A 5-Engel associative algebra whose group of units is not 5-Engel (Q1628496)

Let \(R\) be an associative ring, and denote by \([R]\) the Lie ring obtained by \(R\) by substituting the associative product \(ab\) in \(R\) with the Lie bracket \([a,b]=ab-ba\), \(a\), \(b\in R\). Let \(U(R)\) be the group of units of \(R\). It was proven by \textit{N. Gupta} and \textit{F. Levin} [J. Algebra 81, 225--231 (1983; Zbl 0514.16024)] that if the Lie ring \([R]\) is nilpotent of class \(\le c\) then the group \(U(R)\) is nilpotent of class \(\le c\). The paper under review solves a similar problem assuming that \([R]\) is \(n\)-Engel. Namely the authors construct an example of an associative algebra \(R\) over a field of characteristic different from 2 and 3, such that \([R]\) is 5-Engel but \(U(R)\) is not a 5-Engel group. The algebra they construct is generated by 1, and two nilpotent of class 2 elements. The authors deduce their construction on the following fact. If \(R\) is an associative ring by introducing the product \(a\circ b = a+b+ab\) one gets the so-called \text{adjoint multiplication} on \(R\), and \((R, \circ)\) is a monoid. The adjoint group \(R^\circ\) of \(R\) is the group of units of that monoid. For unitary rings \(R^\circ\cong U(R)\). The isomorphism is given by \(a\mapsto 1+a\). First the authors observe that in case \(R=R^\circ\) the Lie ring \([R]\) is nilpotent of class \(\le c\) or, respectively metabelian if and only if the group \(R^\circ\) is. Then they start with the free associative algebra on two generators \(x\) and \(y\) (without 1), and mod it out by the homogeneous elements of degree \(\ge 8\). In this way they guarantee that the obtained algebra \(B_1\) is nilpotent and thus \(B_1^\circ = B_1\). Furthermore they mod out by all monomials of degree \(> 2\) in \(x\), all monic monomials of degree 7 apart from \(yxy^3xy\) and \(y^2xyxy^2\), all monic monomials in degrees \(<7\) which do not divide the latter two monomials, and finally the polynomials \(2xy^3xy - 5yxyxy^2- 2yxy^3x + 5y^2xyxy\) and \(2yxy^3xy - 5y^2xyxy^2\). Then the obtained algebra \(B\) is nilpotent hence \(B^\circ = B\). The authors prove in their Theorem 1.2 that \([B]\) is 5-Engel but the corresponding adjoint group is not 5-Engel.