集成光学课程课件

《集成光学课程课件》由会员分享，可在线阅读，更多相关《集成光学课程课件（73页珍藏版）》请在金锄头文库上搜索。

1、Integrated Optical DetectorsnDetectors for use in integrated-optic applications must have high sensitivity, short response time, large quantum efficiency and low consumption. nIn this chapter, a number of different detector structures having these performance characteristics are discussed.集成光学课程n9.1

2、 Depletion Layer Photodiodesn9.2 Specialized Photodiode Structuren9.3 Techniques for Modifying Spectral Responsen9.4 Factors Limiting Performance of Integrated Detectors本章学习各种探测器的工作原理，重点掌握耗尽层光电探测器的工作原理集成光学课程9.1 Depletion Layer PhotodiodesThe most common type of semiconductor optical detector, used i

3、n both integrated optic and discrete device applications, is the depletion-layer photodiode. The depletion-layer photodiode is essentially a reverse-biased semiconductor diode in which reverse current is modulated by the electron-hole pairs produced in or near the depletion layer by the absorption o

4、f photons light. The diode is generally operated in the photodiode mode, with relatively large bias voltage.集成光学课程9.1.1 Conventional Discrete PhotodiodeThe simplest type of depletion layer photodiode is the p-n junction diode. The energy band diagram for such a device, with reverse bias voltage Va a

5、pplied is shown in Fig.9.1. The total current of the depletion layer photodiode consists of two components: a drift component originating from carriers generated in region (b) and a diffusion component originating in regions (a) and (c).集成光学课程Fig.9.1 Energy band diagram for a p-n junction diode unde

6、r application of a reverse bias voltage Va集成光学课程Holes and electrons generated in region (b) are separated by the reverse bias field, with holes being swept into the p-region (c) and electrons being swept into n-region (a). Holes generated in the n-region or electrons generated in the p-region have a

7、 certain probability of diffusing to the edge of the depletion region (b), at which point they are swept across by the field. Majority carriers, electrons in (a) or holes in (c) are held in their respective regions by the reverse bias voltage, and are not swept across the depletion layer.集成光学课程In or

8、der to minimize series resistance in a practical photodiode while still maintaining maximum depletion width, usually one region is much more heavily doped than the other. In that case, the depletion layer forms almost entirely on the more lightly doped side of the junction, as shown in Fig.9.2. Such

9、 a device is called a high-low abrupt junction. 集成光学课程Fig.9.2 Energy band diagram for a p+-n (high-low) junction diode under application of a reverse bias voltage Va集成光学课程In GaAs and its ternary alloys, electron mobility is generally much larger than hole mobility. Thus, the p-region is usually made

10、 thinner and much more heavily doped than the n-region, so that the device will be formed mostly in n-type material, and the p-region then serves essentially just as a contact layer. For a device with the high-low junction geometry indicated in Fig.9.2, it can be shown that the total current density

11、 Jtot is given by(9.1.1)集成光学课程where 0 is the total photon flux in photons/cm2s, W is the width of the depletion layer, q is the magnitude of the electronic charge, is the optical interband absorption coefficient, Lp is the diffusion length for holes, Dp is the diffusion constant for holes, and pn0 i

12、s the equilibrium hole density. The last term represents the reverse leakage current (or dark current), which results from thermally generated holes in the n-material. This explains why that term is not proportional to the photon flux 0. 集成光学课程The first term gives the photocurrent, which is proporti

13、onal to 0, and includes current form both the drift of carriers generated within the depletion layer and the diffusion and drift of holes generated within a diffusion length Lp of the depletion layer edge. 集成光学课程The quantum efficiency of the detector, or the number of carriers generated per incident

14、 photon, is given by (9.1.2)which can have any value from zero to one. It should be noted that (9.1.1) and (9.1.2) are based on the tacit assumption that scattering loss and free carrier absorption are negligible small.集成光学课程It can be seen from (9.1.2) that, in order to maximize q, it is desirable t

15、o make the products W and Lp as large as possible. When W and Lp are large enough so that q is approximately equal to one, the diode current is then essentially proportional to 0, because the dark current is usually negligibly small.集成光学课程nIf the interband absorption coefficient is too small compare

16、d to W and Lp, many of the incident photons will pass completely through the active layers of the diode into the substrate, as shown in Fig.9.3. Only those photons absorbed within the depletion layer, of thickness W, have maximum effectiveness in carrier generation. Photons absorbed at depths up to

17、a diffusion length Lp from the depletion layer edge are somewhat effective in generating photo-carriers, in that holes can diffuse into the depletion layer.集成光学课程Fig.9.3 Diagram of a conventional mesa-geometry photodiode with p+-n doping pro photon penetration集成光学课程Photons that penetrate to as depth

18、 greater that (W + Lp) before being absorbed are essentially lost to the photo-generation process because they have such a low statistical probability of producing a hole that can reach the depletion layer and be swept across. Within the semiconductor, the photon flux falls off exponentially with in

19、creasing depth x from the surface. Thus if is not large enough, many photons will penetrate too deeply before being absorbed, thus producing carriers that (on average) will recombine before diffusing far enough to reach the depletion layer. 集成光学课程nInterband absorption is a strong function of wavelen

20、gth in a semiconductor. Thus, it is impossible to design a diode with an ideal W for all wavelengths.nAside from the reduction of quantum efficiency that results from poor matching of , W and Lp, there are some other limitations to depletion layer photodiode performance that are also important.集成光学课

21、程Since W is usually relatively small (in the range of 0.1 to 1.0m), junction capacitance can limit high-frequency response through the familiar R-C time constant. Also, the time required for carriers to diffuse from depths between W and (W + Lp) can limit the high frequency response of a conventiona

22、l photodiode. The waveguide depletion layer photodiode, which is discussed in the next section, significantly mitigates many of these problems of the conventional photodiode.集成光学课程9.1.2 Waveguide PhotodiodesnIf the basic depletion layer photodiode is incorporated into a waveguide structure, as shown

23、 in Fig.9.4, a number of improvements in performance are realized. In this case, the light is incident transversely on the active volume of the detector, rather than being normal to the junction plane. The diode photocurrent density is then given by(9.1.3)集成光学课程Fig.9.4 Diagram of waveguide detector集

24、成光学课程nwhere L is the length of the detector in the direction of light propagation. Since W and L are two independent parameters, the carrier concentration within the detector volume and the basis voltage Va can be chosen so that the L can be made as long as necessary to make L1. Thus 100% quantum ef

25、ficiency can be obtained for any value of , by merely adjusting the length L. For example, for a material with the relatively small value of =30cm-1, a length of L=3mm would give q=0.99988. (Again, it has been tacitly assumed in (9.1.3) that scattering loss and free-carrier absorption are negligible

26、.)集成光学课程n Because a waveguide detector can be formed in a narrow channel waveguide, the capacitance can be very small, even if L is relatively large. This capacitance is about a factor of ten less than that of typical conventional mesa photodiode. Hence, the high frequency response can be expected t

27、o be correspondingly improved.n Because all of the incident photons are absorbed directly within the depletion layer of a waveguide photodetector, not only q is improved, but also the time delay associated with the diffusion of carriers is eliminated. This result is a further improvement in high fre

28、quency response.集成光学课程nDue to the many improvements in performance inherent in the transverse structure of the waveguide detector, as compared to the axial geometry of the conventional mesa photodiode, waveguide detectors should be considered for use in discrete-device applications as well as in opt

29、ical integrated circuits. 集成光学课程9.2 Specialized Photodiode StructurenThere are two very useful photodiode structures that can be fabricated in either a waveguiding or conventional , nonwaveguiding form. These are the Schottky-barrier photodiode and the avalanche photodiode (APD).集成光学课程9.2.1 Schottky

30、-Barrier PhotodiodenThe Schottky-barrier photodiode is simply a depletion layer photodiode in which the p-n junction is replaced by a metal-semiconductor rectifying (blocking) contact. For example, if the p-type layers in the devices of Fig.9.4 were replaced by a metal that forms a rectifying contac

31、t to the semiconductor, Schottky-barrier photodiodes would result. The photocurrent would still be given by (9.1.1) and (9.1.3), and the devices would have essentially the same performance characteristics as their p+-n junction counterparts.集成光学课程The energy band diagrams for a Schottky-barrier diode

32、, under zero bias and under reverse bias, are given in Fig.9.5. It can be seen that the depletion region extends into the n-type material just as in the case of a p+-n junction. The barrier height B depends on the particular metal-semiconductor combination that is used. Typical values for B are abou

33、t 1V.集成光学课程Fig.9.5 a,b. Energy band diagram for a Schottky-barrier diode; (a) zero bias; (b) reverse biased with voltage Va集成光学课程nIn conventional mesa devices, a thin, optically transparent Schottky-barrier contact is often used (rather than a p+-n junction) to enhance short-wave-length response, by

34、 eliminating the strong absorption of these higher energy photons that occurs in the p layer. In a waveguide photodiode, a Schottky-barrier contact is not needed for improved short-wavelength response because the photons enter the active volume transversely. 集成光学课程However, ease of fabrication often

35、makes the Schottky-barrier photodiode the best choice in integrated applications. For example, almost any metal (except for silver) produces a rectifying Schottky-barrier when evaporated onto GaAs at room temperature. Gold, aluminum or platinum are often used. Photodiode masking is adequate to defin

36、e the lateral dimensions during evaporation, and no careful control time and temperature is required, as in the case of diffusion of a shallow p+ layer.集成光学课程9.2.2 Avalanche PhotodiodenThe gain of a depletion layer photodiode (i.e. the quantum efficiency), of either the p-n junction or Schottky-barr

37、ier, can be at most equal to unity, under normal conditions of reverse bias. However, if the device is biased precisely at the point of avalanche breakdown, carrier multiplication due to impact ionization can result in substantial gain in terms of increase in the carrier to photon ratio. In fact, av

38、alanche gains as high as 104 are not uncommon. Typical current-voltage characteristics for an avalanche photodiode are shown in Fig.9.6.集成光学课程Fig.9.6 Response curves for an avalanche photodiode集成光学课程nThe upper curve is for darkened conditions, while the lower one shows the effects of illumination. F

39、or relatively low reverse bias voltage, the diode exhibits a saturated dark current Id0 and a saturated photo-current Iph0. However, when biased at the point of avalanche breakdown, carrier multiplication results in increased dark current Id, as well as increase photo-current Iph. It is possible to

40、define a photo-multiplication factor Mph, given by(9.2.1)集成光学课程and a multiplication factor M, given by(9.2.2)where Vb is the breakdown voltage, and n is an empirically determined exponent depending on the wavelength of light, doping concentration, and, of course, the semiconductor material from whic

41、h the diode is fabricated. For the case of large photo-current Iph0 Id0 the multiplication factor is given by(9.2.3)集成光学课程where I is the total current, given by(9.2.4)R being the series resistance of the diode (including space-charge resistance if significant). The derivation of (9.2.3) assumes that

42、 IRVb. For the case of Id0 and Id being negligibly small compared to Iph0 and Iph, it can be shown that the maximum attainable multiplication factor is given by(9.2.5)集成光学课程nAvalanche photodiodes are very useful detectors, not only because they are capable of high gain, but also because they can be

43、operated at frequencies in excess of 10 GHz. However, not every p-n junction or Schottky-barrier diode can be operated in the avalanche multiplication mode, biased near avalanche breakdown. For example, the field required to produce avalanche breakdown in GaAs is approximately 4x105 V/cm. Hence, for

44、 a typical depletion width of 3m, Vb equals 120V. 集成光学课程Most GaAs diodes will breakdown at much lower voltages due to other mechanisms, such as edge breakdown or microplasms generation at localized defects, thus never reaching the avalanche breakdown condition. In order to fabricate an avalanche pho

45、todiode, extreme care must be taken, beginning with a dislocation-free substrate wafer of semiconductor material.集成光学课程9.3 Techniques for Modifying Spectral ResponsenThe fundamental problem of wavelength incompatibility, which was encountered previously in regard to the design and fabrication of mon

46、olithic laser/waveguides structures, is also very significant with respect to waveguide detectors. An ideal waveguide should have minimal absorption at the wavelength being used. However a detector depends on interband absorption for carrier generation. 集成光学课程nHence, if a detector is monolithically

47、coupled to a waveguide, some means must be provided for increasing the absorption of the photons transmitted by the waveguide within the detector volume. A number of different techniques have been proven effective in this regard.集成光学课程9.3.1 Hybrid StructuresnOne of the most direct approaches to obta

48、ining wavelength compatibility is to use a hybrid structure, in which a detector diode, formed in a relatively narrow bandgap material, is coupled to a waveguide fabricated in wider bandgap material. The two materials are chosen so that photons of the desired wavelength are transmitted freely by the

49、 waveguide, but are strongly absorbed within the detector material. An example of this type of hybrid waveguide/detector is the glass on silicon structure, as shown in Fig.9.7. 集成光学课程Fig.9.7 Hybrid waveguide detector featuring a glass waveguide coupled to a silicon photodiode集成光学课程The diode was form

50、ed by boron diffusion to a depth of about 1m into an n-type, 5 cm silicon substrate. A 1m thick layer of thermally grown SiO2 was used as a diffusion mask. The glass waveguide was then sputter-deposited and silver paint electrodes were added as shown. Total guide loss was measured to be 0.8dB/cm10%

51、for light of 632.8nm wavelength. The efficiency of coupling between the waveguide and the detector was 80%. However, because the light enters the diode in the direction normal to the junction plane rather than parallel to it, this particular waveguide detector geometry does not have many of the adva

52、ntages described in Sect.9.1.2. 集成光学课程Nevertheless, good high frequency response can be expected. These diffused diodes had a capacitance of only 3x10-9F/cm2 when biased with Va equal to 10V. Thus a detector diode of approximately 10m radius, used in conjunction with a 50 load resistance, would have

53、 an RC time constant of about 15ps, implying that modulation of frequencies in excess of 10GHz could be detected.集成光学课程nWhile hybrid detectors offer the possibility of choosing the waveguide and detector materials for optimum absorption characteristics, better coupling efficiency can be obtained wit

54、h monolithic fabrication techniques. Monolithically fabricated waveguide detectors also have the advantages that light enters the device in the plane of the junction rather normal to it.集成光学课程9.3.2 Heteroeptaxial GrowthnPerhaps the most popular method of monolithically integrating a waveguide and de

55、tector is to use heteroepitaxial growth of a relatively narrow bandgap semiconductor at the location where a detector is desired. An example of this approach is given by the InGaAs detector that has been integrated with a GaAs wavelguide, as shown in Fig.9.8. In InxGa(1-x)As, the bandgap can be adju

56、sted to produce strong absorption of light at wavelength in the range from 0.9 to 1.15 m by changing the atomic fraction x of indium.集成光学课程Fig.9.8 Monolithically integrated InGaAs detector in a GaAs waveguide集成光学课程The monolithically waveguide detector structure shown in Fig.9.8 combines an epitaxial

57、ly grown carrier-concentration-reduction type waveguide with a platinum Schottky barrier detector. A 600nm thick layer of pyrolytically deposited SiO2 was used as a mask to etch a 125 m diameter well into the 5 to 20 m thick waveguide, and then grow the InxGa(1-x)As detector material. A quantum effi

58、ciency of 60% was measured for this detector at a wavelength of 1.06m, for low bias voltages. The loss in the waveguide was less than 1dB/cm. At bias voltage greater than about 40V, avalanche multiplication was observed, with multiplication factors as high as 50. 集成光学课程Optimum performance was obtain

59、ed at a wavelength of 1.06 m for an In concentration of x=0.2. The fact that quantum efficiency in this devices did not approach 100% more closely was most likely caused by less than optimum depletion width in the Schottky-barrier diode. The carrier concentration in the waveguide must be very carefu

60、lly controlled in order to make W equal to the waveguide thickness. Defect centers associated with stress at the GaAs-GaInAs interface may have also played a role in reducing q.集成光学课程nIn general, the III-V compound semiconductors and their associated ternary (and quaternary) alloys offer the device

61、designer a wide range of bandgaps, and corresponding absorption edge wavelengths. The relationships between bandgaps, absorption edge wavelengths and lattice constants are shown in Fig.9.9. Dotted portions of the curves correspond to ranges of composition for which the bandgap is indirect. 集成光学课程Fig

62、.9.9 Dependence of absorption edge wavelength and lattice constant on composition for selected III-V alloys集成光学课程Direct bandgap materials generally have interband absorption coefficients greater than 104 cm-1 for wavelengths shorter than the absorption edge, while may be several orders of magnitude

63、less in indirect gap materials. Nevertheless effective detectors can be made in indirect bandgap materials, especially when the waveguide detector geometry is used, so that the length L can be adjusted to compensate for small .The most popular material for the fabrication of detectors in the 1.0 to

64、1.6m wavelength range is GaInAsP.集成光学课程9.3.3 Proton BombardmentnIn Chap.4, proton bombardment was described as a method for producing optical waveguides in a semiconductor by generating carrier-trapping defect centers that resulted in reduced carrier concentration and increased index of refraction.

65、In that case, the waveguides were always sufficiently annealed after proton bombardment to remove the optical absorption associated with the trapping centers. 集成光学课程nHowever, one of the mechanisms responsible for this absorption is the excitation of carriers out of the traps, freeing them to contrib

66、ute to the flow of photon-current. Thus, a photodiode can be fabricated by forming a Schottky-barrier junction over the implanted region, as shown in Fig.9.10. ( A shallow p+-n junction could also be used.) 集成光学课程Fig.9.10 Diagram of a proton-implanted optical detector集成光学课程Photon-current flows when

67、the junction is reverse biased, because carriers liberated by photon-excitation within the depletion layer of the diode are swept across it by the field. Since a substantial number of the trapping centers have energy levels lying within the forbidden gap, the effective bandgap of the semiconductor i

68、s decreased, so that photons having less-than-bandgap energy can be absorbed and take part in the carrier generation process. Thus a proton-bombarded photodiode, made in a given semiconductor, can be respective to photons that would ordinarily not be absorbed in the material.集成光学课程For example, Stoll

69、 et al. have made a detector in GaAs that is sensitive to 1.15m wavelength radiation. The optical waveguide structure consisted of a 3.5m thick n-type epitaxial layer (S-doped, n1016cm-3) grown on a degenerately doped n-type substrate (n1.25x1018cm-3). Prior to proton implantation the optical attenu

70、ation at 1.15 m was measured to be 1.3cm-1, but after implantation with a dose of 2x1015cm-2 300keV protons in the region where a detector was desired, increased to over 300cm-1. A partial annealing of damage at 500C for 30 min was performed to reduce to 15cm-1 in order to allow some optical transmi

71、ssion through the entire length of the implanted region.集成光学课程Then, an Al Schottky-barrier contact was evaporated on top of the implanted region to complete the device. The relative photo-response of the proton implanted detector as a function of wavelength is shown in Fig.9.11, along with the corre

72、sponding curve for a similar detector formed in unimplanted GaAs. Only negligible response occurs in the unimplanted GaAs at wavelengths longer than 900nm, but a substantial absorption tail is observed for the proton bombarded detector, extending to wavelengths as long as 1.3 m. It must be remembere

73、d that, even though may be relatively small at the longer wavelengths, total absorption over the length of the detector can be quite larger. 集成光学课程Fig.9.11 Photoresponse of a proton implanted detector in GaAs集成光学课程9.3.4 Electro-AbsorptionnOne additional method for producing the required shift of the

74、 absorption edge to longer wavelength in a monolithic waveguide detector is electro-absorption, or the Franz-Keldysh effect. When a semiconductor diode is reverse biased, a strong electric field is established within the depletion region. This electric field causes the absorption edge to shift to lo

75、nger wavelength, as shown in Fig.9.12.集成光学课程Fig.9.12 Shift of the absorption edge of GaAs due to the Franz-Keldysh effect. (A) Zero-bias condition; (B) Reverse bias applied to produce a field of 1.35x105 V/cm集成光学课程Curve A shows the normal unbiased absorption edge for n-type GaAs with a carrier conce

76、ntration of 3x1016cm-3. Curve B is a calculated Franz-Keldysh-shifted absorption edge for an applied field of 1.35x105 V/cm, which corresponds to 50V reverse bias across a resulting depletion width of 3.7 m. At a wavelength of 900nm this shift corresponds to an increase in from 25 to 104 cm-1 hardly

77、 a negligible effect! The physical basis for the Franz-Keldysh effect can be understood from the simplified energy-band bending model diagrammed in Fig.9.13. 集成光学课程Fig.9.13 Energy band diagram illustrating the Franz-Keldysh effect. The band bending on the n-side of a p+-n junction ( or a Schottky ba

78、rrier junction) is shown for conditions of strong reverse bias集成光学课程In this diagram x represents distance from the metallurgical junction plane, in the region far from the junction where there is no electric field, photons must have at least the bandgap energy (Ec-Ev) to produce an electronic transi

79、tion as in (a). However, within the depletion region where there field is strong, a transition as in (b) can occur when a photon of less-than-bandgap energy lifts an electron partway into the conduction band, followed by tunneling of the electron through the barrier into the conduction band state. T

80、he states at the conduction band edge are, in fact, broadened into gap so as to produce a change in effective bandgap E, which is given by集成光学课程(9.3.1)where m* is the effective mass of the carrier, q is the magnitude of the electronic charge, and is the electric field strength. The Franz-Keldysh eff

81、ect greatly improves the sensitivity of a detector operating at a wavelength near its absorption edge. 集成光学课程nPerhaps the greatest advantage of electro-absorption detectors is that they can be electrically switched from a low absorption state to a high absorption state by merely increasing the rever

82、se bias voltage. This makes it possible to make emitters and detectors in the same semiconductor material that are wavelength compatible. An example of a device making use of this principle is the emitter/detector terminal shown in Fig.9.14. This device performs the dual function of light emitter, w

83、hen forward biased, and light detector, when strongly reverse biased. 集成光学课程Fig.9.14 An integrated-optic emitter/detector terminal employing the Franz-Keldysh effect集成光学课程Fabricated in series with a waveguide structure, as shown, it can act as a send/receiver tap on an optical transmission line. Bec

84、ause of the large change in produced by the Franz-Keldysh effect, operation can be very efficient. For example, consider the case of a p+-n junction diode in n-type GaAs with carrier concentration equal to 3x1016cm-3, as before. Application of 50V reverse bias changes from 25 to 104cm-1 at a wavelen

85、gth of 900nm. Thus, when forward biased, the diode emits 900nm light into the waveguide. When reverse biased with Va=50V, the diode need a length of only 10-3m to absorb 99.9% of incident 900nm light.集成光学课程When the diode is on standby at zero bias, is just 25cm-1. Hence, for a typical laser length o

86、f 200 m, the insertion loss is only 2dB. Such emitter/detector devices may prove to be very useful in systems employing waveguide transmission lines because they greatly simplify coupling problems, as compared to those encountered when using separate emitters and detectors.集成光学课程9.4 Factors Limiting

87、 Performance of Integrated DetectorsnIn the design of an integrated optical detector, there are a number of mechanisms that can limit performance in various ways. Not all of these are important in every application. However, the designer (or user) should be aware of the limitations associated with d

88、ifferent device types and geometries.集成光学课程9.4.1 High Frequency CutoffnA number of the factors that can limit high frequency response are summarized in Table 9.1. Because of the small area of waveguide photodetectors of the type shown in Fig.9.4, the RC time constant, which most often limits the res

89、ponse of conventional diodes, can be made small enough to allow frequencies of operation well above 10 GHz, as discussed in Sec.9.1.2. In this case, other potentially limiting effects must be considered.集成光学课程Table 9.1 Factors limiting high frequency response of a depletion layer photodiodenRC time

90、constant due to bulk series resistance and junction capacitancenCarrier diffusion time from regions outside of the depletion layernCarrier lifetime and diffusion lengthnCapacitance and inductance of the packagenCarrier drift time across the depletion layernCarrier trapping in deep levels集成光学课程n9.4.2

91、 Linearityn9.4.3 Noise集成光学课程Problemsn1We wish to use a photodiode as a detector for a signal of 900nm wavelength. Which would be the best choice of material for the photodiode, a semiconductor of bandgap =0.5 eV, bandgap=2 eV, or bandgap=1 eV? Why? (assume all three are direct gap and are equivalent

92、 in impurity content, etc.)集成光学课程n2To improve the signal-to-noise of the diode in Problem 9.1, we wish to use a semiconductor low-pass filter which has the following absorption properties at room temperature: for 900 nm radiation, =0.2cm-1; for 700 nm radiation, =103cm-1. How thick must the filter be to attenuate 700 nm background noise by a factor of 104? By what factor is the signal (at 900 nm) attenuated by a filter of this thickness? Neglect reflection at the surfaces.集成光学课程