Looking for very low threshold voltage NMOS and its PMOS counterpart

What is VTO?

I don't think such a part exists. A photodiode might just barely turn on a high-beta bipolar transisor.

What's the open-circuit voltage of the photodiode? What's the current? Is a power supply available? If the pd is back-biased by a supply, you'd have lots of voltage available to turn on a mosfet.

A lithium battery might power a micropower comparator for decades.

VTO: Vertical Take Off

I think they meant V-turn-on?

Strange sub threshold behaviors exist but I know nothing about them.

piglet

That's what amplifiers are for.

Jeroen Belleman

I've seen transistors do that. Just apply 100x rated power

A TO92 is what I think of as a transistor-on-the-half-shell.

Or maybe V-turn-off?

Sorry for the confusion. 'VTO' means threshold voltage for the mosfet. The candidate photodiode is BPW31|34. The datasheet contains the following data: Open circuit voltage : 440 mV In reverse bias, the diode is open circuit, so current should flow. The dark current is 12 pA.

Advanced Linear Devices make mos discretes with turn-on voltages dialable down to zero.

Yes, but the turn-on is trimmed, not made abrupt, so 'down to zero' is more of a gain specification than a full switch threshold. JFETs have been available for zero-threshold for years, and were quite useful for low-leakage input applications.

Your signal (a photocurrent) is fast? And you want to switch according to that current?

There's very little voltage gain in a single MOS device, and what you want might be to avoid current lost to charging the photodiode stray capacitance (i.e. you want near-zero dV/dt); a good receiver for a photodiode with low current output is an op amp with resistor negative feedback, i.e. a transimpedance amplifier. That way, there's negligible delta-V.

Totally in agreement about the op-amp transimpedance amplifier. However, the gain-bandwidth-product of even high end op-amps certainly would not extend beyond some 10s of MHz.