Criteria for Choosing Transparent Conductors

52 MRS BULLETIN/AUGUST 2000

Introduction

Transparent, electrically conductive films

have been prepared from a wide variety of

materials. These include semiconducting

oxides of tin, indium, zinc, and cadmium,

and metals such as silver, gold, and titanium

nitride. In this article, the physical proper-

ties of these materials are reviewed and

compared.

A figure of merit for a transparent con-

ductor may be defined as the ratio of the

electrical conductivity to the optical ab-

sorption coefficient of the film. The mate-

rials having the highest figures of merit

are fluorine-doped zinc oxide and cad-

mium stannate. Physical, chemical, and

thermal durability; etchability; conduc-

tivity; plasma wavelength; work function;

thickness; deposition temperature; uni-

formity; toxicity; and cost are other factors

that may also influence the choice of

transparent conducting material for any

particular application.

Some Applications of

Transparent Conductors

Transparent conductors (TCs) have a

wide variety of uses. Their ability to reflect

thermal infrared heat is exploited to make

energy-conserving windows. These low-

emissivity (“low-e”) windows are the

largest area of current use for TCs. Oven

windows employ TCs to conserve energy

and to maintain an outside temperature

that makes them safe to touch. The electri-

cal conductivity of TCs is exploited in

front-surface electrodes for solar cells and

flat-panel displays (FPDs). Automatically

dimming rear-view mirrors for automobiles

and electrically controlled “smart” windows

incorporate a pair of TCs with an electro-

chromic (EC) material between them.

Electric current is passed through TCs to

defrost windows in vehicles and to keep

freezer display cases frost-free. TCs dissi-

pate static electricity from the windows on

xerographic copiers. Glass touch-control

panels are etched from TC layers. TCs can

also be formed into transparent electro-

magnetic shields, invisible security circuits

on windows, and transparent radio anten-

nas built into automobile windows.

It might appear reasonable to ask which

transparent conducting material is the best.

However, this question does not have a

unique answer, since different TCs are

best suited for different applications. Also,

a given application may constrain the

method of preparation and thereby affect

the choice of material. In the following, we

first summarize the methods for prepar-

ing TCs. Then, we consider the various

materials properties that can be important

in choosing a TC. Finally, we show how

these methods and properties lead to

choices of different TCs that are best for

different applications.

Processes Used in Making

Transparent Conducting Materials

The properties of a TC layer depend

not only on its chemical composition, but

also on the method used for its prepara-

tion. These preparative methods include

physical methods (sputtering, evapora-

tion, pulsed laser deposition) and chemi-

cal methods (chemical vapor deposition,

sol-gel, chemical bath deposition, electro-

plating). Some of the innovations in these

deposition methods are listed in Table I.

Spray pyrolysis was first used commer-

cially more than half a century ago to de-

posit conductive tin oxide films on heated

glass plates in batch processes. Since the

1980s, chemical vapor deposition (CVD) has

been widely adopted in the continuous

production of glass coated with fluorine-

doped tin oxide.

By far the majority of TC

films are currently produced in this way.

Most of this material is used for energy-

conserving “low-e” windows in buildings,

Criteria for Choosing

Transparent

Conductors

Roy G. Gordon

Table I: History of Processes for Making Transparent Conductors.

Materials and Process Reference

Ag by chemical-bath deposition Unknown Venetian

SnO

Sb by spray pyrolysis J.M. Mochel (Corning), 1947

SnO

Cl by spray pyrolysis H.A. McMaster (Libbey-Owens-Ford), 1947

SnO

F by spray pyrolysis W.O. Lytle and A.E. Junge (PPG), 1951

Sn by spray pyrolysis J.M. Mochel (Corning), 1951

Sn by sputtering L. Holland and G. Siddall, 1955

SnO

Sb by CVD H.F. Dates and J.K. Davis (Corning), 1967

SnO

by sputtering A.J. Nozik (American Cyanamid), 1974

SnO

by spray pyrolysis A.J. Nozik and G. Haacke (American Cyanamid), 1976

SnO

F by CVD R.G. Gordon (Harvard), 1979

TiN by CVD S.R. Kurtz and R.G. Gordon (Harvard), 1986

ZnOIn by spray pyrolysis S. Major et al. (Ind. Inst. Tech.), 1984

ZnOAl by sputtering T. Minami et al. (Kanazawa),1984

ZnOIn by sputtering S.N. Qiu et al. (McGill), 1987

ZnOB by CVD P.S. Vijayakumar et al. (Arco Solar), 1988

ZnOGa by sputtering B.H. Choi et al. (KAIST), 1990

ZnOF by CVD J. Hu and R.G. Gordon (Harvard), 1991

ZnOAl by CVD J. Hu and R.G. Gordon (Harvard), 1992

ZnOGa by CVD J. Hu and R.G. Gordon (Harvard), 1992

ZnOIn by CVD J. Hu and R.G. Gordon (Harvard), 1993

SnO

by sputtering H. Enoki et al. (Tohoku), 1992

ZnSnO

by sputtering T. Minami et al. (Kanazawa), 1994

SnO

by pulsed laser deposition J.M. McGraw et al. (Colorado School of Mines and

NREL), 1995

Criteria for Choosing Transparent Conductors

MRS BULLETIN/AUGUST 2000 53

with smaller amounts going into thin-film

photovoltaics and other applications men-

tioned in the first section.

Although indium tin oxide (In

Sn,

ITO) was first made by spray pyrolysis,

sputtering has become the preferred mode

for its production. ITO is mainly used

in FPDs.

Conductive zinc oxide films have been

investigated more recently. One application

in which zinc oxide is used is in photo-

voltaics. Because of its potential lower cost

and easier etchability, zinc oxide may re-

place ITO in display applications.

Materials Properties Relevant

to Transparent Conductors

A number of physical and chemical

properties are related to the performance

of a TC in any given application.

Optical and Electrical Performance

of Transparent Conductors

An effective TC should have high elec-

trical conductivity combined with low ab-

sorption of visible light. Thus an appropriate

quantitative measure of the performance

of TCs is the ratio of the electrical conduc-

tivity



to the visible absorption coeffi-

cient



(1)

in which R

is the sheet resistance in ohms

per square, T is the total visible transmis-

sion, and R is the total visible reflectance.

Thus





is a figure of merit for rating

TCs.

A larger value of





indicates bet-

ter performance of the TC.

Figures of merit for some TCs are given

in Table II. The values given are for the

best samples that we have prepared in our

laboratory by CVD at atmospheric pressure,

except for the indium oxide value, which

is the best that we have measured for a

commercially available film, and the cad-

mium stannate values, which we have taken

from the literature.

The results in Table II show that fluorine-

doped zinc oxide and cadmium stannate

have the best figures of merit of these TCs.

If the electrical and optical properties of a

TC were independent of film thickness,

then the figure of merit





would not de-

pend on film thickness, unlike other figures

of merit that have been proposed.

In fact,

“bulk” properties of TCs, such as



and



do depend somewhat on film thickness.

For example, they depend on crystallite

grain size, which usually increases with film

thickness. The figure of merit therefore

generally increases with film thickness. The

film thicknesses of the samples reported in

Table II were chosen to be typical of those





 R

lnT  R

1

needed for low-resistance applications

such as solar cells.

The results in Table II show that fluorine

doping gives superior performance com-

pared with metallic dopants, in both zinc

oxide and tin oxide. A theoretical under-

standing of this advantage of fluorine can

be obtained by considering that the con-

duction band of oxide semiconductors is

derived mainly from metal orbitals. If a

metal dopant is used, it is electrically ac-

tive when it substitutes for the primary

metal (such as zinc or tin). The conduction

band thus receives a strong perturbation

from each metal dopant, the scattering of

conduction electrons is enhanced, and the

mobility and conductivity are decreased. In

contrast, when fluorine substitutes for oxy-

gen, the electronic perturbation is largely

confined to the filled valence band, and

the scattering of conduction electrons is

minimized.

A theoretical upper limit to the figure of

merit may be estimated from the transport

theory of electrons in metals

given by

(2)

where



is the permittivity of free space,

c is the speed of light in vacuum, n is the

refractive index of the film, m* is the effec-

tive mass of the conduction electrons,



the mobility,



is a visible wavelength of

light, and e is the electronic charge. The re-

fractive index of TCs is close to 2.0 in the

visible region, thus the highest figure of

merit will be obtained from the material

with the highest product of mobility and

effective mass. For zinc oxide,

tin oxide,

and cadmium stannate,

m* is close to 0.3 m,

where m is the free-electron mass. Thus

most of the variation in the figure of merit

is due to differences in mobility. Note that

the free-electron concentration does not

enter into the figure of merit.





 4





nm*







2

The electron mobility is determined

by the electron-scattering mechanisms

that operate in the material. First of all,

some scattering mechanisms, such as scat-

tering of electrons by phonons, are present

in pure single crystals. In tin oxide

and

zinc oxide,

these scattering mechanisms

lead to mobilities of about 250 cm

1

at low doping levels, typically around

3

. Practical TCs need much higher

doping levels, usually 10

3

, in

order to operate at reasonable thicknesses.

For these high doping levels, scattering by

the ionized dopant atoms becomes an-

other important mechanism that alone

limits the mobility to less than about

90 cm

1

In the presence of both

these scattering mechanisms, the mobility

is limited to the value (250

1

 90

1

)

1



66 cm

1

. This maximum mobility is

lowered still further by other scattering

mechanisms such as grain-boundary scat-

tering, present in polycrystalline thin

films. The best TC films, ZnOF and

SnO

, have been prepared with mobili-

ties in the range of 50–60 cm

1

closely approaching the theoretical upper

limit.

Electrical Conductivity

In some applications of TCs, it is critical

that the TC be as thin as possible. For ex-

ample, in high-resolution displays, the re-

quired etched patterns in the TC create

height variations in the device. To keep the

topography as smooth as possible, the thin-

nest possible TC is desired. In this case,

the important material parameter is the

conductivity



. The conductivity increases

with the product of the concentration of

free electrons and the mobility. For metals

such as silver and titanium nitride, the free-

electron concentration is fixed by the

structure and electronic properties of the

solid. For wide-bandgap semiconductors,

Table II: Figures of Merit





for Some Transparent Conductors.

Visible

Absorption

Sheet Resistance Coefficient Figure of Merit

Material (/)



(

1

)

ZnOF 5 0.03 7

SnO

7.2 0.02 7

ZnOAl 3.8 0.05 5

Sn 6 0.04 4

SnO

F 8 0.04 3

ZnOGa 3 0.12 3

ZnOB 8 0.06 2

SnO

Sb 20 0.12 0.4

ZnOIn 20 0.20 0.2

Criteria for Choosing Transparent Conductors

54 MRS BULLETIN/AUGUST 2000

the free-electron concentration is deter-

mined by the maximum number of elec-

tronically active dopant atoms that can be

placed in the lattice. Attempts to place a

larger number of dopant atoms in the lat-

tice simply produce neutral defects that

decrease the mobility. The maximum ob-

tainable electron concentration and the

maximum conductivity in TCs generally are

found to increase in the following order:

ZnOF  SnO

F  ZnOAl  In

Sn 

TiN  Ag.

Plasma Frequency

The plasma frequency for the conduc-

tion electrons in a TC divides the optical

properties. At frequencies higher than the

plasma frequency, the electrons cannot re-

spond, and the material behaves as a trans-

parent dielectric. At frequencies below the

plasma frequency, the TC reflects and ab-

sorbs incident radiation. For most TC ma-

terials, the plasma frequency falls in the

near-infrared part of the spectrum, and

the visible region is in the higher, trans-

parent frequency range. The plasma fre-

quency increases approximately with the

square root of the conduction-electron

concentration. The maximum obtainable

electron concentration and the plasma fre-

quency of TCs generally increase in the

same order as the resistivity, as shown in

Table III. The corresponding plasma wave-

lengths decrease from about 2



m for

ZnOF to 0.7



m (red light) for TiN

(Table III).

Work Function

The work function of a TC is defined as

the minimum energy required to remove

an electron from the conduction band to

the vacuum. Some work functions meas-

ured for TCs are collected in Table IV.

Thermal Stability of

Transparent Conductors

TCs will generally increase in resistance

if heated to a high enough temperature for

a long enough time. For example, some of

the TCs were tested in my laboratory by

heating them in air for 10 min at succes-

sively higher temperatures. Table V gives,

as “stability temperatures,” the tempera-

ture range within which no increase of

10% in sheet resistance was noted.

In each case, the TC remained stable

to temperatures slightly above the opti-

mized deposition temperature. The high-

temperature stability of tin oxide films

allows coated glass to be reheated in order

to strengthen it by tempering. The thermal

stability of tin oxide films is currently

limited more by the softening of glass sub-

strates than by any thermal decomposition

of the SnO

F film.

Cadmium stannate is deposited by sput-

tering at low temperatures in amorphous

form. When annealed at high tempera-

tures, it crystallizes into a crystalline form

that is stable up to at least 1100C.

Minimum Deposition Temperature

When TCs are deposited onto a substrate,

the temperature of the substrate generally

must be maintained at a sufficiently high

temperature in order to develop the re-

quired properties in the TC. The required

temperatures usually increase in the fol-

lowing order: Ag or ITO  ZnO  SnO



SnO

. Thus, silver or ITO may be pre-

ferred for deposition on thermally sensitive

substrates, such as plastic, while cadmium

stannate requires very refractory substrates

to develop its best properties.

Diffusion Barriers between

Transparent Conductors

and Sodium-Containing

Glass Substrates

When TCs are deposited on sodium-

containing glass, such as soda-lime glass,

sodium can diffuse into the TC and increase

its resistance. This effect is particularly

noticeable for tin oxide, because sodium

diffuses rapidly at the high substrate tem-

peratures (often 550C) used for its depo-

sition. It is common to deposit a barrier

layer on the glass prior to the deposition

of tin oxide. Silica is most commonly used

as the barrier layer between soda-lime glass

and tin oxide, even though silica is only

partially effective in blocking the transport

of sodium. The silica layer usually serves a

second purpose, that of eliminating the in-

Table III: Approximate Minimum Resistivities and Plasma Wavelengths for

Some Transparent Conductors.

Resistivity Plasma Wavelength

Material (



 cm) (



Ag 1.6 0.4

TiN 20 0.7

Sn 100 1.0

SnO

130 1.3

ZnOAl 150 1.3

SnO

F 200 1.6

ZnOF 400 2.0

Table IV:Work Functions of Some Transparent Conductors.

Work Function Electron Concentration

Material (eV) (cm

3

)

ZnOF 4.2 2  10

ZnO 4.5 7  10

Sn 4.8 10

SnO

F 4.9 4  10

ZnSnO

5.3 6  10

Taken from Reference 33.

Table V:Thermal Stability of Some Transparent Conductors.

Deposition Stability

Temperature Temperature

Material (C) (C)

LPCVD ZnOB 200 250

APCVD ZnOF 450 500

APCVD SnO

F 650 700

Criteria for Choosing Transparent Conductors

MRS BULLETIN/AUGUST 2000 55

terference colors that would otherwise be

shown by the TC film.

Alumina is a much

more complete barrier against diffusion of

sodium.

Etching Patterns in TCs

For some applications of TCs, such as

displays, heaters, or antennas, parts of the

TC must be removed. Table VI lists some

chemicals that can be used to etch TCs.

Zinc oxide is the easiest material to etch,

tin oxide is the most difficult, and indium

oxide is intermediate in etching difficulty.

Series-connected thin-film solar cells also

need to remove TCs along patterns of lines.

This removal is usually carried out by laser

ablation.

Chemical Durability

The ability of a TC to withstand corrosive

chemical environments is inversely related

to its ease of etching. Tin oxide is the most

resistant, while zinc oxide is readily attacked

by acids or bases. Silver is tarnished by air

and moisture and can be used only in ap-

plications that are hermetically sealed.

Stability in Hydrogen Plasmas

In forming amorphous-silicon solar cells

on TC superstrates, the TC is exposed to a

plasma containing hydrogen atoms. These

plasma conditions rather easily reduce tin

oxide, causing an increase in the optical ab-

sorption by the tin oxide. Zinc oxide is much

more resistant to hydrogen-plasma reduc-

tion and may be preferred for applications

such as amorphous-silicon solar cells.

Mechanical Hardness of TCs

The mechanical durability of TCs is re-

lated to the hardness of the crystals from

which they are formed. Their hardness val-

ues may be ranked using the Mohs scale,

in which higher values represent harder

materials.

Table VII shows that titanium

nitride and tin oxide are even harder than

glass and can be used in applications that

are exposed to contact. Zinc oxide is readily

scratched, but can be handled with care.

Thin silver films are so fragile that they

cannot be touched and can be used only

when coated with protective layers.

Production Costs

The costs of producing a transparent

conducting material depend on the cost

of the raw materials and the processing

of it into a thin layer. The cost of the raw

materials generally increases in this order:

Cd  Zn  Ti  Sn  Ag  In. Indium is

a rare and expensive element that is ob-

tained as a byproduct of the mining of ores

for their content of other metals such as zinc

and lead. There are no “indium mines” be-

cause its concentration in minerals is too

low to allow economic extraction only for

the value of the indium. Thus the supply

of indium cannot be increased significantly

without a large increase in price sufficient

to make “indium mines” profitable.

The costs of the deposition methods

typically increase in the following order:

atmospheric-pressure CVD  vacuum

evaporation  magnetron sputtering 

low-pressure CVD  sol-gel  pulsed laser

deposition. This ranking was estimated by

considering the lowest-cost product made

by each process. The speed of the process

is very important in the cost. Atmospheric-

pressure CVD, vacuum evaporation, and

magnetron sputtering have high deposition

rates and have been scaled up to large areas.

Low-pressure CVD has higher equipment

costs than atmospheric-pressure CVD. Sol-

gel suffers from slow drying and reheating

steps, while pulsed laser deposition is only

suitable for small areas. This ranking can

give only a rough comparison of production

costs because many other factors enter into

a full economic analysis, including the pro-

duction volume and the tolerances for varia-

tions in the properties of the product.

Toxicity

Some of the elements used in TCs are

toxic. This increases the cost of processing

them because of the need to protect work-

ers and prevent the escape of toxic ma-

terials into the environment. Additional

encapsulation during use may be needed,

as well as a provision for recycling at the

end of the product’s lifetime. Toxicity of

the elements generally increases in the fol-

lowing order: Zn  Sn  In  Ag  Cd.

Cadmium compounds are carcinogens

and thus are heavily regulated and even

prohibited from some applications.

Choice of Transparent

Conducting Oxides

It is apparent from the diversity of ap-

plications for TCs that no one material is

most suitable for all uses. Depending on

which material property is of most impor-

tance, different choices are made. Table VIII

summarizes some of the most important

criteria that may influence the choice of a

transparent conducting material.

Examples of Applications and

the TCs Chosen for Them

We will now see how these criteria apply

to a number of applications in which TCs

are used.

Table VIII: Choice of Transparent Conductors.

Property Material

Highest transparency ZnO:F, Cd

SnO

Highest conductivity In

Sn

Lowest plasma frequency SnO

F, ZnOF

Highest plasma frequency Ag, TiN, In

Sn

Highest work function, best contact to p-Si SnO

F, ZnSnO

Lowest work function, best contact to n-Si ZnOF

Best thermal stability SnO

F, TiN, Cd

SnO

Best mechanical durability TiN, SnO

F

Best chemical durability SnO

F

Easiest to etch ZnOF, TiN

Best resistance to H plasmas ZnOF

Lowest deposition temperature In

Sn, ZnOB, Ag

Least toxic ZnOF, Sn O

F

Lowest cost SnO

F

Table VI: Etchants for Transparent

Conductors.

Material Etchant

ZnO Dilute acids

ZnO Ammonium chloride

TiN H

 NH

HCl  HNO

or FeCl

SnO

Zn  HCl

SnO

CrCl

Table VII: Hardness of Some

Transparent Conductors.

Material Mohs Hardness

TiN 9

SnO

6.5

Soda-lime glass 6

5

ZnO 4

Ag low

Criteria for Choosing Transparent Conductors

56 MRS BULLETIN/AUGUST 2000

Low-Emissivity Windows in Buildings

TCs on window glass improve the en-

ergy efficiency of the window because the

free electrons reflect infrared radiation for

wavelengths longer than the plasma wave-

length. The effect is similar to that of the

silver coating in a Thermos bottle. In cold

climates, the plasma wavelength should

be fairly long, about 2



m, so that most of

the solar spectrum is transmitted into heat

inside the building. Fluorine-doped tin

oxide is the best material for this purpose

because it combines a suitable plasma

wavelength with excellent durability and

low cost. Billions of square feet of TC-

coated window glass have been installed

in buildings around the world.

In hot climates, a short plasma wave-

length, 1



m, is desirable, so that the

near-infrared portion of incident sunlight

can be reflected out of the building. The

metals silver and titanium nitride have

sufficiently short plasma wavelengths for

this application. Silver is widely used for

this application despite its poor durability;

it is sealed inside double-glazed panes for

protection from air and moisture. Titanium

nitride is much more durable and can be

used on exposed surfaces, even on single-

glazed windows. The reflective gold color

of TiN-coated glass can frequently be seen

on large office buildings, but it is not popu-

lar for residential windows.

Solar Cells

The front surfaces of solar cells are cov-

ered by transparent electrodes. In single-

crystal silicon cells, a highly doped layer

of the silicon itself serves as the front elec-

trode. In thin-film cells, a TC layer serves

as the front electrode. Cadmium telluride

and some amorphous-silicon solar cells

are grown on a SnO

F-covered glass super-

strate. Thermal stability and low cost are

the primary factors in this choice. The high

work function of SnO

F is also helpful in

making low-resistance electrical contact to

the p-type amorphous-silicon layer. Other

amorphous-silicon cells are grown on flexi-

ble steel or plastic substrates; in this case,

the top TC must be deposited at low tem-

perature on thermally sensitive cells. ITO

or ZnO is chosen for this purpose because

both compounds can be deposited suc-

cessfully at low temperatures (typically

200C).

Flat-Panel Displays

The many different styles of FPDs all

use TCs as a front electrode. Etchability is

a very important consideration in forming

patterns in the TC electrode. The easier

etchability of ITO has favored its use over

tin oxide, which is more difficult to etch.

The low deposition temperature of ITO is

also a factor for color displays in which

the TC is deposited over thermally sensi-

tive organic dyes. Low resistance is another

factor favoring ITO in very finely patterned

displays, since the ITO layer can be made

very thin, thus the etched topography re-

mains fairly smooth. ZnO is lower in cost

and easier to etch than ITO is, so ZnO may

replace ITO in some future displays.

Electrochromic Mirrors and Windows

Automatically dimming rear-view mir-

rors are now installed in millions of auto-

mobiles. They include a pair of SnO



F-coated electrodes with an electrochemi-

cally active organic gel between them. The

main considerations are chemical inertness,

high transparency, and low cost. “Smart”

windows with electrically controllable

transmission are just entering the market-

place. Tin oxide appears to be the material

of choice, for the same reasons that it is

chosen for electrochromic mirrors.

Defrosting Windows

Freezers in supermarkets pass electric

current through TCs on their display win-

dows in order to prevent moisture in the

air from condensing on them and obscur-

ing the view. Low cost and durability are

the main factors that have led to the choice

of tin oxide for this application.

Defrosting windows in airplanes was

the first application of TCs, permitting high-

altitude bombing during World War II. The

discovery of TCs was kept secret until after

the war. Originally tin oxide was used, but

now ITO has replaced it in modern cock-

pits because its lower resistance permits

defrosting larger window areas with rela-

tively low voltage (24 V). Some automobile

windshields use silver or silver-copper

alloy TCs for electrical defrosting because

the 12-V systems in automobiles require

very low resistance, combined with the

legal requirement of a minimum transmis-

sion of 70%. The metal layers are protected

in the windshield by laminating them be-

tween two sheets of glass.

Oven Windows

Tin oxide coatings are placed on oven

windows to improve their safety by low-

ering the outside temperature of the glass

to safe levels. This permits the use of win-

dows even in self-cleaning ovens that reach

very high temperatures. The tin oxide coat-

ing also improves the energy efficiency of

the ovens. The main criteria for this choice

of material are high temperature stability,

chemical and mechanical durability, and

low cost.

Some transparent laboratory ovens are

constructed entirely of TC-coated glass,

which also serves as the electrical resistor

for heating the oven.

Static Dissipation

TCs are placed on glass to dissipate

static charges that can develop on xero-

graphic copiers, television tubes, and CRT

computer displays. Only relatively high

resistances (1 k/) are needed, so the

main concern is mechanical and chemical

durability. Tin oxide is the material of choice

for these applications.

Touch-Panel Controls

Touch-sensitive control panels, such as

those found on appliances, elevator con-

trols, and ATM screens, are formed from

etched TCs on glass. They sense the pres-

ence of a finger either by direct contact or

capacitively through the glass. The dura-

bility and low cost of tin oxide make it a

good choice for these applications.

Electromagnetic Shielding

It is apparently possible to eavesdrop

on computers and communications by de-

tecting electromagnetic signals passing

through windows. These stray signals can

be blocked by TCs with low sheet resis-

tance. Silver and ITO are the best materials

for this purpose.

Invisible Security Circuits

TC-coated glass can be used as part of

invisible security circuits for windows or

on glass over valuable works of art. Some

protection from fading by UV light is also

provided by the TC. Any TC (except for

colored TiN) could be used. Silver/ZnO

multilayers provide the best UV protection.

Improving the Durability of Glass

Some tin oxide coatings are used solely

to take advantage of tin oxide’s extraordi-

nary durability and have nothing to do

with its electrical conductivity. Tin oxide

coatings are used on the windows of bar-

code readers to improve their abrasion re-

sistance. Hydrofluoric acid etches glass,

but does not affect tin oxide. Some vandals

have used hydrogen-fluoride etching kits

(designed to etch identifying marks on auto-

mobile windows) to etch slogans on win-

dows. Tin oxide coatings are used to protect

windows from these attacks.

Conclusions

Transparent conductors have many

applications. There is no one TC that is

best for all applications. Fluorine-doped

tin oxide is the most widely used TC, while

tin-doped indium oxide (ITO) remains pre-

ferred for flat-panel displays. Zinc oxide

has potential for use in more efficient and

less expensive solar cells. All of these com-

Criteria for Choosing Transparent Conductors

MRS BULLETIN/AUGUST 2000 57

monly used transparent conducting mate-

rials and their production methods have

advantages and disadvantages that must

be carefully weighed for each new appli-

cation. The information in this article may

help in choosing the most appropriate trans-

parent conducting material for a new use.

Acknowledgments

This work was supported in part by the

National Renewable Energy Laboratory

under a subcontract for the U.S. Department

of Energy.

References

1. J.M. Mochel, U.S. Patent No. 2,564,706 (1947).

2. H.A. McMaster, U.S. Patent No. 2,429,420

(1947).

3. W.O. Lytle and A.E. Junge, U.S. Patent

No. 2,566,346 (1951).

4. J.M. Mochel, U.S. Patent No. 2,564,707 (1951).

5. L. Holland and G. Siddall, Vacuum III (1955).

6. H.F. Dates and J.K. Davis, U.S. Patent

No. 3,331,702 (1967).

7. A.J. Nozik, U.S. Patent No. 3,811,953 (1974).

8. A.J. Nozik and G. Haacke, U.S. Patent

No. 3,957,029 (1976).

9. R.G. Gordon, U.S. Patent No. 4,146,657

(1979).

10. S.R. Kurtz and R.G. Gordon, Thin Solid Films

139 (1986) p. 277.

11. S. Major, A. Banerjee, and K.L. Chopra, Thin

Solid Films 122 (1984) p. 31.

12. T. Minami, H. Nanto, and S. Takata, Jpn. J.

Appl. Phys., Part 2: Lett. 23 (1984) p. L280; T.

Minami, H. Sato, H. Nanto, and S. Takata, Jpn. J.

Appl. Phys., Part 2: Lett. 24 (1985) p. L781.

13. S.N. Qiu, C.X. Qiu, and I. Shih, Sol. Energy

Mater. 15 (1987) p. 261.

14. P.S. Vijayakumar, K.A. Blaker, R.D. Weiting,

B. Wong, A.T. Halani, and C. Park, U.S. Patent

No. 4,751,149 (1988).

15. B.H. Choi, H.B. Im, J.S. Song, and K.H. Yoon,

Thin Solid Films 193 (1990) p. 712.

16. J. Hu and R.G. Gordon, Sol. Cells 30 (1991)

p. 437.

17. J. Hu and R.G. Gordon, J. Appl. Phys. 71

(1992) p. 880.

18. J. Hu and R.G. Gordon, J. Appl. Phys. 72

(1992) p. 5381.

19. J. Hu and R.G. Gordon, in Microcrystalline

Semiconductors: Materials Science & Devices, ed-

ited by P.M Fauchet, C.C. Tsai, L.T. Canham,

I. Shimizu, and Y. Aoyagi (Mater. Res. Soc.

Symp. Proc. 283, Pittsburgh, 1993) p. 891.

20. H. Enoki, T. Nakayama, and J. Echigoya,

Phys. Status Solidi A 129 (1992) p. 181.

21. T. Minami, H. Sonohara, S. Takata, and H.

Sato, Jpn. J. Appl. Phys., Part 2: Lett. 33 (1994)

p. L1693.

22. J.M. McGraw, P.A. Parilla, D.L. Schulz,

J. Alleman, X. Wu, W.P. Mulligan, D.S. Ginley,

and T.J. Coutts, in Film Synthesis and Growth

Using Energetic Beams, edited by H.A. Atwater,

J.T. Dickinson, D.H. Lowndes, and A. Polman

(Mater. Res. Soc. Symp. Proc. 388, Pittsburgh,

1995) p. 51.

23. P.F. Gerhardinger and R.J. McCurdy, in Thin

Films for Photovoltaic and Related Device Applica-

tions, edited by D. Ginley, A. Catalano, H.W.

Schock, C. Eberspacher, T.M. Peterson, and T.

Wada (Mater. Res. Soc. Symp. Proc. 426, Pitts-

burgh, 1996) p. 399.

24. R.G. Gordon, in Thin Films for Photovoltaic

and Related Device Applications, edited by D.

Ginley, A. Catalano, H.W. Schock, C. Eberspacher,

T.M. Peterson, and T. Wada (Mater. Res. Soc.

Symp. Proc. 426, Pittsburgh, 1996) p. 419; a simi-

lar expression was proposed by V.K. Jain and

A.P. Kulshreshtha, Sol. Energy Mater. 4 (1981)

p. 151, without taking into account the reflection.

25. T.J. Coutts, X. Wu, W.P. Mulligan, and J.M.

Webb, J. Electron. Mater. 25 (1996) p. 935.

26. G. Haacke, J. Appl. Phys. 47 (9) (1976)

p. 4086.

27. For example, see S. Nudelman and S.S. Mitra,

Optical Properties of Solids, Chap. 3 (Plenum

Press, New York, 1969).

28. A.R. Hutson, J. Appl. Phys. 32 (1961) p. 2287.

29. K.J. Button, C.G. Fonstad, and W. Dreybrodt,

Phys. Rev. B 4 (1971) p. 4539.

30. W.P. Mulligan, PhD thesis, Colorado School

of Mines, 1997.

31. G. Bogner, J. Phys. Chem. Solids 19 (1961)

p. 235.

32. J.R. Bellingham, W.A. Phillips, and C.J.

Adkins, J. Mater. Sci. Lett. 11 (5) (1992) p. 263.

33. T. Minami (private communication).

34. M. Hassanein, J. Chem. U.A.R. 9 (1966) p. 275.

35. R.G. Gordon, U.S. Patent No. 4,187,336

(1980); U.S. Patent No. 4,419,386 (1983).

36. J.D. Chapple-Sokol, PhD thesis, Harvard

University, 1988.

37. H.N. Wanka, E. Lotter, and M.B. Shubert, in

Amorphous Silicon Technology–1994, edited by

E.A. Schiff, M. Hack, A. Madan, M. Powell, and

A. Matsuda (Mater. Res. Soc. Symp. Proc. 336,

Pittburgh, 1994) p. 657.

38. D.R. Lide, ed., Handbook of Chemistry and

Physics (CRC Press, Boca Raton, FL, 1999) p. 12.

■■

Circle No. 11 on Inside Back Cover

Advertisers in This Issue

Page No.

Chemat Technology, Inc. 27

EDAX, Inc. 36, 37

High Voltage Engineering Inside front cover

Huntington Mechanical

Laboratories, Inc. Outside back cover

INSPEC/IEE 103

MTS Systems Corp. 44

Structured Materials Industries, Inc. 57

Thermionics Vacuum Products 71

Carl Zeiss, Inc. 66

For free information about the products and services

offered in this issue, check www.mrs.org/publications/

bulletin/advertisers or fill out and fax the inside back

cover to 724-779-4397.