Trapped electron optical absorption band

The possibility of forming PbS was ruled out by the following experiments. Formation of PbS requires generation of HS from cysteine. Illumination of cysteine-modified TiC>2 did not lead to the reduction of cysteine which would result in HS formation, but led to the accumulation of trapped electrons having a broad optical absorption band at 700-800 nm. 

The optical absorption spectrum resembles that of solvated electrons in liquid water and consists of a diffuse featureless band with a broad maximum in the range of 5500-7000 A and possibly an increase in absorption towards the near ultraviolet. Since the good resolution of the e.s.r. spectrum shows that the ground states of all the trapped electrons are very similar, the broad optical absorption band must be caused by large variations in the excited states of different traps. 

Two simple types of structural defects that have been studied in MgO are the O vacancy, or F center, and the cation vacancy, or V center. A neutral O vacancy has two electrons remaining, which may be trapped by the unbalanced Coulombic potential associated with the vacancy. If one of these two electrons is removed, this results in the formation of an F+ center. Both F and F+ centers can be characterized spectroscopically by their optical absorption bands at 5.01 and 4.95 eV, respectively. 

The trapped electron is optically absorbed in the visible part of the spectrum. The spectrum is asymmetric, having a more steep drop on the side of long waves. The similar shape of the absorption band is characteristic of the solvated electrons in the polar liquids. Another feature of the spectrum is large half-width W1/2 of the band. Table II contains some characteristics of the optical absorption band of e tr in the aqueous glasses of electrolytes at 77 °K. Analogous optical bands of e tr in the alkaline and perchlorate glasses are observed in References 6, 8, 10, 18, 41, and 58. 

Trapped electrons are furthermore formed by the deposition of alkali-metal atoms on pure ice at 77°K. (3). The ice samples were microcrystalline or amorphous and from the ESR spectrum which exhibited hyperfine structure one could draw the conclusion that the electron was located in a well defined trap in which it was surrounded by six protons. The optical absorption band had a broad plateau ranging from about 600 to 680 n.m. 

Color centers can be produced in the alkali metal azide by ultraviolet light and ionizing radiation at low temperatures. The phenomenon has been of interest for some time since the defects produced are involved in the process of photochemical decomposition (cf. Chapter 7). In earlier studies [54a, b, c] purely speculative identifications of optical absorption bands with F, V, and aggregate F centers were made by analogy with the alkali halides. The most prominent visible absorption band in each case was attributed to the F center—a defect involving an electron trapped at an azide (N3) vacancy. In the case of NaNa, spin resonance [55] and recent point ion calculations [56] clearly point to the existence of a F center. However, in the case of KN3, spin-resonance studies [54a] point to the existence of molecular centers of type N2 (on low-temperature irradiation) and NJ (on room-temperature irradiation). Infrared absorptions [57] and Raman scattering [58] have been observed in the irradiated alkali azides, which can be correlated with modes associated with these defects. 

The only defects found in the azides which have counterparts in the alkali halides are the F and FJ centers (the F center consists of an electron in an anion vacancy the FJ center is an electron occupying two adjacent anion vacancies). ESR of the F center was observed by Carlson et al. [17] and by King et al. [18] in sodium azide which had been UV-irradiated at 77°K. The observed spectrum consists of 19 hyperfine lines due to the interaction of an electron trapped in an azide vacancy with the nuclear spins (/ = 3/2) of the six nearest-neighbor sodium ions. The ESR signal is correlated with an optical absorption band by thermal and optical bleaching (see below). Bartram et al. [19] have performed calculations of the wave-functions for the F center in sodium azide. Their predictions of the expected hyperfine structure and optical absorptions are in good agreement with experiment. 

Consider as a specific example KCI, a very simple compound indeed. In Chapter 2, its lowest optical absorption band was mentioned to be due to the 3p -> 3p 4s transition on the Cl ion. The excited state can be considered as a hole on the Cl ion (in the 3p shell) and an electron in the direct neighbourhood of the Cl ion, since the outer 4s orbital spreads over the K inns. Now we consider what happens after the absorption process. The hole prefers to bind two Cl" ions forming a Vk centre this centre consists of a C " pseudomolecule on the site of two O" ions in the lattice. The electron circles around the Vk centre. In this way a self-trapped excitnn is formed. An exciton is a state consisting of an electron and a hole bound together. By the relaxation process (Cl" Vk.c) the exciton has lowered its energy and is now trapped in the lattice. 

Similar to H+ implantation, optimized MgO thin films have been implanted with 1.5 MeV Li+ ions for various fluences (lO MO ions cm ). Irradiation of crystalline MgO with energetic metal ions produces stable vacancies and interstitials in the anion sublattice. Elastic collisions with energetic particles also produce cation vacancies, but these defects do not survive because the cation interstitials quickly recombine with the vacancies. Optical absorption bands can monitor these defects induced by ion implantation. Similar observations have been already done on MgO crystals after neutron irradiation (Kappers et al. 1970, Monge et al. 2000). In crystalline MgO material irradiated with Li+ ions, the well-known defects are (1) oxygen vacancies (primarily the one-electron F center), (2) oxygen divacancies Fj, (3) V and V centers (cation vacancies that have trapped one or two holes, respectively) produced by the capture of holes by existing vacancies, and (4) an unidentified defect that absorbs at 2.16 eV (572 nm) (Gonzales et al. 1991). 

There are many electrons in the sample, each in its own trap or in the conduction band (delocalized state), and there is a distribution of trap depths. The trap depths are not all the same because orientational disorder of the molecules provides a variety of polarization potential wells. Furthermore, an electron in a trap seems to couple with vibrational and librational modes of the trapping molecules. The optical absorption band of solvated electrons is very broad (Fig. 8). Part of the broadening might be caused by the distribution of trap depths, and part by the coupling with molecular modes. These parts are sometimes... 

Let us now return to MMCT effects in semiconductors. In this class of compounds MMCT may be followed by charge separation, i.e. the excited MMCT state may be stabilized. This is the case if the M species involved act as traps. A beautiful example is the color change of SrTiOj Fe,Mo upon irradiation [111]. In the dark, iron and molybdenum are present as Fe(III) and Mo(VI). The material is eolorless. After irradiation with 400 nm radiation Fe(IV) and Mo(V) are created. These ions have optical absorption in the visible. The Mo(VI) species plays the role of a deep electron trap. The thermal decay time of the color at room temperature is several minutes. Note that the MMCT transition Fe(III) + Mo(VI) -> Fe(IV) -I- Mo(V) belongs to the type which was treated above. In the semiconductor the iron and molybdenum species are far apart and the conduction band takes the role of electron transporter. A similar phenomenon has been reported for ZnS Eu, Cr [112]. There is a photoinduced charge separation Eu(II) -I- Cr(II) -> Eu(III) - - Cr(I) via the conduction band (see Fig. 18). 

---