
are non-renormalizable, for example φR2, φRµν Rµν , give rise to deviations from GR in strongly
gravitating systems e.g. black holes, neutron stars, and binary pulsars [22].
In this work, we investigate the prospect for constraining UV modifications of GR focusing on
scalar-Gauss-Bonnet gravity (SGB) in which a scalar field couples linearly to the Gauss-Bonnet
topological invariant1
G=R2−4Rµν Rµν +Rµναβ Rµναβ .(1.1)
Such theories are extremely interesting from a theoretical perspective for several reasons. First,
couplings of the form φGrepresent the leading-order scalar-graviton interaction in shift-symmetric
theories [23]. Second, the equations of motion resulting from such couplings are second-order, mean-
ing that the theory does not suffer from an Ostrogradski ghost instability. Third, more generalized
couplings of the form f(φ)Garise naturally in string theory [24,25], may explain the accelerated
expansion of the universe [26,27], and give rise to the novel phenomenomenon of spontaneous
black hole scalarization [25,28,29]. Studying the simplest coupling φGwill lay the foundation for
constraining these more complicated theories.
The effects of SGB gravity are most pronounced in the strong-field regime, and hence black
holes are powerful probes of this theory [30–33]. This is helped by the fact that although smooth
extended objects cannot obtain a scalar charge in the theory, black holes can [23,34]. Furthermore,
certain types of couplings can be restricted based on theoretical arguments alone [35]. It is natural
to wonder how laboratory and Solar System experiments compare against astrophysical tests of
this theory. There currently exists a wide range of tabletop experiments that employ radically
different source mass geometries to test gravity in the weak field regime. Experiments like torsion
balances [36], atom interferometers [37,38], and Casimir force sensors [39–41] have proven in recent
years to be extraordinarily useful thanks to their high accuracy and their ability to be tuned to
search for effects in specific theories (see [42] for a review of laboratory tests of gravity). For
instance, large atom interferometers have performed some of the most precise measurements to
date on Newton’s constant G[38], while miniature ones have proven sensitive to screened modified
gravity theories that are otherwise very difficult to constrain [12,43–46].
Some of the first experimental constraints on SGB gravity were derived from Solar System tests
and focused on the specific case in which the scalar field drives the accelerated expansion of the
universe [26]. The PPN expansion of the theory has been computed, and it was found that the
theory is indistinguishable from GR at second post-Newtonian order [47]. This does not mean that
the theory is impossible to constrain via local tests of gravity, only that the theory does not fit
into the PPN framework. Deviations from GR were computed around point particles in [48], which
were then used to place bounds from Solar System and laboratory tests. In this work we relax the
point particle assumption, which enables us to use a larger range of experimental tests to constrain
the theory. Specifically, we compute the weak field limit of SGB gravity and study the deviations
from GR around extended objects with planar, cylindrical, and spherical symmetry. We find a
1We use units in which c=~= 1 and have defined the reduced Planck mass as MPl ≡(8πG)−1/2.
2