Note on recursion relations for the cut representation
Abstract
In this note, we study the cut representation by combining it with BCFW deformation. As a consequence, the oneloop integrand is expressed in terms of a recursion relation, i.e., point oneloop integrand is constructed using treelevel amplitudes and point oneloop integrands with . By giving explicit examples, we show that the integrand from the recursion relation is equivalent to that from Feynman diagrams or the original cut construction, up to scale free terms.
Keywords:
Scattering Amplitude, Loop Integrand1 Introduction
In a recent work, a new representation of the perturbative matrix, known as cut representation, was proposed Baadsgaard:2015twa (). It allows one to write the integrand of loop amplitude as summation of products of lowerpoint treelevel amplitudes with deformed loop momenta. For generic point oneloop integrand with all massless external legs, the new representation takes the form,
(1) 
where , with , . As will be reviewed shortly, two deformations have been applied to the loop momentum : firstly the dimensional deformation with in extra dimensions, and secondly the scale deformation . The details of the oneloop cut construction was further clarified in Huang:2015cwh (), and generalizations to two loops or more was also illustrated in Baadsgaard:2015twa (). The cut representation circumvented two difficulties in the attempt for recursive construction of loop integrand: canonical definition of loop momentum and the singularities in the forward limit (which will be referred to as forward singularities). On the other hand, the integration over loop momentum with such integrand still requires more systematic investigations.
The cut representation was partly inspired by the work Geyer:2015bja () and finds direct application in the study of writing oneloop amplitudes based on the Riemann sphere Geyer:2015jch (); Baadsgaard:2015hia (); He:2015yua ()^{2}^{2}2In the scattering equation formalism Cachazo:2013gna (); Cachazo:2013hca (); Cachazo:2013iea (); Cachazo:2014nsa (); Cachazo:2014xea (), loop integrands for supergravity and superYangMills amplitude has formerly been proposed Geyer:2015bja (), since in these theories there is no forward singularity., and very recently in an extension to twoloop supersymmetric amplitudes from Riemann sphere Geyer:2016wjx (). Another work also reports similar oneloop integrand expansion while investigating elliptic scattering equations at oneloop level Cardona:2016bpi (), based on an earlier work on the scattering equation Gomez:2016bmv (). The idea in the cut construction also inspires some thoughts in the other approach of constructing oneloop amplitude Cachazo:2015aol (), as well as the construction of twoloop planar integrand of cubic scalar theory Feng:2016nrf (). These works have shown the universality and importance of cut representation for loop integrands in general.
After the discovery of BrittoCachazoFengWitten(BCFW) recursion relations for treelevel amplitudes Britto:2004ap (); Britto:2005fq (), it is very natural to ask if one can construct loop integrands in a similar, recursive way. The key for the progress lies in expressing planar loop integrands from forward limits of tree amplitudes CaronHuot:2010zt (); ArkaniHamed:2010kv (); Boels:2010nw (), which has been very successful for cases without forward singularities, such as superYangMills at one loop and planar SYM to all loops ArkaniHamed:2010kv (). However, for general theories the aforementioned difficulties have only been resolved in the cut construction. These works have indicated clearly that for generic loop integrands, BCFW deformation has to be applied with extra care, especially due to the presence of forward singularities. In the cut construction, the dimensional deformation transforms oneloop integrand into tree diagrams, while the scale deformation has avoided the forward singularities by excluding the tree diagrams that corresponding to oneloop tadpole and massless bubble contributions, which should not be presented in the final amplitude.
Both recursion relations and cut approach to the construction of loop integrands in general theories are promising but with some unsatisfying features: the cut representation has nonstandard propagators, while it is not clear how to remove forward singularities in general in recursion relations. Thus it is natural to see if by combining the two methods to make further progress. In this note, we will initiate the study along this direction. We would like to see if there is another way to deal with forward singularities and how much can we learn about the structure of oneloop integrands from both recursion and cut viewpoints.
This paper is structured as follows. In §2, we illustrate the application of BCFW deformation in the cut construction, and present a recursive formula for oneloop integrand. In §3, we explain the details of the recursive formula by three examples, and confirm the validity of the results by comparing with results from oneloop Feynman diagrams and those from the cut construction. We conclude in §4.
2 The derivation of recursion relation
Let us first recall the original derivation of cut representation in Baadsgaard:2015twa (). After imposing the dimensional deformation as well as the shift for loop momentum, the point oneloop integrand becomes essentially the point treelevel amplitude , on the condition . Then by scale deformation , and by removing diagrams that contribute to oneloop tadpoles and massless bubbles appropriately, one gets the oneloop integrand. Since BCFW recursion has been applied to the computation of ordinary treelevel amplitudes, this naturally motivates us to consider the possibility of constructing the point treelevel amplitude using the recursion. Here we present a derivation of the recursive representation for oneloop integrand following the aforementioned motivation. The derivation will take three steps, as follows.
2.1 Step one: dimensional deformation
Just like the original cut construction Baadsgaard:2015twa (), the first step of the derivation is to reformulate oneloop integrand in terms of treelevel amplitudes. We take the same dimensional deformation as in Baadsgaard:2015twa () and also the loop momentum shifting, to arrive at
(2) 
Some explanations are in order for (2). Firstly, from the dimensional deformation, it is known that is given by those Feynman diagrams with external legs and two extra legs by cutting an internal propagator. Thus is defined on the condition , which says that all in should be understood as the null momentum in higher dimension. Furthermore, is not exactly the full point treelevel amplitude, since in order to reconstruct the oneloop integrand, some diagrams should be excluded. Such treelevel diagrams correspond to oneloop tadpole and massless bubble diagrams with single cuts. From Feynman diagrams one can inspect that, a tadpole after single cut will produce tree diagrams with attaching to the same vertex^{3}^{3}3Here denotes two legs by breaking an internal line., while massless bubble diagram with the massless leg after single cut will produce tree diagrams with (or ) attaching to the same threepoint vertex, and then meeting (or ) in the neighboring vertex. The above scenery would help us to exclude corresponding tree diagrams in the following steps.
Next let us take a look at the contributing tree diagrams to . If the theory under consideration is not colorordered, we shall consider the full point onshell treelevel Feynman diagrams after removing those corresponding to the oneloop tadpole and massless bubbles. While if it is colorordered, the gets contribution from different colorordered tree diagrams, each by breaking an internal line of the propagators. Since there are different color orderings, we can calculate each one independently, for example, using different methods (such as Feynman diagrams or BCFW recursion relations) or different deformations in BCFW recursion relations.
A final remark says that, the loop momentum shifting in expression (2) makes a canonical definition of loop momentum, such that the integrand is irrelevant to the labeling of for internal propagators.
2.2 Step two: BCFW deformation
Now let us turn to , and our aim is to determine it by BCFW deformation. Since it is effectively treelevel amplitude but with forward singularity removed, the analysis on the large behavior would be the same and the computation should be straightforward. Let us, for generality, take two arbitrary momenta (but not ) and perform the standard BCFW deformation
(3) 
Such deformation can be realized when the dimension . In this case, becomes an analytic function of external momenta ’s, loop momentum and a complex variable . As usual, we can consider the contour integration
(4) 
where the contour is a very large circle. This integration leads to
(5) 
where the sum is over all finite pole ’s of , and is possible boundary contribution. It is wellknown for treelevel amplitudes that for YangMills and gravity theories, the BCFW deformation can be chosen such that the boundary contribution vanishes. While for some other theories, the boundary contribution would appear and require more careful analysis ArkaniHamed:2008yf (); Cheung:2008dn (); Feng:2009ei (); Jin:2014qya (); Jin:2015pua (); Feng:2014pia (); Feng:2015qna (); Cheung:2015cba (); Cheung:2015ota (). Here we shall assume for simplicity (but the similar consideration can be generalized to the case with nonzero boundary contributions). Thus the only information we need for computing by means of expression (5) is the pole structure of function .
The BCFW deformation splits a tree amplitude into two parts, with the shifted momenta locating in each part. Assuming is the sum of all momenta in the part containing , and . From we get . Now let us consider the two extra legs . If they are in the same part, will have no dependence on , thus also the pole . We shall denote the corresponding contribution as . While if are separated in two parts, as well as would depend on . We shall denote the corresponding contribution as . So we have
(6) 
For the contribution , we can further organize it into two parts,
(7) 
denotes the contribution where legs are in the part containing , while denotes the contribution where legs are in the part containing . Explicitly, we have
(8) 
where
as well as , , and . Similarly,
(9) 
where
Note that the sum is over all possible splitting of legs and helicities. Also note that inside the bracket we have explicitly labeled all the legs in each part but not the ordering of legs. The colorordering of legs should be understood with respect to their corresponding theories.
Now let us take a more careful look on expressions (8) and (9). Firstly, the part in will be lowerpoint onshell tree diagrams after excluding those corresponding to tadpole and bubble diagrams. This means that when dressing with , they would become lowerpoint oneloop integrand, which can be obtained by any legitimate methods, such as the original cut construction or Feynman diagram method with partial fraction identity. One important implication is that the forward singularities in the type have been automatically removed after using the welldefined oneloop integrands of lower points. Secondly, for , the number of legs in set must be at least one, in order for the amplitude to be nonvanishing. Naively, the number of legs in set could also be zero. However, when it is so, the tree diagrams of are exactly those corresponding to tadpole and massless bubbles, which need to be excluded. So could not be empty set. Similarly for , the number of legs in sets should at least be one.
Now let us analyze the contribution . We can also organize it into two parts,
(10) 
denotes the contribution where leg is in the part containing , while denotes the contribution where leg is in the part containing , explicitly as
(11) 
where
and . While
(12) 
where
Some discussions are in order for expressions (11) and (12). Notice that we have used instead of treelevel amplitude , since in this stage potential contributions coming from corresponding to tadpole and bubble diagrams in should be excluded. Recalling our discussion on the excluded diagrams in the previous subsection, we can conclude that, since are separated into two parts, there could not be diagrams corresponding to oneloop tadpoles, while diagrams corresponding to massless bubbles^{4}^{4}4We need to distinguish massless bubble from massive bubble. The latter is allowed for oneloop diagrams. do exist in and when the set or is empty. In other words, forward singularities corresponding to tadpoles have been avoided in type . Combining the discussions for type , we see that we can remove forward singularities corresponding to tadpoles without using scale deformation as is done in the cut construction. However, forward singularities that corresponding to massless bubbles are more difficult to deal with and we will organize into three contributions
(13) 
denotes the contribution of the case when both and are not empty, so forward singularities corresponding to massless bubbles will not appear and there will be no excluded diagrams. Thus the is exactly the tree amplitude and we have
(14) 
where the sum is over all helicities and possible splitting of external legs with the length of set satisfying . This is to ensure that there is at least one leg in set .
denotes the special case when set . In this case, becomes a threepoint amplitude, and we get explicitly
(15) 
where
and .
denotes the special case when set . In this case, becomes a threepoint amplitude, and we get explicitly
(16) 
where
(17) 
and .
Similarly, we can also organize into three parts,
(18) 
just as it is defined for , but changing . Explicitly, we have
(19) 
and , .
There is an important observation. If we consider the colorordered integrand, we can choose the deformation pair such that are not nearly with the deformed momenta. Thus the contributions of , , and do not exist. As we will discuss in the following subsection, the remaining forward singularities that corresponding to massless bubbles are exactly in those four terms. In other words, with a proper choice of deformation pair, we can naturally avoid forward singularities without further using the scale deformation.
2.3 Step three: scale deformation
In the previous subsection we have expressed as
(20) 
where given in expressions (8), (9) respectively, and , with given in expressions (14), (15), (16), and by changing of . In each expression there would be functions, and we should identify them. The functions are determined by removing tree diagrams that corresponding to tadpole and massless bubbles. In the previous subsections, we have presented some discussions on this point, but the complete resolution will be provided in this subsection. In fact, as we have pointed out, the only left forward singularities are those in terms and . To deal with them, we use the scale deformation.
Before giving a careful discussion, let us take a look on , . When multiplying with in (8), (9), it trivially becomes oneloop integrand of the original cut representation with BCFWdeformed momenta. Thus we can identify them as
(21) 
where , . Similarly,
(22) 
where , . Here ’s are lowerpoint oneloop integrands from cut representation, and ’s are lowerpoint tree amplitudes. In fact, the oneloop integrand in (21) and (22) does not need to be in cut representation, i.e., any representation, such as the one obtained by Feynman diagrams, should be fine. Thus these two terms can be expressed as summation over products of lowerpoint oneloop integrand and tree amplitude. For other two terms , it has already been shown in (14) that they are summation over products of two lowerpoint tree amplitudes. The important point is that for these two terms, the loop momentum is not scaled.
Now let us focus on the special cases , , and specifically take
(23) 
as example. We need to exclude the contribution of massless bubbles from it. In order to do so, let us introduce a scale deformation as is done in the original cut construction. Since , the scale deformation will not change the location of pole . Hence we can write as
(24) 
where .
Let us have a more detailed discussion on the of (24). The onshell condition of is manifestly satisfied for any value of , since (remembering that )
(25) 
Having verified the onshell condition, let us concentrate on the pole structure. We will divide poles into three categories. If the pole does not contain and , then it could either be the sum of some ordinary external legs, or the one containing . For the latter case, we have
(26) 
So this pole is in the scale free form. Similarly, if appears in the numerator, it will give a contribution of in the denominator. Anyway it is also in the scale free form. In other words, these poles does not depend on under the scale deformation.
If the pole contains or , we can always use momentum conservation to rewrite as the leg , so that the pole is in the form containing . For these cases, we can have either leading to a finite pole , or
(27) 
leading to a finite pole
(28) 
Note that both solutions depend on the loop momentum .
If the pole contains both and , then it has no dependence on . This case contains the contribution corresponding to massless bubbles which should be excluded. To see this, let us recall that for the tree diagram that corresponding to massless bubbles with massless external leg , the legs are attached to the same threepoint vertex, then they meet leg in the neighboring vertex. Explicitly for the tree diagrams of , it corresponds to the diagrams where legs and are attached to the same vertex^{5}^{5}5It is easy to see that if we perform the scale deformation , such terms will not contain in the denominator.. This means that the terms corresponding to the massless bubbles are included in the boundary part.
Having understood poles of above three categories, we can now consider the following contour integration
(29)  
where in the second line we have explicitly written down the above mentioned subtle factors in the denominator. Now we consider its various pole contributions,

The pole gives the full undeformed tree amplitude.

There are poles at . Such poles will appear for the propagator when . The other pole can not contribute to pole for generic momentum configuration. From expression (29) we know that the residue at is scale free term and we can ignore them. Note that for this argument to be true, we have assumed the factor in (24) would not provide denominator that breaking the scale free form.

For the pole at , it contains the contribution from massless bubbles, which should be excluded. However, It also contains other contributions which should be included in the final result. But inspecting the expression (29), it can be checked that all such contributions are scale free terms, and we can exclude all the contributions at , letting the result to be valid up to some scale free terms.
With above consideration, we can claim that, the contributions of finite poles are the ones wee need for constructing the oneloop integrands, without the contributions that corresponding to tadpole and massless bubbles, and valid up to some scale free terms. Thus we can write as
(30) 
where , , , and the summation is over all possible splitting of , but with the condition , which means that the set should have more than one external leg.
With above result, we can finally write the as
(31) 
where
and , , , .
Similarly, we have
(32) 
where
and , , , .
We also have
(33) 
To summarize, by BCFW deformation, we have expressed the point oneloop integrand recursively as
(34) 
where , and , are defined as formulas (21), (22) respectively, which are summation of products of lowerpoint tree amplitude with lowpoint oneloop integrand of cut construction. Also, . Among which, are defined in formulas (14), (19) respectively, which are summation of products of two lowerpoint tree amplitudes, and are defined in formulas (31), (32), (33) respectively, which are although products of three lowerpoint tree amplitudes, but one of them is the threepoint amplitude. It is also important to notice how the forward singularities have been removed in various terms by various methods.
3 Some examples
In the previous section, we have presented a recursive formula for oneloop integrand construction, based on the BCFW deformation and cut construction. This new construction shows that there are other ways to write down a welldefined oneloop integrand which is valid up to scale free terms. The recursive formula (34) has given an alternative factorization of oneloop integrand, and it should be equivalent to the result of original cut representation or Feynman diagram method, at least up to some scale free terms. For a better understanding of this recursive formula, in this section, we shall present detailed computation of some oneloop integrands by recursive formula (34), and demonstrate their correspondence with results of original cut construction and Feynman diagram methods.
3.1 The oneloop sixpoint amplitude in scalar theory
In this example we consider the integrand of oneloop sixpoint amplitude in color ordered scalar theory. For this theory, there is no cubic vertex, so the computation is relatively simple since we do not need to use the scale deformation to remove singular terms. After using appropriate BCFW deformation to get rid of boundary contribution, we need to consider contributions from all detectable finite poles of both and . In order to verify the equivalence term by term, we will compute the integrand by Feynman diagram method, the original cut representation and the recursive formula (34).
Using the Feynman rules, we directly get
(35)  
Applying the partial fraction identity