U. S.

DEPARTMENT OF COMMERCE NATIONAL BUREAU OF STANDARDS

RESEARCH PAPER RP1653

Part of Journal of Research of the National Bureau of Standards, Volume 34,
May 1945

METHODS FOR THE EVALUATION OF ANALYTICAL

FILTER PAPERS
By B. W. Scribner and W. K. Wilson

ABSTRACT

As part of an investigation of analytical filter papers for th e purpose of assisting

in the development of standards of quality, improved methods for rate of flow of
water, retention of fine precipitates, and determination of ash were developed.
The use of complicated apparatus for measuring the rate of flow of water was
avoided by applying Darcy's law to the ordinary conical filter. The use of pre-
filtered water in making the measurement was found necessary for satisfactory
reproducibility. As the standard method for determining the ash content of
paper is not sufficiently accurate fo r filter paper, the method was modified i n
that a tare crucible is used and the test specimen wetted and pressed into a com-
pact wad to permit the use of as small a crucible as possible.
The ot her tests were bursting strength of wet paper, thickness, weight per unit
area, a lpha cellulose, copper number, and acidity (pH) . The last three named
are of importance relative to the purity of the cellulose and the stability of acid-
washed papers, as instances of deterioration of the cellulose of such papers have
been observed. Action of acid results in brittling of the papers and in the for-
mation of modified forms of cellulose that are soluble in some solutions.

CONTENTS
Page
I. Introduction _____ __ __ _ __ __ __ _ ___ __ __ __ ___ __ __ _ _. ___ _ _ ___ _ _ _ 453
II. Testing methods _ _ _ __ __ _ __ __ ____ __ ___ __ __ __ ____ ___ __ __ __ __ _ _ ___ _ _ 454
1. Retentiveness to fine precipitates_ __ _ __ __ ___ ____ __ ____ _ _ ____ 454
2. Flow of water __ __ _ _ __ _ ___ __ __ __ __ ___ __ __ __ __ _ __ __ __ _ _ _ __ _ 454
3. Content of ash_ __ _ _ __ _ ___ __ ___ ___ ___ __ __ __ ___ _ 456
4. Bursting strength of wet papeL _'____ _______ . _ __ _ __ . _ _ 457
5. Thickness, weight, and density ____ ___ . ___ _ _ __ ___ _ 457
6. StabilUy_____________________ ____ _____ 457

1. INTRODUCTION
As a part of an investigation at the National Bureau of Standards
to establish standards for analytical filter papers, testing procedures
for determination of their quality have been developed . It is believed
that the various tests are adequate for the evaluation of properties
of importance in ordinary analytical work. The t ests are being
applied to nearly all types and grades of papers, both foreign and
domestic, now available in this country, and experience has indicat ed
that the t ests are suitable for effecting a standardization of quality
for analytical filter papers. When additional testing is completed,
the issuance of a report dealing with quality standards is contemplated.
Improved methods were developed for testing for retention of
barium sulfate, rate of flow of water through the papers, and content
453
-
454 J owrnal of Researoh of the National Bureau of Standards

of ash. A method was also developed for determination of the burst-

ing strength of wet paper. The following additional testing methods
are used. These are standard methods of the Technical Association
of the Pulp and Paper Industry (T APPI), 1 and the Association's
designations of the methods are included.
Thickness (T 411 m) and weight in grams per square meter
(T 410 m). Density is calculated from these values.
Alpha cellulose (T 429 m), copper number (T 430 m) and acidity
as pH (T 435 m). These are related to the purity of the cellulose and
the stability of the papers.
Discussions and descriptions of the methods follow.
II. TESTING METHODS
1. RETENTIVENESS TO FINE PRECIPITATES

The use of a precipitate of barium sulfate is a time-honored practice

for testing filter papers for retentiveness to fine precipitates, and it
appears to be satisfactory for the purpose. However, it is necessary
that the following instructions be followed exactly: !I.>. ~
Dissolve 0.55 g of potassium sulfate in 275 ml of water and add 1.0
ml of concentrated hydrochloric acid. This solution contains about
0.1 g of sulfur. Heat to boiling, and while at the boiling temperature,
add slowly and with continuous stirring 25 ml of 5-percent barium
chloride solution. It is very important to have the specified con-
centration of acid and to keep the solution at the boiling temperature
during the precipitation. Allow to stand without agitation for not
less than 2 hours nor more than 5 hours at It temperature below the
boiling point and not less than 70° C. Prepare at least four circles of
the filter paper in the usual way for filtration, in 50-degree long-
stemmed funnels. Stir the precipitate until it is evenly suspended
in the liquid and filter approximately 50 ml of the mixture through
each circle of paper, collecting the filtrates in 250-ml Erlenmeyer
flasks. Swirl the filtrates to bring any sulfate to the center and view
from above against a black background for the presence of barium
sulfate. The experience at the Bureau is that 0.3 mg or less can be
detected in this manner. The paper is considered unsatisfactory if
any barium sulfate is seen in the filtrate.
2. FLOW OF WATER

A report on the rate of flow of water through filter paper has

already been published as a part of this investigation. 2
Various forms of apparatus were devised by other investigators,
but a satisfactory method was evolved by Bogaty and Carson which
does not require the use of special, complicated apparatus, and has
the advantage of simulating the manner in which filtrations are made.
The filter paper is folded in a cone in the usual way, taking care to
expel air pockets, and, to obviate any variable that might be intro-
duced by the use of a funnel, the cone is suspended freely in a wire
loop. The change in head and filtering area in the use of a cone
introduces no complications because, according to Darcy's law, the
1 Copies of the TAP PI methods can be obtained from the Association at 122 East 42d St., New York
17,N. Y.
, Bogaty, R.o and Carson, F . T ., Measurement of rate of flow of water through filter paper, J . Research
NBS 33, 353 (1944) RP1613.
 Evaluation of Filter Papers 455
rate of filtering of water for a given filter paper is proportional to the
area of the filter and to the head of water. With this relation as a
starting point it was found possible to derive a simple flow constant
applicable to a conical filter, which gives the speed of filtering from a
measurement of the time required to filter half the volume of water
remaining in the cone (starting with any convenient, lmown volume).
The water-filtration coefficient, k, is determined by means of the
equation
1.151 log ~
"k= V
t-t l
in which t-t l is the time interval required to decrease the volume of
water in t4e filter cone from VI to V. If the measurements are so
made that v=vI!2, and tl is taken as zero when the volume in the cone
is VI, k=0.3455 jt for any ~value of VI that may~be~chosen. In order to
avoid very small numbers, the rate of flow through the filter paper is
designated by the coefficient K, where K = 104lc=3455jt. K is the
number of milliliters of water filtered through a square meter of filter
paper in 1 second under a head of 1 centimeter of water.
A phenomenon exhibited by filter paper, and other porous materials,
is that ordinarily the rate of flow of liquids through them becomes
progressively slower with continued flow. As this led to some dis-
crepancies in results obtained for filter paper, considerable study was
made to find the cause of it and means of overcoming its effect. No
conclusive explanation as to cause WaS arrived at, and the same ap-
pears to be true of explanations offered by other inve!5tigators. How-
ever, the difficulty was traced to some unknown condition of water
used in making the test, as it was found that the change in rate can
be almost eliminated by prefiltration of the water. A description of
the method follows.
Prepare distilled water by prefiltering it through filter paper, the
rate of flow of which is at least as low as that of the paper to be tested .
Use two or more filters in series. For each kind of filter paper to b e
tested a volume of prefiltered water equal to three or four times the
volume of the cone will usually be sufficient. If a considerable quan-
tity of water is to be prepared, the filters should be renewed frequently,
because they lose their effectiveness with continued flow of water
through them.
Carefully fold a filter circle in the usual way to form a 50-degree
cone. Place it in a 50-degree funnel and fill it with water prepared
in the manner described. Press down the folds to expel all air pockets
and to make the three plies smooth and in good contact. Allow about
three-fourths of the water to filter through, then pour off the excess.
Save the water to wet the next specimen. Remove the empty cone
and suspend it freely over a burette by supporting the cone in a loop
of wire of such size that it will support the cone at about two-thirds
of the distance from its apex.
Use prefiltered water having a temperature of 23°±2°f C, and
pour into the cone, all at once, a lmown volume which is approxi-
mately two-thirds of the volume of the cone. When one-fifth of the
water has filtered through into the burette, start a stopwatch. When
half the remaining volume has filtered through, stop the stopwatch
and record the time in seconds. Test not less than 10 circles, using
456 J owrnal of Research of the National BU1'eau of Standards
the same water over and over and adding to it from the prepared
supply as found necessary.
The time required to filter half the volume remaining is an inverse
measure of the rate of flow, and is substantially the same for any
given filter paper, irrespective of the size of the test specimen. Con-
venient volumes of water initially poured in for filter circles of the
sizes most generally used are as follows:

Diameter of Initial volume

circle

em ml
11 25
9 15
7 7.5
5. 5 3.5

3. CONTENT OF ASH

The necessity for very accurate determination of ash content,

together with the small amount present, requires special modifications
of the standard method for testing paper. A large test specimen but
as small an ignition crucible as possible are required for good accu-
racy. These conditions are fulfilled by wetting the filter paper with
distilled water, thus permitting compact wadding of the paper for
conservation of space. The use of a tare crucible and special precau-
tions in the testing operations are necessary. The details of the
method are as follows: 3
Use 20-ml platinum crucibles with covers. A tare crucible with
a cover is carried through all operations exactly the same as the
crucibles used for the tests. The tare crucible must have practically
the same weight and surface area as the test crucible.
Heat the crucibles and covers in a muffle furnace at approximately
925° C. Cool in a desiccator and weigh to the nearest 0.1 mg. Dry
about 6 g of the paper to constant weight at 100° to 105° C and weigh
to the nearest 0.01 g. Wet the paper with distilled water containing
not more than 10 parts per million of igneous solids, then roll it into
a compact mass with clean fingers. Put the test specimen in a crucible,
cover it, and place it in a cold muffle furnace. Start the ignition and
after smoking has ceased remove the cover and continue the ignition
for 2 hours at 925 0 C. Replace the cover, cool and weigh the crucible
and ash. The ignition is considered to be complete when the weight
of the ash becomes constant after reignition for 30 minutes. All the
cooling must be done in a desiccator, and the test crucibles and tare
crucible must be allowed to remain on the balance pans the same
length of time. The ash must be protected from air currents at all
times.
Duplicate determinations shall be made, and they shall agree
within 0.003 percent of the weight of the dry paper. Report the
average of the values to the nearest 0.01 mg per ll-cm circle of the
paper and to the nearest 0.001 percent of the weight of the dry paper.
3 The authors acknowledge the assistance of Herbert F. Launer in the development of this method .
 Evaluation of Filter Papers 457
4. BURSTING STRENGTH OF WET PAPER

For determination of the strength of filter paper, resistance of

wetted paper to bursting appears to be a logical choice. With tbe
exception that the paper is tested after wetting it, the t e:;;t is made
according to the standard method of the Technical Association of the
Pulp and Paper Industry, T 03m. The paper is clamped in the
testing instrument over a circular orifice through which pressure is
applied by a compressed fluid until the paper breaks. The pressure
is recorded on a maximum-reading gauge and is reported as "points."
The procedure is as follows:
Clip together a pack of five circles of the filter paper, and immerse
the pack in distilled water at 23° C for 5 minutes. Remove the pack,
allow it to drain suspended for 10 seconds, then immediately determine
the bursting strength. Use enough packs to obtain not less than 10
values. Report the average bursting strength to the nearest one-
quarter point.
5. THICKNESS, WEIGHT, AND DENSITY

The thicknesses of at least 10 separate circles are determined with

a dial micrometer, and the average of the values is reported to the
nearest 0.0002 in. For determination of weight, a pack of not less
than 10 sheets is weighed on a chemical balance, and th e weight is
reported as grams pel' square meter to the nearest 0.1 g. Th e density
is calculated from the thickness and weight, as grams per cubic centi-
meter to the nearest 0.001 g.
6. STABILITY

Analytical filter paper as produced in the paper mill has a high

degree of cellulosic purity, which is a prime requisite for stability of
paper, and every care is taken to keep the paper as free as possible
from extraneous materials, as impurities, of course, would interfere with
analytical determinations. There should be no question, therefore,
of the stability of any of the various types of the papers except
those treated with acid to reduce the ash to a minimum or to harden
the papers. Acids are very deteriorative to cellulose, and unless the
treatment with the strong mineral acids is done very carefully,
degradation of the cellulose with resultant embrittlement of the paper
fibers may occur. Moreover, degraded cellulose is soluble in some
solutions, and therefore may pass into filtrates and cause erroneous
analytical results.
The Bureau has found in its studies 4 of the deterioration of papers
that a high content of alpha cellulose and a low copper number are the
best criteria of the purity of the cellulose of paper fib ers. Alpha is a
term used to designate the part of the cellulose that is insoluble in a
strong solution of sodium hydroxide under definitely prescribed
conditions. It is the pure or unmodified fraction of the cellulose. The
copper number designates the amount of copper precipitated by cellu-
lose under certain conditions from a solution of copper sulfate. It
represents the amount of unstable modified forms of cellulose present.
• A. E. Kimberly and B. W. Scribner, Summary report of National Bureau 01 Standards research on preser-
vationofrecords, Misc. PUb. NBS M154 (1937) .
458 Journal of Research of the National Bureau of Standards

The Bureau has published extensive data showing the close relation
of alpha cellulose, copper number, and acidity as hydrogen ion concen-
tration (pH), to the stability of various other kinds of papers. Some
information of this kind has been obtained for filter papers. For
example, a double acid-washed filter paper that had become brittle
during storage had only 72 percent alpha cellulose, a high copper
number of 3.3, and a pH value of 4.5. Comparable figures for good
paper of the same type and grade are alpha cellulose, 96 percent;
copper number, 0.4; and acidity, 6 pH. For good stability, the pH
value of paper should not be below 5.
In a few other instances noted at the Bureau, papers that had been
double acid-washed were so brittle that they burst when used for
filtration, and all of these were characterized also by a poor condition
of the cellulose. The bursting strength of some of these papers ranged,
for dry paper, from 4 to 14 points as compared with around 20
points for underteriorated paper. All the deteriorated papers had
been stored under good conditions, and some of them had not been in
stock very long.
In view of these experiences with the deteriorated papers, all of
which were made by manufacturers of papers that are usually of the
highest quality, the use of the recommended tests relating to stability
seems desirable for both consumer testing and manufacturing control
of acid-washed papers.
WASHINGTON, January 18, 1945.