Perhaps your instrument is equipped with ATR attachment (it's usually the case with modern spectroscopes). This way you do not need to prepare a pellet, just put your fibres directly under a crystal. No more preparation required.

If you are not able to use the ATR attachment, I am pretty convinced that you could prepare KBr pellet with your natural fibres. I have done that with carbon fibres, by using mortar and pestle and mixing with KBr so it should be possible with your material as well.

I fully agree with Agnieszka.

In case you need to prepare KBr pellets from your fibres you should consider that in contrast to brittle carbon fibres powdering of other fibres may cause problems. In this case you can use a cyro or a vibrating mill. The grinding set of the latter can also be cooled down if necessary.

Just to add,its possible to grind the natural fibres using a mill to obtain powder and i used the same for banana fibres. see attached

Continuation of attachments showing grinding and sieving

1) what information you want to get? Only simple characterization, or identify functional groups?

If is YES, then you should do an analysis in the mid-infrared region, ie in the region between 400-4000 cm-1.

2) What is the shape of the sample: dust, fibers, leaves?

Forms powder, you can mix milligrams of the sample in KBr spectroscopic (150 mg) If you want to grind the fibers and sheets and prepare the sample equal to the powder can too.

After preparation of the well pad transparent, obtain the spectrum in said region, the way you see fit, ie, transmittance, absorbance or reflectance.

2) If in his laboratory have the ATR or DRIFTS accessories you can get the spectra nondestructively samples (leaves and fibers) directly, ie, without having to grind, but in this case you would have to obtain spectra in order REFLECTANCE.