See my comments below. The only error is that the OP reversed the symbol for the battery. Look at the breadboard; he/she built it with the power supply correctl

See my comments below. The only error is that the OP reversed the symbol for the battery. Look at the breadboard; he/she built it with the power supply correctly installed. You're probably right about the datasheet for the opto, but that neither clashes with my statement ("most optocouplers") nor invalidates the broader point (Vcesat is ~300mV; the diode is most definitely off). Aside: I spent several years as the lead engineer on a line of digital isolators (modern replacements for optocouplers: CMOS integrated circuits with much higher speed, better immunity to common-mode transients, and no need to use expensive and generally boutique processes supporting LEDs). Most of the time, opto manufacturers oversell the capabilities of their optos at low voltage, and generally speaking in industrial applications optocouplers are driven much harder than the currents at which forward drop numbers are quoted in datasheets. Aside #2: bipolar transistors absolutely do operate upside-down! Reverse beta is usually far worse than forward beta, because the collector region tends to be much larger than the emitter and as a result there is substantial carrier recombination. In fact, when you want to use a bipolar transistor as a low drop switch, and if you don't care how much base current you have to use to do it, running the transistor upside-down tends to result in lower drop and faster operation. There are a few reasons for this, but it boils down to this: the speed of the bipolar transistor as a switch is limited by charge storage in the base when the device is saturated, or more precisely, the amount of time it takes to remove this charge when trying to turn the device off. The base-collector diode behaves differently than the base-emitter diode because of the difference in emitter and collector doping. As a result, when upside-down, the base-collector junction stores much less charge, the lower beta actually means that the charge is removed faster, and the higher forward drop of the base-emitter diode means that reverse Vcesat is lower than the forward Vcesat. The downside, to which I alluded above, is 10x or more base current to actually drive the transistor. Depending on the quantum efficiency of this phototransistor and its reverse beta, were the OP to use the phototransistor upside-down in this circuit, it might not be able to shunt sufficient current to keep the optoisolator off.