T h e oxidation o f butane is considerably affected b y the nature o f the vessel surface. W i t h a silica vessel, vigorous cornbustion acconlpanied b y multiple cool flarnes occurs between
INTRODUCTIONThere is considerable evidence that the oxidation of hydrocarbons is subject to surface effects. ICinetic experimeilts have revealed a dependence of rate on vessel diameter (1, 2) and on the previous treatment of the vessel walls (3, 4, 5). Although systematic investigations of surface effects have been made with hydrogen (6, 7, 8) ancl methane (4, 9) there is a lack of information about the role of the surface in the oxidatioil of the higher hydrocarbons, where mechanisms not foui~cl with inethaile come into play (10, 11). A better understanding of the part played by the surface in these reactions is desirable for several reasons: it would clarify some of the puzzling anoinalies in the reaction lcinetics (12, 13); it might suggest improvements in petrochemical processes; finally, it might help reveal the mechanism of "ltnoclt" and antiknock action in engines.Although l c n~~l e and its suppression have been subjected to much investigation (14, 15, 16) there is still considerable disagreement about the nature of the process. T h e work of Icing and co-workers (17, 18, 19) and of Walsh and co-worlcers (20, 21, 22) inay be cited as illustrating both the lack of agreement 011 mechanism and the possible iinportance of heterogeneous processes.ICing (17, 18) ascribes the action of antiknoclc agents such as iron carbony1 or lead tetraethyl to depositioil of metals or metal oxides ~vhich are believed to proillote the formation of steam and oxides of carbon; these gases inay then dilute the con~bustible mixture and reduce its inflaminability. 21,22) ascribe the antiltnock action of lead tetraethyl to the formation of a colloidal smolce of lead monoxide, which is assumed to inhibit the lcnoclciilg reaction by virtue of a surface destructioil of free radical chains. Each group of investigators claims supporting evidence from combustion experiments. The positive-catalyst theory favored by King is said to be supported by the results of flow experiments with fuels such as pentane and hyclrogen (18, 19). The negative-catalyst theory favored by Walsh finds support froin static experimellts wit11 fuels such as methane and ethers (21,22).The question of antiknock action has continued to be a contentious one (IG,23,24). I t appeared desirable to establish empirically the effects of various inorganic compounds on the coinbustion behavior of a typical hydrocarbon. For this investigation the fuel chose11 was n-butane, which oxidizes a t collvenient rates between 250' C and 500' C. This compound's oxidation is accompanied by cool flames, two-stage ignition, and other phenomena associated with the lorn-temperature combustion of paraffin hj7drocarboils and related substances (11,25).