3.10

Glazings and Coatings

G Leftheriotis and P Yianoulis, University of Patras, Patras, Greece
© 2012 Elsevier Ltd. All rights reserved.

3.10.1
3.10.1.1
3.10.1.2
3.10.1.3
3.10.1.4
3.10.2
3.10.2.1
3.10.2.1.1
3.10.2.1.2
3.10.2.1.3
3.10.2.1.4
3.10.2.2
3.10.2.2.1
3.10.2.2.2
3.10.2.3
3.10.2.3.1
3.10.2.3.2
3.10.2.3.3
3.10.3
3.10.3.1
3.10.3.2
3.10.3.3
3.10.3.3.1
3.10.3.3.2

3.10.3.3.3
3.10.3.3.4
3.10.3.4
3.10.3.4.1
3.10.3.4.2
3.10.3.4.3
3.10.3.4.4
3.10.3.4.5
3.10.3.4.6
3.10.3.4.7
3.10.3.5
3.10.4
3.10.4.1
3.10.4.1.1
3.10.4.2
3.10.4.2.1
3.10.4.3
3.10.4.3.1
3.10.4.3.2
3.10.4.4
3.10.4.5
3.10.4.5.1
3.10.4.5.2
3.10.4.5.3
3.10.4.5.4
3.10.4.5.5
3.10.4.6
3.10.4.6.1
3.10.4.6.2
3.10.4.7


Introduction
Summary
Historical Development of Glass Manufacture
Modern Windows
Emerging Technologies
Thermal and Optical Properties of Glazing and Coatings
General Considerations
Solar irradiation
Optical properties of a glazing
Definitions of useful terms
Basic laws for solar and thermal radiation
Optical Analysis of Glazing and Coatings
Basic laws for the refraction and transmission of radiation
Combined absorption and reflection for total transmittance
Thermal Properties
Theoretical background
Practical considerations
Other useful terms
Low-Emittance Coatings
General Considerations
Solar Control Versus Thermal Insulation
Deposition Methods
Thermal evaporation
Electron beam gun evaporation
Sputtering
Chemical methods
Types of Coatings
Doped metal oxides
Coatings with metal layers

Use of interface layers
Application of a chemically and mechanically resistant top layer
Development of asymmetrical coatings
Development of Ag-based coatings resistant to high temperatures
Development of coatings with double Ag layers
Conclusions – Resumé
Glass and Windows
Float Glass
Manufacture
Toughened Glass
Manufacture and properties
Use of Glass in Solar Collectors
Light admittance
Weather protection and heat loss suppression
Windows in the Built Environment
Single-Glazed Windows
Clear single glazing
Tinted single glazing
Reflective single glazing
Low-emittance single glazing
Self-cleaning single glazing
Multiple-Glazed Windows
Double glazing
Triple and quadruple glazing for ultrahigh thermal insulation
Window Frames

Comprehensive Renewable Energy, Volume 3

doi:10.1016/B978-0-08-087872-0.00310-3


315

315

315

316

316

316

316

316

317

317

318

319

319

320

321


321

322

322

323

323

323

324

324

324

324

324

325

325

325

327


327

327

327

327

327

327

327

328

328

329

329

329

330

330

331


331

331

331

331

331

332

332

333

333


313


314

Components

3.10.4.7.1
Aluminum
3.10.4.7.2
Wood and wood composites

3.10.4.7.3
Plastics (vinyl, fiberglass, thermoplastics)
3.10.4.7.4
Hybrid
3.10.4.7.5
Effect of frames on the window thermal properties
3.10.4.8
Spacers and Sealants
3.10.4.9
Emerging Technologies
3.10.4.10
Conclusions – Epilogue
3.10.5
Evacuated Glazing
3.10.5.1
Operating Principles
3.10.5.2
Technology and Related Problems
3.10.5.3
The State of the Art
3.10.5.4
Comparison with Conventional Glazing
3.10.5.5
Electrochromic Evacuated Glazing
3.10.5.6
Conclusions
3.10.6
Transparent Insulation
3.10.6.1
Historical Background

3.10.6.2
Optical and Thermal Properties
3.10.6.3
Types of Available Materials
3.10.6.3.1
Granular aerogels
3.10.6.3.2
Monolithic silica aerogel
3.10.6.3.3
Glass capillary structures
3.10.6.4
Conclusions
3.10.7
Chromogenic Materials and Devices
3.10.7.1
Introduction
3.10.7.2
Electrochromics
3.10.7.3
Electrochromic Devices: Principles of Operation and Coloration Mechanisms
3.10.7.4
Materials for Electrochromic Devices
3.10.7.4.1
Transparent electrical conductors
3.10.7.4.2
Active electrochromic film
3.10.7.4.3
Ion storage and protective layers
3.10.7.4.4
Protective layers – magnesium fluoride

3.10.7.5
Performance of a Typical EC Device
3.10.7.6
Photoelectrochromics
3.10.7.7
Gasochromics
3.10.7.8
Thermochromics
3.10.7.9
Metal Hydride Switchable Mirrors
3.10.7.10
Other Switching Devices
3.10.7.10.1
Suspended particle devices
3.10.7.10.2
Polymer-dispersed liquid crystal devices
3.10.7.10.3
Micro-blinds
3.10.7.11
Conclusions – Epilogue
References
Further Reading
Relevant Websites

Glossary
Double, triple, or multiple glazing Glazing with two,
three, or more parallel glass sheets, placed at a short
distance, with the air gap between them hermetically sealed.
Electrochromic windows Windows that can change their
color and appearance with the application of an electrical

potential.
Evacuated glazing Double glazing with vacuum
established in the air gap between the two glass sheets, for
minimizing heat losses.

Float glass Glass produced by the ‘float’ method,
in which the molten glass literally floats on a
tin bath.
Low-e coatings Transparent thin films with
low emittance, usually deposited on the
surface of glass to reduce the emitted thermal
losses.
Transparent insulation Materials that
combine high thermal resistance and visible
transparency.

333

333

334

334

334

335

336


336

336

336

336

337

338

338

339

339

339

339

340

340

341

341


342

342

342

342

343

344

344

344

345

346

346

346

350

350

350


350

351

351

351

351

351

354

354



Glazings and Coatings

315

3.10.1 Introduction
3.10.1.1

Summary

Windows are key elements of a building as they play an important role in many of its functions: They allow the continuity of
indoor/outdoor space by visible light admittance (which is very important esthetically and psychologically). They play a significant
part in the energy balance of the building through ‘solar gains’ (desirable in winter and undesirable in summer) and thermal losses.

They contribute to the daylighting of rooms and present a shield from weather elements (rain, wind, dust, noise). By proper design,
the windows can perform all of the above functions. Glazing also plays a significant role in solar thermal collectors by admitting
solar radiation and by reducing thermal losses to the environment.
The most important breakthrough in the flat glass industry is undoubtedly the development of the float process. It has
revolutionized glass manufacturing and led to the production of high-quality windows. Nowadays, multiple glazing with high
visible transmittance and increased thermal insulation is the state of the art in the fenestration market. The incorporation of various
thin film coatings (such as low emissivity (low-e), reflective, self-cleaning) has added value to the glazing products. Emerging
technologies such as evacuated glazing (EG), aerogels, and chromogenics promise that in the years to come, new improved products
with even better properties will appear.
Nowadays, the technology of windows has advanced to such an extent that optimum performance windows are produced
commercially, each type tailored for a specific need. Different optimization criteria apply for windows depending on climate, use of
the building (residential/commercial), dimensions, and other characteristics. Thus, there is a multitude of solutions available in the
market, ranging from low-performance, inexpensive single glazing to highly insulated triple glazing; and furthermore, to
self-cleaning windows that remove dirt from their exterior surfaces, ‘smart’ electrochromic (EC) windows that alter their color on
demand, and so on. Emerging technologies such as EG, chromogenics, and aerogels promise that in the years to come, new
improved products with even better properties will appear.

3.10.1.2

Historical Development of Glass Manufacture

It is of interest to know that primitive windows were just holes in the walls. In the next step of the development, they were covered
with cloth, wood, or paper and then came the possibility to be closed or opened by the use of appropriate shutters. Later, windows
were built so that they could accomplish a double task: to transmit light and protect the inhabitants from the extreme environ­
mental conditions. Glass was used for this function. The Romans used glass as a material for windows in Alexandria in the second
century AD. They used cast glass windows (with poor optical properties.) for this purpose [1]
The word window appears for the first time in the early thirteenth century, and it was referring to an unglazed hole in a roof. In
English the word fenester was used in parallel until the eighteenth century. Today, to describe the array of windows within a facade,
we use the word fenestration.
Until the seventeenth century, window glass was cut from large disks of crown glass. Larger sheets of glass were made by blowing

large cylinders that were cut open and flattened, and then cut into panes. Most window glass in the early nineteenth century was
made using the cylinder method. The ‘cylinders’ were 2.0–2.5 m long and 250–350 mm in diameter, limiting the width that panes
of glass could be cut, and resulting in windows divided by transoms into rectangular panels. The first advances in automated glass
manufacturing were patented in 1848 by Henry Bessemer, an English engineer. His system produced a continuous ribbon of flat
glass by forming the ribbon between rollers. This was an expensive process, as the surfaces of the glass needed polishing. If the glass
could be set on a perfectly smooth body this would cut costs considerably. Attempts were made to form flat glass on a molten tin
bath, notably in the United States. Patents were awarded in 1902 and 1905 to H. Hill and H. Hitchcock [2], but this process was
unworkable. Before the development of float glass, larger sheets of plate glass were made by casting a large puddle of glass on an
iron surface, and then polishing both sides – a costly process. From the early 1920s, a continuous ribbon of plate glass was passed
through a lengthy series of in-line grinders and polishers, reducing glass losses and cost. Glass of lower quality, sheet glass, was made
by drawing upward a thin sheet from a pool of molten glass, held at the edges by rollers. As it cooled the rising sheet stiffened and
could then be cut. The two surfaces were not as smooth or uniform, and of lower quality than those of float glass. This process was in
use for many years after the development of float glass [3]. Between 1953 and 1957, Sir Alastair Pilkington and Kenneth Bickerstaff
of the UK’s Pilkington Brothers developed the first successful commercial application for forming a continuous ribbon of glass using
a molten tin bath on which the molten glass flows unhindered under the influence of gravity [4]. The success of this process lay in
the careful balance of the volume of glass fed onto the bath, where it was flattened by its own weight [4]. In January 1959, Pilkington
made public its new technology, which led to rapid growth in the production of high-quality glass. Full-scale profitable sales of float
glass were first achieved by Pilkington in 1960. In the Soviet Union, a two-stage molding method was developed in 1969 (USSR
Inventor’s Certificate nos. 230393 and 556593, US patent no. 4081260), and a float glass production line was put into service
manufacturing commercial products. In 1974, PPG Industries (the United States) patented its own method for float glass
production (US patent no. 3843346) [2].
The float method is the standard method for glass production nowadays: Over 90% of flat glass produced worldwide is float
glass. As of 2009, the world float glass market, not including China and Russia, is dominated by four companies: Asahi Glass, NSG/
Pilkington, Saint-Gobain, and Guardian Industries. Other companies include PPG, Central Glass, Hankuk, Visteon, and Cardinal
Glass Industries. The flat glass market is expected to reach 39 million tons by 2010 [3].


316

Components


3.10.1.3

Modern Windows

Modern windows became possible with the perfection of the industrial process for glassmaking and the deposition of appropriate
thin films on transparent surfaces leading to the use of low-e coatings. Low-e coatings are spectrally selective thin films that add
value to plain glass enabling it to perform multiple functions as part of fenestration systems: daylighting of buildings and at the
same time suppression of radiative heat losses. There are two broad categories of coatings: doped metal oxides and metal-based
stacks. The former are less expensive, they can be deposited on glass by spray pyrolysis immediately after it leaves the float line, and
they are better suited for thermal insulation, in cold climates. The latter comprises three to five thin film stacks, which require
advanced equipment for their production (such as sputtering in high vacuum) and accurate thickness control. They are more
expensive, but more versatile: they can be tailored on demand either for solar control or for thermal insulation. Recent advances in
the glazing industry (especially in the metal-based coatings field) have led to widespread production of low-e coatings for
fenestration, automotive, and architectural applications. Furthermore, these coatings exhibit electronic conductivity and are being
used as transparent conductors (TCs) in a multitude of devices, such as light-emitting diodes, displays, dye-sensitized and organic
solar cells, smart switchable windows, and gas sensors. This wide range of applications brings these films in the forefront of high
technology.

3.10.1.4

Emerging Technologies

In recent years, materials science and technology have gained a great impetus. New materials and devices with amazing properties
and functions are being developed. Research teams worldwide continuously come forward with new concepts. These advances
could have a significant impact in the architectural sector as they could bring about a new concept, the ‘dynamic building’, for
example, a building with the capacity to adapt itself to the prevailing weather conditions to save energy and to improve the
occupants’ comfort. Windows play a key role in the dynamic building concept, as they should be dynamic and reversibly altering
their optical properties on demand. To that end, a multitude of materials are being developed, under the collective name of
‘chromogenic materials’. Coming from the Greek ‘χρώμα’ (chroma) and ‘γεννώ’ (genno), their name implies that they ‘create color’.

Indeed, these materials, switch from a transparent state to a colored-absorptive one, or to a reflective-mirror-like one, under the
influence of electrical potential (electrochromics), heat (thermochromics), gases (gasochromics), or light (photochromics and
phototelectrochromics). Furthermore, there is a large variety of chromogenic material at different degrees of maturity. Some others
have found their way to the markets, while yet others are unlikely to ever leave the laboratory bench. Chromogenic devices are
believed to become the smart windows of tomorrow and to eventually dominate the fenestration market, much as float low-e glass
is a standard today. Their widespread use in buildings could improve living conditions of inhabitants, reduce the building energy
consumption – both for cooling and for artificial lighting – and have a positive environmental impact.

3.10.2 Thermal and Optical Properties of Glazing and Coatings
3.10.2.1

General Considerations

We give in this section the basic equations and definitions related to the thermal and optical properties of the solar radiation and the
general environment. Windows are used to permit the entrance of natural light into the buildings (daylighting) and at the same time
to allow visual contact with the outside environment. For these reasons, large windows create a pleasant feeling to the inhabitants.
On the other hand, we can have huge thermal losses through them in cold climates (winter case) and undesired heat gains in hot
climates, especially if they receive direct solar radiation (summer case). However, the solar heat gains are very welcome during
winter and for this reason the appropriate arrangement of windows is a basic element for the bioclimatic design of buildings. In
principle, walls can be insulated thermally very well, but the same is very difficult for windows as they must be transparent. Simple,
single-pane windows may exhibit, in some cases, about 10 times larger heat loses compared to a standard wall of the same surface
area. Advanced double-pane windows have only about 3 times the corresponding losses of a wall or even less. Special products have
been developed for this purpose as we describe them in detail in this chapter. We note here that usually transparent materials are
used, but for special uses both translucent and transparent materials can be used.

3.10.2.1.1

Solar irradiation

We start with some important considerations about the solar irradiation data, as they are needed for the study of optical and thermal

properties of glazing and coatings. For this purpose, we rely mainly on field-measured meteorological data or predictions from
well-known models that are capable for providing such data for the regions that have been modeled. Meteorological hourly data
exist for several locations of most countries. These files usually include solar irradiation, ambient temperature, relative humidity,
wind direction, and speed, and they are very useful for long-term energetic predictions.
The solar irradiation consists of two components: (1) direct and (2) diffuse irradiation. The direct irradiation component
(symbol Ib, from beam radiation) is the solar radiation coming directly from the sun to the point of observation without scattering
or absorption from the molecules and particles of the atmosphere. The diffuse irradiation is the irradiation received after it has been
scattered by these molecules and particles. For example, during a cloudy day the light consists mainly of diffuse irradiation. The
instruments used for the measurement of direct irradiation are called actinometers or pyrheliometers. They consist of a tube directed


Glazings and Coatings

317

toward the sun with collimators inside, which do not permit the diffuse rays into the instrument, and a black absorbing surface at
the bottom of the tube. The solar direct irradiation is absorbed at the base of the tube heating the instrument. Appropriate
thermocouples placed there give a signal in millivolts that is calibrated to give the accurate reading of the direct irradiation
(in W m−2). The apparatus includes a manual tracking mechanism for aiming at the sun. In addition, at the entrance of the tube,
there is a filter wheel so that spectral measurements can also be taken in various spectral regions.
The total irradiation (I or Itot) is measured using an instrument called a pyranometer. It usually consists of black (absorbing) and
white regions and multiple thermoelectric elements connected to them, all under a double glass dome, which is transparent to the
solar radiation. The output is calibrated to give the total solar irradiation. With the same instrument, we measure the diffuse
irradiation (Id) by placing a small disk (or a band) to shadow it at a distance before the pyranometer glass dome. However, if the
direct component Ib (the beam irradiation) is known from the measurement that we have described before using the pyrheliometer,
we can find the diffuse component Id indirectly from the following formula:
Id ¼ I − Ib

½1Š


It is obvious that these three quantities are related and it is usually preferable to measure I and Ib directly and get Id from eqn [1]. If
these quantities are measured on a horizontal plane, we use the corresponding symbols: Idh, Ih, and Ibh, where the index h stands for
horizontal.
When we consider a windowpane, or any other surface having any orientation in space, for many applications, we need to
measure or calculate the total solar irradiation (Ip) on that surface. Ip can be measured placing a pyranometer parallel to the surface.
It can be also calculated from eqn [2], assuming for the diffuse irradiation, in a first approximation, an isotropic distribution over
the sky and then calculating its contribution from the solid angle exposed to the sky dome (second term). In this equation, the first
term on the right-hand side gives the direct component on the plane we are examining. We add the reflected total radiation from the
ground as well (third term) to get the final result as given by [5]






cosθ
1 þ cos β
1 −cos β
þ Ih
ρg
þ Idh
½2Š
Ip ¼ Ibh
2
2
cos θz
In addition to the symbols introduced before, we use θ for the angle of incidence (angle between the beam irradiation and the
normal to the surface), θz for the zenith angle (angle between the vertical and the beam radiation), and ρg the albedo (reflectance of
the ground). The angle between the plane of the pane and the horizontal plane is β (tilt angle).
Equation [2] may be used to find the total radiation received by a surface (as a windowpane) when we have measured the

radiation components on the horizontal plane. Assume that the diffuse radiation has an isotropic distribution as an approximation.
Corrections have been provided for an even better approximation (e.g., see References 5 and 6]). They are based on the fact that the
diffuse solar radiation in an area around the direction of the sun is more intense (circumsolar). In general, we should take three
components for the diffuse radiation from the sky: isotropic, circumsolar, and horizon (coming from a belt near the horizon).

3.10.2.1.2

Optical properties of a glazing

The optical properties of a glazing depend on material properties and on the incidence angle of the irradiation on them. The
incidence angle of the direct irradiation is measured experimentally, but it can also be calculated for a certain place and time by
finding the position of the sun and, consequently, the direction of beam radiation in relation to the surface normal [5]. For the
diffuse radiation, we can proceed by an approximation as we have described before [5]. Modern technology allows the deposition
of thin film coatings on large areas of glass. Low-e coatings can be applied reducing heat loss problems as well as problems with
overheating because they also reflect in the far infrared (IR) radiation [7, 8]. For this reason large-area windows can be used now
in buildings.
The optical properties and the energy performance of a glazing are interrelated. Double-glazed units (DGUs) are common, while
triple-glazed units (TGUs) are rather uncommon, their use being restricted to very harsh environments. The surfaces and panes are
numbered starting with 1 for the outside surface of the outer pane (the surface facing the environment outside that belongs to the
first pane). Then 2 is the inner surface of the outer pane and so on. Also, it is common now to seal the glazed unit using spacers and
to create what is called an insulated glass unit (IGU). The handling of the window is more efficient in this way. In some cases the air
gap is filled with inert, low thermal conductivity, gases such as Ar or Kr. Advanced glazing, creating and maintaining vacuum in the
gap, has been proposed and prototypes have been studied, for extremely low heat transfer [9]. The unit is completed with a frame.
We focus mainly on the glazed part of windows in this chapter.

3.10.2.1.3

Definitions of useful terms

At this point it is useful to introduce some definitions that are commonly used for the optical and thermal properties of glazing and

coatings. We then give the equations for the dependence of various physical quantities on the angle of incidence.
Total solar transmittance (Tsol, expressed as a percent or a number between 0 and 1) is the ratio of the total solar energy in the solar
spectrum (wavelength 300–3000 nm of the solar spectrum) that is allowed to pass through a glazing, to the amount of total solar
energy falling on it. In other words, solar transmittance is the portion of total solar energy that is transmitted through the glazing.
Total solar reflectance (Rsol, expressed usually as a percent or a number between 0 and 1) is the ratio of the total solar energy that is
reflected outward by the glazing system to the amount of total solar energy falling on it. We should note that for windows with


318

Components

different films on the two sides the reflectance will depend on the side of the window surface exposed to the sun. In a similar manner
for DGU and TGU, it depends on the sequence of any existing films.
Total solar absorption or absorptance (Asol, expressed usually as a percent or a number between 0 and 1) is the ratio of the total
solar energy absorbed by a glazing system to the amount of total solar energy falling on it. The Greek letter α (lower case) is also used
as a symbol for the absorption and the same is valid for τ and ρ.
We should point out here that the solar transmittance and solar reflectance can be measured directly. It is usually then easier to
calculate solar absorption from the following basic equation, which is an expression of the energy conservation:
Tsol þ Rsol þ Asol ¼ 1

½3aŠ

Solar transmittance is one of the most important physical parameters as it gives the entry of solar energy through the glazing, or any
protective envelope in general. It affects the total heat transfer, but other factors are needed also to determine the total heat transfer.
Test methods exist that can give the value of transmittance in situ or ex situ. A measurement method of solar transmittance for various
materials can be devised by using the sun (or artificial sun) as the energy source, an enclosure, and a pyranometer. Sometimes, in
addition to transparent, we have to design methods that are appropriate for special cases such as for translucent, patterned, or
corrugated materials. Some methods can be applied at a small sample area, or others may give an average over a large area, as the
need arises. Methods also exist that are used to measure transmittance of glazing materials for various angles of incidence up to

nearly 80° relative to the normal incidence. However, some methods allow measurements of the solar transmittance only at
near-normal incidence.
Visible light transmittance (Tvis) is the ratio of the total visible solar energy (in the range 400–750 nm of the solar spectrum) that
can pass through a glazing system to the amount of total visible solar energy falling on the glazing system.
Visible light reflectance (Rvis) is the percent of total visible light reflected by a glazing system.
Visible light absorption reflectance (Avis) is the percent of total visible light absorbed by a glazing system. Again the sum of the
previous three quantities is 1.
Ultraviolet transmittance (TUV) It is the ratio of the total ultraviolet (UV) solar energy (range 300–400 nm of the solar spectrum)
that is allowed to pass through a glazing system to the amount of total UV solar energy falling on the glazing system.
There are practical reasons for the interest in the absorption of light in the UV region of the solar spectrum because it contributes
to the deterioration and fading of materials (as, e.g., fabrics and furnishing). Obviously we can define the other two quantities for
the UV part of the solar spectrum.
Luminous transmittance (Tlum) is defined as
Tlum ¼

∫ f ðλÞ T ðλÞ dλ
∫ f ðλÞ dλ

½3bŠ

with f(λ) being the relative sensitivity of the human eye in the photopic state (see Figure 1) and T(λ) the transmittance spectrum of a
glazing system. The luminous transmittance is in effect the visible transmittance weighed by the human eye sensitivity. It provides a
quantitative representation of the impression of a glazing system to our vision.
Similarly, we can also define the luminous reflectance (Rlum) and luminous absorption (Alum). Again, the sum of the previous three
quantities is 1.

3.10.2.1.4

Basic laws for solar and thermal radiation


The electromagnetic radiation (solar and thermal that is of interest here) is a flow of photons with energy
E¼hÂf ¼

hÂc
λ

where h is the Planck’s constant (6.626 Â10−34 Js), f the light frequency, λ the wavelength, and c the velocity of light.
The thermal emission of a black body at absolute temperature T is given by Planck’s law:


C1
1
Eλ ¼ 5 ⋅ C = λ T
−1
e 2
λ

½4aŠ

½4bŠ

the numerical values of the constants C1 and C2 being 3.742 Â 10−16 W m2 and 0.014 388 m °K, respectively [10].
Wien’s law gives the wavelength λmax for which the thermal emission has a maximum related to T by
λmax T ¼ 0:2897 ðcm KÞ

½4cŠ

Using this equation we find that the spectral distribution of thermal radiation emitted by bodies at or around ambient temperatures
has a maximum around 10 μm (long-wavelength radiation: symbol ℓ), while the solar spectrum at 0.55 μm (short-wavelength
radiation: symbol s). The solar spectrum is available from measurements with the attenuation caused by atmospheric absorption at

sea level or extraterrestrial from satellites without it. The equivalent temperature of the sun (more specifically of the photosphere
that emits most of the solar spectrum) is at 5900 K. In Figure 1, we show the spectral distribution of solar radiation after passing
through the atmosphere (measured in W m−2 per unit wavelength, left curve). On the right in the same figure, we show the curve for


Glazings and Coatings

50

Solar spectrum at sea level

45

Solar power density (W m–2 μm–1)

1.600

40

1.400

Sensitivity of the photopic vision (a.u.)

35

1.200

30
1.000


Blackbody radiation at 70 °C
25

800
20
600

15

400

10

200
0
0.10

Black body power density (W m–2 μm–1)

1.800

319

5

1.00

10.00

0

100.00

λ (μm)
Figure 1 Spectral distribution of solar radiation after passing through the atmosphere (in W m−2 per unit wavelength, left curve). On the right, we
show thermal radiation at a typical temperature of a body with ε = 1 (black body). Note the different scales on the two sides. The relative efficiency of the
human eye is also shown (in arbitrary units, red line) at the region near the maximum of the solar spectral distribution.

thermal radiation at a typical near ambient temperature of a body with ε = 1 (black body). We note the very different scales on the
two sides. The relative efficiency of the human eye is also shown (in arbitrary units, red line) at the region near the maximum of the
solar spectral distribution.
From Planck’s law we can also find the total power emitted per unit area by integration of the spectral distribution for all
wavelengths (Stefan–Boltzmann law):
∞

∫

E ¼ Eλ dλ¼σ T 4

½4dŠ

0

The constant σ = 5.6697 Â 10−8 W m−2 K−4 is called Stefan–Boltzmann constant.
The blackbody, for example, the perfect absorber and emitter, is an ideal concept. All real materials have a no zero reflectance and
as a result, they emit less radiation than a blackbody. To express this fact, eqn [4d] can be modified for the case of a gray body with
radiation properties independent of wavelength:
E ¼ εσ T 4

½4eŠ


with ε the body ‘emittance’ or ‘emissivity’, which represents the ratio of the electromagnetic radiation emitted by a surface to the
intensity that would be emitted by a black body at the same temperature (T). Emittance can be monochromatic (for a given
wavelength) or directional (at a given angle). In most practical situations, the ‘total hemispherical’ emittance is used by integration
over all wavelengths (total) and over all directions of the hemisphere enclosing the emitting surface (hemispherical).
The radiation exchange between two bodies (1 and 2) depends on their emittance (ε1 and ε2, respectively), their temperature
(T1 and T2, respectively), and their geometry. To express the latter, the configuration factor F1→2 is used. It is a geometrical factor
giving the fraction of radiation emitted by surface 1 that is intercepted by surface 2.
It can be derived easily that for two large parallel surfaces with S1 = S2 = S, we have that the configuration factor F1→2 = 1 and the
exchange of thermal radiation is [5]
À
Á
σ T24 −T14
Q
¼
½4f Š
S
ð1=ε1 þ 1=ε2 Þ−1

3.10.2.2
3.10.2.2.1

Optical Analysis of Glazing and Coatings
Basic laws for the refraction and transmission of radiation

To calculate the optical properties under a given angle of incidence θ1 of a ray of light on glass, we remind the basic equations for the
refraction of light [10]. Snell’s law gives the refraction angle θ2, going from air (medium 1) into medium 2 (glass). C1 and C2 are the


320


Components

values of the velocity of light in media 1 and 2, respectively. The index of refraction of glass relative to air (or vacuum) is n = 1.526 for
the most common type of glass. Symbols n2,1 and n20, n10 are also used for the relative and absolute indices of refraction as shown in
the following equations:


C1
n20
sin θ1
n¼
Snell’s law
n¼
¼
≡ n2 ; 1
½5Š
n10
C2
sin θ2
We note also that the velocity of light C in a material is connected with its index of refraction by the equation (Co is the velocity of
light in vacuum): C = Co/n = λ f.
The reflectivity r (or reflectance) of a surface is the ratio of the intensity of reflected light to that of incident nonpolarized light is
given by the well-known Fresnel equation [5, 6]:
!
1 sin2 ðθ2 − θ1 Þ
tan2 ðθ2 − θ1 Þ
1
Ir
þ
¼ ½rI þ rII Š

½6Š
rðθ1 Þ ¼ ¼
2 sin2 ðθ2 þ θ1 Þ tan2 ðθ2 þ θ1 Þ
2
Ii
For a given material (n known) and θ1 known, from eqn [5] we can find θ2, and from eqn [6] the reflectance r. In this equation, the
two terms in the bracket represent the reflectance for the perpendicular and parallel polarization, respectively. For example, for glass
(n = 1.526) and θ1 = 60°, we have from eqn [5] θ2 = 34.58° and from eqn [6] r (60°) = 0.093, that is, 9.3% of the light beam is
reflected. When θ1 = 0° (perpendicular incidence), θ2 = 0° and from eqn [6] we derive:
rð0Þ ¼

ðn −1Þ2
ðn þ 1Þ2

½7Š

The result then is r(0) = 0.043 (4.3% of the light beam is reflected). Naturally when light goes through a flat glass plate, it is reflected

on both surfaces as it is found from eqns [5] and [6].

If we add the intensity of all rays passing after multiple reflections (method of ray tracing) we find for the transmittance of the plate

for one component of polarization, normal to the plane of incidence:

τ 1 ¼ ð1 − r1 Þ 2

∞
X
n¼0


r12n ¼

ð1 − r1 Þ 2
1 −r1
¼
ð1 −r12 Þ
1 þ r1

The same is derived for the other polarization component (parallel) τ11 (1−r11)/(1+r11).
Then the total transmittance τr is


1 − r11
1 1 − r1
τr ¼
þ
2 1 þ r1 1 þ r11

½8Š

½9Š

For a system of N parallel plates, it can be proved in the same way (method of ray tracing) that the transparency for the combined
system is


1 −r11
1
1 −r1
þ

½10Š
τ rN ¼
2 1 þ ð2N −1Þr1 1 þ ð2N −1Þr11
We have assumed zero absorption up to now and we must consider it next. The system of N parallel plates is very useful in modeling
of the transparency of double and triple glazing.
The results from the last two equations for 0 < θ1 < 50° give transmittance that is almost constant, but for θ1 > 60° the
transmittance is decreasing at a fast rate with the angle of incidence [5].

3.10.2.2.2

Combined absorption and reflection for total transmittance

Now we turn to the combined problem. The absorption for the light traveling the path L/cos θ2, where L is the thickness of the plate
and θ2 the angle of diffraction, the intensity of radiation is found from the differential equation:
dIλ ðxÞ ¼ − Kλ Iλ ðxÞ dx
After integration, we get for the ratio of the transmitted radiation over the total incident radiation:


− KL
τ α ¼ exp
cos θ2

½11Š

½12Š

K is the coefficient of extinction of the material of the plate. We note that in eqn [11] we indicated the dependence on the
wavelength. In eqn [12], we consider a properly weighted value of K in the solar spectrum region, for example, if we consider solar
energy.
We can follow again the ray-tracing method and get a general equation. The exact result is not very useful here. Instead, we can

use the approximate final result [5]:
τ ≈ τr τα

½13Š


Glazings and Coatings

321

For the practical applications, as solar energy, the total transmittance is approximately equal to the product of the two transmit­
tances due to reflection and absorption separately.
At this point it is useful to recall Kirchhoff’s law:
εðλÞ ¼ αðλÞ

½14Š

τðλÞ þ αðλÞ þ ρðλÞ ¼ 1

½15Š

Also the consequence of energy conservation:

Similar equations follow for a weighted integration over some definite spectral distribution, as the solar irradiance, usually denoted by
the index sol, or simply s from the initial of short-wavelength radiation, to distinguish from long-wavelength (or thermal) radiation.
For the absorptance (the ratio of radiant energy absorbed by the pane to that incident upon it) we have
α ≈ 1− τ α

½16Š


ρ¼1 − α − τ ≈ τ α − τ r τ α

½17Š

And for the reflectance

Equations [13], [16], and [17] can be used for more than one plate. In this case, we should use the total thickness of the system for L
and also use eqn [10] instead of eqn [9] for τr.

3.10.2.3
3.10.2.3.1

Thermal Properties
Theoretical background

Our purpose is to establish the energy evaluation of windows and show how we can find the energy gains and losses. In our analysis,
we are using the well-known thermal resistance concept that simplifies calculations. It is based on the solution of the differential
equation for heat conduction:
ρc

∂T
¼ A þ k ∇2 T
∂t

½18Š

where A is the rate of heat production per unit volume in the material, at the point we consider, c the specific heat of the material,
and ρ its density. T is the absolute temperature and t the time. The coefficient of thermal conductivity k depends on the material,
being small for insulating materials and large for good heat conductors. It depends also on temperature, but for relatively small
variations of temperature and the kind of materials we consider it is constant to a very good approximation. The heat flow per unit

time and unit area (vector f) is given by Fourier’s law
f ¼ −kðgradTÞ ¼ −kð∇TÞ

½19Š

Now, if we solve eqn [18] for a problem and find T, then from eqn [19] we can find the heat flow. For the case of a flat wall, or
window, with A = 0 and heat flowing at the direction perpendicular to the wall plane, eqn [18], simplifies to the Laplace equation
∇2T = 0 and we find for the steady state the result:
q¼

T1 −T2
ððx2 −x1 Þ=kSÞ

½20Š

The numerator is the temperature difference between the two faces of the wall or window and is analogous to the voltage difference.
Also q is the current i in the completely analogous electrical problem. The quantity in the denominator is the thermal resistance
R = D/kS, where D = x1 − x2 is the wall (or window) thickness. If we consider the heat flow per unit time and unit area: f = |f| = q/S,
then the thermal resistance per unit area will be r = D/k.
For the thermal flow through a glazing we should, in addition to conduction, consider also radiation and, for the air or gas spaces
between multiple glass panes, convection. For a single glass, we have to consider the internal and external space heat transfer
coefficients hi and ho. Then according to the law of addition of heat resistances if D is the thickness of the glass and k its coefficient of
thermal conductivity, reminding also that the conductance h is the inverse of the corresponding resistance, we have for the total heat
transfer coefficient U, the equation:
1 D
1
þ þ
k
ho
hi


½21Š

1 D1 D2
1
1
þ
þ
þ
þ
hg ho
k1
k2
hi

½22Š

U −1 ¼
For DGU we have:
U −1 ¼

where hg is the gas conductance for the space between the two glass panes. For triple glass or more, the extension is obvious if we
apply the thermal resistance method.


322

Components

The values for hg, hi, and ho in these equations can be determined by using basic experimental or theoretical procedures that are

outside of the scope of this chapter. The interested reader can find details in ASHRAE [11]. It will be sufficient to state that all depend
on the emissivity of the corresponding surfaces involved for each of them and a number of other factors. For hi, which also depends
on the inside radiation and convection and is easier to find, we usually give a typical value: hi = 8 W m−2 K−1. The outside coefficient
may vary considerably because it depends on wind velocity and direction as well on the rest environmental conditions. We may take
the standardized value ho = 23 W m−2 K−1. The gap (gas or air) conductance hg depends on the temperature, the thermal conductivity
of the gas, density, specific heat, viscosity, and the width of the gas space. The inclination to the vertical also affects its value [12].

3.10.2.3.2

Practical considerations

The U-value (or factor) is the overall heat transfer coefficient for a glazing system. It is defined as the rate at which heat is transmitted
through it, per unit surface area per unit temperature difference between its two sides. It is measured in watts per square meter per
degree Kelvin (W m−2 K−1). The U-value is a function of the materials and the detailed construction of the glazing system. U-value
ratings for windows generally have values between 1 and 10 W m−2 K−1. The U-value of a window assembly is affected by the
physical properties of the frame, glass, thin film coatings, and spacers. The lower the U-value, the greater a window’s resistance to
heat flow, and the better is its insulating value. The symbol U (or Uw) is used for the value referring to the whole window, UC the
value at the center of glass, and UF (or Ufr) the value for the frame.
In some cases we may encounter the term R-value, which is the inverse of the U-value. The R-value is usually cited when
discussing wall and ceiling insulation values and rarely for windows and other fenestration products. The higher the R-value, the
better insulated is the wall or window, and it is more effective in keeping out the heat (and cold).
In practice, to facilitate thermal calculations for a window, we consider three zones: glazed, frame, and edge zone (Figure 2). The
edge zone is approximately taken to be about 6 cm wide [11, 13, 14].
Then an average value, 〈U〉, can be found from the following equation:
〈U〉 ¼ Ufr Ar ; fr þ Ued Ar ; ed þ Ugl Ar ; gl

½23Š

where fr stands for frame, ed for edge, gl for the center glazed area, and Ar,x the relative area of x to the total. For example
Ar,fr = Afr/Atotal. It is also common (in Europe) to use a linear heat flow coefficient Ψfr,gl for the edge zone so that the term in the

middle (edge) in the above equation is replaced by Ψfr,gl (Lfr,gl/Atotal), where Lfr,gl is the length of the borderline between the frame
and the edge. Obviously, the units for Ψfr,gl are W m−1 K−1 while those for U are W m−2 K−1.

3.10.2.3.3

Other useful terms

At this point it is useful to mention some other terms that are related to the energy performance of glazing.
Solar gain (or solar heat gain) (SHG) in general refers to the heat increase of a structure (or object) in a space that results from
absorbed solar radiation. Objects intercepting sunlight absorb the radiation and as a result their temperature is increased. Then, of
course, part of the heat is reradiated at far-IR wavelengths. If a glass pane (or other material) is placed between the solar irradiation
and the objects intercepting it, that is, transparent to the shorter wavelengths and not to the longer, then the solar irradiation has as
net result an increase in temperature (solar gain).
This is the general principle on which the greenhouse effect is based and has become well known in the context of global
warming. The amount of solar gain increases with increasing incoming irradiation from the sun and with the ability of the
intervening materials to transmit short-wavelength (solar) radiation and in part to absorb small fraction of it. It is useful to include
a low-emittance coating in order to reflect the long-wavelength (thermal) radiation back into the space protected by glazing. In passive
solar building design, for example, the aim is to maximize solar gain from the building in order to reduce space heating demand
(winter) and to control it in order to minimize cooling requirements (summer). The composition and coatings on glass for the
building glazing can be manipulated to optimize the greenhouse effect, while the pane size, position, and shading can be used to
optimize solar gain. Solar gain can also be transferred to the building by indirect or isolated solar gain systems. Objects having large
thermal capacity are used to smooth out the fluctuations during the day, and to some extent between days.

3

3
2
2
1


1

Figure 2 The areas used for the thermal analysis of a typical window: 1, glazing; 2, spacer; and 3, frame.


Glazings and Coatings

323

Solar heat gain coefficient (SHGC) is the fraction of incident solar irradiation admitted through a window, both directly
transmitted and absorbed, and subsequently released inward. SHGC is given as a number between 0 and 1. The lower a window’s
SHGC, the less solar heat it transmits in the protected space. SHGC is used in the United States.
g-Value is the coefficient commonly used in Europe while shading coefficient (SC) is an older term that is still sometimes used in
the United States. The relationship between SHGC and SC is as follows: SHGC = SC Â 0.87.
SC values are calculated using the sum of the primary solar transmittance and the secondary transmittance. Primary transmit­
tance is the fraction of solar radiation that directly enters a building through a window compared to the total solar insolation, the
amount of radiation that the window receives. The secondary transmittance is the fraction of heat flowing inside the space from the
window, compared to the total solar insolation.
Total solar energy rejected (%) is the percent of incident solar energy rejected by a glazing system. It is to the sum of the solar
reflectance and the part of solar absorption that is reradiated as thermal energy outward.
Shading coefficient (SC) is the ratio of the SHG through a given glazing system to the SHG under the same conditions for clear,
double-glass window. The SC defines the solar control capability of the glazing system and it is useful when discussing the
properties of fenestration and shading devices. In other words, the SC gives the solar energy transmittance through windows.

3.10.3 Low-Emittance Coatings
3.10.3.1

General Considerations

Low-e coatings are thin films that exhibit spectral selectivity: they are highly transparent in the visible (VIS) part of the electro­

magnetic spectrum (from 0.4 to 0.7 μm), highly reflective in the IR (for wavelengths higher than 0.7 μm), and absorbing in the UV
(e.g., below 0.4 μm).
Transparent materials with low-e properties can exist because of the following laws of physics: (1) the Stefan–Boltzmann
law and the Wien displacement law, which state that the heat exchange by radiation between surfaces is characterized by
their thermal emittance and that the maximum of emitted radiation from a body occurs at a specific wavelength, related to
its temperature. For materials at room temperature, this wavelength is about 10 μm away from the visible part of the
spectrum. (2) Drude theory and the Hagen–Rubens law, which state that the free carrier plasma in electrically conductive
materials has a cutout frequency below which all incoming electromagnetic radiation is rejected (e.g., reflected) and that the
IR reflectivity is directly related to the electrical conductivity of the material. Thus, it is possible to decouple the visible light
spectrum from that of thermal radiation and to have surfaces with properties being entirely different with regard to thermal
and visible radiation. Furthermore, it becomes clear that for a film to exhibit low-e properties it is necessary to possess
electronic conductivity.
Low-e coatings were first envisaged for use in transparent heat-insulating glazing. Although they were already known
from the 1960s, the main thrust in their development took place after the petroleum crisis in 1974, as is the case for most
of the renewable energy materials. In the 1980s and 1990s, low-e glass products dominated the markets. Nowadays, use of
low-e glass in architecture is very common, and in many countries it is mandatory by law to increase the energy efficiency
of buildings, to promote rational use of energy, and to reduce CO2 emissions. The most prominent companies in this field
are Pilkington, PPG, Saint-Gobain, AGC (former Glaverbel), Nippon Sheet Glass Co. (NSG), and Guardian. Links to their
Internet sites appear in the webpage list.
Apart from their use in buildings, these coatings have a number of diverse applications, emerging from their electrical
conductivity, which enables them to serve as TCs in a multitude of devices such as dye-sensitized and organic solar cells, smart
switchable windows, gas sensors, light-emitting diodes, and displays. This wide range of applications brings these thin films in the
forefront of high technology.

3.10.3.2

Solar Control Versus Thermal Insulation

In Figure 3 appear the transmittance and reflectance spectra of plain glass and of a typical low-e coating. The intensity of the solar
radiation on the Earth’s surface is also presented for comparison. As can be seen in that figure, the transmittance of plain glass is high

throughout the visible part of the spectrum, up to 2 μm, well into the IR region, apart from a dip at about 1 μm, related to the
amount of Fe2O3 that is present in the glass [15]. It is also evident that in the thermal radiation region (around 10 μm) both
reflectance and transmittance of plain glass remain low, and by Kirchhoff’s law (i.e., α + ρ + τ = 1, ε = α in thermal equilibrium) the
absorptance (and consequently the emittance) of glass is considerable. On the other hand, the transmittance of a typical low-e
coating follows closely the solar spectrum in the visible and diminishes in the IR. The coating reflectance has the opposite behavior:
it remains low in the visible and increases in the IR. By Kirchhoff’s law, at 10 μm, the absorptance (and emittance) of the coating is
low, as it exhibits high reflectance. Thus, the low-e coating acts as a spectrally selective filter, which allows passage of the visible (and
possibly some of the near IR) and rejects the far IR. Depending on the wavelength (λΤ) at which transition from high transmittance
to high reflectance occurs (e.g., when τ ≈ 50% and ρ ≈ 50%), we can distinguish two broad categories of low-e coatings: (1) those
intended for solar control, with λΤ ≈ 0.7 μm, which reject the solar IR spectrum and (2) those intended for thermal insulation, with
λΤ ≈ 2 μm, which transmit the solar IR. The former are suitable for warm climates, the latter being suitable for cold climates in which
solar gains are desirable.


324

Components

100
90

T, R (%)

80
70

Solar radiation of Earth

60


Transmittance of plain glass
Reflectance of plain glass

50

Transmittance of a low-e coated glass
40

Reflectance of a low-e coated glass

30
20
10
0

100

1 000
Wavelength (nm)


10 000


Figure 3 Optical properties of plain glass and a typical low-emissivity (low-e) coating.Adapted from Schaefer C, Brauer G, and Szczyrbowski J (1997)
Low emissivity coatings on architectural glass. Surface and Coatings Technology 93: 37–45, with permission.

3.10.3.3

Deposition Methods


Thin film deposition and optical design of coatings constitute a vast technological field [16–18], and here only the most common
methods are presented.

3.10.3.3.1

Thermal evaporation

The raw material of the film is placed in a crucible and heated in vacuum so that a vapor transfers material to an adjacent surface (the
substrate) at a sufficient rate [19].

3.10.3.3.2

Electron beam gun evaporation

Instead of thermally heating the raw material, which can be ineffective in cases of compounds with low thermal conductivity (as are
dielectrics), an electron beam springing from a metal at high temperature can be deflected into the crucible by a magnetic field [20].
Momentum transfer from the incident electrons heats the material. This method is versatile as the electron beam can be
manipulated (in a way similar to that of a cathode ray tube) to heat uniformly all the material in the crucible. The beam intensity
can also be altered in order to maintain constant deposition rates. Electron guns with multiple crucibles are available for the
sequential deposition of different materials within the same vacuum chamber, without interrupting the vacuum. To improve the
stoichiometry of compounds, reactive evaporation can be used, with the presence of a gas (usually O2) in the vacuum chamber that
is incorporated in the film structure. To enhance the film packing density, ion-assisted evaporation can be used: The substrate is
bombarded by energetic ions of an inert gas (usually Ar) during deposition and the resulting film becomes more compact. The
e-gun method is widely used in the optics industry for the production of antireflection coatings, optical filters, etc.

3.10.3.3.3

Sputtering


In this method, a plasma of inert and/or reactive gases (such as Ar, O2) is created in a low pressure, and energetic ions in the plasma
dislodge chunks of the raw material from a solid plate (known as the target). These chunks are deposited on the substrate [21, 22].
Depending on the gas discharge method used to create the plasma, one can distinguish between radio frequency (RF), medium
frequency (MF), or direct current (DC) sputtering. Furthermore, magnetrons are used to increase the efficiency of the electrons in
ionizing the Ar atoms by trapping the electrons near the target, and thus we have Twin-mag, pulsed magnetron sputtering, etc. The
sputtering method is applied for the deposition of industrial metal-based low-e coatings on large areas.
Evaporation and sputtering are also known as physical vapor deposition (PVD) methods.

3.10.3.3.4

Chemical methods

In order to avoid vacuum that entails expensive (and sensitive) equipment, chemical deposition methods have been developed. A
widespread and relatively simple method is the sol–gel deposition that involves immersion of a substrate in a chemical solution,
withdrawal at a controlled rate, and subsequent heat treatment [23]. Alternatively, the chemical solution can be applied by spray or
by spin coating. In the latter, the substrate is rotating and as a result, the chemical solution is spread evenly on its surface. Chemical
vapor deposition (CVD) uses heat to decompose a vapor of a precursor chemical to make a thin film of a desired composition
[24, 25]. A variation of the CVD technique is called spray pyrolysis; a fluid containing the precursor is then sprayed onto a hot


Glazings and Coatings

325

substrate. This method is used on a large scale for deposition of tin oxide-based films on hot glass as it comes out from the float glass
production and is transferred to the cooling stage. Electrochemical techniques include cathodic electrodeposition from a chemical
solution and anodic conversion of a metallic surface – especially of Al – to form a porous oxide. A disadvantage of the chemical
methods is the inevitable presence of traces in the resulting films of the precursors used, due to limitations in the heat treatment
phase. These traces may impede the film performance or reduce their service life by unwanted side reactions. Furthermore,
environmental and health hazards from the compounds involved can be a concern.


3.10.3.4

Types of Coatings

Two groups of materials are of particular interest for use as low-e coatings: (1) doped metal oxides and (2) film combinations that
incorporate metal layers. The typical thickness of the first group of films is on the order of 10−1 μm. They are hard, compact, strongly
adherent to glass, chemically inert, their luminous and near-IR absorbance can be low, and their thickness does not affect electrical
resistance. The thickness of the second group of films is on the order of 10−2 μm and they are soft, porous, poorly adherent to glass
substrates, and chemically reactive. In this group of films, the electrical resistance is strongly thickness-dependent.

3.10.3.4.1

Doped metal oxides

Low-e coatings based on doped metal oxides comprise a host lattice (usually In2O3, SnO2, or ZnO) that is doped by metal or
halide atoms. The most common representatives of this group of films are tin-doped indium oxide (In2O3:Sn, usually called
ITO), fluorine-doped tin oxide (SnO2:F, usually called TFO), and gallium-doped zinc oxide (ZnO:Ga, usually called GZO).
Doping is accomplished either by adding a higher-valence metal, by replacing some oxygen with fluorine, or by oxygen
vacancies. The compounds mentioned above have wide enough band gaps to allow considerable transmission in the visible
and doping is feasible to a level high enough to render the materials IR reflecting and electrically conducting. In doped metal
oxides, the degree of doping determines the position of the transition wavelength (λΤ): the higher the doping, the more
metal-like the films are and λΤ appears at lower wavelengths. However, in these films, doping cannot be brought to a
sufficiently high level required for solar control, their λΤ usually lies in the near IR and, thus, doped metal oxide coatings are
used mostly for thermal insulation. The main advantage of the doped metal oxides, compared to metal-based films, is the
chemical and mechanical stability, which allows their use on glass surfaces exposed to ambient conditions. This is why these
films are referred to as ‘hard coatings’.
The development of these materials has reached maturity, and numerous commercial products are available in the market for
windows and other architectural applications. Most of these films are prepared by spray pyrolysis and their typical thickness is on
the order of 10−1 μm. Typical properties of such coatings are as follows: Tlum ≈ 90%, Tsol ≈ 75%, ε ≈ 0.20.

Although doped metal oxide coatings are well established, intense research activity continues in the field. The research effort is
directed toward the development of alternative host materials and dopants, as well as multiphase mixtures of known materials, in
order to improve various properties of the coatings, such as electrical conductivity, optical transmission, hardness, and adherence.
[15]. Some of the combinations reported in the literature are the following [15]: ITO:ZnO, ITO:Ti, In2O3–ZnO (IZO), In2O3:Ti,
In2O3:Mo, In2O3:Ga, In2O3:W, In2O3:Zr, In2O3:Nb, In2-2xMxSnxO3 with M being Zn or Cu, ZnO:(Al,F), ZnO:B, ZnO(Ga,B),
Zn0.9Mg0.1O:Ga, and many others.

3.10.3.4.2

Coatings with metal layers

In this type of coatings, the highly reflective metal film (that would otherwise be opaque in the visible) is sandwiched between two
dielectric layers that have an antireflective effect: with an appropriate index of refraction and thickness, the light beams reflected on
the front and back surfaces of each of the dielectric layers are of opposite phase and of nearly equal amplitude. Thus, they interfere
destructively and as a result, the film reflectivity is diminished. One is then led to dielectric/metal/dielectric (D/M/D hereafter)
multilayers. Dielectrics with high refractive indices, usually metal oxides, such as TiO2, ZnO, ZnS, SnO2, Bi2O3, and In2O3, are
suitable. Appropriate metals are the so-called ‘noble’ ones, such as Ag, Au, Cu, and Al (given in order of decreasing performance). Of
all these metals, Ag is the most suitable, due to its low absorption in the visible. Coatings based on Au or Cu have inferior optical
properties and a characteristic golden brown color.
To achieve high transmittance, the metal layer needs to be as thin as possible. The growth mechanism of metal layers on glass
[26–28] imposes the limit: in the initial stages of their development on glass and other dielectric materials, metal films form tiny
nuclei. With material continuously added on the substrate, these nuclei grow via surface diffusion and direct impingement, into
islands that are discontinuous and possess a fractal nature. Further thickening of the metal film leads to large-scale coalescence and
to continuous films. The coalescence thickness is about 15 nm for thermally evaporated Ag films [26, 27] and can be reduced to
about 9 nm using other methods such as sputtering and ion-assisted deposition [27].
In these stacks, the metal layer thickness governs the coating properties. With the metal thickness increasing, the coating
electronic conductivity increases, and its thermal emittance and luminous transmittance decrease. In Figure 4 appears the measured
emittance of evaporated ZnS/Ag/ZnS coatings versus Ag layer thickness [8]. An abrupt decrease of emittance with increasing Ag
thickness can be observed below coalescence, in the range 10–15 nm. For continuous Ag films, thicker than 15 nm, the values of
emissivity continue to decrease with increasing thickness, but at a lower rate. The figure of merit η = Tlum/ε also shown in Figure 4,

indicates that for this type of films, an optimum can be found at around 22 nm of Ag.


326

Components

18

0.12
Measured emittance
Figure of merit (n)

0.10

16

14
0.06
12

η = Tlum/ε

Emittance

0.08

0.04
10


0.02

8

0.00
10

15

20
25
30
Ag layer thickness (nm)

35

40

Figure 4 Emissivity and figure of merit of ZnS/Ag/ZnS coatings vs. Ag thickness.Adapted from Leftheriotis G and Yianoulis. P (1999) Characterization
and stability of low emittance multiple coatings for glazing applications. Solar Energy Materials and Solar Cells 58: 185–197, with permission.

The D/M/D coatings are more versatile than doped metal oxides. It is possible to optimize them either for thermal or for solar
control through proper selection of the thickness of each individual layer with use of standard thin film optics software and the
‘characteristic matrix’ formulation [8]. Typical thicknesses and properties of D/Ag/D films are as follows: 40/20/40 nm for solar
control (Tlum ≈ 85%, Tsol ≈ 50%, ε ≈ 0.05) and 30/10/30 nm for thermal insulation (Tlum ≈ 87%, Tsol ≈ 72%, ε ≈ 0.15). Deposition of
such extremely thin stacks requires exact growth control of each individual layer. Furthermore, optical interference between different
layers causes the whole coating to fail, should only one layer deviate from the desirable thickness. The major disadvantage of D/M/D
coatings is their lack of durability. They are sensitive to environmental exposure and degrade with time as atmospheric gases and Ag
diffuse into the dielectric layer and react with each other [29]. They cannot withstand heat treatment above 200–250 °C [8]. For
these reasons they are referred to as ‘soft coatings’. To resolve these problems and to improve the low-e coatings’ properties intense

research work is being carried out worldwide [15]. Indeed, work has been reported recently on ZnS/Ag/ZnS, ZnO/Ag/ZnO, ZnO/Cu/
ZnO, IZO/Al/GZO, ZnO:Al/Ag/ZnO:Al, TiO2/Ag/TiO2, and many others [15].
Typical optical transmission properties of Ag-based low-e coatings are shown in Figure 5. Therein appear ZnS/Ag/ZnS stacks
optimized for maximum transmittance at 550 nm, ZnS/Cu/Ag/ZnS and ZnS/Al/Ag/ZnS coatings with ultrathin (less than 5 nm) Cu
or Al films added on Ag to improve emittance and provide thermal stability. Finally, five-layer stacks of the form ZnS/Ag/ZnS/Ag/
ZnS have been developed to provide nearly zero emittance (less than 0.02). These coatings and their properties are indicative of the
recent technological developments in the field, which have already been adopted by the glazing industry. Indeed, Ag-based thin
films for energy-efficient fenestration are now highly optimized, and a very large number of products with specified thermal, solar,
and luminous properties are available on the market. Commercial metal-based coatings are prepared by sputtering, as this
technique permits accurate thickness control and in-line coating of large areas. It is remarkable that with the sputtering method,
thickness control approaching atomic precision is feasible in a high-performance production environment and handling of glass
sheets up to 30 m2 in size is made possible. The optimization of commercial Ag-based low-e coatings has been brought about by
several breakthroughs presented in the following paragraphs [8, 30].
90

ZnS/Ag/ZnS
ZnS/Cu/Ag/ZnS
ZnS/Al/Ag/ZnS
ZnS/Ag/ZnS/Ag/ZnS

80
70

T (%)

60
50
40
30
20

10
0
300

400

500

600

700
λ (nm)

800

900

1000

1100

Figure 5 Transmittance spectra of various Ag-based low-emissivity (low-e) coatings.Adapted from Leftheriotis G, Yianoulis P, and Patrikios D (1997)
Deposition and optical properties of optimized ZnS/Ag/ZnS thin films for energy saving applications. Thin Solid Films 306: 92, with permission.


Glazings and Coatings

3.10.3.4.3

327


Use of interface layers

Interfaces are ultrathin, optically neutral layers, used to improve the coating properties: A 5 nm-thick ZnO interface (also called the
‘seed layer’) intervenes between the dielectric film and the Ag layer. It is used as substrate for the Ag layer to enhance the Ag film
uniformity, pushing the coalescence threshold down to 10 nm. Thus, coatings with the lowest thermal emissivity and the highest
luminous transmittance result. Furthermore, a 2–4 nm-thick TiOx interface, also called ‘sacrificial layer’, grown on top of the Ag film
prevents oxidation of the Ag layer during the reactive deposition of the covering oxide layer. It also protects the Ag film from oxygen
permeation that decreases the age resistance of the coating. This development was implemented industrially in the mid-1990s.

3.10.3.4.4

Application of a chemically and mechanically resistant top layer

An oxynitride (such as SiNxOy) improves the mechanical and age resistance of Ag-based low-e coatings. This development was
industrially applied after the implementation of the dual magnetrons as a sputtering tool since the end of the 1990s.

3.10.3.4.5

Development of asymmetrical coatings

Asymmetrical coatings incorporate dielectric layers with different refractive indices. Compared to symmetrical layer structures, such layer
systems result in higher transmittance (due to better antireflection of Ag) and possess neutral color, granting a higher market acceptance.
They also exhibit lower color sensitivity for individual layer thickness variations, resulting in a coating process with less waste. In this field,
fundamentally theoretical and practical developments were performed by Grosse et al. [31]. Since the end of the 1990s, the changeover
from symmetrical to asymmetrical Ag layer systems has been accomplished by degrees by all low-e manufacturers.
The structure of a typical coating incorporating the advances mentioned above is glass/TiO2 (n = 2.5)/ZnO/Ag/TiOx/SnO2
(n = 2.0)/SiNxOy, achieving Tlum = 80%, Tsol = 60%, and ε = 0.10. A double glazing with such a coating and Ar gas filling the air
gap exhibits a U-value equal to 1.1 W m−2 K−1.


3.10.3.4.6

Development of Ag-based coatings resistant to high temperatures

Since the 1980s, Ag layer systems resistant to high temperatures have been developed for the production of bent coated car
windshields at temperatures up to 650 °C. This was achieved with use of suboxide dielectric layers (usually TiOx, x < 2) and/or
ultrathin Ti interface layers that protected the Ag layer against oxidation during annealing. Later, this knowledge was transferred to
the production of heat-strengthened glass. Since the end of the 1990s, heat-resistant Ag-based layer systems have been marketed for
heat-strengthened architectural glass. A specific characteristic of these coatings is that they are opaque and absorbing in the as
prepared state (due to the presence of Ti and suboxide films). They become optically transparent and heat reflective after annealing
at high temperatures, as the opaque films are oxidized and become clear (Ti and TiOx absorb oxygen and are transformed into
TiO2).

3.10.3.4.7

Development of coatings with double Ag layers

Coatings in the form of D/Ag/D/Ag/D were proposed by Berning [26], in the early 1990s. It is essentially two D/Ag/D stacks put
together, with the middle dielectric layer being twice as thick as the other two. The layer structure also contains interface (seed and
sacrificial) layers above and below the Ag films. Insulating glass units with such a coating exhibit reduced heat transmittance
(≈0.1 W m−2 K−1) compared to units with single Ag layer systems, but its selectivity effect on solar radiation is much higher (see
Figure 5). It must be stated that double Ag layer coatings are costly to produce. At the moment, solar control glasses with such
coatings are a trendsetter for car and architectural glazing all over the world.

3.10.3.5

Conclusions – Resumé

Low-e coatings are spectrally selective thin films that add value to plain glass enabling it to perform multiple functions as part of
fenestration systems: daylighting of buildings and at the same time suppression of radiative heat losses. There are two broad

categories of coatings: doped metal oxides and metal-based stacks. The former are less expensive, they can be deposited on glass by
spray pyrolysis immediately after it leaves the float line, and they are better suited for thermal insulation, in cold climates. The latter
comprises three to five thin film stacks that require advanced equipment for their production (such as sputtering in high vacuum)
and accurate thickness control. They are more expensive, but more versatile: they can be tailored on demand either for solar control
or for thermal insulation. Recent advances in the glazing industry (especially in the metal-based coatings field) have led to
widespread production of low-e coatings for fenestration, automotive, and architectural applications. Furthermore, these coatings
exhibit electronic conductivity and are being used as TCs in a multitude of devices, such as light-emitting diodes, displays,
dye-sensitized and organic solar cells, smart switchable windows, and gas sensors. This wide range of applications brings these
films in the forefront of high technology.

3.10.4 Glass and Windows
3.10.4.1

Float Glass

Float glass is the basic glass from which almost all the flat glass products are derived. It is produced by the ‘float’ process that
involves the flotation of molten glass on a bath of liquid tin, producing a perfectly flat surface on both sides. The glass has no wave
or distortion and is nowadays the standard method for glass production: over 90% of flat glass produced worldwide is float glass.


328

Components

3.10.4.1.1

Manufacture

The raw materials used for the production of float glass typically consist of 72.6% sand (silicon dioxide), 13.0% soda (sodium
carbonate), 8.4% limestone (calcium carbonate), 4.0% dolomite, and 1.0% alumina. Another 1.0% of various additives is also

present. These are compounds for the adjustment of the physical and chemical properties of the glass, such as colorants and refining
agents. The raw materials are mixed in a batch mixing process, then fed together with suitable cullet (waste glass), in a controlled
ratio, into a furnace operating at approximately 1500 °C. Common flat glass furnaces are 9 m wide, 45 m long, and contain more
than 1200 tons of glass. Once molten, the temperature of the glass is stabilized to approximately 1200 °C to ensure a homogeneous
specific gravity. The molten glass is fed into a ‘tin bath’, a bath of molten tin (about 3–4 m wide, 50 m long, and 6 cm deep), through
a delivery canal [32]. The amount of glass allowed to pour onto the molten tin is controlled by a gate. Once poured onto the tin
bath, the glass spreads out in the same way that oil spreads out if poured onto a bath of water. In this situation, gravity and surface
tension result in the top and bottom surfaces of the glass becoming approximately flat and parallel. The molten glass does not
spread out indefinitely over the surface of the molten tin. Despite the influence of gravity, it is restrained by surface tension effects
between the glass and the tin. The resulting equilibrium between the gravity and the surface tension defines the equilibrium
thickness of the molten glass (T), given by the relation [33]:
À
Á
T 2 ¼ Sg þ Sgt þ St



2ρt

gρg ρt − ρg



½24Š

with ρg the glass density; ρt the tin density; Sg, Sgt, and St the values of surface tension at the glass–air, glass–tin, and tin–air interfaces,
respectively. For standard soda-lime-silica glass under a protective atmosphere and on clean tin, the equilibrium thickness is
approximately 7 mm.
Tin is suitable for the float glass process because it has a high specific gravity, is cohesive, and is immiscible into the molten glass.
Tin, however, is highly reactive with oxygen and oxidizes in a natural atmosphere to form tin dioxide (SnO2). Known in the

production process as dross, the SnO2 adheres to the glass. To prevent oxidation, the tin bath is provided with a positive
pressure-protective atmosphere consisting of a mixture of nitrogen and hydrogen. The glass flows onto the tin surface forming a
floating ribbon with perfectly smooth surfaces on both sides and an even thickness. As the glass flows along the tin bath, the
temperature is gradually reduced from 1100 °C until the sheet can be lifted from the tin onto rollers at approximately 600 °C. The
glass ribbon is pulled off the bath by rollers at a controlled speed. Variation in the flow speed and roller speed enables glass sheets of
varying thickness to be formed. Top rollers positioned above the molten tin may be used to control both the thickness and the width
of the glass ribbon. Once off the bath, the glass sheet passes through a lehr kiln for approximately 100 m, where it is further cooled
gradually so that it anneals without strain and does not crack from the change in temperature. On exiting the ‘cold end’ of the kiln,
the glass is cut by machines. A block diagram of the float process is shown in Figure 6.

3.10.4.2

Toughened Glass

Toughened glass is physically and thermally stronger than regular glass. The principle of toughened glass relies on the fact that the
faster contraction of the glass surface layer during cooling induces compressive stress in the surface of the glass balanced by tensile
stress in the body of the glass as shown in Figure 7. The greater the surface stress, the smaller the glass particles that will result in case
the glass is broken. Surface compressive stresses increase the glass strength. This is because any surface flaws tend to be pressed
closed by the retained compressive forces, while the core layer remains relatively free of the defects, which could cause crack
initiation. For glass to be considered toughened, the surface compressive stress should be a minimum of 69 MPa. For it to be
considered safety glass, the surface compressive stress should exceed 100 MPa. There are two main types of commercial heat-treated
glass: heat strengthened and fully tempered. Heat-strengthened glass is twice as strong as common glass while fully tempered glass is
typically 4–6 times stronger and can withstand heating in microwave ovens.
Raw material silos

Inert gas

Mixing of raw
materials
Heaters


Float area

Molten glass

Molten tin

Figure 6 Block diagram of a typical float line.

Annealing

Inspection

Cutting


Glazings and Coatings

Tensile stress

329

Compressive stress

Figure 7 Stress distribution of glass due to tempering.

3.10.4.2.1

Manufacture and properties


Toughened glass is made from annealed glass via a thermal tempering process. The glass is placed onto a roller table, taking it
through a furnace that heats it above its annealing point of about 720 °C. The glass is then rapidly cooled by forced convection with
use of air drafts. As a result, the external layers of the glass plate solidify earlier than the internal part. Stresses are generated during
this process as shown in Figure 7. An alternative chemical process involves forcing a surface layer of glass at least 0.1 mm thick into
compression by ion exchange of the sodium ions in the glass surface with the 30% larger potassium ions, by immersion of the glass
into a bath of molten potassium nitrate. Chemical toughening results in increased toughness compared with thermal toughening,
and can be applied to glass objects of complex shape [34].
Toughened glass has several disadvantages, all due to its pronounced stress pattern:
• It must be cut to size or pressed to shape before toughening and cannot be reworked once toughened. The same applies to
polishing the edges or drilling holes.
• It is most susceptible to breakage due to edge damage as the tensile stress is a maximum there. Shattering can also occur
in the event of a hard impact in the middle of the glass pane or if the impact is concentrated (e.g., striking the glass with a
point).
• Using toughened glass can pose a security risk in some situations because of the tendency of the glass to shatter completely
upon hard impact rather than leaving shards in the window frame. In Reference 35, it is stated that “The security value of
tempered glass is questionable. Although it will resist a brick or rock, it is susceptible to sharp instruments such as ice
picks or screwdrivers. When attacked in this manner, tempered glass tends to crumple easily and quietly, leaving no sharp
edges.”
• The surface of toughened glass is not as hard as that of plain glass and is more susceptible to scratching. To prevent this,
toughened glass manufacturers apply various coatings or laminates to the surface of the glass.
• Tempered glass has wavy surfaces, caused by contact with the rollers. This waviness is a significant problem in the manufacturing
of thin film solar cells [36].

3.10.4.3

Use of Glass in Solar Collectors

Glass plays an important role in thermal solar collectors, serving several purposes: It enables light admittance onto the collector
absorber, protects the absorber from weather elements (rain, dust, etc.), and provides some insulation against heat loss from the
collector front surface. An optimum solar collector glazing must perform well all three functions.


3.10.4.3.1

Light admittance

As can be seen in Figure 3, plain float glass exhibits high transmission in the visible and near IR, apart from a reduction in the
wavelength range from 700 to 1500 nm. This dip in transmittance is associated with iron compounds present in the glass, especially
iron trioxide, Fe2O3. The larger the iron content of the glass, the less transparent it becomes. Glass with high iron content (0.5% and
above) has a greenish appearance and poor transmittance. As can be seen in Figure 3, the plain glass reflectance in that range is low,
thus the loss of transmittance is due to absorption. This is very inconvenient from the solar collector point of view, as at this range,
there is a significant amount of solar radiation that cannot reach the absorber. In order to rectify this situation, glass with low iron
content (about 0.02% Fe2O3) has been developed. This improves the glass Tvis value from 88% (a value typical of a 6 mm-thick clear
float glass, see Table 1) to 91%.
Further improvement in the glass transmittance can be effected by reduction of reflectance (which is about 8% in the visible, as
can be seen in Figure 3 and Table 1). There are two types of reflectance: diffuse and specular (mirror-like). The former is caused by
the roughness of the reflecting surface and can be suppressed by a treatment of the glass surfaces called ‘etching’ [5]: The glass is
dipped into a silica-saturated fluorosilicic acid solution, which smooths its surface, reducing Rvis to 2%. The latter (specular)
reflectance is caused by refractive index mismatch between glass and air (1.5 and 1.0 respectively). The use of thin films with
refractive indices in between 1.5 and 1.0 and of appropriate thickness can cause the reflected light beams originating from the air/
film and film/glass surfaces to have equal magnitude and opposite phase. The two beams then cancel out suppressing reflectance
and increasing transmittance. These films are the well-known ‘antireflection’ coatings, used in optical lenses. A glass with anti­
reflection coatings on both sides can achieve Rvis values as low as 1% [5]. However, as the production cost of such coatings is rather
high, they are not very common in solar collector covers.


330

Components

Mid-pane values of optical and thermal performance indicators for various types of coatings


Table 1

Thermal

Visible

No

1
2
3
4
5
6

7
8
9
10
11
12
13
14
15
16
17
18
19
20

21
22
23

Type of glazing
Single Glazing
Clear, 6 mm thick
Tinted, 6 mm thick
Reflective
Tl Low-e 6 mm thick, indoors
Tl Low-e, 10 mm thick, indoors
Self-cleaning, 6 mm thick
Double glazing (all panes 6-mm thick)
OUTDOORS | INDOORS
Clear | Clear
Tinted | Clear
Reflective | Clear
Tl Low-e | Clear
SC Low-e | Clear
Self-cleaning/SC Low-e | Clear
Tinted SC Low-e | Clear
SC Low-e | Tl Low-e
Tl Low-e | Tl Low-e
Triple glazing (all panes 6 mm thick)
Clear | Clear | Clear
Tl Low-e | Clear | Clear
SC Low-e | Clear | Clear
Self-cleaning/SC Low-e | Clear | Tl Low-e
Quadruple glazing (all panes 6 mm thick)
Clear | Clear | Clear | Clear

Tl Low-e | Clear | Clear | Clear
SC Low-e | Clear | Clear | Clear
SC Low-e | Clear | Clear | Tl Low-e

U-value
(WK−1m−2)

Solar

Tvis

Rvis

Tsol

Rsol

Asol

g-value
(SHGC)

SC

88
50
31
67
65
83


8
5
42
25
24
14

79
47
24
58
53
79

7
5
47
19
17
13

14
48
29
23
30
8

0.82

0.60
0.38
0.62
0.58
0.81

0.95
0.68
0.44
0.71
0.66
0.93

5.7
5.7
5.6
3.8
3.7
5.9

79
44
49
73
70
67
49
53
56


14
7
39
17
10
16
39
22
31

64
38
28
55
38
36
28
29
42

12
7
43
15
28
32
42
34
23


24
55
29
30
34
32
30
37
35

0.72
0.48
0.31
0.69
0.43
0.40
0.31
0.34
0.53

0.82
0.55
0.36
0.79
0.49
0.46
0.36
0.40
0.60


2.9
2.7
2.7
1.7
1.4
1.4
1.4
1.3
1.5

2.8
2.6
2.6
1.5
1.1
1.1
1.3
1.1
1.3

70
55
62
54

19
32
14
20


52
39
33
28

15
24
29
35

33
37
38
37

0.63
0.50
0.39
0.35

0.73
0.57
0.45
0.40

1.8
1.3
1.0
0.8


1.7
1.2
0.9
0.7

63
50
56
52

23
35
17
18

43
33
28
26

17
25
31
31

40
42
41
43


0.57
0.45
0.36
0.34

0.65
0.52
0.45
0.39

1.3
1.0
0.9
0.7

1.2
0.9
0.7
0.6

Air filled

Gas filled

SC, solar control; Tl, thermal Insulation.

(|) Symbolizes the gap between glass panes.


3.10.4.3.2


Weather protection and heat loss suppression

The use of tempered (toughened) glass is common in solar collectors to withstand blizzard attacks. Furthermore, appropriate edge
sealing is required to prevent moisture ingress and air infiltration.
As regards heat losses, glass effectively suppresses convective losses to the environment (compared to an unglazed collector) but
as it exhibits significant emittance (∼0.84), it suffers from radiation losses, especially during nighttime. To improve the performance
of glass in this respect, one could use the low-e coatings presented in the previous section. However, these coatings would also
reduce light admittance, as can be seen in Figure 3. A more appropriate solution is to stop the heat from being emitted from the
absorber surface, using ‘selective’ absorbers that exhibit high absorptance and low emittance. The analysis of selective absorbers is
beyond the scope of this chapter.

3.10.4.4

Windows in the Built Environment

A window can be split into two parts: the glass, which occupies 80–90% of its area, and the frame, which is used to support the glass
pane on the building walls and to act as a peripheral seal.
In Table 1 appear the mid-pane values of optical and thermal performance indicators for various types of glazing. They have
been calculated with use of the Pilkington ‘Spectrum’ online tool according to the following standards:
• EN 410 for optical properties in the visible and solar part of the spectrum (Tvis, Rvis, Tsol, Rsol, Asol) and for g-value.
• EN 673 for U-value.
The results derived by this piece of software have been cross-checked and validated with equivalent results from the literature
[37–42].


Glazings and Coatings

331


The values of Table 1 are typical for state-of-the-art commercial products that abound in the market. They are used to compare
different glazing types that are presented in the following sections.

3.10.4.5
3.10.4.5.1

Single-Glazed Windows
Clear single glazing

The simplest type of window consists of a single clear uncoated glass. It provides the highest visible transmittance but exhibits large
thermal losses. Furthermore, such a glazing does not provide sufficient sound insulation and suffers from mist condensation.
Nowadays, the use of such glazing is limited to low-cost solutions or to retrofitting of windows in historical buildings that do not
possess thick enough frames to accommodate double glazing.

3.10.4.5.2

Tinted single glazing

Tinted glass is a normal float glass containing colorants. Colored glass is an important architectural element for the exterior
appearance of facades. It is also used in interior decoration (doors, partitions, staircase panels, mirrors). Its production is the same as
that of clear float glass apart from the addition of appropriate colorants to the standard raw materials. Colorants are mostly metals,
each producing a different color, depending on its nature and its concentration in the glass. Some of the most common colorants
and the colors they produce are the following: iron/green, brown, blue; manganese/purple; chromium/green, yellow, pink;
vanadium/green, blue, gray; copper/blue, green, red; cobalt/blue, green, pink; nickel/yellow, purple; cadmium sulfide/yellow;
titanium/purple, brown; cerium/yellow; carbon and sulfur/amber, brown; selenium/pink, red; gold/red.
Due to its high extinction coefficient, low transmittance, and high absorptance, tinted glass is often called ‘absorptive’. The low
visible transmittance reduces the quantity of daylight admitted indoors. Its primary use in windows is therefore to reduce glare and
excessive solar transmission. As the reduction in light transmission is effected through absorption, such glazing exhibit high SHGCs:
the absorbed radiant energy is initially transformed into heat within the glass, thus raising the glass temperature. A significant
amount of it is then reemitted indoors. Tinted glazing allows a greater reduction in visible transmittance (Tvis) than in SHGC due to

reemission, as can be seen in Table 1. In a practical situation, transmittance in the visible and SHGC are required to increase (winter,
cold climates) or to decrease (summer, hot climates) simultaneously by a similar degree. Thus, single-tinted glazings are far from
achieving optimum performance. To rectify this situation, other, more appropriate solutions have been developed such as spectrally
selective coatings with light blue/green tint having higher visible performance and lower SHG (glazing no. 13 in Table 1 is one such
example).

3.10.4.5.3

Reflective single glazing

A reflective coating can be added to glass to increase the reflectivity of its surface, in order to achieve a considerable reduction in solar
gains. The reflective coating usually consists of thin metallic or metal oxide layers, and comes in various metallic colors such as
bronze, silver, or gold. The SHGC varies with the thickness and reflectivity of the coating, and its location in the glazing system.
While some reflective coatings must be protected by sealing in cavity (e.g., those based on noble metals), others are durable and can
be added on exposed surfaces. It can be seen in Table 1 that a reflective coating changes very little the U-value of single glazing that is
dominated by convection (between glass and the surrounding air) and conduction (through the glass). Similar to that of tinted
glass, the visible transmittance of reflective glass declines more than its SHGC. Architects are generally fond of reflective glazing
because of its glare control and appealing outside appearance. However, the usage is limited by its sun mirror effect that may cause
disturbances on traffic roads and nearby buildings. Also in well-illuminated rooms, the loss of visual privacy and outside views at
night can be a concern to the occupants.

3.10.4.5.4

Low-emittance single glazing

Low-e coatings can be added to float glass to achieve either solar control or thermal insulation. The materials and properties of low-e
coatings have already been covered extensively. The performance of single low-e-coated glazing depends on the position of the
coating (indoors or outdoors). The correct placement of the coating is indoors in order to suppress long-wave radiative heat losses to
the environment. In that configuration, the heat emitted from indoors is reflected back into the room. Otherwise (e.g., with the
low-e coating facing outdoors), the heat would have been absorbed by the glass, raising its temperature, which would have caused

an increase of convective heat losses. The glass thickness also plays a role in the performance of such a glazing, especially in SHGC
(or g-value), as can be seen in Table 1. Low-e-coated glass is very popular in modern architecture over the world, mostly in
conjunction with double glazing. The market share is over 30% of the fenestration products installed in the United States [13].

3.10.4.5.5

Self-cleaning single glazing

Self-cleaning glazing [43, 44] have an additional coating on their external surface that is exposed to weather elements and gets dirty
with time. This coating is a nanostructured TiO2 film (less than 1 μm thick). TiO2 is a semiconductor with a band gap of 3.2 eV.
Under UV light irradiation, electrons and holes are produced in the conduction and valence band of the film, respectively, in
accordance to the following reaction: TiO2 + hv ( > 3.2eV) → e− + h+. The holes and electrons can react with molecules adsorbed on
the film surface as follows: The holes oxidize water molecules to hydroxyl radicals and the electrons react with oxygen to form
peroxide radicals, which in turn can react with electrons and protons to produce hydrogen peroxide. These highly reactive radicals


332

Components

and hydrogen peroxide decompose dirt particles deposited on the glass surface. This phenomenon is called ‘photocatalysis’ and it
was first reported by Fujishima and Honda [45]. Furthermore, TiO2 thin films possess another interesting property: they are water
repellent. Rainwater that lands on the glass surface does not adhere on it. Instead, droplets are formed that drip down the window
washing away the decomposed dirt particles. Hence, the self-cleaning action works in two stages: breakdown of dirt under UV
irradiation and removal of dirt by the rainwater. Self-cleaning glass is particularly suited for highly glazed, tall buildings in which
glass cleaning is time consuming and costly. The reduction of glass cleaning costs brought about by the use of self-cleaning glass can
counterbalance (or even overcome) the increased cost of glass. Nowadays, there are several glass manufacturers that produce
self-cleaning products. However, research is still active in this field, aiming to improve the photocatalytic efficiency by use of visible
light instead of UV, by increase of the TiO2 porosity and by use of alternative materials (such as ZnO, SnO2, and CeO2) and their
combinations [43, 44]. TiO2 is a high index of refraction material and its application on the external glass surface is expected to

increase reflectance. Indeed, in Table 1, it can be seen that windows with self-cleaning glass exhibit slightly lower visible
transmittance, higher reflectance, and slightly lower g-value. No significant effect can be observed in U-values.

3.10.4.6

Multiple-Glazed Windows

Multiple panes with air-sealed cavities can be used to improve the glazing thermal insulation properties without undue reduction in
transmittance and in heat gains. The fabrication of multiple glazing (double, triple, quadruple) poses new challenges to the
manufacturers: the cavity must be air and moisture proof, thus appropriate sealants (called ‘spacers’) must be developed.
Furthermore, the spacers must be able to accommodate the thermal stress and the differential expansion of the two (or more)
glass sheets. They are also required to be thermally insulating, otherwise edge losses may exceed the extra insulation multiple glazing
offer. To minimize heat losses through multiple glazing, one needs to reduce not only the peripheral heat losses (e.g., conduction
through the spacers) but also the heat transferred through the air gap. The latter can be caused either by thermal conductance of the
air (or gas) through an unduly short gap or by convection in case the gap is too large. An optimum of the gap width can be found at
about 14–16 mm [46], as can be seen in Figure 8. At that distance both heat transfer mechanisms are counterbalanced and the heat
loss is minimized. Inert gases have a similar behavior: Ar has an optimum similar to air (14 mm). The optimum for Kr and Xe is 10
and 6 mm, respectively [46]. The optimum distance depends on the thermophysical properties of the various gases (e.g., thermal
conductivity, thermal diffusivity, and viscosity). Different types of multiple glazing are presented next.

3.10.4.6.1

Double glazing

Thermal conductance (W m–2 K–1)

As can be seen in Table 1, the U-value of a double glazing with two clear panes is 49% lower than that of a similar single pane window.
The improvement of the window thermal properties comes at a price of a 12% reduction in SHGC, which is more than acceptable. In
situations where further reduction of solar gains is desirable, tinted and reflective glass can offer a decrease in g-value between 30% and


4.40
4.20

U-value (W K–1m–2 )

4.00
3.80
3.60

3,0
Air
Argon

2,5

Krypton
Xenon

2,0

1,5

1,0

0,5
0

5

10


15

20

25

30

Distance (mm)

3.40
3.20
3.00

Conduction-dominated region

Convection-dominated region

2.80
0

5

10

15
Distance (mm)

20


25

30

Figure 8 U-value vs. distance of glass panes for a double glazing with two clear panes. Calculated using LBNL ‘Window 5.2’. (Inset) Conductive and
convective heat transport in a glazing cavity as a function of distance between panes for air and inert gases. Adapted from Manz H (2008) On minimizing
heat transport in architectural glazing. Renewable Energy 33: 119–128, with permission.


Glazings and Coatings

333

60%. As with single glazing, the U-value of double-glazed windows with tinted or reflective glass are not affected considerably. Low-e
coatings, on the other hand, further improve thermal insulation, halving the clear double glazing U-value. Depending on the type of
low-e coating, different properties can be obtained (e.g., solar control, with suppressed solar gains, or thermal insulation with high solar
gains), each appropriate for a different climate type and application. The placement of low-e coating on the window assembly plays a
marginal role on the overall window properties. Usually these coatings are placed within the air gap for protection reasons. The
outdoors pane is preferred by glass manufacturers to be the coated one, with the indoors pane being clear glass, as this configuration is
more efficient in blocking thermal radiation from the inside (in a similar way as in the case of single glazing but less pronounced). Using
two coatings on both panes is not favored by glass manufacturers. Indeed, as follows by comparison of windows 10 and 15 of Table 1,
the use of two low-e coatings instead of one causes a 23% reduction in SHGC and an 11% reduction in U-value, which are small
compared to the increase of the window cost, which could be 1.5 times higher or more. Of the double glazing appearing in Table 1, the
most effective ones in terms of thermal insulation seem to be those with the solar control coatings.
Further reductions in U-values are possible with use of a less conductive and more viscous inert gas. Manufacturers use Ar, Kr, Xe,
and mixtures of these as filling gases. They are nontoxic, nonreactive, clear, and odorless. As can be seen in Table 1, the effect of
filling the gap with an inert gas is a reduction in U-values in the range of 0.1–0.3 W m−2 K−1 depending on the glazing configuration.
Thus, highly insulating windows benefit more from the use of inert gases, as the relative change in U-value is larger. The biggest
problem with inert gases is that their retention in the glazing is questionable. As with all gases, they tend to diffuse through the seals

and to escape through microcracks in the sealing materials. Keeping the gas within the window unit depends largely on the quality
of the design and construction, materials in use, and assembly, particularly the sealing techniques. As a result of all these
improvements, state-of-the-art double glazings with g-values up to 0.49 and U-values down to 1.1 W m−2 K−1 have been achieved.

3.10.4.6.2

Triple and quadruple glazing for ultrahigh thermal insulation

In heating-dominated climates with extremely low temperatures, the U-values of double glazing are not low enough to ensure
acceptable thermal losses of buildings. In these environments, triple and quadruple glazing are used, having U-values down to
0.6 W m−2 K−1. The price to be paid is the reduction in solar gains, and increase of the window dimensions, weight, and cost. In
high-tech triple and quadruple glazing, Kr or Xe are used as the filling gases to reduce the overall width of the window. These gases
allow placement of the glass panes at shorter distances (see the inset of Figure 8).

3.10.4.7

Window Frames

Frames are very important in glazing systems. Not only do they hold windows in place but they also act as a peripheral seal. The
thermal properties of the frame play a significant role in the overall U-value of the window. Nowadays, a variety of window frames is
available, ranging from wood (the material used traditionally) to aluminum, plastics, and various other composite materials. Next,
the different frame types are presented in order of decreasing thermal conductivity [47–49].

3.10.4.7.1

Aluminum

Aluminum is a light, strong, and durable material that can be easily extruded into complex shapes. Aluminum frames are available
in anodized and factory-baked enamel finishes that are extremely durable and require virtually no maintenance. However,
aluminum as a window frame material is not very efficient in terms of thermal insulation. As a metal, it exhibits high thermal

conductance, greatly raising the overall U-value of a window unit. For this reason (and also for economizing on materials), the
aluminum frame profiles are hollow with complex shapes, in an effort to create air enclaves.
In hot climates, where solar gain suppression is often more important than heat losses, improving the insulating value of the
frame can be much less important than using a higher-performance glazing system. In cold climates, on the other hand, a simple
aluminum frame can easily become cold enough to condense moisture or frost on the surfaces of window frames. Even more than
the problem of heat loss, the condensation problem has led to the development of better insulating aluminum frames. The most
common solution to the heat conduction problem of aluminum frames is to provide a ‘thermal break’ by splitting the frame
components into interior and exterior pieces and use a less conductive material to join them. Current technology with standard
thermal breaks has decreased aluminum frame U-values from roughly 10 to about 6 W m−2 K−1.

3.10.4.7.2

Wood and wood composites

The traditional window frame material is wood, because of its availability and ease of milling into the complex shapes required to
make windows. Wood is favored in many residential applications because of its appearance and traditional place in house design.
From a thermal point of view, wood-framed windows perform well with frame U-values in the range of 2−3 W m−2 K−1. The
disadvantages of wooden frames are that they are not as durable as other materials and require maintenance (with paint or lacquer)
to last longer.
A variation of the wood-framed window is to clad the exterior face of the frame with either vinyl or aluminum, creating a
permanent weather-resistant surface. Clad frames thus have lower maintenance requirements, while retaining the attractive wood
finish on the interior.
Alternatively, composite wood products, such as particle board and laminated strand lumber, in which wood particles and resins
are compressed to form a strong composite material have also been used to produce window frames that have similar thermal


334

Components


properties of wood. New materials have emerged recently, such as wood/polymer composites that are extruded into a series of lineal
shapes for window frame and sash members. These composites are very stable and have the same or better structural and thermal
properties as conventional wood, with better moisture resistance and more decay resistance.

3.10.4.7.3

Plastics (vinyl, fiberglass, thermoplastics)

Vinyl, also known as polyvinyl chloride (PVC), is a very versatile plastic with good insulating properties. Vinyl window frames do
not require painting and have good moisture resistance. Because the color goes all the way through, there is no finish coat that can
be damaged or deteriorate over time – the surface is therefore maintenance-free. Some vinyl window manufacturers are now offering
surface treatments such as laminates (wood veneer, paintable/stainable, maintenance-free) and coatings. The main disadvantage of
vinyl is that, like all polymers, its long chains tend to break up under UV irradiation. Thus, the material ages and loses its properties.
Recent advances have improved resistance to degradation from sunlight and temperature extremes. In terms of thermal perfor­
mance, vinyl frames are comparable with wood, while there are minor differences, depending on the frame construction. Small,
hollow chambers within the frame reduce convection exchange, as does adding an insulating material.
Window frames can also be made of glass-fiber-reinforced polyester, or fiberglass, which is pultruded or extruded into lineal
forms and then assembled into windows. These frames are dimensionally stable and have air cavities (similar to vinyl). When the
cavities are filled with insulation, fiberglass frames have thermal performance superior to wood or vinyl (similar to insulated vinyl
frames). Because the material is stronger than vinyl, it can have smaller cross-sectional shapes and thus less area. Another
polymer-based approach is to use extruded engineered thermoplastics, another family of plastics used extensively in automobiles
and appliances. Like fiberglass, they have some structural and other advantages over vinyl. Usually these high-performance frames
are used with high-performance glazing.

3.10.4.7.4

Hybrid

Manufacturers are increasingly turning to hybrid frame designs that use two or more frame materials to produce a complete window
system. The wood industry has long built vinyl- and aluminum-clad windows to reduce exterior maintenance needs. Vinyl manu­

facturers and others offer interior wood veneers to produce the finish and appearance that many homeowners desire. Split-sash designs
may have an interior wood element bonded to an exterior fiberglass element. We are likely to see an ever-increasing selection of such
hybrid designs as manufacturers continue to try to provide better-performing products at lower cost.

3.10.4.7.5

Effect of frames on the window thermal properties

It is evident that the frame chosen for a given type of glazing must possess similar thermal properties. A highly insulating frame with
a single glazing is a waste and, similarly, a simple, aluminum frame is not suitable for a superinsulating triple glazing. The ideal
frame for a given glazing is one that does not increase the whole window U-value above the mid-pane value. To match the different
types of windows to appropriate frames in terms of thermal properties, simulations of the whole window assembly and its thermal
performance must be carried out. Such results are presented in Table 2 in which appear the U-values of various combinations of
frames and glazing for typical residential windows (0.8 Â 1.2 m). They are simulations obtained with the ‘Window 5.2’ and ‘Therm
5.2’ software developed by Lawrence Berkeley National Laboratory [50, 51]. The mid-pane U-values of Table 1 have also been added
for comparison. It follows from Table 2 that most frame and glazing combinations have U-values higher than the mid-pane ones of
equivalent glazing. At best, insulating frames keep the window U-value equal to the mid-pane one. It is also clear that plain
aluminum frames are not suitable even for a single clear glass. Their use is only justified in hot climates where the reduction of
thermal losses is not a priority. Aluminum frames with thermal break are suitable for single glazing. Wood, composites, and vinyl
are suitable for double glazing with clear glass panes. For the more advanced glazing with low-e coatings either double or triple,
insulated vinyl or fiberglass frames are necessary. It should also be noted that the frame effects are more pronounced in smaller
windows, which have a larger edge-to-area ratio. The frame rebate (e.g., the amount of glazing covered by the frame) also influences
Table 2

U-values of windows with different types of glazing and frames
Type of frame

Type of glazing

Mid-pane


Aluminum

Aluminum with
thermal break

Wood, wood clad,
vinyl, composite

Vinyl insulated,
fiberglass

Clear, 6 mm thick
| INDOORS
Clear | Clear
Tl Low-e | Clear (gas filled)
SC Low-e | Clear (gas filled)
Tl Low-e | Clear | Clear (gas filled)
SC Low-e | Clear | Clear (gas filled)

5.7

6.6

5.7

–

–


2.9
1.5
1.1
1.2
0.9

4.3
3.5
3.4
–
–

3.6
2.8
2.7
–
–

2.9
2.1
1.9
1.6
1.6

–
1.6
1.5
1.3
1.0


OUTDOORS

SC, solar control; TI, thermal insulation.
(|) Symbolizes the gap between glass panes.


Glazings and Coatings

335

the thermal properties of the window: the larger the rebate, the lower the heat loss, as the glass edges that are potential thermal
bridges are ‘buried’ within insulated material and the thermal conduction path through which heat must pass increases [9].

3.10.4.8

Spacers and Sealants

Modern multiple glazing include spacer bars that hold the glass panes at a fixed distance apart and a sealant (called primary seal)
placed around the outer perimeter to hold the window together. The primary seal is usually a thermosetting polyisobutylene or
butyl rubber. The application of an additional layer of sealant to the shoulder of the spacer bar (termed secondary seal) creates an
extra barrier that reduces the overall vapor permeability of the glazing. The secondary seal is polysulfide, polyurethane, silicone, hot
melt butyl, or an epoxy glue. In glazing incorporating inert gases, an additional gas-retention seal (in the form of adhesive tape) is
added [47, 48, 50]. In Figure 9, the sketch of a double glazing is shown. The typical cross section of such a glazing and the various
components discussed above are also highlighted.
The spacer bar has a profile depth of 6–8 mm. Its width can be varied to give a cavity of 6–20 mm. The spacer bar is hollow and a
desiccant is incorporated into it to absorb the moisture trapped between the panes during manufacture of the glazing, as well as
moisture that permeates during its service life.
As in the case of frames, the spacer material is of great importance for the thermal properties of the window.
Aluminum has excellent structural properties, but as a conductor of heat it represents a significant thermal ‘short circuit’ at the
edge of the window, which reduces the benefits of improved glazing. In addition to the increased heat loss, the colder edge is more

prone to condensation.
To address this problem, window manufacturers have developed a series of edge systems, including solutions that depend on
material substitutions as well as new designs. One approach to reducing heat loss has been to replace the aluminum spacer with a
metal that is less conductive, for example, stainless steel, and change the cross-sectional shape of the spacer. These designs are widely
used in windows today.
Another approach is to replace the metal with insulating materials. The most commonly used design incorporates spacer, sealer,
and desiccant in a thermoplastic compound that contains a blend of desiccant materials and a thin, fluted metal shim of aluminum
or stainless steel. Another approach uses an insulating silicone foam spacer that incorporates a desiccant and has a high-strength
adhesive at its edges to bond to glass. The foam is backed with a secondary sealant. Both extruded vinyl and fiberglass spacers have
also been used instead of metals.
There are several hybrid designs that incorporate thermal breaks in metal spacers or use one or more of the elements
described above. Some of these are specifically designed to accommodate three- and four-layer glazings incorporating
stretched plastic films. All are designed to interrupt the heat transfer pathway at the glazing edge between two or more
glazing layers.
The effort to develop ‘warm’ edge materials with improved thermal insulation is justified by the fact that the thermal effect of a
spacer extends beyond its physical size, to a peripheral band 60–70 mm wide. The contribution of this ‘glass edge’ to the total
window U-value depends on the size of the window: glass edge effects are more pronounced in smaller windows, which have a
larger edge-to-area ratio. For a typical residential-size window (0.8 Â 1.2 m), changing from a standard aluminum edge to a
good-quality warm edge can reduce the overall window U-factor by approximately 0.1 W m−2 K−1. Apart from the reduction of
thermal losses, another benefit is the rise in interior surface temperature at the bottom edge of the window, which is the most prone
to condensation part of the glazing.

Air gap

Spacer

Glass pane

Desiccant
Primary seal


Secondary seal
Gas retention seal

Window frame
Frame profile
Figure 9 Sketch of a typical double glazing.


336

Components

3.10.4.9

Emerging Technologies

In a continuous effort to improve the performance of windows, various innovative technologies are emerging. These are as
follows [37]:
• Evacuated glazing. By replacing the air gap with a vacuum, better thermal insulation is expected, as convection and conduction
losses are eliminated. However, new challenges emerge, such as the retention of vacuum, heat transfer through the pillars used to
hold the glass panes apart, accommodation of differential expansion of the two glass panes, sound insulation issues, and more.
• Aerogel glazing. Use of aerogels (microporous insulating material such as granular silica) can lead to highly insulated glazing either
translucent or transparent.
• Switchable glazing. They are intended for dynamic solar control in ‘smart’ buildings that adapt their envelope to changing climatic
conditions. There are many different designs, such as ECs that change their color reversibly from transparent to dark blue upon
application of a DC voltage, gasochromics that perform the same task with use of H2 gas, thermochromics in which the color
change is triggered by heat, liquid crystal, suspended particles, and metal hydride devices with reflectance modulation (they
switch between a mirror and a transparent state), and others.
All these emerging technologies will be presented in the following sections.


3.10.4.10

Conclusions – Epilogue

The most important breakthrough in the flat glass industry is undoubtedly the development of the float process. It has revolutio­
nized glass manufacturing and led to the production of high-quality windows. Nowadays, multiple glazing with high visible
transmittance and increased thermal insulation are the state of the art in the fenestration market. The incorporation of various thin
film coatings (such as low-e, reflective, self-cleaning) have added value to the glazing products. Emerging technologies such as EG,
ECs, and aerogels promise that in the years to come, new improved products with even better properties will appear.

3.10.5 Evacuated Glazing
3.10.5.1

Operating Principles

There are two basic problems with the DGUs that need to be solved, for at least some applications demanding very good insulation.
The first is that the U-value cannot become lower than a limit, which appears to be approximately 1 W m−2 K−1. TGUs and
quadruple-glazed units can be used for this purpose in demanding or very harsh environments. However, this choice introduces the
second difficulty, which is lack of space and the lowering of U-value is still less than desired in the case of DGU. In addition, the TGU
is necessarily more expensive and heavy than is desired for most applications. The problem of space and weight is serious and the
tables in the preceding section show this. The EG was proposed to reply to these challenges.
It was proposed that a solution would be to evacuate the space between the two panes. This could improve the heat-insulating
characteristics to a very low U-value, as there is no convective heat transfer, leaving only the radiative heat transfer. The last can be
reduced to a minimum by using special low-e coatings on the inner two glass surfaces, or at least on one of them. The glass must be
supported internally using the tiny pillars, or else the atmospheric pressure will cause implosion.
We must also stop the glass panes from collapsing and touching each other internally, because of the forces on the outer two
planes (surfaces 1 and 4). The small (typical diameter 0.2–0.4 mm) supporting glass or metal pillars are placed at regular distances
about 20–40 mm from each other in a square arrangement to avoid the problem of reduction of optical clarity (transparency), as
shown in Figure 10. This distance is a critical parameter for the thermal and optical characteristics of the EG.

After this short description of the operating principles, we may state that with the EG, we try to eliminate completely the
convective heat losses in double-glazed windows by evacuating the space between the two glass units.

3.10.5.2

Technology and Related Problems

For the heat transfer, if there is vacuum in the space between the two glass panes, the conductance hg in eqn [22] becomes very small.
The typical EG unit consists of two glass panes sealed hermetically (gastight seal) around their edge so that vacuum can be
maintained in the space between the two glass panes (Figure 10). Low-e coatings are used on the internal glass surfaces (2 and 3 or
only on one of them) to reduce the only remaining mode of heat transfer, the radiative heat transfer.
The EG concept is very advanced and some basic problems must find better solutions. Further research is needed for a smart
design and a breakthrough in the details of its production that will make it attractive for the companies to produce and the
consumers to use it. Transparent low e-coatings (both hard and soft) are used on one or both of the inner surfaces and this is crucial
as otherwise the heat exchange by radiation would wipe out the advantage of the vacuum for extremely low heat transfer.
Benson et al. [52] tried a production method forming the edge seal of the glazing by laser heating in vacuum. During the process,
the entire glass structure was maintained at high temperature to reduce stresses in the glass. Practical difficulties in the sealing
method prevented this technique from being widely implemented. In addition, it was recognized that it would be necessary to


Glazings and Coatings

337

Support pillars
Indium alloy seal

Low-e coating
Figure 10 Construction and operating principle of a typical evacuated glazing with the support pillar array and the indium alloy vacuum seal.


include getters within the device in order to remove gas that was emitted by the glass while it was still hot, after the edge seal had
been completed.
Several factors influence the optimal separation of the glass sheets. On the one hand, we are constrained to have a small
separation, on the order of 1 mm, as otherwise the mechanical energy stored in the EG device could be hazardous in the case of
breakage. On the other hand, there is a low separation limit that is determined by the following factors [53]:
1. Due to the large distributed impedance to gas flow between the closely spaced glass sheets, long evacuation times may occur.
2. Evanescent field radiative coupling between the internal surfaces of the cavity may result in enhanced heat exchange and flow
through the evacuated space [54].
3. Interference fringes due to reflections from the internal surfaces may have visual unpleasant effects.
Clugston and Collins [53] applied a method of using glass, of the same kind as the panes, to seal the two glass panes at the outside
edge. A small tube about 4 mm long, with 1 mm internal diameter, was sealed on an appropriate hole in one glass plate with solder
glass. Evacuation was performed through this tube and the outer end of it is fused to seal the vacuum. They described the
determination of the time required to evacuate such a device through the small pump-out tube. The same researchers have
developed also an approximate analytic method for this study. The numerical results have been validated by experimental
measurements. It was shown that samples of about 1 m2 and internal gap of 0.2 mm could be evacuated in a few minutes through
the small pump-out tube. Therefore, the evacuation time in the production process of EG is determined mainly by the heating cycles
required for the outgassing of the internal glass surfaces. The pressure in the evacuated space should be kept below 1.5 Â 10−3 mbar,
so that the heat flow through the glazing by gas conduction is negligible.
The optimum distance between successive pillars is determined by the following opposite requirements. When the distance is
relatively large, a problem arises because of the tendency of the glass to bend over the glass beads causing disturbing reflections.
Closer bead spacing minimizes the optical anomalies, but denser spacing increases the thermal losses and optical intrusion.

3.10.5.3

The State of the Art

As we have already mentioned, it is necessary to use support pillars to maintain the separation of the glass sheets under the force
from atmospheric pressure. The design of the pillar array also involves careful optimization as we have two competing factors: heat
flow through the pillars and mechanical stresses in glass sheets and pillars. Pillars are about 0.2–0.5 mm in diameter; they are placed
at distances, on a square matrix, that may vary from 25 to 50 mm [7, 9, 53, 55–57]. Experimental and theoretical results enable an

assessment of the design choice of the internal gap in EG. The need to avoid optical interference fringes, and to ensure that
significant evanescent field radiative heat flow does not occur, gives a lower limit of 0.1 mm for the dimension of the gap. The
constraint to keep mechanical energy associated with the evacuated volume small, for safety, imposes an upper limit of 1 mm for the
gap. It is fortunate that the dimensions of practical pump-out tubes can be chosen to give evacuation time of less than 1 min in this
range of gap sizes, for samples of realistic dimensions. This time required to establish vacuum in EG is dominated by the time
required for the outgassing process.
Stable EG units were initially produced around 1990 [52, 55]. All early attempts, as well as the later fabrication of evacuated
experimental glazing units [58], necessitate temperatures above 450 °C for edge sealing. This is an energy-intensive process and
permits the use of only hard low-e coatings. These are essentially metal oxides, as SnO2 with fluorine as a doping material. The hard
coatings have an emissivity in the IR that cannot be lower than 0.15 and this has resulted in their inability to reduce radiative heat
transfer across the unit below a certain limit. In addition, we should emphasize that the emittance of the hard coatings becomes
even higher at the high temperatures used for edge seal fabrication. The emittance is increased irreversibly at very high temperatures.
For example, the emittance after fabrication at temperatures above 450 °C was measured to be 0.25 [56, 57]. There is, on the other
hand, a possibility to use special coatings that have been described earlier in this chapter. The idea to use soft coatings (of the type
D/M/D), which can have very low-e, forces one to be able to seal the double glass using lower temperatures of fabrication (typically
190 °C or lower), as this is near the upper temperature limit that these films can endure [7, 9]. Soft coatings can have much lower