Restricted enveloping algebras with minimal Lie derived length (Q630065)

The Lie derived series of a unital associative algebra \(R\) is defined as \(\delta^{[0]}(R)=R\) and \(\delta^{[n+1]}(R)=[\delta^{[n]}(R),\delta^{[n]}(R)]\), where \([A,A]\) is the span of the Lie commutators \([a,b]=ab-ba\), \(a,b\in A\). One defines also a series of ideals of \(R\) by \(\delta^{(0)}(R)=R\) and \(\delta^{(n+1)}(R)=[\delta^{(n)}(R),\delta^{(n)}(R)]R\). The algebra \(R\) is Lie solvable if \(\delta^{[n]}(R)=0\) and the minimal \(n\) with this property is the Lie derived length \(\text{dl}_{\text{Lie}}(R)\). If \(\delta^{(n)}(R)=0\), then \(R\) is strongly Lie solvable and the minimal \(n\) is the strong Lie derived length \(\text{dl}^{\text{Lie}}(R)\). In the paper under review the authors study the Lie solvability of the restricted enveloping algebra \(u(L)\) of a nonabelian restricted Lie algebra \(L\) over a field \(F\) of positive characteristic \(p\). It is known that for \(p > 2\) the Lie solvability of \(u(L)\) is equivalent to its strong Lie solvability, and this is not true for \(p = 2\). It is also possible to have \(\text{dl}_{\text{Lie}}(u(L)) < \text{dl}^{\text{Lie}}(u(L))\). The second named author established that if \(u(L)\) is Lie solvable then the Lie derived length of \(u(L)\) is at least \(\lceil \log_{2}(p + 1)\rceil\). In the present paper the authors characterize the restricted enveloping algebras whose Lie derived length coincides with this lower bound. The main result shows that the following conditions are equivalent: (1) \(\text{dl}_{\text{Lie}}(u(L))=\lceil \log_{2}(p + 1)\rceil\); (2) \(\text{dl}^{\text{Lie}}(u(L))=\lceil \log_{2}(p + 1)\rceil\); (3) one of the following conditions hold: (i) \(p=2\), \(\dim_F{L'}=2\), \({L'}\) is central, and \({L'}^{[p]}=0\); (ii) \(p=2\), \(\dim_F{L'}=1\), and \({L'}^{[p]}=0\); (iii) \(p>2\), \(\dim_F{L'}=1\), \({L'}\) is central, and \({L'}^{[p]}=0\). Most of the implications follow from previous results of the second and the third named authors. The main efforts are to show that when \(p>2\) the condition (1) implies the case (iii) of (3) and this is achieved in a series of technical lemmas. As a corollary the authors show that if \(p>2\), then \(\text{dl}_{\text{Lie}}(u(L))=\lceil \log_{2}(p + 1)\rceil\) if and only if the class of Lie nilpotency of \(u(L)\) is equal to \(p\).