I am not sure what you mean by "different type of tree". I am attaching an alignment of Escherichia coli 16S rRNA plus Shigella, and some "outgroup" enterobacteria such as Salmonella Citrobacter and Klebsiella as an illustration.

This alignment and tree illustrates that for closely related bacteria such as Escherichia and Shigella, the 16S ribosomal RNA alignment is trivially easy, but the data is only good enough to classify the bacteria in the general ballpark of species, and not good enough to identify true lineages of Shigella or Escherichia.

I am also attaching an alignment of Clostridum botulinum (and related Clostridia bacteria) 16S ribosomal RNA genes. Again the alignment is trivially simple and the tree shows that 16S data is good enough for placing each isolate in the general ballpark of which species it belongs to.

For bacterial phylogeny, use other genes apart from 16S gene study. The use of 16S rRNA gene sequences for determining phylogenetic relationships has limitations. Ribosomal genes can undergo lateral gene transfer across diverse taxonomic groups or be recombined, which could lead to artificial evolutionary data. Consequently, resolving bacterial phylogeny based solely on 16S rRNA gene sequences may not be sufficient (Lerat et al., 2003). Akhurst et al. (2004) were able to demonstrate that a phylogeny based on the gyrB gene provided a more robust tree for determining intrageneric relationships among Photorhabdus. This is due to the lower proportion of invariant nucleotide positions than in the 16S rRNA gene (Akhurst et al., 2004).

In my point of view the best method is maximum likelihhod as it uses more complex evolution model, and much better than NJ. But, if you want to publish in IJSEM, neighbor joining will also work. For alignment clustalW is fine.

There is no best in phylogeny. One has to choose what model fits better to the system under study. You can not use any model which is good for one organism system to any other organism as it is. ML, MP, NJ etc. are all valid and frequently used methods for the phylogenetic reconstruction of the data. And aligners are just a tool, they are equally good. Its the users choice what she/he want to use for the alignment. And with 16S you can easily resolve the phylogeny upto genus level, for species level it could be possible, given your bug does not belong to phyla with complex lineage below genus level.

Maximum likelihhod method