Cleavage planes

Q: Doesn't the presence of cleavage planes mean products out of nanoscale gemstones break easily?

A: No, not really.
Brittleness of macroscale gemstones is mainly a result of the unavoidable presence of defects (internal and on surface)
and continuous covalent connectivity leading to stress concentrations inducing fractures that unboundedly propagate
to a full break long (easily >99%) before the theoretical limit is reached.
Thereby massively reducing toughness.

Interlocking nanoscale gemstones like in gemstone based metamaterials
(will) remove that massive reduction. Even if the many of the cleavage planes are aligned.
★ by stopping fractures at the next very nearby non-bonded interface interlock and
★ by most crystolecules having no internal or external defects
… particularly no stress concentrating surface divest to get a cleavage plane started.

Cleaving planes as only a concern for the very highest performance applications

Cleavage planes in gemstones like diamond and are still well above 50% of the maximum ultimate tensile strength
which is still enormous and sufficient for almost all except the most extreme applications.

Plus there are a number more and less straightforward ways to improve on that
in the rare cases it is really needed. E.g.:

Approaches to mitigate for cleavage planes

Interlocking nano-gemstone crystolecules in different crystal orientations:
Design methodologies following ancient ones used in woodworking where wood has not one but two weak directions. Woodgrain. More recent very much like design methodologies for FFF/FDM 3D printed parts as these often have one particularly weaker direction due to imperfect inter-layer-adhesion.
Side-note: Post assembly of 3D printed parts is still quite underused despite quite a number of benefits including more isotropic and thus overall strength.
See: Limits of in place printing and in place mechanosysnthesis

Varying stacking order:
ABCABC is cubic ABAB is less isotropic hexagonal with a only one major axis insterad of four.
Other stacking patterns are possible too (this happens naturally in moissanite (SiC))
but one would likely want to make it in some less or more periodic systematic pattern.
See: Checkerboard compound

Quasiamorphous and quasicrystalling structures:
Fancy stuff to look at eventually.

Obviously avoid gems with really pronounced cleavage planes.
Some 2D materials like e.g. single crystalline graphite (aka HOPG)
have just mere vdW bonds in one direction.
These are still not weak from a breaking marcoscale parts by human hands perspective
but ~100x weaker than covalent.

Related

• Superelasticity
• Hierarchical intentional breaking interfaces