If you've acquired some duplicates in MongoDB that you want to get rid of, this post from Michael Francis provides a how-to on cleaning them up. The best option, obviously, is not to duplicate things in the first place - you're welcome - but Francis' post is focused on solving the problem after the fact, and he explains some helpful techniques.
The basic idea of Francis' strategy is to hash your documents to find duplicates and store them in a pair of arrays for easy disposal. He has some extra tips and shortcuts depending on how you're working with MongoDB - Node.js and Mongoose makes it easier - but the basics should translate pretty well from language to language.
Check out Francis' full post and see if it can help you clean up your data.