The data from a broad spectrum of investigational techniques strongly and consistently indicate that hydrogen can exist in lower-energy states than previously thought possible. The predicted reaction involves a resonant, nonradiative energy transfer from otherwise stable atomic hydrogen to a catalyst capable of accepting the energy. The product is H(1/p), fractional Rydberg states of atomic hydrogen wherein n = 1/2, 1/3, 1/4,..., 1/p (p <= 137 is an integer) replaces the well-known parameter n = integer in the Rydberg equation for hydrogen excited states. He+, Ar+, and K are predicted to serve as catalysts since they meet the catalyst criterion-a chemical or physical process with an enthalpy change equal to an integer multiple of the potential energy of atomic hydrogen, 27.2 eV. Specific predictions based on closed-form equations for energy levels were tested. For example, two H(1/p) may react to form H-2(1/p) that have vibrational and rotational energies that are p(2) times those of H-2 comprising uncatalyzed atomic hydrogen. Rotational lines were observed in the 145-300 nm region from atmospheric pressure electron-beam-xcited argon-hydrogen plasmas. The unprecedented energy spacing of 4(2) times that of hydrogen established the internuclear distance as 1/4 that of H-2 and identified H-2(1/4). The predicted products of alkali catalyst K are H- (1/4) which form KH*X, a novel alkali halido (X) hydride compound, and H-2 (1/4) which may be trapped in the crystal. The H-1 MAS NMR spectrum of novel compound KH*Cl relative to external tetramethylsilane (TMS) showed a large distinct upfield resonance at -4.4 ppm corresponding to an absolute resonance shift of -35.9 ppm that matched the theoretical prediction of H(1/4) with p = 4. The predicted frequencies of ortho- and para-H-2 (1/4) were observed at 1943 and 2012 cm(-1) in the high-resolution-FTIR spectrum of KH*I having a -4.6 ppm NMR peak assigned to H- (1/4). The 1943/2012 cm(-1)-intensity ratio matched the characteristic ortho-to-pata-peak-intensity ratio of 3:1, and the ortho-para splitting of 69 cm(-1) matched that predicted. KH*Cl having H-(1/4) by NMR was incident to the 12.5 keV electron beam which excited similar emission of interstitial H-2 (1/4) as observed in the argon-hydrogen plasma. KNO3 and Raney nickel were used as a source of K catalyst and atomic hydrogen, respectively, to produce the corresponding exothermic reaction. The energy balance was Delta H = -17 925 kcal/mol KNO3, about 300 times that expected for the most energetic known chemistry of KNO3, and -3585 kcal/mol H-2, over 60 times the hypothetical maximum enthalpy of -57.8 kcal/mol H-2 due to combustion of hydrogen with atmospheric oxygen, assuming the maximum possible H-2 inventory. The reduction of KNO3 to water, potassium metal, and NH3 calculated from the heats of formation only releases -14.2 kcal/mol H-2 which cannot account for the observed heat; nor can hydrogen combustion. But, the results are consistent with the formation of H-(1/4) and H-2(1/4) having enthalpies of formation of over 100 times that of combustion. C) 2007 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.