If you can't find any simple wall jacks that meet the requirements, you could always just install an ethernet switch and use the connectors on the switch. The only drawbacks are that you will need to have power available for the switch, and if you are using certain flavors of Ethernet/IP or Profinet that require IEEE 1588 time protocol compatibility for high-resolution control, those protocols require that the switches be IEEE 1588 compatible (many aren't). There are several companies with hazloc listed ethernet switches such as Moxa or Advantech. Additionally, a few companies have intrinsic safety barrier/hubs for ethernet that can convert to IS, allowing ordinary jacks to be used after the barrier.
P&F has one for example:
http://www.pepperl-fuchs.com/cps/rde/xchg/global/hs.xsl/1815.htm?rdeLocale=en
I haven't used it myself, but P&F isn't known for making stuff that doesn't work.
One other idea that I'm not sure would fly based on NEC requirements, but others might be able to comment on is to simply have a captive cable come out from the wall. Do not terminate the cable in a wall plate jack, just bring 30 feet or however long of cable right out from the wall and terminate it in a plug. The cable would sit unconnected unless plugged into a device, and at that point the device would need the listing. Obviously this assumes that the far end of the cable is plugged into a C1D2 listed hub/switch within the same div/zone, and that the area is dry and relatively clean. I can't remember what NEC says about an unterminated low voltage cable sitting idle, but I would think it is safer than a wall jack that could arc/spark when a cable is plugged/unplugged.