How is modulated the Primary Synchronization Signal in LTE?

PSS is based on 62 values from a Zadoff-Chu sequence. Three different sequences can be used by an LTE Cell, to represent the Physical Cell ID within the Cell ID group.

The formula is the following, $u$ being an index $[0;2]$ corresponding to a Physical Cell ID.

$$\begin{cases} \partial u(n)=e(−\dfrac{j\pi}{63} un(n+1))&\text{ for }n=0,1,…,30 \\ \partial u(n)=e(−j\pi u(n+1)(n+2)63)&\text{ for }n=31,32,…,61 \end{cases}$$

Now my question: is there any kind of QAM or QPSK applied to each of the 62 values we get from the sequence, before they are mapped to the time/frequency grid (Radio frame)?

1. I know that Secondary Synchronization Signal (composed of 31-length M sequence) is modulated by BPSK before being mapped.
2. I've found that and OFDM subcarrier has for formula $s(t) = A\cos(wt + \Phi(t))$. I still haven't found yet where this intervenes.
3. For a 1.4 MHz bandwidth, a 126 points IFFT is performed, if you know what that really means (physically speaking) beyond switching from frequency to time-domain.

In the end I would like to get the proper antenna output signal carrying a PSS, to study it.