How to See and Accurately Measure a Nanoparticle

Speaker / Author: CW Louw Co-author(s): LH Adlem

NMISA Private Bag X34, Lynnwood Ridge, Pretoria, 0040

e-mail: wlouw@nmisa.org Phone: 012 841 4227

Abstract

Nano-scale materials and products either manufactured from or containing “nanoparticles” are now part of our daily lives. Computers, cellular phones and basically all microelectronic-based devices contain nano-materials, or are manufactured utilising single-atom-layer technology. Other everyday commodities such as transparent sunscreens, scratchproof eyeglasses, stain-repellent fabrics, products for gene therapy, etc. (the list is endless) were all made possible by utilising the special properties of nanoparticles.

But can we measure these particles accurately? Accurate dimensional measurements at the nano-scale are a challenge, not to mention measuring the atomic forces and the bio-toxicity effects.

The presentation explores the “beam technology” measurement techniques available at the NMISA to measure nanoparticles such as Scanning Electron Microscopy (SEM), Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) and X-ray Photoelectron Spectroscopy (XPS), shows results of measurements that were performed and then focuses on the traceability of these measurements to the SI (or internationally agreed references). The paper provides an example of a study to improve traceability of particle size distribution (PSD) analysis and deals with the results of an interlaboratory comparison on size characterization of SiO2 airborne nanoparticles using on-line and off-line measurement techniques.

1. Introduction

When engineering new products with “nanoparticles” or using the principles of epitaxial growth, it is the special properties of particles at the nanoscale that are utilised to create a material with special properties. Nanoparticles are effectively a bridge between bulk materials and atomic or molecular structures [1] and size-dependent properties are observed such as quantum confinement in semiconductor particles. The properties of materials change as their size approaches the nanoscale and as the percentage of atoms at the surface of a material becomes significant.

Nanotechnology is thus not simply working at ever smaller dimensions; rather working at the nanoscale enables scientists to utilise the unique physical, chemical, mechanical and optical properties of materials [2]. But the challenge is that by definition these particles are smaller than 100 nm, thus the scientist or engineer must be able to detect particles in the range of 1-100 nm and must be able to accurately measure the dimension and the special properties.

Figure 1. Illustration demonstrating the effect of the increased surface area provided by nanostructured materials [2]

Although the theory of manipulating atoms to build new structures with special properties was already established in the 1960s (epitaxial growth for example), scientists could only detect and measure the structure and properties of the engineered material with X-rays (X-ray diffraction) and optical techniques. In the 1980s and 1990s the development of powerful electron microscopes, able to resolve at the nanometre scale made it possible for scientists to visually see clusters of nanoparticles and the development of techniques such as atomic force microscopy made it possible to measure their atomic forces and to manipulate individual particles.

Today the visualisation of nanometre scale features and particles are routine, but accurate measurement at this scale remains a challenge. Particle size analysers applying laser technology can provide mean particle size, but one of the most common techniques used to characterise the size of groups of or individual particles remains the electron microscope, both in transmission mode (TEM) and surface scanning mode (SEM). Other surface sensitive techniques such as X-ray Photoelectron Spectroscopy (XPS), Auger Electron Spectroscopy (AES), Secondary Ion Mass Spectrometry (SIMS) and Atomic Force Microscopy (AFM) can then be applied to obtain further information such as binding energies (chemical bonding), elemental composition, atomic forces, coverage, distribution, etc.

This paper will report on recent participation of the NMISA in the accurate analysis of nano- particles, that was part of a study performed within the framework of Technical Working Area (TWA) 34 – “Properties of Nanoparticle Populations” of the Versailles Project on Advanced Materials and Standards (VAMAS) in project 3 “Techniques for characterizing size distribution of airborne nanoparticles” [4].

2. Experimental The objectives were to generate airborne SiO2 nanoparticles and to characterise the particle size distribution (PSD) of aerosols with different on-line measurement techniques, and to validate them by an interlaboratory comparison. The sample preparation is described in [4]

and a short summary is given below of the procedures applied by the pilot laboratory. Although not performed by the NMISA, it is deemed important to provide the reader with an understanding of the processes used to prepare such samples. The SEM analysis at the NMISA was performed according to the protocol provided by the pilot laboratory.

Taking into account differences in instrumental performances between the microscopes used, only three relevant constraints were imposed: (1) accelerating voltage: a range of 1 to 30 kV was chosen. Depending on equipment, the high-resolution low-voltage operation mode was preferred to the conventional high-voltage mode, with the aim of avoiding electrical charging effects that necessitate coating with a thin conductive film, (2) a magnification between x250000 and x300000 was chosen. However, it was possible to choose another magnification between x50000 and x300000 depending on the individual constraints of laboratories to take images of about 500 or 1000 nanoparticles, (3) the image magnification calibration was performed with nanoparticle Certified Reference Materials having a mean diameter similar to those of the SiO2 nanoparticles. Such CRMs of nanoparticles in aqueous suspensions were distributed to all participants: (i) Polystyrene Latex Spheres (PSL) with mean diameters of 81 ± 3 nm certified by TEM (3080A, Thermo Fisher Scientific), respectively, (ii) two reference materials of gold nanoparticles with mean diameters, measured using SEM, of 26.9 ± 0.1 nm and 54.9 ± 0.4 nm (RM 8012 and RM 8013, NIST) which do not require a conductive coating. Sample preparation for SEM analysis After particle collection on PMF and before SEM analysis, a thin film of Au, Ag, Pt or Pd was sputter deposited in order to ensure an electrically conductive surface and to avoid surface charging under electron bombardment at high accelerating voltages. Both the PSL calibration and the airborne particle measurements (aerosol OP and DP) were performed under the same conditions to enable comparison of the measurement results with and without coating. This constitutes one way of estimating the measurement uncertainties due to the applied coating. SEM measurement protocol In order to check the uniformity of the particle collection over the whole sample area, a preliminary survey of the PMF sample with the deposited nanoparticles was performed. It was determined that the average distance between neighbouring particles should not be shorter than their average diameter, in order to avoid overlaps which make the correct identification impossible. A magnification of x50000 to x300000 was chosen according to Figure 1.

Figure 1: Working diagram for optimal magnification as a function of the particle diameter For aerosol OP, about 500 particles were measured, while about 1000 particles were measured for aerosol DP. It is important to avoid overlaps between the scanned areas of different SEM microphotographs and it was necessary for the scale bar to be visible on each microphotograph. If coating was used, a thickness correction including measurement uncertainties was applied. SEM measurement uncertainty evaluation

Contributions of the image magnification calibration uncertainty as well as of the coating thickness were taken into account for the evaluation of the measurement uncertainty. Selection of threshold in the greyscale images acquired for an accurate delimitation of the nanoparticles was also implemented as a decisive importance in the evaluation of the diameter measurement uncertainties. It is important to note that the data reduction software required the (re)definition of the pixel size of each particular micrograph.

SEM Results As stated in the section “Sample Preparation for SEM analysis”, it was necessary either to apply a thin conductive coating on the specimen surface to suppress electrical charging of the surface when working at higher accelerating voltages, or to work at low accelerating voltages below 5 kV, due to the fact that the specimens chosen were electrically non-conductive. The classical, first option was used for purposes of inspection of the specimen morphology. The second one was used only at the rather modern SEM where this low-voltage operating mode was available. Most commercial sputter coaters suggest coating thickness values which shall be reached when the coating applied runs under well-defined, recommended controlled conditions. Nevertheless, it was also stated that these given thickness values are rough estimates, so that a calibration of an accompanying witness specimen (silicon wafer) and separate traceable measurement of the coating thickness must be undertaken additionally.

1

10

100

1000

10000

0 50 100 150 200

Expected particles diameter (nm)O

ptim

al M

ag (

x100

0)

Figure 2 : SEM micrographs of aerosol OP specimen of type C (PMF as a support): coated with 10 nm Au (left); note the “cauliflower-like structure” of the Au coating altering the real size of the nanoparticles, and uncoated specimen sampled by the same laboratory in the low-beam voltage mode (right). Au, Pt or Au/Pd have been used as coating materials in this interlaboratory study. When the low-sputtering mode was used, the coating mean grain size reached typically about 2 nm for Au and about 1.7 nm for Au/Pd. The resulting well-known irregular cauliflower-like structure is clearly visible in Fig. 2 (left) compared to uncoated samples Fig. 2 (right). The particle coating thickness was not necessarily the same than the one obtained on a flat surface. Therefore, accurate determination of the applied coating layer thickness was a challenging task which generates the most significant uncertainty contribution to the size measurement that may even exceed 10 nm. 3. Results

Table 1 - SEM operating parameters used during the round-robin test

Operating Parameters Type of SEM LEO 1525 FE-SEM Accelerating Voltage (kV) Varied, but mostly 8 kV Magnification Varied, typically a set at 100 000 x and a set at 50 000 x Image Processing System Raw tiff files were used Image Analysis Software ImageJ. Thresholding with Otsu, made binary, refine on circularity and size Other, like the coating Au/Pd coating Particle size measurement data Table 2: Summary of results of particle size measurements (nm)

30 nm Au 60 nm Au K25 F1 K25 F2 K25 F3 K50 F1 K50 F2 K50 F3

Mean 26.2 57.1 46.1 42.6 77.1 44.6 66.9 77.7 Standard Deviation 4.843 5.357 11.640 15.424 27.221 12.624 17.068 26.989

Count 110 175 900 2903 400 1753 190 822 Expanded uncertainty * 3.60 1.63 1.92 1.64 3.59 1.65 3.76 2.57 * at 95% confidence level, U=k*uc (%) with k=2 (coverage factor)

Comparison of the NMISA results with that of the other Participants Figures 3 and 4 show the comparison of the NMISA results to that of the other participants. (the NMISA participation number is SEM4).

Figure 3: SEM results of the measurement of the mean diameters for aerosol OP samples of each measurement laboratory associated with the sampling laboratory (SMPS2, SMP3 and SMPS4). Error bars indicate the reproducibility and the expanded uncertainty k=2). The solid line represents the total average within a band of ± two interlaboratory standard deviations (dash lines). Filled and empty squares represent measurements with and without coating, respectively.

(a)

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6 7 8

Averaged mean diameter

Averaged mean diameter ± 2 x SD

Mea

n d

iam

eter

(nm

)

SEM6 SEM7 SEM3 SEM5 SEM4 SEM3 SEM7

With coating

Without coating

SMPS4 SMPS2 SMPS3

Sampling laboratory

Reproducibility uncertainty (k = 2 )Expanded uncertainty (k = 2)

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6 7 8 9 10 11

Averaged mean diameter

Averaged mean diameter ± 2 x SD

Mea

n d

iam

eter

(nm

)

SEM6 SEM7 SEM3 SEM1 SEM5 SEM4 SEM1 SEM2 SEM3 SEM7

SMPS4 SMPS2 SMPS3

Sampling laboratory

With coating

Without coating

Reproducibility uncertainty (k = 2 )

Expanded uncertainty (k = 2)

(b)

Figure 4: SEM results of the mean diameters of both aerosol DP samples ((a) first population; (b) second

population) of each measurement laboratory associated with the sampling laboratory (SMPS2, SMP3 and

SMPS4). Error bars indicate the reproducibility SD and expanded uncertainty, (k=2). The solid line represents

the total average within a band of +/- two interlaboratory standard deviations (dash lines). Filled and empty

squares represent measurements with and without coating, respectively.

0

20

40

60

80

100

120

140

0 1 2 3 4 5 6 7 8 9 10 11

Averaged mean diameter

Averaged mean diameter ± 2 x SD

Mea

n d

iam

eter

(nm

)

SMPS4 SMPS2 SMPS3

Sampling laboratory

With coating

Without coating

Reproducibility uncertainty (k = 2 )

Expanded uncertainty (k = 2)

SEM6 SEM7 SEM3 SEM1 SEM5 SEM4 SEM1 SEM2 SEM3 SEM7

4. Conclusion

The NMISA results compares well with that of the other laboratories and fall within a band of ± two interlaboratory standard deviations. The NMISA thus successfully participated in the interlaboratory comparison.

The study provided a valuable learning opportunity to the NMISA and contributed to the confidence of the laboratory in performing accurate particle size analysis at the nanoparticle size level.

The comparison of the SEM results with the other techniques can be obtained from the VAMAS publication [4].

5. References

1. Science Daily - http://www.sciencedaily.com/articles/n/nanoparticle.htm 2. National Nanotechnology Initiative - http://www.nano.gov/nanotech-101/special 3. Applications of nanoparticles - International Journal of Nanotechnology, 6 (2009) 567 4. Motzkus, C. et al., “Size characterization of airborne SiO2 nanoparticles with on-line and

off-line measurement techniques: an interlaboratory comparison study, (accepted in) Journal of Nanoparticle Research, J Nanopart res (2013) 15:1919