WHERE IS THE ISOPROPYL

METHOXYPYRAZINE (IMP) IN

GREEN COFFEE AFFECTED BY

POTATO TASTE?

Susan Jackels, Charles Jackels and Fabrice Lee Seattle University and University of Washington Bothell

Rwanda Antestia-Potato Taste Research Group

Rwanda Collaboration

Colloquium Sponsored by:

Outline

Background: What is potato taste defect?

Results of previous research on IMP

Our research questions

Experiment: SPME GC-MS on sort fractions of a donated PTD

sample

Results: location of IMP in the sample

Conclusions

Tentative implications for PTD

Acknowledgements

Background

In 2001, producing specialty coffee became part of a strategy to

develop the Rwandan economy

Washing stations were implemented to support specialty coffee:

specialty production rose from 1% in 2002 to 27% in 2012

Potato Taste Defect (PTD) threatens this plan.1

1Government of Rwanda, Strategic Plan for Agricultural Transformation in Rwanda (Ministry of Agriculture and Animal Resources,

Kigali, 2004).

What is PTD?

•Potato Taste Defect (PTD) leads to

roasted coffee that smells and tastes

like potato skins

•PTD is not detectable in green

coffee beans by smell or appearance

•PTD is detected in the cup of coffee

•PTD is detected in roasting just

before the first crack

•PTD affects:

– 400,000 smallholder farmers, 86%

of whom are women1

– 25% of Rwandan coffee crops in

20082

– 51% of Cup of Excellence entries

affected by PTD in 2013 1World Bank 2011

2Ngabitsinze, J. C., A. Mukashema, M. Ikirezi and F. Niyitanga.

(October 2011). Planning and costing adaptation of perennial crop

systems to climate change: Coffee and banana in Rwanda. Case

study report. http://pubs.iied.org/pdfs/G03174.pdf#page18 .

TO PREVENT PTD: ALL INSECT

DAMAGED BEANS ARE SORTED OUT

Sorting is not 100% effective

Sometimes PTD is detected

during coffee cupping, or

tasting, before coffee buyer

makes a purchase

Beans may also manifest the

defect during export away from

Rwanda and be noticed upon

arrival

Buyers decline to purchase

lots with PTD

One PTD cup in 40 or fewer

will result in the coffee being

rejected

Previous Research on PTD

Studies of PTD occurred mostly in 1980’s and 1990’s

Most common hypothesis is that PTD originates from

feeding damage by the antestia bug followed by infection

by a bacterium that produces a malodorous metabolite

adhering to the beans

Coffee cherries that survived

attack by the antestia bug. Antestia bug on

coffee cherry

Decaying coffee

cherry after antestia

bug attack

3-isopropyl -2 methoxypyrazine

Intensely odiferous potato skin smelling molecule

Detectable by human nose as low as 2 parts per

trillion

Associated with PTD in 1980s3

Sourced from ground green beans by

solvent extraction and chromatography

Found by GC/MS in both good and PTD

coffees; 30 times higher in PTD coffee

Not detected by nose in whole green

coffee, but is sensed in ground green

coffee and roasted coffee

3Becker, R., Döhla, B., Nitz, S.,Vitzthum, O. G. (1987). Identification of the peasy off-flavour note on central african

coffees. In 12th International Conference on Coffee, Montreux, pp 203-215.

IMP

FROM OUR RESEARCH ON SURFACE

VOLATILES OF GREEN COFFEE

BEANS WE NOW KNOW…

I. The surface volatiles have a profile that distinguishes

PTD from non-PTD coffee, but the profile does not

include IMP.

II. The PTD surface volatile profile consists of tridecane,

dodecane, hexanal – and these same volatiles are

dominant in the antestia bug volatiles. It seems antestia

has left its “scent” on the surface of PTD coffee.

III. No IMP is found on the surface of PTD or non-PTD coffee

beans, even if heated to 90 oC for sampling.

OUR RESEARCH

QUESTIONS:

I. Where is the odiferous IMP found in green coffee beans?

II. Is IMP associated with a particular defect sort fraction,

such as insect damaged, fluorescent , broken or

withered?

III. Where do we look for the proposed “few bad beans” that

are responsible for potato taste?

Objective to find the IMP in

PTD coffee

Previously we studied the surface volatile organic

compounds (SVOCs) in green coffee that define

the chemical nature of PTD on the surface of beans

Surface volatiles should concentrate compounds

deposited through antestia feeding activity and/or

bacterial growth – DID NOT FIND IMP HERE

Now, we study the interior volatile organic

compounds (IVOCs) inside the green coffee beans

Interior volatiles should reflect compounds

produced by the coffee bean itself in response to

stress of antestia feeding activity and/or bacterial

growth

THE SAMPLE NECESSARY FOR

THIS RESEARCH PROJECT WAS

DONATED BY RUTH CHURCH OF

ARTISAN COFFEE IMPORTERS –

THANK YOU!

• This sample is a high-quality specialty coffee from

Burundi that “developed” a very mild potato taste after

being imported to the United States

Sample was cupped and sorted by Mr. Ed Whitman and Mr.

Carlos Serrano of the Rogers Family Company

It had one cup PT out of 120 cups

Excellent quality except for that one cup!

Can we find the PTD and the IMP in this coffee sample?

Method

Starting point: a method previously applied to roasted ground

coffee : heated coffee is sampled by solid phase micro-extraction

(SPME)5

Need to apply to ground green coffee for Interior Volatile Organic

Chemical analysis :

• Freeze green coffee beans at -70 oC and grind, reserving

fraction that passes through a 2 mm sieve

• Collect volatiles of ground coffee at 60 oC with SPME for one

hour

• Separate by gas chromatography and analyze by mass

spectrometry

• SPME has been successfully applied to volatiles,

including IMP, down to nanogram level6

5Mondello, L., et al, (2005) Reliable characterization of coffee bean aroma profiles by headspace solid phase

microextraction-gas chromatography-mass specrometry… J. Separation Science 2005, 28, 1101 – 1109. 6Sala, C, Mestres, M., Marti, M.P., Busto, O., and Guasch, J. (2002) Headspace solid phase micro-extraction analysis of

3-alkyl-2-methoxypyrazines in wines. Journal of Chromatography A, 953, 1-6.

Solid Phase Micro-

Extraction (SPME)

Gas Chromatography-

Mass Spectrometry (GC-MS)

Chromatogram with peaks and mass spectrum for each compound

Compounds identified by 1)comparison of mass spectrum

with NIST05 and FFNSC2 database s(>75% match) and 2) retention

time matches with candidate compound.

Identifying compounds from

GCMS data

With a good MS match (>75% quality), to ensure correct MS

compound identification:

• Check retention times to make sure they correlate for compound

and our GC settings

• Run standards to verify compound identification for key SVOCs

CHROMATOGRAM OF SAMPLE,

FIRST LOOKING AT WHOLE BEANS

5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.000

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

220000

240000

Time-->

Abundance

TIC: 0228_03.D\ data.ms

8.771 23.225

25.120

31.447

31.824

34.353

35.19035.480

37.260

39.122

40.343

40.801

42.555

43.438

45.657

45.920

46.406

47.121

47.600

48.60550.007

50.655

51.882

52.091

53.420

54.054

56.131

56.313

63.341

72.177T

ridecane

Note: Bulk sample does not exhibit a strong PT profile. Tridecane peak is

comparable to other coffee volatiles.

CHROMATOGRAM OF BULK

SAMPLE, GROUND BEANS

5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.000

50000

100000

150000

200000

250000

300000

350000

400000

450000

500000

550000

600000

650000

T ime-->

Abundanc e

T IC: 0203_01.D \ da ta .ms

8 .729

14.853

22.211

23.385

25.050

28.011

31.417

31.700

32.004

32.44232.58433.973

34.317

35.127

35.376

37.211

37.804

39.436

40.462

40.745

42.006

42.525

43.382

45.05545.500

45.594

46.242

47.16647.354

47.934

48.548

50.632

52.150

52.683

53.330

53.971

54.821

55.799

56.082

57.01357.34357.903

58.942

61.552

63.305

67.21767.39968.350

69.342

70.252

72.100

73.449

Trid

eca

ne

Note: Volatile profile is different from surface volatiles. Tridecane peak is very

small, consistent with tridecane on surface but not in interior. No IMP found!

INSECT DAMAGED

FRACTION

Results from chromatograms of ground

fractions: insect damaged, ground

5 .0 0 1 0 .0 0 1 5 .0 0 2 0 .0 0 2 5 .0 0 3 0 .0 0 3 5 .0 0 4 0 .0 0 4 5 .0 0 5 0 .0 0 5 5 .0 0 6 0 .0 0 6 5 .0 0 7 0 .0 00

2 0 0 0 0

4 0 0 0 0

6 0 0 0 0

8 0 0 0 0

1 0 0 0 0 0

1 2 0 0 0 0

1 4 0 0 0 0

1 6 0 0 0 0

1 8 0 0 0 0

2 0 0 0 0 0

2 2 0 0 0 0

2 4 0 0 0 0

2 6 0 0 0 0

2 8 0 0 0 0

3 0 0 0 0 0

3 2 0 0 0 0

3 4 0 0 0 0

3 6 0 0 0 0

3 8 0 0 0 0

4 0 0 0 0 0

T im e -->

A b u n d a n c e

T IC : 0 2 2 1 _ 0 2 .D \ d a ta .m s

6 .8 4 4 8 .6 9 21 8 .1 1 4

2 3 .0 8 5

2 4 .7 3 7

2 4 .9 2 6

3 0 .8 6 8

3 1 .2 5 2

3 1 .5 4 2

3 1 .7 0 4

3 2 .2 7 7

3 4 .1 8 6

3 4 .9 6 2

3 5 .2 4 5

3 7 .0 5 3

3 7 .4 9 8

3 9 .0 5 6

3 9 .7 1 0

3 9 .8 3 14 0 .1 1 5

4 0 .5 8 0

4 2 .3 3 44 3 .2 1 0

4 4 .0 8 74 5 .3 2 1

4 5 .4 3 6

4 5 .8 2 7

4 6 .1 7 84 6 .8 9 3

4 7 .2 5 74 7 .3 7 2

4 8 .3 7 7

4 8 .5 7 24 9 .2 8 0

4 9 .6 4 5

4 9 .7 6 65 0 .4 2 05 1 .7 4 2

5 2 .5 5 8

5 3 .7 9 9

5 4 .6 6 25 5 .0 6 7

5 5 .8 8 3

5 6 .0 7 2

5 6 .9 4 2

6 1 .1 2 46 1 .8 1 2

6 3 .0 7 3

6 6 .9 9 16 8 .1 1 17 1 .8 9 5

7 3 .3 0 4

7 4 .1 2 0

IMP IBMP

Found IMP (quality 91) and isobutylmethoxy pyrazine (IBMP) (qual 83).

IMP is

concentrated

inside beans

of the

insect damaged

fraction.

Results from chromatograms of

fractions: insect damaged, whole

beans

5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.000

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

220000

240000

Time-->

Abundance

TIC: 0228_03.D\ data.ms

8.771 23.225

25.120

31.447

31.824

34.353

35.19035.480

37.260

39.122

40.343

40.801

42.555

43.438

45.657

45.920

46.406

47.121

47.600

48.60550.007

50.655

51.882

52.091

53.420

54.054

56.131

56.313

63.341

72.177

No IMP! No IBMP!

Since no IMP or

IBMP are found

on the surface

of the insect

damaged beans,

they are inside

the beans.

Conclusions from study of

fractions of Burundi sample:

IMP is found inside the beans in the IVOCs, but NOT on the surface

of the beans in the SVOCs

IMP is most concentrated in the insect damaged sort fraction and

less visible in the fluorescent and broken fractions

Bulk sample exhibits only a mild SVOC PTD profile and no IMP

visible in SVOC or IVOC

IMP AND FLUORESCENCE AS

MARKERS FOR PTD COFFEE

IMP is found in the sort fractions associated with insect activity or

damage: insect damaged, fluorescent, broken

Blue fluorescence often occurs on spots that have slight discoloration.

TENTATIVELY,

IMP found in the interior of the beans in

fractions with insect or other damage

supports the connection of PTD and

antestia activity.

But, IMP in the interior of the beans raises

the possibility that the bean itself

produces IMP in response to stress of

antestia or environmental effects!

IMP concentrated in the insect damaged

fraction gives a clue about where to look

for the “bad beans.”

QUESTION: DO THE RWANDAN

PTD SAMPLES ALSO HAVE IMP

INSIDE THE BEANS? YES

5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.000

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

Time-->

Abundance

TIC: 0227_01.D\data.ms

6.836 8.684 23.090

24.958

31.291

31.710

32.33734.205

35.412

37.112

37.577

39.041

40.208

40.66643.297

45.522

45.833

46.230

46.99348.470

49.738

49.859

50.540

51.774

53.912

55.976

60.542

63.186

67.93468.19172.015

73.384

74.132

SG-3387

A PTD

Sample

IMP IBMP

FUTURE WORK

• Collaborate with RFC and Rwandan farmers to continue to

characterize PTD in green coffee

• Pursue the “bad bean” in the insect damaged fraction of

PTD samples

• Obtain some really bad insect damaged samples (primary

sort on the farm) to verify that PTD has been removed

• Continue to study antestia volatiles (further identification

and characterization) to link with PTD

Acknowledgements

Ruth Church of Espresso Elevado for donating

the large sample

Dr. Mario Serracin and Mr. Ed Whitman, Rogers

Family Company for sample characterization

Mr. Steve Miller, Seattle University Chemistry

Department

Rwanda Antestia-Potato Taste Research Group

Rwanda Collaboration

Colloquium Sponsored by:

Questions?