Search Methods for Topologies

Depending on the access level of the user there exist up to three different search methods, the coordination (regular) search, the connectivity search and the vertex based search.
Every method provides basic search options like search by name, by spacegroup number and by transitivity.

Coordination based Search

Unregistered users can in addition to that search by coordination: searching for \(3,4\) yields all \(3,4\) nets, toggling the exclusive checkbox yields nets also featuring other coordinations (like with a wildcard: \(3,4,*\)).

Connectivity based Search

Registered users can also search by connectivity and model penalty: Each connection has to be given as two integer values denoting the respective coordination of the vertex, separated by a hyphen (-). If for example a \(3,4,6\) net with only \(3-4\) and \(4-6\) connections is desired, the connectivity input is \(3-4,4-6\). In this way the query processor knows that \(3,4,6\) nets featuring only these connections (\(3-4\) and \(4-6\)) but no other (\(3-3, 4-4, 6-6, 3-6\)) are to be returned. There is the option to allow for not excluding those nets that do feature those, but also at last one different connection (one of \(3-3, 4-4, 6-6\) or \(3-6\)) by toggling the exclusive checkbox.

Search refinement by Geometry

The connectivity search can be aided by adding a desired amount of geometry constraints via the Add Penalty button below the search form. Geometries can then be selected via a drop-down menu that lists the coordination number and name of the respective geometry. In addition to that, each selected geometry is instantly displayed in a small 3Dmol-js window below the search form:

Vertex based Search

The third method is the vertex based search method. It is the most powerful and flexible search tool, which grants the ability to search for specific connectivity patterns that for example describe sets of vertices that were deconstructed as shown in the section.
There is a certain logic to identify these vertices by their connectivity w.r.t. each other including their connectivity count (stored in the adjacency matrix):

Initially the user has to select the number of vertices, the server then generates the search form with this number. The figure below shows the relevant input for the upper example from the above figure (click here to test the search on your own):

Every field can now, but does not have to be filled. CN is the coordination number. Connected to and Not connected to can be filled with the ids of vertices and Count can be given to define the number of connections per vertex. The latter has to be given in the same order as the respective vertex is mentioned in Count: In the example, the vertex with \(\text{id}=0\) is connected to \(0\) (itself) once and twice to vertex with id \(1\). \(-1\) can be passed as a Count wildcard in case CN is not defined which would limit the sum of the elements of Count to CN.

The input for the second looks as follows (click here to test the search on your own):