No pump or power is used here — the sample wicks through the porous strip by capillary action alone, the same way a paper towel draws up a spill. The liquid front advances quickly at first, then slows as it goes (front position grows roughly with √t), because the pores ahead offer more resistance the farther the fluid has already travelled.
The sample first crosses the conjugate pad, which is pre-dried with colloidal gold nanoparticles (~20–40 nm) coated in detection antibodies. Each gold particle scatters light red because its free surface electrons collectively oscillate with visible light — a surface plasmon resonance that absorbs green wavelengths and reflects red, giving every particle its own microscopic red tint even before any binding happens.
If the target antigen is in the sample, it binds the gold-antibody conjugates as they flow, and that antigen-carrying gold gets captured by a strip of immobilized capture antibody at the test (T) line, forming an antibody-antigen-antibody "sandwich". Enough captured particles pile up in one place for their individual red tints to merge into a visible red band. The control (C) line further downstream always captures any remaining free conjugate, regardless of the sample — a red C line is what confirms the strip itself worked and the liquid actually flowed.
- Positive sample — target antigen present, so nanoparticles are captured at both T and C lines.
- Negative sample — no antigen to bridge the sandwich, so nanoparticles pass the T line unbound and are only captured at C.
- No line at all at C means the test is invalid, however the T line looks, because the strip failed to wick properly.