Clearly, "active" devices require a power source sufficient to create the required brightness from the display. "Passive" devices require much lower power (therefore particularly suitable for batterey operation).

Solid materials generate light through a process known as "luminescence". Various types of luminescence depend on the methods by which it is excited: eg Photoluminescence is excitation due to absorption of photons. Cathodoluminescence is due to bombardment by electrons. Electroluminscence is due to the effect of and electric field. There are many other types but these are all we are interested in this course. Materials designed for theri luminescent properties are called "phosphors". These are old terms now, but alternaitive terms employed are flourescence radiation stop as soon as stimulouse stopped, and Phosphorescence refers to a material that radiates for a time after the stimulous is removed.

We need to consider from what point of veiw we are considering a material, Radiometry or Photometry.
Photometry - This is a measurement of radiation in terms of it's ability to produce the psychological sensation of brightness.
Radiometry - This is the purely physical measure of the radiation which provides the visual stimulous.

We refer to the efficiency of the visual system is teh ratio,

Figure one shows a graph of the human eye's efficiency for detecting light. At 555nm we have 1 Watt (radative unit) which is defined as 680 Lumins (photometric unit). Now we compare the relative units of measurement,

Radiometric units Photometric units
Power (flux), W Luminus power, lm
Irradience, W/m² Illuminance, lm/m² (lux)
Radient intensity, W/sr Luminous intensity, lm/sr (Candela, cd)
Radience, W/m²sr Luminance, lm/m²sr (cd/m²)

It is the meausre of luminance which is really refered to as brightness.

This can be defined in different ways, eg,

Where B_o is the luminescence of the object and b_s is the luminescence of the surroundings.

The eye is capable of detecting differences corresponding to a contrast of ratio of 1.03. (but for white symbols in a black background, a ratio 5 is considered to be the minimum acceptable).

Observation of "flicker" depends on frequency. Critical fusion frequency (CFF) is the frequency of brightness variation for which no flicker is observed by the human eye (for the human eye, CFF is between 20 Hz and 50Hz, depending on brightness.).

Such a display may have the form of a seven segment system (in the configuration of a "figure 8"). The devices associated with each segment share a common anode (or cathode) and there are separate connections to each of the corresponding cathodes (or anodes). Therefore any number with n digits can be formed by repeating this pattern n times and by addressing (7n+1) electrodes.

In order to reduce the number of connections use can made of the coordinate (or matrix) connected method. Each common element in each digit are connected to a common input line, then we can selectively set the anode line

All the anodes of elements in each row(or coloumn) are connected together. Similarly, all the cathodes in each coloumn (row). Therefore a system with (m×n) elements requires only (m+n) connections.
eg, for an array of 12 digits in 7 segmetns format simeple "direct drive" requires the format,

but matrix connection requires 7+12=19 leads.

With matrix format, the required display patter cannot be provided by appling fixed voltages to the leads. Timesharing (or multiplexing) is employed. The required voltage, V, is applied to a particular row and the necessary signal is applied simultaneously to each of the coloumns (depending on whether that particular segment should be 'on' or 'off'). After a time, t, voltage, V, is switched to the next row and a new set of signals are applied to the coloumns. If there are n rows, a time nt will be required to produce the full pattern. To avoid flicker it is necessary for nt to be less than about 1/40 seconds. The observed brightness of each elements will be less than would be the case inder "direct drive" as it will be in the on state for only 1/n of the total 'on' time. Therefore it is necessary for the display to be "overdriven" in order to increase it's time average brightness.

Under matrix conncection each element when in the 'off' state. Ideally there should be no output from such elements. The ideal characteristic for matrix address display devices is shown in figure 4.

Alternatively, use of "memory" mode operation can be made.

(See sheet)

High voltages are required for both the beam acceleration and deflective processes.

High energy electrons lose energy in phospher by ionisation, with production of electron-hole pairs. The depth of penetration of electrons with an increase in the electon beam energy (At 10 KeV, depth is ~ 1µm in 2ns). The Luminescence efficiency with beam energy. (At low beam energy electron-hole pairs are created near the surface of phosphor where there are many non-radiative centres).

The screen is divided into 625 lines (as per the European standard). To avoid flicker , the full screen must be scanned at a rate equal to or greater than 40 Hz (the fusion frequency). This requirement is eased in prcatice by use of "double scanning": Lines 1,3,5,7 and so on are scanned in a time which half of the time to scan the whole screen. The lines 2,4,6,8 and so on are scanned at half the period. The eye responds as if repeat rate is 2/t Hz instead of 1/t Hz. Therefore, a repeat time of 1/25 s can be used (this is the European standard).

Variation in birghtness is determined by bean current. The phosphor must have a decay time less than t to avoid "streaking" (due to image persistence, and where t is the period of full screen scan).

A type of flat CRT proposed was the Aiken system,

Another flat CRT was the Gabor system,

Line scan is provided deflection plates in the usual way. 180 degree beam reversal provides improved utilisation of flat plate area.

Both designs requires large numbers of plates to be switched at high voltages (+ high rate).

If used in conjunction with a channel multiplier plate, the switching can be performed at low voltages (the whole system need be only few cm deep).

Frame plates provided with a ramp voltage which steps from plate to plate in sequence. As each strip is raised to 400V the beam moves further to the right before being defleted.

The gas discharge process:
Consider the current flowing between two electrodes, in a at low pressure.

As the electric field is increased, any charged particles gain sufficient energy to ionize gas atoms or molecules. Ions striking the cathode cause liberation of electrons - these are accelerated and cause further ionization. The mixture of ions and electrons are known as plasma. Radiation from plasma is due to electron transitions between excited states and lower energy states of gas atoms. Characteristic of the gas atom involved deterimines the wavelengths of the emission, ie the colour (Neon is a commonly employed gas).

In a vacuum we would expect the voltage to vary linearly across the container. However, if the container is full of a plasma, the voltage increases quickly (and hence most excitations occur) at the cathode and reaches a high value, the voltage does not increase as much as we pass the middle and towards the anode. This is known as the cathode glow.

Structure and operation of DC systems. Electron systems deposited onto glass. The cathode defines areas of illumination. (see sheet). Such devices have a thick metallic film (eg Nickel, it has a low work function). Anodes may be a conducting/transparent layer (eg ITO, Indium Tin Oxide). Two galss plates sealed together with spacer keeping electrodes approximately 0.5 mm apart. The space between the electrodes is normally evacuated and then filled with gas at low pressure (100-300 Torr).

Structure and operation of AC systems. Electrode construction on galss plates as for DC panels but also material used for cathode not so important. Electrode systems subsequently are coated with an insulating material (eg SiO₂) plus further protective layer of MgO. Plates are sealed together with a spacer (~0.1mm), the gas pressure is 300-600 Torr. Operation involves a "memory" function.

They lay out of such a system is shown in fig 1.

However in practice the construction is like in fig 2.

The overall cell voltage is contributed to by several sources. There is the supply voltage called the sustaining voltage and then there is what is called the wall voltage due to the wall charge. The wall charge comes about because as the external field applied to the cell causes a build up of charge on the side of the insulator we get an internal field due to the wall charge. Hence,

If V_c>V_d then the discharge occurs but build-up of the wall charge rapidly causes extinction of it. V_d is the discharge voltage. The sustaining voltage is never larger than the discharge voltage, hence an extra pulse is required to initiate the process. The extra pulse comes by means of the sum of the sustaining voltage and the wall voltage (see sheet).

The memory function is simply that once a cell has been started it keeps going. So seemingly remembering it was actived. An opposite pulse has to be used to stop the process running, this would remove the charge build up on the insulators.

These are devices which operate using cathode rays. It was found (in the 1960's in japan) that certain materials would flouresce with lower energy cathode rays.

ZnO:Zn G/B 5-7 L/W
ZnS:Mn Yel. 0.5 L/W
ZnCl:Ag Red 0.5 L/W

EL luminescence arising from the application of an electric field. This has been observed in single crystals and in polycrystalline thin films placed between suitable electrodes under alternating current conditions. Also observed in carbon P-N junctions under forward or reverse bias.

EL in the above cases cannot be explained in terms of a single mechanism. Different types of excitation process are responsible for the excitation in the structures referred to above.

1) Quantum Mechanical tunneling

Figure 5 shows the energy routes for excitations of electrons. Figure six shows the effect of applying a potential difference across the region. As is shown there are non-horizontal energy lines, so electrons can gain kinetic energy. For a high potential gradient, it is possible for electrons to tunnel through from the fermi energy level to the conduction bandm, see fig 7.

The photon emission, obviously occurs as an excited electron transits from the conduction band to the valence band and recombines with a hole.

2) Avalanching (impact ionization)

The electron reaches an energy sufficient to excite an electron from the valence band

The hole becomes trapped

The conduction band electron makes a radiative transition.

3) Minority carrier injection

figure 11 shows an unbiased PN junction.

Now, figure 12 shows a forward biased junction. Injected minority carriers recombine with majority carriers.

It is the recombination of the minority carriers that we depend on for photon emission.

Electroluminescence was first discovered in 1936 using ZnS phoshor (powder) held in an insulating "binder" clampled between two parallel plates.

Under AC (~ several hundred volts) ight is emitted in pulses (1 for each ½ cycle).

1) Different colours can be obtained by using different impurities (e.g. rare earths: Sm for red, Tb for green and Tb for blue but luminescent efficiencies are not as high as for Mn doped ZnS).

2) Recently, interest has been shown in other EL materials (eg CaS, SrS).

Figure 1 shows a schematic of a typial LED.
Under forward bias, the current density crossing the junction is given by,

(Where D is the diffusion coefficient, L is the coefficient of dopant concentration fall-off. p_n and n_p arre minority carrier concentration)
Light generated on the side of the junction closest to the surface has a greater chance of being emitted rather than re-absorbed.
MOst of the currrent crossing the junction should be due to electron injection into the p-type material (for the case shown in figure one).
The fraction of the total cureent due to electron injection is,

For many materials used in LEDs, D_n>>D_p and it can be arranged that p_n<<n_p, by making the n-type material more heavily doped than the p side (ie n⁺-p diode). Therefore, I_e~=I_o.

The problem lies with total internal reflection.

Light originating at recombination centres near to the p-n junction will be radiated isotropically but only a fraction of this can penetrate the surface.
According to Snell's law n₁sin(theta₁)=n₂sin(theta₂) when n₁ and n₂ are the refractive indicies in materials 1 and 2 (n₂<n₁).
The critical angle occurs when theta₂=90 degrees. ie,

Therefore the light within a cone with a planar angle of > theta₂ is transmitted.

The only light emitted will be that within the cone which has a planar angle of >theta_c.
Therefore, the fraction, F, is given by area of sphere (radius r) cut by the cone divided by the total area of the sphere.

Figure demonstrates how we use geometry to break down and calculate F.
Consider the circular strip of radius r.sin(theta). The width of this strip is r.dtheta. Therefore, the area of the strip is,

Therefore the area cut by the cone of angle 2.theta_c is,

Therefore,

It has been shown that theta_c=sin^-1(n₂/n₁). For LED material with n=3.4 in air,

so that theta_c~=17 degrees. Therefore light escaping fraction F is,

Emitted light - wavelength requirement.

For transitions between the bottom of the CB and the top of the VB, energy of emitted photon is,

For transitions involving energy levels in the band gap,

for visible radiation, we require lambda in the range 400-700 nm (ie 3.1 eV to 1.8 eV).

We have to consider another factor now, the effect of the nature of the band gap, the band gap can be wither direct of indirect. The need for for concentration of momentum (aswell as energy)) has an effect on recombinaton transitions.
For direct band gaps, then conduction band minimum and valence band maximum (on an E-K plot, see semiconductors) lies along the same k value, this means corresponding carriers have the same momentum.
For indirect band gaps the conduction band minimum and valence band maximum lie at different k value, ie different momentums states for corresponding charge carriers.

We wish now to determine whether the above factors affect the recombination process.

1) Interband transitions
We have to require conservation of energy and momentum. For indirect band gaps, interband transitions means that the momentum of a charge carriers have to change. Momentum is expressed as shown below,

The wave vector (k) for and emitted phton in the visible sprectrum is,

The wave vectors for the electrons involved are in the range, -pi/a to +pi/a where is the crystal lattice spacing (a<=1nm).
As 2pi/lambda<<pi/a a transition involving only an electron and a phton will appear as a nearly vertical transition on an E-K diagram (ie a direct transition).

Non-vertical transitions are possible but these must involve phonons in order to conserve momentum, or k. Phonon energy is very small so that photon energy is still close to the band gap energy, E_g.
But with three particles involved insstead of two, such indirect transitions are less probable.

2) Impurity centre transitions.
When an electron is trapped (localised) it can be seen to have a range of momentum values. The Heisenburg uncertainty principal states,

Where dx is the atomic spacing.

It has been found that as we increasingly dope GaAs with more and more GaP, that is goes from being a direct band gap mateial to an indirect band gap materials.

For GaAs_1-yP_y with y<0.48, there is a direct band gap and the quantum efficiency for radiative transitions is high.
(Quantum efficiency is bumber of photons generated per electron/hole pair).
For y>0.48, the material has an indirect badn gap and the readiative efficiency is reduced because the lattice interactions must participate in the process to conserve momenutm.
For indirect band gap materials, additional recombination centres are incorporated to increase the probability of radiatve transitions.
In GaAsP, incorporation of nitorgen is succesful (Nitrogen is from the same group as for P. If substitutions for P but the different electronic core structure remains in the formation of an electron "trap" just below the conduction band - an interelectronic centre).

(Quantum efficiency = number of photons generated per electron/hole pair)

LCDs are passive systems, they operate on the the principal of modulation of ambiant light in reflection of transmission modes.

LCD cells consist of a layer of liquid crystals (~ 10µm thick) between two glass plates coated with conductive, transparent layers.

The liquid crystal state is a phase of matter exhibitted by a wide range of organic substances. These are composed of molecules which str highly non-spherical (e.g. rod shape).
The molecules can take up particular orientations relative to each other and relative to surfaces with which they are in contact (ie. they becomne ordered). This ordered phase only exists over a particular temperature range. At lower temperatures the material is solid at higher temperature the normal liquid phase exists. There are three broadly different types of liquid crystals: Nematic , Cholerteric , Smetic.

The Nematic phase is the most widely used for device purposes. Nemtic molecules tend to lay parallel to each other but random along each other, there are two types. The homeotropic type lie perpendicular to the surface of the materal, homogeneous lie parallel to the surface of the material.

Smetic molecules lie parallel to each other too, but they are also layered too in thier order. There also the homeotropic and homogeneous varieties too.

Cholerteric molecule show similar layered orientations as Smetic, but each layer is not alignd to other layers. The alignent of the molecules is twisted through space, there is a characteristic depth for a complete twist to occur. A complete twist always takes the same number of molecule layers. This characteristic is called the "characteristic pitch".

Liquid crystals are anisotropic. eg, a dielectric constant can be defined for parallel and perpendicular directions relative to the molecular axis.
If dE=E_par-E_per is positive an electric field will cause molecules to align parallel to the field direction. Similarly, i dE is negative to molecules will align perpendicular to an electric field. Typically dE values range from -0.4 to +10.

E_c is the critical E-field for which the majority of molecules are aligned parallel to the field direction.

(See sheet)

Structure: A nematic material (dE positive) contained between two surfaces treated to produce the homogeneous state but the alignment directions on the two surfaces are mutually perpendicular so that molecular alignment undergoes a 90 degree rotation between the plates.

Effect on light: When polarised light is incident, the strong optical anisotropy causes polarization to undergo a 90 degree rotation (for E<E_c) (see bottom of sheet again).

Figure 4 shows a crosss section for a twisted cell,

1) Require good brightness-voltage discrimination

2) To avoid "flicker", signal repeat rate => 30 Hz with n display lines, device "on" for only 1/n of scan time, T. Therefore we must have fast response time (< T/n).

3) Require high brightness, as time-averaged brightness is much less than steady state.

Clearly factors (ii) and (iii) limit the number of display elements which can be included in the system.

A dichroic dye (the guest) is mixed with the liquid crystal (the host). The dye molecules tend to align with the liquid crystal molecules so that thier alignment can be changed by an applied field. As the light absorption of the dye is dependant on orientation, colour changes can be produced. Different types of GH cell are possible depending on types of LC and types of dye.

eg, GH cell using nematic LC with homogeneous orientation.

To address a particular element, the transistor gate is "opened" by applying suitable voltage to appropriate y line the voltage on the corresponding x line then appears across the device (common connection to all elements is not shown in fig 1). As the LCD consumes very little current, the element will remain on for some time after the FET is switched off.

Each pixel circuit can be represented by a combination of capacitance C (due to the LC device element (LCD!)) and a resistance R associated with the FET) in series. The array is addressed by applying a voltage pulse sequentially to each gate line (which reduces the effective resistance of the FET). The required charge must be transfered to the LCD during this pulse (while the transistor is on). Subsequently, this charge is required to remain on the LCD pixel (without loss) for the remainder of the frame time.

Required properties of FETs
Consider a display consisting of 100 lines. Let full frame time is T=20ms. Therefore, t, address time for each line =20e-3/100=0.2ms. If FET resisitance when on is R_on, the time constant for the circuit is then R_onC. For C to be fully charged, it is required that,

Say t~>5R_onC. Typically C~=1pF, therefor R_on<0.2e-3/5e-12= 4e7 Ohms.

For negligible loss of charge during the off period.

Therefore,