Obras de caída

changes of potential

V; and computing them into one spectrum and comparing this with the result of the coaddition and computation of all the interferograms collected at Vj (e.g. 20 interferograms, 10 collected at V; and 10 at Vf, giving rise to 2 spectra, one representing the species in solution at V; and the other at V f). The second approach

yields spectra with better signal-to-noise, but only in chemically well-behaved systems. When the spectral changes occuring on each cycle are not repeatable, this method of analysis is unsuitable since the resultant interferograms are not of a well-defined product. Indeed, in most other cases it is preferable to analyse each interferogram individually.

The precise synchronisation necessary for this type of experiment requires the potentiostat to be interfaced with the FTIR interferometer. A BBC microcomputer is used for this purpose. The 5 V pulse derived from the white light interferometer at the commencement of every scan is detected by the BBC microcomputer and the time separation between consecutive pulses monitored. A change of potential is initiated by the computer program only after the detection of a longer, data-collection scan. The short, non data-collection scans of the interferometer are ignored. A listing of the BASIC computer program written to perform this task is given in Appendix I. The BASIC program used to control normal spectroelectrochemical experiments is also listed (Appendix II).

The hardware configuration of the experiment is as follows:- The white light interferometer, emitting regular 5 V pulses, is connected to the microcomputer using the printer port. Once the trigger is detected, the microcomputer sends out a signal to the potentiostat via a digital-to-analogue converter, which is connected to the lOMHz bus. The circuit diagram of the digital-to-analogue converter and its amplifier are given in Appendix III.

A gas chromatography software package, with time-resolved spectroscopy capabilities is included in the Bruker data system, and this provides the software interface necessary to control the timing of interferogram acquisitions. By constructing a variety of macro programs within this software, the times between consecutive data- collection scans can be controlled.

It is often found that the spectral data obtained in "modulation" spectroelectrochemical experiments have poor signal-to-noise ratios, which may make

subsequent analysis of the spectra difficult. This is ameliorated by using a BASIC computer program to simulate the spectral bandshapes. In this way, heights, full widths at half-height, positions and Gaus sian/Lorentzian ratios can be determined for all peaks. A listing of this program is included (Appendix IV).

The behaviour of the current, as a function of the potential applied to the working electrode, was monitored using a Y-T chart recorder. This provided valuable additional information related to the spectroelectrochemical experiments. The current­ time graph had two features of interest - the charging current and the reaction current. The former represents the amount of current required to charge the working electrode to the necessary potential and is a function of the size of the electrode; its time dependence is a function of the current output and compliance voltage of the potentiostat. The latter is a function of the current flowing in the cell as a result of the electrochemical reaction taking place. Both observations are useful in determining whether a particular reaction is proceeding in a chemically-reversible fashion. This is indicated by reproducible behaviour by each.

1.3 Ultra-violet/visible Spectroelectrochemistry

1.3.1 Ultra-violet/visible Spectroscopy

Ultra-violetA^isible spectroscopy provides a powerful tool with which to probe the structure of transition metal complexes." In cases where there is partial d-orbital occupancy, many of the visible spectral features are due to crystal field bands. The more intense of these are spin-allowed transitions and the less intense spin-forbidden transitions. Virtually all compounds, however, absorb in the ultra-violet region, such that spectroscopy in this region is a generally applicable technique. Most charge- transfer bands (electronic transitions between the metal and the ligand) and transitions within the ligand itself (usually n -^n or a->a* transitions) occur in the ultra-violet region of the spectrum.

Transition metal spectra were initially explained in terms of crystal field and, subsequently, ligand field theory. These state that the imposition of a non-spherical field of charge around a metal ion results in the removal of the d-orbital degeneracy.

However, crystal field theory treats ligands only as point charges. Metal-ligand interactions are taken into consideration in Molecular Orbital theory. The geometry of the ligand field determines the nature of the d-orbital splitting, whilst the ligand itself will have a significant effect upon the magnitude of this splitting. The latter phenomenon is the basis of the spectrochemical series of ligands.

The intensities of spectroscopic bands conform to the Beer-Lambert Law:-

Absorbance = log(I/Io) = eel

Io=incident intensity I=transmitted intensity

e=molar decadic extinction coefficient (dm^mol'^cm'^) c=concentration (moldm^)

l=path length (cm)

e is a measure of the probability of the electronic transition. Spin-forbidden crystal field transitions will have molar decadic extinction coefficients of the order 1-500 dm^mol^cm \ the more intense, charge transfer bands have values ranging from 1000 to 10 000 dm^mol^cm \

1.32 Instrumentation

An Oriel Instaspec HI system, incorporating multi-channel diode array detection, was used for all measurements in the ultra-violet/visible region. Light was channelled into a 0.25 m monochromator, using fibre optic light guides, and dispersed onto a 1024 pixel diode array. A spectrum could be recorded over 700 nm, if a 150 lines per millimetre grating is used, in a minimum of 0.016 s. The instrumentation was controlled from an IBM-compatible computer with 386sx processor and an enhanced graphics monitor. The computer also allowed a considerable degree of spectral manipulation as well as flexibility in data collection. The source was a 50 W/12 V tungsten lamp, normally operated at 8-9 V.

The IRRAS cell described above (section 1.2.3) was used without further modification, and in an analogous fashion, for ultra-violet/visible studies. However,

a different set o f apparatus was necessary to direct the beam from the source onto the c e ll and then into the detector (Fig. 1.7).

IRRAS

cell

Lens 3 Lens 1 Aperture Lens 2

Source Fibre optic guide to multi-channel detection system Mirror 1 • Mirror 2