1.3.4 Detectors: Photodiode, Photomultiplier, Charge Coupled Device
The fundamental part of a spectroscopy experiment is the detector. It should be sensitive in order to detect the photons also at very low number for unit of time, and avoid having noise signal due to electronics. The most used detectors in UV‐Vis‐IR spectroscopy are the photodiode, the photomultiplier (PMT), and the charge coupled device (CCD). The most sensitive among them is the PMT, which is also characterized by a fast enough time of detection, response time. The photodiode and the PMT are typically used in scanning spectroscopy system, where they are coupled to monochromators that select the wavelength to be revealed. The CCD is typically coupled to a grating and is able to contemporarily detect many wavelengths, thus enhancing the speed of recording of a spectrum. The working principle of these detectors is now briefly summarized [2, 10, 22, 25].
The photodiode is a solid state detector based on the junction of two extrinsic semiconductors with n‐ and p‐doping [14, 25]. In the junction, the charge depletion region is formed that is characterized by the presence of an intrinsic electric field [25]. When photons of energy larger than the bandgap of the semiconductors impinge in this region, an electron–hole couple is generated and is separated, ejecting these charges toward the neutral regions of the two semiconductors. This effect induces a current, if the photodiode is short‐circuited or, equivalently, a voltage if the photodiode is open circuit. Both these signals are proportional to the number of impinging photons per unit of time, or equivalently to the intensity of radiation. A saturation effect for high radiation intensity could occur due to the recombination of electron–holes simultaneously generated in the depletion region, not giving a current signal. Typical spectral sensitivity of the photodiode depends on the semiconductor used and its bandgap. For Si, the spectral range is 200–1100 nm, for Ge 400–1800 nm, and for PbSnTe 2000–18 000 nm [10, 25]. Many different combinations exist nowadays, showing the extension of spectral response from far‐IR to UV [25]. One limit of a photodiode is related to the generation of thermal couples of electron–holes in the depletion region giving rise to a noise signal. To contain this effect, low‐temperature detectors are employed, cooling them with liquid nitrogen (77 K) [22].
The photomultiplier is formed by a photocathode that is an emitting layer that when illuminated with photons ejects electrons by photoelectric effect [2, 10, 12]. A second electrode acts as an anode to accelerate the photoelectrons. Beneath this element is present an array of secondary emitters of electrons, each called dynode. In particular, each electron impinging with opportune energy on a dynode is able to release other electrons. Opportunely accelerating the photoelectron emitted by the photocathode toward the anode, the electron has kinetic energy large enough to release the secondary electrons from the first dynode. By applying the opportune voltage, the secondary electrons are accelerated toward the next dynode and they release other electrons. An avalanche process can be generated this way. In particular, the impinging photon effect is amplified by a factor δ n by an array of n dynodes, each emitting δ electrons characterizing the photomultiplier by a high gain. Materials used for the photocathode are metallic leagues: CsNa2, KSb, Cs3Sb, KCsSb, and also semiconductors: GaAs, InGaAs, that are characterized by a low work function [10, 14]. For the dynode, other leagues are used: Be‐Cu, GaP. A general property of the PMT, apart from its large gain, is that the spectral response depends on the employed material. It has the drawback of the dark current due to spontaneously emitted electrons by the dynodes that are accelerated by the high voltages usually employed to reduce the response time of the PMT. Indeed, this voltage is used to reduce the travel time of the electrons in the array of dynodes and the PMTs are very fast detectors often employed in time‐resolved spectroscopy. These detectors are also used in the photon counting mode that enables to detect very low‐intensity radiation (few events in time). In this procedure, the photoemitted electrons originate a high‐intensity signal in short time, because they are correlated to the arrival event of a photon and they traverse the entire dynodes array, which can be distinguished from the low‐intensity signal continuous in time due to thermal electrons generated with larger probability by few dynodes [10]. Typically, PMT cannot work in the low energy range of photo‐detection since the work function of metals is not low enough.
The CCD detector is constituted by an array of metal‐oxide‐semiconductor (MOS) capacitors [25]. These are typically p‐type doped Si covered by SiO2 insulator and with top metallic gate. Electric contacts on top metal and bottom Si enable the switching of each MOS between the accumulation and depletion or null state [10, 25]. By applying a positive voltage to the gate, the depletion regime is activated in the Si near to the Si–SiO2 interface, implying the accumulation of negative charge. If photons impinge the depleted region in this regime, they increase the number of negative charge trapped. This number is proportional to the intensity of light. In a CCD, the array of MOS is opportunely wired and subjected to a sequence of voltages to let three nearby MOSs constitute a pixel where the central MOS is active and the two edges inactive when the detector is exposed to light. As a consequence, during light exposure, the charge is accumulated only in the central MOS of a pixel. Successively, an applied voltage time sequence opportunely transfers the charge from the central MOS to one of the nearby in a sequence to let the accumulated charge be read by an opportune system. In this way, the array of pixels “traps” the information from the impinging radiation synchronously. Typically, CCDs are formed by arrays of multiples of 256 pixels in a square or rectangular matrix with rows and columns of given spatial extension depending on the construction of MOS. The CCD is usually put after a grating so each pixel is hit by a different wavelength due to the spatial dispersion imposed to the light. The columns of pixels select the wavelength whereas the rows enhance the signal reading for each wavelength. The advantage of a CCD is the recording of all the wavelengths in a spectral range synchronously without the need to change the position of the grating.
1.4 Case Studies
Examples of absorption and photoluminescence measurements will be reported in this paragraph in order to give some hints on the settings to carry out a satisfying experiment and with the aim to let correlate information to reconstruct the electronic energy levels of a given chromophore.
1.4.1 Optical Absorption in Visible‐Ultraviolet Range
A typical experiment of absorption consists in the investigation of the light transmitted by a given sample. The block scheme of the instrumentation has been presented in Section 1.3.1 and it includes a light source and a system to select the wavelength interacting with the sample. Two possible configurations are typically used. In the first, the wavelength of the source is selected by a monochromator and it is sent to the sample and the transmitted light is detected. In the second, all the radiation of the source is sent to the sample and the analysis of the wavelength is done after the sample by a grating and a CCD. Obviously, in each case, a blank or a reference spectrum of the source should be done without the sample to determine the intensity of light arriving at the sample before to traverse it and give rise to the absorption process. In Sections 1.4.1.1, 1.4.1.2 examples of the practice to apply to obtain good spectra are reported.
1.4.1.1 Scanning Device (Bandwidth and Scanning Speed Effects)
In a scanning spectrophotometer, the wavelength is selected before to send the light to the sample. As a consequence, it is necessary to take into consideration the time to select the wavelength and correctly connect it to the time necessary to record the signal passing through the sample by the detector. The latter is typically determined by the detector response time (RT). The time to select the wavelength is the scanning speed (SS) of the spectrophotometer. Both parameters can be adapted by the user, in particular larger response time typically warrants a high signal‐to‐noise ratio and a good analysis of the results. Considering that a monochromator selects a wavelength with a range around it,