GB2564371A - Tomography by dissolution analysis - Google Patents

Abstract

A new approach to achieving chemical mapping on a nanoscale is claimed that can provide 2D and tomographic images of surface and near-surface structure. A grid/scale or graduation is mapped to the surface of the sample and the surface within the grid imaged to generate a first image. Material is then dissolved from the surface of the sample by applying a series of aliquots of solvent then analyzing their contents after removing them while in between exposures to the solvent, the surface is imaged with a form of microscopy to provide at least a second image which is compared to the first. These images can provide a map of the topography of the surface of a sample and by combining the analysis of the remove surface with the imaging data a 2D map of surface composition can be obtained. Furthermore composition as a function of depth may be determined thus enabling tomography.

Description

Tomography by Dissolution Analysis

The present invention relates to a method of chemical imaging by dissolution analysis, producing a representation of the structures of one or more objects or samples by dissolution and/or erosion of at least part of the object or sample with solvent or entrainer. The present invention provides a new approach to chemical mapping down to the nanoscale that can provide 2D and tomographic images of surface and near-surface structure. In particular the observation and recording of the changes and/or relative differences induced by the application of solvent and/or entrainer to the sample enables a representation of the structural composition of the sample to be obtained.

Chemical mapping is routinely used to study the structure and functionality of a wide range of materials. Generally, the desirable capabilities for a chemical imaging method include; the ability to achieve high spatial resolution (preferably down to nanometres), the ability to contend with thick samples and samples with rough surfaces with minimal sample preparation, the ability to use more than one analytical technique to probe the composition of a selected region of the sample (in particular, it is useful to determine whether it contains a mixture of compounds or only one type of molecule), the ability to provide detailed chemical information as a function of depth (preferably tomography).

Although the above list is not exhaustive, as there can be other considerations, these are the principal capabilities most analytical scientists find useful for addressing a wide variety of samples. There is a range of established conventional techniques that can be used for chemical mapping, some of which are commercially available. The ones that operate on a nanoscale fall into two categories: thermomechanical photothermal Nano-IR techniques and tip-enhanced scattering techniques; both of these are non-destructive, an advantage in some cases, but there are fundamental obstacles to either enabling tomography. The former, when used with top-down illumination shows promise (illuminating the sample from below precludes the possibility of looking at thick samples and obtaining depth information). However, current implementations still require that samples are thin and interpreting depth information is not yet possible (we return to this point below when considering tomographic imaging). Tip-enhanced scattering techniques can achieve the highest lateral spatial resolution, circa 20 nm, because the measured signal is dominated by the enhanced field around the tip. By the same token, they probe only the top 20 nm of the sample and cannot, therefore, provide depth information.

It is therefore an aim of the present invention to provide a method of mapping and/or analysis which addresses the abovementioned problems.

It is a yet further aim of the present invention to provide a method of assembling apparatus, the use of which addresses the abovementioned problems.

In a first aspect of the invention there is provided a method of chemically mapping the structure of a sample, said method including the steps of; - Placing or mapping a grid, graduations or scale onto at least part of an image of the surface of the sample; - imaging or mapping at least part of the surface structure of the sample substantially within the grid, graduations or scale using at least one microscopic technique; - dissolving or at least partially removing at least a first material from a surface of the sample by introducing or exposing said surface to a solvent or entrainer; - analyzing the solute, solution and/or contents of the solvents; subsequently imaging or mapping the surface structure of the sample substantially within the grid using at least one microscopic technique; and - comparing the images of the sample

As such, the surface of a sample is imaged in between exposures to a solvent or entrainer and the dissolution or erosion of surface features is mapped. The mapping is typically performed by comparing grid cells or graduation points. These cells can be of arbitrary size including being the size of a pixel.

Preferably the area of the sample that is imaged has a grid, graduations or scale mapped onto at least part of the images of the surface of the sample.

Preferably the maps or grids are aligned by selecting at least one point or cell on the surface and aligning each image accordingly.

In one embodiment the solvent is provided as a series of aliquots.

For high spatial resolution, typically the microscopic technique is atomic force microscopy (AFM). As such, using a depth sensitive mapping or imaging technique a 3 dimensional images or pictures of sample erosion due to the application of solvent can be constructed resulting in chemical tomography.

In one embodiment at least one linescan is performed. Typically at least two substantially parallel linescans are performed, thus providing, after suitable analysis, tomographic slices. Further typically, the linescans are shifted or separated substantially equidistantly, thus ‘body-scanner type’ tomography can be performed.

Typically aliquots of the solute, solvent and/or solution are used for analysis.

J

In one embodiment samples or aliquots are taken from the flow of the solute, solution and/or solvent.

The person skilled in the art will appreciate that a wide range of analytical techniques can be used. In one embodiment combinations of techniques can be employed to analyse the contents of a single solvent aliquot, for example both IR and Raman spectroscopies.

Further analytical techniques include HPLC-MS (including tandem MS). This provides a step-change improvement in analytical specificity compared to the vibrational spectroscopies available for nanoscale imaging such as photothermal IR spectroscopy and tip-enhanced scattering methods.

Thus, with chromatography (the ability to separate mixtures) as a function of depth the present invention provides; the ability to achieve high spatial resolution (down to nanometres); the ability to contend with thick samples and samples with rough surfaces with minimal sample preparation; the ability to use more than one analytical technique to probe the composition of a selected region of the sample; and the ability to provide detailed chemical information as a function of depth (including tomography).

In a second aspect of the invention there is provided a method of producing a representation of the structure of one or more samples, said method including the steps of; - Dividing or designating at least part of the images of the sample into two or more parts, sections or cells; - Exposing said parts, sections or cells to at least one solvent or entrainer for a period of time; - Measuring and/or calculating the height reduction and/or volume reduction of material in said part, section or cell;

Calculating and/or plotting the rate of height and/or volume reduction for said section or cell; and

Comparing the rate of height and/or volume reduction for said sections or cells.

As such, a representation of the structure of the sample can be constructed wherein cells with substantially the same rate of height reduction comprise substantially the same material.

In one embodiment a series of aliquots of solvent or entrainer are applied to the sample.

In one embodiment the solvent or entrainer is supplied as a flow.

In one embodiment the solvent aliquots are pulses of solvent supplied to the sample in a flow of substantially inert (i.e. non-dissolving or non-entraining fluid) liquid typically alternating between a solvent and/or entrainer and the substantially inert fluid.

In one embodiment solvent or entrainer is supplied via or through a tube and/or conduit.

In one embodiment a distal end of the tube and/or conduit is placed on and/or substantially at the surface of the sample.

In one embodiment a droplet of solvent and/or solvent aliquots is delivered on to the surface of the sample.

Typically solvent aliquots are delivered in a localised way by placing a tube on the surface of the sample and delivering a droplet of solvent on to the surface of the sample.

In one embodiment at least part of the solvent, entrainer and/or solvent aliquot is removed or collected from the surface by a tube.

The collecting tube can be the same as that used to dispense the solvent or aliquot or can be a second or further tube.

Typically at least part of the solvent, entrainer and/or aliquot droplet is removed from the surface via a tube by placing a distal end of a tube on, adjacent to and/or otherwise in the vicinity of a droplet.

Further typically the solvent, entrainer and/or aliquot droplet, or at least part thereof, is removed by pumping the same via a tube.

In one embodiment at least part of a first droplet is removed from a first location on the sample to a second location not on the sample.

Typically a pump and a tube is used to pump the droplet fluid into a tube, said fluid staying in the tube until a second location is placed under the tube and/or the tube is moved to a second location. Further typically the fluid in the tube is ejected or otherwise dispensed onto a second location.

In one embodiment the material or fluid ejected onto the sample at the second location not on the sample is analysed. The person skilled in the art will appreciate that the droplet or fluid at the second location could be collected and deposited at third or further sample locations.

In one embodiment the tube is a probe in a scanning probe microscope.

In one embodiment the tube is moved in the x, y and z axes by one or more actuators. Typically said one or more actuators is under computer control.

In one embodiment the tube is used to deliver solvent to a selected area of the sample then the sample is moved into a microscope so that the same area can be imaged. Typically the same tube is used to remove the solvent before imaging occurs.

In one embodiment the sample is moved from the microscope then a tube is used to deliver solvent to a selected area of the sample then the sample is moved back into the microscope so that the same area can be imaged. Typically the same tube is used to remove the solvent before imaging occurs.

In one embodiment the sample remains in the microscope and a tube is moved to be above a selected area on the sample, the tube delivers a droplet of solvent and then the tube is removed and the microscope images the same area of the sample. Typically the same tube is used to remove the solvent before imaging occurs.

In one embodiment the height reduction is measured and/or calculated using one or more imaging techniques.

In one embodiment proximal probe imaging and/or scanning probe is used to measure and/or calculate the height reduction. Typically microscopy is used to measure and/or calculate the height reduction. Further typically, atomic force microscopy (AFM) is used to measure and/or calculate the height reduction.

Typically the rate of height reduction is calculated and/or plotted as a function of time.

In one embodiment the sample is exposed to a substantially continuous flow of solvent and/or entrainer.

In one embodiment the one or more sample includes at least one reference point or reference material. Typically the reference point, or reference material comprising the reference point, is of substantially uniform composition. Further typically the reference material has a uniform composition at least to the depth of interest.

In one embodiment the sample includes a plurality of reference points and/or materials and these.

In one embodiment the reference points and/or materials can be different at different points in the analysis.

Typically the dissolution and/or solvation properties of the reference point or material are known. Typically a solvent is chosen that does not dissolve the reference material. Where the reference material is dissolved, typically from the time of exposure to the solvent or entrainer the height reduction of the reference material or point is calculated.

In one embodiment the solvent is analysed. Typically the solution or entrainer is analysed to quantitatively determine the amount of solute and/or material present in the same. Further typically the quantitve analysis data can be used to supplement and/or in conjunction with the height reduction data.

Typically the solvent or solution analysis if performed by any one or any combination of chromatographic techniques and/or spectroscopic techniques. Further typically high performance liquid chromatography (HPLC) is used to analyse the solutes in the solution / solvent.

Typically the reference points comprise substantially the same material.

In one embodiment at least part of the sample or one or more images thereof is divided into a grid of cells. Typically the cells of substantially uniform size and distribution. Typically any change in height of the cell can be multiplied by the area of the cell to calculate the change in volume.

In one embodiment the differences in height between the reference points and each cell in the grid is calculated. Typically the calculated change in height of the reference points is subtracted from the differences in height between the reference point and each cell in the grid. Further typically this calculation provides:

Height differences relative to the reference height plus calculated height change in the reference = Total height change /nm

Thus, using the above calculation a plot of; total height change and/or volume change as a function of time and/or solvent or entrainer exposure can be obtained.

Typically from the height reduction of a cell, the volume of material that is removed can be calculated.

Typically the cells that have the same rate of height and/or volume loss are categorised as being substantially the same material.

Typically the cells that are deemed the same material are colour coded in an image that shows the distribution of that material over the surface.

Typically the depth of material is determined by, or measured, when the rate of dissolution changes.

Typically the total volume lost of a material is plotted as a function of time.

In one embodiment the cumulative shapes of the plots or the differential of the same are correlated with the appearance or detection of materials in the solvent or entrainer.

In one embodiment the mass of a material lost from the surface is calculated from the volume lost and the density or an estimate of the density of that material.

In one embodiment the mass of material lost from the surface is correlated with the mass of material present in the solvent or entrainer as determined by analysing the solvent using at least one analytical method.

J

In one embodiment the mass lost from the surface as determined by analysing the solvent is used together with the density or an estimate of the density to calculate the height loss or volume loss of a material. Typically by correlating the calculated height loss or volume loss with the measured height or volume loss materials are identified.

In one embodiment at least part of the sample and/or one or more images of the same is divided up into a three dimensional grid or matrix. Typically each cell or grid in the matrix is a voxel.

In one embodiment substantially each voxel in the grid is allocated to one of the materials making up the sample. Typically this is done by observing the change in height difference or volume difference at each step (equivalent to the slope of the plot Total height change v Time). Further typically this provides the composition as a function of depth for each cell.

Typically a 3D map is constructed using this method and from this map tomographic images can be obtained.

In a third aspect of the invention there is provided a method of chemically mapping the structure of a sample, said method including the steps of; imaging or mapping at least part of the surface structure of the sample using at least one microscopic technique; dissolving or at least partially removing at least a first material from a surface of the sample by introducing or exposing said surface to a solvent; analyzing the solute, solution and/or contents of the solvents; subsequently imaging or mapping the surface structure of the same area of the sample using at least one microscopic technique; and comparing the images of the sample

Preferably the area of the sample that is imaged has a grid, graduations or scale mapped onto at least part of the image of the surface of the sample.

Specific embodiments of the invention are now described with reference to the following figures wherein;

Figure 1 shows images of a selected region showing how material is removed from the surface in a series of steps;

Figure 2 shows spectra for the material captured from the blend and plots of % PPMA and cumulative weight loss as a function of time;

The present invention provides a new paradigm for surface characterisation and 3D-chemical mapping down to the nanoscale. It comprises; exposing the surface of a sample to a succession of aliquots of a liquid so that materials are gradually removed from the surface. Using atomic force microscopy or other suitable form of microscopy with suitable image processing software, changes in the topography of the surface are measured. A decrease in height is a measure of the loss of material from the surface of the sample to the liquid, this translates to a measure of the kinetics of the surface dissolution/erosion process, i.e. where we see rapid loss, there is a high rate of dissolution/reaction/erosion, gradual loss equates to slow kinetics (integrating decrease in height over selected areas gives a measure of volume loss). The species taken up by the solvent are analysed by any appropriate analytical method, from this we have a second measure of the kinetics of the dissolution/reaction/erosion process. The AFM data provide the location associated with kinetics, the analytical data provide chemical composition associated with kinetics. By correlating these two sets of data, i.e. rate of volume loss from the surface with quantity measured in each aliquot of the solvent, we can achieve a map of the location of the identified materials in 3D.

We have demonstrated this method using a polymer blend comprising 30% w/w polystyrene (PS) and 70% poly (methylmethacrylate), (PMMA) with analytical information provided by Fourier Transform Infra Red (FTIR).

Sample Dissolution and Chemical Analysis

Figure 1 shows the image of the sample at the start of the experiment and after subsequent exposure to the solvent. The images are of a selected region showing how material is removed from the surface in a series of steps; the scale is 5pm x 5pm. The arrows indicate an area where there is a sudden removal of material resulting in a pit or trough compared to the previous image. At the start, a ‘raspberry-ripple’ type structure is seen with elongated raised features within a flat matrix. After exposure to the solvent, inspection reveals that raised areas are disappearing leaving pits, or trenches, in their place. This is occurring at different rates in different places. The picture is consistent with the progressive removal of a material that is soluble in cyclopentane, leaving behind material that is not soluble.

The spectra for the material captured from the blend is shown in figure 2, which shows at the left; from top to bottom, spectra of pure PS and PMMA then the spectra from the material extracted from the solvent placed on the sample surface. On the right, from top to bottom, °/o of PMMA in the extract (inset is the calibration graph) then the gravimetric data for the blend, squares, and the pure PMMA, diamonds. This data suggests that it is dominated by PS and the quantification confirms that the amount of PMMA remains at about 15%, until after 15 seconds seconds when a relative increase seems to occur due to near-exhaustion of the PS; although the amount removed at this point is small thus the error in this measurement is high, the gravimetric data demonstrate that almost all of the soluble material has been removed by 15 seconds and it is the data before this that are most relevant to the interpretation we offer. The amount of PMMA found in the solvent captured from the blends is greater than the gravimetric curves suggest which implies that some PMMA is entrained with the PS as it leaves the surface.

This result underlines the importance of analysing the solvent; it is not sufficient to assume that only the soluble material is removed and the insoluble component remains. It is easy to envisage how PMMA could be occluded within the PS and, therefore, be removed by the solvent. Nevertheless, in this case, the over-whelming majority of the removed material is PS.

The gravimetric data for PMMA dissolution does exhibit significant scatter and has to be considered as semi-quantitative, however, the slope of the curve provides an estimate of the rate of loss of the PMMA that is adequate to illustrate our method. The slope implies a rate of loss of 1.4 g/s as shown in figure 2.

The measured diameter of the solvent drop was 12.5 mm with a SD of + 0.96 over 9 measurements. With a density of 1.19 g/cm3, this leads to an estimate of the rate of removal of PMMA of 2.4 nm/s. The solvent applied to the pure PS film caused a hole to appear with the first application that then increased in diameter. This behaviour was inconsistent with that of the other samples which remained substantially intact up to 15 seconds. Given that the behaviour of the PS was radically different from that of the PMMA and the blend, the results for the captured material are not presented for comparison. Nevertheless, it is clear that the solubility of PS is orders of magnitude greater than that of PMMA.

Imaging during the Process of Surface Dissolution

We know that there are two materials in the sample shown in figure 1; the images suggests two phases (thus little or no mixing at a molecular level), one of which has formed the raised features. From this image, it appears there are roughly equal amounts of both. This is, perhaps, surprising because the sample comprises only 30% w/w PS. Consequently, we start with a significant degree of uncertainty regarding the sample’s morphology.

As mentioned above, exposure to the solvent changes the sample’s structure and holes, or pits, begin to appear, replacing the raised structures as a consequence of sequential exposures to the solvent. Close inspection of these micrographs reveals that there is significant drift, i.e. the same features appear in different positions at different times. This is to be expected, the sample and stage are subject to forces and temperature changes during the process of exposing the surface to the aliquots of solvent; at high magnification, this inevitably leads to drift between images and even during imaging.

Processing to account for drift in the topographic micrographs For the sake of clarity we will illustrate correcting for drift using only the first and last images, as shown in figure 3 where the top (small images) are topographic images of the selected area after exposure to the solvent for, from left to right, 0, 1, 2, 4, 6, 8, 10, 12 and 15 seconds respectively. It is noted that features drift relative to the frame of the field of view in successive images. The bottom (large images) left, shows topography of the untreated sample and, right, the topography after 15 seconds exposure to the solvent. The centres of the white circles indicate three points that are chosen as being the same position in the two images. The white triangle joins these points for the first image and this shape is located in the same place relative to the edges in the second image; the drift can be clearly seen. All images are 5pm x 5pm.

In this example three points to are selected on each image that are judged to be in the same positions in both images. These are then translated in x and y coordinates so that, as figure 4 shows, the images are substantially coincident.

In figure 4 on the left, the untreated surface is shown. In the middle, the same area after 15 seconds exposure. It is noted that the drift seen in figure 3 has been substantially eliminated by the software as demonstrated by the white triangles (see figure 3 for comparison). These images have been divided into a small number of cells to maintain clarity. Far right, the image is divided into a grid of 50 x 50. The filled-in cells at the centre of the white circles, denote those that were selected as being the same material that will dissolve in the same way. These positions were then used to align the images in the z axis.

The next step is to align the images in the z-axis. The images are devided into a grid of cells (in this case 50 x 50), as shown in figure 4. This is important because operating at the level of a single pixel or point would produce a great deal of noise and too many kinetic plots to process in a reasonable time (see below).

When this is done, cells are chosen that are deemed to be the same material (in this example three) and behave in the same way during exposure to the solvent. Typically the phase chosen would be the least soluble, as in this case, where cells are all sited on the flat areas that change very little, but the method works using any cells provided the behaviour is substantially the same in each.

The images are then shifted so they are now aligned in the x, y and z-axes. Figure 5 shows the change in height for each cell where all plots are shifted to be zero at the start. Zero gradient corresponds to no change relative to the cells selected in figure 4 and a negative gradient represents a relative decrease in height.

In figure 5 the height vs time graph gives the plots of height change from untreated to 15 seconds exposure in each cell. The starting position for each cell has been set to zero. Underneath is a histogram of these height changes where a trimodal distribution can be seen. The dashed line in the bottom plot encompasses all of the positive values; in this case these give a measure of the error in the measurement as no cells increase in height. The other dashed line is a reflection of the red dashed line; together they estimate the error in determining those cells that do not, to a first approximation, change. All cells between a) and b) are coloured blue in the image shown (colour overlaid on topography image, the scale is 5pm x 5pm), cells between b) and c) are coloured green and those between c) and d) are red. This corresponds to blue = greatest change, green = some change and red = no change. Blue and green together map the cells where the raised features have dissolved, i.e. they must be almost entirely composed of PS. The green cells are where the height reduced to the level of the red cells. The blue cells are those where the dissolved phase created a hole or pit. The red cells are PMMA. The white dashed lines on the image are explained in respect of figure 7. A histogram of the gradients is also shown in figure 5. 2D imaging from local kinetic measurements

Continuing with the simplified approach that uses only the first and last images, i.e. the large micrographs shown in figure 3, when inspecting the histogram in Figure 5 we expect there to be no positive gradients if we have selected cells with the lowest solubility from Figure 4; the existence of positive gradients can be regarded as a measure of the noise in the data.

In figure 5, it is shown how symmetry allows us to estimate the distribution for the error in the ‘unchanged’ cells (more accurately, these are cells that have either not changed or have changed in the same way as the reference cells but, for convenience, we will refer to them as ‘unchanged’ cells).

The behaviour appears to be trimodal with one population being unchanged. We can divide the cells into three categories as shown in figure 5. We can see that the blue + green area, representing the dissolved PS phase, is mapped accurately. The cells where there has been a loss of material, down to the level of the unchanged cells, are coloured green. Where the loss of material caused a trench or pit to be formed is shown in dark blue. Although this is a simple example, it illustrates the principle that we can map where dissolution has occurred, analyse the resultant solution, and then assign an identity to each cell. To a first approximation, in this case, all dissolved material is PS (we will refine this observation later).

We have moved beyond what simple inspection affords, the assignment for each cell is based on a quantitative measurement of dissolution behaviour and we can quantify the volume of material dissolved within the field of view. We can also assign colours to the three different types of behaviour seen in the histogram (Figure 5). To a good approximation, the red cells do not dissolve (or more accurately do not dissolve more than the selected cells) and these can ascribed to PMMA, the green cells are those that dissolve to reveal the flat surface of the PMMA and the dark blue cells are those where the PS dips beneath the surface. Inspection of Figure 4 confirms that the mapping is substantially accurate. Our procedure enables subtle differences in dissolution behaviour to be mapped with high spatial resolution. To our knowledge, this is the first nanoscale image where contrast is achieved through differences in quantified localised kinetics. In this case we have focussed on dissolution kinetics but this capability could be ex-tended to any kind of surface reaction that leads to relative de-creases or increases in height.

We can now extend the alignment and analysis to include all of the images shown in figures 1 and 3. The kinetic plots from all the cells for all of the images are given in figure 6. At this point, we are considering 20,000 kinetic measurements.

Inspection of figure 1 indicates that there is little change, then the sudden appearance of a trough or pit; we can map the time at which this sudden change occurs. The coloured map in figure 6 is in good agreement with the behaviour implied by figures 1 and 3. This poses the question, “given that the dissolved material is almost entirely PS, why should there be this large range of behaviours such that the step-change can occur at 1 second to 12 seconds?” This is discussed below in the context of the tomographic images.

Figure 6 shows a top graph that is equivalent to the similar plot shown in figure 5 except all 8 images have been used. Cells that have changed significantly in height after 15 seconds exposure to the solvent are captured by the blue box made of dashed lines (this box is in practice is often drawn by the user using the curser), all of these cells are coloured blue. The red dashed box captures all cells that have changed after 12 seconds, these are coloured red (overwriting the blue); the cells that have changed after 15 seconds but not after 12 seconds remain blue. This process is repeated and, in this way, the cells that have changed significantly from 12 seconds to 15 seconds are identified, then those that have changed from 10 seconds to 12 seconds etc. are identified with corresponding colour coding. The scale of the image is 5pm x 5pm. The white dashed lines are explained in respect of figures 7 and 9.

Construction of Approximate Tomographic Images

To a first approximation, because the polystyrene (PS) is far more soluble in the solvent that the poly methyl methacrylate, (PMMA), comparison of the starting topography with that after the PS has been substantially removed, provides the basis for understanding the structure as a function of depth. In figure 7, two positions marked on the 2D compositional map are shown as a series of linescans for the successive exposures to solvent. The capability to show the same position in each image follows from the x, y and z axes alignment process. In the case of ‘A’, a step-change increase in depth is seen after 2 seconds exposure. In the case of CB’, it occurs after 12 seconds exposure; the reasons for the differences are discussed below. The three step process for constructing an approximate tomographic slide is illustrated in figure 7. The linescan for the untreated sample is coplotted with that for the maximum duration of exposure. The experimenter then decides on the boundaries between the two phases; this choice is guided by the topographic image of the untreated sample. Within the selected limits, the line from the final image is used together with that from the first image, but only the line from the first image is used outside these limits. This then leads to the creation of the first approximation of a tomographic slice.

In figure 7 at the top left, the compositional map taken from figure 5 is shown where red = PMMA and blue and green = PS. The top middle and right show linescans taken from positions A and B in the compositional map for successive exposures to solvent (the same as is shown in figure 6). For A, the step-change in height occurs at 2 sec., for B the change happens after 15 sec. The tomographic slice is created by a three step process shown in the bottom row; 1 - the linescans for the untreated surface and the same position after the step-change occurred are co-plotted, 2 - two positions delineating the edges of the raised feature are selected (the vertical lines in 2) and within these limits the two lines are preserved and outside only the untreated line is used, 3 — the phases are greyscale-coded.

In figure 8 more tomographic slices are shown taken from figure 5. It can be seen that multiple features can be imaged. When a sequence of such maps is obtained by shifting the linescans by equidistant steps, body-scanner type tomography can be performed. In this way we can acquire a detailed understanding of the 3D structure of the sample.

In figure 8 these are in a sequence that is spaced uniformly thus they are the same as the familiar body-scanner type parallel planar tomography.

This construction is based on the simplifying assumption that the PMMA does not dissolve/erode significantly and that there is no inaccessible, occluded PS buried beneath the surface and id thereby afforded protection from the solvent by a thick layer of PMMA. Whether these assumptions are tenable can be tested: the rate of dissolution of the less soluble phase can be measured (and this has been done in this case, see above). The total amount of PS in the blend is known (or, more generally, can be measured); inspection of the 2D and 3D images suggests that the simple construction accounts for the PS known to be in the sample, thus there is little occluded polystyrene.

Construction of tomographic images using kinetic data

Above, a methodology was described for 2D mapping of surface composition (with the option of some depth information) on the basis of measurements of the kinetics of dissolution. This same approach can be applied to mapping composition as a function of depth; this is shown in figure 9.

Histograms of the height change between successive exposures can be obtained. When these differences are small, the entire dissolved layer can be attributed to the least soluble component then greyscale coded accordingly. When there are combinations of small changes and larger changes, the histograms can be used to determine which positions correspond to the less soluble phase and which to the more soluble phase. Some judgement is required because of the noise in the data but, broadly speaking, the gradient (rate of dissolution) of the least soluble phase is known, so faster rates belong to a different category (and by extension of this approach, multiple phases with multiple rates of dissolution can be categorized with a possible additional variable being the type of solvent used for each exposure). This is illustrated by A and B in Figure 9.

In figure 9 the tomographic slices correspond to the lines A and B that feature in figures 6 and 7. The greyscale coding shows the time at which the step-change decrease in height occurs. The histograms represent the decrements between 2 and 4 seconds for A and between 12 and 15 seconds for B. The red lines represent the boundaries; below the greyscale coding is dark grey (PS), above it is light grey (PMMA). Elsewhere the histograms always indicate only a slow rate of dissolution. For A, the kinetics suggest a thin layer of PMMA over the PS while for B the layer is much thicker.

It can be seen that the PMMA layers on either side of the PS domain now have a depth because the decrease in height has been calculated to be 2.4 nm/s, see above. This can be contrasted with the representation in Figures 7 and 8 where it has been assumed that the PMMA is completely insoluble.

Figure 9 gives a more realistic picture of the reach of the experiment; it has not probed far beneath the surface of the PMMA but it has done so to some extent. Furthermore, figure 9A shows that the decrease in dissolution rate, after the rapid dissolution event, identifies a layer of PMMA surrounding the base of the PS structure. This is not seen in Figure 9B, thus it is probable that all of the PS was not dissolved in this case. These tomographic slices illustrate, in a striking way, the differences between the two positions; there is little removal of PS for up to 15 seconds at B even though PS is highly soluble in cyclopentane. At position A, rapid dissolution starts after 2 seconds. The tomographic images clearly suggest an explanation; there is a layer of PMMA over the surface of the entire sample and this layer has different thicknesses in different places. Inspection of the image in figure 6 adds credence to this theory because it is clear that adjacent regions on the same raised feature follow sequentially. The feature located in the top right of the image changes from green (step-change after 1 second) to yellow (step-change after 2 seconds) then to orange (step-change after 4 sec-onds). The feature located bottom right goes from dark red (step-change after 8 second) to mauve (step-change after 6 seconds). We do not see a green area (step-change after 1 second) located next to a dark blue area (step-change after 15 seconds). This is exactly the type of behaviour that would be expected when there is a PMMA layer over the PS because it is very unlikely there would be acute changes in thickness. This theory also goes some way toward explaining the higher-than-expected amount of PMMA entrained by the solvent. It is easy to envisage that a layer on top of a PS feature would be penetrated in one place before the rest, perhaps near the periphery of the PS domain. The subsequent egress of PS could remove associated PMMA.

We have described how 2D Chemical Mapping can be obtained from AFM images taken after a series of exposures to a solvent.

An image can be divided into a grid of cells and the height change or relative height change within each cell can be measured as a function of exposure. In this way the amount of material lost from the surface can be mapped and plotted as a function of time of exposure and local kinetics can be mapped on a nanoscale. By analysing the material dissolved in the solvent and correlating this with the material lost from the surface, 2D nanoscale compositional maps were created. These same principles can be applied as a function of depth. The nature of the material in each cell can be deduced from the rate at which its height decreases.

In a simple analysis, the least soluble phase is assumed not to dissolve and the user identifies boundaries between phases on the basis of an in-spection of the topographic images. Comparison of linescans before exposure to solvent and after sufficient exposure to substantially remove, in this case, all of the PS, leads to an approximate tomographic slice. This method can be very useful for gaining an understanding of the 3D structure of a sample in simple cases. A more sophisticated analysis uses the kinetics of the loss of material as a function of depth. At each exposure, the rate of loss of height is used to categorise the material. This then leads to a tomographic image that does not rely on an assumption that there is a phase that does not dissolve (only that the reference cells behave in a consistent manner) and does not require the user to identify physical boundaries. This approach can map occluded material.

In these examples, a single solvent was used and the composition of the sample was known in advance. However, unknown materials are routinely analysed in great detail using conventional analytical methods. Once the composition is known, solvents can be selected for the chemical imaging by dissolution analysis (CIDA) process. Here we have used only one solvent but, our experiment could be extended by using a second solvent that preferentially dissolves the PMMA; in this way any shielded PS would be exposed. It follows that our approach can be generalised to deal with more complex structures where solvents can be found that enable components to be dissolved at significantly different rates (it is not necessary that there is one component that does not dissolve at all). It is reasonable to expect this requirement will be achievable in most cases as changes in molecular structure will typically imply a change in dissolution behaviour. This is also true of changes in physical structure such as the differences between crystalline and amorphous forms of the same material.

In principle, therefore, the CIDA approach given here can be applied to materials with complex structures. Furthermore, it is not limited to organic materials. Metals could be analysed using different reactive solutions that preferentially etch one material more than another; a similar comment applies to ceramics. Biological systems pose greater challenges but highly specific bonding and bond-breaking reactions can be used and both the additional and removal of molecules could be analysed using the principles of CIDA.

An advantage of CIDA is that it requires no sample preparation. Almost any surface can be exposed to a solvent or a reactive fluid in a controlled way. There is a need that material be removed gradually. If multiple components are removed in a single exposure then interpreting the results could be problematic. Careful selection of solvents and adjusting times of exposure, temperature and rate of flow will, in most cases, enable the removal of material from the surface to be carried out progressively so that every component is removed in a step-wise manner. When multiple components are dissolved in a single step, the analysis of the solvent would detect that this has occurred and, therefore, that the dissolution regime may require refinement.

An important aspect of our method is that any analytical technique can be used. Often combinations of techniques could be employed to analyse the contents of a single solvent aliquot, for example both IR and Raman spectroscopies. A particularly powerful analytical approach is HPLC-MS (including tandem MS etc.). This provides a step-change improvement in analytical specificity compared to the vibrational spectroscopies now available for nanoscale imaging such as photothermal IR spectroscopy (called AFM-IR, Nano-IR or, thermomechncial photothermal microspectroscopy, PTMS) and scattering NSOM Raman and IR spectroscopies.

By adding the ability to analyse composition with chromatography (thus the ability to separate mixtures) as a function of depth, we will have achieved all four of the desirable capabilities listed above.

Experimental

Sample Preparation and Imaging. PS and PMMA were both from Aldrich Chemical Co. (PS: Mw = 2.40 x 10E5 g/mol, PMMA: Mw = 1.03 x 10E5 g/mol (determined by GPC)). The blend was composed of 30% w/w PS and 70% w/w PMMA. The sample was cast from solution in cyclohexane onto a glass substrate, and then conditioned in a vacuum oven at 40°C for one week. The AFM used was a Veeco (now Bruker) Explorer fitted with a high resonance frequency (HRF) silicon probe. The sample was imaged in the tapping mode (TM-AFM). Films of pure PS and PMMA were made in the same way.

Sample Dissolution and Chemical Analysis. For the AFM imaging studies, a single 20 μΐ drop of cyclopentane was delivered onto the sample surface using a fine pipette and, in the case of the AFM imaging experiments, removed using a one second burst from an air-duster after the required exposure time. The mouth of the tube delivering the pulse of compressed air was held 1 cm away from the droplet at a 45° angle. In order to ensure that approximately the same area was imaged each time, the sample remained within the base of the AFM thus it was not possible to collect the removed solvent for subsequent analysis. In a separate series of experiments the solvent was applied in the same way and then substantially removed with the same pipette before applying the air-duster. The reclaimed solvent was evaporated to dryness and the dry weight recorded. Samples were then analyzed by FTIR spectroscopy in the transmission mode. Two drops of cyclopentane were used to dissolve the sample which was then transferred onto a NaCl window and allowed to evaporate for ~ 5 minutes. Reference data were collected for films composed of the pure polymers as well as the blend. A calibration graph for the relative concentrations of the two polymers was obtained using the same procedure, except solutions with known concentrations of PS and PMMA were analyzed. The spectra were obtained from 450 to 4000 cm"1 at 2 cm"1 resolution setting and accumulation scans of 64. The FTIR and gravimetric data are presented in Figure 2. The spectra for the two pure materials each have a strong peak that is located in a region where the other has low absorbance; wavenumber 1729.5 cm"1 for PMMA, corresponding to its carbonyl group, and 698 cm 1 for the PS, corresponding to ring out-of-plane deformation. A calibration graph was created by dividing the absorbance at 1729.5 cm"1 by that at 698 cm"1 and plotting this ratio.

Drawings Key

Figure 1. Images of a selected region showing how material is removed from the surface in a series of steps; the scale is 50m x 5Hm. The red arrows indicate an area where there is a sudden removal of material resulting in a pit or trough compared to the previous image.

Figure 2. Left; from top to bottom, spectra of pure PS and PMMA then the spectra from the material extracted from the solvent placed on the sample surface. Right, from top to bottom, % of PMMA in the extract (inset is the calibration graph) then the gravimetric data for the blend, squares, and the pure PMMA, diamonds.

Figure 3. Top (small images) topographic images of the selected area after exposure to the solvent for, from left to right, 0,1, 2 ,4,6,8,10,12 and 15 seconds; note that features drift relative to the frame of the field of view in successive images. Bottom (large images) left, topography of the untreated sample and, right, the topography after 15 seconds exposure to the solvent. The centres of the white circles indicate three points that are chosen as being the same position in the two images. The white triangle joins these points for the first image and this shape is located in the same place relative to the edges in the second image; the drift can be clearly seen. All images are 50m x 50m.

Figure 4. Left, the untreated surface, middle, the same area after 15 seconds exposure. Note that the drift seen in Figure 3 has been substantially eliminated by the software as demonstrated by the white triangles (see Figure 3 for comparison). These images have been divided into a small number of cells to maintain clarity. Far right, the image is divided into a grid of 50 x 50. The filled-in cells at the centre of the white circles, denote those that were selected as being the same material that will dissolve in the same way. These positions were then used to align the images in the z axis.

Figure 5. The height vs time graph gives the plots of height change from untreated to 15 seconds exposure in each cell. The starting position for each cell has been set to zero. Underneath is a histogram of these height changes; a trimodal distribution can be seen. The red dashed line encompasses all of the positive values; in this case these give a measure of the error in the measurement as no cells increase in height. The mauve dashed line is a reflection of the red dashed line; together they estimate the error in determining those cells that do not, to a first approximation, change. All cells between a) and b) are coloured blue in the image shown (colour overlaid on topography image, the scale is 50m x 50m), cells between b) and c) are coloured green and those between c) and d) are red. This corresponds to blue - greatest change, green - some change and red = no change. Blue and green together map the cells where the raised features have dissolved, i.e. they must be almost entirely composed of PS. The green cells are where the height reduced to the level of the red cells. The blue cells are those where the dissolved phase created a hole or pit. The red cells are PMMA. The white dashed lines on the image are explained in Figure 7.

Figure 6. The top graph is equivalent to the similar plot shown in Figure 5 except all 8 images have been used. Cells that have changed significantly in height after 15 seconds exposure to the solvent are captured by the blue box made of dashed lines (this box is drawn by the user using the curser), all of these cells are coloured blue. The red dashed box captures all cells that have changed after 12 seconds, these are coloured red (overwriting the blue); the cells that have changed after 15 seconds but not after 12 seconds remain blue. This process is repeated and, in this way, the cells that have changed significantly from 12 seconds to 15 seconds are identified, then those that have changed from 10 seconds to 12 seconds etc. are identified with corresponding colour coding. The scale of the image is 50m x 50m. The white dashed lines are explained in Figure 7 and 9.

Figure 7. Top left, compositional map taken from Figure 5; red = PMMA and blue and green = PS.

Top middle and right; linescans taken from positions A and B in the compositional map for successive exposures to solvent (the same as is shown in Figure 6). For A, the step-change in height occurs at 2 sec., for B the change happens after 15 sec. The tomographic slice is created by a three step process shown in the bottom row; 1 - the linescans for the untreated surface and the same position after the step-change occurred are co-plotted, 2 - two positions delineating the edges of the raised feature are selected (the vertical lines in 2) and within these limits the two lines are preserved and outside only the untreated line is used, 3 - the phases are greyscale-coded.

Figure 8. These tomographic slices come from the positions indicated in Figure 5. These are in a sequence that is spaced uniformly thus they are the same as the familiar body-scanner type tomography.

Figure 9. These tomographic slices correspond to the lines A and B that feature in Figures 6 and 7. The greyscale coding shows the time at which the step-change decrease in height occurs. The histograms represent the decrements between 2 and 4 seconds for A and between 12 and 15 seconds for B. The red lines represent the boundaries; below the greyscale coding is dark grey (PS), above it is light grey (PMMA). Elsewhere the histograms always indicate only a slow rate of dissolution .For A, the kinetics suggest a thin layer of PMMA over the PS while for B the layer is much thicker.

Claims (47)

Tomography by Dissolution Analysis Claims

1. A method of chemically mapping the structure of a sample, said method including the steps of; - Placing or mapping a grid, graduations or scale onto at least part of an image of the surface of the sample; - imaging or mapping at least part of the surface structure of the sample substantially within the grid, graduations or scale using at least one microscopic technique; - dissolving or at least partially removing at least a first material from a surface of the sample by introducing or exposing said surface to a solvent or entrainer; - analyzing the solute, solution and/or contents of the solvents; subsequently imaging or mapping the surface structure of the sample substantially within the grid using at least one microscopic technique; and - comparing the images of the sample

2. A method according to claim 1 wherein the maps or grids are aligned by selecting at least one point or cell on the surface and/or surface image and aligning each image accordingly.

3. A method according to claim 1 wherein the solvent is provided as a series of aliquots.

4. A method according to claim 1 Avherein the microscopic technique is atomic force microscopy (AFM).

5. A method according to claim 1 wherein at least one linescan is performed.

6. A method according to claim 5 wherein at least two substantially parallel linescans are performed,

7. A method according to claim 6 wherein the linescans are shifted or separated substantially equidistantly.

8. A method according to claim 1 wherein aliquots of the solute, solvent and/or solution are used for analysis.

9. A method according to claim 1 wherein samples or aliquots are taken from the flow of the solute, solution and/or solvent.

10. A method according to claim 1 wherein analysis is performed using one or more spectroscopic and/or chromatographic techniques.

11. A method according to claim 10 wherein analysis is performed using any one or any combination of IR, Raman, HPLC-MS (including tandem MS), GC-MS.

12. A method of producing a representation of the structure of one or more samples, said method including the steps of; - Dividing or designating at least part of the sample into two or more parts, sections or cells; - Exposing said parts, sections or cells to at least one solvent or entrainer for a period of time; - Measuring and/or calculating the height reduction of material in said part, section or cell; Calculating and/or plotting the rate of height reduction for said section or cell; and Comparing the rate of height reduction for said sections or cells.

13. A method according to claim 12 wherein a series of aliquots of solvent or entrainer are applied to the sample.

14. A method according to claim 12 wherein the solvent or entrainer is supplied as a flow.

15. A method according to claim 13 wherein the solvent aliquots are pulses of solvent supplied to the sample in a flow of substantially inert entrainer or liquid or a sequence of pulses of solvent and inert entrainer or a gradual change in concentration such that the concentration changes from 100% solvent to 100% inert entrainer over a period of time.

16. A method according to claims 12-15 wherein the solvent or entrainer is supplied via or through a tube and/or conduit.

17. A method according to claim 16 wherein a distal end of the tube and/or conduit is placed on and/or substantially at the surface of the sample.

18. A method according to claim 17 wherein a droplet of solvent and/or solvent aliquots is delivered on to the surface of the sample.

19. A method according to any of claims 12-18 wherein at least part o f th e solvent, entrainer and/or solvent aliquot is removed or collected from the surface by a tube.

20. A method according to claim 19 wherein the collecting tube can be the same as that used to dispense the solvent or aliquot or can be a second or further tube.

21. A method according to claim 12 wherein at least part of the solvent, entrainer and/or aliquot droplet is removed from the surface via a tube by placing a distal end of a tube on, adjacent to and/or otherwise in the vicinity of a droplet.

22. A method according to claim 21 wherein the solvent, entrainer and/or aliquot droplet, or at least part thereof, is removed by pumping the same via a tube.

23. A method according to claim 22 wherein at least part of a first droplet is removed from a first location on the sample to a second location that is not on the sample.

24. A method according to claim 23 wherein a pump and a tube is used to pump the droplet fluid into a tube, said fluid staying in the tube until a second location is placed under the tube and/or the tube is moved to a second location.

25. A method according to claim 24 wherein the fluid in the tube is ejected or otherwise dispensed at the second location.

26. A method according to claim 25 wherein the material or fluid ejected onto the sample at the second location is analysed.

27. A method according to claim 16 wherein the tube used to place the droplet or fluid onto the sample has a tip with radius of less than one micron.

28. A method according to claim 16 wherein the tube used to place the fluid onto the sample has a tip with radius of less than ten microns.

29. A method according to claim 16 wherein the tube used to place the fluid onto the sample has a tip with radius of less than one hundred microns.

30. A method according to claim 16 or 19 or 23-25 wherein the tube is a probe in a scanning probe microscope.

31. A method according to claim 30 wherein the tube is moved in the x, y and z axes by one or more actuators.

32. A method according to claim 31 wherein the tube is used to deliver solvent to a selected area of the sample then the sample is moved into a microscope and the same area imaged.

33. A method according to claim 32 wherein the same tube is used to remove the solvent before imaging occurs.

34. A method according to claim 32 wherein the sample is moved from the microscope, then a tube is used to deliver solvent to a selected area of the sample then the sample is moved back into the microscope and the same area imaged.

35. A method according to claim 34 wherein the same tube is used to remove the solvent before imaging occurs.

36. A method according to claim 32 wherein the sample remains in the microscope and a tube is moved to be above a selected area on the sample, the tube delivers a droplet of solvent and then the tube is removed and the microscope images the same area of the sample.

37. A method according to claim 12 wherein the height reduction is measured and/or calculated using one or more imaging techniques.

38. A method according to claim 12 wherein proximal probe imaging and/or scanning probe is used to measure and/or calculate the height reduction.

39. A method according to claim 38 wherein microscopy is used to measure and/or calculate the height reduction.

40. A method according to claim 39 wherein atomic force microscopy (AFM) is used to measure and/or calculate the height reduction.

41. A method according to claim 12 wherein the rate of height reduction is calculated and/or plotted as a function of time.

42. A method according to claim 12 wherein the one or more sample includes at least one reference point or reference material.

43. A method according to claim 42 wherein the reference point, or reference material comprising the reference point, is of substantially uniform composition

44. A method according to claim 43 wherein the sample includes a plurality of reference points and/or materials.

45. A method according to claim 42 wherein the dissolution and/or solvation properties of the reference point or material are known and a solvent is chosen that does not dissolve the reference material.

46. A method according to claim 42 wherein where the reference material is dissolved, the height reduction of the reference material or point is calculated.

47. A method of chemically mapping the structure of a sample, said method including the steps of; imaging or mapping at least part of the surface structure of the sample using at least one microscopic technique; dissolving or at least partially removing at least a first material from a surface of the sample by introducing or exposing said surface to a solvent; analyzing the solute, solution and/or contents of the solvents; subsequently imaging or mapping the surface structure of the same area of the sample using at least one microscopic technique; and comparing the images of the sample