The hybridisation of sulfur in anion SO42- is given to be sp3.
Why is it so? Sulfur has two double bond and two single bond with
four oxygen atoms, meaning there are a total of 12 electrons
surrounding sulfur. sp3 only accomodates up to 8 electrons.
besides, sp3 means that we are assuming there are only four single
S-O bonds, which is not the case since resonance structure for it
exists.
Also given that we know resonance structure exists for the
sulfate ion, it must mean there is an unhypridised p orbital on
sulfur. if sulfur were to be sp3 hybridised, there will not be any
unhybridised p orbital to overlap with other p orbitals of oxygen.
shouldn't the hybridisation of sulfur have beeen spd2? two
unhybridised p orbitals, as well as four spd2 orbitals, so there
will be 4 hybridised orbital to overlap with the ones in oxygen,
and 2 overlap of p orbitals to form two double
bonds?
Edit: in order for resonance structure for sulfate ion
exists, does it mean that one sulfur atom would need to have 4
unhybridised p orbitals to overlap with each p orbitals from 4
oxygen atom? or would 1 unhybridised p orbital for sulfur will do?
[likewise for CO32-, NO3-]
As the electron geometry about the S atom in SO4 2- is tetrahedral
(in order to maximize stabilities by minimizing electron pair
repulsions), hence sp3 hybridized orbitals are required for the
sigma bonding about the S atom.
At A levels, the usual major resonance contributor taught to
students, is that of an expanded octet with no separation of formal
charges (at least in part because formal charges are not usually
taught to most H2 students in the first place), ie. Fig 1
below.
But experimental evidence has indicated that such a resonance
contributor, is actually only a minor resonance contributor. The
major resonance contributor, is actually that of a dipositively
formal charged S atom, singly bonded to 4 O atoms with uninegative
formal charges, ie. Fig 2 below.

The (almost negligible) pi bonding that exists (for the minor
resonance contributors), utilize the d orbitals of the S atom to
overlap (poorly) with the p orbitals of the O atoms.
In other words, the bonding in the SO4 2- ion, is actually covalent
with significant ionic character (due to the formal charges present
in the major resonance contributor).
Experimental evidence has shown that the actual charge on the S
atom is only slightly less than its formal dipositive charge, as
such an S 2+ atom would accordingly be strongly
electron-withdrawing by induction, withdrawing electron-density
from the singly bonded O- atoms, towards the central S 2+
atom.
Concordantly, consequently and most interestingly, you could say
that the (slight) partial double bond character (ie. in terms of
bond length and bond strength) of the S=O bonds in the SO4 2- ion,
is partly (only slightly) due to resonance, and partly (somewhat
moreso) due to induction.
Edit: in order for resonance structure for sulfate ion
exists, does it mean that one sulfur atom would need to have 4
unhybridised p orbitals to overlap with each p orbitals from 4
oxygen atom? or would 1 unhybridised p orbital for sulfur will do?
[likewise for CO32-, NO3-]
1 unhybridized p orbital would be sufficient, but because only one
unhybridized p orbital is available, the sideways overlap with the
p orbitals of the surrounding atoms to form pi bonds is only
*partial*. This is concordant with the observation that in the
resonance hybrid, each C=O bond in CO3 2-, and each N=O bond in NO3
-, only has *partial* double bond character.
One last point : for elements beyond period 2, while their atoms
often appear to achieve an expanded octet (using vacant,
energetically accessible 3d orbitals), which is what is taught at A
levels, however experimental evidence has shown this to be
inaccurate. In point of fact, all molecules/ions with central atoms
that appear to have an expanded octet, actually have as their major
resonance contributor, the central atom with only a stable octet,
and (accordingly) with separation of formal charges. In other
words, in the actual structure or resonance hybrid of the
molecule/ion, the bonding is always covalent with significant ionic
character.