Sorry for maybe too late reply, I hope you were able to solve the column.
Anyway, here is my advice:

1) your column feels way too small for such amount. I prefer to use shorter but wider columns because its faster (you trade number of theoretical plates for their size and since the size has square relation to column diameter unlike the number of plates vs lenght which is linear). I would use 2 or 3 cm diameter and maybe 10-12 cm of silica in the column

2) for such amount, taking 2 ml fractions is overkill. 10 or 20 is good volume here in my opinion - cuts time, glassware use and number of TLCs to do.

3) you dont have to TLC all your fractions, I like to make a grid on TLC plate, check which fractions has SOMETHING and then elute only those containing something. I also run TLC during the column. This way you can figure out when to stop (you obviously stop when your compound of interest is eluted)

4) as one of my advisers told me commonly: "Color is five percent". It is SOME indication but hardly conclusive. Ofc, after you run the same column multiple times you can get more information from the color of fractions ("my compound is eluting AFTER this yellowish band so I can take all the stuff into a beaker before it starts eluting")

5) you can increase polarity of your mobile phase over time to get better separation

6) Based on by experinece, I feel you might need to run more fractions, run TLC on maybe each fifth fractions to get SOME idea but 90 ml of mobile phase doesnt seem to be enough in your case.