product quality with the hp 4440a - infometrix.bizinfometrix.biz/apps/21-0199_hpappsan.pdfthe food...

2
Application Note Headspace Mass Spectrometry © 1999 Infometrix, Inc. Many products can be classified by composi- tional differences in their volatile composition. Analysis of such volatiles is compelling because the instruments involved in headspace analysis are simple to operate, they require little mainte- nance and results are available with rapid turn- around. The applications described in brief below are examples of headspace analysis by the HP 4440A Chemical Sensor. These eight files are shipped with the instrument and are available as part of the download of Pirouette® from the Infometrix web site. Java.dat is an example file that is used in the HP 4440A Primer. This file contains data col- lected from replicates of four Starbucks coffees: Guatemala and Sumatra, both regular and decaf- feinated. The Primer provides a detailed walk- through of the coffee application intended to be a tutorial that can be easily reproduced by an end-use customer. MoreJava.dat continues the training by pro- viding additional coffee samples, some of which are new samples of the Java.dat coffees, some from other parts of the world and some as aged versions of the target coffees. This data set is used in the Primer to show how to automate routine analysis. Belaboring the Coffee application, the Cof- fee.dat data set expands the horizons to include 60 samples of 10 different Starbucks roasts. Each of the coffee types can be distin- guished from one another by comparing the intensities across the mass range. Figure 1. Four differ- ent varieties of ground and roasted coffee are easily distinguished by the HP 4440A. Figure 2. Additional coffee samples along with aged versions and new coffee types are used to show how the system flags unac- ceptable batches. Figure 3. In a compari- son of 10 varieties of Starbucks blends, the distinctiveness of each blend is still evi- dent. -10 0 10 Factor1 -6 -2 2 6 Sumatra D Sumatra Guatemala D Guatemala 51 56 61 -10 10 30 50 70 Response m/z Analysis of Product Quality Using the HP 4440A

Upload: others

Post on 26-Oct-2019

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Product Quality with the HP 4440A - infometrix.bizinfometrix.biz/apps/21-0199_HPAppsAN.pdfthe food samples shown here, there is still enough discrimi-nating power in the HP 4440A to

Application NoteHeadspace Mass Spectrometry

© 1999 Infometrix, Inc.

eP

rease

el-s:af-k-e

an

h

assette

os

in-e

Figure 1. Four differ-ent varieties of ground and roasted coffee areeasily distinguished by the HP 4440A.

Figure 2. Additional coffee samples along with aged versions and new coffee types are used to show how the system flags unac-ceptable batches.

Figure 3. In a comparison of 10 varieties of Starbucks blends, the distinctiveness of each blend is still evi-dent.

Analysis of Product Quality Using the HP 4440A

Many products can be classified by composi-tional differences in their volatile composition.Analysis of such volatiles is compelling becausethe instruments involved in headspace analysisare simple to operate, they require little mainte-nance and results are available with rapid turn-around.

The applications described in brief below arexamples of headspace analysis by the H4440A Chemical Sensor. These eight files ashipped with the instrument and are available part of the download of Pirouette® from thInfometrix web site.

Java.dat is an example file that is used in thHP 4440A Primer. This file contains data colected from replicates of four Starbucks coffeeGuatemala and Sumatra, both regular and decfeinated. The Primer provides a detailed walthrough of the coffee application intended to ba tutorial that can be easily reproduced by end-use customer.

MoreJava.dat continues the training by pro-viding additional coffee samples, some of whicare new samples of the Java.dat coffees,some from other parts of the world and some aged versions of the target coffees. This data is used in the Primer to show how to automaroutine analysis.

Belaboring the Coffee application, the Cof-fee.dat data set expands the horizons tinclude 60 samples of 10 different Starbuckroasts. Each of the coffee types can be distguished from one another by comparing thintensities across the mass range.

-

-10 0 10

Factor1

-6

-2

2

6

Sumatra

D Sumatra

Guatemala

D Guatemala

51 56 61

Variable Name

-10

10

30

50

70

Re

spo

nse

m/z

Page 2: Product Quality with the HP 4440A - infometrix.bizinfometrix.biz/apps/21-0199_HPAppsAN.pdfthe food samples shown here, there is still enough discrimi-nating power in the HP 4440A to

© 1999Infometrix, Inc., 10634 E

s-

--med

is

d-

erevea a

e-oer

-st

ckr-gho-

Figure 4. Although thedistinction is much less than for many of the food samples shown here, there is still enough discrimi-nating power in the HP 4440A to separate premium orange juices.

Figure 5. Coca-Cola and Pepsi (both diet and regular) can be easily differentiated using the HP 4440A; this is a PCA scores plot showing the dis-tinctiveness.

Figure 6. The level of adulteration of olive oil can be tracked using the headspace mass spectrometry system.

Figure 7. Measuring the peroxide value of butter is time consum-ing and is replaced with a five minute analysis with no sam-ple preparation.

Figure 8. Diesel is a complex mixture of hydrocarbons, but various amounts of diesel in kerosene (a similar mixture of hydrocarbons) can be monitored easily.

Evaluation of pure premium orange juice is posible in the OJ.dat data set. Even though thereis no significant difference in the starting material (both Tropicana and Minute Maid use freshsqueeze with no preservatives and no juice froconcentrate) we can see distinctions bassolely on the process.

A comparison of soft drinks by headspace available in the Cola.dat file. Approximately4 mL of the product was poured into a heaspace vial and heated at 80°C to release the vol-atiles. Note that the presence of CO2 does notaffect the analysis.

OliveOil.dat contains data on pure virginolive oils from Spain and Greece plus purhazelnut oil (a common adulterant). There aalso some 5% and 10% hazelnut in Greek olioil mixtures in this data. We suggest doing PCA and then exclude the Spanish oils to doregression for assessing level of purity.

In the Butter.dat set, we utilize theHP 4440A as a replacement for direct measurment of the peroxide value (PV) in order tmonitor the freshness of the butter. The highthe PV number, the more rancid the butter.

data courtesy of Dr. Vincent Shiers, Leatherhead.

The Diesel.dat file evaluates the kerosenecontent in diesel fuel. Both are complex mixtures of paraffinic and naphthenic hydrocarbonof approximately the same molecular weighrange. This is a quality issue, providing a cheon the back-blending that is common in refineies. In addition, there is a safety concern; a hikerosene content increases the risk of automtive fires during accidents.

-Brand 1

Brand 2

-5 0 5 10

Factor1

-1

0

1

Diet Coke

Coke

Diet Pepsi

Pepsi

SpanishHazelnut

Greek + Mix

0 10 20 30

Measured Y

0

10

20

30

Pre

d C

al

5

Measured PV

HP

4440

PV

Diesel

Kerosene

Mixtures

21-1/99. Riverside Dr., Bothell, WA 98011 Phone: (425) 402-1450, Fax: (425) 402-1040, email: [email protected]