SIMS of trna Molecules Encapsulated Between Free-Standing Graphene Sheets

Size: px

Start display at page:

Download "SIMS of trna Molecules Encapsulated Between Free-Standing Graphene Sheets"

James Spencer
6 years ago
Views:

1 SIMS of trna Molecules Encapsulated Between Free-Standing Graphene Sheets Running title: SIMS of Graphene-Encapsulated trna Running Authors: D. S. Verkhoturov et al. Dmitriy S. Verkhoturov Department of Chemistry, Texas A&M University, College Station, TX , USA Sheng Geng Department of Chemistry, Texas A&M University, College Station, TX , USA Stanislav V. Verkhoturov Department of Chemistry, Texas A&M University, College Station, TX , USA Hansoo Kim Microscopy and Imaging Center, Texas A&M University, College Station, TX 77843, USA Emile A. Schweikert a) Department of Chemistry, Texas A&M University, College Station, TX , USA a) Electronic mail: In this study we used Cluster-SIMS method to investigate the preserved transfer Ribonucleic Acid (trna) encapsulated between two free-standing graphene sheets. Single impacts of 50 kev C projectiles generated the emission of trna fragment ions in the transmission direction for mass selection and detection in the TOF mass spectrometer. RNA is extremely unstable and prone to rapid enzymatic degradation by Ribonucleases. Employing graphene to isolate RNA from the environment, we prevent the aforementioned process. Encapsulation was achieved by drop casting a solution of 1

2 trna, prepared using deuterated water, onto one graphene sheet and covering it with another. Event-by-event bombardment/detection mode allowed us to use co-localization analysis method to characterize the trna and its immediate environment. We found that upon drying, trna agglomerated into nano-structures ~60 nm in diameter via formation and subsequent drying of aqua cells. The trna nano-agglomerates had a density of ~42 structures per µm 2 with coverage of ~12% of the surface area. In addition trace amounts of water remained mostly around the trna nano-agglomerates probably in the form of hydration. I. INTRODUCTION One of the challenges in the application of Secondary Ion Mass Spectrometry, SIMS, is the requirement for samples to be vacuum compatible. The purpose of this study was to assess the feasibility of examining hydrated samples via encapsulation using the aqua cell technique. Aqua cells were developed by the Alivisatos group to keep analyte hydrated under high vacuum conditions in Transmission Electron Microscopy (TEM) 2. The term aqua cell refers to nano-sized water pockets trapped between two graphene sheets to isolate it from the vacuum and inhibit their evaporation. The aqua cell TEM experiment required the observation of a single aqua cell with short data collection time per run. In order for the aqua cell SIMS experiment to be successful a sufficient amount of aqua cells have to be present in the sampling area during the entire collection time. The purpose of the present study was to apply the aqua cell technique in order to study non-fixed biological samples with SIMS. For this model experiment transfer Ribonucleic Acid (trna) was chosen for several reasons. First, it is a molecule which can also be viewed as an L shaped nano-object that is ~6 nm by 3 nm in size 14. It consists of 4 types of monomers containing different nitrogenous bases (Adenine, Cytosine, Guanine and Uracil) and has secondary and tertiary structure. Second, trnas belong to an extremely important class of molecules which are notorious for their instability primarily 2

3 due to rapid enzymatic degradation by ribonucleases. 1 These are small and stable enzymes which are readily found in the environment including in a laboratory setting. Thus in addition to hydration of trna molecules, this experiment can also test their preservation from the environment via encapsulation in a graphene sandwich 2 as a byproduct of aqua cell formation. For characterization of the trna encapsulated in graphene sandwich as well as its degree of hydration, a custom built SIMS instrument was used with event-by-event bombardment/detection mode where secondary ions are detected in the transmission direction. 4 The coupling of aqua cell SIMS with the ability of this custom instrument to investigate the co-localization of emitted ions opens new possibilities for studying the interaction of ion cofactors with ribozymes in an aqueous environment in native state conditions. II. EXPERIMENTAL A. Instrument The experiments were run on a custom-built Cluster-SIMS instrument (Fig. 1) consisting of a C 60 ion source and linear time-of-flight, ToF, mass spectrometer operating in the event-by-event bombardment/detection mode. 3 Bombardment was at the level of individual C impacting surface area of ~10 4 µm 2 at 50 kev (impact repetition rate of 1000 impacts/sec) with separate recording of negatively charged secondary ions and electrons emitted in transmission direction from each collision. The detected electrons are used as a start signal to initiate the acquisition of a mass spectrum for a particular 2+ projectile impact event. Target emission area is larger than the diameter of C 60 projectile (0.7 nm). For instance, Molecular Dynamics simulation 4 shows that, in the case of a 4 layer graphene target, the emission area is ~6 nm in diameter. It should be noted that 50 kev C projectiles generate secondary ions in transmission mode only from targets which are less than ~5 nm in thickness. Co-emitted secondary ions from single 3

4 impacts were detected by an eight-anode detector and analyzed by a multichannel timeto-digital converter. The event-by-event bombardment/detection mode allows to select specific impacts, where emitted ions were detected in coincidence 5 with a particular ion of interest. In the present case the impacts involving trna nano-agglomerates were selected at the exclusion of signals from the target holder and support. The data are recorded and processed event-by-event with the custom data acquisition program SAMPI (Surface Analysis and Mapping of Projectile Impacts). 6 FIG. 1. Schematic of experiment: (a) objective lens for secondary ions and electrons, (b) magnetic prism for redirection of electrons toward (c) imaging electron optics, (d) position sensitive detector consisting of dual microchannel plate, phosphor screen and CMOS camera, (e) dual microchannel plate, (f) 8 anode detector. B. Sample Preparation 4

5 The graphene sandwich was prepared using a procedure 2 that we modified as described below. Lyophilized Baker s Yeast transfer RNA (trna) powder was purchased from MP Biomedicals. It was dissolved in deuterated water at 0.07 mg/ml. The presence of Deuterium allows the tracking of the water and gathering of information regarding the co-emission of the Deuterium ion with ions emitted from the analyte and/or substrate. 1 µl of the trna solution (containing 0.07 µg of trna) was drop-cast onto a 4 layer graphene sheet attached to perforated Silicon Nitride substrate (PELCO 3-5 layer graphene on Holey Silicon Nitride with 2.5µm holes in a 0.5 x 0.5 mm window purchased from Ted Pella). Then another similar 4 layer graphene on perforated Silicon Nitride substrate was attached on top of the assembly with graphene facing the drop of the analyte solution. The resulting graphene sandwich, after 10 minutes of drying, would collapse with the generation of nano-aqua cells in the interface area 2. Once inserted into the vacuum, the aqua cells dry within 30 minutes with production of trna nanoagglomerates. III. RESULTS AND DISCUSSION The mass spectrum of ions emitted from the graphene sandwich (Fig. 2) consists of carbon clusters along with trna fragments such as the deprotonated Adenine. 5

6 I/N eff 4.0x10-4 a 3.0x x x10-4 Ribose - COH m=103 Adenine -H m=128 m=134 a (Adenine-H)+PO m= x10-4 I/N 0 2.0x x x10-4 C 10 C 9 C 8 (OH) - PO Mass (amu) b FIG. 2. Mass spectra of graphene sandwich with encapsulated trna nano-agglomerates obtained with 8 million impacts of 50 kev C with detection in the negative mode; (a) in coincidence with emission of PO 3 - ion; (b) total mass spectrum. To create a reference spectrum for trna, a bulk trna sample must be analyzed. 4+ Bulk trna was deposited onto Au substrate and bombarded with 520 kev Au 400 projectiles. Another similar custom made Cluster SIMS device 9 from the laboratory was used. The ion optics of both devices are identical. The two cluster projectiles have similar energy per atom (~1 kev). If one compares mass spectra obtained by 50 kev C and 520 kev Au for the same target in reflection mode, the spectra are qualitatively similar with yields of molecular ions 3-10 times larger for Au projectile. The resultant mass spectrum (Fig. 3) shows which masses correspond to trna fragments. 6

7 2.0x x a I/N eff 1.0x10-3 A(-H), x Au x10-4 I/N 0-1.0x x x mass (amu) b FIG. 3. Reference mass spectra obtained in the negative ion detection mode using 520 kev Au projectiles on a bulk trna target laid on Au substrate; (a) mass spectrum in coincidence with m=128 trna fragment (Cytosine+OH); (b) total mass spectrum. The trna fragments are not dominant in the total spectrum (Fig. 2b) measured for the graphene sandwich due to the small amount of analyte. The coincidence spectrum (Fig. 2a) was computed where the projectile impacts on trna-enriched areas were selected (coincidence with trna phosphate groups, m=79). This spectrum shows a strong increase in peaks pertaining to trna fragments as per the reference spectrum. Below it is shown that this enhancement indicates agglomeration of trna. One should note that intensity values of peaks has been normalized on the total number of impacts for the total mass spectrum and effective number of impacts for the coincidence mass spectrum. The intensities were measured using a modern time-to-digital-converter where each channel corresponds to 120 ps. Thus the width of a peak corresponds to a large number channels. For instance, in the case of (Adenine-H) peak, the number of channels 7

8 is 64. As a result, the normalized intensities on mass spectra have relatively small values. However when yields of ions are calculated, the intensities are integrated for all of the channels belonging to a particular peak (area under the peak). That is to say that ion yields can have significant differences even if the maximum normalized intensity difference is small. To quantify the enhancement of peaks in the coincidental spectra, parameters 7 total yield, Y tot, and effective yield, Y eff, were used as described below through the example of coincidental emission of deprotonated Adenine (m=134) with PO 3 - (m=79), both of which are trna fragments: Y A tot = I A N 0 Equation 1 where I A is the integrated intensity under the deprotonated Adenine peak while N 0 is the total number of projectile bombardment events. The total ensemble of events consists of sub ensembles related to the agglomeration of surface components (in this case trna agglomerates). Similarly: Y A eff = I A N eff Equation 2 where Y A eff is effective yield of deprotonated Adenine and N eff is the sub ensemble of impacts on trna agglomerates. The effective yield of PO 3, Y eff PO3, can be expressed with a similar formula, Y eff PO3 = I PO3. Co-emission of Adenine and PO 3 from trna nano- N eff agglomerates can be described using coincidental effective yield, Y eff A,PO3 = I A,PO3 N eff, which can be factorized 8 as shown below: Y eff A,PO3 = Y eff eff A Y PO3 Equation 3 8

9 By solving for Y A eff and then substituting I PO3 N eff and I A,PO3 N eff for Y eff PO3 and Y eff A,PO3 respectively an expression is obtained which is only dependent on measured intensities that are known from the experiment: Y A eff = I A,PO3 I PO 3 Equation 4 Using Equations 1 and 4 the ratio Y eff tot A Y A can be computed. This ratio represents the enhancement factor mentioned earlier, which shows the degree of agglomeration of trna incorporated into the graphene sandwich. Table 1 has the calculated enhancement factors for several emitted ions in coincidence with PO 3, PO 2 and Deuterium, D. Y eff /Y tot Coincidence ion D PO 2 PO 3 trna fragment m=103 Adenine trna fragment m=181 C 3 C 7 D 2.2 ± ±0.4 * * * 1.2 ± ±0.2 PO ± ± ± ± ± ± ±0.04 PO ± ± ± ± ± ± ±0.04 Table 1. Enhancement factors (Y eff Y tot ) for several ions in coincidence with Deuterium, PO 2, PO 3. *values could not be determined due to a low number of coincidence events (comparable with background signal). Generally, if Y eff Y tot > 1 then the number of impacts on a particular analyte, N eff, is smaller than the total number of impacts, N 0, indicating degree of agglomeration 9. The higher the ratio the greater the degree of agglomeration. In the case of trna, when 9

looking at the enhancement factor of Adenine in coincidence with PO 3, a value of 3.2 for Y eff Y tot represents a high degree of agglomeration, i.e. the trna molecules are agglomerated into small structures.

From the images (Fig 4a) the average size of the trna agglomerates was calculated to be ~60 nm. a) 0.5 µm b) C O Cu Na Si P Ca Cu 20 nm c) C 100 nm O Si Cu FIG. 4. TEM micrographs and EDS spectra were taken using FEI Tecnai G2 F20 ST FE TEM/EDS device.

The micrographs show how the perforated regions of the Silicon Nitride membranes overlap creating windows of various sizes consisting of graphene sandwich and the

10 looking at the enhancement factor of Adenine in coincidence with PO 3, a value of 3.2 for Y eff Y tot represents a high degree of agglomeration, i.e. the trna molecules are agglomerated into small structures. If trna were evenly spread out on the graphene surface, then these ratios would be close to unity. The finding is confirmed by TEM images and EDS (Fig. 4). From the images (Fig 4a) the average size of the trna agglomerates was calculated to be ~60 nm. a) 0.5 µm b) C O Cu Na Si P Ca Cu 20 nm c) C 100 nm O Si Cu FIG. 4. TEM micrographs and EDS spectra were taken using FEI Tecnai G2 F20 ST FE TEM/EDS device. (a) TEM micrographs of the trna nano-agglomerates within the graphene sandwich. The micrographs show how the perforated regions of the Silicon Nitride membranes overlap creating windows of various sizes consisting of graphene sandwich and the trna agglomerates; EDS spectra of (b) trna nano-agglomerate and (c) of the area of graphene outside of trna nano-agglomerates. 10

11 To better understand the characteristics of the agglomeration, surface density and coverage information can be extracted using parameters Y eff and Y tot (Eq. 1, 2). Using the following equation the coverage of trna over the surface of the sample can be calculated: % Coverage of trna = 100 N eff N 0 = N 0 I PO2 I PO3 I PO 2, PO3 Equation 6 PO 3 and PO 2 trna fragments were used in the calculation because their high ionization efficiencies improve their yields, producing significant peaks, which increases accuracy of the result. The estimated coverage from Equation 6 is ~12%. It should be noted that the amount of deposited trna would correspond to about 4.4 layers of trna molecules if they spread evenly within the graphene sandwich. However, in the experiment it was not the case as the coverage of trna agglomerates was ~12%. It may be postulated that evaporation of water, accompanied by rapid collapse of the opposing Silicon Nitride membranes, causes local crystallization of trna. In contrast, the smaller portion of the total amount of trna is trapped in small aqueous pockets between the graphene sheets inside the pores (especially where the holes from the top and bottom Silicon Nitride membranes overlap). As the formed aqua cells continue to evaporate water, the contained trna eventually aggregates into agglomerates which we then observed. Using coverage of trna and the average size of each agglomerate (S aggl ) one can find their surface density using the equation below: trna nano aggl ρ( ) = 1 Coverage Equation 7 μm 2 S aggl The trna nano-agglomerate surface density was determined to be ~42 agglomerates per µm 2. That means there are on average ~42 trna nano-agglomerates, which are ~60 nm 11

12 in diameter, within each µm 2 of graphene sandwich where the pores of the Silicon Nitride membranes overlap. These estimates are consistent with the observations from the TEM images (Fig. 4a). From the agglomeration pattern of trna into small round structures, it can be inferred that the trna solution, upon evaporation, started forming nano-aqua cells. While most of the water was lost after a few minutes under high vacuum conditions, some of it remained as seen from the mass spectrum in Figure x x10-5 m=2 Deuterium trna nano agglomerates in graphene sandwich a I/N 0 1.0x x x10-6 I/N 0 1.0x x10-5 b Control sample: 4 layer graphene on Si 3 N 4 2.0x Mass (amu) 2+ FIG. 5. Mass spectra obtained in the negative ion detection mode using 50 kev C 60 projectiles showing Deuterium peak for (a) the trna graphene sandwich sample and for (b) the graphene-only control sample. The ratio of the Deuterium yield from the trna graphene sandwich sample to the control (graphene-only) was 2.6. Yield of the deuterium peak in the graphene sandwich spectrum is 2.6 times the yield in the control (graphene-only) spectrum. Table 1 shows that Y eff tot D Y D in coincidence with PO 3 has a value of 2.1. This indicates agglomeration of the remaining water molecules around trna, probably in the form of hydration. In contrast, the coemission of graphene fragments, e.g. C 7, with trna fragments, e.g. PO 3, has an 12

13 enhancement factor of 1.0 (Table 1). Thus the effective yield of carbon cluster ions measured for the sub-ensemble of impacts on trna agglomerates is equal to the total yield of those cluster ions. Indeed, trna agglomerates are surrounded by graphene on both sides. When a projectile strikes a trna agglomerate it will always cause emission from the homogeneous surface of graphene located in the transmission direction. Therefore, the presence of a thin layer of trna between graphene sheets does not affect the emission of graphene fragments. It should be noted that in addition to hydration of trna, deuterium could also come from the Hydrogen/Deuterium exchange during sample preparation, however, there is evidence to suggest that this process is not dominant. In the mass spectrum corresponding to the graphene sandwich there is a significant enhancement of the m=134 peak corresponding to deprotonated Adenine (A- H). Previously our lab investigated phages 10 which contained RNA where no deuterated water was used at any point during sample preparation. Using the same device and methodology but with detection in the reflection direction, the mass spectrum of the phages was taken and analyzed. The results for this experiment also showed an abundant peak m=134, corresponding to A-H, which was co-emitted from impacts on the phages. In case of H/D exchange, both Hydrogens of the amino group should get exchanged (after several dissociation/association events) thus abundant peaks of singly deprotonated Adenine with masses 135 and 136 would be expected. However, this was not the case. The lack of H/D exchange can be due to the presence of hydrogen bonds preventing the process from occurring in the time of the experiment. This suggests that the higher order structure of the trna molecules was probably preserved. 13

14 IV. CONCLUSIONS SIMS, in the event-by-event bombardment/detection mode, enables the characterization of trna molecules encapsulated in a graphene sandwich. We found that trna agglomerated into nano-objects due to the formation and subsequent drying of aqua cells. The surface density of the trna nano-agglomerates was ~42 agglomerates per µm 2. In addition, trna was hydrated as some water molecules were retained by the trna graphene sandwich (Fig. 5). Aqua cells open new possibilities for studying the interaction of ion cofactors with ribozymes in an aqueous environment while they are in their native conformation. To perform the experiment we need longer lived aqua cells. The Alivisatos 2 group, whose graphene sandwich preparation procedure we modified, used TEM imaging to study the movement of gold nanoparticles linked by DNA and floating within the aqua cells. For their experiments, following a single aqua cell for a span of 30 minutes was sufficient. However, in our case where statistics on large amounts of aqua cells are needed, the collection time will be longer. Quasi-ambient pressure SIMS 11 is a possible approach to reduce the evaporation rate of aqua cells. Finally, one of the drawbacks of using graphene to produce aqua cells is the permeation 12 of water through the partially oxidized regions of the membrane. To avoid the issue, prior to graphene sandwich preparation, the graphene sheets should be reduced by annealing 12 them in hydrogen atmosphere. Another possibility is the use of alternative 2D membranes, e.g. Boron Nitride. 13 Finally, future experiments should include acquisition of data in the positive ion mode. 14

15 ACKNOWLEDGMENTS This work was supported by the National Science Foundation (Grant CHE ). 1 L. Cui, Z. Chen, Z. Zhu, X. Lin, X. Chen, C. J. Yang, Anal. Chem. 85, 2269 (2013). 2 Q. Chen, J. M. Smith, J. Park, K. Kim, D. Ho, H. I. Rasool, A. Zettl, and A. P. Alivisatos, Nano Letters 13, 4556 (2013). 3 S. V. Verkhoturov, M. J. Eller, R. D. Rickman, S. Della-Negra, E. A. Schweikert, J. Phys. Chem. C 114, 5637 (2010). 4 S. V. Verkhoturov, S. Geng, B. Czerwinski, A. E. Young, A. Delcorte and E. A. Schweikert, J. Chem. Phys (2015) 5 M. A. Park, K. A. Gibson, L. Quinones, E. A. Schweikert, Science 248, 988 (1990). 6 M. J. Eller, S. V. Verkhoturov, S. Della-Negra, E. A. Schweikert, Rev Sci Instrum. 84, (2013). 7 S. Rajagopalachary, S. V. Verkhoturov, and E. A. Schweikert, Nano Letters 8, 1076 (2008). 8 R. D. Rickman, S. V. Verkhoturov, E. S. Parilis, E. A. Schweikert, Phys. Rev. Letters 92, 7601 (2004). 9 C.-K. Liang, S.-T. Chang, S. V. Verkhoturov, L.-C. Chen, K.-H. Chenb, E. A. Schweikert, Int. J. Mass Spectrom. 370, 107 (2014). 10 C.-K. Liang, M. J. Eller, S. V. Verkhoturov, E. A. Schweiker, J. Am. Soc. Mass Spectrom. 26, 1259 (2015). 11 T. Seki, M. Kusakaria, M. Fujiib, T. Aokic, J. Matsuo, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, (2015), doi: /j.nimb R. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva, A. K. Geim, Science 335, 442 (2012). 15

16 13 J. K. Hite, Z. R. Robinson, C. R. Eddy Jr., and B. N. Feigelson, ACS Appl. Mater. Interface 7, (2015). 14 E. Westhof, P. Dumas, D. Moras, Acta Crystallogr. A 44, 112 (1988). 16

Characterization of individual free-standing nanoobjects by cluster SIMS in transmission

Characterization of individual free-standing nanoobjects by cluster SIMS in transmission Running title: Characterization of individual free-standing nano-objects by cluster SIMS in transmission Running