[RASMB] measuring refractive index in the near IR, forlight scattering apps.

[John Philo]

Richard,

I just want to point out that if you are working in normal aqueous solutions
this is unlikely to really be a significant issue compared to the precision
of the measurements. First, as Mattia's formula reveals, the wavelength
dependence for the refractive index of water is rather small (the difference
between 589 and 830 nm is less than 0.4%). This is exactly why Wyatt and the
other manufacturers typically ignore the wavelength difference and simply
use the tabulated values at 589 nm in their software.

Further, the salts and other compounds in most common biological buffers
also have a rather small effect on the total refractive index. For example,
according to the Handbook of Chemistry and Physics, the difference in
refractive index between water and 0.5 M sodium chloride is less than 0.4%.

Thus I think in practice for aqueous solutions the bigger problem for the
absolute calibration of the diffusion coefficients is knowing the viscosity
with sufficient accuracy. And then too there are errors related to
contributions to the scattering from large particle contaminants and/or the
scattering from water and the salts.

If you are working in a mixed solvent or some other situation where the
calibration is unclear (refractive index and/or viscosity unknown), one
approach is to spike that solvent with a latex sphere size standard such as
those used to verify proper DLS function. The ratio of the apparent size of
the spheres compared to the true size (e.g. what you measure for them in
water) will give you a correction factor. This approach will probably
compromise the accuracy slightly (since it involves several measurements
each with some error), but it should work reasonably well.

John

-----Original Message-----
From: rasmb-bounces at rasmb.bbri.org [mailto:rasmb-bounces at rasmb.bbri.org] On
Behalf Of Mattia Rocco
Sent: Tuesday, February 06, 2007 6:08 AM
To: Richard Kingston; rasmb at rasmb.bbri.org
Subject: Re: [RASMB] measuring refractive index in the near IR, forlight
scattering apps.



Hi Richard,

we have faced the same problem here. The first solution was to apply the 
same wavelength dependence law to buffer solutions, i.e. we used the 
tabulated wavelength dispersion of water to derive an equation valid at 
20°C (see eq 1 in the attachment).
Then we derived an equation based on reading the refractive index using a 
standard Abbe refractometer illuminated by a sodium lamp, assuming that the 
dominating line is that at 589.3 nm (eq. 2, here set for 632.8 nm).

In 2003, we wanted to really check these equations, and possibly directly 
read the refractive index at some wavelengths. We contacted Edmund Optics, 
and Olly Simmons  (OllyS at edmundoptics.fsnet.co.uk) devised for us a 
contraption to illuminate the prism of a standard thermostatted Abbe 
refractometer with a single wavelength light, using a light source, fiber 
optic guide, lenses and interferential filters. Here's a list of parts 
required, for a total price of about 700 GBP in 2003:

54700 Model 20 Illuminator
38944 Fiber optic adapter
40639 12" Light guide ¼" diameter
38944 Fiber optic adapter
53047 C-thread light guide mounting adapter
54629 10mm extension barrel
54615 Simple lens mount 12mm
45084 PCX lens 12mm dia x 12mm Fl
54623 12.5mm Thick lens mount
43064 Interference filter 488nm
43077 Interference filter 589nm
43081 Interference filter 632nm
54622 12mm Thick lens mount
45384 PCV 9mm dia x -27mm Fl

We have assembled and used the contraption, and it seems to work fairly 
well. To determine the refractive index at other wavelengths, you can 
construct a curve using measurements at these three wavelengths and then 
extrapolating. Moreover, this direct measurement allows to work at 
different temperatures, without the need to also correct for the 
temperature dependence.

Hope this was helpful.

Best - Mattia




At 16:25 05/02/07 +1300, Richard Kingston wrote:
>Greetings all,
>
>I am seeking advice about measuring the refractive index of a solution,
>which crops up when analyzing light scattering data.
>
>Most who are familiar with the theory of dynamic light scattering
>will  know that the relationship between the measured auto-correlation 
>function and the diffusion constant  involves the refractive index of the 
>medium. To get a real number for the diffusion constant of a protein from 
>DLS, you need the refractive index. Yet the laser light sources used on 
>commercial instruments  (from Wyatt or the now assimilated DynaPro) often 
>operate in the near IR (e.g.  830-840 nm). Measuring the refractive index 
>here is non-trivial. Most lab refractometers are calibrated to give the 
>refractive index at 589 nm (The Sodium D line).
>
>The options seem to be ... in  order of increasing correctness and
>decreasing ease.
>
>1. Pretend your solution is water, and look the numbers up in a table.
>
>2. Measure the refractive index of the solution using a standard Abbe
>refractometer  and ignore the dispersion (wavelength-dependence).
>
>3. Directly measure the refractive index at the wavelength in question.
>
>Option 1 starts looking bad in some circumstances (e.g. when adding 
>salt
>or osmolyte to the solution)
>
>Option 2 is not difficult, but I'm not certain how large an error is
>introduced by  ignoring the wavelength dependence of the  refractive index 
>(say between 589 and 840 nM). Looking at data for pure solvents (water, 
>ethanol, glycerol) it seems like the error introduced could be small.
>
>Option 3 is made difficult by a lack of cheap, accessible 
>instrumentation.
>Recently I came across an interesting idea which would allow measurement 
>of refractive index of solutions in the near IR without the need for 
>expensive equipment. The reference is ...
>
>J Rheims, J Köser and T Wriedt (1997) Refractive-index measurements in 
>the
>near-IR using an Abbe refractometer. Meas. Sci. Technol. 8 601-605
>
>It's easy to find the full text of the article using Google. The basic
>concept is to replace the light source on a conventional Abbe 
>refractometer with an IR laser, and attach an IR sensitive CCD to the 
>refractometer eyepiece tube. Then the instrument must be re-calibrated to 
>make sense of the readings.
>
>This is where it becomes problematic for owners of old Bausch and Lomb
>Abbe-3Ls (surely the most common refractometer ever produced ?) The 
>re-calibration requires knowledge of the glass used in the refracting 
>prism. Yet generally, all you have to go on is the prism series number - 
>engraved on the prism, and on the left of the instrument scale. In my case 
>it reads "538". Although I have a copy of the Bausch and Lomb glass 
>catalog, the company long ago got swallowed by others, and I don't know a 
>way of telling what glass was used in a refracting prism of any given 
>series. That's likely true for most of the Bausch and Lomb Abbe-3Ls except 
>those in current manufacture. So this route is blocked.
>
>There doesn't seem to be a lot of discussion of the general problem in 
>the
>literature, which might indicate that measuring the refractive index in 
>the near IR is unimportant. Or it might not.  I'd appreciate any comments 
>from light scattering specialists or others who might have thought about 
>this issue.
>
>Thanks,
>
>Richard
>
>Richard Kingston, PhD.
>School of Biological Sciences
>The University of Auckland
>Private Bag 92019
>Auckland
>New Zealand.
>
>
>
>_______________________________________________
>RASMB mailing list
>RASMB at rasmb.bbri.org http://rasmb.bbri.org/mailman/listinfo/rasmb


Dr. Mattia Rocco
Proteomica
Istituto Nazionale per la Ricerca sul Cancro (IST)
c/o Centro per le Biotecnologie Avanzate (CBA)
IST c/o CBA, Largo Rosanna Benzi 10
I-16132 Genova, Italy

Phone: +39-0105737-310
Fax: +39-0105737-325
e-mail: mattia.rocco at istge.it