Spin(9) Symmetry in 16-D Geometry

• Spin(9) is a unique rank-4 Lie group defined as the double cover of SO(9), acting as the holonomy group of 16-manifolds like the Cayley plane.
• It is characterized by a canonical invariant 8-form on ℝ¹⁶ and a faithful 16-dimensional spin representation constructed using octonionic techniques.
• Spin(9) informs the construction of maximal vector fields on spheres and underlies key geometric structures in Clifford systems and exceptional holonomy.

Spin(9), the double cover of SO(9), is a rank-4 exceptional compact Lie group which plays a distinguished role in differential geometry, representation theory, and octonionic geometry. It arises as the holonomy group for certain 16-dimensional Riemannian manifolds, most notably the Cayley plane $\mathbb{O}P^{2}$, and governs the symmetries of the octonionic Hopf fibration. Central to its geometry is the unique Spin(9)-invariant 8-form on $\mathbb{R}^{16}$, providing a higher-dimensional analogy to the Kähler and quaternionic 4-forms, and establishing a deep connection between Spin(9) symmetry, calibrations, Clifford systems, and conformal holonomy.

1. The Spin(9) Group and its Spin Representation

Spin(9) is realized as the subgroup of GL(16, $\mathbb{R}$) that preserves a canonical 8-form on $\mathbb{R}^{16}$ (Ornea et al., 2012, Parton et al., 2018). Its fundamental irreducible representation is the 16-dimensional real spin representation, usually identified with the octonionic 2-plane $\mathbb{O}^2 \cong \mathbb{R}^{16}$. The action is constructed using the real Clifford algebra $\mathrm{Cl}_9$ and its unique irreducible real module, and can be given explicitly in terms of octonion right multiplications (Parton et al., 2011, Parton et al., 2018). The spin representation is faithful and irreducible, and underpins all geometric realizations of Spin(9)-symmetry in dimension 16.

The embedding $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$ is characterized by the existence of nine symmetric involutions $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$, satisfying

$$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$

These involutions reflect the Clifford relations for $\mathbb{R}^9$, and generate the Clifford algebra $\mathbb{R}^{16}$0 within $\mathbb{R}^{16}$1 (Parton et al., 2011, Kotrbatý, 2018).

2. The Canonical Spin(9)-Invariant 8-Form

A defining feature of Spin(9) geometry is the canonical, nowhere-vanishing Spin(9)-invariant 8-form $\mathbb{R}^{16}$2 on $\mathbb{R}^{16}$3. It is uniquely (up to scale) preserved by Spin(9) and can be defined both geometrically and algebraically:

• Geometric (Berger’s integral):

$\mathbb{R}^{16}$4

where $\mathbb{R}^{16}$5 are octonionic lines in $\mathbb{R}^{16}$6 ($\mathbb{R}^{16}$7), $\mathbb{R}^{16}$8 is orthogonal projection onto $\mathbb{R}^{16}$9, and $\mathbb{R}$0 is a normalization constant (Ornea et al., 2012, Parton et al., 2018).

• Algebraic (Pfaffian and characteristic polynomial):

Using the skew-symmetric $\mathbb{R}$1 “Kähler matrix” $\mathbb{R}$2 of 2-forms built from the symmetric involutions, the characteristic polynomial

$\mathbb{R}$3

has its quartic coefficient $\mathbb{R}$4 given by

$\mathbb{R}$5

and hence $\mathbb{R}$6 can be written explicitly as a sum of squares of wedge products of these Kähler forms (Parton et al., 2011, Parton et al., 2018).

• Octonion-valued form:

A modern “octonionic” expression presents $\mathbb{R}$7 as a quartic combination of four specifically constructed octonion-valued 4-forms built solely from the coordinate 1-forms of $\mathbb{R}$8 (Kotrbatý, 2018).

$\mathbb{R}$9 is the octonionic analogue of the Kähler 2-form in complex geometry and the Kraines 4-form in quaternionic geometry. For any 16-manifold $\mathbb{R}^{16}$0 with Spin(9)-structure, there is a canonical global 8-form $\mathbb{R}^{16}$1 pulling back locally from $\mathbb{R}^{16}$2 (Ornea et al., 2012, Parton et al., 2018).

3. Spin(9)-Structures and 16-Manifold Geometry

A Spin(9)-structure on an oriented Riemannian 16-manifold $\mathbb{R}^{16}$3 admits multiple equivalent descriptions (Ornea et al., 2012, Parton et al., 2011):

1. Principal bundle reduction: Existence of a principal Spin(9)-subbundle of the orthonormal frame bundle, or equivalently a spinor bundle associated to the 16-dimensional real spin representation.
2. Clifford subbundle: A rank-9 subbundle $\mathbb{R}^{16}$4 locally generated by symmetric involutions $\mathbb{R}^{16}$5 satisfying Clifford relations.
3. Invariant 8-form: A global, nowhere-vanishing 8-form $\mathbb{R}^{16}$6 equivalent at every point to $\mathbb{R}^{16}$7, i.e., the Spin(9)–invariant 8-form.

The existence of a Spin(9)-structure depends only on the conformal class of the metric $\mathbb{R}^{16}$8. The holonomy group of the Levi-Civita connection may reduce to Spin(9), making $\mathbb{R}^{16}$9 an irreducible Riemannian manifold with exceptional holonomy. The only compact irreducible examples are the Cayley projective plane $\mathbb{O}^2 \cong \mathbb{R}^{16}$0 and its noncompact dual (Parton et al., 2011).

4. Octonionic Hopf Fibration and Homogeneous Spin(9) Geometry

Spin(9) acts transitively on the unit sphere $\mathbb{O}^2 \cong \mathbb{R}^{16}$1, with stabilizer Spin(7), yielding the identification $\mathbb{O}^2 \cong \mathbb{R}^{16}$2. The octonionic Hopf fibration,

$\mathbb{O}^2 \cong \mathbb{R}^{16}$3

has fiber the 7-sphere $\mathbb{O}^2 \cong \mathbb{R}^{16}$4 and is homogeneous under the action of Spin(9). The associated bundle maps are

$\mathbb{O}^2 \cong \mathbb{R}^{16}$5

with Spin(7) ⊂ Spin(8) ⊂ Spin(9) (Ornea et al., 2012, Parton et al., 2018).

An important geometric constraint is that every smooth vector field tangent to the fibers of $\mathbb{O}^2 \cong \mathbb{R}^{16}$6 must have a zero. Consequently, there are no $\mathbb{O}^2 \cong \mathbb{R}^{16}$7-subfibrations. This is proved via the Hurwitz–Radon–Adams theorem, as a nowhere-vanishing vertical field would generate ten orthonormal tangent fields on $\mathbb{O}^2 \cong \mathbb{R}^{16}$8, exceeding the known bound of eight (Ornea et al., 2012, Parton et al., 2018).

5. Maximal Systems of Vector Fields on Spheres and the “Fault” of Spin(9)

The existence of the Spin(9) representation on $\mathbb{O}^2 \cong \mathbb{R}^{16}$9 has direct implications for the classical question of the maximal number of linearly independent vector fields on spheres. According to the Hurwitz–Radon–Adams theorem, the maximal number $\mathrm{Cl}_9$0 of linearly independent tangent vector fields on $\mathrm{Cl}_9$1 is given by

$\mathrm{Cl}_9$2

For $\mathrm{Cl}_9$3 ($\mathrm{Cl}_9$4), this gives $\mathrm{Cl}_9$5 (Parton et al., 2011, Parton et al., 2018). Spin(9) symmetry through its spinor representation constructs exactly eight everywhere orthonormal vector fields on $\mathrm{Cl}_9$6 by acting with the complex structures $\mathrm{Cl}_9$7 ($\mathrm{Cl}_9$8) on the unit normal field. Extension to higher spheres proceeds by block and diagonal constructions, providing the structural mechanism, beyond ordinary division algebras, for all spheres $\mathrm{Cl}_9$9 with $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$0 a multiple of $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$1 to admit more than seven independent vector fields (Parton et al., 2011).

6. Locally Conformally Parallel Spin(9) Manifolds

A Riemannian metric $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$2 on a 16-manifold $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$3 is called “locally conformally parallel Spin(9)” (LCP–Spin(9)) if locally, $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$4 is conformal to a metric with holonomy contained in Spin(9): $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$5 for an open cover $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$6 and functions $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$7 (Ornea et al., 2012, Parton et al., 2018).

Key properties of compact LCP–Spin(9) manifolds include (Ornea et al., 2012):

• The universal cover isometric to the metric cone $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$8 with the conic metric, implying that all such $\operatorname{Spin}(9)\subset \mathrm{SO}(16)$9 are finitely covered by $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$0.
• Existence of a canonical 8-dimensional Riemannian foliation, with leaves totally geodesic.
• Under compactness of the foliation leaves, $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$1 fibers over an 8-dimensional orbifold covered by $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$2, with typical fiber covered by $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$3.
• Any compact LCP–Spin(9) manifold is (up to finitely covered diffeomorphism) a quotient $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$4, with $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$5 finite acting freely, and metric structure group lying in the normalizer $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$6.

The canonical 8-form $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$7 on such $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$8 satisfies a conformal divergence relation with respect to the global Lee form $$I_1,\ldots,I_9 \in \operatorname{End}(\mathbb{R}^{16})$$9: $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$0 This framework generalizes earlier quaternionic and complex analogues, linking LCP–Spin(9) geometry to the W₄-component (“vectorial type”) of intrinsic torsion (Parton et al., 2018).

7. Clifford Systems, Grassmannians, and Exceptional Geometries

Spin(9) symmetry and its associated Clifford systems feature prominently in the classification and construction of even Clifford structures on Riemannian manifolds (Parton et al., 2018). A Clifford system $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$1 on $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$2 consists of $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$3 symmetric involutions satisfying $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$4. On $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$5, the nine involutions that define Spin(9) provide the unique irreducible $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$6 Clifford system.

Exceptional symmetric spaces—such as the “Cayley–Rosenfeld planes” $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$7, $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$8, $$I_a^2 = \mathrm{Id}, \qquad I_a I_b = -I_b I_a \quad (a\neq b).$$9, $\mathbb{R}^9$0—carry canonical even Clifford structures of appropriate ranks (9, 10, 12, 16) (Parton et al., 2018). Furthermore, families of oriented Grassmannians $\mathbb{R}^9$1, $\mathbb{R}^9$2, and $\mathbb{R}^9$3 support canonical Clifford structures, constructed naturally from their tautological bundles and the related spin/algebraic data.

These structures, and the associated canonical forms, are central to the study of calibrations, characteristic classes, and curvature invariants in high-dimensional geometry (Parton et al., 2011, Kotrbatý, 2018).

---

References:

(Ornea et al., 2012, Parton et al., 2018, Parton et al., 2011, Parton et al., 2011, Kotrbatý, 2018)