Blogs

Lowest cost equipment in quantitative analytical chemistry typically uses titration but methods are restricted to specific methods having end point detection.  Spectroscopy in various forms allows a much wider range of elemental and molecular analysis, but typically has a much higher capital cost which can be prohibitive for school budgets.  During the 1990s, together with a Victorian company (IEC), we developed an affordable student spectrometer for colorimetric analysis, atomic emission and atomic absorption and about 100 were sold in this state (ref 1).  It’s optical arrangement involved a light source shining directly through a filter, the sample, and onto an all wavelength detector.  For spectral selection we used simple Ilford photographic filters to isolate colours. Colours in molecules are also broad band whereas high resolution inherent in simple atomic spectroscopy (eg Na and K) is put back into the Bunsen burner light source. 

When using photographic filters for selecting wavelength, I had been taught their wide transmission bands would at best give curved calibration lines (ref 2) and we did find this initially when developing the low cost student spectrometer.  High resolving power being needed for linear calibration lines. But in research in science, we are trained to question everything and I recalled from my teaching when I mathematically combined two wavelengths using Beers law, this still gave a linear graph.  To improve resolution using photographic filters, I trialled overlapping two filters but found this made the curve in calibration lines much worse.  This curve was found not to be due to the wide published bandwidth in the visible range of the filter.  When looking at the full wavelength range for these filters using a scanning spectrometer that traversed ultra violet (UV), visible (400 -700 nm) and the near Infra-Red (IR), we found these photographic filters were all transparent when above 800 nm.  Overlapping two filters in an attempt to narrow the visible transmission did not stop any of the near IR band directly illuminating our detector.  By adding a second filter that was transparent for the visual range but opaque to IR, our end result spectrometer gave linear Beers law calibration lines.  This finding was some 30 years after being taught filter spectrometers did not give linear calibration lines, but it was not explained that such curves were due to IR also activating the detector. 

 
An additional feature of the student spectrometer was the opportunity for students to visually look through the solutions and filter to the light source by removing the detector.  It surprised most students that a purple solution of permanganate would absorb best using a green filter.   This is referred to as the complementary colour and even experienced researchers sometimes overlook this conceptual fact. 

At early chemistry levels, the shape of simple molecules like H₂O, NH₃, and CH₄ are taught to be based on tetrahedral structures.   Mathematically, only the electron orbital of the electron in the hydrogen atom can be exactly calculated.  For molecules, orbital energies of the bonding electrons are not precisely calculated but involve assumptions.  The tetrahedral shape comes about by theorising s and p electrons can be promoted to four equivalent hybrid orbitals called sp3.  As the angle for H-O-H in H₂O is about 105o which is less than the tetrahedral angle (109.5o), we were told this is because water has two lone pair electron filled sp3 lobes (like rabbit ears) that repel O-H bonds.  Likewise, NH₃ has only one lone pair and CH₄ assumes the tetrahedral shape as all C-H bonds are sp3 equivalent.

These hybrid orbitals were proposed to help teach the shape of molecules more than 60 years as there was no experimental way of measuring exact energy levels of bonded electrons.  Spectroscopy of molecular energy levels in the gaseous state is restricted by transmission limits of windows in spectrometers and detectors.  Quartz windows allow transmission down to about 190 nm but the absolute energy of bonding electrons in these molecules need to cover the range down to below 50 nm.  One needs a pure vacuum path without windows.

This was achieved in the late 1960s with the advent of photoelectron (PE) spectroscopy.   In its simplest form, a vacuum chamber is constructed with a capillary to a helium electric discharge light source that gives two spectral lines in the vacuum UV light range.  The first of these lines is equivalent to 21.21 ev at 58.4 nm.  This thin beam is shone down the centre of a concentric wire grid connected to a variable voltage in the range up to 40 volts.  Very low pressures of each gas are added in turn to the chamber and the He radiation ejects electrons at 21.21 ev minus the bonding energy of that electron in the molecule.  By varying the voltage to the wire grid, the current flow as the potential is varied suddenly drops when the emitted electrons are repelled. This allows the bonding energy to be experimentally measured rather than calculated from the assumed hybridisation model.  The step height in current change allows the number of electrons in each orbital to be estimated.

From the experimentally measured PE spectra of water vapour, ammonia and methane, it was found that there are no sp3 hybrid orbitals.  The electrons in methane bonding clearly shows one S type bond and three P type bonds, likewise the PE spectrum of water does not show two equivalent lone pairs (ref 3 & 4).

We use models to simplify teaching, for example in crystal structure, atoms are assigned chosen radii to demonstrate models of crystal shape (cubic etc) but we know atoms are not simple ball shaped.   Nevertheless, even 60 years after photoelectron spectroscopy was developed, texts still teach hybrid orbitals without explaining this fundamental assumption does not exist.

Similarly, I found photoelectron spectroscopy helpful for unifying halide substitution in alkane and aromatic derivatives (ref 4) when teaching organic chemistry.  

Dr Ray Hodges FRACI (retired Associate Professor from Monash Gippsland)