Optoelectronics properties investigation of III-V nitride based heterostructures.ppt

Document Sample
Optoelectronics properties investigation of III-V nitride based heterostructures.ppt Powered By Docstoc
					Optoelectronics properties investigation
 of III-V nitride based heterostructures

                  S. Pandey
         Universita di Bologna, Italy

Ø Structures
• InAlN/AlN/GaN heterostructures
• InGaN/GaN heterostructures
Ø Characterization techniques and results
• Basic electrical characterization (I-V, C-V)
• Surface Photovoltage Spectroscopy (SPV)
• Photo-current Spectroscopy (PC)
Ø Summary
Ø Future work
             1. InAlN/AlN/GaN structures
• Introduction and samples
Electrical Characterization results:
• I-V characteristics in back to back Schottky configuration and
   2DEG carrier concentration
• I-V characteristics of Ni/Au Schottky diodes
• C-V characteristics of Ni/Au Schottky diodes
• Mobility calculations and scattering mechanisms
Optoelectronics Characterization results:
• Surface Photovoltage (SPV) analysis
• Photocurrent (PC) analysis
           InAlN/AlN/GaN heterostructures: introduction

    · Strong spontaneous and piezoelectric polarization
    · lattice-matching of AlInN to GaN
    · high 2DEG density
      Applications of III-V nitrides heterostructures
    · HEMTs/HFETs, LEDs, Lasers, Photodetectors etc.

    Samples                                                         1-D Schrödinger-Poisson solution**
Ø    InAlN/AlN/GaN hetrostructures were grown by AIXTRON            for InAlN/AlN/GaN heterostructures
Ø    5 samples with different AlN interlayer thickness (0-7.5 nm)
    have been investigated
Ø    Barrier layer of InAlN thickness is around 15 nm for all
Ø    %In in InAlN layer has been detected around 13- 14 % in all
Ø    GaN template thickness is around 3 μm

     A. Vilalta et al., pss(a) 207, 1105 (2009)
       Tan et al. JAP 68, 4071 (1990)
Electrical Characterizations results
       1. Current-Voltage measurements on Schottky contacts in planar

Ø Two back to back Schottky contacts in planar configuration* have been directly made by
  using In-Ga alloy for all samples
Ø Current-Voltage measurements have been performed at 300K
Ø Change in slope         current transport from barrier layer to AlN/GaN interface

*Carrano et al. JAP 83, 6148 (1998)
 S. Pandey et al. Phys. Solidi. Stat., in press (E-MRS proceeding)
                                             AFM analysis

                            1 nm AlN                        7.5 nm AlN

 Presence of micro/nano sized cracks in case of 7.5 nm AlN interlayer thickness
 Ø Higher value of dislocation density in comparison to other samples
 Ø It results as quite high leakage current and shows ohmic behavior

More detailed analysis on role of cracks :
A. Minj et al. submitted for publication
                                   Our model
Ø Polarization charge density (P) due spontaneous and piezoelectric polarization has
been calculated:

Ø Total electric field for each structures can be calculated from equation:

                                                            Etotal            InAlN

 Ø2DEG density can be calculated from equation:
       Results from I-V model and comparison of results with other

    Ø Calculated 2DEG density values from I-V model* shows excellent agreement
      with values obtained from Hall measurement and C-V**
    Ø Sheet resistance values are in quite good agreement with Hall data

  *Pandey et al., APL, 98, 012111 (2011)
**Gonschorek et al., JAP, 103, 093714 (2008)
  Jena et al. APL 90,182112 (2007)
    2. Current transport properties of Ni/Au Schottky contacts*
Forward and reverse bias characteristics                                  Poole Frenkel emission

From AFM analysis:
On increment of thick AlN interlayer                   Current variation associated with Poole-Frenkel emission
insertion, dislocation density (V-pits)                could be expressed as :
increases from 2 x 109 cm-2to 2.5 x 1010
cm-2 for 1 and 7.5nm thick AlN layer

*Ni/Au Schottky contacts were fabricated by T. Brazzini, ISOM, Madrid
Forward bias extracted parameters                    Calculation of surface donor density

Ø Probably, change in SBH could be due to presence of surface/interface states at metal-
  semiconductor interface?

ØFrom the slope of this, surface/interface state density has been calculated, which is around ~
 2.73x1013 cm-2/eV, which is in good agreement with reported values for AlGaN/GaN

 *Ibbeston e al., Appl. Phys. Lett. 77, 250 (2000)
  S. Pandey et al., manuscript under preparation
      3. Capacitance-Voltage properties on Ni/Au Schottky contacts
   Ø C-V measurements were performed at different frequency (1 KHz-100 KHz)
   Ø 2DEG carrier concentrations values and Thickness of layers were calculated for different samples

        Sample with 0.5 nm AlN interlayer                      Sample with 2.5 nm AlN interlayer

Ø Carrier concentration is calculated of the order of 1013 cm-2 which is comparable to 2DEG
concentration from our I-V model and Hall measurements
Ø Layer thicknesses are in good agreement with reported values by TEM analysis
       4. Mobility Limiting Mechanisms in Polar nitride Semiconductor
                                              Surface roughness: the model
 Possible mechanisms:                         Ø Interface roughness as surface roughness*
 Ø Interface roughness as surface roughness
                                              Ø fluctuations in subband energy levels
 Ø Dislocation scattering
                                              Ø mobility as a function of subband energy levels and
 Ø Coulomb scattering
 Ø Phonon scattering
                                              Ø effect on transport and quantum life time

                          Conduction Band profile and subband
                                  energy fluctuations

      First ground subband energy level

Cao and Jena, APL 97, 222116 (2010)
Liu et al., , APL 97, 262111 (2010)
       Required parameters:

       Ø Surface roughness (by AFM)
       Ø Dislocation density (v-pits density) (by AFM)
       Ø Correlation length (by AFM and STM)
       Ø 2DEG carrier density (by Hall )

Correlation length (Λ) evaluation by AFM and STM maps
one-dimensional height-height correlation function (HHCF)
is calculated as*

            ΛAFM= 60 nm                                            ΛSTM= 2 nm
 *H. Tang et al., Phys. Rev. B 66, 245305 (2002)   *S. Pandey et al., submitted
                         Mobility limiting mechanisms: evaluation
             1. Surface roughness scattering (RSR)

             2. Dislocation scattering*with dislocation density = 2 x V-pits density
                                                                      Mobility vs roughness

                                          Total mobility, µ-1T = µ-1RSR+ µ-1DIS
               Mobility vs roughness for
              different correlation lengths

                                                           Surface roughness is the dominant scattering mechanism for
*D. C. Look et al., PRL 82, 1237 (1999)                    2DEG nitride based heterostructure
Optoelectronics Characterizations results
       Optoelectronics characterization analysis

               1. Surface Phovoltage (SPV)

        SPV set up           SPV vs. λ         EG

                 2. Photocurrent (PC)


lamp     PC set up             IPC vs. λ            efficiency
       (contact based)

                                                                        ΔEG = E G, HET – E G, GaN

                                                                  -Burstein Moss (BM)
                                                                  -Renormalization effect (RN)
  band-to band transitions at the GaN /AlN                      Dependence of ΔEG vs 2DEG density n2D
                                                      S. Pandey et al., Phys. Stati. Solidi, in press (ICNS9 proceeding)
The same effect is not detectable by   Band to band
transmission spectroscopy              transitions at the GaN surface
               2DEG thickness ‘h’ and volume electron density ne


   As the material is degenerate at the hetero interface
    both BM and RN effects should be considered

ØConsideration of calculated strain induced band gap
Ø E* allows for the calculation of m* vs ne. The values
   are in good agreement with the ones reported in**

    Q Yan et al Appl Phys Lett. 95, (2009)                 D. Cavalcoli et al., Appl. Phys. Lett. 98, 142111(2011)
     S. Shokhovets et al, Phys. Rev. B 78, 035207 (2008)
                           2. Photocurrent analysis

Ø Photocurrent measurement was performed at 300 K

          AlInN/AlN/GaN                                     AlInN/GaN

Ø Strained GaN bandgap and 2DEG (1st subband) related transitions are observed
Ø experimental values are comparable with 1-D Schrodinger-Pöisson solver results
      SPV and Photocurrent analysis on lattice matched structures
Ø Samples were grown at 3-5 Lab, France
Ø %In is kept around 17-20% in InxAl1-xN layers

                                              Ø GaN and 2DEG related band to band
                                                 transitions are observed by SPV
                                              Ø 2DEG related transitions are observed by
                                                 PC as it shows maximum peak around
                                              Ø experimental results on first subband
                                                energy level values show good
                                                agreement with simulated ones by
                                                Schrodinger Pöisson solver
InGaN/GaN structures
                SPV on Si-doped InGaN/GaN structures
  Ø Samples were grown by MOCVD at AIXTRON*
  Ø InGaN layer thickness was around 50 nm and %In is around 19%
  Ø Doping concns. are 9x1018 cm-3and 1x1019 cm-3 for sample 1 and 2         GaN (3µm)

 Ø n-type doping is obtained from SPV analysis   Ø effect of doping on energy gap is observed
                                                 Ø no signature of defects related transitions

*More details on growth could be discussed with O. Tuna, AIXTRON
                       Summary and Future work
Ø InAlN/AlN/GaN HEMT structures were widely investigated to understand
  electrical and optoelectronics properties
Ø Presence and role of 2DEG is investigated by optical and electrical measurements
Ø Mobility limiting mechanisms have been identified
Ø Leakage and Schottky diodes issues have been discussed
Ø InGaN/GaN structures have been briefly analyzed

Future Work
ØDeep Level Transient Spectroscopy (DLTS) to be performed to understand
  the activity of different kind of defects/traps level
Ø Role of defects on electrical properties needs to be investigated to improve device
Ø Role of defects on LEDs (InGaN) efficiency droop to be investigated

Shared By: