Repeatability and Reproducibility Studies for the LA-950

Repeatability, reproducibility, and instrument to instrument agreement are important performance characteristics of any analytical instrument. This technical note presents multiple data sets that prove the excellent performance analysts should expect from the HORIBA LA-950 laser diffraction analyzer.

Introduction

Two studies are presented in this technical note; one performed at multiple HORIBA Application Labs using polydisperse standards, and one performed by a customer analyzing their own sample. Both studies tested the repeatability, reproducibility and/or instrument to instrument variation of the LA-950.

Definitions

The word precision is often used as a catch-all to describe the results from any kind of repeated test. Understanding the different types of precision is important because some tests are more diffi cult (and meaningful) than others.

• Repeatability - Measurement variation with a single operator and single instrument on the same sample, over a short amount of time with all other variables held constant (i.e. location). Think of this as taking a sampling, loading it into the LA-950, and taking three consecutive measurements without draining.
• Reproducibility - Measurement variation with either multiple operators on multiple instruments with the same sample (but possibly multiple lots) in multiple locations. Not all of these conditions must be satisfied. This is a much more taxing test than repeatability and is the test performed for this study. When a manufacturer makes a claim about precision, make sure to know which type.
• Intermediate Precision - Measurement variation with multiple operators on either single or multiple instruments, in the same location across multiple days.

The table on the next page summarizes how these three tests differ. This information appears courtesy of ASTM and can be found in ASTM E177, Practice for Use of the Terms Precision and Bias in ASTM Test Methods (1), and E456, Terminology Relating to Quality and Statistics (2).

* This situation can be different instruments meeting the same design requirement.
** Standard test method dependent, typically does not exceed one day.

HORIBA Study

A reproducibility study was performed on 40 unique, randomly selected LA-950 systems; 20 for wet measurements, 20 for dry measurements. Two NIST traceable polydisperse (range of sizes) glass bead reference samples were used in this study. The challenge samples were PS-202 (3-30 µm) and PS-215 (10-100 µm) from Whitehouse Scientific. The PS-202 sample was measured as an aqueous wet dispersion according to the method outlined in Analytical Test Method 102 (3). The PS-215 sample was measured as a dry powder using the PowderJet accessory according to the method outlined in Analytical test method 103 (4). The instrument settings used are shown below.

PS-202

Circulation: 3;
Agitation: 2;
Liquid level: LOW;
Refractive index: STD-GLASS BEADS (1.51-0.00i);
Distribution base: VOLUME;
Form of distribution: Manual (15 iterations);
Data acquisition time LD=5000, LED=5000

PS-215

Refractive index STD-GLASSBEADS (1.51-0.0i);
Distribution Base VOLUME;
Form of distribution Manual (15 iterations);
Data sampling times: LD=50000;
T% for Sampling ; Max T%:= 99%, Min T%= 95%;
Air pressure; 0.3 MPa (3 bar)

Figures 1 and 2 and Tables 1 and 2 show the results for the PS-202 wet measurements and PS-215 dry measurements.

Figure 1. Overlay of 20 wet results from 20 systems.

Figure 2. Overlay of 20 dry results from 20 systems.

Table 1. Results from 20 wet analyses on 20 systems.

Table 2. Results from 20 dry analyses on 20 systems.

Additional statistical information including graphs showing the 1 standard deviation errors bar are shown in Figures 3 and 4 and Tables 3 and 4.

Figure 3. Statistical analysis of 20 wet results on 20 systems.

Figure 4. Statistical analysis of 20 dry results on 20 systems

Table 3. Statistical analysis of 20 wet results on 20 systems.

Table 4. Statistical analysis of 20 dry results on 20 systems

ISO 13320:2009 (5) section 6.4 states that the coefficient of variation (CV %) should be less than 3% at the D50 and less than 5% at the D10 and D90 when testing reproducibility. In the context of the ISO document this pass/fail criteria refers to testing a single instrument. This study was performed across 20 different instruments and still exceeded the ISO guidelines.

Customer Case Study

An existing LA-910 user studied LA-950 performance when considering upgrading to the newer model. To begin the user used two LA-950 systems and one sample (Formulation 1) to test repeatability. The results for these studies are shown in Tables 5 and 6 with size expressed in nm. The CV% values are extremely low, partly due to the nature of the sample which was small, narrow, and easily dispersed, but also because of the high performance level shown by the LA-950 systems.

Table 5. Repeatability of sample Formulation 1 on LA-950 Unit 1

Table 6. Repeatability of sample Formulation 1 on LA-950 Unit 2

Impressed with this performance, the user purchased four LA-950 systems and investigated the instrument to instrument variation using two samples, Formulation 1 and Formulation 2. The results from these studies are shown in Tables 7 and 8. The user was satisfied with these results.

Table 7. Instrument to instrument variation across four LA-950 systems for Formulation 1.

Table 8. Instrument to instrument variation across four LA-950 systems for Formulation 2.

Conclusions

These studies mimic real-world conditions for many users who must reconcile results from multiple operators, units, and locations. This is particularly important for users with units across the world where the challenge of supporting across multiple time zones and languages grows quickly.

In this context the LA-950 Particle Size Analyzer data proves an excellent solution with superb data correlation for realistic (polydisperse) samples. In the HORIBA study this was proven across:

• 40 randomly selected units
• 2 locations
• 6 operators
• and acquired over 6 years (i.e. no drift)

This is accomplished without any unit-matching technique and at normal performance (i.e. no low sensitivity data processing). Such performance is unmatched on the market today.

In the customer study instrument to instrument agreement was proven across four systems measuring their own real world samples.

References and Further Reading

1. ASTM E177-10, Standard Practice for Use of the Terms Precision and Bias in ASTM Test Methods
2. ASTM E456-12, Standard Terminology Relating to Quality and Statistics
3. Analytical Test Method 102, Test Method for PS-202 Polydisperse Glass Bead Standards on Partica LA-950
4. Analytical Test Method 103, Setup of Automatic Dry Measurement Partica LA- 950 with PowderJet
5. ISO13320 Particle size analysis – Laser diffraction methods

This information has been sourced, reviewed and adapted from materials provided by HORIBA.

For more information on this source, please visit HORIBA.