EricDuranzaResearchPaper

Topic: Matrix Trapping

Abstract Matrix trapping is an isolation technique used to isolate very reactive species such as reaction intermediates or radicals. The reactive species is frozen inside of a chemically inert matrix, such as a noble gas, which prevents the targeted species from reacting with anything. It can be difficult to determine the conditions needed to isolate a species as the conditions needed tend to be very exact in what is required. A research group at Furman University in South Carolina took twenty years to determine the method for isolating the H2+ cation. Matrix isolation can even help with determining the progress of a reaction as it can isolate both the intermediates and the products and provide information as to how the reaction proceeds from start to finish. In the reaction trimethylaluminum with methanol, from the inspection of the intermediates formed, you can see how the change of certain variables in the matrix isolation process will affect the progress of the reaction.

Introduction Matrix trapping, also called “matrix isolation”, is a technique that has been developed in order to more closely examine molecules that have been isolated in an inert substance or matrix. This technique is essential when taking a look at unstable systems as it removes the possibility of the species targeted for study from changing to a species that you are not studying. The method involves isolating the species in an inert matrix that is kept rigid at extremely low temperatures. The rigidity of the matrix is used to prevent the species from diffusing throughout the matrix and possibly reacting. The inertness of the matrix is used to prevent the loss of the targeted species through a reaction with the substance comprising the matrix. Because of the need for an inert substance, there are only a few materials outside of noble gases and molecular nitrogen that are inert enough to provide an acceptable matrix. The trapped species is able to be subjected to typical spectroscopic methods of characterization and analysis. Despite having been used for the past three decades, not much is known about the details of the matrix material and the nature of the site at which the targeted species is trapped. For instance, it is unknown whether the matrix is crystalline, amorphous, or in some other form. The most common technique used to deposit the matrix is the “spray-on technique” which involves spraying the sample very slowly onto a low-temperature window. The rate of deposition for this technique is typically only a few millimoles per hour. The second technique is called “pulsed deposition” and involves depositing the sample in a series of pulses. This technique is ideal as it yields a matrix that is optically clearer and gives sharper bands. The reason why this technique results in a superior matrix is due to the annealing of the matrix that takes place as the heat that is transferred to the matrix with each pulse of sample is dissipated. The third method is to simply slowly freeze the gas solution by carefully regulating the temperature of the system [1].

Many studies on the subject of matrix trapping are performed in order to understand how to characterize the matrix and how certain variables affect the formation of the matrix. Such variables are temperature, rate of deposition, and method of deposition. In recent years, molecular dynamics simulations have been used to study issues in condensed matter physics such as defects in crystal growth and glasses. Simulations run using molecular dynamics models are used to investigate the various facets that affect the deposition and formation of the matrix [1]. The dominant theory on where the trapping sites, and thus the targeted molecules, are located is that they are the substitutional sites located in the rare gas solid. For a targeted species that is of similar size to the rare gas atom, a single vacancy at these locations is enough to house the trapped species. For a non-spherical species, the removal of adjacent matrix atoms provides the trapping site. If this removal of atoms occurs, the atoms may assume a shape similar to the targeted species that will fit in the space left by the atoms. This formation of displaced matrix atoms is important when looking for the trapping sites of the targeted species [2].

Trapping of the Molecular Cation H2+ In order to successfully isolate a species needed for characterization via matrix trapping, it may take a variety of different trials to find the necessary environment needed to isolate it. A study performed at Furman University in South Carolina, which aimed to isolate the H2+ cation and characterize its isotopologues using electron spin resonance (ESR), took almost twenty years of trial and error to find the right conditions [3]. The H2+ molecular ion is a difficult target species to characterize spectroscopically as it not only has extremely high reactivity, but since it is a homonuclear diatomic molecule, it cannot possess a dipole and thus does not have a vibrational or rotational spectrum [5]. The difficulty in isolating the cation was a result of the neutral H2 reacting with the present H2+ and forming the H4+ cation [3].

Due to its simplistic nature, the hydrogen molecular ion, H2+ has been the subject of countless and experimental research [4]. The appeal of the molecular cation is that its simplicity provides an opportunity to make accurate theoretical predictions [4]. Through the study of its nuclear hyperfine interaction, data regarding its molecular structure, fundamental symmetries, and nuclear properties have been measured [5]. The early calculations made of the H2+ hyperfine structure were focused on supporting the observations of the molecular ion made in astrophysics [6]. Continued theoretical studies on the hyperfine structure of H2+ improved the accuracy of the already calculated values [7, 8]. A large portion of papers published studying H2+ have been theoretical studies which makes actual experimental studies difficult to find [3]. The earliest experimental studies of the nuclear hyperfine transitions of H2+ in excited vibrational states were performed by K.B. Jefferts. Jefferts bombarded H2 gas with electrons in order to create H2+ ions and then used photodissociation rates to determine the transitions of the molecular ion [9]. The method that Jefferts had used was based on the concept that photodissociation rates of ground-state ions are dependent on F and MF when the sample has been exposed to linearly polarized light [9]. Electric dipole selection rules determined the photodissociation reaction such that:

when the sample is hit by a linearly polarized light [10]. The isolation of the H2+ species was achieved through the use of a 2 K copper deposition target, a 77 K copper heat shield, and a 2 K neon matrix [3]. The key role of the heat shield was that it increased the thermal stability of the neon matrix and all other ESR matrix isolation apparatuses that did not include the heat shield failed to detect any H2+ at all [3]. The following figure details the ESR apparatus used:

Figure 1 - Diagram of Matrix Trapping System [3] The boiling point of the helium was reduced to 2 K by using an open cycle helium cryostat. This helium was used to regulate the temperature of the copper deposition target. The mixtures of neon and hydrogen that were tested varied from ratios of neon to hydrogen 20,000/1 to 100/1. The mixtures were introduced through the gas inlet at a rate of 4 cm3 min-1. A standard deposition time of 100 minutes was used and various rates of deposition were attempted in order to optimize the H2+ ESR absorption. After depositing at 2K, an x-ray source was used operating at 60keV for 15 minutes. After being irradiated, the sample was lowered into the ESR cavity where the sample was photolyzed by a visible light source. Various other methods for ion generation were used, such as open-tube neon discharge photoionization and fast atom beam bombardment (neon seeded with H2). After over 300 experiments, the conditions found to always yield H2+ ESR absorptions under x-irradiation were extreme Ne/H2 dilutions at 2 K using the 77 K copper heat shield. [3]

Using Matrix Trapping to Isolate Reactive Intermediates Matrix trapping can also be a useful tool used to monitor the progress of a reaction. Chemical vapor deposition of thin films that contain aluminum is generally performed using trimethylaluminum as a precursor [12]. Although the reactions of trimethylaluminum with a large variety of electron donors are notorious, much less is known about its reactions in the gas phase which are more centered around chemical vapor deposition processes [13, 14]. Due to (CH3)3Al being an electron-deficient species and thus apt to form donor-acceptor complexes with a range of bases, it would not be unusual to expect the formation of a complex as an intermediate in these reactions [11]. The intermediate complexes would then react to produce the thin film product in a succession of either homogeneous (gas phase) or heterogeneous (reaction taking place on the surface of the substrate) reactions [11]. The reactive intermediates formed in these reactions are perfect subjects of study by matrix isolation [11]. Complexes of (CH3)3Al with a variety of group V and VI alkyl bases have been reported being isolated in argon matrices which helps to elucidate the processes that take place during complex formation[15]. However, bases with an active hydrogen offer a different situation in which methane elimination happens quite easily and further reactions can occur, while also adding to the possibility of forming an initial molecular complex [11]. Such phenomena has been observed during the high temperature reaction and subsequent trapping of reaction mixtures containing (CH3)3Al with H2S and CH3SH [16]. One thing to note is that the reaction of (CH3)3Al with CH3OH which has a considerably more acidic hydrogen than the hydrogen on CH3SH [11]. This reaction has been reported to have possibly formed in solution dimeric, trimeric, or polymeric species [11]. The nature of the initial reaction product has not been determined, as it may be a result of the solvent and various other reaction variables, and there has been no report of the primary monomeric (CH3)3AlOCH3 [11]. As a result, a study was performed to examine the products that are formed during the reaction of (CH3)3Al with CH3OH and its isotopomers in a flowing reactor while trapping the products formed in an argon matrix. Additionally, water was tested in a parallel set of experiments to determine its role [11]. The experiment was carried out in a stainless steel vacuum system utilizing Nupro Teflon-seat valves. A Model 1402B Welch vacuum pump and a Varian M-2 diffusion pump with a liquid nitrogen trap provided the pumping. The cooling for the experiments was supplied by a Model 21 closed cycle refrigerator that was operating at 10 K. Gas samples were deposited from 2L stainless steel vessels through a metering valve onto the cold surface which was a CsI window mounted with indium gaskets to the copper cold block. The deposition of the gaseous samples was oriented perpendicularly to the cold surface. The temperature at the CsI window was measured by using a gold-doped cobalt vs. iron thermocouple and the temperature of the window was regulated by providing a consistent voltage to two 10-W button heaters mounted near the CsI window. [17, 18] Trimethylammonium was pumped into the vacuum system as it was followed by CH3OH, 13CH3OH, CH318OH, CD3OH, CD3OD and H2O which then acted as the “vapor above room temperature liquid after the purification by several freeze-pump-thaw cycles at 77 K” [11]. The argon was used as the matrix gas in all of the experiments as it was, without purification [11].

The samples of the matrix were deposited in two different modes: twin jet and merged jet modes [11]. Twin jet mode is a mode in which each of the two reagents were diluted in argon in separate manifolds and were then sprayed simultaneously onto the cryogenic CsI window from separate nozzles [11]. This mode permits limited and brief mixing and reaction time before matrix deposition occurs [11]. In the merged jet experiments, the two gas samples were prepared in separate manifolds just like the twin jet setup, but the two deposition lines were then joined with a tee at a distance from the CsI window [18]. The length of line in which the two gas samples were merged allowed for greater reaction time than the twin jet deposition but without the static equilibration that occurs with a single jet setup [11]. In order to allow for varied reaction times, the length of the merged region could be adjusted from 10 cm to 250 cm [11]. To induce further reaction, the merged length was also capable of being heated up to 400 C before being trapped at 14 K [11]. The deposition occurred at a flow rate of about 2 mmol h-1 from each manifold for a period of time from 20-24 hours [11]. IR spectra was recorded at various times during deposition including after the deposition period had ended [11]. The type of IR recording the data was either a Perkin-Elmer 983 infrared spectrophotometer over the region of 4000-180 cm-1 with a resolution of 2 cm-1 at 1000 cm-1, or on a Mattson Cygnus FTIR over the range 4000-400 cm-1, at 1 cm-1 resolution [11].

Ab initio calculations were performed on the suspected intermediate species using the Gaussian 94 suite of programs. Stable minima, structures, and vibrational spectra were calculated by using the restricted Hartree-Fock and density functional calculations employing the Becke functional B3LYP. The final calculations were made using full geometry optimization with 6-31G* double zeta basis set only after first calculating initial values with smaller basis sets in order to approximate the energy minima. The calculations were made on a Silicon Graphics Indigo 2 workstation. [11] Before any codeposition experiments, blank spectra of each of the reagents in solid argon was recorded. As had been seen in previous experiments on (CH3)3Al performed by S. Kvisle and E. Rytter, some dimer of the (CH3)3Al was observed once higher sample concentrations were reached. To counteract this, extremely dilute samples of trimethylaluminum to argon were used (Ar/(CH3)3Al=1000-4000) while more concentrated samples of CH3OH were used (200:1 to 500:1). The concentration of the Ar/H2O samples was not explicitly determined, but was in the range of 100:1-500:1 when compared with literature spectra. [11]

An initial twin jet codeposition experiment was conducted with samples of Ar/(CH3)3Al=1500 and Ar/CH3OH=200. This led to no reaction as no new infrared absorption bands were detected and there was no reduction in the intensity of the parent bands. Several more twin jet experiments were conducted at varying concentrations of both samples. No IR bands of the product were observed in any of these experiments. [11] A merged jet codeposition experiment was then performed with the merged length of line of 40 cm and held at room temperature while using samples concentrations of Ar/(CH3)3Al=1500 and Ar/CH3OH=200. The result of this experiment showed many new infrared absorption bands in the IR spectra. [11] The product bands can be observed in the following tables (the issue of “Species B” will be addressed later):

Figure 2 - Table of Absorption Bands for the Product Species [11] As expected, almost all of the absorption bands related to the parent (CH3)3Al had vanished and the parent bands of CH3OH were much lower which shows that reaction had likely gone to completion [11]. A representative of the IR spectra taken is shown below:

Figure 3 - IR Spectra Showing the Difference Between Blank Spectra and Product Formed from the Merged Jet Setup [11] This IR spectra was taken after the merged jet codeposition of a sample of Ar/(CH3)3Al=2000 and a sample of Ar/CH3OH=500 when keeping the merged length of line at room temperature. The noticeable change is that although this spectrum shows various bands close to the bands of the parent compound of (CH3)3Al, the distinct band from (CH3)3Al at 742 cm-1 is absent which indicates that the reaction did in fact proceed to completion. Despite CH4 being present in all of the experiments in small amounts, those using (CH3)3Al had a significant increase in the amount of CH4 relative to the blank experiments, which shows that CH4 is also produced by this reaction. [11]

Following this, a series of 15-20 additional experiments were conducted, using (CH3)3Al and CH3OH in a range of concentrations as indicated previously in merged jet mode. When the concentration of (CH3)3Al was larger than the concentration of CH3OH, the parent bands of CH3OH had disappeared and the parent bands of (CH3)3Al had been greatly reduced. The product bands that are listed in Table 1 persisted throughout the various experiments. In addition to the variation of the concentrations of the reagents, the length of the merged length of line was also varied in a few of these experiments. The change in merged line length did not greatly affect the yield of the product. Once the length of the merged line was increased over 40 cm, there was no definitive evidence of an increase in product yield. Decreasing the length caused the yield of product to slightly drop, but the reaction still seemed to proceed towards completion. [11]

Additional experiments were then performed while heating the merged length of line to above room temperature. The changes seen in the IR spectra were significant as bands that had been previously present at room temperature began to decrease in intensity. At 90 C, there was a noticeable decrease in these bands. When the temperature was increased even higher, the bands were completely absent from the spectra. However, there were product absorption bands that ultimately remained and seemed to have not changed despite the increase in temperature. [11] An example of the effects high temperatures had on the spectra can be seen in the following figure below:

Figure 4 - IR Spectra Showing the Difference Between Room Temperature Versus 250 C [11] This IR spectra was taken after the merged jet codeposition of a sample of Ar/(CH3)3Al=2000 and a sample of Ar/CH3OH=500 with the merged length kept at room temperature as the upper trace, while comparing it to a spectrum of the matrix formed when the merged length was heated to and held at 250 C. [11]

The decrease of some bands at a higher temperature indicates that there are two products present in the matrix at the end of the merged jet experiments: species A (temperature sensitive) and species B (temperature insensitive) [11]. The decomposition of species A at higher temperatures suggests that it is an early intermediate that is converted to either a later intermediate or final product [11]. (CH3)3Al is known to be a strong Lewis Acid that forms complexes with a variety of weak and strong electron pair donors [15, 19]. CH3OH is a moderate Lewis Base as characterized by gas phase proton affinities and reaction chemistry [20]. Thus species A could possibly be the complex of (CH3)3Al and CH3OH [11]. However, the high reactivity of (CH3)3Al towards hydrogen donors would suggest it would react with CH4 as well [11]. So, the two likely candidates for Species A are the complex of (CH3)3Al and CH3OH and the methane elimination product, (CH3)2AlOCH3 [11]. When experiments were run with a 50% mixture of 16O and 18O species, the resulting IR spectra yielded peaks that appeared as doublets [11]. If species A were the complex of (CH3)3Al and CH3OH, which has the possibility of forming the dimer and trimer, then a more complex splitting of the signals would be evident from the IR spectra [11]. Therefore, species A can be concluded to be the elimination product, (CH3)2AlOCH3 [11].

The two candidates for species B are the dimer of (CH3)2AlOCH3 or CH3Al(OCH3)2. It is difficult to tell these two species apart spectroscopically as well as the trimer of (CH3)2AlOCH3. Regarding product yield, a reaction with a second molecule of CH3OH to form the dimethoxy compound would show a strong dependence on the CH3OH concentration which, when looking at the results of the varying concentration of reagents shows that there is no clear or strong reliance on the evolution of species B with the CH3OH concentration. Unfortunately, although the unlikelihood of the dimethoxy compound forming is evident, there is no way to definitively distinguish whether species B is either the dimethoxy or the dimer. [11]

References
 * 1) Cruz, A. and Lopez, G. “The dynamics of matrix trapping. I. Deposition of a pure argon matrix.” J. Chem. Phys. 104, 4294 (1996). []
 * 2) Cruz, A. and Lopez, G. “The dynamics of matrix trapping. II. Simple spherical species trapped in an argon matrix.” J. Chem. Phys. 104, 4301 (1996). []
 * 3) Correnti, M., et al. “Electron spin resonance investigation of H2 +, HD+, and D2 + isolated in neon matrices at 2 K.” J. Chem. Phys. 137, 204308 (2012). []
 * 4) Leach, C. and Moss, R. “Spectroscopy and Quantum Mechanics of the Hydrogen Molecular Cation: A Test of Molecular Quantum Mechanics.” //Ann. Rev. Phys. Chem.// 46, 55 (1995). []
 * 5) Carrington, A., et al. “Microwave electronic spectroscopy, electric field dissociation and photofragmentation of the H+2 ion.” J. Chem. Soc. Faraday Trans. 84, 603(1993). []
 * 6) Stephen, M. and Aufray, J. “Hyperfine Interaction and Quadrupole Coupling Constant in the Hydrogen Molecule Ion.” //J. Chem. Phys.// 31, 1329 (1959). []
 * 7) Babb, J. and Dalgarno, A. “Electron-nuclear coupling in the hyperfine structure of the hydrogen molecular ion.” //Phys. Rev. Lett.// 66, 880 (1991). []
 * 8) Babb, J. and Dalgarno, A. “Spin coupling constants and hyperfine transition frequencies for the hydrogen molecular ion.” //Phys. Rev. A// 46, R5317 (1992). []
 * 9) Jefferts, K. “Rotational Hfs Spectra of H+2 Molecular Ions.” //Phys. Rev. Lett.// 20, 39 (1968). []
 * 10) Jefferts, K. “Hyperfine Structure in the Molecular Ion H+2.” //Phys. Rev. Lett.// 23, 1476 (1969). []
 * 11) Ault, B. and Laboy, J. “Matrix isolation and ab initio study of the reactions of (CH3)3Al with CH3OH and H2O: spectroscopic evidence for (CH3)2AlOCH3.” Journal of Molecular Structure, Volume 475, Issues 2-3, 2 February 1999, Pages 193–202. []
 * 12) Bartram, M., et al. “Chemisorption of trimethylaluminum and ammonia on silica: mechanisms for the formation of aluminum-nitrogen bonds and the elimination of methyl groups bonded to aluminum.” //Chem. Mater.//, 1991, //3// (5), pp 953–960. []
 * 13) Interrante, L., et al. “Preparation and Properties of Aluminum Nitride Films Using an Organometallic Precursor.” J. Electrochem. Soc. 1989volume 136, issue 2, 472-478. []
 * 14) Sauls, F., et al. “Ammine(trimethyl)aluminum formation and pyrolytic methane loss: thermodynamics, kinetics, and mechanism.” //Inorg. Chem.//, 1990, //29// (16), pp 2989–2996. []
 * 15) Ault, B. and Laboy, J. “Matrix isolation study of the 1:1 molecular complexes of (CH3)3Al with group V and VI bases.” Journal of Molecular Structure, Volume 300, 3 December 1993, Pages 351–362. []
 * 16) Ault, B. “Matrix isolation study of the reaction of trimethylaluminum with hydrogen sulfide and mercaptans: synthesis and spectra of molecular aluminum sulfide.” //J. Phys. Chem.//, 1994, //98// (1), pp 77–80. []
 * 17) Ault, B. “Infrared spectra of argon matrix-isolated alkali halide salt/water complexes.” //J. Am. Chem. Soc.//, 1978, //100// (8), pp 2426–2433. []
 * 18) Carpenter, J. and Ault, B. “Infrared matrix isolation characterization of aminoborane and related compounds.” //J. Phys. Chem.//, 1991, //95// (9), pp 3502–3506. []
 * 19) Ault, B. “Matrix isolation study of the reactions of trimethylaluminum with ammonia.” //The Journal of Physical Chemistry// 1992 //96// (20), 7908-7912. []
 * 20) Lias, S., et al. “Evaluated Gas Phase Basicities and Proton Affinities of Molecules; Heats of Formation of Protonated Molecules.” J. Phys. Chem. Ref. Data 13, 695 (1984). []