Report

Selective Bond Dissociation and Rearrangement with Optimally Tailored, Strong-Field Laser Pulses

See allHide authors and affiliations

Science  27 Apr 2001:
Vol. 292, Issue 5517, pp. 709-713
DOI: 10.1126/science.1059133

Abstract

We used strong-field laser pulses that were tailored with closed-loop optimal control to govern specified chemical dissociation and reactivity channels in a series of organic molecules. Selective cleavage and rearrangement of chemical bonds having dissociation energies up to approximately 100 kilocalories per mole (about 4 electron volts) are reported for polyatomic molecules, including (CH3)2CO (acetone), CH3COCF3 (trifluoroacetone), and C6H5COCH3 (acetophenone). Control over the formation of CH3CO from (CH3)2CO, CF3 (or CH3) from CH3COCF3, and C6H5CH3 (toluene) from C6H5COCH3 was observed with high selectivity. Strong-field control appears to have generic applicability for manipulating molecular reactivity because the tailored intense laser fields (about 1013 watts per square centimeter) can dynamically Stark shift many excited states into resonance, and consequently, the method is not confined by resonant spectral restrictions found in the perturbative (weak-field) regime.

The control of organic chemistry with optical fields has been a long-sought-after goal. This objective goes beyond traditional photochemistry, where variation in the product distribution is obtained by tuning a monochromatic radiation source to excited states of a precursor molecule. Here, we focus on whether a suitably tailored laser pulse can guide a distribution of products to a specific channel, characterized by dissociation or rearrangement, as illustrated in Fig. 1. Successful control in this context implies cooperative interaction between a (perhaps complex) time-dependent electric field of the laser pulse and the precursor molecule's dynamics to alter the product distribution in a desirable and variable manner. The objective is to maximize the yield of a specified product by means of a suitable closed-loop laboratory learning procedure that determines the optimal laser pulse. The capability of controlling reactivity, especially of organic species, by this means could have implications in diverse industrial and biological or medicinal settings. For example, microelectronic lithography, the fabrication of gene chips, and photodynamic therapy could all benefit from highly controllable photochemistry.

Figure 1

A schematic of control over chemical reaction pathways with tailored, strong-field laser pulses. By using feedback from the observed reaction products, a learning algorithm can guide an evolving laser pulse shaper to achieve the desired product as best as possible. The experiments presented here focused on dissociation and dissociative rearrangement reactions for ketone molecules.

Since the introduction of the concept of teaching lasers to control photodynamics with closed-loop methods (1), the process has been used for optimization of laser-induced fluorescence (2), stimulated Raman emission from molecules (3), high harmonic generation (4), ultrafast semiconductor switching (5), and management of pulse propagation through nonlinear media (6). The first implementation of this scheme for dissociating bonds was demonstrated in groundbreaking work on organometallic systems (7) and has since been demonstrated in the cleavage of weakly bound cluster species (8). The control of organic reactivity, where bond strengths exceed ∼50 kcal/mol, remains an open and important area of optical control research.

We demonstrate a potentially universal, ultrafast laser excitation procedure to achieve control over molecular reactivity. This procedure is based on closed-loop learning of optimally shaped, strong-field, near-infrared pulses (centered around 800 nm) guided just by the desired product. A strong-field laser pulse (∼1013 W cm−2) implies that the associated electric field magnitude is on the order of the valence electron's binding energy in the molecule. This magnitude is a critical enabling feature for broad-scale application upon consideration of the bandwidth required for the effective control of chemical reactions. Weak-field, continuous wave pulses must be precisely tuned to a molecular resonance for coupling;Fig. 2A represents such an excitation scheme for most conventional photochemical reactions. Ultrafast laser systems allow ∼0.1 eV of bandwidth for a 25-fs pulse (for example, 775 to 825 nm for a near-infrared pulse). The relatively small bandwidth around the laser carrier frequency limits control in the weak-field case (Fig. 2B). However, with strong-field pulses, the nominal eigenstates of the molecule can be placed into resonance through a combination of Stark shifting (up to ∼10 eV in our experiment) (9) and multiphoton excitation (10 to 50 photons using the Ti:sapphire laser centered at 800 nm) (10, 11) (Fig. 2C). The latter two strong-field excitation mechanisms effectively remove bandwidth restrictions. The linear modulation of the strong field provided by the pulse shaper leads to control of the ensuing highly nonlinear molecular dynamics response. Transitions may be moved in and out of resonance as dictated by the pulse structure to optimally achieve the chemical objective. An example of three time-dependent electric fields (shaped laser pulses) producing different products is shown in Fig. 2D.

Figure 2

Laser pulse schemes are shown for creating coherent molecular motion with sources of ever-increasing bandwidth and selectivity capabilities. In each case, the precursor molecule begins in the ground vibronic state, So, and interacts with the electric field, ɛ(t). (A) In the weak field case, the molecule can only respond to the laser if the radiation is in resonance with the relevant molecular levels. (B) In this case, the shorter duration of the pulse helps to overcome resonance constraints. (C) The high-intensity, short-duration pulse creates sufficient dynamic power broadening of the molecular vibronic levels (denoted by yellow bands) and multiphoton excitation to overcome any restrictions on resonant excitation. (D) This case indicates how optimal tailoring of the pulse in (C) can be used to steer the molecular dynamics out one reaction channel versus another for high product selectivity. Controlled dynamics results from a combination of time-dependent ultrafast multiphoton excitation and Stark broadening (approximately several eV) of the molecular eigenstates. Selectivity is achieved by each particular intense laser pulse shape stimulating the necessary molecular dynamics to favor a desired product channel. Colors are as follows in the molecule models: red, oxygen; green, carbon; and blue, hydrogen.

A shaped laser pulse has an electric field with possibly complex time-dependent structure, and suitable manipulation of this structure has the potential to guide a molecule to a specified product. The shaped pulses are produced by spatial light modulation of both the phase and amplitude of the component frequencies in an ultrafast transform-limited pulse (12). It is possible to rapidly synthesize a large number of pulses, each of which will interact with the molecule in a different nontrivial manner, giving rise to a combinatorial photochemistry (13). We used time-of-flight (TOF) mass spectrometry to detect selective bond cleavage or chemical rearrangement after application of each shaped strong-field pulse. The concept is demonstrated here with acetone, trifluoroacetone, and acetophenone.

For strong-field manipulation of polyatomic molecules, the ultimate effect of a laser pulse cannot be calculated by any method at present. However, the molecule rapidly “solves” the Schrödinger equation governing its dynamics in real time when exposed to the laser field (1, 14). Thus, these experiments use the molecule under control as an “analog computer” to guide the shaping of subsequent laser pulses to better achieve the objective. Previous strong-field investigations with unshaped pulses applied to organic molecules have revealed a distribution of products ranging from intact ionization to substantial decomposition (15). The learning algorithm strives to tailor the optimal electric field to synthesize a specific chemical product in competition with the other possible outcomes (Fig. 1). Nonrobust fields are immediately discriminated against in the closed-loop approach (16).

The time-dependent electric field was produced by first dispersing the frequencies comprising the short pulse with a grating, thus transforming the pulse from the time domain into the frequency domain. The light was collimated with a cylindrical lens to produce a ribbon of light (the Fourier plane) 13 mm wide, where the frequencies may be addressed spatially (17). In our experiment, the radiation was discretized into 128 frequency bands, or components, by using a liquid crystal spatial light modulator (SLM). Each of the 128 elements in the SLM is 100 μm wide, with 3 μm of dead space between pixels. The relative phase and amplitude of each component can be altered independently with a pair of SLMs followed by a polarizer. After recombining the modulated frequency components with a second lens/grating pair, one obtains a time-dependent electric field that is unique to the phase and amplitude masks employed.

We can efficiently determine the optimal time-dependent electric field by using a genetic algorithm (18) to control the spatial light modulation. The algorithm creates and manages the 256 voltages (the genome) that determine the phase and amplitude of the component frequencies that control, in turn, the time-dependent structure of the laser pulses. A population of 40 distinct laser pulses (genomes) was propagated to determine the optimal pulse for a desired reaction. The laser pulses are randomly generated on the first cycle around the loop, with the restriction in our experiment that groups of 16 neighboring pixels are tied together to effectively make a genome with 16 variable sites. Each of these sites specifies a retardance that controls the phase and amplitude of the component frequencies. This process creates a search space of 36016 (∼1041) possible control fields given 1° resolution in 360° for the phase delay. Searching through this phase space may seem to be an insurmountable problem, but efficient methods have been developed and many distinct fields likely exist that give good quality control results (19). In subsequent generations, fields are selected for propagation using proportional selection, and new fields are generated by random crossover and mutation with a rate of 6% per field member.

The modulated laser pulses were focused to a 100-μm-diameter spot to interact with the molecules at 300 K and 10−6 torr. Spatial nonuniformity in the focal volume can produce localized laser intensity variations, but the learning algorithm will work to best overcome these differences (20). The intensity variations were further accounted for by using a 500-μm-diameter extraction aperture in the ion optics to select products from the most intense portion of the laser beam. The resulting TOF spectra were measured and signal-averaged eight times for each trial laser pulse shape, requiring ∼40 s to evolve a generation with a 10-Hz repetition rate laser. For these experiments, we found that acceptable convergence rates were obtained when the most fit member was cloned twice in the succeeding generation.

We investigated whether the overall algorithm could determine a suitable control field from both the large search space and nonlinear excitation scheme by beginning with a case where a viable solution was known in advance. A series of reference experiments (21) revealed that intense, transform-limited pulses (5 × 1013 W cm−2, 800 nm, 60 fs) produced a substantial quantity of the CH3CO+ ion from acetone. The evolution of the strong-field mass spectrum is displayed in Fig. 3A as a function of generation when the goal was optimization of the peak at a mass-to-charge ratio of 43 atomic mass units (amu) (CH3CO+). Although the CH3CO+ ion was almost unobservable in the first generation, the signal intensity was sufficient for the algorithm to begin optimization. The experiment produced exponential growth in the early generations, as seen in Fig. 3B. Analysis of the SLM mask revealed that the algorithm was indeed selecting a nearly transform-limited pulse to produce the CH3CO+. The yield increased approximately an order of magnitude by the fifth generation where the enhancement was saturated. In subsequent generations, the signal varied because of changes in the control field introduced by mutation and crossover. This experiment demonstrated that the closed-loop learning process with a genetic algorithm is an efficient method for controlling photodissociation and ionization yields when strong-field excitation is employed. Furthermore, this experiment showed that the mass spectrum of a molecule can be dramatically altered with tailored excitation pulse shapes. This capacity for multidimensional analysis could be of considerable value for determining the identity of complex molecules, because widely varying mass spectral ion distributions (each with different information content) can be obtained as a function of pulse shape. Dissociation need not be a distinct step before ionization occurs. In general, the strong-field pulse will prepare the optimal excited wave function to evolve into the desired products for detection.

Figure 3

(A) A representative mass spectra of acetone for the initial 0th, 3rd, 10th, and 22nd generations when maximization of the CH3CO+ ion from acetone is specified. (B) The CH3CO+ signal as a function of generation of the genetic algorithm. In (B) and the following plots of this type in Figs. 4 and 5, the average signal for the members of the population at each generation is shown.

It is convenient to define two classes of mechanism for the control of strong-field photochemistry: trivial control and nontrivial control. Trivial control refers to simple intensity or pulse duration effects leading to the observed photochemistry; pulse-shaping technology is not necessary when trivial control is satisfactory. Increasing intensity ultimately leads to increasing the dissociative ionization yield in the strong-field mass spectrum (22). Furthermore, increasing the laser pulse duration also alters the degree of dissociation observed in strong-field excitation. The intensity and pulse duration reference experiments for acetone (23) demonstrate that most of the control exerted arises from intensity effects. Cases exhibiting nontrivial control arise from the cooperative interaction of the laser pulse shape with an evolving molecular wave packet. For nontrivial control, the product distribution can be sensitive to the precise time-dependent structure in the pulse. In the strong-field regime, the dynamics could incorporate multiple electronic states in subtle ways to manipulate the dissociation or rearrangement to meet the posed objectives, as suggested by Fig. 2, C and D.

To determine whether the control process is trivial or nontrivial in any particular case, we compared the absolute and relative yields for two different ions as a function of generation in the genetic algorithm to those obtained from the simple intensity and pulse duration reference experiments. Such a closed-loop experiment is shown in Fig. 4 for trifluoroacetone (CH3COCF3), where two different objectives are presented. Maximization of the ratio CF3 +/CH3 + is shown inFig. 4A, and maximization of the ratio CH3 +/CF3 + is shown in Fig. 4B. The ion intensities are normalized to that obtained for the initial random pulses in generation zero, and the ratios are scaled to unity at generation zero. Initially, the absolute ion yields of CH3 + and CF3 + were within a factor of 2 in both experiments. In Fig. 4B, the optimal ratio is obtained on the 13th and again on the 21st generation, with the surrounding forays producing lesser yields. In the closed-loop experiment, the optimal genome was stored (for example, at the 13th generation) and was available for use at later times; thus, in this case, the largest reproducible ratio was ∼2.2. The trend during the maximization in each case is declining total ion signal. The reference experiments (21) for trifluoroacetone reveal no trend in the CF3 +/CH3 + ratio as a function of laser intensity, which eliminated the possibility of trivial intensity control. Considering trivial pulse duration control, the reference experiments reveal that the CF3 +/CH3 + ratio actually decreased with increasing pulse duration, which is opposite to the trend shown in Fig. 4A.

Figure 4

A summary of competing reaction pathways for trifluoroacetone is shown. The average of the ratios (pink) of (A) CF3 +/CH3 + and (B) CH3 +/CF3 + is plotted as a function of generation. Also shown are the relative ion intensities for CH3 + (blue) and CF3 + (green). Maximization of the ratio is the specified goal of the algorithm, and extrema at any point in the learning curve represent the best yield [e.g., a maximum enhancement of ∼2.2 is evident in (B)].

The degree of shaped-pulse control exerted in the case of trifluoroacetone could arise because of the widely different electronegativities available when fluorinated substituents are used. We therefore investigated acetophenone cleavage to determine whether control is possible in the case of hydrocarbon functional groups. The transform-limited, strong-field mass spectrum for acetophenone is shown in Fig. 5A, where several complementary pairs of ions in the transform-limited mass spectrum are observed. The ions at 15 and 105 amu correspond to the species observed after cleavage of the CH3 group (Scheme 1a). The second pair at 77 and 43 amu corresponds to cleavage of the phenyl group (Scheme 1b). Finally, there is an ion at 92 amu, corresponding to toluene. The presence of toluene reveals that a strong-field dissociative rearrangement (compare withFig. 1) of acetophenone is possible. The parent ion with a mass of 120 amu is not visible under these conditions but can be detected when lower intensity pulses are used. We probed whether we can control acetophenone's photochemistry by first optimizing the ratio C6H5CO+/C6H5 +(stipulating preferential cleavage of the methyl species rather than the phenyl species). The measured ratio is plotted as a function of generation in Fig. 5B. The enhancement in this case is a factor of 1.7. The phenylcarbonyl ion remains constant in intensity as the control algorithm progresses, whereas the phenyl ion is suppressed. The optimal pulse is not transform limited and is also complex (24). In these experiments, the C6H5 + and C6H5CO+ ions are in direct competition with each other when a given parent molecule is excited.

Figure 5

(A) A transform-limited strong-field TOF mass spectrum for acetophenone. (B) The relative ion yield for phenylcarbonyl (red) and phenyl (green) and the C6H5CO+/C6H5 +ratio (pink) as a function of generation when maximization of this ratio is the specified goal in the closed-loop experiment. (C) The relative ion yield for phenylcarbonyl and phenyl and the C6H5 +/C6H5CO+ratio as a function of generation when maximization of this ratio is specified. (D) The average signal for toluene, 92 amu, as a function of generation when maximization of the ion signal for this reaction product was specified for optimization. Corresponding electron-impact ionization mass spectrometry revealed no evidence for toluene in the sample.

Scheme 1

The second optimization goal imposed on the system was maximization of the opposite ratio, C6H5 +/C6H5CO+, corresponding to preferential cleavage of the phenyl group. The learning curve for this process, Fig. 5C, again shows that the intensity of the phenylcarbonyl ion remains relatively constant, whereas the intensity of the phenyl ion increases. This is interesting because the energy required to cleave the phenyl–CO(CH3) bond (100 kcal/mol) is ∼15 kcal/mol higher than that for the methyl–CO(C6H5) bond (25). In both experiments, the algorithm enhanced the ratio by either increasing or decreasing the intensity of the phenyl ion while maintaining constant phenylcarbonyl intensity. We attribute the observed control to a nontrivial mechanism for several reasons. In the reference experiments (21), altering the intensity changed the C6H5 +/C6H5CO+ratio with a dynamic range of 1.5, but this process required a change in C6H5CO+ ion intensity of one order of magnitude. The tailored pulses in Fig. 5, B and C, exhibit a dynamic range of at least 4 in this ratio with no change in C6H5CO+ ion intensity. Given the limited amount of amplitude and phase space explored by the experiment (e.g., tying 16 pixels together and the 6% mutation rate), it is likely that the optimal dynamic range from pulse tailoring is much larger. It is extremely unlikely that there are trivial intensity-pulse duration conditions that maintain constant phenylcarbonyl intensity while increasing or decreasing the phenyl intensity. In the reference experiments, we always found that the absolute ion intensity decreased monotonically with increasing pulse duration or decreasing pulse laser energy. Such correlation is not evident in the pulse-shaping experiments.

The goal of laser control of chemical reactivity transcends the simple unimolecular dissociation reactions observed to date (7, 8). Observation of the toluene ion in the strong-field acetophenone mass spectrum suggests that control of molecular dissociative rearrangement may be possible. To test this hypothesis, we specified the goal of maximizing the toluene yield from acetophenone, as shown in Scheme 2. For toluene to be produced from acetophenone, the loss of CO from the parent molecule must be accompanied by the formation of a bond between the phenyl and methyl substituents. Conventional electron-impact mass spectrometric analysis of acetophenone revealed no evidence for toluene in the cracking pattern. The closed-loop control produced an increase in the ion yield at 92 amu of a factor of 4 as a function of generation, as shown in Fig. 5D. As a further test, we specified maximization of the ratio of toluene to phenyl ion and observed a similar learning curve to that in Fig. 5, B and C, with an enhancement in the toluene to phenyl ion ratio of a factor of 3. Again, the final tailored pulse does not resemble the transform-limited pulse (24). In order to confirm the identity of the toluene product, measurements on the deuterated acetophenone molecule C6H5COCD3 were carried out, and the C6H5CD3 + ion was the observed product in an experiment analogous to that shown in Fig. 5D. The observation of optically driven dissociative rearrangement represents a new capability for strong-field chemistry. In such strong-field excitation, the electron dynamics during the pulse is known to be extreme, and substantial disturbance of the molecular eigenstates (compare with Fig. 2) can produce photochemical products, such as novel organic radicals, that are not evident in the weak-field excitation regime. Operating in the strong-field domain opens up the possibility of selectively attaining many new classes of photochemical reaction products.

Scheme 2

Extensive manipulation of mass spectra is possible when shaped, strong-field laser pulses interact with molecules under closed-loop control. The control pulses occur with an intensity of ∼1013 W cm−2, where the radiation substantially disturbs the field-free eigenstates of the molecule. Even in this highly nonlinear regime, the learning algorithm can identify pulse shapes that selectively cleave and rearrange organic functionality in polyatomic molecules. These collective results suggest that closed-loop strong-field laser control may have broad applicability in manipulating molecular reactivity. The relative ease in proceeding from one parent molecule to another should facilitate the rapid exploration of this capability.

REFERENCES AND NOTES

View Abstract

Stay Connected to Science

Navigate This Article