Friday, April 29, 2016

The Tearoom - Part 1 - What the brain does with a story: Charles Rennie Mackintosh Willow Tearoom

In contemporary marketing the mantra is “You are selling a story…not a product.”  

This mantra led me to read a recent article in the New York Times,* which interpreted a paper in Nature,** as saying what happens in the brain when you listen to a story. According to the NYT, the words in the story elicit neural connections that gather information from all parts of the brain, so that you become immersed in the experience as if it were real, and you no longer experience the reality of the outside world. 

So when I heard the word “Tearoom” today I first thought of the perfectly designed and curated Gypsy’s Tearoom, located in a 1760 house in Westminster Maryland, surrounded by green grasses, where I recently enjoyed a classic “afternoon tea” served in the classic manner, with classic teas and classic tea sandwiches and more (will go into these in another post)…

Gypsy's Tearoom, Westminster, Maryland

…when my thoughts leapt to a tearoom in Glasgow. Several years ago, my daughter and I made the pilgrimage to the Charles Rennie Mackintosh Willow Tearoom, as much for its design as for its tea.

The Room de Luxe at Willow Tearoom, Glasgow, Scotland.***

We sat at the table to the far right in this image, where you see someone serving tea. From this vantage point we could survey the whole room, though we couldn’t get a glimpse of the busy noisy grey street below. 

This is the “Room de Luxe,” as it was called when first created: white white white, with the soft purple upholstery and the shocking pink glass squares in the chairs. White for cleanliness, health, and purity, purple for a hint of royalty, and pink, because tearooms were, in contrast to coffee houses, women’s places—a color between playful innocence and discrete sensuality. 

If we had sat at one of the tables shown in the image below, we could have included the street in our purview. And we could have seen people rushing by, very few entering the building where we sat.

Another view of the The Room de Luxe at Willow Tearoom, Glasgow, Scotland.***

The place was sad to us—the spirit of excitement and novelty was gone, of course—art nouveau is hardly nouveau anymore—but at the same time the fixtures themselves don’t age well despite restoration in 1983. As Wikipedia says: “The tea room however has fallen into disrepair and is now lacking its originally intended atmosphere and quality. Fans of Mackintosh are urgently calling for a more sympathetic restoration.”

What a contrast to the 1760 building that houses Gypsy’s Tearoom: it has aged gracefully, and in fact a little aging coupled with a little nostalgia made it all the more attractive!

Gypsy's Tearoom, Westminster, Maryland.



*** Photos of The Room de Luxe at The Willow Tearooms, Glasgow designed by Charles Rennie Mackintosh in collaboration with Margaret MacDonald for Catherine Cranston. Photograph taken on 9 March 2006 by User:Dave souza

Wednesday, April 27, 2016

Receptor rivalry: an experiment.

Recently gave a talk about pairing tea with foods/herbs/spices to a lively group of tearoom people. Started out with my favorite receptor experiment, one that you can do at home:

Mint and Cinnamon. Photo courtesy tea connoisseur Marzi Pecen — thanks Marzi! 

Take a mild cinnamon candy and a mild peppermint (I use Brach’s® cinnamon hard candy and Lifesaver® peppermint candy). First put the cinnamon candy in your mouth, get a good cinnamon flavor going, then take the candy out. Next put the peppermint candy in your mouth: the cinnamon flavor instantly disappears. Then wait—after a short while the mint flavor disappears, and gradually the cinnamon comes back, faintly but perceptible for most people.

What is happening? The cinnamon slowly binds to the hot receptor TRPV1, where it stays. When you put the mint in your mouth, the menthol activates the cool/cold receptors (TRPM8 and TRPA1). Activation of these receptors turns off TRPV1, so you no longer taste the cinnamon. The amount of time that menthol stays on these receptors is short (= short residence time), so the menthol flavor disappears. Meanwhile, the cinnamon sits on its receptor for a longer time, and is still there when you take out the mint. Therefore, once the mint flavor is gone, the cinnamon receptor is no longer inhibited, so you can now taste the cinnamon again.

This is a good demonstration of what can happen when you eat a meal, with items that activate one or another receptor. Once you call your attention to this kind of progression, you will experience it again and again.

More about the talk I gave in the next post.

Tuesday, April 26, 2016

Why is white tea flavor so elusive?

Been thinking a lot about white tea, with its elusive flavor. Specifically, have been fascinated by the compositional differences between white tea and green tea: in fact there are few differences, yet white tea’s aroma is so much more elusive than that of green tea. Why?

To my delight, found a paper that includes white tea in its analyses (most only consider green, oolong, and black).* I thought I may have found the answer…

According to the data in this paper, the difference doesn’t seem to lie in the usual “green” compounds, such as hexanal, with its flavor of cut grass (yes, cut grass emits this chemical), nor in the citrusy flowery compounds such as E-geraniol and linalool, which help give both green and white tea their rose-like and citrus-like aromas. Rather, it seems that the critical difference lies in the presence of nerolidol in green teas and its relative absence in white teas.

Nerolidol has a green, citrus flavor, but most importantly it provides much of the woody flavor that we expect in tea. In perfumery, it is considered a base note that serves as a fixative for other aromas, while providing a subtle addition for blending flower and citrus top notes.** In other words, it is an aroma that has staying power, and may, if the data in this paper are correct, give green teas the flavor “heft” that is missing from white teas. One can almost think of it as a dial that increases the “volume” of the other green ingredients in the tea. 

So I looked further into nerolidol, and found a paper on the nerolidol content of teas, including white tea.*** In this analysis, white tea and green tea have virtually the same amount of nerolidol…so what to think now?

Neroli flower, from the bitter orange tree
(Citrus aurantium subsp. amara or Bigaradia —you can see the green fruit in the upper right corner). 
Nerolidol is one of the characteristic compounds in neroli flower oils. Image from

* Kunbo Wang, Fang Liu, Zhonghua Liu, Jianan Huang, Zhongxi Xu, Yinhua Li, Jinhua Chen, Yushun Gong, Xinghe Yang. Comparison of catechins and volatile compounds among different types of tea using high performance liquid chromatograph and gas chromatograph mass spectrometer: Catechins and volatile compounds. International Journal of Food Science & Technology, ISSN 0950-5423, 07/2011, Volume 46, Issue 7, pp. 1406 - 1412.

*** Chunhua Ma, Yanqin Qu, Yingxue Zhang, Bin Qiu, Yiru Wang, Xi Chen. Determination of nerolidol in teas using headspace solid phase microextraction–gas chromatography. Food Chemistry. Volume 152, 1 June 2014, Pages 285–290.

Friday, April 22, 2016

Some thoughts about cocktail drinking, whether tea-containing or not...

I’ve been often asked about how to create cocktails using tea, and what flavors of tea, spirits, and additions to use. I’m hesitant to answer this question, because I can’t make the experiment myself to see whether flavor reality matches theory. 

That's because I don't drink alcohol, because genetically I lack the enzyme to metabolize it,—even a little bit of alcohol makes me very ill.  Thus I can’t tell from personal experience what cocktails are like, and what would work well. Furthermore, in addition to being intolerant of alcohol for metabolic reasons, I am genetically programmed to be hypersensitive to the burn of alcohol…so all in all, I’m a tea-totaler!

That said, I did analyze some data about consumer cocktail and wine drinking for Tim Hanni, Master of Wine. 

People in the US were invited to fill out an on-line survey between December 2009 and March 2010. Out of 1485 usable responses, 324 preferred sweet wines, so we called them “Sweet” and 341 were tolerant of high alcohol dry wines such as Cabernet Sauvignon, so we called these people “Tolerant.”

From other data, we know that people in the “Sweet” group are more sensitive to the burn of alcohol, and also more sensitive to the bitterness of coffee. In fact, overall they tend to prefer tea to coffee, and if they drink coffee, 64% say they put some flavoring and/or creamer/milk in it, either always or sometimes.

By contrast, people in the Tolerant group, are generally less sensitive to the burn of alcohol can tolerate high alcohol dry wines, and are not put off by the bitterness of coffee. They prefer coffee to tea, and 67% of them drink their coffee black, with no additions.

When we looked at cocktail versus wine consumption, the difference between the two groups was striking. Here’s how they answer the question: “When you are at a bar or non-dining social function/party, what is your primary choice for an adult beverage?” The numbers are the percentages of respondents.

These differences are statistically significant for wine and cocktail drinking, with more than half of the Tolerant group opting for wine, and fewer than a quarter opting for cocktails. Slightly more than a third of the Sweet group chose either wine or cocktails. Considering that this survey was geared towards wine, this proportion of cocktail drinkers among the Sweet group may be an underestimation of the actual proportion among people who are highly sensitive to alcohol’s burn, and tend to prefer tea to coffee. 

To my thinking, these data help me understand why I’ve had so many requests for help with developing tea-containing cocktails!

You can find a summary of the study at:

Thursday, April 21, 2016

Should we enhance the flavor of black tea by treating the plants with an injury compound?

Plants that are injured release volatile compounds that not only help to protect the plant but also “warn” surrounding plants which then produce defense compounds themselves. In previous posts we have discussed the effects of thrips and jassids in enhancing the aromas of second flush Darjeeling and Oriental Beauty Bai Hao teas respectively. Apparently you can have some of the same effects without the intervention of insects, simply by spraying the tea with one of the plant’s defensive volatile compounds, methyl jasmonate.

Shi and colleagues sprayed tea plants with methyl jasmonate in an alcohol solution 24 hours before plucking the bud and two leaves to create the tea. Control plants were sprayed with water. (One wonders what the alcohol might have contributed…though it probably evaporated quickly).*

The gene expression for two enzymes important for black tea aroma increased in the treated tea: β-primeverosidase and polyphenol oxidase. β-primeverosidase releases linalool, linalool oxides, and geraniol from the sugars to which they are bound, thus releasing their aroma. Polyphenol oxidase is important for the formation of theaflavins and thearubigens. Thus an increase in these enzymes may be expected to increase the quality of a black tea.

Indeed, methyl jasmonate noticeably altered the profile of aromatics in the resulting teas, with an increase in linalool and its oxides, which have a characteristic floral aroma, and a decrease in 3-hexen-1-ol, which has a rather bitter “green” smell. Other “green” cool/cold receptor activating chemicals decreased as well, with the result that a panel of tasters found the tea from treated leaves to be more honeyed, sweeter, and more floral. By decreasing the number of compounds that activate the cool/cold receptors, there is less inhibition of the warm/hot receptors, so the black-tea qualities of the tea stood out.

The fact is that we treat tea leaves at and after plucking in order to activate the injury responses in order to yield a more aromatic tea. The question is: should we be nudging the process by “fooling” the leaves into response-to-injury mode before they are even plucked?

Image below: a thrip (tiny brown strip-like object) on a petal of the tea Camellia in Darjeeling—natural aroma creation through insect injury. From a fascinating article about autumn harvest Darjeeling at Hojo Tea:

* Jiang SHI, Li WANG, Cheng-ying MA, Hai-peng LU, Zong-mao CHEN, Zhi LIN.  Aroma changes of black tea prepared from methyl jasmonate treated tea plants. J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2014 15(4):313-321.

Tuesday, April 19, 2016

A graphic demonstration of the differences among green, oolong, and black teas.

Here’s a graph from the paper by Li-Fei Wang and colleagues that I mentioned in yesterday’s post.* It characterizes the aroma chemicals in tea through solid-phase microextraction–gas chromatograph (SPME–GC).  The top graph comes from a green tea (Ouksu tama-ryokucha, which is both steamed and roasted), the middle graph from a Dong-Ding (Tung-Ting) oolong, and the bottom graph from an English breakfast  (black) tea. The height of the peaks indicates the abundance of each chemical in the extract.

SPME–GC separates the chemicals that have a “green” quality—in other words chemicals that activate the cool/cold receptors—from the chemicals that activate the warm and hot receptors. I drew the green line to show the border between the two types of chemicals, primarily cool/cold on the left, and primarily warm/hot on the right.

As you can see, the green tea’s aroma chemicals (top graph) lie predominantly to the left of the line, which corresponds with the fact that green tea pairs well with foods that have a cool/cold aroma such as lemon, mint, and ginger.

As you go to the next graph, from Dong-Ding oolong, you’ll notice a marked difference from the green tea graph, with high peaks to the right as well as to the left. Note too, that there are far more high peaks altogether, consistent with the complexity of oolong flavor. Oolongs can be paired with both cool/cold and warm/hot aroma foods, though pairing is best with foods such as rosemary and strawberries that activate the primarily the warm (not hot) receptors. 

Black teas have very little of the green chemicals, and a somewhat lower quantity of the warm/hot aromas than oolong (bottom graph). The net effect is a far less complex flavor than that of the oolong, but the lack of cool/cold aromas means better pairing with foods with aromas that bind to hot receptors, such as raspberries and chocolate.

* Li-Fei Wang, Joo-Yeon Lee, Jin-Oh Chung, Joo-Hyun Baik, Sung So, Seung-Kook Park. Discrimination of teas with different degrees of fermentation by SPME–GC analysis of the characteristic volatile flavour compounds. Food Chemistry. Volume 109, Issue 1, 1 July 2008, pages 196–206.

Monday, April 18, 2016

Jasmine tea, fake and real…

Have been exploring a fine paper in the journal Food Chemistry by Li-Fei Wang and colleagues that characterizes the aroma chemicals in tea through solid-phase microextraction–gas chromatograph (SPME–GC).*

True high quality jasmine tea is made through a laborious process in which jasmine flowers come into contact with green tea multiple times. Each time, the used flowers are removed, the tea is allowed to dry down, and fresh flowers are added back. In the process the tea/flower pile can reach temperatures as high as 40-48º C, as chemicals in the flowers and tea together undergo metabolic transformations and exchange of volatiles.

A major transformation is the loss of the typical green tea compounds that impart a grassy green flavor. These include 2-methyl butanal, 1-penten-3-ol, and 2-penten-1-ol, and especially n-hexanal. The only major “green” compound left is cis-3-hexenol, which has a fresh, fruity green aroma. At the same time the amount of volatiles that bind to warm and hot receptors increases. Two compounds that smell of jasmine, linalool and benzyl acetate, are relatively increased as well. Further, compounds such as methyl jasmonate are formed in response to stress, which add to the flowery aroma. At the same time, the treatment increases oxidation of the tea so that the result is a tea that resembles an oolong or even a black tea.

Fake jasmine tea is spiked with jasmine chemicals in order to give a jasmine odor. It may smell of jasmine but the jasmine flavor is faint or absent in the brewed tea.

The reason is that “green” chemicals remain in fake jasmine tea after treatment, because no metabolism occurs. These chemicals turn on the cool/cold receptors and turn off the warm receptors to which most jasmine-related compounds, with the exception of linalool, bind. At the same time, these jasmine compounds try to turn off the cool/cold receptors. The net effect is a tea that smells of jasmine but tastes like a weak green tea. 

Here are chromatograms from this paper that show the differences among green, true jasmine, and fake jasmine teas. The peak just to the right of the green line is cis-3-hexenol.
  • The top chromatogram is from green tea. The peaks to the left of the green line correspond chiefly to the “green” compounds in the tea, which bind to cool/cold receptors.
  • The middle chromatogram shows the peaks in true jasmine tea. The peaks to the left of the green line are tiny, while the peaks to the right are much higher. These peaks correspond to jasmine aromas and flavors, and to the tea compounds from leaves with a higher degree of oxidation, akin to oolong or even black teas.
  • The bottom chromatogram corresponds to a fake jasmine tea. Many green tea peaks remain, and the peaks to the right of the chromatogram are much diminished in size relative to the ones in the true jasmine tea.

…that’s why the receptors in your mouth will know it when they meet a fake jasmine tea, even when your nose may be fooled!

* Li-Fei Wang, Joo-Yeon Lee, Jin-Oh Chung, Joo-Hyun Baik, Sung So, Seung-Kook Park. Discrimination of teas with different degrees of fermentation by SPME–GC analysis of the characteristic volatile flavour compounds. Food Chemistry. Volume 109, Issue 1, 1 July 2008, pages 196–206.

Sunday, April 17, 2016

Why do you prefer certain perfumes for yourself?

Before we answer this question, a little about how vertebrates, including humans, choose mates. In vertebrates there is a complex of genes called the major histocompatibility system. This system determines what we will recognize as self and non-self, and governs our immune system, in other words which foreign lifeforms, such as microbes and transplants, we will reject. 

In humans, this system is called the human leukocyte antigen system, or HLA. When a person needs a transplant, he or she gets one that is matched to his or her HLA type, in order to decrease the chances of rejection of the transplant as non-self.

By contrast, it is to a person’s advantage to find a mate who is of a different HLA type, because their offspring will be able to resist a larger number of diseases. Further, if your mate differs from you in HLA type, you are less likely to be related to each other, so inbreeding is decreased. 

The molecules that make up our HLA type are shed into the environment as we shed skin cells, particularly in our sweat. Dogs use the scent of the HLA we shed to track people. Humans can also sense HLA types subliminally, though we don’t know the mechanism behind this ability.

Which brings me to the association between scent preference and HLA. 

First, do we choose perfumes in order to send identification signals? Havlíček and colleagues performed a curious experiment, in which people were asked to discriminate repeatedly among sweat samples from donors.* Each repetition included one sample from one donor (“odd-ball” sample) and two samples from another, and the task was to pick the sample that is different from the other two. For one set of samples, no fragrance was added to the sweat; for another set, a random fragrance was added to the samples; and for the third set, all three samples were spiked with the fragrance preferred by the person giving the “odd-ball” sample. While the participants were able to distinguish the odd-ball sample at a greater than chance rate in all the tests, the success rate was greatest when the sample was spiked with the fragrance preferred by the donor of the sweat. The authors of the study suggest that “naturally occurring variance in body odor is more preserved when blended with fragrances that people choose for themselves, compared with other fragrances. Our data are consistent with the idea that fragrance choices are influenced by fragrance interactions with an individual's own body odor.”

Next, is it our HLA types that guide us in some way in our choice of perfumes? Not every HLA type has been studied for this characteristic, because there is a huge number of possible HLA types. However, there are two tissue types that are very common in European populations, HLA-A1 and HLA-A2, which Wedekind and Füri for testing fragrance preferences.**

Here are the perfume preferences associated with each of these HLA types (derived from the graphs in Wedekind and Füri’s paper):
  • HLA-A1: Like: vetiver, rose, violet, bergamot, tuberose; Dislike: musk, amber, castoreum, amber.
  • HLA-A2: Like: sandal, vetiver, moss, cardamom, cinnamon, musk, amber, tolu, styrax, labdanum; Dislike: orris, rose, cedar, violet, tuberose, bergamot.
Shalimar perfume—probably the choice of women with HLA-A2.
Image from Wikipedia.

As you can see from these two lists, the likes and dislikes of HLA-A1 and HLA-A2 are practically the opposite in their qualities. The scents preferred by people with HLA-A1 are considered light, floral or citrus, and green. These scents activate the cool and cold receptors. By contrast, the scents that HLA-A2 carriers prefer are woody, earthy, and spicy, and activate the warm and hot receptors.

I’ve been typed, and I received HLA-A1 from one parent and HLA-A2 from the other. My choices of scents for myself are in HLA-A1 list, and not the HLA-A2 list, yet I carry both sets of HLA types. How can this be? 

I don’t know, but I have a guess. It comes from a study of women’s preference for male odor.*** Women who were typed and whose parents were typed were asked to smell odors obtained from men whose HLA types matched one of the women’s parents, and from men who did not match either parent. Importantly, the women did not know and did not recognize that the odors came from a human being, so the odors could be considered the equivalent of perfumes. (The title of the paper is misleading on this point.) The result: “A woman's preference was associated only with matches to inherited paternal alleles and not to the paternal alleles that she did not inherit.” In other words, if these samples had been perfumes she would have picked the scents for herself that matched the HLA type she inherited from her father.  

If this finding holds true for perfumes, I could say that in all likelihood I inherited HLA-A1 from my father, and the HLA-A2 haplotype from my mother.

Interestingly, my mother didn’t like the aromas in the HLA-A2 list very much either and in fact disliked musk intensely, while her mother (my maternal grandmother) liked perfumes with the aromas listed for HLA-A2 very much—until a bottle of Shalimar broke in her suitcase!***** So my guess is that my mother inherited HLA-A2 from her mother, and I inherited it from my mother! 

Oh, by the way, HLA type may dictate wine preference, too!****

* Allen C, Havlíček J, Roberts SC.Effect of fragrance use on discrimination of individual body odor. Front Psychol. 2015 Aug 7;6:1115. doi: 10.3389/fpsyg.2015.01115. eCollection 2015.

** Claus Wedekind, Sandra Füri Evidence for MHC-correlated perfume preferences in humans. Proc. R. Soc. Lond. B 1997 264 1471-1479; DOI: 10.1098/rspb.1997.0204. Published 22 October 1997

*** Suma Jacob, Martha K. McClintock, Bethanne Zelano, & Carole Ober. Paternally inherited HLA alleles are associated with women's choice of male odor. Nature Genetics 30, 175 - 179 (2002)

**** Nicola Pirastu, Maarten Kooyman, Michela Traglia, Antonietta Robino, Sara M Willems, Giorgio Pistis, Najaf Amin, Cinzia Sala, Lennart C Karssen, Cornelia M van Duijn, Daniela Toniolo and Paolo Gasparini. Genome-wide association analysis on five isolated populations identifies variants of the HLA-DOA gene associated with white wine liking. European Journal of Human Genetics (2015) 23, 1717–1722; doi:10.1038/ejhg.2015.34; published online 11 March 2015.

***** Shalimar has a bergamot top note, but otherwise is considered a “soft amber” perfume, with “Oriental” notes of incense and vanilla.

Saturday, April 16, 2016

New Fragrances with Tea: Part 4.

This is neither the last nor the least in my series on the Rare Tea Collection from J. Malone:

Golden Needle Tea – A special black tea with leather, spice, sandalwood and benzoin resin notes - 
One of the aspects of the golden needle tea aroma that is missing from the added chemicals is the tea’s marvelously fruity bouquet, which has a raisiny, and to my mind, mango-like, quality. It seems the perfumers wanted to emphasize the tea’s more spicy, warm, sweet, woody, resiny qualities, and left the fruit behind. 

Golden Needle Tea or Dian Hong — 
note the golden color of the buds and the hairy quality of the leaves.
Image from Wikipedia.

Which makes some degree of sense, actually, even though the perfume won’t be true to the tea. From a perfumer’s perspective, this Golden Needle Tea perfume combines woody and leather fragrances, which are heavy on base notes, in other words scents that last for hours. These scents tend to hit the hot receptors, and as a consequence give a sensation of warm sensuality. Were you to add the fruity smells of the tea, they would add a fleeting top note. As people pick their perfumes based on the top notes (what you smell in the bottle, but don’t last on the skin), the relative lack of top notes in this perfume means that you will have to sniff a lot before you sense it, but when you do you will get these warm sensations.

In my next post, I’ll explain why you might like this or prefer another of these perfumes, and also why it is that what you like in a food and a food pairing may be quite different from what you would like in a perfume.

Friday, April 15, 2016

New Fragrances with Tea: Part 3.

The last two of the six fragrances in the Rare Teas Collection* are perhaps the oddest. Today I’m going to talk about the one based on pu’erh, and tomorrow the one based on Yunnan Golden Needle Tea. The reason they are odd is that the perfumers chose some curious ingredients, that can be expected to shift the aromas expected from the teas.

Midnight Black Tea – Pu’erh with vanilla, guaiac wood and labdanum 
Just as with the Oolong perfume, we have to ask which pu’erh did the perfumers choose. Pu’erh comes in two main types: raw and ripened. Both types are made from the broad leaf assamica variety of Camilla sinensis, and for both older as well as younger leaves are picked. The main difference between the two types of pu’erh lies in their handling.  

Raw pu’erh is treated so that initial enzymatic activity due to bruising of the leaves (with the formation of polyphenols) is only partially inhibited, unlike what is done with Chinese green tea. The resulting “maocha” is aged at room temperature, usually in a compressed form, for a matter of years during which time fermentation occurs and the unique aroma profile of each tea develops. 

Ripened pu’erh was designed to shorten the time to readiness of the tea, and involves allowing the maocha to ferment in a warm humid space for about 2 months, where the leaves are turned, dampened, and piled up again repeatedly, much like compost.  Once the tea has fermented, it is. ready for compressing and aging

As you can imagine, there are important differences in the chemistry of these two types, differences so important in fact that the authors of a very recent paper could successfully distinguish between the two types using an “electronic nose.” ** The main differences they found can be divided into two sets. 

The first set has to do with compounds that activate the cool/cold receptors. Raw pu-ehr has 8 times more linalool than does ripened, and also more beta-cyclocitral, which has a minty smell, and safranal, which is green/herbaceous and sweet.

The second set consists of methoxyphenolic compounds—in fact the methods used by the authors could distinguish between the two types of tea based on these compounds alone, with ripened pu’erh having several fold higher levels than raw. And what is the smell of these compounds, you ask? Stale stale stale. Not a smell that you would like in your perfume. In fact people often allow ripened pu’erh to air out for a while to allow these compounds to dissipate before brewing their tea.

Given these differences, it is likely that the perfumers who created Midnight Black Tea used a raw pu’erh in their formulation, but may have missed the opportunity to create a fresher scent (at least for top notes) by not picking up on the compounds in the tea that reach the cool/cold receptors. Or maybe they didn't want that possibility. The rest of the formulation suggests that they didn't.

The exquisite flower of the Cistus ladaniferus plant, from

Labdanum, a resin from the Cistus ladaniferus shrub is a substitute for forbidden ambergris from sperm whales, because it has a tenacious, musky odor. Interestingly, both forms of pu’erh have a significant amount of dihydroactinidiolide, a compound with a musky smell, so tea and labdanum together would shift the aromas to the more warm side of the temperature spectrum. In perfumery, labdanum is frequently combined with vanilla to give a sweet (in my opinion, cloying) aroma—it is combined with vanilla in this perfume as well. 

Guaiac wood also falls into the category of warm, sweet smells with a woody, earthy quality. Pu’erh contains a compound, beta-guaiene, which (as you may have guessed from its name) was first isolated from guaiac wood, and which has a woody, balsamic aroma. The perfumers were probably aiming to enhance that aspect of the pu’erh by adding guaiac wood to the mix.

The result, I imagine, is a perfume heavy with lingering odors, with little if any of the lighter floral and minty aspects of raw pu’erh—a sweet warm musky smell, in other words, often characterized as “oriental” or “amber.” 


** Jing Ye, Wenguang Wang, Chitang Ho, Jun Li, Xiaoyu Guo, Mingbo Zhao, Yong Jiang, and Pengfei Tu. Differentiation of two types of pu-erh teas by using an electronic nose and ultrasound-assisted extraction-dispersive liquid–liquid microextraction-gas chromatography-mass spectrometry. Anal. Methods, 2016, 8, 593-604.

Thursday, April 14, 2016

New Fragrances with Tea: Part 2.

Here are two more of the tea-derived perfumes from the Rare Tea Collection by J. Malone*:

Jade Leaf Tea – Sencha tea with sesame, pomelo and frankincense  
Sencha tea has very “cool” flavors, so pomelo, which has compounds that hit the cool/cold receptors makes sense in this blend. Frankincense has a mixture of compounds: some activate both the cool and cold receptors (for example α- and β-pinene, limonene, and β-phellandrene) and at least one, incense acetate, activates TRPV3, the warm receptor. At the same time, almost all the compounds in sesame activate the warm and hot receptors. My guess is that the initial experience of this perfume will be more green-tea-like, but that top note will quickly dissipate to leave you with a warmer aroma, more reminiscent of a church or temple.
Frankincense, from Wikipedia.

Oolong Tea – Oolong, bitter cocoa, tobacco and tonka bean   
This combination surprises me. My guess is that the perfumers must be using a dark oolong such as Da Hong Pao from the Wu Yi mountains. This tea is highly oxidized for an oolong, and has a chemical profile much more like a black tea. Da Hong Pao has a roasty sweet floral aroma, which is associated with activation of the warm and hot receptors, unlike other oolongs, which activate primarily the warm receptors. Cocoa and tobacco also activate the warm and hot receptors, the hot receptors more strongly than the warm. Tonka bean activates primarily the warm receptors. My guess is that the aroma of the oolong complements these warm/hot aroma palette, but because the aromas of the cocoa, and particularly of tobacco and tonic bean are far from delicate, the oolong will be unrecognizable as such.

Tomorrow we will look at the last two perfumes in this collection, and, as I have done above, discuss what I think is the rationale for the choices of ingredients. So more to come!


Wednesday, April 13, 2016

New Fragrances with Tea: Part 1.

Just saw on World Tea News that Jo Malone fragrances are offering a new series of six tea-based perfumes called The Rare Tea Collection.*

Here’s the description of two of the perfumes in the Collection with my comments, with the rest coming and the next day—note that I have not smelled any of these, so my comments are a guess: 

Silver Needle Tea – White tea with musk, mimosa and rose. 
The odor profile of white tea includes rose notes, so this combination should work well, especially as there aren’t any woody smells, which would be expected to hide any tea aroma completely. Musk is the base note, and gives a  “warm” quality, as does mimosa, which is probably the middle note, and has the smell almost like baby powder. Depending on which scent version is used, the rose may either give a cool top note or a warm one. 

Darjeeling Tea – Black tea with notes of jasmine, freesia and davana. 
This is a fascinating combination. As you probably know, Darjeeling is one of my favorite teas. It is really an oolong, rather than a black tea, and has a great deal of jasmine’s chemicals in it, so jasmine is a natural choice. Freesia provides a curious top note, with a lot of “cool” chemicals, most particularly linalool, which is also abundant in Darjeeling (and virtually all other) Camilla sinensis teas. In fact many consider the linalool scent to be particularly characteristic of freesia.

Davana. Image from

Darjeeling second flush (which they probably used) also has a grape aroma. If so, this is why the perfumers chose davana.  Davana, or Artemisia pallens, has a raisiny smell, courtesy of compounds called davanones. I'm guessing that's why it combines well with Darjeeling. Davana also contains linalool as a minor component, but has large amounts of nerol and geraniol. Together with linalool, these three are the main components of the cool rose smell. Davana also has a strange quality, namely that it smells differently on different people, so who knows what this tea perfume would smell like on your skin?

Two more perfumes from this Rare Tea Collection in my next post!


Tuesday, April 12, 2016

Speaking of color illusions and smell, here is a fascinating experiment:

Participants sniffed bottles of colored (red, green, and yellow) and colorless water while blind-folded and while not. Each nostril was tested separately. Here are the results:

Although no thermal stimulus was present, subjects reported thermal sensations, but only under free viewing conditions. The nature of these sensations depended on the color of the solution, with green inducing cooling and red warming sensations.” Graph from Michael and Rohlion.*

When participants could see the color of the water, they had the illusion of temperature sensations just as if the trigeminal nerve in the nose had been activated. In a follow-up study, when participants were just looking at a bottle of colored water while sniffing a bottle of plain water that they couldn’t see, these researchers saw the same effect, suggesting that the mere presence of a colored water can cause the brain to perceive an illusory state.**

Remember how the oenology students perceived the white-wine-dyed-red  as being spicy and peppery? Spicy and peppery are sensations brought about normally by activation of the TRPV1 “hot” receptor of the trigeminal nerve in the nose. Which raises the question: did these students in fact experience this color-temperature illusion, and—given the lack of the word “warm” in the vocabulary they had to use—chose “spicy” and “peppery” instead?

It appears that the association of color with temperature is so strong in the brain, that we subconsciously ascribe temperature to colored substances. How this association comes about, in other words whether it is innate or learned, is unknown. In favor of innate is the failure to make the same associations between the learned odor of lavender and lavender the color. 

My overall conclusion: rather than laugh at wine experts for being fooled and conclude that wine experts are not truly experts, we should realize that we all make the association between color and other sensory modalities to a greater or lesser degree. The effect is something we seem not to be able to avoid, whether we are expert or naïve.

* George A. Michael, Pauline Rolhion. Cool colors: Color-induced nasal thermal sensations. Neuroscience Letters. Volume 436, Issue 2, 9 May 2008, Pages 141–144.

** George A. Michael, Hélène Galich, Solveig Relland, Sabine Prud’hon. Hot colors: The nature and specificity of color-induced nasal thermal sensations. Behavioural Brain Research. Volume 207, Issue 2, 5 March 2010, Pages 418–428

Friday, April 8, 2016

From a Haha moment to an Aha moment...Part 4:

In my previous post, I mentioned that I was thinking about pairing in terms of color as well flavor, with the thought that color influences perceived flavors through an association process with temperature. Thus we tend to think of red or orange as hot and green or blue as cold. 

In the course of looking up experimental data, I found the following infographic on chili peppers:

Notice how the peppers with the highest Scoville numbers, that are hottest and the richest in capsaicin, are all red. As you go down the infographic to lower Scoville numbers, green colored peppers appear more and more, though the bottom row still contains red-colored peppers.

Capsaicin activates the trigeminal TRPV1 receptor, which is also activated physical temperatures higher than about 40ºC/102ºF. That's why we say that these peppers burn—the brain automatically translates activation of the TRPV1 receptor as some level of burn.

The same activation pattern holds for tea: green tea, which activates the cool and cold receptors, is lighter and greener, while black tea, which activates the warm and hot receptors, is darker and more reddish-brown.

In the white-wine-dyed-red experiment we have been talking about, the most prominent descriptors chosen for the dyed wine were spicy and pepper, which would correspond to activation of the trigeminal TRPV1 receptor. Indeed components of true red wine do in fact activate this receptor, while white wines tend to activate receptors that correspond to cooler temperatures. Yet sniffing the white-wine-dyed-red made the students* experience the TRPV1-type smells. What is happening here at the level of the nervous system?

In upcoming posts I’ll explore some amazing data on color/temperature/odor/flavor associations and the nervous system, and what these data mean for pairing. My next post will be coming on Tuesday—my family is coming for the week-end so I won't have time to write until then.

* They were students of oenology, not full wine experts as the popular press has suggested.

Wednesday, April 6, 2016

From a Haha moment to an Aha moment...Part 3:

In this, the third in a series of posts about the white-wine-dyed-red study, we’ll explore what the students may have actually experienced. As I mentioned in my previous post, Morrot and his colleagues developed a vocabulary of descriptors for white and red wine, and analyzed these words to see how they grouped.* The results led to the conclusion that there was a set of specific words that were most often used to describe white wines, and another set that were most often used to describe red wines. We also noted that Jacques Dupont, wine commentator for La lettre de Gault & Millau, used 16 words exclusively for white wines and only 2 exclusively for red wines, suggesting that descriptors for red wines can also be used for white, but not so much vice versa.

The 54 oenology students who participated in the study were given the Dupont list of words, and were allowed to develop their own vocabulary, using Dupont’s words plus their own, in order to describe the two wines—a Sémillon and Sauvignon blend for the white wine, and a Cabernet-Sauvignon and Merlot blend for the red.  

The students then came back at a later occasion to use these individual vocabularies to describe what they smelled in the white wine and the white-wine-dyed-red.**

Morrot and colleagues present a graph (Figure 2 in the paper—see below) with the descriptors the students chose for white-wine-dyed-red and the white wine, showing how many participants chose each word. Far and away the most commonly used word for the plain white wine was “floral:” 15 students assigned this word to the white wine and 2 assigned to the white-wine-dyed-red. Dupont used this descriptor for white wine 93% of the time, and for red 7%.  

The popular press focused on the use of red fruit words to describe the white-wine-dyed-red, and indeed some participants did in fact use words for red fruits, such as raspberry (10 people), strawberry (6 people), and cherry (7 people). Further, the words the students used for the white-wine-dyed-red were a subset of the words they each had used to describe red wines in the true red wine experiment.

Here are my thoughts:

First, the experimenters made an interesting choice of wine for this experiment: both wines contained Sauvignon grapes. The white wine was a blend of Sauvignon with Sémillion, and the red wine a blend of Cabernet-Sauvignon and Merlot. Both wines therefore can be expected to have contained methoxypyrazines, the key flavor ingredient of Sauvignon wines, both white and red. Methoxypyrazines give the characteristic aroma of green peppers, and stands out in a mixture (a flavor I hate, as I have mentioned in the past). Most people in the US are used to experiencing this aroma in the context of a red wine, where Cabernet Sauvignon is more common than Sauvignon Blanc. I believe this is true in France as well, so the association of this flavor with red wine would be stronger than its association with white. When I saw the types of wine, then, I began to wonder in which direction were the students fooled: were they fooled into thinking the white-wine-dyed-red was actually red wine, or were they assuming the white wine was white because of its color and therefore assigned it white wine descriptors, and that the red wine was the “correct” color for its aroma?

Next, in line with the study of vision-odor detection described in Part 1 of this series, people will tend to detect the presence of odors when the image that they see is congruent with the odor, and fail to smell odors that are incongruent. The two wine blends are not all that different from each other with respect to their chemistry, so the participants would have preferentially detected odors that matched what they were seeing in the glass.

Or was there an effect of the addition of anthocyanins to the white wine that physically altered what the students could detect? The authors of the present study said no, but there are plenty of studies where the non-volatile matrix of a red wine (particularly the anthocyanins), when added to a white wine, make the wine's odor more red-wine-like. One recent study, in particular, combined non-volatile matrices from two white wines (M1 and M2) and two white wines (M3, M4, and M5) with the volatile fractions of these wines (A1 through A5), and had trained panelists sniff these combinations:
"...replacing the original nonvolatile matrix (M1A1) by a nonvolatile matrix from a second white wine (M2A1) has a small sensory effect, but that replacing it by a red wine nonvolatile matrix (M3A1, M4A1, or M5A1) has a deep effect on the sensory properties of the reconstituted wine, changing its aroma to terms related to the “red fruit” aroma family in detriment to the terms typical of white wines observed in the M1A1 and M2A1 samples." ***
This effect was most pronounced for the red wine matrix (M5) that was particularly astringent, and therefore (presumably) contained more anthocyanins. The students in the white-wine-dyed-red study were not trained panelists, and they probably were not aware of the consequences of the addition of anthocyanins when smelling the wines "blind," but with the addition cue of color, they were able to pick out the very real differences in aroma.

What caught my attention even more strongly, however, was that the two most commonly used words for the white-wine-dyed-red were “spicy” (18 people and 2 people for the white wine) and “pepper”**** (16 people and 4 people for the white wine). Neither of these words applies to a specific red fruit; instead they apply to the sensations we get from activating TRPV1, the “hot” receptor, from alcohol. In the study combining matrices with volatiles, adding a red wine matrix to white wine volatiles also enhanced the spiciness and woodiness of the reconstituted wine.

Which raised the question in my mind (and one of my Aha! moments): is the major association of color with odor—in this experiment as well as more generally—not so much based on specific odors, but rather on the trigeminal sensations these odors elicit? 

And is the reason these colors elicit these aromas that, in our day-to-day experience, red plant substances such as anthocyanins actually shift the available odor profile to one that activates the warm and hot receptors?

And if so, can we then develop a scheme for pairing that involves the use of colors to bring out certain flavors?  More about answering this question on Friday, with my next post.

* Morrot G., Brochet F., Dubourdieu D. The Color of Odors. Brain and Language, doi:10.1006/brln.2001.2493. 

** I may have mentioned this before…there is some confusion in the paper concerning whether the students tasted or only smelled the wines. After multiple rereadings, I believe that they only smelled the wines. 

*** María-Pilar Sáenz-NavajasEva CampoLaura CulleréPurificación Fernández-ZurbanoDominique Valentin and Vicente Ferreira. Effects of the Nonvolatile Matrix on the Aroma Perception of Wine. J. Agric. Food Chem.201058 (9), pp 5574–5585. DOI: 10.1021/jf904377p

**** The French distinguish between poivre=pepper and poivron=the pepper fruit (green or red pepper). The word used here was “poivre.” (Incidentally, neither of these words occurred among the most commonly used for red wine in the Dupont list.)