Terahertz-based nanometrology: multispectral imaging of nanoparticles and nanoclusters in suspensions

  • Anis RahmanEmail author
  • Donald A. Tomalia
Research Paper
Part of the following topical collections:
  1. Unifying Concepts for Nanoscience and Nanosystems: 20th Anniversary Issue


Silica nanoparticles suspended in an organic solvent (nanosuspension) have been imaged and characterized via terahertz nanoscanning reconstructive three-dimension (3D) imaging technique. The size of individual silica nanoparticles in the suspension was quantified. In addition, the presence of nanoclusters along with their distribution in the suspension was visualized in 3D. It has also been qualitatively demonstrated that the volume fraction of solvent is significantly higher than that of the silica nanoparticles; an observation consistent with the composition of the nanosuspension in the present investigation. The measured size range of individual nanoparticles was found to be 10–12 nm, while the manufacturer’s specification indicates a nanoparticle size distribution in the range of 10 to 15 nm. However, a typical nanocluster size was determined to be 17.5 nm, thus indicating the presence of nanoparticles less than 10 nm. The nanometrology instrument used in this investigation was based on a dendrimer dipole excitation-based continuous wave terahertz source generating > 200 mW stable terahertz power.


Terahertz multispectral reconstructive imaging Nanoslurry Silica nanoparticles Nanoclusters Nanosuspension Nanoscale instrumentation 


Characterization of nanoparticles in suspension, or as “nanosuspensions” is important in a wide range of areas such as; drug delivery, lubricants, as well as abrasion and polishing protocols. More specifically, nanosuspensions are used for fine scale polishing in the semiconductor industry. Such suspensions are commonly referred to as “slurries,” wherein, silica nanoparticles are suspended in a suitable solvent for the polishing process known as the chemical and mechanical polishing (CMP) protocol. Many different combinations of nanoparticles and suspending media are used depending on the formulation and polishing specifications. However, a common problem in all cases, is to measure the size of the nanoparticles while they are still in the suspension. Many existing methods require extensive optimization, especially for new types of nanoparticles and solvent combinations. Some insights may be gained from the fields of environmental chemistry of natural nanomaterials and from fundamental colloid chemistry (Hassellöv et al. 2008; Wang et al. 2009). A critical requirement for these polishing slurries is that the nanoparticles must remain in suspension without agglomeration since particle aggregation will change the polishing characteristics.

Recently, we reported a method for measuring the size of quantum dots (Rahman et al. 2016a) and carbon nanotubes (CNTs) (Rahman 2016). These nanoparticles were also in solution, but they were first spun onto a silicon wafer. Using this spinning protocol, the nanoparticles formed layers of specific thickness that are determined mainly by the solution concentration and casting speed. Following the approach described in (Rahman 2016), one could quantitate the layer thicknesses and the individual particle size as well as the distribution of the particles’ sizes on a wafer. The whole point of using a spun-on film is to eliminate the solvent and thus solvent effects, if any, on the particle size.

That withstanding, one must ask the question—How would one measure the size and/or the size distribution of nanoparticles, dispersed in a solvent such as the aforementioned nanoslurries? What happens if the nanoparticles remain in solution as opposed to being spun in to a film on a substrate? Can the “terahertz multispectral reconstructive imaging” (TMRI) protocol Rahman et al. (2016a, b) be effective for resolving these issues?

The imaging technique utilized in this report is based on a stable, multispectral, terahertz source, derived from a new mechanism called “dendrimer dipole excitation” (DDE) or the so-called Rahman-Tomalia (RT) effect (Rahman et al. 2017). Earlier reports have described in great detail the systematic engineering of “critical nanoscale design parameters” (CNDPs) manifested by either hard or soft nanoscale quantized building blocks (QBBs) (i.e., hard/soft superatoms) (Tomalia and Khanna 2016; Tomalia 2012) as a strategy for creating new emerging properties at the nanoscale level. Recently, this strategy was used successfully for synthesizing new dendrimer-based, high electro-optical coefficient substrates (Rahman 2010). These hyper-polarizable dendrimer emitters manifesting the RT effect were shown to produce high quality continuous wave (CW) terahertz radiation upon exposure to a suitable pump laser (Rahman et al. 2017). Modification of at least three dendrimer CNDPs, namely, size, surface chemistry, and interior compositions were required to produce these unprecedented dendritic, electro-optical emitters (Rahman and Rahman 2016); see Fig. 1 for a long-term stability of a DDE terahertz source. That withstanding, these DDE emitters have served as outstanding terahertz generators for the successful production of a new class of commercial dendrimer dipole excitation (DDE), time-domain THz spectrometers and other terahertz instruments. These unprecedented DDE-based THz spectrometers are currently produced by Applied Research & Photonics (ARP), Harrisburg, PA, USA, and have been shown to perform critical characterization and quantitation functions for a wide range of applications in the nanoscale world, which are described extensively elsewhere (Rahman et al. 2016a, b; Rahman 2017; Rahman et al. 2016b; Rahman 2011).
Fig. 1

Long-term (uncontrolled) stability of the source before applying to the scanner circuit. The spurious noise in the trace is due to physical disturbance

This brief account focuses on important applications resolved by DDE-based TMRI in the area of nanometrology. More specifically, it describes a facile protocol for real time, non-invasive characterization and quantitation of nanoparticle size/size distribution that may be measured in a variety of solvent environments.

Image formation by the terahertz multispectral reconstructive technique

The use of terahertz multispectral reconstructive imaging (TMRI) and terahertz time-domain spectrometry for characterizing semiconductor wafers and nanomaterials has been described elsewhere (Rahman et al. 2016a, b; Rahman 2017; Tomalia 2012; Tomalia and Khanna 2016). Firstly, TMRI offers an important opportunity to define pixel size (or a voxel size in 3D) by a hardware and software combination, as opposed to being limited by the image sensor chip such as the charge-coupled device (CCD). A digital camera, for example, displays and records the processed signal of an object that is focused on a CCD by means of a lens. The output of the CCD is processed by a built-in processor which displays the image and saves it in a file. In contrast, the reconstructive route eliminates the focusing lens and the CCD. Instead, the object to be imaged is scanned along the three orthogonal axes; the reflected signal (or, equivalently, the transmitted signal) is recorded in a data file and then processed by a suitable algorithm. The procedure for 3D image formation is described extensively elsewhere (Rahman et al. 2016a; Rahman 2017).

This account presents a protocol for measuring nanoparticle size and size distribution using the aforementioned TMRI 3D imaging while the nanoparticles remain dispersed in a solvent. More specifically, this report investigates silica nanoparticles dispersed in isopropyl alcohol (IPA), as a nanoslurry, wherein, preliminary nanoslurry properties are as described in Table 1.
Table 1

Description of silica nanoslurry


Particle size (nm)

SiO2 (wt%)


Viscosity (mPa.s.)

Specific gravity






< 1.0

< 15





7631-86-9. Ref:


The experimental equipment arrangement for characterizing a nanoparticle suspension (nanoslurry) using the TMRI protocol is as illustrated in Fig. 2. A short path-length cuvette was used for placing the nanoslurry suspension in the terahertz beam. A terahertz nanoscanning spectrometer and 3D imaging system (TNS3DI, Applied Research & Photonics, Harrisburg, PA) was used for the current measurements. More complete details of the instrumentation have been described elsewhere (Rahman et al. 2016b; Rahman 2017); however, the relevant technology for this application is briefly stated here. The TNS3DI apparatus is built around the DDE source as described in the context of Fig. 2.
Fig. 2

A small path-length cuvette is mounted on the nanoscanner. The nanosuspension (nanoslurry) is shown by an arrow. Total volume of solution is ~ 1 mL

As exhibited in Fig. 1, the source generates > 200 mW of CW power that is suitable for long-term operation. The output is very stable after the warmup time of ~ 2 h (Fig. 1). The source is capable of running continuously for several days at a time. This power is split into two arms for the built-in terahertz time-domain spectrometer implementation. The arms are termed as the pump arm that generates the terahertz via DDE emitter and the probe arm that detects the terahertz after being transmitted through or reflected from the sample. Both the terahertz source (DDE emitter) and the terahertz sensor are made of the same electro-optic dendrimer. The terahertz detection is carried out by the so-called electro-optic sampling technique, also known as pump-probing (Rahman 2011) technique.

Finally, both of the arms are combined and coupled into a multimode fiber for delivery to the nanoscanner. The nanoscanner deploys an optical circuitry for steering the beam and scanning an object for imaging via terahertz multispectral reconstructive technique described in Rahman et al. (2016b). As outlined in Fig. 2, the nanoslurry sample of the present investigation is kept in a short path-length cuvette. First, the empty cuvette is imaged and then the sample in the cuvette (nanoslurry) is imaged.

Results and discussion

The volume image and a surface image of the empty cuvette are as shown in Fig. 3, wherein, the bare cuvette exhibits an amorphous structure, typical of the plastic material composition.
Fig. 3

One cubic micron volume of the empty cuvette. Image exhibits characteristics of amorphous material

Figure 4 exhibits 3D images of the nanoslurry. Figure 4a is a 3D image of (1 μm)3 volume and b exhibits a single surface extracted from a. As seen from b, the nanoclusters are dispersed homogeneously within the solvent over a macro-scale. However, when zoomed to a nanoscale, the nanodomains become clearly visible (Fig. 5). Here, the nanoclusters are separated by the layers of solvent. The inter cluster separation is not fixed, rather a random distribution of the separation exists. Since the nanoslurry is a homogeneous body of silica nanoparticles in the solvent, the nanoparticle-solvent interaction is at saturation. However, since the nanoparticles have formed into nanoclusters, and since the nanoclusters are not of uniform size, hence, the nanoslurry of the present study is truly a homogeneous medium. There is no evidence of agglomeration of the nanoparticles. Typical single nanoparticle size was determined to be ~ 11 nm, which is within the range specified by the manufacturer (see Table 1; Fig. 6). Still a question may arise, “why the nanoparticles formed clusters that are polydisperse as opposed to being totally uniformly dispersed as individual nanoparticles?”, What determines the nanocluster size?
Fig. 4

a One cubic micron volume of nanoslurry inside the cuvette exhibiting homogeneous distribution of silica nanoparticles within the dispersant. b A single surface (1 μm)2 extracted from a. The nanoclusters are dispersed homogeneously within the dispersant over a macro-scale

Fig. 5

a A volume of (100 nm)3 of the nanoslurry extracted from Fig. 4a. b (100 nm)2 area of the nanosolution extracted from a exhibiting the nanoparticles and the nanoclusters dispersed randomly

Fig. 6

Graphical analysis of a single particle shows a typical silica nanoparticle is ~ 10.75 nm

The formation of clusters or aggregates in suspensions is well documented in the literature. It is known that when the suspension is mechanically stirred, simultaneous aggregation and breakup can occur. Work related to the microscopic nature of aggregation and breakup in such cases have typically centered on detailed analyses of the motion of individual particles or clusters as they approach other particles or clusters to form larger ones—a phenomenon observed in sedimentation, shear, and turbulent flows (Cohen 1992). In the present report, the particles and their nanoclusters are found to remain in suspension wherein the solvent medium prohibits agglomeration into micro structure. As such, nanocluster size formation depends primarily on the concentration of nanoparticles in a given suspension, (i.e., ~ 30% for the present sample (Table 1)), as well as their interaction with the solvent medium. As illustrated in Fig. 7, a typical nanocluster size was found to be ~ 17.5 nm.
Fig. 7

a An example of a typical nanoclusters in a 200 nm × 200 nm image and b profile of a single nanocluster in a shows the size of the nanocluster is ~ 17.5 nm


In this paper, we have provided some details of the terahertz reconstructive imaging technique for the determination of the size and distribution of nanoparticles in suspension. Here, a high power, continuous wave terahertz source based on the Rahman-Tomalia effect (aka dendrimer dipole excitation) is utilized for designing a terahertz nanoscanning spectrometer and 3D imager that is used for all imaging experiments. In particular, a suspension of silica nanoparticles in isopropyl alcohol was investigated. Presence of nanoclusters was imaged along with their distribution in the suspension. It was also found qualitatively that the volume fraction of solvent is significantly higher than that of the silica nanoparticles; an observation consistent with the composition of the nanoslurry of the present investigation. Although the manufacturer’s specification indicates a nanoparticle distribution of 10 to 15 nm, most individual particles were actually found to be within a 11- to 12-nm range. However, a typical nanocluster size was determined to be 17.5 nm, thus indicating the presence of nanoparticles less than 10 nm.


Compliance with ethical standards

This research has been conducted at Applied Research and Photonics, Harrisburg, PA by the initiatives of Dr. Anis Rahman and in consultation with Dr. Donald Tomalia.

Conflict of interest

The authors declare that they have no conflict of interest.


  1. Cohen RD (1992) Development of the cluster-size distribution in flowing suspensions. AICHE J 38:1129–1134CrossRefGoogle Scholar
  2. Hassellöv M, Readman JW, Ranville JF, Tiede K (2008) Nanoparticle analysis and characterization methodologies in environmental risk assessment of engineered nanoparticles. Ecotoxicology 17(344):344–361. CrossRefGoogle Scholar
  3. Rahman A (2010) Stimulated emission of terahertz radiation from electro-optic dendrimer, Proc. SPIE 7601, Terahertz Technology and Applications III, 76010C, DOI:
  4. Rahman A (2011) Dendrimer based terahertz time-domain spectroscopy and applications in molecular characterization. J Mol Struct 1006:59–65CrossRefGoogle Scholar
  5. Rahman A (2016) Simplest method of measuring nanoparticles’ size and size distribution on the surface and on sub-surface layers, Working Paper, December, pp. 1–5, DOI:
  6. Rahman A (2017) Terahertz multispectral imaging of epitaxially grown semiconductors’ lattice defects, In 28th Annual SEMI Advanced Semiconductor Manufacturing Conference (ASMC) 2017, Saratoga Springs, NY, pp. 45–50Google Scholar
  7. Rahman A, Rahman AK (2016) Terahertz spectroscopic analysis and multispectral imaging of epitaxially grown semiconductors with nanometer resolution. Journal of Biosensors & Bioelectronics 7(4):1–6. CrossRefGoogle Scholar
  8. Rahman A, Rahman AK, Yamamoto T, Kitagawa H (2016a) Terahertz sub-nanometer sub-surface imaging of 2D materials. Journal of Biosensors & Bioelectronics 7(3):1–6. CrossRefGoogle Scholar
  9. Rahman A, Rahman AK, Tomalia DA (2016b) Dendrimer dipole excitation: a new mechanism for terahertz generation. Journal of Biosensors & Bioelectronics 7(1):1–6CrossRefGoogle Scholar
  10. Rahman A, Rahman AK, Tomalia DA (2017) Engineering dendrimers to produce dendrimer dipole excitation based terahertz radiation sources suitable for spectrometry, molecular and biomedical imaging. Nanoscale Horiz 2:127–134CrossRefGoogle Scholar
  11. Tomalia DA (2012) Dendritic effects: dependency of dendritic nano-periodic property patterns on critical nanoscale design parameters (CNDPs). New J Chem 36:264–281CrossRefGoogle Scholar
  12. Tomalia DA, Khanna SN (2016) A systematic framework and nanoperiodic concept for unifying nanoscience: hard/soft nanoelements, superatoms, meta-atoms, new emerging properties, periodic property patterns, and predictive mendeleev-like nanoperiodic tables. Chem Rev 116:2705–2774CrossRefGoogle Scholar
  13. Wang XZ, Lande L, Rui FL, Tweedie RJ, Primrose K, Corbett J, McNeil-Watson FK (2009) Online characterisation of nanoparticle suspensions using dynamic light scattering, ultrasound spectroscopy and process tomography. Chem Eng Res Des 87(6):874–884CrossRefGoogle Scholar

Copyright information

© Springer Nature B.V. 2018

Authors and Affiliations

  1. 1.Applied Research & PhotonicsHarrisburgUSA
  2. 2.NanoSynthons LLCMt. PleasantUSA
  3. 3.Department of ChemistryUniversity of PennsylvaniaPhiladelphiaUSA
  4. 4.Department of PhysicsVirginia Commonwealth UniversityRichmondUSA

Personalised recommendations