Flexible Photovoltaic Devices

ENMA466 - Testing Geometry Efficiency

Joseph Ayoub

Sabrina Curtis

Julia Downing

Maria Pascale

Alex Randolph�Haotian Wang

Outline

**Purple = new content. Note: Previously presented slides at the end of the presentation

​

1. Question/Goal
2. Division of Labor
3. Geometries and PN Junction Dimensions
4. Updated Process Flow
5. Fabrication Updates and Upcoming Steps
6. Modeling: COMSOL and ANSYS
7. Power and Efficiency Calculations
8. Short-term To-Dos
9. Characterization Methods
10. Schedule

Question / Goal

End Goal: Develop a flexible/ stretchable photovoltaic device

​

Future Work - Capstone Goal:

Release device from wafer onto a flexible substrate (i.e sillicone) and test device performance while strained (0%, 50%, 100%)

Micro: Develop and characterize multiple unstretched photovoltaic devices of different geometries. Determine the influence of geometry on device performance.

Controls:

Straight line control for each variable, thickness

Variable: Shape of Device (geometric pattern), amplitude, wavelength,and periodicity

​

Division of Labor

Fabrication and Characterization - All

Leads - Maria, Joe, Sabrina

Masks and CAD - Sabrina, Julia

FEM: ANSYS- Julia, Alex COMSOL - Haotian, Sabrina

Scheduling / Gantt - Maria

Process Flow - All

Power/Efficiency Calculations - Haotian, Julia helping

​

Geometry (Top View)

Expectation: Geometries with a more gradual change in direction (sine, horseshoe, curved corner rectangle) will perform better mechanically and electrically.

Straight Line

Rectangular

Curved Corner Rectangular

Horseshoe

Dimensions: Addition and Modification of Si Island

150 um

150 um

25 um

Si = 110 um

N = 40 um

ITO = 15 um

P = 40 um

125 um

150 um

185 um

Updated Fab Process as of 11/1/16

Current step

Fabrication Updates

• Completed: Alex, Joe and Julia performed deposition and patterning of ITO (up to mask 3)
• Tuesday - Sabrina annealed ITO to achieve desired resistivity
• Next step: silicon island etch for device isolation

• Note: discontinuities in horseshoe geometries (not pictured) used for comparison with previous literature

ITO contact

Doped region

Upcoming Process Steps

Modeling - COMSOL and FEM

Goal:

• Investigate various parameters influencing peak stress in different programs to check results

Model:

• PV device with various geometry embedded in a stretchable substrate, such as Sylgard 184 and Kafton tape

Boundary conditions:

• 10000N/m stress will be applied to one end of the substrate, while the other end stay fixed

​

​

Comsol simulation analysis

• Boundary condition: 10N net force, 10000N/m force per length applied at the right boundary, with left boundary fixed. Thickness of the structure is set to 14um, with cross section length for each silicon region equal to 40um.
• PV device will fail in this case
• Have to look at parameters that can attenuate peak stress in the structure

​

Comsol simulation analysis:

Cycle of geometry

• Maximum peak stress decreases with more cycle of geometry

Comsol mechanical simulation analysis:

Thickness of elastomer substrate

• Thickness of substrate is set to 1 mm
• The thicker the substrate, the less the maximum peak stress
• Thick substrate absorp part of the stress in the PV device
• Device will still fail, since tensile stress of silicon is 7GPa

Comsol mechanical simulation analysis:

Strength of elastomer material

• Elastomer has been replaced to silicon rubber(Young’s modulus:0.05GPa, previously used Sylgard 184 with Young’s modulus: 1.32MPa)
• Thickness of substrate is set to 1 mm
• For composite material, the stronger phase takes the majority of stress
• If stronger substrate material is used, stress and strain in PV device will be relaxed

Comsol simulation analysis

Geometry

• Elastomer has been replaced to silicon rubber
• Thickness of substrate is set to 1mm
• Geometry affects stress distribution
• Rectangle PV device is more mechanically durable

Comsol mechanical simulation analysis

Conclusion:

For the curved corner structure:

• The stress concentrate at the corner.
• Thickness of elastomer can effectively affect the stress in the device.
• More geometry cycle will reduce the maximum peak stress.
• Stronger elastomer is beneficial for a stretchable PV device.

For the rectangle structure:

• Rectangle geometry is more mechanically durable than curved corner, while both geometry can be adapted to mechanical test.
• Other parameters need to be investigated in the future, such as size of PV device, crystal orientation of silicon, and test conditions (boundary conditions).

ANSYS Simulation

Only a partial mesh is currently attainable - boundary geometries are too small to mesh in ANSYS for the whole device

ANSYS Simulation Results: Horseshoe Si trace on Kapton tape

Von Mises stresses: maximum 0.57 x 10¹⁰ Pa or 5.7 GPa

ANSYS Simulation Results: Horseshoe Si trace on Kapton tape

Si material model: S (compliance)in units of 10^-11 Pa

Power and Efficiency Calculations

Bottom-up approach:

Power and Efficiency Calculations

Top-Down Approach:

• From Yonggang: It would be easier to do efficiency reduction rather than calculating from bottom-up. For example, the current solar cell efficiency is around 20%. But your system has much thinner n-doped layer so that less light will be absorbed and therefore you have a reduction in efficiency of xxx. This number can be found from the thickness vs. light absorption in silicon solar cell. You need to consider several factors: light absorption, resistivity, etc.

Work Needed in the Near Future

• FEM of three geometries (Alex and Julia complete by 11/11)
• Finish calculations of expected power output of device (Haotian, et al.)
• Next fab steps: patterning n-dopant pits in SiO2 using Mask 1
• Schottky barrier calculations
• Make characterization reservations

​

Potential Characterization Methods (lots)

• Current-voltage (I-V)
• Output
• Turn on voltage
• External Quantum Efficiency (EQE)
• # charge carriers collected to # of incident photons
• Solar Simulation
• Imaging: SEM/EDS; TEM

Abbreviated Schedule

Old Slides

Motivation

• Flexible electronics attractive for wearable applications in military and consumer industry
• limitation is the inability of the metals and conductive polymers to stretch significantly
• Most common approach to make a material both conductive and elastomeric is to embed metal particles into an elastomer
• We are interested in using established fabrication techniques to create thin film-stretchable solar cells.
• potential to be lower cost and higher performing than traditional solar cells

Kim, Dae-Hyeong, et al. "Epidermal electronics." science 333.6044 (2011): 838-843.

Mask as of 10-18-16

Zoomed in cross section

Arrived from Output City (**Scale issue - 3 mm instead of 4 mm)

Fabrication Process

Thin film solar cell: Vertical Process with N or P type�So wafer

Lateral PN Junction: Lateral Process with Silicon-on-insulator

​

Minor mask issue - resolved

Discontinuity in ITO trace → no current

Can add polygon “bridges” around gap

P-N Junction Orientation

P-type

N-type

Gantt Chart

Created by Maria Pascale

.gan file uploaded on the Google Drive folder

Initial Fab Process as of 9/28

Efficiency Considerations

Comment on Yonggang’s comments (in the email thread) here

Updated Fab Process as of 10/10

Further Readings

• http://science.sciencemag.org/content/333/6044/838
• Wearable photovoltaics: http://drexel.edu/materials/news-events/archive/2015/March/capacitive-yarn/
• Wearable photovoltaics: http://nano.materials.drexel.edu/2013/11/nmg-entry-is-finalist-in-nsf-visualization-challenge/
• Solar cell background: http://electronicdesign.com/power-sources/what-s-difference-between-thin-film-and-crystalline-silicon-solar-panels
• Flexible carbon nano tube wire solar cell: http://onlinelibrary.wiley.com/doi/10.1002/anie.201003870/full
• http://journal.frontiersin.org/article/10.3389/fphy.2014.00077/full
• http://www.nature.com/articles/srep17810
• Stretchable solar cell http://onlinelibrary.wiley.com/doi/10.1002/adma.201004426/full
• Polymer solar cell http://www.nature.com/nphoton/journal/v6/n3/abs/nphoton.2012.11.html
• ITO annealing links: http://www.sciencedirect.com/science/article/pii/S0257897212002848 http://www.sciencedirect.com/science/article/pii/S0040609003015918

​

​

​

​