1. Optical Detectors Abdul Rehman

2. Optical Detector Optical detector is an essential component of an optical receiver which converts received optical signal into an electrical signal. Improvement of detector characteristics and performance improves the link performance allowing fewer repeaters. Optical detector has the same required characteristics as that for the optical source Abdul RehmanOptical Communication

3. Requirements High sensitivity at operating wavelength High fidelity Large electrical response to received optical signal (high eff) Fast response/ short response time Low noise High stability Low cost Low bias voltages High reliability Small size

4. lCGeneral Concept If the energy h ν of the incident photon exceeds the band gap energy (h ν > E g ) an electron-hole pair is generated each time a photon is absorbed by the semiconductor Under the influence of an electric field set up by an applied voltage electrons and holes are swept across the semiconductor resulting in a flow of electric current

5.  Responsivity The photo current I p is directly proportional to the incident optical power P in I p  RP in Where R is resposivity of photodetector in units of A/W R  output photo current I p incident optical power P in Responsivity is the ratio of the electrical output to the optical input. Abdul RehmanOptical Communication

6. Three fundamental processes occurring between the two energy states of an atom: (a) absorption; (b) spontaneous emission; and (c) stimulated emission. Abdul RehmanOptical Communication

7. Quantum Efficieny The responsivity can be expressed in terms of Quantum efficiency electron generation rate  photon incidence rate η (Q E ) is a figure given for a photosensitive device (charge-coupled device (C C D ), for example) which is the percentage of photons hitting the photo reactive surface that w ill produce an electron-hole pair. It is an accurate measurement of the device's sensitivity. Energy of a photon is ' hν ' so the photon incident rate may be written in terms of incident optical power and the photon energy as Abdul Rehman Optical Communication P iin hνhν

8.  electron generation rate is given as I p q So Quantum efficiency can be written as η = I p q P in h  h  I p q P in  hνqhνq R [since R  I p P in ] Abdul RehmanOptical Communication

9. hv h λ hc −6−6   x − 34 8 m  R  η q ηq  c  η qλ ηλ = hc q η ⋅ λ ⋅ 10- 6.624*10 −34 Js  2.998*10 8 m s 1.602*10 − 19 C η ⋅ λ ⋅ 10 −6 ⋅ 1.602*10 −19 m  C C 1 6.624*10  2.998*10 Js  s s j s Abdul Rehman  η ⋅ λ 1.24 A W, where λ  is in  m Optical Communication

10. This happens for R ∝  λ until h ν < E g because more photons are present for the same optical power This linear dependence on λ is not continue forever since eventually the photon energy becomes too small to generate electrons. h v   E g, The quantum efficiency η then drops to zero If the facets of the semiconductor slab are assumed to have an antireflection coating, the power transmitted through the slab of width W is p tr  exp(  α W ) p in Where α is absorption coefficient Abdul RehmanOptical Communication

11. The absorbed power is thus given by p abs  p in  p tr   1  exp(  α W)  p in Since each absorbed photon creates an electron-hole pair, the quantum efficiency η is given by η= p abs p in  1  exp   αW   becomes 'zero' when α  0. On the other hand η  approaches 1 if  W  1 Abdul RehmanOptical Communication

12. Photodiode Responsivities

13. Quantum efficiency

14. Optical absorption curve

15. The wavelength c at which α  becomes zero is called cutoff wavelength The material can be used as a photodetector only for c Indirect – bandgap semiconductors: Si, Ge The absorption is not as sharp as for direct band-bandgap materials Direct – bandgap semiconductors : GaAs, InGaAs Abdul RehmanOptical Communication

16. Energy of incident photon must be greater than or equal to bandgap energy of the photodetector material. Therefore photon energy h ν ≥ E g  h c λ  E g  λ ≤ h c E g Thus the threshold for detection commonly known as the long wavelength cutoff point λ c is λ c  h c E g This gives the longest wavelength of light for photodetection Abdul RehmanOptical Communication

17. Direct and Indirect Absorption(senior 425) Si and Ge absorbs light by both direct and indirect optical transitions. Indirect absorption requires the assistance of a photon so that momentum as well as energy are conserved. For direct absorption no photon is involved so transition probability is more likely. Si is only weakly absorbing over the wavelength band of interest in optical fiber communications (800-900 nm). For Ge the threshold for direct absorption occurs above 1530 nm and Ge may be used in the fabrication of the detectors over the whole of the wavelength range of interest. Photodiode material should be chosen with bangap energy slightly less than the photon energy corresponding to the longest operating wavelength of the system.

18. This gives a sufficiently high absorption coefficient to ensure a good response and at the same time limits the thermally generated carriers in order to achieve a low dark current Ge diodes have relatively large dark currents - a disadvantage of Ge To overcome this problem certain alloys of Ge are made where the bandgap is adjusted as per the requirements like InGaAs, GaAlSb Abdul RehmanOptical Communication

19. Bandwidth(agrawallll 135) The bandwidth of photodetector is determined by the speed with which it responds to variation in the incident optical power. The rise time is defined as time over which the current builds up from 10 % to 90 % of its final value when the incident optical power is changed abruptly. The rise time is written as; T r  (ln 9)(  tr   RC ) Abdul Rehman transit time Optical Communication time constant

20. The transit time is added to time constant of equivalent RC circuit because it takes some time before the carriers are collected after their generation through absorption of photons. The maximum collection time just equal to the time an electron takes to traverse the absorption region. The transit time can be reduced by decreasing W but the quantum efficiency begins to decrease significantly for α W  3 Thus, there is a trade-off between the bandwidth and the responsivity (speed versus sensitivity) of a photodetector. Abdul Rehman Optical Communication

21. Thus there is trade-off between the bandwidth and the responsivity (speed versus sensitivity) of a photodetector. The bandwidth is given as Abdul Rehman  f  1 2  tr  RC ) Optical Communication

22. Rise and fall times Abdul RehmanOptical Communication

23. Dark current I d Dark current is the current generated in absence of optical signal It originates from: stray light thermally generated electron – hole pairs I d should be negligible for a good photodetector (I d <10nA) Abdul RehmanOptical Communication

24. – – – – – – – 4.2. Photodetector design Photodetectors can be broadly classified into two categories photoconductive photovoltaic Photoconductive detector homogeneous semiconductor slab as shown in fig before Little current flows when no light is incident Incident light increases conductivity through electron-hole generation and current flow is proportional to the optical power reverse biased p-n junction Photovoltaic detectors solar cells, produce voltage in the presence of light Abdul RehmanOptical Communication

25. p-n photodiodes A reverse bias p-n junction consists of a region, known as depletion region Electron-hole pairs are created through absorption when such p-n junction is illuminated with light on one side Because of the large built-in electric field, electrons and holes generated inside the depletion region accelerate opposite directions and drift to n and p sides respectively. The resulting flow of current is proportional to the incident optical power A reverse bias pn junction acts as a photodetector and is referred as pn photodiode Figure showing structure of a p-n photodiode Light is falling on one side of the photodiode (p-side) The depletion region has width W Abdul RehmanOptical Communication

26. As shown in the figure incident light is absorbed mostly inside the depletion region The responsivity of photodiode is quite high (R~1 A/W) because of high quantum efficiency The electron-hole pairs generated experiences a large electric field and drift rapidly towards the p or n side, depending on the electric charge The resulting current flow because of incident optical power I p  RP in

27. m s 5 m Bandwidth often limited by transit time.If W is the width of the depletion region and v d is the drift velocity, the transit time is given by  tr  WvdWvd W  10  m, V d = 0 5  tr  10  10 − 6 m 10 s  10 − 10 s  100 ps Good enough to 1 Gbit/s Abdul RehmanOptical Communication

28. Both W and V d can be optimized to minimize  tr The depletion-layer width depends on the acceptor and donor concentrations and can be controlled through them. The velocity V d depends on the applied voltage but attains a maximum value called saturation velocity ~ 10 5 m/s that depends on the material used. Abdul RehmanOptical Communication

29. Limiting factor Limiting factor for the bandwidth of p-n photodiode is the presence of a diffusive component in the photocurrent. The physical origin of diffusive component is related to the absorption of incident light outside the depletion region. Electrons generated in the p-region have to diffuse to the depletion-region boundary before they can drift to the n-side, similarly holes generated in the n-region must diffuse to the depletion-region boundary. Diffusion is an inherently slow process; carriers take a nanosecond or longer to diffuse over a distance of about 1 µm. Abdul RehmanOptical Communication

30. Figure shows how the presence of a diffusive component can distort the temporal response of a photodiode. The diffusion contribution can be reduced by decreasing the widths of the p- and n-regions and increasing the depletion- region width so that most of the incident optical power is absorbed inside it. This is the approach adopted for p–i–n photodiodes. Abdul RehmanOptical Communication

31. p-i-n energy-band diagram Abdul RehmanOptical Communication

32. p-i-n photodiodes A simple way to increase the depletion region width is to insert a layer of undoped or lightly doped semiconductor material between the p-n junction Since the middle layer consists of nearly intrinsic material so such a structure is referred as p-i-n photodiode Because of its intrinsic nature the middle i-layer a large electric field exists in the i-layer. The depletion region extends throughout the i-region and its width W can be controlled by changing the i-layer thickness

33. Most power is absorbed in i-region so drift dominates over diffusion The width W depends on a compromise between speed and sensitivity. The responsivity can be increased by increasing W so that quantum efficiency  approaches 100% but at the same time the response time also increases as it takes longer time for carriers to drift across the depletion region. Optimum W is a compromise between responsivity and response time.

34. (PAGE 139 AGRAWAL) For indirect Si and Ge W  20  50  m, for reasonable quantum efficiency The bandwidth of such photodiode is limited by relatively long transit time,  tr  200 ps By contrast in InGaAs,W  3 -  5  m that uses in direct bandgap semiconductor, the transit time is reduced So  10ps and the detector bandwidth f f 1 2 πτ tr  10GHz, τ tr  τ RC 20 GHz possible, even 30 GHz with reduced  Abdul RehmanOptical Communication

35. λ 5-6 Characteristics of common p-i-n diodes Parameter Wavelength Symbol Unit  Si 0.4-1.1 Ge 0,8-1,8 InGaAs 1,0-1,7 Responsivity Quantum efficiency Dark current RηIdRηId A/W % nA 0,4-0,6 75-90 1-10 0,5-0,7 50-55 50-500 0,6-0,9 60-70 1-20 Rise time τrτr ns0,5-10,1-0,5 0,05-0,5 Bandwidth Bias voltage ∆Vb∆Vb GHz V 0,3-0,6 50-100 0,5-3 6-10 1-5 56

36. Intrinsic and extrinsic semiconductors An intrinsic semiconductor is one which is pure enough that impurities do not appreciably affect its electrical behavior. In this case, all carriers are created by thermally or optically excited electrons from the full valence band into the empty conduction band. Thus equal numbers of electrons and holes are present in an intrinsic semiconductor. Electrons and holes flow in opposite directions in an electric field, though they contribute to current in the same direction since they are oppositely charged. Abdul RehmanOptical Communication

37. N-type doping The purpose of n-type doping is to produce an abundance of mobile or "carrier" electrons in the material. To help understand how n-type doping is accomplished, consider the case of silicon (Si). Si atoms have four valence electrons, each of which is covalently bonded with one of four adjacent Si atoms. If an atom with five valence electrons, such as those from group VA of the periodic table (eg. phosphorus (P), arsenic (As), or antimony (Sb)), is incorporated into the crystal lattice in place of a Si atom, then that atom will have four covalent bonds and one unbonded electron. This extra electron is only weakly bound to the atom and can easily be excited into the conduction band. At normal temperatures, virtually all such electrons are excited into the conduction band. Since excitation of these electrons does not result in the formation of a hole, the number of electrons in such a material far exceeds the number of holes. In this case the electrons are the majority carriers and the holes are the minority carriers. Because the five- electron atoms have an extra electron to "donate", they are called donor atoms. Abdul RehmanOptical Communication

38. P-type doping The purpose of p-type doping is to create an abundance of holes. In the case of silicon, a trivalent atom (such as boron) is substituted into the crystal lattice. The result is that one electron is missing from one of the four covalent bonds normal for the silicon lattice. Thus the dopant atom can accept an electron from a neighboring atoms' covalent bond to complete the fourth bond. Such dopants are called acceptors. The dopant atom accepts an electron, causing the loss of one bond from the neighboring atom and resulting in the formation of a "hole." Each hole is associated with a nearby negative-charged dopant ion, and the semiconductor remains electrically neutral as a whole. However, once each hole has wandered away into the lattice, one proton in the atom at the hole's location will be "exposed" and no longer cancelled by an electron. For this reason a hole behaves as a quantity of positive charge. When a sufficiently large number of acceptor atoms are added, the holes greatly outnumber the thermally-excited electrons. Thus, the holes are the majority carriers, while electrons are the minority carriers in P-type materials. Blue diamonds (Type IIb), which contain boron (B) impurities, are an example of a naturally occurring P-type semiconductor. Abdul RehmanOptical Communication

39. Double-hetereostructure The performance of p-i-n photodiode can be improved considerably by using double- heterostructure design In this the middle i-type layer is sandwiched between the p- type and n-type layers of a different semiconductor whose bandgap is chosen such that light is absorbed in the middle i-layer A p-i-n photodiode commonly used for lightwave applications uses InGaAs for the middle layer and InP for the surrounding p-type and n- type layers. Abdul RehmanOptical Communication

40.  − 19 Double – heterostructure design λ c  hc 6.626x10 − 34 x2.998x10 8 E g 1.35x1.602x10  0.92  m InP: E g =1.35 eV  C =0.92  (transparent for wavelengths greater than 0.92  InGaAs: E g =0.75 eV λ C =1.65  lattice matched

41. The middle InGaAs layer absorbs strongly in the wavelength region 1.3-1.6  The diffusive component of the detector current is eliminated completely because light is only absorbed inside the depletion region The quantum efficiency  can be made almost 100% by using an InGaAs layer 4-5  thick Bandwidth as high as 70 GHz were realized using thin absorption layer (W< 1  but only at the expense of a lower quantum efficiency  and responsivity By 1995 photodiodes exhibit bandwidth ∆  GHz and τ RC 

42. How to improve efficiency ?? Fabry-Perot cavity (laser like structure) can be formed around p-i-n structure to enhances quantum efficiency to  FP cavity has a set of longitudinal modes at which the internal optical field is resonantly enhanced through constructive interference The result of such structure is that when the incident wavelength is close to longitudinal mode such photodiode exhibit high sensitivity

43. Another approach for efficient high-speed photodiodes is use of optical waveguide into which the optical signal is edge-coupled It enhances quantum efficiency to  as absorption takes place along the length of the optical waveguide ( ∼ 10  ) 50 Ghz bandwidth was realized in 1992 for waveguide photodiode. The bandwidth could be increased to 110 GHz by adopting mushroom-mesa waveguide structure. In this structure the width of the i-type absorbing layer was reduced to 1.5 µm while the p- and n- type cladding layers were made 6 µm wide In this way both parasitic capacitance and the internal series resistance were minimized, reducing T RC to about 1 ps ∆ f ≈ 172 GHz η ≈ 45%

44. Speed of response The main factors limit the response of photodiode are Drift time of carriers through the depletion region Diffusiontimeofcarriersgeneratedoutsidethe depletion region Time constant incurred by the capacitance of photodiode with its load Abdul RehmanOptical Communication

45. hvhv Avalanche photodiodes All detectors require a certain minimum current to operate reliably The current requirement translate into minimum power requirements through P in  I p R Detectors with large responsivity R are preferred since they require less optical power The responsivity of p-i-n photodiode can takes its maximum value R  q for    1 Abdul RehmanOptical Communication h ν  = --- q R

46. APDs can have much larger values of R as they are designed to provide an internal current gain They are used when the amount of optical power that can be spared for the receiver is limited The physical phenomenon behind the internal gain is known as impact ionization Under certain conditions, an accelerating electron can acquire sufficient energy to generate a new electron-hole pair Abdul RehmanOptical Communication

47. The energetic electron gives a part of its kinetic energy to another electron in the valence band that ends up in the conduction band, leaving behind a hole The net result of impact ionization a single primary electron, generated through absorption of a photon, creates many secondary electrons and holes, all of which contribute to photodiode current Abdul RehmanOptical Communication

48. αe:αe: The primary hole can also generate secondary electron-hole pairs that contribute to the current The generation rate is governed by two parameters, impact-ionization coefficient of electrons,average number of electrons created per length αh:αh: impact-ionization coefficient of average number of holes created per length holes, The numerical value of  e and  h depends on the semiconductor material and on the electric field that accelerates electrons and holes

50. 5   In the last figure E  2  4  10 V cm  α e   α h   1  10 4 cm -1 These values can be realized by applying a high voltage ( ∼ 100 V) to the APD APDs differ in their design from that of p-i-n photodiodes is that an additional layer is added in which secondary electron-hole pairs are generated through impact ionization Abdul RehmanOptical Communication

51. Reverse bias An APD with electrical field distribution inside various layers under reverse bias Abdul RehmanOptical Communication

52. Ref to last figure: under reverse bias a high electric field exists in the p- type layer sandwiched between i-type and and n + type layers This p-type layer is referred to as multiplication layer, since secondary electron-hole pairs are generated here through impact ionization The i-layer still acts as the depletion region in which most of the incident photons are absorbed and primary electron-hole pairs are generated The APD drawbacks include Fabrication difficulties due to their more complex structure and hence increased cost The random nature of the gain mechanism which gives an additional noise contribution The high bias voltage required (50 to 400 V) which are wavelength dependent The variation of gain (multiplication factor) with temperature, thus the temperature compensation is necessary to stabilize the operation of the device The bandwidth of APD depends on M- multiplication factor (decrease with increasing M) Avalanche process takes time to build up

53. λ - - Characteristics of common APDs ParameterSymbolUnitSiGeInGaAs Wavelength Responsivity R APD  A/W 0,4-1,1 80-130 0,8-1,8 3-30 1,0-1,7 2-20 APD gainM100-50050-20010-40 k-factor k A = α h / αeαe 0,02-0,050,7-1,00,5-0,7 Dark current I d nA0,1-150-5001-5 Rise time Bandwidth τr∆τr∆ ns GHz 0,1-2 0,2-1,0 0,5-0,8 0,4-0,7 0,1-0,5 1-3 Bias voltage V b V200-50020-4020-30 Si: very strong for short wavelength

54. Performance of InGaAs APDs can be improved through design modifications The main reason of poor performance of APDs is comparable numerical values of the impact-ionization coefficient  e and  h. The result is that bandwidth is considerably reduced and the noise is relatively high At the same time InGaAs undergoes tunneling breakdown at electric fields of about 1x10 5 V/cm, the value that is below threshold for avalanche multiplication This problem can be solved in heterostructure APDs by using an InP layer for the gain region because quite high electric fields (> 5x10 5 V/cm) can exists in InP without tunneling breakdown The absorption region (i-type InGaAs layer) and the multiplication region (n-type InP layer) are separate in such device so this structure is known as SAM (separate absorption and multiplication regions)  h >  e for InP so the holes initiate the avalanche process Abdul RehmanOptical Communication

55. E g  0,75 eV   SAM APD (separate absorption and multiplication regions) – InP : InGaAs : InGaAs E g  1,35 eV   valenceband step 0,4 eV  InP Holes trapped at interface ε v

56. A problem with SAM APD is related to large bandgap difference between InP (E g = 1.35 eV) and InGaAs (E g = 0.75 eV). Because of this valence band step of 0.4 eV, holes generated in the InGaAs layer are trapped at the heterojunction interface and are slowed before they reach the InP multiplication region Such APD has extremely slow response time and relatively small bandwidth This problem can be solved using SAGM (separate absorption, grading, and multiplication regions)APDs. In SAGM (separate absorption, grading and multiplication) APD, we use another layer (InGaAsP) between absorption and multiplication regions whose bandgap is intermediate to those of InP and InGaAs layers This additional layer improves the bandwidth considerably Gain-bandwidth product (M ∆f) of 100 GHz was demonstrated in 1991 by using a charge region between the grading and multiplication regions

57. SAGM: Separate absorption, grading, multiplication Improved bandwidth M ⋅ ∆f≈70 GHz for M>12 SAGCM n-doped charge layer M  ∆  GHz Abdul RehmanOptical Communication

58. A high performance APDs uses superlattice structure The major limitation of InGaAs APD results from comparable values of  e and  h. In this structure k A  α h α e can be reduced In one structure absorption and multiplication region alternate and consists of thin layers (~ 10 nm) of semiconductor material with different bandgaps Abdul RehmanOptical Communication

59. Its use is less successful for InGaAs/InP material system and the so called staircase APDs were developed In this InGaAsP layer is graded to form a sawtooth kind of structure in the energy-band diagram that looks like staircase under reverse bias Another scheme uses alternate layer of InP and InGaAs for the grading region Superlattice structure can be used for multiplication region so that ∆  for M=10, 10 times more sensitive than p-i-n diode. In MSM (metal semiconductor metal) photodetectors a semiconductor absorbing layer is sandwiched between two metals forming a schottky barrier at each metal-semiconductor interface that prevents flow of electrons from metal to semiconductor This structure result in planner structure with inherently low parasitic capacitance that allows high speed operation (up to 300 GHz) 1.3  MSM photodetectors exhibit 92% quantum efficiency and low dark current The planner MSM structure is most suitable for monolithic integration

60. Abdul RehmanOptical Communication

61. MSM Photodiodes Metal Schottky Barrier Semiconduct or Metal Prevents flow of electrons from metal to semiconductor Integrated 1  m Low dark current Rise time of about 16 ps BW Up to 300 GHz

62. υdυd Bandwidth of p-n photodiode often limited by transit time as ∆ f (bandwidth)  1 2 π ( τ tr  τ RC ) If W is is the width of the depletion region and transit time is given by is the drift velocity then the τ tr  WυdWυd Abdul Rehman Typically W  10  m,  d  10 Optical Communication 5 msms

63. V(E) GaAs Abdul Rehman Si Optical Communication v d  10 5 m E s

64. by Band gap the band gap is the energy difference between the top of the valence band and the bottom of the conduction band in insulators and semiconductors. An intrinsic (pure) semiconductor's conductivity is strongly dependent on the band gap. The only available carriers for conduction are the electrons which have enough thermal energy to be excited across the band gap, which is defined as the energy level difference between the conduction band and the valence band. Band gap engineering is the process of controlling or altering the band gap of a materiallb controlling the composition of certain semiconductor alloys, such as GaAlAs, InGaAs, and InAlAs Abdul RehmanOptical Communication

65. Band gaps Common materials at room temperature Abdul Rehman Ge InN InGaN Si InP GaAs AlGaAs AlAs SiC 6H SiC 4H GaN Diamond 0.67 eV 0.7 eV 0.7 - 3.4eV 1.14 eV 1.34 eV 1.43 eV 1.42 - 2.16 eV 2.16 eV 3.03 eV 3.28 eV 3.37 eV 5.46 - 6.4 eV Optical Communication

66. Photodiode A photodiode is an electronic component and a type of photodetector. It is a p- n junction designed to be responsive to optical input. Photodiodes are provided with either a window or optical fibre connection, in order to let in the light to the sensitive part of the device. Photodiodes can be used in either zero bias or reverse bias. In zero bias, light falling on the diode causes a voltage to develop across the device, leading to a current in the forward bias direction. This is called the photovoltaic effect, and is the basis for solar cells - in fact a solar cell is just a large number of big, cheap photodiodes. Diodes usually have extremely high resistance when reverse biased. This resistance is reduced when light of an appropriate frequency shines on the junction. Hence, a reverse biased diode can be used as a detector by monitoring the current running through it. Circuits based on this effect are more sensitive to light than ones based on the photovoltaic effect. A phototransistor is in essence nothing more than a normal bipolar transistor that is encased in a transparent case so that light can reach the Base-Collector diode. The phototransistor works like a photodiode, but with a much higher sensitivity for light, because the electrons that tunnel through the Base- Collector diode are amplified by the transistor function. A phototransistor has a slower response time than a photodiode however Abdul Rehman Optical Communication