What are you using as shape factors (K) for the thin film and the powder?
What substrate is the thin film sitting on? Was it grown or deposited? Thin films generally tend to have a higher in-built stress, especially influenced by the growth method and the substrate.
While Deby-scherrer formula is relatively simple and straightforward to apply, it does not distinct between the main two components contributing to peak broadening (assuming you already counted for inherent XRD machine broadening):
1. broadening due to crystallite size (grain size)
2. broadening due to internal strain
In order to accurately separate those components, you better apply the Williamson-Hall plot method.
I have a short discussion on this topic in my PhD dissertation "Influence of Fast Neutron Irradiation on Mechanical Properties and Microstructure of
The sintering lead the sample to be more sharp peaks and the grain size of Si thin film become more size because the half-width more smaller by applying Dyby-Sherer equation. You must to repeat the experiment again to assure the result.
Probably, there is another explanation, but first, I would recommend remeasuring the XRD of your sintered film and the powder. In XRD, there is a certain beam spot where your sample is being measured. One possibility is that there just happened to be a few big crystals in the region you measured, which skewed the results of your analysis.