njit-etd2000-029 - New Jersey Institute of Technology

More documents

Recommendations

Info

7 serious problem since the power is converted into heat. Consequently, the amount of heat produced also increases by a factor of four. This excess heat must be dissipated somehow or the temperature of the chip will increase until something catastrophic occurs. As it is, a standard chip produces about the same amount of power as a dim light bulb. This may not sound like terrific amount, but it is actually several times more than the amount of heat produced by the same area of an element on an electric cooker. The usual way to combat this problem is to apply a technique known as constant field scaling. Since the distance between the source and drain has been reduced by the scaling factor of two, we need to reduce the voltage difference by a similar factor in order to obtain the same gradient as before. The switched voltage is reduced by the scaling factor, as is the current flowing through the device. Consequently, the power produced by each device decreases by a factor of four. The greatest problems facing further miniaturization of integrated circuits concern the humble interconnect, the tiny connecting lines that form the electrical links between the devices on the chip. Various materials are used to form these conducting elements, ranging from heavily doped silicon, to silicides and metals, principally aluminum and copper. One of the key problems with interconnects is their resistance to the flow of an electric current. The time required for propagation of electrical signals depends on two factors. These are the resistance of the wire to the flow of current, and the capacitance, or amount of charge stored in the system. The time taken for a signal to propagate along a wire is therefore dependent on what is called the RC factor, the product of resistance and
8 capacitance of the wire. It may seem surprising that the time depends not explicitly on the length of the wire, but there is a strong implicit dependence since both resistance and capacitance vary with the length of the wire. The most obvious approach when reducing the size of components in integrated circuits is to scale the interconnects in the same manner, so that the width and thickness of the connecting wires are decreased accordingly. We can predict the effects of these measures by considering how the resistance and capacitance change under such scaling. Let us examine capacitance first. The interconnect runs across the top surface of the semiconductor below. The amount of charge stored in the interconnect depends inversely on the thickness of the oxide layer separating the interconnect from the semiconductor wafer. We have proposed that the width of the interconnect is reduced by the scaling factor of two, but how does the length of the interconnect change? For the time being let us suppose that it is also subject to the same factor. From these considerations we estimate that the area of the interconnect in contact with the surface of the oxide is a factor of four times smaller in the scaled circuit. However, this partially offset by the fact that the oxide layer is reduced in thickness, and so overall the capacitance decreases by a factor of two. This result seems quite promising, but the effects of scaling on resistance are not so good. As the cross-sectional area of the wires is reduced it becomes increasingly difficult for the current to flow through the wires, and so the resistance increases. If both the thickness and width of the interconnects are scaled, then the cross-sectional area is reduced by a factor of four and the resistance increases accordingly. However, since the resistance is
Page 1 and 2: Copyright Warning & Restrictions Th
Page 3 and 4: ABSTRACT SYNTHESIS AND CHARACTERIZA
Page 5 and 6: SYNTHESIS AND CHARACTERIZATION OF L
Page 7 and 8: APPROVAL PAGE SYNTHESIS AND CHARACT
Page 9 and 10: This thesis is dedicated to my pare
Page 11 and 12: TABLE OF CONTENTS Chapter Page 1 IN
Page 13 and 14: TABLE OF CONTENTS (Continued) 2.6 A
Page 15 and 16: TABLE OF CONTENTS (Continued) Chapt
Page 17 and 18: LIST OF TABLES Table Page 1.1 Evapo
Page 19 and 20: LIST OF FIGURES (Continued) 4.2 Fil
Page 21 and 22: LIST OF FIGURES (Continued) 4.43 SE
Page 23 and 24: 2 of transistors on a chip would co
Page 25 and 26: 4 Typically, light beams do the etc
Page 27: Having the technology to transfer t
Page 31 and 32: 10 side of the chip to the other. O
Page 33 and 34: metal/oxide/semiconductor integrate
Page 35 and 36: 14 properties of the film obtained
Page 37 and 38: 16 Chemical vapor deposition is dis
Page 39 and 40: 18 Figure 1.1 Scheme to show the tr
Page 41 and 42: 20 the low-pressure operation. Dry
Page 43 and 44: 22 • Reaction: The surface cataly
Page 45 and 46: 24 If the deposition process is lim
Page 47 and 48: 26 1.2 Types of CVD Process As ment
Page 49 and 50: 28 Lowering the gas pressure by abo
Page 51 and 52: 30 purity and density, controllable
Page 53 and 54: 32 to block alkali ions. All films
Page 55 and 56: 34 Some of the attractive propertie
Page 57 and 58: 36 1.5 Refractory Materials Ceramic
Page 59 and 60: 38 deposited and then a conducting
Page 61 and 62: 40 Table 1.7 Physical properties of
Page 63 and 64: 42 In Chapter 2, a review of litera
Page 65 and 66: Material Melting Temperature (°C)
Page 67 and 68: 46 The reaction involving BCl 3 for
Page 69 and 70: 48 B2H6 + 2NH3 = 2BN + 6H2 Rand et
Page 71 and 72: 50 ammonia at 250° - 600°C in an
Page 73 and 74: 52 2.3.1 Synthesis by CVD and LPCVD
Page 75 and 76: 54 temperature range of 700-1300 K
Page 77 and 78: 56 The refractive index of c-BN was
Page 79 and 80:
58 the problem to hydrogen content
Page 81 and 82:
60 Many researchers 70,69,82 have s
Page 83 and 84:
62 BN film coatings are useful to i
Page 85 and 86:
64 2.5.1 Dielectric Thin Films With
Page 87 and 88:
66 A family of thermally stable, lo
Page 89 and 90:
68 suggested that the there is a ch
Page 91 and 92:
70 a heavily dependent on process c
Page 93 and 94:
72 Typically thin beryllium or poly
Page 95 and 96:
Figure 2.2 Trends in ULSI miniaturi
Page 97 and 98:
76 Figure 2.3 Basic SCALPEL princip
Page 99 and 100:
78 50 nanometers and smaller. The r
Page 101 and 102:
80 A large number of materials have
Page 103 and 104:
82 2.7 Aim and Scope of the Present
Page 105 and 106:
84 jacket maintained the temperatur
Page 107 and 108:
86 Figure 3.2 Schematic diagram of
Page 109 and 110:
88 contamination and hence requires
Page 111 and 112:
90 known reaction chamber volume. T
Page 113 and 114:
92 all the experiments. The boat wa
Page 115 and 116:
94 limited by sample surface and a
Page 117 and 118:
96 3.5.6 Film Stress Stress develop
Page 119 and 120:
98 3.5.7 X-Ray Diffraction (XRD) A
Page 121 and 122:
100 where hγ , Eg(opt) and K 1 den
Page 123 and 124:
102 technique. However the films we
Page 125 and 126:
104 same film surface. Such measure
Page 127 and 128:
106 the device, an electrical conta
Page 129 and 130:
108 Figure 3.5 Process flow diagram
Page 131 and 132:
110 samples. The recombination-gene
Page 133 and 134:
112 3.7.2 Characterization of B-N-C
Page 135 and 136:
114 In all the experiments, films d
Page 137 and 138:
116 operating in the mass transfer
Page 139 and 140:
118 variation in the NH3/TEAB flow
Page 141 and 142:
120 mode, and a small peak close to
Page 143 and 144:
122 infrared absorption spectra, no
Page 145 and 146:
124 increasing the NH3/TEAB flow ra
Page 147 and 148:
126 tensile behavior with increase
Page 149 and 150:
128 material with the electrons jum
Page 151 and 152:
130 while at a constant temperature
Page 153 and 154:
132 4.4.6 Electrical Characterizati
Page 155 and 156:
134 4.4.6.2 Investigation of LPCVD
Page 157 and 158:
136 Figure 4.17 C-V curve shift alo
Page 159 and 160:
138 Figure 4.18 Capacitance-Voltage
Page 161 and 162:
140 Thus by knowing the refractive
Page 163 and 164:
142 rate of 40 sccm, as the Power w
Page 165 and 166:
144 4.4.7.2 Characterization of B-N
Page 167 and 168:
146 the reactant species in eq. 4.2
Page 169 and 170:
148 Figure 4.25 Variation of Film C
Page 171 and 172:
150 Figure 4.27 Plot of growth rate
Page 173 and 174:
152 4.5.3 Rate Equation Knowing the
Page 175 and 176:
154 4.5.4 Reaction Mechanism In the
Page 177 and 178:
156 high TiC14 coverage. ΘTiC13 ca
Page 179 and 180:
158 Figure 4.30 shows a typical XPS
Page 181 and 182:
160 degraded. Hence, films with thi
Page 183 and 184:
162 Figure 4.35 Variation in film d
Page 185 and 186:
164 Figure 4.37 Variation in film r
Page 187 and 188:
166 emittance increased; the transm
Page 189 and 190:
Figure 4.41 SEM micrograph of TiN d
Page 191 and 192:
Figure 4.4 3 SEM micrograph of TiN
Page 193 and 194:
172 Orientation (hid) Table 4.1 Com
Page 195 and 196:
174 carried out under the same proc
Page 197 and 198:
176 was not in a compound form. It
Page 199 and 200:
REFERENCES 1. R. Turton. 1995. The
Page 201 and 202:
180 36. W. Kern and V. S. Ban, "Che
Page 203:
182 74. L. Imhoff, A. Bouteville, a
show all

njit-etd2000-029 - New Jersey Institute of Technology

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?