# Why is (NH4)2[CuCl4] square planar complex but Cs2[CuCl4] is tetrahedral?

Why is $\ce{(NH4)2[CuCl4]}$ square planar complex but $\ce{Cs2[CuCl4]}$ is tetrahedral even though both have same oxidation number of copper and same ligands?

In solution, the $\ce{[CuCl4]}$2- is expected to exhibit terahedral, or nearly tetrahedral geometry.

The Jahn-Teller theorem states that degenerate orbitals cannot be unequally occupied. Molecules with unequally occupied degenerate orbitals will distort so as to render the orbitals non-degenerate, with more electrons occupying the lower energy states. Generally, this distortion leads to a reduction in symmetry. In the case of Cu (II), it is a d9 cation with a tetrahedral electronic configuration of (e)4(t2)5, so the three degenerate t2 orbitals are unequally occupied, therefore a distortion towards the square planar geometry of (eg)4(b2g)2(a1g)2(b1g)1, with no unequally occupied degenerate orbitals, is electronically favoured. Nonetheless, experimental evidence (1) indicates that this d-orbital stabilization energy is very small. Since copper is a first-row transition metal (not very big), steric effects dominate electronic effects allowing $\ce{[CuCl4]}$2- to assume a tetrahedral geometry. Most spectroscopic methods I've read involve looking at the molecule in crystal form (because that is where this effect gets interesting), but DFT studies of the $\ce{[CuCl4]}$2- hydrate in concentrated aqueous solution indicate that the anion's tetrahedral structure is more stable (2).

The issue of $\ce{(Cs)2[CuCl4]}$ vs $\ce{(NH4)2[CuCl4]}$ comes up when the complex precipitates and crystals are formed. Due to the borderline electronic effects, the $\ce{[CuCl4]}$2- anion exhibits a great degree of structural flexibility. With large cations ($\ce{Cs+}$ radius is 0.174 nm while $\ce{NH4+}$ radius is 0.143 nm) , it crystallizes as a distorted ("squashed") tetrahedron.

$\hskip1.4in$ (3)

Square-planar co-coordination occurs in such a way that you effectively get octahedra with tetragonal distortion (4). I would imagine that this phenomenon is simply a result of the smaller cation allowing the unit cell to crystallize like that. The general idea with unit cells is that the cation is never allowed to "rattle" around, so the structure will find a way to fill in the space. Given the preexisting electronic favourability of square planar geometry, $\ce{(NH4)2[CuCl4]}$ this distortion favoured by close-packing is made even more favourable.

$\hskip2.1in$ (4)

The canonical reference for $\ce{(NH4)2[CuCl4]}$ 's crystal structure is (5) if you want to take a look at the unit cell.

(1) Emeléus & Sharpe. Advances in Inorganic Chemistry and Radiochemistry, vol. 21, p. 128

(3) L. Helmholz & R. F. Kruh, J. Am. Chem. Soc., 1952, 74, 1176

(4) Molecular Structure by Diffraction Methods, vol. 1, The Chemical Society, London, 1973, p. 642

(5) R. D. Willet, J. Chem. Phys., 1964, 1964, 41, 2243

• thanks for pointing out my typos! Yes, I was describing the extreme cases; the tetrahedral form in solution and the electronically favourable square planar case. In ammonium chlorocuprate's unit cell, the ion's geometry is the intermediate, non-inversion symmetric case. If anyone has access to (5), we could describe its symmetry. Sadly, this volume seems to have been misplaced in my library. – gannex Oct 9 '16 at 20:32
• Haha, my VPN gives me access. The article doesn’t include a picture though. To quote the discussion: ‘This structure contains the first instane of a square-planar $\ce{CuC4-}$ ion. The $\ce{Cl-Cu-Cl}$ bond angle is required by symmetry to be $90^\circ$. However, the $\ce{Cu-Cl}$ bond lengths are not required to be equal. Two are $2.300 \pm 0.005~\mathrm{A}$ long, and two are $2.332 \pm 0.004~\mathrm{A}$ long. Each copper atom completes its normal distorted octahedral configuration by forming two long $\ce{Cu-Cl}$ bonds of $2.793 \pm 0.005~\mathrm{A}$ with adjacent ions and thus forming an … – Jan Oct 9 '16 at 20:47
• … infinite two-dimensional network. The $\ce{Cu-Cl-Cu}$ bond angle is linear ($180^\circ \pm 0.2^\circ$) to within experimental error.’ – Jan Oct 9 '16 at 20:48
• is the bond length distortion due to electronic Jahn-Teller effects (as in chemistry.stackexchange.com/questions/43111/…), or is it just a consequence of the crystal packing? I have VPN too, but the journal isn't available in electronic format that far back for me – gannex Oct 9 '16 at 21:05
• They unfortunately don’t go into detail. I would say that JT doesn’t sound unreasonable, but it could also be some weird packing effects. Again, an image is not included, unfortunately. – Jan Oct 9 '16 at 21:06