Reconstructions

PS
*ḫVl- - kind of bird (SED II No. 107)
Not very reliable, scarcely attested.
PS
*ḫlḳ - to perish (CDG 261, Kogan 2015: 356)
PS
*ḫlš - to tear out (Kogan 2015: 558–559)
Contra SED I No. 33, there is no connection between this root and PS *ḫyl ‘to be in labor.’
PS
*ḫimˀ-at- - clarified butter (Fronzaroli 1971: 622, 634, 641; Sima 2000: 240; Kogan 2011: 240)
PS
*ḫaml-, *ḫVlVm- - kind of insect (SED II No. 109)
The reconstruction is not very reliable as it presupposes metathesis combined with a rather vague semantic correspondence.
PS
*ḫam(i)š- - five (Kogan 2009: 73)
PS
*ḫam/nṣ- - waist (SED I No. 132)
Scarcely attested (in Akk. and Arb.); phonetically and semantically highly problematic. Vocalism based on Akk. only, as the Arb. form represents a derived pattern.
PS
*ḫim/nṣ- - fatty tissue around intestins and stomach (SED I No. 133)
Highly problematic. Only Akk. if, as Kaufman asserts ([Kaufman 56]), the Arm. forms are Akkadisms (which is difficult either to prove or disprove). Cf. what may be related with metathesis in MSA: Mhr. ṣanḫ ‘cooking fat; fat put on fire to melt’ [ML 365], Hrs. ṣōnəḫ id. [ibid.], Jib. ṣonḫ ‘fat in an animal; fat put on to melt’ [JL 241].
PS
*ḫVnz(i/u)z- - kind of insect (SED II No. 112)
PS
*ḫV(n)zīr- - pig (SED II No. 111; Kogan 2011: 207)