Institutional Archive of the Naval Postgraduate School

Calhoun: The NPS Institutional Archive DSpace Repository

Theses and Dissertations 1. Thesis and Dissertation Collection, all items


Power recovery of radiation damaged MOCVD grown indium phosphide on silicon solar cells through argon-ion laser annealing

Boyer, Lynn L., IV

Monterey, California. Naval Postgraduate School

This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.

Downloaded from NPS Archive: Calhoun

Calhoun is the Naval Postgraduate School's public access digital repository for th D U DLEY research mate rials and institutional publications c reated by the NPS community. «iit : 3 Calhoun is named for Professor of Mathematics Guy K. Calhoun, NPS's first IHN KN Ox appointed and published scholarly author. a ] LIBRARY Dudley Knox Library / Naval Postgraduate School 411 Dyer Road / 1 University Circle Monterey, California USA 93943





Lynn L. Boyer IV

June , 1996

Thesis Advisor: Sherif Michael



Approved for public release; distribution unlimited.



Public reporting burden for this collection of information is estimated to average | hour per response, including the time for reviewing instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate y Of any other aspect of this collection of information, including suggestions for reducing this burdes, to Washingtun Headquarters Seivices, Directorate for infu nation ] Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork

Reduction Project (0704-0188) Washington DC 20503.




6. AUTHOR(S) Boyer IV, Lynn L.


Naval Postgraduate School REPORT NUMBER Monterey CA_ 93943-5000


SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government.

12a. DISTRIBUTION/A VAILABILITY STATEMENT 12b. DISTRIBUTION CODE Approved for public release; distribution is unlimited.

13. ABSTRACT (maximum 200 words)

This thesis reports the results of a laser annealing technique used to remove defect sites from radiation damaged indium phosphide on silicon MOCVD grown solar cells. This involves the illumination of damaged solar cells with a continuous wave laser to produce a large forward-biased current. The InP/Si cells were irradiated with 1 MeV electrons to a given fluence, and tested for degradation. Light from an argon laser was used to illuminate four cells with an irradiance of 2.5 W/cm2, producing a current density 3 to 5 times larger than AMO conditions. Cells were annealed at 19°C with the laser and at 25°C under AMO conditions. Annealing under laser illumination of n/p-type cells resulted in recoveryof 48%. P/n type cells lost 4 to 12% of the assumed degradaton. Annealing under AMO conditions resulted in power recovery of 70% in n/p type cells. P/n-type cells recovered approximately 16% of lost power. Results indicate that significant power recovery results from the annealing of defects within n/p type InP/Si solar cells.

14. SUBJECT TERMS Indium Phosphide, Solar Cells, Annealing, Lasers 15. NUMBER OF | PAGES |



NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. 239-18 298-102

Approved for public release; distribution unlimited.


Lynn L. Boyer IV Lieutenant, United States Navy B.E., Vanderbilt University, 1989

Submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN ELECTRICAL ENGINEERING from the





This thesis reports the results of a laser annealing technique used to remove defect sites from radiation damaged indium phosphide on silicon MOCVD grown solar cells. This involves the illumination of damaged solar cells with a continuous wave laser to produce a large forward-biased current. The InP/Si cells were irradiated with 1 MeV electrons to a given fluence, and tested for degradation. Light from an argon laser was used to illuminate four cells with an irradiance of 2.5 W/cm2, producing a current density 3 to 5 times larger than AMO conditions. Cells were annealed at 19°C with the laser and at 25°C under AMO conditions. Annealing under laser illumination of n/p-type cells resulted in recoveryof 48%. P/n type cells lost 4 to 12% of the assumed degradaton. Annealing under AMO conditions resulted in power recovery of 70% in n/p type cells. P/n-type cells recovered approximately 16% of lost power. Results indicate that significant power recovery results from the annealing of defects within n/p type InP/Si

solar cells.


iL NER ODOC LG Ie! oye ee err l JU SGU CG 1B UO ce re 5 Aa SEMICONDUCTOR THEORY <9 92sec. od ee el enc: 5

B. SOLAR CELL DIODE CHARACTERISTICS ..................... 12

12 PAN unction-and Dittusion Current’ ...........0...0-.e.-e-; 2

2). \D4¢ pm O10 (3) 01 A 13

gm Diode Ware Curent crewws «> peewee. bya ee a ole 13


emis AIT OV ee myeree ol ey eek bee cs 55 51g. 0 arenes wo 15 POPECRANRESPONSE (2.2. suds sec ean ce We « Jk baw ewe 16 Balnotocenerdte@ CUMeN be ce a. os \s ie jonleen ss gas os ae meee 18

De SOLEAR CEBE PAKAMETERS (3. c:05¢..4.s0.820 0 esse ee ee es 21


B. RADIATION DAMAGE MECHANISMS _ .....................-- 29

1. Inelastic Collisions with Atomic Electrons ................... 29

2. Elastic Collisions with Atomic Nuclei ....................... 29

3. Inelastic Collisions with Atomic Nuclei ...................... 29


IV. ANNEALING AND LASER ILLUMINATION ......................... 35 Ay TEHPRIVIAIZANNEATINGS Se... ssa. ost en ts oe eee sn 35

B. MINORITY CARRIER INJECTION ANNEALING ................ 36 lRonvard Bias :CurmentInjectionen ge. 42s... tots e 37

27 PROCOINCCUONe Ane AMIAG eaern ane iia ne 2. eee tee 38

Ce ASRS ec a ce cos 2 I ee ns 39

VY INDIUM PHOSEMIDE (INE) SOLAR GELLS: jose... 5.5 2s 5s.n es scene. 41 (Se Udita Fett SS MOR Gr gency eee ie ge od Be ake io wD a ee Slee 41 EEN SIS ONG BIS ES ip niga. ites ow ois suas sas ale be cea eters oe 44 VIVPEXPERIMENTAL PROCEDURES AND RESULTS. ....-..6.600.0002-0% 45 AL EB OUIPVIEINMS Ei Werte. we aes os ate een ee ante das Sask yw se 45

1 SO Abs SIMUL mente rare eine ee Qa icp leew s wvueln o 48 2eVicAaSureMment VIClEIS: i. unc cree oe os eI es 49

Bimalhoisieahisats ahi Ovoyy aco) |. See eyan eases: tana oe ann ae sae 50

Peale SS 11) Cl ee cee A Oe fe kee wt ewig 50

|B CSRS Bi 24 Ly Uh Mc oh OM) en ne a2

KEE RIBS |0)) Cd ES eee 56, 3 oe RO an 55

IU CON CIOS IOIN Sarre ries Gude Be PEMA oa co's 59





Figure 2.1 Zincblende Crystalline Structure Common to Many Semiconductors

InchidiignPalRciesape |e ss TRO ee. cs le Figure 2.2 Simplified Energy Band Diagram for a Crystal. [After Ref. 24: p. 14] ....

Figure 2.3 Energy Bands Diagrams for a) an Insulator, b) a Semiconductor, and

eave onductor wAttcminel 4: paSloi| . sees. sat oe Mea the ee ba

Figure 2.4 Energy Bands with the Introduction of a) Donor Impurities and

By Acceplorampurities. [Atter Retoyape 29 |) ee. eee ee. Figure 2.5 Charge Depletion Region in p/n Junction. [Ref. 8: p.12] ............. Figure 2.6 Solar Cell Dark TV Curve for InP/Si n/p Cell #1 Used in This Research. ..

Figure 2.7 Incident Light on a Solar Cell Creates Electrons Hole Pairs, Which Are

Separated by Potential Barrier, Creating a Voltage That Drives a Current

dihrouch an Electrica: Circuit); Retel4p: 5m. > 2. eee sw

Figure 2.8 Solar Spectrum Above and Below the Earth’s Atmosphere and the

5900 K Black BodiGun i. ew.

Figure 2.9 Photongenerated current separated by Depletion Region. [Ref. 8: p.

DES Peer ess cnet en ce ata it As 02 (ns eee wee Figure 2.10 Cross Section and Top Views of a Solar Cell. [Ref. 12: p. 26] .......

Figure 2.11 Light and Dark IV Measurements for a n/p InP on Si Solar Cell. The

Licht TV Curve Pulls the Dark Down: [After Ref. 12: p. 32] ...:.......-. Figure 2.12 Normai Light IV Curve with Cell Parameters Shown. .............. Figure 3.1 Cross Section of the Earth’s Magnetosphere. [Ref. 22:p. 1424] ....... Figure 3.2 Inner and Outer Electron Radiation Belts. [Ref. 1: p.200] ...........

Figure 3.3 Proton Density Distribution in the Van Allen Belts. [Ref. 1: p 201] Figure 3.4 Crystal Damage Displacement: Closely Spaced Interstitials and

Wacancies, Or Pieter Palnsaiinereaap, soaker 16: p- 18l| ayes 3. .

Figure 3.5 Four Recombination/Generation Processes Affecting Solar Cell




Output: a) Electron/Hole Trapping, b) Recombination, c) Generation, and

d) Compensation. Rett 22a 4.0.23 os ee eee ee a) Figure 4.1 Recombination and Energy Release at the Defect Site [Ref. 24: p.36] ... 37 Figure 5.1 Normalized Efficiency of Illuminated InP, Dark InP, Si, and GaAs

Solar Cells [Ref..30ip: 43] ee... 223 eee ee eee, ee 43

Figure 6.1 Block Diagram of Test Equipment Including the Solar Simulator, Test

Cell, Reference Cell, Power Supply, Meters; and P@.3... ae 46 Figure 6.2 Experimental Setup Showing the Solar Simulator, Kiethley

Picoammeter, and the Cell Test Stand. . . see: (eee 47

Figure 6.3 Xenon Arc Lamp and the Sun’s Spectrum at AMO Conditions

[Ref. 3::p.2-6) «2onaa cw ete ne see ee ee 48 Figure 6.4 Solar Cell Biasing and Four-Point Measurement Diagram. ............ 49 Figure 6.5 Test Stand Showing Reference Cell and Test Cell, Micromanipulators

and Probes, Cooling Tubes and Temperature Sensors. ...............5-. 5) Figure 6.6 Argon-Ion Laser with Fiber-Optic Cable Attached. ................. 53 Figure 6.7 InP/Si Test Cell Under Argon-Ion Laser Illumination. ................ 54


Table 3.1 Solar Wind Characteristics at 1 AU from the Sun. [Ref. 2: p.3-1] ...... 26 Table 3.2 Electron-Hole Pair Generation Energies Required for Si, InP, GaAs. [Ref.

A at | rear ar os ed WE Seah « oes Se a RR ees 30 diable >sieinr Solar @ell General Characteristics, [Ref. 23:p.2] ................ 42 Table 6.1 Silicon Reference Cell AMO Output Parameters. .................... 49

Table 6.2 BOL Average Parameter Values for n/p and p/n Type InP/Si Solar Cells. . 52 Table 6.4 Percentage of Cell Degradation from BOL Baseline Due to a 1 MeV

Picceromniucnce O1o-x FONDS e-/CM20 22 hc ete ee hee eae cs es 56 Table 6.5 Percentage of Cell Parameter Recovery from the Degradated Level in

hip sis Olan ells ested in the solar Simulator, “22.2. - 222 2 oe. oe 56 Table 6.6 Assumed Cell Degradation From BOL Baseline Based on Similar Cells

Me steCaimmimerSOlareOlIMU at Ore yi. tisecrt sacs rac) cs eee «inicio eee ess Keres 57 Table 6.7 Cell Recovery for Laser illuminated Test Cells from an Assumed

We SradaviOnesase line ter prre rn saster ee tert case: aval gan, os ote tid eva on Suter Mere meee ee mee 58


I would like to thank my thesis advisor Sherif Michael for his assistance in making this thesis possible.

A very special thanks to Dr. Rob Walters of the Naval Research Laboratory Washington D.C. for providing the solar cells used in this research. His dedication, time and professionalism were crucial to the understanding of several difficult concepts.

A grateful thanks to Dan Sakoda and Ron Phelps of the Space System Engineering Lab for all their computer and thesis assistance.

To my wife Beth: thanks for putting up with all the long hours. You make it all

worth it.



Since the beginnings of the space program over four decades ago, solar cells have powered most of the spacecraft launched into earth orbit. Their high power to weight ratio and durability have made solar cells preferable for satellites electrical power systems. Solar cells have powered a variety of spacecraft, including communications, military, weather, and navigation satellites. Wide mission applications lead to a variety of orbits and differing environments. A major drawback to using solar cells for the electrical power system is radiation degradation in the space environment. Particularly harsh orbits can cause the solar array output power to significantly degrade over mission life. Typical array design compensates for these losses by adding extra cells to meet mission life requirements. Various ongoing research seeks to extend the solar array, and subsequently, spacecraft lifetime.

Silicon solar cells compose the bulk of on-orbit satellite solar panels. In recent years, Gallium Arsenide (GaAs) solar cells have replaced silicon as the preferred solar cell material in many applications. Specifically, superior radiation resistance characteristics of GaAs results in longer array and mission life in high radiation environments. Indium phosphide (InP) solar cells exhibit higher radiation tolerance than GaAs solar cells. This research investigates some radiation resistance properties of Metal Oxide Chemical Vapor Deposition (MOCVD) grown n/p and p/n InP solar cells.

For most solar powered spacecraft, mission life and average power requirements determine array size. Solar arrays must supply sufficient power throughout the life of the spacecraft and at End of Life (EOL) to allow the satellite to accomplish its mission. Solar cell on-orbit degradation requires array oversizing to meet EOL power requirements. Mission requirements determine power requirements and typically remain constant over mission life. Oversized arrays affect excess power at the Beginning of Life (BOL) ofa

satellite. Excess power dissipates as heat. Large oversized arrays can have a significant

impact on the spacecraft’s thermal design. Smaller arrays are more desirable [Ref 1: p. 394-396].

The solar array designer balances mass, area, cost, and risk to design an array. A long and proven history has made silicon arrays the least expensive and most reliable solar power system. Unfortunately, they are larger and less efficient than GaAs and InP arrays. The newer GaAs arrays are gradually replacing silicon as their proven flight time increases. InP arrays are extremely expensive and have minimal flight time. Thus, InP arrays carry a high degree of risk [Ref. 1: p. 396].

Spacecraft power requirements have gone from tens of watts at the beginning of the space program to hundreds of kilowatts. The space station Freedom is estimated to require 300 kW of electrical power before the end of its mission life. Solar array size corresponds directly to solar cell parameters such as efficiency and power output. As power requirements increase, array size must also increase. Larger arrays are costly and heavy. Launch configuration develops into a significant consideration as array size increases. When launching spacecraft, cost is directly proportional to weight. The larger and heavier a spacecraft, the more expensive to lift it into orbit. Smaller arrays are typically lighter and easier to configure for launch. Array configuration is a significant design consideration. Launch vehicle selection seriously influences the cost of a spacecraft mission. Interior space limitations on the launch vehicle can require folding the solar arrays many times. Depending upon the mission and expected degradation, array size can grow to 30%-40% of EOL required minimum size.

The NAVSTAR Global Positioning System (GPS) is in a 55° degree orbit at an altitude of 20,200 km [Ref. 20: p. 13][Ref. 27: p. 1-2]. Because this orbit is in the heart of the radiation belts, significant radiation degradation quickly results. A similar orbit of 20,372 km at 60° accumulates damage equivalent fluence from electrons of 5.26 x 10!3 e/cm2-yr and from protons 2.19 x 10!! e-/cm?-yr for a total fluence of 5.28 x 10!3 e/cm2-yr. For the silicon solar cells used in GPS, this translates to a 9%

degradation per year. Seven years in this environment reduces the array output to


approximately 78% of BOL capabilities [Ref. 20: p. 13-14][Ref. 28: p. 554].

Solar cells placed in the space environment will degrade. One possible solution to oversizing the arrays is on-orbit annealing. Annealing allows some power recovery in radiation-damaged solar cells. Lasers can be used to anneal solar cells under certain conditions. The potential exists for a laser to anneal an on-orbit satellite. If this could be accomplished practically, lower launch weights, simplified power conditioning, and longer mission lives could result [Ref. 13: p. 1].

Silicon solar cells are the most common type of solar cells used to power spacecraft. GaAs solar cells have also been used with a high degree of success to power spacecraft and often are superior to silicon due to their increased radiation resistance. InP solar cells, while in the experimental phase, provide better radiation resistance than GaAs and possibly better annealing characteristics; however, InP has high material cost and a brittle structure. Volume production would lower InP material costs, and manufacturing the InP solar cell on a silicon substrate would solve the brittleness problem. Silicon provides structural support and is less expensive to produce. This thesis examines some radiation and annealing characteristics of InP solar cells grown on a silicon substrate. Both n/p and p/n cell types are radiation damaged and annealed with an argon-ion laser.

Chapter II discusses the basic physics of solar cells. Chapter III describes the space radiation environment. Chapter I'V discusses solar cell annealing, power recovery, and the laser used in this research, while Chapter V describes the InP on silicon cells used in this research. Chapter VI explains the experimental setup, test plan, and results. Finally, Chapter VII presents the conclusions. Data plots for all solar cells can be found

in Appendix B.


This section presents some basic physics required to understand solar cell operation. A review of semiconductors, the p/n junction, and diode characteristics are presented. Silicon is used as an example material to illustrate some ideas. Next, various aspects of solar cell operation including bandgap energy, cell characteristics, current-

voltage plots, solar spectrum, and cell parameters are covered.


Semiconductors are solid crystalline materials characterized by electrical conductivity less than good conductors but greater than insulators. Silicon and germanium are both examples of semiconductors and have four valence, or outer shell, electrons. Group IV materials on the periodic table of elements are semiconductors. To fill their outer shell of electrons to the preferred number of eight electrons, semiconductor atoms will share their valence electrons with other atoms. When an atom shares its valence electrons with another atom, covalent bonds are formed. Silicon, for example, shares its outer four valence electrons with four other silicon atoms and forms a crystal structure.

During manufacturing, covalent bond formation between silicon atoms results in the development of a crystal structure. The potential required to remove a valence electron is much less than that required to remove any of the other electrons. Because of the low potential required, valence electrons can break the covalent bond and become free to move about the crystal. Light, thermal, or radiation sources can supply the required energy to break the covalent bond. At 300 K for example, approximately 1.5 x 10!° free

electrons are moving about one cubic centimeter of pure silicon. These free electrons are

called intrinsic electrons or intrinsic charge carriers. An examination of the thermal characteristics of pure semiconductors reveals an interesting material property: As the temperature increases, the number of intrinsic charge carriers increases, and the material becomes a better conductor. This contrasts with a good conductor such as copper, which becomes less conductive as the temperature increases.

Group IV elements are not the only elements that can behave as semiconductors. Group III and Group V elements can combine to give similar semiconductor properties. Indium, a Group III element has three electrons in its valence band, while phosphorus has five electrons in its valence band. InP forms a crystalline structure that satisfies the atoms need to fill the valence band to eight electrons. Most Group III-V compounds, including InP, form the Zincblende crystal lattice structure. Figure 2.1 shows the Zincblende crystal

structure and the bonds between the atoms.


Figure 2.1 Zincblende Crystalline Structure Common to Many Semiconductors Including InP. [Ref. 8: p. 7]

When an electron breaks its covalent bond, it leaves behind a hole in the crystal lattice structure. Holes are considered positively charged due to the absence of the negative electron which neutralizes the positive charge of the proton in the nucleus. Electrons wandering the crystal can recombine with holes to complete the covalent bonds. The process of an electron recombining with a hole is called recombination. Recombination is an important charge carrier loss mechanism and will be discussed later. Because the semiconductor material is electrically neutral, the number of holes equals the number of free electrons.

Electron band theory explains the behavior and theoretical operation of semiconductors. Electrons in a particular atom can only occupy discrete energy states. Energy levels vary from the ground state to higher ones called excited states [Ref 4: p. 813]. In a crystal, the electron of an atom is affected by the potential of all the atoms in the crystal. The potential interaction causes discrete energy levels to differ significantly from those of an isolated atom. Outer shell electrons can be considered in one of two possible states: the ground state for the outer electron shell is called the valence band, and the excited state is the conduction band. Electrons cannot exist in the forbidden energy gap Separating the two bands.

Most crystal structures have complex energy band structures [Ref. 6: p. 19]. For the purposes of this discussion, the simplified energy band diagram shown in Figure 2.2

will suffice.



Figure 2.2 Simplified Energy Band Diagram for a Crystal. [After Ref. 24: p. 14]

Electrons need to acquire a minimum amount of energy in order to cross the forbidden

energy gap, E.» from the valence band to the conduction band. Similarly, electrons give up

energy to drop down from the conduction band to the valence band. Light, thermal, or radiation sources can supply the required energy for electrons to transition between bands [Ref. 4: p. 813-815]. Figure 2.3 illustrates the relative energy levels between the

conduction valence bands for an insulator, semiconductor, and a conductor.

Band Conduction Band

ee es (OO Snr CREE : Jy y feces etal Cll Yy pants Vi Band Wy Yy Band YY

a) Insulator b) Semiconductor c) Conductor

Figure 2.3 Energy Bands Diagrams for a) an Insulator, b) a Semiconductor, and c) a Conductor. [After Ref. 4: p. 815]

In a conductor, the valence band and conduction bands are close together or overlapping. For an insulator, the valence band and conduction band are far apart, and a significant amount of energy is required for an electron to bridge the gap. Semiconductors lie between the two extremes of conductors and insulators.

Semiconductor crystals become p-type or n-type materials by adding impurity atoms, called dopants, to the crystalline structure. Adding pentavalent dopants (materials with five electrons in their outer shell) creates n-type material. Dopants that add electrons are called donors. Four of the dopant atom’s valence electrons form the covalent bonds of the crystal. The fifth electron is extra and loosely bound to the dopant atom. The extra electron requires little additional energy to transition to the conduction band and freely move about the crystal. The electron, while free to move around the crystal, leaves a

positive space charge at the stationary site of the dopant atom. Considered as a whole,

the semiconductor is electrically neutral. This method of doping reduces the energy required for donor electrons to move freely about the crystal. Figure 2.4 shows the

changes to the energy band structure for n-type and p-type material.

SY Conduction WS Band


Band Y

a) n-type b) p-type Figure 2.4 Energy Bands with the Introduction of a) Donor Impurities and b) Acceptor Impurities. [After Ref. 7: p. 29]

The addition of dopant atoms from a trivalent material with three valence electrons, creates p-type material. The three valence electrons of the dopant atoms form the covalent bonds of the crystal. Where there would normally be a fourth covalent bond, a hole exists. Free electrons in the crystal may combine with the hole and complete the covalent bond which also introduce a negative stationary charge. As with n-type material, the doped semiconductor is electrically neutral. Impurities added to create p-type material are called acceptors.

The addition of impurity atoms to create n-type or p-type material lessens the effect of the forbidden energy gap of an undoped semiconductor. Electrons in doped semiconductors have less potential to overcome the transition between the valence and conduction bands. Dopants add charge carriers to the semiconductor—an extremely

important effect. Charge carrier concentration significantly affects solar cell performance.

As mentioned above, semiconductor conductivity exhibits a proportional temperature dependency. As the temperature increases, the conductivity of a semiconductor increases. At any temperature above absolute zero, sufficient energy exists to break some covalent bonds and ionize electrons to the conduction band where they are free to move about the crystal. The absence of the electrons leaves a hole at the broken bond site. Electrons from neighboring covalent bonds may fill the hole, but in doing so leave a hole at their origin. Thus, holes can propagate though the crystal lattice. Electrons and holes have mobility associated with them. Random motion defines hole and electron mobility. Eventually, holes and electrons will meet and recombine. The recombination rate is proportional to the concentrations of free holes and free electrons. Electron and hole concentrations depend on the ionization rates. At thermal equilibrium, the ionization rate equals the recombination rate, and a constant concentration of charge carriers exists.

The dopant process as mentioned above, increases the concentration of one type of charge carrier. The creation of n-type material increases the concentration of electrons. Because electrons are more populous in n-type material, they are termed the majority charge carriers. Holes become the minority charge carrier. A similar reasoning for p-type material defines holes as the majority charge carriers and electrons as the minority charge carriers [Ref. 8: p. 9-11].

Small impurity concentrations are often used to dope silicon. At 300 K, one free electron exists per 10!2 Si atoms. A typical n-type donor concentration of 1 in 107 atoms increases the majority charge carrier population by a factor of 10° for silicon (10!2/107=105). This large increase shows how majority charge carrier concentration depends on the number of dopant atoms. The temperature contribution to majority charge carrier concentration is small. Minority charge concentration, however, remains a function

of temperature.



The absence of light causes solar cells to behave as diodes. The joining of p-type material with n-type material creates a p-n junction known as a diode. A diode becomes a solar cell by perforating the n-side or p-side metal contact and shining a light on the diode. The dark current-voltage (IV) measurement characterizes some basic physical parameters of a diode/solar cell. A solar cell dark IV measurement records the dark diode current as a function of diode terminal voltage.

1. P/N Junction and Diffusion Current

When the two types of semiconductor materials are brought together, electrons from the n-type material will diffuse to the p-type material where they recombine with holes in the p-type material. Similarly, holes from the p-type material diffuse to the n- type material and recombine with electrons. The diffusion of charge carriers across the junction creates a diffusion current. The diffusion of charge carriers leaves a charge depletion region at the junction. Figure 2.5 illustrates the charge depletion region of a p/n


p-Type Charge Depletion §_0-Type Acceptor Region Donor

S) Fixed Acceptor Ions = Free Electrons

) Fixed Donor Ions + Free Holes

Figure 2.5 Charge Depletion Region in p/n Junction. [Ref. 8: p. 12]

The charge depletion region creates an electrical potential in the opposite direction of the diffusion current. The diffusion current decreases as the electrical potential increases. The diffusion current defines an important parameter used in the complete dark diode equation presented later.

2. Drift Current

The drift current is another important constituent of the dark diode current. Minority charge carriers on both sides of the p/n junction will drift randomly throughout their respective crystal structures. Random n-type hole interaction with the depletion region electric field will cause the hole to quickly sweep across the junction. Electrons in the p-type material may interact with the depletion region electric field and quickly transition to the n-type side of the junction. The two currents added compose the drift current. The drift current acts in the opposite direction of the diffusion current. Unlike the diffusion current, the drift current is independent of the magnitude of the depletion region electrical field. Macroscopically, under open circuit conditions, the drift current is equal and opposite to the diffusion current [Ref 14: p. 7].

3. Diode Dark Current

The total dark diode current is a combination of diffusion current, recombination current, and shunt current. An equation for the p/n junction dark current is given in

Equation 2.1

qV Ge I(V)=I,,(e* -1) + emeeuccmiat ee NE egies ee: a e**T + cosh( = ) where I,,=q ae Qa)

and 14.= = —N,Vi,0M,

I(V) is the dark diode current as a function of terminal voltage, V. On the right side, the

first term is the diffusion current; the second term is the recombination current. The term

V/R,, represents the shunt current. Reference 12 contains a complete derivation of

Equation 2.1. Basically, the first term describes the current due to minority charge carrier drift across the junction. The second term adds the effect of recombination and generation of minority charge carriers due to different loss and gain mechanisms; the shunt current term models diode behavior at small biases. Dark IV curves can be approximated by manipulating the coefficients of Equation 2.1. Final coefficient values provide information about key cell performance parameters including diffusion current, recombination current, shunt resistance, and minority carrier lifetimes [Ref . 12: p. 23]. Equation 2.1 can be fitted

to dark IV data curves such as the one shown in Figure 2.6.

Spire InP/Si np cell #5803-331 8-1 BOL Dark IV’s

Current (A)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Voltage (V)

Figure 2.6 Solar Cell Dark IV Curve for InP/Si n/p Cell #1 Used in This Research.


Solar cells generate useful electric power when illuminated by a light source. This subsection discusses solar cell bandgap energy and the solar spectrum just outside the Earth’s atmosphere. An equation for the photogenerated current is also presented.

1. Bandgap Energy

Solar cells are solid states devices designed to convert sunlight into electrical energy through a process known as the photovoltaic effect. Solar cells are generally classified by material type such as silicon, GaAs, and InP. Each material has a unique bandgap energy associated with it.

Solar cell characteristics are driven from an important characteristic called the bandgap energy. As previously mentioned, the bandgap energy, E, defines the minimum amount of energy required to free an electron from the outer valence band of an atom. Freeing an electron from the outer valence band creates an electron-hole pair. Electrons in solar cells could receive the required bandgap energy if they are bombarded by photons

from sunlight. Photon energy relates to the frequency of light according to Equation 2.2 [Ref. 9: p. 6].

1.24 E,=hxv=hx (—)=—— (2.2) oe’

Where / x v is energy in electron volts, / is Plank’s constant, v is frequency of the light

wave, c is the speed of light, and A is the light wavelength in micrometers. If the energy of the incoming photon is greater than the bandgap energy, EE ? and the photon strikes a valence electron, the electron will absorb the photon energy. The added energy raises the electron from the valence to the conduction band. Once in the conduction band, the electron can move freely through the crystal lattice structure. When the electron leaves

the valence band, it leaves behind a positively charged hole. The absorption of the

photon energy creates an electron-hole pair. Multiple interactions between photons and electrons generate the charge carriers in the solar cell. Only photons with energy greater than the band gap energy will create an electron-hole pair. Photons with energy less than

EP are normally dissipated as heat. Photons with more energy than the bandgap energy

will generate an electron-hole pair, and the excess energy will be dissipated as heat.

Figure 2.7 shows a simplified diagram of a solar cell and the light photon interactions.

Light Generates Light Ils Absorbed Electron and Hole al Back Melal Contact

Electrical Retlected Contacts Top Electrical Grid

Metal Contact

© Electron © Hole

Figure 2.7 Incident Light on a Solar Cell Creates Electrons Hole Pairs, Which Are Separated by Potential Barrier, Creating a Voltage That Drives a Current Through an Electrical Circuit. [Ref. 14: p. 5]

The indium phosphide cells used in this research have a band gap energy of 1.34 eV at 300 K. From Equation 2.2, a bandgap of 1.34 eV equates to a light wavelength of 925 nm. Therefore, light with a wavelength less than or equal to 925 nm will provide enough energy to generate an electron-hole pair.

2. Spectral Response

Sunlight power density near the Earth and outside the Earth’s atmosphere, is

called Air Mass Zero, or AMO. AMO is a common standard used to measure the response of solar cells used in earth orbiting spacecraft. The Earth’s atmosphere has a large filtering effect on solar irradiance. Figure 2.8 shows solar irradiance above and below the Earth’s atmosphere. Significant absorption bands in the atmosphere attenuate portions of the solar spectrum. To test solar cells, solar simulators approximate AMO conditions in the laboratory. The sun’s power output at AMO closely resembles a 5900 K black body

curve as shown in Figure 2.8.










Figure 2.8 Solar Spectrum Above and Below the Earth’s Atmosphere and the 5900 K Black Body Curve.


3. Photogenerated Current

A solar cell is created by replacing the flat metal contact on one side of a diode with a metal grid [Ref. 12: p. 25]. The metal grid allows photons to pass through and strike the semiconductor. The solar cell type is determined by which semiconductor material is illuminated by sunlight. In an n/p type solar cell, the metal grid is on the n-side. Conversely, in a p/n type solar cell the metal grid is on the p-side. Incident photons cause the production of electron-hole pairs in the p-type and n-type material of a solar cell. Minority charge carner migration from the electron-hole pair produces the photogenerated current of the solar cell. Combining the photogenerated current with the dark diode current gives the total output current of the cell.

Photogenerated minority charge carners occurring 1n the depletion region will quickly be swept across the junction and produce useful current. Carners generated in the neutral portion of the cell will diffuse randomly through the cell. Diffusion ends when the carriers reach the cell surface, the junction, or recombine. Charge carriers reaching the junction compose most of the photogenerated current. Figure 2.9 shows a photon striking

a solar cell and the diffusion of the minority charge carrier to the junction.


p-Type Charg ap n-Type Acceptor Region Donor

“af crgacnanmnaneannar a “i-Fied O+ V ® I Figure 2.9 Photongenerated current separated by Depletion Region. [Ref. 8: p. 13]

ay “hi 7 . _ | ae 6 +> > ia ; x a , al ; ; al a | as 4 aqgrn =) " N i ails =i 6 i) , Si ree a lee ' oe e Cort v we ! Pit 1 lotta “I ity Ate aa.


j erable of wre (ewes wl? autcetinwgeress | >. ue eo ipod ie ra tatpitnup of Ganka Re iy si in tes 14 fiat Oelee ry et i

if oth) TL ()4eQ oa Paras ise var |

- oo]

Li oncaaiiog i ‘ply co gettd OURAOTy LIT Piguet ae sy > Gat tt Vee i. @ wo Ya

d= foo METTLE Mee Let @a7) mee ; =. ®.

i ; ee ee id} Gy O64 iWeels 44

rf if s¢2t \A iw


! 210 112% oot Gaee

Those minority charge carriers reaching the end of the cell are lost to surface effects. A back surface field added to the base reduces the loss of charge carriers to surface effects at the base interface. A back surface field is composed of a heavily doped material. This material creates an electric field that repels minority charge carriers from the base surface

toward the junction. Figure 2.10 shows a top and cross sectional view of an n/p solar cell.

Cross section view Top view

metal hu

fingers Bill metal bus bar YY VY

x =O n—region

X =X;

X = Xz

ore C) S wile es

metal fingers

back metal contact Figure 2.10 Cross Section and Top Views of a Solar Cell. [Ref. 12: p. 26]

The back surface field is shown in the cell cross section, while the metal contact grid is shown in the cell top view [Ref. 12: p. 26]. The addition of the photogenerated current term to the dark diode current,

Equation 2.1, results in Equation 2.3:

qv qv kde ee kT (cua 1) vi Ol is | ais | ing mw eR, (2.3) e*“? + cosh(—— )

Where I is the photogenerated current and the other terms are as previously described


[Ref. 12: p. 31]. The photogenerated current opposes the dark diode current and pulls the cell output current into the negative quadrant. Figure 2.11 shows a dark IV and light IV

measurement on a linear current scale.

Spire InP/Si np cell #5803-3318-1 Dark and Light IV’s

30——=—— Dark IV —— Light IV ee


Current (mA) f=)

Voltage (V) Figure 2.11 Light and Dark IV Measurements for a n/p InP on Si Solar Cell. The Light IV Curve Pulls the Dark Down. [After Ref. 12: p. 32]

In the photovoltaic community, the common convention is to place the light [TV curve in the first quadrant as shown in Figure 2.12. This is due to the fact that the direction of the generated current is opposite to the direction of the current in conventional electric

circuits employing diodes.

Spire InP/Si np cell #5803-331 8-1 Initial Light IV

z | Lght1




5 Voc = 0.689 V

O Isc = 25.86 A Pmax = 12.14 mW Eff = 9.0% FF = 0.68

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Voltage (V)

Figure 2.12 Normal Light IV Curve with Cell Parameters Shown.

D. SOLAR CELL PARAMETERS Solar cells are generally characterized by an illuminated current-voltage (IV) curve.

An IV curve for a solar cell is generated by varying the voltage across the terminals from

zero volts to the open circuit voltage (V,,). Examination of the resulting light IV curve


shown in Figure 2.12 reveals the following important cell parameters:

1. Open circuit voltage, V,,.: Voltage across the cell terminals when cell output

current is zero. 2. Short circuit current, I,,: current output of the cell when the voltage across the

cell terminals is zero.

3. Maximum Power, P__,.: The maximum output power of the cell. P__., is the

point on the IV curve where the corner of the largest rectangle that can be drawn

inside the curve intersects with the IV curve. gee occurs on the knee of the curve.

P jax 18 calculated by Equation 2.4: Pax Vu XI (2.4)

Where V,, and I, are the voltage and current at the maximum power point. To compute P jax eXperimentally, I and V are multiplied for each point along the IV curve and the resulting maximum value is P__.. [Ref. 12: p. 35].

Two other parameters are useful when discussing