To charge negatively, the teflon tube that is standard on a tribo gun will have to be replaced by a tube made with an electron donor, or else corona charging used instead of tribocharging.
I calculate that to rise in the atmospheric electric field, a particle needs a charge to mass ratio ("specific charge") greater than 50 millicoulomb/kilogram, which may be another factor requiring corona charging.

I guess to get started we will need a telescopic laser range finder to find out how high the particles actually go.

What about building some macroscopic device that uses sky power? In principle, it could fly if it could exceed a specific charge of 50 mC/kg, with payload. Maybe build it to double as a single, large corner-cube reflector pointing down, to facilitate ground-based tracking. Such a form would be mechanically stable while ascending because it places the center of charge above the widest point. If tracking is done at radar wavelengths, the device could be made of an open honeycomb lattice to reduce weight and air resistance, because relatively long wavelengths do not have the resolution necessary to "see" the holes.

The figure of 50 mC/kg was derived by dividing g, the gravitational acceleration at the Earth's surface (about 10 m/s2), by 200 V/m, and multiplying by 1000 to get the units used in studies of powder-coating physics (and the units analysis checks out).

Extrapolating from data in Meng et al., 2008, http://dx.doi.org/10.1088/0022-3727/41/19/195207 , 2.3-micron-diameter sulfur particles corona charged at 90 kV should fly. However, a ten-fold smaller sulfur particle will have a ten-fold greater specific charge, giving some margin to allow for discharging on the way up.

The diameter of the sulfur particle injected into the stratosphere is unrelated to the diameter of the eventual sulfuric acid droplets it produces upon oxidation in the stratosphere, because one reaction intermediate, sulfur dioxide, is gaseous (complication: it’s also a greenhouse gas).

At this time, my best guess as to how fast the particles would rise is 3 cm/s (because I believe I have seen it), which will take them up to the stratosphere in four to five days.

However, thus far, my calculations have not addressed the fact that the sky electric field weakens with height. At an altitude of 12 km, it is only 5 V/m, versus 100-200 V/m at sea level. The altitude effect will cause the particles to stop ascending and start concentrating at a particular altitude  (a possibly useful effect) where gravitational and coulombic forces are in equilibrium, but is it stratospheric? Unfortunately, no. Even reducing particle diameter 10-fold to 0.23 microns (which uses up our margin for discharge) only gives 4.5 km, less than the minimum height of the stratosphere, 8 km. So, we don't get there, unless we stand on a mountain top in Greenland, but we get interestingly close with what is only the first scheme contemplated. My source for the dependence of electric field on height is figure 20-7a in https://www.ngdc.noaa.gov/stp/space-weather/online-publications/miscellaneous/afrl_publications/handbook_1985/Chptr20.pdf

The weakening of the Earth’s gravity with height is no help, because if you go up to 12 km, the difference is only one-half of one percent.

The problem of particles discharging en route is far from trivial, given that the charge relaxation time constant of air is 15 min at sea level and a tenth of a second at the top of the stratosphere. This is the time required for a 2.7-fold reduction in particle charge (assuming that the particle itself is air). An encapsulated charge structure therefore appears essential. The ideal appears to be a charged particle of supercapacitor dielectric, for example, CCTO, coated in teflon, but this takes us far afield from beautifully simple sulfur aerosols. However, could this be the recipe for the coulombic drone suggested above?

Can we relax some constraints here, given the anticipated economies of a coulombic hoist system? Does injection have to be stratospheric, or will high tropospheric do? If so, how high? Do the light-scattering particles have to be sulfuric acid or can they be electrified mineral dust, pollen, or sea salt?

A further complication is that atmospheric aerosols scavenge ions from the air, which increases the resistivity of the air. By Ohm’s law, this effect will increase the intensity of the atmospheric electric field. This effect in isolation will assist the coulombic hoist, but will probably also reduce the specific charge on the particles, which will act counter to the coulombic hoist process. I am now wondering what the net effect of ion scavenging would be, and the net effect may differ between dispersed and concentrated particle releases. If necessary, the benefits of the scavenging effect could be made more robust by initially releasing heavily charged natural aerosols such as mineral dusts or sea salt to run interference for a subsequent release of a charged sulfur aerosol. The mutual electrostatic repulsion of the released particles will cause them to quickly spread laterally without limit as they rise, which may be what finally kills the scavenging idea dead.

However, other manipulations can be imagined, such as maximizing the winter snowpack in Canada and Russia by seeding supercooled clouds with ice-nucleating proteins isolated from pseudomonas syringae and a few other species of bacteria.

Our task is not to build a cooling system, but a control system having enough power to overwhelm any heating or cooling positive feedbacks that may set in as the result of overshoots or undershoots in temperature control. Focusing narrowly on cooling will trigger a continental glaciation sooner or later.