Atomic Force Microscopy

AFM for the elucidation of solid state syntheses which avoid wastes

Prof. Dr. Gerd Kaupp, Dipl.-Chem. Jens Schmeyers,
Dipl.-Chem. Michael Haak, Cand.-Chem. Andreas Herrmann
Labo 1995 6 57

Article installed with permissions of the publisher
WWW-adaption by M. Haak

The AFM-technique provides new answers to the question of organic solid state reactivity and it uncovers unexpected mechanistic details, which cannot be observed by other analytical procedures. The new concept of phase rebuilding provides comprehension of the longrange anisotropic molecular movements on the basis of crystal packing data. Previously such movements thought impossible and were not forseen, but they now prove essential for the chemical reactivity. Thus a theoretical comprehension of surprisingly efficiant waste-free techniques of syntheses is provided for the first time and such syntheses have been developped up to the kg scale already. Examples for gas/solid- and solid/solid-reactions have been chosen from various branches of organic chemistry. All of them run uniformly and quantitatively and are clearly superior to the corresponding liquid phase reactions with their waste problems. Therfore gas/solid reactions are not only able to detoxicate atmospheric air or exhaust gases but are particularly important in view of their potential as primary action for the avoidance of wastes in industrial chemical production. Both gas/solid- und solid/solid-reactions are particularly simple and cost-effective because they avoid solvents and liquid phases. AFM helps in finding systematizing and predicting further examples by indicating the reasons for the success of these synthese techniques for the future.


Atomic Force Microscopy (AFM) is a very young surface technique. It was invented in 1986 [1] and unlike the 5 years older Scanning Tunneling Microscopy (SFM) it is able to image electrically non-conducting surfaces as well. In favourable cases AFM may resolve single molecules (more rarely single atoms) laterally and vertically [2]. However while the recording of organic molecular steps poses no problems to AFM [3], the imaging of individual atoms is more practical with STM for technical reasons. The longer known STM scans constant tunneling currents, whereas the AFM scans essentially van-der-Waals forces. Meanwhile about 30 different SXM-techniques are available, however, their description would exceed the scope of this contribution.

Imaging of individual atoms is epoch-making. For applied AFM it is important to scan in the nm- and m-range if the surface structures happen to occupy that range. Both physical properties and chemical reactions may depend on submicroscopic surface structures. The requirements for the scanning of large, steep, densely packed, deeply penetrating nanostructures with AFM have been summarized [4]. In Ref.[4] is also shown how changes of the surface during the scanning are avoided, how artefacts are recognized, and how the images look like if nanoliquids form at the surface or if the limits inherent to the AFM techniques are surpassed, an example being the imaging of membrane pores. Waste-free solid state syntheses which do not use solvents are the future production techniques, undoubtedly. Solid-state photolyses are longer used already [3] however, their large scale technical use poses practical problems.

Much more interesting are gas/solid- [5] and solid/solid-syntheses [6]. Both of these deserve puplic support and publicity. The development of such techniques was hampered, because incomplete theories of the so called topochemistry [7] were not able to "understand" the occurrance of such reactions. Despite the fact that most reactions of the new techniques for syntheses proceed quantitatively and avoid all kinds of waste (several examples have been performed in the kg scale) the opinion still prevails, that one should rather dissolve the reactants even at the expense of usually lowered yield and unnecessary problems with waste. Therefor it is very favorable that the investigation of organic gas/solid- [5] and organic solid/solid-reaktions [8] with the AFM uncovered unexpected new theoretical foundations. Thus primary actions may now be accepted which avoid waste problems because there is not only an experimental basis but also a theoretical one.

Atomic Force Microscopy

An atomic force microscope [1] has a surprisingly simple design and it does not require unusually high investments for its purchase and use, at least as long as ambient conditions (1 bar at RT) are prefered, which in any care are favorable for handling volatile organic crystals. Figure 1 shows the basic setup for the most used commercial version of an AFM.

principle of AFM-measurement

A tip, usually made of silica nitride in 4-sided pyramidal form (Si3N4-Standard tip, radius 20-50 mm, apex angle 45, height 2.8 m, base 4x4 m) is connected to a fixed substrate on a cantilever with a very low spring constant (typically 0.1 N/m). The tip is approached to the surface until the van-der-Waals attractive forces are 10-6-10-9 N (in atomic resolution experiments the approach most be closer, down into the repulsive region, and the tip must have an atomic end). The solid surface is then scanned under the tip by mean of a XYZ piezo drive with the recording of about 400x400 points or even more of them. The constant height regulation during the scan is controlled by a feetback loop of high dynamics from a divided photodiode, because the tip is not supposed to hit any obstacles which may be present on the surface. That feature is reached by a laser beam which is focussed to the back of the gold-coated side of the cantilever just opposite to the side of the tip. From there the laser beam is reflected to a mirror which guides it to the split photodiode. Every change in height is recognized and immediately compensated for by the Z-piezo. A surface topography is created. The feedback loop can be based on constant heigh or on constant force between the tip und the surface. Exhausting imaging software packages including drivers for suitable printers are available for all commercial AFM instruments. In this work a NanoScope II AFM from Digital Instruments has been used with NanoScope III software for the imaging and printing with a color video printer. Further details are described in the publications [3-5] and [8-11].

Phase Rebuilding in Solid State Reactions

The topochemical principle [7] has no answer to gas/solid- and solid/solid reactivity, because it believes that it has to consider only those processes, which are presumed to occur with minimal atomic molecular movements in the range of less than 4.2 . However, AFM analysis shows that, with the only exception of extremly rare topotactic photoreactions [3], all chemical crystal reactions require long range anisotropic molecular movements over several or up to hundreds of nm. Those movements are well directed vertically and/or laterally, the reactions working themselves from the surface down into the crystal. Additionally, those movements depend on the crystallographic face wich proves that they are guided by the initial crystal lattice even in the cases which involve movements over distances of several hundred nm. The molecular movements create characteristic surface structures. Right now 8 different geometric types of them are known [3-5,10]. If no molecular movement is possible due to particular features of the crystal packing no chemical reaction does ccur. This is also true if the distances of potentially reactive centers are very short (<4.2 ) and their orientations are favourable. Also numerous positive exceptions to topochemistry (solid state reaction despite large distances of up to 10 ) are now easily comprehended [3]. The experimental results enforce the introduction of the new principle of "phase rebuilding" [3-5, 9-11]. That principle stresses the need for immediate accommodation of new molecules with the crystal lattice present, after their formation in solid state chemical reactions. As the product molecules have shapes that differ from those of the starting molecules, strain is created which has to be released by movements along the easiest ways the crystal packing allows for (face dependancy). Only if chemical transformation has increased so largely that the distorted original lattice cannot be retained for thermodynamic reasons, a phase transformation may occur in which mixed crystals of starting material (low amounts of product) transform into mixed crystals of product (with low amounts of starting material). Thus, it is frequently observed that initially formed surface features disappear on prolonged reaction at the expense of new features which may be considerably larger. If crystal data are available it proves possible in all studied examples to correlate the geometric type of surface feature due to phase rebuilding with the molecular packing. It is thus secure to conclude that anisotropic molecular movements are the decisive criterion for solid state reactivity, though in addition to thermodynamic feasibility, of course.

Gas/Solid-Syntheses Using Acetone Vapor

Acetone is a low boiling volatile solvent of multiple use in research and production. Its vapors in exhaust air have to be removed down under the detection limit. Gas/solid reactions which form valuable and usable products (instead of waste, subject to disposal) are well suited for that purpose. Such a technique has been recently realized on the 200 g scale [9]. In that procedure hydroxylaminiumphosphate and K2HPO4*H2O are reacted with highly diluted acetone vapor at 80 C and a flow rate of 100 ml/min. The Reaction runs to completion, water and the valuable products acetoneoxime and KH2P04 are formed. That procedure is at the same time the most environmentally friendly synthesis of acetoneoxime, which can, of course, also be performed with nearly saturated acetone vapor in chemical production or for student classes. From the numerous preparatively useful gas/solid reactions of acetone with primary amine reagents and aminothiols we discuss here its reaction with hydroxylamine hydrochloride to give acetoneoxime hydrochloride.

Reaction Scheme 1

The AFM-investigation with single crystals of hydoxylamine hydrochloride (from methanol, main face) shows, that steep surface features form in moist air (18 C, rel. humidity 75%) within 200 min. Those are solid surface hydrates which react considerably faster with gaseous acetone than do unhydrated surfaces of the same reactant. Thus, it is clear, that the water of reaction acts autocatalytically if the reaction proceeds from the surface region into the interior of the crystal.

Fig. 2 excellent quality (GIF 244 K) or Fig. 2 (JPEG 55 K)

Figure 2: AFM surface of hydroxylaminehydrochloride; a: main surface after 20 min manipulation in moist air; b: after 200 min hydration in moist air; d: same as b but inverted image; c: after 2 min application of diluted acetone vapor on the surface hydrate of hydroxylaminehydrochloride.

Figure 2 shows both the hydration and the reaction of the surface hydrate with acetone vapor at short exposure. It can be seen, that the hydration forms very steep surface structures with heights of more than 1 m. These features do not exhibit any asymmetries and are thus free of artefacts, at least in their upper parts [4]. However,the structures stand so densely (about one per m2), that the rather large standard tip cannot reach the ground without its pyramidal faces being touched by the features. That circumstance is not seen in Figure 2b (and even less so in 2D-projections of it), but very clearly in the inverted perspective image (Figure 2d) which is most easily obtainable. In the inverted image the tops are pointing down and the depths are pointing up. The sharp edges which appear show that the pyramidal faces and the edges of the tip had contact with the cones (or even columns ?) on the surface while it tried to descend into the depths. Sometimes, if there was simultaneous contact from all sides a true image of the pyramidal tip is seen in Figure 2d (therein enlargement of the Z-scale!). It is to be concluded, that the surface features in Figure 2b are both taller and higher than can be traced with a standard AFM-tip. The advantages of perspective imaging over 2D-projections (which cannot image such tip/surface convolutes) are evident. The reactions of the (hydrated) hydroxylaminehydrochloride surface with acetone vapor leads to massive structures which are more than 4 m high and exhibit the heights and valleys type (Figure 2c). The molecular movements that have occurred are very far-reaching when compared with molecular dimensions. The gas/solid reaction is unusually effective and the stability of the AFM-images during more than 10 consecutive scans proves that no liquid or nanoliquid phases occur [4].

Gas/Solid-Reactions with Amines

Gaseous ammonia and gaseous amines are formed in world-wide agriculture and biological metabolism. They are responsible for new damages to forests by their overfertilizing when entering through the atmosphere. Therefor, gas/solid reactions of these species are of particular importance for atmospheric chemistry and the development of detoxification strategies. In addition to that it is also of importance to create new synthetic techniques which avoid all kinds of waste [10]. Numerous systems have been realized. From these the addition of methylamine to crystalline thiohydantoine with ring opening has been chosen for analysis with the AFM.

Reaction Scheme 2

(N-Methylacetamidomethyl)thiourea is obtained quantitatively and its waste-free formation has synthetic potential. The AFM-scans in Figure 3 show a smooth natural (110)-surface which separates into heights and valleys after treatment with methylamine gas. The direction of preference is along the crystallographic cleavage plane present in the original crystal lattice. Again, we see long range molecular movements as a consequence of phase rebuilding and there is a clearcut relation to the crystal packing.

Fig. 3 excellent quality (GIF 105 K) or Fig. 3 (JPEG 28 K)

Figure 3: Natural (110)-AFM-surface of thiohydantoine; a: fresh; b: after repeated treatment with gaseous methylamine

Gas/Solid Syntheses with Nitrogen Dioxid

Nitrogen dioxid is a toxic gas from (automotive) firings which spoils the environment and which cannot be readily controlled, yet. It is a free radical and has a high reactivity which can be used in numerous gas/solid syntheses which, unlike common solution reactions, proceed quantitatively and waste-free [11].

Reaction Scheme 3

The quantitative tetranitration of tetraphenylethylene is quite spectacular (NO is oxidized with O2 to NO2 and thus recycled). Conventional techniques of synthesis yield 65-70 % yield at most (under optimized conditions) from a highly corrosive nitric acid solution after enormous efforts of purification and with severe problems in waste disposal. The AFM scans in Figure 4 show, that cyclic craters with nicely shaped walls of diverse size form upon phase rebuilding in this reaction. That is seen both in the perspective and in the 2D-projection (Figure 4b, d).

Fig. 4 excellent quality (GIF 319 K) or Fig. 4 (JPEG 54 K)

Figure 4: AFM-surface of tetraphenylethylene; a: fresh natural (10-1)-surface; b/d: after 10 min exposure to 0.2 bar of NO2; c: after 20 min exposure to 0.2 bar of NO2;

Upon longer exponere to NO2 gas the circumwalled craters disappear, no doubt, as a consequence of phase transformation into the product lattice. The initially formed cyclic structures are again comprehended by correlation with the crystal structure of the starting material [11]. The tetraphenylethylene molecules which stand vertical under (10-1) have a limited possibility for partial turns around their long axis. They use it if they grow thicker upon nitration. By doing that they collide with next row molecules which then turn in opposite direction and provoke a third and then a forth molecule to the corresponding turns with finally filling the hole left from the first of these molecules. Thus, a concentric process leaving a hole at its center is described. After such initiation the growing molecules (consecutive nitrations) move upward and present their bottom p-positions for nitration with NO2 gas. The crater diameter is enhanced by participation of neighbouring rows of molecules and those processes proceed until the original lattice breaks down [11]. Again, the crystal structure dictates the long range anisotropic molecular movements during phase rebuilding.
Contrary to solution diazotations, gas/solid diazotation may proceed waste-free and allow elegant applications in syntheses [12]. The solid state diazotation of sulfanilic acid to give its diazonium salt, which is needed for the synthesis of widely used pH-indicator methylorange (via azo coupling with N,N-dimethylaniline) provides particularly impressive AFM-images.

Reaction Scheme 4

Both anhydrous sulfanilic acid or its monohydrate, which forms monoclinic prisms with 1-3 mm long edges, diazotize quantitatively under the action of gasous NO2. The solid product crystals may be ground for IR-pellets etc. They are however, as are other solid diazonium salts, explosive if heavily hit with a hammer on an anvil. The AFM-investigation on the (010) pinacoid of sulfanilic acid monohydrates (at 22C in the air with a rel. humidity of 41%) is shown in Figure 5.

Fig. 5 excellent quality (GIF 256 K) or Fig. 5 (JPEG 49 K)

Figure 5. (010)-AFM surface of sulfanilic acid monohydrate; a: fresh and barely effloresced; b/d: island formation after short diazotation with NO2-gas; c: after advanced diazotation by several treatments with NO2.

The initial surface in Figure 5a shows traces of efflorescence which may be the reason for the volcano like features. These are swallowed after short application of gaseous NO2 to the surface in air. Flat islands which may enclose little lakes are formed (Figure 5b, d). The characteristic island features are still recognizable after multiple application of NO2 (Figure 5c). However, the advanced height-growth is less uniform. A more roughened height scenery with embraced depths is optained. The formation of islands [10] is easily derived from the crystal structure of sulfanilic acid hydrate [13], which is depicted in Figure 6.

 Figure 6 (GIF 22K)

Figure 6. Space filling stereoscopic representation of the molecular packing of sulfanilic acid monohydrate; top: view on the (010)-face showing hydrogen bridges between NH3+ and SO3- groups with additional H-bridging by H2O molecules; bottom: view on (100) which is orthogonal to (010) showing how the molecules alternate in their orientation and how they stand long-elongated under the (010) face which is on top; N: grid; S: shaded; O at S: circles; O of the water: dense circles.

The molecules stand vertically under (010) and their orientation alternates. Thus, a maximum of hydrogen bridges is obtained and the water molecules add further H-bridging. While the upmost layer on (010) is open for attack of the NH2 groups by NO2 of half of the molecules (Figure 6, top), the proceeding of the reaction into the crystal requires, that long elongated molecules exit from their H-bonded lattice positions and move upward in order to make the amino groups available for diazotation which pointed downward from the first layer and upward from the second layer etc. Clearly the reaction is hindered by this particular packing. Therefore the reaction spreads around the initial successfull invasions into row 2 etc. and thus from flat islands. 2 planes with a distance of 15 nm are present in Figure 5b/d.


Waste-free syntheses which succeed by simply mixing of crystalline reactants are particularly attractive for saving ressources [6]. The first AFM-investigation of such a reaction concern the quantitative pinacol rearrangement (elimination of water with rearrangement) of benzopinacol [8].

Reaction Scheme 5

A small crystal of p-toluenesulfonic acid hydrate is placed on the (100)-face of a larger single crystal of benzopinacol for the AFM-investigation. The large crystal is scanned at different distances from the edges of the small one. At 0.5 mm distance (Figure 7a) a moderately stepped surface image is obtained. The same is true at 0.1, 1 and 2 mm distance. Here and in the subsequent AFM-scans it does not make any difference if measurement is done along the b-axis or at 90, 60 or 45 to it. Also higher terrasse steps are found in larger scale 10 m-scans [8].

Fig. 7 excellent quality (GIF 292 K) or Fig. 7 (JPEG 57 K)

Figure 7: AFM-surface of benzopinacol (100); a: fresh in a distance of 0.5 mm from an onlaying crystal of p-toluenesulfonic acid monohydrate b/d: after 12 h at 50 C, at a distance of 0.5 mm; c: after 12 h at 50 C, at a distance of 0.1 mm;

The crystals are heated to 50C for 12 h to initiate the reaction. They are remounted to the AFM and then scanned. One obtaines well-formed craters in 0.5 mm distance from the proton source (Figure 7b,d). Those are 5 mm deep and 150 mm wide. Importantly the reaction has much more proceeded it 0.1 mm distance (Figure 7c). The craters found there are about 11 nm deep and are mostly circumwalled with heights of typically 4 nm. Only at distances of more than 2 mm one finds surface images similar to the one in Figure 7a. The distance dependent reaction and crater formation can be understand again in terms of crystal data [8]. Here catalytic quantities of protons move over 2 mm from the small acid crystal and produce a charge separation, causing the crystals to stick together firmly. Mechanistically the protons combine with OH groups at the front side of molecular layers and water is liberated. Therefor, new protons are formed at the backside of cations after their rearrangement to yield triphenylacetophenone. These newly formed protons react with the frontside OH groups of the next layer etc. [8]. Interestingly, the morphologically dominant (001)-face does not react correspondingly, according to the AFM investigation. That surface remains unchanged after the same treatment and the acid crystal does not stick to that surface of benzopinacol. The reason for that is to be found in the orientation of the OH groups in the upmost layer of benzopinacol. On the (001)-face the hydroxyl-H (but not the free electron pairs) point outward. Therefor, no protonation can occur, inasmuch as there is also steric hindrance [8]. Conversely, on the reactive (100)-face, the free electron pairs are pointing outward and may thus be protonated [8]. This first mechanistic investigation of a crystal/crystal-reaction shows the potential of AFM and indicates a mechanistic understanding of the many further waste-free solid/solid-reactions which occur in most diverse types [6].


[1]  G. Binning, C. Quate, C. Gerber: Atomic force microscope,
Phys. Rev. Lett. 56 (1986) 930.
[2]  J. Frommer: Scanning Tunneling and Force Microscopy in 
Organic Chemistry, Angew. Chem. Int. Ed. Engl. 104 (1992) 1324.
[3]  G. Kaupp: AFM and STM in photochemistry including photon
tunneling, Advances in Photochemistry 19 (1995) 119.
[4]  G. Kaupp, J. Schmeyers, U. Pogodda, M. Haak, T. Marquardt, M. Plagmann:
AFM for the imaging of large and steep submicroscopic features,
artifacts and scraping with asymmetric cantilever tips,
Thin Solid Films, SXM1 special issue, 1995, in press.
[5]  G. Kaupp: Atomic force microscopy in organic gas/solid-reactions:
How do the new phases build up?, Mol. Cryst. Liq. Cryst. 211 (1992) 1, cit. lit.
[6]  F. Toda in Y Ohashi (Hrsg.): Reactivity in Molecular Crystals,
Kap. 4: Solid-to-solid organic reactions, Kodansha/VCH, Tokyo/Weinheim, 1993, S. 177.
[7]  M. D. Cohen, G. M. J. Schmidt, F. I. Sonntag: J. Chem. Soc.
(1964) 2000; G. M. J. Schmidt: ibid. (1964) 2014;
M. D. Cohen: Angew. Chem. 87 (1975) 439; Mol. Cryst. Liq. Cryst. 50 (1979) 1.
[8]  G. Kaupp, M. Haak, F. Toda: Atomic force microscopy and solid
state rearrangement of benzopinacol,
J. Phys. Org. Chem. S 95 H 12, in press.
[9]  G. Kaupp, U. Pogodda, J. Schmeyers: Gas/solid-reactions with
acetone, Chem. Ber. 127 (1994) 2249; cit. lit.
[10] G. Kaupp, J. Schmeyers: Gas/Solid-Reactions of Aliphatic Amines
with Thiohydantoines: Atomic Force Microscopy and New Mechanisms,
Angew. Chem. Int. Ed. Engl. 105 (1993) 1656.
[11] G. Kaupp, J. Schmeyers: Gas/solid-reactions with nitrogen
dioxide, J. Org. Chem. 60 (1995) Heft 16, in press.
[12] A. Herrmann: Diploma Thesis, University of Oldenburg, 1995.
[13] A. I. M. Rae, E. N. Maslen: The crystal structure of
 sulfanilic acid mono-hydrate, Acta Cryst. 15 (1962) 1285.

Supplementary Material:
A comparable illustration about the use of AFM elucidating mechanisms of crystal photolyses can be found in
G. Kaupp: Fachzeitschrift für das Laboratorium GIT 37 (1993) 284-294 und 581-586.

To the Homepage of AK Kaupp (Organische Chemie I)