1. from — de Vaan
from 'the smoking' to 'the purifying'. The development *dhue- > fe- is generally accepted, although there are no completely certain instances. If correct, this preform would be another instance of an e-grade in a substantival ro-formation, cf. Vine 2002, However, I see no way to exclude root identity of februus with febris < PIE *dhegwh-rk That is, *febro- might have meant 'a burning, offering', whence — [de Vaan, s.v. from, p. 222]
2. from — de Vaan
from 'crude (meat)5 > 'crude' rests in the air. Bibl.: WH II: 447f., EM 579, IEW 868-871, Leumann 1977: 73, Risch 1979: 713, Sihier 1995: 158. —> raudus, ruber, rufus rOdo, -ere 'to roar, bellow' [v. Ill; pf. rudivi] (Varro+) Pit. *roudrl*rud-. PIE *h3r(e)ud-H- [pr.] 'to weep, roar5, IE cognates: Skt. rudanti [3p.act], rodisi [2s.act.] 'to weep, bewail', OAv. raosta [aor.], uruddiiata [pr.] 'laments'; OLith. — [de Vaan, s.v. from, p. 542]
3. from — de Vaan
from 'to distinguish, make a decision', -scire can belong to a root meaning 'to cut off, incise'. LIV derives -scio from *sekH- (as in secare) which, in view of its meaning, may well be the same root as the one given by LIV as *sk eh2(i)- 'to cut', which is more likely *skh2~. Thus, the ^-present of Skt. -chyati and Lat. -scid may both go back to the same PIE present. LIV (following Rix 1999) assumes that -scire … — [de Vaan, s.v. from, p. 559]
4. from — de Vaan
from 'to reach' > 'to reach after, seek' > 'swear'. He explains the stem suka- in U. and SPic. as denominal *soikqje/o- to a noun *soiko- 'declaration'. BibL: IEW 897f., Meiser 1986: 87f, Unteimann 2000: 589f, 712f., LIV *sejk-. sucus sucus 'juice* [m. ο] (Ρ1.+) Derivatives: sucidus 'fresh, juicy1 (Varro+), comucidus afresh, juicy' (PL). ι Pit. *souko-. PIE *soulc-o- 'juice' (or *seuk-o~?). IE cognates: W. sugno 'to … — [de Vaan, s.v. from, p. 609]