In situations of high purity compounds it is not advised to determine the concentration of the main compound because of the always present analytical errors. In those situations one determines better all present minor and trace compounds or compounds which likely can be present due to the manufacturing process. Best analytical solution would be to use a multi-element analyzer like ICP-OES or ICP-MS.
Due to the high volatility (dense fumes, completely corrosive) of TiCl4, it seems to be one of the worst substances to perform a single test method for purity. So, i agree with Dr. Chen and surely will be profitable for you to consider the paper he suggested. Nowadays, the tecnology involved in FTIR studies could turn possible a precise sampling of this (elusive) chloride and further, revealing spectrometry. I agree also with all Dr. Cammann remarks, and, with a careful sampling and disolution in ultrapure water (milliQ please...) you would be in excelent position for a proper ICP-MS multi-elemental study, which can assure the potential presence of traces of transition metals (among others). A good bibliographic research can provide you with good advice about this topic, in particular it is could be interesting to consider the following US patents: #3533733, #2412349 and #2600881. Greets from Mendoza, Argentine.