Ehresmann-Schauenburg Hopf Algebroid

• The Ehresmann–Schauenburg Hopf algebroid is a quantum gauge groupoid constructed from a faithfully flat Hopf–Galois extension using a diagonal coinvariant subalgebra.
• It links Hopf–Galois theory, bialgebroids, and noncommutative geometry by incorporating source and target maps, coproducts, and antipode structures.
• Its versatile formulation enables practical applications in gauge symmetry, differential calculi, and deformation theory within quantum principal bundles.

An Ehresmann–Schauenburg Hopf algebroid is the quantum-groupoid object attached to a faithfully flat Hopf–Galois extension $B=P^{coH}\subseteq P$. In the notation used across the literature, it is the diagonal coinvariant algebra

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$

or equivalently $\mathcal C(A,H)$ when the total comodule algebra is denoted by $A$, equipped with source and target maps from $B$, multiplication inherited from $P\otimes P^{op}$, and a coproduct built from the Hopf–Galois translation map. Geometrically, it is the noncommutative analogue of the gauge groupoid $(P\times P)/G\rightrightarrows B$ of a principal bundle; algebraically, it lies at the intersection of Hopf–Galois theory, bialgebroids, and Hopf algebroids in the Schauenburg and Böhm–Szlachányi sense (Dabrowski et al., 2023, Beggs et al., 2024).

1. Construction from a Hopf–Galois extension

The standard input consists of a Hopf algebra $H$, a right $H$-comodule algebra $P$ with coaction

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$0

and the coinvariant subalgebra

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$1

The extension $\mathcal L(P,H)=(P\otimes P)^{co(H)},$2 is Hopf–Galois when the canonical map

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$3

is bijective. Its inverse determines the translation map

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$4

which controls the coproduct and all later structure maps (Dabrowski et al., 2023).

With this data, the Ehresmann–Schauenburg total algebra is the diagonal coinvariant subalgebra

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$5

Equivalent formulations use the translation map, for example

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$6

The algebra structure is the subalgebra structure inside $\mathcal L(P,H)=(P\otimes P)^{co(H)},$7,

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$8

with source and target maps

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$9

The left $\mathcal C(A,H)$0-coring structure is

$\mathcal C(A,H)$1

This is the basic algebraic incarnation of the quantum gauge groupoid (Beggs et al., 2024).

The notation varies across the literature, but the underlying construction is the same.

Notation	Source	Structural emphasis
$\mathcal C(A,H)$2	(Dabrowski et al., 2023)	Ehresmann–Schauenburg bialgebroid
$\mathcal C(A,H)$3	(Han et al., 2022)	left Hopf algebroid
$\mathcal C(A,H)$4	(Beggs et al., 2024)	$\mathcal C(A,H)$5-Hopf algebroid pair/full case
$\mathcal C(A,H)$6	(Kowalzig et al., 22 Jul 2025)	right Hopf algebroid and calculi

2. Structural position inside Hopf algebroid theory

The same object is treated in several axiom systems. In the Hopf–Galois literature it first appears as a left $\mathcal C(A,H)$7-bialgebroid. In Schauenburg’s sense, the relevant Hopf condition is bijectivity of the left Hopf–Galois map

$\mathcal C(A,H)$8

For the Ehresmann–Schauenburg object this condition holds: $\mathcal C(A,H)$9 is a left Hopf algebroid, and the inverse of $A$0 is constructed explicitly from the Hopf–Galois translation map of the original extension (Han et al., 2022).

A complementary result treats the same construction from the opposite side. For

$A$1

the right Hopf–algebroid translation map is

$A$2

and $A$3 is a right Hopf algebroid; the same source also records that the Ehresmann–Schauenburg bialgebroid is known to be a left Hopf algebroid as well (Kowalzig et al., 22 Jul 2025).

A different level of structure arises in the Böhm–Szlachányi framework. There one distinguishes a weaker Hopf condition, formulated through Schauenburg’s canonical map, from a full Hopf algebroid equipped with an invertible antipode. For the Ehresmann–Schauenburg bialgebroid $A$4, the central question is precisely when such an antipode exists. In particular, if the flip map on $A$5 is a right $A$6-comodule endomorphism for the diagonal coaction, then it is an antipode; if $A$7 is commutative, this condition is automatic, and the Ehresmann–Schauenburg bialgebroid becomes a full Hopf algebroid with flip antipode (Dabrowski et al., 2023).

This multiplicity of formulations is not a contradiction. It reflects the fact that the same quantum gauge-groupoid object may be viewed as a left bialgebroid, a left or right Hopf algebroid in the Schauenburg sense, or a full Hopf algebroid in the Böhm–Szlachányi sense once an antipode is available.

3. Antipodes, $A$8-structures, and deformation mechanisms

Antipodes on the Ehresmann–Schauenburg object are highly sensitive to the fibre Hopf algebra and to the geometry of the extension. For a faithfully flat Hopf–Galois extension, the generic structure is a left bialgebroid; when $A$9 is commutative, the flip

$B$0

gives a full Hopf algebroid. More generally, in the quantum projective-space example

$B$1

there are two distinct antipodes: the flip antipode coming from commutativity of $B$2, and a second antipode built from the two $B$3-classes determined by projectors $B$4 and $B$5, with

$B$6

These two antipodes are related by a twist in the convolution algebra, and the twist group is identified with the group of $B$7-comodule algebra automorphisms of the total space (Dabrowski et al., 2023).

The $B$8-theory sharpens this picture. If $B$9 is a $P\otimes P^{op}$0-algebra, $P\otimes P^{op}$1 is a Hopf $P\otimes P^{op}$2-algebra with bijective antipode, and the coaction is unitary, then

$P\otimes P^{op}$3

forms a $P\otimes P^{op}$4-Hopf algebroid pair with

$P\otimes P^{op}$5

If, in addition, $P\otimes P^{op}$6 is commutative, then $P\otimes P^{op}$7 is a full $P\otimes P^{op}$8-Hopf algebroid with

$P\otimes P^{op}$9

A key technical input is the $(P\times P)/G\rightrightarrows B$0-compatibility of the translation map,

$(P\times P)/G\rightrightarrows B$1

which transports the involutive structure from the bundle data to the quantum gauge groupoid (Beggs et al., 2024).

Cocycle deformation provides a second major mechanism. For a cleft Hopf–Galois extension $(P\times P)/G\rightrightarrows B$2, the Ehresmann–Schauenburg bialgebroid is a cocycle twist of the untwisted smash-product case: $(P\times P)/G\rightrightarrows B$3 under suitable centrality and cocommutativity hypotheses. At the left Hopf–algebroid level, this becomes

$(P\times P)/G\rightrightarrows B$4

for cleft extensions of associative type. In that sense, “switching on” the crossed-product cocycle $(P\times P)/G\rightrightarrows B$5 is transported functorially to a Drinfeld cotwist of the associated Ehresmann–Schauenburg Hopf algebroid (Han, 2020, Han et al., 2022).

4. Bisections, cocycles, and nonAbelian symmetry theory

The groupoid interpretation becomes most visible through bisections. For the Ehresmann–Schauenburg Hopf algebroid of a faithfully flat Hopf–Galois extension, the left and right bisection groups reduce to the bundle-automorphism group: $(P\times P)/G\rightrightarrows B$6 The correspondence is explicit. A left bisection $(P\times P)/G\rightrightarrows B$7 determines

$(P\times P)/G\rightrightarrows B$8

whereas a bundle automorphism $(P\times P)/G\rightrightarrows B$9 yields

$H$0

Thus the noncommutative analogue of “bisections of the gauge groupoid are gauge transformations” is literally valid in this setting (Han et al., 2023).

Vertical bisections encode internal gauge symmetry. They identify with a multiplicative cocycle space

$H$1

and bundle automorphisms restricting to the identity on $H$2 correspond to such cocycles. This produces a concrete noncommutative gauge-theoretic interpretation of the first cohomology level (Han et al., 2023).

The second level is a nonAbelian cohomology $H$3 governing cotwists of the Hopf algebroid. Its elements are convolution-invertible normalized $H$4-cocycles on $H$5, and the cotwisted product is

$H$6

For $H$7-commutative $H$8, the abstract Hopf-algebroid cohomology becomes a concrete Hopf-algebraic theory: $H$9 For trivial or cleft bundles, it reduces further to explicit associative-type cocycle data on $H$0 with values in $H$1 (Han et al., 2023).

The same source also introduces coquasi-Ehresmann–Schauenburg bialgebroids, where associativity is controlled by a $H$2-cocycle. This extends the gauge-groupoid picture from strict Hopf algebroids to coquasi-bialgebroid structures, without changing the basic role of $H$3 as the total algebra (Han et al., 2023).

5. Differential calculi, base change, and descent

The augmentation ideal of the Ehresmann–Schauenburg Hopf algebroid is closely tied to universal differential forms on the total space. For

$H$4

one has

$H$5

where $H$6 is the universal first-order calculus on $H$7. As a consequence, left covariant first-order differential calculi on $H$8 are in bijection with right $H$9-covariant first-order differential calculi on $P$0: $P$1 This identifies the differential geometry of the quantum gauge groupoid with the covariant differential geometry of the total space itself (Kowalzig et al., 22 Jul 2025).

The same work places the Ehresmann–Schauenburg example inside a broader Hopf-algebroid theory of covariant calculi, Hopf modules, and Takeuchi–Schneider equivalence. In particular, the explicit right Hopf–algebroid translation map on $P$2 makes the Woronowicz-type classification of left covariant calculi entirely effective for this family (Kowalzig et al., 22 Jul 2025).

Base change is subtler in the noncentral case. Given an algebra map

$P$3

the naive tensor product $P$4 need not inherit an algebra structure. The remedy is a twisting map

$P$5

which defines a twisted tensor-product algebra

$P$6

Under the compatibility conditions stated in the source, $P$7 is again an $P$8-Galois extension, with translation map

$P$9

and its own Ehresmann–Schauenburg bialgebroid

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$00

There is always a bialgebroid morphism from the original $\mathcal L(P,H)=(P\otimes P)^{co(H)},$01 to $\mathcal L(P,H)=(P\otimes P)^{co(H)},$02, and under an additional inner $\mathcal L(P,H)=(P\otimes P)^{co(H)},$03-linearity condition the comparison is refined to a map from the standard base-ring extension

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$04

to $\mathcal L(P,H)=(P\otimes P)^{co(H)},$05; if $\mathcal L(P,H)=(P\otimes P)^{co(H)},$06 is invertible, this map is an isomorphism. The resulting interpretation is that noncentral push-forward transforms the Ehresmann–Schauenburg bialgebroid by a twisted base change rather than by ordinary extension of scalars (Landi et al., 23 Dec 2025).

6. Examples, applications, and scope

The most detailed worked example is the $\mathcal L(P,H)=(P\otimes P)^{co(H)},$07-Hopf–Galois extension

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$08

where $\mathcal L(P,H)=(P\otimes P)^{co(H)},$09. The translation map is

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$10

The corresponding Ehresmann–Schauenburg bialgebroid is generated by two matrix families $\mathcal L(P,H)=(P\otimes P)^{co(H)},$11 and $\mathcal L(P,H)=(P\otimes P)^{co(H)},$12, built from the two $\mathcal L(P,H)=(P\otimes P)^{co(H)},$13-classes $\mathcal L(P,H)=(P\otimes P)^{co(H)},$14 and $\mathcal L(P,H)=(P\otimes P)^{co(H)},$15, with matrix-coalgebra formulas

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$16

This example exhibits both a flip antipode and a $\mathcal L(P,H)=(P\otimes P)^{co(H)},$17-theoretically defined antipode, together with a full twist-group computation $\mathcal L(P,H)=(P\otimes P)^{co(H)},$18 (Dabrowski et al., 2023).

A second explicit family comes from the Hopf fibration over the Podleś sphere. For

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$19

the Ehresmann–Schauenburg bialgebroid is generated by eight coinvariant elements

$\mathcal L(P,H)=(P\otimes P)^{co(H)},$20

with coproduct and counit computed explicitly, and the induced universal calculus on the bialgebroid is written out on generators. This example is the main test case for the classification of covariant calculi (Kowalzig et al., 22 Jul 2025).

The phrase “Ehresmann–Schauenburg Hopf algebroid” does not cover every nearby Hopf-algebroid construction. The Hopf algebroid $\mathcal L(P,H)=(P\otimes P)^{co(H)},$21 attached to partial representations of a Hopf algebra is explicitly described as a Schauenburg/Böhm–Szlachányi-style Hopf algebroid, but it is not a gauge-groupoid construction of Hopf–Galois type (Alves et al., 2013). Conversely, the Hopf algebra object attached to a Lie pair $\mathcal L(P,H)=(P\otimes P)^{co(H)},$22 lives in a derived category and is not a Hopf algebroid in the Schauenburg or Ehresmann sense (Chen et al., 2014). These distinctions matter because the defining feature of the Ehresmann–Schauenburg object is not merely the presence of source, target, coproduct, and antipode, but its origin as the quantum gauge groupoid of a Hopf–Galois extension.

In that precise sense, the modern theory presents the Ehresmann–Schauenburg Hopf algebroid as a flexible but coherent construction: a coinvariant subalgebra of $\mathcal L(P,H)=(P\otimes P)^{co(H)},$23, functorial in Hopf–Galois data, compatible with twists, $\mathcal L(P,H)=(P\otimes P)^{co(H)},$24-structures, calculi, and push-forward procedures, and rich enough to recover automorphism groups, gauge cocycles, and deformation theory directly from the geometry of the underlying quantum principal bundle (Beggs et al., 2024, Han et al., 2023, Kowalzig et al., 22 Jul 2025).