Asymmetric Transport Efficiencies of Positive and Negative Ion Defects in Amorphous Ice Eui-seong Moon, Youngsoon Kim, Sunghwan Shin, and Heon Kang * Department of Chemistry, Seoul National University, 1 Gwanak-ro, Seoul 151-747, Republic of Korea (Received 3 February 2012; published 29 May 2012) Hydronium (H 3 O þ ) ions at an ice surface penetrate into its interior over a substantially longer distance than hydroxide (OH  ) ions. The observation was made by conducting reactive ion scattering and infrared spectroscopic measurements for the acid-base reaction between surface H 3 O þ (or OH  ) and NH 3 (or NH 4 þ ) trapped inside an amorphous ice film at low temperature (< 100 K). The study reveals very different transport efficiencies of positive and negative ion defects in ice. This difference is explained by the occurrence of an efficient proton-relay channel for H 3 O þ , which does not exist for OH  . DOI: 10.1103/PhysRevLett.108.226103 PACS numbers: 68.49.Sf, 42.68.Ge, 66.30.jp, 72.20.Jv Ice is a unique type of solid conductor containing hydro- nium (H 3 O þ ) and hydroxide (OH  ) ions as intrinsic charge carriers. These species exist as positive and nega- tive ion defects, respectively, in the ice lattice where they are in proton-transfer equilibrium with constituting water molecules. It is generally conceived that the coexistence of positive and negative ionic defects is deeply related to certain unusual electrical properties of ice, including the thermoelectric effect [1] and the freezing potential (the Workman-Reynolds effect) [2,3]. The properties of ion defects in ice are an appealing subject of study not only from the perspective of ice physics, but also in a broad range of disciplines in physics, environmental sciences, astrophysics, and chemistry owing to the ubiquity and importance of ice in natural environments. Numerous studies have been performed to understand the charge conduction phenomena of ice [4–13]. The ex- perimental studies range from classical ice conductivity measurements [4] to modern molecular spectroscopic mea- surements that monitor the H/D isotopic exchange kinetics and proton-transfer dynamics [5–11]. The theory of charge conduction in terms of ice molecular structure was ad- vanced by Jaccard [12] and Onsager and Dupuis [13] long ago. Owing to these investigations, the transport mechanism of H 3 O þ in ice is now well understood. H 3 O þ moves via a sequence of proton transfers along the hydrogen bond chain of water (Grotthuss mechanism), and this process may occur even at low temperature. At high temperature, the proton-transfer relay is coupled with Bjerrum defect motion to extend the distance of the proton transfers [5–7]. However, a complete picture for charge conduction in ice will be attainable only when the trans- ports of both positive and negative ion defects are properly understood. The present study explores some open ques- tions related to this issue: What is the relative efficiency of charge transport by positive and negative ion defects? Do they move like ‘‘mirror images’’ in ice with only the proton-transfer directions reversed, or via intrinsically dif- ferent molecular mechanisms? To answer these questions, we have measured the transport distances of H 3 O þ and OH  through an ice film by placing H 3 O þ (or OH  ) at the ice film surface and its counter base (or acid) in the ice interior. The study shows that H 3 O þ travels a substantially longer distance than OH  due to the occurrence of an efficient proton-relay mechanism for H 3 O þ , which does not exist for OH  . We conducted the experiment in an ultrahigh vacuum chamber [14] equipped with instrumentation for reactive ion scattering (RIS), low-energy sputtering (LES), reflection absorption infrared spectroscopy (RAIRS), and temperature programmed desorption (TPD). An H 2 O ice film was grown typically to a thickness of 50 BL (bilayer; 1 BL ¼ 1:14  10 15 water molecules cm 2 ) on a Ru(0001) crystal by using a back-filling method at a slow (  0:1 BL s 1 ) deposition rate. The Ru substrate temperature was main- tained at 70–100 K, which resulted in an amorphous ice film growth [15]. The thickness of the ice film was estimated by performing TPD measurements [16]. HCl and NH 3 vapors were introduced into the chamber through separate leak valves and guided close to the sample surface through tube dosers. Cs atoms were deposited onto the sample by using an alkali metal dispenser (SAES Getters). The chemical species present on the ice film surface were measured by performing RIS, LES, and RAIRS analyses. In the RIS and LES experiments, a Cs þ beam from a low-energy ion gun (Kimball Physics) collided with the sample surface at the incident energy of 30 eV, and the ions emitted from the surface were detected by a quadru- pole mass spectrometer (Extrel) with its ionizer filament switched off. In RIS, neutral species (X) on the surface are picked up by the scattering of Cs þ projectiles to form Cs þ -neutral clusters (CsX þ ). In LES, preformed ionic species (Y þ and Z  ) on the surface are ejected by the Cs þ impact. Thus, RIS and LES signals reveal the identi- ties of neutral (X) and ionic species (Y þ and Z  ), respec- tively, on the surface [14,17]. The probing depth for LES and RIS methods is 1 BL of the ice surface at the employed Cs þ beam energy [17]. Chemical species in the interior of the ice samples were monitored with RAIRS [18,19]. PRL 108, 226103 (2012) PHYSICAL REVIEW LETTERS week ending 1 JUNE 2012 0031-9007= 12=108(22)=226103(5) 226103-1 Ó 2012 American Physical Society