Introduction

We can all recognize intuitively that there must be a zeta potential value for a given papermaking system at which both process and physical properties are at optimum. Over a period of ten years laboratory investigation, done with Zeta Data technology and largely conducted by Professor Anatoly Makhonin, we have confirmed this.

In the course of conducting more than 5000 experiments, we have developed the methodology to continuously optimize process chemistry for almost any relatively clean system, as described in the following text. We define a "relatively clean" system as one that has a conductance in the neighborhood of 1500m S/cm or less.

In 1989 we introduced the first generation streaming potential instrument, called "The On-line Zeta Data", to the international paper industry. (It is now in its 6^th generation.) Simultaneously, we began doing process chemistry development experiments in the lab, using 30-liter stock samples. In 1993, following a suggestion from Professor Allan Springer, we added the drainage function. Zeta Data technology now measures zeta potential, drainage (Specific Filtration Resistance), conductance and temperature.

At this point we were so pleased with the function of the on-line Zeta Data that we decided to miniaturize the key functions and offer a Laboratory Zeta Data. This enabled us to reduce the size of a working stock sample to 1.5 liters. The following described methodology shows how we optimize wet end chemistry. The technical papers published on our Web Site: www.papermaking-chemistry.com demonstrate the substantial benefits derived from operating the headbox at the narrowly defined zeta potential that maximizes drainage.

Laboratory Zeta Data Process Chemistry Development Methodology

1. The first step in laboratory process chemistry development is to prepare stock. It can be in a consistency range anywhere from about 0.2-1.5%. We customarily select a consistency of 0.5% as being convenient for most chemistry experiments.
2. In doing process chemistry work of an academic nature, we tear a handsheet blotter into 25-30 pieces, fill a Waring Blender with tap water almost to the top, throw in the pieces and beat for five minutes at the highest speed.
3. In working on a specific problem with machine stock, we take a sample of thick stock prior to chemical addition, and dilute it with white water to headbox consistency.
4. This procedure is valid up to a conductance of about 1500 m S/cm, after which we lose precision in zeta potential data. At higher conductance, in order to obtain an acceptable standard deviation, at least ten data points can be downloaded to do the calculation in a spreadsheet such as Excel; or if working manually, one can take the median. On the machine, the Distributed Control System (DCS) can be programmed to output a trailing average.
5. The resultant fiber dispersion is decanted into a 2-liter heavy wall beaker to a level of about 1.5 liters. (It is convenient to put a black line on the outside of the beaker at the 1.5 liter level).
6. The beaker is placed under the Lab Zeta Data intake nozzle, and rests on a magnetic stirrer that can optionally be heated to machine temperature, if monitoring stock from the headbox. The chamber assembly is lowered into the beaker until the nozzle is about 1/2 inch from the bottom of the beaker. Positioning a collar with an Allen Wrench, on the vertical rod of the ring stand, enables the chamber position to be reproduced for subsequent experiments.
7. Power is turned on, and the "Continuous" button depressed. The Lab Zeta Data always starts with a discharge cycle, in order to flush out whatever may remain in the chamber from the previous experiment. It is our practice to discard data until we have ten successive zeta potential data points in good agreement, and then to begin collecting experimental data. It is a good idea to cut out a piece of thin, flexible plastic to cover the beaker and keep stock from splashing out on the discharge cycle.
8. Unlike streaming current measurements, which are sensitive to adsorption of contaminants on the measuring surfaces, the LabZD streaming potential hardware is essentially self-cleaning. Therefore, on the occasions when we have difficulty getting ten successive internally consistent data points, we have two choices. We can either continue the process because of the self-cleaning feature, or remove the chamber and clean the electrode surfaces with a Q-tip swab and acetone, as described in the Operating Manual. (We use the "Single Measurement" mode only under rare circumstances.)
9. The drainage measurement hardware consists in part of a transducer mounted in the chamber that measures the liquid level height. Software translates the value to volume, in cubic milliliters, and outputs data as "Drainage, ml". (It correlates with Specific Filtration Resistance, and is a more meaningful indicator of drainage than CSF or SR.) The operator can observe the actual height, as it increases during the experiment, from an LED display in inches.
10. All process and functional chemicals are diluted to a very low concentration prior to addition, e.g. .01%, in order to facilitate mixing. To further maximize mixing, we do not start adding chemicals until the LED readout indicates about 80% of the drainage is complete. This procedure facilitates mixing and thereby helps prevent raw chemical from coming in contact with the electrodes. Should this unhappy event transpire, it is necessary to either clean the electrodes and screen scrupulously with the Q-tip and acetone or let the LabZD run for as much as an hour or longer in order to self-clean.
11. In adding relatively low molecular weight, high charge density chemicals, such as polyamines or wet strength resin, the operator will observe a "kinetic kick". This takes the form of a large zeta potential shift towards the positive side in the second data point after chemical addition, followed by a decay towards the negative side during the remainder of the ten data points that we record between chemical additions. It is the second data point because the first data point, by design, does not reflect the most recent chemical addition.
12. Similarly, in chemistry optimization experiments, the second data point is recorded as the maximum drainage for that particular experiment. This is a kinetic parallel to zeta potential, because it is also followed by a progressive decrease as the microflocculation structure breaks down with agitation. We repeat the experiment with incrementally increasing amounts of chemical until we have enough data to determine the zeta potential that maximizes drainage. This describes how we determine the optimum zeta potential for a given stock.
13. Next we make handsheets to test for physical properties such as sheet ash, sizing and Scott Bond. (N.B. Traditional hand sheet molds are NOT appropriate. Please refer to the Web Site Newsletter of June, 1998 on this subject).
14. On the paper machine, the on-line Zeta Data guides the bumping of cationic scavenger feed rate to continuously maximize drainage. Regardless of variations in stock quality, this ensures continuous maximum process chemistry efficiency.
15. When evaluating the cost-efficiency of cationic scavengers, we add the chemical under test via burette, dropping it in very slowly to ensure thorough mixing before it sees the electrodes.
16. In screening cationic scavengers, the most effective scavenger for that particular stock is the one that gives the highest drainage when the zeta potential is in the optimum range.
17. We also take into consideration the amount of scavenger required to reach +5mV as a valid comparative measure of neutralization efficiency. We calculate the per cent on stock required, and record it as "cationic demand". The procedure is described in detail in a Web Site article entitled "Cationic Scavenger Evaluation".
18. An example of process chemistry optimization is shown on the Web Site www.papermaking-chemistry.com in the paper on "Process Chemistry Optimization, CFS".
19. A semi-final note: When doing the first experiment after an extended period of LabZD idleness, one should run continuous measurement cycles for as long as necessary to thoroughly wet the electrodes and meet our criterion of ten successive internally consistent zeta potential data points, before starting to collect experimental data. (It is OK to run the LabZD over-night with the beaker top covered to avoid splashing out stock.)
20. A final note: Experienced wet end chemists will recognize that conventional retention aids of high molecular weight, low charge density will have a negligible effect on the zeta potential; they are used to improve retention at the expense of formation.

John G. Penniman

Editor's Note: The Newsletter received minor revision on April 16, 1999 in an effort to simplify, clarify and eliminate possible ambiguity in discussing a complex subject.

JGP