Post-Hopf group algebras, Hopf group braces and Rota-Baxter operators on Hopf group algebras

Yan Ning, Xing Wang and Daowei Lu

arXiv: 2507.20176v2 · PDF · 1 August 2025
Categories: math.RA, math-ph, math.MP
Licence: CC BY 4.0 The site shows the text.

Abstract

In this paper, we introduce the notions of Hopf group braces, post-Hopf group algebras and Rota-Baxter Hopf group algebras as important generalizations of Hopf brace, post Hopf algebra and Rota-Baxter Hopf algebras respectively. We also discuss their relationships. Explicitly under the condition of cocomutativity, Hopf group braces, post-Hopf group algebras could be mutually obtained, and Rota-Baxter Hopf group algebras could lead to Hopf group braces.

Items 43

Definition

Definition 2.1

DDFG

A Hopf \pi-brace structure over a family of coalgebras H=\{H_\alpha,\Delta_\alpha,\varepsilon_\alpha\}_{\alpha\in\pi} consisting of the following datum:

a Hopf \pi-algebra structure (H, \cdot, 1, S) ((H,\cdot,S) or H for short),
a Hopf \pi-algebra structure (H, \circ, 1_{\circ}, T) ((H,\circ,T) or H_{\circ} for short),
the compatible condition
g\circ(h\ell)=(g_{(1,\alpha)}\circ h)S_{\alpha}(g_{(2,\alpha)})(g_{(3,\alpha)}\circ \ell),
for all g\in H_\alpha,h\in H_\beta,\ell\in H_\gamma, ~\alpha,\beta,\gamma\in\pi.

Used by Proof, Proof, Proof

Remark

Remark 2.2

FF3I

When the group \pi=\{1\}, we could recover the notion of Hopf braces.
In any Hopf \pi-brace, 1_\circ=1. Indeed, setting g=h=1_\circ in (?) one obtains 1_\circ \ell=\ell for all \ell\in H_\gamma. Similarly, g=\ell=1_\circ yields h1_\circ=h for all h\in H_\beta.
(H,\cdot,\circ) is called cocommutative if H_\alpha is cocommutative for each \alpha\in\pi.

Example

Example 2.3

LVRS

Let (H,\cdot,\circ,S,T) be a Hopf brace and \pi=Aut(H) the group of automorphisms of Hopf brace H. For each \alpha\in\pi, H_\alpha=H as a vector space. We denote the element in H_\alpha by h^\alpha=\alpha(h) for h\in H. Define

\begin{aligned} &\cdot_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_{\alpha\beta},\ g^\alpha\otimes h^\beta\mapsto (g \cdot h)^{\alpha\beta},\\ &\circ_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_{\alpha\beta},\ g^\alpha\otimes h^\beta\mapsto (g \circ h)^{\alpha\beta},\\ &1_\cdot=1_\circ=1,\quad \Delta_\alpha=\Delta,\quad \varepsilon_\alpha=\varepsilon,\\ &S_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}},\ h^\alpha\mapsto S(h)^{\bar{\alpha}},\\ &T_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}},\ h^\alpha\mapsto T(h)^{\bar{\alpha}}. \end{aligned}

Denote \cdot=\{\cdot_{\alpha,\beta}\}_{\alpha,\beta\in\pi},\circ=\{\circ_{\alpha,\beta}\}_{\alpha,\beta\in\pi},S=\{S_\alpha\}_{\alpha\in\pi},T=\{T_\alpha\}_{\alpha\in\pi}), then (H,\cdot,S,\circ,T is a Hopf \pi-brace.

Definition

Definition 2.4

1KP8

Let H=\{H_\alpha\}_{\alpha\in\pi} be a Hopf \pi-algebra, and K a Hopf algebra. Then H is called a left K-\pi-module with the action \rightharpoondown=\{\rightharpoondown_\alpha: K\otimes H_\alpha \rightarrow H_\alpha\}_{\alpha\in\pi} if every H_\alpha is a H-module. And H is called a left K-\pi-module \pi-bialgebra if the following conditions hold:

k\rightharpoondown_{\alpha\beta}(hg)=(k_{(1)}\rightharpoondown_{\alpha}h)(k_{(2)}\rightharpoondown_{\beta}h),
k\rightharpoondown_{e}1_e=\varepsilon_K(k)1_e,
(k\rightharpoondown_{\alpha}h)_{(1,\alpha)} \otimes (k\rightharpoondown_{\alpha}h)_{(2,\alpha)} = (k_{(1)}\rightharpoondown_{\alpha}h_{(1,\alpha)})\otimes(k_{(2)}\rightharpoondown_{\alpha}h_{(2,\alpha)}),
\varepsilon_\alpha(k\rightharpoondown_\alpha h)=\varepsilon_K(k)\varepsilon_\alpha(h).

for all h\in H_\alpha, g\in H_\beta, \alpha,\beta\in\pi, k\in K.

Example

Example 2.5

3J7A

Let H=\{H_\alpha\}_{\alpha\in\pi} be a cocommutative Hopf \pi-algebra, K a cocommutative Hopf algebra, and H a left K-module \pi-bialgebra as above. Then the smash product (H\sharp K=\{H_\alpha\sharp K\}_{\alpha\in\pi},\cdot,\circ) is a cocommutative Hopf \pi-brace with the following structures:

\begin{aligned} (h\sharp k)\cdot_{\alpha\beta}(h'\sharp k') &= hh'\sharp kk', \\ (h\sharp k)\circ_{\alpha\beta}(h'\sharp k') &= h(k_{(1)}\rightharpoondown_{\beta}h')\sharp k_{(2)}k', \\ S_{H_\alpha\sharp K}(h\sharp k) &= S_\alpha(h)\sharp S_K(k), \\ T_{H_\alpha\sharp K}(h\sharp k) &= S_K(k_{(1)})\rightharpoondown_{\bar{\alpha}}S_\alpha(h)\sharp S_K(k_{(2)}), \end{aligned}

for all h\in H_\alpha, h'\in H_\beta, \alpha,\beta\in\pi, k,k'\in K.

Definition

Definition 2.6

J0WA

[Wang] Let K=\{K_\alpha\}_{\alpha\in\pi} a Hopf \pi_1-algebra. Then a vector space M is called a left \pi-K-modulelike object with the action \rightharpoonup=\{\rightharpoonup_\alpha: K_\alpha \otimes M \rightarrow M\}_{\alpha\in\pi_1} if

(k\ell)\rightharpoonup_{\alpha\beta} m = k\rightharpoonup_\alpha(\ell\rightharpoonup_\beta m)\ \hbox{and}\ 1_e \rightharpoonup_e m =m,

for all m\in M, k\in K_\alpha,\ell\in K_\beta, \alpha,\beta\in\pi_1.

Now let H be a Hopf \pi_2-algebra for group \pi_2 and each H_\gamma is a left \pi-K-modulelike object, then H is called a left \pi-K-modulelike bialgebra if the following conditions hold:

k\rightharpoonup_{\alpha}(hg)=(k_{(1,\alpha)}\rightharpoonup_\alpha h)(k_{(2,\alpha)}\rightharpoonup_\alpha g),
k\rightharpoonup_{\alpha}1_H=\varepsilon_\alpha(k)1_H,
(k\rightharpoonup_\alpha h)_{(1,\gamma)} \otimes (k\rightharpoonup_\alpha h)_{(2,\gamma)} = (k_{(1,\alpha)}\rightharpoonup_\alpha h_{(1,\gamma)})\otimes(k_{(2,\alpha)}\rightharpoonup_\alpha h_{(2,\gamma)}),
\varepsilon(k\rightharpoonup_\alpha h)=\varepsilon_{K,\alpha}(k)\varepsilon_{H,\gamma}(h),

for all h\in H_\gamma,g\in H_\delta, k\in K_\alpha.

Example

Example 2.7

9EAG

Let K be a cocommutative Hopf \pi_1-algebra and H a cocommutative left \pi-K-modulelike Hopf \pi_2-algebra. Denote \pi=\pi_1\bigoplus\pi_2, then the smash product (H\natural K=\{H_\gamma\sharp K_\alpha\}_{(\gamma,\alpha)\in\pi},\cdot,\circ) is a cocommutative Hopf \pi-brace with the following structures:

\begin{aligned} (h\natural k)\cdot(h'\natural k') &= hh'\natural kk', \\ (h\natural k)\circ(h'\natural k') &= h(k_{(1,\alpha)}\rightharpoonup_{\alpha}h')\natural k_{(2,\alpha)}k', \\ S_{(\gamma,\alpha)}(h\natural k) &= S_\gamma(h)\natural S_\alpha(k), \\ T_{(\gamma,\alpha)}(h\natural k) &= S_\alpha(k_{(1,\alpha)})\rightharpoonup_{\bar{\alpha}}S_\gamma(h)\natural S_\alpha(k_{(2,\alpha)}), \end{aligned}

for all h\in H_\gamma,h'\in H_\delta, k\in K_\alpha,k'\in K_\beta.

Used by Corollary 3.5

Lemma

Lemma 2.8

K37L

Let (H,\cdot,\circ) be a Hopf \pi-brace. Then

S_{\alpha\beta}(g_{(1,\alpha)}\circ h)g_{(2,\alpha)} = S_{\alpha}(g_{(1,\alpha)})(g_{(2,\alpha)}\circ S_{\beta}(h))

for all g\in H_\alpha, h\in H_\beta,~ \alpha,\beta\in\pi.

Used by Proof

Proof of Lemma 2.8

Equation (?) implies that

\varepsilon_\beta(h)g = g\circ (h_{(1,\beta)}S_{\beta}(h_{(2,\beta)})) = (g_{(1,\alpha)}\circ h_{(1,\beta)})S_{\alpha}(g_{(2,\alpha)})(g_{(3,\alpha)}\circ S_{\beta}(h_{(2,\beta)}))

holds for all g\in H_\alpha, h\in H_\beta,~ \alpha,\beta\in\pi, and hence,

\begin{aligned} S_{\alpha\beta}(g_{(1,\alpha)}\circ h)g_{(2,\alpha)} &= S_{\alpha\beta}(g_{(1,\alpha)}\circ h_{(1,\beta)}\varepsilon_\beta(h_{(2,\beta)}))g_{(2,\alpha)} \\ &= S_{\alpha\beta}(g_{(1,\alpha)}\circ h_{(1,\beta)})\varepsilon_\beta(h_{(2,\beta)})g_{(2,\alpha)}\\ &=S_{\alpha\beta}(g_{(1,\alpha)}\circ h_{(1,\beta)})(g_{(2,\alpha)}\circ h_{(2,\beta)})S_\alpha(g_{(3,\alpha)})(g_{(4,\alpha)}\circ S_\beta(h_{(3,\beta)}))\\ &=\varepsilon_{\alpha\beta}(g_{(1,\alpha)}\circ h_{(1,\beta)})S_\alpha(g_{(2,\alpha)})(g_{(3,\alpha)}\circ S_\beta(h_{(2,\beta)}))\\ &=S_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ S_\beta(h)). \end{aligned}

This completes the proof.

Uses Definition 2.1

Proposition

Proposition 2.9

W8AH

Let (H,\cdot,\circ) be a Hopf \pi-brace, H denote the Hopf \pi-algebra (H, \cdot, S) and H_{\circ} denote the Hopf \pi-algebra (H,\circ,T).

For all g\in H_\alpha, h\in H_\beta, H is a left H_{\circ}-\pi-module like \pi-algebra with
g\rightharpoonup_\alpha h=S_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ h).
For all g\in H_\alpha, h\in H_\beta,
\begin{aligned} &g\circ h=g_{(1,\alpha)}(g_{(2,\alpha)}\rightharpoonup_\alpha h),& \\ &gh=g_{(1,\alpha)}\circ(T_\alpha(g_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}} h).& \end{aligned}
If H is cocommutative, then H is a left H_{\circ}-\pi-module like \pi-bialgebra and
S_\beta(g\rightharpoonup_\alpha h)=g\rightharpoonup_{\alpha} S_\beta(h), ~g\in H_\alpha, h\in H_\beta.

Proof of Proposition 2.9

(1)For g\in H_\alpha, h\in H_\beta,\ell\in H_\gamma, it is clear that, 1\rightharpoonup_\alpha h=S_e(1)(1\circ h)=h and

g\rightharpoonup_\alpha 1=S_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ 1)=S_\alpha(g_{(1,\alpha)})g_{(2,\alpha)}=\varepsilon_\alpha(g)1.

And

\begin{aligned} & & g \rightharpoonup_\alpha (h\rightharpoonup_\beta\ell )\\ &= S_\alpha(g_{(1,\alpha)}) (g_{(2,\alpha)}\circ (S_\beta(h_{(1,\beta)})(h_{(2,\beta)}\circ \ell ))) \\ &\overset{(\ref{eq:Hbrace})}{=}& S_\alpha(g_{(1,\alpha)}) (g_{(2,\alpha)}\circ S_\beta(h_{(1,\beta)})) S_\alpha(g_{(3,\alpha)}) (g_{(4,\alpha)}\circ h_{(2,\beta)}\circ \ell ) \\ &\overset{(\ref{eq:Hbrace2})}{=}& S_{\alpha\beta}(g_{(1,\alpha)}\circ h_{(1,\beta)}) g_{(2,\alpha)} S_\alpha(g_{(3,\alpha)}) (g_{(4,\alpha)}\circ h_{(2,\beta)}\circ \ell ) \\ &= S_{\alpha}(g_{(1,\alpha)}\circ h_{(1,\beta)}) ((g_{(2,\alpha)}\circ h_{(2,\beta)})\circ \ell ) \\ &= (g\circ h)\rightharpoonup_{\alpha\beta} \ell , \end{aligned}

and

\begin{aligned} g\rightharpoonup_{\alpha}(hk) &= S_\alpha(g_{(1,\alpha)}) (g_{(2,\alpha)}\circ(hk)) \\ &= S_\alpha(g_{(1,\alpha)}) (g_{(2,\alpha)}\circ h)S_\alpha(g_{(3,\alpha)}) (g_{(4,\alpha)}\circ k)\\ &= (g_{(1,\alpha)}\rightharpoonup_\alpha h)(g_{(2,\alpha)}\rightharpoonup_\alpha k). \end{aligned}

Therefore, H is a left H_{\circ}-\pi-module \pi-algebra.

(2) For all g\in H_\alpha, h\in H_\beta,

\begin{aligned} g_{(1,\alpha)}(g_{(2,\alpha)}\rightharpoonup_\alpha h) &= g_{(1,\alpha)}S_\alpha(g_{(2,\alpha)})(g_{(3,\alpha)}\circ h) \\ &= \varepsilon_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ h) = g\circ h, \end{aligned}

and

\begin{aligned} g_{(1,\alpha)}\circ(T_\alpha(g_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}} h) &= g_{(1,\alpha)}(g_{(2,\alpha)}\rightharpoonup_\alpha(T_\alpha(g_{(2,\alpha)})\rightharpoonup_{\alpha} h)) \\ &= g_{(1,\alpha)}((g_{(2,\alpha)}\circ T_\alpha(g_{(2,\alpha)}))\rightharpoonup_{e} h) \\ &= g_{(1,\alpha)}\varepsilon_\alpha(g_{(2,\alpha)})(1\rightharpoonup_{e} h) = gh. \end{aligned}

(3) If H is cocommutative, it is straightforward to verify that H is a left H_{\circ}-\pi-module like \pi-bialgebra. For all g\in H_\alpha, h\in H_\beta, we have

\begin{aligned} S_\beta(g\rightharpoonup_\alpha h) &= S_\beta(S_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ h)) \\ &= S_{\alpha\beta}(g_{(1,\alpha)}\circ h)S_{\bar{\alpha}}S_\alpha(g_{(2,\alpha)})\\ &= S_{\alpha\beta}(g_{(1,\alpha)}\circ h)g_{(2,\alpha)} \\ &= S_\alpha(g_{(1,\alpha)})(g_{(2,\alpha)}\circ S_\beta(h)) \\ &= g\rightharpoonup_{\alpha} S_\beta(h). \end{aligned}

These finish the proof.

Uses Definition 2.1, Lemma 2.8

Proposition

Proposition 2.10

A4E9

Let \pi be an abelian group and (H,\cdot,\circ) a cocommutative Hopf \pi-brace, then each H_\alpha is a right H_\circ-\pi-module like coalgebra under the action

a\leftharpoonup_\beta x=T_\beta(a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ a_{(2,\alpha)}\circ x_{(2,\beta)},

for all a\in H_\alpha,x\in H_\beta.

Proof of Proposition 2.10

For all a\in H_\alpha,x\in H_\beta,y\in H_\gamma,

\begin{aligned} &(a\leftharpoonup_\beta x)\leftharpoonup_\gamma y\\ &=(T_\beta(a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ a_{(2,\alpha)}\circ x_{(2,\beta)})\leftharpoonup_\gamma y\\ &=T_{\bar{\gamma}}((T_\beta(a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ a_{(2,\alpha)}\circ x_{(2,\beta)})\rightharpoonup y_{(1,\gamma)})\circ T_\beta(a_{(3,\alpha)}\rightharpoonup_\alpha x_{(3,\beta)})\circ a_{(4,\alpha)}\circ x_{(4,\beta)}\circ y_{(2,\gamma)}\\ &=T_{\beta\gamma}((a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ(T_\beta(a_{(2,\alpha)}\rightharpoonup_\alpha x_{(2,\beta)})\rightharpoonup_{\alpha\beta} ((a_{(3,\alpha)}\circ x_{(3,\beta)})\rightharpoonup y_{(1,\gamma)})))\circ a_{(4,\alpha)}\circ x_{(4,\beta)}\circ y_{(2,\gamma)}\\ &=T_{\beta\gamma}((a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)}) ((a_{(2,\alpha)}\circ x_{(2,\beta)})\rightharpoonup_{\alpha\beta} y_{(1,\gamma)})))\circ a_{(3,\alpha)}\circ x_{(3,\beta)}\circ y_{(2,\gamma)}\\ &=T_{\beta\gamma}(S_\alpha(a_{(1,\alpha)})(a_{(2,\alpha)}\circ x_{(1,\beta)})S_{\alpha\beta}(a_{(3,\alpha)}\circ x_{(2,\beta)})(a_{(4,\alpha)}\circ x_{(3,\beta)}\circ y_{(1,\gamma)}))\circ a_{(5,\alpha)}\circ x_{(4,\beta)}\circ y_{(2,\gamma)}\\ &=T_{\beta\gamma}(S_\alpha(a_{(1,\alpha)})(a_{(2,\alpha)}\circ x_{(1,\beta)}\circ y_{(1,\gamma)}))\circ a_{(3,\alpha)}\circ x_{(2,\beta)}\circ y_{(2,\gamma)}\\ &=T_{\beta\gamma}(a_{(1,\alpha)}\rightharpoonup_\alpha(x_{(1,\beta)}\circ y_{(1,\gamma)}))\circ a_{(2,\alpha)}\circ x_{(2,\beta)}\circ y_{(2,\gamma)}\\ &=a\leftharpoonup_{\beta\gamma} (x\circ y), \end{aligned}

and

\begin{aligned} &\Delta_\alpha(a\leftharpoonup_\beta x)=\Delta_\alpha(T_\beta(a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ a_{(2,\alpha)}\circ x_{(2,\beta)})\\ &=T_\beta(a_{(1,\alpha)}\rightharpoonup_\alpha x_{(1,\beta)})\circ a_{(2,\alpha)}\circ x_{(2,\beta)}\otimes T_\beta(a_{(3,\alpha)}\rightharpoonup_\alpha x_{(3,\beta)})\circ a_{(4,\alpha)}\circ x_{(4,\beta)}\\ &=a_{(1,\alpha)}\leftharpoonup_\beta x_{(1,\beta)}\otimes a_{(2,\alpha)}\leftharpoonup_\beta x_{(2,\beta)}. \end{aligned}

The proof is completed.

Theorem

Theorem 2.11

L4ZS

Let \pi be an abelian group and (H,\cdot,\circ) a cocommutative Hopf \pi-brace. Define c_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_\beta\otimes H_\alpha by

c_{\alpha,\beta}(x\otimes y)=x_{(1,\alpha)}\rightharpoonup_\alpha y_{(1,\beta)}\otimes x_{(2,\alpha)}\leftharpoonup_\beta y_{(2,\beta)},\ \forall x\in H_\alpha,y\in H_\beta,

and \sigma_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_\beta\otimes H_\alpha by

\sigma_{\alpha,\beta}(x\otimes y)=y_{(1,\alpha)}\otimes S_\beta(y_{(2,\beta)})ay_{(3,\alpha)},

then c=\{c_{\alpha,\beta}\} and \sigma=\{\sigma_{\alpha,\beta}\} produce isomorphic representations of the braid group B_n on H_{\alpha_1}\otimes H_{\alpha_2}\otimes\cdots\otimes H_{\alpha_n} for n\geq2 and \alpha_i\in\pi.

Proof of Theorem 2.11

Define f_n,g_n:H_{\alpha_1}\otimes H_{\alpha_2}\otimes\cdots\otimes H_{\alpha_n}\rightarrow H_{\alpha_1}\otimes H_{\alpha_2}\otimes\cdots\otimes H_{\alpha_n} as follows

g_n(a^{\alpha_1}\otimes a^{\alpha_2}\otimes\cdots\otimes a^{\alpha_n})=a^{\alpha_1}_{(1,\alpha_1)}\otimes a^{\alpha_1}_{(2,\alpha_1)}\rightharpoonup_{\alpha_1}a^{\alpha_2}\otimes\cdots\otimes a^{\alpha_1}_{(n,\alpha_1)}\rightharpoonup_{\alpha_1}a^{\alpha_n},

f_2=f:H_\alpha\otimes H_\beta\rightarrow H_\alpha\otimes H_\beta, f(x\otimes y)=x_{(1,\alpha)}\otimes (x_{(2,\alpha)}\rightharpoonup_\alpha y) and recursively f_n=g_n(id\otimes f_{n-1}). Since f is invertible with f^{-1}(x\otimes y)=x_{(1,\alpha)}\otimes (T_\alpha(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}y) and the g_n are invertible, it follows that the g_n are invertible by induction.

Corollary

Corollary 2.12

6OBS

then c=\{c_{\alpha,\beta}\}_{\alpha,\beta\in\pi} is a coalgebra isomorphism and a solution of the braid equation.

Definition

Definition 2.13

L6QB

Let \pi_1 and \pi_2 be two groups. A matched pair of Hopf \pi-algebras is a 4-tuple (H, K,\leftharpoonup,\rightharpoonup), where K is Hopf \pi_1-algebra and H is Hopf \pi_2-algebra, \rightharpoonup_{\alpha}:H_\alpha\otimes K_\gamma\rightarrow K_\gamma and \leftharpoonup_{\gamma}:H_\alpha\otimes K_\gamma \rightarrow H_\alpha are linear maps such that K_\beta is a left \pi_2-H-modulelike coalgebra and H_\alpha is a right \pi_1-K-modulelike coalgebra and the following compatibility conditions hold:

\begin{aligned} x\rightharpoonup_{\gamma\delta}(ab)&=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})((x_{(2,\alpha)}\leftharpoonup_{\gamma}a_{(2,\gamma)})\rightharpoonup_\alpha b)),\\ x\rightharpoonup_{\alpha}1_K&=\varepsilon_\alpha(x)1_K,\\ (xy)\leftharpoonup_{\gamma}a&=(x\leftharpoonup_{\gamma}(y_{(1,\beta)}\rightharpoonup_{\beta}a_{(1,\gamma)}))(y_{(2,\beta)}\leftharpoonup_{\gamma}a_{(2,\gamma)}),\\ 1_H\leftharpoonup_e a&=\varepsilon_\gamma(a)1_H,\\ (x_{(1,\alpha)}\leftharpoonup_\gamma a_{(1,\gamma)})\otimes(x_{(2,\alpha)}\rightharpoonup_\alpha a_{(2,\gamma)})&=(x_{(2,\alpha)}\leftharpoonup_\gamma a_{(2,\gamma)})\otimes(x_{(1,\alpha)}\rightharpoonup_\alpha a_{(1,\gamma)}), \end{aligned}

for all x\in H_\alpha,y\in H_\beta,a\in K_\gamma,b\in K_\delta,\gamma,\delta\in\pi_1,\alpha,\beta\in\pi_2.

Theorem

Theorem 2.14

RPDJ

Let (H, K,\leftharpoonup,\rightharpoonup) be a matched pair of Hopf group algebra and \pi=\pi_1\bigoplus\pi_2, then K\otimes H=\{(K\otimes H)_{(\gamma,\alpha)}=K_\gamma\otimes H_\alpha\}_{(\gamma,\alpha)\in\pi} becomes a Hopf \pi-algebra with the following structures:

\begin{aligned} &(a\otimes x)(b\otimes y)=a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})\otimes(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y,\\ &(a\otimes x)_{(1,(\gamma,\alpha))}\otimes (a\otimes x)_{(2,(\gamma,\alpha))}=(a_{(1,\alpha)}\otimes x_{(1,\gamma)})\otimes (a_{(2,\alpha)}\otimes x_{(2,\gamma)}),\\ &\varepsilon_{(\gamma,\alpha)}(a\otimes x)=\varepsilon_{K,\gamma}(a) \varepsilon_{H,\alpha}(x),\\ &S_{(\gamma,\alpha)}(a\otimes x)=(S_{H,\alpha}(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}S_{K,\gamma}(a_{(2,\gamma)}))\otimes(S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)})), \end{aligned}

for all x\in H_\alpha,y\in H_\beta,a\in K_\gamma,b\in K_\delta.

Proof of Theorem 2.14

For all x\in H_\alpha,y\in H_\beta,z\in H_\mu,a\in K_\gamma,b\in K_\delta,c\in K_\nu,

\begin{aligned} &[(a\otimes x)(b\otimes y)](c\otimes z)\\ &=[a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})\otimes(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y](c\otimes z)\\ &=a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})[((x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y)_{(1,\alpha\beta)}\rightharpoonup_{\alpha\beta}c_{(1,\nu)}]\\ &\quad\otimes[((x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y)_{(2,\alpha\beta)}\leftharpoonup_{\nu}c_{(2,\nu)}]z\\ &=a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})[(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y_{(1,\beta)}\rightharpoonup_{\alpha\beta}c_{(1,\nu)}]\\ &\quad\otimes[(x_{(3,\alpha)}\leftharpoonup_\delta b_{(3,\delta)})y_{(2,\beta)}\leftharpoonup_{\nu}c_{(2,\nu)}]z\\ &=a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})[(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})\rightharpoonup_{\alpha}(y_{(1,\beta)}\rightharpoonup_{\beta}c_{(1,\nu)})]\\ &\quad\otimes[((x_{(3,\alpha)}\leftharpoonup_\beta b_{(3,\beta)})\leftharpoonup_\nu (y_{(2,\beta)}\rightharpoonup_{\beta}c_{(2,\nu)}))(y_{(3,\beta)}\leftharpoonup_{\nu}c_{(3,\nu)})]z\\ &=a(x_{(1,\alpha)}\rightharpoonup_{\alpha} b_{(1,\delta)}(y_{(1,\beta)}\rightharpoonup_{\beta}c_{(1,\nu)}))\\ &\quad\otimes[(x_{(2,\alpha)}\leftharpoonup_{\delta\nu} b_{(2,\delta)} (y_{(2,\beta)}\rightharpoonup_{\beta}c_{(2,\nu)}))(y_{(2,\beta)}\leftharpoonup_{\nu}c_{(2,\nu)})]z\\ &=(a\otimes x)(b(y_{(1,\beta)}\rightharpoonup_\beta c_{(1,\nu)})\otimes(y_{(2,\beta)}\leftharpoonup_\nu c_{(2,\nu)})z)\\ &=(a\otimes x)[(b\otimes y)(c\otimes z)], \end{aligned}

thus K\otimes H is associative.

\begin{aligned} &\Delta_{(\gamma\delta,\alpha\beta)}((a\otimes x)(b\otimes y))\\ &=\Delta_{(\gamma\delta,\alpha\beta)}(a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})\otimes(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y)\\ &=(a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)}))_{(1,\gamma\delta)}\otimes((x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y)_{(1,\alpha\beta)}\\ &\quad \otimes(a(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)}))_{(2,\gamma\delta)}\otimes((x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})y)_{(2,\alpha\beta)}\\ &=a_{(1,\gamma)}(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})_{(1,\delta)}\otimes(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})_{(1,\alpha)}y_{(1,\beta)}\\ &\quad \otimes a_{(2,\gamma)}(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})_{(2,\delta)}\otimes(x_{(2,\alpha)}\leftharpoonup_\delta b_{(2,\delta)})_{(2,\alpha)}y_{(2,\beta)}\\ &=a_{(1,\gamma)}(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})\otimes(x_{(3,\alpha)}\leftharpoonup_\beta b_{(3,\beta)})y_{(1,\beta)}\\ &\quad \otimes a_{(2,\gamma)}(x_{(2,\alpha)}\rightharpoonup_\alpha b_{(2,\delta)})\otimes(x_{(4,\alpha)}\leftharpoonup_\beta b_{(4,\beta)})y_{(2,\beta)}\\ &=a_{(1,\gamma)}(x_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\delta)})\otimes (x_{(2,\alpha)}\rightharpoonup_\alpha b_{(2,\delta)})y_{(1,\beta)}\\ &\quad \otimes a_{(2,\gamma)}(x_{(3,\alpha)}\leftharpoonup_\delta b_{(3,\delta)})\otimes(x_{(4,\alpha)}\leftharpoonup_\delta b_{(4,\delta)})y_{(2,\beta)}\\ &=\Delta_{(\gamma,\alpha)}(a\otimes x)\Delta_{(\delta,\beta)}(b\otimes y), \end{aligned}

and

\begin{aligned} &S_{(\gamma,\alpha)}(a_{(1,\gamma)}\otimes x_{(1,\alpha)})(a_{(2,\gamma)}\otimes x_{(2,\alpha)})\\ &=(S_{H,\alpha}(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}S_{K,\gamma}(a_{(2,\gamma)})\otimes S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)}))(a_{(3,\gamma)}\otimes x_{(3,\alpha)})\\ &=(S_{H,\alpha}(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}S_{K,\gamma}(a_{(2,\gamma)}))((S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)}))_{(1,\bar{\alpha})}\rightharpoonup_{\bar{\alpha}} a_{(3,\gamma)})\\ &\quad\otimes((S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)}))_{(2,\bar{\alpha})}\leftharpoonup_{\alpha} a_{(4,\gamma)})x_{(3,\alpha)}\\ &=(S_{H,\alpha}(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}S_{K,\gamma}(a_{(2,\gamma)}))((S_{H,\alpha}(x_{(1,\alpha)})_{(1,\bar{\alpha})}\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)})_{(1,\bar{\alpha})})\rightharpoonup_{\alpha} a_{(3,\gamma)})\\ &\quad\otimes((S_{H,\alpha}(x_{(1,\alpha)})_{(2,\bar{\alpha})}\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)})_{(2,\bar{\alpha})})\leftharpoonup_{\alpha} a_{(4,\gamma)})x_{(3,\alpha)}\\ &=(S_{H,\alpha}(x_{(3,\alpha)})\rightharpoonup_{\bar{\alpha}}S_{K,\gamma}(a_{(3,\gamma)}))((S_{H,\alpha}(x_{(2,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(2,\gamma)}))\rightharpoonup_{\alpha} a_{(4,\gamma)})\\ &\quad\otimes((S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)}))\leftharpoonup_{\alpha} a_{(5,\alpha)})x_{(4,\alpha)}\\ &=S_{H,\alpha}(x_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}(S_{K,\gamma}(a_{(3,\gamma)})a_{(4,\gamma)})\otimes((S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)}))\leftharpoonup_{\alpha} a_{(5,\alpha)})x_{(3,\alpha)}\\ &=((S_{H,\alpha}(x_{(1,\alpha)})\leftharpoonup_{\bar{\gamma}}S_{K,\gamma}(a_{(1,\gamma)})a_{(2,\gamma)})x_{(2,\alpha)}\\ &=\varepsilon_{K,\alpha}(a) \varepsilon_{H,\alpha}(x). \end{aligned}

Similarly (a_{(1,\gamma)}\otimes x_{(1,\alpha)})S_{\gamma,\alpha}(a_{(2,\gamma)}\otimes x_{(2,\alpha)})=\varepsilon_{K,\gamma}(a) \varepsilon_{H,\alpha}(x).

Proposition

Proposition 2.15

BC3M

Let \pi be an abelian group and (H,\cdot,\circ) be a cocommutative Hopf \pi-brace, then (H_\circ,H_\circ) is a matched pair of cocommutative Hopf \pi-braces with

x\rightharpoonup_\alpha a=S_\alpha(x_{(1,\alpha)})(x_{(2,\alpha)}\circ a),\ x\leftharpoonup_\beta a=T_\beta(x_{(1,\alpha)}\rightharpoonup_\beta a_{(1,\beta)})\circ x_{(2,\alpha)}\circ a_{(2,\beta)},

for all x\in H_\alpha,a\in H_\beta.

Used by Proof

Proof of Proposition 2.15

For all x\in H_\alpha,y\in H_\beta,a\in H_\gamma,b\in H_\delta,

\begin{aligned} &(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})\circ((x_{(2,\alpha)}\leftharpoonup_{\gamma}a_{(2,\gamma)})\rightharpoonup_\alpha b))\\ &=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})_{(1,\gamma)}((x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})_{(2,\gamma)}\rightharpoonup_\gamma((x_{(2,\alpha)}\leftharpoonup_{\gamma}a_{(2,\gamma)})\rightharpoonup_\alpha b))\\ &=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})(((x_{(2,\alpha)}\rightharpoonup_{\alpha}a_{(2,\gamma)})\circ(x_{(3,\alpha)}\leftharpoonup_{\alpha}a_{(3,\gamma)}))\rightharpoonup_{\gamma\alpha} b)\\ &=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})((x_{(2,\alpha)}\rightharpoonup_{\gamma}a_{(2,\gamma)})\circ T_\gamma(x_{(3,\alpha)}\rightharpoonup_\gamma a_{(3,\gamma)})\circ x_{(4,\alpha)}\circ a_{(4,\gamma)})\rightharpoonup_{\gamma\alpha} b)\\ &=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})( (x_{(2,\alpha)}\circ a_{(2,\gamma)})\rightharpoonup_{\alpha\gamma} b)\\ &=(x_{(1,\alpha)}\rightharpoonup_{\alpha}a_{(1,\gamma)})(x_{(2,\alpha)}\rightharpoonup_\alpha (a_{(2,\gamma)}\rightharpoonup_\gamma b))\\ &=x\rightharpoonup_{\alpha}( a_{(1,\gamma)}(a_{(2,\gamma)}\rightharpoonup_\gamma b))\\ &=x\rightharpoonup_{\alpha}(a\circ b), \end{aligned}

and

\begin{aligned} &(a\leftharpoonup_\gamma(b_{(1,\delta)}\rightharpoonup_\delta x_{(1,\alpha)}))\circ(b_{(2,\delta)}\leftharpoonup_\alpha x_{(2,\alpha)}))\\ &=T_\alpha(a_{(1,\gamma)}\circ b_{(1,\delta)}\rightharpoonup_{\gamma\delta} x_{(1,\alpha)})\circ a_{(2,\beta)}\circ(b_{(2,\delta)}\rightharpoonup_\delta x_{(2,\alpha)})\circ T_\alpha(b_{(3,\delta)}\rightharpoonup_\delta x_{(3,\alpha)})\circ b_{(4,\delta)}\circ x_{(4,\alpha)}\\ &=T_\alpha(a_{(1,\gamma)}\circ b_{(1,\delta)}\rightharpoonup_{\gamma\delta} x_{(1,\alpha)})\circ a_{(2,\beta)}\circ b_{(2,\delta)}\circ x_{(2,\alpha)}\\ &=(a\circ b)\leftharpoonup_\alpha x. \end{aligned}

Proposition

Proposition 2.16

S1E0

Let \pi be an abelian group and (H,\circ,\Delta,T) be a cocommutative Hopf \pi-algebra. Assume that (H,H,\rightharpoonup,\leftharpoonup) is a matched pair and that

a\circ b=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})\circ(a_{(2,\alpha)}\leftharpoonup_\beta b_{(2,\beta)}),

for all a\in H_\alpha,b\in H_\beta. Then (H,\cdot,\circ) is a Hopf \pi-brace with

ab=a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}b),\ S_\alpha(a)=a_{(1,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(2,\alpha)}).

Used by Proof

Proof of Proposition 2.16

For all a\in H_\alpha,b\in H_\beta,c\in H_\gamma,

\begin{aligned} \Delta_{\alpha\beta}(ab)&=\Delta_{\alpha\beta}(a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}b))\\ &=\Delta_\alpha(a)\circ\Delta_\beta(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}b)\\ &=\Delta_\alpha(a)\Delta_\beta(b), \end{aligned}

and

\begin{aligned} a_{(1,\alpha)}(a_{(2,\alpha)}\rightharpoonup_{\alpha}b)&=a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}(a_{(3,\alpha)}\rightharpoonup_{\alpha}b))\\ &=a_{(1,\alpha)}\circ((T_\alpha(a_{(2,\alpha)})\circ a_{(3,\alpha)})\rightharpoonup_{e}b))\\ &=a_{(1,\alpha)}\circ(1\circ b)\\ &=a\circ b. \end{aligned}

Since

\begin{aligned} &a\rightharpoonup_\alpha(bc)=a\rightharpoonup_\alpha(b_{(1,\beta)}\circ(T_\beta(b_{(2,\beta)})\rightharpoonup_{\bar{\beta}}c))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})\circ((a_{(2,\alpha)}\leftharpoonup_\beta b_{(2,\beta)})\rightharpoonup_\alpha(T_\beta(b_{(2,\beta)})\rightharpoonup_{\bar{\beta}}c))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})((a_{(2,\alpha)}\rightharpoonup_\alpha b_{(2,\beta)})\rightharpoonup_\beta((a_{(3,\alpha)}\leftharpoonup_\beta b_{(3,\beta)})\rightharpoonup_\alpha(T_\beta(b_{(4,\beta)})\rightharpoonup_{\bar{\beta}}c)))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})((a_{(2,\alpha)}\rightharpoonup_\alpha b_{(2,\beta)})\circ(a_{(3,\alpha)}\leftharpoonup_\beta b_{(3,\beta)})\rightharpoonup_{\alpha\beta}(T_\beta(b_{(4,\beta)})\rightharpoonup_{\bar{\beta}}c)))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})((a_{(2,\alpha)}\circ b_{(2,\beta)})\rightharpoonup_{\alpha\beta}(T_\beta(b_{(4,\beta)})\rightharpoonup_{\bar{\beta}}c))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b_{(1,\beta)})(a_{(2,\alpha)} \rightharpoonup_{\alpha}(b_{(2,\beta)}T_\beta(b_{(4,\beta)})\rightharpoonup_{e}c))\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha b)(a_{(2,\alpha)}\rightharpoonup_\alpha c). \end{aligned}

we have

\begin{aligned} a(bc)&=a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}(bc))\\ &=a_{(1,\alpha)}\circ((T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}b_{(1,\beta)})(T_\alpha(a_{(3,\alpha)})\rightharpoonup_{\bar{\alpha}}c))\\ &=a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}b_{(1,\beta)})\circ (T_\beta(T_\alpha(a_{(3,\alpha)})\rightharpoonup_{\bar{\alpha}}b_{(2,\beta)})\rightharpoonup_{\bar{\beta}}(T_\alpha(a_{(4,\alpha)})\rightharpoonup_{\bar{\alpha}}c))\\ &=(a_{(1,\alpha)}b_{(1,\beta)})\circ(T_\beta(T_\alpha(a_{(3,\alpha)})\rightharpoonup_{\bar{\alpha}}b_{(2,\beta)})\circ T_\alpha(a_{(4,\alpha)})\rightharpoonup_{\bar{\alpha}\bar{\beta}}c)\\ &=(a_{(1,\alpha)}b_{(1,\beta)})\circ(T_{\alpha\beta}(a_{(4,\alpha)}\circ (T_\alpha(a_{(3,\alpha)})\rightharpoonup_{\bar{\alpha}}b_{(2,\beta)})) \rightharpoonup_{\bar{\alpha}\bar{\beta}}c)\\ &=(a_{(1,\alpha)}b_{(1,\beta)})\circ(T_{\alpha\beta}(a_{(3,\alpha)}b_{(2,\beta)}) \rightharpoonup_{\bar{\alpha}\bar{\beta}}c)\\ &=(ab)c. \end{aligned}

For the antipode S,

\begin{aligned} a_{(1,\alpha)}S_\alpha(a_{(2,\alpha)})&=a_{(1,\alpha)}(a_{(2,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(3,\alpha)}))\\ &=a_{(1,\alpha)}\circ(T_\alpha(a_{(2,\alpha)})\rightharpoonup_{\bar{\alpha}}(a_{(3,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(4,\alpha)})))\\ &=a_{(1,\alpha)}\circ((T_\alpha(a_{(2,\alpha)})\circ a_{(3,\alpha)})\rightharpoonup_\eta T_\alpha(a_{(4,\alpha)}))\\ &=a_{(1,\alpha)}\circ T_\alpha(a_{(2,\alpha)})\\ &=\varepsilon_\alpha(a)1, \end{aligned}

and using (2.16) we have

T_\beta(a\rightharpoonup_\alpha b)=(a_{(1,\alpha)}\leftharpoonup_\beta b_{(1,\beta)})\circ T_{\alpha\beta}(a_{(2,\alpha)}\circ b_{(2,\beta)}).

Thus

\begin{aligned} &S_\alpha(a_{(1,\alpha)})a_{(2,\alpha)}\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(2,\alpha)}))\circ(T_{\bar{\alpha}}(a_{(3,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(4,\alpha)}))\rightharpoonup_\alpha a_{(5,\alpha)})\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(2,\alpha)}))\circ((a_{(3,\alpha)}\leftharpoonup_{\bar{\alpha}}T_\alpha(a_{(4,\alpha)}))\circ T_{e}(a_{(5,\alpha)}\circ T_\alpha(a_{(6,\alpha)}))\rightharpoonup_\alpha a_{(7,\alpha)})\\ &=(a_{(1,\alpha)}\rightharpoonup_\alpha T_\alpha(a_{(2,\alpha)}))\circ((a_{(3,\alpha)}\leftharpoonup_{\bar{\alpha}}T_\alpha(a_{(4,\alpha)}))\rightharpoonup_\alpha a_{(5,\alpha)})\\ &=a_{(1,\alpha)}\rightharpoonup_\alpha( T_\alpha(a_{(2,\alpha)})a_{(3,\alpha)})\\ &=\varepsilon_\alpha(a)1. \end{aligned}

For the compatibility condition, we have

\begin{aligned} &(a_{(1,\alpha)}\circ b)S(a_{(2,\alpha)})(a_{(3,\alpha)}\circ c)\\ &=(a_{(1,\alpha)}\circ b)S(a_{(2,\alpha)})a_{(3,\alpha)}(a_{(4,\alpha)}\rightharpoonup_\alpha c)\\ &=(a_{(1,\alpha)}\circ b)(a_{(2,\alpha)}\rightharpoonup_\alpha c)\\ &=(a_{(1,\alpha)}\circ b_{(1,\beta)})\circ(T_{\alpha\beta}(a_{(2,\alpha)}\circ b_{(2,\beta)})\rightharpoonup_{\bar{\alpha}\bar{\beta}}(a_{(3,\alpha)}\rightharpoonup_\alpha c))\\ &=(a_{(1,\alpha)}\circ b_{(1,\beta)})\circ((T_{\beta}(b_{(2,\beta)})\circ T_{\alpha}(a_{(2,\alpha)})\circ a_{(3,\alpha)})\rightharpoonup_{\bar{\beta}} c)\\ &=a\circ (b_{(1,\beta)}\circ(T_{\beta}(b_{(2,\beta)})\rightharpoonup_{\bar{\beta}} c))\\ &=a\circ (b_{(1,\beta)}\circ(bc). \end{aligned}

The proof is completed.

Theorem

Theorem 2.17

W175

Let \pi be an abelian group and (H,\cdot,1,\Delta,\varepsilon,S) a cocommutative Hopf \pi-algebra. The category \mathbf{Br}(H) and \mathbf{M_p}(H) are equivalent.

Proof of Theorem 2.17

Define F:\mathbf{Br}(H)\rightarrow \mathbf{M_p}(H) by F((H,\cdot,\circ))=(H_\circ,H_\circ), where the matched pair (H_\circ,H_\circ) is given in Proposition 2.15. For any morphism in \mathbf{Br}(H), F(f)=f. Clearly F is a functor.

Conversely define G:\mathbf{M_p}(H)\rightarrow\mathbf{Br}(H) by F(H,H)=(H,\cdot,\circ), where (H,\cdot,\circ) is the Hopf brace given in Proposition 2.16, and G(f)=f for any morphism in \mathbf{M_p}(H). It is easy to see that G is a functor. By an routine exercise, \mathbf{Br}(H) and \mathbf{M_p}(H) are equivalent.

Uses Proposition 2.15, Proposition 2.16

Definition

Definition 3.1

AJVG

A post-Hopf \pi-algebra is a pair (H,\triangleright), where H=(\{H_\alpha\}_{\alpha\in\pi},\cdot,1,\Delta,\varepsilon,S) is a Hopf \pi-algebra and \triangleright=\{\triangleright_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_\beta\} is a family of coalgebra homomorphisms satisfying

\begin{aligned} &x\triangleright_{\alpha,\beta\gamma}(yz)=(x_{(1,\alpha)}\triangleright_{\alpha,\beta}y)(x_{(2,\alpha)}\triangleright_{\alpha,\gamma}z),&\\ &x\triangleright_{\alpha,\gamma}(y\triangleright_{\beta,\gamma}z)=(x_{(1,\alpha)}(x_{(2,\alpha)}\triangleright_{\alpha,\beta}y))\triangleright_{\alpha\beta,\gamma}z,& \end{aligned}

for any x\in H_\alpha,y\in H_\beta,z\in H_\gamma, and the linear map \theta_{\alpha,\beta}:H_\alpha\rightarrow {\rm End}(H_\beta) defined by

\theta_{\alpha,\beta,x}(y)=x\triangleright_{\alpha,\beta}y

is convolution invertible in {\rm Hom}(H_\alpha,{\rm End}(H_\beta)), that is, there exists unique \psi_{\alpha,\beta}\in{\rm Hom}(H_\alpha,{\rm End}(H_\beta)) such that

\psi_{\alpha,\beta,x_{(1,\alpha)}}\theta_{\alpha,\beta,x_{(2,\alpha)}}=\theta_{\alpha,\beta,x_{(1,\alpha)}}\psi_{\alpha,\beta,x_{(2,{\alpha,\beta})}}=\varepsilon_\alpha(x)id_{H_\beta}.

Used by Proof, Proof

Remark

Remark 3.2

0BF8

(1) When \pi=\{1\}, a post-Hopf \pi-algebra could be reduced to a post-Hopf algebra introduced in [LST].

(2) A post-Hopf \pi-algebra (H,\triangleright) is called cocommutative if each H_\alpha is cocommutative.

Lemma

Lemma 3.3

RN4A

Let (H,\triangleright) be a post-Hopf \pi-algebra. Then for all x\in H_\alpha,y\in H_\beta, \alpha,\beta\in\pi, we have

\begin{aligned} x\triangleright_{\alpha,e} 1 &= \varepsilon_\alpha(x)1, \\ 1\triangleright_{e,\alpha} x &= x, \\ S_\beta(x\triangleright_{\alpha,\beta} y) &= x\triangleright_{\alpha,\bar{\beta}} S_\beta(y). \end{aligned}

Used by Proof

Proof of Lemma 3.3

Since \triangleright is a coalgebra homomorphism, we have

\begin{aligned} x\triangleright_{\alpha,e} 1 &= (x_{(1,\alpha)}\triangleright_{\alpha,e} 1) \varepsilon(x_{(2,\alpha)}\triangleright_{\alpha,e} 1)\\ &= (x_{(1,\alpha)}\triangleright_{\alpha,e} 1)\cdot (x_{(2,\alpha)}\triangleright_{\alpha,e} 1)\cdot S_e(x_{(3,\alpha)}\triangleright_{\alpha,e} 1)\\ &\stackrel{\ref{eq:P1}}{=}& (x_{(1,\alpha)}\triangleright_{\alpha,e} 1)\cdot S(x_{(2,\alpha)}\triangleright_{\alpha,e} 1)\\ &= \varepsilon_e(x\triangleright_{\alpha,e} 1)1 =\varepsilon_e(x)1. \end{aligned}

By Eq. (3.1), we have \psi_{e,\alpha,1}\theta_{e,\alpha,1}=\theta_{e,\alpha,1}\psi_{e,\alpha,1}=id_{H_\alpha}, which means that \theta_{e,\alpha,1} is a linear automorphism of H_\alpha. On the other hand, we have

\begin{aligned} \theta_{e,\alpha,1}^2(x) = 1\triangleright_{e,\alpha} (1\triangleright_{e,\alpha} x) \stackrel{\eqref{eq:P2}}{=} (1\triangleright_{e,\alpha} 1)\triangleright_{e,\alpha} (x) \stackrel{\eqref{eq:P3}}{=} 1\triangleright_{e,\alpha} x = \theta_{e,\alpha,1}(x). \end{aligned}

Hence, 1\triangleright_{e,\alpha} (x) = \theta_{e,\alpha,1}(x) = x. Finally we have

\begin{aligned} S_\beta(x\triangleright_{\alpha,\beta} y)&=S_\beta(x_{(1,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\varepsilon_\alpha(x_{(2,\alpha)})\varepsilon_\beta(y_{(2,\beta)}) \\ &\stackrel{\ref{eq:P3}}{=}& S_\beta(x_{(1,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\cdot(x_{(2,\alpha)}\triangleright_{\alpha,\beta} \varepsilon_\beta(y_{(2,\beta)})1)\\ &=S_\beta(x_{(1,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\cdot(x_{(2,\alpha)}\triangleright_{\alpha,\beta} (y_{(2,\beta)}\cdot S_\beta(y_3))) \\ &\stackrel{\ref{eq:P1}}{=}& S_\beta(x_{(1,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\cdot(x_{(2,\alpha)}\triangleright_{\alpha,\beta} y_{(2,\beta)})\cdot(x_{(3,\alpha)}\triangleright_{\alpha,\bar{\alpha}} S_\beta(y_{(3,\beta)}))\\ &=\varepsilon_\beta(x_{(1,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})(x_{(2,\alpha)}\triangleright_{\alpha,\beta} S_\beta(y_{(2,\beta)}))\\ &=\varepsilon_\alpha(x_{(1,\alpha)})\varepsilon_\beta(y_{(1,\beta)})(x_{(2,\alpha)}\triangleright S_\beta(y_{(2,\beta)}))=x\triangleright S_\beta(y). \end{aligned}

These finish the proof.

Uses Definition 3.1

Theorem

Theorem 3.4

EWQS

Let (H,\triangleright) be a cocommutative post-Hopf \pi-algebra. Define

\begin{aligned} x *_{\alpha,\beta} y &:=& x_{(1,\alpha)}\cdot (x_{(2,\beta)}\triangleright_{\alpha,\beta} y),\\ T_\alpha(x) &:=& \psi_{\alpha,\bar{\alpha},x_{(1,\alpha)}}(S_\alpha(x_{(2,\alpha)})), \end{aligned}

for all x\in H_\alpha,y\in H_\beta, \alpha,\beta\in\pi. Then H_\triangleright:=(H,*,1,\Delta,\varepsilon,T) is a Hopf \pi-algebra, which is called the subadjacent Hopf \pi-algebra. Furthermore (H_e,\cdot,1,\Delta_e,\varepsilon_e,S_e) is a left H_\triangleright-modulelike \pi-bialgebra under the action \{\triangleright_{\alpha,e}\}_{\alpha\in\pi}.

Used by Corollary 3.5

Proof of Theorem 3.4

Since \triangleright is a coalgebra homomorphism and H is cocommutative, we have

\begin{aligned} \Delta_{\alpha,\beta}(x *_{\alpha,\beta} y)&=\Delta_{\alpha,\beta}(x_{(1,\alpha)}\cdot (x_{(2,\beta)}\triangleright_{\alpha,\beta} y))\\ &=\Delta_\alpha(x_{(1,\alpha)})\cdot \Delta_{\alpha,\beta}(x_{(2,\alpha)}\triangleright_{\alpha,\beta} y)\\ &=(x_{(1,\alpha)}\otimes x_{(2,\alpha)})\cdot((x_{(3,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\otimes(x_{(4,\alpha)}\triangleright_{\alpha,\beta} y_{(2,\beta)}))\\ &=(x_{(1,\alpha)}\cdot (x_{(3,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)}))\otimes (x_{(2,\alpha)}\cdot(x_{(4,\alpha)}\triangleright_{\alpha,\beta} y_{(2,\beta)}))\\ &=(x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)}))\otimes (x_{(3,\alpha)}\cdot(x_{(4,\alpha)}\triangleright_{\alpha,\beta} y_{(2,\beta)}))\\ &=(x_{(1,\alpha)} *_{\alpha,\beta} y_{(1,\beta)})\otimes(x_{(2,\alpha)} *_{\alpha,\beta} y_{(2,\beta)}) \end{aligned}

for all x\in H_\alpha,y\in H_\beta, \alpha,\beta\in\pi, which implies that the comultiplication \Delta is an algebra homomorphism with respect to the multiplication *_{\alpha,\beta}. Moreover, we have

\begin{aligned} \varepsilon_\beta(x *_{\alpha,\beta} y) & =\varepsilon_\beta(x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} y)) \\ & =\varepsilon_\alpha(x_{(1,\alpha)})\varepsilon_\beta(x_{(2,\alpha)}\triangleright_{\alpha,\beta} y)=\varepsilon_\alpha(x)\varepsilon_\beta(y), \end{aligned}

which implies that the counit \varepsilon is also an algebra homomorphism with respect to the multiplication *_{\alpha,\beta}. Since the comultiplication \Delta is an algebra homomorphism with respect to the multiplication \cdot, for all x\in H_\alpha,y\in H_\beta,z\in H_\gamma, \alpha,\beta,\gamma\in\pi, we have

\begin{aligned} (x *_{\alpha,\beta} y)*_{\alpha\beta,\gamma} z&= (x_{(1,\alpha)} *_{\alpha,\beta} y_{(1,\beta)})\cdot((x_{(2,\alpha)} *_{\alpha,\beta} y_{(2,\beta)})\triangleright_{\alpha\beta,\gamma} z)\\ &=(x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)}))\cdot((x_{(3,\alpha)}\cdot (x_{(4,\alpha)}\triangleright_{\alpha,\beta} y_{(2,\beta)}))\triangleright_{\alpha\beta,\gamma} z)\\ &\stackrel{\eqref{eq:P2}}{=}&x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} y_{(1,\beta)})\cdot (x_{(3,\alpha)}\triangleright_{\alpha,\gamma} (y_{(2,\beta)}\triangleright_{\beta,\gamma} z))\\ &\stackrel{\eqref{eq:P1}}{=}&x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} (y_{(1,\beta)}\cdot (y_{(2,\beta)}\triangleright_{\beta,\gamma} z)))\\ &=x *_{\alpha,\beta\gamma} (y*_{\beta,\gamma} z), \end{aligned}

which implies that the multiplication *_{\alpha,\beta} is associative. For any x\in H_\alpha, \alpha\in\pi, by (3.3) and (3.3), we have

\begin{aligned} x *_{\alpha,e} 1&=x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} 1)=x_{(1,\alpha)}\cdot\varepsilon_\alpha(x_{(2,\alpha)})=x,\\ 1 *_{e,\alpha} x&=1\cdot (1\triangleright_{\alpha,\beta} x)=x. \end{aligned}

Since \triangleright is a coalgebra homomorphism and H is cocommutative, we know that

\Delta_\beta\psi_{\alpha,\beta,x}=(\psi_{\alpha,\beta,x_{(1,\alpha)}}\otimes\psi_{\alpha,\beta,x_{(2,\alpha)}})\Delta_\beta,

and T is a coalgebra homomorphism. Also, note that

\begin{aligned} x_{(1,\alpha)}*_{\alpha,\bar{\alpha}} T_\alpha(x_{(2,\alpha)}) &\stackrel{\ref{post-rbb-1}}{=}& x_{(1,\alpha)}\cdot(x_{(2,\alpha)}\triangleright_{\alpha,\bar{\alpha}} T_\alpha(x_{(2,\alpha)}))\\ &\stackrel{\ref{post-rbb-2}}{=}& x_{(1,\alpha)}\cdot(\theta_{\alpha,\bar{\alpha},x_{(2,\alpha)}}(\psi_{\alpha,\bar{\alpha},x_{(3,\alpha)}}(S_\alpha(x_{(4,\alpha)}))))\\ &= x_{(1,\alpha)}\cdot(\varepsilon_\alpha(x_{(2,\alpha)})S_\alpha(x_{(3,\alpha)}))\\ &= \varepsilon_\alpha(x)1, \end{aligned}

and this implies that

\begin{aligned} T_{\bar{\alpha}}T_\alpha(x) &= \varepsilon_\alpha(x_{(1,\alpha)})T_{\bar{\alpha}}T_\alpha(x_{(2,\alpha)}) \\ &= (x_{(1,\alpha)}*_{\alpha,\bar{\alpha}} T_{\alpha}(x_{(2,\alpha)}))*_{e,\alpha} T_{\bar{\alpha}}T_\alpha(x_{(3,\alpha)})\\ &= x_{(1,\alpha)}*_{\alpha,e} (T_{\alpha}(x_{(2,\alpha)})*_{\bar{\alpha},\alpha} T_{\bar{\alpha}}(T_\alpha(x_{(3,\alpha)})))\\ &= x_{(1,\alpha)}*_{\alpha,e} \varepsilon_{\bar{\alpha}}(T_\alpha(x_{(2,\alpha)}))1 = x, \end{aligned}

and

\begin{aligned} T_\alpha(x_{(1,\alpha)}) *_{\bar{\alpha},\alpha} x_{(2,\alpha)} &= T_\alpha(x_{(1,\alpha)}) *_{\bar{\alpha},\alpha} T_{\bar{\alpha}}T_\alpha(x_{(2,\alpha)})\\ &= \varepsilon_{\bar{\alpha}}(T_\alpha(x))1 = \varepsilon_\alpha(x)1. \end{aligned}

Therefore, (H,*,1,\Delta,\varepsilon,T) is a cocommutative Hopf \pi-algebra.

Moreover, we have

(x *_{\alpha,\beta} y)\triangleright_{\alpha\beta,\gamma} z=(x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta} y))\triangleright_{\alpha\beta,\gamma} z=x\triangleright_{\alpha,\gamma} (y\triangleright_{\beta,\gamma} z).

Then by (3.1) and (3.3), (H_e ,\cdot,1) is a left \pi-H_\triangleright-modulelike. Since \triangleright_{(\alpha,\beta)} is also a coalgebra homomorphism, (H_e,\cdot,1,\Delta_e,\varepsilon_e,S_e) is a left \pi-H_\triangleright-modulelike \pi-bialgebra via the action \triangleright_{\alpha,\beta}.

Uses Definition 3.1, Lemma 3.3

Corollary

Corollary 3.5

Z25U

Let (H,\triangleright) be a cocommutative post-Hopf \pi-algebra. We can construct a Hopf \pi-brace H_e\natural H_\triangleright=\{H_e\sharp H_\alpha\}_{\alpha\in\pi} via Theorem 3.4 and Example 2.7.

Uses Theorem 3.4, Example 2.7

Theorem

Theorem 3.6

IU0Z

Let (H,\triangleright) be a cocommutative post-Hopf \pi-algebra. Then (H,\cdot,1,\Delta,\varepsilon,S) and the subadjacent Hopf \pi-algebra (H,*,1,\Delta,\varepsilon,T) form a Hopf \pi-brace. Conversely, any cocommutative Hopf \pi-brace (H,\cdot,\circ) gives a post-Hopf \pi-algebra (H,\triangleright) with \triangleright defined by

x\triangleright_{\alpha,\beta} y=S_\alpha(x_{(1,\alpha)})\cdot_{\bar{\alpha},\beta}(x_{(2,\alpha)}\circ_{\alpha,\beta} y), \quad \forall x\in H_\alpha,y\in H_\beta, \alpha,\beta\in\pi.

Proof of Theorem 3.6

Let (H,\triangleright) be a cocommutative post-Hopf \pi-algebra. We only need to show that the multiplications \cdot and * satisfy the compatibility condition (?), which follows from

\begin{aligned} x*_{\alpha,\beta}(y\cdot z)&=x_{(1,\alpha)}\cdot (x_{(2,\alpha)}\triangleright_{\alpha,\beta\gamma}(y\cdot z))\\ &=x_{(1,\alpha)}\cdot((x_{(2,\alpha)}\triangleright_{\alpha,\beta} y)\cdot (x_{(3,\alpha)}\triangleright_{\alpha,\gamma} z))\\ &=x_{(1,\alpha)}\cdot(x_{(2,\alpha)}\triangleright_{\alpha,\beta} y)\cdot S_\alpha(x_{(3,\alpha)})\cdot x_{(4,\alpha)}\cdot (x_5\triangleright_{\alpha,\gamma} z)\\ &=(x_{(1,\alpha)}*_{\alpha,\beta} y)\cdot S_\alpha(x_{(2,\alpha)})\cdot (x_{(3,\alpha)}*_{\alpha,\beta} z), \end{aligned}

for any x\in H_\alpha,y\in H_\beta,z\in H_\gamma, \alpha,\beta,\gamma\in\pi.

Conversely, it is straightforward but tedious to check that a cocommutative Hopf \pi-brace (H,\cdot,\circ) induces a post-Hopf \pi-algebra (H,\triangleright).

Uses Definition 2.1

Definition

Definition 4.1

03VB

Let H=\{H_\alpha\}_{\alpha\in\pi} be a cocommutative Hopf \pi-algebra. The family of coalgebra homomorphisms B=\{B_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}}\}_{\alpha\in\pi} is called a Rota-Baxter operator on H if for all \alpha,\beta\in\pi, a\in H_\alpha,b\in H_\beta,

B_\alpha(a)B_\beta(b)=B_{\beta\alpha}\big(a_{(1,\alpha)}B_\alpha(a_{(2,\alpha)})bS_{\bar{\alpha}}(B_\alpha(a_{(3,\alpha)}))\big).

The pair (H,B) is called a Rota-Baxter Hopf \pi-algebra.

Used by Proof, Proof

Remark

Remark 4.2

5GSP

When the group \pi=\{1\}, we could recover the notion of Rota-Baxter Hopf algebras.
It is obvious that the antipode S is a Rota-Baxter operator on H.
Let (H,B) be a Rota-Baxter Hopf \pi-algebra, and \varphi a \pi-bialgebra automorphism or antiautomorphism of H. Then B^{(\varphi)}=\{B^{(\varphi)}_\alpha=\varphi_{\bar{\alpha}}\circ B_\alpha\circ\varphi^{-1}_\alpha\}_{\alpha\in\pi} is also a Rota-Baxter operator on H.

Example

Example 4.3

DY4X

Let (H,\Delta,\varepsilon,S,B) be a Rota-Baxter Hopf algebra and \pi=Aut(H) the group of Hopf algebras automorphisms of H. For each \alpha\in\pi, H_\alpha=H as a vector space. We denote the element in H_\alpha by h^\alpha=\alpha(h) for h\in H. Define

\begin{aligned} &m_{\alpha,\beta}:H_\alpha\otimes H_\beta\rightarrow H_{\alpha\beta},\ g^\alpha\otimes h^\beta\mapsto (gh)^{\alpha\beta},\\ &\Delta_\alpha=\Delta,\quad \varepsilon_\alpha=\varepsilon,\\ &S_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}},\ h^\alpha\mapsto S(h)^{\bar{\alpha}},\\ &B_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}},\ h^\alpha\mapsto B(h)^{\bar{\alpha}}, \end{aligned}

then (\{H_\alpha\}_{\alpha\in \pi},\{B_\alpha\}_{\alpha\in \pi}) is a Rota-Baxter Hopf \pi-algebra.

Theorem

Theorem 4.4

7KMA

Let H=\{H_\alpha\}_{\alpha\in\pi} be a cocommutative Hopf \pi-algebra. Suppose that G_\alpha is a Hopf subalgebra of H_e and K_\alpha is a subcoalgebra of H_\alpha for all \alpha\in\pi, such that G=\{G_\alpha\}_{\alpha\in\pi} is a Hopf \pi-algebra and K=\{K_\alpha\}_{\alpha\in\pi} is a Hopf \pi-subalgebra of H_\alpha, and as a Hopf \pi-algebra H=GK, i.e., H_\alpha=G_\alpha K_\alpha for all \alpha\in\pi. Suppose that the product is direct, that is, H is isomorphic to G\otimes K=\{G_\alpha \otimes K_\alpha\}_{\alpha\in\pi} as a family of vector spaces. Define a family of maps B=\{B_\alpha:H_\alpha\rightarrow H_{\bar{\alpha}}\}_{\alpha\in\pi} by

B_\alpha (h h')=\varepsilon_e(h)S_\alpha(h'),

where h_1\in G_\alpha, h_2\in K_\alpha. Then B is a Rota-Baxter operator of H.

Proof of Theorem 4.4

Clearly, B is a family of well-defined linear maps. First we prove that each B_\alpha is a coalgebra map. For x=hg\in H_\alpha, where h\in G_\alpha, g\in K_\alpha we have:

\begin{aligned} \Delta_{\bar{\alpha}}(B_\alpha(x))&=\varepsilon_e(h)\Delta_{\bar{\alpha}}(S_\alpha(g))=\varepsilon_e(h) S_{\bar{\alpha}}(g_{(2,\alpha)})\otimes S_{\bar{\alpha}}(g_{(1,\alpha)})\\ &=\varepsilon_e(h_{(1,e)})S_{\bar{\alpha}}(g_{(1,\alpha)})\otimes \varepsilon_e(h_{(2,e)})S_{\bar{\alpha}}(g_{(2,\alpha)})\\ &=(B_{\bar{\alpha}}\otimes B_{\bar{\alpha}})\Delta_\alpha(x). \end{aligned}

In order to prove that B satisfies (4.1) consider x=hg\in H_\alpha and y=h'g'\in H_\beta, where h\in G_\alpha, g\in K_\alpha, h'\in G_\beta, g'\in K_\beta for all \alpha,\beta\in\pi. We have

\begin{aligned} &&B_{\beta\alpha}(x_{(1,\alpha)}B_\alpha(x_{(2,\alpha)})yS_{\bar{\alpha}}(B_\alpha(x_{(3,\alpha)})))\\ &=B_{\beta\alpha}((h_{(1,e)}g_{(1,\alpha)})(\varepsilon_e(h_{(2,e)})S_\alpha (g_{(2,\alpha)}))(h'g')(\varepsilon_e(h_{(3,e)})S_{\bar{\alpha}}(S_\alpha(g_{(3,\alpha)}))))\\ &=B_{\beta\alpha}(\varepsilon_e(h_{(1,e)})\varepsilon_e(h_{(2,e)})h_{(3,e)}g_{(1,\alpha)}S_\alpha(g_{(2,\alpha)}) h'g' g_{(3,\alpha)})\\ &=B_{\beta\alpha}(hh' g' \varepsilon_e(g_{(1,\alpha)})g_{(2,\alpha)}) = B_{\beta\alpha}(hh'g'g)\\ &=\varepsilon_{e}(hh')S_{\beta\alpha}(g'g)=\varepsilon_e(h)\varepsilon_e(h')S_\alpha(g)S_\beta(g')\\ &=B_\alpha(x)B_\beta(y). \end{aligned}

Since H_\alpha=G_\alpha K_\alpha is spanned by elements of the form hg, where h\in G_\alpha, g\in K_\alpha for all \alpha\in\pi, the equation (4.1) holds for all x,y\in H.

Uses Definition 4.1

Lemma

Lemma 4.5

1GCL

Let (H=\{H_\alpha\}_{\alpha\in\pi},B) be a Rota-Baxter Hopf \pi-algebra. Define

\begin{aligned} & m_{\alpha,\beta}(g,h)=g\circ_B h:=g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})), & \\ & T_\alpha(g):=S_{\bar{\alpha}}(B_\alpha(g_{(1,\alpha)}))S_\alpha(g_{(2,\alpha)})B_\alpha(g_{(3,\alpha)}), & \end{aligned}

for all g\in H_\alpha,h\in H_\beta, ~\alpha,\beta\in\pi.

H_B:=(H,m=\{m_{\alpha,\beta}\}_{\alpha\in\pi},T=\{T_{\alpha,\beta}\}_{\alpha\in\pi}) is a cocommutative Hopf \pi-algebra.
For all h\in H_\alpha, \alpha\in\pi, we have
\begin{aligned} & B_\alpha(h_{(1,\alpha)})B_{\bar{\alpha}}(T_\alpha(h_{(2,\alpha)}))=\varepsilon_\alpha(h)1_e, & \\ & B_{\bar{\alpha}}T_\alpha = S_{\bar{\alpha}}B_\alpha. & \end{aligned}
The operator B is also a Rota-Baxter operator on H_B, that is, (H_B,B) is a Rota-Baxter Hopf \pi-algebra.
The family of maps B:H_B\rightarrow H is a homomorphism of Rota-Baxter Hopf \pi-algebras.

Used by Proof

Proof of Lemma 4.5

(1) First, we prove m=\{m_{\alpha,\beta}\}_{\alpha,\beta\in\pi} is associative. For all g\in H_\alpha, h\in H_\beta, \ell\in H_\gamma, ~\alpha,\beta,\gamma\in\pi,

\begin{aligned} & & (g\circ_B h) \circ_B\ell \\ &= (g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)}))) \circ_B\ell \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h_{(1,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) B_{\beta\alpha}\big( g_{(4,\alpha)} B_\alpha(g_{(5,\alpha)})h_{(2,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(6,\alpha)})) \big) \\ &&\quad \ell S_{\bar{\beta\alpha}}\big(B_{\beta\alpha}( g_{(7,\alpha)} B_\alpha(g_{(8,\alpha)})h_{(3,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(9,\alpha)})) )\big) \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h_{(1,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) B_\alpha(g_{(4,\alpha)}) B_\beta(h_{(2,\beta)})\ell S_{\bar{\beta\alpha}}( B_\alpha(g_{(3,\alpha)}) B_\beta(h_{(3,\beta)}) ) \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h_{(1,\beta)} B_\beta(h_{(2,\beta)})\ell S_{\bar{\beta}}(B_\beta(h_{(3,\beta)})) S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) \\ &= g\circ_B( h_{(1,\beta)} B_\beta(h_{(2,\beta)})\ell S_{\bar{\beta}}(B_\beta(h_{(3,\beta)})) ) \\ &= g\circ_B(h\circ_B \ell). \end{aligned}

Then, we prove m=\{m_{\alpha,\beta}\}_{\alpha,\beta\in\pi} is a \pi-coalgebra homomorphism. For all g\in H_\alpha, h\in H_\beta, ~\alpha,\beta\in\pi,

\begin{aligned} & & (g\circ_B h)_{(1,\alpha\beta)} \otimes (g\circ_B h)_{(2,\alpha\beta)} = (g\circ_B h)_{(1,\beta\alpha)} \otimes (g\circ_B h)_{(2,\beta\alpha)} \\ &= (g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})))_{(1,\beta\alpha)} \otimes (g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})))_{(2,\beta\alpha)} \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h_{(1,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) \otimes g_{(4,\alpha)}B_\alpha(g_{(5,\alpha)})h_{(2,\beta)}S_{\bar{\alpha}}(B_\alpha(g_{(6,\alpha)})) \\ &= (g_{(1,\alpha)}\circ_B h_{(1,\beta)})\otimes(g_{(2,\alpha)}\circ_B h_{(2,\beta)}). \end{aligned}

Last, we prove T=\{T_\alpha\}_{\alpha\in\pi} is an antipode. For all h\in H_\alpha, ~\alpha\in\pi,

\begin{aligned} & & m_{\alpha,\bar{\alpha}}(id_{H_{\alpha}} \otimes T_\alpha)\Delta_\alpha(h) \\ &= h_{(1,\alpha)} \circ_B S_{\bar{\alpha}}(B_\alpha(h_{(2,\alpha)}))S_\alpha(h_{(3,\alpha)})B_\alpha(h_{(4,\alpha)}) \\ &= h_{(1,\alpha)}B_\alpha(h_{(2,\alpha)}) S_{\bar{\alpha}}(B_\alpha(h_{(3,\alpha)}))S_\alpha(h_{(4,\alpha)})B_\alpha(h_{(5,\alpha)}) S_{\bar{\alpha}}(B_\alpha(h_{(6,\alpha)})) \\ &= h_{(1,\alpha)} \varepsilon_\alpha(h_{(2,\alpha)}) S_\alpha(h_{(3,\alpha)}) \varepsilon_\alpha(h_{(4,\alpha)}) 1_e \\ &= \varepsilon_\alpha(h_{(1,\alpha)}) \varepsilon_\alpha(h_{(2,\alpha)}) 1_e \\ &= \varepsilon_\alpha(h)1_e. \end{aligned}

So m_{\alpha,\bar{\alpha}}(id_{H_{\alpha}} \otimes T_\alpha)\Delta_\alpha = \varepsilon_\alpha 1_e. Similarly prove m_{\bar{\alpha},\alpha}(T_\alpha \otimes id_{H_{\alpha}})\Delta_\alpha = \varepsilon_\alpha 1_e. Hence, H with the new multiplication m=\{m_{\alpha,\beta}\}_{\alpha\in\pi} and antipode T=\{T_{\alpha,\beta}\}_{\alpha\in\pi}) is a cocommutative Hopf \pi-algebra

(2) For all h\in H_\alpha, ~\alpha\in\pi, we have

\begin{aligned} & & B_\alpha(h_{(1,\alpha)})B_{\bar{\alpha}}(T_\alpha(h_{(2,\alpha)})) \\ &= B_\alpha(h_{(4,\alpha)}) B_{\bar{\alpha}}(S_{\bar{\alpha}}(B_\alpha(h_{(2,\alpha)}))S_\alpha(h_{(3,\alpha)})B_\alpha(h_{(4,\alpha)})) \\ &= B_e(h_{(1,\alpha)}B_\alpha(h_{(2,\alpha)}) S_{\bar{\alpha}}(B_\alpha(h_{(3,\alpha)}))S_\alpha(h_{(4,\alpha)})B_\alpha(h_{(5,\alpha)}) S_{\bar{\alpha}}(B_\alpha(h_{(6,\alpha)}))) \\ &= B_e(h_{(1,\alpha)} \varepsilon_\alpha(h_{(2,\alpha)}) S_\alpha(h_{(3,\alpha)}) \varepsilon_\alpha(h_{(4,\alpha)}) 1_e) \\ &= \varepsilon_\alpha(h_{(1,\alpha)}) \varepsilon_\alpha(h_{(2,\alpha)}) B_e(1_e) \\ &= \varepsilon_\alpha(h)1_e. \end{aligned}

From the above and B_\alpha(h_{(1,\alpha)})S_{\bar{\alpha}}(B_\alpha(h_{(2,\alpha)}))=\varepsilon_\alpha(B_\alpha(h))=\varepsilon_\alpha(x)1_e, we have B_{\bar{\alpha}}T_\alpha and S_{\bar{\alpha}}B_\alpha are the convolutional inverse for B_\alpha. Hence B_{\bar{\alpha}}T_\alpha = S_{\bar{\alpha}}B_\alpha, for all \alpha\in\pi.

(3) From Eq. (4.1) and (4.5), we have

\begin{aligned} B_{\alpha\beta}(g \circ_B h) &= B_{\beta\alpha}(g \circ_B h) \\ &= B_{\beta\alpha}( g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) ) \\ &= B_\alpha(g) B_\beta(h), \end{aligned}

for all g\in H_\alpha, h\in H_\beta, ~\alpha,\beta\in\pi.

Then we have

\begin{aligned} & & B_{\beta\alpha}\big(g_{(1,\alpha)}\circ_B B_\alpha(g_{(2,\alpha)})\circ_B h \circ_B T_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)}))\big) \\ &= B_{\alpha}(g_{(1,\alpha)}) B_{\bar{\alpha}}(B_\alpha(g_{(2,\alpha)})) B_{\beta}(h) B_{\alpha}(T_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)}))) \\ &= B_{\alpha}(g_{(1,\alpha)}) B_{\bar{\alpha}}(B_\alpha(g_{(2,\alpha)})) B_{\beta}(h) S_{\alpha}(B_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)}))) \\ &= B_{\alpha}(g) \circ_B B_{\alpha}(h). \end{aligned}

for all g\in H_\alpha, h\in H_\beta, ~\alpha,\beta\in\pi. The operator B is also a Rota-Baxter operator on the cocommutative Hopf \pi-algebra H_B.

(4) From (3), it is evident that the family of maps B:H_B\rightarrow H is a homomorphism of Rota-Baxter Hopf \pi-algebras.

Uses Definition 4.1

Theorem

Theorem 4.6

QUMG

Let (H=\{H_\alpha\}_{\alpha\in\pi},\cdot,B) be a Rota-Baxter Hopf \pi-algebra, and the multiplication \circ_B and antipode T defined as (4.5) and (4.5). Then (H,\cdot,\circ_B) is a cocommutative Hopf \pi-brace.

Proof of Theorem 4.6

By Lemma 4.5, (H,\circ_B,T) is a cocommutative Hopf \pi-algebra. For all g\in H_\alpha, h\in H_\beta, \ell\in H_\gamma, ~\alpha,\beta,\gamma\in\pi, we have

\begin{aligned} & & (g_{(1,\alpha)}\circ_B h)S_{\bar{\alpha}}(g_{(2,\alpha)})(g_{(3,\alpha)}\circ_B \ell) \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) S_{\bar{\alpha}}(g_{(4,\alpha)}) g_{(5,\alpha)}B_\alpha(g_{(6,\alpha)})\ell S_{\bar{\alpha}}(B_\alpha(g_{(7,\alpha)})) \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})hS_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) B_\alpha(g_{(4,\alpha)})\ell S_{\bar{\alpha}}(B_\alpha(g_{(5,\alpha)})) \\ &= g_{(1,\alpha)}B_\alpha(g_{(2,\alpha)})h\ell S_{\bar{\alpha}}(B_\alpha(g_{(3,\alpha)})) \\ &= g\circ_B(h\ell). \end{aligned}

Hence (H,\cdot,\circ_B) is a cocommutative \pi-Hopf brace.

Uses Lemma 4.5

Source and licence

The text on this page is from arXiv:2507.20176v2 by Yan Ning, Xing Wang and Daowei Lu, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.

The record of each item is in analysis/objects/2507/2507.20176.jsonl.