Sticky Nanoparticles

Revealing the Adhesion Dynamics of TiO2 Nanoparticles

  • Fig. 1:TEM images taken in sequence while pulling the NPs chain apart. (a) The chain is attached to the tip on the one side and to a tungsten wire to the other side. (a)-(c) The black arrows mark the neck where the chain breaks. (d) Upon re-approaching, the detached chain parts re-connect.  Fig. 1:TEM images taken in sequence while pulling the NPs chain apart. (a) The chain is attached to the tip on the one side and to a tungsten wire to the other side. (a)-(c) The black arrows mark the neck where the chain breaks. (d) Upon re-approaching, the detached chain parts re-connect.
  • Fig. 1:TEM images taken in sequence while pulling the NPs chain apart. (a) The chain is attached to the tip on the one side and to a tungsten wire to the other side. (a)-(c) The black arrows mark the neck where the chain breaks. (d) Upon re-approaching, the detached chain parts re-connect.
  • Fig. 2: Sequence of enlarged TEM images taken upon tensile loading (a)-(e) and re-attachment (f). The NP2 chain consists of several segments, highlighted in different colors. Upon tensile loading the segments rotate around the contact point to the adjacent segment without changing their shape. (d) Prior to the final rupture, local detachments occur (e.g. area marked by green arrow leading to significant shape change. After the final rupture (e) the broken segments "snap-in" at a distance of a few nm and reconnect at a different location (f).
  • Fig. 3: (a) A force-distance curve obtained from manipulation with TiO2 NP aggregates. The force F1 is correlated with the final rupture of the NPs chain whereas the sawtooth-like structure originates from rearrangements including sliding, rotation and local ruptures. (b) Statistical distribution of the force involved in the peaks (black) and that of F1 (red). Both indicate a maximum at 2.5 nN.

Nanoparticles can show different adhesion behavior than their larger counterparts due to the increased surface to volume ratio. In this work, we combine AFM force spectroscopy and in situ manipulation inside a Transmission Electron Microscope to identify the adhesion dynamics of TiO2 nanoparticles. This combination allows us observing TiO2 nanoparticles rearrangements while pulling them apart and measuring the contact forces of individual TiO2 nanoparticles.


Adhesion forces between individual nanoparticles (NPs) play an important role in many different processes, such as fluidization [1,2], agglomeration and coating [3,4], utilized in a wide range of technical fields [5-7]. These interactions strongly depend on the size of the particles. Especially for NPs of the order of 10 nm, subtle effects, such as molecular structure of adsorbate layers or the distribution of terminal groups on the surface can significantly influence and even dominate the adhesion behavior. Therefore, measuring the direct NP-NP interaction is of particular importance. However, the difficulties to directly measure the contact forces hamper gaining fundamental knowledge about the mechanisms of adhesion. So far, atomic force microscopy (AFM) has been used extensively to study the NP-NP interaction. Nevertheless, conventional AFM force spectroscopy enables measuring adhesion force quantitatively but without the capability of observing details of the adhesion dynamics. Moreover, most of the studies rather measured the interaction between a NP and a reference, e.g. functionalized tips with NPs, and not directly the NP-NP interaction [8,9]. In this work, we combine AFM force spectroscopy measurements with in situ manipulation inside a transmission electron microscope (TEM) and correlate the stretching and de-agglomeration behavior of TiO2 NPs directly with adhesion forces measured by AFM force spectroscopy.

Observation of Stretching and Rupturing Behavior of TiO2 NP Agglomerates

For the TEM observation NPs were sprayed on the edge of a tungsten wire, which was mounted onto the hat of the AFM-TEM holder [10].

Figure 1 shows a sequence of the TEM images, taken during the deformation of the NPs. After direct contact with NPs, the AFM tip was decorated with NPs with multiple aggregates and extended branches (fig. 1a). A rearrangement of the NPs occurred during the retraction with a constant speed of about 5 nm/sec. As can be seen in figure 1, the agglomerates first unfold (see area on the right-hand side of the black arrow), when subjected to tensile load, and rearrange to an elongated chain of NP aggregates aligned next to each other. Generally, we could retract more than 150 nm without breaking the chain structure. Just before the final rupture an elongated single chain of NPs formed where most NPs lined up with minimized number of direct neighbors. A sudden rupture occurred associated with the detachment of the contact between two individual primary particles (fig. 1c). Approaching the detached agglomerates again led to their reconnection (fig. 1d). At a distance below 10 nm (difficult to measure precisely because of NPs overlapping), the agglomerates "snapped-in" to the detached counterpart indicating attractive physical forces. As can be seen in figure 1d, the reconnection does not occur at the single point contact from where the agglomerate had detached but preferentially at a location with multiple contacts. Re-stretching the NPs chain showed reversible rupturing behavior but with a different detachment point.

More detailed NPs arrangements during the stretching, rupture and re-attachment can be seen in the enlarged TEM images presented in figure 2: The chain contains several segments of agglomerates (highlighted in different colors), which are connected via a single or double contact with adjoining primary particles forming a neck. In the initial phase of tensile loading, these arrangements mostly tilt around the contact point in order to rearrange to a linear chain. For the segments highlighted in green, red and blue the difference in tilt angles measured in figure 2a and figure 2c are about 15°, 12° and 30°, respectively. In contrast, the shape of these segments remains nearly preserved during this rearrangement (fig. 2a-c). Upon further loading, individual NPs start to rearrange and the initial shape of the segments significantly changes (fig. 2d). For instance the segment colored in green tears off at the location marked by the green arrow and transforms from a double chain, ring-like configuration to a single chain. Multiple local ruptures of this type occurred preferentially prior to the final rupture. Being abrupt they could clearly be observed but the sensitivity of the AFM-TEM setup was not sufficient to quantify the force involved.

Adhesion Forces of TiO2 NPs

Figure 3a shows a force-distance curve measured by means of AFM force spectroscopy. For these measurements, TiO2 NPs were deposited on a flat mica substrate. The force-distance curve exhibits a sawtooth profile with multiple peaks. The final rupture results in a sudden step (labeled as F1 in fig. 3a) in the force-displacement curve and the force drops back to zero. In order to quantitatively analyze the adhesion force, a set of sequential 1024 force-displacement curves was acquired by repeatedly approaching the AFM tip to the NPs film and retracting to full detachment [11]. Figure 3b presents the statistical distribution of the forces determined from the peaks extracted from these force-displacement curves. The statistical distribution of the force F1 correlated with the final rupture was separately analyzed and revealed a log-normal distribution with a maximum at 2.5 nN, which coincides with the maximum of the global distribution shown in figure 3b [11]. Accordingly, the characteristic contact force between two individual primary TiO2 NPs can be concluded to about 2.5 nN.

Remarkably, the forces smaller than 1.2 nN are present in the global distribution but not in the F1 distribution, indicating a minimum threshold in the contact force between two primary NPs. Consequently, the low forces can be attributed to sliding or rotation of the agglomerate segments leading to unfolding to an elongated chain whereas the rearrangements involving 2.5 nN go together with local detachments of two NPs and give rise to the abrupt peaks in the force-distance curve. In an extended study we demonstrated that the size of the agglomerates strongly determined the rearrangements of single segments [12]. While large agglomerates showed multiple rearrangements, agglomerates smaller than 100 nm just detached without several rearrangements. However, the force F1 leading to final rupture remained independent of the agglomerate size [12].

The characteristic contact force of about 2.5 nN is difficult to explain by assuming van der Waals type interactions only but it shows excellent quantitative agreement with the results of Molecular Dynamics simulations of the NP-NP detachment when capillary effects of the water molecules present on the particle surface are taken into account [11]. According to the simulations, when two NPs approach each other, at a center-to-center distance of about 5.5 nm a first water neck builds up and leads to an attraction force that increases roughly linearly up to a value of 1 nN at a separation of about 4.8 nm [11]. Indeed, while re-approaching the fractured NP chain segment inside the TEM a "snap-in" occurred at a distance of several nm, which is significantly higher than the typical distances of van der Waals type interactions (< 1 nm). Although the amount of adsorbed water molecules may be different in vacuum inside a TEM and in atmospheric conditions of an AFM, this observation demonstrates that the water molecules are present in both cases. Hence, both the qualitative and the quantitative adhesion mechanism observed by means of in situ TEM and AFM force spectroscopy, respectively, show excellent agreement and reveal the details of the dynamic behavior of TiO2 NP networks where the molecular nature of the adsorbates plays an important role.


This research was funded by German DFG (SPP 1486) under grants MA 3333/3 and CO 1043/3, Flemish Hercules Stichting (HER/08/25) and KU Leuven STRT1/08/025. We also thank Prof. Lucio Colombi Ciacchi and Jens Laube for valuable discussions.

[1] Seville J.P.K. et al.: Powder Technol. 113, 261-268 (2000)
[2] van Ommen R. et al.: J. Nanoparticle Res. 3, p. 1 (2012)
[3] Jono K. et al.: Powder Technol. 113, 269-277 (2000)
[4] Goulas A. et al.: J. Mater. Chem. 1, p. 4647 (2013)
[5] Kumar A. et al.: Nat. Mater. 7, 236-241 (2008)
[6] O'Regan B. et al.: Nature 353, 737-740 (1991)
[7] Kashiwagi T. et al.: Nat. Mater. 4, 928-933 (2005)
[8] Farshchi-Tabrizi M. et al.: Langmuir 22, p. 2171 (2006)
[9] Ong Q. K. et al.: J. Colloid Interf. Sci. 310, p. 385 (2007)
[10] Nafari A. et al.: J. Microelectromechanical Syst. 17, 328-333 (2008)
[11] Salameh S. et al.: Langmuir 28, p. 11457 (2012)
[12] Salameh S. et al.: Powder Technology p. 345 256 (2014)

Prof. Jin Won Seo

Department of Materials Engineering (MTM)
KU Leuven
Leuven, Belgium

Dr. Samir Salameh
Prof. Lutz Mädler

Foundation Institute of Materials Science (IWT)
Department of Production Engineering
University of Bremen
Bremen, Germany


KU Leuven
Kasteelpark Arenberg 44
3001 Leuven
Phone: +32 16 321272

Register now!

The latest information directly via newsletter.

To prevent automated spam submissions leave this field empty.