The optical source is often considered as an active device in an optical fiber system. The basic function of an optical sourc is to convert electrical energy into optical or light energy. The light put is then coupled into the optical fiber. These optical beams generated by the light source carry the information. In majority of systems, information is put onto the beam by modulating the source input current. Most common source is light emitting diode

Light emission can take place trough two fundamental process known as spontaneous emission and stimulated emission. In the case of spontaneous emission, photons_are emitted in random directions with no phase relationship among them. In stimulated emission, the emitted light is imtiate by an existing photon. The unique characteristic of stimulated emission is that the emitted photon matches the original photon not only in energy but also in its other characteristics, such as the direction of propagation.

Laser emit light through stimulated emission and in contrast LED emits light through emission. Three fundamental processes that can occur between the two energy states of an atom can be seen in figure 5.1

Energy levels E₁ and E₂ correspond to the ground state and excited state of the material. Therefore E₁ and E₂ represent low and high energy level, respectively. The photon energy is given by hf where h=6.626 x 10³⁴ Js is the Planck's constant and f is the frequency.

If electrons in the low energy level are supplied with energy it will move into higher energy level. This is called absorption. If it then drops from high to lower energy level, it will emit photon as the energy lost and this is known as spontaneous emission. If an emitted photon interacts with the atom and generates further photons, the emission is known as stimulated emission.

5.1 Light emitting diode

It is a p-n junction that emits light when forward biased. In a forward biased p-n junction, electrons and holes are injected into the active region, where they recombine to produce light. The electrons recombine with holes on the p-type side, and holes recombine with electrons on the n-type side. The energy lost in the recombination process is given off as light. This is also known as radiative recombination. The emitted light is incoherent, with a relatively wide spectral width and a large angular spread.

By carefully choosing the p and n type materials, it is possible to determine the exact wavelength of the emitted light. This is called as spontaneous emission or electroluminescene. The wavelength of emission is, ~ = (l.²⁴/Eg) ~m where Eg = gap energy in eV. Different material and alloys have different bandgap energies.

Material (~m) Eg (eV)

GamP 0.64-0.68 1.82-1.94

G~s 0.9 1.4

AlGaAs 0.8 - .09 1.4-1.55

InGaAs 1.0-1.3 0.95-1.24

InGaAsP 0.9-1.7 0.73-1.35

Table 5.2 Materials for fabrication of light source with respective emission

wavelength.

The operating wavelength can be chosen for the different materials by varying the proportions of the constituent atoms. The semiconductor materials are made transparent so as not to trap the photons generated.

Different junction profiles can be used in constructing a p-n junction.

1) Homojunction

2) Single heterciunction (SH)

3) Double heterojunction (DH)

A homojunction LED does not confine its emitted radiation very well. Photons radiate from the edges of the junction and from its large planar surface. This makes coupling into small fibers very inefficient. Reasons for this: (a) charge carriers exist over a large area, causing recombination and emission over an extensive region and (b) after the photons are generated they diverge over unrestricted paths. Utilizing heterojunction profiles can solve these problems. These different profiles (SH and DH) are utilized to increase recombination by narrowing the active layer (layer where recombination take place).

Heterojunction is a junction formed by dissimilar semiconductors. The two materials have different bandgap energies and different refractive indices. The changes in bandgap energies create potential barriers for both holes and electrons. The free charges can only meet and recombine in the narrow, well-defined active layer. This is called as carrier confinement. Because the active region has a higher refractive index than the materials on either side, an optic waveguide is formed. Therefore, the light or photons generated will be confined within this region of high refractive index. This is known as optical confinement. The confined emission improves the coupling efficiency, particularly for small fibers. Figure 5.3 Show the different junction pro files used in constructing optical sources.

There are number of ways in which the operational efficiency of the semiconductor LED or Laser may be defined. For example internal quantum efficiency, Tiint is the ratio of photons generated within the semiconductor material to the injected electron. ilint for homojunction LED is = 50%. The total efficiency or the external quantum efficiency, Ti is the ratio of the total number of output photons to the total number of injected electrons.

(1) Optical power generated internally (for a forward bias current I the output power is P~~t) is given by,

                                        I hcl

            P_int =h _int—hf = h _int ---

e el

 

(2) The total efficiency is given by,

h = (P_out/hf)/(IIe)

= (P_out)/(IE_g)

 

(3) The external power efficiency of the device (or device efficiency) Tiep in converting electrical input to optical output is,

Tiep = POUt/P ; P IV which is the d.c. electrical input. = Ti[EgJV]

(4) The optical power emitted into a medium of low refractive index n from the LED. fabricated from a material with refractive index n~ is given approximately by,

POUt = PoutFn²/4n~² ;

where F is the transmission factor of the semiconductor-external interface. Therefore it is possible to estimate the percentage of optical power emitted.

(5) The coupling efficiency of the light from an LED into a step index fiber with numerical aperture NA is,

h _c = (NA)².

 

See Example 5.1

The light output from a GaAs LED is coupled into a step index fiber with a numerical aperture of 0.2, a core refractive index of 1.4 and a diameter larger than the diameter of the device. Assume GaAs to have a refractive index of 3.6 and the transmission factor at the GaAs-air boundary to be 0.68. Estimate:

(a) The coupling efficiency into the fiber when the LED is in close proximity to the fiber core.

(b) The optical loss in dB, relative to the power emitted from the LED, when coupling the light output into the fiber.

(c) The loss relative to the internally generated optical power in the device when coupling the light output into the fiber when there is a small air gap between the LED and the fiber core.

Solution

(a) The coupling efficiency is h _c = (NA)² =0.22 =0.04

(b) Assuming the power coupled into the fiber to be PC and the output power from the LED as POUt, the optical loss in relative to coupling the light into the fiber is PC/ POUt.

The loss in dB is = -10 log PC/ POUt. =-lOlog~C

                                        = 14OB.

(c) When the light is emtting into air,

            2 2

POUt = PintFn /4nx = Pintt0.68(1)2/4(3.6)2 = 0.013 Pint.

 

Assuming a very small air gap, then from (a) the power coupled into the fiber is,

PC =0.04 POUt =0.04 x 0.013 Pint =5.2 x lO⁴Pint

                Loss = -10 log PC/ Pint =32.8 dB

There are five major types of LED. They are Surface Emitter, Edge Emitter, Superluminescent LED, Planar and Dome. The first three are used extensively for optical communications.

5.2.1 Planar LED

This is the simplest structure. This involves a p type diffusion into the n type substrate in order to create the junction. Forward current flow through the junction gives Lambertian spontaneous emission and the device emits light from all surface. In a Lambertian source the power diminishes as cos 0 where 0 is the angle between the viewing direction and the normal surface. However, only a limited amount of light escapes the structure due to total internal reflection, therefore the radiance is low. Figure 5.4 Show the structure of a planar LED.

                5.2.2 Dome LED

The structure is a hemisphere of n-type GaAs formed around a diffused p-type region. The diameter of the dome is chosen to maximize the amount of internal emission reaching the surface within the critical angle of the GaAs-air interface, so that the rays reaching it never exceeds the critical angle. The dome LED has a higher external power efficiency than the planar LED. However the geometry of structure is such that, the dome must be far larger than the active recombination area, which gives a greater effective emission area and thus reduces the radiance. Figure 5.5 Show the structure of a dome LED.

5.2.3 Surface emitter LEDs (SLED)

An example of a DH surface emitting LED is shown in figure 4.6 which is known as a Burrus-type LED. This AlGaAs diode emits at a typical wavelength of 820 nm. To obtain a high radiance output, it is essential to restrict the emission to a small active region within the device. The metal contact is circular and extends through a hole in the 5i0₂ layer. This confines injected charges to a small central portion of the diode. The structure has low thermal impedance in the active region allowing high current densities and giving high radiance emission into the optical fiber. The internal absorption in this device is very low due to the larger bandgap confining layer, and the reflection coefficient at the back crystal face is high giving good forward radiance. The power coupled, PC into the multimode step index fiber may be estimated from the formula,

P_c = (1 - r)AR_D(NA)²

where,

r = Fresnel reflection coefficient at the fiber surface(r = [(n₁ - n)/(n₁ + n)]²).

A = the smaller of the fiher core cross section or the emission area of the source.

RD = radiance of the source.

Methods used to increase the coupling efficiency:

1) Spherical-ended fiber - It consists of a planar emitting structure with a spherical ended fiber attached to the top by epoxy resin. It is essential that the active diameter of the device be less than the fiber core diameter for an increased coupling. The coupling efficiency can be increased 2 to 5 times.

2) Spherical lens between the fiber and source - for efficient coupling, the emission region diameter must be much smaller than the core diameter of the fiber.

3) Integral lens structure on the LED - a favored power coupling strategy for use in surface emitters. In this technique a low absorption lens is formed at the exit face at the substrate material instead of it being fabricated in glass and attached to a planar SLED with epoxy.

4) Tapered GRIN-rod - have been extensively used to couple power from ELEDs to single mode fibers. The small core diameter of the single mode fiber does not allow significant lens coupling gain to be achieved. Therefore a tapered GRIN-rod is positioned between a high power ELED and the fiber.

Figure 4.6 Show the Etched-well, surface emitting LED (Burrus type).

           5.2.4 Edge Emitter LEDs (ELED)

This device has a similar geometry to a conventional contact stripe injection laser. It takes advantage of transparent guiding layers with a very thin active layer (50 to 100 ~m) in order that the light produced in the active layer spread into the transparent guiding layers, reducing self absorption in the active layer. Is a Lambertian source, but the divergence in the plane perpendicular to the junction is small compared to planar LED. Typical beam divergence is 300 in the plane perpendicular to the junction and 1200 in the plane of the junction. Most of the propagating light is emitted at one end face only due to a reflector on the other end face and an antireflection coating on the emitting end face. The coupling efficiency into small numerical aperture fiber is higher in comparison to surface emitter. These LEDs can be made as DH and can achieve very high modulation bandwidth and can be used with SMF.

5.2.5 Superluminescent LED (SLD)

This device offers advantages of: (a) a high output power; (b) a directional output beam; and (c) a narrow linewidth/spectral width. Its operation is very similar to an injection laser but without optical feedback. The injection current is increased until stimulated emission occurs, but because there is high loss at one end of the device, no optical feedback takes place. This allows the device to emit high optical power with a narrow spectral width. There are SLDs which can emit 1 mW power into a 10 ~m SMF with spectral width as low as 30 nm and modulation bandwidth of 350 MHz. The drawback compared with conventional LEDs is the non-linear output characteristics and the increased dependence on temperature of the output.

  5.3 LED Characteristic

5.3.1 Output Power

Power generated is proportional to the forward driving current. The output is more linear than laser, therefore it is very suitable for analog modulation. Since the light is generated through spontaneous emission, the output is incoherent and act as a Lambertian source. Normally, the launched power for LEDs is 100 ~W or less, even though the internal power can easily exceed 10 mW. The ELED exhibits a greater temperature dependence than the SLED and the output of SLD is strongly dependent on the junction temperature.

5.3.2 Output Spectrum

For LED operating between 0.8 - 0.9 ~m, the linewidth is between 25 - 40nm (full width half power point-FWHP). For those operating between 1.1 - 1.7 ~m, the spectral~ width is between 50 - 1 60nm. The increased spectral width of a longer wavelength emitter is compensated by a reduction in material dispersion. The output spectrum also broadens with increase in temperature.

5.3.3 Modulation Response

Normally we use the electrical bandwidth (BW) considering the electrical circuitry involved in optical communication. It is the bandwidth when the electrical signal power has dropped to half ("ower is square of current, therefore (0.707)2 = 0.50) its constant value due to the modulated portion of the optical signal. At 3dB point output of electrical power is reduce by 3 dB with respect to the input electrical power. As optical source operate down to d.c. we only consider that high frequency 3dB point, the modulation bandwidth being the frequency range between zero and this high frequency 3dB point.

Assuming a Gaussian frequency response,

B=f/42

Where, B = electrical BW and f= optical BW

 

Modulation BW of LED is determined by:

1) Doping level in the active layer

2) Carrier lifetime r, the average time for injected charges to recombine.

3) Parasitic capacitance of the device

The major limitations is the carrier lifetime. The ratio of the output power,POUt(o)) and Pdc the d.c. optical output power for the same drive current is;

POUt(O))/PdC = 1/ v(1 + (w t )²

 

Example 5.2

The minority carrier recombination lifetime for an LED is Sns. When a constant dc drive current is applied to the device the optical output power is 0.3 mW.

Determine the optical power when the device is modulated with an rms drive current corresponding to the dc drive current at frequency a) 20 MHz and b) 100 MHz.

Neglecting any parasitic capacitance determine the 3-CIB optical BW. If the frequency response is Gaussian, estimate the electrical BW (3 dB).

Solution

a)Pout(w ) =Pdc/V(1 +(w t )

Pe(20 MHZ) =300x10⁶/V(1 +2~x5x10⁹x20x]⁰⁶)2 =254.2m W

b) Pe(1OOMHz) = 300x 106/ V(1 + 2~x Sx 10⁹x JOOx 106)2 = 9O.9m W

To determine the 3-dB optical BW the 3-dB point occurs when Pe(w )/Pdc = 1/2, therefore,

f=v3 /(2p t )

=v3 /(2p t x 5x10⁹ )=55.1MHz

The electrical bandwidth, B =f / v2 = 39.0 MHz.

Using another approximation, B = 1/(2p t ) = 1/2p x 5ns = 32 MHz.

         5.3.4 Reliability

LEDs are very reliable and lifetime of i0⁵ hours (11 years) are common for good LEDs if operated within the limits ~ower, voltage, current and temperature). Lifetime is the time it takes for the power to reduce to half its initial value. Output power is also function of temperature and a typical decrease is 0.012 dB/⁰C.