Method for Printing a Near Field Photoinduced Stable Structure, and Optical Fiber Tip for Implementing Same

Abstract

The method comprises emitting appropriate polarisation light waves on a photosensitive material for inducing therein a topographic modification through an aperture of at the most 100 mm bound by an opaque area, said layer being arranged at the most 100 mm away from the aperture.

Description

This invention relates to a method for printing a stable structure in relief in a material comprising a photosensitive compound, through radiating said material from one or more localised light sources with a size lower than the wavelength of said radiation.

The increasing development of the computer processing powers and the spread of the information exchange make necessary to get available means allowing for the storage density to be increased on information media.

Generally, the information on the usual media are printed in the form of repeating patterns obtained through index change or through local topographic modification.

One of the major problems lies in increasing the printing resolution.

It is already known to achieve surface topographic modifications, for example in hybrid sol/gel materials containing azobenzene groups.

In the thesis by Philippe Bertrand entitled <<Near field optical microscopy. An optical response at the submicronic scale>> dated 22 Jan. 1999, available in the university library of the Ecole Polytechnique, 91128 Palaiseau, France, an experiment is disclosed where a topographic structure has been photoinduced in a sol/gel layer, using as a light source an optical fibre having the tip obtained through heat sequential stretching and non attacked in the acid.

The printed structure consists of letters in turn consisting in plots of 15 nm height and 40 nm×40 nm side.

A similar result is disclosed in an article entitled <<Photochromic sol-gel films containing dithienylethene and azobenzene derivates: from the optical component design to the optical data storing>>. Mol. Cryst. Liq. Cryst., 2000, vol. 344, pp 77-82.

Generally, both those documents point out that it seems excluded that such a topographic modification originates from a simple thermal effect, as it could be observed from other prior art documents (<<Near-field magnetooptics and high density data storage>>), E. Betzig et al.; Appl. Phys. Letters, vol. 61 (2), pp. 142-144 (1992), in so far as the light polarisation plays a part on distorting the layer surface, and not only the light source intensity.

The first two above-mentioned documents disclose preliminary results and the method being briefly disclosed has the disadvantage of lacking reproducibility.

Indeed, for a readable inscription, the method should allow, on the one hand, to perfectly control the geometry of the written patterns, on the other hand, to avoid any alteration of the layer structure in the pattern circumference, and precisely in the area separating two patterns.

In addition, it would be desirable to obtain patterns having a substantially identical relief.

Too high a fluctuation of the pattern level could indeed impair information reading.

Finally, it would be desirable to control and in particular to increase the relief, i.e. the pattern height, without altering the resolution of the printed array, whilst maintaining the printing stability over time.

An object of the invention is therefore to provide an improved locally printing method, in particular overcoming the above-mentioned problems.

The method according to the invention is a locally printing method for a stable structure in relief comprising a step of emitting appropriate polarisation light waves on a material layer comprising a photosensitive compound adapted to induce a topographic modification of said layer under the effect of said light waves, and is characterized in that said appropriate polarisation light waves are emitted on said layer through an aperture with a maximum size lower than or equal to 100 nm, circumferentially bound by an opaque area preventing the light waves from being propagated outside said aperture, said layer having a surface located at a distance lower than or equal to 100 nm from said aperture.

The above-mentioned opaque area restricts light leakage. It makes it possible to bound a truly localised aperture for the light emission and to control the size thereof. The electromagnetic field is thereby restricted to the vicinity of the surface to be printed and the size of the printed pattern and the resolution of an array comprising such patterns are perfectly controlled.

Because of the size of the above-mentioned aperture and the positioning thereof with respect to the photosensitive material layer, the method allows to print the structure through a near field light radiation.

As is known, it is meant under near field light radiation, an irradiation by the electromagnetic field components having their amplitude that exponentially decreases with the distance to the field generating light source.

It is further altogether surprising to observe that the thus locally bound aperture makes it possible to keep the polarisation of the light being emitted through said aperture.

Preferably, said aperture has a size lower than or equal to 60 nm, and more preferably, a size ranging from 40 to 60 nm.

According to a preferred embodiment of the invention, said appropriate polarisation light waves emitted through said aperture are transmitted by an optical fibre comprising an end portion having the general shape of a tip at the end of which said aperture is located.

More precisely, the tip end has the form of a truncated cone, bound by a globally planar side and a circumferential side and the aperture forms the globally planar side of the truncated cone, whilst the opaque area is the cone circumferential surface.

Preferably, the end portion of the optical fibre has the property of maintaining the polarisation features of the propagating light waves.

To this end, the end portion of the optical fibre has been obtained through heat sequential stretching.

The remainder of the disclosure refers to the accompanying drawings.

FIGS. 1 to 6 illustrate optical fibre ends observed with a scanning electron microscope (MEB) at various steps of the method according to the invention.

FIG. 7 schematically illustrates the device being used for implementing the printing method according to the invention.

FIG. 8 illustrates a top view of an array obtained according to the invention.

FIG. 9 illustrates a sectional view of a relief structure printed on the surface of a layer comprising a photosensitive material.

Preparing optical fibres useful for implementing the method will be now described in further detail.

The general principle of thinning an optical fibre through heat stretching is derived from that developed for biology micropipettes. It consists in locally heating, through infra-red radiation absorption of a CO₂ laser, an optical fibre kept in tension. Above a determined temperature, the heated area of the fibre stretches and gets thinner. This leads, in some conditions, to the fibre breaking and a tip being formed.

The heat sequential stretching allows for the tip morphology to be controlled in situ upon the formation thereof. To this end, the heat stretching of the fibre occurs in a sequence of heating-cooling steps which are controlled by measuring the stretching and the thinning of the fibre.

The sequential stretching consists in heating the optical fibre by a plurality (typically about one hundred) of laser pulses of a varying duration, controlled by an electronic system. The pulse number n, of a τ_n duration, produces a Δ1_n stretching. Between two pulses, a delay is provided so as to allow the fibre to cool down. The process comprises four stages:

• • stage 1 comprises defining the initial stretching conditions,
  • stage 2 achieves some thinning of the fibre through constant steps and imparts a conical morphology to the tip,
  • stage 3 accurately prepares the fibre diameter for forming the tip, and
  • stage 4 only comprises a single laser pulse with a duration such that it achieves breaking the fibre and forming the tip.

During stage 1, the laser pulse duration is incremented by Δτ₁ as long as the stretching generated by a pulse does not exceed a given threshold Δ1₁. When, after the pulse number i, a stretching Δ1_i higher than threshold Δ1_I is measured, stage 1is completed and the initial conditions of the stretching are defined. Duration τ_i of such a pulse is selected as a reference time τ_ref to define the time scale in the whole process; this is the required pulse duration for stretching by length Δ1_I the fibre of a typical initial diameter D₁=125 μm. The stretching threshold is an adjustable parameter the selected typical value of which is Δ1_I=3 μm. In the hereinabove described operating conditions, for a CO₂ laser with a power equal to 14 W, delivering a 2 mm diameter beam, and for the Δ1_I indicated value (+3 μm), the value of the reference pulse duration is typically τ_ref=50 ms. The value of the increment Δτ_I of the pulse duration is selected so as to be small relative to the value of τ_ref: Δτ_I=1 ms.

Stage 2 corresponds to a thinning procedure during which the stretching produced by each pulse is required to be constant. The first parameter being defined is the set value, i.e. the stretching value Δ1_C which is required to be obtained. It is then required, for the pulse number n, to select the value of τ_n, which will produce the stretching Δ1_n=Δ1_C. It is obvious that, as the fibre gets thinner, the pulse duration required for obtaining a fixed value of the stretching decreases. Thus, τ_n will necessarily be lower than τ_n−1.

The decrease of the pulse duration is iteratively controlled. For determining the duration of the nth laser pulse, first the duration τ_n⁰ s defined having the deviation to τ_n−1 identical to that between τ_n−1 and τ_n−2:

τ_n−2−τ_n−1=τ_n−1−τ_n⁰  (1)

Thereafter, the effective duration τ_n of the nth pulse is determined by adding to τ_n⁰ a correction taking into account, on the one hand, the deviation to the set value as measured during the pulse n−1 as well as, on the other hand, the slope with which the stretching (or the deviation to the set value) varies between pulses n−2 and n−1:

τ_n=τ_n⁰+C_rδv_n=τ_n⁰+C_r[(Δ1_C−Δ1_n−1)−(Δ1_n−1−Δ1_n−2)]/Δ1_C  (2)

Such correction, the value of which depends on the measurement of the effect generated by the previous pulses, is one that gives to the method a feedback character. The correction value depends on the coefficient C_r. Such a coefficient plays the part of a feedback gain applied to the error signal δv_n.

Its value is selected so as to keep the error signal as low as possible, i.e. to keep Δ1_n as close as possible to Δ1_C. The set value Δ1_C and the feedback coefficient C_r are also adjustable parameters. In the above-described operating conditions, their respective values are typically Δ1_C=3 μm and C_r=1.3 ms. Stage 2 is completed when the value of the pulse duration calculated according to equation 2 becomes lower than ατ_ref.

Coefficient α is an adjustable parameter a typical value of which is 0.45. At that time, the fibre thinning has made it possible to reach a diameter D₂ per pitch δD₂ (the typical values of D₂ and δD₂ in the above described operating conditions are respectively 20 μm and −1.5 μm). It is then required to continue the method with finer pitches so as to prepare with the utmost accuracy the fibre diameter which will allow for the tip end to be formed with the desired geometrical features.

Stage 3 is then defined during which the pulse duration decreases stepwise. During that stage, this is not only the measurement of the fibre stretching which is used for determining the value of the steps, but also the diameter D₃ as measured in situ through a microscope. As long as the fibre thinning (i.e. the variation δD_n of the fibre diameter) is lower than the threshold δD₃=0.5 μm and the stretching δ1_n of the fibre is lower than threshold δ1_S=1 μm, the procedure is continued without changing the pulse duration. It should be noted that the choice of the threshold value δD₃=0.5 μm is directly related to the final resolution of the microscope being used for measuring the fibre diameter. With a view to the instrumental constraints, such a resolution is hard to improve and could in no way be much lower than 0.5 μm. The stretching threshold value (1 μm) is arbitrarily determined as being a typical value corresponding to the threshold thinning. Passing from one step to another one therefore occurs when one of the measured values, the thinning or the stretching of the fibre, exceeds the threshold value. When the fibre reaches a diameter D₃, stage 3 comes to an end. D₃ is an adjustable parameter. In the above described operating conditions, its typical value is D₃=15 μm.

The feedback principle changes again for stage 4. The duration of the last pulse is not fixed as a function of the result of the previous pulses. The stretching speed V_e of the fibre is permanently measured. As long as the laser is not heating the fibre, V_e is null. The electronic control of the laser supply triggers the pulse which is only interrupted when V_e exceeds an established threshold V_es. The stretching speed threshold is an adjustable parameter the selected typical value of which, according to the above described operating conditions, for forming the desired tips, is V_es=0.002 m.s⁻¹. The resulting tip further to this last pulse, visualised by MEB in FIG. 1 and 50,000× enlarged, has the shape of a truncated cone the end of which has a diameter D₄ ranging from 0.5 μm and 1 μm. The corresponding values of the <<apex>> angle θ range between 40° and 60°.

D4 and θ are illustrated in FIG. 2, where the fibre is identical as that in FIG. 1.

The accurate values for the fibre in FIGS. 1 and 2 are θ=48° and D4=700 nm.

Through such a thinning procedure, the fibre is divided into two identical tips. One of those two twin tips is used as a control tip, the other will be used for producing the optical fibre according to the invention or nanoprobe.

After the heat sequential stretching, the end portion of the optical fibre is subjected to an isotropic chemical etching, preferably with an acid, more particularly hydrofluoric acid.

The chemical etching step may be considered as the last step of the feedback process: the advancing state of the method is measured and, depending on the result of such a measurement, the chemical processing which the tip is subjected to is adjusted. This last step has the feature of not modifying the general shape of the tip, but it makes it possible to accurately adjust the size of its end. At the stretching completion, the size and the shape of the preformed tip, which characterise the advancing state of the method, are measured on MEB photographs of the control tip.

The complementary chemical etching of a preformed tip through heat sequential thinning is based on three principles:

• • the chemical etching is isotropic and thus keeps the truncated cone shape of the tip: the diameter of the higher side of the truncated cone (the diameter of the probe aperture) may be decreased from the value D4 to the selected value D without the angle at the tip apex, θ, being changed;
  • the chemical etching speed may be adjusted independently from the other parameters of the method by selecting the concentration of the hydrofluoric acid solution;
  • the chemical etching duration required for obtaining the desired final diameter only depends on the chemical etching speed and on the values of D4 and θ (measured on the control tip).

If e is defined as being the thickness to be dissolved and the linear etching speed or etching speed in the direction normal to the surface as V=de/dt, the etching time t required for obtaining the final value D for the aperture is [cos(θ/2)(D4−D)]/2V[1−sin(θ/2)].

FIG. 3 shows the end of a tip observed through MEB (50,000× enlargement) before chemical etching and FIG. 4 shows the same tip after an etching in HF 12% at 25° C.

It can be seen that the chemical etching is isotropic and keeps the truncated cone shape of the preformed tips. Moreover, the complementary etching does not measurably affect, at such a scale, the surface roughness. It even seems that some uneven features will be <<rubbed out>>.

In the following step, the circumferential surface of the cone is made opaque.

Such a step is preferably performed through coating at least one metallic layer.

Any coating techniques are possible as long as the coating technique does not affect the geometry of the preformed tip end.

Preferably, the coating occurs under vacuum.

The nature of the metals to be used as well as the coated corresponding thickness are variable and should be such that there will be no passage left for light, and that any light leakage outside the aperture arranged in the globally planar side of the truncated cone will be prevented.

Generally, the total thickness of the metallic layer(s) ranges from 50 to 200 nm.

Preferably, a first chromium based layer is coated followed by an aluminium based layer.

In such a case, the chromium based layer has a thickness ranging from 2 to 20 nm and the aluminium based layer has a thickness ranging from 50 to 180 nm.

Typically, the evaporation occurs under a secondary vacuum from metallic filaments heated through Joule effect.

Within the evaporation enclosure, the end of the tip is provided at an adapted distance, for example 20 cm, above the metallic filaments.

In order that the higher side of the truncated cone formed by the end of the tip does not receive any metal, i.e. in order that the aperture of the probe be maintained, the tip is generally slightly upwardly tilted, for example by a 15° angle relative to the horizontal.

In order to obtain a homogeneous coating, the tip moves upon evaporation according to two rotation movements:

• • a rotation movement around its axis;
  • an azimuthal rotation movement around the vertical, i.e. according to the axis following the average direction of the evaporation flow.

The finally obtained optical fibre is an optical fibre useful as a near field probe comprising an end portion having the general shape of a tip the end of which having the shape of a truncated cone bound by an opaque circumferential surface and a globally planar side with a maximum dimension equal to 100 nm, defining an aperture allowing light waves to pass, said end portion maintaining the polarisation properties of propagating light waves.

FIG. 5 shows a tilted view of such a fibre, being thinned, chemically etched and metallized as previously indicated, 20,000× enlarged and FIG. 6 shows a front view of the same fibre, 103,920× enlarged.

FIG. 5 shows that the metallization processing has not affected the global geometry of the end of the tip.

In FIG. 6, the probe aperture can be seen.

Its diameter is approximately 50 nm. The performed coating can be distinctly seen, in the aperture periphery, which has the form of small compact islands with an average size of 100 nm. An increase of the above-mentioned speeds of metallic evaporation makes it possible to obtain a more regular coating.

The resulting metallic layer prevents any light leakage outside the aperture.

Such a tip is ready to be used in the printing method according to the invention.

The stable structures in relief are printed in material layers comprising a photosensitive compound and generally having a thickness lower than 50 nm.

A preferred photosensitive compound is a compound subjected to an isomerization under the effect of appropriate light waves.

Preferably, such a photosensitive compound is a compound subjected to a cis/trans- or trans/cis-isomerisation under the effect of appropriate polarization light waves, in particular a photochromic compound comprising an azobenzene group.

The photosensitive compound is preferably linked by a covalent bond with the material of the layer in which the stable structure in relief is printed.

Preferably, the covalent bond occurs by means of a flexibilizing property chain and, more particularly, the flexibilizing property chain comprises at least one urethane group.

The particularly suitable material for the layer wherein the structure is printed comprises a polysiloxane matrix.

The polysiloxane matrix is preferably obtained through a sol/gel route from a hydrolysate composition of silane precursors.

The silane precursors are well known in the state of the art and comprise one or more hydrolysable functions such as Si—O-Alkyl, Si—Cl.

One suitable class comprises alkoxysilanes, in particular epoxyalkoxysilanes such as y-glycidoxypropyl trimethoxysilane, epoxycyclohexyl trimethoxysilane, as well as the following alkoxysilanes:

3-aminopropyltriethoxysilane, 3-aminopropyltrimethoxysilane, 3-aminopropylmethyldiethoxysilane, methyltrimethoxysilane, ethyltriethoxysilane, vinyltriethoxysilane, phenyltriethoxysilane, dimethyldiethoxysilane, tetraethoxysilane.

The most suitable alkoxysilanes are tetraalkoxysilanes, and more particularly tetraethoxysilane (TEOS). Such alkoxysilanes play the part of tetrafunctional crosslinking agents and allow for the increase of the structure stiffness.

The photochromic molecules of the azobenzene family being little soluble in matrices of the sol/gel type, a molecule is preferably used with its azobenzene group which has been grafted onto an organic chain ending with a silicon alkoxide.

The polymerisable composition therefore preferably comprises one hydrolysate of at least one precursor represented by formula (I):

A-Z-B

wherein

A represents an azobenzene group or a substituted derivate thereof,

Z is a divalent hydrocarbon chain interrupted by at least one urethane group,

B is a —Si R_(3−n)(X)_n group, n varies from 2 to 3, wherein X represents a hydrolysable group such as

R and R′ identical or different, represent an alkyl group having from 1 to 4 carbon atoms, and

a hydrolysate of at least one precursor represented by formula R_(4−a)Si(X)_a wherein a varies from 2to 4 and X has the same meaning as hereinabove. Preferably, R′ represents C₂H₅ and a=4 and n=3.

The preferred azobenzene precursor is represented by formula:

and will be referred to hereinafter as Si-DR1.

In order to inhibit the formation of aggregates and to give some freedom of movement to the photochromic compound, the composition may also comprise one hydrolysate of a precursor represented by formula (II):

A′-Z′-B′

wherein:

A′ is a cyclic hydrocarbon group optionally comprising an intracyclic heteroatom,

Z′ is a divalent hydrocarbon chain comprising at least one urethane group,

B′ is a group represented by formula —Si R_(3−n)X_n, wherein X, R and n have the same meaning as in formula (I) hereinabove.

The preferred precursor preventing the formation of aggregates is a trialkoxysilane comprising a carbazole unit:

It will be hereinafter referred to as Si—K.

The hydrolysis method of the hydrolysable silanes is conventional and known to the man skilled in the art.

See, for example the following patents: FR 2,702,486, FR 2,704,581 or U.S. Pat. No. 4,211,823.

One could, for example, mix the starting silanes and perform the hydrolysis of the mixture.

Catalysts such as mineral or organic acids may be used.

It is additionally preferred to use, for the hydrolysis, an at least stoechiometric amount of water, i.e. a water molar amount corresponding to the number of function moles able to create silanol groups.

Organic cosolvents such as THF (tetrahydrofurane) may be used.

• • According to a suitable mode,
  • A. a complete hydrolysis is performed of a mixture of silicon alkoxides, previously put in solution in an organic solvent of the alkoxides, by an acid aqueous solution having a pH equal to 1;
  • B. in order to better control the sol evolution during the condensation step, and before shaping the samples, the sol acidity is decreased partially neutralising the acid required for the hydrolysis through the addition of a base;
  • C. the sol is then applied onto a substrate using a centrifugation technique.
  • The film can then be dried at room temperature or be subjected to a thermal processing, for example, through heating at a temperature ranging from 50 to 120° C., for 5 to 30 minutes.

At the end of the coating, a low thickness film is obtained ready to be used.

Printing a stable structure occurs by positioning, through an appropriate device, the aperture of a fibre tip obtained according to the previously described method, in the close vicinity of the layer surface comprising the photosensitive material, at a distance lower than 100 nm and preferably at a distance lower than 50 nm, and more preferably lower than 10 nm.

A light beam is then injected in the fibre tip and irradiation occurs for a time required for obtaining the local topographic modification.

The local topographic modification directly results from the irradiation step and does not require any additional development step, as can be seen from some prior art documents.

Typically, the irradiation general conditions are:

outlet power, at the level of the aperture at the tip end: 0.1 at 1 nW (nanowatt), the irradiation time varying from 1 to 30 seconds.

One of the advantages of the method according to the invention is that it makes it possible to achieve an accurate control of the height of the written plots as a function of the irradiation time and/or the power of the emitted light waves on the material layer comprising the photosensitive compound.

The invention also relates to a method for printing a stable pattern represented by plots in relief in a layer of material comprising a photosensitive compound able to induce a topographic modification of said layer under the effect of light waves emitted through at least one aperture such as previously defined, which is characterized in that:

a) a first plot is printed according to the hereinabove described method,

b) a controlled lateral shift of said aperture is performed,

c) another plot is printed operating as in step a) hereinabove,

d) steps b) and c) are repeated so as to cover with plots a surface corresponding to said pattern.

Generally, the resulting plots have a height of at least 3 nm, preferably from 10 to 50 nm, and a lateral dimension substantially identical to the aperture size and two consecutive plots are spaced apart from less than 200 nm, preferably less than 100 nm, and more preferably less than 50 nm.

In order to more rapidly cover large surfaces, one may advantageously use a mask, comprising an opaque surface, generally parallel to the surface to be printed, and perforated with one or more apertures having the previously described features.

The invention provides numerous advantages.

In particular, in addition to those already mentioned, the invention allows to obtain plot arrays whose altitude variation, in the writing area, remains lower than their mean height, which could not be obtained in the prior art arrays.

The invention is applicable in the most various fields such as storing information on disks, fines markings, biotechnology, molecular electronics.

A particularly interesting application lies in the field of ophthalmic lenses.

It is indeed possible to achieve microstructure arrays, with a pitch lower than 200 nm, for producing anti-reflecting coatings.

Such a microstructure array could be printed directly on the surface of the ophthalmic glass, or on the surface of a mould part and transferred during the glass moulding.

Another possible application is printing distinctive marks such as a logo, invisible to the naked eye, and allowing to identify the glass origin.

Another possible application comprises producing discrete markings for tracking stored glasses or during their manufacturing process.

The following examples illustrate the present invention.

EXAMPLE 1

1) Manufacturing an Optical Fibre According to the Invention

An optical fibre is manufactured comprising an end portion having the general shape of a tip whose end have the form of a truncated cone from a monomode optical fibre sold by 3M and having the following features:

• • fibre diameter, including its protective plastic cover, 250 microns,
  • bare fibre diameter: 125 microns,
  • core diameter: 3.7 microns,
  • profile of the optical properties between core and sheath: index skip type,
  • cutoff wavelength: 460 nm.

Such a fibre is subjected to a heat sequential stretching according to the previously described general method using the following parameters:

• • stage I
    • stretching threshold Δ1₁=3 microns
    • CO₂ laser power: 14 watts
    • diameter of the laser beam: approximately 2 mm
    • value of the reference pulse duration τ_ref: 50 ms
    • valeur of the increment Δτ₁=1 ms
  • stage II
    • Δ1_C=3 microns
    • Cr=1.3 ms
    • α=0.45
  • stage III
    • D3=15 microns
  • stage IV
    • the stretching speed threshold is Ves=0.002 m.s⁻¹

The sequential stretching device that has been used is more particularly described in the hereinabove mentioned thesis by Philippe Bertrand dated 22 Jan. 1999.

In short, the optical fibre is maintained through pinching between two rails, horizontally sliding on ball bearing.

The rail ends are connected to one another by a wire being guided by pulleys and to which is hanging a free weight, or any means allowing to ensure a fixed tension.

The device comprises a laser having its supply adjusted by a TTL control as well as a device allowing for the fibre elongation to be measured.

A microscope equipped with a camera allows for the fibre to be displayed during the stretching.

The features of the resulting fibre preformed tip after heat sequential stretching are as follows:

diameter of the aperture: 477 nm

cone angle: 46°

The tip is subsequently cleaned 3 minutes in a sulfochromic mixture placed in an ultrasonic vessel, then rinsed three minutes in distilled water (still in an ultrasonic vessel).

It is then subjected to a chemical etching with 12% mass HF hydrofluoric acid for a period of time of 2 min. 30 seconds. Then it is rinsed in distilled water and dried with isopropanol.

A metallization step is then performed comprising applying chromium and aluminium layers through evaporation under secondary vacuum from metallic filaments heated by Joule effect on the periphery of the fibre tip.

The first layer is a primer layer made of chromium of 10 nm thick coated on the whole tip circumference.

The chromium evaporates at a 0.23 nm/s speed.

Still through evaporation, a 100 nm thick aluminium layer is coated, at an evaporation speed of 1.82 nm/s.

The tip is kept on rotation whilst the layers are being coated:

• • azimuthal rotation speed: 1 rps
  • axial rotation speed: 5 rps

The ready to be used optical fibre is then obtained.

The size of the aperture of the fibre tip is approximately 55 nm.

2) Manufacturing Photosensitive Samples

The sample is a thin layer coated on a glass substrate.

The layer is obtained through sol/gel route.

2.1—Sol Preparation

The starting reagents are as follows

• • 0.207 g of Si-DRI, i.e. 3.68×10⁻⁴ mol;
  • 0.307 g of Si—K, i.e. 7.37×10⁻⁴ mol;
  • 1.2 cm³ of THF, or tetrahydrofurane, an organic solvent of the alkoxides being used;
  • 0.041 cm³ of TEOS, i.e. 1.84×10⁻⁴ mol;
  • 0.093 cm³, i.e. 5.15×10⁻³ mol, of an aqueous solution having a pH=1, adjusted through addition of hydrochloric acid.

The Si-DRI compound and then the Si—K compound, both under the form of a powder, are weighed in a pill machine.

Then, the solvent (THF) is added in the open air, and the solution is homogenized through magnetic stirring.

Then, the TEOS as well as the acidified water are added, and the mixture is put again under magnetic stirring away from light for two hours.

At the hydrolysis completion, the medium is neutralised adding 0.0987 cm³ of pyridine i.e. 1.22×10⁻³ mol.

In order to reduce the viscosity of the sol being formed, 0.5 cm³ of solvent (THF) are incorporated.

The sol is then filtered (diameter of the filter pores: 0.45 μm), then a 2.5‰ mass surfactant is added.

The sol is then ready to be applied (concentration of mother solution referred to as C₀).

2.2—Coating of the Layers

The coatings occur through centrifugation on glass substrates (preliminarily cleaned microscope plates), for a time of 20 s.

During the rotation movement, the condensation process progresses, the solvent evaporates, and there finally remains a thin layer having its thickness which depends on the rotation speed and the concentration of the sol being used. The thickness typically varies between 10 nm and about 1 μm ; it is determined using a profilometer.

Adjusting the Layer Thickness

A coating of the mother solution with a rotation speed of 3000 rpm allows to obtain a sample approximately 1 μm thick.

In order to obtain layers with lower thicknesses, the mother solution should be diluted and the centrifugation speed should be increased.

The hereinafter table gives the correspondence between the dilution of the mother solution and the thickness of the resulting sample for a rotation speed fixed at 6000 rpm

			Thickness
			(application 20 s,
Composition	Obtention	Concentration	6000 rpm)
A	Mother solution	C0	850 nm
B	0.3 cm3 A	CC0	323 nm
	0.3 cm3 THF	2
C	0.,3 cm3 B	CC0	137 nm
	0.3 cm3 THF	4
D	0.3 cm3 C	CC0	45 nm
	0.3 cm3 THF	8
E	0.3 cm3 D	CC0	22 nm
	0.3 cm3 THF	16
F	0.3 cm3 E	CC0	12 nm
	0.3 cm3 THF	32
G	0.3 cm3 B	CC0	5 nm
	0.3 cm3 THF	64

The thickness measurements are made after placing the freshly coated sample ten minutes in an oven at 100° C. so as to complete the condensation process and to evaporate the residual solvent.

The sample corresponding to E is used for the remainder of the implementation of the method of the invention.

3)—Writing Stable Structures in Relief

Device to be Used

The device to be used is represented in FIG. 7.

An argon-krypton (1) laser is used with a wavelength of 568.2 nm (yellow ray of the krypton) generating a laser beam (3).

At the laser (1) outlet a 2 mm thick optical density (2) or neutral filter G4 is inserted. The laser beam (3) goes through the optical density (2) and is then reflected by a mirror (4), goes through a quarter wave plate (5) which converts the vertical rectilinear polarisation originating from the laser (1) into a circular polarisation, in turn sent to a polarizing filter (6). The filter (6) is typically a model HN22 Polaroïd polarizer.

At the polarizer (6) outlet, the beam impiges on a mirror (7) directing it towards a focus lens (8) (f′=10 mm).

The beam will then be injected into an optical fibre (10), prepared during the above described step 1 and having an end (11) with a thinned, circumferentially metallized tip.

At the lens outlet, the beam goes through a connector (9) corresponding to the non modified end of the optical fibre (10).

The optical fibre (10) goes through a microscope head (12) containing two (non shown) piezoelectric tubes allowing to ensure the shift, in the three space directions, of the end (11) of the tip of the thinned, circumferentially metallized optical fibre (10).

Such a piezoelectric system is disclosed in further detail in the above-mentioned thesis by Philippe Bertrand.

The light beam originating from the (non shown) aperture, at the end (11) of the tip of the fibre then goes through the sample (13) prepared during the above described step 2).

Then, the light beam is collected by a microscope objective (14) typically a Zeiss objective, <<Epiplan>> LD20×/0.40 and detected by a photomultiplier (16), typically a Hamamatsu R943 Ga—As 10 stages, equipped with a polarisation analyzer (15).

Writing an Array

The sample is placed as illustrated in FIG. 7, 5 nm away from the tip of the fibre (11).

A first scanning is preliminarily performed in order to check the cleanliness of the surface of the layer where the array will be written.

To this end, the aperture at the end of the tip is kept 5 nm away and a scanning is performed in a plane.

A 2 μm² surface designed for being printed is then controlled.

Subsequently, acting on the piezoelectric system, the tip is positioned above a determined location of the scanned surface (initial position).

Then the tip is moved in two perpendicular directions, x, y, in a plane parallel to the layer surface, and keeping the aperture 5 nm away from the layer surface.

A first plot line is printed by injecting, for two seconds for each plot, a light beam giving a tip outlet power at the aperture level of 170 pW.

Between each light pulse, the tip is moved along the axis x by a 116 nm distance.

After printing the first line, the tip is again arranged in its initial position and is moved along the 136 nm axis y, followed by printing a new plot line.

The operation is repeated until the array is obtained on the whole surface to be printed.

On the final array, the written plots are on average 5 nm high and 55 nm wide at half height.

The written array is illustrated on FIG. 8.

80 plots are written in a 1 μm² area, without servo light power and it can be observed that the altitude fluctuation is 4 nm for a 5 nm average height of the plots.

The central plot has been deliberately omitted so as to show that writing all the plots does not hinder the non enlightened central area.

EXAMPLE 2

A plot array is printed according to the method as described in example 1, with a 200 nm shift of the tip, along the axis x, between each printing operation, and increasing the radiation time (from 5 to 35 seconds for each 5 second pitch) for a fixed average power at the tip outlet.

FIG. 9 illustrates a sectional view of the resulting array, along one of the lines of the plot array.

It can be observed that the plot height is directly proportional to the irradiation time.

Claims

1-40. (canceled)

41. A method for producing an optical fiber useful as a near field probe, comprising:

(a) subjecting the optical fiber to heat sequential stretching until it breaks to obtain a fiber comprising an end portion, the end portion comprising a tip with an end having the form of a truncated cone bound by a globally planar side and a circumferential surface; and

(b) subjecting the tip of the fiber to isotropic chemical etching, wherein the circumferential surface of the truncated cone is made opaque, such that the globally planar side forms an aperture with a maximum dimension of 100 nm.

42. The method of claim 41, wherein the isotropic chemical etching of the top of the fiber is performed with an acid.

43. The method of claim 42, wherein the acid is hydrofluoric acid HF.

44. The method of claim 41, wherein the circumferential surface of the truncated cone is made opaque by coating it with at least one metallic layer.

45. The method of claim 44, wherein the coating of the metallic layer occurs under vacuum, in the presence of a metallization source.

46. The method of claim 45, wherein, during the coating, the optical fiber is oriented towards the metallization source so as to avoid metal from being coated on the globally planar side of the truncated cone.

47. The method of claim 46, wherein, during coating, the end portion of the optical fiber is subjected to an axial rotation movement and to a rotation movement around an axis carried by the average direction of the evaporation flow.

48. An optical fiber comprising an end portion having the general shape of a tip with the end having the shape of a truncated cone bound by a circumferential surface and a globally planar side defining an aperture allowing light waves to pass, the end portion capable of maintaining polarization properties of propagating light waves, wherein the globally planar side has a maximum dimension of 100 nm and the circumferential surface of the truncated cone is opaque.

49. The optical fiber of claim 48, wherein the truncated cone has an apex angle ranging from 30 to 70°.

50. The optical fiber of claim 49, wherein the truncated cone has an apex angle ranging from 40 to 70°.