Any theories describing Early Earth conditions will necessarily be speculative, since no human was around then. These speculations may be reasonable judgements, but some may be less reasonable than others. Having more than one theoretical possibility allows for a wider experimental approach in understanding the Earth, and how life might have originated. To limit our views to one particular theory would be biased. As such, there are two possibilties as the the early atmospheric composition of Earth: highly reducing, or less reducing. Each view has different implications as to the generation of organic molecules that would have been necessary for life to develop.
A highly reducing atmosphere is one that is predominately composed of molecules (i.e. reducing agents) that readily give up or "donate" electrons to other molecules, thereby "reducing" them. To clarify this, here's an example. A reducing agent, X, is initially reduced and it loses electrons to molecule Y. Molecule Y is initially oxidized, and it gains electrons from reducing agent X. So Y becomes reduced and X becomes oxidized. In a less reducing atmosphere, there are less reducing agents, and more molecules are oxidized, or there are oxidizing agents. At present, Earth's atmosphere is less reducing, since it is made up of 21% oxygen. Oxygen is a highly reactive molecule, so how can it persist in such a high percentage of the atmosphere? I will come to this later.
The view that the early Earth atmosphere was initially highly reducing (and later became less reducing), had been worked out in the 1950s by Urey, and others. Here is the scenario. Early on in Earth's history, there was a lot of volcanic activity (much more than now). Volcanoes initially released gases such as carbon monoxide (CO), hydrogen gas (H2), nitrogen gas (N2) or ammonia (NH3). So the early atmosphere was highly reducing initially. Where did all these gases come from? They came from deep within the Earth, and as such, the gases had to travel through the mantle (the mantle is the region from the Earth's core to the surface). The mantle is made primarily of iron (Fe). Iron comes in three valence states: Fe0, Fe3+ and Fe2+. Early on the mantle was mostly Fe0 (~85%) and very little Fe3+, with the rest as Fe2+. So the ratio of Fe3+ to Fe2+ in the mantle was zero early on. As gases (e.g. H2) moved through the mantle, they would pick up electrons from Fe2+ and become reduced, while Fe2+ would convert to Fe3+. The reduced gases would be less effective at being reducing agents. Since hydrogen gas is very light, it is eventually lost into space. Thus, the atmosphere loses its reducing power over time (since it loses gases that can become reducing agents). Now, at around the same time, oxygen gas (O2) is accumulating in the atmosphere. Remember that O2 is highly reactive, and in the presence of H2, water (H20) is formed. So, the loss of hydrogen gas to space is blocked by an increase in oxygen gas, and water accumulates. Eventually, the hydrogen gas in the Earth's core and mantle runs out. As a result, volcanoes release only CO2 and N2. Nitrogen gas happens to be inert, and is ~79% of Earth's present atmospheric composition. Water can only exist in a stable liquid state when the atmospheric pressure and temperature are in a certain range. The early Earth was initially very hot (and had a very dry surface), but it cooled over time allowing water vapor to condense. The presence of Nitrogen gas in the atmosphere provided sufficient atmospheric pressure for liquid water to exist stably. Ultimately, liquid water became abundant and CO2 emitted from volcanoes allowed for carbon cycling. The significance of these will become clear later.
In 1953, Stanley Miller was the first to show that complex organic molecules could be made from simpler components. In his experiments, he used methane (CH4), ammonia (NH3), H2) and H2 as substrates in a glass apparatus. Miller boiled the system over time, and also used electric discharges as an energy source. Over time the mixture changed color. So, Miller used a chromatograph to analyze the mixture for it's constituents. He discovered that certain amino acids had been formed. Amino acids are the building blocks for proteins, so this discovery that amino acids could be made from much simpler molecules (and within a relatively short time period) pointed to a chemical evolution before life could arise. Later, Miller and Urey used UV light as an energy source with a similar system, and they found that hydrogen cyanide (HCN) could be formed from simpler molecules under reducing conditions. Now, why is the generation of hydrogen cyanide important? Isn't it a toxic material? Yes it is, but conditions in prebiotic times were much different than now. Hydrogen cyanide has been found to react with other HCN molecules and can eventually form adenine -- one of the bases in the genetic code (overall, it takes five molecules of HCN to form one molecule of adenine under UV light). Many other experiments have been since carried out, under slightly different conditions, and the final reaction products end up being various amino acids and sugars, as well as nucleotides. These experiments are significant, in that they demonstrate possible reactions that could have been present during pre-biotic times.
Chemical evolution (during pre-biotic times) began with simple molecules reacting with others, eventually forming more and more complex molecules. Simple chemical reactions can only go so far by themselves. Some form of energy is needed to drive unfavorable reactions, or to aid in driving favorable reactions. There are many possible energy sources that could have been available on early Earth. Solar energy (UV light) is the most obvious source, but isn't UV light damaging to cells? On early Earth, there would have been no ozone layer to protect cells. UV splits an O2 molecule (photolysis), which forms oxygen radicals that can react with other O2 molecules to form ozone (O3). There wouldn't have been enough O2 early on to provide a protective ozone layer. However, an ozone layer would not have been needed in prebiotic times. UV light not only splits oxygen gas molecules, but can split other molecules as well, and so can drive unfavorable chemical reactions. UV would have been quite abundant early on, providing an energy source during chemical evolution. There are other sources of energy, such as electric discharge (i.e. lightening) and nuclear energy (e.g. uranium-238). Today we know that lightening can fix nitrogen gas into a solid form. In Africa, there have been found six natural nuclear reactors -- i.e. deposits of uranimu-238. Remember that man-made nuclear reactors (or bombs) generate enormous amounts of energy. Natural reactors would not have been as powerful in the short term, but could have provided sufficient energy over time to drive chemical reactions. One thing to note about nuclear reactions, is that they require water (a nuclear reaction then, is a "fire" that needs water to drive the reaction). This is significant, in that a water medium can bring in new materials to the reaction. As with UV radiation, nuclear radiation ionizes molecules -- H2O and other molecules would be ionized by a natural nuclear reactor, which could then go on to react with other molecules. Nuclear radiation is also mutagenic, so it would have generated variation in an early genetic system.
There is another source that can generate organic molecules on Earth -- metorites and dust particles from space. Interplanetary Dust Particles (IDPs) are constantly bombarding the planet, and bring in about 3.2x10^5 kg of organics per year in modern times. Meteors large enough survive atmospheric entry to the planet can carry in organic molecules. In some cases, a meteor may heat up internally upon entry to such a high degree that they explode as an airburst, creating a shock wave. A relatively new field, "Shock Chemistry", has shown that organic molecules (e.g. amino acids) can be generated from such shock waves. Large bolloid impacts, such as the one responsible for the K/T boundary, can generate vast amounts of organics in the short term.
Now, remember that two types of atmosphere have been postulated for early Earth: highly reducing, and less reducing. It turns out that the significance of these various energy sources in prebiotic times, is dependent on the type of atmosphere that was present. In a highly reducing atmosphere, about 97% of all organics would be made by photolysis (i.e. by UV light). This means that everything else is less significant, accounting for ~3% of the rest of organics made. It gets much more interesting in a less reducing atmosphere, where photolysis accounts for only 41% of organics, IDPs providing 27%, K/T impacts 27%, electric discharge 4%, and the rest at trace amounts.
Our main problem is the lack of knowledge of early Earth conditions. In a highly reducing atmosphere, only photolysis seems to be significant, whereas in a less reducing atmosphere, there are many more possible energy sources for making organics. One thing is certain: Earth's current atmosphere is "less reducing" and in fact, it is oxidizing. If Earth started out with a highly reducing atmosphere, it eventually became less reducing. If Earth initially had a less reducing atmosphere, it became even less reducing. Either way, Earth now has an oxidizing atmosphere. At some point, the more interesting scenario (i.e. the energy sources in the less reducing atmosphere) must have occurred, adding organics to what had already been generated -- regardless of Earth's initial atmospheric conditions.