The large square panels are deployable radiators. They provide extra heat rejection capability in addition to the fixed radiators that can also be seen in the photographs.
Context part 1 - Thermal design for Geo Comsats
Most geostationary communications satellites are roughly cuboid in shape with the solar arrays and antennas as appendages. The basic thermal design of a satellite of this type is to put insulating blankets on the four sides that face the sun directly during the normal day. In contrast the "North" and "South" panels are usually made of aluminium honeycomb, they are painted black on the inside and have second surface mirrors on the outside. The bulk of the heat dumped by the satelite's electronics then takes this slightly easier route to deep space than going through the blankets. The honeycomb panels usually have embedded or bolted on heat pipes in the inside face in order to spread heat as evenly as possible through the panel so as to prevent local hotspots. The orientation of the satellite to the Sun changes over the year and so at sometimes the sun can be up to 23.5 deg from edge-on. The mirrors help by rejecting optical sunlight.
Context part 2 - Second Surface Mirrors
A second surface mirror, also called an optical solar reflector (OSR), is similar in concept to an ordinary domestic mirrors. The mirror comprises a thin layer of quartz or glass with an aluminized rear surface. This allows them to have good optical reflectivity, $\epsilon_{vis}<0.1$, and high IR emissivity, $\epsilon_{IR}>=0.8$, (as the outer surface is glass).
Deployed radiators
The cuboid shape isn't universal but it is convenient for structural design and packaging into the launch vehicle fairing whilst providing some heat rejection capability. This leads to a potential problem with particularly high power satellites. As far as I am aware the Boeing 702 platform is the only one in regular production that has taken the design solution of creating additional radiator aread for cooling by deploying panels once in-orbit. Clearly, if the heat pipes could spread heat over both of the additional panels, doubling the overall area, then it would greatly increase the possible DC power of the satellite by raising the heat rejection limits. The photographs aren't especially clear though it appears that some of the hardware showing on the reverse, black painted, side of the panel could be a network of heat pipes for this purpose.
Moving further into assumption territory, I assume that the reason that the deployed panels are shown black on one side is to keep a modest environment for the antenna related equipment attached to the East-West faces.
One feature that isn't so clear to me is whether the panels are left fixed after deployment. Clearly there would be a twice daily opportunity to adjust the angle of the radiators so that they shadow the main body panel though my outsider's view is that it may not be worth the extra complexity.
Context part 3 - Heat pipes
Of the variety of things going by the name of "heat pipe" I'm only aware of two types used commonly in satellites:
i) rigid narrow tubes lined with narrow grooves. These wick the working fluid in either direction and it will even out hot spots where ever they form. Their capacity is often limited and they can be zero-g only. The grooves have to be present all the way along the tube to prevent drying out.
ii) loop heat pipes. These have a preferred direction of flow and I think use a slighty different thermodynamic principle (not my territory, that's as far as I'm going!). As I understand it they have a much higher pumping capacity than the normal type and that this enables it to incorporate non-grooved flexible sections. Assuming I have got it right, this latter feature makes them suitable for deployable radiators.
As an aside, in the description of heat pipes I have excluded mechanically pumped fluid loops.
. . 
