You can use any filter that blocks most blue and green light (all visible light other than red) and also passes a lot of near infrared light (somewhere between 700nm and 950nm). Wratten 25 is an old specification that meets these criteria very well. Rosco publishes spectral responses of its filters, and their #19 seems to be similar to the Wratten 25.
If the spectral response of a filter is not known, you can make some observations about its transmission, but without good spectrometer-like equipment you might not really know what your camera is capturing. That might be okay if your goal is just to highlight strong reflectance of NIR compared to visible light (which generally equates to high NDVI values). If you use your unknown filter and repeat some of the observations reported at this site and get similar NDVI results, your filter might be sufficiently similar to Wratten 25 for your purposes.