A review of retention efficiency measurement techniques for sub-30nm liquid filtration

Suwen Liu, Haizheng Zhang, Jennifer Braggin, Entegris Inc., Billerica, MA, USA

Fluorescent quantum dots method measures pore size.

Yield improvements in semiconductor manufacturing are often driven by changes in materials and processes. As linewidths shrink, semiconductor manufacturers must consider all options to improve yield, including point-of-use liquid filtration schemes. Particulate matter suspended in liquids create contamination in semiconductor processes [1], driving the need for point-of-use filtration between the liquid source and the substrate surface. Based on the 2010 International Technology Roadmap for Semiconductors, the critical particle diameter is approaching 10 nm [2], and thus filter membranes must be capable of removing particles in this critical size range.

In this paper, several measurement techniques are compared when trying to determine the retention efficiency of sub-10 nm pore size ultrahigh molecular weight polyethylene (UPE) membranes. In addition to comparing commonly practiced techniques, this paper will also briefly introduce a new method of measurement, specifically the use of fluorescent quantum dots (QDs).

Retention efficiency

Filter membrane media have evolved over time to meet the demands of the semiconductor industry. Advances in membrane media materials, design, and manufacturing have made these membranes more retentive.

The retention efficiency of a liquid filter was historically determined by challenging the membrane with particles of a known size and concentration and measuring particles downstream of the membrane. However, due to the limitation in suitable challenge particles smaller than 10 nm and detectors for them, membrane pore size rating methods for sub-10 nm filters have not been reported. For such tight membranes, the pore-size ratings are usually estimated by bubble-point extrapolation techniques, instead of a particle challenge test [3]. An example of the extrapolation of pore size by bubble point measurement is shown in Figure 1.

Figure 1. Bubble point extrapolation of membrane pore size. [7].

While extrapolation of the pore size is possible, membrane manufacturers and end-users alike struggle to determine a method to directly measure pore size. It is important to review the characteristics of a test designed to measure the pore size and determine if any available techniques can meet this challenge.

Ideal membrane retention efficiency test

A good retention test method should contain two major components: well-defined challenging particles and a sensitive detector that can effectively find them. If a membrane manufacturer or end-user wanted to design the ideal membrane retention efficiency test, the following factors in Table 1 should be considered.

As there are many factors, it is impossible for one method to accurately represent the retention efficiency of every membrane. Therefore, several techniques should be considered and practiced.

Table 2 compares currently available methods for retention efficiency testing, specifically focusing on the use of an optical particle counter (OPC), fluorescence spectroscopy, ultrafine atomization/scanning mobility particle sizing (UFA-MPS), inductively coupled plasma mass spectroscopy (ICP-MS) with gold nanoparticles, and quantum dots.

Polystyrene latex (PSL) beads and OPC

For the last two decades, polystyrene latex (PSL) beads have played an important role in rating liquid filtration products. While this method has been very effective in the past, OPCs currently have a minimum detection limit of 30 nm, and thus are not effective at measuring the retention of a sub-10nm membrane. Another limitation of this method is that it is difficult to produce smaller PSL nanoparticles with a narrow size distribution.

Colloidal silica particles and UFA/SMPS

A new technique was developed recently that allows the measurement of the removal of particles as small as 20 nm diameter from liquids. In this technique, filters are challenged with particles sized ~20 nm diameter. Filter inlet and outlet concentrations are measured using UFA/SMPS [4–5]. Colloidal silica nanoparticles of 18 and 28 nm diam. were used in this technique. The 28nm-diam. colloidal silica has a relative narrow-size distribution. However, it is difficult to get sub-15 nm silica nanoparticles with a narrow distribution.

Fluorescent PSL beads and fluorescence spectrophotometer

Fluorescent (FL) PSL beads have also been developed as a series of fluorescent microspheres having red, blue, and green fluorescent colors in a range of sizes from 2µm down to 25 nm [ 6]. These fluorescent PSL beads can be easily detected by fluorescence microscopy and a fluorescence spectrophotometer. Entegris developed sub-30 nm particle retention tests by using 25 nm fluorescent PSL beads [7]. For this method, the primary limitation is not the detection sensitivity or accuracy of the fluorescence instrument, but the uniformity of the challenge particles. Thirty nanometer and smaller FL PSL beads tend to have a relatively wide-size distribution. Particles smaller than the rated diameter are more likely to be transmitted and to contribute an under-reported retention rating. Therefore, particle uniformity is critical to this method.

Quantum dot and its unique properties

Semiconductor nanocrystals, also called QDs, are artificial nanostructures that can possess varied properties, depending on their material, size, and shape. QDs have proven to be powerful fluorescent probes, especially for long-term, multiplexed, and quantitative imaging and detection [8–11]. For example, QDs can be excited by a single light source and function as broadly tunable fluorescence emitters. They also have exceptional photostability for continuous visualization. More importantly, the wavelength of the light from QDs is largely controlled by their size and material composition, and thus an entire family of distinct colors can be generated by the same material. Figure 2 shows the relations between the CdSe core size (from 2 to 7.7 nm) and related fluorescent emission peak and color (estimated).

Figure 2. Size-dependent properties: CdSe QD size and its related emission peak and color.

Because of the size-dependent photoluminescence tunable across the visible spectrum [12], CdSe (cadmium celenide) nanocrystals have become the most extensively investigated nanocrystals. This is due in large extent to the existence of a successful preparation method for high quality CdSe nanocrystals, which resulted in their commercial availability. The monodispersed size, spherical shape, and unique size-dependent fluorescent properties of CdSe make the QDs a good candidate for the challenge material for rating sub-10nm pore-size membranes.

Gold nanoparticles and ICP/MS

Gold colloidal nanoparticles are good candidates for challenge particles. In 2008, NIST released a set of gold nanoparticles (10, 30, and 60 nm) as reference materials (RMs) with a narrow-size distribution. Combined with the ICP-MS technique, gold nanoparticles were used for rating sub-30 nm filters [13]. However, there are cost concerns. The ICP-MS technique uses an expensive instrument, and the gold particles are expensive as well. In addition, citrtate-stabilized Au nanoparticles strongly interact with certain kinds of membranes even with selective protection ligans for specific membranes. The interaction can lead to nonseiving effects and generate misleading retention rating results.


Several commercially available techniques are viable for measuring sieving retention of advanced UPE membranes. However, each technique has benefits and drawbacks, as briefly described in Table 3.

Retention efficiency techniques should be carefully matched to the membrane and to the end-user's required results. Recently, Entegris has developed a new method that uses fluorescent quantum dots to challenge the tightest commercially available filters. The strong fluorescent properties of the quantum dots make it easy to detect relative pore size at relative low concentrations and at low cost. The Entegris filters, which were rated at 3, 5, and 10 nm using bubble point methods, were further confirmed by using this new QDs challenge test. QDs could be the next-generation challenge particles for small pore-size membranes.


1. Jae-Keun Lee, et al., "Latex sphere retention by microporous membrane in liquid filtration," Journal of the IES 36(1), 26-36 (1993).

2. 2010 ITRS, http://www.itrs.net/Links/2010ITRS/Home2010.htm.

3. Joseph Zahka and Donald Grant, "Predicting the performance efficiency of membrane filters in process liquids based on their pore-size ratings," Microcontamination 9(12), 23-29 (1991).

4. David Blackford and Don Grant, "A proposal for measuring 20-nm particles in high-purity water using a new technology, Ultrapure Water (January 2009).

5. Don Grant, "A new method for determining the size distribution of particles in colloidal suspensions," Particle Society of Minnesota annual meeting (March 19, 2008).

6. Stanley D, Duke, "Particle retention testing of 0.05 to 0.5 micrometer membrane filters," Thermal Scientific Technical Note, TN-020.03.

7. B. Shie, M. Ayubali, and Y. Xiao, "Sub-30nm particle retention test by fluorescence spectroscopy," Semicon China, March 17-19, 2009. 8. P. Alivisatos, "The use of nanocrystals in biogical detection," Nat. Biotechnol. 22(1) 47-52 (2004).

9. X. Gao, et al., "In vivo cancer targeting and imaging with semiconductor quantum dots," Nat. Biotechnol. 22, 969–976 (2004).

10.Suwen Liu, et al., "A new bioimaging carrier for fluorescent quantum dots: phospholipid nanowmulsion mimicking natural lipoprotein core," Drug Deliv. 13, 159-164 (2006).

11. M. Bruchez, et al., "Semiconductor nanocrystals as fluorescent biological labs," Science 281, 2013–2016 (1998).

12. M.A. Marcus, et al., "EXAFS studies of cadmium chalcogenide nanocrystals," Nanostruct. Mater. 1, 323-335 (1992).

13. T. Umeda, et al., "New filter rating method in practice for sub 30 nm lithography process filter," Advances in Resist Materials and Processing Technology XXVII, ed. Robert D. Allen, Proc. SPIE 7639 (SPIE, 2010) 76391F.

Suwen Liu is analytical chemist at Entegris Inc., Billerica, MA, email [email protected]. Haizheng Zhang is manager of the Analytical and Product Evaluation Labs and Jennifer Braggin is strategic applications technologist, both at Entegris.

Solid State Technology | Volume 55 | Issue 8 | October | 2012



Volume 55, Issue 8

Article Archive for Solid State Technology.

© 2012. PennWell Corporation. All Rights Reserved. PRIVACY POLICY | TERMS AND CONDITIONS