[Federal Register Volume 73, Number 121 (Monday, June 23, 2008)]
[Notices]
[Pages 35395-35402]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: E8-14112]


-----------------------------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Centers for Disease Control and Prevention


Notice Regarding Revisions to the Laboratory Protocol To Measure 
the Quantity of Nicotine Contained in Smokeless Tobacco Products 
Manufactured, Imported, or Packaged in the United States

AGENCY: Centers for Disease Control and Prevention (CDC), Department of 
Health and Human Services.

ACTION: Notice and request for public comment.

-----------------------------------------------------------------------

SUMMARY: The uniform protocol for the analysis of nicotine, total 
moisture, and pH in smokeless tobacco products, originally published in 
the Federal Register in 1999 (64 FR 14086), ``Notice Regarding 
Requirement for Annual Submission of the Quantity of Nicotine Contained 
in Smokeless Tobacco Products Manufactured, Imported, or Packaged in 
the United States,'' and revised in the Federal Register on March 14, 
2008 (73 FR 13903), implements the requirement of the Comprehensive 
Smokeless Tobacco Health Education Act (CSTHEA) of 1986 (15 U.S.C. 4401 
et seq., Pub. L. 99-252) that each entity manufacturing, packaging, or 
importing smokeless tobacco products shall annually provide the 
Secretary of Health and Human Services (HHS) with a specification of 
the quantity of nicotine contained in each smokeless tobacco product. 
CDC is re-publishing the notice published in the Federal Register on 
March 14, 2008 (73 FR 13903) concerning the revision of the protocol 
for analysis of nicotine in smokeless tobacco products (hereinafter 
referred to as ``Protocol'') to (1) make a technical change to correct 
the date when the first report of information under the revised 
Protocol is due; (2) solicit public comments concerning a change in the 
Protocol that increased the volume of water in the pH determination 
from 10 mL to 20 mL, and (3) solicit public comments concerning the 
addition of the following commercial smokeless tobacco product 
categories: Dry snuff portion packs, snus, snus portion packs, and 
pellet or compressed.
    The Protocol as published in the Federal Register on March 14, 2008 
(73 FR 13903), remains in effect with the technical correction to the 
date described below.
    Technical change: The language in the March 14, 2008 notice stated 
that ``The first report of information is due June 30, 2008, with 
subsequent submissions due by March 31 of each year.'' The first report 
date of information should be 2009 so that the sentence correctly 
reads: ``The first report of information is due June 30, 2009, with 
subsequent submissions due by March 31 of each year.''

DATES: Written comments concerning the change in the volume of liquid 
in the pH determination and the addition of four commercial smokeless 
tobacco product categories must be received on or before July 23, 2008.

ADDRESSES: Comments should be marked ``Comments on Revised Protocol for 
Analysis of Nicotine'' and mailed to the Centers for Disease Control 
and Prevention, National Center for Chronic Disease Prevention and 
Health Promotion, Office on Smoking and Health, Attention: Matthew 
McKenna, M.D., Director, 4770 Buford Highway NE., MS K-50, Atlanta, 
Georgia 30341-3724. Comments may be e-mailed to: [email protected].

FOR FURTHER INFORMATION CONTACT: Matthew McKenna, M.D. Director, Office 
on Smoking and Health, Telephone: (770) 488-5701.

SUPPLEMENTARY INFORMATION: Several smokeless tobacco product categories 
have entered the U.S. smokeless tobacco market since the implementation 
of the protocol in 1999 including snus, low moisture snuff sold in 
portion pouches, and smokeless tobacco sold in a compressed, pellet 
form. Some of the new smokeless tobacco product categories differ 
physically from previous smokeless tobacco categories.
    After evaluating information that has recently come to the 
attention of the Centers for Disease Control and Prevention's Office on 
Smoking and Health (OSH) regarding low moisture smokeless tobacco 
products packaged in portion pouches, OSH conducted an independent 
comparison of pH measurements in a variety of low and high moisture 
smokeless tobacco products. The results of this comparison, presented 
in Table 1, indicate that there is an acceptable (less than 2%) level 
of change in pH values when measurements are taken with 20 mL 
deionized, distilled water (Condition B) compared to 10 mL of 
deionized, distilled water (Condition A). Increasing the volume of 
water in the mixture ensures that the matrix is sufficiently fluid to 
facilitate ease of measure.

[[Page 35396]]



 Table 1.--Summary of pH Levels Found in Seven Types of Smokeless Tobacco Products: Plug; Loose Leaf or Scrap; Twist; Dry Snuff--Low Moisture/No Pouch;
                                                  Dry Snuff--Low Moisture/Pouch; Snus; and Moist Snuff
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                                pH \a\
                                                                             ---------------------------------------------------------------------------
                  Category                       Smokeless Tobacco Product      Condition A 10 mL \b\      Condition B 20 mL \b\
                                                                             ------------------------------------------------------ pH Change   % Change
                                                                               Mean \c\         SD \d\    Mean \c\         SD \d\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Plug........................................  Days O Work Chew..............       5.06                          us>
                                              Conwood Company's Sun Cured...       5.12                          us>
                                              Levi Garrett Plug Chew........       5.83                          us>
                                              Taylor's Pride Plug Chew......       5.92                          us>
Loose Leaf..................................  Beech-Nut Chew................       5.56                          us>
                                              Redman Chew...................       5.93                          us>
Twist.......................................  Cumberland....................       5.68                          us>
Dry Snuff/No Pouch..........................  Tube Rose Sweet Scotch Snuff..       5.64                          us>
                                              RailRoad Mills Sweet Scotch          5.91                          us>
Dry Snuff/Pouch.............................  Taboka........................       6.44                          us>
                                              Skoal Dry Cinnamon............       6.78                          us>
Snus........................................  Camel Snus Original...........       7.43                          us>
Moist Snuff.................................  Renegades Wintergreen.........       6.45                          us>
                                              Copenhagen Regular............       7.61                          us>
                                              Kodiak Ice Long Cut Regular...       8.13                          us>
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ The standard protocol published in the Federal Register to measure pH in smokeless tobacco products is as follows: 10 mL of deionized distilled
  water is added to 2.00 grams of smokeless tobacco product measuring pH at 5, 15, 30 and 60 minute intervals. Recently introduced low moisture dry
  snuff smokeless tobacco products packed in pouches had a thick paste-like consistency when prepared in 10 mL of deionized distilled water. When 2.00
  grams \e\ of low moisture dry snuff smokeless tobacco products packed in pouches were prepared in 20 mL of deionized distilled water, the sample
  remains suspended in liquid and is well mixed.
\b\ n = 1.
\c\ Average pH from four measured intervals.
\d\ Standard Deviation.
\e\ Accurately weighed: 2.000  .0005 grams.

    OSH has determined that these revisions will improve the 
applicability of the protocol and provide guidance to reporting 
entities and other interested parties for testing of all currently 
marketed categories of smokeless tobacco. The change in the volume of 
liquid in the pH determination facilitates the ease of measure of 
smokeless tobacco pH for all currently marketed smokeless tobacco 
categories (i.e., plug, twist, moist snuff, dry snuff, snus, loose 
leaf, chew, moist snuff in portion pouches, smokeless tobacco 
compressed into a pellet, and dry snuff in portion pouches).

Collection of Information

    This proposed amendment does not call for any new collection of 
information under the Paperwork Reduction Act of 1995 (44 U.S.C. 3501-
3520).

    Dated: June 13, 2008.
James D. Seligman,
Chief Information Officer, Centers for Disease Control and Prevention.

Revised Protocol for Analysis of Nicotine, Total Moisture, and pH in 
Smokeless Tobacco Products

I. Requirements \1\ \2\

A. Reagents \3\

    1. Sodium hydroxide (NaOH), 2N.
    2. Methyl t-butyl ether (MTBE).
    3. (-) -Nicotine (Fluka 72290) >99% purity.4 5
    4. Quinoline (Aldrich).
    5. Standard pH buffers; 4.01, 7.00, and 10.00.
    6. Deionized distilled water.

B. Glassware and Supplies

    1. Volumetric flasks, class A.
    2. Culture tubes, 25 mm x 200 mm, with Teflon-lined screw caps.
    3. Pasteur pipettes.
    4. Repipettors (10 mL and 50 mL).
    5. Linear shaker (configured to hold tubes in horizontal 
position).6 7
    6. Weighing dishes, aluminum.
    7. Teflon-coated magnetic stirring bars.
    8. Polypropylene containers, 50 mL.

C. Instrumentation

    1. Robot Coupe Model RSI 2V Scientific Batch Processor.
    2. Capillary gas chromatograph, Hewlett Packard, Model 6890, with 
split/splitless injector capability, flame ionization detector, and a 
capillary column (Hewlett Packard HP-5, Crosslinked 5% PH ME Siloxane, 
30 m length x 0.32 mm ID, film thickness 0.25 or 0.52 [mu]m).
    3. Orion Model EA 940 pH meter equipped with Orion 8103 Ross 
combination pH electrode.

D. Additional Equipment

    Forced-air oven, Fisher Isotemp[supreg], regulated to 99  1.0 [deg]C. Suggested dimensions: 18 x 18 x 20 inches.

E. Chromatographic Conditions \8\ \9\

    1. Detector temperature: 250 [deg]C.
    2. Injector temperature: 250 [deg]C.
    3. Flow rate at 100 [deg]C--1.7 mL/min; with split ratio of 
40:1.\10\
    4. Injection volume: 2 [mu]l.
    5. Column conditions: 110-185 [deg]C at 10 [deg]C min-1; 
185-240 [deg]C at 6 [deg]C min-1, hold at final temperature 
for 10 min.

F. Sample Preparation \11\

    There are ten different categories of commercial smokeless tobacco 
products:
    1. Dry snuff;

[[Page 35397]]

    2. Moist (wet) snuff;
    3. Moist (wet) snuff portion packs;
    4. Plug;
    5. Twist;
    6. Loose leaf;
    7. Dry snuff portion packs;
    8. Snus;
    9. Snus portion packs; and
    10. Pellet or Compressed.
    Because of their physical characteristics, some of the ten product 
categories must be ground (whole or in part) before nicotine, total 
moisture, and pH analyses can be conducted. The objective of grinding 
the samples is to obtain a homogeneous sample with particles measuring 
approximately 4 mm. Grinding to achieve this particle size should take 
no more than 3 minutes. To ensure proper grinding and an adequate 
amount of the ground sample for analysis, the minimum sample size of 
all commercial products to be ground should not be less than 100 grams.
    To ensure precision of analyses for nicotine, total moisture, and 
pH, the samples that require grinding should be ground using a Robot 
Coupe Model RSI 2V Scientific Batch Processor or its equivalent. This 
is a variable speed (0 to 3000 RPM) processor. The variable speed motor 
is required to ensure proper grinding of the tobacco tissues (and in 
the case of pH determination, the portion pack). Elevated temperatures 
can result in moisture loss and an underestimated value for moisture 
content. Hence, care must be taken during grinding to avoid elevated 
temperatures. The bowl should be cleaned after each grinding to obtain 
accurate results. Freeze- or cryo-grinding is also an acceptable 
grinding method.
    1. Dry snuff: Dry snuff samples do not need to be ground since the 
product is a powder. The sample must be thoroughly mixed before 
weighing for nicotine, total moisture, and pH analysis.
    2. Moist (wet) snuff: Moist (wet) snuff samples do not need to be 
ground. The sample must be thoroughly mixed before weighing for 
nicotine, total moisture, and pH analysis.
    3. Moist (wet) snuff portion packs: The tobacco contents of the 
moist (wet) snuff portion packs do not need to be ground for nicotine, 
total moisture, or pH analysis. The tobacco packaging material (the 
``pouch'') should be separated from the tobacco and ground to obtain 
particles measuring approximately 4 mm for pH analysis. The tobacco of 
the moist (wet) snuff portion pack and the ground pouch are combined 
and thoroughly mixed before pH analysis.
    4. Plug tobacco: Break or cut apart plugs and add in portions to 
grinder at 2000 RPM. Reduce RPM or stop grinding if sample bowl becomes 
warm. Pulse the Robot Coupe, when needed, to complete grinding. Grind 
samples until approximately 4 mm in size. The total grinding time 
should be no more than 3 minutes.
    5. Twist tobacco: Separate twists, add to grinder and grind at 2000 
RPM. Reduce RPM or stop grinding if sample bowl becomes warm. Continue 
grinding until sample particles are approximately 4 mm in size. The 
total time for grinding should be no more than 3 minutes.
    6. Loose leaf: Grind in the same manner as described in 4 and 5 to 
obtain product with particle size of approximately 4 mm.
    7. Dry snuff portion packs: The tobacco contents of the dry snuff 
portion packs do not need to be ground for nicotine, total moisture, or 
pH analysis. The tobacco packaging material (the ``pouch'') should be 
separated from the tobacco and ground to obtain particles measuring 
approximately 4 mm for pH analysis. The tobacco of the dry snuff 
portion pack and the ground pouch are combined and thoroughly mixed 
before pH analysis.
    8. Snus: Snus samples do not need to be ground since the product is 
a powder. The sample must be thoroughly mixed before weighing for 
nicotine, total moisture, and pH analysis.
    9. Snus portion packs: The tobacco contents of the snus portion 
packs do not need to be ground for nicotine, total moisture, or pH 
analysis. The tobacco packaging material (the ``pouch'') should be 
separated from the tobacco and ground to obtain particles measuring 
approximately 4 mm for pH analysis. The tobacco of the snus portion 
pack and the ground pouch are combined and thoroughly mixed before pH 
analysis.
    10. Pellet or compressed: Break apart compressed tobacco pellets 
and add in portions to grinder at 2000 RPM. Reduce RPM or stop grinding 
if sample bowl becomes warm. Pulse the Robot Coupe, when needed, to 
complete grinding. Grind samples until approximately 4 mm in size. The 
total grinding time should be no more than 3 minutes.

II. Nicotine Analysis \12\

A. Calibration Standards

1. Internal Standard (IS)
    Weigh 10.00 grams of quinoline, transfer to a 250 mL volumetric 
flask and dilute to volume with MTBE. This solution will be used for 
calibration of the instrument for the nicotine calibration curve 
(II.A.2), for the standards addition assay (II.B), and for preparation 
of the extracting solution (II.D).
2. Nicotine Calibration Curve
    a. Weigh 1.0000 gram of nicotine into a clean, dry 100 mL 
volumetric flask and dilute to volume with MTBE. This gives a nicotine 
concentration of 10 mg/mL for the stock solution.
    b. Accurately pipette 0.5 mL of IS from stock solution (II.A.1) to 
five clean, dry 50 mL volumetric flasks. To prepare a nicotine standard 
corresponding to a concentration of 0.8 mg/mL, pipette exactly 4.0 mL 
of the nicotine standard (II.A.2.a) to a 50 mL volumetric flask 
containing the internal standard and dilute to volume with MTBE. To 
obtain nicotine concentrations equivalent to 0.6, 0.4, 0.2, and 0.1 mg/
mL, pipette precisely 3.0, 2.0, 1.0, and 0.5 mL, respectively, of the 
nicotine standard into the four remaining flasks and dilute to volume 
with MTBE.
    c. Transfer aliquots of the five standards to auto sampler vials 
and determine the detector response for each standard using gas 
chromatographic conditions described in I.E.
    d. Calculate least squares line for linear equation from these 
standards by obtaining the ratio of Areanicotine/
AreaIS. This ratio will be the Y value and the concentration 
of nicotine will be the X value for determining the linear equation of 
the line (Equation 1):
[GRAPHIC] [TIFF OMITTED] TN23JN08.029

Where:

X = Concentration of nicotine in mg
Y = Areanicotine/AreaIS
a = intercept on the ordinate (y axis)
b = slope of the curve

    The final result will be reported in the following units:
    Concentration of nicotine = mg of nicotine/gram of tobacco sample.
    e. Determine the recovery of nicotine by pipetting 10 mL of the 0.4 
mg/mL nicotine standard to a screw capped tube containing 1.0 mL of 2 N 
NaOH. Cap the tube. Shake the contents vigorously and allow the phases 
to separate. Transfer an aliquot of the organic phase to an injection 
vial and inject. Calculate the concentration of nicotine using the 
equation of the line in II.A.2.d above. This should be repeated two 
more times to obtain an average of the three values. The recovery of 
nicotine can be obtained by using the following equation:
[GRAPHIC] [TIFF OMITTED] TN23JN08.030


[[Page 35398]]



B. Standards Addition Assay

    Prior to analyzing a smokeless tobacco product for nicotine 
content, the testing facility must validate the system to verify that 
matrix bias is not occurring during nicotine extraction. This is done 
by analyzing the nicotine calibration standards in the same vegetable 
matrix as the smokeless tobacco. The first time each smokeless tobacco 
product is tested and whenever a change is made to the product 
formulation (including a change to the tobacco blend or cultivar), the 
Standards Addition Assay will be performed, and documentation of its 
performance and of the nicotine concentrations selected for the 
standard curve (II.B.2) will be submitted to the Centers for Disease 
Control and Prevention.
    1. Using an analytical balance, accurately weigh 1.000  
0.020 gram of the homogeneous, prepared tobacco sample into a culture 
tube. Repeat this five times for a total of 6 culture tubes containing 
the smokeless tobacco product. Record the weight of each sample.
    2. Prepare a five-point standard curve for the Standards Addition 
Assay. The standard curve must consist of nicotine concentrations that 
encompass the range of values expected from adding known concentrations 
of the nicotine standard (II.A.2.a) to a measured quantity of the 
smokeless tobacco product (1.000  0.020 gram, described in 
II.B.1). The sixth culture tube is not supplemented with nicotine and 
serves as an analytical blank. Allow the samples to equilibrate for 10 
minutes.
    3. Pipette 5 mL of 2 N NaOH into each tube. Cap each tube. Swirl to 
wet sample and allow to stand 15 minutes.\13\
    4. Pipette 50 mL of extraction solution (II.D.1) into each tube. 
Cap each tube and tighten.\14\
    5. Place tubes in rack(s), place racks in linear shaker in 
horizontal position and shake for two hours.
    6. Remove rack(s) from shaker and place in vertical position to 
allow the phases to separate.
    7. Allow the solvent and nicotine supplemented samples and the 
blank to separate (maximum 2 hours).
    8. Transfer aliquots of the five standards and the blank from the 
extraction tubes to sample vials and determine the detector response 
for each using gas chromatographic conditions described in I.E.
    9. Subtract the Areanicotine/AreaIS of the 
blank from the Areanicotine/AreaIS of each of the 
standards.
    10. Calculate least squares line for linear equation from the 
corrected standards as described above (Equation 1) in II.A.2.d. The 
final corrected result will be reported in the following units: 
Concentration of nicotine = mg of nicotine/gram of tobacco sample.
    11. Determine the recovery of nicotine by pipetting 10 mL of the 
0.4 mg/mL nicotine standard to a screw capped tube containing 1.0 mL of 
2 N NaOH and 10 mL of extraction solution (II.D.1). Cap the tube and 
tighten. Shake the contents vigorously and allow the phases to 
separate. Transfer an aliquot of the organic phase to an injection vial 
and inject. Calculate the concentration of nicotine using the equation 
of the line above in II.A.2.d. This should be repeated two more times 
to obtain an average of the three values. The recovery of nicotine can 
be obtained by using Equation 2: Recovery = 
Nicotinecalculated/Nicotineactual.
    12. Compare the results of steps II.A.2 and II.B. If they differ by 
a factor of 10% or more, the recovery of nicotine from the aqueous 
matrix is not equivalent to recovery from the vegetable matrix of the 
smokeless tobacco product. In this instance, the nicotine concentration 
of the smokeless tobacco product must be determined from a nicotine 
calibration curve prepared from nicotine standards in a vegetable-based 
matrix.

C. Quality Control Pools

    At least two quality control pools at the high and low ends of the 
expected nicotine values are recommended to be included in each 
analytical run. The pools should be analyzed in duplicate in every run. 
The quality control pools should be available in sufficient quantity to 
last for all analyses of a product.

D. Sample Extraction Procedure \12\

    1. Extraction solution is prepared by pipetting 10 mL of the IS 
from the stock solution (II.A.1) to a 1000 mL volumetric flask and 
diluting to volume with MTBE.
    2. Using an analytical balance, accurately weigh 1.000  
0.020 gram of prepared tobacco sample into culture tube and record 
weight.\15 \ Sample each smokeless tobacco brand name according to the 
provided testing frequency schedule.\19 \ The number of products 
sampled should reflect an acceptable level of precision.\16\ The test 
material is to be representative of the product that is sold to the 
public and therefore should consist of sealed, packaged samples of 
finished product that is ready for commercial distribution. Samples are 
to be analyzed in duplicate.
    3. Pipette 5 mL of 2 N NaOH into the tube. Cap the tube. Swirl to 
wet sample and allow to stand 15 minutes.\13\
    4. Pipette 50 mL of extraction solution into tube, cap tube and 
tighten.\14\
    5. Place tubes in rack(s), place racks in linear shaker in 
horizontal position and shake for two hours.
    6. Remove rack(s) from shaker and place in vertical position to 
allow the phases to separate.
    7. Allow the solvent and sample to separate (maximum 2 hours). 
Transfer an aliquot from the extraction tube to a sample vial and cap.
    8. Analyze the extract using GC conditions as described above (I.E) 
and calculate the concentration of nicotine using the linear 
calibration equation. Correct percent nicotine values for both recovery 
and weight of sample by using Equation 3.\17\
[GRAPHIC] [TIFF OMITTED] TN23JN08.031

    9. Report the final nicotine determination as mg of nicotine per 
gram of the tobacco product (mg nicotine/gram), to an accuracy level of 
two decimal places for each brand name (e.g., Skoal Bandits 
Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, etc.). 
All data should include the mean value with a 95% confidence interval, 
the range of values, the number of samples tested, the number of lots 
per brand name, and the estimated precision of the mean. Information 
will be reported for each manufacturer and variety (including brand 
families and brand variations) and brand name (e.g., Skoal Bandits 
Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, etc.).

[[Page 35399]]

III. Total Moisture Determination

    A. This procedure is a modification of AOAC Method 966.02 (1990) 
and is referred to as ``Total Moisture Determination'' because it 
determines water and tobacco constituents that are volatile at 
temperatures of 99  1.0 [deg]C.
    B. Accurately weigh 5.00 grams of the sample (ground to pass <= 4 
mm screen) \20\ into a weighed moisture dish and place uncovered dish 
in oven.\21\ Sample each smokeless tobacco brand name according to the 
provided testing frequency schedule.\19\ The number of products sampled 
should reflect an acceptable level of precision.\16\ The test material 
is to be representative of the product that is sold to the public and 
therefore should consist of sealed, packaged samples of finished 
product that is ready for commercial distribution. Samples are to be 
analyzed in duplicate.
    C. Do not exceed 1 sample/10 sq in (650 sq cm) shelf space, and use 
only 1 shelf. Dry 3 hr at 99  1.0 [deg]C. Remove from oven, 
cover, and cool in desiccator to room temperature (about 30 min). 
Reweigh and calculate percent moisture.
    D. Report the final moisture determination as a percentage (%), to 
an accuracy level of one decimal place for each brand name (e.g., Skoal 
Bandits Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, 
etc.). All data should include the mean value with a 95% confidence 
interval, the range of values, the number of samples tested, the number 
of lots per brand name, and the estimated precision of the mean. 
Information will be reported for each manufacturer and variety 
(including brand families and brand variations) and brand name (e.g., 
Skoal Bandits Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut 
Wintergreen, etc.).

IV. pH Measurement \12\ \22\

    A. Test samples as soon as possible after they are received. Sample 
each smokeless tobacco brand name according to the provided testing 
frequency schedule.19 The number of products sampled should 
reflect an acceptable level of precision.16 The test 
material is to be representative of the product that is sold to the 
public and therefore should consist of sealed, packaged samples of 
finished product that is ready for commercial distribution. Samples are 
to be analyzed in duplicate.
    B. Accurately weigh 2.00 grams of the sample. Place in a 50 mL 
polypropylene container with 20 mL deionized distilled water.
    C. Place Teflon-coated magnetic stirring bar in container and stir 
mixture continuously throughout testing.
    D. Measure pH of sample after a two-point calibration of the pH 
meter to an accuracy of two decimal places using standard pH buffers 
(4.01 and 7.00 or 7.00 and 10.00) that will encompass the expected pH 
value of the smokeless tobacco product.
    E. The first time pH values are determined for a smokeless tobacco 
product, measure the pH of the smokeless tobacco product at 5, 15, and 
30 minutes. If there is no systematic variation in pH values with time, 
all subsequent pH determinations are made at 5 minutes. If there is 
systematic variation in pH values, continue to measure the pH of the 
smokeless tobacco product until the pH value is stable and does not 
vary more than 10% over 15 minutes. Report the final pH value.
    F. Report the final pH determination to an accuracy level of two 
decimal places for each brand name (e.g., Skoal Bandits Wintergreen, 
Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, etc.). All data 
should include the mean value with a 95% confidence interval, the range 
of values, the number of samples tested, the number of lots per brand 
name, and the estimated precision of the mean. Information will be 
reported for each manufacturer and variety (including brand families 
and brand variations) and brand name (e.g., Skoal Bandits Wintergreen, 
Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, etc.).
    G. Estimate the un-ionized (free) nicotine content with the 
Henderson-Hassel Balch equation (Equation 4), based on measured pH and 
nicotine content.
[GRAPHIC] [TIFF OMITTED] TN23JN08.032

[GRAPHIC] [TIFF OMITTED] TN23JN08.033

pKa = 8.02 (CRC Handbook of Chemistry and Physics, 1989-1990)
[B] = amount of un-ionized (free) nicotine
[BH+] = amount of ionized nicotine

    H. Report the final estimated un-ionized (free) nicotine as a 
percentage (%) of the total nicotine content, to an accuracy level of 
two decimal places and as mg of un-ionized (free) nicotine per gram of 
the tobacco product (mg un-ionized (free) nicotine/gram), to an 
accuracy level of two decimal places for each brand name (e.g., Skoal 
Bandits Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, 
etc.). All data should include the mean value with a 95% confidence 
interval, the range of values, the number of samples tested, the number 
of lots per brand name, and the estimated precision of the mean. 
Information will be reported for each manufacturer and variety 
(including brand families and brand variations) and brand name (e.g., 
Skoal Bandits Wintergreen, Skoal Long Cut Cherry, Skoal Long Cut 
Wintergreen, etc.).

Sample calculation:

Mean total nicotine = 10.30 (mg/g)
Mean pH = 7.50
pKa = 8.02
[GRAPHIC] [TIFF OMITTED] TN23JN08.034

[GRAPHIC] [TIFF OMITTED] TN23JN08.035


[[Page 35400]]


[GRAPHIC] [TIFF OMITTED] TN23JN08.036

[GRAPHIC] [TIFF OMITTED] TN23JN08.037

[GRAPHIC] [TIFF OMITTED] TN23JN08.038

[GRAPHIC] [TIFF OMITTED] TN23JN08.039

[GRAPHIC] [TIFF OMITTED] TN23JN08.040

[GRAPHIC] [TIFF OMITTED] TN23JN08.041

V. Assay Criteria for Quality Assurance

A. Establishing Limits for Quality Control Parameters

    All quality control parameters must be determined within the 
laboratory in which they are to be used. At least 10 within-laboratory 
runs must be performed to establish temporary confidence intervals for 
the quality control parameters. Permanent limits should be established 
after 20 runs and should be reestablished after each additional 20 
runs.

B. Exclusion of Outliers From the Calibration Curve \18\

    The coefficient of determination between Areanicotine/AreaIS and 
nicotine concentration should be equal to 0.99 or higher. Any 
calibration standard having an estimated concentration computed from 
the regression equation (Equation 1) which is different from its actual 
concentration by a factor of 10% can be excluded from the calibration 
curve. Up to two concentrations may be excluded, but caution should be 
used in eliminating values, since bias may be increased in the 
calibration curve. If an outlier value is eliminated, its duplicate 
value must also be discarded to avoid producing a new bias. All 
unknowns must fall within the calibration curve; therefore, duplicate 
values excluded at either end of the calibration curve will restrict 
the useful range of the assay.

C. Quality Control Pools and Run Rejection Rules

    The mean estimated nicotine concentration in a pool should be 
compared with the established limits for that pool based on at least 20 
consecutive runs. An analytical run should be accepted or rejected 
based upon the following set of rules adapted from Westgard et al. 
(1981).
    1. When the mean of one QC pool exceeds the limit of x  
3 standard deviations (SD), then the run is rejected as out of control. 
Here, x and SD represent the overall mean and standard deviation of all 
estimated nicotine concentrations for a particular pool in the runs 
which were used to establish the control limits.
    2. When the mean nicotine concentrations in two QC pools in the 
same run exceed the same direction, then the run must be rejected. The 
same direction is the condition in which both pools exceed either the x 
+ 2 SD or the x - 2 SD limits.
    3. When the mean nicotine concentrations in one or two QC pools 
exceed their x  2 SD limits in the same direction in two 
consecutive runs, then both runs must be rejected.
    4. When the mean nicotine concentrations in two QC pools are 
different by more than a total of 4 SD, then the run must be rejected. 
This condition may occur, for example, when one QC pool is 2 SD greater 
than the mean, and another is 2 SD less than the mean.

Endnotes

    The comments and notes listed below can be described as Good 
Laboratory Practice guidelines; they are described in detail in this 
protocol to ensure minimal interlaboratory variability in the 
determination of nicotine, total moisture, and pH in smokeless 
tobacco.
    \1\ This protocol assumes that the testing facility will 
implement and maintain a stringent Quality Assurance/Quality Control 
program to include, but not be limited to, regular interlaboratory 
comparisons, determination of the quality and purity of purchased 
products, and proper storage and handling of all reagents and 
samples.

[[Page 35401]]

    \2\ When a specific product or instrument is listed, it is the 
product or instrument that was used in the development of this 
method. Equivalent products or instruments may also be used. Use of 
trade names is for identification only and does not constitute 
endorsement by the Public Health Service or the U.S. Department of 
Health and Human Services.
    \3\ All chemicals, solvents, and gases are to be of the highest 
purity.
    \4\ Companies must ensure that the purity of the nicotine base 
is certified by the vendor and that the chemical is properly stored. 
However, nicotine base oxidizes with storage, as reflected by the 
liquid turning brown. If oxidation has occurred, the nicotine base 
should be distilled prior to use in making a standard solution.
    \5\ A suggested method for the determination of nicotine purity 
is CORESTA Recommended Method No. 39.
    \6\ Horizontal shaking will allow more intimate contact of this 
three phase extraction. There is a minimal dead volume in the tube 
due to the large sample size and extraction volume. This 
necessitates horizontal shaking.
    \7\ If a linear shaker is not available, a wrist action shaker 
using 250 mL stoppered Erlenmeyer flasks can be substituted. Values 
for nicotine are equivalent to those obtained from the linear 
shaker.
    \8\ After installing a new column, condition the column by 
injecting a tobacco sample extract on the column, using the 
described column conditions. Injections should be repeated until 
areas of IS and nicotine are reproducible. This will require 
approximately four injections. Recondition column when instrument 
has been used infrequently and after replacing glass liner.
    \9\ Glass liner and septum should be replaced after every 100 
injections.
    \10\ Most older instruments operate at constant pressure. To 
reduce confusion, it is suggested that the carrier gas flow through 
the column be measured at the initial column temperature.
    \11\ The testing facility must ensure that samples are obtained 
through the use of a survey design protocol for sampling ``at one 
point in time'' at the factory or warehouse. The survey design 
protocol must address short-, medium-, and long-term smokeless 
tobacco product variability (e.g., variability over time and from 
container to container of the tobacco product) in a manner 
equivalent to that described for cigarette sampling in Annex C of 
ISO Protocol 8243. Information accompanying results for each sample 
should include, but not be limited to:
    For each product--manufacturer and variety (including brand 
families and brand variations) and brand name (e.g., Skoal Bandits, 
Skoal Long Cut Cherry, Skoal Long Cut Wintergreen, etc.):
    1. Product ``category,'' e.g., loose leaf, plug, twist, dry 
snuff, moist (wet) snuff, etc.
    2. Lot number.
    3. Lot size.
    4. Number of randomly sampled, sealed, packaged (so as to be 
representative of the product that is sold to the public) smokeless 
tobacco products selected (sampling fraction) for nicotine, 
moisture, and pH determination.
    5. Documentation of method used for random sample selection.
    6. ``Age'' of product when received by testing facility and 
storage conditions prior to analysis.
    \12\ Extraction of nicotine and pH determination must be 
performed with reagents and samples at a room temperature of 22-25 
[deg]C. Room temperature should not vary more than 1[deg]C during 
extraction of nicotine or pH determination.
    \13\ Use non-glass 10 mL repipette for transferring NaOH 
solution.
    \14\ Use 50 mL repipette for transferring MTBE.
    \15\ For dry snuff, use 0.500  0.010 gram sample.
    \16\ The testing facility is referred to ISO Procedure 8243 for 
a discussion of sample size and the effect of variability on the 
precision of the mean of the sample (ISO 8243, 1991).
    \17\ When analyzing new smokeless tobacco products, extract 
product without IS to determine if any components co-elute with the 
IS or impurities in the IS. This interference could artificially 
lower calculated values for nicotine.
    \18\ The calculated nicotine values for all samples must fall 
within the low and high nicotine values used for the calibration 
curve. If not, prepare a fresh nicotine standard solution and an 
appropriate series of standard nicotine dilutions. Determine the 
detector response for each standard using chromatographic conditions 
described in I.E.
    \19\ The testing frequency for each smokeless tobacco brand name 
(e.g., Skoal Bandits Wintergreen, Skoal Long Cut Cherry, Skoal Long 
Cut Wintergreen, etc.) is based on the manufacturing duration (refer 
to table below). Each smokeless tobacco brand name will be sampled 
and tested for nicotine, total moisture, and pH no fewer than twice 
and no more than four times during a calendar year.

------------------------------------------------------------------------
                                                                  Test
               Manufacturing duration in weeks                 frequency
                                                                   *
------------------------------------------------------------------------
up to and including 4........................................          2
up to and including 28.......................................          3
up to and including 52.......................................          4
------------------------------------------------------------------------
* Use a statistical program to determine random sampling dates based on
  the total manufacturing duration during a calendar year. Sampling
  dates should fall on actual manufacturing days for the product when
  test material that is representative of the product that is sold to
  the public (consisting of sealed, packaged samples) is available. If a
  statistically determined sampling date falls on a day that does not
  meet this criterion, sample the product on the next date that does
  meet the criteria.

    For smokeless tobacco brand names with episodic production 
during a calendar year, the total number of sampling dates is 
determined by the sum of the individual test frequencies, not to 
exceed four. For the purpose of the Protocol, episodic production is 
defined as manufacturing intervals separated by periods of 30 or 
more days when the smokeless tobacco brand name is not manufactured.
    Example 1: Within a single calendar year a smokeless tobacco 
brand name is manufactured from January 1 to March 31 and from 
September 1 to December 15. The testing frequency for the first 
manufacturing interval is 3 and for the second manufacturing 
interval is 3. The Protocol allows that each smokeless tobacco brand 
name be tested for nicotine, total moisture, and pH no more than 
four times during a calendar year. Therefore, 4 random sampling 
dates, as described in the footnote to the above table, are 
determined for the smokeless tobacco brand name. The values for 
nicotine, moisture, and pH determinations, and unionized (free) 
nicotine calculations and the mean of the 4 data points for that 
smokeless tobacco brand name are reported.
    Example 2: Within a single calendar year a smokeless tobacco 
brand name is manufactured from April 5 to May 3 and from September 
1 to December 15. The testing frequency for the first manufacturing 
interval is 2 and for the second manufacturing interval is 3. The 
values for nicotine, moisture, and pH determinations, and unionized 
(free) nicotine calculations and the mean of the 4 data points for 
that smokeless tobacco brand name are reported.
    Example 3: Within a single calendar year a smokeless tobacco 
brand name is manufactured from January 1 to January 15 and from 
September 1 to September 22. The testing frequency for the first 
manufacturing interval is 2 and for the second manufacturing 
interval is 2. Four random sampling dates are selected to fall 
within the 6 weeks of manufacturing for the smokeless tobacco brand 
name. The values for nicotine, moisture, and pH determinations, and 
unionized (free) nicotine calculations and the mean of the 4 data 
points for that smokeless tobacco brand name are reported.

    \20\ The method is a modification of AOAC Method 966.02 (1990) 
in that the ground tobacco passes through a 4 mm screen rather than 
a 1 mm screen.
    \21\ When drying samples, do not dry different products (e.g., 
moist (wet) snuff, dry snuff, loose leaf) in the oven at the same 
time since this will produce errors in the moisture determinations.
    \22\ The method is a modification of a method published by 
Henningfield et al. (1995).

References

    AOAC (Association of Official Analytical Chemists). Official 
Methods of Analysis. 966.02: Moisture in Tobacco. (1990) Fifth 
Edition. K. Helrich (ed). Association of Official Analytical 
Chemists, Inc., Suite 400, 2200 Wilson Boulevard, Arlington, 
Virginia 22201 USA.
    CORESTA (Centre de Coop[eacute]ration pour les Recherches 
Scientifiques relatives au Tabac). Recommended Method No. 39: 
Determination of the purity of nicotine and nicotine salts by 
gravimetric analysis--Tungstosilic acid method. November 1994. 87-
90.
    CRC Handbook of Chemistry and Physics. R.C. Weast, D.R. Lide, 
M.J. Astle, and W.H. Beyer (eds). 70th ed. Boca Raton, Florida: CRC 
Press (1989-1990) D-162.
    Henningfield, J.E., Radzius A., Cone E.J. (1995). Estimation of 
available nicotine content of six smokeless tobacco products. 
Tobacco Control 4:57-61.

[[Page 35402]]

    ISO (International Organization for Standardization). IOS 8243: 
Cigarettes--Sampling. (1991). Second Edition. Prepared by Technical 
Committee ISO/TC 126, Tobacco and tobacco products. International 
Organization for Standardization, Case Postale 56, CH-1211 Genve 20, 
Switzerland.
    Westgard, J.O., Barry P., Hunt M., and Groth T. (1981). A multi-
rule Shewhart chart for quality control in clinical chemistry. 
Clinical Chemistry 27:493.

[FR Doc. E8-14112 Filed 6-20-08; 8:45 am]
BILLING CODE 4163-18-P