Perhaps you can talk to these guys?
http://www.com-power.com/antennas.html

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

"... from 100 kHz to 1 GHz. I am hoping that there would be one directional-type antenna that would cover that range..."

Typically one would need a collection of E3 antennas to cover that sort range.

To answer the next obvious question, by referencing the "Antenna Factor" data included with a calibrated E3 antenna, one may translate the measured signal level into actual field strength.

Thanks, VE1BLL. That is very helpful.

directional at 0.1 Mhz is a huge antenna. You won't find a directional antenna, law of physics. AHSystems has the frequency you need http://www.ahsystems.com/catalog/currentprobes.php

Exception to the rule: Ferrite 'Loop-stick' antennas can cover 0.1 MHz, and are directional (cardioid pattern with a lovely null). Used for MF-DF systems.

But I don't believe that they make very good EMC antennas. And certainly not for transmit mode!

If I were making a Pre-EMC Test, I'd make my own antenna. Assuming you have a box radiating problem, just a coax. center conductor poking out of the cable may be adequate. Let the length of the center conductor be extended by adding a spool of wire, or maybe wrap wire around a ferrite.
The thought is to have a small, thin antenna that you can place close to your leaking box and pick up more energy than getting a larger antenna away from the box.

Placing your box inside another metal box is a good way to "pseudo amplify" any leakage from your box. Place your hand-made antenna sniffer inside the home made metal box and you can measure leakage more readily.

Precompliance testing is done to assure, to a reasonable level, that you will pass the EMI leakage testing if you send the unit to a calibrated test lab (and spend lots of money doing so). So you not only need an antenna that covers your frequency range, you need one that is "calibrated"--usually meaning you have to buy a new one. Ones you find on ebay are sometimes junk, and their initial published calibrations are no longer to be trusted.

Here are some ideas:
3103 conical log spiral, or 3143B Biconilogic antenna.

http://www.ets-lindgren.com/DefenseAntennas

If you do not have the bux to buy one of these, I would get a cheapo discone antenna, like:
http://www.ebay.com/itm/DISCONE-BROAD-BAND-ANTENNA...

And when you find a spur that seems close to failing your leakage level, double check the level with a quarterwave monopole over a ground plane (sma connector soldered to a FR4 pcb ground plane in the middle, and a length of stiff wire long enough to resonate at that one frequency. You know the theoretical gain of such an antenna.

www.MaguffinMicrowave.com

Maguffin Microwave wireless design consulting