@User: In the Sky130 process, if you want to lower the threshold, you use a low threshold device. This is done by threshold-shifting implants.

@Tim Edwards What did you exactly mean by threshold-shifting implants ?

N+ and P+ implants that shift the threshold by additional doping of the diffusion at the source and drain. From the HVT to the LVT to the native transistors (blocking the initial implant rather than adding an additional one), there is a range of thresholds on the standard 1.8V FET devices of around 1V.

@Tim Edwards how does additional doping at source and drain LOWER the threshold voltage ?

You can dope both ways, N or P, into diffusion that is already doped N or P. But I don't know how the masks are derived (I think that's in the documentation but I haven't looked at it closely), so it's also possible that the low-threshold doping first blocks the initial doping and then adds a specific doping after that. There are a number of ways to do it.

@Tim Edwards I am confused, how is this related to threshold voltage ?

From S. M. Sze, "VLSI Technology": Vt = phi_ms - (Qf / Cox) + 2 * psi_b - (Qb / Cox), where phi_ms is the metal-silicon work function of the gate material, Qf is the charge in the gate oxide, psi_b is the substrate Fermi potential, Qb is the depletion region charge in the substrate, and Cox is the capaitance of the gate. Qb is proportional to the square root of the channel doping. While doping the polysilicon gate will change the threshold, the most common threshold adjustment is to dope the channel under the gate.

@promach Vt threshold implants are typically done with so-called channel implants. These are done before gate patterning so can actually seen as locally changing the bulk doping.