Bifacial CdSCdTeZnTe device characterization

1
Bifacial CdS/CdTe/ZnTe device characterization

• Darshini Desai, Steven Hegedus
• Institute of Energy Conversion
• University of Delaware
• Newark DE, USA
• Development of the transparent contact device and
  characterization techniques driven by need for
  better quantitative metrics and understanding of
  CdS/CdTe device operation

2
Motivation and Goals

• Motivation
• A transparent back contact made with
  controlled Cu doped ZnTe can be used as
• - A source of controlled Cu doping
• - A robust back contact
• - Transparent interconnect for tandem cell
  applications
• - A novel device characterization tool
• Goal
• Development of ZnTeCu based transparent,
  ohmic and stable contact for CdTe/CdS solar cells
  to enable bifacial device characterization and
  provide alternative source of Cu doping.

3
Transparent (Semi) ZnTeCu Film Deposition

• Films grown by galvanic deposition with Zn as
  anode and substrate as cathode
• Source electrolytes ZnSO4, TeO2, CuSO4
• Deposition time 1.5 mins, pH3 and T68C
• Triethanolamine(TEA) used to regulate amount of
  Cu in the solution. More TEA results in less free
  Cu and hence higher film transparency.
• Selected 20 drops1E-8M for device fabrication

4
JV results summary
5
JV data with Bifacial Illumination

• VT 154.4.2 back-wall JV has no blocking barrier,
  unlike the dark and front-wall
• JV Suggests photoconductive back contact
• Both devices Higher fill-factor for back-wall
  illumination due to lower I2R loss

6
Spectral Response Analysis and Modeling

• Fit modeled Phillips 1st WCPVEC 1994 to
  measured SR with W2.5?m and L 0.8?m Increasing
  L from 0.5?m to 1.5?m results in better
  collection for all wavelengths, for back wall SR
• Front wall SR unaffected by L, back wall SR
  sensitive to L 400-800nm and W800-860nm

7
Voc vs. T vs. Intensity A fundamental
characterization tool

• J-V at low T often distorted by blocking diode
  and/or very high Rs
• Cannot analyze J-V curve with single diode model
• JO at Voc so no blocking, Rs
• Voclinear in -T, log in I
• Voc upper limit is fundamental parameter
• Slope dV/dT indicative of recombination mech.
• JL depends on V? field driven collection
• JL(V)JL0 ? (V)

8
Voc-T linear fit and extrapolation above 220K

• Voc(0K)1.5 V indep. of intensity

9
Voc saturation below 200K

• After 5 days stress
• Voc(300K) decreased 0.03V
• Voc(150K) decreased 0.06V

• Initial results
• Saturation independent of T, L
• No recombination?

10
Bifacial Voc vs. T at 100, 10 light

• Front illumination Voc independent of
  temperature and light intensity below 220K
• Back illumination Voc continues to increase
  irrespective of front-wall Voc
• Transient increase in Voc mins observed at
  temperatures below 110K from dark to 100
• This behavior seen in ALL CdTe cells
  irrespective of the back contact or the
  manufacturer

11
Bifacial Jsc vs. T Intensity and Spectral
effects
Jsc(mA/cm2)
Jsc(mA/cm2)

• Front illumination Jsc decreases below 220K.
  Virtually no change above 220K
• Back illumination Jsc increases below 220K .
  Opposite effect
• Mobility increases at lower temperature ?
• Temperature dependence of field/depletion region
  which causes better collection from back but
  reduces field in front?
• Similar temperature behavior seen in FS samples

12
Maximum achievable Voc Band diagram model
SnO2
CdS/CdTexS1-x
CdTe
Electron fermi level pinned at SnO2 interface
Ec - 0.3eV
Hole fermi level pinned at Cu defect Ev 0.3eV
13
Observations

• Voc saturation below 220K observed for gt 30
  samples
• Is the Voc _at_ 100K the Built-in voltage Maximum
  achievable limit ??
• - Unclear whether back-wall Voc really
  saturates due to system limitations
• Fermi-level pinning at SnO2 interface alters the
  quasi-fermi level profile
• dEg/dT (- 1.70 mV/K) so CdTe bandgap at 100K
  1.7eV. If this hypothesis valid then the maximum
  achievable Voc is
• 1.7 - 0.3- 0.3 1.1 eV
• Typically _at_ 100K maximum Saturated Voc
  0.95-1.05V achieved
• Other issues with low temperature measurement
• - Blocking contact at lower temperatures
• - Defect/Trap energy level positions not
  known
• - Slow transient increase after light
  exposure
• Interpretation of Voc complicated by large Jsc
  temperature dependence

14
Conclusions and Future Work

• Bifacial device characterization enhances
  conventional characterization tools and helps to
  isolate properties of front and back junction
• Temperature dependent Voc and Jsc measurements
  fundamental to quantify transport in CdTe/CdS
  solar cells
• No spectral dependence of Voc or Jsc
• Bifacial spectral response model fit to data with
  W 2.5?m and L 0.8?m the back-wall spectral
  response sensitive to L and W
• In future intend to make cells with thinner CdTe
  to evaluate drift and diffusion limited transport
• We intend do Voc-T measurement on cells with
  CdZnTe window layer to further investigate the
  possibility of interface recombination limited
  transport