Chapter 5 Microwave Transistors and Tunnel Diodes 5-0 INTRODUCTION Microwave solid-state devices are becoming increasingly important at microwave frequencies. These devices can be broken down into four groups. In the first group are the microwave bipolar junction transistor (BIT), the heterojunction bipolar tran- sistor (HBT), and the tunnel diodes. This group is discussed in this chapter. The second group includes microwave field-effect transistors (FETs) such as the junc- tion field-effect transistors (JFETS), metal-semiconductor field-effect transistors (MESFETs), high electron mobility transistors (HEMTs), metal-oxide-semiconduc- tor field-effect transistors (MOSFETS), the metal-oxide-semiconductor transistors and memory devices, and the charge-coupled devices (CCDs). This group is de- scribed in Chapter 6. The third group, which is characterized by the bulk effect of the semiconductor, is called the transferred electron device (TED). These devices include the Gunn diode, limited space-charge-accumulation diode (LSA diode), in- dium phosphide diode (InP diode), and cadmium telluride diode (CdTe diode). This group is analyzed in Chapter 7. The devices of the fourth group, which are operated by the avalanche effect of the semiconductor, are referred to as avalanche diodes: the impact ionization avalanche transit-time diodes (IMPATT diodes), the trapped plasma avalanche triggered transittime diodes (TRAPATT diodes), and the barrier injected transit-time diodes (BARITT diodes). The avalanche diodes are studied in Chapter 8. All those microwave solid-state devices are tabulated in Table 5-0-1 In studying microwave solid-state devices, the electrical behavior of solids is the first item to be investigated. In this section it will be seen that the transport of charge through a semiconductor depends not only on the properties of the electron but also on the arrangement of atoms in the solids. Semiconductors are a group of substances having electrical conductivities that are intermediate between metals and insulators. Since the conductivity of the semiconductors can be varied over wide 166‘Sec. 5.0 Introduction 167 TABLE 5-0-1 MICROWAVE SOLID-STATE DEVICES Mon cn JFET Field-effect, MESFET, HEMT transistors MOSFET NMOS, PMOS, CMOS Memories cco Microwave solid-state Gunn diode Gevices | Transferred electron LSA diodes devices InP diodes Cilfe dies Avalanche tans —[IMMPATT diode time devices TRAPATT diode E Read diode BARITT diode ranges by changes in their temperature, optical excitation, and impurity content, they are the natural choices for electronic devices. The properties of important semiconductors are tabulated in Table 5-0-2. ‘The energy bands of a semiconductor play a major role in their electrical be~ havior. For any semiconductor, there is a forbidden energy region in which no al- Jowable states can exist. The energy band above the forbidden region is called the ‘conduction band, and the bottom of the conduction band is designated by E-. The ‘energy band below the forbidden region is called the valence band, and the top of TABLE 5-0-2 PROPERTIES OF IMPORTANT SEMICONDUCTORS Bandgap Mobility at 300°K energy (eV) (em'/V= 8) Relative ee dielectric Semiconductor oK 300K Holes Electrons constant ¢ 551 3.47 1600 1800 5s Ge 039 o.803 1900) 3900 16 si 116 Lia 450 1600 8 Also 17s 163 20 200 nL Gasb 0.30 0.67 1400 5000 1s Gass 12 143 400 8500 Ba GaP 240 226 15 110 10 Insb 0.26 180 750 78,000 7 InAs 0.46 033 460 33,000 Ws InP. 134 129 180 4600 my cas 256 242 0 300 10 Case 13s 170 800 0 200 3.20 200 9 208 370 30 16s 8168 Microwave Transistors and Tunnel Diodes Chap. 5 the valence band is designated by E, . The separation between the energy of the low- est conduction band and that of the highest valence band is called the Bandgap en- ergy Eq, Which is the most important parameter in semiconductors. Electron energy is conventionally defined as positive when measured upward whereas the hole energy is positive when measured downward. A simplified band di- agram is shown in Fig. 5-0-1 QO ‘Conduction band KR = Distance Figure $01. Energy-band diagram. In the 1970s, it seemed that microwave transistors would be useful for generat- ing power up to about $ GHz. Since their inception, avalanche diodes have produced in excess of 4 W continuous wave (CW) at 5 GHz. Gunn diodes had been considered ‘only for local oscillators or low-power transmitter applications, but recent results in dicate that a single Gunn diode can generate an output power of 1 W at X band. At higher microwave frequencies, and even well into the millimeter range, limited space-charge-accumulation diodes (LSAS) can provide the highest peak power of any solid-state device, up to 250W in C band, 100 W in X band, and 50 W in Ku band, Since the pulsed Gunn and TRAPATT diodes are essentially transittime devices, their operating frequency is approximately determined by the thickness of the active layer in the diode. An operating frequency of 10 GHz requires an active layer thickness on the order of 10 4m (microns). Thus only a limited voltage can be applied to such a thin layer because of breakdown limitations. Consequently, the 1000, Peak power (watts) ¢ x . © S10 29 igure $0.2. Peak-power levels Frequency (GH2) achieved by microwave diode,Sec. 5.1 Microwave Bipolar Transistors 169 peak power capability of both the pulsed Gunn diodes and the TRAPATT diodes is ‘reatly limited at higher frequencies. On the other hand, the peak power capability of an LSA diode is approximately proportional to the square of the thickness of the active layer because its operating frequency is independent of the thickness of the active layer. Thus the LSA diode is capable of producing higher peak power than ei- ther the pulsed Gunn diodes or the TRAPATT diodes. Figure 5-0-2 shows peak power versus frequency for these three devices. state microwave power sources are widely used in radar, communica- tions, navigational and industrial electronics, and medical and biological equipment. Representative applications for microwave solid-state devices are listed in Table 5-0-3, ‘TABLE 5-0-3 APPLICATIONS OF MICROWAVE SOLID-STATE DEVICES. Devices Applications Advantages ‘Transistor L-band transmis for telemetry Low cost, low power supply, relia systems and phased array radar high CW power output light weight systems Land S-band transmitters for communications systems TED , X-, and Ku-band ECM amplifiers Low power supply (12 V), low cost, for wideband systerns light weight, reliable, low noise, high X+ and Ku-band transmitters for radar ain systems, such as trafic control IMPATT Transmitters for millimeter-wave Low power supply, low cost, reliable, Communications systems high CW power output, ight weight TRAPATT S-band pulsed transmitters for phased High peak and average power, reliable, aay radar systems low power supply, low cost BARITT Low cost, low power supply, reliable, Tow noise 5-1 MICROWAVE BIPOLAR TRANSISTORS ‘The invention of the transistor (contraction for transfer resistor) by William Shock- Jey,and his coworkers at Bell Laboratory in 1948 had a revolutionary impact on elec tronic technology in general and on solid-state devices in particular. Since then tran- sistors and related semiconductor devices have replaced vacuum tubes for lower- power sources. Microwave power transistor technology has advanced significantly during the past three decades. The microwave transistor is a nonlinear device, and its principle of operation is similar to that of the low-frequency device, but requirements for dimensions, process control, heat sinking, and packaging are much more severe. For microwave applications, the silicon (Si) bipolar transistors dominate for frequency range from UHF to about $ band (about 3 GHz). As the technology im- proves, the upper frequency limit for these devices is continuously being extended, and at the present time the devices are capable of producing useful power up to 22 GHz, The majority of bipolar transistors of current interest are fabricated from silicon, although GaAs devices offer prospects for improvements in operating fre-170 Microwave Transistors and Tunnel Diodes Chap. 5 quency, in high temperatures, and in radiation hardness. The Si bipolar transistor is inexpensive, durable, integrative, and offers gain much higher than available with competing field-effect devices. It has moderate noise figure in RF amplifiers and lif noise characteristics that are about 10-20 dB superior to GaAs MESFETS. For these reasons, the Si bipolar transistors dominate in amplifier applications for the lower microwave frequencies and are often the devices of choice for local oscillators. 5-1-1 Physical Structures All microwave transistors are now planar in form and almost all are of the silicon np-n type. The geometry can be characterized as follows: (a) interdigitated, (b) overlay, and (c) matrix (also called mesh or emitter grid) as shown in Fig. 5-1-1 ‘The interdigitated type is for a small signal and power, but the overlay type and matrix type are for small power only. The figure of merit for the three surface geometries shown in Fig. 5-1-1 is listed in Table 5-1-1 [1] Emitter (a) Interdigitated (b) Overlay }<— Emitter metalization }-——— P* base ditfusion Base metalization (© Matrix Figure 5-1-1 Surface geometries of microwave power transistor. (From H. Sobo! ‘and F. Sterzer [1 reprinted by permission of IEEE. Inc.) For high-frequency applications, the n-p-n structure is preferred because the electron mobility (Wn = 1500 cm’/V -s) is much higher than the hole mobility (a, = 450 cm’/V » 3), Figure 5-1-2 shows an example of the densities for an n-p-n transistor. The density unit is in cm’/V s, Although there are many ways of fabricating a transistor, diffusion and ion im- plantation are generally used. For example, the structure would typically start with a lightly doped n-type epitaxial layer as the collector. The base region would be formed by counter-doping the base region p-type by diffusion. The emitter wouldSec. 5.1 Microwave Bipolar Transistors 171 ‘TABLE 5-1-1 FIGURE OF MERIT (M) OF VARIOUS SURFACE GEOMETRIES Suter omer nd ce w-Z Loveday iaazp=st fo Ho 2¢ +» + eae Ww + E+ p) enter 1 Hf obs “pete 1 iene Wr we i ! wr) ius (E+ w) wo e+e) Emitter | P pase pt ‘Source: From H. Sobol and F. Sterzer (1: reprinted by permission of IEEE, Ine be formed by a shallow heavily doped n-type diffusion or by ion implantation. The emitter and base contacts are generally located on the semiconductor surface in an interdigital, planar arrangement, The interdigital geometry always provides for n+ 1 base fingers, where n is the number of emitter fingers. The number of fingers varies with the application, with more fingers required as the output power capabil- teniter Be Collstor ’ y netrxaor | ngetrx to | ag 17x 0! presto | peareaott | ples 7x 108 Figure $-1-2 Carrier densities of an mp-m transistor.172 Microwave Transistors and Tunnel Diodes Chap. 5 ity of the transistor increases. Additional fingers, however, increase the device para- sitics and degrade the noise and upper frequency capability of the devices. Figure 5-1-3 shows two schematic diagrams for a bipolar junction transistor (BIT): (a) the cross section of a discrete n-p-n planar BJT and (b) the cross section of a chip-type n-p-n integrated BIT. ‘The p-n-p bipolar junction transistor is a complementary structure of the n-p-n BIT by interchanging p for n and n for p. The p-n-p BIT is basically fabricated by first forming an n-type layer in the p-type substrate; then a p*-type region is devel- Insulator ayer So Enter contact base layer Collector contact’ ‘memite layer, ‘ecolector layer (2) Diserote mp plnar BIT Enter contact ‘Base contact, (6) Iterated chiptype mn BIT Figure 1-3. Schematic diagrams of ipl junction tanssers. (Aer. NoonSec. 5.1 Microwave Bipolar Transistors 173 ‘oped in the n layer. Finally, metallic contacts are introduced to the p* region and layer through the windows opened in the oxide layer and to the p region at the bottom. 5-1-2 Bipolar Transistor Configurations In general, there are two types of bipolar transistors: p-n-p and n-p-n. In practical applications, a transistor can be connected as three different configurations: common base (CB), common emitter (CE), and common collector (CC), depending on the polarities of the bias voltages connected to its terminals. Common-base configuration. The common-base (CB) configuration refers to the one where the emitter (input circuit) and collector (output circuit) terminals are common to the base as shown in Fig. 5-1-4. ‘ [ c ‘ j | f Yea vos Yow 4 Yew £ e & 2 { i 3 Yew Vos Yew ot o (2) €8 connguraton (09 CB contgration Figure §-1-4 | Common-base configurations for p-n-p and n-p- transistors The CB configuration is also called the grounded-base configuration. For a p- rp transistor, the largest current components are caused by holes. Holes flow from the emitter to the collector and down toward ground out of the base terminal. In an n-p-n transistor all current and voltage polarities are negative to those in a p-n-p transistor. The CB configuration of a transistor is usually used in amplifier applica- tions. Its input voltage Ves and output current Zc can be expressed in terms of the out- put voltage Vee and input current fe as fos = some function (Veu fr) (S-1-1) Ic = some function (Veo , I) (6-1-2) Common-emitter configuration. Most transistors have their emitter, rather than their base, as the terminal to both input and output networks. Such @174 Microwave Transistors and Tunnel Diodes Chap. 5 configuration is known as common-emitter (CE) ot grounded-emitter configuration as shown in Fig. 5-1-5. In the CE configuration, the input current Jy and the output voltage Vee are dependent variables, whereas the input voltage Var and output current Jc can be writ- ten as Vee = some function (Ve , Is) (5-1-3) Te = some function (Vee , In) (5-1-4) ‘The CE configuration is commonly used as a switch or pulse transistor amplifier This is because the transistor is open at the cutoff mode and is closed at the satura- tion mode. ¢ ¢ * m a P I Ie e Ye Yee Ye Vee ¢ io I Ve Vee ‘ne Yor Figure 5-1-5 Common-emitter coniguntions for p-mp and mp-n transistors, Common-collector configuration. Another transistor-circuit configuration is called the common-collector ot grounded-collector configuration as shown in Fig. 5-1-6. In a common-collector (CC) configuration, the output voltage of the load is taken from the emitter terminal instead of the collector as in the common-base and ‘common-emitter configurations. When the transistor is cut off, no current will flow in the emitter terminal at the load. When the transistor is operating in a saturation mode, the load current reaches toward its maximum, Therefore the CC configura- tion transistor can also be used as a switch or pulse amplifier. The significant differ-Sec. 5.1 Microwave Bipolar Transistors 15 s ‘ a * 5 ? = ? Var Yer Vm Vex 2 8 R R ‘ue "Vee Ta Ue Figure §-1-6 Common-collector configuration for p-n-p and n-p-n transistors. fence, however, between common-emitter and common-collector configurations is that the common-collector amplifier has no voltage gain. Hybrid-pl equivalent model. The hybrid-pi equivalent model is commonly used in the normal active mode of the common-emitter configuration for small- signal operation. Fig. 5-1-7 shows the hybrid-pi equivalent model of the common- emitter configuration. For a small-signal operation, the nonlinear or ac parameters of a hybrid-pi equivalent model can be expressed as (5-1-5) (5-1-6) (-1-7) (5-1-8) (5-1-9) (5-1-10)176 Microwave Transistors and Tunnel Diodes Chap. § (8) Hybrid mode Figure $-1-7Hybrid-pi equivalent (e) Simplied hybrid model at low frequency model ‘When the dimensions of a bipolar junction transistor become very small, their Z, ¥, or H parameters cannot be measured because the input and output terminals cannot bbe openly and shortly realized. Therefore, the S parameters are commonly mea- sured. In transistor design, it is necessary to convert the S parameters into ¥ parameters for the network component computations. ‘An incremental change of the emitter voltage AY’. at the input terminal will induce an incremental change of the collector current Ai at the output terminal. Then the mutual conductance (or transconductance) of a small-signal transistor is defined by die & esl, G-1-1) From the diode junction theory the thermal equilibrium density at the junction is ‘equal to the minority density times the forward-bias voltage factor. That is, (0) = moe" (S-1-12)Sec. 5.1 Microwave Bipolar Transistors wt and lil (5-1-13) Substitution of Eq. (5-1-12) into Eq. (5-1-13) and differentiation of the resultant yield Lid Pie te (5-1-14) where Vr = 26 mV at 300° K is the voltage equivalent of temperature AS the width of the base region is very narrow, Cy > C,, the diffusion capaci- tance in the base charge-storage is given by a Coe = Me (5-1-15) and the total charges stored in the base is o, = wea (5-1-16) ‘Then the diffusion capacitance is expressed by gAWonp(0) Wi Co, = Apel) = 5 (5-1-17) ‘The voltage across the diffusion capacitance can be written as 2 a te Vie = 56 (S-1-18) ‘The small-signal input conductance of the emitter junction looking at the input of the base is defined as mks m= len fa BOR Vr Ieee hee (eel) ‘where he is the linear or de common-emitter current gain factor and ‘Gm is the mutual conductance. Example 5-1-1: Equivalent Elements of a Hybrid-Pi Common-Emitter Circuit AA Si n-p-n bipolar transistor has the following parameters: Collector current: Common-emitter current gain factor: Operational temperature: Cross-sectional area:178 Microwave Transistors and Tunnel Diodes Chap. 5 ‘Compute: (a) the mutual conductance gq; (b) the input conductance ge and resistance Ry; (©) the electron diffusion coefficient D, ; and (d) the diffusion capacitance Ci. Solution 1a, The mutual conductance is tho i Ths ap conten on fate fe 003 ; 92am xi mo R= 521 ohms eT een tn oat is The diffusion capacitance is 10 Ix 46 Ch = ent = 0.23 x = 2.76 8 5-1-3 Principles of Operation ‘The bipolar junction transistor (BIT) is an active three-terminal device which is commonly used as an amplifier or switch. Its principles of operation are discussed in this section. Modes of operation. A bipolar transistor can operate in four different modes depending on the voltage polarities across the two junctions: normal (active) ‘mode, saturation mode, cutoff mode, and inverse (or inverted) mode as shown in Fig. 5-1-8. 1, Normal Mode. If the emitter junction of an n-p-n transistor is forward-biased and the collector is reverse-biased, the transistor is operated in the normal mode as shown in Fig. 5-1-8(a). The term forward bias means that the posi- tive polarity of the bias voltage is connected to the p side and the negative po- larity to the n side for a p-n junction; the opposite obtains for reverse bias, Most transistor amplifiers are operated in normal mode, and its common-base current gain alpha is known as the normal alpha aw 2. Saturation Mode. When both transistor junctions are forward-biased, the transistor is in its saturation mode with very low resistance, and acts like a short circuit, as shown in Fig. 5-1-8(b). 3. Cutoff Mode. If both transistor junctions are reverse-biased the transistor is operated in its cutoff mode. As the current is cut off, the transistor acts like anSec. 5.1 Microwave Bipolar Transistors 179 (a) Noxmal moe igure 5-1-8 Operational modes of an n-p-n transistor. ‘open circuit. Both the cutoff and saturation modes of a transistor are used as switching devices for the OFF and ON states. Fig. 5-1-8(c) shows the cutoff ‘mode bias-voltage connection. 4, Inverse Mode. When the emitter is reverse-biased and the collector is forward-biased, the transistor is operated in the inverse (or inverted) mode, and its current gain is designated as the inverse alpha ay. If the transistor is symmetric, the normal alpha av is nearly equal to the inverse alpha a. The ‘two current gains, however, are not actually equal because of their unequal dopings. The inverse mode is shown in Fig. 5-1-8(d). In practice, the inverse ‘mode is not commonly used except as a multiemitter transistor in TTL (transistor-transistor logic) logic gate. Current flow in normal mode. When a transistor is properly biased, the holes and electrons in the transistor will follow the field direction in motion. Figure 5-1-9 shows the current flow of an mp-n transistor. ‘The current flow in an ideal n-p-n bipolar junction transistor is analyzed under the following assumptions: 1. The resistivities of the semiconductor regions are low 2. The injected current densities are low 3. The space-charge layer widening effects can be ignored 4. The current and voltage of n-p junction diodes follow the basic equation of 1= L(e" 1) (S-1-20)180 ‘Microwave Transistors and Tunnel Diodes Chap. 5 Hctron nox Hote Seema Je Figure $-1-9 Current ow in an n-p-n transistor. For a common-base n-p-n transistor, the emitter junction is forward-biased and the collector junction is reverse-biased as shown in Fig. 5-1-9. Consequently, the emit- ter current /e consists of electron current /ae crossing from the emitter into the base and the hole current Jne ctossing from the base into the emitter. Since the doping of the emitter is much larger than the doping of the base, the hole current is negligible. However, not all the electrons crossing the emitter junction Je reach the collector junction Jc because some of them combine with the holes in the p-type base. If Jac is the electron current at the collector junction Jc, there must be @ recombination ‘current Iu: ~ [ac leaving the base. When the emitter is open-circuited, then fe = 0 and Inc = 0. AS a result, the collector current /c is equal to the reverse saturation current [ce because the junction between base and collector is reverse-biased. From Fig. 5-1-9, we have In = Tye = (Int ~ Ine) ~ Heo + Ing (5-1-21) Te = be + he - In (5-1-22) and Te = leo ~ Ine (5-1-23) For an n-p-n transistor, Ico consists of holes moving across the collector junction Je from right to left (collector to base) and electrons crossing Jc in the opposite direc- tion. Since the reference direction for Jco in Fig. 5-1-9 is assumed from left to right, then, for an n-p-n transistor, Ico is positive for forward-biased Jc junction and co is, negative for reverse-biased Jc junction. The saturation current /co at the Jc junction of an n-p-n transistor is given by _AqDy , AgD;pas (2 Db Ww ¥ Le Agi WNa A Les Ico ) (5-1-24)Sec. 5.1 Microwave Bipolar Transistors 181 ‘Also from Fig. 5-1-9, the sum of the three terminal currents should be zero and it is le + lot lo (5-1-25) Equation (5-1-25) can be verified by adding together Eqs, (5-1-21) through (5-1-23) Current flow in common-base n-p-n transistor. In a common-base configuration of an n-p-n transistor as shown in Fig. 5-1-9, the emitter n-p junction is forward-biased and the collector p-n junction is reverse biased. Their total current flow can be found from the basic diffusion equation. The steady-state diffusion equa- tion for an n-p-n transistor at low-level injection is given by dn 1, = AqD, & (5-125; 4D. Te « sa) and Pry _ mi= te 2g (5-1-26) dx? T where D, = clectron diffusion constant minority electron carrier density in the p-type base layer equilibrium minority electron carrier density in the p-type base layer x = distance measured from the base region as shown in Fig. 5-1-10 74 = electron lifetime Figure 51-10 Minority-carrier den = ties under normal ative bias with ne ae ee csletor ible recombination The general solution of Eq. (5-1-26) can be written as Cres + Crea (5-1-27) ‘where C, and C: are constants to be determined by the boundary conditions L, = Vimy is the electron diffusion length Myo = niNa is the mass-action law The boundary conditions at the edge of the emitter depletion layer in the base side with a forward-biased emitter junction is 1,(0) = nywee/"T (5-1-28)182 Microwave Transistors and Tunnel Diodes Chap. 5 where Ve = forward-biased voltage across the emitter junction Vr = 26 X 10” V at 300° k is the voltage equivalent of temperature ‘The boundary condition at the depletion-layer edge of the reverse-biased collector junction is usually assumed to be ndW) = 0 (5-129) ‘The two boundary conditions are shown in Fig. 5-1-10. ‘The general solution of Eq. (5-1-26) can be written as (Ww - »/L) __ sinh (x/Ly) sinh (W/L.) +m( an ule)| oe) In almost all transistors, the base width is made very narrow (W < L,) so that the minority-carrier recombination in the base is negligible, As a result, the boundary conditions specify the two end points of the base carrier concentration with a straight line as shown in Fig. S-I-11, (Note: sinh y * y, cosh y = coth y = Wy, and sech y = 1 — y%/2 fory <1.) Myla) = nolee"F = of Lo os 06 oa Figure $-1-11 Minority carrier disti= ° W butions in base region a8 function of Distance + Wie. (fier S. Sze (3].) For W W is proportional to the gradient of the minority carrier density and is expressed as = AgD ste Te rari Tye lisa, coth (uta Bio DP wal (5-1-39) ‘The electron current which reaches the collector at x = W is _ AgDatipn 1 vee = = Me 1 er 9 + cant] - veltr — ) —AQDant © _ AqDant yey, (e' dD NW NW e (S-1-40) The collector current Ic can also be expressed as184 Microwave Transistors and Tunnel Diodes Chap. 5 be ~ la = la 1-12) Ta) (5-1-41) where Le = diffusion length of the emitter and exp (Vs/Vs) > 1 is assumed Similarly, the hole currents are lem Aa epee err 1) fea May (5-1-42) and Ire = MPP fore = xe (5-1-43) where De and Dc are the hole diffusion constants in the emitter and collector, respectively Le = diffusion length of the collector The current flow in an n-p-n transistor as described so far is an ideal model, and its recombination-generation current is not counted. If the recombination cur- rent is considered, the current flow is the sum of the drift, the diffusion, and the recombination-generation currents. That is, 1 = Ly (drift) + Ly Giffusion) + [ (recombination-generation) _(5-1-44) ‘The recombination-generation current can be computed ftom the following equation ty = BA ere (5-1-45) depletion-layer width effective minority-carrier lifetime in the depletion layer Figure 5-1-12 shows current-voltage (IV) characteristics of an ideal n-p-n bipolar transistor for a common-base configuration. There are three regions for the I-V characteristics of an n-p-n bipolar transis- tor: 1, Active Region: In this region the emitter junction is forward-biased and the collector junction is reverse-biased. The collector current Jc is essentially in- dependent of collector voltage and depends only on the emitter current Je. When the emitter current is zero, the collector current is equal to the reverse saturation current Lc. 2, Saturation Region: In this region, as shown on the left side of Fig. 5-1-12, both emitter and collector junctions are forward-biased. The electron current flows from the n side across the collector junction to the p-type base. As a re- sult, the collector current increases sharply.Sec. 5.1. Microwave Bipolar Transistors 185 stor CUTER fe. WA © Cotlctortorboewotage ey in vl Figure $-1-12 Current Voltage (LV) characteristics of an n-p-mtransistr. 3. Cutoff Region: In this region the emitter and collector junctions are both reverse-biased. Consequently, the emitter current is cut off to zero, as shown in the lower right side of Fig. 5-1-12. Example 5-1-2: I-V Characteristies of an n-p-n Transistor A silicon n-p-n transistor at 300° K has the following parameters: Base width: w= 10% cm Diffusion length in emitter: Le = 10™* em Diffusion length in collector: Le = 5 x 10-*em Base resistivity: pe = 0.15 Dem Emitter resistivity Pe = 0.006 -cm Collector resistivity be = 16 Q-cm Emitter junction voltage: Ve=05V Collector junction voltage: Ve=06V 2x 10°? em? Cross-section area: ri Find: a. The impurity densities in the emitter, base, and collector regions bb. The mobilities in the emitter, base, and collector regions lengths in the emitter, base, and collector regions106 Microwave Transistors and Tunnel Diodes Chap. § 4. The equilibrium densities inthe emitter, base, and collector regions . The terminal currents Solution ‘a, The impurity densities are read from Fig. A-l in Appendix A as Ne =1x 10 cm? in the n-type emitter region 15 x 10° cm in the p-type base region Nac = 3x 10 cm"? in the n-type collector region . The mobilities are read from Fig. A-2 in Appendix A as Moe = 80 cmiVes in the emitter Hae = 108 cm/V-s in the emitter Hoe = 400 cmi/V-s in the base Hac = 1600 cm'/V-s in the collector ‘¢. The diffusion constants are computed to be Dae = UreVr = 80 X 26 10 = 2.08 cms 73 emis 400 x 26 x 10 = 10.4 emis = 416 cms d. The equilibrium densities are ‘try =n? [Nag = (1.5 % 10"P/(1.5 x 10") = 1,5 x 10? em”? ni Na = (1.5 * 10)/(1 10%) = 2.5% 10 cm”? ni Nac = (1-5 * 10%/(3 x 10") = 7.5 x 10° cm"? ‘e. The terminal currents are computed as follows: From Eq. (S-1-39), the electron current in the emitter is Aut uy 5 ADO ers =p — 4) Na tN, 2X 10 X 1.6 x 10" X 2.73 x (1S x 10°F 15x 107% 10 % exp[0.5/(26 x 10°*)] 13.104 x 10" x 2.248 x 10 0.2946 mA, From Eq. (5-1-42), the hole current in the emitter is has pe = AAP1B0O gr — 1) = AAPL grr — 1) Le LNs = 2X10 X 1.6 x 10°" x 2.08 x (1.5 107 = i moat x (2.248 x 10" ~ 1) = 14.976 x 10° x 2.248 x 10° = 0.337 wa From Eq. (5-1-24), the reverse saturation current in the collector isSoc.5.1 Microwave Bipolar Transistors 187 ~AqD.ni _ AgD, Po» NW Te 13.108 x 10° = 14,976 x 10°" =1312 pA From Eq. (5-1-40), the electron current which reaches the collector is leo lag = — MBO grrr w 13,104 x 10-2 X 2.248 X 108 = =0.2946 ma ‘The emitter current is Je ~Ipe + hae = ~33.67 10" ~ 0.295 * 10° = 0.295 ma ‘The collector current is de = “Heo ~ Ine = 1.312 % 10° = (0.295 x 10) .295 mA ‘The current in the base terminal is I= bye ~ (lat ~ Lae) + Teo = 33.67 X 10" ~ (-29.46 X 10° + 29.46 x 10°9) + 1.312 x 10-8 = 0.337 wA ‘Note: The recombination-generation currents in the space-charge regions are not counted, 5-1-4 Amplification Phenomena Bipolar transistors are usually used for signal amplification. The amplification phe- ‘nomena can be described from the common-base and common-emitter transistors. Common-base n-p-n transistor. The ratio of the output current to the in- put current for a small signal in a bipolar junction transistor is known as the current ain alpha a, oF hy». The current gain of a common-base p-n-p transistor is defined by the current components crossing the emitter and collector junctions as (8-1-46) where Ico = collector-junction reverse saturation current with zero emitter current Since Jc and Ir have opposite signs, the alpha a, as defined, as always positive. Typ- ical numerical values of a are between 0.9 and 0.995.198 Microwave Transistors and Tunnel Diodes Chap. 5 The emitter efficiency (or injection efficiency) is defined as current of injected carriers at Je total emitter current (5-1-47) die =( natty Tne + Tye + Toe D Nabe: where /,e = injected electron diffusion current at emitter junction Je J, = injected hole diffusion current at emitter junction Je ‘The transport factor B* is defined as injected carrier current reaching Jc eo ee )- 1-48) - injected carrier current at Je Cre) ol yn Vine Ine 2D, ww where Inc = injected electron diffusion current at Jc Ie = injected electron diffusion current at Je Ds Ve D, is the electron diffusion length At the collector we have. Te Teo — be (S-1-49) ‘Then the current gain can be expressed as am pty (6-1-50) Te In the normal active mode, the collector current is given by Te = ~als ~ leo (5-1-51) The current /co is the current crossing the p-n junction, and it is expressed in Eq (5-1-24). Then the complete expression of Zc for any Ve and Je is Te = ~ale + Ico( — e%) (S-1-52) where fo = —Ico is replaced. Example §-1-3: Silicon Bipolar Transistor A silicon n-p-n bipolar transistor operates in common-base mode at 300° K and has the following parameters: 5x 10% em? Sx 10% em-* x10" em~? Silicon intrinsic density: ‘Acceptor density in base region: Donor density in emitter region Hole lifetime:Sec. 5.1 Microwave Bipolar Transistors 189 Electron lifetime: nel us Cross section: A= 10+ em? Base width: w= 107 em Emitter length: Le = 107 cm Determine: a, The mobilities 4. and jy , The diffusion coefiicients D, and D, cc. The emitter efficiency factor y 4. The transport factor B . The current gain « Solution 1, The mobilities are read from Fig. A-2 in Appendix A as be = 200 cmi/V-s for Nu = 5 X 10" em"? Hy = 500 cmiV-s for Ny = 5 X 10" cm . The diffusion coefficients are Dy = HaVp = 200 X 26 X 10° = 5.20 emis D, = 500 X 26 X 10°? = 13.0 emis c. The emitter efficiency factor is 13.0 x 5 x 10! x 10°)? y= (+ goxs tess) 70% 4. The transport factor is ze Ue B=) eres OM fe. The current gain is = By = 0.904 x 0.997 = 0.90 Common-emitter n-p-n transistor. In the active region of a common- emitter n-p-n transistor, the emitter junction is forward-biased and the collector Junction is reverse-biased. The base current is In = -(le + he) Combining this equation with Eq. (5-1-51), we have ale T-a 1-a xleo (5-1-54) Ie In the cutoff region, if Js = 0, then Je = — Ic and the collector current is given by190 Microwave Transistors and Tunnel Diodes Chap. 5 ‘The actual collector current with the collector junction reverse-biased and the base Junction open-circuited is designated by the symbol Icro. We define the common- emitter current gain B ot hy as le + Ico B=he= (5-1-56) Since Ie + Ie + Je + Teo = 0, we have B (5-1-57) 5-1-5 Power-Frequency Limitations “The question then arises as to whether microwave power transistors have any Timita- tions on their frequency and output power. The answer is yes. Several authors have discussed this subject. Early [4] first introduced the power-frequency limitations in- herent in (1) the limiting velocity of carriers in semiconductors and (2) the maxi- ‘mum fields attainable in semiconductors without the onset of avalanche multiplica- tion, These basic ideas were later developed and discussed in detail by Johnson (5), ‘who made three assumptions: 1. There is a maximum possible velocity of carriers in a semiconductor. This the “saturated drift velocity ¢,” and is on the order of 6 x 10° em/s for elec- trons and holes in silicon and germanium. 2. There is a maximum electric field Ex that can be sustained in a semiconductor without having dielectric breakdown. This field is about 10° Viem in germa- nium and 2x 10° Viem in silicon. 3. The maximum current that a microwave power transistor can carry is limited by the base width. With these three postulates Johnson derived four basic equations for the power- frequency limitations on microwave power transistors. First equation: —Voltage-Frequency Limitation: p X 10! Vis for silicon 1 x 10" Vis for germanium Vafr = ae a (5-1-58) is the charge-carrier transit-time cutoff frequency is the average time for a charge carrier moving at an average velocity v to traverse the emitter-collector distance L Vn = Endo is the maximum allowable applied voltage oe maximum possible saturated drift velocity En= maximum electric field With the carriers moving at a velocity v, of 6 X 10° emvs, the transit time canSec. 5.1 Microwave Bipolar Transistors 191 be reduced even further by decreasing the distance L. The lower limit on Z can be reached when the electric field becomes equal to the dielectric breakdown field. However, the present state of the art of microwave transistor fabrication limits the emitter-collector length L to about 25 um for overlay and matrix devices and to nearly 250 jum for interdigitated devices. Consequently, there is an upper limit on cutoff frequency. In practice, the attainable cutoff frequency is considerably less than the maximum possible frequency indicated by Eq. (5-1-58) because the satu- rated velocity o, and the electric field intensity wili not be uniform. Second equation: Current Frequency Limitation: (aX, fr = Fete §-1-59) where /, = maximum current of the device 1 1 X= Gs Iafeee Co = collector-base capacitance It should be noted that the relationship 2xfr» ~ 2mfrt = 1 is used in deriving Eq. (5- 1-59) from Eq. (5-1-58). In practice, no maximum current exists because the area of the device cannot be increased without bound. If the impedance level is zero, the maximum current through a velocity-saturated sample might be infinite. How- ever, the limited impedance will limit the maximum current for a maximum attain- able power is the reactive impedance Third equation: Power-Frequency Limitation: E, (PAX) fr (5-1-60) 2m ‘This equation was obtained by multiplying Eq. (S-1-58) by Eq. (S-1-59) and replac- ing Valu by Pa. It is significant that, for a given device impedance, the power capacity of a device must be decreased as the device cutoff frequency is increased. For a given product of Em, (that is, a given material), the maximum power that can be delivered to the carriers traversing the transistor is infinite if the cross section of the transistor can be made as large as possible. In other words, the value of the reac- tance X- must approach zero, Thus Eq, (51-60) allows the results to be predicted. Figure 5-1-13 shows a graph of Eq. (5-1-60) and the experimental results reported from manufacturers (6) Fourth equation: Power Gain-Frequency Limitation: Ents a (5-1-61) (GuVaVn)'*f where Gy = maximum available power gain Va. = KT/e is the thermal voltage K = Boltzmann's constant, 1.38 x 10°? Ji*K192 Microwave Transistors and Tunnel Diodes Chap. 5 tot 1 Absolute maxinum ‘ohason) 1 a N. > < ¢ SS 102 Experimental reaut 10 L Ne Figure $-1-13 (P,X,)"” versus f; 01 m7 i 10 (After B. C. De Loach (6) reprinted by frinGite permission of Academie Pres) T = absolute temperature in degrees Kelvin ¢ = electron charge (1.60 x 10° C) ‘The maximum available power gain of a transistor was derived by Johnson (5] G.= (ey oe (5-1-62) where Zou and Zi are the output and input impedances, respectively. If the electrode series resistances are assumed to be zero, the ratio of the output impedance to the input impedance can be written as Yas Go Zn Con where Cis is the input capacitance and Cou is the output (base-collector) capacitance. When the maximum total carrier charges Qu move to the collector in a carrier-base transit time 7» and with a thermal voltage Va, the input capacitance C., and the emit- ter diffusion capacitance C, are related by (5-1-63) Ca = Cy = = 2B (S-1-64) ‘The output capacitance is given by (8-1-65) Substitution of Eqs. (5-1-58), (5-1-64), and (5-1-65) in Eq. (5-1-62) yields Eq. (5-1-61). The actual performance of a microwave transistor will fall far short of that predicted by Eq. (5-1-61). At present the high-frequency limit of a 28-V silicon zn-p-n transistor operating at the 1-W level is approximately 10 Ghz. Typical power gains of microwave transistors lie in the 6- to 10-4B range.Sec. 52 Heterojunction Bipolar Transistors (HBTs) 193 Example $-1-4: | Power-Frequency Limitation A certain silicon microwave transistor has the following parameters: Reactance: X10 ‘Transittime cutoff frequency: fr = 4 GHz Maximum electric field: Ex = 1.6 X 10° Viem Saturation drift velocity b= 4x 10 emis Determine the maximum allowable power that the transistor can carry. Solution, From Eq. (5-1-60) the maximum allowable power is 1_ (Env 1 1.6 x 10° x 4 x 105)? _ t= a(S) waco 5-2 HETEROJUNCTION BIPOLAR TRANSISTORS (HBTs) Bipolar transistors can be constructed as homojunction or heterojunction types of transistors. When the transistor junction is jointed by two similar materials such as silicon to silicon or germanium to germanium, it is a homojunction transistor. The transistor junction formed by two different materials, such as Ge to GaAs, is called a heterojunction transistor. So far only the ordinary homojunction transistors have been discussed. In this section we study the heterojunction transistors. 52-1 Physical Structures When the lattice constants of two semiconductor materials are matched, they can be formed together as a heterojunction transistor. This lattice condition is very impor- tant because the lattice mismatch could introduce a large number of interface states and degrade the heterojunction operation. Currently, Ge and GaAs are the two ma- terials commonly used for heterojunction structures because their lattice constants (a = 5.646 A for Ge and a = 5.653 A for GaAs) are matched to within 1%. Since ‘each material may be either p type or m type, there are four possible heterojunction ‘combinations: 1, p-Ge to p-GaAs junction 2, p-Ge to n-GaAs junction 3. n-Ge to p-GaAs junction 4, n-Ge to n-GaAs junction Fig. 5-2-1 shows the mode! diagram of a heterojunction transistor formed by n-Ge, p-GaAs, and n-GaAs materials.Problems 205, SUGGESTED READINGS IEEE Proceedings. 70, No. |, January 1982, Special issue on very fast solid-state technol- ony. Lino, S. ¥., Semiconductor Electronic Devices. Prentice-Hall, Inc., Englewood Cliffs, N.t 1990. NAVON, D. H., Semiconductor Microdevices and Materials. Holt, Kinehart ana Winston, New York, 1986, Sze, S. M., Semiconductor Devices: Physics and Technology. John Wiley & Sons, New York, 1985. PROBLEMS Microwave Bipolar Transistors 5-1. A GaAs n-p-n transistor at 300° K has the following parameters: Base width W=2x 10% em Diffusion length in emitter. Le = 2 x 10 em Diffusion length in collector Le = 4 x 10 em Base resistivity Pe = 20 A-cm Emitter resistivity pe = 1 O-cm Collector resistivity pe = 10 M-cm Emitter junction voltage Ve=04V Collector junction volage Vo = 0.5 V Cross-section area A= 0.01 cm? the impurity densities in the emitter, base and collector the mobilities in the emitter, base, and collector the diffusion lengths in the emitter, base, and collector |. the equilibrium densities in the emitter, base, and collector €. the coupling coefficients the emitter and collector currents 5-2, Derive Eq. (5-1-0) from Eq. (5-1-27). 3. Derive Eq. (5-1-31) from Eq. (5-1-30) 5-4, Verify Eq. (51-34). 56S. Derive Eqs. (5-1-39),(S-1-40), and (5-1-81) 5-6. A certain silicon bipolar transistor has a maximum electric field intensity Ex of 3 X 10° volsem and its carrier as a saturated drift velocity v, of 4 10° emisec ‘The emitter-collector length L is 4 microns a. Calculate the maximum allowable applied vokage. 9. Compute the transit time fora charge to transverse the emitter-collector length L. €. Determine the maximum possible transit frequency.206 Microwave Transistors and Tunnel Diodes Chap. 5 5-7. A bipolar transistor has voltage-frequency and current-frequency limitations as shown in Eqs, (5-1-58) and (5-1-59). Derive the power-frequency limitation as shown in Eq, (5-1-60) 5-8, A bipolar transistor has a voltage-current-power limitation. Derive the power gain- frequency relationship as shown in Eq. (S-1-61). Heterojunction Bipolar Transistors 5.9. An n-Ge-p-GaAs-n-GaAs heterojunction transistor at 300° K has the following parameters: Donor density in the n-Ge region: Ny = 2 X 10" em” ‘Acceptor density in p-GaAs region: Nz = 3 X 10" cm™? Hole lifetime: = 4X 10% see Bias voltage at emitter junction: 150 Cross section: A= 4x 10cm Compute: a, The built-in voltage in the p-GaAs side b. The hole mobility (read the value from Fig, 2-4-2) ‘c. The hole diffusion constant 4. The minority hole density in the n-Ge region . The minority electron density in the p-GaAs region f. The hole diffusion length &. The emitter junction current S10. An n-Ge-p-GaAs-n-GaAs heterojunction transistor at 300° K has the following parameters Donor density in the n-Ge region: N= 5x 10" cm” ‘Acceptor density in the p-GaAs region: Hole lifetime: Bias voltage across n-p junction: Cross section: Calculate: a, The built-in voltage in the p-GaAs side b. The hole mobility (read from Fig. A-2 in Appendix A) €. The hole diffusion constant 4. The minority hole density in the n-Ge region €. The minority electron density in the p-GaAs region f. The hole diffusion length . The emitter-junction current Describe the electronic applications of an HBT. Tunnel Diodes S-11. A negative-resistance device is connected through a 1-KO resistor in series and a 0.01-q4F capacitor in parallel to a combination of a supply voltage V of 10 V and a sig-