Wafer doping

Figure 12. Photograph (4y magnification) of an electrochemically deuterium-doped rutile wafer. Doping was accomplished by current-controlled cathodic aging an undoped TiOz wafer in LiOD, D20 for 3 days at 10 mA. This resulted in a shattering of the sample in the region exposed to the electrolyte.

Silicon wafers doped with phosphorus (resistivity 0.1 D. cm) have been used as substrates. A composite aluminum+silicon (Al+Si) films were fabricated by magnetron sputtering of a compound target containing 45 at. % of Si and 55 at. % of Al. The films thickness was varied in the range of 20-140 nm. Then... 

In order to investigate the photochemistry of CI2 on a solid substrate, multilayered CI2 condensed on an Si wafer (doped by at a rate of 3 x 10 cm ) was irradiated by excimer laser light ( 10 mJ/cm ) at 1 93, 248, and 352 nm. The substrate was cooled to 100 K. Photoproducts ejected from the substrate were analyzed by a mass spectrometer (12). 

Flat n -type (100) Si wafers doped with antimony up to 4T0 cm were used as initial substrates. Curvature radii of the Si wafers were measured to be more than... 

Wafer doping (p-type) Wafer doping (n-type) Decrease Increases Increases Increases Increases... 

These processes are considerably more complex in actual CMOS fabrication. First, the lower layers of a CMOS stmcture typically have a twin-tub design which includes both PMOS and NMOS devices adjacent to each other (see Fig. 3b). After step 1, a mask is opened such that a wide area is implanted to form the -weU, followed by a similar procedure to create the -weU. Isolation between active areas is commonly provided by local oxidation of sihcon (LOCOS), which creates a thick field oxide. A narrow strip of lightly doped drain (LDD) is formed under the edges of the gate to prevent hot-carrier induced instabiUties. Passivation sidewalls are used as etch resists. A complete sequence of fabrication from wafer to packaged unit is shown in Figure 10. 

To achieve the lowest possible delay a bipolar switching transistor developed by IBM minimizes parasitic resistances and capacitances. It consists of self-aligned emitter and base contacts, a thin intrinsic base with an optimized collector doping profile, and deep-trench isolation (36). Devices must be isolated from each other to prevent unwanted interactions in integrated circuits. While p—n junctions can be used for isolation, IBM s approach etches deep trenches in the siUcon wafer which are filled with Si02 to provide electrical insulation. 

Doping of the siHcon can have a large effect on the etch rate, and layers of different materials such as Si02 and Si N can have different etch rates. Eor pressure sensors, thin diaphragms of Si or related materials are etched into the wafer (see Pressure measurements). 

Figure 14-12. Various types of OFETs. (a) Inverted coplanar on a highly doped Si wafer, (b) inverted coplanar on a neutral substrate, (c) inverted staggered oil a neutral substrate, (d) inverted staggered using the dielectric layer as the substrate.

The patterned wafer might next be placed in a diffusion furnace, where a first doping step is performed to deposit phosphoras or boron into... 

The process of substituting elements for the silicon is called doping, while the elements are referred to as dopants. The amount of dopant that is required in practical devices is very small, ranging from about 100 dopant atoms per million silicon atoms downward to 1 per billion. Dopants are usualty added to the silicon after the crystal growth process, when an integrated circuit is being formed on the surface of the wafer. 

The fact that pure silicon is not a conductor but is "doped to make it a good semi-conductor might seem odd, particularly since considerable effort goes into obtciining very pure silicon. Nonetheless, the undoped silicon is not conductive. This allows one to form discrete areas of n-type and p-type silicon on the same wafer and to position these in conjunction with each other in a manner so that a current amplifying device results. An example of such a device is shown in the following diagram ... 

Figure 13.5 (a) Fluorescence micrograph of the self-spreading lipid bilayer doped with a dye molecule. The lipid bilayer spread on an oxidized silicon wafer from a deposited lipid aggregate illustrated on the left, (b) A schematic drawing of the selfspreading lipid bilayer from the lipid aggregate. Adapted from Ref [48] with permission. 

In the framework of CUORICINO [41] and CUORE [42] experiments (see Section 16.5), Ge crystal wafers of natural isotopic composition have been doped by neutron irradiation, and the heavy doping led to materials close to the metal insulator transition. Several series of NTD wafers with different doping have been produced [43], After an implantation and metallization process on both sides of the wafers, thermistors of different sizes can be obtained by cutting the wafers and providing electrical contacts. 

Here we will report measurements on the heat capacity of two NTD Ge 34B wafers, one non-metallized (only doped and annealed) and the other metallized (by B+ implantation and Au deposition). The comparison of data obtained from the non-metallized NTD Ge wafer and from the wafer with electrical contacts revealed an excess heat capacity, which can be attributed to the implantation process with B ions [44],... 

NTD wafers were produced by irradiating natural ultra pure Ge crystals by means of a flux of thermal neutrons (see Section 15.2.2). To realize the electrical contacts, both sides of the wafers (disks, 3 cm in diameter, 3 mm thick) were doped by implantation with B ions to a depth of 200nm. The implanted layers are doped to such a high concentration that the semiconductor becomes metallic. Then a layer of Pd (about 20 nm) and Au (about 400 nm) was sputtered onto the both sides of the wafers. Finally, the wafers were annealed at 200°C for 1 h. The wafers are cut to produce thermistors of length 3 mm between the metallized ends (3x3x1 mm3 typical size) the electrical contacts are made by ball bonding with Au wires. 

Fig. 12.16. Heat capacity per unit volume of the two neutron transmutation doped (NTD) Ge wafers.

In the literature [55], typical energies involved in the nuclear quadrupole moments -crystalline electric field gradient interactions range up to A E 2x 10-25 J. The measured AE seems to confirm the hypothesis that the excess specific heat of the metallized wafer is due to boron doping of the Ge lattice. 

---