I'm not sure this question is research level question. Sorry in advance.


  1. $k$ is a commutative ring.
  2. $A$ is an augmented $k$-algebra.
  3. $A^e$ is defined as the $k$-algebra $A\otimes_{k}A^{op}$. It is naturally augmented $k$-algebra.


  1. $k$ (as left $A$-module) is quasi-isomorphic to a perfect complex. $k\in \mathbf{Perf}(A)$.
  2. $k$ (as left $A^e$-module) is quasi-isomorphic to a perfect complex. $k\in \mathbf{Perf}(A^e)$.
  3. $A$ (viewed as left $A^e$-module in a standard way) is quasi-isomorphic to a perfect complex. $A\in \mathbf{Perf}(A^e)$.

Question Let $\langle A\rangle$ be the thick subcategory of the category of perfect complexes $\mathbf{Perf}(A^e)$ generated by the left $A^e$-module $A$ (where $A$ is viewed as $A^e$-module in standard way). Is it clear that $k$ (viewed as $A^e$-module via the augmentation $A^e\rightarrow k$ ) is an object of $\langle A\rangle$ ?



Let $k$ be a field, and let $A$ be the algebra of upper triangular $2\times 2$ matrices over $k$, with augmentation map $\pmatrix{a&b\\0&c}\mapsto a$.

$A$ and $A^e$ have finite global dimension, so all modules have finite projective dimension, and are therefore quasi-isomorphic to perfect complexes.

Let $\mathcal{D}=\mathcal{D}(A^e)$ be the derived category of $A^e$-modules. The category $$k^\perp=\{X\in\mathcal{D}\mid \operatorname{Hom}_{\mathcal{D}}(k,X[t])=0\text{ for all $t\in\mathbb{Z}$}\}$$ is a thick subcategory of $\mathcal{D}$. I claim (proof below) that $A\in k^{\perp}$. So $\langle A\rangle\subseteq k^{\perp}$ for all $t\in\mathbb{Z}$. But clearly $k\not\in k^{\perp}$.

Proof of claim: Let $e_{11}=\pmatrix{1&0\\0&0}$ and $e_{22}=\pmatrix{0&0\\0&1}$, idempotent elements of $A$. Then $k=Ae_{11}$ is projective as a left $A$-module, and as a right $A$-module $k$ has a projective resolution $$0\to e_{22}A\to e_{11}A\to k\to0,$$ where the first nonzero map is $\pmatrix{0&0\\0&c}\mapsto\pmatrix{0&c\\0&0}$. So as an $A^e$-module, $k$ has a projective resolution $$0\to Ae_{11}\otimes_k e_{22}A\to Ae_{11}\otimes_k e_{11}A\to k\to0.$$ Applying the functor $\operatorname{Hom}_{A^e}(-,A)$ to the projective terms of this resolution gives the map $e_{11}Ae_{11}\to e_{11}Ae_{22}$ where $\pmatrix{a&0\\0&0}\mapsto\pmatrix{0&a\\0&0}$. The kernel and cokernel of this map are both zero, so $\operatorname{Ext}^t(k,A)=0$ for all $t\geq0$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.