Research ArticleAPPLIED SCIENCES AND ENGINEERING

Electrospray sample injection for single-particle imaging with x-ray lasers

See allHide authors and affiliations

Science Advances  03 May 2019:
Vol. 5, no. 5, eaav8801
DOI: 10.1126/sciadv.aav8801

Figures

  • Fig. 1 ES aerosol injector.

    (A) Design of the ES aerosol injector. In the aerosolization chamber, the ES nebulizer generates droplets that are neutralized with a 210Po alpha emitter. The ES nebulizer is operated in an atmosphere of N2 and CO2 at 1 bar. The aerosol is transported through two nozzle-skimmer assemblies, where excess gas is pumped away. At a reduced pressure of 1 to 10 mbar, the aerosol enters the aerosol lens stack, which focuses it to a narrow particle beam entering the experimental chamber, which is held at a pressure of 10−6 to 10−5 mbar to match requirements for XFEL imaging. (B) Size distributions of initial droplets for ES (green) and GDVN (blue) aerosols determined by RSM (top) and XFEL diffraction (bottom). The results of the two sizing methods are comparable within the limits of reproducibility expected for the manually manufactured nozzles and variations in operational parameters, such as pressures, voltage, and flow rate. (C) RSM size distributions of aerosolized particles from carboxysome sample (purple) and from its buffer solution (red). Data collected on electrosprayed particles are shown in the first panel (median, 95 nm; FWHM, 14 nm), and data collected on particles injected by GDVN at two different pressure configurations (Table 2) are shown in the second (median, 102 nm; FWHM, 17 nm) and third panels (median, 105 nm; FWHM, 17 nm). Dashed lines indicate the detection limit.

  • Fig. 2 XFEL diffraction data of biological particles injected with the ES aerosol injector.

    (A) Simulated and measured diffraction patterns of carboxysomes and (B) their size distribution (median, 90 nm; FWHM, 13 nm) determined from the measured diffraction patterns. (C and D) Simulated and measured diffraction patterns of TBSV particles (C, singles; D, clusters of two) and (E) their size distribution (median, 30 nm; FWHM, 1 nm) determined from the measured diffraction patterns. Insets in (A), (C), and (D) show 2D projection images reconstructed from the respective diffraction patterns. The edge length of the insets corresponds to 220 nm.

  • Fig. 3 Injection of Rubisco proteins.

    (A) Radial averages of 14,361 background-subtracted diffraction patterns recorded during injection of sample (1), 14,343 during injection of buffer solution (2), 14,367 during injection of only gas (3), and 6993 during a dark run (4). (B) Diffraction patterns of two intense sample hits. (C) Radial averages (orange lines) of the diffraction patterns shown in (B) and fits (black lines) to a sphere model that best match the data. Light orange areas indicate the confidence intervals of the data (±1 SD). The fit values for intensity and sphere diameter are annotated. (D) STEM image of Rubisco proteins injected onto a TEM sample support film. Detected particles are highlighted in red. (E) The red histogram shows the distribution of particle diameters derived from (D). The black line shows the fit of our droplet occupancy model to the data. The good match indicates that the electrosprayed proteins were successfully transferred into the interaction region. (F) DMA data of electrosprayed Rubisco proteins at three concentrations. Our droplet occupancy model (black) was fitted to the measured size histograms (red). The agreement shows that, by changing concentration, we specifically control the protein cluster composition.

  • Fig. 4 PRTFs for reconstructed projection images shown in Fig. 2 (A, C, and D).

    The dashed lines indicate the value e−1, often used as threshold for judging the reproducibly of the retrieved phases.

Tables

  • Table 1 Aerosolization parameters.

    Characteristic parameters for sample aerosolization with ES and a GDVN assuming an average droplet occupancy of 1.

    Sample
    flow rate
    Droplet
    size
    Sample
    concentration
    Particle rate
    ES0.06 μl/min150 nm5 × 1014/ml5.7 × 108/s
    GDVN2 μl/min1000 nm2 × 1012/ml0.6 × 108/s
  • Table 2 Datasets used for this study. ID, inner diameter; n.a., not available.

    Datasets used for this study. ID, inner diameter; n.a., not available..

    MeasurementDataset
    name
    Run
    #
    Photon
    energy (eV)
    Detector
    distance (mm)
    Sample
    concentration
    Liquid flow
    (μl/min)
    Gas flow
    (SLM)
    Capillary
    ID
    (μm)
    Voltage
    (kV)
    Sucrose (ES)
    (Fig. 1B, bottom panel)
    AMO
    L3416
    386703705 v/v %0.06CO2 0.15
    N2 1.30
    402.20
    Sucrose (GDVN)
    (Fig. 1B, bottom panel)
    AMO
    L3116
    1428003700.1 v/v %0.7He 0.4n.a.n.a.
    Sucrose (ES)
    (Fig. 1B, top panel)
    RSM337n.a.n.a.12 v/v %0.06CO2 0.20
    N2 1.45
    n.a.n.a.
    Sucrose (GDVN)
    (Fig. 1B, top panel)
    RSM385n.a.n.a.0.1 v/v %0.44He 0.4n.a.n.a.
    Carboxysomes (ES)
    (Fig. 1C, top panel)
    RSM301n.a.n.a.1 × 1013 ml−10.06CO2 0.15
    N2 1.20
    402.50
    Carboxysomes (GDVN 1)
    (Fig. 1C, middle panel)
    RSM305n.a.n.a.1 × 1012 ml−10.59He 0.4n.a.n.a.
    Carboxysomes (GDVN 2)
    (Fig. 1C, bottom panel)
    RSM309n.a.n.a.1 × 1012 ml−10.59He 0.6n.a.n.a.
    Carboxysomes (ES)
    (Fig. 2, A to C )
    AMO
    L3416
    51–568003701 × 1013 ml−10.06CO2 0.15 N2 1.30402.15
    TBSV (ES)
    (Fig. 2, C to E)
    AMO
    L3416
    132–135 137–1428002593 × 1014 ml−10.06CO2 0.15 N2 1.30302.25
    Rubisco (sample)
    (Fig. 3A, panel 1)
    AMO
    L3416
    2528001308 × 1014 ml−10.06CO2 0.15 N2 1.30302.25
    Rubisco (buffer)
    (Fig. 3A, panel 2)
    AMO
    L3416
    203800130n.a.0.06CO2 0.15 N2 1.30302.15
    Rubisco (gas)
    (Fig. 3A, panel 3)
    AMO
    L3416
    256800130n.a.0.00CO2 0.15 N2 1.30n.an.a.
    Rubisco (dark)
    (Fig. 3A, panel 4)
    AMO
    L3416
    257800130n.a.n.a.n.a.n.a.n.a.

Navigate This Article