About Mechanisms in Inorganic Chemistry

(Chemistry in the School of Natural and Environmental Sciences, University of Newcastle upon Tyne, England)

History

These web pages were originally the results of two undergraduate web programming projects, by Mr. J.N.R. Butler in 2003 on dissociative substitution mechanisms in octahedral transition metal complexes, and by Mr. J. Marshall in 2004 on associative substitution mechanisms in square planar complexes.

In each case the aim was to produce animated molecular models showing reactions in progress, to illustrate topics in Inorganic Mechanisms undergraduate courses.  The models could be used with a projector as lecture aids, or in subsequent private study by the students taking those courses.

Prof. R.A. Henderson devised the scopes of the projects, with a view to using the models in his teaching, and advised the project students on chemical content.  Dr. B.W. Tattershall devised the required computer programs to present the models, and taught and supervised the model-making and website parts of the projects.

The Idea

The key feature of these pages is that they each can show two mouse-rotatable images, side by side, of ball-and-stick models.  The models consist of a number of animation frames to show the progress of the reaction, with the left hand model going from a starting point to the reaction intermediate, and the right hand model going from the intermediate to the finishing point.   When the first animation is complete, the second continues from that point automatically.   A clickable reaction scheme is used to switch the models on or off, so allowing a particular route to be chosen in the more complex schemes.

The Method

In the original projects, the Chime/Rasmol browser plugin was used to show the models.   It could read model files consisting of concatenations of sets of .xyz atomic coordinates, with one set for each animation frame.   User control was by the clickable schemes and HTML form buttons to change the mode from Animate to First Frame Only or vice-versa, and all such choices were interpreted by a server-side .php script to redeliver the appropriate modified page.

Update to use Jmol/JSmol instead of Chime

Following the disappearance of the Chime plugin, Dr. B.W. Tattershall in 2018 undertook the conversion of the pages to use the Jmol/JSmol applet instead.   This could read the original .xyz animation files and use Chime/Rasmol script files to create the same appearance of the models.   A bonus was that Jmol buttons could be used client side to pause or play, etc., the two models individually, allowing more convenient user interaction for exploring the reaction path.   The original HTML form / .php script mechanism for model and mode selection was retained as well, though all such user interactions result in reloading of the Jmol models and the animations, if switched on, restarting at the beginning. Reloading JSmol applets is a lot slower than just downloading the relatively small .xyz files required by the client-side Chime plugin, but is still quite tolerable for the originally intended uses. An advantage is that the user needs only a standard browser such as Chrome or Firefox to use the JSmol option.   IE11 is also adequate for these models.

The scripts developed for these pages could be adapted quite easily to show other reaction mechanisms as two consecutive animations, provided the molecular models required for the animation frames may calculated

Why Use Rotatable Animated Models?

For some teaching purposes, molecular models are necessary. Pre-prepared rotatable computer models are preferable to physical model-kit models in that there is no barrier of effort or cost to the student, and misleading mistakes in model building are avoided.   They are particularly useful for students to explore symmetry, as in the case of
Ring Flipping of Cyclohexane or where a structural rearrangement is difficult to draw convincingly or to show by manipulating a physical model.   An example of that is the Racemisation Reaction of P3Se4I

For substitution reactions in transition metal complexes, however, the need is not so obvious.   A well-taught student should be able to draw these structures and changes clearly and without undue effort on paper, to show the salient points of the mechanisms.   The benefit of the present web pages is quite different:   it is to show the subject as attractive and interesting to students who are accustomed to professionally produced computer graphics in many areas of everyday life.

We make no claim that our models represent accurate reaction paths.   However, they convey the desired ideas about the reactions to students, and we believe they are interesting

What Models are Used?

We need to ask 'To what extent do we wish to represent reality?', and for this purpose we need to inquire what reality is in this context.   Clearly, a two-dimensional image on a computer screen, of model-kit plastic balls and sticks, is a long way from what is being represented at the molecular level.   We could have shown abstract models of starting point, transition state, and final state, as might be drawn in a textbook, and generated animation frames between these points.   A much more comprehensive web site about substitution mechanisms has appeared, since our original one, at the University of Valencia and appears to take this approach in producing an attractive and instructive product.

Alternatively, it would be possible to calculate a minimum energy reaction path along the reaction coordinate, using ab initio methods for a real example.   That was what was done, albeit at a quite low level, and with some interpolations, for Ring Flipping of Cyclohexane.   However, for transition metal complexes of interest, this would be a major computational chemistry effort, well beyond the scope of an undergraduate project at the time.   Even if this was done, one could still ask how real it was, even if calculated at a high level of theory.   In reality, there would be a spread of molecules to higher energies than on the minimum energy path.

For the present pages, we chose to use as animation frames, the points in Gaussian geometry optimisations, going towards the geometries of stable molecules from starting points in which we pulled away a substituent to a suitable distance.   For dissociation steps, which would not go spontaneously in the gas phase, the sequence of frames for the reverse, association, step, was used in the reverse order.

For the sake of computational speed, the calculations were done at the pm3 semi-empirical level.   While realistic ligands could be modelled, use of pm3 in Gaussian 98 meant that the only transition metal available was zinc, so this was used even though zinc might not display the chemistry found for other metals which give less labile complexes.   The identity of the metal in the models is not obvious to the user unless they look for it.

Square Planar Complexes

The substitution in square planar complexes presented the particular difficulty that we wished to model a trigonal bipyramidal geometry, in which halide ligands stayed in equatorial positions.  The minimum energy configuration for zinc, as for a main groups element, is with halides in the apical positions.  To deal with this, we modelled a diethylenetriamine complex, in which the chelate tridentate ligand would preferentially occupy both apical positions, leaving the remaining equatorial positions for the halides.  We then converted the reaction animations, frame by frame, to have ammonia ligands rather than the original triamine, using a Z matrix editor.  We found the program Molden extremely useful for this purpose, as well as, throughout the projects, for building initial structures and extracting coordinates for each optimisation step.