That would tie in with what's been discovered so far. The 2 pins are clearly the emitter. The OP suggested that applying a voltage across the 2 pins associated with the emitter (see post #20) that they could see the LED light up. The OP didn't say if they had a current limiting resistor to the emitter. Hopefully their particular device uses a visible LED, and the "lighting up" wasn't the demise of the IR LED!
That also ties in with what we've been saying, that the sensor is a simple sensor that simply provides pulses out in response to the beam being interrupted.
What the OP is yet to determine is which of the 4 pins on the photodiode side is supply, GND and signal out. It may be of benefit for the OP to draw out how those discrete components around the sensor connect to it.