Photophosphorylation is ATP formation driven by a proton-motive force generated by photosynthetic electron transfer across the thylakoid membrane.

Linear electron flow builds that gradient in several ways:

• water oxidation by PSII releases protons into the thylakoid lumen;
• electron transfer through the plastoquinone/cytochrome b6f region moves additional proton equivalents toward the lumen;
• reduction of NADP$^+$ to NADPH consumes reducing equivalents on the stromal side.

Together these processes make the lumen relatively proton-rich compared with the stroma.

ATP synthase uses proton return

Thylakoid ATP synthase allows protons to move downhill from lumen to stroma. That favorable electrochemical movement is coupled to

$$\mathrm{ADP+P_i\rightarrow ATP+H_2O}.$$

ATP is therefore produced on the stromal side, where it can be used by carbon-fixation reactions.

The energy-conversion chain is

light-driven electron flow
        ↓
H+ accumulation in thylakoid lumen
        ↓
proton-motive force
        ↓
H+ return through ATP synthase
        ↓
ATP made in stroma

Light does not phosphorylate ADP directly

A photon excites a photosystem; photosynthetic electron flow then builds an electrochemical gradient; ATP synthase converts that gradient into chemical free energy.

Thus

$$\text{light}\not\rightarrow\text{ATP directly}.$$

The membrane gradient is the intermediate energy store.

This is the same chemiosmotic principle used in respiration, but the upstream energy source differs: mitochondria commonly build the gradient using favorable oxidation of reduced fuel carriers, whereas chloroplasts use photon-driven electron transfer.