Duct flows constitute an important category of modern thermal engineering. Optimizing efficiency has become a significant objective in the 21st century in, for example, heating ventilation and air-conditioning (HVAC), coolant or heat transfer fluid flows in a nuclear power reactor, heat exchanger design etc, and this has been achieved by either new materials (improved thermal insulation properties) constituting the duct walls, novel geometric designs or improved working fluids. Nanotechnology has infiltrated into duct design in parallel with many other fields of mechanical, medical and energy engineering. Motivated by the excellent potential of nanofluids, a subset of materials engineered at the nanoscale, in the present work, a new mathematical model is developed for natural convection in a rectangular vertical duct containing nanofluid. e left and right walls of the duct are maintained at constant and unequal temperatures, while the front and rear walls of the duct are insulated. Thermo-solutal (double-diffusive) natural convection of aqueous nanofluid containing various metallic nanoparticles (e. g. copper, titanium oxide) or carbon-based nanoparticles (e. g. diamond, silicon oxide) is simulated. The Tiwari-Das nanoscale volume fraction model is used in addition to the Brinkman and Maxwell models for defining the properties of the nanofluid. The partial differential conservation equations for mass, momentum and energy are non-dimensionalized via appropriate transformations and the resulting boundary value problem is solved with a second-order accurate implicit finite difference technique employing Southwell-Over-Relaxation (SOR). esh independence tests are conducted. Extensive visualization of the solutions for velocity, temperature, nanoparticle concentration (volume fraction) are presented for five different nanoparticles (silicon oxide, diamond, copper, titanium oxide and silver), thermal Grashof number, nanoparticle species (solutal) Grashof number, volume fraction of nanoparticles (i.e. percentage doping), Dufour number, Soret number, Prandtl number, Schmidt number and duct aspect ratio.