CA2237182A1 - Two-step gated holographic recording in photorefractive materials using cw lasers - Google Patents

Abstract

Ferroelectric materials are disclosed as reversible holographic recording media (25) for use in two-photon recording systems. The ferroelectric materials disclosed herein provide long-lived electronic states intermediate the ferroelectric material's valence and conduction bands. These intermediate states have a sufficiently long life (on the order of 1 to 100 milliseconds) that low-power continuous wave ("cw") laser (1) can be used to record interference patterns on them. Thus, two-photon holographic recording systems are also disclosed which do not require high-power, short pulse length, mode-locked or Q-switched lasers. Rather, the disclosed holographic recording systems employ visible and near IR cw lasers such as diode lasers. The disclosed two-photon holographic recording systems provide for absorption of a first photon which excites electrons of holographic recoding media to an intermediate state. Thereafter, upon absorption of a second photon, the electrons are promoted to the media's conduction band where they are arranged according to the interference pattern provided by the recording system.

Description

CA 02237182 1998-0~-08 W O 97120317 PCT~US96/188~2 TWO-STFP GATFn HOLOGRAPHIC RECORDING IN PHOTOREFR~CTIVE
MATERIALS USrNG CW LASE3RS
This invention was made with government support under agreement number MDA972-g4-2-0008 (ARPA Order No. A576) awarded by the Advanced Research ProjectsAgency. The government has certain rights in this invention.
Description Terhniç~l Field This invention relates to holographic recording media. More particularly, the invention relates to holographic recording media including poled single crystal ferroelectric m~tPri~l~
lQ
13ack~round Art Holographic recording systems hold forth the promise of very high data storage c~en.citif~s and parallel recording and reading capabilities. It was recognized by the early 1960s that holographic recording media could, in theory, provide a much greater data storage density 15 than m~gnetic lecc,lding media and other conventional recording media. Such high densities are attributable, in part, to the holographic storage media's ability to store information in three dimensions. Thus, the theoretical storage capacity for a volume hologram varies appr ~;m~tPly as media volume divided by the cube of the recording wavelength (V/~3). So, for a wavelength of 500 nanometers, the storage capacity for a volume hologram is on the 20 order of 1012 to 1013 bits/cm3, compared to 108 bits/cm2 for two-dimensional optical storage media, such as CD-ROMs.
In holographic recording systerns, a source of monochlulllillic coherent radiation is split into (1) an "object beam" which is spatially modulated by a two-dimensional grid (or data source) cont~inin,~ the information to be recorded, and (2) a "reference beam" which is 25 unmor~ These two beams are then directed onto a single region of a holographic recording mP(lillm, where they interact to generate an hll~lr~;;r~l-ce pattern. The holographic recording medium is made from a light sensitive material that records the resulting interference pattern, and thereby stores the information provided on the grid. An image of the recorded information can be constructed from the recorded interference pattern by illl.,ni.,~f;,.g 30 the mP~ m with a "read beam." The read beam will have the same ch~.;Lt;li~Lics as the reference beam used to record the hologram.
Ferroelectric materials have been investig~tPd as potential holographic recording media because they can m~int~in electric dipole flom~in~c~ even in the absence of an applied electric CA 02237182 1998-0~-08 WO 97nO317 PCT~US96tl8802 field. In such m~t~.ri~l~, in~ ent radiation from an optical i~ rer~nce pattern ~lu~l~otes electrons from illllmin~te~l areas to a conduction band where they diffuse away from the illumin~t~.tl areas. Some of these mobile electrons fall from the conduction band back into stationary traps. The electrons in the illllmin~t~ areas continue to be promoted to the 5 conduction band at a much greater rate than the electrons in the dark areas. Thus, during illumination, the concentration of electrons gradually increases in the dark areas and gradually decreases in the light areas. When the radiation inlt,rer~llce pattern is removed, the electrons are no longer promoted to the conduction band by radiation, but do remain trapped at the same spatial locations that they ~cs~lm~d during illl-.lli..,.~ion. Thus, the ferroelectric m~tf~.ri~l can 10 record an optical interference pattern in the form of a spatial distribution of de~,l,uns. In some ferroelectric materials (SOm~tim~os referred to as "photorefractive" m~teri~l~), this spatial distribution of electrons causes a corresponding spatial distribution in ,Gr,dl;Live index which can be read by directing a read beam onto the recorded ferroelectric material as mentioned above.
In some materials, the charge carriers IC;S~ ible for the photorefractive effectdescribed above are mobile "holes" excited to the valence band by the laser light. For simplicity, the relll~;ll;llg (1i~cus~ion is limited to electrons. However, it should be understood that the phenomena described herein can be extended to holes in straight forward manner.
Ferroelectrics are particularly attractive holographic storage media because they can be 20 reversibly recorded; i.e., they can be erased and Ic,~,corded many times. Many co.l.peling holographic r~cc,ldillg media store information only irreversibly in the manner of a CD-ROM.
Unfortunately, ferroelectric materials often can be too easily erased -- particularly during the process of reading the recorded hllelrl;rellce pattern. As mentioned, to read the recorded medium, radiation must be directed onto it. And the reading radiation must be of the 25 same wavelength as the radiation used to record the data image. Thus, the photon energy of the read beam will cause some of the electrons in the ferroelectric recording medium to reenter the conduction band (just as they did during recording) where they diffuse to a uniform distribution, thereby erasing the interference pattern.
In an effort to overcome this problem, it has been proposed to use a "two-photon"
30 recording pl~ce-lu.~, so named because it requires two photons to excite an electron to the ferroelectric's conduction band. It has been shown that two-photon recording can be accompli~h~-l by ill~ g the ferroelectric with a "gating" beam at one wavelength and a read/write beam at a second wavelength. See, for example, D. von der Linde et al., "Multiphoton photorefractive processes for optical storage in LiNb03" Appl. Phys. Lett. 25, 155 (1974). The two wavelengths are chosen such that photons at both wavelengths have insufficient energy, by themselves, to promote electrons to the conduction band. However, the sum of the photon energies for the two wavelengths is sufficient to promote electrons to the CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802conduction band. In application, a first photon excites an ele~ 11 to an "illLt. . . ~flis~t~"
electronic state close to the conduction band. Then while the electron is l~ o,a,ily residing in such int~,rn~ t~. energy state, a second photon (typically of a diîr~.tn~ wavelength than the first wavelength) promotes it to the conduction band where it diffuses and becomes trapped to S record the hll~lrt;lG~Ice pattern as described above.
This two-photon process provides a more stable holograrn than the previously-described "single-photon" process in which only a single photon is required to promote an electron into the conduction band. In single-photon systems, the photons provided by the read beam promote the electrons making up the recorded interference pattern to the conduction 10 band where they redistribute themselves and thereby erase the stored information. In the two-photon recording systems, however, the read beam will have an intensity and photon energy chosen to make single-photon promotion impossible and two-photon promotion unlikely.
Thus, a holographic recording m~-lium recorded by a two-photon process can be read many times without erasure.
It was realized early on that the two-photon process' reliance on an intermediate energy state below the conduction band posed its own problem. Without special IIG<~ IP~I of the ferroelectric recording m~-lium, such states were ~LlG-Ilely short lived. As such, the light sources required to promote electrons to the conduction band would have be of extremely high power (at least on the order of a gigawatt/cm2) and therefore of extremely short pulse duration 20 (on the order of 10 picoseconds). Such constraints would be unworkable for commercial systems. To .onh~n~e the two-photon transition probability and thereby relax these constraints, subsequent work employed ferroelectric materials doped with transition metal ions (e.g., iron, ~;hl~lluulll, and copper ions). Such ions provided relatively long-lived int~,rmf,~ te electronic states (typically on the order of 100 nanoseconds) close to the ferroelectric's conduction band.
While two-photon It;co~ g procedures relying upon such doped ferroelectrics havebeen recognized as an advance in the march toward a c~ leLcial holographic It;co~ g system, they still require a light source having an ~ln~ cept~bly high power and short pulse length. In U.S. Patent No. 3,922,061 issued to Glass et al. -- which tl~ r,ribes some of the early work on two-photon recording -- it is stated that such light source should be a mode-30 locked or Q-switched laser having "a ;~.i--peak intensity of 1 megawatt/cm2". The patent further suggests that for some two-photon systems, lasers of 10 to 100 gigawatt/cm2 may be required (see column 9, lines 1-12). Unfortunately, such power requirements coupled with mode-locking or Q-~wiL~ lg are incompatible with a low cost, reliable storage system.
Very recently, it has been proposed to use ferroelectric materials doped with rare earth metal ions having 4f excited states that give rise to absorptions in the near infra-red and visible spectral regions (e.g., praseodymium, neodyllliulll, and thulium ions). Such ions provide nifi~ ntly longer-lived intermt~ f~ electronic states (typically on the order of 0.1 to 1 CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802 millicecond) than the transition metal ions used in previous systems. This allowed t~,vo-photon holographic recording to be accomplished with inexpensive low power continuos wave ("cw") lasers such as diode lasers. Thus, commercial reversible holographic recording systems may now be within reach. Holographic recording systems employing such rare earth S doped ferroelectric materials are ~les~riheA in US Patent application serial no. 08/S38,704 (Attorney docket no. SRIlP009/P3554), filed on October 3, 1995, narning Bai et al. as inventors, and entitled "GATED RECORDING OF HOLOGRAMS USING RARE-EAl~TH DOPED FERROELECTRIC MATERIALS." That application is incol~ol~t~d herein by reference in its entirety and for all ~ul~oses.
While the rare earth doped ferroelectrics may become h~ holographic lGcordi"g media, other recording media might also be useful, particularly media that could efficiently record at a wide range of laser wavelengths. It should be recognized that to attain optimal .t;cordi"g efficiency in a doped ferroelectric recording mt tiium, one of the two photon sources should be provided at a wavelength in resonance with an electronic transition in the dopant 15 atom. This is not always practical, as inexpensive laser sources generally produce radiation only at discreet wavelengths that are not n~cçs~s~rily in resonance with available dopant ion transitions.
Thus, it would be desirable to provide a reversible holographic recording system that can efficiently record at wavelengths provided with various types of continuous wave lasers.

~icrlosure of the Invention To meet this need, and take advantage of an observed ~.~GILy of certain ferroelectrics, the present invention provides a holographic recording system that employs a continuous wave laser for reading from and writing to a ferroelectric recording medium. Unlike the 2~ doped ferroelectric media described above, the ferroelectric m~ linm of this invention co~ ls, at most, a very low concentration of optically active hl-~u~ilies (i.e., illl~u-ilies such as transition metals or rare earth metals which have electronic transitions in resonance with the wavelength of the continuous wave }aser). Preferably, the ferroelectric material contains defects or other features that provide long-lived intermP~ tP states on the order of 1 to 100 30 milli.~econds. It has been observed that such intermr~ tP states can be used in two-photon recording processes employing a continuous wave laser producing radiation of an intensity less than about 1000 W/cm2. Further, it has been observed that, in many cases, such intermP.tli~tP states allow recording with radiation over a wide range of wavelength~ in the visible and infrared electr~m~gn~tic spectral regions. Thus, the present invention can be 35 practiced without l~ a laser that produces radiation of a wavelength in resonance with a narrow absorption peak of a dopant atom.

CA 02237l82 l998-0~-08 In the systems of this invention, like other two-photon recording systems, absorption of a first photon excites electrons of the recording mPAium to an in~PrmPAi~tP state. Thereafter, upon absorption of a second photon, the electrons are promoted to the meAillm's conduction band where they diffuse before becoming trapped in an arrangement corresponding to the S hlLG relGIlce pattern provided by the recording system. Unlike most prior two-photon systems, the first and second photons can be provided by low-power commercially available cw lasers such as the type used in colllulGlcially available optical systems such as CD readers.
In one aspect, the present invention provides a holographic recording system that can be chala~;lGl;~ed as including the following elemPnt~ (1) a holographic recording mP~itlm 10 including a ferroelectric material that contains no more than about 0.01% atomic of an optically active hll~uliLy; (2) a first radiation source for providing coherent monochromatic radiation which is first divided into a reference beam and a spatially modulated object beam and then recombined to form an inter~erence pattern on a first region of the holographic recording mPriitlm; and (3) a second radiation source for providing a gating beam which is optically 15 coupled to a second region of the holographic recording mPAitlm. The first and second regions of the lGcoldi-lg mPf~ m should be at least partially coextensive with one another.
Further, the first and second radiation sources should be chosen so that their respective photons together promote electrons of the holographic recording media to a conduction band by a two-photon process. In this manner, the system records the illL~lr~,~Gnce pattern in the 20 holographic recording m--Aillm In general, one of the radiation sources should be chosen to produce radiation of a frequency that promotes electrons to the recording meflit]m's tP. levels.
Many variations on this basic theme may be provided. For example, in some cases radiation from the first and second radiation sources will have the same wavelength. In such 25 cases, the two-photons nlscec~ to promote an electron to the conduction band may be provided by a single radiation source -- i.e., the first and second radiation sources are the sarne.
This is known as a "one-color two-photon" system. More commonly, two dirr~ ,nt radiation sources ("two colors") will be employed: one of which provides the object and reference beams, and the other of which provides photons used exclusively to promote electrons to the 30 conduction band (the gating beam). In plGrtll~,d embodiments, the gating beam will have a shorter wavelength than the reference/object beam. This makes erasure more difficult during subse~uent read operations -- which employ relatively low energy photons at the reference beam wavelength.
Preferably, the first radiation source is a continuous wave laser (as opposed to a more 35 expensive Q-switched or mode-locked laser) having a power density of less than about 1000 W/cm2 and more preferably in range of 20 to 200 W/cm2. Further, the first radiation source preferably produces radiation in the visible or near infrared regions of the eleclru~ gnP.tic spectrum. In especially preferred embodiments, the first radiation source is a diode laser. The CA 02237182 1998-0~-08 W O 97/20317 PCT~US96118802second radiation source may be either a laser or an incoherent radiation source. In the latter case, it may require a filter to block high energy photons that would promote electrons to the conduction band by a single photon.
In preferred embo-lim~nt~, the ferroelectric l~;cordillg media contains no more than 5 about 100 parts per million (on a per mole basis), and more plcf~,.dbly no more than about 10 part per million, of any optically active ç~trin~ic hll~uliLy. Suitable ferroelectric materials for use with the present invention include lithium niobate, pot~c~ m lithium niobate, lithium t~nt~ tt~, barium tit~n~te, ~Ll~llliulll barium niobate (SBN), lead barium niobate (PBN), and barium strontium pol~iulll sodium niobate (BSKNN). Preferably, the ferroelectric has been 10 processed to introduce defects which produce long-lived intermediate states. Such prcce~ing may include subjecting the ferroelectric to a reduction process to introduce oxygen v~n-ies.
~ltt-.rn~tively, defects can be introduced by doping with an optically inactive dopant.
In another aspect, the present invention provides a method of writing to a holographic recording medium (a ferroelectric material co--l;~;--i--g no more than about 0.01% atomic of an 15 optically active hlll!uli~y). ~uch method can be characterized as including the following steps:
(I ) sepa~atillg a first radiation beam of a first wavelength into a l~;fe~ ce beam and a spatially modulated object beam; (2) colllbhlillg the leçclcllce beam and the spatially mo-l--l~t~(l object beam to form an interference pattern on a first region of the holographic recording m~-linm;
and (3) directing a gating radiation beam CO~ illg at least a second wavelength onto a second 20 region at least partially coextensive with the first region of the holographic recording mPriil-m Photons of the first and second wavelengths together promote electrons of the holographic recording media to a conduction band by a two-photon process such that the hllelr~ ce pattern is recorded in the holographic recording medium. In general, the character of the recording media and the gating, reference, and object beams used in this method are as 25 described above in the context of the system aspect of this invention.
The methods of this invention allow for angular and frequency, and phase multiplexing. In angular multiplexing, the step of combining the reference beam and the spatially mo~ ted object beam to form an hlLelr~"ellce pattern is conrlllctPcl at a defined first angle with respect to the recording medium. Thereafter, the reference beam and a second 30 spatially modulated object beam are combined at a second defined angle, different from the first defined angle, to form a second int~,.r~ ce pattern which is recorded on the holographic recording m~ m In this manner two or more "slices" of two-dimensional data are recorded at slightly different angles on the same holographic recording m~ m In frequency multiplexing, the method includes steps of (I) St;~dtillg a third radiation 35 beam of a third wavelength, different from the first wavelength, into a second reference beam and a second spatially modulated object beam; and (2) c~lllbhlillg the second reference beam and the second spatially modulated object beam on the ho}ographic recording medium. This is done in a manner that produces an iulGlrGlGIlce pattern of radiation from the third radiation bearn that is recorded in the holographic r cor~ g mPAi~lm together with the inlelr~l.,.lce pattern of the first radiation beam.
In phase multiplexing, the reference beam is spatially phase mofl~ t~.A as compared to - S the spatially uniform reference bean used in the above two methods. In general, a set of orthogonal phase codes can be constructed so that the total number of the objects recorded in a m~flillm can be equal to the number of phase codes. A detailed discussion can be found in U.S. Patent 3,612,641 by C.C. Eagler~leld which is incorporated herein by rGrGIGllce for all purposes.
~hese and other features and advantages of the present invention will be presented in more detail in the following detailed (l~osrription of the invention and the associated figures.
Rrief Description of the Draw;np~
Figure 1 is a schematic representation of holographic storage apparatus to read and write information on a holographic storage m~rlinm in accordance with this invention.
Figure 2 is a schematic lGylGs~ ~ion of an object, reference, and gating beam incident on a holographic storage element during a writing process in accordance with the invention, and a scattered beam which would result during a reading process.
Figure 3a is a block representation of the experimental setup used for one-color/two-photon experiments discussed below.
Figure 3b is a block representation of the e,,yt;lhllGlllal setup used for two-color/two-photon e~yelillRllL~ discussed below.
Figure 4 is a graph co. . .~ g the diffraction efficiency of a hologram versus readJwrite beam wavelength for holograms recorded in one-color and two-color e~yc~ GI~ts.
Figure S is a graph showing how the diffraction efficiency of a two-photon, two-color holographic recording decreases with increasing delay between a writing pulse and a gating pulse.

Best Modes for Carrying out the Invention l . Holographic Recording Figure 1 prGse~ a preferred holographic data recording a~ dlus of this invention, such as may be used with a high speed CO1~ ULG1. A first coherent light source 1 puts out a beam of monocl"~,matic light 2 of a first wavelength (~1)- The bearn of monochromatic light CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/188022 is made incident on a beam splitter 3 which reflects approximately half of themonochromatic light to form an object beam (Eo) 7 and passes the rem~in~Pr of the monochromatic light to form a reference beam (ER) 5. The object beam 7 may be directed through various optical elements, such as, for example, a beam expanding element 13, and 5 then towards a representation of data to be stored 15, shown in this illustration as a two-t1imPncional sheet of alternating spaces of light and dark meant to represent a page of binary information. The object beam 7, which is scattered from the data r~lGsel"~lion 15, is collected and recollim~tPA by a second set of optical elements, such as, for example, a second lens system, represented here by a second single lens 17, and a deflector element 19. The 10 scattered object beam 7 now contains the information in data representation 15 in the forrn of amplitude and phase variations with respect to the reference beam 5.
Although the data object is shown as a two-dimensional sheet 15 of binary information in Figure 1, holographic data storage systems are not, in general, so limited. For example, the information to be stored may be analog as well as digital: although digital information will be 15 a~ opliate for most current co~ uLillg applications. In addition, the object to be recorded may take many forms -- each of which spatially modulates incident radiation to produce the object beam. Such spatial modulators are conventional in the art and incl~(le, for example, liquid crystal modulators, electro-optic modulators, m~gnPto-optic modulators, and acousto-optic mo~ t~rc Such modulators are discussed in "The Physics of Optical Recording" by 20 K. Schwartz, Springer-Verlag, 1993 which is inc("~o~aled herein by reference for all purposes.
Both the object beam 7 and the reference beam 5 are made incident on a holographic storage ...~ . 25 and interact so as to create an interference pattern in the storage mP~ lm The holographic storage m~ lm 25 comprises a ferroelectric material co~ ing at most low 25 concentrations of optically active i",~u,ilies as described in more detail below. A second (unmo~ erl) light beam, known as a gating beam (EG) 23, of a second wavelength (~2) not equal to ~1 and preferably less than ~1, from a second light source 21 is directed simllltslneously onto the storage mP-lillm 25 with the object beam 7 and reference beam 5. A
photon of light with wavelength ~2 provides enough energy to excite an electron in the storage 30 mPAinm from a low Iying defect state to an illle;llllediate state closer to the medium's conduction band, but not enough energy for an electron to directly enter the conduction band.
Then a photon of light with wavelength ~1 provides the electron in the intermP.~ tP state with enough energy to enter the ferroelectric m~t~ri~l's conduction band, where it can move from the regions of high light intensity and become trapped in the regions of low intensity. After 35 the illl~. . .i. .i.l ion from the first and/or second light beams is removed, and electrons are no longer optically promoted to the conduction band, and thus remain spatially trapped in of the ferroelP~ctric material, thereby creating a local electric field within the ferroelectric m~tPri~l which leads to an in~lncP l refractive index change similar to the spatial variations of intensity CA 02237182 1998-0~-08 WO 97nO317 PCT~US96/18802 produced by the interference pattern. As noted, this phenomenon is som~timPs referred to as the photorefractive effect. It should be understood that holographic recording may also be effected by an analogoug mech:~nicm in which holes (positive mobile charge carriers) are excited to a ferroelectric's valence band where such holes move away from the incident 5 r~ tion In some preferred embodiments, in contrast to the above-described embodiment, photons from the reference and object beams 5 and 7 will excite electrons tO the inte~ne~i~tp states, and the gating beam 23 photons will then promote the excited electrons to the conduction band. Either way, ~2 and ~ should be chosen in conjunction with the ferroelectric 10 material so that neither ~2 and ~I can, by itself, promote an electron to the ferroelectric's conduction band, but the sum of the photon energies at the two wavelengths is sufficient to promote an electron to the conduction band. In PspP~çi~lly p.er~,"ed embo~imPntc, at least the object/reference beam can be produced by a cornmercially available serniconductor laser. That is, first coherent light source 1 is a semiconductor laser.
To ensure that a mP~ningful il~t~;lrt;lc~nce pattern is produced, the object and reference beams should be phase locked. Thus, these beams generally should be produced by a single coherent laser r~ n source. Suitable lasers for producing the object and reference beams in accordance with this invention include semiconductor lasers, argon ion lasers, Nd:YAG lasers, etc. In especially preferred embodiments, the laser wavelength is in the red or near infrared region of 620 to 1000 nm. In general, such lasers may be operated at relatively low intensities For example, it has been found that the ferroelectric m~tPri~lc used with this invention require a threshold intensity of about 1000 W/cm2 to effect the transition. More preferably, the laser intensity employed in this invention will be between about 20 and 200 W/cm2. This is well within the realm of ~-u~ tly available diode lasers, such as those used in compact disc players and laser printers, which can emit several hundred milliwatts of coherent, cw-near infra-red and visible radiation. The ability to use small diode lasers represents a cignifi(e:m~ advance towards the comrnercial feasibility of holographic date storage.
While the object and reference beams should be monoch~ atic and coherent, the gating beam need not be. Thus, the gating beam need not be produced by a laser and need not even be monochromatic. In general, the gating beam source need only produce radiation in a wavelength range which will assist the promotion of electrons into the conduction band by a two-photon mrch~nicm (in conjunction with radiation from the object and lere~llce beams).
However, the gating beam should not include wavelengths which tend to promote electrons to the conduction band by a single photon mech~ni.cm Thus, it may be necessary to block some of the gating beam's shorter wavelength photons from striking the recording medium (by, e.g., a notch filter).

CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802The optimal gating beam wavelength will vary depending upon the band gap size ofthe ferroelectric recording mP~ m In especially ~lG~.I~;d embodillle~ " the maximum gating beam wavelength has a photon energy of between about .5 and .75 times the band gap energy. By way of example, a 515 nm gating beam will be suitable for use with lithium S niobate lccoldillg media and a 630 nm gating beam will be suitable for use with ~,L~ lliu barium niobate recording media.
As noted, the int~rm~ te states of the ferroelest~ recording media will have a defined lifetime. Thus, it is important that in pulsed recording systems, the write beam pulse follow the gating beam pulse (or vice versa) within a time frame defined by the lifetime of the 10 interm~.d,~ state. For example, it has been found that some ferroelectric materials provide interm~.-liatP. state lifetimes of about 30-50 mi~ econds. Thus, the gating beam pulse should precede the write beam pulse by no more than 30-50 milli~econds (~sllming that the gating beam is used to promote electrons to the intern~ te level).
In preferred embodiments, the gating beam intensity should be at most about 500 W/cm2. More preferably, the intensity should be between about 5 and 200 W/cm2, and most preferably between about 10 and 100 W/cm2. Further, the gating beam intensity will typically be between about 0.1 and 1 times the reference/obiect beam hlLt;nsiLy. Suitable gating beam sources include, for example, xenon lamps, halogen lamps, argon ion lasers, Nd:YAG lasers, etc. As noted, it may be nPce~ry in some cases to filter the radiation from these sources to 20 meet the above con ~LI~illL~.
Various modifications may be made to the above system without departing from thescope of this invention. For example in some embodiments, a single light source may be employed as a source of the object, reference, and gating beams, such that ~2 = ~1. Such systems, sometimes referred to as "one color, two-photon" systems, can have the general 25 arrangement as shown in Figure 1, but without the use of second radiation source 21. Of course, the photon energy in such systems must be below t'ne energy required to directly promote clc.,llons into the conduction band on its own. Further, the beam used to record the hologram should generally have a higher intensity than the beam used to read the hologram.
This reduces the likelihood of promoting electrons to the conduction band during reading --30 and thereby erasing the hologram. It should be noted that such one-color two-photon systems of this invention resist erasure far better than comparable one-color single photon systems.
This is because the erasure rate during reading is ~l~ollional to intensity in single photon ,.
systems and is proportional to intensity squared in two photon systems. Thus, reducing the read beam intensity (in comparison to the write beam intensity) in two photon systems 35 reduces erasure rate much more dr~m~fic:~lly than in single photon systems.
In another ~ltorn~tive embodiment, not shown, a reference beam, an object beam, and a gating beam are all generated by a first coherent light source of a first wavelength. The gating CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802beam is produced by passing radiation from the first light source through a frec~uency doubler.
Some fraction of the radiation exiting the frequency doubler will have a wavelength that is one-half that of the first wavelength. This short wavelength radiation serves as the gating beam which passes with the reference or object beam through a~r~liate optics and onto the 5 holographic recording media. Preferably, the radiation exiting the frequency doubler will be directed onto a beam splitter that transmits both colllponents and reflects oniy one. The reflected monochromatic portion then forrns the object beam, and the L~ ...ill~ cu~l.ponelll forms the reference beam (long wavelength) and gating beam (short wavelength).
R~ rning again to Figure l, the stored information can be read by blocking the object 10 beam 7 and scattering off of the recorded interference pattern the reference beam 5 or its equivalent in terrns of wavelength and angle of inçit1PIlre with the storage mP-1inm 25. This scattering creates a scattered beam (Es) 27 which passes through a lens 28 to produce an holographic image of the data l~ s~ alion which then is captured by a detector array 29 such as a charge-coupled device (CCD). Output from the detector array can be converted 31 into a 15 serial binary forrn 33 suitable to input into standard coln~uLel- central processing units 35. It should be noted that the reading process is illhe~ ly parallel. That is, the individual bits of data (in the case of a digital recording) are all read .~imlllt~nP.ously and provided as a two-dimensional array. In conventional single CPU colll~u~ g systems, the information in this array typically must be at least partially seri~li7PA for use with the col"~uLel. However, in 20 more advanced parallel processing co~ uL~;l systems, there may be no need to serialize the data image. In fact, holographic storage media should be very efficient memory devices for massively parallel col"~u~....
The systems of this invention may be used to record volume holograms. Such holograms include various "slices" of recorded hlrolllla~ion ove~ illg one another within the 25 recording medium. Typically, the various slices of information are each recorded at a dirrt;le angle by a process known as angular multiplexing. As illustrated in Figure 2, angular multiplexing is accomplished by storing multiple images within a given recording m~Aillm volume by varying the angles of incidence, ~o and ~R of an object beam 45 and a reference beam 43, respectively, on a holographic storage m~flillm 41. ~o and ~3R may be controlled by 30 any number of means. A deflector element 51, for example, can be used to control the angle at which the obJect beam 45 (or the reference beam, not shown) is inci~1ent on the storage m~ Tm All~.llaLively, the orientation of the storage ml tlillm could be manipulated to vary ~o and ~R. not shown. The angular resolution of a volume hologram, which determines the number of holograms that can be accommodated in the m~ lm, increases with the thickness 35 of the storage m~ m To attain good angular resolution, the thickness of a volume holographic recording medium should be on the order of 0.1 to l cm.
In addition to angular multiplexing, a technique known as fre~uency multiplexing may be used to record multiple interference gratings in a single holographic recording mPrlil-m , CA 02237182 1998-0~-08 W O 97/20317 PCTrUS96/18802 Frequency multiplexing allows the gratings to be overlaid on one another by storing separate data repr~sellt~fions at dirr~ wavelengths. Thus, a first data ~ ;se~ lion can be stored on a holographic recording mP~ lm using reference and object beams of one wavelength, and a second data representation can be stored on top of the first data r~pl~;se..lalion using reference 5 and object beams of a dirrelenl wavelength.
Another technique known as phase multiplexing may also be used to record multiple interference patterns in a single m~ m In this technique, the lGr~ ,nce beam is spatially phase mo~ t~l with a set of specially c~esignf~cl orthogonal phase codes so that an interference pattern recorded using a reference beam with a particular code can be read out only by the same reference beam. A clet~ discussion can be found in U.S. Patent 3,612,641 by C.C. Bagler~leld, which was previously incul~o~ d by reference.
A holographic recording prepared in accordance with this invention may be erased by exposing the recording medium to photons having an energy sufficient to promote electrons to the medium's conduction band. Preferably, the radiation will be sufficiently intense that the erasure process can be colllplctcd in a short time. For most ferroelectrics (at least those having a band gap of 4 eV or less), intense ultraviolet radiation will serve this purpose. Often, however, an intense focused white light source will be sufficient to erase the recording m~ lm Further in cases where the gating radiation wavelength is si~nifiç~n~ly shorter than the read/write beam wavelength, a high intensity gating beam itself may actually be used to erase recorded data. After a medium is erased, it can, of course, be rerecorded in the manner described above.

2. Holographic Storage Media: Ferroelectrics with Long-lived Intern e~ te States The holographic recording media of this invention are, as noted, made from photorefractive ferroelectric materials. As used herein, the term "ferroelectric" generally refers to crystals exhibiting an electric dipole moment even in the absence of an external electric field.
Thus, in the ferroelectric state, the center of positive charge of a clystal does not coincide with the center of negative charge. Further, a plot of polarization versus electric field for the ferroelectric state shows a hysteresis loop. A crystal in a normal dielectric state usually does not show .ci~nifiç~nt hysteresis when the electric field is slowly increased and then reversed.
Suitable pho~ Grld ;live, ferroelectric m~t~n~ for use in the recording media of this invention include (13 perovskites such as BaTiO3, CaTiO3, KNbO3, and KTaxNb1_x03; (2) oxides such as LiNbO3, LiTaO3; (3) complex oxides with a tungsten bronze structure such as SrxBal xNb206 (SBN) (4) non-oxide sulfur iodides such as SbSI, SbSeI, and BiSI; (5) bismuth germanium compounds such as Bil2GeO20 and Bil2SiO20; and (6) PLZI
ceramics such as PbLaZrTi. Examples of prere,~t;d photorefractive ferroelectrics include w o 97nO317 PCT~US96/18802 lithium niobate, pot~ . lithium niobate, lithium t~nt~l~te, barium tit~n~t~, strontium barium niobate (SBN), lead barium niobate (P13N), and barium strontium potassium sodium niobate (BSKNN). Within this group, the compounds ~ iulll barium niobate, barium titanate, and lithium niobate have been found to ~e,rc~ quite well.
~ 5 Generally, the ferroelectric recording medium should contain no more than about 0.01% atomic of an optically active h~ ;ty. More preferably, the l~co~ g medium should contain no more than about 100 parts per million (per mole basis), more preferably no more than about 10 part per million, and most preferably no more than about 1 part per million of an optically active illl~Ulity. Thus, an important feature of the present invention is that the ferroelectric recording media remains undoped by an optically active hll~uliLy. An optically active iln~3uliLy is an illl~Ulit~y that tends to absorb either the gating beam radiation, the readJwrite bearn radiation or both. Such illl~ulilies should be avoided because they absorb radiation that could otherwise be used to promote electrons to the ferroelectric's con~ Cction band.
It has been surprisingly observed that some such undoped ferroelectric m71teri~1s support intermediate states with long lifetimes between about 1 and 100 milli~econds For comparison, transition metal elements such as iron and copper provide intorm~ t~-. electronic state life times on the order of about 0.1 microseconds and rare earth elements such as praseodymium provide interm~ t~ electronic state life times on the order of about 0.1 to 1 microseconds. It has also been observed that the intermP~ t~. states in ferroelectrics of this invention give rise to absorptions over a wide spectral range in the near infra-red and visible spectral regions (e.g., between about 400 and 1550 nanolllel~l~). While not wishing to be bound by theory, it is believed that such intermediate states are attributable to defects in the ferroele&tric crystal lattice that may result from atomic vacancies in the lattice, h~l~uliLies, etc.
Ferroelectr;~ single crystals suitable for use with this invention can be pr~p~ed according to conventional methods known in the art or can be custom made by certain vendors such as Virgo Optics of Port Richey, FL, Deltronic Crystal Tn-ln~tries, Inc. of Dover NJ, and Fujien C~tp~ll Crystals, Inc. of Fuzhou, Fujian Peoples Republic of China. Depending upon the nature of the material, growth may be by deposition from the vapor phase (e.g., ~UU~ g, evaporation, ablation, chemical vapor deposition), by bulk process, such as by melt growth, from flux, etc. In general, melt growth involves fusing inorganic components in the correct ratios to forrn the ferroelectric and then pulling a single crystal from the melt. Such techniques are described in, for example, "Development and modification of photorefractive plupt;l~ies in the tungsten bronze family crystals" by Neurgaonkar in Optical Engineering, Vol. 26, pg. 392 et se~., May 1987 which is incorporated herein by reference for all purposes.
To increase the number of long-lived interm~ t~ states in ferroelectric crystalline lattices, the ferroelectric can be specially processed. For example, to increase the llulllber of CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802atomic v~c~ncie~"c, the lattice may be bol"l~alded with a particle beam or an X-ray beam. This will kick out some atoms in the lattice and thereby introduce vacancy defects. In oxygen cont:~inin~ ferroelectrics, vacancies may also be introduced by subjecting a sarnple to re~ cing conditions such as a high t~lllpelalllre anneal in a vacuum or inert atmosphere (e.g., an argon 5 atmosphere). Preferably, the anneal is con(l~ .te~l at a te~ eldlul~ of between about 300 and 1000 ~C (but not above a phase transition) for between about 0.1 and 1 hours. It has been reported that such annealing can indeed affect the absorption ~"~e,Lies of ferroelectrics. See Sweeney et al., "Oxygen Vacancies in Lithium Niobate" Appl. Phys. Lett. 43 (4), pp. 336-338 ( 1983) which is incorporated herein by reference for all purposes.
Defects may also be introduced by doping the ferroelectric material with optically inactive dopant atoms such as certain transition metals. Such dopant atoms should be of a size and chemical nature to locally disrupt the lattice structure, and thereby introduce defects.
Further, the dopant atoms should not absorb radiation at the wavelength of the read/write beam.
The thickness of the holographic recording media should be at le~t several tin~es ~, where ~ is the wavelength of the radiation used to generate the hllelrelt;llce pattern. In further preferred embodiments, the holographic recording media will have a thickness of between about 0.1 and 10 millim.oters. In the example set forth below, it was found that holographic recording media having dimensions of S millim~ters by 5 millim~.ters by 5 millimt-ters 20 performed quite well.
Further, the recording m~ m should be coated with an anti-reflective material such as is employed to coat lens. Such materials include, for exarnple, m~nesillm fluoride, magnesium oxide, and beryllium oxide. ~n general, such anti-reflective coatings should be provided on all sides of the holographic recording medium through which radiation will pass.
25 Typically, it will be convenient and desirable to coat all outer surfaces of the medium.
Still further, it will generally be n~ce ~ry to pole the single crystal ferroelectric recording mPrlillm before recording a grating on it. Conventional poling may be carried out generally by use of an applied electric field ~ il";~ d during cooling of a m~t~ri~l through its Curie point to some lower ~ e~ is described in Ferroelectrics, 4, 189 (1972).

3. Examples The first two examples below involved e~ e,h~ performed on a 0.5 x 0.5 x 0.5 cm sample of single-crystal lithium niobate (LiNbO3) supplied by Deltronic Crystal Tn~ tri~s of Dover, New Jersey. Chemical analysis of the sample showed all major transition metal CA 02237182 1998-0~-08 W O 97/20317 PCTAUS96/18802inlpulilies to be below 1 part per million. The sample was poled to a single domain by the supplier.

P~ mrle I ~One-Color Two-Photon Recordin~) For these experiments, output from a laser 64 was split by a beam splitter 66 (reflectance = 50%) to produce a reference laser beam (ER)60 and an object laser beam (Eo) 62, as shown schematically in Figure 3a. These beams were then recombined to cross in the poled single-crystal sample of LiNbO3 74 at an angle of approximately 5~ where they produced an interference pattern. Specifically, the object beam 62 was reflected off of mirrors 70 and 72 and thereby directed onto sarnple 74, while the reference beam 60 was refl~ctec1 off of mirror 68 and thereby directed onto sample 74. The c-axis of the sample was oriented in parallel with the two tr~ncmitting surfaces of the sample and was in the polarization plane of the object and reference beams The holographic grating was written with a single one second laser pulse from laser 64 at a power density of about 2 W/mm2. The grating was read by blocking the object beam 62 so that only the reference beam 60 would be incident on the sample and diffracted by the holographic grating to produce a scattered beam (Es) 76. Thus, the holographic grating was read with continuous radiation from laser 64 at about one-half the laser intensity of the recording laser beams. A silicon photodiode 78 (EG&G Model. FND-100) was used tomonitor the intensity of the diffracted beam. The diffraction efficiency was ~lP~lllced as 1l =
IE~/Eol2. After all measurements n~ce~s~ry to deduce 1l were completed, radiation from a halogen lamp was focused on the sample to erase the gratings and allow reuse.
The above-described reading and writing steps were conducted at a series of wavelengths. A tunable ring-dye laser (Coherent Model 699-21) with a wavelength range of 580 to 650 nm was used in ~GIiln~nls conducted at 580 nm, 600 nm, 62û nm, and 650 nm.
A tunable Ti:sapphire laser (Coherent 899-01) was used in e~e.in,e~ , c-nrluçt~-~l at 690 nm, 720 nm, 755 nm, 795 nm, 835 nm, 875 nm, 915 nm, and 995 nm.
It was found that for wavelengths of at least about 620 nm, the dirrldclion efficiency r is ~-u~ullional to the fourth power of laser power density (i.e., ll oc I4). This is consistent with a two-photon process. It was also found that for wavelengths from 580 to 620 nm, the diffraction efficiency rl is ~upo-lional to I2-5 to 3-5. This suggests a mixture of one and two-photon processes (1l oc I2 for one-photon processes). Thus, longer wavelengths (in the region of 620 nm or greater) can likely be used in holographic recording systems of this invention without causing rapid erasure during reading.

CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802 For fixed laser power densities, the efficiency 1l was found to fal} exponentially with wavelength at wavelengths greater than about 620 nm. This is illustrated in a plot of dirrla ;Lion efficiency versus laser wavelength for a one-color process (curve 120) in Figure 4.
The data shown in this plot were taken with laser beams having a power density of 2 W/mm2 5 and a pulse length of l second. As can be seen, at wavelengths of at least about 795 nm (and power densities of 2 W/mm2), the efficiency Tl was less than 10-6, which is comparable to the sc~lL~ g from background radiation.

Fx~mple 2 (Two-Color Two-Photon Recording) In two-color experiments, the expe.h"G.. l~l al~palalus was es~t~nt~ ly identical to that used in the one-color e~fillle~ , except that second laser 83 was employed to produce a gating beam (EG) 85. Specifically, as shown schemzltiç~lly in Figure 3b, the experimental ayl~alus employed an unmodulated gating laser beam (EG) 85 which was directed on a poled single-crystal LiNbO3 sample 94 simultaneously with an object (Eo) 88 and a reference (ER) 86 beam to write the grating in the sarnple material. The object (Eo) 88 and reference (E~) 86 beams were provided from a tunable laser 80 whose beam was split by a beam splitter 82.
The object beam 88 was reflected off mirrors 9(~ and 92 and onto sample 94. The reference beam 86 was reflected off mirror 84 and onto sample 94. The gating beam 85 was the 514.5 nm fixed-wavelength output of an argon ion laser 83 (Coherent Model 200) polarized 20 perpe.n-lic ll~r to the c-axis of the LiNbO3 sample. It had a t~ m~ter of about 0.2 millimPt~r and a power density of about 0.8 W/mm2. As in the one-color experiments, the grating was read by blocking the object beam 88 so that only the reference beam 86 would be incident on the sample and diffracted by the holographic grating to produce a scattered beam (Es) 96. The intensity of beam 96 was measured by a silicon photodiode 98.
As in the one-color experiments, the reading and writing steps were cnn~l--cte-1 at a series of wavelengths with a tunable ring-dye laser and a tunable Ti:sa~yhi.G laser. In each case, the diffraction efficiency ~ was measured. A plot rl versus wavelength for the two-color experiments is shown as curve 126 in Figure 4. As before, the grating was recorded with a single one second pulse from laser beams (object 88 and reference 86 beams) having a power density of 2 W/mm2. The gating beam (EG) 85 was also provided as a one second pulse at a power density of 0.8 W/mm2 as mentioned above. The writing and gating pulses were provided .~imnlt~neously .
The gating beam provides high energy photons used to promote electrons to intPrm.orli~t~ states close to the conduction band. Some electrons residing in these interm~Ai~t~
states are then promoted into the conduction band by photons from the writing beam. In a one color situation, in contrast, the writing beam must promote the electrons first to the W O 97/20317 PCT~US96/18802 intermP~liz~te states and then to the conduction band. Thus, in a two-color process, electrons are more efficiently fixed (with regard to the writing beam intensity) in the recorded illle.re~ ce pattern, so that the signal strength in the dirrla~ d intensity (Is) is improved.
Evidence of this is seen in Figure 4 which shows the dirrla~;lion efficiency 1l from the two-5 color experiment to be lO00 times greater than that from the one-color ~ elilllt;llL at 755 nm (for the object and reference beams). Further, the enhancement factor (defined as ratio of the diffraction efficiency with the gating beam and without the gating beam) grew from 1 at 580 nm to lO00 at 755 nm.
In the two-color e~cfi~ nt~, like the one-color e~ ..t~, the diffraction efficiency 11 was found to decrease with wavelength. As shown in Pigure 4, curve 126, the decrease was exponential with a rate about one-half that of the one-color case.
The erasure rate of holograms recorded by two-color experiments was determined as follows. For holograms recorded at 755 nm with the ~i.ct:~nce of the 514.5 nm gating beam under the same conditions as described above, the diffraction efficiency rl retained 60% of its 15 original value after 30 min~ltes of reading (under the reading conditions described above).
Further, a gated recording made at 835 nm retained 94% of its original diffraction efficiency after 20 minutes. A much more rapid erasure is expected for a one-photon recording process.
For example, it can be shown by straight forward calculation that for holograms recorded at 755 nm under the same conditions, but without a gating beam, the diffraction efficiency 20 would drop to 60% of its initial value after only 5.3 seconds of reading. Thus, the gated grating written at 755 nm as described above should be 340 times more resistant to erasure from reading than the llng~Pd grating. This assumes that for the ideal one-photon process, the grating build up is exponential (i.e., llWriting(t) = (l - e~~It)2) and that the grating erasure is also exponential (i.e-, llerasure(t) = llo(e ~ ) ) As noted, two-photon recording at low powers (e.g., in the neighborhood of lO00 W/cm2 or lower) requires a ferroelectric material having long Iived intPrmerli~t~ states. Such states were eollrilll,ed by the formation of gated gratings at low write laser powers as described above. The lifetime of these intermP~ t~ states was determined as follows. The writing beams and the gating beam were chopped to l 0 rnillisecond pulses and the resulting 30 diffraction efficiency rl was measured as a function of the delay of the writing pulses from the gating pulses. As shown in Figure 5, between delays of about lO to 80 milli~econds, the diffraction efficiency rl decreases dramatically, implying an interrn~ tP. level lifetime of l~;lweell about 30 and 50 milliseconds (~uming that ~ is proportional to (exp(-t/~o))2 where J ~0 iS the interrnPdi~tP level lifetime).

CA 02237182 1998-0~-08 W O 97/20317 PCT~US96/18802Ex:~m~le 3 Further one and two-color experim~nt~ were performed with undoped and poled 60%-strontium, 40%-barium niobate (Sro 6Bao 4Nb2O6; also known as SBN:60), a ferroelectric material hereinafter referred to as SBN. The SBN sample used in these experiments was 5 supplied by Deltronic Crystal TnAllctries of Dover, New Jersey.
The grating measurements obtained with the SBN were essentially the same as those obtained with the LiNbO3 crystal. However, because the SBN band gap is narrower (3.4 eV
as opposed to 3.9 eV for LiNbO3), the one-photon effect is dominant throughout the visible region, and the observable gated response shifted to longer wavelengths. Thus the gating 10 beam illlt;;n.,ily was reduced to 0.1 Wlmm2. Other conditions were identical to those described above. At a 915 nm reference beam wavelength, the gated diffraction efficiency r was 0.7%, while at 995 nm, that efficiency had dropped to about 0.1%. Further, at 915 nm, the enhancement factor (the ratio of two-color to one-color diffraction efficiencies) was greater than 300. It should be noted that the 514.5 nm gating beam produces a rather large one-15 photon photorefractive effect, and hence is not an optimal gating wavelength for SBN. GivenSBN's band gap, a gating beam of about 630 nm or larger would produce many fewer single photon transitions and would therefore be more a~ u~liate for two photon transitions.
Fxample 4 Experiments were also pe,r,lllled with a barium titanate crystal supplied by Virgo 20 Optics of Port Richie, FL. Under identical conditions used with SBN, a gated diffraction efficiency of approximately 0.1% was obtained at 915 nm, which represents an enhancement factor about 20. The band gap of this crystal is only 3.1 eV. The optimal gating wavelength is thus essim~t~A to be greater than 690 nm.

4. Conclusion Although the foregoing invention has been described in some detail for purposes of clarity of underst~nAing, it will be a~ucllt that certain changes and modifications may be practiced within the scope of the appended claims. For instance, although the specification has described holographic photorefractive memory, the ferroelectric m~t~ri~l~ of this invention will be useful in all photolcrldeli~re applications involving signal proces~ing, routing, switching, and optical interconnections (see, for example, "Selected Papers on Optical Cu~ ulhlg" SPIE
Milestone Series, H. John Caulfield and Gregory Gheen, editors, SPIE Optical Engineering Press, 1989, which is incorporated herein by reference for all purposes). In addition, the reader will understand that varying certain recording conditions may require a corresponding adj-l~fm~nt in certain of the above discussed pd dllleters. For example, increasing the lclll~eld~lc of the ferroelectric recording m~Aillm will decrease the lifetime of the medium's WO 97/20317 PCT~US96/18802interm~ t.- states. Thus, the time frame in which a write pulse must follow a gating pulse will have to be decreased by a col.es~ ding amount.

Claims (33)

Claims WHAT IS CLAIMED IS:

1. A holographic recording system comprising:
a holographic recording medium including a ferroelectric material that contains no more than about 0.01% atomic of an optically active impurity;
a first radiation source for providing coherent monochromatic radiation which is first divided into a reference beam and a spatially modulated object beam and then recombined to form an interference pattern on a first region of said holographic recording medium; and a second radiation source for providing a gating beam which is optically coupled to a second region of said holographic recording medium which second region is at least partially coextensive with said first region of the holographic recording medium, whereby photons from the first and second radiation source together promote electrons of the holographic recording media to a conduction band by a two-photon process such that the interference pattern is recorded in said holographic recording medium, wherein at least one of the first and second radiation sources is a continuous wave laser, and wherein said optically active impurity is active with respect to at least one of the coherent monochromatic radiation and the gating beam.

2. The holographic recording system of claim 1 wherein the continuous wave laser provides radiation at an intensity of at most about 1000 watts/cm2.

3. The holographic recording system of claim 1 wherein the continuous wave laser provides radiation at an intensity of between about 20 and 200 watts/cm2.

4. The holographic recording system of claim 1 wherein the first radiation source is a continuous wave laser.

5. The holographic recording system of claim 4 wherein the continuous wave laser is a diode laser.

6. The holographic recording system of claim 1 wherein the second radiation source is a continuous wave laser.

7. The holographic recording system of claim 6 wherein at least one of the firstand second radiation sources is a diode laser.

8. The holographic recording system of claim 1 wherein the second radiation source is an incoherent radiation source.

9. The holographic recording system of claim 1 wherein the first radiation source provides coherent monochromatic radiation of a wavelength within the red or infrared regions of the electromagnetic spectrum.

10. The holographic recording system of claim 1 wherein the ferroelectric material contains no more than about 10 parts per million of optically active impurity.

11. The holographic recording system of claim 1 wherein the ferroelectric material has been subjected to a reduction process to introduce oxygen vacancies in the ferroelectric material.

12. The holographic recording system of claim 1 wherein the ferroelectric material is doped with an optically inactive dopant, thereby providing defects in the ferroelectric material.

13. The holographic recording system of claim 1 wherein the ferroelectric material is selected from the group consisting of lithium niobate, potassium lithium niobate, lithium tantalate, barium titanate, strontium barium niobate, lead barium niobate, and barium strontium potassium sodium niobate.

14. The holographic recording system of claim 1 wherein the first radiation source produces radiation of a first wavelength and the second radiation source produces radiation having at least a second wavelength, and wherein the first and second wavelengths are different from one another.

15. The holographic recording system of claim 14 wherein the second wavelength is shorter than the first wavelength.

16. The holographic recording system of claim 1 wherein the first radiation source produces radiation of a first wavelength and the second radiation source produces radiation having at least a second wavelength, and wherein the first and second wavelengths are the same.

17. The holographic recording system of claim 1 wherein photons from the first radiation source have a first defined photon energy and at least some photons from the second radiation source have at least a second defined photon energy, and wherein the sum of the first and second defined photon energies is at least as great enough to promote electrons or holes into the conduction or valence bands respectively, but neither the first nor the second defined photon energies is great enough to promote electrons or holes into the conduction or valence bands respectively.

18. A method of writing to a holographic recording medium containing a ferroelectric material that contains no more than about 0.01% atomic of an optically active impurity, the method comprising the following steps:
separating a first radiation beam of a first wavelength into a reference beam and a spatially modulated object beam;
combining the reference beam and the spatially modulated object beam to form an interference pattern on a first region of said holographic recording medium; and directing a gating radiation beam containing at least a second wavelength onto asecond region at least partially coextensive with said first region of the holographic recording medium, whereby photons of the first and second wavelengths together promote electrons of the holographic recording media to a conduction band by a two-photon process such that the interference pattern is recorded in said holographic recording medium, wherein at least one of the first radiation beam and the gating radiation beam is produced by a continuous wave laser, and wherein said optically active impurity is active with respect to at least one of the first radiation beam and the gating radiation beam.

19. The method of claim 18 wherein the continuous wave laser provides radiation at an intensity of at most about 1000 watts/cm2.

20. The holographic recording system of claim 18 wherein the continuous wave laser provides radiation at an intensity of between about 20 and 200 watts/cm2.

21. The method of claim 18 wherein the first radiation beam is produced by a continuous wave laser.

22. The method of claim 18 wherein the gating radiation beam is produced by a continuous wave laser.

23. The method of claim 18 wherein the ferroelectric material contains no more than about 10 parts per million of optically active impurity.

24. The method of claim 18 wherein the ferroelectric material is selected from the group consisting of lithium niobate, potassium lithium niobate, lithium tantalate, barium titanate, strontium barium niobate, lead barium niobate, and barium strontium potassium sodium niobate.

25. The method of claim 18 wherein the ferroelectric material has been subjected to a reduction process to introduce oxygen vacancies in the ferroelectric material.

26. The method of claim 18 wherein the ferroelectric material is doped with an optically inactive dopant, thereby providing defects in the ferroelectric material.

27. The method of claim 18 wherein the first and second wavelengths are different from one another.

28. The method of claim 27 wherein said second wavelength is shorter than said first wavelength.

29. The method of claim 18 further comprising a step of erasing the recorded interference pattern.

30. The method of claim 29 wherein the step of erasing is performed by focusing radiation of the second wavelength onto the recorded interference pattern.

31. The method of claim 29 wherein the step of erasing is performed by focusing white light onto the recorded interference pattern.

32. The method of claim 18 wherein the first radiation beam and the gating radiation beam are pulsed, and wherein the gating radiation beam pulses precede the first radiation beam pulses by no more than about the lifetime of an intermediate state of the holographic recording medium.

33. The method of claim 32 wherein the gating radiation pulse precedes the firstradiation pulse by no more than about 30-50 milliseconds.