Most typically the n(C-O) around 1100 cm-1 of the ether linkages but also of the n(OH) of the hdyroxyl groups around 3400 cm-1. Take care, that also amounts of bound water (n(OH) around 3400 cm-1 and d(OH) at 1640 cm-1 may contribute to cellulose IR spectra.

Hi,

There is a nice paper with FTIR assigments for cellulose crystalline and amorphous one :

D. Ciolacu, F. Ciolacu, V.I. Popa, AMORPHOUS CELLULOSE – STRUCTURE AND CHARACTERIZATION, Cellul. Chem. Technol. 45 (2011) 13–21.

But concerning the band shape and exact band maxima postions it's best to measure your own standard of cellulose (for ATR or DR diffuse reflectance in many cases we observe some artifacts typical for the measuring technique like bands shifts or their intensity wavenumber dependency) so best way is to compare with results obtained in the same measuring technique).

Thanx Maja Mroczkowska and Martin Müller