Molecular Memory Operations

One component of this molecular memory is a photochromic N-(4-aminophenyl)fulgimide, and the other is a highly fluorescing oxazine dye. The absorption spectra are shown in Fig. 2. The fulgimide has two photochromic forms: the closed, polar form that is transformed by excitation with 530-nm light (14–17) and the open, nonpolar form that absorbs at ≈400 nm. The dye component has its maximum absorption at ≈650 nm and fluoresces intensely at 700 nm only when the fulgimide component is in its nonpolar form. The time-resolved spectra and kinetics of both components are presented, and the reaction mechanism of this photoconversion is described also.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Spectra of the components. Solid line, oxazine dye; dashed line, fulgimide, open form; dotted line, fulgimide, closed form.

The operation of this molecular memory is as follows. When the fulgimide component is in the polar, closed form, write form, it is assigned the 0 computer code. After excitation with a 530-nm pulse it is transformed to the nonpolar, read form that corresponds to 1 in the computer code and absorbs at 400 nm. This transformation is the information-storage process. Reading, or accessing, the information is achieved by exciting the dye at 650 nm, which induces the written molecules to emit fluorescence at ≈700 nm. Note that only the written molecules will fluoresce strongly because they are the only ones that provide the dye with a nonpolar environment. The fluorescing molecules, bits, are imaged on the surface of a charge-coupled device, which is coupled to the processor. The reading process is a true, nondestructible read-out process, because the 650-nm photon energy used for read-out is much lower than the 530-nm minimum energy required to write and more so than the 400 nm needed to erase the stored information. Therefore, while reading neither writing nor erasing is possible. To erase the stored information, the written bits are illuminated with 400 nm, which initiates the photoreaction that converts the open form of the fulgimide moiety to the “write,” closed, polar structure. We have studied this interconversion between the two photochromic forms and found it be very efficient and to occur with a rate of ≈4 × 10¹¹ s^-1. A single molecule may indeed perform the entire set of write, read, and erase operations, and the stored data may be accessed by single-photon counting techniques. The fulgimide and dye components alone can perform >10⁴ cycles with very small loss of information. We have ascertained that the write and read forms of the fulgimide and the dye are stable between -55 and +55°C, and they are also not damaged by the picosecond laser pulses and He/Ne used to store, access, and erase the information.

A very important characteristic of any storage device is the rate by which the stored information may be accessed. In this case it depends on the rate of the photochromic transformation from the closed, write to the open, read forms and the dye fluorescence lifetime, which are analogous to the rates of writing and accessing the information, respectively. The data presented show that both are extremely fast with lifetimes in the picosecond and nanosecond range, respectively. These rates may be contrasted with the diffusion-controlled or triplet-singlet emission rates of other proposed materials.

Spectroscopy and Kinetics

The energy-level diagram for the photoreaction responsible for the conversion of the nonpolar, open form to the polar, closed form is shown in Fig. 3. The rate of population growth of the v = 0 of the first excited singlet state, corresponding to vibrational relaxation after excitation with a 100-fs, 400-nm pulse, was found to be pulse-limited or ≈200 fs. It is rather difficult to ascertain the exact lifetime, because although the full-width half-maximum of the pulse is 100 fs, most of the photons that populate the excited state are spread at two and four times the 100-fs width of the Gaussian pulse distribution, making extrapolation to lifetimes less than the 100-fs width rather difficult to assess.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Photoreaction energy-level diagram of the composite molecule.

The rates of formation and decay of all excited states were measured by means of an ultrafast time-resolved laser system similar to those described by us previously (18). The absorption spectra of the fulgimide and dye moieties are shown in Fig. 2. The time-resolved spectra for the formation of the polar form and the decay of the nonpolar form, manifested in the photochromic transformation of the fulgimide molecule, are shown in Fig. 4. A shift in the band maxima observed during the first 3 ps after excitation is due to index dispersion of the broad continuum used for monitoring the changes in the absorption spectra. We performed experiments with various laser-pulse bandwidths, intensities, and time widths, which confirm that the observed shift can be accounted for by dispersion. This assignment agrees also with the study on furyl fulgides (19). The appearance of a shoulder band at ≈580 nm at 3.5 ps after excitation that smooths out after 10 ps suggests the possibility of an intermediate excited state designated by I in the energy-level diagram (Fig. 3). At the present time we are not certain of the cause of this shoulder and the possible existence and assignment of this intermediate state. The kinetics of the fulgimide plotted in the form of optical density vs. time at 520 nm is shown in Fig. 5. This band grows with a rate of 5 × 10¹² s^-1 followed by a decay with a rate of ≈3 × 10¹¹ s^-1. The optical density does not decay to 0 but reaches a plateau at 0.035 ΔA and then remains constant. This plateau is due to the fact that even though the nonpolar form does not absorb at 520 nm, the polar form has an appreciable S₀ → S₁ absorption at this wavelength. We used the fulgimide extinction coefficient value of 1.5 × 10⁴ and the excited-state transient absorption data to calculate the S₁ → S_n extinction coefficient value at 520 nm. This value was found to be three times less than that of the ground-state transition, namely ≈4.5 × 10³ s^-1.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

(A) Fulgimide, open to closed form transient spectra. (B) Transient spectra of oxazine. (C) Transient spectra of open to closed form of composite molecule.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Plot of ΔA vs. time. Solid line, composite molecule; dotted line, oxazine; dashed line, fulgimide.

The dye component absorption and fluorescence spectra are shown in Fig. 2, where the dominant bands are due to strong absorption at 650 nm and intense fluorescence at 700 nm. The excited-state dye-relaxation mechanism has also been studied. The 650-nm upper singlet state relaxes to the ground state by means of either a radiationless transition with a rate of 5 × 10¹⁰ s^-1 or fluorescence that has a quantum yield of 0.005 and a fluorescence decay rate of ≈5 × 10⁸ s^-1 when dissolved in propanol. The quantum yield of fluorescence is found to increase by ≈2 orders of magnitude in solid polymethylmethacrylate (PMMA) matrices. Time-resolved spectra of the oxazine dye are displayed in Fig. 4, and the growth of the excited state and the decay of the dye are summarized in Fig. 5, in which the change in optical density, ΔA, at 540 nm is plotted against time. The ΔA decays to 0 in this case, because the oxzine dye, when alone, returns to its normal ground-state absorption at 540 nm. It was known previously and we have made in-depth studies to confirm that the fluorescence intensity of this dye is very much higher in nonpolar than in polar environments. Table 1 lists several solvents and solid PMMA, their dielectric constants, and the corresponding fluorescence quantum yields of the composite molecule. It was found that the fluorescence quantum yield varied by ≈2 orders of magnitude from methanol to solid PMMA where it was found to be 0.16. These data suggest that in addition to polarity, viscosity and the hardness of the matrix also play a dominant role in the fluorescence quantum yield.

We performed the same kinetic and spectroscopic studies using our fulgimide/oxazine composite nondestructible read-out molecular memory molecule, the absorption spectra of which are shown in Fig. 1. Comparison with the absorption spectra of the fulgimide and oxazine dye alone (Fig. 2) shows that the spectra of the composite molecule are essentially a superposition of the spectra of its two components. The spectrum is identical to that of each component in the areas where only one component absorbs; however, the spectrum is different from each of the two components in the 200- to 400-nm region, because both the dye and fulgimide components absorb in this region. Analysis of the spectrum of the composite molecule in the 200- to ≈400-nm region shows that it is a superposition of the spectra of the two components. The fluorescence spectra emitted by the composite molecule are also the same as that of oxazine alone with the very notable exception that intense fluorescence is observed only when the fulgimide component is in its open, nonpolar structure. When the fulgimide component is converted to its closed, polar structure, the fluorescence of the oxazine moiety decreases by a factor of 4 and as much as a factor of 10 in molecules with long chains attached to the dye group. The absorption wavelength of the dye moiety and the strong dependence of fluorescence on the structure of the fulgimide group are shown to have a pivotal effect on the ability of this molecule to operate as an erasable, nondestructible read-out molecular memory.

Because of the relatively high absorption intensity of the dye moiety in the 600-nm region, the absorption cross section of the dye and fulgimide at 540 nm are approximately equal (see Fig. 2). A histogram of the photoinduced events, occurring in the composite molecule, is depicted in Fig. 4C in the form of transient spectra vs. time after excitation with a 400-nm, 100-fs pulse. We also measured the depopulation and growth of the ground-state kinetics and the formation of a new transient absorption band that appears immediately after excitation. The bleaching in the 650-nm region is similar to that observed for the dye group alone, whereas the absorption band in the 500- to 600-nm region is due mostly to the fulgimide conversion from the open to the closed form. Comparison of these transient spectra with that of the fulgimide and oxazine alone, Fig. 4 A and B, respectively, make it evident that the kinetic behavior of the composite molecule is practically the same as that of the fulgimide and oxazine components alone. The growth, decay, and intermediate states of the photochromic forms of the fulgimide and the dye are also present in the transient spectra of the composite molecule. Analysis of the transient spectra (Fig. 4C) reveals the growth of a new band at 540 nm within the pulse time width, which decays within 200 fs. It is assigned as it was previously to excited-state vibrational relaxation in S₁. The population decay from the first electronic excited state relaxes with a biexponential decay, with rate values of 3 × 10¹¹ and 7 × 10¹⁰ s^-1. These rates agree quite well with the decay rates that we measured for the fulgimide and oxazine, respectively. These data are also shown in Fig. 5 in the form of ΔA vs. t at 540 nm.

To show the potential of this molecular memory as a medium for write, nondestructive read-out, and erase devices, we dissolved the composite molecule in PMMA and fabricated a 2-cm-diameter, 0.5-cm-thick optically polished disk. Using 530nm, second harmonic pulses from an Nd/YAG picosecond laser, we stored the image of the letter “A” on the surface, which can also be easily stored in the bulk of the disk, by means of two-photon processes. Accessing the stored information was achieved by illuminating the disk with a He/Ne 632-nm laser that induced fluorescence from the written molecules, bits, that were imaged onto a charge-coupled device coupled to a processor (see Fig. 6A). This image was read many times without any detected loss in intensity. Fig. 6B shows the same written area, bits, of the disk after the image was erased by illumination with 400-nm light. No detectable fluorescence is observed after being illuminated with the same 632-nm He/Ne laser, because the fulgimide component has been converted to the polar form. Fig. 6C shows an image rewritten on the erased area after illumination with a 530-nm laser pulse that converted the fulgimide to the nonpolar form. Fig. 6 provides strong evidence that this composite molecule that we designed, synthesized, and studied indeed operates well as a write/erase molecular memory with nondestructive read-out. This is accomplished because the write, read, and erase forms have their own distinct, widely separated absorption bands, and the fluorescence intensity strongly depends on the polarity of the fulgimide component. Uses for this molecule may also be found in microfluidics as a smart biological indicator because of its fluorescence intensity changes as a function of polarity, pH, and viscosity and as an optical communications broad-band optical switch because of its changes in the refractive index between the two forms. We have also designed other molecular systems with more advanced properties for optical storage and other applications.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

Images, written in disk, of the composite molecule dispersed in PMMA matrix. (A) Written image. (B) Erased. (C) Rewritten.