Strong-field ionization of water: Nuclear dynamics revealed by varying the pulse duration

Polyatomic molecules in strong laser fields can undergo substantial nuclear motion within tens of femtoseconds. Ion imaging methods based on dissociation or Coulomb explosion therefore have difficulty faithfully recording the geometry dependence of the field ionization that initiates the dissociation process. Here we compare the strong-field double ionization and subsequent dissociation of water (both ${\mathrm{H}}_2\mathrm{O}$ and ${\mathrm{D}}_2\mathrm{O}$) in 10-fs and 40-fs 800-nm laser pulses. We find that 10-fs pulses turn off before substantial internuclear motion occurs, whereas rapid internuclear motion can take place during the double ionization process for 40-fs pulses. The short-pulse measurements are consistent with a simple tunnel ionization picture whose predictions help interpret the motion observed in the long-pulse measurements.

Figure 1

(a) A schematic rendering of the ionization process and detection geometry. A focused femtosecond laser pulse (green), with polarization ${\widehat{\epsilon }}_{\mathrm{laser}}$ along the $\widehat{x}$ direction, removes electrons from the three highest occupied molecular orbitals (HOMO, HOMO-1, and HOMO-2) of neutral water. Preferential ionization of each orbital in a particular orientation with respect to the laser polarization induces geometric alignment as indicated by the molecular axis (${\widehat{x}}_{\mathrm{m}}/{\widehat{y}}_{\mathrm{m}}/{\widehat{z}}_{\mathrm{m}}$) that is parallel to ${\widehat{\epsilon }}_{\mathrm{laser}}$. After ionization, ionic fragments are accelerated toward the detector screen by an electrostatic lens stack (orange) that applies a constant electric field ${\widehat{E}}_{\mathrm{VMI}}$ in the $\widehat{z}$ direction. The net result of this scheme is an image that forms on the detector plane capturing the $\widehat{x}$ and $\widehat{y}$ momenta of each fragment as well as a distribution of fragment arrival times indicative of the $\widehat{z}$ momentum. (b) The $\widehat{x}$ and $\widehat{y}$ momentum distribution of ${\mathrm{D}}^{+}$ ions captured in coincidence with ${\mathrm{OD}}^{+}$, following ionization with a 40-fs pulse. The laser polarization axis is drawn as a green arrow. (c) The same distribution as in (b), but following ionization with a 10-fs pulse. (d) The kinetic energy release (KER) distribution plotted for both pulse durations. (e) The angular distribution of ${\mathrm{D}}^{+}$ ions where $\theta =\arccos ({\widehat{\epsilon }}_{\mathrm{laser}}·{\widehat{p}}_{{\mathrm{D}}^{+}}$) represents the angle between the laser polarization axis and the ${\mathrm{D}}^{+}$ vector momentum ($p_{{\mathrm{D}}^{+}}$), as drawn in panel (b).

Figure 2

(a) 2D momentum distributions of the fragments in the three-body breakup channel of the heavy water dication, ${\mathrm{D}}^{+}/{\mathrm{D}}^{+}/\mathrm{O}$, following ionization by the 40-fs pulse. Here, the bisector between the vector momenta of the two ${\mathrm{D}}^{+}$ ions is fixed to be along the horizontal such that the vector momentum of each ion can be decomposed into a component parallel to the bisector $p_{z_{\mathrm{m}}}$ and a component orthogonal to the bisector $p_{y_{\mathrm{m}}}$. (b) The same as in (a) but following ionization with the 10-fs pulse. (c) The KER distribution for the case of 40 fs (solid red line) and 10 fs (dashed blue line). Light red shading indicates the KER filter used to discriminate between the ${\mathrm{D}}^{+}/{\mathrm{D}}^{+}/\mathrm{O}$ dication channel (solid red line, shaded) and contamination from the higher-KER ${\mathrm{D}}^{+}/{\mathrm{D}}^{+}/{\mathrm{O}}^{+}$ trication channel (light red markers, unshaded). Here, the 40-fs data were been resized to match the peak height of the 10-fs data. (d) The distribution in angle $\theta =\arccos ({\widehat{\epsilon }}_{\mathrm{laser}}·{\widehat{y}}_{\mathrm{m}}$), measured between the laser polarization axis (${\widehat{\epsilon }}_{\mathrm{laser}}$) and the ${\widehat{y}}_{\mathrm{m}}$ molecular axis. (e) The distribution in angle $\beta =\arccos ({\widehat{p}}_{{\mathrm{D}}^{+}}^{\left(1\right)}·{\widehat{p}}_{{\mathrm{D}}^{+}}^{\left(2\right)}$), measured between the two ${\mathrm{D}}^{+}$ vector momenta ($p_{{\mathrm{D}}^{+}}^{\left(1\right)}$ and $p_{{\mathrm{D}}^{+}}^{\left(2\right)}$), as drawn in panel (a).

Figure 3

(a) A schematic drawing of a definite laser polarization (${\widehat{\epsilon }}_{\mathrm{laser}}$) plotted in the molecular frame with each of its projections along the molecular axes (${\widehat{x}}_{\mathrm{m}}/{\widehat{y}}_{\mathrm{m}}/{\widehat{z}}_{\mathrm{m}}$) labeled. (b) The angle-dependent ionization yield for ${\mathrm{D}}_2{\mathrm{O}}^{+}$ (calculated via TD-RIS) represented as a 3D probability distribution of the laser polarization in the molecular frame as seen in (a). (c) The time-resolved alignment distribution of ${\mathrm{D}}_2{\mathrm{O}}^{+}$ as characterized by the value of the probability distribution $P(t,\widehat{\epsilon })$ along each molecular axis: $\widehat{\epsilon }={\widehat{x}}_{\mathrm{m}}$ (solid line), $\widehat{\epsilon }={\widehat{y}}_{\mathrm{m}}$ (dashed line), and $\widehat{\epsilon }={\widehat{z}}_{\mathrm{m}}$ (dotted line). Here, $P(t,\widehat{\epsilon })$ was initiated with the distribution displayed in (b) at $t/\tau =-0.5$ where $t/\tau =0$ is the time at which the molecule experiences peak field intensity, $I_0=600$ $\mathrm{TW}/{\mathrm{cm}}^2$, and $\tau =40$ fs is the ionizing pulse duration (FWHM). (d) The 3D alignment distribution of ${\mathrm{D}}_2{\mathrm{O}}^{+}$ after evolving in a strong field from $t/\tau =-0.5$ to $t/\tau =0.5$. As in (b), this distribution is represented by the 3D probability distribution of the laser polarization in the molecular frame as seen in (a). Note that the limits of all three axes are double those of (b) to capture the whole distribution. (e) The same as in (c) but for $\tau =10$ fs and $I_0=400$ $\mathrm{TW}/{\mathrm{cm}}^2$. (f) The same as in (d) but for $\tau =10$ fs and $I_0=400$ $\mathrm{TW}/{\mathrm{cm}}^2$. Note that the limits of all three axes are identical to those of panel (b).

Figure 4

The angular distribution of both two-body and three-body decay channels, as measured by the angle between the assumed ${\widehat{y}}_{\mathrm{m}}$ molecular axis and the laser polarization ${\epsilon }_{\mathrm{laser}}$. For both (a) the long-pulse data (in red) and (b) the short-pulse data (in blue), the two-body dissociation channel (${\mathrm{D}}^{+}/{\mathrm{OD}}^{+}$) is plotted as a thin, solid line and the three-body dissociation channel (${\mathrm{D}}^{+}/{\mathrm{D}}^{+}/\mathrm{O}$) is plotted as a thick, dashed line. For the two-body channel, $\theta$ represents the angle between the dissociation axis and the laser polarization. For the three-body channel $\theta =\arccos ({\widehat{\epsilon }}_{\mathrm{laser}}·{\widehat{y}}_{\mathrm{m}}$).

Figure 5

The KER distribution for the two-body decay pathway of doubly ionized (a) water ${\mathrm{H}}_2{\mathrm{O}}^{2+}\to {\mathrm{H}}^{+}/{\mathrm{OH}}^{+}$ and (b) deuterated water ${\mathrm{D}}_2{\mathrm{O}}^{2+}\to {\mathrm{D}}^{+}/{\mathrm{OD}}^{+}$. In each case, the 40-fs data are plotted in solid red lines, and the 10-fs data are plotted in dashed blue lines. Both long and short-pulse data are further subdivided between dissociation aligned along the laser polarization axis: ${\theta }_{\parallel }=0^{\circ }\pm {30}^{\circ }$ (thick lines) and dissociation aligned perpendicular to the laser polarization axis: ${\theta }_{\perp }={90}^{\circ }\pm {30}^{\circ }$ (thin lines). All distributions are independently normalized to integrate to 1.