# Is a wedge of spheres an $E_\infty$ ring spectrum?

The wedge sum $\bigvee_{k \in 2 \mathbb{Z}} S^{k}$ is an $A_\infty$-ring spectrum: the connective cover is the free $A_\infty$-ring on the sphere $S^2$, if I'm not mistaken, and then one inverts the element in $\pi_2$. It is also homotopy commutative. Can it be made into an $E_\infty$-ring spectrum? More generally, given an $E_\infty$-ring spectrum $R$, when can $\bigvee_{k \in 2 \mathbb{Z}} \Sigma^k E$ be made into an $E_\infty$-ring?

-
I'm pretty sure it is not, why would it be? Sorry not to have time to say more. Essentially, even with the $2$, the commutativity is only homotopical, not as strict as would be required. – Peter May Nov 2 '12 at 3:11
Here's a point of terminology that will help you find references: an $H_\infty$ structure on $\bigvee_{k\in d\mathbb{Z}}\Sigma^kE$ is essentially what is called an $H^d_\infty$ structure in the $H_\infty$ book (Springer LNM 1176). Of course $H_\infty$ is weaker than $E_\infty$ but this is a start. – Neil Strickland Nov 2 '12 at 7:07

Ah, tracked it down. Here is an argument. I should mention that Peter once pointed me towards an original source due to McClure, or page 238, Prop. 6.1, of SLN 1176 (the $H_\infty$ book).

Suppose you had such a ring object $R$. We examine its mod-2 homology. This has several features:

• It is the ring $\mathbb Z/2[t^{\pm 1}]$.

• It has trivial action of the Steenrod operations $P_r$ (this is dual to the cohomology action), because it's a wedge of spheres.

• From the $H_\infty$ book, it has Dyer-Lashof operations $Q^s$. These satisfy $Q^{|x|} x = x^2$, and the Nishida relations $$P_r Q^s = \sum \binom{s - r}{r - 2i} Q^{s-r+i}P_i.$$

In particular, these together would say $$0 = P_2 Q^4 t = \binom{2}{2} Q^{2}P_0 t + \binom{2}{0} Q^{4}P_2 t = Q^2 t = t^2.$$

If $E$ is a commutative $MU$-algebra, then you can use the map from $MU$ to its periodic version $MUP$ to produce $R \wedge_{MU}MUP$, which is 2-periodified. Unfortunately, we know very few genuine $MU$-algebras. Barry Walker proved that complex K-theory is one of them, based on Matthew Ando's work on studying $H_\infty$ structures on Lubin-Tate cohomology theories. My understanding is that the problem is still open for almost all of the Lubin-Tate cohomology theories.

-
Thank you! LNM 1176 is on my reading list, but I haven't gotten very far in it. These types of methods seem like quite a useful tool for answering such questions. – Akhil Mathew Nov 2 '12 at 11:32
Perhaps worth pointing out: the ring spectrum in question can be made E_2, since it can be described as the Thom spectrum of a 2-fold loop map Omega^2(BU(1)) -> Omega^2(BU) = Z x BU. – Jacob Lurie Nov 4 '12 at 16:03