There could be research questions where very negligible research has been done. That, in such questions there would be issue in getting proper references.
There are many variables, so it is hard to give an exact number. As Hein says, more is better.
For my last five journal articles, I had 20, 28, 39, 50 and 50 references, respectively.
My rule of thumb is that at least half should be from within the last five years, so that you can show you are using the most recent literature as the basis for your study.
This issue raised by Bidhan Mohapatra has been well responded to by Hein Retter that there are no fast and hard rules about the upper or lower limits that are set for authors.
I would like to suggest that if possible, make your lowest limit to be five references, ten to fifteen (10-15) for average, and the upper limit could be between 30 to 50 references.
What is vital is the context and content of the article. The articles that are informed by more references preferred in the context of quality. You need to be your own judge as well.
You may also learn through experience that the peer reviewers of your article may suggest additional inputs or updates that you the author may be required to perform to improve on the quality of your manuscript.
It is a good idea to read some publications by the same journal which you intend to submit your article to as it could give you some rough idea about your major concerns. Read also some details about the guidelines meant for the prospective authors which each journal publisher usually provide.