Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?

We present FTIR studies of the 193 nm photolysis of fully deuterated formic acid (DCOOD) isolated in solid parahydrogen at 1.9 K which show evidence of transient HDO rovibrational satellite peaks. The S1 and S2 satellite peaks are readily detected for α-type (1₀₁ ← 0₀₀) rovibrational transitions...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:2012
Автори: Wonderly, W.R., Anderson, D.T.
Формат: Стаття
Мова:English
Опубліковано: Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України 2012
Назва видання:Физика низких температур
Теми:
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/117419
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies? / W.R. Wonderly , D.T. Anderson // Физика низких температур. — 2012. — Т. 38, № 8. — С. 853-859. — Бібліогр.: 31 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-117419
record_format dspace
spelling nasplib_isofts_kiev_ua-123456789-1174192025-02-09T09:45:34Z Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies? Wonderly, W.R. Anderson, D.T. Low Temperature Spectroscopy and Radiation Effects We present FTIR studies of the 193 nm photolysis of fully deuterated formic acid (DCOOD) isolated in solid parahydrogen at 1.9 K which show evidence of transient HDO rovibrational satellite peaks. The S1 and S2 satellite peaks are readily detected for α-type (1₀₁ ← 0₀₀) rovibrational transitions of HDO either during or immediately after photolysis. Intensity measurements show the HDO b-type (1₁₁ ← 0₀₀) rovibrational transitions have satellite peaks as well, but due to the greater linewidth of these absorptions, the satellite peaks cannot be spectroscopically resolved from the monomer transition and are therefore difficult to detect. These newly identified HDO satellite peaks may result from the HDO photoproduct being formed next to an H atom or a vacancy in the parahydrogen solid. The development of the infrared spectroscopy of these satellite peaks can provide a new means to study radiation effects on low-temperature hydrogen solids doped with chemical species. W.R.W. was supported partially by the Research Experiences for Undergraduates (REU) Program of the National Science Foundation under Award Number CHE 08-51931. The authors thank the National Science Foundation for its generous support through Grant CHE 08-48330. 2012 Article Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies? / W.R. Wonderly , D.T. Anderson // Физика низких температур. — 2012. — Т. 38, № 8. — С. 853-859. — Бібліогр.: 31 назв. — англ. 0132-6414 PACS: 61.80.–x, 67.80.f, 67.80.dj https://nasplib.isofts.kiev.ua/handle/123456789/117419 en Физика низких температур application/pdf Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
topic Low Temperature Spectroscopy and Radiation Effects
Low Temperature Spectroscopy and Radiation Effects
spellingShingle Low Temperature Spectroscopy and Radiation Effects
Low Temperature Spectroscopy and Radiation Effects
Wonderly, W.R.
Anderson, D.T.
Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
Физика низких температур
description We present FTIR studies of the 193 nm photolysis of fully deuterated formic acid (DCOOD) isolated in solid parahydrogen at 1.9 K which show evidence of transient HDO rovibrational satellite peaks. The S1 and S2 satellite peaks are readily detected for α-type (1₀₁ ← 0₀₀) rovibrational transitions of HDO either during or immediately after photolysis. Intensity measurements show the HDO b-type (1₁₁ ← 0₀₀) rovibrational transitions have satellite peaks as well, but due to the greater linewidth of these absorptions, the satellite peaks cannot be spectroscopically resolved from the monomer transition and are therefore difficult to detect. These newly identified HDO satellite peaks may result from the HDO photoproduct being formed next to an H atom or a vacancy in the parahydrogen solid. The development of the infrared spectroscopy of these satellite peaks can provide a new means to study radiation effects on low-temperature hydrogen solids doped with chemical species.
format Article
author Wonderly, W.R.
Anderson, D.T.
author_facet Wonderly, W.R.
Anderson, D.T.
author_sort Wonderly, W.R.
title Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
title_short Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
title_full Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
title_fullStr Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
title_full_unstemmed Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
title_sort transient hdo rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies?
publisher Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України
publishDate 2012
topic_facet Low Temperature Spectroscopy and Radiation Effects
url https://nasplib.isofts.kiev.ua/handle/123456789/117419
citation_txt Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies? / W.R. Wonderly , D.T. Anderson // Физика низких температур. — 2012. — Т. 38, № 8. — С. 853-859. — Бібліогр.: 31 назв. — англ.
series Физика низких температур
work_keys_str_mv AT wonderlywr transienthdorovibrationalsatellitepeaksinsolidparahydrogenevidenceofhydrogenatomsorvacancies
AT andersondt transienthdorovibrationalsatellitepeaksinsolidparahydrogenevidenceofhydrogenatomsorvacancies
first_indexed 2025-11-25T12:21:44Z
last_indexed 2025-11-25T12:21:44Z
_version_ 1849764937434398720
fulltext © William R. Wonderly and David T. Anderson, 2012 Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8, pp. 853–859 Transient HDO rovibrational satellite peaks in solid parahydrogen: evidence of hydrogen atoms or vacancies? William R. Wonderly and David T. Anderson Department of Chemistry, University of Wyoming, Laramie, WY 82071-3838, USA E-mail: danderso@uwyo.edu Received March 18, 2012 We present FTIR studies of the 193 nm photolysis of fully deuterated formic acid (DCOOD) isolated in solid parahydrogen at 1.9 K which show evidence of transient HDO rovibrational satellite peaks. The S1 and S2 satel- lite peaks are readily detected for a-type (101 ← 000) rovibrational transitions of HDO either during or imme- diately after photolysis. Intensity measurements show the HDO b-type (111 ← 000) rovibrational transitions have satellite peaks as well, but due to the greater linewidth of these absorptions, the satellite peaks cannot be spec- troscopically resolved from the monomer transition and are therefore difficult to detect. These newly identified HDO satellite peaks may result from the HDO photoproduct being formed next to an H atom or a vacancy in the parahydrogen solid. The development of the infrared spectroscopy of these satellite peaks can provide a new means to study radiation effects on low-temperature hydrogen solids doped with chemical species. PACS: 61.80.–x Physical radiation effects, radiation damage; 67.80.ff Molecular hydrogen and isotopes; 67.80.dj Defects, impurities, and diffusion. Keywords: solid hydrogen, impurities, diffusion. 1. Introduction In preliminary studies of the 193 nm in situ photolysis of different isotopomers of formic acid (HCOOH and DCOOD) in solid parahydrogen (pH2) at 1.9 K, we identi- fied satellite peaks in close proximity to the a-type rovibra- tional peaks of the H2O and HDO photoproducts [1]. Spe- cifically, we observe the S1 satellite peak approximately 1 cm–1 lower and the S2 peak around 2 cm–1 higher in energy than the monomer a-type R(0) rovibrational peak for HDO and H2O, but not D2O. Even when the sample is maintained at 1.9 K after photolysis the intensity of these satellite peaks decay slowly with a time constant of 121(7) min [1]. Photoexcitation of HCOOH in the gas phase at 193 nm leads to direct dissociation to give HCO + OH as the dominant photochannel [2,3]. In the preliminary studies we used the lack of satellite peaks for D2O to as- sign the observed satellite peaks to H···HDO or H···H2O radical clusters that form as by-products of the 193 nm photochemistry via reactions of the OD or OH photofrag- ments with the pH2 host [1]. However, at that time we were unable to definitively explain why we did not observe the analogous satellite peaks associated with the b-type R(0) rovibrational transitions of either HDO or H2O. Fur- ther, the spectroscopic data is also consistent with the HDO or H2O photoproducts being produced next to a va- cancy left by the reacting pH2 molecule. Therefore, in this work we more fully develop our understanding of the in- frared (IR) spectroscopy of the HDO satellite peaks. The rovibrational bands of asymmetric top molecules such as H2O or HDO are classified by whether the transi- tion dipole moment responsible for the absorption has a pro- jection along the A, B, or C inertial axes of the molecule [4]. For H2O the symmetry axis corresponds to the B axis and the A axis lies in the plane of the molecule, so the ν2 bend fundamental is a b-type band and the ν3 asymmetric stretch is an a-type band [5]. For HDO where the symmetry is lowered by isotopic substitution, the ν2 bend is now a mixed a/b-type band since the transition dipole has pro- jections along both inertial axes. Furthermore, HDO and H2O freely rotate in solid pH2 such that at the low temper- atures where solid pH2 is stable, the overwhelming majori- ty of HDO or H2O molecules populate only the lowest rotational level, namely a cK KJ = 000. Under these condi- tions, the rovibrational bands collapse into a single rovibra- tional transition which is b-type (111 ← 000) for the ν2 bend and a-type (101 ← 000) for the ν3 asymmetric stretch of H2O, but both b- and a-type rovibrational transitions are observed for HDO for the ν2 and ν3 vibrational fundamen- William R. Wonderly and David T. Anderson 854 Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 tals [5]. In this paper we show that satellite peaks are also present for the b-type rovibrational transitions of HDO, but due to the greater full width half maximum of the b-type transitions of HDO in solid pH2, the satellite peaks cannot be resolved from the monomer transition and thus are much more difficult to detect using FTIR. The assignment of the carriers of these HDO satellite peaks is important if we are to develop new methods to study the in situ photochemistry of chemically doped pH2 quantum solids. If the HDO satellite peaks presented in this work are due to H···HDO radical clusters then these satellite peaks could be used to study the mechanism by which in situ photochemical reactions proceed and also provide a new spectroscopic handle on H atoms in solid pH2. Indeed, H atoms were first detected and shown to be mobile in solid hydrogen samples at liquid helium temper- atures nearly 30 years ago [6–8]. Much of the subsequent progress in our understanding of H atoms trapped in solid pH2 has been provided by the electron spin resonance (ESR) studies of Kumada and co-workers [9–14]. These studies led first to the determination that the H atoms dif- fuse via a chemical diffusion mechanism by which the H atom moves by sequential H + H2 → H2 + H chemical exchange reactions [13]. Through a variety of ESR measure- ments, both the rates of H atom diffusion and H + H → H2 recombination can be studied separately [12]. These mea- surements show that for highly enriched pH2 samples (i.e., low orthohydrogen (oH2) concentrations) and at low tem- perature, the recombination rate is too slow to be explained by the diffusion rate which suggests that the H atoms do not recombine efficiently under these conditions [12]. Fushita- ni and Momose used FTIR spectroscopy [15] to measure the H atom diffusion rate by studying the H + NO → HNO tunneling reaction after the in situ photolysis of NO in so- lid pH2 at 5.2 K. In these studies the H atoms are produced as by-products of the NO photolysis, and the H-atom diffu- sion rate is determined from the growth in the IR absorp- tion intensity of HNO. Andrews and co-workers studied the induced IR transitions in solid hydrogen (and enriched pH2 solids) produced by the presence of an H atom either by laser ablating metal atoms into the solids or by direct condensation of hydrogen gas subjected to tesla coil dis- charge in a quartz tube [16,17]. In these studies H atom induced peaks in pH2 were identified at 4151.8 cm–1 which show intensity half-lives compatible with the ESR mea- surements. The other possibility that we consider is the HDO satel- lite peaks are caused by HDO produced next to a vacancy. As discussed in our earlier work, we only observe HDO satellite peaks for irradiated samples suspected to produce OD photofragments that can react with the pH2 host to form HDO. In this case the products of the OD + pH2 → → HDO + H chemical reaction get trapped next to each other in the solid, and may lead to the formation of H···HDO radical clusters. However, if the H atom escapes somehow, then the HDO molecule might be trapped next to the vacancy left behind by the pH2 molecule that reacted. This would mean that the in situ photolysis pro- duces vacancies and these vacancies may play a role in subsequent H atom reactions that occur after photolysis. Indeed vacancies are central to the original proposal by Andreev and Lifshitz that under certain conditions quan- tum crystals may show a new super state of matter, a so- called supersolid [18]. A number of studies have explored defect (H atom, oH2, 2H )− diffusion in hydrogen crystals, but none have been able to directly quantify the number of vacancies [19,20]. If the HDO satellite peaks reflect HDO next to vacancies, then in situ photolysis of certain precur- sor molecules can lead to the generation of nonequilibrium vacancy populations at low temperature which may influ- ence the diffusion of chemical impurities after photolysis. 2. Experimental details Our experimental apparatus is described in detail else- where [1,21]; briefly, we grow millimeters thick, chemi- cally doped pH2 crystals via co-deposition of independent gas streams of dopant (e.g., DCOOD) and pH2 host onto a BaF2 optical substrate maintained at approximately 2.5 K inside a sample-in-a-vacuum liquid helium bath cryostat. The host in these studies is enriched to approximately 99.97% pure pH2 levels using a variable temperature or- tho/para catalytic converter operated near 14.0 K during deposition. FTIR spectroscopy at 0.05 cm–1 resolution is performed on the sample using a normal incidence trans- mission optical setup. The measured integrated intensities of specific solid pH2 IR absorptions allow us to determine the IR path length through the sample [22], which permits the concentration of species within the crystal to be deter- mined using Beer’s law. Photolysis is achieved using the 193 nm output of an ArF excimer laser (Gam Laser EX5) configured to pass through the sample at an angle of 45° with respect to the FTIR beam. This optical setup permits FTIR spectra to be recorded within the photolyzed region of the crystal either during or immediately after 193 nm irradiation. The DCOOD concentration is the average val- ue determined from the ν1 (OD stretch) and ν6 (deforma- tion) peaks with integrated absorption intensities of 36.9 and 56.8 km/mol, respectively [23]. 3. Experimental results and discussion Shown in Fig. 1 is IR difference spectra of the a-type (101 ← 000) and b-type (111 ← 000) rovibrational transi- tions of the ν2 mode of HDO for a DCOOD doped pH2 sample that is subjected to 193 nm photolysis. The differ- ence IR spectra (after minus before) shown as traces (a) through (d) in Fig. 1 are generated using a “before” spec- trum recorded just prior to photolysis, and a series of “after” spectra recorded with 3 min acquisition times Transient HDO rovibrational satellite peaks in solid parahydrogen Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 855 (0.05 cm–1 and 16 co-added scans) centered at 1.5, 4.8, 8.2, and 11.5 min after the start of the photolysis laser. All these spectra are recorded during the 193 nm photolysis of a 30 ppm DCOOD doped pH2 solid (99.97% pH2) at 1.93 K. We estimate that the solid pH2 contains 6.4 ppm of DCOOH, 1.3 ppm of HCOOD, and 0.7 ppm of HCOOH pri- or to photolysis due to isotopic scrambling of the DCOOD. As can be seen in Fig. 1, both the a-type and b-type rovi- brational transitions for the ν2 bend of HDO increase in intensity with photolysis. These two rovibrational transi- tions are labeled a-type and b-type, respectively, to indi- cate the orientation of the transition dipole moment with respect to the inertial axis of the HDO molecule and to account for the selection rules for that particular rovibra- tional transition. However, the S1 and S2 satellite peaks are only clearly detected for the a-type rovibrational tran- sition while no satellite peaks are detected for the b-type transition. As discussed in the Introduction, this observa- tion confused us in preliminary studies where we specu- lated the lack of b-type satellite peaks was due either to some sort of selection rules or related to the breadth of the two rovibrational absorptions [1]. These two rovibrational transitions for HDO monomers isolated in solid pH2 have been well characterized previously [5,24,25], however the reason that the two peaks have such different full width half maximum (fwhm) values has not been considered in detail [5,25]. The greater fwhm of the b-type transition (1.8 cm–1) compared to the a-type absorption (0.27 cm–1) is likely due to faster rotational relaxation of the 111 rota- tional state compared to the 101 state, but a detailed study of these differences is beyond the scope of the present work. Here we only point out that if the satellite peaks for the b-type rovibrational transitions come at comparable shifts from the monomer rovibrational transition, then the satellite peaks will be difficult to detect since they will not be cleanly resolved, as is the case for the a-type transition. To clearly show that S1 and S2 satellite peaks are de- tected for a-type rovibrational peaks of all the fundamental modes of HDO, we show the analogous difference spectra for the ν1 and ν3 HDO rovibrational peaks in Fig. 2. These two a-type transitions show that the shifts of the satellite peaks from the monomer rovibrational transition are quite similar for both rovibrational transitions (and with the ν2 a-type peak in Fig. 1). Specifically, the S1 peak always comes at lower wavenumbers and the S2 at higher wave- numbers compared to the HDO monomer transition. This seems to indicate that the shift of the satellite peaks from the corresponding monomer transition depends on the up- per rotational state, which is the same for all these transi- tions, and not too strongly on the specific vibrational state. The spectra shown in Fig. 2 do not have as high a signal- to-noise (S/N) ratio as the previously published spectra [1], and this is due in part to the greater noise of the MCT de- Fig. 1. (Color online) A series of IR difference absorption spectra (after minus before) displaying the growth of the ν2 HDO a-type and b-type rovibrational peaks during photolysis of a DCOOD doped pH2 sample. The “before” spectrum is recorded just before photolysis is started, and the “after” spectra are recorded at 1.93 K during the 14.4 min (37 mW at 250 Hz) UV irradiation of the sample. Each FTIR scan has a 3 min acquisition time (0.05 cm–1 and 16 co-added scans) and traces (a) through (d) were recorded at increasing times (1.5, 4.8, 8.2, 11.5 min) after photolysis is started. Note the growth of the S1 and S2 satellite peaks close to the a-type transition. The sample is 0.26(1) cm thick with an initial DCOOD concentration of 30(5) ppm. Spectra are offset vertically for ease of comparison. 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Δl og ( /) 10 0I I a-type b-typeS1 S2 (a) (b) (c) (d) 1410 1420 1430 1440 Wavenumber, cm–1 Fig. 2. (Color online) The same series of IR difference spectra shown in Fig. 1 showing the ν1 and ν3 HDO a-type rovibrational peaks. Note that both HDO rovibrational transitions show S1 and S2 satellite peaks. 0.6 0.8 0.4 0.2 0 (a) (b) (c) (d) Wavenumber, cm–1 S1 S2 S1 S2 ν1 HDO ν3 HDO 2728 2733 3708 3713 Δl og ( /) 10 0I I William R. Wonderly and David T. Anderson 856 Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 tector used in this study compared to the InSb detector used previously. However, we used the MCT detector spe- cifically to record the ν2 bend rovibrational transitions shown in Fig. 1 since this mode has both relatively strong a-type and b-type rovibrational peaks. These ν2 HDO ab- sorptions cannot be recorded with the InSb detector be- cause the peaks are below the band gap for that detector. Further, the two ν2 HDO transitions shown in Fig. 1 are well separated and do not overlap with any other HDO, D2O, or H2O peaks allowing us to measure the integrated intensity of both of these features. As we will show in this paper, this makes the ν2 HDO fundamental an excellent spectroscopic probe to determine the reason for the appar- ent lack of b-type satellite peaks. Another advantage of HDO compared to H2O or D2O is that HDO does not have complications in the intensity analysis due to different nuc- lear spin states. The in situ photolysis of DCOOD in solid pH2 also produces growth in the intensities of D2O and H2O rovibrational peaks as well [1]. However, for H2O and D2O these species are also produced in excited nuclear spin states which relax with time. As we will show, this complicates the intensity analysis of the H2O satellite peaks. Shown in Fig. 3 are rovibrational transitions out of the lowest rotational state for the two vibrational modes of D2O. In this case, neither the b-type ν2 bend nor the a-type ν3 asymmetric stretch shows evidence of satellite peaks. As we reported earlier [1], this is what we expect if the carriers of the satellite peaks are only formed when one of the photoproducts chemically reacts with the pH2 host. Starting from a DCOOD precursor molecule, the only way to generate D2O is via the molecular photochannel that produces CO + D2O. Indeed, gas phase measurements on HCOOH indicate the molecular photoproducts (i.e., CO + + H2O and CO2 + H2) are important at 193 nm and strong- ly favor production of CO + H2O channel [2,3]. In fact, photochemical studies of HCOOH in Ar and Xe matrices show that only the molecular photochannels are observed in these condensed phase studies [26]. While some isotopic scrambling does occur to the DCOOD precursor molecule prior to photolysis, which can complicate the determina- tion of the mechanistic origins of the HDO or H2O photo- products, the D2O must come directly from the DCOOD precursor. This is consistent with the absence of any satel- lite peaks detected for the D2O molecule because there is no way to form D2O from reactions of the OD photofrag- ment with the pH2 host. Thus, while not observing satellite peaks for the b-type ν2 transition is inconclusive, the lack of satellite peaks for the a-type transition strongly supports our assignment of why the satellite peaks form. For exam- ple, in other experiments using our InSb detector and thus with higher S/N in the ν3 region, we have not observed any hint of a-type D2O satellite peaks. Finally for completeness, we show in Fig. 4 the analog- ous rovibrational peaks for H2O. Once again, we only ob- serve satellite peaks centered around the a-type rovibra- tional transition and not for the b-type peak of the ν2 bend of H2O. Tracking the mechanistic origins of the H2O satel- lite peaks is less clear than for D2O since some of the sig- nal could result from the photochemistry of minor isoto- pomers present in the pH2 solid. Further, it seems that some of the H2O satellite peak intensity comes from DCOOD via photoproduction of OD followed by exchange reactions with the pH2 host or secondary photolysis that ultimately produces OH. Thus, while it is difficult to tease out contributions from photolysis of DCOOH from DCOOD, the data indicates that some of the H2O satellite peaks are 0.15 0.25 0.20 0.30 0.10 0 Δl og ( /) 10 0I I Δl og ( /) 10 0I I Wavenumber, cm–1 ν2 2 D O ν3 2 D O 1195 1200 2788 2793 0.6 0.5 0.4 0.3 0.2 0.1 0 a-type b-type (a) (b) (c) (d) Fig. 3. (Color online) The same series of IR difference spectra shown in Fig. 1 now displaying the b-type ν2 D2O and a-type ν3 D2O rovibrational peaks. Note that neither of the D2O rovibra- tional peaks show evidence of S1 and S2 satellite peaks. Fig. 4. (Color online) The same series of IR difference spectra shown in Fig. 1 now displaying the b-type ν2 H2O and a-type ν3 H2O rovibrational peaks. Note that only the a-type rovibrational peak shows clear S1 and S2 satellite peaks. 0.3 0.2 0.4 0.1 0 Wavenumber, cm–1 ν2 2 H O ν3 2 H O 1629 1634 3763 3768 0.8 0.6 0.4 0.2 0 a-type b-type S1 S2 Δl og ( /) 10 0I I Δl og ( /) 10 0I I (a) (b) (c) (d) Transient HDO rovibrational satellite peaks in solid parahydrogen Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 857 produced from photolysis of DCOOD. In the case of DCOOD photolysis only the O atom comes from the pre- cursor, and all the H atoms in the cluster come from reac- tions with the pH2 solid. An important clue as to why the b-type satellite peaks are not detected is provided by careful intensity analysis of the ν2 HDO rovibrational peaks. The satellite peaks rely on the transition strength of the HDO (H2O) rovibrational transition to be detected, and therefore satellite peaks should be evident for all the rovibrational transitions of HDO (H2O), not just the a-type transitions. Shown in Fig. 5,a are plots of the integrated intensity of various ab- sorption features versus time for the experiment depicted in the earlier spectral figures. The grey vertical bar in Fig. 5,a indicates the time and duration of the 193 nm pho- tolysis exposure for this experiment. As can be seen in Fig. 5,a, before photolysis the intensity of the HDO mo- nomer b-type and a-type rovibrational transitions are both quite small simply due to the low concentration of HDO prior to photolysis. During photolysis, the intensity of the HDO monomer transitions both increase as well as the S1 and S2 satellite peaks. The plot of the b-type transition intensity rapidly increases during photolysis, and then re- mains constant after photolysis. This is the expected beha- vior for the photolytic production of HDO if no interme- diate species are detected. However, for the a-type trans- ition the intensity is observed to continue to increase slightly after photolysis is complete (see Fig. 5,a). This is due to the mechanism by which the HDO monomer is pro- duced and because we resolve the a-type satellite peaks; we can monitor the intensity or concentration of these two species separately. The HDO photoproduct cannot be gen- erated directly from the DCOOD precursor, instead some of the photoexcited DCOOD molecules dissociate along the DCO + OD radical photochannel. The nascent OD ei- ther reacts directly with the pH2 host, or as we suspect, becomes thermally equilibrated and then reacts with the pH2 host via a tunneling reaction mechanism that pro- duces HDO trapped next to an H atom or a vacancy. The OH + H2 → H2O + H hydrogen abstraction reaction is well characterized and is exothermic (–7450 K) and has a low barrier (3060 K) [27,28]. The presence of the H atom or vacancy next to the HDO molecule slightly perturbs the 101 ← 000 a-type rovibrational transition such that it lifts the three-fold upper state MJ degeneracy and splits the transition into two peaks (e.g., S1 and S2). Once the satel- lite peaks are produced via this mechanism, the intensity starts to decrease as the HDO molecules in this metastable solvation site decay back to a well isolated HDO molecule in the ground rotational state. Due to the low temperature, relaxation of the quantum solid around the newly formed HDO molecule by either mechanism (movement of the H atom or vacancy away from the HDO molecule) would be expected to occur with such a slow time constant (e.g., 121(7) min). Consistent with this assignment, the sum of the S1 and S2 satellite peaks intensities decay with time after photolysis while the a-type HDO monomer transition increases. Furthermore, the decrease in the sum of the two satellite peak integrated intensities matches quantitatively the increase in the intensity of the a-type HDO mono- mer transition. Thus, if the sum of the satellite intensities (S1 + S2) is added to the intensity of the a-type HDO mo- nomer transition, the resulting intensity profile (a-type sum) matches qualitatively the b-type where the contribu- tions between the satellite and monomer peaks cannot be separately measured using FTIR spectroscopy. This im- plies that the broad b-type absorption profile includes con- tributions from the satellite peaks, but these contributions are just not spectroscopically resolved. To further show that b-type satellite peaks are also present in the spectra presented in Fig. 1, but just not spec- troscopically resolved, we show in Fig. 5,b integrated in- tensity correlation plots. If we make a plot of the a-type intensity versus the b-type intensity (blue circles), we ob- serve a nonlinear correlation. This indicates that the po- Fig. 5. (Color online) (a) A plot of the integrated intensities of the ν2 HDO b-type, a-type, S1 + S2, and sum of the a-type and S1 + S2 satellite peaks (a-type sum) versus time. The grey shaded area represents the timing of the 193 nm photolysis during the experi- ment. Note the temporal profile of the a-type sum and b-type is very similar. (b) A correlation plot (blue circles) of the integrated intensity of the a-type transition versus the b-type transition gen- erated from the data shown in (a). The corresponding correlation plot (red circles) for a-type sum versus b-type displays a strong linear correlation with a fitted slope of 0.6492(14) determined from a least squares fit of the data to y = ax (solid red line). 0.25 0.20 0.15 0.10 0.05 0 0 50 100 150 200 Time, min 0 0.05 0.10 0.15 0.20 0.15 0.10 0.05In te ns ity , c m –1 In te ns ity , c m –1 (b) a-type a-type b-type a-type sum a-type sum S1 + S2 (a) B-type intensity, cm –1 William R. Wonderly and David T. Anderson 858 Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 pulation or concentration of the species measured by a- type transition does not match exactly the population probed by the b-type transition. More specifically, after photolysis the a-type HDO monomer transition continues to increase in intensity while the b-type HDO intensity remains constant (vertical scatter of data at the high b-type intensity). This simply shows that the b-type transition reflects the total of satellite and monomer peaks, while for the a-type transition these two contributions are spectros- copically resolved and can be detected separately. If in- stead we make a correlation plot of the a-type sum (mo- nomer plus satellites) versus the b-type transition intensity, then we observe a strong linear correlation with a fitted slope of 0.6489(17). Indeed, this is what is expected based on the transitions strengths of a-type and b-type HDO ν2 bend transitions, which show that the transition dipole has a larger projection along the inertial B axis [29]. This also shows that the approximation that the transition strength is the same for the monomer and satellite peaks is valid. Thus, the satellite peaks are present for the b-type transi- tions, but just not spectrally resolved. To further test this explanation, we looked at the details of the b-type lineshape right after photolysis is complete, when the satellite peaks have the greatest intensity, and after the sample has been allowed to fully equilibrate and the satellite peaks have decayed to zero. Shown in Fig. 6 are expanded views of the b-type transitions of D2O (a) and HDO (b) for the ν2 bend vibrational mode (blue trace). Each transition is then least-squares fit to a pseudo-Voigt lineshape [30] and the fitted lineshape is also shown in Fig. 6 (red trace), so that the difference lineshape can be examined at greater detail. For D2O, as shown previously, there are no satellite peaks produced and the difference between the two lineshapes just shows overall growth in the intensity of this feature. This can easily be explained since some of the D2O molecules are produced in excited nuclear spin states during photolysis and thus right after photolysis is complete, the intensity of the b-type transition of D2O out of the lowest rotational level (000) increases in intensity as the para-D2O molecules in excited rotational levels (101) convert to the lower energy ortho-D2O spin state [1]. Indeed, we have shown in these earlier studies that the growth is first-order with a rate constant consistent with the literature value for the nuclear spin conversion rate constant of D2O [1,5]. However, the difference spec- trum generated for the analogous b-type HDO rovibration- al transition shows a qualitatively different behavior. In- stead in this case, the peak appears to narrow slightly after photolysis is complete. This is shown in Fig. 6,b by the difference spectrum that increases at the center line but has two dips near the wings. This behavior is consistent with there being two contributions to the observed lineshape whose relative concentrations change with time after the photolysis is complete. However, the total integrated inten- sity remains constant because the absorption strengths of both the satellite and monomer transitions are the same. Interestingly, the effect of these changes in relative con- centrations is just at the level that can be detected using this spectroscopic data. This again is consistent with ap- proximately 15% of the b-type signal being due to satellite peaks right after photolysis is complete, which then decays with time. It does not appear that the contributions of the satellite peaks can be extracted solely from the b-type ro- vibrational data, but fortunately the satellite peaks can easi- ly be resolved using the narrower a-type rovibrational tran- sition. In fact, using the gas phase absorption strength for the HDO ν2 bend (56.91 km/mol) [31] appropriately scaled for the weaker a-type rovibrational transition, we estimate using Beer’s law that concentration of HDO molecules in this metastable solvation state is 1.1 ppm right after photo- lysis is complete under the conditions of this experiment. 4. Summary In preliminary studies of the 193 nm in situ photolysis of DCOOD in solid pH2 at temperatures below 2 K we identified satellite peaks for the a-type rovibrational transi- tions of all three fundamental modes of HDO which are produced during photolysis and decay in intensity after photolysis is complete. We assigned these satellite features to H···HDO radical clusters that form by reactions of OD with the pH2 host. However, in these initial studies we were confused as to why we did not observe the analogous Fig. 6. (Color online) (a) Expanded view of the absorption spec- tra (blue) of the b-type ν2 D2O and (b) ν2 HDO rovibrational transitions and least squares fits to a pseudo-Voigt lineshape (red). Spectra are offset vertically for ease of comparison. The top spectrum is recorded 496.5 min after the middle spectrum, which is recorded immediately after photolysis with a 3 min acquisition time. The bottom difference spectra (top — middle) is generated from these two spectra with the intensity multiplied by a factor of two (2×). For D2O the peak intensity increases with time, but for HDO the difference lineshape is more complex due to contribu- tions from satellite peaks (see text for details). 0.10 0.05 0 2 2 1195 1200 1430 1435 (a) (b) Δl og ( /) 10 0I I Wavenumber, cm–1 Transient HDO rovibrational satellite peaks in solid parahydrogen Low Temperature Physics/Fizika Nizkikh Temperatur, 2012, v. 38, No. 8 859 satellite transitions for b-type HDO rovibrational transi- tions. In the present paper, we use the ν2 bend of HDO to show that indeed the satellite peaks are present for b-type transitions, but are difficult to detect due to the greater fwhm of the b-type transitions of HDO in solid pH2. Plots of integrated intensity versus time after photolysis show that both the HDO monomer and satellite peaks contribute to the b-type 111 ← 000 rovibrational absorption intensity, while for the a-type transition the two contributions are spectrally resolved and can therefore be separated. The present study also shows that the absorption strength of the a-type HDO monomer and satellite peaks are equal, such that the gas phase absorption strength can be used to calcu- late the approximate concentration of HDO molecules in this metastable state. The additional spectral data presented in this paper is consistent with the carriers of the satellite peaks being ei- ther H···HDO radical clusters or HDO next to a substitu- tional vacancy and represents a significant step forward in the FTIR spectroscopy of these satellite peaks. The IR spectroscopy shows that the satellite spectra (splitting and relative intensities) are very similar for all three vibrational modes of HDO and thus primarily reflect the lifting of the (2J + 1) spatial degeneracy of the upper 101 HDO rotation- al state. The ground rotational state (000) is non-degene- rate. However, a complete assignment of the satellite peaks is still lacking. Indeed it may prove difficult to distinguish between the two potential assignments presented here us- ing just IR spectroscopy. We hope the preset experimental findings stimulate theoretical calculations of the two pre- sented scenarios: (i) rotational motion of HDO next to an H atom in solid pH2 and (ii) rotation of HDO next to a vacancy in solid pH2. In either case the identification of these HDO satellite peaks in UV irradiated DCOOD doped pH2 solids offer new potential insights into the details of radiation effects in chemically doped quantum solids. We are currently using the spectroscopy developed here to stu- dy the details of the photolysis of DCOOD in solid pH2 aimed at correlating the decay in the satellite peaks with the growth of other species within the solid. Acknowledgments W.R.W. was supported partially by the Research Expe- riences for Undergraduates (REU) Program of the National Science Foundation under Award Number CHE 08-51931. The authors thank the National Science Foundation for its generous support through Grant CHE 08-48330. 1. K.A. Kufeld, W.R. Wonderly, L.O. Paulson, S.C. Kettwich, and D.T. Anderson, J. Phys. Chem. Lett. 3, 342 (2012). 2. H. Su, Y. He, F. Kong, W. Fang, and R. Liu, J. Chem. Phys. 113, 1891 (2000). 3. E. Martinez-Nunez, S.A. Vazquez, I. Borges, Jr., A.B. Rocha, C.M. Estevez, J.F. Castillo, and F.J. Aoiz, J. Phys. Chem. A109, 2836 (2005). 4. P.F. Bernath, Spectra of Atoms and Molecules, Oxford Uni- versity Press, Inc., Oxford (2005). 5. M.E. Fajardo, S. Tam, and M.E. DeRose, J. Mol. Struct. 695, 111 (2004). 6. A.V. Ivliev, A.Y. Katunin, I.I. Lukashevich, V.V. Sklyarev- skii, V.V. Suraev, N.I. Filippov, and V.A. Shevtsov, JETP Lett. 36, 472 (1982). 7. A.S. Iskovskhih, A.Y. Katunin, I.I. Lukashevich, V.V. Sklya- revskii, V.V. Suraev, N.I. Filippov, and V.A. Shevtsov, Sov. Phys. JETP 64, 1085 (1986). 8. T. Miyazaki, K.P. Lee, K. Fueki, and A. Takeuchi, J. Phys. Chem. 88, 4959 (1984). 9. T. Kumada, N. Kitagawa, T. Noda, J. Kumagai, Y. Aratono, and T. Miyazaki, Chem. Phys. Lett. 288, 755 (1998). 10. T. Miyazaki, S. Mori, T. Nagasaka, J. Kumagai, Y. Aratono, and T. Kumada, J. Phys. Chem. A104, 9403 (2000). 11. T. Kumada, S. Mori, T. Nagasaka, J. Kumagai, and T. Miya- zaki, J. Low Temp. Phys. 122, 265 (2001). 12. T. Kumada, M. Sakakibara, T. Nagasaka, H. Fukuta, and J. Ku- magai, J. Chem. Phys. 116, 1109 (2002). 13. T. Kumada, Phys. Rev. B68, 052301 (2003). 14. T. Kumada, Y. Shimizu, T. Ushida, and J. Kumagai, Radiat. Phys. Chem. 77, 1318 (2008). 15. M. Fushitani and T. Momose, Fiz. Nizk. Temp. 29, 985 (2003) [Low Temp. Phys. 29, 740 (2003)]. 16. L. Andrews and X. Wang, J. Phys. Chem. A108, 3879 (2004). 17. L. Andrews and X. Wang, J. Chem. Phys. 121, 4724 (2004). 18. A.F. Andreev and I.M. Lifshitz, Sov. Phys. JETP 29, 1107 (1969). 19. D. Zhou, C.M. Edwards, and N.S. Sullivan, Phys. Rev. Lett. 62, 1528 (1989). 20. F. Operetto and F. Pederiva, Phys. Rev. B75, 064201 (2007). 21. L.O. Paulson and D.T. Anderson, J. Phys. Chem. A113, 1770 (2009). 22. M.E. Fajardo, in: Physics and Chemistry at Low Tempera- tures, L. Khriachtchev (ed.), Pan Stanford Publishing Pte. Ltd., Singapore (2011), p. 167. 23. I. Yokoyama, Y. Miwa, and K. Machida, J. Am. Chem. Soc. 113, 6458 (1991). 24. M.E. Fajardo and S. Tam, J. Chem. Phys. 115, 6807 (2001). 25. M.E. Fajardo and C.M. Lindsay, J. Chem. Phys. 128, 014505 (2008). 26. J. Lundell and M. Räsänen, J. Phys. Chem. 99, 14301 (1995). 27. M. Alagia, N. Balucani, P. Casavecchia, D. Stranges, G.G. Volpi, D.C. Clary, A. Kliesch, and H.-J. Werner, Chem. Phys. 207, 389 (1996). 28. M.D. Wheeler, D.T. Anderson, and M.I. Lester, Int. Rev. Phys. Chem. 19, 501 (2000). 29. S.A. Clough, Y. Beers, G.P. Klein, and L.S. Rothman, J. Chem. Phys. 59, 2254 (1973). 30. A.L. Stancik and E.B. Brauns, Vib. Spectrosc. 47, 66 (2008). 31. D.J. Swanton, G.B. Bacskay, and N.S. Hush, J. Chem. Phys. 84, 5715 (1986).