For explanations to the experimental fact that unstable acetylene and methy
lacetylene radical anions (A(-) and MA(-)) were observed in the trans-bent
forms, we carried out a theoretical study based on calculations and theoret
ical analysis. The calculations were performed using (DFT) B3LYP, MP2, QCIS
D, and MRSDCI methods in conjunction with large basis sets, including diffu
se functions on no centers (Bn), on carbons (Bn+), and on both carbons and
hydrogens (B++). The trans-bent A(-) was located in the optimization calcul
ations with all the basis sets used. It is characterized as a minimum-energ
y structure and is predicted to be more stable than the cis-bent form, and
the isotropic proton hyperfine coupling constant (a(H)) values calculated o
n the trans-bent A(-) are in agreement with experiment. All these results c
onfirm that A(-) exists in the trans-bent form.. The trans-bent form of MA(
-) was located only at some of the levels of the calculations with the Bn a
nd Bn+ basis sets. For the trans-bent MA(-), the B3LYP and QCISD calculatio
ns with the Bn basis sets predict a(H) values in agreement with those of th
e experiment. In the optimization calculations using the Bn++ basis sets, w
e got linear structures of MA(-) instead of the bent ones. Using Guerra's 6
-311++G(d,p) basis, which contains moderate diffuse functions, we located t
he trans-bent MA(-), but the a(H) results were not good. Theoretical analys
is invokes the Renner-Teller theory. The trans-bending potential energy cur
ves Of the Ag-2 and (2)Bg states of A(-) were calculated. The two curves co
nverge to the (2)IIg state of linear A(-) when the bending angle goes to 18
0 degrees, which describes the Renner-Teller splitting of the (2)IIg state
of linear A(-) along the trans-bending coordinate. There exists a local min
imum at a (trans) bending angle not equal 180 degrees along the curve of th
e lower state 2Ag, and the existence of the trans-bent A(-) is considered a
s a consequence of the Renner-Teller effect.