Abstract
Nitrous oxide has been used as a specific scavenger of aquated electrons produced by photo-detachment from Fe$^{2+}_{aq.}$(\rightarrow$^{hv}$Fe$^{3+}_{aq.}$ + e$^-_{aq.}$) and Fe(CN)$^{4-}_6$(\rightarrow$^{hv}$Fe(CN)$^{3-}_6$ + e$^-_{aq.}$) by light of wavelength 2537 $\overset{\circ}{\mathrm A}$. In each case $\phi$(N$_2$) increases to a limiting value as [N$_2$O] is increased; it is 0.07 for the former process and 0.66 for the latter. Use of isopropanol instead of N$_2$O in the latter case indicated that no H atoms are formed in the primary act. Competition experiments indicate that $k(e^-_{aq.} + H^+)/k(e^-_{aq.} + N_2O) = 1.86 \pm 0.12$ when Fe$^{2+}_{aq.}$ is used as the electron source and 1.55 \pm 0.1 when lanthanum ferrocyanide is used. For the K$_4$Fe(CN)$_6$ solutions, 2$\phi$(N$_2$) = $\phi$(Fe(CN)$^{3-}_6$) unless KCN is present when $$OH + CN^- \rightarrow OH^- + CN$$ and $$Fe(CN)^{3-}_6 + CN \rightarrow C_2N_2 + Fe(CN)_5H_2O^{3-}(+CN^- \rightarrow Fe(CN)^{4-}_6 + H_2O)$$ occur. The slopes of the Bronsted-Bjerrum plot for log$_{10}$k(e$^-_{aq.}$ + NO$^-_2$) is much larger (2 to 2.7) than expected or is obtained from radiation chemical experiments and is explained by a model in which e$^-_{aq.}$ reacts with NO$^-_2$ while still within the relaxed ion atmosphere of the ferrocyanide ion. Other experimental and theoretical evidence is added in support of this hypothesis.