ecse-6290 semiconductor devices and models ii spring, 2010 s. sawyer 1-1 ecse-6290 semiconductor...

ECSE-6290 Semiconductor Devices and Models II Spring, 2010 S. Sawyer 1-1

ECSE-6290Semiconductor Devices and Models II

Lecture 20: Laser Diodes

Shayla M. Sawyer

Bldg. CII, Room 8225

Rensselaer Polytechnic Institute

Troy, NY 12180-3590

Tel. (518)276-2164

FAX (518)276-2990

e-mail: [email protected]


Lecture Outline

• Introduction• Main Concepts

– Stimulated Emission– Population Inversion– Optical Gain– Optical Resonator– Threshold Current

• Laser Diode Types


Concepts: Population Inversion•Switch from two energy levels to two separate continuous bands•Electron concentration as a function of energy, determined by Fermi-Dirac distribution and Density of States

EquilibriumPopulation Inversion

T=0K

Population Inversion T>0K


hEg

Optical gain EF n EF p

Optical absorption

0

Energy

Ec

Ev

CB

VB

(a) The density of states and energy distribution of electrons and holes inthe conduction and valence bands respectively at T 0 in the SCLunder forward bias such that EFn EFp > Eg. Holes in the VB are emptystates. (b) Gain vs. photon energy.

Density of states

Electronsin CB

Holes in VB= Empty states

EF n

EF p

eV

At T > 0

At T = 0

(a) (b)

© 1999 S.O. Kasap, Optoelectronics (Prentice Hall)

Concepts: Population Inversion and Optical Gain

Photons above EFn-EFp are absorbed


p+ n+

EF n

(a)

Eg

Ev

Ec

Ev

Ho les in V BElectro ns in C B

Junction

Electro nsEc

p+

Eg

V

n+

(b)

EF n

eV

EF p

The energy band diagram of a degenerately doped p-n with no bias. (b) Banddiagram with a sufficiently large forward bias to cause population inversion andhence stimulated emission.

In v ers io nreg io n

EF p

Ec

Ec

eVo


Concepts: Population Inversion and Optical Gain


Laser Physics: Population Inversion•Necessary requirements and conditions for lasing

•In order for equation below to be positive FC>FV and EFn>EFp (population inversion)

•Photon energy must be larger than the bandgap•For current pumped laser diode the quantity (EFn-EFp) is equal to the bias voltage, bias is limited to the built up potential of the junction (ψBn+ ψBp)

R st R ab B 21 EN ph F C F V N c N v

d

E g Bn Bp q

For homojunction: one

side must be doped to degeneracy

FC E( )1

1 expE EFn

kT

FV E( )1

1 expE EFp

kT

Fermi-Dirac


LElectrode

Current

GaAs

GaAsn+

p+

Cleaved surface mirror

Electrode

Active region(stimulated emission region)

A schematic illustration of a GaAs homojunction laserdiode. The cleaved surfaces act as reflecting mirrors.

L


Concepts: Optical Resonator

• Need to build up stimulated emissions by a optical resonator

• Provided by cleaved and polished ends of the crystal


Concepts: Optical Resonator•Requirement is a structure to trap the light and build up intensity inside

•Like a Fabry-Perot etalon, two parallel walls perpendicular to the junction•Longitudinal modes: multiple resonant frequencies

•Separation of modes in wavelength and frequency

m

2 n r

L

ddm

m

2

2 L n r

m

vc

2 L n r

m

Only multiples of the half wavelength can exist in the cavity

Output spectrum is determined by the optical cavity, and optical gain vs.

wavelength characteristics


Concepts: Optical Gain•Optical gain (g) due to stimulated emission is compensated by optical loss due to absorption (α)

•Net gain/loss as a function of distance

•For a given system R1, R2, and α are fixed, the only parameter to vary is overall gain•To keep gain positive

z( ) exp g Z

R 1 R 2 exp g 2 L 1

g th 1

2 Lln

1

R 1 R 2

Threshold gain


Concepts: Threshold Current• The relationship between optical gain and bias current can be described by the equation

Linear increase of optical gain with bias current

gg 0

J 0

J in

dJ 0

=

Nominal current density

Threshold value


Typical output optical power vs. diode current (I) characteristics and the correspondingoutput spectrum of a laser diode.

Laser

LaserOptical Power

Optical Power

I0

LEDOptical Power

Ith

Spontaneousemission

Stimulatedemission

Optical Power


Concepts: Threshold Current

• Lasing oscillations occur only when the optical gain in the medium can overcome the photon losses from the cavity

• Cavity modes vary with increasing current


Laser Types

homostructure single heterostructure

double heterostructure


Refractiveindex

Photondensity

Activeregion

n ~ 5%

2 eV

Holes in VB

Electrons in CB

AlGaAsAlGaAs

1.4 eV

Ec

Ev

Ec

Ev

(a)

(b)

pn p

Ec

(a) A doubleheterostructure diode hastwo junctions which arebetween two differentbandgap semiconductors(GaAs and AlGaAs).

2 eV

(b) Simplified energyband diagram under alarge forward bias.Lasing recombinationtakes place in the p-GaAs layer, theactive layer

(~0.1 m)

(c) Higher bandgapmaterials have alower refractiveindex

(d) AlGaAs layersprovide lateral opticalconfinement.

(c)

(d)


GaAs

Both Carrier and Optical Confinement!


Example: Modes in an optical cavity length

Consider an AlGaAs based heterostructure laser diode which has an optical cavity of length 200 microns. The peak radiation is at 870 nm and the refractive index of GaAs is about 3.7.

a) What is the mode integer m of the peak radiation and the separation between modes of the cavity?

b) If the optical gain vs. wavelength characteristics has a FWHM wavelength width of about 6 nm, how many modes are there within this bandwidth?

c) How many modes are there if the cavity length is 20 μm?


Schematic illustration of the the structure of a double heterojunction stripecontact laser diode

Oxide insulator

Stripe electrode

SubstrateElectrode

Active region where J > Jth.(Emission region)

p-GaAs (Contacting layer)

n-GaAs (Substrate)

p-GaAs (Active layer)

Currentpaths

L

W

Cleaved reflecting surfaceEllipticallaserbeam

p-AlxGa

1-xAs (Confining layer)

n-AlxGa

1-xAs (Confining layer) 12 3

Cleaved reflecting surface

Substrate


• Stripe contact increases current density in the active region.

• The widths of the active region or the optical gain region is defined by current density from the stripe

Gain guided: optical gain is highest where

current density is greatest


Oxide insulation

n-AlGaAs

p+-AlGaAs (Contacting layer)

n-GaAs (Substrate)

p-GaAs (Active layer)n-AlGaAs (Confining layer)

p-AlGaAs (Confining layer)

Schematic illustration of the cross sectional structure of a buriedheterostructure laser diode.

Electrode


• Active layer is surrounded by lower index AlGaAs and behaves like a dielectric waveguide

• Ensures that photons are confined to the active or optical gain region

• Increases rate of stimulated emission

Index guided: optical power confined to waveguide


778 780 782

Po = 1 mW

Po = 5 mW

Relative optical power

(nm)

Po = 3 mW

Output spectra of lasing emission from an index guided LD.At sufficiently high diode currents corresponding to highoptical power, the operation becomes single mode. (Note:Relative power scale applies to each spectrum individually andnot between spectra)


•Index guided LD: changes from multiple mode to single mode with increasing optical power

•Gain guided: remain multimode even at high diode currents


Corrugateddielectric structure

Distributed Braggreflector

(a) (b)

AB

q(B/2n) =

Active layer

(a) Distributed Bragg reflection (DBR) laser principle. (b) Partially reflected wavesat the corrugations can only constitute a reflected wave when the wavelengthsatisfies the Bragg condition. Reflected waves A and B interfere constructive whenq(B/2n) = .


• Frequency selective dielectric mirrors a cleaved surfaces.

• Only allow a single mode to exist

• Periodic corrugated structure that interfere constructively when the wavelength corresponds to twice the corrugation periodicity (Bragg wavelengths)

Single Frequency Solid State Lasers: DBR laser


Active layer

Corrugated grating

Guiding layer

(a)

(a) Distributed feedback (DFB) laser structure. (b) Ideal lasing emission output. (c)Typical output spectrum from a DFB laser.

Optical power

(nm)

0.1 nm

Ideal lasing emission

B(b) (c)


• The corrugated layer, called the guiding layer, is now next to the active layer

• In the DFB structure traveling wave are reflected partially and periodically as they propogate.

Single Frequency Solid State Lasers: DFB laser


A quantum well (QW) device. (a) Schematic illustration of a quantum well (QW) structure in which athin layer of GaAs is sandwiched between two wider bandgap semiconductors (AlGaAs). (b) Theconduction electrons in the GaAs layer are confined (by ² Ec) in the x-direction to a small length d sothat their energy is quantized. (c) The density of states of a two-dimensional QW. The density of statesis constant at each quantized energy level.

AlGaAs AlGaAs

GaAs

yz

x

d

Ec

Ev

d

E1

E2

E3

g(E)Density of states

E

BulkQW

n = 1

Eg2 Eg1

E n = 2² E c

BulkQW

² E v

(a) (b) (c)

Dy

D z


• Constant 2D density of states means a large concentration of electron can easily occur at E1 (and holes at the minimum valence band energy)

• Population inversion occurs quickly without the need for a large current to bring a large number of electrons

• Benefits: Threshold current reduced, linewidth is narrower


Active layer Barrier layerEc

Ev

E

A multiple quantum well (MQW) structure.Electrons are injected by the forward currentinto active layers which are quantum wells.



A simplified schematic illustration of a vertical cavitysurface emitting laser (VCSEL).

Contact

Surface emission

Dielectric mirror

Contact

Substrate

/4n1

Active layer

/4n2 Dielectric mirror


•Optical cavity axis along the direction of current flow rather than perpendicular to current flow•Radiation emerges from the surface of the cavity rather than from its edge•Reflectors at the edges of the cavity are dielectric mirrors•20-30 layers for mirror, MQW active region


Summary• Laser concepts: Stimulated Emission,

Population Inversion, Optical Resonator, Optical Gain, and Laser Threshold Current

• Types of laser diodes include– Homostructure

– Single heterostructure

– Double heterostructure

– Gain guided (stripe geometry)

– Index guided (buried heterostructure)

– Distributed Bragg reflection

– Distributed feedback

– Quantum well, multiple quantum well

– Vertical cavity surface emitting


ECSE-6290Semiconductor Devices and Models II

Lecture 21: Photodetectors

Prof. Shayla M. Sawyer

Bldg. CII, Room 8225

Rensselaer Polytechnic Institute

Troy, NY 12180-3590

Tel. (518)276-2164

FAX (518)276-2990

e-mail: [email protected]


Lecture Outline• Introduction to Photodetectors

• Photodiodes– General

– p-i-n and p-n

– Metal-Semiconductor

– Avalanche

• Metal-Semiconductor-Metal Photodetector

• Quantum Well Infrared Photodetector• Summary


Introduction• Photodetectors are semiconductor devices that

can detect optical signals through electronic processes– Three main processes:

• Carrier generation by incident light

• Carrier transport and/or multiplication by current-gain mechanism

• Extraction of carriers as terminal current to provide the output signal

– Desired: High sensitivity, high response speed, minimum noise, compact size, low biasing voltage and current


Introduction• Wavelength relation to transition energy

• ΔE is the transition of energy levels– Depending on photodetector type can be:

• Energy gap of the semiconductor

• Barrier height as in a metal semiconductor photodiode

• Transition energy between impurity level and band edge as an extrinsic photoconductor

hc

E

1.24

E eV( )

Often minimum wavelength for detection


Introduction• Important Factors/Parameters

– Absorption Coefficient

– Response Speed

– Quantum Efficiency

– Responsivity

– Gain

– Noise

– Detectivity


Introduction• Absorption

coefficient– Determines whether

light can be absorbed for photoexcitation

– Determines where light is absorbed

• High value means near surface

• Low value means deeper penetration


Introduction• Response Speed

– Shorter carrier lifetime yields fast response at the expense of higher dark current (noise)

– Depletion width should be shortened to reduce transit time at the expense of capacitance

• Quantum Efficiency– Number of carriers produced per photon

– Iph is the photocurrent, Φ is the photon flux (=Popt/hv) and Popt is the optical power

I ph

q

I ph

q

hv

P opt


Introduction• Responsivity: Photocurrent generated per incident

optical power

• Gain and response time for common photodetectors

RI ph

P opt

q

hv

m 1.24

A

W


Introduction• Noise ultimately determines minimum

detectable signal strength– Sources of noise

• Dark current

• Thermal noise

• Shot noise

• Flicker noise

• Generation recombination noise

– Figure of Merit Noise Equivalent Power

– Detectivity A is the Area

B is the Bandwidth

D* AB

NEP

NEP-incident rms optical power required to produce a signal-to-noise ratio of one in a 1 Hz bandwidth (minimum detectable light power)


Introduction

• Detectivity– The signal-to-noise ratio when one watt of light power is

incident on a detector of area 1 cm2 measured over 1 Hz bandwidth

– Normalized to area, noise is generally proportional to the square root of area

– Detectivity depends on

• Detector sensitivity

• Spectral Response

• Noise

– Is a function of wavelength, modulation frequency and

bandwidth

D* AB

NEP


Photodiodes: General• Photodiodes have depleted

region with a high electric field that separates photogenerated electron-hole pairs

– Tradeoff between speed of response and quantum efficiency (depletion layer: transit time, absorbance area)

– Reverse biasing often employed to reduce carrier transit time and lower diode capacitance

– All photodiodes except Avalanche has a maximum gain of one

a) p-i-n photodiodes b) pn photodiode c) Metal-i-n photodiode d) Metal-semiconductor photodiode e) Point contact photodiode


Photodiodes: General• Important characteristics

– Quantum efficiency• Absorption coefficient strong dependence on

wavelength

• Long wavelength cutoff given by energy gap of semiconductor

• Short wavelength cutoff given by large value of α (surface where recombination is likely)

– Response Speed• Limited by

– Drift time in the depletion region

– Diffusion of carriers

– Capacitance of detection region

• Optimized when the depletion layer is chosen so the transit time is on the order of one half the modulation period

WD ~ 1/α


Photodiodes: General• Device Noise

– Shot noise• IP average photocurrent, IB background radiation, ID dark current due to

thermal generation of electron hole pairs in the depletion region

– Thermal noise where• Rj Junction resistance

• Ri Input resistance of amplifier

• RL External load resistor

Important characteristicsQuantum efficiency

optical absorption coefficient is a strong function of wavelengths so there is a limit to the appreciable amount of photocurrent for a given wavelengths

cutoff long wavelength is established by energy gap of the semiconductorshort wavelenth cutoff comes from values of that are large so radiatino is absorbed near the surface where

recombination is likely

Response SpeedLimited by drift time in teh depletion region, diffusion of carriers, and capacitance of detection regionMinimize diffusion effect junction should be formed close tho the surfacewidth of dpletion region must we wide but not too wide oro transit time effects will limit the frequency responseToo thin and excessive capacitance C will reust in a large RLC time constantOptimum compromise occurs when the depeltion layer is chose so tha tthe transit time is of the order of one half

the modulation period

Device Noise

Shot noise<i s

2 2 q I P I B I D B

<i T2

4kT BR eq

1

R eq

1

R j

1

R L

1

R i


Photodiodes: General• Signal to Noise for 100% modulated signal with

average power Popt

• Minimum optical power required to obtain a given signal-to-noise ratio is (setting Ip=0)

• Noise equivalent power (S/N=1; B=1 Hz)

S

N

i p2

<i s2

<i T2

1

2

q P opt

hv

2

2 q I P I B I D B4kTB

R eq

NEPhv

2 I eq

q

I eq I B I D2kT

qR eq

P optmin2 h

S

N

I eq B

q=


Photodiodes: p-i-n and p-n • Depletion layer thickness

(intrinsic layer) can be tailored to optimize the quantum efficiency and frequency response

• Total photocurrent density through reverse biased depletion layer

• Total current density is the sum of Idr inside the depletion region and Idiff outside the depletion region

J tot J dr J diff

J tot q 0 1exp W D

1 L p

q p no D p

L p


Photodiodes: p-i-n and p-n • Quantum efficiency

– Reduced from unity from• Reflection R

• Light absorbed outside the depletion region

– High quantum efficiency, low R and αWD>>1 is desirable

• For WD>>1/ α transit time delay may be considerable

AJ tot

q

P opt

hv

1 R( ) 1exp W D

1 L p


Photodiodes: p-i-n and p-n • Frequency Response

– Phase difference between photon flux and photocurrent will appear when incident light intensity is modulated rapidly

– Assume light is absorbed at surface, applied voltage is high enough to ensure saturation velocity

– Response time is limited by the carrier transit time through the depletion layer

– Compromise for high frequency response and quantum efficiency

– Absorption region of thickness 1/α to 2/α

– Large portion of light is absorbed within the depletion region


Photodiodes: p-i-n and p-n • Frequency Response

– 3-dB frequency

• Illustrates trade off between response speed and quantum efficiency at various wavelengths by adjusting the depletion width

• Smaller WD, shorter transit time, higher speed, but reduced η

f 3dB2.4

2 t r

0.4 s

W D0.4 s

Shows internal quantum efficiency of the Si p-i-n photodiode as a function of the 3-dB frequency and depletion width


Photodiodes: p-i-n and p-n • p-n photodiode

– Thin depletion layer means some light can be absorbed outside

– Light more than a diffusion length outside do not contribute at all to photocurrent

• Reduces quantum efficiency

• Diffusion process is slow– Time require to diffuse a

distance x

• Lower response speed than p-i-n

• Neutral region contributes to noise

t4 x

2

2

D p


Photodiodes: Heterojunction• Advantages

– Large bandgap material can be transparent and used as a window for transmission of incoming optical power

• Quantum efficiency is not dependent on distance of junction from surface

– Unique material combinations so quantum efficiency and response speed can be optimized for a given optical wavelength

– Reduced dark currentJ.H. Jang et al., Journal of Lightwave

Technology, Vol. 20, No. 3, March 2002.


Photodiodes: Metal-Semiconductor

• Operates in two modes– hυ>Eg : radiation produces electron hole pairs (similar to pin

photodiode)

– hυ<Eg : photoexcited electrons surmount barrier

Has a threshold of qΦB, when it gets to the energy gap value the quantum efficiency jumps to a much high value


Photodiodes: Metal-Semiconductor• Quantum Efficiency in two modes

– hυ>Eg : radiation produces electron hole pairs (similar to p-i-n)

– hυ<Eg : photoexcited electrons surmount barrier, internal photoemission

Internal photoemission has typical quantum efficiencies of less than 1%

AJ tot

q

P opt

hv

1 R( ) 1exp W D

1 L p

C F

h qB 2h

CF is the Fowler emission coefficient


Photodiodes: Metal-Semiconductor• Configurations

– Advantageous for band-to-band

• Diode illuminated through thin metal contact with antireflection coating

• Use low doping i layer similar to p-i-n

• Point contact diode reduces active volume, drift time and capacitance are small

– Very high modulation frequencies


Photodiodes: Metal-Semiconductor• Main advantages

– High speed and long wavelength detection capability without having to use a semiconductor with a small energy gap

– Not limited by charge storage of minority diffusion current

– Ultrafast Schottky barrier photodiodes beyond 100 GHz have been reported

– Useful in the visible and UV


Photodiodes: Avalanche • Operate at high reverse bias voltages where avalanche

multiplication takes place– Creates internal current gain

– Can respond to light modulated at microwave frequencies• Current gain-bandwidth product of an APD can be higher than 300

GHz

– High gain comes at the price of noise

• Low frequency avalanche gain

M 1

0

W D

x n exp

x

W D

x' n p

d

d

1αn and αp are electron and hole ionization rates


Photodiodes: Avalanche • For equal ionization coefficients (α =αn=αp)

multiplication takes the simple form

• In a practical device, the dc multiplication at high light intensities is limited by series resistance and space-charge effect

M1

1 W D

M ph

I I MD

I P I D1

VR IR s

VB

n

1

Breakdown when αWD=1

I total multiplied currentIP unmultiplied currentID unmultiplied dark currentIMD multiplied dark currentVR Reverse bias voltageVB Breakdown voltage


Photodiodes: Avalanche • Current gain mechanism

multiples the signal current, background current and dark current indiscriminately

• Signal to noise power ratio

S

N

1

2

q P opt

h

2

2 q I p I B I D F M( ) B4kTB

R eq M2


Photodiodes: Avalanche • The previous equation can

be solved for the minimum optical power Popt required to produce a given S/N with avalanche gain

• NEP can be improved by the reduction of Ieq by the gain M

I eq I B I D F M( )2kT

q R eq M2

P optmin2 h

S

N

I eq B

q=


Photodiodes: Avalanche • At high light power p-i-n

detectors have larger SNR than APD

• There is a significant advantage of APDs for low light signals due to multiplication factor

• Optimization required for any system depending on optical flux SNR comparison between 1 mm2

diode detectors with a bandwidth 50Hz


Photodiodes: Avalanche • Material comparison

– Heterojunction avalanche photodiodes of III-V alloys advantages

• Wavelength response can be tuned

• High absorption coefficients of direct bandgap III-V alloys makes quantum efficiency high even if a narrow depletion width is used to provide a high speed response

• Heterostructure window layer can be grown

– Surface recombination loss minimized

– High speed performance


Photodiodes: Avalanche • Material comparison

– AlGaAs/GaAs, AlGaSb/GaSb, InGaAs/InP, and InGaAsP/InP

• Shown improvements in speed and quantum efficiency over Si and Ge

– Separation absorption and multiplication

• Higher bandgap material in the multiplication region

• Low bandgap material for light absorption

• Breakdown voltage is expected to vary as Eg

2/3 so dark current due to tunneling is suppressed


Metal-Semiconductor-Metal Photodetectors• Basically two Schottky barriers

connected back-to-back on a coplanar surface

• Added a thin barrier enhancement layer to reduce dark current

• Metal contacts have shape of interdigitated stripes– Light is received at the gap

between metal contacts

– For more complete light absorption, active layer has a thickness slightly larger than absorption length


Metal-Semiconductor-Metal Photodetectors• Typical operation

– Photocurrent first rises with voltage then becomes saturated

– Increase of photocurrent at low bias is due to expansion of the depletion region in the reverse biased Schottky junction and internal quantum efficiency improves

– Voltage at which photocurrent saturates corresponds to flat band condition

• Electric field at anode becomes zero

• Estimated by 1D depletion equation

V FBqN

2 s

s2


Metal-Semiconductor-Metal Photodetectors• Main Drawback

– High dark current due to Schottky barrier junction

– Barrier enhancement layer can be added to reduce dark current of narrow energy gap semiconductor such as InGaAs

• Layer thickness ranges from 30 to 100nm

• Can be graded in composition to avoid carrier trapping


Metal-Semiconductor-Metal Photodetectors

• Primary advantages

– High speed and compatibility with FET technology

– Planar structure easy to integrate on single chip

– Very low capacitance per area• Advantageous for detectors requiring

large light sensitive areas

• Compared to p-i-n capacitance is about half


Quantum Well Infrared Photodetectors

• Structure of QWIP using GaAs/AlGaAs heterostructure– QW layers 5nm doped to

n-type in 1017 range

– Barrier layers are undoped and have a thickness 30-50nm

– Periods 20 to 50



• Incident light normal to surface has zero absorption– Intersubband transition

require electric field have components normal to QW plane

– Two methods• Polished facet

• Grating to refract light

Figure 716



• Intersubband excitation– Three types of transitions

• Bound to bound (escape well by tunneling)

• Bound to continuum (escape well because first state is above barrier: easier)

• Bound to miniband in superlattice



• I-V of QWIP is similar to photodetectors

• Quantum efficiency is different since light absorption and carrier generation occur only in quantum wells

– Nop number of optical passes, Nw number of quantum wells, Lw is the length, P is the polarization correction factor

– Ep is the escape probability and is a function of bias

I ph q ph Ga

1 R( ) 1 exp N op N w L w E p P

G a1

N w C p

Cp is the capture probabilityof electron traversing quantum well

tp transit time across single period of structurett transit time across entire QWIP active length L

C p

t p

t t

N w

Ga is optical gain



• Dark current is due to thermionic emission over the quantum well barriers and thermionic field emission (thermally assisted tunneling) near the barrier peaks– To limit dark current, the QWIP has to be operated at low

temperatures in the range of 4-77 K

• Can be applied in focal plane arrays for 2D imaging– High speed capability and fast response

– Coupling of light to the photodetector is difficult


Summary• Photodiodes have depleted region with a high electric field that separates

photogenerated electron-hole pairs

• Width of depletion layer determines tradeoff between speed and quantum efficiency

• P-n photodiodes have lower response speed and higher noise than a p-i-n photodiode

• Heterojunction photodiodes can move light absorption region away from the surface due to transparence of larger bandgap materials

• MS photodiodes operate in two modes, used because no minority carriers that increases speed

• Avalanche photodiodes have high gain but at the cost of noise, better for signals of low light intensity

• MSM photodectors have high speed and large area but have high dark noise

• QWIP uses quantum wells and various intersubband transitions for electrons, and are often operated at low temperatures

ecse-6290 semiconductor devices and models ii spring, 2010 s. sawyer 1-1 ecse-6290 semiconductor...

Documents

optical gain slide

semiconductor devices

optical gain optical

optical gain photons

optical cavity

optical loss

positive threshold gain

population inversion