Presentation is loading. Please wait.

Presentation is loading. Please wait.

Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs A.Fox, M. Marso CWRF2008 CERN.

Similar presentations


Presentation on theme: "Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs A.Fox, M. Marso CWRF2008 CERN."— Presentation transcript:

1 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs A.Fox, M. Marso CWRF2008 CERN

2 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20082 Research Center Juelich? Where: »between Cologne>>>>>Aachen areas of research : »M aterial »E nvironment »I nformation »L ife »E nergie »Established more than 50 years

3 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20083 Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Conclusion Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis introduction

4 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20084 Application example AlGaN/GaN HFETs are used in rf-generators, telecommunication and other applications where power devices at higher frequencies, voltages and temperatures are needed.

5 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20085 AlGaN/GaN Heterostructure and 2DEG This 2DEG charge plate creates the area with very high mobility and sheet concentration of carriers Formation of 2DEG in AlGaN/GaN heterostructure heterostructure is the layer system with different band gap material e.g. with AlGaN grown on GaN. High Electron Mobility Transistor D Ohmic contact S Ohmic contact G Schottky contact Ohmic contacts are controlled by the Schottky Gate conntact

6 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20086 How to improve power performance P out = I max * (V breakdown -V knee ) /8 for increasing P out an output voltage- and current increase is needed Example I-V characteristic I max increases due to increase carrier density and velocity V breakdown increases e.g. -with fieldplate technology -wide bandgap ClassA operating point I max Max. output power: f(I ds,V ds ) Opt. R L =(V b -V s )/I max V knee V breakdown

7 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20087 Theory of HFETs: DC behavior Drain Current (I ds ) How do the DC measures translate into device geometry and material properties? Transconductance (g m ) h W LGLG qn 0 r absolute charge underneath the gate equals charge in active region of 2DEG: insertion of expression into saturated current formula......yields: differentiation yields intrinsic transconductance (robertson2001a) the formulas tell us to: charge carrier concentration mobility, velocity IdId VdVd

8 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20088 Advanced Material Properties , W/cm-K v sat, 10 7 cm/s v peak, 10 7 cm/s , cm 2 /Vs E BR, 10 5 V/cm E G, eV GaN6H SiC4H SiC InGaAs * GaAsSi Important for high output power: - high breakdown field and voltage, i.e. wide bandgap -high thermal conductivity Important for high f T and f max : Fast carriers (i.e. µ 0, v peak, v sat ) Frank Schwierz Fachgebiet Festkörperelektronik, Technische Universität Ilmenau, Germany

9 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF20089 HFET performance improvement *Juraj Bernat Fabrication and characterisation of AlGaN HEMT (Diss Aachen) Enlarging of the Source - Drain distance is limited degradation of dc and transport properties due to long S_D distance

10 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF HFET performance improvement Fieldplate technology: Known since 1969 on Si Higher breakdown voltage Different electric field distribution between Gate and Drain *Juraj Bernat Fabrication and characterisation of AlGaN HEMT (Diss Aachen) Performance improved by Fieldplate technology on AlGaN-HEMT

11 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Fieldplate Technology - Results Increase of the output power is due to modification of the electric field on GaN cap Confirmed by: A. Koudymov, IEEE Electr. Device L. 26, Oct * Juraj Bernat Fabrication and characterisation of AlGaN HEMT (Diss Aachen) Simulation: by ATLAS, Silvaco

12 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Experiment Unpassiv. HFET Simultaneous fabrication of unpassivated & passivated SiC/GaN/AlGaN HFETs and MOSHFETs: 10 nm SiO 2 layer for passivated HFETs and MOSHFETs (PECVD), L G = µm, L W =200 µm (2-fingers), Source Drain Gate SD SiO 2 ~10 nm passivated HFET SiO 2 ~10 nm MOS HFET Gero Heidelberger: Technology related issues regarding fabrication 4H-SiC 3 µm GaN 30 nm Al 0.28 Ga 0.72 N G

13 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF HFET- technology Mesa etching –ECR RIE or Ar + sputter –Depth: 250 ~ 300 nm Ohmic contacts –Ti/Al/Ni/Au –Annealing: 850ºC, 30sec Schottky contacts –Ni/Au Pads –Ti/Au Fabrication of our HFET Devices

14 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF HFET Layout SEM picture of AlGaN/GaN HFET wg=200 µm Lg=0.3 µm Hetero Field Effect Transistor SEM picture of AlGaN/GaN HFET with airbridge technology >> increasing I max Detailed view of HFET Device as fabricated in our lab prepared for on-wafer measurements with 100 µm pitch

15 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

16 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF S-parameter Measurement System Important for microwave design and characterization Basics for parameter extraction and simulation Our measurement system: –Frequency up to 110 GHz –Network Analyzer –On wafer measurements –Control System and Calibration (LRM)

17 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF GHz Measurement System We are well prepared for high quality s-parameter measurements HP8510XF NWA-System: -Network Analyser HP8510C - RF source 45 MHz 50GHz Synth. sweeper HP LO source 45MHz to 20GHz Synth. Sweeper HP mm-wave controller -Cascade probestation -Connectors: 1 mm coaxial -Coplanar Probetips 110 GHz -LRM Calibration, attanuation DUT mm-wave controller FZ-Jülich, ISG1, A. Fox

18 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Results of S-parameter measurements drain source voltage: 20V gate length: 500 nm SiO 2 layer thickness: 10nm cut off frequency (f t ) is the frequency where current gain h 21 =0 [db] MOSHFETs show significantly higher cut off frequencies

19 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Variation of f t vs. gate length Higher cut-off frequencies (f t ) are observed for MOSHFETs for all kinds of gate lengths. f t increases with decreasing gate lengths.

20 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

21 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF RF-Power measurement Load Pull Measurement System step attenuator Microwave Tuner System slotted coaxial line computer controlled slug Tuner cal. deembeding Focus Microwave Ref.plane resonanz.

22 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF RF-Power measurement system FZ-Jülich, ISG1, A. Fox Large signal measurements were performed on wafer at 7 GHz by source and load pull measurements Load-tuner Source-tuner Networkanalyzer HP8510B DUT Bias supply

23 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Power performance will increase with further increase of V ds for passivated HFET Load pull measurement

24 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Power measurement at differend SiO 2 layer Comparison of P out for HFET and MOSHFET with different SiO 2 dielectric 7 GHz Performance improvement by optimising the dielectric layer thickness

25 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Comparison of Power measurements The RF output power increases from 2.9 W/mm (HFET) to 6.7 W/mm (MOSHFET) with no fieldplate implemented SiO 2 layer thickness: 10nm W G = 200 µm, L G = 0.7 µm f = 7 GHz V g = -2.5 V HFET V g = -2.5 V passiv. HFET V g = -7 V MOSHFET

26 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

27 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF C gs and g m change with the dielectric layer Modelling Equivalent circuit for the AlGaN/GaN MOSHFET S G D S f t =f( g m /C gs ) Oxyde deplition

28 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Extracted Parameters C gs, g m C gs decreases for MOSHFETs due to additional SiO 2 layer v sat increases with passivation From theory:

29 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Modelling Comparison of g m /C gs ratio g m /C gs is in agreement with the increase of cut off frequency from h 21 for the MOSHFET compared to HFET f t =f( g m /C gs )

30 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Modelling MOSHFET V g =-1V V ds =19V Frequency: 2 to 50GHz S22 S11 Good agreement between measurement and optimized model Measurement: blue curves Model: red curves Comparison of measurements and model S21 S12 Parameterextraction done by TOPAS, IMST

31 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

32 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF DC analysis DC design circuit DC simulator bias sweep DUT with the modeled parameters Advanced Design System, Agilent DUT

33 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF DC analysis: Results The curves are comparable to DC-measurements and reflect the intrinsic behavior Output I-V- curves Transfer I-V- curves Transconductance

34 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF S-parameter analysis S-parameter design circuit S-parameter simulator frequency sweep DUT

35 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF S-parameter analysis measured: red curves simulated: blue curves Comparison of Transistor Scattering Parameters S11 S21 S22 different fitting at dif. biaspoints needs further working point

36 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

37 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Specifying and generating desired load and source reflection coefficients (impedance) Biasing the device and running a simulation Calculating desired responses (delivered power, PAE, etc.) Load-pull simulation

38 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation Load pull simulation varies the load reflection coefficient presented to a device... DUT Load- Pull design circuit Harmonic-Balance-simulator for nonlinear circuit behavior ? ADS © Agilent

39 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation …to find the optimal value to maximize power or PAE, etc. This data shows the simulation results when the reflection coefficient is the value selected by the markers location.

40 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation Harmonic Balance simulator Input power sweep Non linear circuit simulation results in P out and PAE of the DUT DUT Matched Input impedance Matched Output impedance

41 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Outline Measurements RF-power Simulation Comparison of simulated and measured RF-power performance Modelling S-Parameter RF-power Load-Pull Simulation DC-, S-parameter analysis

42 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Comparison of simulated and measured PAE MOSHFET, SiO 2 thickness: 10nm Curves are acceptable between measurement and simulation of PAE

43 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Comparison of simulated and measured output power MOSHFET, SiO 2 thickness: 10nm Good agreement between measurement and simulation of P out

44 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Conclusion Increase of RF-performance by introduction of the MOSHFET-technology. - MOS-HFET superior to unpassivated and passivated HFET regarding DC-, RF-, and power-performance. Increase of output power, PAE and cut-off frequency. Simulations verify the measured large signal results. Further work: alternative dielectric material and additional fieldplate implementation

45 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF CWRF2008 Thank you!

46 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF ENDE Vortrag

47 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF RF-Power measurement (Load-Pull ) Comparison of RF-Power performance for passivated HFET and MOSHFET PAE=(Pout-Pin)/Pdc MOSHFETHFET MOSHFETs show significantly higher output Power and PAE

48 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation matched load impedance DUT Harmonic Balance Input power sweep Load-Pull design circuit matched source impedance Non linear circuit simulation results in P out and PAE of the DUT

49 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation DUT Detail circuit of Load-Pull design circuit

50 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Load pull simulation varies the load reflection coefficient presented to a device... Load pull simulation … to find the optimal value to maximize output-power

51 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Modeling From physical structure to equivalent circuit

52 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF CV-measurement

53 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Modeling Comparision of measured and simulated S-Parameter S1 1 S1 2 S22

54 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Output Power To reach maximum P out : –High I d sat –High V br –Small V knee Small R s increased output power by increasing Id and Ud

55 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Reduction of Gate Leakage Current MOSHFET using 11nm thick SiO 2 gate isolation: 30% increase of the drain saturation current Reduction of gate leakage current from to A/mm *Juraj Bernat Fabrication and characterisation of AlGaN HEMT (Diss Aachen)

56 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Wide bandgap semiconductors Common semiconductors E G around 1 eV SiE G = 1.1 eV GaAsE G = 1.4 eV Wide bandgap semiconductors E G > 2 eV 4H SiCE G = 3.2 eV 6H SiCE G = 3.0 eV GaNE G = 3.4 eV AlNE G = 6.1 eV AlGaN E G = eV CE G = 5.5 eV Narrow bandgap semiconductors E G << 1 eV InAsE G = 0.35 eV InSbE G = 0.17 eV

57 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF SiC and GaN based Transistors Cree, Inc : Pout= 32.2 W/mm, PAE= 54,8%, Fe_doped and Field plate GaN Pout=4.1W/mm, PAE=72%

58 Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF Heterostructure and 2DEG A heterostructure is the layer system where two semiconductors with different band gaps Eg are grown one on the other In the thermodynamical equilibrium,when both semiconductors are "connected" together, the Fermi-level energy (EF ) of the Semiconductor I and Semiconductor II must be in the line what cause the discontinuity in the conductance (EC) and valence (EV ) band and the band bending. This results in the formation of the triangular quantum well where carriers are fixed in one axis and the 2DEG is formed.


Download ppt "Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs A.Fox, M. Marso CWRF2008 CERN."

Similar presentations


Ads by Google