Should I notify people I mention in the acknowledgements section?

shall I communicate their inclusion in the acknowledgements section?

Yes, why not? It helps at building connections and improving relationships. Besides, if the acknowledged people are from a different field than yours, they might never know you've acknowledged them otherwise.

what will be the correct etiquette for such an email?

I don't think there is any need to be overly formal. Typically I would write something simple, along the lines of (names and facts have been changed to protect the innocent):

Dear John,

thank you very much for helping me with the problem of packing holes. I've included an acknowledgment in a paper I recently wrote on the topic.

The paper is titled "Packing the unpackable and stacking the unstackable" and has been submitted to the Transactions on Painstaking Stacking. You can find a preprint of the paper at the address (link to, e.g., arXiv).

Kind regards/Sincerely/Cheers,

Massimo

The above example is meant to be sent just after the first submission. In more critical cases, when I'm not sure whether the acknowledgment would be well-received or not, or if I think I might have written a wrong detail (e.g., the affiliation), I typically send a copy of the paper before the submission, asking for feedback.

IMHO, definitely before first submission. It would be a nice idea to send them a "Thank You" email and let them know that you've finished the paper and in order to show your appreciation you've included their names in the acknowledgements section. So If they do not feel comfortable with this they could tell you even before the paper is submitted.

In math at least, you do not usually explicitly inform people that you have included them in the acknowledgements. You should of course thank them for their help, and it is common to send them a copy of the paper when you have a more-or-less final version (often this is the version you submit--in math the refereeing process takes a long time, plus your acknowledge-ees (sp?) may have additional comments/suggestions), or notify them when you post it on the arXiv. (This is regardless of whether they have seen a preliminary version or not.)

To add a little more on the question in the edit: sometimes I will send a pre-submission version if I really want to see if anyone has comments before I submit (students probably at least want to get comments from advisors before submitting), but usually I send out copies of the paper/post on the arXiv about the same time I submit. I don't typically actively notify acknowledge-ees when a paper has been accepted or published (though I will update the arXiv), unless the final version ends up being significantly different in a way that would be of interest to them. However, sometimes I see colleagues I have acknowledged, they will inquire about the status of the paper.