionC16-2 An E-nose system with on-chip increment

100、al learning for noninvasive diagnosis of diseases A low-power system solution integrates ADC-free analog front-end circuit with a two-layer SNN asynchronous processor An asynchronous training method to reduce redundant synapse weight update and storage costSlide 4320.3:An RFID-inspired One-step Pack

101、aged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference1 of 61An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics Yan-Ting Hsiao1,Ya-Chen Tsai1,Wei Foo1

102、,Hung-Yu Hou1,Yun-Chun Su2,Yueting Lily Li1,Jun-Chau Chien11University of California,Berkeley,CA,2National Taiwan University,Taipei City,Taiwan 20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State

103、Circuits Conference2 of 61Outline Introduction System block diagram and circuit details Measurements results Packaging and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International So

104、lid-State Circuits Conference3 of 61Outline Introduction System block diagram and circuit details Measurements results Packaging and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE Intern

105、ational Solid-State Circuits Conference4 of 61The Need for Point-of-Care DiagnosticsMolecular Diagnostics Workflow at Centralized LaboratoriesSampleTransferComplex procedure yet highly accurate.Analyzer a panel of bio-markers simultaneously.Require sample transportation leading to delayed results.In

106、frastructure costs(controlled space,electricity needs)limits accessibility.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference5 of 61Successful Point-of-Care Diagnosticssource:

107、cdc.govsource: Patient-side diagnostics.Rapid results,e.g.48 hrs)Electrical connections(wirebonds needed);limited fluidic routing space.Our prior works20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid

108、-State Circuits Conference9 of 61Challenge#2:Sample Delivery and PreparationB.Shen,et al.,T-BioCAS,2022Plasma extraction(avoid cellular interference)Sample delivery20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE Intern

109、ational Solid-State Circuits Conference10 of 61Recent CMOS Microfluidics PlatformsComplex packaging flowWired powering and data transmissionCostly custom electrodesElectro-osmosis drivenC.Zhu,et.al.,T-BIOCAS,2021Complex packaging flowWired powering and data transmissionCost-efficient fabricationPump

110、-drivenQ.Liu,et.al.,ISSCC,2024Simplified packagingWired powering and data transmissionCostly custom electrodesMagnetic field-drivenD.Lee,et.al.,T-BIOCAS,202420.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International

111、 Solid-State Circuits Conference11 of 61Our Solution:(1)RFID-inspired Wireless IC Avoid wirebonds.Single-step packaging flow.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference

112、12 of 61Our Solution:(2)Vacuum Microfluidics Automates sample delivery and plasma extraction.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference13 of 61Our Solution:(3)Multi-mo

113、dal analysis Dual pH and current readouts for different signaling assays.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference14 of 61Put it all Together!One-step packaging flowW

114、ireless powering and data transmissionCost-efficient fabricationAir pressure-drivenAccurate,cost-efficient,and rapid home-use solution!20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits

115、Conference15 of 61Outline Introduction System block diagram and circuit details Measurements results Packaging and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-Stat

116、e Circuits Conference16 of 61System Block Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference17 of 61System Block Diagram2-turn on-chip coilIn M6Width=20 mGap=10 mL=18 n

117、HQmax=7.4700 MHz20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference18 of 61System Block Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Micro

118、fluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference19 of 61System Block Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference20

119、of 61System Block Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference21 of 61System Block Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vac

120、uum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference22 of 61System OperationWhen wireless power comes in,the BGR and DLDO automatically turns on,and POR signal makes system initialization.POR=Power-on ResetEOP=End-of-power20.3:An RFID-inspired One-s

121、tep Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference23 of 61System OperationIn DL mode,the ALDO and MLDO are activated,and system configuration parameters are transmitted to the FSM via PWM-ASK downlink.P

122、OR=Power-on ResetEOP=End-of-power20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference24 of 61System OperationIn sensing mode,a 2.5 MHz system clock is generated via downlink,ena

123、bling the system to operate in sync with the clock.POR=Power-on ResetEOP=End-of-power20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference25 of 61System OperationPOR=Power-on Res

124、etEOP=End-of-powerLSK=Load-Shift-KeyingThe Miller encoded data is then transmitted to the reader via LSK,with a on-chip 1.8MHz clock.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Co

125、nference26 of 61Power Management Rectifier and BGRRectifierBGR Generates a supply voltage(VRECT)greater than 2V.Provides many current bias and VBGRof 1.25V.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International

126、Solid-State Circuits Conference27 of 61Power Management LDO An additional feedforward cancellation path in the LDOs to create notch filtering at sensing frequency 2.5 MHz Isolating VRECTchange for clock DL in sensing mode.J.Guo et al.,A-SSCC,201320.3:An RFID-inspired One-step Packaged Multi-mode Bio

127、-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference28 of 61Data Transmission Downlink(DL)Provide programming bits for controlling chip in DL mode.We use a wireless 2.5 MHz clock instead of an on-chip clock for greater precision su

128、pporting amperometry and SWV sensing.DL CircuitPWM-ASK=Pulse Width Modulation-Amplitude-Shift-KeyingDATAmodulationDATAmodulationDATAmodulation20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Ci

129、rcuits Conference29 of 61UL data is Miller-coded using a clock from the on-chip 1.8MHz oscillator and data stored in memory.UL Data is read out by the external reader at a speed of 0.9 Mbps.Data Transmission Uplink(UL)Receive data!LSK=Load-Shift-Keying.20.3:An RFID-inspired One-step Packaged Multi-m

130、ode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference30 of 61Sensor Readout PotentiostatThe 10b DAC and amplifier A1 form a potentiostat,enabling electrochemical measurements.The rail-to-rail structure allows the RE to operat

131、e over a wider range.Buffer equalizes RE/CE voltagesA120.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference31 of 61Sensor Readout Amperometric SensorThe core amplifier A2 is curr

132、ent-reuse structure Better gm,low noise,but limited output range.Source-follower buffer is for CMFB maintain high RoutC-TIA senses the redox currentA220.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-

133、State Circuits Conference32 of 61Sensor Readout Amperometric SensorThe low leakage switch is to prevent the redox current from flowing through unwanted paths.VXis locked to Vin VSB 0,low leakage current when switch is off.Low leakage reset switches20.3:An RFID-inspired One-step Packaged Multi-mode B

134、io-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference33 of 61To prevent output saturation,we set a threshold voltage VTH.When the output exceeds this value,it will be reset.Sensor Readout Auxiliary ResetY.Chen et al.,TBCAS,201320.

135、3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference34 of 61Sensor Readout pH&Temperature SensorTemp.Sensor10b SAR ADCIPTAT is from BGR.SNDR/SFDR(dB):58.0/68.3ENOB(bit):9.34 DNL(LS

136、B):+0.74/-0.98INL (LSB):+0.77/-0.82pH SensorAl2O3membrane pH sensor20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference35 of 61Outline Introduction System block diagram and circ

137、uit details Measurements results Packaging and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference36 of 61Die PhotoAfter ENIG PlatingGold Surface

138、 Electrodefor Amperometry SensingTSMC 180nm CMOS*ENIG:electroless nickel immersion gold20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference37 of 61Initialization Operation Resul

139、tAfter powering on,the system enters the initialization state via the POR signal.System Operation Diagram20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference38 of 61DL&UL Operat

140、ion ResultIf DL data transmission is correct,the rising edge is detected.An off-chip reader can clearly detect UL data in the range of a few mV.Down-Link Data TransmissionUp-Link Data TransmissionEOPDATAmodulation20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidi

141、cs for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference39 of 61Temperature SensorTemperature benchmarked against the NI USB-TC01.Expected higher error due to air convection in open space environments.Heated by a resistive heater.20.3:An RFID-inspired One-step Package

142、d Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference40 of 61Amperometric Sensor Auxiliary ResetOriginal dynamic range(DR)is 72.0 dB.With signal-folding circuit DR is 129.8 dB(when RST is re-timed with 10 sec)Signal-

143、folding functionThe max.current is constrained by the time(Here is 10 sec).20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference41 of 610.24 pArms4.27 pArms54.15 pArms Amperometr

144、ic Sensor NoiseWith a 5-nF electrode capacitance,the amperometry front-end circuit achieves 0.24 pArms noise at 100-Hz filter.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conferenc

145、e42 of 61Amperometric Sensor HRP/TMB Measurements DemoHRP/TMB binding releases electronsOur system showed higher sensitivity than commercial potentiostatInvitrogen Cat.KHA003120.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025

146、IEEE International Solid-State Circuits Conference43 of 61pH Sensor Standard SolutionThe Al2O3membrane is more sensitive than CVD SiO2.Sensitivity of this work is 91.7 mV/pH from pH 4 to 6 and 172 mV/pH from pH 6 to 9.20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microf

147、luidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference44 of 61pH Sensor Glucose/Avidin-GOx Measurement DemoGlucose and Avidin-conjugated GOx binding releases protons,altering pHOur system showed higher sensitivity than commercial potentiostatFluctuation comes f

148、rom the thin Al2O3layer 20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference45 of 61Outline Introduction System block diagram and circuit details Measurements results Packaging

149、and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference46 of 61Assembly20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuu

150、m Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference47 of 61Assembly20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference48 of 61

151、Vacuum Microfluidics20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference49 of 61Vacuum Microfluidics20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum M

152、icrofluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference50 of 61Vacuum Microfluidics20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conferenc

153、e51 of 61Use case20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference52 of 61Use case20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics fo

154、r Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference53 of 61Use case20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference54 of 61Use case20.3:An RF

155、ID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference55 of 61Use case20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics

156、 2025 IEEE International Solid-State Circuits Conference56 of 61Vacuum microfluidic video demo20 x speed upReagent 1Reagent 2BloodSensing AreaTrench filterSerpentine channelsReservoir20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnosti

157、cs 2025 IEEE International Solid-State Circuits Conference57 of 61Outline Introduction System block diagram and circuit details Measurements results Packaging and vacuum microfluidics Conclusion20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Car

158、e Diagnostics 2025 IEEE International Solid-State Circuits Conference58 of 61Comparison table20.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference59 of 61ConclusionsA CMOS/microf

159、luidics point-of-care device with near-field wireless power/comm.enables simple one-step packaging/assembly.We employ self-powered vacuum microfluidics for pumpless sample delivery.Demonstrates multi-modal bio-sensing targeting electrochemical ELISA and pH-based ELISA.20.3:An RFID-inspired One-step

160、Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference60 of 61Acknowledgement Taiwan Semiconductor Manufacturing Company(TSMC)National Science Foundation(NSF)2406340 National Institutes of Health(NIH)National I

161、nstitute of Biomedical Imaging and Bioengineering(NIBIB)Trailblazer Award R21EB03615420.3:An RFID-inspired One-step Packaged Multi-mode Bio-analyzer with Vacuum Microfluidics for Point-of-Care Diagnostics 2025 IEEE International Solid-State Circuits Conference61 of 61Thank you for the attention20.4:

162、MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference1 of 27MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular

163、-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor RegenerationMarco Saif*,Fuze Jiang*,Shao-Wei Lo,Adam Wang,Zhikai Huang,Jiachen Wang,Hangxing Liu,Chih-Jen Shih,Thomas Burger,Hua WangETH Zrich,Zrich,Switzerland*Equally credited authors(ECAs)20.4:MEMS-Fr

164、ee 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference2 of 27Outline Background and Motivation Proposed PlatformChip architecture CM

165、OS designMaterial and in-house processing Measurement resultsElectrical measurementsGas-measurements Conclusion20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE

166、International Solid-State Circuits Conference3 of 27BackgroundCC BY-NCAgriculture and Crop MonitoringCC BYMedicalDiagnosticsIndustrial SafetyCheapReusabilityHigh dynamic rangeHigh selectivityHigh sensitivityDense arrayLow powerPatrick J.Lynch,medical illustrator,CC BY 2.5,via Wikimedia CommonsSPI204

167、8:32 SerializerBias CircuitClk Buffers2048:32 SerializerBias Circuit45m x 45mNon-Overlapping CLKVREFx 4096VDD45m x 45mx 128Non-Overlapping CLK 2nd order SD ADCJouleHeaterMOF 1MOF 1MOF 1MOF 1MOF 1MOF 1MOF 1MOF 2MOF 2MOF 2MOF 2MOF 2MOF 2MOF 2Remote Sensing with Electronic NoseMobile Molecular Sensing

168、with High Sensitivity and SpecificityMOFs Modified E-noseReal-time MonitoringAccidents and Emergency ResponseGas sensing pixelTemperature sensing pixelSystem ArchitectureImproved Sensing Interface Massive sensor node In-pixel digitizationProposed designConventionalLess sensor node Rapid response Fas

169、t desorption Fully regenerationbaselineresponse Slow desorption Poor regenerationresidues MOF20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Soli

170、d-State Circuits Conference4 of 27MotivationHigh temperature operationCan be used in dense arrayHigh sensitivityHigh DRRoom temperature operationCan be used in dense arrayLow sensitivityHigh DRHigh temperature operationLarge areaHigh sensitivityLow DRF.Ciciotti,IEEE Sensors19T.Chou,TCAS II18Y.Lee,JS

171、SCC23MOXPolymer-carbonMEMS20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference5 of 27MotivationHigh temperature operat

172、ionCan be used in dense arrayHigh sensitivityHigh DRRoom temperature operationCan be used in dense arrayLow sensitivityHigh DRHigh temperature operationLarge areaHigh sensitivityLow DRF.Ciciotti,IEEE Sensors19T.Chou,TCAS II18Y.Lee,JSSCC23Room temperature operationCan be used in dense arrayHigh sensi

173、tivityHigh DRReusabilityMOXPolymer-carbonMEMSMOF20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference6 of 27Chip archit

174、ecture NH3CH3OHC2H5OHC3H6O2048:32 SerializerBias CircuitClk Buffers2048:32 SerializerBias CircuitMOF 1MOF 2SPIMOF 1MOF 1MOF 1MOF 1MOF 1MOF 1MOF 2MOF 2MOF 2MOF 2MOF 2MOF 2VctrlVHeater45m x 45mNon-Overlapping CLKVREFx 4096JouleHeater Gas sensing pixel45m x 45mx 128Non-Overlapping CLK 2nd order SD ADC

175、Temperature sensing pixelBJT-based temperature sensor20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference7 of 27Chip a

176、rchitecture NH3CH3OHC2H5OHC3H6O2048:32 SerializerBias CircuitClk Buffers2048:32 SerializerBias CircuitMOF 1MOF 2SPIMOF 1MOF 1MOF 1MOF 1MOF 1MOF 1MOF 2MOF 2MOF 2MOF 2MOF 2MOF 2VctrlVHeater45m x 45mNon-Overlapping CLKVREFx 4096JouleHeater Gas sensing pixel45m x 45mx 128Non-Overlapping CLK 2nd order SD

177、 ADC Temperature sensing pixelBJT-based temperature sensor20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference8 of 27G

178、as sensing pixelCK2CK1DCK1Timing Diagram VDDCLKCLKCLKCLKVIPVINCLKComparatorHeater OFFHeater ONVOUTVDDVINVIP2-stage amplifierCK1DCK2CK2CALCK1DCK2CK2CK1DCK2CK1CK1DCK2CK1CK2CK2CK1DCK2C/V ConveterVIPCsCVADCVREFVREFVREFVREFVREFVCMVCMVCMVADC=VIPCC+CsPassive C-to-V conversion through capacitive divider.20.

179、4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference9 of 27Gas sensing pixelCK2CK1DCK1Timing Diagram VDDCLKCLKCLKCLKVIPVI

180、NCLKComparatorHeater OFFHeater ONVOUTVDDVINVIP2-stage amplifierCK1DCK2CK2CALCK1DCK2CK2CK1DCK2CK1CK1DCK2CK1CK2CK2CK1DCK2C/V ConveterVIPCsCVADCVREFVREFVREFVREFVREFVCMVCMVCMVADC=VIPCC+CsPassive C-to-V conversion through capacitive divider.BP SD ADC to digitize the output.20.4:MEMS-Free 4096-Pixel CMOS

181、E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference10 of 27Gas sensing pixelCK2CK1DCK1Timing Diagram VDDCLKCLKCLKCLKVIPVINCLKComparatorHeater OFFHea

182、ter ONVOUTVDDVINVIP2-stage amplifierCK1DCK2CK2CALCK1DCK2CK2CK1DCK2CK1CK1DCK2CK1CK2CK2CK1DCK2C/V ConveterVIPCsCVADCVREFVREFVREFVREFVREFVCMVCMVCMVADC=VIPCC+CsPassive C-to-V conversion through capacitive divider.BP SD ADC to digitize the output.Class AB amplifier for large output swing.20.4:MEMS-Free 4

183、096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference11 of 27Temperature sensing pixel 1CK1DCK2CK1CK1CK2CK1DCK2CK2CK1VREFVCMVREFVCMHybr

184、id active-passive LP SDSD ADCTemperature sensorVDD71Start-up Circuit VIPLP SDSD ADC modela VREFCK1DCK2CK2CK1VREFBJT-based temperature sensor.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast

185、Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference12 of 27Temperature sensing pixel 1CK1DCK2CK1CK1CK2CK1DCK2CK2CK1VREFVCMVREFVCMHybrid active-passive LP SDSD ADCTemperature sensorVDD71Start-up Circuit VIPLP SDSD ADC modela VREFCK1DCK2CK2CK1VREFBJT-based temperature sensor.Hy

186、brid active-passive LP SD ADC for low power and area.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference13 of 27MOF d

187、epositionMOFs mappingSEMEDXMOFDepositionEtching the passivation layer to expose the electrodes.Deposit the MOF material on top of the electrodes.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for F

188、ast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference14 of 27MOF depositionMOFs mappingMultiple MOFs materials selective patterning on chipTwo MOFs patternSEMEDXCo EDX mappingNi EDX mappingMOFDepositionMOF-2MOF-1Etching the passivation layer to expose the electrodes.Deposit

189、 the MOF material on top of the electrodes.Multiple MOF materials on the same chip for multi-species gas detection.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 I

190、EEE International Solid-State Circuits Conference15 of 27Gas sensor pixel electrical results 5 4 4 384 PixelsBP ADC has maximum SNDR of 60.14 dB(OSR=850).The serializer allows reading multiple pixels together.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organ

191、ic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference16 of 27Gas sensor pixel electrical results 5 4 4 VDDCLKCLKCLKCLKVIPVINCLKCK1DCK2CK2CALCK1DCK2CK2CK1DCK2CK1CK1DCK2CK1CK2CK2CK1DCK2C/V ConveterVIP Heater OFFH

192、eater ONCsCVADCVREFVREFVREFVREFVREFOn-Chip HeaterOn-Chip Heater PerformanceComparatorVCMVCMVCMVctrlVHeater 5 4 4 26.2 oC42.5 oC384 PixelsBP ADC has maximum SNDR of 60.14 dB(OSR=850).The serializer allows reading multiple pixels together.The Heater can increase chip temperature to 42.5 C for sensor r

193、eusability.384 Pixels20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference17 of 27Temperature sensor pixel results25303

194、5404550550.7800.7850.7900.7950.800Voltage(V)Temperature(C)Temperature sensor results repeated 5 times.ADC spectrum shows 40 dB/decade.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor

195、Regeneration 2025 IEEE International Solid-State Circuits Conference18 of 27Gas measurements0 ppm NH320 ppm NH3C/CC/C1010Mapping results for before and after adding NH3.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermo

196、dynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference19 of 270801602403200.00.20.40.6 Heat-off Heat-on(40oC)D DC/CTime(s)ThermodynamicAdsorptionDesorption10 ppmBackgroundGas measurements0 ppm NH320 ppm NH3C/CC/C1010Mapping results for before and afte

197、r adding NH3.Sensor reusability using the heater.The sensor has faster desorption with heating.The sensor can reach baseline only with heating.Important for sensing low concentrations after exposure to high concentrations.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selectiv

198、e Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference20 of 27Gas measurements02468100.20.40.60.8 Heat-off Heat-on(40oC)D DC/CCyclesRegeneration 0 ppm NH320 ppm NH3C/CC/C1010Mapping results for befo

199、re and after adding NH3.Sensor is stable for 10 heating cycles.Heating the chip(with continuous NH3flow)allows the reading to approach baseline.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fa

200、st Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference21 of 270123456D DC/CGas typeSelectivityIPATHFEthanolMethanolAcetoneDMSOAmmonia173 ppmGas measurements0 ppm NH320 ppm NH3C/CC/C1010Mapping results for before and after adding NH3.High selectivity making it suitable for e-n

201、ose applications.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference22 of 27Gas measurements02468100.000.020.040.060.

202、080.10D DC/Cppm(NH3)Limit-of-DetectionLoD:5 ppm0.11010000481216Sensitivity and Linearityppm(NH3)D DC/CLinear regionHigh sensitivity of 0.012(C/C)/ppm.Limit of Detection of 5 ppm.Wide dynamic range from 5 ppm to 1000 ppm.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective

203、Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference23 of 27SOAa Electrode areaThis WorkChiuTBCAS 14 12 ChouTCAS II 18 13 LEEISSCC 20 14BurnsCICC 24 15FaruqeTBCAS 24 16Technology(nm)40901801006565Se

204、nsor surface modificationMetal-Organic-Framework(MOF)Polymer and CarbonPolymer and CarbonPolymerMolecularly Imprinted Polymers(MIP)Sol-gelMEMS requiredNoNoNoYesNoNoReadout arch.BP SDSD CDCR-SARR-SARSH-SAW Stiffness sensorWhB+PiplineSAR CDCNumber of pixels4096 CDC8364 Gas110128 Temp.Sensor1 Temp.1 Hu

205、midityPower/pixel(W)6.5127038597024.40.12Pixel area(mm2)0.0020250.16a0.032a0.220.540.0251LoD(ppm)5-10000.02-On-chip electrodesYesYesYesYesNoNoOn-chip temperature controlYesNoYesYesNoNo20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and

206、In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference24 of 27SOAa Electrode areaThis WorkChiuTBCAS 14 12 ChouTCAS II 18 13 LEEISSCC 20 14BurnsCICC 24 15FaruqeTBCAS 24 16Technology(nm)40901801006565Sensor surface modificationMetal-Organ

207、ic-Framework(MOF)Polymer and CarbonPolymer and CarbonPolymerMolecularly Imprinted Polymers(MIP)Sol-gelMEMS requiredNoNoNoYesNoNoReadout arch.BP SDSD CDCR-SARR-SARSH-SAW Stiffness sensorWhB+PiplineSAR CDCNumber of pixels4096 CDC8364 Gas110128 Temp.Sensor1 Temp.1 HumidityPower/pixel(W)6.5127038597024.

208、40.12Pixel area(mm2)0.0020250.16a0.032a0.220.540.0251LoD(ppm)5-10000.02-On-chip electrodesYesYesYesYesNoNoOn-chip temperature controlYesNoYesYesNoNo20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation fo

209、r Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference25 of 27ConclusionHigh density CMOS E-Nose with 4096 pixels.Molecular-Selective MOF Sensing for high selectivity.Low-power sensing of 6.5 W at room temperature.On-chip heater for reusability.High DR from 5 ppm to 1000

210、ppm.High sensitivity of 0.012 C/C)/ppm.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference26 of 27AcknowledgementThe

211、authors would like to thank Globalfoundries for designkit support and chip fabrication.The research is in part sponsored by the Swiss State Secretariat for Education,Research,and Innovation(SERI)under the SwissChips initiative,Swiss National Science Foundation for MICA Project under the grant number

212、 207914,and ETH internal grants.20.4:MEMS-Free 4096-Pixel CMOS E-Nose Gas Sensor Array with Molecular-Selective Metal-Organic-Framework Sensing and In-Pixel Thermodynamic Modulation for Fast Sensor Regeneration 2025 IEEE International Solid-State Circuits Conference27 of 27Thank you!Welcome to atten

213、d our demo(DS 2Tonight 5:00 7:00 PM)20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference1 of 33Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transisto

214、rs Based Electrochemical Sensing SoCYuan Ma1*,Shangbin Liu1*,Yahao Song1,Chao Xie1,Yuwei Zhang2,Chao Sun2,Xiaoyan Ma2,Lan Yin1and Milin Zhang11Tsinghua University,Beijing,ChinaBeijing Ningju Technology,Beijing,China20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transisto

215、rs Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference2 of 33OutlineIntroductionKey TechnologiesExperimental ResultsConclusion20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE Inte

216、rnational Solid-State Circuits Conference3 of 33IntroductionInflammatory cytokines crucial biomarkers of chronic diseases Autoimmune disorders Cardiovascular diseases CancerCommon analytes TNF-IL-6Alzheimers diseaseArthritisChronic WoundsLiver CancerLung CancerMyocarditis20.5 Millimeter-sized 0.1pM-

217、LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference4 of 33Detection of inflammatory cytokinesInflammatory cytokinesLong-termReal-timeCompact sizeAlzheimers diseaseArthritisChronic WoundsLiver CancerLung

218、 CancerMyocarditisElectrochemical Sensor20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference5 of 33Detection of inflammatory cytokinesInflammatory cytokinesLong-termReal-timeC

219、ompact sizeAlzheimers diseaseArthritisChronic WoundsLiver CancerLung CancerMyocarditisElectrochemical SensorChallenges for implantableHigh sensitivitymulti-channelsFully wireless&Low power20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Se

220、nsing SoC 2025 IEEE International Solid-State Circuits Conference6 of 33Existing Electrochemical sensingThree-electrode(3E)system The surface contact between the solution and the electrode limits the current sensitivity.WERECEe-target+-J.Chein et al.,ISSCC 202020.5 Millimeter-sized 0.1pM-LoD Wireles

221、s 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference7 of 33Existing Electrochemical sensingIon-Sensitive Field-Effect Transistor(ISFET)The variation of threshold voltage shifts with the concentration of H+ions in c

222、ontact with the extended gate of the ISFET.+-SGDH+H+H+H+H+Q.Lin et al ISSCC 2023 20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference8 of 33Existing Electrochemical sensingOrg

223、anic electrochemical transistor(OECT)Volume gating based on organic semiconductors,resulting in much higher transconductance.The high sensitivity of OECTs requires relatively high static current.GDS-+-+target+-0+-0PEDOT:PSSRivnay et al Nature 2018+-S+-CCHRSD,G20.5 Millimeter-sized 0.1pM-LoD Wireless

224、 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference9 of 33The proposed OECT-based System on chip Millimeter-sizedHigh sensitivityWide dynamic range16-channelFully wirelessLow powerRectifierBGRLDOPower ManagementVCO

225、PDLFFilterFIFOOOKDigiClk PLLOOK Mod.Analog Front-end CircuitsCh1Ch16ADCPGAPattern GenLSVCVCADACCC+R1/2/3IoffsetBias33/181Backscatter Control&20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State C

226、ircuits Conference10 of 33OutlineIntroductionKey TechnologiesThe proposed CC+R potentiostat with active control OECTThe Dual-mode CC+R/C analog frond-endThe active full-wave rectifier with backscatter amplitude modulationExperimental ResultsConclusion20.5 Millimeter-sized 0.1pM-LoD Wireless 16-chann

227、el Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference11 of 33Sensing principle of the OECTFix and OECT operate in the saturation regionTest=,22:capacitance per unit volume of the channel,:Effective gate voltage,=,+-S+-CCHRS=0

228、.6D,=0.2G20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference12 of 33Sensing principle of the OECT=,22,=,2.3(1+)2l logarithmic variation with the concentration of the target a

229、nalyteSelectivity:Specific aptamer on the gate DC Operating Point:Scan the transfer curve to optimize gm/idsTargetIncreasing+-S+-CCHRS=0.6D,=0.2G20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-Sta

230、te Circuits Conference13 of 33Potentiostat for OECTCurrent MirrorR/C-TIAThis work(CC+R)Static Current-+Active control+-+Noise-+Power-+-+OPA+-OTAMCGCfRfR/C-TIA,2+12,2+4-+RfOPACfThis work(CC+R)Current Mirror(CM),212,2+420.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transis

231、tors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference14 of 33Potentiostat for OECTCurrent MirrorR/C-TIAThis work(CC+R)Static Current-+Active control+-+Noise-+Power-+-+OPA+-OTAMCGCfRfR/C-TIA,2+12,2+4-+RfOPACfThis work(CC+R)Current Mirror(CM),212,2+420.5 Millim

232、eter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference15 of 33Potentiostat for OECTCurrent MirrorR/C-TIAThis work(CC+R)Static Current-+Active control+-+Noise-+Power-+-+OPA+-OTAMCGCfRfR/C-T

233、IA,2+12,2+4-+RfOPACfThis work(CC+R)Current Mirror(CM),212,2+420.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference16 of 33CC+R potentiostat with active controlSampling phaseMCG

234、turn on with 7 kbpsSampling rate 24Hz/ChRobustBit error rate(BER)15%Head Num.ch 1ch 2ch 3ch 4ToFToFcode=1code=01 Frame Data=16 bit*(Channel+2)20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State

235、Circuits Conference29 of 33Measurement results of TNF-,IL-6 20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference30 of 33State-of-the-art Comparison20.5 Millimeter-sized 0.1pM-

236、LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference31 of 33OutlineIntroductionKey TechnologiesThe proposed CC+R potentiostat with active control OECTThe Dual-mode CC+R/C analog frond-endThe active full-

237、wave rectifier with backscatter amplitude modulationExperimental ResultsConclusion20.5 Millimeter-sized 0.1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference32 of 33ConclusionDual-mode potentiostat

238、front-endA bidirectional CC+R potentiostat with active control reduces the duty-cycle of the OECT amplification process with static current.Reuse the PGAs sampling capacitor extended the dynamic range.Ultrasonic wireless power and backscatter wireless communicationAn active rectifier with backscatte

239、r amplitude modulation features a wide input range of received pulse amplitudes,tunable ultrasound frequencies,and variable time-of-flight(ToF).Electrochemical SoC4.3x4.1x4 mm3wireless implant Integrated 16-channel OECT Limit of detection(LoD):0.1 pMInput current range:184 dB20.5 Millimeter-sized 0.

240、1pM-LoD Wireless 16-channel Organic Electrochemical Transistors Based Electrochemical Sensing SoC 2025 IEEE International Solid-State Circuits Conference33 of 33Thank you for your attention!20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energ

241、y Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference1 of 76Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis ActuationRunqing Cai,Muhamm

242、ad Dilawar Khan Niazi,Yucheng Ai,Jin Zhu,Linsheng Wu,Xuyang LuShanghai Jiao Tong University20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circu

243、its Conference2 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis pumpTWEO pump designOn-chip driving circuit Primary-side-only Control Based on PT Theory Measurement Results Conclusion20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-

244、Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference3 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis pumpTWEO pump designOn-chip driving circuit Primary-side-only Control Based on PT Theory Measurement Results C

245、onclusion20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference4 of 76Evolution of MicrorobotsResonant magnetic microrobotElectroni

246、cally integrated microrobotUsing CMOS technologyM.Z.Miskin,et al.Nature 2020D.R.Frutiger,et al.IJRR 2009M.F.Reynolds,et al.Science Robotics 202220.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosi

247、s Actuation 2025 IEEE International Solid-State Circuits Conference5 of 76Evolution of MicrorobotsM.Z.Miskin,et al.Nature 2020D.R.Frutiger,et al.IJRR 2009M.F.Reynolds,et al.Science Robotics 2022Using bio-inspired actuatorNot fully compatible with CMOS technology20.6:Fully Integrated Self-Propelling

248、Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference6 of 76Challenges to Silicon-based MicrorobotsBatteryEnergy requirementWireless energy harvesting Viscosity dominat

249、edNeed effective actuator Robots in fluidsReynolds numberFeature lengthViscosityInertialFor microrobots20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid

250、-State Circuits Conference7 of 76Actuation of Microrobots+-+Bare surfaceSurface with absorbatesRequire post processingM.Z.Miskin,et al.Nature 2020C.Zhu,et al.ISSCC 2021P.Yen,et al.Scientific Reports2021Easy to fabricateFragileRobustSurface Electrochemical ActuationAC Electroosmosis20.6:Fully Integra

251、ted Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference8 of 76Energy Harvesting of MicrorobotsRequire specific processPoor permeabilityProcess friendl

252、yEfficiency varies with positionGood permeabilityEnergy wastingLEDRobotRF sourceRobotPhotovoltaicsMagnetic coupling20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE Interna

253、tional Solid-State Circuits Conference9 of 76Chip Overview1.3mm1.8mm20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference10 of 76C

254、hip OverviewOn-chip travelling-wave electroosmosis pump20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference11 of 76Chip OverviewM

255、agnetic coupled energy harvesting circuitParity-time symmetry efficiency compensation method+20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Cir

256、cuits Conference12 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis PumpTWEO pump designOn-chip driving circuit Primary-side-only Control Based on PT Theory Measurement Results Conclusion20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry

257、On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference13 of 76On-chip TWEO Pump:StructureSilicon substrateAluminum metal electrode12m12m20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symm

258、etry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference14 of 76On-chip TWEO Pump:PrincipleSubstrateElectrodeElectrodeSolution20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On

259、-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference15 of 76On-chip TWEO Pump:PrincipleSubstrate+-ElectrodeElectrodeSolution20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip

260、 Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference16 of 76On-chip TWEO Pump:PrincipleSubstrate+-ElectrodeElectrode+-+Solution20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip E

261、nergy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference17 of 76On-chip TWEO Pump:Control Signal09018027009018027009018027009018027020.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip En

262、ergy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference18 of 76On-chip TWEO Pump:Control Signal090180270090180270Time interval 1Substrate:Electrode with positive potential:Electrode with negative potential+Direction of net force of ions-20

263、.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference19 of 76On-chip TWEO Pump:Control Signal090180270Substrate090180270:Electrode w

264、ith positive potential:Electrode with negative potentialTime interval 2Direction of net force of ions+-20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid

265、-State Circuits Conference20 of 76On-chip TWEO Pump:Control Signal090180270Substrate090180270:Electrode with positive potential:Electrode with negative potentialTime interval 3+Direction of net force of ions-20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symm

266、etry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference21 of 76On-chip TWEO Pump:Control Signal090180270Substrate090180270:Electrode with positive potential:Electrode with negative potentialTime interval 4+Direction of net f

267、orce of ions-20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference22 of 76On-chip TWEO Pump:Flow Diagram Steady flow is generated

268、by the vortex of TWEO pump 0901802700 m60 m120 mCOMSOL simulation of the flow diagram20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Co

269、nference23 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis PumpTWEO pump designOn-chip driving circuit Primary-side-only Control Based on PT Theory Measurement Results Conclusion20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip

270、Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference24 of 76On-chip TWEO Pump Driving CircuitVDDCurrentbias10 kHzosc.20-30 nWVBD Q:4VcompResetShifter registerD QD QD QReset logic090180270Buffer17-25 nWTWEO pumpDriving signals are gene

271、rated by a relaxation oscillator and a shifter register20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference25 of 76On-chip TWEO P

272、ump Driving CircuitVDDCurrentbias10 kHzosc.20-30 nWVBD Q:4VcompResetShifter registerD QD QD QReset logic090180270Buffer17-25 nWTWEO pumpVDDVDDGNDGNDCurrent biasRelaxation oscillatorVBVBVcompResetTotal power consumption 1WDesigned to work under large VDD variation from 1.2V to 2.3V20.6:Fully Integrat

273、ed Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference26 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis pumpTWEO pump designOn-chip

274、driving circuit Primary-side-only Control Based on PT Theory Measurement Results Conclusion20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circu

275、its Conference27 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2Power SourceLoad20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circu

276、its Conference28 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2Power SourceLoadTX CoilRX Coil20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Sol

277、id-State Circuits Conference29 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-DCRFinRFoutRF+DCRF-20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE Internati

278、onal Solid-State Circuits Conference30 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-DCVarying coupling coefficient kVarying impedanceVarying efficiency20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Trav

279、elling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference31 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-DCTX IMNRX IMNIMN:Impedance matching network20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry O

280、n-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference32 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-DCTX IMNRX IMNT-type matchingT-type matching20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS wi

281、th Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference33 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-DCTX IMNRX IMNArea wastingIf the receiver side is a chip,thenLack of inducto

282、rHigh loss20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference34 of 76Basic Magnetic Coupled Serial-Parallel ModelkC1L1L2C2PAAC-D

283、CTX IMNRX IMNA primary-side-only control method should be adoptedPrimary sideSecondary side20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circu

284、its Conference35 of 76Parity-Time Symmetry Model:ConceptGainLossk20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference36 of 76Pari

285、ty-Time Symmetry Model:ConceptGainLosskEnergy20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference37 of 76Parity-Time Symmetry Mod

286、el:ConceptGainLosskEnergyParity TransformationGainLosskEnergy20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference38 of 76Parity-T

287、ime Symmetry Model:ConceptGainLosskEnergyParity TransformationGainLosskEnergyTime Transformation+20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State

288、 Circuits Conference39 of 76Parity-Time Symmetry Model:ConceptGainLosskEnergyParity TransformationGainLosskEnergyTime Transformation+Equal System20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmos

289、is Actuation 2025 IEEE International Solid-State Circuits Conference40 of 76Parity-Time Symmetry Model:PropertyGainLosskWhat if k varies in such system?20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Elect

290、roosmosis Actuation 2025 IEEE International Solid-State Circuits Conference41 of 76Parity-Time Symmetry Model:PropertyGainLosskIf k is large enough,energy can all go to the loss partEnergy20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy

291、Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference42 of 76Parity-Time Symmetry Model:PropertyGainLosskIf k is not large enough,energy will stay in the gain partEnergyEnergy20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with

292、 Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference43 of 76Parity-Time Symmetry Model:PropertyGainLosskkk0Real part of eigenvalueImag part of eigenvalue20.6:Fully Integrated Self-Propelling Micro

293、robot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference44 of 76Parity-Time Symmetry Model:PropertyGainLosskkk0Strong CouplingStrong CouplingReal part of eigenvalueImag part of

294、 eigenvaluePT symmetryPT symmetry20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference45 of 76Parity-Time Symmetry Model:PropertyG

295、ainLosskkkReal part of eigenvalueImag part of eigenvalue0No lossVaryingStrong CouplingStrong Coupling20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-S

296、tate Circuits Conference46 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRGainLosskGainLossk20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid

297、-State Circuits Conference47 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRGainLossReal part of the eigenvalueSteady-state frequencyImag part of the eigenvalueEnergy loss during the transmissionk20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symme

298、try On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference48 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRGainLossStrong CouplingPossible frequency variationMaintain constant efficiencykFrequencyk20.6:Fully Integrat

299、ed Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference49 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutImag(Zout)Real(Zout)Frequency(

300、MHz)k200-204804604400.050.100.150.200.25Frequency(MHz)k500-504804604400.050.100.150.200.25Ideal case:20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-S

301、tate Circuits Conference50 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutImag(Zout)Real(Zout)Frequency(MHz)k200-204804604400.050.100.150.200.25Frequency(MHz)k500-504804604400.050.100.150.200.25Ideal case:Zero phase lineGood for impedance matching20.6:Fully Integrated Self-Propelling

302、 Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference51 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutReal case:Under high frequency,Inductance(nH)QFr

303、equency(MHz)Frequency(MHz)040080012001600040080012001600100500-5050nH6420Q=620.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference5

304、2 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutReal case:Under high frequency,Inductance(nH)QFrequency(MHz)Frequency(MHz)040080012001600040080012001600100500-5050nH6420Q=6Rcoil20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip E

305、nergy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference53 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutReal case:RcoilRF+RF-Under high frequency,DC20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-G

306、Hz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference54 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutReal case:RcoilRF+DCRF-Under high frequency,Crec20.6:Fully Integrated Self-Propellin

307、g Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference55 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutReal case:RcoilCrecImbalanced coil designTX Coi

308、lRX Coil20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference56 of 76Connection Between PT Theory and WPT ModelC1L1L2C2-RRkZoutRea

309、l case:RcoilCrecImag(Zout)Real(Zout)Frequency(MHz)Frequency(MHz)k200-204600.100.200.300.40-40-60420380k0-504600.100.200.300.40420380Zero phase line50Smaller variation of impedance20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvestin

310、g and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference57 of 76+0+j0.5+j2+j1-j0.5-j1-j2Connection Between PT Theory and WPT ModelC1L1kZoutReal case:L2C2RRcoilCrecPAPA target regionZoutbefore PTPT Zoutafter PTVariation of kPA IMNAchieve good PA efficienc

311、y20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference58 of 76Wireless Power Transfer Circuit ArchitectureInput matchingControlRF

312、sourcePhase detectorBufferBufferAC-DCCircuit+TWEOpumpPhase detectorCpCsLsTX coilRX coilFB coilZoutRFink1k2On chip20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE Internati

313、onal Solid-State Circuits Conference59 of 76Wireless Power Transfer Circuit ArchitectureInput matchingControlRF sourcePhase detectorBufferBufferAC-DCCircuit+TWEOpumpPhase detectorCpCsLsTX coilRX coilFB coilZoutRFinVk1k2On chip20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GH

314、z Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference60 of 76Wireless Power Transfer Circuit ArchitectureInput matchingControlRF sourcePhase detectorBufferBufferAC-DCCircuit+TWEOpumpPhase detectorCpCsLsTX

315、 coilRX coilFB coilZoutRFinVk1k2On chipI20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference61 of 76Wireless Power Transfer Circu

316、it ArchitectureInput matchingControlRF sourcePhase detectorBufferBufferAC-DCCircuit+TWEOpumpPhase detectorCpCsLsTX coilRX coilFB coilZoutRFinVk1k2PhaseVIIPhaseOn chip20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travell

317、ing Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference62 of 76On chipWireless Power Transfer Circuit ArchitectureInput matchingControlRF sourcePhase detectorBufferBufferAC-DCCircuit+TWEOpumpPhase detectorCpCsLsTX coilRX coilFB coilZoutRFinVk1k2PhaseVIIPhaseVILower

318、frequencyLower k1Phasetttttt000PTPhase detectorPhase20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference63 of 76On-chip Energy Ha

319、rvesting CircuitStacked MIM cap(2.8fF/m2)and poly cap(5fF/m2)NwellMIM cap.Poly cap.M3MIMM2 On-chip coil50nHRectifierOn-chip storage capacitor5nFDiode-based full-bridge rectifier20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting

320、and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference64 of 76Outline Introduction On-chip Travelling-Wave Electroosmosis pumpTWEO pump designOn-chip driving circuit Primary-side-only Control Based on PT Theory Measurement Results Conclusion20.6:Fully In

321、tegrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference65 of 76Measurement Setup for TWEO PumpPower meterReceiving chipTWEO pumpDC supplyRF sourc

322、ePC20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference66 of 76Measured TWEO Pump 0s6s100 m100 mFaster flow speed can be generate

323、d under lower concentrationAverage power 50WPower Consumption at 1.8V VDD(W)Avg.Power Consumption at 1.8V VDD(W)Time(s)050100 150 200250300-50510152025303540Without Load0.001mmol/L KCl0.01mmol/L KCl0.1mmol/L KCl5.511.018.324.40.001mmol/L 0.01mmol/L9.8 m/s8.3 m/sFlow speed at 2500Hz and 1.8VFlow Diag

324、ram20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference67 of 76Measured TWEO PumpMeasured at 2.3V and 0.001mmol/LMeasured at 1.8V

325、 and 0.001mmol/L20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference68 of 76Measurement Setup for WPTRF sourceDC supplyReceiving

326、chipChip packagePhase detectorTX boardOscilloscope20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference69 of 76With 100K load0.100

327、.200.300.400.500.600.703753803853903954004054104150.450.640.830.26With frequency variationFixed 390 MHzFixed 385 MHzFixed 395 MHzPT symmetry regionPT broken regionFrequency0.00DC-DC Efficiency(%)Varied Frequency(MHz)Coil Gap(mm)Measured WPT EfficiencyMeasured at fixed driving power RFIN and VDSBy ch

328、anging the frequency,efficiency can be maintained within a wider rangeWith 15K load0.100.150.200.250.300.350.400.450.50With frequency variationFixed 390 MHzFixed 385 MHzPT symmetry regionPT broken region375380385390395400405410415Frequency0.450.640.830.26DC-DC Efficiency(%)Varied Frequency(MHz)Coil

329、Gap(mm)20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE International Solid-State Circuits Conference70 of 76Measurement WPT PerformanceWith 100K loadFor certain coil gap

330、and load resistance,the PT frequency barely changes under different harvested voltages1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0375380385390395400405410415At coil gap of 0.38mmAt coil gap of 0.30mmAt coil gap of 0.46mmMeasured PT Frequency(MHz)Feedback WaveformTime(ns)-3-2-101230.0-0.1-0.20.10.2V(V)I(V)-1-

331、2012-3-2-10123V(V)I(V)0.0-0.2-0.40.20.40-224-4-3-2-101230.0-0.1-0.20.10.2At PT frequencyV(V)I(V)-1-2012Harvested Voltage(V)20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025 IEEE

332、 International Solid-State Circuits Conference71 of 76Measured Movement0s1s2s3s4s5s6s7s500 m500 m500 m500 m500 m500 m500 m500 m20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosis Actuation 2025

333、IEEE International Solid-State Circuits Conference72 of 76Chip Die PhotoThe chip size is 1.3mm*1.8mm and is fabricated in 180nm SiGe processStorage Cap.RectifierTWEO ElectrodesBias&Osc.RX CoilStorage Cap.RectifierRX CoilBias&Osc.TWEO Electrodes20.6:Fully Integrated Self-Propelling Microrobot in 180nm CMOS with Sub-GHz Parity-Time Symmetry On-Chip Energy Harvesting and Travelling Wave Electroosmosi